https://orcid.org/0000-0002-1958-2800
Abstract
Evidence suggests that metabolic adaptation occurs after bariatric surgery such that resting energy expenditure (REE) declines more than accounted for by body weight or body composition changes in adults. Little is known about REE and metabolic adaptation among adolescents after bariatric surgery.
To examine changes in REE and metabolic adaptation among adolescents at 12 months (12M) after bariatric surgery.
Pediatric hospital, Canada.
Adolescents undergoing Roux-en-Y gastric bypass (RYGB) or sleeve gastrectomy (SG) were followed. Bioelectrical impedance analysis and indirect calorimetry were completed to measure body composition and REE, respectively. Predicted REE was calculated using the Mifflin equation before and after bariatric surgery and a predictive equation using preoperative data.
Among 20 patients (15 girls), the mean age and body mass index at surgery were 17.2 ± 0.8 years and 48.7 ± 7.4 kg/m2, respectively. REE had decreased by 548.3 kcal/d at 12M postoperatively (P < 0.001). Metabolic adaptation, determined by two procedures, was negative and significantly different from baseline (P < 0.05). When stratified by surgery type, REE change at 12M was not significantly different (RYGB, −494.0 ± 260.9 kcal/d, n = 11; SG, −614.6 ± 344.4 kcal/d, n = 9; P = 0.384). Among 13 patients with REE data at 6 and 12M, no statistically significant difference was found (P = 0.368).
Predicted and measured REE was 19% and 25% lower at 12M, respectively, irrespective of bariatric surgery type. Metabolic adaptation might predispose adolescents to weight regain after bariatric surgery and warrants careful nutritional management and counseling.
The high prevalence and increase in severe obesity among children and adolescents (1, 2) has raised concern for increased health complications in adulthood. Obesity elevates the risk of developing medical comorbidities, including cardiovascular disease, hypertension, dyslipidemia, obstructive sleep apnea, and type 2 diabetes (3), contributing to substantial morbidity and premature mortality (4, 5). Childhood obesity prevention efforts have included family-based lifestyle modifications (6); however, these efforts have been ineffective in preventing the increase in obesity prevalence globally. Thus, better treatments to improve long-term health are needed. Without more effective treatment strategies, obesity will persist and continue to have major economic effects on the health care system (7).
Accumulating evidence has supported bariatric surgery as an efficacious intervention among adolescents to promote weight loss and reduce medical comorbidities, and, as a result, has become more commonly offered to this particular population (8). However, after excess weight loss, weight regain can be a chronic challenge for people living with obesity (9). Clinical findings have demonstrated a wide variation in weight loss and weight regain after bariatric surgery (10–14). In a study inclusive of 50 adolescents who had undergone bariatric surgery, Ryder et al. (14) reported weight regain in 54% of the adolescents. From 1 year after surgery to ≥5 years of follow-up, body mass index had increased by 30% in the group of adolescents who regained weight and had decreased by 3% in those who maintained their weight loss (14). Overall, the contributing factors to weight regain susceptibility and trends have been unclear (14); therefore, the prevention of this occurrence is minimized. The possible challenges with preventing weight regain could be associated with decreased leptin levels after fat loss, which drives energy intake (15). Evidence has also shown that muscle loss is a potential factor that could predict weight regain (16). A better understanding of weight loss trajectories and weight regain after bariatric surgery in adolescents will aid in the development of strategies to improve health outcomes after surgery.
Weight regain after excess weight loss could be partially related to changes in resting energy expenditure (REE) and metabolic adaptation, which refers to a reduced REE associated with weight loss not accounted for by changes in body weight or body composition (17–19). This is determined by a greater decline in measured REE using indirect calorimetry compared with predicted REE using predictive equations.
Among adults, bariatric surgery contributing to weight loss has shown substantial changes in REE (17, 20–27) and metabolic adaptation (17, 21, 22, 26). Wilms et al. (26) found the measured REE was reduced by 20.4% at 12 months (12M) after Roux-en-Y gastric bypass (RYGB) in adults. Similarly, Bettini et al. (17) reported a reduced REE by 27.3% at 12M and metabolic adaptation of −199 ± 238 kcal/d, calculated as the difference between the measured and predicted REE after sleeve gastrectomy (SG). Conflicting results have been reported by studies of adults, with a few studies showing no substantial changes in REE postoperatively after adjustment for changes in weight or body composition (20, 28, 29). These findings suggest that no metabolic adaptation occurs after bariatric surgery (20, 28, 29).
