25: Depression

Depression is common, disabling, and often unrecognized in general medical practice. Even when recognized, physicians frequently do not provide systematic, longitudinal evidence-based management. And from the perspective of the patient, stigma and other psychosocial barriers such as negative expectations often diminish adherence to treatment recommendations. Despite robust documentation that depression is quite treatable and the widespread availability of evidence-based guidelines, overall outcomes remain poor.

This chapter focuses on the core knowledge and skills needed by general medical practitioners to effectively assess and manage major depressive disorder (MDD). We briefly address other related mood disorders: dysthymic disorder, adjustment disorder with depressed mood, depression secondary to general medical conditions, bipolar disorder/bipolar depression, and melancholia. We emphasize the routine use of a brief patient self-assessment tool, the nine-item Patient Health Questionnaire (PHQ-9), for diagnostic and ongoing management purposes. Widespread adoption of this one practice innovation may provide the pivotal leverage needed to improve outcomes for depression.

Major depressive disorder (MDD) is associated with considerable disability, morbidity, and mortality. Epidemiologic studies demonstrate that depression causes more disability and social and role impairment than most other chronic illnesses, such as diabetes, arthritis, hypertension, and coronary artery disease. The World Health Organization has identified major depression as the fourth leading cause of disability worldwide and projects that it will become the second leading cause of worldwide disability by 2020. Major depression is also a well-documented and common comorbidity in many other chronic conditions: for example, heart disease, stroke, diabetes mellitus, cancer, Parkinson disease, arthritis, pulmonary disease, and others. Furthermore, when present as a comorbidity, depression accounts for significant increases in disability, morbidity, and mortality.

The etiologic and sustaining relationships between depression and these other conditions appear to be bidirectional. For example, preexisting depression has been established as a predictor of future atherosclerotic coronary artery disease, cerebrovascular disease, diabetes, and osteoporosis, and conversely, significant physical illness predicts higher prevalence of major depression compared with individuals without physical illness. Depressed patients with heart disease (coronary artery disease, congestive heart failure) have worse medical outcomes including increased risk of reinfarction (after myocardial infarction [MI]) and up to a three-fold increase in all-cause mortality (especially after MI), even after controlling for all other identifiable and measurable cardiac risks (such as overeating, sedentary lifestyle, smoking, and other predictors of poor outcome). Patients with diabetes and depression have worse glycemic control, more microvascular and macrovascular complications and greater all-cause mortality.

Major depression is associated with adverse health habits, such as smoking, poor diet, overeating, and sedentary lifestyle, these in turn contribute to the onset of general medical illness and/or poor outcomes in illness. Conversely, functional impairment stemming from these chronic illnesses predispose to development of new depression. From an etiologic perspective, variables such as genetic vulnerability, childhood adversity (neglect and abuse), and stressful life events all contribute to the development of depression itself as well as to lifestyle risks such as obesity, sedentary behavior, and smoking that themselves predispose to chronic general medical illnesses.

Chronic care of general medical illness requires self-management behaviors to optimize treatment, for example, special diets, exercise regimens, medication changes, daily glucose monitoring, blood pressure checks, and limiting potentially harmful behaviors, such as smoking and risky drinking. Studies show that depression adversely impacts self-management, at least partly due to the fact that depressed patients are more likely to be nonadherent than nondepressed patients. Depressed diabetic patients have decreased adherence to diet and suffer more lapses in refills of oral hypoglycemic medications. Depressed patients with heart disease or stroke show decreased adherence to treatment recommendations such as taking daily aspirin and participating in exercise rehabilitation programs. This nonadherence in post-MI patients predicts increased rehospitalizations and mortality. Major depressive disorder, with or without general medical comorbidity generally is a chronic, recurring illness, with varying cycles of exacerbation and remission. Furthermore, these exacerbations or new episodes of depression tend to occur more frequently and with greater severity as the patient ages.

Dysthymic disorder is a less severe but more chronic form of depressive illness that is also associated with significant disability, and is even more likely to go undiagnosed than major depression. This disorder is diagnosed when depressed mood and at least two other symptoms of depression have been present “more than half the days” during the previous 2 years. Dysthymic disorder has been shown to respond to treatment with antidepressant medication. Patients with dysthymic disorder have an increased risk of experiencing a major depressive episode.

Adjustment disorder with depressed mood involves a reaction to an identifiable stressor, such as divorce or job loss. It presents with a sad or depressed mood, a level of impairment greater than expected for most individuals facing that specific stressor and is diagnosed within the first 6 months after the stressor has occurred. A “normal” reaction to a distressing life event should not be diagnosed as an adjustment disorder. When a stressor precipitates a depressive condition that meets the severity and symptom criteria for major depression, the diagnosis of major depression is made, regardless of the condition’s etiologic relationship to an identifiable stressor.

Mood disorders due to a general medical condition (or substance) refers to psychiatric syndromes judged to result from the direct physiologic consequence of a general medical condition (e.g., hypothyroidism), substance use (e.g., amphetamine withdrawal), or medication (e.g., interferon). Treatment focuses on resolution of the underlying general medical problem or withdrawal of the offending medication, although specific psychiatric treatment may also be useful.

Bipolar disorder is a common and severe mental illness, occurring in about 3–4% of the general population, causing significant disability, and carrying 80–85% genetic heritability. Bipolar I disorder refers to patients with a history of at least one episode meeting full criteria for major depression and at least one other distinct episode meeting criteria for mania. Other bipolar spectrum disorders such as bipolar II disorder (a condition marked by episodes of major depression and at least one documented episode of hypomania, not mania) and cyclothymic disorder (no episodes meeting full criteria for either major depression or mania/hypomania) are probably much more common, thought to occur in approximately 4–6% of the population.

Bipolar depression refers to an episode of illness meeting criteria for major depression in a patient with a history of either mania or hypomania. It can be very difficult for physicians to differentiate an episode of major depression from an episode of bipolar depression because the two conditions are phenotypically identical, that is they present the same signs and symptoms. It is extremely important, however, to distinguish between major depression and bipolar depression because the two conditions, though they present with the same symptoms, should be treated very differently. Studies in primary care indicate approximately 25% of patients presenting with symptoms that suggest the diagnosis of major depression actually suffer from bipolar depression.

Evidence for the best treatment of bipolar depression, however, is still somewhat limited and controversial. Only two medications, quetiapine and a combination product (olanzapine and fluoxetine) have received Food and Drug Administration (FDA) approval for the treatment of bipolar depression. Lamotrigine has FDA approval for maintenance treatment of bipolar depression, but is not yet approved for acute bipolar depression. Many experts recommend treating bipolar depression with a combination of a mood stabilizer (e.g., lithium, valproate, or carbamazepine) plus an antidepressant. Experts agree, however, that bipolar depression should not be treated with an antidepressant alone. Such “unopposed” treatment (i.e., antidepressant without mood stabilizer) increases the likelihood of precipitating an affective switch from depression to mania. Failure to elicit a history of mania or hypomania by the physician can lead, therefore, to potentially serious consequences by (mis)treating the condition with an antidepressant alone, thereby increasing the chances of a “switch” into mania or hypomania, with attendant risks of erratic or irrational behavior, poor judgment in social, occupational, economic, or interpersonal situations, psychosis, and even suicide. The section below on differential diagnosis provides physicians with some guidelines to help them make the diagnosis of bipolar depression in patients presenting with the symptom profile of major depression.

Melancholia is a severe form of major depression, often but not necessarily, associated with psychosis (“psychotic depression”). The symptom profile of melancholia includes lack of mood reactivity to usually pleasurable stimuli, loss of pleasure in all (or almost all) activities, distinct quality of depressed mood, diurnal variation in mood (worse in the morning), early morning awakening, marked psychomotor retardation or agitation, significant weight loss, and excessive or inappropriate guilt. This subtype of major depression is most closely associated with the so-called “biological markers” of depression, that is, nonsuppression of endogenous cortisol after administration of exogenous dexamethasone (DST), blunted thyroid-stimulating hormone (TSH) response to exogenous IV administration of thyrotropin-releasing hormone (TRH), and characteristically abnormal sleep physiology (including prolonged sleep latency, early-onset rapid eye movement [REM], increased total REM, and increased REM density). It appears that presence of melancholic features may predict preferentially better response to tricyclic medications (vs. the selective serotonin reuptake inhibitors [SSRIs]). The presence of melancholic features predicts a positive response to electroconvulsive therapy (ECT).

Epidemiologic studies demonstrate a lifetime prevalence of major depression in 7–12% of men and 20–25% of women. The point prevalence of major depression in a community sample is 2.3–3.2% for men and 4.5–9.3% for women. Reasons for these gender differences have not been fully elucidated, but both biological and sociocultural factors are involved. Numerous studies report a 10–15% prevalence of major depression in ambulatory medical settings with a substantially higher rate (20–40%) in patients with coexisting medical problems, particularly in those with diseases associated with strong biological or psychological predispositions to depression (e.g., stroke, Parkinson disease, traumatic brain injury, diabetes, coronary atherosclerotic disease, pancreatic cancer, and other terminal illnesses).

Prevalence of depression varies among age groups. Recent data point to a cohort effect through which current “baby boomers” experience the highest rates of depression of any previous generation. Although the most current epidemiologic findings show a surprisingly low 1-year prevalence rate of major depression in the elderly (1–2%), the rate of major or minor depression in elderly patients who seek treatment in primary care practices is 5%, with rates ranging from 15% to 25% in nursing home residents. Major depression is often misdiagnosed in elderly primary care patients as signs of aging, and cognitive impairment may also complicate accurate diagnosis. Some medications commonly prescribed in the elderly population may actually precipitate the onset of depression or cause symptoms like fatigue and poor concentration, which may mimic depressive symptoms.

Because the usual age of onset of depression is under the age of 40 years, an apparent first episode of depression in an older patient should prompt a thorough evaluation to exclude other underlying disease and/or medication effects.

Major depressive disorder represents a heterogeneous group of disorders. It is likely that future research will eventually provide diagnostic specificity to these disorders, leading to more targeted and effective treatments. For present purposes, however, the clinical manifestation of a major depressive episode should be considered a final common psychobiological pathway among multiple candidate etiologic determinants. Dramatic advances, however, in genetic, anatomic, physiologic, and immunologic studies already point the way toward a more precise biological understanding of this common and disabling condition.

