Prediabetes: Epidemiology, diagnosis, and management

Abstract

Prediabetes describes impaired glucose regulation below the diagnostic threshold for diabetes and represents an important stage for prevention. In 2024, an estimated 635 million adults worldwide had impaired glucose tolerance and 488 million had impaired fasting glucose, indicating a large at-risk population. Prediabetes increases the risk of type 2 diabetes and cardiometabolic complications, but this risk varies across individuals. Diagnosis is mainly based on fasting plasma glucose, the oral glucose tolerance test, and glycated hemoglobin, which assess different aspects of glucose metabolism and do not identify identical populations. This narrative review summarizes the definition, epidemiology, screening, diagnosis, and management of prediabetes. Lifestyle intervention, particularly dietary modification and regular physical activity, remains the main approach. Pharmacological treatment may be considered for selected high-risk individuals, including those with obesity, marked dysglycemia, or a history of gestational diabetes. Metabolic and bariatric surgery may reduce diabetes risk in patients with obesity who meet surgical indications, but it is not a general treatment for prediabetes. Better risk stratification, individualized intervention, and long-term follow-up are needed to improve diabetes prevention and cardiometabolic outcomes.

Key Words: Prediabetes; Diagnosis and prevention; Lifestyle intervention; Pharmacological therapy; Metabolic surgery

Core Tip: Prediabetes is a heterogeneous dysglycemic state with variable risk of progression to type 2 diabetes. Fasting plasma glucose, oral glucose tolerance testing, and glycated hemoglobin identify partly different populations, making risk stratification important. Lifestyle intervention remains the foundation of management, while pharmacological therapy and metabolic and bariatric surgery are appropriate only for selected high-risk individuals. This narrative review integrates diagnostic heterogeneity, glycemic phenotype, and intervention intensity to support more individualized diabetes prevention.

INTRODUCTION

Prediabetes represents impaired glucose regulation before the onset of overt diabetes and includes impaired fasting glucose (IFG), impaired glucose tolerance (IGT), and elevated glycated hemoglobin (HbA1c) according to commonly used diagnostic criteria[1]. Its public health burden is considerable. In 2024, an estimated 635 million adults worldwide had IGT, and 488 million had IFG[2,3]. These figures indicate that a large population has already entered a stage in which diabetes prevention remains possible but requires timely recognition.

The clinical relevance of prediabetes extends beyond its potential progression to type 2 diabetes mellitus (T2DM). Approximately 10% of individuals with prediabetes in the United States develop diabetes each year[4], and prediabetes has also been linked to higher risks of cardiovascular events, mortality, and metabolic abnormalities[5]. However, risk is not uniform across all individuals. Differences in age, body weight, ethnicity, baseline glycemic status, and other metabolic factors may influence disease progression and response to intervention. For this reason, prediabetes should not be managed simply as a single diagnostic label. Early identification needs to be followed by risk assessment and appropriate intervention.

Current diagnosis mainly depends on fasting plasma glucose (FPG), the oral glucose tolerance test, and HbA1c[1]. These tests assess different aspects of glucose metabolism and do not fully classify the same populations. This diagnostic inconsistency affects prevalence estimates, risk prediction, and the interpretation of prevention trials. It also partly explains why intervention strategies need to be matched to individual risk profiles rather than applied uniformly to all people with prediabetes.

Lifestyle intervention remains the first-line strategy for diabetes prevention. In the Diabetes Prevention Program, intensive lifestyle intervention reduced the incidence of diabetes by 58%, whereas metformin reduced it by 31% compared with placebo[6]. Long-term follow-up from the Da Qing Diabetes Prevention Outcome Study also showed sustained benefits of lifestyle intervention in individuals with IGT[7]. Pharmacological therapy and metabolic surgery may be useful in selected high-risk populations, especially those with obesity or marked metabolic abnormalities, but their long-term efficacy, safety, cost, and adherence issues should be weighed carefully. Based on these considerations, this narrative review examines the definition, epidemiology, screening and diagnosis, and management of prediabetes. Particular attention is given to diagnostic heterogeneity and to how glycemic phenotype and diagnostic method may inform progression risk and the intensity of preventive intervention.

METHODS

This narrative review was informed by a structured PubMed search. The search covered the period from database inception to the final literature search, with no restrictions on publication year, language, or article type. PubMed was the only electronic database searched. The core search was (“Prediabetic State”[MeSH] OR prediabet*[Title/Abstract] OR “impaired fasting glucose”[Title/Abstract] OR “impaired glucose tolerance”[Title/Abstract]). Topic-specific searches combined the core search with terms related to epidemiology (epidemiology, prevalence, incidence, burden); diagnosis and screening [diagnosis, screening, FPG, oral glucose tolerance testing (OGTT), HbA1c, continuous glucose monitoring (CGM)]; exercise (exercise, physical activity, aerobic, resistance); diet (diet, nutrition, Mediterranean, ketogenic, personalized diet); pharmacological treatment (metformin, acarbose, voglibose, thiazolidinediones, liraglutide, semaglutide, tirzepatide); and metabolic and bariatric surgery (MBS) (bariatric surgery, metabolic surgery, RYGB, sleeve gastrectomy (SG), gastric banding, duodenal switch).

Titles and abstracts were reviewed for relevance to the clinical domains covered in this review, and full texts were consulted when needed. Studies unrelated to prediabetes were excluded unless they provided directly relevant evidence where prediabetes-specific evidence was limited. Guidelines, systematic reviews and meta-analyses, randomized controlled trials, large cohort studies, and relevant landmark studies were prioritized. Prediabetes-specific evidence was preferred when available, while evidence from obesity, mixed dysglycemia, or established T2DM populations was treated as indirect evidence. When findings differed, greater weight was given to higher-level evidence and larger studies, while differences in study population, diagnostic criteria, intervention duration, and outcomes were considered. Findings from small or single-center studies were interpreted cautiously. As this was a narrative review, no formal systematic-review screening protocol, risk-of-bias assessment, or evidence grading was used.

DEFINITION OF PREDIABETES

In clinical practice, prediabetes is usually identified through IFG, IGT, or elevated HbA1c. The American Diabetes Association (ADA) classifies FPG of 5.6-6.9 mmol/L, 2-hour plasma glucose of 7.8-11.0 mmol/L during a 75-g OGTT, and HbA1c of 5.7%-6.4% as prediabetic ranges[1]. These criteria capture related but not identical disturbances in glucose metabolism. IFG and IGT also differ in their underlying metabolic defects. IFG is characterized mainly by hepatic insulin resistance and impaired early-phase insulin secretion. In contrast, IGT is more strongly associated with skeletal muscle insulin resistance and defects in both early- and late-phase insulin secretion. Individuals with combined IFG and IGT generally show more severe hepatic and peripheral insulin resistance together with greater β-cell dysfunction[8,9]. These differences help explain the heterogeneity of prediabetes.

This stage is often clinically silent and may go undetected because most individuals have no obvious symptoms. Longitudinal studies suggest an annual progression rate to diabetes of approximately 5%-10%[4]. In the Rotterdam Study, the remaining lifetime risk of progression from prediabetes to diabetes was estimated to be 74% among individuals aged 45 years[10]. Emerging evidence suggests that microvascular changes and increased cardiovascular risk may already be present during prediabetes[11]. In addition, oral health assessments have reported greater periodontal disease severity in individuals with prediabetes, which may have potential diagnostic or screening value[12].

