Next Article in Journal
Nurse Coaching in Oncology Care to Reduce Stress: A Systematic Review and Meta-Analysis
Previous Article in Journal
Sleep Quality, Glycemic Control, and Daytime Dysfunction in Type 2 Diabetes
Previous Article in Special Issue
Mindful Eating and Healthy Lifestyle Behaviors Among Women with and Without Regular Exercise Habits
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Healthcare
Volume 14
Issue 7
10.3390/healthcare14070839
healthcare-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Review

Screening Tools for Early Identification of Adults at High Risk of Type 2 Diabetes: A Scoping Review

by
Christos Christakis
1,
Dimitra Saliari
1,
Antonis Zampelas
2 and
Odysseas Androutsos
1,*
1
Laboratory of Clinical Nutrition and Dietetics, Department of Nutrition and Dietetics, University of Thessaly, Argonafton 1C, 42132 Trikala, Greece
2
Laboratory of Dietetics and Quality of Life, Department of Food Science and Human Nutrition, School of Food and Nutritional Sciences, Agricultural University of Athens, 75 Iera Odos Street, 11855 Athens, Greece
*
Author to whom correspondence should be addressed.
Healthcare 2026, 14(7), 839; https://doi.org/10.3390/healthcare14070839 (registering DOI)
Submission received: 13 February 2026 / Revised: 17 March 2026 / Accepted: 24 March 2026 / Published: 25 March 2026
(This article belongs to the Special Issue Psychophysiological Responses to Physical Activity, Exercise and/or Nutrition Interventions in Individuals with Cardiometabolic Health Impairments)
Browse Figures
Versions Notes

Highlights

What are the main findings?
  • Fifty-eight screening tools for early identification of adults at high risk for type 2 diabetes mellitus were identified across diverse populations and settings.
  • Most tools rely on simple demographic and anthropometric variables, showing moderate to good discriminatory ability with low implementation cost.
What are the implications of the main findings?
  • Easily applicable screening tools can support large-scale early detection and prevention strategies for type 2 diabetes mellitus, particularly in primary care.
  • Integration of validated risk scores into digital and electronic health platforms may enhance population-level screening and reduce healthcare burden.

Abstract

Background/Objectives: Global estimates suggest that approximately 43% of individuals living with diabetes remain undiagnosed, underscoring the need for early identification of adults at high risk of type 2 diabetes mellitus (T2DM) to support timely preventive interventions. This scoping review aimed to map and summarize existing non-invasive screening tools for identifying adults at high risk of T2DM. Methods: PubMed (MEDLINE), Web of Science, ScienceDirect, and Scopus were searched in accordance with the PRISMA extension for Scoping Reviews (PRISMA-ScR). Studies published between 1995 and 2026 that described screening tools for adult populations were included. Results: A total of 58 studies describing screening tools were identified. The tools were developed and applied across diverse populations and ethnic groups. Most were questionnaire-based, easy to administer, and low cost. Commonly included variables comprised demographic characteristics, anthropometric measures, lifestyle factors, and clinical indicators associated with increased T2DM risk. Substantial heterogeneity was observed in tool structure and reported predictive components. Conclusions: This scoping review provides an overview of available screening tools for the early identification of adults at high risk of T2DM. The mapped evidence may inform future validation studies and support context-specific implementation in public health and clinical practice settings, including integration into digital platforms.

1. Introduction

Diabetes mellitus (DM) is a serious chronic disease and one of the most important public health challenges worldwide. It is defined as a heterogeneous, multifactorial metabolic disorder characterized by chronic hyperglycemia resulting from defects in insulin secretion and/or action [1]. Type 2 Diabetes mellitus (T2DM), characterized by insulin resistance and progressive β-cell dysfunction [2], accounts for approximately 90–95% of DM cases and primarily affects adults [3]. Because DM may remain undiagnosed for prolonged periods, affected individuals are at increased risk of developing serious long-term complications, mainly due to vascular damage [1]. Early identification of individuals at increased risk is therefore a key component of primary prevention strategies. It is important to distinguish between diagnostic testing, which confirms the presence of diabetes, and screening or risk prediction tools, which aim to estimate the likelihood of developing the disease in asymptomatic populations. Such tools may help reduce the high proportion of undiagnosed cases and support timely lifestyle or clinical interventions.
DM is among the leading causes of mortality worldwide. According to World Health Organization data, diabetes remains among the leading causes of mortality worldwide [4]. The global prevalence of DM has increased steadily over recent decades, rising from 382 million cases in 2013 to 589 million in 2025 [5,6,7]. This upward trend is accompanied by a substantial economic burden on healthcare systems and national economies through both direct and indirect costs [8]. The total global cost of DM has risen dramatically over time, with projections indicating further increases in the coming decades [8]. In addition to system-level costs, individuals and families often experience significant financial hardship related to treatment expenses and productivity losses [9].
Despite growing awareness, a considerable proportion of cases remain undiagnosed. The proportion of diagnosed and treated individuals varies substantially across countries [10]. In high-income settings, up to 30–50% of cases may remain undiagnosed [11], while socioeconomically disadvantaged populations are disproportionately affected [12]. In 2025, approximately 43% of individuals living with DM were unaware of their condition, with particularly high rates of undiagnosed cases reported in low- and middle-income countries [7,13]. Early detection and appropriate management have been associated with improved outcomes in certain populations [14].
Given the rapid development of multiple screening tools over recent decades, along with differences in included variables, validation processes, and application contexts, the existing evidence is characterized by considerable heterogeneity. In such cases, a scoping review is considered an appropriate methodological approach, as it aims to map the extent, range, and characteristics of available evidence rather than to assess comparative effectiveness. Therefore, the purpose of the present study was to identify and summarize the available tools for detecting adults at high risk of developing T2DM.
Specifically, this scoping review sought to address the following research questions:
  • What screening tools have been developed or validated for the early identification of adults at high risk of type 2 diabetes?
  • In which populations and settings have these screening tools been applied?
  • What key characteristics and reported implementation features of these screening tools have been described in the literature?

2. Materials and Methods

This study is a scoping review conducted in accordance with the PRISMA Extension for Scoping Reviews (PRISMA-ScR) guidelines. A scoping review approach was chosen to map the available evidence on screening tools for early identification of adults at high risk of type 2 diabetes, given the heterogeneity of study designs, populations, and outcome measures. The PRISMA-ScR checklist is available in the Supplementary Materials (Table S1).

2.1. Study Selection Methodology

Studies included in this scoping review were identified through electronic searches of international scientific databases, specifically PubMed (MEDLINE), Web of Science, ScienceDirect and Scopus. The review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) framework, specifically the PRISMA extension for Scoping Reviews (PRISMA ScR) [15], to ensure transparency and methodological rigor.

2.2. Inclusion and Exclusion Criteria

Eligible studies were those reporting on adult populations—either healthy individuals or individuals undiagnosed with T2DM or prediabetes—that described the development or primary validation of screening tools for early risk identification.
The following studies were excluded: (a) studies evaluating or re-validating existing screening tools in populations different from the original derivation cohort; (b) studies involving non-adult populations; (c) studies focusing exclusively on individuals already diagnosed with T2DM; and (d) meta-analyses and systematic reviews. Secondary validation studies conducted in populations other than the original derivation sample were excluded to maintain the focus on identifying and mapping the initial development and core characteristics of screening tools. Inclusion of such studies would have shifted the review toward comparative cross-population performance and effectiveness analysis, which falls beyond the scope of a scoping review.

2.3. Search Strategy

The electronic search was performed without time-frame restrictions, covering the period from 1995 to 7 January 2026 to ensure maximum coverage of available data. In addition, a supplementary search was performed in Google Scholar by screening the first 200 results based on title relevance, and the bibliographic references of all eligible studies were screened to identify additional relevant sources, which were reviewed and evaluated using the same methodological approach.
The review questions and eligibility criteria were developed using the Population–Concept–Context (PCC) framework, which is recommended for scoping reviews. This is a well-established methodological approach for formulating research questions and conducting structured searches in the scientific literature (Table 1).
The search strategy was structured according to the Population–Concept–Context (PCC) framework. The Population component included terms referring to adult individuals, such as “adults,” “individuals,” and “undiagnosed”. The Concept component captured terms related to screening and risk assessment tools, including “test,” “risk score,” and “screen*.” The Context component encompassed terms associated with early detection and high-risk identification, such as “prediabetes,” “early diagnosis,” “high risk,” and “predict*.” The database-specific Boolean search strings are provided in Supplementary Table S2.
The comprehensive literature search across four electronic databases, namely Scopus (n = 987), Web of Science Core Collection (n = 278), ScienceDirect (n = 149), and PubMed/MEDLINE (n = 221), yielded a total of 1635 records for screening. In addition, a supplementary search was performed. No additional eligible records were identified through this supplementary search; therefore, the total number of records included in the PRISMA flow diagram remained unchanged. After screening, 58 articles that met the predefined inclusion criteria were finally included in the review (Figure 1).

2.4. Selection and Analysis Process

Study selection was performed independently by two reviewers to enhance transparency and minimize selection bias. Discrepancies were resolved through discussion with a third reviewer. Ιn detail, the studies identified through the search process were assessed for eligibility by two independent researchers (C.C. and D.S.). In the first phase the titles and abstracts were screened to exclude studies that did not meet the inclusion criteria. Studies deemed potentially eligible were then subjected to full-text evaluation. In cases where consensus on study inclusion could not be reached, a third reviewer (O.A.), who was aware of the initial assessments, reviewed the study and made the final decision.

