Advertisement

IDF diabetes Atlas: Global estimates of undiagnosed diabetes in adults for 2021

Published:December 05, 2021DOI:https://doi.org/10.1016/j.diabres.2021.109118

      Abstract

      Aims

      To provide up-to-date estimates of undiagnosed diabetes mellitus (UDM) prevalence - both globally, and by region/country, for the year 2021.

      Methods

      Data sources reporting diabetes prevalence were identified through a systematic search in the peer-reviewed and grey literature. The prevalence of undiagnosed diabetes was estimated from the data from each country where data was available. For countries without in-country data, the prevalence of undiagnosed diabetes was approximated by extrapolating the average of the estimates from countries with data sources within the same International Diabetes Federation (IDF) region and World Bank income grouping. We then applied these stratified prevalence estimates of UDM from each country to the number of adults in each strata and summed the counts to generate the number of adults with UDM (aged 20–79 years) for 215 countries and territories.

      Results

      In 2021, almost one in two adults (20–79 years old) with diabetes were unaware of their diabetes status (44.7%; 239.7 million). The highest proportions of undiagnosed diabetes (53.6%) were found in the Africa, Western Pacific (52.8%) and South-East Asia regions (51.3%), respectively. The lowest proportion of undiagnosed diabetes was observed in North America and the Caribbean (24.2%).

      Conclusions

      Diabetes surveillance needs to be strengthened to reduce the prevalence of UDM, particularly in low- and middle-income countries.

      Abbreviation:

      IDF (International Diabetes Federation), WHO (World Health Organization), T2DM (type 2 diabetes mellitus), T1DM (type 1 diabetes mellitus), UDM (undiagnosed diabetes mellitus), FBG (fasting plasma glucose), OGTT (oral glucose tolerance test), HbA1c (glycated haemoglobin), CI (confidence intervals), AFR (Africa), EUR (Europe), MENA (Middle East and North Africa), NAC (North America and the Caribbean), SACA (South and Central America), SEA (South-East Asia), WP (Western Pacific), LIC (low-income countries), MIC (middle-income countries), HIC (high-income countries)

      Keywords

      1. Introduction

      The prevalence of diabetes is increasing worldwide. The International Diabetes Federation (IDF) estimates that 536.6 million people are living with diabetes (diagnosed or undiagnosed) in 2021, and this number is projected to increase by 46%, reaching 783.2 million by 2045 [
      • Sun Hong
      • Saeedi Pouya
      • Karuranga Suvi
      • Pinkepank Moritz
      • et al.
      IDF diabetes Atlas: Global, regional and country-level diabetes prevalence estimates for 2021 and projections for 2045.
      ]. As previous IDF estimates and other studies have shown, approximately 50% of all individuals with diabetes are unaware of their condition [
      • Beagley J.
      • Guariguata L.
      • Weil C.
      • Motala A.A.
      Global estimates of undiagnosed diabetes in adults.
      ,

      Saeedi P, Petersohn I, Salpea P, Malanda B, Karuranga S, Unwin N, 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. https://doi.org/10.1016/j.diabres.2019.107843.

      ,
      • Cho N.H.
      • Shaw J.E.
      • Karuranga S.
      • Huang Y.
      • da Rocha Fernandes J.D.
      • Ohlrogge A.W.
      • et al.
      IDF Diabetes Atlas: Global estimates of diabetes prevalence for 2017 and projections for 2045.
      ]. From a clinical perspective, earlier identification during the asymptomatic stage is important to permit earlier initiation of treatment to prevent or delay the development of micro- and macrovascular complications.
      Studies have shown that a person may spend 5–6 years in an asymptomatic phase of pre-diabetes and type 2 diabetes mellitus (T2DM) before being diagnosed [
      • Whicher C.A.
      • O’Neill S.
      • Holt R.I.G.
      Diabetes in the UK: 2019.
      ], during which time micro- and macrovascular complications may develop [
      • Gedebjerg A.
      • Almdal T.P.
      • Berencsi K.
      • Rungby J.
      • Nielsen J.S.
      • Witte D.R.
      • et al.
      Prevalence of micro- and macrovascular diabetes complications at time of type 2 diabetes diagnosis and associated clinical characteristics: A cross-sectional baseline study of 6958 patients in the Danish DD2 cohort.
      ]. In 2013, it was estimated that 60% of people with T2DM were asymptomatic at the time of diagnosis [

      England’s Prescribing Data n.d. https://openprescribing.net/ (accessed October 6, 2021).

