Treatment Considerations for in Patients with Significant Medical Comorbidity

David M. McCoy, MD Nashville, Tennessee

In addition to being a strong psychological stressor in favors the use of selective reuptake inhibitors itself, medical illness is associated with risk factors that and other new because they have fewer predispose patients to develop coexisting depression. anticholinergic, cardiac, or cognitive adverse effects. Patients with conditions such as cancer, cardiovascular Depressed medically ill patients clearly benefit from anti­ disease, and neurologic disorders are particularly prone depressant therapy. Because mental health influences to depression because these illnesses are severe, prognosis and treatment outcome, primary care physi­ chronic, and often fatal. Because an may cians should maintain a high index of suspicion for exacerbate an underlying illness, leading to more seri­ depression in patients with significant medical illness ous side effects, agents with a poor tolerability profile or and aggressively treat the condition when indicated. that act at multiple receptor sites should be avoided. In many cases, this precludes the use of tricyclic antide­ KEY WORDS. Depression; comorbidity; drug therapy. J pressants and monoamine oxidase inhibitors, and Fam Pract 1996; 43(suppl):S35-S44

ajor depression is responsible for and treatment of depression in the family medicine more days in nonpsychiatric hospitals setting. and more days home from work than any other chronic illness except for RISK FACTORS severe, unstable coronary artery dis­ ease.1M Even though at least 8 million people in the Some medical disorders have clearly been associat­ United States become depressed in any given year,2 ed with higher incidences of depression. These only one third to one half of those with mtyor depres­ include cancer (1% to 50%),7" stroke (25% to 30%),”" sive disorder are properly diagnosed:' Family physi­ coronary heart disease (18% to 50%),1214 and cians have an increased burden to care for these Parkinson’s disease (4% to 70%).1WW7 These epi­ patients because 3 of every 4 mental health visits demiologic data concur with the results from other present initially to the primary care setting.1 studies, demonstrating that severe or chronic dis­ There is a trend toward increased prevalence of eases are associated with a higher prevalence of major depression in primary care clinics (5% to 10%) depression.618 Medically ill patients who are over 70 and inpatient medical wards (10% to 14%) compared years of age,19 hospitalized,20'21 or institutionalized22 with community samples (2% to 4%) (Figure l).6 In are also more likely to develop depression. view of these statistics, it is not surprising that patients with chronic medical conditions, such as DIAGNOSIS cancer, chronic lung disease, and arthritis, have a sig­ nificantly higher' lifetime prevalence of psychiatric It is expected that patients will react to a serious medical illness with some degree of demoralization disorders (46% to 60%) compared with patients with or dysphoria. It is important to note, however, that none of these conditions (33%; Pc.001).6 Thus, the noncompliance with or disregard for medical treat­ prevalence of depression in medically ill primary ment by patients is a significant indicator of depres­ care patients underscores the need for recognition sion.18 The diagnosis of depression should also be Revised, submitted September 30, 1996. considered when somatic symptoms are out of pro­ Address correspondence to David M. McCoy, MD, 394 portion to the illness or when the symptoms have a Harding PI, Suite 201, Nashville, TN 37211.

The Journal of Family Practice, Vol. 43, No. 6 (Suppl), 199(1 S 35 © 1996 Appleton & Lange/ISSN 0094-3509 DEPRESSION IN PATIENTS WITH MEDICAL COMORBIDITY

FIGURE 1 pie, , anorexia, and fatigue Relationship between the prevalence of depression and medical illness. Based are features that can be attributed to on data from Katon and Schulberg,5 either a medical illness or depres­ sion.24 Due to a 15 10%-14% General Medical Condition is a disor­ der described in the DSM-IV as a per­ o sistent mood disturbance caused by M<7> the direct physiological effects of a 5%-10% Q.S 10 I general medical condition. Symptoms O Q associated with the mood disturbance o 'ro* may include a depressed mood, 2 diminished interest in activities, or I O<1) 2%-4% elevated or irritable mood.23 Specific _0) symptoms that differentiate depres­ > sion from manifestations of medical k.o CL illnesses include suicidal feelings, a sense of failure or hopelessness, the Community Primary Care Clinic Inpatient Medical Ward feeling of being punished, diminished social interest, crying, indecision, and dissatisfaction (Table 1).2M7 Psycho­ significant impact on functional status.18 Somatic motor retardation (with the exception of hypothy­ complaints that are exaggerated relative to the roidism and Parkinson’s disease) and feelings of guilt patient’s medical illness or that are persistent or and self-reproach are other features of depression unresponsive to treatment are an indication that that are not usual consequences of medical illness­ depression may underlie the medical illness.20 The es.4' presence of a prior depressive episode or a positive According to DSM-IV, a Substance-Induced Mood family history for affective disorder should also raise Disorder is a persistent mood disturbance that is the level of suspicion for depression. According to associated with the direct physiological effects of a the Diagnostic and Statistical Manual of Mental substance (ie, a drug of abuse, a , other Disorders, 4th edition (DSMTV),23 patients may be somatic treatment for depression, or toxin expo­ diagnosed with a Major Depressive Disorder if at sure).23 Thus, reviewing the medication profile can least five symptoms are present over a 2-week peri­ uncover potential iatrogenic sources of depression. od. One of the diagnostic symptoms must be either a Although a myriad of drugs have side-effect profiles depressed mood or the loss of interest or pleasure. that resemble depression,28 drugs of abuse (eg, opi­ Symptoms that may be present nearly every day ate, marijuana, and alcohol), antihypertensives (eg, include a depressed mood, diminished interest or pleasure in most activities, significant weight loss or TABLE 1 weight gain (eg, change of more than 5% of body Discriminators of Depression from Medical Illness weight), insomnia or hypersomnia, psychomotor agi­ tation or retardation, fatigue or loss of energy, feel­ • Guilt and self-reproach ings of worthlessness or inappropriate guilt, • Psychomotor retardation decreased ability to think or concentrate, or recur­ • Recurrent suicidal thoughts (except for dialysis and HIV) rent thoughts of death or suicidal ideation. • Sense of failure • Loss of social interest • Feeling punished D ifferential D ia g n o s is Depression may coexist with medical illness. • Persistent dissatisfaction • Indecision However, depression may also be a psychological • Recurrent or persistent episodes of crying response or a direct effect of a physical illness.

