Evidence-Based Reviews

Alysha, age 15, is an “A” student and top athlete who feels her parents push her to be perfect. After getting a B on a test, she feels overwhelmed by shame and guilt. She locks herself in the bathroom and begins cutting her arm with a razor blade.

Self-injurious behavior (SIB) in adolescents can be associated with internalizing, externalizing, and substance abuse disorders (Table 1). For most practitioners, such as Alysha, a major goal of SIB is to relieve intolerable stress and negative affect.¹

Although this secretive, highly addictive, learned behavior can be difficult to control, some clinical approaches can help these distressed teens and their parents. This article examines the dynamics of SIB, the association between suicidal ideation and SIB, and recommended treatments such as substitute behaviors and dialectic behavioral therapy.

Table 1

Common Axis I disorders among teens with SIB

Axis I disorder	Prevalence*
Any internalizing disorder	52%
Major depressive disorder	42
Posttraumatic stress disorder	24
Generalized anxiety disorder	16
Any externalizing disorder	63
Conduct disorder	49
Oppositional defiant disorder	45
Any substance use disorder	60
Alcohol abuse	18
Alcohol dependence	17
Nicotine dependence	39
Marijuana abuse	13
Marijuana dependence	30
Other substance abuse	3
Other substance dependence	6
*Among a sample of 89 adolescents who engaged in SIB
Source: Nock MD, Joiner TE Jr, Gordon KL, et al. Non-suicidal self-injury among adolescents: diagnostic correlates and relationship to suicide attempts. Psychiatry Res 2006;144(1):68.
Reprinted with permission from Elsevier.

Growing problem in adolescents

SIB is the deliberate infliction of harm to oneself, either internally or externally, without suicidal intent.¹ This behavior is also known as impulsive self-injury, non-suicidal self-injury, self-mutilation, cutting, and self-harm. Once reported primarily in adults with borderline personality disorder, SIB is becoming common among adolescents:

• Among 663 teens in community-sample survey, 46% engaged in some form of SIB in the past year, and 28% engaged in serious, repetitive behaviors.²
  
• 13% to 23% of teenagers have engaged in SIB, according to literature review.³
  
• Children as young as 9 have presented with SIB.
  

Not just cutting. Besides cutting, other methods of self-injury include:

• burning
  
• swallowing objects or substances
  
• hitting oneself with the fist or against an object
  
• cutting circulation to a digit
  
• picking at skin
  
• pulling out hair.
  

Individuals with chronic illnesses can engage in SIB by not complying with treatment, such as a diabetic taking too much or too little insulin or an epileptic not taking medication.

Socially sanctioned behaviors, such as body piercing and tattoos, usually are not considered SIB. They can be used as SIB, however, by teens who impulsively self-pierce or tattoo without appropriate hygiene or anesthetic agents. Their purpose is not to make a fashion statement but to produce pain or discomfort. Cultural behaviors that cause scarring as a rite of passage, such as the Native American Sun Dance, are not considered SIB.¹

Despite increasing prevalence among adolescents, SIB remains a solitary behavior. Based on my clinical experience, teens may share ideas about SIB, but they generally don’t practice it in groups.

SIB psychodynamics

Adults and adolescents with SIB frequently have:

• difficulty regulating emotions
  
• unstable interpersonal relations
  
• limited coping strategies.¹
  

Adolescents with SIB frequently experience anger, and their self-harm can result from turning this anger inward because they are unable to express it toward others. This is seen in a patient who describes a cutting episode that occurred while you were on vacation but sees no connection between your absence and the SIB.

Shame is also common and can be a major barrier to diagnosing SIB. Adolescents who are ashamed of the behavior will go to great lengths to hide it from others, including clinicians. Despite the shame, many adolescents feel unable to stop engaging in SIB because it fulfills a powerful need.

Adults and adolescents who practice SIB most often report their behavior is motivated by affect regulation and tension release.⁴ Some adolescents engage in a different, manipulative form of SIB not to relieve tension but as a threat to prevent loss (Box).⁴

Behavioral reinforcement. An acute stressor—such as parental limits on behavior, feelings of rejection or abandonment by peers, or failing to achieve an unrealistic goal—triggers an escalating, intolerable affect. By experimentation or accident, an adolescent discovers that SIB provides rapid relief of the intolerable state—a calmness that may last for minutes, hours, or days. This relief reinforces the behavior, and the adolescent repeats SIB when faced with the next stressor.

Most individuals with SIB report a similar sequence of events. There is a trigger event, usually involving a real or perceived feeling of loss, rejection, or abandonment. The adolescent tries to resist the impulse to self-harm, feels escalating emotional distress, engages in SIB, and feels immediate relief.⁵

 

Inducing pain or bleeding as a means of relieving intolerable stress may be an attempt to:

• turn emotional pain into more manageable physical pain
  
• direct anger that cannot be expressed at others onto oneself
  
• punish oneself for perceived misdeeds.⁴
  

SIB tends to escalate over time. Those who engage in it may require more frequent or intensive self-injury to achieve relief. Because the emotional state improves quickly and the adolescent feels a sense of control over the behavior, SIB rapidly can become habitual and difficult to interrupt.

Box

CASE 2: Lingering effects of trauma

Melissa, age 13, has been sexually abused by her foster brother. She briefly returns to the home where her foster brother still resides. He has torn up her room, and her mother—who supports the foster brother—has left the room that way for her to find. Melissa returns to her current residence, cuts her arm 15 times, and pierces her tongue.

In therapy, she denies being angry or upset and does not know why she cut her arm or pierced her tongue.

An adolescent with a history of childhood trauma might have difficulty identifying and expressing internal states and developing trusting relations with others. He or she also can have high levels of anger and shame.

Even an adolescent who experienced neglect or loss of attachment object without overt physical or sexual abuse might have trouble establishing a therapeutic alliance because of difficulty with trust.⁶ He or she may have little or no capacity to identify emotional states, which limits insight into the behavior.⁵

Dissociation. Some adolescents use SIB to try to feel something or to bring themselves back from a dissociative state. They report feeling lost, alone, and disconnected from others and themselves. Some report seeing blood as a way of reconnecting with being alive.

Dissociation may occur in adolescents with a history of trauma, particularly in childhood. Abused individuals who engage in SIB may be identifying with the aggressor, attempting to cut away internalized negative images of the abuser or to control anger they are unable to acknowledge.⁶

Clues prompt further assessment

SIB assessment begins with screening for the behavior. If you find any indicators that suggest SIB (Table 2), question the adolescent about self-injury. Many adolescents want help—and will respond accurately to questions about self-injury—but need to be asked. They usually won’t volunteer the information during an initial evaluation.

Once you have identified SIB, explore the behavior. Document:

• number, location, and age of injuries
  
• depth of cuts (if applicable)
  
• signs of infection.
  

Determine if the adolescent needs medical intervention. Discuss how the adolescent causes the injury, and what precipitates it. If possible, obtain some form of safety contract in which the adolescent agrees to not engage in SIB and to notify a designated adult if he or she feels like engaging in SIB or has done so.

Suicide screening. By definition, SIB does not include an intention to die, and most teenagers with SIB will deny suicidal intent. However, because the line between SIB and passive suicide can be thin, careful screening and ongoing monitoring for suicidal ideation and behavior in teens with SIB is essential. In one study:

• 70% of adolescents who engaged in SIB had made one suicide attempt
  
• 55% had made multiple attempts.⁷
  

A separate study found suicidal ideation and depression are keys to identifying adolescents with SIB at risk for suicide attempts.⁸ Because SIB and suicidal ideation/behaviors can co-occur, SIB safety contracts must cover both.

 

Address the parents’ concerns. During your assessment, also focus on the adolescent’s parents. They often are highly distressed, confused, and angry. They typically learn about the SIB from their child’s school counselors or peers and feel betrayed and guilty. They may want to be excessively intrusive and punitive and need support, information, and guidance to address their child’s safety.

Table 2

SIB: Spotting behavioral clues

Adolescents Wearing long sleeves or pants in warm weather Becoming increasingly isolated from family and peers Spending long periods of time in the bathroom or bedroom Deteriorating school performance Decreasing hygiene
Parents Finding blood on teen’s clothing or in unusual places in the home Discovering sharp instruments are missing or placed in unusual locations

Treatment recommendations

Always take SIB in adolescents seriously, and not as something they will “outgrow.” Adolescents with SIB need help modulating affect, stabilizing interpersonal relationships, and developing more adaptive coping strategies and problem-solving skills. Underlying dynamics—especially childhood trauma—must be explored and resolved.

Few evidence-based studies have evaluated SIB treatment in adolescents. Clinicians have extrapolated suggested interventions from the adult literature; however, much of this data was obtained from treating adult women with borderline personality disorder.

No medications are FDA-approved for treating SIB. Use pharmacologic interventions to treat underlying disorders, such as depression or anxiety, so that patients are better able to participate in other therapeutic interventions.

Dialectical behavioral therapy (DBT) is the only therapeutic entity shown in controlled trials to successfully treat SIB. Weekly individual psychotherapy and skills training groups focus on:

• regulating emotions
  
• tolerating distress
  
• improving interpersonal relationships
  
• reducing identity confusion and maladaptive cognitions.⁹
  

Other types of therapy—including psychoanalysis, self psychology, object relations, and interpersonal approaches—have a similar understanding of impulsive SIB and employ similar approaches.

Substitute behaviors. The treatment goal is for patients to substitute less destructive behaviors in response to intense emotional states. Some can use techniques such as snapping a rubber band or rubbing ice against the skin, both of which cause discomfort without injury. Patients can listen to music, create art, write in journals, or engage in other physical activities. Each patient has to find a different behavior that works.

Prevention. Although SIB can be done with any object, most adolescents have a preferred method for causing self-injury and may have a “kit” of equipment. Identify and remove any tools the adolescent uses for self-injury. Because SIB is a highly ritualistic behavior, denying access to the preferred tools can help reduce self-injury frequency and convey that the behavior is unacceptable.

One individual should be designated to monitor the adolescent for SIB. Adolescents are seeking trust and do not respond well to constant questions about their behavior. Because some parents can become intrusive, monitoring may be best assigned to the adolescent’s therapist or a less emotional parent.

CASE 3: Scars provoke relapse

Claudia, a musically talented teen with SIB, withdraws from her choir when they choose costumes with short sleeves. She had not engaged in SIB in >1 year but has scarring and cheloid from the cuts. Years later she starts cutting again after laser treatments fail to remove the scars. She is frustrated because she will always have to wear long sleeves.

Risk of relapse. Therapy for SIB tends to be intense and difficult, with frequent relapses. To overcome SIB, the adolescent must want to stop and work hard at other coping strategies.

Treatment is essential, however, because this behavior can last for decades and leave scars that might interfere with future goals. The longer the adolescent has been dependent on the behavior, the more difficult it is to treat.

Related resources

• Levenkron S. Cutting: understanding and overcoming self-mutilation. New York: W.W. Norton & Company; 1998.
  
• Favazza AR. Bodies under siege: self-mutilation and body modification in culture and psychiatry. 2nd ed. Baltimore: The Johns Hopkins University Press; 1996.
  

Disclosure

The author reports no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Alysha, age 15, is an “A” student and top athlete who feels her parents push her to be perfect. After getting a B on a test, she feels overwhelmed by shame and guilt. She locks herself in the bathroom and begins cutting her arm with a razor blade.

Self-injurious behavior (SIB) in adolescents can be associated with internalizing, externalizing, and substance abuse disorders (Table 1). For most practitioners, such as Alysha, a major goal of SIB is to relieve intolerable stress and negative affect.¹

Although this secretive, highly addictive, learned behavior can be difficult to control, some clinical approaches can help these distressed teens and their parents. This article examines the dynamics of SIB, the association between suicidal ideation and SIB, and recommended treatments such as substitute behaviors and dialectic behavioral therapy.

Table 1

Common Axis I disorders among teens with SIB

Axis I disorder	Prevalence*
Any internalizing disorder	52%
Major depressive disorder	42
Posttraumatic stress disorder	24
Generalized anxiety disorder	16
Any externalizing disorder	63
Conduct disorder	49
Oppositional defiant disorder	45
Any substance use disorder	60
Alcohol abuse	18
Alcohol dependence	17
Nicotine dependence	39
Marijuana abuse	13
Marijuana dependence	30
Other substance abuse	3
Other substance dependence	6
*Among a sample of 89 adolescents who engaged in SIB
Source: Nock MD, Joiner TE Jr, Gordon KL, et al. Non-suicidal self-injury among adolescents: diagnostic correlates and relationship to suicide attempts. Psychiatry Res 2006;144(1):68.
Reprinted with permission from Elsevier.

Growing problem in adolescents

SIB is the deliberate infliction of harm to oneself, either internally or externally, without suicidal intent.¹ This behavior is also known as impulsive self-injury, non-suicidal self-injury, self-mutilation, cutting, and self-harm. Once reported primarily in adults with borderline personality disorder, SIB is becoming common among adolescents:

• Among 663 teens in community-sample survey, 46% engaged in some form of SIB in the past year, and 28% engaged in serious, repetitive behaviors.²
  
• 13% to 23% of teenagers have engaged in SIB, according to literature review.³
  
• Children as young as 9 have presented with SIB.
  

Not just cutting. Besides cutting, other methods of self-injury include:

• burning
  
• swallowing objects or substances
  
• hitting oneself with the fist or against an object
  
• cutting circulation to a digit
  
• picking at skin
  
• pulling out hair.
  

Individuals with chronic illnesses can engage in SIB by not complying with treatment, such as a diabetic taking too much or too little insulin or an epileptic not taking medication.

Socially sanctioned behaviors, such as body piercing and tattoos, usually are not considered SIB. They can be used as SIB, however, by teens who impulsively self-pierce or tattoo without appropriate hygiene or anesthetic agents. Their purpose is not to make a fashion statement but to produce pain or discomfort. Cultural behaviors that cause scarring as a rite of passage, such as the Native American Sun Dance, are not considered SIB.¹

Despite increasing prevalence among adolescents, SIB remains a solitary behavior. Based on my clinical experience, teens may share ideas about SIB, but they generally don’t practice it in groups.

SIB psychodynamics

Adults and adolescents with SIB frequently have:

• difficulty regulating emotions
  
• unstable interpersonal relations
  
• limited coping strategies.¹
  

Adolescents with SIB frequently experience anger, and their self-harm can result from turning this anger inward because they are unable to express it toward others. This is seen in a patient who describes a cutting episode that occurred while you were on vacation but sees no connection between your absence and the SIB.

Shame is also common and can be a major barrier to diagnosing SIB. Adolescents who are ashamed of the behavior will go to great lengths to hide it from others, including clinicians. Despite the shame, many adolescents feel unable to stop engaging in SIB because it fulfills a powerful need.

Adults and adolescents who practice SIB most often report their behavior is motivated by affect regulation and tension release.⁴ Some adolescents engage in a different, manipulative form of SIB not to relieve tension but as a threat to prevent loss (Box).⁴

Behavioral reinforcement. An acute stressor—such as parental limits on behavior, feelings of rejection or abandonment by peers, or failing to achieve an unrealistic goal—triggers an escalating, intolerable affect. By experimentation or accident, an adolescent discovers that SIB provides rapid relief of the intolerable state—a calmness that may last for minutes, hours, or days. This relief reinforces the behavior, and the adolescent repeats SIB when faced with the next stressor.

Most individuals with SIB report a similar sequence of events. There is a trigger event, usually involving a real or perceived feeling of loss, rejection, or abandonment. The adolescent tries to resist the impulse to self-harm, feels escalating emotional distress, engages in SIB, and feels immediate relief.⁵

 

Inducing pain or bleeding as a means of relieving intolerable stress may be an attempt to:

• turn emotional pain into more manageable physical pain
  
• direct anger that cannot be expressed at others onto oneself
  
• punish oneself for perceived misdeeds.⁴
  

SIB tends to escalate over time. Those who engage in it may require more frequent or intensive self-injury to achieve relief. Because the emotional state improves quickly and the adolescent feels a sense of control over the behavior, SIB rapidly can become habitual and difficult to interrupt.

Box

CASE 2: Lingering effects of trauma

Melissa, age 13, has been sexually abused by her foster brother. She briefly returns to the home where her foster brother still resides. He has torn up her room, and her mother—who supports the foster brother—has left the room that way for her to find. Melissa returns to her current residence, cuts her arm 15 times, and pierces her tongue.

In therapy, she denies being angry or upset and does not know why she cut her arm or pierced her tongue.

An adolescent with a history of childhood trauma might have difficulty identifying and expressing internal states and developing trusting relations with others. He or she also can have high levels of anger and shame.

Even an adolescent who experienced neglect or loss of attachment object without overt physical or sexual abuse might have trouble establishing a therapeutic alliance because of difficulty with trust.⁶ He or she may have little or no capacity to identify emotional states, which limits insight into the behavior.⁵

Dissociation. Some adolescents use SIB to try to feel something or to bring themselves back from a dissociative state. They report feeling lost, alone, and disconnected from others and themselves. Some report seeing blood as a way of reconnecting with being alive.

Dissociation may occur in adolescents with a history of trauma, particularly in childhood. Abused individuals who engage in SIB may be identifying with the aggressor, attempting to cut away internalized negative images of the abuser or to control anger they are unable to acknowledge.⁶

Clues prompt further assessment

SIB assessment begins with screening for the behavior. If you find any indicators that suggest SIB (Table 2), question the adolescent about self-injury. Many adolescents want help—and will respond accurately to questions about self-injury—but need to be asked. They usually won’t volunteer the information during an initial evaluation.

Once you have identified SIB, explore the behavior. Document:

• number, location, and age of injuries
  
• depth of cuts (if applicable)
  
• signs of infection.
  

Determine if the adolescent needs medical intervention. Discuss how the adolescent causes the injury, and what precipitates it. If possible, obtain some form of safety contract in which the adolescent agrees to not engage in SIB and to notify a designated adult if he or she feels like engaging in SIB or has done so.

Suicide screening. By definition, SIB does not include an intention to die, and most teenagers with SIB will deny suicidal intent. However, because the line between SIB and passive suicide can be thin, careful screening and ongoing monitoring for suicidal ideation and behavior in teens with SIB is essential. In one study:

• 70% of adolescents who engaged in SIB had made one suicide attempt
  
• 55% had made multiple attempts.⁷
  

A separate study found suicidal ideation and depression are keys to identifying adolescents with SIB at risk for suicide attempts.⁸ Because SIB and suicidal ideation/behaviors can co-occur, SIB safety contracts must cover both.

 

Address the parents’ concerns. During your assessment, also focus on the adolescent’s parents. They often are highly distressed, confused, and angry. They typically learn about the SIB from their child’s school counselors or peers and feel betrayed and guilty. They may want to be excessively intrusive and punitive and need support, information, and guidance to address their child’s safety.

Table 2

SIB: Spotting behavioral clues

Adolescents Wearing long sleeves or pants in warm weather Becoming increasingly isolated from family and peers Spending long periods of time in the bathroom or bedroom Deteriorating school performance Decreasing hygiene
Parents Finding blood on teen’s clothing or in unusual places in the home Discovering sharp instruments are missing or placed in unusual locations

Treatment recommendations

Always take SIB in adolescents seriously, and not as something they will “outgrow.” Adolescents with SIB need help modulating affect, stabilizing interpersonal relationships, and developing more adaptive coping strategies and problem-solving skills. Underlying dynamics—especially childhood trauma—must be explored and resolved.

Few evidence-based studies have evaluated SIB treatment in adolescents. Clinicians have extrapolated suggested interventions from the adult literature; however, much of this data was obtained from treating adult women with borderline personality disorder.

No medications are FDA-approved for treating SIB. Use pharmacologic interventions to treat underlying disorders, such as depression or anxiety, so that patients are better able to participate in other therapeutic interventions.

Dialectical behavioral therapy (DBT) is the only therapeutic entity shown in controlled trials to successfully treat SIB. Weekly individual psychotherapy and skills training groups focus on:

• regulating emotions
  
• tolerating distress
  
• improving interpersonal relationships
  
• reducing identity confusion and maladaptive cognitions.⁹
  

Other types of therapy—including psychoanalysis, self psychology, object relations, and interpersonal approaches—have a similar understanding of impulsive SIB and employ similar approaches.

Substitute behaviors. The treatment goal is for patients to substitute less destructive behaviors in response to intense emotional states. Some can use techniques such as snapping a rubber band or rubbing ice against the skin, both of which cause discomfort without injury. Patients can listen to music, create art, write in journals, or engage in other physical activities. Each patient has to find a different behavior that works.

Prevention. Although SIB can be done with any object, most adolescents have a preferred method for causing self-injury and may have a “kit” of equipment. Identify and remove any tools the adolescent uses for self-injury. Because SIB is a highly ritualistic behavior, denying access to the preferred tools can help reduce self-injury frequency and convey that the behavior is unacceptable.

One individual should be designated to monitor the adolescent for SIB. Adolescents are seeking trust and do not respond well to constant questions about their behavior. Because some parents can become intrusive, monitoring may be best assigned to the adolescent’s therapist or a less emotional parent.

CASE 3: Scars provoke relapse

Claudia, a musically talented teen with SIB, withdraws from her choir when they choose costumes with short sleeves. She had not engaged in SIB in >1 year but has scarring and cheloid from the cuts. Years later she starts cutting again after laser treatments fail to remove the scars. She is frustrated because she will always have to wear long sleeves.

Risk of relapse. Therapy for SIB tends to be intense and difficult, with frequent relapses. To overcome SIB, the adolescent must want to stop and work hard at other coping strategies.

Treatment is essential, however, because this behavior can last for decades and leave scars that might interfere with future goals. The longer the adolescent has been dependent on the behavior, the more difficult it is to treat.

Related resources

• Levenkron S. Cutting: understanding and overcoming self-mutilation. New York: W.W. Norton & Company; 1998.
  
• Favazza AR. Bodies under siege: self-mutilation and body modification in culture and psychiatry. 2nd ed. Baltimore: The Johns Hopkins University Press; 1996.
  

Disclosure

The author reports no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Alysha, age 15, is an “A” student and top athlete who feels her parents push her to be perfect. After getting a B on a test, she feels overwhelmed by shame and guilt. She locks herself in the bathroom and begins cutting her arm with a razor blade.

Self-injurious behavior (SIB) in adolescents can be associated with internalizing, externalizing, and substance abuse disorders (Table 1). For most practitioners, such as Alysha, a major goal of SIB is to relieve intolerable stress and negative affect.¹

Although this secretive, highly addictive, learned behavior can be difficult to control, some clinical approaches can help these distressed teens and their parents. This article examines the dynamics of SIB, the association between suicidal ideation and SIB, and recommended treatments such as substitute behaviors and dialectic behavioral therapy.

Table 1

Common Axis I disorders among teens with SIB

Axis I disorder	Prevalence*
Any internalizing disorder	52%
Major depressive disorder	42
Posttraumatic stress disorder	24
Generalized anxiety disorder	16
Any externalizing disorder	63
Conduct disorder	49
Oppositional defiant disorder	45
Any substance use disorder	60
Alcohol abuse	18
Alcohol dependence	17
Nicotine dependence	39
Marijuana abuse	13
Marijuana dependence	30
Other substance abuse	3
Other substance dependence	6
*Among a sample of 89 adolescents who engaged in SIB
Source: Nock MD, Joiner TE Jr, Gordon KL, et al. Non-suicidal self-injury among adolescents: diagnostic correlates and relationship to suicide attempts. Psychiatry Res 2006;144(1):68.
Reprinted with permission from Elsevier.

Growing problem in adolescents

SIB is the deliberate infliction of harm to oneself, either internally or externally, without suicidal intent.¹ This behavior is also known as impulsive self-injury, non-suicidal self-injury, self-mutilation, cutting, and self-harm. Once reported primarily in adults with borderline personality disorder, SIB is becoming common among adolescents:

• Among 663 teens in community-sample survey, 46% engaged in some form of SIB in the past year, and 28% engaged in serious, repetitive behaviors.²
  
• 13% to 23% of teenagers have engaged in SIB, according to literature review.³
  
• Children as young as 9 have presented with SIB.
  

Not just cutting. Besides cutting, other methods of self-injury include:

• burning
  
• swallowing objects or substances
  
• hitting oneself with the fist or against an object
  
• cutting circulation to a digit
  
• picking at skin
  
• pulling out hair.
  

Individuals with chronic illnesses can engage in SIB by not complying with treatment, such as a diabetic taking too much or too little insulin or an epileptic not taking medication.

Socially sanctioned behaviors, such as body piercing and tattoos, usually are not considered SIB. They can be used as SIB, however, by teens who impulsively self-pierce or tattoo without appropriate hygiene or anesthetic agents. Their purpose is not to make a fashion statement but to produce pain or discomfort. Cultural behaviors that cause scarring as a rite of passage, such as the Native American Sun Dance, are not considered SIB.¹

Despite increasing prevalence among adolescents, SIB remains a solitary behavior. Based on my clinical experience, teens may share ideas about SIB, but they generally don’t practice it in groups.

SIB psychodynamics

Adults and adolescents with SIB frequently have:

• difficulty regulating emotions
  
• unstable interpersonal relations
  
• limited coping strategies.¹
  

Adolescents with SIB frequently experience anger, and their self-harm can result from turning this anger inward because they are unable to express it toward others. This is seen in a patient who describes a cutting episode that occurred while you were on vacation but sees no connection between your absence and the SIB.

