Evidence-Based Reviews

Autism and related pervasive developmental disorders (PDD) are increasingly being identified in children—and even in some adolescents and adults. As a result, psychiatry now recognizes that these debilitating disorders are more common than was once believed, with a prevalence as high as 1 in 250.

An accurate diagnosis can help families take advantage of the variety of treatments being offered and investigated for affected individuals (Box).¹ As psychiatrists who primarily see patients with autism and PDD, we recommend a three-step diagnostic approach that includes:

• a comprehensive initial assessment to rule out medical or neurologic illnesses that can mimic or are associated with autism
• differentiating PDDs from other psychiatric disorders with similar symptoms
• distinguishing among the five PDD subtypes described in DSM-IV.²

Step 1: Comprehensive initial assessment

Assessment for possible PDD begins with a comprehensive history and examination. Most patients will be assessed in childhood, but milder symptoms of autism or Asperger’s disorder may go unrecognized initially and not be brought to a clinician’s attention until adolescence or even adulthood.

As PDDs are childhood-onset disorders, the logical approach emphasizes the developmental course and onset of symptoms. By definition, children with autism show evidence of the disorder by age 3. However, the diagnosis can often be made as early as age 18 to 24 months, when typically developing children exhibit a number of social and communicative milestones that are absent in autism.

History A thorough description of the mother’s pregnancy, labor, and delivery (if known) can help you determine whether intrauterine or perinatal events could be related to the patient’s presenting problem. These include infections and exposure to exogenous substances (e.g., alcohol) during the pregnancy, as well as complications during pregnancy and delivery (e.g., maternal bleeding, neonatal hypoxia).

A complete description of the child’s development including major milestones (e.g., sitting without support, walking, first words) can help distinguish among certain diagnoses and estimate the extent of developmental delay. Ask the parents what first concerned them about their child’s development, as children with autism most often present with delays in social or language milestones. Any developmental regression in acquired skills may implicate other neurologic processes and help distinguish among the subtypes of PDD.

Symptoms Review the symptoms of autism at length in all patients in whom you suspect a PDD. It is important to assess these symptoms in the context of the child’s overall developmental level. For example, a child with known mental retardation should be compared with peers of similar age and cognitive impairment.

Box

Approximately 75% of persons with autism are diagnosed with mental retardation. A review of intellectual abilities and level of adaptive functioning can suggest the degree of mental retardation. A detailed family history is important because autism and other syndromes associated with mental retardation have varying degrees of heritability.

A thorough medical history, review of systems, and physical exam (with focus on the neurologic exam) can suggest the presence of medical conditions that could mimic or be associated with autism. The symptoms of autism are traditionally divided into three domains: social, communication, and repetitive behavior/narrow interests (Table 1):

• Social impairment includes problems with nonverbal behaviors such as eye contact, facial expressions, and “body language;” failure to develop peer relationships; lack of spontaneous seeking to share enjoyment, interests, or achievements with other people; and lack of social or emotional reciprocity.
• Communication impairment includes language delay, decreased communication with others, conversational impairment, echolalia, and lack of imaginative or social imitative play.
• Impairments in behavior, interests, and activity take the form of all-encompassing preoccupations, “need for sameness” and compulsive rituals, motor stereotypies, and preoccupation with parts of objects.

After the preliminary history and exam, you should have an initial impression as to whether a PDD may be present. Conditions that could mimic or co-exist with autism will then need to be evaluated (Table 2).

Hearing and vision testing Every child presenting with a language or cognitive delay should have an adequate hearing assessment, including an audiogram at the very least. If results are equivocal, then brainstem auditory-evoked responses are indicated to establish the auditory system’s integrity. Vision screening should be completed, and the child should be referred to an ophthalmologist if a problem is suspected.

IQ testing A child with cognitive delays, learning problems, or suspected PDD should be referred to an experienced psychologist for intelligence testing. Public school systems often provide this service. Intelligence testing can document mental retardation and offer important information about the child’s strengths and weaknesses in learning.

Speech and language assessment A speech and language pathologist should evaluate children with PDD and/or language problems for articulation, prosody, receptive and expressive language, and pragmatics.

Lab and genetic tests Laboratory investigation in a child with cognitive delays should include routine blood chemistries, CBC, thyroid function tests, and lead level. Screen for fragile X syndrome, as its symptoms overlap those of autism, and for disorders of amino acid and organic acid metabolism. Finally, consider a chromosome karyotype, especially in patients with dysmorphic features on physical exam.

Table 1

THE THREE DOMAINS OF IMPAIRMENT IN AUTISM

EEG Obtain a sleep-deprived electroencephalogram (EEG) in children with a history of significant language regression, episodic symptoms, or other indicators of possible seizures. Ideally, the EEG should include monitoring during sleep to help rule out acquired epileptic aphasia (Landau-Kleffner syndrome), a rare disorder associated with late-onset language regression. MRI of the brain is not routine but should be considered if indicated by the history or neurologic exam.

Consultations Consider consulting colleagues in neurology and medical genetics, especially when patients present with definite neurologic signs and symptoms or obvious dysmorphic features. These medical specialists are trained to screen for complex and rare syndromes that are sometimes associated with features of PDD.

Step 2. Is it PDD or another psychiatric disorder?

Psychiatric disorders that can be mistaken for PDD are listed in Table 3. The central feature of all PDDs is disturbance in social relatedness, and a diagnosis of PDD requires a history of significant social impairment.

Problems with social reciprocity in PDD are qualitatively different from the social impairment seen in other psychiatric disorders. For example, a child with attention-deficit/hyperactivity disorder (ADHD) and oppositional defiant disorder (ODD) may have few friends because of a tendency to act impulsively and get into frequent conflicts with others. These social difficulties would not be considered indicative of PDD, as they typify those seen in children with ADHD and ODD.

Mental retardation Although mental retardation occurs in approximately 75% of persons with autism, most patients with mental retardation do not have autism. In assessing an individual with mental retardation for symptoms of PDD, consider the overall developmental level. It is not uncommon for individuals with mental retardation to have mild social problems, a history of language delay, and even motor stereotypies. These symptoms are considered indicative of PDD only when they are more severe than would be expected for the patient’s developmental level.

Table 2

TESTING OPTIONS FOR PATIENTS WITH PDD

Table 3

DIFFERENTIAL DIAGNOSIS OF PDD

Other symptoms that are relatively specific to autism and PDD include lack of appropriate eye-to-eye gaze, abnormal speech prosody, echolalia, pronominal reversal, and narrow and circumscribed interests. The presence of these symptoms in excess should increase your suspicion of comorbid PDD.

RAD Reactive attachment disorder presents with abnormal social relatedness that can sometimes be confused with milder PDDs, especially in patients with comorbid mental retardation. In RAD, however, a history of severe neglect or abuse is thought to have caused the abnormal social relatedness. Placing the child in a caring and secure environment should improve many of the social deficits.

Language disorders are distinguished from PDDs by the absence of marked social impairment and lack of restricted interests and repetitive behaviors. In addition, children with primary language disorders often have intact nonverbal communication skills and make other attempts to communicate (e.g., through gesture, eye contact).

Stereotypic movement disorder can be seen in individuals with and without comorbid mental retardation. It is not diagnosed in the presence of autism, as these movements are thought to be part of the underlying disorder. The lack of social and communication impairments distinguishes stereotypic movement disorder from PDD.

ADHD Many children with autism and other PDDs have interfering symptoms of inattention, hyperactivity, and impulsivity. We usually do not give them an additional diagnosis of ADHD, as these symptoms are common in PDD. The pathophysiology of these symptoms may be different in ADHD and PDD, as evidenced by the frequent report of adverse effects following stimulant treatment of children with autism.²

Social phobia In higher functioning individuals with PDD, excessive social anxiety can sometimes be confused with social phobia. In social phobia, however, individuals usually do not exhibit marked problems with social relatedness and are able to interact normally with persons they know well and in some situations.

OCD Obsessive-compulsive disorder can occur in individuals with PDD but must be distinguished from the abnormal preoccupations and ritualistic behavior characteristic of autism. In autism, these activities often differ in quality from obsessions and compulsions.³ Furthermore, they usually are not associated with distress, and repetitive behaviors are not linked to a specific obsession.

Selective mutism is usually easy to distinguish from PDD because the affected child is typically able to talk in certain environments, such as at home. Also, the onset of selective mutism follows a period of normal social and communicative development.

Schizophrenia Autism was historically conceptualized as a type of childhood schizophrenia but is now thought to be distinct from the psychotic disorders. Schizophrenia with onset in childhood is much more rare than autism. Its onset usually occurs after several years of normal development, though some children with schizophrenia may have symptoms that resemble PDD early in their illness.⁴ Autistic persons may at times present with symptoms of a thought disorder. A diagnosis of schizophrenia usually is not made without evidence of prominent delusions and hallucinations.

Personality disorders PDDs are sometimes difficult to distinguish from personality disorders with similar features (e.g., schizotypal personality, schizoid personality). The social impairment in autistic and Asperger’s disorders is generally of earlier onset and greater severity than that seen in personality disorders. Those with personality disorders also typically lack stereotyped language or repetitive behaviors that are common in PDDs.

Table 4

CHARACTERISTIC FEATURES OF DSM-IV SUBTYPES OF PERVASIVE DEVELOPMENTAL DISORDERS

Social awkwardness Finally, some of the PDDs that allow higher functioning (e.g., Asperger’s disorder and PDD not otherwise specified [NOS]) need to be distinguished from normal social awkwardness that can be common, especially in adolescence. The social impairment in PDD is marked and interferes with normal functioning and development.

Step 3. Which PDD is it?

DSM-IV describes five subtypes of PDD (autistic disorder, Asperger’s disorder, Rett’s disorder, childhood disintegrative disorder, and PDD NOS) that have in common problems with reciprocal social interaction (Table 4). For psychiatrists making the diagnosis, it is probably most difficult to differentiate the two most common types: autistic and Asperger’s disorders.

Autistic disorder is the prototypical PDD that is associated with abnormalities in reciprocal social interaction, qualitative impairments in communication, and narrow interests and repetitive behaviors (Table 1). By definition, symptoms of the disorder manifest by age 3.

Asperger’s disorder has several features that distinguish it from autistic disorder:

• Children with Asperger’s disorder do not have language delays. By definition, a child who has not developed single words by age 2 cannot be diagnosed with Asperger’s disorder.
• Early cognitive development in Asperger’s disorder is normal. Children with Asperger’s disorder are much more likely to have normal or above-average intellectual functioning than children with autistic disorder.
• Circumscribed interests and intense preoccupations are more common than motor stereotypies.
• Affected children may show verbal abilities that greatly exceed their visual-spatial skills. This may be apparent on individualized intelligence testing (i.e., verbal IQ > performance IQ) and clinically in the form of good language abilities but lagging fine-motor development (e.g., clumsiness).

Rett’s disorder occurs almost exclusively in girls, whereas autistic disorder is more common in boys. Following a brief period of normal development, affected girls experience deceleration of head growth, loss of previously acquired purposeful hand skills with subsequent development of stereotyped hand-wringing movements, loss of social engagement, onset of trunk and gait ataxia, and severe language and cognitive impairment. Genetic testing for mutations at MECP2 will be positive in most patients having all features of the classic phenotype.⁵

Childhood disintegrative disorder is thought to be more rare than autistic disorder. Following at least 2 years of normal development, affected children lose previously acquired skills. These can include play skills, language, social skills, bowel or bladder control, and motor skills. The children show impairments in social interaction, communication, and behavior of the type common to autistic disorder and often have severe mental retardation. The late onset of severe regression in development often prompts extensive neurologic evaluation, but a specific etiology is usually not found. The disorder is not diagnosed if full diagnostic criteria for autistic disorder are met (including onset of symptoms before age 3).

PDD NOS is diagnosed in many patients who are determined to have a significant impairment in social relatedness but do not meet full criteria for a specific PDD. Recent epidemiologic studies suggest that PDD NOS may be more common than autistic disorder.⁶ It may also be an appropriate designation for children with other proposed diagnostic constructs, such as nonverbal learning disabilities⁷ and multiple complex developmental disorders.⁸

Unfortunately, children diagnosed with PDD NOS instead of the better-recognized term “autism” may be denied appropriate social and financial services. When you inform others about the diagnosis of PDD NOS, it is important to emphasize to parents, schools, and funding agencies that PDD NOS is related to autism and should be considered one of the “autism spectrum disorders.” Children with PDDs or autism spectrum disorders will often benefit from similar treatment and educational interventions. In addition, their needs are equivalent for adequate insurance coverage and funding for specialized treatments.

Related resources

• Cohen DJ, Volkmar FR (eds). Handbook of autism and pervasive developmental disorders (2nd ed). New York: Wiley, 1997.
• National Institute of Mental Health. Autism booklet. http://www.nimh.nih.gov/publicat/autism.cfm
• Autism Society of America http://www.autism-society.org Online Asperger Syndrome Information and Support (OASIS) http://www.udel.edu/bkirby/asperger

Autism and related pervasive developmental disorders (PDD) are increasingly being identified in children—and even in some adolescents and adults. As a result, psychiatry now recognizes that these debilitating disorders are more common than was once believed, with a prevalence as high as 1 in 250.

An accurate diagnosis can help families take advantage of the variety of treatments being offered and investigated for affected individuals (Box).¹ As psychiatrists who primarily see patients with autism and PDD, we recommend a three-step diagnostic approach that includes:

• a comprehensive initial assessment to rule out medical or neurologic illnesses that can mimic or are associated with autism
• differentiating PDDs from other psychiatric disorders with similar symptoms
• distinguishing among the five PDD subtypes described in DSM-IV.²

Step 1: Comprehensive initial assessment

Assessment for possible PDD begins with a comprehensive history and examination. Most patients will be assessed in childhood, but milder symptoms of autism or Asperger’s disorder may go unrecognized initially and not be brought to a clinician’s attention until adolescence or even adulthood.

As PDDs are childhood-onset disorders, the logical approach emphasizes the developmental course and onset of symptoms. By definition, children with autism show evidence of the disorder by age 3. However, the diagnosis can often be made as early as age 18 to 24 months, when typically developing children exhibit a number of social and communicative milestones that are absent in autism.

History A thorough description of the mother’s pregnancy, labor, and delivery (if known) can help you determine whether intrauterine or perinatal events could be related to the patient’s presenting problem. These include infections and exposure to exogenous substances (e.g., alcohol) during the pregnancy, as well as complications during pregnancy and delivery (e.g., maternal bleeding, neonatal hypoxia).

A complete description of the child’s development including major milestones (e.g., sitting without support, walking, first words) can help distinguish among certain diagnoses and estimate the extent of developmental delay. Ask the parents what first concerned them about their child’s development, as children with autism most often present with delays in social or language milestones. Any developmental regression in acquired skills may implicate other neurologic processes and help distinguish among the subtypes of PDD.

Symptoms Review the symptoms of autism at length in all patients in whom you suspect a PDD. It is important to assess these symptoms in the context of the child’s overall developmental level. For example, a child with known mental retardation should be compared with peers of similar age and cognitive impairment.

Box

Approximately 75% of persons with autism are diagnosed with mental retardation. A review of intellectual abilities and level of adaptive functioning can suggest the degree of mental retardation. A detailed family history is important because autism and other syndromes associated with mental retardation have varying degrees of heritability.

A thorough medical history, review of systems, and physical exam (with focus on the neurologic exam) can suggest the presence of medical conditions that could mimic or be associated with autism. The symptoms of autism are traditionally divided into three domains: social, communication, and repetitive behavior/narrow interests (Table 1):

• Social impairment includes problems with nonverbal behaviors such as eye contact, facial expressions, and “body language;” failure to develop peer relationships; lack of spontaneous seeking to share enjoyment, interests, or achievements with other people; and lack of social or emotional reciprocity.
• Communication impairment includes language delay, decreased communication with others, conversational impairment, echolalia, and lack of imaginative or social imitative play.
• Impairments in behavior, interests, and activity take the form of all-encompassing preoccupations, “need for sameness” and compulsive rituals, motor stereotypies, and preoccupation with parts of objects.

After the preliminary history and exam, you should have an initial impression as to whether a PDD may be present. Conditions that could mimic or co-exist with autism will then need to be evaluated (Table 2).

Hearing and vision testing Every child presenting with a language or cognitive delay should have an adequate hearing assessment, including an audiogram at the very least. If results are equivocal, then brainstem auditory-evoked responses are indicated to establish the auditory system’s integrity. Vision screening should be completed, and the child should be referred to an ophthalmologist if a problem is suspected.

IQ testing A child with cognitive delays, learning problems, or suspected PDD should be referred to an experienced psychologist for intelligence testing. Public school systems often provide this service. Intelligence testing can document mental retardation and offer important information about the child’s strengths and weaknesses in learning.

Speech and language assessment A speech and language pathologist should evaluate children with PDD and/or language problems for articulation, prosody, receptive and expressive language, and pragmatics.

Lab and genetic tests Laboratory investigation in a child with cognitive delays should include routine blood chemistries, CBC, thyroid function tests, and lead level. Screen for fragile X syndrome, as its symptoms overlap those of autism, and for disorders of amino acid and organic acid metabolism. Finally, consider a chromosome karyotype, especially in patients with dysmorphic features on physical exam.

Table 1

THE THREE DOMAINS OF IMPAIRMENT IN AUTISM

EEG Obtain a sleep-deprived electroencephalogram (EEG) in children with a history of significant language regression, episodic symptoms, or other indicators of possible seizures. Ideally, the EEG should include monitoring during sleep to help rule out acquired epileptic aphasia (Landau-Kleffner syndrome), a rare disorder associated with late-onset language regression. MRI of the brain is not routine but should be considered if indicated by the history or neurologic exam.

Consultations Consider consulting colleagues in neurology and medical genetics, especially when patients present with definite neurologic signs and symptoms or obvious dysmorphic features. These medical specialists are trained to screen for complex and rare syndromes that are sometimes associated with features of PDD.

Step 2. Is it PDD or another psychiatric disorder?

Psychiatric disorders that can be mistaken for PDD are listed in Table 3. The central feature of all PDDs is disturbance in social relatedness, and a diagnosis of PDD requires a history of significant social impairment.

Problems with social reciprocity in PDD are qualitatively different from the social impairment seen in other psychiatric disorders. For example, a child with attention-deficit/hyperactivity disorder (ADHD) and oppositional defiant disorder (ODD) may have few friends because of a tendency to act impulsively and get into frequent conflicts with others. These social difficulties would not be considered indicative of PDD, as they typify those seen in children with ADHD and ODD.

Mental retardation Although mental retardation occurs in approximately 75% of persons with autism, most patients with mental retardation do not have autism. In assessing an individual with mental retardation for symptoms of PDD, consider the overall developmental level. It is not uncommon for individuals with mental retardation to have mild social problems, a history of language delay, and even motor stereotypies. These symptoms are considered indicative of PDD only when they are more severe than would be expected for the patient’s developmental level.

Table 2

TESTING OPTIONS FOR PATIENTS WITH PDD

Table 3

DIFFERENTIAL DIAGNOSIS OF PDD

Other symptoms that are relatively specific to autism and PDD include lack of appropriate eye-to-eye gaze, abnormal speech prosody, echolalia, pronominal reversal, and narrow and circumscribed interests. The presence of these symptoms in excess should increase your suspicion of comorbid PDD.

RAD Reactive attachment disorder presents with abnormal social relatedness that can sometimes be confused with milder PDDs, especially in patients with comorbid mental retardation. In RAD, however, a history of severe neglect or abuse is thought to have caused the abnormal social relatedness. Placing the child in a caring and secure environment should improve many of the social deficits.

Language disorders are distinguished from PDDs by the absence of marked social impairment and lack of restricted interests and repetitive behaviors. In addition, children with primary language disorders often have intact nonverbal communication skills and make other attempts to communicate (e.g., through gesture, eye contact).

Stereotypic movement disorder can be seen in individuals with and without comorbid mental retardation. It is not diagnosed in the presence of autism, as these movements are thought to be part of the underlying disorder. The lack of social and communication impairments distinguishes stereotypic movement disorder from PDD.

ADHD Many children with autism and other PDDs have interfering symptoms of inattention, hyperactivity, and impulsivity. We usually do not give them an additional diagnosis of ADHD, as these symptoms are common in PDD. The pathophysiology of these symptoms may be different in ADHD and PDD, as evidenced by the frequent report of adverse effects following stimulant treatment of children with autism.²

Social phobia In higher functioning individuals with PDD, excessive social anxiety can sometimes be confused with social phobia. In social phobia, however, individuals usually do not exhibit marked problems with social relatedness and are able to interact normally with persons they know well and in some situations.

OCD Obsessive-compulsive disorder can occur in individuals with PDD but must be distinguished from the abnormal preoccupations and ritualistic behavior characteristic of autism. In autism, these activities often differ in quality from obsessions and compulsions.³ Furthermore, they usually are not associated with distress, and repetitive behaviors are not linked to a specific obsession.

Selective mutism is usually easy to distinguish from PDD because the affected child is typically able to talk in certain environments, such as at home. Also, the onset of selective mutism follows a period of normal social and communicative development.

Schizophrenia Autism was historically conceptualized as a type of childhood schizophrenia but is now thought to be distinct from the psychotic disorders. Schizophrenia with onset in childhood is much more rare than autism. Its onset usually occurs after several years of normal development, though some children with schizophrenia may have symptoms that resemble PDD early in their illness.⁴ Autistic persons may at times present with symptoms of a thought disorder. A diagnosis of schizophrenia usually is not made without evidence of prominent delusions and hallucinations.

Personality disorders PDDs are sometimes difficult to distinguish from personality disorders with similar features (e.g., schizotypal personality, schizoid personality). The social impairment in autistic and Asperger’s disorders is generally of earlier onset and greater severity than that seen in personality disorders. Those with personality disorders also typically lack stereotyped language or repetitive behaviors that are common in PDDs.

Table 4

CHARACTERISTIC FEATURES OF DSM-IV SUBTYPES OF PERVASIVE DEVELOPMENTAL DISORDERS

Social awkwardness Finally, some of the PDDs that allow higher functioning (e.g., Asperger’s disorder and PDD not otherwise specified [NOS]) need to be distinguished from normal social awkwardness that can be common, especially in adolescence. The social impairment in PDD is marked and interferes with normal functioning and development.

Step 3. Which PDD is it?

DSM-IV describes five subtypes of PDD (autistic disorder, Asperger’s disorder, Rett’s disorder, childhood disintegrative disorder, and PDD NOS) that have in common problems with reciprocal social interaction (Table 4). For psychiatrists making the diagnosis, it is probably most difficult to differentiate the two most common types: autistic and Asperger’s disorders.

Autistic disorder is the prototypical PDD that is associated with abnormalities in reciprocal social interaction, qualitative impairments in communication, and narrow interests and repetitive behaviors (Table 1). By definition, symptoms of the disorder manifest by age 3.

Asperger’s disorder has several features that distinguish it from autistic disorder:

• Children with Asperger’s disorder do not have language delays. By definition, a child who has not developed single words by age 2 cannot be diagnosed with Asperger’s disorder.
• Early cognitive development in Asperger’s disorder is normal. Children with Asperger’s disorder are much more likely to have normal or above-average intellectual functioning than children with autistic disorder.
• Circumscribed interests and intense preoccupations are more common than motor stereotypies.
• Affected children may show verbal abilities that greatly exceed their visual-spatial skills. This may be apparent on individualized intelligence testing (i.e., verbal IQ > performance IQ) and clinically in the form of good language abilities but lagging fine-motor development (e.g., clumsiness).

Rett’s disorder occurs almost exclusively in girls, whereas autistic disorder is more common in boys. Following a brief period of normal development, affected girls experience deceleration of head growth, loss of previously acquired purposeful hand skills with subsequent development of stereotyped hand-wringing movements, loss of social engagement, onset of trunk and gait ataxia, and severe language and cognitive impairment. Genetic testing for mutations at MECP2 will be positive in most patients having all features of the classic phenotype.⁵

Childhood disintegrative disorder is thought to be more rare than autistic disorder. Following at least 2 years of normal development, affected children lose previously acquired skills. These can include play skills, language, social skills, bowel or bladder control, and motor skills. The children show impairments in social interaction, communication, and behavior of the type common to autistic disorder and often have severe mental retardation. The late onset of severe regression in development often prompts extensive neurologic evaluation, but a specific etiology is usually not found. The disorder is not diagnosed if full diagnostic criteria for autistic disorder are met (including onset of symptoms before age 3).

PDD NOS is diagnosed in many patients who are determined to have a significant impairment in social relatedness but do not meet full criteria for a specific PDD. Recent epidemiologic studies suggest that PDD NOS may be more common than autistic disorder.⁶ It may also be an appropriate designation for children with other proposed diagnostic constructs, such as nonverbal learning disabilities⁷ and multiple complex developmental disorders.⁸

Unfortunately, children diagnosed with PDD NOS instead of the better-recognized term “autism” may be denied appropriate social and financial services. When you inform others about the diagnosis of PDD NOS, it is important to emphasize to parents, schools, and funding agencies that PDD NOS is related to autism and should be considered one of the “autism spectrum disorders.” Children with PDDs or autism spectrum disorders will often benefit from similar treatment and educational interventions. In addition, their needs are equivalent for adequate insurance coverage and funding for specialized treatments.

Related resources

• Cohen DJ, Volkmar FR (eds). Handbook of autism and pervasive developmental disorders (2nd ed). New York: Wiley, 1997.
• National Institute of Mental Health. Autism booklet. http://www.nimh.nih.gov/publicat/autism.cfm
• Autism Society of America http://www.autism-society.org Online Asperger Syndrome Information and Support (OASIS) http://www.udel.edu/bkirby/asperger

Autism and related pervasive developmental disorders (PDD) are increasingly being identified in children—and even in some adolescents and adults. As a result, psychiatry now recognizes that these debilitating disorders are more common than was once believed, with a prevalence as high as 1 in 250.

An accurate diagnosis can help families take advantage of the variety of treatments being offered and investigated for affected individuals (Box).¹ As psychiatrists who primarily see patients with autism and PDD, we recommend a three-step diagnostic approach that includes:

• a comprehensive initial assessment to rule out medical or neurologic illnesses that can mimic or are associated with autism
• differentiating PDDs from other psychiatric disorders with similar symptoms
• distinguishing among the five PDD subtypes described in DSM-IV.²

Step 1: Comprehensive initial assessment

Assessment for possible PDD begins with a comprehensive history and examination. Most patients will be assessed in childhood, but milder symptoms of autism or Asperger’s disorder may go unrecognized initially and not be brought to a clinician’s attention until adolescence or even adulthood.

As PDDs are childhood-onset disorders, the logical approach emphasizes the developmental course and onset of symptoms. By definition, children with autism show evidence of the disorder by age 3. However, the diagnosis can often be made as early as age 18 to 24 months, when typically developing children exhibit a number of social and communicative milestones that are absent in autism.

History A thorough description of the mother’s pregnancy, labor, and delivery (if known) can help you determine whether intrauterine or perinatal events could be related to the patient’s presenting problem. These include infections and exposure to exogenous substances (e.g., alcohol) during the pregnancy, as well as complications during pregnancy and delivery (e.g., maternal bleeding, neonatal hypoxia).

A complete description of the child’s development including major milestones (e.g., sitting without support, walking, first words) can help distinguish among certain diagnoses and estimate the extent of developmental delay. Ask the parents what first concerned them about their child’s development, as children with autism most often present with delays in social or language milestones. Any developmental regression in acquired skills may implicate other neurologic processes and help distinguish among the subtypes of PDD.

