Effect modification of the association between plasma glucose and diabetic kidney disease by hypersensitive C-reactive protein in patients with diabetes mellitus

Background Prior studies showed activation inflammatory biomarkers, e.g., hypersensitive C-reactive protein (hs-CRP), were associated with diabetic kidney disease (DKD) susceptibility; inflammatory gene expression profiles in the diabetic mice might be critical features determining susceptibility of DKD. The aim of this investigation was to explore effect modification of hs-CRP on the association between hyperglycemia and DKD in type 2 diabetes mellitus (T2DM). Methods We consecutively collected 812 patients with T2DM in a cross-sectional study. Multivariable logistic regression models was used to estimate odd ratios and 95% confidence intervals of plasma glucose for DKD, with adjustment for the potential confounders. Interaction between plasma glucose and hs-CRP was tested by likelihood ratio tests. Results The median age of the participants was 54 years (interquartile: 46–63), 58% were male and 26.2% experienced DKD. There seemed to be a nonlinear effect modification on the association between fast plasma glucose (FPG) and DKD by hs-CRP (P for linear interaction = 0.052). For participants with hs-CRP value lower than 3 mg/L, there no existed an interaction effect (P interaction = 0.3). In contrast, a significant interaction effect was noted among individuals with hs-CRP value higher than 3 mg/L (P interaction = 0.003).The marginal effect, i.e., log (odds), of FPG on DKD linearly rose with hs-CRP level increasing among these subjects. In terms of effect modification on the relationship between 2-hour plasma glucose (2 h-PG) and DKD by hs-CRP, it appeared to be linear. The higher the hs-CRP value, the stronger the strength of their associations. Johnson-Neyman plot showed when hs-CRP level was lower than 2.4 mg/L the marginal effect of 2 h-PG on DKD was nonsignificant (boundary of 95% confidence interval included zero) and it became significant as hs-CRP level higher than 2.4 mg/L (boundary of 95% confidence interval excluded zero) . Conclusions The associations of 2 h-PG as well as FPG with DKD were modified by hs-CRP in individuals with T2DM. It appeared when hs-CRP level was higher than 3 mg/L in these patients, the association strength of both FPG and 2 h-PG with DKD linearly rose with hs-CRP level increasing. A prospective study is warranted to confirm this finding.