Intended for healthcare professionals

Clinical Review State of the Art Review

Emerging treatments for severe obesity in children and adolescents

BMJ 2016; 354 doi: (Published 29 September 2016) Cite this as: BMJ 2016;354:i4116
  1. Nicole Coles, pediatric endocrine fellow1,
  2. Catherine Birken, staff physician and associate professor of pediatrics2,
  3. Jill Hamilton, staff physician and professor of pediatrics1
  1. 1Division of Endocrinology, Hospital for Sick Children, University of Toronto, Toronto, ON, Canada, M5G 1X8
  2. 2Division of Pediatric Medicine, Hospital for Sick Children, University of Toronto, Toronto, Canada
  1. Correspondence to: J Hamilton Jill.hamilton{at}


Severe obesity in childhood is increasing in prevalence and is associated with considerable morbidity. Studies into pediatric obesity have focused largely on interventions that do not necessarily target the unique biologic or psychological underpinnings for the weight gain in the individual child or adolescent. Outcomes show modest improvement and are of questionable benefit for patients with severe obesity. Although weight is a commonly used outcome, other psychological and metabolic parameters including normalization of physical activity and eating behaviors should be primary outcome goals. The durability of weight loss is often limited by physiologic systems that are evolutionarily designed to promote weight gain. Drug therapies for children are limited, as is their effect on weight and metabolism. Existing drugs that are incidentally found to cause weight loss through off-target effects are being actively investigated for obesity indications. Bariatric surgery results in the most significant weight reduction, but it is associated with potential morbidity and long term data are not available for adolescents undergoing this procedure. As understanding of the biologic and psychosocial contributors to eating behaviors and body weight regulation increases, multifaceted and targeted behavioral, pharmacological, and surgical treatment algorithms should be developed and applied to target the underlying pathways involved for the individual child or adolescent with severe obesity.


Although international epidemiological studies suggest that the overall prevalence of pediatric obesity in developed countries may have plateaued, a discernible shift has occurred toward an increasing prevalence of severe obesity, representing 2.5-6% of children aged 2-19 years.1 2 3 4 Societal and environmental changes promoting consumption of calorie dense, processed foods and an increasingly sedentary lifestyle are the driving forces for the obesity epidemic, but they do not explain why some individuals are more vulnerable than others to these factors. Although single gene disorders leading to severe obesity are rare, genetics plays an important role in the establishment of body weight set point and dictates the physiologic responses to environmental moderators such as nutrition, activity, and other social behaviors. Twin study designs show that more than half of body mass index (BMI) heritability can be explained by cumulative genetic factors.5 Other factors, including the effects of copy number variants and epigenetic modification of DNA expression in fetal or early life development, may also contribute to susceptibility to obesity.6

Weight patterns in early years persist over time, with childhood obesity tracking into adulthood.7 Longitudinal cohort studies show that obesity during childhood is associated with medical comorbidity, and excess weight in childhood independently increases the risk of mortality related to the development of cardiovascular and metabolic disease in later life.8 9 10 It has also been linked with adverse effects on social, psychological, and academic development.11 Obese children are at high risk of bullying, discrimination, lower health related quality of life, and impaired mental health.1 12 13 Stigmatization of severely obese children by healthcare providers has also been described, with a negative effect on patients’ outcomes.14 These medical and social consequences of severe obesity pose a substantial threat to children’s health and wellbeing.

Existing lifestyle interventions and medical treatments have shown only a modest effect, particularly for young people with severe obesity.15 16 At the same time, our understanding of the pathophysiology of obesity has grown. Consequently, treatment has now shifted toward acknowledging the diversity of underlying mechanisms leading to weight gain and evaluating multiple psychosocial and health related outcomes.

This review will explore recent developments in our understanding of severe childhood obesity, discussing epidemiology, underlying causes, and weight regulation. It will focus on emerging, targeted medical and surgical treatments, reviewing the indications, limitations, and evidence behind their use.