EPIDEMIOLOGY OF PREDIABETES

The epidemiology of prediabetes should be understood in the broader context of diabetes prevention. As an intermediate stage of dysglycemia, prediabetes affects a large population worldwide and contributes substantially to the future burden of T2DM. Its distribution is shaped by demographic aging, urbanization, dietary transition, physical inactivity, obesity, and unequal access to screening and preventive care. These factors do not affect all populations equally. Differences across socioeconomic settings, ethnic groups, and age categories may influence both the prevalence of prediabetes and the risk of progression to T2DM. Therefore, understanding the epidemiological patterns of prediabetes is important for identifying high-risk populations and developing targeted prevention strategies.

Socioeconomic disparities in prediabetic conditions

The distribution of prediabetes differs across countries and income groups. These differences are related not only to population aging and urbanization, but also to unequal access to screening, preventive care, and long-term lifestyle support. Available estimates suggest that most individuals with prediabetes live in low- and middle-income countries, with 72.2% reported in these regions, especially in countries with a medium human development index[13]. Regional evidence also shows variation by development level. A meta-analysis from the Eastern Mediterranean region reported a higher prevalence of prediabetes in countries with a medium human development index than in those with a low index[14]. At the global level, high-income countries currently have the highest estimated prevalence of IGT and IFG, but the largest relative increase is projected to occur in low-income countries by 2045[15].

This pattern is also reflected in national surveys. In China, a national survey conducted from 2015 to 2017 using the 2018 ADA diagnostic criteria reported a prediabetes prevalence of 35.2% among adults, corresponding to about 478 million individuals[16]. In England, repeated cross-sectional data from the Health Survey for England showed that the prevalence of prediabetes increased from 11.6% in 2003 to 35.3% in 2011[17]. A similar problem exists in the United States, where prediabetes is common among older adults but awareness remains low. One report noted that 48.3% of adults aged 65 years or older met criteria for prediabetes based on HbA1c or fasting glucose, whereas only 11.6% were aware of their condition[18]. These findings suggest that socioeconomic status affects not only the burden of prediabetes, but also its detection and subsequent prevention.

Ethnic variations in prediabetes epidemiology

Substantial ethnic differences in the prevalence of prediabetes have been reported across populations worldwide. Many Asian populations have been reported to have higher prediabetes rates than some other regions, which may relate to dietary patterns and lifestyle changes associated with modernization. Gardner et al[19] conducted a large-scale study from 2005 to 2017 that included 400000 adults who met criteria for prediabetes. They reported that South Asian individuals had a 1.8-fold higher risk of progression from prediabetes to T2DM than European White individuals, independent of conventional risk factors[19]. This pattern has also been observed in multi-ethnic populations that include people with South Asian ancestry.

Age-related patterns in prediabetes also appear to differ by ethnicity. United States national surveillance data from 2009 to 2012 indicated that 37% of adults aged ≥ 20 years and 51% of those aged ≥ 65 years met glycemic criteria for prediabetes. The age-associated increase also differed across ethnic groups, with stronger age-prediabetes associations reported in European or White populations than in some other groups.

Age-stratified prevalence patterns of prediabetes

Prediabetes displays a strong age-dependent epidemiological pattern, with increasing age being an independent risk factor[11,13]. Longitudinal surveillance data show a progressive increase in prevalence across age groups. An analysis of the 2011-2012 National Health and Nutrition Examination Survey data reported age-specific prevalence rates of 28.2% in individuals aged 20-44 years and 49.5% in those aged ≥ 65 years, based on composite glycemic criteria (2-hour plasma glucose, FPG, or HbA1c)[20]. This pattern was also observed in the 2005-2008 National Health and Nutrition Examination Survey data, which showed prediabetes in 35% of adults aged ≥ 20 years and 50% of those aged ≥ 65 years, based on fasting glucose/HbA1c thresholds[4].

The relationship between age and prediabetes prevalence reflects well-established biological aging processes, including gradual β-cell dysfunction and progressive insulin resistance. Notably, individuals aged ≥ 65 years have a 2.1-fold higher risk of prediabetes compared to those aged < 45 years, independent of body mass index adjustments.

DIAGNOSIS AND SCREENING OF PREDIABETES

Prediabetes is an intermediate dysglycemic state, with an annual progression rate to diabetes of approximately 5%-10%, and the cumulative risk increases over time[4]. In addition to future diabetes, prediabetes has been associated with a higher burden of cardiometabolic risk and early microvascular or neurological abnormalities, although the strength of evidence differs across outcomes[13]. Because prediabetes is usually asymptomatic, screening can identify individuals at increased metabolic risk before overt diabetes develops. However, the value of screening depends on subsequent risk assessment and effective intervention, rather than detection alone; current evidence has not fully established a direct population-level benefit of screening in the absence of follow-up management[21].

Diagnostic criteria for prediabetes differ across guidelines and organizations. At present, FPG, the 75-g OGTT, and HbA1c are the main diagnostic or screening tools used in clinical practice. These methods reflect different aspects of glucose metabolism and differ in feasibility, sensitivity, and clinical interpretation. Continuous CGM can provide additional information on glucose variability and post-load glucose excursions, but it has not yet been established as a routine diagnostic tool for prediabetes. The main advantages, disadvantages, and typical clinical uses of these tests are summarized in Table 1.

Table 1 Main advantages, disadvantages, and clinical uses of tests for prediabetes screening and diagnosis.

Test	Advantages	Disadvantages	Main purpose
OGTT	Reference standard for diagnosing IGT; detects post-load dysglycemia[1,117]	Time-consuming; requires fasting and at least 2 h of testing; multiple blood samples; affected by recent diet, acute illness, stress, and medications	Diagnose IGT; assess glucose tolerance; clarify borderline or inconclusive cases
FPG	Simple, inexpensive, widely available; requires only one blood sample[1,117]	May miss isolated IGT or post-load hyperglycemia; affected by short-term physiological variation, acute illness, and stress	Diagnose IFG; initial screening in routine practice
HbA1c	No fasting or timed glucose loading required; reflects glycemic exposure over past 2-3 month; relatively low short-term variability[1,23,24]	Affected by altered red cell turnover, hemoglobin variants, anemia, blood loss, kidney disease, pregnancy, HIV, and some medications	Screening and diagnosis according to specific guidelines; risk stratification and follow-up monitoring
CGM	Provides continuous glucose profiles; captures glycemic variability and postprandial excursions; avoids repeated venipuncture[27,28]	High cost and limited access; sensor-related discomfort; interstitial glucose may lag behind blood glucose; diagnostic thresholds for prediabetes are not standardized[28]	Mainly used in diabetes management; may be useful for research or selected high-risk individuals, but not routine prediabetes diagnosis

OGTT: Oral glucose tolerance test; IGT: Impaired glucose tolerance; FPG: Fasting plasma glucose; IFG: Impaired fasting glucose; HbA1c: Glycated hemoglobin; CGM: Continuous glucose monitoring; HIV: Human immunodeficiency virus.

