Data on whether physical activity (PA) levels are related to nonalcoholic fatty liver disease (NAFLD) when considering body mass index (BMI) are scarce. We assessed whether PA affects the development or resolution of NAFLD in conjunction with BMI changes.
Overall, 130,144 participants who underwent health screening during 2011-2016 were enrolled. According to the PA level in the Korean version of the validated International PA Questionnaire Short Form, participants were classified into the inactive, active, and health-enhancing PA (HEPA) groups.
In participants with increased BMI, the hazard ratio (HR) and 95% confidence interval after multivariable Cox hazard model for incident NAFLD was 0.97 (0.94-1.01) in the active group and 0.94 (0.89-0.99) in the HEPA group, whereas that for NAFLD resolution was 1.03 (0.92-1.16) and 1.04 (0.88-1.23) (reference: inactive group). With increased BMI, high PA affected only new incident NAFLD. PA enhancement or maintenance of sufficient PA prevented new incident NAFLD. In participants with decreased BMI, the HRs were 0.98 (0.90-1.07) and 0.88 (0.78-0.99) for incident NAFLD and 1.07 (0.98-1.17) and 1.33 (1.18-1.49) for NAFLD resolution in the active and HEPA groups, respectively. With decreased BMI, high PA reduced incident NAFLD and increased NAFLD resolution. Maintenance of sufficient PA led to a considerable resolution of NAFLD.
In this large longitudinal study, PA prevented incident NAFLD regardless of BMI changes. For NAFLD resolution, sufficient PA was essential along with BMI decrease. Maintaining sufficient PA or increasing the PA level is crucial for NAFLD prevention or resolution.