Antidepressants and the Risk of Falls Among Nursing Home Residents.

Thapa PB, Gideon P, Cost TW, et al (Division of Pharmacoepidemiology, Department of Preventive Medicine, Vanderbilt University School of Medicine, Nashville, Tenn; Cost Consulting, Nashville, Tenn; Pharmaceutical Consulting Services, Murfreesboro, Tenn), N Engl J Med. 1988;339:875-882.

The authors wanted to determine whether the use of certain types of antidepressant drugs commonly prescribed to residents of nursing homes would result in different rates of falls. Tricyclic antidepressants have been related to increased rates of falls, whereas selective serotonin reuptake inhibitors have fewer psychomotor effects and are believed to be safer for elderly patients who are frail. A third antidepressant type that was studied was trazodone, an atypical antidepressant often used in long-term care because of its sedative properties.

This was a retrospective study of 2,428 elderly residents selected from 53 participating nursing homes in Tennessee. In order to be included in the study, the person must have been a new user of antidepressants (ie, no antidepressant use during the previous 90 days), a resident in the facility for 30 days or more, and at least 65 years of age. The exclusion criteria included users of multiple types of antidepressants, subjects for whom the primary reason for starting therapy was somatic (ie, migraine, peripheral neuropathy, pain), and patients who used nonstudy antidepressants. Of the 2,428 subjects, 665 were new users of tricyclic antidepressants, 612 used selective serotonin reuptake inhibitors, and 304 used trazodone. A group of 847 control subjects who did not use antidepressants were selected for each pair of subjects who used antidepressants. The other selection criteria for this group were identical to those of the other study participants.

Baseline data collection included admission date, age, sex, race, body mass index, ambulatory status, number of dependent activities of daily living, incontinence, cognitive impairment, use of physical restraints, previous falls, primary reason for antidepressant therapy, and the use of other drugs. These variables were factored into statistical analyses to determine their impact on study outcomes. Primary study outcomes included number of subject falls during the study period, rates of falls, and rate ratios (control subject served as a reference). The authors did not define the term fall, although falls were subclassified as "injuries" if they required medical treatment (fractures, joint dislocations, sutured lacerations, or head injuries with altered consciousness).

Higher rates of falls were found for users of an antidepressant compared with control subjects, even after statistically controlling for other variables. The rate of falls was greatest for users of tricyclic antidepressants, followed by users of selective serotonin reuptake inhibitors, and then by users of trazodone. The adjusted rate ratio of falls for users of tricyclic antidepressants was also significantly greater than the adjusted rate ratio for users of selective serotonin reuptake inhibitors or users of trazodone. There were no statistically significant differences, however, among rate ratios for injurious falls for users of each of the 3 of antidepressant types. Elevated rates of falls persisted over the course of therapy, and increased rates were found as daily doses of antidepressants were increased. Excluding potentially atypical residents did not change these findings.

The authors concurred with previous associations between use of tricyclic antidepressants and increased rates of falls. Subjects using selective serotonin reuptake inhibitors also had a higher rate of falls than control subjects, although these drugs appeared safer for use with residents with more severe cardiovascular disease. Taking trazodone resulted in a 20% increase in falls. The researchers concluded that rate of falls is increased either as a result of antidepressant use or as a result of depression or its cofactors. Thus, the authors recommended that patients receiving any type of antidepressant should be considered at increased risk for falls.

Sondra E Dunkle, EdD, PT Pocatello, ID