Circulating Estrogens and Androgens Associated With Breast Cancer Risk in Premenopausal Women

Get Permission

In a study reported in The Lancet Oncology, Timothy Key, DPhil, of Oxford University, and colleagues in the Endogenous Hormones and Breast Cancer Collaborative Group analyzed data from seven prospective studies to determine associations between sex hormones and risk of breast cancer in premenopausal women.1 They found that estradiol, calculated free estradiol, estrone, androstenedione, dehydroepiandrosterone sulphate (DHEAS), and testosterone concentrations were positively associated with risk of breast cancer and that luteal phase progesterone and sex hormone–binding globulin were not.

Study Details

Individual participant data for prediagnostic sex hormone and sex hormone–binding globulin concentrations were contributed from seven prospective studies, with analyses restricted to women who were premenopausal and younger than 50 years at blood collection and to women with breast cancer diagnosed before age 50 years. Most of the women in the analysis were of white European ethnic origin. Odds ratios for breast cancer associated with hormone concentrations were calculated for cases and controls matched for age, date of blood collection, and day of cycle, with stratification by study and further adjustment for cycle phase.

Estrogens and Androgens Associated With Risk

The analyses included data for up to 767 women with breast cancer and 1,699 controls. Breast cancer risk was significantly associated with a doubling in concentrations of estradiol (OR = 1.19, 95% confidence interval [CI] = 1.06–1.35), calculated free estradiol (OR = 1.17, 95% CI = 1.03–1.33), estrone (OR = 1.27, 95% CI = 1.05–1.54), androstenedione (OR = 1.30, 95% CI = 1.10–1.55), DHEAS (OR = 1.17, 95% CI = 1.04–1.32), and testosterone (OR = 1.18, 95% CI = 1.03–1.35), but not calculated free testosterone (OR = 1.08, 95% CI = 0.97–1.21), luteal phase progesterone (OR = 1.00, 95% CI = 0.92–1.09), or sex hormone–binding globulin (OR = 1.07, 95% CI = 0.94–1.23).

Further adjustment for age at menarche, age at first full-term pregnancy, number of full-term pregnancies, and body mass index had little effect on any of these odds ratios, except that the association with doubling of calculated free testosterone with risk became significant (OR = 1.14, 95% CI = 1.01–1.28).

There was no evidence suggesting that any of the odds ratios varied by the time between blood collection and diagnosis. For estrogens and androgens, the odds ratios were larger for estrogen receptor–positive tumors than for estrogen receptor-negative tumors (eg, 1.25 vs 1.09 for estradiol, 1.26 vs 0.90 for estrone, 1.45 vs 1.11 for androstenedione, and 1.13 vs 1.03 for testosterone), but none of the differences was significant (all P > .19 for heterogeneity). For estradiol according to phase of the menstrual cycle, the odds ratios for a doubling in concentration were 1.25 (95% CI = 1.06–1.48) for follicular samples, 1.20 (0.81–1.79) for midcycle samples, and 1.13 (0.92–1.37) for luteal samples (P = .732 for heterogeneity).

Association of Sex Hormones With Risk Factors in Controls

Cross-sectional analyses in control women showed several associations of sex hormones with breast cancer risk factors. Compared with women with a body mass index of less than 22.5 kg/m2, those with a body mass index of 30 kg/m2 or more had lower mean concentrations of estradiol (by 17%), luteal phase progesterone (by 28%), and sex hormone–binding globulin (by 46%), and body mass index was positively associated with higher mean concentrations of calculated free estradiol (by 10%), estrone (by 16%), DHEAS (by 8%), testosterone (by 7%), and calculated free testosterone (by 63%), with means adjusted for age, study, and cycle phase.

Concentrations of sex hormones were lower in older women than in younger women and sex hormone–binding globulin was higher in older women. Parity was inversely associated with calculated free testosterone, but was not associated with concentrations of the other sex hormones or sex hormone–binding globulin, and none of the hormones or sex hormone–binding globulin was associated with age at menarche or family history of breast cancer.

Compared with never-smokers, current smokers of at least 15 cigarettes per day had increased concentrations of androstenedione (by 21%), DHEAS (by 12%), testosterone (by 12%), and calculated free testosterone (by 13%). Compared with women who did not consume alcohol, women with an alcohol intake of at least 20 g/d had increased concentrations of androstenedione (by 13%), DHEAS (by 16%), testosterone (by 23%), and calculated free testosterone (by 23%). Adjustment of the analyses by smoking for alcohol and of the analyses by alcohol for smoking had no marked effect on the results.

Women who had used hormonal contraceptives had lower concentrations of estradiol (by 7%), estrone (by 7%), androstenedione (by 5%), and sex hormone–binding globulin (by 4%) compared with those who had not.

The investigators concluded: “This collaborative analysis noted a positive association between sex hormones and breast cancer risk in premenopausal women. Whether or not this association is causal is not known, but plausible biological mechanisms exist that could explain such an effect…. The magnitude of the reported association was modest, but the true association could be substantially larger because of measurement error in the assessment of long-term premenopausal hormone levels.” ■

Disclosure: The study was funded by Cancer Research UK. The study authors reported no potential conflicts of interest.


1. Endogenous Hormones and Breast Cancer Collaborative Group: Sex hormones and risk of breast cancer in premenopausal women: A collaborative reanalysis of individual participant data from seven prospective studies. Lancet Oncol 14:1009-1019, 2013.

Related Articles

COMMENTARY: Evidence-Based Opportunity to Personalize Breast Cancer Risk: The Data Are Building

The worldwide data from prospective studies of the relationship between levels of endogenous sex hormones and breast cancer risk in postmenopausal women show multiple and complex relationships.1 Nine prospective studies (different from those reported here) of women not taking exogenous sex hormones ...