2.5. Data Recording

Detailed information from each eligible study was extracted independently by the two researchers using a predefined data extraction table, which was initially piloted on a random sample of studies prior to full data collection. Extracted variables included: (a) key study characteristics (title, authors, year of publication); (b) population characteristics (sample size, geographical location, age range, and proportion of women); (c) components and variables included in each screening tool; (d) validation procedures and reported performance metrics (e.g., AUC, sensitivity, specificity, predictive values); and (e) additional comments or observations.
Data were recorded only when explicitly reported in the original publications, and no attempt was made to impute or estimate missing information. Consequently, variability in reporting completeness across studies is reflected in the descriptive synthesis.
In accordance with established scoping review methodology (PRISMA-ScR framework), a formal risk-of-bias or methodological quality appraisal of the included studies was not conducted. The objective of the present review was to map the breadth, characteristics, and reported features of existing screening tools rather than to critically evaluate internal validity or comparative effectiveness.

3. Results

Table 2 presents the 58 screening tools for adults at high risk of developing T2DM, identified based on the predefined study criteria, and recognized according to the screening variables used. For transparency and traceability, Table 2 presents the identified tools in chronological order. However, the instruments may also be conceptually grouped into three broad methodological categories: (i) anthropometric- and questionnaire-based tools, (ii) tools incorporating biochemical parameters, and (iii) algorithmic or machine learning–based models. The outcome targeted by each screening tool (e.g., type 2 diabetes, prediabetes, or impaired glucose regulation) is also presented. Table 3 summarizes the reported accuracy of each tool in the respective studies.
The screening variables utilized across all 58 studies for early identification of T2DM covered a broad range of demographic, anthropometric, lifestyle, and clinical factors. The most frequently used variables were age, sex, body mass index (BMI), obesity, family history of diabetes, hypertension or blood pressure, and waist circumference. Several models also incorporated lifestyle factors such as physical activity, smoking, alcohol consumption, and dietary habits, while others included education, socioeconomic status, and ethnicity. Clinical and biochemical markers, including fasting glucose, lipid levels, glycated hemoglobin (HbA1c), and history of gestational diabetes or hyperglycemia were also considered in some studies (Figure 2).
Performance metrics were reported heterogeneously across studies. Some articles presented discrimination measures such as AUC values, others reported sensitivity and specificity at predefined cut-off points, while a smaller number provided additional indicators including predictive values or calibration measures. The absence of standardized reporting across studies reflects variability in study design and validation objectives rather than selective extraction, and limits direct quantitative comparison between tools. In addition to variability in the type of performance metrics reported, heterogeneity was also observed in threshold selection. In several studies, more than one cut-off value was evaluated, resulting in different sensitivity and specificity profiles. However, detailed justification for the selection of optimal thresholds was not consistently provided.
Across the 58 eligible studies, the accuracy of screening tools for T2DM varied, with area under the curve (AUC) values ranging from 0.68 to 0.92. Using commonly applied thresholds (excellent: AUC > 0.80; good: 0.75–0.80; moderate: 0.70–0.75; limited: <0.70), the majority of tools fell within the good-to-moderate range, with sensitivity and specificity typically between 65–85%. Negative predictive values were consistently high (>90%), whereas positive predictive values remained relatively low. Reported cut-off values also varied across tools and populations.
The screening tools were applied to populations of diverse nationalities and geographical regions. The highest number of tools was reported in China (n = 12, including Taiwan) and the United States (n = 10). Overall, the tools were tested in 26 countries (Figure 3).

4. Discussion

This scoping review mapped the available evidence on screening tools for the early identification of adults at high risk of T2DM. The findings demonstrate substantial heterogeneity in the types of tools identified, the populations and settings in which they were applied, and the characteristics reported across studies. Most tools were simple questionnaires, presented either as decision trees or scoring systems in which the final score indicates the probability of incident T2DM. Many were described as user-friendly and low cost, features that may facilitate their implementation in primary healthcare settings. As reported in the included studies, their primary objective is early risk identification to support preventive strategies.
The identified tools incorporate multiple variables to enhance sensitivity and specificity according to the target population. These variables are generally related to established T2DM risk factors and can be grouped into four broad categories: (a) individual factors (age, sex, BMI, waist circumference, and physical activity); (b) genetic and family factors (family history of DM or hypertension); (c) clinical and metabolic factors (hypertension and gestational diabetes); and (d) socioeconomic and lifestyle factors (educational level, socioeconomic status, dietary habits, alcohol consumption, smoking status, and ethnicity). Age was the only variable included in all screening tools, followed by obesity indicators (BMI and waist circumference) and family history of T2DM, which appeared in 54 and 44 studies, respectively. Hypertension (either as a diagnosis or antihypertensive treatment) and family history of hypertension were also reported in 44 studies, sex in 33 studies, and physical activity in 21 studies. Less frequently included variables comprised ethnicity, biochemical markers, dietary habits, smoking status, and alcohol consumption.
From an implementation perspective, the identified tools differ considerably in feasibility requirements. Questionnaire-based tools relying solely on self-reported demographic and anthropometric variables can be administered rapidly, require no laboratory infrastructure, and are suitable for large-scale community or primary care screening. In contrast, tools incorporating biochemical markers such as fasting glucose or lipid parameters entail additional laboratory costs and logistical considerations, making them more appropriate for structured clinical environments. Although many tools are described as “low cost,” the resource implications vary: anthropometric-only tools may be considered minimal- or near-zero-cost approaches, whereas models requiring laboratory measurements impose greater financial and organizational demands. Therefore, selection of a screening instrument should be aligned with available healthcare infrastructure and screening objectives.
Findings from the included studies suggest that the selection of variables largely depends on the characteristics of the target population. In countries with pronounced economic inequality, such as China, India, and Indonesia, simple questionnaires that do not require biochemical testing are commonly used to minimize costs and facilitate large-scale implementation [23,24,26,28,39,40,44,48,49,52,58,59,62,65]. In settings marked by socioeconomic disparities, variables such as educational level and socioeconomic status are more frequently incorporated [44,54,55,59]. Overall, each tool appears to be designed to achieve acceptable predictive performance within its intended population.
Among the 58 screening tools identified, only 10 included ethnicity as a variable. This was particularly evident in countries characterized by high levels of multiculturalism, such as the United States, Canada, and the United Kingdom, where substantial proportions of the population consist of immigrants [25,31,35,36,41,43,45,63,64]. In the context of increasing global migration, consideration of cultural and demographic diversity may warrant further attention in future tool development.
Population-specific considerations merit further attention. Gestational diabetes history was incorporated in a small number of tools, although women with prior gestational diabetes represent a well-recognized high-risk group for future T2DM. This observation suggests that sex-specific or reproductive history–informed screening approaches remain relatively underrepresented within the mapped literature. Age-related variability also warrants consideration. Although age was included in all identified tools, relatively few studies explicitly examined whether discriminatory ability differed across age strata. Given that T2DM risk increases substantially with advancing age, calibration and threshold optimization for specific age groups may enhance screening precision. Finally, despite representation across 26 countries, certain populations remain underrepresented, including indigenous groups, low-income regions, and specific ethnic minorities. Targeted development and validation of screening tools for these populations may improve equity and contextual applicability.
Although most studies focused on non-invasive screening approaches, a subset of tools incorporated biochemical parameters, including fasting glucose, high-density lipoprotein (HDL), low-density lipoprotein (LDL), triglycerides (TG), and insulin. In certain cases, the addition of fasting glucose and lipid measurements improved the identification of individuals at very high risk of T2DM but did not substantially enhance detection among those at moderate or lower risk [36]. Moreover, in one study, insulin measurement demonstrated lower predictive value compared with glucose-based indicators [63], while its higher cost and technical complexity may limit its practical applicability in routine screening settings. Overall, these observations align with WHO recommendations emphasizing screening approaches that balance predictive performance with affordability and feasibility, particularly in resource-limited healthcare settings.
Where discrimination metrics were reported, performance varied across tools. Several instruments demonstrated AUC values exceeding 0.80, indicating strong discriminative ability, whereas others fell within moderate ranges (0.70–0.75), and a smaller proportion reported lower discrimination. Differences in study design, outcome definitions, and cut-off selection limit direct comparability across tools. In particular, heterogeneity in outcome targets warrants attention: while some instruments were designed to detect undiagnosed T2DM, others aimed to identify prediabetes or broader dysglycemia endpoints. These distinctions are clinically meaningful, as screening for prediabetes primarily supports primary prevention and lifestyle intervention strategies, whereas identification of undiagnosed T2DM may prompt confirmatory diagnostic testing and immediate clinical management.
Importantly, even tools with moderate AUC values may retain substantial utility in population-based screening programs, particularly when high negative predictive value is prioritized. The consistently high negative predictive values reported across several tools suggest suitability for initial population-level screening, where the primary objective is to exclude individuals at low risk and identify those requiring further assessment. Conversely, lower positive predictive values indicate that confirmatory diagnostic testing remains necessary before clinical decision-making, reinforcing the role of these instruments as preliminary risk stratification tools rather than diagnostic tests. Variation in cut-off selection further influences clinical and public health application: lower thresholds typically increase sensitivity at the expense of specificity, potentially leading to higher referral rates for confirmatory testing, whereas higher thresholds prioritize specificity but may miss individuals at moderate risk. The absence of standardized threshold recommendations across populations may therefore complicate direct implementation without local validation.
The use of electronic databases and digital platforms for automated risk identification is increasingly explored as a complementary strategy. Public availability of screening tools through websites or mobile applications may facilitate broader access and earlier engagement. For example, the German screening tool [27] enables individuals to assess their risk electronically. Digital applications allow users to enter personal data, which may be integrated with electronic health records, thereby supporting large-scale preventive initiatives at relatively low cost, particularly when combined with artificial intelligence (AI) approaches [52]. However, this strategy presents challenges, including the need to adapt or redesign tools for digital environments. Tools originally developed for traditional use may not retain equivalent performance when transferred to electronic platforms [75]. Beyond transferability, digital implementation introduces additional practical considerations. Effective deployment of web-based or mobile screening platforms requires user-friendly interface design, clear risk communication, and adaptation to varying levels of health literacy. Integration with electronic health record systems may enhance scalability and automated risk flagging; however, such integration necessitates data interoperability standards and robust data protection frameworks. Issues related to privacy, informed consent, and secure data storage are particularly relevant when personal health information is processed digitally. Moreover, disparities in access to digital technologies—the so-called “digital divide”—may limit equitable implementation, particularly in low-resource or older populations. These factors should be considered when translating screening tools from traditional to digital environments.
In addition to traditional regression-based risk scores, a subset of the identified tools employed machine-learning (ML) approaches for T2DM risk prediction [52,63]. ML-based models can capture complex nonlinear relationships between predictors and may improve predictive performance compared with conventional statistical methods. However, increased algorithmic complexity may reduce interpretability, particularly in black-box models where the contribution of individual predictors is less transparent [76]. In clinical practice, model transparency and explainability remain important for physician acceptance, patient communication, and responsible implementation. Among the studies included in this review, several ML-based models employed different algorithmic architectures, including artificial neural networks, support vector machines, and decision tree algorithms [46,52,62,63]. These models typically relied on structured training and validation procedures, such as train–test splitting or external validation datasets, to evaluate predictive performance. Nevertheless, compared with traditional questionnaire-based screening tools, ML models often require more complex data inputs and computational resources, which may limit their applicability in large-scale population screening. A summary of the methodological characteristics of the machine-learning–based models identified in this review is presented in Supplementary Table S3.
Beyond the initial development of screening models, their application to populations different from those in which they were originally derived requires independent validation, including reassessment of sensitivity and specificity [23,48,54]. These studies represent secondary external validation analyses of previously developed screening tools. As the present review focused on studies proposing or developing screening instruments, such validation studies were not included in the primary synthesis presented in Table 2. Nevertheless, external validation studies provide valuable insight into the cross-population applicability of screening tools. For example, evaluation of the FINDRISC tool in a Greek cohort of 7900 adults aged 35–75 years demonstrated a sensitivity of 81% and specificity of 60% [77], supporting its contextual usability outside its original setting. Similarly, the American Diabetes Association (ADA) Diabetes Risk Test has been validated in Middle Eastern populations, yielding a sensitivity of 78.9% and specificity of 82%, although performance characteristics differed from those observed in the original derivation cohort [78]. Further evidence from an Iranian population showed that the Australian Diabetes Risk Assessment Tool (AUSDRISK) achieved an AUC of 0.67, while FINDRISC demonstrated an AUC of 0.62 for identifying undiagnosed T2DM, indicating moderate discrimination and the potential need for recalibration [79]. Collectively, these findings underscore that screening tool performance may vary across demographic, cultural, and healthcare contexts, highlighting the importance of local validation prior to widespread implementation. In addition, the selection and validation of risk score thresholds represent an important component of practical implementation. Although most studies reported cut-off values used to classify individuals as high risk, the methodological justification for these thresholds was not consistently described. Ideally, threshold determination should be based on ROC curve analysis and evaluation of sensitivity–specificity trade-offs, followed by validation in independent populations and, where appropriate, recalibration to reflect local epidemiological characteristics. Establishing such transparent threshold validation procedures may facilitate the reliable application of screening tools in clinical practice and population-based prevention strategies.
Across the broader literature, anthropometric variables have been extensively investigated as practical and cost-effective indicators of metabolic and functional health. Enhanced anthropometric indices have been reported to improve predictive accuracy for metabolic syndrome beyond traditional measures [80,81], while age-adjusted waist circumference thresholds have been proposed as low-cost approaches for dysglycemia screening [82]. Anthropometric assessment has also been applied in geriatric and metabolic research to estimate muscle mass and support early detection of sarcopenia using both simplified equations and data-driven models [83,84]. More recently, incorporation of anthropometric data into machine learning models has been explored to improve risk prediction for prediabetes [85]. This broader body of evidence provides contextual support for the frequent inclusion of anthropometric variables in the screening tools mapped in the present review.
The public health relevance of early risk identification is further illustrated by epidemiological data from national contexts. In Greece, for example, surveys have reported a considerable prevalence of T2DM among adults, with a substantial proportion of cases remaining undiagnosed [86,87]. Although such epidemiological findings were not systematically analyzed within this scoping review, they provide contextual background that underscores the importance of effective screening approaches.
While previous research has evaluated individual T2DM risk models or specific screening tools, recent systematic reviews have primarily focused on predictive performance, validation procedures, or comparative accuracy within selected models [88,89,90]. In contrast, the present scoping review offers a structured mapping of the breadth and diversity of screening tools developed over the past decades. By identifying 58 distinct tools and documenting their variables, target populations, and implementation characteristics, this review highlights the methodological heterogeneity of the field. Rather than assessing comparative effectiveness, this mapping approach clarifies patterns of variable selection, structural design, and population targeting, thereby contributing to a more comprehensive understanding of the current landscape of T2DM risk assessment tools and their contextual applicability across populations.
Finally, several limitations should be acknowledged. In accordance with scoping review methodology, the objective of the present study was to map the breadth and characteristics of available screening tools rather than to conduct formal methodological quality appraisal or quantitative synthesis of outcomes. Consequently, no structured risk-of-bias assessment was undertaken. Considerable variation was observed in the reporting of validation procedures and performance metrics across studies: while some tools provided discrimination measures such as AUC values and sensitivity/specificity estimates, including external validation in independent populations, others reported limited or incomplete performance data. This heterogeneity restricts direct comparison and precludes definitive conclusions regarding comparative robustness or clinical superiority. Furthermore, the exclusion of secondary validation studies in new populations limits conclusions regarding cross-population generalizability. Cross-population validation is particularly important given ethnic, cultural, and healthcare system differences that may influence tool performance. Future systematic reviews focusing specifically on external validation data would provide valuable guidance for adaptation and contextual implementation. Nevertheless, the review provides a comprehensive mapping of 58 T2DM risk assessment tools across diverse geographical and population contexts. Furthermore, the mapped literature also revealed important gaps, including limited reporting on implementation settings, resource requirements, and real-world applicability. These gaps highlight priorities for future research rather than conclusions regarding tool performance.

