CHAPTER 35—DEPRESSION AND OTHER MOOD DISORDERS
CLINICAL PRESENTATION AND DIAGNOSIS
MANAGEMENT OF BIPOLAR DISORDER
Prevalence studies of community residents demonstrate surprisingly low rates of depressive disorders among those aged 65 years and older. Only 1% to 2% of women and fewer than 1% of men interviewed with standardized instruments meet diagnostic criteria for major depression. Both current and lifetime prevalence rates for older persons are lower than those for middle-aged persons; furthermore, these relatively low rates persist after accounting for possible premature death and institutionalization, both of which may be associated with depression. Similarly, the incidence of first-episode major depression decreases after age 65. Data demonstrating that older persons are less likely to recognize depression and to endorse depressed mood offer one explanation for the lower prevalence and incidence of depressive syndromes among older community residents.
However, the prevalence of depressive symptoms that do not meet the threshold for a clinical diagnosis is substantial in seniors, with most studies reporting rates in the range of 15%. These subsyndromal states are not inconsequential. “Minor” or subsyndromal depression has been associated with increased use of health services, excess disability, and poor health outcomes, including higher mortality rates.
The prevalence rates of both major and subsyndromal depression are related to the setting in which older persons are seen and methods used to identify cases. Thus, major depression has been identified in 6% to 10% of older persons in primary care clinics and 12% to 20% of nursing-home residents. More varied rates of 11% up to 45% have been reported among elderly patients requiring inpatient medical care. The reported prevalence rates of minor depression in outpatient medical settings have been varied as well, with rates from 8% to over 40% reported. Studies that count symptoms due to physical illness toward a diagnosis of depression may inflate prevalence rates among medical patients because of symptom overlap. In the psychiatric clinic setting, major depression is the most common disorder seen among elderly patients and accounts for more than 40% of outpatient caseloads and inpatient psychiatry admissions.
Symptoms used to diagnose major depression are listed in Table 35.1, which summarizes the criteria for major depression from the fourth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-IV). Although aging does not markedly affect the overall phenomenology of major depression, some differences between younger and older depressed patients have been reported. Older patients are more preoccupied with somatic symptoms and less frequently report depressed mood and guilty preoccupations. Among patients who do not acknowledge sustained feelings of sadness, the demonstration of the second core symptom, a persistent loss of pleasure and interest in previously enjoyable activities (anhedonia), is necessary for a diagnosis of major depression. Thus, one need not exhibit depressed mood to meet criteria for major depressive disorder. The core symptoms are particularly important to detect in primary care settings and have been found to identify most medically ill patients who also meet full diagnostic criteria. Furthermore, these core symptoms are less likely to overlap with those of a medical illness.
The diagnosis of major depression in older persons is complicated by the overlap among symptoms of major depression with those of physical illness. Patients with serious medical illness may be preoccupied with thoughts about death or feel worthless because of concomitant disability. The DSM-IV criteria require that the depressive symptoms are not a direct effect of a general medical condition or medication used to treat it, but this distinction based on cause may be difficult to make reliably. Alternative diagnostic criteria have been suggested for medical patients, including an inclusive approach that counts all symptoms regardless of cause. Inclusive approaches result in the highest prevalence rates. Furthermore, a depressed person’s thoughts of death and worthlessness appear to differ from those of a patient with a serious medical illness. In depression, these thoughts are not based on a realistic assessment of prognosis or overall self-worth. The perspective of depressed patients is influenced by the feelings of sadness or guilt that accompany the disordered mood. The alternative diagnosis of mood disorder due to a general medical condition should be used for patients with depression that appears to result directly from a specific medical condition (eg, hypothyroidism).
The recognition of psychotic depression has particular relevance to primary care clinicians. Patients with psychotic depression have sustained irrational beliefs (delusions) in association with their depression, and these beliefs not infrequently focus on physical or medical preoccupations, such as the belief that one’s bowels are “blocked with cancer.” Psychotic depression may be suspected when the irrational belief focuses on somatic symptoms or around fears of a serious physical condition such as cancer when there is no medical source identified to support the belief.
Screening for mood symptoms is an important part of a comprehensive assessment of all older patients regardless of socioeconomic or ethnic status, although cultural variation may influence how emotional states are expressed. Since the diagnosis of depression hinges on a qualitative change from a patient’s normal mood state, it is essential to obtain information from involved family members or caregivers to determine whether a clinically meaningful syndrome has developed. Recent large-scale studies of mental health in the elderly age group (eg, the Primary Care Research in Substance Abuse and Mental Health for the Elderly, PRISM-E) have shown that among older persons with depression, there is a preponderance of women and ethnic minorities. This suggests that a higher degree of vigilance for mood disorders is warranted in these groups.
