When two Z-scores meet—analysis of exercise capacity of children and adolescents with Kawasaki disease by a new Z-score model of coronary artery and a new Z-score evaluating peak oxygen consumption

Kawasaki disease (KD), which was first described in 1967 by Dr. Tomisaku Kawasaki [1], is now the most common form of pediatric vasculitis in children [2]. After Japan and Korea, Taiwan has the third-highest incidence of KD in the world (69 in 1,00,000 children aged < 5 years) [3]. The condition is characterized by systemic inflammation in medium-sized arteries, multiple organs, and tissues during the acute febrile phase, with a predilection for the coronary artery (CA) [4]. An estimated 25% and 5% of untreated and treated children with KD, respectively, may develop CA aneurysms (CAAs) [5]. Although 80% of the CAAs tend to regress within 5 years after the onset of KD, 1% of the CAAs eventually lead to progressive arterial stenosis and thrombosis [6]. These cardiac sequelae may cause myocardial infarction or sudden death in the acute phase, post-acute phase, or even in adulthood [7]. Therefore, both the American Heart Association (AHA) [2] and the Japanese Ministry of Health and welfare [8] recommend routine echocardiographic coronary examination for children with KD to evaluate the presence of CAAs.

Earlier, CAAs were diagnosed based on the CA size as per the Japanese Ministry of Health and Welfare recommendations [9]. These criteria were controversial due to the lack of a correlation with body habitus and the non-differentiation between the right and left CAs. The CA Z-score system, describing how many standard deviations above or below a size or age-specific population mean a given measurement lies [10], can allow more efficient discrimination of CAAs and show better correlation with clinical outcomes [11]. CA Z-score system is derived from a large heterogeneous population of children undergoing echocardiography. Several different CA Z-score regression Eqs. [11], both linear [12] and exponential [13] functions with body surface area (BSA), have been proposed to provide an objective basis for determining CA size abnormalities. In Taiwan, Lin et al. established reference ranges for the CA diameters after evaluating a nationwide cohort of 412 healthy children aged < 6 years [10]. However, there is a lack of available norms for CA for Taiwanese children aged > 6 years. A well-known and acceptable [14, 15] CA Z-score equation for children aged from 0 month to 18.9 years has been established by Kobayashi et al. to estimate the sex-specific Z-score of each internal CA diameter (ZSP version 4) [16]. This equation showed better goodness of fit compared to previously reported regression models. Our team had also used it in our previous study and found that KD children with higher CA Z-score had significantly lower peak exercise capacity (EC) than those with lower CA Z-score [17]. Therefore, we chose ZSP version 4 for our analysis.

In previous studies based on the findings of direct cardiopulmonary exercise testing (CPET), children with a history of KD showed comparable EC [18‐20] but lower myocardial flow reserve and higher total coronary resistance compared to their healthy peers [20]. However, adolescents with KD history had significantly lower aerobic metabolism capacity and peak exercise load capacity than their peers [21, 22]. Although, most of the above-mentioned studies used peak oxygen consumption (peak VO₂), the current gold standard marker of exercise capacity, there are still some limitations that should be addressed when using peak VO₂. Peak VO₂ may vary with age, maturity, and sex. It is developmentally divergent and shows a strong correlation with body size and body composition [23]. Conventionally, we scale peak VO₂ by simply dividing it (mL/min) by body mass (mL/kg/min) to obtain peak metabolic equivalent (peak MET). However, the peak VO₂ should be scaled for size using the general allometric equation to derive the appropriate size power function (y/xb), thereby providing a more appropriate interpretation of size-related (xb) changes in physiologic function (y) [24]. Recently, a new Z-score equation based on allometric scaled peak VO₂ values was developed for Southern Chinese children and adolescents (Peak VO2 Z-score). Peak VO₂ Z-score improves the evaluation of cardiopulmonary fitness, allowing comparisons across ages and sex and will likely provide a better metric for tracking temporal changes in children and adolescents, regardless of body size and age [25]. Therefore, by combining the two Z-scores together, we aimed to compare the difference of cardiopulmonary function indicated by peak VO₂ Z-score between KD children and adolescents with different CA Z-score in our study.

One hundred twelve patients with KD qualified the inclusion criteria. Of these ten patients were excluded (one each with moderate and severe cardiac valvular disease, two with significant arrhythmia, and six with incomplete pulmonary function test/ PCET data). Therefore, 102 children and adolescents with KD were included in this study. Among them, 87 (85.3%) had CA Z-score < 2.5 (KD group 1) and 15 (14.7%) had CA Z-score ≥ 2.5 (KD group 2).

