Objective   To access the association between alcohol consumption and mor tality in men in urban Shanghai.

  Methods   A total of 18 244 male r esidents of urban Shanghai aged 45 to 64 years were enrolled in the study during January 1, 1986 through September 30, 1989, and were actively followed up viaannual visits.Cox proponional hazards model was used to estimate the relative risks(RR).

  Results   By the end of follow-up in 2002, the cohort was followed up with invo lvement of 235 762 person-years, aver aged 12.9 years per subject.3 365 deaths including 1 381 from cancer were registered during the follow-up period.Compared with lifelong non-drinkers, the RR of overall mortality among those who drank < 15g rams, 15-29 grams ethanol per day was 0.80(95%CI): 0.72-0.89)and 0.87(95%CI: 0.78-0.97), respectively; the RR among those drinking 90 grams or more ethanol per day was 1.25(95%CI: 1.04-1.51).Among light drinkers the mortality of ischemic heart disease and chronic obstructive pulmonary disease decreased significantly.Drinking 70 grams or more ethanol per day was significantly associated with increased risks of death from esophageal cancer(RR 5.08), colorectal cancer(RR 2.57), and stroke(RR 1.57).The risk of death from hepatic cimhosis also rose significantly in those who drank 30 grams or more ethanol per day(RR 1.89).

  Conclusion   Light drinking is associated with lower overall mortality in mid dle-aged and elderly men, while heavy drinking is associated with increased risk of death from esophageal cancer, colorectal cancer, stroke, and hepatic cirrhosis.