If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Diabetes is an increasingly important condition globally and robust estimates of its prevalence are required for allocating resources.
Methods
Data sources from 1980 to April 2011 were sought and characterised. The Analytic Hierarchy Process (AHP) was used to select the most appropriate study or studies for each country, and estimates for countries without data were modelled. A logistic regression model was used to generate smoothed age-specific estimates which were applied to UN population estimates for 2011.
Results
A total of 565 data sources were reviewed, of which 170 sources from 110 countries were selected. In 2011 there are 366 million people with diabetes, and this is expected to rise to 552 million by 2030. Most people with diabetes live in low- and middle-income countries, and these countries will also see the greatest increase over the next 19 years.
Discussion
This paper builds on previous IDF estimates and shows that the global diabetes epidemic continues to grow. Recent studies show that previous estimates have been very conservative. The new IDF estimates use a simple and transparent approach and are consistent with recent estimates from the Global Burden of Disease study. IDF estimates will be updated annually.
Diabetes mellitus is one of the most common chronic diseases in nearly all countries, and continues to increase in numbers and significance, as economic development and urbanization lead to changing lifestyles characterised by reduced physical activity, and increased obesity. Estimates of the current and future burden of diabetes are important in order to allocate community and health resources, to emphasise the role of lifestyle, and encourage measures to counteract trends for increasing prevalence.
There have been several previous estimates of the number of persons with diabetes [
National, regional, and global trends in fasting plasma glucose and diabetes prevalence since 1980: systematic analysis of health examination surveys and epidemiological studies with 370 country-years and 2.7 million participants.
Guariguata L, Whiting DR, Weil C, Unwin NC. The International Diabetes Federation diabetes atlas methodology for estimating global and national prevalence of diabetes in adults. Diabetes Res Clin Pract 2011; [this issue].
]. Briefly, the PubMed database and Google Scholar were searched for sources published between January 1980 and April 2011, using the search terms: (‘diabetes’ or ‘impaired glucose tolerance’) and ‘prevalence’ and (country name or region/continent); ‘cardiovascular risk factors’ and (country name or region/continent).
Studies were identified that reported on the prevalence of diabetes for at least three adult age-groups. A number of other avenues were explored in the search for relevant data. Relevant citations from each article were obtained, and diabetes researchers in each IDF geographical region were consulted. Studies were classified by a number of criteria: method of diagnosis; population size; study type (e.g. population-based, clinic-based, diabetes registry, medical records review); representativesness; age of the data source; type of data source (e.g. peer-reviewed publication, national report).
Scores for each criterion were determined by an expert committee using the Analytic Hierarchy Process (AHP) [
] and applied to each potential data source. AHP is a means by which criteria from different domains may be compared, e.g. study size versus age of the study. Two thresholds, lower and upper, were determined. Studies that scored below the lower threshold were discarded and studies that scored above the upper threshold were selected. For each country, where there were no studies above the upper threshold, the top-scoring study between the lower and upper threshold along with any other studies with a score that was within 10 percentage points of the top-scoring study were selected. In countries where more than one study was selected, a weighted average, based on the score, was calculated for each age-specific prevalence.
If no studies were selected for a country, data from countries within a ‘data region’ were used as a proxy. A data region was defined as a combination of IDF region, World Bank country income group [
Estimates were made of the total diabetes population, including those who were newly diagnosed in surveys, and those with type 1 diabetes.
2.2 Statistical methods
For each data source, prevalence was modelled using a logistic regression, with age as a quadratic term to allow flattening for older ages. Rates were calculated separately for males and females, and for urban and rural populations. Where data were not available for one setting or for one sex, these were estimated from prevalence ratios from other sources within the data region. The proportion of people found in population-based prevalence surveys to have undiagnosed diabetes was estimated for each combination of IDF region and World Bank country income group. These proportions were used to adjust estimates from self-report surveys of diabetes.
The smoothed age- and sex-specific prevalences for urban and rural settings were then applied to each national population distribution for the years 2011 and 2030 (using the United Nations Population estimates [
]) to estimate national prevalence and numbers of adults with diabetes. The age-specific prevalences of each country were also applied to the world population distribution to determine age- and sex-adjusted prevalences for each country. Calculations were performed using the R statistical program [
Countries were assigned to regions based on the IDF membership regions [Africa (AFR); Middle East and North Africa (MENA); Europe (EUR); North America and Caribbean (NAC); South and Central America (SACA); Southeast Asia (SEA); Western Pacific (WP)]. Countries were also assigned to World Bank income groups: low-income, lower middle-income, upper middle-income and high income [
A total of 565 data sources were reviewed, of which 170 sources from 110 countries were selected.
