Latent class patterns of medication attitudes and their correlates among older adults in the United States

Polypharmacy and potentially inappropriate medication use are highly prevalent among older adults and contribute to adverse health outcomes. Medication attitudes, such as concerns, confidence, and willingness to change, may play a critical role in shaping deprescribing decisions, yet population-level heterogeneity in these attitudes and their correlates is poorly understood. This study aimed to identify latent classes of medication attitudes among U.S. older adults and to examine sociodemographic, health-related, medication-related, and social correlates of latent class membership. A cross-sectional analysis was conducted using nationally representative data from Round 6 of the National Health and Aging Trends Study (NHATS). The sample included 1,951 community-dwelling adults aged ≥ 65 years who completed the Medication Attitudes module. Seven medication attitude items were subject to latent class analysis. Survey-weighted multinomial logistic regression examined correlates of class membership across sociodemographic, health, medication management, and social domains. Three distinct latent classes were identified: Confident & Willing (48.8%), Low Concern & Low Engagement (29.3%), and High Concern but Conflict about Change (21.9%). Compared with the Confident & Willing class, membership in the High Concern but Conflict class was associated with lower education (associate degree or higher vs. < high school: RRR = 0.51, 95% CI 0.30–0.87), poorer self-rated health (RRR = 1.52, 95% CI 1.03–2.24), greater difficulty tracking medications (RRR = 2.02, 95% CI 1.17–3.48), and fewer social activities (RRR = 0.80 per activity, 95% CI 0.70–0.93). The Low Concern & Low Engagement class was more likely to include women (RRR = 1.71, 95% CI 1.25–2.32) and individuals with dementia (RRR = 3.89, 95% CI 1.92–7.91), despite lower medication burden. Medication attitudes among older adults cluster into distinct profiles with different implications for deprescribing. Older adults in the High Concern but Conflict about Change class appeared particularly vulnerable, with lower education, poorer health, greater difficulty tracking medications, and lower social activity. Identifying these attitudinal subgroups may help address an important gap in deprescribing research by informing more tailored, capacity-sensitive, and patient-centered strategies.