      ]. A large cross-sectional study in Denmark found that 35% of previously undiagnosed participants had complications at the time of diagnosis, of whom 12% had microvascular, 17% had macrovascular, and 6% showed both micro- and macrovascular complications [
      • Gedebjerg A.
      • Almdal T.P.
      • Berencsi K.
      • Rungby J.
      • Nielsen J.S.
      • Witte D.R.
      • et al.
      Prevalence of micro- and macrovascular diabetes complications at time of type 2 diabetes diagnosis and associated clinical characteristics: A cross-sectional baseline study of 6958 patients in the Danish DD2 cohort.
      ]. Of those with microvascular complications, retinopathy was the most frequent (13%), followed by neuropathy (4%) and nephropathy (3%). Ischemic heart disease was the most common macrovascular complication noted (15%), followed by atherosclerotic cerebrovascular (5%) and peripheral arterial disease (2%). However, since Denmark has a highly developed health system with a strong capacity for screening and diagnosis, these data may not be generalisable to countries lacking similar sophisticated health systems [
      • Ruta L.M.
      • Magliano D.J.
      • LeMesurier R.
      • Taylor H.R.
      • Zimmet P.Z.
      • Shaw J.E.
      Prevalence of diabetic retinopathy in Type 2 diabetes in developing and developed countries.
      ,
      • Heydari I.
      • Radi V.
      • Razmjou S.
      • Amiri A.
      Chronic complications of diabetes mellitus in newly diagnosed patients.
      ,
      • Katulanda P.
      • Ranasinghe P.
      • Jayawardena R.
      • Constantine G.R.
      • Sheriff M.H.R.
      • Matthews D.R.
      The prevalence, patterns and predictors of diabetic peripheral neuropathy in a developing country.
      ]. Indeed, without the same established medical infrastructure and resources necessary for early detection, middle- and low-income countries in other regions of the world may be characterised by higher numbers of people diagnosed with diabetes after the onset of complications. Maintaining good glycaemic control is the mainstay for the prevention of diabetes complications [
      Prospective UDS group
      Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33).
      ,
      • Nathan D.M.
      The Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications Study at 30 Years: Overview.
      ,
      • Turner R.C.
      • Millns H.
      • Neil H.A.W.
      • Stratton I.M.
      • Manley S.E.
      • Matthews D.R.
      • et al.
      Risk factors for coronary artery disease in non-insulin dependent diabetes mellitus: United Kingdom prospective diabetes study (UKPDS: 23).
      ]. Thus, better systems for screening which identify individuals with undiagnosed diabetes earlier in the disease process – thus allowing the opportunity for treatment – will reduce morbidity and mortality associated with this disease.
      Accurate measurement of the burden of undiagnosed diabetes (UDM) is also critical to monitor public health efforts related to diabetes screening and diagnosis. The IDF first produced estimates of UDM in 2011 [

      IDF Diabetes Atlas Group. IDF Diabetes Atlas, the Fifth Edition. Brussels, Belgium: International Diabetes Federation; 2011. https://doi.org/10.1007/978-90-481-3271-3.

      ], providing a global-scale quantification of this burden. Such estimates are important as they provide the necessary data for governments and health care systems to benchmark diabetes prevention activities.
      The objective of this study was to provide up-to-date estimates of UDM prevalence - both globally, and by region/country - for the year 2021.

      2. Methods

      The prevalence of UDM is estimated from population-based surveys using different objective measures and diagnostic criteria (fasting plasma glucose (FBG), oral glucose tolerance test (OGTT) and/or glycated haemoglobin (HbA1c)). Participants who report not knowing they have been diagnosed with diabetes may have diabetes upon testing and be classified as having UDM, or “previously undiagnosed”, or “newly diagnosed” diabetes. People with undiagnosed type 1 diabetes (T1DM) are not likely be detected in epidemiological cross-sectional surveys due to a different clinical presentation which is normally acute; hence, studies reporting on undiagnosed diabetes prevalence typically focus on T2DM only.
      The IDF methodology for estimating prevalence of diabetes and undiagnosed diabetes has previously been described [
      • Beagley J.
      • Guariguata L.
      • Weil C.
      • Motala A.A.
      Global estimates of undiagnosed diabetes in adults.
      ,
      • Guariguata L.
      • Whiting D.
      • Weil C.
      • Unwin N.
      The International Diabetes Federation diabetes atlas methodology for estimating global and national prevalence of diabetes in adults.
      ]. In brief, data sources reporting diabetes prevalence are identified from multiple sources including a systematic literature review, and a grey literature search comprising official government reports, the World Health Organization STEPwise and other non-peer reviewed sources such as government reports. The data sources were scored for quality and those meeting predefined criteria were selected as previously described [
      • Beagley J.
      • Guariguata L.
      • Weil C.
      • Motala A.A.
      Global estimates of undiagnosed diabetes in adults.
      ]. In the 10th edition of the IDF Diabetes Atlas, all studies reporting undiagnosed diabetes that met the selection criteria were included regardless of the method used to diagnose diabetes [
      • Cho N.H.
      • Shaw J.E.
      • Karuranga S.
      • Huang Y.
      • da Rocha Fernandes J.D.
      • Ohlrogge A.W.
      • et al.
      IDF Diabetes Atlas: Global estimates of diabetes prevalence for 2017 and projections for 2045.
      ]. The total proportion of UDM was extracted or calculated where possible from the numbers of newly diagnosed diabetes cases and the total number of diabetes cases. For countries with only one data source on undiagnosed diabetes, the extracted UDM proportions were used. If there were multiple estimates presented per country, the arithmetic mean of the estimates from each data source was calculated. However, in countries without in-country data sources, the undiagnosed proportion was obtained by the average of the estimates from countries with data sources within the same IDF and World Bank Income regions.
      The country-wide proportion of UDM, i.e. numbers of people with undiagnosed diabetes as a proportion of the total diabetes population (undiagnosed + diagnosed), were multiplied within age-, sex-, and setting-specific strata by the number of people with diabetes in each stratum to estimate the number of UDM. The prevalence of UDM (i.e. the percentage of the population with UDM) was calculated by dividing the age- and sex-specific number of people with UDM by the age- and sex-specific adult (20–79 years) population. We then applied these stratified prevalence estimates of UDM from each country to the number of adults in each strata and summed the counts to give the total number of adults with UDM (aged 20–79 years) for 215 countries and territories. Confidence intervals (CI) of the UDM prevalence estimates were derived from CIs of total diabetes prevalence by multiplying them by the proportion of UDM [

      Saeedi P, Petersohn I, Salpea P, Malanda B, Karuranga S, Unwin N, 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. https://doi.org/10.1016/j.diabres.2019.107843.

      ].

