|Home | About | Journals | Submit | Contact Us | Français|
The objective of this study is to update evidence-based best practice guidelines for pediatric/adolescent weight loss surgery (WLS). We performed a systematic search of English-language literature on WLS and pediatric, adolescent, gastric bypass, laparoscopic gastric banding, and extreme obesity published between April 2004 and May 2007 in PubMed, MEDLINE, and the Cochrane Library. Keywords were used to narrow the search for a selective review of abstracts, retrieval of full articles, and grading of evidence according to systems used in established evidence-based models. In light of evidence on the natural history of obesity and on outcomes of WLS in adolescents, guidelines for surgical treatment of obesity in this age group need to be updated. We recommend modification of selection criteria to include adolescents with BMI ≥ 35 and specific obesity-related comorbidities for which there is clear evidence of important short-term morbidity (i.e., type 2 diabetes, severe steatohepatitis, pseudotumor cerebri, and moderate-to-severe obstructive sleep apnea). In addition, WLS should be considered for adolescents with extreme obesity (BMI ≥ 40) and other comorbidities associated with long-term risks. We identified >1,085 papers; 186 of the most relevant were reviewed in detail. Regular updates of evidence-based recommendations for best practices in pediatric/adolescent WLS are required to address advances in technology and the growing evidence base in pediatric WLS. Key considerations in patient safety include carefully designed criteria for patient selection, multidisciplinary evaluation, choice of appropriate procedure, thorough screening and management of comorbidities, optimization of long-term compliance, and age-appropriate fully informed consent.
Evidence-based best practice guidelines for pediatric/adolescent weight loss surgery (WLS) have been previously described (1). Earlier guidelines focused on patient safety, criteria for eligibility, informed consent, psychological maturity, and surgeon and program credentialing (2). This report covers key updates relating to pediatric/adolescent WLS.
Rapidly increasing prevalence of obesity among children and adolescents is associated with substantial medical and psychosocial morbidity (3,4). It is important that health professionals assess obesity and initiate action plans (5). Some WLS procedures may be indicated for carefully selected, extremely obese adolescents (5). Children with BMI >99th percentile become obese adults (BMI ≥ 30) (3) with more health complications and a higher mortality rate (6–9) than those who become obese in adulthood.
WLS in the mature adolescent may reduce the risk of morbidity and early mortality from obesity-related disease (10–12). To date, the current evidence base is not sufficient to determine which WLS procedures are optimal for adolescents. However, early evidence of safety and efficacy exists for two procedures (13–16). This report, which updates best practice guidelines in pediatric/adolescent WLS, focuses on prevention of early mortality and comorbidities from obesity, patient selection criteria, and long-term outcomes of WLS.
We searched PubMed, MEDLINE, and the Cochrane Library for articles published between April 2004 and May 2007 on WLS and pediatrics, adolescents, gastric bypass, laparoscopic adjustable band, and extreme obesity. The system used to grade the quality of the evidence has already been described (2). More than 1,085 papers were identified; 186 of the most relevant were reviewed in detail. These included randomized controlled trials, prospective and retrospective cohort studies, meta-analyses, case reports, prior systematic reviews, and expert opinion. The focus of the recommendations and the process used to develop them are described in our prior report (2).
Data indicate that patient safety and weight loss outcomes for extremely obese adolescents who undergo WLS are comparable to, or better than, those seen in adults (13,15–18). Ten case series of WLS in adolescents have been published since 2004 (see Table 1) (17–26).
In the United States, Roux-en-Y gastric bypass (RYGB) for weight loss dates back to the 1960s for adults and the 1980s for adolescents (27,28). Only three new studies on RYGB in adolescents have been published since our last review in 2004 (17,21,24). One of these (17) is a controlled multicenter study that compared laparoscopic RYGB (LRYGB) with a 1-year, family- based pediatric behavioral treatment program. In the LRYGB group, BMI decreased from 56.5 to 35.8, with significant resolution of comorbidities; there was no significant BMI change in the comparison group.
