Table 4.

HR (95% CI) of breast cancer for IGF-I and IGFBP-3 levels according to attained age during follow-up

Quartiles* cases/person-years and HR (95% CI)
Doubling hormone concentration§PtrendP
Q1Q2Q3Q4
IGF-I
    <5014/68519/89510/9579/736
Reference0.98 (0.46-2.07)0.50 (0.21-1.18)0.60 (0.25-1.45)0.52 (0.24-1.16)0.11
    50-5934/1,18224/1,26530/1,17528/1,127
Reference0.70 (0.39-1.24)0.85 (0.49-1.46)0.96 (0.55-1.67)0.97 (0.56-1.68)0.92
    60+40/1,87950/1,67545/1,70868/1,954
Reference1.43 (0.90-2.28)1.24 (0.77-1.99)1.61 (1.04-2.51)1.47 (0.99-2.19)0.060.06
IGFBP-3
    <5015/75216/8889/87510/654
Reference0.86 (0.41-1.82)0.52 (0.22-1.23)0.79 (0.34-1.83)0.52 (0.13-2.03)0.35
    50-5935/1,12527/1,17731/1,24821/1,119
Reference0.69 (0.39-1.22)0.81 (0.47-1.40)0.61 (0.33-1.11)0.52 (0.21-1.25)0.14
    60+39/1,82341/1,72148/1,67468/1,834
Reference1.01 (0.62-1.64)1.34 (0.83-2.18)1.62 (1.03-2.55)2.32 (1.14-4.71)0.020.02
  • * Quartiles were adjusted for variations between batches and by age and menopausal status at time of blood collection, according to procedures described in Materials and Methods.

  • Breast cancer cases and person-years.

  • Estimates from the Cox regression model adjusted for country of birth, age at menarche, parity, duration of lactation, oral contraceptive use, menopausal status at baseline, hormone replacement therapy use, physical activity, alcohol consumption, energy from diet, smoking, and level of education and stratified according to BMI category. Estimates were based on the following number of women (cases): IGF-I, 2,016 (371); IGFBP-3, 1,969 (360).

  • § Estimates from the model including the pseudo-continuous variable log 2–transformed.

  • Test for log-linear trend using the pseudo-continuous variable log 2–transformed.

  • Test for homogeneity in the HRs for hormone concentration (pseudo-continuous log 2–transformed) in the three age groups.