Morbid obesity (body mass index [BMI] ≥40 kg/m²) affects over 1 million adolescents and young adults (1), and being obese as an adolescent is predictive of remaining obese into adulthood (2- 3). Little is known about the reversibility of cardiac abnormalities in morbidly obese adolescents. The purpose of this study was to determine whether cardiac abnormalities reverse with significant weight loss in adolescents. An important model for this evaluation is bariatric surgery, because rapid and profound weight loss occurs over a short period of time, such that plasticity, if present, might be demonstrated.

All morbidly obese (BMI ≥99th percentile) (4- 5) adolescents (≤19 years old) undergoing bariatric surgery at Cincinnati Children’s Hospital Medical Center were eligible for inclusion. All patients were physician-referred and then evaluated by a multidisciplinary team before undergoing surgery (6).

To be included in the analysis, echocardiograms had to be obtained pre-operatively and at least 4 to 18 months post-operatively, and image quality had to be visually discernible for measurements. The number of subjects included (n = 38) who had both an adequate pre-operative and post-operative echocardiogram is less than the total number of operations performed (n = 67) on this age group during this timeframe. When multiple echocardiograms had been performed on a subject, the 2 echocardiograms chosen for analysis were the pre-operative echocardiogram closest to the surgery date and the post-operative echocardiogram with the longest follow-up period. This analysis represents all patients who met the aforementioned criteria between August 2002 and December 2006.

Patients with congenital heart disease were excluded.

Institutional review board-approval for retrospective analysis of these clinical data was required.

Patient weight, height, race, gender, age, and blood pressure were recorded at the time of the echocardiogram. Weights were obtained without shoes, in light clothing, with a digital scale. Heights were measured with a calibrated wall-mounted stadiometer. Body mass index was calculated as weight (kg)/height × height (m²). Because the Centers for Disease Control and Prevention growth curves for standard weight and BMI are based on a normal population distribution, the accurate calculation of the outermost BMI percentiles less than the third percentile or less than the 97th percentile (z-scores below −2 [2.3rd percentile] or above 2 [97.9th percentile]) are felt to be beyond the capabilities of the growth curve data. Therefore, to account for the morbid obesity of these participants, ideal body weight (IBW) was also calculated for each patient and their weight was indexed to IBW as “percent over IBW.” The IBW was defined as the weight at the 50th percentile of BMI for the patient’s gender and age. Percent over IBW was then calculated as: (100 × [patient’s weight in kg/IBW]) − 100.

Heart rate was obtained from averaging 3 RR intervals during the echocardiogram. The rate-pressure product (an indirect index of myocardial oxygen consumption) was calculated as heart rate × average systolic blood pressure (7- 8).

The cardiovascular (CV) system was assessed with a GE Vivid 5 or 7 (Milwaukee, Wisconsin) or a Philips Sonos 5500 (Andover, Massachusetts) ultrasound system. All echocardiographic images were obtained with the patient in the left decubitus position to obtain the following images: parasternal long- and short-axis and apical 4-chamber. Measurements were performed off-line with a Cardiology Analysis System (Digisonics, Houston, Texas). A total of 3 to 5 measurements for each variable were obtained per patient.

Measurements were obtained by 2-dimensional directed M-mode for left ventricular end-diastolic dimension (LVED), left ventricular end-systolic dimension (LVES), short-axis left ventricular (LV) end-diastolic cavity area, end-diastolic septal thickness (IVSd), and posterior wall thickness (LVPWd) (9). Left ventricular mass (LVM) was calculated with the method of Devereux et al. (10).$LVM(g):=(0.8)(1.04)×[(LVED+LVPWd+IVSd)3−LVED3)]+0.6$

The LV mass was indexed by dividing by height in meters^2.7 as described by de Simone (11). Relative wall thickness (RWT) was calculated by: RWT = (LVPWd + IVSd)/LVED. Cardiac geometry was subdivided on the basis of LVM and RWT into: concentric hypertrophy (increased LVM and increased RWT), eccentric hypertrophy (increased LVM and normal RWT), concentric remodeling (normal LVM and increased RWT), and normal geometry (normal LVM and normal RWT) (12- 13). For RWT, a common adult threshold of >0.43 cm was used to determine abnormal RWT (14- 15). For indexed LVM, an adult threshold of 51 g/m^2.7 was used, owing to the association with predicting future cardiac morbidity in adults (16). For statistical analysis, patients were stratified by LV geometry subtype.