Diagnosing depression in medically ill patients is fur­ HIV denotes human immunodeficiency virus. Based on data from ther complicated by medication side effects or symp­ Camaron25; Cassem26; and Clark, Cavanaugh, Gibbons.27 toms of the illness that mimic depression. For exarn-

S 3 6 The Journal of Family Practice, Vol. 43, No. 6 (Suppl), 1996 d e p r essio n in pa tien ts w ith m e d ic a l comorbidity

- TABLE 2 ______reserpine, propranolol, and methyl- dopa), and corticosteroids are fre­ That May Induce Depression quently implicated (Table 2). If drug- induced depression is suspected, Cardiovascular drugs Sedatives/tranquilizers a-methylclopa Barbiturates substituting an alternative agent or Reserpine discontinuing the drug altogether Benzodiazepines Propranolol Ethanol will quickly uncover a drug-related Guanethidine Major and minor .29 C lonidine Age may be helpful in determin­ Thiazide diuretics Anti-inflammatory/ D igitalis ing whether depression or a physi­ anti-infective agents Hydralazine Nonsteroidal anti-inflammatory agents cal illness is the primary disorder. Procainamide Ampicillin Depression is more likely to devel­ Cycloserine op in patients who are young (usu­ Antineoplastics Dapsone ally in their third decade of life) Azathioprine Griseofulvin 6-Azouridine Isoniazid compared with older patients who L-asparaginase Co-trimoxazole develop medical illnesses (usually Bleom ycin Ethambutol in their fifth decade of life).25 C isplatin Disulfiram Depression and medical illness may Cyclophosphamide Sulfonamides develop concurrently, but it is also Doxorubicin Metoclopramide Mithramycin Metronidazole possible that depression is masking Vinblastine Nalidixic acid a newly developing medical condi­ Vincristine tranquilizers Nitrofurantoin tion, such as Huntington’s disease, Procaine penicillin multiple sclerosis, Cushing’s dis­ Anti-Parldnsonian drugs Streptomycin Tetracycline ease, Parkinson’s disease, systemic Amantadine Bromocriptine lupus, and human immunodeficien­ Levodopa cy virus (HIV) encephalopathy.26 Amphetamines (withdrawal) For example, spontaneous crying is Antipsychotic drugs indicative of severe depression, but Haloperidol Cocaine (withdrawal) Methylphenidate it can also be a manifestation of Fluphenazine neurologic disease.30 To reduce the Sedative and antianxiety drugs Hormones potential for misdiagnosis, case Benzodiazepines Oral contraceptives finding for depression in medically Barbiturates ACTH (corticotropin) Glucocorticoids ill patients should be based primar­ Choral hydrate Anabolic steroids ily on the psychological state. Ethanol Physical symptoms should be used Others only to support the diagnosis of Carbamazepine Cim etidine depression after more specific Ethosuximide Disulfiram Methyserglde screening methods have been uti­ Phenobarbital Phenytoin Phenylephrine lized. Ranitidine Prim idone

Ot h e r A f f e c t iv e D is o r d e r s Reproduced from Valente et al, 1994, with permission.-s Adjustment Disorder Medically ill patients with depres­ have been removed, the adjustment disorder sive features, such as flat mood, tearfulness, or resolves within 6 months. Treatment consists of hopelessness, who do not meet the DSMTV crite­ identifying the stressor, relating the symptoms to ria for major depression may be suffering from the stressor, and providing social support.' an adjustment disorder. Adjustment disorder Because depressive features of affective disor­ with depressed mood develops due to a maladap­ ders can develop into a full-blown major depres­ tive response to a stressful life event (eg, marital sion, patients should be continually assessed for difficulties or a disabling general medical condi­ changes in symptom severity. tion).31 Once the stressor or its consequences

The Journal of Family Practice, Vol. 43, No. 6 (Suppl), 1996 S 37 DEPRESSION IN PATIENTS WITH MEDICAL COMORBIDITY

