Photo courtesy of the CDC

The American Society of Clinical Oncology (ASCO) has issued recommendations addressing the needs of sexual and gender minority (SGM) populations with cancer.

The recommendations are designed to focus attention on the challenges facing the SGM community—including discrimination and greater risk of anxiety and depression, resulting in disparate care—and concrete steps that can help minimize health disparities among SGM individuals.

The recommendations were published in a policy statement in the Journal of Clinical Oncology.

“Sexual and gender minorities face unique challenges related to cancer risk, discrimination, and other psychosocial issues,” said ASCO President Daniel F. Hayes, MD.

“Compounding these challenges is the fact that providers may have a lack of knowledge and sensitivity about the health risks and health needs facing their SGM patients.”

SGMs include individuals who are lesbian, gay, bisexual, transgender, and intersex (also referred to as those with differences in sex development).

ASCO’s policy statement notes that SGM populations bear a disproportionate cancer burden stemming from several factors, including:

• Lower rates of cancer screening, in part due to lower rates of insurance coverage, exclusion from traditional screening campaigns, and previous experience of discrimination in the healthcare system
• A hesitancy on the part of SGM patients to disclose their sexual orientation to providers due to a fear of stigmatization, which can create additional barriers to care.

ASCO’s statement calls for a coordinated effort to address health disparities among SGM populations, including:

• Increased patient access to culturally competent support services
• Expanded cancer prevention education for SGM individuals
• Robust policies prohibiting discrimination
• Adequate insurance coverage to meet the needs of SGM individuals affected by cancer
• Inclusion of SGM status as a required data element in cancer registries and clinical trials
• Increased focus on SGM populations in cancer research.

“Our objective was to raise awareness among oncology providers, patients, policy makers, and other stakeholders about the cancer care needs of SGM populations and the barriers that SGM individuals face in getting the highest-quality care,” said Jennifer J. Griggs, MD, lead author of the policy statement and a professor at the University of Michigan in Ann Arbor.

“To address these barriers, a coordinated effort is needed to enhance education for patients and providers, to improve outreach and support, and to encourage productive policy and legislative action.”

Photo courtesy of the CDC

The American Society of Clinical Oncology (ASCO) has issued recommendations addressing the needs of sexual and gender minority (SGM) populations with cancer.

The recommendations are designed to focus attention on the challenges facing the SGM community—including discrimination and greater risk of anxiety and depression, resulting in disparate care—and concrete steps that can help minimize health disparities among SGM individuals.

The recommendations were published in a policy statement in the Journal of Clinical Oncology.

“Sexual and gender minorities face unique challenges related to cancer risk, discrimination, and other psychosocial issues,” said ASCO President Daniel F. Hayes, MD.

“Compounding these challenges is the fact that providers may have a lack of knowledge and sensitivity about the health risks and health needs facing their SGM patients.”

SGMs include individuals who are lesbian, gay, bisexual, transgender, and intersex (also referred to as those with differences in sex development).

ASCO’s policy statement notes that SGM populations bear a disproportionate cancer burden stemming from several factors, including:

• Lower rates of cancer screening, in part due to lower rates of insurance coverage, exclusion from traditional screening campaigns, and previous experience of discrimination in the healthcare system
• A hesitancy on the part of SGM patients to disclose their sexual orientation to providers due to a fear of stigmatization, which can create additional barriers to care.

ASCO’s statement calls for a coordinated effort to address health disparities among SGM populations, including:

• Increased patient access to culturally competent support services
• Expanded cancer prevention education for SGM individuals
• Robust policies prohibiting discrimination
• Adequate insurance coverage to meet the needs of SGM individuals affected by cancer
• Inclusion of SGM status as a required data element in cancer registries and clinical trials
• Increased focus on SGM populations in cancer research.

“Our objective was to raise awareness among oncology providers, patients, policy makers, and other stakeholders about the cancer care needs of SGM populations and the barriers that SGM individuals face in getting the highest-quality care,” said Jennifer J. Griggs, MD, lead author of the policy statement and a professor at the University of Michigan in Ann Arbor.

“To address these barriers, a coordinated effort is needed to enhance education for patients and providers, to improve outreach and support, and to encourage productive policy and legislative action.”

Photo courtesy of the CDC

The American Society of Clinical Oncology (ASCO) has issued recommendations addressing the needs of sexual and gender minority (SGM) populations with cancer.

The recommendations are designed to focus attention on the challenges facing the SGM community—including discrimination and greater risk of anxiety and depression, resulting in disparate care—and concrete steps that can help minimize health disparities among SGM individuals.

The recommendations were published in a policy statement in the Journal of Clinical Oncology.

“Sexual and gender minorities face unique challenges related to cancer risk, discrimination, and other psychosocial issues,” said ASCO President Daniel F. Hayes, MD.

“Compounding these challenges is the fact that providers may have a lack of knowledge and sensitivity about the health risks and health needs facing their SGM patients.”

SGMs include individuals who are lesbian, gay, bisexual, transgender, and intersex (also referred to as those with differences in sex development).

ASCO’s policy statement notes that SGM populations bear a disproportionate cancer burden stemming from several factors, including:

• Lower rates of cancer screening, in part due to lower rates of insurance coverage, exclusion from traditional screening campaigns, and previous experience of discrimination in the healthcare system
• A hesitancy on the part of SGM patients to disclose their sexual orientation to providers due to a fear of stigmatization, which can create additional barriers to care.

ASCO’s statement calls for a coordinated effort to address health disparities among SGM populations, including:

• Increased patient access to culturally competent support services
• Expanded cancer prevention education for SGM individuals
• Robust policies prohibiting discrimination
• Adequate insurance coverage to meet the needs of SGM individuals affected by cancer
• Inclusion of SGM status as a required data element in cancer registries and clinical trials
• Increased focus on SGM populations in cancer research.

“Our objective was to raise awareness among oncology providers, patients, policy makers, and other stakeholders about the cancer care needs of SGM populations and the barriers that SGM individuals face in getting the highest-quality care,” said Jennifer J. Griggs, MD, lead author of the policy statement and a professor at the University of Michigan in Ann Arbor.

“To address these barriers, a coordinated effort is needed to enhance education for patients and providers, to improve outreach and support, and to encourage productive policy and legislative action.”

IN THIS ARTICLE

• Diagnosis: questions to ask
• Treatment and management
• Follow-up care

An obese 43-year-old Hispanic man presents to the emergency department (ED) with complaints of severe pain and swelling in his right hand. The patient states that he felt a bite on his hand as he was planting flowers and laying down potting soil near a tree and decorative rocks in his yard. He did not seek immediate medical treatment because the pain was minimal.

As the hours passed, though, the pain increased, and he began to notice tightness in his hand. Twelve hours after the initial bite, the pain became intolerable and his hand swelled to double its normal size, such that he could no longer bend his fingers. He then sought treatment at the ED.

The patient denies previous drug use but indicates that he smokes 1.5 packs of cigarettes daily and drinks alcohol occasionally in social settings. He has no known drug or food allergies. His history is remarkable for hypertension and hyperlipidemia, treated with simvastatin (40 mg/d) and lisinopril (10 mg/d), respectively.

The physical examination reveals an arterial blood pressure of 152/84 mm Hg; heart rate, 76 beats/min; respiratory rate, 18 breaths/min^-1; and temperature, 99ºF. His height is 5 ft 8 in and weight, 297 lb. Cardiovascular examination reveals no irregular heart rhythm, and S₁ and S₂ are heard, with no murmurs or gallops. He denies chest pain and palpitations. Respiratory examination reveals clear breath sounds that are equal and unlabored. He denies shortness of breath or coughing. The patient states that he had nausea earlier that day, but it has subsided.

Dermatologic examination reveals severe erythema and 3+ edema in the patient’s right hand. A 3-cm, irregularly shaped, red, hemorrhagic blister is observed close to the thumb on the posterior side of the right hand. There are two small holes in the center and slight bruising around the lesion. The right hand is hard and warm to the touch upon palpation, and the patient rates his pain as severe (10 out of 10).

The symptoms of severe pain and swelling and the early observation of bruising and hemorrhagic blistering raise suspicion for venomous spider bite (ICD-10 code: T63.331A). Laboratory work-up, including complete blood count, electrolytes, kidney function studies, and urinalysis, is performed. The results are inconclusive, and the reported symptoms and objective assessment are used to make the diagnosis of spider bite.

DISCUSSION

The brown recluse spider (Loxosceles reclusa) is notorious for its bite, which can result in dermonecrosis within 24 to 48 hours. It inhabits the lower Midwest, south central, and southeastern regions of the United States and is not endemic in the West, Northeast, Mid-Atlantic, or Coastal South. Brown recluse spiders are nonaggressive and prefer warm, dark, dry habitats, dwelling under rocks, logs, woodpiles, and debris, as well as in attics, sheds, basements, boxes, travel bags, and motor vehicles.^1,2 They can survive for months without food and can withstand temperatures ranging from 46.4°F to 109.4°F.³ They build irregular, cottony webs that serve as housing but are not used to capture prey.³ (Note that webs found strung along walls, ceilings, outdoor vegetation, and in other exposed areas are nearly always associated with other types of spiders.) The brown recluse is nocturnal, seeking insect prey, either alive or dead.

Brown re­cluse spiders range in size from 6 mm to 20 mm; they have a violin-shaped pattern on the cephalothorax and long legs that allow them to move quickly (see Figure 1). A distinguishing feature is their six eyes, arranged in three pairs (most spiders have eight eyes).

Venom production is influenced by the size and sex of the spider as well as ambient temperature.⁴ The venom contains at least eight enzyme and protein components, including the most active enzyme, sphingomyelinase D.³ This enzyme causes dermonecrosis, platelet aggregation, and complement-mediated hemolysis in vitro, and it may also be responsible for the ulcerating and systemic effects observed in humans.⁵ Sphingomyelinase D has been shown to induce grossly visible tissue necrosis in rabbit tissue within 24 hours after envenomation.³

CLINICAL PRESENTATION

The brown recluse spider bite may be imperceptible at the time of envenomation, requiring no medical attention. Depending on a person’s sensitivity level and the amount of venom injected, however, a mild stinging sensation at the site may be felt, which is usually accompanied by redness and inflammation that may disappear within seconds or last for a couple of hours.⁶

Within two to eight hours, severe pain may occur, progressing to a burning sensation.⁵ The bite site will become pale, due to venom-induced vasoconstriction, with increasing erythema and swelling in the surrounding tissue.⁵ This extreme pain could be due to absorption of the venom by the muscle tissues; if untreated, further tissue damage can occur. Within 12 to 24 hours, there is painful edema with induration and an irregular area of ecchymosis and ische­mia.⁷ Occasionally, the site will develop red, white, and blue hemorrhagic blisters, with the blue ischemic portion centrally located and the red erythematous areas on the periphery.⁸ In almost half of all cases, the lesion is associated with nonspecific systemic symptoms, such as generalized pruritus and rash, headache, nausea, vomiting, and low-grade fever in the first 24 to 48 hours.⁷

Three days after envenomation, the wound will expand and deepen, with skin breakdown noted not sooner than 72 hours after the bite (see Figure 2).^7,8 After five to seven days, the cutaneous lesion forms a dry necrotic eschar with a well-­demarcated border. Within two to three weeks after the bite, the necrotic tissue should detach, and the wound should develop granulated tissue that indicates healing.⁸ Complete healing can take weeks or months, depending on the extent and depth of the wound, with scarring possible in severe cases.⁷

Severe systemic illness (ie, systemic loxoscelism)—rare in the US—is a potential complication of the brown recluse spider bite.⁴ It manifests with fever, malaise, vomiting, headache, and rash; in rare instances, it results in death.⁷

Diagnosis

Brown recluse bite is diagnosed based on history and clinical presentation and, when possible, identification of the spider. However, patients often do not realize they have been bitten before they develop symptoms, making it impossible to confirm the etiology of the lesion. It is often helpful to ask the following questions during the assessment

• Did you feel the bite take place?
• Did you see or capture the spider? If so, can you describe it?
• Where were you when the spider bit you?
• Did you recently clean any clutter or debris?

Furthermore, patients who recall seeing a spider after being bitten typically do not bring the arachnid to their health care facility. Another complicating factor is the numerous possible causes of necrotic skin lesions that can be mistaken for spider bites.⁵ The differential diagnosis can include allergic dermatitis, cellulitis, methicillin-resistant Staphylococcus aureus (MRSA) infection, skin abscesses, other arthropod bites, necrotizing fasciitis, or bee sting.

TREATMENT AND MANAGEMENT

One of the most important factors in successful treatment is timeliness of medical attention after the initial bite; because the most damaging tissue effects occur within the first three to six hours after envenomation, intervention during this time is imperative.⁸ Initial treatment of cutaneous brown recluse spider bite is often conservative, given the variation in clinical presentation, inability to predict the future extent of lesions, and lack of evidence-based treatment options.⁹ The goals of therapy are to ensure that skin integrity is maintained, infection is avoided, and circulation is preserved.¹⁰

Nonpharmacologic treatments for brown recluse spider bite consist of cleaning the wound, treating the bite area with “RICE” (rest-ice-compression-elevation) therapy during the first 72 hours to reduce tissue damage, and ensuring adequate hydration.^1,10-13 The affected area should be cleaned thoroughly; infected wounds require topical antiseptics and sterile dressings. Applying a cold compress to the bite area at 20-minute intervals during the first 72 hours after envenomation has been shown to reduce tissue damage.¹⁰ Heat should not be applied to the area, as it may increase tissue damage.

Pharmacologic treatment. Patients who experience systemic symptoms such as nausea, vomiting, pain, fever, and pruritus should be provided antipyretics, hydration, and analgesics for symptomatic relief, as needed.⁹ Antihistamines and benzodiazepines have been found to be useful in relieving symptoms of anxiety and pruritus. To help manage mild pain, OTC NSAIDs are recommended.¹⁰

If the date of the last tetanus shot is unknown, a prophylactic tetanus booster (tetanus/diphtheria [Td] or TDaP) should be administered.¹⁰ The prophylactic use of cephalosporins to treat infection is indicated in patients with tissue breakdown.¹

Among the more controversial treatment choices are use of corticosteroids and dapsone, prescribed frequently in the past. Use of oral corticosteroids for cutaneous forms of spider bite is not supported by current evidence.^5,10,14 Research does, however, support their role in the treatment of bite-induced systemic illness, particularly for preventing kidney failure and hemolysis in children.^1,15

Dapsone, prescribed for the necrotic lesions, may be useful in limiting the inflammatory response at the site of envenomation.^1,3 However, human studies have shown conflicting results with dapsone administration, with some demonstrating no improvement in patient outcomes.⁸ The risks of dapsone’s many adverse effects, including dose-related hemolysis, sore throat, pallor, agranulocytosis, aplastic anemia, and cholestatic jaundice, may outweigh its benefits.^1,12 Furthermore, dapsone treatment is restricted in patients with G6PD (glucose-6-phosphate dehydrogenase) deficiency because of their increased risk for hemolytic anemia.¹ Accordingly, dapsone is recommended only for moderate-to-severe or rapidly progressing cases in adults.¹

FOLLOW-UP CARE

A patient's follow-up care should be assessed individually, based on the nature of his/her reaction to the bite. In all instances, however, ask the patient to report worsening of symptoms and changes in the skin around the bite area; if systemic symptoms develop, patients should proceed to the ED. If, after six to eight weeks, the necrotic lesion is large and has stabilized in size, consider referring to a wound care clinic for surgical excision of the eschar.⁹

To avoid future spider bites, advise patients to clear all clutter, move beds away from the wall, remove bed skirts or ruffles, avoid using underbed storage containers, avoid leaving clothing on the floor in piles, and check shoes before dressing.⁵

OUTCOME FOR THE CASE PATIENT

Initial supportive treatment for this patient included cleaning the bite area with antiseptic soap and water. A cold compress was applied to the bite area at 20-minute intervals, and the right hand was elevated. Hydrocodone bitartrate/acetaminophen (5/325 mg qid) was administered to alleviate pain. The patient was also given a tetanus booster because the date of his last immunization was unknown.

After two hours of monitoring, the patient was no longer able to move his hand, swelling around the affected area increased, and the bite site began to appear necrotic. Cephalexin (500 mg bid) was ordered along with dapsone (100 mg/d). The patient was referred for consultation with wound care and infectious disease specialists because of possible tissue necrosis.

CONCLUSION

Brown recluse spider bites are uncommon, and most are unremarkable and self-healing. Patients who present following a brown recluse bite typically can be managed successfully with supportive care (RICE) and careful observation. In rare cases, however, bites may result in significant tissue necrosis or even death.

The diagnosis is typically based on thorough physical examination, with attention to the lesion characteristics and appropriate questions about the spider and the development of the lesion over time. Diagnosis through identification of the spider seldom occurs, since patients typically do not capture the spider and bring it with them for identification. The geographic region where the bite occurs is an important factor as well, since brown recluse envenomation is higher on the differential diagnosis of necrotic skin lesions in areas where these spiders are endemic (the lower Midwest, south central, and southeastern regions of the US).

IN THIS ARTICLE

• Diagnosis: questions to ask
• Treatment and management
• Follow-up care

An obese 43-year-old Hispanic man presents to the emergency department (ED) with complaints of severe pain and swelling in his right hand. The patient states that he felt a bite on his hand as he was planting flowers and laying down potting soil near a tree and decorative rocks in his yard. He did not seek immediate medical treatment because the pain was minimal.

As the hours passed, though, the pain increased, and he began to notice tightness in his hand. Twelve hours after the initial bite, the pain became intolerable and his hand swelled to double its normal size, such that he could no longer bend his fingers. He then sought treatment at the ED.

The patient denies previous drug use but indicates that he smokes 1.5 packs of cigarettes daily and drinks alcohol occasionally in social settings. He has no known drug or food allergies. His history is remarkable for hypertension and hyperlipidemia, treated with simvastatin (40 mg/d) and lisinopril (10 mg/d), respectively.

The physical examination reveals an arterial blood pressure of 152/84 mm Hg; heart rate, 76 beats/min; respiratory rate, 18 breaths/min^-1; and temperature, 99ºF. His height is 5 ft 8 in and weight, 297 lb. Cardiovascular examination reveals no irregular heart rhythm, and S₁ and S₂ are heard, with no murmurs or gallops. He denies chest pain and palpitations. Respiratory examination reveals clear breath sounds that are equal and unlabored. He denies shortness of breath or coughing. The patient states that he had nausea earlier that day, but it has subsided.

Dermatologic examination reveals severe erythema and 3+ edema in the patient’s right hand. A 3-cm, irregularly shaped, red, hemorrhagic blister is observed close to the thumb on the posterior side of the right hand. There are two small holes in the center and slight bruising around the lesion. The right hand is hard and warm to the touch upon palpation, and the patient rates his pain as severe (10 out of 10).

The symptoms of severe pain and swelling and the early observation of bruising and hemorrhagic blistering raise suspicion for venomous spider bite (ICD-10 code: T63.331A). Laboratory work-up, including complete blood count, electrolytes, kidney function studies, and urinalysis, is performed. The results are inconclusive, and the reported symptoms and objective assessment are used to make the diagnosis of spider bite.

DISCUSSION

The brown recluse spider (Loxosceles reclusa) is notorious for its bite, which can result in dermonecrosis within 24 to 48 hours. It inhabits the lower Midwest, south central, and southeastern regions of the United States and is not endemic in the West, Northeast, Mid-Atlantic, or Coastal South. Brown recluse spiders are nonaggressive and prefer warm, dark, dry habitats, dwelling under rocks, logs, woodpiles, and debris, as well as in attics, sheds, basements, boxes, travel bags, and motor vehicles.^1,2 They can survive for months without food and can withstand temperatures ranging from 46.4°F to 109.4°F.³ They build irregular, cottony webs that serve as housing but are not used to capture prey.³ (Note that webs found strung along walls, ceilings, outdoor vegetation, and in other exposed areas are nearly always associated with other types of spiders.) The brown recluse is nocturnal, seeking insect prey, either alive or dead.

Brown re­cluse spiders range in size from 6 mm to 20 mm; they have a violin-shaped pattern on the cephalothorax and long legs that allow them to move quickly (see Figure 1). A distinguishing feature is their six eyes, arranged in three pairs (most spiders have eight eyes).

Venom production is influenced by the size and sex of the spider as well as ambient temperature.⁴ The venom contains at least eight enzyme and protein components, including the most active enzyme, sphingomyelinase D.³ This enzyme causes dermonecrosis, platelet aggregation, and complement-mediated hemolysis in vitro, and it may also be responsible for the ulcerating and systemic effects observed in humans.⁵ Sphingomyelinase D has been shown to induce grossly visible tissue necrosis in rabbit tissue within 24 hours after envenomation.³

CLINICAL PRESENTATION

The brown recluse spider bite may be imperceptible at the time of envenomation, requiring no medical attention. Depending on a person’s sensitivity level and the amount of venom injected, however, a mild stinging sensation at the site may be felt, which is usually accompanied by redness and inflammation that may disappear within seconds or last for a couple of hours.⁶

Within two to eight hours, severe pain may occur, progressing to a burning sensation.⁵ The bite site will become pale, due to venom-induced vasoconstriction, with increasing erythema and swelling in the surrounding tissue.⁵ This extreme pain could be due to absorption of the venom by the muscle tissues; if untreated, further tissue damage can occur. Within 12 to 24 hours, there is painful edema with induration and an irregular area of ecchymosis and ische­mia.⁷ Occasionally, the site will develop red, white, and blue hemorrhagic blisters, with the blue ischemic portion centrally located and the red erythematous areas on the periphery.⁸ In almost half of all cases, the lesion is associated with nonspecific systemic symptoms, such as generalized pruritus and rash, headache, nausea, vomiting, and low-grade fever in the first 24 to 48 hours.⁷

Three days after envenomation, the wound will expand and deepen, with skin breakdown noted not sooner than 72 hours after the bite (see Figure 2).^7,8 After five to seven days, the cutaneous lesion forms a dry necrotic eschar with a well-­demarcated border. Within two to three weeks after the bite, the necrotic tissue should detach, and the wound should develop granulated tissue that indicates healing.⁸ Complete healing can take weeks or months, depending on the extent and depth of the wound, with scarring possible in severe cases.⁷

Severe systemic illness (ie, systemic loxoscelism)—rare in the US—is a potential complication of the brown recluse spider bite.⁴ It manifests with fever, malaise, vomiting, headache, and rash; in rare instances, it results in death.⁷

Diagnosis

Brown recluse bite is diagnosed based on history and clinical presentation and, when possible, identification of the spider. However, patients often do not realize they have been bitten before they develop symptoms, making it impossible to confirm the etiology of the lesion. It is often helpful to ask the following questions during the assessment

• Did you feel the bite take place?
• Did you see or capture the spider? If so, can you describe it?
• Where were you when the spider bit you?
• Did you recently clean any clutter or debris?

Furthermore, patients who recall seeing a spider after being bitten typically do not bring the arachnid to their health care facility. Another complicating factor is the numerous possible causes of necrotic skin lesions that can be mistaken for spider bites.⁵ The differential diagnosis can include allergic dermatitis, cellulitis, methicillin-resistant Staphylococcus aureus (MRSA) infection, skin abscesses, other arthropod bites, necrotizing fasciitis, or bee sting.

TREATMENT AND MANAGEMENT

One of the most important factors in successful treatment is timeliness of medical attention after the initial bite; because the most damaging tissue effects occur within the first three to six hours after envenomation, intervention during this time is imperative.⁸ Initial treatment of cutaneous brown recluse spider bite is often conservative, given the variation in clinical presentation, inability to predict the future extent of lesions, and lack of evidence-based treatment options.⁹ The goals of therapy are to ensure that skin integrity is maintained, infection is avoided, and circulation is preserved.¹⁰

Nonpharmacologic treatments for brown recluse spider bite consist of cleaning the wound, treating the bite area with “RICE” (rest-ice-compression-elevation) therapy during the first 72 hours to reduce tissue damage, and ensuring adequate hydration.^1,10-13 The affected area should be cleaned thoroughly; infected wounds require topical antiseptics and sterile dressings. Applying a cold compress to the bite area at 20-minute intervals during the first 72 hours after envenomation has been shown to reduce tissue damage.¹⁰ Heat should not be applied to the area, as it may increase tissue damage.

Pharmacologic treatment. Patients who experience systemic symptoms such as nausea, vomiting, pain, fever, and pruritus should be provided antipyretics, hydration, and analgesics for symptomatic relief, as needed.⁹ Antihistamines and benzodiazepines have been found to be useful in relieving symptoms of anxiety and pruritus. To help manage mild pain, OTC NSAIDs are recommended.¹⁰

If the date of the last tetanus shot is unknown, a prophylactic tetanus booster (tetanus/diphtheria [Td] or TDaP) should be administered.¹⁰ The prophylactic use of cephalosporins to treat infection is indicated in patients with tissue breakdown.¹

Among the more controversial treatment choices are use of corticosteroids and dapsone, prescribed frequently in the past. Use of oral corticosteroids for cutaneous forms of spider bite is not supported by current evidence.^5,10,14 Research does, however, support their role in the treatment of bite-induced systemic illness, particularly for preventing kidney failure and hemolysis in children.^1,15

Dapsone, prescribed for the necrotic lesions, may be useful in limiting the inflammatory response at the site of envenomation.^1,3 However, human studies have shown conflicting results with dapsone administration, with some demonstrating no improvement in patient outcomes.⁸ The risks of dapsone’s many adverse effects, including dose-related hemolysis, sore throat, pallor, agranulocytosis, aplastic anemia, and cholestatic jaundice, may outweigh its benefits.^1,12 Furthermore, dapsone treatment is restricted in patients with G6PD (glucose-6-phosphate dehydrogenase) deficiency because of their increased risk for hemolytic anemia.¹ Accordingly, dapsone is recommended only for moderate-to-severe or rapidly progressing cases in adults.¹

FOLLOW-UP CARE

A patient's follow-up care should be assessed individually, based on the nature of his/her reaction to the bite. In all instances, however, ask the patient to report worsening of symptoms and changes in the skin around the bite area; if systemic symptoms develop, patients should proceed to the ED. If, after six to eight weeks, the necrotic lesion is large and has stabilized in size, consider referring to a wound care clinic for surgical excision of the eschar.⁹

To avoid future spider bites, advise patients to clear all clutter, move beds away from the wall, remove bed skirts or ruffles, avoid using underbed storage containers, avoid leaving clothing on the floor in piles, and check shoes before dressing.⁵

OUTCOME FOR THE CASE PATIENT

Initial supportive treatment for this patient included cleaning the bite area with antiseptic soap and water. A cold compress was applied to the bite area at 20-minute intervals, and the right hand was elevated. Hydrocodone bitartrate/acetaminophen (5/325 mg qid) was administered to alleviate pain. The patient was also given a tetanus booster because the date of his last immunization was unknown.

After two hours of monitoring, the patient was no longer able to move his hand, swelling around the affected area increased, and the bite site began to appear necrotic. Cephalexin (500 mg bid) was ordered along with dapsone (100 mg/d). The patient was referred for consultation with wound care and infectious disease specialists because of possible tissue necrosis.

CONCLUSION

Brown recluse spider bites are uncommon, and most are unremarkable and self-healing. Patients who present following a brown recluse bite typically can be managed successfully with supportive care (RICE) and careful observation. In rare cases, however, bites may result in significant tissue necrosis or even death.

The diagnosis is typically based on thorough physical examination, with attention to the lesion characteristics and appropriate questions about the spider and the development of the lesion over time. Diagnosis through identification of the spider seldom occurs, since patients typically do not capture the spider and bring it with them for identification. The geographic region where the bite occurs is an important factor as well, since brown recluse envenomation is higher on the differential diagnosis of necrotic skin lesions in areas where these spiders are endemic (the lower Midwest, south central, and southeastern regions of the US).

IN THIS ARTICLE

• Diagnosis: questions to ask
• Treatment and management
• Follow-up care

An obese 43-year-old Hispanic man presents to the emergency department (ED) with complaints of severe pain and swelling in his right hand. The patient states that he felt a bite on his hand as he was planting flowers and laying down potting soil near a tree and decorative rocks in his yard. He did not seek immediate medical treatment because the pain was minimal.

As the hours passed, though, the pain increased, and he began to notice tightness in his hand. Twelve hours after the initial bite, the pain became intolerable and his hand swelled to double its normal size, such that he could no longer bend his fingers. He then sought treatment at the ED.

The patient denies previous drug use but indicates that he smokes 1.5 packs of cigarettes daily and drinks alcohol occasionally in social settings. He has no known drug or food allergies. His history is remarkable for hypertension and hyperlipidemia, treated with simvastatin (40 mg/d) and lisinopril (10 mg/d), respectively.

The physical examination reveals an arterial blood pressure of 152/84 mm Hg; heart rate, 76 beats/min; respiratory rate, 18 breaths/min^-1; and temperature, 99ºF. His height is 5 ft 8 in and weight, 297 lb. Cardiovascular examination reveals no irregular heart rhythm, and S₁ and S₂ are heard, with no murmurs or gallops. He denies chest pain and palpitations. Respiratory examination reveals clear breath sounds that are equal and unlabored. He denies shortness of breath or coughing. The patient states that he had nausea earlier that day, but it has subsided.

Dermatologic examination reveals severe erythema and 3+ edema in the patient’s right hand. A 3-cm, irregularly shaped, red, hemorrhagic blister is observed close to the thumb on the posterior side of the right hand. There are two small holes in the center and slight bruising around the lesion. The right hand is hard and warm to the touch upon palpation, and the patient rates his pain as severe (10 out of 10).

The symptoms of severe pain and swelling and the early observation of bruising and hemorrhagic blistering raise suspicion for venomous spider bite (ICD-10 code: T63.331A). Laboratory work-up, including complete blood count, electrolytes, kidney function studies, and urinalysis, is performed. The results are inconclusive, and the reported symptoms and objective assessment are used to make the diagnosis of spider bite.

DISCUSSION

The brown recluse spider (Loxosceles reclusa) is notorious for its bite, which can result in dermonecrosis within 24 to 48 hours. It inhabits the lower Midwest, south central, and southeastern regions of the United States and is not endemic in the West, Northeast, Mid-Atlantic, or Coastal South. Brown recluse spiders are nonaggressive and prefer warm, dark, dry habitats, dwelling under rocks, logs, woodpiles, and debris, as well as in attics, sheds, basements, boxes, travel bags, and motor vehicles.^1,2 They can survive for months without food and can withstand temperatures ranging from 46.4°F to 109.4°F.³ They build irregular, cottony webs that serve as housing but are not used to capture prey.³ (Note that webs found strung along walls, ceilings, outdoor vegetation, and in other exposed areas are nearly always associated with other types of spiders.) The brown recluse is nocturnal, seeking insect prey, either alive or dead.

Brown re­cluse spiders range in size from 6 mm to 20 mm; they have a violin-shaped pattern on the cephalothorax and long legs that allow them to move quickly (see Figure 1). A distinguishing feature is their six eyes, arranged in three pairs (most spiders have eight eyes).

Venom production is influenced by the size and sex of the spider as well as ambient temperature.⁴ The venom contains at least eight enzyme and protein components, including the most active enzyme, sphingomyelinase D.³ This enzyme causes dermonecrosis, platelet aggregation, and complement-mediated hemolysis in vitro, and it may also be responsible for the ulcerating and systemic effects observed in humans.⁵ Sphingomyelinase D has been shown to induce grossly visible tissue necrosis in rabbit tissue within 24 hours after envenomation.³

CLINICAL PRESENTATION

The brown recluse spider bite may be imperceptible at the time of envenomation, requiring no medical attention. Depending on a person’s sensitivity level and the amount of venom injected, however, a mild stinging sensation at the site may be felt, which is usually accompanied by redness and inflammation that may disappear within seconds or last for a couple of hours.⁶

Within two to eight hours, severe pain may occur, progressing to a burning sensation.⁵ The bite site will become pale, due to venom-induced vasoconstriction, with increasing erythema and swelling in the surrounding tissue.⁵ This extreme pain could be due to absorption of the venom by the muscle tissues; if untreated, further tissue damage can occur. Within 12 to 24 hours, there is painful edema with induration and an irregular area of ecchymosis and ische­mia.⁷ Occasionally, the site will develop red, white, and blue hemorrhagic blisters, with the blue ischemic portion centrally located and the red erythematous areas on the periphery.⁸ In almost half of all cases, the lesion is associated with nonspecific systemic symptoms, such as generalized pruritus and rash, headache, nausea, vomiting, and low-grade fever in the first 24 to 48 hours.⁷

Three days after envenomation, the wound will expand and deepen, with skin breakdown noted not sooner than 72 hours after the bite (see Figure 2).^7,8 After five to seven days, the cutaneous lesion forms a dry necrotic eschar with a well-­demarcated border. Within two to three weeks after the bite, the necrotic tissue should detach, and the wound should develop granulated tissue that indicates healing.⁸ Complete healing can take weeks or months, depending on the extent and depth of the wound, with scarring possible in severe cases.⁷

Severe systemic illness (ie, systemic loxoscelism)—rare in the US—is a potential complication of the brown recluse spider bite.⁴ It manifests with fever, malaise, vomiting, headache, and rash; in rare instances, it results in death.⁷

Diagnosis

Brown recluse bite is diagnosed based on history and clinical presentation and, when possible, identification of the spider. However, patients often do not realize they have been bitten before they develop symptoms, making it impossible to confirm the etiology of the lesion. It is often helpful to ask the following questions during the assessment

• Did you feel the bite take place?
• Did you see or capture the spider? If so, can you describe it?
• Where were you when the spider bit you?
• Did you recently clean any clutter or debris?

Furthermore, patients who recall seeing a spider after being bitten typically do not bring the arachnid to their health care facility. Another complicating factor is the numerous possible causes of necrotic skin lesions that can be mistaken for spider bites.⁵ The differential diagnosis can include allergic dermatitis, cellulitis, methicillin-resistant Staphylococcus aureus (MRSA) infection, skin abscesses, other arthropod bites, necrotizing fasciitis, or bee sting.

TREATMENT AND MANAGEMENT

One of the most important factors in successful treatment is timeliness of medical attention after the initial bite; because the most damaging tissue effects occur within the first three to six hours after envenomation, intervention during this time is imperative.⁸ Initial treatment of cutaneous brown recluse spider bite is often conservative, given the variation in clinical presentation, inability to predict the future extent of lesions, and lack of evidence-based treatment options.⁹ The goals of therapy are to ensure that skin integrity is maintained, infection is avoided, and circulation is preserved.¹⁰

Nonpharmacologic treatments for brown recluse spider bite consist of cleaning the wound, treating the bite area with “RICE” (rest-ice-compression-elevation) therapy during the first 72 hours to reduce tissue damage, and ensuring adequate hydration.^1,10-13 The affected area should be cleaned thoroughly; infected wounds require topical antiseptics and sterile dressings. Applying a cold compress to the bite area at 20-minute intervals during the first 72 hours after envenomation has been shown to reduce tissue damage.¹⁰ Heat should not be applied to the area, as it may increase tissue damage.

Pharmacologic treatment. Patients who experience systemic symptoms such as nausea, vomiting, pain, fever, and pruritus should be provided antipyretics, hydration, and analgesics for symptomatic relief, as needed.⁹ Antihistamines and benzodiazepines have been found to be useful in relieving symptoms of anxiety and pruritus. To help manage mild pain, OTC NSAIDs are recommended.¹⁰

If the date of the last tetanus shot is unknown, a prophylactic tetanus booster (tetanus/diphtheria [Td] or TDaP) should be administered.¹⁰ The prophylactic use of cephalosporins to treat infection is indicated in patients with tissue breakdown.¹

Among the more controversial treatment choices are use of corticosteroids and dapsone, prescribed frequently in the past. Use of oral corticosteroids for cutaneous forms of spider bite is not supported by current evidence.^5,10,14 Research does, however, support their role in the treatment of bite-induced systemic illness, particularly for preventing kidney failure and hemolysis in children.^1,15

Dapsone, prescribed for the necrotic lesions, may be useful in limiting the inflammatory response at the site of envenomation.^1,3 However, human studies have shown conflicting results with dapsone administration, with some demonstrating no improvement in patient outcomes.⁸ The risks of dapsone’s many adverse effects, including dose-related hemolysis, sore throat, pallor, agranulocytosis, aplastic anemia, and cholestatic jaundice, may outweigh its benefits.^1,12 Furthermore, dapsone treatment is restricted in patients with G6PD (glucose-6-phosphate dehydrogenase) deficiency because of their increased risk for hemolytic anemia.¹ Accordingly, dapsone is recommended only for moderate-to-severe or rapidly progressing cases in adults.¹

FOLLOW-UP CARE

A patient's follow-up care should be assessed individually, based on the nature of his/her reaction to the bite. In all instances, however, ask the patient to report worsening of symptoms and changes in the skin around the bite area; if systemic symptoms develop, patients should proceed to the ED. If, after six to eight weeks, the necrotic lesion is large and has stabilized in size, consider referring to a wound care clinic for surgical excision of the eschar.⁹

To avoid future spider bites, advise patients to clear all clutter, move beds away from the wall, remove bed skirts or ruffles, avoid using underbed storage containers, avoid leaving clothing on the floor in piles, and check shoes before dressing.⁵

OUTCOME FOR THE CASE PATIENT

Initial supportive treatment for this patient included cleaning the bite area with antiseptic soap and water. A cold compress was applied to the bite area at 20-minute intervals, and the right hand was elevated. Hydrocodone bitartrate/acetaminophen (5/325 mg qid) was administered to alleviate pain. The patient was also given a tetanus booster because the date of his last immunization was unknown.

After two hours of monitoring, the patient was no longer able to move his hand, swelling around the affected area increased, and the bite site began to appear necrotic. Cephalexin (500 mg bid) was ordered along with dapsone (100 mg/d). The patient was referred for consultation with wound care and infectious disease specialists because of possible tissue necrosis.

CONCLUSION

Brown recluse spider bites are uncommon, and most are unremarkable and self-healing. Patients who present following a brown recluse bite typically can be managed successfully with supportive care (RICE) and careful observation. In rare cases, however, bites may result in significant tissue necrosis or even death.

The diagnosis is typically based on thorough physical examination, with attention to the lesion characteristics and appropriate questions about the spider and the development of the lesion over time. Diagnosis through identification of the spider seldom occurs, since patients typically do not capture the spider and bring it with them for identification. The geographic region where the bite occurs is an important factor as well, since brown recluse envenomation is higher on the differential diagnosis of necrotic skin lesions in areas where these spiders are endemic (the lower Midwest, south central, and southeastern regions of the US).

From the Maimonides Medical Center (Dr. Karam) and the SUNY Downstate Medical Center (Dr. Karam and Dr. McFarlane), Brooklyn, NY.

Abstract

• Objective. To discuss the epidemic of diabetes highlighting the natural history of the disease and the major clinical trials aimed at diabetes prevention in different prediabetic populations around the world.
• Results. Diabetes prevention studies have evaluated various interventions including lifestyle modifications, metformin, alpha-glucosidase inhibitors, thiazolidinediones, nateglinide, and xenical as well as the renin-angiotensin aldosterone system (RAS) inhibitors. Lifestyle modifications seem to be the safest, most effective, and most sustainable intervention to prevent diabetes. Except for metformin, the potential diabetes prevention benefits of the studied pharmacologic agents are limited by safety concerns or lack of durable efficacy or tolerability. RAS blockade and fibrates have a favorable glycemic effect, and, when indicated, are reasonable treatment options for hypertension and hyperlipidemia in prediabetic patients.
• Conclusion. As recommended by American Diabetes Association guidelines, patients with prediabetes should be referred to an intensive diet and physical activity behavioral counseling program; diet and activity goals include a loss of 7% of body weight and at least 150 minutes of moderate physical activity per week. Metformin therapy for diabetes prevention should be considered as well.

Key words: prediabetes; type 2 diabetes mellitus, diabetes prevention, lifestyle modifications.

Diabetes mellitus has reached pandemic proportions across the globe. The International Diabetes Federation (IDF) estimates that in 2015 around 415 million people, or 1 in 11 adults, had diabetes, compared to 285 million in 2010, with 5 million deaths, or 1 death every 6 seconds, occurring because of diabetes or diabetes complications [1]. In the United States, an estimated 29.1 million Americans, or 9.3% of the population, have diabetes, 27.8% of them undiagnosed [2]. The prevalence of diabetes increases significantly with age, affecting around 16.2% of American adults aged 45 to 64 years and 25.9% of adults aged 65 years or older [2]. The Centers for Disease Control and Prevention (CDC) estimates that, with current trends, as many as 1 in 3 American adults could have diabetes by 2050 [3].

