Table 4. Associations Between Treatment Group and Week 12 Estradiol Concentrations by Baseline Factor Subgroups.
| Subgroup | No. | Vaginal estrogen vs placebo, % difference (95% CI)a | P value for interactionb |
|---|---|---|---|
| All participants | 174 | 23.8 (6.9 to 43.3) | |
| Baseline estradiol, pg/mL | |||
| <3 | 59 | 16.3 (−6.9 to 45.4) | .32 |
| 3 to <4 | 52 | 36.2 (7.2 to 73.0) | |
| ≥4 | 63 | 2.3 (−17.4 to 26.7) | |
| BMI | |||
| <25 | 82 | 27.6 (2.9 to 58.1) | .78 |
| 25 to <30 | 54 | 20.3 (−8.2 to 57.6) | |
| ≥30 | 38 | 23.2 (−10.0 to 68.5) | |
| Time since menopause, y | |||
| <5 | 25 | 11.7 (−27.6 to 72.2) | .14 |
| 5 to <10 | 54 | 10.6 (−15.2 to 44.1) | |
| ≥10 | 93 | 40.3 (13.9 to 72.7) |
Abbreviation: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared).
SI conversion: to convert estradiol to picomoles per liter, multiply by 3.671.
Percent difference derived by exponentiating estimates.
Estimates and 95% CIs for treatment group effect were calculated from linear models of log-transformed week 12 estradiol concentration as a function of treatment group, the subgroup of interest, and their interaction; P values for interaction were calculated from a separate linear model with log-transformed week 12 estradiol concentration as a function of treatment group, linear trend over the subgroup of interest, and their interaction. Models were adjusted for baseline estradiol level, clinical center, age, and BMI, as appropriate. Time since menopause subgroup models was not adjusted for age due to the close relationship between time since menopause and age.