Almanac A1C

Our Perspective on the WHO GLP-1 Guidelines and AI Integration

Posted

by

A1C Medical Team

Table of Contents
Keywords: , , , , ,

Introduction

Obesity is now recognized as a chronic, relapsing disease driven by complex interactions between biology, behaviour and environment, rather than a simple consequence of individual lifestyle choice. Global prevalence has surpassed one billion people and continues to rise across all regions and income levels, contributing to escalating rates of type 2 diabetes, cardiovascular disease, certain cancers, osteoarthritis, sleep apnea, and multiple obesity-related complications. The associated economic burden is substantial, with projections suggesting, trillions of dollars in annual direct and indirect costs and the potential for obesity to consume a substantial share of national health expenditures, thereby threatening the sustainability and capacity of health system worldwide. In this context, there is a pressing need for scalable, lifelong management strategies that move beyond episode counselling to structured. Chronic care models embedded in primary care and universal health coverage frameworks.

In response to these challenges, the World Health Organization (WHO) has issues a global guideline on the use of glucagon-like-peptide-1 (GLP-1) based therapies for adults living with obesity, marking a major milestone in the evolution of obesity care. The guideline issues a conditional recommendation for the long-term use of GLP-1 therapies as pharmacologic treatment for obesity, reflecting moderate and certainty of evidence for clinically meaningful weight loss and broad metabolic and cardiovascular evidence for clinically meaningful weight loss and broad metabolic and cardiovascular benefits, alongside important considerations regarding cost, health-system readiness, and equity of access. A second conditional recommendations supports providing intensive behavioural therapy (IBT) in adults prescribed GLP-1 agents, positioning structured behavioural interventions like encompassing diet, physical activity, and ongoing support as a core component of a comprehensive multimodal clinical algorithm rather than an optional adjunct.

Within this emerging framework, the integration of AI-enabled IBT with GLP-1 pharmacotherapy offers a pragmatic strategy to address key limitations of current care models. Traditional lifestyle programs and clinic-based counselling are often constrained by time, workforce capacity, geography, and variability in quality, leading to gaps in long-term adherence, suboptimal dose titration, and insufficient support for behaviour change between visits. Digital health tools such as mobile applications, telehealth platforms, wearables and decision support system can extend the reach of IBT by enabling continuous monitoring, structured goal setting, and timely feedback on diet, physical activity, and adherence to pharmacotherapy while tailoring recommendations to individual preference, comorbidities and real-world data streams. Embedding AI-driven personalization within WHO’s multimodal paradigm may therefore enhance the durability and equity of GLP-1-based obesity treatment, transforming pharmacologic efficacy into sustained, system level impact.

Conceptual framework: GLP-1 Therapies within a Multimodal Obesity Ecosystem

GLP-1 receptor agonists and newer dual or multi-receptor incretin agonists exert their effects through complementary central and peripheral mechanisms that extend well beyond glucose lowering alone. By enhancing glucose-dependent insulin secretion, suppressing glucagon, delaying gastric emptying, and acting on hypothalamic and brainstem circuits regulating appetite, these agents reduce hunger, increase satiety, and lower energy intake, resulting is sustained weight loss in adults living with obesity. Large randomized trials and meta-analyses have demonstrated that , at adequate doses and treatment durations, GLP-1 based therapies improve multiple cardiometabolic risk factors, including glycemia, blood pressure, and atherogenic lipids, and can reduce major adverse cardiovascular events and other obesity-related comorbidities such as heart failure with preserved ejection fraction, obstructive sleep apnoea, and metabolic dysfunction associated steatohepatitis. However, “medication-only” approaches have clear limitations: treatment effects are attenuated by suboptimal adherence and early discontinuation, access is constrained by cost and supply and weight regain and persistent behavioural drivers of obesity often remain undressed when pharmacotherapy is not embedded within structured lifestyle and system-level intervention [1,2,3].

The recent WHO guideline on the use and indications of GLP-1 therapies explicitly situates these medicines within a broader “obesity ecosystem” that must link pharmacologic innovation to prevention, lifestyle intervention and health system reform. In this framework, obesity is framed as a chronic, relapsing disease requiring lifelong, person-centred care, with GLP-1 agents will depend on strengthening primary care based on chronic care platforms, developing transparent prioritization criteria for high-risk groups, and implementing financing, supply-chain, and service delivery models that prevent widening of health inequities between and within countries. WHO therefore calls for an integrated ecosystem in which pharmacologic advances are coupled with upstream policies targeting food systems and obesogenic environments, digital and community based delivery of behavioural support, and universal health coverage arrangements that comprehensive obesity prevention and care universally available, accessible, affordable, and sustainable [1,4].

Intensive Behavioural Therapy in the WHO GLP-1 Guideline

The WHO guideline defines intensive behavioural therapy (IBT) as a structured, multi-component intervention designed to amplify and sustain the therapeutic effects of GLP-1 pharmacotherapy through evidence-based lifestyle modification. Core elements of IBT, as outlined in the guideline’s second recommendation and good practice statements, include structured goal setting for both physical activity and dietary intake, energy-intake restriction through various means, weekly counselling sessions delivered by trained health professionals or task-share personnel, and routine assessment of progress using standardized metrics such as weight, waist circumference, behavioural adherence, and patient-reported outcomes. the guideline specifies that IBT should be provided as part of comprehensive, multimodal clinical algorithm rather than as a stand-alone or optional add-on, emphasizing that context appropriate counselling on behavioural and lifestyle changes, but not limited to physical activity and healthy dietary practices should be offered as an initial step toward more intensive, structured interventions for all individuals prescribed GLP-1 therapies. This framing positions IBT not merely as general advice, but as a formalized component of chronic obesity care that requires adequate health-system capacity, trained workforce, and integration within primary care platforms and referral pathways [1,4].