A diagnosis of mania requires a distinct period of persistently elevated mood lasting for 1 or more weeks and three additional symptoms that may include inflated self-esteem or grandiosity, hypersexuality, increased activity, decreased need for sleep, pressured speech, racing thoughts or flight of ideas, and distractibility. Grandiose or paranoid delusions may be present. Although the criteria for diagnosing bipolar disorder in younger and older patients are identical, some differences in phenomenology have been noted. Elderly patients with bipolar disorder are more likely to have a mixture of depression and marked irritability. Pressured speech that tends to go off on tangents is common, although the severity of thinking disturbance is less pronounced than in young adults and flight of ideas is less common. Hypersexuality and grandiosity may be present but are less prominent, as well. Manic-like syndromes in late life are also distinguished by a greater likelihood of confusion, often a reflection of an underlying cognitive disturbance, such as an incipient dementia.
Approximately 30% to 50% of patients with late-life depression are experiencing a recurrence of episodes that began in early adulthood. Despite the associations between late-onset depression and chronic medical illness, disability, and psychosocial stresses, most episodes develop without an identifiable precipitant. Because late-onset depression is less likely than typical recurrent depression to be familial, a positive family history is less useful for establishing a diagnosis.
Medical disorders that may imitate depression are particularly important to consider in elderly patients because of the increased vulnerability of this population to physical illnesses. Hyperthyroidism merits special consideration, because older hyperthyroid patients may present with apathy and diminished energy that mimics depression. Although apathy may be commonly associated with the core depressive symptom of anhedonia, apathy may also occur in persons who retain a reactive capacity for experiencing pleasure. The distinction of apathy from true depression becomes particularly important among patients with Parkinson’s disease, carcinoma of the pancreas, and dementia, because depressive syndromes that occur commonly in these disorders are responsive to antidepressant medications.
Major depression may be a prodrome of dementia such as Alzheimer’s disease or may develop after the onset of cognitive decline. Differentiation may be confounded because depression is commonly accompanied by symptoms of impaired concentration, indecisiveness, and lack of motivation that are also associated with dementia. An older person with depression may both report memory loss and poor concentration and be unable to perform simple cognitive tests without having dementia. Differentiation from dementia requires evidence that the cognitive impairment, developed concurrently with the onset of depression, and reverses with improvement in mood. Conversely, patients with true dementia may have symptoms that imitate depression. These include loss of interest, apathy, psychomotor retardation, and disrupted sleep. A diagnosis of major depression requires the presence of at least one core symptom and the persistence of symptoms for at least 2 weeks; approximately 20% of patients with early Alzheimer’s disease meet these criteria.
Studies have described a syndrome of “vascular” depression that is associated commonly with anhedonia, executive dysfunction, and the absence of guilty preoccupations. The syndrome is thought to result from prefrontal and subcortical lesions due to microinfarcts. These patients tend to have a late age of onset, risk factors for vascular disease, prefrontal or subcortical white matter hyperintensities on T2 weighted magnetic resonance imaging, and evidence of non-amnestic neuropsychologic deficits in tasks requiring initiation, persistence, and self-monitoring. Execution of tasks requiring planning and shifting between cognitive sets is impaired. Depressed patients with such defects in executive function may experience a remission of disturbed mood in response to standard pharmacotherapies but remain substantially disabled because of the cognitive dysfunction.
Bereavement is common among older persons. Feelings of sadness, disturbed sleep, and diminished appetite are common among bereaved persons but are generally time limited and resolve within a few months. Up to 14% of bereaved adults develop a major depression within 2 years of the loss, which occurs with a lower frequency among bereaved elderly persons. Bereavement that has evolved into a major depression is characterized by morbid preoccupations with guilt or suicidal ideas beyond transient thoughts of joining the deceased that would be expected in association with the loss. Also, bereavement is not associated with sustained functional impairment, so a marked decline in self-care may be a sign of a depressive syndrome.
Major depression is often a recurrent disorder in both younger and older adults. Evidence from observational studies in which treatment was not provided systematically indicate that up to one third of patients run a chronic course and another third have partial recovery with residual disability. Patients who suffer their first episode in later life take longer to recover. Depressed elderly persons also use health services, including outpatient visits, specialist consultations, and laboratory assessments, more than nondepressed elderly persons do.
Comorbid medical illness is associated with more refractory mood symptoms, and medical morbidity and mortality are themselves increased in patients who have suffered a depressive episode. The relationship between depression and increased mortality rates from cardiovascular causes is particularly striking. Depressed patients have an increased risk of cardiovascular mortality generally, and the development of depression in patients following a myocardial infarction, heart failure, or cardiac bypass surgery increases mortality from cardiovascular events.
