Small, unrepaired atrial septal defects are considered a benign lesion with good prognosis. Recently, clinical and register-based studies discovered increased long-term mortality and morbidity. The nature of these findings is not fully understood. Therefore, MRI was performed to evaluate cardiac function at rest and during exercise.

Adults with open or spontaneously closed atrial septal defects and healthy, matched controls underwent MRI for evaluation of cardiac chamber volume. Quantitative flow scans measured blood flow in the ascending aorta and the proximal pulmonary artery at rest and during increasing supine exercise.

In total, 15 open defects (39 ± 11 years) and 15 matched controls (38 ± 12 years) were included, along with 20 spontaneously closed (36 ± 13 years) and 20 controls (36 ± 11 years). Cardiac chamber volumes and flow measurements at rest were comparable between groups, as were heart rates and workloads during exercise. At maximal exercise, open defects reached 31% lower cardiac index and had 38% higher retrograde flow in the pulmonary artery than their controls, p < 0.01. Shunt ratio remained unchanged during exercise, 1.2 ± 0.2. Closed defects reached 18% lower cardiac index, p = 0.02, with comparable pulmonary retrograde flow. Maximal cardiac index was inversely correlated with increasing age for patients only.

Adults with a small, open or spontaneously closed atrial septal defects exhibit markedly lower exercise capacity compared with healthy peers. Moreover, open defects exhibit higher retrograde flows with increasing exercise. Finally, increasing age is related to poorer results in patients but not healthy controls. Longitudinal studies are necessary in order to determine potential accelerated worsening of physical capacity along with age-related changes in patients.

Clinical studies have revealed decreased exercise capacity in adults with small, unrepaired ventricular septal defects. Increasing shunt ratio and growing incompetence of the aortic and pulmonary valve with retrograde flow during exercise have been proposed as reasons for the previously found reduced exercise parameters. With MRI, haemodynamic shunt properties were measured during exercise in ventricular septal defects.

Patients with small, unrepaired ventricular septal defects and healthy peers were examined with MRI during exercise. Quantitative flow scans measured blood flow through ascending aorta and pulmonary artery. Scans were analysed post hoc where cardiac index, retrograde flows, and shunt ratio were determined.

In total, 32 patients (26 ± 6 years) and 28 controls (27 ± 5 years) were included. The shunt ratio was 1.2 ± 0.2 at rest and decreased to 1.0 ± 0.2 at peak exercise, p < 0.01. Aortic cardiac index was lower at peak exercise in patients (7.5 ± 2 L/minute/m2) compared with controls (9.0±2L l/minute/m2), p<0.01. Aortic and pulmonary retrograde flow was larger in patients during exercise, p < 0.01. Positive correlation was demonstrated between aortic cardiac index at peak exercise and previously established exercise capacity for all patients (r = 0.5, p < 0.01).

Small, unrepaired ventricular septal defects revealed declining shunt ratio with increasing exercise and lower aortic cardiac index. Patients demonstrated larger retrograde flow both through the pulmonary artery and the aorta during exercise compared with controls. In conclusion, adults with unrepaired ventricular septal defects redistribute blood flow during exercise probably secondary to a more fixed pulmonary vascular resistance compared with age-matched peers.