Postprandial serum C-peptide to plasma glucose ratio predicts future insulin therapy in Japanese patients with type 2 diabetes

Abstract

Type 2 diabetes is a progressive disease, and most patients with type 2 diabetes eventually need insulin therapy to maintain glycemic control. Recent studies have shown that early insulin treatment improved beta cell function as well as insulin resistance in patients with type 2 diabetes[1], suggesting that early introduction of insulin treatment may delay disease progression and be able to maintain good glycemic control thereafter. However, since predictive biomarkers for future insulin therapy have not been established, the introduction of insulin treatment is often delayed in clinical settings [2]. We have recently reported that lower 2 h postprandial C-peptide immunoreactivity (CPR) to plasma glucose ratio (postprandial CPR index; PCPRI) was associated with subsequent need for insulin treatment in Japanese subjects with type 2 diabetes who were admitted to our hospital between 2000 and 2007 [3]. In that study, however, since most patients were started on insulin treatment during the baseline admission, the time-to-event relationship of insulin initiation could not be accounted for in the analysis. Here, we report a sub analysis of this cohort in whom insulin therapy had not been introduced at baseline (N = 190), to establish the predictive value of PCPRI for future insulin therapy using univariate and multivariate Cox proportional hazard models. In addition, the prognostic performance of PCPRI was estimated using receiver operating characteristic (ROC) curve analysis. Baseline characteristics of the subjects were as follows; male 57.9 %, age 65 ± 13 years (mean ± SD), duration of diabetes 10 ± 9 years, BMI 25.3 ± 4.7, HbA1c 9.5 ± 1.8 % (80 ± 20 mmol/mol), fasting CPR 2.14 ± 0.95 ng/ml, fasting CPRI 1.56 ± 0.74, PCPR 5.62 ± 2.59 ng/ml, PCPRI 2.72 ± 1.62 and urinary CPR 78.4 ± 54.7 lg/day. Of the subjects, 52.1 % were treated with a sulfonylurea at baseline. During the observation period (mean 4.2 ± 2.3 years, median 3.0 years), insulin therapy was introduced in 34 subjects (18 %). Higher PCPRI at baseline was associated with a lower risk of future insulin therapy (hazard ratio (HR) for 1 unit increase in PCPRI was 0.48, 95 % CI: 0.32 to 0.73, P \ 0.001). However, neither fasting CPRI nor urinary CPR predicted future insulin therapy (P = 0.30 and P = 0.72, respectively). After adjustment for age, sex, duration of diabetes, family history of diabetes, BMI, HbA1c and sulfonylurea use at baseline, PCPRI remained an independent predictor of future insulin therapy (HR = 0.42, 95 % CI: 0.25 to 0.70, P = 0.001). The optimal cutoff value of PCPRI for predicting insulin therapy was 2.16, with 70.6 % sensitivity and 64.1 % specificity (area under the curve (AUC) of ROC curve = 0.698, 95 % CI: 0.609 to 0.788, P \ 0.001). In the present study, PCPRI predicted future insulin therapy in patients with type 2 diabetes. Interestingly, neither fasting CPR index nor urinary CPR was significantly associated with future insulin therapy, suggesting that PCPRI is the superior marker for future insulin therapy Communicated by Renato Lauro.