Old age is associated with an increased risk for suicide. Persons aged 65 and over represent less than 13% of the population but account for 25% of suicides. The likelihood that an attempt will be lethal approximates 25%, a rate far greater than in young adults. “Psychologic autopsy” studies that use interviews with relatives and physicians to reconstruct a suicide victim’s premorbid mental state have demonstrated that as many as 75% of older adults who commit suicide were suffering from depression. The vast majority had seen a primary care physician within the previous month. In addition to depression, risk factors for late-life suicide include a comorbid physical illness, living alone, male gender, and alcoholism. Older men and women commit suicide violently; firearms and hanging are the methods most commonly employed.
The course of treatment for major depression can be conceptualized as follows: acute treatment to reverse the current episode, continuation treatment to prevent relapse, and maintenance treatment to prevent recurrence. Continuation treatment to stabilize the recovery involves ongoing antidepressant therapy for an additional 6 months. Maintenance treatment (3 years or longer) is provided to patients with a history of recurrent depression. The duration of maintenance therapy should be based on the frequency and severity of previous episodes. Recurrent episodes complicated by suicidal ideas or attempts warrant lifelong treatment. Psychotherapy, antidepressants, and electroconvulsive therapy (ECT) are empirically proven treatments for depression in older persons.
Initiatives have been developed to increase the likelihood that patients with major depression treated in primary care settings will receive antidepressant treatment that is adequate in dose and duration. The increased attention to this problem results from findings that most elderly patients prefer being treated for depression by their primary care physicians. Studies have shown that the majority of patients who are prescribed antidepressants by their primary care physicians do not obtain refills of their initial prescription. Educational programs and the use of depression nurse specialists have been developed to increase recognition and treatment intensity.
Table 35.2 lists the variety and characteristics of psychotherapies advocated for older adults. Research demonstrating the efficacy of psychotherapy for major depression in older adults has included cognitive-behavioral therapy, interpersonal psychotherapy, and problem-solving therapy. Problem-solving therapy involves working with the patient to identify practical life difficulties that are causing distress and providing guidance to help the patient identify solutions. The treatment is delivered generally in six to eight meetings spaced 1 to 2 weeks apart. Cognitive and interpersonal psychotherapy are also time-limited but less highly structured. Psychotherapy for minor depression has been promising, with efficacy demonstrated particularly in persons who have suffered a loss. Also, caregivers of older persons may develop minor or major depressive syndromes that benefit from psychotherapy. Psychotherapy may be combined with an antidepressant, and the combination has been associated with a longer period of remission following recovery from the acute episode. Psychotherapy combined with antidepressant medication is recommended for all patients with severe or suicidal depression.
The choice of agent depends on the patient’s comorbid medical conditions, the side-effect profile of the antidepressant, and the individual patient’s sensitivity to these effects. Potential interactions with other medications also should be considered. Prominent complaints of sleep disturbance, anxiety, or psychomotor retardation will direct the practitioner toward agents that are, respectively, more sedative or more activating. And because the SSRIs are now the mainstay of antidepressant prescribing, a hierarchy of choices is available in published protocols designed for either primary care or older patient practices. Sertraline and citalopram are often preferred because adverse effects and drug interactions may be less likely than with some other SSRIs, such as fluoxetine and paroxetine.
A number of researchers have explored various strategies to optimally manage depressive syndromes in older persons, recognizing that a systematic approach is often helpful. One such treatment strategy, developed by Steffans and colleagues at Duke University, is called the Duke Somatic Treatment Algorithm for Geriatric Depression. The graded approach in this algorithm begins with a 6- to 12-week trial of the SSRI sertraline for the older depressed patient who has not previously been treated. The only stipulation at this point is that the patient’s illness not be severe enough to warrant an intervention such as ECT (eg, an imminent risk of suicide, life-threatening food refusal, psychosis). For an initial SSRI trial, it is recommended that the dose be gradually titrated upward at weekly intervals for the first 3 weeks, as tolerated. If the SSRI is not well tolerated, the clinician may proceed to the venlafaxine trial described below. For patients who have had a previous successful response to an antidepressant, that particular agent would be the medication of choice.
If, after 6 to 12 weeks, a satisfactory response is not evident, then a trial of sustained-release venlafaxine is recommended at a dosage of 150 mg daily for 6 to 12 weeks. If the symptoms do not remit with venlafaxine, other options may be considered at this point, such as augmentation with another medication. For example, if there is a partial response to the SSRI but with residual low energy and low interest, then the SSRI may be augmented with sustained-release bupropion for an additional 6 weeks. Other options in the context of incomplete response may include augmenting with lithium carbonate or switching to a tricyclic antidepressant or monoamine oxidase inhibitor (MAOI). However, at this juncture it might be helpful to obtain consultation from a psychiatrist or other mental health professional.
