Hba1c Variability Is An Independent Predictor Of Incident Hip Fractures In Chinese People With Type 2 Diabetes

Abstract

Background and aims: Both type 2 diabetes and osteoporosis are becoming more prevalent with ageing population. Type 2 diabetes is associated with a 1.3-fold increase in the risk of hip fractures, which in turn are associated with significant morbidity and mortality. Higher glycated haemoglobin (HbA1c) has been shown to be associated with higher risk of hip fracture. However, little is known whether long-term glycaemic variability predicts incident hip fractures. We hypothesized that HbA1c variability could predict incident hip fractures in type 2 diabetes. Materials and methods: Chinese individuals with type 2 diabetes aged ≥60 years were identified from electronic health records in Hong Kong between 2008 and 2012, and observed for incident hip fractures. Diagnosis of hip fracture was defined by International Classification of Diseases, Ninth Revision (ICD-9) 820. Baseline HbA1c variability was determined using standard deviation (HbA1c-SD), adjusted standard deviation (HbA1c-AdjSD), and coefficient of variation (HbA1c-CV) of HbA1c measurements in the 5 years preceding the entry date. Multivariable Cox regression analysis was used to evaluate the associations between HbA1c variability and incident hip fractures. Results: 83282 participants were included (53.8% women), with mean age of 71.3±7.5 years, body mass index (BMI) of 25.2±3.8 kg/m2, duration of diabetes of 11.7±7.7 years, and baseline HbA1c of 7.33±1.23%. Over a median follow-up of 6.8 years, 3066 participants had incident hip fractures, a cumulative incidence of 5.44 per 1000 person-years. Individuals with incident hip fractures were older, more likely women and ever-smokers, had lower BMI, longer duration of diabetes, higher prevalence of severe hypoglycaemia (all p<0.001), more comorbidities (hypertension, cardiovascular diseases and renal impairment) (p<0.05), but less dyslipidaemia (p<0.001). On multivariable Cox regression models, higher baseline HbA1c (adjusted hazard ratio [aHR] 1.04), higher mean HbA1c (aHR 1.10), and higher HbA1c variability (HbA1c-SD [aHR 1.26], HbA1c-AdjSD [aHR 1.29] and HbA1c-CV [aHR 1.02]) were significantly associated with increased risk of incident hip fractures (all p<0.001) in addition to age, women, smoking history, duration of type 2 diabetes, severe hypoglycaemia, renal impairment and cardiovascular diseases, BMI and dyslipidaemia (all p<0.05). Comparing with the first quartile of HbA1c-CV, the risk of incident hip fractures significantly increased in quartiles 3 and 4 (19% p=0.002; 48%, p<0.001 respectively) even after adjustment for risk factors at baseline. In the sensitivity analysis, we also demonstrated a similar predictive ability of all indices of HbA1c variability for incident hip fractures, showing significant aHR of similar magnitude in participants with a mean HbA1c <7, 7.0-7.9, and ≥8.0% . Conclusion: We provided the first report of the significance of HbA1c variability as an independent predictor of incident hip fractures in type 2 diabetes, in addition to baseline and mean HbA1c. HbA1c variability remained a significant independent predictor of incident hip fractures regardless of overall glycaemic control as indicated by HbA1c. These suggest the important benefit of achieving glycaemic stability in addition to target HbA1c from the perspective of bone health.