Perioperative morbidity was generally consistent with that seen in numerous adult studies and meta-analyses. In the three case series, no deaths occurred in the perioperative period. In Inge et al. (17), an 18-year-old adolescent with a BMI of 80 and multiple comorbidities died 9 months after RYGB from complications of Clostridium difficile colitis (17). Thus, based on current literature, RYGB (not minigastric bypass or loop gastric bypass) is recommended as a safe operation in adolescents, with outcomes similar to those observed in adults. However, every effort should be made to avoid vitamin deficiency (29) and to maximize postoperative compliance (30) because adolescence is a time of increased growth and development, and decreased compliance (28).
A number of new reports for adjustable gastric band (AGB) have been published in the past 4 years. Because of its relative safety, AGB offers an effective and attractive treatment option in carefully selected patients. It also has a lower risk of postoperative vitamin deficiencies compared with RYGB or biliopancreatic diversion (BPD). Between 2005 and 2007, five case series were reported in adolescents (18,20,23,25,26). The larger studies included 221 patients between the ages of 9 and 19. Patients had a mean preoperative BMI of 43–48, and they lost 37–63% of excess body weight during follow-up periods that ranged from 6 months to 7 years. However, caution must be used when interpreting reported excess weight loss. Few adolescent WLS studies indicate how data are calculated, and estimation of ideal weight for children differs considerably from adults.
Complication rates were 6–10%, with no deaths. Reoperation rates, including band removal, were 8–10% (20,23,25). Long-term weight loss outcomes and precise descriptions of changes in comorbidities following WLS are still lacking. In one study, at least 80% of adolescents had sustained weight loss 5 years after AGB, but their numbers were small, and the fraction lost to follow-up was not provided (20). Nonetheless, we can recommend AGB placement as safe, and more effective than behavioral interventions for a selected population of adolescents. However, we suggest limiting widespread use until more robust long-term safety and efficacy outcomes and the Food and Drug Administration trial results are available. Weight loss devices should only be used in pediatric populations in the setting of a controlled clinical trial after investigational device exemption and institutional review board approval.
Papadia et al. (19) describe a case series on BPD in adolescents. Outcomes suggest that risks outweigh potential benefits of greater weight loss with BPD, duodenal switch, and other procedures that cause significant malabsorption compared with RYGB or AGB. This is particularly true in light of well-described compliance issues that may increase risks for late protein malnutrition and nutritional complications surrounding pregnancy.
Laparoscopic sleeve gastrectomy is a new operation that produces significant initial weight loss with low operative risk in adults. Short-term data suggest that laparoscopic sleeve gastrectomy may be a safe alternative, with fewer nutritional risks than RYGB (20,31,32). Until techniques are standardized and proof of longer-term efficacy becomes available, we recommend that this operation be considered investigational and only be offered to adolescents within the context of a controlled prospective study.
A steep rise in prevalence of type 2 diabetes is occurring worldwide in parallel with an increasing rate of obesity in children and adolescents (33). Children and adolescents with type 2 diabetes are at increased risk for obesity-related comorbidities, including hypertension (HTN), dyslipidemia, and nonalcoholic fatty liver disease. Moderately good evidence suggests that young patients with type 2 diabetes have rapidly progressive disease. Nephropathy, in particular, progresses rapidly; 30–40% of patients develop microalbuminuria within 5 years of diagnosis (34,35). Progressive retinopathy and atherosclerotic disease have also been documented within 5 years of diagnosis of type 2 diabetes in young adults (36). In this age group, glycemic control with medical treatment is often poor (37). Early data suggest that diabetes may completely reverse in adolescents who undergo LRYGB (T.H. Inge, unpublished data). Thus, established type 2 diabetes is a strong indication for WLS in adolescents (38).