SSRIs and tricyclic antidepressants are comparably effective for treating mild to moderate major depression, but SSRIs are often better tolerated. Table 35.3 provides a detailed summary of dosing, formulations, precautions, and advantages of the individual SSRIs. Controversy continues about whether the most severe melancholic form of depression that generally requires treatment in a psychiatric hospital responds better to a tricyclic antidepressant. Melancholic patients typically suffer from marked appetite and weight loss, diurnal mood variation that is worse in the morning, early morning awakening, and either retardation of motor movement or agitation.
Although the SSRIs are generally free of severe side effects, a small proportion of elderly patients develop hyponatremia that is due to the syndrome of inappropriate antidiuretic hormone secretion during SSRI treatment, particularly at higher doses. Rare patients are unable to tolerate SSRIs because of induced anxiety, sleep disturbance, or agitation. This reaction appears to be more common in older persons and may in some cases be related to the inhibition of the metabolism of other medications, discussed later in this section. Beginning at lower doses may reduce the incidence of anxiety with activating SSRIs such as fluoxetine and may decrease the gastrointestinal adverse effects of nausea or diarrhea that occur with SSRIs. Sexual side effects occur commonly with all SSRIs, and many patients may not wish to continue SSRI treatment because of them, so inquiring about sexual function is an important part of follow-up care. Although SSRIs have been associated with mild weight loss initially, long-term use with many of these medications has been associated with weight gain. Pseudoparkinsonism and other movement disorders may occur, particularly if these medications are used in combination with other drugs that block dopamine (eg, metoclopramide).
SSRIs inhibit various P-450 hepatic cytochrome isoenzymes that metabolize most medications. Fluoxetine and paroxetine are potent inhibitors of the P-450 2D6 isoenzyme responsible for the metabolism of most psychiatric medications, as well as dextromethorphan, codeine, tricyclic antidepressants, and metoprolol. Sertraline is a weak inhibitor of both the 2D6 and 3A4 isoenzymes. Inhibition of 2D6 can decrease the analgesic effects of codeine (and related compounds) and tramadol by preventing their conversion to their active metabolites, morphine and O-desmethyltramadol. Fluoxetine’s active metabolite, norfluoxetine, also inhibits the 3A4 isozymes. The SSRI fluvoxamine inhibits the 3A4, 2D6, and 1A2 systems. Because P-450 1A2 is required to metabolize theophylline, olanzapine, and phenacetin, concurrent use of fluvoxamine may result in markedly increased concentrations of these medications. Given its potential for multiple interactions, the use of fluvoxamine in older patients is best limited unless monitored very closely.
SSRIs may increase the anticoagulant effects of medications such as warfarin, potentially through cytochrome isoenzyme inhibition or the inhibition of platelet activity. Careful monitoring of prothrombin times or another suitable index of blood clotting is indicated following the introduction of an SSRI to patients being treated with warfarin.
SSRIs have been considered particularly safe for older persons because these antidepressants do not cause orthostatic hypotension, arrhythmias, or marked sedation. However, evidence indicates that nursing-home residents treated with SSRIs are at increased risk for falling and suffering hip fractures. It is not clear whether these findings are due to patient selection in that patients considered at greatest risk for falling may be treated with SSRIs preferentially. Alternatively, SSRIs may contribute to falls by causing a subtle balance disturbance or pseudoparkinsonism in vulnerable patients. Or depression itself may be a risk factor.
With the introduction of serotonergic reuptake inhibitors, reports emerged of a serotonergic withdrawal syndrome following discontinuation lasting 2 to 3 weeks and characterized by lightheadedness, insomnia, agitation, nausea, headache, and sensory disturbances. Mood disturbance may also occur. The shorter-acting SSRIs (sertraline, paroxetine) appear to induce this syndrome, but venlafaxine and other SSRIs have also been implicated. Therefore, these agents should be tapered rather than stopped abruptly.
The tricyclic antidepressants nortriptyline and desipramine are the most appropriate for use in older persons. Table 35.4 provides a detailed summary of dosing, formulations, precautions, and advantages of these agents. They are effective in the most severe forms of depression but are associated with anticholinergic side effects that may be problems in older patients. For nortriptyline, therapeutic response is associated with blood levels between 50 and 150 ng/mL and for desipramine, levels above 120 ng/mL. Over 60% of patients with nonpsychotic major depression or with depression that is not associated with dementia respond within 6 weeks to levels in these ranges. Although 5% of the population requires lower dosing because of the absence of the enzyme required to metabolize secondary amine tricyclics, most patients achieve target concentrations at dosages of 50 to 75 mg per day of nortriptyline and 100 to 150 mg per day of desipramine.
