Introduction
The increasing prevalence of severe obesity in adolescents forms a critical public health challenge, characterised by a worrying increase in obesity related comorbidities. Historically associated with adults, these comorbidities are diagnosed in younger populations at increasing rates. Reference Bentham, Di Cesare, Bilano, Bixby, Zhou and Stevens1 A comorbidity of important concern is maladaptive change in left ventricular geometry, particularly left ventricular hypertrophy. Reference Daniels, Pratt, Hayman, Pratt and Hayman2–Reference Avelar, Cloward, Walker, Farney, Strong and Pendleton4 Such changes potentially increase the risk of future cardiovascular disease, thus demanding urgent medical attention and interventional strategies in this vulnerable demographic.
Bariatric surgery is proven to be a viable intervention for adolescents with severe obesity, providing significant improvements in weight and obesity related comorbidities, in particular when changes in lifestyle and pharmacotherapy are insufficient. Reference Inge, Courcoulas, Jenkins, Michalsky, Helmrath and Brandt5 However, the impact of significant weight loss on cardiac geometry in this vulnerable demographic group remains underexplored.
Cardiac remodelling in obesity, manifesting as left ventricular hypertrophy or other structural changes, is a maladaptive response with potential long-term consequences, including heart failure and cardiac arrhythmias. Reference Daniels, Pratt, Hayman, Pratt and Hayman2,Reference Michalsky, Inge, Jenkins, Xie, Courcoulas and Helmrath6,Reference Azevedo, Polegato, Minicucci, Paiva and Zornoff7 The degree to which these cardiac changes are reversible, in particular as a response to bariatric surgery in adolescents, is a crucial question. This study aims to fill the existing knowledge gap by evaluating changes in cardiac geometry using echocardiographic assessments in severely obese adolescents undergoing bariatric surgery in addition to multidisciplinary lifestyle intervention, providing insights into potential reversibility and implications for long-term cardiovascular health.
Methods
This study was conducted as part of a randomised controlled trial for adolescents eligible for bariatric surgery (BASIC trial, NCT01172899). Reference Roebroek, Paulus, Van Mil, Vreugdenhil, Winkens and Nederkoorn8 The study population of the BASIC trial consists of severely obese adolescents who all have been treated extensively for their severe obesity by conservative methods for at least 12 months without effect. Patients in the intervention arm received laparoscopic adjustable gastric banding in addition to multidisciplinary lifestyle intervention, while patients in the control arm received only multidisciplinary lifestyle intervention.
Participants
All patients included in the BASIC trial who received a baseline cardiac ultrasound examination were used for analysis at baseline and follow-up. Detailed information regarding the BASIC trial study design, inclusion and exclusion criteria, and randomisation process was published previously. Reference Cole and Lobstein9 In summary, inclusion criteria were age 14–16 years; sex- and age-adjusted BMI ≥ 40 kg/m2 (or ≥ 35 kg/m2 combined with presence of obesity-associated comorbidity); and participation in combined lifestyle interventions during at least 12 months without adequate weight loss (defined as 5% total body weight loss). In order to maintain a homogeneous study population with regard to pubertal status, girls were excluded if they were premenarchal, boys if their bone age was < 15 years. All participants were subjected to standardised comprehensive baseline measurements and investigations in order to exclude (subclinical) conditions causing obesity.
