Assessing Antidepressant Safety in the Elderly

Although evidence shows that selective serotonin reuptake inhibitor (SSRI) antidepressants cause less orthostasis and interfere less with psychomotor function than do tricyclic antidepressants (TCAs), a recent pharmacoepidemiologic study found them comparable in increasing elderly patients' risk for falling (Thapa et al., 1998). For patients also receiving multiple cardiovas-cular medications, however, the study did substantiate a lower risk for falls with SSRIs relative to TCAs.

The study was a retrospective analysis of the records of 180 days of treatment for 2,428 nursing home residents who were either antidepressant nonusers or had newly prescribed antidepressants (665 received TCAs or other heterocyclic antidepressants, 612 SSRIs and 304 received trazodone [Desyrel]).

In the 1,460 person-years studied, 3,524 falls were documented. The elderly institutionalized patients receiving SSRIs were found to have had 80% more falls, and those receiving a TCA had twice the number of falls, than did matched patients not receiving an antidepressant. These findings are consistent with a study published earlier that found no significant difference between TCAs and SSRIs in the associated increased rates of hip fractures subsequent to falls (Liu et al., 1998).

These studies implicate the different antidepressant classes with a similarly heightened risk of elderly patients falling. More important, they illustrate the disparities between drug responses observed in the elderly and those assumed to occur from studies with younger adults. In an editorial accompanying the Thapa et al. study, Jerry Avorn, M.D., of the Harvard Medical School, noted that conventional trials conducted against placebo in younger patients can provide sufficient information to approve an antidepressant for the market. But, he added, "It leaves the practitioner hungry for more clinically relevant data, which may be very difficult to find" (Avorn, 1998).

Drug studies that are conducted with younger adults are difficult to apply to the elderly because of differences in drug disposition and response, and the greater susceptibility of the elderly to side effects. "How might a drug's effect change when it is used in older patients, particularly frail ones?" Avorn asked. "Its risks may be higher, its efficacy lower."

The pressing need for depression treatment studies to be conducted in the elderly is evident from the high prevalence of depression in this age group that can be responsive to appropriate antidepressanttreatment (NIH, 1991), and the increased mortality associated with untreated depressive symptoms. In a study conducted by the Osteoporotic Fractures Research Group published in the October 1998 Archives of Internal Medicine, elderly women with six or more symptoms of depression had a 24% mortality rate over seven years. A matched sample with no depressive symptoms had a 7% mortality rate over the same period, and those with three to five symptoms had a 17% rate (Whooley and Browner, 1998). Adjusting for potential confounding factors such as history of myocardial infarction, stroke or diabetes, the investigators found depressive symptoms correlated with an increased risk of death from cardiovascular diseases and noncancer, noncardiovascu-lar diseases, but not with deaths from cancer.

Clinical data from depression studies in the elderly can guide effective antidepressant treatment and monitoring to reduce morbidity and mortality from both the depressive disorder and from adverse reactions to the medication. Such data are anticipated, for example, from the first clinical trial of an antidepressant in the treatment of depression in patients 75 years of age and older, now being conducted with citalopram(Drug information on citalopram) (Celexa) in 200 patients in 10 centers in the United States. In addition to multiple instruments to gauge antidepressant efficacy, the drug safety in this trial is being determined through adverse event monitoring, laboratory measurements, physical examination, and vital sign and electrocardiogram monitoring. This trial will also include magnetic resonance imaging of each patient at baseline.

Factoring Falls

Purushottam B. Thapa, M.S., M.P.H., and colleagues from the division of pharmaco-epidemiology, department of preventive medicine at Vanderbilt University School of Medicine in Nashville, Tenn., identified their study cohort from computerized records of two pharmacies consulting with 80 nursing homes. They approached the 55 facilities having the most frequent use of SSRIs. Of those, 54 agreed to participate and 53 had records of adequate quality.

Patients over 65 years of age were identified as new users of antidepressants when beginning the medication for other than somatic symptoms such as migraine, after a period of at least 90 days without antidepressant medication. Nonuser controls were randomly selected from residents meeting all other criteria for cohort eligibility. The number of falls during the study period were ascertained from incident reports and medical records.