Association Between Body Mass Index and Breast Cancer Incidence in Non-Hispanic White and Black Women: Methodological Considerations From a Secondary Analysis of the Vitamin D and Omega-3 Trial Cohort

Introduction: Obesity and age are established risk factors for breast cancer, but racial differences in the association between body mass index (BMI) and breast cancer incidence remain incompletely understood. This study examined the association between BMI and breast cancer incidence among non-Hispanic White and Black women in the VITamin D and OmegA-3 TriaL (VITAL) cohort and secondarily considered methodological issues that may arise when clinical trial data are analyzed observationally. Methods: We conducted a secondary observational analysis of women enrolled in the VITAL cohort, treating trial participants as a cohort rather than comparing randomized treatment arms. After excluding men and women of races/ethnicities other than non-Hispanic White or Black, 11,377 women aged 55-90 years at baseline with no previous history of breast cancer were included. Using SAS OnDemand software (SAS Institute, Cary, NC), we examined the association between BMI and the incidence of breast cancer with Cox proportional hazards regression models. Results: The study sample of 218 women with breast cancer included 84% non-Hispanic White women (n = 183) and 16% Black women (n = 35). Age and White race were each positively related to incident breast cancer in unadjusted models (both p < 0.001). There was a positive nonlinear relationship between BMI and breast cancer (p = 0.0072, df = 1, x2 = 7.23). On more detailed examination, the increasing breast cancer risk was found mainly across lower levels of BMI and younger ages in these non-Hispanic White and Black women. Conclusion: Breast cancer incidence showed a nonlinear association with BMI and age, increasing across lower ranges but attenuating at higher BMI levels and older ages. Because this was a secondary analysis of a selected trial cohort, the unexpected attenuation should be interpreted cautiously rather than as evidence of a physiologic plateau.