Mediation effects of inflammation on the association of physical activity and chronic kidney disease: evidence from questionnaire and device-measured assessments

Background: The association between physical activity (PA) and chronic kidney disease (CKD) was initially explored; however, it remained unknown whether PA affected the incidence of CKD through the inflammation pathway. Moreover, objective PA measured by accelerometers was rarely considered for this association. Methods: This study was performed in a large-scale prospective cohort with two different sub-cohorts: the International Physical Activity Questionnaire (IPAQ)-measured cohort (N=314,694); and the device-measured cohort (N=79,454). Cox models were conducted to assess the association of PA with incident CKD. The mediating role of inflammation in such association was investigated by four inflammation metrics [C-reactive protein (CRP), white blood cell (WBC), a low-grade inflammation (INFLA) score, and monocyte to high-density lipoprotein cholesterol ratio (MHR)]. Results: In the questionnaire-measured cohort, compared to low PA, moderate and high PA reduced the risk of CKD by approximately 28.0% (95% CI 24.4[~]31.5%) or 37.6% (34.4[~]40.7%), respectively. Inflammation significantly mediated this association, with the mediation proportion was 4.1% (3.0[~]5.1%), 1.4% (1.1[~]1.7%), 9.8% (7.7[~]11.9%), and 1.4% (1.1[~]1.7%) for CRP, WBC, INFLA score, and MHR, respectively. Evidence from the device-measured cohort further strengthened the robustness of our findings, but the effects were somewhat attenuated, with the mediation proportion being 2.2% (1.2[~]3.2%), 0.8% (0.2[~]1.3%), 4.3% (2.5[~]6.0%), and 1.3% (0.6[~]2.1%) for CRP, WBC, INFLA score, and MHR, respectively. Conclusions: Our study reveals suggestive evidence for the association of active PA with reduced CKD risk and further demonstrates the mediating role of inflammation in such association, providing a novel perspective for the early prevention of CKD.