In this retrospective multi-center study, 150 healthy children and adolescents were included. Those subjects were identified from those who underwent complete echocardiograms for clinical indications (heart murmur, atypical chest pain, syncope, and family history of congenital heart diseases) from databases of the following institutions: (1) Pediatric Echocardiographic Laboratory of the Magna Graecia University Hospital of Catanzaro (Italy); (2) Pediatric Echocardiographic Laboratory of the Royal Brompton Hospital of London (United Kingdom); (3) Pediatric Echocardiographic Laboratory of the Department of Women's and Children University Hospital of Padua (Italy). Retrospective use of clinical data for this study was approved by the local competent clinical director in charge.

Demographic data such as gender, age, weight, height, and BSA have been collected for each subject. Moreover, non-invasive blood pressure values have been measured for every individual at the time of echocardiographic examination.

Inclusion criteria were: age between 0.5 and 18 years and normal cardiac structure and function at cardiovascular examination. Children were selected among those referred to our ambulatory for different clinical indications (heart murmur, chest pain of unknown origin, or as a volunteer among relatives of patients with simple congenital heart disease). All the selected subjects have an electrocardiogram and a complete echocardiography examination excluding any cardiovascular abnormalities. Exclusion criteria were: poor acoustic window, poor compliance, marked sinus arrhythmia or frequent extra-systole, any cardiac and extracardiac abnormality, and the presence of any genetic syndrome or disease.

Table 1 summarizes the main characteristics of the study population. A total of one hundred fifty children and adolescents were included with an overall mean age of 10.6 ± 4.5 (age range 0.5–17.9 years). Of those, 91 were males (mean age 11.3 ± 4.4, age range 0.7–17.9 years) and 59 were females (mean age 9.5 ± 4.5 years, age range 0.5–16.6 years). Seven patients (4.6%) were 6-to-12-month-old, 16 (10.6%) patients were 1-to-5-year-old, 33 (22%) were 6-to-9-year-old, 39 (26%) were 10-to-13-year-old, and 55 (36%) were 14-to-18-year-old. Significant differences in height (p = 0.002), weight (p = 0.003), BSA (p = 0.001), diastolic blood pressure (p = 0.040), and LV end-diastolic diameter (LVEDD) (p = 0.020) between male and female subjects were observed. There were no significant differences in LV ejection fraction (EF) and LV global longitudinal strain (GLS) between sexes.

Two-dimensional MW parameters acquired from the study population are shown in Table 2 and in Figure 1. The lowest expected values were 1,723 mmHg% in males and 1,682 mmHg% in females for GWI, 2,089 mmHg% and 2,106 mmHg% for GCW, and 95.9 and 95.5% for GWE, respectively. The highest expected value for GWW was 78 mmHg% in men and 90 mmHg% in women. GWI, GCW, and GWE were slightly lower in females than in males, while the opposite occurred for GWW, with no statistical significance.

There were significant linear correlations between GWI and GCW with both systolic blood pressure (SBP) and GLS (p < 0.001 for all). No significant correlations were observed between GWE and GWW with SBP and between GWW and GLS (p = NS); there was a significant correlation between GWE and GLS (p = 0.020).

As to demographic variables, GCW, GWW, and GWE correlated significantly with age (R = −0.267, p < 0.001, R = −0.203, p = 0.012, and R = 0.162, p = 0.047, respectively); GCW and GWW with BSA (R = −0.289, p = 0.001, and R = −0.178, p = 0.035, respectively).

On both univariable and multivariable analyses, we observed significant associations between either GWI or GCW with SBP (β coefficient = 0.446, p < 0.001; β coefficient = 0.456, p < 0.001, respectively) and LV GLS (β coefficient = −0.268, p = 0.001; β coefficient = −0.233, p = 0.003, respectively).

No associations were found at multivariable analysis between any of the MW parameters and age, sex, or BSA.