Shame is also common and can be a major barrier to diagnosing SIB. Adolescents who are ashamed of the behavior will go to great lengths to hide it from others, including clinicians. Despite the shame, many adolescents feel unable to stop engaging in SIB because it fulfills a powerful need.

Adults and adolescents who practice SIB most often report their behavior is motivated by affect regulation and tension release.⁴ Some adolescents engage in a different, manipulative form of SIB not to relieve tension but as a threat to prevent loss (Box).⁴

Behavioral reinforcement. An acute stressor—such as parental limits on behavior, feelings of rejection or abandonment by peers, or failing to achieve an unrealistic goal—triggers an escalating, intolerable affect. By experimentation or accident, an adolescent discovers that SIB provides rapid relief of the intolerable state—a calmness that may last for minutes, hours, or days. This relief reinforces the behavior, and the adolescent repeats SIB when faced with the next stressor.

Most individuals with SIB report a similar sequence of events. There is a trigger event, usually involving a real or perceived feeling of loss, rejection, or abandonment. The adolescent tries to resist the impulse to self-harm, feels escalating emotional distress, engages in SIB, and feels immediate relief.⁵

 

Inducing pain or bleeding as a means of relieving intolerable stress may be an attempt to:

• turn emotional pain into more manageable physical pain
  
• direct anger that cannot be expressed at others onto oneself
  
• punish oneself for perceived misdeeds.⁴
  

SIB tends to escalate over time. Those who engage in it may require more frequent or intensive self-injury to achieve relief. Because the emotional state improves quickly and the adolescent feels a sense of control over the behavior, SIB rapidly can become habitual and difficult to interrupt.

Box

CASE 2: Lingering effects of trauma

Melissa, age 13, has been sexually abused by her foster brother. She briefly returns to the home where her foster brother still resides. He has torn up her room, and her mother—who supports the foster brother—has left the room that way for her to find. Melissa returns to her current residence, cuts her arm 15 times, and pierces her tongue.

In therapy, she denies being angry or upset and does not know why she cut her arm or pierced her tongue.

An adolescent with a history of childhood trauma might have difficulty identifying and expressing internal states and developing trusting relations with others. He or she also can have high levels of anger and shame.

Even an adolescent who experienced neglect or loss of attachment object without overt physical or sexual abuse might have trouble establishing a therapeutic alliance because of difficulty with trust.⁶ He or she may have little or no capacity to identify emotional states, which limits insight into the behavior.⁵

Dissociation. Some adolescents use SIB to try to feel something or to bring themselves back from a dissociative state. They report feeling lost, alone, and disconnected from others and themselves. Some report seeing blood as a way of reconnecting with being alive.

Dissociation may occur in adolescents with a history of trauma, particularly in childhood. Abused individuals who engage in SIB may be identifying with the aggressor, attempting to cut away internalized negative images of the abuser or to control anger they are unable to acknowledge.⁶

Clues prompt further assessment

SIB assessment begins with screening for the behavior. If you find any indicators that suggest SIB (Table 2), question the adolescent about self-injury. Many adolescents want help—and will respond accurately to questions about self-injury—but need to be asked. They usually won’t volunteer the information during an initial evaluation.

Once you have identified SIB, explore the behavior. Document:

• number, location, and age of injuries
  
• depth of cuts (if applicable)
  
• signs of infection.
  

Determine if the adolescent needs medical intervention. Discuss how the adolescent causes the injury, and what precipitates it. If possible, obtain some form of safety contract in which the adolescent agrees to not engage in SIB and to notify a designated adult if he or she feels like engaging in SIB or has done so.

Suicide screening. By definition, SIB does not include an intention to die, and most teenagers with SIB will deny suicidal intent. However, because the line between SIB and passive suicide can be thin, careful screening and ongoing monitoring for suicidal ideation and behavior in teens with SIB is essential. In one study:

• 70% of adolescents who engaged in SIB had made one suicide attempt
  
• 55% had made multiple attempts.⁷
  

A separate study found suicidal ideation and depression are keys to identifying adolescents with SIB at risk for suicide attempts.⁸ Because SIB and suicidal ideation/behaviors can co-occur, SIB safety contracts must cover both.

 

Address the parents’ concerns. During your assessment, also focus on the adolescent’s parents. They often are highly distressed, confused, and angry. They typically learn about the SIB from their child’s school counselors or peers and feel betrayed and guilty. They may want to be excessively intrusive and punitive and need support, information, and guidance to address their child’s safety.

Table 2

SIB: Spotting behavioral clues

Adolescents Wearing long sleeves or pants in warm weather Becoming increasingly isolated from family and peers Spending long periods of time in the bathroom or bedroom Deteriorating school performance Decreasing hygiene
Parents Finding blood on teen’s clothing or in unusual places in the home Discovering sharp instruments are missing or placed in unusual locations

Treatment recommendations

Always take SIB in adolescents seriously, and not as something they will “outgrow.” Adolescents with SIB need help modulating affect, stabilizing interpersonal relationships, and developing more adaptive coping strategies and problem-solving skills. Underlying dynamics—especially childhood trauma—must be explored and resolved.

Few evidence-based studies have evaluated SIB treatment in adolescents. Clinicians have extrapolated suggested interventions from the adult literature; however, much of this data was obtained from treating adult women with borderline personality disorder.

No medications are FDA-approved for treating SIB. Use pharmacologic interventions to treat underlying disorders, such as depression or anxiety, so that patients are better able to participate in other therapeutic interventions.

Dialectical behavioral therapy (DBT) is the only therapeutic entity shown in controlled trials to successfully treat SIB. Weekly individual psychotherapy and skills training groups focus on:

• regulating emotions
  
• tolerating distress
  
• improving interpersonal relationships
  
• reducing identity confusion and maladaptive cognitions.⁹
  

Other types of therapy—including psychoanalysis, self psychology, object relations, and interpersonal approaches—have a similar understanding of impulsive SIB and employ similar approaches.

Substitute behaviors. The treatment goal is for patients to substitute less destructive behaviors in response to intense emotional states. Some can use techniques such as snapping a rubber band or rubbing ice against the skin, both of which cause discomfort without injury. Patients can listen to music, create art, write in journals, or engage in other physical activities. Each patient has to find a different behavior that works.

Prevention. Although SIB can be done with any object, most adolescents have a preferred method for causing self-injury and may have a “kit” of equipment. Identify and remove any tools the adolescent uses for self-injury. Because SIB is a highly ritualistic behavior, denying access to the preferred tools can help reduce self-injury frequency and convey that the behavior is unacceptable.

One individual should be designated to monitor the adolescent for SIB. Adolescents are seeking trust and do not respond well to constant questions about their behavior. Because some parents can become intrusive, monitoring may be best assigned to the adolescent’s therapist or a less emotional parent.

CASE 3: Scars provoke relapse

Claudia, a musically talented teen with SIB, withdraws from her choir when they choose costumes with short sleeves. She had not engaged in SIB in >1 year but has scarring and cheloid from the cuts. Years later she starts cutting again after laser treatments fail to remove the scars. She is frustrated because she will always have to wear long sleeves.

Risk of relapse. Therapy for SIB tends to be intense and difficult, with frequent relapses. To overcome SIB, the adolescent must want to stop and work hard at other coping strategies.

Treatment is essential, however, because this behavior can last for decades and leave scars that might interfere with future goals. The longer the adolescent has been dependent on the behavior, the more difficult it is to treat.

Related resources

• Levenkron S. Cutting: understanding and overcoming self-mutilation. New York: W.W. Norton & Company; 1998.
  
• Favazza AR. Bodies under siege: self-mutilation and body modification in culture and psychiatry. 2nd ed. Baltimore: The Johns Hopkins University Press; 1996.
  

Disclosure

The author reports no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Ms. K, age 25, is 6 weeks pregnant and is taking medications for generalized anxiety disorder (GAD). When she was diagnosed with GAD at age 19, her symptoms included 6 months of excessive anxiety—insomnia, fatigue, difficulty with concentration, and psychomotor agitation—without mood symptoms. These symptoms interfered greatly with her schoolwork and other daily activities.

For 6 years Ms. K has been taking the selective serotonin reuptake inhibitor (SSRI) paroxetine, 15 mg/d, and the benzodiazepine clonazepam, 0.5 mg as needed, with good symptom control. Now that she is pregnant and her primary care doctor has refused to continue these medications, she is seeking treatment and advice.

Not enough is known about how to safely treat anxiety disorders during pregnancy, and physicians are not sure what to do with patients such as Ms. K. Without evidence-based guidelines, we feel anxious about potential risks to mother and fetus as we try to provide appropriate drug therapy.

To help you and your patients weigh the risks and benefits of perinatal treatments for anxiety disorders, this article briefly summarizes the evidence on:

• anxiety disorders’ natural history during pregnancy
  
• how untreated maternal anxiety affects the fetus
  
• nonpharmacologic therapies for anxiety disorders
  
• a plan to manage fetal risks by staggering SSRI and benzodiazepine use during the first and third trimesters.
  

Anxiety during pregnancy

Nearly one-third of women experience an anxiety disorder during their lives, with peak onset during childbearing years.^1,2 Compared with research on perinatal depression, far fewer studies have examined anxiety disorders’ onset, presentation, prevalence, and treatment.¹

The literature includes no studies of the course of preexisting GAD or posttraumatic stress disorder (PTSD) and no evidence that symptoms of preexisting obsessive-compulsive disorder (OCD) change during pregnancy. Some studies of panic disorder show symptoms improving during pregnancy, whereas others do not (Table 1).¹

One small study done in late pregnancy found a significant association between the prevalence of an anxiety disorder, maternal primiparity, and comorbid medical conditions. Thus, a woman in her first pregnancy may be at increased risk to develop an anxiety disorder if she has a comorbid medical condition.³ As in the case of Ms. K, however, continuation of preexisting anxiety appears more likely than onset of a new anxiety disorder during pregnancy.

Table 1

How pregnancy affects the course of 4 anxiety disorders

Disorder	Prevalence	Effect
Generalized anxiety disorder (GAD)	8.5% of women experience GAD during the third trimester, compared with a 5% prevalence in the general population	No studies have reported on the course of GAD in pregnant women with preexisting disorder
Obsessive-compulsive disorder (OCD)	2% to 12% of OCD outpatients of childbearing age report onset during pregnancy	Preexisting OCD usually shows no change during pregnancy but may worsen postpartum
Panic disorder (PD)	1.3% to 2% in pregnant women, compared with 1.5% to 3.5% in the general population	Panic symptoms in women with preexisting PD may improve during pregnancy and worsen postpartum
Posttraumatic stress disorder (PTSD)	2.3% to 7.7% in pregnant women and 0% to 6.9% postpartum, compared with 1% to 14% in the community	No studies have reported on the course of PTSD in pregnant women with preexisting disorder
Source: References 1,2

Fetal risks from maternal anxiety

Fetal risk from severe maternal anxiety is not zero. Offspring born to high-anxiety mothers exhibit neurobehavioral differences compared with offspring of calmer mothers. Changes in high-anxiety mothers’ offspring include:

• altered EEG activation and vagal tone
  
• increased time in deep sleep and less time in active alert states
  
• lower performance on the Brazelton Neonatal Behavior Assessment Scale.⁴
  

A cohort study by Teixeira et al⁵ found an association between maternal anxiety in pregnancy and uterine artery resistance, suggesting a possible mechanism by which a mother’s psychologic state may affect fetal development. High anxiety and self-reported life stress during pregnancy also are associated consistently with abnormal, high-frequency heart rate variability in infants—a finding linked with negative infant behavior and later adult hostility.⁶

Exposure to maternal high anxiety has been associated with mental developmental delays in infants and increased risk for behavioral and emotional problems in young children.^7-10 Anxiety may not directly cause intrauterine growth retardation and preterm delivery, but it is significantly associated with prenatal tobacco, alcohol, and narcotics use—which predicts these and other negative neonatal outcomes.¹¹

Anxiety during pregnancy is a risk factor for postnatal depressive symptoms, independent of depressed mood and family or marital stressors during pregnancy.¹² Mothers with postpartum depression appear less able to respond sensitively and competently to their newborns, and these infants may be at increased risk of behavioral, emotional, and cognitive problems.

 

^7,13

CASE CONTINUED: ‘Stay the course’

Ms. K worries that she could not tolerate recurrence of her anxiety symptoms and wishes to continue both medications. Her husband concurs, but they want to minimize potential risks to their baby. You discuss the options for treating anxiety symptoms during pregnancy, including medications, psychotherapy, and behavioral treatments.

Treatment decisions

Ideally you’ll begin treating anxiety disorders in women of childbearing age with preconception psychoeducation. Explaining the risks of medications if she were to become pregnant and asking about the contraception she is using are de rigueur. Psychotherapy is low risk to the fetus and is considered first choice for treating mild to moderate anxiety in women of childbearing age who plan to become pregnant (Box).^1,14-17

Box

Psychotherapy alone is inadequate, however, for the many patients—such as Ms. K—who present already pregnant with a history of moderate to severe anxiety. Adjunctive psychotropic therapy—along with various nonmedication therapies—is warranted for patients whose social or occupational functioning would be substantially impaired by suboptimal control of anxiety during pregnancy.

Because Ms. K wishes to continue taking paroxetine and clonazepam, what can you tell her about the risks and benefits of SSRIs and benzodiazepines during pregnancy?

SSRIs in pregnancy

Teratogenicity. Compared with benzodiazepines, SSRIs have been considered agents of choice for use during pregnancy because of a lower risk of teratogenic effects.¹⁵ Paroxetine, however, appears to pose a greater risk for teratogenicity than other SSRIs.

An increased risk for fetal ventricular and/or atrial septal defects has been associated with first-trimester exposure to paroxetine, but no other SSRI.¹⁸ First trimester exposure to paroxetine at doses averaging 25 mg/d has been associated with statistically significant risks of major congenital anomalies (2-fold increase) and major cardiac anomalies (3-fold increase),¹⁹ although other studies have failed to reproduce this finding. A meta-analysis of 7 studies by Bar-Oz et al²⁰ found an association between first-trimester paroxetine exposure and a significant increase in risk for cardiac malformations (odds ratio [OR] 1.72; 95% CI,1.22-2.42).

The overall rate of fetal malformations from SSRIs appears to be low, although most studies have examined only fluoxetine or paroxetine. Some studies have reported various malformations with fluoxetine or sertraline, but others have not. In Finland, a population-based study found no increase in rate of major congenital malformations in offspring of 1,782 women who filled prescriptions for SSRIs during pregnancy, compared with the general population rate of 1% to 3%.²¹

Neurobehavioral effects. SSRI exposure during fetal life has shown no long-term neurobehavioral effects. A blinded prospective study by Nulman et al²² found no differences in global IQ scores, language development, or behavioral development among children age ≤5 who were exposed in utero to fluoxetine (n=40) or a tricyclic antidepressant (n=46), compared with unexposed children of nondepressed mothers (n=36). Similarly, using reports from teachers and clinical measures of internalizing behaviors, Misri et al¹⁰ found no increase in depression, anxiety, or withdrawal in 4-year-olds with prenatal exposure to SSRIs (n=22), compared with nonexposed children (n=14).

Pulmonary hypertension. SSRI exposure in later pregnancy may increase the rate of persistent pulmonary hypertension of the newborn (PPHN), which occurs in 1 to 2 infants per 1,000 live births. PPHN showed a statistically significant association with late prenatal SSRI exposure (OR 6.1) in a study that controlled for maternal smoking, body mass index, and diabetes.²³ PPHN occurred in approximately 1% of infants exposed to SSRIs in late pregnancy. PPHN rates were not affected by maternal depression/anxiety, non-SSRI antidepressant exposure throughout pregnancy, or SSRI exposure during early pregnancy only.

 

Toxicity and withdrawal syndromes. Infants of women who continue to take SSRIs just before delivery can develop toxicity or withdrawal syndromes. Occurrence of either syndrome depends on SSRI half-life, serum concentration, and the pharmacodynamics of other medications given during pregnancy and labor.²⁴

Discontinuation syndromes can occur in SSRI-exposed neonates within a few hours or days after birth and last up to 1 month after delivery, depending on the infant’s susceptibility.²⁵ Nearly two-thirds of suspected SSRI-induced neonatal withdrawal syndromes have been associated with paroxetine, although all SSRIs appear be associated with some risk.²⁶ Several trials, including a recent prospective study, found prenatal antidepressant use associated with lower gestational age at birth and increased risk of preterm birth.²⁷

A prospective study compared the effects of maternal SSRI use on behavioral state, sleep, motor activity, and heart rate variability in 17 exposed vs 17 nonexposed matched neonates. In the first 1 to 2 weeks of life, SSRI-exposed neonates showed:

• greater tremulousness
  
• less flexible and dampened state regulation
  
• more time in uninterrupted REM sleep
  
• more frequent startles or sudden arousals
  
• greater generalized motor activity
  
• greater autonomic dysregulation.²⁸
  

In a cohort study of 60 neonates exposed to SSRIs in utero, 30% met diagnostic criteria for neonatal abstinence syndrome. The most common discontinuation symptoms were:

• tremor (37/60)
  
• GI disturbances (34/60)—including exaggerated sucking, poor feeding, regurgitation, vomiting, and loose stools
  
• sleep disturbance (21/60).
  

Other symptoms included irritability, constant crying, shivering, increased tone, convulsions, jitteriness, poor gaze control, vomiting, myoclonus, and lethargy.²⁵

Recommendations. The perception that SSRIs have low fetal toxicity has guided prescribing practices in recent years. Newer evidence shows, however, that fetal exposure to SSRIs may have some adverse effects, including lower birth weight and early delivery. First-trimester paroxetine use has been associated with increased risk for fetal ventricular and/or atrial septal defects.

Discuss these risks with the patient when you consider starting or continuing SSRI use during pregnancy.²⁴ If you prescribe an SSRI, use the minimum effective dosage and avoid paroxetine during pregnancy.¹⁸

To reduce the risk for PPHN, early delivery, and neonatal withdrawal syndromes, taper and discontinue the SSRI during the third trimester. Restarting the SSRI soon after delivery is the most effective way to prevent recurrence of anxiety symptoms or postpartum depression.

Benzodiazepines

Teratogenicity. Like SSRIs, benzodiazepines cross the placenta to the fetus.²⁹ Benzodiazepine teratogenicity remains controversial.⁸ Some—but not all—data show a small but significant increased risk for major malformations/oral cleft malformations with first-trimester benzodiazepine exposure.

A Medline literature search from 1966 to 2000 found not enough information to determine whether potential benefits of benzodiazepines to the mother outweigh risks to the fetus.²⁹ An ambitious meta-analysis of >1,400 studies by Dolovich et al³⁰ found a small association between fetal exposure to benzodiazepines and major malformations/cleft palate, but only in pooled data from case-controlled studies. No association was found between fetal exposure to benzodiazepines and malformations/cleft palate in pooled data from cohort studies.

A 32-month, hospital-based surveillance program of 28,565 births found no increase in the rate of major malformations in 43 infants exposed to clonazepam monotherapy—33 (77%) in the first trimester.³¹ Thus, the risk of major malformations/cleft palate with the use of benzodiazepines in the first trimester appears to be low.

Toxicity and withdrawal syndromes. Neonatal benzodiazepine toxicity and withdrawal syndromes have been reported in studies and case reports. Although these syndromes occur, they do not affect all infants with late third-trimester benzodiazepine exposure. Prevalence rates have not been calculated.³²

• Neonatal toxicity (“floppy infant syndrome”)—characterized by hypothermia, lethargy, poor respiratory effort, and feeding difficulties—occurs after maternal benzodiazepine use just before delivery.⁸
  
• Neonatal withdrawal may be caused by very late, third trimester exposure to benzodiazepines. Symptoms—which can persist ≤3 months after delivery—include restlessness, irritability, abnormal sleep patterns, suckling difficulties, growth retardation, hypertonia, hyperreflexia, tremulousness, apnea, diarrhea, and vomiting.^8,29
  

Recommendations. When possible, avoid benzodiazepines in the first trimester because of possible teratogenicity and then again late in the third trimester before delivery because of neonatal withdrawal syndromes. To reduce as much as possible the small risk of a benzodiazepine-related fetal malformation/cleft palate, wean the mother from benzodiazepines before conception. After the first trimester, the benzodiazepine can be restarted if necessary.²⁹

To minimize neonatal withdrawal, gradually taper the mother’s benzodiazepine before delivery.²⁹ Because the baby’s due date is calculated to be ±2 weeks before delivery, begin this taper 3 to 4 weeks before the due date and discontinue at least 1 week before delivery. Breastfeeding while taking benzodiazepines is not recommended because of the risk of over-sedating the infant.

 

A rational approach

Both benzodiazepines and SSRIs are associated with low but demonstrated risks to the fetus when used during pregnancy (Table 2).^{19,20,23,25,30,33} Use these medications to manage a patient’s anxiety only if the clinical benefit to the mother justifies the potential risks to the fetus.²⁹

A staggered combination of SSRIs during the first 2 trimesters and benzodiazepines during the last 2 trimesters can help balance the risks and benefits of pharmacotherapy of anxiety disorders during pregnancy (Table 3).

Frankly discuss with your patient the risks and benefits in the context of her perceived need for symptom control to sustain her level of functioning. You could document this discussion in the progress note as “R, B, A, and pt C,” signifying that risks, benefits, and alternatives were discussed, and the patient consented. If possible, include the patient’s husband, partner, or parent in this discussion.

Table 2

Risks of SSRIs vs benzodiazepines during pregnancy stages

Pregnancy stage when given	Fetal risk	SSRIs	Benzodiazepines
First trimester*	Teratogenicity	Paroxetine use associated with 2-fold increased risk of major congenital anomalies and 3-fold increased risk of major cardiac anomalies;19 meta-analysis calculated significant risk of cardiac malformations (odds ratio 1.72; population prevalence = 13.4/1,000 births)20,33	Meta-analysis of case control studies showed increased risk of major malformations/cleft palate (odds ratio 3.01; population prevalence = 10 to 20/1,000 births); no association seen in cohort studies30
Third trimester	PPHN	Case control study showed 3.7% of infants with PPHN were exposed to SSRIs vs 0.7% of controls; adjusted odds ratio 6.1, absolute risk to exposed population = 6 to 12/1,000 births)23
Perinatal and long-term effects	Toxicity/withdrawal syndromes	Cohort study of 60 infants concluded prevalence of discontinuation syndromes is 30% in neonates with third trimester SSRI exposure25	Neonatal toxicity (“floppy infant syndrome”) and neonatal withdrawal reported with maternal benzodiazepine use in late third trimester; prevalence unknown†
	Preterm birth, serotonin withdrawal syndromes, CNS effects, long-term neurobehavioral effects	Unknown†	Unknown†
PPHN: persistent pulmonary hypertension of the newborn; SSRIs: selective serotonin reuptake inhibitors
* Available data indicate that first-trimester exposure to SSRIs (other than paroxetine) and benzodiazepines may increase the relative risk for congenital anomalies, but the absolute risk of having a child with an anomaly is small.
† Some case reports, but published literature is insufficient to determine prevalence or magnitude of risk.

Table 3

Staggered, combination therapy for anxiety disorders during pregnancy

Pregnancy stage	Recommended to manage risks to mother and fetus
First trimester	SSRI (not paroxetine) No benzodiazepines Nondrug therapies*
Second trimester	SSRI (not paroxetine) Can use benzodiazepine if needed Nondrug therapies*
Third trimester	Taper off SSRI by 1 to 2 months before due date Can use benzodiazepine until 2 weeks before due date Nondrug therapies*
SSRI: selective serotonin reuptake inhibitor
* Nondrug therapies can include prenatal exercise, sleep hygiene, relaxation, and psychotherapy (cognitive-behavioral therapy, interpersonal therapy, supportive therapy, family/couples therapy)

CASE CONTINUED: CBT plus medication

Ms. K and her husband are open to adding weekly cognitive-behavioral therapy (CBT) for anxiety as long as she can continue her medications. You discuss the evidence regarding potential neonatal risks with paroxetine and clonazepam treatment. Because Ms. K is 6 weeks pregnant, you outline a plan for a rapid cross-taper off paroxetine and onto fluoxetine, 10 to 30 mg/d, explaining that paroxetine might pose a greater first-trimester risk of major congenital malformations and cardiac malformations. You discuss possible side effects of fluoxetine and explain a plan to taper off fluoxetine during the third trimester to reduce the risk of PPHN, early delivery, and withdrawal in the newborn.

Because Ms. K has been taking clonazepam at only 0.5 mg 1 to 2 times per week, you instruct her to stop taking the benzodiazepine for the next 6 weeks until she is through her first trimester. You also reassure her that she can use clonazepam after the first trimester, if necessary, as long as she agrees to taper off completely 1 to 2 weeks before to her due date.

You refer her to a CBT therapist and emphasize the importance of CBT, relaxation, and sleep hygiene—as well as support from her husband, family, and friends—to reduce her stress and facilitate the medication taper during her third trimester. You plan to see her monthly and co-manage her care with the CBT therapist and Ob/Gyn. You document this discussion in her medical record as evidence of informed consent.

Related resources

• Baby Center. Managing stress and anxiety during pregnancy. Patient information. www.babycenter.com/0_managing-stress-and-anxiety-during-pregnancy_1683.bc.
  
• Organization of Teratology Information Specialists (OTIS). www.otispregnancy.org.
  

