Table 3.
Multivariable analyses of primary and secondary outcomes according to body mass index class at randomization, overall and stratified by age at randomization
| BMI |
||||||
|---|---|---|---|---|---|---|
| Normal weight | Overweight |
Obese |
||||
| Endpoint | Reference | (s)HRa (95% CI) | P-value | (s)HRa (95% CI) | P-value | P interaction |
| Disease-free survival | ||||||
| All patients (n = 1781 patients, 706 events) | 1.00 | 1.16 (0.97 to 1.38) | .10 | 1.26 (1.03 to 1.54) | .03 | |
| .009 | ||||||
| <60 years (n = 1023 patients, 323 events) | 1.00 | 1.29 (1.00 to 1.67) | .05 | 1.83 (1.36 to 2.46) | <.001 | |
| ≥60 years (n = 758 patients, 383 events) | 1.00 | 1.04 (0.82 to 1.33) | .72 | 0.94 (0.72 to 1.23) | .63 | |
| Overall survival | ||||||
| All patients (n = 1781 patients, 484 events) | 1.00 | 1.20 (0.97 to 1.48) | .10 | 1.16 (0.91 to 1.48) | .23 | |
| .07 | ||||||
| <60 years (n = 1023 patients, 191 events) | 1.00 | 1.46 (1.05 to 2.04) | .03 | 1.62 (1.09 to 2.42) | .02 | |
| ≥60 years (n = 758 patients, 293 events) | 1.00 | 1.02 (0.77 to 1.34) | .90 | 0.93 (0.68 to 1.26) | .62 | |
| Breast cancer-specific mortalityb | ||||||
| All patients (n = 1781 patients, 247 events) | 1.00 | 1.25 (0.93 to 1.68) | .15 | 1.36 (0.97 to 1.91) | .07 | |
| .56 | ||||||
| <60 years (n = 1023 patients, 119 events) | 1.00 | 1.44 (0.96 to 2.18) | .08 | 1.32 (0.78 to 2.25) | .31 | |
| ≥60 years (n = 758 patients, 128 events) | 1.00 | 1.04 (0.67 to 1.60) | .88 | 1.28 (0.83 to 1.99) | .27 | |
| Other-cause mortalityb | ||||||
| All patients (n = 1781 patients, 237 events) | 1.00 | 1.12 (0.83 to 1.52) | .45 | 1.01 (0.71 to 1.43) | .96 | |
| .02 | ||||||
| <60 years (n = 1023 patients, 72 events) | 1.00 | 1.39 (0.79 to 2.45) | .25 | 2.01 (1.11 to 3.33) | .02 | |
| ≥60 years (n = 758 patients, 165 events) | 1.00 | 1.00 (0.70 to 1.44) | .99 | 0.75 (0.50 to 1.13) | .17 | |
The full multivariable models of the total study population are displayed in Supplementary Tables 6 to 9. Analyses were adjusted for age (≥60 years vs <60 years), history of cardiovascular disease (yes vs no), smoking history (yes vs no), tumor status (≥pT2 vs pT1), nodal status (pN positive vs pN negative), histology (lobular vs other), histologic grade (histologic grade 3 vs histologic grade 1 and 2), hormone receptor status (estrogen receptor-positive or progesterone receptor-positive vs estrogen receptor-positive and progesterone receptor-positive), and previous chemotherapy (yes vs no). Age was excluded as a confounding factor in the stratified analyses by age. BMI = body mass index; CI = confidence interval; (s)HR = (subdistribution) hazard ratio.
In the analyses of breast cancer-specific mortality and other-cause mortality, we reported sHR instead of HR.