Symptoms Review the symptoms of autism at length in all patients in whom you suspect a PDD. It is important to assess these symptoms in the context of the child’s overall developmental level. For example, a child with known mental retardation should be compared with peers of similar age and cognitive impairment.

Box

Approximately 75% of persons with autism are diagnosed with mental retardation. A review of intellectual abilities and level of adaptive functioning can suggest the degree of mental retardation. A detailed family history is important because autism and other syndromes associated with mental retardation have varying degrees of heritability.

A thorough medical history, review of systems, and physical exam (with focus on the neurologic exam) can suggest the presence of medical conditions that could mimic or be associated with autism. The symptoms of autism are traditionally divided into three domains: social, communication, and repetitive behavior/narrow interests (Table 1):

• Social impairment includes problems with nonverbal behaviors such as eye contact, facial expressions, and “body language;” failure to develop peer relationships; lack of spontaneous seeking to share enjoyment, interests, or achievements with other people; and lack of social or emotional reciprocity.
• Communication impairment includes language delay, decreased communication with others, conversational impairment, echolalia, and lack of imaginative or social imitative play.
• Impairments in behavior, interests, and activity take the form of all-encompassing preoccupations, “need for sameness” and compulsive rituals, motor stereotypies, and preoccupation with parts of objects.

After the preliminary history and exam, you should have an initial impression as to whether a PDD may be present. Conditions that could mimic or co-exist with autism will then need to be evaluated (Table 2).

Hearing and vision testing Every child presenting with a language or cognitive delay should have an adequate hearing assessment, including an audiogram at the very least. If results are equivocal, then brainstem auditory-evoked responses are indicated to establish the auditory system’s integrity. Vision screening should be completed, and the child should be referred to an ophthalmologist if a problem is suspected.

IQ testing A child with cognitive delays, learning problems, or suspected PDD should be referred to an experienced psychologist for intelligence testing. Public school systems often provide this service. Intelligence testing can document mental retardation and offer important information about the child’s strengths and weaknesses in learning.

Speech and language assessment A speech and language pathologist should evaluate children with PDD and/or language problems for articulation, prosody, receptive and expressive language, and pragmatics.

Lab and genetic tests Laboratory investigation in a child with cognitive delays should include routine blood chemistries, CBC, thyroid function tests, and lead level. Screen for fragile X syndrome, as its symptoms overlap those of autism, and for disorders of amino acid and organic acid metabolism. Finally, consider a chromosome karyotype, especially in patients with dysmorphic features on physical exam.

Table 1

THE THREE DOMAINS OF IMPAIRMENT IN AUTISM

EEG Obtain a sleep-deprived electroencephalogram (EEG) in children with a history of significant language regression, episodic symptoms, or other indicators of possible seizures. Ideally, the EEG should include monitoring during sleep to help rule out acquired epileptic aphasia (Landau-Kleffner syndrome), a rare disorder associated with late-onset language regression. MRI of the brain is not routine but should be considered if indicated by the history or neurologic exam.

Consultations Consider consulting colleagues in neurology and medical genetics, especially when patients present with definite neurologic signs and symptoms or obvious dysmorphic features. These medical specialists are trained to screen for complex and rare syndromes that are sometimes associated with features of PDD.

Step 2. Is it PDD or another psychiatric disorder?

Psychiatric disorders that can be mistaken for PDD are listed in Table 3. The central feature of all PDDs is disturbance in social relatedness, and a diagnosis of PDD requires a history of significant social impairment.

Problems with social reciprocity in PDD are qualitatively different from the social impairment seen in other psychiatric disorders. For example, a child with attention-deficit/hyperactivity disorder (ADHD) and oppositional defiant disorder (ODD) may have few friends because of a tendency to act impulsively and get into frequent conflicts with others. These social difficulties would not be considered indicative of PDD, as they typify those seen in children with ADHD and ODD.

Mental retardation Although mental retardation occurs in approximately 75% of persons with autism, most patients with mental retardation do not have autism. In assessing an individual with mental retardation for symptoms of PDD, consider the overall developmental level. It is not uncommon for individuals with mental retardation to have mild social problems, a history of language delay, and even motor stereotypies. These symptoms are considered indicative of PDD only when they are more severe than would be expected for the patient’s developmental level.

Table 2

TESTING OPTIONS FOR PATIENTS WITH PDD

Table 3

DIFFERENTIAL DIAGNOSIS OF PDD

Other symptoms that are relatively specific to autism and PDD include lack of appropriate eye-to-eye gaze, abnormal speech prosody, echolalia, pronominal reversal, and narrow and circumscribed interests. The presence of these symptoms in excess should increase your suspicion of comorbid PDD.

RAD Reactive attachment disorder presents with abnormal social relatedness that can sometimes be confused with milder PDDs, especially in patients with comorbid mental retardation. In RAD, however, a history of severe neglect or abuse is thought to have caused the abnormal social relatedness. Placing the child in a caring and secure environment should improve many of the social deficits.

Language disorders are distinguished from PDDs by the absence of marked social impairment and lack of restricted interests and repetitive behaviors. In addition, children with primary language disorders often have intact nonverbal communication skills and make other attempts to communicate (e.g., through gesture, eye contact).

Stereotypic movement disorder can be seen in individuals with and without comorbid mental retardation. It is not diagnosed in the presence of autism, as these movements are thought to be part of the underlying disorder. The lack of social and communication impairments distinguishes stereotypic movement disorder from PDD.

ADHD Many children with autism and other PDDs have interfering symptoms of inattention, hyperactivity, and impulsivity. We usually do not give them an additional diagnosis of ADHD, as these symptoms are common in PDD. The pathophysiology of these symptoms may be different in ADHD and PDD, as evidenced by the frequent report of adverse effects following stimulant treatment of children with autism.²

Social phobia In higher functioning individuals with PDD, excessive social anxiety can sometimes be confused with social phobia. In social phobia, however, individuals usually do not exhibit marked problems with social relatedness and are able to interact normally with persons they know well and in some situations.

OCD Obsessive-compulsive disorder can occur in individuals with PDD but must be distinguished from the abnormal preoccupations and ritualistic behavior characteristic of autism. In autism, these activities often differ in quality from obsessions and compulsions.³ Furthermore, they usually are not associated with distress, and repetitive behaviors are not linked to a specific obsession.

Selective mutism is usually easy to distinguish from PDD because the affected child is typically able to talk in certain environments, such as at home. Also, the onset of selective mutism follows a period of normal social and communicative development.

Schizophrenia Autism was historically conceptualized as a type of childhood schizophrenia but is now thought to be distinct from the psychotic disorders. Schizophrenia with onset in childhood is much more rare than autism. Its onset usually occurs after several years of normal development, though some children with schizophrenia may have symptoms that resemble PDD early in their illness.⁴ Autistic persons may at times present with symptoms of a thought disorder. A diagnosis of schizophrenia usually is not made without evidence of prominent delusions and hallucinations.

Personality disorders PDDs are sometimes difficult to distinguish from personality disorders with similar features (e.g., schizotypal personality, schizoid personality). The social impairment in autistic and Asperger’s disorders is generally of earlier onset and greater severity than that seen in personality disorders. Those with personality disorders also typically lack stereotyped language or repetitive behaviors that are common in PDDs.

Table 4

CHARACTERISTIC FEATURES OF DSM-IV SUBTYPES OF PERVASIVE DEVELOPMENTAL DISORDERS

Social awkwardness Finally, some of the PDDs that allow higher functioning (e.g., Asperger’s disorder and PDD not otherwise specified [NOS]) need to be distinguished from normal social awkwardness that can be common, especially in adolescence. The social impairment in PDD is marked and interferes with normal functioning and development.

Step 3. Which PDD is it?

DSM-IV describes five subtypes of PDD (autistic disorder, Asperger’s disorder, Rett’s disorder, childhood disintegrative disorder, and PDD NOS) that have in common problems with reciprocal social interaction (Table 4). For psychiatrists making the diagnosis, it is probably most difficult to differentiate the two most common types: autistic and Asperger’s disorders.

Autistic disorder is the prototypical PDD that is associated with abnormalities in reciprocal social interaction, qualitative impairments in communication, and narrow interests and repetitive behaviors (Table 1). By definition, symptoms of the disorder manifest by age 3.

Asperger’s disorder has several features that distinguish it from autistic disorder:

• Children with Asperger’s disorder do not have language delays. By definition, a child who has not developed single words by age 2 cannot be diagnosed with Asperger’s disorder.
• Early cognitive development in Asperger’s disorder is normal. Children with Asperger’s disorder are much more likely to have normal or above-average intellectual functioning than children with autistic disorder.
• Circumscribed interests and intense preoccupations are more common than motor stereotypies.
• Affected children may show verbal abilities that greatly exceed their visual-spatial skills. This may be apparent on individualized intelligence testing (i.e., verbal IQ > performance IQ) and clinically in the form of good language abilities but lagging fine-motor development (e.g., clumsiness).

Rett’s disorder occurs almost exclusively in girls, whereas autistic disorder is more common in boys. Following a brief period of normal development, affected girls experience deceleration of head growth, loss of previously acquired purposeful hand skills with subsequent development of stereotyped hand-wringing movements, loss of social engagement, onset of trunk and gait ataxia, and severe language and cognitive impairment. Genetic testing for mutations at MECP2 will be positive in most patients having all features of the classic phenotype.⁵

Childhood disintegrative disorder is thought to be more rare than autistic disorder. Following at least 2 years of normal development, affected children lose previously acquired skills. These can include play skills, language, social skills, bowel or bladder control, and motor skills. The children show impairments in social interaction, communication, and behavior of the type common to autistic disorder and often have severe mental retardation. The late onset of severe regression in development often prompts extensive neurologic evaluation, but a specific etiology is usually not found. The disorder is not diagnosed if full diagnostic criteria for autistic disorder are met (including onset of symptoms before age 3).

PDD NOS is diagnosed in many patients who are determined to have a significant impairment in social relatedness but do not meet full criteria for a specific PDD. Recent epidemiologic studies suggest that PDD NOS may be more common than autistic disorder.⁶ It may also be an appropriate designation for children with other proposed diagnostic constructs, such as nonverbal learning disabilities⁷ and multiple complex developmental disorders.⁸

Unfortunately, children diagnosed with PDD NOS instead of the better-recognized term “autism” may be denied appropriate social and financial services. When you inform others about the diagnosis of PDD NOS, it is important to emphasize to parents, schools, and funding agencies that PDD NOS is related to autism and should be considered one of the “autism spectrum disorders.” Children with PDDs or autism spectrum disorders will often benefit from similar treatment and educational interventions. In addition, their needs are equivalent for adequate insurance coverage and funding for specialized treatments.

Related resources

• Cohen DJ, Volkmar FR (eds). Handbook of autism and pervasive developmental disorders (2nd ed). New York: Wiley, 1997.
• National Institute of Mental Health. Autism booklet. http://www.nimh.nih.gov/publicat/autism.cfm
• Autism Society of America http://www.autism-society.org Online Asperger Syndrome Information and Support (OASIS) http://www.udel.edu/bkirby/asperger

Adolescents with depressive disorders tend to arrive in psychiatrists’ offices when their behavior has already been identified as problematic. Suicide attempts, academic failure, substance abuse, and family conflicts can all lead to teen psychiatric referrals. Other times, subtler changes in behavior may lead a family doctor or pediatrician to suspect depression and to send an adolescent to you for a psychiatric consultation.

The psychiatrist’s task is challenging. Adolescents are usually brought in by their worried parents and may not want to talk to a psychiatrist. Or they may be unable to accurately describe their internal states. Even people who know an adolescent well may not discern the emotions that drive his or her behavior. Adding to the mix are the recurrent nature of major depression in adolescents and the likelihood of complicating comorbid psychiatric conditions (Box).^1-6

Based on clinical evidence, we offer advice to help you promptly identify and effectively treat adolescents with depressive disorders. We also provide preliminary information on two studies examining medication treatment, psychotherapy, and combined treatment for teens with major depression.

Adolescent depression disorders

Symptoms of depression in adolescents are similar to those in adults, and it is appropriate for psychiatrists to use DSM-IV diagnostic criteria for making the diagnosis. The three primary depressive disorders for both adults and adolescents are major depressive disorder (MDD), dysthymic disorder, and depressive disorder not otherwise specified (NOS).

Box

Although the symptoms that make up the diagnostic criteria are similar for adults and teens, the behavioral manifestations and response to treatment may differ. The adolescent may present as irritable and angry, rather than overly sad. Impairments in functioning are likely to be related to decline in school performance, social withdrawal, or increased conflicts with peers and family.

As for treatment, certain antidepressant medications of proven efficacy in adults (i.e., tricyclics) do not seem to work for adolescents.

MDD is a time-limited episode of depressive symptoms severe enough to cause functional impairment, such as decline in school performance, social withdrawal, or increased conflicts with peers and family. Symptoms must be present at least 2 weeks.

Dysthymia is a chronic depression that is less severe than MDD and lasts 1 year or longer without sustained remission. It often begins early in childhood and may include periods of increased symptoms consistent with major depression (sometimes called “double depression”).

Depressive disorder NOS is a category of depression that, though clinically significant, does not meet the full criteria for severity, duration, or level of impairment of MDD or dysthymia.

Unless otherwise specified, the terms “depression” and “depressive disorder” in this article are used generically to include all three of these disorders.

Depressive disorders must be differentiated from bipolar disorder, which is characterized by least one prior episode of mania (for bipolar type I) or hypomania (for bipolar type II). The clinical picture of bipolar disorder in youths may differ from that seen in adults. For example, bipolar youth often present with dysphoric mood interspersed with frequent, short periods of intense emotional lability and irritability, rather than “classic” euphoria.

Diagnosis

To diagnose a depressive disorder in an adolescent, information is typically obtained from multiple sources, most commonly the teenage patient and at least one of the parents. Because several sources are involved, however, the information may be conflicting. For instance, the adolescent may contradict a parent’s report that he or she is having difficulties in school or has a substance abuse problem.

Interviewing skills and clinical judgment are required of the clinician in these situations. It is important to:

• obtain a complete description of the adolescent’s behavior and mood over time and as accurate a description as possible of when changes occurred
• assess comorbid conditions (particularly anxiety, attention-deficit/hyperactivity disorder [ADHD], conduct disorder, and substance abuse)
• differentiate between unipolar MDD and bipolar disorder
• evaluate the risk of suicide.

Table 1

DEPRESSION IN ADOLESCENTS AND ADULTS

Interviewing Standardized diagnostic interviews, such as the Schedule for Affective Disorders and Schizophrenia for children (K-SADS), are commonly used to research adolescent depression but require special training and approximately 1 to 2 hours to administer. As an alternative, clinicians generally develop their own “semi-structured interview” to try to collect all the relevant information required for an accurate diagnosis.⁷

The interview should be conducted with the adolescent and the parent(s), first separately for ease of disclosure then together to reconcile any differences in the information they report. Open-ended questions and time for building rapport may facilitate disclosure from a reticent adolescent. At times, however, one must make the diagnosis by relying more on reports from others who know the child well. Building a trusting therapeutic relationship then becomes part of ongoing treatment.

Standardized measures In addition to the interview, standardized self-report and other-report measures can help:

• The Child Depression Rating Scale-Revised, commonly used in clinical research, can also be used in practice to quantify symptom severity and document treatment response. A score above 40 usually indicates major depression; a score below 28 indicates remission of depression.^8-10
• The Beck Depression Inventory (BDI), a 21-item self-report questionnaire for adults, has reasonable reliability and validity for adolescents. Its modest specificity suggests that it may measure general distress and dysphoria, which is not specific to depressive disorders. The language may be too difficult for some younger adolescents and those with poor reading comprehension skills.
• The Children’s Depression Inventory, a version of the BDI for prepubertal children, can be considered for adolescents whose cognitive and/or reading skills are less mature.
• Achenbach’s Child Behavior Checklists and other standardized questionnaires can screen for comorbid psychopathology.

Assessing psychosocial stress, such as conflicts with parents or peers, school problems, or risk-taking behavior, is also important. Depressed youth often have family members with histories of depression, alcoholism, anxiety, and other psychiatric diagnoses. History of sexual abuse has been linked to depression.³ The depressed adolescent’s impaired functioning in school and at home may cause secondary stress, increasing the burden of illness and need for treatment.

Suicide risk Although suicide remains rare among adolescents in general, the rate of suicide among this age group has risen dramatically over the past decade, particularly among younger teens and preteens. In 1997, suicide was the third leading cause of death in adolescents after accidental injuries and homicide.

Adolescents with depressive disorders are at increased risk for suicide, and boys are more likely than girls to attempt and complete suicide. It is therefore imperative to assess and document suicide risk for each adolescent who presents with depressive symptoms.

After establishing a rapport, the most effective screening is a straightforward conversation with the adolescent about suicidal ideation, intent, and behavior. Assess the social context of support and psychopathology in the family, availability and accessibility of lethal suicide methods (e.g., firearms in the home), and presence of events that could influence imitative suicidal behavior (e.g., a friend’s suicide).⁶

Treatment

Approaches to adolescent depression include (in increasing order of intensity and complexity) watchful monitoring, nonspecific supportive therapy, pharmacotherapy, specific psychotherapy (i.e., cognitive-behavioral or interpersonal therapy), and combined treatment (e.g., psychotherapy plus pharmacotherapy, adolescent psychotherapy plus family therapy).

There are no clear-cut guidelines as to whether pharmacologic or psychosocial therapy should be offered first.¹¹ In the community, patient and family preferences, past treatment response, and the clinician’s background and expertise influence the choice of treatment. As with adults, adolescents deemed at high risk for suicidal behavior must receive immediate attention from mental health professionals and must be monitored, usually in an inpatient setting.

Watchful monitoring means to wait and see if the youth improves spontaneously.

In some studies, nearly one-half (48%) of adolescents with depression were found to go into spontaneous remission within 8 weeks.¹² Watchful monitoring, however, would leave most patients still depressed, and no predictors of spontaneous remission have been identified.

Table 2

SSRI DOSAGES FOUND TO BE EFFECTIVE IN ADOLESCENTS WITH MDD

Because of the risks of suicide and social and academic impairment, monitoring alone is acceptable only for a few weeks, and only in cases where depression is mild and uncomplicated. In any case, “monitoring” requires that you periodically reassess the teen and be available for consultation between assessments.

Nonspecific supportive therapy Most psychotherapy provided in the community probably is nonspecific (i.e., not theoretically driven or conducted according to a treatment manual) and supportive (i.e., aimed at providing encouragement). This approach is known to be less effective than specific psychotherapies or antidepressant pharmacotherapy, but we have virtually no data comparing it with lack of treatment.

Nonspecific supportive therapy can be considered a reasonable first-step treatment for depressed teens without complicating risk factors.¹³ Specific treatment is indicated, however, if the adolescent does not improve in a few weeks.

Pharmacotherapy A few placebo-controlled clinical trials have studied the efficacy of selective serotonin reuptake inhibitors (SSRIs) in outpatient adolescents with major depression. It must be noted that practically all the available data relate to major depression, and no systematic studies have been done in dysthymia and other types of depression in this population.

The SSRIs fluoxetine,^8,9 citalopram,¹⁰ paroxetine,¹⁴ and sertraline¹⁵ can decrease symptoms of adolescent depression over 2 to 3 months when given at dosages similar to those used in adults (Table 2). At this time, there are no data that suggest the SSRI dosage must be different in younger (12-year-old) compared with older (18-year-old) adolescents, or in girls compared with boys.

The response rate (adolescents who were substantially improved at end of treatment) ranged from 52 to 65% with SSRI medication and 33 to 48% with a placebo. This means that one would need to treat about six adolescents in order to add one to those who would improve by taking a placebo. Thus, 6 is the number needed to treat (NNT), a common index used to make decisions in evidence-based medicine. As a comparison, the NNT is 1.5 for stimulant treatment of ADHD, indicating that stimulants are more effective in ADHD (i.e., the difference between an active drug and a placebo is greater) than SSRIs are in depression.

It must be noted that receiving a placebo in clinical trials of depression does not equal absence of treatment. Typically, research participants assigned to a placebo have weekly clinical contacts, so placebo treatment could be defined as “nonspecific clinical management.” In any case, the NNT for SSRIs in adolescent depression does not appear to be substantially different from that found in adults.

There is no evidence that tricyclic antidepressants, or TCAs, are more effective than a placebo in adolescents with depression. Considering TCAs’ side effects and potential cardiotoxicity in overdose, their use is not recommended in depressed adolescents. Other antidepressants, such as bupropion and venlafaxine, have been reported to be effective in open studies and clinical observations.

Antidepressants can trigger mania in persons predisposed to bipolar disorder. Before prescribing antidepressant medications, carefully evaluate depressed adolescents for present and past symptoms of mania or hypomania and check family history for possible bipolar disorder. During treatment, monitor the adolescent for signs that suggest emerging mania.

SSRIs and stimulant medications are commonly combined, as depression and ADHD often coexist. Clinical experience indicates that these medications can be safely combined in most cases, although no systematic studies of this combination have been performed in large, representative samples of teens. As with antidepressant monotherapy, adolescent patients must be monitored for signs of mania or hypomania.

Specific psychotherapy Among psychotherapies for depression, the best evidence of efficacy exists for cognitive-behavioral therapy (CBT), which attempts to correct cognitive distortions associated with depression and promotes healthy behaviors. In two clinical studies, CBT was more effective in depressed adolescents than no treatment (i.e., wait list), supportive therapy, or systemic behavior family therapy. About two-thirds of patients treated with CBT achieved remission from depressive symptoms.^16,17

Interpersonal therapy (IPT) has been adapted for use in depressed adolescents.¹⁸ Its efficacy is supported by one controlled study that found greater improvement after 12 weekly IPT sessions than after once-monthly clinical contact.¹⁹ There is no convincing evidence that family therapies are effective in treating depression in adolescents or add much to the benefit of CBT.¹⁷

Combined treatment Research on combined treatments for adolescent depression has been limited. One study found that adding family therapy to CBT does not improve the adolescent’s depressive symptoms.⁸ Although no data are available on the combination of SSRI medication and CBT (or IPT), the National Institute of Mental Health is funding two studies on this important issue:

• The Treatment of Adolescents with Depression Study (TADS) is comparing the efficacy of fluoxetine plus CBT with that of fluoxetine or CBT alone in adolescents with major depression. The study, coordinated by John March, MD, of Duke University Medical Center, is being conducted with 432 teen subjects at 13 clinical sites.
• The Treatment of Resistant Depression in Adolescents (TORDIA) study is comparing the efficacy of antidepressant medication plus CBT to that of antidepressants alone in adolescents who have not improved on initial treatment with an SSRI. TORDIA, coordinated by David Brent, MD, University of Pittsburgh, is being conducted at six clinical sites.

Related resources

• National Institute of Mental Health. Child and Adolescent Mental Health. http://www.nimh.nih.gov/publicat/childmenu.cfm
• Depression in children and adolescents: A fact sheet for physicians. http://www.nimh.nih.gov/publicat/depchildresfact.cfm
• NIMH-funded treatment studies in adolescent depression:

Drug brand names

• Bupropion • Wellbutrin
• Citalopram • Celexa
• Fluoxetine • Prozac
• Paroxetine • Paxil
• Sertraline • Zoloft
• Venlafaxine • Effexor

Disclosure

The authors report no affiliation or financial relationship with any of the companies whose products are mentioned in this article. The opinions and assertions contained in this article are the private views of the authors and are not to be construed as official or as reflecting the views of the Department of Health and Human Services or the National Institutes of Health.

Adolescents with depressive disorders tend to arrive in psychiatrists’ offices when their behavior has already been identified as problematic. Suicide attempts, academic failure, substance abuse, and family conflicts can all lead to teen psychiatric referrals. Other times, subtler changes in behavior may lead a family doctor or pediatrician to suspect depression and to send an adolescent to you for a psychiatric consultation.

The psychiatrist’s task is challenging. Adolescents are usually brought in by their worried parents and may not want to talk to a psychiatrist. Or they may be unable to accurately describe their internal states. Even people who know an adolescent well may not discern the emotions that drive his or her behavior. Adding to the mix are the recurrent nature of major depression in adolescents and the likelihood of complicating comorbid psychiatric conditions (Box).^1-6

Based on clinical evidence, we offer advice to help you promptly identify and effectively treat adolescents with depressive disorders. We also provide preliminary information on two studies examining medication treatment, psychotherapy, and combined treatment for teens with major depression.

Adolescent depression disorders

Symptoms of depression in adolescents are similar to those in adults, and it is appropriate for psychiatrists to use DSM-IV diagnostic criteria for making the diagnosis. The three primary depressive disorders for both adults and adolescents are major depressive disorder (MDD), dysthymic disorder, and depressive disorder not otherwise specified (NOS).

Box

Although the symptoms that make up the diagnostic criteria are similar for adults and teens, the behavioral manifestations and response to treatment may differ. The adolescent may present as irritable and angry, rather than overly sad. Impairments in functioning are likely to be related to decline in school performance, social withdrawal, or increased conflicts with peers and family.

As for treatment, certain antidepressant medications of proven efficacy in adults (i.e., tricyclics) do not seem to work for adolescents.

MDD is a time-limited episode of depressive symptoms severe enough to cause functional impairment, such as decline in school performance, social withdrawal, or increased conflicts with peers and family. Symptoms must be present at least 2 weeks.

Dysthymia is a chronic depression that is less severe than MDD and lasts 1 year or longer without sustained remission. It often begins early in childhood and may include periods of increased symptoms consistent with major depression (sometimes called “double depression”).

Depressive disorder NOS is a category of depression that, though clinically significant, does not meet the full criteria for severity, duration, or level of impairment of MDD or dysthymia.

Unless otherwise specified, the terms “depression” and “depressive disorder” in this article are used generically to include all three of these disorders.

Depressive disorders must be differentiated from bipolar disorder, which is characterized by least one prior episode of mania (for bipolar type I) or hypomania (for bipolar type II). The clinical picture of bipolar disorder in youths may differ from that seen in adults. For example, bipolar youth often present with dysphoric mood interspersed with frequent, short periods of intense emotional lability and irritability, rather than “classic” euphoria.

Diagnosis

To diagnose a depressive disorder in an adolescent, information is typically obtained from multiple sources, most commonly the teenage patient and at least one of the parents. Because several sources are involved, however, the information may be conflicting. For instance, the adolescent may contradict a parent’s report that he or she is having difficulties in school or has a substance abuse problem.

Interviewing skills and clinical judgment are required of the clinician in these situations. It is important to:

• obtain a complete description of the adolescent’s behavior and mood over time and as accurate a description as possible of when changes occurred
• assess comorbid conditions (particularly anxiety, attention-deficit/hyperactivity disorder [ADHD], conduct disorder, and substance abuse)
• differentiate between unipolar MDD and bipolar disorder
• evaluate the risk of suicide.

Table 1

DEPRESSION IN ADOLESCENTS AND ADULTS

Interviewing Standardized diagnostic interviews, such as the Schedule for Affective Disorders and Schizophrenia for children (K-SADS), are commonly used to research adolescent depression but require special training and approximately 1 to 2 hours to administer. As an alternative, clinicians generally develop their own “semi-structured interview” to try to collect all the relevant information required for an accurate diagnosis.⁷

The interview should be conducted with the adolescent and the parent(s), first separately for ease of disclosure then together to reconcile any differences in the information they report. Open-ended questions and time for building rapport may facilitate disclosure from a reticent adolescent. At times, however, one must make the diagnosis by relying more on reports from others who know the child well. Building a trusting therapeutic relationship then becomes part of ongoing treatment.