Sources and selection criteria

We identified studies by searching PubMed, Medline, and Embase databases from January 2010 to December 2015. The aim was to appraise recent literature with a focus on novel or emerging themes in the treatment of severe obesity in children and adolescents. We included only studies published in English and conducted in a pediatric age group (0-18 years). We used the following search terms: “severe obesity”, “morbid obesity”, “childhood obesity”, “adolescent obesity”, “therapeutics”, “obesity treatment”, and “drug therapy”. We also identified additional references by using review articles. We reviewed and selected citations according to their relevance and study quality. We included all study types but prioritized study designs offering the highest level of evidence (that is, systematic reviews and randomized control trials). We searched to find ongoing clinical trials concerning treatments for pediatric obesity.

Definition of severe childhood obesity

Obesity is defined in the clinical setting by using body mass index (measured as kg/m2)—a calculation that allows for adjustment of weight on the basis of height and can be plotted on a growth chart for age and sex. The World Health Organization,17 the Centers for Disease Control and Prevention (CDC),18 and the International Obesity Task Force (IOTF)19 have each used different reference data and thresholds to develop their own definitions of severe obesity (fig 1).9 20 21 22 23 24


Fig 1 Proposed definitions of severe obesity. BMI=body mass index; SDS=standard deviation score

With the exception of the WHO birth to 5 years growth chart, which was developed from a healthy cohort of infants from six countries after longitudinal collection of heights and weights, all other BMI charts were developed from cross sectional population datasets in time periods before the obesity epidemic.

One problem with the use of percentile curves in the CDC’s and WHO’s BMI charts in children with severe obesity is that outlier data have been removed and upper centiles smoothed such that a change in weight may not result in a significant change in centile at the extremes.25 26 By contrast, the IOTF’s BMI reference uses a definition that predicts an adult BMI associated with increased health risks.19 Although studies have shown some correlation with BMI and cardiometabolic risk when using the traditional CDC definition of above the 99th centile, more recent studies have proposed a new definition based on health risks that are further stratified into three risk tiers.9 27 28 This may result in improved classification; however, a general caveat is that these are defined using population data and may not apply to health risk at the individual level.

Although BMI is one of the most widely used markers of adiposity, it does not differentiate fat from fat-free muscle, and nor does it identify the distribution of adiposity, which is more closely linked to cardiometabolic risk.29 Thus, it is a screening tool but it does not necessarily indicate clinical health risks. Furthermore, in the prepubertal child, body composition changes with reduced fat mass measured by dual energy x ray absorptiometry, suggesting that BMI may yield even weaker predictions for adverse health outcomes during certain periods of children’s growth.30

Waist:height ratio has been explored as a surrogate marker of central adiposity and metabolic risk, and recent studies have shown a superior association of a waist:height ratio of 0.5 or greater with adverse lipid profiles, compared with BMI.31 32 However, a paucity of research remains to support clinical uptake of other feasible measures of adiposity including waist:height ratio, and its use in patients with severe obesity has been cautioned owing to technical challenges with the measurement.21 Assessment of BMI remains the recommended approach to identify children and adolescents with severe obesity who need further screening for comorbidities and management.21

Appetite regulation

Understanding the physiology of short term and long term appetite regulation is useful to identify the rare pathological causes of severe obesity and the rationale for various pharmacologic treatments. Appetite regulation is a precise biological process that involves integration of complex mechanisms from the central nervous system, with peripherally derived signals of the body’s nutritional and energy status. Input is received from sensory properties of food, mechanical, and chemical receptors in the gastrointestinal tract, circulating metabolites, and the autonomic nervous system, as well as hormones from the gut.33 At the start of a meal, hormonal signals are released from the small and large bowel (PYY, GLP-1, GLP-2, CCK, and OXM), stomach (ghrelin), and pancreas (insulin, glucagon).34 Insulin and leptin play dual roles as long term feedback signals reflecting the overall body energy status (fig 2).34


