The mean (95% CI) decrease in BMI across studies in which this endpoint was assessed was 29% (24–37.5%). Mean weight loss, where measured, was 27% (25–29%). Mean percent excess weight loss (%EWL), when reported, was 90.5% (85–96%) for 1 year post-surgical procedure, and 41.5% (36–47%) at 3 to 5 years post-surgical procedure.

The mean (95% CI) decrease in hemoglobin A1c concentration in diabetic patients was from a baseline of 6.8% (6.4–7.3%) to 5.5% (4.7–6.3%). Additionally, including type 2 diabetes, among the studies that measured overall co-morbidity outcomes, the mean percent (95% CI) resolution of such conditions from baseline was 80.5% (66–95%). Such co-morbidities included dyslipidemia, kidney disease, OSA, osteoarthritis, hypertension, and non-alcoholic fatty liver disease.

Most patients retained symptomatic or non-symptomatic status from baseline to 2 years, and remission was more prevalent than the development of new symptomology across studies. Anxiety, depression, disruptive behavior, and anger all decreased significantly at 2 years, contrasted by increases in mood, self-esteem, and self-concept at the same time point, reflected in IWQOL-Kids, BDI, and perceived confidence scores.

The dramatic efficacy of bariatric surgery on both weight loss and co-morbidity resolution that is well-documented in adult patients is echoed in available research for children and adolescents. Studies show weight loss of up to 96 percent of excess body weight in the short-term (up to 1 year) and 39.4 percent sustained in the longer-term (up to 6 years), complete resolution or substantial improvement in type 2 diabetes, decreased appetite and increased satiety, and significant resolution of musculoskeletal pain (26, 28–33).

One large-scale, longitudinal assessment currently being conducted is Teen-LABS (40, 50), a consortium comprised of six clinical sites and a data-coordinating center, gathering information from participating clinical sites to develop a database of surgical outcomes in adolescents. Primary measured surgical outcomes include percent change in BMI, and remission from baseline type 2 diabetes and hypertension, with secondary measured outcomes being incidence of iron and B12 deficiency, and occurrence of abdominal reoperations.

Many of the studies reviewed herein utilize data currently available from this project (4, 26, 28, 33, 46). In one such study, Inge et al. (33) reported a decrease in mean weight of 27 percent, a decrease in kidney dysfunction in 86 percent of patients who had such dysfunction at baseline, remission of hypertension in 72 percent, and resolution of dyslipidemia in 66 percent. Weight-related quality of life also showed significant improvement. The researchers did find an increase in iron deficiency in over half of the participants, and 13 percent underwent revision surgeries. This study looked at outcomes after 3 years of follow-up; consequently, it is one of the longer-term sources of data available.

Inge et al. also performed a secondary study looking exclusively at adolescents with type 2 diabetes, comparing data from Teen-LABS to that from TODAY, a study in which participants were randomized to receive metformin therapy only, metformin therapy in combination with rosiglitazone, or an intensive lifestyle intervention (26, 51). A cohort of 30 subjects in the Teen-LABS study showed a decrease in mean hemoglobin A1c concentration from 6.8% at baseline to 5.5% over the course of 2 years, while that in the 63 subjects in the TODAY study increased from 6.4 to 7.8% (26). Further, mean BMI in the Teen-LABS group decreased by 29%, while mean BMI in the TODAY group increased by 3.7% (26). These results offer compelling support for bariatric surgery in these cohorts; however, for accurate comparison, it is necessary to examine prospective studies to eliminate possible confounders and biases.

One such prospective cohort, and another source of long-term data is the ongoing study by Zitsman et al. which recently posted 5-year follow-up data for all 137 study participants regarding BMI and %EWL endpoints (47). Researchers found a mean post-laparoscopic banding decrease in %EWL of 35.5 percent (39.4% standard deviation), and a percent BMI decrease of 36.6 (37.5% standard deviation) at 5 years. Similar results are reported by Olbers et al. in a Swedish nationwide study that also followed participants over 5 years; however, in this cohort, the surgical method evaluated was Roux-en-Y gastric bypass (42), rather than gastric banding. Participants in this study had a mean (95% CI) weight reduction of 36.8 kg (32.8–40.9 kg), and a mean decrease in BMI of 13.1 kg/m² (14.5–11.8 kg/ m²). The study also reported significant co-morbidity resolution and a decrease in cardiovascular risk factors over the 5 years; however, researchers noted weight loss of <10 percent in nine participants (11%). These longitudinal studies reflect an even greater degree of efficacy than previously seen in Teen-LABS (40, 50), and again highlight the need for further long-term data.

