RELATIONSHIP BETWEEN EARLY METABOLIC CONTROL AND THE DEVELOPMENT OF MICROALBUMINURIA - A LONGITUDINAL-STUDY IN CHILDREN WITH TYPE-1 (INSULIN-DEPENDENT) DIABETES-MELLITUS

The cumulative incidence of microalbuminuria from onset up to 15 years of Type 1 (insulin-dependent) diabetes mellitus and the relative importance of age, duration, blood pressure and metabolic control for subsequent microalbuminuria was studied in 156 children. Urinary albumin excretion and HbA1c were followed at 3-month intervals from onset and systolic and diastolic blood pressure at the same interval from 5 years of diabetes. Persistent microalbuminuria over 20 mug/min developed in 17 children. The cumulative incidence by duration was 24.2% at 15 years of diabetes. Eleven patients developed microalbuminuria after more than 5 years. Among these, first 5-year mean HbA1c was 8.4 +/- 1.3% vs 7.2 +/- 1.1% in normoalbuminuric children (p < 0.001). The crude relative risk of developing microalbuminuria with a first 5-year mean HbA1c greater than 7.5% was 4.49 (95 % confidence interval 1.13-17.84). The age- and duration-adjusted relative risk was 3.51% (0.90-14.42). The year before transition to microalbuminuria neither mean HbA1c nor yearly mean blood pressures, sex or age at onset of diabetes differed from normoalbuminuric children at the same diabetes duration. Age and duration were higher (p = 0.04). The relative importance of early vs later hyperglycaemia, yearly blood pressures, age, age at onset and duration of diabetes for increased albumin excretion rate after more than 5 years, was shown in a multiple regression analysis where the first 5-year mean HbA1c was the only independent predictor (p = 0.02). Six patients had an onset of microalbuminuria before 5 years of diabetes. In this group age, age at onset and yearly mean HbA1c levels did not differ from normoalbuminuric children at the same duration. We conclude that one-fourth of diabetic youth under the age of 21 years develop microalbuminuria within 14 years of diagnosis. Some patients have an early onset of microalbuminuria, not necessarily related to hyperglycaemia, while in later onset cases early hyperglycaemia is strongly related to subsequent microalbuminuria.