III. Specific Clinical Features Influencing the Treatment Plan

text A A A

A. Psychiatric Factors

1. Depressive symptoms
a. Suicidal ideation and behaviors

Because suicide is the worst outcome of major depressive disorder, a patient's risk for suicide must be assessed repeatedly over the course of treatment. For patients who report suicidal ideation, intention, or planning, close surveillance is necessary (see Sections II.A.3 and II.A.4). Psychiatrists should consider greater intensity of treatment for suicidal patients, including hospitalization when warranted and/or combined treatment with pharmacotherapy and psychotherapy. In patients at high risk for suicide and in whom a particularly rapid antidepressant response is required, consideration should be given to the use of ECT (239, 243, 521). Patients with major depressive disorder who present to an emergency department with suicidal ideation, or who have made a suicide attempt, should be triaged to determine their level of safety and establish the appropriate level and setting of care. Upon entrance to the emergency department, they should be searched to permit removal of potentially dangerous items, such as weapons and personal belongings that could cause harm (e.g., sharp objects, belts, shoes, medications). Factors to consider in determining the nature and intensity of treatment include (but are not limited to) access to and lethality of suicide means, past and family history of suicidal behavior, co-occurring substance abuse, the availability and adequacy of social supports, and the nature of the doctor-patient alliance.

To lower the risk of suicide, the psychiatrist should also treat modifiable risk factors, such as anxiety (especially panic attacks), insomnia, agitation, psychotic symptoms, and substance abuse (22), in addition to treating the major depressive episode. Among inpatients who died by suicide, Busch et al. (522) observed that a great majority were admitted for indications other than suicidal ideation, and anxiety and agitation were common, suggesting that such symptoms should be addressed if they are present. Family members can also play an important role in detecting and preventing suicidal behaviors. When permitted by the patient, the psychiatrist should educate those close to the patient concerning appropriate interventions and encourage communication.

There has been a growing controversy about the risk of suicidal ideas and behaviors (sometimes referred to as "suicidality") after initiation of antidepressant treatment. Although information on such risk continues to evolve, a predictive relationship to suicide has never been demonstrated. Clinical experience has long suggested that patients may develop the energy and capacity to act on self-destructive plans made earlier in the course of their illness if neurovegetative and psychomotor symptoms respond to antidepressant treatment before mood improves. More recently, meta-analyses of data from clinical trials have shown statistically significant increases in suicidal thoughts or behaviors in individuals age 25 years or younger who are treated with antidepressant medication, compared with placebo, with an approximately 1.5- to 2.5-fold increase in the relative risk (26, 523–530). In percentage terms, it is estimated that one to three of 100 individuals age 25 years or younger could potentially have an increase in suicidal thoughts or behaviors with antidepressant treatment (26), although there has been no evidence of increased mortality in the study subjects as a result of suicide (531). To alert clinicians to the need for vigilance and communication during the initial phase of antidepressant treatment, the FDA has issued a black-box warning pertaining to children, adolescents, and young adults that advises of this increase in the risk of suicidal thinking and behavior; information on the warning is available on the FDA Web site at http://www.fda.gov/cder/drug/antidepressants/default.htm. In making decisions about treatment, this awareness of a potential increase in suicidal thinking and behavior in children, adolescents, and young adults must be balanced against the negative effects, including suicide, of untreated depression (532) as well as the demonstrated benefits of antidepressant treatment (523, 533–535). For adults age 65 years or older, a review of the evidence from clinical trials showed a decrease in the risk of suicidal thinking or behaviors with antidepressant treatment, with no change in risk detected for other adults (age 25 to 64 years) (536).

For additional details about the treatment of suicidal individuals, clinicians are encouraged to consult APA's Practice Guideline for the Assessment and Treatment of Patients With Suicidal Behaviors (22).

b. Major depressive disorder–related cognitive dysfunction

Cognitive inefficiency and impairment are common features of major depressive disorder. Many depressed patients report slowed thoughts, poor concentration, distractibility, and reduced capacity to process information. They also display diminished attention to self-care and to their environment. Transient cognitive impairment, especially involving attention, concentration, and memory storage and retrieval, are demonstrable through neuropsychological testing (537). In extreme cases, especially in elderly patients, these deficits are so prominent that patients appear to have dementia. For individuals who exhibit symptoms of a dementia syndrome, it is crucial that any underlying depressive disorder be identified and treated.

Among depressed patients, the differential diagnosis of cognitive dysfunction includes degenerative dementias (such as Alzheimer's disease and Pick's disease) and reversible causes (such as vitamin B12 deficiency, folate deficiency, testosterone deficiency, substance use). Several clinical features help distinguish major depressive disorder–related cognitive dysfunction from other dementia syndromes. When performing cognitive tasks, depressed patients generally exert less effort and report greater incapacity than do patients with dementia. The latter, especially in more advanced stages, typically do not recognize their cognitive failures, since insight is impaired. In contrast, depressed patients may report being unable to think or remember. Patients with major depressive disorder–related cognitive dysfunction lack the signs of cortical dysfunction (i.e., aphasia, apraxia, agnosia) encountered in dementias such as Alzheimer's disease (538, 539). Nevertheless, distinguishing dementia from depression-related cognitive dysfunction can be difficult, particularly as the two may coexist. For further discussion of the co-occurrence of dementia and depression, the reader may also wish to consult APA's Practice Guideline for the Treatment of Patients With Alzheimer's Disease and Other Dementias, Second Edition (539).

The detection of major depressive disorder–related cognitive dysfunction alerts the psychiatrist to the need for treatment of the underlying major depressive disorder, which should in turn reduce the signs and symptoms of the cognitive dysfunction. Although initially reversible, major depressive disorder–related cognitive dysfunction increasingly appears to be a harbinger of subsequent dementia (540, 541). In addition, research suggests that certain types of executive cognitive dysfunction predict greater disability and limited treatment response in geriatric patients with depression (542, 543).

2. DSM depressive subtypes
a. Psychotic features

Major depressive disorder is sometimes accompanied by hallucinations or delusions, which may be congruent or incongruent with the depressed mood. Recognition of psychosis is essential among patients with major depressive disorder as it is often undetected, resulting in ineffective treatment (544–546). Psychotic features constitute a risk factor for recurrent major depressive disorder and recurrent psychosis and hence indicate the need for maintenance treatment.

Electroconvulsive therapy is highly effective in treating psychotic depression (241) and can be considered as a first-line treatment option whenever major depressive disorder is associated with psychotic features (239, 243, 547). Pharmacotherapy can also be used as a first-line treatment option for major depressive disorder with psychotic features. Psychotic depression typically responds better to the combination of an antipsychotic and an antidepressant medication rather than treatment with either component alone (547–549), although some research has shown comparable responses for antidepressive treatment or antipsychotic treatment alone (550, 551). Lithium augmentation is helpful for some patients who have not responded to combined treatment with antidepressant and antipsychotic medication (262, 552).

b. Catatonic features

A catatonic syndrome sometimes occurs in the context of major depressive disorder (553–556) and is characterized by at least two of the following manifestations: immobility, as evidenced by catalepsy or stupor; extreme agitation; extreme negativism; peculiarities of voluntary movement, as evidenced by posturing, stereotyped movements, mannerisms, or grimacing; and echolalia or echopraxia (556, 557). The presence of catatonia should prompt a thorough differential diagnosis as it can also occur in association with general medical conditions and with several other psychiatric disorders, including bipolar disorder and schizophrenia (556, 558, 559). Catatonic signs often dominate the clinical presentation and may be so severe as to be life-threatening, compelling the consideration of urgent somatic treatment. Patients with catatonic features may also need supportive medical interventions including hydration, nutrition, prophylaxis against deep vein thrombosis, turning to prevent bed sores, and passive range of motion to prevent contractures. Intravenous administration of a benzodiazepine (e.g., lorazepam, diazepam) or a barbiturate (e.g., amobarbital) may induce rapid relief (242, 553) and can be followed by continued oral administration for patients who show an initial response. When such intervention is ineffective, the clinician should consider the urgent use of ECT (239, 242). The efficacy of ECT in catatonia is well documented in the case series literature (239) and is usually apparent after a few treatments. After catatonic manifestations recede, antidepressant medication treatments may be needed during acute and maintenance phases of treatment. In addition to antidepressant medications, ongoing treatment may include ECT, lithium, antipsychotics, or a combination of these approaches, depending upon the patient's condition. Patients with catatonia may have an increased susceptibility to neuroleptic malignant syndrome when exposed to antipsychotic medications (560), and this should be considered in planning treatment.

c. Melancholic features

Melancholic features describe characteristic somatic symptoms, such as the loss of interest or pleasure in all, or almost all, activities or a lack of reactivity to usually pleasurable stimuli. Other symptoms include worsened depression in the morning, early morning awakening, and significant anorexia or weight loss, among others (16). Major depressive disorder with melancholic features is responsive to both pharmacotherapy and ECT. Serotonin norepinephrine reuptake inhibitors and TCAs may have an advantage over SSRIs for this patient population (561, 562). Psychotherapy may be less appropriate for patients with melancholia (563), particularly if the symptoms prevent engagement with the therapist (e.g., lack of interest in activities). Major depressive disorder with melancholic features may also be associated with an added risk of suicide (564) and an increased risk of subsequent recurrence despite the use of maintenance pharmacotherapy (509).

d. Atypical features

Major depressive disorder with atypical features is characterized by a pattern of marked mood reactivity and at least two additional symptoms, including leaden paralysis, a long-standing pattern of interpersonal rejection sensitivity, significant weight gain or increase in appetite, and hypersomnia (the latter two of which are considered reversed vegetative symptoms) (16). The phrase "atypical features" distinguishes this depressive subtype from the more classical "endogenous" presentation of depression, but it does not connote an uncommon or unusual form of depression. Atypical features are more common in women, are associated with an earlier age at onset of depression and a greater degree of associated anxiety disorders, and frequently have a more chronic, less episodic course, with only partial interepisode recovery (565, 566). These atypical features are also common in the depressed phase of bipolar I disorder and bipolar II disorder (567, 568), indicating a need for careful screening for manic or hypomanic episodes in patients who present with atypical depressive symptoms.