Obstructive sleep apnea (OSA) and obesity hypoventilation syndrome are common among children with extreme obesity, cause substantial morbidity, and respond to WLS. Between 8 and 20% of children and adolescents with obesity have moderate-to-severe OSA, and ~15% have central sleep apnea, often associated with episodes of severe oxygen desaturation during sleep (<85%) (ref. 39). Adolescents presenting for WLS tend to have extreme obesity, and among this group, the prevalence of OSA is ~55% (ref. 39). Consequences of OSA include learning difficulties, hyperactivity, and cardiovascular abnormalities. In a small case series using pre- and postoperative polysomnography, OSA significantly improved or resolved in most adolescents after WLS (40). These findings are consistent with outcomes in adults after WLS. Thus, moderate or severe OSA (e.g., apnea–hypopnea index (AHI) >15) is a strong indication for early WLS in adolescents.
Studies using histological diagnoses show that 38% of obese children and adolescents have steatosis compared with 5% of lean subjects; ~9% have NASH compared with 1% of the lean population (41). Although steatosis and NASH may progress to cirrhosis, the risk of disease progression is not well understood. There is good evidence that WLS can decrease the overall amount of steatosis (42) and many of the inflammatory markers associated with fatty liver disease (43). Dixon et al. (44) demonstrated regression in fibrosis with WLS at 2 years. Several drug trials are underway to treat pediatric nonalcoholic fatty liver disease, but weight loss is currently the only treatment option available for NASH in adolescents. Therefore, severe and progressive NASH (as opposed to steatosis alone or mild NASH) should be considered a strong indication for early WLS in adolescent patients.
WLS is considered the long-term procedure of choice among adults with pseudotumor cerebri (45,46). As with WLS in adults, symptoms of pseudotumor cerebri in adolescents improve several months after WLS (47,48). Thus, pseudotumor cerebri is a strong indication for early WLS in adolescents.
Childhood obesity (BMI at ages 4–17) is associated with left ventricular hypertrophy in young adults (ages 20–38) (ref. 49). Skinfold thickness and blood pressure measured in childhood and adolescence predict decreased carotid artery elasticity in adulthood (50). These factors likely predict long-term risk for cardiovascular disease (CVD), but evidence of short-term morbidity from these risk factors is lacking. WLS clearly improves these risk factors. Shargorodsky et al. (51) found that patients between 16 and 55 years old with AGB showed improvement in their metabolic milieu 4 months postoperatively, and high-risk patients (≥2 CVD risk factors) showed improvement in small artery elasticity (51). WLS improves HTN and significantly improves CVD risk factors in adults (22). In adolescents, CVD risk factors are not as strong indications for early WLS.
Indicators such as high waist circumference and triglycerides in childhood (9–10 years) predict the metabolic syndrome in young adulthood (18–19 years) (52). But unlike adults, metabolic syndrome in adolescents is ill-defined and unstable during this period of major physiologic changes, and the diagnosis may have less clinical utility than it does in adults (53). Buchwald et al. (22) found that WLS may result in improvement of metabolic and inflammatory parameters, including hyperinsulinemia, insulin resistance, and lipid metabolism. These conditions (i.e., hyperinsulinemia, insulin resistance, dyslipidemia, and HTN) are common among adolescents with obesity and are associated with long-term cardiovascular risk. Nonetheless, these indications are not strong enough to recommend early WLS in adolescents.
Research clearly shows that obesity has a negative impact on quality of life (QOL) for adolescents (54–58), and the degree of obesity is directly related to the perceived impairments in emotional, social, physical, and school functioning experienced by adolescents (55,57). Strauss and Pollack (56) demonstrated that teens with obesity are more likely to be socially marginalized than their normal-weight peers. Ball et al. (54)) found that being overweight as a young adult had a lasting impact on life satisfaction and aspirations. Research demonstrated improved psychosocial status in adults following WLS (59).