Dysthymia chotherapy or to medication alone. is another affective disorder that com­ Selecting an antidepressant is more of a challenge monly occurs in patients with medical illness. in the setting of medical illness because care must be Dysthymia is a milder, but more chronic, form of taken to avoid agents that exacerbate the medical depression that is present for a majority of days dur­ condition. Other factors to consider include poten­ ing at least a 2-year period in adults.23 Dysthymia usu­ tial antidepressant drug interactions and the effect of ally begins insidiously during childhood or adoles­ impaired renal, hepatic, or gastrointestinal (GI) func­ cence and, in children, continues for a least 1 year.33 tion on drug metabolism and elimination. In some Even though it is considered to be less severe than cases, antidepressant side effects can be exploited to major depression, untreated dysthymia causes a provide a therapeutic benefit to both the depressive greater degree of physical and social impairment and medical illnesses. Antidepressant therapy compared with many other chronic medical illness­ should be tailored to the individual patient based on es, such as hypertension and diabetes.14 tolerability, prior response, and the potential clinical advantage that can be derived from the side effects IMPACT ON MEDICAL ILLNESS of the drug. Based on these factors, treatment rec­ ommendations for various medical conditions are Depressed medically ill patients perceive their gen­ provided below (Table 3). eral health as poor and tend to amplify somatic symptoms of their medical illness. As patients N eurological D is o r d e r s become less cooperative and motivated, the recov­ It has been postulated that some central nervous sys­ ery period is prolonged and the likelihood for a pos­ tem (CNS) disorders are associated with neurotrans­ itive outcome is decreased. Unrecognized depres­ mitter imbalances that precipitate depression. sion may appear to be a worsening medical condi­ Regardless of the mechanism, depression is a com­ tion that leads to costly and unnecessary diagnostic mon consequence of neurological illnesses, such as tests and medications. Even if the medical condition Parkinson’s disease, multiple sclerosis, and stroke;1 is properly treated, severe limitations in physical Because of the impact on quality of life and ability to function may persist if the depression does not function normally, the choice of an antidepressant in remit.35 This can eventually lead to increased mortal­ this patient population is mainly directed toward ity either by suicide or by worsening medical dis­ avoiding agents with anticholinergic and orthostatic ease3^* and to an increased risk of substance abuse.6 properties. Other consequences of untreated depression include Stroke patients are usually elderly and have suf­ social and family dysfunction, failure to advance in fered some degree of cerebral insult. Thus, tricyclic school, and loss of productivity at work. antidepressants (TCAs) are less desirable because patients are more sensitive to the anticholinergic, TREATMENT OPTIONS IN THE sedative, and orthostatic side effects.40 The MEDICALLY ILL monoamine oxidase inhibitors (MAOIs) also are not used because of orthostasis. Besides headache, the Treatment options for depressed patients with other commonly reported side effects of the selec­ severe medical comorbidity include nonpharmaco- tive serotonin reuptake inhibitors (SSRIs), namely logic approaches, medication, and a combination nausea, sleep disturbance, and sexual dysfunction, of medication and . One advantage do not exacerbate neurologic conditions and are of nonpharmacologic therapy is the lack of side safer to administer in stroke patients. The lower inci­ effects that are commonly associated with antide­ dence of anticholinergic and orthostatic effects asso­ pressant medications. Psychotherapy (eg, cogni­ ciated with the secondary amine TCAs (eg, tive, interpersonal, and behavioral therapy) may be desipramine and nortriptyline) also makes these effective in patients with mild to moderate depres­ agents an option. sion. However, it should be noted that many Stroke patients or other patients with insult to the patients are unable to comply with or complete the brain should not receive agents with notable - full course of psychotherapy.39 Patients who do not inducing properties, such as , maprotiline, respond to nonpharmacologic therapy may or clomipramine.41 For cases in which hepatic respond to a combination of medication and psy­ enzyme-inducing antiepileptics (eg, phenytoin and

S 3 8 The Journal of Family Practice, Vol. 43, No. 6 (Suppl), 1996 DEPRESSION IN PATIENTS WITH MEDICAL COMORBIDITY

Antidepressant Treatment Recommendations for Various Medical Conditions

Medical Condition Antidepressant of Choice Drugs to Avoid

Neurological disorders Parkinson’s disease Bupropion, TCAs, MAOIs, SSRis (if coprescribed with selegiline)

Stroke SSRis, possibly secondary amines TCAs, MAOIs, bupropion Alzheimer’s SSRis, nefazodone, venlafaxine TCAs, bupropion, and if dementia accompanied with agitation Cancer Severe cancer pain TCAs, but watch for constipation MAOIs requiring opiates Cachexia , , trazodone, Fluoxetine nefazodone, TCAs, venlafaxine

Nausea No specific advantages Bupropion, SSRis, venlafaxine Endocrine disorders No specific advantages No specific disadvantages Cardiac disease SSRis, bupropion, nefazodone TCAs, venlafaxine in patient with high blood pressure, SSRis with type 1C antiarrhythmics (eg, encainide and flecainide)

Allergic conditions No specific advantages Coprescribing SSRis, nefazodone, and fluvoxamine with terfenadine, astemizole

MAOIs denotes monoamine oxidase inhibitors; SSRis, selective serotonin reuptake inhibitors; TCAs, tricyclic antidepressants. Based on data from Haig7; C um m ings” ; Reynolds"0; Cunningham4'; Schowersky and deVries'0; Raskind"; Guze and Gitlin"; Maguire et al” ; Jackson et al*; Glassman et al47; Veith et a!48; McElroy et al49; Raghebw; and DeVaneT