Fig 2 Physiology of appetite regulation. (1) Regulation of appetite is mediated by the hypothalamus and brainstem, which act to integrate peripheral hormonal and autonomic inputs. At the start of a meal, hormonal signals are released from the small and large bowel (PYY, GLP-1, GLP-2, CCK, and OXM), stomach (ghrelin), pancreas (insulin, liver), and adipose tissue (leptin). All are anorexigenic, with the exception of ghrelin which is orexigenic. (2) Hormonal signals (directly and indirectly via the autonomic nervous system) relay to anorexigenic (POMC, CART) and orexigenic neurons (AgRP, NPY) in the ARC and to neurons in the brainstem (DVN and NTS). (3) Neurotransmitters from the hypothalamus then relay a signal to the second order neurons in the PVN within the hypothalamus. (4) The brainstem directs sympathetic and parasympathetic output to the periphery, affecting metabolism and energy expenditure. (5) Higher level inputs from other cortical areas involved in executive function are integrated with information from reward centers driving hedonic eating behaviors, motivation, and sensory input. AgRP=agouti-related peptide; AP=area postrema; ARC=arcuate nucleus; CART=cocaine and amphetamine regulated transcript; CCK=cholecystokinin; DVN=dorsal motor nucleus of the vagus; GABA=gamma-aminobutyric acid; GLP-1=glucagon-like peptide 1; GLP-2=glucagon-like peptide 2; LH=lateral hypothalamus; NPY=neuropeptide Y; NTS=nucleus of the tractus solitarius; OXM=oxyntomodulin; POMC=pro-opiomelanocortin; PP=pancreatic polypeptide; PVH=paraventricular hypothalamus; PYY=peptide YY

The hypothalamus and brainstem act to integrate these inputs and signal to the arcuate nucleus, where orexigenic neurons release agouti related peptide (AgRP) and neuropeptide Y (NPY) and anorexigenic neurons release promelanocortin (POMC), and cocaine and amphetamine regulated transcript (CART).35 These neurotransmitters transmit signals to the second order neurons in the paraventricular nucleus within the hypothalamus. Cortical inputs from areas involved in executive function are integrated with information from reward centers driving hedonic eating behaviors, motivation, and sensory input. The output leads to efferent signaling to the periphery to direct food intake, satiety, substrate metabolism, and energy expenditure (fig 2).35

Body weight “set point” is a term used to describe the homeostatic neuroendocrine pathways, largely centered in nuclei of the hypothalamus and hindbrain, which maintain adult body weight within a narrow range.36 Prolonged exposure to higher energy intake with resultant increased weight can lead to a recalibration of this set point in genetically susceptible individuals, and input from hedonic areas of the brain will further override the set point leading to ongoing weight gain.

In obesity, this dysregulated system is further exacerbated by systemic inflammation, which is often associated with excess intra-abdominal and visceral lipid storage as well as adverse cardiometabolic health.37 38 Evidence from animal models supports the hypothesis that pro-inflammatory signaling induced by a high fat diet may mediate the phenomenon of leptin resistance commonly seen in obese people.39 As adiposity increases, higher circulating concentrations of leptin, insulin, and free fatty acids lead to resistance in peripheral tissues and in the hypothalamus, with resultant reduced satiety and a disruption of the homeostatic mechanism controlling body weight set point.36 Neuro-inflammation may in turn trigger cytokine signaling affecting the immune system, resulting in a state of innate immune reactivity.40 Collectively, this signaling damages hypothalamic neurons, astrocytes, and glial cells, further dysregulating energy homeostasis.41

Causes of severe obesity in children

The hypothalamus plays an important role in regulating body weight, so it is unsurprising that many of the known monogenic and syndromic forms of severe obesity are associated with hypothalamic dysfunction (fig 3).42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 In general, these conditions are rare, the most common being a homozygous mutation in melanocortin 4 receptor (MC4R), which is estimated to explain up to 5% of severe, early onset obesity.42 60 Clinically, children with this mutation present with increased linear growth and rapid weight gain associated with hyperphagia beginning in early childhood.43


Fig 3 Genetic causes of severe obesity. ACTH=adrenocorticotropic hormone; GnRH=gonadotropic releasing hormone; MSH=melanocyte stimulation hormone; TRH= thyroid releasing hormone; WAGR syndrome=Wilms tumor, anirida, genitourinary anomalies, and intellectual disability syndrome

Prader-Willi syndrome is a well described condition, occurring in about one in 25 000 live births.54 It is characterized by developmental delay, low resting energy expenditure, altered body composition with low muscle mass and increased fat mass, neonatal hypotonia, and poor feeding shifting to unrelenting hyperphagia beginning at approximately age 3 years.55 With increasing clinical use of genetic techniques, other variations such as deletions or duplications have been associated with severe obesity and variable neurocognitive delays.61 For example, pathogenic deletions at the 16p11.2 locus are linked to the shared phenotype of autism, neurodevelopmental changes, and severe obesity.56