Despite the biological plausibility and clinical rationale for combining pharmacotherapy with behavioural support, the WHO guideline development group assigned a conditional recommendation grade to the pairing of GLP-1 agents with IBT, reflecting low certainty in the evidence that IBT meaningfully enhances the efficacy of GLP-1 therapies across the approved agents (liraglutide, semaglutide, and tirzepatide). Systematic reviews conducted to inform the guideline identified limited high-quality data directly comparing GLP-1 monotherapy versus GLP-1 plus IBT, as well as heterogeneity in the definition, intensity, and delivery models of behavioural interventions across trials, making it difficult to isolate the incremental benefit attributable to IBT. Furthermore, the evidence base did not adequately address patient-level variability in weight-outcome priorities, the equity and feasibility implications of implementing resource intensive weekly counselling in diverse health-system contexts, or the long-term durability of combined treatment effects beyond the typical 6-12 month trial endpoints. These uncertainties have important implications for both research design and implementation science: future pragmatic trials should test scalable, real-world delivery models of IBT (including digital and AI-enabled formats), assess subgroup responsiveness and cost-effectiveness across settings, employ longer follow-up periods to capture relapse and maintenance phases, and incorporate implementation outcomes such as fidelity, reach, and sustainability to guide health-system integration and policy decisions [1,4].

AI-Enabled Delivery of Intensive Behavioural Therapy

The delivery of intensive behavioural therapy at scale is constrained by traditional face-to-face models that require specialised staff, facilities, and fixed scheduling, creating barriers to access for many adults with obesity, particularly in under-resourced settings, rural areas, and populations with time or mobility constraints. Digital health tools including telehealth platforms, mobile health (mHealth) applications, wearable activity trackers, and continuous glucose monitoring (CGM) devices offer the potential to extend IBT reach, enhance personalization and provide real-time feedback on behaviours central to weight management. Telehealth-delivered behavioural weight loss interventions have demonstrated non-inferiority to in-person programmes, with randomised trials showing equivalent weight loss when standard face-to-face sessions were replaced by videoconferencing during the COVID-19 pandemic, and high patient and provider satisfaction with remote delivery models. Mobile applications incorporating goal setting, self-monitoring, tailored feedback and structured coaching have shown modest but significant effects on weight loss and BMI reduction in meta-analyses, with greater efficacy when combined with clinician oversight or multi-component interventions. Wearable devices and CGM systems further enable continuous, passive monitoring of physical activity, energy expenditure, sleep, and interstitial glucose levels, generating granular, real-time data streams that support adaptive, individualized guidance and may improve adherence by providing immediate, personalized biofeedback on the metabolic impact of specific foods and behaviours [5,6,7,8,9,10,11,12,13,14,15,16,17,18].

Artificial intelligence and machine learning can augment these digital platforms by automating data integration, identifying patterns in complex multi-dimensional datasets (activity logs, dietary intake, fasting adherence, glucose excursions, self‑reported symptoms), generating dynamic risk stratification and treatment response predictions, and delivering automated yet contextually responsive coaching workflows that adapt to individual progress and preferences. A proposed AI‑assisted care model for pairing GLP‑1 therapy with IBT would therefore integrate multiple data inputs like wearable-derived step counts and heart rate variability, app-based diet and fasting logs, self-monitoring of symptoms and medication adherence, and periodic clinical measurements (weight, waist circumference, blood pressure, laboratory values) into a unified dashboard accessible to both patient and clinician. Machine learning algorithms can then use these inputs to personalize dietary recommendations (for example, identifying individuals food that trigger glucose spikes and tailoring meal timing around fasting windows), adjust physical activity prescriptions to account for adherence trends and metabolic responses, flag early warning signs of non-response or adverse events requiring clinical attention, and deliver just in time adaptive interventions such as motivational messages, reminders, or content modules matched to the patient’s current readiness and barriers. Critically, this AI-enabled model must be designed with robust clinician oversight, ensuring that automated workflows support rather than replace human judgement, that data privacy and algorithmic fairness are protected, and that the system enhances rather than exacerbates health inequities by remaining accessible, culturally appropriate, and responsive to diverse populations and health-system contexts [4,5,6,7,10,12,16,18,19].

Structured Goal Setting for Physical Activity

Evidence-based physical activity guidelines for adults with obesity and related metabolic disease emphasize a multimodal approach that combines aerobic exercise, resistance training, and when appropriate, functional or high-intensity interval training, tailored to individual fitness levels, comorbidities, and treatment goals. Current consensus recommendations, including those from the American College of Sports Medicine, the World Health Organization, and metabolic disease expert panels , advise a minimum of 150 minutes per week of moderate-intensity aerobic exercise (or 75 minutes per week of vigorous-intensity exercise) to initiate weight loss and improve cardiometabolic health, with higher volumes of 300 minutes or more per week recommended to achieve and maintain clinically significant weight reduction. Aerobic training such as brisk walking, cycling, swimming or structured cardiovascular programs has been shown to reduce total body mass, fat mass, visceral adiposity, blood pressure and insulin resistance, and to improve cardiorespiratory fitness and lipid profiles in adults with obesity and metabolic syndrome. In parallel, resistance training performed at moderate to high intensity on at least two days per week is critical for preserving or increasing lean body mass during energy restriction, enhancing muscle strength and functional capacity, and improving insulin sensitivity and  glucose homeostasis, even in the absence of large reductions in body weight. Meta-analyses demonstrate that combined aerobic and resistance exercise programs produce the greatest overall benefits across body composition outcomes (reducing body fat percentage and whole-body fat mass while maintaining or increasing lean mass), glycemic control and cardiometabolic risk markers, supporting the integration of both modalities into comprehensive obesity treatment plans [20,21,22,23,24,25,26,27].

AI- assisted digital platforms can operationalize these evidence-based recommendations by facilitating the creation, monitoring, and dynamic adaptation of SMART (Specific, Measurable, Achievable, Realistic, Time-bound) physical activity goals that are individualize to each patient’s baseline fitness, comorbidities, preferences, and evolving response to treatment. At the initial assessment, AI algorithms can integrate baseline data, including body composition, cardiovascular risk profile, musculoskeletal limitations, patient-reported exercise history and preferences, and wearable-derived activity patterns to generate personalized exercise prescriptions that balance efficacy with safety and adherence. For example, a patient with obesity and type 2 diabetes may receive an initial goal of accumulating 150 minutes per week of moderate-intensity walking in bouts of at least 10 minutes, paired with two supervised resistance training sessions targeting major muscle groups, with progression threshold clearly defined (e.g., increase walking duration by 10% per week if adherence exceeds 80% and no adverse symptoms are reported). The AI system can then continuously monitor adherence through wearable devices and patient self-report, track intermediate outcomes such as step counts, heart rate variability, exercise session completion, and symptom logs, and automatically adjust goals in response to real-world performance: escalating intensity or volume when targets are consistently met offering modified or alternative activities when barriers emerge (e.g,, substituting water-based exercise for patients with joint pain), and flagging patients who show declining adherence or worsening symptoms for timely clinician review. This adaptive, data-driven approach to goal setting not only enhances personalization and engagement but also provides a scalable mechanism to translate static physical activity guidelines into dynamic, responsive behaviour-change support that evolves alongside the patient’s journey through GLP-1 therapy, weight loss, and metabolic improvement [23,28,29,30].

