NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Depression Guideline Panel. Depression in Primary Care: Detection and Diagnosis (Volume 1: Detection and Diagnosis). Rockville (MD): Agency for Health Care Policy and Research (AHCPR); 1993 Apr. (AHCPR Clinical Practice Guidelines, No. 5.1.)

  • This publication is provided for historical reference only and the information may be out of date.

This publication is provided for historical reference only and the information may be out of date.

Cover of Depression in Primary Care: Detection and Diagnosis (Volume 1: Detection and Diagnosis)

Depression in Primary Care: Detection and Diagnosis (Volume 1: Detection and Diagnosis).

Show details

5Guideline: Depression Associated with Medications

Various medications have long been reported to cause or to be associated with mood symptoms or formal disorders as side effects. The development of depression in some patients taking reserpine formed one of the bases for the biologic theories of depression. Most of the evidence rests on case reports of medications "causing" mood symptoms. Table 7 lists those agents reported to have been associated with depression in some patients.

Table 7. Medications reportedly associated with depression.


Table 7. Medications reportedly associated with depression.

It is essential to recognize that idiosyncratic reactions to medications do occur. Even without data to suggest a causal relationship between a drug and mood symptoms, good clinical judgment dictates that it should be stopped or changed if a patient develops depression after beginning use. However, such an event does not suggest that the particular medication should not be used in other patients who appropriately require it, but may have either a depression or a propensity for depression. That is, the reaction should be regarded as truly idiosyncratic and should not form the basis for a general conclusion about the medication in all patients.


Various blood pressure medications have been linked to depressive symptomatology, although much of the evidence for the association has been weak or equivocal.

Goodwin and Bunney (1971) found a 5 to 20 percent risk of depression in patients treated with reserpine. Depression was related to the dosage; more severe depressions were reported in patients receiving more than 0.5 mg per day. A history of depression was associated with a higher incidence of depressive symptoms and a higher incidence of severe depression. These data suggest that, if reserpine is used, the dosage should not exceed 0.5 mg per day, and the drug should be avoided in patients with a history of mood disorders.

Beta-adrenergic blocking agents have been associated with depressive symptomatology, with lethargy being the main symptom reported. In one study, depression was found to be no greater, and perhaps less, in the beta-blocker group than in controls (Bartels, Glasser, Wang, et al., 1988). Another study evaluated various beta-blockers at varying dosages (Carney, Rich, teVelde, et al., 1987). The treated group experienced a 21 percent incidence of depression, but the control group of hypertensive patients on other medications had a 33 percent incidence of depression. Bant (1978) compared hypertensive patients on various medications and a control group of medically ill (nonhypertensive) patients. He found a "high" prevalence of depression in both groups, but no preponderance in the hypertensive patients. Hypertensive patients with a personal or family history of psychiatric illness had a higher level of depression. The Veterans Administration Cooperative Study used patient symptom questionnaires and reported only a 1 percent incidence of depression with propranolol, which is consistent with that found in the general population (VA Cooperative Study Group on Hypertensive Agents, 1982). On the other hand, patients with a personal or family history of depression may be prone to develop depression when treated with propranolol. Data are not available to recommend avoiding propranolol in such patients, but the clinician should be alert to the possibility of depressive reactions in patients beginning this treatment. There is no clear-cut contraindication for its use in any particular patient group.

Studies reviewed fail to show a causal relationship between alpha-methyldopa given in typical dosages and depression. While alpha-methyldopa can be used in patients with a history of depression, they appear to be at greater risk for depression while they are taking it.

A review of 44 studies evaluating the use of clonidine for hypertension found a 1.5 percent incidence of depression (Paykel, Fleminger, and Watson, 1982). This finding indicates that clonidine is rarely associated with depression.

The psychiatric side effects of calcium-channel blockers have been reported to include depression. While most of the data are from case reports, some comprehensive, double-blind, crossover studies have evaluated the potential use of verapamil in the treatment of bipolar disorder. Since there are no controlled reports of depressive disorders associated with calcium-channel blockers, they can be used in depressed patients. At present, there is no evidence that angiotensin I converting enzyme (ACE) inhibitors are associated with depression.


Glucocorticoids are well known to cause depression or psychosis. Lewis and Smith (1983) found that 5 percent of steroid-treated patients experienced severe psychiatric reactions (mostly mood disorders), which usually occurred early in the course. Risk factors were female gender, a diagnosis of systemic lupus erythematosus, and high doses of prednisone.

