AbstractBackgroundGreater socioeconomic deprivation (SED) has been shown to be associated with a higher risk of dementia. There is an urgent need to investigate whether modifiable lifestyle factors, such as physical activity (PA), can help address the SED disparity in dementia. The current study aims to examine whether PA moderates the association between SED and dementia risk.MethodData were obtained from UK Biobank, a nationwide cohort study with over half a million participants recruited during 2006‐2010. We excluded those diagnosed with dementia at baseline. SED was represented by the Townsend deprivation index (TDI) at recruitment, with a higher TDI score indicating a higher degree of deprivation. PA was assessed by an International Physical Activity Questionnaire, and participants were classified into low, moderate, and high PA groups according to their total amount of PA measured by metabolic equivalents. Dementia was defined from self‐reported verbal interviews, hospital episode records, and death register data. Cox proportional hazard models were used to examine the association of TDI and PA with dementia risk, adjusting for baseline age, sex, ethnicity, self‐reported overall health, and assessment centers.ResultThis study included 399,270 participants (209,358 [52.44%] females, and 363,116 [90.94%] Whites) aged 38‐73 years (mean = 56.32, standard deviation [SD] = 8.12), with a mean follow‐up period of 12.95 years (SD = 1.85). When categorized by TDI tertile, those in the high‐deprivation group (hazard ratio [HR] = 1.39, 95% confidence interval [CI] = 1.24‐1.56, p<0.001) had an elevated risk of dementia, relative to the participants in the low‐deprivation group. Greater PA attenuated the association between TDI and dementia risk (HR[moderate‐PA*TDI] = 0.96, 95% CI = 0.93‐0.99, p = 0.021; HR[high‐PA*TDI] = 0.97, 95% CI = 0.93‐1.00, p = 0.080). In the low PA group, the high TDI group faced 1.46 times (95% CI = 1.13‐1.89, p = 0.004) higher risk of dementia than the low TDI group, while the associations were smaller in moderate (HR = 1.31, 95% CI = 1.10‐1.57, p = 0.003) and high (HR = 1.43, 95% CI = 1.17‐1.74, p<0.001) PA groups.ConclusionGreater PA might, to some extent, help mitigate the detrimental effect of SED disparity in dementia risk, but further research is warranted to explore additional modifiable factors associated with health disparity elimination.