In a study of 12 months' duration, Kelly et al. (31) reported a significant decrease in leptin levels post-surgical procedure, with a concurrent increase in adiponectin levels but no significant change in resistin levels. Further, researchers found that interleukin-6 (IL-6) levels decreased significantly, as did oxidative LDL cholesterol levels, leading to improved lipid profiles. This study population was small, encompassing 52 subjects in two separate cohorts, one of which included participants from Teen-LABS, but the duration suggests that longer follow-up could affect measured hormone outcomes.

In their study examining energy expenditure post-bariatric surgery, Butte et al. (39) noted not only significant decreases in total, activity, resting, and walking energy expenditure, measured by 24-h room calorimetry, but also a parallel and significant decline in serum insulin, and decreases in fasting serum TSH and total serum triiodothyronine (T3), indicating alterations in thyroid function post-surgery. Leptin also decreased significantly, as did urinary norepinephrine associated with the alterations in energy expenditure and basal metabolic rate (BMR), and substrate utilization increased significantly for carbohydrates, fats, and proteins. Additionally, adiponectin increased, and levels of inflammatory marker CRP decreased significantly (P = 0.01). No significant changes were seen in resistin levels, echoing results produced by Kelly et al. Again, this study had a small population of only 11 subjects and five controls; however, the findings at 1 year closely mirror those seen previously, strengthening the evidence for improved energy and metabolic profiles post-bariatric surgery in these young patients, and suggesting the need for further study.

Recent evidence suggests that bariatric surgery has a significant and positive impact on the psychological co-morbidities of obese teenagers. A study by Schmitt et al. examining the outcomes of laparoscopic adjustable gastric banding, in addition to noting positive physiological outcomes (Tables 1–3), also gathered data on the perceived reasons for the onset of obesity in the adolescent patients (45). According to the responses, more than half of the participants attributed the start of their weight gain to a single, deleterious life event (45), such as divorce or a parental suicide, and the same proportion presented with a psychiatric history at baseline. In addition, study participants noted teasing (n = 11), decreases in self-esteem (n = 7), body dissatisfaction (n = 14), anxiety (n = 5), and attempted suicide (n = 1) as triggers of current psychological distress. This study found an improvement in depressive symptoms and social isolation at 1 year post-surgical intervention, and results from Psychosocial PedsQL evaluations noted statistically significant improvement at both 1 and 2 year endpoints in emotional functioning, social functioning, and school functioning (45). However, the study was performed in a specialty obesity center with a cohort of only 16 subjects, so it is not necessarily generalizable to the population as a whole.

Jarvholm et al. examined a cohort of Swedish adolescents, aged 13–18, in an attempt to determine if the positive psychological outcomes observed in research available for adult patients were mirrored in these younger patients at 2-years post-surgery (44). According to Jarvholm et al. (44), anxiety, depression, anger, and disruptive behavior all decreased significantly at 2 years post-bariatric surgery in a cohort of 88 adolescents, using Beck Youth Inventory (BYI) scores. Additionally, self-esteem and self-concept improved significantly at 1 year post-surgery, with stabilization occurring by the 2-year mark.

Hunsaker et al. (43) found less dramatic improvement over 2 years, with most of the subjects maintaining their symptomatic psychological status; however, remission or improvement was more common than the occurrence of new symptomology. These findings were mirrored by Zeller et al. (46); however, this study utilized Youth Self-Report (YSR) and Adult Self-Report (ASR) examinations, which have the potential to introduce bias. The study had a follow-up period that lasted as long as 6 years, however, making it one of the longest-term studies currently available for analysis. It is important to note the difficulty of evaluating psychological improvement, as studies measure according to clinical diagnosis, which does not always apply uniformly across individuals. Further, self-reported methods are often employed, resulting in potential bias. More study, particularly long-term and objective, is needed to clearly assess the mental health changes, though initial results in both adult and adolescent studies lend strong evidence to the efficacy of bariatric surgery on psychological endpoints.

