Synthesized cardiometabolic, pulmonary, and sleep outcomes after bariatric surgery in adolescents with severe obesity: A systematic review and meta-analysis

Abstract

Background

To synthesize cardiometabolic, pulmonary, sleep-related, anthropometric, and safety outcomes after metabolic and bariatric surgery in adolescents with severe obesity.

Methods

This was a systematic review and meta-analysis conducted according to PRISMA 2020 and prospectively registered in PROSPERO (CRD420261354795). Five databases were searched for studies of adolescents aged approximately 10–19 years undergoing recognized metabolic or bariatric procedures. Two reviewers independently performed study selection, data extraction, and risk-of-bias assessment. Random-effects meta-analysis was used when studies were sufficiently comparable.

Results

Of 5110 records, 147 reports representing 101 unique studies or cohorts published from 1995 to 2026 were included. Follow-up ranged from 30 days to approximately 15 years. Across three independent randomized cohorts, surgery reduced body mass index by 9.39 kg/m² at 2 years versus non-surgical care (95% CI −13.90 to −4.89; I² = 88%; low certainty). Randomized evidence indicated lower triglycerides (mean difference −0.36 mmol/L, 95% CI −0.57 to −0.14; low certainty), whereas blood-pressure effects were indeterminate (very low certainty). Observational cohorts reported improvement or remission of type 2 diabetes, hypertension, dyslipidemia, and obstructive sleep apnea. Weight reduction was durable after gastric bypass and sleeve gastrectomy but smaller and less durable after gastric banding. Repeat objective sleep testing was reported in three studies and postoperative spirometry in one. No included report documented use of a validated asthma-control instrument, ambulatory blood-pressure monitoring, or vascular imaging as a pre- and postoperative outcome; diastolic function was reported once; diastolic function was reported once. Perioperative mortality and serious 30-day complications were uncommon, whereas reoperation and micronutrient deficiency remained important during extended follow-up.

Conclusions

Metabolic and bariatric surgery produces large body mass index reductions and multisystem improvement in adolescents with severe obesity; serious short-term complications and perioperative mortality were uncommon, but lifelong nutritional and multidisciplinary surveillance remains important. Most effect estimates remain uncertain, particularly for objective long-term sleep and pulmonary outcomes.

Graphical abstract

1. Introduction

Severe obesity during adolescence is a chronic, relapsing disease with consequences that extend across the life course. Although overall trends vary by region, severe forms of pediatric obesity have continued to rise and are associated with early cardiometabolic disease, impaired quality of life, and increased risk of premature adult morbidity and mortality [1]. The burden is not confined to excess body weight. Adolescents may present with hypertension, dyslipidemia, insulin resistance or type 2 diabetes, inflammatory and hepatic abnormalities, respiratory limitation, and sleep-disordered breathing. Obstructive sleep apnea is particularly common among children and adolescents with obesity and is itself associated with adverse cardiometabolic and neurocognitive outcomes [2].

These complications are biologically and clinically interconnected. Excess adiposity can alter chest-wall mechanics and ventilatory reserve, increase upper-airway collapsibility, and contribute to nocturnal hypoxemia and fragmented sleep. Recurrent hypoxemia, sympathetic activation, and sleep disruption may then aggravate blood-pressure and glucose dysregulation. Consequently, improvement in weight, pulmonary physiology, and sleep may occur together, but the magnitude and timing of change need not be uniform across domains. Evidence from broader bariatric populations suggests that surgery can improve pulmonary function and sleep architecture, although such syntheses have not been restricted to adolescents [3]. Adolescent-specific studies have also shown that obstructive sleep apnea may remit after surgery, while persistent disease remains clinically important and may not be explained by early weight loss alone [4].

For adolescents with severe obesity and clinically significant comorbidity, metabolic and bariatric surgery is now recognized as an evidence-based treatment within multidisciplinary care. Contemporary guidance supports consideration of surgery for appropriately selected adolescents with class II obesity and a major comorbidity or class III obesity, without requiring attainment of a particular Tanner stage or linear-growth milestone [5]. Sleeve gastrectomy and Roux-en-Y gastric bypass are the most commonly used procedures, while adjustable gastric banding has declined because of concerns about durability and reoperation. Prospective adolescent cohorts have reported substantial weight reduction accompanied by remission or improvement of type 2 diabetes, hypertension, dyslipidemia, and abnormal kidney function [6]. Five-year data from the AMOS cohort likewise showed sustained weight loss and improvement in comorbidities relative to non-surgical care [7].

The potential benefits of surgery must be interpreted alongside short- and long-term harms. Follow-up studies have identified nutritional deficiencies, additional abdominal procedures, and the need for lifelong clinical surveillance. In a five-year comparison, adolescents and adults experienced similar weight loss after gastric bypass, while adolescents had favorable remission of some cardiometabolic conditions but also substantial micronutrient deficiencies [8]. A systematic review of studies with at least five years of follow-up similarly concluded that durable body mass index reduction and comorbidity remission were supported by evidence of limited certainty and that long-term complications were incompletely reported [9]. These findings make safety and follow-up integral to any assessment of multisystem benefit.

Despite the expanding evidence base, existing syntheses have generally answered narrower questions. The updated Cochrane review was restricted to randomized trials and found only one small trial, illustrating why long-term adolescent evidence cannot be characterized adequately from randomized evidence alone [10]. Other reviews have focused on long-term weight and selected comorbidities [9], whereas pulmonary and sleep meta-analyses have largely drawn from adult or mixed-age populations [3]. Cardiometabolic, pulmonary, sleep, and safety outcomes have therefore rarely been synthesized together in a way that preserves adolescent specificity, distinguishes unique cohorts from companion publications, and evaluates the durability of change across clinically meaningful follow-up intervals.

This systematic review and meta-analysis was undertaken to evaluate the effects of metabolic and bariatric surgery on cardiometabolic, pulmonary, and sleep-related outcomes in adolescents with severe obesity, while also examining anthropometric outcomes and adverse events specified in the registered protocol. The review further aimed to compare effects across procedure types and follow-up durations, explore associations between weight loss and multisystem outcomes when data permitted, and identify limitations in adolescent-specific and long-term evidence. The PICOS question was: among adolescents aged approximately 10–19 years with severe obesity, what changes in protocol-relevant cardiometabolic, pulmonary, sleep, anthropometric, and safety outcomes follow recognized metabolic or bariatric surgery, compared with preoperative status, non-surgical management, or another surgical procedure, in randomized, non-randomized interventional, cohort, or eligible pre–post studies?

2. Materials and methods

2.1. Protocol and reporting

The review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 statement [11]. The protocol was prospectively registered in PROSPERO (CRD420261354795). The registration and protocol defined the population, eligible surgical procedures, outcomes, study designs, risk-of-bias methods, and planned synthesis. Any departure from the registered methods is identified, justified, and reported transparently. One descriptive analysis was added after peer review: a post hoc audit of reporting gaps in clinically relevant measures within the prespecified outcome domains (Section 2.11). This analysis did not alter the eligibility criteria, prespecified outcomes, or quantitative synthesis and is identified throughout as post hoc.

2.2. Eligibility criteria

2.2.1. Population

Studies were eligible if they enrolled adolescents aged approximately 10–19 years, or used an adolescent definition specified by the investigators, with severe obesity undergoing bariatric or metabolic surgery. Mixed-age studies were eligible only when data for the adolescent subgroup were separately extractable. Adult-only studies, mixed populations without separable adolescent data, populations without severe obesity, and surgery performed for an indication other than obesity were excluded.

2.2.2. Interventions and comparators

Eligible interventions were sleeve gastrectomy, Roux-en-Y gastric bypass, adjustable gastric banding, one-anastomosis gastric bypass, and other recognized metabolic or weight-loss surgical procedures. Lifestyle-only, pharmacological-only, and endoscopic non-surgical procedures were excluded. Eligible comparators included preoperative baseline, non-surgical management, usual care, and another bariatric procedure. A concurrent comparator was not required; eligible single-group pre–post studies were retained.

2.2.3. Outcomes

At least one protocol-relevant outcome was required. Outcomes comprised body mass index and other weight-loss measures; obstructive sleep apnea remission; apnea-hypopnea index; oxygen saturation and sleep architecture; glycated hemoglobin, insulin resistance, glucose status, lipids, blood pressure, and inflammatory markers; forced expiratory volume in 1 s, forced vital capacity, and other respiratory outcomes; and adverse events, complications, hospital stay, and other specified perioperative outcomes. Cardiopulmonary exercise testing was not treated as an independent review domain, and CPET-only studies were not added outside the registered scope.

Study designs. Randomized controlled trials, non-randomized interventional studies, prospective or retrospective cohort studies, and eligible pre–post studies were included. Case reports, case series with fewer than 10 surgically treated participants, reviews, meta-analyses, editorials, and conference abstracts without a complete report were excluded. Only published English-language studies were sought, consistent with the registered record. The standard full-text exclusion labels were wrong population, wrong intervention, no relevant outcomes, wrong study design, insufficient data, duplicate study, and conference abstract only.

2.3. Information sources and search strategy

A comprehensive search was conducted from database inception to August 11, 2026 in MEDLINE/PubMed through the NCBI PubMed interface, Embase (Embase.com), the Cochrane Central Register of Controlled Trials (CENTRAL; Cochrane Library), the Science Citation Index through the Web of Science Core Collection (Clarivate), and Scopus (Elsevier). No publication-date restrictions were applied, although eligibility was restricted to English-language publications in accordance with the registered protocol. Strategy development initially used four concepts: population, obesity severity, intervention, and outcomes. Before the final searches were executed, the outcome concept was removed to avoid restricting retrieval according to outcome reporting, and outcome eligibility was assessed during screening and data extraction. Severe or morbid obesity was retained as a required population concept. Because requiring explicit obesity-severity terminology may have reduced sensitivity for reports that did not use these terms in searchable fields, this feature of the search strategy was considered when interpreting the completeness of retrieval. Controlled vocabulary was used in MEDLINE/PubMed; the Embase, Scopus, Web of Science, and CENTRAL strategies used free-text terms in title, abstract, and keyword fields. Scopus, Web of Science, and CENTRAL specified named bariatric procedures rather than relying on a generic intervention term. Complete executed strategies, platforms, search dates, and final cumulative retrieval counts are provided in Supplement 1. Trial- and study-registry searching was undertaken as prespecified, together with backward reference-list screening and forward citation tracking of included studies and relevant reviews. Registry-specific executed strings, per-registry yields, and update-specific yields were not retained in the project record and are therefore reported as unavailable in Supplement 1 rather than reconstructed retrospectively. The searches were first executed on April 17, 2026 and were rerun in full on July 7, 2026 and again on August 11, 2026, the last of these being the date on which every source was last searched; the study-selection tables were finalized thereafter.

2.4. Study selection

All 5110 retrieved records were imported into a single review library. Duplicate candidates were identified using DOI, title, and publication-year matching supplemented by fuzzy title comparison. Every candidate duplicate was manually reviewed before removal. In total, 556 records were merged and 1212 duplicate records were removed; 21 candidate pairs were judged not to be duplicates. Manual review also reinstated seven records that had been incorrectly flagged as duplicates; all seven subsequently underwent screening and were excluded on study design. After deduplication, 3898 records remained for screening. Screening for adolescent populations relied on title and abstract text alone. Database-assigned index keywords were not used for this purpose, because an audit showed that they flagged several hundred adult-only bariatric records as adolescent. Records for which no complete abstract could be obtained were not screenable and were excluded at this stage. Two reviewers, Purushottam Lal (PL) and Nitin Kumar (NK), independently screened titles and abstracts in Rayyan's blinded mode using a prespecified checklist. Records considered potentially eligible by either reviewer proceeded to full-text assessment. The same reviewers independently assessed the full-text reports against the registered eligibility criteria. Disagreements were resolved through discussion and, when consensus could not be reached, through adjudication by Surendra Gupta (SG). Each excluded full-text report was assigned one primary exclusion reason using the seven standardized labels. No study authors were contacted for additional information. The study-selection process is summarized using a PRISMA 2020 flow diagram (Fig. 1).

Fig. 1.

PRISMA 2020 flow diagram showing study identification, screening, eligibility assessment, and inclusion.

2.5. Data extraction and management of companion reports

Two reviewers independently extracted data using a standardized, piloted form. Extracted variables included publication details, country and setting, study design, recruitment period, sample size, age and sex distribution, definition and severity of obesity, baseline comorbidities, procedure type, comparator, follow-up duration, outcome definitions, numerical results, attrition, funding, and conflicts of interest. Disagreements were resolved by consensus or third-reviewer adjudication. Reported data were not inferred when absent, and study authors were not contacted for additional information.

Multiple publications from the same underlying cohort were linked before synthesis. Cohort identity was assessed using study name, center, recruitment dates, sample size, procedures, participant characteristics, and author group. For each outcome and follow-up interval, the most complete report was selected as the primary source; companion publications were used only when they provided additional outcomes, longer follow-up, or otherwise non-overlapping information. Participants were counted once in any meta-analysis.

2.6. Outcome grouping and follow-up

Outcomes were organized into cardiometabolic, pulmonary, sleep, anthropometric, and safety domains. Follow-up was grouped as short term (up to 12 months), mid term (more than 12 months to less than 5 years), and long term (5 years or longer), while preserving study-reported time points. Results from randomized or comparative studies, uncontrolled cohorts, and pre–post studies were synthesized separately when design differences could alter interpretation.

2.7. Risk of bias within studies

Two reviewers, Purushottam Lal (PL) and Nitin Kumar (NK), independently assessed risk of bias for each outcome result included in the synthesis, with disagreements resolved by discussion and, when necessary, adjudication by Surendra Gupta (SG). The appraisal instrument was matched to study design. Randomized trials were evaluated using the Cochrane Risk of Bias 2 (RoB 2) tool across bias arising from the randomization process, deviations from intended interventions, missing outcome data, outcome measurement, and selection of the reported result. Non-randomized studies of interventions that included a concurrent comparator, such as surgery versus medical or lifestyle treatment or one procedure versus another, were evaluated using ROBINS-I across bias due to confounding, participant selection, intervention classification, deviations from intended interventions, missing data, outcome measurement, and selection of the reported result; the effect of interest and an appropriate target trial were specified before each assessment. Uncontrolled cohort studies and single-group pre–post studies were appraised using the JBI Critical Appraisal Checklist for Cohort Studies, analytical cross-sectional studies using the JBI Checklist for Analytical Cross-Sectional Studies, and case series of at least 10 participants using the JBI Checklist for Case Series. For each JBI assessment, item-level responses were summarized into an overall category ranging from low to high risk of bias.

For uncontrolled single-group designs, particular attention was paid to secular trends, maturation, regression to the mean, co-interventions, attrition, consistency of outcome measurement over time, and selective reporting, because a single preoperative and a single postoperative measurement cannot reliably separate the effect of surgery from other time-related causes. Such evidence was treated as providing weaker causal support, was synthesized separately from controlled studies, and was interpreted with corresponding caution; results considered too unreliable for meaningful interpretation were not carried into any quantitative synthesis. Domain- or item-level and overall judgments were recorded with supporting rationale, and risk-of-bias judgments informed interpretation, sensitivity analyses, and certainty-of-evidence ratings.

2.8. Effect measures and data synthesis

Meta-analysis was undertaken only when studies were sufficiently comparable in population, procedure, outcome definition, follow-up, and design. Continuous outcomes measured on a common scale were summarized as mean differences with 95% confidence intervals, and standardized mean differences were used when conceptually similar outcomes were measured on different scales. Dichotomous comparative outcomes were summarized as risk ratios with 95% confidence intervals. Single-group event outcomes, including remission and adverse-event frequencies, were summarized as proportions when appropriate. Pooling of such proportions was made conditional on the contributing cohorts having ascertained the outcome in a comparable way, on the numerator and denominator being explicit, and on the denominator representing all affected participants rather than only those reassessed. Where remission was established by different methods across cohorts, the frequencies were tabulated and described rather than combined, because a pooled proportion drawn from incompatible definitions would misrepresent its own precision. Change scores and post-intervention values were not combined without a defensible transformation or separate analysis.

Random-effects models were specified because clinical and methodological heterogeneity was expected. Statistical heterogeneity was evaluated using the chi-square test and the I² statistic, with interpretation also considering the direction, magnitude, and clinical sources of inconsistency. Analyses were conducted in Review Manager Web (RevMan Web; Cochrane), using the inverse-variance method for continuous and generic effect estimates and the Mantel–Haenszel method for dichotomous outcomes, with between-study variance estimated by the DerSimonian–Laird method. A continuity correction of 0.5 was applied to comparative outcomes containing a zero cell, and studies with no events in either group were retained in the narrative synthesis only. Quantitative pooling was possible only for randomized evidence, for four outcomes at 2 years; observational outcomes did not meet the comparability threshold and were summarized using a structured narrative synthesis organized by outcome domain, design, procedure, and follow-up. As sensitivity analyses for the primary pooled outcome, a Hartung–Knapp-adjusted confidence interval and a leave-one-out estimate were computed separately.

2.9. Subgroup, meta-regression, and sensitivity analyses

Prespecified subgroup analyses considered procedure type, baseline obesity severity, baseline obstructive sleep apnea severity, and follow-up duration. Where the number and distribution of studies were adequate, meta-regression was planned to explore associations of weight loss with sleep and cardiometabolic improvement. Sensitivity analyses were planned by study design, risk of bias, alternative statistical assumptions, and exclusion of studies at greatest risk of participant overlap. Because no outcome yielded a sufficient set of independent, comparable studies, subgroup analyses and meta-regression could not be undertaken.

2.10. Reporting bias and certainty of evidence

For any meta-analysis containing at least 10 studies, small-study effects were to be examined using funnel plots alongside consideration of selective outcome reporting and protocol–publication discrepancies. No outcome reached this threshold, so formal assessment of small-study effects was not performed. The certainty of evidence for each principal outcome was evaluated using the GRADE framework across risk of bias, inconsistency, indirectness, imprecision, and publication bias, and summary-of-findings entries were prepared for the pooled outcomes (see Section 3.10).

2.11. Post hoc audit of unreported outcome measures

This descriptive analysis was not specified in the registered protocol and was added in response to peer review. Its purpose was to identify clinically relevant measures that were absent from, or inconsistently reported in, the included literature within the prespecified cardiometabolic, sleep, and pulmonary domains. It did not alter study eligibility, introduce a new clinical outcome domain, or contribute to any quantitative synthesis. For sleep outcomes they were the polysomnographic definition of obstructive sleep apnea, including whether the total apnea–hypopnea index, the obstructive apnea–hypopnea index, or the obstructive apnea index was used; whether pediatric or adult respiratory scoring rules were applied; and reporting of the oxygen desaturation index, oxygen-saturation nadir, carbon dioxide monitoring, arousal index, sleep architecture, and validated sleep questionnaires. For pulmonary outcomes they were validated asthma-control instruments (the Asthma Control Test, Childhood Asthma Control Test, or Asthma Control Questionnaire), asthma exacerbations, the definition used for asthma and its remission, spirometry, fractional exhaled nitric oxide, evaluation for obesity hypoventilation, and echocardiographic assessment of pulmonary hypertension.

The full texts of all 147 included reports were electronically searched for terms corresponding to each candidate measure after text extraction; one scanned report required optical character recognition before searching. All identified matches were subsequently reviewed in context by the review team. A measure was counted as reported only when it was used as an outcome before and after surgery or, for definitions, when it was the stated basis for classifying participants. Mentions confined to introductions, discussions, or reference lists were disregarded. Supplement files and trial registrations were not searched. Findings are presented descriptively by report and cohort (see Section 3.11); no measure was pooled, and no certainty rating was assigned.

3. Results

3.1. Study selection

The searches yielded 5110 records: 2565 from MEDLINE/PubMed, 1456 from Scopus, 843 from Embase, 168 from the Science Citation Index via the Web of Science Core Collection, and 78 from the Cochrane Central Register of Controlled Trials. After removal of 1212 confirmed duplicates, 3898 records underwent title-and-abstract screening, of which 3643 were excluded. Full texts were sought for 255 records; 13 could not be retrieved, leaving 242 reports for eligibility assessment.

Ninety-five reports were excluded at full-text review: wrong population (n = 63), wrong study design (n = 14), no relevant outcomes (n = 10), conference abstract only (n = 3), insufficient data (n = 2), wrong intervention (n = 2), and duplicate study (n = 1). A total of 147 reports met the eligibility criteria. After linking companion publications to their underlying research programmes and cohorts, these represented 101 unique studies or cohorts (Supplement 2). The study-selection process is summarized in Fig. 1.

3.2. Study and participant characteristics

The 147 included reports were published between 1995 and 2026 and represented 101 unique studies or cohorts. Several were companion publications from major research programmes or databases, including Teen-LABS, AMOS/AMOS2, BASIC, FABS/FABS-5+, IMPROVE-T2D, BOLD, PCORnet, and MBSAQIP, and were linked to their underlying cohorts to prevent duplicate participant counting. Studies were conducted across North America, South America, Europe, the Middle East, Asia, and Australia and included single- and multicentre cohorts, registries, and population-level databases. Six reports presented randomized evidence: four from the BASIC trial, one from AMOS2, and one Australian trial of adjustable gastric banding; most remaining reports used cohort, registry-based, matched comparative, or uncontrolled pre–post designs [12], [13], [14], [15]. Individual report characteristics are summarized in Table 1, with cohort and companion-publication linkages detailed in Supplement 2.

Table 1.

Characteristics of the 147 included reports.

