Dose-response relationships between drinking and serum tests in Japanese men aged 40-59 years.

Alcohol intake per Japanese adult has been increasing year by year. To show biological effects of drinking, the dose-response relationships between alcohol use and serum indices were analyzed in 5919 Japanese men aged 40-59 years. The subjects were classified into nine groups: a nondrinking (ND, n = 1827) group and eight drinking (1D-8D) groups, by self-reported drinking habit. The 1D (the lightest drinking, n = 699), 5D (n = 942), and 8D (the heaviest drinking, n = 46) groups consumed alcohol less than 30 g per week, 25-30 g alcohol per day, and 100 g alcohol per day or more, respectively. Ten serum indices, total cholesterol, HDL cholesterol, triglyceride, total/HDL cholesterol ratio, LDL cholesterol, gamma-glutamyl transpeptidase, aspartate aminotransferase, alanine aminotransferase, glucose, and uric acid, were used. The dose-response analysis was statistically controlled for age, body mass index, smoking, and habitual exercise, and showed that drinking, even a small amount of alcohol, always had both beneficial and adverse effects on humans. However, alcohol less than 30 g per day may be tolerable for middle-aged Japanese men, because it improved serum lipids profile but did not induce apparent liver cell damage, hyperglycemia, or hyperuricemia.