Abstract
BACKGROUND: Physiological age (PA) derived from clinical indicators including blood-based biomarkers and tests of physiological function can be compared with chronological age to examine disparities in health between older adults of the same age. Though education interacts with sex to lead to inequalities in healthy ageing, their combined influence on longitudinally-measured PA has not been explored. We derived PA based on longitudinally-measured clinical indicators and examined how sex and education interact to inform PA trajectories. METHODS: Three waves of clinical indicators (2004/05-2012/13) drawn from the English Longitudinal Study of Ageing (ages 50-100 years) were used to estimate PA, which was internally validated by confirming associations with incident chronic conditions, functional limitations, and memory impairment after adjustment for chronological age and sex. Joint models were used to construct PA trajectories in 8,891 ELSA participants to examine sex and educational disparities in PA. FINDINGS: Among the least educated participants, there were negligible sex differences in PA until age 60 (sex difference [men-women] age 50=-0.6 years [95% confidence interval=-2.2-0.6]; age 60=0.4 [-0.6-1.4]); at age 70, women were 1.5 years (0.7-2.2) older than men. Among the most educated participants, women were 3.8 years (1.6-6.0) younger than men at age 50, and 2.7 years (0.4-5.0) younger at age 60, with a non-significant sex difference at age 70. INTERPRETATION: Higher education provides a larger midlife buffer to physiological ageing for women than men. Policies to promote gender equity in higher education may contribute to improving women's health across a range of ageing-related outcomes.