An extreme pediatric form of obesity, ROHHAD (rapid onset obesity with hypothalamic dysfunction, hypoventilation, and autonomic dysregulation) is a rare but life threatening condition characterized by rapid weight gain beginning after 18 months of age, obesity, hypopituitarism, central hypoventilation, and autonomic dysfunction, as well as behavioral changes.62 Interestingly, non-secreting paragangliomas have been found in 30-100% of patients with ROHHAD.62 63 Evidence to suggest that ROHHAD may be an autoimmune paraneoplastic inflammatory condition is supported by findings of hypothalamic lymphocyte infiltration in an autopsy of a child who died suddenly with ROHHAD, an inability to identify causative genes in this condition, and case reports of clinical improvement in children treated with anti-inflammatory and immune modulating drugs.63 64 65

Other inflammatory or traumatic conditions affecting the hypothalamus or congenital central nervous system malformations affecting the midline can also disrupt the normal hypothalamic homeostatic mechanisms, leading to significant weight gain.66 Tumors, notably craniopharyngiomas, located in the sellar region commonly lead to severe obesity in up to 50% of patients, both before and after diagnosis and treatment.67

Cushing’s syndrome, either due to pituitary or adrenal overproduction of cortisol or, more commonly, due to exogenous treatment with supraphysiologic doses of glucocorticoids for other conditions, can also lead to hyperphagia and very rapid weight gain in a central distribution accompanied by poor linear growth.68 Lastly, the use of atypical antipsychotic drugs in children leads to extremely rapid weight gain, doubling the prevalence of overweight/obesity compared with drug-naive children and significantly increasing the risk of cardiometabolic abnormalities such as type 2 diabetes and dyslipidemia.69

In summary, key clinical features to suggest that severe obesity may be related to an underlying pathological condition include young age of onset, extremely rapid weight gain, poor or advanced linear growth, dysmorphic features, neurocognitive delay, obesity very discordant to parents and siblings, and central nervous system symptoms to suggest an infiltrative or inflammatory condition.

Emerging treatments

Emerging behavioral interventions

Current guidelines on the management of obesity recommend a multidisciplinary approach to implementing family based lifestyle interventions. This includes goal setting to increase healthy nutrition and physical activity, reduction of sedentary time, self monitoring, stimulus control, and positive reinforcement.20 However, the efficacy of this approach in children with severe obesity is unclear.16 Critics argue that intensive lifestyle modification, particularly in isolation, is inadequate for children with severe obesity as it reduces BMI only modestly by 1-2.70 However, stabilization for children who may have been on an upward trajectory of weight gain may prevent further morbidity. Another commonly cited challenge is the greater rate of attrition in lifestyle programs among children with a higher BMI.71 Qualitative studies have reported on how the stigma and physical limitations imposed by severe obesity may provide additional obstacles to achieving effective exercise.72

A few studies of lifestyle intervention published over the past five years have focused on children or adolescents with severe obesity.73 74 75 76 77 78 79 80 A randomized controlled trial from the Netherlands evaluated outpatient versus inpatient treatment of six months’ duration in 90 children and adolescents with severe obesity aged 8-18 years.76 Both groups showed a reduced BMI, but the BMI Z score reduction was significantly greater at six months in the inpatient group (−0.26 (12); P=0.04). There was no longer a significant difference between the ambulatory and inpatient groups over the follow-up period of 30 months, but both groups maintained BMI Z scores below that at study entry. A randomized controlled trial that compares the duration of inpatient stay for young people with severe obesity is under way; it will include an economic analysis.81 82 An observational study that included 487 children aged 6-15 years with severe obesity showed that a lifestyle intervention based on guideline recommendations was successful in achieving a significantly greater reduction in BMI Z score compared with obese children (mean −0.24 (SD 0.38) v −0.16 (0.3); P=0.021).73 This effect was more marked in severely obese children younger than 10 years, suggesting that age may be a predictor of weight loss, a finding noted in other behavioral intervention studies.74 75

Overall, behavioral interventions for children and adolescents show modest reductions in BMI over the short term, with improved efficacy in younger children. Studies have also shown improvements in myocardial mechanics, insulin sensitivity, and markers of inflammation in adolescents with severe obesity following completion of exercise programs.83 84 85

Durability of weight loss is a recognized challenge; however, a recent study following children with severe obesity enrolled in an outpatient lifestyle program found significant improvements in BMI Z score (−0.23 (0.32); P=0.01) and other cardiovascular risk factors at two years.77 Further research is needed to test adjunctive therapies that can be applied after six months of a behavioral intervention, when physiological responses to reduction of body weight promote regain.

