METABOLIC & BARIATRIC SURGERY

Vol. 2, Issue 11   |   September 22, 2026

Weight loss, glucose control, and nutritional health need separate measures of success. This issue examines what our outcome measures capture, what they miss, and how that distinction should shape care.

A — Metabolic Surgery Research

Kim, Sheu, Nimeri et al. | Obesity Surgery, 2026 | 10.1007/s11695-026-08913-w

Kim and colleagues used the 2023 MBSAQIP registry to compare weight-loss measures after primary sleeve gastrectomy and gastric bypass. Among 182,826 patients, 136,406 had paired six-month BMI measurements, and 27,517 had paired BMI and HbA1c data. Percent total weight loss (%TWL) had almost no relationship to starting BMI, whereas excess-weight measures depended substantially on it. For six-month glycemic target attainment in an AUC analysis of 4,220 patients with baseline HbA1c of at least 7.0%, weight measures alone yielded AUCs of 0.596–0.630, compared with 0.768 for a model using baseline HbA1c, insulin use, and age.

Programs should report %TWL for weight outcomes and measure metabolic outcomes separately. A weight-loss threshold can miss a patient whose glucose control improves, or falsely reassure the team about someone whose diabetes remains uncontrolled. This analysis addresses prediction, not whether weight loss contributes biologically to glycemic improvement. Selective HbA1c follow-up and the six-month observation period limit generalizability; target attainment here should not be labeled durable diabetes remission.

KEY FINDING

Among 4,221 patients with baseline HbA1c of at least 7.0%, classification by a 20% TWL threshold and by six-month HbA1c below 7.0% disagreed in 47.4%. Weight and glycemic outcomes need separate places on the postoperative scorecard.

Olmi, Delcarro, Ciccarese et al. | Surgical Endoscopy, 2026 | 10.1007/s00464-026-13391-3

Olmi and colleagues reviewed 400 consecutive primary sleeve gastrectomies performed at one center in 2019, with five-year endoscopy available for 356 patients. All 115 patients with an intraoperatively identified hiatal hernia completed endoscopy; 75 had undergone posterior cruroplasty, and 40 had not. Esophagitis occurred in 4 of 75 patients with a repaired hernia (5.3%), compared with 9 of 40 patients with an unrepaired hernia (22.5%). The association remained after stratification by baseline clinical and endoscopic status, with a Mantel–Haenszel common odds ratio of 0.14 (95% CI, 0.04–0.57).

The long follow-up and endoscopic endpoint make this a useful contribution to counseling about reflux after sleeve gastrectomy. Repair was selective: defects of at least 2 cm were routinely repaired, while smaller defects were managed at the surgeon's discretion. The exact defect size was not retained, and patient-level multivariable adjustment was not possible; therefore, the stratified association does not establish causation or an optimal repair threshold. These results support considering repair of an identified hernia when feasible, but do not promise prevention of reflux injury or equivalence to gastric bypass.

CLINICAL IMPLICATION

Esophagitis was less common when an identified hernia was repaired, but this was an observational comparison. Hiatal repair should not be presented as a guarantee against later esophageal disease.

Eldegwy, Ragab, Alafifi et al. | Obesity Surgery, 2026 | 10.1007/s11695-026-08909-6

Eldegwy and colleagues pooled four randomized and five nonrandomized sleeve gastrectomy studies involving 1,665 patients. Tranexamic acid was given to 842 patients, while 823 received placebo or no intervention. Pooled mean differences favored tranexamic acid by 4.56 minutes in operative duration, 0.12 days in hospital stay, and 0.20 g/dL in hemoglobin decline. Transfusion occurred in 2 of 793 versus 9 of 771 patients, and reoperation for bleeding in 1 of 416 versus 6 of 430; neither comparison was statistically significant. No thromboembolic events were reported in the included studies.

The distinction between smaller laboratory changes and fewer clinically consequential bleeds matters when deciding whether to change a pathway. The hemoglobin-decline benefit was significant in the randomized subgroup but not in the nonrandomized subgroup, and dosing protocols varied. The analysis does not establish that routine tranexamic acid prevents transfusion or return to the operating room, nor does the absence of reported thromboembolism exclude an uncommon safety risk. Programs considering its use should review dosing, contraindications, and thromboprophylaxis within an agreed perioperative protocol rather than adopt a universal regimen from this synthesis.

