BA7 – Endogenous Estradiol and the Effects of Estrogen Therapy on Major Outcomes of WHI

This page provides study documentation for BA7. For description of the specimen results, see Specimen Results Description (open to public). Data sets of the specimen results are included in the existing WHI datasets located on the WHI Data on this site (sign in and a completed Data Distribution Agreement are required; see details on the Data site). 

Investigator Names and Contact Information

Steven R Cummings, M.D, California Pacific Medical Center, San Francisco


The investigators will study how baseline levels of the estrogen molecule, estradiol, and of sex hormone binding globulin (SHBG), a protein that binds to testosterone and estradiol, relate to treatment effects of hormone therapy on coronary heart disease (CHD), stroke, blood clots, fractures, breast cancer, dementia, and mild cognitive impairment. The team will test for interaction of baseline hormone levels with treatment effects of hormone therapy.

Postmenopausal women with extremely low endogenous estradiol (E2) levels have different risks of major diseases (breast cancer, fracture, stroke, cognitive impairment) than women with high levels. Estrogen therapy (estrogen alone (E-alone) and estrogen plus progestin (E+P)) is likely to have different effects on these conditions in those with extremely low levels than those with relatively high E2 levels. For example, estrogen therapy may substantially increase the risk of stroke in women with low E2 levels and low risk of stroke but may not increase the risk of stroke in women who have high E2 levels and relatively high stroke risk at the start. Therefore, we will test the hypothesis that the effects of estrogen therapies depend on baseline E2 levels.

We will use the efficient case-only design to test this hypothesis for the main outcomes of WHI (hormone therapy trials (WHI-HT)) and most important causes of disability in postmenopausal women: stroke, coronary heart disease (CHD), fractures, breast cancer, dementia and cognitive impairment. For the bioavailable E2 index (E2/SHBG), E2 will be assayed by a lab whose E2 results are as or more accurate than other available assays; the lab was also selected by WHI to make hormone measurements for core WHI studies. The experienced investigators include internationally recognized experts in women’s health and estrogen and all of the conditions under study. The team includes several leading WHI investigators, including leaders of WHI’s core studies of biomarkers for breast cancer (W10) and hip fracture (W9) that will provide E2 data to reduce overlap and sample use.  Our team (and senior statistician) developed and validated the case-only design and has led subgroup analyses from several large trials and multicenter studies involving hormone therapies and endogenous hormones. 
We minimize specimen use and costs by sharing breast cancer and hip fracture E2 data from other WHI studies, using 0.65 ml (instead of 1.0 ml) for the assay, using a random sample instead of all nonvertebral fractures, permitting use of serum or plasma (whichever is more plentiful), and using the case-only design.  We have worked out a feasible timeline for providing assay results condition by condition, allowing analyses to begin 3 months after specimen batches are shipped.
The results will advance scientific understanding of interactions between endogenous and exogenous estrogen. We will also determine whether a screening measurement of E2 would find women who benefit and some who have particularly great harm from estrogen therapies. These results would affect millions of women in the U.S. and beyond who still start and continue postmenopausal estrogen therapy.
Objectives and hypotheses
Our main objective is to determine whether the effects of estrogen therapy on important conditions (main aims of WHI) depends on a woman’s endogenous level of biologically available estradiol (E2).[1]
Our primary aim:  We will use the efficient case-only design to test the hypotheses that the baseline endogenous concentration of bioavailable E2 will significantly modify the effect of estrogen therapy (E-alone and E+P) on the risk of:
1. Fractures
Hip fracture
Nonvertebral fractures
Clinical vertebral fractures
2. Cardiovascular diseases
Coronary heart disease (CHD)
Venous thromboembolic events (VTE)
3. Breast Cancer
4. Dementia and mild cognitive impairment
Secondary aim:  we will estimate the absolute risk of these conditions within levels of bioavailable E2.  Note: the case-only design identifies interactions but, by itself, does not allow estimates of the absolute risk of a condition (for example, 2% annual risk of stroke) by E2 level.  These analyses require that we measure the bioavailable E2 index in a small (400 women) random sample of women in the WHI’s E-alone and E+P (WHI-HT) trials. 
Although there is a dearth of clinical evidence, we recognize that sex hormone binding globulin (SHBG) might have direct effects on tissues that are independent of E2[2]. Thus, in an exploratory aim, we will test whether levels of SHBG also modify the effects of E-alone and E+P on these conditions, independently of E2 levels.
Our proposal minimizes the use of WHI specimen by using the case-only design, using E2 data from WHI studies W9 (led by Dr. Cauley) and W10 (led by Dr. Cummings), using only a sample of cases of nonvertebral fracture, and permitting use of plasma for E2 in case serum is scarce (SHBG should not be measured in plasma).
Some of the publications related to this ancillary study are:  1033. 
For a complete, up-to-date list of WHI papers related to this ancillary study, please use the searchable Bibliography section of this website. To search for papers by study number, access the Simple Search, and enter the study number in the “Related Studies” field. 



1 -  Fink HA, Ensrud KE, Nelson DB, et al.: Disability after clinical fracture in postmenopausal women with low bone density: the fracture intervention trial (FIT). Osteoporos Int 2003; 14(1): 69-76.