barbiturates) are added to preexisting TCA therapy, hypotension increases the risk for falls in elderly increased TCA doses may be needed to compensate patients with an unsteady gait. Another limitation of for decreased serum concentrations. The SSRIs, par­ TCA therapy in patients with Parkinson's disease is a ticularly fluvoxamine and fluoxetine, may inhibit drug interaction with levodopa. It is important to cytochrome P450 3A4 and 2C, resulting in increased optimize levodopa absorption so that the maximum levels of carbamazepine and phenytoin, respective­ amount is available for conversion to dopamine. ly.52 Therefore, selection of an agent in patients with Consequently, drugs that decrease the absorption of seizure potential should take into consideration the levodopa, such as the TCAs, should be avoided. Another strategy that can be used to optimize antidepressant side-effect and drug-interaction pro­ dopamine concentrations in the CNS is to avoid files. dopaminergic blocking agents, such as amoxapine. The pathophysiology of Parkinson’s disease is Conversely, antidepressants that block dopamine attributed to decreased dopamine, norepinephrine, uptake (eg, bupropion and venlafaxine) may have and serotonin and to increased acetylcholine levels dual therapeutic benefit by being effective for in the CNS. The anticholinergic effects of the TCAs depression in addition to augmenting other copre­ may improve impaired motor function, but they can scribed drug therapies for Parkinson’s disease, such also aggravate other Parkinsonian symptoms, such as levodopa and amantadine. Improvement of sero­ as cognitive deficits, constipation, somnolence, and tonin imbalances by the SSRis may also be advanta­ urinary retention.15 Some patients already receiving geous in ameliorating Parkinsonian as well as anticholinergic agents, such as benztropine and tri- depressive symptoms. Case report data suggests a hexyphenidel, may not be able to tolerate the addi­ possible association between fluoxetine and tive anticholinergic symptoms when TCAs are extrapyramidal symptoms, which may offset this coprescribed. In addition, TCA-induced orthostatic

The Journal of Family Practice, Vol. 43, No. 6 (Suppl), 1996 S 39 DEPRESSION IN PATIENTS WITH MEDICAL COMORBIDITY

clinical benefit.53'56 Drug treatment options for with opiates is common clinical practice even in Parkinson’s disease include the monoamine oxidase patients with no depressive symptoms. This combi­ B inhibitor selegiline. Addition of a serotonergic nation has been effectively used to provide equipo- drug to a regimen with an MAOI could potentiate a tent pain relief with a lower opioid dose. Patients serotonergic syndrome leading to delirium, coma with insomnia or diarrhea caused by radiation thera­ and, in some cases, death. This interaction has py may also benefit from the sedative and anti­ occurred after concomitant use of sertraline and cholinergic side effects of TCAs.7 In patients with selegiline.42 Because other antidepressants have sim­ cachexia, fluoxetine, which has been associated ilar effects on serotonin receptors in the brain, they with weight loss, should be avoided. Instead, some should all be suspected of having the same drug physicians have found that weight-neutral agents interaction potential with selegiline. (eg, sertraline and paroxetine)44 or weight-promoting Depression is not an expected psychological con­ agents (eg, doxepin, , and ) sequence of Alzheimer’s dementia. Distinguishing are useful. between the diagnosis of depression as opposed to Although the TCAs are effective for treating dementia is confusing because many of the charac­ depression and other symptoms associated with teristic symptoms overlap (eg, sleep and appetite cancer, side effects can still be problematic and changes, loss of interest in usual activities, and psy­ result in a high rate of noncompliance.4sa chomotor agitation or retardation).57 If depression is Anticholinergic side effects, such as constipation present, improved motivation and functional and and dry mouth, can be troublesome for patients cognitive status will become evident with treatment. recovering from GI surgery or stomatitis. Agitated patients may benefit from the sedative Electrocardiographic changes associated with properties of TCAs or trazodone. However, drugs TCAs may potentiate the cardiotoxicity of doxoru­ with sedative properties may actually worsen cogni­ bicin. Therapeutic doses of meperidine used with­ tive impairment in patients with dementia.41 This is a in 14 days of an MAOI have resulted in an opiate result of the anticholinergic properties of TCAs, overdosage syndrome characterized by symptoms which induce confusion, short-term memory loss, of coma, severe respiratory depression, cyanosis, poor concentration and attention, and disorientation and hypotension.65 As a result, MAOIs should be and delirium.43 Because of the activating properties administered with caution, if at all, to cancer of fluoxetine58450 and bupropion,6162 some primary patients who are receiving opiates for chronic care physicians avoid using them in patients with pain management.41 The use of morphine or fen- dementia accompanied by agitation. In the elderly, tanyl is preferred over meperidine in patients the SSRIs may be better choices because their selec­ receiving MAOIs. tive mechanism of action results in milder, better tol­ Nausea is a frequent complication of cancer erated side effects. Other agents that do not have resulting from GI oncologic processes or emeto- negative cognitive effects include nefazodone and genic chemotherapy. This undesirable aspect of venlafaxine. the disease needs to be taken into consideration when prescribing antidepressants associated C a n c e r with nausea, such as the SSRIs, bupropion, and Depression in cancer patients may be due to biolog­ venlafaxine. Nausea associated with SSRI thera­ ic changes caused by the disease.® On the one hand, py is generally transient, dose-dependent, and a developing cerebral metastasis or paraneoplastic tends to resolve after several weeks of therapy. syndrome can cause CNS or endocrine disturbances Patients who are particularly troubled by nausea that present as a subclinical depression. Also, the may choose to take their dose in the evening or emotional stress of coping with cancer may be the switch to an alternate SSRI and then to an alter­ catalyst for depression. Cancer patients at increased nate class of antidepressants. Although TCAs risk for depression include those who need multiple may be appropriate in some situations, overall, courses of poorly tolerated chemotherapy, who have the SSRIs, trazodone, nefazodone, or venlafaxine recurring disease, or who have a poor prognosis.18 are preferred in cancer patients because of the TCAs are clinically useful in this patient population lower risk for complicating side effects and the because their therapeutic benefits extend beyond decreased potential for lethality in overdose situ­ their efficacy in depression. Coprescribing TCAs ations.