The potential for side effects that are due to the anticholinergic and sedative properties of tricyclics limits their use in older patients. These medications are particularly inappropriate for patients who are sensitive to constipation or vulnerable to orthostatic hypotension and for men with benign prostatic hyperplasia. Also, tricyclic antidepressants have a quinidine-like effect that delays ventricular conduction. Among patients with a pretreatment first-degree heart block, 10% may develop a second-degree block during treatment. Patients with bundle branch block or prolonged QTc interval are at risk for developing a ventricular arrhythmia.
Table 35.5 provides a detailed summary of dosing, formulations, precautions, and advantages of other antidepressants, including a stimulant. Bupropion is generally safe, free of sexual side effects, and well tolerated when used at recommended doses. Bupropion has been associated with a 0.4% risk of seizures, which is much higher when recommended doses are exceeded, and is contraindicated in persons with a seizure disorder. Bupropion appears to act by increasing the activity of dopamine and norepinephrine and therefore has stimulant-like qualities. Because of the activating property of bupropion, dosing of the short-acting form in the morning and mid-afternoon may help to avoid insomnia. Long-acting bupropion has a smaller seizure risk and is less likely to cause insomnia. Doses are increased gradually by the use of a twice-a-day regimen to achieve target doses of 150 to 300 mg per day of the short-acting form. The dosage range of sustained-release bupropion is 150 to 300 mg per day given as a single daily dose.
Venlafaxine acts as an SSRI at lower doses while also inhibiting the reuptake of norepinephrine at the high end of the therapeutic range of 75 to 225 mg per day. Venlafaxine is effective for both generalized anxiety and major depression and may treat both effectively at dosages of 75 to 150 mg per day. A dose-response relationship for severe depression has been demonstrated, and patients with melancholia may require dosages of 225 mg per day or greater. Initial dosing should use gradual increases to minimize the early side effect of nausea. The noradrenergic properties of higher doses may explain the association between doses of 225 mg and higher with hypertension. Patients requiring doses at the high end of the therapeutic range should have blood-pressure monitoring. This side effect does not occur in patients with hypertension controlled by a β-blocker. Venlafaxine should also be discontinued several days prior to ECT to avoid hypertension during ECT. Venlafaxine, like other SSRIs, has been associated with decreased sexual functioning. Venlafaxine should be discontinued by gradual tapering to avoid the risk of flu-like discontinuation symptoms.
Duloxetine is a new antidepressant medication that has been approved for the treatment of both depression and neuropathic pain secondary to diabetes mellitus. Duloxetine is a serotonin- and norepinephrine-reuptake inhibitor, and its pharmacodynamic characteristics are generally like those of venlafaxine, although it is structurally unique. Because of its effects of increasing neural sphincter activity and bladder capacity, it has also been approved for the treatment of urinary incontinence and has been shown to reduce stress incontinence in women. These features may make duloxetine advantageous for the older patient.
Mirtazapine is a norepinephrine, 5-HT2, and 5-HT3 antagonist. The total daily dosage is 15 to 45 mg. Mirtazapine is given as a single bedtime dose to capitalize on its sedative properties and is available in rapidly dissolving form for patients who have difficulty swallowing pills. Mirtazapine is also associated with increased appetite and weight gain. This side-effect profile has contributed to the use of mirtazapine before bed in nursing-home residents with depression and dementia who demonstrate nighttime agitation and weight loss. Mirtazapine is also free of sexual side effects.
The MAOIs are effective antidepressants that have been in use for nearly half a century. Orthostatic hypotension is a common adverse effect of MAOIs and in older patients may increase the risk of falling. Concomitant ingestion of a food product rich in tyramine (eg, cheese) or of pseudoephedrine or pressor amines can cause a life-threatening hypertensive crisis. The use of MAOIs with an SSRI or meperidine can cause a fatal serotonin syndrome associated with delirium and hyperthermia.
Although methylphenidate and other stimulants have been used for decades to treat major depression, controlled data demonstrating their efficacy are limited. However, these agents may have a role in reversing the apathy and lack of energy seen in some patients with dementia or disabling medical conditions. Benzodiazepines may be useful temporary adjuncts to treat anxiety and sleep disturbance associated with a major depression, but older persons are particularly sensitive to the adverse effects of these medications on cognition and gait stability.
