Background
The pediatric obesity epidemic had resulted in an exponential rise of Type-2 diabetes mellitus (T2DM). It is well established that in children and adolescents with obesity, ventricular chamber size, left ventricular (LV) wall thickness, and LV mass are elevated compared with lean pears despite no significant changes in ejection fraction (EF) [
1]. Furthermore, myocardial geometry and function as assessed by a two-dimensional and three-dimensional speckle tracking echocardiography (2D-STE, 3D-STE) is impaired in obesity with significant reductions in LV circumferential strain, longitudinal strain, and strain rate [
1‐
4]. In obese states, high body mass index (BMI), insulin resistance and hyperinsulinemia are the most important predictors of subclinical myocardial deformational changes [
5].
In states of abnormal glucose tolerance (prediabetes) and/or T2DM, myocardial deformational changes are worsened due to the hyperglycemia, increased fatty acid acids, activation of the renin angiotensin system, microangiopathy and increased oxidative stress. In the context of obese adolescents with abnormal glucose tolerance (dysglycemia), LV mass, systolic blood pressure, and resting heart rate are observed to be elevated when compared to lean peers. The study of obese adolescents who have glucose dysregulation (dysglycemia) merits investigation of subclinical myocardial changes in the context of pediatric obesity [
6,
7]. However, it is unknown whether dysglycemia in comorbid adolescent obesity has effects on myocardial deformation parameters that are more pronounced when compared to obesity alone.
2D-STE and 3D-STE have recently emerged as reliable techniques for quantification of myocardial deformation in multiple imaging planes due to its angle independence and high interobserver reliability [
8,
9]. High temporal resolution during image acquisition allows for the determination of strain rate, or the rate of myocardial deformation, which correlates with LV peak elastance, a load independent measure of LV function, and diastolic ventricular filling [
8‐
10]. In pediatric populations, control values for myocardial longitudinal strain and strain rate from 2D-STE are well established [
10]. LV global longitudinal strain (GLS), the most validated of myocardial deformation indices, is a measure of subendocardial longitudinal myofiber function and is susceptible to ischemia and fibrosis [
11]. Changes in LV GLS are well documented in ischemic cardiomyopathy, although not relied upon clinically [
12]. In addition to its association with ischemia, decreased LV GLS has been reported early in the disease process in adolescents with T1DM and adults with T2DM [
13,
14]. Furthermore, the degree of LV GLS impairment correlates well with HbA1c levels, suggesting that glycemic control may be on the main risk factors for impairment of myocardial mechanics [
13,
14]. While exercise induced LVEF depression has been observed in a subpopulation of young adults with DM, clinical diabetic cardiomyopathy is not observed in these subjects due to lack of inotropic and microvascular abnormalities; rather, these changes are likely due to impaired ventricular-vascular coupling [
15]. Together, these data suggest that LV myocardial deformation parameters and ventricular vascular coupling ratio may serve to be useful subclinical indices of myocardial dysfunction in the context of dysglycemia.
To our knowledge, no studies have investigated myocardial strain and strain rate in adolescents with prediabetes, and few studies have demonstrated changes in LV GLS in T2DM [
13]. The objective of this observational cross-sectional study is therefore to compare myocardial mechanics from 2D-STE, namely longitudinal strain, circumferential strain, strain rate, and ventricular-vascular coupling index in obese-normoglycemic and obese-dysglycemic adolescents. We hypothesized the following: (1) Obese adolescents with dysglycemia would have impaired strain parameters (LS and CS) when compared to obese adolescents with normoglycemia and lean controls; (2) In obese adolescents, changes in strain (LS and CS) and strain rate (LSR and CSR) would reflect changes in ventricular vascular coupling.
Discussion
Our study demonstrates that dysglycemia worsens changes in myocardial deformation parameters as noted in obese adolescents. Despite the observed changes in myocardial deformation, ventricular vascular coupling appears to be preserved in our cohort of obese, insulin resistant adolescents. Thus, clinically, our study highlights that subtle changes in myocardial function are manifest in adolescents with dysglycemia and may be useful for predicting adverse cardiovascular outcomes. “Traditional” diastolic parameters may not be sensitive to detect early stages of diastolic dysfunction. Left atrial enlargement is a reliable diastolic parameter but requires some degree of chronicity of left ventricular diastolic dysfunction to manifest. Image quality in these obese patients may limit the reliability of some of these diastolic parameters.
