Self-Reported Prescription Drug Use for Pain and for Sleep and Incident Frailty

OBJECTIVES: We aimed to estimate incident frailty risks of prescription drugs for pain and for sleep in older US adults. DESIGN: Longitudinal cohort. SETTING: Health and Retirement Study. PARTICIPANTS: Community-living respondents aged 65 years and older, excluding individuals who received recent treatment for cancer (N = 14 208). Our longitudinal analysis sample included respondents who were not frail at baseline and had at least one follow-up wave with complete information on both prescription drug use and frailty, or date of death (N = 7201). MEASUREMENTS: Prescription drug use for pain and sleep, sociodemographics, other drug and substance use, and Burden frailty model components. Multivariable drug use stratified hazard models with death as a competing risk evaluated frailty risks associated with co-use and single use of prescription drugs for pain and for sleep. RESULTS: Proportions endorsing prescription drug use were 22.1% for pain only, 6.8% for sleep only, and 7.7% for both indications. Burden frailty model prevalence was 41.0% and varied significantly by drug use. Among non-frail individuals at baseline, proportions endorsing prescription drug use were 14.9%, 5.6%, and 2.2% for the three indications. Prescription drug use was associated with increased risk of frailty (co-use adjusted subhazard ratio [sHR] = 1.95; 95% confidence interval [CI] = 1.6-2.4; pain only adjusted sHR = 1.58; CI = 1.4-1.8; sleep-only adjusted sHR = 1.35; CI = 1.1-1.6; no use = reference group). Cumulative incidence of frailty over 8 years for the four groups was 60.6%, 50.9%, 45.8%, and 34.1%. Sensitivity analyses controlling for chronic diseases associated with persistent pain resulted in minor risk reductions. CONCLUSION: Prescription pain and sleep drug use is significantly associated with increased incidence of frailty. Research to estimate effects of pain and sleep indications and of drug class–specific dosage and duration on incident frailty is indicated before advocating deprescribing based on these findings. © 2019 The American Geriatrics Society

cited By 0