Endometrial hyperplasia was assessed by endometrial biopsies. All endometrial biopsies were read centrally by 2 primary pathologists. Participants were considered to have a diagnosis of hyperplasia if both pathologists read hyperplasia (simple hyperplasia with or without atypia or complex hyperplasia with or without atypia). If the both pathologists disagreed on the presence of hyperplasia, a third pathologist was consulted, with the final diagnosis determined by the majority opinion.

Endometrial hyperplasia at screening was an exclusion criterion and participants who had hyperplasia were not included in the analysis. Therefore this data is not available.

Endometrial hyperplasia was assessed by endometrial biopsies. All endometrial biopsies were read centrally by 2 primary pathologists. Participants were considered to have a diagnosis of hyperplasia if both pathologists read hyperplasia (simple hyperplasia with or without atypia or complex hyperplasia with or without atypia). If the both pathologists disagreed on the presence of hyperplasia, a third pathologist was consulted, with the final diagnosis determined by the majority opinion.

Efficacy evaluable (EE) analysis population for Year 1: all participants who were randomized and took at least 1 dose of test article, who had a screening endometrial biopsy with readings by at least 2 blinded central pathologists, had a biopsy during Month 12, or had hyperplasia diagnosed before Month 12 and had no major protocol violations.

BMD measurements of the anteroposterior lumbar spine were acquired by dual-energy x-ray absorptiometry (DXA), twice during screening in participants who entered the osteoporosis substudy. The second scan was to be performed on the same day as the first; however, the participant was to be removed completely from the table after the first scan and repositioned for the second scan. An average of the 2 readings was reported.

Modified intent-to-treat (MITT) population for BMD of lumber spine included all randomized participants took at least 1 dose of test article, and had a baseline and at least 1 on-therapy evaluation of BMD (scans acquired more than 60 days after the test article administration was stopped were excluded) at Year 1.

BMD measurements of the anteroposterior lumbar spine were acquired by DXA, twice at Month 12 in participants who entered the osteoporosis substudy. The second scan was to be performed on the same day as the first; however, the participant was to be removed completely from the table after the first scan and repositioned for the second scan. An average of the 2 readings was reported.

MITT population for BMD of lumber spine: all randomized participants who took at least 1 dose of test article, and had a baseline and at least 1 on-therapy evaluation of BMD (scans acquired more than 60 days after test article administration was stopped were excluded) at Year 1. Missing values imputed using last observation carried forward (LOCF).

BMD measurements of the total hip were acquired by DXA, twice during screening in participants who entered the osteoporosis substudy. The second scan was to be performed on the same day as the first; however, the participant was to be removed completely from the table after the first scan and repositioned for the second scan. An average of the 2 readings was reported.

MITT population for BMD of total hip included all randomized participants who took at least 1 dose of test article, and had a baseline and at least 1 on-therapy evaluation of BMD (scans acquired more than 60 days after the test article administration was stopped were excluded) at Year 1. Missing values were imputed using LOCF method.

BMD measurements of the total hip were acquired by DXA, twice at Month 12 in participants who entered the osteoporosis substudy. The second scan was to be performed on the same day as the first; however, the participant was to be removed completely from the table after the first scan and repositioned for the second scan. An average of the 2 readings was reported.

MITT population for BMD of total hip included all randomized participants who took at least 1 dose of test article, and had a baseline and at least 1 on-therapy evaluation of BMD (scans acquired more than 60 days after the test article administration was stopped were excluded) at Year 1. Missing values were imputed using LOCF method.

Percentage of days with breast pain in each 4-week period (for example, Week 1 to 4, 5 to 8) calculated as the number of days on which a participants reported breast pain divided by total number of days with data recorded multiplied by 100. Data was collected every day after randomization up to Year 1 and was analyzed in 4 weeks intervals. Data for screening was not analyzed since data were collected only for 7 days at screening which was not considered comparable to 4-week post-baseline data.

MITT population for breast pain: all randomized participants who took at least 1 dose of test article, and had data available at least for 5 of 7 days at screening and 20 days for at least 1 post-baseline interval. n=participants evaluable at specified time periods for each arm, respectively.

Data was collected every day after randomization up to Year 1 and was analyzed in 4 weeks intervals. Data for screening was not analyzed since data were collected only for 7 days at screening which was not considered comparable to 4-week post-baseline data.

MITT population for uterine bleeding or spotting included all randomized participants who had received at least 1 dose of test article and had at least 1 day of on-therapy bleeding data. Imputation=LOCF. n=participants evaluable for this measure at specified time periods for each arm, respectively.

Endometrial hyperplasia was assessed by endometrial biopsies. All endometrial biopsies were read centrally by 2 primary pathologists. Participants were considered to have a diagnosis of hyperplasia if both pathologists read hyperplasia (simple hyperplasia with or without atypia or complex hyperplasia with or without atypia). If the both pathologists disagreed on the presence of hyperplasia, a third pathologist was consulted, with the final diagnosis determined by the majority opinion.

EE analysis population for Year 2 included all randomized participants who took at least 1 dose of test article, participated in study extension, had a screening endometrial biopsy with readings by at least 2 blinded central pathologists, had biopsy during Month 24, or had hyperplasia diagnosed before Month 24 and had no major protocol violations.

BMD measurements of the anteroposterior lumbar spine were acquired by DXA, twice at Month 24 in participants who entered the osteoporosis substudy. The second scan was to be performed on the same day as the first; however, the participant was to be removed completely from the table after the first scan and repositioned for the second scan. An average of the 2 readings was reported.

MITT population for BMD of lumber spine: all participants who took at least 1 dose of test article, participated in study extension, and had a baseline and at least 1 on-therapy evaluation of BMD (scans acquired more than 60 days after the test article administration was stopped were excluded) at Year 2. Missing values imputed using LOCF method.

BMD measurements of the total hip were acquired by DXA, twice at Month 24 in participants who entered the osteoporosis substudy. The second scan was to be performed on the same day as the first; however, the participant was to be removed completely from the table after the first scan and repositioned for the second scan. An average of the 2 readings was reported.

MITT population for BMD of total hip: all participants who took at least 1 dose of test article, participated in study extension, and had a baseline and at least 1 on-therapy evaluation of BMD (scans acquired more than 60 days after the test article administration was stopped were excluded) at Year 2. Missing values imputed using LOCF method.