Joint effect of changes in physical activity and weight on incident non-alcoholic fatty liver disease

Background: Increases in physical activity (PA) and weight have opposite effects on the risk of non-alcoholic fatty liver disease (NAFLD), but their joint effect remains unknown. We examined the dose-effect of PA increase for NAFLD prevention and the amount of PA increase required to offset the deleterious effect of weight gain. Methods: We analysed 47 058 participants who were extracted from the Taiwan MJ cohort, aged 20-50 years, without NAFLD at baseline, and followed at 1-5 years between 1997 and 2016. The outcome was incident NAFLD, diagnosed by ultrasonography. PA was measured by metabolic equivalents (METs) and duration (hour/week). We used flexible parametric survival models to estimate the HRs of annual change in PA and body mass index (BMI), controlling for their interaction and baseline covariates. Results: During 138 646 person-years of follow-up, 12 836 participants (40.6% men and 20.1% women) developed incident NAFLD. The HR (95% CI) of annual PA increase of 1 MET-hour/week was 0.88 (0.85-0.92) after controlling for weight change. Moreover, 28 min/week of moderate-intensity PA could neutralise NAFLD risk elevated by annual BMI increase of 0.1 kg/m2 at the end of year 3. We also observed an extra 35% risk reduction when PA increase (1 MET-hour/week) and weight loss (0.1 kg/m2) occurred simultaneously. Conclusions: Annual PA increase of 1 MET-hour/week was associated with a 12% lower NAFLD risk. PA increase can counteract the harmful effect of weight gain and there is a synergistic effect from PA increase and weight loss.