The practitioner is advised to monitor patients on steroids carefully for development of mood syndromes (both manic and depressive types). If they occur, treatment with steroid taper and use of neuroleptics or ECT is indicated. Tricyclic antidepressants may be less useful and may be ineffective or actually aggravate symptoms (Hall, Popkin, and Kirkpatrick, 1978). In patients for whom steroids are medically necessary, lithium prophylaxis can be tried (Falk, Mahnke, and Poskanzer, 1979).

Anabolic steroids used for athletic enhancement deserve comment. Pope and Katz (1988) found that 22 percent of young bodybuilders who returned their questionnaires suffered from a full mood syndrome, 12 percent had psychotic symptoms, and 12 percent suffered depression when cycled off the drug. The greatest use of these drugs seems to be among young men and women focused on enhancing their physiques. Practitioners should be aware of these drugs and ask about them in all who present with a mood syndrome.

Case reports of depression associated with the use of oral contraceptives are available, but many of the better documented studies found little relationship. Slap (1981) found that 9 of 12 studies reported some depressive symptoms in 15 to 56 percent of oral contraceptive users. However, many of these studies were not well controlled or did not use standard criteria to define depression. One study found a 6.6 percent incidence of depression in those who used oral contraceptives versus a 2 percent incidence in those who had never used them (Herzberg and Coppen, 1970); a 10 percent incidence rate was found in a followup study (Herzberg, Johnson, and Brown, 1970). These agents are not contraindicated in depressed or formerly depressed patients; however, careful observation is recommended when initiating treatment.

Histamine-2 Receptor Blockers

Multiple case reports have linked histamine-2 receptor blockers (cimetidine and ranitidine) to depressive, manic, and psychotic behaviors. When closely evaluated, however, these cases occur in patients who are otherwise severely ill with multiple system failure or renal or hepatic insufficiency. The panel recommends individualized dosage reduction if the drug is to be used in patients known to have renal or hepatic impairment.


Numerous investigators have evaluated the depressant effects of phenobarbital and carbamazepine. Many, however, do not define depression in standard ways. "Depression" is usually attributed to the generalized psychomotor slowing and sedation noted in association with these drugs. In a carefully documented study of patients on phenobarbital and carbamazepine using depression criteria, 40 percent of 64 patients on anticonvulsant drugs had clinical major depression (34 percent with a family history of mood disorder) (Robertson, Trimble, and Townsend, 1987). Patients on phenobarbital were more depressed than those on carbamazepine. There are no comparative controlled studies.

The panel suggests close monitoring of all patients on phenobarbital for signs of depression, especially those with a known personal or family history of depression. No current data suggest that drug levels predict the risk of depression in these patients, and carbamazepine has actually been an effective prophylactic agent for both depressive and manic episodes in bipolar disorder (Ballenger and Post, 1980).


Another therapeutic agent long associated with mood symptoms is levodopa (L-dopa). Most of the studies reviewed did not use currently accepted criteria for diagnosis of depression and employed variable assessment scales.

In an evaluation of 33 Parkinson's disease patients on various forms of therapy (anticholinergics, amantadine, L-dopa, and L-dopa plus carbidopa), 22 were depressed (12 had a history of mood disorders) and 11 had no mood symptoms (none had a history) (Mindham, Marsden, and Parkes, 1976). There was no general relationship between L-dopa dosage and psychiatric disturbance, though those on L-dopa may have had increased mood symptoms. Given the frequency of depression in untreated Parkinson's disease, the effects of these medications remain unclear. However, there is no clear evidence that L-dopa increases the incidence of depression in these patients when there is no history of depression.


Mood disorders associated with antibiotics have largely been noted in case reports, without controlled studies. Dapsone has been reported to cause depression and anger. In these cases, symptoms have resolved within days of discontinuing the drug, and the symptoms have recurred on rechallenge. Isoniazid is reported to be related to psychotic syndromes and delirium. Amphotericin B shows a dose-related delirium and electroencephalograph changes when injected intrathecally. By and large, antibiotic use should be based on medical indications rather than on concern about mood disorders.


Digoxin has been studied only with regard to its toxicity and the related delirium. There are no reports of mood disorders per se related to digoxin. Procainamide has been linked in several case reports to mania.


  • PubReader
  • Print View
  • Cite this Page

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...