The pediatric data on REE after bariatric surgery are more limited. One study conducted among adolescents with obesity reported a decrease in basal metabolic rate by 18.2% at 1.5 months, which remained unchanged at 6 months (6M) and 12M after RYGB (30). The investigators did not determine or discuss metabolic adaptation among the adolescents included in their study (30). From the limited data, it could be speculated that reductions in REE might be similar between adolescents and adults after bariatric surgery; however, any metabolic adaptation similarities or differences are unknown. In theory, stage of puberty could potentially affect energy expenditure via differences in substrate metabolism (31, 32); however, the patients in this study were postpubertal. To the best of our knowledge, no studies have reported on metabolic adaptation or adaptive thermogenesis in children or adolescents compared with adults.
In addition, to the best of our knowledge, no other studies of adolescents have reported on changes in REE after bariatric surgery to treat obesity. Therefore, the primary objective of the present study was to examine changes in REE and metabolic adaptation among adolescents at 12M after bariatric surgery. Secondary analyses were completed to investigate correlations between changes in REE with body composition parameters at 12M, and to investigate changes in REE and metabolic adaptation after laparoscopic RYGB compared with laparoscopic SG at 12M postoperatively in adolescents.
Materials and Methods
The SickKids Team Obesity Management Program (STOMP) is an interdisciplinary clinic at the Hospital for Sick Children that has been providing care to children and adolescents with severe obesity since January 2010. All adolescents from the STOMP clinic who had undergone RYGB or SG from October 2010 to October 2017 and had attended their follow-up appointments at 12M postoperatively were included in the present study. The patients and families from the STOMP clinic provided written informed consent for evaluation of the anthropometric and metabolic data collected during the clinical assessments. The research ethics board at the Hospital for Sick Children approved the present study.
The patients completed anthropometric, body composition, and indirect calorimetry measurements at the following time points: baseline (before surgery) and at 6M and 12M postoperatively. During the appointments, the patients met with a multidisciplinary team that included physicians, nurses, dieticians, a psychologist, an exercise therapist, and a social worker. Diet and exercise was advised on an individual basis before and after bariatric surgery by trained personnel in accordance with the recommended guidelines (33, 34). The standing height was measured using a stadiometer to the nearest 0.1 cm (Holtain Ltd., Crymych, UK). The body weight was measured and recorded within 0.1 kg using a standing scale (model no. 5002, Scaletronix, Carol Stream, IL). The percentage of total weight loss (%TWL) was calculated from the follow-up data. To estimate the fat mass (FM) and fat-free mass (FFM), hand-to-foot bioelectrical impedance analysis (BIA) was performed with the participants supine and in a fasting state (IMP SFB7 Multifrequency BIA, Impedimed, Pinkenba, Australia), with the analyzer fixed at a frequency BIA of 50 kHz at 800 mÅ. The patients were asked to fast overnight and advised to avoid both food and water. The BIA equation that was included in the device software was not used. Instead, resistance and reactance measurements from the BIA were entered into the equations developed by Gray et al. (35), because these equations most accurately predicted the FFM and FM in adolescents with obesity compared with other BIA equations (36).
REE was measured in the fasting condition with indirect calorimetry using a ventilated canopy connected to a metabolic cart (Vmax Encore V29C; Sensormedics Corp., Yorba Linda, CA). The system was calibrated with a known concentration of gases before every measurement, in accordance with the supplier’s instructions. Oxygen uptake and carbon dioxide production were averaged at 1-minute intervals. The measured REE was calculated using the averaged VO2 and VCO2 values according to the Weir equation (37). To calculate the predicted REE, the Mifflin equation was used, which has been shown to best predict REE in adolescents with severe obesity (38, 39). As a second method to determine the predicted REE, preoperative data were used to compute a predictive REE equation using linear regression analysis, similar to other studies (17, 22, 40). Metabolic adaptation was defined as the difference between the measured REE and predicted REE determined using the Mifflin equation or the predictive REE equation from our sample. We have referred to the two methods as metabolic adaption (Mifflin equation) and metabolic adaptation (residuals). Metabolic adaptation was considered present if the value was negative and differed from that at baseline (18, 40).
