Objective: To determine the effect of the timing of initiation and the duration of postmenopausal estrogen therapy on bone mineral density (BMD).
Design: Cross-sectional study.
Setting: White, middle-class to upper middle-class community-dwelling women.
Participants: A total of 740 women aged 60 to 98 years who participated in a study of osteoporosis.
Measurements: Questionnaire, validated medication use, and height and weight. Bone mineral density at the ultradistal radius and midshaft radius using single-photon absorptiometry (SPA) and at the hip and lumbar spine using dual-energy x-ray absorptiometry (DEXA).
Results: Of the 740 women, 69% had used oral estrogen after menopause and 30% were current users. Five groups of estrogen use were identified: never users, past users who started at menopause, past users who started after age 60 years, current users who started after age 60 years, and current users who started at menopause. At all 4 bone sites, current users who started at menopause had the highest BMD levels, which were significantly higher than never users or past users who started at menopause (with 10 years' duration of use). These differences persisted after controlling for all major risk factors for osteoporosis. Among current users, there was no significant difference in BMD levels at any site between those who started estrogen at menopause (with 20 years of use) and those who started after age 60 years (with 9 years of use).
Conclusions: Estrogen initiated in the menopausal period and continued into late life is associated with the highest bone density. Nevertheless, estrogen begun after age 60 years and continued appears to offer nearly equal bone-conserving benefit.