5. Conclusions

This scoping review provides a comprehensive mapping of screening tools developed for the early identification of adults at high risk of type 2 diabetes. The findings reveal substantial heterogeneity in tool design, variable selection, and implementation context, reflecting adaptation to diverse demographic and healthcare settings. While most tools emphasize feasibility and low-cost application, differences in structure and reporting highlight the absence of standardized approaches across populations. Although this review did not evaluate comparative performance or diagnostic accuracy, it clarifies the current landscape of T2DM risk assessment and identifies important gaps related to validation, reporting consistency, and real-world implementation. Addressing these gaps through cross-population validation studies, harmonized reporting standards, and evaluation of digital integration strategies may strengthen the role of screening tools in primary prevention and public health planning. Policymakers and healthcare planners may therefore consider aligning screening tool selection with local resource availability and population characteristics to optimize implementation and population-level impact.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/healthcare14070839/s1, Table S1. PRISMA-ScR-Fillable-Checklist [15]; Table S2. Complete search strategies used in each database; Table S3. Machine-learning–based models [46,52,62,63].

Author Contributions

Conceptualization, O.A.; writing—original draft preparation, C.C. and D.S.; writing—review and editing, C.C., O.A. and A.Z. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

All data analyzed during this study are included in this manuscript.

Acknowledgments

During the preparation of this manuscript the authors used ChatGPT (OpenAI, version GPT-5.3) for the purpose of language refinement. The authors have reviewed and edited the output and take full responsibility for the content of this publication.

Conflicts of Interest

The authors declare no conflicts of interest.

Abbreviations

The following abbreviations are used in this manuscript:
DMDiabetes mellitus
T2DMType 2 diabetes mellitus
PDMPrediabetes
AUCArea Under the Curve
AROC/ROC(Area under the) Receiver Operating Characteristic
CIConfidence Interval
PPVPositive Predictive Value
NPVNegative Predictive Value
LRLogistic Regression
MLMachine Learning
BMIBody Mass Index
WCWaist Circumference
WHRWaist-to-Hip Ratio
BPBlood Pressure
HRHeart Rate
PPPulse Pressure
FPGFasting Plasma Glucose
HbA1cGlycated Hemoglobin
HDLHigh-Density Lipoprotein
LDLLow-Density Lipoprotein
TGTriglycerides
TCTotal Cholesterol
PRISMAPreferred Reporting Items for Systematic Reviews and Meta-Analyses
PRISMA-ScRPRISMA Extension for Scoping Reviews
PICOPopulation, Intervention, Comparison, Outcome
WHOWorld Health Organization
IDFInternational Diabetes Federation
EPICEuropean Prospective Investigation into Cancer and Nutrition
TUFTübingen Family Study
MeSyBePoMetabolic Syndrome Berlin Potsdam
CKDChronic kidney disease

