Abstract
Association between diabetes mellitus (diabetes) and risk of knee osteoarthritis (KOA) is confounded by high body mass index (BMI), a strong risk factor for both conditions. We evaluated the association between diabetes and incidence of total knee replacement (TKR) due to severe KOA in the Singapore Chinese Health Study, a prospective cohort of 63,257 Chinese men and women, aged 45-74 years at recruitment in 1993-1998, and re-interviewed in 1999-2004. Height, weight, lifestyle factors and history of diabetes were obtained through in-person interviews at recruitment and re-interview. Incident cases of TKR were identified via record linkage with nationwide hospital discharge database. Subjects with/without prevalent diabetes had comparable BMI (24.0 kg/m(2) versus 23.0 kg/m(2)). After an average of 14-years, 1,973 subjects had TKR attributable to KOA. Compared to subjects without diabetes, hazard ratio (HR) of TKR for subjects with diabetes was 0.63 [95% confidence interval (CI), 0.52-0.75] after controlling for BMI and other risk factors. An inverse association was also observed between incident diabetes at re-interview and subsequent risk of TKR (HR = 0.74; 95% CI = 0.58-0.94). The inverse diabetes-TKR risk association was similar by gender and across three categories of BMI. Our study does not support diabetes as a risk factor of KOA.