Type 2 diabetes mellitus (T2DM) accounts for the majority of prevalent and newly diagnosed diabetes in the world, and is strongly linked to overweight and inactivity in adults [4]. T2DM is increasingly being diagnosed in pediatric patients, in whom type 1 diabetes has historically been predominant; it now accounts for approximately 30% of newly diagnosed diabetes in children aged 10 to 19 years, exceeding 50% in certain ethnicities such as non-Hispanic black and American Indian/Alaska Native children [2].

These alarming trends have spurred significant research and public efforts aimed at reducing the prevalence of diabetes by preventing T2DM. Indeed, insulin resistance and abnormal carbohydrate metabolism progress over many years prior to the diagnosis of diabetes and manifest with different clinical and biochemical features. Both the pathophysiology and the natural history of T2DM offer clinicians an opportunity to identify patients at risk for developing the disease and to implement prevention strategies. This article outlines the risk factors and diagnostic criteria for prediabetes, describes the studies that have explored diabetes prevention through lifestyle changes, pharmacotherapy, or surgery, and reviews recommendations for managing patients at risk.

Risk Factors and Screening for T2DM

The American Diabetes Association (ADA) recommends screening all adults for prediabetes by assessing for diabetes risk factors [8]. Glucose testing is recommended in individuals aged 45 years or older, and should be considered in adults of any age who are overweight or obese (body mass index [BMI] ≥ 25 kg/m² or ≥ 23 kg/m² in Asian Americans) and have 1 or more additional risk factors for diabetes. Testing also should be considered in children and adolescents who are overweight or obese and who have 2 or more additional risk factors. If tests are normal, repeat testing carried out at a minimum of 3-year intervals is suggested [8].

Prediabetes

Abnormalities in glucose metabolism progress along a continuum through various stages before T2DM develops. Years before the development of overt diabetes, and especially in the presence of excessive visceral fat, cellular sensitivity to insulin gradually decreases, leading to a compensatory increased insulin secretion [9]. With time, and under continuous increased demand, pancreatic beta cell function declines and ultimately fails to overcome insulin resistance and maintain a normal glucose metabolism, resulting in prediabetes followed by the development of diabetes. This early beta cell dysfunction was illustrated by the decreased beta cell volume observed on autopsy of obese patients with IFG or T2DM, when compared to obese individuals with normal glucose tolerance [10]. It is estimated that around 40% to 70% of beta cell function is already lost by the time diabetes is clinically diagnosed. This relatively slow pathophysiologic process allows the identification of at-risk patients well before their blood glucose levels reach the diabetic diagnostic thresholds, and therefore presents an opportunity for prevention.

Diagnostic Criteria

The ADA guidelines released in 2003 define prediabetes as IFG (fasting blood glucose [FBG] levels of 100–125 mg/dL), IGT (glucose levels of 140–199 mg/dL at 2 hours during an oral glucose tolerance test [OGTT] following an oral load of 75 g of dextrose), or both. Additionally, hemoglobin A1C (A1C) was introduced as a diagnostic tool for prediabetes in 2010, with values between 5.7% and 6.4% indicating prediabetes [8]. Most of these thresholds were chosen due to their association with increased rates of complications, notably retinopathy and cardiovascular disease.

A combined report from the World Health Organization (WHO) and the IDF published in 2006 defined intermediate hyperglycemia as IFG, but with a higher cutoff for FBG (110–125 mg/dL) than the ADA’s definition, and/or IGT (2-hour OGTT glucose level of 140–199 mg/dL) [11]. The rationale for a higher cut-point for IFG is the concern about the increased prevalence of IFG and its impact on individuals and health systems and the more favorable cardiovascular risk profile and decreased risk of progression to diabetes in the group of patients with FBG of 100 to 110 mg/dL when compared to the group with FBG of 110 to 125 mg/dL. The report does not recommend the use of A1C in the diagnosis of diabetes or intermediate hyperglycemia because of a lack of global consistency and the potential for other factors that can be prevalent in some developing countries, such as hemoglobinopathies and anemia, to interfere with the assay.

Prevalence and Progression to Diabetes

According to CDC data from 2014, up to 86 million American adults, more than 1 in 3, have prediabetes, and 9 out of 10 of these individuals are undiagnosed [2]. It is estimated that approximately 25% of people diagnosed with either IFG or IGT progress to diabetes mellitus over a 3- to 5-year period [12]. If observed for longer periods, most prediabetic persons will probably develop diabetes. The highest rate of progression to diabetes is observed in patients with both IFG and IGT, older age, overweight, or other diabetic risk factors.

Complications

In addition to increasing the risk for progression to diabetes, prediabetes is independently associated with microvascular and macrovascular complications and increased risk of death, prior to the actual onset of diabetes. The DECODE study demonstrated significantly increased mortality in 2766 individuals with IGT after 7 years of follow-up, when compared to normoglycemic patients; this effect was more prominent in participants with IGT than in participants with IFG [13]. In the Australian Diabetes, Obesity and Lifestyle Study, IFG was found to be an independent predictor for cardiovascular mortality after adjustment for age, sex, and other traditional cardiovascular risk factors [14].

Similarly, a recent meta-analysis demonstrated that the presence of IFG was significantly associated with future risk for coronary heart disease (CHD), with the risk increase starting when fasting plasma glucose was as low as 100 mg/dL; however, this finding may have been confounded by the presence of undetected IGT or other cardiovascular risk factors [15]. Another recent systematic review of 53 prospective cohort studies with 1,611,339 participants showed that prediabetes (IFG or IGT) was associated with an increased risk of composite cardiovascular disease, CHD, stroke, and all-cause mortality [16].

The association between retinopathy and prediabetes has been described in multiple reports and this association has helped guide authors on selected thresholds for diagnosis of prediabetes. For example, in 1 study, the incidence of retinopathy in individuals with IGT was 12% among Pima Indians [17]. Similarly, in a follow-up study of the Diabetes Prevention Program, 8% of prediabetic participants who remained nondiabetics had evidence of retinopathy [18].

Neuropathy also has been observed in prediabetes. A noninvasive neurologic evaluation of individuals with IGT revealed subclinical neural dysfunction suggestive of cardiovascular autonomic neuropathy [19]. At the clinical level, a study that evaluated 100 patients with chronic idiopathic axonal neuropathy of unknown etiology found IFG in 36 and IGT in 38 patients, underscoring the role of abnormal glucose metabolism in these patients [20].

Nephropathy may also be more prevalent in those with prediabetes. In a 1999–2006 National Health and Nutrition Examination Survey analysis, the adjusted prevalence of chronic kidney disease, defined by estimated glomerular filtration rate (eGFR) of 15 to 59 mL/min per 1.73 m² or albumin-creatinine ratio ≥ 30 mg/g, was 17.1% in individuals with IFG, compared to 11.8% in individuals with normal fasting glucose [21].

Lifestyle Modifications

The alarming rapid increase in the prevalence of T2DM has been linked to a parallel rising epidemic of overweight, obesity, and lack of physical activity. Therefore, lifestyle changes aiming at weight reduction seemed to be a natural individual and public health strategy to prevent diabetes, and such strategies have been the focus of many randomized controlled trials around the world. As anticipated, weight loss, exercise, and diet have all been shown, separately or in combination, to be effective in decreasing the incidence of T2DM in high-risk patients [22–27]. Furthermore, and well beyond the benefit observed during the trials, follow-up studies revealed a sustained reduction of diabetes incidence in intervention groups several years after cessation of the intervention [28–32] (Table 2).

The Da Quing Diabetes Prevention Study (DQDPS), published in 1997, is one of the earliest prospective diabetes prevention trials [22]. This 6-year study conducted in 33 clinics in China from 1986 through 1992 included 577 participants with IGT who were randomly assigned to 1 of 4 groups: (1) diet (high vegetables, low sugar/alcohol) only, (2) exercise, (3) diet plus exercise, and (4) standard of care. At 6 years, diabetes incidence was significantly reduced by 46% in the exercise group, 31% in the diet group, and 42% in the diet plus exercise group compared to standard care. In 2006, 14 years after the end of the trial and 20 years after the initial enrollment, the cumulative incidence of diabetes was significantly lower in the intervention group at 80%, compared to 93% in the control group, and the annual incidence of diabetes was 7% and 11%, respectively, with a 43% lower incidence of diabetes over the 20-year period in the combination lifestyle changes group [28]. The preventive benefit of lifestyle changes persisted 2 decades after the initial randomization despite the standardization of treatment for all groups over the 14 years following the study, suggesting a strong and longitudinal preventive effect of the initial lifestyle modifications. In a follow-up study of the DQDPS conducted in 2009, at 23 years of follow-up, the cumulative incidences of cardiovascular mortality and all-cause mortality were significantly lower in the intervention group (11.9% versus 19.6%, and 28.1% versus 38.4%, respectively), highlighting the long-term clinical benefits of lifestyle intervention in patients with IGT [29].

Similarly, the Finnish Diabetes Prevention Study (FDPS), published in 2001, enrolled 522 middle-aged overweight participants with IGT [23]. The participants randomly assigned to the intervention group received individualized counseling designed to reduce weight, decrease total intake of fat and saturated fat, increase intake of fiber, and increase physical activity. The control group received standard therapy. At 4 years of follow-up, the cumulative incidence of diabetes was 11% in the intervention group and 23% in the control group, with a statistically significant 58% reduction in risk for progression to diabetes. A follow-up of the FDPS was published in 2006 [31]. Participants who did not progress to diabetes in the initial 4-year study were further followed for a median of 3 years. Interestingly, lifestyle changes were maintained by the intervention group participants despite the cessation of the individual counseling, leading to a 36% relative reduction in diabetes incidence during the post-intervention follow-up period alone (4.6 vs 7.2 per 100 person-years, P = 0.041) and a 43% cumulative diabetes incidence reduction over the 7-year follow-up, demonstrating, one more time, the sustained efficacy of lifestyle changes.

In the United States, the Diabetes Prevention Program (DPP) trial is a landmark NIH-sponsored multicenter randomized controlled trial published in 2002, and one of the largest diabetes prevention studies with lifestyle changes to date [24]. A total of 3234 participants with prediabetes, defined as an IFG or IGT, were randomly assigned to an intensive lifestyle modification program, metformin 850 mg twice daily, or matching placebo. Lifestyle changes included a low-fat (< 25% of caloric intake), 1200- to 1800-calorie diet and exercise for 150 minutes a week, with a 7% body weight reduction goal and a very well structured curriculum and professional support group. The study was discontinued early (at 3 years) as the data demonstrated the superiority of lifestyle changes, with a 58% reduction in diabetes incidence in the lifestyle intervention group and a 31% reduction in the metformin group when compared to placebo (cumulative incidence of diabetes at 3 years of 28.9%, 21.7 %, and 14.4% in the placebo, metformin, and lifestyle intervention groups, respectively). Lifestyle changes were significantly more effective than metformin and were consistently effective in men and women across age, BMI, and ethnic groups.

The DPPOS (DPP Outcome Study) was a 10-year follow-up of the DPP study published in 2009 where all participants were offered group-implemented lifestyle changes and were followed for an additional 5.7 years [32]. Unlike the Finnish follow-up study, diabetes incidence was similar in the 3 treatment groups in the follow-up period. However, the cumulative incidence of diabetes remained significantly the lowest in the original lifestyle group, with a 34% cumulative risk reduction in the lifestyle group and an 18% reduction in the metformin group at 10 years when compared to placebo. Interestingly, unlike most other studies of weight-reducing interventions, in the DPPOS, patients in the lifestyle changes and metformin groups maintained weight loss at 10 years’ follow-up.

In Japan, a diabetes prevention study assigned 458 male participants with IGT to a standard intervention group or an intensive intervention group receiving detailed lifestyle modification counseling every 3 to 4 months during hospital visits [25]. The cumulative 4-year incidence of diabetes was 9.3% in the control group versus 3.0% in the intervention group, and the reduction in diabetes risk was 67.4% (P < 0.001), with body weight reductions of 0.39 kg and 2.18 kg, respectively (P < 0.001). Of note, participants with higher FBG at baseline developed diabetes at a higher rate than those with lower values. This study suggested that lifestyle change counseling conducted in an outpatient clinic setting can be very effective in preventing diabetes.

Indian adults are thought to be more insulin resistant at a younger age and at a lower BMI than Caucasians. To assess whether the DPP findings can be replicated in an Indian population, the Indian Diabetes Prevention Program (IDPP) trial randomized a total of 531 participants with IGT to 4 groups: control, lifestyle modification, metformin, and lifestyle modifications with metformin [26]. The 3-year cumulative incidences of diabetes were 55.0%, 39.3%, 40.5%, and 39.5%, respectively, showing again a significant relative reduction in progression to diabetes of 28.5% with lifestyle changes, 26.4% with metformin, and 28.2% with both lifestyle changes and metformin, as compared with the control group.

In a Japanese unmasked, multicenter, randomized controlled trial published in 2011, 641 overweight adults with IFG were randomized to a frequent intervention group, receiving individual counseling and support for lifestyle modifications 9 times over 36 months, or a control group, receiving counseling 4 times over the same period. The 3-year cumulative incidence of T2DM was significantly lower in the frequent intervention group than in the control group (12.2% vs 16.6%) [27]. Interestingly, in a posthoc subgroup analysis, the protective effect was more prominent in patients with underlying associated IGT or elevated A1C, but was not observed in patients with isolated IFG, suggesting a possible prognostic value of an additional A1C or oral glucose tolerance test in individuals with IFG.

Diet

The diet followed in the major diabetes prevention trials discussed above has typically been a weight-reducing diet with decreased fat intake (eg, DPP, Finnish trial) and increased fiber intake (eg, Da Quing, DPP, Finnish trials). However, there has been more emphasis recently on the importance of the quality rather than the quantity of fats in preventing diabetes. For example, in a Spanish study, a non–calorie-restricted traditional Mediterranean diet, enriched with high-fat foods of vegetable origin (olive oil, nuts) decreased the incidence of diabetes by 52% in individuals at high cardiovascular risk after a median follow-up of 4.0 years, and in the absence of significant changes in body weight or physical activity among the groups [33]. These findings were reproduced by other studies. A recent meta-analysis examining the relation between intake of fruits and vegetables and the incidence of diabetes revealed that higher intake of fruit, especially berries, and green, leafy vegetables, yellow vegetables, cruciferous vegetables, or their fiber is associated with a lower risk of T2DM [34].

Exercise

Exercise is thought to improve insulin sensitivity and promote peripheral glucose uptake in normal individuals. Long-term moderate exercise, similar to the exercise recommended in DPP and FDPS, results in increased translocation of insulin-responsive glucose transporter (GLUT-4) from intracellular stores to the cell surface, facilitating glucose uptake [35]. A systematic review of 10 prospective cohort studies published in 2007 showed that, after adjustment for BMI, moderate-intensity physical activity was significantly associated with reduced diabetes incidence [36]. In the FDPS, participants who achieved at least 4 hours of exercise per week had a significant 80% decrease in incidence of diabetes, and this effect was observed even in the group that did not lose weight [23]. In the DQDPS, the greatest reduction in diabetes incidence was observed in the exercise group [22].

In a recent NIH-funded trial designed to examine the relative contribution of exercise alone to the overall beneficial effect of lifestyle changes in the DPP study, a total of 237 adults with IFG were randomly assigned to 4 different groups: low-amount moderate intensity exercise (similar to exercise followed in DPP), high-amount moderate intensity exercise, high-amount vigorous intensity exercise, and a combination of diet, weight loss, and low-amount moderate exercise. Only the diet and exercise group experienced a decrease in fasting glucose, whereas similar improvements in glucose tolerance were observed in both the diet and exercise group and the high-amount moderate-intensity exercise group, suggesting that such an exercise regimen may be as effective as a more intensive multicomponent approach involving diet, exercise, and weight loss for preventing diabetes [37].

Weight Loss

Weight reduction in prediabetic individuals has been consistently associated with reduced incidence of diabetes. Furthermore, the amount of weight loss needed to achieve this benefit seems to be relatively modest and a realistic goal to set for patients. Indeed, in the DPP trial, an average weight loss of only 5.6 kg was associated with a 58% lower incidence of diabetes [24]. Moreover, on further analysis of the DPP trial, and among weight, diet, and exercise, diabetes prevention correlated most strongly with weight loss, with an estimated 16% diabetes risk reduction for every single kilogram of weight reduction [38]. Similarly, within the same lifestyle intervention group in the FDPS, the participants who were able to achieve an initial body weight loss greater than 5% at 1 year had a nearly 70% relative risk reduction in progression to diabetes, when compared to their peers in the intervention group who had less or no weight loss [23].

In summary, numerous randomized controlled studies from various populations have proved that lifestyle modifications, including healthy diet, moderate weight loss, and moderate-intensity exercise, represent a very effective strategy to prevent diabetes in patients at risk, mostly patients with IGT, and this protective effect seems to be sustained over time.

Pharmacologic Interventions

Metformin

Metformin is an antidiabetic agent that works mostly at the liver site by suppressing hepatic glucose production and inhibiting production and oxidation of free fatty acids (FFA), thereby reducing FFA-induced insulin resistance and promoting peripheral glucose uptake [39]. This effect has the potential of preserving beta cell function by reducing the demand for insulin secretion.

In the DPP trial, metformin, although generally less effective than lifestyle changes, was associated with a significant 31% reduction in diabetes incidence (cumulative incidence of 22% in metformin group vs 29% in placebo group) and significant weight reduction (average of 2 kg) [24]. Further analysis of the DPP results showed that metformin efficacy, compared to placebo, was greater in patients who were younger, had higher BMI, and had higher FBG levels. In addition, a DPP substudy of 350 women with history of gestational diabetes and IGT revealed that this group of women, who had a higher risk of progression to diabetes (71% at 3 years) when compared to women with no history of gestational diabetes, despite similar baseline glucose levels, had similar diabetes risk reduction of 50% with both metformin and lifestyle changes [40].

In the IDPP study, both lifestyle changes and metformin reduced significantly and similarly the incidence of diabetes in adults with IGT, with no observed added benefit from combining both interventions [26]. It has not been clear, however, how much of this effect of metformin is a result of pharmacologic properties masking hyperglycemia or a true protective and preventive effect. In a washout study in which 1274 DPP participants who did not progress to diabetes underwent an OGTT after 1 to 2 weeks of discontinuing metformin or placebo, the incidence of diabetes was still reduced by 25% in the metformin group, after the washout period, compared to a 31% risk reduction in the primary DPP analysis, suggesting a partially sustained rather than temporary effect of metformin [41]. In the DPPOS long-term follow-up study, metformin (850 mg twice daily as tolerated) was continued in the group initially assigned to metformin in addition to lifestyle counseling [32]. Although the progression to diabetes was similar in all groups during the 5.7-year follow-up period, the cumulative incidence of diabetes at 10 years was still reduced in the metformin group by 18% when compared to control group. Furthermore, the weight loss associated with metformin was also interestingly sustained at 10 years. A meta-analysis echoed this beneficial effect of metformin observed in the DPP trial, reporting a relative risk reduction of new-onset diabetes of 40% with the use of metformin [42].

In summary, metformin has been shown to be effective in preventing diabetes in patients at risk, especially persons with younger age, higher BMI, and history of gestational diabetes and in native Asian Indians. The protective effect of metformin seems to be sustained over the long term in follow-up studies.

Thiazolidinediones

Thiazolidinediones (TZDs) are antidiabetic agents that have been evaluated in diabetes prevention trials. TZDs are peroxisome proliferator-activated gamma receptor (PPAR-γ) agonists that work by augmenting conversion of preadipocytes to adipocytes, which in turn increase adiponectin levels, promoting insulin sensitivity [43]. In addition to their antihyperglycemic properties, TZDs are thought to have a direct protective effect on beta cells, potentially translating into prevention and delay of diabetes [44].

The first study to demonstrate diabetes prevention with a TZD was the TRIPOD study (Troglitazone in Prevention of Diabetes), in which 266 Hispanic women with a history of gestational diabetes were randomly assigned to troglitazone or placebo [45]. Troglitazone use was significantly associated with reduction of progression to diabetes at 1.5-year follow-up when compared to placebo (relative risk reduction of 55%), with a decrease of endogenous insulin requirement at 3 months of therapy and sustained benefit after discontinuation of the TZD, suggesting an effect on beta cell preservation.

Moreover, troglitazone was an investigational drug in the DPP trial from 1996 to 1998, at which time it was discontinued because of associated fatal liver failure in a DPP participant. In the DPP trial, troglitazone was asso-ciated with a remarkable 75% decrease in progression to diabetes at 1 year. Troglitazone was withdrawn from the US market in 2000 because of its association with severe hepatotoxicity.

The international DREAM (Diabetes REduction Assessment with ramipril and rosiglitazone Medications) trial randomly assigned more than 5000 participants with IFG and/or IGT to rosiglitazone, ramipril, or placebo in a 2 × 2 factorial design [46]. In participants receiving rosiglitazone, the risk for progression to diabetes was reduced by 60% and the likelihood of regression to normoglycemia was increased by 71% when compared to placebo. However, the use of rosiglitazone was associated with an increased risk of new-onset congestive heart failure and a mean weight gain of 2.2 kg, thought to reflect increased subcutaneous gluteal fat deposition, with an observed decreased waist-to-hip ratio.

Interestingly, in a passive follow-up of the DREAM study conducted a median 1.6 years after the end of the trial and 4.3 years after randomization, participants treated with rosiglitazone had a 39% lower incidence of diabetes compared to placebo participants, and 17% more of them regressed from prediabetes to normoglycemia [47]. Nonetheless, there was no difference between the 2 groups when the analysis was restricted to the passive follow-up period, suggesting a time-limited exposure to rosiglitazone reduces the longer-term incidence of diabetes by likely delaying but not reversing the underlying disease process.

The third large trial assessing the efficacy of a TZD in preventing diabetes was the Actos Now for the prevention of diabetes (ACT NOW) trial, which was a randomized, double-blinded study that assigned 602 patients with IGT to pioglitazone 45 mg daily or placebo [48]. Over a median follow-up of 2.6 years, pioglitazone was associated with a 72% lower annual rate of progression to diabetes (2.1% compared to 7.6 % in placebo group), and a higher rate of conversion to normal glucose tolerance (48%). In addition, pioglitazone had favorable effects on fasting and 2-hour blood glucose, A1C level, diastolic blood pressure, carotid intima thickness, and HDL cholesterol. As in the DREAM trial, an increased incidence of edema and weight gain was observed with pioglitazone.

Unlike the strong evidence supporting TZDs as an approach to diabetes prevention in the US trials, the Indian Diabetes Prevention Program-2 (IDPP-2) trial, which randomized 497 participants with IGT to lifestyle modifications with pioglitazone versus lifestyle modifications with placebo, did not demonstrate a significant reduction in diabetes at 3 years’ follow-up, suggesting a possible ethnicity-related variation in the effect of the medication [49]. In 2011, the French and German medications regulatory agency withdrew pioglitazone from the market because of a potential increase in incidence of bladder cancer with the cumulative use of more than 28 g of pioglitazone. In the United States, the Food and Drug Administration is performing an extensive review of data and advises against the use of pioglitazone in patients with a history of bladder cancer.

In summary, TZDs demonstrated significant efficacy in preventing diabetes in many patients at risk, but their safety concerns, particularly the associated new onset of congestive heart failure and potential increased risk of bladder cancer, might outweigh this benefit.

Combination Metformin and Thiazolidinediones

As metformin and rosiglitazone both have preventive benefits in diabetes, and rosiglitazone is associated with numerous side effects at a higher dose, a combination of metformin and low-dose rosiglitazone was evaluated in in the CAnadian Normoglycemia Outcomes Evaluation (CANOE) trial [50]. A total of 207 patients with IGT were randomly assigned to receive combination metformin (500 mg twice daily) and rosiglitazone (2 mg daily) versus placebo for a median of 3.9 years. The combination therapy was associated with a 66% relative risk reduction of progression to diabetes.

Alpha-glucosidase Inhibitors

Alpha-glucosidase inhibitors are antidiabetic agents that slow oral carbohydrate intestinal absorption, subsequently improving postprandial hyperglycemia, which can eventually reduce glucose toxicity of pancreatic beta cells. In addition, they have been shown to improve insulin sensitivity in individuals with IGT [51] and have been found to exert a favorable protective effect in a prediabetic population [52]. In a multicenter placebo-controlled randomized trial, the Study to Prevent Non-Insulin Dependent Diabetes Mellitus (STOP-NIDDM), 1429 participants with IGT were randomly assigned to receive acarbose 100 mg 3 times a day or placebo for 3 years [53]. As expected, diabetes incidence was significantly decreased by 25% in the acarbose group (relative risk of 32.4% vs 41.5% in acarbose and placebo group, respectively), and acarbose significantly increased reversion to normal glucose tolerance (P < 0.0001). Furthermore, the use of acarbose was associated with a statistically significant 49% decrease in the rate of any cardiovascular event, highlighting the cardiovascular protective effect of improving postprandial hyperglycemia with acarbose. This study had many limitations: a high percentage of participants discontinued treatment (31% in the acarbose group and 19% in the placebo group), most likely related to increased gastrointestinal adverse effects of acarbose. In addition, the diabetes prevention effect does not seem to be sustained: during a 3-month wash-out period where all patients received placebo, incidence of diabetes in the initial intervention group was higher than in the initial placebo group.

In a Japanese multicenter randomized double-blind trial, 1780 patients with IGT were randomly assigned to receive the alpha-glucosidase inhibitor voglibose or placebo [54]. An interim analysis at 48 weeks revealed a significantly lower risk of progression to diabetes in the voglibose group.

Combination Metformin and Acarbose

In a 6-year multicenter British study, the Early Diabetes Intervention Trial (EDIT), 631 participants with IFG were randomly assigned, in a factorial design, to double-blind treatment with acarbose or placebo and simultaneously to metformin or placebo [55]. At 3 years, there was a nonsignificant risk reduction of 8% and 37% in progression to 2 successive fasting plasma glucose values of 140 mg/dL or more in the acarbose and metformin groups, respectively, but a significantly lower 2-hour OGTT glucose in the acarbose group and significantly lower FBG in the metformin group. Interestingly, at 6 years of follow-up, there was no significant difference in relative risk of progression to diabetes with acarbose, metformin, or combination therapy [56]. However, unlike metformin or combination therapy, acarbose was associated with a significant relative risk reduction of diabetes (0.66, P = 0.046) in the subgroup of patients with IGT at baseline, suggesting a possible differential protective effect of certain agents in patients with IGT or IFG.

Nateglinide

Nateglinide is a short-acting insulin secretagogue that is mostly used in the treatment of postprandial hyperglycemia in diabetic patients. The protective effect of nateglinide in a prediabetic population was examined in the NAVIGATOR study (the NAteglinide and Valsartan in Impaired Glucose Tolerance Outcomes Research), a large prospective multinational, randomized, double-blind, placebo-controlled trial. Nateglinide (30–60 mg 3 times daily) and valsartan (80–160 mg daily) versus placebo were used in a 2×2 factorial design in 9306 participants with IGT and increased risk of cardiovascular events [57]. At 5 years, nateglinide did not reduce the cumulative incidence of diabetes or cardiovascular outcomes, when compared to placebo, whereas risk of hypoglycemia was significantly increased in the intervention group.

Liraglutide

Liraglutide is an injectable glucagon-like peptide-1 (GLP-1) receptor agonist used to treat T2DM, and recently approved as a weight-reducing agent at the dose of 3 mg injected subcutaneously. GLP-1 receptor agonists work by stimulating insulin secretion in a glucose-dependent manner, suppressing glucagon secretion, inducing satiety, and slowing gastric emptying. In the international double-blind SCALE (Satiety and Clinical Adiposity-Liraglutide Evidence) trial, 3731 nondiabetic patients, among whom 61.2% had prediabetes, were randomly assigned to liraglutide 3 mg subcutaneous injection daily or placebo, in addition to diet and exercise [58]. Liraglutide was associated with lower glucose levels on OGTT and lower A1C values at the end of the study (56 weeks), with this decrease especially prominent in prediabetic patients. Significantly fewer participants in the liraglutide group (4/2219) compared to the placebo group (14/1225) developed diabetes at 56 weeks, nearly all of whom (except for 1 in the placebo group) had prediabetes at the beginning of the study. Of note, the liraglutide group had a mean 8.4-kg weight reduction by week 56, compared to 2.8 kg in the placebo group.

Insulin

Insulin has also been investigated as a possible diabetes prevention agent, given the assumed protective effect insulin could exert on beta cell reserve. In the landmark international Outcome Reduction with Initial Glargine Intervention (ORIGIN) trial, 12,537 participants (mean age 63.5 years) with cardiovascular risk factors plus IFG, IGT, or type 2 diabetes were randomly assigned to receive insulin glargine (with a target FBG ≤ 95 mg/dL) or standard care and were monitored for cardiovascular outcomes and other secondary endpoints including incidence of diabetes [59]. After a median follow-up of 6.2 years, and 3 months after discontinuation of therapy, among the 1456 participants without baseline diabetes, new diabetes was diagnosed in 30% of participants receiving glargine versus 35% of those receiving standard therapy. However, rates of severe hypoglycemia and modest weight gain were higher in the insulin group, calling in to question the benefit/risk balance with the use of basal insulin for diabetes prevention.

ACE Inhibitors and ARBs

A possible diabetes preventive effect was observed with renin-angiotensin system (RAS) blockade agents in secondary analysis of several hypertension trials, such as with ramipril in the Heart Outcomes Prevention Evaluation study, captopril (compared to diuretics and beta blockers) in the CAptopril Prevention Project, lisinopril (compared to amlodipine and chlorthalidone) in the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial, losartan (compared to atenolol) in the Losartan Intervention For Endpoint reduction in hypertension study), and multiple other randomized controlled trials [60–64]. Therefore, 2 major trials were designed to examine, as a primary outcome, the effect of RAS inhibition on diabetes prevention in a population at risk. The DREAM trial randomly assigned, in a 2 × 2 factorial design, 5269 relatively healthy participants with IGT and/or IFG to rosiglitazone, ramipril, or placebo [65]. Although the use of ramipril at a dose of 15 mg daily for 3.5 years did not prevent diabetes significantly, it was associated with a 9%, nonsignificant decrease in new-onset of diabetes and a 16%, significant increase in regression of IFG and IGT to normoglycemia, as well as a significant decrease in OGTT 2-hour glucose level (135.1 vs 140.5 mg/dL) with no improvement in FBG.

Similarly, in the NAVIGATOR trial that examined the effect of nateglinide and valsartan on the prevention of diabetes in 9306 participants with IGT and increased risk of cardiovascular events, valsartan significantly but slightly reduced the incidence of diabetes at 5 years, by 14%, when compared to placebo (33% versus 37%, respectively), with no significant reduction in cardiovascular outcome [66]. Unlike in the DREAM study, the patients enrolled in the NAVIGATOR trial had established cardiovascular disease or cardiovascular risk factors and assumable elevated RAS activation level. This baseline population difference might explain the more significant effect of RAS inhibition in the NAVIGATOR trial.

Given the positive glycemic effect of ACE inhibitors and ARBs, their use should be encouraged in prediabetic patients when indicated for treatment of high blood pressure or cardiovascular disease. Different mechanisms could explain this favorable glycemic impact: inhibition of the post-receptor insulin signaling abnormalities, increased blood flow to the skeletal muscle facilitating insulin action, enhanced differentiation of pre-adipocytes into mature adipocytes, and increased pancreatic islet blood perfusion leading to appropriate insulin release and possible partial PPAR-γ activity [67].

Xenical

Xenical is a gastrointestinal lipase inhibitor approved for use for weight reduction and maintenance. A possible diabetes prevention benefit of xenical was initially suggested by a retrospective analysis of xenical treatment effects on obese patients with IGT [68]. This finding was subsequently confirmed by a multicenter randomized placebo-controlled study, XENical in the prevention of Diabetes in Obese Subjects (XENDOS), where 3305 obese subjects, with normal glucose tolerance or IGT were randomly assigned to either xenical 120 mg 3 times a day or placebo, in addition to lifestyle changes for all participants [69]. In the group of patients with IGT (694 subjects), xenical treatment was associated with a 45% risk reduction of progression to diabetes at 4 years (18.8% versus 28.8% in placebo), whereas participants with baseline normal glucose tolerance had no significant change in incidence of diabetes. On the other hand, weight reduction at 4 years was significantly greater in all patients who received xenical (5.8 kg in intervention group vs 3 kg in control group). The beneficial effect of xenical in diabetes prevention seems to be additive to the benefit of weight loss. As in many weight reduction trials, this study was limited by the high discontinuation rate in both groups (48% in xenical group and 66% in control group), probably related to insufficient clinical response.

Fibric Acid Derivatives (Bezafibrate)

Bezafibrate, a nonselective ligand/activator for PPAR-α, was found to reduce not only triglycerides, but also FPG, fructosamine, and A1C levels significantly in T2DM patients with hyperlipidemia [70]. Different mechanisms of glucose lowering have been suggested with bezafibrate: nonselective activation of PPAR-γ, improving insulin sensitivity, and enhancing glucose disposal in adipose tissue and skeletal muscles [71]. Furthermore, bezafibrate treatment was associated with decreased incidence of diabetes in patients with IFG and in obese non-diabetic patients with normal glycemic levels [72,73]. In a posthoc analysis of the Bezafibrate Infarction Prevention study, 303 patients with IFG received either 400 mg of bezafibrate daily or placebo [73]. Over a mean follow-up of 6.2 years, development of diabetes was less prevalent (54.4% vs 42.3%, relative risk reduction of 22%) and delayed (mean 10 months) in the bezafibrate group compared to placebo. Multivariate analysis identified bezafibrate as an independent predictor of decreased risk of new diabetes development, regardless of BMI and lipid profile.

Surgery

Over the past decade, bariatric surgery has become one of the most effective interventions for inducing and sustaining weight reduction in severely obese patients, leading to a significant benefit in diabetes prevention or remission. The Swedish Obese Subject Study is a large ongoing prospective nonrandomized cohort study that between 1987 and 2001 enrolled 4047 nondiabetic obese participants who underwent gastric surgery or were matched obese control, with diabetes incidence measured at 2, 10 and 15 years [74–76]. At 15 years, analysis of the available cohort of the initial group showed that T2DM developed in 392 of 1658 control participants and in 110 of 1771 bariatric-surgery participants, corresponding to incidence rates of 28.4 and 6.8 cases per 1000 person-years, respectively (P < 0.001). The treatment effects on the incidence of T2DM were at least as strong after 2 years and 10 years of follow-up as after 15 years. This effect was most prominent among the 591 patients who had IFG at baseline, with a number needed to treat as low as 1.3. The surgery group maintained an average 20-kg weight loss at 15 years.

In another study of the effects of bariatric surgery, 150 of 152 obese participants with IGT who underwent gastric bypass achieved and maintained a normal glycemic profile at 14 years of follow-up [77]. Similarly, in a follow-up of 136 obese participants with IGT, 109 of whom underwent bariatric surgery, 1 participant in the surgical group developed diabetes, as compared with 6 out of 27 in the control group [78]. In a meta-analysis including studies involving 22,094 patients who underwent bariatric surgery, 76.8% had complete resolution of their diabetes [79]. The rapid improvement of glycemic profile after bariatric surgery is thought to be due to oral intake restriction as well as acute hormonal changes related to the exclusion of the upper gastrointestinal tract (eg, incretin and ghrelin levels variations) [80].

Conclusions and Recommendations

The natural history of T2DM allows identification of patients at risk for diabetes and implementation of prevention strategies, which seems to be a public health need given the alarming increase in diabetes incidence. Indeed, the onset of T2DM is typically preceded by many years of beta cell dysfunction translating into carbohydrate metabolism abnormalities such as IFG and IGT, providing an excellent window of opportunity to identify persons at risk and prevent progression to diabetes. Numerous randomized controlled trials established lifestyle modifications, including dietary changes, moderate weight loss, and moderate intensity physical activity, as safe and effective interventions to prevent diabetes. This protective effect has been consistently shown to be sustained for more than 10 years after the initial intervention. Pharmacologic agents such as metformin, thiazolidinediones, alpha-glucosidase inhibitors, xenical, liraglutide, and insulin have also been associated with diabetes prevention in patients at risk. However, except for metformin, safety concerns or lack of durable efficacy or tolerability seem to outweigh their potential diabetes prevention benefit.

Given their favorable glycemic effect, RAS blockade and fibrates should be considered, when indicated, as reasonable treatment options for hypertension and hyperlipidemia in prediabetic patients. Bariatric surgery has been associated with a dramatic reduction in diabetes incidence in obese prediabetic patients and can be considered an alternative prevention measure in patients with severe obesity and prediabetes.

The recently updated ADA guidelines recommend referring patients with prediabetes to an intensive diet and physical activity behavioral counseling program; diet and activity goals should adhere to the tenets of the DPP, with a loss of 7% of body weight and at least 150 minutes of moderate physical activity (eg, brisk walking) per week [8]. Metformin therapy for diabetes prevention should be considered in patients with prediabetes, especially in those with BMI greater than 35 kg/m², those younger than 60 years of age, women with history of gestational diabetes, and/or those with rapidly rising A1C despite lifestyle modifications. Monitoring for development of diabetes, at least annually, and screening for and treatment of modifiable cardiovascular risk factors are suggested in patients with prediabetes [8].

Many lessons have been learned through the studies of diabetes prevention interventions. The challenge that remains is how to apply these interventions, especially the lifestyle modifications, in real world medical practice, at both the individual and public health level.

Corresponding author: Jocelyne Karam, MD, 4802 10th Avenue, Brooklyn, NY 11219, [email protected].

Financial disclosures: None reported.

From the Maimonides Medical Center (Dr. Karam) and the SUNY Downstate Medical Center (Dr. Karam and Dr. McFarlane), Brooklyn, NY.

Abstract

• Objective. To discuss the epidemic of diabetes highlighting the natural history of the disease and the major clinical trials aimed at diabetes prevention in different prediabetic populations around the world.
• Results. Diabetes prevention studies have evaluated various interventions including lifestyle modifications, metformin, alpha-glucosidase inhibitors, thiazolidinediones, nateglinide, and xenical as well as the renin-angiotensin aldosterone system (RAS) inhibitors. Lifestyle modifications seem to be the safest, most effective, and most sustainable intervention to prevent diabetes. Except for metformin, the potential diabetes prevention benefits of the studied pharmacologic agents are limited by safety concerns or lack of durable efficacy or tolerability. RAS blockade and fibrates have a favorable glycemic effect, and, when indicated, are reasonable treatment options for hypertension and hyperlipidemia in prediabetic patients.
• Conclusion. As recommended by American Diabetes Association guidelines, patients with prediabetes should be referred to an intensive diet and physical activity behavioral counseling program; diet and activity goals include a loss of 7% of body weight and at least 150 minutes of moderate physical activity per week. Metformin therapy for diabetes prevention should be considered as well.

Key words: prediabetes; type 2 diabetes mellitus, diabetes prevention, lifestyle modifications.

Diabetes mellitus has reached pandemic proportions across the globe. The International Diabetes Federation (IDF) estimates that in 2015 around 415 million people, or 1 in 11 adults, had diabetes, compared to 285 million in 2010, with 5 million deaths, or 1 death every 6 seconds, occurring because of diabetes or diabetes complications [1]. In the United States, an estimated 29.1 million Americans, or 9.3% of the population, have diabetes, 27.8% of them undiagnosed [2]. The prevalence of diabetes increases significantly with age, affecting around 16.2% of American adults aged 45 to 64 years and 25.9% of adults aged 65 years or older [2]. The Centers for Disease Control and Prevention (CDC) estimates that, with current trends, as many as 1 in 3 American adults could have diabetes by 2050 [3].

Type 2 diabetes mellitus (T2DM) accounts for the majority of prevalent and newly diagnosed diabetes in the world, and is strongly linked to overweight and inactivity in adults [4]. T2DM is increasingly being diagnosed in pediatric patients, in whom type 1 diabetes has historically been predominant; it now accounts for approximately 30% of newly diagnosed diabetes in children aged 10 to 19 years, exceeding 50% in certain ethnicities such as non-Hispanic black and American Indian/Alaska Native children [2].

These alarming trends have spurred significant research and public efforts aimed at reducing the prevalence of diabetes by preventing T2DM. Indeed, insulin resistance and abnormal carbohydrate metabolism progress over many years prior to the diagnosis of diabetes and manifest with different clinical and biochemical features. Both the pathophysiology and the natural history of T2DM offer clinicians an opportunity to identify patients at risk for developing the disease and to implement prevention strategies. This article outlines the risk factors and diagnostic criteria for prediabetes, describes the studies that have explored diabetes prevention through lifestyle changes, pharmacotherapy, or surgery, and reviews recommendations for managing patients at risk.