Structured Dietary Goal Setting with Allulose and Food-Based Strategies

D-allulose, a rare monosaccharide and C-3 epimer of D-fructose, is emerging as a promising functional ingredient in the dietary management of obesity and metabolic disease due to its unique metabolic and endocrine properties. Despite sharing the same molecular formula as glucose (C6H12O6), allulose is poorly absorbed in the small intestine and provides minimal caloric value (approximately 0.2-0.4kcal/g compared to 4 kcal/g for sucrose), yet it retains approximately 70% of the sweetness of table sugar, making it an attractive low-calorie sweetener for sugar substitution without compromising palatability. Clinical trials in healthy adults and individuals with impaired glucose tolerance or type 2 diabetes have demonstrated that co-ingestion of 5-10 grams of allulose with carbohydrate, containing meals significantly attenuates postprandial glucose excursions in a dose-dependent manner, with reductions in both peak glucose levels and overall glucose area under the curve at 30 to 120 minute post-ingestion. Longer-term supplementation studies have further shown that allulose-enriched diets reduce body weight gain, improve insulin sensitivity, prevent hepatic triglyceride accumulation, and promote favourable changes in adiposity and metabolic markers compared to isocaloric diets containing conventional sugars or non-nutritive sweeteners [31,32,33,34].

The mechanisms underlying allulose’s glycemic and metabolic benefits are multi-factorial and complementary to the pharmacologic actions of GLP-1 therapies. First, allulose inhibits intestinal a- glucosidase activity, thereby slowing the digestion and absorption of complex carbohydrates and reducing the rate of glucose entry into the bloodstream following mixed-meal ingestion. Second, allulose and glucose complete for shared intestinal transporters (GLUT2 and GLUT5), such that the presence of allulose delays glucose absorption and effectively blunts postprandial glycemia through substrate competition. Third, allulose stimulates hepatic glycogen synthesis and promotes translocation of hepatic glucokinase, thereby enhancing glucose clearance from the circulation and buffering blood glucose fluctuations. Fourth, and of particular relevance in the context of GLP-1 based obesity treatment, allulose potently stimulates endogenous GLP-1 secretion from intestinal L-cells through a novel mechanism involving intestinal distension: because allulose is slowly and incompletely absorbed, it increases luminal volume and mechanically distends the intestine, triggering GLP-1 release via stretch-sensitive pathways independent of nutrient metabolism. This effect is dose-dependent, positively correlated with intestinal content volume, and can be augmented by co-administration with other poorly absorbable solutes or carbonated beverages, suggesting that strategic meal structuring with allulose may synergistically enhance the satiety, appetite suppression, and glycemic benefits of exogenous GLP-1 pharmacotherapy. Practice guidance for substituting added sugars with allulose includes replacing 5-10 grams of table sugar or high-fructose corn syrup with an equivalent amount of allulose is sweetened beverages (coffee, tea, smoothies), baked goods (muffins, cookies, cakes), breakfast cereals, yogurt, and desserts, with attention to adjusting recipes for allulose’s lower sweetness intensity and different moisture retention properties compared to sucrose [31,32,34].

Within an AI-enabled behavioural therapy framework paired with GLP-1 pharmacotherapy, structured dietary goals can leverage both allulose substitution and evidence-based meal structuring principles to optimize adherence, metabolic outcomes, and long-term weight maintenance. Digital platforms can guide patients to systematically reduce intake of ultra-processed foods (UPFs), industrially formulated products that are energy-dense, nutrient-poor, high in refined sugars, unhealthy fats, and sodium, and low in fiber and protein, which have been consistently associated with increased obesity risk, higher BMI, greater waist circumference, and excess energy intake in cross-sectional and prospective cohort studies. AI-driven meal planning tools can provide personalized, actionable recommendations such as replacing sugar-sweetened beverages and confectionery with allulose-sweetened alternatives, restructuring meals to prioritize high-protein foods (lean meats, fish, legumes, dairy)and high-fiber sources (vegetables, whole grains, fruits) that enhance satiety and GLP-1 secretion, and delivering behavioural nudges, such as pre-meal reminders, shopping list generation, and recipe suggestions that reduce reliance on UPFs while capitalizing on GLP-1-induced reductions in hunger and food cravings. By integrating continuous glucose monitoring data, AI systems can further personalize allulose dosing and meal composition in real time, identifying individual glycemic responses to specific foods and iteratively refining dietary goals to achieve optimal postprandial glucose control, sustained energy balance, and alignment with patient preferences and cultural food practices. This technology-enabled, food-based approach transforms abstract dietary guidelines into concrete, individualized, and dynamically adaptive eating patterns that synergize with pharmacotherapy to support durable weight loss and metabolic health [35,36,37,38,39,40,41].

Energy Intake Restriction: Intermittent, Fat and Sardine Fasting

Intermittent fasting (IF) encompasses a diverse range of dietary strategies that manipulate the timing and frequency of energy intake rather than focusing exclusively on caloric quantity or macronutrient composition, with the most commonly studied patterns including alternate-date fasting (ADF), time-restricted eating (TRE), the 5:2 diet (two non-consecutive days of severe energy restriction per week), and weekly whole-day fasts (WDF). A recent network meta-analysis of 99 randomized controlled trials involving 6,582 adults across varying health conditions demonstrated that all IF and continuous energy restriction (CER) strategies significantly reduced body weight compared with ad-libitum diets, with moderate certainty evidence that alternate-day fasting produced superior weight loss compared to CER (mean difference -1.29kg, 95% CI -1.99 to -0.59) in trials with durations less than 24 weeks. In moderate to long term trials (³24 weeks), ADF, TRE, and CER all showed small but significant reductions in body weight (mean difference range -1.88 to -3.63 kg) versus ad-libitum, with no statistically significant differences between IF strategies and CER in sustained weight reduction, though IF approaches demonstrated preferential benefits on certain cardiometabolic markers including total cholesterol, triglycerides and non-HDL cholesterol. Systematic reviews and meta-analyses have further shown that IF, particularly when implemented for durations exceeding 12 weeks produces clinically meaningful reductions in waist circumference, fat mass, body mass index, fasting insulin, LDL-cholesterol and preserving fat-free mass, with evidence suggesting that these metabolic improvements may occur through mechanisms beyond simple caloric deficit, including enhanced mitochondrial oxidative capacity, reduced oxidative stress, improved circadian rhythm alignment, and induction of ketosis [42,43,44,45,46,47,48,49,50].