Measurements
All measurements within one patient were carried out during a single visit at baseline and one year. Body height and weight were measured using a stadiometer and digital scale, respectively, with patients dressed in underwear. A tape measure was used for standardised measurement of body circumferences at the neck level, abdominal level, and hip level. BMI was calculated as [body weight]/[body height*body height] in kg/m2, and BMI z scores were calculated using Cole’s LMS method. Reference Cole and Lobstein9 Daytime blood pressure was measured while the patient was resting, during a period of 60 to 90 min with intervals of 3 min between measurements, using the Mobil-O-Graph® NG (I.E.M. GmbH, Stolberg, Germany). Prehypertension and hypertension were defined according to the fourth report from the National High Blood Pressure Education Program, and blood pressure z scores were calculated according to the method described in that same report. 10
A fasting blood draw was performed to measure serum glucose, insulin, cholesterol, triglycerides, free fatty acids, glycated haemoglobin (HbA1C), and leptin. HOMA-IR was calculated according to the method described by Mathews et al., i.e., ((fasting insulin (microU/l)/(fasting glucose (mmol/l))/22.5, where insulin was converted from pmol/l to microU/l by dividing by 6.945. Reference Matthews, Hosker, Rudenski, Naylor, Treacher and Turner11
Video-assisted 12-channel polysomnography (Brain RT, OSG, Rumst, Belgium) was performed at the paediatrics department of Maastricht University Medical Center. The scoring of sleep stages and respiratory-related events was performed by a single specialised analyst using the American Academy of Sleep Medicine (AASM) 2012 updated guidance for scoring paediatric respiratory events. Reference Berry, Budhiraja, Gottlieb, Gozal, Iber and Kapur12
All cardiac ultrasounds were planned with a single paediatric cardiologist, who reported interventricular septal thickness, left ventricular end-diastolic posterior wall thickness, left ventricular end-diastolic dimension (LVEDD), and left ventricular end-systolic dimension. In the absence of this paediatric cardiologist, another experienced paediatric cardiologist was consulted to perform the ultrasound examination.
Left ventricular mass was calculated using the Devereux formula, where left ventricular mass (in grams) is equal to 0.8(1.04((LVEDD + IVST + LPWT)3 − LVEDD3)) + 0.6. Reference Devereux, Alonso, Lutas, Gottlieb, Campo and Sachs13 Left ventricular mass was subsequently indexed by dividing it by height to the power of 2.7, which was previously described as the optimal height exponent between children and adults for indexing left ventricular mass. Reference De Simone, Devereux, Roman, Alderman and Laragh14,Reference Daniels, Kimball, Morrison, Khoury, Witt and Meyer15 Thus, left ventricular mass index is equal to left ventricular mass divided by height2.7 (height in metres). Left ventricular hypertrophy was defined as a left ventricular mass index ≥ 38.6 g/m2.7, in accordance with the 95th percentile of left ventricular mass in a cohort of 192 healthy children aged 6 to 17 years old. Reference Daniels, Kimball, Morrison, Khoury, Witt and Meyer15
Cardiac geometry was further specified using the relative wall thickness and left ventricular mass index, according to the method of Ganau et al., where relative wall thickness was calculated as 2*LVPWT/LVEDD. Reference Ganau, Devereux, Roman, de Simone, Pickering and Saba16 In children, a relative wall thickness > 0.41, corresponding to the 95th percentile of relative wall thickness in children and adolescents, corresponds to concentric changes, and a relative wall thickness ≤ 0.41 corresponds to normal geometry or eccentric changes. Reference Daniels, Meyer, Liang and Bove17 We defined normal geometry and concentric left ventricular remodelling as a normal left ventricular mass index (< 38.6 g/m2.7) with a relative wall thickness that is ≤ 0.41 or > 0.41, respectively. We defined left ventricular hypertrophy as a left ventricular mass index ≥ 38.6 g/m2.7 and further specified this as eccentric left ventricular hypertrophy if relative wall thickness was ≤ 0.41 and concentric left ventricular hypertrophy when relative wall thickness was > 0.41.
Statistical analysis:
Numerical data are presented as mean ± standard deviation and range where appropriate. Categorical data are presented as number (percentage). Demographic and clinical variables were compared between patients in the intervention and control arm, using an independent samples t-test for numerical variables and chi-square or Fisher’s exact test for categorical variables. Differences between baseline and follow-up within one treatment arm were evaluated using a paired samples T-test. All assumptions were checked using plots (scatterplots for linearity and histograms for normality). Mean differences between baseline and follow-up for relevant anthropometric and cardiac ultrasonographic variables were calculated and compared between the intervention and control arm using an independent samples t-test. As sensitivity analyses, Mann–Whitney U-tests were performed as well. IBM SPSS Statistics for Windows (version 28.0; Armonk, NY, USA) was used for the aforementioned statistical analyses. A two-sided p-value ≤ 0.05 was considered statistically significant.
