Background Maintaining wellbeing in later life is a key component of healthy ageing. Identifying modifiable factors associated with positive mental wellbeing in later life is thus an important priority, especially in the context of global population ageing. We aimed to investigate associations of monitored free-living and self-reported leisure time physical activity (LTPA) with positive mental wellbeing in early old age.

Methods 1976 men and women from the MRC National Survey of Health and Development whose positive mental wellbeing and physical activity had been assessed at age 60–64 years were included in analyses. Wellbeing was assessed using the Warwick-Edinburgh Mental Wellbeing Scale (WEMWBS). Continuous free-living activity was assessed by combined heart rate and movement sensors worn for 2–5 days and LTPA was self-reported. Using linear regression models we tested the associations of five parameters of physical activity (physical activity energy expenditure (PAEE, kJ/kg/day), time spent in activities of light-moderate (1.5–3 metabolic equivalents (METs)) and moderate-vigorous (>3 METs) intensity, participation in LTPA (i.e. exercise and sports) in the previous month and number of hours/week spent walking for leisure) with WEMWBS scores. Adjustments were made for sex, long-term limiting illness, socio-economic circumstances, smoking status and personality traits.

Results Participation in LTPA and more time spent walking for leisure were both associated with higher WEMWBS scores and these associations remained after adjustment for potential confounders. In fully-adjusted models: those people who walked for leisure for one hour or more per week had mean WEMWBS scores 1.58 (95% CI 0.57–2.59) higher than those that reported no walking for leisure; and those people who reported participating in LTPA at least 5 times per month had mean WEMWBS scores 1.60 (95% CI 0.61–2.59) higher than those that reported no LTPA. When these analyses were rerun using multiply imputed datasets to allow for missing data on covariates conclusions remained the same. There was no evidence of associations between any of the free-living physical activity measures (i.e. PAEE and time spent in different intensities of activity) and WEMWBS scores in sex-adjusted models.

Conclusion Our results suggest that specific types of volitional activity (namely, leisure time physical activity and walking for leisure) are more beneficial for mental wellbeing in early old age than incidental free-living physical activity. Our finding of an association between walking and positive mental wellbeing provides further justification, alongside evidence of its positive impact on physical health and function, for its promotion among older adults.