Shortening fraction (SF = [(LVED − LVES)/LVED]) and heart rate-corrected velocity of circumferential fiber shortening (VCF) (VCF = SF/LVETc) were calculated where LVETc is the heart rate corrected ejection time. Left ventricular end-systolic meridional wall stress (WS) was determined by (17- 18):$WS(g/cm2)=(1.35)(P)(LVES)/((4)(LVPWs)×(1+LVPWs/LVES))$ where P = LV end-systolic pressure (mm Hg), and LVPWs = end-systolic LV posterior wall thickness (cm).

Finally, contractility was assessed by calculating VCF difference (i.e., VCF diff = measured VCF − the predicted VCF for the measured WS) (19- 20).

There is no single measurement for evaluation of LV diastolic function. Instead, multiple measures are generally evaluated, and therefore the following were included in this analysis:

Laparoscopic Roux-en-Y gastric bypass was performed on all patients as described in (Figure 1) (24).

Analyses were performed with SAS version 9.1 (SAS Institute, Inc., Cary, North Carolina). In preliminary analyses, all variables were examined for missing values and outliers with cross (e.g., scattergrams) and range checks. Continuous variables were tested for normality with the Kolmogorov-Smirnov D statistic as well as visual inspection of the univariate distributions. Chi-square statistics were used to determine difference in proportions between categorical variables. Descriptive statistics (mean, SD) were generated for the demographic characteristics. To compare differences before and after surgery for normally distributed continuous variables, Student t tests for paired data were performed. The Wilcoxon signed-rank test, the nonparametric equivalent, was used to assess changes in the following variables: height, age, heart rate, Ea ratio, and Ea-sept. Statistical significance was determined at a p value of <0.05. Spearman’s and Pearson’s correlations were calculated to further determine which factors were associated with measures of change in cardiac structure and change in cardiac function after weight loss. Specifically, the dependent variables were change in: indexed LVM, E/A, Ea/Aa-lat, E/Ea-lat, LA size, and indexed LA size. The variables included in correlation analysis as potential independent risk factors were: gender, race, and change in: weight, BMI z-score, BMI, kilograms over ideal body weight (IBW), percent over IBW, heart rate, systolic blood pressure z score (SBP-z), diastolic blood pressure z score (DBP-z), indexed LVM, IVSd, LVPWd, LVED, RWT, SF, and VCF difference. Multivariate regression models were built with the results from Pearson’s analysis to explore potential determinants of change in indexed LVM. These models were built by selecting either clinically or statistically significant independent variables from the univariate correlation analyses. The following independent variables were included in the model: gender, race, change in percent over IBW, change in SBP-z, and change in DBP-z. As an additional analysis, the pre-operative values for indexed LV mass and LV geometry were added to the aforementioned independent variables. Analysis of covariance was performed to further elaborate upon the results of the regression analyses, where appropriate. No regression models were built for changes in diastolic function, because there were no significant correlations in univariate analyses.

A total of 38 morbidly obese adolescents were evaluated (Table 1). The average age at surgery was 16 ± 1 years (range 13 to 19 years). There were 29 girls and 9 boys, of whom 33 were Caucasian and 5 were African American. As expected, BMI, weight, and percent over IBW significantly decreased after gastric bypass. Follow-up echocardiograms were performed at a mean post-operative period of 10 ± 3 months.

An effect was seen between pre-operative and post-operative measurements of hemodynamic variables, including a significant decrease in both heart rate and systolic blood pressure (p < 0.0001) (Table 1). In addition, the rate-pressure product significantly improved (pre-operative mean 9,832 ± 2,843 vs. post-operative 6,937 ± 1,637, p < 0.001). These changes demonstrate a shift toward a more favorable autonomic status and possibly reflect a decreased cardiac workload (7- 8).

