Successful ageing, education, and functional trajectories in later life: a longitudinal latent variable modelling analysis

Abstract

Listen

Translate article

BackgroundThe MRC Cognitive Function and Ageing Study (MRC CFAS) is a longitudinal population-based cohort of British adults aged 65 years and older. This study assessed associations between education and longitudinal successful ageing trajectories. MethodsA representative sample of individuals aged 65 years or more, from five UK sites, was eligible for inclusion. Study participants were individuals (n=1141) from the 2-year, 4-year, and 6-year MRC CFAS follow-up data collection periods. Mean age was 76·4 years (SD 6·5), and 723 (63·4%) were women. The primary outcome was an index consisting of items identified by systematic reviews of operational definitions and lay perspectives of successful ageing: self-reported psychological and social components (ie, perspective, engagement, and self-rated health) and physiological components (ie, cognitive and physical functioning). Growth mixture modelling, a person-centred longitudinal latent variable modelling procedure, was used to identify heterogeneous functional trajectories across three MRC CFAS periods (data version 9.0). Models were estimated using maximum likelihood, with missing observations assumed to be missing at random. Unadjusted and adjusted (for age, sex, and socioeconomic status [SES]) logistic regression was used to assess the association between years in full-time education (≤9, 10–11, ≥12) and functional trajectory. FindingsA three-class model provided the best fit, capturing a high functioning trajectory (HFT), moderate functioning trajectory (MFT), and low functioning trajectory (LFT). Individuals in the HFT were significantly younger, and consisted of significantly more men and of fewer individuals with low SES, than individuals in the MFT (p<0·0001) and LFT (p<0·0001). Logistic regression models adjusted for age, sex, and SES indicated that individuals in the HFT spent significantly more years in full-time education than did individuals in the MFT (odds ratio 1·68, 95% CI 1·18–2·38, p=0·004) and LFT (1·59, 1·07–2·36, p=0·02), as did unadjusted models (1·53, 1·16–2·03, p=0·003, vs 1·75, 1·30–2·36, p<0·0001). InterpretationEducation is independently associated with a high psychosocial-physiological functioning trajectory—ie, successful ageing. These results provide evidence for the long-term, beneficial effects of education in a sample of older British adults, with implications for policy, such as promotion of continued education, and for public health, such as fostering lifelong wellbeing. FundingMedical Research Council and Department of Health (grant no G9901400).