Meta-analysis for combining relative risks of alcohol consumption and prostate cancer

Prostate. 2000 Jan;42(1):56-66. doi: 10.1002/(sici)1097-0045(20000101)42:1<56::aid-pros7>;2-p.


Background: Prostate cancer has become the most common cancer among men in the United States, but little is known about factors associated with prostate cancer incidence.

Methods: A meta-analysis of studies published prior to July 1998 was conducted to pool relative risk (RR) estimates from the existing literature on the association between prostate cancer and alcohol consumption, in an attempt to determine whether there is an association, and if so, what its magnitude is.

Results: The overall pooled RR estimate was 1. 05 for both fixed and random effects models, based on six cohort studies and 27 case-control studies. The RR estimate varied little by study design. Among types of consumption, the highest risk was found for beer (RR = 1.27), but this was based on only eight studies that reported type of alcohol consumed. A linear dose-response was fit to the 15 studies reporting amount of alcohol consumed, finding a RR of 1.05 (95% confidence interval (CI), 0.91-1.20) for each additional drink of alcohol per day or a RR of 1.21 for 4 drinks per day. When the average drinks per day consumed in the 15 studies were used to estimate the overall risk for all 33 studies, a RR of 1.02 was found for each additional drink of alcohol per day.

Conclusions: Overall, no association between prostate cancer and alcohol consumption was seen. While some categories of consumption showed an increased risk, the studies reporting such categories appeared to be biased towards reporting a positive association among the categories.

Publication types

  • Meta-Analysis
  • Research Support, Non-U.S. Gov't

MeSH terms

  • Alcohol Drinking*
  • Case-Control Studies
  • Cohort Studies
  • Dose-Response Relationship, Drug
  • Humans
  • Male
  • Prostatic Neoplasms / etiology*
  • Risk Factors