Drug brand names

• Clonazepam • Klonopin
  
• Paroxetine • Paxil
  
• Fluoxetine • Prozac
  
• Sertraline • Zoloft
  

Disclosures

The authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Ms. K, age 25, is 6 weeks pregnant and is taking medications for generalized anxiety disorder (GAD). When she was diagnosed with GAD at age 19, her symptoms included 6 months of excessive anxiety—insomnia, fatigue, difficulty with concentration, and psychomotor agitation—without mood symptoms. These symptoms interfered greatly with her schoolwork and other daily activities.

For 6 years Ms. K has been taking the selective serotonin reuptake inhibitor (SSRI) paroxetine, 15 mg/d, and the benzodiazepine clonazepam, 0.5 mg as needed, with good symptom control. Now that she is pregnant and her primary care doctor has refused to continue these medications, she is seeking treatment and advice.

Not enough is known about how to safely treat anxiety disorders during pregnancy, and physicians are not sure what to do with patients such as Ms. K. Without evidence-based guidelines, we feel anxious about potential risks to mother and fetus as we try to provide appropriate drug therapy.

To help you and your patients weigh the risks and benefits of perinatal treatments for anxiety disorders, this article briefly summarizes the evidence on:

• anxiety disorders’ natural history during pregnancy
  
• how untreated maternal anxiety affects the fetus
  
• nonpharmacologic therapies for anxiety disorders
  
• a plan to manage fetal risks by staggering SSRI and benzodiazepine use during the first and third trimesters.
  

Anxiety during pregnancy

Nearly one-third of women experience an anxiety disorder during their lives, with peak onset during childbearing years.^1,2 Compared with research on perinatal depression, far fewer studies have examined anxiety disorders’ onset, presentation, prevalence, and treatment.¹

The literature includes no studies of the course of preexisting GAD or posttraumatic stress disorder (PTSD) and no evidence that symptoms of preexisting obsessive-compulsive disorder (OCD) change during pregnancy. Some studies of panic disorder show symptoms improving during pregnancy, whereas others do not (Table 1).¹

One small study done in late pregnancy found a significant association between the prevalence of an anxiety disorder, maternal primiparity, and comorbid medical conditions. Thus, a woman in her first pregnancy may be at increased risk to develop an anxiety disorder if she has a comorbid medical condition.³ As in the case of Ms. K, however, continuation of preexisting anxiety appears more likely than onset of a new anxiety disorder during pregnancy.

Table 1

How pregnancy affects the course of 4 anxiety disorders

Disorder	Prevalence	Effect
Generalized anxiety disorder (GAD)	8.5% of women experience GAD during the third trimester, compared with a 5% prevalence in the general population	No studies have reported on the course of GAD in pregnant women with preexisting disorder
Obsessive-compulsive disorder (OCD)	2% to 12% of OCD outpatients of childbearing age report onset during pregnancy	Preexisting OCD usually shows no change during pregnancy but may worsen postpartum
Panic disorder (PD)	1.3% to 2% in pregnant women, compared with 1.5% to 3.5% in the general population	Panic symptoms in women with preexisting PD may improve during pregnancy and worsen postpartum
Posttraumatic stress disorder (PTSD)	2.3% to 7.7% in pregnant women and 0% to 6.9% postpartum, compared with 1% to 14% in the community	No studies have reported on the course of PTSD in pregnant women with preexisting disorder
Source: References 1,2

Fetal risks from maternal anxiety

Fetal risk from severe maternal anxiety is not zero. Offspring born to high-anxiety mothers exhibit neurobehavioral differences compared with offspring of calmer mothers. Changes in high-anxiety mothers’ offspring include:

• altered EEG activation and vagal tone
  
• increased time in deep sleep and less time in active alert states
  
• lower performance on the Brazelton Neonatal Behavior Assessment Scale.⁴
  

A cohort study by Teixeira et al⁵ found an association between maternal anxiety in pregnancy and uterine artery resistance, suggesting a possible mechanism by which a mother’s psychologic state may affect fetal development. High anxiety and self-reported life stress during pregnancy also are associated consistently with abnormal, high-frequency heart rate variability in infants—a finding linked with negative infant behavior and later adult hostility.⁶

Exposure to maternal high anxiety has been associated with mental developmental delays in infants and increased risk for behavioral and emotional problems in young children.^7-10 Anxiety may not directly cause intrauterine growth retardation and preterm delivery, but it is significantly associated with prenatal tobacco, alcohol, and narcotics use—which predicts these and other negative neonatal outcomes.¹¹

Anxiety during pregnancy is a risk factor for postnatal depressive symptoms, independent of depressed mood and family or marital stressors during pregnancy.¹² Mothers with postpartum depression appear less able to respond sensitively and competently to their newborns, and these infants may be at increased risk of behavioral, emotional, and cognitive problems.

 

^7,13

CASE CONTINUED: ‘Stay the course’

Ms. K worries that she could not tolerate recurrence of her anxiety symptoms and wishes to continue both medications. Her husband concurs, but they want to minimize potential risks to their baby. You discuss the options for treating anxiety symptoms during pregnancy, including medications, psychotherapy, and behavioral treatments.

Treatment decisions

Ideally you’ll begin treating anxiety disorders in women of childbearing age with preconception psychoeducation. Explaining the risks of medications if she were to become pregnant and asking about the contraception she is using are de rigueur. Psychotherapy is low risk to the fetus and is considered first choice for treating mild to moderate anxiety in women of childbearing age who plan to become pregnant (Box).^1,14-17

Box

Psychotherapy alone is inadequate, however, for the many patients—such as Ms. K—who present already pregnant with a history of moderate to severe anxiety. Adjunctive psychotropic therapy—along with various nonmedication therapies—is warranted for patients whose social or occupational functioning would be substantially impaired by suboptimal control of anxiety during pregnancy.

Because Ms. K wishes to continue taking paroxetine and clonazepam, what can you tell her about the risks and benefits of SSRIs and benzodiazepines during pregnancy?

SSRIs in pregnancy

Teratogenicity. Compared with benzodiazepines, SSRIs have been considered agents of choice for use during pregnancy because of a lower risk of teratogenic effects.¹⁵ Paroxetine, however, appears to pose a greater risk for teratogenicity than other SSRIs.

An increased risk for fetal ventricular and/or atrial septal defects has been associated with first-trimester exposure to paroxetine, but no other SSRI.¹⁸ First trimester exposure to paroxetine at doses averaging 25 mg/d has been associated with statistically significant risks of major congenital anomalies (2-fold increase) and major cardiac anomalies (3-fold increase),¹⁹ although other studies have failed to reproduce this finding. A meta-analysis of 7 studies by Bar-Oz et al²⁰ found an association between first-trimester paroxetine exposure and a significant increase in risk for cardiac malformations (odds ratio [OR] 1.72; 95% CI,1.22-2.42).

The overall rate of fetal malformations from SSRIs appears to be low, although most studies have examined only fluoxetine or paroxetine. Some studies have reported various malformations with fluoxetine or sertraline, but others have not. In Finland, a population-based study found no increase in rate of major congenital malformations in offspring of 1,782 women who filled prescriptions for SSRIs during pregnancy, compared with the general population rate of 1% to 3%.²¹

Neurobehavioral effects. SSRI exposure during fetal life has shown no long-term neurobehavioral effects. A blinded prospective study by Nulman et al²² found no differences in global IQ scores, language development, or behavioral development among children age ≤5 who were exposed in utero to fluoxetine (n=40) or a tricyclic antidepressant (n=46), compared with unexposed children of nondepressed mothers (n=36). Similarly, using reports from teachers and clinical measures of internalizing behaviors, Misri et al¹⁰ found no increase in depression, anxiety, or withdrawal in 4-year-olds with prenatal exposure to SSRIs (n=22), compared with nonexposed children (n=14).

Pulmonary hypertension. SSRI exposure in later pregnancy may increase the rate of persistent pulmonary hypertension of the newborn (PPHN), which occurs in 1 to 2 infants per 1,000 live births. PPHN showed a statistically significant association with late prenatal SSRI exposure (OR 6.1) in a study that controlled for maternal smoking, body mass index, and diabetes.²³ PPHN occurred in approximately 1% of infants exposed to SSRIs in late pregnancy. PPHN rates were not affected by maternal depression/anxiety, non-SSRI antidepressant exposure throughout pregnancy, or SSRI exposure during early pregnancy only.

 

Toxicity and withdrawal syndromes. Infants of women who continue to take SSRIs just before delivery can develop toxicity or withdrawal syndromes. Occurrence of either syndrome depends on SSRI half-life, serum concentration, and the pharmacodynamics of other medications given during pregnancy and labor.²⁴

Discontinuation syndromes can occur in SSRI-exposed neonates within a few hours or days after birth and last up to 1 month after delivery, depending on the infant’s susceptibility.²⁵ Nearly two-thirds of suspected SSRI-induced neonatal withdrawal syndromes have been associated with paroxetine, although all SSRIs appear be associated with some risk.²⁶ Several trials, including a recent prospective study, found prenatal antidepressant use associated with lower gestational age at birth and increased risk of preterm birth.²⁷

A prospective study compared the effects of maternal SSRI use on behavioral state, sleep, motor activity, and heart rate variability in 17 exposed vs 17 nonexposed matched neonates. In the first 1 to 2 weeks of life, SSRI-exposed neonates showed:

• greater tremulousness
  
• less flexible and dampened state regulation
  
• more time in uninterrupted REM sleep
  
• more frequent startles or sudden arousals
  
• greater generalized motor activity
  
• greater autonomic dysregulation.²⁸
  

In a cohort study of 60 neonates exposed to SSRIs in utero, 30% met diagnostic criteria for neonatal abstinence syndrome. The most common discontinuation symptoms were:

• tremor (37/60)
  
• GI disturbances (34/60)—including exaggerated sucking, poor feeding, regurgitation, vomiting, and loose stools
  
• sleep disturbance (21/60).
  

Other symptoms included irritability, constant crying, shivering, increased tone, convulsions, jitteriness, poor gaze control, vomiting, myoclonus, and lethargy.²⁵

Recommendations. The perception that SSRIs have low fetal toxicity has guided prescribing practices in recent years. Newer evidence shows, however, that fetal exposure to SSRIs may have some adverse effects, including lower birth weight and early delivery. First-trimester paroxetine use has been associated with increased risk for fetal ventricular and/or atrial septal defects.

Discuss these risks with the patient when you consider starting or continuing SSRI use during pregnancy.²⁴ If you prescribe an SSRI, use the minimum effective dosage and avoid paroxetine during pregnancy.¹⁸

To reduce the risk for PPHN, early delivery, and neonatal withdrawal syndromes, taper and discontinue the SSRI during the third trimester. Restarting the SSRI soon after delivery is the most effective way to prevent recurrence of anxiety symptoms or postpartum depression.

Benzodiazepines

Teratogenicity. Like SSRIs, benzodiazepines cross the placenta to the fetus.²⁹ Benzodiazepine teratogenicity remains controversial.⁸ Some—but not all—data show a small but significant increased risk for major malformations/oral cleft malformations with first-trimester benzodiazepine exposure.

A Medline literature search from 1966 to 2000 found not enough information to determine whether potential benefits of benzodiazepines to the mother outweigh risks to the fetus.²⁹ An ambitious meta-analysis of >1,400 studies by Dolovich et al³⁰ found a small association between fetal exposure to benzodiazepines and major malformations/cleft palate, but only in pooled data from case-controlled studies. No association was found between fetal exposure to benzodiazepines and malformations/cleft palate in pooled data from cohort studies.

A 32-month, hospital-based surveillance program of 28,565 births found no increase in the rate of major malformations in 43 infants exposed to clonazepam monotherapy—33 (77%) in the first trimester.³¹ Thus, the risk of major malformations/cleft palate with the use of benzodiazepines in the first trimester appears to be low.

Toxicity and withdrawal syndromes. Neonatal benzodiazepine toxicity and withdrawal syndromes have been reported in studies and case reports. Although these syndromes occur, they do not affect all infants with late third-trimester benzodiazepine exposure. Prevalence rates have not been calculated.³²

• Neonatal toxicity (“floppy infant syndrome”)—characterized by hypothermia, lethargy, poor respiratory effort, and feeding difficulties—occurs after maternal benzodiazepine use just before delivery.⁸
  
• Neonatal withdrawal may be caused by very late, third trimester exposure to benzodiazepines. Symptoms—which can persist ≤3 months after delivery—include restlessness, irritability, abnormal sleep patterns, suckling difficulties, growth retardation, hypertonia, hyperreflexia, tremulousness, apnea, diarrhea, and vomiting.^8,29
  

Recommendations. When possible, avoid benzodiazepines in the first trimester because of possible teratogenicity and then again late in the third trimester before delivery because of neonatal withdrawal syndromes. To reduce as much as possible the small risk of a benzodiazepine-related fetal malformation/cleft palate, wean the mother from benzodiazepines before conception. After the first trimester, the benzodiazepine can be restarted if necessary.²⁹

To minimize neonatal withdrawal, gradually taper the mother’s benzodiazepine before delivery.²⁹ Because the baby’s due date is calculated to be ±2 weeks before delivery, begin this taper 3 to 4 weeks before the due date and discontinue at least 1 week before delivery. Breastfeeding while taking benzodiazepines is not recommended because of the risk of over-sedating the infant.

 

A rational approach

Both benzodiazepines and SSRIs are associated with low but demonstrated risks to the fetus when used during pregnancy (Table 2).^{19,20,23,25,30,33} Use these medications to manage a patient’s anxiety only if the clinical benefit to the mother justifies the potential risks to the fetus.²⁹

A staggered combination of SSRIs during the first 2 trimesters and benzodiazepines during the last 2 trimesters can help balance the risks and benefits of pharmacotherapy of anxiety disorders during pregnancy (Table 3).

Frankly discuss with your patient the risks and benefits in the context of her perceived need for symptom control to sustain her level of functioning. You could document this discussion in the progress note as “R, B, A, and pt C,” signifying that risks, benefits, and alternatives were discussed, and the patient consented. If possible, include the patient’s husband, partner, or parent in this discussion.

Table 2

Risks of SSRIs vs benzodiazepines during pregnancy stages

Pregnancy stage when given	Fetal risk	SSRIs	Benzodiazepines
First trimester*	Teratogenicity	Paroxetine use associated with 2-fold increased risk of major congenital anomalies and 3-fold increased risk of major cardiac anomalies;19 meta-analysis calculated significant risk of cardiac malformations (odds ratio 1.72; population prevalence = 13.4/1,000 births)20,33	Meta-analysis of case control studies showed increased risk of major malformations/cleft palate (odds ratio 3.01; population prevalence = 10 to 20/1,000 births); no association seen in cohort studies30
Third trimester	PPHN	Case control study showed 3.7% of infants with PPHN were exposed to SSRIs vs 0.7% of controls; adjusted odds ratio 6.1, absolute risk to exposed population = 6 to 12/1,000 births)23
Perinatal and long-term effects	Toxicity/withdrawal syndromes	Cohort study of 60 infants concluded prevalence of discontinuation syndromes is 30% in neonates with third trimester SSRI exposure25	Neonatal toxicity (“floppy infant syndrome”) and neonatal withdrawal reported with maternal benzodiazepine use in late third trimester; prevalence unknown†
	Preterm birth, serotonin withdrawal syndromes, CNS effects, long-term neurobehavioral effects	Unknown†	Unknown†
PPHN: persistent pulmonary hypertension of the newborn; SSRIs: selective serotonin reuptake inhibitors
* Available data indicate that first-trimester exposure to SSRIs (other than paroxetine) and benzodiazepines may increase the relative risk for congenital anomalies, but the absolute risk of having a child with an anomaly is small.
† Some case reports, but published literature is insufficient to determine prevalence or magnitude of risk.

Table 3

Staggered, combination therapy for anxiety disorders during pregnancy

Pregnancy stage	Recommended to manage risks to mother and fetus
First trimester	SSRI (not paroxetine) No benzodiazepines Nondrug therapies*
Second trimester	SSRI (not paroxetine) Can use benzodiazepine if needed Nondrug therapies*
Third trimester	Taper off SSRI by 1 to 2 months before due date Can use benzodiazepine until 2 weeks before due date Nondrug therapies*
SSRI: selective serotonin reuptake inhibitor
* Nondrug therapies can include prenatal exercise, sleep hygiene, relaxation, and psychotherapy (cognitive-behavioral therapy, interpersonal therapy, supportive therapy, family/couples therapy)

CASE CONTINUED: CBT plus medication

Ms. K and her husband are open to adding weekly cognitive-behavioral therapy (CBT) for anxiety as long as she can continue her medications. You discuss the evidence regarding potential neonatal risks with paroxetine and clonazepam treatment. Because Ms. K is 6 weeks pregnant, you outline a plan for a rapid cross-taper off paroxetine and onto fluoxetine, 10 to 30 mg/d, explaining that paroxetine might pose a greater first-trimester risk of major congenital malformations and cardiac malformations. You discuss possible side effects of fluoxetine and explain a plan to taper off fluoxetine during the third trimester to reduce the risk of PPHN, early delivery, and withdrawal in the newborn.

Because Ms. K has been taking clonazepam at only 0.5 mg 1 to 2 times per week, you instruct her to stop taking the benzodiazepine for the next 6 weeks until she is through her first trimester. You also reassure her that she can use clonazepam after the first trimester, if necessary, as long as she agrees to taper off completely 1 to 2 weeks before to her due date.

You refer her to a CBT therapist and emphasize the importance of CBT, relaxation, and sleep hygiene—as well as support from her husband, family, and friends—to reduce her stress and facilitate the medication taper during her third trimester. You plan to see her monthly and co-manage her care with the CBT therapist and Ob/Gyn. You document this discussion in her medical record as evidence of informed consent.

Related resources

• Baby Center. Managing stress and anxiety during pregnancy. Patient information. www.babycenter.com/0_managing-stress-and-anxiety-during-pregnancy_1683.bc.
  
• Organization of Teratology Information Specialists (OTIS). www.otispregnancy.org.
  

Drug brand names

• Clonazepam • Klonopin
  
• Paroxetine • Paxil
  
• Fluoxetine • Prozac
  
• Sertraline • Zoloft
  

Disclosures

The authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Ms. K, age 25, is 6 weeks pregnant and is taking medications for generalized anxiety disorder (GAD). When she was diagnosed with GAD at age 19, her symptoms included 6 months of excessive anxiety—insomnia, fatigue, difficulty with concentration, and psychomotor agitation—without mood symptoms. These symptoms interfered greatly with her schoolwork and other daily activities.

For 6 years Ms. K has been taking the selective serotonin reuptake inhibitor (SSRI) paroxetine, 15 mg/d, and the benzodiazepine clonazepam, 0.5 mg as needed, with good symptom control. Now that she is pregnant and her primary care doctor has refused to continue these medications, she is seeking treatment and advice.

Not enough is known about how to safely treat anxiety disorders during pregnancy, and physicians are not sure what to do with patients such as Ms. K. Without evidence-based guidelines, we feel anxious about potential risks to mother and fetus as we try to provide appropriate drug therapy.

To help you and your patients weigh the risks and benefits of perinatal treatments for anxiety disorders, this article briefly summarizes the evidence on:

• anxiety disorders’ natural history during pregnancy
  
• how untreated maternal anxiety affects the fetus
  
• nonpharmacologic therapies for anxiety disorders
  
• a plan to manage fetal risks by staggering SSRI and benzodiazepine use during the first and third trimesters.
  

Anxiety during pregnancy

Nearly one-third of women experience an anxiety disorder during their lives, with peak onset during childbearing years.^1,2 Compared with research on perinatal depression, far fewer studies have examined anxiety disorders’ onset, presentation, prevalence, and treatment.¹

The literature includes no studies of the course of preexisting GAD or posttraumatic stress disorder (PTSD) and no evidence that symptoms of preexisting obsessive-compulsive disorder (OCD) change during pregnancy. Some studies of panic disorder show symptoms improving during pregnancy, whereas others do not (Table 1).¹

One small study done in late pregnancy found a significant association between the prevalence of an anxiety disorder, maternal primiparity, and comorbid medical conditions. Thus, a woman in her first pregnancy may be at increased risk to develop an anxiety disorder if she has a comorbid medical condition.³ As in the case of Ms. K, however, continuation of preexisting anxiety appears more likely than onset of a new anxiety disorder during pregnancy.

Table 1

How pregnancy affects the course of 4 anxiety disorders

Disorder	Prevalence	Effect
Generalized anxiety disorder (GAD)	8.5% of women experience GAD during the third trimester, compared with a 5% prevalence in the general population	No studies have reported on the course of GAD in pregnant women with preexisting disorder
Obsessive-compulsive disorder (OCD)	2% to 12% of OCD outpatients of childbearing age report onset during pregnancy	Preexisting OCD usually shows no change during pregnancy but may worsen postpartum
Panic disorder (PD)	1.3% to 2% in pregnant women, compared with 1.5% to 3.5% in the general population	Panic symptoms in women with preexisting PD may improve during pregnancy and worsen postpartum
Posttraumatic stress disorder (PTSD)	2.3% to 7.7% in pregnant women and 0% to 6.9% postpartum, compared with 1% to 14% in the community	No studies have reported on the course of PTSD in pregnant women with preexisting disorder
Source: References 1,2

Fetal risks from maternal anxiety

Fetal risk from severe maternal anxiety is not zero. Offspring born to high-anxiety mothers exhibit neurobehavioral differences compared with offspring of calmer mothers. Changes in high-anxiety mothers’ offspring include:

• altered EEG activation and vagal tone
  
• increased time in deep sleep and less time in active alert states
  
• lower performance on the Brazelton Neonatal Behavior Assessment Scale.⁴
  

A cohort study by Teixeira et al⁵ found an association between maternal anxiety in pregnancy and uterine artery resistance, suggesting a possible mechanism by which a mother’s psychologic state may affect fetal development. High anxiety and self-reported life stress during pregnancy also are associated consistently with abnormal, high-frequency heart rate variability in infants—a finding linked with negative infant behavior and later adult hostility.⁶

Exposure to maternal high anxiety has been associated with mental developmental delays in infants and increased risk for behavioral and emotional problems in young children.^7-10 Anxiety may not directly cause intrauterine growth retardation and preterm delivery, but it is significantly associated with prenatal tobacco, alcohol, and narcotics use—which predicts these and other negative neonatal outcomes.¹¹

Anxiety during pregnancy is a risk factor for postnatal depressive symptoms, independent of depressed mood and family or marital stressors during pregnancy.¹² Mothers with postpartum depression appear less able to respond sensitively and competently to their newborns, and these infants may be at increased risk of behavioral, emotional, and cognitive problems.

 

^7,13

CASE CONTINUED: ‘Stay the course’

Ms. K worries that she could not tolerate recurrence of her anxiety symptoms and wishes to continue both medications. Her husband concurs, but they want to minimize potential risks to their baby. You discuss the options for treating anxiety symptoms during pregnancy, including medications, psychotherapy, and behavioral treatments.

Treatment decisions

Ideally you’ll begin treating anxiety disorders in women of childbearing age with preconception psychoeducation. Explaining the risks of medications if she were to become pregnant and asking about the contraception she is using are de rigueur. Psychotherapy is low risk to the fetus and is considered first choice for treating mild to moderate anxiety in women of childbearing age who plan to become pregnant (Box).^1,14-17

Box

Psychotherapy alone is inadequate, however, for the many patients—such as Ms. K—who present already pregnant with a history of moderate to severe anxiety. Adjunctive psychotropic therapy—along with various nonmedication therapies—is warranted for patients whose social or occupational functioning would be substantially impaired by suboptimal control of anxiety during pregnancy.

Because Ms. K wishes to continue taking paroxetine and clonazepam, what can you tell her about the risks and benefits of SSRIs and benzodiazepines during pregnancy?

SSRIs in pregnancy

Teratogenicity. Compared with benzodiazepines, SSRIs have been considered agents of choice for use during pregnancy because of a lower risk of teratogenic effects.¹⁵ Paroxetine, however, appears to pose a greater risk for teratogenicity than other SSRIs.

An increased risk for fetal ventricular and/or atrial septal defects has been associated with first-trimester exposure to paroxetine, but no other SSRI.¹⁸ First trimester exposure to paroxetine at doses averaging 25 mg/d has been associated with statistically significant risks of major congenital anomalies (2-fold increase) and major cardiac anomalies (3-fold increase),¹⁹ although other studies have failed to reproduce this finding. A meta-analysis of 7 studies by Bar-Oz et al²⁰ found an association between first-trimester paroxetine exposure and a significant increase in risk for cardiac malformations (odds ratio [OR] 1.72; 95% CI,1.22-2.42).

The overall rate of fetal malformations from SSRIs appears to be low, although most studies have examined only fluoxetine or paroxetine. Some studies have reported various malformations with fluoxetine or sertraline, but others have not. In Finland, a population-based study found no increase in rate of major congenital malformations in offspring of 1,782 women who filled prescriptions for SSRIs during pregnancy, compared with the general population rate of 1% to 3%.²¹

Neurobehavioral effects. SSRI exposure during fetal life has shown no long-term neurobehavioral effects. A blinded prospective study by Nulman et al²² found no differences in global IQ scores, language development, or behavioral development among children age ≤5 who were exposed in utero to fluoxetine (n=40) or a tricyclic antidepressant (n=46), compared with unexposed children of nondepressed mothers (n=36). Similarly, using reports from teachers and clinical measures of internalizing behaviors, Misri et al¹⁰ found no increase in depression, anxiety, or withdrawal in 4-year-olds with prenatal exposure to SSRIs (n=22), compared with nonexposed children (n=14).

