Systematic Review and Meta-Analysis of Bariatric Surgery for Pediatric Obesity

Jonathan R. Treadwell, PhD; Fang Sun, MD, PhD; Karen Schoelles, MD, SM


Annals of Surgery. 2008;248(5):763-776. 

In This Article


In considering bariatric surgery for pediatric patients, 3 unique issues arise: informed consent, interference with physical growth/maturation, and compliance with postsurgical diets.[27] Regarding informed consent, Inge et al(2004)[36] stated that one important ethical consideration is whether the pediatric patient has decisional capacity. Determining decisional capacity often requires consultation with the family, and patients without such capacity should not be treated surgically. Even with good decisional capacity when surgery is elected, some pediatric bariatric patients may later regret the decision to undergo surgery. If so, bariatric procedures that are more easily reversed (such as LAGB) may receive greater consideration in the pediatric population.

Another concern is the potential for bariatric surgery to interfere with physical growth and/or sexual maturation. In our review, only one study formally evaluated the growth of patients after surgery,[29] and the authors considered the data inconclusive. We note that the overall average age of patients in the included studies was 16.7, and our analysis of CDC growth charts suggests that an average boy of this age has completed 98.4% of his growth to age 20, and the corresponding percentage for a girl is 99.6%. These percentages suggest that, among the pediatric patients who have received bariatric surgery, the impact of surgery on height is largely moot. The impact on other aspects of maturation, however, is unclear.

Compared with adults, pediatric patients may have lower levels of compliance with postsurgical dietary regimens, dietary supplements, and exercise recommendations. One study included in our review reported that only 13% of pediatric patients continued taking nutritional supplements as instructed.[29] No other included studies examined the issue. To adequately address concerns about low compliance, additional evidence is needed from future studies.

Turning to surgical outcomes, the amount of weight loss after bariatric surgery among pediatric patients appears to be clinically significant. The F.D.A. defines clinically significant weight loss for pharmacotherapy as 5% of body weight.[72] Trials in diabetes prevention in obese patients have targeted 7% of body weight as a meaningful weight loss goal.[45,46] In a pediatric patient of average age, height and BMI in the included studies, the 5% and 7% goals correspond to BMI reductions of only 2.7 and 3.7 units, respectively. Our meta-analyses of BMI reductions (Fig. 1) show that postsurgical weight loss far exceeds these targets. The BMI reduction appears to be larger after RYGBP than after LAGB, however the RYGBP patients had larger presurgical BMIs (~52 versus~46), and some of the weight loss difference may be because of the baseline difference.

A more direct measure of clinical impact, however, is the postsurgical status of obesity-associated comorbidities. In the pediatric population, some studies have reported the postsurgical resolution of comorbidities such as diabetes and hypertension. This evidence, however, is sparsely reported. Future decisions about bariatric surgery in this population will be aided by improved reporting of key adolescent health outcomes including comorbidities, bone growth, postsurgical compliance, and physical and sexual maturation. We also found limited evidence on quality-of-life improvements after surgery (only one study's data met the inclusion criteria), and no studies of pediatric patients have followed patients long enough to determine whether bariatric surgery extends survival. Two recent studies suggested that, among obese adults, those who chose bariatric surgery lived longer than those who did not.[73,74]

The applicability of that finding to pediatric patients is unclear. Two LAGB studies have directly compared the postsurgical outcomes of adolescents and adults (these adolescents were included in our meta-analysis of BMI reduction after LAGB).[48,58] In a study by Dolan,[58] 2 years after LAGB, the mean BMI for the 17 adolescents (median age 17; range 12-19) had dropped from 42.2 to 30.1, whereas for the 17 matched adults (median age 41; range 23-70) it had dropped from 41.8 to 33.1. Similarly, Dillard et al (2007)[48] reported similar weight loss between adolescents and adults. These findings suggest that the BMI impact of surgery does not depend on age, but additional research is necessary to permit firm conclusions.

The Teen Longitudinal Assessment of Bariatric Surgery(Teen-LABS) is an ongoing four-center study devoted to prospective examination of the outcomes of pediatric bariatric surgery.[75] The plan is to enroll 200 patients, all aged 19 or less, and examine outcomes after bariatric surgery (mostly RYGB between 2007-2009) (see for details). Outcomes to be collected include not only weight loss but medical comorbidities, early (30 days) and late (1-2 years) complications, and psychosocial status (eating behavior, depressive symptoms, and health-related quality of life). Also, researchers will compare these results to the surgical outcomes of 200 adults who had a history of severe obesity before the age of 18 but did not receive surgery at that time. This comparison may shed light on the health implications of postponing surgery.

In general, pediatric patients who undergo bariatric surgery have previously had unsuccessful weight loss with nonsurgical methods. The included studies, however, did not provide details about which methods were attempted, the intensity of those methods, the duration of attempted treatment, or the success criteria. The United Kingdom National Institute for Clinical Excellence (NICE) published a systematic review in December 2006 of obesity treatments, including nonsurgical weight loss treatments in pediatric patients.[68] The trials reviewed, however, do not shed light on the anticipated efficacy of nonsurgical methods in patients who qualify for bariatric surgery. Those trials generally involved patients who were younger and less obese (eg, BMI ~30, age ~12 as opposed to BMI ~50 and age ~17), and some studies followed patients for only a short period of time (eg, 9 months). The single trial in our review that compared surgical to nonsurgical approaches did not include patient groups of comparable weight (and probably not of comparable comorbidity burden).

The NICE review was one of 5 other systematic reviews addressing the use of bariatric surgery for pediatric patients, all published between 2003 and 2006.[76,77,78,79,80] Of the 27 included articles in our review, 13 were not included by any of the other reviews, mostly because of later publication dates. Eight other publications were included by other reviews but not ours. We excluded these either because the bariatric procedure was outdated;[81,82,83,84,85] because authors combined data on different procedures;[86] because it was a case study;[87] or because it focused on a single adverse event.[88]

In the absence of randomized controlled studies, decision makers - including clinicians, patients and policy makers - are forced to make choices based on the information that is currently available. In the interest of making nonrandomized surgical studies more useful for this purpose, we encourage improvements in the reporting of data collection, patient follow-up and outcomes. Elapsed follow-up times for study patients vary over a wide range as do time points chosen for evaluation. Overlap of patient populations in subsequent publications documenting longer follow-up and/or additional patient experience may not always be readily apparent. Right-censoring (ie, patients with more recently performed surgery) is often not distinguished from loss to follow-up. Slim et al (2003)[89] have developed and validated a quality assessment tool for surgical publications which merits consideration by authors in this field. Table 5 lists several items in the instrument that seem particularly relevant to the bariatric surgery literature.

We believe that the present review contributes to the literature in several ways. First, we included 13 publications not previously available. Second, we rated the strength of the evidence using a rating system to incorporate the quality, quantity and consistency of the evidence. Third, we performed meta-analyses to estimate the overall impact on BMI, separately for different procedures. Fourth, we considered the clinical significance of BMI changes by comparison to established targets. Finally, we performed multiple sensitivity analyses to show that the conclusions about weight loss do not depend critically on analytic assumptions. The limitations of our review reflect the lack of reporting of long-term data on a sufficient number of participants, of comorbidity burden and resolution, and of compliance with postsurgical recommendations. Future studies of bariatric surgery in obese pediatric patients should make an effort to capture and report this important information.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.