      3. Results

      Among 457 data sources reviewed, 111 met the quality threshold and were selected to estimate the prevalence of undiagnosed diabetes (Table S1, Supplementary Materials). The data sources include estimates from 68 of 215 (31.6%) countries and areas worldwide. For countries with either low quality or no in-country data on undiagnosed diabetes (147 countries, 68.4%), the prevalence of undiagnosed diabetes was approximated. The UDM proportions by data region as well as data sources can be found in Table 1.
      Table 1Study sources, characteristics by data region and proportion of undiagnosed diabetes (UDM) in adults (20–79 years), 2021.
      Data regionNo. of studiesCountries in data region providing dataCountries (no. of studies)Study sample representationStudy diagnosticStudy range UDM proportions (%) reportedData region UDM proportion (%)Data region UDM prevalence (in million)
      AFR65 of 48; 10%national(6)OGTT(3);

      FBG(2);

      HbA1c(1)
      36.0–86.753.612.7
      AFR-HIC11 of 1; 100%Seychelles(1)national(1)OGTT(1)46.046.00.003
      AFR-MIC43 of 23; 13%Comoros(1); Kenya(2); South Africa(1)national(4)OGTT(2); FBG(1); HbA1c(1)36.0 – 52.850.57.5
      AFR-LIC11 of 24; 4%Mozambique(1)national(1)FBG(1)86.758.85.2
      EUR2318 of 59; 31%national(19); regional(4)OGTT(11); FBG(9); HbA1c(2); Combined(1)*16.7 – 74.035.721.9
      EUR-HIC1613 of 40; 33%Cyprus(1); Estonia(1);

      Faroe Islands(1); Finland(1); France(2); Germany(1); Hungary(1); Malta(1); Poland(1); Portugal(1); Romania(2); Spain(1);

      United Kingdom(2)
      national(15); regional(1)OGTT(7); FBG(7); HbA1c(1); Combined(1)*16.7 – 65.231.111.3
      EUR-MIC75 of 18; 28%Bulgaria(1); Russian Federation(2); Turkey(2); Turkmenistan(1); Uzbekistan(1)national(4); regional(3)OGTT(4); FBG(2); HbA1c(1)25.8 – 74.042.410.5
      EUR-LIC00 of 1; 0%43.90.1
      MENA1812 of 21; 57%national(14); regional(1); local(3)OGTT(4); FBG(13); HbA1c(1)17.4 – 76.137.627.3
      MENA-HIC54 of 6; 67%Kuwait(2); Oman(1); Saudi Arabia(1); United Arab Emirates (1)national(5)FBG(4); HbA1c(1)33.1 – 64.045.53.2
      MENA-MIC95 of 11; 45%Egypt(1); Iraq(1); Iran (Islamic Republic of) (3); Jordan(2);Pakistan (2)national(7); regional(1); local(1)OGTT(3); FBG(6)17.4 – 62.036.221.2
      MENA-LIC43 of 4; 75%Afghanistan(2);

      Sudan(1); Yemen(1)
      national(2); local(2)OGTT(1); FBG(3)22.2 – 76.141.13.0
      NAC107 of 23; 30%national(9);

      regional(1)
      OGTT(4); FBG(4); HbA1c(2)12.5 – 49.024.112.2
      NAC-HIC22 of 14; 14%Canada(1); USA(1)national(2)OGTT(1); HbA1c(1)12.5 – 37.814.75.2
      NAC-MIC74 of 9; 44%Belize(2); Jamaica(2); Mexico(2); Suriname(1)national(7)OGTT(2); FBG(4); HbA1c(1)23.9 – 49.047.06.8
      NAC-LIC11 of 1; 100%Haiti(1)regional(1)OGTT(1)29.429.40.2
      SACA97 of 19; 37%National(1); regional(6); local(2)OGTT(4); FBG(4); HbA1c(1)20.0 – 53.032.910.7
      SACA-HIC00 of 5; 0%31.21.4
      SACA-MIC97 of 14; 50%Brazil(1); Colombia(2); Ecuador(1); Guatemala(1); Honduras(1); Nicaragua(2); Venezuela (Bolivarian Republic of) (1)national(1); regional(6); local(2)OGTT(4); FBG(4); HbA1c(1)20.0 – 53.033.19.3
      SEA43 of 7; 43%national(4)OGTT(4)25.3 – 53.151.246.2
      SEA-HIC21 of 1; 100%Mauritius(2)national(2)OGTT(2)25.3 – 46.635.70.1
      SEA-MIC22 of 6; 33%India(1); Sri Lanka(1)national(2)OGTT(2)35.8 – 53.151.346.1
      WP4116 of 37; 42%national(29);

      regional(10);

      local(2)
      OGTT(20); FBG(19); HbA1c(1); Combined(1)*25.0 – 75.852.9108.7
      WP-HIC96 of 14; 40%Australia(1); China, Hong Kong (1); Japan(1); Nauru(1); Republic of Korea (3); Singapore(2)national(9)OGTT(2); FBG(6); Combined(1)*25.0 – 54.242.88.7
      WP-MIC3210 of 22; 45%Cambodia(2); China(14); Indonesia(1); Kiribati(1); Lao People’s Democratic Republic(1); Malaysia(4); Marshall Islands(1); Myanmar(2); Philippines(1); Thailand(5)national(20); regional(10); local(2)OGTT(18); FBG(13); HbA1c(1)27.1 – 75.853.998.9
      WP-LIC00 of 1; 0%55.81.0
      HIC3527 of 81; 33%national(34), regional(1)OGTT(13); FBG(17); HbA1c(3); Combined (2)*12.5–65.228.829.9
      MIC7036 of 103; 35%national(45); regional(20); local(5)OGTT(13 + 22 = 35); FBG(5 + 25 = 30); HbA1c(1 + 4 = 5)17.4–75.848.4200.4
      LIC65 of 31; 16%national(3); regional(1); local(2)OGTT(2); FBG(4)22.2–86.750.59.5
      World11168 of 215; 32%national (82)

      regional (22)

      local (7)
      OGTT (50)

      FBG (51)

      HbA1c (8)