Pulmonary hypertension. SSRI exposure in later pregnancy may increase the rate of persistent pulmonary hypertension of the newborn (PPHN), which occurs in 1 to 2 infants per 1,000 live births. PPHN showed a statistically significant association with late prenatal SSRI exposure (OR 6.1) in a study that controlled for maternal smoking, body mass index, and diabetes.²³ PPHN occurred in approximately 1% of infants exposed to SSRIs in late pregnancy. PPHN rates were not affected by maternal depression/anxiety, non-SSRI antidepressant exposure throughout pregnancy, or SSRI exposure during early pregnancy only.

 

Toxicity and withdrawal syndromes. Infants of women who continue to take SSRIs just before delivery can develop toxicity or withdrawal syndromes. Occurrence of either syndrome depends on SSRI half-life, serum concentration, and the pharmacodynamics of other medications given during pregnancy and labor.²⁴

Discontinuation syndromes can occur in SSRI-exposed neonates within a few hours or days after birth and last up to 1 month after delivery, depending on the infant’s susceptibility.²⁵ Nearly two-thirds of suspected SSRI-induced neonatal withdrawal syndromes have been associated with paroxetine, although all SSRIs appear be associated with some risk.²⁶ Several trials, including a recent prospective study, found prenatal antidepressant use associated with lower gestational age at birth and increased risk of preterm birth.²⁷

A prospective study compared the effects of maternal SSRI use on behavioral state, sleep, motor activity, and heart rate variability in 17 exposed vs 17 nonexposed matched neonates. In the first 1 to 2 weeks of life, SSRI-exposed neonates showed:

• greater tremulousness
  
• less flexible and dampened state regulation
  
• more time in uninterrupted REM sleep
  
• more frequent startles or sudden arousals
  
• greater generalized motor activity
  
• greater autonomic dysregulation.²⁸
  

In a cohort study of 60 neonates exposed to SSRIs in utero, 30% met diagnostic criteria for neonatal abstinence syndrome. The most common discontinuation symptoms were:

• tremor (37/60)
  
• GI disturbances (34/60)—including exaggerated sucking, poor feeding, regurgitation, vomiting, and loose stools
  
• sleep disturbance (21/60).
  

Other symptoms included irritability, constant crying, shivering, increased tone, convulsions, jitteriness, poor gaze control, vomiting, myoclonus, and lethargy.²⁵

Recommendations. The perception that SSRIs have low fetal toxicity has guided prescribing practices in recent years. Newer evidence shows, however, that fetal exposure to SSRIs may have some adverse effects, including lower birth weight and early delivery. First-trimester paroxetine use has been associated with increased risk for fetal ventricular and/or atrial septal defects.

Discuss these risks with the patient when you consider starting or continuing SSRI use during pregnancy.²⁴ If you prescribe an SSRI, use the minimum effective dosage and avoid paroxetine during pregnancy.¹⁸

To reduce the risk for PPHN, early delivery, and neonatal withdrawal syndromes, taper and discontinue the SSRI during the third trimester. Restarting the SSRI soon after delivery is the most effective way to prevent recurrence of anxiety symptoms or postpartum depression.

Benzodiazepines

Teratogenicity. Like SSRIs, benzodiazepines cross the placenta to the fetus.²⁹ Benzodiazepine teratogenicity remains controversial.⁸ Some—but not all—data show a small but significant increased risk for major malformations/oral cleft malformations with first-trimester benzodiazepine exposure.

A Medline literature search from 1966 to 2000 found not enough information to determine whether potential benefits of benzodiazepines to the mother outweigh risks to the fetus.²⁹ An ambitious meta-analysis of >1,400 studies by Dolovich et al³⁰ found a small association between fetal exposure to benzodiazepines and major malformations/cleft palate, but only in pooled data from case-controlled studies. No association was found between fetal exposure to benzodiazepines and malformations/cleft palate in pooled data from cohort studies.

A 32-month, hospital-based surveillance program of 28,565 births found no increase in the rate of major malformations in 43 infants exposed to clonazepam monotherapy—33 (77%) in the first trimester.³¹ Thus, the risk of major malformations/cleft palate with the use of benzodiazepines in the first trimester appears to be low.

Toxicity and withdrawal syndromes. Neonatal benzodiazepine toxicity and withdrawal syndromes have been reported in studies and case reports. Although these syndromes occur, they do not affect all infants with late third-trimester benzodiazepine exposure. Prevalence rates have not been calculated.³²

• Neonatal toxicity (“floppy infant syndrome”)—characterized by hypothermia, lethargy, poor respiratory effort, and feeding difficulties—occurs after maternal benzodiazepine use just before delivery.⁸
  
• Neonatal withdrawal may be caused by very late, third trimester exposure to benzodiazepines. Symptoms—which can persist ≤3 months after delivery—include restlessness, irritability, abnormal sleep patterns, suckling difficulties, growth retardation, hypertonia, hyperreflexia, tremulousness, apnea, diarrhea, and vomiting.^8,29
  

Recommendations. When possible, avoid benzodiazepines in the first trimester because of possible teratogenicity and then again late in the third trimester before delivery because of neonatal withdrawal syndromes. To reduce as much as possible the small risk of a benzodiazepine-related fetal malformation/cleft palate, wean the mother from benzodiazepines before conception. After the first trimester, the benzodiazepine can be restarted if necessary.²⁹

To minimize neonatal withdrawal, gradually taper the mother’s benzodiazepine before delivery.²⁹ Because the baby’s due date is calculated to be ±2 weeks before delivery, begin this taper 3 to 4 weeks before the due date and discontinue at least 1 week before delivery. Breastfeeding while taking benzodiazepines is not recommended because of the risk of over-sedating the infant.

 

A rational approach

Both benzodiazepines and SSRIs are associated with low but demonstrated risks to the fetus when used during pregnancy (Table 2).^{19,20,23,25,30,33} Use these medications to manage a patient’s anxiety only if the clinical benefit to the mother justifies the potential risks to the fetus.²⁹

A staggered combination of SSRIs during the first 2 trimesters and benzodiazepines during the last 2 trimesters can help balance the risks and benefits of pharmacotherapy of anxiety disorders during pregnancy (Table 3).

Frankly discuss with your patient the risks and benefits in the context of her perceived need for symptom control to sustain her level of functioning. You could document this discussion in the progress note as “R, B, A, and pt C,” signifying that risks, benefits, and alternatives were discussed, and the patient consented. If possible, include the patient’s husband, partner, or parent in this discussion.

Table 2

Risks of SSRIs vs benzodiazepines during pregnancy stages

Pregnancy stage when given	Fetal risk	SSRIs	Benzodiazepines
First trimester*	Teratogenicity	Paroxetine use associated with 2-fold increased risk of major congenital anomalies and 3-fold increased risk of major cardiac anomalies;19 meta-analysis calculated significant risk of cardiac malformations (odds ratio 1.72; population prevalence = 13.4/1,000 births)20,33	Meta-analysis of case control studies showed increased risk of major malformations/cleft palate (odds ratio 3.01; population prevalence = 10 to 20/1,000 births); no association seen in cohort studies30
Third trimester	PPHN	Case control study showed 3.7% of infants with PPHN were exposed to SSRIs vs 0.7% of controls; adjusted odds ratio 6.1, absolute risk to exposed population = 6 to 12/1,000 births)23
Perinatal and long-term effects	Toxicity/withdrawal syndromes	Cohort study of 60 infants concluded prevalence of discontinuation syndromes is 30% in neonates with third trimester SSRI exposure25	Neonatal toxicity (“floppy infant syndrome”) and neonatal withdrawal reported with maternal benzodiazepine use in late third trimester; prevalence unknown†
	Preterm birth, serotonin withdrawal syndromes, CNS effects, long-term neurobehavioral effects	Unknown†	Unknown†
PPHN: persistent pulmonary hypertension of the newborn; SSRIs: selective serotonin reuptake inhibitors
* Available data indicate that first-trimester exposure to SSRIs (other than paroxetine) and benzodiazepines may increase the relative risk for congenital anomalies, but the absolute risk of having a child with an anomaly is small.
† Some case reports, but published literature is insufficient to determine prevalence or magnitude of risk.

Table 3

Staggered, combination therapy for anxiety disorders during pregnancy

Pregnancy stage	Recommended to manage risks to mother and fetus
First trimester	SSRI (not paroxetine) No benzodiazepines Nondrug therapies*
Second trimester	SSRI (not paroxetine) Can use benzodiazepine if needed Nondrug therapies*
Third trimester	Taper off SSRI by 1 to 2 months before due date Can use benzodiazepine until 2 weeks before due date Nondrug therapies*
SSRI: selective serotonin reuptake inhibitor
* Nondrug therapies can include prenatal exercise, sleep hygiene, relaxation, and psychotherapy (cognitive-behavioral therapy, interpersonal therapy, supportive therapy, family/couples therapy)

CASE CONTINUED: CBT plus medication

Ms. K and her husband are open to adding weekly cognitive-behavioral therapy (CBT) for anxiety as long as she can continue her medications. You discuss the evidence regarding potential neonatal risks with paroxetine and clonazepam treatment. Because Ms. K is 6 weeks pregnant, you outline a plan for a rapid cross-taper off paroxetine and onto fluoxetine, 10 to 30 mg/d, explaining that paroxetine might pose a greater first-trimester risk of major congenital malformations and cardiac malformations. You discuss possible side effects of fluoxetine and explain a plan to taper off fluoxetine during the third trimester to reduce the risk of PPHN, early delivery, and withdrawal in the newborn.

Because Ms. K has been taking clonazepam at only 0.5 mg 1 to 2 times per week, you instruct her to stop taking the benzodiazepine for the next 6 weeks until she is through her first trimester. You also reassure her that she can use clonazepam after the first trimester, if necessary, as long as she agrees to taper off completely 1 to 2 weeks before to her due date.

You refer her to a CBT therapist and emphasize the importance of CBT, relaxation, and sleep hygiene—as well as support from her husband, family, and friends—to reduce her stress and facilitate the medication taper during her third trimester. You plan to see her monthly and co-manage her care with the CBT therapist and Ob/Gyn. You document this discussion in her medical record as evidence of informed consent.

Related resources

• Baby Center. Managing stress and anxiety during pregnancy. Patient information. www.babycenter.com/0_managing-stress-and-anxiety-during-pregnancy_1683.bc.
  
• Organization of Teratology Information Specialists (OTIS). www.otispregnancy.org.
  

Drug brand names

• Clonazepam • Klonopin
  
• Paroxetine • Paxil
  
• Fluoxetine • Prozac
  
• Sertraline • Zoloft
  

Disclosures

The authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Mr. G, age 38, is an investment banker referred for evaluation of an alcohol use disorder. Three years ago his internist diagnosed Mr. G with major depression and prescribed a selective serotonin reuptake inhibitor. Mr. G’s mood has improved, but his drinking is out of control and is affecting his work and marriage.

Mr. G describes his father as an alcoholic and says he has noticed worrisome similarities in himself. Since his teenage years, he recalls always being able to drink more than his peers. Amnesia episodes began in college during heavy drinking days and now occur almost weekly. Most recently his driver’s license was suspended after he was arrested for driving while impaired by alcohol.

He has attempted to stop drinking 3 times in the last 6 months and feels frustrated because he continues to relapse. During his last quit attempt, he remained abstinent for 3 months and believes his mood was unchanged during that time.

For motivated patients such as Mr. G, National Institute on Alcohol Abuse and Alcoholism (NIAAA) guidelines (updated in 2007) consider medications first-line treatment for alcohol dependence, along with psychotherapies and mutual-help groups such as Alcoholics Anonymous.¹ Medications with evidence of efficacy include FDA-approved disulfiram, naltrexone, and acamprosate, and off-label topiramate.

Each drug’s pharmacology is different; some may be beneficial during early abstinence, whereas others are more effective for maintaining abstinence. Because many physicians have had little or no experience using these medications,^2,3 we discuss dosing recommendations and side effect profiles—important clinical differences to guide drug selection and administration.

Box

CASE CONTINUED: Which came first?

When Mr. G presented with depressive symptoms, his internist informed him that alcohol’s effects can mimic depression and advised him to cut back his consumption. Mr. G temporarily reduced his drinking but continued to experience depressed mood, sleep disturbances, difficulty concentrating, fatigue, and poor appetite. The internist then prescribed escitalopram, 10 mg/d, and Mr. G says his depressive symptoms improved. He has not attempted suicide or required psychiatric hospitalization.

After a thorough evaluation—including a detailed assessment of his drinking history, other substance use, and mood symptoms—you diagnose Mr. G with alcohol dependence without physiological dependence and a primary major depressive disorder (MDD).

Mr. G’s diagnosis of alcohol dependence is based on evidence in the last 12 months of tolerance, repeated loss of control over the amount he drinks, multiple failed attempts to stop drinking, repeated negative consequences to his work productivity and personal relationships, and continued drinking despite knowing that alcohol consumption sometimes intensifies his mood symptoms (Box). Physiological dependence is unlikely because he did not experience withdrawal during a recent period of abstinence.

The mood component. Particularly in psychiatric patients, alcohol dependence often coexists with and affects the treatment of other psychopathologies:

• 1 in 3 adults experience an alcohol use disorder during their lifetimes.⁴
  
• 1 in 3 adults with an alcohol use disorder have a comorbid psychiatric disorder.⁵
  
• Alcohol dependence doubles the risk of major depression and triples the lifetime risk of any mood disorder.⁴
  

A thorough evaluation—both medical and psychiatric—is necessary to distinguish a primary mood disorder from a substance-induced mood disorder. Aspects of a patient’s history that may support a substance-induced disorder diagnosis include:

• Mood symptoms appear after the onset of a substance use disorder.
  
• Mood symptoms are absent during abstinence.
  
• Mood symptoms are consistent with the effects of the drug being used.⁶
  

Substance-induced mood symptoms usually become less intense and eventually resolve as abstinence is maintained. Primary mood symptoms may not be affected by abstinence, and interventions—including pharmacotherapy—are more likely to be required.

In evaluating Mr. G’s drinking history, you determined that his mood symptoms began before he started consuming alcohol regularly and have persisted during periods of abstinence. Thus, primary MDD is a more likely diagnosis than substance-induced MDD.

 

CASE CONTINUED: Planning treatment

A combination of behavioral therapy and pharmacotherapy is appropriate for treating Mr. G’s alcohol dependence. When you discuss the diagnosis with him, he endorses a goal of abstinence. For behavioral therapy, he says he would like to try Alcoholics Anonymous, which helped a friend “turn his life around.”

He has become accustomed to taking escitalopram once daily but is hesitant to take any medication that requires more frequent dosing. He also worries that medication might impair his work performance, which requires extensive periods of concentration. Your goal—as you consider available medications—is to develop a treatment plan that incorporates Mr. G’s preferences and addresses his concerns.

Disulfiram

Disulfiram, an irreversible aldehyde dehydrogenase inhibitor, is indicated for maintaining enforced sobriety in patients with chronic alcohol dependence (Table 1). Aldehyde dehydrogenase inhibition disrupts alcohol-to-acetate metabolism, which leads to acetaldehyde accumulation. If a disulfiram-treated patient ingests alcohol, increased acetaldehyde levels lead to the unpleasant “disulfiram-ethanol reaction,” with diaphoresis, flushing, nausea, vomiting, headache, tachycardia, and hypotension. The reaction’s severity is proportional to the disulfiram dose and amount of alcohol consumed.

Patients taking disulfiram must abstain from all alcohol, including over-the-counter cold remedies and mouthwashes containing alcohol. Advise patients that ingesting small amounts of alcohol can induce symptoms, even days after taking disulfiram.

Efficacy. Clinical trial results with disulfiram have been mixed. In the largest controlled study to date, Fuller et al⁷ found no significant difference in rates of total abstinence, time to first drink, employment, or social stability measures at 1 year among 605 men who received counseling plus disulfiram, 250 mg/d, or placebo. Other disulfiram studies have found a modest decrease in the frequency of drinking but no effect on abstinence rates.^8,9

Because medication adherence is the strongest predictor of outcome with disulfiram,¹⁰ monitoring for adherence and stressing its importance to patients may increase the drug’s efficacy. Disulfiram may be most effective in highly motivated patients with stable social support or as an adjuvant to an outpatient treatment program.

Administration. Disulfiram is available in 250-mg tablets and is usually dosed from 125 to 500 mg/d. Treatment can begin after patients abstain from alcohol for ≥12 hours and have a serum alcohol concentration of zero.

Side effects. Drowsiness is a common complaint with disulfiram; this adverse effect is frequently self-limited and can be reduced by evening dosing.

Subclinical liver enzyme elevations have been reported in 25% of patients taking disulfiram.¹¹ Although rare, potentially fatal hepatotoxicity has been reported^12,13 (with a dose as low as 200 mg/d¹²), typically occurring early in treatment and associated with jaundice and fever. One study estimated the risk of dying of hepatotoxicity caused by disulfiram to be 1 in 30,000 patients/year.¹⁴

A recent Swedish study¹³ reviewed data from 1966 through 2002 and found 82 cases of drug-induced liver injury associated with disulfiram. By comparing these findings with sales figures from 1972 to 2002, the authors report an incidence of disulfiram-induced liver injury of about 1 case per 1.3 million estimated average daily doses.

Order liver function tests at baseline, then retest 10 to 14 days after starting disulfiram and again approximately 4 weeks later. Thereafter, monitoring once every 3 to 6 months is generally sufficient in patients without liver disease symptoms.

Other serious adverse events associated with disulfiram therapy include optic neuritis, peripheral neuritis, cholestatic hepatitis, seizures, and arrhythmias. Psychosis also can occur, generally with dosages ≥500 mg/d. Avoid concomitant use of disulfiram and metronidazole, which can cause acute psychosis.

Disulfiram-related inhibition of cytochrome P-450 can increase serum levels and toxicity risk of medications metabolized in the liver, such as warfarin, phenytoin, and isoniazid. Patients taking concomitant warfarin and disulfiram require close monitoring for increases in the international normalized ratio (INR).

Contraindications. Disulfiram is contraindicated in patients with ischemic heart disease and those who are pregnant. Also avoid disulfiram in patients with cerebrovascular disease, diabetes mellitus, psychosis, or cognitive impairment.

Recommendation. Disulfiram is a valid option for treating alcohol dependence in a select group of highly motivated patients who are medically and psychiatrically stable and in whom adherence can be closely monitored.

Table 1

Disulfiram: Fast facts

Mechanism: Acetaldehyde accumulates when aldehyde dehydrogenase is inhibited
FDA-approved for alcohol dependence: Yes
Dosing: 125 to 500 mg once daily
Effect: Aversive reaction to alcohol
Potential side effects: Liver toxicity, seizures, arrhythmia, peripheral neuropathy, psychosis
Contraindications: Concurrent alcohol consumption, severe cardiac disease, psychosis, pregnancy
Comments: Many drug interactions, including warfarin, metronidazole, and phenytoin; monitor liver function for toxicity

 

Naltrexone

Naltrexone is a μ-opioid receptor antagonist thought to reduce alcohol’s reinforcing effects by interfering with β-endorphin pathways. It is indicated for treating alcohol dependence and has been shown to reduce relapse and number of drinking days in alcohol-dependent patients (Table 2).¹⁵

Naltrexone also has reduced alcohol consumption in healthy volunteers, social drinkers, and other nondependent drinkers.¹⁶ Its effect may have a genetic component, as suggested by greater benefit in persons with a family history of alcoholism.¹⁷

Efficacy. Most studies investigated naltrexone as part of a comprehensive treatment program that included behavioral therapies.¹⁸ Recently, the randomized, placebo-controlled Combining Medications and Behavioral Interventions (COMBINE) study found that using either naltrexone or behavioral therapy improved abstinence, and combing naltrexone with behavioral therapy was not more effective than either treatment alone.¹⁹

Administration. Oral naltrexone is usually started at 25 mg and increased over 2 to 3 days to 50 or 100 mg/d. The standard dose is 50 mg/d, although the COMBINE study reported efficacy at 100 mg/d.¹⁹

Oral naltrexone is most helpful for patients who adhere to 70% to 90% of the medication.²⁰ The extended-release form (a 380-mg IM dose given every 4 weeks) provides an option to monitor adherence.²¹

Side effects. Naltrexone toxicity can cause hepatocellular injury. Do not administer this drug to patients with acute hepatitis or end-stage liver disease. When prescribing naltrexone, check patients’ liver function monthly for the first 3 months, then once every 3 months thereafter.²² Less serious, common side effects include nausea, myalgia, and headache.

Naltrexone antagonizes opioid receptors and causes withdrawal symptoms in patients who are physically dependent on opioids. Therefore, do not give naltrexone to patients who require opioids for chronic pain. If your patient is using an opioid but could switch to other pain medication, discontinue the opioid for at least 7 days and consider a urine toxicology or naltrexone challenge before starting naltrexone.

Urine drug tests are inexpensive and easy to use but have limitations. Many standard “dipsticks” will detect heroin, morphine, and codeine but not oxycodone, hydrocodone, or other synthetic opioids. Specific tests are available to detect oxycodone, hydrocodone, hydromorphone, buprenorphine, and methadone. Some synthetic opioids (such as fentanyl) remain difficult to detect, however, because of their low concentration and rapid metabolism.

Table 2

Naltrexone: Fast facts

Mechanism: Opioid receptor antagonism interferes with β-endorphin pathways
FDA-approved for alcohol dependence: Yes
Dosing: Oral 50 mg once daily (recent evidence suggests safety and efficacy at 100 mg once daily); IM 380 mg once every 4 weeks
Effect: Decreases frequency and severity of relapse
Potential side effects: Nausea, myalgia, headache, dizziness
Contraindications: Opioid use, acute hepatitis, liver failure
Comments: Monitor liver function; once-monthly dosing may improve adherence

Acamprosate

Acamprosate is structurally similar to gamma-aminobutyric acid (GABA) and is thought to inhibit the glutamatergic system. This attenuation by acamprosate reduces the glutamatergic hyperactivity normally seen after chronic alcohol exposure.

Acamprosate is indicated for relapse prevention in patients with alcohol dependence who have stopped drinking (Table 3). Multiple randomized studies have demonstrated its efficacy in improving abstinence rates as compared with placebo.^23,24 Other studies, however, failed to show improved abstinence with acamprosate.²⁵ Some of the negative studies included patients who recently relapsed or had only a few days of abstinence before starting acamprosate.^19,26 Therefore, acamprosate might be most effective when used to maintain abstinence and less effective—if at all—to initiate abstinence.

Administration. Acamprosate is available in 333-mg tablets, with a recommended dosage of 666 mg tid. Side effects tend to be transient and mild. Reduce the dose to 333 mg tid for patients with moderate renal insufficiency (creatinine clearance [CrCl] 30 to 50 mL/min), and do not use acamprosate in patients with severe renal insufficiency (CrCl

Side effects. Acamprosate was well-tolerated in clinical trials; diarrhea and other GI side effects were the most commonly reported adverse events.

In some placebo-controlled studies, patients taking acamprosate reported more frequent suicidal thoughts and attempts compared with patients taking placebo.²⁷ These events were extremely rare, and no direct relationship with acamprosate therapy has been established. Nonetheless, monitor patients for depression and suicidal thoughts during acamprosate therapy.

Table 3

Acamprosate: Fast facts

Mechanism: Structurally similar to GABA; thought to inhibit the glutamatergic system
FDA-approved for alcohol dependence: Yes
Dosing: 666 mg tid
Effect: Increases abstinence
Potential side effects: Nausea, diarrhea, suicidal thoughts
Contraindications: Severe renal disease
Comments: Monitor patients for suicidal thoughts and depression

Topiramate

Topiramate potentiates GABA and inhibits excitatory glutamate transmission—properties believed to lead to decreased dopamine release at the nucleus accumbens in response to alcohol consumption. Although topiramate is not FDA-approved for alcohol dependence, limited data comparing this anticonvulsant with placebo have shown a reduction in drinking and increased abstinence (Table 4).^28,29

 

Administration. Start with 25 mg/d and increase over several weeks to 300 mg/d, given in divided doses.

Side effects include dizziness, paresthesia, somnolence, difficulty concentrating, and weight loss. Because topiramate is excreted renally, reduce doses by 50% in patients with CrCl

Other renal side effects include an elevated risk of nephrolithiasis. Topiramate’s inhibition of carbonic anhydrase can reduce bicarbonate levels, leading to a nonanion gap metabolic acidosis.

Table 4

Topiramate: Fast facts

Mechanism: Potentiates GABA and inhibits glutamate receptor subtypes
FDA-approved for alcohol dependence: No
Dosing: 300 mg/d in divided doses
Effect: Decreases craving and drinking
Potential side effects: Metabolic acidosis, psychomotor slowing, dizziness, difficulty concentrating, paresthesia, weight loss, nephrolithiasis, hyperammonemia with concomitant use of valproic acid
Contraindications: None known other than hypersensitivity (as with all drugs)
Comments: Dose titration requires several weeks; avoid abrupt withdrawal; may reduce effectiveness of oral contraceptives

CASE CONTINUED: Implementing a treatment plan

You start Mr. G on oral naltrexone, 25 mg/d, and titrate to 100 mg/d. Although no optimum treatment duration has been established, you plan to follow NIAAA recommendations that Mr. G use naltrexone at least 3 months, with the possibility of continuing 1 year or longer if he responds well.¹

You schedule weekly visits for the first month to monitor for side effects and to make any necessary modifications in behavioral and pharmacologic treatment. You also continue escitalopram, 10 mg, which has successfully controlled Mr. G’s MDD symptoms.