Results
Sixty patients were included and randomised to the intervention and control arms. After inclusion, one patient in the intervention arm was still excluded, due to a prolactinoma, resulting in 29 patients in the intervention arm and 30 in the control arm. Baseline characteristics of this study population are shown in Table 1.
Baseline characteristics of the overall study population with a comparison between the intervention and control group. Between brackets: standard deviation
* chi squared test.
The number of patients that received a cardiac ultrasonographic examination both at baseline and one-year follow-up was limited to 18 (intervention group n = 10, control group n = 8). The group that did receive both ultrasound examinations (n = 18) had similar characteristics as the group that did not (n = 41), and no statistically significant differences between these groups were found (Table 2).
Baseline characteristics of patients who did and who did not receive both ultrasounds (at baseline and follow-up). Between brackets: standard deviation
*chi squared test.
The prevalences of different cardiac geometries before and after surgery in the intervention and control groups are shown in Table 3. Scatter plots depicting left ventricular mass index plotted against relative wall thickness for both the baseline and follow-up are shown in Figures 1 and 2. The horizontal and vertical dashed black lines represent the respective thresholds for left ventricular mass index and relative wall thickness. Values exceeding these lines respectively indicate left ventricular hypertrophy and concentricity. Coloured overlays indicate four quadrants of left ventricular geometry, namely normal, eccentric left ventricular hypertrophy, concentric left ventricular hypertrophy and concentric left ventricular remodelling.
Scatter plot showing relative wall thickness (RWT) by left ventricular mass index (LVMI) by study arm, at baseline. Note: this includes all patients who received an ultrasound examination at baseline (n = 43), including those who did not receive a follow-up ultrasound. A left ventricular mass ≥ 38.6 g/m2.7 is hypertrophy. A RWT of >0.41 is either concentric hypertrophy or concentric remodelling.
Scatter plot showing relative wall thickness (RWT) by left ventricular mass index (LVMI) by study arm, at one-year follow-up. Note: this chart includes all patients who received an ultrasound examination at follow-up (n = 26), including those who did not receive a baseline ultrasound. A left ventricular mass ≥ 38.6 g/m2.7 is hypertrophy. A RWT of >0.41 is either concentric hypertrophy or concentric remodelling.
The prevalence of different geometries in the overall study population, intervention group and control group at baseline and one-year follow-up
Normal = LVMI and RWT within physiological range. cRML = concentric remodelling. eLVH = eccentric left ventricular hypertrophy. cLVH = concentric left ventricular hypertrophy.
Table 4 combines the outcomes of paired samples T-tests and independent samples T-tests comparing anthropometric and cardiac ultrasonographic findings for the intervention and control group between baseline and follow-up. Paired samples T-tests within the intervention group showed a reduction of BMI by 5.64 kg/m2 (95% CI: −8.37 – −2.91), BMI z score by 0.49 (95% CI: −0.07 – −0.01), relative wall thickness by 0.04 (95% CI: −0.07 – −0.00), and left ventricular mass index by 8.21 g/m2.7 (95% CI: −15.43 – −0.98). In the control group, we did not observe a statistically significant change in these variables. When comparing the mean changes to determine intervention effects, we found statistically significant intervention effects for BMI and BMI Z-score, but not for cardiac ultrasonographic variables. Although not significant, an intervention effect was found for relative wall thickness, which decreased by 0.04 in the intervention group and 0.00 in the control group (intervention effect −0.04; 95% CI: −0.07 – 0.00), suggesting amelioration of concentricity.
