Differences in screen-detected type 2 diabetes in two targeted community-based screening campaigns in Denmark Authors: Nuri Cayuelas Mateu, Peter Rossing, Kirsten Piepgras Neergaard, Tanja Thybo Published: 2026 ABSTRACT AbstractObjectiveTargeted type 2 diabetes screening campaigns aim to improve patients’ quality of life and life expectancy and reduce healthcare costs, although different setups may yield varying detection outcomes. The aim of this study is to assess whether at-home self-sampling improves the detection rates of prediabetes and diabetes compared with standard mobile unit screening.Research design and methodsBetween June 2022 and November 2024, the Danish Diabetes Association offered free capillary HbA1c tests to individuals at high risk of diabetes based on the Leicester’s Risk Assessment score. Two community-based screening setups were applied: (a) a mobile unit using a point-of-care device and (b) at-home self-sampling screening tests using mailed Hemoglobin Capillary Collection System analyzed by high-performance liquid chromatography.ResultsIn total, 4051 individuals were tested via the mobile unit and 5609 via the at-home setup. In the mobile unit setup, the proportion of individuals aged ≥65 years was 54.3%, whereas in the at-home setup, they constituted 38.1%. HbA1c ≥48 mmol/mol (≥6.5%) was found in 3.1% of participants in the mobile unit and 3.5% in the at-home setup, while HbA1c values of 42–47 mmol/mol (6.0–6.4%) occurred in 9.1% and 15.2%, respectively.ConclusionsMobile units and at-home self-sampling are complementary screening approaches, finding similar rates of screen detection with HbA1c ≥48 mmol/mol (≥6.5%), though the latter is more efficient in detecting individuals with HbA1c values of 42–47 mmol/mol (6.0–6.4%). WHAT IS ALREADY KNOWN ON THIS TOPIC Early detection of type 2 diabetes substantially reduces complications, increases life expectancy, and lowers healthcare costs, supporting systematic screening recommendations. Emerging self-sampling technologies enable at-home blood collection outside conventional facilities, potentially overcoming known barriers to participation. Early detection of type 2 diabetes substantially reduces complications, increases life expectancy, and lowers healthcare costs, supporting systematic screening recommendations. Emerging self-sampling technologies enable at-home blood collection outside conventional facilities, potentially overcoming known barriers to participation. WHAT THIS STUDY ADDS Targeted mobile unit and at-home self-sampling type 2 diabetes screening campaigns are feasible, effective, and complementary. Mobile units and at-home self-sampling are complementary screening approaches, finding similar rates of screen detection with HbA1c ≥48 mmol/mol (≥6.5%), though the latter is more efficient in detecting individuals with HbA1c values of 42–47 mmol/mol (6.0%–6.4%). Targeted mobile unit and at-home self-sampling type 2 diabetes screening campaigns are feasible, effective, and complementary. Mobile units and at-home self-sampling are complementary screening approaches, finding similar rates of screen detection with HbA1c ≥48 mmol/mol (≥6.5%), though the latter is more efficient in detecting individuals with HbA1c values of 42–47 mmol/mol (6.0%–6.4%). HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE, OR POLICY By implementing a diverse array of screening strategies tailored to different population segments, healthcare providers can identify a broader spectrum of individuals with type 2 diabetes at earlier stages. Hence, reducing the share of diabetes complications and excess mortality and consequently lowering the direct and indirect economic burden of type 2 diabetes. By implementing a diverse array of screening strategies tailored to different population segments, healthcare providers can identify a broader spectrum of individuals with type 2 diabetes at earlier stages. Hence, reducing the share of diabetes complications and excess mortality and consequently lowering the direct and indirect economic burden of type 2 diabetes. Type 2 diabetes (T2D) often develops insidiously and may remain undiagnosed for years until complications emerge.1 In Denmark, one-third of individuals with newly diagnosed T2D already present with diabetes-related complications at the time of diagnosis.2 3 Across Europe, approximately 36% of people living with T2D are undiagnosed.4 In Denmark alone, about 342 000 individuals are diagnosed with T2D,5 while it is estimated that around 100 000 live with undetected T2D and about 480 000 with prediabetes (HbA1c 42–47 mmol/mol, 6.0-6.4%).3 The risk of complications increases substantially with longer duration of undetected diabetes. Evidence shows that early detection and initiation of optimal management can reduce microvascular complications by 37% with only a 1%-point reduction in HbA1c6 and lower cardiovascular disease incidence by 38% over 5 years.7 Early detection also increases life expectancy by up to eight years8 9 and reduces healthcare costs. These benefits underpin recommendations for systematic screening to diagnose diabetes before complications occur.9 10 Screening strategies are generally categorized into three main approaches: universal or population-based outreach, opportunistic screening of individuals attending healthcare services for other reasons and targeted screening either within healthcare systems or in community-based settings.11 