References

  1. Ramachandran, A.; Snehalatha, C. Diabetes mellitus. In Public Health Nutrition; Gibney, M.J., Marghetts, B.M., Kearney, J.M., Arab, L., Eds.; Blackwell Science: New Jersey, NJ, USA, 2004; pp. 330–340. [Google Scholar]
  2. Leslie, R.D.; Palmer, J.; Schloot, N.C.; Lernmark, A. Diabetes at the crossroads: Relevance of disease classification to pathophysiology and treatment. Diabetologia 2016, 59, 13–20. [Google Scholar] [CrossRef] [PubMed]
  3. World Health Organization. Global Report on Diabetes; WHO: Geneva, Switzerland, 2016; Available online: https://www.who.int/publications/i/item/9789241565257 (accessed on 5 January 2021).
  4. World Health Organization. WHO Mortality Database; WHO: Geneva, Switzerland, 2016; Available online: https://www.who.int/data/data-collection-tools/who-mortality-database (accessed on 5 January 2021).
  5. International Diabetes Federation. IDF Diabetes Atlas, 6th ed.; IDF: Brussels, Belgium, 2013. [Google Scholar]
  6. International Diabetes Federation. IDF Diabetes Atlas, 8th ed.; IDF: Brussels, Belgium, 2017. [Google Scholar]
  7. International Diabetes Federation. IDF Diabetes Atlas, 11th ed.; IDF: Brussels, Belgium, 2025; Available online: https://diabetesatlas.org/resources/idf-diabetes-atlas-2025/ (accessed on 12 February 2026).
  8. Bommer, C.; Heesemann, E.; Sagalova, V.; Manne-Goehler, J.; Atun, R.; Bärnighausen, T.; Vollmer, S. The global economic burden of diabetes in adults aged 20–79 years: A cost-of-illness study. Lancet Diabetes Endocrinol. 2017, 5, 423–430. [Google Scholar] [CrossRef] [PubMed]
  9. Smith-Spangler, C.M.; Bhattacharya, J.; Goldhaber-Fiebert, J.D. Diabetes, its treatment, and catastrophic medical spending in 35 developing countries. Diabetes Care 2012, 35, 319–326. [Google Scholar] [CrossRef]
  10. World Health Organization. Tracking Universal Health Coverage: First Global Monitoring Report; WHO: Geneva, Switzerland, 2015; Available online: https://www.who.int/publications/i/item/9789241564977 (accessed on 5 January 2021).
  11. Beagley, J.; Guariguata, L.; Weil, C.; Motala, A.A. Global estimates of undiagnosed diabetes in adults. Diabetes Res. Clin. Pract. 2014, 103, 150–160. [Google Scholar] [CrossRef] [PubMed]
  12. World Health Organization. Diabetes: Equity and social determinants. In Equity, Social Determinants, and Public Health Programmes; WHO: Geneva, Switzerland, 2010. [Google Scholar]
  13. Saeedi, P.; Petersohn, I.; Salpea, P.; Malanda, B.; Karuranga, S.; Unwin, N.; Colagiuri, S.; Guariguata, L.; Motala, A.A.; Ogurtsova, K.; et al. Global and regional diabetes prevalence estimates for 2019 and projections for 2030 and 2045: Results from the International Diabetes Federation Diabetes Atlas, 9th edition. Diabetes Res. Clin. Pract. 2019, 157, 107843. [Google Scholar] [CrossRef]
  14. Chatterjee, S.; Khunti, K.; Davies, M.J. Type 2 diabetes. Lancet 2017, 389, 2239–2251. [Google Scholar] [CrossRef]
  15. Tricco, A.C.; Lillie, E.; Zarin, W.; O’Brien, K.K.; Colquhoun, H.; Levac, D.; Moher, D.; Peters, M.D.J.; Horsley, T.; Weeks, L.; et al. PRISMA Extension for Scoping Reviews (PRISMA-ScR): Checklist and explanation. Ann. Intern. Med. 2018, 169, 467–473. [Google Scholar] [CrossRef]
  16. Peters, M.D.J.; Godfrey, C.M.; McInerney, P.; Munn, Z.; Tricco, A.C.; Khalil, H. Scoping Reviews. In JBI Manual for Evidence Synthesis; Aromataris, E., Lockwood, C., Porritt, K., Pilla, B., Jordan, Z., Eds.; JBI: Adelaide, Australia, 2024; Chapter 10. [Google Scholar] [CrossRef]
  17. Herman, W.H.; Smith, P.J.; Thompson, T.J.; Engelgau, M.M.; Aubert, R.E. A new and simple questionnaire to identify people at increased risk for undiagnosed diabetes. Diabetes Care 1995, 18, 382–387. [Google Scholar] [CrossRef]
  18. Ruige, J.B.; de Neeling, J.N.D.; Kostense, P.J.; Bouter, L.M.; Heine, R.J. Performance of an NIDDM screening questionnaire based on symptoms and risk factors. Diabetes Care 1997, 20, 491–496. [Google Scholar] [CrossRef]
  19. Baan, C.A.; Ruige, J.B.; Stolk, R.P.; Witteman, J.C.; Dekker, J.M.; Heine, R.J.; Feskens, E.J. Performance of a predictive model to identify undiagnosed diabetes in a health care setting. Diabetes Care 1999, 22, 213–219. [Google Scholar] [CrossRef]
  20. Griffin, S.J.; Little, P.S.; Hales, C.N.; Kinmonth, A.L.; Wareham, N.J. Diabetes risk score: Towards earlier detection of type 2 diabetes in general practice. Diabetes Metab. Res. Rev. 2000, 16, 164–171. [Google Scholar] [CrossRef]
  21. Lindström, J.; Tuomilehto, J. The diabetes risk score: A practical tool to predict type 2 diabetes risk. Diabetes Care 2003, 26, 725–731. [Google Scholar] [CrossRef] [PubMed]
  22. Glumer, C.; Carstensen, B.; Sandbæk, A.; Lauritzen, T.; Jørgensen, T.; Borch-Johnsen, K. A Danish diabetes risk score for targeted screening: The Inter99 study. Diabetes Care 2004, 27, 727–733. [Google Scholar] [CrossRef] [PubMed]
  23. Ramachandran, A.; Snehalatha, C.; Vijay, V.; Wareham, N.J.; Colagiuri, S. Derivation and validation of diabetes risk score for urban Asian Indians. Diabetes Res. Clin. Pract. 2005, 70, 63–70. [Google Scholar] [CrossRef]
  24. Mohan, V.; Deepa, R.; Deepa, M.; Somannavar, S.; Datta, M. A simplified Indian diabetes risk score for screening for undiagnosed diabetic subjects. J. Assoc. Physicians India 2005, 53, 759–763. [Google Scholar]
  25. Schmidt, M.I.; Duncan, B.B.; Bang, H.; Pankow, J.; Ballantyne, C.; Golden, S.; Folsom, A.; Chambless, L. Identifying individuals at high risk for diabetes: The Atherosclerosis Risk in Communities study. Diabetes Care 2005, 28, 2013–2018. [Google Scholar] [CrossRef] [PubMed]
  26. Aekplakorn, W.; Bunnag, P.; Woodward, M.; Sritara, P.; Cheepudomwit, S.; Yamwong, S.; Yipintsoi, T.; Rajatanavin, R. Risk score for predicting incident diabetes in the Thai population. Diabetes Care 2006, 29, 1872–1877. [Google Scholar] [CrossRef]
  27. Schulze, M.B.; Hoffmann, K.; Boeing, H.; Linseisen, J.; Rohrmann, S.; Möhlig, M.; Pfeiffer, A.F.H.; Spranger, J.; Thamer, C.; Häring, H.U.; et al. An accurate risk score based on anthropometric, dietary, and lifestyle factors to predict the development of type 2 diabetes. Diabetes Care 2007, 30, 510–515. [Google Scholar] [CrossRef]
  28. Al-Lawati, J.A.; Tuomilehto, J. Diabetes risk score in Oman: A tool to identify prevalent type 2 diabetes among Arabs of the Middle East. Diabetes Res. Clin. Pract. 2007, 77, 438–444. [Google Scholar] [CrossRef]
  29. Wilson, P.W.; Meigs, J.B.; Sullivan, L.; Fox, C.S.; Nathan, D.M.; D’Agostino, R.B. Prediction of incident diabetes mellitus in middle-aged adults: The Framingham Offspring Study. Arch. Intern. Med. 2007, 167, 1068–1074. [Google Scholar] [CrossRef]
  30. Cabrera de León, A.; Coello, S.D.; Rodríguez Pérez, M.C.; Medina, M.B.; Almeida González, D.; Díaz, B.B.; de Fuentes, M.M.; Aguirre-Jaime, A. A simple clinical score for type 2 diabetes mellitus screening in the Canary Islands. Diabetes Res. Clin. Pract. 2008, 80, 128–133. [Google Scholar] [CrossRef]
  31. Heikes, K.E.; Eddy, D.M.; Arondekar, B.; Schlessinger, L. Diabetes risk calculator: A simple tool for detecting undiagnosed diabetes and pre-diabetes. Diabetes Care 2008, 31, 1040–1045. [Google Scholar] [CrossRef] [PubMed]
  32. Rahman, M.; Simmons, R.K.; Harding, A.H.; Wareham, N.J.; Griffin, S.J. A simple risk score identifies individuals at high risk of developing type 2 diabetes: A prospective cohort study. Fam. Pract. 2008, 25, 191–196. [Google Scholar] [CrossRef]
  33. Balkau, B.; Lange, C.; Fezeu, L.; Tichet, J.; de Lauzon-Guillain, B.; Czernichow, S.; Fumeron, F.; Froguel, P.; Vaxillaire, M.; Cauch, S.; et al. Predicting diabetes: Clinical, biological, and genetic approaches: Data from the Epidemiological Study on the Insulin Resistance Syndrome (DESIR). Diabetes Care 2008, 31, 2056–2061. [Google Scholar] [CrossRef] [PubMed]
