Modeling GFR trajectories in diabetic nephropathy.

Area of Science:

• Nephrology
• Endocrinology
• Biostatistics

Background:

• Diabetic nephropathy is a major complication of type 2 diabetes, characterized by progressive decline in kidney function.
• Microalbuminuria and macroalbuminuria are key indicators of kidney damage in diabetic patients.
• Understanding the longitudinal changes in glomerular filtration rate (GFR) is crucial for predicting disease progression.

Purpose of the Study:

• To compare the longitudinal changes in glomerular filtration rate (GFR) in Pima Indians with type 2 diabetes and nephropathy.
• To identify distinct GFR trajectories based on the progression from microalbuminuria to macroalbuminuria.
• To analyze GFR decline using novel statistical methods with minimal assumptions.

Main Methods:

• Longitudinal study over 8 years involving Pima Indians with type 2 diabetes and nephropathy.
• Serial measurement of GFR using urinary iothalamate clearances.
• Modeling of individual GFR courses using smoothing and regression cubic B-splines.
• Group comparisons utilizing a permutation approach to Hotelling's T(2) statistic on fitted GFR vectors.

Main Results:

• GFR profiles of progressors (microalbuminuria to macroalbuminuria) significantly differed from nonprogressors (P = 0.003).
• No significant baseline clinical differences were observed between progressors and nonprogressors.
• The GFR decline trajectory after progressing to macroalbuminuria did not differ from those newly screened with macroalbuminuria (P = 0.27).

Conclusions:

• The study demonstrates distinct GFR decline trajectories in subgroups of diabetic Pima Indians with microalbuminuria.
• These trajectories are related to the extent of glomerular barrier dysfunction, indicated by the progression from microalbuminuria to macroalbuminuria.
• Novel statistical techniques reveal subgroup-specific GFR decline patterns without imposing simplifying models.