Exploring the dose-response relationship between alcohol consumption and the risk of several alcohol-related conditions: a meta-analysis

Objective. To compare the strength of the evidence provided by the epidemio logical literature on the association between alcohol consumption and the r isk of sh cancers (oral cavity, oesophagus, colorectum, liver, larynx, brea st), hypertension, cerebrovascular diseases, gastric and duodenal ulcer, li ver cirrhosis and other chronic liver diseases, pancreatitis and injures an d adverse effects. Methods. A search of the epidemiological literature from 1966 to 1998 was performed by several bibliographic databases. Meta-regres sion models were fitted considering fixed and random models and linear and non-linear effects of alcohol intake on the risk of each condition. The eff ects of some characteristics of the studies including an index of their qua lity were considered as putative sources of heterogeneity of the estimates. Publication bias was also investigated by asymmetry of funnel plots. Resul ts. Of the 397 initially reviewed studies, 200 were selected for meta-analy sis. Since qualitative characteristics of the studies were often significan t sources of heterogeneity among them, the estimates of the pooled dose-res ponse slopes were based only on the 123 studies with higher quality score a nd/or reporting adjusted estimates of relative risks. Higher alcohol-relate d risks were found for liver cirrhosis, neoplasms of the upper respiratory and digestive tracts, haemorrhagic stroke and injuries and adverse effects. Weaker bur significant associations were found for colorectum, liver and b reast cancers, essential hypertension and chronic pancreatitis. For all the se conditions, low intakes, corresponding to daily consumption of two drink s or two glasses of wine (25 g/day), have shown significant risks. Ischaemi c 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 outc ome 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 g reat need for well-conducted epidemiological studies performed in several c ountries, 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.