In our comparative cross-sectional study design, diastolic dysfunction was found to significantly higher in adolescents with dysglycemia when compared to their normoglycemic counterparts. We found significant impairments in global and average LS when compared to lean age matched controls. Furthermore, we observed that this impairment in global and average LS parameters were exaggerated in obese adolescents with dysglycemia compared to their normoglycemic counterparts. Of note in this study distinctive differences were observed in CS in dysglycemic states, implying greater abnormalities in myocardial deformation in dysglycemia than previously appreciated. This defects in CS has not been previously reported in obese insulin resistant adolescents.
Myocardial strain is well known to reflect ventricular vascular coupling, as diminished myocardial strain is known to decrease myocardial efficiency. However, in our study despite the observed changes in myocardial strain, VVI was observed to preserved. Nonetheless, VVI was additionally observed to correlate with circumferential strain rate, a load independent index of LV peak elastance and diastolic ventricular filling. Clinically, VVI is a known prognostic marker and indicator of aortic compliance, LV function, and LV performance. Together with current information regarding the prognostic value of VVI, our results suggest that ventricular efficiency is intimately dependent on diastolic ventricular function and may be a subclinical marker for changes in LV function later in life.
Prior studies have shown that in obese adolescents, insulin resistance alters myocardial substrate utilization and increases sympathetic tone, preload, and fatty acid metabolism, which ultimately results in impaired LV contractility and may lead to clinical heart failure in later life [
21]. Additionally, many studies have shown that myocardial deformation as assessed by strain and strain rate are altered prior to clinically relevant ventricular dysfunction [
1,
4,
22]. Singh et al. recently demonstrated that LV GLS and early diastolic strain were significantly decreased in obese adolescents and in obese subjects with nonalcoholic fatty liver disease (NAFLD) than those without NAFLD compared with appropriate controls. Furthermore, both GLS and early diastolic strain were observed to be independently associated with the homeostatic model of insulin resistance (HOMA-IR) [
23]. Sanchez et al. reported that LS was found to be abnormally depressed in obese adolescents (− 13.35%) compared with lean controls (− 18.8%) [
22]. Our study demonstrates similar findings in global and average LS in a population with similar demographics. As previously reported, traditional cardio metabolic risk factors, namely SBP, LDL, fasting glucose, and insulin, were not predictive of these LV functional changes in pediatric populations. Thus, this study confirmed previous reports that beyond conventional echocardiography parameters, obesity was associated with decreased myocardial strain, as assessed by 2D-STE [
11,
24].
Studies of adults with T2DM have shown that cumulative effect of diabetes exposure (those with diabetes diagnosis in childhood and/or early adulthood) and BMI had greatest impact on the adverse LV remodeling [
25,
26]. Bjornstad et al. were the first pediatric study to analyze strain cardiac strain from 2D-STE in adolescents with T2DM and found that compared to lean and to obese controls, adolescents with T2D had significantly lower CS [
27]. Our results showcase similar findings as adolescents with dysglycemia, which included a small T2DM cohort, were observed to have a depressed CS, despite differences in study demographics, namely differences in ethnicity and a more obese cohort in our group (BMI Z = 2.49) compared to Bjornstad’s group (BMI Z = 2.01) [
28,
29]. Our study thus underscores that dysglycemia attenuates unfavorable cardiac remodeling in obese adolescents, as demonstrated by changes in LS and CS. Briefly, LS is considered a measure of subendocardial injury and fibrosis, while CS is a measure of mid wall fiber damage [
9]. In early ventricular dysfunction, a depressed LS is followed by an elevated CS to maintain LVEF. However, progressive dysglycemic conditions result in a decline in CS, indicating mid-wall fiber injury and damage to deeper myocardial layers. Consequently, changes in CS have been known to precede incident heart failure. In our study, we observed significant changes in both LS and CS in obese adolescents with dysglycemia compared with normoglycemic counterparts. Higher rates of dysglycemia and insulin resistance are reported in male adolescents as reflected in our study [
30,
31]. Furthermore, in our cohort, no significant differences were observed in myocardial strain parameters, by gender. Bjornstad reported gender specific differences in adolescents with Type 1 diabetes (19 males and 22 females; average age 15 years). Compared to controls, adolescents with type 1 diabetes had significantly lower CS (− 20.9 vs. − 22.7%, P = 0.02), but not LS (P = 0.83). Boys with T1D had significantly lower LS than girls with T1D (− 17.5 vs. − 19.7%, P = 0.047), adjusted for Tanner stage [
32]. In a study of adults (n = 277; age 56.1 years) with metabolic syndrome showed that global LS were lower in males than females as correlated with hs CRP and epicardial adipose tissue [
33]. Along with the observed correlation between CS and HbA1c, we speculate that impaired glycemic control may be play a role in myocardial wall fiber injury.