Statistical analysis
The data were analyzed using statistical analysis software (IBM SPSS Statistics, version 20.0.0; IBM Corp., Armonk, NY). Descriptive statistics are expressed as the mean ± SD. All variables were checked for normality using the Shapiro-Wilk test. Repeated measures ANOVA with the Fisher least significant difference post hoc analysis was used to identify differences at baseline and 6M and 12M postoperatively for REE and metabolic adaptation. Paired t tests were used to compare the pre- and postoperative outcomes at 12M. Pearson r correlations or Spearman rho correlations were conducted to examine the correlations between the measured REE at 12M and %TWL and body composition parameters at 12M. Independent t tests were used to test for differences in outcomes by surgery type (RYGB vs SG) at 12M after bariatric surgery. Statistical significance was defined as P ≤ 0.05.
Results
From October 2010 to October 2017, 34 adolescent patients had undergone bariatric surgery. Four patients were excluded from the analysis (three had undergone laparoscopic adjustable gastric banding and one had completed their follow-up appointments at another institution). In addition, not all the patients completed the body composition and/or REE measurements at each anticipated follow-up time point. Thus, 28 of the 30 patients had REE measured before surgery, 18 had REE measured at 6M, and 22 had REE measured at 12M. When examining the differences for measured REE and metabolic adaptation (Mifflin equation), 20 patients had both baseline and 12M data available and 13 patients had data at 6M available. The body composition measurements were collected for 19 of the 20 patients who had also completed the REE measurements at 12M. Therefore, 19 patients had metabolic adaptation determined using the predictive equation at baseline and at 12M postoperatively. Using the preoperative data (n = 26), a predictive REE equation (y = 10.733 × FM + 12.727 × FFM + 595.071) was computed to determine the predicted REE at 6M and 12M. The FM and FFM explained 73% of the variance in the model (r2 = 0.730). Sex was not a statistically significant predictor of REE in the model (P = 0.362) and was not included in the equation. The patient characteristics are reported in Table 1.
Patient Characteristics
| Characteristic | Baseline (n = 20) | 6 Months (n = 13) | 12 Months (n = 20) |
|---|---|---|---|
| Sex | |||
| Male | 5 | 3 | 5 |
| Female | 15 | 10 | 15 |
| Surgical procedure | |||
| RYGB | 11 | 8 | 11 |
| SG | 9 | 5 | 9 |
| Age, y | 17.2 ± 0.8 | 18.0 ± 0.6 | 18.3 ± 0.9 |
| Weight, kg | 137.2 ± 27.1 | 92.4 ± 21.8a | 94.9 ± 22.1a,b |
| Height, cm | 167.5 ± 10.1 | 165.2 ± 8.6 | 167.4 ± 10.3 |
| BMI, kg/m2 | 48.7 ± 7.4 | 33.7 ± 6.8a | 33.7 ± 6.1a,b |
| FM, kg | 71.4 ± 16.9 (n = 20) | 37.0 ± 13.6 (n = 8) | 36.8 ± 13.4a (n = 19) |
| FFM, kg | 65.4 ± 10.9 (n = 20) | 51.4 ± 4.7 (n = 8) | 56.3 ± 8.7a,b (n = 19) |
| %TWL | NA | 28.4 ± 8.4 | 30.6 ± 9.8b |
| Characteristic | Baseline (n = 20) | 6 Months (n = 13) | 12 Months (n = 20) |
|---|---|---|---|
| Sex | |||
| Male | 5 | 3 | 5 |
| Female | 15 | 10 | 15 |
| Surgical procedure | |||
| RYGB | 11 | 8 | 11 |
| SG | 9 | 5 | 9 |
| Age, y | 17.2 ± 0.8 | 18.0 ± 0.6 | 18.3 ± 0.9 |
| Weight, kg | 137.2 ± 27.1 | 92.4 ± 21.8a | 94.9 ± 22.1a,b |
| Height, cm | 167.5 ± 10.1 | 165.2 ± 8.6 | 167.4 ± 10.3 |
| BMI, kg/m2 | 48.7 ± 7.4 | 33.7 ± 6.8a | 33.7 ± 6.1a,b |
| FM, kg | 71.4 ± 16.9 (n = 20) | 37.0 ± 13.6 (n = 8) | 36.8 ± 13.4a (n = 19) |
| FFM, kg | 65.4 ± 10.9 (n = 20) | 51.4 ± 4.7 (n = 8) | 56.3 ± 8.7a,b (n = 19) |
| %TWL | NA | 28.4 ± 8.4 | 30.6 ± 9.8b |
Abbreviations: BMI, body mass index; NA, not applicable.
