RESUMO
BACKGROUND: Combination therapy with alendronate and estrogen for 2 years increases bone mineral density at the spine and hip more than does therapy with either agent alone. Changes in bone mineral density after discontinuation of therapy have not been compared directly. OBJECTIVE: To determine the rate of bone loss when therapy with alendronate, estrogen, or both agents is discontinued. DESIGN: Double-blind, placebo-controlled discontinuation trial. SETTING: 18 U.S. centers. PATIENTS: 244 postmenopausal, hysterectomized women 44 to 77 years of age. INTERVENTION: 2 years of therapy with alendronate, 10 mg/d (n = 92); conjugated estrogen, 0.625 mg/d (n = 143); alendronate and conjugated estrogen (n = 140); or placebo (n = 50). At year 3, women were allocated into five groups: Twenty-eight women continued to take placebo and 44 women continued to take combination therapy, but 50 women taking alendronate, 81 taking conjugated estrogen, and 41 taking combination therapy were switched to placebo. MEASUREMENTS: Bone mineral density and biochemical markers of bone turnover. RESULTS: Women taking alendronate or combination therapy who were switched to placebo for year 3 of the study maintained bone mass. Bone mineral density in these women was 4.1% (CI, 2.6% to 5.7%) and 6.6% (CI, 5.0% to 8.2%) higher, respectively, at the spine (P < 0.001 for both treatment comparisons) and 3.5% (CI, 2.3% to 4.6%) and 3.0% (CI, 1.8% to 4.2%) higher, respectively, at the trochanter (P < 0.001 for both treatment comparisons) than that in women previously taking estrogen who were switched to placebo. In contrast, women who were taking estrogen and were switched to placebo during year 3 experienced a 4.5% decrease at the spine (95% CI, -5.0% to -4.0%) and a 2.4% decrease at the trochanter (CI, -2.7% to -2.1%) (P < 0.001 for both changes). Compared with women who took placebo for 3 years, women who took estrogen for 2 years and were then switched to placebo had a bone mineral density that was 2.9% higher (CI, 1.2% to 4.6%) at the spine (P < 0.05) and 2.9% higher (CI, 1.6% to 4.2%) at the trochanter (P < 0.001). Changes in biochemical markers during year 3 did not differ among the groups that discontinued active treatment. CONCLUSIONS: Accelerated bone loss is seen after withdrawal of estrogen therapy but not after withdrawal of alendronate or combination therapy. The differential effects after withdrawal of therapy should be considered in the management of postmenopausal osteoporosis.