Low protein intake in the population: low risk of kidney function decline but high risk of mortality.

OBJECTIVE: This population-based study investigated low protein intake, mortality, and kidney function decline. DESIGN: Observational longitudinal cohort study. SUBJECTS: Target cohort consisted of 4,679 adults participating in 1988-1992 and 2001-2007 examinations of the Gubbio Study (baseline and follow-up). Data collection included overnight urine urea nitrogen (UUN) and other variables at baseline, serum creatinine at baseline and follow-up, and mortality from baseline to follow-up. Three hundred seventy-two persons were excluded for missing data. UUN in the lowest 20% of the distribution was defined as low and used as index of low protein intake. Estimated glomerular filtration rate (eGFR, mL/minute × 1.73 m2) was used as kidney function index. INTERVENTION: None (observational study). MAIN OUTCOME MEASURE: Mortality and eGFR decline are the main outcome measures, and eGFR decline was defined as eGFR change from baseline to follow-up ≤ mean-1 standard deviation (Z-score ≤ -1). RESULTS: Eight hundred seventy-one deaths occurred over 15.9 ± 4.0 years of observation (417 from cardiovascular disease and 276 from neoplastic disease). Low UUN associated with mortality (hazard ratio, HR = 1.31, 95% confidence interval, CI = 1.12/1.53) due to association with mortality from neoplastic disease (HR = 1.33, 95% CI = 1.02/1.76). Mortality-corrected follow-up response rate was 79.9% (n = 2845). Baseline to follow-up eGFR change was -9.9 ± 10.1, and eGFR decline was found in 454 examinees. Low UUN associated with eGFR decline only in subgroup with baseline eGFR <90 (n = 1441, odds ratio = 0.44, 95% CI = 0.22/0.85). Low baseline eGFR interacted with the association between low UUN and eGFR decline (P = .024). CONCLUSION: Low protein intake predicted higher mortality in the whole population and lower incidence of eGFR decline only in subgroup with reduced kidney function.