Association between reported help and subsequent functional status for the same activity among older adults in China: a three-wave CHARLS study

Abstract Background Functional limitation in later life is dynamic, but studies of care receipt commonly use person-level measures that cannot determine whether help reported for a specific activity is associated with the subsequent course of that same activity. We examined the activity-specific association between reported help and subsequent transition out of severe functional limitation among older adults in China. Methods We conducted a three-wave longitudinal analysis of the China Health and Retirement Longitudinal Study using data from 2015, 2018, and 2020. Adults aged ≥ 60 years contributed one record for each of ten basic or instrumental activities rated in 2018 as difficulty requiring help or inability to perform. The exposure was the 2018 report that anyone had ever helped with the same activity, and the outcome was transition to no or mild difficulty by 2020. Participant-clustered logistic regression adjusted for activity, baseline severity, sociodemographic characteristics, functional burden, and health characteristics. A secondary analysis additionally adjusted for the four-level same-activity functional response in 2015. The primary estimand was conditional on survival, participation, and an observed 2020 functional outcome. Results Among 5,676 person–activity observations with observed outcomes, crude transition to no or mild difficulty occurred in 71.5% of activities without reported help and 45.1% of those with reported help. The primary adjusted analysis included 5,651 observations from 2,177 participants. Model-standardised probabilities were 59.5% (95% CI 55.8–63.3) without reported help and 47.8% (95% CI 46.0–49.7) with reported help, corresponding to a risk difference of − 11.7 percentage points (95% CI − 15.6 to − 7.8) and an adjusted odds ratio of 0.530 (95% CI 0.429–0.654). Among observations with 2015 functional history, prior severe limitation was more common with reported help than without it (33.7% vs 18.1%). Additional adjustment for 2015 same-activity status modestly attenuated the association (adjusted odds ratio 0.552, 95% CI 0.447–0.684). A mortality-inclusive composite analysis yielded a similar estimate (adjusted odds ratio 0.537, 95% CI 0.440–0.655). Conclusions Severe activity-specific limitation remained dynamic, but activities with reported help were less likely to transition to no or mild difficulty. Reported help may mark more established functional dependence and a less favourable subsequent functional course.