In the treatment of major depressive disorder with atypical features, MAOIs have greater efficacy than TCAs (122, 569–572). Some data also support the use of SSRIs, bupropion, and CBT in this patient population (573–577). Electroconvulsive therapy is also effective in treating patients with atypical features (578). The presence and severity of specific symptoms as well as safety considerations should help guide the choice of treatment for major depressive disorder with atypical features. For example, if a patient does not wish to, cannot, or appears unlikely to adhere to the dietary and medication precautions associated with MAOI treatment, the clinician should consider alternative antidepressant medication or psychotherapy.

e. Seasonal pattern

A seasonal pattern of major depressive disorder is characterized by a regular temporal relationship between particular periods of the year and the onset and remission of symptoms, which is not the result of seasonally related psychosocial stressors (e.g., seasonal unemployment, significant anniversaries). The most common presentation in the northern hemisphere is the regular appearance of symptoms between early October and late November and regular remission from mid-February to mid-April. Episodes of major depressive disorder with seasonal pattern frequently have atypical features such as hypersomnia and overeating. Some of these patients experience manic or hypomanic episodes as well; hence, it is important to diagnose bipolar disorder when appropriate.

The entire range of treatments for major depressive disorder may be used to treat major depressive disorder with seasonal pattern, either in combination with or as an alternative to light therapy. As a primary treatment, light therapy may be recommended as a 1- to 2-week time-limited trial (395), primarily for outpatients with clear seasonal patterns. For patients with more severe forms of major depressive disorder with seasonal pattern, the use of light therapy is considered adjunctive to pharmacological intervention. In terms of specific antidepressive agents, the extended release formulation of bupropion is FDA approved for use with patients who have major depressive disorder with seasonal pattern.

3. Co-occurring psychiatric disorders

Co-occurring psychiatric disorders generally complicate treatment. Patients with major depressive disorder who also have other psychiatric disorders have greater symptom severity and are more challenging to treat than patients with major depressive disorder alone. Yet the presence of co-occurring Axis I or Axis II disorders should not lead clinicians to conclude that patients are untreatable. Furthermore, other Axis I or Axis II disorders may masquerade as major depressive disorder or may seem to co-occur with the depression. In these cases, the other apparent disorders evanesce with successful treatment of the underlying major depressive disorder.

a. Dysthymic disorder

Dysthymic disorder is a chronic mood disorder with symptoms that fall below the threshold for major depressive disorder. Because of this, it may escape notice and may be inadequately treated. Nonetheless, it can cause significant suffering and disability. In some patients, both dysthymic disorder and major depressive disorder (so-called "double depression") may be diagnosed.

In the treatment of dysthymic disorder and chronic major depressive disorder, there is demonstrated efficacy for antidepressant medications, including TCAs, SSRIs, other newer agents, and MAOIs. Unfortunately, clinical trials provide little evidence of the relative efficacies of particular agents (105, 579). In general, pharmacotherapy of dysthymic disorder resembles that for episodes of major depressive disorder; responses to antidepressant medications by patients with dysthymic and chronic major depressive disorders have been comparable to the responses by patients with major depressive disorder episodes (580). In "double depression," antidepressant medication can reverse not only the acute major depressive episode but also the co-occurring dysthymic disorder (581).

Patients with dysthymic disorder, as well as patients with chronic and severe major depressive disorder, typically have a better response to the combination of pharmacotherapy and psychotherapy than to either alone (294, 295), although results of combined treatment studies have been mixed and complicated by methodological problems (582).

b. Anxiety disorders

As a group, anxiety disorders are the most commonly occurring psychiatric disorders in patients with major depressive disorder (583). A 2005 epidemiological study found that among individuals with major depressive disorder, 62% also met the criteria for generalized anxiety disorder, 52% for social phobia, 50% for posttraumatic stress disorder (PTSD), 48% for panic disorder, 43% for specific phobia, and 42% for obsessive-compulsive disorder (584). In addition, agitation and anxiety, including panic attacks, are frequent co-occurring symptoms of major depressive disorder. The appearance of anxiety and agitation in patients in a major depressive episode, particularly when accompanied by racing or ruminative thoughts, should alert the clinician to the possibility of a mixed mood state of a bipolar spectrum disorder (585).

In studies of major depressive disorder with a co-occurring anxiety disorder, both depressive symptoms and anxiety symptoms respond to antidepressant medication treatment (586). However, TCAs and SSRIs may initially worsen rather than alleviate anxiety symptoms, including panic attacks; patients should be so advised, and these medications should be introduced at low doses and slowly increased when treating such patients. Adjunctive antipanic agents, such as benzodiazepines, may be necessary as well. Selective serotonin reuptake inhibitors are beneficial for patients with co-occurring depression and social anxiety disorder (587) and co-occurring depression and PTSD (588). Bupropion is comparable to SSRIs in the treatment of patients with major depressive disorder and low to moderate levels of anxiety (82), but studies vary as to whether bupropion is (589) or is not (590) effective in the treatment of panic disorder. Because benzodiazepines are not antidepressants and carry their own adverse effects and toxicity, including abuse and dependence, benzodiazepines should not be the primary pharmacological agents for patients with major depressive disorder who have co-occurring anxiety symptoms. These agents may be used adjunctively with other antidepressive treatments, however (591). Psychotherapies such as CBT, behavioral therapy, and IPT may also be used to treat anxiety symptoms in the context of major depressive disorder (591, 592).

Obsessive-compulsive symptoms are also common in patients with major depressive episodes. In addition, obsessive-compulsive disorder may appear as a co-occurring condition in some patients with major depressive disorder. Clomipramine and SSRIs have demonstrated efficacy in managing obsessive-compulsive symptoms in addition to treating depression (593595).

c. Dementia

Patients with dementia are predisposed to depression, and the psychiatrist should therefore screen for depression in this population, although this is sometimes challenging (539). One screening tool is the Cornell Scale for Depression in Dementia, which incorporates self-report with caregiver and clinician ratings of depressive symptoms (596). Treatment of major depressive disorder in the cognitively impaired patient requires careful supervision and monitoring of the patient's pharmacotherapy; this may entail education of home health aides, nursing home staff, and others. Antidepressants are likely to be efficacious in treatment of depressive symptoms, but they do not improve cognition, and data on antidepressant use in patients with dementia are limited (597–599). Individuals with dementia are particularly susceptible to the adverse effects of muscarinic blockade on memory and attention. Therefore, individuals with dementia generally do best when given antidepressant medications with the lowest possible degree of anticholinergic effect, e.g., bupropion, fluoxetine, sertraline, trazodone, and, of the tricyclic agents, desipramine or nortriptyline (600). Alternatively, some patients do well when given stimulants in small doses. Electroconvulsive therapy is also effective in major depressive disorder superimposed on dementia. It should be used if medications are associated with an excessive risk of adverse effects, are not tolerated, or if immediate resolution of the major depressive disorder episode is medically indicated (such as when it interferes with the patient's acceptance of food). In individuals with dementia, ECT treatment may be associated with a transient worsening of the patient's cognitive status (239, 601, 602). APA's Practice Guideline for the Treatment of Patients With Alzheimer's Disease and Other Dementias, Second Edition (539) contains more information about the treatment of depression and dementia.

d. Substance use disorders

Major depressive disorder frequently occurs with alcohol or other substance abuse or dependence. Therefore, the psychiatrist should obtain a detailed history of the patient's substance use. With the patient's permission, family members, friends, or co-workers can be collateral sources of information. If the evaluation reveals a substance use disorder, this should be addressed in treatment. A patient with major depressive disorder who has a co-occurring substance use disorder is more likely to require hospitalization, more likely to attempt suicide, and less likely to adhere to treatment than a patient with major depressive disorder of similar severity uncomplicated by substance use (603–608).

Detoxifying patients before initiating antidepressant medication therapy is advisable when possible (110). Antidepressants may be used to treat depressive symptoms following initiation of abstinence if symptoms do not improve over time. It is difficult to identify patients who should begin a regimen of antidepressant medication therapy soon after initiation of abstinence, because depressive symptoms may have been induced by intoxication and/or withdrawal of the substance. A family history of major depressive disorder, a history of major depressive disorder preceding alcohol or other substance abuse, or a history of major depressive disorder during periods of sobriety raises the likelihood that the patient might benefit from antidepressant medication, which may then be started early in treatment. Comparing the temporal pattern of symptoms with the periods of use and abstinence of the substance may help to clarify the patient's diagnosis. Repeated, longitudinal psychiatric assessments may be necessary to distinguish substance-induced depressive disorder from co-occurring major depressive disorder, particularly because some individuals with substance use disorders reduce their substance consumption once they achieve remission of a co-occurring major depressive disorder.

