Aortic stenosis and mitral regurgitation as predictors of atrial fibrillation during 11 years of follow-up
1 Department of Cardiology, Malmö University Hospital, Malmö, Sweden
2 Department of Clinical Sciences, Malmö University Hospital, Malmö, Sweden
3 Department of Cardiology and Center for Integrative Electrocardiology, Lund University (CIEL), Lund, Sweden
4 Department of Cardiology, Lund University and Heart Health Group, Malmö, Sweden
BMC Cardiovascular Disorders 2012, 12:92 doi:10.1186/1471-2261-12-92Published: 18 October 2012
There is limited information about any association between the onset of atrial fibrillation (AF) and the presence of valvular disease.
We retrospectively examined 940 patients in sinus rhythm, examined by echocardiography in 1996. During 11 years of follow-up, we assessed the incidence of AF and outcome defined as valvular surgery or death, in relation to baseline valvular function. AS (aortic stenosis) severity at baseline examination was assessed using peak transaortic valve pressure gradient.
In univariate analysis, the risk of developing AF was related to AS (significant AS versus no significant AS; hazard ratio (HR) 3.73, 95% confidence interval (CI) 2.39-5.61, p<0.0001) and mitral regurgitation (MR) (significant MR versus no significant MR; HR 2.52, 95% CI 1.77-3.51, p<0.0001). Also the risk of valvular surgery or death was related to AS (HR 3.90, 95% CI 3.09-4.88, p<0.0001) and MR (HR 2.07, 95% CI 1.67-2.53, p<0.0001). In multivariate analyses, adjusting for sex, age, other valvular abnormalities, left ventricular ejection fraction and left atrial size − AS was independently related to both endpoints, whereas MR was not independently related to either endpoint.
AS, but not MR, was independently predictive of development of AF and combined valvular surgery or death. In patients with combined AS and MR, the grade of AS, more than the grade of MR, determined the risk of AF and combination of valvular surgery or death. Further studies using contemporary echocardiographic quantification of aortic stenosis are warranted to confirm these retrospective data based on peak transaortic valve pressure gradient.