Exercise testing as a long-term predictor of the development of microalbuminuria in normoalbuminuric IDDM patients.

Abstract

To determine the association between exercise-induced albuminuria and the development of microalbuminuria over 10 years in subjects with insulin-dependent diabetes mellitus (IDDM) who were initially normoalbuminuric. Thirty-two patients with IDDM and a resting urinary albumin/creatinine ratio (UA/UC) < 2.1 mg/mmol (< 15 micrograms/min) were exercised after water loading on a treadmill for 20 min at double their resting heart rate. UA/UC was determined before and after exercise. The exercise test was considered positive if the UA/UC was > 4.3 mg/mmol (> 30 micrograms/min). Results were compared with resting UA/UC after a 10-year follow-up. Persistent microalbuminuria was defined as a UA/UC > 2.1 mg/mmol (> 15 micrograms/min) in each of two early-morning urine collections. Five patients developed persistent microalbuminuria after 10 years, and four patients were predicted by a positive exercise test. Two patients with positive exercise tests did not develop persistent microalbuminuria. The sensitivity of the exercise test for the development of microalbuminuria was 80% (95% confidence interval [CI] 65.8-94.2%) and the specificity was 92.9% (95% CI 83.9-100%). The postexercise UA/UC was positively associated with the UA/UC after 10 years (P = 0.005, R2 = 0.31). This association was independent of HbA1, systolic blood pressure, body mass index, and duration of diabetes, but HbA1 remained an independent predictor (P = 0.02) of UA/UC at follow-up. Exercise testing may be useful for identifying normoalbuminuric IDDM patients who are susceptible to the later development of microalbuminuria.