Universal screening has shown limited effectiveness in reducing cardiovascular morbidity or mortality,9 10 12 partly due to methodological challenges such as the need for very large cohorts and comprehensive longitudinal health records.13 In contrast, opportunistic and targeted screening approaches have demonstrated effectiveness in reducing morbidity and mortality, are cost-effective, and improve quality-adjusted life years.9 13–15 Accordingly, these strategies are endorsed by the World Health Organization and the National Institute for Health and Care Excellence.16 17 In Denmark, opportunistic screening is increasingly common, with 59% of 70-year-olds having at least one HbA1c measurement in 2018 and about 70% of 84-year-olds.18 In addition, multiple targeted initiatives for early detection are established both within the healthcare sector19 and in community-based settings.20 Community programs have often been organized through pharmacies21 or in socioeconomically disadvantaged urban areas in collaboration with local organizations.20 Community-based screening programs, however, have demonstrated a consistent age and gender gap in participation and adherence. Men, particularly younger men, are less likely to engage,22 despite yielding higher detection rates of prediabetes and diabetes when screened compared to women.23 Emerging technologies, including novel blood collection and storage methods, along with increasing digitalization, permit blood sampling to take place outside of conventional healthcare facilities. Self-sampling at home has recently shown to be feasible and cost-effective in a pilot study in Denmark.24 Such methods offer a potential means of overcoming barriers such as transportation, work obligations, and mobility limitations and may increase participation across demographic and socioeconomic groups. The aim of this study is to compare screen detection rates, as well as the sex and age distribution of the screen-detected across community-based at-home self-sampling and a standard community-based mobile unit screening approach. Research design and methods Between June 2022 and November 2024, the Danish Diabetes Association offered free capillary HbA1c tests to individuals at high risk of developing diabetes through two screening setups: (a) a mobile screening unit conducted in 22 major Danish cities with a total of 58 stops, using Afinion 2 point-of-care (POCT) analyzer to screen for T2D on-site and (b) at-home self-sampling screening using mailed Hemoglobin Capillary Collection System (HCCS) using Tosoh, a high-performance liquid chromatography (HPLC) for analyzing the blood samples. Both screening setups applied a two-stage screening strategy. Initially, all participants completed the Leicester Risk Assessment (LRA) questionnaire, and individuals identified as high-risk (score ≥16) were subsequently invited to undergo HbA1c testing. LRA is a validated tool recommended by NICE, used for the identification of those at risk of T2D.25 26 The survey consists of seven questions on age, sex, body mass index (BMI), waist size, ethnicity, family history of diabetes, and anti-hypertensive use. LRA has shown to perform better than the Finnish Diabetes Risk Score and the Danish Diabetes Risk Score in the prediction of unknown T2D and prediabetes in Denmark.27 Besides being used by, among others, UK Diabetes, LRA is also used by the Danish Diabetes Association as a tool for the identification of individuals at risk of T2D. To prioritize participant anonymity and data privacy, no baseline demographic and health characteristics obtained through the LRA screening were retained for any individuals, regardless of their risk score or subsequent participation. While this approach was intended to lower the threshold for engagement, it precludes any formal assessment of participation bias, as no comparative data exist for the non-responding group. Mobile screening unit The mobile screening unit, designed as a bus with an outdoor activity area and an informal, welcoming layout, toured 22 large Danish cities across 58 stops. The locations and dates were selected to coincide with events attracting high footfall, such as health fairs and farmers’ markets. The unit was staffed by a project manager, a dietitian, and one or two diabetes nurses. In addition to providing general information on T2D, the unit offered free capillary HbA1c testing to individuals identified as high-risk. HbA1c was measured using a handheld POC analyzer, which provided results within three minutes. This enabled on-site discussion of results between participants and healthcare professionals. POC HbA1c testing in general, and particularly the Afinion 2 device that was used in the mobile unit, has demonstrated good comparability performance to standard laboratory methods. Moreover, Afinion 2 demonstrates only a small negative bias (-0.7 to -1.3 mmol/mol (-0.06 to -0.12%)) overall and an even smaller bias at mid-range (45-49 mmol/mol (6.3-6.6%)) in comparison with Tosoh G8/SRMPs and meets stringent IFCC/NGSP criteria.28 29 All participants received both verbal and written information about the screening initiative and its purpose prior to testing and provided written informed consent. Blood samples were collected solely for the purpose of assessing HbA1c and documenting participation and detection rates. Samples were destroyed immediately after producing test results. Participants with an