Abstract

Objetive: This study was designed to quantify the crude and adjusted effects of estimated glomerular filtration rate (eGFR) on N-terminal-pro-brain-natriuretic peptide (proBNP) measured with three immunoassays and brain natriuretic peptide (BNP) in elderly individuals.

Design: Cross sectional study.

Setting: Four-hundred-seventy-four elderly outpatients with suspected heart failure (prevalence: 13 %) from the primary care.

Main outcome measures: The effects of eGFR on proBNP, measured with three different immunoassays (Roche Diagnostics, Oslo and Copenhagen), and BNP (Shionogi) concentrations were evaluated by multiple linear regression models.

Results: In univariate analyses the effect of a 10 % decrease in eGFR on proBNP concentrations was a 15 % (95 % Confidence Interval: 11-18 %), 9 % (5-13 %) and 21 % (14-28 %) increase. In multivariate models the effect was a 7 % (3-11 %), 4 % (2-6%) and 13 % (4-20%) increase. The effect of a 10 % decrease in eGFR on BNP concentrations (Shionogi) was a 10 % (5-15 %) (univariate) and a 4 % (1-9 %) (multivariate) increase.

Conclusions: The effect of eGFR on proBNP measured with three different immunoassays and BNP is modest and within the same range. The effect of eGFR on proBNP and BNP concentrations is reduced substantially after adjustment for important clinical and echocardiographic confounders. These findings should be considered before renal function per se is offered as an explanation for increased proBNP or BNP levels.