For this analysis, follow-up time began in 1980 and terminated with the diagnosis of any type of cancer, death, or June 1, 2008, whichever came first. Cox proportional hazards models were used to compute hazard ratios as estimates for age-adjusted and multivariable-adjusted RR and 95% CI's. The underlying time variables for the Cox model are questionnaire year and age. Additional covariates in the model were chosen to represent possible confounders and commonly accepted breast cancer risk factors and included menopausal status, age at menarche, parity, age at first birth , body mass index, family history of breast cancer in a first-degree relative, breastfeeding, cigarette smoking, and self-report of benign breast disease . All variables except age at menarche and breastfeeding were updated from follow-up questionnaires. For postmenopausal women, terms were also included for age at menopause, type of menopause, and duration/type of hormone therapy use. We included dummy variables for missing covariate data, which comprised less than 5% of total person-time (except for missing breastfeeding which was 9.5%). Tests for trend were calculated using alcohol consumption as a continuous variable.