Fat fasting and sardine fasting represent novel, highly restrictive variants of very low-calorie ketogenic diets (VLCKDs) that merge principles of macronutrient manipulation, prolonged ketosis, and food-based simplicity to achieve rapid metabolic shifts without complete food abstention. Fat fasting typically involves consuming 80-90% of calories from dietary fat (primarily medium-chain triglycerides or long-chain fats such as olive oil, avocado, nuts, and butter), while restricting total energy intake to 800-1,200 kcal per day for short durations (3-7 days), thereby inducing rapid hepatic ketogenesis, mobilization of stored body fat, appetite suppression via ketone signalling, and metabolic benefits such as reduced liver fat, improved insulin sensitivity, and preservation of lean body mass despite significant caloric restriction. Sardine fasting, a more recent and empirically documented mono-food protocol, involves consuming exclusively canned sardines (typically 3-5 tins per day, providing approximately 1,000-1,500 kcal) along with water, black coffee or tea, and optional low-carbohydrate condiments (mustard, hot sauce, olive oil, salt) for periods of 3-7 days or intermittently (e.g., one week per month). Sardines provide a nutrient-dense matrix rich in high quality protein (preserving muscle mass during energy deficit), long-chain omega-3 polyunsaturated fatty acids (eicosapentaenoic acid and docosahexaenoic acid, supporting anti-inflammatory and cardioprotective pathways), fat soluble vitamins (D and E), minerals (calcium, selenium, phosphorus), and bioactive compounds (creatine, coenzyme Q10), while remaining extremely low in carbohydrates, thus mimicking fasting-induced ketosis and delivering metabolic benefits including rapid weight loss (2-6 pounds over 3-7 days), elevations in blood ketone levels, ultra-high omega03 indices, enhanced fat oxidation, upregulation of fibroblast growth factor 21 (FGF-21, a longevity-associated metabolic hormone), preserved lean mass, and subjective reports of increased energy, mental clarity, and reduced hunger without the muscle catabolism and micronutrient depletion risk associated with water only fasting [51,52,53,54,55,56,57,58].

When positioned within a GLP-1 plus intensive behavioural therapy framework, fat fasting and sardine fasting can serve a structured, short-term, food-based energy restriction protocols to break weight-loss plateaus, re-establish ketosis, amplify early treatment response, or provide metabolic “resets” during maintenance phases of GLP-1 therapy. However, their implementation requires careful attention to safety, micronutrient adequacy, and patient selection. Sardine fasting protocols should be limited to well-nourished adults without contraindications to very low-carbohydrate diets (e.g., those with adequate baseline lean mass, no history of disordered eating, normal kidney and liver function and no medications requiring carbohydrate intake), with consideration of supplementing nutrients not adequately provided by sardines alone, such as vitamin C, magnesium, and fiber. Similarly, fat fasting should be avoided in individuals with gallbladder disease, severe hyperlipidemia, pancreatitis, or insulin-dependent diabetes without close medical supervision, and long0term adherence to very high-fat ketogenic patterns has been associated in preclinical models with unfavourable metabolic adaptations including hyperlipidemia, hepatic steatosis, glucose intolerance, and impaired insulin secretion, underscoring the importance of short‑term, intermittent application rather than continuous use. Within an AI‑assisted IBT platform paired with GLP‑1 therapy, these advanced fasting strategies can be prescribed selectively, with digital monitoring of adherence, symptoms (nausea, fatigue, gastrointestinal distress), ketone levels (via fingerstick or breath monitoring), body composition changes, and real‑time clinician alerts for adverse events, ensuring that energy restriction remains safe, personalized, and aligned with each patient’s metabolic phenotype, treatment goals, and tolerance for dietary restriction [4,10,49,51,52,55,56,58].

Weekly Counselling and Routine Assessment: Human-AI Hybrid Model

Weekly counselling and routine assessment are central operational elements of intensive behavioural therapy (IBT) in the WHO GLP‑1 guideline, which frames obesity as a chronic, relapsing disease requiring structured, longitudinal support rather than sporadic lifestyle advice. In a WHO‑aligned model, weekly IBT sessions are typically delivered by a multidisciplinary team (physician, nurse, dietitian, psychologist, health coach) and focus on four core functions: structured counselling on diet and physical activity consistent with national guidelines; systematic review of previously agreed SMART goals for energy intake, food quality, physical activity, and medication adherence; troubleshooting barriers such as environmental triggers, psychosocial stressors, financial constraints, or treatment side‑effects; and iterative adjustment of the treatment plan, including modification of GLP‑1 dose, fasting protocols, allulose substitution strategies, physical activity prescriptions, and relapse‑prevention plans. Each session reinforces person‑centred care by eliciting patient values and preferences, monitoring for weight stigma and discrimination, and integrating management of obesity‑related comorbidities (e.g., type 2 diabetes, hypertension, sleep apnoea) into a unified clinical algorithm rather than siloed disease programmes. Consistent with WHO good practice statements, these encounters are embedded within primary‑care or community‑based platforms and designed to be scalable through task‑sharing, standardized curricula, and the use of digital tools to extend reach in resource‑constrained settings [1,4,,59].