Risk Factors and Screening for T2DM

The American Diabetes Association (ADA) recommends screening all adults for prediabetes by assessing for diabetes risk factors [8]. Glucose testing is recommended in individuals aged 45 years or older, and should be considered in adults of any age who are overweight or obese (body mass index [BMI] ≥ 25 kg/m² or ≥ 23 kg/m² in Asian Americans) and have 1 or more additional risk factors for diabetes. Testing also should be considered in children and adolescents who are overweight or obese and who have 2 or more additional risk factors. If tests are normal, repeat testing carried out at a minimum of 3-year intervals is suggested [8].

Prediabetes

Abnormalities in glucose metabolism progress along a continuum through various stages before T2DM develops. Years before the development of overt diabetes, and especially in the presence of excessive visceral fat, cellular sensitivity to insulin gradually decreases, leading to a compensatory increased insulin secretion [9]. With time, and under continuous increased demand, pancreatic beta cell function declines and ultimately fails to overcome insulin resistance and maintain a normal glucose metabolism, resulting in prediabetes followed by the development of diabetes. This early beta cell dysfunction was illustrated by the decreased beta cell volume observed on autopsy of obese patients with IFG or T2DM, when compared to obese individuals with normal glucose tolerance [10]. It is estimated that around 40% to 70% of beta cell function is already lost by the time diabetes is clinically diagnosed. This relatively slow pathophysiologic process allows the identification of at-risk patients well before their blood glucose levels reach the diabetic diagnostic thresholds, and therefore presents an opportunity for prevention.

Diagnostic Criteria

The ADA guidelines released in 2003 define prediabetes as IFG (fasting blood glucose [FBG] levels of 100–125 mg/dL), IGT (glucose levels of 140–199 mg/dL at 2 hours during an oral glucose tolerance test [OGTT] following an oral load of 75 g of dextrose), or both. Additionally, hemoglobin A1C (A1C) was introduced as a diagnostic tool for prediabetes in 2010, with values between 5.7% and 6.4% indicating prediabetes [8]. Most of these thresholds were chosen due to their association with increased rates of complications, notably retinopathy and cardiovascular disease.

A combined report from the World Health Organization (WHO) and the IDF published in 2006 defined intermediate hyperglycemia as IFG, but with a higher cutoff for FBG (110–125 mg/dL) than the ADA’s definition, and/or IGT (2-hour OGTT glucose level of 140–199 mg/dL) [11]. The rationale for a higher cut-point for IFG is the concern about the increased prevalence of IFG and its impact on individuals and health systems and the more favorable cardiovascular risk profile and decreased risk of progression to diabetes in the group of patients with FBG of 100 to 110 mg/dL when compared to the group with FBG of 110 to 125 mg/dL. The report does not recommend the use of A1C in the diagnosis of diabetes or intermediate hyperglycemia because of a lack of global consistency and the potential for other factors that can be prevalent in some developing countries, such as hemoglobinopathies and anemia, to interfere with the assay.

Prevalence and Progression to Diabetes

According to CDC data from 2014, up to 86 million American adults, more than 1 in 3, have prediabetes, and 9 out of 10 of these individuals are undiagnosed [2]. It is estimated that approximately 25% of people diagnosed with either IFG or IGT progress to diabetes mellitus over a 3- to 5-year period [12]. If observed for longer periods, most prediabetic persons will probably develop diabetes. The highest rate of progression to diabetes is observed in patients with both IFG and IGT, older age, overweight, or other diabetic risk factors.

Complications

In addition to increasing the risk for progression to diabetes, prediabetes is independently associated with microvascular and macrovascular complications and increased risk of death, prior to the actual onset of diabetes. The DECODE study demonstrated significantly increased mortality in 2766 individuals with IGT after 7 years of follow-up, when compared to normoglycemic patients; this effect was more prominent in participants with IGT than in participants with IFG [13]. In the Australian Diabetes, Obesity and Lifestyle Study, IFG was found to be an independent predictor for cardiovascular mortality after adjustment for age, sex, and other traditional cardiovascular risk factors [14].

Similarly, a recent meta-analysis demonstrated that the presence of IFG was significantly associated with future risk for coronary heart disease (CHD), with the risk increase starting when fasting plasma glucose was as low as 100 mg/dL; however, this finding may have been confounded by the presence of undetected IGT or other cardiovascular risk factors [15]. Another recent systematic review of 53 prospective cohort studies with 1,611,339 participants showed that prediabetes (IFG or IGT) was associated with an increased risk of composite cardiovascular disease, CHD, stroke, and all-cause mortality [16].

The association between retinopathy and prediabetes has been described in multiple reports and this association has helped guide authors on selected thresholds for diagnosis of prediabetes. For example, in 1 study, the incidence of retinopathy in individuals with IGT was 12% among Pima Indians [17]. Similarly, in a follow-up study of the Diabetes Prevention Program, 8% of prediabetic participants who remained nondiabetics had evidence of retinopathy [18].

Neuropathy also has been observed in prediabetes. A noninvasive neurologic evaluation of individuals with IGT revealed subclinical neural dysfunction suggestive of cardiovascular autonomic neuropathy [19]. At the clinical level, a study that evaluated 100 patients with chronic idiopathic axonal neuropathy of unknown etiology found IFG in 36 and IGT in 38 patients, underscoring the role of abnormal glucose metabolism in these patients [20].

Nephropathy may also be more prevalent in those with prediabetes. In a 1999–2006 National Health and Nutrition Examination Survey analysis, the adjusted prevalence of chronic kidney disease, defined by estimated glomerular filtration rate (eGFR) of 15 to 59 mL/min per 1.73 m² or albumin-creatinine ratio ≥ 30 mg/g, was 17.1% in individuals with IFG, compared to 11.8% in individuals with normal fasting glucose [21].

Lifestyle Modifications

The alarming rapid increase in the prevalence of T2DM has been linked to a parallel rising epidemic of overweight, obesity, and lack of physical activity. Therefore, lifestyle changes aiming at weight reduction seemed to be a natural individual and public health strategy to prevent diabetes, and such strategies have been the focus of many randomized controlled trials around the world. As anticipated, weight loss, exercise, and diet have all been shown, separately or in combination, to be effective in decreasing the incidence of T2DM in high-risk patients [22–27]. Furthermore, and well beyond the benefit observed during the trials, follow-up studies revealed a sustained reduction of diabetes incidence in intervention groups several years after cessation of the intervention [28–32] (Table 2).

The Da Quing Diabetes Prevention Study (DQDPS), published in 1997, is one of the earliest prospective diabetes prevention trials [22]. This 6-year study conducted in 33 clinics in China from 1986 through 1992 included 577 participants with IGT who were randomly assigned to 1 of 4 groups: (1) diet (high vegetables, low sugar/alcohol) only, (2) exercise, (3) diet plus exercise, and (4) standard of care. At 6 years, diabetes incidence was significantly reduced by 46% in the exercise group, 31% in the diet group, and 42% in the diet plus exercise group compared to standard care. In 2006, 14 years after the end of the trial and 20 years after the initial enrollment, the cumulative incidence of diabetes was significantly lower in the intervention group at 80%, compared to 93% in the control group, and the annual incidence of diabetes was 7% and 11%, respectively, with a 43% lower incidence of diabetes over the 20-year period in the combination lifestyle changes group [28]. The preventive benefit of lifestyle changes persisted 2 decades after the initial randomization despite the standardization of treatment for all groups over the 14 years following the study, suggesting a strong and longitudinal preventive effect of the initial lifestyle modifications. In a follow-up study of the DQDPS conducted in 2009, at 23 years of follow-up, the cumulative incidences of cardiovascular mortality and all-cause mortality were significantly lower in the intervention group (11.9% versus 19.6%, and 28.1% versus 38.4%, respectively), highlighting the long-term clinical benefits of lifestyle intervention in patients with IGT [29].

Similarly, the Finnish Diabetes Prevention Study (FDPS), published in 2001, enrolled 522 middle-aged overweight participants with IGT [23]. The participants randomly assigned to the intervention group received individualized counseling designed to reduce weight, decrease total intake of fat and saturated fat, increase intake of fiber, and increase physical activity. The control group received standard therapy. At 4 years of follow-up, the cumulative incidence of diabetes was 11% in the intervention group and 23% in the control group, with a statistically significant 58% reduction in risk for progression to diabetes. A follow-up of the FDPS was published in 2006 [31]. Participants who did not progress to diabetes in the initial 4-year study were further followed for a median of 3 years. Interestingly, lifestyle changes were maintained by the intervention group participants despite the cessation of the individual counseling, leading to a 36% relative reduction in diabetes incidence during the post-intervention follow-up period alone (4.6 vs 7.2 per 100 person-years, P = 0.041) and a 43% cumulative diabetes incidence reduction over the 7-year follow-up, demonstrating, one more time, the sustained efficacy of lifestyle changes.

In the United States, the Diabetes Prevention Program (DPP) trial is a landmark NIH-sponsored multicenter randomized controlled trial published in 2002, and one of the largest diabetes prevention studies with lifestyle changes to date [24]. A total of 3234 participants with prediabetes, defined as an IFG or IGT, were randomly assigned to an intensive lifestyle modification program, metformin 850 mg twice daily, or matching placebo. Lifestyle changes included a low-fat (< 25% of caloric intake), 1200- to 1800-calorie diet and exercise for 150 minutes a week, with a 7% body weight reduction goal and a very well structured curriculum and professional support group. The study was discontinued early (at 3 years) as the data demonstrated the superiority of lifestyle changes, with a 58% reduction in diabetes incidence in the lifestyle intervention group and a 31% reduction in the metformin group when compared to placebo (cumulative incidence of diabetes at 3 years of 28.9%, 21.7 %, and 14.4% in the placebo, metformin, and lifestyle intervention groups, respectively). Lifestyle changes were significantly more effective than metformin and were consistently effective in men and women across age, BMI, and ethnic groups.

The DPPOS (DPP Outcome Study) was a 10-year follow-up of the DPP study published in 2009 where all participants were offered group-implemented lifestyle changes and were followed for an additional 5.7 years [32]. Unlike the Finnish follow-up study, diabetes incidence was similar in the 3 treatment groups in the follow-up period. However, the cumulative incidence of diabetes remained significantly the lowest in the original lifestyle group, with a 34% cumulative risk reduction in the lifestyle group and an 18% reduction in the metformin group at 10 years when compared to placebo. Interestingly, unlike most other studies of weight-reducing interventions, in the DPPOS, patients in the lifestyle changes and metformin groups maintained weight loss at 10 years’ follow-up.

In Japan, a diabetes prevention study assigned 458 male participants with IGT to a standard intervention group or an intensive intervention group receiving detailed lifestyle modification counseling every 3 to 4 months during hospital visits [25]. The cumulative 4-year incidence of diabetes was 9.3% in the control group versus 3.0% in the intervention group, and the reduction in diabetes risk was 67.4% (P < 0.001), with body weight reductions of 0.39 kg and 2.18 kg, respectively (P < 0.001). Of note, participants with higher FBG at baseline developed diabetes at a higher rate than those with lower values. This study suggested that lifestyle change counseling conducted in an outpatient clinic setting can be very effective in preventing diabetes.

Indian adults are thought to be more insulin resistant at a younger age and at a lower BMI than Caucasians. To assess whether the DPP findings can be replicated in an Indian population, the Indian Diabetes Prevention Program (IDPP) trial randomized a total of 531 participants with IGT to 4 groups: control, lifestyle modification, metformin, and lifestyle modifications with metformin [26]. The 3-year cumulative incidences of diabetes were 55.0%, 39.3%, 40.5%, and 39.5%, respectively, showing again a significant relative reduction in progression to diabetes of 28.5% with lifestyle changes, 26.4% with metformin, and 28.2% with both lifestyle changes and metformin, as compared with the control group.

In a Japanese unmasked, multicenter, randomized controlled trial published in 2011, 641 overweight adults with IFG were randomized to a frequent intervention group, receiving individual counseling and support for lifestyle modifications 9 times over 36 months, or a control group, receiving counseling 4 times over the same period. The 3-year cumulative incidence of T2DM was significantly lower in the frequent intervention group than in the control group (12.2% vs 16.6%) [27]. Interestingly, in a posthoc subgroup analysis, the protective effect was more prominent in patients with underlying associated IGT or elevated A1C, but was not observed in patients with isolated IFG, suggesting a possible prognostic value of an additional A1C or oral glucose tolerance test in individuals with IFG.

Diet

The diet followed in the major diabetes prevention trials discussed above has typically been a weight-reducing diet with decreased fat intake (eg, DPP, Finnish trial) and increased fiber intake (eg, Da Quing, DPP, Finnish trials). However, there has been more emphasis recently on the importance of the quality rather than the quantity of fats in preventing diabetes. For example, in a Spanish study, a non–calorie-restricted traditional Mediterranean diet, enriched with high-fat foods of vegetable origin (olive oil, nuts) decreased the incidence of diabetes by 52% in individuals at high cardiovascular risk after a median follow-up of 4.0 years, and in the absence of significant changes in body weight or physical activity among the groups [33]. These findings were reproduced by other studies. A recent meta-analysis examining the relation between intake of fruits and vegetables and the incidence of diabetes revealed that higher intake of fruit, especially berries, and green, leafy vegetables, yellow vegetables, cruciferous vegetables, or their fiber is associated with a lower risk of T2DM [34].

Exercise

Exercise is thought to improve insulin sensitivity and promote peripheral glucose uptake in normal individuals. Long-term moderate exercise, similar to the exercise recommended in DPP and FDPS, results in increased translocation of insulin-responsive glucose transporter (GLUT-4) from intracellular stores to the cell surface, facilitating glucose uptake [35]. A systematic review of 10 prospective cohort studies published in 2007 showed that, after adjustment for BMI, moderate-intensity physical activity was significantly associated with reduced diabetes incidence [36]. In the FDPS, participants who achieved at least 4 hours of exercise per week had a significant 80% decrease in incidence of diabetes, and this effect was observed even in the group that did not lose weight [23]. In the DQDPS, the greatest reduction in diabetes incidence was observed in the exercise group [22].

In a recent NIH-funded trial designed to examine the relative contribution of exercise alone to the overall beneficial effect of lifestyle changes in the DPP study, a total of 237 adults with IFG were randomly assigned to 4 different groups: low-amount moderate intensity exercise (similar to exercise followed in DPP), high-amount moderate intensity exercise, high-amount vigorous intensity exercise, and a combination of diet, weight loss, and low-amount moderate exercise. Only the diet and exercise group experienced a decrease in fasting glucose, whereas similar improvements in glucose tolerance were observed in both the diet and exercise group and the high-amount moderate-intensity exercise group, suggesting that such an exercise regimen may be as effective as a more intensive multicomponent approach involving diet, exercise, and weight loss for preventing diabetes [37].

Weight Loss

Weight reduction in prediabetic individuals has been consistently associated with reduced incidence of diabetes. Furthermore, the amount of weight loss needed to achieve this benefit seems to be relatively modest and a realistic goal to set for patients. Indeed, in the DPP trial, an average weight loss of only 5.6 kg was associated with a 58% lower incidence of diabetes [24]. Moreover, on further analysis of the DPP trial, and among weight, diet, and exercise, diabetes prevention correlated most strongly with weight loss, with an estimated 16% diabetes risk reduction for every single kilogram of weight reduction [38]. Similarly, within the same lifestyle intervention group in the FDPS, the participants who were able to achieve an initial body weight loss greater than 5% at 1 year had a nearly 70% relative risk reduction in progression to diabetes, when compared to their peers in the intervention group who had less or no weight loss [23].

In summary, numerous randomized controlled studies from various populations have proved that lifestyle modifications, including healthy diet, moderate weight loss, and moderate-intensity exercise, represent a very effective strategy to prevent diabetes in patients at risk, mostly patients with IGT, and this protective effect seems to be sustained over time.

Pharmacologic Interventions

Metformin

Metformin is an antidiabetic agent that works mostly at the liver site by suppressing hepatic glucose production and inhibiting production and oxidation of free fatty acids (FFA), thereby reducing FFA-induced insulin resistance and promoting peripheral glucose uptake [39]. This effect has the potential of preserving beta cell function by reducing the demand for insulin secretion.

In the DPP trial, metformin, although generally less effective than lifestyle changes, was associated with a significant 31% reduction in diabetes incidence (cumulative incidence of 22% in metformin group vs 29% in placebo group) and significant weight reduction (average of 2 kg) [24]. Further analysis of the DPP results showed that metformin efficacy, compared to placebo, was greater in patients who were younger, had higher BMI, and had higher FBG levels. In addition, a DPP substudy of 350 women with history of gestational diabetes and IGT revealed that this group of women, who had a higher risk of progression to diabetes (71% at 3 years) when compared to women with no history of gestational diabetes, despite similar baseline glucose levels, had similar diabetes risk reduction of 50% with both metformin and lifestyle changes [40].

In the IDPP study, both lifestyle changes and metformin reduced significantly and similarly the incidence of diabetes in adults with IGT, with no observed added benefit from combining both interventions [26]. It has not been clear, however, how much of this effect of metformin is a result of pharmacologic properties masking hyperglycemia or a true protective and preventive effect. In a washout study in which 1274 DPP participants who did not progress to diabetes underwent an OGTT after 1 to 2 weeks of discontinuing metformin or placebo, the incidence of diabetes was still reduced by 25% in the metformin group, after the washout period, compared to a 31% risk reduction in the primary DPP analysis, suggesting a partially sustained rather than temporary effect of metformin [41]. In the DPPOS long-term follow-up study, metformin (850 mg twice daily as tolerated) was continued in the group initially assigned to metformin in addition to lifestyle counseling [32]. Although the progression to diabetes was similar in all groups during the 5.7-year follow-up period, the cumulative incidence of diabetes at 10 years was still reduced in the metformin group by 18% when compared to control group. Furthermore, the weight loss associated with metformin was also interestingly sustained at 10 years. A meta-analysis echoed this beneficial effect of metformin observed in the DPP trial, reporting a relative risk reduction of new-onset diabetes of 40% with the use of metformin [42].

In summary, metformin has been shown to be effective in preventing diabetes in patients at risk, especially persons with younger age, higher BMI, and history of gestational diabetes and in native Asian Indians. The protective effect of metformin seems to be sustained over the long term in follow-up studies.

Thiazolidinediones

Thiazolidinediones (TZDs) are antidiabetic agents that have been evaluated in diabetes prevention trials. TZDs are peroxisome proliferator-activated gamma receptor (PPAR-γ) agonists that work by augmenting conversion of preadipocytes to adipocytes, which in turn increase adiponectin levels, promoting insulin sensitivity [43]. In addition to their antihyperglycemic properties, TZDs are thought to have a direct protective effect on beta cells, potentially translating into prevention and delay of diabetes [44].

The first study to demonstrate diabetes prevention with a TZD was the TRIPOD study (Troglitazone in Prevention of Diabetes), in which 266 Hispanic women with a history of gestational diabetes were randomly assigned to troglitazone or placebo [45]. Troglitazone use was significantly associated with reduction of progression to diabetes at 1.5-year follow-up when compared to placebo (relative risk reduction of 55%), with a decrease of endogenous insulin requirement at 3 months of therapy and sustained benefit after discontinuation of the TZD, suggesting an effect on beta cell preservation.

Moreover, troglitazone was an investigational drug in the DPP trial from 1996 to 1998, at which time it was discontinued because of associated fatal liver failure in a DPP participant. In the DPP trial, troglitazone was asso-ciated with a remarkable 75% decrease in progression to diabetes at 1 year. Troglitazone was withdrawn from the US market in 2000 because of its association with severe hepatotoxicity.

The international DREAM (Diabetes REduction Assessment with ramipril and rosiglitazone Medications) trial randomly assigned more than 5000 participants with IFG and/or IGT to rosiglitazone, ramipril, or placebo in a 2 × 2 factorial design [46]. In participants receiving rosiglitazone, the risk for progression to diabetes was reduced by 60% and the likelihood of regression to normoglycemia was increased by 71% when compared to placebo. However, the use of rosiglitazone was associated with an increased risk of new-onset congestive heart failure and a mean weight gain of 2.2 kg, thought to reflect increased subcutaneous gluteal fat deposition, with an observed decreased waist-to-hip ratio.

Interestingly, in a passive follow-up of the DREAM study conducted a median 1.6 years after the end of the trial and 4.3 years after randomization, participants treated with rosiglitazone had a 39% lower incidence of diabetes compared to placebo participants, and 17% more of them regressed from prediabetes to normoglycemia [47]. Nonetheless, there was no difference between the 2 groups when the analysis was restricted to the passive follow-up period, suggesting a time-limited exposure to rosiglitazone reduces the longer-term incidence of diabetes by likely delaying but not reversing the underlying disease process.

The third large trial assessing the efficacy of a TZD in preventing diabetes was the Actos Now for the prevention of diabetes (ACT NOW) trial, which was a randomized, double-blinded study that assigned 602 patients with IGT to pioglitazone 45 mg daily or placebo [48]. Over a median follow-up of 2.6 years, pioglitazone was associated with a 72% lower annual rate of progression to diabetes (2.1% compared to 7.6 % in placebo group), and a higher rate of conversion to normal glucose tolerance (48%). In addition, pioglitazone had favorable effects on fasting and 2-hour blood glucose, A1C level, diastolic blood pressure, carotid intima thickness, and HDL cholesterol. As in the DREAM trial, an increased incidence of edema and weight gain was observed with pioglitazone.

Unlike the strong evidence supporting TZDs as an approach to diabetes prevention in the US trials, the Indian Diabetes Prevention Program-2 (IDPP-2) trial, which randomized 497 participants with IGT to lifestyle modifications with pioglitazone versus lifestyle modifications with placebo, did not demonstrate a significant reduction in diabetes at 3 years’ follow-up, suggesting a possible ethnicity-related variation in the effect of the medication [49]. In 2011, the French and German medications regulatory agency withdrew pioglitazone from the market because of a potential increase in incidence of bladder cancer with the cumulative use of more than 28 g of pioglitazone. In the United States, the Food and Drug Administration is performing an extensive review of data and advises against the use of pioglitazone in patients with a history of bladder cancer.

In summary, TZDs demonstrated significant efficacy in preventing diabetes in many patients at risk, but their safety concerns, particularly the associated new onset of congestive heart failure and potential increased risk of bladder cancer, might outweigh this benefit.

Combination Metformin and Thiazolidinediones

As metformin and rosiglitazone both have preventive benefits in diabetes, and rosiglitazone is associated with numerous side effects at a higher dose, a combination of metformin and low-dose rosiglitazone was evaluated in in the CAnadian Normoglycemia Outcomes Evaluation (CANOE) trial [50]. A total of 207 patients with IGT were randomly assigned to receive combination metformin (500 mg twice daily) and rosiglitazone (2 mg daily) versus placebo for a median of 3.9 years. The combination therapy was associated with a 66% relative risk reduction of progression to diabetes.

Alpha-glucosidase Inhibitors

Alpha-glucosidase inhibitors are antidiabetic agents that slow oral carbohydrate intestinal absorption, subsequently improving postprandial hyperglycemia, which can eventually reduce glucose toxicity of pancreatic beta cells. In addition, they have been shown to improve insulin sensitivity in individuals with IGT [51] and have been found to exert a favorable protective effect in a prediabetic population [52]. In a multicenter placebo-controlled randomized trial, the Study to Prevent Non-Insulin Dependent Diabetes Mellitus (STOP-NIDDM), 1429 participants with IGT were randomly assigned to receive acarbose 100 mg 3 times a day or placebo for 3 years [53]. As expected, diabetes incidence was significantly decreased by 25% in the acarbose group (relative risk of 32.4% vs 41.5% in acarbose and placebo group, respectively), and acarbose significantly increased reversion to normal glucose tolerance (P < 0.0001). Furthermore, the use of acarbose was associated with a statistically significant 49% decrease in the rate of any cardiovascular event, highlighting the cardiovascular protective effect of improving postprandial hyperglycemia with acarbose. This study had many limitations: a high percentage of participants discontinued treatment (31% in the acarbose group and 19% in the placebo group), most likely related to increased gastrointestinal adverse effects of acarbose. In addition, the diabetes prevention effect does not seem to be sustained: during a 3-month wash-out period where all patients received placebo, incidence of diabetes in the initial intervention group was higher than in the initial placebo group.

In a Japanese multicenter randomized double-blind trial, 1780 patients with IGT were randomly assigned to receive the alpha-glucosidase inhibitor voglibose or placebo [54]. An interim analysis at 48 weeks revealed a significantly lower risk of progression to diabetes in the voglibose group.

Combination Metformin and Acarbose

In a 6-year multicenter British study, the Early Diabetes Intervention Trial (EDIT), 631 participants with IFG were randomly assigned, in a factorial design, to double-blind treatment with acarbose or placebo and simultaneously to metformin or placebo [55]. At 3 years, there was a nonsignificant risk reduction of 8% and 37% in progression to 2 successive fasting plasma glucose values of 140 mg/dL or more in the acarbose and metformin groups, respectively, but a significantly lower 2-hour OGTT glucose in the acarbose group and significantly lower FBG in the metformin group. Interestingly, at 6 years of follow-up, there was no significant difference in relative risk of progression to diabetes with acarbose, metformin, or combination therapy [56]. However, unlike metformin or combination therapy, acarbose was associated with a significant relative risk reduction of diabetes (0.66, P = 0.046) in the subgroup of patients with IGT at baseline, suggesting a possible differential protective effect of certain agents in patients with IGT or IFG.

Nateglinide

Nateglinide is a short-acting insulin secretagogue that is mostly used in the treatment of postprandial hyperglycemia in diabetic patients. The protective effect of nateglinide in a prediabetic population was examined in the NAVIGATOR study (the NAteglinide and Valsartan in Impaired Glucose Tolerance Outcomes Research), a large prospective multinational, randomized, double-blind, placebo-controlled trial. Nateglinide (30–60 mg 3 times daily) and valsartan (80–160 mg daily) versus placebo were used in a 2×2 factorial design in 9306 participants with IGT and increased risk of cardiovascular events [57]. At 5 years, nateglinide did not reduce the cumulative incidence of diabetes or cardiovascular outcomes, when compared to placebo, whereas risk of hypoglycemia was significantly increased in the intervention group.

Liraglutide

Liraglutide is an injectable glucagon-like peptide-1 (GLP-1) receptor agonist used to treat T2DM, and recently approved as a weight-reducing agent at the dose of 3 mg injected subcutaneously. GLP-1 receptor agonists work by stimulating insulin secretion in a glucose-dependent manner, suppressing glucagon secretion, inducing satiety, and slowing gastric emptying. In the international double-blind SCALE (Satiety and Clinical Adiposity-Liraglutide Evidence) trial, 3731 nondiabetic patients, among whom 61.2% had prediabetes, were randomly assigned to liraglutide 3 mg subcutaneous injection daily or placebo, in addition to diet and exercise [58]. Liraglutide was associated with lower glucose levels on OGTT and lower A1C values at the end of the study (56 weeks), with this decrease especially prominent in prediabetic patients. Significantly fewer participants in the liraglutide group (4/2219) compared to the placebo group (14/1225) developed diabetes at 56 weeks, nearly all of whom (except for 1 in the placebo group) had prediabetes at the beginning of the study. Of note, the liraglutide group had a mean 8.4-kg weight reduction by week 56, compared to 2.8 kg in the placebo group.

Insulin

Insulin has also been investigated as a possible diabetes prevention agent, given the assumed protective effect insulin could exert on beta cell reserve. In the landmark international Outcome Reduction with Initial Glargine Intervention (ORIGIN) trial, 12,537 participants (mean age 63.5 years) with cardiovascular risk factors plus IFG, IGT, or type 2 diabetes were randomly assigned to receive insulin glargine (with a target FBG ≤ 95 mg/dL) or standard care and were monitored for cardiovascular outcomes and other secondary endpoints including incidence of diabetes [59]. After a median follow-up of 6.2 years, and 3 months after discontinuation of therapy, among the 1456 participants without baseline diabetes, new diabetes was diagnosed in 30% of participants receiving glargine versus 35% of those receiving standard therapy. However, rates of severe hypoglycemia and modest weight gain were higher in the insulin group, calling in to question the benefit/risk balance with the use of basal insulin for diabetes prevention.

ACE Inhibitors and ARBs

A possible diabetes preventive effect was observed with renin-angiotensin system (RAS) blockade agents in secondary analysis of several hypertension trials, such as with ramipril in the Heart Outcomes Prevention Evaluation study, captopril (compared to diuretics and beta blockers) in the CAptopril Prevention Project, lisinopril (compared to amlodipine and chlorthalidone) in the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial, losartan (compared to atenolol) in the Losartan Intervention For Endpoint reduction in hypertension study), and multiple other randomized controlled trials [60–64]. Therefore, 2 major trials were designed to examine, as a primary outcome, the effect of RAS inhibition on diabetes prevention in a population at risk. The DREAM trial randomly assigned, in a 2 × 2 factorial design, 5269 relatively healthy participants with IGT and/or IFG to rosiglitazone, ramipril, or placebo [65]. Although the use of ramipril at a dose of 15 mg daily for 3.5 years did not prevent diabetes significantly, it was associated with a 9%, nonsignificant decrease in new-onset of diabetes and a 16%, significant increase in regression of IFG and IGT to normoglycemia, as well as a significant decrease in OGTT 2-hour glucose level (135.1 vs 140.5 mg/dL) with no improvement in FBG.

Similarly, in the NAVIGATOR trial that examined the effect of nateglinide and valsartan on the prevention of diabetes in 9306 participants with IGT and increased risk of cardiovascular events, valsartan significantly but slightly reduced the incidence of diabetes at 5 years, by 14%, when compared to placebo (33% versus 37%, respectively), with no significant reduction in cardiovascular outcome [66]. Unlike in the DREAM study, the patients enrolled in the NAVIGATOR trial had established cardiovascular disease or cardiovascular risk factors and assumable elevated RAS activation level. This baseline population difference might explain the more significant effect of RAS inhibition in the NAVIGATOR trial.

Given the positive glycemic effect of ACE inhibitors and ARBs, their use should be encouraged in prediabetic patients when indicated for treatment of high blood pressure or cardiovascular disease. Different mechanisms could explain this favorable glycemic impact: inhibition of the post-receptor insulin signaling abnormalities, increased blood flow to the skeletal muscle facilitating insulin action, enhanced differentiation of pre-adipocytes into mature adipocytes, and increased pancreatic islet blood perfusion leading to appropriate insulin release and possible partial PPAR-γ activity [67].

Xenical

Xenical is a gastrointestinal lipase inhibitor approved for use for weight reduction and maintenance. A possible diabetes prevention benefit of xenical was initially suggested by a retrospective analysis of xenical treatment effects on obese patients with IGT [68]. This finding was subsequently confirmed by a multicenter randomized placebo-controlled study, XENical in the prevention of Diabetes in Obese Subjects (XENDOS), where 3305 obese subjects, with normal glucose tolerance or IGT were randomly assigned to either xenical 120 mg 3 times a day or placebo, in addition to lifestyle changes for all participants [69]. In the group of patients with IGT (694 subjects), xenical treatment was associated with a 45% risk reduction of progression to diabetes at 4 years (18.8% versus 28.8% in placebo), whereas participants with baseline normal glucose tolerance had no significant change in incidence of diabetes. On the other hand, weight reduction at 4 years was significantly greater in all patients who received xenical (5.8 kg in intervention group vs 3 kg in control group). The beneficial effect of xenical in diabetes prevention seems to be additive to the benefit of weight loss. As in many weight reduction trials, this study was limited by the high discontinuation rate in both groups (48% in xenical group and 66% in control group), probably related to insufficient clinical response.

Fibric Acid Derivatives (Bezafibrate)

Bezafibrate, a nonselective ligand/activator for PPAR-α, was found to reduce not only triglycerides, but also FPG, fructosamine, and A1C levels significantly in T2DM patients with hyperlipidemia [70]. Different mechanisms of glucose lowering have been suggested with bezafibrate: nonselective activation of PPAR-γ, improving insulin sensitivity, and enhancing glucose disposal in adipose tissue and skeletal muscles [71]. Furthermore, bezafibrate treatment was associated with decreased incidence of diabetes in patients with IFG and in obese non-diabetic patients with normal glycemic levels [72,73]. In a posthoc analysis of the Bezafibrate Infarction Prevention study, 303 patients with IFG received either 400 mg of bezafibrate daily or placebo [73]. Over a mean follow-up of 6.2 years, development of diabetes was less prevalent (54.4% vs 42.3%, relative risk reduction of 22%) and delayed (mean 10 months) in the bezafibrate group compared to placebo. Multivariate analysis identified bezafibrate as an independent predictor of decreased risk of new diabetes development, regardless of BMI and lipid profile.

Surgery

Over the past decade, bariatric surgery has become one of the most effective interventions for inducing and sustaining weight reduction in severely obese patients, leading to a significant benefit in diabetes prevention or remission. The Swedish Obese Subject Study is a large ongoing prospective nonrandomized cohort study that between 1987 and 2001 enrolled 4047 nondiabetic obese participants who underwent gastric surgery or were matched obese control, with diabetes incidence measured at 2, 10 and 15 years [74–76]. At 15 years, analysis of the available cohort of the initial group showed that T2DM developed in 392 of 1658 control participants and in 110 of 1771 bariatric-surgery participants, corresponding to incidence rates of 28.4 and 6.8 cases per 1000 person-years, respectively (P < 0.001). The treatment effects on the incidence of T2DM were at least as strong after 2 years and 10 years of follow-up as after 15 years. This effect was most prominent among the 591 patients who had IFG at baseline, with a number needed to treat as low as 1.3. The surgery group maintained an average 20-kg weight loss at 15 years.

In another study of the effects of bariatric surgery, 150 of 152 obese participants with IGT who underwent gastric bypass achieved and maintained a normal glycemic profile at 14 years of follow-up [77]. Similarly, in a follow-up of 136 obese participants with IGT, 109 of whom underwent bariatric surgery, 1 participant in the surgical group developed diabetes, as compared with 6 out of 27 in the control group [78]. In a meta-analysis including studies involving 22,094 patients who underwent bariatric surgery, 76.8% had complete resolution of their diabetes [79]. The rapid improvement of glycemic profile after bariatric surgery is thought to be due to oral intake restriction as well as acute hormonal changes related to the exclusion of the upper gastrointestinal tract (eg, incretin and ghrelin levels variations) [80].

Conclusions and Recommendations

The natural history of T2DM allows identification of patients at risk for diabetes and implementation of prevention strategies, which seems to be a public health need given the alarming increase in diabetes incidence. Indeed, the onset of T2DM is typically preceded by many years of beta cell dysfunction translating into carbohydrate metabolism abnormalities such as IFG and IGT, providing an excellent window of opportunity to identify persons at risk and prevent progression to diabetes. Numerous randomized controlled trials established lifestyle modifications, including dietary changes, moderate weight loss, and moderate intensity physical activity, as safe and effective interventions to prevent diabetes. This protective effect has been consistently shown to be sustained for more than 10 years after the initial intervention. Pharmacologic agents such as metformin, thiazolidinediones, alpha-glucosidase inhibitors, xenical, liraglutide, and insulin have also been associated with diabetes prevention in patients at risk. However, except for metformin, safety concerns or lack of durable efficacy or tolerability seem to outweigh their potential diabetes prevention benefit.

Given their favorable glycemic effect, RAS blockade and fibrates should be considered, when indicated, as reasonable treatment options for hypertension and hyperlipidemia in prediabetic patients. Bariatric surgery has been associated with a dramatic reduction in diabetes incidence in obese prediabetic patients and can be considered an alternative prevention measure in patients with severe obesity and prediabetes.

The recently updated ADA guidelines recommend referring patients with prediabetes to an intensive diet and physical activity behavioral counseling program; diet and activity goals should adhere to the tenets of the DPP, with a loss of 7% of body weight and at least 150 minutes of moderate physical activity (eg, brisk walking) per week [8]. Metformin therapy for diabetes prevention should be considered in patients with prediabetes, especially in those with BMI greater than 35 kg/m², those younger than 60 years of age, women with history of gestational diabetes, and/or those with rapidly rising A1C despite lifestyle modifications. Monitoring for development of diabetes, at least annually, and screening for and treatment of modifiable cardiovascular risk factors are suggested in patients with prediabetes [8].

Many lessons have been learned through the studies of diabetes prevention interventions. The challenge that remains is how to apply these interventions, especially the lifestyle modifications, in real world medical practice, at both the individual and public health level.

Corresponding author: Jocelyne Karam, MD, 4802 10th Avenue, Brooklyn, NY 11219, [email protected].

Financial disclosures: None reported.

From the Maimonides Medical Center (Dr. Karam) and the SUNY Downstate Medical Center (Dr. Karam and Dr. McFarlane), Brooklyn, NY.

Abstract

• Objective. To discuss the epidemic of diabetes highlighting the natural history of the disease and the major clinical trials aimed at diabetes prevention in different prediabetic populations around the world.
• Results. Diabetes prevention studies have evaluated various interventions including lifestyle modifications, metformin, alpha-glucosidase inhibitors, thiazolidinediones, nateglinide, and xenical as well as the renin-angiotensin aldosterone system (RAS) inhibitors. Lifestyle modifications seem to be the safest, most effective, and most sustainable intervention to prevent diabetes. Except for metformin, the potential diabetes prevention benefits of the studied pharmacologic agents are limited by safety concerns or lack of durable efficacy or tolerability. RAS blockade and fibrates have a favorable glycemic effect, and, when indicated, are reasonable treatment options for hypertension and hyperlipidemia in prediabetic patients.
• Conclusion. As recommended by American Diabetes Association guidelines, patients with prediabetes should be referred to an intensive diet and physical activity behavioral counseling program; diet and activity goals include a loss of 7% of body weight and at least 150 minutes of moderate physical activity per week. Metformin therapy for diabetes prevention should be considered as well.

Key words: prediabetes; type 2 diabetes mellitus, diabetes prevention, lifestyle modifications.

Diabetes mellitus has reached pandemic proportions across the globe. The International Diabetes Federation (IDF) estimates that in 2015 around 415 million people, or 1 in 11 adults, had diabetes, compared to 285 million in 2010, with 5 million deaths, or 1 death every 6 seconds, occurring because of diabetes or diabetes complications [1]. In the United States, an estimated 29.1 million Americans, or 9.3% of the population, have diabetes, 27.8% of them undiagnosed [2]. The prevalence of diabetes increases significantly with age, affecting around 16.2% of American adults aged 45 to 64 years and 25.9% of adults aged 65 years or older [2]. The Centers for Disease Control and Prevention (CDC) estimates that, with current trends, as many as 1 in 3 American adults could have diabetes by 2050 [3].

Type 2 diabetes mellitus (T2DM) accounts for the majority of prevalent and newly diagnosed diabetes in the world, and is strongly linked to overweight and inactivity in adults [4]. T2DM is increasingly being diagnosed in pediatric patients, in whom type 1 diabetes has historically been predominant; it now accounts for approximately 30% of newly diagnosed diabetes in children aged 10 to 19 years, exceeding 50% in certain ethnicities such as non-Hispanic black and American Indian/Alaska Native children [2].

These alarming trends have spurred significant research and public efforts aimed at reducing the prevalence of diabetes by preventing T2DM. Indeed, insulin resistance and abnormal carbohydrate metabolism progress over many years prior to the diagnosis of diabetes and manifest with different clinical and biochemical features. Both the pathophysiology and the natural history of T2DM offer clinicians an opportunity to identify patients at risk for developing the disease and to implement prevention strategies. This article outlines the risk factors and diagnostic criteria for prediabetes, describes the studies that have explored diabetes prevention through lifestyle changes, pharmacotherapy, or surgery, and reviews recommendations for managing patients at risk.

Risk Factors and Screening for T2DM

The American Diabetes Association (ADA) recommends screening all adults for prediabetes by assessing for diabetes risk factors [8]. Glucose testing is recommended in individuals aged 45 years or older, and should be considered in adults of any age who are overweight or obese (body mass index [BMI] ≥ 25 kg/m² or ≥ 23 kg/m² in Asian Americans) and have 1 or more additional risk factors for diabetes. Testing also should be considered in children and adolescents who are overweight or obese and who have 2 or more additional risk factors. If tests are normal, repeat testing carried out at a minimum of 3-year intervals is suggested [8].

Prediabetes

Abnormalities in glucose metabolism progress along a continuum through various stages before T2DM develops. Years before the development of overt diabetes, and especially in the presence of excessive visceral fat, cellular sensitivity to insulin gradually decreases, leading to a compensatory increased insulin secretion [9]. With time, and under continuous increased demand, pancreatic beta cell function declines and ultimately fails to overcome insulin resistance and maintain a normal glucose metabolism, resulting in prediabetes followed by the development of diabetes. This early beta cell dysfunction was illustrated by the decreased beta cell volume observed on autopsy of obese patients with IFG or T2DM, when compared to obese individuals with normal glucose tolerance [10]. It is estimated that around 40% to 70% of beta cell function is already lost by the time diabetes is clinically diagnosed. This relatively slow pathophysiologic process allows the identification of at-risk patients well before their blood glucose levels reach the diabetic diagnostic thresholds, and therefore presents an opportunity for prevention.

