Temporal association between loneliness and polypharmacy: Analyses of the Canadian Longitudinal Study of Aging

PURPOSE: Although loneliness is cross-sectionally associated with polypharmacy, its association over time is unknown. We examined whether loneliness was associated with incident polypharmacy and greater medication use over time, and whether this association was attenuated by mood and anxiety disorders. METHODS: We analyzed baseline and 3-year follow-up data from the Canadian Longitudinal Study of Aging (CLSA) comprehensive cohort. Loneliness was defined as responding "all the time" or "occasionally" to the question, "How often did you feel lonely?" We ascertained the association between loneliness at baseline and polypharmacy (defined as ≥5 medications) and change in number of medications at follow-up from multivariable modified Poisson and linear regression models, with and without adjusting for mood and anxiety disorders. RESULTS: We included 29,968 participants (mean age 62.9 years, 50.9 % female); 3244 (10.8 %) were lonely at baseline. Among participants without polypharmacy at baseline (n = 16,075), participants who were lonely were more likely to have incident polypharmacy at follow-up than those who were not lonely (1453 [51.1 %] vs. 10,188 [41.3 %], aRR 1.16 [1.05-1.28]). Participants who were lonely increased medication use more than those who were not lonely (mean change 0.06 [standard deviation 3.0] vs. 0.04 [2.7], adjusted difference = 0.13 [0.04-0.23]). These associations were attenuated and became non-significant after adjusting for mood and anxiety disorders. CONCLUSIONS: Loneliness was associated with an increase in risk of polypharmacy and the number of medications used over time, but mood disorders and anxiety attenuated this effect. Using multimodal interventions targeting mood disorders and anxiety may help mitigate loneliness-related polypharmacy.