Sleep disorders such as obstructive sleep apnea (OSA), circadian rhythm disorders, and poor sleep quality show a high prevalence in obese adolescents (17). He et al. found that habitual sleep variability is significantly associated with abdominal and visceral obesity (17), and even more compelling, that 20 percent of the association with visceral adiposity can be attributed to carbohydrate intake, an effect more clearly delineated in a recent review by St-Onge et al. (18). Additionally, due to a delay in melatonin release during the adolescent years, teenagers are more prone to sleep restriction resulting from the delayed sleep onset, combined with the early start of the typical school day (17). Such short sleep duration can lead to weight gain and eventual obesity, due to an increase in neuronal reward responses to unhealthy foods, potential alterations in hormonal activity (leptin, ghrelin, GLP-1, orexins), and the increased time awake that leads to higher caloric intake (18–20). Similarly, in a randomized, crossover, sleep restriction-extension paradigm, Beebe DW et al. reported that adolescents with chronic sleep restriction tend to consume high glycemic index foods, and that such sleep restriction could cause lasting changes in dietary habits that increase the risk of obesity into adulthood (52).

Many obese adolescents also experience varying degrees of OSA (37), and according to findings by Talib et al. OSA is closely related to BMI Z-score (3.7 vs. 3.4; P = 0.003) (53).

Mounting evidence suggests that bariatric surgery is instrumental in reversing or minimizing the impact on sleep caused by obesity (36, 37), and can even be beneficial in stabilizing hormonal factors that interfere with proper sleep structure (38). According to findings by Amin et al. (38), apnea hypoxia index (AHI) decreased post-surgery from 15.8 events per hour at baseline to 9.1 events per hour at 5 weeks post-surgery. In the same cohort, leptin levels decreased significantly and orexin levels increased significantly, demonstrating an early alteration in hormone activity following surgical intervention. Ashrafian et al. (37) found similar improvements in OSA in their recent meta-analysis, noting significantly stronger results in those who received bariatric surgery vs. traditional weight loss interventions. In one of the few studies examined that focused on gastric banding, Zitsman et al. (34) utilized polysomnography to diagnose 35.8% of study subjects (49 patients) with OSA, and the researchers noted a resolution of OSA in 17 of those patients post-surgery. It is important to note, however, that complication rates in this study were more prevalent than in studies investigating sleeve and bypass techniques, and improvement in OSA was measured by self-report rather than formal re-testing. Few studies are available that examine the changes in OSA following bariatric surgery in adolescents specifically—and even fewer look at the long-term efficacy; consequently, more data is needed prior to definitive conclusions regarding the positive outcomes noted here.

As with the vast majority of medical procedures, socioeconomic status (SES) plays a role in patient eligibility due to quality of insurance, cost of support structures and specialists, proximity to high-quality facilities, and availability of transportation/lodging when necessary. Because obesity has been strongly linked to those of lower SES and has a disproportionate impact on Black and Hispanic adolescents when compared with non-Hispanic whites (11, 41, 54–56), considering not only SES, but also race, ethnicity, and geographic location is important when examining the accessibility of bariatric surgery for some of the most at-risk populations. In reviewing the current literature, very few studies specifically focus on these areas with regard to bariatric surgery outcomes in adolescents.

Gullick et al. (56) examined both race and SES in relation to bariatric surgery outcomes, finding similar improvements in co-morbidities, but lower weight loss in African American subjects than European Americans at 1, 2, and 5 year endpoints. The study, however, enrolled 74.4 percent European Americans, and only 25.6 percent African Americans, and it also studied adult, rather than adolescent, subjects. Based on the evidence, it appears that SES may be a bigger predictor of outcomes than ethnicity; however, not enough data exist to make a definitive conclusion (55, 56). What can be assessed is that those of lower SES or living in rural areas further from reputable medical centers, have more post-operative complications and lower success rates than those with better access to resources (11).

Perez et al. found that age-adjusted severe obesity rates were higher in minority adolescent groups: black adolescents have a rate of 8.6% and Hispanic adolescents a rate of 8.4%, while white adolescents have less than half that rate at 4.1% (P = 0.03) (57). The researchers also found that white adolescents utilize bariatric surgery at a much higher rate than black or Hispanic patients, and that there is a paradoxical relationship with regard to insurance type (57): minority patients are even less likely to undergo weight loss surgical procedures if they carry Medicaid rather than private insurance; however, white patients with Medicaid are much more likely to undergo such procedures than those insured by private carriers. Though the reasons for such dichotomy are as of yet unclear, what Perez et al. pointed out was that the solution is far more complex than a simple expansion of government-sponsored health insurance. As with many complicated issues, there are several components at play here, including cultural context/norms, health education, and physical access to care. Each must be examined specifically within the context of the adolescent population to determine future public health measures that could mitigate such disparity and ensure a more equitable distribution of care.