Blunted stroke volume response to exercise in adolescent children born premature. A Cardiac MRI Study

In order to meet the increased metabolic demand imposed by exercise, heart rate (HR) increases linearly throughout incremental exercise, and stroke volume (SV) increases in a curvilinear fashion, with most of the augmentation occurring between rest and 50% VO 2max . In healthy children, the SV augmentation from rest to maximal exercise is not well documented due to the invasive nature of gold standard cardiac output (Q) measurements, though it has been shown that healthy children augment their stroke volume from rest to submaximal and maximal exercise. We sought to determine the SV response to submaximal and maximal exercise in adolescent children born premature using a bioimpedance technique and cardiac magnetic resonance imaging (cMRI), and to validate a thoracic bioimpedance technique (Physioflow) in children by comparing resting and 70% Pmax SV between cMRI and Physioflow. Methods Seven children born preterm (PT) (age 12–13, birthweight <1500 g, gestational age 24–31 weeks) and 6 age‐matched children born full term (CT) (gestational age 38–40 weeks) underwent incremental maximal exercise testing on a cycle ergometer, with continuous measurement of oxygen consumption and HR with a metabolic cart, and cardiac output (Q) and SV were determined using thoracic bioimpedance. SV and Q were indexed to body surface area (BSA (m 2 ), SVi and Qi, respectively). One hour later, the children performed supine exercise on an MRI‐compatible stepper at 70% of their maximal power (70%Pmax) attained on the upright bike. SV at rest and 70%Pmax from the cMRI was done using Segment software, obtaining flow measurements through the aorta. Statistical analysis was done using t‐tests. Results Using thoracic bioimpedance, there was no difference in SVi at rest between PTs and CTs (51.2 ± 8.4 v 47.9 ± 4.3 ml/m 2 , p=0.369, respectively), but PTs had significantly lower SVi at 70%Pmax than CTs (55.4 ± 9.2 v 71.8 ± 13.4 ml/m 2 , p=0.015) and at max (50.5 ± 9.4 v 68.9 ±14.7 ml/m 2 , p=0.012). In the cMRI, PT SVi did not differ significantly from CTs at rest (47.4 ± 4.0 v 48.6 ± 3.8 ml/m 2 , p=0.73, respectively), and at 70%Pmax PTs had a significantly lower SVi than CTs (47.6 ± 5.7 v 59.0 ± 4.8 ml/m 2 , p=0.04, respectively). In both measurement techniques, the change in SV from rest to submaximal or maximal exercise was not significant in PT, but significant in the CT. There were no differences between resting and maximal HRs between groups. PT had lower relative VO 2max compared to CT (38.3 ± 9.3 v 51.5 ± 7.3 ml/kg/min, p=0.03). The calculated arterio‐venous O 2 difference (a‐vO 2 ) using the Fick equation was not different in PT compared to CT (13.2 ± 6.5 v 13.2 ± 2.3 ml/dL, p=0.98). Discussion The finding that SVi does not augment from rest to submaximal and maximal exercise in adolescent children born preterm was confirmed using two techniques, a thoracic bioimpedance method and cardiac MRI. While term‐born children exhibited a SV profile from rest to submaximal to maximal exercise consistent with what others have shown in the literature, the profile in preterm children was consistent with that of children with cardiac disease. We were able to demonstrate that the lower VO 2max in preterm children is driven by the inhibited SV response to exercise by showing that the arteriovenous oxygen difference was nearly identical between the two groups at maximal exercise. We have demonstrated that adolescents born preterm have lasting cardiac effects that limit their aerobic capacity, and may have other detrimental effects. Support or Funding Information National Institutes of Health: 1R01 HL086897 (M.W.E.) and R01 HL38149 (M.P.), and UW CVRC T32‐HL 07936.