Intra-observer and inter-observer indices of variability for MW parameters are shown in Table 3. Intra-observer and inter-observer analyses demonstrated good repeatability and reproducibility of MW parameters (Table 3; Figure 2). The Bland-Altman plot showed a low bias with narrow limits of agreement and no evidence of relevant heteroscedasticity for all MW indices assessed (Figure 2). Figure 3 reports a practical example of calculation of non-invasive Myocardial Work, along with summary normative results.

It is well known that longitudinal deformation parameters are influenced by changes in loading conditions, which may limit their accuracy (7). For example, it has been previously demonstrated that an increase in afterload may lead to a systolic impairment in presence of preserved or even increased MW (7).

Myocardial work has been measured invasively and used as a marker of ventricular contractility for decades (8–10). Russell et al. have introduced a novel method for calculating non-invasive MW (11, 12), which takes into consideration speckle-tracking analysis and LV pressure estimated from brachial artery cuff pressure and good reproducibility (13).

Myocardial work parameters have been shown to be reliable predictors of response to cardiac resynchronization therapy (CRT) in patients with heart failure with reduced ejection fraction (18). Chan et al. have demonstrated that MW indices are significantly increased in patients with hypertension compared to controls, regardless of normal GLS values (19).

Most recently, a previous study from our group demonstrated MW indices as being sensitive and early indicators of myocardial ischemia during transient acute coronary occlusion (20).

Our study sought to provide norms for future studies investigating the clinical utility of MW parameters in children and adolescents, especially in those with congenital or acquired cardiac disease.

Non-invasive MW parameters have already shown insightful diagnostic yield in two studies conducted on pediatric population. Children with Turner syndrome demonstrated increased MW compared to age-matched controls, as shown by Oberhoffer et al. (21). Results from this study may relate to the early development of arterial stiffening widely reported in such patients.

A recent study from our group evaluated the diagnostic performance of non-invasive MW indices in a cohort of children and adolescents with Kawasaki disease (KD) with and without coronary dilatation (12). It was thus demonstrated that non-invasive MW indices are reduced in children with KD, even in the absence of coronary aneurysm, and that they can detect subtle myocardial abnormalities better than GLS and LV EF.

The feasibility of non-invasive MW in our healthy pediatric population was optimal, as 2D imaging quality is usually greater in pediatrics than adults. Thus, we had to exclude no subject from this study because of poor image quality. On the other side, intra- and inter-observer variability for MW indices were relatively higher than the previous rates published in adults (22). This can be partially explained by the fact that we did not perform any specific acquisition for valvular timings, and this may be particularly relevant in pediatric patients in the presence of a relatively high heart rate. Moreover, measurement of non-invasive blood pressure in children could be challenging, especially in pre-school children or newborns.

This is a retrospective study. Thus, our findings should be further validated in large prospective studies.

Measurement of non-invasive MW, in this study, is derived from speckle-tracking analysis, carrying with itself the same well-known limitations of the speckle-tracking techniques. Indeed, while acoustic windows are usually more echogenic in younger patients, poor image quality can be related to the lack of children's compliance, the relatively high heart rate, and lung artifacts.

Age distribution of patients may have influenced statistical analysis. In particular, since we did not set up a fixed number of individuals per age class to minimize selection biases, different age classes do not have the same size. For the same reason, though we do report data for all the different age classes, these should be considered as informative explorative data, as some age subgroups included a limited number of individuals and their relative normative data may not be sufficiently robust.

Our results do not apply to subjects from 0.0 to 0.5 years of age. In fact, these were not included in this study because they represent a different physiological background. In particular, relevant maturational changes in myocardial deformations compared to older children are to be expected. For this reason, the 0–0.5 years age groups merit a dedicated study to find normal values.

Finally, the non-invasive MW evaluation currently can only be performed by means of General Electric echocardiography machines and EchoPAC software. This can reduce its broad utilization; moreover, we might speculate vendor-specific differences in GMW-parameters when other manufacturers will upgrade their software with these functionalities. Would other vendors develop their MW indices? This would most probably need a dedicated normative study.