Overview comparing anthropometric and cardiac ultrasonographic variables between baseline and follow-up for the intervention (n = 10) and control (n = 8) arms. Between brackets: standard deviation
Statistical significance of change within a treatment arm was evaluated using a paired samples T-test, with bold 95% confidence intervals indicating significance. Intervention effects were determined via independent samples t-tests assessing the difference in mean change between the intervention and control arms. The respective Mann–Whitney U P-values for the change scores of the eight included variables from BMI to LVMI were 0.001, 0.001, 0.203, 0.408, 0.460, 0.083, 1.000, and 0.515.
Since sample sizes were small, a line chart visualising the evolution of left ventricular mass index in the individual patients of the intervention and control groups is provided in Figure 3. The dashed red horizontal line is the threshold for left ventricular hypertrophy. This illustration suggests a global trend towards either improved or stagnant left ventricular mass index in the intervention group. By contrast, in the control group, multiple patients had a worsening of their left ventricular mass index, some had stable left ventricular mass index, and some had improvements of left ventricular mass index. Importantly, the three patients in the control group who had a left ventricular mass index improve from above left ventricular hypertrophy threshold to below left ventricular mass index threshold to 23.65, 28.70, and 33.61 g/m2.7 were not found to have a relevant increase in body length or a significant decrease in BMI.
A line chart visually depicting the evolution of the left ventricular mass index for the 11 patients in the intervention group and 8 patients in the control group who received both a baseline and control ultrasound. The red horizontal line indicates the level of left ventricular hypertrophy.
Discussion
In our defined population of severely obese adolescents who received both a baseline and follow-up ultrasound one year after laparoscopic adjustable gastric banding + multidisciplinary lifestyle intervention or multidisciplinary lifestyle intervention alone (control), we observed a significant BMI change in the intervention group compared to the control group of −6.75 kg/m2 (95% CI: −10.28 – −3.23). For left ventricular mass index and relative wall thickness, we saw no significant intervention effect, though both values did significantly decrease within the intervention arm: relative wall thickness −0.04 (95% CI: −0.07 – 0.00) and left ventricular mass index −8.21 g/m2.7 (95% CI: −15.43 – −0.98). At baseline, 54.5% of patients in the intervention arm had an abnormal cardiac morphology (concentric left ventricular remodelling 9.1%, eccentric left ventricular hypertrophy 36.4%, and concentric left ventricular hypertrophy 9.1%), compared to 13.3% (only eccentric left ventricular hypertrophy) at one-year follow-up.
Studies evaluating the reversibility of cardiac changes with weight loss in severely obese adolescents are sparse and limited to the studies by Ippisch and Michalsky. Reference Ippisch, Inge, Daniels, Wang, Khoury and Witt18,Reference Michalsky, Raman, Teich, Schuster and Bauer19 Despite methodological differences, especially applicable to the study of Michalsky, our findings seem consistent.