S. No.	Author & Year	Study Design	Population	Intervention	Comparator	Outcomes Measured	Key Findings	Limitations
1	Carbajo et al., 2019 [16]	Retrospective cohort study (pre–post follow-up)	39 adolescents aged 13–19 years with morbid/severe obesity (mean BMI 48.2 kg/m2)	One-Anastomosis Gastric Bypass (OAGB)	Preoperative baseline vs postoperative follow-up (1, 2, and 5 years)	Anthropometric (BMI, weight loss), cardiometabolic (T2DM, hypertension, dyslipidemia remission), sleep-related (sleep apnea/hypopnea syndrome prevalence), safety outcomes (complications, nutritional deficiencies)	Mean BMI decreased from 48.2 to 25.9 kg/m2 at 5 years; total weight loss 38.8%; 100% remission of T2DM, hypertension, and dyslipidemia; no malnutrition or growth disorders reported; one case of iron-deficiency anemia requiring treatment.	Retrospective design; small sample size (n = 39); single-center study; no non-surgical control group; low baseline prevalence of some comorbidities may limit interpretation of remission rates.
2	El-Matbouly et al., 2018 [17]	Retrospective cohort study with 1- and 5-year follow-up	91 morbidly obese adolescents aged 10–19 years (mean age 17 ± 1.5 years; mean BMI 48 ± 7.5 kg/m2)	Laparoscopic Sleeve Gastrectomy (LSG)	Preoperative baseline compared with postoperative follow-up (1 and 5 years)	Anthropometric (BMI, weight loss, %EWL, %TWL), cardiometabolic (HbA1c, diabetes remission, hypertension), sleep-related (obstructive sleep apnea), safety outcomes (complications), psychological outcomes (body image)	Significant reduction in BMI (48.0–30.3 kg/m2 at 5 years); %EWL 78%; complete remission of all prediabetes cases and 75% remission of diabetes at 5 years; 64% resolution of OSA; hypertension resolved; significant improvement in body image satisfaction and psychological well-being.	Retrospective design; single-center study; no control group; limited assessment of some comorbidities; BIQ response rate only 58%; follow-up loss at 5 years.
3	Zitsman et al., 2015 [18]	Prospective longitudinal cohort study	137 morbidly obese adolescents (mean age 17.0 ± 1.2 years; mean BMI 48.3 kg/m2; ages 14–19 years)	Laparoscopic Adjustable Gastric Banding (LAGB)	Preoperative baseline compared with postoperative follow-up (up to 3 years)	Anthropometric (weight, BMI, %EWL, %EBMIL), cardiometabolic (diabetes, insulin resistance, metabolic syndrome, hypertension, dyslipidemia, PCOS), sleep-related (OSA), psychological outcomes, safety outcomes	Mean BMI decreased from 48.3 to 39.3 kg/m2 at 36 months; %EWL increased to 41.2% and %EBMIL to 41.1% at 3 years; improvement/resolution of multiple comorbidities including insulin resistance (40.3%), metabolic syndrome (37.5%), hypertension (33.3%), dyslipidemia (31.7%), OSA (34.7%), and PCOS (50%); no perioperative mortality.	Follow-up attendance declined over time; some outcomes (OSA, asthma) based on self-report rather than repeat objective testing; 22% required additional operations for complications; 20% underwent band removal/conversion; no non-surgical control group
4	Perera et al., 2026 [19]	Prospective cohort study (36-month follow-up)	20 adolescents aged 15–17 years with severe obesity (BMI ≥35 kg/m2); 85% had prediabetes or type 2 diabetes	Very-low-calorie diet (VLCD) plus laparoscopic sleeve gastrectomy (n = 7)	VLCD/lifestyle intervention only (n = 13)	Renal outcomes (UACR, eGFR), anthropometric outcomes (weight), glycemic outcomes (HbA1c), lipid profile, liver enzymes	Bariatric surgery group achieved sustained weight loss (∼31% at 36 months), 83–86% reduction in albuminuria (UACR), improved HbA1c (8.1%–5.8%), improved triglycerides and liver enzymes. Lifestyle-only group showed minimal long-term improvement. eGFR remained stable in both groups.	Small sample size (n = 20), non-randomized design, potential selection bias, single-center study, predominantly Māori/Pacific population limiting generalizability, missing data related to COVID-19 disruptions.
5	Savas et al., 2025 [20]	Multicenter observational cohort; data recorded prospectively and analyzed retrospectively	169 adolescent patients from six Turkish bariatric and metabolic surgery centers; mean age 17 years (range 11–19); mean baseline BMI 43.9 kg/m2. Eligibility generally required BMI >40 kg/m2 or BMI 35–40 kg/m2 with an adiposity-related comorbidity.	Bariatric surgery: laparoscopic sleeve gastrectomy (LSG; n = 159), Roux-en-Y gastric bypass (RYGB; n = 5), one-anastomosis gastric bypass (OAGB; n = 3), and laparoscopic adjustable gastric banding (LAGB; n = 2).	No formal non-surgical comparator; preoperative baseline compared with postoperative follow-up. This is eligible under the protocol.	Anthropometric: postoperative BMI, excess weight loss (EWL), total weight loss (TWL). Cardiometabolic/comorbidities: hypertension and type 2 diabetes, among other obesity-related comorbidities. Safety: perioperative and postoperative complications, length of hospital stay, and ICU admission.	At 6 months, mean BMI decreased to approximately 30.1 kg/m2, with substantial EWL and TWL; the decrease in BMI was statistically significant. Approximately 75% of patients experienced remission or clinical improvement of associated comorbidities. Complications occurred in approximately 1.8%, with no ICU admissions.	Retrospective analysis despite prospective data collection; susceptibility to selection bias; no formal control group; limited reporting of procedure-specific outcomes; lack of detailed assessment of comorbidity remission; dyslipidemia was not assessed; patients with diabetes were not stratified by disease duration or medication use; limited emphasis on long-term outcomes despite a reported mean follow-up of approximately 40 months.
6	Nadler et al., 2008 [21]	Prospective cohort study (database review of prospectively collected data)	73 adolescents aged 13–17 years with morbid obesity (mean BMI 47.6 ± 7.0 kg/m2)	Laparoscopic Adjustable Gastric Banding (LAGB)	Preoperative baseline compared with postoperative follow-up (up to 2 years)	Anthropometric outcomes (% excess weight loss, BMI), cardiometabolic outcomes (hypertension, dyslipidemia, impaired glucose tolerance), sleep outcomes (obstructive sleep apnea), safety outcomes (surgical and nutritional complications), compliance outcomes	Mean %EWL was 35.1% at 6 months, 56.7% at 1 year, and 60.9% at 2 years. Among 50 comorbid conditions assessed at 1 year, 70% resolved and 18% improved. Hypertension and impaired glucose tolerance resolved completely in affected patients. Overall follow-up compliance was 89.5% at 2 years.	Single-center study; no non-surgical control group; only 21 patients had detailed comorbidity analysis; follow-up numbers decreased over time; predominantly Caucasian population.
7	Holterman et al., 2010 [22]	Prospective longitudinal cohort study	20 morbidly obese adolescents aged 14–17 years (mean BMI 50 ± 10 kg/m2)	Laparoscopic Adjustable Gastric Banding (LAGB)	Preoperative baseline compared with postoperative follow-up (12 and 18 months)	Anthropometric (BMI loss, %EWL), cardiometabolic (hypertension, dyslipidemia, insulin resistance, metabolic syndrome), quality of life, safety outcomes	%EWL increased to 34% at 12 months and 41% at 18 months; hypertension resolved in 100% of affected patients; dyslipidemia resolved in 46% (12 months) and 67% (18 months); insulin resistance resolved in 72% by 18 months; metabolic syndrome resolved in 82% by 18 months; significant improvement in quality-of-life scores.	Small sample size (n = 20); interim analysis rather than completed 5-year trial; no non-surgical control group; 20% of enrolled participants did not complete on-site follow-up; limited long-term outcome data.
8	Talib et al., 2025 [15]	Randomized Controlled Trial (BASIC Trial)	59 adolescents aged 14–16 years with severe obesity (BMI ≥40 kg/m2 or ≥35 kg/m2 with comorbidity)	Laparoscopic Adjustable Gastric Banding (LAGB) + Combined Lifestyle Intervention (CLI)	Combined Lifestyle Intervention (CLI) alone	Anthropometric outcomes (weight, BMI, BMI z-score, waist circumference), hepatic outcomes (hepatorenal index [HRI], liver span, ALAT)	At 1 year, BMI decreased by 5.24 kg/m2 in the LAGB group versus an increase of 0.37 kg/m2 in controls (p < .001). Hepatic steatosis resolution (HRI <1.05) occurred in 21.4% of LAGB patients versus 4.3% of controls. Liver size decreased significantly after surgery.	Small sample size; one-year follow-up only; hepatic steatosis outcomes did not reach statistical significance; some missing ultrasound data; HRI is an indirect measure of steatosis.
9	Altieri et al., 2018 [23]	Retrospective national database cohort	Adolescents with severe obesity undergoing bariatric surgery at accredited centers	Metabolic/bariatric surgery	Procedure-level and/or cohort comparisons as reported	30-day perioperative safety outcomes, including complications, readmission, reoperation, leak, and mortality	Eligible safety outcomes are reported; extract numerical estimates from the full report during formal data extraction.	Administrative database study; short-term follow-up.
10	Mayer Lacrosniere et al., 2023 [24]	Prospective observational cohort study	26 adolescents aged 13–19 years with severe obesity (mean age 17.4 years; mean BMI 52.0 kg/m2)	Roux-en-Y Gastric Bypass (RYGB)	Preoperative baseline vs 2-year postoperative follow-up	Anthropometric outcomes (BMI, weight, %EWL), cardiometabolic outcomes (T2DM, hypertension, insulin resistance, dyslipidemia), sleep outcomes (OSA), safety outcomes (surgical and nutritional complications)	Mean BMI decreased from 52.0 to 34.7 kg/m2 at 24 months (−33.1%); hypertension resolved in 100% (2/2); insulin resistance resolved in 88% (15/17); triglyceride and LDL abnormalities resolved in 100%; OSA improved, with one patient cured; no mortality reported.	Single-center study; small sample size (n = 26); no control group; follow-up limited to 2 years; quality-of-life outcomes not systematically assessed.
11	Sugerman et al., 2003 [25]	Retrospective cohort study (20-year institutional database review)	33 severely obese adolescents aged 12–17.9 years (mean age 16 years; mean BMI 52 ± 11 kg/m2)	Bariatric surgery (primarily Roux-en-Y Gastric Bypass, Long-Limb Gastric Bypass, Distal Gastric Bypass; a few gastroplasty procedures)	Preoperative baseline compared with postoperative follow-up (1–14 years)	Anthropometric outcomes (BMI, weight loss, %EWL), cardiometabolic outcomes (type 2 diabetes, hypertension), sleep outcomes (sleep apnea), obesity-related comorbidities, safety outcomes	Significant and sustained weight loss up to 14 years; mean BMI decreased from 52 to 36 kg/m2 at 1 year and remained significantly lower long term; type 2 diabetes and sleep apnea resolved completely; hypertension improved substantially; marked improvement in self-image and social functioning.	Small sample size; retrospective design; mixed bariatric procedures; incomplete long-term follow-up; quality-of-life outcomes not formally measured; study spans two decades with evolving surgical techniques.
12	Abu-Abeid et al., 2003 [26]	Retrospective cohort study	11 adolescents aged 11–17 years (mean age 15.7 years) with severe obesity (mean BMI 46.6 kg/m2) who failed non-surgical therapy	Laparoscopic Adjustable Gastric Banding (LAGB)	Preoperative baseline vs postoperative follow-up (mean 23 months)	Anthropometric outcomes (BMI), cardiometabolic outcomes (triglycerides, cholesterol), pulmonary outcomes (pulmonary hypertension, CPAP dependence), safety outcomes, psychological well-being	Mean BMI decreased from 46.6 to 32.1 kg/m2; marked improvement in obesity-related comorbidities; normalization of elevated triglycerides; resolution/improvement of pulmonary hypertension and sleep-disordered breathing; no operative or late complications reported.	Small sample size (n = 11); retrospective single-center study; no control group; relatively short follow-up; limited metabolic measurements (no HbA1c, insulin, or glucose tolerance testing).
13	Capella & Capella, 2003 [27]	Retrospective cohort study	19 adolescents aged 13–17 years with severe obesity (mean BMI 49 kg/m2; range 38–67 kg/m2) and at least one adiposity-related comorbidity	Vertical Banded Gastroplasty–Roux-en-Y Gastric Bypass (VBG-RYGBP)	Preoperative baseline vs postoperative follow-up (up to 10 years)	Anthropometric outcomes (BMI, %EWL), cardiometabolic outcomes (hypertension, type 2 diabetes, dyslipidemia), sleep outcomes (sleep apnea), pulmonary outcomes (pulmonary hypertension), safety outcomes and comorbidity resolution	Mean BMI decreased from 49 to 28 kg/m2 at mean 5.5-year follow-up; average %EWL approximately 80% at 3–6 years; all major obesity-related comorbidities resolved; no mortality and minimal morbidity; only one surgical failure (35% EWL).	Small sample size (n = 19); retrospective single-center design; no control group; incomplete long-term follow-up beyond 6 years for some patients; psychosocial outcomes assessed qualitatively rather than with validated instruments.
14	Papadia et al., 2007 [28]	Retrospective cohort study	68 adolescents (52 females, 16 males) aged 14–18 years with severe obesity (mean BMI 46 kg/m2)	Biliopancreatic Diversion (BPD)	Preoperative baseline vs longest postoperative follow-up (mean 11 years)	Anthropometric outcomes (% excess weight loss), cardiometabolic outcomes (hypertension, dyslipidemia, hyperglycemia, type 2 diabetes), safety outcomes (protein malnutrition, reoperations, mortality, pregnancy outcomes)	Mean excess weight loss was 78% at longest follow-up; hypertension decreased from 49% to 9%; all dyslipidemia and diabetes resolved; operative mortality was zero; long-term weight loss maintained for over a decade.	Retrospective design; single-center experience; malabsorptive procedure no longer commonly used in adolescents; high rate of protein malnutrition (16%) and reoperations (20%); absence of a control group.
15	Goldberg et al., 2017 [29]	Retrospective cohort study	Retrospective cohort study	Sleeve Gastrectomy (SG) and Laparoscopic Adjustable Gastric Banding (LAGB)	Comparison between SG and LAGB groups	Safety/nutritional outcomes (iron, ferritin, vitamin B12, folate, hemoglobin, hematocrit), weight loss-adjusted analyses	No patients developed anemia while receiving routine supplementation. SG patients had significantly lower folate levels at 3 and 6 months postoperatively compared with LAGB patients. Folate and B12 deficiencies improved with targeted supplementation. No significant differences in hemoglobin, hematocrit, or iron levels.	Retrospective design; substantial loss to follow-up by 12 months; nutritional outcomes only; no detailed cardiometabolic or sleep outcomes reported.
16	Widhalm et al., 2008 [30]	Retrospective cohort study	10 adolescents (mean age 17.3 years; mean BMI 49.1 kg/m2) with morbid obesity who failed conservative treatment	Bariatric surgery (7 LAGB, 2 sleeve gastrectomy, 1 gastric bypass)	Preoperative baseline vs postoperative follow-up (mean 41 months)	Anthropometric outcomes (BMI, weight loss), cardiometabolic outcomes (hypertension, lipids, CRP, hepatic steatosis), psychological outcomes, safety outcomes	Mean BMI reduction of 10.3 kg/m2 after 41 months; improvements in hypertension, CRP, hepatic steatosis, cholesterol, and triglycerides; gastric bypass produced greater weight loss than gastric banding; several patients required revisional surgery due to inadequate weight loss or regain.	Very small sample size; retrospective design; heterogeneous surgical procedures; poor follow-up compliance; limited quantitative reporting of comorbidity outcomes; substantial psychological morbidity persisted after surgery.
17	Widhalm et al., 2011 [31]	Retrospective cohort study	18 adolescents with severe obesity (mean age 17.7 ± 2.6 years; mean BMI 51.7 ± 8.0 kg/m2)	Bariatric surgery (8 LAGB, 9 primary gastric bypass, 1 sleeve gastrectomy; 4 LAGB patients later converted to gastric bypass)	Preoperative baseline vs long-term postoperative follow-up (mean 42 months)	Anthropometric outcomes (weight, BMI), cardiometabolic outcomes (hypertension, dyslipidemia, impaired glucose tolerance), safety outcomes (iron deficiency, revisional surgery)	Mean body weight decreased from 156.2 kg to 122.7 kg and BMI from 51.8 to 40.5 kg/m2 (p < .0001). Gastric bypass achieved greater sustained weight loss than LAGB. Four of eight LAGB patients required conversion to gastric bypass because of weight regain.	Retrospective design; small sample size; mixed procedures; no control group; limited quantitative reporting of comorbidity outcomes; likely overlap with previous Vienna cohort publications.
18	Talib et al., 2025 [32]	Randomized controlled trial (follow-up analysis of BASIC trial)	Severely obese adolescents aged 14–16 years; cardiac follow-up available in 18 participants (10 intervention, 8 control)	Laparoscopic Adjustable Gastric Banding (LAGB) + multidisciplinary lifestyle intervention	Multidisciplinary lifestyle intervention alone	Anthropometric outcomes (BMI, BMI z-score), cardiac outcomes (left ventricular mass index [LVMI], relative wall thickness [RWT], cardiac geometry)	Significant BMI reduction after surgery. LVMI and RWT improved significantly within the surgical group, suggesting reversal of adverse cardiac remodeling and left ventricular hypertrophy.	Small cardiac imaging sample (n = 18); substantial missing echocardiographic follow-up data; single-center study; limited power to detect between-group differences.
19	Michalsky et al., 2018 [33]	Prospective multicenter cohort study (Teen-LABS)	242 adolescents with severe obesity (mean age 17.1 ± 1.6 years; median BMI 51 kg/m2)	Bariatric surgery: Roux-en-Y gastric bypass (n = 161), sleeve gastrectomy (n = 67), adjustable gastric banding (n = 14)	Preoperative baseline compared with postoperative follow-up (6 months, 1, 2, and 3 years)	Blood pressure, dyslipidemia, diabetes, insulin resistance (HOMA-IR), hs-CRP, BMI, weight loss	Significant and sustained improvements in cardiovascular risk factors, including reductions in elevated blood pressure, dyslipidemia, diabetes, insulin resistance, and systemic inflammation. Greater weight loss predicted greater cardiovascular risk factor remission.	Significant and sustained improvements in cardiovascular risk factors, including reductions in elevated blood pressure, dyslipidemia, diabetes, insulin resistance, and systemic inflammation. Greater weight loss predicted greater cardiovascular risk factor remission.
20	Kelly et al., 2016 [34]	Prospective longitudinal cohort study (two adolescent cohorts; pre–post design)	52 adolescents with severe obesity (mean age 16.5 years) undergoing bariatric surgery	Roux-en-Y Gastric Bypass (RYGB) and Vertical Sleeve Gastrectomy (VSG)	Preoperative baseline vs. postoperative follow-up (3, 6, and 12 months)	Anthropometric: BMI, weight; Cardiometabolic: glucose, insulin, HOMA-IR; Additional inflammatory biomarkers: IL-6, leptin, adiponectin, MCP-1, TNF-α, oxidized LDL (oxLDL)	Significant reductions in BMI, insulin resistance, IL-6, leptin, and oxLDL; significant increase in adiponectin after bariatric surgery. Findings suggest improvement in cardiometabolic risk profile among adolescents with severe obesity.	No control group; relatively small sample size; possible insufficient power for some biomarkers; limited biomarker panel; unable to account for diet and physical activity influences.
21	Messiah et al., 2013a [35]	Prospective multicenter longitudinal database study (BOLD)	890 adolescents aged 11–19 years with morbid obesity undergoing bariatric surgery	Gastric Bypass (n = 454) and Adjustable Gastric Banding (n = 436)	Comparison between bariatric procedures and preoperative baseline values	Weight, BMI, hypertension, diabetes, hyperlipidemia, GERD, obstructive sleep apnea, asthma, depression, menstrual irregularities, complications, readmissions, reoperations	Gastric bypass produced significantly greater weight loss than gastric banding (48.6 kg vs 19.8 kg at 12 months). Marked improvements occurred in hypertension, diabetes, hyperlipidemia, GERD, obstructive sleep apnea, asthma, and other obesity-related comorbidities.	High loss to follow-up at 12 months, reliance on registry data, possible underreporting of complications, limited co-morbidity severity scoring system, lack of dietary/physical activity data.
22	Frederick et al., 2025 [36]	Retrospective cohort analysis of prospectively collected MBSAQIP registry data from 2020 to 2022	Adolescent subgroup: 2829 adolescents aged 14–19 years undergoing primary metabolic and bariatric surgery; 793 Black and 2036 White adolescents. Adult participants were also studied, but adolescent data are reported separately and are extractable. Median preoperative BMI was 48.6 kg/m2 in Black adolescents and 45.4 kg/m2 in White adolescents.	Primary laparoscopic or robotic metabolic and bariatric surgery: sleeve gastrectomy, gastric bypass, or duodenal switch. Among adolescents, 87.7% underwent sleeve gastrectomy and approximately 12.2% underwent gastric bypass.	Comparison between Black and White adolescent bariatric surgery patients; adult groups were additionally analyzed separately.	Anthropometric: baseline BMI and 30-day change in BMI. Cardiometabolic: diabetes, HbA1c, hypertension, and hyperlipidemia. Sleep-related: obstructive sleep apnea. Safety: pulmonary embolism, VTE/DVT, surgical-site infection, anastomotic leak, transfusion, sepsis, bowel obstruction, stroke, reoperation, readmission, intervention, ED visits, and 30-day follow-up.	Black adolescents had a higher preoperative median BMI than White adolescents (48.6 vs 45.4 kg/m2). They were more likely to receive no VTE prophylaxis (7.1% vs 3.5%) and had a higher incidence of postoperative pulmonary embolism (0.3% vs 0%). Anastomotic/deep-space complications were also more frequent in Black adolescents, while most other short-term outcomes were broadly similar. At 30 days, Black adolescents showed a slightly greater BMI decrease than White adolescents (3.02 vs 2.8 kg/m2).	Retrospective database analysis; only 30-day postoperative outcomes available; no assessment of sustained weight loss or long-term comorbidity remission; limited preoperative risk-factor data; inability to assess institutional variations, referral patterns, or causal relationships; analysis restricted to Black and White participants; potential race misclassification because race may not always be self-reported.
23	Inge et al., 2018 [37]	Retrospective multicenter comparative cohort study using EHR data (PCORnet)	544 adolescents aged 12–19 years with severe obesity undergoing bariatric surgery	Sleeve Gastrectomy (SG), Roux-en-Y Gastric Bypass (RYGB), Adjustable Gastric Banding (AGB)	Head-to-head comparison of SG, RYGB, and AGB	Percent BMI change at 1, 3, and 5 years; proportion achieving BMI-loss thresholds; perioperative adverse events	RYGB and SG produced substantial and durable BMI reduction; AGB was markedly less effective. One-year BMI change: RYGB −31%, SG −28%, AGB −10%. Three-year BMI change: RYGB −29%, SG −25%. No perioperative deaths occurred.	Non-randomized design; substantial loss to follow-up at later time points; limited long-term AGB sample size; no detailed comorbidity outcome assessment.
24	Blumenfeld et al., 2020 [38]	Retrospective national registry-based comparative cohort study	259 adolescents aged 12–19 years with severe obesity	Laparoscopic Sleeve Gastrectomy (LSG)	Laparoscopic Adjustable Gastric Banding (LAGB)	BMI change, total weight loss, percentage total weight loss, postoperative complications, hospitalizations, interventions, mortality	LSG achieved significantly greater BMI reduction (15.4 vs 10.3 kg/m2), greater total weight loss (43.6 vs 28.8 kg), and higher likelihood of achieving BMI <30 kg/m2 at 1 year. No mortality was reported.	Retrospective design; non-randomized procedure selection; substantial exclusion due to missing follow-up data; limited follow-up duration.
25	Zitsman et al., 2011 [39]	Retrospective matched cohort study	115 adolescents (14–19 years) with severe obesity undergoing LAGB	Laparoscopic Adjustable Gastric Banding (LAGB)	BMI- and sex-matched adult LAGB cohort (n = 115)	Weight loss, BMI change, excess weight loss (%EWL), hypertension, dyslipidemia, metabolic syndrome, GERD, menstrual irregularities/PCOS, asthma, complications	Adolescents achieved weight loss comparable to adults. Significant improvements were observed in hypertension, dyslipidemia, metabolic syndrome, menstrual irregularities, and asthma. No deaths occurred	Retrospective design; substantial loss to follow-up; single-center study; limited long-term follow-up data; adult comparison group not ideal for pooled pediatric analyses.
26	Hervieux et al., 2016 [40]	Prospective bicentric comparative cohort study	36 adolescents (mean age 16.7 ± 1.3 years) with severe obesity	Laparoscopic Adjustable Gastric Banding (LAGB)	53 young adults (18–25 years) undergoing LAGB	BMI, excess weight loss (EWL), insulin resistance (HOMA), glycemia, HbA1c, lipid profile, quality of life, complications, reoperations	Adolescents achieved significantly greater EWL than young adults at 12 and 24 months. Insulin resistance resolved by 24 months, dyslipidemia improved, and no major postoperative morbidity occurred beyond band-related complications.	Small sample size; two-center design; short-term follow-up (2 years); no non-surgical control group; long-term band-related complications not assessed.
27	Zhu et al., 2023 [41]	Multicenter retrospective non-randomized comparative cohort study	202 adolescents (13–19 years) with severe obesity and type 2 diabetes	Bariatric surgery (LSG and LRYGB)	Standard medical treatment	BMI, total weight loss (TWL%), fasting blood glucose (FBG), HbA1c, HOMA-IR, diabetes remission, medication use, complications	Bariatric surgery produced significantly greater weight loss, better glycemic control, and markedly higher diabetes remission rates than medical therapy. Diabetes remission reached 76.1% after surgery versus 6.5% with medical treatment.	Retrospective design; non-randomized treatment allocation; relatively small sample size; follow-up limited to 2 years.
28	Inge et al., 2018 [42]	Secondary comparative analysis of two prospectively studied multicenter cohorts: Teen-LABS and TODAY	93 adolescents with severe obesity and type 2 diabetes: 30 surgically treated Teen-LABS participants (mean age 16.9 years; mean baseline BMI 54.4 kg/m2) and 63 medically treated TODAY participants (mean age 15.3 years; mean baseline BMI 40.5 kg/m2)	Bariatric surgery in Teen-LABS: predominantly Roux-en-Y gastric bypass and sleeve gastrectomy; Table 1 reports 23 gastric bypass, 6 sleeve gastrectomy, and 1 adjustable gastric banding procedure	Medical therapy in TODAY: metformin alone or combined with rosiglitazone or intensive lifestyle intervention; insulin added for disease progression	Anthropometric: weight, BMI, percentage BMI change, waist circumference. Glycemic: HbA1c, fasting glucose, fasting insulin, diabetes remission/glycemic control. Cardiometabolic: elevated blood pressure, systolic and diastolic BP, dyslipidemia, total cholesterol, LDL-C, HDL-C, triglycerides. Renal: eGFR and urinary albumin-creatinine ratio. Safety: subsequent operations, readmissions, hospitalizations, and clinical adverse events	At 2 years, the surgical group had a 29.0% BMI reduction, whereas BMI increased by 3.7% in the medical group. HbA1c fell from 6.8% to 5.5% after surgery but increased from 6.4% to 7.8% with medical therapy. At 2 years, 94% of surgical participants versus 38% of medically treated participants had HbA1c <6.5%. Surgery was also associated with improvements in elevated blood pressure, dyslipidemia, and abnormal kidney-function measures. However, 7 of 30 surgical participants (23%) experienced complications requiring a subsequent operation and/or readmission related or possibly related to bariatric surgery.	Nonrandomized comparison between two independently conducted cohorts with different original study objectives; substantial baseline differences, including higher BMI and cardiovascular risk burden in the surgical group; small surgical sample (n = 30); few sleeve-gastrectomy patients; missing follow-up visits; only 2 years of follow-up; potential residual confounding despite statistical adjustment
29	Aryannezhad et al., 2021 [43]	Prospective matched case–control study	118 adolescents (11–18 years) with severe obesity	Bariatric surgery (Sleeve Gastrectomy or One-Anastomosis Gastric Bypass)	236 matched young adults (19–29 years)	BMI, weight, waist circumference, excess weight loss (EWL%), body composition, cardiometabolic risk factors, comorbidity resolution, micronutrient deficiencies, complications	Bariatric surgery produced substantial weight loss and improvement in cardiometabolic risk factors in adolescents, with outcomes comparable to young adults at 1 year. Safety outcomes were also similar between groups.	One-year follow-up only; no non-surgical control group; high missing rates for micronutrient data; lack of psychological outcome assessment.
30	Elhag & El Ansari, 2021 [44]	Retrospective longitudinal cohort study	146 adolescents aged ≤18 years with severe obesity	Laparoscopic Sleeve Gastrectomy (LSG)	Preoperative baseline vs postoperative follow-up	BMI, weight, EWL%, TWL%, blood pressure, lipid profile, fasting glucose, HbA1c, liver enzymes, uric acid, prediabetes remission, T2DM remission, hypertension remission	Significant and durable weight loss maintained up to 9 years. BMI decreased from 45.6 to 30.2 kg/m2 at 9 years. High remission rates observed for dyslipidemia, prediabetes, T2DM, elevated liver enzymes, and hypertension.	Retrospective design; decreasing sample size at longer follow-up; no control group; quality-of-life outcomes unavailable; year-9 cardiometabolic analyses limited by small numbers.
31	Teeple et al., 2012 [45]	Retrospective cohort study	15 morbidly obese adolescents undergoing bariatric surgery	Roux-en-Y Gastric Bypass (RYGB)	Preoperative baseline vs postoperative follow-up	BMI, weight, %EWL, HbA1c, insulin, C-peptide, HOMA-IR, insulin sensitivity, lipid profile, blood pressure, complications	Significant weight loss and marked improvements in insulin resistance, HbA1c, insulin sensitivity, and lipid profile within 2 years after surgery.	Small sample size (n = 15); retrospective design; loss to follow-up; short follow-up duration (2 years); single-center study.
32	Nadler et al., 2012 [46]	Prospective cohort study (retrospective review of prospective registry)	23 adolescents aged 14–19 years with morbid obesity	Laparoscopic Sleeve Gastrectomy (LSG)	Preoperative baseline vs postoperative follow-up	BMI, % excess weight loss (%EWL), obesity-related comorbidities, complications, reoperations, length of stay	LSG resulted in approximately 40% EWL at 1 year with substantial improvement or resolution of obesity-related comorbidities and very low perioperative morbidity.	Small sample size; short follow-up (mean 10.9 months); single-center study; incomplete long-term follow-up for some comorbidity assessments.
33	Chu et al., 2019 [47]	Retrospective cohort study	28 adolescents (13–19 years) with severe obesity undergoing bariatric surgery in Canada	Bariatric surgery (Roux-en-Y gastric bypass or sleeve gastrectomy) Preoperative baseline compared with p	Preoperative baseline compared with postoperative follow-up at 3, 6, 12, and 24 months	Weight, BMI, % excess weight loss (%EWL), remission of obesity-related comorbidities including obstructive sleep apnea, hypertension, dyslipidemia, impaired glucose tolerance, T2DM, NAFLD, and PCOS	Mean %EWL was 33.6% at 3 months, 55.0% at 12 months, and 55.1% at 24 months. Early weight loss at 3 months significantly predicted weight loss at 12 and 24 months. A threshold of ≥30% EWL at 3 months predicted successful long-term weight loss (≥50% EWL). Remission of several cardiometabolic and sleep-related comorbidities was observed after surgery.	Small sample size (n = 28); retrospective design; no non-surgical control group; predominantly Caucasian cohort; findings not generalizable to adjustable gastric banding; limited power for analysis of comorbidity outcomes.
34	de la Cruz-Muñoz et al., 2012 [48]	Retrospective observational pre–post cohort study	71 adolescents aged 16–19 years with severe obesity; 80% Hispanic, 77% female; mean baseline BMI 46.1 kg/m2	Bariatric surgery: Roux-en-Y gastric bypass (92%) or adjustable gastric banding (8%)	No non-surgical comparator; preoperative measurements served as baseline control. Secondary comparison between gastric bypass and adjustable gastric banding groups.	Weight, BMI, BMI z-score, BMI percentile, weight z-score, weight percentile, estimated weight loss, perioperative and postoperative complications	Significant reductions in all anthropometric measures at 1 year (P < .001). Mean BMI decreased from 46.1 to 35.4 kg/m2 and mean weight decreased from 128.4 to 99.1 kg. Gastric bypass produced significantly greater weight loss than adjustable gastric banding (BMI reduction 11.3 vs 3.9 kg/m2). Weight loss outcomes were similar between Hispanic and non-Hispanic adolescents. No perioperative complications occurred; only three cases of iron-deficiency anemia due to noncompliance with supplements.	Retrospective design; single-center study; follow-up limited to approximately 1 year; no non-surgical control group; primarily anthropometric outcomes reported; limited cardiometabolic outcome data; small adjustable gastric band subgroup (n = 6).
35	Mizrahi Reuveni et al., 2026 [49]	Retrospective cohort study with 5-year follow-up	278 adolescents (<18 years) with severe obesity; 152 bariatric surgery, 126 nutritional intervention	Laparoscopic Sleeve Gastrectomy (LSG)	Nutritional intervention (≥3 dietitian visits)	BMI, weight, BMI SDS, weight SDS, obesity class transition, hemoglobin, vitamin B12, vitamin D, folic acid, TSH, revision surgery, nutritional deficiency outcomes	Bariatric surgery produced sustained weight loss (BMI 44 → 32 kg/m2 over 5 years) compared with minimal change in the nutritional intervention group (45.5 → 44 kg/m2). Significant improvements in obesity class were observed, but surgery was associated with lower hemoglobin and vitamin B12 levels, emphasizing the need for long-term nutritional monitoring.	Retrospective design, non-matched groups, missing long-term data, no sleep or pulmonary outcomes, no detailed cardiometabolic markers (HbA1c/lipids), potential selection bias.
36	Messiah et al., 2013b [50]	Prospective longitudinal database cohort study	827 adolescents (11–19 years) with severe obesity	Gastric bypass surgery or adjustable gastric banding	Ethnic groups (Non-Hispanic White, Non-Hispanic Black, Hispanic) and baseline values	Weight, BMI, BMI percentile, estimated weight loss (EWL), postoperative complications, readmissions	Bariatric surgery produced substantial weight loss (>30 kg on average) and significant BMI reduction at 1 year, with no significant differences in outcomes by ethnicity.	High loss to follow-up at 12 months (70%); limited outcome measures beyond anthropometry; procedures analyzed together; observational design.
37	Silva et al., 2012 [51]	Observational pre–post cohort study (single-center)	14 extremely obese adolescents (10 females, 4 males); age 13.5–17.5 years; mean BMI 46.1 ± 3.15 kg/m2	Laparoscopic Adjustable Gastric Banding (LAGB)	No external comparator; preoperative values served as control	BMI, weight, percentage excess weight loss (%EWL), systolic and diastolic blood pressure, lipid profile (HDL, LDL, TG, total cholesterol), insulin resistance (HOMA-IR), postoperative complications, mortality	Significant and sustained weight loss over 3 years (BMI reduced from 46.1 to 33.7 kg/m2); %EWL reached 48.1% at 3 years; hypertension resolved in all affected patients within 1 year; marked improvement in insulin resistance with normalization of HOMA-IR by 3 years; HDL increased significantly; no mortality and few manageable complications.	Small sample size (n = 14); single-center study; non-randomized design; no control group; predominantly cardiometabolic outcomes with no sleep or pulmonary outcomes; limited generalizability; some loss of follow-up by 3 years (12/14 patients available).
38	Ryder et al., 2021 [52]	Secondary analysis of two multicenter longitudinal cohorts (Teen-LABS and TODAY); comparative observational study with 5-year follow-up	93 adolescents with severe obesity and type 2 diabetes; Teen-LABS: n = 30 (mean age 16.9 years, BMI 54.4 kg/m2); TODAY: n = 63 (mean age 15.4 years, BMI 40.5 kg/m2)	Metabolic Bariatric Surgery (MBS): Roux-en-Y gastric bypass (n = 24) or Vertical Sleeve Gastrectomy (n = 6)	Medical therapy from the TODAY trial (metformin alone, metformin + rosiglitazone, or intensive lifestyle intervention; insulin added for glycemic progression)	Estimated 30-year cardiovascular disease (CVD) event risk (Framingham score), BMI, blood pressure, lipid profile, HbA1c, diabetes remission, antihypertensive medication use, insulin sensitivity, kidney disease markers	Despite higher baseline CVD risk, adolescents undergoing bariatric surgery experienced marked reductions in estimated 30-year CVD event risk at 1 year and sustained through 5 years. BMI decreased from 54.4 to 42.9 kg/m2, HDL increased, blood pressure improved, antihypertensive use declined, and 85.4% achieved diabetes remission by 5 years. In contrast, medically treated adolescents showed worsening cardiometabolic risk and increasing predicted CVD risk over time.	Not a randomized comparison between surgery and medical therapy; secondary analysis; small surgical cohort (n = 30); cardiovascular outcomes were estimated risk scores rather than actual events; baseline differences between cohorts; no pulmonary or sleep outcomes reported; potential residual confounding.
39	Kalra et al., 2008 [53]	Retrospective pre–post observational study	19 adolescents with severe obesity undergoing bariatric surgery; mean age 16.5 ± 0.35 years; mean BMI 60.3 ± 2.11 kg/m2; 66% female; 74% had obstructive sleep apnea (OSA) at baseline	Laparoscopic Roux-en-Y Gastric Bypass (RYGB) with assessment before and after surgical weight loss	No external comparator; preoperative polysomnography served as baseline control	Sleep architecture (sleep efficiency, sleep stages, arousal index, REM latency, REM cycles), obstructive sleep apnea (AHI), respiratory disturbance index, total sleep time, oxygen saturation-related sleep outcomes	Mean weight loss was 66.4 kg within approximately 1 year. Sleep efficiency improved significantly (73.1%–80.2%, p = .01), stage 1 sleep decreased (6.0%–3.0%, p = .02), arousal index improved (11.3–7.6 events/h, p = .01), and AHI decreased markedly from 20.5 to 6.1 events/h (p = .01), demonstrating substantial improvement in sleep fragmentation and OSA severity after surgery.	Retrospective design; small sample size (n = 19); single-center study; no non-surgical control group; short follow-up (∼0.9 years); limited ability to assess long-term sleep outcomes or perform multivariable analyses.
40	Torres et al., 2024 [54]	Retrospective statewide cohort	2241 adolescents aged 12–19 years undergoing metabolic/bariatric surgery	LAGB, RYGB, or sleeve gastrectomy	Comparison across bariatric procedure types; nonsurgical comparator not required	In-hospital complications, 30-day readmission, length of stay, revision/conversion, and durability	Reports eligible short- and long-term safety/durability outcomes.	Retrospective administrative database; residual confounding.
41	Phan et al., 2013 [55]	Prospective cohort study	51 morbidly obese adolescents aged 14–18 years; mean age 16.1 years; mean BMI 50.4 kg/m2; 86.3% female	Laparoscopic Adjustable Gastric Banding (LAGB) following a 2-week low-calorie meal replacement diet (LCMRD)	No external comparator; analysis examined association between preoperative LCMRD weight loss and postoperative weight loss after LAGB	Weight loss during LCMRD, weight loss at 3, 6, and 12 months after surgery, percent total weight loss (%TWL)	Mean weight loss during the 2-week LCMRD was 5.7 kg, and mean postoperative weight loss at 1 year was 17.4 kg. However, weight loss during the preoperative LCMRD was not significantly associated with weight loss at 3, 6, or 12 months after LAGB. The study concluded that short-term preoperative dietary weight loss is not a useful predictor of postoperative surgical weight loss in adolescents.	Small sample size (n = 51); single-center study; predominantly female and Caucasian population limiting generalizability; no control group; follow-up limited to 1 year; focused only on weight-loss outcomes without cardiometabolic, sleep, pulmonary, or safety outcomes.
42	Elhag et al., 2018 [56]	Retrospective cohort study	79 adolescents (from 102 eligible patients with complete follow-up data), aged 13–17 years; mean age 15.99 ± 1.07 years; BMI ≥40 kg/m2 or ≥35 kg/m2 with comorbidities	Laparoscopic Sleeve Gastrectomy (LSG)	No external comparator; preoperative values compared with postoperative outcomes at mean follow-up of 24.2 months	Anthropometric outcomes (weight, BMI, excess weight loss); nutritional parameters (vitamin D, vitamin B12, ferritin, albumin, iron, trace elements); cardiometabolic outcomes (TC, LDL, HDL, TG, ALT, AST, fasting blood glucose, HbA1c, T2DM, prediabetes, uric acid, PTH)	Significant weight loss was achieved (weight reduction 51.8 kg; BMI reduction 17.0 kg/m2; EWL 81.1%). Cardiometabolic risk factors improved substantially: LDL decreased, HDL increased, triglycerides decreased, HbA1c improved, and liver enzymes normalized. Prediabetes remission was 100%, and type 2 diabetes remission was 80%. Nutritional deficiencies persisted postoperatively, particularly vitamin D deficiency (89.3%), hypoalbuminemia (38%), ferritin deficiency (33.3%), and anemia among females. No significant trace element deficiencies were observed.	Retrospective single-center design; no control group; follow-up available for only 79/102 patients (77.5%); lack of dietary adherence and supplement compliance data; no sleep or pulmonary outcomes assessed; longer-term (>2 years) outcomes unavailable.
43	Derderian et al., 2020 [57]	Prospective multicenter observational cohort study (Teen-LABS) with 5-year follow-up	217 adolescents aged 13–19 years with severe obesity undergoing bariatric surgery; mean age 16.7 years; mean BMI 52.4 kg/m2	Bariatric surgery: Roux-en-Y Gastric Bypass (65.9%), Sleeve Gastrectomy (27.6%), Adjustable Gastric Banding (6.5%)	No external comparator; longitudinal pre–post assessment of excess skin outcomes over 5 years	Abdominal pannus severity (grades 0–5), excess skin symptoms, intertriginous infections, cellulitis, ulcerations, need for body contouring surgery, postoperative skin-related complications	At surgery, 90.3% of adolescents had an abdominal pannus and 7.4% had pannus-related symptoms. During follow-up, 22.1% experienced excess-skin-related symptoms. Higher preoperative pannus severity predicted greater postoperative pannus reduction but also a higher risk of postoperative symptoms. Most changes in pannus severity occurred within the first 24 months and then stabilized through 5 years. Intertriginous infections were the most common complication.	Study focused on excess skin outcomes rather than cardiometabolic, pulmonary, or sleep outcomes; symptom reporting was self-reported; relatively few participants experienced symptoms, limiting power for subgroup analyses; limited ethnic diversity; body contouring surgery recipients were excluded from analysis.
44	Levaillant et al., 2023 [58]	Prospective observational cohort study	42 adolescents with severe obesity undergoing LAGB; mean age 16.0 ± 1.1 years; 83% female; mean BMI 41.8 ± 4.6 kg/m2	Laparoscopic Adjustable Gastric Banding (LAGB)	No external comparator; participants classified according to successful versus unsuccessful weight loss at 24 months	Weight, BMI, % excess weight loss (%EWL), blood pressure, insulin resistance, NAFLD, glucose metabolism, obstructive sleep apnea, bone mineral density, vitamin/mineral deficiencies, surgical complications, psychosocial and behavioral predictors of weight loss success	Mean %EWL was 34.1% ± 38.9% at 24 months with a mean BMI reduction of 5.6 kg/m2. Insulin resistance decreased from 40% at baseline to 17% at 12 months and 15% at 24 months. NAFLD prevalence decreased from 64% to 21% at 24 months. Most obesity-related comorbidities improved and no major complications occurred. Preoperative weight loss before surgery was associated with successful postoperative outcomes (OR 1.16; p = .007), whereas higher BMI at surgery predicted failure (OR 0.83; p = .033). No other psychosocial or clinical factors significantly predicted success.	Small sample size (n = 42); single-center study; substantial loss to follow-up by 24 months; no control group; only LAGB evaluated; self-reported behavioral measures may introduce bias; limited long-term follow-up beyond 3 years.
45	Henfridsson et al., 2019 [59]	Prospective multicenter longitudinal cohort study (AMOS study) with matched non-surgical controls at 5 years	85 adolescents with severe obesity (67% female); mean age 16.0 ± 1.2 years; mean BMI 45.5 ± 6.1 kg/m2; age 14–19 years	Laparoscopic Roux-en-Y Gastric Bypass (LRYGB)	Matched adolescents with severe obesity receiving non-surgical treatment from the Swedish Childhood Obesity Treatment Register (BORIS)	Weight, BMI, % total body weight loss, % excess BMI loss, body composition (fat mass, fat-free mass, muscle mass), energy intake, dietary energy density, protein intake, dietary composition, alcohol intake	Significant and sustained weight loss was observed after LRYGB. Weight decreased by 31%, 33%, and 28% at 1, 2, and 5 years, respectively, while controls gained 13% over 5 years. BMI decreased from 45.5 to 32.3 kg/m2 at 5 years. Fat mass decreased by 41%, with substantial reductions in body fat. Energy intake remained 10% lower than baseline at 5 years and was significantly lower than controls. Adequate protein intake was associated with better preservation of muscle mass. Boys preserved muscle mass better than girls.	Study focused primarily on dietary intake and body composition rather than detailed cardiometabolic, pulmonary, or sleep outcomes; non-randomized design; control group differed somewhat at baseline; dietary intake based on self-report; no detailed data on HbA1c, lipids, hypertension, lung function, or OSA outcomes.