Diagnostic Criteria

The ADA guidelines released in 2003 define prediabetes as IFG (fasting blood glucose [FBG] levels of 100–125 mg/dL), IGT (glucose levels of 140–199 mg/dL at 2 hours during an oral glucose tolerance test [OGTT] following an oral load of 75 g of dextrose), or both. Additionally, hemoglobin A1C (A1C) was introduced as a diagnostic tool for prediabetes in 2010, with values between 5.7% and 6.4% indicating prediabetes [8]. Most of these thresholds were chosen due to their association with increased rates of complications, notably retinopathy and cardiovascular disease.

A combined report from the World Health Organization (WHO) and the IDF published in 2006 defined intermediate hyperglycemia as IFG, but with a higher cutoff for FBG (110–125 mg/dL) than the ADA’s definition, and/or IGT (2-hour OGTT glucose level of 140–199 mg/dL) [11]. The rationale for a higher cut-point for IFG is the concern about the increased prevalence of IFG and its impact on individuals and health systems and the more favorable cardiovascular risk profile and decreased risk of progression to diabetes in the group of patients with FBG of 100 to 110 mg/dL when compared to the group with FBG of 110 to 125 mg/dL. The report does not recommend the use of A1C in the diagnosis of diabetes or intermediate hyperglycemia because of a lack of global consistency and the potential for other factors that can be prevalent in some developing countries, such as hemoglobinopathies and anemia, to interfere with the assay.

Prevalence and Progression to Diabetes

According to CDC data from 2014, up to 86 million American adults, more than 1 in 3, have prediabetes, and 9 out of 10 of these individuals are undiagnosed [2]. It is estimated that approximately 25% of people diagnosed with either IFG or IGT progress to diabetes mellitus over a 3- to 5-year period [12]. If observed for longer periods, most prediabetic persons will probably develop diabetes. The highest rate of progression to diabetes is observed in patients with both IFG and IGT, older age, overweight, or other diabetic risk factors.

Complications

In addition to increasing the risk for progression to diabetes, prediabetes is independently associated with microvascular and macrovascular complications and increased risk of death, prior to the actual onset of diabetes. The DECODE study demonstrated significantly increased mortality in 2766 individuals with IGT after 7 years of follow-up, when compared to normoglycemic patients; this effect was more prominent in participants with IGT than in participants with IFG [13]. In the Australian Diabetes, Obesity and Lifestyle Study, IFG was found to be an independent predictor for cardiovascular mortality after adjustment for age, sex, and other traditional cardiovascular risk factors [14].

Similarly, a recent meta-analysis demonstrated that the presence of IFG was significantly associated with future risk for coronary heart disease (CHD), with the risk increase starting when fasting plasma glucose was as low as 100 mg/dL; however, this finding may have been confounded by the presence of undetected IGT or other cardiovascular risk factors [15]. Another recent systematic review of 53 prospective cohort studies with 1,611,339 participants showed that prediabetes (IFG or IGT) was associated with an increased risk of composite cardiovascular disease, CHD, stroke, and all-cause mortality [16].

The association between retinopathy and prediabetes has been described in multiple reports and this association has helped guide authors on selected thresholds for diagnosis of prediabetes. For example, in 1 study, the incidence of retinopathy in individuals with IGT was 12% among Pima Indians [17]. Similarly, in a follow-up study of the Diabetes Prevention Program, 8% of prediabetic participants who remained nondiabetics had evidence of retinopathy [18].

Neuropathy also has been observed in prediabetes. A noninvasive neurologic evaluation of individuals with IGT revealed subclinical neural dysfunction suggestive of cardiovascular autonomic neuropathy [19]. At the clinical level, a study that evaluated 100 patients with chronic idiopathic axonal neuropathy of unknown etiology found IFG in 36 and IGT in 38 patients, underscoring the role of abnormal glucose metabolism in these patients [20].

Nephropathy may also be more prevalent in those with prediabetes. In a 1999–2006 National Health and Nutrition Examination Survey analysis, the adjusted prevalence of chronic kidney disease, defined by estimated glomerular filtration rate (eGFR) of 15 to 59 mL/min per 1.73 m² or albumin-creatinine ratio ≥ 30 mg/g, was 17.1% in individuals with IFG, compared to 11.8% in individuals with normal fasting glucose [21].

Lifestyle Modifications

The alarming rapid increase in the prevalence of T2DM has been linked to a parallel rising epidemic of overweight, obesity, and lack of physical activity. Therefore, lifestyle changes aiming at weight reduction seemed to be a natural individual and public health strategy to prevent diabetes, and such strategies have been the focus of many randomized controlled trials around the world. As anticipated, weight loss, exercise, and diet have all been shown, separately or in combination, to be effective in decreasing the incidence of T2DM in high-risk patients [22–27]. Furthermore, and well beyond the benefit observed during the trials, follow-up studies revealed a sustained reduction of diabetes incidence in intervention groups several years after cessation of the intervention [28–32] (Table 2).

The Da Quing Diabetes Prevention Study (DQDPS), published in 1997, is one of the earliest prospective diabetes prevention trials [22]. This 6-year study conducted in 33 clinics in China from 1986 through 1992 included 577 participants with IGT who were randomly assigned to 1 of 4 groups: (1) diet (high vegetables, low sugar/alcohol) only, (2) exercise, (3) diet plus exercise, and (4) standard of care. At 6 years, diabetes incidence was significantly reduced by 46% in the exercise group, 31% in the diet group, and 42% in the diet plus exercise group compared to standard care. In 2006, 14 years after the end of the trial and 20 years after the initial enrollment, the cumulative incidence of diabetes was significantly lower in the intervention group at 80%, compared to 93% in the control group, and the annual incidence of diabetes was 7% and 11%, respectively, with a 43% lower incidence of diabetes over the 20-year period in the combination lifestyle changes group [28]. The preventive benefit of lifestyle changes persisted 2 decades after the initial randomization despite the standardization of treatment for all groups over the 14 years following the study, suggesting a strong and longitudinal preventive effect of the initial lifestyle modifications. In a follow-up study of the DQDPS conducted in 2009, at 23 years of follow-up, the cumulative incidences of cardiovascular mortality and all-cause mortality were significantly lower in the intervention group (11.9% versus 19.6%, and 28.1% versus 38.4%, respectively), highlighting the long-term clinical benefits of lifestyle intervention in patients with IGT [29].

Similarly, the Finnish Diabetes Prevention Study (FDPS), published in 2001, enrolled 522 middle-aged overweight participants with IGT [23]. The participants randomly assigned to the intervention group received individualized counseling designed to reduce weight, decrease total intake of fat and saturated fat, increase intake of fiber, and increase physical activity. The control group received standard therapy. At 4 years of follow-up, the cumulative incidence of diabetes was 11% in the intervention group and 23% in the control group, with a statistically significant 58% reduction in risk for progression to diabetes. A follow-up of the FDPS was published in 2006 [31]. Participants who did not progress to diabetes in the initial 4-year study were further followed for a median of 3 years. Interestingly, lifestyle changes were maintained by the intervention group participants despite the cessation of the individual counseling, leading to a 36% relative reduction in diabetes incidence during the post-intervention follow-up period alone (4.6 vs 7.2 per 100 person-years, P = 0.041) and a 43% cumulative diabetes incidence reduction over the 7-year follow-up, demonstrating, one more time, the sustained efficacy of lifestyle changes.

In the United States, the Diabetes Prevention Program (DPP) trial is a landmark NIH-sponsored multicenter randomized controlled trial published in 2002, and one of the largest diabetes prevention studies with lifestyle changes to date [24]. A total of 3234 participants with prediabetes, defined as an IFG or IGT, were randomly assigned to an intensive lifestyle modification program, metformin 850 mg twice daily, or matching placebo. Lifestyle changes included a low-fat (< 25% of caloric intake), 1200- to 1800-calorie diet and exercise for 150 minutes a week, with a 7% body weight reduction goal and a very well structured curriculum and professional support group. The study was discontinued early (at 3 years) as the data demonstrated the superiority of lifestyle changes, with a 58% reduction in diabetes incidence in the lifestyle intervention group and a 31% reduction in the metformin group when compared to placebo (cumulative incidence of diabetes at 3 years of 28.9%, 21.7 %, and 14.4% in the placebo, metformin, and lifestyle intervention groups, respectively). Lifestyle changes were significantly more effective than metformin and were consistently effective in men and women across age, BMI, and ethnic groups.

The DPPOS (DPP Outcome Study) was a 10-year follow-up of the DPP study published in 2009 where all participants were offered group-implemented lifestyle changes and were followed for an additional 5.7 years [32]. Unlike the Finnish follow-up study, diabetes incidence was similar in the 3 treatment groups in the follow-up period. However, the cumulative incidence of diabetes remained significantly the lowest in the original lifestyle group, with a 34% cumulative risk reduction in the lifestyle group and an 18% reduction in the metformin group at 10 years when compared to placebo. Interestingly, unlike most other studies of weight-reducing interventions, in the DPPOS, patients in the lifestyle changes and metformin groups maintained weight loss at 10 years’ follow-up.

In Japan, a diabetes prevention study assigned 458 male participants with IGT to a standard intervention group or an intensive intervention group receiving detailed lifestyle modification counseling every 3 to 4 months during hospital visits [25]. The cumulative 4-year incidence of diabetes was 9.3% in the control group versus 3.0% in the intervention group, and the reduction in diabetes risk was 67.4% (P < 0.001), with body weight reductions of 0.39 kg and 2.18 kg, respectively (P < 0.001). Of note, participants with higher FBG at baseline developed diabetes at a higher rate than those with lower values. This study suggested that lifestyle change counseling conducted in an outpatient clinic setting can be very effective in preventing diabetes.

Indian adults are thought to be more insulin resistant at a younger age and at a lower BMI than Caucasians. To assess whether the DPP findings can be replicated in an Indian population, the Indian Diabetes Prevention Program (IDPP) trial randomized a total of 531 participants with IGT to 4 groups: control, lifestyle modification, metformin, and lifestyle modifications with metformin [26]. The 3-year cumulative incidences of diabetes were 55.0%, 39.3%, 40.5%, and 39.5%, respectively, showing again a significant relative reduction in progression to diabetes of 28.5% with lifestyle changes, 26.4% with metformin, and 28.2% with both lifestyle changes and metformin, as compared with the control group.

In a Japanese unmasked, multicenter, randomized controlled trial published in 2011, 641 overweight adults with IFG were randomized to a frequent intervention group, receiving individual counseling and support for lifestyle modifications 9 times over 36 months, or a control group, receiving counseling 4 times over the same period. The 3-year cumulative incidence of T2DM was significantly lower in the frequent intervention group than in the control group (12.2% vs 16.6%) [27]. Interestingly, in a posthoc subgroup analysis, the protective effect was more prominent in patients with underlying associated IGT or elevated A1C, but was not observed in patients with isolated IFG, suggesting a possible prognostic value of an additional A1C or oral glucose tolerance test in individuals with IFG.

Diet

The diet followed in the major diabetes prevention trials discussed above has typically been a weight-reducing diet with decreased fat intake (eg, DPP, Finnish trial) and increased fiber intake (eg, Da Quing, DPP, Finnish trials). However, there has been more emphasis recently on the importance of the quality rather than the quantity of fats in preventing diabetes. For example, in a Spanish study, a non–calorie-restricted traditional Mediterranean diet, enriched with high-fat foods of vegetable origin (olive oil, nuts) decreased the incidence of diabetes by 52% in individuals at high cardiovascular risk after a median follow-up of 4.0 years, and in the absence of significant changes in body weight or physical activity among the groups [33]. These findings were reproduced by other studies. A recent meta-analysis examining the relation between intake of fruits and vegetables and the incidence of diabetes revealed that higher intake of fruit, especially berries, and green, leafy vegetables, yellow vegetables, cruciferous vegetables, or their fiber is associated with a lower risk of T2DM [34].

Exercise

Exercise is thought to improve insulin sensitivity and promote peripheral glucose uptake in normal individuals. Long-term moderate exercise, similar to the exercise recommended in DPP and FDPS, results in increased translocation of insulin-responsive glucose transporter (GLUT-4) from intracellular stores to the cell surface, facilitating glucose uptake [35]. A systematic review of 10 prospective cohort studies published in 2007 showed that, after adjustment for BMI, moderate-intensity physical activity was significantly associated with reduced diabetes incidence [36]. In the FDPS, participants who achieved at least 4 hours of exercise per week had a significant 80% decrease in incidence of diabetes, and this effect was observed even in the group that did not lose weight [23]. In the DQDPS, the greatest reduction in diabetes incidence was observed in the exercise group [22].

In a recent NIH-funded trial designed to examine the relative contribution of exercise alone to the overall beneficial effect of lifestyle changes in the DPP study, a total of 237 adults with IFG were randomly assigned to 4 different groups: low-amount moderate intensity exercise (similar to exercise followed in DPP), high-amount moderate intensity exercise, high-amount vigorous intensity exercise, and a combination of diet, weight loss, and low-amount moderate exercise. Only the diet and exercise group experienced a decrease in fasting glucose, whereas similar improvements in glucose tolerance were observed in both the diet and exercise group and the high-amount moderate-intensity exercise group, suggesting that such an exercise regimen may be as effective as a more intensive multicomponent approach involving diet, exercise, and weight loss for preventing diabetes [37].

Weight Loss

Weight reduction in prediabetic individuals has been consistently associated with reduced incidence of diabetes. Furthermore, the amount of weight loss needed to achieve this benefit seems to be relatively modest and a realistic goal to set for patients. Indeed, in the DPP trial, an average weight loss of only 5.6 kg was associated with a 58% lower incidence of diabetes [24]. Moreover, on further analysis of the DPP trial, and among weight, diet, and exercise, diabetes prevention correlated most strongly with weight loss, with an estimated 16% diabetes risk reduction for every single kilogram of weight reduction [38]. Similarly, within the same lifestyle intervention group in the FDPS, the participants who were able to achieve an initial body weight loss greater than 5% at 1 year had a nearly 70% relative risk reduction in progression to diabetes, when compared to their peers in the intervention group who had less or no weight loss [23].

In summary, numerous randomized controlled studies from various populations have proved that lifestyle modifications, including healthy diet, moderate weight loss, and moderate-intensity exercise, represent a very effective strategy to prevent diabetes in patients at risk, mostly patients with IGT, and this protective effect seems to be sustained over time.

Pharmacologic Interventions

Metformin

Metformin is an antidiabetic agent that works mostly at the liver site by suppressing hepatic glucose production and inhibiting production and oxidation of free fatty acids (FFA), thereby reducing FFA-induced insulin resistance and promoting peripheral glucose uptake [39]. This effect has the potential of preserving beta cell function by reducing the demand for insulin secretion.

In the DPP trial, metformin, although generally less effective than lifestyle changes, was associated with a significant 31% reduction in diabetes incidence (cumulative incidence of 22% in metformin group vs 29% in placebo group) and significant weight reduction (average of 2 kg) [24]. Further analysis of the DPP results showed that metformin efficacy, compared to placebo, was greater in patients who were younger, had higher BMI, and had higher FBG levels. In addition, a DPP substudy of 350 women with history of gestational diabetes and IGT revealed that this group of women, who had a higher risk of progression to diabetes (71% at 3 years) when compared to women with no history of gestational diabetes, despite similar baseline glucose levels, had similar diabetes risk reduction of 50% with both metformin and lifestyle changes [40].

In the IDPP study, both lifestyle changes and metformin reduced significantly and similarly the incidence of diabetes in adults with IGT, with no observed added benefit from combining both interventions [26]. It has not been clear, however, how much of this effect of metformin is a result of pharmacologic properties masking hyperglycemia or a true protective and preventive effect. In a washout study in which 1274 DPP participants who did not progress to diabetes underwent an OGTT after 1 to 2 weeks of discontinuing metformin or placebo, the incidence of diabetes was still reduced by 25% in the metformin group, after the washout period, compared to a 31% risk reduction in the primary DPP analysis, suggesting a partially sustained rather than temporary effect of metformin [41]. In the DPPOS long-term follow-up study, metformin (850 mg twice daily as tolerated) was continued in the group initially assigned to metformin in addition to lifestyle counseling [32]. Although the progression to diabetes was similar in all groups during the 5.7-year follow-up period, the cumulative incidence of diabetes at 10 years was still reduced in the metformin group by 18% when compared to control group. Furthermore, the weight loss associated with metformin was also interestingly sustained at 10 years. A meta-analysis echoed this beneficial effect of metformin observed in the DPP trial, reporting a relative risk reduction of new-onset diabetes of 40% with the use of metformin [42].

In summary, metformin has been shown to be effective in preventing diabetes in patients at risk, especially persons with younger age, higher BMI, and history of gestational diabetes and in native Asian Indians. The protective effect of metformin seems to be sustained over the long term in follow-up studies.

Thiazolidinediones

Thiazolidinediones (TZDs) are antidiabetic agents that have been evaluated in diabetes prevention trials. TZDs are peroxisome proliferator-activated gamma receptor (PPAR-γ) agonists that work by augmenting conversion of preadipocytes to adipocytes, which in turn increase adiponectin levels, promoting insulin sensitivity [43]. In addition to their antihyperglycemic properties, TZDs are thought to have a direct protective effect on beta cells, potentially translating into prevention and delay of diabetes [44].

The first study to demonstrate diabetes prevention with a TZD was the TRIPOD study (Troglitazone in Prevention of Diabetes), in which 266 Hispanic women with a history of gestational diabetes were randomly assigned to troglitazone or placebo [45]. Troglitazone use was significantly associated with reduction of progression to diabetes at 1.5-year follow-up when compared to placebo (relative risk reduction of 55%), with a decrease of endogenous insulin requirement at 3 months of therapy and sustained benefit after discontinuation of the TZD, suggesting an effect on beta cell preservation.

Moreover, troglitazone was an investigational drug in the DPP trial from 1996 to 1998, at which time it was discontinued because of associated fatal liver failure in a DPP participant. In the DPP trial, troglitazone was asso-ciated with a remarkable 75% decrease in progression to diabetes at 1 year. Troglitazone was withdrawn from the US market in 2000 because of its association with severe hepatotoxicity.

The international DREAM (Diabetes REduction Assessment with ramipril and rosiglitazone Medications) trial randomly assigned more than 5000 participants with IFG and/or IGT to rosiglitazone, ramipril, or placebo in a 2 × 2 factorial design [46]. In participants receiving rosiglitazone, the risk for progression to diabetes was reduced by 60% and the likelihood of regression to normoglycemia was increased by 71% when compared to placebo. However, the use of rosiglitazone was associated with an increased risk of new-onset congestive heart failure and a mean weight gain of 2.2 kg, thought to reflect increased subcutaneous gluteal fat deposition, with an observed decreased waist-to-hip ratio.

Interestingly, in a passive follow-up of the DREAM study conducted a median 1.6 years after the end of the trial and 4.3 years after randomization, participants treated with rosiglitazone had a 39% lower incidence of diabetes compared to placebo participants, and 17% more of them regressed from prediabetes to normoglycemia [47]. Nonetheless, there was no difference between the 2 groups when the analysis was restricted to the passive follow-up period, suggesting a time-limited exposure to rosiglitazone reduces the longer-term incidence of diabetes by likely delaying but not reversing the underlying disease process.

The third large trial assessing the efficacy of a TZD in preventing diabetes was the Actos Now for the prevention of diabetes (ACT NOW) trial, which was a randomized, double-blinded study that assigned 602 patients with IGT to pioglitazone 45 mg daily or placebo [48]. Over a median follow-up of 2.6 years, pioglitazone was associated with a 72% lower annual rate of progression to diabetes (2.1% compared to 7.6 % in placebo group), and a higher rate of conversion to normal glucose tolerance (48%). In addition, pioglitazone had favorable effects on fasting and 2-hour blood glucose, A1C level, diastolic blood pressure, carotid intima thickness, and HDL cholesterol. As in the DREAM trial, an increased incidence of edema and weight gain was observed with pioglitazone.

Unlike the strong evidence supporting TZDs as an approach to diabetes prevention in the US trials, the Indian Diabetes Prevention Program-2 (IDPP-2) trial, which randomized 497 participants with IGT to lifestyle modifications with pioglitazone versus lifestyle modifications with placebo, did not demonstrate a significant reduction in diabetes at 3 years’ follow-up, suggesting a possible ethnicity-related variation in the effect of the medication [49]. In 2011, the French and German medications regulatory agency withdrew pioglitazone from the market because of a potential increase in incidence of bladder cancer with the cumulative use of more than 28 g of pioglitazone. In the United States, the Food and Drug Administration is performing an extensive review of data and advises against the use of pioglitazone in patients with a history of bladder cancer.

In summary, TZDs demonstrated significant efficacy in preventing diabetes in many patients at risk, but their safety concerns, particularly the associated new onset of congestive heart failure and potential increased risk of bladder cancer, might outweigh this benefit.

Combination Metformin and Thiazolidinediones

As metformin and rosiglitazone both have preventive benefits in diabetes, and rosiglitazone is associated with numerous side effects at a higher dose, a combination of metformin and low-dose rosiglitazone was evaluated in in the CAnadian Normoglycemia Outcomes Evaluation (CANOE) trial [50]. A total of 207 patients with IGT were randomly assigned to receive combination metformin (500 mg twice daily) and rosiglitazone (2 mg daily) versus placebo for a median of 3.9 years. The combination therapy was associated with a 66% relative risk reduction of progression to diabetes.

Alpha-glucosidase Inhibitors

Alpha-glucosidase inhibitors are antidiabetic agents that slow oral carbohydrate intestinal absorption, subsequently improving postprandial hyperglycemia, which can eventually reduce glucose toxicity of pancreatic beta cells. In addition, they have been shown to improve insulin sensitivity in individuals with IGT [51] and have been found to exert a favorable protective effect in a prediabetic population [52]. In a multicenter placebo-controlled randomized trial, the Study to Prevent Non-Insulin Dependent Diabetes Mellitus (STOP-NIDDM), 1429 participants with IGT were randomly assigned to receive acarbose 100 mg 3 times a day or placebo for 3 years [53]. As expected, diabetes incidence was significantly decreased by 25% in the acarbose group (relative risk of 32.4% vs 41.5% in acarbose and placebo group, respectively), and acarbose significantly increased reversion to normal glucose tolerance (P < 0.0001). Furthermore, the use of acarbose was associated with a statistically significant 49% decrease in the rate of any cardiovascular event, highlighting the cardiovascular protective effect of improving postprandial hyperglycemia with acarbose. This study had many limitations: a high percentage of participants discontinued treatment (31% in the acarbose group and 19% in the placebo group), most likely related to increased gastrointestinal adverse effects of acarbose. In addition, the diabetes prevention effect does not seem to be sustained: during a 3-month wash-out period where all patients received placebo, incidence of diabetes in the initial intervention group was higher than in the initial placebo group.

In a Japanese multicenter randomized double-blind trial, 1780 patients with IGT were randomly assigned to receive the alpha-glucosidase inhibitor voglibose or placebo [54]. An interim analysis at 48 weeks revealed a significantly lower risk of progression to diabetes in the voglibose group.

Combination Metformin and Acarbose

In a 6-year multicenter British study, the Early Diabetes Intervention Trial (EDIT), 631 participants with IFG were randomly assigned, in a factorial design, to double-blind treatment with acarbose or placebo and simultaneously to metformin or placebo [55]. At 3 years, there was a nonsignificant risk reduction of 8% and 37% in progression to 2 successive fasting plasma glucose values of 140 mg/dL or more in the acarbose and metformin groups, respectively, but a significantly lower 2-hour OGTT glucose in the acarbose group and significantly lower FBG in the metformin group. Interestingly, at 6 years of follow-up, there was no significant difference in relative risk of progression to diabetes with acarbose, metformin, or combination therapy [56]. However, unlike metformin or combination therapy, acarbose was associated with a significant relative risk reduction of diabetes (0.66, P = 0.046) in the subgroup of patients with IGT at baseline, suggesting a possible differential protective effect of certain agents in patients with IGT or IFG.

Nateglinide

Nateglinide is a short-acting insulin secretagogue that is mostly used in the treatment of postprandial hyperglycemia in diabetic patients. The protective effect of nateglinide in a prediabetic population was examined in the NAVIGATOR study (the NAteglinide and Valsartan in Impaired Glucose Tolerance Outcomes Research), a large prospective multinational, randomized, double-blind, placebo-controlled trial. Nateglinide (30–60 mg 3 times daily) and valsartan (80–160 mg daily) versus placebo were used in a 2×2 factorial design in 9306 participants with IGT and increased risk of cardiovascular events [57]. At 5 years, nateglinide did not reduce the cumulative incidence of diabetes or cardiovascular outcomes, when compared to placebo, whereas risk of hypoglycemia was significantly increased in the intervention group.

Liraglutide

Liraglutide is an injectable glucagon-like peptide-1 (GLP-1) receptor agonist used to treat T2DM, and recently approved as a weight-reducing agent at the dose of 3 mg injected subcutaneously. GLP-1 receptor agonists work by stimulating insulin secretion in a glucose-dependent manner, suppressing glucagon secretion, inducing satiety, and slowing gastric emptying. In the international double-blind SCALE (Satiety and Clinical Adiposity-Liraglutide Evidence) trial, 3731 nondiabetic patients, among whom 61.2% had prediabetes, were randomly assigned to liraglutide 3 mg subcutaneous injection daily or placebo, in addition to diet and exercise [58]. Liraglutide was associated with lower glucose levels on OGTT and lower A1C values at the end of the study (56 weeks), with this decrease especially prominent in prediabetic patients. Significantly fewer participants in the liraglutide group (4/2219) compared to the placebo group (14/1225) developed diabetes at 56 weeks, nearly all of whom (except for 1 in the placebo group) had prediabetes at the beginning of the study. Of note, the liraglutide group had a mean 8.4-kg weight reduction by week 56, compared to 2.8 kg in the placebo group.

Insulin

Insulin has also been investigated as a possible diabetes prevention agent, given the assumed protective effect insulin could exert on beta cell reserve. In the landmark international Outcome Reduction with Initial Glargine Intervention (ORIGIN) trial, 12,537 participants (mean age 63.5 years) with cardiovascular risk factors plus IFG, IGT, or type 2 diabetes were randomly assigned to receive insulin glargine (with a target FBG ≤ 95 mg/dL) or standard care and were monitored for cardiovascular outcomes and other secondary endpoints including incidence of diabetes [59]. After a median follow-up of 6.2 years, and 3 months after discontinuation of therapy, among the 1456 participants without baseline diabetes, new diabetes was diagnosed in 30% of participants receiving glargine versus 35% of those receiving standard therapy. However, rates of severe hypoglycemia and modest weight gain were higher in the insulin group, calling in to question the benefit/risk balance with the use of basal insulin for diabetes prevention.

ACE Inhibitors and ARBs

A possible diabetes preventive effect was observed with renin-angiotensin system (RAS) blockade agents in secondary analysis of several hypertension trials, such as with ramipril in the Heart Outcomes Prevention Evaluation study, captopril (compared to diuretics and beta blockers) in the CAptopril Prevention Project, lisinopril (compared to amlodipine and chlorthalidone) in the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial, losartan (compared to atenolol) in the Losartan Intervention For Endpoint reduction in hypertension study), and multiple other randomized controlled trials [60–64]. Therefore, 2 major trials were designed to examine, as a primary outcome, the effect of RAS inhibition on diabetes prevention in a population at risk. The DREAM trial randomly assigned, in a 2 × 2 factorial design, 5269 relatively healthy participants with IGT and/or IFG to rosiglitazone, ramipril, or placebo [65]. Although the use of ramipril at a dose of 15 mg daily for 3.5 years did not prevent diabetes significantly, it was associated with a 9%, nonsignificant decrease in new-onset of diabetes and a 16%, significant increase in regression of IFG and IGT to normoglycemia, as well as a significant decrease in OGTT 2-hour glucose level (135.1 vs 140.5 mg/dL) with no improvement in FBG.

Similarly, in the NAVIGATOR trial that examined the effect of nateglinide and valsartan on the prevention of diabetes in 9306 participants with IGT and increased risk of cardiovascular events, valsartan significantly but slightly reduced the incidence of diabetes at 5 years, by 14%, when compared to placebo (33% versus 37%, respectively), with no significant reduction in cardiovascular outcome [66]. Unlike in the DREAM study, the patients enrolled in the NAVIGATOR trial had established cardiovascular disease or cardiovascular risk factors and assumable elevated RAS activation level. This baseline population difference might explain the more significant effect of RAS inhibition in the NAVIGATOR trial.

Given the positive glycemic effect of ACE inhibitors and ARBs, their use should be encouraged in prediabetic patients when indicated for treatment of high blood pressure or cardiovascular disease. Different mechanisms could explain this favorable glycemic impact: inhibition of the post-receptor insulin signaling abnormalities, increased blood flow to the skeletal muscle facilitating insulin action, enhanced differentiation of pre-adipocytes into mature adipocytes, and increased pancreatic islet blood perfusion leading to appropriate insulin release and possible partial PPAR-γ activity [67].

Xenical

Xenical is a gastrointestinal lipase inhibitor approved for use for weight reduction and maintenance. A possible diabetes prevention benefit of xenical was initially suggested by a retrospective analysis of xenical treatment effects on obese patients with IGT [68]. This finding was subsequently confirmed by a multicenter randomized placebo-controlled study, XENical in the prevention of Diabetes in Obese Subjects (XENDOS), where 3305 obese subjects, with normal glucose tolerance or IGT were randomly assigned to either xenical 120 mg 3 times a day or placebo, in addition to lifestyle changes for all participants [69]. In the group of patients with IGT (694 subjects), xenical treatment was associated with a 45% risk reduction of progression to diabetes at 4 years (18.8% versus 28.8% in placebo), whereas participants with baseline normal glucose tolerance had no significant change in incidence of diabetes. On the other hand, weight reduction at 4 years was significantly greater in all patients who received xenical (5.8 kg in intervention group vs 3 kg in control group). The beneficial effect of xenical in diabetes prevention seems to be additive to the benefit of weight loss. As in many weight reduction trials, this study was limited by the high discontinuation rate in both groups (48% in xenical group and 66% in control group), probably related to insufficient clinical response.

Fibric Acid Derivatives (Bezafibrate)

Bezafibrate, a nonselective ligand/activator for PPAR-α, was found to reduce not only triglycerides, but also FPG, fructosamine, and A1C levels significantly in T2DM patients with hyperlipidemia [70]. Different mechanisms of glucose lowering have been suggested with bezafibrate: nonselective activation of PPAR-γ, improving insulin sensitivity, and enhancing glucose disposal in adipose tissue and skeletal muscles [71]. Furthermore, bezafibrate treatment was associated with decreased incidence of diabetes in patients with IFG and in obese non-diabetic patients with normal glycemic levels [72,73]. In a posthoc analysis of the Bezafibrate Infarction Prevention study, 303 patients with IFG received either 400 mg of bezafibrate daily or placebo [73]. Over a mean follow-up of 6.2 years, development of diabetes was less prevalent (54.4% vs 42.3%, relative risk reduction of 22%) and delayed (mean 10 months) in the bezafibrate group compared to placebo. Multivariate analysis identified bezafibrate as an independent predictor of decreased risk of new diabetes development, regardless of BMI and lipid profile.

Surgery

Over the past decade, bariatric surgery has become one of the most effective interventions for inducing and sustaining weight reduction in severely obese patients, leading to a significant benefit in diabetes prevention or remission. The Swedish Obese Subject Study is a large ongoing prospective nonrandomized cohort study that between 1987 and 2001 enrolled 4047 nondiabetic obese participants who underwent gastric surgery or were matched obese control, with diabetes incidence measured at 2, 10 and 15 years [74–76]. At 15 years, analysis of the available cohort of the initial group showed that T2DM developed in 392 of 1658 control participants and in 110 of 1771 bariatric-surgery participants, corresponding to incidence rates of 28.4 and 6.8 cases per 1000 person-years, respectively (P < 0.001). The treatment effects on the incidence of T2DM were at least as strong after 2 years and 10 years of follow-up as after 15 years. This effect was most prominent among the 591 patients who had IFG at baseline, with a number needed to treat as low as 1.3. The surgery group maintained an average 20-kg weight loss at 15 years.

In another study of the effects of bariatric surgery, 150 of 152 obese participants with IGT who underwent gastric bypass achieved and maintained a normal glycemic profile at 14 years of follow-up [77]. Similarly, in a follow-up of 136 obese participants with IGT, 109 of whom underwent bariatric surgery, 1 participant in the surgical group developed diabetes, as compared with 6 out of 27 in the control group [78]. In a meta-analysis including studies involving 22,094 patients who underwent bariatric surgery, 76.8% had complete resolution of their diabetes [79]. The rapid improvement of glycemic profile after bariatric surgery is thought to be due to oral intake restriction as well as acute hormonal changes related to the exclusion of the upper gastrointestinal tract (eg, incretin and ghrelin levels variations) [80].

Conclusions and Recommendations

The natural history of T2DM allows identification of patients at risk for diabetes and implementation of prevention strategies, which seems to be a public health need given the alarming increase in diabetes incidence. Indeed, the onset of T2DM is typically preceded by many years of beta cell dysfunction translating into carbohydrate metabolism abnormalities such as IFG and IGT, providing an excellent window of opportunity to identify persons at risk and prevent progression to diabetes. Numerous randomized controlled trials established lifestyle modifications, including dietary changes, moderate weight loss, and moderate intensity physical activity, as safe and effective interventions to prevent diabetes. This protective effect has been consistently shown to be sustained for more than 10 years after the initial intervention. Pharmacologic agents such as metformin, thiazolidinediones, alpha-glucosidase inhibitors, xenical, liraglutide, and insulin have also been associated with diabetes prevention in patients at risk. However, except for metformin, safety concerns or lack of durable efficacy or tolerability seem to outweigh their potential diabetes prevention benefit.

Given their favorable glycemic effect, RAS blockade and fibrates should be considered, when indicated, as reasonable treatment options for hypertension and hyperlipidemia in prediabetic patients. Bariatric surgery has been associated with a dramatic reduction in diabetes incidence in obese prediabetic patients and can be considered an alternative prevention measure in patients with severe obesity and prediabetes.

The recently updated ADA guidelines recommend referring patients with prediabetes to an intensive diet and physical activity behavioral counseling program; diet and activity goals should adhere to the tenets of the DPP, with a loss of 7% of body weight and at least 150 minutes of moderate physical activity (eg, brisk walking) per week [8]. Metformin therapy for diabetes prevention should be considered in patients with prediabetes, especially in those with BMI greater than 35 kg/m², those younger than 60 years of age, women with history of gestational diabetes, and/or those with rapidly rising A1C despite lifestyle modifications. Monitoring for development of diabetes, at least annually, and screening for and treatment of modifiable cardiovascular risk factors are suggested in patients with prediabetes [8].

Many lessons have been learned through the studies of diabetes prevention interventions. The challenge that remains is how to apply these interventions, especially the lifestyle modifications, in real world medical practice, at both the individual and public health level.

Corresponding author: Jocelyne Karam, MD, 4802 10th Avenue, Brooklyn, NY 11219, [email protected].

Financial disclosures: None reported.

Study Overview

Objective. To assess the risks associated with magnetic resonance imaging (MRI) in patients with a pacemaker or implantable cardioverter-defibrillator (ICD) that is “non–MRI-conditional.”

Design. Prospective cohort study using the multicenter MagnaSafe Registry.

Setting and participants. Patients were included in the registry if they were 18 years of age or older and had a non–MRI-conditional pacemaker or ICD generator, from any manufacturer, that was implanted after 2001, with leads from any manufacturer, and if the patient’s physician determined that nonthoracic MRI at 1.5 tesla was clinically indicated. Exclusion criteria included an abandoned or inactive lead that could not be interrogated, an MRI-conditional pacemaker, a device implanted in a nonthoracic location, or a device with a battery that was near the end of its battery life. In addition, pacing-dependent patients with an ICD were also excluded.

Main outcome measures. The primary outcomes of the study were death, generator or lead failure requiring immediate replacement, loss of capture (for pacing-dependent patients with pacemakers), new-onset arrhythmia, and partial or full generator electrical reset. The secondary outcomes were changes in device settings including: a battery voltage decrease of 0.04V or more, a pacing lead threshold increase of 0.5V or more, a P-wave amplitude decrease of 50% or more, an R-wave amplitude decrease of 25% or more and of 50% or more, a pacing lead impedance change of 50 ohms or more, and a high-voltage (shock) lead impedance change of 3 ohms or more.

Main results. Between April 2009 and April 2014, clinically indicated nonthoracic MRI was performed in a total of 1000 pacemaker cases (818 patients) and 500 ICD cases (428 patients) across 19 centers in the United States. The majority (75%) of the MRI examinations were performed on the brain or the spine. The mean time patients spent within the magnetic field was 44 minutes. Four patients reported symptoms of generator-site discomfort; one patient with an ICD was removed from the scanner when a sensation of heating was described at the site of the generator implanted and did not complete the examination.

Regarding primary outcomes, no deaths, lead failures, losses of capture, or ventricular arrhythmias occurred during MRI. One ICD device was left in the active mode for anti-tachycardia therapy (a protocol violation) and the generator could not be interrogated after MRI and required immediate replacement. Four patients had atrial fibrillation and 2 patients had atrial flutter during or immediately after the MRI. All 6 patients returned to sinus rhythm within 49 hours after MRI. No ventricular arrhythmias were noted. There were also 6 cases of partial generator electrical reset with no clinical significance.

Regarding secondary outcomes, a decrease of 50% or more in P-wave amplitude was detected in 0.9% of pacemaker leads and in 0.3% of ICD leads; a decrease of 25% or more in R-wave amplitude was detected in 3.9% of pacemaker leads and in 1.5% of ICD leads, and a decrease of 50% or more in R-wave amplitude was detected in no pacemaker leads and in 0.2% of ICD leads. An increase in pacing lead threshold of 0.5 V or more was detected in 0.7% of pacemaker leads and in 0.8% of ICD leads. A pacing lead impedance change of 50 ohms or more was noted in 3.3% of pacemakers and in 4.2% of ICDs.

Conclusion. Device or lead failure did not occur in any patient with a non–MRI-conditional pacemaker or ICD who underwent clinically indicated nonthoracic MRI at 1.5 tesla when patients were appropriately screened and had the cardiac device reprogrammed in accordance with the protocol. Substantial changes in device settings were infrequent and did not result in clinical adverse events.

Commentary

It is estimated that 2 million people in the United States and an additional 6 million worldwide have an implanted non–MRI-conditional cardiac pacemaker or ICD [1]. At least half of patients with such devices are predicted to have a clinical indication for MRI during their lifetime after device implantation [2]. The use of MRI poses concerns due to the potential for magnetic field–induced cardiac lead heating, which could result in myocardial thermal injury and detrimental changes in pacing properties [3,4].

In this study, Russo and colleagues assessed the risks for patients with a non-MRI-conditional pacemaker or ICD receiving an MRI scan using a pre-scanning protocol. If the patient was asymptomatic and had an intrinsic heart rate of at least 40 beats per minute, the device was programmed to a no-pacing mode (ODO or OVO). Symptomatic patients or those with an intrinsic heart rate of less than 40 beats per minute were determined to be pacing-dependent, and the device was reprogrammed to an asynchronous pacing mode (DOO or VOO). All bradycardia and tachycardia therapies were inactivated before the MRI. Based on this standardized protocol, no major adverse outcomes occurred. All pacemaker or ICD device were reprogrammed in accordance with the pre-specified protocol except one case where the ICD device was left in the active mode for anti-tachycardia therapy (a protocol violation) and the generator could not be interrogated after MRI and required immediate replacement. In addition to patient safety, the authors also measure the functionality of the devices pre-MRI and post-MRI. One of these measurements were battery voltage changes, a small decrease was noted for both pacemakers and ICDs as expected. The radiofrequency energy generated during MRI scanning creates a temporary decrease in battery voltage, which had resolved in all pacemaker cases although some ICD voltage decreases of 0.04 V or more had not resolved by the end of the 6 month post-MRI follow-up.

Several limitations exist. The study registry included devices and leads from different manufacturers, but did not report outcomes by manufacturer. While overall it appears to be safe to conduct an MRI study for patients who have non–MRI-conditional devices, this study did not provide enough information for patients younger than 18 years of age, patients who required repeat MRI studies, MRI examinations of the thorax, or higher MRI field strengths—the newer 3 tesla high-resolution MRI machines.

