Patterns and Predictors of Sexual Activity Among Women in the Hormone Therapy Trials of the Women's Health Initiative

Gass, Margery L.S. MD, NCMP; Cochrane, Barbara B. PhD, RN; Larson, Joseph C. MS; Manson, JoAnn E. MD, DrPH, NCMP; Barnabei, Vanessa M. MD, PhD, NCMP; Brzyski, Robert G. MD, PhD; Lane, Dorothy S. MD, MPH; LaValleur, June MD; Ockene, Judith K. PhD, MEd, MA; Mouton, Charles P. MD, MS; Barad, David H. MD, MS


Menopause. 2011;18(11):1160-1171. 

In This Article


The WHI-HT trials included 27,347 postmenopausal women aged 50 to 79 years at baseline who were enrolled between 1993 and 1998 at 40 clinical centers across the United States.[29] Women with severe vasomotor symptoms were discouraged from participating in the trial because it was thought that, if they were assigned to placebo rather than HT, they would drop out of the study. Women who had no history of hysterectomy before baseline were assigned to the EPT trial and were randomized to daily conjugated equine estrogens 0.625 mg plus 2.5 mg medroxyprogesterone acetate or placebo. Those who had undergone a previous hysterectomy were assigned to the ET trial and were randomized to either daily 0.625 mg conjugated equine estrogens or placebo. Both EPT and ET interventions were stopped earlier than planned (after a mean of 5.2 and 6.8 years, respectively) because of the identified risks with no overriding benefit. The primary outcomes and the details of the WHI study design, eligibility criteria, recruitment, and screening have been published.[29,30,31]

All participants in the HT trials completed questionnaires on demographic characteristics, health, risk factors (exposures), quality of life, and symptoms at baseline and 1 year after randomization. An 8.6% random subsample of participants, oversampled for minority participants to maintain minority representation, also completed the questionnaires during follow-up years 3 and 6. Several WHI-HT trial publications provided descriptions of these questionnaires and findings in HT trial participants.[23,26,27,29,30,31] Sexual activity items were adapted from miscellaneous sources[32] and are listed in Figure 1. Sexual activity/orientation was categorized as "never had sex," "lifetime versus mature (after age 45 y) lesbian," "heterosexual," "bisexual," or "prefer not to answer."[33]

Figure 1.

Women's Health Initiative sexual activity items.

A pelvic examination and a Papanicolaou test were performed by licensed clinicians at baseline for all HT trial participants and annually in EPT trial participants throughout the study. These clinicians were blinded to treatment assignment. As part of the pelvic examination procedure, the clinicians evaluated the condition of the external genitalia, the vagina, and the cervix. Assessment data included the following: for the external genitalia, loss of adipose tissue and thinning of hair; for the vagina, atrophy, smooth, pale, or friable with contact; for the cervix, flush with vaginal vault and friable with contact. For the purpose of this analysis, vulvovaginal atrophy was defined as present if an overall clinician impression of atrophy was recorded. The pelvic examination also included a documentation of the presence of a cystocele, rectocele, or uterine prolapse. The degree of pelvic relaxation was scored by the clinician as none, in the vagina, at the introitus, or outside the vagina.

Statistical Analysis

Data from the EPT and ET trials were combined except for analyses that required data from follow-up pelvic examinations (ie, vaginal atrophy and pelvic relaxation), which were performed only in EPT trial participants. All statistics were calculated in SAS Version 9.1 (SAS Institute, Cary, NC).

χ2 analyses were used to evaluate the baseline characteristics of respondents associated with sexual activity (nonrespondents were not included in these analyses). In addition to these analyses, a logistic model incorporated sexual activity (active/inactive) as a function of the baseline covariate of interest and was adjusted for age, body mass index (BMI), married/intimate relationship status, hysterectomy status, income, life satisfaction, vaginal dryness, and hormone use at enrollment. Baseline and year 1 covariates associated with stopping sexual activity at year 1 among those women who were sexually active at baseline were evaluated using χ2 tests. New diagnoses of cardiovascular disease, breast cancer, any cancer, arthritis for the participant, and major accident/disaster as well as the effect of partner illness, death, or separation during the preceding year were also assessed. The frequency distributions of the responses to various sexual activity questions at baseline and year 1 and in the 8.6% subsample in years 3 and 6 were calculated and stratified by partner status and age group. A linear model trend test across age groups was performed at each time point.

Logistic regression models of sexual activity at year 1 were analyzed as a function of the baseline covariate of interest in EPT trial participants (ET trial participants did not have follow-up pelvic examinations), adjusting for other covariates, including treatment assignment, sexual activity at baseline, age, smoking, and BMI. This model was additionally adjusted for treatment assignment in the WHI Dietary Modification trial component. Repeated-measures models of sexual activity as a function of HT treatment assignment were analyzed overall and by partner status for all active and placebo participants and then only for those who were adherent. Adherence was defined as taking 80% or greater of the assigned study pill regimen.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: