BackgroundIt has been shown that diabetes is associated with insufficient physical activity among middle-aged and older adults, but the association between different physical activity levels (PAL) and diabetes incidence needs to be further explored.ObjectiveThis study aims to explore the correlation and dose–response relationship between different PAL and the diabetes incidence in middle-aged and older adults.MethodsUtilizing data from the 2018 China Health and Retirement Longitudinal Study (CHARLS), this cross-sectional analysis included 17,226 middle-aged and older adults aged 45 and above. Binary logistic regression models and restricted cubic spline (RCS) were used to explore the correlation and dose–response relationship between different PAL and the incidence of diabetes in the total middle-aged and older adults population as well as in subgroups. Sensitivity analyses were also performed to verify the robustness of the findings.ResultsIn the entire study population, compared with the lowest PAL, participants in the third and fourth quartiles PAL saw diabetes incidence significantly reduced by 16% (p = 0.005) and 33% (p