46	Khidir et al., 2018 [60]	Retrospective comparative cohort study with 5-year follow-up	91 adolescents (14–19 years; mean age 17 ± 1.5 years) with morbid obesity undergoing LSG; compared with 139 adults. Mean adolescent baseline BMI 47.6 ± 7.5 kg/m2.	Laparoscopic Sleeve Gastrectomy (LSG)	Adult patients undergoing LSG (n = 139) followed over the same period; no non-surgical comparator.	BMI, % excess weight loss (%EWL), % total weight loss (%TWL), % BMI loss (%BMIL), diabetes remission, prediabetes remission, HbA1c, postoperative complications, hospital stay, mortality.	Adolescents achieved superior long-term weight-loss outcomes compared with adults. At 5 years, adolescents achieved %TWL of 35.8% versus 26.3% in adults and %EWL of 78% versus 55% in adults. BMI decreased from 47.6 to 30.3 kg/m2. All prediabetic adolescents normalized HbA1c levels. Among diabetic adolescents, 62.5% achieved complete remission and the remainder improved; importantly, no adolescent experienced diabetes relapse at 5 years, whereas 13% of adults relapsed. Postoperative complications were uncommon (4.4%), and no mortality occurred.	Retrospective design; single-country cohort (Qatar); no non-surgical control group; assessment focused mainly on weight and diabetes outcomes; limited evaluation of other cardiometabolic comorbidities, sleep apnea, pulmonary outcomes, and quality-of-life measures; follow-up attrition present.
47	Guan et al., 2023 [61]	Prospective, nonrandomized matched comparative cohort study with 5-year follow-up	87 Chinese adolescents aged 13–18 years with severe obesity; 44 underwent MBS and 43 were matched nonsurgical controls. All participants met MBS eligibility criteria of BMI >40 kg/m2 or >35 kg/m2 with related metabolic complications. Mean baseline age was 16.5 ± 1.6 years in the surgical group and 15.9 ± 2.3 years in controls.	Metabolic bariatric surgery (n = 44): laparoscopic Roux-en-Y gastric bypass (LRYGB, n = 15) or laparoscopic sleeve gastrectomy (LSG, n = 29)	Matched nonsurgical control group (n = 43) receiving lifestyle modification consisting of calorie restriction, aerobic physical activity (150 min/week), behavioral therapy, counseling sessions, and follow-up	Anthropometric: weight, BMI, BMI change, total weight loss (TWL), height and height change. Metabolic: hypertension, type 2 diabetes mellitus, dyslipidemia and remission rates. Nutritional/safety: anemia, folate deficiency, vitamin B12 deficiency, vitamin D deficiency, early complications, cholecystitis, and revision surgery. Psychosocial/QOL: SF-36 physical and mental health domains, education and employment status.	At 5 years, surgery resulted in significantly lower BMI (28.7 vs 38.1 kg/m2), greater BMI reduction (−13.0 vs −1.2 kg/m2), and greater TWL (30.5% vs −5.6%) compared with controls. Surgical patients achieved remission rates of 54.5% for hypertension, 87.5% for T2DM, and 93.8% for dyslipidemia. Physical and mental QOL outcomes were generally better after surgery. However, surgery was associated with increased nutritional deficiencies, including anemia (20.5%), folate deficiency (13.6%), vitamin B12 deficiency (25.0%), and vitamin D deficiency (70.5%) at 5 years. No 30-day postoperative complications occurred; one patient developed acute cholecystitis and three underwent revision surgery for weight regain.	Small sample size (n = 87); single-center study; nonrandomized treatment allocation with potential selection bias and residual confounding; control-group follow-up was mainly obtained by telephone, limiting evaluation of obesity-related comorbidities; limited generalizability beyond Chinese/East Asian adolescents; some baseline imbalance in absolute body weight; possible concerns regarding nutritional adherence and measurement consistency.
48	de la Cruz-Muñoz et al., 2010 [62]	Retrospective cohort study	78 adolescents aged 16–19 years with morbid obesity; 76.9% Hispanic, 19.2% non-Hispanic White, 76.9% female; mean baseline BMI 45.8 kg/m2	Bariatric surgery: predominantly Laparoscopic Roux-en-Y Gastric Bypass (91.0%), with Adjustable Gastric Banding (7.7%) and Sleeve Gastrectomy (1.3%)	No non-surgical comparator; longitudinal comparison of postoperative outcomes with baseline values over 4 years	Weight change, BMI change, long-term weight-loss maintenance, postoperative complications, nutritional monitoring (iron deficiency anemia, vitamin and mineral surveillance)	Significant and sustained weight loss was observed through 4 years. Non-Hispanic White adolescents lost 113.8 lb and 18.6 BMI units, while Hispanic adolescents lost 99.4 lb and 16.2 BMI units over 4 years. Most weight loss occurred during the first year, followed by a plateau at 12–18 months with maintenance through year 4. Only two minor postoperative complications occurred. Three patients developed iron-deficiency anemia related to noncompliance with supplementation. Overall, bariatric surgery was considered safe and effective across ethnic groups.	Retrospective design; substantial missing follow-up data over time; outcomes focused mainly on weight and BMI; limited reporting of cardiometabolic comorbidities, sleep outcomes, pulmonary outcomes, and quality-of-life measures; predominantly Hispanic cohort may limit generalizability; very small numbers for gastric banding and sleeve gastrectomy subgroups.
49	Strauss et al., 2001 [63]	Retrospective case series/cohort study	10 adolescents aged 15–17 years with morbid obesity; mean preoperative weight 148 ± 37 kg; all were >100% above ideal body weight and had failed multiple weight-loss attempts. Seven had severe obesity-related comorbidities including sleep apnea and hypertension.	Roux-en-Y Gastric Bypass (RYGB)	No comparator; preoperative values compared with postoperative outcomes.	Weight loss, BMI change, sleep apnea, hypertension, adiposity-related comorbidity resolution, postoperative complications, nutritional deficiencies (iron, folate, vitamin D), pregnancy outcomes, reoperations.	Mean weight loss was 53.6 ± 25.6 kg, representing approximately 62% of excess weight. Nine of ten adolescents lost >30 kg. Severe sleep apnea resolved in affected patients, and hypertension normalized. Obesity-related comorbidities resolved in all adolescents. Nutritional complications included iron-deficiency anemia (5 patients) and transient folate deficiency (3 patients). Four adolescents required later operative intervention for complications such as cholelithiasis, hernia, or bowel obstruction. No early postoperative complications or mortality occurred.	Very small sample size (n = 10); retrospective single-center study; no control group; long study period with heterogeneous follow-up durations; limited quantitative cardiometabolic data (lipids, HbA1c, insulin resistance not reported); potential attrition and reliance on self-reported long-term weights in some patients.
50	Shehata et al., 2025 [64]	Retrospective cohort study with 5-year follow-up	264 adolescents with morbid obesity (mean age 15.0 ± 3.0 years; 74% female) aged 10–19 years who underwent LSG; mean baseline BMI 48.4 ± 6.95 kg/m2.	Laparoscopic Sleeve Gastrectomy (LSG); subgroup with clinically significant weight regain (n = 62) received adjunct semaglutide beginning postoperative year 3.	No non-surgical control group. Comparisons were made longitudinally over 5 years and between weight-regain and non-weight-regain subgroups.	Weight, BMI, %TWL, %EWL, HbA1c, GLP-1, GIP, type 2 diabetes, hypertension, dyslipidemia, gastric volume, weight regain, semaglutide response.	Mean weight decreased from 132.8 kg to 90.1 kg and BMI from 48.4 to 32.9 kg/m2 at 5 years. %EWL peaked at 68.5% at year 2 and remained 63.3% at year 5. HbA1c improved from 7.12% to 5.36%. Type 2 diabetes prevalence decreased from 36.4% to 5.7%, hypertension from 48.9% to 6.8%, and dyslipidemia from 43.2% to 4.6%. Clinically significant weight regain occurred in 23.5% of patients. Semaglutide improved %EWL in weight-regain patients from 34.1% to 68.0% after treatment. Early postoperative GLP-1 and GIP surges attenuated over time, particularly in patients experiencing weight regain.	Retrospective single-center design; no randomized comparator; semaglutide subgroup not randomized; no pulmonary or sleep outcomes reported; hormonal associations with weight regain were exploratory and descriptive rather than based on formal correlation analyses; potential selection bias and limited generalizability.
51	Howard et al., 2022 [65]	Retrospective national claims cohort	1132 adolescents aged ≤19 years with severe obesity insured by Medicaid	Sleeve gastrectomy or Roux-en-Y gastric bypass	Sleeve gastrectomy versus Roux-en-Y gastric bypass	Mortality, complications, hospitalization, emergency department use, reoperation, and revision up to 5 years	Reports protocol-eligible adverse-event outcomes and procedure comparisons.	Claims-based outcome ascertainment; mean age approximately 18.6–18.7 years.
52	Beamish et al., 2026 [66]	Prospective multicentre observational cohort (Teen-LABS)	228 adolescents (mean age 17 years; mean BMI 53 kg/m2) with severe obesity	Roux-en-Y Gastric Bypass (RYGB) and Vertical Sleeve Gastrectomy (VSG)	Procedure comparison and 5-year follow-up	Healthcare utilisation events, hospital admissions, reoperations, complications, mortality	68% experienced ≥1 healthcare utilisation event over 5 years; 36% had surgery-related events; 39% required hospital admission. Early postoperative complications predicted increased long-term healthcare utilisation. No major difference between RYGB and VSG for overall safety events.	No non-surgical control group; observational design; some events not independently adjudicated; potential under-capture of outside-hospital events.
53	Nunez Lopez et al., 2017 [67]	Retrospective cohort study (BOLD database, 2007–2014)	1539 adolescents aged 12–19 years with severe obesity undergoing bariatric surgery in the United States	Gastric bypass, sleeve gastrectomy, adjustable gastric banding, other bariatric procedures	Preoperative versus 12-month postoperative outcomes; subgroup analyses by race, sex, and procedure type	BMI change, net weight loss, % total weight loss, remission of hypertension, obstructive sleep apnea, diabetes, dyslipidemia, GERD, asthma, depression	Significant weight loss observed after surgery. OSA, hypertension, diabetes, and other obesity-related comorbidities improved substantially. Race and sex did not significantly affect 12-month weight loss or remission rates. Minority adolescents underwent bariatric surgery at disproportionately lower rates despite higher severe obesity prevalence.	Retrospective design; socioeconomic variables unavailable; follow-up attrition; database limited to accredited bariatric centers; long-term outcomes beyond 12 months incompletely captured.
54	Hoeltzel et al., 2022 [68]	Retrospective cohort analysis of prospectively collected MBSAQIP database data from 2015 to 2018; Level III evidence	5068 adolescents aged 10–19 years who underwent bariatric surgery. Mean BMI was 47.3 kg/m2; 78.5% were female. Age distribution: 10–14 years, n = 78 (1.5%); 15–17 years, n = 939 (18.5%); and 18–19 years, n = 4051 (79.9%). The most prevalent comorbidities included hypertension (26.7%), obstructive sleep apnea (16.1%), GERD (12.4%), and diabetes (12.1%).	Minimally invasive bariatric surgery: sleeve gastrectomy (SG), n = 4122 (81.3%), and Roux-en-Y gastric bypass (RYGB), n = 946 (18.7%). Procedures included laparoscopic and robotic-assisted approaches.	SG versus RYGB; outcomes were also analyzed according to BMI categories and age. No nonsurgical comparator was included.	Safety: 30-day mortality, overall intraoperative or postoperative complications, organ/space SSI, superficial and deep SSI, wound disruption, blood transfusion, sepsis, DVT, UTI, pulmonary embolism, unplanned intubation, prolonged ventilation, and pneumonia. Healthcare utilisation: 30-day readmission and reoperation. Perioperative: operating time and length of stay. The study also assessed associations between BMI, age, procedure type, and short-term outcomes.	Overall 30-day mortality was 0.04% (2 deaths), complications occurred in 1.2%, readmission in 3.5%, and reoperation in 1.1%. Higher BMI and younger age were not associated with significantly increased odds of readmission, reoperation, or complications. Compared with SG, RYGB was associated with higher odds of readmission (OR 1.66, 95% CI 1.15–2.38; p = 0.006) and complications (OR 2.27, 95% CI 1.28–4.02; p = 0.005). The proportion of adolescents undergoing SG increased from 73.9% in 2015 to 84.3% in 2018.	Retrospective observational database study; potential misclassification and underreporting of complications; participating MBSAQIP centers are voluntary and do not constitute a weighted nationally representative sample, creating possible selection bias; outcomes limited to 30 days; no assessment of long-term weight loss or comorbidity remission; findings may not represent the entire clinically eligible adolescent population; 80% of participants were aged 18–19 years; inclusion criterion was BMI ≥30 kg/m2, meaning a small proportion of participants did not necessarily have severe obesity by standard definitions.
55	Mohammed et al., 2021 [69]	Retrospective cohort study	72 adolescents (13–19 years) with severe obesity; mean BMI 47.9 kg/m2	Sleeve Gastrectomy	Preoperative versus 12-month postoperative outcomes	%TWL, %EWL, remission of type 2 diabetes, pre-diabetes, hypertension, sleep apnea, complications	Mean %TWL 34.9%; mean %EWL 73.5%. Complete remission of type 2 diabetes, pre-diabetes, and obstructive sleep apnea; 80% remission of hypertension. Long-term complication rate 18%. Higher baseline BMI was associated with lower %EWL. Better adherence to follow-up was associated with improved outcomes.	Single-center retrospective design; relatively small sample size; 12-month follow-up only.
56	Schmoke et al., 2024 [70]	Single-center retrospective review of prospectively collected data; comparative observational cohort study	358 adolescents who underwent laparoscopic sleeve gastrectomy between June 2010 and June 2023 at an academic children's hospital in New York City. Median age was 16.8 years (IQR 15.7–17.8), median baseline BMI was 44.6 kg/m2 (IQR 41.1–49.8), and 68% were female. The pre-COVID cohort included 245 patients and the COVID-era cohort included 113 patients.	Laparoscopic sleeve gastrectomy (LSG)	Pre-COVID cohort (surgery before March 1, 2019; n = 245) versus COVID-era cohort (surgery after March 1, 2020; n = 113). Patients undergoing surgery between March 1, 2019 and March 1, 2020 were excluded because their follow-up crossed between the two periods.	Anthropometric outcomes: absolute weight loss, percentage weight loss, absolute BMI change, and percentage BMI change at 6 and 12 months after surgery. Multivariable analyses also assessed associations with age, gender, ethnicity, baseline BMI, and COVID period.	There were no significant differences between pre-COVID and COVID-era cohorts in weight loss, percentage weight loss, or BMI change at either 6 or 12 months. At 6 months, weight loss was 22.5 vs 21.6 kg, percentage weight loss was 18% vs 18%, and BMI change was 8.4 vs 8.0, respectively. At 12 months, weight loss was 32.3 vs 33.9 kg, percentage weight loss was 25% vs 28%, and BMI change was 12.0 vs 12.3. After multivariable adjustment, surgery during the COVID-19 pandemic was not associated with significantly different weight-loss or BMI outcomes.	Single-center observational study, limiting generalizability; possible residual confounding; inability to assess individual socioeconomic and behavioral factors such as family income, food access, and exercise opportunities; substantial loss of weight data at 12 months, particularly in the COVID cohort (35% data availability vs 59% pre-COVID); telehealth limited objective weight measurement because some families lacked scales; patients returning outside the defined 12-month window were excluded; no nonsurgical comparator; no cardiometabolic, comorbidity-remission, quality-of-life, or safety outcomes were reported as primary study outcomes.
57	Dobbs et al., 2026 [71]	Prospective longitudinal single-arm pre–post study; IMPROVE-T2D cohort	16 adolescents aged 14–19 years with obesity and confirmed youth-onset T2D; mean age 16.6 ± 1.9 years and baseline BMI 47 ± 2 kg/m2	Vertical sleeve gastrectomy (VSG)	No nonsurgical control group	Weight, BMI, body composition, HbA1c, diabetes medication use, lipids, blood pressure, liver enzymes, insulin sensitivity, insulin secretion, β-cell function, and adverse events	At 12 months, weight decreased by 31.0 kg (23%), BMI fell from 47.0 to 36.2 kg/m2, and HbA1c decreased from 6.6% to 5.7%. 86% were off diabetes medications, and substantial improvements occurred in insulin sensitivity, β-cell function, lipids, blood pressure, and liver enzymes.	Small sample size; n = 16 at baseline and n = 14 at 3 and 12 months; no nonsurgical control group; short 12-month follow-up; explicitly extends a previous 3-month report of the same IMPROVE-T2D cohort, creating an overlap risk.
58	Silva et al., 2017 [72]	Prospective longitudinal cohort	64 adolescents aged 15–19 years with severe obesity (BMI ≥40 kg/m2)	Roux-en-Y Gastric Bypass (RYGB)	Baseline vs 30, 180, and 365 days after surgery	BMI, waist circumference, glucose, insulin resistance (HOMA-IR), lipid profile, hypertension, vitamin A status, β-carotene, night blindness	Significant reductions in BMI (45.8 → 27.6 kg/m2), waist circumference, glucose, insulin resistance, triglycerides, LDL cholesterol, and hypertension prevalence (65.6%→15.6%). However, vitamin A deficiency, β-carotene deficiency, and night blindness worsened despite supplementation.	Single-center study; relatively small sample; focused on nutritional deficiencies; follow-up limited to 1 year.
59	Derderian et al., 2020 [73]	Prospective multicenter observational cohort (Teen-LABS)	192 adolescents undergoing RYGB or VSG	Metabolic and bariatric surgery	≥20% TBWL vs <20% TBWL at 5 years	Weight loss, hypertension, dyslipidemia, type 2 diabetes, musculoskeletal pain, renal outcomes	Greater weight loss was associated with better remission of composite dyslipidemia. Remission of type 2 diabetes, hypertension, musculoskeletal pain, and renal dysfunction occurred regardless of achieving ≥20% weight loss.	Post-hoc analysis; possible type II error due to subgroup sizes; OSA not assessed; arbitrary weight-loss cut-off.
60	Collins et al., 2007 [74]	Retrospective cohort study	11 adolescents aged 15–18 years; mean BMI 50.5 kg/m2	Laparoscopic Roux-en-Y Gastric Bypass (LRYGB)	Preoperative vs postoperative follow-up (mean 11.5 months)	% Excess Weight Loss (%EWL), diabetes, hypertension, OSA, PCOS, quality of life (BAROS), complications	%EWL reached 72.9% at 12 months (overall mean 60.8%). Approximately 70% of obesity-related comorbidities improved or resolved. Diabetes improved/resolved in 5 of 6 patients, hypertension improved/resolved in all 6 patients, OSA improved in both affected patients, and quality-of-life scores demonstrated substantial improvement. No mortality occurred.	Very small sample size (n = 11); retrospective design; short follow-up; single-center experience.
61	Burgwardt et al., 2025 [75]	Observational program cohort	Pediatric/adolescent bariatric surgery patients	Robotic bariatric surgery within a pediatric program	As reported; comparator is not mandatory	Perioperative complications and other surgical safety outcomes	Reports protocol-eligible perioperative safety outcomes.	Program experience; outcome detail should be verified during extraction.
62	Nadler et al., 2009 [76]	Prospective FDA-approved cohort study	45 adolescents aged 14–18 years with morbid obesity (BMI 48 ± 6.4 kg/m2)	Laparoscopic Adjustable Gastric Banding (LAGB)	Baseline vs 1-year and 2-year follow-up	BMI, %EWL, waist circumference, body composition, hypertension, impaired glucose tolerance, dyslipidemia, obstructive sleep apnea, asthma, GERD, fasting insulin, HbA1c, lipids, liver enzymes	%EWL reached 46% at 1 year and 47% at 2 years. Fifty-five percent of comorbidities completely resolved and 29% improved. Significant reductions in BMI, waist circumference, fasting insulin, HbA1c, triglycerides, liver enzymes, and improvement in HDL cholesterol. Most weight loss occurred through reduction in android fat mass. OSA, hypertension, GERD, asthma, impaired glucose tolerance, and dyspnea improved substantially.	Single-arm cohort; no control group; moderate sample size; some loss to follow-up; focused on LAGB, a procedure less commonly performed today.
63	Schmitt et al., 2016 [77]	Prospective cohort study	16 adolescents (≥14 years) with severe obesity; median BMI 43.0 kg/m2	Laparoscopic Adjustable Gastric Banding (LAGB)	Baseline vs 1-year and 2-year follow-up	BMI, excess body weight loss (EBWL), insulin resistance, glucose intolerance, hypertension, obstructive sleep apnea, NAFLD, quality of life (PedsQL™), psychosocial outcomes, complications	BMI decreased from 43.0 to 33.0 kg/m2 at 2 years. EBWL reached 49.2%. Hypertension, sleep apnea, glucose intolerance, and NAFLD resolved. Insulin resistance improved markedly. Quality of life and psychosocial functioning improved. No severe surgical complications occurred.	Small sample size (n = 16); single-center study; only 10 patients completed 2-year follow-up; limited statistical power for psychosocial analyses.
64	Zitsman et al., 2011 [78]	Prospective cohort study	100 adolescents aged 14–19 years; median BMI 48.7 kg/m2	Laparoscopic Adjustable Gastric Banding (LAGB)	Baseline vs 6- and 12-month follow-up	Weight loss, BMI change, % excess weight loss (%EWL), hypertension, dyslipidemia, obstructive sleep apnea, asthma, menstrual irregularities/PCOS, CRP, complications	Mean %EWL was 20.9% at 6 months and 31.9% at 12 months. Average BMI decreased from 48.7 to 41.9 kg/m2 at 12 months. Improvements occurred in hypertension, dyslipidemia, OSA symptoms, asthma, menstrual irregularities, and inflammatory markers. No mortality occurred.	Short-term follow-up; only 47 patients had 12-month data; no control group; repeat sleep studies were not performed to objectively confirm OSA resolution.
65	Peña et al., 2017 [79]	Prospective single-center cohort study	21 adolescents aged 14–18 years with severe obesity; mean BMI 47.3 ± 8.4 kg/m2	Laparoscopic Adjustable Gastric Banding (LAGB)	Baseline vs follow-up to 48 months	BMI, weight loss, pre-diabetes, liver function, lipid abnormalities, quality of life (PedsQL), postoperative complications, admissions	Significant BMI reduction at all follow-up time points. Median maximum BMI loss was 10 kg/m2. Pre-diabetes resolved in affected participants, liver function improved significantly, and quality-of-life scores improved for both adolescents and parents. No life-threatening complications occurred.	Small sample size; single-center study; incomplete long-term follow-up for some participants; limited objective sleep outcome assessment.
66	Fielding & Duncombe, 2005 [80]	Retrospective cohort study	41 adolescents aged 12–19 years; mean BMI 42.4 ± 8.2 kg/m2	Laparoscopic Adjustable Gastric Banding (LAGB)	Baseline vs follow-up up to 5 years	BMI, excess weight loss (EWL), diabetes, hypertension, sleep apnea, mobility-related outcomes, complications	Mean BMI decreased from 42.4 to 29 kg/m2 at 3 years and remained stable through 5 years. Mean EWL reached 70% at 3 years. Eighty-three percent of patients were no longer obese (BMI <30 kg/m2). Diabetes, hypertension, and sleep apnea resolved. No operative mortality occurred.	Retrospective design; no control group; limited standardized reporting of metabolic outcomes; quality-of-life outcomes largely descriptive.
67	Dolan et al., 2003 [81]	Prospective cohort study	17 adolescents aged 12–19 years with morbid obesity	Laparoscopic Adjustable Gastric Banding (LAGB)	Preoperative vs postoperative follow-up	Weight loss, BMI, % excess weight loss, complications	Median BMI decreased from 44.7 to 30.2 kg/m2 at 24 months. Median weight loss was 35.6 kg and median %EWL was 59.3%. Thirteen of 17 patients (76.5%) lost ≥50% excess weight. Two complications occurred (one band slip, one leaking port) and no mortality was reported.	Small sample size (n = 17); single-surgeon, single-center study; no control group; follow-up limited to medium-term outcomes; limited assessment of obesity-related comorbidities.
68	Zeinoddini et al., 2014 [82]	Prospective cohort study	12 morbidly obese adolescents (<18 years; mean BMI 46.0 kg/m2)	Laparoscopic Gastric Plication (LGP)	Preoperative vs postoperative follow-up	%EWL, %EBMIL, obesity-related comorbidities (DM, prediabetes, dyslipidemia, prehypertension), sleep apnea symptoms, complications	Mean %EWL was 68.2% and mean %EBMIL was 79.0% at 2 years. All medical comorbidities improved/remitted. Sleep apnea symptoms resolved. No mortality occurred; one patient required reoperation for obstruction.	Small sample size (n = 12); single-surgeon study; no control group; limited power; long-term durability uncertain.
69	Göthberg et al., 2014 [83]	Nationwide prospective cohort study (AMOS)	81 adolescents aged 13–18 years with severe obesity (BMI ≥40 kg/m2 or ≥35 kg/m2 with comorbidity)	Laparoscopic Roux-en-Y Gastric Bypass (RYGB)	Comparison with 81 conventionally treated adolescents and 81 adults undergoing RYGB	Weight, BMI, excess weight loss, cardiovascular and metabolic risk factors, inflammatory markers, quality of life, adverse events	Weight decreased from 133 kg to 89 kg at 2 years. Mean weight loss was 32%, and 95% achieved >50% excess weight loss. Significant improvements occurred in cardiovascular/metabolic risk factors, inflammation, and quality of life. No postoperative mortality occurred.	Non-randomized design; psychosocially complex cohort; some adverse-event follow-up beyond 2 years was partly retrospective; long-term nutritional adherence was suboptimal.
70	Olbers et al., 2017 [7]	Prospective nationwide non-randomized controlled study	81 adolescents aged 13–18 years with severe obesity (BMI ≥40 kg/m2 or ≥35 kg/m2 with comorbidity)	Laparoscopic Roux-en-Y Gastric Bypass (RYGB)	80 conservatively treated adolescents and 81 adult RYGB patients	BMI, weight, glucose homeostasis, HbA1c, insulin, lipids, blood pressure, inflammatory markers, liver enzymes, quality of life, adverse events, nutritional deficiencies	Mean BMI decreased by 13.1 kg/m2 and body weight by 36.8 kg over 5 years. Significant improvements occurred in glucose metabolism, insulin resistance, dyslipidemia, hypertension, inflammatory markers, and quality of life. All baseline type 2 diabetes cases were in remission at 5 years.	Non-randomized design; crossover of controls to surgery during follow-up; only RYGB evaluated; nutritional deficiencies and reoperations were common after surgery.
71	Lainas et al., 2020 [84]	Retrospective analysis of prospectively collected cohort	84 adolescents aged 15–17 years with severe obesity	Laparoscopic Sleeve Gastrectomy (LSG)	Preoperative vs postoperative follow-up	BMI, %EWL, %TWL, diabetes, insulin resistance, hypertension, dyslipidemia, OSAS, arthralgia, GERD, quality of life, complications, mortality	BMI decreased from 43.7 to 29.8 kg/m2 at 1 year and 28.8 kg/m2 at 2 years. Mean %EWL was 76.2% at 1 year and 79.1% at 2 years. Significant remission of insulin resistance, hypertension, OSAS, and arthralgia. Quality-of-life scores improved significantly. One major complication (pneumonia); no mortality.	Retrospective analysis; single-center study; loss to follow-up at 2 years (28.8%); no control group.
72	Peacey et al., 2026 [85]	Retrospective cohort study based on a prospectively maintained bariatric database; single-centre, single-surgeon observational study	63 adolescents aged 13–19 years who underwent primary sleeve gastrectomy between December 2011 and December 2024 in Brisbane, Australia. Mean age was 16 years; 78% were female. Mean preoperative BMI was 42.7 kg/m2 (reported as 42.1 kg/m2 in Table 1), and mean weight was 122.5 kg. All patients had previously failed lifestyle management involving diet, exercise, and education.	Primary laparoscopic sleeve gastrectomy (SG) delivered within a multidisciplinary program involving a surgeon, dietitian, psychologist, and nurse practitioner	No concurrent nonsurgical or alternative-surgery comparator group; preoperative measurements served as the baseline for within-group postoperative comparisons	Anthropometric: weight, BMI, excess weight loss, and longitudinal BMI trajectory up to 5 years. Safety: 30-day and later surgical complications, staple-line leak, subsequent cholecystectomy, conversion to RYGB, mortality, and length of stay. Mental health/psychiatric: baseline psychiatric and mental-health disorders, including ASD, ADHD, bipolar disorder, OCD, depression, anxiety, self-harm and suicidal ideation; weight-related bullying was also recorded. Metabolic/nutritional: T2DM/prediabetes, high cholesterol, low ferritin, vitamin D deficiency, and vitamin B12 deficiency at baseline. Follow-up: postoperative follow-up rates from 1 month to 5 years.	Mean BMI decreased from 42.7 kg/m2 before surgery to 30.6 at 6 months, 27.5 at 1 year, 28.0 at 2 years, and 30.4 kg/m2 at 5 years. Excess weight loss was 72% at 6 months, 90% at 1 year, 85% at 2 years, and 70% at 5 years. Of the cohort, 41% had at least one pre-existing psychiatric and/or mental-health condition; seven had a history of self-harm or suicidal ideation. There was no mortality, but one patient developed a staple-line leak requiring two laparoscopic washouts and eventual conversion to RYGB; six patients subsequently underwent laparoscopic cholecystectomy. Follow-up declined substantially over time, reaching only 16% at 5 years.	Single-centre, single-surgeon private-practice cohort; retrospective analysis of prospectively collected data; no control or comparator group; substantial risk of selection bias; very poor long-term follow-up, creating major attrition and survivorship bias; only 16% follow-up at 5 years; postoperative psychological, mental-health and quality-of-life outcomes were not collected, meaning the study cannot determine whether surgery improved or worsened mental health; limited metabolic and comorbidity follow-up outcomes; participants ranged from 13 to 19 years, potentially including some individuals outside a strict ≤18-year adolescent age criterion; and generalizability is limited by the private Australian clinical setting.
73	Dargan et al., 2018 [86]	Retrospective analysis of prospectively collected cohort data	13 adolescents aged 16–19 years with morbid obesity (mean BMI 46.2 ± 6.3 kg/m2)	Laparoscopic Sleeve Gastrectomy (LSG)	Preoperative vs postoperative follow-up	BMI, weight loss, %EWL, body composition, diabetes remission, hypertension, dyslipidaemia, sleep apnea, complications, hospital stay	At 1 year, %EWL was 64.3%, BMI decreased from 46.2 to 31.2 kg/m2, and mean weight decreased by 39.3 kg. Remission/improvement occurred in diabetes, hypertension, and dyslipidaemia. No complications, reoperations, readmissions, or mortality were reported.	Small sample size (n = 13); retrospective analysis; substantial loss to follow-up (only 5 patients at 1 year and 3 at 3 years); single-center study.
74	Khoursheed et al., 2019 [87]	Retrospective cohort study	22 adolescents (10–19 years), mean age 16.4 ± 2.1 years, mean BMI 45.1 ± 8.1 kg/m2	Sleeve Gastrectomy (SG)	Preoperative vs postoperative follow-up	Weight loss, BMI loss, %EWL, type 2 diabetes remission, hypertension, OSA, complications, reoperations	Mean BMI decreased by 13.8 ± 10.5 kg/m2. Mean %EWL was 50.9 ± 27.1%. Both patients with type 2 diabetes achieved complete remission without medication. One of two hypertensive patients discontinued medication. OSA improved symptomatically. No leaks, reoperations, or mortality occurred. One patient developed vitamin deficiency related to vomiting/non-compliance.	Small sample size (n = 22); retrospective single-surgeon study; median follow-up only 9 months; incomplete long-term follow-up; limited reporting of lipid outcomes.
75	Mostafa et al., 2022 [88]	Prospective cohort study	32 adolescents (<19 years) with morbid obesity (BMI ≥35 kg/m2) and type 2 diabetes mellitus; mean age 17.0 ± 0.97 years; mean BMI 49.2 ± 3.4 kg/m2	Laparoscopic Sleeve Gastrectomy (LSG)	Preoperative vs postoperative follow-up (12 months)	HbA1c, fasting blood glucose (FBS), diabetes remission, excess weight loss (EWL), postoperative complications	27/32 patients (84.3%) achieved diabetes remission without medication. Mean HbA1c decreased from 9.6% to 6.5%, and mean FBS decreased from 157.9 to 116.7 mg/dL at 12 months. Mean excess weight loss reached 42 kg (≈65% EWL). No major postoperative complications were reported.	Short follow-up (12 months); no control group; relatively small sample size; exclusively diabetic adolescents, limiting generalizability to all adolescents with severe obesity.
76	Goldenshluger et al., 2023 [89]	Retrospective analysis of prospectively collected cohort data	31 adolescents (<18 years at surgery) with severe obesity; mean age 16.3 years; mean BMI 44.9 kg/m2	Laparoscopic Sleeve Gastrectomy (LSG)	Preoperative vs long-term postoperative follow-up (mean 10.8 years)	BMI, weight loss, hypertension, dyslipidemia, fasting glucose, obstructive sleep apnea, GERD, nutritional deficiencies, revisional surgery, patient satisfaction	Mean BMI decreased from 44.9 to 30.1 kg/m2 at 10 years. Mean %TWL was 32.3%. Hypertension resolved in 88.9% of affected patients. Impaired fasting glucose resolved. OSA remitted in all affected patients. Long-term complications included GERD (22.6%), iron deficiency anemia (19%), vitamin deficiencies, alopecia (48.4%), and cholecystectomy (22.6%).	Retrospective design; 32.6% lost to follow-up; small sample size; missing long-term lipid data for some patients; single-center study.
77	Iossa et al., 2018 [90]	Retrospective multicenter cohort study	41 adolescents with morbid obesity (mean age 16 years; mean BMI 49 kg/m2), including 7 with syndromic obesity	Laparoscopic Sleeve Gastrectomy (LSG)	Preoperative vs postoperative follow-up	BMI, %EWL, T2DM, hypertension, OSA, NASH, complications	Mean %EWL reached 59.4% at 24 months. Remission rates were 71% for T2DM, 75% for hypertension, and 61% for OSA. Overall comorbidity remission rate was 78.2%. No mortality occurred and postoperative complications were uncommon (4.9%).	Retrospective design; relatively small sample size; inclusion of syndromic obesity may increase heterogeneity; no control group.
78	Tuna et al., 2020 [91]	Retrospective cohort study using prospectively collected data	16 adolescents (<19 years), median age 17.5 years, median BMI 48.6 kg/m2; all had ≥1 adiposity-related comorbidity	Laparoscopic Sleeve Gastrectomy (LSG)	Preoperative vs postoperative follow-up	BMI, %TWL, %EWL, insulin resistance, hypertension, dyslipidemia, obstructive sleep apnea (OSA), NAFLD, hyperuricemia, complications, nutritional deficiencies	Mean %TWL was 32.5% and mean %EWL was 68%. Median BMI decreased from 48.6 to 31.7 kg/m2. Resolution occurred in 100% of insulin resistance, OSA, and hyperuricemia cases; hypertension remitted in 66% and dyslipidemia/NAFLD improved in ∼60%. No perioperative complications or mortality occurred.	Small sample size (n = 16); retrospective design; single-center experience; median follow-up only 18.5 months; few patients reached 2–3 years follow-up.
79	Boza et al., 2012 [92]	Retrospective cohort study	51 obese adolescents ≤19 years; mean age 18 ± 1.45 years; mean BMI 38.5 ± 3.7 kg/m2	Laparoscopic Sleeve Gastrectomy (LSG)	Preoperative vs postoperative follow-up	BMI, %EWL, insulin resistance, dyslipidemia, type 2 diabetes, hypertension, quality of life, complications	%EWL was 94.6% at 6 months, 96.2% at 1 year, and 92.9% at 2 years. BMI decreased from 38.5 to 26.3 kg/m2 at 2 years. Remission rates were 96.2% for insulin resistance, 100% for hypertension, 58% for dyslipidemia, and 50% for T2DM. All patients reported good or very good quality of life after surgery. One staple-line leak occurred (1.9%) and was successfully treated.	Retrospective design; single-center study; relatively short follow-up (maximum 2 years); no control/comparator group; limited reporting of some comorbidity outcomes such as fatty liver resolution.
80	Franco et al., 2017 [93]	Retrospective descriptive case series	22 adolescents aged 14–19 years with severe obesity (mean BMI 46.5 kg/m2)	Laparoscopic Sleeve Gastrectomy (LSG)	Preoperative vs postoperative follow-up (6, 12, 18, and 24 months)	BMI, weight loss, waist circumference, blood pressure, lipid profile, fasting glucose, insulin, HOMA-IR, type 2 diabetes, hepatic steatosis, metabolic syndrome, complications	Mean weight loss was 34.5 kg at 12 months with 60% EWL. BMI decreased by 12.3 kg/m2. Dyslipidemia improved in 67.8%, hepatic steatosis significantly decreased, hypertension improved, and type 2 diabetes achieved complete remission. Significant reductions in insulin resistance (HOMA-IR) were observed. No deaths, readmissions, or major postoperative complications occurred.	Retrospective design; small sample size (n = 22); no control group; some loss to follow-up over time; follow-up limited to 24 months.
81	Pedroso et al., 2015 [94]	Retrospective analysis of a prospectively collected database	174 morbidly obese adolescents (137 LAGB, 37 VSG); mean age ∼17 years	Vertical Sleeve Gastrectomy (VSG)	Laparoscopic Adjustable Gastric Banding (LAGB)	Weight, BMI, %EWL, %BMI loss, lipid profile, HbA1c, fasting glucose, CRP, complications	At 24 months VSG achieved greater %EWL than LAGB (70.9% vs 35.5%, p = .004) and greater %BMI loss (32.3% vs 16.4%, p = .004). Both procedures improved HDL, HbA1c, fasting glucose, and CRP. Complications were more frequent after LAGB (23.4%) than VSG, although one VSG patient died from mesenteric venous thrombosis.	Retrospective non-randomized design; small VSG sample (n = 37); substantial loss to follow-up by 24 months; unequal group sizes; short-term follow-up for VSG.
82	Qureshi et al., 2021 [95]	Single-center prospective longitudinal observational comparative cohort study conducted from 2015 to 2020, with outcomes assessed for up to 24 months	352 adolescents referred for metabolic and bariatric surgery evaluation. Mean age was 15.6 ± 1.4 years; 69% were female; 38% were Hispanic; median baseline weight was approximately 135 kg and BMI 48.4 kg/m2. Of these, 79 (22%) underwent LSG within 6 months and 273 (78%) did not. Referral criteria generally included age >14 years with severe obesity: ≥120% of the 95th BMI percentile, BMI z-score ≥2.33, BMI >35 kg/m2 with serious obesity-related comorbidities, or BMI >40 kg/m2 without previously identified complications.	Laparoscopic sleeve gastrectomy (LSG), n = 79	NoLSG group, n = 273: adolescents referred to the same MBS program who did not undergo LSG within 6 months of initial referral. They received nutritional, social, psychological, and lifestyle counseling recommendations but no bariatric surgery during the defined exposure period.	Anthropometric: body weight, BMI, percentage total weight change, percentage total BMI change at 3, 6, 12, 18, and 24 months. Comorbidity resolution: obstructive sleep apnea, prediabetes/T2DM, hypertension, hyperlipidemia, and PCOS. Safety: perioperative complications, reoperation, readmissions, endoscopy/dilation, cholecystectomy, leaks, mortality, and length of hospital stay. Healthcare pathway: proportion proceeding to surgery, reasons for not undergoing surgery, and predictors of high weight-loss performance after LSG.	At 24 months, LSG patients achieved 21.4% total weight loss and 22.3% total BMI loss, whereas NoLSG patients had 3.7% total weight gain and 3.0% BMI gain. At 12 months, the LSG group showed substantial comorbidity resolution: OSA 38%, prediabetes/T2DM 54%, hypertension 53%, hyperlipidemia 29%, and PCOS 33%; the NoLSG group showed no documented resolution of these conditions. Safety outcomes included one postoperative bleed requiring additional laparoscopy and washout, five early readmissions for nausea/vomiting, four endoscopies with dilation, and five subsequent cholecystectomies. There were no leaks or deaths. Only 22% of referred adolescents underwent LSG within 6 months.	Nonrandomized observational design with substantial risk of selection bias and confounding; intervention and comparator groups were self-selected rather than randomized; the NoLSG group was heterogeneous and included patients who did not return, were nonadherent, had insurance denial, had psychological contraindications, or remained in the preoperative pathway; 29 patients still pursuing surgery after 6 months were classified as NoLSG; substantial differential attrition, with only 28.8% of eligible LSG patients and 21.8% of NoLSG patients having weight data at 24 months; single-center design limits generalizability; comorbidity resolution in the NoLSG group depended on provider documentation; inconsistent behavioral assessment and psychological monitoring; and baseline weight timing differed between groups, with the initial bariatric visit used for NoLSG but day-of-surgery weight used for LSG.
83	Barnett et al., 2005 [96]	Retrospective cohort study	14 adolescents (13–17 years) with morbid obesity; mean BMI 55.1 kg/m2	Bariatric surgery (VBG n = 7, RYGB n = 5, JIB n = 3)	Preoperative vs postoperative follow-up	BMI, excess weight loss, hypertension, obstructive sleep apnea, diabetes, asthma, musculoskeletal symptoms, complications	Mean BMI decreased from 58.5 to 32.1 kg/m2 among patients with ≥9 months follow-up (45% reduction). Average excess weight loss was 64%. Hypertension resolved in 100%, OSA resolved in 100%, and diabetes resolved in the single affected patient. No perioperative deaths occurred.	Small sample size; retrospective design; heterogeneous procedures; variable follow-up; only 9 patients available for long-term analysis; includes obsolete jejunoileal bypass procedures.
84	Paulus et al., 2016 [97]	Retrospective cohort study	10 adolescents ≤18 years with morbid obesity (BMI >40 kg/m2 or >35 kg/m2 with comorbidity)	Laparoscopic Adjustable Gastric Banding (LAGB)	Preoperative vs postoperative follow-up	BMI loss, excess BMI loss (%EBMIL), hypertension, musculoskeletal complaints, asthma/dyspnea, adverse events, reoperations, band removal	Median BMI loss was 10.7 kg/m2 after a median follow-up of 64 months. Six of ten patients achieved >50% excess BMI loss. Hypertension resolved in both affected patients. Four of ten patients required band removal, resulting in a 40% treatment failure rate. No major adverse events or mortality occurred.	Very small sample size (n = 10); retrospective design; 30% loss to follow-up; no control group; substantial reoperation/band-removal rate; limited cardiometabolic data (no diabetes or dyslipidemia cases).
85	Yska et al., 2021 [98]	Retrospective single-center longitudinal cohort study	56 adolescents with severe obesity who underwent LAGB between 2008 and 2019 at a French pediatric center. Mean age was 16.5 ± 0.9 years (range 14.4–18.6); 77% were female. Mean baseline weight was 127 ± 18 kg and mean BMI was 45 ± 5 kg/m2. Surgical eligibility required bone age ≥14 years and BMI >40 kg/m2 or >35 kg/m2 with significant comorbidity, following failure to achieve significant weight loss after 12 months of multidisciplinary lifestyle management.	Laparoscopic adjustable gastric banding (LAGB)	No concurrent nonsurgical or alternative-surgery comparator group; participants served as their own baseline controls for preoperative-to-postoperative comparisons.	Anthropometric: body weight, BMI, BMI change, percentage excess weight loss (%EWL), and proportion achieving ≥50% EWL. Cardiometabolic: insulin resistance, prediabetes, dyslipidemia, prehypertension, fasting glucose, insulin, HbA1c, HOMA index, and lipid profile. Hepatic: liver steatosis, ALT, AST, GGT, ultrasound findings, and liver-biopsy findings in a subset. Nutritional: iron, ferritin, vitamin B12, and nutritional deficiencies. Safety/surgical: band removal, second bariatric procedures, gastrointestinal intolerance, band slippage, perforation, erosion, reoperation, cholecystectomy, and mortality. Follow-up: pediatric and postpediatric follow-up rates through as long as 9 years.	Mean BMI decreased by 9 kg/m2 at 12 months, 11 kg/m2 at 24 months, 12 kg/m2 at 36 months, and 11 kg/m2 at last contact. Mean EWL was 47% at 1 year, peaked at 62% at 3 years, and was 55% at last contact. Comorbidities improved substantially: insulin resistance decreased from 54% at baseline to 4% at 3 years, prediabetes from 48% to 14%, dyslipidemia from 44% to 7%, and steatosis from 86% to 46%; prehypertension did not significantly improve. However, 17 patients (30%) had the gastric band removed, 12 (21%) underwent another bariatric procedure, and one patient died 40 months after surgery due to gastric hemorrhage associated with band erosion. No important nutritional deficiencies were reported. Follow-up declined from 96% at 3 years to 54% at 6 years and 29% at 9 years.	Retrospective, single-center observational design; small sample size (n = 56); no concurrent comparator or control group; substantial long-term attrition, with 39% lost to follow-up at last contact; follow-up fell to only 29% at 9 years; patients lost to follow-up may have had poorer outcomes, potentially causing overestimation of LAGB effectiveness; heterogeneous follow-up duration and missing data across time points; postpediatric information was sometimes obtained through informal contact; comorbidity prevalence was calculated only among participants available at each time point, creating selection bias; long-term comorbidity outcomes beyond 3 years were unavailable; only eight participants had liver biopsies; age range extended to 18.6 years; and this may overlap with earlier publications from the same French adolescent LAGB cohort.
86	Stafler et al., 2025 [99]	Prospective single-center cohort study	17 adolescents with morbid obesity (mean age 17.1 years; mean BMI 45.5 kg/m2) undergoing bariatric surgery	Laparoscopic Sleeve Gastrectomy (LSG)	Preoperative vs postoperative follow-up	Lung Clearance Index (LCI), spirometry (FEV1, FVC), plethysmography, diffusion capacity, 6-min walk test, BMI change, OSA status	BMI decreased from 45.5 to 36.5 kg/m2. LCI improved significantly from 7.0 to 6.5 (p = .009), indicating improved airway function. FEV1, TLC, FRC, RV and FEF25–75 also improved after surgery. Changes in LCI correlated with BMI reduction.	Small sample size (n = 17); no control group; short follow-up (median 4 months); high attrition; respiratory outcomes only; not designed to evaluate long-term weight or cardiometabolic outcomes.
87	Dumont et al., 2018 [100]	Retrospective single-center observational cohort	97 adolescents; mean age 17.2 ± 0.7 years; mean BMI 44.9 ± 6.1 kg/m2	Laparoscopic Adjustable Gastric Banding (LAGB)	Preoperative vs postoperative follow-up	BMI, weight loss, EWL, TWL, hypertension, type 2 diabetes, dyslipidemia, sleep apnea syndrome (SAS), asthma, adverse events, band removal	Mean follow-up 56 months. BMI decreased from 44.2 to 35.2 kg/m2 (−9.0 kg/m2). Mean EWL was 46.6% and TWL 20.0%. Resolution rates were 100% for hypertension, 100% for T2DM, 50% for sleep apnea, and 100% for asthma. No mortality or perioperative morbidity occurred. Band removal was required in 25% of patients.	Retrospective design; 21.6% lost to follow-up; no control group; substantial band-removal rate (25%); single-center experience.
