Arthritis is associated with an increased risk of incident diabetes in Chinese adults: A nationwide cohort study and updated meta-analysis.

Abstract

To prospectively examine the association between arthritis and type 2 diabetes (T2D) in the Chinese population and confirm this association through a comprehensive meta-analysis of cohort studies. Data were from the China Health and Retirement Longitudinal Study which was started in 2011-2013 and followed up in 2013-2014 and 2015-2016. Arthritis was defined as self-reported physician diagnosis at baseline, and incident T2D was determined by self-reported physician diagnosis, fasting blood glucose ≥7.0mmol/L or glycosylated haemoglobin ≥6.5% during the follow-ups. Cox proportional hazards regression models were used to assess the association between arthritis and risk for T2D. A meta-analysis was conducted to pool our effect estimate and those from other cohort studies using a random-effects model. Eleven thousand four hundred and eight participants (47.9% men; mean age: 59.3years) were included in final analyses. During a 4-year follow-up, 981 participants reported incident T2D. Compared with individuals without arthritis, those with arthritis at baseline had an 18% higher risk for incident T2D (multivariable-adjusted hazard ratio: 1.18; 95% confidence interval: 1.04, 1.34). In the meta-analysis of 13 cohort studies including ours, a total of 2,473,514 participants were included with 121,851 incident diabetes. The pooling HR was 1.32 (95% CI: 1.21, 1.44) for the association between arthritis and diabetes. Arthritis was associated with an increased risk of incident diabetes in Chinese adults, and the positive association was confirmed in the meta-analysis of cohort studies. Our work can inform clinical trials to assess the effectiveness of arthritis treatments in reducing risk of diabetes.