Objective: To compare the strength of the evidence provided by the epidemiological literature on the association between alcohol consumption and the risk of six cancers (oral cavity, oesophagus, colorectum, liver, larynx, breast), hypertension, cerebrovascular diseases, gastric and duodenal ulcer, liver cirrhosis and other chronic liver diseases, pancreatitis and injures and adverse effects.
Methods: A search of the epidemiological literature from 1966 to 1998 was performed by several bibliographic databases. Meta-regression models were fitted considering fixed and random models and linear and non-linear effects of alcohol intake on the risk of each condition. The effects of some characteristics of the studies including an index of their quality were considered as putative sources of heterogeneity of the estimates. Publication bias was also investigated by asymmetry of funnel plots.
Results: Of the 397 initially reviewed studies, 200 were selected for meta-analysis. Since qualitative characteristics of the studies were often significant sources of heterogeneity among them, the estimates of the pooled dose-response slopes were based only on the 123 studies with higher quality score and/or reporting adjusted estimates of relative risks. Higher alcohol-related risks were found for liver cirrhosis, neoplasms of the upper respiratory and digestive tracts, haemorrhagic stroke and injuries and adverse effects. Weaker but significant associations were found for colorectum, liver and breast cancers, essential hypertension and chronic pancreatitis. For all these conditions, low intakes, corresponding to daily consumption of two drinks or two glasses of wine (25 g/day), have shown significant risks. Ischaemic stroke and gastric and duodenal ulcer seem independent of alcohol intake. The area in which the study was performed, the study's design and the outcome variable differently affected the slopes.
Conclusions: The small number of sufficiently reliable studies, the strong indications of heterogeneity across them and the suspicion of publication bias suggest that there is a great need for well-conducted epidemiological studies performed in several countries, to examine the dose-response relationship between alcohol intake and the risk of several alcohol-related conditions, as well as the role of drinking pattern in determining the risk.