Applications for Clinical Practice

This multicenter prospective cohort study provides strong evidence that patients with a non–MRI-conditional pacemaker or defibrillator can receive nonthoracic MRI studies at 1.5 tesla when a straight pre-scanning device interrogation is performed per the standardized protocol.

—Ka Ming Gordon Ngai, MD, MPH

Study Overview

Objective. To assess the risks associated with magnetic resonance imaging (MRI) in patients with a pacemaker or implantable cardioverter-defibrillator (ICD) that is “non–MRI-conditional.”

Design. Prospective cohort study using the multicenter MagnaSafe Registry.

Setting and participants. Patients were included in the registry if they were 18 years of age or older and had a non–MRI-conditional pacemaker or ICD generator, from any manufacturer, that was implanted after 2001, with leads from any manufacturer, and if the patient’s physician determined that nonthoracic MRI at 1.5 tesla was clinically indicated. Exclusion criteria included an abandoned or inactive lead that could not be interrogated, an MRI-conditional pacemaker, a device implanted in a nonthoracic location, or a device with a battery that was near the end of its battery life. In addition, pacing-dependent patients with an ICD were also excluded.

Main outcome measures. The primary outcomes of the study were death, generator or lead failure requiring immediate replacement, loss of capture (for pacing-dependent patients with pacemakers), new-onset arrhythmia, and partial or full generator electrical reset. The secondary outcomes were changes in device settings including: a battery voltage decrease of 0.04V or more, a pacing lead threshold increase of 0.5V or more, a P-wave amplitude decrease of 50% or more, an R-wave amplitude decrease of 25% or more and of 50% or more, a pacing lead impedance change of 50 ohms or more, and a high-voltage (shock) lead impedance change of 3 ohms or more.

Main results. Between April 2009 and April 2014, clinically indicated nonthoracic MRI was performed in a total of 1000 pacemaker cases (818 patients) and 500 ICD cases (428 patients) across 19 centers in the United States. The majority (75%) of the MRI examinations were performed on the brain or the spine. The mean time patients spent within the magnetic field was 44 minutes. Four patients reported symptoms of generator-site discomfort; one patient with an ICD was removed from the scanner when a sensation of heating was described at the site of the generator implanted and did not complete the examination.

Regarding primary outcomes, no deaths, lead failures, losses of capture, or ventricular arrhythmias occurred during MRI. One ICD device was left in the active mode for anti-tachycardia therapy (a protocol violation) and the generator could not be interrogated after MRI and required immediate replacement. Four patients had atrial fibrillation and 2 patients had atrial flutter during or immediately after the MRI. All 6 patients returned to sinus rhythm within 49 hours after MRI. No ventricular arrhythmias were noted. There were also 6 cases of partial generator electrical reset with no clinical significance.

Regarding secondary outcomes, a decrease of 50% or more in P-wave amplitude was detected in 0.9% of pacemaker leads and in 0.3% of ICD leads; a decrease of 25% or more in R-wave amplitude was detected in 3.9% of pacemaker leads and in 1.5% of ICD leads, and a decrease of 50% or more in R-wave amplitude was detected in no pacemaker leads and in 0.2% of ICD leads. An increase in pacing lead threshold of 0.5 V or more was detected in 0.7% of pacemaker leads and in 0.8% of ICD leads. A pacing lead impedance change of 50 ohms or more was noted in 3.3% of pacemakers and in 4.2% of ICDs.

Conclusion. Device or lead failure did not occur in any patient with a non–MRI-conditional pacemaker or ICD who underwent clinically indicated nonthoracic MRI at 1.5 tesla when patients were appropriately screened and had the cardiac device reprogrammed in accordance with the protocol. Substantial changes in device settings were infrequent and did not result in clinical adverse events.

Commentary

It is estimated that 2 million people in the United States and an additional 6 million worldwide have an implanted non–MRI-conditional cardiac pacemaker or ICD [1]. At least half of patients with such devices are predicted to have a clinical indication for MRI during their lifetime after device implantation [2]. The use of MRI poses concerns due to the potential for magnetic field–induced cardiac lead heating, which could result in myocardial thermal injury and detrimental changes in pacing properties [3,4].

In this study, Russo and colleagues assessed the risks for patients with a non-MRI-conditional pacemaker or ICD receiving an MRI scan using a pre-scanning protocol. If the patient was asymptomatic and had an intrinsic heart rate of at least 40 beats per minute, the device was programmed to a no-pacing mode (ODO or OVO). Symptomatic patients or those with an intrinsic heart rate of less than 40 beats per minute were determined to be pacing-dependent, and the device was reprogrammed to an asynchronous pacing mode (DOO or VOO). All bradycardia and tachycardia therapies were inactivated before the MRI. Based on this standardized protocol, no major adverse outcomes occurred. All pacemaker or ICD device were reprogrammed in accordance with the pre-specified protocol except one case where the ICD device was left in the active mode for anti-tachycardia therapy (a protocol violation) and the generator could not be interrogated after MRI and required immediate replacement. In addition to patient safety, the authors also measure the functionality of the devices pre-MRI and post-MRI. One of these measurements were battery voltage changes, a small decrease was noted for both pacemakers and ICDs as expected. The radiofrequency energy generated during MRI scanning creates a temporary decrease in battery voltage, which had resolved in all pacemaker cases although some ICD voltage decreases of 0.04 V or more had not resolved by the end of the 6 month post-MRI follow-up.

Several limitations exist. The study registry included devices and leads from different manufacturers, but did not report outcomes by manufacturer. While overall it appears to be safe to conduct an MRI study for patients who have non–MRI-conditional devices, this study did not provide enough information for patients younger than 18 years of age, patients who required repeat MRI studies, MRI examinations of the thorax, or higher MRI field strengths—the newer 3 tesla high-resolution MRI machines.

Applications for Clinical Practice

This multicenter prospective cohort study provides strong evidence that patients with a non–MRI-conditional pacemaker or defibrillator can receive nonthoracic MRI studies at 1.5 tesla when a straight pre-scanning device interrogation is performed per the standardized protocol.

—Ka Ming Gordon Ngai, MD, MPH

Study Overview

Objective. To assess the risks associated with magnetic resonance imaging (MRI) in patients with a pacemaker or implantable cardioverter-defibrillator (ICD) that is “non–MRI-conditional.”

Design. Prospective cohort study using the multicenter MagnaSafe Registry.

Setting and participants. Patients were included in the registry if they were 18 years of age or older and had a non–MRI-conditional pacemaker or ICD generator, from any manufacturer, that was implanted after 2001, with leads from any manufacturer, and if the patient’s physician determined that nonthoracic MRI at 1.5 tesla was clinically indicated. Exclusion criteria included an abandoned or inactive lead that could not be interrogated, an MRI-conditional pacemaker, a device implanted in a nonthoracic location, or a device with a battery that was near the end of its battery life. In addition, pacing-dependent patients with an ICD were also excluded.

Main outcome measures. The primary outcomes of the study were death, generator or lead failure requiring immediate replacement, loss of capture (for pacing-dependent patients with pacemakers), new-onset arrhythmia, and partial or full generator electrical reset. The secondary outcomes were changes in device settings including: a battery voltage decrease of 0.04V or more, a pacing lead threshold increase of 0.5V or more, a P-wave amplitude decrease of 50% or more, an R-wave amplitude decrease of 25% or more and of 50% or more, a pacing lead impedance change of 50 ohms or more, and a high-voltage (shock) lead impedance change of 3 ohms or more.

Main results. Between April 2009 and April 2014, clinically indicated nonthoracic MRI was performed in a total of 1000 pacemaker cases (818 patients) and 500 ICD cases (428 patients) across 19 centers in the United States. The majority (75%) of the MRI examinations were performed on the brain or the spine. The mean time patients spent within the magnetic field was 44 minutes. Four patients reported symptoms of generator-site discomfort; one patient with an ICD was removed from the scanner when a sensation of heating was described at the site of the generator implanted and did not complete the examination.

Regarding primary outcomes, no deaths, lead failures, losses of capture, or ventricular arrhythmias occurred during MRI. One ICD device was left in the active mode for anti-tachycardia therapy (a protocol violation) and the generator could not be interrogated after MRI and required immediate replacement. Four patients had atrial fibrillation and 2 patients had atrial flutter during or immediately after the MRI. All 6 patients returned to sinus rhythm within 49 hours after MRI. No ventricular arrhythmias were noted. There were also 6 cases of partial generator electrical reset with no clinical significance.

Regarding secondary outcomes, a decrease of 50% or more in P-wave amplitude was detected in 0.9% of pacemaker leads and in 0.3% of ICD leads; a decrease of 25% or more in R-wave amplitude was detected in 3.9% of pacemaker leads and in 1.5% of ICD leads, and a decrease of 50% or more in R-wave amplitude was detected in no pacemaker leads and in 0.2% of ICD leads. An increase in pacing lead threshold of 0.5 V or more was detected in 0.7% of pacemaker leads and in 0.8% of ICD leads. A pacing lead impedance change of 50 ohms or more was noted in 3.3% of pacemakers and in 4.2% of ICDs.

Conclusion. Device or lead failure did not occur in any patient with a non–MRI-conditional pacemaker or ICD who underwent clinically indicated nonthoracic MRI at 1.5 tesla when patients were appropriately screened and had the cardiac device reprogrammed in accordance with the protocol. Substantial changes in device settings were infrequent and did not result in clinical adverse events.

Commentary

It is estimated that 2 million people in the United States and an additional 6 million worldwide have an implanted non–MRI-conditional cardiac pacemaker or ICD [1]. At least half of patients with such devices are predicted to have a clinical indication for MRI during their lifetime after device implantation [2]. The use of MRI poses concerns due to the potential for magnetic field–induced cardiac lead heating, which could result in myocardial thermal injury and detrimental changes in pacing properties [3,4].

In this study, Russo and colleagues assessed the risks for patients with a non-MRI-conditional pacemaker or ICD receiving an MRI scan using a pre-scanning protocol. If the patient was asymptomatic and had an intrinsic heart rate of at least 40 beats per minute, the device was programmed to a no-pacing mode (ODO or OVO). Symptomatic patients or those with an intrinsic heart rate of less than 40 beats per minute were determined to be pacing-dependent, and the device was reprogrammed to an asynchronous pacing mode (DOO or VOO). All bradycardia and tachycardia therapies were inactivated before the MRI. Based on this standardized protocol, no major adverse outcomes occurred. All pacemaker or ICD device were reprogrammed in accordance with the pre-specified protocol except one case where the ICD device was left in the active mode for anti-tachycardia therapy (a protocol violation) and the generator could not be interrogated after MRI and required immediate replacement. In addition to patient safety, the authors also measure the functionality of the devices pre-MRI and post-MRI. One of these measurements were battery voltage changes, a small decrease was noted for both pacemakers and ICDs as expected. The radiofrequency energy generated during MRI scanning creates a temporary decrease in battery voltage, which had resolved in all pacemaker cases although some ICD voltage decreases of 0.04 V or more had not resolved by the end of the 6 month post-MRI follow-up.

Several limitations exist. The study registry included devices and leads from different manufacturers, but did not report outcomes by manufacturer. While overall it appears to be safe to conduct an MRI study for patients who have non–MRI-conditional devices, this study did not provide enough information for patients younger than 18 years of age, patients who required repeat MRI studies, MRI examinations of the thorax, or higher MRI field strengths—the newer 3 tesla high-resolution MRI machines.

Applications for Clinical Practice

This multicenter prospective cohort study provides strong evidence that patients with a non–MRI-conditional pacemaker or defibrillator can receive nonthoracic MRI studies at 1.5 tesla when a straight pre-scanning device interrogation is performed per the standardized protocol.

—Ka Ming Gordon Ngai, MD, MPH

Study Overview

Objective. To assess the prevalence and determinants of patient–oncologist discordance in opinion of prognosis, and evaluate how often patients are aware of this discordance.

Design. Cross-sectional study.

Setting and participants. The study included 236 adult patients with advanced cancer and their 38 oncologists at academic and community oncology practices in Rochester, New York, and Sacramento, California. Inpatients and those already enrolled in hospice were excluded.

Main outcome measures. Patients and their oncologists independently reported their ratings of 2-year survival probability on a postindex visit multiple-choice questionnaire (response options included 100%, about 90%, about 75%, about 50%, about 25%, about 10%, and 0%). Prognostic discordance was defined as a more than 1 category difference between the patient and physician prognostic ratings. All patients were asked to report how they believed their oncologist would rate their 2-year survival probability. Those who correctly perceived their oncologist’s rating of their prognosis (within 1 category) were defined as knowing their oncologist’s opinion, and the rest were defined as not knowing their oncologist’s opinion. This distinction was used to categorize patients whose self-rating of their prognosis was discordant from their oncologist’s rating as either knowingly discordant or unknowingly discordant. Patient characteristics including age, sex, race/ethnicity, education, income, aggressiveness of cancer, self-efficacy with health care communication, recall of prognostic discussion with the oncologist, and end-of-life treatment preferences were evaluated as potential determinants of prognostic discordance.

Main results. 68% of patients rated their 2-year survival probability discordantly from their oncologists. Among these, 96% rated their prognosis more optimistically than their oncologists, and 89% were unaware that their opinions differed from that of their oncologists. Prognostic discordance was more common among nonwhite compared to white patients (95% versus 65%, P = 0.03). The prevalence of prognostic discordance did not significantly differ based on the other patient characteristics studied. Among patients whose prognostic ratings were discordant from their oncologist’s, 99% reported that they wanted to be involved in treatment decision making, and 70% were interested in involving palliative care when the end of life became near.

Conclusion. Patient–oncologist discordance about prognosis was common, particularly among nonwhite patients. In cases of prognostic discordance, patients rarely knew that their opinion differed from that of their oncologist, suggesting a lack of successful communication of prognostic information.

Commentary

Prior studies have noted that patients with advanced cancer perceive prognosis more optimistically than their physicians [1–3]. In a large national prospective observational study, the majority of patients receiving chemotherapy for metastatic (stage IV) lung or colorectal cancer inaccurately believed that chemotherapy was likely to be curative, potentially compromising their ability to make informed treatment decisions [4]. In the present study by Gramling et al, the authors confirm the observation that patients are more optimistic about prognosis than their oncologists, and furthermore demonstrate that most patients are unaware of the discrepancy, suggesting a failure of communication. As in prior studies, racial disparity in prognostic understanding was observed, with nonwhite patients being more likely to have overly optimistic views of their prognosis [4,5].

While the perceived 2-year survival probability is a somewhat arbitrary measure of prognostic opinion, it provides a useful representation of how one views the expected trajectory of disease. A high perceived likelihood of 2-year survival implies a view that long-term disease control can be achieved, whereas a low perceived likelihood of 2-year survival implies acknowledgement of terminal illness. This study effectively contrasts patient and physician opinions of 2-year survival probability, but it does not discriminate among clinically relevant differences in opinions within the 0–2 year prognostic range. For example, a patient whose oncologist believes his prognosis is < 6 months may be an appropriate candidate for hospice, but the patient may be unprepared to make the transition to hospice if he believes his prognosis is closer to a year or more. While the patient and oncologist may agree that 2-year survival is unlikely, they may have differing beliefs about the appropriateness of certain interventions based on their discrepant short-term prognostic views. Additional studies looking at perceived probabilities of short-term survival may be helpful in assessing patients’ readiness to transition to symptom-focused care when medically appropriate.

The authors designate 7 categories of 2-year survival probability (100%, about 90%, about 75%, about 50%, about 25%, about 10%, and 0%). The differences in percentage between prognostic categories are not evenly distributed, and therefore definition of discordance is non-uniform. The smaller percentage difference at the highest and lowest ends of the scale may result in overestimation of discordance at these extremes. For example, a patient rating her 2-year survival probability at 100% would be defined as having a discordant viewpoint from an oncologist rating her 2-year survival at 75% (as would be realistic for a diagnosis such as metastatic colon cancer). Given the imprecise nature of prognostication, the views of the patient and oncologist in this example are arguably similar, and perhaps should not be categorized as discordant.

As noted, patients already enrolled in hospice were excluded from the study, thus omitting a key group of patients whose prognostic views are more likely to be concordant with their physicians’ views. This group may be better captured in a prospective study of prognostic discordance among newly diagnosed advanced cancer patients after initial oncology consultation, allowing for inclusion of those who make an early transition to hospice.

Applications for Clinical Practice

Although clinicians tend to overestimate prognosis, their predictions correlate with outcomes in advanced cancer [6], and may therefore provide a useful framework for patients to understand the likely course of their disease. However, physicians often avoid explicit discussion of prognosis by shrouding prognostic information in discussions of radiographic findings, and quickly transitioning to discussion of treatment options [7]. Patients and families rarely inquire about prognosis, further limiting disclosure of prognostic information [7,8]. Even when prognostic information is explicitly stated, patients may misinterpret the information [4], potentially adversely affecting their ability to participate in shared decision making.

Some useful approaches for successfully communicating prognostic information may include asking patients what information they wish to hear before it is disclosed, providing prognostic data for patients with similar disease states while acknowledging individual variability, clearly defining the intent of proposed therapy (ie, curative versus noncurative), and asking the patient to restate information in order to assess understanding. Early involvement of palliative care specialists may help reinforce understanding about prognosis and goals of therapy, facilitate advance care planning, and reduce aggressive interventions at the end of life [2]. Ongoing research is directed at identifying effective interventions to improve communication between patients with advanced cancer and their oncologists [9].

—Irene M. Hutchins, MD
Scripps Cancer Center, La Jolla, CA

Study Overview

Objective. To assess the prevalence and determinants of patient–oncologist discordance in opinion of prognosis, and evaluate how often patients are aware of this discordance.

Design. Cross-sectional study.

Setting and participants. The study included 236 adult patients with advanced cancer and their 38 oncologists at academic and community oncology practices in Rochester, New York, and Sacramento, California. Inpatients and those already enrolled in hospice were excluded.

Main outcome measures. Patients and their oncologists independently reported their ratings of 2-year survival probability on a postindex visit multiple-choice questionnaire (response options included 100%, about 90%, about 75%, about 50%, about 25%, about 10%, and 0%). Prognostic discordance was defined as a more than 1 category difference between the patient and physician prognostic ratings. All patients were asked to report how they believed their oncologist would rate their 2-year survival probability. Those who correctly perceived their oncologist’s rating of their prognosis (within 1 category) were defined as knowing their oncologist’s opinion, and the rest were defined as not knowing their oncologist’s opinion. This distinction was used to categorize patients whose self-rating of their prognosis was discordant from their oncologist’s rating as either knowingly discordant or unknowingly discordant. Patient characteristics including age, sex, race/ethnicity, education, income, aggressiveness of cancer, self-efficacy with health care communication, recall of prognostic discussion with the oncologist, and end-of-life treatment preferences were evaluated as potential determinants of prognostic discordance.

Main results. 68% of patients rated their 2-year survival probability discordantly from their oncologists. Among these, 96% rated their prognosis more optimistically than their oncologists, and 89% were unaware that their opinions differed from that of their oncologists. Prognostic discordance was more common among nonwhite compared to white patients (95% versus 65%, P = 0.03). The prevalence of prognostic discordance did not significantly differ based on the other patient characteristics studied. Among patients whose prognostic ratings were discordant from their oncologist’s, 99% reported that they wanted to be involved in treatment decision making, and 70% were interested in involving palliative care when the end of life became near.

Conclusion. Patient–oncologist discordance about prognosis was common, particularly among nonwhite patients. In cases of prognostic discordance, patients rarely knew that their opinion differed from that of their oncologist, suggesting a lack of successful communication of prognostic information.

Commentary

Prior studies have noted that patients with advanced cancer perceive prognosis more optimistically than their physicians [1–3]. In a large national prospective observational study, the majority of patients receiving chemotherapy for metastatic (stage IV) lung or colorectal cancer inaccurately believed that chemotherapy was likely to be curative, potentially compromising their ability to make informed treatment decisions [4]. In the present study by Gramling et al, the authors confirm the observation that patients are more optimistic about prognosis than their oncologists, and furthermore demonstrate that most patients are unaware of the discrepancy, suggesting a failure of communication. As in prior studies, racial disparity in prognostic understanding was observed, with nonwhite patients being more likely to have overly optimistic views of their prognosis [4,5].

While the perceived 2-year survival probability is a somewhat arbitrary measure of prognostic opinion, it provides a useful representation of how one views the expected trajectory of disease. A high perceived likelihood of 2-year survival implies a view that long-term disease control can be achieved, whereas a low perceived likelihood of 2-year survival implies acknowledgement of terminal illness. This study effectively contrasts patient and physician opinions of 2-year survival probability, but it does not discriminate among clinically relevant differences in opinions within the 0–2 year prognostic range. For example, a patient whose oncologist believes his prognosis is < 6 months may be an appropriate candidate for hospice, but the patient may be unprepared to make the transition to hospice if he believes his prognosis is closer to a year or more. While the patient and oncologist may agree that 2-year survival is unlikely, they may have differing beliefs about the appropriateness of certain interventions based on their discrepant short-term prognostic views. Additional studies looking at perceived probabilities of short-term survival may be helpful in assessing patients’ readiness to transition to symptom-focused care when medically appropriate.

The authors designate 7 categories of 2-year survival probability (100%, about 90%, about 75%, about 50%, about 25%, about 10%, and 0%). The differences in percentage between prognostic categories are not evenly distributed, and therefore definition of discordance is non-uniform. The smaller percentage difference at the highest and lowest ends of the scale may result in overestimation of discordance at these extremes. For example, a patient rating her 2-year survival probability at 100% would be defined as having a discordant viewpoint from an oncologist rating her 2-year survival at 75% (as would be realistic for a diagnosis such as metastatic colon cancer). Given the imprecise nature of prognostication, the views of the patient and oncologist in this example are arguably similar, and perhaps should not be categorized as discordant.

As noted, patients already enrolled in hospice were excluded from the study, thus omitting a key group of patients whose prognostic views are more likely to be concordant with their physicians’ views. This group may be better captured in a prospective study of prognostic discordance among newly diagnosed advanced cancer patients after initial oncology consultation, allowing for inclusion of those who make an early transition to hospice.

Applications for Clinical Practice

Although clinicians tend to overestimate prognosis, their predictions correlate with outcomes in advanced cancer [6], and may therefore provide a useful framework for patients to understand the likely course of their disease. However, physicians often avoid explicit discussion of prognosis by shrouding prognostic information in discussions of radiographic findings, and quickly transitioning to discussion of treatment options [7]. Patients and families rarely inquire about prognosis, further limiting disclosure of prognostic information [7,8]. Even when prognostic information is explicitly stated, patients may misinterpret the information [4], potentially adversely affecting their ability to participate in shared decision making.

Some useful approaches for successfully communicating prognostic information may include asking patients what information they wish to hear before it is disclosed, providing prognostic data for patients with similar disease states while acknowledging individual variability, clearly defining the intent of proposed therapy (ie, curative versus noncurative), and asking the patient to restate information in order to assess understanding. Early involvement of palliative care specialists may help reinforce understanding about prognosis and goals of therapy, facilitate advance care planning, and reduce aggressive interventions at the end of life [2]. Ongoing research is directed at identifying effective interventions to improve communication between patients with advanced cancer and their oncologists [9].

—Irene M. Hutchins, MD
Scripps Cancer Center, La Jolla, CA

Study Overview

Objective. To assess the prevalence and determinants of patient–oncologist discordance in opinion of prognosis, and evaluate how often patients are aware of this discordance.

Design. Cross-sectional study.

Setting and participants. The study included 236 adult patients with advanced cancer and their 38 oncologists at academic and community oncology practices in Rochester, New York, and Sacramento, California. Inpatients and those already enrolled in hospice were excluded.

Main outcome measures. Patients and their oncologists independently reported their ratings of 2-year survival probability on a postindex visit multiple-choice questionnaire (response options included 100%, about 90%, about 75%, about 50%, about 25%, about 10%, and 0%). Prognostic discordance was defined as a more than 1 category difference between the patient and physician prognostic ratings. All patients were asked to report how they believed their oncologist would rate their 2-year survival probability. Those who correctly perceived their oncologist’s rating of their prognosis (within 1 category) were defined as knowing their oncologist’s opinion, and the rest were defined as not knowing their oncologist’s opinion. This distinction was used to categorize patients whose self-rating of their prognosis was discordant from their oncologist’s rating as either knowingly discordant or unknowingly discordant. Patient characteristics including age, sex, race/ethnicity, education, income, aggressiveness of cancer, self-efficacy with health care communication, recall of prognostic discussion with the oncologist, and end-of-life treatment preferences were evaluated as potential determinants of prognostic discordance.

Main results. 68% of patients rated their 2-year survival probability discordantly from their oncologists. Among these, 96% rated their prognosis more optimistically than their oncologists, and 89% were unaware that their opinions differed from that of their oncologists. Prognostic discordance was more common among nonwhite compared to white patients (95% versus 65%, P = 0.03). The prevalence of prognostic discordance did not significantly differ based on the other patient characteristics studied. Among patients whose prognostic ratings were discordant from their oncologist’s, 99% reported that they wanted to be involved in treatment decision making, and 70% were interested in involving palliative care when the end of life became near.

Conclusion. Patient–oncologist discordance about prognosis was common, particularly among nonwhite patients. In cases of prognostic discordance, patients rarely knew that their opinion differed from that of their oncologist, suggesting a lack of successful communication of prognostic information.

Commentary

Prior studies have noted that patients with advanced cancer perceive prognosis more optimistically than their physicians [1–3]. In a large national prospective observational study, the majority of patients receiving chemotherapy for metastatic (stage IV) lung or colorectal cancer inaccurately believed that chemotherapy was likely to be curative, potentially compromising their ability to make informed treatment decisions [4]. In the present study by Gramling et al, the authors confirm the observation that patients are more optimistic about prognosis than their oncologists, and furthermore demonstrate that most patients are unaware of the discrepancy, suggesting a failure of communication. As in prior studies, racial disparity in prognostic understanding was observed, with nonwhite patients being more likely to have overly optimistic views of their prognosis [4,5].

While the perceived 2-year survival probability is a somewhat arbitrary measure of prognostic opinion, it provides a useful representation of how one views the expected trajectory of disease. A high perceived likelihood of 2-year survival implies a view that long-term disease control can be achieved, whereas a low perceived likelihood of 2-year survival implies acknowledgement of terminal illness. This study effectively contrasts patient and physician opinions of 2-year survival probability, but it does not discriminate among clinically relevant differences in opinions within the 0–2 year prognostic range. For example, a patient whose oncologist believes his prognosis is < 6 months may be an appropriate candidate for hospice, but the patient may be unprepared to make the transition to hospice if he believes his prognosis is closer to a year or more. While the patient and oncologist may agree that 2-year survival is unlikely, they may have differing beliefs about the appropriateness of certain interventions based on their discrepant short-term prognostic views. Additional studies looking at perceived probabilities of short-term survival may be helpful in assessing patients’ readiness to transition to symptom-focused care when medically appropriate.

The authors designate 7 categories of 2-year survival probability (100%, about 90%, about 75%, about 50%, about 25%, about 10%, and 0%). The differences in percentage between prognostic categories are not evenly distributed, and therefore definition of discordance is non-uniform. The smaller percentage difference at the highest and lowest ends of the scale may result in overestimation of discordance at these extremes. For example, a patient rating her 2-year survival probability at 100% would be defined as having a discordant viewpoint from an oncologist rating her 2-year survival at 75% (as would be realistic for a diagnosis such as metastatic colon cancer). Given the imprecise nature of prognostication, the views of the patient and oncologist in this example are arguably similar, and perhaps should not be categorized as discordant.

As noted, patients already enrolled in hospice were excluded from the study, thus omitting a key group of patients whose prognostic views are more likely to be concordant with their physicians’ views. This group may be better captured in a prospective study of prognostic discordance among newly diagnosed advanced cancer patients after initial oncology consultation, allowing for inclusion of those who make an early transition to hospice.

Applications for Clinical Practice

Although clinicians tend to overestimate prognosis, their predictions correlate with outcomes in advanced cancer [6], and may therefore provide a useful framework for patients to understand the likely course of their disease. However, physicians often avoid explicit discussion of prognosis by shrouding prognostic information in discussions of radiographic findings, and quickly transitioning to discussion of treatment options [7]. Patients and families rarely inquire about prognosis, further limiting disclosure of prognostic information [7,8]. Even when prognostic information is explicitly stated, patients may misinterpret the information [4], potentially adversely affecting their ability to participate in shared decision making.

Some useful approaches for successfully communicating prognostic information may include asking patients what information they wish to hear before it is disclosed, providing prognostic data for patients with similar disease states while acknowledging individual variability, clearly defining the intent of proposed therapy (ie, curative versus noncurative), and asking the patient to restate information in order to assess understanding. Early involvement of palliative care specialists may help reinforce understanding about prognosis and goals of therapy, facilitate advance care planning, and reduce aggressive interventions at the end of life [2]. Ongoing research is directed at identifying effective interventions to improve communication between patients with advanced cancer and their oncologists [9].

—Irene M. Hutchins, MD
Scripps Cancer Center, La Jolla, CA

Judge Neil Gorsuch has moved one step closer to becoming the next U.S. Supreme Court Justice.

The U.S. Senate Committee on the Judiciary approved Judge Gorsuch’s nomination by a 11-9 vote on April 3. The vote was a strict party line vote with 11 Republicans voting in favor of Judge Gorsuch and 9 Democrats voting against him.

[email protected]

On Twitter @legal_med

Judge Neil Gorsuch has moved one step closer to becoming the next U.S. Supreme Court Justice.

The U.S. Senate Committee on the Judiciary approved Judge Gorsuch’s nomination by a 11-9 vote on April 3. The vote was a strict party line vote with 11 Republicans voting in favor of Judge Gorsuch and 9 Democrats voting against him.

[email protected]

On Twitter @legal_med

Judge Neil Gorsuch has moved one step closer to becoming the next U.S. Supreme Court Justice.

The U.S. Senate Committee on the Judiciary approved Judge Gorsuch’s nomination by a 11-9 vote on April 3. The vote was a strict party line vote with 11 Republicans voting in favor of Judge Gorsuch and 9 Democrats voting against him.

[email protected]

On Twitter @legal_med

SAN ANTONIO – At 5-year follow-up, outcomes were slightly better on most measures for transvaginal uterosacral ligament suspension versus transvaginal sacrospinous ligament fixation for apical prolapse, but the differences were not statistically significant, according to the first randomized trial to compare the two techniques.

Quality of life improvements were durable, but the overall 5-year success rate – defined as the absence of descent of the vaginal apex more than one-third into the vagina; anterior or posterior vaginal wall descent beyond the hymen; bothersome vaginal bulge symptoms; and further treatment for prolapse – was 39% in the 109 women randomized to bilateral uterosacral ligament suspension (ULS) and 30% in the 109 women randomized to unilateral sacrospinous ligament fixation (SSLF).

* The meeting sponsor information was updated 6/9/2017.

[email protected]
 

SAN ANTONIO – At 5-year follow-up, outcomes were slightly better on most measures for transvaginal uterosacral ligament suspension versus transvaginal sacrospinous ligament fixation for apical prolapse, but the differences were not statistically significant, according to the first randomized trial to compare the two techniques.

Quality of life improvements were durable, but the overall 5-year success rate – defined as the absence of descent of the vaginal apex more than one-third into the vagina; anterior or posterior vaginal wall descent beyond the hymen; bothersome vaginal bulge symptoms; and further treatment for prolapse – was 39% in the 109 women randomized to bilateral uterosacral ligament suspension (ULS) and 30% in the 109 women randomized to unilateral sacrospinous ligament fixation (SSLF).

* The meeting sponsor information was updated 6/9/2017.

[email protected]
 

SAN ANTONIO – At 5-year follow-up, outcomes were slightly better on most measures for transvaginal uterosacral ligament suspension versus transvaginal sacrospinous ligament fixation for apical prolapse, but the differences were not statistically significant, according to the first randomized trial to compare the two techniques.

Quality of life improvements were durable, but the overall 5-year success rate – defined as the absence of descent of the vaginal apex more than one-third into the vagina; anterior or posterior vaginal wall descent beyond the hymen; bothersome vaginal bulge symptoms; and further treatment for prolapse – was 39% in the 109 women randomized to bilateral uterosacral ligament suspension (ULS) and 30% in the 109 women randomized to unilateral sacrospinous ligament fixation (SSLF).

* The meeting sponsor information was updated 6/9/2017.

[email protected]
 

From the Department of Preventive Medicine and Environmental Health, University of Kentucky College of Public Health, Lexington, KY.

Abstract

• Objective: To review the classification, epidemiology, clinical presentation, and evaluation of patients with chronic obstructive pulmonary disease (COPD).
• Methods: Review of the literature.
• Results: While smoking remains the most important risk factor for COPD in much of the developed world, other risk factors, including genetic factors and occupational or environmental exposures, remain important. COPD is the third leading cause of death in the United States. In 2011, 13.7 million adults aged ≥ 25 years were diagnosed with COPD in the United States, and as many as 12 million adults may have COPD that is undiagnosed. In 2010, COPD was responsible for an estimated 10.3 million physician office visits and 1.5 million emergency room visits and was estimated to be the second leading cause of disability-adjusted life years lost among the US population. COPD has primary, secondary, and tertiary prevention strategies. The treatment of COPD has improved in recent years, with new therapies improving patient quality of life.
• Conclusion: COPD remains a serious public health problem that is often underdiagnosed, particularly in its early stages.

Key words: Chronic obstructive pulmonary disease; epidemiology; mortality; smoking; evaluation.

Chronic obstructive pulmonary disease (COPD) is characterized by fixed airflow obstruction with breathing-related symptoms, such as chronic cough, exertional dyspnea, expectoration, and wheeze [1]. These symptoms may occur in conjunction with airway hyperresponsiveness and overlap with other chronic diseases such as asthma. Although COPD is a nonspecific term referring to a set of conditions that develop progressively as a result of a number of different disease processes, it most commonly refers to chronic bronchitis and emphysema. These conditions can be present with or without significant physical impairment. Despite being a very common disease and the third leading cause of death in the United States [2], COPD often is a silent and unrecognized disease, particularly in its early phases [3], and may go untreated.

In this article, we review the classification, epidemiology, clinical presentation, and assessment of patients with COPD.

Definition and Classification

Several different definitions have existed for COPD [4–8]. The Global Initiative for Chronic Obstructive Lung Disease (GOLD), an international collaboration of leading experts in COPD launched in the late 90s with a goal to develop evidence-based recommendations for diagnosis and management of COPD [4], currently defines COPD as “a common, preventable and treatable disease that is characterized by persistent respiratory symptoms and airflow limitation that is due to airway and/or alveolar abnormalities usually caused by significant exposure to noxious particles or gases” [4].

Severity of COPD has typically been determined using the degree of lung function impairment, although the wisdom of this approach has been questioned, 

Previous definitions of COPD differentiated between chronic bronchitis, asthma, and emphysema, acknowledging that there is frequently overlap between these disease entities [12,13]. The GOLD definition of COPD does not differentiate between chronic bronchitis and emphysema but does note that although asthma and COPD can coexist [4], the largely reversible airflow limitation in asthma merits different therapeutic approaches than the largely irreversible airflow limitation of COPD. The overlap of asthma and COPD in a significant proportion of patients has been the focus of recent work [14].

Epidemiology

Prevalence of COPD

The Behavioral Risk Factor Surveillance System (BRFSS) is an ongoing national random-digit-dialed telephone survey of landline and cellphone households designed to measure behavioral risk factors for the noninstitutionalized adult population of the US [15]. An affirmative response to the following question was defined as physician-diagnosed COPD: “Have you ever been told by a doctor or other health professional that you have chronic obstructive pulmonary disease (COPD), emphysema, or bronchitis?”[16]. Based on 2011 BRFSS survey, 13.7 million adults aged ≥ 25 years were estimated to have a self-reported physician diagnosis of COPD in the United States. The greatest age-adjusted prevalence was found to be clustered along the Ohio River Valley and the southern states [16].

The National Health Interview Survey (NHIS) is an annually conducted, nationally representative survey of the civilian noninstitutionalized population aged 18 years and older. A positive response to one or both of the following questions was used to define COPD: “Have you ever been told by a doctor or other health professional that you had emphysema?” and “During the past 12 months, have you been told by a doctor or other health professional that you had chronic bronchitis?” Age-adjusted COPD prevalence estimates showed significant interyear variation during 1999–2011 period, and were higher in women than in men with the highest prevalence noted in 2001 for both genders [16].

The NHIS estimates for COPD have 2 important limitations. First, these estimates depend on the proper recognition and diagnosis of COPD by both the study participants and their health care providers. This would tend to bias the estimates toward counting fewer cases than actually exist. A bias in the opposite direction, however, is that the term chronic bronchitis in this survey is not precisely defined and could be interpreted as recurrent episodes of acute bronchitis. The finding that “chronic bronchitis” has been reported in 3% to 4% of children supports the presence of this potential bias. The second limitation is that this survey is not able to validate, through physiologic evaluation, whether airway obstruction is present or absent.

These limitations were addressed, in part, by separate nationally representative US surveys that include an examination component, such as the National Health and Nutrition Examination Surveys (NHANES) [17]. An analysis of these data from 1988–1994 and 2007–2012 [18] demonstrated that over 70% of people with evidence of obstruction (based on an FEV₁/FVC < 70%) did not have a diagnosis of lung disease (COPD or asthma). In addition, people with evidence of obstruction had a higher risk of mortality whether or not they had diagnosed lung disease [18].

Evaluation of “reversibility” of the airway obstruction requires the administration of bronchodilator, which is not a part of most population-based studies. A subset of participants in the NHANES 2007–2012 survey received a bronchodilator, with a decrease in the estimated prevalence of obstruction from 20.9% to 14.0% [19]. However, a closer look at similar data from a study where all people got a bronchodilator reveal that only a small proportion of people with “reversibility” actually had a significant response to the bronchodilator [20]. In a clinic-based study of subjects with COPD who were aged 69 years and older, 31% demonstrated reversibility, defined as a 15% improvement (from baseline) in FVC and FEV₁ following administration of an inhaled bronchodilator [21]. In this study, subjects with more severe obstruction were more likely to have reversibility but would also be more likely to continue to have diminished lung function after maximum improvement was obtained, thus being classified as having “partial reversibility.”

The presence of significant reversibility or partial reversibility in patients with COPD [15] and nonreversible airflow obstruction in asthma patients [22] demonstrates that these diseases can coexist or, alternatively, that there is overlap and imprecision in the ways that these diseases are clinically diagnosed.

Morbidity and Mortality

COPD is a leading cause of disease morbidity and mortality in the United States. The National Center for Health Statistics (NCHS) conducts ongoing surveillance of several health indicators nationally. The NCHS collects physician office visit data using the National Ambulatory Medical Care Survey [23], emergency department visit data and hospital outpatient data using the National Hospital Ambulatory Medical Care Survey [24], hospitalization data using the National Hospital Discharge Survey [25], and death data using the mortality component of the National Vital Statistics System [26]. The following data include the number and rate of COPD events in adults in the United States (using International Classification of Diseases, 9th Revision, Clinical Modification [ICD-9-CM], codes 490, 491, 492 and 496) in these data sets for the most recent years available.

In 2010, COPD was responsible for an estimated 10.3 million physician office visits, with a resulting age-adjusted rate of 494.8 per 10,000 US civilian population [16]. COPD was also responsible for an estimated 1.5 million emergency room visits, with a resulting age-adjusted rate of 72 visits per 10,000 population [16].

COPD is a leading cause of hospitalization in US adults, particularly in older populations. In 2010, almost 699,000 hospitalizations, were attributed to COPD. The age-adjusted rate of COPD hospitalizations (as the primary cause of hospitalization) was 32.2 per 10,000 population in 2010 [16].

Deaths due to or associated with COPD have not significantly changed since 1999. While the age-adjusted death rate among men declined during 1999–2010 (P = 0.001), the rate among women has not changed significantly (P = 0.127). In 2010, 63, 778 men and 69, 797 women aged ≥ 25 years died of COPD [26]. One of the limitations of using the mortality component of the National Vital Statistics System is that it is based on the underlying cause of death as reported on the death certificate; however, many decedents with COPD listed on the death certificate have their death attributed to another cause [27]. The significance of COPD as a contributor to death is undefined when it is present with diseases more likely to be attributed as the underlying cause of death, such as myocardial infarction or lung cancer [28].

COPD is a very costly disease, with estimated direct medical costs in 2004 of $20.9 billion. The estimated indirect costs related to morbidity (loss of work time and productivity) and premature mortality is an additional $16.3 billion, for a total of $37.2 billion [29]. Because COPD may be present but not listed as the underlying cause of death or the primary reason for hospitalization, these cited estimates may underestimate the true cost of COPD. For example, in another analysis of COPD costs in the US, the total for 2010 was estimated at $32.1 billion [30], but could be up to $100 billion [31] depending on the assumptions surrounding comorbid disease.