Ippisch et al. conducted a prospective cohort study comparing BMI and cardiac ultrasonographic findings one year after Roux-en-Y gastric bypass surgery. They consecutively included all adolescents (age ≤ 19 years) who were to receive Roux-en-Y gastric bypass and who had a preoperative ultrasound examination. Comparably to our study, they calculated left ventricular mass using the Devereux formula and indexed it by height in metres to the power of 2.7. However, they used the adult threshold of left ventricular mass index of 51 g/m2.7. Of 67 potentially eligible adolescents, only 38 had the follow-up ultrasound examination and were thus eligible for examination. Of these, 29 were females and the age range was 13 to 19 years (average ± SD 16 ± 1 years), with an average pre- and postoperative BMI of 60 ± 9 and 40 ± 8 kg/m2, respectively. The pre- and postoperative left ventricular mass index values were 54 ± 13 and 42 ± 10 g/m2.7 (p < 0.001), respectively. The pre- and postoperative relative wall thickness values were 0.41 ± 0.07 and 0.33 ± 0.08 (p < 0.001), respectively. Interestingly, the baseline values of BMI, left ventricular mass index, and relative wall thickness were much higher than in our population, where we observed a baseline BMI of 44.27 ± 0.54 kg/m, left ventricular mass index of 40.41 ± 8.26, and relative wall thickness of 0.37 ± 0.047. Moreover, their degree of weight loss was higher, likely due to both the significantly higher baseline BMI and the known higher efficacy of Roux-en-Y gastric bypass compared to laparoscopic adjustable gastric banding. Compared to our study, Ippisch et al. specified left ventricular geometrical subtypes, reporting uniform improvement of all subtypes. The slight majority of abnormal geometries was concentric left ventricular hypertrophy, of which the prevalence decreased from 28% preoperatively to 3% postoperatively. In our study, the majority of abnormal geometries was eccentric left ventricular hypertrophy (48.8% for the overall baseline), possibly reflecting the higher baseline BMI in Ippisch’s study, which may be related to comorbidities that predispose to concentric morphology, e.g. hypertension.
Michalsky et al. performed a retrospective analysis of adolescents undergoing weight loss surgery and who had cardiac magnetic resonance imaging examinations. Reference Michalsky, Raman, Teich, Schuster and Bauer19 They analysed ten subjects (9 female), with a mean age and BMI of 17.4 ± 1.9 years and 50.33 ± 10.21 kg/m2, respectively. They did not calculate the left ventricular mass index, but only reported left ventricular mass, which was calculated using software analysis. An average BMI reduction of 31% was achieved (resultant BMI 34.6 ± 4.16 kg/m2, p < 0.01) within one year after either Roux-en-Y gastric bypass or laparoscopic adjustable gastric banding. Only four patients received both the pre- and postoperative MRI, yielding a reduction in left ventricular mass of 8 g (range 2–12 g). Despite the differences in methodology (e.g. different method of calculating left ventricular mass, absence of height indexing, different imaging modality), heterogeneous population, and quality issue (e.g. retrospective nature and low sample size), Michalsky’s study suggests a similar effect of weight loss on cardiac mass.
Other data concerning reversibility of left ventricular geometrical changes in the severely obese are entirely limited to adults and were meta-analysed by Sargsyan et al. in 2023. Reference Sargsyan, Chen, Aggarwal, Fadel, Fehervari and Ashrafian20 Their proportional analysis of 45 studies with 2361 patients demonstrated a 12.2% decrease (p < 0.001) in left ventricular mass index after surgery, which is at least directionally consistent with our findings. Unfortunately, the vast majority of these studies are observational, and a significant number is retrospective. Moreover, only two studies evaluated paediatric populations, as aforementioned by Ippisch et al. and Michalsky et al.
The mechanisms explaining the amelioration of both left ventricular form and function after weight loss are diverse and may be grossly categorised as haemodynamic and nonhaemodynamic factors, the latter including metabolic and endocrine changes. Reference Toprak, Wang, Chen, Paul, Srinivasan and Berenson21,Reference Murdolo, Angeli, Reboldi, Di Giacomo, Aita and Bartolini22
Haemodynamic factors include, on the one hand, a reduction of preload or total circulating volume reflecting an improvement in body mass, and on the other hand, a reduction of afterload through both a decrease of body mass and conduit artery stiffness. Reference Kindel and Strande23,Reference Vasan24 The reduction of left ventricular mass index in our study population likely reflects haemodynamic factors not relating to blood pressure, considering that the systolic and diastolic blood pressures in our study population were normal as described earlier, with z scores of 0.7 (± 1.0) and 0.3 (± 0.8), respectively. Reference Talib, Roebroek, Paulus, van Loo, Winkens and Bouvy3