Co-occurring substance use, especially with stimulant drugs, raises the risk of deleterious interactions with MAOIs, although few such events have been reported (609). Benzodiazepines and other sedative-hypnotics carry the potential for abuse or dependence and should rarely be prescribed to patients with co-occurring substance use disorders, except as part of a brief detoxification regimen. Hepatic dysfunction and hepatic enzyme induction frequently complicate pharmacotherapy of patients with alcoholism and other substance abuse. These conditions may require careful monitoring of blood levels (as appropriate for the medication), therapeutic effects, and side effects to avoid the opposing risks of either psychotropic medication intoxication or underdosing.

For individuals with nicotine dependence who wish to stop smoking, bupropion and nortriptyline treatment increase smoking cessation rates by about twofold (109) and would be useful to consider when selecting a specific antidepressive agent (110).

e. Personality disorders

For patients who exhibit symptoms of both major depressive disorder and a personality disorder, psychiatrists should consider appropriate treatment for each. Major depressive disorder should generally be the initial target; if evidence of a personality disorder persists when the depressive symptoms have resolved, psychotherapeutic and adjunctive pharmacotherapeutic approaches may be helpful (610–615). Patients with virtually any personality disorder exhibit a less satisfactory antidepressant medication treatment response, in terms of both social functioning and residual major depressive disorder symptoms, than do individuals without personality disorders (616). Personality disorders tend to interfere with treatment adherence and development of a psychotherapeutic relationship. Furthermore, many personality disorders increase the risk of episodes and increase time to remission of major depressive disorder (617, 618). Patients with various personality disorders also showed high rates of new-onset major depressive episodes in a large prospective study (619) and were at higher risk of attempting suicide than patients without a co-occurring personality disorder (620).

Axis II diagnoses should be made with caution during a major depressive episode, as depressive symptoms may exaggerate or mimic personality traits. Treatment of the depressive disorder for these patients can cause the apparent personality disorder symptoms to remit or greatly diminish. Depressed patients may believe that their current symptoms have been present from early life, when in fact they only began with the current episode. Such misperceptions often hinder accurate diagnosis. Some Cluster C personality disorders (e.g., avoidant, dependent, obsessive-compulsive) may reflect the residual impact of recurrent depressive symptoms and may remit with maintenance therapy (621).

Patients with borderline personality disorder have a particularly high rate of major depressive disorder: 20% in a community sample (622) and 50% in clinical samples (623). About 10%–15% of patients with major depressive disorder have co-occurring borderline personality disorder (624), and the percentage increases significantly in hospital and partial hospital samples. Patients with borderline personality disorder often exhibit mood lability, rejection sensitivity, inappropriate intense anger, and depressive "mood crashes." These symptoms are also common in patients with depression, particularly with atypical features, complicating the diagnosis of these disorders. Antidepressants are in general less effective in treating depressive episodes in patients with major depressive disorder and borderline personality disorder (625–629), with an overall response rate to all therapies of 20% (630).

For patients with major depressive disorder and borderline personality disorder, the personality disorder must also be addressed in treatment. Symptoms of both disorders can initially be treated with an SSRI or SNRI. Behavioral impulsivity and dyscontrol can also be treated with low-dose antipsychotics, lithium, and some antiepileptic medications. Monoamine oxidase inhibitors, although efficacious, are not recommended due to the risk of serious side effects and the difficulties with adherence to dietary restrictions. Psychotherapeutic approaches such as dialectical behavioral therapy and psychodynamic psychotherapy have been useful in treatment of borderline personality disorder as well. In patients with borderline personality disorder particular attention must be paid to the maintenance of therapeutic rapport, which is frequently disrupted, and to the risk of self-harm and suicide, which occurs in 8%–10% of such individuals. More information about the treatment of borderline personality disorder can be found in APA's Practice Guideline for Treatment of Patients With Borderline Personality Disorder (610).

f. Eating disorders

Eating disorders are also common in patients with major depressive disorder (631). Selective serotonin reuptake inhibitors are the best studied medications for treatment of eating disorders, with fluoxetine having the most evidence for the effective treatment of bulimia nervosa (170). Antidepressants may be less effective in patients who are severely underweight or malnourished, and normalizing weight should take priority in these patients. Although SSRIs and psychotherapy are widely used for patients with anorexia nervosa, the evidence base on which these practices rest is modest. Patients with chronic anorexia nervosa have in general been less responsive to formal psychotherapy. On the other hand, evidence strongly supports the use of CBT in the treatment of bulimia nervosa. Other therapies (e.g., IPT, group therapies, family therapy) and medications such as SSRIs have also been studied and have demonstrated effectiveness for this disorder. Bupropion should be avoided in individuals with eating disorders due to the increased risk of seizures in these patients. Electroconvulsive therapy has not generally been useful in treating eating disorder symptoms. Although there are few data to guide treatment of co-occurring major depressive disorder and eating disorders, it is reasonable to optimize treatment of both disorders based on these and other considerations. More information about the treatment of eating disorders can be found in APA's Practice Guideline for the Treatment of Patients With Eating Disorders, Third Edition (170).


B. Demographic and Psychosocial Variables

1. Major psychosocial stressors

Major depressive disorder may follow a substantial adverse life event, especially one that involves the loss of an important relationship or life role. This is particularly true in initial episodes of depression, with psychosocial stressors being less associated with the onset of recurrent episodes (632). Lower socioeconomic status, nonmarried status, unemployment, urbanization, and violent trauma seem to increase the risk of developing major depressive disorder, whereas religious belief may decrease it (633–635). Among those exposed to trauma, the prevalence of major depressive disorder seems to be higher among persons who develop PTSD than among those who do not (636). A recent meta-analysis underlined that among refugees, PTSD was diagnosed in one of 10 and one in 20 had major depressive disorder (637).

Marital discord has been identified as a potent risk factor in women for the development of depression (638, 639). Problems in the family setting may become an ongoing stressor that hampers the patient's response to treatment. Ambivalent, abusive, rejecting, or highly dependent family relationships may predispose an individual to major depressive disorder. The psychiatrist should screen for such factors and consider family therapy, as indicated, for these patients. Family therapy may be conducted in conjunction with individual and pharmacological therapies. Even for instances in which there is no apparent family dysfunction, it is important to provide the family with education about the nature of the illness and to enlist the family's support and cooperation with the treatment.

The psychiatrist may choose to treat a major depressive episode with an antidepressant, even if a major stressor preceded the episode. Nonetheless, attention to the relationship of both prior and concurrent life events to the onset, exacerbation, or maintenance of major depressive disorder symptoms is an important aspect of the overall treatment approach and may enhance the therapeutic alliance, help to prevent relapse, and guide the current treatment. A close relationship between a life stressor and major depressive disorder suggests the potential utility of a psychotherapeutic intervention coupled, as indicated, with somatic treatment.

2. Bereavement

Bereavement is a particularly severe stressor that can trigger a major depressive episode. However, grief, the natural response to bereavement, resembles depression, and this sometimes causes confusion. Psychiatrists treating bereaved individuals should differentiate symptoms of normal acute grief, complicated grief, and major depressive disorder, as each of these disorders requires a unique management plan. Normal grief should be treated with support and psychoeducation about symptoms and the course of mourning; complicated grief requires a targeted psychotherapy, with or without concomitant medication (535, 640); and major depressive disorder should be treated with medication and/or depression-focused psychotherapy.

Acute grief is the universal reaction to loss of a loved one, and it is a highly dysphoric and disruptive state (641). Acute grief is characterized by prominent yearning and longing for the person who died, recurrent pangs of sadness and other painful emotions, preoccupation with thoughts and memories of the person who died, and relative lack of interest in other activities and people. Despite the similarity with depression, only about 20% of bereaved people meet the criteria for major depressive disorder. Successful mourning leads to resolution of acute grief over a period of about 6 months. Integrated grief remains as a permanent state in which there is ongoing sadness about the loss often accompanied by ongoing feelings of yearning for the person who died. However, when the death is accepted, and grief integrated, the person is again interested in his or her own life and other people.

Complicated grief is a recently recognized syndrome in which symptoms of acute grief are prolonged, associated with intense and persistent yearning and longing for the deceased person, and complicated by guilty or angry ruminations related to the death and/or avoidance behavior. Studies of individuals bereaved by the attacks of September, 11, 2001, have demonstrated that complicated grief is a distinct condition from either major depressive disorder or PTSD (642, 643). It is important to note that treatment for depression is not effective in relieving symptoms of complicated grief (640).

Although DSM-IV-TR excludes the diagnosis of major depressive disorder during the first 2 months following bereavement, major depressive disorder in the wake of bereavement can impede the course of mourning. Bereavement-related depression responds to antidepressant medication and should be treated; otherwise it is likely to become chronic and impairing (644). There is no indication that depression in the context of bereavement differs from other major depressive episodes, and data indicate that chronicity of bereavement-related depression over 13 months is similar to chronicity of depression in other contexts (644).