To the best of our knowledge, this is the first study to compare the peak VO₂ Z-score between KD children and adolescents with different CA Z-scores. We found that after the acute stage of KD, although most KD children and adolescents were able to exert sufficient effort to reach peak performance during CPET, there was still mildly decreased peak PD% as compared to normal reference. We also observed that KD children and adolescents with MAX-Z < 2.5 had higher peak EC, including peak VO₂ Z-score, peak PD%, peak MET, and PRPP compared to those with MAX-Z ≥ 2.5.

Few studies have investigated the exercise performance and aerobic capacity in patients with KD. Allen et al. used CPET with leg ergometer to evaluate the performance of KD patients. They found no difference in total work rate, mean power, maximal HR, and maximal VO₂ between KD children and control subjects [33]. In the study by Rhodes et al. (1996), participants performed the CPET via leg ergometer and they found that KD patients had similar maximal VO₂, peak workload, and AT compared to the control participants [34]. Wang et al. (2008) observed that KD children had similar maximal HR but lower maximal VO₂ and maximal SBP compared to healthy peers during the CPET with treadmill [34]. Our previous study also found no significant difference in the aerobic metabolism and peak exercise load capacities of KD children and control group [20]. However, the healthy controls in all of the studies mentioned above were collected from previously reported normal participants from the database of each institution, matched for sex, age, and BMI, or BSA. Therefore, the results may have been affected by potential selection and population bias. In the present study, the mean peak PD% of all the KD children and adolescents was 90.11 ± 13.33%. Given that peak PD% was the percentage of measured peak MET to predicted peak MET by an age- and sex-specific reference established based on a large cohort of Southern Chinese children and adolescents, we believe that the results may be more accurate than the above-mentioned studies. KD children and adolescents had relatively lower peak MET than their healthy peers. The results of a recent study by Yang et al. who recruited age- and sex-matched healthy volunteers through poster advertising were consistent with our results; they reported that adolescents with KD history had significantly lower peak VO₂/kg (approximately 7.93%) than controls [21].

In our previous study, we also found that the PRPP of KD patients was significantly lower than that of controls and the Max-Z of CA showed a significant inverse correlation with PRPP [20]. These findings indicated that KD patients may still have compromised coronary perfusion during exercise after the acute stage and it is crucial to examine the impact of pathological change in CA with EC among KD patients. However, there are few studies about the EC of KD patients with different CA Z-score or CA aneurysm. Allen et al. found no differences in maximal voluntary work rate and maximal VO₂ between KD patients with and without aneurysms [33]. Paridon et al. divided the KD children and adolescents into three groups (one group with no objective evidence of CAA, one group with resolved CAA, and the other one with persistent CAA) and found that the maximal VO₂ was normal after acute KD regardless of the status of CA [35]. Both studies were conducted before 1995, when the idea and application of CA Z-score was not mature and commonly used. The presence of CAA was directly recognized by cardiologist via echocardiography or angiography. Our previous study used ZSP version 4 to calculate CA Z-score and found that children with KD who had higher Max-Z had significantly lower peak MET and PRPP than those with a lower Max-Z. Consistent with our previous study, in the present study, children and adolescents with Max-Z ≥ 2.5 had significantly higher peak VO₂ Z-score and peak PD% than those with Max-Z < 2.5. Since the allometric scaling peak VO₂ Z-score equations were developed for different sex and age groups, which were effective in removing the influence of body mass, height, and age on peak VO₂, these findings indicate that the effects of KD on the cardiovascular system persists for years after the acute stage, which might influence the cardiopulmonary function of patients with KD. Moreover, given that PRPP is a good indicator of coronary flow reserve, it is plausible that the weaker performance during CPET may have resulted from the compromised coronary perfusion during peak exercise in KD participants with Max-Z ≥ 2.5. We assumed that the difference in peak EC between the KD group 1 and 2 might be attributed to CA-related factors such as, CA endothelial dysfunction, compliance of CA lumen, and different CA resistance to inflammatory status [18].

Indeed, the long term effects of KD on the circulatory system may persist even without the presence of CAA. Iemura et al. performed ultrasound cardiography and Tsuchihashiet al. used selective coronary angiography (CAG) to examine the structural changes in CA after the acute stage. Both these studies demonstrated the presence of an abnormal vascular structure even though small CAAs in the acute phase of KD reverted to a normal appearance in the convalescent phase at the previous site of an acute CAA [36, 37]. Some reports have described patients with angiographically normal CA after acute KD who later developed cardiovascular disorders in their early adulthood [38, 39]. Therefore, regular follow-up of KD patients with echocardiography and CPET after the acute stage is crucial.