      Combined (2)*
      12.5–86.744.7239.7
      AFR – Africa, EUR – Europe, MENA – Middle East and North Africa, NAC – North America and the Caribbean, SACA – South and Central America, SEA – South-East Asia, WP – Western Pacific, OGTT – oral glucose tolerance test, FBG – fasting plasma glucose, HbA1c – glycated haemoglobin, LIC – low-income countries, MIC – middle-income countries, HIC – high-income countries.
      * - In two data sources from the UK

      Bluett. NHS Research Scotland (NRS) Diabetes Register 2021.

      and Australia

      Diabetes Australia. National Diabetes Service Scheme. 2021. https://www.ndss.com.au/.

      data were based on available diabetes registries in these countries, but the UDM proportions were obtained from previous high-quality publications on UDM.
      Almost all data was based on either OGTT or FPG (101, 91%), but less than half (50, 45%) of the 111 sources used an OGTT to diagnose UDM. Only eight studies were based on HbA1c. The majority (82, 74%) of all data sources were nationally representative.
      The low-income countries (LICs) had the lowest proportion of original source data (5 of 31; 16%) compared to high-income countries (HICs) (27 of 81; 33%) and middle-income countries (MICs) (36 of 103; 35%). By IDF region (Africa – AFR; Europe – EUR; Middle East and North Africa – MENA; North America and the Caribbean – NAC; South and Central America – SACA; South-East Asia – SEA; and Western Pacific – WP), MENA had the highest proportion of countries with in-country source data (57%). In all other regions, less than half of countries had primary data on UDM. Only 10% of countries in AFR had primary data on UDM.
      In 2021, globally, almost one in two (44.7%; 239.7 million) adults (20–79 years old) are unaware that they have diabetes. The global estimates compared with previous IDF Atlas versions are shown in Fig. 1. While the UDM proportion was stable around 45–50% over the past 10 years, the absolute number of people living with UDM grew steadilyfrom 182.6 million in 2011 to 239.7 million in 2021.
      Figure thumbnail gr1
      Fig. 1Global proportion of undiagnosed diabetes and number of undiagnosed diabetes cases in each IDF Diabetes Atlas Edition, 5th-10th, 2011–2021. M – million.
      In 2021, the highest proportion of UDM was found in the AFR (53.6%), WP (52.8%) and SEA Regions (51.3%), respectively (Fig. 2). The lowest proportion of UDM was found in NAC (24.2%). The highest proportion of UDM (58.8%) was found in low-income countries in AFR and the lowest (14.7%) in the high-income countries in NAC. Globally, 87.5% of all people with UDM lived in low- and middle-income countries. In high-income countries, 28.8% of people with diabetes were previously undiagnosed (Fig. 2).
      Figure thumbnail gr2
      Fig. 2Proportion of undiagnosed diabetes by IDF and World Bank regions in adults (20–79 years), 2021. AFR – Africa, EUR – Europe, MENA – Middle East and North Africa, NAC – North America and the Caribbean, SACA – South and Central America, SEA – South-East Asia, WP – Western Pacific, LIC – low-income countries, MIC – middle-income countries, HIC – high-income countries.
      The number of people with UDM and the proportion of UDM varies widely by country (Table S2, Supplementary Materials). In the IDF data, between 12.5% (the USA) and 86.7% (Mozambique) of people with diabetes were undiagnosed. The countries with the highest numbers of people with UDM are the same countries with the largest number of people with diabetes: China (72.8 million), India (39.4 million) and Indonesia (14.3 million). Among countries with source data on the proportion of UDM, Mozambique (86.7%, representing 0.3 million people), Uzbekistan (74.0%, 1.0 million), Indonesia (73.7% 14.3 million) and Afghanistan (71.4% 1.1 million) have the highest proportion of people with UDM.

