Pro-atrial natriuretic peptide and prediction of atrial fibrillation and stroke: The Malmö Preventive Project.

Background The increasing prevalence of atrial fibrillation and novel therapeutic tools to prevent cardioembolic stroke has increased the need for risk markers. Objectives This study explored the relationship between the midregional sequence of pro-atrial natriuretic peptide (MR-proANP) levels with the risk of atrial fibrillation and stroke, and whether measurement of MR-proANP improves the prediction of these outcomes. Methods MR-proANP was measured in fasting blood samples of 5130 subjects (69% men, mean age 69.2 ± 6.2 years) without a history of atrial fibrillation or stroke from the general population. The incidence of atrial fibrillation and stroke was monitored over a median follow-up of 5.6 years. C-statistics and net reclassification improvement was used to assess the predictive ability of MR-proANP in addition to conventional risk factors. Results Log-normalized MR-proANP was significantly associated with the incidence of atrial fibrillation ( n = 362; hazard ratio (HR); 95% confidence interval (CI) per 1 standard deviation (SD) 2.05, 1.86-2.27) and stroke from all causes ( n = 195; HR 1.30; 95% CI 1.12-1.50). The HR for stroke events related to atrial fibrillation was 1.79 (95% CI 1.25-2.58) per 1 SD. MR-proANP significantly improved the prediction of atrial fibrillation when added to a risk score of conventional risk factors (C statistic 0.69 vs. 0.75), mainly by down-classifying subjects who did not develop atrial fibrillation. A smaller improvement in predictive ability was observed for stroke (C statistic 0.66 vs. 0.68). Conclusion High plasma levels of MR-proANP are associated with the incidence of atrial fibrillation and stroke in the middle-aged and elderly population. MR-proANP may be useful to identify individuals with an increased risk of atrial fibrillation.