Introduction: Diastolic dysfunction (DDFx) was shown to contribute to heart failure symptoms. The standard echocardiographic measures of diastolic function cannot be applied to patients in atrial fibrillation (AF) during the echocardiographic examination. Left ventricular/atrial speckle-tracking echocardiography (STE) may provide rhythm independent clues to diastolic dysfunction. We aimed at finding the STE predictors of dyspnea in patients presenting in AF for their echocardiographic examination.
Methods: Forty-four echocardiographic exams in patients with AF were analyzed retrospectively by STE. Various patients’ demographic and clinical characteristics were collected. Each patient underwent a thorough echocardiographic examination that included left ventricular/atrial speckle-tracking echocardiographic evaluation. The different echocardiographic and myocardial mechanics characteristics were also collected into the database. Predictors of dyspnea were analyzed in the multivariate logistic regression model.
Results: Patients in AF during the study were 72.7±13.9 years old, 62.8% male, with average BMI of 30±7.1. Prevalence of hypertension was 37.8%, diabetes mellitus 16.1%, hypercholesterolemia 25% and 7.8% smoked. Dyspnea was reported by 14/44 patients (31.8%). There was no difference in age, gender or prevalence of different co-morbidities between AF patients with or without dyspnea complains. None of the patient characteristics and various STE parameters were found to correlate with dyspnea in the study population except increased end diastolic left ventricular volume (EDV) and conduit volume (CV) that were found to predict dyspnea in patients with AF during the exam (p<0.05 for both). These variables didn’t remain independent predictors in a multivariate logistic regression analysis when corrected for each other.
Conclusions: Patients characteristics as well as most of the echocardiographic parameters didn’t correlate with dyspnea in patients presenting in AF except increased EDV and CV that predicted heart failure symptoms in a univariate analysis but not in a multivariate regression model.