Recurrent major depression has been shown to have a heritability of 35–40%, and genetic linkage studies have begun to identify specific regions of the genome thought to be candidates for carrying depression susceptibility. One particularly interesting candidate gene is the serotonin transporter gene (5-HTT), which makes functional sense since many antidepressants seem to work through binding to the 5-HTT protein. Recent gene by environment studies have explored the relationship between heterozygous and homozygous states and response to negative life events. While the results are not clear, this type of exploration illustrates the potential impact of genetic and environmental interactions.

Twin studies on depression in women indicate that genetic factors play the strongest etiologic role in depression, followed by recent (as opposed to early environmental) negative life events. Animal studies, on the other hand, demonstrate that early environmental stress predisposes to biological abnormalities associated with depression that may not emerge until adult life.

Postmortem pathologic studies, along with functional and structural imaging studies, converge in locating anatomic loci of depressive illness in the hippocampus, the dorsolateral prefrontal cortex, the anterior cingulate cortex, and the amygdala. Animal and human studies confirm volumetric decreases in the hippocampus in depressive illness in individuals with a history of adverse childhood events. Antidepressant medications appear to induce neurogenesis (increases in volume) in the hippocampus, possibly through increases in brain-derived neurotrophic factor (BDNF).

From the physiologic point of view, a considerable body of emerging evidence now conceptually and experimentally points to dysregulation of distributed brain networks and second messenger abnormalities as the underlying neurobiological abnormalities in recurrent mood disorders. This contrasts with earlier theories postulating that straightforward monoamine neurotransmitter deficits (e.g., decreases in norepinephrine, serotonin, and/or dopamine) serve as the biological substrate of depression. Lastly, immunologic studies have consistently found abnormalities of cytokines associated with depressive illness. Advances in all these biological correlates of depression hold great promise for the development of more specific and more effective treatments of depressive disorders in the not too distant future.

High Stress & Low Support

From a societal perspective, significant life stress, and/or lack of social support predisposes to development of MDD. Life stress that involves loss, for example, death of a parent or spouse, the end of a relationship, and events involving loss of self-esteem, such as termination from a job, create particular vulnerability for depression. Low social support, both independently and in the face of significant stress also predisposes to depressive disorder. Low perceived social support, that is, the extent to which an individual believes himself or herself to lack a supportive social network, creates a higher risk than any absolute or objective measure. (It is worth noting that these same risk factors of high stress and low support tend to increase risk for all illnesses, whether psychiatric or general medical illnesses.)

The stress caused by natural disasters also increases the vulnerability of survivors to depression. While the psychiatric impact of such disasters includes increased prevalence rates of posttraumatic stress disorder (PTSD), substance abuse, and other conditions, the increased rate of depression itself is significant and measurable. For example, children and adults in the tsunami-affected areas of southwestern Thailand showed significantly increased and persistent rates of depression—ranging from 6% to 30% depending upon level of exposure and level of life disruption. Similarly, war has always been a stressor with major mental health consequences. For example, the US ground troops in Iraq have shown an increase in prevalence of depression from 11.4% prior to deployment to 15.2% after deployment.

With increased life expectancy and the aging of the population in the United States, spousal caregiving of persons with disability, including dementia, is increasing. Caregivers (most often the female partner) of spouses with chronic neurodegenerative illnesses (e.g., Alzheimer disease) experience extreme physical and emotional burden. The role of caregiver presents a situation of both high stress and increasingly low support (as the caregiver progressively loses any emotionally meaningful relationship with the patient). Up to 40% of caregivers of patients with progressive dementia suffer from significant depressive symptoms or major depression.

Postpartum “blues” typically occurs in 50–80% of women within 1–5 days of childbirth and lasts up to 1 week. This “normal” reaction should be distinguished from postpartum depression, which occurs in 10–15% of women in the first 3–6 months after childbirth. Postpartum psychosis, which occurs in 0.5–2.0/1000 deliveries and typically begins 2–3 days after delivery, and is most common in individuals with a personal or family history of bipolar disorder. Postpartum psychosis is a highly acute psychiatric illness that usually requires mood stabilizers such as neuroleptic medications or lithium, and psychiatric referral (see Chapter 16).

The criteria for major depression require that five of nine symptoms be present for a 2-week period (Table 25-1). One of these nine symptoms must be either a persistent depressed mood (present most of the day, nearly every day) or pervasive anhedonia (loss of interest or pleasure in living).

Clinicians should realize that a depressed mood is not synonymous with major depression and is neither necessary nor sufficient for a diagnosis of major depression. Sadness (or tearfulness) does not constitute major depression (four other symptoms described in the following paragraph are necessary), and, conversely, major depression can be diagnosed without the presence of depressed mood (if pervasive anhedonia is present), a presentation that is more common in the elderly.

Organizing these nine symptoms into clusters of four hallmarks can facilitate clinical evaluation: (1) depressed mood; (2) anhedonia; (3) physical symptoms (sleep disorder, appetite problem, fatigue, and psychomotor changes); and (4) psychological symptoms (difficulty concentrating or indecisiveness, guilt or low self-esteem, and thoughts of death). Physical symptoms predict a favorable response to biological intervention. For example, when middle insomnia is present (awaking at 3 or 4 a.m. with an inability to return to sleep) and when a diurnal variation in mood is present (feeling more depressed in the morning), patients are more likely to respond to biological interventions.

The Fallacy of “Good Reasons”

Depression is often mistakenly believed to be an “expected” result of stressful life events. Studies of individuals under stress (e.g., terminal cancer or natural disaster) do show rates of major depression above the general population rate, but these rates do not exceed 50%. Although sad or depressed affect is an expected accompaniment of a stressful event, the full syndrome of major depression certainly does not appear in everyone. Thus, life stressors may seem to provide “good reasons” for sadness, but a stressful event, in itself, should not be considered a rationale to withhold depression treatment. If a major depressive syndrome emerges following a stressful life situation, the medical provider should treat it appropriately. This is similar to treating other medical conditions that have a “good reason,” such as hypovolemic shock from a gunshot wound—the etiology is evident but the condition clearly requires treatment.

The term reactive depression has historically suggested a mild syndrome without a biological substrate, resulting from a psychosocial precipitant, and treatable with psychotherapy alone. None of these assumptions is true. Stressful events can precipitate very severe depressions; a biologically predisposed individual may suffer major depression in response to a minor life event; a major depression following a life stressor may develop a biological substrate; and a major depression from a life stress may respond to biological therapy as well as or better than it responds to psychotherapy. Thus, the presence or absence of identifiable precipitants is irrelevant to the diagnosis of major depression, which can be treated pharmacologically whether or not the condition resulted in part from psychological stressors. Depression is a frequent, but not an inevitable consequence of a stressful life event. When major depression is present, it should be treated aggressively.

The Confound of Overlapping Etiology

A comorbid general medical condition (such as cancer or Parkinson disease) may seemingly “cause” many of the physical symptoms of major depression, such as fatigue, anorexia, or psychomotor retardation. These symptoms may lead clinicians to discount their relevance and thus disregard the possibility of a treatable depression. However, a physical symptom may be “caused” by a physical illness (e.g., cancer) or by a major depressive illness or both. It is important to include these symptoms in the initial diagnostic approach to depression in the medically ill, and exclude them only if they are clearly and fully accounted for by the physical illness. Although this “inclusive” approach might seem to result in the overdiagnosis of major depression; studies in stroke, Parkinson disease, hospitalized elderly, and traumatic brain injury indicate that the problem of overdiagnosis is quite low (around 2%) and the underdiagnosis of depression is more of a problem.

Build Trust by Responding to Distress

The medical interview holds the key to the assessment of major depression. Efficient assessment involves attention to data-gathering as well as rapport-building functions of the interview. Physicians should always be alert for nonverbal cues of depression: for example, a sad mood may be communicated by downcast eyes, slow speech, wrinkled brow, or tearful affect. When a depressed mood is detected or emotional distress is suspected, physicians should first respond empathically to this distress, by demonstrating a caring attitude, and using attentive silence or direct reflective and empathic statements, such as “I can see you’re having some trouble,” or “It sounds like you’ve been under a lot of stress lately,” or “You seem pretty down right now.” Responding directly to the patient’s distress in this way builds trust and encourages the patient to more openly share his or her feelings that may underlie a depressive illness.

Use Direct, but Open-Ended Questioning

Use open-ended questions and facilitation techniques to provide patients with the opportunity to discuss the issues that may be troubling (see Chapter 1). In gathering data for assessment, physicians should focus on anhedonia (e.g., “What do you enjoy doing these days?”) and depressed mood (e.g., “How has your mood been the last few weeks?” or “Have you been feeling sad, blue, or down in the dumps?”). These simple questions can effectively uncover an underlying depression in most patients, despite the fact that most depressed patients in the general medical setting initially present with chief complaints more related to physical and bodily symptoms (e.g., headache, fatigue, and insomnia).

Involve the Family

Optimal assessment and management of the depressed patient is enhanced by involvement of one or more significant other(s). A spouse, a partner, a parent, or others can help the physician gather useful information regarding the patient’s mood, activities, behaviors, and history. In fact, because of stigma, denial, and other psychosocial barriers, other persons often provide much more accurate information regarding depressive illness than the self-report of the patient himself or herself.

The Patient Health Questionnaire: Screening, Assessment, Engagement, & Monitoring

The US Preventive Services Task Force (USPSTF) supports screening for depression “in clinical practices that have systems in place to assure accurate diagnosis, effective treatment, and careful follow-up.” Although many tools to screen for depression are available, USPSTF recommends the use of a straightforward two-item (“yes/no”) screener for major depression that is as effective as longer screening instruments. When administered verbally, clinicians can ask:

1. “Over the past 2 weeks, have you ever felt down, depressed, or hopeless?” and

2. “Over the past 2 weeks, have you felt little interest or pleasure in doing things?”

A positive answer to either of these two questions requires a full diagnostic assessment for major depression. This screener has over 90% sensitivity but relatively low specificity (many false positives). Alternatively, many experts now advocate administration of the full PHQ-9 (Appendix 25-A) to “red flag” patients, that is, patients likely to be at high risk of major depression. “Red flag” patients generally include those with chronic medical illness (e.g., diabetes), and patients with persistent unexplained medical complaints. This one-step approach, combining screening and assessment, can simplify operational strategies.