      3. Discussion

      Globally we estimate that 239.7 million people are unaware of their diabetes, with great variation in the proportions of undiagnosed diabetes across regions and countries.
      There are no other studies to our knowledge estimating the global prevalence of UDM. The World Health Organization (WHO) [
      • World Health Organization
      ], and the Global Burden of Disease study [
      • Liu J.
      • Ren Z.-H.
      • Qiang H.
      • Wu J.
      • Shen M.
      • Zhang L.
      • et al.
      Trends in the incidence of diabetes mellitus: results from the Global Burden of Disease Study 2017 and implications for diabetes mellitus prevention.
      ] both present estimates on global prevalence of diabetes and for each country but no estimates of UDM are reported [
      • Liu J.
      • Ren Z.-H.
      • Qiang H.
      • Wu J.
      • Shen M.
      • Zhang L.
      • et al.
      Trends in the incidence of diabetes mellitus: results from the Global Burden of Disease Study 2017 and implications for diabetes mellitus prevention.
      ]. The IDF Diabetes Atlas estimates are the only source of global data on undiagnosed diabetes prevalence. The value of this data is that it provides a way to monitor the effectiveness of public health efforts related to diabetes screening and diagnosis.
      Earlier diagnosis coupled with earlier treatment and adequate glycaemic control can slow the development and progression of diabetes complications. Complication rates among persons with UDM and persons with newly diagnosed diabetes are higher in countries with developing economies than in countries with developed economies. For microvascular complications, a review on the trends of diabetes complications worldwide showed that, for example, the prevalence of diabetic retinopathy is greater in countries with developing economies compared to those with developed economies [
      • Dal Canto E.
      • Ceriello A.
      • Rydén L.
      • Ferrini M.
      • Hansen T.B.
      • Schnell O.
      • et al.
      Diabetes as a cardiovascular risk factor: An overview of global trends of macro and micro vascular complications.
      ]. In Sweden, coronary heart disease mortality due to diabetes has declined since 1998 [
      • Rawshani A.
      • Rawshani A.
      • Franzén S.
      • Eliasson B.
      • Svensson A.-M.
      • Miftaraj M.
      • et al.
      Mortality and Cardiovascular Disease in Type 1 and Type 2 Diabetes.
      ], whereas age-standardised death rates among people with diabetes in India, a middle-income country, increased in the years from 1990–2016 [
      • Tandon N.
      • Anjana R.M.
      • Mohan V.
      • Kaur T.
      • Afshin A.
      • Ong K.
      • et al.
      The increasing burden of diabetes and variations among the states of India: the Global Burden of Disease Study 1990–2016.
      ].
      Given the long asymptomatic phase of T2DM, reducing UDM proportions worldwide can only be achieved via screening programs from readily accessible and equitable healthcare systems. In this regard, targeted approaches - where those at highest risk of diabetes are identified and screened - are certainly feasible and more cost-effective [
      • Najafi B.
      • Farzadfar F.
      • Ghaderi H.
      • Hadian M.
      Cost effectiveness of type 2 diabetes screening: A systematic review.
      ]. Several non-invasive or minimally invasive tools that assess the risk of having undiagnosed or future diabetes have been developed, adapted for use in diverse populations and recommended by the WHO [
      • World Health Organization
      ]: FINRISK [
      • Lindström J.
      • Eriksson J.G.
      • Valle T.T.
      • Aunola S.
      • Cepaitis Z.
      • Hakumäki M.
      • et al.
      Prevention of Diabetes Mellitus in Subjects with Impaired Glucose Tolerance in the Finnish Diabetes Prevention Study: Results From a Randomized Clinical Trial.
      ], ColDRISC [
      • Barengo N.C.
      • Tamayo D.C.
      • Tono T.
      • Tuomilehto J.
      A Colombian diabetes risk score for detecting undiagnosed diabetes and impaired glucose regulation.
      ], AUSDRISK [
      • Chen L.
      • Magliano D.J.
      • Balkau B.
      • Colagiuri S.
      • Zimmet P.Z.
      • Tonkin A.M.
      • et al.
      AUSDRISK: an Australian Type 2 Diabetes Risk Assessment Tool based on demographic, lifestyle and simple anthropometric measures.
      ], and IDRS (Indian Diabetes Risk Score) [
      • Mohan V.
      • Sandeep S.
      • Deepa M.
      • Gokulakrishnan K.
      • Datta M.
      • Deepa R.
      A diabetes risk score helps identify metabolic syndrome and cardiovascular risk in Indians ? the Chennai Urban Rural Epidemiology Study (CURES-38).
      ]. Broader application and implementation of these tools in country-specific guidelines may help to improve screening rates and reduce the proportion of UDM. Nevertheless, large-scale screening programs are generally only acceptable to health care systems when cost effectiveness has been demonstrated and where the health systems can cope with the long-term monitoring and care necessary for managing diabetes. Data pertaining to the cost effectiveness of diabetes screening programs are mixed [
      • Najafi B.
      • Farzadfar F.
      • Ghaderi H.
      • Hadian M.
      Cost effectiveness of type 2 diabetes screening: A systematic review.
      ,
      • Roberts S.
      • Barry E.
      • Craig D.
      • Airoldi M.
      • Bevan G.
      • Greenhalgh T.
      Preventing type 2 diabetes: systematic review of studies of cost-effectiveness of lifestyle programmes and metformin, with and without screening, for pre-diabetes.
      ], and difficult to generalise given that a broad array of country-specific factors may determine their outcomes. Further complexity arises when considering that screening programs should not simply be aimed at reducing the UDM proportion per se; rather, their effectiveness should be based on capacity to achieve reductions in long-term morbidity/mortality. For example, a screening program targeted at older people may meaningfully reduce numbers of people with undiagnosed diabetes due to high diabetes prevalence at older ages; however, it would miss those with younger-onset T2DM, who may have the highest lifetime incidence of complications due to longer diabetes duration. Furthermore, in countries with limited resources, it would not be desirable for screening costs to ‘compete’ with treatment costs. Clearly, further work on how to best target individuals for screening (i.e., those individuals most at risk of future diabetes complications; not necessarily just diabetes alone) is required to equip health care systems with the tools necessary to develop and implement cost-effective screening programs. This will be especially important in the context of the current pandemic, given that substantive healthcare resources have been necessarily re-directed toward COVID-19 in the last two years. Indeed, whether this has already negatively impacted awareness/screening/treatment of non-communicable diseases such as diabetes will be of major interest in coming years.

