C-reactive protein partially mediates the inverse association between coffee consumption and risk of type 2 diabetes: The UK Biobank and the Rotterdam study cohorts*

Background: Coffee is among the most consumed beverages worldwide. Coffee consumption has been associated with lower risk of type 2 diabetes mellitus (T2D), but underlying mechanisms are not well understood. We aimed to study the role of classic and novel-T2D biomarkers with anti-or pro -inflammatory activity in the association between habitual coffee intake and T2D risk. Furthermore, we studied differences by coffee types and smoking status in this association. Methods: Using two large population-based cohorts, the UK-Biobank (UKB; n = 145,368) and the Rot-terdam Study (RS; n = 7111), we investigated associations of habitual coffee consumption with incident T2D and repeated measures of insulin resistance (HOMA-IR), using Cox proportional hazards and mixed effect models, respectively. Additionally, we studied associations between coffee and subclinical inflammation biomarkers including C-reactive protein (CRP) and IL-13, and adipokines, such as adipo-nectin and leptin, using linear regression models. Next, we performed formal causal mediation analyses to investigate the role of coffee-associated biomarkers in the association of coffee with T2D. Finally, we evaluated effect modification by coffee type and smoking. All models were adjusted for sociodemo-graphic, lifestyle and health-related factors. Results: During a median follow-up of 13.9 (RS) and 7.4 (UKB) years, 843 and 2290 incident T2D cases occurred, respectively. A 1 cup/day increase in coffee consumption was associated with 4% lower T2D risk (RS, HR = 0.96 [95%CI 0.92; 0.99], p = 0.045; UKB, HR = 0.96 [0.94; 0.98], p < 0.001), with lower HOMA-IR (RS, log-transformed f3 =-0.017 [-0.024;-0.010], p < 0.001), and with lower CRP (RS, log -transformed f3 =-0.014 [-0.022;-0.0 05], p = 0.002; UKB, f3 =-0.011 [-0.012;-0.0 09], p < 0.001). We also observed associations of higher coffee consumption with higher serum adiponectin and IL-13 concentrations, and with lower leptin concentrations. Coffee-related CRP levels partially mediated the inverse association of coffee intake with T2D incidence (average mediation effect RS f3 = 0.105 (0.014; 0.240), p = 0.016; UKB f3 = 6.484 (4.265; 9.339), p < 0.001), with a proportion mediated by CRP from 3.7%