Medications for alcohol dependence generally have been studied as adjuncts to behavioral therapies. The COMBINE study of 1,383 alcohol-abstinent patients found naltrexone with medical management or cognitive-behavioral therapy alone to be equivalent in efficacy.¹⁹ The medical management provided a supportive environment, encouraged medication compliance, provided empathy to build a therapeutic relationship, and promoted self-help groups as an adjunct to treatment.³⁰ Thus, medication has a role in treating alcohol dependence, but behavioral therapy remains an important part of comprehensive substance abuse treatment.

When choosing medications, consider the agents’ clinically relevant differences:

• Naltrexone and—less conclusively—topiramate have shown benefit for alcoholdependent patients starting treatment and for relapse prevention.
  
• Acamprosate may help prevent relapse in abstinent patients.
  
• Disulfiram remains a valid option in highly motivated patients with social support available to ensure medication adherence.
  

Because Mr. G is starting therapy after recent alcohol use, medications such as naltrexone and topiramate that have shown benefit early in treatment (in addition to relapse prevention) are preferred. Of these 2 drugs, naltrexone is the better choice for Mr. G—who is concerned about his work performance—because difficulty concentrating is a common side effect of topiramate. Oral naltrexone would be preferred as initial therapy for Mr. G because he has expressed comfort taking once-daily oral medication. The more expensive oncemonthly naltrexone depot formulation could be a second-line treatment if adherence becomes an issue.

Hypersensitivity is considered a contraindication for any medication. Mr. G tolerates well an initial dose of 25 mg/d, followed by increases to 50 mg and then 100 mg over several days. You titrate oral naltrexone to 100 mg/d—even though it is commonly prescribed at 50 mg/d—because recent evidence suggests efficacy and safety at the higher dosage.¹⁹

Related resources

• National Institutes of Health. Helping patients who drink too much: a clinician’s guide. Bethesda, MD: National Institute on Alcohol Abuse and Alcoholism; 2007. NIH Publication 07-3769. www.niaaa.nih.gov/Publications/EducationTrainingMaterials/guide.htm.
  
• American Society of Addiction Medicine. www.asam.org.
  
• American Academy of Addiction Psychiatry. www.aaap.org.
  

Drug brand names

• Acamprosate • Campral
  
• Disulfiram • Antabuse
  
• Escitalopram • Lexapro
  
• Naltrexone, oral • ReVia
  
• Naltrexone, extended-release • Vivitrol
  
• Topiramate • Topamax
  
• Valproic acid • Depakote
  

Disclosure

The authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Mr. G, age 38, is an investment banker referred for evaluation of an alcohol use disorder. Three years ago his internist diagnosed Mr. G with major depression and prescribed a selective serotonin reuptake inhibitor. Mr. G’s mood has improved, but his drinking is out of control and is affecting his work and marriage.

Mr. G describes his father as an alcoholic and says he has noticed worrisome similarities in himself. Since his teenage years, he recalls always being able to drink more than his peers. Amnesia episodes began in college during heavy drinking days and now occur almost weekly. Most recently his driver’s license was suspended after he was arrested for driving while impaired by alcohol.

He has attempted to stop drinking 3 times in the last 6 months and feels frustrated because he continues to relapse. During his last quit attempt, he remained abstinent for 3 months and believes his mood was unchanged during that time.

For motivated patients such as Mr. G, National Institute on Alcohol Abuse and Alcoholism (NIAAA) guidelines (updated in 2007) consider medications first-line treatment for alcohol dependence, along with psychotherapies and mutual-help groups such as Alcoholics Anonymous.¹ Medications with evidence of efficacy include FDA-approved disulfiram, naltrexone, and acamprosate, and off-label topiramate.

Each drug’s pharmacology is different; some may be beneficial during early abstinence, whereas others are more effective for maintaining abstinence. Because many physicians have had little or no experience using these medications,^2,3 we discuss dosing recommendations and side effect profiles—important clinical differences to guide drug selection and administration.

Box

CASE CONTINUED: Which came first?

When Mr. G presented with depressive symptoms, his internist informed him that alcohol’s effects can mimic depression and advised him to cut back his consumption. Mr. G temporarily reduced his drinking but continued to experience depressed mood, sleep disturbances, difficulty concentrating, fatigue, and poor appetite. The internist then prescribed escitalopram, 10 mg/d, and Mr. G says his depressive symptoms improved. He has not attempted suicide or required psychiatric hospitalization.

After a thorough evaluation—including a detailed assessment of his drinking history, other substance use, and mood symptoms—you diagnose Mr. G with alcohol dependence without physiological dependence and a primary major depressive disorder (MDD).

Mr. G’s diagnosis of alcohol dependence is based on evidence in the last 12 months of tolerance, repeated loss of control over the amount he drinks, multiple failed attempts to stop drinking, repeated negative consequences to his work productivity and personal relationships, and continued drinking despite knowing that alcohol consumption sometimes intensifies his mood symptoms (Box). Physiological dependence is unlikely because he did not experience withdrawal during a recent period of abstinence.

The mood component. Particularly in psychiatric patients, alcohol dependence often coexists with and affects the treatment of other psychopathologies:

• 1 in 3 adults experience an alcohol use disorder during their lifetimes.⁴
  
• 1 in 3 adults with an alcohol use disorder have a comorbid psychiatric disorder.⁵
  
• Alcohol dependence doubles the risk of major depression and triples the lifetime risk of any mood disorder.⁴
  

A thorough evaluation—both medical and psychiatric—is necessary to distinguish a primary mood disorder from a substance-induced mood disorder. Aspects of a patient’s history that may support a substance-induced disorder diagnosis include:

• Mood symptoms appear after the onset of a substance use disorder.
  
• Mood symptoms are absent during abstinence.
  
• Mood symptoms are consistent with the effects of the drug being used.⁶
  

Substance-induced mood symptoms usually become less intense and eventually resolve as abstinence is maintained. Primary mood symptoms may not be affected by abstinence, and interventions—including pharmacotherapy—are more likely to be required.

In evaluating Mr. G’s drinking history, you determined that his mood symptoms began before he started consuming alcohol regularly and have persisted during periods of abstinence. Thus, primary MDD is a more likely diagnosis than substance-induced MDD.

 

CASE CONTINUED: Planning treatment

A combination of behavioral therapy and pharmacotherapy is appropriate for treating Mr. G’s alcohol dependence. When you discuss the diagnosis with him, he endorses a goal of abstinence. For behavioral therapy, he says he would like to try Alcoholics Anonymous, which helped a friend “turn his life around.”

He has become accustomed to taking escitalopram once daily but is hesitant to take any medication that requires more frequent dosing. He also worries that medication might impair his work performance, which requires extensive periods of concentration. Your goal—as you consider available medications—is to develop a treatment plan that incorporates Mr. G’s preferences and addresses his concerns.

Disulfiram

Disulfiram, an irreversible aldehyde dehydrogenase inhibitor, is indicated for maintaining enforced sobriety in patients with chronic alcohol dependence (Table 1). Aldehyde dehydrogenase inhibition disrupts alcohol-to-acetate metabolism, which leads to acetaldehyde accumulation. If a disulfiram-treated patient ingests alcohol, increased acetaldehyde levels lead to the unpleasant “disulfiram-ethanol reaction,” with diaphoresis, flushing, nausea, vomiting, headache, tachycardia, and hypotension. The reaction’s severity is proportional to the disulfiram dose and amount of alcohol consumed.

Patients taking disulfiram must abstain from all alcohol, including over-the-counter cold remedies and mouthwashes containing alcohol. Advise patients that ingesting small amounts of alcohol can induce symptoms, even days after taking disulfiram.

Efficacy. Clinical trial results with disulfiram have been mixed. In the largest controlled study to date, Fuller et al⁷ found no significant difference in rates of total abstinence, time to first drink, employment, or social stability measures at 1 year among 605 men who received counseling plus disulfiram, 250 mg/d, or placebo. Other disulfiram studies have found a modest decrease in the frequency of drinking but no effect on abstinence rates.^8,9

Because medication adherence is the strongest predictor of outcome with disulfiram,¹⁰ monitoring for adherence and stressing its importance to patients may increase the drug’s efficacy. Disulfiram may be most effective in highly motivated patients with stable social support or as an adjuvant to an outpatient treatment program.

Administration. Disulfiram is available in 250-mg tablets and is usually dosed from 125 to 500 mg/d. Treatment can begin after patients abstain from alcohol for ≥12 hours and have a serum alcohol concentration of zero.

Side effects. Drowsiness is a common complaint with disulfiram; this adverse effect is frequently self-limited and can be reduced by evening dosing.

Subclinical liver enzyme elevations have been reported in 25% of patients taking disulfiram.¹¹ Although rare, potentially fatal hepatotoxicity has been reported^12,13 (with a dose as low as 200 mg/d¹²), typically occurring early in treatment and associated with jaundice and fever. One study estimated the risk of dying of hepatotoxicity caused by disulfiram to be 1 in 30,000 patients/year.¹⁴

A recent Swedish study¹³ reviewed data from 1966 through 2002 and found 82 cases of drug-induced liver injury associated with disulfiram. By comparing these findings with sales figures from 1972 to 2002, the authors report an incidence of disulfiram-induced liver injury of about 1 case per 1.3 million estimated average daily doses.

Order liver function tests at baseline, then retest 10 to 14 days after starting disulfiram and again approximately 4 weeks later. Thereafter, monitoring once every 3 to 6 months is generally sufficient in patients without liver disease symptoms.

Other serious adverse events associated with disulfiram therapy include optic neuritis, peripheral neuritis, cholestatic hepatitis, seizures, and arrhythmias. Psychosis also can occur, generally with dosages ≥500 mg/d. Avoid concomitant use of disulfiram and metronidazole, which can cause acute psychosis.

Disulfiram-related inhibition of cytochrome P-450 can increase serum levels and toxicity risk of medications metabolized in the liver, such as warfarin, phenytoin, and isoniazid. Patients taking concomitant warfarin and disulfiram require close monitoring for increases in the international normalized ratio (INR).

Contraindications. Disulfiram is contraindicated in patients with ischemic heart disease and those who are pregnant. Also avoid disulfiram in patients with cerebrovascular disease, diabetes mellitus, psychosis, or cognitive impairment.

Recommendation. Disulfiram is a valid option for treating alcohol dependence in a select group of highly motivated patients who are medically and psychiatrically stable and in whom adherence can be closely monitored.

Table 1

Disulfiram: Fast facts

Mechanism: Acetaldehyde accumulates when aldehyde dehydrogenase is inhibited
FDA-approved for alcohol dependence: Yes
Dosing: 125 to 500 mg once daily
Effect: Aversive reaction to alcohol
Potential side effects: Liver toxicity, seizures, arrhythmia, peripheral neuropathy, psychosis
Contraindications: Concurrent alcohol consumption, severe cardiac disease, psychosis, pregnancy
Comments: Many drug interactions, including warfarin, metronidazole, and phenytoin; monitor liver function for toxicity

 

Naltrexone

Naltrexone is a μ-opioid receptor antagonist thought to reduce alcohol’s reinforcing effects by interfering with β-endorphin pathways. It is indicated for treating alcohol dependence and has been shown to reduce relapse and number of drinking days in alcohol-dependent patients (Table 2).¹⁵

Naltrexone also has reduced alcohol consumption in healthy volunteers, social drinkers, and other nondependent drinkers.¹⁶ Its effect may have a genetic component, as suggested by greater benefit in persons with a family history of alcoholism.¹⁷

Efficacy. Most studies investigated naltrexone as part of a comprehensive treatment program that included behavioral therapies.¹⁸ Recently, the randomized, placebo-controlled Combining Medications and Behavioral Interventions (COMBINE) study found that using either naltrexone or behavioral therapy improved abstinence, and combing naltrexone with behavioral therapy was not more effective than either treatment alone.¹⁹

Administration. Oral naltrexone is usually started at 25 mg and increased over 2 to 3 days to 50 or 100 mg/d. The standard dose is 50 mg/d, although the COMBINE study reported efficacy at 100 mg/d.¹⁹

Oral naltrexone is most helpful for patients who adhere to 70% to 90% of the medication.²⁰ The extended-release form (a 380-mg IM dose given every 4 weeks) provides an option to monitor adherence.²¹

Side effects. Naltrexone toxicity can cause hepatocellular injury. Do not administer this drug to patients with acute hepatitis or end-stage liver disease. When prescribing naltrexone, check patients’ liver function monthly for the first 3 months, then once every 3 months thereafter.²² Less serious, common side effects include nausea, myalgia, and headache.

Naltrexone antagonizes opioid receptors and causes withdrawal symptoms in patients who are physically dependent on opioids. Therefore, do not give naltrexone to patients who require opioids for chronic pain. If your patient is using an opioid but could switch to other pain medication, discontinue the opioid for at least 7 days and consider a urine toxicology or naltrexone challenge before starting naltrexone.

Urine drug tests are inexpensive and easy to use but have limitations. Many standard “dipsticks” will detect heroin, morphine, and codeine but not oxycodone, hydrocodone, or other synthetic opioids. Specific tests are available to detect oxycodone, hydrocodone, hydromorphone, buprenorphine, and methadone. Some synthetic opioids (such as fentanyl) remain difficult to detect, however, because of their low concentration and rapid metabolism.

Table 2

Naltrexone: Fast facts

Mechanism: Opioid receptor antagonism interferes with β-endorphin pathways
FDA-approved for alcohol dependence: Yes
Dosing: Oral 50 mg once daily (recent evidence suggests safety and efficacy at 100 mg once daily); IM 380 mg once every 4 weeks
Effect: Decreases frequency and severity of relapse
Potential side effects: Nausea, myalgia, headache, dizziness
Contraindications: Opioid use, acute hepatitis, liver failure
Comments: Monitor liver function; once-monthly dosing may improve adherence

Acamprosate

Acamprosate is structurally similar to gamma-aminobutyric acid (GABA) and is thought to inhibit the glutamatergic system. This attenuation by acamprosate reduces the glutamatergic hyperactivity normally seen after chronic alcohol exposure.

Acamprosate is indicated for relapse prevention in patients with alcohol dependence who have stopped drinking (Table 3). Multiple randomized studies have demonstrated its efficacy in improving abstinence rates as compared with placebo.^23,24 Other studies, however, failed to show improved abstinence with acamprosate.²⁵ Some of the negative studies included patients who recently relapsed or had only a few days of abstinence before starting acamprosate.^19,26 Therefore, acamprosate might be most effective when used to maintain abstinence and less effective—if at all—to initiate abstinence.

Administration. Acamprosate is available in 333-mg tablets, with a recommended dosage of 666 mg tid. Side effects tend to be transient and mild. Reduce the dose to 333 mg tid for patients with moderate renal insufficiency (creatinine clearance [CrCl] 30 to 50 mL/min), and do not use acamprosate in patients with severe renal insufficiency (CrCl

Side effects. Acamprosate was well-tolerated in clinical trials; diarrhea and other GI side effects were the most commonly reported adverse events.

In some placebo-controlled studies, patients taking acamprosate reported more frequent suicidal thoughts and attempts compared with patients taking placebo.²⁷ These events were extremely rare, and no direct relationship with acamprosate therapy has been established. Nonetheless, monitor patients for depression and suicidal thoughts during acamprosate therapy.

Table 3

Acamprosate: Fast facts

Mechanism: Structurally similar to GABA; thought to inhibit the glutamatergic system
FDA-approved for alcohol dependence: Yes
Dosing: 666 mg tid
Effect: Increases abstinence
Potential side effects: Nausea, diarrhea, suicidal thoughts
Contraindications: Severe renal disease
Comments: Monitor patients for suicidal thoughts and depression

Topiramate

Topiramate potentiates GABA and inhibits excitatory glutamate transmission—properties believed to lead to decreased dopamine release at the nucleus accumbens in response to alcohol consumption. Although topiramate is not FDA-approved for alcohol dependence, limited data comparing this anticonvulsant with placebo have shown a reduction in drinking and increased abstinence (Table 4).^28,29

 

Administration. Start with 25 mg/d and increase over several weeks to 300 mg/d, given in divided doses.

Side effects include dizziness, paresthesia, somnolence, difficulty concentrating, and weight loss. Because topiramate is excreted renally, reduce doses by 50% in patients with CrCl

Other renal side effects include an elevated risk of nephrolithiasis. Topiramate’s inhibition of carbonic anhydrase can reduce bicarbonate levels, leading to a nonanion gap metabolic acidosis.

Table 4

Topiramate: Fast facts

Mechanism: Potentiates GABA and inhibits glutamate receptor subtypes
FDA-approved for alcohol dependence: No
Dosing: 300 mg/d in divided doses
Effect: Decreases craving and drinking
Potential side effects: Metabolic acidosis, psychomotor slowing, dizziness, difficulty concentrating, paresthesia, weight loss, nephrolithiasis, hyperammonemia with concomitant use of valproic acid
Contraindications: None known other than hypersensitivity (as with all drugs)
Comments: Dose titration requires several weeks; avoid abrupt withdrawal; may reduce effectiveness of oral contraceptives

CASE CONTINUED: Implementing a treatment plan

You start Mr. G on oral naltrexone, 25 mg/d, and titrate to 100 mg/d. Although no optimum treatment duration has been established, you plan to follow NIAAA recommendations that Mr. G use naltrexone at least 3 months, with the possibility of continuing 1 year or longer if he responds well.¹

You schedule weekly visits for the first month to monitor for side effects and to make any necessary modifications in behavioral and pharmacologic treatment. You also continue escitalopram, 10 mg, which has successfully controlled Mr. G’s MDD symptoms.

Medications for alcohol dependence generally have been studied as adjuncts to behavioral therapies. The COMBINE study of 1,383 alcohol-abstinent patients found naltrexone with medical management or cognitive-behavioral therapy alone to be equivalent in efficacy.¹⁹ The medical management provided a supportive environment, encouraged medication compliance, provided empathy to build a therapeutic relationship, and promoted self-help groups as an adjunct to treatment.³⁰ Thus, medication has a role in treating alcohol dependence, but behavioral therapy remains an important part of comprehensive substance abuse treatment.

When choosing medications, consider the agents’ clinically relevant differences:

• Naltrexone and—less conclusively—topiramate have shown benefit for alcoholdependent patients starting treatment and for relapse prevention.
  
• Acamprosate may help prevent relapse in abstinent patients.
  
• Disulfiram remains a valid option in highly motivated patients with social support available to ensure medication adherence.
  

Because Mr. G is starting therapy after recent alcohol use, medications such as naltrexone and topiramate that have shown benefit early in treatment (in addition to relapse prevention) are preferred. Of these 2 drugs, naltrexone is the better choice for Mr. G—who is concerned about his work performance—because difficulty concentrating is a common side effect of topiramate. Oral naltrexone would be preferred as initial therapy for Mr. G because he has expressed comfort taking once-daily oral medication. The more expensive oncemonthly naltrexone depot formulation could be a second-line treatment if adherence becomes an issue.

Hypersensitivity is considered a contraindication for any medication. Mr. G tolerates well an initial dose of 25 mg/d, followed by increases to 50 mg and then 100 mg over several days. You titrate oral naltrexone to 100 mg/d—even though it is commonly prescribed at 50 mg/d—because recent evidence suggests efficacy and safety at the higher dosage.¹⁹

Related resources

• National Institutes of Health. Helping patients who drink too much: a clinician’s guide. Bethesda, MD: National Institute on Alcohol Abuse and Alcoholism; 2007. NIH Publication 07-3769. www.niaaa.nih.gov/Publications/EducationTrainingMaterials/guide.htm.
  
• American Society of Addiction Medicine. www.asam.org.
  
• American Academy of Addiction Psychiatry. www.aaap.org.
  

Drug brand names

• Acamprosate • Campral
  
• Disulfiram • Antabuse
  
• Escitalopram • Lexapro
  
• Naltrexone, oral • ReVia
  
• Naltrexone, extended-release • Vivitrol
  
• Topiramate • Topamax
  
• Valproic acid • Depakote
  

Disclosure

The authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Mr. G, age 38, is an investment banker referred for evaluation of an alcohol use disorder. Three years ago his internist diagnosed Mr. G with major depression and prescribed a selective serotonin reuptake inhibitor. Mr. G’s mood has improved, but his drinking is out of control and is affecting his work and marriage.

Mr. G describes his father as an alcoholic and says he has noticed worrisome similarities in himself. Since his teenage years, he recalls always being able to drink more than his peers. Amnesia episodes began in college during heavy drinking days and now occur almost weekly. Most recently his driver’s license was suspended after he was arrested for driving while impaired by alcohol.

He has attempted to stop drinking 3 times in the last 6 months and feels frustrated because he continues to relapse. During his last quit attempt, he remained abstinent for 3 months and believes his mood was unchanged during that time.

For motivated patients such as Mr. G, National Institute on Alcohol Abuse and Alcoholism (NIAAA) guidelines (updated in 2007) consider medications first-line treatment for alcohol dependence, along with psychotherapies and mutual-help groups such as Alcoholics Anonymous.¹ Medications with evidence of efficacy include FDA-approved disulfiram, naltrexone, and acamprosate, and off-label topiramate.

Each drug’s pharmacology is different; some may be beneficial during early abstinence, whereas others are more effective for maintaining abstinence. Because many physicians have had little or no experience using these medications,^2,3 we discuss dosing recommendations and side effect profiles—important clinical differences to guide drug selection and administration.

Box

CASE CONTINUED: Which came first?

When Mr. G presented with depressive symptoms, his internist informed him that alcohol’s effects can mimic depression and advised him to cut back his consumption. Mr. G temporarily reduced his drinking but continued to experience depressed mood, sleep disturbances, difficulty concentrating, fatigue, and poor appetite. The internist then prescribed escitalopram, 10 mg/d, and Mr. G says his depressive symptoms improved. He has not attempted suicide or required psychiatric hospitalization.

After a thorough evaluation—including a detailed assessment of his drinking history, other substance use, and mood symptoms—you diagnose Mr. G with alcohol dependence without physiological dependence and a primary major depressive disorder (MDD).

Mr. G’s diagnosis of alcohol dependence is based on evidence in the last 12 months of tolerance, repeated loss of control over the amount he drinks, multiple failed attempts to stop drinking, repeated negative consequences to his work productivity and personal relationships, and continued drinking despite knowing that alcohol consumption sometimes intensifies his mood symptoms (Box). Physiological dependence is unlikely because he did not experience withdrawal during a recent period of abstinence.

The mood component. Particularly in psychiatric patients, alcohol dependence often coexists with and affects the treatment of other psychopathologies:

• 1 in 3 adults experience an alcohol use disorder during their lifetimes.⁴
  
• 1 in 3 adults with an alcohol use disorder have a comorbid psychiatric disorder.⁵
  
• Alcohol dependence doubles the risk of major depression and triples the lifetime risk of any mood disorder.⁴
  

A thorough evaluation—both medical and psychiatric—is necessary to distinguish a primary mood disorder from a substance-induced mood disorder. Aspects of a patient’s history that may support a substance-induced disorder diagnosis include:

• Mood symptoms appear after the onset of a substance use disorder.
  
• Mood symptoms are absent during abstinence.
  
• Mood symptoms are consistent with the effects of the drug being used.⁶
  

Substance-induced mood symptoms usually become less intense and eventually resolve as abstinence is maintained. Primary mood symptoms may not be affected by abstinence, and interventions—including pharmacotherapy—are more likely to be required.

In evaluating Mr. G’s drinking history, you determined that his mood symptoms began before he started consuming alcohol regularly and have persisted during periods of abstinence. Thus, primary MDD is a more likely diagnosis than substance-induced MDD.

 

CASE CONTINUED: Planning treatment

A combination of behavioral therapy and pharmacotherapy is appropriate for treating Mr. G’s alcohol dependence. When you discuss the diagnosis with him, he endorses a goal of abstinence. For behavioral therapy, he says he would like to try Alcoholics Anonymous, which helped a friend “turn his life around.”

He has become accustomed to taking escitalopram once daily but is hesitant to take any medication that requires more frequent dosing. He also worries that medication might impair his work performance, which requires extensive periods of concentration. Your goal—as you consider available medications—is to develop a treatment plan that incorporates Mr. G’s preferences and addresses his concerns.

Disulfiram

Disulfiram, an irreversible aldehyde dehydrogenase inhibitor, is indicated for maintaining enforced sobriety in patients with chronic alcohol dependence (Table 1). Aldehyde dehydrogenase inhibition disrupts alcohol-to-acetate metabolism, which leads to acetaldehyde accumulation. If a disulfiram-treated patient ingests alcohol, increased acetaldehyde levels lead to the unpleasant “disulfiram-ethanol reaction,” with diaphoresis, flushing, nausea, vomiting, headache, tachycardia, and hypotension. The reaction’s severity is proportional to the disulfiram dose and amount of alcohol consumed.

Patients taking disulfiram must abstain from all alcohol, including over-the-counter cold remedies and mouthwashes containing alcohol. Advise patients that ingesting small amounts of alcohol can induce symptoms, even days after taking disulfiram.