OGTT

The OGTT is performed by administering 75 g of anhydrous glucose dissolved in water to adults after an overnight fast (for children, 1.75 g/kg body weight, up to a maximum of 75 g). Plasma glucose is typically measured at baseline and 2 hours; additional time points (e.g., 30 minutes, 60 minutes, and 90 minutes) may be collected in research settings to characterize glucose dynamics. OGTT is considered the reference standard for diagnosing IGT, one category of prediabetes[1]. IGT is defined as a 2-hour plasma glucose of 7.8-11.0 mmol/L during a 75-g OGTT. OGTT is widely used in both clinical practice and epidemiological studies. It primarily evaluates glucose tolerance, and when glucose is measured together with insulin or C-peptide, OGTT-derived indices can also provide information on β-cell function and insulin sensitivity. OGTT is commonly used in high-risk populations, including pregnant women (for gestational diabetes screening/diagnosis) and individuals with a family history of diabetes. However, OGTT is time-consuming and may have limited acceptability because it requires fasting and repeated blood sampling. Results may also be unreliable during acute illness or in some pancreatic disorders (e.g., acute pancreatitis). In addition, OGTT results can be influenced by recent diet and physical activity, intercurrent illness or stress, and medications (e.g., glucocorticoids).

FPG

FPG is measured after at least 8 hours of fasting, with water permitted before blood collection. According to ADA criteria, FPG levels of 5.6-6.9 mmol/L are diagnostic of IFG[1]. FPG is simple, inexpensive, widely available, and generally more acceptable to patients than the OGTT. These advantages make it useful as an initial screening test in routine clinical practice.

However, FPG reflects glucose concentration at a single fasting time point and cannot capture post-load glucose abnormalities. As a result, reliance on FPG alone may miss individuals with isolated IGT or postprandial hyperglycemia. For individuals with borderline findings or high metabolic risk, and in settings where post-load glucose assessment is clinically important, OGTT can provide complementary diagnostic information and improve the classification of prediabetes.

HbA1c

HbA1c was introduced by Rahbar[22] in the late 1960s, but it was not widely used for clinical testing until 1976. HbA1c levels between 5.7% and 6.4% are diagnostic of prediabetes according to ADA. However, the World Health Organization (WHO) has not yet included HbA1c in the diagnostic criteria for prediabetes[23]. In 2009, the International Expert Committee recommended the use of HbA1c as the preferred screening method[24]. HbA1c reflects the average blood glucose level over the past 2-3 months[23], and its diagnostic rate is higher compared to FPG. HbA1c detection can help prevent complications, such as retinopathy[25]. Additionally, HbA1c testing does not require repeated measurements, leading to better patient compliance compared to OGTT.

However, HbA1c levels can be influenced by factors such as abnormal red blood cell turnover and certain medications, including antiretroviral drugs[26]. Given that HbA1c can be influenced by individual factors, it is suitable as a screening tool, but not as the sole diagnostic measure. For patients with abnormal red blood cell turnover (e.g., blood loss or anemia), the diagnosis should rely on OGTT combined with FPG.

CGM

In recent years, CGM has provided a dynamic approach to glucose assessment through a subcutaneous sensor. Unlike HbA1c, FPG, or OGTT, CGM can capture glucose variability, nocturnal glucose changes, and postprandial glucose excursions, thereby offering information that is not reflected by conventional static tests[27]. These features may be useful for research and for selected high-risk individuals with suspected early dysglycemia. However, CGM has not been established as a routine diagnostic tool for prediabetes. Its wider application is limited by cost, access, sensor-related discomfort or skin reactions, possible differences between interstitial and blood glucose, and the lack of standardized diagnostic thresholds for prediabetes[28]. These barriers may be more relevant in resource-limited settings, where device availability, reimbursement, and access to clinicians experienced in CGM interpretation may be restricted. Therefore, CGM should currently be regarded as a supplementary tool rather than a replacement for FPG, OGTT, or HbA1c in prediabetes screening.

In summary, FPG, OGTT, and HbA1c reflect different aspects of glucose metabolism and therefore do not identify identical populations. FPG mainly reflects fasting glucose regulation, whereas OGTT detects abnormalities after a glucose load and can identify isolated IGT. HbA1c reflects longer-term glycemic exposure and may also be affected by factors unrelated to glucose metabolism, such as altered red blood cell turnover and hemoglobin variants. These differences help explain the limited overlap among individuals identified by the three diagnostic approaches. Combining FPG with OGTT can therefore improve the identification and classification of prediabetes in selected individuals. CGM provides additional information on glucose variability, but its routine use for prediabetes screening remains limited. Selected guideline recommendations for prediabetes screening and pharmacological prevention are compared in Table 2.

Table 2 Comparison of selected guideline recommendations for prediabetes screening and pharmacological prevention.

Guideline	Screening recommendation	Prediabetes/high-risk criteria	Pharmacological prevention
ADA 2026	Screen from age 35; earlier with overweight/obesity plus risk factors	FPG 5.6-6.9 mmol/L; 2-hour OGTT 7.8-11.0 mmol/L; HbA1c 5.7%-6.4%	Consider metformin in selected high-risk adults (age 25-59, BMI ≥ 35 kg/m2, higher glycemia, or previous GDM)
USPSTF 2021[118]	Adults aged 35-70 years with overweight/obesity	Prediabetes by FPG, HbA1c, or OGTT	Offer/refer to effective preventive interventions; metformin may benefit selected high-risk groups
NICE PH38[119]	Risk assessment, then FPG or HbA1c in people at increased risk	FPG 5.5-6.9 mmol/L or HbA1c 6.0%-6.4%	Consider metformin if lifestyle intervention is ineffective or not feasible, especially BMI > 35 kg/m2

ADA: American Diabetes Association; USPSTF: United States Preventive Services Task Force; NICE: National Institute for Health and Care Excellence; FPG: Fasting plasma glucose; OGTT: Oral glucose tolerance test; HbA1c: Glycated hemoglobin; BMI: Body mass index; GDM: Gestational diabetes mellitus.

INTERVENTIONS FOR PREDIABETES

Prediabetes should not be viewed as a uniform risk state. The main goal of management is to prevent or delay progression to T2DM. After diagnosis, the glycemic phenotype and the degree of dysglycemia can help identify individuals at greater risk of progression. Risk is higher in people with combined IFG and IGT or with glycemic values near the diabetes threshold. Obesity and a history of gestational diabetes may further increase risk. Consistent with current guideline recommendations, lifestyle intervention remains the foundation of management, while metformin may be considered in selected high-risk adults[29]. Based on the evidence reviewed here, we interpret combined IFG and IGT, glycemic values near the diabetes threshold, obesity, and previous gestational diabetes as features that may justify closer follow-up and greater preventive intensity. In individuals with obesity, additional weight-management treatment should follow established obesity-specific indications, and MBS should be considered only in those who meet surgical criteria[30]. This risk-stratified sequence represents the authors’ interpretation of the reviewed evidence rather than a formal guideline algorithm. An integrated clinical framework for diagnosis, risk stratification, follow-up, lifestyle intervention, and escalation to pharmacological or obesity-directed treatment is summarized in Figure 1. The following sections discuss exercise, dietary modification, pharmacological treatment, and metabolic surgery.