      Limitations

      The principal limitation in generating accurate estimates for UDM is the lack of high-quality data suitable for inclusion in our estimates. While 457 of the 1,241 sets of data in the IDF Diabetes Atlas database contained data on UDM, only 111 fulfilled the necessary quality criteria to generate estimates for UDM. Furthermore, only 32% of countries and areas worldwide provided high quality in-country source data on UDM, and this proportion varied by income status: i.e. 33% and 35% of high and middle-income countries, respectively, compared with 16% of low-income countries. While the aggregated average proportion of UDM might be stable in countries with data, the estimates for countries without in-country source data sources are obviously less reliable. Unfortunately, there are no statistical methods that can overcome lack of data, so well-designed population-based cross-sectional studies are urgently needed, particularly in low-income countries. Moreover, the proportions of previously diagnosed and undiagnosed diabetes are not universally reported in cross-sectional surveys focused on measuring diabetes prevalence. Recently, guidelines for diabetes epidemiological studies, with specific recommendations for resource-limited settings, were published by the IDF and provide a set of recommendations for conducting high-quality surveys to estimate the burden of diabetes [
      • Aschner P.
      • Karuranga S.
      • James S.
      • Simmons D.
      • Basit A.
      • Shaw J.E.
      • et al.
      The International Diabetes Federation’s guide for diabetes epidemiological studies.
      ]
      The prevalence of UDM measured in a screening study is related to many factors and depends on the underlying diabetes prevalence, screening coverage, diagnostic methods, performance of a health care system and general awareness of diabetes in the public and among health care professionals. The prevalence of undiagnosed diabetes varies by age, gender, urban or rural setting, and ethnicity [
      • Coppell K.J.
      • Mann J.I.
      • Williams S.M.
      • Jo E.
      • Drury P.L.
      • Miller J.C.
      • et al.
      Prevalence of diagnosed and undiagnosed diabetes and prediabetes in New Zealand: Findings from the 2008/09 adult nutrition survey.
      ]. Thus, the relative proportion of diagnosed and undiagnosed diabetes can vary substantially across different subgroups. The majority of the data sources did not provide the prevalence of UDM by subgroup, limiting our ability to provide stratified estimates.
      With the data available to us, we were unable to distinguish whether UDM prevalence varie d by diagnostic method, and our global estimates are based on mixed diagnosis definitions. Nevertheless, the three main methods (OGTT, FBG, HbA1c) for assessing diabetes status are based on different biochemical and physiological processes and can yield different results [
      • Galaviz K.I.
      • Narayan K.M.V.
      • Lobelo F.
      • Weber M.B.
      Lifestyle and the Prevention of Type 2 Diabetes: A Status Report.
      ]. Moreover, they can detect different subgroups of people with diabetes [
      • Bahijri S.
      • Al-Raddadi R.
      • Ajabnoor G.
      • Jambi H.
      • Al Ahmadi J.
      • Borai A.
      • et al.
      Dysglycemia risk score in Saudi Arabia: A tool to identify people at high future risk of developing type 2 diabetes.
      ,

      Costa B, Barrio F, Cabré JJ, Piñol JL, Cos FX, Solé C, et al. Shifting from glucose diagnostic criteria to the new HbA 1c criteria would have a profound impact on prevalence of diabetes among a high-risk Spanish population. Diabet Med 2011;28:1234–7. https://doi.org/10.1111/j.1464-5491.2011.03304.x.

      ] and those groups do not always overlap – another reason why studies from one country utilizing different diagnostic methods may report different proportions of undiagnosed diabetes. Likewise, the definition of previously diagnosed diabetes may vary, with some studies relying on self-report only, and others confirming diagnoses using more reliable methods such as medical record review, blood tests, and/or medication prescription data review.
      Our approach to estimate undiagnosed diabetes on a global scale does not account for many of the complex determinants involved in undiagnosed diabetes. It draws directly from high-quality studies in the existing literature, but these are few and far between so that the estimates can provide only a broad understanding of the true burden of UDM. More sophisticated approaches of estimation based on evidence gathered from high-quality epidemiological surveys are needed built on socio-economics characteristics, health care system types and performance, rapidity of disease progression, health equity and other factors. Unfortunately, there are no such models in the literature. The estimates would benefit greatly from an approach that could integrate multiple determinants to provide an accurate estimate of the prevalence of UDM.
      While most population-based studies of diabetes generally undertake only one test for the purposes of surveillance, the clinical diagnosis of diabetes is generally based on two positive test results in the screening setting [
      • World Health Organization
      ,

      American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care 2010;33. https://doi.org/10.2337/dc10-S062.

      ]. It is important to highlight that the IDF Diabetes Atlas results on UDM are largely based on the single test screening strategy and may not reflect the prevalence of undiagnosed diabetes in clinical settings. These issues highlighted by the definitions of ‘undiagnosed diabetes’ warrant further examination as they are important both for quality improvement in clinical practice and for monitoring of the diabetes epidemic.

      4. Conclusion

      In 2021, the global prevalence of undiagnosed diabetes remains high. Almost half of all people with diabetes (44.7%; 239.7 million) were unaware that they have the condition. Moreover, the disparity between high income and low to middle income regions and the prevalence of undiagnosed diabetes is still pronounced. Further improvements in diabetes surveillance systems and in the implementation of tools and strategies to detect undiagnosed diabetes at a population level are needed, particularly in low- and middle-income countries.

      Funding

      The 10th edition of the IDF Diabetes Atlas was supported by the following sponsors: Pfizer-MSD Alliance, Novo Nordisk, Sanofi.

      Declaration of Competing Interest

      The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

      Acknowledgements

      IDF Diabetes Atlas, Tenth Edition Committee.
      Dianna J Magliano, Edward J Boyko, Beverley Balkau, Noël C Barengo, Elizabeth Barr, Abdul Basit, Christian Bommer, Gillian Booth, Bertrand Cariou, Juliana Chan, Hongzhi Chen, Lei Chen, Tawanda Chivese, Dana Dabalea, Hema Divakar, Daisy Duan, Bruce B Duncan, Michael Fang, Ghazal Fazli, Courtney Fischer, Kathryn Foti, Laercio Franco, Edward Gregg, Leonor Guariguata, Akhil Gupta, Anthony Hanley, Jessica L Harding, William H Herman, Cheri Hotu, Cecilia Høgfeldt, Elbert Huang, Adam Hulman, Steven James, Alicia J Jenkins, Seung Jin Han, Calvin Ke, Emma L Klatman, Shihchen Kuo, Jean Lawrence, Dinky Levitt, Xia Li, Lorraine Lipscombe, Paz Lopez-Doriga Ruiz, Andrea Luk, Ronald C Ma, Jayanthi Maniam, Louise Maple-Brown, Jean-Claude Mbanya, Natalie McGlynn, Fernando Mijares Diaz, Hiliary Monteith, Ayesha Motala, Estelle Nobecourt, Graham D Ogle, Katherine Ogurstova, Richard Oram, Bige Ozkan, Emily Papadimos, Chris Patterson, Meda Pavkov, Cate Pihoker, Justin Porter, Camille Powe, Ambady Ramachandran, Gojka Roglic, Mary Rooney, Julian Sacre, Elizabeth Selvin, Baiju Shah, Jonathan E Shaw, David Simmons, Caroline Stein, Jannet Svensson, Olive Tang, Justin Echouffo Tcheugui, Jincy Varghese, Amelia Wallace, Pandora L Wander, Donald Warne, Mahmoud Werfalli, Sarah Wild, Jencia Wong, Yuting Xie, Xilin Yang, Lili Yuen, Philip Zeitler, Ping Zhang, Sui Zhang, Xinge Zhang, Zhiguang Zhou.

