Associations of variability in body weight and glucose levels with the risk of hip fracture in people with diabetes

Diabetes is associated with a high risk of fragility fracture. However, there are controversies regarding the effect of fluctuations in metabolic parameters on the risk of fracture. We aimed to investigate the associations of body weight or glucose variability or their combination with the risk of hip fracture in people with diabetes.A population-based cohort study with 480,539 subjects over 40 years who had undergone three or more health examinations was performed. The degree of variability was evaluated using variability independent of the mean (VIM, 100 × standard deviation / meanbeta), coefficient of variation (CV), and average real variability (ARV, average of the absolute differences between consecutive values). High variability was defined as having values in the highest quartile. Cox proportional hazards models were used to estimate the risk of hip fracture.There were 2834 hip fracture events (0.59%) during the mean follow-up of 8.1 years. After multivariable adjustment for age, sex, alcohol consumption, smoking, regular exercise, income, glucose, body mass index, hemoglobin, estimated glomerular filtration rate, diabetes duration, diabetes treatment with multiple agents, and osteoporosis, the HRs (95% CI) of hip fracture were 1.36 (1.24-1.50) and 1.29 (1.16-1.43) for high body weight VIM and high glucose VIM, respectively. The HR (95% CI) of both high VIM group was 1.63 (1.44-1.83), suggesting an additive effect of variabilities in body weight and glucose. The results were consistent when using CV and ARV and in various sensitivity analyses.High variability in body weight and glucose levels is associated with an increased incidence rate and risk of hip fracture in people with diabetes.