P < 0.001 compared with baseline.
P < 0.05 compared with 6 months.
Patient Characteristics
| Characteristic | Baseline (n = 20) | 6 Months (n = 13) | 12 Months (n = 20) |
|---|---|---|---|
| Sex | |||
| Male | 5 | 3 | 5 |
| Female | 15 | 10 | 15 |
| Surgical procedure | |||
| RYGB | 11 | 8 | 11 |
| SG | 9 | 5 | 9 |
| Age, y | 17.2 ± 0.8 | 18.0 ± 0.6 | 18.3 ± 0.9 |
| Weight, kg | 137.2 ± 27.1 | 92.4 ± 21.8a | 94.9 ± 22.1a,b |
| Height, cm | 167.5 ± 10.1 | 165.2 ± 8.6 | 167.4 ± 10.3 |
| BMI, kg/m2 | 48.7 ± 7.4 | 33.7 ± 6.8a | 33.7 ± 6.1a,b |
| FM, kg | 71.4 ± 16.9 (n = 20) | 37.0 ± 13.6 (n = 8) | 36.8 ± 13.4a (n = 19) |
| FFM, kg | 65.4 ± 10.9 (n = 20) | 51.4 ± 4.7 (n = 8) | 56.3 ± 8.7a,b (n = 19) |
| %TWL | NA | 28.4 ± 8.4 | 30.6 ± 9.8b |
| Characteristic | Baseline (n = 20) | 6 Months (n = 13) | 12 Months (n = 20) |
|---|---|---|---|
| Sex | |||
| Male | 5 | 3 | 5 |
| Female | 15 | 10 | 15 |
| Surgical procedure | |||
| RYGB | 11 | 8 | 11 |
| SG | 9 | 5 | 9 |
| Age, y | 17.2 ± 0.8 | 18.0 ± 0.6 | 18.3 ± 0.9 |
| Weight, kg | 137.2 ± 27.1 | 92.4 ± 21.8a | 94.9 ± 22.1a,b |
| Height, cm | 167.5 ± 10.1 | 165.2 ± 8.6 | 167.4 ± 10.3 |
| BMI, kg/m2 | 48.7 ± 7.4 | 33.7 ± 6.8a | 33.7 ± 6.1a,b |
| FM, kg | 71.4 ± 16.9 (n = 20) | 37.0 ± 13.6 (n = 8) | 36.8 ± 13.4a (n = 19) |
| FFM, kg | 65.4 ± 10.9 (n = 20) | 51.4 ± 4.7 (n = 8) | 56.3 ± 8.7a,b (n = 19) |
| %TWL | NA | 28.4 ± 8.4 | 30.6 ± 9.8b |
Abbreviations: BMI, body mass index; NA, not applicable.
P < 0.001 compared with baseline.
P < 0.05 compared with 6 months.
To date, the mean %TWL among the patients with completed follow-up data available at 12M (n = 27) was 24.2% ± 7.9% and 28.5% ± 10.3% at 6M and 12M, respectively. The %TWL was significantly greater statistically at 12M than at 6M (P = 0.001). When stratified by surgery type, the RYGB group (n = 12) had a significantly greater %TWL statistically compared with the SG group (n = 15) at 6M (RYGB, 27.7% ± 7.2%; SG, 21.4% ± 7.6%; P = 0.038) but not at 12M (RYGB, 32.1% ± 8.2%; SG, 25.7% ± 11.1%; P = 0.107).
