Natriuretic peptide levels in atrial fibrillation: a prospective hormonal and Doppler-echocardiographic study.

OBJECTIVES The objective was to determine the independent association between atrial fibrillation (A-Fib) and activation of natriuretic peptides. BACKGROUND The association of A-Fib with activation of N-terminal atrial and brain natriuretic peptides (N-ANPs and BNPs, respectively) is uncertain but of great importance for the diagnostic utilization of natriuretic peptides. This uncertainty is related to the lack of appropriate controls, with left ventricular (LV) and atrial overload similar to A-Fib. METHODS We prospectively measured N-terminal atrial and BNPs and endothelin-1 levels in 100 patients and 14 age- and gender-matched control subjects. The 32 patients with A-Fib were compared with 68 patients in sinus rhythm and similar LV and atrial overload (due to mitral regurgitation or LV dysfunction) measured simultaneously with hormonal levels with comprehensive Doppler echocardiography. RESULTS Patients with A-Fib compared with those in sinus rhythm had similar symptoms, comorbid conditions, cardioactive medications, pulmonary pressure, left atrial volume, and LV ejection fraction and filling characteristics but demonstrated higher N-ANP levels (2,613 +/- 1,681 vs. 1,654 +/- 1,323 pg/ml, p = 0.007) even after adjustment for the underlying cardiac disease (p < 0.0001). Conversely, BNP levels were similar in both groups (165 +/- 163 vs. 160 +/- 269 pg/ml, p = 0.9). In multivariate analysis, a higher N-ANP level was associated with A-Fib (p = 0.0003), symptom class (p < 0.0001) and endothelin-1 level (p = 0.032) independently of left atrial volume and LV ejection fraction. Conversely, BNP showed no independent association with and was most strongly associated with LV ejection fraction (p < 0.0001). CONCLUSIONS Atrial fibrillation is an independent determinant of higher N-ANP levels and blurs its association with LV dysfunction. Conversely, the BNP is not independently associated with A-Fib and is strongly determined by LV dysfunction, for which it is an independent marker.