Efficacy. Clinical trial results with disulfiram have been mixed. In the largest controlled study to date, Fuller et al⁷ found no significant difference in rates of total abstinence, time to first drink, employment, or social stability measures at 1 year among 605 men who received counseling plus disulfiram, 250 mg/d, or placebo. Other disulfiram studies have found a modest decrease in the frequency of drinking but no effect on abstinence rates.^8,9

Because medication adherence is the strongest predictor of outcome with disulfiram,¹⁰ monitoring for adherence and stressing its importance to patients may increase the drug’s efficacy. Disulfiram may be most effective in highly motivated patients with stable social support or as an adjuvant to an outpatient treatment program.

Administration. Disulfiram is available in 250-mg tablets and is usually dosed from 125 to 500 mg/d. Treatment can begin after patients abstain from alcohol for ≥12 hours and have a serum alcohol concentration of zero.

Side effects. Drowsiness is a common complaint with disulfiram; this adverse effect is frequently self-limited and can be reduced by evening dosing.

Subclinical liver enzyme elevations have been reported in 25% of patients taking disulfiram.¹¹ Although rare, potentially fatal hepatotoxicity has been reported^12,13 (with a dose as low as 200 mg/d¹²), typically occurring early in treatment and associated with jaundice and fever. One study estimated the risk of dying of hepatotoxicity caused by disulfiram to be 1 in 30,000 patients/year.¹⁴

A recent Swedish study¹³ reviewed data from 1966 through 2002 and found 82 cases of drug-induced liver injury associated with disulfiram. By comparing these findings with sales figures from 1972 to 2002, the authors report an incidence of disulfiram-induced liver injury of about 1 case per 1.3 million estimated average daily doses.

Order liver function tests at baseline, then retest 10 to 14 days after starting disulfiram and again approximately 4 weeks later. Thereafter, monitoring once every 3 to 6 months is generally sufficient in patients without liver disease symptoms.

Other serious adverse events associated with disulfiram therapy include optic neuritis, peripheral neuritis, cholestatic hepatitis, seizures, and arrhythmias. Psychosis also can occur, generally with dosages ≥500 mg/d. Avoid concomitant use of disulfiram and metronidazole, which can cause acute psychosis.

Disulfiram-related inhibition of cytochrome P-450 can increase serum levels and toxicity risk of medications metabolized in the liver, such as warfarin, phenytoin, and isoniazid. Patients taking concomitant warfarin and disulfiram require close monitoring for increases in the international normalized ratio (INR).

Contraindications. Disulfiram is contraindicated in patients with ischemic heart disease and those who are pregnant. Also avoid disulfiram in patients with cerebrovascular disease, diabetes mellitus, psychosis, or cognitive impairment.

Recommendation. Disulfiram is a valid option for treating alcohol dependence in a select group of highly motivated patients who are medically and psychiatrically stable and in whom adherence can be closely monitored.

Table 1

Disulfiram: Fast facts

Mechanism: Acetaldehyde accumulates when aldehyde dehydrogenase is inhibited
FDA-approved for alcohol dependence: Yes
Dosing: 125 to 500 mg once daily
Effect: Aversive reaction to alcohol
Potential side effects: Liver toxicity, seizures, arrhythmia, peripheral neuropathy, psychosis
Contraindications: Concurrent alcohol consumption, severe cardiac disease, psychosis, pregnancy
Comments: Many drug interactions, including warfarin, metronidazole, and phenytoin; monitor liver function for toxicity

 

Naltrexone

Naltrexone is a μ-opioid receptor antagonist thought to reduce alcohol’s reinforcing effects by interfering with β-endorphin pathways. It is indicated for treating alcohol dependence and has been shown to reduce relapse and number of drinking days in alcohol-dependent patients (Table 2).¹⁵

Naltrexone also has reduced alcohol consumption in healthy volunteers, social drinkers, and other nondependent drinkers.¹⁶ Its effect may have a genetic component, as suggested by greater benefit in persons with a family history of alcoholism.¹⁷

Efficacy. Most studies investigated naltrexone as part of a comprehensive treatment program that included behavioral therapies.¹⁸ Recently, the randomized, placebo-controlled Combining Medications and Behavioral Interventions (COMBINE) study found that using either naltrexone or behavioral therapy improved abstinence, and combing naltrexone with behavioral therapy was not more effective than either treatment alone.¹⁹

Administration. Oral naltrexone is usually started at 25 mg and increased over 2 to 3 days to 50 or 100 mg/d. The standard dose is 50 mg/d, although the COMBINE study reported efficacy at 100 mg/d.¹⁹

Oral naltrexone is most helpful for patients who adhere to 70% to 90% of the medication.²⁰ The extended-release form (a 380-mg IM dose given every 4 weeks) provides an option to monitor adherence.²¹

Side effects. Naltrexone toxicity can cause hepatocellular injury. Do not administer this drug to patients with acute hepatitis or end-stage liver disease. When prescribing naltrexone, check patients’ liver function monthly for the first 3 months, then once every 3 months thereafter.²² Less serious, common side effects include nausea, myalgia, and headache.

Naltrexone antagonizes opioid receptors and causes withdrawal symptoms in patients who are physically dependent on opioids. Therefore, do not give naltrexone to patients who require opioids for chronic pain. If your patient is using an opioid but could switch to other pain medication, discontinue the opioid for at least 7 days and consider a urine toxicology or naltrexone challenge before starting naltrexone.

Urine drug tests are inexpensive and easy to use but have limitations. Many standard “dipsticks” will detect heroin, morphine, and codeine but not oxycodone, hydrocodone, or other synthetic opioids. Specific tests are available to detect oxycodone, hydrocodone, hydromorphone, buprenorphine, and methadone. Some synthetic opioids (such as fentanyl) remain difficult to detect, however, because of their low concentration and rapid metabolism.

Table 2

Naltrexone: Fast facts

Mechanism: Opioid receptor antagonism interferes with β-endorphin pathways
FDA-approved for alcohol dependence: Yes
Dosing: Oral 50 mg once daily (recent evidence suggests safety and efficacy at 100 mg once daily); IM 380 mg once every 4 weeks
Effect: Decreases frequency and severity of relapse
Potential side effects: Nausea, myalgia, headache, dizziness
Contraindications: Opioid use, acute hepatitis, liver failure
Comments: Monitor liver function; once-monthly dosing may improve adherence

Acamprosate

Acamprosate is structurally similar to gamma-aminobutyric acid (GABA) and is thought to inhibit the glutamatergic system. This attenuation by acamprosate reduces the glutamatergic hyperactivity normally seen after chronic alcohol exposure.

Acamprosate is indicated for relapse prevention in patients with alcohol dependence who have stopped drinking (Table 3). Multiple randomized studies have demonstrated its efficacy in improving abstinence rates as compared with placebo.^23,24 Other studies, however, failed to show improved abstinence with acamprosate.²⁵ Some of the negative studies included patients who recently relapsed or had only a few days of abstinence before starting acamprosate.^19,26 Therefore, acamprosate might be most effective when used to maintain abstinence and less effective—if at all—to initiate abstinence.

Administration. Acamprosate is available in 333-mg tablets, with a recommended dosage of 666 mg tid. Side effects tend to be transient and mild. Reduce the dose to 333 mg tid for patients with moderate renal insufficiency (creatinine clearance [CrCl] 30 to 50 mL/min), and do not use acamprosate in patients with severe renal insufficiency (CrCl

Side effects. Acamprosate was well-tolerated in clinical trials; diarrhea and other GI side effects were the most commonly reported adverse events.

In some placebo-controlled studies, patients taking acamprosate reported more frequent suicidal thoughts and attempts compared with patients taking placebo.²⁷ These events were extremely rare, and no direct relationship with acamprosate therapy has been established. Nonetheless, monitor patients for depression and suicidal thoughts during acamprosate therapy.

Table 3

Acamprosate: Fast facts

Mechanism: Structurally similar to GABA; thought to inhibit the glutamatergic system
FDA-approved for alcohol dependence: Yes
Dosing: 666 mg tid
Effect: Increases abstinence
Potential side effects: Nausea, diarrhea, suicidal thoughts
Contraindications: Severe renal disease
Comments: Monitor patients for suicidal thoughts and depression

Topiramate

Topiramate potentiates GABA and inhibits excitatory glutamate transmission—properties believed to lead to decreased dopamine release at the nucleus accumbens in response to alcohol consumption. Although topiramate is not FDA-approved for alcohol dependence, limited data comparing this anticonvulsant with placebo have shown a reduction in drinking and increased abstinence (Table 4).^28,29

 

Administration. Start with 25 mg/d and increase over several weeks to 300 mg/d, given in divided doses.

Side effects include dizziness, paresthesia, somnolence, difficulty concentrating, and weight loss. Because topiramate is excreted renally, reduce doses by 50% in patients with CrCl

Other renal side effects include an elevated risk of nephrolithiasis. Topiramate’s inhibition of carbonic anhydrase can reduce bicarbonate levels, leading to a nonanion gap metabolic acidosis.

Table 4

Topiramate: Fast facts

Mechanism: Potentiates GABA and inhibits glutamate receptor subtypes
FDA-approved for alcohol dependence: No
Dosing: 300 mg/d in divided doses
Effect: Decreases craving and drinking
Potential side effects: Metabolic acidosis, psychomotor slowing, dizziness, difficulty concentrating, paresthesia, weight loss, nephrolithiasis, hyperammonemia with concomitant use of valproic acid
Contraindications: None known other than hypersensitivity (as with all drugs)
Comments: Dose titration requires several weeks; avoid abrupt withdrawal; may reduce effectiveness of oral contraceptives

CASE CONTINUED: Implementing a treatment plan

You start Mr. G on oral naltrexone, 25 mg/d, and titrate to 100 mg/d. Although no optimum treatment duration has been established, you plan to follow NIAAA recommendations that Mr. G use naltrexone at least 3 months, with the possibility of continuing 1 year or longer if he responds well.¹

You schedule weekly visits for the first month to monitor for side effects and to make any necessary modifications in behavioral and pharmacologic treatment. You also continue escitalopram, 10 mg, which has successfully controlled Mr. G’s MDD symptoms.

Medications for alcohol dependence generally have been studied as adjuncts to behavioral therapies. The COMBINE study of 1,383 alcohol-abstinent patients found naltrexone with medical management or cognitive-behavioral therapy alone to be equivalent in efficacy.¹⁹ The medical management provided a supportive environment, encouraged medication compliance, provided empathy to build a therapeutic relationship, and promoted self-help groups as an adjunct to treatment.³⁰ Thus, medication has a role in treating alcohol dependence, but behavioral therapy remains an important part of comprehensive substance abuse treatment.

When choosing medications, consider the agents’ clinically relevant differences:

• Naltrexone and—less conclusively—topiramate have shown benefit for alcoholdependent patients starting treatment and for relapse prevention.
  
• Acamprosate may help prevent relapse in abstinent patients.
  
• Disulfiram remains a valid option in highly motivated patients with social support available to ensure medication adherence.
  

Because Mr. G is starting therapy after recent alcohol use, medications such as naltrexone and topiramate that have shown benefit early in treatment (in addition to relapse prevention) are preferred. Of these 2 drugs, naltrexone is the better choice for Mr. G—who is concerned about his work performance—because difficulty concentrating is a common side effect of topiramate. Oral naltrexone would be preferred as initial therapy for Mr. G because he has expressed comfort taking once-daily oral medication. The more expensive oncemonthly naltrexone depot formulation could be a second-line treatment if adherence becomes an issue.

Hypersensitivity is considered a contraindication for any medication. Mr. G tolerates well an initial dose of 25 mg/d, followed by increases to 50 mg and then 100 mg over several days. You titrate oral naltrexone to 100 mg/d—even though it is commonly prescribed at 50 mg/d—because recent evidence suggests efficacy and safety at the higher dosage.¹⁹

Related resources

• National Institutes of Health. Helping patients who drink too much: a clinician’s guide. Bethesda, MD: National Institute on Alcohol Abuse and Alcoholism; 2007. NIH Publication 07-3769. www.niaaa.nih.gov/Publications/EducationTrainingMaterials/guide.htm.
  
• American Society of Addiction Medicine. www.asam.org.
  
• American Academy of Addiction Psychiatry. www.aaap.org.
  

Drug brand names

• Acamprosate • Campral
  
• Disulfiram • Antabuse
  
• Escitalopram • Lexapro
  
• Naltrexone, oral • ReVia
  
• Naltrexone, extended-release • Vivitrol
  
• Topiramate • Topamax
  
• Valproic acid • Depakote
  

Disclosure

The authors report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Diagnosis and treatment of neuroleptic malignant syndrome (NMS) are controversial because this potentially life-threatening syndrome is rare and its presentation varies. These factors make it difficult to evaluate treatments in controlled clinical trials, and data about the relative efficacy of specific interventions are scarce. It may be possible, however, to develop rational treatment guidelines using empiric clinical data.^1,2

This article examines the evidence related to 6 controversial aspects of NMS diagnosis and treatment:

• most-reliable risk factors
• NMS as a spectrum disorder
• what causes NMS
• NMS triggered by first-generation vs second-generation antipsychotics
• first-line interventions
• restarting antipsychotics after an NMS episode.

1. Are there reliable risk factors for NMS?

In small case-controlled studies, agitation, dehydration, and exhaustion were the most consistently found systemic factors believed to predispose patients taking antipsychotics to NMS (Table 1).^3-5 Catatonia and organic brain syndromes may be separate risk factors.^1,6

Preliminary studies also have implicated dopamine receptor abnormalities caused by genetic polymorphisms or effects of low serum iron.^1,7,8 Pharmacologic studies have suggested that higher doses, rapid titration, and IM injections of antipsychotics are associated with increased NMS risk.^3,5 Some studies suggest that 15% to 20% of NMS patients have a history of NMS episodes.^1,2 In addition, high-potency first-generation antipsychotics (FGAs)—especially haloperidol—are assumed to carry higher risk than low-potency drugs and second-generation antipsychotics (SGAs), although this hypothesis remains difficult to prove.^9-11

These risk factors, however, are not practical for estimating NMS risk in a given patient because they are relatively common compared with the low risk of NMS occurrence. For the vast majority of patients with psychotic symptoms, the benefits of properly indicated antipsychotic pharmacotherapy will outweigh the risks.

Table 1

Systemic

Agitation

Dehydration

Exhaustion

Low serum iron concentrations (normal: 60 to 170 mcg/dL)

Diagnoses

History of NMS

Catatonia

Organic brain syndromes

Central nervous system

Dopamine receptor dysfunction

Basal ganglia dysfunction

Sympathetic nervous system dysfunction

Pharmacologic treatment*

Intramuscular or intravenous injections

High-potency dopamine antagonists

Rapid dose titration

High doses

FGAs compared with SGAs (?)

*For individual patients, these common risk factors must be weighted again the benefits of antipsychotic therapy FGAs: first-generation antipsychotics; SGAs:second-generation antipsychotics; NMS: neuroleptic malignant syndromeSource: References 1-5

2. Is NMS related to parkinsonism, catatonia, or malignant hyperthermia?

Parkinsonsim. Some researchers have described NMS as an extreme parkinsonian crisis resulting from overwhelming blockade of dopamine pathways in the brain.^1,2,12 In this view, NMS resembles the parkinsonian-hyperthermia syndrome that can occur in Parkinson's disease patients following abrupt discontinuation or loss of efficacy of dopaminergic therapy, which can be treated by reinstituting dopaminergic agents.¹³ Evidence to support this view includes:

• Parkinsonian signs are a cardinal feature of NMS.
• Withdrawal of dopamine agonists precipitates the syndrome.
• All triggering drugs are dopamine receptor antagonists.
• Risks of NMS correlates with drugs' dopamine receptor affinity.
• Dopaminergic agonists may be an effective treatment.
• Lesions in dopaminergic pathways produce a similar syndrome.
• Patients with NMS have demonstrated low cerebrospinal fluid concentrations of the dopamine metabolite homovanillic acid.¹⁴

Catatonia. Fink et al¹⁵ and others^16-18 have persuasively argued that NMS represents a form of drug-induced malignant catatonia. Evidence supporting this includes:

• The 2 disorders share neuropsychiatric symptoms.
• Catalonic signs are common in NMS.¹⁹
• Malignant catatonia and NMS share physiologic and labratory signs.²⁰
• Reintroduction of antipsychotics can acutely worsen both conditions.
• Benzodiazepines and electroconvulsive therapy (ECT) are effective treatments for both disorders.^15-18

Lee²¹ examined the relationship between catatonic features and treatment response of 14 NMS patients. Most patients with catatonic symptoms responded to benzodiazepines, whereas none of those did who had an extrapyramidal-hyperthermic presentation without catatonia. Lee concluded that NMS is heterogeneous and may occur in catatonic and noncatatonic forms that differ in treatment response.

Document

file8nCZdF

Malignant hyperthermia. Some clinicians have compared NMS with malignant hyperthermia caused by inhalational anesthetics and succinylcholine.^1,2 Evidence includes

• similar clinical signs of rigidity, hyperthermia, and hypermetabolism
• similar psychologic and labratory signs, such as rhabdomyolysis
• hyperthermia in both responding to dantrolene.

Although the 2 are similar in presentation, malignant hyperthermia occurs intraoperatively and reflects a pharmacogenetic disorder of calcium regulation in skeletal muscle. Additionally, rigidity in malignant hyperthermia does not respond to peripheral-acting muscle relaxants.^1,22 Evidence suggests that patients who have previously experienced an NMS episodes are not at risk for malignant hyperthermia.²²

3. What is the pathophysiology of NMS?

NMS pathophysiology is complex and likely involves interplay between multiple central and systemic pathways and neurotransmitters. As described above, compelling evidence suggests that dopamine blockade plays a central role.¹²

Dopamine blockade in the hypothalamus is believed to contribute to thermoregulatory failure, and blockade in the nigrostriatal system likely contributes to muscle rigidity and hypermetabolism. The loss of dopaminergic input to the anterior cingulate-medial orbitofrontal circuit and the lateral orbitofrontal circuit likely con-tributes to the mental status changes and catatonic features seen in NMS.¹²

 

Some researchers have proposed competing or complementary hypotheses, however. For example, Gurrera²³ proposed that patients who are prone to developing NMS have a vulnerability to a hyperactive and dysregulated sympathetic nervous system, and this trait—together with dopamine system disruption induced by dopamine-blocking agents—produces NMS. Other investigators have implicated serotonin, norepinephrine, gamma-aminobutyric acid and glutaminergic mechanisms.^1,12,24,25

4. Are FGAs or SGAs more likely to cause NMS?

NMS is assumed to occur less frequently in patients treated with SGAs than in those receiving FGAs, although this hypothesisis unproven. Isolated reports of NMS have been associated with nearly every SGA.^9-11 It is difficult to prove FGA vs SGA liabilities because:

• NMS is rare.
• Dosing practices may be more conser-vative now than in the past.
• Most clinicians are aware of the earlysigns of NMS.

In an epidemiological study of a large database, Stubner et al²⁶ found that patients receiving SGAs had a lower risk of NMS than those treated with haloperidol.²⁶ In this study, the overall rate of NMS was 0.02%.

NMS hotline data. We recently examined which medication classes were implicated in 111 NMS cases reported to the Neuroleptic Malignant Syndrome Information Service hotline (1-888-NMS-TEMP) between 1997 and 2006 (Figure). We included only cases of definite or probable NMS (as diagnosed by hotline consultants) in which a single antipsychotic was administered. Slightly more cases were attributed to FGAs (51%) than SGAs (45%). The remaining cases were attributed to neuroleptics used in medical settings (such as promethazineor prochlorperazine). Because they are now prescribed less often, FGAs accounted for a disproportionate number of NMS cases reported to the hotline. Haloperidol accounted for the majority of FGA cases and 44% of all cases. If we had excluded haloperidol and compared the NMS risk of SGAs to only intermediate- or low-potency FGAs, the relative advantage of SGAs would have been lost. On the other hand, it is clear that SGAs still carry a risk for NMS. Analyses suggest that the SGA-associated classic features of NMS—fever, muscle rigidity, and autonomic and mental status changes—are retained in patients receiving SGAs, although some may not develop the severe rigidity and extreme temperatures common in patients receiving FGAs.^9-11 The milder clinical characteristics associated with SGAs may reflect more conservative prescribing patterns or increased awareness and earlier recognition of NMS, which would prevent fulminant presentations.

5. What is the evidence for specific NMS treatments?

NMS is rare, its presentation varies, and its progression is unpredictable. These factors make it difficult to evaluate treatments in controlled clinical trials, and data about the relative efficacy of specific interventions are scarce.

Even so, the notion that NMS represents an extreme variant of drug-induced parkinsonism or catatonia suggests that specific NMS treatments could be based on symptom severity or stage of presentation. We propose a treatment guideline basedon theoretical mechanisms and anecdotal data (Algorithm).^2,27-29

Support. After immediate withdrawal of the offending medication, supportive therapy is the cornerstone of NMS treatment.^1,2,27

For patients presenting with mild signs and symptoms, supportive care and careful clinical monitoring may be sufficient. Extreme hyperthermia demands volume resuscitation and cooling measures, intensive medical care, and careful monitoring for complications.

Treatment. Despite a lack of consensus on drug treatments for uncomplicated NMS, approximately 40% of patients with acute NMS receive pharmacologic treatments.²

Lorazepam, 1 to 2 mg parenterally, is a reasonable first-line therapy for NMS, especially in individuals with catatonic features.^{4,15-18,21,30,31} Some investigators recommend higher doses.¹⁵ Benzodiazepines are preferred if sedation is required in agitated NMS patients.^4,15-18

Dopaminergic agents such as bromocriptine and amantadine enhance dopaminergic transmission to reverse parkinsonian symptoms and have been reported to reduce time to recovery and halve mortality rates when used alone or in conjunction with other treatments.^13,27,32,33 Rapid discontinuation of these agents can result in rebound symptoms, although this may be true for any specific drug treatment of NMS.^1,31,32

Dantrolene uncouples excitation-contraction coupling by enhancing calcium sequestration in sarcoplasmic reticulumin skeletal muscle and has been used to treat NMS hypermetabolic symptoms. Some reviews found improvement in up to 80% of NMS patients treated with dantrolene monotherapy.^27,32-35 Compared with supportive care, time to recovery may be reduced—and mortality decreased by almost one-half—when dantrolene is used alone or in combination with other medications.

Not all case reports have shown that dantrolene, benzodiazepines, ordopaminergic agonists are effective in treating NMS.^31,36 In our opinion, only advanced NMS cases—with extreme temperature elevations, severe rigidity, and evidence of systemic hypermetabolism—benefit from dantrolene treatment.^1,2

ECT has been used successfully to reduce mortality from NMS and other catatonic-spectrum disorders. It is usually employed after supportive therapy and psychopharmacologic interventions fail.^{2,15,16,27,37} ECT for acute NMS typically consists of a series of 6 to 10 treatments with bilateral electrode placement. Daily ECT may be needed initially.¹⁵

 

6. Are antipsychotics contraindicated following an NMS episode?

The rate of NMS recurrence on retreatment with an antipsychotic has varied.38 We estimate that up to 30% of patients may be at risk of NMS recurrence when rechallenged with an antipsychotic.1 By following proper precautions (Table 2), however, you can safely treat most patients who require continued antipsychotic therapy.^1,2 When you restart treatment, a lower-potency antipsychotic from a different chemical class may be a safer option than retrying the triggering agent, according to retrospective analyses of limited available data. A patient who develops NMS on a FGA might benefit from an SGA trial, although some risk of recurrence remains.^1,10

Current Psychiatry 2007;6(8):89-95.
Drug brand names

• Amantadine • Symmetrel
• Bromocriptine • Parlodel
• Chlorpromazine • Thorazine
• Dantrolene • Dantrium
• Fluphenazine • Prolixin
• Haloperidol • Haldol
• Lorazepam • Ativan
• Loxapine • Loxitane
• Perphenazine • Trilafon
• Prochlorperazine • Compazine, Compro
• Promethazine • Phenergan
• Thioridazine • Mellaril

Disclosure

Dr. Strawn is an American Psychiatric Institute for Research and Education (APIRE)/Janssen Scholar.

Dr. Keck has received research support from or served as a consultant to Abbott Laboratories, American Diabetes Association, AstraZeneca Pharmaceuticals, Bristol-Myers Squibb, GlaxoSmithKline, Eli Lilly and Company, Janssen Pharmaceutica, National Institute of Mental Health, National Institute of Drug Abuse, Pfizer, Stanley Medical Research Institute, and UCB Pharma.

Dr. Caroff has received research support from Bristol-Myers Squibb, Ortho-McNeil Neurologics, and Pfizer.

Diagnosis and treatment of neuroleptic malignant syndrome (NMS) are controversial because this potentially life-threatening syndrome is rare and its presentation varies. These factors make it difficult to evaluate treatments in controlled clinical trials, and data about the relative efficacy of specific interventions are scarce. It may be possible, however, to develop rational treatment guidelines using empiric clinical data.^1,2

This article examines the evidence related to 6 controversial aspects of NMS diagnosis and treatment:

• most-reliable risk factors
• NMS as a spectrum disorder
• what causes NMS
• NMS triggered by first-generation vs second-generation antipsychotics
• first-line interventions
• restarting antipsychotics after an NMS episode.

