Objectives: We investigated the incidence of new non-vertebral fractures during HRT or low-dose vitamin (Vit) D3 supplementation in a 5-year prospective trial.
Methods: A total of 464 early postmenopausal women, (a subgroup of the Kuopio Osteoporosis Study, n = 13,100) were randomized to four groups: (1) HRT, a sequential combination of 2 mg estradiol valerate and 1 mg cyproterone acetate; (2) Vit D (300 IU/day and 100 IU/day during the fifth years); (3) HRT + Vit D; and (4) placebo. Lumbar (L2-4) and femoral neck bone mineral densities (BMD) were determined by dual X-ray absorptiometry (DXA) at baseline, after 2.5 and 5 years of treatment. All new symptomatic non-vertebral, radiographically defined fractures were recorded.
Results: Altogether, 368 women (79%) completed the 5 year treatment. In all, 32 women had 39 non-vertebral fractures during a mean of 4.3 year follow-up (HRT 4, Vit D 10, HRT + Vit D 8 and placebo 17). The reduction in the incidence of new non-vertebral fractures was significant in women with HRT alone (P = 0.032) when adjusted by baseline BMD and previous fractures; observed also with the intention-to-treat principle (P = 0.048). When the HRT groups were pooled, HRT showed a significantly lower incidence of new non-vertebral fractures (P = 0.042) than women receiving placebo and also after adjusting as above (P = 0.016); both in valid-case and in the intention-to-treat analysis. In the Vit D group, the fracture incidence was non-significantly decreased (P = 0.229) in comparison with the placebo group. The estimated risk of new non-vertebral fractures among women treated with HRT alone was 0.29 (95% CI, 0.10-0.90) and with Vit D 0.47 (95% CI, 0.20-1.14) and with HRT + Vit D 0.44 (95% CI, 0.17-1.15), in comparison with the placebo group (adjusted by femoral BMD and previous fractures).
Conclusions: This study is the first prospective trial confirming the beneficial effect of HRT on prevention of peripheral fractures in non-osteoporotic postmenopausal women. The effect of low-dose Vit D remains to be proved.