Figure 1 Clinical algorithm for the diagnosis, risk stratification, and management of prediabetes. FPG: Fasting plasma glucose; OGTT: Oral glucose tolerance test; HbA1c: Glycated hemoglobin; IFG: Impaired fasting glucose; IGT: Impaired glucose tolerance.

Exercise interventions for prediabetes

Health behaviors, particularly physical activity, play an important role in the development of T2DM[31]. In the Finnish Diabetes Prevention Study, 522 overweight adults with IGT were followed for a mean of 3.2 years. The cumulative incidence of diabetes at 4 years was 11% in the lifestyle-intervention group and 23% in the control group, corresponding to an absolute risk reduction of 12 percentage points[32]. In Indian Diabetes Prevention Programme-1, 531 Asian Indian adults with IGT were followed for a median of 30 months. The 3-year cumulative incidence of diabetes was 39.3% with lifestyle modification and 55.0% with standard care, corresponding to an absolute risk reduction of 15.7 percentage points[33]. Exercise may improve insulin sensitivity, glycemic control, and lipid metabolism. Changes in inflammatory pathways and the gut microbiota have also been reported, although their contribution to the glycemic benefits of exercise remains uncertain[34]. Exercise interventions are widely used to improve abnormal glycemic status in individuals with prediabetes. During the prediabetes stage, regular exercise may lower metabolic risk and may prevent or delay progression to diabetes. In addition, improved physical activity may mitigate cardiometabolic risk factors that contribute to complications once diabetes develops.

Current guidelines support regular physical activity as a core strategy for diabetes prevention. The WHO recommends 150-300 minutes/week of moderate-intensity aerobic activity or 75-150 minutes/week of vigorous-intensity activity, together with muscle-strengthening activities at least twice weekly[35]. The ADA also recommends intensive lifestyle intervention for adults at high risk of T2DM, with a physical activity goal of at least 150 minutes/week of moderate-intensity activity[29]. Evidence from randomized trials further suggests that different exercise modalities may improve glycemic control in people with prediabetes[36]. Therefore, exercise prescriptions should be individualized according to fitness level, metabolic status, comorbidities, and adherence. The following sections summarize aerobic exercise, resistance training, and flexibility and balance training.

Aerobic exercise: Aerobic exercise is one of the most commonly used forms of lifestyle intervention for people with prediabetes. It relies on repeated, rhythmic movement of large muscle groups, such as brisk walking, cycling, swimming, or jogging[37-39]. These activities are relatively easy to prescribe and can be adjusted according to age, baseline fitness, joint condition, and personal preference. For individuals who have difficulty with weight-bearing exercise, cycling or swimming may be more acceptable options.

Clinical studies suggest that aerobic training can improve glucose regulation in people with prediabetes. Reported benefits include better insulin sensitivity, improved fasting glucose or OGTT-related measures, and favorable changes in beta-cell related indices[40-43]. Some studies also suggest that regular aerobic activity may help reduce the risk of progression from prediabetes to T2DM, although the magnitude of benefit varies across populations and training protocols[40-42]. In addition to glycemic outcomes, improvements in exercise capacity, physical function, and confidence in maintaining activity may support long-term lifestyle change[44,45].

Moderate-intensity aerobic activity is generally appropriate for most individuals with prediabetes. A practical target is at least 150 minutes/week of moderate-intensity activity, such as brisk walking, with exercise distributed across the week[29]. The duration and intensity can be increased gradually according to baseline fitness and exercise tolerance. For individuals with limited mobility or joint problems, lower-impact activities such as cycling or swimming may be more suitable. Exercise frequency, duration, intensity, comorbidities, and individual preferences should all be considered when developing a sustainable program.

Resistance training: Resistance training can improve muscle strength, body composition, insulin sensitivity, and glycemic control in individuals with prediabetes[34,39,40,46]. Common approaches include free weights, resistance machines, elastic bands, and bodyweight exercises. Muscle-strengthening exercise should generally be performed at least twice weekly and should involve the major muscle groups[35,37]. Training intensity and progression should be adjusted according to baseline fitness, age, comorbidities, and previous exercise experience. Resistance training can be used alone or combined with aerobic exercise, depending on individual needs and preferences.

However, adherence to resistance training often declines over time in real-world settings. Adherence may be limited by a lack of professional guidance, difficulty integrating training into daily routines, and low confidence or motivation[45]. Strategies such as exercising with a partner, using structured programs, and improving access to suitable training environments may help maintain participation[45]. Patients should also be informed that benefits often require sustained engagement rather than short-term effort[47].

Flexibility exercises and balance training: Flexibility exercises can improve joint range of motion and mobility, while balance training can improve gait and postural stability and may reduce the risk of falls[37]. These activities are particularly useful for older adults or individuals with limited mobility, but they should be considered complementary rather than primary strategies for glycemic control. Flexibility and balance activities can be performed at least twice weekly and may be incorporated into a broader exercise program according to individual functional status and preferences[48]. Common options include stretching, yoga, Tai Chi, and similar low-impact activities.

Dietary interventions for prediabetes

Dietary modification is one of the most widely recommended non-pharmacological strategies for prediabetes. Current evidence does not support a single dietary pattern as optimal for all individuals. Different diets may improve glucose metabolism through weight reduction, reduced postprandial glucose excursions, improved insulin sensitivity, and favorable changes in cardiometabolic risk factors. Therefore, dietary intervention should be assessed not only by short-term metabolic effects, but also by safety, adherence, individual preference, and long-term sustainability. The main metabolic effects and limitations of commonly studied dietary patterns are shown in Table 3.

Table 3 Effects of dietary interventions relevant to prediabetes on body weight, glycemic control, blood lipids, and sustainability.

Diet	Key features	Body weight	Blood lipids			Blood glucose	HbA1c	Sustainability
			HDL-C	LDL-C	Triglycerides
Ketogenic diet	High-fat, very low-carbohydrate pattern; macronutrient composition varies, and classical regimens are often approximately 70% fat, 10% carbohydrate, and 20% protein[120]	Usually decreases in overweight/obese adults and in T2DM studies; evidence specifically in prediabetes remains limited[49,50,53,54]	Variable; may increase in some studies[53,54]	Variable; LDL-C may increase in some individuals and should be monitored[49,120]	Usually decreases[49,53,54]	May improve fasting or overall glycemia, mainly supported by T2DM and mixed dysglycemia studies[49,53,54]	May decrease in the short term; in Keto-Med, HbA1c reduction was not clearly superior to Mediterranean diet[49]	Restrictive pattern; long-term adherence may be challenging and may be lower than with the Mediterranean diet[49,51]
Mediterranean diet	Plant-based dietary pattern emphasizing vegetables, fruits, whole grains, legumes, nuts, olive oil as the main fat source, moderate fish intake, and limited saturated fat[57,121]	May support weight loss or weight maintenance when combined with energy control and lifestyle management[71,121,122]	May improve or maintain HDL-C[122]	Generally neutral or favorable, depending on baseline diet and fat quality[121-123]	May decrease[121,122]	Associated with improved glycemic control and lower diabetes risk in clinical and epidemiological studies[59,60,71,121-124]	May decrease; benefits are reported in T2DM and prediabetes-related studies[49,121,122]	Generally more compatible with long-term dietary practice than highly restrictive regimens[49,71]
PPT diet	Personalized postprandial glucose response-targeting diet based on individual-level data and machine-learning prediction[70]	Not primarily a weight-loss diet; modest decrease reported in some intervention studies without prescribed calorie restriction[72]	May improve, but evidence is limited and not consistently superior to Mediterranean diet[71,72]	No consistent advantage over Mediterranean diet has been established[71,72]	May decrease, but evidence remains limited[71,72]	Targets postprandial excursions; one prediabetes trial showed better reduction in postprandial hyperglycemia than Mediterranean diet[71]	Greater HbA1c reduction than Mediterranean diet was reported in one prediabetes trial[71]	Personalization may support adherence, but implementation depends on data collection, algorithm access, and patient engagement[70,71]

