Regional and age-related variations in the proportions of hip fractures and major fractures among postmenopausal women: the Global Longitudinal Study of Osteoporosis in Women

We examined variations in proportions of hip fractures and major fractures among postmenopausal women using the Global Longitudinal Study of Osteoporosis in Women (GLOW). The proportion of major fractures that were hip fractures varied with age and region, whereas variations in the proportion of fractures that were major fractures appeared modest.IntroductionIn many countries, the World Health Organization fracture risk assessment tool calculates the probability of major fractures by assuming a uniform age-associated proportion of major fractures that are hip fractures in different countries. We further explored this assumption, using data from the GLOW.MethodsGLOW is an observational population-based study of 60,393 non-institutionalized women aged ≥55 years who had visited practices within the previous 2 years. Main outcome measures were self-reported prevalent fractures after the age of 45 years and incident fractures during the 2 years of follow-up.ResultsThe adjusted proportion of prevalent and incident major fractures after the age of 45 years that were hip fractures was higher in North America (16%, 17%) than in northern (13%, 12%) and southern Europe (10%, 10%), respectively. The proportion of incident major fractures that were hip fractures increased more than five-fold with age, from 6.6% among 55–59-year-olds to 34% among those aged ≥85 years. Regional and age-associated variations in the proportion of all incident fractures that were major fractures were less marked, not exceeding 16% and 28%, respectively.ConclusionsThe data suggest that there may be regional differences in the proportion of major fractures that are hip fractures in postmenopausal women. In contrast, the regional and age-related variations in the proportion of fractures that are major fractures appear to be modest. However, because of the limited number of fractures in our sample, further studies are necessary to confirm these findings.