Effect of high-sensitivity C-reactive protein on the development of diabetes as demonstrated by pooled logistic-regression analysis of annual health-screening information from male Japanese workers.

Aim: The aim of this study was to determine the relationship between the development of diabetes mellitus and high-sensitivity C-reactive protein (HsCRP) adjusted for various potential confounders.

Methods: This 5-year prospective cohort study was conducted at a Japanese steel factory and involved male workers who had received annual health screenings between 2005 and 2010. The 7392 male participants were aged 19-75 years. The study endpoint, the development of diabetes mellitus, was defined as HbA(1c) greater or equal to 6.5% or the use of antidiabetic medication. The association between variables was investigated using pooled logistic regression adjusted for various covariates such as age, baseline body mass index (BMI) and increase in BMI from baseline, blood biochemistry, job schedule and job-related stress.

Results: The incidence rate of diabetes development per 1000 person-years was 13.9. Multivariate analysis showed a significant relationship between the development of diabetes and elevated levels of baseline HsCRP and increases in levels from baseline. The Odds ratios for a 2.9-fold (±1 geometric standard deviation) increase in baseline HsCRP and increase in HsCRP level from baseline were 1.18 [95% confidence interval (CI): 1.03-1.34; P=0.018] and 1.21 (95% CI: 1.03-1.41; P=0.018), respectively.

Conclusion: The present study has indicated that HsCRP is an independent predictor for the development of diabetes in men, together with various confounders such as BMI, type of job schedule and job-related stress.