Standardized measures In addition to the interview, standardized self-report and other-report measures can help:

• The Child Depression Rating Scale-Revised, commonly used in clinical research, can also be used in practice to quantify symptom severity and document treatment response. A score above 40 usually indicates major depression; a score below 28 indicates remission of depression.^8-10
• The Beck Depression Inventory (BDI), a 21-item self-report questionnaire for adults, has reasonable reliability and validity for adolescents. Its modest specificity suggests that it may measure general distress and dysphoria, which is not specific to depressive disorders. The language may be too difficult for some younger adolescents and those with poor reading comprehension skills.
• The Children’s Depression Inventory, a version of the BDI for prepubertal children, can be considered for adolescents whose cognitive and/or reading skills are less mature.
• Achenbach’s Child Behavior Checklists and other standardized questionnaires can screen for comorbid psychopathology.

Assessing psychosocial stress, such as conflicts with parents or peers, school problems, or risk-taking behavior, is also important. Depressed youth often have family members with histories of depression, alcoholism, anxiety, and other psychiatric diagnoses. History of sexual abuse has been linked to depression.³ The depressed adolescent’s impaired functioning in school and at home may cause secondary stress, increasing the burden of illness and need for treatment.

Suicide risk Although suicide remains rare among adolescents in general, the rate of suicide among this age group has risen dramatically over the past decade, particularly among younger teens and preteens. In 1997, suicide was the third leading cause of death in adolescents after accidental injuries and homicide.

Adolescents with depressive disorders are at increased risk for suicide, and boys are more likely than girls to attempt and complete suicide. It is therefore imperative to assess and document suicide risk for each adolescent who presents with depressive symptoms.

After establishing a rapport, the most effective screening is a straightforward conversation with the adolescent about suicidal ideation, intent, and behavior. Assess the social context of support and psychopathology in the family, availability and accessibility of lethal suicide methods (e.g., firearms in the home), and presence of events that could influence imitative suicidal behavior (e.g., a friend’s suicide).⁶

Treatment

Approaches to adolescent depression include (in increasing order of intensity and complexity) watchful monitoring, nonspecific supportive therapy, pharmacotherapy, specific psychotherapy (i.e., cognitive-behavioral or interpersonal therapy), and combined treatment (e.g., psychotherapy plus pharmacotherapy, adolescent psychotherapy plus family therapy).

There are no clear-cut guidelines as to whether pharmacologic or psychosocial therapy should be offered first.¹¹ In the community, patient and family preferences, past treatment response, and the clinician’s background and expertise influence the choice of treatment. As with adults, adolescents deemed at high risk for suicidal behavior must receive immediate attention from mental health professionals and must be monitored, usually in an inpatient setting.

Watchful monitoring means to wait and see if the youth improves spontaneously.

In some studies, nearly one-half (48%) of adolescents with depression were found to go into spontaneous remission within 8 weeks.¹² Watchful monitoring, however, would leave most patients still depressed, and no predictors of spontaneous remission have been identified.

Table 2

SSRI DOSAGES FOUND TO BE EFFECTIVE IN ADOLESCENTS WITH MDD

Because of the risks of suicide and social and academic impairment, monitoring alone is acceptable only for a few weeks, and only in cases where depression is mild and uncomplicated. In any case, “monitoring” requires that you periodically reassess the teen and be available for consultation between assessments.

Nonspecific supportive therapy Most psychotherapy provided in the community probably is nonspecific (i.e., not theoretically driven or conducted according to a treatment manual) and supportive (i.e., aimed at providing encouragement). This approach is known to be less effective than specific psychotherapies or antidepressant pharmacotherapy, but we have virtually no data comparing it with lack of treatment.

Nonspecific supportive therapy can be considered a reasonable first-step treatment for depressed teens without complicating risk factors.¹³ Specific treatment is indicated, however, if the adolescent does not improve in a few weeks.

Pharmacotherapy A few placebo-controlled clinical trials have studied the efficacy of selective serotonin reuptake inhibitors (SSRIs) in outpatient adolescents with major depression. It must be noted that practically all the available data relate to major depression, and no systematic studies have been done in dysthymia and other types of depression in this population.

The SSRIs fluoxetine,^8,9 citalopram,¹⁰ paroxetine,¹⁴ and sertraline¹⁵ can decrease symptoms of adolescent depression over 2 to 3 months when given at dosages similar to those used in adults (Table 2). At this time, there are no data that suggest the SSRI dosage must be different in younger (12-year-old) compared with older (18-year-old) adolescents, or in girls compared with boys.

The response rate (adolescents who were substantially improved at end of treatment) ranged from 52 to 65% with SSRI medication and 33 to 48% with a placebo. This means that one would need to treat about six adolescents in order to add one to those who would improve by taking a placebo. Thus, 6 is the number needed to treat (NNT), a common index used to make decisions in evidence-based medicine. As a comparison, the NNT is 1.5 for stimulant treatment of ADHD, indicating that stimulants are more effective in ADHD (i.e., the difference between an active drug and a placebo is greater) than SSRIs are in depression.

It must be noted that receiving a placebo in clinical trials of depression does not equal absence of treatment. Typically, research participants assigned to a placebo have weekly clinical contacts, so placebo treatment could be defined as “nonspecific clinical management.” In any case, the NNT for SSRIs in adolescent depression does not appear to be substantially different from that found in adults.

There is no evidence that tricyclic antidepressants, or TCAs, are more effective than a placebo in adolescents with depression. Considering TCAs’ side effects and potential cardiotoxicity in overdose, their use is not recommended in depressed adolescents. Other antidepressants, such as bupropion and venlafaxine, have been reported to be effective in open studies and clinical observations.

Antidepressants can trigger mania in persons predisposed to bipolar disorder. Before prescribing antidepressant medications, carefully evaluate depressed adolescents for present and past symptoms of mania or hypomania and check family history for possible bipolar disorder. During treatment, monitor the adolescent for signs that suggest emerging mania.

SSRIs and stimulant medications are commonly combined, as depression and ADHD often coexist. Clinical experience indicates that these medications can be safely combined in most cases, although no systematic studies of this combination have been performed in large, representative samples of teens. As with antidepressant monotherapy, adolescent patients must be monitored for signs of mania or hypomania.

Specific psychotherapy Among psychotherapies for depression, the best evidence of efficacy exists for cognitive-behavioral therapy (CBT), which attempts to correct cognitive distortions associated with depression and promotes healthy behaviors. In two clinical studies, CBT was more effective in depressed adolescents than no treatment (i.e., wait list), supportive therapy, or systemic behavior family therapy. About two-thirds of patients treated with CBT achieved remission from depressive symptoms.^16,17

Interpersonal therapy (IPT) has been adapted for use in depressed adolescents.¹⁸ Its efficacy is supported by one controlled study that found greater improvement after 12 weekly IPT sessions than after once-monthly clinical contact.¹⁹ There is no convincing evidence that family therapies are effective in treating depression in adolescents or add much to the benefit of CBT.¹⁷

Combined treatment Research on combined treatments for adolescent depression has been limited. One study found that adding family therapy to CBT does not improve the adolescent’s depressive symptoms.⁸ Although no data are available on the combination of SSRI medication and CBT (or IPT), the National Institute of Mental Health is funding two studies on this important issue:

• The Treatment of Adolescents with Depression Study (TADS) is comparing the efficacy of fluoxetine plus CBT with that of fluoxetine or CBT alone in adolescents with major depression. The study, coordinated by John March, MD, of Duke University Medical Center, is being conducted with 432 teen subjects at 13 clinical sites.
• The Treatment of Resistant Depression in Adolescents (TORDIA) study is comparing the efficacy of antidepressant medication plus CBT to that of antidepressants alone in adolescents who have not improved on initial treatment with an SSRI. TORDIA, coordinated by David Brent, MD, University of Pittsburgh, is being conducted at six clinical sites.

Related resources

• National Institute of Mental Health. Child and Adolescent Mental Health. http://www.nimh.nih.gov/publicat/childmenu.cfm
• Depression in children and adolescents: A fact sheet for physicians. http://www.nimh.nih.gov/publicat/depchildresfact.cfm
• NIMH-funded treatment studies in adolescent depression:

Drug brand names

• Bupropion • Wellbutrin
• Citalopram • Celexa
• Fluoxetine • Prozac
• Paroxetine • Paxil
• Sertraline • Zoloft
• Venlafaxine • Effexor

Disclosure

The authors report no affiliation or financial relationship with any of the companies whose products are mentioned in this article. The opinions and assertions contained in this article are the private views of the authors and are not to be construed as official or as reflecting the views of the Department of Health and Human Services or the National Institutes of Health.

Adolescents with depressive disorders tend to arrive in psychiatrists’ offices when their behavior has already been identified as problematic. Suicide attempts, academic failure, substance abuse, and family conflicts can all lead to teen psychiatric referrals. Other times, subtler changes in behavior may lead a family doctor or pediatrician to suspect depression and to send an adolescent to you for a psychiatric consultation.

The psychiatrist’s task is challenging. Adolescents are usually brought in by their worried parents and may not want to talk to a psychiatrist. Or they may be unable to accurately describe their internal states. Even people who know an adolescent well may not discern the emotions that drive his or her behavior. Adding to the mix are the recurrent nature of major depression in adolescents and the likelihood of complicating comorbid psychiatric conditions (Box).^1-6

Based on clinical evidence, we offer advice to help you promptly identify and effectively treat adolescents with depressive disorders. We also provide preliminary information on two studies examining medication treatment, psychotherapy, and combined treatment for teens with major depression.

Adolescent depression disorders

Symptoms of depression in adolescents are similar to those in adults, and it is appropriate for psychiatrists to use DSM-IV diagnostic criteria for making the diagnosis. The three primary depressive disorders for both adults and adolescents are major depressive disorder (MDD), dysthymic disorder, and depressive disorder not otherwise specified (NOS).

Box

Although the symptoms that make up the diagnostic criteria are similar for adults and teens, the behavioral manifestations and response to treatment may differ. The adolescent may present as irritable and angry, rather than overly sad. Impairments in functioning are likely to be related to decline in school performance, social withdrawal, or increased conflicts with peers and family.

As for treatment, certain antidepressant medications of proven efficacy in adults (i.e., tricyclics) do not seem to work for adolescents.

MDD is a time-limited episode of depressive symptoms severe enough to cause functional impairment, such as decline in school performance, social withdrawal, or increased conflicts with peers and family. Symptoms must be present at least 2 weeks.

Dysthymia is a chronic depression that is less severe than MDD and lasts 1 year or longer without sustained remission. It often begins early in childhood and may include periods of increased symptoms consistent with major depression (sometimes called “double depression”).

Depressive disorder NOS is a category of depression that, though clinically significant, does not meet the full criteria for severity, duration, or level of impairment of MDD or dysthymia.

Unless otherwise specified, the terms “depression” and “depressive disorder” in this article are used generically to include all three of these disorders.

Depressive disorders must be differentiated from bipolar disorder, which is characterized by least one prior episode of mania (for bipolar type I) or hypomania (for bipolar type II). The clinical picture of bipolar disorder in youths may differ from that seen in adults. For example, bipolar youth often present with dysphoric mood interspersed with frequent, short periods of intense emotional lability and irritability, rather than “classic” euphoria.

Diagnosis

To diagnose a depressive disorder in an adolescent, information is typically obtained from multiple sources, most commonly the teenage patient and at least one of the parents. Because several sources are involved, however, the information may be conflicting. For instance, the adolescent may contradict a parent’s report that he or she is having difficulties in school or has a substance abuse problem.

Interviewing skills and clinical judgment are required of the clinician in these situations. It is important to:

• obtain a complete description of the adolescent’s behavior and mood over time and as accurate a description as possible of when changes occurred
• assess comorbid conditions (particularly anxiety, attention-deficit/hyperactivity disorder [ADHD], conduct disorder, and substance abuse)
• differentiate between unipolar MDD and bipolar disorder
• evaluate the risk of suicide.

Table 1

DEPRESSION IN ADOLESCENTS AND ADULTS

Interviewing Standardized diagnostic interviews, such as the Schedule for Affective Disorders and Schizophrenia for children (K-SADS), are commonly used to research adolescent depression but require special training and approximately 1 to 2 hours to administer. As an alternative, clinicians generally develop their own “semi-structured interview” to try to collect all the relevant information required for an accurate diagnosis.⁷

The interview should be conducted with the adolescent and the parent(s), first separately for ease of disclosure then together to reconcile any differences in the information they report. Open-ended questions and time for building rapport may facilitate disclosure from a reticent adolescent. At times, however, one must make the diagnosis by relying more on reports from others who know the child well. Building a trusting therapeutic relationship then becomes part of ongoing treatment.

Standardized measures In addition to the interview, standardized self-report and other-report measures can help:

• The Child Depression Rating Scale-Revised, commonly used in clinical research, can also be used in practice to quantify symptom severity and document treatment response. A score above 40 usually indicates major depression; a score below 28 indicates remission of depression.^8-10
• The Beck Depression Inventory (BDI), a 21-item self-report questionnaire for adults, has reasonable reliability and validity for adolescents. Its modest specificity suggests that it may measure general distress and dysphoria, which is not specific to depressive disorders. The language may be too difficult for some younger adolescents and those with poor reading comprehension skills.
• The Children’s Depression Inventory, a version of the BDI for prepubertal children, can be considered for adolescents whose cognitive and/or reading skills are less mature.
• Achenbach’s Child Behavior Checklists and other standardized questionnaires can screen for comorbid psychopathology.

Assessing psychosocial stress, such as conflicts with parents or peers, school problems, or risk-taking behavior, is also important. Depressed youth often have family members with histories of depression, alcoholism, anxiety, and other psychiatric diagnoses. History of sexual abuse has been linked to depression.³ The depressed adolescent’s impaired functioning in school and at home may cause secondary stress, increasing the burden of illness and need for treatment.

Suicide risk Although suicide remains rare among adolescents in general, the rate of suicide among this age group has risen dramatically over the past decade, particularly among younger teens and preteens. In 1997, suicide was the third leading cause of death in adolescents after accidental injuries and homicide.

Adolescents with depressive disorders are at increased risk for suicide, and boys are more likely than girls to attempt and complete suicide. It is therefore imperative to assess and document suicide risk for each adolescent who presents with depressive symptoms.

After establishing a rapport, the most effective screening is a straightforward conversation with the adolescent about suicidal ideation, intent, and behavior. Assess the social context of support and psychopathology in the family, availability and accessibility of lethal suicide methods (e.g., firearms in the home), and presence of events that could influence imitative suicidal behavior (e.g., a friend’s suicide).⁶

Treatment

Approaches to adolescent depression include (in increasing order of intensity and complexity) watchful monitoring, nonspecific supportive therapy, pharmacotherapy, specific psychotherapy (i.e., cognitive-behavioral or interpersonal therapy), and combined treatment (e.g., psychotherapy plus pharmacotherapy, adolescent psychotherapy plus family therapy).

There are no clear-cut guidelines as to whether pharmacologic or psychosocial therapy should be offered first.¹¹ In the community, patient and family preferences, past treatment response, and the clinician’s background and expertise influence the choice of treatment. As with adults, adolescents deemed at high risk for suicidal behavior must receive immediate attention from mental health professionals and must be monitored, usually in an inpatient setting.

Watchful monitoring means to wait and see if the youth improves spontaneously.

In some studies, nearly one-half (48%) of adolescents with depression were found to go into spontaneous remission within 8 weeks.¹² Watchful monitoring, however, would leave most patients still depressed, and no predictors of spontaneous remission have been identified.

Table 2

SSRI DOSAGES FOUND TO BE EFFECTIVE IN ADOLESCENTS WITH MDD

Because of the risks of suicide and social and academic impairment, monitoring alone is acceptable only for a few weeks, and only in cases where depression is mild and uncomplicated. In any case, “monitoring” requires that you periodically reassess the teen and be available for consultation between assessments.

Nonspecific supportive therapy Most psychotherapy provided in the community probably is nonspecific (i.e., not theoretically driven or conducted according to a treatment manual) and supportive (i.e., aimed at providing encouragement). This approach is known to be less effective than specific psychotherapies or antidepressant pharmacotherapy, but we have virtually no data comparing it with lack of treatment.

Nonspecific supportive therapy can be considered a reasonable first-step treatment for depressed teens without complicating risk factors.¹³ Specific treatment is indicated, however, if the adolescent does not improve in a few weeks.

Pharmacotherapy A few placebo-controlled clinical trials have studied the efficacy of selective serotonin reuptake inhibitors (SSRIs) in outpatient adolescents with major depression. It must be noted that practically all the available data relate to major depression, and no systematic studies have been done in dysthymia and other types of depression in this population.

The SSRIs fluoxetine,^8,9 citalopram,¹⁰ paroxetine,¹⁴ and sertraline¹⁵ can decrease symptoms of adolescent depression over 2 to 3 months when given at dosages similar to those used in adults (Table 2). At this time, there are no data that suggest the SSRI dosage must be different in younger (12-year-old) compared with older (18-year-old) adolescents, or in girls compared with boys.

The response rate (adolescents who were substantially improved at end of treatment) ranged from 52 to 65% with SSRI medication and 33 to 48% with a placebo. This means that one would need to treat about six adolescents in order to add one to those who would improve by taking a placebo. Thus, 6 is the number needed to treat (NNT), a common index used to make decisions in evidence-based medicine. As a comparison, the NNT is 1.5 for stimulant treatment of ADHD, indicating that stimulants are more effective in ADHD (i.e., the difference between an active drug and a placebo is greater) than SSRIs are in depression.

It must be noted that receiving a placebo in clinical trials of depression does not equal absence of treatment. Typically, research participants assigned to a placebo have weekly clinical contacts, so placebo treatment could be defined as “nonspecific clinical management.” In any case, the NNT for SSRIs in adolescent depression does not appear to be substantially different from that found in adults.

There is no evidence that tricyclic antidepressants, or TCAs, are more effective than a placebo in adolescents with depression. Considering TCAs’ side effects and potential cardiotoxicity in overdose, their use is not recommended in depressed adolescents. Other antidepressants, such as bupropion and venlafaxine, have been reported to be effective in open studies and clinical observations.

Antidepressants can trigger mania in persons predisposed to bipolar disorder. Before prescribing antidepressant medications, carefully evaluate depressed adolescents for present and past symptoms of mania or hypomania and check family history for possible bipolar disorder. During treatment, monitor the adolescent for signs that suggest emerging mania.

SSRIs and stimulant medications are commonly combined, as depression and ADHD often coexist. Clinical experience indicates that these medications can be safely combined in most cases, although no systematic studies of this combination have been performed in large, representative samples of teens. As with antidepressant monotherapy, adolescent patients must be monitored for signs of mania or hypomania.

Specific psychotherapy Among psychotherapies for depression, the best evidence of efficacy exists for cognitive-behavioral therapy (CBT), which attempts to correct cognitive distortions associated with depression and promotes healthy behaviors. In two clinical studies, CBT was more effective in depressed adolescents than no treatment (i.e., wait list), supportive therapy, or systemic behavior family therapy. About two-thirds of patients treated with CBT achieved remission from depressive symptoms.^16,17

Interpersonal therapy (IPT) has been adapted for use in depressed adolescents.¹⁸ Its efficacy is supported by one controlled study that found greater improvement after 12 weekly IPT sessions than after once-monthly clinical contact.¹⁹ There is no convincing evidence that family therapies are effective in treating depression in adolescents or add much to the benefit of CBT.¹⁷

Combined treatment Research on combined treatments for adolescent depression has been limited. One study found that adding family therapy to CBT does not improve the adolescent’s depressive symptoms.⁸ Although no data are available on the combination of SSRI medication and CBT (or IPT), the National Institute of Mental Health is funding two studies on this important issue:

• The Treatment of Adolescents with Depression Study (TADS) is comparing the efficacy of fluoxetine plus CBT with that of fluoxetine or CBT alone in adolescents with major depression. The study, coordinated by John March, MD, of Duke University Medical Center, is being conducted with 432 teen subjects at 13 clinical sites.
• The Treatment of Resistant Depression in Adolescents (TORDIA) study is comparing the efficacy of antidepressant medication plus CBT to that of antidepressants alone in adolescents who have not improved on initial treatment with an SSRI. TORDIA, coordinated by David Brent, MD, University of Pittsburgh, is being conducted at six clinical sites.

Related resources

• National Institute of Mental Health. Child and Adolescent Mental Health. http://www.nimh.nih.gov/publicat/childmenu.cfm
• Depression in children and adolescents: A fact sheet for physicians. http://www.nimh.nih.gov/publicat/depchildresfact.cfm
• NIMH-funded treatment studies in adolescent depression:

Drug brand names

• Bupropion • Wellbutrin
• Citalopram • Celexa
• Fluoxetine • Prozac
• Paroxetine • Paxil
• Sertraline • Zoloft
• Venlafaxine • Effexor

Disclosure

The authors report no affiliation or financial relationship with any of the companies whose products are mentioned in this article. The opinions and assertions contained in this article are the private views of the authors and are not to be construed as official or as reflecting the views of the Department of Health and Human Services or the National Institutes of Health.

Adolescents with depressive disorders tend to arrive in psychiatrists’ offices when their behavior has already been identified as problematic. Suicide attempts, academic failure, substance abuse, and family conflicts can all lead to teen psychiatric referrals. Other times, subtler changes in behavior may lead a family doctor or pediatrician to suspect depression and to send an adolescent to you for a psychiatric consultation.

The psychiatrist’s task is challenging. Adolescents are usually brought in by their worried parents and may not want to talk to a psychiatrist. Or they may be unable to accurately describe their internal states. Even people who know an adolescent well may not discern the emotions that drive his or her behavior. Adding to the mix are the recurrent nature of major depression in adolescents and the likelihood of complicating comorbid psychiatric conditions (Box).^1-6

Based on clinical evidence, we offer advice to help you promptly identify and effectively treat adolescents with depressive disorders. We also provide preliminary information on two studies examining medication treatment, psychotherapy, and combined treatment for teens with major depression.

Adolescent depression disorders

Symptoms of depression in adolescents are similar to those in adults, and it is appropriate for psychiatrists to use DSM-IV diagnostic criteria for making the diagnosis. The three primary depressive disorders for both adults and adolescents are major depressive disorder (MDD), dysthymic disorder, and depressive disorder not otherwise specified (NOS).

Box

Although the symptoms that make up the diagnostic criteria are similar for adults and teens, the behavioral manifestations and response to treatment may differ. The adolescent may present as irritable and angry, rather than overly sad. Impairments in functioning are likely to be related to decline in school performance, social withdrawal, or increased conflicts with peers and family.

As for treatment, certain antidepressant medications of proven efficacy in adults (i.e., tricyclics) do not seem to work for adolescents.

MDD is a time-limited episode of depressive symptoms severe enough to cause functional impairment, such as decline in school performance, social withdrawal, or increased conflicts with peers and family. Symptoms must be present at least 2 weeks.

Dysthymia is a chronic depression that is less severe than MDD and lasts 1 year or longer without sustained remission. It often begins early in childhood and may include periods of increased symptoms consistent with major depression (sometimes called “double depression”).

Depressive disorder NOS is a category of depression that, though clinically significant, does not meet the full criteria for severity, duration, or level of impairment of MDD or dysthymia.

Unless otherwise specified, the terms “depression” and “depressive disorder” in this article are used generically to include all three of these disorders.

Depressive disorders must be differentiated from bipolar disorder, which is characterized by least one prior episode of mania (for bipolar type I) or hypomania (for bipolar type II). The clinical picture of bipolar disorder in youths may differ from that seen in adults. For example, bipolar youth often present with dysphoric mood interspersed with frequent, short periods of intense emotional lability and irritability, rather than “classic” euphoria.

Diagnosis

To diagnose a depressive disorder in an adolescent, information is typically obtained from multiple sources, most commonly the teenage patient and at least one of the parents. Because several sources are involved, however, the information may be conflicting. For instance, the adolescent may contradict a parent’s report that he or she is having difficulties in school or has a substance abuse problem.

Interviewing skills and clinical judgment are required of the clinician in these situations. It is important to:

• obtain a complete description of the adolescent’s behavior and mood over time and as accurate a description as possible of when changes occurred
• assess comorbid conditions (particularly anxiety, attention-deficit/hyperactivity disorder [ADHD], conduct disorder, and substance abuse)
• differentiate between unipolar MDD and bipolar disorder
• evaluate the risk of suicide.

Table 1

DEPRESSION IN ADOLESCENTS AND ADULTS

Interviewing Standardized diagnostic interviews, such as the Schedule for Affective Disorders and Schizophrenia for children (K-SADS), are commonly used to research adolescent depression but require special training and approximately 1 to 2 hours to administer. As an alternative, clinicians generally develop their own “semi-structured interview” to try to collect all the relevant information required for an accurate diagnosis.⁷

The interview should be conducted with the adolescent and the parent(s), first separately for ease of disclosure then together to reconcile any differences in the information they report. Open-ended questions and time for building rapport may facilitate disclosure from a reticent adolescent. At times, however, one must make the diagnosis by relying more on reports from others who know the child well. Building a trusting therapeutic relationship then becomes part of ongoing treatment.

Standardized measures In addition to the interview, standardized self-report and other-report measures can help:

• The Child Depression Rating Scale-Revised, commonly used in clinical research, can also be used in practice to quantify symptom severity and document treatment response. A score above 40 usually indicates major depression; a score below 28 indicates remission of depression.^8-10
• The Beck Depression Inventory (BDI), a 21-item self-report questionnaire for adults, has reasonable reliability and validity for adolescents. Its modest specificity suggests that it may measure general distress and dysphoria, which is not specific to depressive disorders. The language may be too difficult for some younger adolescents and those with poor reading comprehension skills.
• The Children’s Depression Inventory, a version of the BDI for prepubertal children, can be considered for adolescents whose cognitive and/or reading skills are less mature.
• Achenbach’s Child Behavior Checklists and other standardized questionnaires can screen for comorbid psychopathology.

Assessing psychosocial stress, such as conflicts with parents or peers, school problems, or risk-taking behavior, is also important. Depressed youth often have family members with histories of depression, alcoholism, anxiety, and other psychiatric diagnoses. History of sexual abuse has been linked to depression.³ The depressed adolescent’s impaired functioning in school and at home may cause secondary stress, increasing the burden of illness and need for treatment.

Suicide risk Although suicide remains rare among adolescents in general, the rate of suicide among this age group has risen dramatically over the past decade, particularly among younger teens and preteens. In 1997, suicide was the third leading cause of death in adolescents after accidental injuries and homicide.

Adolescents with depressive disorders are at increased risk for suicide, and boys are more likely than girls to attempt and complete suicide. It is therefore imperative to assess and document suicide risk for each adolescent who presents with depressive symptoms.

After establishing a rapport, the most effective screening is a straightforward conversation with the adolescent about suicidal ideation, intent, and behavior. Assess the social context of support and psychopathology in the family, availability and accessibility of lethal suicide methods (e.g., firearms in the home), and presence of events that could influence imitative suicidal behavior (e.g., a friend’s suicide).⁶

Treatment

Approaches to adolescent depression include (in increasing order of intensity and complexity) watchful monitoring, nonspecific supportive therapy, pharmacotherapy, specific psychotherapy (i.e., cognitive-behavioral or interpersonal therapy), and combined treatment (e.g., psychotherapy plus pharmacotherapy, adolescent psychotherapy plus family therapy).

There are no clear-cut guidelines as to whether pharmacologic or psychosocial therapy should be offered first.¹¹ In the community, patient and family preferences, past treatment response, and the clinician’s background and expertise influence the choice of treatment. As with adults, adolescents deemed at high risk for suicidal behavior must receive immediate attention from mental health professionals and must be monitored, usually in an inpatient setting.

Watchful monitoring means to wait and see if the youth improves spontaneously.

In some studies, nearly one-half (48%) of adolescents with depression were found to go into spontaneous remission within 8 weeks.¹² Watchful monitoring, however, would leave most patients still depressed, and no predictors of spontaneous remission have been identified.

Table 2

SSRI DOSAGES FOUND TO BE EFFECTIVE IN ADOLESCENTS WITH MDD

Because of the risks of suicide and social and academic impairment, monitoring alone is acceptable only for a few weeks, and only in cases where depression is mild and uncomplicated. In any case, “monitoring” requires that you periodically reassess the teen and be available for consultation between assessments.

Nonspecific supportive therapy Most psychotherapy provided in the community probably is nonspecific (i.e., not theoretically driven or conducted according to a treatment manual) and supportive (i.e., aimed at providing encouragement). This approach is known to be less effective than specific psychotherapies or antidepressant pharmacotherapy, but we have virtually no data comparing it with lack of treatment.

Nonspecific supportive therapy can be considered a reasonable first-step treatment for depressed teens without complicating risk factors.¹³ Specific treatment is indicated, however, if the adolescent does not improve in a few weeks.

Pharmacotherapy A few placebo-controlled clinical trials have studied the efficacy of selective serotonin reuptake inhibitors (SSRIs) in outpatient adolescents with major depression. It must be noted that practically all the available data relate to major depression, and no systematic studies have been done in dysthymia and other types of depression in this population.

The SSRIs fluoxetine,^8,9 citalopram,¹⁰ paroxetine,¹⁴ and sertraline¹⁵ can decrease symptoms of adolescent depression over 2 to 3 months when given at dosages similar to those used in adults (Table 2). At this time, there are no data that suggest the SSRI dosage must be different in younger (12-year-old) compared with older (18-year-old) adolescents, or in girls compared with boys.

The response rate (adolescents who were substantially improved at end of treatment) ranged from 52 to 65% with SSRI medication and 33 to 48% with a placebo. This means that one would need to treat about six adolescents in order to add one to those who would improve by taking a placebo. Thus, 6 is the number needed to treat (NNT), a common index used to make decisions in evidence-based medicine. As a comparison, the NNT is 1.5 for stimulant treatment of ADHD, indicating that stimulants are more effective in ADHD (i.e., the difference between an active drug and a placebo is greater) than SSRIs are in depression.

It must be noted that receiving a placebo in clinical trials of depression does not equal absence of treatment. Typically, research participants assigned to a placebo have weekly clinical contacts, so placebo treatment could be defined as “nonspecific clinical management.” In any case, the NNT for SSRIs in adolescent depression does not appear to be substantially different from that found in adults.

There is no evidence that tricyclic antidepressants, or TCAs, are more effective than a placebo in adolescents with depression. Considering TCAs’ side effects and potential cardiotoxicity in overdose, their use is not recommended in depressed adolescents. Other antidepressants, such as bupropion and venlafaxine, have been reported to be effective in open studies and clinical observations.

Antidepressants can trigger mania in persons predisposed to bipolar disorder. Before prescribing antidepressant medications, carefully evaluate depressed adolescents for present and past symptoms of mania or hypomania and check family history for possible bipolar disorder. During treatment, monitor the adolescent for signs that suggest emerging mania.

SSRIs and stimulant medications are commonly combined, as depression and ADHD often coexist. Clinical experience indicates that these medications can be safely combined in most cases, although no systematic studies of this combination have been performed in large, representative samples of teens. As with antidepressant monotherapy, adolescent patients must be monitored for signs of mania or hypomania.

Specific psychotherapy Among psychotherapies for depression, the best evidence of efficacy exists for cognitive-behavioral therapy (CBT), which attempts to correct cognitive distortions associated with depression and promotes healthy behaviors. In two clinical studies, CBT was more effective in depressed adolescents than no treatment (i.e., wait list), supportive therapy, or systemic behavior family therapy. About two-thirds of patients treated with CBT achieved remission from depressive symptoms.^16,17

Interpersonal therapy (IPT) has been adapted for use in depressed adolescents.¹⁸ Its efficacy is supported by one controlled study that found greater improvement after 12 weekly IPT sessions than after once-monthly clinical contact.¹⁹ There is no convincing evidence that family therapies are effective in treating depression in adolescents or add much to the benefit of CBT.¹⁷

Combined treatment Research on combined treatments for adolescent depression has been limited. One study found that adding family therapy to CBT does not improve the adolescent’s depressive symptoms.⁸ Although no data are available on the combination of SSRI medication and CBT (or IPT), the National Institute of Mental Health is funding two studies on this important issue:

• The Treatment of Adolescents with Depression Study (TADS) is comparing the efficacy of fluoxetine plus CBT with that of fluoxetine or CBT alone in adolescents with major depression. The study, coordinated by John March, MD, of Duke University Medical Center, is being conducted with 432 teen subjects at 13 clinical sites.
• The Treatment of Resistant Depression in Adolescents (TORDIA) study is comparing the efficacy of antidepressant medication plus CBT to that of antidepressants alone in adolescents who have not improved on initial treatment with an SSRI. TORDIA, coordinated by David Brent, MD, University of Pittsburgh, is being conducted at six clinical sites.

Related resources

• National Institute of Mental Health. Child and Adolescent Mental Health. http://www.nimh.nih.gov/publicat/childmenu.cfm
• Depression in children and adolescents: A fact sheet for physicians. http://www.nimh.nih.gov/publicat/depchildresfact.cfm
• NIMH-funded treatment studies in adolescent depression:

Drug brand names

• Bupropion • Wellbutrin
• Citalopram • Celexa
• Fluoxetine • Prozac
• Paroxetine • Paxil
• Sertraline • Zoloft
• Venlafaxine • Effexor

Disclosure

The authors report no affiliation or financial relationship with any of the companies whose products are mentioned in this article. The opinions and assertions contained in this article are the private views of the authors and are not to be construed as official or as reflecting the views of the Department of Health and Human Services or the National Institutes of Health.

Adolescents with depressive disorders tend to arrive in psychiatrists’ offices when their behavior has already been identified as problematic. Suicide attempts, academic failure, substance abuse, and family conflicts can all lead to teen psychiatric referrals. Other times, subtler changes in behavior may lead a family doctor or pediatrician to suspect depression and to send an adolescent to you for a psychiatric consultation.

The psychiatrist’s task is challenging. Adolescents are usually brought in by their worried parents and may not want to talk to a psychiatrist. Or they may be unable to accurately describe their internal states. Even people who know an adolescent well may not discern the emotions that drive his or her behavior. Adding to the mix are the recurrent nature of major depression in adolescents and the likelihood of complicating comorbid psychiatric conditions (Box).^1-6

Based on clinical evidence, we offer advice to help you promptly identify and effectively treat adolescents with depressive disorders. We also provide preliminary information on two studies examining medication treatment, psychotherapy, and combined treatment for teens with major depression.

Adolescent depression disorders

Symptoms of depression in adolescents are similar to those in adults, and it is appropriate for psychiatrists to use DSM-IV diagnostic criteria for making the diagnosis. The three primary depressive disorders for both adults and adolescents are major depressive disorder (MDD), dysthymic disorder, and depressive disorder not otherwise specified (NOS).

Box

Although the symptoms that make up the diagnostic criteria are similar for adults and teens, the behavioral manifestations and response to treatment may differ. The adolescent may present as irritable and angry, rather than overly sad. Impairments in functioning are likely to be related to decline in school performance, social withdrawal, or increased conflicts with peers and family.

As for treatment, certain antidepressant medications of proven efficacy in adults (i.e., tricyclics) do not seem to work for adolescents.

MDD is a time-limited episode of depressive symptoms severe enough to cause functional impairment, such as decline in school performance, social withdrawal, or increased conflicts with peers and family. Symptoms must be present at least 2 weeks.

Dysthymia is a chronic depression that is less severe than MDD and lasts 1 year or longer without sustained remission. It often begins early in childhood and may include periods of increased symptoms consistent with major depression (sometimes called “double depression”).

Depressive disorder NOS is a category of depression that, though clinically significant, does not meet the full criteria for severity, duration, or level of impairment of MDD or dysthymia.

Unless otherwise specified, the terms “depression” and “depressive disorder” in this article are used generically to include all three of these disorders.

Depressive disorders must be differentiated from bipolar disorder, which is characterized by least one prior episode of mania (for bipolar type I) or hypomania (for bipolar type II). The clinical picture of bipolar disorder in youths may differ from that seen in adults. For example, bipolar youth often present with dysphoric mood interspersed with frequent, short periods of intense emotional lability and irritability, rather than “classic” euphoria.

Diagnosis

To diagnose a depressive disorder in an adolescent, information is typically obtained from multiple sources, most commonly the teenage patient and at least one of the parents. Because several sources are involved, however, the information may be conflicting. For instance, the adolescent may contradict a parent’s report that he or she is having difficulties in school or has a substance abuse problem.

Interviewing skills and clinical judgment are required of the clinician in these situations. It is important to:

• obtain a complete description of the adolescent’s behavior and mood over time and as accurate a description as possible of when changes occurred
• assess comorbid conditions (particularly anxiety, attention-deficit/hyperactivity disorder [ADHD], conduct disorder, and substance abuse)
• differentiate between unipolar MDD and bipolar disorder
• evaluate the risk of suicide.

Table 1

DEPRESSION IN ADOLESCENTS AND ADULTS

Interviewing Standardized diagnostic interviews, such as the Schedule for Affective Disorders and Schizophrenia for children (K-SADS), are commonly used to research adolescent depression but require special training and approximately 1 to 2 hours to administer. As an alternative, clinicians generally develop their own “semi-structured interview” to try to collect all the relevant information required for an accurate diagnosis.⁷

The interview should be conducted with the adolescent and the parent(s), first separately for ease of disclosure then together to reconcile any differences in the information they report. Open-ended questions and time for building rapport may facilitate disclosure from a reticent adolescent. At times, however, one must make the diagnosis by relying more on reports from others who know the child well. Building a trusting therapeutic relationship then becomes part of ongoing treatment.

Standardized measures In addition to the interview, standardized self-report and other-report measures can help:

• The Child Depression Rating Scale-Revised, commonly used in clinical research, can also be used in practice to quantify symptom severity and document treatment response. A score above 40 usually indicates major depression; a score below 28 indicates remission of depression.^8-10
• The Beck Depression Inventory (BDI), a 21-item self-report questionnaire for adults, has reasonable reliability and validity for adolescents. Its modest specificity suggests that it may measure general distress and dysphoria, which is not specific to depressive disorders. The language may be too difficult for some younger adolescents and those with poor reading comprehension skills.
• The Children’s Depression Inventory, a version of the BDI for prepubertal children, can be considered for adolescents whose cognitive and/or reading skills are less mature.
• Achenbach’s Child Behavior Checklists and other standardized questionnaires can screen for comorbid psychopathology.

Assessing psychosocial stress, such as conflicts with parents or peers, school problems, or risk-taking behavior, is also important. Depressed youth often have family members with histories of depression, alcoholism, anxiety, and other psychiatric diagnoses. History of sexual abuse has been linked to depression.³ The depressed adolescent’s impaired functioning in school and at home may cause secondary stress, increasing the burden of illness and need for treatment.

Suicide risk Although suicide remains rare among adolescents in general, the rate of suicide among this age group has risen dramatically over the past decade, particularly among younger teens and preteens. In 1997, suicide was the third leading cause of death in adolescents after accidental injuries and homicide.

Adolescents with depressive disorders are at increased risk for suicide, and boys are more likely than girls to attempt and complete suicide. It is therefore imperative to assess and document suicide risk for each adolescent who presents with depressive symptoms.

After establishing a rapport, the most effective screening is a straightforward conversation with the adolescent about suicidal ideation, intent, and behavior. Assess the social context of support and psychopathology in the family, availability and accessibility of lethal suicide methods (e.g., firearms in the home), and presence of events that could influence imitative suicidal behavior (e.g., a friend’s suicide).⁶

Treatment

Approaches to adolescent depression include (in increasing order of intensity and complexity) watchful monitoring, nonspecific supportive therapy, pharmacotherapy, specific psychotherapy (i.e., cognitive-behavioral or interpersonal therapy), and combined treatment (e.g., psychotherapy plus pharmacotherapy, adolescent psychotherapy plus family therapy).

There are no clear-cut guidelines as to whether pharmacologic or psychosocial therapy should be offered first.¹¹ In the community, patient and family preferences, past treatment response, and the clinician’s background and expertise influence the choice of treatment. As with adults, adolescents deemed at high risk for suicidal behavior must receive immediate attention from mental health professionals and must be monitored, usually in an inpatient setting.

Watchful monitoring means to wait and see if the youth improves spontaneously.

In some studies, nearly one-half (48%) of adolescents with depression were found to go into spontaneous remission within 8 weeks.¹² Watchful monitoring, however, would leave most patients still depressed, and no predictors of spontaneous remission have been identified.

Table 2

SSRI DOSAGES FOUND TO BE EFFECTIVE IN ADOLESCENTS WITH MDD

Because of the risks of suicide and social and academic impairment, monitoring alone is acceptable only for a few weeks, and only in cases where depression is mild and uncomplicated. In any case, “monitoring” requires that you periodically reassess the teen and be available for consultation between assessments.

Nonspecific supportive therapy Most psychotherapy provided in the community probably is nonspecific (i.e., not theoretically driven or conducted according to a treatment manual) and supportive (i.e., aimed at providing encouragement). This approach is known to be less effective than specific psychotherapies or antidepressant pharmacotherapy, but we have virtually no data comparing it with lack of treatment.

Nonspecific supportive therapy can be considered a reasonable first-step treatment for depressed teens without complicating risk factors.¹³ Specific treatment is indicated, however, if the adolescent does not improve in a few weeks.

Pharmacotherapy A few placebo-controlled clinical trials have studied the efficacy of selective serotonin reuptake inhibitors (SSRIs) in outpatient adolescents with major depression. It must be noted that practically all the available data relate to major depression, and no systematic studies have been done in dysthymia and other types of depression in this population.

The SSRIs fluoxetine,^8,9 citalopram,¹⁰ paroxetine,¹⁴ and sertraline¹⁵ can decrease symptoms of adolescent depression over 2 to 3 months when given at dosages similar to those used in adults (Table 2). At this time, there are no data that suggest the SSRI dosage must be different in younger (12-year-old) compared with older (18-year-old) adolescents, or in girls compared with boys.

The response rate (adolescents who were substantially improved at end of treatment) ranged from 52 to 65% with SSRI medication and 33 to 48% with a placebo. This means that one would need to treat about six adolescents in order to add one to those who would improve by taking a placebo. Thus, 6 is the number needed to treat (NNT), a common index used to make decisions in evidence-based medicine. As a comparison, the NNT is 1.5 for stimulant treatment of ADHD, indicating that stimulants are more effective in ADHD (i.e., the difference between an active drug and a placebo is greater) than SSRIs are in depression.

It must be noted that receiving a placebo in clinical trials of depression does not equal absence of treatment. Typically, research participants assigned to a placebo have weekly clinical contacts, so placebo treatment could be defined as “nonspecific clinical management.” In any case, the NNT for SSRIs in adolescent depression does not appear to be substantially different from that found in adults.

There is no evidence that tricyclic antidepressants, or TCAs, are more effective than a placebo in adolescents with depression. Considering TCAs’ side effects and potential cardiotoxicity in overdose, their use is not recommended in depressed adolescents. Other antidepressants, such as bupropion and venlafaxine, have been reported to be effective in open studies and clinical observations.

Antidepressants can trigger mania in persons predisposed to bipolar disorder. Before prescribing antidepressant medications, carefully evaluate depressed adolescents for present and past symptoms of mania or hypomania and check family history for possible bipolar disorder. During treatment, monitor the adolescent for signs that suggest emerging mania.

SSRIs and stimulant medications are commonly combined, as depression and ADHD often coexist. Clinical experience indicates that these medications can be safely combined in most cases, although no systematic studies of this combination have been performed in large, representative samples of teens. As with antidepressant monotherapy, adolescent patients must be monitored for signs of mania or hypomania.

Specific psychotherapy Among psychotherapies for depression, the best evidence of efficacy exists for cognitive-behavioral therapy (CBT), which attempts to correct cognitive distortions associated with depression and promotes healthy behaviors. In two clinical studies, CBT was more effective in depressed adolescents than no treatment (i.e., wait list), supportive therapy, or systemic behavior family therapy. About two-thirds of patients treated with CBT achieved remission from depressive symptoms.^16,17

Interpersonal therapy (IPT) has been adapted for use in depressed adolescents.¹⁸ Its efficacy is supported by one controlled study that found greater improvement after 12 weekly IPT sessions than after once-monthly clinical contact.¹⁹ There is no convincing evidence that family therapies are effective in treating depression in adolescents or add much to the benefit of CBT.¹⁷

Combined treatment Research on combined treatments for adolescent depression has been limited. One study found that adding family therapy to CBT does not improve the adolescent’s depressive symptoms.⁸ Although no data are available on the combination of SSRI medication and CBT (or IPT), the National Institute of Mental Health is funding two studies on this important issue:

• The Treatment of Adolescents with Depression Study (TADS) is comparing the efficacy of fluoxetine plus CBT with that of fluoxetine or CBT alone in adolescents with major depression. The study, coordinated by John March, MD, of Duke University Medical Center, is being conducted with 432 teen subjects at 13 clinical sites.
• The Treatment of Resistant Depression in Adolescents (TORDIA) study is comparing the efficacy of antidepressant medication plus CBT to that of antidepressants alone in adolescents who have not improved on initial treatment with an SSRI. TORDIA, coordinated by David Brent, MD, University of Pittsburgh, is being conducted at six clinical sites.

Related resources

• National Institute of Mental Health. Child and Adolescent Mental Health. http://www.nimh.nih.gov/publicat/childmenu.cfm
• Depression in children and adolescents: A fact sheet for physicians. http://www.nimh.nih.gov/publicat/depchildresfact.cfm
• NIMH-funded treatment studies in adolescent depression:

Drug brand names

• Bupropion • Wellbutrin
• Citalopram • Celexa
• Fluoxetine • Prozac
• Paroxetine • Paxil
• Sertraline • Zoloft
• Venlafaxine • Effexor

Disclosure

The authors report no affiliation or financial relationship with any of the companies whose products are mentioned in this article. The opinions and assertions contained in this article are the private views of the authors and are not to be construed as official or as reflecting the views of the Department of Health and Human Services or the National Institutes of Health.

Adolescents with depressive disorders tend to arrive in psychiatrists’ offices when their behavior has already been identified as problematic. Suicide attempts, academic failure, substance abuse, and family conflicts can all lead to teen psychiatric referrals. Other times, subtler changes in behavior may lead a family doctor or pediatrician to suspect depression and to send an adolescent to you for a psychiatric consultation.

The psychiatrist’s task is challenging. Adolescents are usually brought in by their worried parents and may not want to talk to a psychiatrist. Or they may be unable to accurately describe their internal states. Even people who know an adolescent well may not discern the emotions that drive his or her behavior. Adding to the mix are the recurrent nature of major depression in adolescents and the likelihood of complicating comorbid psychiatric conditions (Box).^1-6

Based on clinical evidence, we offer advice to help you promptly identify and effectively treat adolescents with depressive disorders. We also provide preliminary information on two studies examining medication treatment, psychotherapy, and combined treatment for teens with major depression.

Adolescent depression disorders

Symptoms of depression in adolescents are similar to those in adults, and it is appropriate for psychiatrists to use DSM-IV diagnostic criteria for making the diagnosis. The three primary depressive disorders for both adults and adolescents are major depressive disorder (MDD), dysthymic disorder, and depressive disorder not otherwise specified (NOS).

Box

Although the symptoms that make up the diagnostic criteria are similar for adults and teens, the behavioral manifestations and response to treatment may differ. The adolescent may present as irritable and angry, rather than overly sad. Impairments in functioning are likely to be related to decline in school performance, social withdrawal, or increased conflicts with peers and family.

As for treatment, certain antidepressant medications of proven efficacy in adults (i.e., tricyclics) do not seem to work for adolescents.

MDD is a time-limited episode of depressive symptoms severe enough to cause functional impairment, such as decline in school performance, social withdrawal, or increased conflicts with peers and family. Symptoms must be present at least 2 weeks.

Dysthymia is a chronic depression that is less severe than MDD and lasts 1 year or longer without sustained remission. It often begins early in childhood and may include periods of increased symptoms consistent with major depression (sometimes called “double depression”).

Depressive disorder NOS is a category of depression that, though clinically significant, does not meet the full criteria for severity, duration, or level of impairment of MDD or dysthymia.

Unless otherwise specified, the terms “depression” and “depressive disorder” in this article are used generically to include all three of these disorders.

Depressive disorders must be differentiated from bipolar disorder, which is characterized by least one prior episode of mania (for bipolar type I) or hypomania (for bipolar type II). The clinical picture of bipolar disorder in youths may differ from that seen in adults. For example, bipolar youth often present with dysphoric mood interspersed with frequent, short periods of intense emotional lability and irritability, rather than “classic” euphoria.

Diagnosis

To diagnose a depressive disorder in an adolescent, information is typically obtained from multiple sources, most commonly the teenage patient and at least one of the parents. Because several sources are involved, however, the information may be conflicting. For instance, the adolescent may contradict a parent’s report that he or she is having difficulties in school or has a substance abuse problem.

Interviewing skills and clinical judgment are required of the clinician in these situations. It is important to:

• obtain a complete description of the adolescent’s behavior and mood over time and as accurate a description as possible of when changes occurred
• assess comorbid conditions (particularly anxiety, attention-deficit/hyperactivity disorder [ADHD], conduct disorder, and substance abuse)
• differentiate between unipolar MDD and bipolar disorder
• evaluate the risk of suicide.

Table 1

DEPRESSION IN ADOLESCENTS AND ADULTS

Interviewing Standardized diagnostic interviews, such as the Schedule for Affective Disorders and Schizophrenia for children (K-SADS), are commonly used to research adolescent depression but require special training and approximately 1 to 2 hours to administer. As an alternative, clinicians generally develop their own “semi-structured interview” to try to collect all the relevant information required for an accurate diagnosis.⁷

The interview should be conducted with the adolescent and the parent(s), first separately for ease of disclosure then together to reconcile any differences in the information they report. Open-ended questions and time for building rapport may facilitate disclosure from a reticent adolescent. At times, however, one must make the diagnosis by relying more on reports from others who know the child well. Building a trusting therapeutic relationship then becomes part of ongoing treatment.

Standardized measures In addition to the interview, standardized self-report and other-report measures can help:

• The Child Depression Rating Scale-Revised, commonly used in clinical research, can also be used in practice to quantify symptom severity and document treatment response. A score above 40 usually indicates major depression; a score below 28 indicates remission of depression.^8-10
• The Beck Depression Inventory (BDI), a 21-item self-report questionnaire for adults, has reasonable reliability and validity for adolescents. Its modest specificity suggests that it may measure general distress and dysphoria, which is not specific to depressive disorders. The language may be too difficult for some younger adolescents and those with poor reading comprehension skills.
• The Children’s Depression Inventory, a version of the BDI for prepubertal children, can be considered for adolescents whose cognitive and/or reading skills are less mature.
• Achenbach’s Child Behavior Checklists and other standardized questionnaires can screen for comorbid psychopathology.

Assessing psychosocial stress, such as conflicts with parents or peers, school problems, or risk-taking behavior, is also important. Depressed youth often have family members with histories of depression, alcoholism, anxiety, and other psychiatric diagnoses. History of sexual abuse has been linked to depression.³ The depressed adolescent’s impaired functioning in school and at home may cause secondary stress, increasing the burden of illness and need for treatment.

Suicide risk Although suicide remains rare among adolescents in general, the rate of suicide among this age group has risen dramatically over the past decade, particularly among younger teens and preteens. In 1997, suicide was the third leading cause of death in adolescents after accidental injuries and homicide.

Adolescents with depressive disorders are at increased risk for suicide, and boys are more likely than girls to attempt and complete suicide. It is therefore imperative to assess and document suicide risk for each adolescent who presents with depressive symptoms.

After establishing a rapport, the most effective screening is a straightforward conversation with the adolescent about suicidal ideation, intent, and behavior. Assess the social context of support and psychopathology in the family, availability and accessibility of lethal suicide methods (e.g., firearms in the home), and presence of events that could influence imitative suicidal behavior (e.g., a friend’s suicide).⁶

Treatment

Approaches to adolescent depression include (in increasing order of intensity and complexity) watchful monitoring, nonspecific supportive therapy, pharmacotherapy, specific psychotherapy (i.e., cognitive-behavioral or interpersonal therapy), and combined treatment (e.g., psychotherapy plus pharmacotherapy, adolescent psychotherapy plus family therapy).

There are no clear-cut guidelines as to whether pharmacologic or psychosocial therapy should be offered first.¹¹ In the community, patient and family preferences, past treatment response, and the clinician’s background and expertise influence the choice of treatment. As with adults, adolescents deemed at high risk for suicidal behavior must receive immediate attention from mental health professionals and must be monitored, usually in an inpatient setting.

Watchful monitoring means to wait and see if the youth improves spontaneously.

In some studies, nearly one-half (48%) of adolescents with depression were found to go into spontaneous remission within 8 weeks.¹² Watchful monitoring, however, would leave most patients still depressed, and no predictors of spontaneous remission have been identified.

Table 2

SSRI DOSAGES FOUND TO BE EFFECTIVE IN ADOLESCENTS WITH MDD

Because of the risks of suicide and social and academic impairment, monitoring alone is acceptable only for a few weeks, and only in cases where depression is mild and uncomplicated. In any case, “monitoring” requires that you periodically reassess the teen and be available for consultation between assessments.

Nonspecific supportive therapy Most psychotherapy provided in the community probably is nonspecific (i.e., not theoretically driven or conducted according to a treatment manual) and supportive (i.e., aimed at providing encouragement). This approach is known to be less effective than specific psychotherapies or antidepressant pharmacotherapy, but we have virtually no data comparing it with lack of treatment.

Nonspecific supportive therapy can be considered a reasonable first-step treatment for depressed teens without complicating risk factors.¹³ Specific treatment is indicated, however, if the adolescent does not improve in a few weeks.

Pharmacotherapy A few placebo-controlled clinical trials have studied the efficacy of selective serotonin reuptake inhibitors (SSRIs) in outpatient adolescents with major depression. It must be noted that practically all the available data relate to major depression, and no systematic studies have been done in dysthymia and other types of depression in this population.

The SSRIs fluoxetine,^8,9 citalopram,¹⁰ paroxetine,¹⁴ and sertraline¹⁵ can decrease symptoms of adolescent depression over 2 to 3 months when given at dosages similar to those used in adults (Table 2). At this time, there are no data that suggest the SSRI dosage must be different in younger (12-year-old) compared with older (18-year-old) adolescents, or in girls compared with boys.