88	Järvholm et al., 2023 [13]	Multicentre, open-label, randomized controlled trial; 1:1 allocation; intention-to-treat analysis; 2-year follow-up	50 adolescents aged 13–16 years with severe obesity and BMI ≥35 kg/m2 who had received obesity treatment for at least 1 year, including ≥6 months at a specialized pediatric obesity unit. Mean age was 15.8 years and mean baseline BMI was 42.6 kg/m2. Participants were randomized to MBS (n = 25) or intensive non-surgical treatment (n = 25).	Metabolic and bariatric surgery (MBS): primarily laparoscopic Roux-en-Y gastric bypass (23 participants) and sleeve gastrectomy (2 participants). Participants also received postoperative follow-up and multivitamin/mineral supplementation.	Intensive non-surgical obesity treatment beginning with an 8-week low-calorie diet of approximately 880 kcal/day, followed by multidisciplinary lifestyle treatment targeting approximately 1500 kcal/day, 60 min of moderate-to-vigorous physical activity daily, reduced sedentary behavior, and frequent clinical contacts.	Primary outcome: 2-year change in BMI. Anthropometric: body weight, BMI, relative BMI change, proportion achieving ≥5% and ≥10% weight reduction. Cardiometabolic: glucose, insulin, HbA1c, oral glucose tolerance, triglycerides, HDL, LDL, blood pressure, high-sensitivity CRP, and ALT. Quality of life: physical and mental RAND-36 domains and obesity-specific OP-14. Eating behavior: binge eating, uncontrolled eating, emotional eating, and cognitive restraint. Safety: surgical adverse events, vitamin and mineral levels, gastrointestinal symptoms, bone mineral density/content, growth, excess skin, and pregnancies. Mental health/risk behavior: anxiety, depression, anger, disruptive behavior, self-concept, suicidal ideation, alcohol use, drug use, and smoking.	At 2 years, mean BMI decreased by 12.6 kg/m2 in the MBS group versus 0.2 kg/m2 in the intensive non-surgical group, giving a between-group difference of −12.4 kg/m2 (95% CI −15.5 to −9.3; p < .0001). Relative body-weight change was −28.7% after MBS versus +0.4% with non-surgical treatment. At 2 years, 63% of surgical participants had BMI <30 kg/m2 versus 4% of controls. MBS produced greater improvements in several cardiometabolic risk factors, physical functioning, general health, binge eating, and uncontrolled eating. Surgical adverse events were few and mild, but MBS resulted in a greater reduction in bone mineral density. No clinically relevant between-group differences were found in vitamin/mineral concentrations or overall mental health at 2 years.	Small sample size (n = 50); open-label design with no participant or staff masking after allocation; short follow-up of only 2 years for evaluating lifelong bariatric outcomes; predominantly female sample; conducted exclusively at three Swedish tertiary university centers, potentially limiting generalizability; five participants in the non-surgical group crossed over to bariatric surgery during the second year, with an additional participant undergoing surgery abroad and being lost to follow-up; secondary outcomes were not adjusted for multiple testing and therefore require confirmation; most surgical participants underwent Roux-en-Y gastric bypass, limiting procedure-specific conclusions for sleeve gastrectomy; no objective measures of calorie intake or physical activity; and the study was not powered to detect uncommon or long-term adverse events.
89	Zolfaghari et al., 2025 [101]	Prospective observational cohort study using non-random, consecutive sampling; 1-year postoperative follow-up	219 adolescents aged 10–19 years with severe obesity from the Tehran Obesity Treatment Study (TOTS): 182 underwent SG and 37 underwent GB. Mean age was 17.0 ± 1.8 years in the SG group and 17.8 ± 1.1 years in the GB group. Eligibility required BMI >99th percentile, or BMI ≥95th to <99th percentile with at least one obesity-related complication.	Sleeve gastrectomy (SG), n = 182	Gastric bypass (GB), n = 37	Anthropometric: BMI, EWL%, total weight loss, body-fat percentage and mass, fat-free mass, waist and hip circumference, waist-to-height ratio. Metabolic: systolic and diastolic blood pressure, fasting glucose, triglycerides, total cholesterol, LDL-C, HDL-C, AST/SGOT, ALT/SGPT, and alkaline phosphatase. Nutritional: vitamin D, vitamin B12, calcium, phosphorus, zinc, copper, iron, hemoglobin, ferritin, and albumin. Follow-up occurred at 1, 3, 6, and 12 months depending on outcome.	Both procedures produced substantial weight loss. At 3 months, BMI was significantly lower after SG than GB (36.1 ± 4.0 vs 38.5 ± 4.8 kg/m2; p = .030), but no significant BMI difference remained at 6 or 12 months. At 12 months, EWL was similar (80.0% SG vs 80.9% GB; p = .850). GB had significantly lower total cholesterol and LDL-C at 12 months. At 6 months, SG had higher HDL-C and lower AST and ALT than GB. No significant between-group differences in measured micronutrient status were detected at 12 months.	Non-randomized surgical selection creates substantial risk of confounding by indication; severe group-size imbalance (SG n = 182 vs GB n = 37); considerable and differential loss of available observations during follow-up, especially for micronutrients; only 1 year of follow-up; no information on physical activity; dietary intake was not recorded; supplementation adherence was not assessed; baseline age differed significantly between groups and baseline ALT also differed; no information on the severity/stage of fatty liver disease; no reported multivariable adjustment for potential confounders in the main comparisons; and adverse surgical events were not systematically reported as study outcomes.
90	Cozacov et al., 2014 [102]	Retrospective review of a prospectively collected database	18 adolescents aged 11–19 years; mean age 17.5 years; mean BMI 47.2 kg/m2	Laparoscopic Sleeve Gastrectomy (LSG) and Roux-en-Y Gastric Bypass (RYGB)	Comparison between LSG and RYGB; also preoperative vs postoperative follow-up	Weight, BMI, hypertension, sleep apnea, diabetes, GERD, depression, hypercholesterolemia, complications, quality of life	Mean BMI decreased from 47.2 to 30.1 kg/m2 at mean 55.2 months follow-up. Overall, 13 of 16 comorbidities (81%) resolved. Sleep apnea resolved in all patients with follow-up, hypertension resolved in 100%, diabetes resolved in 100%, and hypercholesterolemia resolved in 100%. No postoperative complications or mortality were reported.	Small sample size (n = 18); retrospective design; incomplete follow-up (83% available >1 year); quality-of-life data collected only postoperatively; limited statistical power for comparison between LSG and RYGB.
91	Angrisani et al., 2005 [103]	Retrospective multicenter cohort study	58 adolescents aged 15–19 years with severe obesity; mean BMI 46.1 kg/m2	Laparoscopic Adjustable Gastric Banding (Lap-Band System)	Preoperative vs postoperative follow-up (1, 3, 5, and 7 years)	BMI, % Excess Weight Loss (EWL), sleep apnea prevalence, hypertension, diabetes, dyslipidemia, complications, band removal, reoperations	No mortality. Mean BMI decreased from 46.1 to 29.7 kg/m2 at 7 years. Mean EWL reached 55.6% at 7 years. Postoperative complication rate 10.3%; band removal rate 10.3%. Authors concluded Lap-Band was an effective and conservative surgical option for severely obese adolescents.	Retrospective design; no control group; relatively small sample size; decreasing follow-up numbers over time (only 10 patients at 7 years); limited reporting of postoperative cardiometabolic and sleep outcome changes.
92	Kalra et al., 2005 [104]	Retrospective cohort study with pre-post analysis	35 extremely overweight adolescents (23 girls, 12 boys), mean age 17.6 years, mean BMI 57.0 kg/m2	Laparoscopic Roux-en-Y Gastric Bypass	Baseline vs postoperative follow-up after significant weight loss	OSA prevalence, Apnea-Hypopnea Index (AHI), arousal index, mean and minimum oxygen saturation, weight, BMI	OSA prevalence was 55% (19/34). Among 10 patients with follow-up polysomnography, mean weight loss was 58 kg and BMI decreased from 60.8 to 41.6 kg/m2. Median AHI improved from 9.1 to 0.65 events/hour, and minimum oxygen saturation improved significantly (82.9%–91.7%). Most patients experienced resolution or marked improvement of OSA following surgery.	Retrospective design; small sample size; only 10 of 19 OSA patients completed postoperative sleep studies; follow-up duration was short (∼5 months); no non-surgical control group.
93	Shah et al., 2024 [105]	Retrospective comparative cohort study using the 2016–2021 MBSAQIP database, with propensity-score adjustment and additional retrospective analysis of center-specific databases from three specialized centers	MBSAQIP cohort: 4755 patients aged ≤18 years undergoing MBS, including 47 preteens aged <13 years and 4708 teens aged 13–18 years. Mean preoperative BMI was 46.9 ± 7.0 kg/m2 in preteens and 47.0 ± 12.5 kg/m2 in teens. Center-specific cohort: 41 preteens aged <13 years, likely substantially overlapping with the 47 preteens in MBSAQIP	Metabolic and bariatric surgery. All 47 preteens in MBSAQIP underwent laparoscopic sleeve gastrectomy (LSG). Among teens, 3971 (84%) underwent LSG, 602 (13%) underwent RYGB, and approximately 3% underwent other/unlisted procedures. All 41 preteens in the center-specific analysis underwent LSG.	Primary comparison: preteens aged <13 years undergoing MBS versus teens aged 13–18 years undergoing MBS. For risk-adjusted analyses, preteens were compared with a propensity-score-matched subset of teens, with adjustment for age, sex, race, and propensity-score quintile.	Anthropometric: change in BMI at 30 days in MBSAQIP; BMI, absolute BMI loss, percentage BMI loss, and percentage excess BMI loss at 3, 6, and 12 months in center-specific preteen data. Safety: postoperative complications, 30-day unplanned readmission, 30-day reoperation, and mortality. Baseline comorbidities: T2DM, insulin-dependent T2DM, sleep apnea, GERD, hypertension, hyperlipidemia, renal insufficiency, DVT/PE history, and other conditions.	Preteens and teens had comparable risk-adjusted 30-day BMI reductions: 4.2 kg/m2 (95% CI 3.0–5.4) versus 4.6 kg/m2 (95% CI 4.4–4.7), respectively (p = .6). No preteen experienced a complication, unplanned readmission, reoperation, or mortality within 30 days, compared with rates of 0.5%, 2.9%, 0.8%, and 0.1%, respectively, among teens; none of these differences was statistically significant. In center-specific preteen data, mean BMI decreased by 7 ± 3 kg/m2 at 3 months, 10 ± 3 kg/m2 at 6 months, and 9 ± 4 kg/m2 at 12 months, corresponding to percentage BMI reductions of 16 ± 7%, 23 ± 7%, and 20 ± 8%, respectively.	Retrospective observational design; only 47 preteens in the MBSAQIP cohort despite 4755 total participants; rare adverse events and small preteen sample limit statistical power; MBSAQIP comparative outcomes are primarily limited to 30 days; no long-term teen comparator data; center-specific weight outcomes were available only for preteens; substantial missing data at later follow-up, with only 19 of 41 center-specific preteens having 12-month data; likely substantial overlap between the 47 MBSAQIP preteens and 41 center-specific patients; heterogeneous data-collection methods across the three centers; no standardized assessment of comorbidity remission; potential data-entry errors and selection bias inherent in administrative databases; three centers accounted for most preteen cases, limiting generalizability; MBSAQIP lacks several pediatric-specific clinical variables; and substantial potential cohort overlap exists with other MBSAQIP adolescent bariatric studies.
94	Poliakin et al., 2020 [106]	Retrospective registry cohort	Adolescent subgroup compared with young adults after bariatric surgery	Metabolic/bariatric surgery	Adolescents versus young adults	30-day complications, reoperation, readmission, and mortality	Reports protocol-eligible perioperative safety outcomes for adolescents.	Administrative registry; short-term follow-up.
95	Ogle et al., 2021 [107]	Prospective multicenter observational cohort (Teen-LABS)	228 adolescents ≤19 years with severe obesity; mean baseline BMI 52.6 kg/m2	Roux-en-Y Gastric Bypass (RYGB) and Vertical Sleeve Gastrectomy (VSG)	Younger adolescents (13–15 years) vs older adolescents (16–19 years); 5-year follow-up	BMI change, hypertension remission, dyslipidemia remission, type 2 diabetes remission, micronutrient deficiencies, quality of life	Five years after surgery, BMI decreased by 22.2% in younger adolescents and 24.6% in older adolescents. Hypertension remission was 77% vs 67%, dyslipidemia remission 61% vs 58%, and type 2 diabetes remission 83% vs 87% in younger versus older adolescents. Quality of life improved significantly in both groups. Younger adolescents were less likely to develop elevated transferrin and vitamin D deficiency.	Observational (non-randomized) design; no non-surgical control group; predominantly female and White population may limit generalizability; some analyses based on small subgroup numbers (especially type 2 diabetes).
96	Lee et al., 2012 [108]	Retrospective comparative cohort study	55 adolescents aged 15–19 years with morbid obesity (LRYGB n = 32; LAGB n = 23)	Laparoscopic Roux-en-Y Gastric Bypass (LRYGB)	Laparoscopic Adjustable Gastric Banding (LAGB)	% Excess Weight Loss (EWL), diabetes remission, dyslipidemia resolution, nutritional deficiencies, complications, reoperations	LRYGB produced significantly greater weight loss than LAGB at all follow-up points. At 24 months, EWL was 83.4% after LRYGB versus 29.7% after LAGB (p < .01). All three adolescents with type 2 diabetes in the LRYGB group achieved remission. Dyslipidemia resolved in both LRYGB patients and in one of two LAGB patients. Complication rates requiring intervention were similar between groups.	Retrospective design; small sample size; substantial loss to follow-up (only 16% LRYGB and 30% LAGB patients available at 24 months); non-randomized treatment allocation; limited long-term outcome data.
97	Holterman et al., 2012 [109]	Prospective longitudinal cohort study	18 adolescents with severe obesity (11 morbidly obese; 7 superobese)	Laparoscopic Adjustable Gastric Banding (LAP-BAND)	Morbidly obese (BMI <50 kg/m2) vs superobese (BMI ≥50 kg/m2); 18-month follow-up	Weight loss, BMI change, insulin resistance (HOMA-IR), dyslipidemia, NAFLD/NASH, inflammatory markers (CRP, leptin, IL-6), quality of life	Both groups achieved similar weight loss (∼30 kg at 18 months) and BMI reduction (∼23% MO vs 21% SO). Resolution of insulin resistance was greater in MO adolescents (78%) than SO adolescents (29%). Dyslipidemia resolved in 67% of MO patients versus 17% of SO patients. Significant improvements in health-related quality of life occurred in both groups.	Small sample size (n = 18); single-center study; short-term follow-up (18 months); non-randomized design; findings may not be generalizable; limited power for subgroup analyses.
98	Inge et al., 2014 [110]	Prospective multicenter observational cohort (Teen-LABS)	242 adolescents ≤19 years with severe obesity; median BMI 50.5 kg/m2	Bariatric surgery (RYGB, VSG, AGB)	Procedure-specific comparison of perioperative outcomes	Major complications, minor complications, reoperations, readmissions, baseline comorbidities	No deaths occurred within 30 days. Major complications occurred in 8% of patients and minor complications in 15%. Reoperations and readmissions were largely procedure-related. Baseline prevalence of dyslipidemia (74%), sleep apnea (57%), hypertension (45%), and fatty liver disease (37%) demonstrated substantial obesity-related disease burden before surgery.	Observational design; follow-up limited to 30 days; no long-term weight-loss or cardiometabolic outcome data in this report; procedure selection was not randomized.
99	Varela et al., 2007 [111]	Retrospective comparative cohort study (administrative database analysis)	309 adolescents aged 12–18 years undergoing bariatric surgery	Gastric bypass, laparoscopic adjustable gastric banding, and laparoscopic gastroplasty	Adults undergoing bariatric surgery (n = 55,192)	Adults undergoing bariatric surgery (n = 55,192)	Adolescent bariatric surgery demonstrated a significantly lower overall 30-day complication rate than adults (5.5% vs 9.8%). No in-hospital mortality occurred among adolescents. Gastric bypass was the most common procedure. Restrictive procedures (LAGB/gastroplasty) had 0% morbidity and mortality in adolescents.	Administrative database study; limited clinical detail; no BMI or long-term outcome data; complications only captured during hospitalization; no assessment of weight loss or cardiometabolic improvement.
100	Price et al., 2020 [112]	Prospective longitudinal observational cohort (Teen-LABS secondary analysis)	88 adolescents (mean age 16.9 years) with severe obesity undergoing bariatric surgery; diabetes excluded	Vertical Sleeve Gastrectomy (VSG) or Roux-en-Y Gastric Bypass (RYGB) with objectively measured physical activity	Preoperative baseline vs. 6, 12, 24, and 36 months after surgery; comparison of more-active vs. less-active trajectories	Cardiometabolic: fasting insulin, HOMA-IR, fasting glucose. Anthropometric: BMI (used in adjustment). Cardiovascular: resting heart rate.	Greater physical activity after bariatric surgery was independently associated with significantly greater reductions in fasting insulin and HOMA-IR beyond weight loss alone. All more-active participants achieved and maintained normalization of insulin sensitivity through 36 months, whereas improvements declined over time in less-active participants. More-active participants also demonstrated greater reductions in resting heart rate.	Secondary analysis of the Teen-LABS cohort; relatively small sample (especially more-active group, n = 13); incomplete physical activity data for all original participants; no dietary intake data; limited resting heart rate follow-up beyond 6 months; potential self-selection bias.
101	Price et al., 2019 [113]	Prospective longitudinal observational cohort (Teen-LABS secondary analysis)	108 adolescents (mean age 17.0 years) with severe obesity undergoing bariatric surgery	Roux-en-Y Gastric Bypass (RYGB) or Vertical Sleeve Gastrectomy (VSG) with objectively measured physical activity	Preoperative baseline vs. postoperative follow-up (6 months, 1, 2, and 3 years); comparison between more-active and less-active physical activity trajectories	Cardiometabolic: LDL-C, non-HDL-C, HDL-C, triglycerides. Anthropometric: BMI, iliac waist circumference (IWC), weight-loss maintenance.	Greater physical activity after bariatric surgery was independently associated with significantly greater reductions in LDL-C and non-HDL-C, higher rates of dyslipidemia resolution, and approximately 11% greater BMI reduction at 3 years compared with less-active participants. HDL-C improvement was associated mainly with surgery rather than physical activity.	Secondary analysis of the Teen-LABS cohort; small more-active group (n = 14); limited StepWatch participation; relatively few VSG participants; no individual dietary intake data; potential self-selection bias; overlapping cohort with other Teen-LABS publications requiring careful handling during meta-analysis.
102	Khen-Dunlop et al., 2016 [114]	Prospective observational cohort (single-center)	49 adolescents with severe obesity (mean age 16.2 years; mean BMI 42.5 kg/m2)	Laparoscopic Adjustable Gastric Banding (LAGB) with multidisciplinary postoperative follow-up	Preoperative baseline vs. 6-, 12-, and 24-month postoperative follow-up	Anthropometric: weight, BMI, excess weight loss (EWL). Cardiometabolic: HOMA-IR, insulin resistance. Safety: band slippage, reoperations, explantation, mortality, other complications.	Significant reductions in weight and BMI were observed, with EWL increasing to 59.1% at 24 months. Insulin resistance prevalence declined from 100% to 17%, and HOMA-IR normalized by 24 months. Greater postoperative follow-up compliance (more clinic visits) was significantly associated with greater BMI reduction and EWL. Six patients (12%) required band removal, and one late postoperative death due to catastrophic aorto-gastric fistula was reported.	Single-center study with a relatively small sample size; no control group; follow-up limited mainly to 24 months for outcome analysis; evaluated only LAGB (less commonly used currently); postoperative compliance may reflect patient motivation, introducing potential confounding; rare but serious adverse event limits interpretation of long-term safety.
103	Burgos et al., 2008 [115]	Prospective observational cohort (single-center)	12 adolescents (≤18 years; mean age 16.3 ± 1.2 years) with morbid obesity (mean BMI 45 ± 6.7 kg/m2)	Open resection Roux-en-Y gastric bypass	Preoperative baseline vs. postoperative follow-up (mean 22.9 ± 8.4 months)	Anthropometric: BMI. Cardiometabolic: hypertension, dyslipidemia, type 2 diabetes, insulin resistance (HOMA), hyperinsulinemia. Safety: postoperative complications, incisional hernia, mortality, nutritional deficiencies.	Mean BMI decreased from 45.0 to 27.7 kg/m2 during follow-up. Hypertension, dyslipidemia, type 2 diabetes, and insulin resistance resolved or improved. Adolescents demonstrated normal growth after surgery. No perioperative mortality or early complications occurred; one patient (8.3%) developed an incisional hernia, and no clinically significant nutritional deficiencies were observed.	Very small adolescent sample (n = 12); single-center study; no control group; open surgical technique (less commonly performed today); relatively short follow-up (∼23 months); publication includes an elderly cohort, although adolescent outcomes are reported separately.
104	Chu et al., 2019 [116]	Prospective longitudinal observational cohort	20 adolescents (15 females; mean age 17.2 ± 0.8 years; mean BMI 48.7 ± 7.4 kg/m2) with severe obesity	Roux-en-Y Gastric Bypass (RYGB) or Sleeve Gastrectomy (SG)	Preoperative baseline vs. 12-month postoperative follow-up; secondary comparison between RYGB and SG	Anthropometric: BMI, body composition (fat mass, fat-free mass), weight loss. Metabolic: resting energy expenditure (REE), predicted REE, metabolic adaptation.	Measured REE decreased by approximately 548 kcal/day (≈25%) at 12 months, while predicted REE decreased by ~ 19%. Significant negative metabolic adaptation was observed beyond that expected from weight loss alone. REE reduction and metabolic adaptation were similar between RYGB and SG, suggesting metabolic adaptation may contribute to future weight regain regardless of procedure type.	Small sample size (n = 20); no non-surgical control group; body composition measured by bioelectrical impedance rather than DXA; incomplete follow-up data for some participants; relatively short follow-up (12 months), limiting assessment of long-term metabolic adaptation and weight regain.
105	Ippisch et al., 2008 [117]	Prospective longitudinal observational cohort	38 adolescents (13–19 years; mean age 16 ± 1 years) with morbid obesity (mean BMI 60 ± 9 kg/m2)	Laparoscopic Roux-en-Y Gastric Bypass (RYGB)	Preoperative baseline vs. postoperative follow-up (mean 10 ± 3 months)	Anthropometric: weight, BMI. Cardiometabolic/Cardiovascular: left ventricular mass index (LVMI), left ventricular hypertrophy (LVH), cardiac geometry, systolic and diastolic function, blood pressure, heart rate, rate-pressure product.	Mean BMI decreased from 60 ± 9 to 40 ± 8 kg/m2. Left ventricular mass index significantly decreased (54 ± 13 to 42 ± 10 g/m2·7), concentric LVH prevalence declined from 28% to 3%, normal cardiac geometry increased from 36% to 79%, diastolic function significantly improved (reduced mitral E/Ea ratio), and cardiac workload (rate-pressure product) decreased following bariatric surgery. Systolic function remained normal throughout follow-up.	Single-center study; relatively small sample (n = 38); only 38 of 67 eligible surgical patients had analyzable paired echocardiograms, introducing potential selection bias; short follow-up (∼10 months); no non-surgical control group; only RYGB evaluated, limiting generalizability to other bariatric procedures.
106	Mora, Diefenbach & Michalsky, 2022 [118]	Retrospective comparative observational study; Level III evidence	115 adolescents undergoing robotic-assisted VSG; mean age 17.2 ± 1.8 years; mean BMI 49.3 ± 7.5 kg/m2	Robotic-assisted vertical sleeve gastrectomy	BMI <50 kg/m2 (n = 64) vs BMI ≥50 kg/m2 (n = 51)	30-day complications, hospital readmission, postoperative vomiting, stricture, leak, hospital length of stay; also operative, access, and docking times	No significant difference in overall complications between BMI groups (p = .133); mean LOS was 1.8 days in both groups; major complications were rare
107	Pepper et al., 2016 [119]	Retrospective comparative cohort (single-center)	28 adolescents with severe obesity (14 robotic SG; 14 laparoscopic SG)	Robotic Sleeve Gastrectomy (RSG)	Laparoscopic Sleeve Gastrectomy (LSG)	Anthropometric: BMI, weight loss, % excess weight loss (%EWL), % total weight loss (%TWL). Perioperative/Safety: operative time, hospital length of stay (LOS), hospital charges, complications, readmissions, transfusion, re-intervention.	RSG and LSG produced similar early weight-loss outcomes at 30 days, with no significant differences in BMI reduction, %EWL, or %TWL. RSG had a longer operative time (132 vs. 100 min) but a shorter hospital stay (69.6 vs. 75.9 h). Hospital charges were significantly higher with RSG. No significant differences were observed in postoperative complications, readmissions, leaks, strictures, or wound infections, indicating that robotic SG was as safe and effective as conventional laparoscopic SG in the short term.	Retrospective design; very small sample size (n = 28); single-center and single-surgeon experience; only 30-day follow-up, preventing assessment of long-term weight loss or cardiometabolic outcomes; potential learning-curve bias due to early robotic experience; limited generalizability.
108	Kumar et al., 2018 [120]	Prospective longitudinal observational cohort (pilot study)	12 female adolescents with severe obesity (mean age 16.7 years; mean BMI 51.6 kg/m2)	Laparoscopic Roux-en-Y Gastric Bypass (RYGB)	Preoperative baseline vs. 6- and 12-month postoperative follow-up	Anthropometric: BMI, body weight. Cardiometabolic: HOMA-IR, high-sensitivity C-reactive protein (hs-CRP), leptin, total adiponectin, high-molecular-weight adiponectin. Novel biomarker: circulating spexin.	BMI decreased by 28.3% at 6 months and 33.8% at 12 months, with corresponding weight losses of 28.4% and 33.8%, respectively. Spexin concentrations increased significantly by 36.2% at 6 months and remained elevated at 12 months. Significant improvements were observed in insulin resistance (HOMA-IR ↓76.5% at 12 months), leptin, hs-CRP, and adiponectin. Higher spexin levels were associated with lower HOMA-IR independent of BMI change, suggesting a potential role of spexin in postoperative cardiometabolic improvement.	Pilot study with small sample size (n = 12); included only female participants, limiting generalizability; no nonsurgical control group; follow-up limited to 12 months; body composition, dietary intake, and physical activity were not assessed; potential overlap with other Cincinnati bariatric cohorts should be evaluated during systematic review data extraction.
109	DuCoin et al., 2015 [121]	Retrospective matched comparative cohort (single center)	14 adolescents aged 17–19 years (mean age 18.4 ± 0.8 years; mean BMI 47.2 ± 11.1 kg/m2) matched to 14 adults (>20 years)	Laparoscopic Roux-en-Y Gastric Bypass (LRYGB)	Matched adult patients (>20 years) with similar sex, BMI, comorbidities, and follow-up	Anthropometric: BMI, percentage excess weight loss (%EWL). Cardiometabolic: remission/improvement of diabetes mellitus, hypertension, hypercholesterolemia, obstructive sleep apnea. Safety: postoperative complications, revisional surgery for weight regain, readmissions.	Adolescents achieved 43.1%, 70.5%, 69.8%, and 54.8% EWL at 6, 12, 24, and 36 months, respectively, with no significant difference compared with matched adults. Approximately 80% of obesity-related comorbidities improved or resolved in the adolescent group. No major postoperative complications occurred during follow-up. One adolescent required revisional surgery for weight regain, similar to the adult cohort, supporting that LRYGB is safe and effective in late adolescents.	Small sample size (n = 14 adolescents); retrospective design; single surgeon and single-center experience; participants limited to older adolescents (17–19 years), reducing generalizability to younger adolescents; relatively short and variable follow-up; limited statistical power for comparisons.
110	Shehata et al., 2025 [122]	Retrospective longitudinal cohort (single-center)	63 adolescents (10–19 years; mean age 15.4 ± 2.39 years; mean BMI 48.1 ± 10.6 kg/m2) with severe obesity	Laparoscopic Sleeve Gastrectomy (LSG)	Preoperative baseline vs. postoperative follow-up at 1, 3, 5, and 7 years	Anthropometric: body weight, BMI, % total weight loss (%TWL), % excess weight loss (%EWL), % excess BMI loss (%EBMIL), waist/thigh/arm circumference, gastric volumetry. Cardiometabolic: type 2 diabetes (HbA1c), hypertension, dyslipidemia, obstructive sleep apnea (OSA), gastroesophageal reflux disease (GERD). Safety: postoperative complications, revisional surgery, weight regain, fertility outcomes.	Mean BMI decreased from 48.1 to 29.3 kg/m2 over 7 years. %TWL peaked at 32.9% (3 years) and remained 29.1% at 7 years, demonstrating durable weight loss despite modest regain. Significant remission of obesity-related comorbidities was observed: diabetes decreased from 34.9% to 11.1%, hypertension from 60.3% to 12.7%, OSA from 39.7% to 6.4%, dyslipidemia from 42.9% to 3.2%, and GERD from 25.4% to 4.8%. Overall complication rate was 7.9%, with 14.3% undergoing revisional surgery during long-term follow-up.	Retrospective, single-center study; modest sample size (n = 63); 14 patients lost to follow-up, introducing potential attrition bias; no non-surgical control group; psychosocial and quality-of-life outcomes were not assessed; fertility analysis was based on a very small subgroup (13 married participants), limiting generalizability.
111	Holterman et al., 2007 [123]	Prospective longitudinal cohort (FDA-approved IDE trial)	10 female adolescents aged 14–17 years with morbid obesity (mean BMI 50 ± 13 kg/m2)	Laparoscopic Adjustable Gastric Banding (LAGB)	Preoperative baseline vs. postoperative follow-up at 3, 6, and 9 months	Anthropometric: BMI, percentage excess weight loss (%EWL). Cardiometabolic: hypertension, insulin resistance (HOMA-IR), dyslipidemia, metabolic syndrome, nonalcoholic steatohepatitis (NASH). Psychological: Pediatric Quality of Life (PedsQL), Beck Depression Inventory (BDI). Safety: operative time, hospital stay, dehydration, pouch dilatation, port revision, readmission.	Mean %EWL reached 30 ± 16% at 9 months, with all participants losing weight. Hypertension resolved in 100% of affected patients, metabolic syndrome resolved in 80%, dyslipidemia resolved in 71%, and depression resolved in all three affected participants. Significant improvements were observed in quality of life, with minimal perioperative morbidity (two dehydration episodes, one pouch dilatation, one port revision, no major surgical complications or nutritional deficiencies).	Very small sample size (n = 10); all participants were female, limiting generalizability; short follow-up (9 months); no nonsurgical control group; single-center study; interim analysis of an ongoing FDA trial, creating a potential overlap with later publications from the same cohort.
112	Nadler et al., 2007 [124]	Prospective observational cohort (single-center; prospectively maintained database)	53 adolescents aged 13–17 years (41 females, 12 males; mean age 15.9 ± 1.3 years; mean BMI 47.6 ± 6.7 kg/m2) with morbid obesity	Laparoscopic Adjustable Gastric Banding (LAGB)	Preoperative baseline vs. postoperative follow-up (6, 12, 18, and 24 months)	Anthropometric: BMI, % excess weight loss (%EWL), body weight. Safety: operative time, hospital stay, band slippage, hiatal hernia, wound infection, hair loss, iron deficiency, nephrolithiasis, cholelithiasis, gastroesophageal reflux, reoperations.	Mean %EWL was 37.5% at 6 months, 62.7% at 1 year, and 48.5% at 18 months, demonstrating substantial short-term weight loss. There were no intraoperative complications or deaths. Overall operative complication rate was 9.4%, including two band slips requiring laparoscopic repositioning, two hiatal hernia repairs, and one wound infection. Nutritional complications were generally mild (hair loss and iron deficiency) and managed conservatively.	Single-center observational study; no nonsurgical control group; attrition reduced the number of participants available at later follow-up (53 at baseline, 18 at 1 year, 6 at 18 months, 2 at 2 years); limited reporting of cardiometabolic outcomes; relatively short follow-up for most participants; high potential overlap with other FDA trial/LAGB cohorts from the same period, requiring careful duplicate assessment.
113	Fischer et al., 2020 [125]	Observational cross-sectional follow-up analysis of the prospective FABS/FABS-5+ cohort	88 participants originally enrolled as adolescents with severe obesity; 58 underwent RYGB and 30 received non-operative weight management; mean follow-up ≈8 years	Roux-en-Y Gastric Bypass (RYGB)	Non-operative weight-management cohort	Anthropometric: BMI, percent BMI change from baseline. Behavioral: daily sitting time (IPAQ).	At approximately 8 years of follow-up, the surgical group achieved a mean 29% reduction in BMI, whereas the non-operative group experienced an 8% increase in BMI. Median daily sitting time was 5.1 h/day after RYGB versus 7.0 h/day in the non-operative group (difference not statistically significant). Across both groups, each additional hour of sitting per day was associated with a 1.4 percentage-point smaller BMI reduction, indicating that greater sedentary behavior was independently associated with poorer long-term weight outcomes.	Secondary analysis of the FABS/FABS-5+ cohort, creating a high risk of overlapping participants with previous FABS publications; cross-sectional assessment of sitting time at follow-up (no baseline sitting-time measurement); self-reported physical activity (IPAQ) introduces recall bias; predominantly White population and RYGB-only cohort limit generalizability; no cardiometabolic or safety outcomes were evaluated.
114	Liang et al., 2024 [126]	Single-center retrospective matched cohort study	30 pediatric/adolescent patients with severe obesity: 15 aged 11–13 years and 15 matched older adolescents aged 17–18 years; median baseline BMI 51.8 and 50.9 kg/m2, respectively	Laparoscopic vertical sleeve gastrectomy	Younger cohort aged ≤13 years vs matched older adolescent cohort aged 17–18 years; also preoperative vs postoperative comparisons	% excess BMI loss; % total body weight loss; HbA1c; total cholesterol; triglycerides; OSA/AHI; surgical complications; reoperation; 30-day readmission; hospital LOS	Both cohorts achieved substantial weight loss through 1 year. At 1 year, excess BMI loss was 54% vs 44%, and total body weight loss was 27% vs 20%. HbA1c significantly improved in both groups. No immediate postoperative complications or 30-day readmissions occurred.	Single-center retrospective design; small sample size (15 per group); follow-up limited to 1 year; missing follow-up data; very limited postoperative sleep-study data; some patients had not yet reached 1-year follow-up at manuscript submission; study period overlapped with COVID-19.
115	Hornack et al., 2019 [127]	Retrospective observational comparative cohort with longitudinal follow-up	63 adolescents with severe obesity (17 with CI/DD and 46 without CI/DD); mean age 17.7 years; mean baseline BMI 51.2 kg/m2	Laparoscopic Sleeve Gastrectomy (LSG)	Adolescents with CI/DD compared with adolescents without CI/DD, all undergoing LSG	Anthropometric: BMI, BMI change, percentage excess BMI loss (%EBMIL), percentage total weight loss (%TWL) at 3, 6, 12, and 24 months after surgery.	Weight loss after LSG was similar in adolescents with and without CI/DD. Presence of CI/DD or lower IQ did not significantly affect BMI reduction, %EBMIL, %TWL, or weight-loss trajectory during the first 24 months. Mean %EBMIL increased from approximately 37% at 3 months to 58% at 24 months, and latent growth modeling demonstrated that cognitive functioning was not a predictor of postoperative weight loss. These findings suggest that adolescents with CI/DD benefit from LSG to a similar extent as their peers without cognitive impairment.	Small CI/DD subgroup (n = 17); substantial loss to follow-up by 24 months (only 22% had 24-month data); single-center study; follow-up limited to 2 years; focused primarily on anthropometric outcomes without detailed cardiometabolic or long-term safety outcomes; potential overlap with other publications from the same bariatric program should be evaluated.
116	Schutte & Zitsman, 2022 [128]	Retrospective observational cohort (single-center)	142 adolescents with severe obesity undergoing LSG (mean age 16.0 ± 1.4 years; mean BMI 47.7 ± 9.1 kg/m2)	Laparoscopic Sleeve Gastrectomy (LSG)	Patients with intraoperative splenic ischemia (n = 44) versus those without splenic ischemia (n = 98)	Safety: incidence of splenic ischemia, division of segmental polar vessels, left shoulder pain, fever, leukocytosis, tachycardia, length of hospital stay, postoperative complications.	Splenic ischemia occurred in 31% (44/142) of adolescent LSG procedures. Division of segmental polar vessels increased the risk of splenic ischemia by 2.3-fold (p = .0055). However, splenic ischemia was not associated with increased postoperative fever, leukocytosis, tachycardia, shoulder pain, prolonged hospital stay, or other clinically significant complications, indicating that focal splenic ischemia is generally an incidental intraoperative finding without clinical consequence.	Retrospective single-center design; assessment relied on operative video quality, which may have underestimated ischemia in some patients; follow-up was limited to the inpatient postoperative period with no long-term assessment of splenic function; focused exclusively on safety outcomes without reporting weight-loss or cardiometabolic results; potential overlap with other publications from the same bariatric center should be evaluated during data extraction.
117	Furbetta et al., 2020 [129]	Retrospective longitudinal observational cohort (single-center; prospectively maintained database)	59 adolescents (≤19 years; mean age 17.7 ± 1.5 years; 64.4% female; mean BMI 40.9 ± 6.4 kg/m2) with morbid obesity	Laparoscopic Adjustable Gastric Banding (LAGB)	Preoperative baseline compared with postoperative follow-up at 1, 5, 10, and 15 years; no external comparator	Anthropometric: BMI, percentage excess weight loss (%EWL), percentage total weight loss (%TWL). Cardiometabolic: remission of diabetes, hypertension, sleep apnea. Psychological/Musculoskeletal: anxiety/depression, joint pain. Safety: postoperative complications, mortality, reoperations, band removal, conversion to other bariatric procedures.	LAGB resulted in sustained long-term weight loss, with mean %EWL of 42.6% at 1 year, 61.7% at 5 years, 48.1% at 10 years, and 55.8% at 15 years. Comorbidity resolution was excellent, with 100% remission of diabetes, 78% remission of hypertension, 69% remission of sleep apnea, 75% improvement of joint pain, and 57% remission of anxiety/depression. There was no perioperative mortality or major early complications. The overall reoperation rate was 30.5%, primarily because of port-related problems, while 13.5% required band removal and 3.4% underwent conversion to another bariatric procedure.	Retrospective single-center study; no nonsurgical control group; follow-up attrition increased over time (62.5% attendance at 15 years); relatively small number of patients at the longest follow-up (n = 5); reoperation rate was relatively high due to device-related issues; high likelihood of overlap with earlier publications from the same Italian LAGB cohort, requiring careful deduplication during evidence synthesis.
118	Tashiro et al., 2022 [130]	Retrospective comparative cohort (secondary analysis of a prospectively maintained institutional database)	328 adolescents aged 10–19 years with severe obesity (71.6% female; mean age 16.7 ± 1.88 years) undergoing LSG; 76 had a parental history of bariatric surgery and 252 did not	Laparoscopic Sleeve Gastrectomy (LSG)	(1) Parental history of bariatric surgery vs. no parental history; (2) Age <16 years vs. ≥16 years	Anthropometric: BMI, ΔBMI, %ΔBMI, % excess BMI loss (%EBMIL), weight class distribution at baseline, 3, 6, and 12 months	Adolescents with a parental history of bariatric surgery were heavier at presentation and at surgery, but achieved similar postoperative weight loss (%EBMIL, ΔBMI, and %ΔBMI) compared with those without a parental history. Likewise, adolescents <16 years and ≥16 years experienced comparable weight-loss trajectories through 12 months after LSG, supporting that younger age should not be considered a barrier to bariatric surgery. The authors suggest that earlier intervention may reduce long-term metabolic consequences of obesity.	Retrospective single-center study; follow-up attrition increased over time (12-month data available for only 51.8% of eligible participants); focused primarily on anthropometric outcomes without detailed cardiometabolic or complication analyses; did not evaluate parental weight-loss success or behavioral factors; high likelihood of participant overlap with previous Children's National Hospital LSG publications, requiring careful deduplication during evidence synthesis.
119	El Chaar et al., 2021 [131]	Retrospective observational registry cohort (MBSAQIP database)	1983 adolescents (≤19 years) with severe obesity undergoing primary bariatric surgery	Laparoscopic Sleeve Gastrectomy (LSG) and Laparoscopic Roux-en-Y Gastric Bypass (LRYGB)	LSG vs LRYGB; secondary comparison with adult bariatric patients (>19 years)	Safety outcomes: 30-day serious adverse events (SAEs), organ-space infection (OSI), reoperation, re-intervention, readmission, operative time, length of hospital stay	LSG demonstrated significantly lower rates of SAEs (2.9% vs 6.5%), re-intervention (1.2% vs 3.0%), and readmission (2.6% vs 5.6%) compared with LRYGB. Overall 30-day outcomes in adolescents were comparable to adults, supporting the short-term safety of bariatric surgery in adolescents.	Retrospective registry-based study; follow-up limited to 30 days; long-term outcomes unavailable; no adjustment for all potential confounders; limited procedural and institutional details; registry-dependent data quality.
120	Ryder et al., 2020 [132]	Prospective multicenter longitudinal observational cohort (Teen-LABS)	215 adolescents (<20 years) with severe obesity undergoing metabolic and bariatric surgery; compared with age-matched adolescents with normal weight, overweight, obesity, and severe obesity without surgery	Metabolic bariatric surgery (71% Roux-en-Y gastric bypass; 29% vertical sleeve gastrectomy)	Preoperative vs postoperative follow-up (1–5 years) and nonsurgical comparison cohorts (normal weight, overweight, obesity, severe obesity)	Anthropometric: BMI, weight; Cardiometabolic: predicted 30-year cardiovascular disease (CVD) event risk, systolic blood pressure, total cholesterol, HDL cholesterol, diabetes prevalence, antihypertensive medication use	Adolescents undergoing bariatric surgery experienced marked reductions in BMI (52.9–37.0 kg/m2 at 1 year), predicted 30-year CVD event risk (7.9%–4.0% at 1 year, sustained through 5 years), diabetes prevalence (13.5%–2.2%), systolic blood pressure, and improved HDL cholesterol. The estimated number needed to treat to prevent one CVD event over 30 years was 36.5, indicating substantial long-term cardiometabolic benefit.	Predicted CVD outcomes were model-based (Framingham 30-year risk) rather than observed cardiovascular events; the Framingham model was developed for adults and adapted for adolescents by assigning an age of 20 years; this publication represents a secondary analysis of the Teen-LABS cohort, so participant overlap should be considered during evidence synthesis.
121	Ballinger et al., 2025 [133]	Retrospective registry-based cross-sectional cohort analysis of MBSAQIP PUF data	2229 adolescents aged 10–17 years undergoing MBS; mean age 16.1 years; 69% female; mean BMI 47.25 kg/m2	Metabolic and bariatric surgery, predominantly sleeve gastrectomy, with some Roux-en-Y gastric bypass; laparoscopic and robotic approaches	Comparisons by age, year, sex, race, surgical approach, and comorbidity status; no nonsurgical comparator	30-day overall complications, dehydration, ED visits, readmission, reoperation, reintervention, SSI, DVT, ICU admission, GI bleeding, blood transfusion, mortality; baseline BMI and comorbidities	Overall 30-day complication rate was 2.9%; dehydration was the most common complication; ED visits occurred in 4.9%; mean readmission, reoperation, and reintervention rates were 2.4%, 0.4%, and 0.5%; no 30-day deaths occurred.	Retrospective registry analysis; only 30-day follow-up; no meaningful long-term weight-loss or comorbidity-remission outcomes; MBSAQIP centers may not represent all U.S. bariatric care; coding changes affected dehydration reporting from 2020 onward; potential data-entry errors; substantial potential for participant overlap with other MBSAQIP publications.
122	Kachornvitaya et al., 2025 [134]	Retrospective observational cohort (MBSAQIP registry, 2020–2023)	3056 adolescents (13–18 years) with severe obesity undergoing primary bariatric surgery	Sleeve Gastrectomy (SG) and Roux-en-Y Gastric Bypass (RYGB)	Adolescents with BMI ≥50 kg/m2 versus BMI <50 kg/m2	Cardiometabolic: diabetes, hypertension, obstructive sleep apnea; Safety: serious complications, anastomotic leak, postoperative bleeding, reoperation, non-operative intervention, readmission, pneumonia, AKI, VTE, mortality, operative time	Metabolic bariatric surgery was safe in adolescents with BMI ≥50 kg/m2, with no increase in 30-day serious complications compared with those with BMI <50 kg/m2 (1.3% vs 1.5%). Adolescents with BMI ≥50 kg/m2 had a higher burden of diabetes, hypertension, and obstructive sleep apnea, but no 30-day mortality. RYGB and non-insulin-dependent type 2 diabetes were independent predictors of serious complications.	Retrospective registry-based study; limited to 30-day postoperative outcomes; no long-term weight loss, comorbidity remission, nutritional, psychosocial, or quality-of-life outcomes; no center- or surgeon-level data; potential overlap with other MBSAQIP studies.