HDL-C: High-density lipoprotein cholesterol; LDL-C: Low-density lipoprotein cholesterol; HbA1c: Glycated hemoglobin; PPT: Personalized postprandial glucose response-targeting diet; T2DM: Type 2 diabetes mellitus.

Ketogenic diet: The ketogenic diet is a very-low-carbohydrate, high-fat dietary pattern that promotes ketone body production. It has been studied as a strategy for weight reduction and glycemic improvement in people with obesity, T2DM, and prediabetes, although evidence specifically in prediabetes remains limited[49,50]. Studies in T2DM suggest that ketogenic diets may reduce body weight, triglycerides, and HbA1c, although lipid responses vary among individuals[51-54]. In the Keto-Med randomized crossover trial, the ketogenic diet did not show a clear advantage over the Mediterranean diet in HbA1c reduction and was more restrictive[49]. Therefore, its potential metabolic benefits should be balanced against lipid changes, nutritional adequacy, and long-term adherence.

Mediterranean diet: In the Seven Countries Study, Mediterranean countries were observed to have lower cardiovascular mortality and longer life expectancy[55,56]. Keys later described this regional dietary pattern as the Mediterranean diet[57]. The Mediterranean diet is characterized by a high intake of plant-based foods and a fat profile dominated by olive oil and other unsaturated fatty acids. It typically emphasizes vegetables, legumes, fruits, whole grains, nuts, and moderate fish intake, while wine is consumed in moderation in some traditional settings[58].

Growing evidence suggests that the Mediterranean diet is beneficial for metabolic health. The Mediterranean diet has been associated with a lower risk of T2DM, and improvements in glycemic and cardiometabolic measures have also been reported[59]. The magnitude of these benefits appears to be related to dietary adherence[60]. However, much of this evidence comes from diabetes-prevention cohorts or patients with established T2DM rather than exclusively from prediabetes-specific populations. Chronic inflammation is considered an important component of obesity, and several mechanisms have been proposed to explain the association between the Mediterranean diet and lower diabetes risk, although their relative contributions remain uncertain[61]. A 22-year follow-up study reported that higher olive oil intake was modestly associated with a lower risk of T2DM in women[62]. This finding supports a possible role of dietary fat quality in the metabolic benefits of the Mediterranean diet, although these benefits should be interpreted in the context of the overall dietary pattern.

Many Mediterranean diet foods such as fruits, vegetables, whole grains, legumes, tea, cocoa, and wine contain polyphenols[63]. Polyphenols have antioxidant properties and have been proposed to influence pathways related to oxidative stress and insulin resistance[64,65]. For example, green tea has been reported to reduce endothelial inflammation in experimental settings, but clinical findings have been inconsistent across studies[66,67]. Similarly, resveratrol related findings have been mixed, and the evidence for several other foods and supplements remains insufficient in clinical studies[67,68]. One study compared a ketogenic diet and a Mediterranean diet using a randomized crossover design with two 12-week phases. The ketogenic diet led to a greater reduction in triglycerides but increased low-density lipoprotein cholesterol, and it was also associated with lower fiber intake and reduced intake of some essential nutrients. In contrast, the Mediterranean diet increased fiber intake and showed a larger reduction in average glucose, while HbA1c improved similarly in both diets. The authors also reported higher dietary sustainability with the Mediterranean diet[49].

Personalized postprandial-targeting diet: The personalized postprandial-targeting (PPT) diet is a personalized dietary approach developed using machine learning algorithms based on individual-level data. Because individuals respond heterogeneously to common diets, personalized dietary strategies have been proposed to improve postprandial glycemic responses. Similar trends toward personalization have also been observed in pharmacological strategies[69].

The PPT diet concept was proposed by Zeevi et al[70]. In 2015, they developed a machine-learning model integrating blood parameters, dietary habits, and gut microbiota to predict individualized postprandial glucose responses to real-life meals. This proof-of-concept study showed substantial inter-individual variation in postprandial glucose responses and suggested that personalized dietary recommendations may help reduce postprandial glycemic excursions[70]. However, its generalizability requires further validation.

Because there are relatively few studies on the PPT diet, current evidence for diabetes prevention and management is based on a limited number of reports. In a randomized trial in individuals with prediabetes, the PPT diet improved several measures of glycemic control compared with the Mediterranean diet, including postprandial glucose responses[71]. A separate pilot trial in patients with newly diagnosed T2DM also reported improvements in glycemic and metabolic outcomes with a personalized diet[72]. However, the number of clinical studies remains limited, and larger independent trials are needed before this approach can be recommended routinely for prediabetes.

Authors have noted that larger studies are needed to validate clinical efficacy, compare the PPT diet with other dietary patterns, and assess its impact on microbiome composition[72]. In 2023, investigators further evaluated the effects of the PPT diet vs the Mediterranean diet on the oral and gut microbiome, metabolites, and cytokines in 200 individuals with prediabetes, aiming to clarify microbiome-related mechanisms[73]. They reported changes in the microbiome, cardiometabolic features, and immune-related measures following the dietary interventions[73]. The PPT diet was associated with increased gut microbiome diversity and richness, and baseline microbiome diversity was negatively correlated with HbA1c. After 6 months, changes in microbial taxa were associated with changes in clinical outcomes, and mediation analyses suggested that microbiome shifts may partially contribute to the observed clinical effects[73]. These findings suggest a possible microbiome-related mechanism, but they do not establish that microbiome changes mediate the clinical benefits of the PPT diet. Routine implementation may also be limited by the need for individual-level data collection, digital or algorithmic infrastructure, professional dietary support, and associated costs, which may reduce accessibility in resource-limited settings.

Pharmacological interventions

Pharmacological interventions for prediabetes are mainly aimed at delaying or preventing progression to T2DM. Current evidence supports selective drug use in individuals with higher metabolic risk, rather than routine treatment for all people with prediabetes[6,29]. The populations studied differ across drug classes. Metformin, alpha-glucosidase inhibitors, and thiazolidinediones have been evaluated directly in individuals with prediabetes, IFG, or IGT, whereas evidence for incretin-based therapies is derived mainly from people with overweight or obesity and prediabetes. The main agents studied include metformin, alpha-glucosidase inhibitors, thiazolidinediones, glucagon-like peptide-1 (GLP-1) receptor agonists, and glucose-dependent insulinotropic polypeptide [glucose-dependent insulinotropic polypeptide (GIP)]/GLP-1 dual receptor agonists, with their evidence base and main clinical considerations summarized in Table 4.

