Practical aspects of screening for and monitoring microalbuminuria in diabetes mellitus

Watts, Gerald Francis (1990) Practical aspects of screening for and monitoring microalbuminuria in diabetes mellitus. University of Southampton, Doctoral Thesis.

Record type: Thesis (Doctoral)

Abstract

Laboratory methods for measuring low concentrations of albumin in urine were compared: immunoturbidimetry was the most practicable assay. Side-room tests capable of detecting a urinary albumin concentration (U_A) of 15 mg/l were assessed; the latex bead test was the most appropriate for screening purposes. Reference values for urinary albumin excretion were derived from 120 healthy, adult subjects: U_A was measured by radioimmunoassay; multivariate regression analysis (MRA) showed that U_A was influenced by age and urinary albumin:creatinine ratio (U_A/U_C) by surface area. Overnight and daytime urine samples were obtained from 172 normotensive, insulin-dependent diabetics (IDDs) without clinical proteinuria. Microalbuminuria (M) was defined as the overnight albumin excretion rate (U_AV): the prevalence of M > 30 μg/min was 18% the most valid screening test was the overnight U_A/U_C. In a cross-section study MRA revealed that only 23% of the variability of overnight U_AV could be explained by age, neuropathy and retinopathy; neither blood pressure, glycosylated haemoglobin (HbA_1), nor creatinine clearance were significantly associated with M. Seven measurements of albumin excretion were carried out over 18 months: overnight measurements were more reproducible than daytime ones; the overnight U_A/U_C had the least intra-patient variation. The pattern of M was related to the level of albuminuria at initial screening: established M was commonest in patients with an initial U_AV > 70 μg/min (U_A/U_{C >} 10.2 mg/mmol); those with an initial U_AV ≤10 ug/min (U_A/U_C ≤1.4 mg/mmol) did not develop M > 70 μg/min. In a 4-year prospective study MRA revealed that the initial U_A/U_C and HbA_1 explained 55% of the variability of U_A/U_C at follow-up; a history of hospitalisation and smoking were less important associations. An exercise test for provoking M in IDDs was described. The protocol was based on a change in heart rate and was well tolerated. Patients with a positive test (U_AV > 15 μg/min) had a greater frequency of M and a greater progression of U_A/U_C over 4 years in resting samples. Based on the above findings, practical recommendations are made for screening and monitoring diabetic patients for M.