Albumin to creatinine ratio: a screening test with limitations.

Abstract

The aim of this study is to assess the effects of age on (1) the ability of a spot albumin-creatinine ratio (ACR) to accurately predict 24-hour albumin excretion rate (AER), and (2) the performance of spot ACR as a screening test for microalbuminuria. Three hundred fourteen patients with diabetes aged 18 to 84 years attending a tertiary outpatient clinic underwent one 24-hour urine collection and, immediately after completion, provided one fasting spot morning urine sample. Twenty-four-hour AER and spot ACR were determined. Performance of spot ACR was assessed according to age and sex. Fifty-three percent of men and 32% of women had an AER of 20 microg/min or greater. Multiple regression analysis showed age was an independent predictor of spot ACR. For an AER of 20 microg/min for patients in the age range of 40 to 80 years, there was an increase in corresponding values for spot ACR from 18.2 mg/g (95% confidence interval [CI], 15.6 to 21.3) to 32.5 mg/g (95% CI, 27.5 to 38.4) in men and from 22.1 mg/g (95% CI, 18.0 to 27.1) to 56.4 mg/g (95% CI, 47.2 to 67.4) in women. Using ACR cutoff values of 22.1 mg/g or greater and 30.9 mg/g or greater in conventional units (equivalent to > or =2.5 and > or =3.5 mg/mmol in SI units) in men and women, the spot ACR provided high sensitivities (men, 95.7%; women, 93.35%) and had excellent receiver operator characteristic curves, respectively. However, the spot ACR false-positive rate increased with age from 15.9% (age, 40 to 65 years) to 31.8% (>65 years) in men and from 10.5% (age, 45 to 65 years) to 28.3% (>65 years) in women. Spot ACR is a good screening test for microalbuminuria, but a poor predictor of quantitative AER, and should not be used as a diagnostic test. The increase in spot ACR relative to 24-hour AER with age supports the use of sex- and age-adjusted ACR cutoff values.