KEY FINDING

The pooled signal favored tranexamic acid for selected perioperative measures, but not for transfusion or reoperation. Rare-event safety remains unresolved.

Altwaheih, Atieh, Abdel-Jaleel et al. | Obesity Surgery, 2026 | 10.1007/s11695-026-08934-5

Altwaheih and colleagues combined 12 comparative studies of SADI-S and Roux-en-Y gastric bypass, reporting 3,317 adults with severe obesity. Eleven studies were observational, and one was randomized, with primary, revisional, and mixed surgical populations represented. Pooled total weight loss favored SADI-S at 1 and 5 years, but not at 2 years, and no clear differences emerged in major comorbidity remission or early complications. Evidence certainty was predominantly low or very low; lower internal-hernia risk was the exception, rated moderate certainty. The apparent reoperation and long-term complication advantages were statistically fragile: changing a single participant's event status could render the results non-significant for either outcome.

These findings support discussing SADI-S as an option, not declaring it the superior operation across outcomes. The one-year total weight-loss finding remained significant in only one of five leave-one-out analyses, and the randomized trial lacked primary outcomes for about one-fifth of participants. A nonsignificant calcium-deficiency signal from two studies reinforces the need to consider nutritional follow-up alongside weight-loss goals. With no poolable comparative mortality data, the review cannot establish a survival advantage or equivalent mortality risk.

CLINICAL IMPLICATION

A possible weight-loss advantage does not establish better metabolic remission or overall safety. Counsel patients with the uncertainty of the comparative evidence in view.

B — Science of Obesity Metabolism

Aguree and Meng | Current Obesity Reports, 2026 | 10.1007/s13679-026-00753-0

Aguree and Meng review the mechanisms and clinical assessment of altered iron handling in obesity, drawing on human and experimental studies identified through June 2026. Inflammation can increase hepcidin, reduce iron export through ferroportin, and leave less iron available in circulation even when stores are preserved. Ferritin may rise with that same inflammation, obscuring the distinction between stored iron and iron available for erythropoiesis. The review therefore favors interpreting several iron indices together rather than using ferritin as a stand-alone answer.

For a bariatric team, an apparently reassuring ferritin should not end the evaluation when symptoms, anemia, or other iron indices suggest deficiency. Transferrin saturation and an inflammatory marker help interpret the result; soluble transferrin receptor or reticulocyte hemoglobin may help clarify difficult cases where available. Elevated ferritin also does not establish tissue iron overload, particularly in a patient with metabolic liver disease. This is a clinical synthesis, not a trial validating a new treatment algorithm, and assay-dependent thresholds require care.

CLINICAL IMPLICATION

Normal or high ferritin levels can coexist with restricted iron availability. Interpret it alongside circulating iron indices and inflammatory status before concluding that iron supply is adequate.

Wang, Fan, Gao et al. | Current Obesity Reports, 2026 | 10.1007/s13679-026-00761-0

Wang and colleagues review how mechanical loading and metabolic disturbances contribute to obesity-associated osteoarthritis. Their synthesis considers leptin signaling, adipose dysfunction, and the possibility that obesity leaves persistent biological changes after weight loss. They propose combining weight reduction with interventions directed at this residual biological memory. The proposal is a research strategy, not evidence that a tested combination prevents or reverses joint damage in patients.

For patients seeking bariatric care partly because of joint disease, weight and orthopedic outcomes should be discussed separately. Persistent pain or structural disease need not mean that obesity treatment has failed, and a lower weight cannot be promised to restore damaged cartilage. The review raises useful questions about earlier prevention and continued joint care after weight reduction. Its discussion of agents such as semaglutide and metformin does not establish an indication for prescribing either as a proven treatment for obesity-related joint memory.

CLINICAL IMPLICATION

Weight reduction and recovery from established joint disease are different endpoints. Memory-targeted treatment remains a hypothesis requiring clinical testing.

