Characterization of hot flashes reported by healthy postmenopausal women receiving raloxifene or placebo during osteoporosis prevention trials

Objective: Raloxifene, a selective estrogen receptor modulator, is estrogen-like in the skeleton and cardiovascular system and antiestrogenic in reproductive tissues. In contrast to estrogens, raloxifene is not indicated for the treatment of hot flashes. This study was designed to examine the characteristics of hot flashes among healthy postmenopausal women participating in osteoporosis prevention trials who were receiving raloxifene or placebo. Methods: Adverse event data from three randomized, double-blind trials (N = 876) comparing raloxifene 60 mg/day with placebo for 30 months were integrated and analyzed. Two of the three trials (one European, two North American) were identically designed and were open to healthy postmenopausal women ages 45 through 60 without regard to prior hysterectomy. The third trial was multinational, was open to women ages 40 through 60, and all enrollees had prior hysterectomy at baseline. Women were questioned in general terms about the occurrence of adverse events at 3-6-month intervals. Treatment-emergent adverse events pertaining to hot flashes were included in the current study. Results: At baseline, 13% of women randomly assigned to placebo and 13% assigned to raloxifene reported prevalent hot flashes. After 30 months, the cumulative incidence of hot flashes was 21% for placebo and 28% for raloxifene (P = 0.022), with the difference in incidence rate confined to the first 6 months of therapy. There was no difference between placebo and raloxifene in reported maximum severity of or early discontinuations as a result of hot flashes (13% per group for both outcomes). Among women whose hot flashes had stopped completely during the 30-month study period, the median total duration of the event prior to becoming symptom-free was 246 days for placebo and 205 days for raloxifene. Among all women reporting a hot flash, the extrapolated total duration of hot flashes was the same for women treated with either raloxifene or placebo. No subgroup-by-therapy interactions were detected. Multivariable regression analysis revealed several factors that were independently weakly predictive of hot flashes. Conclusions: Raloxifene slightly affects the incidence but not the natural history of hot flashes in healthy postmenopausal women seeking prevention therapy. (C) 2000 Elsevier Science Ireland Ltd. All rights reserved.