Competent to care? Grandchild care and frailty trajectory among Chinese middle-aged and older adults through a person-environment fit lens

Abstract Objectives This study investigated how grandchild care intensity and community environment influenced frailty trajectories among Chinese middle-aged and older adults, examining whether effects vary by initial frailty status and are moderated by community environment. Methods Using data from the China Health and Retirement Longitudinal Study (2013, 2015, 2018) combined with 2011 community data, this study included 13,531 respondents (40,593 observations) across 442 communities. Marginal structural models with stabilized inverse probability of treatment weighting were employed to mitigate selection bias. Weighted regressions with community-clustered robust standard errors were constructed, with subgroup analyses stratified by baseline frailty status (non-frail, pre-frail, frail). Results While grandchild care showed a null overall effect in the general population, specific impacts varied significantly by baseline frailty status. Intensive caregiving significantly decelerated frailty progression for non-frail adults, whereas moderate caregiving accelerated longitudinal decline among frail adults. Caregiving showed non-significant impacts on frailty trajectory among pre-frail adults. A supportive community environment exerted protective main effects across all subgroups, with the frail group being most affected. It also magnified the protective effects of intensive caregiving on baseline frailty for the non-frail group. Discussion Using a person–environment fit perspective, this study reveals that the health impacts of grandchild care depend on the fit between individuals’ pre-existing health status and caregiving demands, and the role of community becomes more salient as competence declines. Findings highlight the critical need to prioritize vulnerable frail caregivers in health interventions and underscore community development as a key strategy for facilitating healthy aging.