Abstract
OBJECTIVES: Since the ageing population demands a response to ensure older people remain healthy and active, we studied the dynamics of a recently proposed healthy ageing phenotype. We drew the phenotype's trajectories and tested whether their levels and rates of change are influenced by health behaviours, comorbidities and socioeconomic positions earlier in the life course. DESIGN AND OUTCOMES: The English Longitudinal Ageing Study, a prospective, nationally representative sample of people aged ≥50 years, measured a set of eight biomarkers which make up the outcome of the healthy ageing phenotype three times over nearly a decade (N2004=5009, N2008=5301, N2013=4455). A cluster of health behaviours, comorbidities and socioeconomic positions were also measured repeatedly. We assessed the phenotype's distribution non-parametrically, then fitted linear mixed models to phenotypic change and further examined time interactions with gender and socioeconomic position. We ran additional analyses to test robustness. RESULTS: Women had a wider distribution of the healthy ageing phenotype than men had. The phenotype declined annually by -0.242 (95% confidence interval [CI]: -0.352, -0.131). However, there was considerable heterogeneity in the levels and rates of phenotypic change. Women started at higher levels, then declined more steeply by -0.293 (CI: -0.403, -0.183) annually, leading to crossover in the trajectories. Smoking and physical activity assessed on the Allied Dunbar scale were strongly associated with the trajectories. CONCLUSION: Though marked by secular decline, the trajectories of the healthy ageing phenotype showed distinct socioeconomic gradients. The trajectories were also susceptible to variations in health behaviours, strengthening the case for serial interventions to attain healthy and active ageing.