Abstract
Objectives: We aimed to investigate the effect of physical activity (PA) on non-alcoholic fatty liver disease (NAFLD) prevalence and long-term survival, particularly in some specific population such as those with different socioeconomic status (SES). Methods: Multivariate regression and interaction analyses were conducted to deal with confounders and interacting factors. Results: Active PA was associated with lower prevalence of NAFLD in both cohorts. Individuals with active-PA had better long-term survival compared to those with inactive-PA in both cohorts, and the results were only statistically significant in NAFLD defined by US fatty liver index (USFLI). We found clear evidence that the beneficial role of PA was more obvious in individuals with better SES, and the statistical significances were presented in both two hepatic steatosis index (HSI)-NAFLD cohorts from the NHANES III and NHANES 1999-2014. Results were consistent in all sensitivity analyses. Conclusion: We demonstrated the importance of PA in decrease the prevalence and mortality of NAFLD, and highlights the need for improving SES simultaneously to increase the protective effect of PA.