The REE and metabolic adaptation values at baseline and 6M and 12M postoperatively are shown in Table 2. Repeated measures ANOVA showed a statistically significant time effect for REE (P < 0.001) and metabolic adaptation (Mifflin equation) (P = 0.043). In 13 patients, the REE had decreased by 475.1 kcal/d (22%) at 6M (P < 0.001) and 545.5 kcal/d (26%) at 12M (P < 0.001) after bariatric surgery compared with the REE at baseline. The metabolic adaptation (Mifflin equation) at 12M (−103.7 ± 214.5 kcal/d) was negative, and the difference was statistically significant from that at baseline (20.4 ± 52.7 kcal/d; P = 0.043). A trend was found at 6M (−86.4 ± 284.7 kcal/d; P = 0.051). No statistically significant differences were found in REE (P = 0.368) or metabolic adaptation (Mifflin equation; P = 0.775) between 6M and 12M of follow-up. Repeated measures ANOVA was not performed for metabolic adaptation (residuals) because only 8 patients could be included to compare the values at baseline and 6M and 12M postoperatively. Therefore, paired t tests were used to detect differences in REE and metabolic adaptation (Mifflin and residuals) between baseline and 12M postoperatively (Fig. 1), which consisted of a larger number of patients with indirect calorimetry data compared with the data from 6M and 12M.
REE and Metabolic Adaptation at Baseline Before Bariatric Surgery and 6M and 12M After Surgery
| Variable | Baseline (n = 20) | 6 Months (n = 13) | 12 Months (n = 20) |
|---|---|---|---|
| Measured REE, kcal/d | 2217.2 ± 380.4 | 1644.1 ± 359.3a | 1669.0 ± 271.8a |
| Predicted REE, Mifflin equation, kcal/d | 2204.3 ± 362.6 | 1730.5 ± 292.5 | 1776.3 ± 315.1 |
| Metabolic adaptation, Mifflin equation, kcal/d | 12.9 ± 192.8 | −86.4 ± 284.7 | −107.3 ± 184.5b |
| Predicted REE, regression equation, kcal/d | 2163.7 ± 241.9 (n = 19) | 1646.7 ± 191.5 (n = 8) | 1706.5 ± 241.9 (n = 19) |
| Metabolic adaptation, residuals, kcal/d | 29.3 ± 193.1 (n = 19) | −70.1 ± 182.7 (n = 8) | −59.8 ± 174.4b (n = 19) |
| Variable | Baseline (n = 20) | 6 Months (n = 13) | 12 Months (n = 20) |
|---|---|---|---|
| Measured REE, kcal/d | 2217.2 ± 380.4 | 1644.1 ± 359.3a | 1669.0 ± 271.8a |
| Predicted REE, Mifflin equation, kcal/d | 2204.3 ± 362.6 | 1730.5 ± 292.5 | 1776.3 ± 315.1 |
| Metabolic adaptation, Mifflin equation, kcal/d | 12.9 ± 192.8 | −86.4 ± 284.7 | −107.3 ± 184.5b |
| Predicted REE, regression equation, kcal/d | 2163.7 ± 241.9 (n = 19) | 1646.7 ± 191.5 (n = 8) | 1706.5 ± 241.9 (n = 19) |
| Metabolic adaptation, residuals, kcal/d | 29.3 ± 193.1 (n = 19) | −70.1 ± 182.7 (n = 8) | −59.8 ± 174.4b (n = 19) |
P < 0.001 compared with baseline.
P < 0.05 compared with baseline.