3. Culture and ethnicity

An appreciation of cultural and ethnic variables is important to the accurate diagnosis of major depressive disorder and in the selection and conduct of psychotherapy and pharmacotherapy (645–647). Although major depressive disorder is seen across cultural and ethnic groups, and the age at onset, gender differences, and prevalence of co-occurring conditions are similar across cultures, the actual incidence and prevalence of depression vary (648–656). Furthermore, some evidence suggests that patients of different cultures express depressive symptoms differently, particularly somatic and psychomotor symptoms (657). Specific cultural variables may also influence the assessment of major depressive disorder symptoms. For example, in some cultures, depressive symptoms may be more likely to be attributed to physical diseases (658). In addition, language barriers can impede accurate psychiatric diagnosis and effective treatment (659), and, even when speaking the same language, individuals of different cultures may use different psychological terms to describe their symptoms (6, 7). In addition, the importance of individual experience should not be underestimated in the therapeutic relationship (660). The assessment and treatment process can also be influenced by religious beliefs (5). Individuals with high levels of religious involvement may have diminished rates of major depressive disorder (661, 662).

Differences in the utilization of psychiatric services by some cultural and ethnic groups have been well documented. Relative to Caucasians, African Americans and Latinos appear less likely to receive treatment for mood disorders (663–665).

Several studies have underscored the lower frequency of use of antidepressant drugs (and more specifically, SSRIs) (648, 666–669), ECT (670), and psychotherapy (671, 672) in minority groups, compared with Caucasians. If treatment for depression is initiated, African Americans are disproportionately more likely to receive pharmacotherapy (672), to drop out of treatment (673), and to develop chronic symptoms (674) than are Caucasian patients. These differences in mental health service use by minority populations appear to have a number of potential causes. Cultures and ethnicities may differ in the degree to which psychiatric illness is stigmatized (675) and in the preferences of individuals for treatment (676–678). For example, studies have found that Hispanic individuals were more likely to prefer counseling than whites, whereas African Americans varied across studies in their relative preference for counseling rather than pharmacotherapy (6, 679). Service use by minority populations is more affected by financial constraints (including those related to insurance) and social barriers (e.g., stigma) than the use of comparable services by Caucasians (664, 680, 681). In addition, pharmacological factors may play a role in patient preferences and adherence, as ethnic groups may differ in their relative rates of metabolism (682–684) and side effects and response to antidepressant medications (685–688).

4. Older age

The combined prevalence of major depression, dysthymic disorder, and "minor" depression in individuals over age 60 years has been reported to be as high as 25%, and major depressive disorder has been reported to be present in 14%–42% of nursing home residents (689). Elderly patients typically display more vegetative signs and cognitive disturbance but report less subjective dysphoria than younger patients. Major depressive disorder may consequently be misattributed to physical illness, dementia, or the aging process itself. For older adults with chronic illness or physical disability, including those expected to remain in a long-term care facility, depression may be erroneously regarded as expected or inevitable, and therefore untreatable (690). As a result, it is common for major depressive disorder to be undiagnosed and untreated among older adults.

As in all depressed individuals, a suicide risk assessment is an essential element of the evaluation process in older individuals. Although older adults constitute only 13% of the U.S. population, they account for 19% of all suicides, with elderly white men having the highest rates of completed suicides (691). This increase in suicide risk with aging in some demographic groups should be taken into consideration when estimating suicide risk and developing a plan to reduce such risk.

Several general medical conditions common among older adults are risk factors for depression. In addition, the presence of depression often exacerbates the course of the co-occurring medical condition and is a risk factor for poor outcomes. For example, elderly patients who are depressed and recovering from hip fractures have poorer functional outcomes from rehabilitative care and are less likely to return to full ambulation, compared with older adults with hip fractures who are not depressed (692–694). There is also frequent co-occurrence of major depressive disorder and cardiovascular disease; 25% or more of those with cardiovascular disease also have major depressive disorder, and co-occurring depression increases the morbidity and mortality of cardiovascular illness (695–697). The term vascular depression has been used to describe depression occurring in late life in patients with clinical evidence of cerebrovascular disease (698), although at this time it has not been established as a unique subtype of depression. In addition, major depressive disorder is a complication of cerebral infarction (see Section III.C.3). There is also a high prevalence of major depressive disorder among patients with dementia, and mood symptoms may precede cognitive symptoms and diagnosis of dementia (see Section III.A.3.C).

Just as patients with medical conditions should be screened for depression, patients exhibiting symptoms of depression should be thoroughly evaluated for the presence of co-occurring medical conditions, as major depressive disorder and general medical illnesses frequently coexist, especially in elderly patients (696, 699). This evaluation should include a systematic review of the patient's medications. Some medications have been reported to induce depressive symptoms (e.g., beta-blockers), although they may simply be producing lethargy and fatigue that mimic depression (700, 701). Consequently, the psychiatrist must carefully assess whether a given medication is contributing to depressive symptoms before prematurely altering what may be a valuable treatment. Patients undergoing their first major depressive episode in old age should be assessed for an undiagnosed neurological or other general medical disorder that may be responsible for the depressive symptoms. Similarly, frequently co-occurring symptoms of major depressive disorder, such as lassitude or pain, may mimic symptoms of a general medical condition. Pain in older adults, especially from orthopedic sources, may contribute significantly to the presence of depression in this population (702).

Once the patient has been thoroughly assessed, the treatment considerations for depressed geriatric patients are essentially the same as for younger patients. A meta-analysis has shown that SSRIs; TCAs; and cognitive-behavioral, behavioral, and psychodynamic therapies are all superior to placebo in the acute treatment for depression in subjects older than age 55 years (703–708). In addition, treatments for depression have been shown to be effective in nursing home populations (709, 710), as well as in inpatient and traditional outpatient settings of care. However, compared with younger individuals, older patients may be more likely to experience relapses and less likely to achieve a full response to treatment with antidepressant medications (711–714). In contrast, ECT is not only effective as a treatment for depression among older individuals, but those over age 65 years actually have better response rates than younger patients (715). Consideration should also be given to combined treatment with pharmacotherapy and psychotherapy, as evidence for the efficacy of stand-alone psychotherapy, such as CBT, is weak (716). Although the role of stimulants for antidepressant monotherapy is very limited, these compounds have some role in apathetic major depressive disorder in elderly patients with complicating general medical conditions. Late-life depression associated with vascular disease has also been found to respond well to treatment with antidepressants (717) and, in one unreplicated study, to TMS (718). Furthermore, in a recent randomized controlled trial, administering escitalopram prophylactically to patients who had experienced a stroke resulted in lower rates of depression at 12 months (334). Psychosocial factors are also frequent contributors to depression among older adults and should be addressed as part of the treatment plan (719, 720).

There are several considerations in prescribing medications for elderly patients. As with any patient, the psychiatrist should attempt to use as few medications as possible, and this is especially important given the complexity and multiplicity of issues in elderly patients. It is often useful to use medications that address several issues at once, such as choosing mirtazapine for a depressed, elderly patient with weight loss and insomnia. Elderly patients typically require a lower oral dose than younger patients to yield a particular blood level, and they tolerate a given blood level less well. Nevertheless, the blood levels at which antidepressant medications are maximally effective for elderly patients appear to be the same as those for younger patients (721, 722). Dose regimens should be adjusted for age-related metabolic changes, with close attention paid to hepatic and renal metabolic function. For patients who are receiving other medications, careful attention should be paid to potential drug interactions (160, 161, 723–725) (Tables 4 and 5).

Elderly patients are particularly prone to orthostatic hypotension and cholinergic blockade; for this reason, SSRIs, SNRIs, and other antidepressants should be considered over MAOIs or TCAs. Among the TCAs, desipramine and nortriptyline should be considered over amitriptyline, imipramine, and doxepin, due to increasing sensitivity to side effects and toxicity with advancing age. Treatment with SSRIs causes the syndrome of inappropriate antidiuretic hormone secretion (SIADH) at a higher rate in elderly patients than in younger patients (726, 727). Patients taking SSRIs have a three times greater risk of developing SIADH than patients taking other antidepressants, with the greatest risk among elderly patients (728).

For maintenance treatment, one study has shown that antidepressant medication (nortriptyline) and IPT are effective for elderly patients with recurrent major depressive disorder (315), yet a trend toward superior response was observed for combined pharmacotherapy and IPT, compared with pharmacotherapy alone. Another study demonstrated that paroxetine (but not monthly psychotherapy) was effective as maintenance therapy for elderly patients (729). For older patients with particularly severe or treatment-resistant depressive illness, addition of maintenance ECT to nortriptyline appears to improve treatment outcome (730). Among elderly patients who have had prior depression, the risk of developing another episode of major depressive disorder is substantially increased in those who develop or report sleep disturbance (731). Sleep disturbances may function as independent predictors of depression and are not simply prodromal depressive symptoms.

A body of systematic evidence suggests a particular value for various forms of collaborative or team-based care for elderly patients. Such care combines, for example, specialty mental health consultation/intervention with primary care management or community-based outreach and monitoring of care (732, 733). Older adults with depression can benefit from integration of mental health services in the setting where they typically receive their general medical care. It has been shown that support for algorithm-driven depression care processes within the primary care outpatient practice can lead to increased treatment adherence and improved clinical outcomes, including a reduction in mortality (734).

5. Gender

As part of the diagnostic assessment of a woman with major depressive disorder, there should be a detailed inquiry regarding reproductive life history and mood symptoms associated with reproductive life events, such as menses, use of oral contraceptive agents, peripartum, infertility, menopause, and pregnancy loss due to abortions, miscarriages, and perinatal losses. Although associations between reproductive factors and major depressive disorder are neither widespread nor consistent, some women may be particularly vulnerable to fluctuations in gonadal hormone levels (735). The perimenopausal transition has been identified as a high-risk period for new-onset major depressive disorder, with high variability of sex hormones as a risk factor (736, 737). Women in the perimenopausal transition may benefit from the use of serotonergic antidepressants, for mood and also for somatic symptoms such as hot flashes (738).