The PHQ-9 is an assessment and severity tool that has been validated for use in general medical as well as specialty psychiatric settings. A score of 10 or more has been demonstrated to have 88% sensitivity and 88% specificity for the diagnosis of major depression. Furthermore, the tool can also be used effectively as a severity tool to track patients’ symptom severity and improvement over time. The instrument and scoring key are in Appendices 25-A and 25-B. Pfizer holds the copyright for this tool, but it can be used in the public domain for clinical purposes or research. Readers are also referred to the web site of the MacArthur Foundation Initiative on Depression in Primary Care for materials related to the use of this instrument (www.depression-primarycare.org).

Patient Barriers: Somatic Presentations & Stigma

Patients in general medical practice with major depression typically present with physical complaints, such as pain (headache, backache), fatigue, insomnia, dizziness, or gastrointestinal (GI) problems. Many of these patients are willing to acknowledge feelings of depressed mood and to consider the possibility that biologically mediated depression may also cause or exacerbate their physical problems. There are many somatically preoccupied patients, however, who do not experience any depressed affect at all, and, without this self-appraisal of sadness, they are reluctant to consider the possibility that depression may contribute to their physical problems. In these patients, evaluating both general medical and psychiatric problems simultaneously saves time, expense, and frustration for both the physician and the patient. In addition, many patients and families (particularly in some cultures) are reluctant to accept the diagnosis of depression because of associated social stigma. Physicians can help overcome this barrier by understanding and explaining to patients and families that depression is a common and treatable illness, like other medical illnesses. They may explain that depression represents a chemical imbalance that can, like diabetes and other medical conditions, be corrected or managed with adequate treatment.

Clinician Barriers

Depression is often undetected or is not adequately treated in the medical setting. Some physicians avoid depression diagnoses because they harbor the same stigmatizing attitudes toward depression that many of their patients feel. In addition, inadequate knowledge and skill, lack of time, reluctance to “open up” new domains of emotional distress, and misaligned financial incentives all operate as barriers to physician recognition and treatment. However, early recognition of behavioral and psychiatric disorders is time efficient in the long run while minimizing the cost and risk of extended, unnecessary workups for nonspecific physical complaints.

Suicide risk must be evaluated in all patients with symptoms of depression. It is one of the top 10 causes of death in all age groups, and one of the top 3 causes in young adults and teenagers. Risk factors for completed suicide include gender (elderly white males are at highest risk), alcoholism, psychosis, chronic physical illness, lack of social support, recent humiliation, and use of generally lethal methods (e.g., gun rather than overdose of pills). Increased risk of suicide has also been noted among depressed adolescents and among gay and lesbian patients. Explicit suicidal intent, hopelessness, and a well-formulated plan indicate relatively higher risk. Many patients who eventually commit suicide visit a primary care physician in the weeks before they take their lives.

Physicians are sometimes reluctant to explore suicidal ideation in the mistaken belief that asking about suicide may actually increase a patient’s risk. To the contrary, assessment of suicidal tendencies usually reassures patients, reduces anxiety for both patient and provider, and facilitates partnership in suicide prevention.

The assessment of suicidal ideation is best approached gradually, with general questions like, “Do you sometimes feel that life is not worth living?” and then asking more specifically about a history of suicide attempts, any specific current plans, hopelessness, and any specific current intentions. Once a patient reveals suicidal ideation, the physician must consider psychiatric consultation and hospitalization. If outpatient management is considered, some experts suggest that physicians consider use of a “no suicide contract.” The no suicide contract involves asking the patient to promise that he or she will contact the physician (or other appropriate caregiver) if there is a danger of losing control of a suicidal impulse. In using such a “contract,” however, physicians need to realize that there is no convincing empirical evidence to support its validity. In fact, other experts specifically advise against its use, arguing that a mechanistic pursuit of obtaining a “contract” can functionally undermine an open relationship and provide physicians with a false sense of security. The main utility of the “contract” may be as tool to discuss the strength of the individual’s suicidal ideas.

In all cases, when treating depression, physicians must evaluate suicidality at the initiation of treatment and throughout the treatment program.

Routine use of the PHQ-9 can aid in assessing suicide risk at the initiation of treatment and, of equal importance, can also aid in the recognition of any subsequent or treatment-emergent suicide risk. Because the risk of suicide sometimes increases within the first few weeks of treatment and can emerge at any point in the subsequent treatment, regular and routine use of the PHQ-9 can function as an efficient and effective suicide reassessment tool. The Columbia-Suicide Severity Rating Scale is a useful clinician-administered instrument (http://www.cssrs.columbia.edu/about_cssrs.html).

There are no specific diagnostic signs of depression. A careful medical history and physical examination are required for the evaluation of depression at all ages, but especially in the elderly. Some medical “mimics” of depression (e.g., hypothyroidism and Cushing syndrome) present with classic physical signs.

No laboratory studies can be used to diagnose major depression reliably or specifically. A general laboratory screen (complete blood count, chemistry profile, urinalysis, TSH, and vitamin levels), however, may be useful in selected patients to rule out other conditions that may mimic or exacerbate depression. In treatment-resistant cases, or when indicated, brain imaging, an electroencephalogram (EEG) or lumbar puncture (LP) can be considered, but these studies are not part of the standard work up. Patients over the age 40 years usually require an electrocardiogram (ECG) to rule out conduction disturbances or bradycardia if treatment with a tricyclic antidepressant (TCA) or if citalopram at doses greater than 40 mg/day are being considered since those doses may be associated with prolonged QT intervals and arrhythmias

Mental Disorders

Other mental disorders often present with symptoms similar to depression; in addition, depression often presents in combination with other mental disorders. Thus, knowledge of other mental disorders common in medical practice is essential for adequate assessment and management of depression. In the presence of psychiatric comorbidity, effective treatment of depression may lead to improvement in the other condition as well. Modifications of treatment, however, may be necessary depending on the particular comorbidity present.

Major Depressive Disorder versus Bipolar Depression

One of the most important, yet difficult, differential diagnostic questions facing the physician is to distinguish MDD from bipolar depression (discussed earlier). The clinical signs and symptoms of the two disorders are identical and the differential diagnosis hinges on one and basically only one crucial historical question: Did the patient ever experience clinical mania or hypomania? The symptoms of a manic episode are listed in Table 25-2, the most common of which include an elated or irritable mood, racing thoughts, poor judgment in interpersonal, sexual, or financial situations, and excess energy. Criteria for hypomania are the same, but are less intense and not disruptive of normal functioning. To uncover a possible history of mania/hypomania, the physician should ask about any personal or family history of treatment of mania/hypomania/bipolar disorder and whether the patient has ever experienced any distinct period when he or she experienced racing thoughts, markedly decreased need for sleep, especially high energy or unusually poor judgment, out of character to the patient’s usual behavior. The Mood Disorders Questionnaire (MDQ) can be added to the physician’s diagnostic armamentarium as a guide (Table 25-3), though its sensitivity and specificity are not high enough to rely on as a stand-alone tool. Even if there is no history of mania/hypomania, if the individual has a strong family history of bipolar disorder, the clinician must consider that at least one-third of bipolar patients have depression as their index mood episode. Such a history should help make the clinician alert to potential mood switches with antidepressant treatment. In addition, such patients may also be resistant to antidepressant strategies and require mood stabilizers instead.

Anxiety Disorders

Anxiety and depression commonly co-occur in medical patients. Most patients with depression suffer from anxiety symptoms or a formal anxiety disorder and most patients with an anxiety disorder have depressive symptoms or meet criteria for major depression. The most common anxiety disorders in medical outpatients are PTSD, generalized anxiety disorder (GAD), panic disorder (PD), and obsessive-compulsive disorder (OCD). Central features of these disorders include pervasive and disabling anxiety (e.g., GAD), discrete panic attacks (e.g., PD), or the presence of unreasonable, uncontrollable, and repetitive behaviors (e.g., compulsions such as hand-washing or checking the stove) or the presence of intrusive, uncontrollable, and disturbing thoughts (e.g., sexual and violent obsessions in a patient who has no history or risk of such behaviors) (e.g., OCD). Treatment of the major depression by itself, however, often helps to resolve or improve these other coexisting conditions (see Chapter 26), especially since many antidepressant medications have proven safe and effective for treating some common anxiety disorders.

Somatic Symptom Disorders

Depression often presents with unexplained bodily complaints. It can therefore be challenging to differentiate between a depressive illness and somatic symptom disorder which includes previously used terms such as medically unexplained symptoms (see Chapter 28). Depressive disorders are highly treatable, but somatic symptom disorders can be more chronic and refractory to treatment. Somatic symptom disorders are usually best managed conservatively with a focus on improved functioning, whereas depression should be treated aggressively with the goal of complete recovery. Any of the somatic symptom disorders (formerly diagnosed as conversion, somatization, hypochondriasis, body dysmorphic disorder, and somatoform pain disorder) can present comorbidly with major depression. Approximately 50% of patients with persistent unexplained physical complaints suffer from depression. Effective treatment of major depression usually improves the severity, intensity, and functional impairment of a comorbid somatic symptom disorder.

Substance Abuse

Patients with alcohol dependence or other substance use disorders commonly present with major depression. Likewise, all patients diagnosed with depression should be screened for comorbid substance use disorders, particularly alcohol. Major depression even in the context of substance dependence deserves treatment in its own right, as long as the treatment plan attends to potential complications from continuing substance abuse. Furthermore, effective treatment of depression may help ameliorate the alcohol problem itself and does not generally lead to increased complications. Some patients will benefit from referral to a specialty mental health or substance abuse facility for treatment of their substance dependence and some severely ill patients will need treatment in settings with the capacity to treat “dual diagnosis” patients.

Personality Disorders

Personality disorders represent enduring character patterns that are deeply ingrained and are not generally amenable to alteration (see Chapter 29). They often complicate the diagnosis and management of mood disorders. Because patients with personality disorders can be difficult and demanding, physicians often try to minimize contact with them. Unfortunately, this may lead to avoidance of emotional issues and failure to diagnose depression. Effective treatment of major depression often improves functioning when the depression coexists with a personality disorder, even if the underlying personality disorder is itself not fundamentally changed.

Dementia

In its early stages, dementia can be often difficult to distinguish from depression. Depression often leads to reversible cognitive impairment in the form of decreased concentration, memory difficulties, impaired decision-making ability, difficulty planning and organizing, and difficulty getting started on tasks. These are also impairments that can result from an insidious and irreversible neurodegenerative process. Likewise, the effect of dementia on a person’s functioning can lead to a depressed mood. When diagnostic uncertainty exists, because the cognitive impairments of MDD are reversible, clinicians should treat the depressive component (with both medication and counseling) and observe for changes in the patient’s cognitive symptom cluster.

