Skip to main content
. 2010 Nov 18;12(6):R98. doi: 10.1186/bcr2779

Table 4.

Odds ratios (ORs) and confidence intervals (CIs) of breast cancer by prediagnostic serum testosterone and SHBG concentrationa

Hormone Number of cases/Number of controls Unadjusted Adjusted


OR (95% CI) P b OR (95% CI) P b
All participants
Testosterone 0.005 0.006
 9.4-20.0 ng/dl 17/49 1.0 1.0
 20.1-26.9 ng/dl 27/40 1.9 (0.9 to 4.2) 2.1 (0.9 to 4.8)
 27.0-34.3 ng/dl 22/45 1.3 (0.6 to 2.9) 1.5 (0.6 to 3.4)
 34.4-65.5 ng/dl 32/34 2.9 (1.4 to 6.3) 3.3 (1.5 to 7.5)
Bioavailable testosterone 0.002 0.002
 3.0-8.0 ng/dl 19/46 1.0 1.0
 8.1-10.8 ng/dl 21/46 1.4 (0.6 to 3.1) 1.7 (0.7 to 4.2)
 10.9-13.8 ng/dl 24/45 1.4 (0.6 to 3.0) 1.7 (0.7 to 4.0)
 14.0-30.4 ng/dl 34/31 3.2 (1.4 to 7.1) 4.2 (1.6 to 10.9)
SHBG 0.34 0.48
 11.3-42.2 nmol/L 25/41 1.0 1.0
 42.5-57.9 nmol/L 27/40 1.0 (0.5 to 2.0) 1.3 (0.6 to 2.8)
 58.3-77.6 nmol/L 25/42 0.9 (0.5 to 1.9) 1.1 (0.5 to 2.8)
 78.1-198.0 nmol/L 21/45 0.7 (0.3 to 1.5) 0.8 (0.3 to 2.0)

aCases and controls were matched on age, date, hour, and days since last menses at blood collection. All adjusted models include terms for age at blood collection, height, BMI, age at menarche, number of term pregnancies, smoking status, and history of breast cancer in a first-degree relative. Hormone cut-points define quartiles estimated from all participants.

bP values (two-sided) were calculated by using conditional logistic regression with log-transformed hormone and SHBG concentrations entered as a linear term to test for trend by using a Wald test.