Curtin, Jackson, Lee et al. | Science, 2026 | 10.1126/science.aea4287

Curtin and colleagues combined adipocyte experiments, mouse tumor models, human breast-tissue analyses, and patient-derived tumor organoids to study an obesity–breast cancer mechanism. They identified 9S-HODE, a lipid released by lean mammary adipocytes, as a trigger of ferroptosis in breast cancer cells. Intratumoral delivery of 9S-HODE reduced tumor growth in obese mice and inhibited growth in three patient-derived xenograft organoid lines. Human breast tissue 9-HODE concentrations were inversely related to donor BMI, providing an observational link to the experimental results.

The study suggests that loss of a protective adipose signal may contribute to cancer biology, in addition to signals that actively promote tumor growth. That is a mechanistic finding, not evidence for giving 9S-HODE to patients or changing breast-cancer treatment. It also does not show that bariatric surgery restores this pathway or that the pathway explains a clinical reduction in cancer risk after surgery. Human dosing, delivery, and safety remain questions for subsequent research.

KEY FINDING

Restoring a specific adipocyte-derived lipid suppressed tumor growth in mouse models. The human tissue and organoid findings support further investigation, not claims of clinical efficacy.

Ogawa, Nishikage, Nomura et al. | Diabetology International, 2026 | 10.1007/s13340-026-00926-1

In an editorial, Ogawa and colleagues distinguish compliance with prescribing rules from the clinical safety of the weight-loss process itself. They discuss reports of ketosis or ketoacidosis in young women using tirzepatide without documented appropriate supervision, including nutritional counseling. The authors argue that dietary restriction and inadequate supervision may contribute to harm alongside medication-related risks. These cases illustrate possible hazards; they do not estimate their frequency or prove which factor caused each event.

For integrated bariatric and obesity medicine programs, medication reconciliation should include medications obtained outside the treating clinic and the patient's actual food intake. A prescription within an approved indication does not remove the need to assess nutritional adequacy, weight-loss goals, and emerging symptoms. The same clinical attention is appropriate when medication is used after surgery. This editorial supports supervised care, but its case-based reasoning should not be used to imply that appropriately managed GLP-1 therapy is broadly unsafe.

CLINICAL IMPLICATION

Assess the nutritional and clinical consequences of weight reduction, not only the medication and its dose. Prescribing oversight and ongoing follow-up address different parts of patient safety.

C — Metabolic Innovation and Technology

Patra, Bhandarkar, Kapoor and Kalra | Advances in Therapy, 2026 | 10.1007/s12325-026-03775-1

Patra and colleagues survey digital and procedural technologies in a narrative review of obesity care. They discuss machine-learning risk prediction and adipose-tissue imaging alongside robotic surgery, virtual-reality behavioral interventions, and remote patient support. The procedural discussion separates endoscopic sleeve gastroplasty from robotic-assisted operations and describes mixed comparative results for robotics. Across the review, heterogeneous interventions and limited real-world trial evidence prevent treating technology as a single intervention with a uniform benefit.

A practical starting point for a bariatric program is to name the problem a tool is supposed to solve, such as missed follow-up or limited access to counseling. An app that collects data without a workable feedback process may add tasks without improving care. The review's discussion of cost, access, and privacy supports evaluating implementation as carefully as technical performance. Robotic platforms and digital behavioral tools require separate assessments; enthusiasm for one should not substitute for outcome evidence about the other.

CLINICAL IMPLICATION

Choose a measurable clinical objective before adding a technology. Patient access, clinician oversight, and data protection should be considered alongside efficacy in the adoption decision.

Editorial

Samer G. Mattar, MD, FACS, FASMBS, DABOM | Section A Editor, MBS Digest | Guest editorial

SMART technologies, large language models (LLMs), and virtual-reality (VR)– based behavioral interventions are reshaping obesity and metabolic syndrome care by enabling continuous, adaptive, and personalized management. Yet their rapid integration into clinical practice also introduces important limitations, risks, and implementation challenges that must be addressed to ensure equitable, safe, and effective adoption.

SMART systems

SMART systems (sensor-enabled, mobile, adaptive, real-time tools) are already transforming how clinicians monitor and intervene on metabolic risk. Continuous glucose monitors, wearable accelerometers, and smart scales generate streams of physiologic and behavioral data that can be integrated into clinical decision support platforms. These technologies allow early detection of weight-trajectory inflections, dietary lapses, sleep disturbances, or rising insulin resistance long before they manifest as overt disease.