123	Lazzati et al., 2021 [135]	Nationwide retrospective descriptive cohort study using the French PMSI administrative database, 2008–2018	1027 adolescents aged <18 years; mean age 16.34 ± 0.99 years; 74.8% female; predominantly BMI ≥40 kg/m2	AGB (n = 549), SG (n = 298), gastric bypass (n = 179), BPD (n = 1)	Older adolescents aged 18–19 years and adults aged ≥20 years	30-day surgical/endoscopic reoperation, ICU admission, mortality, hospital LOS; long-term AGB removal and SG revision	In <18 adolescents, surgical reoperation was 1.0%, endoscopic reoperation 0.3%, ICU admission 0.2%, and mortality 0%; procedure-specific safety data are separately extractable; long-term revision outcomes were also assessed up to 10 years	Retrospective administrative database; no postoperative weight-loss or cardiometabolic-remission data; no outpatient clinical outcomes; no BMI percentiles for younger patients; nutritional deficiencies not captured; mixed bariatric procedures; potential overlap with other French PMSI studies; internal inconsistency in reported 10-year AGB removal percentages
124	Olbers et al., 2012 [136]	Prospective multicenter nationwide observational intervention study (AMOS)	81 adolescents aged 13–18 years with severe obesity (BMI ≥40 kg/m2 or ≥35 kg/m2 with comorbidities); mean age 16.5 years, mean baseline BMI 45.5 ± 6.1 kg/m2	Laparoscopic Roux-en-Y Gastric Bypass (RYGB)	Two matched comparison groups: 81 adults undergoing RYGB and 81 adolescents receiving conventional obesity treatment	Anthropometric: BMI, body weight, % weight loss, % excess BMI loss. Cardiometabolic: fasting glucose, insulin, HbA1c, triglycerides, HDL, LDL, CRP, blood pressure, liver enzymes. Quality of life: SF-36. Nutritional: iron, ferritin, vitamin B12, homocysteine. Safety: postoperative complications, reoperations, psychiatric events, nutritional deficiencies, adverse events.	At 2 years, adolescents achieved a 32% mean weight loss, with BMI decreasing from 45.5 to 30.2 kg/m2 and 93% achieving >50% excess BMI loss. Weight loss was similar to adults undergoing RYGB (31%) and significantly greater than the 3% weight gain observed in conventionally treated adolescents. Marked improvements occurred in insulin resistance, HbA1c, lipid profile, inflammatory markers, blood pressure, liver function, and quality of life. There was no mortality and 100% follow-up at 2 years. Adverse events occurred in 33% of patients, including 15% requiring additional surgical procedures (mainly internal hernia repair and cholecystectomy), psychiatric events in 7%, and nutritional deficiencies (particularly ferritin and vitamin B12) related to poor supplement adherence.	Non-randomized study; comparison groups were matched rather than randomized; biochemical and quality-of-life outcomes were unavailable for the comparison groups; relatively short follow-up (2 years) for assessment of long-term durability; psychiatric adverse events and poor adherence to vitamin supplementation emphasize the need for intensive long-term multidisciplinary follow-up; later AMOS publications report follow-up of the same cohort, so duplicate data extraction should be avoided.
125	Zeller et al., 2011 [137]	Prospective longitudinal observational cohort	16 adolescents (14–17 years) with severe obesity undergoing bariatric surgery	Roux-en-Y gastric bypass (RYGB)	Preoperative vs postoperative follow-up at 6, 12, 18, and 24 months	Anthropometric: BMI, % excess weight loss; Psychosocial: health-related quality of life (HRQOL), depressive symptoms, self-concept	Mean BMI decreased from 59.9 to 38.4 kg/m2 at 24 months (≈35% BMI reduction; 63% excess weight loss). Psychosocial functioning, HRQOL, depressive symptoms, and self-concept improved markedly during the first postoperative year, although some attenuation and mild weight regain occurred during the second year.	Small single-center cohort (n = 16); no nonsurgical control group; predominantly White participants; primary emphasis on psychosocial outcomes; limited generalizability and only 2-year follow-up.
126	Singendonk et al., 2018 [138]	Prospective observational cohort	15 adolescents (14–18 years) with severe obesity (BMI >40 kg/m2 or ≥35 kg/m2 with comorbidities)	Laparoscopic adjustable gastric banding (LAGB)	Preoperative vs 6-month postoperative assessments	Anthropometric: BMI, weight; Gastrointestinal: gastric emptying, esophageal motility, esophageal bolus transport, appetite hormones, hunger/fullness, dysphagia symptoms	At 6 months, LAGB resulted in significant weight and BMI reduction, increased postprandial fullness, decreased hunger, and increased dysphagia symptoms after solid food. There were no significant changes in gastric emptying, esophageal motility, esophageal bolus transport, esophageal emptying, or appetite-regulatory hormones.	Small sample size (n = 15); single-center study; short follow-up (6 months); primarily evaluated gastrointestinal physiology rather than long-term clinical outcomes; no nonsurgical control group.
127	Censani et al., 2014 [139]	Prospective observational cohort	135 adolescents (14–18 years) with severe obesity evaluated for bariatric surgery; 56 had 12-month postoperative follow-up data; 5 had MC4R mutations (4 underwent surgery)	Laparoscopic adjustable gastric banding (LAGB) or vertical sleeve gastrectomy (VSG)	Adolescents with MC4R mutations versus adolescents without MC4R mutations (including matched-control analysis)	Anthropometric: percentage excess weight loss (%EWL), BMI change, absolute weight loss, percent total weight loss	MC4R mutations were identified in 3.7% of adolescents. At 12 months, adolescents with heterozygous MC4R mutations achieved weight loss comparable to controls, with no statistically significant differences in %EWL, BMI change, or absolute weight loss. The findings suggest that heterozygous MC4R mutation carriers benefit from bariatric surgery similarly to non-carriers.	Small number of mutation carriers (n = 4 underwent surgery); substantial loss to follow-up (only 56/117 surgical patients had 12-month data); single-center study; predominantly restrictive procedures (mainly LAGB); limited statistical power to detect differences between groups.
128	de Boom et al., 2026 [140]	Randomized controlled trial (BASIC trial) with long-term follow-up	59 adolescents aged 14–16 years with severe obesity (BMI ≥40 kg/m2 or ≥35 kg/m2 with adiposity-related comorbidity)	Laparoscopic adjustable gastric banding (LAGB) plus multidisciplinary lifestyle intervention (MLI)	Multidisciplinary lifestyle intervention (MLI) alone	Anthropometric: BMI, weight, % total weight loss (%TWL); Safety: band removal, reoperation, revisional surgery, complications, band survival	LAGB produced greater weight loss during the first 2 years, but no significant long-term advantage over lifestyle intervention after a median 8.6 years. Band removal occurred in 52% of patients, mainly because of weight-loss failure or intolerance (pain, vomiting, dysphagia, reflux), with most subsequently converted to Roux-en-Y gastric bypass or sleeve gastrectomy. These findings demonstrate limited long-term durability of LAGB in adolescents.	High loss to follow-up (70% long-term follow-up achieved); self-reported long-term weights; crossover from control to surgery reduced between-group differences; high band removal rate; long-term comparison confounded by revisional surgery.
129	Roebroek et al., 2024 [14]	Randomized controlled trial (BASIC trial)	59 adolescents (14–16 years) with severe obesity who had inadequate response to ≥12 months of multidisciplinary lifestyle intervention	Laparoscopic adjustable gastric banding (LAGB) + multidisciplinary lifestyle intervention	Multidisciplinary lifestyle intervention (MLI) alone	Anthropometric: weight loss (%), BMI z-score, body weight; Cardiometabolic: fasting glucose, HbA1c, fasting insulin, HOMA-IR, lipid profile, blood pressure; Safety: postoperative complications	At 12 months, the surgery group achieved 11.2% mean weight loss, while the control group gained 1.7% body weight. Bariatric surgery significantly improved BMI z-score, fasting insulin, insulin resistance (HOMA-IR), and triglycerides, with only a small number of mostly minor postoperative complications.	Short-term follow-up (12 months); single-center trial; prolonged recruitment period (2011–2019); evaluated LAGB, a procedure now used less frequently; same cohort as the BASIC 8-year follow-up publication, so duplicate participants should be accounted for during evidence synthesis.
130	Fleming et al., 2025 [141]	Prospective multicenter observational longitudinal cohort	14 adolescents <19 years; median age 18.5 years; median baseline BMI 48.8 kg/m2	LAGB	Preoperative baseline; no nonsurgical control	BMI, HTN, T2D, dyslipidemia, micronutrients, WRQOL, general QOL, band removal, revision and conversion	Initial 10% BMI reduction at 1 year was not durable; no significant BMI improvement at 10 years; cardiometabolic risk and QOL were largely unchanged; 4/14 underwent band removal	Very small sample; limited statistical power; observational design; variable postoperative management; same underlying cohort as the previous 5-year Teen-LABS LAGB report
131	Inge et al., 2016 (Teen-LABS Consortium) [6]	Prospective multicenter observational cohort (Teen-LABS)	228 adolescents ≤19 years with severe obesity (161 RYGB, 67 SG); mean age 17 ± 1.6 years, mean BMI 53 kg/m2, 75% female	Roux-en-Y Gastric Bypass (RYGB) and Sleeve Gastrectomy (SG)	Baseline versus postoperative follow-up at 6 months, 1, 2, and 3 years; no nonsurgical control group	Anthropometric: weight, BMI, % weight loss. Cardiometabolic: type 2 diabetes, prediabetes, hypertension, dyslipidemia, kidney function. Quality of life: Impact of Weight on Quality of Life–Kids (IWQOL-Kids). Nutritional: ferritin, vitamin B12, vitamin A, vitamin D, folate. Safety: additional abdominal procedures, endoscopy, mortality.	At 3 years, mean weight loss was 27%, with BMI decreasing from 53 to 38 kg/m2. Remission occurred in 95% of participants with type 2 diabetes, 86% with abnormal kidney function, 76% with prediabetes, 74% with hypertension, and 66% with dyslipidemia. Weight-related quality of life improved significantly. However, 57% developed hypoferritinemia, 8% vitamin B12 deficiency, and 13% required one or more additional intra-abdominal procedures. There was one death (0.4%), occurring 3.3 years after RYGB in a participant with pre-existing type 1 diabetes due to complications of a hypoglycemic event.	Observational (non-randomized) design without a nonsurgical control group; relatively small SG subgroup (n = 67) limited procedure-specific comparisons; missing laboratory data for some participants; follow-up limited to 3 years for this report; high risk of overlap with subsequent Teen-LABS publications, requiring careful avoidance of duplicate participant inclusion during evidence synthesis.
132	Dewberry et al., 2020 [142]	Prospective multicenter observational cohort (Teen-LABS)	14 adolescents (mean age 18.2 ± 0.4 years) with severe obesity undergoing LAGB	Laparoscopic adjustable gastric banding (LAGB)	Preoperative vs postoperative follow-up through 5 years	Anthropometric: BMI, % total body weight loss, % excess weight loss; Cardiometabolic: dyslipidemia, elevated blood pressure, type 2 diabetes; Quality of life: IWQOL, SF-36; Safety: band removal, reoperations, complications; Micronutrients	Participants experienced initial weight loss but substantial weight regain by 5 years, with 43% exceeding their baseline BMI and 21% requiring band removal. Improvements were observed in dyslipidemia, elevated blood pressure, type 2 diabetes, and weight-related quality of life despite limited long-term weight loss, indicating poor long-term durability of LAGB in adolescents.	Very small sample (n = 14); no control group; only LAGB recipients; observational design; participant overlap with other Teen-LABS publications; findings may not generalize to current bariatric procedures (sleeve gastrectomy or Roux-en-Y gastric bypass).
133	Davidson et al., 2017 [143]	Prospective pre–post observational cohort	10 male adolescents with severe obesity (baseline mean age 17.4 ± 1.6 years) undergoing VSG	Vertical sleeve gastrectomy (VSG)	Baseline vs 1 year after surgery (lean adolescent reference group for laboratory comparison)	Anthropometric: weight, BMI; Cardiometabolic: HDL-C, LDL-C, triglycerides, fasting glucose, fasting insulin, HOMA-IR, systolic blood pressure; Mechanistic: HDL subspecies, cholesterol efflux capacity, HDL anti-oxidative function	At 1 year, BMI decreased by 32%, HDL-C increased by 23%, triglycerides decreased by 25%, fasting insulin decreased by 71%, HOMA-IR decreased by 75%, and systolic blood pressure improved. Surgery also restored large HDL particles and improved HDL function, suggesting improved cardiovascular risk profiles after VSG.	Very small sample (n = 10); males only; single-center study; no non-surgical control group; primary focus was HDL biology rather than clinical outcomes; possible participant overlap with other Cincinnati adolescent bariatric cohorts requiring duplicate assessment.
134	O'Brien et al., 2010 [12]	Randomized controlled trial (single-centre, Melbourne Australia; unblinded, parallel-group; ANZCTR12605000160639)	50 adolescents aged 14–18 years (mean 16.5–16.6y), BMI >35 (mean 40.4–42.3 kg/m^2, mean 99.2–99.3 BMI percentile); recruited from the community May 2005–Sept 2008; excluded intellectual disability (n = 3) and Prader-Willi syndrome (n = 1)	Laparoscopic adjustable gastric banding (LAP-BAND, Allergan), n = 25 (24 completed)	Supervised lifestyle intervention (individualized 800–2000 kcal/d diet,≥10,000 steps/day, structured exercise, behavioural modification, personal trainer), n = 25 (18 completed)	Primary: weight loss (>50% excess weight loss). Secondary: BMI, BMI z-score, waist circumference, metabolic syndrome (Jolliffe/Janssen criteria), HOMA insulin sensitivity and beta-cell function, blood pressure, lipids, quality of life (Child Health Questionnaire CF-50), adverse events	At 24 months (intention-to-treat for weight/BMI; completer's analysis for other outcomes, n = 24 banding/n = 18 lifestyle): banding group lost 34.6 kg (95% CI 30.2–39.0) vs 3.0 kg (95% CI 2.1–8.1) in lifestyle group; 84% vs 12% achieved >50% excess weight loss. Metabolic syndrome resolved in all 24 banding completers vs 4/18 lifestyle completers (P = 0.008 within banding group; between-group P = 0.025). Between-group differences at 24 months (banding minus lifestyle, mean [95% CI]): waist circumference −24.7 cm (−33.1 to −16.3, P = 0.001); systolic BP +7.8 mmHg (−4.5 to 20.1, P = 0.21); diastolic BP +0.9 mmHg (−5.9 to 7.7, P = 0.79); HOMA insulin sensitivity +74.6% (33–116, P = 0.001); triglycerides −20 mg/dL (−59 to 18, P = 0.29); HDL +5.4 mg/dL (3.5–14, P = 0.22). Quality of life (CHQ) improved in physical functioning, general health, self-esteem, family activities, and change-in-health domains for the banding group. Adverse events: 48% of banding group (12/25) had 13 events, 8 requiring revisional surgery in 7 patients (28%; proximal pouch dilatation x6, tubing injury x2), 1 cholecystectomy, 1 psychiatric admission, 2 pregnancies, 1 lost to follow-up; 44% of lifestyle group (11/25) had 18 events, including 1 participant with 8 hospital admissions (depression/intracranial hypertension), 1 cholecystectomy, 7 lost to follow-up (6 of the 7 withdrawals explicitly attributed to weight gain), 2 pregnancies.	JAMA 2010; 303(6):519–526, PMID 20145228. Single-centre, unblinded, no stratification/blocking; allocation-concealment procedure not described; baseline imbalance in systolic BP (122.0 vs 132.8 mmHg, P = 0.01) and HOMA beta-cell value (P = 0.048) favouring the lifestyle group at entry. Weight/BMI analyzed by intention-to-treat via mixed-effects model assuming noninformative dropout, but 6 of 7 lifestyle withdrawals were explicitly attributed to weight gain -- directly contradicting the noninformative-dropout assumption. All other outcomes (BP, lipids, insulin, HOMA, QoL) used completer's analysis only (n = 24/18), with no ITT sensitivity analysis and no adjustment for multiple secondary comparisons. RoB 2 overall high risk of bias for the 2-year BMI/weight-change result and for the completer-analyzed secondary cardiometabolic outcomes The between-group BMI difference used in the meta-analysis feasibility file (−11.4 kg/m^2) is a derived approximation combining the two groups' separately reported within-group changes and CIs (12.7 [11.3–14.2] vs 1.3 [0.4–2.9]); the article does not report a single pooled between-group BMI estimate with its own CI, so this remains an approximation, not a directly extracted figure. added following formal PICOS re-screening on Aug 04, 2026 (documented in Correction_Audit).
135	Raziel et al., 2014 [144]	Prospective single-centre pre-post cohort (uncontrolled); Assuta Medical Center, Tel Aviv, Israel	32 adolescents (20 female, 12 male) aged 14–18 years (mean 16.75y), operated Jan 2006–Dec 2011; mean weight 121.88 kg (range 83–178), mean BMI 43.23 kg/m^2 (range 35–54); all met Israel Ministry of Health guidelines for adolescent bariatric surgery; 34 comorbid conditions identified in 20/32 patients (62.5%): diabetes mellitus, hypertension, obstructive sleep apnea, hyperlipidemia, asthma, low back pain	Laparoscopic sleeve gastrectomy (LSG), sole bariatric procedure in all patients	Preoperative baseline vs postoperative follow-up (no nonsurgical control)	Anthropometric: %excess weight loss (%EWL) at 1, 3, 6, 9, 12, 24, 36, 48, 60 months. Comorbidity: resolution/improvement of diabetes, hypertension, OSA, hyperlipidemia, asthma, low back pain within 1 year. Safety: operative time, perioperative and later complications. Quality of life: BAROS questionnaire at 1–2 years.	Mean operative time 60 min; 2 complications (6.25%): 1 early staple-line leak (postoperative day 5, treated by re-resection/drainage) and 1 late acute cholecystitis (10 months, elective cholecystectomy); no deaths. Mean %EWL at 1/3/6/9/12/24/36/48/60 months: 27.9%/41.1%/62.6%/79.2%/81.7%/71%/75%/102.9%/101.6% (denominators declining from 29 patients at 1 month to 2 at 60 months). Of 34 comorbidities, 28 (82.4%) completely resolved and 6 (17.6%) improved within 1 year (diabetes 2/2 resolved; hypertension 5/7 resolved, 2 improved; OSA 2/2 resolved; hyperlipidemia 6/8 resolved, 2 improved; asthma 3/5 resolved, 2 improved; low back pain 10/10 resolved). BAROS obtained in 22/31 (71%) at 1–2 years: high self-satisfaction and social-activity scores; only mild increase in physical-activity score despite weight loss.	IMAJ 2014; 16:37–41. Full text reviewed Aug 04, 2026 (primary source). Single-centre, uncontrolled pre-post design; small sample with markedly declining denominators beyond 24 months (n = 8 at 24mo down to n = 2 at 60mo), limiting interpretation of longer-term %EWL figures; comorbidity resolution reported in aggregate across conditions rather than with denominator-specific risk-of-bias context; BAROS response rate 71% with no comparison of responders vs non-responders. Identified via full-text retrieval search on Aug 04, 2026 (previously logged as full text not found, Jul 13, 2026 list); added following formal PICOS re-screening (documented in Correction_Audit).
136	Dobbs et al., 2025 [145]	Prospective longitudinal single-arm pre-post observational study (IMPROVE-T2D study; NCT03620773)	14 adolescents aged 14–19 years (mean 16.8 ± 1.4y; 50% female; 78.6% Hispanic/Latino), antibody-negative youth-onset type 2 diabetes, severe obesity (BMI 46.7 ± 2 kg/m^2), diabetes duration 17.6 ± 13.8 months, age at diabetes diagnosis 15.9 ± 1.4y, baseline HbA1c 6.6 ± 0.2%; recruited from pediatric metabolic/bariatric surgery clinic, Children's Hospital Colorado	Vertical sleeve gastrectomy (VSG; ∼75–80% gastric resection)	Preoperative baseline vs 3 months postoperative (no nonsurgical control; within-subject pre-post)	Glycemic control (HbA1c, diabetes medication use), insulin sensitivity (HOMA-IR, Matsuda Index, Oral Minimal Model static/dynamic SI), insulin secretion (C-peptide-model insulin secretion rate, Disposition Index), anthropometric (weight, BMI, body composition via air-displacement plethysmography), cardiometabolic (lipids, liver enzymes, blood pressure, resting heart rate), resting energy expenditure/substrate oxidation, incretins (GLP-1, PYY, glucagon), safety/adverse events	At 3 months: weight −25.2 kg (19%, P < 0.001), BMI -8.7 kg/m^2 (18%, P < 0.001), fat mass −14.1 kg (P = 0.005), fat-free mass −4.5 kg (P = 0.034). HbA1c decreased 6.6 ± 0.2% to 5.7 ± 0.2% (P = 0.003); 12/14 (85.7%) no longer required diabetes medication. HOMA-IR improved 8.3 ± 1.5 to 2.2 ± 1.5 (P = 0.009); Matsuda Index 2.7 ± 0.4 to 3.5 ± 0.4 (P = 0.048); Disposition Index 0.14 ± 0.1 to 0.39 ± 0.1 (P = 0.027); insulin secretion rate AUC roughly doubled over the first 60 min of a mixed-meal tolerance test (P < 0.001). Total cholesterol, LDL, triglycerides, ALT, and AST all significantly improved; HDL unchanged. Resting heart rate fell from 94 ± 5 to 74 ± 5 bpm (P = 0.018). All statistically significant findings remained significant after adjusting for percent weight loss. Adverse events: 1 laparoscopic cholecystectomy for symptomatic cholelithiasis at ∼2 months; no hypoglycemia during MMTT or at home.	Journal of the Endocrine Society 2025; 9(3):bvaf020 (open access), Clinical Trials Registration NCT03620773. Full text reviewed Aug 04, 2026 (primary source). Very small sample (n = 14), not powered for sex-difference analyses; short follow-up (3 months only); no nonsurgical comparator group; authors note this is an early, weight-loss-independent-effect analysis and plan longer-term (1-year) follow-up of the same cohort -- watch for a later companion publication from IMPROVE-T2D reporting extended follow-up on the same participants, which would need to be treated as a companion report rather than an independent study. Closest verified match to your list entry "Vertical Sleeve Gastrectomy Improves Glycemia and Insulin Sensitivity in Youth with Type 2 Diabetes"; please do a final title check against your own record. Identified via full-text retrieval search on Aug 04, 2026 (previously logged as full text not found).
137	Beamish et al., 2023 [146]	Prospective observational cohort (AMOS study; 3 specialized adolescent obesity centers, Sweden); DXA body-composition sub-study with pre-post (baseline vs 1-, 2-, and 5-year) follow-up, no non-surgical comparator	Adolescent Morbid Obesity Surgery (AMOS) cohort, recruited 2006–2009. Eligibility: age 13–18 years (baseline mean 16.3 [SD 1.3] years); BMI ≥40 kg/m2, or ≥35 kg/m2 with comorbidity; Tanner stage ≥ III; attained peak height velocity; ≥1 year in a formal weight-loss programme. Exclusions: severe psychiatric disorder, ongoing drug abuse, syndromic/monogenic/brain-injury-related obesity. Of 100 adolescents assessed, 81 underwent RYGB; 9 exceeded the DXA scanner weight limit; 59 had a useable baseline DXA scan (41 female, 18 male) and form the analytic cohort for this report (n = 58 at 1 year, n = 56 at 2 years, n = 51 at 5 years).	Laparoscopic Roux-en-Y Gastric Bypass (RYGB)	None - single-arm pre-post design (baseline vs 1-, 2-, and 5-year postoperative follow-up); no non-surgical control group	Body composition by DXA (BMI, total fat mass, lean body mass, gynoid fat, android fat, subcutaneous adipose tissue [SAT], visceral adipose tissue [VAT]) at baseline, 1, 2, and 5 years; cardiometabolic laboratory measures (ALP, ALT, HbA1c, fasting insulin, LDL, HDL, triglycerides, hs-CRP) and blood pressure (systolic/diastolic) at the same time points; regression of percent change in adiposity depots against percent change in cardiometabolic risk factors.	BMI fell from 43.3 (SD 4.6) kg/m2 at baseline to 30.0 (4.2) kg/m2 at 1 year (−30.3%, 95% CI -32.5 to −28.0, P < 0.001) and regained to 31.6 (5.8) kg/m2 by 5 years (+5.5%, 95% CI 1.3 to 9.6, P = 0.011 vs year 1). VAT mass fell from 1.9 (0.81) kg to 0.64 (0.35) kg at 1 year (−65.1%, 95% CI -68.7 to −61.8, P < 0.001), the largest relative reduction of any depot, and regained to 0.73 (0.55) kg by 5 years (+25.3%, 95% CI 2.1 to 48.6, P = 0.034), remaining well below baseline. Total fat mass fell 49.3% at 1 year (P < 0.001) and regained 16.0% from year 1–5 (P = 0.004). Lean body mass fell 11.4% at 1 year (P < 0.001) and was the only depot without a statistically significant regain by 5 years (+1.2%, 95% CI 0.3 to 2.7, P = 0.105). Cardiometabolic measures improved from baseline to 1 year and were largely maintained at 5 years: HbA1c 5.3%→5.1%→5.2%; fasting insulin 210 → 50.1→63.6 pmol/L; LDL 101 → 75.9→85.1 mg/dL; HDL 43.5 → 56.0→58.9 mg/dL; triglycerides 119 → 81.4→82.0 mg/dL; hs-CRP 7.4 → 2.8→1.9 mg/L; systolic BP 125 → 117→112 mmHg; diastolic BP 76.4 → 70.7→69.8 mmHg (all mean, baseline/1-year/5-year). Correlation and regression analyses found minimal association between depot-specific adiposity change and cardiometabolic risk change: percent change in VAT at 1 year correlated only with triglycerides (slope 0.21 mg/dL/kg, P = 0.034) and fasting insulin (slope 44 pmol/L/kg, P = 0.027). In adjusted linear regression, each 1% reduction in VAT at 1 year was associated with a 0.43% reduction in fasting insulin (95% CI 0.05–0.82, P = 0.027); each 1% reduction in BMI was associated with a 0.24% reduction in systolic BP (95% CI 0.05–0.42, P = 0.011); each 1% reduction in total fat mass was associated with a 0.46% reduction in fasting insulin (95% CI 0.12–0.81, P = 0.008) and a 0.20% reduction in systolic BP (95% CI 0.04–0.37, P = 0.015). Authors conclude adiposity measures decreased substantially after RYGB but poorly predicted the magnitude of cardiometabolic risk-factor improvement.	Excluded the heaviest adolescents at baseline because they exceeded the DXA scanner's weight limit (9 of 81 post-RYGB patients); >25% of the analytic cohort either had unusable VAT scans (poor image quality) or did not return for 5-year scanning, despite no patient being formally lost to follow-up; predominantly Northern European ancestry, limiting transferability; small male subgroup (n = 18/59, 30.5%); findings should not be extrapolated to laparoscopic sleeve gastrectomy. COHORT-OVERLAP FLAG (confirmed from full text): this report explicitly identifies its population as the AMOS (Adolescent Morbid Obesity Surgery) study, recruited 2006–2009 at the same Swedish centers, and cites Olbers et al., 2012 (2-year AMOS outcomes) and Olbers et al., 2017 (5-year AMOS outcomes, Lancet Diabetes Endocrinol) as the parent cohort reports. If those AMOS publications are already included, this article must be treated as a companion/sub-study reporting a DXA body-composition subset of the SAME cohort (n = 59 of the larger AMOS sample), not as an independent participant sample - flag in the cohort map to avoid double-counting participants or effect sizes in any pooled analysis. Distinct from AMOS2 (Jarvholm et al., 2023), which is a separate randomised trial of surgery vs intensive non-surgical treatment. Full text obtained and reviewed (Surg Obes Relat Dis 2023; 19:1154–1161); this record supersedes the prior web-summary-derived version of this entry.
138	Gosher et al., 2026 [147]	Nationwide retrospective registry-based cohort study (pre-post follow-up)	460 adolescents aged <18 years at surgery (mean 16.9 ± 0.8 years) with severe obesity per Israeli Ministry of Health criteria (BMI≥40 kg/m2 with mild comorbidity or≥35 kg/m2 with severe comorbidity such as T2DM or severe OSA); mean baseline BMI 46.4 ± 6.8 kg/m2 (range 34.9–70.6); all 33 Israeli bariatric centers, October 2017–December 2024; Israel National Bariatric Surgery Registry (INBSR)	Primary metabolic/bariatric surgery: sleeve gastrectomy n = 403 (87.6%), one-anastomosis gastric bypass n = 44, adjustable gastric banding n = 7, Roux-en-Y gastric bypass n = 5, SADI-S n = 1	Preoperative baseline vs postoperative follow-up (6 months–4 years); descriptive comparison across procedure types	Anthropometric (%EWL by procedure and %TWL with denominators at 6 months, 1, 2, 3, and 4 years); safety outcomes (conversion/revisional surgery and reasons); baseline comorbidity prevalence (OSA, T2DM, MAFLD, dyslipidemia, hypertension)	Sleeve gastrectomy achieved mean %EWL of 56.3% at 6 months, 70.7% at 2 years, and 68.5% at 4 years; OAGB reached 80.2% at 3 years; whole-cohort %TWL was 30.9 ± 9.3% at 1 year (n = 196) and 29.9 ± 11.7% at 4 years (n = 90). Sixteen patients (all sleeve gastrectomy) required conversion/revisional surgery in adulthood, 81.3% for recurrent weight gain. Dietitian follow-up declined from 47.6% at 1 year to 14.5% at 4 years.	Retrospective registry analysis dependent on reporting accuracy; substantial long-term follow-up attrition; no postoperative comorbidity-remission or granular complication data; possible participant overlap with other Israeli registry-based reports.
139	Thareja et al., 2026 [148]	Retrospective cohort analysis of prospectively collected multicenter registry data (MBSAQIP Participant Use Files 2015–2023)	5432 adolescents aged≤19 years (median 17.0, IQR 16.0–18.0; 75% female) with severe obesity (median BMI 46 kg/m2, IQR 41–52); MBSAQIP-accredited centers, North America, 2015–2023; baseline comorbidities: OSA 22%, T2DM 16%, hypertension 6.7%	Sleeve gastrectomy n = 4672 (86%), Roux-en-Y gastric bypass n = 598 (11%), gastric band placement n = 68, gastric band revision n = 3, other n = 91	Internal comparisons across pandemic phases (pre-COVID, COVID, recovery) and racial/ethnic groups; no non-surgical comparator (not required)	Safety/adverse events (composite 30-day complications from 17 MBSAQIP-defined events, 30-day readmission, 30-day reoperation, length of stay); 30-day mortality not analyzable owing to registry coding inconsistency	Overall 30-day complication rate was 1.0% with median length of stay 1–2 days. Complications among Black adolescents declined from 3.9% pre-COVID to 0.8% during recovery (p = .004), although Firth-penalized analysis showed higher complication odds for Black vs White patients (OR 3.21, 95% CI 1.55–6.71); readmissions among White patients fell from 3.8% to 2.0% (p = .032). Sleeve gastrectomy use rose from 80% to 92% of adolescent cases and cohort composition shifted toward Hispanic and Black patients (p < .001).	Thirty-day outcomes only; no socioeconomic variables; sparse events precluded fully adjusted models; overlaps other MBSAQIP-based adolescent analyses (companion registry report - do not double-count in pooling).
140	Luna et al., 2026 [149]	Prospective observational cohort study (adolescent vs adult comparison; pre-post at 0, 6, and 12 months)	60 adolescents aged 15 years–19 years 11 months (Tanner stage ≥ IV; bone age≥13.5 y girls/≥14.5 y boys) with severe obesity defined as BMI-for-age≥99.9th percentile (baseline BMI 47.10 ± 8.35 kg/m2), compared with 60 adults aged 20–49 years (BMI≥40 kg/m2); private bariatric clinic, Rio de Janeiro, Brazil (UFRJ, ethics protocol 011/15)	Roux-en-Y gastric bypass in all participants, with standardized supplementation (vitamin D3 1800 IU/day, calcium carbonate 250 mg/day)	Adult RYGB group (developmental-stage comparison); within-group preoperative baseline	Anthropometric (BMI, waist circumference, waist-to-hip ratio at 0/6/12 months); safety/nutritional outcomes (25(OH)D and vitamin D inadequacy, ionized calcium/hypocalcemia, phosphorus/hypophosphatemia, PTH/secondary hyperparathyroidism), all reported separately for adolescents	Adolescent BMI decreased from 47.10 ± 8.35 to 28.61 ± 7.73 kg/m2 at 12 months (BMI reduction −18.49 vs −15.01 kg/m2 in adults, p = .002). Vitamin D inadequacy persisted in 78.3% of adolescents at 12 months despite supplementation; hypocalcemia (63.3% vs 0%) and hypophosphatemia (76.7% vs 10%) were more frequent in adolescents than adults (both p < .001), and secondary hyperparathyroidism reached 65.0% at 12 months (PTH 47.97–63.82 pg/mL).	Complete-case analysis restricted to participants with full follow-up and≥80% supplementation adherence (selection bias); single private clinic; recruitment period and mean ages not reported; possible companion reports from the same UFRJ cohort should be checked before extraction.
141	Safar et al., 2026 [150]	Retrospective single-center cohort study (pre-post, non-comparative)	26 adolescents who underwent surgery (median age at surgery 17 years, range 14–19; of 157 assessed and 38 referred within the CEASO transitional-care pathway) with severe obesity (referral criteria BMI >35 kg/m2 with major comorbidity or >40 kg/m2; median baseline BMI 46 kg/m2, range 37–88); McGill University Health Centre, Montreal, Canada, 2018–2024	Sleeve gastrectomy n = 25 (96%), Roux-en-Y gastric bypass n = 1; one planned second-stage single-anastomosis duodenal switch 11 months after sleeve gastrectomy	Preoperative baseline vs postoperative follow-up (median 37 months, range 14–93)	Anthropometric (%TWL and BMI change at 1, 2, and≥3 years); cardiometabolic (T2DM and hypertension remission, HbA1c, ALT); sleep (OSA remission/improvement); safety (30-day complications, readmission, reoperation, length of stay, de novo GERD, eating disorders)	Median %TWL was 28% (range 5–49) with BMI change of −14 kg/m2 at last follow-up (91% follow-up at≥3 years). T2DM remission occurred in 5/8 (63%; median HbA1c 7.1%–5.6%), hypertension remission in 4/6 (67%), and OSA remission in 5/15 (33%) with improvement in a further 6/15 (40%); median ALT fell from 89 to 26 U/L. There were no 30-day complications, readmissions, or reoperations; median stay was 24 h; de novo GERD occurred in 4/25 (16%) after sleeve gastrectomy.	Small single-center retrospective series (n = 26) dependent on clinical documentation; OSA remission assigned clinically (symptom-free, off CPAP) without postoperative polysomnography; nutritional and quality-of-life outcomes not assessed.
142	Breaux, 1995 [151]	Retrospective single-surgeon case series (pre-post) with sleep-apnea vs non-sleep-apnea subgroup comparison	22 children and adolescents aged 8–18 years (20/22 aged≥11 years; sleep-apnea group n = 11, mean age 14.2 years; non-apnea group n = 11, mean age 16.4 years, range 14–18) with clinically severe obesity: sleep-apnea group mean BMI 67.8 kg/m2 (range 51–105; 341% of ideal body weight), non-apnea group mean BMI 56.4 kg/m2 (range 41.3–94; 254% IBW); Cooper Green Hospital, Birmingham, Alabama, USA; operations 1983–1995	Vertical banded gastroplasty n = 5 (one later converted to RYGB), Roux-en-Y gastric bypass n = 14, biliopancreatic diversion n = 4 (BPD reserved for super-obese sleep-apnea patients)	Preoperative baseline vs postoperative follow-up (mean 32 months in the sleep-apnea group; mean 50 months in the non-apnea group); descriptive comparison of sleep-apnea vs non-apnea subgroups	Anthropometric (BMI, %EWL); sleep-related (resolution of sleep apnea, nocturnal low oxygen saturation on repeat sleep studies); safety outcomes (operative mortality, complications, nutritional deficiencies, reoperation/revision, late deaths, hospital stay)	Mean BMI decreased from 67.8 to 46.5 kg/m2 (45% EWL; 82% follow-up) in the sleep-apnea group at mean 32 months and from 56.4 to 35.5 kg/m2 (59% EWL; 100% follow-up) in the non-apnea group at mean 50 months. All nine sleep-apnea patients with documented postoperative sleep studies had resolution of sleep apnea, with mean nocturnal low oxygen saturation improving from 73% to 93%, achieved at a mean loss of 29.8 kg (27% of excess weight). There were no operative deaths; complications included protein deficiency in three BPD patients, vitamin A/D and folate deficiencies, one laryngeal edema requiring tracheostomy, and one incisional hernia; two late deaths (15 months and 3.5 years) were not attributed to the operations.	Dated single-surgeon series using partly obsolete procedures (VBG, open BPD); two participants aged 8–9 years are inseparably pooled with the adolescents (minor population deviation - flag for sensitivity analysis); the first two sleep-apnea patients underwent concomitant UPPP (co-intervention confounding OSA attribution); sleep-apnea resolution documented in 9/11 with follow-up; OCR-degraded historical scan; no cardiometabolic outcome data.
143	Zitsman et al., 2020 [152]	Prospective single-center longitudinal cohort study (FDA Investigational Device Exemption G050149; pre-post, 5-year follow-up) - companion report of the Columbia adolescent LAGB cohort (Zitsman 2011; Zitsman 3-year report, Surg Obes Relat Dis 2015; 11:101–9)	137 adolescents enrolled at ages 14–17 years (mean 16.0 ± 1.2 at enrollment; 17.0 ± 1.2 at surgery; enrollment bins include 13–14 y n = 11 and 18–19 y n = 5, all within 10–19 years; 94 female/43 male) with BMI >40 kg/m2 or >35 kg/m2 with serious comorbidity (obstructive sleep apnea, metabolic syndrome); mean baseline BMI 48.3 ± 8.2 kg/m2; Morgan Stanley Children's Hospital of New York-Presbyterian/Columbia University Irving Medical Center, USA	Laparoscopic adjustable gastric banding n = 137 (pars flaccida technique, single surgeon); 16 patients later converted to other procedures (13 sleeve gastrectomy, 3 RYGB)	Preoperative baseline vs postoperative follow-up through 60 months (weight change also compared with each patient's 6-month preoperative lifestyle program)	Anthropometric (weight, BMI, %EWL, %EBMIL through 60 months); cardiometabolic (remission/relapse of diabetes, insulin resistance, metabolic syndrome, hypertension, dyslipidemia, PCOS); sleep (OSA prevalence and remission); pulmonary (asthma prevalence and remission); safety (adverse events, reoperations, band removal, deaths)	Weight loss peaked at 36 months (%EWL 40.6 ± 35.2; %EBMIL 41.6 ± 34.9) and was %EWL 35.5 ± 39.4 and %EBMIL 36.6 ± 37.5 at 60 months; 23/127 (18%) achieved≥50% EWL. Comorbidity remission occurred in 85.5% for insulin resistance, 60% for diabetes, 46.7% for hypertension, 43.2% for asthma (19 patients), 38.8% for OSA (19/49), and 18.3% for dyslipidemia; weight loss >8.1 kg was associated with remission of dyslipidemia (OR 14.51), insulin resistance (OR 5.20), and asthma (OR 9.34). Complications requiring additional surgery occurred 80 times in 63 patients; 33/127 (26%) had undergone band removal by 5 years; three deaths occurred during the study period, none related to the band, surgery, or weight loss.	Substantial attrition (only 37 attended the 60-month visit; self-reported weights for many non-attenders); postoperative OSA and asthma status based on clinical report without routine repeat polysomnography; internal inconsistencies to reconcile at extraction (band removals 33 vs 36; reoperation rate 46% vs 50%); companion publication of the included Zitsman cohort - count participants once and use this report for the 60-month time point; publication year printed as 2019 (online first; resolve 2019 vs 2020 print citation).
144	Lee et al., 2026 [153]	Single-center retrospective cohort study (pre-post)	24 consecutive adolescents aged 19 years and younger (range 14–19; median 17; 19 female) with severe obesity per investigator-specified South Korean criteria (BMI >35 kg/m2, or >30 kg/m2 with≥1 severe adiposity-related comorbidity, after≥6 months of failed supervised lifestyle intervention); mean baseline BMI 40.5 kg/m2 (range 31.4–57.7); fatty liver disease 95.8%; Incheon Sejong Hospital, South Korea; 2020–2025	Laparoscopic sleeve gastrectomy, n = 24	Preoperative baseline vs postoperative follow-up (6 and 12 months); 100% follow-up	Anthropometric (%TWL, %EWL, %EBMIL, body composition, abdominal circumference); cardiometabolic/hepatic (triglycerides, HDL/LDL/total cholesterol, fasting glucose, fasting C-peptide, uric acid, AST/ALT, FibroScan CAP and LSM); safety (perioperative complications, operative time)	At 12 months, %TWL was 43.34% and %EWL 84.12%, with 51.1% fat-mass reduction. Triglycerides fell from 167.3 ± 131.3 to 67.9 ± 18.1 mg/dL (p = .001), fasting C-peptide from 4.0 ± 1.5 to 2.4 ± 1.0 (p = .009), and ALT from 49.4 ± 28.1 to 16.0 ± 18.5 U/L at 6 months (p < .001); hepatic CAP decreased from 337 to 246 dB/m and LSM from 7.1 to 5.1 kPa. No major postoperative complications occurred (0.0%).	Small single-institution retrospective cohort with 12-month follow-up and no control arm; Asian-adapted severe-obesity thresholds mean some individuals had BMI <35 with comorbidity (minimum 31.4) - flag for sensitivity analysis; %EWL/%EBMIL internally inconsistent with reported weights (prefer %TWL/BMI); printed C-peptide units (mg/dL) likely ng/mL; OSA reported at baseline only.
145	Rodrigues de Oliveira Filho et al., 2026 [154]	Multicenter retrospective cohort study (chart review) with age-group comparison	324 patients under 23 years undergoing sleeve gastrectomy at three US freestanding children's hospitals (Massachusetts General, Stanford, Yale; COSMIC collaborative), March 2013–September 2024. Adolescent subgroups fully extractable: <15 years n = 55 (mean 13.78 ± 0.93) and 15–17.99 years n = 168 (mean 16.80 ± 0.82); mean BMI at surgery 48 ± 9 and 49 ± 9 kg/m2 (ASMBS pediatric class II/III framework); the 18-23-year subgroup (n = 101) falls outside the review population	Laparoscopic sleeve gastrectomy, n = 324 (adolescent subgroups n = 223)	None (pre-post; internal comparison across age groups <15, 15–17.99, and 18+ years)	Anthropometric (weight, BMI, %BMI change from consultation to surgery, %BMI loss and %TWL at 6 and 12 months, proportion with <20% BMI loss); baseline comorbidity prevalence (OSA 71–78%, prediabetes ∼30%, T2DM 16–20%, hypertension 25–32%, hepatic steatosis 29–39%)	Patients <15 years gained +7.6% BMI from consultation to surgery versus +2% (15–17.99 y) and +1.2% (18+), p < .001, persisting in a GLP-1-free sensitivity analysis. Postoperative %BMI loss was similar across age groups at 6 months (17.8 ± 6.5 vs 19.3 ± 6.5) and 12 months (23.6 ± 9.6 vs 21.8 ± 9.3, p = .347); 12-month %TWL was 22 ± 9% (<15 y) and 21 ± 14% (15–17.99 y); 34–41% of each group achieved <20% BMI loss.	Retrospective design with ∼30% missing 12-month data; anthropometric domain only (no postoperative comorbidity or safety outcomes); no pubertal staging; possible participant overlap with prior single-center MGH/Stanford sleeve reports (Griggs 2023, cited) and MBSAQIP-based analyses - extract only the two adolescent age-group columns and cross-check overlap before pooling.
146	Pastore et al., 2022 [155]	Single-center observational pre-post study (prospectively collected samples/database)	24 adolescents aged 13–17 years (mean 15.3 ± 2.6; Tanner stage >1–2; 12 male/12 female) with severe obesity (BMI≥40 kg/m2; median baseline BMI 45.0 [IQR 42.5–50.5]) and biopsy-proven MAFLD after failed lifestyle intervention; Bambino Gesu Children's Hospital, Rome, Italy; evaluated 2014–2016	Laparoscopic sleeve gastrectomy (40F bougie), n = 24	None (pre-post: baseline vs 12 months)	Anthropometric (BMI, weight, %TWL); cardiometabolic/hepatic (fasting glucose, insulin, HOMA-IR, lipids, AST/ALT/GGT, paired liver histology: steatosis, ballooning, inflammation, fibrosis, NAS/NASH; TNF-alpha, IL-1beta, IL-6); nutritional/safety (total homocysteine, glutathione; postoperative complications narratively reported as absent)	At 12 months, median BMI fell from 45.0 to 31.0 kg/m2 (∼29% TWL, p < .0001) and HOMA-IR from 7.4 ± 3.9 to 2.5 ± 1.2 (p < .0001); HDL rose from 44 to 51 mg/dL. Histology improved substantially: NASH prevalence 50%–8.3%, severe steatosis to 0%, and F2 fibrosis 45.8%–8.7% (p = .004); ALT fell from 38.6 to 17.2 IU/L. Total homocysteine increased from 15.7 ± 4.1 to 21.1 ± 9.3 μmol/L (p = .0146), indicating mild post-LSG hyperhomocysteinemia requiring nutritional surveillance.	Small single-center sample without a control group; folate/B12 not monitored; 12-month follow-up; follow-up histology in 23/24; no formal adverse-event table. Almost certainly overlaps the Bambino Gesu LSG cohort (Manco 2017, Nobili 2018, Franchitto 2020 - same institution, surgeons, and 2014–2016 era); none of those is currently included, but if added, count shared outcomes once and use this report for its unique outcomes (homocysteine, glutathione, TNF-alpha).
147	Yodoshi et al., 2026 [156]	Prospective, nonrandomized, parallel-arm controlled cohort study (target-trial emulation with overlap-weighted analysis; ClinicalTrials.gov NCT02412540)	42 adolescents aged 12–19 years with severe obesity (BMI≥120% of the 95th percentile for age and sex; BMI 35–69 kg/m2) and biopsy-confirmed MASLD (NAS≥3): VSG n = 17 (median age 16.0 [14.0–17.0]; median BMI 51 [42–56]; %BMIp95 214) and comprehensive lifestyle intervention n = 25 (median BMI 38); Cincinnati Children's Hospital Medical Center, USA; consecutively enrolled 2015–2019; 52-week follow-up	Laparoscopic vertical sleeve gastrectomy, n = 17	Comprehensive lifestyle intervention (structured multidisciplinary program,≥26 contact hours over 52 weeks), n = 25	Cardiometabolic/hepatic (paired liver histology: NAS improvement, MASH resolution, fibrosis change, MASLD resolution; ALT, AST, GGT, HbA1c, T2DM resolution, lipids, MRI-PDFF, MRE stiffness); anthropometric (percent weight change, %BMIp95, BMI z-score); safety (deaths, gastric leak, 30-day minor complications, hospitalization, emergency visits)	In overlap-weighted adjusted analyses, VSG versus lifestyle intervention showed risk differences of 0.68 (95% CI 0.45–0.91) for NAS improvement, 0.36 (0.12–0.59) for fibrosis improvement, and 0.49 (0.26–0.72) for complete MASLD resolution, with mean differences of −19.6% (−30.0 to −9.1) for weight change, −54.0 U/L for ALT, and −8.0% for MRI-PDFF. Among completers, MASH resolved in 8/8 after VSG versus 4/11 with lifestyle intervention (p = .01), and T2DM resolved in 5/6 versus 0/1. One VSG participant (6%) had a major perioperative complication (gastric leak); minor 30-day complications 29% vs 0% (p = .007); no deaths.	Nonrandomized treatment selection with marked baseline imbalance disfavoring VSG (median BMI 51 vs 38; T2DM 41% vs 16%) - prefer the overlap-weighted adjusted estimates; 36% missing week-52 paired histology (COVID-related attrition); single center; enrollment 2015–2019 postdates Teen-LABS with no stated cross-cohort membership (Inge/Helmrath co-authors)

