Do non-antiarrhythmic drugs have enough pleiotropic power to reduce atrial fibrillation?