AN EXAMINATION OF THE COVARIABILITY OF SUBCLINICAL ALBUMINURIA IN INSULIN-DEPENDENT DIABETES-MELLITUS - IMPLICATIONS FOR MONITORING MICROALBUMINURIA

We examined the covariability of subclinical albuminuria in a cohort of 160 insulin-dependent diabetics who underwent simultaneous measurements of overnight urinary albumin excretion rate (UAV) and several clinical variables on 7 occasions over a period of 18 months. After allowing for within-patient correlation between visits, significant within-patient associations were found between UAV and variations in tobacco consumption (regression coefficient=0.006, S.E.=0.002, P=0.02), insulin dose (regression coefficient=0.003, S.E.=0.001, P=0.01) and creatinine clearance (regression coefficient=0.776, S.E.=0.047, P < 0.0001); the association between UAV and creatinine clearance was not influenced by adjusting for urine flow rate. No significant associations were found between UAV and changes in blood pressure, glycaemic control or other variables. We then focused on 33 patients with intermittent microalbuminuria, defined as a UAV > 30 mu g/min on at least one occasion, but not exceeding 3 consecutive occasions. These patients had a total of 52 episodes of microalbuminuria which were significantly associated at the 5% level with increases in creatinine clearance (P=0.02), but not with changes in other variables. Our findings did not differ in patients with intermittent microalbuminuria defined with reference to a cut-off UAV of 20 mu g/min. We conclude that changes in glomerular filtration rate (as reflected by creatinine clearance) may partly account for the within-patient variability in subclinical albuminuria in insulin-dependent diabetics, and that variables routinely measured in the clinic are not useful for deciding when to monitor patients for microalbuminuria.