The use of interactive information technologies and applications (“apps”) for smart phones and tablets has not been tested in young people with severe obesity. Motivational interviewing is a behavioral technique that seeks to facilitate intrinsic motivation in a client by using an empathetic, goal oriented approach to resolve ambivalence toward behavior change. It has been studied in several chronic medical conditions including childhood obesity but has not been extensively evaluated in severe obesity.86 87 Research is also needed to identify psychosocial characteristics of individuals and families that respond best to specific behavioral interventions.

Emerging medical therapies

Given the numerous and redundant pathways favoring the promotion of weight gain, drugs that affect multiple mechanisms in parallel may be a useful addition to behavioral lifestyle treatment in this population.88 Existing drugs can generally be classified as being directed at one of the following primary targets: nutrient processing, enteric hormonal regulation, or central modulation of appetite, satiety, and food behaviors. Compared with adults, far fewer pediatric trials of drugs for reducing weight have been conducted, partly owing to the developmental changes in growth affecting nutritional requirements and concerns about adverse effects. This review will focus on drugs studied in children that remain available on the market and those being tested in the pediatric age range (0-18 years).

Nutrient processing

Orlistat remains the only drug approved for treatment of obesity in the pediatric age group. It acts to inhibit gastric and pancreatic lipase and diminishes absorption of triglycerides by approximately 30%.89 Several clinical trials have evaluated its use in pediatric weight management, most of which have shown a positive, albeit modest, effect.90 91 92 93 94 The largest controlled trial randomized 539 adolescents with obesity to placebo or 120 mg of orlistat three times a day. At 54 weeks, BMI was significantly reduced by 0.55 in the treated group compared with an increase in BMI of 0.31 in the control group (P=0.001).93

The effects of orlistat on secondary cardiometabolic outcomes and measures of insulin resistance are not consistent across studies.93 95 In general, orlistat seems to be a safe drug and has minimal systemic absorption.96 Unfortunately, compliance is affected by gastrointestinal side effects including loose and oily stools, flatulence, and fecal urgency, which occur in 9-50% of patients.93 Theoretical concerns have been raised about impaired absorption of lipid soluble vitamins and increased risk of cholelithiasis in patients taking orlistat.97

Enteric hormonal regulation

Gastrointestinal tract hormones play a critical role in peripheral energy metabolism in addition to communicating satiety and hunger cues to the hypothalamus, rendering these a logical drug target. Initially marketed for pediatric patients with type 2 diabetes, metformin has been used off label in the clinical treatment of adolescent obesity.98 Metformin’s mechanism of action to reduce weight is not clear but is likely multifactorial. In diabetes, metformin lowers glucose predominantly by reducing hepatic gluconeogenesis and to a lesser extent by increasing insulin sensitivity.99 Its effect on weight may be mediated by increasing GLP-1, lowering dipeptidyl peptidase-4 activity, altering the bile acid pool and the gut microbiome, enhancing lipid oxidation, and promoting central satiety.100 101 102

A systematic review published in 2014 summarized the results of 14 randomized trials in obese pediatric patients without diabetes; it found that treatment with metformin in conjunction with lifestyle therapy was associated with a reduction in BMI of 1.16 (95% confidence interval 0.73 to 1.6; P=0.003), compared with the untreated groups. Subgroup analyses found a significantly greater reduction in patients with baseline BMI 35 or higher (−1.23, −1.66 to −0.79) than in those with BMI below 35 (−1.05, −1.81 to −0.29).98 However, two trials in the review found no significant difference after 12 months of treatment.98 Metformin was also associated with gastrointestinal side effects in 26% of patients. This may contribute to lowering of appetite, but these side effects may be limited with gradual dose titration. No pediatric studies have reported derangements in liver or kidney function or lactic acidosis.103 Overall, evidence suggests that metformin is safe in achieving weight loss in children with obesity; however, the magnitude of its effect is small and durability of weight loss is questionable.