The term “pseudodementia” is used to describe treatable depression presenting with features of cognitive impairment. While it is useful for clinicians to understand that depression may be associated with cognitive impairment, clinicians should avoid use of this terminology. In actual practice, some patients for whom the concept of “pseudodementia” is considered may actually suffer from both conditions. Furthermore, late-onset depression (even reversible late-onset depression) is itself predictive of future dementia.

Depression due to General Medical Conditions or Medications

Approximately 10–15% of all depression is considered to be the direct physiologic result of a medical illness, such as hypothyroidism or hyperthyroidism, pancreatic cancer, Parkinson disease, or stroke (Table 25-4). Because there are no clear criteria to help guide clinicians in their evaluation, this diagnosis is ultimately made by clinical inference, considering the timing of the depression in relation to the physical illness. Limited data seem to indicate that standard treatments for major depression are effective in these cases.

Similarly, depression can be caused by exogenous medications (Table 25-5). Approximately 50% of patients treated with interferon develop a full major depressive episode and the history of previous depression or the presence of even mild current depressive symptoms can reliably predict the development of subsequent depression during treatment with interferon. Evidence now supports the prophylactic use of antidepressants before initiating interferon treatment, especially in patients with histories of depression. No medication has been noted to “cause” depression in all patients, so it is therefore crucial to carefully evaluate the patient’s clinical history and link the onset of depressive symptoms to the initiation of new medications or changes in the current regimen as well as considering psychosocial stressors to obtain the clearest picture of depression causation.

Use Communication Skills, PHQ-9 Scores, & the Family for Engagement

Building on the patient’s trust, the working alliance, and the doctor–patient relationship, the physician should integrate his or her understanding of the patient’s symptoms and life situation into a clear presentation of the diagnosis to engage the patient for collaborative management.

In addition, the process of the clinician’s communication with a depressed patient should be informed by an appreciation of the patient’s slower rate of cognitive processing. Information should be given in small chunks, allowing adequate attentive silence for the patient to assimilate and respond. In addition, the physician should verify adequate assimilation of information by asking the patient to summarize what has just been communicated. The following statement would be appropriate:

“Doctors sometimes have a hard time communicating clearly with their patients. Just to be sure I have been able to make my thoughts clear to you, would you mind summarizing what we just went over.”

By using the PHQ-9, practitioners can integrate these results into the engagement process. The first step is to check some of the symptoms with the patient to verify the validity of specific responses. After verifying that the actual score on the PHQ-9 appears to accurately capture the patient’s symptoms, discuss the diagnosis and engage the patient for collaborative treatment planning. Because some patients may feel shame or embarrassment with a diagnosis of depression, it may be helpful for some patients to explain that major depression is a common biological disorder. Drawing a picture of a synapse and neurotransmitters also may be helpful for some.

A crucial part of the engagement process involves instilling hope that this painful illness is treatable. The treatment involves mobilizing patient resources and/or external resources such as medication and social support. Indicate that depressive symptoms can be relieved and note that “others may notice improvements in you before you notice the change yourself.” When adverse life circumstances play a role in the cause of major depression, the physician can acknowledge the role played by the stressors, but must also help the patient understand that treatment of the depression can help him or her cope better with life’s adversities.

The physician should attempt to involve the family or other support systems (however, defined) in treatment planning to help ensure adherence to the collaborative plans. Keep in mind, however, that family and/or friends may at times sabotage or undermine treatment and impede recovery. Approximately 40% of depressed patients suffer from some degree of discord in the relationship that contributes to illness onset and/or becomes a barrier to recovery. When discord is detected, it should be acknowledged and included in the overall treatment.

Treatment Goals: Defining Clinically Significant Improvement, Response, & Remission

A wide variety of biological and psychological therapies have been shown in randomized clinical trials to be effective for treatment of depression. The concept of “effectiveness,” however, hinges on the demonstration of a “response” to treatment defined as “50% improvement in symptoms.” A 50% improvement is important clinically, but this criterion is broad enough to leave significant residual symptoms unresolved. The concept of “remission,” therefore, has been introduced to underscore the importance of helping the depressed patient reach full return of function and achieve relative absence of all depressive symptoms.

The following sections review broad treatment guidelines, psychotherapies, biological treatments, self-management support (SMS), and the need for objective monitoring of symptoms to help the patient attain full remission. Using the Case Illustration of Mrs. Gladstone, a 5-point decrease in her PHQ score from 16 to 11, would be considered a clinically significant improvement (the “minimally important clinical difference”). A decrease of 50%, from 16 to 8, would be considered a treatment “response,” but the score of 8 reflects the persistence of residual symptoms of depression. Not until her PHQ score goes below 5, can Mrs. Gladstone be considered to be in remission.

General Treatment Guidelines: Close Monitoring, Persistence in Treatment Trials, & Chronic Reassessment

Research indicates a disappointingly low overall 50% response rate and a 33% remission rate for the first antidepressant a physician selects for the treatment of depression. Using logical strategies to switch or combine antidepressants after the first failure, about 70% of all patients will eventually respond to treatment. And we also know now that depression is a chronic disease. At least 50% of patients with a major depressive episode will have a second episode, and patients who have had two or more episodes of major depression have a 75–90% likelihood of recurrence.

These results suggest the following 7-point treatment program to encourage close monitoring, persistence of treatment trials, and chronic care for patients suffering from major depression. These points are all aligned with current evidence-based guidelines.

1. Use the PHQ-9 to help confirm the diagnosis and establish baseline severity.

2. For those treated with medication, patients should have an “early follow-up” contact, visit or phone call (1–3 weeks) to check on adherence to treatment, side effects, or potential change in symptoms for better or worse (e.g., emergent suicidality).

3. The first goal of treatment should be achieving a 5-point decrease in PHQ-9, which can be considered a clinically significant improvement. General target goal (which is very difficult to attain) should be achievement of a 5-point decrease in PHQ-9 score every month until a score below 5 is attained.

4. Patients should be evaluated with a repeat PHQ-9 once per month until remission is achieved (PHQ-9 <5). In general, the treatment approach should be modified if the patient does not show continuing improvement in PHQ-9 scores. (Modifications could include, for example, adding medication to psychotherapy or the converse, increasing the dose of a medication, switching medication, augmenting medication, considering couple therapy, reevaluating the diagnosis, and so on.)

5. Once patients reach remission, treatment should be continued (at the current dose of medication(s), or decreased frequency of psychotherapy) for at least 6–12 months.

6. For patients achieving remission with medications, evidence is clear that maintenance therapy requires the same dose of medication used for recovery. For patients achieving remission with psychotherapy, evidence is less clear, but there is some indication that intermittent “refresher” maintenance psychotherapy sessions can help prevent recurrences.

7. Patients with a history of depression should receive PHQ-9 “check-ups” every 6–12 months for the rest of their lives.

Choosing Initial Treatments: Respect Patient Preferences

Evidence-based treatment for depression includes antidepressant medication, several forms of psychotherapy, a combination of medication and psychotherapy and, in refractory cases, ECT. Electroconvulsive therapy (ECT) remains the single most efficacious treatment available for severe depression, and can be lifesaving.

Management of depressive illness, like any other chronic illness, should be collaborative. After providing basic education, physicians should offer patients realistic treatment options and respect patient preferences. Patients tend to be more adherent to a treatment if they have selected it. Patients may need to be reminded that, even if they are pessimistic about treatment, such a perception is due to depression affecting their outlook. If they can stick with the treatment they have an excellent chance of responding to treatment. They should also be told that without treatment, they are likely to continue to suffer from depressive symptoms for a long time.

Individual Office Counseling by the Physician

Counseling by the primary care physician anchors all treatment modalities (pharmacotherapy, care management, psychotherapy) and provides the relational framework for continuity with the patient’s other health care needs. Physician counseling provides the context for discussing the diagnosis, instilling hope for change, mobilizing the patient’s self-management, and planning treatment. The time frame for such counseling can be a few minutes to an extended 30-minute visit.

Some physicians may want to engage in more specific counseling. The patient, however, should be notified that this interaction should not be considered formal psychotherapy, unless the practitioner has been adequately trained. Psychotherapeutic situations invariably arouse strong emotions in both patients and physicians. When complex interpersonal issues or strong feelings emerge during office counseling, the physician should seek supervision from a trained therapist or consultation from a colleague.

The acronym SPEAK was developed by one of the authors (JFC) to help physicians in counseling depressed patients (Table 25-6). The five components of SPEAK (Schedule, Pleasurable activities, Exercise, Assertiveness, and Kind thoughts about oneself) are grounded in core elements of evidence-based psychotherapies of depression including behavioral, interpersonal, and cognitive approaches. They provide a framework both for patient education and for ongoing supportive counseling by the physician. The SPEAK summary in Table 25-6 can be given to the patient as a handout to use as a component of guided self-help (see below).

Including the Patient’s Partner

For patients in a significant relationship, including the partner in conjoint counseling can place the patient’s symptoms and recovery within a relational context. Engaging the partner as an ally in the patient’s self-management activities can increase adherence as well as external motivation in the context of a caring relationship. It can also help the partner understand the patient’s symptoms as part of the illness being treated. In working with couples, practitioners can utilize some of the following interventions, all of which now have an emerging evidence base. Again, make it clear to patient and their partner that these interventions are not a substitute for therapy by a mental health specialist.

1. Listen attentively to the patient and their partner; discuss their current life stresses and emotional difficulties. Allowing straightforward “ventilation” can be therapeutic in itself and you will learn the extent to which discord in the relationship may be a factor in the depression. In addition, by expressing empathy to each partner, the physician can model appropriate interpersonal behavior for the discordant couple and build mutual support and trust.

2. Straightforward education about the biological nature of depression is a necessary component of all treatment, but in the context of discord in a couple it can help both partners view the depression in a more “forgiving” manner. In a common, but painful and maladaptive “dance,” the discordant partner typically expresses a blaming attitude toward the depressed partner, which the patient accepts wholeheartedly because it resonates with his or her own highly negative cognitive self-appraisal. Similarly, an educated spouse is more likely to be able to view the depressed individual’s depressive irritability as a symptom needing to be addressed, rather than a personal attack. By reframing the illness as an involuntary biological condition, the practitioner helps build the relationship and aids medication adherence (if antidepressants are part of the treatment plan).