Table 4 Pharmacological interventions for prediabetes.

Drug category	Representative agents	Action profile	Applicable population	Potential limitations	Clinical implication
Biguanides	Metformin	Reduces hepatic glucose production and improves insulin sensitivity; possible intestinal effects[75]	High-risk prediabetes, especially younger adults, higher BMI, elevated fasting glucose or HbA1c, or previous gestational diabetes[74]	Less effective than intensive lifestyle intervention; preventive effect weakens over time[6,76]	Best-established option; suitable as an adjunct for selected high-risk individuals
α-glucosidase inhibitors	Acarbose, voglibose	Delay intestinal carbohydrate digestion and absorption; reduce postprandial glucose excursions[80]	Mainly IGT or postprandial dysglycemia[77,78]	Gastrointestinal adverse effects, meal-time dosing, adherence burden, and lack of clear cardiovascular benefit[79,80]	May delay diabetes onset in IGT-dominant prediabetes, but not generalizable to all forms of prediabetes
Thiazolidinediones	Rosiglitazone, pioglitazone	Improve insulin sensitivity through PPARγ activation	Selected high-risk non-diabetic populations with IFG or IGT[81,83,84,125]	Weight gain, edema, heart failure risk, fracture risk, hepatotoxicity or cardiovascular safety concerns[81,83,84,125]	Strong proof of concept for targeting insulin resistance, but safety concerns limit routine use
GLP-1 receptor agonists	Liraglutide, semaglutide	Promote weight loss, reduce appetite, delay gastric emptying, and regulate insulin and glucagon secretion[89]	Prediabetes with overweight or obesity[85-87]	Gastrointestinal adverse effects, injection burden, high cost, and partial loss of benefit after withdrawal[85,87,90]	Most relevant for obesity-related prediabetes; not routine therapy for all patients
GIP/GLP-1 dual receptor agonists	Tirzepatide	Activates GIP and GLP-1 receptors; associated with weight loss, reduced appetite, improved insulin secretion and insulin sensitivity[91]	Obesity with prediabetes[92]	Long-term injections, weight regain after withdrawal, gastrointestinal adverse effects, cost and access issues[92,93]	Promising for obesity-related prediabetes, but evidence should not be generalized to all prediabetic individuals
Other agents	Orlistat, SGLT2 inhibitors and selected non-core agents	Weight-loss related, cardiorenal-metabolic, insulin secretory, vascular-metabolic, or supplementation-based approaches[94-98]	Selected high-risk or disease-specific populations	Evidence is narrower or less consistent; some agents show limited or negative effects	Secondary or context-dependent options rather than core pharmacological strategies

DPP: Diabetes prevention program; IFG: Impaired fasting glucose; IGT: Impaired glucose tolerance; GLP-1: Glucagon-like peptide-1; GIP: Glucose-dependent insulinotropic polypeptide; PPARγ: Peroxisome proliferator-activated receptor gamma; SGLT2: Sodium-glucose cotransporter 2.

Metformin: Metformin has the strongest clinical evidence among pharmacological interventions for prediabetes. In the Diabetes Prevention Program, 3234 overweight or obese adults with elevated fasting and 2-hour post-load glucose were followed for an average of 2.8 years. The estimated 3-year cumulative incidence of diabetes was 28.9% with placebo, 21.7% with metformin, and 14.4% with intensive lifestyle intervention, corresponding to absolute risk reductions of 7.2% and 14.5% points, respectively[6]. Further analyses showed that the benefit was more evident in younger adults, individuals with higher body mass index (BMI), elevated fasting glucose or HbA1c, and women with previous gestational diabetes[74]. This may be related to its main actions in reducing hepatic glucose production and improving insulin sensitivity, together with possible intestinal effects[75]. Long-term follow-up suggested that the preventive effect could persist, although it became weaker over time[76]. Therefore, metformin is more suitable as an adjunctive option for high-risk individuals with prediabetes rather than routine treatment for all patients.

α-glucosidase inhibitors: Alpha-glucosidase inhibitors have been studied mainly in people with IGT, rather than in all forms of prediabetes. This distinction is important because these agents act primarily on postprandial glucose by delaying intestinal carbohydrate digestion and absorption. In the STOP-NIDDM trial, acarbose reduced the progression from IGT to T2DM[77]. Voglibose showed a similar preventive effect in Japanese individuals with IGT when added to lifestyle modification[78]. In the ACE trial, acarbose reduced incident diabetes in Chinese patients with coronary heart disease and IGT, but did not reduce major cardiovascular events[79]. These findings suggest that alpha-glucosidase inhibitors may be useful for delaying diabetes onset in prediabetes characterized by postprandial dysglycemia. Their use is limited by gastrointestinal adverse effects, meal-time dosing, adherence burden, and the lack of clear cardiovascular benefit[80].

Thiazolidinediones: Thiazolidinediones have been tested for diabetes prevention in high-risk non-diabetic populations, mainly because they improve insulin sensitivity through peroxisome proliferator-activated receptor gamma activation. Early evidence came from troglitazone studies in women with previous gestational diabetes and from the troglitazone arm of the Diabetes Prevention Program, but this drug was later withdrawn because of hepatotoxicity[81,82]. In the DREAM trial, adults with IFG and/or IGT and no previous cardiovascular disease were followed for a median of 3 years. Diabetes developed in 10.6% of participants receiving rosiglitazone and 25.0% receiving placebo, corresponding to an absolute risk reduction of 14.4% points[83]. In ACT NOW, 602 overweight adults with IGT and additional risk factors for diabetes were followed for a median of 2.4 years. Diabetes developed in 5.0% of the pioglitazone group and 16.7% of the placebo group, corresponding to an absolute risk reduction of 11.7 percentage points[84]. These findings show that targeting insulin resistance can markedly delay diabetes onset in selected prediabetic populations. However, this class is not commonly used for routine diabetes prevention because of safety concerns, including weight gain, edema, heart failure risk, bone fracture risk, and drug-specific concerns such as troglitazone-related liver toxicity and rosiglitazone-related cardiovascular controversy[83,84].

GLP-1 receptor agonists: GLP-1 receptor agonists are not routine pharmacological therapy for all individuals with prediabetes. Their evidence in prediabetes is mainly derived from people with overweight or obesity. In the SCALE Obesity and Prediabetes trial, adults with prediabetes and a BMI ≥ 30 kg/m², or ≥ 27 kg/m² with obesity-related comorbidities, received liraglutide 3.0 mg or placebo for up to 160 weeks. Diabetes was diagnosed in 2% and 6% of participants, respectively, corresponding to an absolute risk reduction of approximately 4 percentage points[85]. For semaglutide 2.4 mg, pooled analyses from the STEP program showed improved glycemic status and a higher rate of reversion to normoglycemia among participants with baseline prediabetes[86]. STEP 10 further confirmed that semaglutide 2.4 mg produced greater weight loss and more frequent return to normoglycemia than placebo in adults with obesity and prediabetes[87]. In SELECT, semaglutide reduced progression to biochemical diabetes in adults with overweight or obesity and cardiovascular disease but without diabetes[88]. These effects are mainly related to weight reduction, appetite suppression, delayed gastric emptying, and glucose-dependent regulation of insulin and glucagon secretion[89]. The main limitations include gastrointestinal adverse effects, injection burden, high cost, variable access and reimbursement, and partial loss of metabolic benefit after treatment withdrawal[90]. These factors may restrict long-term use and contribute to unequal access to treatment. Therefore, GLP-1 receptor agonists may be considered mainly in obesity-related prediabetes, rather than as general treatment for all prediabetic individuals.