However, as promising as SMART systems are, their implementation is not without notable drawbacks. High-resolution physiologic data can overwhelm both patients and clinicians, particularly when systems lack streamlined clinical decision support or when alerts are excessive. Data accuracy varies across consumer-grade devices, raising concerns about clinical reliability and potential misinterpretation. Cost and digital-literacy barriers may exacerbate disparities in populations already disproportionately affected by obesity. Many SMART systems require sustained engagement, yet long-term adherence to wearables declines sharply after 3–6 months. Privacy concerns remain a major barrier to patient trust. Without robust regulatory frameworks and interoperability standards, SMART technologies risk creating fragmented, inequitable, and clinician-burdensome ecosystems.

Large language models

Large language models (LLMs) bring a different kind of intelligence; one that is conversational, contextual, and scalable. Their ability to synthesize complex guidelines, personalize behavioral coaching, and maintain engagement makes them powerful adjuncts in obesity care.

LLMs can deliver tailored nutrition guidance, motivational interviewing–style coaching, and adaptive cognitive-behavioral strategies. They can help patients interpret SMART-device data, troubleshoot barriers, and maintain adherence. Importantly, they can scale high-quality counseling to populations that lack access to obesity specialists.

Future iterations may integrate multimodal data from wearables, labs, and imaging to generate personalized metabolic phenotyping, thereby enabling treatment plans that evolve dynamically. When paired with clinician oversight, LLMs could become continuous metabolic-health copilots, extending care beyond the clinic walls.

However, LLMs also pose risks. They may generate inaccurate or overly confident recommendations, particularly when interpreting complex physiologic data or when patients describe symptoms ambiguously. Without clinician oversight, LLM-driven guidance may inadvertently reinforce harmful behaviors or misinterpret SMART-device trends. As has been the case in other clinical implementations, bias embedded in training data may disproportionately affect marginalized populations, leading to inappropriate dietary guidance, misaligned behavioral strategies, or culturally insensitive recommendations. And once again, privacy concerns are heightened when LLMs process sensitive metabolic-health data, all while the regulatory landscape for AI-mediated clinical decision support remains underdeveloped.

A further concern is patient over-reliance on conversational agents. Excessive dependence may reduce human-clinician engagement, potentially undermining therapeutic alliance, which is an essential component of long-term obesity management.

Virtual reality exposure therapy: rewiring behavior

Virtual Reality (VR) adds an immersive behavioral dimension in the management of patients with metabolic syndrome. Exposure-based VR therapy has shown promise in treating binge-eating disorder, emotional eating, and maladaptive food-environment responses. By simulating high-risk settings, such as buffets, grocery stores, and social gatherings, VR allows patients to practice coping strategies in controlled, repeatable environments.

VR can also enhance the adoption of physical activities by transforming exercise into engaging, gamified experiences. It can also support stress-reduction protocols, thereby improving sleep and autonomic balance, both of which are key drivers of insulin resistance.

Yet VR adoption faces significant logistical and clinical barriers. Hardware costs, motion sickness, accessibility limitations, and the need for dedicated physical space can restrict implementation in both clinical and community settings.

VR exposure therapy requires careful clinical supervision; poorly designed environments may inadvertently trigger maladaptive eating behaviors or emotional distress. Evidence for long-term efficacy remains limited, and sustained engagement with VR platforms is variable. Additionally, VR systems generate sensitive behavioral and biometric data, raising concerns about privacy, data ownership, and secure storage.

A unified future: promise tempered by pragmatic considerations

The integration of SMART devices, LLMs, and VR offers a powerful framework for continuous, adaptive metabolic care. Yet their promise must be balanced with thoughtful implementation strategies that address cost, equity, privacy, clinician workload, and patient safety. Clinicians must remain central by providing oversight, contextualizing recommendations, and ensuring that intelligent systems augment rather than replace human judgment.

Despite these challenges, the future is bright. I see an exciting future for the management of metabolic syndrome within a hybrid model, one in which human expertise is amplified by intelligent, immersive technologies, deployed within ethical, equitable, and evidence-based frameworks.

Closing the gaps in knowledge...

Production Editor John D. Scott MD FACS | [email protected]

MBS Digest | September 22, 2026 | Vol. 2, Issue 11
For educational purposes only.