S 4 0 The Journal of Family Practice, Vol. 43, No. 6 (Suppl), 1996 DEPRESSION^ PATIENTS WITH MEDICAL COMORBIDITY

En d o c r in e D is o r d e r s rationale to treat depression in patients with preex­ Corticosteroid abnormalities have a number of bio­ isting heart disease.16 Even after controlling for other logical effects on tire CNS that commonly precipitate clinical variables, such as the degree of cardiac depressive disorders. Almost every endocrine dis­ impairment and previous Mis, depression itself is a ease has been associated with depression, but only a significant risk factor for dying in the 6 months after few (eg, Cushing’s syndrome, Addison’s disease, an MI (Figure 2).® It is recommended that TCAs be hyperthyroidism, hypothyroidism, and hyperpro­ avoided in this patient population because they slow lactinemic amenorrhea) cause symptoms that reach cardiac conduction and cause greater postural blood the intensity of a major depressive disorder.66 pressure changes." Another precaution associated Clinical response to treatment and specific disease- with TCA therapy is a potential drag interaction wit h related features can be used to determine whether warfarin. Because anticoagulation with warfarin in endocrine disease or depression is the primary dis­ the post-MI phase is becoming common practice, order. For example, depression in Cushing’s syn­ primary care physicians will need to be more cog­ drome responds poorly to antidepressant drags but nizant of this drag interaction and the associated improves rapidly with steroid inhibitors. increased risk for hemorrhagic events. Furthermore, patients with Cushing’s syndrome Patients with cardiac conduction disease are par­ rarely experience major psychiatric disturbances ticularly prone to the unpredictable and even fatal other than affective disorders (eg, depression and antiarrhythmic effects of TCAs. An important impli­ related and irritable mood). Because stress­ cation of this antiarrhythmic effect is the potential ful life events can induce both depression and pitu­ for high TCA concentrations to induce ventricular itary-adrenal imbalances, it is possible

that these disorders may coexist with FIGURE 2 no physiological link.67 It is recognized that hormone imbalances can cause Depression is a significant risk factor for mortality in the 6 months after depression, but physiologic changes myocardial infarction (Ml). Reproduced from Frasure-Smith et al, 1993, with permission.68 can also be indirectly responsible for depressive symptoms; for instance, contending with the social stigma of acromegaly or hirsutism may be the source of depression. Therefore, not all cases of depression coexisting with endocrine disorders will resolve once the hormonal imbalances are corrected. It is possible that personal distress may be the actual source of depression, in which case, treating the condition with an antidepressant is appropriate.

Ca r d ia c D is e a s e Loss of control over important aspects of life, such as employment, independence, or anxiety about recur­ rent cardiac events, contributes to the prevalence of depression in patients with myocardial infarction (MI), con­ gestive heart failure, coronary bypass surgery, and other cardiac conditions. Because untreated depression pro­ longs recovery and increases the risk of cardiac death, there is a compelling

The Journal of Family Practice, Vol. 43, No. 6 (Suppl), 1996 S 41 DEPRESSION IN PATIENTS WITH MEDICAL COMORBIDITY

arrhythmias. Thus, it is advisable that other antide­ histaminergic, and a-adrenergic receptors. $1 pressants be used in place of TC As so that the poten­ SSRIs should not be administered concomitantly to tial for cardiac complications can be avoided, espe­ patients being treated with type IC antiarrhyth* cially in patients with suicidal tendencies. TCAs (eg, encainide, flecainide, and propafenone) taken with sympathomimetic amines can also result because SSRIs block the metabolism of these antiar- in tachycardia, arrhythmias, and hypertension. rhythmics (via inhibition of cytochrome P450 2D6 Because some of these products are available as resulting in high concentrations and, possibly, treat' over-the-counter purchases in the form of deconges­ ment-emergent arrhythmias. Bupropion may be a tants or weight-loss products, patients can precipi­ better choice in these circumstances. However, tate this reaction inadvertently if not properly these cases are seldom managed in the family prac! warned. More severe, life-threatening hypertensive tice setting because these patients often require spe­ crises can occur if MAOIs are taken with sympath­ cialized care from cardiologists. omimetic amines or tyramine-containing foods. Therefore, it is best to avoid the likelihood of these CONCLUSIONS serious events occurring by selecting antidepres­ sants that are not associated with adverse cardiac Treating depression in patients with significant med­ events, such as the SSRIs, bupropion, or nefazodone. ical comorbidity is further complicated by physio­ TCA-induced orthostatic effects also increase in logic changes, drug interactions, and drug side frequency and magnitude with worsening left-ven­ effects that are not present in physically healthy but tricular function. In one study, approximately 50% depressed patients. In the past, physicians have of patients with congestive heart failure experi­ often hesitated to prescribe antidepressant therapy enced such severe orthostatic hypotension with for fear that side effects would exacerbate the under­ imipramine therapy (drops in systolic blood pres­ lying medical condition. Today, newer antidepres­ sure of up to 44 mm Hg) that they withdrew from sants with specific mechanisms of action represent a the 3-week trial.47 Nortriptyline causes less ortho­ significant advance in the treatment of medically ill static hypotension compared with imipramine, patients. Because they are well tolerated and do not amitriptyline, and desipramine,47'48'69 70 and is a ratio­ increase morbidity, these agents offer a rational nal choice in the rare cases when a TCA is consid­ treatment option for the medically ill depressed ered necessary. patient. Other patients who need careful consideration Drug interactions can also determine the are those with borderline or unstable hypertension. choice of an antidepressant. Because there are Higher doses of venlafaxine (eg, 37.5 mg or greater) now several antidepressants available from which have been associated with small but clinically signif­ to choose, drug interactions can often be avoided. icant increases in blood pressure.4971 The likelihood Although this review covered only the disease for this adverse effect to occur is unpredictable states in which depression occurs most common­ because it is not related to any patient-specific char­ ly, family physicians should be aware that antide­ acteristic, such as age, sex, or renal or hepatic func­ pressants interact with other drugs that are com­ tion.72 Consequently, alternatives to venlafaxine monly prescribed. For example, nefazodone and should be considered for hypertensive patients who fluvoxamine are contraindicated for therapy with cannot tolerate additional increases in blood pres­ terfenadine and astemizole because of the risk of sure. TCAs should also be avoided in hypertensive fatal arrhythmias. The mechanism for this drug patients because they interact with many antihyper­ interaction has been implicated to be the same as tensive agents to either increase (eg, guanethidine, that which has occurred when ketoconazole is clonidine, and methyldopa) or decrease (eg, p-block- adm inistered w ith these antihistamines. er, calcium channel blocker, diuretics, angiotensin­ Fluvoxamine also increases theophylline levels, converting enzymes) blood pressure.50 SSRIs and which requires close monitoring during concomi­ bupropion are considered the first-line treatment for tant therapy. Thus, it is good practice to continu­ patients with cardiac disease because they have a ally review each patient’s medication profile more favorable side-effect profile. They have mini­ before beginning or changing his or her antide­ mal effects on pulse, blood pressure, and cardiac pressant therapy. conduction due to their low affinity for muscarinic, Depression should not be dismissed by the