Approximately 50% of patients with major depression demonstrate a robust response to treatment within 6 weeks. An additional 15% to 25% who have begun to improve achieve remission if treatment is continued for an additional 4 to 6 weeks. Some published studies using SSRIs have reported 6-week response rates of 40% that increases to more than 60% when treatment extends to 12 weeks. However, noticeable improvement within the first 2 weeks of treatment predicts the likelihood of successful recovery with the initial medication. The admonition to “start low, go slow” in treating the older patient is too often interpreted as “start low, stay low,” which can lead to mistaken labeling of the patient as a nonresponder. A less than 30% reduction in symptom severity by the sixth week of treatment at the indicated dose defines nonresponse. A 50% reduction in symptom severity indicates partial response. In either case the dose should be “optimized” by increasing to the upper limit of the final recommended dose range, depending on patient tolerance. Patients with a genuine nonresponse should be switched from an SSRI to a non-SSRI. Patients experiencing a partial response to an SSRI may receive augmentation (combination) therapy with a non-SSRI.
The options of switching to an antidepressant from another class or consulting a psychiatrist are more appropriate than initiating unfamiliar combination treatments.
Psychotic depression is often resistant to standard antidepressant regimens. Aggressive pharmacotherapy is required, with best results obtained in young adults when high doses of antidepressants and antipsychotic medications are combined. However, the effectiveness of combination treatment for older patients with psychotic depression remains uncertain. Available evidence suggests that most elderly patients who have depression with pronounced psychotic features either cannot tolerate adequate doses of conventional medications or do not respond to them. Therefore, ECT has become the standard treatment for late-life psychotic depression.
Most persons with bipolar disorder have a history of episodes in early adulthood and often receive chronic treatment with an anti-manic medication. Occasionally this regimen may need adjustment if a new episode of mania emerges. Evidence to support an array of medications for mania has increased substantially in the past decade (see Table 35.6). In addition to lithium and established mood stabilizers for mania, newer antipsychotic and anticonvulsant medications are used in managing manic symptoms, although controlled trials in elderly patients are lacking.
Lithium carbonate is highly effective for the acute and maintenance treatment of classical manic episodes. It is important to recognize that much lower plasma levels are recommended for older patients than for younger patients. For example, plasma ranges of 0.8 to 1.2 mEq/L are used to treat acute mania in younger adults, and levels of 0.6 to 0.8 mEq/L are used for mania prophylaxis. The use of lithium for treating elderly patients is complicated by the 30% decrease in renal functioning that accompanies normal aging and the increased sensitivity of older persons to neurologic side effects from this medication, particularly if structural brain disease is present.
The decreased renal clearance of lithium in the older patient is addressed simply by reducing the total daily dose. Furthermore, lower concentrations may be adequate in older adults; some patients respond acutely to levels in the range of 0.6 to 0.8 mEq/L or less. Levels of 0.4 to 0.6 mEq/L may provide effective prophylaxis. Also, older persons develop neurologic side effects at concentrations that are well tolerated in young adults. A fine resting tremor, diarrhea, myoclonus, an intention tremor, or pseudoparkinsonism may develop with levels in the standard therapeutic range. Sensitivity to these effects and to increased confusion is increased among patients with dementia and when lithium is coadministered with an antipsychotic medication. Clinicians are advised to balance the therapeutic effects of a concentration at the low end of the usual target range against side effects that develop in a particular patient. Consideration should be given to a time lag of 1 to 2 weeks that may occur between achieving a steady-state dose and either therapeutic efficacy or the development of neurologic side effects.
Drugs that inhibit lithium excretion may raise concentrations by up to 50% and result in toxicity if the lithium dose is not decreased. Many of the nonsteroidal anti-inflammatory agents, including ibuprofen, have this effect. Thiazide and potassium-sparing diuretics, angiotensin-converting enzyme inhibitors, and angiotensin-receptor blockers raise lithium levels as well. The effect of furosemide is less clear, but lithium levels should be checked a few days after this diuretic is added in an older patient receiving lithium. Both dehydration and salt depletion are known to raise lithium levels. With adequate monitoring, bipolar patients who have benefited from lithium need not be deprived of this medication because of advanced age.
Valproic acid (valproate, divalproex) has been approved for the treatment of bipolar disorder and is comparable in efficacy to lithium. Target concentrations of 50 to 100 μg/mL used to treat seizure disorders appear to be appropriate. Divalproex has been used effectively to treat agitation in patients with dementia and may be particularly useful for the treatment of behavioral disturbance associated with mood lability. However, upward titration in the frail elderly person must be carried out carefully because dementia patients may be particularly vulnerable to developing excessive sedation during valproic acid treatment. As with other anticonvulsants used to treat mania, valproic acid is associated with allergic skin rashes in 5% to 10% of patients. Decreased platelet counts occur at higher doses, and platelet functioning may be diminished. Patients with hepatic disease may develop liver toxicity during valproic acid treatment. Patients with low albumin concentrations will also have a higher proportion of free drug in their blood and are at greater risk for toxicity.