Where reported, adolescent sample sizes ranged from 10 to 5432 participants. Participants were generally aged 10–19 years with severe obesity, most commonly defined as BMI ≥40 kg/m² or ≥35 kg/m² with an adiposity-related comorbidity. Baseline BMI values were generally approximately 40–60 kg/m². Female participants predominated in many larger cohorts, although sex distribution was not consistently reported.

Procedures included sleeve gastrectomy, Roux-en-Y gastric bypass, adjustable gastric banding, one-anastomosis gastric bypass, biliopancreatic diversion with or without duodenal switch, single-anastomosis duodeno-ileal bypass with sleeve gastrectomy, and older gastroplasty procedures. Comparators included preoperative status, non-surgical management, alternative bariatric procedures, and other participant groups. Follow-up ranged from perioperative or 30-day assessment to approximately 15 years.

Anthropometric and safety outcomes were most frequently reported, followed by cardiometabolic outcomes. Sleep-related outcomes, principally obstructive sleep apnea, were less frequently reported, and objective pulmonary outcomes were uncommon [6], [7], [136].

Metabolic or bariatric surgery was the intervention of interest; routine perioperative medications and nutritional supplements were not considered separate interventions unless specifically evaluated or relevant to a safety outcome. Details of procedures, comparators, devices, medications, and supplementation regimens are provided in Table 1 where reported.