GIP/GLP-1 dual receptor agonists: Tirzepatide is currently the main clinical representative of GIP/GLP-1 dual receptor agonists. By activating both glucose-dependent insulinotropic polypeptide and GLP-1 receptors, it can reduce appetite and body weight, while also improving insulin secretion and insulin sensitivity[91]. Evidence for its use in prediabetes mainly comes from people with obesity. In SURMOUNT-1, adults with obesity and prediabetes received tirzepatide for 176 weeks. Diabetes developed in 1.3% of participants in the tirzepatide group and 13.3% in the placebo group, corresponding to an absolute risk reduction of 12.0 percentage points over 176 weeks, together with sustained weight reduction[92]. This finding supports a preventive effect in obesity-related prediabetes, but it does not mean that tirzepatide should be used broadly in all individuals with prediabetes. Its clinical value should be interpreted in the context of excess adiposity, long-term injectable treatment, cost, and tolerability. In addition, SURMOUNT-4 showed weight regain after tirzepatide withdrawal, suggesting that the benefit may depend on continued treatment or effective long-term weight maintenance strategies[93]. Gastrointestinal adverse effects and limited real-world durability remain practical concerns.

Other agents have also been examined in people with prediabetes or in non-diabetic populations at high metabolic risk, but their evidence is less central to routine prediabetes management. In the XENDOS study, orlistat reduced diabetes incidence in individuals with obesity, largely through weight loss, although gastrointestinal adverse effects may limit adherence[94]. Sodium-glucose cotransporter 2 inhibitors have been linked to a lower risk of new-onset diabetes in adults with prediabetes and cardiorenal disease, but this evidence comes from specific high-risk populations rather than trials designed for general prediabetes[95]. Findings for several other interventions are more limited. Nateglinide did not reduce diabetes incidence in individuals with IGT, valsartan produced only a modest reduction without cardiovascular benefit, and vitamin D supplementation has shown mixed results across randomized trials and meta-analyses[96-98].

Taken together, pharmacological treatment for prediabetes should be selective. Metformin remains the best-established option, alpha-glucosidase inhibitors are more relevant to IGT, and incretin-based therapies are most applicable when prediabetes is accompanied by obesity. For most individuals, drug therapy should be considered an adjunct to lifestyle intervention rather than a replacement for dietary change, physical activity, and long-term risk management.

Surgical interventions

MBS is an established treatment for obesity and obesity-related metabolic complications[30]. In patients with obesity and prediabetes, MBS has been associated with remission of prediabetes and reduced long-term risk of T2DM[99-101]. The Swedish Obese Subjects study showed that bariatric surgery markedly reduced the long-term incidence of T2DM, especially in participants with IFG at baseline[101]. Borges-Canha et al[99] further reported that the remission rates of prediabetes after bariatric surgery were 82%, 73%, 66%, and 58% from the first to fourth postoperative year, respectively. Therefore, MBS may be considered an obesity-centered intervention for selected patients with prediabetes, but it is not a routine treatment for all individuals with prediabetes. Access may also be limited by the need for specialized surgical centers, multidisciplinary assessment, perioperative care, long-term nutritional follow-up, and substantial upfront costs, with these barriers being particularly relevant in resource-limited settings. Common procedures include SG, Roux-en-Y gastric bypass (RYGB), laparoscopic adjustable gastric banding (LAGB), and biliopancreatic diversion with duodenal switch (BPD/DS). The main characteristics, potential indications, procedural features, and common complications of these procedures are summarized in Table 5.

Table 5 Summary of metabolic and bariatric surgical procedures relevant to obesity-associated prediabetes.

Procedure	Mechanism/nature	Potentially suitable population	Procedure summary	Main limitations or complications
RYGB	Combined restrictive and malabsorptive procedure	Patients with obesity and marked metabolic abnormalities; evidence in prediabetes is mainly derived from obesity cohorts or indirect T2DM data[30,100,101]	Creation of a small gastric pouch with jejunal anastomosis, bypassing the distal stomach, duodenum and proximal jejunum[102]	Nutritional deficiencies, anastomotic leak, hemorrhage, dumping symptoms, and need forlong-term follow-up[102,104,126]
LAGB	Restrictive procedure	Patients requiring a less invasive, adjustable, and reversible procedure; now less commonly used because of weaker long-term weight-loss and metabolic effects[106,108]	Placement of an adjustable silicone band around the upper stomach to create a small gastric pouch[106]	Band slippage, port-related problems, erosion, insufficient weight loss, band removal, or conversion to other procedures[106,127]
SG/VSG	Mainly restrictive procedure with metabolic effects	Commonly performed procedure; may be suitable for patients with obesity and prediabetes, although direct prediabetes evidence remains limited[30,110,111]	Resection of most of the stomach along the greater curvature, leaving a tubular gastric sleeve[109]	GERD, staple-line leak, bleeding, and possible nutritional deficiencies[109,110,112]
BPD/DS	Combined restrictive and strongly malabsorptive procedure	Mainly considered for severe or super obesity, especially BMI ≥ 50 kg/m2, or as a revisional option after insufficient response to other procedures[113-115]	Sleeve gastrectomy with preservation of the pylorus and intestinal bypass[115]	Higher risk of nutritional deficiencies, protein-calorie malnutrition, fat-soluble vitamin deficiency, diarrhea, malodorous flatus, anastomotic leak, and hemorrhage[115,116]

The evidence for several procedures is derived mainly from obesity or type 2 diabetes mellitus cohorts and should not be directly generalized to all individuals with prediabetes. RYGB: Roux-en-Y gastric bypass; LAGB: Laparoscopic adjustable gastric banding; SG/VSG: Sleeve gastrectomy/vertical sleeve gastrectomy; BPD/DS: Biliopancreatic diversion with duodenal switch; GERD: Gastroesophageal reflux disease; T2DM: Type 2 diabetes mellitus.

RYGB surgery: RYGB is a mixed restrictive and malabsorptive procedure. It creates a small gastric pouch and connects it to the jejunum, bypassing the distal stomach, duodenum, and proximal jejunum[102]. This reconstruction reduces food intake and part of nutrient absorption, and also changes gut hormone responses involved in glucose metabolism. RYGB is a commonly used metabolic and bariatric procedure for obesity. In patients with established T2DM, Girundi[103] reported that 410 of 468 patients achieved diabetes remission 18 months after RYGB. This result was obtained from patients with established T2DM, so it should be regarded as indirect evidence when discussing prediabetes. More directly, a 15-year single-institution cohort study found that bariatric surgery was associated with a lower risk of progression from prediabetes to T2DM, with a greater preventive effect after RYGB than after SG in that cohort[100]. However, the anatomical changes caused by RYGB also bring long-term problems, including nutritional deficiency and gastrointestinal symptoms. In a Danish survey of 1429 patients at a median of 4.7 years after RYGB, 52.4% reported symptoms related to dumping syndrome[104]. Therefore, although RYGB is effective in improving glucose metabolism, its use in prediabetes should be limited to patients with obesity who meet surgical indications, and long-term nutritional monitoring is necessary.