Serum biomarkers of inflammation, namely adipokines and leptins, have been associated myocardial deformational change [
29]. Adiponectin is an adipokine secreted by adipocytes and has been established to have a cardio protective effect, and elevated adiponectin has been associated with lower risk of myocardial infarction in men [
34]. In adiponectin knockout mouse models, abnormal cardiac remodeling and cardiac concentric hypertrophy have been observed, indicating adiponectin mediated signaling may play a role in preventing abnormal ventricular remodeling [
35,
36]. In our cohort we demonstrated a negative linear relationship between average LS rate and adiponectin (r = − 0.4), indicating the serum adiponectin levels correlate with early subendocardial fiber damage.
Previously, Li et al. reported that LS at the cardiac apex correlated with VVI in adults with T2DM with either normal or depressed ejection fraction. Our study demonstrated similar findings. However, Li et al. demonstrated abnormal VVI (1.55 ± 0.53), indicating impaired ventricular-vascular coupling, a finding not observed in our cohort. We speculate that changes in VVI in pediatric populations likely are less pronounced compared to adult populations due to lack of other known comorbidities that abnormally alter ventricular-vascular coupling.
Of note is the correlation observed in our study between VVI and circumferential strain rate, a proxy for LV peak elastance. To our knowledge, this is the first study to have correlated VVI, an index of mechanical efficiency and myocardial deformation, namely circumferential strain rate. In the broader context of myocardial mechanics and function, strain rate is a load independent marker of myocardial deformation and can be utilized for clinical assessment of regional wall motion abnormalities [
37‐
39]. While strain rate is unable to quantify myocardial contractility due to its load independence, it remains a clinically useful tool for assessment of passive expansion and recoil of the myocardium [
39]. Clinically, measurements of strain rate have correlated well with improvements in wall motion deformation post coronary bypass and percutaneous revascularization and for identification of wall motion abnormalities in hypertrophic cardiomyopathy and cardiac amyloidosis [
39‐
41]. In T2DM, myocardial steatosis and myocardial triglyceride content have demonstrated to be independent predictors of both LV and RV longitudinal strain and strain rate, suggesting that subclinical changes in myocardial deformation assessed by strain rate are influenced by extra myocardial signaling factors [
42]. In the metabolic phenotype of heart failure with preserved ejection fraction (HFpEF), similar reductions in diastolic strain rate have been observed [
43,
44]. Together with the observed depression in CSR in dysglycemia in study, these data suggest broader implications for the importance of myocardial strain rate in understanding the relationship between dysglycemia and abnormalities in myocardial deformation, particularly in HFpEF in patients with T2DM or metabolic syndrome,
Limitations
Our study has several limitations. Due to the cross-sectional design, causal relationships cannot be established among the observed associations. Additionally, the predominance of Hispanic subjects, about 60% of the cohort, limits generalizability of the findings. However, the study presented simultaneously provides novel data in for pediatric Hispanic subjects, a population known to be a high risk for diabetes. The reduced sample size in this study reduced power required to detect differences between groups based on OGTT results. Though we had age and gender matched control group, we recognize that the control group was predominantly Caucasian.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.