Results
Of 10,739 participants in the estrogen-alone trial, 8,879 women were included in the analysis. We excluded 1,860 participants for either having missing information or having self-reported BSO in an age bracket different from hysterectomy.
Participants' demographics, by analysis group, are presented in Table 1 .
Women who had BSO and were on HT for the entire time from BSO to study enrollment were more likely to be younger (52% for the ≥59 y group vs 29%-34% for other groups) and to have lower parity (82% with ≥1 live births vs 87%-93% for other groups). They also had a shorter time from hysterectomy to screening compared with the other groups.
Older age, higher parity, higher BMI, higher waist-to-hip ratio, and non–African-American race/ethnicity were associated with increased odds of developing cystocele or rectocele.
Participants who retained their ovary ("no BSO") were divided into three groups by HT use and analyzed for the presence of cystocele or rectocele. Overall, 39% of the "no-BSO" group had cystocele or rectocele at screening, with 7% having grade 2 to grade 3 cystocele or rectocele. In both univariable and multivariable analyses, cystocele or rectocele rates were similar in women who retained their ovary and never used HT, used HT some of the time, or used HT for the entire time since menopause (P = 0.23). Because HT had no effects on cystocele or rectocele in women who retained their ovaries, these women were grouped together for further analysis.
Women who retained their ovaries (no BSO) had higher rates of cystocele or rectocele at screening (39%) compared with all women who had BSO: women on HT since hysterectomy (31%), partial HT users (36%), and women not on HT since hysterectomy (35%; Table 2 ). Only 238 women had BSO and reported HT use for the entire time since hysterectomy.
Table 3 summarizes ORs for cystocele or rectocele at screening for the different groups. Overall, when adjusting for HT use, women who retained their ovaries ("no BSO") were 15% more likely to have cystocele or rectocele than women who had BSO (OR, 1.15; 95% CI, 1.04-1.26). Our crude data showed that women who retained their ovaries had higher odds of developing cystocele or rectocele compared with women who had BSO and used HT for the entire time (OR, 1.42; 95% CI, 1.07-1.88) and women who had BSO and did not use HT at all (OR, 1.18; 95% CI, 1.04-1.33; Table 3 ).
A multivariable logistic model was then fit ( Table 4 ), controlling for age, race/ethnicity, parity, BMI, waist-to-hip ratio, physical activity, smoking, alcohol, asthma, emphysema, and constipation. After controlling for these variables, we found that women who retained their ovaries had higher adjusted odds of developing cystocele or rectocele only comparedwith womenwho hadBSO andnosubsequent HT (OR, 1.23; 95% CI, 1.07-1.41). All other comparisons were nonsignificant.
When looking at the difference between unadjusted and adjusted data, the largest single change in OR was observed in the comparison between Bno BSO[ and BBSO, on HT for the entire time since BSO,[ with the addition of parity (≥1 live births vs 0 live births) to the unadjusted model. The addition of this variable reduced the OR from 1.42 to 1.27 because women who had BSO and were on HT for the entire time from BSO to study enrollment were more likely to have lower parity (82% with ≥1 live births vs 87%-93% for other groups).
Neither time at risk (defined as the time when the woman was not taking estrogen since BSO or since menopause) nor total time from BSO to screening was a significant predictor of cystocele or rectocele at screening.
Because prolapse symptoms usually appear when the leading edge of the prolapse is at or outside the hymen, we performed another analysis of the data using only grade 2 or grade 3 prolapse (WHI classification) as positive outcome. Also with this definition of prolapse, older age, higher parity, higher BMI, higher waist-to-hip ratio, and non-African-American race/ethnicity were associated with increased odds of developing cystocele or rectocele. Seven percent of women who retained their ovaries versus 6% of women who had BSO and never used hormones (not significant) had grade 2 or grade 3 cystocele or rectocele. No significant differences were found between BSO/HT use groups, with an overall P value of 0.2.