3. Encourage or facilitate joint problem solving. Helping the depressed patient and their partner identify two or three troubling life problems they face and helping them develop a shared, concrete behavioral solution can help build their alliance and contribute to recovery. However, continuing or persistent discord may undermine treatment or recovery and should be referred for formal couples counseling.

When involving the partner in assessment or management of depression, it is essential to obtain informed consent first.

Psychotherapy

The psychotherapy strategies commonly used in treating depression include cognitive behavioral therapy (CBT), interpersonal therapy (IPT), psychodynamic therapy (PDT), problem-solving therapy (PST), hypnotherapy, and family therapy (including couple therapy). Some of these therapies can be administered in individual or group sessions (CBT, IPT, and PST). In addition, some therapies can be offered via the Internet (real-time text messaging, email, or video conference) or by telephone. Psychotherapy is typically provided by a licensed mental health therapist (psychiatrist, clinical psychologist, clinical social worker, psychiatric nurse practitioner).

Cognitive behavioral therapy attempts to identify and challenge cognitive distortions (e.g., “I fail at everything”) that are pessimistic or self-critical thoughts which contribute to and maintain the depressed mood. Once these negative cognitive patterns are identified, the patient learns to challenge them and to develop counter-messages that are more evidence based and positive. In addition, CBT teaches “behavioral activation” to counter the inertia of depression (e.g., engaging in pleasurable activities in the presence of anhedonia, or engaging in physical exercise in the absence of motivation). Evidence supports the effectiveness of CBT in treating depression, both as a single problem and as a comorbidity with somatic illnesses. Cognitive behavioral therapy combined with pharmacotherapy has higher effectiveness than CBT alone in treating older adults, particularly in patients with chronic, treatment-resistant depression.

There is also evidence that several variants of CBT, such as mindfulness-based cognitive therapy (MBCT) and cognitive behavioral analysis system of psychotherapy (CBASP) may be effective in the treatment of depression. Mindfulness-based cognitive therapy utilizes mindfulness meditation to teach patients to recognize and let go of ruminative thinking by enhancing “metacognitive awareness,” which is a nonjudgmental observation of the passage of thoughts and feelings and how they are constructed. Patients undergoing this approach learn to experience negative thoughts and feelings as mental events, rather than as the self (see Chapter 7).

Within the domain of CBT, two methods of therapy have emerged in recent years with a promising evidence base for treatment of depression: Dialectical behavior therapy (DBT) and acceptance and commitment therapy (ACT). Dialectical behavior therapy was originally developed to treat the parasuicidal behavior of patients suffering from borderline personality disorder. The treatment is dialectical in that its goal is to synthesize the alternative principles of acceptance and change. Acceptance strategies are similar to the notion of unconditional positive regard in client-centered therapy. Change strategies have the objective of changing thoughts and overt behavior. Dialectical behavior therapy addresses four skill sets: mindfulness, interpersonal effectiveness, emotion regulation, and distress tolerance. Group DBT treatment has shown positive effects in reducing symptoms in patients with treatment-resistant depression.

Acceptance and commitment therapy conceptualizes psychopathology as stemming from unhealthy efforts to control emotions, thoughts, memories, and other private experiences and a lack of clarity about one’s core values and the ability to act in accordance with them. The goal of ACT is to diminish the role of thought while encouraging the patient to embrace experience directly without suppression. At the same time it encourages behavior consistent with one’s chosen values. Compared to traditional cognitive therapy, ACT has shown equal effectiveness in reducing symptoms of depression.

Interpersonal therapy focuses on one or more of several interpersonal situations that may relate to initiating or sustaining depression: interpersonal conflict, role transition, grief, or social skills deficits. Assertiveness training is one component of IPT that develops patients’ capacity to be direct in communicating feelings, thoughts, and preferences in their current relationships. In the treatment of depression, IPT has been shown to be more effective than placebo and equivalent to medication in adults. In late life depression, medication was more effective. Overall, CBT and IPT have the broadest evidence base for effectiveness in treatment of depression.

Psychodynamic therapy focuses on interpersonal relationships and unconscious feelings, desires, strivings, and thoughts to treat symptom disorders. It involves using the therapeutic alliance to gain an understanding of the ways maladaptive defense mechanisms have developed and contribute to depression and then working through these mechanisms in an effort to build more adaptive and less punishing defense mechanisms. A meta-analysis of short-term dynamic psychotherapy (STDP) found that it had significant effects pre-to posttreatment for depressive symptoms and that these changes endured at 1-year follow-up. Although other forms of psychotherapy showed a greater effect immediately posttreatment, there were no differences at 1-year follow-up.

Problem-solving therapy teaches patients to break down larger life problems into smaller elements and to identify specific steps to address these elements. It has been shown to be as effective as antidepressant medication in the treatment of depressed primary care patients. Meta-analysis has shown a modest effect of PST on depression, with greater effects noted in group treatment with PST.

The evidence from many randomized clinical trials is that these psychotherapies are as effective as antidepressant medication (in mild-to-moderate MDD) in achieving a significant reduction of symptoms (over 50% response rate) after 10–16 weeks of treatment. Although the response to antidepressants is generally more in the first 4 weeks, response to psychotherapy catches up and by 12 weeks the efficacy of medication and psychotherapy are roughly similar. There is some reason to believe that antidepressant medication and psychotherapy of depression achieve their benefits through different neurobiological mechanisms, so it is not surprising that combination treatments (medication and psychotherapy) have generally been found to be more effective than either one alone. For individuals with severe depression, antidepressants are generally important mainstays of treatment and more effective than psychotherapy alone.

Cognitive behavioral therapy is generally considered the treatment of choice for depression in children and adolescents, though one medication (fluoxetine) is FDA approved for the treatment of child and adolescent depression. The clinical benefits of psychotherapy are usually seen in 8–12 weeks, and the duration of treatment is typically 6–16 sessions. Meta-analysis has shown that psychotherapy is moderately effective in the short run, especially for adolescents aged 12–18 years, but that after 6 months psychotherapy does not show enduring effects.

Guided self-help involves the patient working at home with a standardized psychological treatment, such as CBT, with guidance from a physician or mental health therapist, who provides initial support and occasional coaching. Therapist support can be face-to-face or via telephone or email. The standardized treatment can be in various formats—book, audio CD, video, or Internet. Remote psychotherapy via phone or Internet also has shown promise for the treatment of depression.

Considerations for choosing psychotherapy in the acute phase of treatment are shown in Table 25-7. Besides being an adjunct to medication for the treatment of depression, psychotherapy may also have a role in prophylactic maintenance therapy for the prevention or delay of future depressive episodes in patients susceptible to recurrences. For example, MBCT has been shown to reduce depressive relapse rates at levels similar to maintenance antidepressants in several randomized controlled trials. Psychotherapy can also be useful for women with major depression who want to become pregnant and bear a child in a medication-free condition or for other patients with major depression who must be medication free for limited periods. In mild-to-moderate depressions, psychotherapy alone may be as effective as psychotherapy combined with antidepressants.

Marital and Couple Therapy

As mentioned above, marital or couple therapy may be an appropriate choice of psychotherapy, especially when marital problems are entwined with depression. Marital therapy for depressed women who are also in troubled relationships can be effective in alleviating the depression as well as in decreasing the marital discord.

Marital therapy is no more or less effective than individual therapy, even when depression is associated with marital distress. Although individual therapy for depression can be effective in decreasing depressive symptoms even in the presence of marital discord, if the discord is not addressed relapse into depression is more likely. Moreover, it is useful for the physician to get some idea of what the patient believes caused the depression. If the patient believes that the marital discord caused the depression, then addressing the discord is even more important for the prevention of relapse.

Exercise & Physical Activity

Evidence from randomized clinical trials is inconclusive on the benefits of moderate exercise in the treatment of depression. In the Standard Medical Intervention and Long-term Exercise (SMILE) study, 16 weeks of aerobic exercise training was comparable to that of standard pharmacotherapy with sertraline and combined exercise and medication. In a 10-month continuation study, remitted subjects in the exercise group had significantly lower relapse rates than subjects in the medication group. A more recent meta-analysis of exercise compared with other therapies for depression found that exercise did not differ from CBT and antidepressant medication. When rigorous methodology was applied to these studies, the affect of exercise was only moderate.

Self-Help Books

For motivated, literate patients, self-help books that increase understanding of depression and provide guidance on self-management strategies can be an effective adjunct to treatment. For patients without access to psychotherapy, the use of self-help materials can be a helpful substitute. A meta-analysis of studies examining the use of self-help books in the United Kingdom found that one book, Feeling Good, had a large treatment effect compared to delayed treatment.

Evidence demonstrates that antidepressant medications are effective for the treatment of major depression and dysthymia. There are no compelling data, however, to support the use of antidepressant medication for patients with adjustment disorders or other minor depressive disorders. The initial treatment of choice for most patients with an adjustment disorder or other minor depressive episode is “watchful waiting,” which consists of physician support, office counseling, and close observation, with repeat assessment to document improvement, remission, lack of improvement, or possible transformation into major depression. Physicians can also support patient self-management goal setting and problem solving with Ultra-Brief Personal Action Planning (UB-PAP) (discussed earlier). Patients with minor depression who do not improve after 3–6 months of watchful waiting can be treated empirically with a trial of antidepressant medication or referred for psychotherapy.

Communicating With Patients About Medication

Patients initiating antidepressant medication for the first time sometimes are concerned about the stigma associated with their use. Some fear getting “hooked” on the medicine or that it will change their personality in some way. It is important to explain that antidepressants are nonaddictive medicines that can restore the natural balance of neurotransmitters and normalize neural pathways in the brain. It can be helpful to inform the patient about the possibility of experiencing some of the known side effects of the particular antidepressant and to refer to those side effects as an indicator of the drug’s potency to achieve the desired results. Finally, it is important to build a realistic expectation about the tempo of therapeutic effectiveness by letting the patient know that even though the symptoms of depression may persist for a week or two after starting the antidepressant, the healing process has already begun.

Adherence to treatment has been shown to be enhanced when the physician–patient alliance is optimized, when the patient’s attitudes toward and experience with treatment are thoroughly explored, and when the patient’s opinion in treatment decisions is explicitly valued.

Pharmacotherapy Options

The cyclic antidepressants include the TCAs, SSRIs, serotonin norepinephrine reuptake inhibitors (SNRIs), and a noradrenaline and specific serotonin antagonist (NaSSa). The cyclic antidepressants are generally equally effective; however, due to the adverse effect profile of the TCAs, they have been supplanted by the SSRIs, SNRIs, and NaSSa agents. Nevertheless, TCAs are still commonly employed as second-line antidepressants in patients who do not respond to first-line agents and for augmentation therapy.

Antidepressants illicit immediate first-dose pharmacological effects on the catecholamine reuptake system; however, these agents only begin to provide some symptom improvement after 1–2 weeks of treatment. An adequate trial of an antidepressant requires a minimum of 4–6 weeks at appropriate doses. Cyclic antidepressant response rates approximate 50–70%; hence, physicians should expect that up to half of the patients they treat with the first antidepressant may not respond. Patients with no response or minimal response should be switched to another agent. Patients with partial response can be managed with augmenting agents. If the first SSRI trial fails, patients may be switched to another SSRI (e.g., bupropion, venlafaxine, mirtazapine, or a TCA). Of note, 25–50% of patients who fail to respond to the first trial of an SSRI will respond to a different SSRI on a second trial. A list of antidepressant side effects, mechanisms of action, and dosages is listed in Table 25-8.

Tricyclic Antidepressants

The TCAs have fallen out of favor predominantly because of their adverse effects and toxicity profile in the overdose setting. Common antimuscarinic effects include dry mouth, blurred vision, constipation, urinary retention, and sinus tachycardia; histamine receptor blockade effects include sedation, drowsiness, and weight gain; and alpha-1 receptor blockade effects include orthostatic hypotension and sedation. The TCAs also possess quinidine-like effects, such as QRS and QTc prolongation, which can predispose to ventricular dysrhythmias and torsades de pointes—especially in the overdose setting, or in patients with preexisting cardiac disease. A TCA overdose is a medical emergency, with significant risk of lethality. Tricyclic antidepressants lower the seizure threshold, with the potential for refractory seizures in overdose. To minimize the risk of lethal overdose in patients with suicidal ideation, principles of prudent care suggest that only limited amounts of TCAs should be prescribed at any one time (e.g., 7–10 days supply, especially in higher doses). Secondary amines, such as desipramine and nortriptyline, are favored over tertiary amines, such as amitriptyline and imipramine, because of a lower propensity to induce antimuscarinic adverse effects, less orthostatic hypotension, and less sedation. The adverse effects of the TCAs are dose related, and can be minimized by starting patients on low doses, and titrating upward slowly. Though the newer antidepressants seem to be generally as effective as the TCAs for the broad category of major depression, these older agents may be more effective for the more severe depressive disorders and, especially depression with melancholic features.

Selective Serotonin Reuptake Inhibitors

There are five SSRIs approved for major depression in the United States—fluoxetine, paroxetine, sertraline, citalopram, and escitalopram. The SSRIs have replaced the TCAs as first-line agents for major depression because they are generally better tolerated and are not life-threatening in overdose. Selective serotonin reuptake inhibitors are virtually free of antimuscarinic, antihistaminic, and antiadrenergic side effects. All SSRIs are equally effective; therefore, choice of any particular SSRI is generally based on side effect profiles, pharmacokinetics, drug interactions, dosage forms, cost, formulary availability, and so on. Table 25-9 demonstrates comparative SSRI adverse effects. As a class, many SSRIs have also been shown to be effective for PD, GAD, OCD, social anxiety disorder, PTSD, premenstrual dysphoric disorder, and bulimia nervosa. Fluoxetine is the only SSRI with an FDA indication for the treatment of childhood or adolescent depression and fluoxetine paroxetine, and sertraline have FDA indications for the treatment of childhood and adolescent OCD. Vilazodone, a new SSRI, is also a partial agonist at the 5HT1A receptors and appears to have less sexual dysfunction and potential for weight gain. With the exception of fluoxetine and paroxetine, SSRIs show very little inhibition of liver isoenzyme systems. Because of a common SSRI discontinuation syndrome, marked by GI and autonomic signs and symptoms, physicians taking patients off these medications should taper them slowly. Withdrawal from paroxetine seems to be more problematic than other SSRIs. Brief use of the long-acting SSRI fluoxetine (for several days) can be used to help “cover” SSRI withdrawal effects when discontinuing other SSRIs. In the overdose setting, SSRIs can cause dose-related toxicity, manifesting with nausea and vomiting, tremor, myoclonus, dysrhythmias, and seizures—plausibly all manifestations of the serotonin syndrome.

Gastrointestinal Adverse Effects

The SSRIs inhibit serotonin reuptake, thereby activating 5-HT₃ receptors, which can cause nausea and vomiting in 10–15% of patients. Due to desensitization of the 5-HT₃ receptors, tolerance to nausea and vomiting usually occurs within 1–2 weeks. Short-term use of cyproheptadine (5-HT3 receptor antagonist) may decrease uncomfortable nausea and vomiting. To minimize GI irritation, nausea, and vomiting, it is useful to administer the SSRIs with food. Dose-related diarrhea is also a common adverse effect of the SSRIs, with the possible exception of paroxetine which possesses mild-to-moderate anticholinergic effects and can cause constipation.

Weight Gain & Weight Loss

Both weight gain and weight loss have been reported with the SSRIs. The incidence of SSRI-induced weight gain greater than 5–10% of body weight can approach 25%, especially in females and, especially in patients using paroxetine (possibly due to its anticholinergic effects). Addition of bupropion may decrease SSRI-induced weight gain. Anorexia and weight loss have also been reported during early treatment, especially in overweight individuals with carbohydrate cravings, underweight depressed patients, bulimics, or patients using fluoxetine.

Syndrome of Inappropriate Antidiuretic Hormone Secretion

A clinically relevant and underappreciated adverse effect of the cyclic antidepressants is the syndrome of inappropriate antidiuretic hormone secretion (SIADH). Syndrome of inappropriate antidiuretic hormone secretion presents with serum hyponatremia and hypoosmolality, and urinary hypernatremia and hyperosmolality. This syndrome is caused by an excessive release of antidiuretic hormone, presenting with signs and symptoms such as nausea, vomiting, weakness, fatigue, confusion, and seizures. Syndrome of inappropriate antidiuretic hormone secretion has been reported with all antidepressants, and is reversible upon dose reduction or discontinuation. Management also includes fluid restriction and demeclocycline. Syndrome of inappropriate antidiuretic hormone secretion is more common in the context of tobacco use.

Bleeding Diatheses

Many physicians, including psychiatrists are still unaware of recent reports (mostly retrospective case-controlled studies) associating SSRIs with a plethora of bleeding diatheses including GI bleeding, epistaxis, petechiae, purpura, and ecchymosis. Approximately 95% of serotonin is stored in the platelets, and serotonin is one of many chemotactic factors responsible for platelet aggregation. SSRIs possess an aspirin-like effect and inhibit platelet aggregation, but the potency and duration of this effect is not yet well elucidated. When SSRIs are administered with nonsteroidal antiinflammatory drugs (NSAIDs), the risk of NSAID-induced gastropathy can increase significantly. Physicians should be cautious when prescribing an SSRI in patients with a history of GI bleeding and in patients who are already taking other antiplatelet medications. In the context of concurrent NSAIDs, aspirin, or other antiplatelet therapy, physicians should consider adding a proton pump inhibitor for prophylaxis of gastropathy.

Sexual Dysfunction

Decreased libido, orgasmic difficulties or anorgasmia, penile anesthesia, and erectile dysfunction (impotence) are very common adverse effects associated with the SSRIs in men and women. Sexual side effects occur in 25–75% of all patients, a higher prevalence than reported in package inserts because of underreporting and the short duration of clinical trials. The only available cyclic antidepressants that are less likely to cause sexual dysfunction are bupropion and mirtazapine. Sexual dysfunction in males, especially erectile dysfunction has been successfully managed with phosphodiesterase type-5 antagonists, such as sildenafil, tadalafil, and vardenafil. Other medications such as bupropion, yohimbine, amantadine, buspirone, bethanechol, neostigmine, and Ginkgo biloba have all been utilized with mixed results, with no clear evidence of efficacy from randomized, placebo-controlled clinical trials. Some clinicians will attempt a drug holiday, and skip the dose of the SSRI 24 hours prior to sexual activity (not effective with fluoxetine because of its long half-life). Because sexual dysfunction has such a significant impact on quality of life, this problem may be the most common limiting factor in use of SSRIs.

Serotonin Syndrome

Poison control centers report over 27,000 toxic exposures to SSRIs annually, with 15% of these reporting a serotonin syndrome. The serotonin syndrome can occur in overdose or when two or more serotonergic agents are combined. Table 25-9 lists common serotonergic agents. The serotonin syndrome presents as a clinical triad of autonomic dysfunction (e.g., hyperthermia, labile blood pressure), neuromuscular dysfunction (e.g., clonus, hyperreflexia), and mental status changes (e.g., agitation, delirium). This spectrum of clinical findings can be mild and transient and include akathisia, tremor, and altered mental status; however, it can progress to life-threatening symptoms such as sustained clonus, muscular hypertonicity, and hyperthermia approaching 40°C. Severe serotonin syndrome is a medical emergency. Although it is important for physicians to be aware of the signs, symptoms, and risk factors for the serotonin syndrome, it is also important for physicians treating depression to understand that pharmacological combination and augmentation treatment strategies often include the cautious prescription of two or more medications with proserotonin effects. Bupropion does not have serotonin effects.

Venlafaxine

Venlafaxine is a bicyclic antidepressant, approved in 1994 as the first dual-acting antidepressant agent to inhibit the reuptake of both norepinephrine and serotonin. These agents are often referred to as SNRIs. At lower doses, venlafaxine exhibits relatively more serotoninergic activity; however, at doses above 150 mg daily the adrenergic and serotonergic effects become more balanced. Venlafaxine is also FDA approved for the management of generalized and social anxiety disorder and PD. The extended release form of the product (XR) has become preferred because it can be used once daily and has much lower propensity to cause nausea, vomiting, and increases in blood pressure (3% of patients on doses of 375 mg experience sustained elevation in blood pressure). Venlafaxine shows very low inhibition of liver isoenzyme systems and has much lower protein binding (about 30%) than most other antidepressants that are very highly protein bound. Discontinuation of venlafaxine is associated with an SSRI-like discontinuation syndrome, requiring gradual taper (or “covering” for withdrawal effects with fluoxetine). Desvenlafaxine is a newer preparation consisting of a metabolite of the desmethyl metabolite of venlafaxine. It offers simpler dosing at a starting dose of 50 mg that is the same as the full continuation dose. This is a modest advantage in that venlafaxine typically requires gradual increases in dosing to avoid GI side effects.