An important metabolic factor that likely mediates left ventricular hypertrophy and remodelling is dyslipidaemia. Reference Sundström, Lind, Vessby, Andrén, Aro and Lithell25–Reference Zhang, Li, Shi, Zhang and Sun27 Studies, including our baseline study of this population, have established an independent association between increased left ventricular mass and facets of dyslipidaemia such as higher LDL cholesterol, lower HDL cholesterol, and higher triglyceride levels. Reference Talib, Roebroek, Paulus, van Loo, Winkens and Bouvy3,Reference Aung, Sanghvi, Piechnik, Neubauer, Munroe and Petersen28 A causal mechanism is likely, which for cholesterol centres around the Rho signalling protein family that regulates the process of cell hypertrophy. Reference Shimokawa and Takeshita29–Reference Viana Gonçalves, Cerdeira, Poletti Camara, Dias Garcia, Ribeiro Pereira Lima Brigagão and Bessa Veloso Silva32 Rho proteins are post-translationally modified by products of the metabolism of cholesterol by HMG-CoA reductase. Indeed, animal models inhibiting HMG-CoA reductase using statins have resulted in regression of cardiac hypertrophy. Reference Barbosa, Costa, Awata, Singh, Alves and Bruder-Nascimento31,Reference Patel, Nagueh, Tsybouleva, Abdellatif, Lutucuta and Kopelen33
An Important endocrine factor that is associated with adverse left ventricular geometry is insulin resistance. Reference Sciacqua, Cimellaro, Mancuso, Miceli, Cassano and Perticone34,Reference Mureddu, Greco, Rosato, Cella, Vaccaro and Contaldo35 Multiple studies, including our baseline study, have described independent associations that are likely causal and that pertain to both eccentric and concentric patterns of remodelling and hypertrophy. Reference Talib, Roebroek, Paulus, van Loo, Winkens and Bouvy3,Reference Verdecchia, Reboldi, Schillaci, Borgioni, Ciucci and Telera36–Reference El Tantawy, Fadel, Abdelrahman, Nabhan, Ibrahim and Fattouh41 This may be the result of direct trophic effects of insulin on cardiomyocytes, characterised by the induction of protein synthesis. Reference Mohan, Dihoum, Mordi, Choy, Rena and Lang42
Although in our previous study we described independent associations between interventricular septal thickness and LDL-c, HDL-c, triglyceride levels and Homeostatic Model Assessment for Insulin Resistance (HOMA-IR), we did not perform such an analysis in this study, owing to sample size limitations. Reference Talib, Roebroek, Paulus, van Loo, Winkens and Bouvy3
This is the first RCT evaluating the effect of weight loss surgery on cardiac geometry in severely obese adolescents. We have analysed a homogeneous population of post-pubertal adolescents who received either laparoscopic adjustable gastric banding + multidisciplinary lifestyle intervention or continued multidisciplinary lifestyle intervention and who were ultrasonographically analysed by a single experienced paediatric cardiologist.
One limitation of our study was the degree of missing ultrasonographic data. This was due to the unavailability of the cardiologist at sampling dates, which is related to the complexity of the BASIC Trial, where a variety of clinical examinations (e.g. DEXA scan, glucose challenge, polysomnography, abdominal ultrasonography) had to be planned at each hospital visit. Potentially, this reduces overall comparability between baseline and follow-up. There is no reason to suggest bias regarding which patients did and did not receive both a baseline and follow-up ultrasound. Reassuringly, the baseline demographic, anthropometric and ultrasonographic variables were similar for patients who did and who did not receive both ultrasound examinations, yielding no statistically significant differences in these variables.
Although we established statistically significant improvements in left ventricular mass index within the intervention arm, the sample size limitations have hampered our ability to determine intervention effects. We suggest that future studies prioritise sample sizes and completeness of cardiac data in order to further elucidate the determinants of amelioration of left ventricular geometry after weight loss.
In conclusion, our findings suggest that even moderate weight loss through bariatric surgery might be associated with an improvement of left ventricular geometry in severely obese adolescents.
Funding Statement
Open access funding provided by Maastricht University.
Open access
