J Clin Epidemiol Vol. 50, No. 7, pp. 857–863, 1997 0895-4356/ 97/ $17.00 Copyright  1997 Elsevier Science Inc. PII S0895-4356(97)00118-2 PHARMACOEPIDEMIOLOGY REPORT Potential Adverse Outcomes of Psychotropic and Narcotic Drug Use in Canadian Seniors Erika M. Ebly, 1 David B. Hogan, 1,2,* and Tak S. Fung 3 Departments of 1 Clinical Neurosciences, 2 Medicine, and 3 University Computing Services, University of Calgary, Calgary, Alberta, Canada T2N 4N1 ABSTRACT. Potential adverse outcomes (falls, impaired cognition, impaired self-care) of psychotropic (benzo- diazepines, antidepressants, antipsychotics) and narcotic medication use were examined in a large sample of Canadians 65 years of age and older. We examined rates of reported falls, Digit Symbol Substitution Test (DSST) scores, and mean number of self-care impairments for those consuming psychotropic and/or narcotic drugs as compared to non-users. Use of psychotropic drugs and/or narcotics was associated with an increased prevalence of falls, lower DSST scores, and/or number of self-care difficulties. For example, in subjects who were cognitively normal, the frequency of falls was 60% greater in benzodiazepine users and 120% greater in users of antidepres- sants as compared to non-users. Adverse outcomes increased with the number of classes of psychotropic or narcotic medications used. These effects were most pronounced in cognitively normal subjects where the preva- lence of falls increased from 13.9% in those consuming none of the medication classes to 42.6% for users of two or more classes. In subjects with mild dementia the use of these medications was not associated with any significant differences in the outcomes measured. Regression models showed that benzodiazepine, antidepressant, and narcotic use remained significant explanatory variables for potential adverse outcomes even after simulta- neously considering the effects of several other variables. Although further work is necessary, our results suggest that individuals with better cognitive function may be at particular risk for adverse effects with use of these medications. j clin epidemiol 50;7:857–863, 1997.  1997 Elsevier Science Inc. KEY WORDS. Elderly, psychotropic drugs, narcotics, falls, cognitive impairment, self-care ability INTRODUCTION and narcotics [6]. Most of these reports have looked at drug use among fallers rather than examining the prevalence of Adverse outcomes from medications are an important cause falls among users of particular drugs [26]. Drugs causing cog- of morbidity and mortality in seniors [1]. Psychotropic drugs nitive impairment (e.g., delirium) are potentially an impor- (sedatives-hypnotics, anxiolytics, antidepressants, antipsy- tant source of morbidity in the elderly [8,27–30]. Impaired chotics) and narcotics are among the most common classes self-care abilities have been found in anxiolytic, sedative, of medications associated with adverse effects in this popu- and anti-depressant users [31]. lation. In this study we used data collected as part of the Cana- High rates of psychotropic and narcotic drug use in the dian Study of Health and Aging (CSHA) to: (1) describe elderly have been described [2–5]. Because of potential side the prevalence of psychotropic and narcotic use in a large effects [6–9] their use requires the weighing of benefits and sample of Canadians 65 years and older; (2) compare the risks, coupled with careful monitoring. A number of these prevalence of falls, cognitive impairment, and impaired self- drugs have been categorized by some as being inappropriate care with the use of different classes of these drugs and with for use by the elderly at any time [10–12]. Several studies specific drugs within these classes; and (3) evaluate whether have documented that potentially inappropriate prescribing drug use was a significant explanatory variable or predictor of psychotropic medications and narcotics is a common oc- for falls, cognitive difficulties, and impairment in self-care. currence [3,10,11,13]. Falls, impaired cognition, and im- The type of data collected in the CSHA permitted evalua- paired self-care abilities are adverse outcomes associated tion of these endpoints for the group overall and stratified with the use of these medications. Falls are the most com- by degree of cognitive impairment. mon cause of injury in the elderly [14]. Falls and/or fractures have been associated with the use of benzodiazepines [15– METHODS 20], antidepressants [18,21,22], antipsychotics [18,23–25] The CSHA was a population-based study of 10,263 subjects designed to determine the prevalence of dementia and ex- * Address reprint requests to: David B. Hogan, M.D., Health Sciences Cen- amine other aspects of aging in Canadians 65 years of age tre, 3330 Hospital Drive NW, Calgary, Alberta, Canada T2N 4N1. Accepted for publication on 23 October 1996. and over [32].