GLP-1 agonists are also used in type 2 diabetes in adults and have been found to have a favorable side effect of weight loss. GLP-1 acts locally in the gastrointestinal tract to increase glucose dependent insulin, reduce glucagon secretion from the pancreatic β cell, slow gastric emptying, and increase free plasma leptin concentrations by lowering its soluble receptor concentration.104 Centrally, GLP-1 acts in the brainstem nucleus of the solitary tract and in the hypothalamus via anorexigenic POMC neurons to reduce appetite and intake.105 Exenatide, a GLP-1 agonist, given as a subcutaneous injection twice daily, has been studied in two small trials of adolescents with severe obesity and has shown beneficial effect with a BMI reduction of 1.13 (0.24 to 2.03) and 1.7 (0.4 to 3.0) (table 1).106 107 In pediatric studies, these drugs are safe and generally well tolerated, with nausea reported as the most common side effects, occurring in 36-62% of patients.106 107 Liraglutide, a GLP-1 analog given by subcutaneous injections once daily, has also been evaluated in adolescents, but the randomized controlled trial remains unpublished.108 This new evidence base may predict its increasing use in clinical practice in the future.

Table 1

Summary of clinical trials for obesity drugs in children and adolescents

View this table:

Central nervous systems targets

Several drugs targeting central mechanisms of weight gain have been evaluated and approved in adults for the management of obesity, but fewer have been studied in children and adolescents. Lorcaserin is a serotonin receptor agonist that targets 2C receptors on POMC neurons, inducing satiety.109 Highly selective activation of this anorexigenic pathway leads to reduced energy intake and weight loss, without the significant cardiovascular and psychiatric adverse effects seen with previous serotonin agonists.110 Two clinical trials using lorcaserin in children and adolescents have just been completed, with study results pending.111 112 Currently, the use of these drugs is restricted to clinical research studies.

Specific anti-epileptic drugs used to treat patients with seizure disorders have incidentally been found to cause weight loss. Topiramate has been historically used as an anti-epileptic drug affecting inhibitory GABAergic pathways, but it has more recently been evaluated in the management of binge eating disorders.113 It has been parenthetically noted to cause weight loss in treated adolescents.114 Its mechanism of action is unclear but may be related to inhibition of carbonic anhydrase, an enzyme involved in de novo adipogenesis and/or a reduction in concentrations of neuropeptide Y.115 116 Alternatively, the side effect of dysgeusia may lead to a decreased appetite.103 The use of this anti-epileptic drug remains largely within a research setting, but clinical trials are under way to study the use of this drug in adolescents with severe obesity.117

Zonisamide is a novel anti-epileptic drug that may also promote weight loss. It has been extensively studied in children for treatment of epilepsy; however, there are case reports and case series retrospectively documenting a weight loss effect in pediatric patients. A review of 103 patients treated with zonisamide (average maximum dose of 375 mg/day) showed that 35% of patients experienced weight loss of more than 5%, an effect that was more common in overweight patients.118 Similar to topiramate, the most common adverse effects are largely central nervous system related and include somnolence, dizziness, headache, nausea, and poor appetite.119 These anti-epileptic drugs may represent a preferred therapy for adolescents with a seizure disorder and comorbid obesity, but further studies are needed to evaluate them more broadly as an obesity treatment.

Emerging surgical therapy

Over the past 20 years, bariatric surgery has become an important treatment option for adults with obesity and is increasingly being offered to adolescents.120 National trends in the United States show a fivefold increase in the number of adolescent procedures from 1997 and 2003.121 Rates of bariatric surgery for adolescents vary by center, country, and healthcare infrastructure.122 123 The laparoscopic Roux-en-Y gastric bypass (RYGB) is the most frequently used procedure in adults and has been extensively studied in adolescents. Laparoscopic adjustable gastric banding (LAGB) and sleeve gastrectomy have also been evaluated in this age group. Laparoscopic gastric plication is another emerging surgical technique not yet well studied in pediatric patients.124 Randomized trials comparing the different surgical techniques are scarce, with no clear consensus on the optimal surgical technique in adolescents.125