3.3. Risk of bias within studies

Among the six randomized-trial reports assessed using RoB 2, one (16.7%) was judged at low overall risk of bias, one (16.7%) raised some concerns, and four (66.7%) were at high risk. Four reports originated from the BASIC trial, with the remaining two from AMOS2 and an Australian trial of adjustable gastric banding. Missing outcome data was the principal concern, with four reports rated at high risk in this domain. Other concerns included incomplete reporting of allocation concealment, baseline imbalance, complete-case analyses, substantial or differential attrition, lack of reported assessor blinding for imaging outcomes, and unavailable prespecified analysis plans for secondary outcomes. Outcome-specific judgments are presented in Table 2, with domain-level distributions shown in Fig. 2A–B.

Table 2.

Appraisal tool assigned to each of the 147 included reports.

Appraisal tool	Reports (n)	Per cent	Basis for selection	Table
RoB 2 (Cochrane risk-of-bias tool for randomized trials, version 2)	6	4.1	Randomized trial reports; the effect of assignment to intervention was appraised for one nominated outcome per report.	2A
ROBINS-I (Risk Of Bias In Non-randomized Studies of Interventions)	19	12.9	Non-randomized studies making a between-intervention comparison, whether surgery against non-surgical care or one procedure against another.	2B
JBI Critical Appraisal Checklist for Cohort Studies	117	79.6	Longitudinal single-arm, pre-post, or registry cohorts in which no between-intervention effect estimate was the object of appraisal.	2C
JBI Critical Appraisal Checklist for Case Series	3	2.0	Uncontrolled descriptive series of consecutive cases with no defined comparison group.	2D
JBI Critical Appraisal Checklist for Analytical Cross-sectional Studies	2	1.4	Exposure and outcome ascertained at a single time point, with no longitudinal follow-up of the analyzed sample.	2E
Total	147	100.0

Fig. 2.

Summary of risk-of-bias assessments: (A) Cochrane RoB 2 assessments of six randomized-trial reports; (B) ROBINS-I assessments of 19 comparative non-randomized results.

Report-level tool assignments for all 147 reports are listed in the Tool Assignment sheet of the accompanying risk-of-bias workbook. Where a randomized trial contributed more than one report, each report was appraised separately for its own nominated outcome and time point rather than once at trial level.

D1 bias arising from the randomization process; D2 bias due to deviations from intended interventions; D3 bias due to missing outcome data; D4 bias in measurement of the outcome; D5 bias in selection of the reported result.

D1 confounding; D2 classification of interventions; D3 selection of participants into the study; D4 deviations from intended interventions; D5 missing data; D6 measurement of outcomes; D7 selection of the reported result. L = low, M = moderate, S = serious, C = critical. Footnote marks carried forward from the previous workbook version: * Zhu 2023 [41], missing-data domain; † Henfridsson 2019 [59], outcome-measurement domain; ‡ Guan 2023 [61], outcome-measurement domain.

Eighty-seven reports (74.4%) were judged at moderate-to-high or high risk of bias. The report-level listing, with the design, sample and principal concerns for each of the 117 cohorts, is given as Supplement 3 and in Table 2C JBI Cohort sheet of the risk-of-bias workbook.

Q1 clear inclusion criteria; Q2 condition measured in a standard, reliable way for all participants; Q3 valid methods for identification of the condition; Q4 consecutive inclusion; Q5 complete inclusion; Q6 demographics reported; Q7 clinical information reported; Q8 outcomes or follow-up results reported; Q9 presenting site demographics reported; Q10 statistical analysis appropriate. Item-level rationales are held in the risk-of-bias workbook.

Strauss et al., 2001. Consecutive and complete institutional inclusion with detailed per-patient reporting, offset by partly self-reported follow-up anthropometry, highly variable follow-up intervals and purely descriptive analysis in a very small series.

Franco et al., 2017. Well-specified eligibility, standardised measurement and appropriate repeated-measures analysis are undermined by exclusion of participants lacking follow-up, an inconsistent baseline denominator, sparse demographic reporting and severe attrition by 24 months.

Breaux, 1995. Consecutive single-surgeon inclusion with full complication reporting, but no prespecified eligibility criteria, mixed anthropometric reference standards, incomplete sleep-apnoea follow-up, obsolete and heterogeneous procedures, and descriptive analysis with an unreconciled baseline BMI discrepancy.

Q1 inclusion criteria clearly defined; Q2 subjects and setting described in detail; Q3 exposure measured validly and reliably; Q4 objective, standard criteria used for measurement of the condition; Q5 confounding factors identified; Q6 strategies to deal with confounding stated; Q7 outcomes measured validly and reliably; Q8 appropriate statistical analysis. Item-level rationales are held in the risk-of-bias workbook.

Fischer et al., 2019. Every checklist item is satisfied, but the overall rating is held at moderate because the exposure was measured only once, concurrently with the outcome, so temporality cannot be established; sitting time is self-reported; the arms differ at baseline in age, race and BMI; and the sample overlaps other FABS/FABS-5+ publications.

Ballinger et al., 2025. All eight items are satisfied within a large audited registry with prespecified outcome definitions and adjusted analyses; residual concern relates to coverage restricted to accredited centers, the mid-series change in the dehydration definition, the 30-day outcome ceiling and probable participant overlap with other MBSAQIP analyses.

All 19 results assessed using ROBINS-I had substantial methodological limitations: 11 (57.9%) were at serious overall risk of bias and eight (42.1%) at critical risk. Confounding was the predominant concern, with all studies rated serious or critical, while participant selection, deviations from intended interventions, and missing data also frequently contributed to higher risk. Intervention classification was generally well handled, whereas outcome measurement and selective reporting raised fewer but persistent concerns. Common limitations included non-random treatment selection, inadequate control of baseline prognostic differences, selection based on follow-up availability, differential co-interventions, attrition, completer-only analyses, and selective reporting. Domain-level and overall judgments are presented in Table 2 and summarized in Fig. 2B.

The five studies assessed using other design-specific JBI tools also had methodological limitations. The two analytical cross-sectional studies were rated from low-to-moderate to moderate risk of bias, and the three case series from moderate to moderate-to-high risk. Principal concerns included cross-sectional temporality, self-reported or incompletely standardized measurements, residual confounding, small uncontrolled samples, heterogeneous procedures, and incomplete or variable follow-up. Detailed risk-of-bias assessments are presented in Table 2A, Table 2B, Table 2C, Table 2D, Table 2E.

Table 2A.

RoB 2 judgments for the six randomized-trial reports.

Trial report (outcome; time point)	Trial	D1	D2	D3	D4	D5	Overall
Talib et al., 2025 [15] Hepatic steatosis measured using HRI (1 year)	BASIC	Some concerns	Some concerns	High	Some concerns	Some concerns	High
Talib et al., 2025 [32] Left ventricular mass index (1 year)	BASIC	Some concerns	Low	Some concerns	Some concerns	Some concerns	Some concerns
Järvholm et al., 2023 [13] Change in BMI (2 years)	AMOS2	Low	Low	Low	Low	Low	Low
de Boom et al., 2026 [140] Long-term BMI and weight loss (Median 8.6 years)	BASIC	Some concerns	Some concerns	High	Some concerns	Some concerns	High
Roebroek et al., 2024 [14] Percentage total bodyweight change (12 months)	BASIC	Some concerns	Some concerns	High	Low	Low	High
O'Brien et al., 2010 [12] Change in BMI/body weight (primary) and secondary cardiometabolic outcomes (24 months)	O'Brien RCT	Some concerns	Some concerns	High	Some concerns	Some concerns	High

Table 2B.

ROBINS-I judgments for the 19 comparative non-randomized results.

No.	Study	Outcome-level result assessed	D1	D2	D3	D4	D5	D6	D7	Overall
1	Perera et al., 2026 [19]	UACR trajectory through 36 months: sleeve gastrectomy plus VLCD versus VLCD/lifestyle support	S	L	S	S	M	M	M	Serious
2	Inge et al., 2018 [37], PCORnet	Percentage BMI change at 1 and 3 years: RYGB versus sleeve gastrectomy	S	L	S	M	S	M	M	Serious
3	Blumenfeld et al., 2020 [38]	One-year BMI change: sleeve gastrectomy versus adjustable gastric banding	S	L	C	M	C	L	M	Critical
4	Zhu et al., 2023 [41]	Type 2 diabetes remission at 2 years: bariatric surgery versus medical treatment	S	L	S	M	L*	M	M	Serious
5	Inge et al., 2018 [42], Teen-LABS/TODAY	Change in HbA1c from baseline to 2 years: surgery versus intensive medical therapy	S	L	S	M	M	L	M	Serious
6	Mizrahi Reuveni et al., 2026 [49]	Five-year BMI trajectory: sleeve gastrectomy versus nutritional intervention	S	L	S	S	S	M	M	Serious
7	Ryder et al., 2021 [52]	Estimated 30-year cardiovascular-event risk at 5 years	S	L	S	M	M	S	M	Serious
8	Henfridsson et al., 2019 [59]	Five-year energy intake and body-composition differences: RYGB versus non-surgical care	S	L	C	S	S	S†	M	Critical
9	Guan et al., 2023 [61]	Five-year BMI difference: metabolic bariatric surgery versus lifestyle intervention	S	L	M	S	L	L‡	M	Serious
10	de la Cruz-Muñoz et al., 2010 [62]	Four-year BMI change after bariatric surgery; implicit no-surgery comparator	C	L	S	S	C	L	M	Critical
11	Howard et al., 2022 [65]	Five-year all-cause hospitalization: sleeve gastrectomy versus RYGB	S	L	M	M	S	M	M	Serious
12	Olbers et al., 2017, AMOS [7]	Five-year BMI change: adolescent RYGB versus conservative treatment	S	L	C	C	M	L	L	Critical
13	Pedroso et al., 2015 [94]	Percentage excess weight loss at 24 months: sleeve gastrectomy versus gastric banding	S	L	S	S	C	L	M	Critical
14	Qureshi et al., 2021 [95]	Percentage body-weight change at 24 months: sleeve gastrectomy versus no surgery within 6 months	S	M	S	S	C	L	M	Critical
15	Zolfaghari et al., 2025 [101]	Twelve-month total cholesterol: sleeve procedure versus gastric bypass	S	L	M	M	S	L	S	Serious
16	Cozacov et al., 2014 [102]	Latest postoperative BMI: sleeve gastrectomy versus RYGB	S	M	S	S	S	M	M	Critical
17	Lee et al., 2012 [108]	Percentage excess weight loss at 24 months: RYGB versus adjustable gastric banding	S	L	S	S	C	L	M	Critical
18	Pepper et al., 2016 [119]	Postoperative hospital length of stay: robotic versus conventional laparoscopic sleeve gastrectomy	S	L	M	L	L	L	M	Serious
19	Yodoshi et al., 2026 [156]	NAS improvement, MASH resolution, fibrosis change and percent weight change at 52 weeks: vertical sleeve gastrectomy versus comprehensive lifestyle intervention (overlap-weighted target-trial emulation)	S	L	M	M	S	M	L	Serious

Table 2C.

JBI cohort appraisals distribution across the 117 cohort reports.

JBI overall judgment	Reports (n)	Per cent	Typical profile
Low	2	1.7	Prospective multicentre cohorts with protocol-defined outcome ascertainment, near-complete follow-up and adjusted analysis.
Low-to-moderate	7	6.0	Large accredited-registry analyses of standardized short-term outcomes, with residual confounding and clustering the main open questions.
Moderate	21	17.9	Prospective or registry cohorts with a defined follow-up schedule but no comparator, or with partial outcome ascertainment.
Moderate-to-high	50	42.7	Single-centre or retrospective cohorts without a comparator, with complete-case selection, modest samples, or probable participant overlap with companion reports.
High	37	31.6	Very small samples, unclear or severe attrition, subjective or self-reported outcomes, and no confounder control; long-term denominators frequently collapse.
Total	117	100.0

Table 2D.

JBI case-series appraisals (three reports, ten items).

Report	Design and sample	Q1	Q2	Q3	Q4	Q5	Q6	Q7	Q8	Q9	Q10	Overall
Strauss et al., 2001 [63]	Retrospective single-surgeon RYGB case series; n = 10; ages 15–17 years; operations 1985–1999; Robert Wood Johnson Medical School/St Peter's University Hospital, New Brunswick, NJ, USA	Yes	Unclear	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Unclear	Moderate
Franco et al., 2017 [93]	Retrospective descriptive LSG case series; n = 22 analyzed; ages 14–19 years; operations 2007–2014; Instituto da Crianca, University of Sao Paulo, Brazil	Yes	Yes	Yes	Unclear	No	Unclear	Yes	Yes	Yes	Yes	Moderate-to-high
Breaux, 1995 [151]	Retrospective single-surgeon case series (VBG, RYGB, BPD); n = 22; ages 8–18 years; operations 1983–1995; Cooper Green Hospital, Birmingham, AL, USA	Unclear	Unclear	Yes	Yes	No	Yes	Yes	Yes	Yes	No	Moderate-to-high

Table 2E.

JBI analytical cross-sectional appraisals (two reports, eight items).

Report	Design and sample	Q1	Q2	Q3	Q4	Q5	Q6	Q7	Q8	Overall
Fischer et al., 2020 [125]	Observational cross-sectional analysis nested in the prospective FABS/FABS-5+ cohort; n = 88 (RYGB 58, non-operative 30); single follow-up visit at a mean of 8 years	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Moderate
Ballinger et al., 2025 [133]	Retrospective registry-based cross-sectional analysis of MBSAQIP Participant Use Files 2017–2022; n = 2229 adolescents aged 10–17 years; 30-day outcome window	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Low-to-moderate

Among the 117 cohort reports assessed using the JBI Cohort Checklist, two (1.7%) were categorized as low risk, seven (6.0%) as low-to-moderate risk, 21 (17.9%) as moderate risk, 50 (42.7%) as moderate-to-high risk, and 37 (31.6%) as high risk; thus, 87 reports (74.4%) were rated moderate-to-high or high risk. Common concerns included absence of an appropriate comparator, limited group comparability, small or selected samples, inadequate control of confounding, incomplete follow-up and missing data, and reliance on surrogate or self-reported outcomes. Findings from studies with greater methodological limitations were therefore interpreted cautiously. Study-level assessments are presented in Table 2, with detailed assessments provided in Supplement 3.

3.4. Cardiometabolic outcomes

Cardiometabolic outcomes were reported across six randomized reports (AMOS2, four BASIC publications, and one Australian banding trial), comparative observational studies, and uncontrolled cohorts, with follow-up from approximately 10 months to 10 years (Table 3). Across randomized trials, surgery improved glucose and insulin indices, lipids, blood pressure, and inflammatory markers, although individual trials were small and underpowered for secondary outcomes. In the O'Brien trial, banding improved insulin sensitivity and waist circumference relative to supervised lifestyle treatment, but the evidence was at high risk of bias from informative dropout.

Table 3.

Cardiometabolic outcomes by procedure and follow-up.

Study	Design	Procedure	n	Follow-up	Key cardiometabolic findings
Järvholm 2023 [13] (AMOS2)	RCT	MBS (mostly RYGB) vs non-surgical	50	2 y	Greater improvement in glucose/insulin indices, lipids, BP, hs-CRP, ALT; underpowered for individual secondary outcomes
Roebroek 2024 [14](BASIC)	RCT	LAGB + lifestyle vs lifestyle	52	12 mo	↓ fasting insulin, HOMA-IR, triglycerides
O'Brien 2010 [12]	RCT	LAGB vs supervised lifestyle	50	24 mo	↑ insulin sensitivity (HOMA +74.6%, 95% CI 33–116, P = 0.001); ↓ waist −24.7 cm (P = 0.001); BP and triglycerides NS
Inge 2018 [42](Teen-LABS vs TODAY)	Comparative NRS	RYGB/VSG vs medical	NR	2 y	HbA1c 6.8 → 5.5% (surgery) vs 6.4 → 7.8% (medical); HbA1c<6.5% 94% vs 38%; BP, dyslipidemia, kidney improved
Zhu 2023 [41]	Comparative NRS	Surgery vs medical	NR	≤2 y	T2D remission 76.1% vs 6.5%
Perera 2026 [19]	Comparative NRS	SG + VLCD vs lifestyle	7 (surg)	36 mo	HbA1c 8.1 → 5.8% (surgery); albuminuria ↓ ∼83–86%; eGFR stable
Inge 2016 [6] (Teen-LABS)	Prospective cohort	RYGB/VSG	242	3 y	Remission: T2D 95%, prediabetes 76%, HTN 74%, dyslipidemia 66%, kidney 86%
Olbers 2012 [136] (AMOS)	Prospective cohort	RYGB	81	2 y	Improved triglycerides, HDL/LDL, BP
Olbers 2017 [7] (AMOS)	Prospective cohort	RYGB	81	5 y	Sustained glucose/HbA1c/insulin-resistance improvement; all baseline T2D in remission
El-Matbouly 2018; Elhag 2021 [17], [44]	Retrospective cohort	SG	NR	5–9 y	Sustained improvement/remission of prediabetes and T2D
Davidson 2017 [143]	Prospective	VSG (male)	10	1 y	↑ HDL 23%, ↓ TG 25%, ↓ fasting insulin 71%, ↓ HOMA-IR 75%
Dewberry 2020 [142] (Teen-LABS)	Cohort	LAGB	14	5 y	Some improvement dyslipidemia/BP/T2D despite regain and band removal
Fleming 2025 [141]	Cohort	LAGB	14	10 y	Cardiometabolic risk largely unchanged; weight loss not sustained
Kelly 2016 [34]	Prospective	RYGB/VSG	NR	1 y	↓ IL-6, leptin, oxidized LDL, insulin resistance; ↑ adiponectin
Michalsky 2018 [33](Teen-LABS)	Cohort	RYGB/VSG	NR	3 y	Sustained ↓ hs-CRP and CV risk factors
Kumar 2018 [120]	Pilot prospective	RYGB (female)	12	12 mo	↓ HOMA-IR 76.5%, ↓ hs-CRP, ↓ leptin; ↑ adiponectin
Ippisch 2008 [117]	Prospective	RYGB	NR	∼10 mo	LV mass index 54 → 42 g/m2·7; concentric LVH 28 → 3%; normal geometry 36 → 79%
Talib 2025 [15](BASIC)	RCT sub-study	MBS	18 (imaging)	1 y	↑ LV mass index and relative wall thickness within surgical group
Ryder 2020 [132] (Teen-LABS)	Cohort (modelled)	RYGB/VSG	NR	5 y	Predicted 30-y CVD risk 7.9%→4.0% at 1 y, sustained to 5 y
Safar 2026 [150]	Retrospective cohort	SG/RYGB	26	median 37 mo	T2D remission 5/8 (63%), HbA1c 7.1 → 5.6%; HTN remission 4/6 (67%); median ALT 89 → 26 U/L
Zitsman 2020 [152]	Prospective cohort	LAGB	137	5 y	Remission: insulin resistance 85.5%, T2D 60%, HTN 46.7%, dyslipidemia 18.3%
Gosher 2026 [147](INBSR)	Registry cohort	SG/OAGB/LAGB/RYGB	460	≤4 y	Baseline T2D, dyslipidemia, HTN, MAFLD recorded; postoperative remission not reported
Yodoshi 2026 [156]	Prospective controlled cohort	VSG vs comprehensive lifestyle intervention	42 (17 VSG)	52 wk	Adjusted risk difference for complete MASLD resolution 0.49 (0.26–0.72) and for fibrosis improvement 0.36 (0.12–0.59); ALT −54.0 U/L (−80.8 to −27.1); MRI-PDFF −8.0%; T2D resolved in 5/6 vs 0/1
Pastore 2022 [155]	Pre–post cohort	LSG	24	12 mo	HOMA-IR 7.4 → 2.5; HDL 44 → 51 mg/dL; ALT 38.6 → 17.2 IU/L; NASH 50%→8.3%; F2 fibrosis 45.8%→8.7%
Lee 2026 [153]	Retrospective cohort	LSG	24	12 mo	Triglycerides 167.3 → 67.9 mg/dL; fasting C-peptide 4.0 → 2.4; ALT 49.4 → 16.0 U/L; hepatic CAP 337 → 246 dB/m, LSM 7.1 → 5.1 kPa

Comparative and cohort evidence generally showed improvement in glycemic control, with high rates of type 2 diabetes remission sustained to 3–5 years in Teen-LABS and AMOS and to 5–9 years in retrospective sleeve-gastrectomy cohorts. Improvements were also reported in lipids, blood pressure, inflammatory markers, renal measures, and cardiac geometry, while Teen-LABS modelling suggested a sustained reduction in predicted 30-year cardiovascular risk. Outcomes after adjustable gastric banding were less durable, with early metabolic improvements largely attenuated by 10 years (Table 3).

Randomized cardiometabolic outcomes were pooled when at least two independent cohorts provided sufficiently comparable data (Fig. 3A–C). Observational outcomes were not pooled because group data, variances, follow-up time points, and cohort linkage were insufficiently comparable and were therefore synthesized narratively and in Table 3.

Fig. 3.

Exploratory random-effects meta-analyses of cardiometabolic outcomes at 2 years following metabolic and bariatric surgery versus non-surgical care in adolescents with severe obesity: (A) systolic blood-pressure change (mmHg); (B) diastolic blood-pressure change (mmHg); and (C) triglyceride change (mmol/L). Each analysis includes two independent randomized cohorts.

3.5. Sleep outcomes

Sleep outcomes were reported exclusively in non-randomized studies; no randomized trial evaluated postoperative obstructive sleep apnea (OSA), apnea–hypopnea index (AHI), oxygenation, or sleep architecture. Objective preoperative and postoperative sleep studies were available in only three reports. Two small Roux-en-Y gastric bypass cohorts [53] showed marked reductions in AHI with improved oxygenation and sleep architecture, although participant overlap was possible. A historical series (Breaux 1995) reported OSA resolution in all nine reassessed patients but was limited by obsolete procedures and concomitant uvulopalatopharyngoplasty in two patients [151]. The principal comparative study (Qureshi 2021) reported 38% OSA resolution after sleeve gastrectomy versus none among non-surgical referrals, subject to selection bias and confounding [95]. Other sleeve-gastrectomy and banding cohorts reported improvement or remission, but most relied on clinical or self-reported outcomes without repeat polysomnography (Table 4). Overall, surgery was associated with improvement in OSA severity and related sleep measures, but evidence was limited by small uncontrolled cohorts, inconsistent remission definitions, and possible cohort overlap.

Table 4.

Sleep outcomes by procedure and follow-up.

Study	Design	Procedure	n	Follow-up	Sleep findings	Assessment
Kalra 2005 [104]	Prospective cohort	RYGB	35 (10 OSA, PSG)	∼5 mo	Median AHI 9.1 → 0.65/h; min SpO2 82.9 → 91.7%	Polysomnography
Kalra 2008 [53]	Prospective cohort	RYGB	19	∼1 y	Mean AHI 20.5 → 6.1/h; sleep efficiency 73.1 → 80.2%; arousal index 11.3 → 7.6/h	Polysomnography
Qureshi 2021 [95]	Comparative NRS	SG vs non-surgical	NR	12 mo	OSA resolution 38% vs 0%	Documented/clinical
Iossa 2018 [90]	Cohort	SG	NR	24 mo	61% remission	Clinical
El-Matbouly 2018 [17]	Retrospective cohort	SG	NR	5 y	64% resolution	Clinical
Mohammed 2021 [69]	Cohort	SG	NR	12 mo	Complete OSA remission	NR
Tuna 2020 [91]	Cohort	SG	NR	median 18.5 mo	Complete OSA remission	NR
Shehata 2025 [64]	Cohort	SG	NR	7 y	OSA prevalence 39.7 → 6.4%	Clinical
Goldenshluger 2023 [89]	Cohort	SG	NR	∼10.8 y	Remission in all affected	Clinical
Zitsman 2015 [18]	Prospective cohort	LAGB	137	3 y	OSA improvement/resolution 34.7%	Partly self-report
Dumont 2018 [100]	Cohort	LAGB	NR	56 mo	50% resolution	NR
Furbetta 2020 [129]	Cohort	LAGB	NR	long-term	69% remission	NR
Collins 2007 [74]	Cohort	RYGB	2 (affected)	∼1 y	OSA improved in both	Clinical
Zeinoddini 2014 [82]	Prospective cohort	Gastric plication	12	NR	Sleep-apnea symptom resolution	Symptom-based
Breaux 1995 [151]	Case series	VBG/RYGB/BPD	22 (11 OSA)	mean 32 mo	Sleep apnea resolved in all 9 with repeat studies; nocturnal low SpO2 73 → 93%	Repeat sleep studies
Zitsman 2020 [152]	Prospective cohort	LAGB	137 (49 OSA)	5 y	OSA remission 19/49 (38.8%)	Clinical report
Safar 2026 [150]	Retrospective cohort	SG/RYGB	26 (15 OSA)	median 37 mo	Remission 5/15 (33%); improvement 6/15 (40%)	Clinical (no repeat PSG)
Barnett 2005; Sugerman 2003; Cozacov 2014; Strauss 2001; Fielding & Duncombe 2005; Schmitt 2016; Zitsman 2011 [25], [39], [63], [77], [96], [102]	Small cohorts	RYGB/LAGB/mixed	NR	varied	Resolution/improvement reported; denominators and criteria often unclear	Mostly clinical/self-report

A proportional meta-analysis of OSA remission was not performed because outcome ascertainment and denominators were not sufficiently comparable. Breaux (1995) reported remission among nine reassessed patients using repeat sleep studies, whereas Zitsman et al. (2020) and Safar et al. (2026) used clinical remission definitions without postoperative polysomnographic confirmation; Breaux also excluded two affected patients who were not reassessed [150], [151]. Pooling these estimates would therefore imply a degree of comparability and precision unsupported by the underlying data. Remission frequencies are reported individually in Table 4.

3.6. Pulmonary outcomes

Pulmonary outcomes were reported only in non-randomized studies and were the least frequently assessed domain; no randomized trial reported spirometry. The only study with objective postoperative lung-function testing (Stafler 2025; 17 sleeve-gastrectomy patients) reported improvements in lung clearance index, FEV₁, and lung volumes at 4 months that correlated with BMI reduction, but was limited by its small sample and short follow-up [99]. Other cohorts reported improvement in asthma, dyspnea, or pulmonary hypertension based primarily on clinical records or self-report rather than repeat objective testing (Table 5).

Table 5.

Pulmonary outcomes by procedure and follow-up.

Study	Design	Procedure	n	Follow-up	Pulmonary findings	Method
Stafler 2025 [99]	Prospective	SG	17	4 mo	↓ BMI 45.5 → 36.5; lung clearance index 7.0 → 6.5 (p = 0.009); FEV1, TLC, FRC, RV, FEF25–75 improved	Objective (spirometry/MBW)
Nadler 2009 [76]	Prospective cohort	LAGB	45	NR	Asthma and dyspnea improved (denominators/criteria unclear)	Clinical
Messiah 2013; Nunez Lopez 2017; Zitsman 2011, 2015 [35], [39], [67]	Registry/institutional	RYGB/LAGB	NR	varied	Asthma improvement	Clinical/self-report
Dumont 2018; Paulus 2016 [97], [100]	Cohort	LAGB	NR	long-term	Complete asthma resolution (few affected)	Clinical
Abu-Abeid 2003 [26]	Cohort	LAGB	NR	NR	Pulmonary hypertension improved/resolved; ↓ CPAP dependence	Clinical
Capella & Capella 2003 [27]	Cohort	VBG + RYGB	NR	long-term	Pulmonary hypertension resolved	Clinical
Hoeltzel 2022 [68]	MBSAQIP	SG/RYGB	5068	30 d	Overall complications 1.2%; RYGB > SG (OR 2.27); pulmonary events not separately reported	Registry
Kachornvitaya 2025 [134]	MBSAQIP	SG/RYGB	NR	30 d	Serious complications 1.3% (BMI≥50) vs 1.5%	Registry
Frederick 2025 [36]	MBSAQIP	SG/RYGB	NR	30 d	PE 0.3% (Black) vs 0% (White); very few events	Registry
Lainas 2020 [84]	Cohort	SG	84	periop	1 pneumonia (only major complication)	Clinical
Zitsman 2020 [152]	Prospective cohort	LAGB	137 (19 asthma)	5 y	Asthma remission 43.2%; weight loss >8.1 kg associated with remission (OR 9.34)	Clinical report
Thareja 2026 [148]	MBSAQIP	SG/RYGB	5432	30 d	Composite 30-day complications 1.0%; respiratory events not separately reported	Registry

Perioperative respiratory complications were uncommon in large MBSAQIP analyses; pulmonary embolism, pneumonia, and prolonged ventilation were infrequently and inconsistently reported, and overlapping registry populations precluded independent pooling. Overall, pulmonary function and respiratory comorbidities may improve after surgery and serious perioperative respiratory events appear uncommon, but the evidence remains sparse and methodologically limited. No pooled estimate was calculated, and cardiopulmonary exercise testing (CPET)-only evidence was excluded according to the protocol.

3.7. Anthropometric outcomes

Anthropometric outcomes were the most frequently reported, using body weight, BMI, %TWL, %EWL, and %EBMIL, from 30 days to 15 years; because definitions differed, these metrics were not treated as interchangeable, and companion reports were mapped to their cohorts (Table 6). In the AMOS2 randomized trial, mean BMI fell 12.6 kg/m² at 2 years versus 0.2 kg/m² with non-surgical treatment (between-group −12.4 kg/m², 95% CI −15.5 to −9.3, p < 0.0001), with 63% versus 4% reaching a BMI below 30. The BASIC banding trial showed a smaller effect that was no longer evident at 8.6 years, complicated by crossover and 52% band removal, and the O'Brien banding trial likewise showed substantially greater weight loss with banding than with a supervised lifestyle programme at 2 years (BMI −12.7 vs −1.3 kg/m²), at high risk of bias from informative dropout. Comparative and prospective cohorts consistently showed greater and more durable weight loss after Roux-en-Y gastric bypass and sleeve gastrectomy (roughly 25–31% BMI or weight reduction sustained to 3–9 years) than after banding, which produced smaller, less durable loss with frequent band removal (Table 6).

Table 6.

Anthropometric outcomes by procedure and follow-up.