Since women are often caretakers in families, psychosocial stresses such as caring for an ill husband, child, or parent must be carefully assessed. Treating depressed mothers is associated with improved prognosis for their children as well (739). Maternal remission from depression was associated after 3 months with significantly decreased diagnoses and symptoms in their children, compared with children of mothers whose depression had not remitted. Thus, treating depressed mothers may crucially benefit both the patients and their children.

The patient's gender also factors into other treatment considerations for major depressive disorder. For example, the risks of certain adverse effects from treatments may also differ by gender. When prescribing trazodone to men, it is important to provide education about the risk of priapism (174). Older men typically have prostatic hypertrophy, making them particularly sensitive to anticholinergic effects of some antidepressants on the bladder outlet. While both men and women may experience decreased libido or anorgasmia while taking SSRIs, men may also experience ejaculatory dysfunction (740). Some women who are taking birth control pills require higher doses of TCA medications because of the induction of the hepatic enzymes responsible for medication metabolism. Similarly, medications that induce hepatic enzymes, such as anticonvulsants used as adjunctive treatment, reduce the effectiveness of contraceptives.

6. Pregnancy and postpartum

Major depressive disorder during pregnancy and postpartum presents unique treatment considerations. During these periods, approximately 10% to 15% of perinatal women will experience major depressive disorder, which is at least as common as rates reported for women in nonreproductive states (741, 742). Evaluation and communication of risks and benefits of antidepressants during pregnancy and breast-feeding is challenging and must include the risks of untreated maternal mood disorder, the limited body of research that informs safety of antidepressants, and the general lack of prospective long-term data following antidepressant exposure in utero and through lactation.

Depression-focused psychotherapy or other nonmedication therapies may be considered first for some women, and psychotherapy should be considered as part of the treatment plan whenever possible. As childbearing is a life stressor with psychosocial repercussions that may be amenable to psychotherapy, psychotherapy may serve to minimize medication exposure in some women. Although there is little controlled research, psychotherapies appear efficacious in antenatal and postpartum depression, with IPT being the best studied (743, 744). Therefore, depression-focused psychotherapies such as IPT and CBT should be considered in the treatment plan as a first-line treatment or in combination with medication to minimize medication exposure, especially if the individual has experienced a good response to a particular psychotherapy in the past or strongly prefers to avoid medication.

a. Depression during pregnancy

Psychiatrists should be familiar with the management of major depressive disorder in the context of pregnancy (745). More than 80% of women in the United States will have children (746), and about half of pregnancies are unplanned (747). Therefore, pregnancies—including unplanned pregnancies—are likely to occur during the course of treatment of major depressive disorder, as it is often a chronic and/or recurrent condition that is a major cause of disability during the reproductive years and disproportionately affects women, compared with men. In consideration of the high prevalence of both unplanned pregnancy and major depressive disorder in women, the risks and benefits of antidepressants and untreated maternal depression during pregnancy should be discussed with all female patients who have reproductive potential. Whenever possible, a pregnancy should be planned in consultation with a treating psychiatrist, who may wish to consult with a specialist in perinatal psychiatry. For women who are pregnant or planning to become pregnant, decisions about treatment for depression require weighing multiple benefits and risks for the woman as well as for the fetus. Making such decisions may require several discussions and will generally involve discussions with the patient's partner as well as her obstetrician.

Antidepressant medications carry some reported risks in pregnancy (see below), but so does untreated depression. Suicide risk, marital discord, the inability to engage in appropriate obstetrical care, and difficulty caring for other children must also be considered. There are also serious and well-characterized risks to the fetus of exposure to maternal major depressive disorder, including the possibility of low birth weight secondary to poor maternal weight gain (or frank weight loss) and increased risk of obstetrical complications such as premature delivery (748).

Antidepressant efficacy has not been determined for pregnant women, and questions remain as to whether medications have equivalent efficacy during pregnancy, compared with the nonpregnant state. Some safety data are available, but the findings often conflict, making data interpretation challenging and difficult to apply to the care of individual patients. Nevertheless, antidepressant medication should be considered and discussed as an option with pregnant women who have moderate to severe major depressive disorder.

For women who are in remission from major depressive disorder and receiving maintenance medication and/or for women deemed to be at high risk for a recurrence if the medication is discontinued, the risks of treatment with medications must also be weighed against the risks of alternative treatment options and untreated depression. Relapse rates for women with a history of major depressive disorder are high during pregnancy, especially if antidepressants are discontinued (749).

+ 1. Risks of antidepressants during pregnancy

The impact of the duration and timing of antidepressant exposure during pregnancy requires further study. Wisner et al. (750) reviewed the risks associated with the use of antidepressants during pregnancy, and a growing body of complicated literature has followed. Overall, risk of teratogenicity with antidepressants following first trimester exposure appears to be low, although some rare birth defects have been observed to occur at higher rates with use of specific SSRIs (751, 752). First-trimester paroxetine exposure has been associated with cardiac malformations, a finding that resulted in changes in FDA labeling for paroxetine from C to D (753); labeling changes have been made to all SSRI antidepressants with regard to this possibility when used during pregnancy. There have been conflicting results regarding whether first-trimester paroxetine exposure and cardiac teratogenicity are associated (754, 755). For fluoxetine, one study (756) found a higher incidence of three or more minor physical anomalies in infants exposed to fluoxetine than in a control group, and fetuses exposed to fluoxetine after 25 weeks' gestation had lower birth weights, which were associated with lower maternal weight gain. Although one case-control study reported a potential association between late antenatal SSRI use and the rare but serious condition of persistent pulmonary hypertension in the newborn (PPHN) (757), two subsequent retrospective chart review studies showed no such association (758, 759). An additional case-control study (760) showed a marginally significant increase in the relative risk of PPHN with SSRI use, but the estimated rate of PPHN occurrence was 1.5 per 1,000 births, which is less than the rate of 2.7 per 1,000 births reported by others (758) in non-SSRI exposed infants. Therefore, while many physicians are concerned about the reported association between SSRIs and PPHN, the preponderance of evidence from published studies on this topic does not support an association. Use of SSRIs may also be associated with prematurity (761, 762), although untreated depression and stress during pregnancy may also contribute to the risk of prematurity (748, 763). Some naturalistic studies and health care utilization studies suggest that antidepressants are associated with shorter length of gestation (761, 762), but there have been no randomized studies of the treatment of antenatal major depressive disorder that would adequately control for untreated maternal depression, antidepressant use, and confounding variables related to treatment selection. With late pregnancy antidepressant use, some but not all studies show a risk of medical complications such as prematurity and a transient neonatal withdrawal/adaptation syndrome (761, 764). The syndrome in the neonate appears to be associated with antidepressant use in the third trimester, has been reported in babies exposed in utero to TCAs and SSRIs, and includes transient symptoms such as jitteriness, tremor, difficulty with feedings, and other symptoms (764). Several studies, involving relatively small samples, have examined effects of antidepressant exposure during pregnancy on subsequent childhood behavior and development and found minimal (765) or no (766–768) effects on language, cognition, or motor or behavioral development independent of maternal depressed mood during pregnancy and the child's early life (766, 767).

+ 2. Implementation of pharmacotherapy during pregnancy

No controlled trials inform the use of antidepressants during pregnancy. Dose requirements may change during pregnancy because of changes in volume of distribution, hepatic metabolism, protein binding, and gastrointestinal absorption. Pharmacokinetic changes in late pregnancy may result in lower blood levels, with clinical implications, although more study is needed to develop monitoring and dosing guidelines. The limited data in this area demonstrate that pregnant women metabolize TCAs and SSRIs more rapidly in late pregnancy (769–774).

A number of factors influence antidepressant choice during pregnancy. If a woman has had a history of a good response to or is already taking a particular antidepressant, it is logical to consider that antidepressant among first-line treatments in an effort to minimize the number of different medication exposures. Using a single agent is also preferable to using several medications concomitantly. Because paroxetine use is classified as having a higher level of risk than other SSRIs, it should not be considered a first-line treatment when selecting a new antidepressant for a pregnant patient. Fluoxetine has the longest half-life and is more likely to be demonstrated at high levels in newborns after in utero exposure. Sertraline has been demonstrated to have lower cord blood levels than other SSRIs, although the clinical significance of this is unknown (775). Although there are few data for bupropion and safety in pregnancy, its benefits for smoking cessation may make it especially useful in women who have major depressive disorder and who smoke cigarettes, as tobacco is a known teratogen. Given these data, it is recommended that consideration be given to using an antidepressant with some available safety information that has been studied in pregnant women. For women who discontinue medication during pregnancy and are deemed at risk for postpartum depression, medication can be restarted following delivery.

Electroconvulsive therapy is also recommended as a treatment option for major depressive disorder during pregnancy (239). The current literature supports the safety for mother and fetus, as well as the efficacy of ECT during pregnancy (239). The psychiatrist should consider ECT for pregnant patients with moderate to severe depression who are unresponsive to or unsuitable for pharmacotherapy, for pregnant patients with major depressive disorder with psychotic features, and for pregnant patients electing to use this modality as a matter of preference after having weighed the relative risks and benefits of ECT and other treatment options. For details on the use of ECT during pregnancy, refer to The Practice of Electroconvulsive Therapy: Recommendations for Treatment, Training, and Privileging (A Task Force Report of the American Psychiatric Association) (239).

b. Postpartum depression

Major depressive disorder with postpartum onset is defined in DSM-IV-TR as a major depressive episode with onset within 4 weeks of delivery. However, the occurrence and course of major depressive disorder in childbearing women is heterogeneous, and definitions of postpartum depression may evolve with continued research (16, 776). In major depressive disorder with postpartum onset, anxiety symptoms are more prevalent than in major depressive disorder occurring at other times (777). It is not uncommon for women with postpartum depression to experience obsessions and/or compulsions, and obsessions may often involve thoughts of harming the baby, which must be differentiated from postpartum psychosis. Psychiatrists should provide psychoeducation about major depressive disorder to pregnant and postpartum women and their families to improve the detection of major depressive disorder during pregnancy and the postpartum period.

Several other psychiatric conditions may follow childbirth (778). The transient 7- to 10-day depressive condition referred to as "postpartum blues" is by definition too mild to meet the criteria for major depressive disorder and does not require medication. In addition to providing reassurance, psychiatrists should encourage mothers who experience postpartum blues to increase psychosocial support and obtain help with the care of the infant. Puerperal psychosis is a more severe disorder complicating one to two of 1,000 births. Although postpartum psychosis is rare, women with this disorder may have homicidal impulses toward the newborn; for this reason, careful assessment of homicidal as well as suicidal ideation, intention, and plans is important. Postpartum psychosis must always be treated as a psychiatric emergency, with hospitalization considered for the safety of the mother and baby (779). Many patients who have had episodes of this type ultimately prove to have bipolar disorder (780).

The woman's parenting skills for both the newborn baby and any other children in her care must be carefully assessed. Untreated maternal major depressive disorder, and specifically postpartum depression, have negative consequences for children, with adverse effects on attachment and child development (781, 782). Major depressive disorder can seriously interfere with the new mother's ability to provide physically and emotionally appropriate care for her baby and other children. The psychiatrist should work with the patient to develop a plan to manage this effect, such as enlisting family members to assist with child care.

Antidepressants are often prescribed for postpartum depression, according to the same principles delineated for other types of major depressive disorder, despite a limited number of controlled studies. Two placebo-controlled trials of SSRIs (fluoxetine and paroxetine) have been published for the treatment of postpartum depression, with fluoxetine appearing more efficacious than placebo and paroxetine being comparable to placebo on primary outcome measures of depressive symptoms (783, 784). Wisner et al. (785) did not find a difference in response and remission rates in a randomized controlled trial of sertraline versus nortriptyline for postpartum major depressive disorder. Open studies of other antidepressants in postpartum women suggest efficacy, although some studies included only a small number of participants (786). Paroxetine alone and paroxetine plus CBT both produced a significant change from baseline in one study, but there was no placebo-only group for comparison (787).

Patients and clinicians are often concerned about the risks of possible exposure to antidepressants during breast-feeding. These risks, however, must be weighed against the well-known, and at times profound, risks to the woman and her children of untreated postpartum depression. Mothers should be counseled regarding the relative risks and benefits when making these treatment decisions. Antidepressant medications are considered compatible with breast-feeding, but long-term data are not available regarding risks and benefits. Although there have been some suspected case reports of adverse effects in breast-feeding infants exposed to maternal antidepressants, most studies show low levels of exposure via breast milk, with the exception of fluoxetine, which appears to have a dose-related risk for detectable levels in infant sera (788, 789). At this time, there are no studies which have determined a "safe" amount and duration of antidepressant exposure in the fetus and newborn. However, exposure to antidepressants via breast milk is considered substantially lower than in-utero exposure. Women who elect to breast-feed while taking antidepressants should be supported in doing so, given the widely known health benefits (e.g., immune system effects) to infants who are breast-fed. Similarly, women who elect to bottle-feed should also be supported in this decision. Some women will not accept treatment with antidepressant medication while they are breast-feeding. Depression-focused psychotherapy can be recommended instead.

7. Family history

Major depressive disorder is one and one-half to three times as common among those with a first-degree biological relative affected with the disorder as in the general population. In addition, the rates of depression, anxiety, and other disorders are increased more than two- to sixfold in the offspring of depressed parents. A family history of depression is associated with an earlier age at onset of depression (790), and children of depressed parents are more likely to have depression with a chronic and recurrent course (791). Furthermore, a family history of recurrent major depressive disorder increases the chances that the patient's own illness will be recurrent and that the patient will not fully recover between episodes (792). A family history of bipolar I disorder, bipolar II disorder, or acute psychosis probably increases the chances that the patient's own major depressive disorder is a manifestation of bipolar rather than unipolar depression, and that antidepressant medication therapy may incite a switch to mania (793). Patients with such a family history should be questioned particularly closely regarding a prior history of mania or hypomania and should be carefully observed for signs of a switch to mania during treatment with antidepressant medication.

There are no real predictors of response to individual antidepressants, yet in the absence of other information clinicians sometimes rely on family history of therapeutic benefit to select a specific medication for a family member. Although it does not have specific support in the literature, this practice appears reasonable.


C. Treatment Implications of Co-occurring General Medical Conditions

In patients with co-occurring medical conditions, there is a higher prevalence of major depressive disorder than in the general population. Furthermore, co-occurring medical conditions in patients with major depressive disorder are associated with poorer outcome (794, 795). A number of medical conditions are known to cause mood symptoms, such as stroke, hypothyroidism, carcinoma of the pancreas, and many others. Apart from directly causing depressive symptoms, debilitating, painful, and chronic medical conditions often constitute an ongoing stressor that predisposes patients to depressive episodes. Nevertheless, a depressive episode, in any context, is never a "normal" response to illness and consequently warrants treatment.

In addition to the increased risk of major depressive disorder with general medical conditions, depressive episodes increase the risk of certain general medical conditions, such as heart disease. (796). Due to the interrelationship between depression and medical illness, it is very important to recognize and treat depressive symptoms in medically ill patients, and vice versa. The psychiatrist should also attend to the potential for interactions between antidepressants and the co-occurring medical conditions as well as any nonpsychiatric medications that the patient may be taking.

1. Hypertension

The presence of treated or untreated hypertension may influence the choice of an antidepressant, as a few antidepressant medications have been associated with increases in blood pressure. With SNRIs such as venlafaxine and duloxetine, dose-dependent elevations in blood pressure are usually mild, although more severe hypertension has also been observed (166, 797). However, another study found no increase in hypertension with duloxetine dosed up to 80 mg/day (798). Hypertension induced by SNRIs may respond to a decrease in the medication dose, or an alternative antidepressant medication may be considered. Alternatively, for a patient with well-controlled depressive symptoms, it may be preferable to add an antihypertensive agent rather than risk a depressive relapse or recurrence with medication tapering.

Antihypertensive agents and antidepressant medications may interact to either intensify or counteract the effect of the antihypertensive therapy (799). The action of antihypertensive agents that block alpha receptors (e.g., prazosin) may be intensified by antidepressant medications that block these same receptors, notably the TCAs and trazodone. Tricyclic antidepressants may antagonize the therapeutic actions of guanethidine, clonidine, or alpha-methyldopa. Concomitant antihypertensive treatment, especially with diuretics, increases the likelihood that TCAs, trazodone, or MAOIs will induce symptomatic orthostatic hypotension.

Side effects of antihypertensive agents, such as fatigue or sexual dysfunction, may also confound the evaluation and interpretation of depressive symptoms. It has also been thought that beta-blockers, especially propranolol, may account for depressive symptoms in some patients, but this association has been questioned (700, 701).

2. Cardiac disease

Depression increases the risk of cardiovascular disease (800). In addition, patients who are depressed following a myocardial infarction have an increased rate of mortality, compared with patients without depression (801–803). Following an acute myocardial infarction, the decreased survival rates of depressed patients may in part be due to lower heart rate variability in these patients, compared with nondepressed patients (804). Particularly in patients with a history of major depressive disorder (805), there is evidence that the depressive symptoms associated with cardiac illness respond to antidepressants (717, 806, 807). However, studies in which the attempt has been made to influence cardiac-related mortality through treatment of depression have shown mixed results (808–811).

A depressed patient with a history of a cardiac problem should be monitored for the emergence of cardiac symptoms, ECG changes, persistent tachycardia, or orthostatic blood pressure decrements. Consultation with the patient's cardiologist before and during antidepressant medication treatment may be advisable, especially for patients who have recently had a myocardial infarction. Increases in heart rate and blood pressure may be associated with the use of agents with noradrenergic properties such as SNRIs and stimulants; thus, changes in heart rate and blood pressure should be assessed after treatment with these agents is instituted in patients with coronary artery disease, hypertension, or congestive heart failure. Although TCAs have been used effectively to treat major depressive disorder in patients with some forms of ischemic heart disease (812), psychiatrists should take particular care in using TCAs for patients with a history of ventricular arrhythmia, subclinical sinus node dysfunction, conduction defects (including asymptomatic conduction defects), prolonged QT intervals, or a recent history of myocardial infarction (184, 185, 813–818). Selective serotonin reuptake inhibitors, SNRIs, and bupropion appear to be safer for patients with preexisting cardiac disease; several SSRIs have been used safely in patients with cardiac disease in large clinical trials (807, 809, 810, 819). Electroconvulsive therapy can also be used safely in individuals with cardiac disease or arrhythmias, although specialist consultation and modifications in ECT technique or anesthesia may be indicated (239, 245, 820–822). Monamine oxidase inhibitors do not adversely affect cardiac conduction, rhythm, or contraction but may induce orthostatic hypotension and have risks relating to drug-food and drug-drug interactions.