1. Are there reliable risk factors for NMS?

In small case-controlled studies, agitation, dehydration, and exhaustion were the most consistently found systemic factors believed to predispose patients taking antipsychotics to NMS (Table 1).^3-5 Catatonia and organic brain syndromes may be separate risk factors.^1,6

Preliminary studies also have implicated dopamine receptor abnormalities caused by genetic polymorphisms or effects of low serum iron.^1,7,8 Pharmacologic studies have suggested that higher doses, rapid titration, and IM injections of antipsychotics are associated with increased NMS risk.^3,5 Some studies suggest that 15% to 20% of NMS patients have a history of NMS episodes.^1,2 In addition, high-potency first-generation antipsychotics (FGAs)—especially haloperidol—are assumed to carry higher risk than low-potency drugs and second-generation antipsychotics (SGAs), although this hypothesis remains difficult to prove.^9-11

These risk factors, however, are not practical for estimating NMS risk in a given patient because they are relatively common compared with the low risk of NMS occurrence. For the vast majority of patients with psychotic symptoms, the benefits of properly indicated antipsychotic pharmacotherapy will outweigh the risks.

Table 1

Systemic

Agitation

Dehydration

Exhaustion

Low serum iron concentrations (normal: 60 to 170 mcg/dL)

Diagnoses

History of NMS

Catatonia

Organic brain syndromes

Central nervous system

Dopamine receptor dysfunction

Basal ganglia dysfunction

Sympathetic nervous system dysfunction

Pharmacologic treatment*

Intramuscular or intravenous injections

High-potency dopamine antagonists

Rapid dose titration

High doses

FGAs compared with SGAs (?)

*For individual patients, these common risk factors must be weighted again the benefits of antipsychotic therapy FGAs: first-generation antipsychotics; SGAs:second-generation antipsychotics; NMS: neuroleptic malignant syndromeSource: References 1-5

2. Is NMS related to parkinsonism, catatonia, or malignant hyperthermia?

Parkinsonsim. Some researchers have described NMS as an extreme parkinsonian crisis resulting from overwhelming blockade of dopamine pathways in the brain.^1,2,12 In this view, NMS resembles the parkinsonian-hyperthermia syndrome that can occur in Parkinson's disease patients following abrupt discontinuation or loss of efficacy of dopaminergic therapy, which can be treated by reinstituting dopaminergic agents.¹³ Evidence to support this view includes:

• Parkinsonian signs are a cardinal feature of NMS.
• Withdrawal of dopamine agonists precipitates the syndrome.
• All triggering drugs are dopamine receptor antagonists.
• Risks of NMS correlates with drugs' dopamine receptor affinity.
• Dopaminergic agonists may be an effective treatment.
• Lesions in dopaminergic pathways produce a similar syndrome.
• Patients with NMS have demonstrated low cerebrospinal fluid concentrations of the dopamine metabolite homovanillic acid.¹⁴

Catatonia. Fink et al¹⁵ and others^16-18 have persuasively argued that NMS represents a form of drug-induced malignant catatonia. Evidence supporting this includes:

• The 2 disorders share neuropsychiatric symptoms.
• Catalonic signs are common in NMS.¹⁹
• Malignant catatonia and NMS share physiologic and labratory signs.²⁰
• Reintroduction of antipsychotics can acutely worsen both conditions.
• Benzodiazepines and electroconvulsive therapy (ECT) are effective treatments for both disorders.^15-18

Lee²¹ examined the relationship between catatonic features and treatment response of 14 NMS patients. Most patients with catatonic symptoms responded to benzodiazepines, whereas none of those did who had an extrapyramidal-hyperthermic presentation without catatonia. Lee concluded that NMS is heterogeneous and may occur in catatonic and noncatatonic forms that differ in treatment response.

Document

file8nCZdF

Malignant hyperthermia. Some clinicians have compared NMS with malignant hyperthermia caused by inhalational anesthetics and succinylcholine.^1,2 Evidence includes

• similar clinical signs of rigidity, hyperthermia, and hypermetabolism
• similar psychologic and labratory signs, such as rhabdomyolysis
• hyperthermia in both responding to dantrolene.

Although the 2 are similar in presentation, malignant hyperthermia occurs intraoperatively and reflects a pharmacogenetic disorder of calcium regulation in skeletal muscle. Additionally, rigidity in malignant hyperthermia does not respond to peripheral-acting muscle relaxants.^1,22 Evidence suggests that patients who have previously experienced an NMS episodes are not at risk for malignant hyperthermia.²²

3. What is the pathophysiology of NMS?

NMS pathophysiology is complex and likely involves interplay between multiple central and systemic pathways and neurotransmitters. As described above, compelling evidence suggests that dopamine blockade plays a central role.¹²

Dopamine blockade in the hypothalamus is believed to contribute to thermoregulatory failure, and blockade in the nigrostriatal system likely contributes to muscle rigidity and hypermetabolism. The loss of dopaminergic input to the anterior cingulate-medial orbitofrontal circuit and the lateral orbitofrontal circuit likely con-tributes to the mental status changes and catatonic features seen in NMS.¹²

 

Some researchers have proposed competing or complementary hypotheses, however. For example, Gurrera²³ proposed that patients who are prone to developing NMS have a vulnerability to a hyperactive and dysregulated sympathetic nervous system, and this trait—together with dopamine system disruption induced by dopamine-blocking agents—produces NMS. Other investigators have implicated serotonin, norepinephrine, gamma-aminobutyric acid and glutaminergic mechanisms.^1,12,24,25

4. Are FGAs or SGAs more likely to cause NMS?

NMS is assumed to occur less frequently in patients treated with SGAs than in those receiving FGAs, although this hypothesisis unproven. Isolated reports of NMS have been associated with nearly every SGA.^9-11 It is difficult to prove FGA vs SGA liabilities because:

• NMS is rare.
• Dosing practices may be more conser-vative now than in the past.
• Most clinicians are aware of the earlysigns of NMS.

In an epidemiological study of a large database, Stubner et al²⁶ found that patients receiving SGAs had a lower risk of NMS than those treated with haloperidol.²⁶ In this study, the overall rate of NMS was 0.02%.

NMS hotline data. We recently examined which medication classes were implicated in 111 NMS cases reported to the Neuroleptic Malignant Syndrome Information Service hotline (1-888-NMS-TEMP) between 1997 and 2006 (Figure). We included only cases of definite or probable NMS (as diagnosed by hotline consultants) in which a single antipsychotic was administered. Slightly more cases were attributed to FGAs (51%) than SGAs (45%). The remaining cases were attributed to neuroleptics used in medical settings (such as promethazineor prochlorperazine). Because they are now prescribed less often, FGAs accounted for a disproportionate number of NMS cases reported to the hotline. Haloperidol accounted for the majority of FGA cases and 44% of all cases. If we had excluded haloperidol and compared the NMS risk of SGAs to only intermediate- or low-potency FGAs, the relative advantage of SGAs would have been lost. On the other hand, it is clear that SGAs still carry a risk for NMS. Analyses suggest that the SGA-associated classic features of NMS—fever, muscle rigidity, and autonomic and mental status changes—are retained in patients receiving SGAs, although some may not develop the severe rigidity and extreme temperatures common in patients receiving FGAs.^9-11 The milder clinical characteristics associated with SGAs may reflect more conservative prescribing patterns or increased awareness and earlier recognition of NMS, which would prevent fulminant presentations.

5. What is the evidence for specific NMS treatments?

NMS is rare, its presentation varies, and its progression is unpredictable. These factors make it difficult to evaluate treatments in controlled clinical trials, and data about the relative efficacy of specific interventions are scarce.

Even so, the notion that NMS represents an extreme variant of drug-induced parkinsonism or catatonia suggests that specific NMS treatments could be based on symptom severity or stage of presentation. We propose a treatment guideline basedon theoretical mechanisms and anecdotal data (Algorithm).^2,27-29

Support. After immediate withdrawal of the offending medication, supportive therapy is the cornerstone of NMS treatment.^1,2,27

For patients presenting with mild signs and symptoms, supportive care and careful clinical monitoring may be sufficient. Extreme hyperthermia demands volume resuscitation and cooling measures, intensive medical care, and careful monitoring for complications.

Treatment. Despite a lack of consensus on drug treatments for uncomplicated NMS, approximately 40% of patients with acute NMS receive pharmacologic treatments.²

Lorazepam, 1 to 2 mg parenterally, is a reasonable first-line therapy for NMS, especially in individuals with catatonic features.^{4,15-18,21,30,31} Some investigators recommend higher doses.¹⁵ Benzodiazepines are preferred if sedation is required in agitated NMS patients.^4,15-18

Dopaminergic agents such as bromocriptine and amantadine enhance dopaminergic transmission to reverse parkinsonian symptoms and have been reported to reduce time to recovery and halve mortality rates when used alone or in conjunction with other treatments.^13,27,32,33 Rapid discontinuation of these agents can result in rebound symptoms, although this may be true for any specific drug treatment of NMS.^1,31,32

Dantrolene uncouples excitation-contraction coupling by enhancing calcium sequestration in sarcoplasmic reticulumin skeletal muscle and has been used to treat NMS hypermetabolic symptoms. Some reviews found improvement in up to 80% of NMS patients treated with dantrolene monotherapy.^27,32-35 Compared with supportive care, time to recovery may be reduced—and mortality decreased by almost one-half—when dantrolene is used alone or in combination with other medications.

Not all case reports have shown that dantrolene, benzodiazepines, ordopaminergic agonists are effective in treating NMS.^31,36 In our opinion, only advanced NMS cases—with extreme temperature elevations, severe rigidity, and evidence of systemic hypermetabolism—benefit from dantrolene treatment.^1,2

ECT has been used successfully to reduce mortality from NMS and other catatonic-spectrum disorders. It is usually employed after supportive therapy and psychopharmacologic interventions fail.^{2,15,16,27,37} ECT for acute NMS typically consists of a series of 6 to 10 treatments with bilateral electrode placement. Daily ECT may be needed initially.¹⁵

 

6. Are antipsychotics contraindicated following an NMS episode?

The rate of NMS recurrence on retreatment with an antipsychotic has varied.38 We estimate that up to 30% of patients may be at risk of NMS recurrence when rechallenged with an antipsychotic.1 By following proper precautions (Table 2), however, you can safely treat most patients who require continued antipsychotic therapy.^1,2 When you restart treatment, a lower-potency antipsychotic from a different chemical class may be a safer option than retrying the triggering agent, according to retrospective analyses of limited available data. A patient who develops NMS on a FGA might benefit from an SGA trial, although some risk of recurrence remains.^1,10

Current Psychiatry 2007;6(8):89-95.
Drug brand names

• Amantadine • Symmetrel
• Bromocriptine • Parlodel
• Chlorpromazine • Thorazine
• Dantrolene • Dantrium
• Fluphenazine • Prolixin
• Haloperidol • Haldol
• Lorazepam • Ativan
• Loxapine • Loxitane
• Perphenazine • Trilafon
• Prochlorperazine • Compazine, Compro
• Promethazine • Phenergan
• Thioridazine • Mellaril

Disclosure

Dr. Strawn is an American Psychiatric Institute for Research and Education (APIRE)/Janssen Scholar.

Dr. Keck has received research support from or served as a consultant to Abbott Laboratories, American Diabetes Association, AstraZeneca Pharmaceuticals, Bristol-Myers Squibb, GlaxoSmithKline, Eli Lilly and Company, Janssen Pharmaceutica, National Institute of Mental Health, National Institute of Drug Abuse, Pfizer, Stanley Medical Research Institute, and UCB Pharma.

Dr. Caroff has received research support from Bristol-Myers Squibb, Ortho-McNeil Neurologics, and Pfizer.

Diagnosis and treatment of neuroleptic malignant syndrome (NMS) are controversial because this potentially life-threatening syndrome is rare and its presentation varies. These factors make it difficult to evaluate treatments in controlled clinical trials, and data about the relative efficacy of specific interventions are scarce. It may be possible, however, to develop rational treatment guidelines using empiric clinical data.^1,2

This article examines the evidence related to 6 controversial aspects of NMS diagnosis and treatment:

• most-reliable risk factors
• NMS as a spectrum disorder
• what causes NMS
• NMS triggered by first-generation vs second-generation antipsychotics
• first-line interventions
• restarting antipsychotics after an NMS episode.

1. Are there reliable risk factors for NMS?

In small case-controlled studies, agitation, dehydration, and exhaustion were the most consistently found systemic factors believed to predispose patients taking antipsychotics to NMS (Table 1).^3-5 Catatonia and organic brain syndromes may be separate risk factors.^1,6

Preliminary studies also have implicated dopamine receptor abnormalities caused by genetic polymorphisms or effects of low serum iron.^1,7,8 Pharmacologic studies have suggested that higher doses, rapid titration, and IM injections of antipsychotics are associated with increased NMS risk.^3,5 Some studies suggest that 15% to 20% of NMS patients have a history of NMS episodes.^1,2 In addition, high-potency first-generation antipsychotics (FGAs)—especially haloperidol—are assumed to carry higher risk than low-potency drugs and second-generation antipsychotics (SGAs), although this hypothesis remains difficult to prove.^9-11

These risk factors, however, are not practical for estimating NMS risk in a given patient because they are relatively common compared with the low risk of NMS occurrence. For the vast majority of patients with psychotic symptoms, the benefits of properly indicated antipsychotic pharmacotherapy will outweigh the risks.

Table 1

Systemic

Agitation

Dehydration

Exhaustion

Low serum iron concentrations (normal: 60 to 170 mcg/dL)

Diagnoses

History of NMS

Catatonia

Organic brain syndromes

Central nervous system

Dopamine receptor dysfunction

Basal ganglia dysfunction

Sympathetic nervous system dysfunction

Pharmacologic treatment*

Intramuscular or intravenous injections

High-potency dopamine antagonists

Rapid dose titration

High doses

FGAs compared with SGAs (?)

*For individual patients, these common risk factors must be weighted again the benefits of antipsychotic therapy FGAs: first-generation antipsychotics; SGAs:second-generation antipsychotics; NMS: neuroleptic malignant syndromeSource: References 1-5

2. Is NMS related to parkinsonism, catatonia, or malignant hyperthermia?

Parkinsonsim. Some researchers have described NMS as an extreme parkinsonian crisis resulting from overwhelming blockade of dopamine pathways in the brain.^1,2,12 In this view, NMS resembles the parkinsonian-hyperthermia syndrome that can occur in Parkinson's disease patients following abrupt discontinuation or loss of efficacy of dopaminergic therapy, which can be treated by reinstituting dopaminergic agents.¹³ Evidence to support this view includes:

• Parkinsonian signs are a cardinal feature of NMS.
• Withdrawal of dopamine agonists precipitates the syndrome.
• All triggering drugs are dopamine receptor antagonists.
• Risks of NMS correlates with drugs' dopamine receptor affinity.
• Dopaminergic agonists may be an effective treatment.
• Lesions in dopaminergic pathways produce a similar syndrome.
• Patients with NMS have demonstrated low cerebrospinal fluid concentrations of the dopamine metabolite homovanillic acid.¹⁴

Catatonia. Fink et al¹⁵ and others^16-18 have persuasively argued that NMS represents a form of drug-induced malignant catatonia. Evidence supporting this includes:

• The 2 disorders share neuropsychiatric symptoms.
• Catalonic signs are common in NMS.¹⁹
• Malignant catatonia and NMS share physiologic and labratory signs.²⁰
• Reintroduction of antipsychotics can acutely worsen both conditions.
• Benzodiazepines and electroconvulsive therapy (ECT) are effective treatments for both disorders.^15-18

Lee²¹ examined the relationship between catatonic features and treatment response of 14 NMS patients. Most patients with catatonic symptoms responded to benzodiazepines, whereas none of those did who had an extrapyramidal-hyperthermic presentation without catatonia. Lee concluded that NMS is heterogeneous and may occur in catatonic and noncatatonic forms that differ in treatment response.

Document

file8nCZdF

Malignant hyperthermia. Some clinicians have compared NMS with malignant hyperthermia caused by inhalational anesthetics and succinylcholine.^1,2 Evidence includes

• similar clinical signs of rigidity, hyperthermia, and hypermetabolism
• similar psychologic and labratory signs, such as rhabdomyolysis
• hyperthermia in both responding to dantrolene.

Although the 2 are similar in presentation, malignant hyperthermia occurs intraoperatively and reflects a pharmacogenetic disorder of calcium regulation in skeletal muscle. Additionally, rigidity in malignant hyperthermia does not respond to peripheral-acting muscle relaxants.^1,22 Evidence suggests that patients who have previously experienced an NMS episodes are not at risk for malignant hyperthermia.²²

3. What is the pathophysiology of NMS?

NMS pathophysiology is complex and likely involves interplay between multiple central and systemic pathways and neurotransmitters. As described above, compelling evidence suggests that dopamine blockade plays a central role.¹²

Dopamine blockade in the hypothalamus is believed to contribute to thermoregulatory failure, and blockade in the nigrostriatal system likely contributes to muscle rigidity and hypermetabolism. The loss of dopaminergic input to the anterior cingulate-medial orbitofrontal circuit and the lateral orbitofrontal circuit likely con-tributes to the mental status changes and catatonic features seen in NMS.¹²

 

Some researchers have proposed competing or complementary hypotheses, however. For example, Gurrera²³ proposed that patients who are prone to developing NMS have a vulnerability to a hyperactive and dysregulated sympathetic nervous system, and this trait—together with dopamine system disruption induced by dopamine-blocking agents—produces NMS. Other investigators have implicated serotonin, norepinephrine, gamma-aminobutyric acid and glutaminergic mechanisms.^1,12,24,25

4. Are FGAs or SGAs more likely to cause NMS?

NMS is assumed to occur less frequently in patients treated with SGAs than in those receiving FGAs, although this hypothesisis unproven. Isolated reports of NMS have been associated with nearly every SGA.^9-11 It is difficult to prove FGA vs SGA liabilities because:

• NMS is rare.
• Dosing practices may be more conser-vative now than in the past.
• Most clinicians are aware of the earlysigns of NMS.

In an epidemiological study of a large database, Stubner et al²⁶ found that patients receiving SGAs had a lower risk of NMS than those treated with haloperidol.²⁶ In this study, the overall rate of NMS was 0.02%.

NMS hotline data. We recently examined which medication classes were implicated in 111 NMS cases reported to the Neuroleptic Malignant Syndrome Information Service hotline (1-888-NMS-TEMP) between 1997 and 2006 (Figure). We included only cases of definite or probable NMS (as diagnosed by hotline consultants) in which a single antipsychotic was administered. Slightly more cases were attributed to FGAs (51%) than SGAs (45%). The remaining cases were attributed to neuroleptics used in medical settings (such as promethazineor prochlorperazine). Because they are now prescribed less often, FGAs accounted for a disproportionate number of NMS cases reported to the hotline. Haloperidol accounted for the majority of FGA cases and 44% of all cases. If we had excluded haloperidol and compared the NMS risk of SGAs to only intermediate- or low-potency FGAs, the relative advantage of SGAs would have been lost. On the other hand, it is clear that SGAs still carry a risk for NMS. Analyses suggest that the SGA-associated classic features of NMS—fever, muscle rigidity, and autonomic and mental status changes—are retained in patients receiving SGAs, although some may not develop the severe rigidity and extreme temperatures common in patients receiving FGAs.^9-11 The milder clinical characteristics associated with SGAs may reflect more conservative prescribing patterns or increased awareness and earlier recognition of NMS, which would prevent fulminant presentations.

5. What is the evidence for specific NMS treatments?

NMS is rare, its presentation varies, and its progression is unpredictable. These factors make it difficult to evaluate treatments in controlled clinical trials, and data about the relative efficacy of specific interventions are scarce.

Even so, the notion that NMS represents an extreme variant of drug-induced parkinsonism or catatonia suggests that specific NMS treatments could be based on symptom severity or stage of presentation. We propose a treatment guideline basedon theoretical mechanisms and anecdotal data (Algorithm).^2,27-29

Support. After immediate withdrawal of the offending medication, supportive therapy is the cornerstone of NMS treatment.^1,2,27

For patients presenting with mild signs and symptoms, supportive care and careful clinical monitoring may be sufficient. Extreme hyperthermia demands volume resuscitation and cooling measures, intensive medical care, and careful monitoring for complications.

Treatment. Despite a lack of consensus on drug treatments for uncomplicated NMS, approximately 40% of patients with acute NMS receive pharmacologic treatments.²

Lorazepam, 1 to 2 mg parenterally, is a reasonable first-line therapy for NMS, especially in individuals with catatonic features.^{4,15-18,21,30,31} Some investigators recommend higher doses.¹⁵ Benzodiazepines are preferred if sedation is required in agitated NMS patients.^4,15-18

Dopaminergic agents such as bromocriptine and amantadine enhance dopaminergic transmission to reverse parkinsonian symptoms and have been reported to reduce time to recovery and halve mortality rates when used alone or in conjunction with other treatments.^13,27,32,33 Rapid discontinuation of these agents can result in rebound symptoms, although this may be true for any specific drug treatment of NMS.^1,31,32

Dantrolene uncouples excitation-contraction coupling by enhancing calcium sequestration in sarcoplasmic reticulumin skeletal muscle and has been used to treat NMS hypermetabolic symptoms. Some reviews found improvement in up to 80% of NMS patients treated with dantrolene monotherapy.^27,32-35 Compared with supportive care, time to recovery may be reduced—and mortality decreased by almost one-half—when dantrolene is used alone or in combination with other medications.

Not all case reports have shown that dantrolene, benzodiazepines, ordopaminergic agonists are effective in treating NMS.^31,36 In our opinion, only advanced NMS cases—with extreme temperature elevations, severe rigidity, and evidence of systemic hypermetabolism—benefit from dantrolene treatment.^1,2

ECT has been used successfully to reduce mortality from NMS and other catatonic-spectrum disorders. It is usually employed after supportive therapy and psychopharmacologic interventions fail.^{2,15,16,27,37} ECT for acute NMS typically consists of a series of 6 to 10 treatments with bilateral electrode placement. Daily ECT may be needed initially.¹⁵

 

6. Are antipsychotics contraindicated following an NMS episode?

The rate of NMS recurrence on retreatment with an antipsychotic has varied.38 We estimate that up to 30% of patients may be at risk of NMS recurrence when rechallenged with an antipsychotic.1 By following proper precautions (Table 2), however, you can safely treat most patients who require continued antipsychotic therapy.^1,2 When you restart treatment, a lower-potency antipsychotic from a different chemical class may be a safer option than retrying the triggering agent, according to retrospective analyses of limited available data. A patient who develops NMS on a FGA might benefit from an SGA trial, although some risk of recurrence remains.^1,10

Current Psychiatry 2007;6(8):89-95.
Drug brand names

• Amantadine • Symmetrel
• Bromocriptine • Parlodel
• Chlorpromazine • Thorazine
• Dantrolene • Dantrium
• Fluphenazine • Prolixin
• Haloperidol • Haldol
• Lorazepam • Ativan
• Loxapine • Loxitane
• Perphenazine • Trilafon
• Prochlorperazine • Compazine, Compro
• Promethazine • Phenergan
• Thioridazine • Mellaril

Disclosure

Dr. Strawn is an American Psychiatric Institute for Research and Education (APIRE)/Janssen Scholar.

Dr. Keck has received research support from or served as a consultant to Abbott Laboratories, American Diabetes Association, AstraZeneca Pharmaceuticals, Bristol-Myers Squibb, GlaxoSmithKline, Eli Lilly and Company, Janssen Pharmaceutica, National Institute of Mental Health, National Institute of Drug Abuse, Pfizer, Stanley Medical Research Institute, and UCB Pharma.

Dr. Caroff has received research support from Bristol-Myers Squibb, Ortho-McNeil Neurologics, and Pfizer.

Accurate and early diagnosis of Alzheimer’s disease (AD) is evolving, and—although not yet definitive—is no longer one of exclusion. With a careful in-office work-up and routine assessment tools, you can accurately identify >90% of patients with late-onset AD.¹

AD is by far the most common cause of dementia in older patients. To help you make the diagnosis, this state-of-the-art article discusses:

• AD’s clinical presentation and course
  
• the role of neuropsychological tests for assessing cognitive and functional status
  
• neuropsychiatric and medical findings that differentiate AD from other dementia causes
  
• indications for structural neuroimaging with CT or MRI.
  

Presentation and course

Variability. AD’s gradual onset and progression are characterized by prominent memory loss, anomia, constructional apraxia, anosognosia, and personality changes with affect deregulation, behavioral disturbance, and distorted perception.¹ Amnesia—particularly deficits in anterograde episodic memory—is the most common presentation, but the disease course is heterogeneous and may be affected by:

• patient age at onset
  
• illness severity at diagnosis
  
• comorbid medical and neuropsychiatric illnesses
  
• premorbid cerebral reserves (amount of brain damage a person can sustain before reaching a threshold for the clinical expression of dementia).^1-3
  

Box 1

Staging illness severity. AD has 3 clinical stages of cognitive dysfunction:

Mild AD. An individual or close companion may notice increased forgetfulness and word-finding difficulties, a tendency to lose or misplace things, repeated questioning, and some disorientation. Motor skills are intact.

Moderate AD. Cognitive decline continues, memory deteriorates, and self-care ability is markedly impaired. The individual may undergo personality changes, confuse time and place, have trouble communicating and recognizing family members or friends, develop agitation, begin to wander, and experience delusions and hallucinations.

Severe AD. An individual with late-stage disease has severe impairment and can be bedridden, incontinent, and unable to under-stand or speak. Full-time care is required.

