Abstract / Summary
Dyslipidemia is a major risk factor for cardiovascular complications in patients with type 2 diabetes mellitus (T2DM), and early risk assessment may help inform follow-up and lipid monitoring. This study aimed to develop and externally validate a nomogram based on Cox proportional hazards regression to predict the risk of incident LDL-C elevation (LDL-C ≥ 2.6 mmol/L) in patients with T2DM in primary care settings. We conducted a multicenter retrospective cohort study involving 1,893 patients with T2DM from five community health service centers in Guangzhou, China. The development cohort (n = 1,072), drawn from a single center, was randomly divided into a training set (n = 750) and an internal validation set (n = 322), while four geographically distinct centers contributed an independent external validation cohort (n = 821). Candidate predictors were screened using least absolute shrinkage and selection operator (LASSO) regression with 10-fold cross-validation, and the selected variables were used to construct a multivariable Cox regression model. A nomogram was developed to estimate the 1-year, 3-year, and 5-year probabilities of incident LDL-C elevation. Model performance was assessed using the concordance index (C-index), time-dependent receiver operating characteristic (ROC) curves, calibration plots, and decision curve analysis (DCA). Seven predictors were retained in the final nomogram: diabetes duration, fasting blood glucose control, regular exercise, body mass index, total bilirubin, triglycerides, and low-density lipoprotein cholesterol. The model showed moderate discrimination across the training, internal validation, and external validation cohorts. The 5-year AUCs were 0.819 (95%CI: 0.780–0.859), 0.810 (95%CI: 0.752–0.868) and 0.772 (95%CI: 0.718–0.826), respectively. Calibration and decision-curve analyses were also performed, and Kaplan–Meier curves differed across the three score-defined risk groups (P < 0.001). We developed and externally validated a nomogram that predicts the 1-year, 3-year, and 5-year risk of incident LDL-C elevation in community-dwelling patients with T2DM without major cardiovascular or renal complications. Because all incorporated predictors are simple, routinely available and easily obtainable, this model may, after further validation and clinical-impact evaluation, be considered for risk stratification in similar community settings, particularly in resource-limited community settings. Not applicable.