Bariatric surgery promotes weight loss through multiple, diverse mechanisms, which accounts for its effectiveness compared with medical treatments.126 All procedures lead to mechanical restriction of food intake by reducing or bypassing the stomach. The RYGB also leads to interference with absorption of nutrients, and the gastric band may increase vagal afferent activity to promote satiety.126

Understanding of the changes that occur after bariatric surgery in enteric hormonal and neural signaling pathways comprising the “gut-brain neuroendocrine axis” is growing. Weight loss triggers expected changes in leptin, insulin, and adiponectin, but bariatric surgery directly affects other regulators of energy metabolism such as glucagon, ghrelin, GLP-1, and PPY.127 128 Improvements in metabolic health occur immediately after the RYGB and sleeve, indicating that hormonal changes occur acutely and are not simply a reflection of weight loss.129 Other proposed drivers of surgical weight loss include shifts in the microbial composition of the gut, differential metabolism of bile acids, and changes to branched chain amino acids.130 131

The RYGB has been associated with the most serious potential surgical risks, including complications such as anastomotic leakage, infection, thromboembolism, and bleeding; however, these are more rarely reported (5-6%) in adolescent patients.132 Nutritional deficiencies can also occur, but their risk is mitigated by the use of appropriate vitamin supplementation. In general, the LAGB carries less significant risk and lower rates of complications, although surgical revision may be needed in the longer term.133

Expert consensus guidelines vary between countries but suggest that bariatric surgery should be restricted to adolescents with BMI of 35 or above with major comorbidities (type 2 diabetes mellitus, moderate to severe sleep apnea, pseudotumor cerebri, or severe non-alcoholic fatty liver disease) or with BMI of 40 or above and other comorbidities (hypertension, glucose intolerance, insulin resistance, or dyslipidemia).134 135 Patients should be carefully evaluated and followed by a multidisciplinary healthcare team (physician/surgeon, nurse, mental health specialist, dietitians, and exercise counselors) expert in pediatric medical and psychosocial suitability for surgery. Surgery should be performed in centers with high volumes of surgery or in conjunction with bariatric surgeons who perform these procedures routinely in adults.134 136 Existing guidelines provide a framework, but they are largely based on criteria in adults and have not been comprehensively evaluated in the pediatric age range.

A meta-analysis of 23 studies consisting primarily of uncontrolled retrospective case series reported the outcomes of pediatric bariatric surgery. It found a decrease in BMI of 13.5 (95% confidence interval 11.9 to 14.1) at one year, approximately one third of initial body weight, with the largest reduction occurring in patients who underwent RYGB.133 Importantly, this weight loss occurred concurrently with improvements in metabolic health and other obesity related comorbidities including hypertension, sleep apnea, dyslipidemia, and type 2 diabetes.133

A second systematic review of six studies evaluating short term psychological outcomes after bariatric surgery reported improvements in depression and quality of life.137 A third systematic review included one randomized control trial in 50 adolescents comparing LAGB with an intensive lifestyle control intervention. Bariatric surgery resulted in an average BMI reduction of 12.7 (11.3 to 14.2) compared with 1.3 (0.4 to 2.9) in the intensive lifestyle group.138

Longer term outcome data are emerging from prospective, observational cohorts of adolescents.139 140 141 142 143 Prospective results from a large US, multicenter group (Teen-LABS), performing RYGB and gastric sleeve procedures showed major perioperative complications in 8% of patients and a 5% rate of major inpatient morbidity.144 Follow-up of more than 300 adolescents in this cohort shows a weight reduction of 27%, remission of many cardiometabolic comorbidities, and improvements in quality of life measures at three years after surgery. In this group, 95% of study participants with type 2 diabetes had remission at three years. This rate is much higher than other estimates previously reported in adult cohorts, highlighting the potential importance of early surgical intervention.145 146 The most significant long term risks in this cohort included micronutrient deficiencies in 57% and risk of reoperation (for example, internal hernia repair, lysis of adhesions, and cholecystectomy) in 13%.145

Bariatric surgery remains the most effective treatment for severe obesity and has the potential to significantly alter the health trajectory of adolescents. However, it raises ethical and medical concerns that are unique to the pediatric population. Some of these concerns include the short and long term effects on growth, nutrition, mental health, and quality of life, as well as the ethics of consent in children with evolving capacity for decision making.147 There have been reports of children undergoing these procedures in early childhood,148 which highlights the urgency for ongoing evaluation.

