Abstract
BACKGROUND: Few chronic kidney disease (CKD) risk prediction models have been investigated in low- and middle-income areas worldwide. We developed new risk scores for predicting incident CKD in low- and middle-income rural Chinese populations. METHODS: Data from the Handan Eye Study, which was a village-based cohort study and conducted from 2006 to 2013, were utilized as part of this analysis. The present study utilized data generated from 3266 participants who were ≥ 30 years of age. Two risk models for predicting incident CKD were derived using two-thirds of the sample cohort (selected randomly) using stepwise logistic regression, and were subsequently validated using data from the final third of the sample cohort. In addition, two simple point systems for incident CKD were generated according to the procedures described in the Framingham Study. CKD was defined as reduced renal function (estimated glomerular filtration rate (eGFR) < 60 mL/min/1.73m(2)) or the presence of albuminuria (urinary albumin-to-creatinine ratio (UACR) ≥30 mg/g). RESULTS: The Simple Risk Score included waist circumference, systolic blood pressure (SBP), diabetes, sex, and education. The Best-fit Risk Score included urinary albumin-to-creatinine ratio, SBP, C-reactive protein, triglyceride, sex, education, and diabetes. In the validation sample, the areas under the receiver operating curve of the Simple Risk Score and Best-fit Risk Score were 0.717 (95% CI, 0.689-0.744) and 0.721 (95% CI, 0.693-0.748), respectively; the discrimination difference between the score systems was not significant (P = 0.455). The Simple Risk Score had a higher Youden index, sensitivity, and negative predictive value, with an optimal cutoff value of 14. CONCLUSIONS: Our Simple Risk Score for predicting incident CKD in a low- and middle-income rural Chinese population will help identify individuals at risk for developing incident CKD.