The data sources and the results for the 80 most populous countries (those with 2011 adult population of greater than 6.399 million, with a combined 2011 population of 94% of the world adult population) are shown in Table 1, Table 2. There were 70 of the 80 countries that had their own data (98 separate studies). Details of prevalence and case numbers for all 216 countries can be found in the online Appendix at www.idf.org/diabetesatlas.
Table 1Data sources for the 80 most populous countries.
Mesure des facteurs de risque des maladies non transmissibles au Niger (approche Step“wise” de l’OMS), tech. rep. Ministere de la Sante Publique, Republique du Niger; 2008.
Malawi National STEPS Survey for Chronic Non-Communicable Diseases and their risk factors, tech. rep. Malawi Ministry of Health, World Health Organization; 2010.
Zimbabwe NCDs risk factors surveillance report 2005, tech. rep. Ministry of Health and Child Welfare, University of Zimbabwe, World Health Organization; 2005.
Malawi National STEPS Survey for Chronic Non-Communicable Diseases and their risk factors, tech. rep. Malawi Ministry of Health, World Health Organization; 2010.
Mesure des facteurs de risque des maladies non transmissibles au Niger (approche Step“wise” de l’OMS), tech. rep. Ministere de la Sante Publique, Republique du Niger; 2008.
Myocardial infarction and complications. Longitudinal observation of a population of 280,000 women and men-Project POL-MONICA Krakow. I. Genesis and objectives of the WHO MONICA Project.
A first national prevalence estimate of diagnosed and undiagnosed diabetes in France in 18- to 74-year-old individuals: the French Nutrition and Health Survey 2006/2007.
A first national prevalence estimate of diagnosed and undiagnosed diabetes in France in 18- to 74-year-old individuals: the French Nutrition and Health Survey 2006/2007.
Prevalence rate of diabetes mellitus and impaired fasting glycemia in Hungary: cross-sectional study on nationally representative sample of people aged 20–69 years.
Myocardial infarction and complications. Longitudinal observation of a population of 280,000 women and men-Project POL-MONICA Krakow. I. Genesis and objectives of the WHO MONICA Project.
Myocardial infarction and complications. Longitudinal observation of a population of 280,000 women and men-Project POL-MONICA Krakow. I. Genesis and objectives of the WHO MONICA Project.
Myocardial infarction and complications. Longitudinal observation of a population of 280,000 women and men-Project POL-MONICA Krakow. I. Genesis and objectives of the WHO MONICA Project.
Myocardial infarction and complications. Longitudinal observation of a population of 280,000 women and men-Project POL-MONICA Krakow. I. Genesis and objectives of the WHO MONICA Project.
Prevalence of diabetes mellitus and cardiovascular risk factors in the adult population of the autonomous region of Madrid (Spain): the PREDIMERC study.
Ministry of Health Algeria, World Health Organization. Algérie Enquête STEPS 2003, tech. rep. Ministry of Health Algeria, World Health Organization; 2005.
Ellabany E, Abel-Nasser M. Community based survey study on non-communicable diseases and their risk factors, Egypt, 2005–2006, tech. rep. Egypt Ministry of Health and Population, and WHO; 2006.
Chronic non-communicable diseases risk factor survey in Iraq, 2006, tech. rep. Iraq Ministry of Health and Ministry of planning and Development Cooperation, World Health Organization; 2006.
Chronic non-communicable diseases risk factor survey in Iraq, 2006, tech. rep. Iraq Ministry of Health and Ministry of planning and Development Cooperation, World Health Organization; 2006.
Chronic non-communicable diseases risk factor survey in Iraq, 2006, tech. rep. Iraq Ministry of Health and Ministry of planning and Development Cooperation, World Health Organization; 2006.
National Center for Health Statistics. National Health and Nutrition Survey (NHANES), tech. rep. Centers for Disease Control and Prevention (CDC), Hyattsville, MD; 2010.
National Center for Health Statistics. National Health and Nutrition Survey (NHANES), tech. rep. Centers for Disease Control and Prevention (CDC), Hyattsville, MD; 2011.
