Revisiting reciprocal links between well-being and functioning in aging: A STARTS and constrained modeling approach

In aging models, well-being and functioning are two pillars that are closely intertwined. Given older adults’ susceptibility to compromised health, understanding whether well-being improvements could enhance future functioning or vice versa may be important for promoting healthy aging. However, few empirical studies have examined these reciprocal relationships over long periods using representative older adult samples and longitudinal models designed to infer causality. We used 12 years of panel data from older U.S. adults (National Health and Aging Trends Study; N = 12,427) to examine within-person reciprocal effects between well-being (emotional well-being, purpose) and functioning (self-reported health, memory). We employed variations on the Stable Trait, Autoregressive Trait, State (STARTS) model for each variable to disentangle between-person stable trait and/or occasion-specific state variances. We found very high within-person stability for well-being and functioning variables and little evidence for reciprocal effects over time. The robustness of these results and their sensitivity to different assumptions were assessed by comparing simpler models. Our findings underscore the utility of sophisticated longitudinal modeling in aging research and contribute to theoretical discussions on the causal relationship between well-being and functioning.