S 4 2 The Journal of Family Practice, Vol. 43, No. 6 (Suppl), 1996 d e p r e s s io n in patients w ith m e d ic a l comorbidity

patient and primary care physician as an under­ 19. Koenig HG, Meador KG, Cohen HJ, Blazer DG. Depression in standable or normal reaction to the individual’s dis­ MedlgftjSTg^ie11™18 "lediCaI fflnea8- Arch Imom ease. If untreated, it can hinder the recovery process 20. Cavanaugh S, Clark DC, Gibbons RD. Diagnosing depression in the hospitalized medically ill. Psychosomatics 1983; 24•809- and increase morbidity and mortality. The burden of 15. a mood disorder to the patient, to his or her family, 21. Moffic HS, Paykel ES. Depression in medical in-patients. Br J and to society is unnecessary because depression is 1975; 126:34653. 22. Rovner BW, German PS, Brant U , Clark R, Burton L, Folstein a highly treatable disease. Approximately 80% to MF. Depression and mortality in nursing homes. JAMA 1991; 90% of cases resolve with appropriate treatment.73 23. DSM-IV. Mood disorders. Diagnostic and statistical manual of Therefore, family physicians should not hesitate to mental disorders. 4th ed. Washington, DC: American provide counseling, support, and medication when Psyclriatric Association, 1994:317-91. 24. Neese JB. Depression in the general hospital. Nurs Clin North depression is present. An increased awareness of the Am 1991; 26:613-22. indirect or covert manifestations of depression in the 25. Cameron OG. Guidelines for diagnosis and treatment of depression in patients with medical illness. J Clin Psychiatry medically ill will enable rapid diagnosis and effective 1990; 51(suppl 7):49-54. therapy. 26. Cassem EH. Depression and anxiety secondary to medical ill­ ness. Psychol Clin North Am 1990; 13:597-612. REFERENCES 27. Clark DC, Cavanaugh SV, Gibbons RD. The core symptoms of 1. Wells KB, Stewart A, Hays RD, et al. The functioning and well­ depression in medical and psychiatric patients. J Nerv Ment Dis 1983; 171:70613. being of depressed patients. Results from the Medical Outcomes Study. JAMA 1989; 262:914-19. 28. Valente SM, Saunders JM, Cohen MZ. Evaluating depression among patients with cancer. Cancer Pract 1994; 2:6671. 2. Regier DA, Narrow WE, Rae DS, Manderscheid RW, Locke BZ, Goodwin FK. The defacto US mental and addictive disorders 29. Chamey DS, Miller HL, Licinio J, Salomon R. Treatment of service system. Epidemiologic Catchment Area prospective 1- depression. In: Schatzberg AF, Nemeroff CB, eds. Textbook of year prevalence rates of disorders and services. Arch Gen psychopharmacology. Washington, DC: American Psychiatric Psychiatry 1993; 50:85-94. Press, 1995:576601. 30. 3. Perez-Stable EH, Miranda J, Munoz RF, Ying YW. Depression in Green RL, McAllister TW, Bemat JL. A study of crying in med­ medical outpatients: underrecognition and misdiagnosis. Arch ically and surgically hospitalized patients. Am J Psychiatry 1987; 144:442-7. Intern Med 1990; 150:1083-8. 31. DSM-IV. Appendix B. Adjustment disorders. Diagnostic and 4. Schurman RA, Kramer PD, Mitchell JB. The hidden mental statistical manual of mental disorders. 4th ed. Washington, DC: health network: treatment of mental illness by nonpsychiatrist American Psychiatric Association, 1994:623-7. physicians. Arch Gen Psychiatry 1985; 42:89-94. 32. Rodin G, Voshart K Depression in the medically ill: an 5. Katon W, Schulberg H. Epidem iology o f depression in primary overview. Am J Psychiatry 1986; 143:696705. care. Gen Hosp Psychiatry 1992; 14:237-47. 