The response rate (adolescents who were substantially improved at end of treatment) ranged from 52 to 65% with SSRI medication and 33 to 48% with a placebo. This means that one would need to treat about six adolescents in order to add one to those who would improve by taking a placebo. Thus, 6 is the number needed to treat (NNT), a common index used to make decisions in evidence-based medicine. As a comparison, the NNT is 1.5 for stimulant treatment of ADHD, indicating that stimulants are more effective in ADHD (i.e., the difference between an active drug and a placebo is greater) than SSRIs are in depression.

It must be noted that receiving a placebo in clinical trials of depression does not equal absence of treatment. Typically, research participants assigned to a placebo have weekly clinical contacts, so placebo treatment could be defined as “nonspecific clinical management.” In any case, the NNT for SSRIs in adolescent depression does not appear to be substantially different from that found in adults.

There is no evidence that tricyclic antidepressants, or TCAs, are more effective than a placebo in adolescents with depression. Considering TCAs’ side effects and potential cardiotoxicity in overdose, their use is not recommended in depressed adolescents. Other antidepressants, such as bupropion and venlafaxine, have been reported to be effective in open studies and clinical observations.

Antidepressants can trigger mania in persons predisposed to bipolar disorder. Before prescribing antidepressant medications, carefully evaluate depressed adolescents for present and past symptoms of mania or hypomania and check family history for possible bipolar disorder. During treatment, monitor the adolescent for signs that suggest emerging mania.

SSRIs and stimulant medications are commonly combined, as depression and ADHD often coexist. Clinical experience indicates that these medications can be safely combined in most cases, although no systematic studies of this combination have been performed in large, representative samples of teens. As with antidepressant monotherapy, adolescent patients must be monitored for signs of mania or hypomania.

Specific psychotherapy Among psychotherapies for depression, the best evidence of efficacy exists for cognitive-behavioral therapy (CBT), which attempts to correct cognitive distortions associated with depression and promotes healthy behaviors. In two clinical studies, CBT was more effective in depressed adolescents than no treatment (i.e., wait list), supportive therapy, or systemic behavior family therapy. About two-thirds of patients treated with CBT achieved remission from depressive symptoms.^16,17

Interpersonal therapy (IPT) has been adapted for use in depressed adolescents.¹⁸ Its efficacy is supported by one controlled study that found greater improvement after 12 weekly IPT sessions than after once-monthly clinical contact.¹⁹ There is no convincing evidence that family therapies are effective in treating depression in adolescents or add much to the benefit of CBT.¹⁷

Combined treatment Research on combined treatments for adolescent depression has been limited. One study found that adding family therapy to CBT does not improve the adolescent’s depressive symptoms.⁸ Although no data are available on the combination of SSRI medication and CBT (or IPT), the National Institute of Mental Health is funding two studies on this important issue:

• The Treatment of Adolescents with Depression Study (TADS) is comparing the efficacy of fluoxetine plus CBT with that of fluoxetine or CBT alone in adolescents with major depression. The study, coordinated by John March, MD, of Duke University Medical Center, is being conducted with 432 teen subjects at 13 clinical sites.
• The Treatment of Resistant Depression in Adolescents (TORDIA) study is comparing the efficacy of antidepressant medication plus CBT to that of antidepressants alone in adolescents who have not improved on initial treatment with an SSRI. TORDIA, coordinated by David Brent, MD, University of Pittsburgh, is being conducted at six clinical sites.

Related resources

• National Institute of Mental Health. Child and Adolescent Mental Health. http://www.nimh.nih.gov/publicat/childmenu.cfm
• Depression in children and adolescents: A fact sheet for physicians. http://www.nimh.nih.gov/publicat/depchildresfact.cfm
• NIMH-funded treatment studies in adolescent depression:

Drug brand names

• Bupropion • Wellbutrin
• Citalopram • Celexa
• Fluoxetine • Prozac
• Paroxetine • Paxil
• Sertraline • Zoloft
• Venlafaxine • Effexor

Disclosure

The authors report no affiliation or financial relationship with any of the companies whose products are mentioned in this article. The opinions and assertions contained in this article are the private views of the authors and are not to be construed as official or as reflecting the views of the Department of Health and Human Services or the National Institutes of Health.

Recent practice guidelines generally do not position tricyclic antidepressants (TCAs) as first-line therapies because of concerns about side effects and safety issues.^1-3 Yet these agents have important clinical uses—both for approved and off-label indications—and diverse pharmacologic properties that distinguish them from each other as well as from many of the “newer antidepressants.”

Eleven TCAs are available in the United States (Table 1). Here’s what the evidence says about when and how to use them to treat a variety of psychiatric disorders, along with our recommendations on how to reduce the risk of side effects.

Indications and off-label uses

TCAs’ pharmacologic properties make them very versatile (Box 1).⁴ Major depression and anxiety disorders (such as obsessive compulsive disorder [OCD]) are the principal FDA-approved indications for TCAs. Other approved indications are anxiety (doxepin) and childhood enuresis (imipramine). Common off-label uses supported by scientific literature include panic disorder, social phobia, insomnia, posttraumatic stress disorder (PTSD), generalized anxiety disorder (GAD), attention-deficit/hyperactivity disorder (ADHD), migraine headache, chronic pain syndromes, premature ejaculation, substance abuse disorders, and eating disorders, to name a few.

Major depression To treat major depressive disorder, the American Psychiatric Association recommends selective serotonin reuptake inhibitors (SSRIs), bupropion, venlafaxine, and the secondary amine TCAs desipramine and nortriptyline as “optimal” first-line therapy for most patients.¹ The Texas Medication Algorithm Project recommends SSRIs, bupropion, nefazodone, venlafaxine, or mirtazapine as first-line therapy and lists TCAs as second- or third-line therapy for patients with partial or no response to initial therapy.²

Table 1

THERAPEUTIC DOSAGE RANGE

Box 1

Although recommended as first-line therapies, SSRIs and other newer antidepressants exhibit no greater efficacy than TCAs in treating major depression.^5,6 Three major meta-analyses^7-9 reported no significant difference in efficacy between the two antidepressant classes. In the largest,⁹ published by the U.S. Department of Health and Human Services, 50% of inpatients and 52% of outpatients responded to TCAs, whereas 54% of inpatients and 47% of outpatients responded to SSRIs. Compared with SSRIs, TCAs also may be associated with higher rates of remission.^7,8

Clomipramine—approved for OCD—has been used for decades to treat depression and resistant depression. Two meta-analyses by the Danish University Antidepressant Group^7,8 showed that clomipramine produced a “significantly better therapeutic effect” compared with citalopram and paroxetine. In three major studies,^5,7,8 TCAs showed greater effectiveness than SSRIs in treating melancholic depression.

The anticholinergic side effects of TCAs have been perceived to cause higher patient dropout and discontinuation rates, but studies have shown mixed results. In one meta-analysis comparing SSRIs and TCAs, the difference in dropout rates due to adverse effects was less significant than previously reported. When total dropout rates for any reason were examined, the difference was less than was originally expected.¹⁰

OCD In clinical practice, most patients with OCD are started on an SSRI. Clomipramine is another valuable option, however, especially for patients who fail one or more therapeutic trials with SSRIs. Clomipramine may produce significant therapeutic benefit in patients with OCD, possibly because of its potent 5-HT reuptake properties. In one study, patients with OCD symptoms who received clomipramine improved more than those receiving SSRIs when each class was compared with placebo.¹¹

Panic disorder Although not approved for panic disorder, imipramine and desipramine provide effective treatment, even in nondepressed patients.¹² Doses and plasma levels are the same as those used for treating depression. Start low (e.g., imipramine, 10 to 20 mg/d; desipramine, 10 to 25 mg/d) and increase gradually over several weeks to typical therapeutic dosages (Table 1). Do not escalate too rapidly, as this may increase anxiety or precipitate a panic attack. Uses in children TCAs have been used to treat children with ADHD, OCD, enuresis, and depression. The American Academy of Child and Adolescent Psychiatry (AACAP) does not recommend TCAs as first-line treatment for youths requiring pharmacotherapy for depressive disorders but acknowledges that some youths with depression may respond better to TCAs than to other medications.³

Imipramine is the only TCA indicated for nocturnal enuresis, but its exact mechanism of action is not known. The benefit may be secondary to imipramine’s anticholinergic effect or to changes in sleep architecture. Recommended bedtime doses for children with enuresis are:

• 25 to 50 mg under age 12;
• up to 75 mg age 12 and older.

TCAs are an option for children with ADHD, especially if ADHD is present with comorbid depression or anxiety disorder. However, because at least four cases of sudden cardiac death have been reported in children taking desipramine, it is prudent to monitor cardiac function when children are started on TCAs. AACAP guidelines recommend obtaining a baseline ECG, resting blood pressure, and pulse (supine or sitting, then standing), with regular monitoring of the child’s weight during TCA therapy.

Combination therapy Few controlled studies have tested TCAs as combination therapy. Trazodone can be used as an adjunct to SSRIs and monoamine oxidase inhibitors (MAOIs) for patients with insomnia. TCAs are used as an adjunct to treat resistant depression and OCD^13,14 (e.g., clomipramine added to fluvoxamine to treat OCD).¹⁵ Serum level monitoring is recommended when TCAs are used as adjuncts.

Other uses TCAs play a role in the prophylactic treatment of premature ejaculation and migraine headaches, probably because of their serotonergic (5HT2) effect. In patients with migraine and depression, a trial of a TCA such as amitriptyline may be warranted.

TCAs are widely used to manage neuropathic pain,¹⁶ although the exact mechanism of action is unknown. Because depression is commonly associated with pain, the effect may result from the agents’ action on depression, or TCAs may possess direct analgesic action.

Trazodone, 25 to 100 mg at bedtime, is widely used for insomnia, alone or as an adjunct with other classes of antidepressants—even in combination with MAOIs. Nortriptyline has been shown to be safe and effective for post-stroke and geriatric types of depression.¹⁷

TCAs are also used for a host of other medical, psychiatric, and neurologic problems, such as social phobia, PTSD, GAD, substance abuse disorders, and eating disorders.¹⁸ Only a few controlled studies have tested TCAs for these indications.

Side effects

Although TCAs have shown efficacy in many clinical situations, their use is associated with potentially serious side effects, which may include anticholinergic effects, sedation, weight gain, CNS toxicity, orthostatic hypotension, cardiovascular toxicity, delirium, and risk of suicide by overdose. The risk of side effects can be reduced with careful prescribing practices (Box 2).

Anticholinergic effects, sedation TCAs vary in their anticholinergic activity (Table 2). Tertiary amine TCAs such as amitriptyline and protriptyline may cause dry mouth, constipation, urinary hesitancy, and blurred vision in some patients, and confusion in elderly or demented patients. Secondary amine TCAs such as nortriptyline or desipramine are less anticholinergic and less likely to cause these side effects.

Peripheral anticholinergic side effects can be managed with bethanacol—a cholinergic drug—in dosages of 25 to 50 mg tid or qid. Dry mouth can be treated with pilocarpine, 5 mg bid to qid, or oral bethanacol (5- to 10-mg tablets sublingually), artificial saliva drops, sugarless candy/gum, or mouthwash.

Box 2

Table 2

BIOCHEMICAL EFFECTS OF TRICYCLICS AND OTHER ANTIDEPRESSANTS

Table 3

APPROXIMATE THERAPEUTIC PLASMA LEVEL RANGES

Sedation is a common side effect caused by the antihistaminic properties of some TCAs (Table 2). Agents with this effect (e.g., doxepin), when given once daily at bedtime, can benefit patients with concomitant sleep disturbance.

Weight gain Patients being treated with TCAs often gain weight, most likely because of carbohydrate craving associated with H2 blockade. Patient education, monitoring of weight, and dietary counseling may be necessary during TCA use. Regular exercise is recommended, although depressed patients often lack the motivation and energy to exercise.

Cardiovascular toxicity All TCAs have potential cardiovascular effects (Table 3), which are seen on an ECG as increased PR, QRS, or QTc intervals, especially at higher dosages or in patients with pre-existing cardiac disease.¹⁹ Orthostatic hypotension is among the most common cardiovascular side effects, particularly in the elderly, and may result in falls or other injuries.

A baseline ECG is recommended before starting TCA therapy, especially in depressed patients with cardiac conduction delays (primary or secondary to concomitant medications). Routine ECGs should be used to monitor patients:

• with pre-existing cardiac disease;
• taking higher-than-recommended dosages of a TCA;
• taking other medications that may affect cardiac conduction.

When in doubt, obtaining a cardiology consultation is recommended.

CNS effects Signs of CNS toxicity include confusion, memory impairment, delirium, seizures, coma, and eventual respiratory depression.²⁰ Risk factors include toxic TCA plasma levels, elderly patient age, and concomitant use of other medications, such as psychotropics (neuroleptics), anticholinergics, and antihistamines. CNS toxicity may be difficult to diagnose, as it may initially resemble worsening of depressed mood. Confusion or worsening of memory or cognitive function are predictors of CNS toxicity.

A patient with anticholinergic-related delirium should be monitored on a medical unit. A trial of physostigmine may be warranted to confirm the diagnosis.

Neurologic symptoms Maprotiline and clomipramine have been associated with an increased risk of seizures. Use reduced dosages in patients with a history of seizures or concomitant use of medications that may increase maprotiline or clomipramine levels or decrease the seizure threshold (e.g., other antidepressants, withdrawal from benzodiazepines). Amoxapine has been associated with extrapyramidal symptoms secondary to a neuroleptic metabolite.

Overdose The severity of adverse side effects often depends upon drug concentration. An overdose of as little as a 2-week supply of a TCA can cause potentially fatal arrhythmias.

Priapism Trazodone has been associated with priapism, which has an incidence rate of 1/6,000 in patients taking this drug. Priapism is a medical emergency that requires prompt treatment.

Administration

TCAs are well-absorbed following oral administration and reach peak plasma levels in 2 to 6 hours. The average half-life of approximately 24 hours often allows therapeutic levels to be achieved with once-daily dosing at bedtime (Table 3).

TCAs are metabolized primarily by the cytochrome P-450 2D6 (CYP2D6) isoenzyme. Patients differ by a factor of 30- to 40-fold in their rate of metabolism of some TCAs, depending on individual genotypes.

The 7 to 9% of Caucasians classified as poor metabolizers at CYP2D6 require much lower-than-usual dosages of secondary amine TCAs.²¹ he 50% of Asians who are intermediate metabolizers require one-half the usual dosage. Tailor therapy to the individual patient, starting low, checking plasma levels, and monitoring for side effects.

For patients who would benefit from TCAs’ H1 anxiolytic effects, start with bid or tid dosing and later switch to once daily after achieving efficacy.

When using TCAs, start low and go slow to maximize efficacy and minimize side effects, especially in the elderly patient. The goal is to treat the patient, not to achieve target serum levels. The most favorable results have been demonstrated when patients are maintained at the dosages to which they respond when their disorder is in the acute stage.¹⁷

Symptoms usually improve after 2 weeks of TCA therapy, and up to 6 weeks may be required for a clinically significant effect. For many depressed patients, symptoms of insomnia, anxiety, and poor appetite improve within the first few days.

Monitoring serum levels

Serum levels are useful for monitoring treatment, compliance, and toxicity.²² Because the most accurate TCA serum levels are found at steady state, check the level after the patient has been on TCA therapy for at least 5 days and 8 to 12 hours after the last dose. Serum levels may also guide dosage increases in patients who exhibit a partial response to therapy.

Table 4

DRUGS THAT INTERACT WITH TCAs

Nortriptyline has a therapeutic window between 50 and 150 ng/mL. Patients who do not respond to serum levels higher than 150 ng/mL may respond when the dosage is reduced and the serum level falls to within that range. If response is adequate and without side effects, however, there is no need to reduce the dose. Other, less clear therapeutic windows have been described for other TCAs (Table 3).

TCA metabolism is affected by age and interaction with other drugs. Closely monitor serum levels in elderly and medically compromised patients, especially during dosage increases. Pay particular attention to other medications that may affect serum levels²³ (Table 4). SSRIs, particularly fluoxetine and paroxetine, may affect TCA plasma levels because of their potent inhibition of CYP2D6.

Individualize dosing and serum level monitoring. For example, the patient who is clinically tolerating a TCA and has a normal ECG does not require a serum level measurement, unless medically indicated.

Discontinuing tricyclics

When discontinuing a TCA, taper the dosage no more rapidly than 25 to 50 mg every 2 to 3 days. Abrupt discontinuation can cause cholinergic rebound, with symptoms such as nausea, cramping, headache, vomiting, and sweating. “Rebound hypomania” or “mania” have been reported with abrupt cessation of TCAs,²⁴ especially in patients with bipolar disorder. If the etiology of rebound symptoms is unclear (i.e., medical versus psychiatric), re-administering the discontinued TCA should relieve the symptoms and confirm a diagnosis of discontinuation.

Related resources

• Schatzberg AF, Cole JO, DeBattista C. Antidepressants. Manual of clinical psychopharmacology (3rd ed). Washington, DC: American Psychiatric Press, 1997.
• Janicak PG, Davis JM, Preskorn SH, Ayd Jr. FJ. Principles and practice of psychopharmacotherapy (2nd ed). Baltimore: Lippincott Williams & Wilkins, 1997.
• PDR Psychotropic Prescribing Guide (2nd ed). Montvale, NJ: Medical Economics, 1999.
• National Library of Medicine, MedlinePlus Health Information: Antidepressants, tricyclics (http://www.nlm.nih.gov/medlineplus/druginfo/antidepressantstricyclicsystem202055.html)

Drug brand names

See table 1 for tricyclic drug brand names. Others mentioned in this article include:

• Bupropion • Wellbutrin
• Cimetidine • Tagamet
• Citalopram • Celexa
• Disulfiram • Antabuse
• Fluoxetine • Prozac
• Fluvoxamine • Luvox
• Guanethidine • Ismelin
• Methylphenidate • Concerta, Ritalin
• Mirtazapine • Remeron
• Nefazodone • Serzone
• Paroxetine • Paxil
• Phenytoin • Dilantin

Disclosure

Dr. Zajecka reports that he receives grant/research support from, serves as a consultant to, and is on the speaker’s bureau of Bristol-Myers Squibb and Eli Lilly and Co.; receives grant/research support from Cephalon Inc., GlaxoSmithKline, Lichtwer Pharma, Merck and Co., MIICRO Inc., Otsuka Pharmaceuticals, Parke-Davis, Pfizer Inc., and Wyeth-Ayerst Pharmaceuticals; serves as a consultant to Abbott Laboratories; and is on the speaker’s bureau of Abbott Laboratories, Pfizer/Roerig, GlaxoSmithKline, Pharmacia, and Wyeth-Ayerst Pharmaceuticals.

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

Recent practice guidelines generally do not position tricyclic antidepressants (TCAs) as first-line therapies because of concerns about side effects and safety issues.^1-3 Yet these agents have important clinical uses—both for approved and off-label indications—and diverse pharmacologic properties that distinguish them from each other as well as from many of the “newer antidepressants.”

Eleven TCAs are available in the United States (Table 1). Here’s what the evidence says about when and how to use them to treat a variety of psychiatric disorders, along with our recommendations on how to reduce the risk of side effects.

Indications and off-label uses

TCAs’ pharmacologic properties make them very versatile (Box 1).⁴ Major depression and anxiety disorders (such as obsessive compulsive disorder [OCD]) are the principal FDA-approved indications for TCAs. Other approved indications are anxiety (doxepin) and childhood enuresis (imipramine). Common off-label uses supported by scientific literature include panic disorder, social phobia, insomnia, posttraumatic stress disorder (PTSD), generalized anxiety disorder (GAD), attention-deficit/hyperactivity disorder (ADHD), migraine headache, chronic pain syndromes, premature ejaculation, substance abuse disorders, and eating disorders, to name a few.

Major depression To treat major depressive disorder, the American Psychiatric Association recommends selective serotonin reuptake inhibitors (SSRIs), bupropion, venlafaxine, and the secondary amine TCAs desipramine and nortriptyline as “optimal” first-line therapy for most patients.¹ The Texas Medication Algorithm Project recommends SSRIs, bupropion, nefazodone, venlafaxine, or mirtazapine as first-line therapy and lists TCAs as second- or third-line therapy for patients with partial or no response to initial therapy.²

Table 1

THERAPEUTIC DOSAGE RANGE

Box 1

Although recommended as first-line therapies, SSRIs and other newer antidepressants exhibit no greater efficacy than TCAs in treating major depression.^5,6 Three major meta-analyses^7-9 reported no significant difference in efficacy between the two antidepressant classes. In the largest,⁹ published by the U.S. Department of Health and Human Services, 50% of inpatients and 52% of outpatients responded to TCAs, whereas 54% of inpatients and 47% of outpatients responded to SSRIs. Compared with SSRIs, TCAs also may be associated with higher rates of remission.^7,8

Clomipramine—approved for OCD—has been used for decades to treat depression and resistant depression. Two meta-analyses by the Danish University Antidepressant Group^7,8 showed that clomipramine produced a “significantly better therapeutic effect” compared with citalopram and paroxetine. In three major studies,^5,7,8 TCAs showed greater effectiveness than SSRIs in treating melancholic depression.

The anticholinergic side effects of TCAs have been perceived to cause higher patient dropout and discontinuation rates, but studies have shown mixed results. In one meta-analysis comparing SSRIs and TCAs, the difference in dropout rates due to adverse effects was less significant than previously reported. When total dropout rates for any reason were examined, the difference was less than was originally expected.¹⁰

OCD In clinical practice, most patients with OCD are started on an SSRI. Clomipramine is another valuable option, however, especially for patients who fail one or more therapeutic trials with SSRIs. Clomipramine may produce significant therapeutic benefit in patients with OCD, possibly because of its potent 5-HT reuptake properties. In one study, patients with OCD symptoms who received clomipramine improved more than those receiving SSRIs when each class was compared with placebo.¹¹

Panic disorder Although not approved for panic disorder, imipramine and desipramine provide effective treatment, even in nondepressed patients.¹² Doses and plasma levels are the same as those used for treating depression. Start low (e.g., imipramine, 10 to 20 mg/d; desipramine, 10 to 25 mg/d) and increase gradually over several weeks to typical therapeutic dosages (Table 1). Do not escalate too rapidly, as this may increase anxiety or precipitate a panic attack. Uses in children TCAs have been used to treat children with ADHD, OCD, enuresis, and depression. The American Academy of Child and Adolescent Psychiatry (AACAP) does not recommend TCAs as first-line treatment for youths requiring pharmacotherapy for depressive disorders but acknowledges that some youths with depression may respond better to TCAs than to other medications.³

Imipramine is the only TCA indicated for nocturnal enuresis, but its exact mechanism of action is not known. The benefit may be secondary to imipramine’s anticholinergic effect or to changes in sleep architecture. Recommended bedtime doses for children with enuresis are:

• 25 to 50 mg under age 12;
• up to 75 mg age 12 and older.

TCAs are an option for children with ADHD, especially if ADHD is present with comorbid depression or anxiety disorder. However, because at least four cases of sudden cardiac death have been reported in children taking desipramine, it is prudent to monitor cardiac function when children are started on TCAs. AACAP guidelines recommend obtaining a baseline ECG, resting blood pressure, and pulse (supine or sitting, then standing), with regular monitoring of the child’s weight during TCA therapy.

Combination therapy Few controlled studies have tested TCAs as combination therapy. Trazodone can be used as an adjunct to SSRIs and monoamine oxidase inhibitors (MAOIs) for patients with insomnia. TCAs are used as an adjunct to treat resistant depression and OCD^13,14 (e.g., clomipramine added to fluvoxamine to treat OCD).¹⁵ Serum level monitoring is recommended when TCAs are used as adjuncts.

Other uses TCAs play a role in the prophylactic treatment of premature ejaculation and migraine headaches, probably because of their serotonergic (5HT2) effect. In patients with migraine and depression, a trial of a TCA such as amitriptyline may be warranted.

TCAs are widely used to manage neuropathic pain,¹⁶ although the exact mechanism of action is unknown. Because depression is commonly associated with pain, the effect may result from the agents’ action on depression, or TCAs may possess direct analgesic action.

Trazodone, 25 to 100 mg at bedtime, is widely used for insomnia, alone or as an adjunct with other classes of antidepressants—even in combination with MAOIs. Nortriptyline has been shown to be safe and effective for post-stroke and geriatric types of depression.¹⁷

TCAs are also used for a host of other medical, psychiatric, and neurologic problems, such as social phobia, PTSD, GAD, substance abuse disorders, and eating disorders.¹⁸ Only a few controlled studies have tested TCAs for these indications.

Side effects

Although TCAs have shown efficacy in many clinical situations, their use is associated with potentially serious side effects, which may include anticholinergic effects, sedation, weight gain, CNS toxicity, orthostatic hypotension, cardiovascular toxicity, delirium, and risk of suicide by overdose. The risk of side effects can be reduced with careful prescribing practices (Box 2).

Anticholinergic effects, sedation TCAs vary in their anticholinergic activity (Table 2). Tertiary amine TCAs such as amitriptyline and protriptyline may cause dry mouth, constipation, urinary hesitancy, and blurred vision in some patients, and confusion in elderly or demented patients. Secondary amine TCAs such as nortriptyline or desipramine are less anticholinergic and less likely to cause these side effects.

Peripheral anticholinergic side effects can be managed with bethanacol—a cholinergic drug—in dosages of 25 to 50 mg tid or qid. Dry mouth can be treated with pilocarpine, 5 mg bid to qid, or oral bethanacol (5- to 10-mg tablets sublingually), artificial saliva drops, sugarless candy/gum, or mouthwash.

Box 2

Table 2

BIOCHEMICAL EFFECTS OF TRICYCLICS AND OTHER ANTIDEPRESSANTS

Table 3

APPROXIMATE THERAPEUTIC PLASMA LEVEL RANGES

Sedation is a common side effect caused by the antihistaminic properties of some TCAs (Table 2). Agents with this effect (e.g., doxepin), when given once daily at bedtime, can benefit patients with concomitant sleep disturbance.

Weight gain Patients being treated with TCAs often gain weight, most likely because of carbohydrate craving associated with H2 blockade. Patient education, monitoring of weight, and dietary counseling may be necessary during TCA use. Regular exercise is recommended, although depressed patients often lack the motivation and energy to exercise.

Cardiovascular toxicity All TCAs have potential cardiovascular effects (Table 3), which are seen on an ECG as increased PR, QRS, or QTc intervals, especially at higher dosages or in patients with pre-existing cardiac disease.¹⁹ Orthostatic hypotension is among the most common cardiovascular side effects, particularly in the elderly, and may result in falls or other injuries.

A baseline ECG is recommended before starting TCA therapy, especially in depressed patients with cardiac conduction delays (primary or secondary to concomitant medications). Routine ECGs should be used to monitor patients:

• with pre-existing cardiac disease;
• taking higher-than-recommended dosages of a TCA;
• taking other medications that may affect cardiac conduction.

When in doubt, obtaining a cardiology consultation is recommended.

CNS effects Signs of CNS toxicity include confusion, memory impairment, delirium, seizures, coma, and eventual respiratory depression.²⁰ Risk factors include toxic TCA plasma levels, elderly patient age, and concomitant use of other medications, such as psychotropics (neuroleptics), anticholinergics, and antihistamines. CNS toxicity may be difficult to diagnose, as it may initially resemble worsening of depressed mood. Confusion or worsening of memory or cognitive function are predictors of CNS toxicity.

A patient with anticholinergic-related delirium should be monitored on a medical unit. A trial of physostigmine may be warranted to confirm the diagnosis.

