Dietary Fats and All-cause and Breast Cancer–specific Mortality among Women with Breast Cancer: The Western New York Exposures and Breast Cancer Study

Results of studies of prediagnostic dietary fat intake in association with mortality from breast cancer have been inconclusive. In a systematic review of diet and breast cancer survival, associations of prediagnostic intake of total and saturated fat with all-cause mortality were classified as “limited-suggestive” (1). While dietary fat subtypes [saturated (SFA), polyunsaturated (PUFA), and monounsaturated (MUFA) fatty acids] may have different biological effects (2, 3), there are few studies examining fat subtypes with breast cancer mortality (4–6). We report here on the association of prediagnostic, self-reported dietary intake of total fat and fat subtypes with breast cancer-specific and all-cause mortality among women with breast cancer in a population-based study.

Included in this prospective analysis were women with primary, incident, pathologically confirmed breast cancer and complete dietary data, cases from the Western New York Exposures and Breast Cancer (WEB) study, a population-based case–control study (7). Participants with stage 4 disease (n = 37) were excluded. Follow-up for breast cancer–specific and all-cause mortality was through the National Death Index until December 31, 2018. Survival time was calculated as time from date of diagnosis until date of death or December 31, 2018, whichever came first. This study was conducted in accordance with the ethical principles that have their origin in the Declaration of Helsinki. Written informed consent was obtained from all participants; the study was approved by the Institutional Review Boards of the University at Buffalo, State University of New York, and participating hospitals.

Usual prediagnostic intake of total fat and fat subtypes were estimated from a food frequency questionnaire completed at baseline for 12–24 months prior to diagnosis (7). Dietary fat intake measures were adjusted for total energy, calculated as percent of total calories. HRs and 95% confidence intervals (CI) were estimated using Cox proportional hazards models for tertiles of intake of percent of kilocalories from total fat and fat subtypes (SFA, MUFA, and PUFA) in relation to all-cause and breast cancer–specific mortality. Covariates, selected a priori based on existing research, were included in the final model if addition changed HRs by 10% or more for a model including age and kilocalories. Final models were adjusted for age, tumor stage, ER status, education, body mass index (BMI), race/ethnicity, menopausal status, and smoking status (current, former, never). Tumor grade was not included in this analysis. Additional analyses included examination of interaction by menopausal status and ER status and models stratified by tumor stage (stage I, II, and III).

The deidentified data for these analyses can be shared on the basis of reasonable requests to the corresponding author upon approval by the WEB study research committee.

Median follow-up time was 18.75 years. Among 793 women with invasive breast cancer, 327 (41.2%) had died. Characteristics associated with mortality in this study population have been reported previously (8). Comparisons of participant characteristics by fat intake among those alive and deceased at the end of the follow-up are shown in Table 1. Among the deceased, those with higher compared with lower fat intake had had more births. Among those still alive, age, age at menopause, years of education, weight, and BMI were different by fat intake without a consistent direction of association.

In adjusted models, total dietary fat and fat subtypes were not associated with survival (Table 2). There was no evidence of interaction by menopausal status or ER status. Furthermore, there were no associations between risk of mortality and dietary fat intake within strata of stage.

Examination of dietary fat subtypes in additional to total dietary fat intakes potentially could explain inconsistent results regarding associations of dietary fat with breast cancer mortality. Fatty acids are hypothesized to influence carcinogenesis because of effects on circulating estrogen, oxidative stress, and inflammation (2, 3); differences in effects by fat subtypes are less clear (3). High- and low-fat diets may differ in the proportion of fat subtypes as well as in the dietary source of the fat. In a secondary data analysis of the Women's Health Initiative dietary intervention trial, participants following the low-fat eating pattern had lower mortality after breast cancer; it was not clear, however, whether the effect resulted from the fat intake or to other, associated changes in diet such as changed consumption of fruits and vegetables or animal products (9).

We found no association of prediagnostic intake of total fat or SFA, with either all-cause or breast cancer–specific mortality, consistent with some, but not all, studies of prediagnostic total dietary fat (1, 5). We found no associations of either MUFA or PUFA intake with risk of all-cause or breast cancer–specific mortality, consistent with most recent studies (6).

Strengths of this study include the study sample of population-based, incident, histologically confirmed breast cancer cases, the length of follow-up, and the use of validated tools for diet assessment. Limitations include the potential for residual confounding which may persist despite efforts to control it. Our analysis was limited to a report of usual prediagnostic intake, which likely includes nondifferential misclassification. Furthermore, the measured timepoint might not be the critical window for exposure relative to survival. In addition, we could not account for change, if any, in diet following diagnosis. Finally, other dietary components such as the source of dietary fat and total energy intake may also impact survival; we adjusted for energy intake but not other factors.

In our cohort, intake of dietary fat and its subtypes were not associated with mortality among women with breast cancer.

D.S. Meyer reports grants from Army Medical Research and Material Command, NCI, and National Institute on Alcohol Abuse and Alcoholism during the conduct of the study. J.L. Freudenheim reports grants from NIH and DOD CDMRP during the conduct of the study. No disclosures were reported by the other authors.

D.S. Meyer: Formal analysis, writing–original draft. A.E. Millen: Writing–review and editing. J. Nie: Data curation, formal analysis, writing–review and editing. M. Trevisan: Writing–review and editing. J.L. Freudenheim: Conceptualization, supervision, funding acquisition, methodology, writing–review and editing.

The work was supported in part by grant DAMD17-96-1-6202US from the U.S. Army Medical Research and Materiel Command awarded to J.L. Freudenheim and M. Trevisan, grant R01CA92040 from the NCI awarded to J.L. Freudenheim and M. Trevisan, and grant P50 AA09802 from the National Institute on Alcohol Abuse and Alcoholism awarded to J.L. Freudenheim and M. Trevisan.

The publication costs of this article were defrayed in part by the payment of publication fees. Therefore, and solely to indicate this fact, this article is hereby marked “advertisement” in accordance with 18 USC section 1734.