33. Ferentz KS. Recognizing and treating dysthymia in the prima­ 6. Wells KB, Golding JM, Bum am MA. Psychiatric disorder in a ry care patient [special report]. Postgrad Med 1995; Dec:13-20. sample of the general population with and without chronic Hays RD, Wells KB, Sherboume CD, Rogers W, Spritzer K. medical conditions. Am J Psychiatry 1988; 145:976-81. 34. Functioning and well-being outcomes of patients with depres­ 7. Haig RA. Management o f depression in patients with advanced sion compared with chronic general medical illnesses. Arch cancer. Med J Aust 1992; 156:499-503. Gen Psychiatry 1995; 52:11-19. 8. McDaniel JS, M usselm an DL, Porter MR, Reed DA, Nemeroff 35. Schubert DSP, Taylor C, Lee S, Mentari A, Tamaklo W. Physical CB. Depression in patients with cancer. Diagnosis, biology, and consequences of depression in the stroke patient. Gen Hosp treatment. Arch Gen Psychiatry 1995; 52:89-99. Psychiatry 1992; 14:69-76. 9. Clothier J, Grotta J. Recognition and management of post­ 36. Beskow J. Depression and suicide. Pharmacopsychiatry 1990; stroke depression in the elderly. Clin Geriatr Med 1991; 7:493- 23(suppl l):3-8. 506. 37. Bruce ML, Leaf PJ. Psychiatric disorders and 16month mor­ 10. McNamara ME. Psychological factors affecting neurological tality in a community sample of older adults. Am J Public conditions: depression and stroke, multiple sclerosis, Health 1989; 79:727-30. Parkinson’s disease, and epilepsy. Psychosomatics 1991; 38. Murphy E, Smith R, Lindesay J, Slattery J. Increased mortality 32:255-67. rates in late-life depression. Br J Psychiatry 1988; 152:347-53. 11. Robinson RG, Starr LB, Kubos KL, Price TR. A two-year longi­ 39. iepression Guideline Panel. Depression in primary care: vol- tudinal study of post-stroke mood disorders: findings during me 2. Treatment of mqjor depression. Clinical Practice the initial evaluation. Stroke 1983;14:736-41. luideline, number 5. Rockville, Md: US Department of Health 12. Carney RM, Rich MW, Tevelde A Saini J, Clark K, Jaffe AS. nd Human Services, Public Health Service, Agency for Health Mqjor depressive disorder in coronary artery disease. Am J )are Policy Research; 1993. AHCPR publication No. 93-0551. Cardiol 1987; 60:1273-5. 40. ieynolds CF III. TYeatment of depression in special popula- 13. Fielding R. Depression and acute myocardial infarction: a ions. J Clin Psychiatry 1992; 53(suppl 9):4653. review and reinterpretation. Soc Sci Med 1991; 32:1017-27. 41. 'unningham LA. Depression in the medically ill: choosing an 14. Roose SP, Dalack GW, Woodring S. Death, depression, and ntidepressant. J Clin Psychiatry 1994; 55(9, suppl A):967. heart disease. J Clin Psychiatry 1991; 52(suppl 6):34-9. 42. .chowersky 0, deVries JD. Interaction of fluoxetine and selegi- 15. Cummings JL. Depression and Parkinson’s disease: a review. ne. Can J Psychiatry 1990; 53:571-2. Am J Psychiatry 1992; 149:443-54. 43. laskind MA. Management of late-life depression and the 16. Guze BH, Barrio JC. The etiology of depression in Parkinson’s oncognitive behavioral disturbances of Alzheimer’s disease. disease patients. Psychosomatics 1991; 32(4):390-5. 'sychol Clin North Am 1993; 16:81627. 17. Mayberg HS, Solomon DH. Depression in Parkinson’s disease: 44. luze BH, Gitlin M. New antidepressants and the treatment of a biochem ical and organic viewpoint. Adv Neurol 1995; 65:49- epression. J Fam Pract 1994; 38:49-57. 60. 45. laguire P, Hopwood P, Tarrier N, Howell T. Treatment of 18. Katon W, Sullivan MD. Depression and chronic medical illness. epression in cancer patients. Acta Psychiatr Med 1985; J Clin Psychiatry 1990; 51(suppl 6):3-ll. 2(suppl):814.

The Journal of Family Practice, Vol. 43, No. 6 (Suppl), 1996 S 4 3 DEPRESSION IN PATIENTS WITH MEDICAL COMORBIDITY