Several recent studies also suggest significant improvement in postoperative QOL after AGB and RYGB in adolescents (13,18,20). Behavioral interventions that focus on obese teens and their families generally have low success in achieving long-term weight loss (60). Based on this nascent data, WLS may bring important benefits to emotional health and QOL in extremely overweight adolescents.
There is a significant incidence of depression among overweight and obese adolescents. Studies consistently demonstrate that many obese adolescents seeking weight management treatment present with depression (58,60–62). For example, Zeller et al. (62) found that 53% of adolescents were mildly depressed, 30% self-reported clinically significant depressive symptoms, and 45% were clinically depressed based on their mothers’ reports. In addition, only 21% of those seeking surgery were currently engaged in any form of psychological treatment (58). Available data indicate that preoperative depression does not adversely affect short-term (1–2 years) weight loss after WLS (63). Therefore, presence of depression is not an exclusion criterion for WLS.
Binge eating and self-induced purging occur in 5–30% of obese adolescents seeking WLS. Such preoperative eating disturbances do not appear to affect weight loss outcome after WLS, at least in the short term. Therefore, presence of eating disturbances is not an exclusion criterion, but treatment must be initiated and the patient considered stable prior to surgery.
The presence of eating disturbances is not an exclusion criterion for WLS, but adolescents with such disorders should be treated prior to surgery (category B).
Compared with lower levels of obesity, Freedman et al. (3) show increasing metabolic risks associated with higher BMI for age, especially ≥99th BMI percentile. They recommend more aggressive weight control strategies for this group. Using the Learning Management System database from the Centers for Disease Control (64), we calculated the BMI percentile values that correspond to BMI cut points recommended for use in adults (i.e., 35 and 40). A BMI of 35 between the ages 18 and 20 corresponds to a BMI percentile of 99.1–98.4 in men and 97.7–96.8 in women. In contrast to adults, a BMI of 35 at age 16 corresponds to a BMI percentile of 99.2 in boys and 98.4 in girls, although at age 12 it corresponds to a BMI percentile of 99.4 in boys and 99.3 in girls.
Because the average BMI increases with increasing age, a more conservative approach to younger patients is achieved by using a fixed BMI cut point. All adolescent boys, and most girls who are under age 18 and have a BMI of 35, are above the 99th BMI percentile (3). Therefore, BMI thresholds used for selecting adults for WLS also identify adolescents at substantially increased risk for short- and long-term medical comorbidities.
The benefits of WLS outweigh the risks in adolescents with extreme obesity and associated comorbidities. However, selection for surgery during adolescence should be closely linked to obesity-related comorbidities.
If short-term health consequences are likely to have a negative effect on long-term health, and if significant benefit is expected from WLS, we recommend WLS at a BMI cut point of ≥35. This is the case for patients with established type 2 diabetes mellitus, pseudotumor cerebri, moderate-to-severe sleep apnea (AHI > 15), and severe steatohepatitis. If adolescents have less severe comorbidities or risk factors for long-term disease, and if there is no proven disadvantage of waiting until adulthood, we recommend a BMI cut point ≥40 for WLS. Those comorbidities include, among others, HTN, milder forms of OSA, impaired QOL, insulin resistance, glucose intolerance, or dyslipidemia.
Considerations other than BMI and comorbidities must remain an important part of medical decision making for adolescents. These include, but are not limited to, physical and psychological maturity, treatment and stability of psychological comorbidities, adequacy of prior weight loss attempts, firm evidence of ability to comply with follow-up medical care, and the desire of the patient to have surgery. Table 2 provides a summary of updated recommendations on selection criteria for WLS in adolescents.
Patients with uncontrolled psychosis (presence of hallucinations and delusions), bipolar disorder (extreme mood lability), or substance use disorders can be considered for WLS on a case-by-case basis after they have been in remission for 1 year (category C).
