Oxford University Press, International Journal of Epidemiology, 3(50), p. 841-851, 2021
DOI: 10.1093/ije/dyaa271
Full text: Download
Abstract Background Despite increasing life expectancy (LE), cross-sectional data show widening inequalities in disability-free LE (DFLE) by socioeconomic status (SES) in many countries. We use longitudinal data to better understand trends in DFLE and years independent (IndLE) by SES, and how underlying transitions contribute. Methods The Cognitive Function and Ageing Studies (CFAS I and II) are large population-based studies of those aged ≥65 years in three English centres (Newcastle, Nottingham, Cambridgeshire), with baseline around 1991 (CFAS I, n = 7635) and 2011 (CFAS II, n = 7762) and 2-year follow-up. We defined disability as difficulty in activities of daily living (ADL), dependency by combining ADLs and cognition reflecting care required, and SES by area-level deprivation. Transitions between disability or dependency states and death were estimated from multistate models. Results Between 1991 and 2011, gains in DFLE at age 65 were greatest for the most advantaged men and women [men: 4.7 years, 95% confidence interval (95% CI) 3.3–6.2; women: 2.8 years, 95% CI 1.3–4.3]. Gains were due to the most advantaged women having a reduced risk of incident disability [relative risk ratio (RRR):0.7, 95% CI 0.5–0.8], whereas the most advantaged men had a greater likelihood of recovery (RRR: 1.8, 95% CI 1.0–3.2) and reduced disability-free mortality risk (RRR: 0.4, 95% CI 0.3–0.6]. Risk of death from disability decreased for least advantaged men (RRR: 0.7, 95% CI 0.6–0.9); least advantaged women showed little improvement in transitions. IndLE patterns across time were similar. Conclusions Prevention should target the most disadvantaged areas, to narrow inequalities, with clear indication from the most advantaged that reduction in poor transitions is achievable.