Background & Aims Roles of alcohol consumption in non-alcoholic fatty liver disease are still controversial, although several cross-sectional studies have suggested the beneficial effect of light to moderate drinking on fatty liver. We analyzed the longitudinal relationship between drinking pattern and fatty liver. Methods We included 5297 Japanese individuals (3773 men and 1524 women) who underwent a baseline study in 2003 and follow-up at least once from 2004 to 2006. Generalized estimating equation was used to estimate any association between drinking pattern and fatty liver assessed by ultrasonography. Results At baseline, 1179 men (31.2%) and 235 women (15.4%) had fatty liver; 2802 men (74.2%) and 436 women (28.6%) reported alcohol consumption. At the latest follow-up, 348 of 2594 men (13.4%) and 101 of 1289 women (7.8%) had newly developed fatty liver; 285 of 1179 men (24.2%) and 70 of 235 women (29.8%) demonstrated a remission of fatty liver. In men, drinking 0.1-69.9 g/week (odds ratio, 0.79 [95% confidence interval, 0.68-0.90]), drinking 70.0-139.9 g/week (0.73 [0.63-0.84]), drinking 140.0-279.9 g/week (0.69 [0.60-0.79]), and drinking ≥280.0 g/week (0.68 [0.58-0.79]) were inversely associated with fatty liver after adjusting for obesity, exercise, and smoking. In women, drinking 0.1-69.9 g/week (0.71 [0.52-0.96]) and drinking 70.0-139.9 g/week (0.67 [0.45-0.98]) were inversely associated with fatty liver after the adjustment. Conclusions Light to moderate alcohol consumption, or even somewhat excessive amounts especially in men, was likely to protect most individuals against fatty liver over time.
- Drinking pattern
- Generalized estimating equation
- Non-alcoholic fatty liver disease
ASJC Scopus subject areas