REE and Metabolic Adaptation at Baseline Before Bariatric Surgery and 6M and 12M After Surgery
| Variable | Baseline (n = 20) | 6 Months (n = 13) | 12 Months (n = 20) |
|---|---|---|---|
| Measured REE, kcal/d | 2217.2 ± 380.4 | 1644.1 ± 359.3a | 1669.0 ± 271.8a |
| Predicted REE, Mifflin equation, kcal/d | 2204.3 ± 362.6 | 1730.5 ± 292.5 | 1776.3 ± 315.1 |
| Metabolic adaptation, Mifflin equation, kcal/d | 12.9 ± 192.8 | −86.4 ± 284.7 | −107.3 ± 184.5b |
| Predicted REE, regression equation, kcal/d | 2163.7 ± 241.9 (n = 19) | 1646.7 ± 191.5 (n = 8) | 1706.5 ± 241.9 (n = 19) |
| Metabolic adaptation, residuals, kcal/d | 29.3 ± 193.1 (n = 19) | −70.1 ± 182.7 (n = 8) | −59.8 ± 174.4b (n = 19) |
| Variable | Baseline (n = 20) | 6 Months (n = 13) | 12 Months (n = 20) |
|---|---|---|---|
| Measured REE, kcal/d | 2217.2 ± 380.4 | 1644.1 ± 359.3a | 1669.0 ± 271.8a |
| Predicted REE, Mifflin equation, kcal/d | 2204.3 ± 362.6 | 1730.5 ± 292.5 | 1776.3 ± 315.1 |
| Metabolic adaptation, Mifflin equation, kcal/d | 12.9 ± 192.8 | −86.4 ± 284.7 | −107.3 ± 184.5b |
| Predicted REE, regression equation, kcal/d | 2163.7 ± 241.9 (n = 19) | 1646.7 ± 191.5 (n = 8) | 1706.5 ± 241.9 (n = 19) |
| Metabolic adaptation, residuals, kcal/d | 29.3 ± 193.1 (n = 19) | −70.1 ± 182.7 (n = 8) | −59.8 ± 174.4b (n = 19) |
P < 0.001 compared with baseline.
P < 0.05 compared with baseline.
(A) REE at baseline (black bars) and 12M (white bars) after bariatric surgery (n = 20). Paired t test results showed REE was significantly different after bariatric surgery (P < 0.001). *Statistical significance was set at P < 0.05. (B) Metabolic adaptation (Mifflin equation) at baseline (black bars) and 12M (white bars) after bariatric surgery (n = 20). Paired t test results showed metabolic adaptation was significantly different statistically after bariatric surgery (P = 0.008). *Statistical significance was set at P < 0.05. (C) Metabolic adaptation (residuals) at baseline (black bars) and 12M (white bars) after bariatric surgery (n = 19). Paired t test results showed metabolic adaptation was significantly different statistically after bariatric surgery (P = 0.028). *Statistical significance was set at P < 0.05.
Correlations showed that a greater reduction in REE at 12M was associated with a smaller %TWL at 12M (r = −0.726; P < 0.001), a greater change in FFM at 12M (r = 0.629; P = 0.004), and a greater change in FM at 12M (r = 0.933; P < 0.001). However, no statistically significant correlation was found for metabolic adaptation (Mifflin equation and residuals) at 12M with the %TWL at 12M, change in FFM at 12M, or change in FM at 12M (P > 0.05). Similar results were found for the change in REE at 6M and the change in FM and FFM (P < 0.001). However, %TWL at 6M was not significantly associated (r = −0.442; P = 0.076). To further explore associations, the change in REE from 6M to 12M was examined. A greater reduction in REE from 6M to 12M was associated with a smaller %TWL from 6M to 12M (r = −0.726; P < 0.001). In eight patients, a greater reduction in REE from 6M to 12M postoperatively was positively associated with a greater reduction in FM from 6M to 12M postoperatively (ρ = 0.905; P = 0.002) but not with FFM at 6M to 12M postoperatively (ρ = 0.227; P = 0.588). This suggests that greater reductions in FM, which could be linked to decreased leptin levels, could be more important to consider when investigating change in REE after bariatric surgery.
The data from 12M postoperatively were used to compare measured REE and metabolic adaptation between the two surgical procedures (Fig, 2). The 20 patients included in the present analysis did not show a statistically significant difference in %TWL at 12M when stratified by surgery group (RYGB, 32.6% ± 8.3%; SG, 28.2% ± 11.4%; P = 0.330), similar to the results found for all 27 patients.
