Variation in bone mineral density and fractures over 20 years among Canadians: a comparison of the Canadian Multicenter Osteoporosis Study and the Canadian Longitudinal Study on Aging

Abstract

Summary: International variations in osteoporosis and fracture rates have been reported, with temporal trends differing between populations. We observed higher BMD and lower fracture prevalence in a recently recruited cohort compared to that of a cohort recruited 20 years ago, even after adjusting for multiple covariates. Purpose: We explored sex-specific differences in femoral neck bone mineral density (FN-BMD) and in prevalent major osteoporotic fractures (MOF) using two Canadian cohorts recruited 20 years apart. Methods: We included men and women aged 50–85 years from the Canadian Multicentre Osteoporosis Study (CaMos, N = 6,479; 1995–1997) and the Canadian Longitudinal Study on Aging (CLSA, N = 19,534; 2012–2015). We created regression models to compare FN-BMD and fracture risk between cohorts, adjusting for important covariates. Among participants with prevalent MOF, we compared anti-osteoporosis medication use. Results: Mean (SD) age in CaMos (65.4 years [8.6]) was higher than in CLSA (63.8 years [9.1]). CaMos participants had lower mean body mass index and higher prevalence of smoking (p < 0.001). Adjusted linear regression models (estimates [95%CI]) demonstrated lower FN-BMD in CaMos women (− 0.017 g/cm2 [− 0.021; − 0.014]) and men (− 0.006 g/cm2 [− 0.011; 0.000]), while adjusted odds ratios (95%CI) for prevalent MOF were higher in CaMos women (1.99 [1.71; 2.30]) and men (2.33 [1.82; 3.00]) compared to CLSA. In women with prevalent MOF, menopausal hormone therapy use was similar in both cohorts (43.3% vs 37.9%, p = 0.076), but supplements (32.0% vs 48.3%, p < 0.001) and bisphosphonate use (5.8% vs 17.3%, p < 0.001) were lower in CaMos. The proportion of men with MOF who received bisphosphonates was below 10% in both cohorts. Conclusion: Higher BMD and lower fracture prevalence were noted in the more recently recruited CLSA cohort compared to CaMos, even after adjusting for multiple covariates. We noted an increase in bisphosphonate use in the recent cohort, but it remained very low in men.