      IDF executive office

      Mikkel Pape Dysted, Sanju Gautam, Bruno Helman, Suvi Karuranga, Lorenzo Piemonte, Moritz Pinkepank, Adilson Randi, Phil Riley, Pouya Saeedi, Agus Salim, Hong Sun, Beatriz Yáñez Jiménez, Katherine Wallis, Margaux Ysebaert.

      Contributors

      GLOBODIAB Research Consortium.

      Data

      The list of studies on which estimates in the IDF Diabetes Atlas are based and those considered but not used can be found at: www.diabetesatlas.org.

      Corporate sponsors

      The 10th edition of the IDF Diabetes Atlas was supported by an educational grant from the Pfizer-MSD Alliance, with the additional support of Sanofi and Novo Nordisk.

      Appendix A. Supplementary data

      The following are the Supplementary data to this article:

      References

        • Beagley J.
        • Guariguata L.
        • Weil C.
        • Motala A.A.
        Global estimates of undiagnosed diabetes in adults.
        Diabetes Res Clin Pract. 2014; 103: 150-160https://doi.org/10.1016/j.diabres.2013.11.001
      1. Saeedi P, Petersohn I, Salpea P, Malanda B, Karuranga S, Unwin N, 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. https://doi.org/10.1016/j.diabres.2019.107843.

        • Cho N.H.
        • Shaw J.E.
        • Karuranga S.
        • Huang Y.
        • da Rocha Fernandes J.D.
        • Ohlrogge A.W.
        • et al.
        IDF Diabetes Atlas: Global estimates of diabetes prevalence for 2017 and projections for 2045.
        Diabetes Res Clin Pract. 2018; 138: 271-281https://doi.org/10.1016/j.diabres.2018.02.023
        • Whicher C.A.
        • O’Neill S.
        • Holt R.I.G.
        Diabetes in the UK: 2019.
        Diabet Med. 2020; 37: 242-247https://doi.org/10.1111/dme.v37.210.1111/dme.14225
        • Gedebjerg A.
        • Almdal T.P.
        • Berencsi K.
        • Rungby J.
        • Nielsen J.S.
        • Witte D.R.
        • et al.
        Prevalence of micro- and macrovascular diabetes complications at time of type 2 diabetes diagnosis and associated clinical characteristics: A cross-sectional baseline study of 6958 patients in the Danish DD2 cohort.
        J Diabetes Complications. 2018; 32: 34-40https://doi.org/10.1016/j.jdiacomp.2017.09.010
      2. England’s Prescribing Data n.d. https://openprescribing.net/ (accessed October 6, 2021).

        • Ruta L.M.
        • Magliano D.J.
        • LeMesurier R.
        • Taylor H.R.
        • Zimmet P.Z.
        • Shaw J.E.
        Prevalence of diabetic retinopathy in Type 2 diabetes in developing and developed countries.
        Diabet Med. 2013; 30: 387-398https://doi.org/10.1111/dme.12119
        • Heydari I.
        • Radi V.
        • Razmjou S.
        • Amiri A.
        Chronic complications of diabetes mellitus in newly diagnosed patients.
        Int J Diabetes Mellit. 2010; 2: 61-63https://doi.org/10.1016/j.ijdm.2009.08.001
        • Katulanda P.
        • Ranasinghe P.
        • Jayawardena R.
        • Constantine G.R.
        • Sheriff M.H.R.
        • Matthews D.R.
        The prevalence, patterns and predictors of diabetic peripheral neuropathy in a developing country.
        Diabetol Metab Syndr. 2012.1–8.; 2012: 4https://doi.org/10.1186/1758-5996-4-21
        • Prospective UDS group
        Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33).
        Lancet. 1998; 352: 837-853https://doi.org/10.1016/S0140-6736(98)07019-6
        • Nathan D.M.
        The Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications Study at 30 Years: Overview.
        Diabetes Care. 2014; 37: 9-16https://doi.org/10.2337/dc13-2112
        • Turner R.C.
        • Millns H.
        • Neil H.A.W.
        • Stratton I.M.
        • Manley S.E.
        • Matthews D.R.
        • et al.
        Risk factors for coronary artery disease in non-insulin dependent diabetes mellitus: United Kingdom prospective diabetes study (UKPDS: 23).
        Br Med J. 1998; 316: 823-828
      3. IDF Diabetes Atlas Group. IDF Diabetes Atlas, the Fifth Edition. Brussels, Belgium: International Diabetes Federation; 2011. https://doi.org/10.1007/978-90-481-3271-3.

