Predictors and generation of risk equations for albuminuria progression in type 2 diabetes

Background: Diabetes is the most common cause of end-stage renal disease and is associated with increased mortality. Although only a proportion of type 2 diabetic subjects develop albuminuria or progress, it is not currently possible to identify those patients who will develop this complication or who will progress. Aim: The aim of the study was to identify baseline risk factors for the development and progression of albuminuria in a cohort with type 2 diabetes and use this data to generate risk equations. Patients and methods: Type 2 diabetic subjects who had albumin-creatinine ratio (ACR) measurement in 2007 - 2008 were recruited and followed-up for 8 years. Results: 260 patients were included in the study. Of all the normoalbuminuric and microalbuminuric patients, 24.3% progressed. Baseline HbA1c, white cell count (WCC), smoking, and duration of diabetes were associated with progression of albuminuria stage in univariate analysis. Duration of diabetes (p = 0.034) was independently associated with progression in binary logistic regression. Baseline HbA1c (p = 0.002), age (p = 0.01), serum creatinine (p = 0.02), serum potassium (p = 0.04), serum urea (p = 0.0004), WCC (p = 0.02), serum triglycerides (p = 0.02), systolic blood pressure (p = 0.02), and duration of diabetes (p = 0.003) were positively correlated with percentage change (% change) in ACR, whilst baseline estimated glomerular filtration rate (eGFR) (p = 0.03), serum sodium (p = 0.04), hemoglobin (p = 0.0006), and hematocrit (p = 0.0002) were negatively correlated in Spearman correlation. Duration of diabetes (p = 0.025) and baseline HbA1c (p = 0.02) were independently associated with % change in ACR in multivariate analysis. Based on these results, novel risk equations were generated. Conclusions: We have identified baseline characteristics associated with progression of renal disease in type 2 diabetic subjects and generated equations to estimate the risk of progression. If validated in other populations, these equations might be useful in predicting risk of progression in clinical practice.