The Danish osteoporosis prevention study (DOPS): project design and inclusion of 2000 normal perimenopausal women

Abstract

Objective: In 1990 we initiated a 20 year, partly randomised study (Danish Osteoporosis Prevention Study, DOPS) in order to (a) evaluate clinical, biochemical and osteodensitometric variables as predictors of low bone mass and future osteoporotic fractures, and (b) test the hypothesis, that hormone replacement therapy (HRT) initiated shortly after menopause reduces the risk of later osteoporotic fractures. This report describes study design and baseline characteristics of the DOPS-cohort. Methods: The study design is pragmatic, attempting to mimic the normal clinical situation. Several HRT alternatives are available according to clinical need. It was considered futile, impractical and unethical to use placebo for 20 years. Instead the study focus on hard endpoints (fractures) confirmed by independent persons (peripheral fractures) or by methods which allow investigator blinding (spinal X-rays). Statistical evaluation will focus on intention to treat analyses evaluating the decision of HRT and it’s feasibility. With a compliance of 60% we will have sufficient statistical power (88%) to detect a fracture reduction of 40% in the treatments group. Clinical risk factors, current daily intakes of macronutrients, vitamins and minerals, anthropometric variables, biochemical variables (including bone markers and 25-hydroxyvitamin D), regional bone mineral density (BMD) and total body composition were assessed in all participants at entry and at various follow up intervals. Results: 2016 study participants were recruited by direct mailing to a random sample of 45–58 years old women. In the randomised arm 501 were allocated to HRT and 505 to no treatment. In the non-randomised arm 219 preferred HRT and 791 preferred no treatment. Post-randomisation analysis revealed a slight but significant difference in age (50.01 versus 50.44 years) but no difference in menopausal age, prevalence of hysterectomy, educational level, BMI, serum bone alkaline phosphatase, serum osteocalcin, urine hydroxyproline or serum 25-hydroxyvitamin D. In the non-randomised arm women preferring HRT were closer to menopause, had a higher prevalence of hysterectomy, were better educated, were leaner, and had lower bone turnover than the women, who refused HRT. Conclusion: It is possible to include a sufficient number of perimenopausal women in a randomised 20 year study on the antifracture effect of HRT.