Liver enzymes, alcohol consumption and the risk of diabetes: the Suita study.

Abstract

We aimed to investigate the combined impact of liver enzymes and alcohol consumption on the diabetes risk. Data on 5972 non-diabetic participants aged 30-79years from the Suita study were analyzed. Diabetes incidence was surveyed every 2years. Current daily alcohol consumption was defined as light drinking (< 23.0g ethanol/day in men and < 11.5g in women), moderate drinking (23.0-45.9g and 11.5-22.9g), and heavy drinking (≥ 46.0g and ≥ 23.0g). The nondrinkers category included both never-drinkers and former drinkers. During the median follow-up of 13years, 597 incident diabetes cases were diagnosed. Higher levels of γ-glutamyltransferase (GGT), alanine aminotransferase (GPT), and aspartate aminotransferase (GOT) were associated with an increased diabetes risk, and current light drinkers had a lower risk of diabetes than nondrinkers. No sex differences were observed in these associations. Compared to nondrinkers having the lowest quartiles of liver enzymes, nondrinkers and current moderate/heavy drinkers having the highest quartiles had an increased risk of diabetes. However, no association was observed for current light drinkers having the highest quartiles of liver enzymes; the multivariable hazard ratios (95% CIs) in current light drinkers with the highest quartile of liver enzymes were 1.27 (0.68-2.37) for GGT, 1.05 (0.59-1.89) for GPT, and 0.76 (0.40-1.47) for GOT, respectively. High liver enzymes were associated with an increased diabetes risk. No increased diabetes risk was observed in current light drinkers, even in these who had high levels of liver enzymes.