Table 2.

Results from epidemiologic studies of proposed biomarkers and breast cancer risk in postmenopausal women

Proposed biomarker in bloodType of study design, study results (±/NA) and number of analyses
Classification
Cross-sectional
Case-control
Nested case-control or case-cohort
Cohort
Pooled analysis*
+NA+NA+NA+NA+NA
Estrone2 (241, 242)2 (70, 243)1 (59)Convincing
Estradiol2 (241, 242)3 (70, 72, 243)1 (244)1 (59)Probable
Testosterone1 (242)1 (241)3 (70, 72, 243)1 (244)1 (59)Probable
Androstenedione1 (241)2 (70, 72)1 (243)1 (59)Probable
SHBG2 (241, 242)2 (70, 72)2 (243, 244)1 (59)Possible
Leptin2 (125, 126)4 (245-248)1 (249)Possible
Adiponectin4 (125, 204, 245, 250)2 (219, 251)1 (252)Possible
TNF-α1 (156)Hypothesized
Interleukin-61 (156)Hypothesized
CRP2 (156, 205)Hypothesized
Insulin1 (85)2 (86, 87)3 (88, 92, 93)1 (94)Possible
Glucose1 (88)1 (93)2 (94, 95)Possible
C-peptide2 (89, 90)3 (87, 92, 98)1 (91)2 (96, 97)Possible
Body weight, BMIReviews of the epidemiologic literature support positive associations between postmenopausal breast cancer risk and BMI and/or body weight (3, 4, 31, 33-38).Convincing
  • NOTE: +, positive association and P ≤ 0.05; −, negative association and P ≤ 0.05; NA, no association and P > 0.05.

  • * Original studies included in pooled analyses were exclusive from studies referenced elsewhere in the table. The pooled analysis by Key et al. (59) included only prospective studies in postmenopausal women: six studies on estrone, nine on estradiol, seven on SHBG, seven on testosterone, and five on androstenedione. An update on one cohort included the pooled analysis (253) and was subsequently published (254) but was not included in the table above to avoid double counting.