Multicenter study of the prevalence of diabetes mellitus and impaired glucose tolerance in the urban Brazilian population aged 30–69 yr. the Brazilian Cooperative Group on the Study of Diabetes Prevalence.
Multicenter study of the prevalence of diabetes mellitus and impaired glucose tolerance in the urban Brazilian population aged 30–69 yr. the Brazilian Cooperative Group on the Study of Diabetes Prevalence.
Encuesta Demografica y de Salud, Republica Dominicana 2007, tech. rep. Centro de Estudios Sociales y Demograficos (CESDEM), Macro International Inc.; 2008.
Iniciativa Centroamericana de Diabetes (CAMDI). Encuesta de diabetes, hipertension y factores de riesgo de enfermedaded cronicas. Villa Nueva, Guatemala 2006, tech. rep. Pan American Health Organization (OPS), Centers for Disease Control and Prevention (CDC), Instituto de Nutricion de Centro America y Panama; 2007.
Multicenter study of the prevalence of diabetes mellitus and impaired glucose tolerance in the urban Brazilian population aged 30–69 yr. the Brazilian Cooperative Group on the Study of Diabetes Prevalence.
University of Health Sciences and Ministry of Health. Prevalence of non-communicable disease risk factors in Cambodia, tech. rep. University of Health Sciences and Ministry of Health; 2010.
Hong Kong Cardiovascular Risk Factor Prevalence Study Steering Committee
Fibrinogen, other cardiovascular risk factors and diabetes mellitus in Hong Kong: a community with high prevalence of type 2 diabetes mellitus and impaired glucose tolerance.
The prevalence of diabetes, association with cardiovascular risk factors and implications of diagnostic criteria (ADA 1997 and WHO 1998) in a 1996 community-based population study in Hong Kong Chinese. Hong Kong Cardiovascular Risk Factor Steering Committee, American Diabetes Association.
University of Health Sciences and Ministry of Health. Prevalence of non-communicable disease risk factors in Cambodia, tech. rep. University of Health Sciences and Ministry of Health; 2010.
The highest regional prevalence (Table 3) for 2011 (after age standardization to the world population) was for MENA, followed by the NAC and WP. The AFR region is expected to have the largest proportional increase in adult diabetes numbers by 2030, followed by MENA, though WP will continue to have the world's highest number of adults with diabetes, due primarily to the number of people with diabetes in China. Every region will have an increase in numbers well in excess of adult population growth, and total numbers with diabetes are likely to increase by 50.7% over the 19 years.
Table 3Prevalence of diabetes and estimated diabetes numbers by region among adults aged 20–79 years for the years 2011 and 2030.
The 10 countries by prevalence are dominated by pacific island states and countries in the Middle East (Table 4). Table 5 shows the 10 countries with the largest numbers of people with diabetes. As might be expected, the countries with the largest populations have the highest number of persons with diabetes. Only Pakistan and Nigeria of the world's 10 most populous countries are not among the 10 countries with the highest diabetes numbers (replaced by Mexico and Egypt) for 2011.
Table 4Top 10 countries for diabetes prevalence in 2011 and 2030.
The numbers of people with diabetes differ substantially by World Bank country income group. Fig. 1 shows current estimated numbers of people with diabetes by age-group for 2011 and 2030. The rate of the increase in numbers with diabetes is inversely related to current income status, with the greatest increase expected in low-income countries (92%), followed by lower-middle income countries (57%), upper-middle income countries (46%) and finally higher income countries (25%). This compares to adult populations which are expected to increase by 58%, 29%, 19%, and 7% respectively.
Fig. 1Numbers of adults with diabetes by World Bank income group in 2011 and 2030, according to age. LIC: low-income countries; LMIC: lower middle-income countries; UMIC: upper middle-income countries; HIC: high-income countries.
The largest increases are expected in the older age groups in low and lower-middle income countries, with numbers more than doubling for the over 60-year age-group. In high-income countries, an increase (42%) is only expected among the over 60s, with almost no change predicted for the younger age-groups. Currently, the greatest number of people worldwide with diabetes is in the 40–59-year-old age-group, and this is predicted to remain so in 2030, although there will almost as many people with diabetes in the 60–79-year-old age-group.
The overall total predicted increase in numbers with diabetes from 2011 to 2030 is 50.7%, at an averaged annual growth of 2.7%, which is 1.7 times the annual growth of the total world adult population.
Forty-eight percent of the anticipated absolute global increase of 186 million people with diabetes is projected to occur in India and China alone.