Another manifestation of the importance of COPD is its effect on the burden of disease in a population determined using disability-adjusted life-years (DALYs). DALYs for a disease or condition are calculated as the sum of the years of life lost due to premature mortality in the population and the years of life lost due to disability [32]. In 2010, COPD was estimated to be the second leading cause of DALYs lost among the North American population [33]. Worldwide, COPD is expected to move up from being the twelfth leading cause of DALYs lost in 1990 to the fifth leading cause in 2020 [34].

Gender Differences

Smoking-related diseases such as COPD and lung cancer are continuing to increase among women in the United States [35,36], while they have plateaued or are decreasing among men [27,37]. Some evidence has emerged that compared with men at a similar level of tobacco smoking, women may be more likely to develop COPD [38] or that the severity of COPD in women may be increased [39–41].

In the Lung Health Study, which evaluated patients with mild COPD, more women than men demonstrated increased airway responsiveness, although this difference was thought to be related to airway caliber rather than gender [42]. Adult women are more likely to both develop and die of asthma than are men [43–45]. In NHANES III, whereas women reported more physician-diagnosed COPD and asthma than men, men and women had similar rates of decreased lung function, and a similar proportion of both men and women with low lung function had undiagnosed lung disease [3]. The current evidence is inadequate to determine whether women who smoke are more likely to develop COPD or have more severe COPD than men, although this question is being studied by various groups.

Risk Factors and Etiology

Smoking is the dominant risk factor for the development and progression of COPD; however, not all smokers develop COPD, and COPD does occur in persons who have never smoked [1], suggesting that other factors are important in the etiology of COPD. Alpha1-antitrypsin deficiency is an important cause of COPD in a very small percentage of cases [46]. Other undefined genetic factors certainly play an important role in COPD development [38]. The role of infections in both the development and progression of COPD is receiving increased attention, including the role of adenoviral infections in emphysema [47–49].

Occupational and environmental exposures to various pollutants (eg, particulate matter, agricultural dusts) are also important factors in the development of COPD [50,51]. Exposure to indoor air pollutants such as smoke from solid biomass fuels is a major risk factor for COPD especially among women and children in low- and middle-income countries [52,53]. Occupational exposure to fumes and dusts remains an important cause for COPD globally [53,54]. Exposure to outdoor air pollution is associated with a risk of development of COPD as well as exacerbation of the existing disease [53,55].

Clinical Presentation

COPD is heterogeneous in its presentation. Based on data from NHANES III, 44% of patients with severe airflow limitation (FEV₁ < 50% of predicted) may not report symptoms [3]. Among patients with severe airflow limitation who do report symptoms, the symptoms reported most frequently include wheezing (64%) and shortness of breath (65%).

In recent years, COPD has been increasingly recognized as a systemic illness, with effects on nutritional status, muscle wasting, and depression [56–58]. A large proportion of patients probably have components of chronic bronchitis, asthma, and emphysema occurring together. Although some of this overlap may be related to misdiagnosis, some of it may be a measure of the presence of airflow limitation reversibility, as described above. Better defining individuals in these groups may ultimately help tailor better interventions.

Some of the barriers to COPD diagnosis and subsequent treatment often include insufficient knowledge and awareness about COPD especially among primary care physicians, misdiagnosis of COPD as other respiratory diseases such asthma, as well as patient-related barriers involving lack of awareness of early symptoms of COPD and considering them to be related to aging or smoking [59].

Evaluation

The evaluation of a patient with suspected COPD is oriented toward establishing the correct diagnosis and, once this has occurred, determining the extent of the impairment such that therapy can be appropriately targeted.

Components in the evaluation of COPD are listed in Table 3. Every patient with suspected COPD should undergo a thorough history and physical examination. The history should pay particular attention to the following: exposure to risk factors; past history of asthma or allergic disease; family history of COPD; presence of comorbid diseases; effect of disease on the patient’s life, including ability to work and mental health status; and possibilities for reducing risk factors, especially smoking cessation [4]. The physical examination is rarely diagnostic in COPD because most physical abnormalities do not occur until the advanced stages of the disease. Physical examination findings in 

Pulmonary function testing is a critical part of the evaluation of suspected COPD. Whereas most patients with COPD can be managed by a primary care physician, patients with moderate or severe COPD should be evaluated by a specialist [4].

Once the diagnosis of moderate or severe COPD has been established, further testing, including chest radiograph, arterial blood gas determination, screening for α1-antitrypsin deficiency, 6-minute walk testing or exercise oxymetry may be indicated based on the patient’s history and/or clinical findings. Data from computed tomography scans are useful in some advanced cases.

Prognosis of COPD is often influenced by presence of various comorbidities including extrapulmonary, such as osteoporosis, metabolic syndrome, and depression that may be seen as parts of multimorbidity associated with aging [60,61]. Therefore, it is advised to look for comorbidities in COPD patients with any severity of airflow obstruction and treat them accordingly [4].

Therapy for COPD targets reducing risk factors, improving symptoms, and decreasing the risk of exacerbations [10]. Interventions include smoking cessation, vaccinations, decreasing exposures to occupational and environmental pollutants, pulmonary rehabilitation, bronchodilators, and corticosteroids. Select patients with advanced COPD may benefit from other interventions, such as surgical reduction of lung size, lung transplant, the phosphodiesterase inhibitor roflumilast and chronic treatment with antibiotics such as macrolides.

Conclusion

COPD is a common disease that is a leading cause of morbidity and mortality, both in the United States and worldwide. Most cases of COPD are attributable to smoking. Although its incidence among men has plateaued, it continues to increase among women. COPD, particularly in its early stages, is under-diagnosed in the United States. An increased awareness among physicians of the prevalence of mild COPD and the importance of spirometry in diagnosing the disease is important in combating the disease.

Corresponding author: David M. Mannino, MD, Department of Preventive Medicine and Environmental Health, University of Kentucky College of Public Health, 111 Washington Avenue, Lexington, KY 40536, [email protected].

 Financial disclosures: Dr. Mannino has received fees from GlaxoSmithKline, Novartis, AstraZeneca, Sunovion, and Boehringer Ingelheim for advisory board services.

From the Department of Preventive Medicine and Environmental Health, University of Kentucky College of Public Health, Lexington, KY.

Abstract

• Objective: To review the classification, epidemiology, clinical presentation, and evaluation of patients with chronic obstructive pulmonary disease (COPD).
• Methods: Review of the literature.
• Results: While smoking remains the most important risk factor for COPD in much of the developed world, other risk factors, including genetic factors and occupational or environmental exposures, remain important. COPD is the third leading cause of death in the United States. In 2011, 13.7 million adults aged ≥ 25 years were diagnosed with COPD in the United States, and as many as 12 million adults may have COPD that is undiagnosed. In 2010, COPD was responsible for an estimated 10.3 million physician office visits and 1.5 million emergency room visits and was estimated to be the second leading cause of disability-adjusted life years lost among the US population. COPD has primary, secondary, and tertiary prevention strategies. The treatment of COPD has improved in recent years, with new therapies improving patient quality of life.
• Conclusion: COPD remains a serious public health problem that is often underdiagnosed, particularly in its early stages.

Key words: Chronic obstructive pulmonary disease; epidemiology; mortality; smoking; evaluation.

Chronic obstructive pulmonary disease (COPD) is characterized by fixed airflow obstruction with breathing-related symptoms, such as chronic cough, exertional dyspnea, expectoration, and wheeze [1]. These symptoms may occur in conjunction with airway hyperresponsiveness and overlap with other chronic diseases such as asthma. Although COPD is a nonspecific term referring to a set of conditions that develop progressively as a result of a number of different disease processes, it most commonly refers to chronic bronchitis and emphysema. These conditions can be present with or without significant physical impairment. Despite being a very common disease and the third leading cause of death in the United States [2], COPD often is a silent and unrecognized disease, particularly in its early phases [3], and may go untreated.

In this article, we review the classification, epidemiology, clinical presentation, and assessment of patients with COPD.

Definition and Classification

Several different definitions have existed for COPD [4–8]. The Global Initiative for Chronic Obstructive Lung Disease (GOLD), an international collaboration of leading experts in COPD launched in the late 90s with a goal to develop evidence-based recommendations for diagnosis and management of COPD [4], currently defines COPD as “a common, preventable and treatable disease that is characterized by persistent respiratory symptoms and airflow limitation that is due to airway and/or alveolar abnormalities usually caused by significant exposure to noxious particles or gases” [4].

Severity of COPD has typically been determined using the degree of lung function impairment, although the wisdom of this approach has been questioned, 

Previous definitions of COPD differentiated between chronic bronchitis, asthma, and emphysema, acknowledging that there is frequently overlap between these disease entities [12,13]. The GOLD definition of COPD does not differentiate between chronic bronchitis and emphysema but does note that although asthma and COPD can coexist [4], the largely reversible airflow limitation in asthma merits different therapeutic approaches than the largely irreversible airflow limitation of COPD. The overlap of asthma and COPD in a significant proportion of patients has been the focus of recent work [14].

Epidemiology

Prevalence of COPD

The Behavioral Risk Factor Surveillance System (BRFSS) is an ongoing national random-digit-dialed telephone survey of landline and cellphone households designed to measure behavioral risk factors for the noninstitutionalized adult population of the US [15]. An affirmative response to the following question was defined as physician-diagnosed COPD: “Have you ever been told by a doctor or other health professional that you have chronic obstructive pulmonary disease (COPD), emphysema, or bronchitis?”[16]. Based on 2011 BRFSS survey, 13.7 million adults aged ≥ 25 years were estimated to have a self-reported physician diagnosis of COPD in the United States. The greatest age-adjusted prevalence was found to be clustered along the Ohio River Valley and the southern states [16].

The National Health Interview Survey (NHIS) is an annually conducted, nationally representative survey of the civilian noninstitutionalized population aged 18 years and older. A positive response to one or both of the following questions was used to define COPD: “Have you ever been told by a doctor or other health professional that you had emphysema?” and “During the past 12 months, have you been told by a doctor or other health professional that you had chronic bronchitis?” Age-adjusted COPD prevalence estimates showed significant interyear variation during 1999–2011 period, and were higher in women than in men with the highest prevalence noted in 2001 for both genders [16].

The NHIS estimates for COPD have 2 important limitations. First, these estimates depend on the proper recognition and diagnosis of COPD by both the study participants and their health care providers. This would tend to bias the estimates toward counting fewer cases than actually exist. A bias in the opposite direction, however, is that the term chronic bronchitis in this survey is not precisely defined and could be interpreted as recurrent episodes of acute bronchitis. The finding that “chronic bronchitis” has been reported in 3% to 4% of children supports the presence of this potential bias. The second limitation is that this survey is not able to validate, through physiologic evaluation, whether airway obstruction is present or absent.

These limitations were addressed, in part, by separate nationally representative US surveys that include an examination component, such as the National Health and Nutrition Examination Surveys (NHANES) [17]. An analysis of these data from 1988–1994 and 2007–2012 [18] demonstrated that over 70% of people with evidence of obstruction (based on an FEV₁/FVC < 70%) did not have a diagnosis of lung disease (COPD or asthma). In addition, people with evidence of obstruction had a higher risk of mortality whether or not they had diagnosed lung disease [18].

Evaluation of “reversibility” of the airway obstruction requires the administration of bronchodilator, which is not a part of most population-based studies. A subset of participants in the NHANES 2007–2012 survey received a bronchodilator, with a decrease in the estimated prevalence of obstruction from 20.9% to 14.0% [19]. However, a closer look at similar data from a study where all people got a bronchodilator reveal that only a small proportion of people with “reversibility” actually had a significant response to the bronchodilator [20]. In a clinic-based study of subjects with COPD who were aged 69 years and older, 31% demonstrated reversibility, defined as a 15% improvement (from baseline) in FVC and FEV₁ following administration of an inhaled bronchodilator [21]. In this study, subjects with more severe obstruction were more likely to have reversibility but would also be more likely to continue to have diminished lung function after maximum improvement was obtained, thus being classified as having “partial reversibility.”

The presence of significant reversibility or partial reversibility in patients with COPD [15] and nonreversible airflow obstruction in asthma patients [22] demonstrates that these diseases can coexist or, alternatively, that there is overlap and imprecision in the ways that these diseases are clinically diagnosed.

Morbidity and Mortality

COPD is a leading cause of disease morbidity and mortality in the United States. The National Center for Health Statistics (NCHS) conducts ongoing surveillance of several health indicators nationally. The NCHS collects physician office visit data using the National Ambulatory Medical Care Survey [23], emergency department visit data and hospital outpatient data using the National Hospital Ambulatory Medical Care Survey [24], hospitalization data using the National Hospital Discharge Survey [25], and death data using the mortality component of the National Vital Statistics System [26]. The following data include the number and rate of COPD events in adults in the United States (using International Classification of Diseases, 9th Revision, Clinical Modification [ICD-9-CM], codes 490, 491, 492 and 496) in these data sets for the most recent years available.

In 2010, COPD was responsible for an estimated 10.3 million physician office visits, with a resulting age-adjusted rate of 494.8 per 10,000 US civilian population [16]. COPD was also responsible for an estimated 1.5 million emergency room visits, with a resulting age-adjusted rate of 72 visits per 10,000 population [16].

COPD is a leading cause of hospitalization in US adults, particularly in older populations. In 2010, almost 699,000 hospitalizations, were attributed to COPD. The age-adjusted rate of COPD hospitalizations (as the primary cause of hospitalization) was 32.2 per 10,000 population in 2010 [16].

Deaths due to or associated with COPD have not significantly changed since 1999. While the age-adjusted death rate among men declined during 1999–2010 (P = 0.001), the rate among women has not changed significantly (P = 0.127). In 2010, 63, 778 men and 69, 797 women aged ≥ 25 years died of COPD [26]. One of the limitations of using the mortality component of the National Vital Statistics System is that it is based on the underlying cause of death as reported on the death certificate; however, many decedents with COPD listed on the death certificate have their death attributed to another cause [27]. The significance of COPD as a contributor to death is undefined when it is present with diseases more likely to be attributed as the underlying cause of death, such as myocardial infarction or lung cancer [28].

COPD is a very costly disease, with estimated direct medical costs in 2004 of $20.9 billion. The estimated indirect costs related to morbidity (loss of work time and productivity) and premature mortality is an additional $16.3 billion, for a total of $37.2 billion [29]. Because COPD may be present but not listed as the underlying cause of death or the primary reason for hospitalization, these cited estimates may underestimate the true cost of COPD. For example, in another analysis of COPD costs in the US, the total for 2010 was estimated at $32.1 billion [30], but could be up to $100 billion [31] depending on the assumptions surrounding comorbid disease.

Another manifestation of the importance of COPD is its effect on the burden of disease in a population determined using disability-adjusted life-years (DALYs). DALYs for a disease or condition are calculated as the sum of the years of life lost due to premature mortality in the population and the years of life lost due to disability [32]. In 2010, COPD was estimated to be the second leading cause of DALYs lost among the North American population [33]. Worldwide, COPD is expected to move up from being the twelfth leading cause of DALYs lost in 1990 to the fifth leading cause in 2020 [34].

Gender Differences

Smoking-related diseases such as COPD and lung cancer are continuing to increase among women in the United States [35,36], while they have plateaued or are decreasing among men [27,37]. Some evidence has emerged that compared with men at a similar level of tobacco smoking, women may be more likely to develop COPD [38] or that the severity of COPD in women may be increased [39–41].

In the Lung Health Study, which evaluated patients with mild COPD, more women than men demonstrated increased airway responsiveness, although this difference was thought to be related to airway caliber rather than gender [42]. Adult women are more likely to both develop and die of asthma than are men [43–45]. In NHANES III, whereas women reported more physician-diagnosed COPD and asthma than men, men and women had similar rates of decreased lung function, and a similar proportion of both men and women with low lung function had undiagnosed lung disease [3]. The current evidence is inadequate to determine whether women who smoke are more likely to develop COPD or have more severe COPD than men, although this question is being studied by various groups.

Risk Factors and Etiology

Smoking is the dominant risk factor for the development and progression of COPD; however, not all smokers develop COPD, and COPD does occur in persons who have never smoked [1], suggesting that other factors are important in the etiology of COPD. Alpha1-antitrypsin deficiency is an important cause of COPD in a very small percentage of cases [46]. Other undefined genetic factors certainly play an important role in COPD development [38]. The role of infections in both the development and progression of COPD is receiving increased attention, including the role of adenoviral infections in emphysema [47–49].

Occupational and environmental exposures to various pollutants (eg, particulate matter, agricultural dusts) are also important factors in the development of COPD [50,51]. Exposure to indoor air pollutants such as smoke from solid biomass fuels is a major risk factor for COPD especially among women and children in low- and middle-income countries [52,53]. Occupational exposure to fumes and dusts remains an important cause for COPD globally [53,54]. Exposure to outdoor air pollution is associated with a risk of development of COPD as well as exacerbation of the existing disease [53,55].

Clinical Presentation

COPD is heterogeneous in its presentation. Based on data from NHANES III, 44% of patients with severe airflow limitation (FEV₁ < 50% of predicted) may not report symptoms [3]. Among patients with severe airflow limitation who do report symptoms, the symptoms reported most frequently include wheezing (64%) and shortness of breath (65%).

In recent years, COPD has been increasingly recognized as a systemic illness, with effects on nutritional status, muscle wasting, and depression [56–58]. A large proportion of patients probably have components of chronic bronchitis, asthma, and emphysema occurring together. Although some of this overlap may be related to misdiagnosis, some of it may be a measure of the presence of airflow limitation reversibility, as described above. Better defining individuals in these groups may ultimately help tailor better interventions.

Some of the barriers to COPD diagnosis and subsequent treatment often include insufficient knowledge and awareness about COPD especially among primary care physicians, misdiagnosis of COPD as other respiratory diseases such asthma, as well as patient-related barriers involving lack of awareness of early symptoms of COPD and considering them to be related to aging or smoking [59].

Evaluation

The evaluation of a patient with suspected COPD is oriented toward establishing the correct diagnosis and, once this has occurred, determining the extent of the impairment such that therapy can be appropriately targeted.

Components in the evaluation of COPD are listed in Table 3. Every patient with suspected COPD should undergo a thorough history and physical examination. The history should pay particular attention to the following: exposure to risk factors; past history of asthma or allergic disease; family history of COPD; presence of comorbid diseases; effect of disease on the patient’s life, including ability to work and mental health status; and possibilities for reducing risk factors, especially smoking cessation [4]. The physical examination is rarely diagnostic in COPD because most physical abnormalities do not occur until the advanced stages of the disease. Physical examination findings in 

Pulmonary function testing is a critical part of the evaluation of suspected COPD. Whereas most patients with COPD can be managed by a primary care physician, patients with moderate or severe COPD should be evaluated by a specialist [4].

Once the diagnosis of moderate or severe COPD has been established, further testing, including chest radiograph, arterial blood gas determination, screening for α1-antitrypsin deficiency, 6-minute walk testing or exercise oxymetry may be indicated based on the patient’s history and/or clinical findings. Data from computed tomography scans are useful in some advanced cases.

Prognosis of COPD is often influenced by presence of various comorbidities including extrapulmonary, such as osteoporosis, metabolic syndrome, and depression that may be seen as parts of multimorbidity associated with aging [60,61]. Therefore, it is advised to look for comorbidities in COPD patients with any severity of airflow obstruction and treat them accordingly [4].

Therapy for COPD targets reducing risk factors, improving symptoms, and decreasing the risk of exacerbations [10]. Interventions include smoking cessation, vaccinations, decreasing exposures to occupational and environmental pollutants, pulmonary rehabilitation, bronchodilators, and corticosteroids. Select patients with advanced COPD may benefit from other interventions, such as surgical reduction of lung size, lung transplant, the phosphodiesterase inhibitor roflumilast and chronic treatment with antibiotics such as macrolides.

Conclusion

COPD is a common disease that is a leading cause of morbidity and mortality, both in the United States and worldwide. Most cases of COPD are attributable to smoking. Although its incidence among men has plateaued, it continues to increase among women. COPD, particularly in its early stages, is under-diagnosed in the United States. An increased awareness among physicians of the prevalence of mild COPD and the importance of spirometry in diagnosing the disease is important in combating the disease.

Corresponding author: David M. Mannino, MD, Department of Preventive Medicine and Environmental Health, University of Kentucky College of Public Health, 111 Washington Avenue, Lexington, KY 40536, [email protected].

 Financial disclosures: Dr. Mannino has received fees from GlaxoSmithKline, Novartis, AstraZeneca, Sunovion, and Boehringer Ingelheim for advisory board services.

From the Department of Preventive Medicine and Environmental Health, University of Kentucky College of Public Health, Lexington, KY.

Abstract

• Objective: To review the classification, epidemiology, clinical presentation, and evaluation of patients with chronic obstructive pulmonary disease (COPD).
• Methods: Review of the literature.
• Results: While smoking remains the most important risk factor for COPD in much of the developed world, other risk factors, including genetic factors and occupational or environmental exposures, remain important. COPD is the third leading cause of death in the United States. In 2011, 13.7 million adults aged ≥ 25 years were diagnosed with COPD in the United States, and as many as 12 million adults may have COPD that is undiagnosed. In 2010, COPD was responsible for an estimated 10.3 million physician office visits and 1.5 million emergency room visits and was estimated to be the second leading cause of disability-adjusted life years lost among the US population. COPD has primary, secondary, and tertiary prevention strategies. The treatment of COPD has improved in recent years, with new therapies improving patient quality of life.
• Conclusion: COPD remains a serious public health problem that is often underdiagnosed, particularly in its early stages.

Key words: Chronic obstructive pulmonary disease; epidemiology; mortality; smoking; evaluation.

Chronic obstructive pulmonary disease (COPD) is characterized by fixed airflow obstruction with breathing-related symptoms, such as chronic cough, exertional dyspnea, expectoration, and wheeze [1]. These symptoms may occur in conjunction with airway hyperresponsiveness and overlap with other chronic diseases such as asthma. Although COPD is a nonspecific term referring to a set of conditions that develop progressively as a result of a number of different disease processes, it most commonly refers to chronic bronchitis and emphysema. These conditions can be present with or without significant physical impairment. Despite being a very common disease and the third leading cause of death in the United States [2], COPD often is a silent and unrecognized disease, particularly in its early phases [3], and may go untreated.

In this article, we review the classification, epidemiology, clinical presentation, and assessment of patients with COPD.

Definition and Classification

Several different definitions have existed for COPD [4–8]. The Global Initiative for Chronic Obstructive Lung Disease (GOLD), an international collaboration of leading experts in COPD launched in the late 90s with a goal to develop evidence-based recommendations for diagnosis and management of COPD [4], currently defines COPD as “a common, preventable and treatable disease that is characterized by persistent respiratory symptoms and airflow limitation that is due to airway and/or alveolar abnormalities usually caused by significant exposure to noxious particles or gases” [4].

Severity of COPD has typically been determined using the degree of lung function impairment, although the wisdom of this approach has been questioned, 

Previous definitions of COPD differentiated between chronic bronchitis, asthma, and emphysema, acknowledging that there is frequently overlap between these disease entities [12,13]. The GOLD definition of COPD does not differentiate between chronic bronchitis and emphysema but does note that although asthma and COPD can coexist [4], the largely reversible airflow limitation in asthma merits different therapeutic approaches than the largely irreversible airflow limitation of COPD. The overlap of asthma and COPD in a significant proportion of patients has been the focus of recent work [14].

Epidemiology

Prevalence of COPD

The Behavioral Risk Factor Surveillance System (BRFSS) is an ongoing national random-digit-dialed telephone survey of landline and cellphone households designed to measure behavioral risk factors for the noninstitutionalized adult population of the US [15]. An affirmative response to the following question was defined as physician-diagnosed COPD: “Have you ever been told by a doctor or other health professional that you have chronic obstructive pulmonary disease (COPD), emphysema, or bronchitis?”[16]. Based on 2011 BRFSS survey, 13.7 million adults aged ≥ 25 years were estimated to have a self-reported physician diagnosis of COPD in the United States. The greatest age-adjusted prevalence was found to be clustered along the Ohio River Valley and the southern states [16].

The National Health Interview Survey (NHIS) is an annually conducted, nationally representative survey of the civilian noninstitutionalized population aged 18 years and older. A positive response to one or both of the following questions was used to define COPD: “Have you ever been told by a doctor or other health professional that you had emphysema?” and “During the past 12 months, have you been told by a doctor or other health professional that you had chronic bronchitis?” Age-adjusted COPD prevalence estimates showed significant interyear variation during 1999–2011 period, and were higher in women than in men with the highest prevalence noted in 2001 for both genders [16].

The NHIS estimates for COPD have 2 important limitations. First, these estimates depend on the proper recognition and diagnosis of COPD by both the study participants and their health care providers. This would tend to bias the estimates toward counting fewer cases than actually exist. A bias in the opposite direction, however, is that the term chronic bronchitis in this survey is not precisely defined and could be interpreted as recurrent episodes of acute bronchitis. The finding that “chronic bronchitis” has been reported in 3% to 4% of children supports the presence of this potential bias. The second limitation is that this survey is not able to validate, through physiologic evaluation, whether airway obstruction is present or absent.

These limitations were addressed, in part, by separate nationally representative US surveys that include an examination component, such as the National Health and Nutrition Examination Surveys (NHANES) [17]. An analysis of these data from 1988–1994 and 2007–2012 [18] demonstrated that over 70% of people with evidence of obstruction (based on an FEV₁/FVC < 70%) did not have a diagnosis of lung disease (COPD or asthma). In addition, people with evidence of obstruction had a higher risk of mortality whether or not they had diagnosed lung disease [18].

Evaluation of “reversibility” of the airway obstruction requires the administration of bronchodilator, which is not a part of most population-based studies. A subset of participants in the NHANES 2007–2012 survey received a bronchodilator, with a decrease in the estimated prevalence of obstruction from 20.9% to 14.0% [19]. However, a closer look at similar data from a study where all people got a bronchodilator reveal that only a small proportion of people with “reversibility” actually had a significant response to the bronchodilator [20]. In a clinic-based study of subjects with COPD who were aged 69 years and older, 31% demonstrated reversibility, defined as a 15% improvement (from baseline) in FVC and FEV₁ following administration of an inhaled bronchodilator [21]. In this study, subjects with more severe obstruction were more likely to have reversibility but would also be more likely to continue to have diminished lung function after maximum improvement was obtained, thus being classified as having “partial reversibility.”

The presence of significant reversibility or partial reversibility in patients with COPD [15] and nonreversible airflow obstruction in asthma patients [22] demonstrates that these diseases can coexist or, alternatively, that there is overlap and imprecision in the ways that these diseases are clinically diagnosed.

Morbidity and Mortality

COPD is a leading cause of disease morbidity and mortality in the United States. The National Center for Health Statistics (NCHS) conducts ongoing surveillance of several health indicators nationally. The NCHS collects physician office visit data using the National Ambulatory Medical Care Survey [23], emergency department visit data and hospital outpatient data using the National Hospital Ambulatory Medical Care Survey [24], hospitalization data using the National Hospital Discharge Survey [25], and death data using the mortality component of the National Vital Statistics System [26]. The following data include the number and rate of COPD events in adults in the United States (using International Classification of Diseases, 9th Revision, Clinical Modification [ICD-9-CM], codes 490, 491, 492 and 496) in these data sets for the most recent years available.

In 2010, COPD was responsible for an estimated 10.3 million physician office visits, with a resulting age-adjusted rate of 494.8 per 10,000 US civilian population [16]. COPD was also responsible for an estimated 1.5 million emergency room visits, with a resulting age-adjusted rate of 72 visits per 10,000 population [16].

COPD is a leading cause of hospitalization in US adults, particularly in older populations. In 2010, almost 699,000 hospitalizations, were attributed to COPD. The age-adjusted rate of COPD hospitalizations (as the primary cause of hospitalization) was 32.2 per 10,000 population in 2010 [16].

Deaths due to or associated with COPD have not significantly changed since 1999. While the age-adjusted death rate among men declined during 1999–2010 (P = 0.001), the rate among women has not changed significantly (P = 0.127). In 2010, 63, 778 men and 69, 797 women aged ≥ 25 years died of COPD [26]. One of the limitations of using the mortality component of the National Vital Statistics System is that it is based on the underlying cause of death as reported on the death certificate; however, many decedents with COPD listed on the death certificate have their death attributed to another cause [27]. The significance of COPD as a contributor to death is undefined when it is present with diseases more likely to be attributed as the underlying cause of death, such as myocardial infarction or lung cancer [28].

COPD is a very costly disease, with estimated direct medical costs in 2004 of $20.9 billion. The estimated indirect costs related to morbidity (loss of work time and productivity) and premature mortality is an additional $16.3 billion, for a total of $37.2 billion [29]. Because COPD may be present but not listed as the underlying cause of death or the primary reason for hospitalization, these cited estimates may underestimate the true cost of COPD. For example, in another analysis of COPD costs in the US, the total for 2010 was estimated at $32.1 billion [30], but could be up to $100 billion [31] depending on the assumptions surrounding comorbid disease.

Another manifestation of the importance of COPD is its effect on the burden of disease in a population determined using disability-adjusted life-years (DALYs). DALYs for a disease or condition are calculated as the sum of the years of life lost due to premature mortality in the population and the years of life lost due to disability [32]. In 2010, COPD was estimated to be the second leading cause of DALYs lost among the North American population [33]. Worldwide, COPD is expected to move up from being the twelfth leading cause of DALYs lost in 1990 to the fifth leading cause in 2020 [34].

Gender Differences

Smoking-related diseases such as COPD and lung cancer are continuing to increase among women in the United States [35,36], while they have plateaued or are decreasing among men [27,37]. Some evidence has emerged that compared with men at a similar level of tobacco smoking, women may be more likely to develop COPD [38] or that the severity of COPD in women may be increased [39–41].

In the Lung Health Study, which evaluated patients with mild COPD, more women than men demonstrated increased airway responsiveness, although this difference was thought to be related to airway caliber rather than gender [42]. Adult women are more likely to both develop and die of asthma than are men [43–45]. In NHANES III, whereas women reported more physician-diagnosed COPD and asthma than men, men and women had similar rates of decreased lung function, and a similar proportion of both men and women with low lung function had undiagnosed lung disease [3]. The current evidence is inadequate to determine whether women who smoke are more likely to develop COPD or have more severe COPD than men, although this question is being studied by various groups.

Risk Factors and Etiology

Smoking is the dominant risk factor for the development and progression of COPD; however, not all smokers develop COPD, and COPD does occur in persons who have never smoked [1], suggesting that other factors are important in the etiology of COPD. Alpha1-antitrypsin deficiency is an important cause of COPD in a very small percentage of cases [46]. Other undefined genetic factors certainly play an important role in COPD development [38]. The role of infections in both the development and progression of COPD is receiving increased attention, including the role of adenoviral infections in emphysema [47–49].

Occupational and environmental exposures to various pollutants (eg, particulate matter, agricultural dusts) are also important factors in the development of COPD [50,51]. Exposure to indoor air pollutants such as smoke from solid biomass fuels is a major risk factor for COPD especially among women and children in low- and middle-income countries [52,53]. Occupational exposure to fumes and dusts remains an important cause for COPD globally [53,54]. Exposure to outdoor air pollution is associated with a risk of development of COPD as well as exacerbation of the existing disease [53,55].

Clinical Presentation

COPD is heterogeneous in its presentation. Based on data from NHANES III, 44% of patients with severe airflow limitation (FEV₁ < 50% of predicted) may not report symptoms [3]. Among patients with severe airflow limitation who do report symptoms, the symptoms reported most frequently include wheezing (64%) and shortness of breath (65%).

In recent years, COPD has been increasingly recognized as a systemic illness, with effects on nutritional status, muscle wasting, and depression [56–58]. A large proportion of patients probably have components of chronic bronchitis, asthma, and emphysema occurring together. Although some of this overlap may be related to misdiagnosis, some of it may be a measure of the presence of airflow limitation reversibility, as described above. Better defining individuals in these groups may ultimately help tailor better interventions.

Some of the barriers to COPD diagnosis and subsequent treatment often include insufficient knowledge and awareness about COPD especially among primary care physicians, misdiagnosis of COPD as other respiratory diseases such asthma, as well as patient-related barriers involving lack of awareness of early symptoms of COPD and considering them to be related to aging or smoking [59].

Evaluation

The evaluation of a patient with suspected COPD is oriented toward establishing the correct diagnosis and, once this has occurred, determining the extent of the impairment such that therapy can be appropriately targeted.

Components in the evaluation of COPD are listed in Table 3. Every patient with suspected COPD should undergo a thorough history and physical examination. The history should pay particular attention to the following: exposure to risk factors; past history of asthma or allergic disease; family history of COPD; presence of comorbid diseases; effect of disease on the patient’s life, including ability to work and mental health status; and possibilities for reducing risk factors, especially smoking cessation [4]. The physical examination is rarely diagnostic in COPD because most physical abnormalities do not occur until the advanced stages of the disease. Physical examination findings in 

Pulmonary function testing is a critical part of the evaluation of suspected COPD. Whereas most patients with COPD can be managed by a primary care physician, patients with moderate or severe COPD should be evaluated by a specialist [4].

Once the diagnosis of moderate or severe COPD has been established, further testing, including chest radiograph, arterial blood gas determination, screening for α1-antitrypsin deficiency, 6-minute walk testing or exercise oxymetry may be indicated based on the patient’s history and/or clinical findings. Data from computed tomography scans are useful in some advanced cases.

Prognosis of COPD is often influenced by presence of various comorbidities including extrapulmonary, such as osteoporosis, metabolic syndrome, and depression that may be seen as parts of multimorbidity associated with aging [60,61]. Therefore, it is advised to look for comorbidities in COPD patients with any severity of airflow obstruction and treat them accordingly [4].

Therapy for COPD targets reducing risk factors, improving symptoms, and decreasing the risk of exacerbations [10]. Interventions include smoking cessation, vaccinations, decreasing exposures to occupational and environmental pollutants, pulmonary rehabilitation, bronchodilators, and corticosteroids. Select patients with advanced COPD may benefit from other interventions, such as surgical reduction of lung size, lung transplant, the phosphodiesterase inhibitor roflumilast and chronic treatment with antibiotics such as macrolides.

Conclusion

COPD is a common disease that is a leading cause of morbidity and mortality, both in the United States and worldwide. Most cases of COPD are attributable to smoking. Although its incidence among men has plateaued, it continues to increase among women. COPD, particularly in its early stages, is under-diagnosed in the United States. An increased awareness among physicians of the prevalence of mild COPD and the importance of spirometry in diagnosing the disease is important in combating the disease.

Corresponding author: David M. Mannino, MD, Department of Preventive Medicine and Environmental Health, University of Kentucky College of Public Health, 111 Washington Avenue, Lexington, KY 40536, [email protected].

 Financial disclosures: Dr. Mannino has received fees from GlaxoSmithKline, Novartis, AstraZeneca, Sunovion, and Boehringer Ingelheim for advisory board services.

Individuals with rheumatoid arthritis who have antifibrillar collagen type II antibodies appear to have a unique disease phenotype characterized by acute but transient inflammation around the time of diagnosis, as well as a favorable prognosis, according to findings from a retrospective cohort study.

The antifibrillar collagen type II (anti-CII) phenotype also differed from the phenotype of individuals with anti–cyclic citrullinated peptide 2 (anti-CCP2) antibodies who typically exhibit inflammatory markers and higher disease activity later in the disease course, reported Vivek Anand Manivel of Uppsala (Sweden) University and his colleagues (Ann Rheum Dis. 2017 Mar 23. doi: 10.1136/annrheumdis-2016-210873).

Suze777/Thinkstock

[email protected]

Individuals with rheumatoid arthritis who have antifibrillar collagen type II antibodies appear to have a unique disease phenotype characterized by acute but transient inflammation around the time of diagnosis, as well as a favorable prognosis, according to findings from a retrospective cohort study.

The antifibrillar collagen type II (anti-CII) phenotype also differed from the phenotype of individuals with anti–cyclic citrullinated peptide 2 (anti-CCP2) antibodies who typically exhibit inflammatory markers and higher disease activity later in the disease course, reported Vivek Anand Manivel of Uppsala (Sweden) University and his colleagues (Ann Rheum Dis. 2017 Mar 23. doi: 10.1136/annrheumdis-2016-210873).

Suze777/Thinkstock

[email protected]

Individuals with rheumatoid arthritis who have antifibrillar collagen type II antibodies appear to have a unique disease phenotype characterized by acute but transient inflammation around the time of diagnosis, as well as a favorable prognosis, according to findings from a retrospective cohort study.

The antifibrillar collagen type II (anti-CII) phenotype also differed from the phenotype of individuals with anti–cyclic citrullinated peptide 2 (anti-CCP2) antibodies who typically exhibit inflammatory markers and higher disease activity later in the disease course, reported Vivek Anand Manivel of Uppsala (Sweden) University and his colleagues (Ann Rheum Dis. 2017 Mar 23. doi: 10.1136/annrheumdis-2016-210873).

Suze777/Thinkstock

[email protected]

From the Primary Care Institute, Gainesville, FL.

Abstract

• Objective: To provide primary care practitioners with the knowledge required to identify and address older adult suicide risk in their practice.
• Methods: Review of the literature and good clinical practices.
• Results: Primary care practitioners play an important role in older adult suicide prevention and must have knowledge about older adult suicide risk, including risk factors and warning signs in this age-group. Practitioners also must appropriately screen for and manage suicide risk. Older adults, particularly older men, are at high risk for suicide, though they may be less likely to report suicide ideation. Additionally, older adults frequently see primary care practitioners within a month prior to death by suicide. A number of older adult–specific risk factors are reviewed, and appropriate screening and intervention for the primary care setting are discussed.
• Conclusion: Primary care practitioners are uniquely qualified to address a broad range of potential risk factors and should be prepared to identify risk factors and warning signs for older adult suicide, ask appropriate questions to screen for suicide risk, and intervene to prevent suicide.

Key words: suicide; older adults; risk factors; screening; safety planning.

Primary care practitioners play an important role in older adult suicide prevention and have a responsibility to identify and address suicide risk among older adults. To do so, practitioners must understand the problem of older adult suicide, recognize risk factors for suicide in older adults, screen for suicide risk, and appropriately assess and manage suicide risk. Primary care practitioners may face challenges in completing these tasks; the goal of this article is to assist practitioners in addressing these challenges.

Suicide in Older Adults

The United States has recently seen increases in suicide rates across the lifespan; from 1999 to 2014, the suicide rate rose by 24% across all ages [3]. Among both men and women aged 65 to 74, the suicide rate increased in this time period [3]. The high suicide rate among older adults is particularly important to address given the increasing numbers of older adults in the United States. By 2050, the older adult population in the United States is expected to reach 88.5 million, more than double the older adult population in 2010 [4]. Additionally, the generation that is currently aging into older adulthood has historically had higher rates of suicide across their lifespan [5]. Given that suicide rates also increase in older adulthood for men, the coming decades may evidence even higher rates of suicide among older adults than previously and it is critical that older adult suicide prevention becomes a public health priority.

It is also essential to discuss other suicide-related outcomes among older adults, including suicide attempts and suicide ideation. This is critical particularly because the ratio of suicide attempts to deaths by suicide in this age-group is 4 to 1 [1]. This is in contrast to the ratio of attempts to deaths across all ages, which is 25 suicide attempts per death by suicide [1]. This means that suicide prevention must occur before a first suicide attempt is made; suicide attempts cannot be used a marker of elevated suicide risk in older adults or an indication that intervention is needed. Intervention is required prior to suicide risk becoming elevated to the point of a suicide attempt.

It is also critical to recognize that despite the fact that suicide rates rise with age, reports of suicide ideation decrease with age [7,8]. Across all ages, 3.9% of Americans report past-year suicide ideation; however, only 2.7% of older adults report thoughts of suicide [9]. The discrepancy with the increasing rates of death by suicide with age suggest that suicide risk, and thereby opportunities for intervention, may be missed in this age-group [10].

However, older adults may be more willing to report death ideation, as research has found that over 15% of older adults endorse death ideation [11–13]. Death ideation is a desire for death without a specific desire to end one’s own life, and is an important suicide-related outcome, as older adults with death ideation appear the same as those with suicide ideation in terms of depression, hopelessness, and history of suicidal behavior [14]. Additionally, older adults with death ideation had more hospitalizations, more outpatient visits, and more medical issues than older adults with suicide ideation [15]. Therefore, death ideation should be taken as seriously as suicide ideation in older adults [14]. In sum, the high rates of death by suicide, the likelihood of death on a first or early suicide attempt, and the discrepancy between decreasing reports of suicide ideation and increasing rates of death by suicide among older adults indicate that older adult suicide is an important public health problem.

