Alcohol and Risk of Breast Cancer

Alcohol Increasse the risk of Breast Cancer

Citation: European Journal of Cancer Supplements (2007), Vol. 5, Issue 4, p. 161.

Y. Li¹, D. Baer1, G.D. Friedman², N. Udaltsova², A.L. Klatsky²

¹Kaiser Permanente Medical Care Program, Oakland Medical Center, Oakland, USA
²Kaiser Permanente Medical Care Program, Division of Research, Oakland, USA

Background: Drinking of alcoholic beverages has consistently been linked in population studies to increased risk of female breast cancer (BrCa), but data are relatively sparse about an independent role of choice of beverage type.

Materials and Methods: We did a cohort study of 70,033 women (59% white, 26% African-American, 10% Asian-American) who supplied information about demographics and habits at health examinations in 1978–85. Through 2004 BrCa was subsequently diagnosed in 2,829 women; the role of specific beverage types was studied among women taking more than one drink per month per month (n = 37,879 with 1,509 BrCa). We used Cox proportional hazards models adjusted for age, ethnicity, body mass index, education, and smoking, both with and without control for total alcohol. In one type of analysis women whose habits indicated a preponderant beverage choice were compared to women with no clear preference with these numbers: 10,570 wine (W), 3,783 liquor (L), 2,702 beer (B), and 20,824 no preference (N). In another type of analysis we examined the independent association of frequency (per day per week) of drinking each beverage type (W, L, B, and type of wine). Finally, we looked at the role of total alcohol (<1 drink/day referent) within beverage preference strata.

Results: Controlled for total alcohol, the RR's for comparison of preference groups to non-preferrers (N) were: W = 1.06 (0.94–1.20), L = 1.02 (0.87–1.21), and B = 1.02 (0.81–1.29). Also controlled for total alcohol, the RR's for frequency (per day per week) of the major types were: W = 1.02 (0.99–1.04), L = 1.01 (0.98–1.04), B = 1.01 (0.97–1.06). With wine type subsetted into red, white, etc., no disparities in BrCa risk were seen; e.g., for both red W and white W, RR per day per week = 1.01. For total alcohol (vs <1 drink per day) the RR of BrCa was 1.1 for women reporting 1–2 drinks/day and 1.3 for women reporting 3+ drinks/day (p for trend <0.001), with similar trends for total alcohol within the W, L, B, and N strata. All results were similar in subgroups stratified by age or ethnicity.

Conclusion: These data show the relation of alcohol intake to increased BrCa risk is independent of beverage choice, indicating that ethyl alcohol is the likely culprit.