Abstract
BACKGROUND & AIMS: Moderate alcohol intake is associated with reduced prevalence or incidence of fatty liver. However, whether or not the association is independent of dietary patterns remains unclear. We aimed to evaluate the cross-sectional association of alcohol intake with fatty liver after accounting for dietary patterns and obesity. METHODS: We assessed 4579 adults aged 30-79 years who participated in routine clinical examinations in St. Luke's International Hospital, Japan (January to March, 2015). We assessed their habitual diet using diet-history questionnaire, estimated alcohol intake, and derived dietary pattern variables using factor analysis. Fatty liver was ascertained using ultrasonography. Linear and U-shaped associations of alcohol intake with fatty liver were evaluated using Poisson regression, and a post hoc analysis was conducted after detecting potential outliers for alcohol intake and excluding them using sex-specific statistics (median plus 2 × interquartile range). RESULTS: Fatty liver was ascertained in 1120 participants (24.5%). Whereas no significant association of alcohol intake with fatty liver was observed when potential outliers of alcohol intake were included (p = 0.25), a significant U-shaped association was observed after excluding the outliers with and without adjustment for dietary patterns (p = 0.003 and 0.02, respectively). The lowest prevalence was estimated when alcohol consumption was approximately 7% of energy, with a prevalence ratio of 0.72 (95% confidence interval = 0.59-0.86) compared to non-drinkers. The association became imprecise and attenuated toward the null after further adjustment for body mass index (p = 0.06). CONCLUSIONS: Alcohol intake showed a U-shaped association with fatty liver prevalence. This association was independent of underlying dietary patterns, while it was sensitive to excessive alcohol intake and obesity status, providing clinical implications for the prevention of fatty liver.