Effect of Exercise on Urinary N-Acetyl-Beta-D-Glucosaminidase Activity and Albumin Excretion in Children with Type I Diabetes Mellitus

Urinary N-acetyl-beta-D-glucosaminidase (NAG), a proximal tubule lysosomal enzyme, has been used as an indicator of subtle renal injury. Since it has been positively and significantly correlated with hemoglobin A1c and microalbuminuria, it has been suggested that this enzyme may also reflect metabolic control. Albumin excretion is exacerbated in adult diabetic individuals during exercise; such exercise-induced albuminuria may be a forerunner of diabetic nephropathy. Metabolic control, degree of exertion, and duration of diabetes have been suggested to influence this increase in albuminuria during exercise. Studies of children are few and have produced inconsistent results. Thus we studied 28 insulin-dependent diabetic children ranging in age from 5 yr to 16 yr and 27 age-matched controls using treadmill exercise; two exercise periods consisting of (1) graded increases in speed and grade at 3-min intervals until exhaustion and (2) a constant speed and grade necessary to produce ⅔-¾ maximal heart rate for 30 min were performed. Capillary blood glucose, urinary NAG/creatinine (cr) ratios (U_NAG/U_cr) and urinary albumin/creatinine ratio (U_alb/U_cr) were measured before and after each exercise period; hemoglobin A1c was also measured. The latter averaged 11.8 ± 0.6% (mean ± SEM); contrary to previous studies, this was not correlated with pre- or postexercise U_NAG/U_cr. During both exercise periods, blood glucose dropped 271 ± 19 mg/dl to 213 ± 21 mg/dl (period 1) and 230 ± 22 mg/dl to 157 ± 21 mg/dl (period 2). U_alh/U_cr was slightly but not significantly elevated at rest in the diabetic subjects (0.020 ± 0.005 versus 0.016 ± 0.003 mg/mg) (P = NS); however, U_NAG/U_cr was significantly higher (7.3 ± 0.7 versus 2.9 ± 0.3) (P < 0.05). Regardless of age, duration of disease, or metabolic control as measured by hemoglobin A_lc, no significant differences in albumin excretion were noted between diabetic subjects and controls after exercise; nor did U_alh excretion increase significantly in either group during either exercise period. However, U_NAG/U_cr was elevated in all groups in the diabetic subjects compared with the controls (P < 0.05). Furthermore, during both exercise periods U_NAG/U_cr consistently decreased in the controls (− 6% and −8% for periods 1 and 2, respectively), whereas U_NAG activity consistently rose in the diabetic subjects (27% and 36%); the differences were significant (P < 0.05 and P < 0.01). Thus, U_alh excretion is not significantly different in children with insulin-dependent diabetes compared with controls before or after either maximal or submaximal exercise. However, U_NAG activity showed significant differences before exercise and a significantly different response to exercise in the diabetic subjects and controls. Such enzymuria appears to be a more sensitive indicator of exercise-induced renal abnormalities than does albuminuria in children. Whether such changes will be useful in predicting early nephropathy or selecting those patients at risk for nephropathy requires further study.