CarbamazepineOL is another antiepileptic medication that has been used to treat mania. Leukopenia and thrombocytopenia occurs in 5% to 10% of patients within the first 2 weeks of treatment. However, life-threatening agranulocytosis and aplastic anemia occur rarely.
Lamotrigine is a newer anticonvulsant that has been approved for the treatment of depression in the context of bipolar affective disorder, although trials specifically in elderly patients are not available. Titration should be conducted very carefully in patients who may also be taking hepatic cytochrome isoenzyme-inducing drugs or valproic acid. In the absence of these, titration begins at 25 mg per day for 2 weeks, then 50 mg per day for 2 weeks, then 100 mg per day for 1 week, then 200 mg per day for usual maintenance. As with most of the anticonvulsants, for patients with moderate liver dysfunction the doses should be reduced by approximately 50%, and with more severe dysfunction the doses should be reduced by approximately 75%.
For patients in whom ECT is anticipated, anticonvulsants agents should be tapered and discontinued if at all possible because of the likelihood of interfering with an adequate seizure duration during the course of treatments.
Most antipsychotic medications are observed clinically to exert some anti-manic effects; however, olanzapine has been approved by the U.S. Food and Drug Administration for the acute treatment of mania. Concerns about olanzapine-associated metabolic syndrome need to be considered. Mania in late life may be associated with confusion and cognitive dysfunction; the anticholinergic effects of olanzapine may exacerbate these symptoms. Risperidone and quetiapine are also FDA approved for the acute treatment of mania.
ECT is highly effective for the treatment of major depression and mania, with response rates exceeding 70% in older adults. ECT is the first-line treatment for patients at serious risk for suicide or life-threatening poor intake due to a major depression. Patients with delusional depression may demonstrate paranoia about their food or caregivers, precluding pharmacologic treatment because of the unreliable oral intake. Also, delusional depression is less responsive to standard medication regimens. Therefore, ECT is generally the first-line treatment for these patients and is associated with response rates that approximate 80%.
The cognitive side effects of ECT are the principal factor limiting its acceptance. Anterograde amnesia or the inability to learn new information may be pronounced initially, particularly during bilateral ECT, but it improves rapidly following completion of treatment. Retrograde amnesia is more persistent, and the recall of events that immediately preceded ECT may be lost permanently. Although patients may complain that ECT has had a long-term effect on their memory, longitudinal studies have failed to demonstrate lasting cognitive effects; furthermore, improved memory, perhaps owing to recovery from depression, has been reported. There are few absolute medical contraindications other than the presence of increased intracranial pressure or unstable angina. Patients with coronary artery disease or cerebral vascular disease may be administered ECT safely by the use of appropriate pharmacologic management of the autonomic responses that may occur during treatment. Nevertheless, a recent myocardial infarction or cerebral vascular event and unstable coronary artery disease increase the risk of complications. Right unilateral treatment produces fewer cognitive side effects than bilateral treatment but is less effective unless doses markedly exceeding a patient’s seizure threshold are used.
The selection of ECT over aggressive pharmacotherapy is generally made by weighing the risk of waiting for a medication to work against the burden of hospital treatment, any medical conditions that may complicate general anesthesia, and patient biases against this treatment. After a course of ECT, patients should be treated with continuation pharmacotherapy following recovery. Patients who have failed intensive antidepressant treatment before receiving ECT have lower acute response rates and are more likely to relapse subsequently, even when antidepressant continuation treatment with a new medication is provided. Although maintenance ECT is sometimes used to prevent relapse, the burden that maintenance ECT places on patients and their families may limit its usefulness for the long-term management of late-life major depression.
■ Alexopoulos GS, Meyers BS, Young RC, et al. Executive dysfunction and long-term outcomes of geriatric depression. Arch Gen Psychiatry. 2000;57(3):285–290.
This study assessed relapse and recurrence rates among elderly patients with major depression during 16 weeks of continuation treatment with therapeutic concentrations of nortriptyline and during up to 2 years of maintenance treatment with nortriptyline or placebo. All patients had achieved remission with nortriptyline. Patients who demonstrated abnormal initiation and perseveration scores indicative of executive dysfunction had significantly higher relapse rates during continuation treatment and higher recurrence rates during either nortriptyline or placebo maintenance. Memory impairment did not affect the course of remission. The data are consistent with the hypothesis that disruption of prefrontal pathways worsens the treatment response and course of geriatric depression.