Duloxetine

Approved by the FDA in 2004, duloxetine is a bicyclic SNRI antidepressant indicated for MDD, GAD, and diabetic peripheral neuropathic pain, fibromyalgia, and chronic musculoskeletal pain. Duloxetine, like venlafaxine can produce modest increases in blood pressure, nausea, sweating, insomnia, dizziness, and sexual dysfunction. Duloxetine is metabolized by CYP-1A2 and -2D6, and is a -2D6 inhibitor. Of clinical importance, physicians should be aware that when duloxetine is administered with 1A2 inhibitors, such as fluvoxamine or ciprofloxacin, or 2D6 inhibitors, such as paroxetine, duloxetine serum concentrations can increase considerably. Duloxetine should not be used in patients with a creatinine clearance less than 30 mL/min due to significant accumulation of the parent (two-fold) and numerous metabolites (nine-fold). Duloxetine has been associated with hepatotoxicity; in controlled trials, 1% of the patients experienced elevations in liver enzyme tests three times the upper limit of normal. Duloxetine is best avoided in patients with a history of alcohol abuse or hepatic disease, and in patients receiving concomitant hepatotoxic agents.

Mirtazapine

Mirtazapine, a tetracyclic noradrenaline and specific serotonin antidepressant (an NaSSa agent), is a potent serotonin-2 (5-HT2), serotonin-3 (5-HT3), and central alpha-2-adrenergic receptor antagonist. Advantages of mirtazapine include the following: it is less likely to cause sexual dysfunction, may reduce sleep latency and prolong sleep duration, may have anxiolytic effects, is relatively devoid of alpha-1 blocking effects such as orthostatic hypotension, and may cause less GI adverse effects than the SSRIs. Mirtazapine is a potent appetite stimulant, causing substantial weight gain in many patients. This can be advantageous in patients who are cachectic and nutritionally deficient, such as debilitated elderly, and cancer and HIV patients. However, in obese, diabetic, and patients with cardiovascular diseases, mirtazapine-induced weight gain is undesirable, and has been associated with deleterious effects on the lipoprotein profile (increased cholesterol and triglycerides). Patients should be warned about this side effect and often refuse to take it for this reason. Mirtazapine possesses significant antihistaminergic and minimal antimuscarinic effects, with the potential for considerable drowsiness, dry mouth, and constipation. Mirtazapine has a very low incidence of transient rise in liver enzymes (2%) and severe neutropenia or reversible agranulocytosis (0.1% incidence). It may be prudent for physicians to check white counts at initiation of the medication, semiannually, and with occurrence of infectious illnesses. Mirtazapine is unlikely to have clinically significant effects on the metabolism of other medications through liver enzyme systems.

Bupropion

The monocyclic antidepressant bupropion inhibits the reuptake of norepinephrine and dopamine into presynaptic neurons and is referred to as a norepinephrine dopamine reuptake inhibitor (NDRI). Bupropion does not inhibit serotonin reuptake and does not cause the serotonin syndrome. Bupropion is also indicated for smoking cessation under the trade name Zyban. Bupropion may enhance energy and motivation due to norepinephrine- and dopamine-induced activating effects. However, bupropion’s activating effect can also cause agitation, irritability, aggression, vivid dreams, nightmares, and insomnia. Despite this pronorepinephrine and prodopamine effect, only 2% of patients on this medication demonstrate elevations in blood pressure. Bupropion may be used for sexual dysfunction, it may promote weight loss, and it has a lower propensity to induce mania in patients with a bipolar diathesis.

Bupropion is contraindicated in patients with seizure disorders, bulimia, anorexia, and alcohol withdrawal. Bupropion-induced seizure disorders are dose related, and total daily doses above 450 mg, or single doses above 150 mg of immediate-release dosage or 200 mg of the sustained-release dosage forms increase the risk of seizures and should be avoided. Up to 450 mg of the XL form of the medication can be prescribed once a day, which is a major advantage for improving adherence. Bupropion has been shown to inhibit selective liver isoenzyme systems.

Black Box Warning on Antidepressant-Induced Suicide Ideation in Children & Adolescents

In 2004, the FDA issued a “black box” warning that the use of antidepressant medication in children and adolescents may be associated with an increase in suicide ideation and suicide attempts. In 2007, the “black box” warning was expanded to include young adults up to the age of 24 years. In brief, the warning results from a pooled analysis of 4400 subjects in 24 placebo-controlled drug trials with nine different antidepressant medications in children and adolescents for treating depression or OCD. The study found prevalence rates of treatment-emergent suicidal thoughts or suicide attempts in 4% of all patients in the active medication group, compared to 2% of patients taking placebo (a statistically significant difference). There were no completed suicides in any of the studies. Similar pooled analyses in adults found no difference between active drug and placebo in studies totaling more than 50,000 patients. The black box recommends that when prescribing antidepressant medication in children, adolescents, and young adults, physicians should warn patients and families to report immediately any increase in agitation or thoughts of suicide, and physicians should monitor patients through regular and frequent follow-up visits.

Physicians should also be aware of some important general epidemiologic findings. In the decade before this warning was issued, suicide attempts and completed suicides decreased significantly in children and adolescents, somewhat paralleling the increased use of antidepressant medications in the population. Since 2004, however, use of these medications in children and adolescents has decreased by 25%, and of interest and concern, rates of suicide attempts and completions in children and adolescents have increased for the first time in over a decade. Furthermore, large observational studies have documented that the risk of a suicide attempt actually decreases after patients begin taking medication and that communities with higher rates of antidepressant use have, on average, lower rates of suicide.

Drug Interactions

Whenever prescribing a new medication, practitioners must consider the possibility of drug–drug interactions. In addition, the use of the new drug in combination with another drug the patient is already taking may create a new condition of potential risk (e.g., additive sedative risks or potential serotonin syndrome). Because antidepressants are in such common use, questions of potential antidepressant drug interactions become issues of daily importance to most physicians, especially in the medically ill or the elderly.

Physicians should routinely use an electronic database to check potential drug interactions whenever they prescribe a new medication. They should understand the basic principles of enzyme substrates and enzyme inhibitors to be able to better interpret the results they receive from such an electronic database check on drug interactions.

All antidepressant medications are metabolized by cytochrome P450 liver enzyme subsystems. That is, they are substrates for one or more major P450 isoenzyme systems 1A2, 2C9, 2C19, 2D6, and 3A4. While all antidepressants are metabolized in the liver, only some of the SSRIs are moderate inhibitors of either 2C19 (fluoxetine and sertraline) or 2D6 (fluoxetine and paroxetine). Fluoxetine and paroxetine, for example, are moderate 2D6 inhibitors. As such, these drugs inhibit the metabolism of drugs metabolized by the 2D6 subsystem, for example, several beta-blockers, some antipsychotic medications, most TCAs, some benzodiazepines, and the over-the-counter cough suppressant dextromethorphan. When dextromethorphan is combined with fluoxetine or paroxetine, its plasma levels can increase significantly, which could theoretically increase the likelihood of a serotonin syndrome. Also, some SSRIs may inhibit the hepatic 2D6-mediated conversion of several prodrug-like opioids into their active analgesic metabolites, such as codeine to morphine, oxycodone to oxymorphone, and hydrocodone to hydromorphone. Clinically, this inhibition could theoretically decrease the analgesic effect of the opiates listed above. Among the SSRIs, sertraline, citalopram, and escitalopram present the least inhibition of 2D6, while mirtazapine and venlafaxine also have little effect on 2D6.

Clinicians should evaluate the probability, outcome, and risk of drug–drug interactions prior to prescribing antidepressants. In most settings, an alternative agent that does not interact can be prescribed. When patients are receiving drugs that can interact with each other, patients should be counseled on the signs and symptoms of toxicity, and clinicians should heighten monitoring techniques. Patients should be advised that using only one pharmacy allows the pharmacist to maintain a complete, comprehensive, and accurate medication profile and improves the likelihood that the pharmacist will also detect potential problematic drug–drug interactions.

Pharmacogenomics

Pharmacogenomics refers to the interaction of multiple genes in determining drug response. For example, genetic polymorphisms in the serotonin transporter gene have been associated with differential SSRI response rates. Some individuals have been shown to have differential activity levels of 2C19 and 2D6 isoenzymes—some patients are “extensive” metabolizers (high activity) and some are “poor” metabolizers (low enzyme activity). Of interest to researchers is that the prevalence rates of “poor metabolizers” vary by ethnicity: 30% of Asians, 10% of African Americans, and 5% of whites are poor metabolizers. From a clinical perspective, given the same dose of drug, poor metabolizers are likely to have much higher blood levels of the medication than extensive metabolizers. Why some individuals are more sensitive to lower doses of medication and/or respond better to lower doses, may be secondary to this genetically determined rate of metabolism, though how to apply this information clinically has not been clearly established.

The FDA has licensed and accredited several laboratories to provide DNA testing as well as blood tests for determination of CYP-2C9, -2C19, -2D6, and -1A2 genotypes. Although these tests are currently available, they have not yet gained widespread acceptance. The next decade will probably witness great progress in pharmacogenomics, such that testing and genomic-based medication algorithms will become part of the routine practice of medicine.

Antidepressant Withdrawal

Abrupt withdrawal of SSRIs, usually within 24–48 hours may cause a discontinuation syndrome consisting of a constellation of somatic, neurologic, and psychological symptoms. These symptoms include GI complaints, headache, fever, malaise, vivid dreams, myalgias, paresthesias, “electric shock-like” sensations, worsened mood, irritability, anxiety, confusion, and forgetfulness. These symptoms can be extremely uncomfortable, even leading to emergency room visits, but they are not life threatening. The incidence of the SSRI discontinuation syndrome may be as high as 40%, but the most common and most severe withdrawal syndrome is most commonly associated with paroxetine, probably due to its short half-life (21 hours) and the absence of any active metabolites. All of the SSRIs except for fluoxetine can cause this syndrome. Fluoxetine is unlikely to cause the discontinuation syndrome because of its long half-life (4–14 days, plus its active metabolite norfluoxetine), and hence has a self-tapering discontinuation mechanism. Venlafaxine, duloxetine, trazodone, and mirtazapine have also been associated with the discontinuation syndrome. Treatment of the discontinuation syndrome consists of reinstituting the SSRI and taper more slowly, or substitute the SSRI with one or two doses of fluoxetine (10–20 mg). In summary, all known antidepressants, with the possible exception of fluoxetine and bupropion should be withdrawn very gradually to avoid the discontinuation syndrome.

