Table 5. Multivariatea relative risks (RR) and 95% confidence intervals (CI) of epithelial ovarian cancer according to height and birthweight, by study population.
Nurses' Health Study (NHS)
|
Nurses' Health Study II (NHSII)
|
NHS and NHSII combined
|
|||||||
---|---|---|---|---|---|---|---|---|---|
Cases | Person- years | Multivariatea RR (95% CI) | Cases | Person-years | Multivariatea RR (95% CI) | Cases | Person-years | Multivariatea RR (95% CI) | |
Height, m | |||||||||
<1.6 | 148 | 572 360 | 1.0 (REF) | 18 | 314 588 | 1.0 (REF) | 166 | 886 961 | 1.0 (REF) |
1.6 to <1.65 | 217 | 734 892 | 1.16 (0.94–1.43) | 36 | 448 775 | 1.41 (0.80–2.49) | 253 | 1 183 667 | 1.19 (0.98–1.45) |
1.65 to <1.7 | 198 | 710 542 | 1.11 (0.89–1.37) | 35 | 452 681 | 1.39 (0.78–2.47) | 233 | 1 163 223 | 1.14 (0.93–1.39) |
1.7 to <1.75 | 135 | 394 339 | 1.39 (1.10–1.75) | 32 | 308 771 | 1.96 (1.10–3.50) | 167 | 703 111 | 1.46 (1.17–1.81) |
⩾1.75 | 37 | 116 768 | 1.27 (0.88–1.82) | 16 | 129 593 | 2.35 (1.19–4.63) | 53 | 246 361 | 1.43 (1.05–1.96) |
P for trendb | 0.01 | 0.01 | 0.001 | ||||||
P for heterogeneityc=0.22 | |||||||||
Birthweight, lbs | |||||||||
<5.5 | 13 | 34 925 | 0.99 (0.55–1.78) | 2 | 36 519 | 0.68 (0.16–2.84) | 15 | 71 444 | 0.93 (0.54–1.60) |
5.5–6.9 | 69 | 165 868 | 1.08 (0.79–1.48) | 29 | 303 063 | 1.19 (0.74–1.92) | 98 | 468 931 | 1.11 (0.85–1.44) |
7.0–8.4 | 101 | 248 159 | 1.0 (REF) | 45 | 527 664 | 1.0 (REF) | 146 | 775 823 | 1.0 (REF) |
⩾8.5 | 43 | 72 412 | 1.32 (0.92–1.90) | 12 | 149 557 | 0.83 (0.44–1.58) | 55 | 221 969 | 1.17 (0.85–1.60) |
P for trendd | 0.58 | 0.45 | 0.92 | ||||||
P for heterogeneityc=0.35 |
Multivariate analyses for height adjusted for age (continuous), parity (continuous), duration of oral contraceptive use (continuous), tubal ligation history (yes/no), and body mass index at the age of 18 years (<20, 20–20.9, 21–22.9, 23–24.9, and ⩾25 kg m−2). Multivariate analyses for birthweight adjusted for same factors except height (<1.6, 1.6 to <1.65, 1.65 to <1.7, 1.7 to <1.75, and ⩾1.75 m) instead of body mass index at the age of 18 years.
P value from multivariate model with height modelled as medians of categories.
P for heterogeneity by cohort assessed using the DerSimonian and Laird random effects model.
P value from multivariate model with birthweight categories modelled as ordinal.