The purpose of this project is to determine the natural course of renal disease in type 2 diabetes mellitus in the Pima Indians and to identify the underlying pathogenetic mechanisms involved in the initiation and progression of renal disease in this type of diabetes. This project, in part, represents an extension of work previously reported as Project Number Z01 DK 69037. Glomerular function was measured over a 4-year period in 194 Pima Indians selected to represent stages in the development and progression of diabetic renal disease. Eighty-six of these subjects had type 2 diabetes and micro- or macroalbuminuria. Subjects underwent serial determinations of albuminuria and GFR. Statistical models, including linear regression, a generalized additive model, a tree-based model, and artificial neural networks were used to identify predictors of GFR decline. Data were randomly assigned to a training set (n=60) or a testing set (n=26). Performance of the statistical models generated on the training set was evaluated on the testing set by correlation coefficients between observed and predicted values, root mean squared prediction errors, mean absolute deviation, and percent variation explained. Each of the statistical models successfully predicted long-term GFR based on baseline data. Proteinuria remained the single most important predictor of long-term renal function. Other determinants included initial GFR, blood pressure, plasma renin activity, serum lipid concentrations, age, body mass index, and duration of diabetes. Thirty-six of these subjects were then followed for an additional 4 years with serial measures of GFR. GFR slope was computed using smoothing and regression cubic B-splines with a basis width of 18 months. Subjects whom initially had microalbuminuria and remained microalbuminuric during the study had the slowest decline in renal function during the study. Those that progressed from micro- to macroalbuminuria had an intermediate rate of decline, and those with new-onset macroalbuminuria at baseline had the most rapid rate of decline. Our data suggest that a protracted course of microalbuminuria represents a more benign course of kidney disease, whereas the initiating diabetic renal injury or some aspect of the progression process is more severe in subjects who manifest an early onset of macroalbuminuria. Our efforts in the coming year will focus on the pathobiology of kidney disease. We wish to identify, in particular, the role of the glomerular podocyte in the development and progression of diabetic kidney disease.
Showing the most recent 10 out of 25 publications