  34. Chien, K.; Cai, T.; Hsu, H.; Su, T.; Chang, W.; Chen, M.; Lee, Y.; Hu, F.B. A prediction model for type 2 diabetes risk among Chinese people. Diabetologia 2009, 52, 443–450. [Google Scholar] [CrossRef] [PubMed]
  35. Hippisley-Cox, J.; Coupland, C.; Robson, J.; Sheikh, A.; Brindle, P. Predicting risk of type 2 diabetes in England and Wales: Prospective derivation and validation of QDScore. BMJ 2009, 338, b880. [Google Scholar] [CrossRef]
  36. Kahn, H.S.; Cheng, Y.J.; Thompson, T.J.; Imperatore, G.; Gregg, E.W. Two risk-scoring systems for predicting incident diabetes mellitus in U.S. adults aged 45 to 64 years. Ann. Intern. Med. 2009, 150, 741–751. [Google Scholar] [CrossRef]
  37. Bang, H.; Edwards, A.M.; Bomback, A.S.; Ballantyne, C.M.; Brillon, D.; Callahan, M.A.; Teutsch, S.M.; Mushlin, A.I.; Kern, L.M. Development and validation of a patient self-assessment score for diabetes risk. Ann. Intern. Med. 2009, 151, 775–783. [Google Scholar] [CrossRef]
  38. Sun, F.; Tao, Q.; Zhan, S. An accurate risk score for estimation 5-year risk of type 2 diabetes based on a health screening population in Taiwan. Diabetes Res. Clin. Pract. 2009, 85, 228–234. [Google Scholar] [CrossRef]
  39. Gao, W.G.; Dong, Y.H.; Pang, Z.C.; Nan, H.; Wang, S.; Ren, J.; Zhang, L.; Tuomilehto, J.; Qiao, Q. A simple Chinese risk score for undiagnosed diabetes. Diabet. Med. 2010, 27, 274–281. [Google Scholar] [CrossRef]
  40. Xin, Z.; Yuan, J.; Hua, L.; Ma, Y.H.; Zhao, L.; Lu, Y.; Yang, J.K. A simple tool detected diabetes and prediabetes in rural Chinese. J. Clin. Epidemiol. 2010, 63, 1030–1035. [Google Scholar] [CrossRef]
  41. Gray, L.J.; Taub, N.A.; Khunti, K.; Gardiner, E.; Hiles, S.; Webb, D.R.; Srinivasan, B.T.; Davies, M.J. The Leicester risk assessment score for detecting undiagnosed type 2 diabetes and impaired glucose regulation for use in a multiethnic UK setting. Diabet. Med. 2010, 27, 887–895. [Google Scholar] [CrossRef]
  42. Chen, L.; Magliano, D.J.; Balkau, B.; Colagiuri, S.; Zimmet, P.Z.; Tonkin, A.M.; Mitchell, P.; Phillips, P.J.; Shaw, J.E. AUSDRISK: An Australian type 2 diabetes risk assessment tool based on demographic, lifestyle and simple anthropometric measures. Med. J. Aust. 2010, 192, 197–202. [Google Scholar] [CrossRef] [PubMed]
  43. Robinson, C.A.; Agarwal, G.; Nerenberg, K. Validating the CANRISK prognostic model for assessing diabetes risk in Canada’s multi-ethnic population. Chronic Dis. Inj. Can. 2011, 32, 19–31. [Google Scholar] [CrossRef] [PubMed]
  44. Soewondo, P.; Pramono, L.A. Prevalence, characteristics, and predictors of pre-diabetes in Indonesia. Med. J. Indones. 2011, 20, 283–294. [Google Scholar] [CrossRef]
  45. Gray, L.J.; Davies, M.J.; Hiles, S.; Taub, N.; Webb, D.; Srinivasan, B.; Khunti, K. Detection of impaired glucose regulation and/or type 2 diabetes mellitus using primary care electronic data in a multiethnic UK community setting. Diabetologia 2012, 55, 959–966. [Google Scholar] [CrossRef]
  46. Shankaracharya; Odedra, D.; Samanta, S.; Vidyarthi, A.S. Computational intelligence-based diagnosis tool for the detection of prediabetes and type 2 diabetes in India. Rev. Diabet. Stud. 2012, 9, 55–62. [Google Scholar] [CrossRef]
  47. Heianza, Y.; Arase, Y.; Hsieh, S.D.; Saito, K.; Tsuji, H.; Kodama, S.; Tanaka, S.; Ohashi, Y.; Shimano, H.; Yamada, N.; et al. Development of a new scoring system for predicting the 5-year incidence of type 2 diabetes in Japan: The Toranomon Hospital Health Management Center Study 6 (TOPICS 6). Diabetologia 2012, 55, 3213–3223. [Google Scholar] [CrossRef]
  48. Lee, Y.H.; Bang, H.; Kim, H.C.; Kim, H.M.; Park, S.W.; Kim, D.J. A simple screening score for diabetes for the Korean population: Development, validation, and comparison with other scores. Diabetes Care 2012, 35, 1723–1730. [Google Scholar] [CrossRef]
  49. Zhou, X.; Qiao, Q.; Ji, L.; Ning, F.; Yang, W.; Weng, J.; Shan, Z.; Tian, H.; Ji, Q.; Lin, L.; et al. Nonlaboratory-based risk assessment algorithm for undiagnosed type 2 diabetes developed on a nationwide diabetes survey. Diabetes Care 2013, 36, 3944–3952. [Google Scholar] [CrossRef] [PubMed]
  50. Heianza, Y.; Arase, Y.; Saito, K.; Hsieh, S.D.; Tsuji, H.; Kodama, S.; Tanaka, S.; Ohashi, Y.; Shimano, H.; Yamada, N.; et al. Development of a screening score for undiagnosed diabetes and its application in estimating absolute risk of future type 2 diabetes in Japan: Toranomon Hospital Health Management Center Study 10 (TOPICS 10). J. Clin. Endocrinol. Metab. 2013, 98, 1051–1060. [Google Scholar] [CrossRef]
  51. Gray, L.J.; Barros, H.; Raposo, L.; Khunti, K.; Davies, M.J.; Santos, A.C. The development and validation of the Portuguese risk score for detecting type 2 diabetes and impaired fasting glucose. Prim. Care Diabetes 2013, 7, 11–18. [Google Scholar] [CrossRef]
  52. Choi, S.B.; Kim, W.J.; Yoo, T.K.; Park, J.S.; Chung, J.W.; Lee, Y.; Kang, E.S.; Kim, D.W. Screening for prediabetes using machine learning models. Comput. Math. Methods Med. 2014, 2014, 618976. [Google Scholar] [CrossRef]
  53. Dugee, O.; Janchiv, O.; Jousilahti, P.; Sakhiya, A.; Palam, E.; Nuorti, J.P.; Peltonen, M. Adapting existing diabetes risk scores for an Asian population: A risk score for detecting undiagnosed diabetes in the Mongolian population. BMC Public Health 2015, 15, 938. [Google Scholar] [CrossRef] [PubMed]
  54. Yan, F.; Cha, E.; Lee, E.T.; Mayberry, R.M.; Wang, W.; Umpierrez, G. A self-assessment tool for screening young adults at risk of type 2 diabetes using Strong Heart Family Study data. Diabetes Educ. 2016, 42, 607–617. [Google Scholar] [CrossRef] [PubMed]
  55. Zhang, M.; Zhang, H.; Wang, C.; Ren, Y.; Wang, B.; Zhang, L.; Yang, X.; Zhao, Y.; Han, C.; Pang, C.; et al. Development and validation of a risk-score model for type 2 diabetes: A cohort study of a rural adult Chinese population. PLoS ONE 2016, 11, e0152054. [Google Scholar] [CrossRef] [PubMed]
  56. Barengo, N.C.; Tamayo, D.C.; Tono, T.; Tuomilehto, J. A Colombian diabetes risk score for detecting undiagnosed diabetes and impaired glucose regulation. Prim. Care Diabetes 2017, 11, 86–93. [Google Scholar] [CrossRef]
  57. Chen, X.; Wu, Z.; Chen, Y.; Wang, X.; Zhu, J.; Wang, N.; Jiang, Q.; Fu, C. Risk score model of type 2 diabetes prediction for rural Chinese adults: The Rural Deqing Cohort Study. J. Endocrinol. Investig. 2017, 40, 1115–1123. [Google Scholar] [CrossRef]
  58. Zhou, H.; Li, Y.; Liu, X.; Xu, F.; Li, L.; Yang, K.; Qian, X.; Liu, R.; Bie, R.; Wang, C. Development and evaluation of a risk score for type 2 diabetes mellitus among middle-aged Chinese rural population based on the RuralDiab Study. Sci. Rep. 2017, 7, 42685. [Google Scholar] [CrossRef]
  59. Wen, J.; Hao, J.; Liang, Y.; Li, S.; Cao, K.; Lu, X.; Lu, X.; Wang, N. A non-invasive risk score for predicting incident diabetes among rural Chinese people: A village-based cohort study. PLoS ONE 2017, 12, e0186172. [Google Scholar] [CrossRef]
  60. Sulaiman, N.; Mahmoud, I.; Hussein, A.; Elbadawi, S.; Abusnana, S.; Zimmet, P.; Shaw, J. Diabetes risk score in the United Arab Emirates: A screening tool for the early detection of type 2 diabetes mellitus. BMJ Open Diabetes Res. Care 2018, 6, e000489. [Google Scholar] [CrossRef]
  61. Félix-Martínez, G.J.; Godínez-Fernández, J.R. Screening models for undiagnosed diabetes in Mexican adults using clinical and self-reported information. Endocrinol. Diabetes Nutr. 2018, 65, 380–386. [Google Scholar] [CrossRef]
  62. Pei, D.; Zhang, C.; Quan, Y.; Guo, Q. Identification of potential type II diabetes in a Chinese population with a sensitive decision tree approach. J. Diabetes Res. 2018, 2018, 4248218. [Google Scholar] [CrossRef]