(A) REE stratified by surgical procedure: RYGB (black bars; n = 11) and laparoscopic SG (white bars; n = 9). Independent t test showed no statistically significant differences between groups (P = 0.384). Statistical significance was set at P < 0.05. (B) Metabolic adaptation (Mifflin equation) at 12M after surgery stratified by surgical procedure: RYGB (black bars; n = 11) and SG (white bars; n = 9). Independent t test showed no statistically significant differences between groups (P = 0.063). Statistical significance was set at P < 0.05. (C) Metabolic adaptation (residuals) at 12M after surgery stratified by surgical procedure: RYGB (black bars; n = 11) and SG (white bars; n = 8). Independent t test showed no statistically significant differences between groups (P = 0.369). Statistical significance was set at P < 0.05.
Discussion
The key findings from the present study were that REE decreased in adolescents after bariatric surgery and that metabolic adaptation was negative and significantly different from baseline values. No difference in measured REE at 6M and 12M was identified in a smaller group of the patients. The secondary findings suggested that the change in REE and metabolic adaptation at 12M were not different between the adolescents who had undergone RYGB compared with those who had undergone SG. However, because of the smaller sample when the patients were divided by surgical procedure, the probability of a type II error was greater, and future research is warranted to confirm that no difference exists for change in REE and metabolic adaptation between the RYGB and SG groups.
Previous studies of adults have all reported evidence of a decline in REE postoperatively (17, 20–29, 40–43), with most studies showing REE reductions of ~20% to 30% postoperatively (17, 20, 23–26, 28, 40–42). Moehlecke et al. (23) reported a mean reduction in REE at 6M postoperatively in adults of 405 kcal/d (17.6%), which was slightly lower but close to our findings for adolescents at 6M. Our results did not show differences in REE or metabolic adaptation at 12M postoperatively when stratified by the surgical procedure performed. The findings from the adult studies comparing SG and RYGB (24, 25, 28, 41) corroborate these results. To the best of our knowledge, only one study has evaluated REE in adolescents after bariatric surgery (30). In 11 subjects who had undergone RYGB, REE was lowered by ~500 kcal/d at 12M after surgery compared with REE at baseline (30). Metabolic adaptation was not calculated (30). Greater reductions in REE after bariatric surgery could be a factor contributing to suboptimal long-term weight loss and weight regain.
A smaller number of adult studies have specifically reported on metabolic adaptation after bariatric surgery (17, 20–22, 25–28, 40, 43). Unlike REE, which was been reported to decrease in all studies, metabolic adaptation has been shown to be negative (17, 21, 25, 27, 40), positive (26, 43), or remain unchanged (20, 22, 28, 29, 42). For example, among adults at 12M postoperatively, the metabolic adaptation was −199 kcal/d after SG (17) and −150 kcal/d after RYGB and SG combined (25). In contrast, Wilms et al. (26) reported that metabolic adaptation was 1.6% greater postoperatively than at baseline. Furthermore, Das et al. (28) concluded no metabolic adaptation had occurred in a cohort of adults because the reduction in REE was predicted by the decreases in FFM and FM.
In adult studies showing negative metabolic adaptation, the adaptation occurred 1 to 6 months after bariatric surgery (22, 27, 40). In our subgroup analysis, we found that metabolic adaptation was reduced (trend) at 6M after surgery and did not show a substantial reduction from 6M to 12M after surgery. In addition, Wolfe et al. (27) reported that the metabolic adaptation at 6M was substantially different from baseline but not substantially different from 6M to 24M in adults.
The reason for the conflicting data on metabolic adaptation is unclear but could have been influenced by variability between individuals (42) and relatively small sample sizes and/or the methods chosen (18). For example, metabolic adaptation has been calculated by comparing REE adjusted for FFM or FM at baseline and follow-up (20), using predicted values from a regression analysis (21, 25, 40), and by comparing the measured vs predicted REE at baseline and follow-up (17, 26, 42, 43). Future work using more consistent procedures for metabolic adaptation to allow for comparisons of data between studies is needed. We used two methods for determining metabolic adaptation in our study and found a negative and substantial metabolic adaptation at 12M using both procedures. The magnitude of the negative metabolic adaptation determined by the predicted REE values from the regression analysis was more modest than when determined by the predicted values from the Mifflin equation.
