Background: Elevated high-sensitivity C-reactive protein (hs-CRP), a marker of low-grade systemic inflammation, may be involved in the etiology of type 2 diabetes mellitus (T2DM). However, whether inflammation precedes development of T2DM independent of cigarette smoking and obesity remains to be confirmed. Methods: We studied 4213 civil servants in a local government in Japan aged 35-66years at baseline in 2002, who donated blood samples and were followed 6years. Hazard ratios (HR) of T2DM according to the hs-CRP quartiles [range Q1: 0.02-0.18 (reference), Q2: 0.18-0.33, Q3: 0.33-0.67 and Q4: 0.67-9.62mg/L) were estimated by Cox proportional hazards model adjusted for gender, age, body mass index, alcohol intake, smoking status (current, past and never), number of cigarettes per day, physical activity, family history of diabetes (Model 1) and variables in Model 1+glucose (Model 2). Results: The geometric mean [95% confidence interval (CI)] of hs-CRP was 0.36mg/L (0.34-0.37). During the follow-up, 156 new T2DM cases were confirmed. In total sample, Model 2 HRs (95% CIs) for hs-CRP quartiles Q2-Q4 compared with Q1 were 0.69 (0.36-1.26), 1.47 (0.91-2.39) and 1.78 (1.10-2.88), respectively (p for linear trend=0.014). Stratified analysis revealed that a statistically significant association was observed only in normal weight non-current smokers with Model 2 HRs (CIs) being 0.79 (0.29-2.17), 2.63 (1.25-5.56) and 3.19 (1.49-6.86) for Q2-Q4 compared with Q1, respectively (p for linear trend=0.0006). The relationship did not change materially after further adjusting for log-homeostasis model assessment or exclusion of past smokers. Conclusions: These findings imply that higher hs-CRP itself or existence of chronic systemic inflammation precedes onset of T2DM independent of obesity and smoking. © 2013 John Wiley &amp
Sons, Ltd.