After weight loss there were significant improvements in indexed LVM, LVPW, interventricular septal thickness, and RWT (Table 2). In univariate analysis, change in indexed LVM correlated with change in percent over IBW (r = 0.41, p = 0.01). Multiple stepwise regression analysis with backward stepping was performed to evaluate predictors of change in indexed LVM. The following variables were included in the model as potential independent predictors: gender, race, change in SBP-z and DBP-z, and change in percent over IBW. The only predictor of change in indexed LVM was change in percent over IBW (r = 0.41, p = 0.01). This suggests that, as weight improved, indexed LVM improved (Figure 2). As an additional analysis, the pre-operative values for indexed LV mass and LV geometry were added to the aforementioned independent variables. Change in percent over IBW and type of pre-operative LV geometry were the only 2 independent risk factors associated with change in indexed LVM at follow-up (p = 0.001 and p = 0.002, respectively, model R² = 0.49). Analysis of covariance showed that the subjects with pre-operative concentric LVH and eccentric LVH geometry had a significantly greater mean decrease in indexed LVM when compared with those with normal geometry.

Concentric LVH was seen in 28% pre-operatively, whereas post-operatively only 3% had concentric LVH. The percentage with normal LV geometry improved from 36% pre-operatively to 79% at follow-up. The overall geometry distribution was significantly improved in the post-operative period (p < 0.0001) (Figure 3). In addition, none of the patients with normal LV geometry pre-operatively regressed into abnormal LV geometry at follow-up. All changes in LV geometry after weight loss were uniformly in the direction of improvement.

Systolic function and contractility were normal pre-operatively and did not significantly change at follow-up (Table 2). There were no significant changes in end-systolic wall stress or stress-velocity relationship.

Compared with normative diastolic data in adolescents ages 14 to 18 years old (25), the pre-operative values for E/Ea-lateral and Ea-lateral for this cohort were abnormal (Table 3). The normal range for E/Ea-lateral ratio is 4.7 ± 1.3, and Ea-lateral velocity is 20.6 ± 3.8 as previously described by Eidem et al. (25). Post-operatively, there were no significant changes in LA diameter; however, there were significant improvements in other variables of diastolic function, including: improved E/A ratio, Ea/Aa ratio, and E/Ea lateral ratios. The lack of improvement in LA diameter might be secondary to small sample size or lack of sufficient time for remodeling of LA chamber size. In addition, the LA diameter was limited to m-mode measurements and did not include planimetry. The improvements in Doppler indexes of diastolic function suggest that some features of LV diastolic function are capable of improving after significant weight loss in morbidly obese adolescents. Univariate correlations were performed to analyze the change in Doppler indexes of diastolic function after weight loss. Change in diastolic Doppler indexes were not correlated with change in: weight, BMI z-score, percent over IBW, heart rate, SBP-z, DBP-z, indexed LVM, IVSd, LVPWd, LVED, RWT, SF, and VCF difference.

The significant findings of this study are that abnormal cardiac geometry and altered diastolic function in morbidly obese adolescents improve with significant weight loss. These results are important, because reversal of abnormal cardiac geometry might improve predictors of future CV morbidity in these young people.

Over one-quarter of adolescents in this data set had already developed high-risk concentric hypertrophy before bariatric surgery. In 1 adult study of hypertensive patients, 53% with concentric hypertrophy had CV events, compared with 10% of those with eccentric hypertrophy (26).

In this adolescent population, concentric hypertrophy largely reversed. In addition, prevalence of normal LV geometry improved from 36% pre-operatively to 79% after surgical weight loss. In a study in adults, improvements in indexed LVM (g/m) were seen after surgical weight loss, but no significant change was observed in LVED or posterior wall and septal thicknesses (27). By comparison, this study shows the posterior wall and septal thicknesses in adolescents can significantly improve after weight loss, demonstrating remodeling of the myocardium. This might be an argument for intervention at an earlier age. Although likely related to age and duration of obesity, these differences between adults and adolescents might also be due to sample size, variable weight changes, or variability in length of follow-up.

Because these adolescents are anthropometrically similar to adults, we chose adult thresholds for determining LV geometry subtypes. But as our results demonstrate, even by adult standards, a significant number of adolescents seem to be at increased risk for future cardiac morbidity on the basis of LV mass and geometry. Using pediatric criteria would have reflected an even higher prevalence of abnormal cardiac geometries (15).