3. Stroke

Depression is observed in approximately one-third to one-half of individuals in the weeks to months following a stroke, with a substantial proportion developing major depressive disorder (334, 823, 824). Although conclusions of meta-analyses are mixed (825, 826), some research suggests that antidepressant treatment immediately following a stroke may reduce rates of depression (334) and possibly mortality (827). Among psychotherapeutic approaches to preventing depression after stroke, problem-solving therapy has been best studied, but findings are inconsistent (334, 825).

When depression develops after a stroke, it has detrimental effects on quality of life (823). In addition, the presence of depression 1 month following a stroke has been associated with an increase in subsequent mortality (828). Use of screening tools such as the PHQ-9 may help to identify depressive episodes after stroke (829), and care management may improve outcomes once poststroke depression is recognized (830). Psychotherapies have not been well studied as treatments for poststroke depression; however, a meta-analysis of randomized trials that have been conducted did not show efficacy (825). Findings on the therapeutic effects of antidepressants in post-stroke depression have been mixed, perhaps due to the substantial heterogeneity of study populations and designs (831, 832). Although a meta-analysis did not show any difference in the rate of depressive remission with antidepressant treatments compared with placebo (832), patients receiving an antidepressant did show more improvement in depressive symptoms (831, 832) and a greater proportion were classified as treatment responders (831). Individual randomized controlled trials have shown therapeutic benefits for several SSRIs, including fluoxetine, sertraline, and citalopram (833–836), and for the TCA nortriptyline (837, 838); however, not all studies have shown benefit for some of these agents (837, 839–841). Nevertheless, for individuals with poststroke depressive symptoms, a trial of antidepressant therapy may be considered, with SSRIs being better tolerated and having fewer contraindications in this older and more medically ill population (842, 843). However, in individuals who are receiving concomitant treatment with anticoagulant (e.g., warfarin) or antiplatelet (e.g., dipyridamole, clopidogrel, aspirin) medications, it is important to consider the potential for an increased bleeding risk due to drug-drug interactions with antidepressants (844, 845).

4. Parkinson's disease

Major depressive disorder occurs to some degree in 40%–50% of patients with Parkinson's disease. Patients with Parkinson's disease experience alterations of serotonergic and noradrenergic systems that may induce depression. There is no evidence favoring any particular antidepressant medication from the standpoint of therapeutic efficacy and safety for patients with Parkinson's disease complicated by major depressive disorder (846). A meta-analysis of placebo-controlled studies identified a clear benefit for both active treatment and placebo, but it did not find differences between them (847). Although SSRIs can be used, there is some risk of worsening of Parkinson's disease symptoms (increases in "off" time and exacerbation of tremor) with agents that are primarily serotonergic (848). Bupropion, in contrast, exerts a beneficial effect on the symptoms of Parkinson's disease in some patients but may also induce psychotic symptoms, perhaps because of its agonistic action in the dopaminergic system (849). Noradrenergic agents and SNRIs may also be preferable to SSRIs. Nonselective MAOIs (e.g., tranylcypromine, phenelzine, isocarboxazid) may adversely interact with l-dopa products (850). Selegiline, also known as l-deprenyl, is a selective type B MAOI recommended in the treatment of Parkinson's disease. Selegiline loses its specificity for MAO B in doses greater than 10 mg/day. As a result, it may induce serotonin syndrome when given in higher doses in conjunction with serotonin-enhancing antidepressant medications. The theoretical benefits of the antimuscarinic effects of some of the tricyclic agents in the treatment of patients with major depressive disorder with Parkinson's disease are offset by the memory impairment that may result. Amoxapine, an antidepressant medication with dopamine-receptor-blocking properties, should be avoided for patients who have Parkinson's disease. Lithium may, in some instances, induce or exacerbate parkinsonian symptoms. Electroconvulsive therapy exerts a transient beneficial effect on the symptoms of idiopathic Parkinson's disease in many patients (851, 852); however, it might occasionally worsen l-dopa-induced dyskinesias and induce a transient interictal delirium (853), which necessitates reductions in doses of dopamine agonist medications (239).

5. Epilepsy

The prevalence of depression in individuals with epilepsy appears to be increased in secondary and tertiary care center samples, although in population-based studies this increase is not well established (854). On the other hand, major depressive disorder significantly increases the risk of unprovoked seizures even after the adjustment of age, sex, length of medical follow-up, and medical therapies for depression (855). In addition, some antidepressant drugs, such as TCAs and bupropion, lower the seizure threshold and have a dose-dependent epileptogenic potential. This seizure risk is intermediate for immediate-release formulations of bupropion, maprotiline, and TCAs (in particular, clomipramine) and low for sustained-release formulations of bupropion (17, 197, 856).

Major depressive disorder in patients with seizure disorders can usually be safely and effectively managed according to the same principles outlined for patients without seizures. In particular, SSRIs and SNRIs are not likely to increase the risk of developing seizures (856–858). However, blood levels of TCAs may be increased by several antiepileptic drugs, increasing the side-effect burden of the anticholinergic and other side effects of tricyclics (859).

Some anticonvulsants appear useful for treatment and prophylaxis of mood disorders (e.g., carbamazepine, valproate, lamotrigine). Thus, in patients with depression and epilepsy, consideration can be given to concomitant prescription of an anticonvulsant (or elevating the dose of an existing anticonvulsant). Nevertheless, anticonvulsant compounds may also have a negative effect on mood for some patients (859). For example, barbiturates and possibly vigabatrin have been associated with an increased risk for depression (860). In addition, a recent FDA statement noted that increased rates of depression and suicide risk may be associated with anticonvulsants (861).

6. Obesity

Many individuals with major depressive disorder will be overweight or obese, given the high prevalence of excess weight in the general population (862). In addition, rates of depression may be increased in obese individuals, particularly among women and in those with a body mass index (BMI) greater than 40 (863). Individuals with obesity resulting from binge eating disorder also have higher rates of depression (170). In the subgroup of patients with atypical depression, increased eating and weight gain are symptomatic of the depressive disorder (864). For other patients, the lack of motivation and energy that occur with depression can make it difficult to maintain an exercise regimen or nutritional dietary habits. In addition, treatment with many antidepressant medications appears to lead to weight gain (865) and also makes it more difficult to lose weight in a structured weight management program (866).

In treating individuals with major depressive disorder who are overweight or obese, the effects of treatment on weight should be considered in selecting a therapeutic approach. If pharmacotherapy is used, the selection of an antidepressant medication should include consideration of its relative tendency to contribute to weight gain, which is generally greatest with mirtazapine, TCAs (tertiary TCAs more so than secondary TCAs), and MAOIs and less prominent with SSRIs and SNRIs (865). Bupropion is generally weight neutral and has been associated with modest weight reduction when used to treat major depressive disorder in obese adults (867, 868). Longitudinal monitoring of weight, either by direct measurement or patient report, can permit monitoring of BMI, as well as early intervention if weight gain becomes a problem with antidepressant treatment. Patients' concerns about weight gain may also contribute to poor adherence, and monitoring of weight can facilitate such discussions. The impact of weight on medication dosing should also be considered. In one study, greater relative body weight was associated with a lesser likelihood of response to a fixed dose trial of an antidepressant (869), perhaps suggesting a need for increases in medication dose with increasing body weight.

Psychotherapeutic approaches to treatment avoid the potential for medication-induced weight gain and may also have modest benefits in weight management. Cognitive-behavioral therapy has shown efficacy in the treatment of binge eating disorder (170, 870) and could potentially be used in addressing obesity (871) and medication-induced weight gain (872).

The increasing use of surgical treatments for obesity also has implications for the treatment of patients with major depressive disorder. Depression is common among bariatric surgical candidates and, in and of itself, is not a contraindication for surgery (873–877). Long-term follow-up studies show improvements in co-occurring general medical conditions (878), as well as decreases in depressive symptoms and improved quality of life with weight loss (879–881). However, weight loss after surgery may be less pronounced in individuals with a lifetime diagnosis of major depressive disorder (882) or in those with severe psychiatric illness that has required hospitalization (883). Close follow up is important following bariatric surgery in order to assess for changes in psychiatric symptoms, assist patients in the psychological and psychosocial adaptation to weight loss, and adjust medication regimens. Particularly following jejunoileal bypass or biliopancreatic diversion, but also following gastric bypass procedures, altered dissolution (884) and absorption of medication may require adjusting the dose of medication or changing from a slow-release to an immediate-release formulation (875).

7. Diabetes

Diabetes mellitus is common in the general population, particularly in overweight or obese individuals (885). However, it is not clear whether an association exists between diabetes and major depressive disorder, as meta-analyses and epidemiologic studies yield mixed results (886–888). Some patients may have reduced adherence to diet and medications when depressed (889), but there are inconsistent findings on whether successful treatment of depression (with medication, psychotherapy, and/or collaborative care) improves glycemic control (889–896). However, when initiating antidepressant therapy or making significant dosing adjustments, it is useful to collaborate with the patient's primary care physician in monitoring diabetic control because fluctuations in fasting blood glucose may occur. Some evidence suggests that use of TCAs may be associated with worsened glycemic control, and other antidepressants (such as SSRIs) may be preferable to TCAs for patients with diabetes (896, 897).

