Psychoactive Medications and Adverse Outcomes Among Older Adults Receiving Hemodialysis

Julie H. Ishida, MD, MAS; Charles E. McCulloch, PhD; Michael A. Steinman, MD; Barbara A. Grimes, PhD; Kirsten L. Johansen, MD


J Am Geriatr Soc. 2019;67(3):449-454. 

In This Article

Abstract and Introduction


Background: Guidelines recommend avoidance of several psychoactive medications such as hypnotics in older adults due to their adverse effects. Older patients on hemodialysis may be particularly vulnerable to complications related to use of these agents, but only limited data are available about the risks in this population.

Objectives: To evaluate the association between the use of psychoactive medications and time to first emergency department visit or hospitalization for altered mental status, fall, and fracture among older patients receiving hemodialysis.

Design: Observational cohort study.

Setting: National registry of patients receiving hemodialysis (US Renal Data System).

Participants: A total of 60 007 adults 65 years or older receiving hemodialysis with Medicare Part D coverage in 2011.

Measurements: The predictors were use of sedative-hypnotics and anticholinergic antidepressants (modeled as separate time-varying exposures). The outcomes were time to first emergency department visit or hospitalization for altered mental status, fall, and fracture (modeled separately).

Results: Overall, 17% and 6% used sedative-hypnotics and anticholinergic antidepressants, respectively, in 2011. In multivariable-adjusted Cox regression, anticholinergic antidepressant use was associated with a 25%, 27%, and 39% higher hazard of altered mental status, fall, and fracture, respectively, compared with no use. Use of sedative-hypnotics was not associated with adverse outcomes.

Conclusion: Anticholinergic antidepressants were associated with adverse outcomes in older hemodialysis patients, and alternative treatments should be considered. Sedative-hypnotics were not associated with the risks evaluated in this study, but further investigation of the harms of this class of agents is warranted before their recommendation as a treatment option for insomnia in this population.


Psychoactive medications (eg, sedative-hypnotics, antidepressants with anticholinergic properties) are associated with adverse events including cognitive impairment, falls, and fracture in older adults.[1–4] The Beers Criteria, a set of guidelines for identifying potentially inappropriate medication use among older adults, recommend avoidance of these categories of psychoactive medications based on their adverse side effect profiles.[5]

A substantial proportion of patients receiving hemodialysis are older than 65 years,[6] and they may represent a subset of older patients who are particularly vulnerable to medication complications due to a high medication burden, multiple comorbidities, and reduced clearance of drug metabolites.[7,8] At the same time that they may be at risk for complications, they also experience a high prevalence of the comorbidities for which psychoactive medications are indicated (eg, insomnia, depression, anxiety).[9–12] Thus, it is important to understand the potential adverse consequences of the use of these agents and whether providers should consider alternative strategies for managing these symptoms within this multimorbid population. However, studies examining this topic are limited to relatively small cohort studies in which antidepressant use was an independent predictor of falls,[13] benzodiazepines or zolpidem, a nonbenzodiazepine hypnotic, were associated with mortality, and zolpidem was associated with hip fracture.[14]

Our research aim was to describe the extent of use of sedative-hypnotics and anticholinergic antidepressants and their association with major adverse outcomes (ie, altered mental status, fall, and fracture) among older (ie, ≥65 y) hemodialysis patients in the United States. We hypothesized that these agents would be associated with the adverse outcomes of interest.