Neurologic symptoms Maprotiline and clomipramine have been associated with an increased risk of seizures. Use reduced dosages in patients with a history of seizures or concomitant use of medications that may increase maprotiline or clomipramine levels or decrease the seizure threshold (e.g., other antidepressants, withdrawal from benzodiazepines). Amoxapine has been associated with extrapyramidal symptoms secondary to a neuroleptic metabolite.

Overdose The severity of adverse side effects often depends upon drug concentration. An overdose of as little as a 2-week supply of a TCA can cause potentially fatal arrhythmias.

Priapism Trazodone has been associated with priapism, which has an incidence rate of 1/6,000 in patients taking this drug. Priapism is a medical emergency that requires prompt treatment.

Administration

TCAs are well-absorbed following oral administration and reach peak plasma levels in 2 to 6 hours. The average half-life of approximately 24 hours often allows therapeutic levels to be achieved with once-daily dosing at bedtime (Table 3).

TCAs are metabolized primarily by the cytochrome P-450 2D6 (CYP2D6) isoenzyme. Patients differ by a factor of 30- to 40-fold in their rate of metabolism of some TCAs, depending on individual genotypes.

The 7 to 9% of Caucasians classified as poor metabolizers at CYP2D6 require much lower-than-usual dosages of secondary amine TCAs.²¹ he 50% of Asians who are intermediate metabolizers require one-half the usual dosage. Tailor therapy to the individual patient, starting low, checking plasma levels, and monitoring for side effects.

For patients who would benefit from TCAs’ H1 anxiolytic effects, start with bid or tid dosing and later switch to once daily after achieving efficacy.

When using TCAs, start low and go slow to maximize efficacy and minimize side effects, especially in the elderly patient. The goal is to treat the patient, not to achieve target serum levels. The most favorable results have been demonstrated when patients are maintained at the dosages to which they respond when their disorder is in the acute stage.¹⁷

Symptoms usually improve after 2 weeks of TCA therapy, and up to 6 weeks may be required for a clinically significant effect. For many depressed patients, symptoms of insomnia, anxiety, and poor appetite improve within the first few days.

Monitoring serum levels

Serum levels are useful for monitoring treatment, compliance, and toxicity.²² Because the most accurate TCA serum levels are found at steady state, check the level after the patient has been on TCA therapy for at least 5 days and 8 to 12 hours after the last dose. Serum levels may also guide dosage increases in patients who exhibit a partial response to therapy.

Table 4

DRUGS THAT INTERACT WITH TCAs

Nortriptyline has a therapeutic window between 50 and 150 ng/mL. Patients who do not respond to serum levels higher than 150 ng/mL may respond when the dosage is reduced and the serum level falls to within that range. If response is adequate and without side effects, however, there is no need to reduce the dose. Other, less clear therapeutic windows have been described for other TCAs (Table 3).

TCA metabolism is affected by age and interaction with other drugs. Closely monitor serum levels in elderly and medically compromised patients, especially during dosage increases. Pay particular attention to other medications that may affect serum levels²³ (Table 4). SSRIs, particularly fluoxetine and paroxetine, may affect TCA plasma levels because of their potent inhibition of CYP2D6.

Individualize dosing and serum level monitoring. For example, the patient who is clinically tolerating a TCA and has a normal ECG does not require a serum level measurement, unless medically indicated.

Discontinuing tricyclics

When discontinuing a TCA, taper the dosage no more rapidly than 25 to 50 mg every 2 to 3 days. Abrupt discontinuation can cause cholinergic rebound, with symptoms such as nausea, cramping, headache, vomiting, and sweating. “Rebound hypomania” or “mania” have been reported with abrupt cessation of TCAs,²⁴ especially in patients with bipolar disorder. If the etiology of rebound symptoms is unclear (i.e., medical versus psychiatric), re-administering the discontinued TCA should relieve the symptoms and confirm a diagnosis of discontinuation.

Related resources

• Schatzberg AF, Cole JO, DeBattista C. Antidepressants. Manual of clinical psychopharmacology (3rd ed). Washington, DC: American Psychiatric Press, 1997.
• Janicak PG, Davis JM, Preskorn SH, Ayd Jr. FJ. Principles and practice of psychopharmacotherapy (2nd ed). Baltimore: Lippincott Williams & Wilkins, 1997.
• PDR Psychotropic Prescribing Guide (2nd ed). Montvale, NJ: Medical Economics, 1999.
• National Library of Medicine, MedlinePlus Health Information: Antidepressants, tricyclics (http://www.nlm.nih.gov/medlineplus/druginfo/antidepressantstricyclicsystem202055.html)

Drug brand names

See table 1 for tricyclic drug brand names. Others mentioned in this article include:

• Bupropion • Wellbutrin
• Cimetidine • Tagamet
• Citalopram • Celexa
• Disulfiram • Antabuse
• Fluoxetine • Prozac
• Fluvoxamine • Luvox
• Guanethidine • Ismelin
• Methylphenidate • Concerta, Ritalin
• Mirtazapine • Remeron
• Nefazodone • Serzone
• Paroxetine • Paxil
• Phenytoin • Dilantin

Disclosure

Dr. Zajecka reports that he receives grant/research support from, serves as a consultant to, and is on the speaker’s bureau of Bristol-Myers Squibb and Eli Lilly and Co.; receives grant/research support from Cephalon Inc., GlaxoSmithKline, Lichtwer Pharma, Merck and Co., MIICRO Inc., Otsuka Pharmaceuticals, Parke-Davis, Pfizer Inc., and Wyeth-Ayerst Pharmaceuticals; serves as a consultant to Abbott Laboratories; and is on the speaker’s bureau of Abbott Laboratories, Pfizer/Roerig, GlaxoSmithKline, Pharmacia, and Wyeth-Ayerst Pharmaceuticals.

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

Recent practice guidelines generally do not position tricyclic antidepressants (TCAs) as first-line therapies because of concerns about side effects and safety issues.^1-3 Yet these agents have important clinical uses—both for approved and off-label indications—and diverse pharmacologic properties that distinguish them from each other as well as from many of the “newer antidepressants.”

Eleven TCAs are available in the United States (Table 1). Here’s what the evidence says about when and how to use them to treat a variety of psychiatric disorders, along with our recommendations on how to reduce the risk of side effects.

Indications and off-label uses

TCAs’ pharmacologic properties make them very versatile (Box 1).⁴ Major depression and anxiety disorders (such as obsessive compulsive disorder [OCD]) are the principal FDA-approved indications for TCAs. Other approved indications are anxiety (doxepin) and childhood enuresis (imipramine). Common off-label uses supported by scientific literature include panic disorder, social phobia, insomnia, posttraumatic stress disorder (PTSD), generalized anxiety disorder (GAD), attention-deficit/hyperactivity disorder (ADHD), migraine headache, chronic pain syndromes, premature ejaculation, substance abuse disorders, and eating disorders, to name a few.

Major depression To treat major depressive disorder, the American Psychiatric Association recommends selective serotonin reuptake inhibitors (SSRIs), bupropion, venlafaxine, and the secondary amine TCAs desipramine and nortriptyline as “optimal” first-line therapy for most patients.¹ The Texas Medication Algorithm Project recommends SSRIs, bupropion, nefazodone, venlafaxine, or mirtazapine as first-line therapy and lists TCAs as second- or third-line therapy for patients with partial or no response to initial therapy.²

Table 1

THERAPEUTIC DOSAGE RANGE

Box 1

Although recommended as first-line therapies, SSRIs and other newer antidepressants exhibit no greater efficacy than TCAs in treating major depression.^5,6 Three major meta-analyses^7-9 reported no significant difference in efficacy between the two antidepressant classes. In the largest,⁹ published by the U.S. Department of Health and Human Services, 50% of inpatients and 52% of outpatients responded to TCAs, whereas 54% of inpatients and 47% of outpatients responded to SSRIs. Compared with SSRIs, TCAs also may be associated with higher rates of remission.^7,8

Clomipramine—approved for OCD—has been used for decades to treat depression and resistant depression. Two meta-analyses by the Danish University Antidepressant Group^7,8 showed that clomipramine produced a “significantly better therapeutic effect” compared with citalopram and paroxetine. In three major studies,^5,7,8 TCAs showed greater effectiveness than SSRIs in treating melancholic depression.

The anticholinergic side effects of TCAs have been perceived to cause higher patient dropout and discontinuation rates, but studies have shown mixed results. In one meta-analysis comparing SSRIs and TCAs, the difference in dropout rates due to adverse effects was less significant than previously reported. When total dropout rates for any reason were examined, the difference was less than was originally expected.¹⁰

OCD In clinical practice, most patients with OCD are started on an SSRI. Clomipramine is another valuable option, however, especially for patients who fail one or more therapeutic trials with SSRIs. Clomipramine may produce significant therapeutic benefit in patients with OCD, possibly because of its potent 5-HT reuptake properties. In one study, patients with OCD symptoms who received clomipramine improved more than those receiving SSRIs when each class was compared with placebo.¹¹

Panic disorder Although not approved for panic disorder, imipramine and desipramine provide effective treatment, even in nondepressed patients.¹² Doses and plasma levels are the same as those used for treating depression. Start low (e.g., imipramine, 10 to 20 mg/d; desipramine, 10 to 25 mg/d) and increase gradually over several weeks to typical therapeutic dosages (Table 1). Do not escalate too rapidly, as this may increase anxiety or precipitate a panic attack. Uses in children TCAs have been used to treat children with ADHD, OCD, enuresis, and depression. The American Academy of Child and Adolescent Psychiatry (AACAP) does not recommend TCAs as first-line treatment for youths requiring pharmacotherapy for depressive disorders but acknowledges that some youths with depression may respond better to TCAs than to other medications.³

Imipramine is the only TCA indicated for nocturnal enuresis, but its exact mechanism of action is not known. The benefit may be secondary to imipramine’s anticholinergic effect or to changes in sleep architecture. Recommended bedtime doses for children with enuresis are:

• 25 to 50 mg under age 12;
• up to 75 mg age 12 and older.

TCAs are an option for children with ADHD, especially if ADHD is present with comorbid depression or anxiety disorder. However, because at least four cases of sudden cardiac death have been reported in children taking desipramine, it is prudent to monitor cardiac function when children are started on TCAs. AACAP guidelines recommend obtaining a baseline ECG, resting blood pressure, and pulse (supine or sitting, then standing), with regular monitoring of the child’s weight during TCA therapy.

Combination therapy Few controlled studies have tested TCAs as combination therapy. Trazodone can be used as an adjunct to SSRIs and monoamine oxidase inhibitors (MAOIs) for patients with insomnia. TCAs are used as an adjunct to treat resistant depression and OCD^13,14 (e.g., clomipramine added to fluvoxamine to treat OCD).¹⁵ Serum level monitoring is recommended when TCAs are used as adjuncts.

Other uses TCAs play a role in the prophylactic treatment of premature ejaculation and migraine headaches, probably because of their serotonergic (5HT2) effect. In patients with migraine and depression, a trial of a TCA such as amitriptyline may be warranted.

TCAs are widely used to manage neuropathic pain,¹⁶ although the exact mechanism of action is unknown. Because depression is commonly associated with pain, the effect may result from the agents’ action on depression, or TCAs may possess direct analgesic action.

Trazodone, 25 to 100 mg at bedtime, is widely used for insomnia, alone or as an adjunct with other classes of antidepressants—even in combination with MAOIs. Nortriptyline has been shown to be safe and effective for post-stroke and geriatric types of depression.¹⁷

TCAs are also used for a host of other medical, psychiatric, and neurologic problems, such as social phobia, PTSD, GAD, substance abuse disorders, and eating disorders.¹⁸ Only a few controlled studies have tested TCAs for these indications.

Side effects

Although TCAs have shown efficacy in many clinical situations, their use is associated with potentially serious side effects, which may include anticholinergic effects, sedation, weight gain, CNS toxicity, orthostatic hypotension, cardiovascular toxicity, delirium, and risk of suicide by overdose. The risk of side effects can be reduced with careful prescribing practices (Box 2).

Anticholinergic effects, sedation TCAs vary in their anticholinergic activity (Table 2). Tertiary amine TCAs such as amitriptyline and protriptyline may cause dry mouth, constipation, urinary hesitancy, and blurred vision in some patients, and confusion in elderly or demented patients. Secondary amine TCAs such as nortriptyline or desipramine are less anticholinergic and less likely to cause these side effects.

Peripheral anticholinergic side effects can be managed with bethanacol—a cholinergic drug—in dosages of 25 to 50 mg tid or qid. Dry mouth can be treated with pilocarpine, 5 mg bid to qid, or oral bethanacol (5- to 10-mg tablets sublingually), artificial saliva drops, sugarless candy/gum, or mouthwash.

Box 2

Table 2

BIOCHEMICAL EFFECTS OF TRICYCLICS AND OTHER ANTIDEPRESSANTS

Table 3

APPROXIMATE THERAPEUTIC PLASMA LEVEL RANGES

Sedation is a common side effect caused by the antihistaminic properties of some TCAs (Table 2). Agents with this effect (e.g., doxepin), when given once daily at bedtime, can benefit patients with concomitant sleep disturbance.

Weight gain Patients being treated with TCAs often gain weight, most likely because of carbohydrate craving associated with H2 blockade. Patient education, monitoring of weight, and dietary counseling may be necessary during TCA use. Regular exercise is recommended, although depressed patients often lack the motivation and energy to exercise.

Cardiovascular toxicity All TCAs have potential cardiovascular effects (Table 3), which are seen on an ECG as increased PR, QRS, or QTc intervals, especially at higher dosages or in patients with pre-existing cardiac disease.¹⁹ Orthostatic hypotension is among the most common cardiovascular side effects, particularly in the elderly, and may result in falls or other injuries.

A baseline ECG is recommended before starting TCA therapy, especially in depressed patients with cardiac conduction delays (primary or secondary to concomitant medications). Routine ECGs should be used to monitor patients:

• with pre-existing cardiac disease;
• taking higher-than-recommended dosages of a TCA;
• taking other medications that may affect cardiac conduction.

When in doubt, obtaining a cardiology consultation is recommended.

CNS effects Signs of CNS toxicity include confusion, memory impairment, delirium, seizures, coma, and eventual respiratory depression.²⁰ Risk factors include toxic TCA plasma levels, elderly patient age, and concomitant use of other medications, such as psychotropics (neuroleptics), anticholinergics, and antihistamines. CNS toxicity may be difficult to diagnose, as it may initially resemble worsening of depressed mood. Confusion or worsening of memory or cognitive function are predictors of CNS toxicity.

A patient with anticholinergic-related delirium should be monitored on a medical unit. A trial of physostigmine may be warranted to confirm the diagnosis.

Neurologic symptoms Maprotiline and clomipramine have been associated with an increased risk of seizures. Use reduced dosages in patients with a history of seizures or concomitant use of medications that may increase maprotiline or clomipramine levels or decrease the seizure threshold (e.g., other antidepressants, withdrawal from benzodiazepines). Amoxapine has been associated with extrapyramidal symptoms secondary to a neuroleptic metabolite.

Overdose The severity of adverse side effects often depends upon drug concentration. An overdose of as little as a 2-week supply of a TCA can cause potentially fatal arrhythmias.

Priapism Trazodone has been associated with priapism, which has an incidence rate of 1/6,000 in patients taking this drug. Priapism is a medical emergency that requires prompt treatment.

Administration

TCAs are well-absorbed following oral administration and reach peak plasma levels in 2 to 6 hours. The average half-life of approximately 24 hours often allows therapeutic levels to be achieved with once-daily dosing at bedtime (Table 3).

TCAs are metabolized primarily by the cytochrome P-450 2D6 (CYP2D6) isoenzyme. Patients differ by a factor of 30- to 40-fold in their rate of metabolism of some TCAs, depending on individual genotypes.

The 7 to 9% of Caucasians classified as poor metabolizers at CYP2D6 require much lower-than-usual dosages of secondary amine TCAs.²¹ he 50% of Asians who are intermediate metabolizers require one-half the usual dosage. Tailor therapy to the individual patient, starting low, checking plasma levels, and monitoring for side effects.

For patients who would benefit from TCAs’ H1 anxiolytic effects, start with bid or tid dosing and later switch to once daily after achieving efficacy.

When using TCAs, start low and go slow to maximize efficacy and minimize side effects, especially in the elderly patient. The goal is to treat the patient, not to achieve target serum levels. The most favorable results have been demonstrated when patients are maintained at the dosages to which they respond when their disorder is in the acute stage.¹⁷

Symptoms usually improve after 2 weeks of TCA therapy, and up to 6 weeks may be required for a clinically significant effect. For many depressed patients, symptoms of insomnia, anxiety, and poor appetite improve within the first few days.

Monitoring serum levels

Serum levels are useful for monitoring treatment, compliance, and toxicity.²² Because the most accurate TCA serum levels are found at steady state, check the level after the patient has been on TCA therapy for at least 5 days and 8 to 12 hours after the last dose. Serum levels may also guide dosage increases in patients who exhibit a partial response to therapy.

Table 4

DRUGS THAT INTERACT WITH TCAs

Nortriptyline has a therapeutic window between 50 and 150 ng/mL. Patients who do not respond to serum levels higher than 150 ng/mL may respond when the dosage is reduced and the serum level falls to within that range. If response is adequate and without side effects, however, there is no need to reduce the dose. Other, less clear therapeutic windows have been described for other TCAs (Table 3).

TCA metabolism is affected by age and interaction with other drugs. Closely monitor serum levels in elderly and medically compromised patients, especially during dosage increases. Pay particular attention to other medications that may affect serum levels²³ (Table 4). SSRIs, particularly fluoxetine and paroxetine, may affect TCA plasma levels because of their potent inhibition of CYP2D6.

Individualize dosing and serum level monitoring. For example, the patient who is clinically tolerating a TCA and has a normal ECG does not require a serum level measurement, unless medically indicated.

Discontinuing tricyclics

When discontinuing a TCA, taper the dosage no more rapidly than 25 to 50 mg every 2 to 3 days. Abrupt discontinuation can cause cholinergic rebound, with symptoms such as nausea, cramping, headache, vomiting, and sweating. “Rebound hypomania” or “mania” have been reported with abrupt cessation of TCAs,²⁴ especially in patients with bipolar disorder. If the etiology of rebound symptoms is unclear (i.e., medical versus psychiatric), re-administering the discontinued TCA should relieve the symptoms and confirm a diagnosis of discontinuation.

Related resources

• Schatzberg AF, Cole JO, DeBattista C. Antidepressants. Manual of clinical psychopharmacology (3rd ed). Washington, DC: American Psychiatric Press, 1997.
• Janicak PG, Davis JM, Preskorn SH, Ayd Jr. FJ. Principles and practice of psychopharmacotherapy (2nd ed). Baltimore: Lippincott Williams & Wilkins, 1997.
• PDR Psychotropic Prescribing Guide (2nd ed). Montvale, NJ: Medical Economics, 1999.
• National Library of Medicine, MedlinePlus Health Information: Antidepressants, tricyclics (http://www.nlm.nih.gov/medlineplus/druginfo/antidepressantstricyclicsystem202055.html)

Drug brand names

See table 1 for tricyclic drug brand names. Others mentioned in this article include:

• Bupropion • Wellbutrin
• Cimetidine • Tagamet
• Citalopram • Celexa
• Disulfiram • Antabuse
• Fluoxetine • Prozac
• Fluvoxamine • Luvox
• Guanethidine • Ismelin
• Methylphenidate • Concerta, Ritalin
• Mirtazapine • Remeron
• Nefazodone • Serzone
• Paroxetine • Paxil
• Phenytoin • Dilantin

Disclosure

Dr. Zajecka reports that he receives grant/research support from, serves as a consultant to, and is on the speaker’s bureau of Bristol-Myers Squibb and Eli Lilly and Co.; receives grant/research support from Cephalon Inc., GlaxoSmithKline, Lichtwer Pharma, Merck and Co., MIICRO Inc., Otsuka Pharmaceuticals, Parke-Davis, Pfizer Inc., and Wyeth-Ayerst Pharmaceuticals; serves as a consultant to Abbott Laboratories; and is on the speaker’s bureau of Abbott Laboratories, Pfizer/Roerig, GlaxoSmithKline, Pharmacia, and Wyeth-Ayerst Pharmaceuticals.

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

Outpatient treatment of substance abuse is changing as research and experience teach us more about the nature of addictive illness and the principles of recovery. The recommended approach now emphasizes ease of access, chronic rather than acute treatment, and collaboration rather than confrontation.

As psychiatrists, we should be familiar with these changes, so that we can offer our addicted patients effective treatment and referrals. In particular, those of us who lead multidisciplinary teams in mental health clinics and outpatient programs need strategies that will help us make sound clinical decisions on detoxification, medical or psychiatric stabilization, and rehabilitation goals.

A new protocol that addresses these needs is being developed by a consensus panel of the Center for Substance Abuse and Treatment (Box 1).^1,2 Two of us (RFF, RR) are members of that panel, and we all are recognized experts in the outpatient treatment of addicted individuals, with combined experience of more than 70 years. Based on available evidence and expert opinion, we offer you 12 principles of outpatient substance dependence treatment that can help you achieve the most favorable results (Table 1).

Box 1

Principle 1: Open the doors wider

Outpatient clinics were once considered inappropriate for addicted individuals with significant psychosocial problems (such as homelessness) or co-occurring psychiatric disorders. Successful outpatient treatment was thought possible only for high-functioning, employed addicts who were free of significant psychiatric comorbidity.

Today, it is accepted that outpatients with a wide range of biopsychosocial problems can be effectively treated IF they receive case management and housing support and their co-occurring medical and psychiatric conditions are stabilized.

Efforts to ease the addicted patient into treatment should begin the moment a potential patient or family member seeks help. The pleasure that substance abusers derive from drug use makes them typically ambivalent about stopping their compulsive behaviors, and delays or obstacles to admission lead to “no shows” and drop-outs.³ Admissions increase when patients are given appointments the day they call for help.

From the initial outpatient encounter, the patient should feel like a welcomed participant who is responsible for his or her recovery. Access to outpatient programs increases when:

• child-care assistance is provided as needed;
• hours of operation are designed for the patient’s (rather than the staff’s) convenience;
• transportation assistance is provided, particularly for adolescents;
• the treatment plan is flexible and individualized to meet each patient’s specific needs.

Principle 2: Do a comprehensive initial evaluation

The open-door approach is most successful when the psychiatrist performs a comprehensive initial psychiatric and medical evaluation and works closely with a specialized treatment team. The initial medical and psychiatric evaluation is beyond the scope of this article and has been previously reviewed.⁴ Determining the need for medical detoxification is a priority during this phase of treatment.

Drug use patterns The treating physician should maintain a high index of suspicion for conditions associated with drug use. Cocaine causes seizures and cardiac arrhythmias, as well as vasoconstriction that leads to tissue necrosis (i.e., myocardial infarction, stroke, spontaneous abortion, and renal failure). Alcohol abuse affects brain, liver, cardiac, and endocrine tissue. Heroin produces acute overdose through respiratory depression, and its IV route of administration increases the risk of AIDS, viral hepatitis, pneumonia, sepsis, and endocarditis.

Function Structured interviews such as the Addiction Severity Index (ASI)⁵ can be used to assess functional impairment. Because addicted patients may be reluctant to disclose sensitive personal information, it is important to collect collateral information from family and friends, laboratory tests, and medical records.

Psychiatric concerns Many psychiatric syndromes are caused by substance abuse. Cocaine intoxication is often associated with psychosis (paranoia, auditory and tactile hallucinations), panic anxiety, and aggressiveness, whereas cocaine withdrawal produces depressed mood. Depression is also highly associated with chronic alcohol and opiate dependence. Withdrawal from opioids, alcohol, and sedatives produces anxiety.

Patients with bipolar disorder, major depression, panic anxiety, and schizophrenia who present with co-occurring addiction require coordinated and simultaneous stabilization of their addictive and psychiatric disorders.

Table 1

12 PRINCIPLES OF EFFECTIVE OUTPATIENT ADDICTION TREATMENT

Principle 3: Build on existing motivation

That an addicted patient must “hit bottom” before successful treatment may begin is a common misconception. In truth, studies find similar outcomes in individuals who enter treatment voluntarily and those who are externally pressured or legally coerced.⁶ Regardless of patients’ motivation when they enter treatment, they are likely to alternate over time between being more and less motivated. For this reason, it is necessary to remind them why they sought treatment and to use existing external pressures.

For example, with the patient’s consent it would be valuable to maintain contact with a parole officer who mandated a patient’s substance-abuse treatment. Likewise, patients entering treatment under threat of divorce should be asked to consent to family contact and should receive family therapy. Families often provide useful clinical information and can exert powerful influence when the patient’s motivation wanes. A patient’s refusal to allow contact with family (or other important sources of collateral information) often represents resistance that should be explored clinically and addressed.

With open dialogue, resistance to treatment can be reduced with education, peer groups, and family therapy. Motivational enhancement and interviewing work described by Miller, Procaska, and DiClemente^7,8 is designed to reduce treatment resistance in a respectful and clinically effective manner while avoiding confrontation that might provoke dropout and relapse.

Principle 4: Forge a therapeutic alliance

A therapeutic alliance produces positive outcomes in substance-dependent outpatients.^9-11 A recent National Institute on Drug Abuse (NIDA) therapy manual notes that a therapeutic alliance exists when the patient perceives that:

• the clinician accepts and respects him or her;
• the patient’s problems can be overcome by working together with the clinician;
• the clinician understands what the patient is hoping to get out of treatment.¹²

Clinicians can help forge this therapeutic alliance by being active listeners, by being empathic and nonjudgmental, and by approaching treatment as an active collaboration.¹²

Principle 5: Make retention a priority

It is simple but true: you cannot treat a patient who has dropped out.

Treatment retention is associated with better outcomes^13,14 and is a key indicator of the performance of an outpatient treatment program. High drop-out rates are demoralizing to patients who remain in treatment and to the clinicians who must document so-called “3-day treatments.” Because admission and initial evaluation of patients is labor-intensive, premature attrition is costly and time-consuming. Strategies to increase retention are listed in Table 2.

Table 2

STRATEGIES TO RETAIN PATIENTS IN TREATMENT

Clinicians can improve retention by tolerating different rates of change and levels of motivation. Individuals adopt new behaviors at different rates. You might become frustrated when patients do not immediately “buy” a particular version of recovery. Patients, however, often drop out when they feel they are being “force-fed” recommendations for sacrifice and major lifestyle changes that make no sense to them (at least not at the moment).

Principle 6: Provide ongoing care

Addiction is a complex biopsychosocial problem that requires long-term treatment. Even after extended abstinence, substance abusers experience craving and are vulnerable to relapse.

Addicts often enter outpatient treatment with psychosocial, medical, and psychiatric problems. Transformation from active addiction to full functioning in society requires sustained and conscientious effort by the patient, support system, and treatment team. Like asthma, diabetes, and other chronic diseases, addiction requires ongoing care.¹⁵

Unlike other chronic conditions, however, addiction is pleasure-reinforced, and addicted individuals are particularly at risk for relapse. Ongoing care may interrupt a relapse or at least interrupt it sooner than if no ongoing treatment were provided.

Substance dependence treatment for less than 90 days is of little or no use, and treatment lasting significantly longer often is indicated, according to the NIDA.¹⁶ When patients complete an intensive treatment phase, they should be evaluated for readiness to transfer to less-intensive care, with gradual transition from several therapeutic contacts per week, to weekly contact, to semimonthly contact, and so on. The concept of “graduation” should clearly convey not an ending but a “commencement” or beginning, as it does in college.

