Objectives: To investigate the association between smoking and invasive breast cancers characterized by their estrogen receptor status in a large prospective study of mainly premenopausal women.
Methods: 112,844 women aged 25-42 years in 1989 were followed 10 years; questionnaire information on medical illnesses and risk factors was collected biennially and information on diet was collected in 1991 and 1995. During this period of follow-up (1,077,536 person-years), 1009 incident breast cancer cases were documented.
Results: In the multivariate-adjusted models, smoking status was not significantly related to overall breast cancer risk: compared with never smokers, the relative risks (RRs) were 1.18 [95% confidence interval (CI) 1.02-1.36] for past smokers and 1.12 (95% CI 0.92-1.37) for current smokers. Increasing duration of smoking before the first pregnancy was associated with a greater risk of breast cancer, although little increase was seen in the highest category: compared with never smokers, RRs were 1.42 (95% CI 1.10-1.83) for 15-19 years of smoking and 1.10 (95% CI 0.80-1.52) for >/=20 years of smoking (P for trend = 0.01). Smoking was related most strongly to the risk of estrogen receptor-positive breast cancers. For women who had smoked for >/=20 years, the RR of estrogen receptor-positive cancer was 1.37 (95% CI 1.07-1.74) and the RR of estrogen receptor-negative cancer was 1.04 (95% CI 0.71-1.53). For smoking before age 15, the RRs were 1.49 (95% CI 1.03-2.17) for estrogen receptor-positive cancer and 1.19 (95% CI 0.69-2.08) for estrogen receptor-negative cancer.
Conclusions: Our results suggest that longer duration of smoking may be related to the risk of estrogen receptor-positive breast cancer but possibly less so for estrogen receptor-negative breast cancer.