Context:

Hormone replacement therapy (HRT) is widely used to prevent osteoporosis and fractures.

Objective:

To systematically review and evaluate studies of the use of estrogen and selective estrogen receptor modulators (SERMs) to prevent bone loss and fractures in healthy postmenopausal women.

Data Sources:

Studies with English language abstracts identified in MEDLINE, HealthSTAR, and Cochrane Library databases from 1994 to May 2001. Reference lists of key articles and meta-analyses were also reviewed.

Study Selection:

Only randomized controlled trials and cohort studies were included.

Data Extraction:

Studies meeting inclusion criteria were formally abstracted and quality was rated using criteria developed by the U.S. Preventive Services Task Force (USPSTF).

Data Synthesis:

For bone density outcomes, randomized controlled trials consistently indicated improved bone density with estrogen use. These findings were similar between prevention and treatment trials, opposed and unopposed regimens, oral and transdermal forms of estrogen, and types of progestins. For fracture outcomes, 5 randomized controlled trials of estrogen were identified and none met criteria to be ranked as a good-quality study. A primary prevention trial indicated a significant decrease in risk of nonvertebral fractures in 1 of 2 estrogen arms. None of the other trials indicated a significant risk reduction with estrogen, but all had important methodologic limitations. A recent meta-analysis of 22 trials of estrogen, many with unpublished data, reported an overall 27% reduction in nonvertebral fractures (relative risk [RR], 0.73; confidence interval [CI], 0.56–0.94). Six good-quality cohort studies were identified, and 3 of 4 studies reported 20% to 35% reductions in adjusted relative risks for hip fractures among ever users.

In randomized controlled trials, raloxifene increased bone density at lumbar spine, hip, and wrist sites, and tamoxifen modestly increased bone density at the hip and spine, although results were inconsistent between studies. The largest trial of raloxifene reported a protective effect for vertebral fractures (RR, 0.59; 95% CI, 0.05–0.70), but not for nonvertebral fractures. A large trial of tamoxifen indicated no statistically significant fracture benefit.

Conclusions:

The results of these studies support a benefit for estrogen in improving bone density and protecting against fractures, although the evidence for fractures is based on one arm of one randomized controlled trial and good-quality cohort studies. Trials indicate that raloxifene increases bone density and protects against vertebral fractures.