4. Discussion
The estimates presented here are based on methods that are an evolution of the methods developed by Shaw et al. [
] and suggest that in 2011 there will be 366 million people worldwide with diabetes, with considerable disparity between populations and regions. The pattern of diabetes varies considerably according to countries’ income status. For countries classified by the World Bank as being high-income countries, most people with diabetes are aged over 60 years, whereas for low- and middle-income countries most people with diabetes are of working age, between 40 and 60 years.
This difference is likely to still be present in 2030, although less marked, as the average age of low- and middle-income countries’ populations will increase slightly more than in high-income countries. Population growth, ageing of populations, and urbanization with associated lifestyle change is likely to lead to a 50.7% increase in worldwide numbers with diabetes by 2030.
Over the last 15 years several global estimates of the prevalence of diabetes have been produced, and these are presented in Fig. 2. In 1998, King et al. [
]. Each estimate has been based on the latest data available and each subsequent estimate has been higher than the previous report. While there were some differences in the methods the main reason for the increase has been the availability of newer data that incorporate real increases in the incidence of diabetes that go beyond what is predicted by the models.
Fig. 2Estimates and projections of global diabetes prevalence, including the estimates presented in this paper.
National, regional, and global trends in fasting plasma glucose and diabetes prevalence since 1980: systematic analysis of health examination surveys and epidemiological studies with 370 country-years and 2.7 million participants.
]. Their estimate of the global prevalence of diabetes, 346 (302–394) million in 2008, is very similar to our estimate, with ours being well-within their uncertainty intervals. The similarity in the global estimate does, however, mask some differences at the country-level, with GBD estimates for countries in Africa appearing to be rather high, and some trends far from intuitive. For example, the GBD project estimates that the prevalence of diabetes in Tanzania in the 1980s, when most of the population living as subsistence farmers and the country was in the middle of a long period of severe economic difficulty, was almost as high as it is now in the USA.
Counter-intuitive country-level estimates notwithstanding, the latest GBD and IDF global estimates are very close to each other, with the IDF estimates continuing the trend in the GBD estimates.
These data suggest that the prevalence of diabetes is increasing, as a consequence of increasing incidence due to demographic changes such as ageing, and as a result of risk factors such as obesity and sedentary life becoming more common, and also a result of better health care improving longevity of people with diabetes. They also indicate that previous modelled estimates have generally been under-estimating the prevalence for any given year.
The IDF approach is deliberately simple and conservative. We do not model changes in incidence, and have based our projections for 2030 on predicted demographic changes: urbanization and ageing. Urbanization is associated with a more sedentary lifestyle tending to increase diabetes prevalence [
], so to some extent is a proxy for lifestyle changes. Other than the impact of urbanization, we have not attempted to directly account for the effects of changes in risk factors (e.g. obesity), as accurately assessing the relationship between risk factors and diabetes is difficult across the diverse global population. Thus, if the prevalence of obesity and other risk factors continue to rise, it is likely that the estimates presented here will be lower than the actual prevalence in 2030.
Three major factors affect the accuracy of these estimates: the availability and quality of data, and the representativeness of the data sources chosen. Recently there have been a number of data sources that provide information on diabetes prevalence with 117 out of the 170 studies being multi-regional, population-based studies (although only 41 have used OGTT).
The region with the fewest multi-region population-based data sources is South and Central America, with just 6, of which only 1 used OGTT. Europe, Western Pacific, Middle East and North Africa had the most multi-region, population-based data sources with 29, 26 and 24, respectively.
The global pattern is dominated by countries with large populations, and these data highlight the extent to which demographic changes in India, China and Brazil are likely to affect the total numbers with diabetes in the future. Each of these countries has had relatively recent national surveys, so that the likelihood of unrepresentative data is reduced.
In summary, these results update the global estimates of diabetes prevalence and show that diabetes is continuing to be an increasing international health burden. The estimates here are higher than previous estimates, and are consistent with recent estimates using more complex methods. The estimates have been generated using a conservative approach and as a result may be an under-estimate.
Ageing and the changes that are associated with urbanization, globalisation and development are increasingly adding to the burden of diabetes in all countries, and particularly in low- and middle-income countries where resources for dealing with the associated clinical problems are most scarce.
Conflict of interest
There are no conflicts of interest.