Artificial intelligence can operationalize routine assessment within this human–AI hybrid IBT model by transforming raw patient‑generated health data into actionable insights for both clinicians and patients. Standardized metrics, including weight, and waist circumference, blood pressure, heart rate, fasting and postprandial glycaemia, medication adherence, physical activity (steps, minutes of moderate‑to‑vigorous activity), sleep, and patient‑reported outcomes (PROs) such as hunger, satiety, mood, fatigue, and treatment satisfaction can be collected via connected scales, wearables, continuous glucose monitors, mobile apps, and brief electronic questionnaires between visits. AI‑driven platforms aggregate these data into clinician dashboards that present longitudinal trajectories, risk scores, and alerts, enabling the care team to rapidly identify patterns such as weight‑loss plateaus, declining adherence to GLP‑1 injections or fasting protocols, deterioration in glycaemic control, rising blood pressure, or worsening mood and sleep quality. Machine‑learning algorithms can be trained to flag non‑response (e.g., <5% weight loss after 3–6 months despite adequate medication exposure), emerging side‑effects (e.g., gastrointestinal symptoms temporally linked to dose escalation), and psychosocial risk (e.g., depressive symptom trajectories, increased binge‑eating episodes, or social isolation signals) to prompt proactive outreach or treatment modification. Within weekly sessions, clinicians and coaches then use these AI‑generated summaries to guide shared decision making reviewing progress in a data-driven yet empathetic manner, prioritizing issues that algorithms have algorithms have highlighted, and co‑creating updated goals and action plans that remain aligned with WHO’s person‑centred, equity‑focused vision for GLP‑1–based obesity care [1,4,5,10,16,59,60].

Health System, Equity and Implementation Considerations

Embedding GLP‑1 therapies combined with AI‑enabled intensive behavioural therapy (IBT) into primary care requires deliberate health‑system strengthening across workforce, information, and financing domains. The WHO guideline emphasizes that effective obesity care depends on trained multidisciplinary teams, robust supply chains, patient registries, and integrated chronic‑care platforms rather than isolated specialist services. Workforce training must therefore extend beyond pharmacology to include competencies in obesity as a chronic disease, shared decision‑making, structured behavioural counselling, digital‑tool use, and interpretation of AI‑generated risk scores. Primary‑care‑based registries and standardized referral pathways are needed to identify eligible patients, track longitudinal outcomes (weight, comorbidities, treatment exposure), and coordinate escalation to specialized services such as bariatric surgery or mental health care. Digital infrastructure including secure electronic health records, interoperability with mobile apps, wearables, and telehealth platforms, and governance frameworks for data privacy and algorithm oversight is essential to support AI‑driven IBT workflows at scale, while sustainable financing (through universal health coverage schemes, national insurance, or targeted subsidy programmes) is required to ensure that GLP‑1 medicines and associated digital services are affordable and integrated into essential‑care benefit packages [1,4,59].

Applying an equity lens is critical because GLP‑1 availability and cost, as well as access to digital technologies, are unevenly distributed within and between countries. WHO highlights a “triple challenge” of constrained production capacity, high prices, and health‑system preparedness, warning that without intentional design, GLP‑1 therapies risk widening existing disparities in obesity and non‑communicable disease outcomes. AI‑enabled IBT can help expand access through task‑sharing models (e.g., nurse‑ or community health worker–led digital coaching supported by decision‑support algorithms), low‑cost smartphone or SMS‑based interventions, and telehealth that reaches rural or underserved populations, but these approaches must be accompanied by strategies to address digital divides such as limited connectivity, device affordability, and varying levels of digital literacy. Algorithm development should proactively mitigate bias by using diverse training datasets, auditing model performance across demographic and socioeconomic subgroups, and involving people living with obesity in co‑design to ensure cultural relevance and acceptability. In parallel, transparent, evidence‑based prioritization frameworks are needed to direct scarce GLP‑1 supplies toward high‑risk groups such as individuals with obesity plus established cardiovascular disease, diabetes, or chronic kidney disease, while gradually expanding eligibility as production, affordability, and delivery capacity improve, aligning pharmacologic innovation with the goal of universal, equitable access to comprehensive obesity care [1,4,10,16,59,60].

Research Agenda and Future Directions

The WHO guideline underscores that the current evidence base for GLP‑1 therapies, particularly when combined with intensive behavioural therapy (IBT) is limited by relatively short follow‑up, highly selected trial populations, and uncertainty regarding long‑term safety, durability of weight loss, and real‑world implementation. Priority research questions therefore include: the long‑term cardiometabolic, renal, and neurocognitive outcomes of sustained GLP‑1 plus AI‑supported IBT; the comparative effectiveness and safety of different dietary strategies (e.g., intermittent fasting variants, ketogenic or fat‑focused protocols, sardine‑based or other food‑based fasting mimetics, and allulose‑enriched diets) used adjunctively with GLP‑1s; and the identification of predictors of differential response, encompassing genetics, baseline metabolic phenotypes, microbiome signatures, behavioural profiles, and social determinants of health. Parallel economic and implementation research is needed to assess cost‑effectiveness, budget impact, and affordability of GLP‑1 plus digital IBT across diverse health‑system settings, including low‑ and middle‑income countries where resource constraints and competing priorities may limit scale‑up [1,4,42,48].

Addressing these gaps will require a shift from traditional explanatory trials to pragmatic and adaptive study designs that test multimodal clinical algorithms under real‑world conditions. Such trials should evaluate integrated care pathways that combine GLP‑1 pharmacotherapy with structured dietary patterns (including allulose substitution and fasting protocols), evidence‑based physical activity prescriptions, and AI‑enabled digital tools for monitoring, coaching, and decision support, while explicitly incorporating health‑system constraints such as workforce availability, digital infrastructure, reimbursement models, and medication supply limits. Adaptive platform trials and learning health‑system frameworks could allow iterative refinement of these algorithms, dropping ineffective components, tailoring intensity to patient risk strata, and testing step‑up or step‑down strategies for GLP‑1 dosing and behavioural support while embedding equity outcomes (access, adherence, and benefit across socioeconomic and demographic groups) as core endpoints rather than secondary considerations. Generating this type of context‑sensitive, implementation‑oriented evidence is essential to transform GLP‑1–based obesity treatment from a promising pharmacologic innovation into a scalable, fair, and sustainable global obesity‑care ecosystem [1,4,5,10].

Conclusion

Effective management of obesity in the GLP-1 ear requires moving beyond a narrow pharmacocentric paradigm toward a genuinely multimodal person-centred model of chronic disease car. GLP-1 based therapies and related incretin agents have transformed expectations by producing clinically meaningful weight loss and broad cardiometabolic benefits in high-risk adults, yet medication alone cannot resolve a condition rooted in complex biological, behavioural, social, and environmental determinants. Pairing pharmacotherapy with structured behavioural interventions including goal-directed nutritional strategies, physical activity prescription, and including goal-directed nutritional strategies, physical activity prescription, and longitudinal counselling, supported by digital and AI-enabled tools, provides a concrete way to operationalize the WHO vision of an integrated “obesity ecosystem” that embeds prevention, treatment and equity within routine health system functions.