The Z-score describes how many standard deviations above or below a size or age-specific population mean a given measurement lies. In the context of KD, both western (2017 AHA [2]) and eastern (2020 Japanese Circulation Society Joint Working Group [8]) guidelines recently have endorsed the use of CA Z-score system to define coronary abnormalities and classify CAAs. CA Z-score systems were shown to improve risk classifications of CAAs and predict the clinical prognosis [11]. The application of peak VO₂ Z-score for better evaluation of cardiopulmonary fitness is a new concept [24, 25]. It is important to eliminate the effect of body size on CPET parameters to obtain body size–independent reference values in children and adolescents whose aerobic capacity are strongly influenced by body size and pubertal stage [40]. Adequate peak VO₂ Z-score equation is independent of body size and pubertal stage. Use of peak VO₂ Z-score allows us to compare across ages and sex and might provide a better metric for tracking changes over time [25]. By combining these two concepts of Z-score in this current study, we found that KD children and adolescents with MAX-Z < 2.5 had higher peak VO₂ Z-score than those with MAX-Z ≥ 2.5. This result was in accordance with our previous study showing that KD children with higher Max-Z had significantly lower peak MET [17]. Moreover, it provides a more definitive evidence of the influence of KD on the EC since the peak VO₂ Z-score is independent of sex, age, body size, puberty, and BMI.

Last but not the least, the author wanted to emphasize that KD populations should still engage in exercise normally even though there might be compromised CA flow during the peak exercise effort. The RER is defined as the ratio of VCO₂ and VO₂ consumption measured via respiratory gas analysis. Peak RER ≥ 1.10 is considered the minimal requirement to perform sufficient effort during CPET [28]. The peak RER in all KD participants was 1.16 ± 0.09. Even the participants in the KD group 2 had peak RER of 1.11 ± 0.09. This means that most KD children in our study, irrespective of KD-1 or KD-2 group, could reach peak exercise testing value. Moreover, a physical activity requiring more than 6 METs is considered vigorous according to the definition of ACSM, and the average peak MET in the KD group 1 and 2 of our study were 10.21 ± 1.55 and 9.15 ± 1.89, respectively. There findings suggest that all the KD participants in our study may engage in normal vigorous daily activities.

Some limitations of this study should be considered. First, this was a retrospective study. Although we tried our best to perform the CPET soon (within days) after complete transthoracic echocardiographic examination, there was still some variability in the timing of follow-up. Second, the number of KD patients who had Max-Z ≥ 2.5 (n = 15) was lower than the minimum required sample size (n = 17). Although this distribution was in line with the previous study [5, 7], the results should be viewed in light of this limitation. Third, the subjects were recruited from a single medical center in southern Taiwan. A larger cross-national study is required for further evaluation. Last, the ZSP version 4 is an equation based on the data from Japan. Given the current lack of a well-accepted equation for CA Z-score based on Taiwanese children aged > 6 years, the ZSP version 4 might be the most appropriate and available one. However, the Max-Z by ZSP version 4 calculator may not fully reflect the CA condition of Chinese KD children and adolescents. This situation also applies to peak VO₂ Z-score. Although the peak VO₂ Z-score equation used in this study was derived from data of Southern Chinese children and adolescents, there might be still differences between the Taiwanese and Cantonese.

In this study, KD children and adolescents showed slightly reduced peak EC than their healthy peers even though they could reach maximal exercise effort during CPET. In addition, KD children and adolescents with Max-Z < 2.5 had higher peak EC, including peak VO₂ Z-score, peak PD%, peak MET, and PRPP, than those with MAX-Z ≥ 2.5. Incorporating both the CA Z-score and peak VO₂ Z-score, we could offer clinicians a more precise tool to evaluate cardiopulmonary fitness in children and adolescents, irrespective of age and body size. This metric can lead to enhanced clinical decisions for KD patients, by providing a comprehensive view of cardiopulmonary health. Moreover, our results of peak EC might be due to CA difference since the pulmonary function indices were comparable in the two groups. It is important to promote cardiovascular health of all KD patients after the acute stage owing to the potential long term pathological effects on CA. Monitoring the cardiovascular risk of KD children with Max-Z ≥ 2.5 is imperative.