Acknowledgements
The IDF Diabetes Atlas project was supported by the following sponsors: Lilly Diabetes, Merck and Co, Inc., Novo Nordisk A/S supported through an unrestricted grant by the Novo Nordisk Changing Diabetes initiative, and Sanofi.
References
King H.
Rewers M.
Global estimates for prevalence of diabetes mellitus and impaired glucose tolerance in adults. WHO Ad Hoc Diabetes Reporting Group.
National, regional, and global trends in fasting plasma glucose and diabetes prevalence since 1980: systematic analysis of health examination surveys and epidemiological studies with 370 country-years and 2.7 million participants.
Guariguata L, Whiting DR, Weil C, Unwin NC. The International Diabetes Federation diabetes atlas methodology for estimating global and national prevalence of diabetes in adults. Diabetes Res Clin Pract 2011; [this issue].
Mesure des facteurs de risque des maladies non transmissibles au Niger (approche Step“wise” de l’OMS), tech. rep. Ministere de la Sante Publique, Republique du Niger; 2008.
Malawi National STEPS Survey for Chronic Non-Communicable Diseases and their risk factors, tech. rep. Malawi Ministry of Health, World Health Organization; 2010.
Zimbabwe NCDs risk factors surveillance report 2005, tech. rep. Ministry of Health and Child Welfare, University of Zimbabwe, World Health Organization; 2005.
Myocardial infarction and complications. Longitudinal observation of a population of 280,000 women and men-Project POL-MONICA Krakow. I. Genesis and objectives of the WHO MONICA Project.
A first national prevalence estimate of diagnosed and undiagnosed diabetes in France in 18- to 74-year-old individuals: the French Nutrition and Health Survey 2006/2007.
Prevalence rate of diabetes mellitus and impaired fasting glycemia in Hungary: cross-sectional study on nationally representative sample of people aged 20–69 years.
Prevalence of diabetes mellitus and cardiovascular risk factors in the adult population of the autonomous region of Madrid (Spain): the PREDIMERC study.
Ministry of Health Algeria, World Health Organization. Algérie Enquête STEPS 2003, tech. rep. Ministry of Health Algeria, World Health Organization; 2005.
Ellabany E, Abel-Nasser M. Community based survey study on non-communicable diseases and their risk factors, Egypt, 2005–2006, tech. rep. Egypt Ministry of Health and Population, and WHO; 2006.
Chronic non-communicable diseases risk factor survey in Iraq, 2006, tech. rep. Iraq Ministry of Health and Ministry of planning and Development Cooperation, World Health Organization; 2006.
National Center for Health Statistics. National Health and Nutrition Survey (NHANES), tech. rep. Centers for Disease Control and Prevention (CDC), Hyattsville, MD; 2010.
National Center for Health Statistics. National Health and Nutrition Survey (NHANES), tech. rep. Centers for Disease Control and Prevention (CDC), Hyattsville, MD; 2011.
Multicenter study of the prevalence of diabetes mellitus and impaired glucose tolerance in the urban Brazilian population aged 30–69 yr. the Brazilian Cooperative Group on the Study of Diabetes Prevalence.
Encuesta Demografica y de Salud, Republica Dominicana 2007, tech. rep. Centro de Estudios Sociales y Demograficos (CESDEM), Macro International Inc.; 2008.
Iniciativa Centroamericana de Diabetes (CAMDI). Encuesta de diabetes, hipertension y factores de riesgo de enfermedaded cronicas. Villa Nueva, Guatemala 2006, tech. rep. Pan American Health Organization (OPS), Centers for Disease Control and Prevention (CDC), Instituto de Nutricion de Centro America y Panama; 2007.
University of Health Sciences and Ministry of Health. Prevalence of non-communicable disease risk factors in Cambodia, tech. rep. University of Health Sciences and Ministry of Health; 2010.
Hong Kong Cardiovascular Risk Factor Prevalence Study Steering Committee
Fibrinogen, other cardiovascular risk factors and diabetes mellitus in Hong Kong: a community with high prevalence of type 2 diabetes mellitus and impaired glucose tolerance.
The prevalence of diabetes, association with cardiovascular risk factors and implications of diagnostic criteria (ADA 1997 and WHO 1998) in a 1996 community-based population study in Hong Kong Chinese. Hong Kong Cardiovascular Risk Factor Steering Committee, American Diabetes Association.
00591-2&id=gr1.jpg){kind=link}
00591-2&id=gr2.jpg){kind=link}