Prescriptions for psychotropic medications are common among older adults. A large Danish study showed that older age was associated with a higher probability of long-acting benzodiazepines use as well as use of three or more psychotropic drugs.[1] According to a Finnish study, two out of three community-dwelling older adults are prescribed at least one CNS medicine.[2] In the United States, nearly one in five community-dwelling elderly use psychotropic medications, primarily antidepressants followed by anti-anxiety agents.[3] Although psychotropic medications may help in the management of mental health conditions in older adults, the medications also carry risks for deleterious effects. For example, use of sedatives, hypnotics, antidepressants, and benzodiazepines may be associated with falls.[4] Adverse effects of anxiolytic and tricyclic medications may include psychomotor retardation, postural hypotension, and anticholinergic effects, which may cause blurred vision and cognitive impairment.[5]

Epidemiologic studies suggest that psychotropic medication use is more common in whites,[6] women,[3,7] and individuals who report poorer health.[3] In this research we ask if location, in particular degree of urban residential status, may also play a role in prescribing patterns for psychotropic medications.

In the United States, access to medical services varies by place. Geriatric and mental health services are less available in rural areas.[8–10] For example, the number of geriatricians per 10,000 older adults is lower in rural areas (from 1.48 in the most urban areas to 0.80 in the most rural). Similarly, board-certified internal medicine physicians are more plentiful in urban counties and not as accessible in less urban areas (27.29 vs. 3.85 per 10,000 older adults). The presence of family physicians varies less across the rural-urban continuum (ranging from 22.02 to 14.27 per 10,000 older adults) but still suggests less access to medical care in the rural sector.

Variation in psychotropic medication use may also reflect differences between older adults living in more and less urban areas, such as differences in medical status or sociodemographic features, or perhaps differences in the kinds of psychotropic medications prescribed. This research sought to determine (i) if differences in psychotropic medication use were evident across the more urban-less urban continuum, (ii) whether differences in psychotropic medication use persist in analyses that adjust for potential differences between more and less urban populations, and (iii) whether these differences were apparent for all classes of psychotropic drugs or only particular classes.

In 2010–2011 we enrolled a large group of community-dwelling seniors in Falls-Free PA, a research study comparing falls incidence among older adults completing Pennsylvania’s Healthy Steps for Older Adults, a falls prevention program (n=814), and a comparator group of older adults from the same senior services sites who did not complete the program (n=1015). Both groups completed baseline interviews by telephone after providing informed consent, and all were followed up to a year in monthly automated or in-person telephone interviews to track falls. Exclusion criteria included inability to provide informed consent, language other than English or Spanish, and inability to participate in telephone follow-up; otherwise all seniors attending PA Department of Aging programs were eligible.

Details of recruitment and follow-up for Falls-Free PA are reported elsewhere.[11] Briefly, the Falls Free-PA study enrolled participants in 19 counties. A total of 2459 older adults were ascertained at Healthy Steps senior center sites and agreed to give contact information to the research team. 90.5% (n=2219) signed informed consent, and 1835 subsequently completed baseline assessment and were eligible for follow-up. We excluded 74 participants who did not provide medication information or addresses that could be geocoded. A total of 1761 were included in analyses. The University of Pittsburgh Institutional Review Board approved the research.

Participants in Falls-Free PA were mostly white (88.2%) and female (79.2%), with a mean age (sd) of 75.5 (8.4) years. 38.6% of the sample reported education beyond high school, mostly 1–2 years of college or trade school. About half the sample reported solo residence. 22.8% of the sample resided in less urban areas, defined as counties with less than 250,000 residents. The mean number (sd) of prescriptions among respondents was 3.7 (3.0). A third reported five or more prescription medications.

To assess potential selection bias in the final analytic sample, we compared respondents who reported prescription drug information to the 4% of the sample who refused or did not supply information for other reasons (Table 1). Participants who declined or were unable to provide prescription drug information were older (77.7 vs. 74.9, p = .006) and less likely to have post-high school education (23.3% vs. 39.2%, p = .006). While minorities were overrepresented among participants not providing drug information, differences did not achieve significance. Level of urban residence was not associated with willingness to provide prescription drug information.

In this sample of older adults ascertained in senior centers, prescriptions for psychotropic medication prescriptions ranged from 13.3% in the most highly urbanized counties to 22.2% in less urban areas. Using the threshold of 250,000 people to separate counties, we found that 19.7% of older adults in less urban areas reported a psychotropic medication compared to 14.2% in more urban areas. These figures are in accord with findings from other studies.[3] Despite growing awareness of the potential risks of psychotropic medication use, about one in five older adults in less urban areas receive prescriptions for these classes of medication.

We found that gender, age, and medical conditions, both psychiatric conditions and total count, were associated with use of psychotropic medications. Consistent with other studies, women were twice as likely to report use of the medications.[3] Participants under age 75 were more likely to report use of psychotropic medications than people over age 75. This finding contradicts findings of Nordic studies showing that older age is associated with greater use of benzodiazepines, antidepressants, anxiolytics, antipsychotics, and hypnotics.[14, 15] A possible reason for these differences may be sources of data. The Nordic studies relied on administrative pharmacy claims, while medication information in the Falls Free-PA sample was established in an interview.

After controlling for sociodemographic and disease status measures, urban residence was still significantly associated with use of psychotropic medications. Residents of less urban counties were nearly twice as likely to report use compared to more urban counties. We did not find as strong a trend for polypharmacy. As noted earlier, the mean number of prescription medications was only marginally higher among less urban residents: 4.0 vs. 3.7.

Possible explanations for the greater use of psychotropic medications in less urban areas include less access to geriatricians, psychiatrists, and mental health care specialists generally.[10,16] In less urban settings, family physicians and general internists are more likely to address mental health among older adults than specialists.[9] One study found that medical generalists score lower than geriatricians on appropriate prescribing and proactive assessments for geriatric and psychological syndromes, which may help explain the greater rates of psychotropic medication prescriptions in these areas.[17] More generally, medical generalists are unlikely to receive geriatrics training.[18] Recognizing this problem, interventions have been developed to help lower inappropriate prescribing for older adults.[19]

Limitations of our study should be noted. The use of a senior center sample does not allow us to generalize results to older adults as a whole, nor does it provide an accurate representation of the entire older U.S. population. Importantly, the Falls Free-PA sample does not adequately represent the most rural localities. We relied on participants’ self-reports on prescription medications and co-morbidity, which might be less accurate than collecting pharmacy claims or other data from participants’ medical files. Finally, we were unable to collect information on sources of medical care, and, most critically, who prescribed psychotropic medications.

This study confirms the high prevalence of psychotropic medication use, with less urban residence as a risk factor for psychotropic medication use. Our findings highlight the need for a proactive approach that involves assessment of the appropriateness of psychotropic medications. Findings regarding psychotropic medications in this sample raise two key concerns. First, while 15.4% of respondents were taking a psychotropic medication, 2.5% were taking medications involving two or more classes of psychotropic medication. Use of multiple classes of psychotropic drugs is a concern, since the effectiveness of combining different types of psychotropic medications is questionable.[20] Second, of the 35 medications tracked in this study, 26 appear on the 2012 Beers list as potentially inappropriate medications for patients aged 65 years or older (Appendix),[13] and less urbanized residence was associated with greater use of these Beers-list medications. Our results suggest that residents of less urbanized may be at higher risk for potentially inappropriate prescribing of psychotropic medications.