Table 1.

Reported associations between BMI and invasive breast cancer risk among pre- and postmenopausal women

PremenopausalPostmenopausal
BMI, kg/m2CasesRRa,b (95% CI)CasesRR (95% CI)
Prospective cohort studies
 Cecchini and colleagues (3), data from 2 large RCTs with BMI measured at baseline and protocol-defined mammography<25431.00 (ref)1941.00 (ref)
25–29.9451.59 (1.05–2.42)2281.07 (0.88–1.30)
≥30381.70 (1.10–2.63)3311.14 (0.94–1.38)
P = 0.01P = 0.17
 White and colleagues (6), Multiethnic Cohort Study, self-reported baseline BMI, no control for mammography, analysis shown for nonusers of MHT<20.0630.90 (0.69–1.18)
20.0–24.93161.00 (ref)
25–29.93961.35 (1.17–1.57)
≥303291.60 (1.36–1.87)
P < 0.0001
 Harris and colleagues (7), Nurses Health Study II, a large prospective cohort using self-reported BMI assessed in 1993, no control for mammography<20.51321.00 (ref)
20.6–22.01280.98 (0.76–1.25)
22.1–23.91290.94 (0.74–1.20)
24.0–27.41350.94 (0.74–1.20)
≥27.5960.75 (0.57–0.99)
P = 0.03
 Reeves and colleagues (8), MWS—a very large U.K. cohort within a national mammography program, BMI self-reported at baseline; analyses shown among never users of MHT<22.52710.96 (0.85–1.08)8790.85 (0.80–0.91)
22.5–24.93521.00 (0.90–1.11)1,3361.00 (0.95–1.06)
25–27.42390.93 (0.82–1.05)1,2621.10 (1.04–1.16)
27.5–29.51510.99 (0.84–1.16)8781.21 (1.13–1.29)
≥301660.79 (0.68–0.92)1,2741.29 (1.22–1.36)
Per 10 kg/m20.86 (0.73–1.00)1.40 (1.31–1.49)
 Ahn and colleagues (9), NIH-AARP cohort study using self-reported baseline BMI, no control for mammography, analysis shown for nonusers of MHT15.0–18.460.64 (0.28–1.45)
18.5–22.41341.00 (ref)
22.5–24.91791.19 (0.95–1.49)
25.0–27.41971.35 (1.08–1.68)
27.5–29.91361.52 (1.29–1.94)
30.0–34.91751.55 (1.22–1.96)
35.0–39.9771.89 (1.40–2.55)
≥40442.08 (1.44–2.99)
P < 0.001
 Palmer and colleagues (10), Black Women's Health Study, a large prospective cohort using baseline self-reported BMI, no control for mammography, analyses shown for never users of MHT<251571.00 (ref)321.00 (ref)
25–291690.88 (0.71–1.10)530.80 (0.51–1.24)
30–34980.89 (0.68–1.15)420.88 (0.55–1.42)
≥35710.72 (0.54–0.96)380.94 (0.58–1.54)
 Lahmann and colleagues (11), EPIC, a large multinational European prospective cohort study, BMI measured at baseline, postmenopausal, no control for mammography, results shown among non-MHT users<21.61321.00 (ref)981.00 (ref)
21.6–23.51140.95 (0.73–1.23)1271.02 (0.78–1.33)
23.6–25.6850.78 (0.59–1.04)2061.35 (1.06–1.73)
25.7–28.7750.80 (0.59–1.09)2411.38 (1.08–1.76)
≥28.8680.82 (0.59–1.14)2391.36 (1.06–1.75)
P = 0.100P = 0.002
 Morimoto and colleagues (12), WHI observational study, BMI measured at baseline, no control for mammography, analyses shown for never users of MHT≤22.6371.00 (ref)
>22.6–24.9541.52 (0.95–2.42)
>24.9–27.4551.40 (0.87–2.32)
>27.4–31.1661.70 (1.08–2.68)
≥31.11032.52 (1.62–3.93)
P < 0.001
 van den Brandt and colleagues (17), pooled analysis of 7 prospective cohort studies, including 2 screening cohorts, BMI self-reported at baseline, no direct control for mammography<211581.0 (ref)3631.0 (ref)
21–22.92231.24 (0.97–1.57)6321.14 (0.99–1.33)
23–24.91311.03 (0.78–1.35)6991.15 (1.00–1.34)
25–26.9821.08 (0.79–1.48)5641.26 (1.09–1.47)
27–28.9470.97 (0.66–1.44)4011.43 (1.21–1.67)
29–30.9320.96 (0.60–1.52)2241.21 (1.01–1.46)
31–32.9100.55 (0.26–1.15)1401.29 (1.03–1.60)
≥33200.58 (0.34–1.00)1851.27 (1.03–1.55)
P = 0.007P = 0.001
Case–control studies
 John and colleagues (13), a multi-ethnic case–control study using self-reported and measured BMI collected after diagnosis, no control for mammography<252981.00 (ref)
25–29.91950.65 (0.49–0.85)
≥301790.60 (0.45–0.79)
P < 0.01
 Berstad and colleagues (14), Women's CARE, a large case–control study in AA and Whites, BMI self-reported 5 years before diagnosis/reference date, no control for mammography<251,3421.0 (ref)9181.0 (ref)
25–29.94720.92 (0.79–1.08)5790.95 (0.82–1.11)
30–34.91680.86 (0.68–1.09)2540.95 (0.77–1.16)
≥351150.81 (0.61–1.06)1490.96 (0.75–1.24)
P = 0.05P = 0.59
 Ogundiran and colleagues (15), a case–control study in Nigerian women, BMI measured after diagnosis<21.01531.0 (ref)1001.0 (ref)
21.0–23.91720.89 (0.64–1.24)1151.04 (0.63–1.71)
24.0–27.91700.74 (0.53–1.04)1390.88 (0.55–1.41)
≥281870.70 (0.50–0.98)1510.76 (0.48–1.21)
P = 0.027P = 0.15
 Boyd and colleagues (16), nested case–control studies conducted within screened populations using self-reported BMI≤21.79861 (ref)1591 (ref)
21.79–23.30540.69 (0.4–1.1)1641.05 (0.8–1.4)
23.30–25.02490.79 (0.5–1.3)1590.95 (0.7–1.3)
25.02–27.64420.68 (0.4–1.2)1701.02 (0.8–1.4)
>27.64510.76 (0.5–1.3)1801.17 (0.9–1.6)
P = 0.37P = 0.43

Abbreviations: RCTs, randomized-controlled trials; RR, relative risk; ref, reference (comparator) group; BMI, body mass index; MHT, menopausal hormone therapy; AA, African American.

  • aRRs were derived from multivariate proportional hazards models or conditional or unconditional logistic regression models, as reported by each study.

  • bAll P values associated with RRs are for tests of linear trend with BMI.