Antidepressant “Tolerance”

Despite the absence of clear evidence, clinicians commonly observe a tolerance phenomenon or poop-out syndrome that may occur in 10–20% of patients who have successfully responded to antidepressant therapy and remain compliant. Possible mechanisms for this observed tachyphylaxis include adaptation of central nervous system receptors via upregulation and decreased receptor density, disease severity or exacerbation, loss of a placebo effect, unrecognized rapid cycling, accumulation of a less potent or competitively antagonistic metabolite, or drug interactions. Enzyme-inducing medications, such as St. John wort, phenytoin, barbiturates, carbamazepine, oxcarbazepine, rifampin, and rifabutin, may increase the hepatic metabolism of non-CYP-2D6 substrate antidepressants, thus decreasing plasma levels. For example, phenobarbital has been associated with a 25% decrease in the area under the paroxetine serum concentration–time curve; St. John wort has been associated with a 22% decrease in the area under the serum concentration–time curve of amitriptyline and a 41% decrease in its metabolite nortriptyline; and phenobarbital can decrease the plasma levels of mirtazapine by 60%. Additionally, clinicians should be suspicious of a lack of compliance in all patients with the poop-out syndrome. Management of the poop-out syndrome includes increasing antidepressant doses, switching to an alternative agent, or adding augmentation therapy.

Augmentation Therapy

After an antidepressant at maximal doses has produced only a partial response, with persistent residual symptoms, experts usually recommend consideration of augmentation therapy, which consists of adding another agent to the antidepressant currently in use. Quantitative assessment with the PHQ-9 can help clinicians make decisions about when to initiate augmentation therapy.

Clinicians can use the following guideline.

When a depressed patient experiences a significant clinical improvement (as manifested by a drop in PHQ-9 score by 5 points or more), but fails to reach remission (PHQ <5) within a minimum of 1 month after the dose of the initial antidepressant is raised to its maximum, then consider augmentation therapy.

The only two augmentation agents with a substantial evidence base include lithium carbonate (300–1200 mg) and triiodothyronine (T3, 25–50 μg). More commonly used agents, however, include use of a second antidepressant from a different class (e.g., bupropion, venlafaxine, and mirtazapine), buspirone (a 5-HT1A receptor agonist, approved for GAD), pindolol (beta-blocker), atypical antipsychotics (e.g., aripiprazole and olanzapine are FDA approved for augmentation), and psychostimulants (e.g., methylphenidate, dextroamphetamine, and modafinil).

Are antidepressant medications effective in the medically ill and the elderly? There have been surprisingly few randomized-controlled trials supporting the safety and efficacy of antidepressant medication in the medically ill and elderly. However, a recent systematic review of studies in specific medical populations, such as post-MI, poststroke, cancer, diabetes, cancer, Alzheimer disease, as well as other recent studies in general medical populations and elderly medical populations, suggest the safety and efficacy of antidepressant medications in these populations. Physicians should be aware, however, that most studies suggest that concurrent general medical illness is associated with somewhat lower overall depression response and remission rates compared to patients without concurrent medical illness.

Of particular importance, the SADHART study (Sertraline Antidepressant Heart Attack Trial) documented the safety of SSRI use (sertraline) immediately after heart attack, even in patients with severe cardiac disease and in patients using multiple other medications. Other recent studies now also document the safety and effectiveness of citalopram and of mirtazepine for depression after heart attack.

For many of the reasons enumerated earlier, the SSRIs and other new agents have become the agents of choice in the elderly and in the medically ill. When tricyclic agents are used, however, the safer ones are nortriptyline and desipramine, which are preferred because of their relatively lower anticholinergic and antiadrenergic side effects. Because of their quinidine-like effects, however, all TCAs are relatively contraindicated in patients with unstable cardiac disease or with recent MI. In depressed patients with diabetes, use of TCAs, compared to SSRIs has been associated with impaired glycemic control. Because of this finding, as well as their other autonomic side effects (constipation, postural hypotension, and so on), use of tricyclic medications such as amitriptyline should be used only with caution in diabetics, including for the treatment of neuropathic pain. For treatment of painful diabetic neuropathy, physicians could consider the preferential use of nortriptyline (a metabolite of amitriptyline) or one of the SNRIs (duloxetine or venlafaxine).

Dosing strategies in the elderly and medically ill should adhere to the general well-known guideline “start low and go slow.” Pharmacokinetically, these agents are metabolized more slowly, resulting in accumulation and toxicity. Increased pharmacodynamic effects in the elderly may result from lower albumin levels, leading to higher levels of unbound drug.

Because depression is associated with increased morbidity, mortality, and cost in patients with concurrent general medical illness, it is reasonable to ask whether improvement in depression in medical patients leads to improvements in these other domains. There are limited and somewhat conflicting data on these questions, but some recent and notable studies point to potential decreases in morbidity, mortality, and overall health care costs with adequate treatment of depression. The SADHART study showed a trend (but not statistically significant) toward decreased rate of incident severe cardiac events and death in the sertraline group compared to placebo. A naturalistic outcome report of patients in the Enhancing Recovery in Coronary Heart Disease (ENRICHD) psychotherapy study noted that depressed patients treated with SSRIs (at physician/patient discretion, outside of protocol) had a statistically significant decrease in cardiac morbidity and mortality compared with the depressed patients who did not receive this treatment. A 2-year study of the collaborative care of patients with both depression and diabetes has shown an overall decrease in total health costs (general medical and psychiatric) in patients successfully treated for depression.

Electroconvulsive Therapy

Electroconvulsive therapy is still the most effective means available for the treatment of refractory depression. It is the treatment of choice for patients with psychotic depression, depression refractory to pharmacotherapy, and for some patients who are acutely suicidal. Despite prejudices and fears about ECT, new methods of administration have proven it to be a safe and effective treatment. In fact, ECT can be safer than antidepressant medication in the elderly. Some short-term memory loss is common, but research indicates that this reverts to normal in most patients. In some cases, ECT can be lifesaving and should not be denied to patients because of poor understanding or unrealistic fear. Electroconvulsive therapy does not lead to permanent remission of depression in patients susceptible to recurrence. Thus, patients with recurrent depression who receive ECT should receive either prophylactic medication after a course of therapy (as an outpatient) or maintenance ECT.

Light Therapy

Light therapy has been used successfully in the treatment of seasonal affective disorder (SAD), a condition in which recurrent depressive episodes occur during autumn and winter alternating with nondepressive episodes during spring and summer. Phototherapy is based on the principle that the presentation of artificial light at a similar strength to natural sunlight will prevent the biological changes that mediate SAD during the winter.

Other Brain Stimulation Interventions

With increasing awareness of alterations in brain networks, a number of direct brain stimulation interventions have been explored recently. These include approaches such as repetitive transcranial magnetic stimulation (rTMS), deep brain stimulation, and magnetic seizure therapy. Of these, only rTMS has an FDA-approved device. Even that approval has a narrow indication of only being used in individuals who have failed to respond to one antidepressant. Whether rTMS will be proved to have broader utility remains to be evaluated. Although its mechanism is unclear, high frequency stimulation is delivered over the left dorsolateral prefrontal cortex. Deep brain stimulation is available on an experimental basis for refractory depression. It focuses on decreasing activation of Brodmann Area 25, in the subgenual cingulate gyrus. This intervention can have immediate effects, but is a neurosurgical procedure with attendant risks, so that its application appears limited. Magnetic seizure therapy is being researched as an alternative to ECT and shows preliminary promise.

A generation of randomized, controlled health services trials indicates that systematic implementation of a “collaborative care model,” or “chronic illness care” (or “chronic disease management”) model of health care for depressed medical outpatients improves key processes and outcomes of care such as detection of depression; adequacy of depression treatment; patient adherence with treatment; improved clinical outcomes as assessed by measures such as PHQ-9 scores; and improved patient satisfaction. Key components of most of these new models generally include use of structured tools for assessment and management; utilization of care managers to help educate patients, support self-management, coordinate care and ensure follow-up; use of formal evidence-based guidelines and decision support tools; and finally, integration of behavioral health specialists into the medical team for consultation and ongoing support.

The criteria for referral to a mental health specialist depend a great deal on the experience and expertise of the primary treating physician. Patients should be informed that consultation with a mental health specialist may be necessary if the depression does not fully remit. This can make referral at a later stage much more acceptable. Because partial remission is a common occurrence, physicians should make every effort to establish clear indications of predepression functioning, and if full return to baseline functioning does not occur, referral to a specialist should be made. A baseline pretreatment PHQ-9 should be obtained and repeated at regular intervals to insure that the patient is responding appropriately to treatment. Patients with persistently elevated PHQ-9 scores (20 or above) or whose scores are not reduced to at least 50% of pretreatment levels, should be evaluated by a mental health specialist. Other indications for referral include patients with active suicidality, bipolar disorder, or psychosis; for evaluation of any patient in whom the diagnosis is in doubt; and for advice about pharmacological management of patients with treatment-resistant depression. The primary practitioner should communicate with the mental health specialist and provide the following information: the nature of the depressive symptoms; baseline premorbid functioning; baseline and follow-up PHQ-9 scores; other treatments, including medication, that have been tried previously or are concurrent; and patient understanding and expectations about psychotherapy. Communication should be maintained with the mental health therapist about whether therapy is completed, since failure to follow through on a referral and/or premature discontinuation of psychotherapy is even more common than failure to comply with pharmacotherapy.

Of all the psychiatric conditions relevant to the general physician, major depression is arguably the most common and most important. Major depression is common, serious, often disabling, and associated with increased morbidity and mortality in general medical conditions. It is often missed and often inadequately treated. Recognition and adequate treatment is associated with response and remission and can often be provided adequately in the general medical sector. Increased understanding of the biopsychosocial etiology of major depression, the biopsychosocial manifestations of major depression, and the biopsychosocial treatment of major depression holds great promise for improving outcomes in this debilitating condition.