Postmenopausal hormone replacement therapy (HRT) is known to reduce the incidence of both osteoporosis and hip fractures. However, many questions remain regarding the specific characteristics of HRT, such as the optimal dosage and duration of treatment, the influence of the patient's age when therapy is initiated or discontinued, and the effectiveness of alternative regimens of HRT. Michaëlsson and colleagues conducted a case-control study to assess the risk of hip fracture in patients receiving HRT. The study focused specifically on whether dosage and duration of therapy, route of administration and addition of progestins to the regimen have any influence on the incidence of hip fracture.

The authors used hospital discharge records to identify all female residents of six Swedish counties who were born after 1914 and treated for fractures of the proximal femur any time between 1993 and 1995. A total of 1,610 women were included in the study. Exclusion criteria included incorrect diagnosis, old fracture, blindness, birth outside of Sweden, a diagnosis of severe alcohol misuse, senile dementia or psychosis, death within three months of the fracture, or fracture attributable to severe trauma or malignancy. Data on demographics, reproductive history, use of HRT, dietary habits and smoking history were collected using a comprehensive questionnaire that was completed an average of 95 days after the fracture occurred. Hormone replacement therapy regimens were categorized by dosage. Dosages of 1 mg of estradiol, 0.325 mg of conjugated estrogens, 5 μg of ethinyl estradiol or 25 μg of transdermal estradiol were considered low-dose regimens. Higher amounts were classified as high-dose regimens. Each study patient was matched with two control subjects from the same county of residence.

Complete data were obtained from 1,328 (82.5 percent) of the study patients and 3,312 (81.6 percent) of the control subjects. Use of HRT was associated with a 6 percent (range: 3 to 9 percent) reduction in risk for hip fracture for every year of HRT, regardless of type. Current users of any type of HRT had a 9 percent (range: 5 to 13 percent) reduction in risk of hip fracture. For current users, regimens that combined estrogen and progestin provided the greatest reduction in risk (8 to 24 percent) compared with estrogen therapy alone (1 to 12 percent). In addition, current users of combination therapy experienced a substantial protective effect compared with those who had never received HRT. However, the benefit of combination therapy over estrogen therapy alone was less apparent among former users. The odds ratios showed a substantial net reduction in the risk of hip fracture. Current users of estrogen alone had an odds ratio of 0.48, and former users had an odds ratio of 0.76.

Both oral and transdermal routes of administration appeared to be equally effective. Hormone replacement therapy that was started up to nine years after menopause provided the same protective effect in reduction of hip fracture as therapy started much sooner. However, this protective effect diminished when HRT was discontinued. After five years without HRT, the protective effect diminished by up to 48 percent.

The authors conclude that HRT provides substantial protection against hip fracture in women. Protection appears to be dosage-related and increases with duration of therapy. The minimally effective dosage of estrogen without progestin to prevent post-menopausal bone loss in most women is 2 mg of estradiol, 0.625 mg of conjugated estrogens or the equivalent daily. Initiating therapy several years after menopause is effective, but protection diminishes when therapy is discontinued.