Suicide Prevention Strategies

Many suicide prevention strategies to date have focused on indicated prevention, which concentrates on individuals already identified at high risk (eg, those with suicide ideation or who have made a suicide attempt) [16]. However, because older adults may not report suicide ideation or survive a first suicide attempt, indicated prevention is likely not enough to be effective in older adult suicide prevention. A multilevel suicide prevention strategy [17] is required to prevent older adult suicide [18]. Older adult suicide prevention must include indicated prevention but must also include selective and universal prevention [16]. Selective prevention focuses on groups who may be at risk for suicide (eg, individuals with depression, older adults) and universal prevention focuses on the entire population (eg, interventions to reduce mental health stigma) [16]. To prevent older adult suicide, crisis intervention is critical, but suicide prevention efforts upstream of the development of a suicidal crisis are also essential.

The Importance of Primary Care

Research indicates that primary care is one of the best settings in which to engage in older adult suicide prevention [18]. Older adults are significantly less likely to receive specialty mental health care than younger adults, even when they have depressive symptoms [19]. Additionally, among older adults who died by suicide, 58% had contact with a primary care provider within a month of their deaths, compared to only 11% who had contact with a mental health specialist [20]. Among older adults who died by suicide, 67% saw any provider in the 4 weeks prior to their death [21]. Approximately 10% of older adults saw an outpatient mental health provider, 11% saw a primary care physician for a mental health issue, and 40% saw a primary care physician for a non-mental health issue [21]. Therefore, because older adults are less likely to receive specialty mental health treatment and so often seen a primary care practitioner prior to death by suicide, primary care may be the ideal place for older adult suicide risk to be detected and addressed, especially as many older adults visit primary care without a mental health presenting concern prior to their death by suicide.

Additionally, older adults may be more likely to disclose suicide ideation to primary care practitioners, with whom they are more familiar, than physicians in other settings (eg, emergency departments). Research has shown that familiarity with a primary care physician significantly increases the likelihood of patient disclosure of psychosocial issues to the physician [22]. Primary care providers also have a critical role as care coordinators; many older adults also see specialty physicians and use the emergency department. In fact, older adults are more likely to use the emergency department than younger adults, but emergency departments are not equipped to navigate the complex care needs of this population [23]. Primary care practitioners are important in ensuring that health issues of older adults are addressed by coordinating with specialists, hospitals (eg, inpatient stays, emergency department visits, surgery) and other health services (eg, home health care, physical therapy). Approximately 35% of older adults in the United States experience a lack of care coordination [24], which can negatively impact their health and leave issues such as suicide ideation unaddressed. Primary care practitioners may be critical in screening for mental health issues and suicide risk during even routine visits because of their familiarity with patients, and also play an important role in coordinating care for older adults to improve well-being and to ensure that critical issues, such as suicide ideation, are appropriately addressed.

Primary care practitioners can also be key in upstream prevention. Primary care practitioners are in a unique role to address risk factors for suicide prior to the development of a suicidal crisis. Because older adults frequently see primary care practitioners, such practitioners may have more opportunities to identify risk factors (eg, chronic pain, depression). Primary care practitioners are also trained to treat a broad range of conditions, providing the skills to address many different risk factors.

Finally, primary care is a setting in which screening for depression and suicide ideation among older adults is recommended. The US Preventive Services Task Force recommends screening for depression in all adults and older adults and provides recommended screening instruments, some of which include questions about self-harm or suicide risk [25]. However, this same group has concluded that there is insufficient evidence to support a recommendation for suicide risk screening [26]. Despite this, the Joint Commission recently released an alert that recommends screening for suicide risk in all settings, including primary care [27]. The Joint Commission requirement for ambulatory care that is relevant to suicide is PC.04.01.01: The organization has a process that addresses the patient’s need for continuing care, treatment, or services after discharge or transfer; behavioral health settings have additional suicide-specific requirements. The recommendations, though, go far beyond this requirement for primary care. The Joint Commission specifically notes that primary care clinicians play an important role in detecting suicide ideation and recommends that primary care practitioners review each patient’s history for suicide risk factors, screen all patients for suicide risk, review screenings before patients leave appointments, and take appropriate actions to address suicide risk when needed [27]. Further details are available in the Joint Commission’s Sentinel Event Alert titled, “Detecting and treating suicide ideation in all settings” [27]. Given these recommendations, primary care is an important setting in which to identify and address suicide risk.

Risk Factors for Older Adult Suicide

Numerous reviews exist that cover many risk factors for suicide in older adults [18,28]. This article will focus briefly on risk factors that are likely to be recognized and potentially addressed by primary care practitioners. Risk factors that apply across the lifespan can be recalled through a mnemonic: IS PATH WARM [29]. These risk factors include suicide Ideation, Substance abuse, Purposelessness, Anxiety (including agitation and poor sleep), feeling Trapped, Hopelessness, social Withdrawal, Anger or rage, Recklessness (ie, engaging in risky activities), and Mood changes. The National Suicide Prevention Lifeline also includes being in unbearable physical pain, perceiving one’s self as a burden to others, and seeking revenge on others as risk factors [30]. More specific to older adults, Conwell notes 5 categories or domains of risk factors with strong research support: psychiatric symptoms, somatic illness, functional impairment, social integration, and personality traits and coping [18,31].

Affective or mood disorders, particularly depression and depressive symptoms, are some of the most well-studied and strongest risk factors for older adult suicide [31]; 71% to 97% of all older adults who die by suicide have psychiatric illnesses [28]. Mood disorders, including major depressive episodes, are most consistently linked to older adult suicide risk; there is evidence as well for anxiety disorders and substance abuse disorders as risk factors, though it is somewhat mixed [28]. Therefore, screening for depression, anxiety, and substance abuse may be key to recognizing potential suicide risk. However, depression and anxiety do not present similarly in younger and older adults [32,33]. Depressive symptoms in older adults may be more somatic (eg, agitation, gastrointestinal symptoms) [32] and may reflect more anhedonia than mood changes [33]. Anxiety in older adults tends to be reported as stress or tension, whereas younger adults report feeling anxious or worried [33]. Additionally, substance abuse is often underrecognized, underdiagnosed, and undertreated in older adults [34]. Proactive screening for substance abuse is important as it may not interfere with work or other obligations in older adults, and therefore substance abuse may not be identified by older adults or others in their lives.

Physical illness may also be a risk factor for suicide [28,31]. Numerous diagnoses have been linked to suicide risk, including cancers, neurodegenerative diseases (eg, amyotrophic lateral sclerosis, Huntington disease), spinal cord injury, cardiovascular disease, and pulmonary disease [28,35]. However, overall illness burden (ie, number of chronic illnesses) [28] and self-perceived health [36] appear to be stronger risk factors than any specific illness. Additionally, authors have suggested that illness itself may not be a particularly strong risk factor, but the effect of illness on depressive symptoms [35], functioning, pain, or hopelessness due to the potential for decline over time [28] may increase suicide risk in older adults. Pain itself has been identified as a risk factor for suicide, as have perceptions of burden to others, hopelessness, and functional impairment [28].

In terms of functional impairment, research has shown that impairment in completing instrumental activities of daily living is associated with higher risk for death by suicide, and cognitive impairment may also be associated with elevated suicide risk [28]. However, there are some discrepant findings regarding the role of dementia in suicide risk, which may reflect medical and psychiatric comorbidities, as well as different stages of dementia or levels of cognitive impairment (eg, hopelessness about cognitive decline may increase suicide risk shortly after diagnosis, whereas lack of insight may decrease risk later in the course of the illness) [37]. Related to functional or cognitive impairment is perceived burdensomeness (ie, the perception that one is a liability or burden to others, to the point that others would be better off if one was gone) [38], which may also be associated with suicide risk in older adults [39,40]. Researchers have found that the interaction between perceived burdensomeness and thwarted belongingness (ie, a belief that one lacks reciprocal caring relationships and does not belong) identified older adults who were likely experiencing suicide ideation but did not report it [41]. These findings indicate that perceived burdensomeness and thwarted belongingness may be key in identifying older adults at risk for suicide.

Thwarted belongingness has also been linked to suicide ideation in older adults [41]. In fact, studies suggest that social integration is especially important for reducing suicide risk in this population [28,31,42]. A larger social network, living with others, and being active in the community are each protective against suicide [28]. Bereavement, which can reduce social connectedness and acts as a significant life stressor, is also an important risk factor [31]. Retirement may also reduce social connectedness, and employment changes have been identified as a suicide risk factor for older adults [28]. Retirement has been linked to risk for death by suicide in this population [43], and may not only serve to reduce social connectedness, but for some older adults may also be a significant role loss or loss of sense of purpose that can influence suicide risk.

Finally, rigid personality traits or coping styles are a risk factor for suicide among older adults [28,31]. As older adults face potential losses, health changes, and functional decline, effective positive coping strategies and flexibility are key to maintaining well-being. If older adults are unable to flexibly cope with these challenges, their risk for suicide increases [28].

In addition to risk factors, which confer suicide risk but do not necessarily suggest that an older adult is thinking about suicide, warning signs exist that indicate that suicide risk is imminent. These include suicidal communication (ie, talking or writing about suicide), seeking access to means, and making preparations for suicide (eg, ensuring a will is in place, giving away prized possessions). One important note is that discussing and preparing for death may be developmentally appropriate for older adults, particularly those with chronic illnesses; however, such appropriate preparation is critically different from talking about suicide or a desire for death.

Additionally, a lack of planning for the future may be a warning sign. For example, older adults who decline to schedule medical follow-up or do not wish to refill needed prescriptions may be exhibiting warning signs that should be addressed. Similarly, not following needed medical regimens (eg, an older adult with diabetes no longer taking insulin) is also a warning sign. Other, potentially more subtle warning signs may include significant changes in mood, sleep, or social interactions. Older adults may become agitated and sleep less when they are considering suicide, or may feel more at ease after they have made the decision to die by suicide and their sleep or mood may improve. Withdrawing from valued others may also be a warning sign. Finally, recent major changes (eg, loss of a spouse, moving to an assisted living facility) may be triggers for suicide risk and can serve as warning signs themselves.

Specific Screening Strategies

Given the numerous risk factors and warning signs for older adult suicide, as well as the time limitations that primary care practitioners face [44,45], it would be impractical to comprehensively assess each older adult who presents at a primary care practice. Therefore, more specific screening is necessary. Most importantly, every older adult should be screened for suicide ideation and death ideation at every visit. Screening at every visit is critical because suicide ideation may develop at any point. Previous research has included screening of over 29,000 older adults in 11 primary care settings for suicide ideation, risk of alcohol misuse, and mental health disorders [15], suggesting that suicide risk screening is feasible. Other studies have also successfully used widespread screening for depression and suicide ideation among older adults in primary care [46–48]. Additionally, in an emergency department setting, universal suicide risk screening has been associated with significantly improved risk detection [49], indicating that improved screening may be beneficial in identifying suicide risk. Importantly, asking about suicide does not cause thoughts of suicide [50]. Additionally, it is a myth that those who talk about suicide ideation will not act on these thoughts [51].

When primary care practitioners inquire about suicide ideation, they should also ask about death ideation; though some may believe that death ideation is not as significant in terms of suicide risk as suicide ideation, recall that research has not found differences in previous suicide attempts or current hopelessness among older adults with death ideation versus suicide ideation [14]. Therefore, screening for death ideation should be completed as part of every suicide risk screening.

Screening can take many forms. Screening may be oral; asking an older adult if he or she is having thoughts of suicide or is experiencing a desire to die is a brief, 2-question screening that may provide valuable information (eg, “Are you having thoughts about your own death or wanting to die?”, “Are you having thoughts of killing yourself or thinking about suicide?”). This screening may be conducted by medical assistants, nurses, care managers, or physicians, with the patient’s responses documented. Importantly, a standard procedure should be implemented to ensure older adults are consistently asked about suicide risk at each visit, but do not feel inundated by such questions from numerous staff.

If verbal questions are asked, they must be asked appropriately. Euphemisms or indirect language should not be used during a screening; older adults should be directly asked about thoughts of death and suicide, not simply asked questions such as, “Have you ever had thoughts of harming or hurting yourself?” A question like this does not adequately assess current suicide risk, as it does not assess current thoughts, nor does it specifically inquire about suicide ideation (ie, killing one’s self). It is also important to phrase questions in a manner that invites honest responses and conveys an openness to listening. For example, asking, “You’re not thinking about suicide, are you?” suggests that the practitioner wants the older adult to say no and is not comfortable with the older adult endorsing suicide ideation. Open questions that allow endorsement or denial (eg, “Are you having thoughts of killing yourself?”) imply that the practitioner is receptive to either an endorsement or denial of suicide ideation.

Alternatively, a written screening can be used; older adults may complete a questionnaire prior to their appointment or while waiting to see their practitioner. Such an assessment may be a brief screening (eg, using similar yes/no questions to an oral screening), or may be a standardized measure. For example, the Geriatric Suicide Ideation Scale [52] is a 31-item self-report measure that provides scores for suicide ideation, death ideation, loss of personal and social worth, and perceived meaning in life. Though there are not standard cutoffs that suggest high versus low suicide risk, responses can be reviewed to identify whether older adults are reporting suicide ideation or death ideation, and can also be compared to norms (ie, average scores) from other older adults [52]. This measure also has the benefit of 2 subscales that do not specifically require reporting thoughts of suicide or death (ie, loss of personal and social worth, perceived meaning in life), which may give practitioners an indication of an older adult’s suicide risk even if the older adult is not comfortable disclosing suicide ideation, as has been shown in previous research [7,8].

Similarly, the Geriatric Depression Scale, which has a validated 15-item version [53], does not directly ask about suicide ideation but has a 5-item subscale that has been found to be highly correlated with reported suicide ideation [54]. When administered to older adult primary care patients, this subscale was an effective measure of suicide ideation; a score of ≥ 1 was the best cutoff for determining whether an older adult reported suicide ideation [55].

Additionally, as noted previously, the interaction between perceived burdensomeness and thwarted belongingness may identify older adults who are potentially experiencing, but not reporting, suicide ideation [41]. The Interpersonal Needs Questionnaire [56] is the validated assessment for both perceived burdensomeness and thwarted belongingness. Perceived burdensomeness is assessed via 6 self-report items, and thwarted belongingness is assessed via 9 self-report items on this measure [56]. There are not specific cutoffs that determine high versus low perceived burdensomeness or thwarted belongingness, but older adults’ responses can provide information about their experiences of these constructs. Administration of the Interpersonal Needs Questionnaire can provide information about potential risk for suicide among older adults who may otherwise deny thoughts of suicide or death.

If the screening for suicide ideation or death ideation is positive (ie, the older adult endorses thoughts of suicide or death), the treating primary care practitioner must then follow up with additional questions to determine current level of suicide risk. To make this determination, at a minimum, follow-up questions should focus on whether the older adult has any intent to die by suicide (eg, “Do you have any intent to act on your thoughts of suicide?”), as well as whether he or she has a plan to die by suicide (eg, “Have you begun formulating a plan to die by suicide?”). When asking about a plan, it is important to determine how specific the plan is. For example, an older adult with a specific method identified and date selected to implement the plan is at much higher risk than an older adult with a relatively vague idea. It is also critical to assess for the older adult’s access to means for suicide. If an older adult has a specific plan and has the capability to carry out the plan (eg, plans to overdose on prescription medication and has large quantities of medication or high-lethality medication at home), he or she is more likely to die by suicide than an older adult who does not have access to means (eg, only has small quantities of low-lethality medication available). A general assessment of risk factors and previous suicidal behavior (ie, any previous suicide attempts) also informs decisions about level of risk and interventions.

After a screening or assessment is completed, a risk determination must be made and documented. Acute suicide risk can be categorized as low, moderate, or high. It is not appropriate to say that there is “no” suicide risk present. Low risk occurs when there is no current suicide ideation, no plan to die by suicide, and no intent to act on suicidal thoughts, especially when the patient has no history of suicidal behavior and few risk factors [57]. Moderate risk is evident when there is current suicide ideation, but no specific plan to die by suicide or intent to act on suicidal thoughts. There are likely warning signs or risk factors, which may include previous suicidal behaviors, present in moderate suicide risk [57]. High risk is indicated by current suicide ideation with plan to die by suicide and suicidal intent. There are significant warning signs and risk factors present; there may also be a recent suicide attempt, though this is not a requirement for a high risk determination [57]. Undetermined suicide risk occurs when a practitioner cannot accurately assess risk, but concern regarding suicide is present; this is primarily used when a patient refuses to answer questions about suicide. Undetermined risk should be treated as at least moderate risk. Because research shows that death ideation has similar outcomes to suicide ideation in older adults [14], death ideation should also be factored into determinations of suicide risk; reports of death ideation may indicate low or moderate risk in older adults, dependent upon other risk factors, suicidal intent, and plan.

After a risk determination is made, it must be documented in the medical record. The level of risk and rationale for that determination must be included [58]. Stating only the level of risk without a rationale (ie, the older adult’s responses to questions) is not adequate, and documenting only the older adult’s responses without a determination of risk is also not sufficient. Finally, it is critical to document the intervention that occurred or steps taken after the level of risk was determined.

Critically, stating only that there was no indication of suicide risk is inadequate. For example, documenting “No evidence of suicide risk” is not appropriate. This documentation does not indicate that the older adult was specifically asked about suicide ideation, death ideation, suicidal intent, or plan to die by suicide. It also does not indicate a level of suicide risk. Examples of appropriate documentation include:

 Patient was asked about suicide risk. She denied current suicide ideation but reported death ideation. She denied any current suicidal intent or plan. She also denied any previous suicide attempts. Therefore, acute suicide risk was deemed to be low. Provided patient with wallet card about the National Suicide Prevention Lifeline. Also called the Friendship Line while in the room with the patient to connect her with services. Finally, provided a brief list of local mental health professionals to patient; the patient reported she would like to see Dr. Smith. Called and left a message for Dr. Smith with referral information with patient during appointment.

Patient was asked about suicide risk. He reported both death ideation and suicide ideation. He also reported a nonspecific plan (ie, causing a single-vehicle motor vehicle accident, with no specific plan for the motor vehicle accident or timeframe) and denied any intent to act on his thoughts of suicide. He reported one previous suicide attempt, at age 22, by overdose on over-the-counter medication. He reported that this attempt did not require medical attention. Therefore, acute suicide risk was determined to be moderate. Patient was introduced to the behavioral health specialist, who met with the patient during the appointment to conduct further assessment and intervention.

Specific Intervention Strategies

Despite the fact that the pace of the primary care setting often does not allow for time-intensive intervention, there are ways to address suicide risk in this setting. Importantly, no-suicide contracts should not be used at any time [59,60]. No-suicide contracts are documents that patients who are experiencing suicide ideation are required to sign that state that they will not die by suicide while under the care of the practitioner. These contracts have no evidence of effectiveness, and some researchers argue that they may in fact damage the relationship with patients and serve the practitioner’s needs more than the patient’s needs [59].

One of the best options for older adults at low acute suicide risk is to provide resources and referrals. The National Suicide Prevention Lifeline can be reached at 1-800-273-TALK (8255); trained counselors are available to speak to patients at all times. Wallet cards with information about the National Suicide Prevention Lifeline are available at no charge from the US Substance Abuse and Mental Health Services Administration online store. The Friendship Line is another service available free to adults ages 60 and older, 24 hours per day, 7 days per week; this line can be reached at 1-800-971-0016. The Friendship Line, which is managed by the Institute on Aging, also provides outreach calls to older adults who may be isolated or lonely, increasing connectedness and potentially reducing suicide risk.

Having a ready list of local mental health professionals with expertise in geriatrics and suicide risk to provide to the patient is also beneficial. Recall, though, that older adults are less likely to seek out and receive mental health services [19]; therefore, connecting the patient with resources or referrals during the appointment is critical. If the practitioner does not have time to do this, having a medical assistant or other staff member that the patient knows engage in this step may be appropriate. For example, the patient can call the Friendship Line or National Suicide Prevention Lifeline while in the room with the practitioner, which may reduce anxiety or stigma about doing so and connect the patient with services. Similarly, calling a local mental health professional to make a referral during the appointment may increase the likelihood that the older adult will follow up on the referral.

The most ideal method of intervention for moderate or high acute suicide risk is a warm handoff to a behavioral or mental health specialist. As primary care and behavioral health become more integrated and financially viable as reimbursement through the Centers for Medicare and Medicaid Services improves [61], it is becoming increasingly likely that such a specialist will be on-site and available. Research has found that collaborative care in primary care reduces suicide risk in older adults [46–48,62]. Mental health specialists can conduct more comprehensive assessments and spend more time intervening to reduce suicide risk among older adults with death or suicide ideation. If an on-site behavioral health specialist is not available, older adults at high suicide risk may need to be referred to an emergency department for further evaluation and follow-up. Each state has its own laws and procedures regarding this process, which should be incorporated into a practice’s procedures for addressing high suicide risk. The procedure often involves ensuring that the older adult is accompanied at all times (ie, not left alone in a room), alerting emergency services (usually via phone call to an emergency line, such as 911), and completion of paperwork by a practitioner asserting that the patient is a danger to self. Police or other emergency personnel are then responsible for transporting the patient for further evaluation and determination of whether hospitalization is required.

If more time is available, either via the treating primary care practitioner or other patient care staff in the office, other brief interventions may be beneficial. First, means safety discussions are critical, particularly for older adults with plans for suicide or access to highly lethal means. In such discussions, patients are encouraged to restrict access to the methods that they may use to die by suicide. Plans for restricting access are developed, and when possible, a support person is enlisted to ensure that the plans are carried out. For example, if an older adult has access to firearms (eg, keeps a loaded weapon in his or her nightstand), he or she is encouraged to restrict his or her access to it. Ideally, this is through removing the weapon from the home, either permanently or until suicide risk reduces (eg, giving it to a friend, turning it over to police), but more safe storage may also be an option if the older adult is not willing to remove the weapon from the home. This may mean using a gun lock or storing the weapon in a gun safe, storing ammunition separately from an unloaded weapon, removing the firing pin, or otherwise disassembling the weapon. Means safety counseling has been shown to be effective in reducing suicide rates [63] and is acceptable to patients [64]. Studies indicate that over 90% of individuals who make a suicide attempt and survive do not go on to die by suicide [65]; therefore, reducing access to highly lethal means during a suicidal crisis may be key in reducing suicide rates. Though an in-depth review of means safety counseling is outside the scope of this article, readers are directed to Bryan, Stone, and Rudd’s article for a practical overview of means safety discussions [66].

Second, safety planning is a brief intervention that may be beneficial in the primary care setting [67,68]. The goal of a safety plan is to create an individualized plan to remain safe during a suicidal crisis. Means safety discussion is the last of 6 steps in the safety plan [68]. The first 5 steps include identifying warning signs, using internal coping strategies, social connectedness as distraction, social support for the crisis, and professionals that can be used as resources. When patients can identify specific, individualized warning signs that occur prior to a crisis, they can then use strategies to cope and prevent the crisis from worsening. Coping strategies that are encouraged are first internal (ie, those that can be done without relying on anyone else), such as exercise or journaling. If those do not improve the patient’s mood, then he or she is encouraged to use people or social settings as a distraction (eg, people watching at the mall, calling an acquaintance to chat), and if he or she is still feeling bad, encouraged to get social support for the crisis (eg, calling a family member to discuss the crisis and get support). Finally, if all of these steps are not effective, the older adult is encouraged to reach out to professional supports, such as a mental health provider, the National Suicide Prevention Lifeline, or 911 (or go to an emergency room). Readers are encouraged to review Stanley and Brown’s articles for comprehensive details about safety planning as an intervention [67,68]. Additionally, an article with specific adaptations for safety planning with older adults is forthcoming [69].

Conclusion

Older adults, particularly older men, are at high risk for suicide [1,2], and primary care practitioners are a critical component of older adult suicide prevention. Older adults frequently see primary care practitioners within a month prior to death by suicide [20,21]; primary care practitioners are uniquely qualified to address a broad range of potential risk factors, and may have more interactions and familiarity with older adults at risk for suicide than other medical professionals [20–22]. Primary care practitioners should be prepared to identify risk factors and warning signs for older adult suicide, ask appropriate questions to screen for suicide risk, and intervene to prevent suicide. Screening can consist of standardized written questionnaires or oral questioning, and interventions may include providing resources and referrals, discussions about means safety, safety planning, and handoff to a mental health specialist. Interventions for suicide risk are likely feasible and acceptable in primary care [46–48]. Primary care practitioners have an important role to play in older adult suicide prevention, and must be prepared to interact with older adults who may be at risk for suicide.

Corresponding author: Danielle R. Jahn, PhD, Primary Care Institute, 605 NE 1st St, Gainesville, FL 32605, [email protected].

Financial disclosures: None reported.

From the Primary Care Institute, Gainesville, FL.

Abstract

• Objective: To provide primary care practitioners with the knowledge required to identify and address older adult suicide risk in their practice.
• Methods: Review of the literature and good clinical practices.
• Results: Primary care practitioners play an important role in older adult suicide prevention and must have knowledge about older adult suicide risk, including risk factors and warning signs in this age-group. Practitioners also must appropriately screen for and manage suicide risk. Older adults, particularly older men, are at high risk for suicide, though they may be less likely to report suicide ideation. Additionally, older adults frequently see primary care practitioners within a month prior to death by suicide. A number of older adult–specific risk factors are reviewed, and appropriate screening and intervention for the primary care setting are discussed.
• Conclusion: Primary care practitioners are uniquely qualified to address a broad range of potential risk factors and should be prepared to identify risk factors and warning signs for older adult suicide, ask appropriate questions to screen for suicide risk, and intervene to prevent suicide.

Key words: suicide; older adults; risk factors; screening; safety planning.

Primary care practitioners play an important role in older adult suicide prevention and have a responsibility to identify and address suicide risk among older adults. To do so, practitioners must understand the problem of older adult suicide, recognize risk factors for suicide in older adults, screen for suicide risk, and appropriately assess and manage suicide risk. Primary care practitioners may face challenges in completing these tasks; the goal of this article is to assist practitioners in addressing these challenges.

Suicide in Older Adults

The United States has recently seen increases in suicide rates across the lifespan; from 1999 to 2014, the suicide rate rose by 24% across all ages [3]. Among both men and women aged 65 to 74, the suicide rate increased in this time period [3]. The high suicide rate among older adults is particularly important to address given the increasing numbers of older adults in the United States. By 2050, the older adult population in the United States is expected to reach 88.5 million, more than double the older adult population in 2010 [4]. Additionally, the generation that is currently aging into older adulthood has historically had higher rates of suicide across their lifespan [5]. Given that suicide rates also increase in older adulthood for men, the coming decades may evidence even higher rates of suicide among older adults than previously and it is critical that older adult suicide prevention becomes a public health priority.

It is also essential to discuss other suicide-related outcomes among older adults, including suicide attempts and suicide ideation. This is critical particularly because the ratio of suicide attempts to deaths by suicide in this age-group is 4 to 1 [1]. This is in contrast to the ratio of attempts to deaths across all ages, which is 25 suicide attempts per death by suicide [1]. This means that suicide prevention must occur before a first suicide attempt is made; suicide attempts cannot be used a marker of elevated suicide risk in older adults or an indication that intervention is needed. Intervention is required prior to suicide risk becoming elevated to the point of a suicide attempt.

It is also critical to recognize that despite the fact that suicide rates rise with age, reports of suicide ideation decrease with age [7,8]. Across all ages, 3.9% of Americans report past-year suicide ideation; however, only 2.7% of older adults report thoughts of suicide [9]. The discrepancy with the increasing rates of death by suicide with age suggest that suicide risk, and thereby opportunities for intervention, may be missed in this age-group [10].

However, older adults may be more willing to report death ideation, as research has found that over 15% of older adults endorse death ideation [11–13]. Death ideation is a desire for death without a specific desire to end one’s own life, and is an important suicide-related outcome, as older adults with death ideation appear the same as those with suicide ideation in terms of depression, hopelessness, and history of suicidal behavior [14]. Additionally, older adults with death ideation had more hospitalizations, more outpatient visits, and more medical issues than older adults with suicide ideation [15]. Therefore, death ideation should be taken as seriously as suicide ideation in older adults [14]. In sum, the high rates of death by suicide, the likelihood of death on a first or early suicide attempt, and the discrepancy between decreasing reports of suicide ideation and increasing rates of death by suicide among older adults indicate that older adult suicide is an important public health problem.

Suicide Prevention Strategies

Many suicide prevention strategies to date have focused on indicated prevention, which concentrates on individuals already identified at high risk (eg, those with suicide ideation or who have made a suicide attempt) [16]. However, because older adults may not report suicide ideation or survive a first suicide attempt, indicated prevention is likely not enough to be effective in older adult suicide prevention. A multilevel suicide prevention strategy [17] is required to prevent older adult suicide [18]. Older adult suicide prevention must include indicated prevention but must also include selective and universal prevention [16]. Selective prevention focuses on groups who may be at risk for suicide (eg, individuals with depression, older adults) and universal prevention focuses on the entire population (eg, interventions to reduce mental health stigma) [16]. To prevent older adult suicide, crisis intervention is critical, but suicide prevention efforts upstream of the development of a suicidal crisis are also essential.

The Importance of Primary Care

Research indicates that primary care is one of the best settings in which to engage in older adult suicide prevention [18]. Older adults are significantly less likely to receive specialty mental health care than younger adults, even when they have depressive symptoms [19]. Additionally, among older adults who died by suicide, 58% had contact with a primary care provider within a month of their deaths, compared to only 11% who had contact with a mental health specialist [20]. Among older adults who died by suicide, 67% saw any provider in the 4 weeks prior to their death [21]. Approximately 10% of older adults saw an outpatient mental health provider, 11% saw a primary care physician for a mental health issue, and 40% saw a primary care physician for a non-mental health issue [21]. Therefore, because older adults are less likely to receive specialty mental health treatment and so often seen a primary care practitioner prior to death by suicide, primary care may be the ideal place for older adult suicide risk to be detected and addressed, especially as many older adults visit primary care without a mental health presenting concern prior to their death by suicide.

Additionally, older adults may be more likely to disclose suicide ideation to primary care practitioners, with whom they are more familiar, than physicians in other settings (eg, emergency departments). Research has shown that familiarity with a primary care physician significantly increases the likelihood of patient disclosure of psychosocial issues to the physician [22]. Primary care providers also have a critical role as care coordinators; many older adults also see specialty physicians and use the emergency department. In fact, older adults are more likely to use the emergency department than younger adults, but emergency departments are not equipped to navigate the complex care needs of this population [23]. Primary care practitioners are important in ensuring that health issues of older adults are addressed by coordinating with specialists, hospitals (eg, inpatient stays, emergency department visits, surgery) and other health services (eg, home health care, physical therapy). Approximately 35% of older adults in the United States experience a lack of care coordination [24], which can negatively impact their health and leave issues such as suicide ideation unaddressed. Primary care practitioners may be critical in screening for mental health issues and suicide risk during even routine visits because of their familiarity with patients, and also play an important role in coordinating care for older adults to improve well-being and to ensure that critical issues, such as suicide ideation, are appropriately addressed.

Primary care practitioners can also be key in upstream prevention. Primary care practitioners are in a unique role to address risk factors for suicide prior to the development of a suicidal crisis. Because older adults frequently see primary care practitioners, such practitioners may have more opportunities to identify risk factors (eg, chronic pain, depression). Primary care practitioners are also trained to treat a broad range of conditions, providing the skills to address many different risk factors.

Finally, primary care is a setting in which screening for depression and suicide ideation among older adults is recommended. The US Preventive Services Task Force recommends screening for depression in all adults and older adults and provides recommended screening instruments, some of which include questions about self-harm or suicide risk [25]. However, this same group has concluded that there is insufficient evidence to support a recommendation for suicide risk screening [26]. Despite this, the Joint Commission recently released an alert that recommends screening for suicide risk in all settings, including primary care [27]. The Joint Commission requirement for ambulatory care that is relevant to suicide is PC.04.01.01: The organization has a process that addresses the patient’s need for continuing care, treatment, or services after discharge or transfer; behavioral health settings have additional suicide-specific requirements. The recommendations, though, go far beyond this requirement for primary care. The Joint Commission specifically notes that primary care clinicians play an important role in detecting suicide ideation and recommends that primary care practitioners review each patient’s history for suicide risk factors, screen all patients for suicide risk, review screenings before patients leave appointments, and take appropriate actions to address suicide risk when needed [27]. Further details are available in the Joint Commission’s Sentinel Event Alert titled, “Detecting and treating suicide ideation in all settings” [27]. Given these recommendations, primary care is an important setting in which to identify and address suicide risk.

Risk Factors for Older Adult Suicide

Numerous reviews exist that cover many risk factors for suicide in older adults [18,28]. This article will focus briefly on risk factors that are likely to be recognized and potentially addressed by primary care practitioners. Risk factors that apply across the lifespan can be recalled through a mnemonic: IS PATH WARM [29]. These risk factors include suicide Ideation, Substance abuse, Purposelessness, Anxiety (including agitation and poor sleep), feeling Trapped, Hopelessness, social Withdrawal, Anger or rage, Recklessness (ie, engaging in risky activities), and Mood changes. The National Suicide Prevention Lifeline also includes being in unbearable physical pain, perceiving one’s self as a burden to others, and seeking revenge on others as risk factors [30]. More specific to older adults, Conwell notes 5 categories or domains of risk factors with strong research support: psychiatric symptoms, somatic illness, functional impairment, social integration, and personality traits and coping [18,31].

Affective or mood disorders, particularly depression and depressive symptoms, are some of the most well-studied and strongest risk factors for older adult suicide [31]; 71% to 97% of all older adults who die by suicide have psychiatric illnesses [28]. Mood disorders, including major depressive episodes, are most consistently linked to older adult suicide risk; there is evidence as well for anxiety disorders and substance abuse disorders as risk factors, though it is somewhat mixed [28]. Therefore, screening for depression, anxiety, and substance abuse may be key to recognizing potential suicide risk. However, depression and anxiety do not present similarly in younger and older adults [32,33]. Depressive symptoms in older adults may be more somatic (eg, agitation, gastrointestinal symptoms) [32] and may reflect more anhedonia than mood changes [33]. Anxiety in older adults tends to be reported as stress or tension, whereas younger adults report feeling anxious or worried [33]. Additionally, substance abuse is often underrecognized, underdiagnosed, and undertreated in older adults [34]. Proactive screening for substance abuse is important as it may not interfere with work or other obligations in older adults, and therefore substance abuse may not be identified by older adults or others in their lives.

Physical illness may also be a risk factor for suicide [28,31]. Numerous diagnoses have been linked to suicide risk, including cancers, neurodegenerative diseases (eg, amyotrophic lateral sclerosis, Huntington disease), spinal cord injury, cardiovascular disease, and pulmonary disease [28,35]. However, overall illness burden (ie, number of chronic illnesses) [28] and self-perceived health [36] appear to be stronger risk factors than any specific illness. Additionally, authors have suggested that illness itself may not be a particularly strong risk factor, but the effect of illness on depressive symptoms [35], functioning, pain, or hopelessness due to the potential for decline over time [28] may increase suicide risk in older adults. Pain itself has been identified as a risk factor for suicide, as have perceptions of burden to others, hopelessness, and functional impairment [28].

In terms of functional impairment, research has shown that impairment in completing instrumental activities of daily living is associated with higher risk for death by suicide, and cognitive impairment may also be associated with elevated suicide risk [28]. However, there are some discrepant findings regarding the role of dementia in suicide risk, which may reflect medical and psychiatric comorbidities, as well as different stages of dementia or levels of cognitive impairment (eg, hopelessness about cognitive decline may increase suicide risk shortly after diagnosis, whereas lack of insight may decrease risk later in the course of the illness) [37]. Related to functional or cognitive impairment is perceived burdensomeness (ie, the perception that one is a liability or burden to others, to the point that others would be better off if one was gone) [38], which may also be associated with suicide risk in older adults [39,40]. Researchers have found that the interaction between perceived burdensomeness and thwarted belongingness (ie, a belief that one lacks reciprocal caring relationships and does not belong) identified older adults who were likely experiencing suicide ideation but did not report it [41]. These findings indicate that perceived burdensomeness and thwarted belongingness may be key in identifying older adults at risk for suicide.

Thwarted belongingness has also been linked to suicide ideation in older adults [41]. In fact, studies suggest that social integration is especially important for reducing suicide risk in this population [28,31,42]. A larger social network, living with others, and being active in the community are each protective against suicide [28]. Bereavement, which can reduce social connectedness and acts as a significant life stressor, is also an important risk factor [31]. Retirement may also reduce social connectedness, and employment changes have been identified as a suicide risk factor for older adults [28]. Retirement has been linked to risk for death by suicide in this population [43], and may not only serve to reduce social connectedness, but for some older adults may also be a significant role loss or loss of sense of purpose that can influence suicide risk.

Finally, rigid personality traits or coping styles are a risk factor for suicide among older adults [28,31]. As older adults face potential losses, health changes, and functional decline, effective positive coping strategies and flexibility are key to maintaining well-being. If older adults are unable to flexibly cope with these challenges, their risk for suicide increases [28].

In addition to risk factors, which confer suicide risk but do not necessarily suggest that an older adult is thinking about suicide, warning signs exist that indicate that suicide risk is imminent. These include suicidal communication (ie, talking or writing about suicide), seeking access to means, and making preparations for suicide (eg, ensuring a will is in place, giving away prized possessions). One important note is that discussing and preparing for death may be developmentally appropriate for older adults, particularly those with chronic illnesses; however, such appropriate preparation is critically different from talking about suicide or a desire for death.

Additionally, a lack of planning for the future may be a warning sign. For example, older adults who decline to schedule medical follow-up or do not wish to refill needed prescriptions may be exhibiting warning signs that should be addressed. Similarly, not following needed medical regimens (eg, an older adult with diabetes no longer taking insulin) is also a warning sign. Other, potentially more subtle warning signs may include significant changes in mood, sleep, or social interactions. Older adults may become agitated and sleep less when they are considering suicide, or may feel more at ease after they have made the decision to die by suicide and their sleep or mood may improve. Withdrawing from valued others may also be a warning sign. Finally, recent major changes (eg, loss of a spouse, moving to an assisted living facility) may be triggers for suicide risk and can serve as warning signs themselves.

Specific Screening Strategies

Given the numerous risk factors and warning signs for older adult suicide, as well as the time limitations that primary care practitioners face [44,45], it would be impractical to comprehensively assess each older adult who presents at a primary care practice. Therefore, more specific screening is necessary. Most importantly, every older adult should be screened for suicide ideation and death ideation at every visit. Screening at every visit is critical because suicide ideation may develop at any point. Previous research has included screening of over 29,000 older adults in 11 primary care settings for suicide ideation, risk of alcohol misuse, and mental health disorders [15], suggesting that suicide risk screening is feasible. Other studies have also successfully used widespread screening for depression and suicide ideation among older adults in primary care [46–48]. Additionally, in an emergency department setting, universal suicide risk screening has been associated with significantly improved risk detection [49], indicating that improved screening may be beneficial in identifying suicide risk. Importantly, asking about suicide does not cause thoughts of suicide [50]. Additionally, it is a myth that those who talk about suicide ideation will not act on these thoughts [51].

When primary care practitioners inquire about suicide ideation, they should also ask about death ideation; though some may believe that death ideation is not as significant in terms of suicide risk as suicide ideation, recall that research has not found differences in previous suicide attempts or current hopelessness among older adults with death ideation versus suicide ideation [14]. Therefore, screening for death ideation should be completed as part of every suicide risk screening.

Screening can take many forms. Screening may be oral; asking an older adult if he or she is having thoughts of suicide or is experiencing a desire to die is a brief, 2-question screening that may provide valuable information (eg, “Are you having thoughts about your own death or wanting to die?”, “Are you having thoughts of killing yourself or thinking about suicide?”). This screening may be conducted by medical assistants, nurses, care managers, or physicians, with the patient’s responses documented. Importantly, a standard procedure should be implemented to ensure older adults are consistently asked about suicide risk at each visit, but do not feel inundated by such questions from numerous staff.

If verbal questions are asked, they must be asked appropriately. Euphemisms or indirect language should not be used during a screening; older adults should be directly asked about thoughts of death and suicide, not simply asked questions such as, “Have you ever had thoughts of harming or hurting yourself?” A question like this does not adequately assess current suicide risk, as it does not assess current thoughts, nor does it specifically inquire about suicide ideation (ie, killing one’s self). It is also important to phrase questions in a manner that invites honest responses and conveys an openness to listening. For example, asking, “You’re not thinking about suicide, are you?” suggests that the practitioner wants the older adult to say no and is not comfortable with the older adult endorsing suicide ideation. Open questions that allow endorsement or denial (eg, “Are you having thoughts of killing yourself?”) imply that the practitioner is receptive to either an endorsement or denial of suicide ideation.

