Abstract
BACKGROUND: Urinary markers were tested as predictors of macroalbuminuria or microalbuminuria in patients with type 1 diabetes. STUDY DESIGN: Nested case-control of participants in the Diabetes Control and Complications Trial (DCCT). SETTING & PARTICIPANTS: 87 cases of microalbuminuria were matched to 174 controls in a 1:2 ratio, while 4 cases were matched to 4 controls in a 1:1 ratio, resulting in 91 cases and 178 controls for microalbuminuria. 55 cases of macroalbuminuria were matched to 110 controls in a 1:2 ratio. Controls were free of micro-/macroalbuminuria when their matching case first developed micro-/macroalbuminuria. PREDICTORS: Urinary N-acetyl-beta-d-glucosaminidase (NAG), pentosidine, advanced glycation end product (AGE) fluorescence, and albumin excretion rate (AER). OUTCOMES: Incident microalbuminuria (2 consecutive annual AERs > 40 but < or = 300 mg/d) or macroalbuminuria (AER > 300 mg/d). MEASUREMENTS: Stored urine samples from DCCT entry and 1-9 years later when macro- or microalbuminuria occurred were measured for the lysosomal enzyme NAG and the AGE pentosidine and AGE fluorescence. AER and adjustor variables were obtained from the DCCT. RESULTS: Submicroalbuminuric AER levels at baseline independently predicted microalbuminuria (adjusted OR, 1.83; P < 0.001) and macroalbuminuria (adjusted OR, 1.82; P < 0.001). Baseline NAG excretion independently predicted macroalbuminuria (adjusted OR, 2.26; P < 0.001) and microalbuminuria (adjusted OR, 1.86; P < 0.001). Baseline pentosidine excretion predicted macroalbuminuria (adjusted OR, 6.89; P = 0.002). Baseline AGE fluorescence predicted microalbuminuria (adjusted OR, 1.68; P = 0.02). However, adjusted for NAG excretion, pentosidine excretion and AGE fluorescence lost the predictive association with macroalbuminuria and microalbuminuria, respectively. LIMITATIONS: Use of angiotensin-converting enzyme inhibitors was not directly ascertained, although their use was proscribed during the DCCT. CONCLUSIONS: Early in type 1 diabetes, repeated measurements of AER and urinary NAG excretion may identify individuals susceptible to future diabetic nephropathy. Combining the 2 markers may yield a better predictive model than either one alone. Renal tubule stress may be more severe, reflecting abnormal renal tubule processing of AGE-modified proteins, in individuals susceptible to diabetic nephropathy.