Staging informs treatment. In clinical trials, patients with mild-to-moderate AD consistently show small improvements in cognitive and global function when treated with acetylcholinesterase inhibitors (AChEIs) such as donepezil, rivastigmine, and galantamine.⁴ Donepezil also is approved for use in severe AD.

Memantine is indicated for symptomatic treatment of moderate-to-severe AD. It differs in mechanism of action from the AChEIs and is thought to inhibit cytotoxic overstimulation of glutamatergic neurons.⁴ For moderately advanced AD, memantine appears to be beneficial alone or in combination with AChEIs.

Dementia assessment

Clinical assessment has low sensitivity for early-phase AD and compromised specificity in advanced stages, where all dementia subtypes are similar and comorbidities may confuse the picture. Promising surrogate biomarkers and other diagnostic tools are being developed (Box 1),^5-8 but definitive AD diagnosis still requires post-mortem histopathologic examination of the cerebral cortex.

Box 2

 

History and physical exam. Depending on the AD stage at presentation, patients might not be a reliable source of information. For a realistic and unbiased history and evaluation, assess the patient separately and obtain collateral information from reliable informants.

In typical cases, the history guides the physical/neurologic examination. Advancing age and family history are confirmed risk factors for AD; others may include:

• female gender (after age 80)
  
• cardiovascular disease (such as cerebral infarcts, hypertension, elevated cholesterol/homocysteine, smoking, and diabetes mellitus)
  
• history of head trauma, especially with loss of consciousness.
  

Assess premorbid functioning and existing medical conditions. Apraxia, aphasia, and cortical visual impairment may reflect focal signs of atypical AD; consider other neurologic signs in the context of clinical data.

Early and accurate diagnosis of AD is challenging in patients with mixed dementias, comorbid neurologic diseases, or atypical features. Patients with these presentations may require referral to an expert clinician, extensive workup, or longitudinal follow-up before the diagnosis becomes clear.

Neuropsychological testing. Most mental status tests examine orientation, attention/concentration, learning, memory, language, and constructional praxis. The Folstein Mini-Mental State Examination (MMSE)⁹ is the most widely used and well-validated mental status test. A score of 10 to 20 on the MMSE is generally considered as moderate AD, and 10 Other mental status testing options include:

• Blessed Information-Memory-Concentration (BIMC)
  
• Blessed Orientation-Memory-Concentration (BOMC)
  
• Short Test of Mental Status (STMS)
  
• Saint Louis University Mental Status (SLUMS).^11,12
  

Neuropsychological tests have limitations, but they can supplement clinical cognitive assessment by detecting milder cases and may help answer questions about a patient’s ability to drive or live alone (Box 2).^13,14

Reversible causes. If the patient is generally healthy, a core of laboratory tests is recommended in the diagnostic workup (Table 1).^6,15 Other options include:

• CSF examination for atypical presentations, such as unusually rapid symptom progression, altered consciousness, or other neurologic manifestations
  
• EEG to differentiate delirium, seizure disorders, encephalopathies, or a rapidly progressing dementia such as CreutzfeldtJakob disease.
  

Only 1% of dementia causes are considered reversible,¹⁶ but keep them in mind in the AD differential diagnosis (Table 2). Depression, vitamin B12 deficiency, medication side effects, and hypothyroidism are common comorbidities in elderly patients, particularly in those with suspected dementia. Correcting these problems might or might not reverse the dementia.

Because delirium may be the initial presentation of AD or reversible causes, re-evaluate patients for dementia after delirium clears.

Neuroimaging. Structural neuroimaging with a noncontrast CT or MRI is appropriate in the initial evaluation of patients with dementia.¹⁷ More routinely, it is used to exclude rare but potentially correctable dementia causes, such as space-occupying lesions.¹⁸ Hippocampal and entorhinal volume are measured most often in discriminating AD from non-demented aging and other dementias.¹⁹

Positron emission tomography (PET) using fluorine-18-labeled deoxyglucose (FDG) may help differentiate characteristic patterns of cerebral hypometabolism in the temporoparietal lobes in AD from fronto-temporal dementia (FTD) and other less common dementias, particularly during the earliest stages of the disease.¹⁹ Medi-care reimbursement for brain PET is limited to differentiating FTD from AD.

Table 1

Recommended lab tests for Alzheimer’s disease workup

Test	Rationale
CBC	Anemia and signs of infection
Vitamin B12	Related to reversible dementia, anemia
Folate	Related to reversible dementia, anemia
Homocysteine	More accurate than individual B12/folate tests
C-reactive protein	Ongoing inflamatory reaction
Thyroid function	Hypothyroidism (reversible dementia)
Liver function	Metabolic causes of cognitive impairment
Renal function	Uremia, metabolic causes of cognitive impairment
Electrolytes	Hypo/hypernatremia as a cognitive impairment cause
Glucose	Recurrent hypoglycemia, diabetes mellitus
Lipid panel	Vascular dementia risk factor
Baseline ECG	Cardiac abnormalities as vascular risk factors
STS (optional)	Neurosyphilis
CBC: complete blood count; ECG: electrocardiogram;
STS: serologic test for syphilis
Source: Adapted from references 6,15

Table 2

Detecting causes of potentially reversible cognitive impairment

Cause	Examples	Suggested tests
Space-occupying lesions	Subdural hematoma, benign tumors, hydrocephalus	CT/MRI without contrast
Infectious diseases	AIDS dementia complex, syphilis, Lyme disease	Serologic tests
Endocrinopathies/ metabolic/autoimmune disorders	Hypothyroidism, Cushing’s disease, uremia, hepatic encephalopathy, Wilson’s disease, recurrent hypoglycemia, chronic hypocalcemia, multiple sclerosis, disseminated SLE, sarcoidosis	Thyroid panel, renal and liver function tests, electrolytes, slit lamp test, serum ceruloplasmin
Psychiatric	Depression, alcohol dependence	Geriatric Depression Scale, assess vitamin deficiency states
Nutritional deficiencies	Vitamin B12, thiamine (Wernicke-Korsakoff syndrome), pyridoxine, niacin (pellagra)	Vitamin B12, homocysteine
Medication effects	Benzodiazepines, barbiturates, anticholinergics, opioid analgesics, antihypertensives, antiarrhythmics, antidepressants, anticonvulsants, cardiac drugs such as digitalis and derivatives (among others)	Review patients’ medications for drugs that can cause cognitive changes
Others	Autoimmune diseases, heavy metals, illicit drugs, obstructive sleep apnea	Drug screens and specific tests

Diagnostic criteria

NINCDS-ADRDA. Neuropsychological AD assessment criteria developed by the National Institute of Neurological and Communicative Disorders and Stroke and Alzheimer’s Disease and Related Disorders Association (NINCDS-ADRDA) classify AD as probable, possible, or definite:

 

Possible AD is considered when a patient has an atypical onset, presentation, or course and other secondary illnesses capable of producing dementia are not believed to be the cause.

Probable AD is diagnosed when dementia is established by clinical exam and con- firmed with cognitive testing, where ≥2 cognitive domains are progressively affected; includes gradual memory loss not caused by another systemic or brain disease, with age of onset between 40 and 90 years.

Definite AD requires histopathologic evidence of AD in addition to fulfilling criteria for probable AD.20

DSM-IV-TR. Similar but broader DSM-IVTR criteria describe an insidious progressive cognitive decline that affects recent memory and ≥1 other cognitive domain (apraxia, aphasia, agnosia, or executive functioning). This cognitive decline impairs social and occupational function, represents a change from a higher level, and is not due to other causes such as delirium.²¹

NINCDS-ADRDA and DSM-IV-TR criteria have comparable sensitivity and specificity for clinical AD diagnosis. Neither requires neuropathologic or genetic assessment (Box 3).^15,17,22-²⁴ Neuroimaging and other tests may be required to rule out other brain diseases that may cause dementia.

Other causes of dementia

Mild cognitive impairment (MCI) may represent a prodromal state for the earliest clinical manifestations of dementia. Symptoms include memory complaints but generally preserved activities of daily living.

Originally introduced to define a progressive, single-symptom amnestic syndrome, MCI has evolved into a classification of amnestic and non-amnestic MCI with single or multiple domains.²⁵ Amnestic MCI is the most specifically correlated with AD.²⁶ Neurobiologic similarities between amnestic MCI and clinically diagnosed AD include:

• neuropsychiatric symptoms, such as apathy, mood disturbance, irritability and anxiety
  
• over-representation of the APOE ε4 allele
  
• volumetric loss in the entorhinal cortex and hippocampus as measured by MRI
  
• Glucose hypometabolism in AD-typical regions as measured by FDG-PET
  
• neuronal loss in vulnerable brain regions.²⁶
  

Most patients with MCI go on to meet AD criteria within 5 to 10 years, and approximately 80% of those originally diagnosed with MCI prove to have AD at post-mortem histopathologic examination.^26,27

Dementia with Lewy bodies (DLB) is the second most common dementing disorder in late life—after Alzheimer’s dementia— and two-thirds of DLB cases overlap with AD. Core DLB clinical features include early recurrent visual hallucinations, fluctuating cognition, spontaneous parkinsonism, and sensitivity to conventional antipsychotics.^15,28

Parkinson’s disease (PD) and DLB may represent a clinicopathologic continuum, and substantial overlap exists among AD, DLB, and PD in underlying disease process and clinical presentation.^15,29 Hallucinations, depression, delusions, and delusional misidentification are seen more often in patients with DLB than AD.¹⁵

Vascular dementia (VaD) was once thought to account for 15% to 20% of dementing illnesses, but discrete VaD is now viewed as much less common. Whatever the underlying vasculopathy, vascular lesions often co-exist with other causes of dementia—usually AD (in 77% of presumed VaD cases).³⁰

Compared with AD, patients with VaD have a more subcortical dementia with difficulty retrieving words, organizing and solving complex problems, “absent-mindedness,” and psychomotor slowing with relatively preserved language skills. VaD is thought to have a more abrupt onset than AD and “stepladder” deterioration.

Frontotemporal dementia (FTD)—such as Pick’s disease—is associated with focal atrophy of the frontal and/or temporal lobes. Mean onset is age 52 to 56, and FTD is less common than AD, VaD, or DLB.

FTD often presents with gradual personality changes (with inappropriate responses or activities) or language changes (with severe naming difficulty and problems with word meaning).³¹ Features that may help differentiate FTD from AD include:

• disinhibition/apathy with personality change
  
• affect disregulation
  
• behavioral disturbance (frontal type) and expressive/receptive language changes (semantic or primary progressive aphasia) with relatively mild memory loss.^32,33
  

Unlike AD, memory usually is unaffected in early FTD, with problems largely secondary to poor concentration and relating to difficulties with working (immediate) memory.

Other neurodegenerative diseases that might present with dementia include PD, Huntington’s disease, progressive supra-nuclear palsy, corticobasal degeneration, and Creutzfeldt-Jakob disease.³³

Box 3

 

Related resources

• Morris JC. Dementia update 2005. Alzheimer Dis Assoc Disord 2005;19:100-17.
  
• Boeve BF. A review of the non-Alzheimer dementias. J Clin Psychiatry 2006;67(12)1985-2001.
  
• Medscape. Alzheimer’s disease resource center. www.medscape.com/resource/alzheimers.
  

Drug brand names

• Donepezil • Aricept
  
• Memantine • Namenda
  
• Galantamine • Razadyne
  
• Rivastigmine • Exelon
  

Disclosure

Dr. Gebretsadik reports no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Dr. Grossberg receives grant/research support from Abbott Laboratories, Bristol-Myers Squibb, Forest Laboratories, Eli Lilly and Company, Novartis, Pfizer Inc., Wyeth, Elan, Myriad, Ono Pharmaceutical, and the Alzheimer’s Disease Cooperative Study Consortium. He is a consultant to Bristol-Myers Squibb, Forest Laboratories, GlaxoSmithKline, Janssen Pharmaceutica, Novartis, AstraZeneca, Wyeth, Pfizer Inc., Takeda, and Sepracor.

Accurate and early diagnosis of Alzheimer’s disease (AD) is evolving, and—although not yet definitive—is no longer one of exclusion. With a careful in-office work-up and routine assessment tools, you can accurately identify >90% of patients with late-onset AD.¹

AD is by far the most common cause of dementia in older patients. To help you make the diagnosis, this state-of-the-art article discusses:

• AD’s clinical presentation and course
  
• the role of neuropsychological tests for assessing cognitive and functional status
  
• neuropsychiatric and medical findings that differentiate AD from other dementia causes
  
• indications for structural neuroimaging with CT or MRI.
  

Presentation and course

Variability. AD’s gradual onset and progression are characterized by prominent memory loss, anomia, constructional apraxia, anosognosia, and personality changes with affect deregulation, behavioral disturbance, and distorted perception.¹ Amnesia—particularly deficits in anterograde episodic memory—is the most common presentation, but the disease course is heterogeneous and may be affected by:

• patient age at onset
  
• illness severity at diagnosis
  
• comorbid medical and neuropsychiatric illnesses
  
• premorbid cerebral reserves (amount of brain damage a person can sustain before reaching a threshold for the clinical expression of dementia).^1-3
  

Box 1

Staging illness severity. AD has 3 clinical stages of cognitive dysfunction:

Mild AD. An individual or close companion may notice increased forgetfulness and word-finding difficulties, a tendency to lose or misplace things, repeated questioning, and some disorientation. Motor skills are intact.

Moderate AD. Cognitive decline continues, memory deteriorates, and self-care ability is markedly impaired. The individual may undergo personality changes, confuse time and place, have trouble communicating and recognizing family members or friends, develop agitation, begin to wander, and experience delusions and hallucinations.

Severe AD. An individual with late-stage disease has severe impairment and can be bedridden, incontinent, and unable to under-stand or speak. Full-time care is required.

Staging informs treatment. In clinical trials, patients with mild-to-moderate AD consistently show small improvements in cognitive and global function when treated with acetylcholinesterase inhibitors (AChEIs) such as donepezil, rivastigmine, and galantamine.⁴ Donepezil also is approved for use in severe AD.

Memantine is indicated for symptomatic treatment of moderate-to-severe AD. It differs in mechanism of action from the AChEIs and is thought to inhibit cytotoxic overstimulation of glutamatergic neurons.⁴ For moderately advanced AD, memantine appears to be beneficial alone or in combination with AChEIs.

Dementia assessment

Clinical assessment has low sensitivity for early-phase AD and compromised specificity in advanced stages, where all dementia subtypes are similar and comorbidities may confuse the picture. Promising surrogate biomarkers and other diagnostic tools are being developed (Box 1),^5-8 but definitive AD diagnosis still requires post-mortem histopathologic examination of the cerebral cortex.

Box 2

 

History and physical exam. Depending on the AD stage at presentation, patients might not be a reliable source of information. For a realistic and unbiased history and evaluation, assess the patient separately and obtain collateral information from reliable informants.

In typical cases, the history guides the physical/neurologic examination. Advancing age and family history are confirmed risk factors for AD; others may include:

• female gender (after age 80)
  
• cardiovascular disease (such as cerebral infarcts, hypertension, elevated cholesterol/homocysteine, smoking, and diabetes mellitus)
  
• history of head trauma, especially with loss of consciousness.
  

Assess premorbid functioning and existing medical conditions. Apraxia, aphasia, and cortical visual impairment may reflect focal signs of atypical AD; consider other neurologic signs in the context of clinical data.

Early and accurate diagnosis of AD is challenging in patients with mixed dementias, comorbid neurologic diseases, or atypical features. Patients with these presentations may require referral to an expert clinician, extensive workup, or longitudinal follow-up before the diagnosis becomes clear.

Neuropsychological testing. Most mental status tests examine orientation, attention/concentration, learning, memory, language, and constructional praxis. The Folstein Mini-Mental State Examination (MMSE)⁹ is the most widely used and well-validated mental status test. A score of 10 to 20 on the MMSE is generally considered as moderate AD, and 10 Other mental status testing options include:

• Blessed Information-Memory-Concentration (BIMC)
  
• Blessed Orientation-Memory-Concentration (BOMC)
  
• Short Test of Mental Status (STMS)
  
• Saint Louis University Mental Status (SLUMS).^11,12
  

Neuropsychological tests have limitations, but they can supplement clinical cognitive assessment by detecting milder cases and may help answer questions about a patient’s ability to drive or live alone (Box 2).^13,14

Reversible causes. If the patient is generally healthy, a core of laboratory tests is recommended in the diagnostic workup (Table 1).^6,15 Other options include:

• CSF examination for atypical presentations, such as unusually rapid symptom progression, altered consciousness, or other neurologic manifestations
  
• EEG to differentiate delirium, seizure disorders, encephalopathies, or a rapidly progressing dementia such as CreutzfeldtJakob disease.
  

Only 1% of dementia causes are considered reversible,¹⁶ but keep them in mind in the AD differential diagnosis (Table 2). Depression, vitamin B12 deficiency, medication side effects, and hypothyroidism are common comorbidities in elderly patients, particularly in those with suspected dementia. Correcting these problems might or might not reverse the dementia.

Because delirium may be the initial presentation of AD or reversible causes, re-evaluate patients for dementia after delirium clears.

Neuroimaging. Structural neuroimaging with a noncontrast CT or MRI is appropriate in the initial evaluation of patients with dementia.¹⁷ More routinely, it is used to exclude rare but potentially correctable dementia causes, such as space-occupying lesions.¹⁸ Hippocampal and entorhinal volume are measured most often in discriminating AD from non-demented aging and other dementias.¹⁹

Positron emission tomography (PET) using fluorine-18-labeled deoxyglucose (FDG) may help differentiate characteristic patterns of cerebral hypometabolism in the temporoparietal lobes in AD from fronto-temporal dementia (FTD) and other less common dementias, particularly during the earliest stages of the disease.¹⁹ Medi-care reimbursement for brain PET is limited to differentiating FTD from AD.

Table 1

Recommended lab tests for Alzheimer’s disease workup

Test	Rationale
CBC	Anemia and signs of infection
Vitamin B12	Related to reversible dementia, anemia
Folate	Related to reversible dementia, anemia
Homocysteine	More accurate than individual B12/folate tests
C-reactive protein	Ongoing inflamatory reaction
Thyroid function	Hypothyroidism (reversible dementia)
Liver function	Metabolic causes of cognitive impairment
Renal function	Uremia, metabolic causes of cognitive impairment
Electrolytes	Hypo/hypernatremia as a cognitive impairment cause
Glucose	Recurrent hypoglycemia, diabetes mellitus
Lipid panel	Vascular dementia risk factor
Baseline ECG	Cardiac abnormalities as vascular risk factors
STS (optional)	Neurosyphilis
CBC: complete blood count; ECG: electrocardiogram;
STS: serologic test for syphilis
Source: Adapted from references 6,15

Table 2

Detecting causes of potentially reversible cognitive impairment

Cause	Examples	Suggested tests
Space-occupying lesions	Subdural hematoma, benign tumors, hydrocephalus	CT/MRI without contrast
Infectious diseases	AIDS dementia complex, syphilis, Lyme disease	Serologic tests
Endocrinopathies/ metabolic/autoimmune disorders	Hypothyroidism, Cushing’s disease, uremia, hepatic encephalopathy, Wilson’s disease, recurrent hypoglycemia, chronic hypocalcemia, multiple sclerosis, disseminated SLE, sarcoidosis	Thyroid panel, renal and liver function tests, electrolytes, slit lamp test, serum ceruloplasmin
Psychiatric	Depression, alcohol dependence	Geriatric Depression Scale, assess vitamin deficiency states
Nutritional deficiencies	Vitamin B12, thiamine (Wernicke-Korsakoff syndrome), pyridoxine, niacin (pellagra)	Vitamin B12, homocysteine
Medication effects	Benzodiazepines, barbiturates, anticholinergics, opioid analgesics, antihypertensives, antiarrhythmics, antidepressants, anticonvulsants, cardiac drugs such as digitalis and derivatives (among others)	Review patients’ medications for drugs that can cause cognitive changes
Others	Autoimmune diseases, heavy metals, illicit drugs, obstructive sleep apnea	Drug screens and specific tests

Diagnostic criteria

NINCDS-ADRDA. Neuropsychological AD assessment criteria developed by the National Institute of Neurological and Communicative Disorders and Stroke and Alzheimer’s Disease and Related Disorders Association (NINCDS-ADRDA) classify AD as probable, possible, or definite:

 

Possible AD is considered when a patient has an atypical onset, presentation, or course and other secondary illnesses capable of producing dementia are not believed to be the cause.

Probable AD is diagnosed when dementia is established by clinical exam and con- firmed with cognitive testing, where ≥2 cognitive domains are progressively affected; includes gradual memory loss not caused by another systemic or brain disease, with age of onset between 40 and 90 years.

Definite AD requires histopathologic evidence of AD in addition to fulfilling criteria for probable AD.20

DSM-IV-TR. Similar but broader DSM-IVTR criteria describe an insidious progressive cognitive decline that affects recent memory and ≥1 other cognitive domain (apraxia, aphasia, agnosia, or executive functioning). This cognitive decline impairs social and occupational function, represents a change from a higher level, and is not due to other causes such as delirium.²¹

NINCDS-ADRDA and DSM-IV-TR criteria have comparable sensitivity and specificity for clinical AD diagnosis. Neither requires neuropathologic or genetic assessment (Box 3).^15,17,22-²⁴ Neuroimaging and other tests may be required to rule out other brain diseases that may cause dementia.

Other causes of dementia

Mild cognitive impairment (MCI) may represent a prodromal state for the earliest clinical manifestations of dementia. Symptoms include memory complaints but generally preserved activities of daily living.

Originally introduced to define a progressive, single-symptom amnestic syndrome, MCI has evolved into a classification of amnestic and non-amnestic MCI with single or multiple domains.²⁵ Amnestic MCI is the most specifically correlated with AD.²⁶ Neurobiologic similarities between amnestic MCI and clinically diagnosed AD include:

• neuropsychiatric symptoms, such as apathy, mood disturbance, irritability and anxiety
  
• over-representation of the APOE ε4 allele
  
• volumetric loss in the entorhinal cortex and hippocampus as measured by MRI
  
• Glucose hypometabolism in AD-typical regions as measured by FDG-PET
  
• neuronal loss in vulnerable brain regions.²⁶
  

Most patients with MCI go on to meet AD criteria within 5 to 10 years, and approximately 80% of those originally diagnosed with MCI prove to have AD at post-mortem histopathologic examination.^26,27

Dementia with Lewy bodies (DLB) is the second most common dementing disorder in late life—after Alzheimer’s dementia— and two-thirds of DLB cases overlap with AD. Core DLB clinical features include early recurrent visual hallucinations, fluctuating cognition, spontaneous parkinsonism, and sensitivity to conventional antipsychotics.^15,28

Parkinson’s disease (PD) and DLB may represent a clinicopathologic continuum, and substantial overlap exists among AD, DLB, and PD in underlying disease process and clinical presentation.^15,29 Hallucinations, depression, delusions, and delusional misidentification are seen more often in patients with DLB than AD.¹⁵

Vascular dementia (VaD) was once thought to account for 15% to 20% of dementing illnesses, but discrete VaD is now viewed as much less common. Whatever the underlying vasculopathy, vascular lesions often co-exist with other causes of dementia—usually AD (in 77% of presumed VaD cases).³⁰

Compared with AD, patients with VaD have a more subcortical dementia with difficulty retrieving words, organizing and solving complex problems, “absent-mindedness,” and psychomotor slowing with relatively preserved language skills. VaD is thought to have a more abrupt onset than AD and “stepladder” deterioration.

Frontotemporal dementia (FTD)—such as Pick’s disease—is associated with focal atrophy of the frontal and/or temporal lobes. Mean onset is age 52 to 56, and FTD is less common than AD, VaD, or DLB.

FTD often presents with gradual personality changes (with inappropriate responses or activities) or language changes (with severe naming difficulty and problems with word meaning).³¹ Features that may help differentiate FTD from AD include:

• disinhibition/apathy with personality change
  
• affect disregulation
  
• behavioral disturbance (frontal type) and expressive/receptive language changes (semantic or primary progressive aphasia) with relatively mild memory loss.^32,33
  

Unlike AD, memory usually is unaffected in early FTD, with problems largely secondary to poor concentration and relating to difficulties with working (immediate) memory.

Other neurodegenerative diseases that might present with dementia include PD, Huntington’s disease, progressive supra-nuclear palsy, corticobasal degeneration, and Creutzfeldt-Jakob disease.³³

Box 3

 

Related resources

• Morris JC. Dementia update 2005. Alzheimer Dis Assoc Disord 2005;19:100-17.
  
• Boeve BF. A review of the non-Alzheimer dementias. J Clin Psychiatry 2006;67(12)1985-2001.
  
• Medscape. Alzheimer’s disease resource center. www.medscape.com/resource/alzheimers.
  

Drug brand names

• Donepezil • Aricept
  
• Memantine • Namenda
  
• Galantamine • Razadyne
  
• Rivastigmine • Exelon
  

Disclosure

Dr. Gebretsadik reports no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

Dr. Grossberg receives grant/research support from Abbott Laboratories, Bristol-Myers Squibb, Forest Laboratories, Eli Lilly and Company, Novartis, Pfizer Inc., Wyeth, Elan, Myriad, Ono Pharmaceutical, and the Alzheimer’s Disease Cooperative Study Consortium. He is a consultant to Bristol-Myers Squibb, Forest Laboratories, GlaxoSmithKline, Janssen Pharmaceutica, Novartis, AstraZeneca, Wyeth, Pfizer Inc., Takeda, and Sepracor.