There is no empiric evidence supporting the establishment and use of a multidisciplinary team for adults or adolescents undergoing WLS, but this approach is rational and well-established as the standard of care (65–67). Experts agree that having a multidisciplinary team improves preoperative selection and education as well as postoperative outcomes. This is especially true in pediatric and adolescent programs. The ideal team would include a minimum of four or five professionals who are colocated and have at least one face-to-face meeting preoperatively to prepare a treatment plan for each patient. Primary team members should include a surgeon; pediatric specialist; registered dietitian; mental health specialist; and coordinator. Specialists in pediatric physical therapy, pulmonology, gynecology, endocrinology, infectious diseases, cardiology, sleep disorders, gastroenterology, radiology, psychiatry, and hematology should be available for consultation as needed.
Establishing a WLS program in a free-standing children’s hospital entails significant expense (for extensive training and equipment that can accommodate extremely obese patients) (68). Low operative volume as well as surgeon and institutional inexperience with WLS may also pose significant risks (69,70). The volume of appropriate adolescent candidates for WLS is unlikely to be high enough in any one center to allow a pediatric surgeon to gain the experience required to minimize complications of WLS. Therefore, the pediatric/adolescent WLS program would ideally be colocated with an adult WLS program to allow for sharing of equipment (e.g., large computed tomography scanners, Hoyer lifts), and for an experienced weight loss surgeon to work with a pediatric surgeon in these complex, high-risk cases. Partnership with adult programs will also enable a seamless transition to support group and lifelong postoperative monitoring.
Patients with higher BMI and more significant medical illness are at increased risk during WLS. Access to WLS earlier in life may reduce the risk. Because younger patients will generally have fewer advanced comorbidities, early WLS may also decrease risk of perioperative mortality. One longitudinal study compared mortality among groups of extremely obese patients <40 years old. One group received WLS, the other did not. Among those treated with WLS, 3% died in the 13-year follow- up period compared with 13.8% of those who did not have surgery (71). This study suggests that surgery in early adulthood may reduce the risk of death from obesity. However, it does not directly show that WLS during adolescence confers additional benefit compared with WLS during early adulthood.
Psychosocial outcomes after WLS have not been adequately studied, particularly in adolescents. Data suggest short-term improvements in depression, eating disturbances, and QOL after WLS (72). It is unknown whether these improvements are sustained long-term. Some long-term data in adults indicate that mood and eating disturbances may recur after initial improvement. It is therefore important that all adolescents undergoing WLS should receive careful follow-up, and that appropriate treatment be instituted should mood, eating, or substance use disorders occur after WLS (73).
The majority of patients undergoing WLS will develop some nutritional deficiency; therefore, strict preoperative (for those deficient in one or more nutrients) and postoperative adherence to multivitamin and mineral supplementation is critical for preventing severe complications (29,74,75). Non-compliance with medical regimens is particularly common among adolescents with chronic illnesses (76). Therefore, adolescents undergoing WLS should be carefully assessed for ability to comply with medical regimens and follow-up care. Consistent attendance and compliance with medical interventions is an important measure of whether a patient and family are likely to comply with postoperative care.
Low levels of iron, vitamin B12, vitamin D, and calcium are common problems after RYGB (77). Adolescents may also be at particular risk for thiamine deficiency (29). Adolescence is a critical time for bone mass accumulation, with up to 50% of adult total bone mass achieved during this period. Calcium and vitamin D are vital for optimal bone mineral accrual in the developing skeleton (78,79).
There are no randomized controlled trials on pregnancy before vs. after WLS, but some data show that pregnancy after RYGB and AGB is safe (80). In 1998, Wittgrove et al. (81) found less risk of gestational diabetes, macrosomia, and cesarean section post-RYGB than with pregnancy while obese (81). Dao et al. (82) reported no significant episodes of malnutrition, adverse fetal outcomes, or pregnancy complications within the first year after WLS. There are no studies on outcomes of pregnancy after WLS in the adolescent population, However, T.H. Inge (unpublished data) reported a twofold increase in teen pregnancy in his female LRYGB patients.