46. Jackson WK, Roose SP, Glassman AH. Cardiovascular toxicity 60. Rothschild AJ, Locke CA Reexposure to fluoxetine after ^ of antidepressant medication. Psychopathology 1987; 20(suppl ous suicide attempts by three patients: the role of aka - l):64-74. Clin Psychiatry 1991; 52:491-3. 47. Glassman AH, Johnson LL, Giardina EGV, Walsh T, Roose SP, 61. Rudorfer MV, Potter WZ. Antidepressants. A compar* Cooper TB, Bigger JT. The use of imipramine in depressed review of the clinical pharmacology and therapeutic use off patients with congestive heart failure. JAMA 1983; 250:1997- “new er” versus the “older” drugs. Drugs 1989; 37:713-8. 2001. 62. Golden RN, Bebchuk JM, Leatherman ME. Trazodone ar- 48. Veith RC, Raskind MA, Caldwell JH, Bames RF, Gumbrecht G, other antidepressants. In: Schatzberg AF, Nemeroff CB e4 Ritchie JL. Cardiovascular effects of tricyclic antidepressants Textbook of psychopharmacology. Washington, DC- Amenr ' in depressed patients with chronic heart disease. N Engl J Med Psychiatric Press, 1995:195-213. 1982; 306:954-9. 63. Whitlock FA, Siskind M. Depression and cancer: a follow m 49. McElroy SL, Keck PE Jr, Friedman LM. Minimizing and man­ study. Psychol Med 1979; 9:747-52. aging antidepressant side effects. J Clin Psychiatry 1995; 56(suppl 2):49-55. 64. Popkin M, Callies A, Mackenzie T. The outcome of antidepres sant use in the m edically ill. Arch Gen Psychiatry 1985; 42:ii(j 50. Ragheb M. Drug interactions in psychiatric practice. Int 3. Pharmacopsychiatry 1981; 16:92-118. 65. Krishnan KRR. Monoamine oxidase inhibitors. In: Schatzben 51. DeVane CL. Pharmacokinetics of the newer antidepressants: AF, N em eroff CB, eds. Textbook o f psychopharmacolog: clinical relevance. Am J Med 1994; 97(suppl 6A):13S-23S. Washington, DC: American Psychiatric Press, 1995:575601. 52. Nemeroff CB, DeVane CL, Pollock BG. Newer antidepressants 66. Fava GA Sonino N, Morphy MA. Mgjor depression associated and the cytochrome P450 system. Am J Psychiatry 1996; with endocrine disease. Psychiatr D evel 1987; 4:32148. 153:311-20. 67. Fava GA. Affective disorders and endocrine disease 53. Baldwin D, Fineberg N, Montgomery S. Fluoxetine, fluvoxam- Psychosomatics 1994; 35:341-53. ine, and extrapyramidal tract disorders. Int Clin Psychopharm acol 1991; 6:51-8. 68. Frasure-Smith N, Lesperance F, Tal^jic M. Depression follow­ ing myocardial infarction. Impact on 6-month survival. JAIL! 54. Bouchard RH, Pourcher E, Vincent P. Fluoxetine and 1993; 270:1819-25. extrapyramidal side effects [letter]. Am J Psychiatry 1989; 146:1352-3. 69. Roose SP, Glassman AH, Siris SG, Walsh T, Bruno RL, Wright 55. Budman CL, Bruun RD. Persistent dyskinesia in a patient LB. Com parison o f imipramine- and nortriptyline-induced receiving fluoxetine [letter]. Am J Psychiatry 1991; 148:1403. orthostatic hypotension: a meaningful difference. J Clin Psychopharm acol 1981; 1:316-9. 56. Caley CF, Friedman JH. Does fluoxetine exacerbate Parkinson’s disease? J Clin Psychiatry 1992; 53:278-82. 70. R oose SP, Glassm an AH, Giardina EGV, et al. Nortriptyline in depressed patients with left ventricular impairment. JAMA 57. Katz IR, Streini J, Parm elee P. Prevention o f depression, recur­ 1986; 256:3253-7. rences, and complications in late life. Prev Med 1994; 23:743- 50. 71. Ellingrod VL, Perry PJ. Venlafaxine: a heterocyclic antidepres­ sant. Am J Hosp Pharm 1994; 51:303346. 58. Amsterdam JD, Homig-Rohan M, Maislin G. Efficacy of alpra­ zolam in reducing fluoxetine-induced jitteriness in patients 72. Feighner JP. The role of venlafaxine in rational antidepressant with m^jor depression. J Clin Psychiatry 1994; 55:394-400. therapy. J Clin Psychiatry 1994; 55(9, suppl A):62-8. 59. Beasley CM Jr, Sayler ME, Weiss AM, Potvin JH. Fluoxetine: 73. Regier DA, Boyd JH, Burke JD Jr, et al. One-month prevalence activating and sedating effects at multiple fixed doses. J Clin of mental disorders in the United States. Arch Gen Psychialn Psychopharmacol 1992; 12:328-33. 1988; 45:977-86.

DISCUSSION pseudodementia is a common way for depression to Dr Richardson: Most patients in primary care come present in the elderly. I think as long as you use a safe in with a hybrid of medical problems, and it is easy agent and monitor the patient’s response, a trial with to forget about the depressive component. an antidepressant is justified. Sometimes we shy away from working up depres­ sion because we are consumed with the patient’s Dr McCoy: These examples emphasize the point that multiple problems. However, it could be that depres­ family doctors should remember that depression in sion is presenting in the form of somatic symptoms. medical illness can occur coincidentally, can be a true If we would simply spend a few minutes to screen for symptom of the medical disorder, or can be a sec­ depression, we might save ourselves a lot of time in ondary or functional reaction to impairment. There the long run. It could be that once we treat the are quick and easy screening methods that can be depression appropriately, all the patient’s physical used to discriminate depression from medical illness, symptoms will disappear. and I think it would really be worth our time to per­ form them. Once we identify and treat the depres­ Dr De Wester: I think an example of this is in the sion, it could curtail the number of patients who demented patient. These patients have been on mul­ repeatedly return to the office with illegitimate med­ tiple medications chronically, and we always meet a ical complaints and save us from having to perform lot of resistance anytime we try to change anything unnecessary tests and procedures. It may also or add anything new to their medication regimen. A improve the condition of patients with true medical common misconception is that dementia is a natural illnesses because they start feeling better about them­ process of growing old, but we all know that selves and less handicapped by their medical illness.

S 4 4 The Journal of Family Practice, Vol. 43, No. 6 (Suppl), 1996