Realizing this vision will depend on coordinated action from clinicians, health-tech innovators, and policymakers to design, implement and evaluate scalable care models that integrate GLP-1 therapy with intensive behavioural support an digital technologies. Frontline providers and multidisciplinary teams must adopt chronic-care workflows that incorporate AI-assisted risk stratification, remote monitoring and adaptive coaching while safeguarding person-centredness and shared decision making. Technology developers should prioritize interoperable, low-burden solutions that augment clinical judgement, reduce inequities in access to counselling, and are rigorously tested in diverse populations and settings. Policymakers and payers, in turn, are tasked with creating regulatory, reimbursement, and data-governance frameworks that support equitable access to GLP-1 medicines and digitally delivered IBT, alongside investment in implementation research and real-world evaluation. Aligning these efforts can convert the current pharmacological breakthrough into durable population level gains in obesity outcomes, advancing universal, affordable and sustainable obesity care.

References

  1. Celletti F, Farrar J, De Regil L. World Health Organization Guideline on the Use and Indications of Glucagon-Like Peptide-1 Therapies for the Treatment of Obesity in Adults. JAMA [Internet]. 2025 Dec 1; Available from: https://jamanetwork.com/journals/jama/fullarticle/2842199?guestAccessKey=c54b0a37-ccd5-4ecc-b541-764413e030c5&utm_source=for_the_media&utm_medium=referral&utm_campaign=ftm_links&utm_content=tfl&utm_term=120125
  2. Ezpeleta M, Cienfuegos S, Lin S, Pavlou V, Gabel K, Tussing-Humphreys L, et al. Time-restricted eating: Watching the clock to treat obesity. Cell Metabolism [Internet]. 2024 Jan 3; Available from: https://www.sciencedirect.com/science/article/abs/pii/S1550413123004540?via%3Dihub
  3. Neale T. WHO Endorses Treating Obesity in Adults With GLP-1s [Internet]. TCTMD.com. 2025 [cited 2025 Dec 9]. Available from: https://www.tctmd.com/news/who-endorses-treating-obesity-adults-glp-1s
  4. WHO issues global guideline on the use of GLP-1 medicines in treating obesity [Internet]. Who.int. World Health Organization: WHO; 2025. Available from: https://www.who.int/news/item/01-12-2025-who-issues-global-guideline-on-the-use-of-glp-1-medicines-in-treating-obesity
  5. Teke J, Msiska M, Adanini OA, Egbon E, Osborne A, Olawade DB. Artificial intelligence for obesity management: A review of applications, opportunities, and challenges. Obesity Medicine [Internet]. 2025 Oct 15;58:100657. Available from: https://www.sciencedirect.com/science/article/pii/S2451847625000776
  6. Heerman WJ, Rothman RL, Sanders LM, Schildcrout JS, Flower KB, Delamater AM, et al. A Digital Health Behavior Intervention to Prevent Childhood Obesity. JAMA [Internet]. 2024 Nov 3;332(24). Available from: https://jamanetwork.com/journals/jama/fullarticle/2825869
  7. KOO TH, LEONG XB, NG JK, TAN NKB, MOHAMED M. Emerging Trends in Telehealth and AI-Driven Approaches for Obesity Management: A New Perspective. Malaysian Journal of Medical Sciences. 2025 Feb 28;32(1):26–34.
  8. Yang X, Liu X, Guo Y. Continuous Glucose Monitoring‐Guided Self‐Management Leads to Diabetes Remission in a Newly Diagnosed Overweight Patient With Type 2 Diabetes: A Brief Report. Health Care Science. 2025 Sep 28;4(5):350–4.
  9. Bertoli S, Paolo Capodaglio, Colosimo S, Silvana R, Gilardini L, Bruno A, et al. Effectiveness of a Digital Therapy on 6-Month Weight Loss in People With Obesity: The Digital Therapy to Promote Weight Loss in Patients With Obesity by Increasing Their Adherence to Treatment (DEMETRA) Randomized Clinical Trial. PubMed. 2025 Oct 21;27:e72054–4.
  10. Kalantar HT, Kalantar AT, Alali M, Alali N, Naji H. Effectiveness of Telehealth and Wearable Device-Based Interventions for Managing Childhood and Adolescent Obesity: A Systematic Review and Meta-Analysis. Cureus. 2025 Oct 14;
  11. Constantino AK. FDA approves first-ever glucose monitoring system for weight loss from Signos [Internet]. CNBC. 2025. Available from: https://www.cnbc.com/2025/08/20/fda-approves-signos-glucose-monitoring-for-weight-loss.html
  12. New college study finds personalized dietary recommendations combined with continuous glucose monitoring helps individuals with prediabetes lose more weight [Internet]. College of Public Health. 2023. Available from: https://publichealth.gmu.edu/news/2024-09/new-college-study-finds-personalized-dietary-recommendations-combined-continuous
  13. Kim J, Kang M, Song kwangho, Ahn H, Park YK. Application of personalized nutrition counseling according to glycemic response in obese adults: A randomized dietary intervention study. Nutrition. 2025 Mar;131:112641.
  14. Raedeh Basiri, Cheskin LJ. Personalized Nutrition Therapy without Weight Loss Counseling Produces Weight Loss in Individuals with Prediabetes Who Are Overweight/Obese: A Randomized Controlled Trial. Nutrients. 2024 Jul 11;16(14):2218–8.
  15. Ross KM, Carpenter CA, Arroyo KM, Shankar MN, Yi F, Qiu P, et al. Impact of transition from face‐to‐face to telehealth on behavioral obesity treatment during the COVID‐19 pandemic. Obesity. 2022 Mar 11;30(4):858–63.
  16. Pujia C, Ferro Y, Mazza E, Maurotti S, Montalcini T, Pujia A. The Role of Mobile Apps in Obesity Management: Systematic Review and Meta-Analysis. Journal of Medical Internet Research. 2025 May 6;27:e66887.
  17. Balázs Erdős, O’Donovan SD, Adriaens ME, Anouk Gijbels, Trouwborst I, Jardon KM, et al. Leveraging continuous glucose monitoring for personalized modeling of insulin-regulated glucose metabolism. Scientific reports. 2024 Apr 5;14(1).
