Background: Body mass index (BMI) has been associated with breast cancer outcomes. However, few studies used clinical trial settings where treatments and outcomes are consistently evaluated and documented. There are also limited data assessing how patient/disease characteristics and treatment may alter the BMI-breast cancer association. Methods: We evaluated 15,538 breast cancer participants from four NSABP protocols. B-34 studied early-stage breast cancer patients (N = 3,311); B-30 and B-38 included node-positive breast cancer patients (N = 5,265 and 4,860); and B-31 studied node-positive and HER2-positive breast cancer patients (N = 2,102). We used Cox proportional hazards regression to calculate adjusted hazards ratios (HR) for risk of death and recurrence, and conducted separate analyses by estrogen receptor (ER) status and treatment group. Results: In B-30, increased BMI was significantly related to survival. Compared with BMI <25, HRs were 1.04 for BMI 25 to 29.9 and 1.18 for BMI ≥ 30 (P = 0.02). Separate analyses indicated the significant relationship was only in ER-positive disease (P = 0.002) and the subgroup treated with doxorubicin/cyclophosphamide (P = 0.005). There were no significant trends across BMI for the other three trials. Similar results were found for recurrence. Increased BMI was significantly related to recurrence in B-30 (P = 0.03); and the significant relationship was only in ER-positive breast cancers (P = 0.001). Recurrence was also significant among ER-positive disease in B-38 (P = 0.03). Conclusions: In our investigation, we did not find a consistent relationship between BMI at diagnosis and breast cancer recurrence or death. Impact: This work demonstrates that the heterogeneity of breast cancer between different breast cancer populations and the different therapies used to treat them may modify any association that exists between BMI and breast cancer outcome.
ASJC Scopus subject areas