TABLE 2.
Hazard ratios (HRs) and 95% CIs for additional breast cancer events after 7.3 y of follow-up according to baseline dietary quartiles (Qs) in the subgroup of women without hot flashes
No. of participants (no. of additional breast cancer events) |
|||||
Dietary component | All | Intervention | Comparison | HR1 (95% CI) | P for trend2 |
All | 896 (179) | 446 (72) | 450 (107) | 0.7 (0.52, 0.95) | |
Vegetables–fruit | 0.01 | ||||
Q1: ≤4.9, servings/d | 249 (61) | 124 (26) | 125 (35) | 0.8 (0.48, 1.35) | |
Q2: >4.9–6.7 servings/d | 222 (37) | 108 (18) | 114 (19) | 0.99 (0.51, 1.94) | |
Q3: >6.7–8.9 servings/d | 210 (45) | 110 (18) | 100 (27) | 0.56 (0.31, 1.03) | |
Q4: >8.9 servings/d | 215 (36) | 104 (10) | 111 (26) | 0.41 (0.19, 0.86) | |
Fiber | 0.02 | ||||
Q1: ≤15.6 g/d | 227 (45) | 113 (19) | 114 (26) | 0.82 (0.45, 1.48) | |
Q2: >15.6–19.9 g/d | 226 (42) | 117 (20) | 109 (22) | 0.79 (0.42, 1.47) | |
Q3: >19.9–25.2 g/d | 216 (42) | 107 (17) | 109 (25) | 0.99 (0.52, 1.89) | |
Q4: >25.2 g/d | 227 (50) | 109 (16) | 118 (34) | 0.48 (0.26, 0.87) | |
Energy from fat | 0.06 | ||||
Q1: ≤23.8%/d | 229 (34) | 117 (11) | 112 (23) | 0.42 (0.2, 0.88) | |
Q2: >23.8–28.6%/d | 214 (51) | 106 (25) | 108 (26) | 0.88 (0.5, 1.55) | |
Q3: >28.6–33.4%/d | 228 (51) | 112 (19) | 116 (32) | 0.69 (0.38, 1.26) | |
Q4: >33.4%/d | 225 (43) | 111 (17) | 114 (26) | 0.75 (0.4, 1.43) | |
Fiber-to-fat ratio3 | 0.01 | ||||
Q1: ≤0.25 | 221 (37) | 104 (15) | 117 (22) | 0.82 (0.42, 1.63) | |
Q2: >0.25–0.36 | 237 (50) | 119 (22) | 118 (28) | 0.81 (0.45, 1.43) | |
Q3: >0.36–0.54 | 221 (52) | 123 (24) | 98 (28) | 0.69 (0.39, 1.23) | |
Q4: >0.54 | 217 (40) | 100 (11) | 117 (29) | 0.38 (0.19, 0.77) |
Derived from Cox model adjusted for stage and grade of original tumor and antiestrogen therapy.
Linear trend test for intervention effect across quartiles of baseline dietary pattern in multiple Cox model; likelihood ratio tests for group quartile interaction were not significant for any diet component.
Fiber (g/d) divided by fat (g/d).