[Skip to Content]
[Skip to Content Landing]
January 20, 1989

Meta-analysis of Alcohol and Risk of Breast Cancer-Reply

Author Affiliations

Harvard School of Public Health Boston

Harvard School of Public Health Boston

JAMA. 1989;261(3):383. doi:10.1001/jama.1989.03420030057028

In Reply.—  Dr Rosenberg has two specific complaints about our meta-analysis. The first is that we were wrong to have made an assumption about the distribution of alcohol consumption among subjects in her study.1 Whether the presence of a slight degree of exposure misclassification merits disregarding the results of Rosenberg et al is a decision we will leave to the reader. We repeated our meta-analysis of the casecontrol data on alcohol consumption in relation to risk of breast cancer with the data of Rosenberg et al excluded and got a result virtually identical to that originally published. We would like to remind Dr Rosenberg that of the two control groups she included in her study, we included in the meta-analysis only the one that gave the smaller estimates of relative risk associated with alcohol.Dr Rosenberg's second complaint pertains to our exclusion of the study by Miller et al.