Reference ranges for natriuretic peptides for diagnostic use are dependent on age, gender and heart rate

Abstract

Background: Plasma natriuretic peptide levels may be useful in the diagnosis of heart failure. The available natriuretic peptide assays differ markedly in their performance characteristics. In addition, plasma levels are influenced by a number of factors including age and gender.

Aims: The aim of this study was to describe, in a healthy population, the influence of clinical and echocardiographic parameters on three natriuretic peptide moieties.

Methods: 1360 individuals were screened for the presence of left ventricular systolic dysfunction. We identified a cohort (n=720) of men aged 45-80 years (n=417) and women aged 55-80 years (n=303). None had history of cardiovascular disease or were taking any cardiovascular medication. All had normal echocardiographic and ECG findings. B-type (BNP), N-terminal pro-B-type (N-BNP) and N-terminal pro-Atrial (N-ANP) natriuretic peptides were assayed using in-house immunoluminometric assays.

Results: Of the considered clinical variables, only gender and heart rate (each P<0.005) were independently associated with levels of all three natriuretic peptides. Plasma levels of N-ANP (15%), BNP (25%) and N-BNP (75%) were higher in women compared to men. An increase in heart rate of 10 bpm corresponded to a reduction of 9% in N-ANP or BNP and a 15% reduction in N-BNP. Each 10 years of age was associated with 16% and 74% increase in ANP and N-BNP levels, respectively, but no increase in plasma BNP. Left ventricular ejection fraction could be assessed in 582 (81%) subjects and correlated positively with N-ANP (r(s)=6.48x10(-3), P<0.001) and BNP (r(s)=2.4x10(-2), P<0.001) but not N-BNP (P=0.405). No parameter of diastolic function was associated with any peptide level.

Conclusions: In this healthy population, plasma levels of N-ANP, BNP and N-BNP were variably influenced by clinical covariates. While all three peptides were higher in women, only N-ANP and N-BNP were influenced by age. Levels of all peptides were inversely correlated with heart rate. Using this immunoluminometric assay, plasma BNP is not influenced by age, in contrast to N-ANP and N-BNP. In constructing normal ranges for diagnostic use, covariates such as age and gender must be considered, in addition to the format of assay being used.