This unexpected finding suggests that there may be an increased risk of pregnancy in adolescents undergoing WLS. For this reason, we recommend that all female adolescents be informed preoperatively about increased fertility following weight loss, and the possible risks associated with pregnancy during the first 18 months after WLS. These patients should be counseled to avoid pregnancy during this period, and offered contraception.
The process of informed consent in the adolescent who is referred for WLS is associated with certain medical, legal, and ethical issues. As part of a carefully considered risk–benefit decision, it is important for the care team, patient, and family to recognize and consider the specific risks of WLS, and particularly those relevant to the younger patient. The key facts to recognize and consider are a majority of adolescent obesity tracks into adulthood; risk factors for adult obesity are increasing age, higher BMI, and parental obesity (3); WLS is far more effective than behavior modification, and family-based therapy is generally more effective than unsupervised diet and exercise (17); some dieting behaviors and obesity both carry a risk of morbidity and mortality, and these long-term risks must be weighed against operative mortality and morbidity associated with WLS. Knowledge and understanding of these issues by patient and family alike should be formally assessed as part of the informed consent process.
Problems arise when the adolescent and the parents disagree about WLS. Parents and adolescents differ in their perceptions of the impact of obesity on their lives (55,57,60). Parents tend to more strongly endorse the negative medical and psychosocial impact that obesity is having on their children. One must be extremely careful to recognize when overt or subtle coercion is responsible for a child’s assent to surgery. Without an empirically valid method of assessing an adolescent’s capacity to make an informed decision about WLS, the clinical team must consider the adolescent’s cognitive, social, and emotional development, and support his or her independent role in the decision-making process (83).
Approximately 4% of US children suffer from extreme obesity (99th percentile of BMI for age) (3). There are currently no firmly established criteria for selecting adolescent patients who will benefit most from WLS. Patients who should undergo surgery are those who have the highest risk of continuing to suffer from obesity as adults, and those who have already developed comorbidities. The major risk factors for childhood obesity tracking into adult obesity include parental obesity, increasing age, and increasing BMI (3).
Our task group carefully considered several possible BMI-related selection criteria, and the available evidence for short-and long-term medical risks at a variety of BMI thresholds. We concluded that adult BMI criteria for WLS (≥35 with significant comorbidities or ≥40 with less serious comorbidities) are also appropriate for selecting adolescents who are most likely to benefit from WLS, provided that these thresholds are closely linked to established medical comorbidities and that all other selection criteria are rigorously met. This recommendation differs from that in our previous report (i.e., BMI cut points of 40 and 50) (ref. 14).
The evidence base on which these recommendations are made has limitations that need to be addressed by future research. Currently, little research effort focuses specifically on interventions that could treat or reverse extreme obesity for young people. However, federally sponsored multicenter studies have recently started (http://www.cincinnatichildrens.org/teen-labs). Recommendations are, therefore, largely based on cohort studies, nonrandomized clinical trials, case series or reports (i.e., categories B and C), and expert opinion (category D). All programs performing WLS in adolescents should participate in rigorous long-term data collection to improve the evidence base in this field.
Dr Lenders and the Nutrition and Fitness for Life (CL, AM, MM) at the Boston Medical Center thank the American Society for Nutrition (Physician Nutrition Specialist Award), the Carl and Ruth Shapiro Family Foundation, and the New Balance Foundation for their clinical and educational activities support. We thank Frank Hu for advice in manuscript preparation, Leslie Kirle for administrative support, and Rita Buckley for research and editorial services. This report on WLS was prepared for the Betsy Lehman Center for Patient Safety and Medical Error Reduction (Commonwealth of Massachusetts, Boston, MA). Manuscript preparation was supported, in part, by the Boston Obesity Nutrition Research Center Grant P30-DK- 46200 and the Center for Healthy Living, Division of Nutrition, Harvard Medical School.
The authors declared no conflict of interest.
To review task group appendices, go to www.mass.gov/dph and search “Weight Loss Surgery.”