Reversal in LV geometry patterns and improvement in LV mass carries prognostic importance. Data suggest that in hypertensive adults, LVM >51 g/m^2.7 indicates a >4-fold risk of CV morbidity (16). The adolescent population in our analysis had an average indexed LVM of 54 g/m^2.7 (maximum of 86 g/m^2.7). The LVH of this magnitude suggests that adolescents with morbid obesity are at elevated risk for future adverse cardiac events.

Interestingly, even in the face of continued obesity, the indexed LVM improved with weight loss and the prevalence of concentric LVH nearly resolved. One hypothesis to explain these observations is that weight loss translates into a reduction in cardiac workload that thereby decreases LVM. This possibility is supported by the reduction in heart rate, decrease in systolic blood pressure, and decrease in rate-pressure product.

Initial increases in LV mass might be physiologic (28), and in obesity this might be due to increased body mass, yet the concern is that long-term exposure to LVH might result in poor outcomes. In 1970, the Framingham Heart Study reported the risk of cardiac morbidity and mortality in men with significant LVH, even after adjustment for hypertension (29- 30). Results from another study showed that improvements in LVH were associated with lower CV morbidity and mortality, also independent of blood pressure (31). In addition, Devereux et al. (32) reported that lower LVM was associated with better outcomes, as demonstrated by lower rates of CV death, myocardial infarction, and stroke.

Without weight loss intervention, these morbidly obese children (BMI ≥99th percentile) will likely remain obese in adulthood (2- 3). In recent adult studies, bariatric surgery for severe obesity is associated with decreased overall mortality (33- 34). Although this type of evidence is currently lacking in adolescent studies, at the very least, significant LVH should be considered a relative indication for more aggressive weight loss interventions.

In this analysis, there were also several measures of diastolic function that improved after weight loss, including mitral Ea-lat velocity and increased E/Ea-lat ratio, with post-operative results comparing more favorably with normal adolescents (25). The adult published reports examining diastolic function after gastric bypass have shown conflicting findings (27,35- 36). Again, the improvements seen in this adolescent population might reflect advantages to earlier age at intervention. Additional studies are needed to determine the effect of other confounding factors, including: sample size, length of follow-up, and magnitude of weight loss.

This study was limited by the analysis of only 38 of the 67 patients who were ≤19 years old and underwent bariatric surgery during this study period. The remaining 29 patients were not evaluated, because either no pre- or post-operative images were available for comparison or the image quality was insufficient for analysis. Of note, there were no significant differences between pre-operative and follow-up weights between those included versus excluded (pre-operative BMI 60 vs. 60 kg/m² and follow-up BMI 41 vs. 40 kg/m², respectively). Therefore it is doubtful that a selection bias was introduced on the basis of weight. It is recognized that other comorbidities associated with morbid obesity could be playing a role in the observed changes of cardiac structure and function, including changes in: metabolic parameters, obstructive sleep apnea, dyslipidemia, body fat distribution, and hypertension. Although patients receiving medications were not excluded from the analysis, there were only 5 of the 38 that were taking cardiac medications. Two of those were receiving lipid-lowering medications, and 3 were receiving antihypertensive drugs. All were able to discontinue medications within 1 to 5 months post-operatively, with the exception of 1 patient taking an antilipidemic drug. Four patients were taking glucophage pre-operatively, and all were discontinued within 6 weeks post-operatively. There was 1 type I diabetic patient controlled on an insulin pump, with a glycosylated hemoglobin (HbA1c) value of 6.2%. The population as a whole was not hypertensive before or after surgery despite significant obesity.

We demonstrate for the first time in morbidly obese adolescents the plasticity of pathologic LV geometry patterns with weight loss. The abnormalities of cardiac structure and function that can be reversed include: abnormal LV mass, high-risk forms of LV geometry, cardiac workload, and abnormal diastolic function. Weight loss interventions during adolescence might be more likely to change cardiac parameters than adults who have been exposed to long-term effects of obesity. This might be an argument for earlier intervention at younger ages in severely obese young people.

Long-term follow-up studies in adolescents are needed to determine whether these improvements in LVM and cardiac geometry persist and whether these findings translate into long-term reduction in their future CV morbidity during adulthood.