Urine protein/creatinine ratio as a mortality risk predictor in non-diabetics with normal renal function.

Objective: Determine the relative mortality in apparently healthy adults with various levels of urinary protein measured by urine protein/creatinine (p/c) ratio.

Method: By use of the Social Security Death Master File, mortality in 2010 was determined for 7.5 million life insurance applicants age 20 to 89 providing urine samples between 1992 and 2006. Relative mortality by Cox regression for bands of p/c ratios was determined using age and sex as covariates and with an age split at 60 after excluding those with hematuria (> 3 red cells/hpf), diabetes, evidence of blood sugar elevation, or eGFR < 60 mL/min.

Results: After the exclusions noted above, relative mortality increased to 160% beginning at a p/c ratio of 0.11 mg/mg and rose steadily above that value regardless of sex and age. Most of this risk was not explained by a history of hypertension or elevated systolic blood pressure. Albumin testing identified roughly a third of urine samples with elevated p/c ratios as not containing albumin; those cases appeared to be associated with much lower risk as long as the p/c ratio was < or = 1.0 mg/mg.

Conclusion: Low levels of proteinuria identified as urine protein/creatinine ratios of 0.11 mg/mg or higher (much lower than the usual lower cut-off value of 0.21) are associated with substantial excess mortality risk, even after excluding diabetics and those with reduced kidney function or hematuria.