A total of 4793 diabetes patients were enrolled from seven communities in Shanghai, China, in 2018. CVD was defined as a self-reported combination of previous diagnoses, including
coronary heart disease,
myocardial infarction and
stroke. DR was examined using fundus photographs. Logistic regression analyses were performed, and multiple imputed data were used to obtain stabilized estimates.
RESULTS: Prevalence of CVD increased with increasing
C-peptide levels (Q1, Q2, Q3 and Q4: 33%, 34%, 37% and 44%, respectively; Pfor trend < 0.001), whereas DR prevalence decreased with increasing
C-peptide quartiles (Q1, Q2, Q3 and Q4: 21%, 19%, 15% and 12%, respectively; Pfor trend < 0.001). On logistic regression analysis,
C-peptide levels were significantly associated with CVD prevalence (1.27, 95% CI: 1.13-1.42; P < 0.001) and
C-peptide quartiles (Q1: reference; Q2: 1.31, 95% CI: 1.00-1.70; Q3: 1.53, 95% CI: 1.16-2.01; Q4: 1.76, 95% CI: 1.32-2.34; Pfor trend < 0.001). Given the interaction between
C-peptide and BMI and the association between
C-peptide and CVD (Pfor interaction = 0.015), study participants were divided into two subgroups based on BMI which revealed that the association persisted despite different BMI statuses. However, DR prevalence decreased with increasing
C-peptide levels (0.73, 95% CI: 0.62-0.86; P < 0.001) and quartiles (Q1: reference; Q2: 1.00, 95% CI: 0.76-1.33; Q3: 0.69, 95% CI: 0.50-0.94; Q4: 0.51, 95% CI: 0.36-0.72; Pfor trend < 0.001).
CONCLUSION:
C-peptide was positively associated with CVD, but inversely associated with DR progression. The association between
C-peptide and CVD could be due to associated metabolic risk factors.