LAGB: LAGB is a restrictive bariatric procedure in which an adjustable silicone band is placed around the upper part of the stomach to create a small gastric pouch. The degree of restriction can be modified by injecting or removing fluid through a subcutaneous port, thereby slowing gastric emptying, increasing satiety, and reducing food intake[105,106]. Compared with other bariatric procedures, LAGB has the advantages of no gastrointestinal transection, adjustability, reversibility, and a relatively low rate of early postoperative complications. However, compared with RYGB or SG, LAGB generally shows less durable or less predictable effects on long-term weight loss and metabolic improvement[106-108]. In a multicenter retrospective cohort study from South Korea including 261 patients who underwent LAGB, RYGB, or SG, diabetes remission occurred in 40.0% of patients after LAGB, compared with 65.9% after RYGB and 30.2% after SG; dyslipidemia remission was observed in 66.7%, 100%, and 73.7% of patients, respectively[107]. Because these outcomes were reported in patients with established diabetes, they should be regarded as indirect evidence when considering LAGB for prediabetes. These data suggest that LAGB may improve metabolic abnormalities, but its glycemic effect appears less consistent than that of RYGB. In addition, long-term complications such as band slippage, erosion, esophageal dilation, band removal, and reoperation remain important limitations[108]. For these reasons, the use of LAGB has declined over the past decade, and it is now less commonly selected than SG or RYGB.

SG: SG, also known as vertical SG, removes most of the stomach along the greater curvature and leaves a tubular gastric sleeve. It mainly reduces gastric capacity and food intake, but its effect on glucose metabolism may also involve changes in gut hormone responses and other weight loss-dependent or independent mechanisms[109]. SG is generally considered a restrictive procedure, but postoperative nutritional deficiencies may still occur, especially when food intake is limited or nutritional supplementation is insufficient[110]. Evidence directly involving prediabetes is still limited. In adolescent patients with morbid obesity, Elhag et al[110] reported 100% remission of prediabetes cases two years after LSG, but this result came from a small subgroup and should be interpreted cautiously. Another retrospective study of 202 overweight and obese patients undergoing SG between 2016 and 2018 found that 32 patients (16%) had prediabetes before surgery, and glucose-related indices improved significantly during the first postoperative year[111]. SG may also affect gastroesophageal reflux, which should be considered when choosing the procedure[112].

BPD/DS: BPD/DS is a combined restrictive and malabsorptive bariatric procedure that includes SG, preservation of the pylorus, and intestinal bypass. Because of its strong weight-loss and metabolic effects, it is mainly considered for patients with severe or super obesity, and may also be used as a revisional option after insufficient response to other bariatric procedures[113-115]. In a retrospective study of 211 patients with super obesity, Skogar and Sundbom[116] found that BPD/DS produced greater weight reduction and better metabolic control than RYGB, including a stronger effect on diabetes. However, this evidence was not derived from a prediabetes-specific population. BPD/DS also carries a higher risk of nutritional deficiencies and gastrointestinal adverse effects, so long-term nutritional monitoring is required[115,116].

CLINICAL IMPLICATIONS, LIMITATIONS, AND FUTURE DIRECTIONS

The evidence supporting different approaches to prediabetes management varies in strength, duration, and clinical applicability. Lifestyle intervention has the most established evidence for preventing or delaying T2DM, with benefits reported during long-term follow-up, although maintaining dietary and physical activity changes remains difficult for many individuals[6,7]. Metformin also has substantial evidence for diabetes prevention, particularly in selected high-risk groups, but its effect is smaller than that of intensive lifestyle intervention and may become less pronounced over time[6,76]. Newer incretin-based therapies have shown encouraging results, especially in people with obesity, but cost, treatment burden, adverse effects, and loss of benefit after treatment withdrawal may affect their long-term feasibility[90,93]. Evidence for ketogenic and personalized dietary approaches is less established in prediabetes, with uncertainties regarding long-term adherence and generalizability[49,71]. MBS can provide substantial metabolic benefit in appropriately selected patients with obesity, but procedural risks and the need for long-term follow-up limit its broader use[30,101].

Equity and implementation should also be considered when translating these interventions into routine practice. CGM requires device availability, repeated sensor use, and appropriate interpretation, while personalized dietary approaches depend on data collection, digital infrastructure, and professional support. Incretin-based therapies may be limited by high cost, reimbursement, treatment availability, and the need for long-term therapy. MBS requires specialized centers, multidisciplinary care, and prolonged postoperative follow-up. These barriers may disproportionately affect people in resource-limited settings and should be considered alongside clinical efficacy when selecting preventive strategies.

Diagnostic heterogeneity also has practical implications for screening and follow-up. FPG, OGTT, and HbA1c do not identify identical populations, and individuals meeting only one criterion may differ in metabolic phenotype and future diabetes risk. Therefore, an isolated abnormal result should be interpreted in the context of the overall risk profile rather than as a uniform risk category. Screening may help identify individuals at increased risk, but its clinical benefit depends on appropriate confirmation, risk assessment, and access to effective preventive intervention; the population-level benefit of screening alone remains uncertain[1,21].

This review should also be considered in light of the characteristics of a narrative approach. Formal risk-of-bias assessment and evidence grading were not undertaken, as the aim was to provide a broad clinical synthesis rather than a systematic evaluation of intervention effects. We sought to strengthen the reliability of the review by prioritizing guidelines, systematic reviews, randomized trials, and large cohort studies where available. Nevertheless, differences across studies in population characteristics, diagnostic criteria, intervention duration, and outcome measures make direct comparison between management strategies difficult. These factors should be considered when interpreting the overall evidence summarized in this review. Future studies should clarify how different prediabetes phenotypes and diagnostic criteria relate to long-term outcomes and treatment response. More pragmatic studies are also needed to determine how risk-stratified screening, follow-up, and intervention strategies can be implemented effectively in routine clinical practice.

CONCLUSION

Prediabetes should be viewed not only as a diagnostic category, but also as a stage at which future diabetes risk can still be modified. Its clinical assessment is complicated by differences among FPG, OGTT, and HbA1c, since these measures do not identify exactly the same populations. This diagnostic heterogeneity makes risk stratification important when selecting prevention strategies.

For most individuals with prediabetes, dietary modification and regular physical activity remain the preferred first-line approach. Their benefit, however, depends on whether patients can maintain these changes with adequate follow-up and practical support. Pharmacological treatment is more appropriate for selected high-risk groups, such as individuals with obesity, marked dysglycemia, or other metabolic risk factors. MBS may further reduce diabetes risk in patients with obesity who already meet surgical indications, but it should not be regarded as a general treatment for prediabetes. Future prevention strategies need to move beyond simple detection and pay closer attention to risk classification, long-term adherence, and individualized intervention.

ACKNOWLEDGEMENTS

We are grateful to our colleagues at Jiangsu Medical College for their support during the preparation of this manuscript.