Unfortunately, the long-term approach to substance dependence is undermined by managed care organizations’ insistence on brief treatments. Also, regulations that view addiction as an acute episode may require that patient charts be closed at the end of intensive treatment. Such failures to appreciate the chronic nature of addiction undermine access to treatment and service delivery and contribute to recidivism and medical, social, criminal, and economic consequences associated with active addiction.

Principle 7: Match services with treatment needs

Outcomes improve when treatment services meet individual needs.¹⁷ Medical, psychiatric, psychosocial, legal, and housing problems can distract patients from the work of therapy. It is important to match each patient’s problems and needs with appropriate treatment settings, interventions, and services, according to the NIDA’s Principles of Drug Addiction Treatment.¹⁶ Creating flexible, responsive programs means more work for the treatment team, but it enhances the quality of care.

Treating concomitant psychiatric illness often requires innovations in outpatient treatment programs. For example, psychiatrists must avoid undermining treatment by inappropriately prescribing addictive agents such as benzodiazepines. At the same time, drug counselors may benefit from education about the potential benefits of medications in treating co-occuring disorders and craving. Coordinated delivery of psychiatric and rehabilitative treatment requires open communication in regularly scheduled multidisciplinary team meetings.

Principle 8: Monitor abstinence

Routine urine drug screens, Breathalyzer tests—administered at least weekly— and/or other laboratory tests to confirm self-reported abstinence can improve treatment outcomes. Regular drug and alcohol monitoring provides an objective indicator of progress, serves as a deterrent, and can help motivate the patient to withstand drug urges.

Individuals attempting abstinence from one substance have better outcomes if they abstain from all addictive substances¹⁸ (although tobacco use is controversial and requires further research). Even so, patients often continue to use addictive substances during treatment. Patients struggling with abstinence should not be discharged from treatment programs for manifesting the symptoms for which they are seeking treatment.

Substance-dependent individuals progress at different rates during treatment, and creative strategies to enhance motivation and retention can ultimately produce positive results. Outpatient clinics should consider different treatment tracks for patients at different stages of readiness for recovery.

Box 2

Principle 9: Use 12-step and other community supports

Patients who participate in 12-step programs and treatments have better outcomes than those who do not.^19-21 Still, patients in early recovery may find it difficult to join community-based support groups, such as Alcoholics Anonymous (AA). Patients are often ambivalent about—or strongly opposed to—joining AA because of embarrassment, negative experiences, or inadequate preparation for joining a 12-step fellowship. Substance abusers who are ambivalent about recovery often dispute 12-step directives on total abstinence, sweeping lifestyle changes, and the need to “give up control” over treatment recommendations. Common issues in early recovery include:

• how to select a 12-step home group and a sponsor;
• how to overcome uneasiness associated with being in a 12-step group;
• how to address any discomfort the patient may feel with the religious nature of 12-step meetings.²²

Patients’ resistance to 12-step treatment should be explored and addressed. Sometimes all they need is encouragement and help in finding a sponsor. Those with more difficult concerns may need a different approach. For example, social phobia is common in alcoholics and can be exacerbated by 12-step meetings; symptoms often respond to beta-adrenergic blockade. Patients with schizophrenia and those with paranoid features often do poorly in 12-step treatment if their paranoid symptoms cannot be successfully managed.

Principle 10: Manage medications

To avoid drug interactions, all prescribers involved with the patient’s care should coordinate their medication management efforts. Many substance abusers suffer from co-occurring psychiatric conditions²³ for which psychiatric medications are standard treatment.^24,25 In addition, medical detoxification is often necessary for heroin, alcohol, and sedative/hypnotic-dependent individuals. These treatments, which are beyond the scope of this article, are best integrated with drug rehabilitation.

Various medications for addiction have been reported to improve outcome:

• Agonist treatment with methadone, a long-acting opioid, can reduce heroin use, crime, and the risk of illnesses such as AIDS and viral hepatitis that are associated with IV drug use.
• Buprenorphine, a partial opioid receptor agonist with similar benefits, may soon be approved for the treatment of opiate dependence in outpatient settings.²⁶
• Naltrexone, an opioid receptor antagonist, has long been proposed as a treatment for opiate dependence and has been shown to be effective in alcoholism.²⁷

No effective pharmacologic treatment is available for cocaine dependence, although this is the focus of extensive government-sponsored research.

Principle 11: Educate about addiction and recovery

A wealth of accurate, free information about addiction and recovery is available through Web sites (Box 2) and other sources.

Ideally, outpatients in early recovery should be oriented in how to refuse offers of addictive substances, stress management, relapse prevention, information about the biology of addiction, 12-step fellowship integration, and appropriate use of medications.

Principle 12: Involve families in treatment

Treatment outcomes improve when addicts’ families are involved in the recovery process.^28,29 Some family members enable addictive behavior by purchasing drugs for their relatives or providing money for this purpose, while other families are knowledgeable about treatment and can be a vital force supporting the recovery process

Treatment can help modify unhealthy behavior patterns that some families develop to compensate for a substance abuser’s actions. Because substance use disorders often run in families, try to assess not only the identified patient but also others in the patient’s life.

Related resources

• Center for Substance Abuse Treatment (http://www.health.org/)
• Miller WR. Enhancing motivation for change in substance abuse treatment. CSAT treatment improvement protocol #35. U.S. Department of Health and Human Services, 1999.
• Rawson R. Treatment for stimulant use disorders. CSAT treatment improvement protocol #33. U.S. Department of Health and Human Services, 1999.
• Mercer DE, Woody GE. An individual drug counseling approach to treat cocaine addiction: The Collaborative Cocaine Treatment Study model. NIDA manual #3. Therapy manuals for drug addiction. U.S. Department of Health and Human Services, 1999.

Disclosure

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

Outpatient treatment of substance abuse is changing as research and experience teach us more about the nature of addictive illness and the principles of recovery. The recommended approach now emphasizes ease of access, chronic rather than acute treatment, and collaboration rather than confrontation.

As psychiatrists, we should be familiar with these changes, so that we can offer our addicted patients effective treatment and referrals. In particular, those of us who lead multidisciplinary teams in mental health clinics and outpatient programs need strategies that will help us make sound clinical decisions on detoxification, medical or psychiatric stabilization, and rehabilitation goals.

A new protocol that addresses these needs is being developed by a consensus panel of the Center for Substance Abuse and Treatment (Box 1).^1,2 Two of us (RFF, RR) are members of that panel, and we all are recognized experts in the outpatient treatment of addicted individuals, with combined experience of more than 70 years. Based on available evidence and expert opinion, we offer you 12 principles of outpatient substance dependence treatment that can help you achieve the most favorable results (Table 1).

Box 1

Principle 1: Open the doors wider

Outpatient clinics were once considered inappropriate for addicted individuals with significant psychosocial problems (such as homelessness) or co-occurring psychiatric disorders. Successful outpatient treatment was thought possible only for high-functioning, employed addicts who were free of significant psychiatric comorbidity.

Today, it is accepted that outpatients with a wide range of biopsychosocial problems can be effectively treated IF they receive case management and housing support and their co-occurring medical and psychiatric conditions are stabilized.

Efforts to ease the addicted patient into treatment should begin the moment a potential patient or family member seeks help. The pleasure that substance abusers derive from drug use makes them typically ambivalent about stopping their compulsive behaviors, and delays or obstacles to admission lead to “no shows” and drop-outs.³ Admissions increase when patients are given appointments the day they call for help.

From the initial outpatient encounter, the patient should feel like a welcomed participant who is responsible for his or her recovery. Access to outpatient programs increases when:

• child-care assistance is provided as needed;
• hours of operation are designed for the patient’s (rather than the staff’s) convenience;
• transportation assistance is provided, particularly for adolescents;
• the treatment plan is flexible and individualized to meet each patient’s specific needs.

Principle 2: Do a comprehensive initial evaluation

The open-door approach is most successful when the psychiatrist performs a comprehensive initial psychiatric and medical evaluation and works closely with a specialized treatment team. The initial medical and psychiatric evaluation is beyond the scope of this article and has been previously reviewed.⁴ Determining the need for medical detoxification is a priority during this phase of treatment.

Drug use patterns The treating physician should maintain a high index of suspicion for conditions associated with drug use. Cocaine causes seizures and cardiac arrhythmias, as well as vasoconstriction that leads to tissue necrosis (i.e., myocardial infarction, stroke, spontaneous abortion, and renal failure). Alcohol abuse affects brain, liver, cardiac, and endocrine tissue. Heroin produces acute overdose through respiratory depression, and its IV route of administration increases the risk of AIDS, viral hepatitis, pneumonia, sepsis, and endocarditis.

Function Structured interviews such as the Addiction Severity Index (ASI)⁵ can be used to assess functional impairment. Because addicted patients may be reluctant to disclose sensitive personal information, it is important to collect collateral information from family and friends, laboratory tests, and medical records.

Psychiatric concerns Many psychiatric syndromes are caused by substance abuse. Cocaine intoxication is often associated with psychosis (paranoia, auditory and tactile hallucinations), panic anxiety, and aggressiveness, whereas cocaine withdrawal produces depressed mood. Depression is also highly associated with chronic alcohol and opiate dependence. Withdrawal from opioids, alcohol, and sedatives produces anxiety.

Patients with bipolar disorder, major depression, panic anxiety, and schizophrenia who present with co-occurring addiction require coordinated and simultaneous stabilization of their addictive and psychiatric disorders.

Table 1

12 PRINCIPLES OF EFFECTIVE OUTPATIENT ADDICTION TREATMENT

Principle 3: Build on existing motivation

That an addicted patient must “hit bottom” before successful treatment may begin is a common misconception. In truth, studies find similar outcomes in individuals who enter treatment voluntarily and those who are externally pressured or legally coerced.⁶ Regardless of patients’ motivation when they enter treatment, they are likely to alternate over time between being more and less motivated. For this reason, it is necessary to remind them why they sought treatment and to use existing external pressures.

For example, with the patient’s consent it would be valuable to maintain contact with a parole officer who mandated a patient’s substance-abuse treatment. Likewise, patients entering treatment under threat of divorce should be asked to consent to family contact and should receive family therapy. Families often provide useful clinical information and can exert powerful influence when the patient’s motivation wanes. A patient’s refusal to allow contact with family (or other important sources of collateral information) often represents resistance that should be explored clinically and addressed.

With open dialogue, resistance to treatment can be reduced with education, peer groups, and family therapy. Motivational enhancement and interviewing work described by Miller, Procaska, and DiClemente^7,8 is designed to reduce treatment resistance in a respectful and clinically effective manner while avoiding confrontation that might provoke dropout and relapse.

Principle 4: Forge a therapeutic alliance

A therapeutic alliance produces positive outcomes in substance-dependent outpatients.^9-11 A recent National Institute on Drug Abuse (NIDA) therapy manual notes that a therapeutic alliance exists when the patient perceives that:

• the clinician accepts and respects him or her;
• the patient’s problems can be overcome by working together with the clinician;
• the clinician understands what the patient is hoping to get out of treatment.¹²

Clinicians can help forge this therapeutic alliance by being active listeners, by being empathic and nonjudgmental, and by approaching treatment as an active collaboration.¹²

Principle 5: Make retention a priority

It is simple but true: you cannot treat a patient who has dropped out.

Treatment retention is associated with better outcomes^13,14 and is a key indicator of the performance of an outpatient treatment program. High drop-out rates are demoralizing to patients who remain in treatment and to the clinicians who must document so-called “3-day treatments.” Because admission and initial evaluation of patients is labor-intensive, premature attrition is costly and time-consuming. Strategies to increase retention are listed in Table 2.

Table 2

STRATEGIES TO RETAIN PATIENTS IN TREATMENT

Clinicians can improve retention by tolerating different rates of change and levels of motivation. Individuals adopt new behaviors at different rates. You might become frustrated when patients do not immediately “buy” a particular version of recovery. Patients, however, often drop out when they feel they are being “force-fed” recommendations for sacrifice and major lifestyle changes that make no sense to them (at least not at the moment).

Principle 6: Provide ongoing care

Addiction is a complex biopsychosocial problem that requires long-term treatment. Even after extended abstinence, substance abusers experience craving and are vulnerable to relapse.

Addicts often enter outpatient treatment with psychosocial, medical, and psychiatric problems. Transformation from active addiction to full functioning in society requires sustained and conscientious effort by the patient, support system, and treatment team. Like asthma, diabetes, and other chronic diseases, addiction requires ongoing care.¹⁵

Unlike other chronic conditions, however, addiction is pleasure-reinforced, and addicted individuals are particularly at risk for relapse. Ongoing care may interrupt a relapse or at least interrupt it sooner than if no ongoing treatment were provided.

Substance dependence treatment for less than 90 days is of little or no use, and treatment lasting significantly longer often is indicated, according to the NIDA.¹⁶ When patients complete an intensive treatment phase, they should be evaluated for readiness to transfer to less-intensive care, with gradual transition from several therapeutic contacts per week, to weekly contact, to semimonthly contact, and so on. The concept of “graduation” should clearly convey not an ending but a “commencement” or beginning, as it does in college.

Unfortunately, the long-term approach to substance dependence is undermined by managed care organizations’ insistence on brief treatments. Also, regulations that view addiction as an acute episode may require that patient charts be closed at the end of intensive treatment. Such failures to appreciate the chronic nature of addiction undermine access to treatment and service delivery and contribute to recidivism and medical, social, criminal, and economic consequences associated with active addiction.

Principle 7: Match services with treatment needs

Outcomes improve when treatment services meet individual needs.¹⁷ Medical, psychiatric, psychosocial, legal, and housing problems can distract patients from the work of therapy. It is important to match each patient’s problems and needs with appropriate treatment settings, interventions, and services, according to the NIDA’s Principles of Drug Addiction Treatment.¹⁶ Creating flexible, responsive programs means more work for the treatment team, but it enhances the quality of care.

Treating concomitant psychiatric illness often requires innovations in outpatient treatment programs. For example, psychiatrists must avoid undermining treatment by inappropriately prescribing addictive agents such as benzodiazepines. At the same time, drug counselors may benefit from education about the potential benefits of medications in treating co-occuring disorders and craving. Coordinated delivery of psychiatric and rehabilitative treatment requires open communication in regularly scheduled multidisciplinary team meetings.

Principle 8: Monitor abstinence

Routine urine drug screens, Breathalyzer tests—administered at least weekly— and/or other laboratory tests to confirm self-reported abstinence can improve treatment outcomes. Regular drug and alcohol monitoring provides an objective indicator of progress, serves as a deterrent, and can help motivate the patient to withstand drug urges.

Individuals attempting abstinence from one substance have better outcomes if they abstain from all addictive substances¹⁸ (although tobacco use is controversial and requires further research). Even so, patients often continue to use addictive substances during treatment. Patients struggling with abstinence should not be discharged from treatment programs for manifesting the symptoms for which they are seeking treatment.

Substance-dependent individuals progress at different rates during treatment, and creative strategies to enhance motivation and retention can ultimately produce positive results. Outpatient clinics should consider different treatment tracks for patients at different stages of readiness for recovery.

Box 2

Principle 9: Use 12-step and other community supports

Patients who participate in 12-step programs and treatments have better outcomes than those who do not.^19-21 Still, patients in early recovery may find it difficult to join community-based support groups, such as Alcoholics Anonymous (AA). Patients are often ambivalent about—or strongly opposed to—joining AA because of embarrassment, negative experiences, or inadequate preparation for joining a 12-step fellowship. Substance abusers who are ambivalent about recovery often dispute 12-step directives on total abstinence, sweeping lifestyle changes, and the need to “give up control” over treatment recommendations. Common issues in early recovery include:

• how to select a 12-step home group and a sponsor;
• how to overcome uneasiness associated with being in a 12-step group;
• how to address any discomfort the patient may feel with the religious nature of 12-step meetings.²²

Patients’ resistance to 12-step treatment should be explored and addressed. Sometimes all they need is encouragement and help in finding a sponsor. Those with more difficult concerns may need a different approach. For example, social phobia is common in alcoholics and can be exacerbated by 12-step meetings; symptoms often respond to beta-adrenergic blockade. Patients with schizophrenia and those with paranoid features often do poorly in 12-step treatment if their paranoid symptoms cannot be successfully managed.

Principle 10: Manage medications

To avoid drug interactions, all prescribers involved with the patient’s care should coordinate their medication management efforts. Many substance abusers suffer from co-occurring psychiatric conditions²³ for which psychiatric medications are standard treatment.^24,25 In addition, medical detoxification is often necessary for heroin, alcohol, and sedative/hypnotic-dependent individuals. These treatments, which are beyond the scope of this article, are best integrated with drug rehabilitation.

Various medications for addiction have been reported to improve outcome:

• Agonist treatment with methadone, a long-acting opioid, can reduce heroin use, crime, and the risk of illnesses such as AIDS and viral hepatitis that are associated with IV drug use.
• Buprenorphine, a partial opioid receptor agonist with similar benefits, may soon be approved for the treatment of opiate dependence in outpatient settings.²⁶
• Naltrexone, an opioid receptor antagonist, has long been proposed as a treatment for opiate dependence and has been shown to be effective in alcoholism.²⁷

No effective pharmacologic treatment is available for cocaine dependence, although this is the focus of extensive government-sponsored research.

Principle 11: Educate about addiction and recovery

A wealth of accurate, free information about addiction and recovery is available through Web sites (Box 2) and other sources.

Ideally, outpatients in early recovery should be oriented in how to refuse offers of addictive substances, stress management, relapse prevention, information about the biology of addiction, 12-step fellowship integration, and appropriate use of medications.

Principle 12: Involve families in treatment

Treatment outcomes improve when addicts’ families are involved in the recovery process.^28,29 Some family members enable addictive behavior by purchasing drugs for their relatives or providing money for this purpose, while other families are knowledgeable about treatment and can be a vital force supporting the recovery process

Treatment can help modify unhealthy behavior patterns that some families develop to compensate for a substance abuser’s actions. Because substance use disorders often run in families, try to assess not only the identified patient but also others in the patient’s life.

Related resources

• Center for Substance Abuse Treatment (http://www.health.org/)
• Miller WR. Enhancing motivation for change in substance abuse treatment. CSAT treatment improvement protocol #35. U.S. Department of Health and Human Services, 1999.
• Rawson R. Treatment for stimulant use disorders. CSAT treatment improvement protocol #33. U.S. Department of Health and Human Services, 1999.
• Mercer DE, Woody GE. An individual drug counseling approach to treat cocaine addiction: The Collaborative Cocaine Treatment Study model. NIDA manual #3. Therapy manuals for drug addiction. U.S. Department of Health and Human Services, 1999.

Disclosure

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

Outpatient treatment of substance abuse is changing as research and experience teach us more about the nature of addictive illness and the principles of recovery. The recommended approach now emphasizes ease of access, chronic rather than acute treatment, and collaboration rather than confrontation.

As psychiatrists, we should be familiar with these changes, so that we can offer our addicted patients effective treatment and referrals. In particular, those of us who lead multidisciplinary teams in mental health clinics and outpatient programs need strategies that will help us make sound clinical decisions on detoxification, medical or psychiatric stabilization, and rehabilitation goals.

A new protocol that addresses these needs is being developed by a consensus panel of the Center for Substance Abuse and Treatment (Box 1).^1,2 Two of us (RFF, RR) are members of that panel, and we all are recognized experts in the outpatient treatment of addicted individuals, with combined experience of more than 70 years. Based on available evidence and expert opinion, we offer you 12 principles of outpatient substance dependence treatment that can help you achieve the most favorable results (Table 1).

Box 1

Principle 1: Open the doors wider

Outpatient clinics were once considered inappropriate for addicted individuals with significant psychosocial problems (such as homelessness) or co-occurring psychiatric disorders. Successful outpatient treatment was thought possible only for high-functioning, employed addicts who were free of significant psychiatric comorbidity.

Today, it is accepted that outpatients with a wide range of biopsychosocial problems can be effectively treated IF they receive case management and housing support and their co-occurring medical and psychiatric conditions are stabilized.

Efforts to ease the addicted patient into treatment should begin the moment a potential patient or family member seeks help. The pleasure that substance abusers derive from drug use makes them typically ambivalent about stopping their compulsive behaviors, and delays or obstacles to admission lead to “no shows” and drop-outs.³ Admissions increase when patients are given appointments the day they call for help.

From the initial outpatient encounter, the patient should feel like a welcomed participant who is responsible for his or her recovery. Access to outpatient programs increases when:

• child-care assistance is provided as needed;
• hours of operation are designed for the patient’s (rather than the staff’s) convenience;
• transportation assistance is provided, particularly for adolescents;
• the treatment plan is flexible and individualized to meet each patient’s specific needs.

Principle 2: Do a comprehensive initial evaluation

The open-door approach is most successful when the psychiatrist performs a comprehensive initial psychiatric and medical evaluation and works closely with a specialized treatment team. The initial medical and psychiatric evaluation is beyond the scope of this article and has been previously reviewed.⁴ Determining the need for medical detoxification is a priority during this phase of treatment.

Drug use patterns The treating physician should maintain a high index of suspicion for conditions associated with drug use. Cocaine causes seizures and cardiac arrhythmias, as well as vasoconstriction that leads to tissue necrosis (i.e., myocardial infarction, stroke, spontaneous abortion, and renal failure). Alcohol abuse affects brain, liver, cardiac, and endocrine tissue. Heroin produces acute overdose through respiratory depression, and its IV route of administration increases the risk of AIDS, viral hepatitis, pneumonia, sepsis, and endocarditis.

Function Structured interviews such as the Addiction Severity Index (ASI)⁵ can be used to assess functional impairment. Because addicted patients may be reluctant to disclose sensitive personal information, it is important to collect collateral information from family and friends, laboratory tests, and medical records.

Psychiatric concerns Many psychiatric syndromes are caused by substance abuse. Cocaine intoxication is often associated with psychosis (paranoia, auditory and tactile hallucinations), panic anxiety, and aggressiveness, whereas cocaine withdrawal produces depressed mood. Depression is also highly associated with chronic alcohol and opiate dependence. Withdrawal from opioids, alcohol, and sedatives produces anxiety.

Patients with bipolar disorder, major depression, panic anxiety, and schizophrenia who present with co-occurring addiction require coordinated and simultaneous stabilization of their addictive and psychiatric disorders.

Table 1

12 PRINCIPLES OF EFFECTIVE OUTPATIENT ADDICTION TREATMENT

Principle 3: Build on existing motivation

That an addicted patient must “hit bottom” before successful treatment may begin is a common misconception. In truth, studies find similar outcomes in individuals who enter treatment voluntarily and those who are externally pressured or legally coerced.⁶ Regardless of patients’ motivation when they enter treatment, they are likely to alternate over time between being more and less motivated. For this reason, it is necessary to remind them why they sought treatment and to use existing external pressures.

For example, with the patient’s consent it would be valuable to maintain contact with a parole officer who mandated a patient’s substance-abuse treatment. Likewise, patients entering treatment under threat of divorce should be asked to consent to family contact and should receive family therapy. Families often provide useful clinical information and can exert powerful influence when the patient’s motivation wanes. A patient’s refusal to allow contact with family (or other important sources of collateral information) often represents resistance that should be explored clinically and addressed.

With open dialogue, resistance to treatment can be reduced with education, peer groups, and family therapy. Motivational enhancement and interviewing work described by Miller, Procaska, and DiClemente^7,8 is designed to reduce treatment resistance in a respectful and clinically effective manner while avoiding confrontation that might provoke dropout and relapse.

Principle 4: Forge a therapeutic alliance

A therapeutic alliance produces positive outcomes in substance-dependent outpatients.^9-11 A recent National Institute on Drug Abuse (NIDA) therapy manual notes that a therapeutic alliance exists when the patient perceives that:

• the clinician accepts and respects him or her;
• the patient’s problems can be overcome by working together with the clinician;
• the clinician understands what the patient is hoping to get out of treatment.¹²

Clinicians can help forge this therapeutic alliance by being active listeners, by being empathic and nonjudgmental, and by approaching treatment as an active collaboration.¹²

Principle 5: Make retention a priority

It is simple but true: you cannot treat a patient who has dropped out.

Treatment retention is associated with better outcomes^13,14 and is a key indicator of the performance of an outpatient treatment program. High drop-out rates are demoralizing to patients who remain in treatment and to the clinicians who must document so-called “3-day treatments.” Because admission and initial evaluation of patients is labor-intensive, premature attrition is costly and time-consuming. Strategies to increase retention are listed in Table 2.

Table 2

STRATEGIES TO RETAIN PATIENTS IN TREATMENT

Clinicians can improve retention by tolerating different rates of change and levels of motivation. Individuals adopt new behaviors at different rates. You might become frustrated when patients do not immediately “buy” a particular version of recovery. Patients, however, often drop out when they feel they are being “force-fed” recommendations for sacrifice and major lifestyle changes that make no sense to them (at least not at the moment).

Principle 6: Provide ongoing care

Addiction is a complex biopsychosocial problem that requires long-term treatment. Even after extended abstinence, substance abusers experience craving and are vulnerable to relapse.

Addicts often enter outpatient treatment with psychosocial, medical, and psychiatric problems. Transformation from active addiction to full functioning in society requires sustained and conscientious effort by the patient, support system, and treatment team. Like asthma, diabetes, and other chronic diseases, addiction requires ongoing care.¹⁵

Unlike other chronic conditions, however, addiction is pleasure-reinforced, and addicted individuals are particularly at risk for relapse. Ongoing care may interrupt a relapse or at least interrupt it sooner than if no ongoing treatment were provided.

Substance dependence treatment for less than 90 days is of little or no use, and treatment lasting significantly longer often is indicated, according to the NIDA.¹⁶ When patients complete an intensive treatment phase, they should be evaluated for readiness to transfer to less-intensive care, with gradual transition from several therapeutic contacts per week, to weekly contact, to semimonthly contact, and so on. The concept of “graduation” should clearly convey not an ending but a “commencement” or beginning, as it does in college.

Unfortunately, the long-term approach to substance dependence is undermined by managed care organizations’ insistence on brief treatments. Also, regulations that view addiction as an acute episode may require that patient charts be closed at the end of intensive treatment. Such failures to appreciate the chronic nature of addiction undermine access to treatment and service delivery and contribute to recidivism and medical, social, criminal, and economic consequences associated with active addiction.

Principle 7: Match services with treatment needs

Outcomes improve when treatment services meet individual needs.¹⁷ Medical, psychiatric, psychosocial, legal, and housing problems can distract patients from the work of therapy. It is important to match each patient’s problems and needs with appropriate treatment settings, interventions, and services, according to the NIDA’s Principles of Drug Addiction Treatment.¹⁶ Creating flexible, responsive programs means more work for the treatment team, but it enhances the quality of care.

Treating concomitant psychiatric illness often requires innovations in outpatient treatment programs. For example, psychiatrists must avoid undermining treatment by inappropriately prescribing addictive agents such as benzodiazepines. At the same time, drug counselors may benefit from education about the potential benefits of medications in treating co-occuring disorders and craving. Coordinated delivery of psychiatric and rehabilitative treatment requires open communication in regularly scheduled multidisciplinary team meetings.

Principle 8: Monitor abstinence

Routine urine drug screens, Breathalyzer tests—administered at least weekly— and/or other laboratory tests to confirm self-reported abstinence can improve treatment outcomes. Regular drug and alcohol monitoring provides an objective indicator of progress, serves as a deterrent, and can help motivate the patient to withstand drug urges.

