A Prospective Population-Based Study of Microalbuminuria as a Predictor of Mortality in NIDDM

• DOI: 10.2337/diacare.16.7.996
• Publication Date: 2007-03-05T22:51:27Z
• Authors: Andrew Neil; Michael Hawkins; Michael Potok; Margaret Thorogood; David Cohen; Jim Mann
• Keywords: Microalbuminuria; Standardized mortality ratio; Univariate analysis

To assess prospectively the relationship between microalbuminuria and mortality in a geographically defined population of NIDDM patients to determine relative importance as risk factor for mortality.A survey known diabetes undertaken 1982 identified cohort 249 patients. Follow-up information was available 246 who contributed 1498 person-yr exposure were followed up mean period 6.1 yr. The median age at entry 68 yr (range 28-89 yr), duration 7 1-41 yr). At baseline, clinical examination performed random daytime urine specimen obtained measurement urinary albumin concentration.UAC results 236 patients: 45 (19%) had UAC > 15- < 40 mg/L; 36 (15%) 40-200 10 (4%) 200 145 (61%) normal or = 15 mg/L. During follow-up period, 93 died. All-causes mortality, expressed standardized ratio (SMR 149) coronary heart disease (CHD SMR 166) significantly increased. This excess significant women (all-causes 194, CHD 234) but not men 118, 128). On univariate analysis, systolic blood pressure only association with concentration (P 0.0002). An age-stratified log-rank test conducted effect potential explanatory variables on survival. Survival distributions different 0.045), intermittent claudication 0.012), severity retinopathy, lens opacity 0.001) 0.013) diastolic approached significance 0.051). After adjusting effects these potentially confounding by predictors early multivariate survival analysis age, mg/L (relative 2.2, 95% confidence interval 1.3-3.7), more severe retinopathy 3.4, 1.9-6.0), 2.4, 1.6-3.8).The findings from this population-based confirm predictive power NIDDM. In contrast prospective studies conventional cardiovascular factors NIDDM, consistent evidence indicates that is an independent predictor regardless collection procedure used.