Methodological Biases in Estimating the Relationship Between Alcohol Consumption and Breast Cancer: The Role of Drinker Misclassification Errors in Meta‐Analytic Results

Abstract

While the great majority of studies of the alcohol-breast cancer link include misclassification errors, only misclassification of occasional drinkers was found to bias risk estimates significantly. Estimates based on error-free studies confirmed that low, hazardous and harmful levels of alcohol use each significantly increase the risk of breast cancer.