Individuals attempting abstinence from one substance have better outcomes if they abstain from all addictive substances¹⁸ (although tobacco use is controversial and requires further research). Even so, patients often continue to use addictive substances during treatment. Patients struggling with abstinence should not be discharged from treatment programs for manifesting the symptoms for which they are seeking treatment.

Substance-dependent individuals progress at different rates during treatment, and creative strategies to enhance motivation and retention can ultimately produce positive results. Outpatient clinics should consider different treatment tracks for patients at different stages of readiness for recovery.

Box 2

Principle 9: Use 12-step and other community supports

Patients who participate in 12-step programs and treatments have better outcomes than those who do not.^19-21 Still, patients in early recovery may find it difficult to join community-based support groups, such as Alcoholics Anonymous (AA). Patients are often ambivalent about—or strongly opposed to—joining AA because of embarrassment, negative experiences, or inadequate preparation for joining a 12-step fellowship. Substance abusers who are ambivalent about recovery often dispute 12-step directives on total abstinence, sweeping lifestyle changes, and the need to “give up control” over treatment recommendations. Common issues in early recovery include:

• how to select a 12-step home group and a sponsor;
• how to overcome uneasiness associated with being in a 12-step group;
• how to address any discomfort the patient may feel with the religious nature of 12-step meetings.²²

Patients’ resistance to 12-step treatment should be explored and addressed. Sometimes all they need is encouragement and help in finding a sponsor. Those with more difficult concerns may need a different approach. For example, social phobia is common in alcoholics and can be exacerbated by 12-step meetings; symptoms often respond to beta-adrenergic blockade. Patients with schizophrenia and those with paranoid features often do poorly in 12-step treatment if their paranoid symptoms cannot be successfully managed.

Principle 10: Manage medications

To avoid drug interactions, all prescribers involved with the patient’s care should coordinate their medication management efforts. Many substance abusers suffer from co-occurring psychiatric conditions²³ for which psychiatric medications are standard treatment.^24,25 In addition, medical detoxification is often necessary for heroin, alcohol, and sedative/hypnotic-dependent individuals. These treatments, which are beyond the scope of this article, are best integrated with drug rehabilitation.

Various medications for addiction have been reported to improve outcome:

• Agonist treatment with methadone, a long-acting opioid, can reduce heroin use, crime, and the risk of illnesses such as AIDS and viral hepatitis that are associated with IV drug use.
• Buprenorphine, a partial opioid receptor agonist with similar benefits, may soon be approved for the treatment of opiate dependence in outpatient settings.²⁶
• Naltrexone, an opioid receptor antagonist, has long been proposed as a treatment for opiate dependence and has been shown to be effective in alcoholism.²⁷

No effective pharmacologic treatment is available for cocaine dependence, although this is the focus of extensive government-sponsored research.

Principle 11: Educate about addiction and recovery

A wealth of accurate, free information about addiction and recovery is available through Web sites (Box 2) and other sources.

Ideally, outpatients in early recovery should be oriented in how to refuse offers of addictive substances, stress management, relapse prevention, information about the biology of addiction, 12-step fellowship integration, and appropriate use of medications.

Principle 12: Involve families in treatment

Treatment outcomes improve when addicts’ families are involved in the recovery process.^28,29 Some family members enable addictive behavior by purchasing drugs for their relatives or providing money for this purpose, while other families are knowledgeable about treatment and can be a vital force supporting the recovery process

Treatment can help modify unhealthy behavior patterns that some families develop to compensate for a substance abuser’s actions. Because substance use disorders often run in families, try to assess not only the identified patient but also others in the patient’s life.

Related resources

• Center for Substance Abuse Treatment (http://www.health.org/)
• Miller WR. Enhancing motivation for change in substance abuse treatment. CSAT treatment improvement protocol #35. U.S. Department of Health and Human Services, 1999.
• Rawson R. Treatment for stimulant use disorders. CSAT treatment improvement protocol #33. U.S. Department of Health and Human Services, 1999.
• Mercer DE, Woody GE. An individual drug counseling approach to treat cocaine addiction: The Collaborative Cocaine Treatment Study model. NIDA manual #3. Therapy manuals for drug addiction. U.S. Department of Health and Human Services, 1999.

Disclosure

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

Many fully trained psychiatrists, even psychopharmacologists, have rarely prescribed a monoamine oxidase inhibitor (MAOI), afraid of the possible consequences. One Boston-area psychiatrist who plans to retire soon has 12 patients who have been doing well for years on MAOIs. Before he called us, no psychiatrist in the area was willing to accept these patients who intended to stay on MAOIs.

Despite their potential side effects, the older MAOIs (phenelzine, tranylcypromine, isocarboxazid, and selegiline) have all been proven effective in depression; some studies have found them more effective than tricyclic antidepressants (TCAs) or selective serotonin reuptake inhibitors (SSRIs).^1,2 In our experience, MAOIs succeed in at least one-half the depressed patients who have failed on other antidepressants, and they are highly useful in treatment-resistant depressions and related disorders. Further, hypertensive crises are rare, and dietary restrictions are often exaggerated.

Drawing from the evidence and from our nearly 50 years’ experience in the use of MAOIs, we hope to acquaint or re-acquaint you with how to use these agents safely and to maximum effect.

When to use an MAOI

Reviews, meta-analyses, and more than 250 controlled studies have addressed MAOIs in psychiatric conditions.

The bulk of the literature reports a response rate of about 70% in various types of depression and anxiety.³ Unfortunately, because of the absence of major federal or industry funding, the findings are difficult to integrate, with most investigators testing different MAOIs on different subpopulations. This type of fragmented research, however, is not unique to MAOIs.

We at McLean have found MAOIs useful as second-line agents—and occasionally as first-line agents—in treating depression. We also use them as third-, fourth-, or fifth-line agents because we see many patients who have failed to respond to or tolerate several other antidepressants or whose response to other antidepressants has faded.

MAOIs are effective for a range of clinical presentations in inpatients and outpatients—depressive disorders (endogenous and atypical, unipolar and bipolar, neurasthenic and phobic), anxiety disorders (panic disorder and social phobia), and conditions as far afield as borderline personality disorder and the negative symptoms of schizophrenia.⁴

In controlled trials, many investigators have shown an MAOI to be more effective in some clinical conditions than a tricyclic and than a placebo.⁵ For example:

• A group at Columbia University studied phenelzine in various permutations of atypical depression (rejection sensitivity, overeating, oversleeping, mood reactivity, severe lack of physical energy).⁶
• Davidson et al studied isocarboxazid in anxious depression.⁷
• Thase and Himmelhoch studied tranylcypromine in anergic and bipolar depressions.⁸
• Liebowitz followed up with studies of phenelzine and tranylcypromine in social phobia.⁹
• At Yale, early studies were done with phenelzine in posttraumatic stress disorder (PTSD).¹⁰
• The British have generally argued for use of MAOIs in mixed anxiety and depression.¹¹
• The magisterial text Manic Depressive Illness recommends MAOIs as first-line treatment for bipolar depression.¹²

All of these studies yielded clinically and statistically substantial results with modest adverse effects.

Box 1

Characteristics of each agent

In the United States, a psychiatrist interested in using an MAOI for depressed patients can choose from among four agents. These older MAOIs are irreversible blockers of the enzymes MAO-A and MAO-B (Box 1).^13-16

Tranylcypromine has been found effective in patients hospitalized with endogenous depressions.¹⁷ Most clinicians experienced with MAOIs prefer this agent because it triggers weight gain to a much lesser degree than other MAOIs. Tranylcypromine can cause insomnia and a greater increase in tyramine sensitivity than other MAOIs, which therefore brings a greater risk of hypertensive crisis.

Phenelzine has the advantage that the effective dosage for depression is probably now known—more than 1 mg/kg body weight—although in three earlier large negative controlled studies, the effective dosage was not known and too little was given for too short a period. Weight gain and hypotension may be more common with this drug than with other MAOIs.

Isocarboxazid is the least studied but has been shown to be effective, with an average 41% drug-placebo difference in two studies.¹⁶ One of us (JOC) preferred it for several years, based on good responses in a handful of treatment-resistant patients.

Selegiline is used in low dosages for parkinsonism. At higher dosages (more than 20 mg/d), selegiline is a good antidepressant but is no longer MAO-B selective—inhibiting both the MAO-B and MAO-A enzymes and thus requiring the usual dietary restrictions. One of us (JAB) routinely maintains bipolar patients with mild baseline depression or lethargy on a mood stabilizer plus a low dosage (5 to 10 mg/d) of selegiline. With this combination, patients are more satisfied and compliant with the mood stabilizer regimen and require no dietary restrictions.

Overall, there may be a special clinical benefit to phenelzine in the more anxious, easily distressed patient and to tranylcypromine or selegiline in the more sluggish, lethargic patient. But probably the soundest way to choose MAOIs is if other, more standard treatments have not worked.

Combinations with other antidepressants

We have found all antidepressants that do not involve significant serotonin reuptake inhibition (e.g., bupropion, trazodone, and tricyclics other than clomipramine) can be safely administered with MAOIs. Combination therapy is worth considering because it may be effective when other approaches have failed.

For a patient switching from an SSRI, a trial of one of these other antidepressants may be appropriate before undertaking MAOI therapy, simply because it avoids the SSRI washout period. Normally, patients should be off fluoxetine for 5 weeks, other SSRIs or clomipramine for 2 weeks, and venlafaxine for 1 week before starting an MAOI.

With the exception of clomipramine, TCAs generally are safe with MAOIs, although the Physician’s Desk Reference warns against adding a TCA to an MAOI. We know of one death and one case of delirium when parenteral imipramine was given to a patient who had been on a MAOI for some time unbeknownst to the MD giving the injection. Both amitriptyline (and by extension, nortriptyline) and trimipramine have been given along with an MAOI in controlled studies with no special adverse effects.¹⁸ Both drugs, and probably doxepin, are good hypnotics and seem safe in combination, although one should begin with a low dosage (e.g., 25 mg hs) if the patient is improving on an MAOI but is bothered by insomnia.

Table 1

RECOMMENDED MAOI DOSAGE RANGES

Trazodone is frequently employed as a remedy for MAOI-induced insomnia.¹⁹ Mirtazapine can be used safely in combination with MAOIs, and the agent has been reported to acutely treat serotonin syndrome.²⁰

Determining dosage

Failure to understand the required effective dosages of MAOIs led to many early treatment failures and fueled clinicians’ perception that drugs in this class were ineffective. The senior author had to hospitalize a depressed man who had failed to improve on 30 mg/d of phenelzine for 10 days. The author later learned that a dose of more than 1 mg/kg per day for 3 to 4 weeks was necessary for response.

A rule of thumb is to start with one MAOI pill the first day and increase the dosage by one pill every 4 to 7 days until these levels are reached:

• 1 mg/kg/d for phenelzine;
• 40 mg/d for tranylcypromine and isocarboxazid;
• 45 mg/d for selegiline.

Reduce the dosage if side effects occur, or increase it if improvement is not noted in 1 to 2 months. We’ve tried starting patients on the 1 mg/kg dose of phenelzine the first day and elicited rapid and severe insomnia. Other intolerable side effects can also emerge from too-rapid dosage escalation. Clinical experience suggests the dosage ranges shown in Table 1.

Avoiding MAOI-related hypertension

We believe two factors have contributed to disuse of MAOIs by U.S. psychiatrists and slow development of new MAOIs here. These are the fear of severe hypertensive crisis and often vastly inflated MAOI dietary restrictions. In our practice, we provide patients receiving MAOIs with a reasonable list developed at the University of Toronto that explains foods to be avoided and alternatives that are allowed (Table 2).²¹

Table 2

DIETARY INSTRUCTIONS WITH MAOIs

In experience, tyramine-associated hypertensive crisis in patients receiving the older MAOIs is rare, often very painful, and time-limited. Data on incidence of hypertensive crises with any MAOI is inadequate, except for a negligible incidence with low-dose selegiline and moclobemide (a reversible MAOI available in Canada but not in the United States). Most reported cases have involved tranylcypromine,²² which causes the greatest increase in sensitivity to tyramine, the basis of the dietary interaction with MAOIs.²³ Specifically with tranylcypromine, transient hypertension can occur in the absence of dietary indiscretion or drug interaction.^24,25

Food-associated hypertensive reactions are more common but unpredictable. A patient at McLean who had eaten cheese without trouble for 4 years while taking tranylcypromine in Europe mysteriously developed a severe headache after one bit of cheddar in the United States; she never tried an MAOI again. Another patient who took tranylcypromine for months at dosages exceeding 100 mg/d (but never took the tyramine restriction seriously) suffered a frightening pulsatile headache after a Chinese restaurant meal with soy sauce.

In our experience, oral adrenergic drugs (e.g., pseudoephedrine) can cause symptomatic hypertensive crises in patients taking MAOIs. Some over-the-counter cold preparations may contain an adrenergic decongestant, but it may not be listed clearly on the bottle. Cough syrup with codeine is preferable to OTC cough syrups that contain dextromethorphan. Patients should be urged to check labels and ask the pharmacist.

Advise patients taking MAOIs to purchase a blood pressure cuff or finger sphygmomanometer and learn how to take their own blood pressure. If they experience a “normal” headache and their systolic blood pressure is not elevated by at least 30 mm Hg, they are not having a “cheese reaction” and do not need to worry. MAOI headaches are typically unmistakable (they feel as if one’s head is splitting apart), and blood pressure is markedly elevated.

We give patients starting on MAOIs 10-mg tablets of nifedipine and advise them that if symptoms suggesting a hypertensive crisis appear, they should bite into one tablet to release the fluid inside and then swallow it. We tell patients to repeat this in 15 minutes if the headache is not receding; if the headache persists, they should visit the emergency ward or the internist’s office for observation.

Although there has been concern about the risk of MI or stroke with the hypotensive effect of nifedipine, we believe it is still the best option for acute severe hypertension in patients who do not have chronic hypertension. Chlorpromazine tablets (50 mg) also will stop the headache and lower blood pressure but will leave the patient groggy for about 24 hours, with possible extrapyramidal symptoms.

In the rare instance that a hypertensive crisis occurs, the “official” labeling recommendation is to give IV phentolamine, but we find emergency rooms either no longer stock the drug or do not remember to do this. Send any patient who is hurting and panicky to an ER, and call to suggest what the attending might do (i.e., IV phentolamine or oral nifedipine or chlorpromazine, or the emergency physician’s preference for hypertensive crisis).

Avoiding other MAOI-related side effects

Orthostatic hypotension is the most common early side effect of phenelzine and isocarboxazid. Clinically significant hypotension is best managed via dosage adjustment. If that fails, add Na Cl, 1 to 2 grams bid with increased fluids. If that doesn’t work, administer fludrocortisone, 0.1 to 0.4 mg/d. If that fails, administer d-amphetamine, starting with 2.5 to 5 mg/d, and observe carefully for hypertension.

An altered diurnal sleep cycle is another MAOI-related side effect. Some patients get extremely drowsy in the late afternoon but then may have insomnia at night.²⁶ Attempts to alter this odd cycle by changing the timing of administration have generally not helped. We add a stimulant—pemoline (18.75 to 35.5 mg), d-amphetamine (2.5 to 5 mg), methylphenidate (5 to 10 mg), or modafinil (50 to 100 mg)—either in the morning or during afternoon sleepiness, with good effect and no hypertensive crises. We supervise the effect on blood pressure of the first few doses. Another helpful alternative has been to administer trazodone at night to initiate sleep.

Box 2

MAOIs are adrenergic agonists, and dry mouth and constipation can occur as side effects; these can be treated with bethanechol. Edema also occurs, mainly with isocarboxazid and phenelzine, and responds in most cases to modest diuretic doses, such as hydrochlorothiazide, 50 mg/d.

The now-recognized trouble reaching orgasm on SSRIs was first reported to us by women on phenelzine, and this side effect is as difficult to treat with MAOIs as it is with other antidepressants. We have found that sexual dysfunction is less likely to occur with tranylcypromine and selegiline than other MAOIs.

Toward safer MAOIs

Newer reversible MAOIs are in development, and two agents are available (although not in the United States). Delivery systems that reduce the risk of hypertensive crisis also may soon be available (Box 2).^27,28

Related resources

• Amsterdam JD, Chopra M. Monoamine oxidase inhibitors revisited. Psychiatric Ann2001;31(6):361-70.
• Treatment with antidepressants. In: Janicak PG, Davis JM, Preskorn SH, Ayd FJ, eds. Principles and practice of psychopharmacotherapy. 3rd ed. Philadelphia: Lippincott, Williams & Wilkins, 2001:215-326.
• Kennedy SH, ed. Clinical Advances in Monoamine Oxidase Inhibitor Therapies. Progress in Psychiatry Series. Washington, DC: American Psychiatric Press, 1994.

Drug brand names

• Isocarboxazide • Marplan
• Moclobemide • Aurorix, Manerix
• Phenelzine • Nardil
• Selegiline • Eldepryl
• Tranylcypromine • Parnate

Disclosure

Dr. Cole reports that he has served as a consultant to SmithKline Beecham Pharmaceuticals, manufacturer of tranylcypromine, and to Somerset Pharmaceuticals, manufacturer of selegiline in the United States. He also has participated in clinical studies of selegiline.

Dr. Bodkin reports that he has served as a consultant to Somerset Pharmaceuticals and has been principal investigator in several multicenter trials of selegiline.

Many fully trained psychiatrists, even psychopharmacologists, have rarely prescribed a monoamine oxidase inhibitor (MAOI), afraid of the possible consequences. One Boston-area psychiatrist who plans to retire soon has 12 patients who have been doing well for years on MAOIs. Before he called us, no psychiatrist in the area was willing to accept these patients who intended to stay on MAOIs.

Despite their potential side effects, the older MAOIs (phenelzine, tranylcypromine, isocarboxazid, and selegiline) have all been proven effective in depression; some studies have found them more effective than tricyclic antidepressants (TCAs) or selective serotonin reuptake inhibitors (SSRIs).^1,2 In our experience, MAOIs succeed in at least one-half the depressed patients who have failed on other antidepressants, and they are highly useful in treatment-resistant depressions and related disorders. Further, hypertensive crises are rare, and dietary restrictions are often exaggerated.

Drawing from the evidence and from our nearly 50 years’ experience in the use of MAOIs, we hope to acquaint or re-acquaint you with how to use these agents safely and to maximum effect.

When to use an MAOI

Reviews, meta-analyses, and more than 250 controlled studies have addressed MAOIs in psychiatric conditions.

The bulk of the literature reports a response rate of about 70% in various types of depression and anxiety.³ Unfortunately, because of the absence of major federal or industry funding, the findings are difficult to integrate, with most investigators testing different MAOIs on different subpopulations. This type of fragmented research, however, is not unique to MAOIs.

We at McLean have found MAOIs useful as second-line agents—and occasionally as first-line agents—in treating depression. We also use them as third-, fourth-, or fifth-line agents because we see many patients who have failed to respond to or tolerate several other antidepressants or whose response to other antidepressants has faded.

MAOIs are effective for a range of clinical presentations in inpatients and outpatients—depressive disorders (endogenous and atypical, unipolar and bipolar, neurasthenic and phobic), anxiety disorders (panic disorder and social phobia), and conditions as far afield as borderline personality disorder and the negative symptoms of schizophrenia.⁴

In controlled trials, many investigators have shown an MAOI to be more effective in some clinical conditions than a tricyclic and than a placebo.⁵ For example:

• A group at Columbia University studied phenelzine in various permutations of atypical depression (rejection sensitivity, overeating, oversleeping, mood reactivity, severe lack of physical energy).⁶
• Davidson et al studied isocarboxazid in anxious depression.⁷
• Thase and Himmelhoch studied tranylcypromine in anergic and bipolar depressions.⁸
• Liebowitz followed up with studies of phenelzine and tranylcypromine in social phobia.⁹
• At Yale, early studies were done with phenelzine in posttraumatic stress disorder (PTSD).¹⁰
• The British have generally argued for use of MAOIs in mixed anxiety and depression.¹¹
• The magisterial text Manic Depressive Illness recommends MAOIs as first-line treatment for bipolar depression.¹²

All of these studies yielded clinically and statistically substantial results with modest adverse effects.

Box 1

Characteristics of each agent

In the United States, a psychiatrist interested in using an MAOI for depressed patients can choose from among four agents. These older MAOIs are irreversible blockers of the enzymes MAO-A and MAO-B (Box 1).^13-16

Tranylcypromine has been found effective in patients hospitalized with endogenous depressions.¹⁷ Most clinicians experienced with MAOIs prefer this agent because it triggers weight gain to a much lesser degree than other MAOIs. Tranylcypromine can cause insomnia and a greater increase in tyramine sensitivity than other MAOIs, which therefore brings a greater risk of hypertensive crisis.

Phenelzine has the advantage that the effective dosage for depression is probably now known—more than 1 mg/kg body weight—although in three earlier large negative controlled studies, the effective dosage was not known and too little was given for too short a period. Weight gain and hypotension may be more common with this drug than with other MAOIs.

Isocarboxazid is the least studied but has been shown to be effective, with an average 41% drug-placebo difference in two studies.¹⁶ One of us (JOC) preferred it for several years, based on good responses in a handful of treatment-resistant patients.

Selegiline is used in low dosages for parkinsonism. At higher dosages (more than 20 mg/d), selegiline is a good antidepressant but is no longer MAO-B selective—inhibiting both the MAO-B and MAO-A enzymes and thus requiring the usual dietary restrictions. One of us (JAB) routinely maintains bipolar patients with mild baseline depression or lethargy on a mood stabilizer plus a low dosage (5 to 10 mg/d) of selegiline. With this combination, patients are more satisfied and compliant with the mood stabilizer regimen and require no dietary restrictions.

Overall, there may be a special clinical benefit to phenelzine in the more anxious, easily distressed patient and to tranylcypromine or selegiline in the more sluggish, lethargic patient. But probably the soundest way to choose MAOIs is if other, more standard treatments have not worked.

Combinations with other antidepressants

We have found all antidepressants that do not involve significant serotonin reuptake inhibition (e.g., bupropion, trazodone, and tricyclics other than clomipramine) can be safely administered with MAOIs. Combination therapy is worth considering because it may be effective when other approaches have failed.

For a patient switching from an SSRI, a trial of one of these other antidepressants may be appropriate before undertaking MAOI therapy, simply because it avoids the SSRI washout period. Normally, patients should be off fluoxetine for 5 weeks, other SSRIs or clomipramine for 2 weeks, and venlafaxine for 1 week before starting an MAOI.

With the exception of clomipramine, TCAs generally are safe with MAOIs, although the Physician’s Desk Reference warns against adding a TCA to an MAOI. We know of one death and one case of delirium when parenteral imipramine was given to a patient who had been on a MAOI for some time unbeknownst to the MD giving the injection. Both amitriptyline (and by extension, nortriptyline) and trimipramine have been given along with an MAOI in controlled studies with no special adverse effects.¹⁸ Both drugs, and probably doxepin, are good hypnotics and seem safe in combination, although one should begin with a low dosage (e.g., 25 mg hs) if the patient is improving on an MAOI but is bothered by insomnia.

Table 1

RECOMMENDED MAOI DOSAGE RANGES

Trazodone is frequently employed as a remedy for MAOI-induced insomnia.¹⁹ Mirtazapine can be used safely in combination with MAOIs, and the agent has been reported to acutely treat serotonin syndrome.²⁰

Determining dosage

Failure to understand the required effective dosages of MAOIs led to many early treatment failures and fueled clinicians’ perception that drugs in this class were ineffective. The senior author had to hospitalize a depressed man who had failed to improve on 30 mg/d of phenelzine for 10 days. The author later learned that a dose of more than 1 mg/kg per day for 3 to 4 weeks was necessary for response.

A rule of thumb is to start with one MAOI pill the first day and increase the dosage by one pill every 4 to 7 days until these levels are reached:

• 1 mg/kg/d for phenelzine;
• 40 mg/d for tranylcypromine and isocarboxazid;
• 45 mg/d for selegiline.

Reduce the dosage if side effects occur, or increase it if improvement is not noted in 1 to 2 months. We’ve tried starting patients on the 1 mg/kg dose of phenelzine the first day and elicited rapid and severe insomnia. Other intolerable side effects can also emerge from too-rapid dosage escalation. Clinical experience suggests the dosage ranges shown in Table 1.

Avoiding MAOI-related hypertension

We believe two factors have contributed to disuse of MAOIs by U.S. psychiatrists and slow development of new MAOIs here. These are the fear of severe hypertensive crisis and often vastly inflated MAOI dietary restrictions. In our practice, we provide patients receiving MAOIs with a reasonable list developed at the University of Toronto that explains foods to be avoided and alternatives that are allowed (Table 2).²¹

Table 2

DIETARY INSTRUCTIONS WITH MAOIs

In experience, tyramine-associated hypertensive crisis in patients receiving the older MAOIs is rare, often very painful, and time-limited. Data on incidence of hypertensive crises with any MAOI is inadequate, except for a negligible incidence with low-dose selegiline and moclobemide (a reversible MAOI available in Canada but not in the United States). Most reported cases have involved tranylcypromine,²² which causes the greatest increase in sensitivity to tyramine, the basis of the dietary interaction with MAOIs.²³ Specifically with tranylcypromine, transient hypertension can occur in the absence of dietary indiscretion or drug interaction.^24,25

Food-associated hypertensive reactions are more common but unpredictable. A patient at McLean who had eaten cheese without trouble for 4 years while taking tranylcypromine in Europe mysteriously developed a severe headache after one bit of cheddar in the United States; she never tried an MAOI again. Another patient who took tranylcypromine for months at dosages exceeding 100 mg/d (but never took the tyramine restriction seriously) suffered a frightening pulsatile headache after a Chinese restaurant meal with soy sauce.

In our experience, oral adrenergic drugs (e.g., pseudoephedrine) can cause symptomatic hypertensive crises in patients taking MAOIs. Some over-the-counter cold preparations may contain an adrenergic decongestant, but it may not be listed clearly on the bottle. Cough syrup with codeine is preferable to OTC cough syrups that contain dextromethorphan. Patients should be urged to check labels and ask the pharmacist.

Advise patients taking MAOIs to purchase a blood pressure cuff or finger sphygmomanometer and learn how to take their own blood pressure. If they experience a “normal” headache and their systolic blood pressure is not elevated by at least 30 mm Hg, they are not having a “cheese reaction” and do not need to worry. MAOI headaches are typically unmistakable (they feel as if one’s head is splitting apart), and blood pressure is markedly elevated.

We give patients starting on MAOIs 10-mg tablets of nifedipine and advise them that if symptoms suggesting a hypertensive crisis appear, they should bite into one tablet to release the fluid inside and then swallow it. We tell patients to repeat this in 15 minutes if the headache is not receding; if the headache persists, they should visit the emergency ward or the internist’s office for observation.

Although there has been concern about the risk of MI or stroke with the hypotensive effect of nifedipine, we believe it is still the best option for acute severe hypertension in patients who do not have chronic hypertension. Chlorpromazine tablets (50 mg) also will stop the headache and lower blood pressure but will leave the patient groggy for about 24 hours, with possible extrapyramidal symptoms.

In the rare instance that a hypertensive crisis occurs, the “official” labeling recommendation is to give IV phentolamine, but we find emergency rooms either no longer stock the drug or do not remember to do this. Send any patient who is hurting and panicky to an ER, and call to suggest what the attending might do (i.e., IV phentolamine or oral nifedipine or chlorpromazine, or the emergency physician’s preference for hypertensive crisis).

Avoiding other MAOI-related side effects