        • Guariguata L.
        • Whiting D.
        • Weil C.
        • Unwin N.
        The International Diabetes Federation diabetes atlas methodology for estimating global and national prevalence of diabetes in adults.
        Diabetes Res Clin Pract. 2011; 94: 322-332https://doi.org/10.1016/j.diabres.2011.10.040
        • World Health Organization
        Global Report on Diabetes. vol. 978. Switzerland, Geneva2016
        • Liu J.
        • Ren Z.-H.
        • Qiang H.
        • Wu J.
        • Shen M.
        • Zhang L.
        • et al.
        Trends in the incidence of diabetes mellitus: results from the Global Burden of Disease Study 2017 and implications for diabetes mellitus prevention.
        BMC Public Health. 2020; 20https://doi.org/10.1186/s12889-020-09502-x
        • Dal Canto E.
        • Ceriello A.
        • Rydén L.
        • Ferrini M.
        • Hansen T.B.
        • Schnell O.
        • et al.
        Diabetes as a cardiovascular risk factor: An overview of global trends of macro and micro vascular complications.
        Eur J Prev Cardiol. 2019; 26: 25-32https://doi.org/10.1177/2047487319878371
        • Rawshani A.
        • Rawshani A.
        • Franzén S.
        • Eliasson B.
        • Svensson A.-M.
        • Miftaraj M.
        • et al.
        Mortality and Cardiovascular Disease in Type 1 and Type 2 Diabetes.
        Http://DxDoiOrg/101056/NEJMoa1608664. 2017; 376: 1407-1418https://doi.org/10.1056/NEJMoa1608664
        • Tandon N.
        • Anjana R.M.
        • Mohan V.
        • Kaur T.
        • Afshin A.
        • Ong K.
        • et al.
        The increasing burden of diabetes and variations among the states of India: the Global Burden of Disease Study 1990–2016.
        Lancet Glob Heal. 2018; 6: e1352-e1362https://doi.org/10.1016/S2214-109X(18)30387-5
        • Najafi B.
        • Farzadfar F.
        • Ghaderi H.
        • Hadian M.
        Cost effectiveness of type 2 diabetes screening: A systematic review.
        Med J Islam Repub Iran. 2016; 30: 326
        • Lindström J.
        • Eriksson J.G.
        • Valle T.T.
        • Aunola S.
        • Cepaitis Z.
        • Hakumäki M.
        • et al.
        Prevention of Diabetes Mellitus in Subjects with Impaired Glucose Tolerance in the Finnish Diabetes Prevention Study: Results From a Randomized Clinical Trial.
        J Am Soc Nephrol. 2003; 14: S108-S113https://doi.org/10.1097/01.ASN.0000070157.96264.13
        • 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: 86-93https://doi.org/10.1016/j.pcd.2016.09.004
        • Chen L.
        • Magliano D.J.
        • Balkau B.
        • Colagiuri S.
        • Zimmet P.Z.
        • Tonkin A.M.
        • et al.
        AUSDRISK: an Australian Type 2 Diabetes Risk Assessment Tool based on demographic, lifestyle and simple anthropometric measures.
        Med J Aust. 2010; 192: 197-202https://doi.org/10.5694/mja2.2010.192.issue-410.5694/j.1326-5377.2010.tb03478.x
        • Mohan V.
        • Sandeep S.
        • Deepa M.
        • Gokulakrishnan K.
        • Datta M.
        • Deepa R.
        A diabetes risk score helps identify metabolic syndrome and cardiovascular risk in Indians ? the Chennai Urban Rural Epidemiology Study (CURES-38).
        Diabetes, Obes Metab. 2007; 9: 337-343https://doi.org/10.1111/j.1463-1326.2006.00612.x
        • Roberts S.
        • Barry E.
        • Craig D.
        • Airoldi M.
        • Bevan G.
        • Greenhalgh T.
        Preventing type 2 diabetes: systematic review of studies of cost-effectiveness of lifestyle programmes and metformin, with and without screening, for pre-diabetes.
        BMJ Open. 2017; 7: e017184https://doi.org/10.1136/bmjopen-2017-01718410.1136/bmjopen-2017-017184.supp1
        • Aschner P.
        • Karuranga S.
        • James S.
        • Simmons D.
        • Basit A.
        • Shaw J.E.
        • et al.
        The International Diabetes Federation’s guide for diabetes epidemiological studies.
        Diabetes Res Clin Pract. 2021; 172: 108630https://doi.org/10.1016/j.diabres.2020.108630
        • Coppell K.J.
        • Mann J.I.
        • Williams S.M.
        • Jo E.
        • Drury P.L.
        • Miller J.C.
        • et al.
        Prevalence of diagnosed and undiagnosed diabetes and prediabetes in New Zealand: Findings from the 2008/09 adult nutrition survey.
        N Z Med J. 2013; 126: 23-42
        • Galaviz K.I.
        • Narayan K.M.V.
        • Lobelo F.
        • Weber M.B.
        Lifestyle and the Prevention of Type 2 Diabetes: A Status Report.
        Am J Lifestyle Med. 2018; 12: 4-20https://doi.org/10.1177/1559827615619159
        • Bahijri S.
        • Al-Raddadi R.
        • Ajabnoor G.
        • Jambi H.
        • Al Ahmadi J.
        • Borai A.
        • et al.
        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: 844-855https://doi.org/10.1111/jdi.13213
      4. Costa B, Barrio F, Cabré JJ, Piñol JL, Cos FX, Solé C, et al. Shifting from glucose diagnostic criteria to the new HbA 1c criteria would have a profound impact on prevalence of diabetes among a high-risk Spanish population. Diabet Med 2011;28:1234–7. https://doi.org/10.1111/j.1464-5491.2011.03304.x.

      5. American Diabetes Association. Diagnosis and classification of diabetes mellitus. Diabetes Care 2010;33. https://doi.org/10.2337/dc10-S062.

      6. Bluett. NHS Research Scotland (NRS) Diabetes Register 2021.

      7. Diabetes Australia. National Diabetes Service Scheme. 2021. https://www.ndss.com.au/.

        • Sun Hong
        • Saeedi Pouya
        • Karuranga Suvi
        • Pinkepank Moritz
        • et al.
        IDF diabetes Atlas: Global, regional and country-level diabetes prevalence estimates for 2021 and projections for 2045.
        Diabetes Research and Clinical Practice. 2021; (In press)https://doi.org/10.1016/J.DIABRES.2021.109119