8. Sleep apnea

The possible contribution of obstructive sleep apnea (OSA) to depressive symptoms is an important consideration, particularly in patients who are obese, report excessive daytime sleepiness, or have treatment-resistant depressive symptoms. Symptoms such as fatigue and poor sleep quality can occur in sleep apnea as well as in major depressive disorder, requiring a careful assessment to distinguish whether either or both disorders are present. Underrecognition of OSA is common, and rates of OSA appear to be increasing with the increasing prevalence of obesity (898). Although the prototypical sleep apnea patient is likely to be obese with a history of snoring, sleep apnea may still be present even in the absence of these findings (899). Individuals with OSA or excessive daytime sleepiness appear to have greater rates of depression than comparison groups (900–902), although the rates of depressive symptoms and major depressive disorder diagnosis fluctuate across studies (903). In addition, epidemiological findings suggest an increasing likelihood of depression with increasing sleep-related breathing disorder severity (904). With initiation of continuous positive airway pressure treatment, improvement in OSA symptoms has been associated with decreased depressive symptoms (905, 906), in addition to reductions in OSA-associated health risks (898). Consequently, recognition and treatment of OSA is important among individuals with major depressive disorder. Identification of OSA is also important to treatment planning, as use of sedating medications can exacerbate OSA and worsen daytime sleepiness, with associated complications (907).

9. Human immunodeficiency virus and hepatitis C infections

According to the Centers for Disease Control and Prevention, more than one million individuals in the United States were living with HIV infection by the end of 2003 (908), with increased rates among individuals with psychiatric disorders, including substance use disorders (909). Estimates suggest that at least one-fifth of infected individuals have unrecognized infection (910), necessitating increased efforts to identify HIV infection (911), given the availability of effective treatment (912). Consequently, clinicians treating patients with major depressive disorder should consider screening for HIV.

Rates of major depressive disorder are increased among individuals with HIV infection, compared with HIV-negative individuals (913). When treating major depressive disorder in patients who also have HIV, antidepressant medications can be used safely and effectively (914, 915), although high placebo response rates and high study attrition rates have sometimes confounded interpretation of research findings (916, 917). When antidepressants are used, SSRIs are better tolerated than TCAs (918, 919). In individuals who are receiving treatment with antiretroviral agents, it is important to check for potential drug-drug interactions when choosing a medication regimen (920). Significant interactions can also occur if St. John's wort is taken by patients receiving antiretroviral medications. Although few studies have been conducted in patients who meet diagnostic criteria for major depressive disorder, individual and group psychotherapies using interpersonal, cognitive-behavioral, and psychoeducational approaches have also been associated with reductions in depressive symptoms among patients with HIV infection (316, 921–926).

Persons with mental illness also have elevated rates of infection with hepatitis C (927), and infection with hepatitis C is commonly present in individuals with HIV infection (928). Among individuals with hepatitis C, depressive symptoms are common, and many patients fully meet the criteria for major depressive disorder (929). Treatment of hepatitis C with interferon appears to be associated with a further increase in the risk for depression, although findings vary depending upon the study population, concomitant medications (e.g., ribavirin), and the type of interferon used for therapy (930, 931). The increase in depressive symptoms with interferon treatment may also be more prominent in patients with greater levels of pretreatment depression (932). This suggests a need for careful monitoring if patients with current major depressive disorder are administered interferon, particularly since many patients treated with interferon have unrecognized or insufficiently treated depression (933). Studies in which antidepressant medications were administered concomitantly with interferon have shown inconsistent prophylactic effects (934, 935). However, antidepressant therapy does seem to be effective when used to treat depression that develops in the course of interferon therapy for hepatitis C infection (936, 937). Consequently, major depressive disorder should not be viewed as a contraindication to the treatment of hepatitis C infection, particularly given the severe long-term hepatic complications associated with chronic infection (938).

10. Pain syndromes

Pain syndromes and major depressive disorder frequently co-occur. Although the reported prevalence of pain among depressed patients varies with cultural differences and study design, one-half to two-thirds of depressed individuals will typically note some type of pain (702, 939–941). Conversely, in primary care settings, individuals with pain symptoms are about twice as likely to be depressed as those without pain, and the rates of depression are further increased if pain is chronic or involves multiple types of pain (940, 942). It is important to note that individuals with co-occurring pain and depression tend to have worse treatment outcomes and poorer overall functioning than those with either condition alone (940, 942–944). Consequently, every patient with depression should be assessed for the presence, nature, location, and severity of pain complaints.

Overall, antidepressant treatment has been associated with reductions in pain symptoms among individuals with psychogenic or somatoform pain disorders (945). However, among trials of second-generation antidepressants in individuals with co-occurring pain and depression, duloxetine, venlafaxine, and paroxetine seem to be of comparable but relatively minor benefit (939, 946, 947).

Neuropathic pain is commonly associated with diabetic peripheral neuropathy but may also have other etiologies such as postherpetic neuralgia. For neuropathic pain in general, evidence-based guidelines recommend the use of TCAs or SNRIs (948, 949). Of the antidepressant medications, TCAs have been found to be most effective in decreasing pain associated with both postherpetic neuralgia (950, 951) and diabetic neuropathy (949, 952) but have no apparent effect on HIV-associated neuropathic pain (949). Given the greater tolerability of SNRI antidepressants, these agents may sometimes be chosen before a TCA for a patient with co-occurring depression and neuropathic pain. In addition, if a TCA is used, therapeutic drug monitoring may be helpful, given the wide variability of TCA blood levels across individuals (953).

Similar effects have been found for the use of antidepressants to prevent migraine and tension-type headaches. In patients with and without co-occurring depression, TCAs show greater efficacy than SSRIs (954, 955), but SNRIs also have some evidence for efficacy (956, 957). In individuals with tension-type headaches, addition of stress management therapy may augment the effects of TCA treatment (958).

Antidepressant treatment is also recommended for individuals with fibromyalgia, as it is associated with reductions in pain and often leads to improvements in function, with the best evidence available for amitriptyline (959). Beneficial effects are observed in those with or without co-occurring major depressive disorder (960–962). Although evidence from controlled trials is more limited for nonpharmacological approaches than for antidepressant treatment, education, exercise, and CBT are generally recommended for the treatment of fibromyalgia in combination with antidepressant medication (963, 964).

In the context of rheumatoid arthritis, antidepressant medications also appear to be effective in reducing pain as well as treating depressive symptoms (965–967). Evidence for psychosocial treatment is less consistent, with mindfulness meditation and emotion regulation therapy being associated with reduced pain and enhanced coping in individuals with rheumatoid arthritis and depression (968), but CBT having mixed results (966, 968). In individuals with co-occurring depression and osteoarthritis, collaborative depression care has been associated with reduced pain severity, improved function, and enhanced quality of life in those with low pain scores at baseline but had no effect when compared with usual treatment in those with severe arthritis pain (969, 970).

For individuals with chronic low back pain, there are conflicting opinions about the utility of antidepressant medications in reducing pain or improving function, even in the presence of co-occurring depression (971, 972). Nevertheless, antidepressant medications may still be indicated to treat depression on the basis of individual circumstances.

Since depressed patients with concurrent pain are often treated by primary care physicians and other medical specialists with a variety of potent analgesic medications, including narcotics, psychiatrists treating such patients are advised to be in contact with these other physicians initially and on a regular ongoing basis as indicated. The purposes of such contacts are to review the entire treatment plan, to assure that all prescribing physicians are aware of the full extent of pharmacological interventions, to coordinate specific prescribing areas and responsibilities so that patients do not receive prescriptions for the same medications or have their doses for given medications adjusted by several different prescribing clinicians, and to set up a mechanism and plan whereby all prescribing clinicians consistently keep one another informed about changes in their treatment plans and prescriptions.

11. Obstructive uropathy

Enlarged prostate size and other causes of bladder outlet obstruction are relative contraindications to the use of antidepressant medication compounds with antimuscarinic effects. For this reason, tertiary amine TCAs are best avoided in these patients. Benzodiazepines, trazodone, and MAOIs may also retard bladder emptying. The antidepressant medications with the least propensity to do this are SSRIs, SNRIs, and bupropion. If a TCA is chosen, the secondary amine desipramine is the least likely to cause urinary hesitancy or retention.

12. Glaucoma

Medications with anticholinergic potency may precipitate acute narrow-angle glaucoma in susceptible individuals (i.e., those with shallow anterior chambers) (973). Patients with glaucoma receiving local miotic therapy may be treated with antidepressant medications, including those possessing anticholinergic properties, provided that their intraocular pressure is monitored during antidepressant medication treatment. Prescription of agents lacking anticholinergic activity avoids this risk. Other agents sometimes used in psychiatry, e.g., topiramate and related sulfa-based medications, may cause acute angle closure glaucoma by ciliary body edema, a different mechanism (974).


Citing articles are presented as examples only. In non-demo SCM6 implementation, integration with CrossRef’s "Cited By" API will populate this tab (http://www.crossref.org/citedby.html).
Related Content
Manual of Clinical Psychopharmacology, 7th Edition > Chapter 3.  >
Manual of Clinical Psychopharmacology, 7th Edition > Chapter 8.  >
The American Psychiatric Publishing Textbook of Psychiatry, 5th Edition > Chapter 8.  >
The American Psychiatric Publishing Textbook of Psychiatry, 5th Edition > Chapter 9.  >
The American Psychiatric Publishing Textbook of Psychiatry, 5th Edition > Chapter 11.  >
Psychiatric News
PubMed Articles
  • Print
  • PDF
  • E-mail
  • Chapter Alerts
  • Get Citation