■ Denihan A, Kirby M, Bruce I, et al. Three-year prognosis of depression in the community-dwelling elderly. Br J Psychiatry. 2000;176(5):453–457.
This study builds on prognostic research of clinical samples of elderly persons with major depression by investigating community residents at baseline and 3 years later to determine their course of illness. Late-life depression was found to be associated with higher-than-expected mortality rates and persistent impairment. Poor physical health was found to be associated with a poor outcome, and having been treated with antidepressants was found to significantly improve the prognosis.
■ Fava M, Rush AJ, Trivedi MH, et al. Background and rationale for the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) study. Psychiatr Clin N Am. 2003;26(2):457–494.
Results from the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) study, once complete, will provide evidence to guide both switch and augmentation strategies. The study is being conducted in both primary care and specialty practices in public and private settings. Citalopram is the first choice, followed by switching to either sertraline, bupropion, or venlafaxine. Cognitive-behavioral therapy is also an option with or without medications. The first choice for augmentation includes bupropion sustained release, cognitive therapy, or addition of the antianxiety agent buspirone. Additional steps in treatment include mirtazapine or nortriptyline as well as augmentation strategies. STAR*D will provide data on the “switch or augment” strategy by anticipating partial responses and genuine treatment resistance.
■ Kennedy GJ, Marcus P. Use of antidepressants in older patients with co-morbid medical conditions: guidance from studies of depression in somatic illness. Drugs Aging. 2005;22(4):273–287.
In this review the authors survey the antidepressant literature for the eight most common disabling somatic illnesses in late life. They observe that the increasing number of antidepressants and adjunctive medications have increased the practitioner’s perplexity when confronted with a very old, very depressed patient. However, a growing body of evidence from antidepressant studies in the context of age-related somatic illnesses allows reasonable inferences to guide the diagnosis and treatment of depression in the oldest-old patient. They advise practitioners to assess the benefits of prescribed medication within the first days rather than first weeks of treatment and to escalate the antidepressant to the established therapeutic range rather than seek the lowest effective dose. Patients who experience no benefit whatsoever within the first weeks of treatment despite being within the therapeutic range should be prescribed an alternative.
■ Mulsant BH, Alexopoulos GS, Reyolds CF 3rd, et al. Pharmacologic treatment of depression in older primary care patients: the PROSPECT algorithm. Int J Geriatr Psychiatry. 2001;16(6):585–592.
The Prevention of Suicide in Primary Care Elderly: Collaborative Trial (PROSPECT) algorithm adapted existing treatment guidelines for depression in older adults to primary care settings. The aim of the study was to prevent suicide through aggressive treatment of late-life depression in primary care patients. Strict definitions for nonresponse and partial response were linked to unambiguous options. Augmentation is preferred for partial responders, switch for nonresponders. Citalopram is the first choice, but if the patient is taking another antidepressant at substandard dose or duration, that medication should be “optimized” to the recommended dose for 12 weeks before choosing switch or augmentation. Psychotherapy is included as augmentation for partial responders or as sole therapy for those who prefer drug-free treatment. Psychotherapy as augmentation is also recommended at the outset for severe depression. Nonresponders may be switched to nortriptyline. Lithium and nortriptyline as well as the newer antidepressants may be used to augment the treatment of partial responders.
■ Steffens DC, McQuoid DR, Krishnan KR. The Duke Somatic Treatment Algorithm for Geriatric Depression (STAGED) approach. Psychopharmacol Bull. 2002;36(2):58–68.
The Duke Somatic Treatment Algorithm for Geriatric Depression (STAGED) offers five stages of treatment that are based on the patient’s past experience with antidepressants. Nearly 90% of patients achieved partial remission and nearly two thirds full recovery, as defined by predetermined symptom assessment measures. For persons with either a partial response or no response, the choices included both switch and augmentation. Through stage 4 the newer antidepressants were employed before nortriptyline singly or combined with either lithium or a selective serotonin-reuptake inhibitor was prescribed. Cumulative response rates took 9 months to plateau, but ranged from 95% (stage 1) to 81% (stage 5). The authors acknowledge that their practice site is a university-based psychiatric research setting with access to laboratory and diagnostic procedures as well as an inpatient unit, resources that are not available to every primary care site. Indeed, nearly one quarter of patients studied received electroconvulsive therapy. In summary, as applied by expert practitioners in an ideal setting with aggressive follow-through, STAGED gets nearly every patient better and almost two thirds well.
Gary J. Kennedy, MD