Table 3.

Odds ratios (95% CIs) of invasive breast cancer according to quartiles of estrogen metabolites, premenopausal women, New York University Women's Health Study (377 cases, 377 controls)

CharacteristicOR (95% CI)Ptrend
Quartile 1Quartile 2Quartile 3Quartile 4
2-OHE1, pg/mL<158158-235236-333>333
    Unadjusted*1.00 (reference)1.22 (0.79-1.89)0.89 (0.57-1.39)1.37 (0.84-2.25)0.48
    Adjusted1.00 (reference)1.25 (0.78-2.02)0.93 (0.56-1.53)1.27 (0.74-2.18)0.65
16α-OHE1, pg/mL<293293-356357-415>415
    Unadjusted*1.00 (reference)1.17 (0.78-1.75)0.89 (0.60-1.34)1.31 (0.87-1.99)0.42
    Adjusted1.00 (reference)1.27 (0.80-2.02)1.14 (0.72-1.78)1.41 (0.89-2.24)0.21
2-OHE1:16α-OHE1 ratio<0.4530.453-0.6350.636-0.936>0.936
    Unadjusted*1.00 (reference)0.91 (0.59-1.38)1.07 (0.70-1.65)1.18 (0.74-1.88)0.38
    Adjusted1.00 (reference)0.83 (0.53-1.31)1.04 (0.65-1.65)1.13 (0.68-1.87)0.51

Abbreviation: OR, odds ratio.

  • *Conditional logistic regression (matching variables: age, date of blood donation, day and phase of menstrual cycle).

  • Conditional logistic regression (matching variables: age, date of blood donation, day and phase of menstrual cycle) adjusting for first-degree family history of breast cancer (negative, one affected relative ≥45 years old, one affected relative <45, or more than one affected relative), ever smoking (no, yes), BMI (log2 transformed), age at menarche (log transformed), parity/age at first full-term pregnancy (<20, 20-24, 25-29, 30+, nulliparous).