The underlying mechanisms contributing to metabolic adaptation are not well understood. Decreased circulating leptin levels (22, 42), decreased thyroid hormones linked with blunted sympathetic nervous system activity, and decreased catecholamines associated with weight loss have been potential mechanisms described (44–47). Data are available to suggest leptin supplementation to levels before weight loss can reverse the declines in energy expenditure and circulating thyroid hormones that occur with weight loss (46, 47). Leptin could affect energy expenditure and skeletal muscle function via potential changes in mitochondria (48) and uncoupling protein 2 expression (49). Changes in uncoupling protein 2 could also correspond to weight loss by moderating substrate oxidation (49). The results from other studies have supported the presence of links between metabolic adaptation and reduced weight loss (17, 26), FM loss (17), and lipid oxidation (24, 42), which could impede weight loss. Thus, changes in substrate oxidation could be connected with changes in metabolic adaptation. Overall, the mechanisms to explain the changes in REE and metabolic adaptation remain poorly understood. Future work investigating mediators of REE and metabolic adaptation are warranted.
The present study had a few limitations. We used the Mifflin equation, which includes weight, height, and age but not FM and FFM in its calculation of predicted REE. However, the Mifflin equations best predicted REE among adolescents with severe obesity compared with other available equations (39). BIA was used to estimate the body composition parameters. Compared with reference methods such as dual x-ray absorptiometry, BIA could underestimate FM in children and adolescents (50). However, BIA measurements have shown excellent reproducibility (50), which was an advantage for our study because body composition was assessed at more than one time point. Furthermore, BIA was more cost effective, accessible, and efficient to complete in the clinical setting. Another limitation of our study was the absence of a nonoperative control group. Thus, the present study was limited in determining whether changes in REE and metabolic adaptation result primarily from the bariatric surgery or from excessive weight loss, in general. However, it would be difficult to have adolescents with obesity lose the same amount of weight using conventional lifestyle interventions. Owing to inadequate time to complete all procedures during the appointments and missed follow-up appointments, not all of the patients completed the REE and/or body composition tests at baseline and after surgery at the anticipated follow-up points. Nevertheless, we were able to identify that changes in REE, and not metabolic adaptation, were associated with changes in FM and FFM in the adolescents at 12M after bariatric surgery.
The strengths of our study included the longitudinal nature of our study design, which allowed for the collection of data on REE and metabolic adaptation at various time points in adolescents who had undergone bariatric surgery. To the best of our knowledge, only one other study has reported on REE after bariatric surgery in adolescents (30). No previous studies have investigated differences in REE between RYGB and SG nor the metabolic adaptation after bariatric surgery in adolescents.
Conclusion
We have demonstrated a substantial reduction in REE after bariatric surgery at 12M among adolescents. Metabolic adaptation occurred at 6M and 12M after bariatric surgery. These changes could predispose some adolescents to weight regain. Thus, careful monitoring of energy expenditure changes and accompanying caloric reductions postoperatively, especially at 6M and 12M, would optimize health outcomes for these patients. Future investigations to elucidate the factors and mechanisms related to changes in REE and metabolic adaptation would be beneficial for optimizing nutritional and lifestyle counseling for obese adolescents.
Abbreviations:
- 6M
6 months
- 12M
12 months
- BIA
bioelectrical impedance analysis
- FFM
fat-free mass
- FM
fat mass
- REE
resting energy expenditure
- RYGB
Roux-en-Y gastric bypass
- SG
sleeve gastrectomy
- STOMP
SickKids Team Obesity Management Program
- %TWL
percentage of total weight loss
Acknowledgments
We thank the patients and families involved in the STOMP and their dedicated clinical care team. Without them, our study would not have been possible.
Financial Support: L.C. was supported by the SickKids Restracomp Fellowship (2018) and is currently supported by the Banting & Best Diabetes Centre Fellowship in Diabetes Care (2018). J.H. is supported by the SickKids University of Toronto Mead Johnson Chair in Nutritional Science, which provides unrestricted research funds.
Disclosure Summary: The authors have nothing to disclose.
