Table 4. Relative risks (RR) of breast cancer by total activity, stratified by BMI and by oral contraceptive (OC) use, among premenopausal women in the NHS II cohort, 1989–1999.
|
BMI <25 kg m−2 |
BMI 25–29.9 kg m−2 |
BMI ⩾30 kg m−2 |
|||||||
|---|---|---|---|---|---|---|---|---|---|
| Person years (thousands) | Cases | Multivariatea RR (95% CI) | Person-years (thousands) | Cases | Multivariatea RR (95% CI) | Person-years (thousands) | Cases | Multivariatea RR (95% CI) | |
| Total Activity (MET-h/week−1) | |||||||||
| <3 | 56.7 | 52 | 1.00 | 24.1 | 20 | 1.00 | 24.4 | 22 | 1.00 |
| 3–8.9 | 123.1 | 120 | 1.06 (0.7601.47) | 50.8 | 49 | 1.12 (0.66–1.88) | 42.4 | 40 | 0.95 (0.56–1.60) |
| 9–7.9 | 137.6 | 120 | 0.95 (0.69–1.32) | 51.4 | 50 | 1.14 (0.67–1.92) | 35.4 | 27 | 0.76 (0.43–1.34) |
| 18–26.9 | 90.0 | 88 | 1.03 (0.77–1.52) | 31.0 | 33 | 1.28 (0.73–2.23) | 18.5 | 10 | 0.56 (0.26–1.18) |
| ⩾27 | 170.6 | 144 | 1.04 (0.72–1.36) | 47.9 | 36 | 0.98 (0.57–1.70) | 24.9 | 35 | 1.53 (0.89–2.63) |
| P for linear trend | 0.89 | 0.57 | 0.04 | ||||||
| OC use=never | OC use=past | OC use=current | |||||||
| Total Activity (MET-h/week−1) | |||||||||
| <3 | 16.0 | 11 | 1.00 | 78.1 | 79 | 1.00 | |||
| 3–8.9 | 32.5 | 28 | 1.23 (0.63–2.56) | 160.8 | 158 | 0.94 (0.72–1.23) | 28.9 | 25 | 1.00b |
| 9–17.9 | 33.1 | 33 | 1.48 (0.74–2.95) | 165.7 | 140 | 0.81 (0.61–1.07) | 22.6 | 21 | 1.07 (0.59–1.92) |
| 18–26.9 | 20.5 | 20 | 1.50 (0.71–3.16) | 102.1 | 100 | 0.95 (0.71–1.28) | 15.2 | 10 | 0.75 (0.35–1.57) |
| ⩾27 | 36.1 | 26 | 1.19 (0.58–2.42) | 173.4 | 175 | 1.04 (0.79–1.36) | 29.5 | 14 | 0.59 (0.30–1.16) |
| P for linear trend | 0.80 | 0.46 | 0.14 | ||||||
a
All RRs were adjusted for age, BMI, height, alcohol intake, age at menarche, parity, age at first birth, history of benign breast disease, and mother or sister with breast cancer. Analyses stratified by BMI were additionally adjusted for OC use.
b
Reference category is <9 MET-h week−1 due to small numbers (four cases and 9038 person-years) in the <3 MET-h week−1 category.