Other important considerations include the effect of bariatric surgery on premorbid patterns of disordered eating such as binge eating, increased postoperative fertility in young female patients, and ensuring equitable access to care.149 Further research is needed to assess the experience of adolescents after surgery and to compare the relative efficacy of different surgical techniques, criteria for selecting patients, durability of the procedures over time, and long term complication rates.150


The National Health and Research Council of Australia published Clinical practice guidelines for the management of overweight and obesity in adults, adolescents and children in Australia in 2013, which includes a section for post-pubertal adolescents with severe obesity.151 This guideline provides a clear, evidenced based approach but is not specifically focused on severe pediatric obesity. The American Heart Association has published a scientific statement entitled Severe obesity in children and adolescents: identification, associated health risks, and treatment approaches,21 which outlines the most recent evidence for different medical and surgical treatment of young people with severe obesity. This statement does not provide explicit therapeutic recommendations but critically reviews the evidence behind different treatment modalities. Recent best practice guidelines from the American Society for Metabolic and Bariatric Surgery Pediatric Committee and the Interdisciplinary European Guidelines on Metabolic and Bariatric Surgery both provide recommendations around the surgical management of young people with severe obesity.134 136 The former is focused exclusively on the care of adolescents and concerns unique to pediatric bariatric surgery.134


Despite substantial advances in our understanding of severe obesity in children and adolescents, sizeable gaps in our knowledge remain. Current evidence suggests that existing behavioral interventions for severe obesity are modestly effective in reducing weight, with incremental short term improvements seen with adjunctive drugs for weight loss. Bariatric surgery is effective in reducing weight, but more research is needed in the pediatric age range to determine who will benefit most from these procedures. Development of safe, effective, multimodal, and individualized behavioral and drug treatments that target the fundamental and diversified mechanisms underlying weight gain represents a critical challenge facing the field. Studies have largely focused on BMI outcomes, a surrogate measure that may not reflect the state of physical and psychological health of the individual. Recognition of these factors has led to a call for more patient centered outcomes and quality of life data to better understand the personal effect of obesity treatment.152 153

Research questions

  • What is the spectrum of the contribution of hedonistic and homeostatic mechanisms regulating body weight in children and adolescents with severe obesity?

  • What is the optimal combination of strategies to individualize obesity treatment for the pediatric patient by age and developmental stage?

  • What strategies can be used to enhance durability of short term improvements in body mass index and cardiometabolic health measures after behavioral lifestyle interventions?

  • What characteristics of patients and bariatric surgery procedures result in the greatest long term efficacy and safety in adolescents undergoing surgery?

Patient involvement

Attempts have been made to better understand the adolescent patient’s experience with obesity and its medical treatment.154 155 156 In determining areas of importance for this review, we sought the patient’s perspective by conducting group discussions with families attending a weight management program. The adolescents were very interested in technology and have made the recommendation that, “researchers need to move more quickly developing and doing research in apps for healthy living that are designed for us.” An adolescent who had undergone bariatric surgery wanted to share his experience: “I think bariatric surgery should be a treatment option available for teens to help prevent poor health in adulthood. However, teens should understand surgery is a ‘tool’ not a ‘cure.’ Over two years, I have lost 120 lb [54 kg] and my life is much better. I have some challenges with eating, but the positives far outweigh the negatives. I can keep up with friends now—things have opened up for me and my future looks good.”


  • We acknowledge the work of Matan Berson in helping to conceptualize and design the figures for this manuscript.

  • Contributors: All three authors were involved in drafting and revising the manuscript. All approved the final version.

  • Competing interests: We have read and understood the BMJ policy on declarations of interest and declare the following interests: JH is supported by the Mead Johnson chair in nutritional science.

  • Provenance and peer review: Commissioned; externally peer reviewed.

  • Patient consent: Consent was obtained to include the patient perspective.


View Abstract