Pre- and postdiagnostic diet in relation to mortality among breast cancer survivors in the CPS-II Nutrition Cohort.

Purpose: Due to the limited evidence on the role of diet and cause-specific mortality among breast cancer survivors, current nutrition guidelines for this population are consistent with those for cancer prevention. We evaluated whether diets consistent with the American Cancer Society recommendations for cancer prevention were associated with risk of death in breast cancer survivors.
Methods: Participants reported information on diet and other factors at baseline in 1992-1993 and twice during follow-up. A nine-point score reflecting concordance with diet recommendations was calculated. Multivariable-adjusted relative risks (RR) and 95 % confidence intervals (CI) for diet score in relation to overall and cause-specific mortality were computed using Cox proportional hazards regression methods.
Results: Among 4,452 women diagnosed with locally and regionally staged breast cancer after baseline and until 2011, 1,204 died during follow-up through 2012 (398 from breast cancer). Prediagnostic diet score was not associated with mortality from any cause. Postdiagnostic diet score was associated with neither breast cancer-specific mortality (RR 1.44, 95 % CI 0.90-2.30 for scores 6-9 vs 0-2) nor cardiovascular disease mortality (RR 0.81, 95 % CI 0.47-1.39), but compared to a score of 0-2, a score of 6-9 was associated with a borderline lower risk of other causes of death (RR 0.78, 95 % CI 0.56-1.07, p _trend  = 0.03; per two-point increase in score RR 0.88, 95 % CI 0.79-0.99). Of diet score components, only limiting red and processed meat consumption was associated with statistically significantly lower risk of total, CVD, and other non-breast cancer mortality.
Conclusions: Diets consistent with guidelines for cancer prevention were not associated with breast cancer-specific mortality. However, their association with other causes of mortality underscores the importance of consuming a healthy diet in this population.