The role of diet in epigenetic ageing over long follow-up periods and the possible moderating role of physical activity have remained unclear. We examined (1) whether dietary habits over follow-ups of 17–32 years are associated with level or change of epigenetic ageing over a seven-year follow-up and (2) whether physical activity moderates these associations. The prospective population-based Young Finns Study data (n = 1039) was used. Epigenetic ageing was measured in 2011 and 2018 using AgeDev Pheno , AgeDev Grim , and DunedinPACE. Food frequency questionnaires were used in 2001, 2007, 2011 and 2018 to calculate five diet indices: Mediterranean Diet Index, Findiet Index, Alternative Healthy Eating Index (AHEI) Dietscore (additionally used in 1986) and Baltic Sea Diet Index. The applied physical activity index included e.g. frequency and intensity of exercise. Covariates included cardiovascular and metabolic factors, other health behaviours, and socioeconomic factors. More favourable scores in (1) all diet indices except Dietscore were associated with decelerated AgeDev Grim cross-sectionally (β = –0.08 to –0.06, p = 0.003–0.022), (2) all diet indices averaged over follow-ups of 17–32 years were associated with slower epigenetic ageing in all three epigenetic clocks (β = –0.01 to –0.23, p = 2e–5–0.042), and (3) AHEI and Findiet index were most consistently associated with a decelerated change in AgeDev Grim and AgeDev Pheno over a seven-year follow-up. Modest interaction effects were also observed: among those with high physical activity, epigenetic ageing was approximately similar irrespective of diet healthiness, whereas among those with low physical activity, more favourable diet index scores were associated with less accelerated epigenetic ageing. Healthier eating over the follow-up was associated with decelerated epigenetic across different diet indices. In terms of biological ageing, having a healthy diet may be especially crucial for those with low levels of physical activity.
Ida et al. (Wed,) studied this question.