  18. Evolve: Menopause & CardioMetabolic Clinic [Internet]. Evolve: Menopause & CardioMetabolic Clinic. 2025 [cited 2025 Dec 9]. Available from: https://www.evolvelifeclinic.com/news/how-continuous-glucose-monitoring-cgm-personalized-nutrition-improve-weight-loss-and-metabolic-health
  19. Lee H, Hwang J, Yon DK, Rhee SY. Multimodal and Multidimensional Artificial Intelligence Technology in Obesity. Journal of Obesity & Metabolic Syndrome. 2025 Sep 9;34(4):394–404.
  20. Fernández-Alonso M, Bejarano G, Creel DB, Kohl HW, Messiah SE, Altieri MS, et al. Expert-based physical activity guidelines for metabolic and bariatric surgery patients: a systematic review of randomized controlled trials. Surgery for Obesity and Related Diseases [Internet]. 2025 May [cited 2025 Aug 28];21(5):606–14. Available from: https://www.soard.org/article/S1550-7289(24)00911-0/abstract?utm_source=chatgpt.com
  21. Willis LH, Slentz CA, Bateman LA, Shields AT, Piner LW, Bales CW, et al. Effects of aerobic and/or resistance training on body mass and fat mass in overweight or obese adults. Journal of Applied Physiology [Internet]. 2012 Dec 15;113(12):1831–7. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3544497/
  22. Hang S, Xiaoyu L, Jue W, Yingli L, Li Z. Effects of resistance training and aerobic training on improving the composition of middle-aged adults with obesity in an interventional study. Scientific Reports [Internet]. 2025 Sep 30;15(1). Available from: https://www.nature.com/articles/s41598-025-11076-w
  23. Chomiuk T, Niezgoda N, Mamcarz A, Śliż D. Physical activity in metabolic syndrome. Frontiers in physiology [Internet]. 2024 Feb 19;15. Available from: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC10910017/#:~:text=Inadequate energy balance resulting from
  24. Liu X, Li Q, Lu F, Zhu D. Effects of aerobic exercise combined with resistance training on body composition and metabolic health in children and adolescents with overweight or obesity: systematic review and meta-analysis. Frontiers in Public Health. 2024 Aug 9;12.
  25. Lopez P, Taaffe DR, Galvão DA, Newton RU, Nonemacher ER, Wendt VM, et al. Resistance training effectiveness on body composition and body weight outcomes in individuals with overweight and obesity across the lifespan: A systematic review and meta‐analysis. Obesity Reviews. 2022 Feb 21;23(5).
  26. Joao Victor Esteves, Stanford KI. Exercise as a tool to mitigate metabolic disease. American journal of physiology Cell physiology. 2024 Jul 9;
  27. Johnson NA, Sultana RN, Brown WJ, Bauman AE, Gill T. Physical Activity in the Management of Obesity in adults: a Position Statement from Exercise and Sport Science Australia. Journal of Science and Medicine in Sport. 2021 Jul 31;24(12).
  28. Hurley L, Nezami BT, Sciamanna C, Tate DF. Personal goal setting eHealth component associated with improved weight loss at 6 months: A mixed methods secondary analysis. Digital Health. 2024 Jan 1;10.
  29. Zhu Y, Long Y, Wang H, Kun Pyo Lee, Zhang L, Stephen Jia Wang. Digital Behavior Change Intervention Designs for Habit Formation: Systematic Review. JMIR Journal of medical internet research/Journal of medical internet research. 2024 May 24;26(19):e54375–5.
  30. Anderton K. How Utilizing SMART Goals Helps Improve Patient Care [Internet]. Thoroughcare.net. 2023. Available from: https://www.thoroughcare.net/blog/patients-develop-healthy-behaviors-smart-goals
  31. Tani Yuma, Tokuda M, Nishimoto N, Yokoi H, Izumori K. Allulose for the attenuation of postprandial blood glucose levels in healthy humans: A systematic review and meta-analysis. PLOS ONE. 2023 Apr 6;18(4):e0281150–0.
  32. Cayabyab KB, Shin MJ, Heimuli MS, Kim IJ, D’Agostino DP, Johnson RJ, et al. The Metabolic and Endocrine Effects of a 12-Week Allulose-Rich Diet. Nutrients [Internet]. 2024 Jan 1;16(12):1821. Available from: https://www.mdpi.com/2072-6643/16/12/1821
  33. gov. 2025 [cited 2025 Dec 9]. Available from: https://clinicaltrials.gov/study/NCT02988999?intr=erythritol&rank=8
  34. Franchi F, Yaranov DM, Rollini F, Rivas A, Rios JR, Been L, et al. Effects of D-allulose on glucose tolerance and insulin response to a standard oral sucrose load: results of a prospective, randomized, crossover study. BMJ Open Diabetes Research and Care [Internet]. 2021 Feb 1 [cited 2022 Mar 11];9(1):e001939. Available from: https://drc.bmj.com/content/9/1/e001939
  35. Shim JS. Ultra-Processed Food Consumption and Obesity: A Narrative Review of Their Association and Potential Mechanisms. Journal of Obesity & Metabolic Syndrome. 2025 Jan 17;34(1):27–40.
  36. Al Hourani H, Shhadeh HA, Al-Jawaldeh A. Association between consumption of ultra processed foods and obesity among Jordanian children and adolescents. Scientific Reports. 2025 Mar 18;15(1).
  37. Alomari WD, Almoraie NM. Ultra-processed food intake and its association with obesity risk factors, Mediterranean diet, and nutrient intake of adults. Frontiers in Nutrition. 2025 Jul 24;12.
  38. New college study finds personalized dietary recommendations combined with continuous glucose monitoring helps individuals with prediabetes lose more weight [Internet]. College of Public Health. 2023. Available from: https://publichealth.gmu.edu/news/2024-09/new-college-study-finds-personalized-dietary-recommendations-combined-continuous
  39. Balázs Erdős, O’Donovan SD, Adriaens ME, Anouk Gijbels, Trouwborst I, Jardon KM, et al. Leveraging continuous glucose monitoring for personalized modeling of insulin-regulated glucose metabolism. Scientific reports. 2024 Apr 5;14(1).