HbA1c level of ≥42 mmol/mol (≥6.0%) were offered immediate medical advice from the diabetes nurse and were advised to consult their general practitioner (GP) for confirmation and, if needed, initiation of treatment. At-home self-sampling screening In parallel with the mobile screening unit, a social media awareness campaign was conducted for about 4 months, highlighting the risks of undiagnosed T2D and promoting the Danish Diabetes Association’s online risk test (LRA). Individuals scoring as high risk (score ≥16) were offered a free at-home capillary blood sampling kit mailed directly to their residence. Each kit included an HCCS for sample collection, along with two single-use 1.5 mm lancets. The HCCS integrates collection, preparation, and transport and ensures sample stability for at least 13 days at room temperature. Participants also received written instructions on how to collect and return their sample using a prepaid envelope, as well as information on how to access their results. HbA1c analysis was performed at the Steno Diabetes Center Copenhagen laboratory using a Tosoh, an HPLC analyzer, which is considered the gold standard due to its high precision and reliability.30 Prior to study initiation, the HCCS was selected among three candidate systems based on laboratory evaluation demonstrating accuracy and durability comparable to venous sampling. Test results were recorded in the Danish laboratory database (LABKA) and made accessible to both participants and their GP through the national electronic health record system within one to two weeks. Participants with an HbA1c ≥48 mmol/mol (6.5%) were contacted by a physician prior to being granted access to their results. To further evaluate post-screening follow-up of the at-home self-sampling setup, individuals who completed the LRA test on the Danish Diabetes Association’s website during the campaign period were surveyed regarding whether they had subsequently consulted their GP for diagnostic confirmation and possible treatment initiation. Ethics All participants provided written informed consent prior to HbA1c testing. The consent form included information regarding study purpose, data collection, storage, and analysis, as well as the requirement to provide contact details, date, signature, and in the at-home self-sampling screening setup, and also the participants' personal identification number. Participants were informed that they could withdraw consent and discontinue participation at any time. All data were anonymized and processed in accordance with Danish legislation and the General Data Protection Regulation (EU 2016/679). The Danish National Committee on Health Research Ethics evaluated the study and determined that, while it constituted a health science research project according to Section 21 of the Committee Act, the interventions were considered so minimal that protocol submission and full committee approval were not required (Komitéloven §14, section 2, journal numbers H-21000371 (03-02-2021) and H-22021655 (20-05-2022)). The informed consent letter also provided participants with information on the ethical aspects of diabetes screening. While a capillary HbA1c test carries no physical risk, being screened for an asymptomatic disease may cause stress or anxiety. Participants were therefore advised to consider whether they felt prepared to receive test results before consenting to participate. Participants found to have an HbA1c ≥48 mmol/mol (6.5%) received in-person counseling from a nurse in the mobile screening unit or, in the case of at-home self-sampling, were contacted directly by a project physician before being granted access to test results. This ensured that individuals with values in the diagnostic range for diabetes were given appropriate information on the meaning of the result and guidance on referral to their GP for confirmation and potential treatment initiation. Statistical analysis Descriptive statistical analyses were conducted to examine participation and detection rates as well as HbA1c results by sex and age groups. The detection rates in the two screening setups were compared with estimates of the prevalence of T2D and prediabetes in Denmark using one-sample proportion test. The national reference estimates were derived from prior evidence of undiagnosed diabetes and prediabetes prevalence in Denmark in 20113 and adjusted for the trend in T2D prevalence during the intervening period.5 Further, the sex and age distributions among the participants, screening-detected, and individuals diagnosed with T2D through the Danish Healthcare System (DHS) in 2023 and 2024 were compared using χ2 test to evaluate the profile of the screen-detected. In cases where expected cell counts were less than five, p values were determined using chi-square test with Monte Carlo simulation based on 2000 replicates to ensure robust results. Additionally, the sex distribution of the screen-detected was compared against the national reference estimates for undiagnosed diabetes and prediabetes using one-sample proportion test. The sex and age distribution of those diagnosed through DHS were derived from the digital dashboard Diabetestal.nu,5 while the estimated sex distribution of individuals with undiagnosed diabetes and prediabetes was derived from prior evidence in Denmark in 2011.3 Finally, a probabilistic sensitivity analysis evaluated potential assay bias between the Afinion 2 and Tosoh methods, which could affect measurements near diagnostic thresholds and influence detection rates. Assay bias was simulated using a normal distribution (mean -1.0 mmol/mol, 95% CI -1.3 to -0.7 mmol/mol). Integer rounding errors were modeled assuming a uniform distribution of true values within rounded intervals. Following probabilistic reclassification, the adjusted mobile unit prediabetes prevalence was compared against the at-home arm using a χ2 test. All tests were two-sided, and a p value <0.05 was considered statistically significant. Results Between June 2022 and November 2024, a total of 9660 individuals at high risk based on LRA had their HbA1c tested. In the mobile unit setting, a total of 4051 individuals were identified as being at high risk based on the LRA test, and all consented to HbA1c testing. In contrast, in the at-home self-sampling setting, 10 583 individuals were identified as being at high risk based on the LRA test and requested a self-sampling kit; approximately 53% returned an adequate blood sample, resulting in a total of 5609 participants. The sex distribution of the participants was similar across the two screening modalities. Men accounted for 50.1% of participants in the at-home setup and 49.1% in the mobile unit, with no statistically significant difference between setups (p=0.34). This indicates a balanced sex distribution in both screening approaches. The age distribution differed significantly between the two modalities (p<0.001). In the mobile unit, the proportion of individuals aged ≥65 years was 54.3%, whereas in the at-home setup, they constituted 38.1%. In contrast, the at-home setup included a higher proportion of participants aged 40–64 years, accounting for 54.9%, whereas the corresponding proportion in the mobile unit was 39.1%. The proportion of younger participants aged 18–39 years was similar across the two setups. These findings indicate distinct age profiles between the screening approaches, with the mobile unit reaching an older population and the at-home setup predominantly engaging middle-aged individuals. Screen-detected with T2D The proportion of individuals with HbA1c values ≥48 mmol/mol (≥6.5%) was 3.5% in the at-home self-sampling setup and 3.1% in the mobile unit. The difference between the two screening modalities was not statistically significant (p=0.26), indicating comparable detection yields across the two approaches. However, the detection rates in both setups statistically and substantially exceeded the estimated national prevalence of undiagnosed T2D of 2.3% (p<0.001 and p=0.001, respectively). These findings were further supported by the probabilistic sensitivity analysis. After correcting for potential assay bias and rounding errors in the mobile unit setup, the expected diabetes prevalence in the mobile unit arm shifted slightly to 3.79%, while the difference between the two modalities remained non-significant (p=0.83). The differences between the detection rate in the two setups and the estimated prevalence of undiagnosed T2D in the general population are expected and likely reflect the use of a risk-based screening strategy. Participants in the present study were selected based on an elevated risk according to the LRA, resulting in a population enriched for established risk factors such as older age and higher BMI and a family history of diabetes. In contrast, population-based estimates include a substantial proportion of low-risk individuals, which lowers the overall prevalence. Thus, the higher prevalence observed in the two screening setups should be interpreted in the context of targeted screening. Among individuals screen-detected with HbA1c values ≥48 mmol/mol (≥6.5%), the at-home setup yielded more men (66%) than the DHS population (57%; p=0.018), aligning with national reference estimates, where men comprise 64% of undiagnosed diabetes in Denmark (p=0.61). Conversely, the mobile unit’s male proportion (60%) did not differ significantly from the DHS (p=0.66), or national reference estimate for undiagnosed diabetes (p=0.32). Thus, the higher male representation across both alternative approaches—significantly so in the at-home setup—reflects the higher underlying national prevalence of undiagnosed diabetes among men. By age, individuals under 40 years were underrepresented in both screening setups compared with the DHS data, with young women being entirely absent from the mobile unit. In contrast, the at-home setup preferentially reached middle-aged individuals (40–64 years), while older individuals (65–79 years) were markedly overrepresented in the mobile unit for both sexes (figures 1 and 2). These age-specific distribution deviations were statistically significant for both men (p=0.04) and women (p=0.012), demonstrating that the two screening approaches are highly complementary as they capture undetected T2D cases in completely different segments of the population. The age distribution among males from the study population detected with HbA1c values ≥48 mmol/mol (≥6.5%) and the males detected with type 2 diabetes through the Danish healthcare system (DHS). Source: Diabetestal.nu 2023 and 2024 and own calculations. The age distribution among females from the study population detected with HbA1c ≥48 mmol/mol (≥6.5%) and the females detected with type 2 diabetes through the Danish Healthcare System (DHS). Source: Diabetestal.nu 2023 and 2024 and own calculations. Screen-detected with prediabetes Values of HbA1c 42–47 mmol/mol (6.0-6.4%) were identified in 15.2% in the at-home setup and 9.1% in the mobile unit setup (figure 3). The prediabetes prevalence in the at-home self-sampling setup was statistically higher than both the mobile unit prevalence and the updated national estimates of prediabetes in the population of 11.3% (p<0.001 and p<0.001, respectively). Conversely, the mobile unit prevalence was significantly lower than the national estimate (p<0.001). These findings show differential participation patterns between screening modalities (figure 3). The share of adults detected with HbA1c ≥48 mmol/mol (≥6.5%) and type 2 HbA1c values of 42–47 mmol/mol (6–6.4%) and estimates of the share of adults living with unknown type 2 diabetes and pre-diabetes in Denmark. Source: *Jørgensen ME, Ellervik C, Ekholm O et al (3) and own calculations. These results were further confirmed in the probabilistic sensitivity analysis. After adjusting for assay bias and rounding errors, the expected prediabetes prevalence in the mobile unit arm shifted to 12.9%. However, the difference between the two screening arms remained highly statistically significant (p<0.002), confirming that measurement uncertainty alone cannot account for the higher detection rate in the at-home setup. Among those with HbA1c values of 42–47 mmol/mol (6.0%–6.4%), 50% of at-home participants and 56% of mobile unit participants were men. This difference in sex distribution was not statistically significant (p=0.10), indicating only minor variation between the two screening modalities. However, while the male proportion in the at-home arm aligned perfectly with the estimated male proportion of individuals with prediabetes in the general population (p=0.995), the male proportion in the mobile unit arm was significantly overrepresented compared with the national estimate (p=0.023). In contrast, the age composition of individuals with HbA1c 42–47 mmol/mol (6.0%–6.4%) differed significantly between the two setups (p=0.002). Individuals aged ≥65 years constituted a larger proportion of the prediabetes cases in the mobile unit, whereas those aged 40–64 years were more prevalent in the at-home setup. Notably, women aged 18–39 constituted a larger proportion of cases in the at-home group compared with the mobile unit. Discussion Prior evidence demonstrates that community-based two-stage T2D screening can reach population groups that may not otherwise engage with healthcare-based opportunistic screening.31 Consistent with previous findings,20–22 the two screening modalities tested here—mobile unit screening and at-home self-sampling—reached partly different population groups and can therefore be considered complementary rather than competing strategies. The mobile unit, staffed by healthcare professionals and using POC HbA1c devices, engaged older adults and ensured that results were discussed face-to-face with trained staff. This setting maximized test validity and provided immediate health advice, in line with evidence from similar interventions.32 In contrast, the at-home self-sampling approach proved particularly effective in reaching middle-aged individuals (40–64 years). Moreover, it outperformed the mobile unit in detecting individuals with pre-diabetes (15.2% vs 9.1%), a difference that remained highly significant even after controlling for potential assay bias and rounding errors. This higher detection yield is partially driven by a recruitment-related self-selection. The digital at-home setup was hosted on the official website of the Danish Diabetes Association, naturally attracting individuals who were already health-conscious or suspected they were at high risk. Conversely, since individuals with diagnosed T2D in Denmark receive regular blood glucose monitoring free of charge in primary care, they had little incentive to use the online self-test. This highlights the core strength of digital screening methods as they offer an anonymous, low-barrier, and highly accessible screening for younger, disproportionately high-risk individuals to test themselves in the privacy of their own homes, overriding the expected age-dependent background prevalence. Importantly, to our knowledge, this is the first study to evaluate an at-home HbA1c self-sampling kit for targeted diabetes screening, highlighting its potential for scalability and flexibility. Despite these strengths, neither setup was efficient in detecting individuals under 40 years of age, particularly women, compared with national survey data. This under-representation can only partly be explained by more effective detection of younger women through existing healthcare pathways. For example, a recent Danish pilot program reported that only 44% of women attended the recommended oral glucose tolerance test at 3 months postpartum and just 49% at 12 months, highlighting limited follow-up after gestational diabetes.33 A more likely explanation is the design of the LRA, which favors older age groups and men, while excluding risk factors particularly pertinent to younger women. Specifically, neither prior gestational diabetes (GDM) nor polycystic ovary syndrome (PCOS) are included. In Denmark, approximately 6% of pregnant women develop GDM, of whom 11% progress to T2D within 7 years.34 Likewise, PCOS affects about 17% of women and is associated with a fourfold increased risk of T2D, most often diagnosed before the age of 40.35 36 These women, while at elevated risk, may not be identified by the LRA if other risk factors are absent, underscoring the need to refine existing risk assessment tools for younger populations. This limitation is particularly concerning since early-onset T2D (<40 years) is associated with more aggressive disease progression, including greater insulin resistance, faster development of microvascular and macrovascular complications, and higher risks of adverse outcomes.37 Raising awareness and developing screening strategies suited to younger adults with specific risk factors is therefore critical.38 The overrepresentation of men with pre-diabetes in the mobile unit, contrasted with the high proportion of men with manifest, undiagnosed T2D in the at-home setup, likely reflects distinct health-seeking behaviors and self-selection between modalities. Because pre-diabetes is largely asymptomatic and men use primary healthcare less frequently than women, early metabolic abnormalities often go undetected. The mobile unit provided an opportunistic, walk-in public service requiring minimal individual initiative. This low-threshold approach effectively engaged men unlikely to participate in preventive screening, yielding higher rates of unrecognized pre-diabetes. Conversely, the at-home self-sampling program required active online registration and procedure completion. This format may preferentially attract individuals with a higher perceived diabetes risk, successfully capturing men who have already progressed to manifest T2D. Several limitations should be noted. First, although participants were informed that screening was intended for individuals without known diabetes or pre-diabetes, we could not independently verify their prior status; some may have used the campaigns as a way to confirm known results. This is, however, a common limitation for community-based screening.39 Second, two different methods were used in the two setups, that is, Afinion 2 assay in the mobile unit setup and Tosoh in the at-home setup. Although some POC HbA1c systems, including Afinion 2, have shown good agreement with laboratory methods in independent evaluations, performance varies across devices and small method-dependent biases remain possible.40 Third, a general challenge of community and home-based screening is ensuring continuity of care. Our setup did not allow us to determine whether participants contacted their GP after receiving a result indicating pre-diabetes or diabetes. The return rate of the at-home test kit was approximately 53%. While non-return may have influenced the observed detection rate, our primary objective was to evaluate a flexible screening approach rather than estimate the prevalence of undiagnosed T2D. Importantly, some individuals who did not return the kit reported practical barriers to the self-test procedure and instead chose assessment by their GP, while others underwent opportunistic testing in general practice before returning the kit. More generally, non-response in screening programs may reflect an interplay of sociodemographic barriers, including lower health literacy and practical constraints, as well as avoidant behavioral patterns such as health anxiety or reluctance to confront a potential diagnosis. Nevertheless, as outcome data were unavailable for non-responders, the potential impact on the observed detection rate remains uncertain. In contrast, the 3.5 and 15.2% detection rate reflects the true clinical yield of implementing this modality in practice. Conclusion This study provides novel evidence on the feasibility and diagnostic yield of a combined digital and in-person community-based screening setup targeted at T2D screening using at-home HbA1c self-sampling and a mobile unit. These approaches engaged different population groups and together broadened the reach of community-based detection efforts. HbA1c testing with self-collected blood at home proved technically reliable, user-centered, and suitable for integration into digital health infrastructures. Similarly, mobile units facilitated immediate counseling in accessible community locations. Both approaches can complement opportunistic and healthcare-based screening, extending detection to individuals who may not regularly seek healthcare services. Importantly, screening strategies should be regarded as complementary rather than competing setups. By implementing multiple approaches tailored to specific population subgroups, healthcare systems can improve early detection, reduce diabetes-related complications and mortality, and ultimately lower the societal and economic burden of T2D. A key consideration beyond the scope of this study is the comparative cost-effectiveness of mobile clinic–based screening versus at-home self-sampling. As mobile units require substantial logistical and financial resources, it remains unclear whether these investments translate into improved health outcomes compared with lower-cost, scalable at-home approaches. Future research should therefore evaluate the cost-effectiveness and long-term impact of these screening strategies, both individually and in combination, to better inform their optimal role in population-level diabetes prevention. Further, future research should focus on refining risk assessment tools to better identify young adults at risk—particularly women with GDM or PCOS—and on ensuring robust pathways for follow-up care after screening.