  63. Abbas, H.T.; Alic, L.; Erraguntla, M.; Ji, J.X.; Abdul-Ghani, M.; Abbasi, Q.H.; Qaraqe, M.K. Predicting long-term type 2 diabetes with support vector machine using oral glucose tolerance test. PLoS ONE 2019, 14, e0219636. [Google Scholar] [CrossRef]
  64. Srugo, S.; Groh, M.; Jiang, Y.; Morrison, H.; Villeneuve, P. Evaluating the utility of self-reported questionnaire data to screen for dysglycemia in young adults: Findings from the US National Health and Nutrition Examination Survey. Prev. Med. 2019, 120, 50–59. [Google Scholar] [CrossRef] [PubMed]
  65. Wu, J.; Hou, X.; Chen, L.; Chen, P.; Wei, L.; Jiang, F.; Bao, Y.; Jia, W. Development and validation of a non-invasive assessment tool for screening prevalent undiagnosed diabetes in middle-aged and elderly Chinese. Prev. Med. 2019, 119, 145–152. [Google Scholar] [CrossRef] [PubMed]
  66. Bahijri, S.; Al-Raddadi, R.; Ajabnoor, G.; Jambi, H.; Al Ahmadi, J.; Borai, A.; Barengo, N.C.; Tuomilehto, J. Dysglycemia risk score in Saudi Arabia: A tool to identify people at high future risk of developing type 2 diabetes. J. Diabetes Investig. 2020, 11, 844–855. [Google Scholar] [CrossRef]
  67. Lowe, J.; Ke, C.; Singh, K.; Gobin, R.; Lebovic, G.; Ostrow, B. Development and validation of a new diabetes risk score in Guyana. Diabetes Ther. 2020, 11, 873–883. [Google Scholar] [CrossRef]
  68. Buccheri, E.; Dell’Aquila, D.; Russo, M. Artificial intelligence in health data analysis: The Darwinian evolution theory suggests an extremely simple and zero-cost large-scale screening tool for prediabetes and type 2 diabetes. Diabetes Res. Clin. Pract. 2021, 174, 108722. [Google Scholar] [CrossRef]
  69. Abbas, M.; Mall, R.; Errafii, K.; Lattab, A.; Ullah, E.; Bensmail, H.; Arredouani, A. Simple risk score to screen for prediabetes: A cross-sectional study from the Qatar Biobank cohort. J. Diabetes Investig. 2021, 12, 988–997. [Google Scholar] [CrossRef] [PubMed]
  70. Cho, E.; Min, D.; Lee, H.S. Development and Validation of an Undiagnosed Diabetes Screening Tool: Based on the Korean National Health and Nutrition Examination Survey (2010–2016). Healthcare 2021, 9, 1138. [Google Scholar] [CrossRef]
  71. Henjum, S.; Hjellset, V.T.; Andersen, E.; Flaaten, M.Ø.; Morseth, M.S. Developing a risk score for undiagnosed prediabetes or type 2 diabetes among Saharawi refugees in Algeria. BMC Public Health 2022, 22, 720. [Google Scholar] [CrossRef]
  72. Sadek, K.; Abdelhafez, I.; Al-Hashimi, I.; Al-Shafi, W.; Tarmizi, F.; Al-Marri, H.; Alzohari, N.; Balideh, M.; Carr, A. Screening for diabetes and impaired glucose metabolism in Qatar: Models’ development and validation. Prim. Care Diabetes 2022, 16, 69–77. [Google Scholar] [CrossRef]
  73. Dong, W.; Tse, T.Y.E.; Mak, L.I.; Wong, C.K.H.; Wan, Y.F.E.; Tang, H.M.E.; Chin, W.Y.; Bedford, L.E.; Yu, Y.T.E.; Ko, W.K.W.; et al. Non-laboratory-based risk assessment model for case detection of diabetes mellitus and pre-diabetes in primary care. J. Diabetes Investig. 2022, 13, 1374–1386. [Google Scholar] [CrossRef]
  74. Alkattan, A.; Al-Zeer, A.; Alsaawi, F.; Alyahya, A.; Alnasser, R.; Alsarhan, R.; Almusawi, M.; Alabdulaali, D.; Mahmoud, N.; Al-Jafar, R.; et al. The utility of a machine learning model in identifying people at high risk of type 2 diabetes mellitus. Expert. Rev. Endocrinol. Metab. 2024, 19, 513–522. [Google Scholar] [CrossRef]
  75. Stiglic, G.; Pajnkihar, M. Evaluation of major online diabetes risk calculators and computerized predictive models. PLoS ONE 2015, 10, e0142827. [Google Scholar] [CrossRef]
  76. Rudin, C. Stop explaining black box machine learning models for high stakes decisions and use interpretable models instead. Nat. Mach. Intell. 2019, 1, 206–215. [Google Scholar] [CrossRef] [PubMed]
  77. Makrilakis, K.; Liatis, S.; Grammatikou, S.; Perrea, D.; Stathi, C.; Tsiligros, P.; Katsilambros, N. Validation of the Finnish diabetes risk score (FINDRISC) questionnaire for screening for undiagnosed type 2 diabetes, dysglycaemia and the metabolic syndrome in Greece. Diabetes Metab. 2011, 37, 144–151. [Google Scholar] [CrossRef]
  78. Aldayel, F.A.; Belal, M.A.; Alsheikh, A.M. The Validity of the American Diabetes Association’s Diabetes Risk Test in a Saudi Arabian Population. Cureus 2021, 13, e18018. [Google Scholar] [CrossRef] [PubMed]
  79. Mahmoodzadeh, S.; Jahani, Y.; Najafipour, H.; Sanjari, M.; Shadkam-Farokhi, M.; Shahesmaeili, A. External Validation of Finnish Diabetes Risk Score and Australian Diabetes Risk Assessment Tool Prediction Models to Identify People with Undiagnosed Type 2 Diabetes: A Cross-sectional Study in Iran. Int. J. Endocrinol. Metab. 2022, 20, e127114. [Google Scholar] [CrossRef] [PubMed]
  80. Tian, T.; Zhang, J.; Zhu, Q.; Xie, W.; Wang, Y.; Dai, Y. Predicting value of five anthropometric measures in metabolic syndrome among Jiangsu Province, China. BMC Public Health 2020, 20, 1317. [Google Scholar] [CrossRef]
  81. Wu, Y.; Li, H.; Tao, X.; Fan, Y.; Gao, Q.; Yang, J. Optimised anthropometric indices as predictive screening tools for metabolic syndrome in adults: A cross-sectional study. BMJ Open 2021, 11, e043952. [Google Scholar] [CrossRef]
  82. Buccheri, E.; Dell’Aquila, D.; Russo, M. Stratified analysis of the age-related waist circumference cut-off model for the screening of dysglycemia at zero cost. Obes. Med. 2022, 31, 100398. [Google Scholar] [CrossRef]
  83. Buccheri, E.; Dell’Aquila, D.; Russo, M.; Chiaramonte, R.; Musumeci, G.; Vecchio, M. Can artificial intelligence simplify the screening of muscle mass loss? Heliyon 2023, 9, e16323. [Google Scholar] [CrossRef]
  84. Buccheri, E.; Dell’Aquila, D.; Russo, M.; Chiaramonte, R.; Vecchio, M. Appendicular skeletal muscle mass in older adults can be estimated with a simple equation using a few zero-cost variables. J. Geriatr. Phys. Ther. 2024, 47, E149–E158. [Google Scholar] [CrossRef]
  85. De Silva, K.; Jonsson, D.; Demmer, R.T. A combined strategy of feature selection and machine learning to identify predictors of prediabetes. J. Am. Med. Inform. Assoc. 2019, 27, 396–406. [Google Scholar] [CrossRef]
  86. Magriplis, E.; Dimakopoulos, I.; Karageorgou, D.; Mitsopoulou, A.V.; Bakogianni, I.; Micha, R.; Michas, G.; Ntouroupi, T.; Tsaniklidou, S.M.; Argyri, K.; et al. Aims, design and preliminary findings of the Hellenic National Nutrition and Health Survey (HNNHS). BMC Med. Res. Methodol. 2019, 19, 37. [Google Scholar] [CrossRef]
  87. Magriplis, E.; Panagiotakos, D.; Papakonstantinou, E.; Mitsopoulou, A.V.; Karageorgou, D.; Dimakopoulos, I.; Bakogianni, I.; Chourdakis, M.; Micha, R.; Michas, G.; et al. Prevalence of type 2 diabetes mellitus in a representative sample of Greek adults and its association with modifiable risk factors: Results from the Hellenic National Nutrition and Health Survey. Public Health 2021, 197, 75–82. [Google Scholar] [CrossRef]
  88. Cheng, W.H.G.; Mi, Y.; Dong, W.; Tse, E.T.Y.; Wong, C.K.H.; Bedford, L.E.; Lam, C.L.K. Non-Laboratory-Based Risk Prediction Tools for Undiagnosed Pre-Diabetes: A Systematic Review. Diagnostics 2023, 13, 1294. [Google Scholar] [CrossRef] [PubMed]
  89. Liu, Y.; Feng, W.; Lou, J.; Qiu, W.; Shen, J.; Zhu, Z.; Hua, Y.; Zhang, M.; Billong, L.F. Performance of a prediabetes risk prediction model: A systematic review. Heliyon 2023, 9, e15529. [Google Scholar] [CrossRef] [PubMed]
  90. Li, M.; Tan, A.; Song, L.; Qiu, X.; Zhou, L.; Wen, Z. Performance of diabetes risk prediction models: A systematic review and meta-analysis. Endocr. Connect. 2025, 14, e250353. [Google Scholar] [CrossRef] [PubMed]
Healthcare 14 00839 g001
Figure 1. Scoping review flow diagram.
Figure 1. Scoping review flow diagram.
Healthcare 14 00839 g001
Healthcare 14 00839 g002
Figure 2. Frequency of variables included across the 58 identified T2DM screening tools, presented in descending order. The “Others” category includes heterogeneous variables reported in fewer studies.
Figure 2. Frequency of variables included across the 58 identified T2DM screening tools, presented in descending order. The “Others” category includes heterogeneous variables reported in fewer studies.
Healthcare 14 00839 g002
Healthcare 14 00839 g003
Figure 3. World map showing countries where the identified T2DM screening tools were developed or applied. Highlighted countries correspond to the geographic settings reported in the included studies (Canary Islands are represented under Spain).
Figure 3. World map showing countries where the identified T2DM screening tools were developed or applied. Highlighted countries correspond to the geographic settings reported in the included studies (Canary Islands are represented under Spain).
Healthcare 14 00839 g003
Table 1. PCC framework used to structure the search strategy [16].
Table 1. PCC framework used to structure the search strategy [16].
PCC ElementDescriptionSearch Terms (Examples)
PopulationAdults at risk of developing type 2 diabetesadult*, population*, individual*, participant*, patient*, undiagnosed
ConceptScreening tools, risk scores, questionnaires, and predictive models used to identify individuals at high risk of developing type 2 diabetes.screening tool*, risk score*, risk assessment*, questionnaire*, prediction model*, diabetes risk
ContextEarly identification of individuals at high risk of type 2 diabetes in general or clinical populationstype 2 diabetes, T2DM, prediabetes, dysglycemia, early diagnosis, risk identification
Table 2. Screening variables reported in the 58 eligible studies. The category “Others” includes heterogeneous predictors reported infrequently across studies, such as comorbidities, medication history (e.g., antihypertensive treatment), and specific biochemical or metabolic indicators not consistently categorized across tools. Outcome refers to the primary endpoint targeted for the development of each screening tool.
Table 2. Screening variables reported in the 58 eligible studies. The category “Others” includes heterogeneous predictors reported infrequently across studies, such as comorbidities, medication history (e.g., antihypertensive treatment), and specific biochemical or metabolic indicators not consistently categorized across tools. Outcome refers to the primary endpoint targeted for the development of each screening tool.
StudyCountryScreening Variables
SexAgeHeightWeightObesityBMIPhysical ActivityFamily History of DiadetesHypertension/Blood PressureWaist CircumferenceWaist-to-Hip RatioEthnicity/RaceDietary HabitsSmoking StatusAlcohol ConsumptionEducational LevelSocioeconomic StatusGestational DiabetesOutcomeOthers
1Herman et al., 1995 [17]USA ✓✓✓ ✓✓✓ Type 2 DiabetesMacrosomic infant
2Ruige et al., 1997 [18]The Netherlands✓✓ ✓ ✓✓ Type 2 DiabetesPain during walking with need to slow down, shortness of breath when walking with people of the same age, frequent thirst, and reluctance to use a bicycle for transportation
3Baan et al., 1999 [19]The Netherlands✓✓ ✓ ✓✓✓ Type 2 Diabetes
4Griffin et al., 2000 [20]UK✓✓ ✓ ✓✓ ✓ Type 2 DiabetesSteroid medication
5Lindstrom and Tuomilehto 2003 [21]Finland ✓ ✓✓ ✓✓ ✓ Type 2 Diabetes History of high blood glucose
6Glumer et al., 2004 [22]Denmark✓✓ ✓✓✓✓ Type 2 Diabetes
7Ramachandran et al., 2005 [23]India ✓ ✓✓✓ ✓ Type 2 Diabetes
8Mohan et al., 2005 [24]India ✓ ✓✓ ✓ Type 2 Diabetes
9Schmidt et al., 2005 [25]USA ✓✓ ✓✓✓ ✓ Type 2 Diabetes
10Aekplakorn et al., 2006 [26]Thailand ✓ ✓ ✓✓✓ Type 2 Diabetes
11Schulze et al., 2007 [27]Germany ✓✓ ✓ ✓✓ ✓✓✓ Type 2 Diabetes
12Al-Lawati and Tuomilehto 2007 [28]Oman ✓ ✓ ✓✓✓ Type 2 Diabetes
13Wilson et al., 2007 [29]USA✓✓ ✓ ✓ Type 2 DiabetesMetabolic syndrome traits (simple clinical model), 2 h post–oral glucose tolerance test
14Cabrera de León et al., 2008 [30]Canary Islands ✓ ✓✓ ✓Type 2 DiabetesWaist-to-height ratio
15Heikes et al., 2008 [31]USA✓✓✓✓ ✓✓✓✓✓ ✓ ✓Type 2 Diabetes and PrediabetesTaking cholesterol medication, high cholesterol
16Rahman et al., 2008 [32]UK✓✓ ✓ ✓✓ ✓ Type 2 Diabetes Prescription of steroids
17Balkau et al., 2008 [33]France✓✓ ✓ ✓✓ Type 2 DiabetesFasting glucose, HDL-C
18Chien et al., 2009 [34]Taiwan ✓ ✓ Type 2 Diabetes Elevated fasting glucose, triacylglycerol, white blood cell count and a higher HDL-C
19Hippisley-Cox et al., 2009 [35]UK✓✓ ✓ ✓✓ ✓ Type 2 Diabetes Townsend deprivation score, cardiovascular disease, and current use of corticosteroids.
20Kahn et al., 2009 [36]USA ✓ ✓ ✓✓✓ ✓ ✓✓ Type 2 DiabetesGlucose, triglycerides, low HDL-C concentration, short stature, high uric acid, rapid pulse
21Bang et al., 2009 [37]USA✓✓ ✓ ✓✓✓ Type 2 Diabetes
22Sun et al., 2009 [38]Taiwan✓✓ ✓ ✓✓✓ ✓ ✓ Type 2 Diabetes FPG
23Gao et al., 2010 [39]China ✓ ✓ ✓ Type 2 Diabetes
24Xin et al., 2010 [40]China ✓ ✓ ✓✓ ✓ Type 2 Diabetes and Prediabetes
25Gray et al., 2010 [41]UK ✓ ✓ ✓✓✓ ✓ Type 2 Diabetes and Impaired Glucose Regulation
26Chen et al., 2010 [42]Australia✓✓ ✓✓✓✓ ✓ ✓ Type 2 Diabetes History of high blood glucose level
27Robinson et al., 2011 [43]Canada✓✓✓✓ ✓✓✓✓✓ ✓✓ ✓ Type 2 DiabetesHistory of high blood glucose level
28Soewondo and Pramono 2011 [44]Indonesia✓✓ ✓ ✓✓ ✓ ✓✓ Prediabetes
29Gray et al., 2011 [45]UK✓✓ ✓ ✓✓ ✓ Type 2 Diabetes and Impaired Glucose Regulation
30Shankaracharya et al., 2012 [46]India✓✓ ✓✓✓✓✓ ✓✓✓ ✓ Type 2 Diabetes and PrediabetesHigh cholesterol
31Heianza et al., 2012 [47]Japan✓✓ ✓ ✓ ✓ Type 2 Diabetes FPG, HbA1c
32Lee et al., 2012 [48]South Korea ✓ ✓✓✓ ✓✓ Type 2 Diabetes
33Zhou et al., 2013 [49]China✓✓ ✓ ✓✓✓ Type 2 Diabetes
34Heianza et al., 2013 [50]Japan✓✓ ✓ ✓✓ ✓ Type 2 Diabetes
35Gray et al., 2013 [51]Portugal✓✓ ✓ ✓ Type 2 Diabetes and Impaired Fasting Glucose
36Choi et al., 2014 [52]South Korea✓✓ ✓ ✓✓✓ ✓ Prediabetes
37Dugee et al., 2015 [53]Mongolia✓✓✓✓ ✓✓ ✓✓ ✓✓ Type 2 DiabetesHistory of elevated glucose level
38Yan et al., 2016 [54]USA✓✓ ✓✓ ✓✓✓ ✓✓✓✓ Type 2 Diabetes Indian Health Service clinic checkup, past and current health history, cardiometabolic risk, TC, TG, HDL-C, LDL-C, FPG level
39Zhang et al., 2016 [55]China✓✓ ✓✓✓ ✓ ✓✓✓✓✓ Type 2 DiabetesPP, HR, TC, TG, HDL-C, LDL-C, FPG
40Barengo et al., 2017 [56]Colombia✓✓ ✓✓✓✓ ✓ Type 2 Diabetes and Impaired Glucose RegulationFasting glucose, 2 h glucose, history of hyperglycemia, impaired glucose regulation
41Chen et al., 2017 [57]China ✓ ✓✓ ✓✓ ✓ Type 2 DiabetesImpaired fasting glucose
42Zhou et al., 2017 [58]China✓✓ ✓✓✓✓✓ Type 2 DiabetesHistory of dyslipidemia
43Wen et al., 2017 [59]China ✓ ✓✓✓ ✓✓ ✓ ✓ Type 2 DiabetesFPG
44Sulaiman et al., 2018 [60]United Arab Emirates ✓ ✓ ✓✓ ✓ Type 2 Diabetes
45Félix-Martínez and Godínez-Fernández 2018 [61]Mexico ✓✓ ✓✓✓ Type 2 Diabetes
46Pei et al., 2019 [62]China✓✓ ✓✓✓✓ ✓ Type 2 DiabetesHistory of cardiovascular disease or stroke, work-related stress
47Abbas et al., 2019 [63]USA ✓ ✓ ✓ Type 2 Diabetes Plasma glucose and insulin concentrations before glucose intake and at three time-points thereafter (30, 60 and 120 min)
48Srugo et al., 2019 [64]USA ✓ ✓✓✓ ✓ ✓ ✓ DysglycemiaDelivered infant birthweight
49Wu et al., 2019 [65]China ✓ ✓ ✓✓✓ Type 2 Diabetes
50Bahijri et al., 2020 [66]Saudi Arabia✓✓ ✓ ✓ Type 2 DiabetesHistory of hyperglycemia
51Lowe et al., 2020 [67]Guyana✓✓ ✓ ✓✓ ✓ Type 2 DiabetesKnown diabetes, current treatment
52Buccheri et al., 2021 [68]USA ✓ ✓ Type 2 Diabetes and Prediabetes
53Abbas et al., 2021 [69]Qatar✓✓ ✓ ✓✓ Prediabetes
54Cho et al., 2021 [70]South Korea✓✓ ✓✓ ✓ ✓ Type 2 DiabetesHealth check-ups
55Henjum et al., 2022 [71]Algeria ✓ ✓ ✓ Type 2 Diabetes and Prediabetes
56Sadek et al., 2022 [72]Qatar✓✓ ✓ ✓ ✓ Type 2 Diabetes and impaired glucose metabolismHistory of hyperlipidemia
57Dong et al., 2022 [73]China ✓ ✓✓ ✓✓✓ ✓ Type 2 Diabetes and PrediabetesSleep duration
58Alkattan et al., 2024 [74]Saudi Arabia✓✓ ✓ ✓ ✓Type 2 DiabetesDyslipidemia, cardiovascular disorders, hyperplasia of the prostate, ophthalmic disorders, connective tissue disorders, thyroid orders, gastrointestinal disorders, CKD, spondylopathies, refractive errors, nutritional deficiencies, genitourinary infections, and polyneuropathies
HDL-C: high-density lipoprotein cholesterol; FPG: fasting plasma glucose; HbA1c: gylcated hemoglobin; LDL-C: low-density lipoprotein cholesterol; HR: heart rate; TC: total cholesterol; TG: triglyceride; PP: pulse pressure; CKD: chronic kidney disease. ✓ indicates that the variable is included in the screening tool. Narrative text is used in the “Others” column to describe screening variables that are specific to individual models and not commonly included across screening tools.
Table 3. Accuracy of the screening tools identified in the 58 eligible studies.
Table 3. Accuracy of the screening tools identified in the 58 eligible studies.
StudyCountryAccuracy of Screening Tools
AUCSensitivity (%)Specificity (%)(%) Positive Predictive Value(%) Negative Predictive ValueComments
1Herman et al., 1995 [17]USA 796510
2Ruige et al., 1997 [18]The Netherlands 72566.598
3Baan et al., 1999 [19]The Netherlands0.74 (0.70–0.78) and 0.68 (0.64–0.72) Two predictive models compared
4Griffin et al., 2000 [20]UK807772
5Lindstrom and Tuomilehto 2003 [21]Finland 78 and 8177 and 7613 and 5 Two cohort analyses
6Glumer et al., 2004 [22]DenmarkThe AROC curve value was 0.804 (95% CI, 0.765–0.838) for the first half of the Inter99 population, 0.761 (95% CI, 0.720–0.803) for the second half of the Inter99 population, and 0.803 (95% CI, 0.721–0.876) for the ADDITION pilot study.7672 False-negative risk comparison
7Ramachandran et al., 2005 [23]India 76.6, 72.4 and 73.759.9, 59.0 and 61.09.4, 8.3 and 12.297.9, 97.6 and 96.9Cohort-specific risk threshold
8Mohan et al., 2005 [24]India0.698 (95% CI, 0.663–0.733)72.560.117.095.1Overall model accuracy reported
9Schmidt et al., 2005 [25]USA0.75 and 0.78, respectively40–8750–86 Metabolic syndrome–based score evaluation
10Aekplakorn et al., 2006 [26]Thailand0.817674
11Schulze et al., 2007 [27]Germany0.84 in the EPIC-Potsdam and 0.82 in the EPIC-Heidelberg studies.0.56 in the TUF and 0.45 in the MeSyBePo studies Cohort-based risk gradient analysis
12Al-Lawati and Tuomilehto 2007 [28]Oman 78.6 and 62.873.4 and 78.2 Cohort-specific cut-off threshold
13Wilson et al., 2007 [29]USA0.85
14Cabrera de León et al., 2008 [30]Canary Islands Age-stratified screening performance
15Heikes et al., 2008 [31]USA0.8588 and 7575 and 6514 and 4999.3 and 85Diabetes and prediabetes detection models
16Rahman et al., 2008 [32]UK0.745
17Balkau et al., 2008 [33]France0.713 for men and 0.827 for women
18Chien et al., 2009 [34]Taiwan 5278
19Hippisley-Cox et al., 2009 [35]UK Sex-specific risk model performance
20Kahn et al., 2009 [36]USA0.71 (95% CI, 0.69–0.73) and 0.79 (95% CI, 0.77–0.81)69 and 7464 and 71 Basic vs. enhanced scoring systems
21Bang et al., 2009 [37]USA 796710 Positive likelihood ratio reported
22Sun et al., 2009 [38]Taiwan0.848 (95% CI, 0.829–0.868)72.26% and 76.64%82.84% and 76.00%19.87 and 19.47 Separate cohort validation
23Gao et al., 2010 [39]China 85.6 and 75.521.1 and 43.6 Sex-specific model evaluation
24Xin et al., 2010 [40]China 74.6 and 65.371.6 and 72.523.6 and 33.296.0 and 90.7Diabetes and combined outcome models
25Gray et al., 2010 [41]UK 7281
26Chen et al., 2010 [42]Australia0.78 (95% CI, 0.76–0.81)74.067.712.7 Independent cohort validation
27Robinson et al., 2011 [43]Canada0.80672.768.1
28Soewondo and Pramono 2011 [44]Indonesia
29Gray et al., 2012 [45]UK70.1 (95% CI, 68.4–71.7)
30Shankaracharya et al., 2012 [46]India 99.599.07 Classification accuracy reported
31Heianza et al., 2012 [47]Japan0.771 (95% CI, 0.758–0.784)72.768.16.498.8
32Lee et al., 2012 [48]South Korea0.7381546 Validation dataset performance metrics
33Zhou et al., 2013 [49]China0.748 (95% CI, 0.739–0.756) in the exploratory population, 0.725 (95% CI, 0.683–0.767) in validation 1, and 0.702 (95% CI, 0.680–0.724) in validation 292.3 and 86.835.5 and 38.8 Dual validation cohorts
34Heianza et al., 2013 [50]Japan0.887 (95% CI, 0.871–0.903)83.779.0
35Gray et al., 2013 [51]Portugal0.74 (95% CI, 0.72–0.77)73.2 and 69.155.5 and 63.327.0 and 38.090.2 and 86.2Cross-sectional and prospective validation
36Choi et al., 2014 [52]South Korea 74 and 7070 and 61
37Dugee et al., 2015 [53]Mongolia0.77 (95% CI, 0.71–0.82)81.458.910.998.1Defined cut-off threshold
38Yan et al., 2016 [54]USA0.68
39Zhang et al., 2016 [55]China0.768 (95% CI, 0.760–0.776)66.774.0
40Barengo et al., 2017 [56]Colombia0.74 (95%CI, 0.70–0.79)736710.697.9
41Chen et al., 2017 [57]China0.705 and 0.75470.5 and 63.160.4 and 75.92.5 and 2.599.3 (noninvasive model)Noninvasive vs. laboratory model
42Zhou et al., 2017 [58]China0.723 (95% CI, 0.710–0.735)67.967.8
43Wen et al., 2017 [59]China0.68674.3258.82
44Sulaiman et al., 2018 [60]United Arab Emirates0.82 (95% CI, 0.78–0.86).75.47045.389.6
45Félix-Martínez and Godínez-Fernández 2018 [61]Mexico0.70 and 0.660.74 and 0.760.62 and 0.55 National survey–derived models
46Pei et al., 2019 [62]China94.2%, 94.0%, 94.2%, and 94.8%.
47Abbas et al., 2019 [63]USA 80.09 Average model accuracy reported
48Srugo et al., 2019 [64]USA0.69 and 0.92 Sex-specific prediction models
49Wu et al., 2019 [65]China0.654 (95% CI, 0.629–0.680) and 0.684 (95% CI, 0.662–0.705) Sex-stratified model analysis
50Bahijri et al., 2020 [66]Saudi Arabia0.76 (95% CI, 0.73–0.79)0.70.7 Alternative cut-off thresholds evaluated
51Lowe et al., 2020 [67]Guyana0.81288.243.738.190.4Accuracy reported for defined threshold
52Buccheri et al., 2021 [68]USA 6573
53Abbas et al., 2021 [69]Qatar80% (95% CI, 0.78–0.83)86.257.949.589.75
54Cho et al., 2021 [70]South Korea0.760 (95% CI, 0.752–0.767)82.758.2
55Henjum et al., 2022 [71]Algeria0.8289652897
56Sadek et al., 2022 [72]Qatar0.870 (95% CI, 0.843–0.896)84.676.228.797.7
57Dong et al., 2022 [73]China0.812 and 0.822 Logistic regression vs. ML model comparison
58Alkattan et al., 2024 [74]Saudi Arabia0.803 (95% CI, 0.779–0.826)
AUC: area under the curve; CI: confidence interval; EPIC: European Prospective Investigation into Cancer and Nutrition; TUF: Tübingen Family Study; MeSyBePo: Metabolic Syndrome Berlin Potsdam; AROC: area under the receiver operating characteristic; ML: machine learning; Blank cells indicate information not reported in the original source study and should not be interpreted as zero or not applicable.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Christakis, C.; Saliari, D.; Zampelas, A.; Androutsos, O. Screening Tools for Early Identification of Adults at High Risk of Type 2 Diabetes: A Scoping Review. Healthcare 2026, 14, 839. https://doi.org/10.3390/healthcare14070839

AMA Style

Christakis C, Saliari D, Zampelas A, Androutsos O. Screening Tools for Early Identification of Adults at High Risk of Type 2 Diabetes: A Scoping Review. Healthcare. 2026; 14(7):839. https://doi.org/10.3390/healthcare14070839

Chicago/Turabian Style

Christakis, Christos, Dimitra Saliari, Antonis Zampelas, and Odysseas Androutsos. 2026. "Screening Tools for Early Identification of Adults at High Risk of Type 2 Diabetes: A Scoping Review" Healthcare 14, no. 7: 839. https://doi.org/10.3390/healthcare14070839

APA Style

Christakis, C., Saliari, D., Zampelas, A., & Androutsos, O. (2026). Screening Tools for Early Identification of Adults at High Risk of Type 2 Diabetes: A Scoping Review. Healthcare, 14(7), 839. https://doi.org/10.3390/healthcare14070839

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Article metric data becomes available approximately 24 hours after publication online.
Back to TopTop