Alternatively, a written screening can be used; older adults may complete a questionnaire prior to their appointment or while waiting to see their practitioner. Such an assessment may be a brief screening (eg, using similar yes/no questions to an oral screening), or may be a standardized measure. For example, the Geriatric Suicide Ideation Scale [52] is a 31-item self-report measure that provides scores for suicide ideation, death ideation, loss of personal and social worth, and perceived meaning in life. Though there are not standard cutoffs that suggest high versus low suicide risk, responses can be reviewed to identify whether older adults are reporting suicide ideation or death ideation, and can also be compared to norms (ie, average scores) from other older adults [52]. This measure also has the benefit of 2 subscales that do not specifically require reporting thoughts of suicide or death (ie, loss of personal and social worth, perceived meaning in life), which may give practitioners an indication of an older adult’s suicide risk even if the older adult is not comfortable disclosing suicide ideation, as has been shown in previous research [7,8].

Similarly, the Geriatric Depression Scale, which has a validated 15-item version [53], does not directly ask about suicide ideation but has a 5-item subscale that has been found to be highly correlated with reported suicide ideation [54]. When administered to older adult primary care patients, this subscale was an effective measure of suicide ideation; a score of ≥ 1 was the best cutoff for determining whether an older adult reported suicide ideation [55].

Additionally, as noted previously, the interaction between perceived burdensomeness and thwarted belongingness may identify older adults who are potentially experiencing, but not reporting, suicide ideation [41]. The Interpersonal Needs Questionnaire [56] is the validated assessment for both perceived burdensomeness and thwarted belongingness. Perceived burdensomeness is assessed via 6 self-report items, and thwarted belongingness is assessed via 9 self-report items on this measure [56]. There are not specific cutoffs that determine high versus low perceived burdensomeness or thwarted belongingness, but older adults’ responses can provide information about their experiences of these constructs. Administration of the Interpersonal Needs Questionnaire can provide information about potential risk for suicide among older adults who may otherwise deny thoughts of suicide or death.

If the screening for suicide ideation or death ideation is positive (ie, the older adult endorses thoughts of suicide or death), the treating primary care practitioner must then follow up with additional questions to determine current level of suicide risk. To make this determination, at a minimum, follow-up questions should focus on whether the older adult has any intent to die by suicide (eg, “Do you have any intent to act on your thoughts of suicide?”), as well as whether he or she has a plan to die by suicide (eg, “Have you begun formulating a plan to die by suicide?”). When asking about a plan, it is important to determine how specific the plan is. For example, an older adult with a specific method identified and date selected to implement the plan is at much higher risk than an older adult with a relatively vague idea. It is also critical to assess for the older adult’s access to means for suicide. If an older adult has a specific plan and has the capability to carry out the plan (eg, plans to overdose on prescription medication and has large quantities of medication or high-lethality medication at home), he or she is more likely to die by suicide than an older adult who does not have access to means (eg, only has small quantities of low-lethality medication available). A general assessment of risk factors and previous suicidal behavior (ie, any previous suicide attempts) also informs decisions about level of risk and interventions.

After a screening or assessment is completed, a risk determination must be made and documented. Acute suicide risk can be categorized as low, moderate, or high. It is not appropriate to say that there is “no” suicide risk present. Low risk occurs when there is no current suicide ideation, no plan to die by suicide, and no intent to act on suicidal thoughts, especially when the patient has no history of suicidal behavior and few risk factors [57]. Moderate risk is evident when there is current suicide ideation, but no specific plan to die by suicide or intent to act on suicidal thoughts. There are likely warning signs or risk factors, which may include previous suicidal behaviors, present in moderate suicide risk [57]. High risk is indicated by current suicide ideation with plan to die by suicide and suicidal intent. There are significant warning signs and risk factors present; there may also be a recent suicide attempt, though this is not a requirement for a high risk determination [57]. Undetermined suicide risk occurs when a practitioner cannot accurately assess risk, but concern regarding suicide is present; this is primarily used when a patient refuses to answer questions about suicide. Undetermined risk should be treated as at least moderate risk. Because research shows that death ideation has similar outcomes to suicide ideation in older adults [14], death ideation should also be factored into determinations of suicide risk; reports of death ideation may indicate low or moderate risk in older adults, dependent upon other risk factors, suicidal intent, and plan.

After a risk determination is made, it must be documented in the medical record. The level of risk and rationale for that determination must be included [58]. Stating only the level of risk without a rationale (ie, the older adult’s responses to questions) is not adequate, and documenting only the older adult’s responses without a determination of risk is also not sufficient. Finally, it is critical to document the intervention that occurred or steps taken after the level of risk was determined.

Critically, stating only that there was no indication of suicide risk is inadequate. For example, documenting “No evidence of suicide risk” is not appropriate. This documentation does not indicate that the older adult was specifically asked about suicide ideation, death ideation, suicidal intent, or plan to die by suicide. It also does not indicate a level of suicide risk. Examples of appropriate documentation include:

 Patient was asked about suicide risk. She denied current suicide ideation but reported death ideation. She denied any current suicidal intent or plan. She also denied any previous suicide attempts. Therefore, acute suicide risk was deemed to be low. Provided patient with wallet card about the National Suicide Prevention Lifeline. Also called the Friendship Line while in the room with the patient to connect her with services. Finally, provided a brief list of local mental health professionals to patient; the patient reported she would like to see Dr. Smith. Called and left a message for Dr. Smith with referral information with patient during appointment.

Patient was asked about suicide risk. He reported both death ideation and suicide ideation. He also reported a nonspecific plan (ie, causing a single-vehicle motor vehicle accident, with no specific plan for the motor vehicle accident or timeframe) and denied any intent to act on his thoughts of suicide. He reported one previous suicide attempt, at age 22, by overdose on over-the-counter medication. He reported that this attempt did not require medical attention. Therefore, acute suicide risk was determined to be moderate. Patient was introduced to the behavioral health specialist, who met with the patient during the appointment to conduct further assessment and intervention.

Specific Intervention Strategies

Despite the fact that the pace of the primary care setting often does not allow for time-intensive intervention, there are ways to address suicide risk in this setting. Importantly, no-suicide contracts should not be used at any time [59,60]. No-suicide contracts are documents that patients who are experiencing suicide ideation are required to sign that state that they will not die by suicide while under the care of the practitioner. These contracts have no evidence of effectiveness, and some researchers argue that they may in fact damage the relationship with patients and serve the practitioner’s needs more than the patient’s needs [59].

One of the best options for older adults at low acute suicide risk is to provide resources and referrals. The National Suicide Prevention Lifeline can be reached at 1-800-273-TALK (8255); trained counselors are available to speak to patients at all times. Wallet cards with information about the National Suicide Prevention Lifeline are available at no charge from the US Substance Abuse and Mental Health Services Administration online store. The Friendship Line is another service available free to adults ages 60 and older, 24 hours per day, 7 days per week; this line can be reached at 1-800-971-0016. The Friendship Line, which is managed by the Institute on Aging, also provides outreach calls to older adults who may be isolated or lonely, increasing connectedness and potentially reducing suicide risk.

Having a ready list of local mental health professionals with expertise in geriatrics and suicide risk to provide to the patient is also beneficial. Recall, though, that older adults are less likely to seek out and receive mental health services [19]; therefore, connecting the patient with resources or referrals during the appointment is critical. If the practitioner does not have time to do this, having a medical assistant or other staff member that the patient knows engage in this step may be appropriate. For example, the patient can call the Friendship Line or National Suicide Prevention Lifeline while in the room with the practitioner, which may reduce anxiety or stigma about doing so and connect the patient with services. Similarly, calling a local mental health professional to make a referral during the appointment may increase the likelihood that the older adult will follow up on the referral.

The most ideal method of intervention for moderate or high acute suicide risk is a warm handoff to a behavioral or mental health specialist. As primary care and behavioral health become more integrated and financially viable as reimbursement through the Centers for Medicare and Medicaid Services improves [61], it is becoming increasingly likely that such a specialist will be on-site and available. Research has found that collaborative care in primary care reduces suicide risk in older adults [46–48,62]. Mental health specialists can conduct more comprehensive assessments and spend more time intervening to reduce suicide risk among older adults with death or suicide ideation. If an on-site behavioral health specialist is not available, older adults at high suicide risk may need to be referred to an emergency department for further evaluation and follow-up. Each state has its own laws and procedures regarding this process, which should be incorporated into a practice’s procedures for addressing high suicide risk. The procedure often involves ensuring that the older adult is accompanied at all times (ie, not left alone in a room), alerting emergency services (usually via phone call to an emergency line, such as 911), and completion of paperwork by a practitioner asserting that the patient is a danger to self. Police or other emergency personnel are then responsible for transporting the patient for further evaluation and determination of whether hospitalization is required.

If more time is available, either via the treating primary care practitioner or other patient care staff in the office, other brief interventions may be beneficial. First, means safety discussions are critical, particularly for older adults with plans for suicide or access to highly lethal means. In such discussions, patients are encouraged to restrict access to the methods that they may use to die by suicide. Plans for restricting access are developed, and when possible, a support person is enlisted to ensure that the plans are carried out. For example, if an older adult has access to firearms (eg, keeps a loaded weapon in his or her nightstand), he or she is encouraged to restrict his or her access to it. Ideally, this is through removing the weapon from the home, either permanently or until suicide risk reduces (eg, giving it to a friend, turning it over to police), but more safe storage may also be an option if the older adult is not willing to remove the weapon from the home. This may mean using a gun lock or storing the weapon in a gun safe, storing ammunition separately from an unloaded weapon, removing the firing pin, or otherwise disassembling the weapon. Means safety counseling has been shown to be effective in reducing suicide rates [63] and is acceptable to patients [64]. Studies indicate that over 90% of individuals who make a suicide attempt and survive do not go on to die by suicide [65]; therefore, reducing access to highly lethal means during a suicidal crisis may be key in reducing suicide rates. Though an in-depth review of means safety counseling is outside the scope of this article, readers are directed to Bryan, Stone, and Rudd’s article for a practical overview of means safety discussions [66].

Second, safety planning is a brief intervention that may be beneficial in the primary care setting [67,68]. The goal of a safety plan is to create an individualized plan to remain safe during a suicidal crisis. Means safety discussion is the last of 6 steps in the safety plan [68]. The first 5 steps include identifying warning signs, using internal coping strategies, social connectedness as distraction, social support for the crisis, and professionals that can be used as resources. When patients can identify specific, individualized warning signs that occur prior to a crisis, they can then use strategies to cope and prevent the crisis from worsening. Coping strategies that are encouraged are first internal (ie, those that can be done without relying on anyone else), such as exercise or journaling. If those do not improve the patient’s mood, then he or she is encouraged to use people or social settings as a distraction (eg, people watching at the mall, calling an acquaintance to chat), and if he or she is still feeling bad, encouraged to get social support for the crisis (eg, calling a family member to discuss the crisis and get support). Finally, if all of these steps are not effective, the older adult is encouraged to reach out to professional supports, such as a mental health provider, the National Suicide Prevention Lifeline, or 911 (or go to an emergency room). Readers are encouraged to review Stanley and Brown’s articles for comprehensive details about safety planning as an intervention [67,68]. Additionally, an article with specific adaptations for safety planning with older adults is forthcoming [69].

Conclusion

Older adults, particularly older men, are at high risk for suicide [1,2], and primary care practitioners are a critical component of older adult suicide prevention. Older adults frequently see primary care practitioners within a month prior to death by suicide [20,21]; primary care practitioners are uniquely qualified to address a broad range of potential risk factors, and may have more interactions and familiarity with older adults at risk for suicide than other medical professionals [20–22]. Primary care practitioners should be prepared to identify risk factors and warning signs for older adult suicide, ask appropriate questions to screen for suicide risk, and intervene to prevent suicide. Screening can consist of standardized written questionnaires or oral questioning, and interventions may include providing resources and referrals, discussions about means safety, safety planning, and handoff to a mental health specialist. Interventions for suicide risk are likely feasible and acceptable in primary care [46–48]. Primary care practitioners have an important role to play in older adult suicide prevention, and must be prepared to interact with older adults who may be at risk for suicide.

Corresponding author: Danielle R. Jahn, PhD, Primary Care Institute, 605 NE 1st St, Gainesville, FL 32605, [email protected].

Financial disclosures: None reported.

From the Primary Care Institute, Gainesville, FL.

Abstract

• Objective: To provide primary care practitioners with the knowledge required to identify and address older adult suicide risk in their practice.
• Methods: Review of the literature and good clinical practices.
• Results: Primary care practitioners play an important role in older adult suicide prevention and must have knowledge about older adult suicide risk, including risk factors and warning signs in this age-group. Practitioners also must appropriately screen for and manage suicide risk. Older adults, particularly older men, are at high risk for suicide, though they may be less likely to report suicide ideation. Additionally, older adults frequently see primary care practitioners within a month prior to death by suicide. A number of older adult–specific risk factors are reviewed, and appropriate screening and intervention for the primary care setting are discussed.
• Conclusion: Primary care practitioners are uniquely qualified to address a broad range of potential risk factors and should be prepared to identify risk factors and warning signs for older adult suicide, ask appropriate questions to screen for suicide risk, and intervene to prevent suicide.

Key words: suicide; older adults; risk factors; screening; safety planning.

Primary care practitioners play an important role in older adult suicide prevention and have a responsibility to identify and address suicide risk among older adults. To do so, practitioners must understand the problem of older adult suicide, recognize risk factors for suicide in older adults, screen for suicide risk, and appropriately assess and manage suicide risk. Primary care practitioners may face challenges in completing these tasks; the goal of this article is to assist practitioners in addressing these challenges.

Suicide in Older Adults

The United States has recently seen increases in suicide rates across the lifespan; from 1999 to 2014, the suicide rate rose by 24% across all ages [3]. Among both men and women aged 65 to 74, the suicide rate increased in this time period [3]. The high suicide rate among older adults is particularly important to address given the increasing numbers of older adults in the United States. By 2050, the older adult population in the United States is expected to reach 88.5 million, more than double the older adult population in 2010 [4]. Additionally, the generation that is currently aging into older adulthood has historically had higher rates of suicide across their lifespan [5]. Given that suicide rates also increase in older adulthood for men, the coming decades may evidence even higher rates of suicide among older adults than previously and it is critical that older adult suicide prevention becomes a public health priority.

It is also essential to discuss other suicide-related outcomes among older adults, including suicide attempts and suicide ideation. This is critical particularly because the ratio of suicide attempts to deaths by suicide in this age-group is 4 to 1 [1]. This is in contrast to the ratio of attempts to deaths across all ages, which is 25 suicide attempts per death by suicide [1]. This means that suicide prevention must occur before a first suicide attempt is made; suicide attempts cannot be used a marker of elevated suicide risk in older adults or an indication that intervention is needed. Intervention is required prior to suicide risk becoming elevated to the point of a suicide attempt.

It is also critical to recognize that despite the fact that suicide rates rise with age, reports of suicide ideation decrease with age [7,8]. Across all ages, 3.9% of Americans report past-year suicide ideation; however, only 2.7% of older adults report thoughts of suicide [9]. The discrepancy with the increasing rates of death by suicide with age suggest that suicide risk, and thereby opportunities for intervention, may be missed in this age-group [10].

However, older adults may be more willing to report death ideation, as research has found that over 15% of older adults endorse death ideation [11–13]. Death ideation is a desire for death without a specific desire to end one’s own life, and is an important suicide-related outcome, as older adults with death ideation appear the same as those with suicide ideation in terms of depression, hopelessness, and history of suicidal behavior [14]. Additionally, older adults with death ideation had more hospitalizations, more outpatient visits, and more medical issues than older adults with suicide ideation [15]. Therefore, death ideation should be taken as seriously as suicide ideation in older adults [14]. In sum, the high rates of death by suicide, the likelihood of death on a first or early suicide attempt, and the discrepancy between decreasing reports of suicide ideation and increasing rates of death by suicide among older adults indicate that older adult suicide is an important public health problem.

Suicide Prevention Strategies

Many suicide prevention strategies to date have focused on indicated prevention, which concentrates on individuals already identified at high risk (eg, those with suicide ideation or who have made a suicide attempt) [16]. However, because older adults may not report suicide ideation or survive a first suicide attempt, indicated prevention is likely not enough to be effective in older adult suicide prevention. A multilevel suicide prevention strategy [17] is required to prevent older adult suicide [18]. Older adult suicide prevention must include indicated prevention but must also include selective and universal prevention [16]. Selective prevention focuses on groups who may be at risk for suicide (eg, individuals with depression, older adults) and universal prevention focuses on the entire population (eg, interventions to reduce mental health stigma) [16]. To prevent older adult suicide, crisis intervention is critical, but suicide prevention efforts upstream of the development of a suicidal crisis are also essential.

The Importance of Primary Care

Research indicates that primary care is one of the best settings in which to engage in older adult suicide prevention [18]. Older adults are significantly less likely to receive specialty mental health care than younger adults, even when they have depressive symptoms [19]. Additionally, among older adults who died by suicide, 58% had contact with a primary care provider within a month of their deaths, compared to only 11% who had contact with a mental health specialist [20]. Among older adults who died by suicide, 67% saw any provider in the 4 weeks prior to their death [21]. Approximately 10% of older adults saw an outpatient mental health provider, 11% saw a primary care physician for a mental health issue, and 40% saw a primary care physician for a non-mental health issue [21]. Therefore, because older adults are less likely to receive specialty mental health treatment and so often seen a primary care practitioner prior to death by suicide, primary care may be the ideal place for older adult suicide risk to be detected and addressed, especially as many older adults visit primary care without a mental health presenting concern prior to their death by suicide.

Additionally, older adults may be more likely to disclose suicide ideation to primary care practitioners, with whom they are more familiar, than physicians in other settings (eg, emergency departments). Research has shown that familiarity with a primary care physician significantly increases the likelihood of patient disclosure of psychosocial issues to the physician [22]. Primary care providers also have a critical role as care coordinators; many older adults also see specialty physicians and use the emergency department. In fact, older adults are more likely to use the emergency department than younger adults, but emergency departments are not equipped to navigate the complex care needs of this population [23]. Primary care practitioners are important in ensuring that health issues of older adults are addressed by coordinating with specialists, hospitals (eg, inpatient stays, emergency department visits, surgery) and other health services (eg, home health care, physical therapy). Approximately 35% of older adults in the United States experience a lack of care coordination [24], which can negatively impact their health and leave issues such as suicide ideation unaddressed. Primary care practitioners may be critical in screening for mental health issues and suicide risk during even routine visits because of their familiarity with patients, and also play an important role in coordinating care for older adults to improve well-being and to ensure that critical issues, such as suicide ideation, are appropriately addressed.

Primary care practitioners can also be key in upstream prevention. Primary care practitioners are in a unique role to address risk factors for suicide prior to the development of a suicidal crisis. Because older adults frequently see primary care practitioners, such practitioners may have more opportunities to identify risk factors (eg, chronic pain, depression). Primary care practitioners are also trained to treat a broad range of conditions, providing the skills to address many different risk factors.

Finally, primary care is a setting in which screening for depression and suicide ideation among older adults is recommended. The US Preventive Services Task Force recommends screening for depression in all adults and older adults and provides recommended screening instruments, some of which include questions about self-harm or suicide risk [25]. However, this same group has concluded that there is insufficient evidence to support a recommendation for suicide risk screening [26]. Despite this, the Joint Commission recently released an alert that recommends screening for suicide risk in all settings, including primary care [27]. The Joint Commission requirement for ambulatory care that is relevant to suicide is PC.04.01.01: The organization has a process that addresses the patient’s need for continuing care, treatment, or services after discharge or transfer; behavioral health settings have additional suicide-specific requirements. The recommendations, though, go far beyond this requirement for primary care. The Joint Commission specifically notes that primary care clinicians play an important role in detecting suicide ideation and recommends that primary care practitioners review each patient’s history for suicide risk factors, screen all patients for suicide risk, review screenings before patients leave appointments, and take appropriate actions to address suicide risk when needed [27]. Further details are available in the Joint Commission’s Sentinel Event Alert titled, “Detecting and treating suicide ideation in all settings” [27]. Given these recommendations, primary care is an important setting in which to identify and address suicide risk.

Risk Factors for Older Adult Suicide

Numerous reviews exist that cover many risk factors for suicide in older adults [18,28]. This article will focus briefly on risk factors that are likely to be recognized and potentially addressed by primary care practitioners. Risk factors that apply across the lifespan can be recalled through a mnemonic: IS PATH WARM [29]. These risk factors include suicide Ideation, Substance abuse, Purposelessness, Anxiety (including agitation and poor sleep), feeling Trapped, Hopelessness, social Withdrawal, Anger or rage, Recklessness (ie, engaging in risky activities), and Mood changes. The National Suicide Prevention Lifeline also includes being in unbearable physical pain, perceiving one’s self as a burden to others, and seeking revenge on others as risk factors [30]. More specific to older adults, Conwell notes 5 categories or domains of risk factors with strong research support: psychiatric symptoms, somatic illness, functional impairment, social integration, and personality traits and coping [18,31].

Affective or mood disorders, particularly depression and depressive symptoms, are some of the most well-studied and strongest risk factors for older adult suicide [31]; 71% to 97% of all older adults who die by suicide have psychiatric illnesses [28]. Mood disorders, including major depressive episodes, are most consistently linked to older adult suicide risk; there is evidence as well for anxiety disorders and substance abuse disorders as risk factors, though it is somewhat mixed [28]. Therefore, screening for depression, anxiety, and substance abuse may be key to recognizing potential suicide risk. However, depression and anxiety do not present similarly in younger and older adults [32,33]. Depressive symptoms in older adults may be more somatic (eg, agitation, gastrointestinal symptoms) [32] and may reflect more anhedonia than mood changes [33]. Anxiety in older adults tends to be reported as stress or tension, whereas younger adults report feeling anxious or worried [33]. Additionally, substance abuse is often underrecognized, underdiagnosed, and undertreated in older adults [34]. Proactive screening for substance abuse is important as it may not interfere with work or other obligations in older adults, and therefore substance abuse may not be identified by older adults or others in their lives.

Physical illness may also be a risk factor for suicide [28,31]. Numerous diagnoses have been linked to suicide risk, including cancers, neurodegenerative diseases (eg, amyotrophic lateral sclerosis, Huntington disease), spinal cord injury, cardiovascular disease, and pulmonary disease [28,35]. However, overall illness burden (ie, number of chronic illnesses) [28] and self-perceived health [36] appear to be stronger risk factors than any specific illness. Additionally, authors have suggested that illness itself may not be a particularly strong risk factor, but the effect of illness on depressive symptoms [35], functioning, pain, or hopelessness due to the potential for decline over time [28] may increase suicide risk in older adults. Pain itself has been identified as a risk factor for suicide, as have perceptions of burden to others, hopelessness, and functional impairment [28].

In terms of functional impairment, research has shown that impairment in completing instrumental activities of daily living is associated with higher risk for death by suicide, and cognitive impairment may also be associated with elevated suicide risk [28]. However, there are some discrepant findings regarding the role of dementia in suicide risk, which may reflect medical and psychiatric comorbidities, as well as different stages of dementia or levels of cognitive impairment (eg, hopelessness about cognitive decline may increase suicide risk shortly after diagnosis, whereas lack of insight may decrease risk later in the course of the illness) [37]. Related to functional or cognitive impairment is perceived burdensomeness (ie, the perception that one is a liability or burden to others, to the point that others would be better off if one was gone) [38], which may also be associated with suicide risk in older adults [39,40]. Researchers have found that the interaction between perceived burdensomeness and thwarted belongingness (ie, a belief that one lacks reciprocal caring relationships and does not belong) identified older adults who were likely experiencing suicide ideation but did not report it [41]. These findings indicate that perceived burdensomeness and thwarted belongingness may be key in identifying older adults at risk for suicide.

Thwarted belongingness has also been linked to suicide ideation in older adults [41]. In fact, studies suggest that social integration is especially important for reducing suicide risk in this population [28,31,42]. A larger social network, living with others, and being active in the community are each protective against suicide [28]. Bereavement, which can reduce social connectedness and acts as a significant life stressor, is also an important risk factor [31]. Retirement may also reduce social connectedness, and employment changes have been identified as a suicide risk factor for older adults [28]. Retirement has been linked to risk for death by suicide in this population [43], and may not only serve to reduce social connectedness, but for some older adults may also be a significant role loss or loss of sense of purpose that can influence suicide risk.

Finally, rigid personality traits or coping styles are a risk factor for suicide among older adults [28,31]. As older adults face potential losses, health changes, and functional decline, effective positive coping strategies and flexibility are key to maintaining well-being. If older adults are unable to flexibly cope with these challenges, their risk for suicide increases [28].

In addition to risk factors, which confer suicide risk but do not necessarily suggest that an older adult is thinking about suicide, warning signs exist that indicate that suicide risk is imminent. These include suicidal communication (ie, talking or writing about suicide), seeking access to means, and making preparations for suicide (eg, ensuring a will is in place, giving away prized possessions). One important note is that discussing and preparing for death may be developmentally appropriate for older adults, particularly those with chronic illnesses; however, such appropriate preparation is critically different from talking about suicide or a desire for death.

Additionally, a lack of planning for the future may be a warning sign. For example, older adults who decline to schedule medical follow-up or do not wish to refill needed prescriptions may be exhibiting warning signs that should be addressed. Similarly, not following needed medical regimens (eg, an older adult with diabetes no longer taking insulin) is also a warning sign. Other, potentially more subtle warning signs may include significant changes in mood, sleep, or social interactions. Older adults may become agitated and sleep less when they are considering suicide, or may feel more at ease after they have made the decision to die by suicide and their sleep or mood may improve. Withdrawing from valued others may also be a warning sign. Finally, recent major changes (eg, loss of a spouse, moving to an assisted living facility) may be triggers for suicide risk and can serve as warning signs themselves.

Specific Screening Strategies

Given the numerous risk factors and warning signs for older adult suicide, as well as the time limitations that primary care practitioners face [44,45], it would be impractical to comprehensively assess each older adult who presents at a primary care practice. Therefore, more specific screening is necessary. Most importantly, every older adult should be screened for suicide ideation and death ideation at every visit. Screening at every visit is critical because suicide ideation may develop at any point. Previous research has included screening of over 29,000 older adults in 11 primary care settings for suicide ideation, risk of alcohol misuse, and mental health disorders [15], suggesting that suicide risk screening is feasible. Other studies have also successfully used widespread screening for depression and suicide ideation among older adults in primary care [46–48]. Additionally, in an emergency department setting, universal suicide risk screening has been associated with significantly improved risk detection [49], indicating that improved screening may be beneficial in identifying suicide risk. Importantly, asking about suicide does not cause thoughts of suicide [50]. Additionally, it is a myth that those who talk about suicide ideation will not act on these thoughts [51].

When primary care practitioners inquire about suicide ideation, they should also ask about death ideation; though some may believe that death ideation is not as significant in terms of suicide risk as suicide ideation, recall that research has not found differences in previous suicide attempts or current hopelessness among older adults with death ideation versus suicide ideation [14]. Therefore, screening for death ideation should be completed as part of every suicide risk screening.

Screening can take many forms. Screening may be oral; asking an older adult if he or she is having thoughts of suicide or is experiencing a desire to die is a brief, 2-question screening that may provide valuable information (eg, “Are you having thoughts about your own death or wanting to die?”, “Are you having thoughts of killing yourself or thinking about suicide?”). This screening may be conducted by medical assistants, nurses, care managers, or physicians, with the patient’s responses documented. Importantly, a standard procedure should be implemented to ensure older adults are consistently asked about suicide risk at each visit, but do not feel inundated by such questions from numerous staff.

If verbal questions are asked, they must be asked appropriately. Euphemisms or indirect language should not be used during a screening; older adults should be directly asked about thoughts of death and suicide, not simply asked questions such as, “Have you ever had thoughts of harming or hurting yourself?” A question like this does not adequately assess current suicide risk, as it does not assess current thoughts, nor does it specifically inquire about suicide ideation (ie, killing one’s self). It is also important to phrase questions in a manner that invites honest responses and conveys an openness to listening. For example, asking, “You’re not thinking about suicide, are you?” suggests that the practitioner wants the older adult to say no and is not comfortable with the older adult endorsing suicide ideation. Open questions that allow endorsement or denial (eg, “Are you having thoughts of killing yourself?”) imply that the practitioner is receptive to either an endorsement or denial of suicide ideation.

Alternatively, a written screening can be used; older adults may complete a questionnaire prior to their appointment or while waiting to see their practitioner. Such an assessment may be a brief screening (eg, using similar yes/no questions to an oral screening), or may be a standardized measure. For example, the Geriatric Suicide Ideation Scale [52] is a 31-item self-report measure that provides scores for suicide ideation, death ideation, loss of personal and social worth, and perceived meaning in life. Though there are not standard cutoffs that suggest high versus low suicide risk, responses can be reviewed to identify whether older adults are reporting suicide ideation or death ideation, and can also be compared to norms (ie, average scores) from other older adults [52]. This measure also has the benefit of 2 subscales that do not specifically require reporting thoughts of suicide or death (ie, loss of personal and social worth, perceived meaning in life), which may give practitioners an indication of an older adult’s suicide risk even if the older adult is not comfortable disclosing suicide ideation, as has been shown in previous research [7,8].

Similarly, the Geriatric Depression Scale, which has a validated 15-item version [53], does not directly ask about suicide ideation but has a 5-item subscale that has been found to be highly correlated with reported suicide ideation [54]. When administered to older adult primary care patients, this subscale was an effective measure of suicide ideation; a score of ≥ 1 was the best cutoff for determining whether an older adult reported suicide ideation [55].

Additionally, as noted previously, the interaction between perceived burdensomeness and thwarted belongingness may identify older adults who are potentially experiencing, but not reporting, suicide ideation [41]. The Interpersonal Needs Questionnaire [56] is the validated assessment for both perceived burdensomeness and thwarted belongingness. Perceived burdensomeness is assessed via 6 self-report items, and thwarted belongingness is assessed via 9 self-report items on this measure [56]. There are not specific cutoffs that determine high versus low perceived burdensomeness or thwarted belongingness, but older adults’ responses can provide information about their experiences of these constructs. Administration of the Interpersonal Needs Questionnaire can provide information about potential risk for suicide among older adults who may otherwise deny thoughts of suicide or death.

If the screening for suicide ideation or death ideation is positive (ie, the older adult endorses thoughts of suicide or death), the treating primary care practitioner must then follow up with additional questions to determine current level of suicide risk. To make this determination, at a minimum, follow-up questions should focus on whether the older adult has any intent to die by suicide (eg, “Do you have any intent to act on your thoughts of suicide?”), as well as whether he or she has a plan to die by suicide (eg, “Have you begun formulating a plan to die by suicide?”). When asking about a plan, it is important to determine how specific the plan is. For example, an older adult with a specific method identified and date selected to implement the plan is at much higher risk than an older adult with a relatively vague idea. It is also critical to assess for the older adult’s access to means for suicide. If an older adult has a specific plan and has the capability to carry out the plan (eg, plans to overdose on prescription medication and has large quantities of medication or high-lethality medication at home), he or she is more likely to die by suicide than an older adult who does not have access to means (eg, only has small quantities of low-lethality medication available). A general assessment of risk factors and previous suicidal behavior (ie, any previous suicide attempts) also informs decisions about level of risk and interventions.

After a screening or assessment is completed, a risk determination must be made and documented. Acute suicide risk can be categorized as low, moderate, or high. It is not appropriate to say that there is “no” suicide risk present. Low risk occurs when there is no current suicide ideation, no plan to die by suicide, and no intent to act on suicidal thoughts, especially when the patient has no history of suicidal behavior and few risk factors [57]. Moderate risk is evident when there is current suicide ideation, but no specific plan to die by suicide or intent to act on suicidal thoughts. There are likely warning signs or risk factors, which may include previous suicidal behaviors, present in moderate suicide risk [57]. High risk is indicated by current suicide ideation with plan to die by suicide and suicidal intent. There are significant warning signs and risk factors present; there may also be a recent suicide attempt, though this is not a requirement for a high risk determination [57]. Undetermined suicide risk occurs when a practitioner cannot accurately assess risk, but concern regarding suicide is present; this is primarily used when a patient refuses to answer questions about suicide. Undetermined risk should be treated as at least moderate risk. Because research shows that death ideation has similar outcomes to suicide ideation in older adults [14], death ideation should also be factored into determinations of suicide risk; reports of death ideation may indicate low or moderate risk in older adults, dependent upon other risk factors, suicidal intent, and plan.

After a risk determination is made, it must be documented in the medical record. The level of risk and rationale for that determination must be included [58]. Stating only the level of risk without a rationale (ie, the older adult’s responses to questions) is not adequate, and documenting only the older adult’s responses without a determination of risk is also not sufficient. Finally, it is critical to document the intervention that occurred or steps taken after the level of risk was determined.

Critically, stating only that there was no indication of suicide risk is inadequate. For example, documenting “No evidence of suicide risk” is not appropriate. This documentation does not indicate that the older adult was specifically asked about suicide ideation, death ideation, suicidal intent, or plan to die by suicide. It also does not indicate a level of suicide risk. Examples of appropriate documentation include:

 Patient was asked about suicide risk. She denied current suicide ideation but reported death ideation. She denied any current suicidal intent or plan. She also denied any previous suicide attempts. Therefore, acute suicide risk was deemed to be low. Provided patient with wallet card about the National Suicide Prevention Lifeline. Also called the Friendship Line while in the room with the patient to connect her with services. Finally, provided a brief list of local mental health professionals to patient; the patient reported she would like to see Dr. Smith. Called and left a message for Dr. Smith with referral information with patient during appointment.

Patient was asked about suicide risk. He reported both death ideation and suicide ideation. He also reported a nonspecific plan (ie, causing a single-vehicle motor vehicle accident, with no specific plan for the motor vehicle accident or timeframe) and denied any intent to act on his thoughts of suicide. He reported one previous suicide attempt, at age 22, by overdose on over-the-counter medication. He reported that this attempt did not require medical attention. Therefore, acute suicide risk was determined to be moderate. Patient was introduced to the behavioral health specialist, who met with the patient during the appointment to conduct further assessment and intervention.

Specific Intervention Strategies

Despite the fact that the pace of the primary care setting often does not allow for time-intensive intervention, there are ways to address suicide risk in this setting. Importantly, no-suicide contracts should not be used at any time [59,60]. No-suicide contracts are documents that patients who are experiencing suicide ideation are required to sign that state that they will not die by suicide while under the care of the practitioner. These contracts have no evidence of effectiveness, and some researchers argue that they may in fact damage the relationship with patients and serve the practitioner’s needs more than the patient’s needs [59].

One of the best options for older adults at low acute suicide risk is to provide resources and referrals. The National Suicide Prevention Lifeline can be reached at 1-800-273-TALK (8255); trained counselors are available to speak to patients at all times. Wallet cards with information about the National Suicide Prevention Lifeline are available at no charge from the US Substance Abuse and Mental Health Services Administration online store. The Friendship Line is another service available free to adults ages 60 and older, 24 hours per day, 7 days per week; this line can be reached at 1-800-971-0016. The Friendship Line, which is managed by the Institute on Aging, also provides outreach calls to older adults who may be isolated or lonely, increasing connectedness and potentially reducing suicide risk.

Having a ready list of local mental health professionals with expertise in geriatrics and suicide risk to provide to the patient is also beneficial. Recall, though, that older adults are less likely to seek out and receive mental health services [19]; therefore, connecting the patient with resources or referrals during the appointment is critical. If the practitioner does not have time to do this, having a medical assistant or other staff member that the patient knows engage in this step may be appropriate. For example, the patient can call the Friendship Line or National Suicide Prevention Lifeline while in the room with the practitioner, which may reduce anxiety or stigma about doing so and connect the patient with services. Similarly, calling a local mental health professional to make a referral during the appointment may increase the likelihood that the older adult will follow up on the referral.

The most ideal method of intervention for moderate or high acute suicide risk is a warm handoff to a behavioral or mental health specialist. As primary care and behavioral health become more integrated and financially viable as reimbursement through the Centers for Medicare and Medicaid Services improves [61], it is becoming increasingly likely that such a specialist will be on-site and available. Research has found that collaborative care in primary care reduces suicide risk in older adults [46–48,62]. Mental health specialists can conduct more comprehensive assessments and spend more time intervening to reduce suicide risk among older adults with death or suicide ideation. If an on-site behavioral health specialist is not available, older adults at high suicide risk may need to be referred to an emergency department for further evaluation and follow-up. Each state has its own laws and procedures regarding this process, which should be incorporated into a practice’s procedures for addressing high suicide risk. The procedure often involves ensuring that the older adult is accompanied at all times (ie, not left alone in a room), alerting emergency services (usually via phone call to an emergency line, such as 911), and completion of paperwork by a practitioner asserting that the patient is a danger to self. Police or other emergency personnel are then responsible for transporting the patient for further evaluation and determination of whether hospitalization is required.

If more time is available, either via the treating primary care practitioner or other patient care staff in the office, other brief interventions may be beneficial. First, means safety discussions are critical, particularly for older adults with plans for suicide or access to highly lethal means. In such discussions, patients are encouraged to restrict access to the methods that they may use to die by suicide. Plans for restricting access are developed, and when possible, a support person is enlisted to ensure that the plans are carried out. For example, if an older adult has access to firearms (eg, keeps a loaded weapon in his or her nightstand), he or she is encouraged to restrict his or her access to it. Ideally, this is through removing the weapon from the home, either permanently or until suicide risk reduces (eg, giving it to a friend, turning it over to police), but more safe storage may also be an option if the older adult is not willing to remove the weapon from the home. This may mean using a gun lock or storing the weapon in a gun safe, storing ammunition separately from an unloaded weapon, removing the firing pin, or otherwise disassembling the weapon. Means safety counseling has been shown to be effective in reducing suicide rates [63] and is acceptable to patients [64]. Studies indicate that over 90% of individuals who make a suicide attempt and survive do not go on to die by suicide [65]; therefore, reducing access to highly lethal means during a suicidal crisis may be key in reducing suicide rates. Though an in-depth review of means safety counseling is outside the scope of this article, readers are directed to Bryan, Stone, and Rudd’s article for a practical overview of means safety discussions [66].

Second, safety planning is a brief intervention that may be beneficial in the primary care setting [67,68]. The goal of a safety plan is to create an individualized plan to remain safe during a suicidal crisis. Means safety discussion is the last of 6 steps in the safety plan [68]. The first 5 steps include identifying warning signs, using internal coping strategies, social connectedness as distraction, social support for the crisis, and professionals that can be used as resources. When patients can identify specific, individualized warning signs that occur prior to a crisis, they can then use strategies to cope and prevent the crisis from worsening. Coping strategies that are encouraged are first internal (ie, those that can be done without relying on anyone else), such as exercise or journaling. If those do not improve the patient’s mood, then he or she is encouraged to use people or social settings as a distraction (eg, people watching at the mall, calling an acquaintance to chat), and if he or she is still feeling bad, encouraged to get social support for the crisis (eg, calling a family member to discuss the crisis and get support). Finally, if all of these steps are not effective, the older adult is encouraged to reach out to professional supports, such as a mental health provider, the National Suicide Prevention Lifeline, or 911 (or go to an emergency room). Readers are encouraged to review Stanley and Brown’s articles for comprehensive details about safety planning as an intervention [67,68]. Additionally, an article with specific adaptations for safety planning with older adults is forthcoming [69].

Conclusion

Older adults, particularly older men, are at high risk for suicide [1,2], and primary care practitioners are a critical component of older adult suicide prevention. Older adults frequently see primary care practitioners within a month prior to death by suicide [20,21]; primary care practitioners are uniquely qualified to address a broad range of potential risk factors, and may have more interactions and familiarity with older adults at risk for suicide than other medical professionals [20–22]. Primary care practitioners should be prepared to identify risk factors and warning signs for older adult suicide, ask appropriate questions to screen for suicide risk, and intervene to prevent suicide. Screening can consist of standardized written questionnaires or oral questioning, and interventions may include providing resources and referrals, discussions about means safety, safety planning, and handoff to a mental health specialist. Interventions for suicide risk are likely feasible and acceptable in primary care [46–48]. Primary care practitioners have an important role to play in older adult suicide prevention, and must be prepared to interact with older adults who may be at risk for suicide.

Corresponding author: Danielle R. Jahn, PhD, Primary Care Institute, 605 NE 1st St, Gainesville, FL 32605, [email protected].

Financial disclosures: None reported.