Abstract
Background: After two-years of follow-up of 263 apparently healthy 18- to 30-year-old men in communities affected by Mesoamerican nephropathy (MeN), we identified three distinct case groups: a subgroup with (i) established renal dysfunction (case-group 1); individuals with (ii) a rapid decline in kidney function (case-group 2); and individuals with (iii) stable kidney function (non-cases). This paper investigates whether local tests are potentially useful for the timely identification of these case groups. Methods: Creatinine levels were measured in local laboratories every six months for two years. Aliquots were sent to a centralized laboratory for measurements of cystatin C and creatinine levels. We investigated agreement between the locally and centrally measured creatinine-based Chronic Kidney disease Epidemiology Collaboration (CKD-EPI) equation for estimating the Glomerular Filtration Rate (eGFR). A logistic regression analysis was used to assess predictive factors for case groups 1 and 2 compared to non-cases. Predictive variables were locally measured eGFR, and urinary neutrophil gelatinase-associated lipocalin (uNGAL) levels. The discrimination performance of the model was assessed using the area under the receiver operating characteristic curve (AUC). Results: Considerable variation in local eGFR measurements was observed. The prediction model for case-group 1 included baseline kidney function and with or without uNGAL (AUC = 0.98, 95% confidence interval (CI), 0.91–1.00). The prediction model for case-group 2 also required eGFRScr at six and twelve months after baseline, with or without uNGAL levels (AUC = 0.88; 95% CI 0.80–0.99). Conclusions: Established renal dysfunction was detected at a single time point using local measurements and uNGAL. For the detection of a rapid decline in kidney function over time, at least 2 more measurements at six and twelve months are needed.