  40. Hamano S, Sawada M, Aihara M, Sakurai Y, Sekine R, Satoshi Usami, et al. Ultra‐processed foods cause weight gain and increased energy intake associated with reduced chewing frequency: A randomized, open‐label, crossover study. Diabetes Obesity and Metabolism [Internet]. 2024 Sep 12;26(11). Available from: https://pubmed.ncbi.nlm.nih.gov/39267249/
  41. WHO issues global guideline on the use of GLP-1 medicines in treating obesity [Internet]. Who.int. World Health Organization: WHO; 2025. Available from: https://www.who.int/news/item/01-12-2025-who-issues-global-guideline-on-the-use-of-glp-1-medicines-in-treating-obesity
  42. Semnani-Azad Z, Khan TA, Chiavaroli L, Chen V, Bhatt HA, Chen A, et al. Intermittent fasting strategies and their effects on body weight and other cardiometabolic risk factors: systematic review and network meta-analysis of randomised clinical trials. BMJ. 2025 Jun 18;389:e082007.
  43. Rodríguez MM, Muñoz AM, Franco Hoyos K. Intermittent fasting as a nutritional tool. BMJ. 2025 Jun 18;389:r1156.
  44. Wang B, Wang C, Li H. The impact of intermittent fasting on body composition and cardiometabolic outcomes in overweight and obese adults: a systematic review and meta-analysis of randomized controlled trials. Nutrition Journal. 2025 Jul 30;24(1).
  45. Zhang S, Sun B, Sun L, Zou S, Chen Q. Effect of intermittent fasting on obesity and metabolic indices in patients with metabolic syndrome: a systematic review and meta analysis. BMC Endocrine Disorders. 2025 May 14;25(1).
  46. Chen S, Zhang X, Kortas J, Liu H. Effects of time-restricted eating on body composition and metabolic parameters in overweight and obese women: a systematic review and meta-analysis. Frontiers in Nutrition [Internet]. 2025 Sep 16;12. Available from: https://pmc.ncbi.nlm.nih.gov/articles/PMC12479299/
  47. Bakhsh J, Salvy SJ, Vidmar AP. Intermittent fasting as a treatment for obesity in young people: a scoping review. npj Metabolic Health and Disease. 2024 Dec 30;2(1).
  48. Sun ML, Yao W, Wang XY, Gao S, Varady KA, Forslund SK, et al. Intermittent fasting and health outcomes: an umbrella review of systematic reviews and meta-analyses of randomised controlled trials. EClinicalMedicine (Oxford). 2024 Apr 1;70:102519–9.
  49. Nikhila Chelikam, Akella SA, Manisha Rakesh Lakhanpal, Simmy Lahori, Singh A, Zafar S, et al. Role of Ketogenic Diets and Intermittent Fasting in Neurologic Diseases, Cancers, and Obesity: A Systematic Review of Human Studies. Journal of Endocrinology and Metabolism [Internet]. 2024;14(3):103–27. Available from: https://jofem.org/index.php/jofem/article/view/944/284284712?form=MG0AV3
  50. Xie Y, Zhou K, Shang Z, Bao D, Zhou J. The Effects of Time-Restricted Eating on Fat Loss in Adults with Overweight and Obese Depend upon the Eating Window and Intervention Strategies: A Systematic Review and Meta-Analysis. Nutrients. 2024 Oct 5;16(19):3390–0.
  51. Gallop MR, Vieira RFL, Mower PD, Matsuzaki ET, Liou W, Smart FE, et al. A long-term ketogenic diet causes hyperlipidemia, liver dysfunction, and glucose intolerance from impaired insulin secretion in mice. Science Advances. 2025 Sep 19;11(38).
  52. Dabbs R. Sardine Fasting Is the Latest Health Trend – But Experts Warn It’s a Terrible Idea [Internet]. Men’s Health. 2025. Available from: https://www.menshealth.com/uk/nutrition/food-drink/a69014131/sardine-fasting-diet/
  53. Zhu H, Bi D, Zhang Y, Kong C, Du J, Wu X, et al. Ketogenic diet for human diseases: the underlying mechanisms and potential for clinical implementations. Signal Transduction and Targeted Therapy [Internet]. 2022 Jan 17;7(1). Available from: https://www.nature.com/articles/s41392-021-00831-w
  54. Biesiekierska M, Strigini M, Śliwińska A, Pirola L, Balcerczyk A. The Impact of Ketogenic Nutrition on Obesity and Metabolic Health: Mechanisms and Clinical Implications. Nutrition Reviews [Internet]. 2025 Feb 26; Available from: https://academic.oup.com/nutritionreviews/advance-article/doi/10.1093/nutrit/nuaf010/8043247
  55. Bardolf D. Man ate 1,000 sardines in a month, lost 6 pounds and “smelled like a fish market” [Internet]. Fox News. 2025. Available from: https://www.foxnews.com/food-drink/man-ate-1000-sardines-month-lost-6-pounds-smelled-like-fish-market
  56. Masood W, Uppaluri KR, Annamaraju P, Khan Suheb MZ. Ketogenic diet [Internet]. National Library of Medicine. StatPearls Publishing; 2023. Available from: https://www.ncbi.nlm.nih.gov/books/NBK499830/
  57. Nikhila Chelikam, Akella SA, Manisha Rakesh Lakhanpal, Simmy Lahori, Singh A, Zafar S, et al. Role of Ketogenic Diets and Intermittent Fasting in Neurologic Diseases, Cancers, and Obesity: A Systematic Review of Human Studies. Journal of Endocrinology and Metabolism [Internet]. 2024;14(3):103–27. Available from: https://jofem.org/index.php/jofem/article/view/944/284284712?form=MG0AV3
  58. Rizzo S. All Things Ketones, How to Boost Cognition, Sardine Fasting, Diet Rules, & More | Dr. Dom D’Agostino on The Tim Ferriss Show Ep. 825 • Podcast Notes [Internet]. Podcast Notes. 2025 [cited 2025 Dec 9]. Available from: https://podcastnotes.org/tim-ferris-show/all-things-ketones-how-to-boost-cognition-sardine-fasting-diet-rules-more-dr-dom-dagostino-on-the-tim-ferriss-show-ep-825/
  59. Marthe Isaksen Aukan, Larsen MA, Melan TI, Salvesen ØO. Beyond weight loss: digital therapeutic for behavioral change and psychological well-being for individuals with overweight and obesity in a primary healthcare setting—A randomized controlled pilot study. Frontiers in Digital Health. 2025 Sep 16;7.
  60. Flølo TN, Tørris C, Riiser K, Almendingen K, Chew HSJ, Fosså A, et al. Digital health interventions to treat overweight and obesity in children and adolescents: An umbrella review. Obesity Reviews. 2025 Feb 19;