Study	Design	Procedure	n	Follow-up	Weight/BMI outcome
Järvholm 2023 [13](AMOS2)	RCT	MBS vs non-surgical	50	2 y	BMI −12.6 vs −0.2 kg/m2 (Δ −12.4, 95% CI −15.5 to −9.3); 63% vs 4% BMI<30
Roebroek 2024/Talib 2025 (BASIC) [14], [15]	RCT	LAGB vs lifestyle	52	1 y	Weight −11.2% vs +1.7%; BMI −5.24 vs + 0.37 kg/m2
de Boom 2026 [140](BASIC)	RCT (long-term)	LAGB	39	8.6 y	Initial advantage lost; 52% band removal
O'Brien 2010 [12]	RCT	LAGB vs supervised lifestyle	50	24 mo	BMI −12.7 vs −1.3 kg/m2; markedly greater %EWL with banding
Inge 2018 [37](PCORnet)	Comparative	RYGB/SG/LAGB	NR	1–3 y	BMI −31%/−28%/−10% at 1 y; −29%/−25% at 3 y
Blumenfeld 2020 [38]	Registry	SG vs LAGB	NR	1 y	BMI −15.4 vs −10.3 kg/m2; weight −43.6 vs −28.8 kg
Lee 2012 [108]	Comparative	RYGB vs LAGB	NR	24 mo	EWL 83.4% vs 29.7%
Inge 2016 [6](Teen-LABS)	Prospective cohort	RYGB/SG	242	3 y	Weight −27%; BMI 53 → 38 kg/m2
Ogle 2021 [107](Teen-LABS)	Cohort	RYGB/SG	NR	5 y	BMI −22.2% (younger) vs −24.6% (older)
Olbers 2012/Henfridsson 2019 [59], [136](AMOS)	Cohort	RYGB	81	2–5 y	Weight −32% (2 y); BMI 45.5 → 30.2; 93% > 50% EBMIL; −28% at 5 y
El-Matbouly 2018 [17]	Cohort	SG	NR	5 y	BMI 48.0 → 30.3; 78% EWL
Khidir 2018 [60]	Cohort	SG	NR	5 y	35.8% TWL; 78% EWL; BMI 47.6 → 30.3
Raziel 2014 [144]	Uncontrolled cohort	SG	31	60 mo	%EWL trajectory with attrition (n 31 → 2 over 1–60 mo); BAROS good/excellent 71%
Elhag 2021 [44]	Cohort	SG	NR	9 y	BMI 45.6 → 30.2
Shehata 2025 [64]	Cohort	SG	NR	7 y	BMI 48.1 → 29.3; %TWL 32.9% (3 y)→29.1% (7 y)
Holterman 2010; Nadler 2007/2008; Zitsman 2015 [18], [22], [124]	Cohort	LAGB	NR	1–3 y	34–63% EWL (12–24 mo); 41.2% EWL (3 y)
Furbetta 2020 [129]	Retrospective	LAGB	NR	5–15 y	EWL 61.7% (5 y), 48.1% (10 y), 55.8% (15 y; n = 5)
Dewberry 2020/Fleming 2025 [141], [142] (Teen-LABS)	Cohort	LAGB	14	5–10 y	43% exceeded baseline BMI at 5 y; 10% BMI loss lost by 10 y
Carbajo 2019 [16]	Cohort	OAGB	NR	5 y	BMI 48.2 → 25.9; 38.8% TWL
Capella & Capella 2003; Papadia 2007 [27], [28]	Cohort	VBG + RYGB/BPD	NR	long-term	∼78–80% EWL
Gosher 2026 [147](INBSR)	Registry cohort	SG/OAGB/LAGB/RYGB	460	6 mo–4 y	SG %EWL 56.3% (6 mo), 70.7% (2 y), 68.5% (4 y); cohort %TWL 30.9% (1 y), 29.9% (4 y)
Luna 2026 [149]	Prospective cohort	RYGB	60	12 mo	BMI 47.10 → 28.61 kg/m2; reduction −18.49 vs −15.01 kg/m2 in adults (p = 0.002)
Safar 2026 [150]	Retrospective cohort	SG/RYGB	26	median 37 mo	Median %TWL 28% (range 5–49); BMI change −14 kg/m2
Zitsman 2020 [152]	Prospective cohort	LAGB	137	5 y	%EWL 40.6% (3 y) → 35.5% (5 y); 23/127 (18%) achieved ≥50% EWL
Breaux 1995 [151]	Case series	VBG/RYGB/BPD	22	32–50 mo	BMI 67.8 → 46.5 (apnea group); 56.4 → 35.5 (non-apnea); 45–59% EWL
Rodrigues de Oliveira Filho 2026 [154] (COSMIC)	Multicentre retrospective cohort	SG	223 (adolescent subgroups)	12 mo	%BMI loss 23.6% (<15 y) and 21.8% (15–17.99 y), P = 0.347; 12-month %TWL 22% and 21%; 34–41% achieved <20% BMI loss
Lee 2026 [153]	Retrospective cohort	LSG	24	12 mo	%TWL 43.3%; fat-mass reduction 51.1%
Pastore 2022 [155]	Pre–post cohort	LSG	24	12 mo	Median BMI 45.0 → 31.0 kg/m2 (≈29% TWL)
Yodoshi 2026 [156]	Prospective controlled cohort	VSG vs lifestyle	42 (17 VSG)	52 wk	Adjusted percent weight change −19.6% (−30.0 to −9.1) and %BMIp95 −27.6% (−42.6 to −12.7) versus lifestyle intervention

One-anastomosis gastric bypass and older gastroplasty or biliopancreatic-diversion procedures were represented by small retrospective cohorts. Additional observational cohorts generally showed substantial weight reduction after surgery, although long-term estimates were limited by declining follow-up and heterogeneous outcome definitions (Table 6).

The randomized 2-year BMI comparison was pooled across three independent cohorts (Fig. 4), showing greater BMI reduction with surgery than with non-surgical care (mean difference −9.39 kg/m², 95% CI −13.90 to −4.89; I² = 88%). Observational anthropometric outcomes were not pooled because definitions, group-level variances, follow-up time points, and cohort linkage were insufficiently comparable.

Fig. 4.

Random-effects meta-analysis of BMI change (kg/m²) at 2 years following metabolic and bariatric surgery versus non-surgical care in adolescents with severe obesity. Three independent randomized cohorts contributed: BASIC (de Boom 2026), AMOS2 (Järvholm 2023), and O'Brien 2010. The pooled mean difference was −9.39 kg/m² (95% CI −13.90 to −4.89; I² = 88%) [12], [13], [140].

3.8. Safety and adverse events

Safety outcomes were reported across randomized trials, prospective cohorts, case series, administrative databases, and national registries, with follow-up from 30 days to 15 years; overlapping populations were treated as companion reports (Table 7). Perioperative and 30-day mortality was uncommon, with no 30-day deaths in Teen-LABS or several national studies and two deaths (0.04%) among 5068 adolescents in one MBSAQIP cohort. Early complications were also generally uncommon: Teen-LABS reported 8% major and 15% minor 30-day complications, whereas MBSAQIP analyses reported rates of 1.2–2.9%, with sleeve gastrectomy generally associated with fewer serious events, reinterventions, and readmissions than Roux-en-Y gastric bypass.

Table 7.

Safety, adverse events, and nutritional outcomes by procedure and follow-up.

Domain	Study	Procedure	n	Follow-up	Finding
Mortality	Inge 2014 [110] (Teen-LABS)	RYGB/SG	–	30 d	No 30-day deaths
Mortality	Hoeltzel 2022 [68](MBSAQIP)	SG/RYGB	5068	30 d	2 deaths (0.04%)
Mortality	Inge 2016 [6] (Teen-LABS)	RYGB	242	3 y	1 death (0.4%; hypoglycemia, type 1 diabetes)
Mortality	Yska 2021 [98]	LAGB	NR	40 mo	1 death (band-erosion hemorrhage)
Early complications	Inge 2014 [110] (Teen-LABS)	RYGB/SG	–	30 d	Major 8%, minor 15%
Early complications	Varela 2007 [111]	Mixed	–	30 d	5.5% (adolescents) vs 9.8% (adults)
Early complications	Hoeltzel 2022 [68	SG/RYGB	5068	30 d	Complications 1.2%; readmission 3.5%; reoperation 1.1%; RYGB > SG
Early complications	El Chaar 2021 [131]	SG vs RYGB	–	30 d	Serious AE 2.9% vs 6.5%; lower reintervention/readmission after SG
Early complications	Ballinger 2025 [133]	SG/RYGB	–	30 d	Complications 2.9%; ED visits 4.9%; dehydration commonest
Early complications	Kachornvitaya 2025 [134]	SG/RYGB	–	30 d	Serious 1.3% (BMI≥50) vs 1.5%
Early complications	Schutte & Zitsman 2022 [128]	SG	142	intraop	Splenic ischemia 31% (no clinical sequelae)
Early complications	Dobbs 2025 [145] (IMPROVE-T2D)	SG	14	3 mo	1 cholecystectomy (symptomatic cholelithiasis)
Reoperation/revision	Olbers 2012 [136] (AMOS)	RYGB	81	2 y	AEs 33%; 15% additional surgery (hernia, cholecystectomy)
Reoperation/revision	Inge 2016 [6] (Teen-LABS)	RYGB/SG	242	3 y	13% additional intra-abdominal procedure
Reoperation/revision	Beamish 2026 [66]	RYGB/SG	–	5 y	68% ≥ 1 healthcare event; 36% surgery-related
Reoperation/revision	Zitsman 2015 [18]	LAGB	137	3 y	22% reoperation; 20% removal/conversion
Reoperation/revision	Dumont 2018; Yska 2021 [98], [100]	LAGB	–	long-term	Band removal 25–30%
Reoperation/revision	Furbetta 2020 [129]	LAGB	–	long-term	Reoperation 30.5%; removal 13.5%; conversion 3.4%
Reoperation/revision	O'Brien 2010 [12]	LAGB	24	24 mo	Band revisional procedures in a substantial minority
Reoperation/revision	de Boom 2026 [140] (BASIC)	LAGB	–	8.6 y	52% band removal
Reoperation/revision	Fleming 2025 [141]	LAGB	14	10 y	4/14 band removal
Reoperation/revision	Shehata 2025 [64]	SG	NR	7 y	Complications 7.9%; revision 14.3%
Reoperation/revision	Goldenshluger 2023 [89]	SG	NR	10 y	GERD 22.6%; cholecystectomy 22.6%
Nutrition	Inge 2016 [6] (Teen-LABS)	RYGB/SG	242	3 y	Hypoferritinemia 57%; vitamin B12 deficiency 8%
Nutrition	Olbers 2012 [136] (AMOS)	RYGB	81	2 y	↓ ferritin and B12 (adherence-related)
Nutrition	Silva 2017 [72]	RYGB	NR	1 y	Worsening vitamin A/β-carotene; night blindness
Nutrition	Goldberg 2017 [29]	SG vs LAGB	NR	3–12 mo	No anemia on supplementation; low folate after SG
Nutrition	Zolfaghari 2025 [101]	SG vs RYGB	NR	1 y	No between-procedure micronutrient difference
Nutrition	Goldenshluger 2023 [89]	SG	NR	10 y	Iron-deficiency anemia 19%; alopecia 48.4%
Nutrition	Carbajo 2019 [16]	OAGB	NR	5 y	No generalized malnutrition; 1 iron-deficiency anemia
Nutrition	Papadia 2007 [28]	BPD	NR	long-term	Protein malnutrition 16%; 20% reoperation
Perioperative	Thareja 2026 [148]	SG/RYGB	5432	30 d	Composite complications 1.0%; median stay 1–2 d; higher odds for Black vs White patients (OR 3.21, 1.55–6.71)
Perioperative	Safar 2026 [150]	SG/RYGB	26	30 d	No complications, readmissions, or reoperations; median stay 24 h
Reoperation	Zitsman 2020 [152]	LAGB	137	5 y	80 additional operations in 63 patients; band removal 33/127 (26%)
Reoperation	Gosher 2026 [147]	SG	460	≤4 y	16 conversion/revisional procedures in adulthood; 81.3% for recurrent weight gain
Late complication	Safar 2026 [150]	SG	25	median 37 mo	De novo GERD 4/25 (16%)
Nutrition	Luna 2026 [149]	RYGB	60	12 mo	Vitamin D inadequacy 78.3%; hypocalcemia 63.3%; hypophosphatemia 76.7%; secondary hyperparathyroidism 65.0%
Nutrition	Breaux 1995 [151]	BPD/RYGB/VBG	22	32–50 mo	Protein deficiency in 3 BPD patients; vitamin A/D and folate deficiencies
Perioperative	Yodoshi 2026 [156]	VSG	17	52 wk	One gastric leak (6%); minor 30-day complications 29% vs 0% with lifestyle intervention (P = 0.007); no deaths
Perioperative	Lee 2026 [153]	LSG	24	12 mo	No major postoperative complications (0.0%)
Nutrition	Pastore 2022 [155]	LSG	24	12 mo	Total homocysteine 15.7 → 21.1 μmol/L (P = 0.0146); folate and vitamin B12 not monitored

Reoperations and revisional procedures increased with longer follow-up and varied by procedure, occurring most frequently after adjustable gastric banding, with band removal, conversion, or revision reported in 20–52% of participants across cohorts. Sleeve gastrectomy had lower revision rates but was associated with late gastroesophageal reflux, cholelithiasis, and cholecystectomy in some cohorts. Nutritional deficiencies were also reported, particularly iron/ferritin and vitamin B12 deficiency after gastric bypass, with additional procedure-specific deficiencies and protein malnutrition reported less frequently (Table 7).

Overall, perioperative mortality and serious 30-day complications were uncommon, particularly after sleeve gastrectomy, whereas longer follow-up identified clinically important reoperations, nutritional deficiencies, and other procedure-specific complications. Safety outcomes were not pooled because definitions, surveillance periods, denominators, and registry populations differed or overlapped.

3.9. Subgroup, sensitivity, and reporting-bias analyses

Quantitative synthesis was limited to randomized evidence. Three independent randomized cohorts (O'Brien et al., 2010; Järvholm et al., 2023; and BASIC, represented at 2 years by de Boom et al., 2026) contributed to the primary meta-analysis of BMI change, and two cohorts contributed to each exploratory analysis of systolic blood pressure, diastolic blood pressure, and triglyceride change (Fig. 3, Fig. 4) [12], [13], [140]. For multiple BASIC publications, only one effect per cohort, outcome, and time point was included. Observational outcomes were not pooled because independent studies with comparable definitions, aligned follow-up, and complete effect estimates were insufficient; consequently, the planned meta-regression of weight reduction against sleep and cardiometabolic improvement was not performed. OSA remission was also not pooled because outcome ascertainment differed substantially across studies, as described in Section 3.5.

Prespecified subgroup analyses by baseline obesity severity and OSA severity could not be performed because stratified data were insufficient. Formal subgroup analysis by procedure was also not feasible with only three cohorts. Although the primary BMI analysis showed substantial heterogeneity (I² = 88%), apparent differences between procedures were not tested formally and should not be interpreted as subgroup effects.

Sensitivity analyses were limited to the primary BMI outcome. Excluding the O'Brien trial, which was at high overall risk of bias and required derivation of the between-group effect and standard error, attenuated the pooled estimate to −8.22 kg/m² (95% CI −16.48 to 0.04). Hartung–Knapp adjustment produced a wider interval (−20.6 to 1.9) that crossed the null, underscoring uncertainty with only three heterogeneous cohorts.

Formal assessment of small-study effects was not performed because no meta-analysis included the protocol-specified minimum of 10 independent studies; publication bias therefore could not be excluded. Outcome-reporting concerns were considered within study-level risk-of-bias assessments, but a review-wide estimate of selective reporting was not possible.

3.10. Certainty of evidence

Certainty was assessed using GRADE for the four pooled randomized outcomes and narratively for unpooled domains, for which incomplete effect estimates, variance measures, and cohort-linkage information precluded reliable formal assessment. Summary-of-findings ratings are presented in Table 8.

Table 8.

GRADE summary of findings — pooled randomized outcomes at 2 years, metabolic and bariatric surgery versus non-surgical care in adolescents with severe obesity.

Outcome (2 years; surgery vs non-surgical care)	Participants (cohorts)	Effect, MD (95% CI)	Heterogeneity	Certainty (GRADE)	Reasons for downgrading
BMI change (kg/m2)	141 (3 RCTs: O'Brien 2010 [12]; AMOS2 [13]; BASIC [14], [15])	−9.39 (−13.90 to −4.89)	τ2 = 13.74; I2 = 88%; P = 0.0003	⊕⊕◯◯ LOW a,b	Risk of bias (−1): two of three cohorts were at high overall risk, principally because of missing outcome data and complete-case analyses. Inconsistency (−1): substantial heterogeneity (I2 = 88%), with a smaller effect in BASIC; the direction of effect was consistent across cohorts.
Systolic BP change (mmHg)	89 (2 RCTs: O'Brien 2010 [12]; AMOS2)	+1.92 (−6.06 to +9.90)	τ2 = 11.78; I2 = 31%; P = 0.228	⊕◯◯◯ VERY LOW a,d	Risk of bias (−1). Imprecision (−2): 95% CI compatible with both lower and higher systolic blood pressure; two small cohorts.
Diastolic BP change (mmHg)	89 (2 RCTs: O'Brien 2010 [12]; AMOS2)	−0.92 (−5.48 to +3.65)	τ2 = 0; I2 = 0%; P = 0.481	⊕◯◯◯ VERY LOW a,d	Risk of bias (−1). Imprecision (−2): 95% CI compatible with both lower and higher diastolic blood pressure; two small cohorts.
Triglyceride change (mmol/L)	89 (2 RCTs: O'Brien 2010 [12]; AMOS2)	−0.36 (−0.57 to −0.14)	τ2 = 0; I2 = 0%; P = 0.496	⊕⊕◯◯ LOW a,e	Risk of bias (−1). Imprecision (−1): estimate based on only two small randomized cohorts.

GRADE Working Group grades: High — further research is very unlikely to change confidence in the estimate; Moderate; Low; Very low. MD = mean difference; random-effects (DerSimonian–Laird), generic inverse variance.

• a.
  Two of the three randomized cohorts (O'Brien 2010; BASIC) were at high overall risk of bias, principally from missing outcome data and completer-only analyses; the O'Brien BMI effect and standard error were derived from group-level confidence intervals [12].
• b.
  Substantial statistical heterogeneity (I² = 88%) reflecting a smaller effect of adjustable gastric banding (BASIC) than of predominantly Roux-en-Y gastric bypass (AMOS2); all cohorts favored surgery in direction.
• c.
  Additional consideration (not an applied downgrade): Additional consideration (not an applied downgrade): The Hartung–Knapp-adjusted interval (−20.6 to +1.9) crossed the null, indicating additional uncertainty in the precision of the pooled estimate. This was not applied as a separate GRADE downgrade.
• d.
  Confidence intervals include both appreciable benefit and appreciable harm.
• e.
  Publication bias was not formally assessable for any outcome (fewer than 10 studies); no outcome was downgraded for indirectness, although two of three BMI cohorts evaluated adjustable gastric banding, a procedure now uncommon in adolescents.

For the primary pooled outcome, BMI change at 2 years, randomized evidence began at high certainty and was downgraded one level for risk of bias, because two of the three contributing cohorts were at high overall risk driven principally by missing outcome data, and one level for inconsistency because heterogeneity was substantial (I² = 88%). Overall certainty was therefore rated low. The direction of effect favored surgery across all three cohorts, although the magnitude remained uncertain. The Hartung–Knapp sensitivity analysis produced a substantially wider confidence interval that crossed the null and was considered an additional indication of uncertainty in the precision of the pooled estimate, but was not applied as a separate GRADE downgrade.

For unpooled anthropometric and cardiometabolic outcomes, confidence was limited by predominantly observational evidence, non-random treatment allocation, attrition, inconsistent outcome definitions, treatment crossover, and overlapping cohorts. Nevertheless, comparative studies generally showed a consistent direction of anthropometric and cardiometabolic improvement, with greater uncertainty regarding effect magnitude and long-term durability.

Certainty was lower for sleep and pulmonary outcomes. No randomized trial evaluated postoperative OSA outcomes, objective preoperative and postoperative sleep studies were available in only three reports, and most other sleep outcomes relied on clinical diagnoses, symptoms, or heterogeneous remission definitions. Pulmonary evidence was still more limited, with only one small prospective cohort providing objective postoperative lung-function testing. Small samples, lack of nonsurgical comparators, inconsistent outcome ascertainment, and limited follow-up reduced confidence in these domains.

For safety outcomes, large prospective cohorts and registries consistently indicated that perioperative mortality and serious 30-day complications were uncommon. Confidence in procedure-specific comparisons and late adverse events was lower because of non-random procedure selection, overlapping registry populations, heterogeneous definitions and surveillance periods, and attrition. Rare serious harms could not be estimated precisely.

Across domains, certainty was principally limited by risk of bias, heterogeneity, imprecision, incomplete long-term follow-up, and possible participant overlap. Publication bias could not be formally assessed because no meta-analysis included at least 10 independent studies.

3.11. Reporting gaps in clinically relevant measures

The post hoc audit identified substantial gaps and inconsistencies in the reporting of clinically relevant measures across the included literature (Table 9). Because the audit was based on published reports and available supplementary material, absence from a publication should not be interpreted as evidence that a measure was never collected.

Table 9.

Guideline-based measures and their reporting across the 147 included reports (post hoc audit).

Domain and measure	Reports measuring it before and after surgery	What the included reports did instead, or comment
CARDIOMETABOLIC
Metabolic syndrome, adolescent-specific definition	5 reports, 3 cohorts (O'Brien 2010; Holterman 2007, 2010; Zitsman 2011 [comparative], 2015) [12], [39], [123]	Three frameworks: adolescent criteria linked to ATP III (O'Brien 2010); IDF (Holterman 2007; Zitsman 2015); fixed adult cut-offs (Zitsman 2011). Resolution defined inconsistently. Khen-Dunlop 2016 reported baseline prevalence only [12], [18], [39], [114], [123].
Echocardiographic left ventricular structure	3 reports, 3 cohorts (Ippisch 2008; Talib 2025; Franco 2017) [15], [93], [117]	LV mass index and geometry (Ippisch 2008; Talib 2025, 18 paired examinations); LVH as a category only (Franco 2017). Teeple 2012 performed echocardiography before surgery only [32], [45], [93], [117].
Diastolic function	1 report (Ippisch 2008) [117]	Tissue-Doppler mitral E/Ea (lateral) 7.7 to 6.3. No report of left atrial size, myocardial strain, or right ventricular indices.
Ambulatory blood-pressure monitoring	None	Office or resting daytime measurements only.
Blood pressure classified against pediatric norms	Inconsistent	Fourth Report percentiles (Talib 2025); 2017 AAP thresholds (Levaillant 2023); 99th percentile for resolution (Zitsman 2015). Adult thresholds of 140/90 mm Hg elsewhere (e.g., Furbetta 2020); pediatric criteria stated to be unavailable in a registry analysis (Nunez Lopez 2017) [15], [18], [58], [67], [129].
Carotid intima–media thickness, pulse-wave velocity, flow-mediated dilation	None	Not measured in any included report.
SLEEP
Polysomnography before and after surgery	3 reports, 2 cohorts (Kalra 2005, 2008; Breaux 1995) [104], [151]	Other reports relied on symptom resolution, discontinuation of positive airway pressure, or self-report.
Polysomnographic definition of OSA	Thresholds varied	AHI ≥1/h or clinical CPAP need (Lee 2026); AHI >2 with PSQ >33 (Iossa 2018); AHI ≥5 or >5 (Kalra 2005, 2008; Holterman 2007; Furbetta 2020); AHI >5 labelled severe (Lacrosniere 2023); obstructive AHI ≥15 (Levaillant 2023) [24], [58], [90], [104], [123], [129], [153]. Most reports gave no definition.
Obstructive AHI and obstructive apnea index	None as a pre–post outcome	Obstructive AHI at baseline only (Kalra 2005) or as a diagnostic threshold (Levaillant 2023) [58], [104]. Obstructive apnea index never reported.
Pediatric respiratory scoring rules	None with pre–post sleep results	Event rules described by Kalra 2008 (events ≥2 breaths; hypopnea 50–80% airflow fall with ≥4% desaturation; Rechtschaffen–Kales staging). AASM 2012 pediatric rules stated for BASIC trial studies (Talib 2025), without sleep results. Not stated elsewhere [15].
Oxygen desaturation index; CO2 during sleep	None	Oxygen-saturation nadir reported in the three reports with repeat studies.
Sleep architecture and arousal index	1 report (Kalra 2008) [53]	Sleep efficiency, stage 1 sleep, and arousal index.
Validated sleep questionnaire at follow-up	1 report (Iossa 2018 [90]; Pediatric Sleep Questionnaire)	Berlin questionnaire and Epworth Sleepiness Scale used before surgery only (Dargan 2018; Dumont 2018) [86], [100].
PULMONARY
Asthma Control Test, Childhood Asthma Control Test, or Asthma Control Questionnaire	None	Asthma status recorded from medication use or clinical records.
Asthma exacerbations, unscheduled care, systemic corticosteroid courses	None	Not reported.
Definition of asthma and its remission	Medication-based where stated	Symptoms requiring pharmacologic treatment (Nunez Lopez 2017); remission as cessation of bronchodilator therapy (Zitsman 2015). Unstated in most reports [18], [67].
Spirometry and fractional exhaled nitric oxide	1 report (Stafler 2025) [99]	Participants with asthma excluded. Dumont 2018 performed spirometry before surgery only, when dyspnea was present [100].
Evaluation for obesity hypoventilation	None	Recorded as a baseline comorbidity (Sugerman 2003) [25]; no postoperative blood gas or bicarbonate.
Echocardiographic assessment of pulmonary hypertension	None	Clinical improvement or resolution only (Abu-Abeid 2003; Capella & Capella 2003) [26], [27].

AAP, American Academy of Pediatrics; AASM, American Academy of Sleep Medicine; AHI, apnea–hypopnea index; ATP III, Adult Treatment Panel III; CPAP, continuous positive airway pressure; IDF, International Diabetes Federation; LV, left ventricular; LVH, left ventricular hypertrophy; OSA, obstructive sleep apnea; PSQ, Pediatric Sleep Questionnaire. Counts refer to the 147 included reports; companion reports of one cohort are counted separately and the number of cohorts is given where it differs.

3.11.1. Cardiometabolic outcomes

Metabolic syndrome was reported before and after surgery in five reports from three cohorts, using heterogeneous definitions; no report used a pediatric percentile-based definition. Echocardiographic cardiac structure was assessed longitudinally in three reports, while diastolic function was reported in only one. No report assessed left atrial size, myocardial strain, or right ventricular function. Ambulatory blood-pressure monitoring and vascular measures such as carotid intima–media thickness, pulse-wave velocity, and flow-mediated dilation were not reported. Blood-pressure definitions also varied between pediatric percentile-based and adult thresholds.

3.11.2. Sleep outcomes

Repeat polysomnography with numerical results was available in only three reports from two cohorts, with heterogeneous definitions of obstructive sleep apnea. Pediatric polysomnographic scoring rules were rarely specified, and the obstructive apnea–hypopnea index was reported only as a baseline value or diagnostic threshold. Oxygen-saturation nadir was reported in the three repeat-study reports, whereas detailed sleep architecture was reported in only one. No report provided an oxygen desaturation index or carbon-dioxide measurement during sleep. Validated sleep questionnaires were rarely used during postoperative follow-up.

3.11.3. Pulmonary outcomes

No study used a validated asthma-control instrument or reported asthma exacerbations, unscheduled care, or systemic corticosteroid courses. Asthma outcomes were generally based on symptoms or medication use. Only one study reported preoperative and postoperative spirometry and fractional exhaled nitric oxide, and that study excluded participants with asthma. Obesity hypoventilation and pulmonary hypertension were reported clinically without objective postoperative physiological assessment.

4. Discussion

4.1. Summary of evidence

This review synthesized 147 reports representing 101 unique studies or cohorts across cardiometabolic, sleep, pulmonary, anthropometric, and safety outcomes after metabolic and bariatric surgery in adolescents with severe obesity. Three independent randomized cohorts showed greater BMI reduction with surgery than with non-surgical care at 2 years (MD −9.39 kg/m², 95% CI −13.90 to −4.89), although certainty was low because of risk of bias and substantial heterogeneity. Triglycerides also improved with low-certainty evidence, whereas blood-pressure effects remained uncertain. Observational evidence supported sustained weight reduction and improvement or remission of multiple cardiometabolic conditions, particularly after sleeve gastrectomy and Roux-en-Y gastric bypass; outcomes after adjustable gastric banding were less durable and revision was frequent. Sleep outcomes generally improved, but objective polysomnographic evidence was sparse, and pulmonary evidence was insufficient for firm conclusions. Perioperative mortality and serious short-term complications were uncommon, whereas longer follow-up identified reoperations, nutritional deficiencies, and procedure-specific complications. Overall, the evidence supports substantial multisystem benefit after surgery, but the magnitude and long-term durability of several effects remain uncertain.

4.2. Interpretation in relation to previous evidence

These findings extend earlier syntheses. The Cochrane review of surgery for pediatric obesity identified only one randomized trial and could draw limited conclusions [10]; inclusion of O'Brien, AMOS2, and BASIC allowed the present review to synthesize three independent randomized cohorts. The direction of the findings is consistent with previous long-term observational reviews and major prospective cohorts reporting sustained BMI reduction and improvement in weight-related conditions [6], [7], [8], [9].

The present review extends this evidence by jointly examining cardiometabolic, sleep, and pulmonary outcomes in adolescents. In contrast to the extensive anthropometric and cardiometabolic literature, objective sleep and pulmonary evidence was sparse. This distinction is clinically important because improvement in weight does not establish resolution of sleep or pulmonary disease; persistent obstructive sleep apnea despite postoperative weight reduction has been reported in adolescents [4].

4.3. Clinical implications

The findings are consistent with guidelines supporting metabolic and bariatric surgery as an option for appropriately selected adolescents with severe obesity within multidisciplinary care [5]. Preoperative assessment should include evaluation for sleep-disordered breathing, and adolescents with documented obstructive sleep apnea may warrant objective postoperative reassessment rather than assuming resolution from symptoms or weight reduction alone. Long-term cardiometabolic and nutritional surveillance remains important because remission is not universal and iron, vitamin B12, and other deficiencies may occur after surgery. Procedure-specific counseling should also acknowledge the less durable outcomes and greater revision burden observed with adjustable gastric banding. Although perioperative risk was generally low, longer-term reoperations and other complications reinforce the need for continued specialist follow-up.

4.4. Research implications

Future research should prioritize standardized outcome definitions, objective measurement, transparent cohort identification, and longer controlled follow-up. A core outcome set with prespecified definitions and assessment time points would improve comparability across studies, particularly for diabetes, hypertension, dyslipidemia, anthropometric outcomes, and obstructive sleep apnea. Sleep and pulmonary outcomes require objective postoperative assessment, including polysomnography and spirometry, rather than reliance on clinician-recorded or symptom-based resolution. Transparent identification of shared cohorts would also reduce the risk of participant double-counting across companion publications.

Longer comparative follow-up remains a major evidence gap. Randomized evidence beyond 2 years was extremely limited, while long-term estimates largely came from uncontrolled cohorts with declining retention. Adequately powered comparisons of contemporary procedures should incorporate cardiometabolic, sleep, pulmonary, nutritional, and safety outcomes. The post hoc audit (Table 9) identifies specific priorities for more standardized outcome reporting. No included report documented use of a validated asthma-control instrument, ambulatory blood-pressure monitoring, or vascular imaging as a pre- and postoperative outcome, and diastolic function was reported in only one cohort. Reporting of sleep outcomes was similarly incomplete: objective postoperative polysomnography was uncommon, definitions of obstructive sleep apnea varied, and several clinically informative indices were rarely or not reported. Priority measures include validated asthma-control instruments and exacerbation frequency; serial spirometry when pulmonary outcomes are assessed; polysomnography using age-appropriate scoring criteria with consistent reporting of obstructive respiratory indices, oxygenation, and carbon dioxide measures; ambulatory blood-pressure monitoring where feasible; and echocardiographic assessment that includes relevant structural and diastolic measures. Standardized assessment of these outcomes at prespecified follow-up intervals would strengthen the evidence for clinically important effects that remain poorly characterized.

4.5. Strengths and limitations

Strengths of this review include prospective PROSPERO registration, duplicate independent screening and risk-of-bias assessment with adjudication, a comprehensive multi-database search, explicit linkage of companion publications to unique cohorts, separation of randomized and observational evidence, and pooling only when independence and clinical comparability could be established.

Several limitations should be considered. Most evidence was observational and susceptible to confounding, selection bias, and attrition, while long-term follow-up was frequently incomplete. Outcome definitions were heterogeneous, pulmonary evidence was sparse, and safety data partly originated from overlapping registry populations, leaving some possibility of residual participant duplication. Randomized meta-analyses included only a few small cohorts, and one effect estimate required derivation from reported group data. Eligibility was restricted to English-language publications. Strategy development initially used four concepts: population, obesity severity, intervention, and outcomes. Before execution of the final searches, the outcome concept was removed to avoid restricting retrieval according to outcome reporting, and outcome eligibility was instead assessed during screening and data extraction. Severe or morbid obesity was retained as a required population concept. Requiring explicit obesity-severity terminology may have reduced sensitivity for otherwise eligible reports that did not use these terms in searchable fields. Complete executed database strategies are reported above to permit assessment of this limitation. Although registry searching was undertaken as prespecified, registry-specific executed strategies, per-registry yields, and update-specific yields were not retained, limiting complete reproducibility of these supplementary searches. Publication bias could not be formally assessed because no meta-analysis included at least 10 independent studies. Finally, the audit of reporting gaps was post hoc and based on published reports; absence of a measure from a publication does not establish that it was not collected.

5. Conclusions

In adolescents with severe obesity, metabolic and bariatric surgery produces substantial reductions in BMI; pooled randomized evidence indicates a reduction of approximately 9 kg/m² at 2 years compared with non-surgical care, although certainty in the magnitude of effect is low. Improvements in glycemic control, lipids, blood pressure, and sleep-disordered breathing were also observed, with substantial and generally sustained weight reduction through mid-term follow-up after gastric bypass and sleeve gastrectomy.

Serious short-term complications and perioperative mortality were uncommon, but reoperation, micronutrient deficiencies, and device-related failure after gastric banding remained important during longer-term follow-up, supporting the need for lifelong multidisciplinary surveillance. A major evidence gap is the scarcity of objective, controlled, long-term data on sleep and pulmonary outcomes. Until adolescent trials and prospective cohorts assess these domains as rigorously as anthropometric outcomes, the magnitude and durability of the multisystem effects of surgery will remain uncertain. Future studies should also address the identified reporting gaps through standardized outcome assessment, including validated asthma-control instruments, consistently reported polysomnographic indices using age-appropriate pediatric scoring criteria, ambulatory blood-pressure monitoring, and more comprehensive echocardiographic assessment, including diastolic function.

• •Metabolic and bariatric surgery produces substantial BMI reduction and improvement in multiple cardiometabolic outcomes in adolescents with severe obesity, although certainty in several effect estimates remains limited.
• •Perioperative mortality and serious short-term complications are uncommon, but lifelong nutritional and multidisciplinary surveillance is warranted because late reoperations and nutritional deficiencies occur.
• •Objective long-term assessment of sleep and pulmonary outcomes remains a major evidence gap and should be prioritized in future adolescent studies.

Author contributions (CRediT)

P.L.: conceptualization, methodology, investigation (title/abstract and full-text screening, data extraction, risk-of-bias assessment), data curation, writing — original draft. N.K.: investigation (title/abstract and full-text screening, data extraction, risk-of-bias assessment), validation, writing — review and editing. S.G.: conceptualization, methodology, supervision, adjudication of screening and risk-of-bias disagreements, writing — review and editing. S.A.: methodology, validation, writing — review and editing. A.G.: formal analysis, software, visualization, data curation, project administration, writing — review and editing. All authors meet the ICMJE criteria for authorship, approved the final manuscript, and agree to be accountable for all aspects of the work.

Ethics review

Not applicable; this review synthesized previously published data.

Data availability

The screening records, data-extraction workbooks, risk-of-bias assessments, and analysis inputs are available from the corresponding author on reasonable request. PROSPERO registration: CRD420261354795.

Declaration of AI

During preparation of this manuscript, the authors used generative AI tools for language editing, consistency checking, and document formatting. All study-selection decisions, data extraction, risk-of-bias assessments, statistical analyses, interpretation, and conclusions were performed and verified by the authors, who take full responsibility for the final manuscript.

Source of funding

This review received no specific funding and was undertaken in the authors’ own time.

Declaration of competing interests

The authors declare no conflicts of interest.

Appendix A. Supplementary data

The following are the Supplementary data to this article: