Pediatric pulmonary hypertension (PH) is associated with various vascular, cardiac, pulmonary, and systemic conditions. Without treatment, the disease leads to right ventricular dysfunction, right ventricular failure, and death. While therapies have improved in recent years (1, 2), long-term outcomes remain poor. Patients report low quality of life in the face of significant morbidity and mortality risk (3, 4). Exercise intolerance is common in PH patients, significantly impacting quality of life and prognosis. While cardiopulmonary status affects exercise performance, the contribution of peripheral factors, including skeletal muscle dysfunction, is increasingly recognized (5). Skeletal muscle dysfunction is associated with worse 6-min walk distance (6 MWD) in adults with idiopathic pulmonary arterial hypertension (IPAH) (6). Similarly, we previously reported marked deficits in densitometry (DXA)-measured leg lean mass, a surrogate marker of skeletal muscle, in youth with PH in association with inactivity and worse 6 MWD (7). In other studies, skeletal muscle atrophy, impaired peripheral oxygen extraction, and reduced muscle contractility suggest that PH patients exhibit a generalized “myopathy” similar to patients with left heart failure (8–11). The skeletal muscle pump is critically important to augmentation of systemic venous return and pulmonary blood flow with erect exercise (12). Skeletal muscle dysfunction may critically hinder this mechanism in PH patients. Therefore, we sought to expand on our prior findings by characterizing leg lean mass Z-score (LLMZ) and muscle strength in a different cohort of pediatric PH patients and to explore the associations between muscle deficits, self-reported physical activity, and measures of exercise performance on cardiopulmonary exercise test (CPET) and exercise cardiac magnetic resonance (eCMR).

Demographic and clinical characteristics of the 25 participants are displayed in Table 1. The cohort was 56% female and majority white. Mean height Z-score was –0.24 ± 1.24 while mean BMI Z-score was 0.19 ± 0.93. Most patients (64%) were classified as WSPH Group 1 (pulmonary arterial hypertension), but 24% of patients were classified as WSPH group 3 (PH due to lung disease). Of the 16 participants in WSPH Group 1, there were 10 with idiopathic or heritable PAH and 6 with PAH after repair of congenital heart disease. No participants with congenital heart disease had unrepaired cyanotic lesions or significant residual shunts. More patients were WHO functional class I but patients in functional class II enrolled as well. The most common medications were tadalafil and ambrisentan. Twenty-eight percent of the cohort met criteria for vitamin D deficiency. Mean brain natriuretic peptide and hemoglobin levels were within normal limits for our laboratory. Mean tricuspid annular plane systolic excursion Z-score was –2.5 ± 3.2 reflecting decreased right ventricular function. Participants achieved an average 6 MWD of 636 ± 113 meters reflecting good functional status. Cardiac catheterization data were available in 23 participants (90%) at median 1.42 years prior to the study visit. Mean pulmonary artery pressure was 33.1 ± 12.0 mm Hg and indexed pulmonary vascular resistance was 6.2 ± 3.0 indexed Wood units. PAQ scores were lower than reported for healthy populations. Average quality of life score was less than 70 for both participants and parents/guardians.

Skeletal muscle mass and strength data are shown in Table 2. Mean LLMZ was markedly decreased at –0.96 ± 1.14, equivalent to the 17th percentile. Dominant (D-HGZ) and non-dominant (ND-HGZ) were also very low with mean D-HGZ of –1.33 ± 1.37 equivalent to the 9th percentile.

Table 3 includes CPET and eCMR data. Baseline oxygen saturation was 97.7 ± 1.5% with decrease to 91.5 ± 8.3% (range 71–100) with CPET. On CPET, participants performed lower than predicted compared to healthy reference data. Percent predicted peak VO₂ was approximately 71% while percent predicted VO₂ at VAT was approximately 80%. Participants only achieved a peak work rate 59% predicted for demographics and body size. Reference data are not available for work rate at VAT. Resting and exercise CMR data are displayed. Right and left ventricular cardiac index and systemic blood flow increased with exercise.

LLMZ was associated with disease-specific factors. LLMZ was positively associated with hemoglobin (r = 0.44, p = 0.03) and negatively associated with brain type natriuretic peptide (r = –0.40, p = 0.05), mean pulmonary artery pressure (r = –0.54, p < 0.01), and indexed pulmonary vascular resistance (r = –0.50, p = 0.02) but was not associated with tricuspid annular plane systolic Z-score. LLMZ was higher in those with higher PAQ scores (r = 0.50, p = 0.01).

LLMZ, muscle strength, and PAQ score were associated with performance on CPET and eCMR (Table 4). LLMZ, both D-HGZ and ND-HGZ, and PAQ score were associated with percent predicted peak VO₂, percent predicted VO₂ at VAT, and percent predicted work (Figures 1–3). LLMZ was associated with O₂ pulse on CPET (r = 0.41, p = 0.04) while D-HGZ was not. D-HGZ was associated with most recent 6 MWD (r = 0.44, p = 0.03). D-HGZ, ND-HGZ, and PAQ score were also associated with change in systemic flow from rest to exercise on eCMR, while LLMZ was not (r = 0.51, p = 0.1).

The associations between lower extremity strength by Biodex and exercise performance were variable. Ankle extension was associated with change in systemic flow from rest to exercise (r = 0.69, p = 0.04). Both knee extension and flexion were associated with peak work (not percent predicted) (r = 0.52, p = 0.02 for extension; r = 0.48, p = 0.03 for flexion), possibly because Z-scores are not available for the Biodex variables.

Results from multiple linear regression models testing the effect of skeletal muscle and physical activity variables on exercise performance are shown in Table 5. LLMZ and PAQ score were positively associated with peak VO₂, while handgrip Z-score and PAQ score were positively associated with peak work rate. No factors were associated with the change in systemic flow from rest to exercise on eCMR.

There were no differences in LLMZ, PAQ, CPET, or eCMR findings in participants with Group 2 PH vs. other classifications. The study findings were unchanged when participants with Group 2 PH were excluded. There were some differences in LLMZ and exercise performance based on medication regimen. Participants on subcutaneous treprostinil had lower LLMZ (–2.02 ± 1.03 vs. –0.76 ± 1.06, p = 0.04), percent predicted VO₂ (0.49 ± 0.11 vs. 0.75 ± 0.21, p = 0.02), percent predicted VO₂ at VAT (0.54 ± 0.13 vs. 0.86 ± 0.17, p < 0.01), and percent predicted work (0.41 ± 0.10 vs. 0.63 ± 0.20, p = 0.05) compared to those not on subcutaneous treprostinil. Interestingly, participants on oral treprostinil demonstrated greater change in systemic flow with exercise [4.56 ± 1.56 (n = 3 with cCMR) vs. 2.02 ± 0.85 (n = 7 with eCMR), p < 0.01] and higher PAQ (2.3 ± 0.6 vs. 1.8 ± 0.6, p = 0.06), D-HGZ (–0.30 ± 1.62 vs. –1.67 ± 1.14, p = 0.03), and ND-HGZ (–0.02 ± 1.74 vs. –1.45 ± 1.28, p = 0.04) compared to those participants not on oral treprostinil. Participants on ambrisentan demonstrated lower percent predicted VO2 at VAT (0.73 ± 0.22 vs. 0.91 ± 0.13, p = 0.04) compared to those not on ambrisentan.

No adverse events occurred with exercise testing.

In this study, we demonstrated marked deficits in skeletal muscle mass and strength in association with worse exercise performance on CPET and eCMR in youth with PH. Lower skeletal muscle mass was associated with physical inactivity. In regression models, skeletal mass, strength, and physical activity were positively associated with exercise performance on CPET. This study builds on our prior work in which lower LLMZ was associated with lower 6 MWD in youth with PH (7). These pilot data add to our appreciation of musculoskeletal abnormalities in pediatric PH and the potential impact of these deficits on exercise performance. Increased understanding of the peripheral determinants of exercise performance may identify targets for intervention trials.

The general “myopathy” seen in adult PH patients is an area of active investigation, but few studies have examined this issue in pediatric PH. PH is increasingly understood to be a systemic condition with metabolic, inflammatory, genetic, and epigenetic contributions (42). As one of the extra-cardiopulmonary manifestations of this condition, skeletal muscle dysfunction can be grouped into structural deficits, functional impairment, and molecular abnormalities with evidence from both human and animal studies (42). Structural deficits include reduced muscle fiber cross sectional area, lower proportion of type I fibers, decreased capillary density, lipid inclusion, and loss of mitochondrial structure (8, 10, 43). Functional impairment can manifest as reduced muscle strength, decreased muscular endurance, impaired oxygenation in the microcirculation, reduced type I fiber tension, and sarcomeric dysfunction (6, 8, 9, 44, 45). Molecular abnormalities have been described including reduced oxidative capacity, increased protein degradation and decreased protein synthesis, impaired angiogenesis, and impaired mitochondrial function (11, 46, 47). The spectrum of muscular abnormalities in children with PH is unknown. As pediatric PH often occurs in the context of congenital heart disease, developmental lung diseases, genetic differences, and other syndromic conditions, the findings may be even more severe. To our knowledge, no other investigators have explored decreased muscle mass and strength in children with PH. The data are essential to understanding the burden of disease across the lifespan and identifying appropriate timepoints for intervention. Future studies should explore the breath of these findings.

We demonstrated associations between muscle deficits and performance on CPET. CPET is a comprehensive assessment of a patient’s exercise performance that utilizes inspiratory and expiratory gas exchange to quantify peak oxygen consumption, carbon dioxide production, and minute ventilation (48). Peak VO₂ is the most common indicator of a patient’s cardiorespiratory fitness. Adult patients with PH demonstrate decreased peak VO₂, higher VE/VCO₂ (ventilatory equivalents of carbon dioxide) indicating ventilatory inefficiency, lower arterial CO₂ tension and end-tidal CO₂ tension, lower O₂ pulse (VO₂/heart rate, a surrogate for stroke volume), and lower systemic oxygen saturation (49–51). Worse peak VO₂ is associated with more symptoms of dyspnea and fatigue which limit quality of life. Both low peak VO₂ and low 6 MWD are associated with mortality in adults with PAH and are incorporated into REVEAL 2.0 and European Society of Cardiology/European Respiratory Society risk stratification tools (52, 53). However, completion of a CPET requires significant developmental skills and reference values relative to outcomes are lacking in pediatric PH (54, 55). The lack of adverse events in this pilot study suggests that select older children and teenagers with PH can safely perform CPET and eCMR. Peak VO₂ is included in pediatric PAH disease severity guidelines (56) and perhaps exercise parameters will be incorporated into novel pediatric risk stratification tools which are being developed. To our knowledge, this is the first study to express PH participants’ CPET performance relative to reference values from more than 1,800 healthy United States children published by members of our study team (28). Participants in our study generally performed at 60–80% predicted, suggesting children with PH already face marked limitations prior to progressive decline in adulthood. Additionally, this is the first report of eCMR in pediatric PH, although we have previously described eCMR in patients with single ventricle physiology after Fontan palliation (21). Higher PAQ and handgrip Z-scores were associated with greater change in systemic flow from rest to exercise, raising compelling questions regarding physical activity, strength, and the ability to augment systemic flow on exertion. Future, larger studies should also explore the correlation between eCMR performance and PH disease-specific outcomes.

While peak VO₂ is a common outcome reported in research and clinical care, CPET measures of submaximal performance including VO₂ at VAT and peak work rate relative to respiratory exchange ratio as well as measures of anaerobic strength may better reflect a patient’s daily activities and may also be targets for improvement. PH patients rapidly reach anaerobic threshold when performing activities of daily living such as doing laundry and folding clothes, cooking and setting the table, and walking outside with moderate effort (57, 58). These activities are often performed with short bursts of activity. The associations between muscle deficits, VO₂ at VAT, and peak work rate in our study suggest that skeletal muscle deficits may impact everyday activity performance for PH patients and improvements in muscle mass or function could improve quality of life and other outcomes over longer periods. These findings are also consistent with our previous report associating lower LLMZ with lower 6 MWD, another test of submaximal exercise performance (7).

Cardiopulmonary rehabilitation improves symptoms and functional status in adult PH but the mechanisms underlying those improvements are incompletely understood. Some studies have demonstrated improvement in functional status and exercise performance without change in pulmonary hemodynamics. Data such as those from Bauer et al. (6) and Mainguy et al. (10) have supported development of exercise rehabilitation programs to improve muscle function and strength. In a separate study, Mainguy demonstrated increased 6 MWD, increased exercise duration on CPET, and improved VE/CO₂ in association with decreased number of quadriceps type IIx fibers in 5 adult IPAH patients after a 12-week rehabilitation program (59). De Man also demonstrated increased quadriceps strength and endurance, increased quadriceps capillarization, and higher oxidative enzyme activity in 19 adult IPAH patients after a 12-week rehabilitation program (60). There is only one published report of exercise training in pediatric PH. Zöller et al. demonstrated increased treadmill running distance and improved VO₂ at VAT after a 16-week home exercise program in 9 children with PH and functional class I and II (61). Skeletal muscle characterization was not part of that study. While PH patients perform less moderate to vigorous physical activity compared to healthy peers (62), no studies have tested interventions to improve activity. To our knowledge, no pediatric PH exercise trials have described changes in physical activity and/or skeletal muscle performance with training. Future studies should determine whether low skeletal muscle mass is a marker of disease severity or a modifiable determinant of exercise performance in youth with PH. It is not known whether increasing physical activity can augment skeletal muscle characteristics and improve exercise performance in youth with PH.

A primary limitation of this study is the lack of control group. It is not clear whether the relationships between LLMZ, PAQ score, and exercise performance are different in PH patients vs. healthy youth. Additionally, causal inferences cannot be made from these data given the study’s cross-sectional design. The small number of participants with certain medication regimens does not allow us to draw conclusions regarding relationships between medications, underlying disease status, and the study outcomes. The relationship between disease progression, muscle deficits, and exercise performance is also unknown. The study was also limited by lack of comparison reference data for lower extremity strength. The lack of association between leg strength and exercise performance may have been because we were unable to express strength relative to gender and body size. Finally, physical activity was measured by self-report. The analyses would have been strengthened by quantification of physical activity by wearable accelerometer. However, despite these limitations, this study has generated preliminary data for future observational studies and interventional trials.

Youth with PH demonstrate inactivity and marked skeletal muscle deficits in association with worse exercise performance on CPET and eCMR. Future studies should determine whether low skeletal muscle mass is a marker of disease severity or a modifiable target for exercise interventions. Interventions that improve skeletal muscle mass and function could improve exercise performance in this population.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by the Children’s Hospital of Philadelphia Institutional Review Board. Written informed consent to participate in this study was provided by the participants or their legal guardian/next of kin.

CA designed the study, recruited participants, supervised study procedures, analyzed and interpreted data, and wrote the manuscript. MM and SP supervised exercise testing procedures, interpreted exercise data, and reviewed the final manuscript. MH, KW, and MF supervised exercise cardiac magnetic resonance procedures, analyzed and interpreted the cardiac magnetic resonance data, and reviewed the final manuscript. BZ analyzed and interpreted anthropometry and densitometry data and reviewed the final manuscript. All authors contributed to the article and approved the submitted version.