IDF Diabetes Atlas 8th Edition: 425 million have diabetes globally, with 629 million projected by 2045; 1 in 2 undiagnosed – November 14, 2017

Executive Highlights

  • IDF just published the eighth edition of its Diabetes Atlas: Today, there are 425 million people ages 20-79 with diabetes, up 2.4% (10 million) from 2015. One in two are undiagnosed. Diabetes prevalence is projected to reach 629 million in 2045, which actually reflects a tempering of the 2015 Atlas’ projection for 642 million people with diabetes in 2040. Someone dies from diabetes every eight seconds (four million in a year).
  • One in eight global healthcare dollars ($727 billion) goes toward diabetes, up from $673 billion in 2015, up 8%. The US is responsible for 48% of these costs and the increase in US costs from $274 billion in 2015 to $348 billion in 2017 shows that US increases drive 100% of growth. Our rough calculations show that direct+indirect global costs easily break $1 trillion. By 2045, IDF very conservatively expects spending to increase 7% to $776 billion in direct costs. We are sure this estimate is low, when considered alongside a 48% increase in prevalence; the IDF estimates on cost have always struck us as low.
  • China (114 million), India (73 million), and the US (30 million) sit atop the rankings for the most people with diabetes, but moving forward, India seems to be in the most trouble. IDF estimates that India will gain >60 million people with diabetes by 2045 while the US and China will add six million each (these are probably low).

This morning, IDF published its eighth edition, 150-page Diabetes Atlas, and as we all feared, but expected, the pandemic has worsened in the past two years: By the Atlas’ estimates, there are now 425 million people in the world living with diabetes – that’s 1 in 11 adults ages 20-79 years old – a 2.4% increase from 415 million in 2015. More concerning, it has been estimated that half of these people, or 212.4 million, are undiagnosed. By 2045, IDF projects that we’ll be faced with a global burden of 629 million people with diabetes (1 in 10 adults). While this would still spell utter devastation for health systems and global productivity, it actually reflects an improvement from 2015, when researchers suggested that 642 million people could have diabetes by 2040. The prediction is now for 13 million fewer people, five years down the line. Perhaps this is a faint signal that we’re beginning to bend the curve (or will be able to), but clearly nowhere near enough and more likely, this reduction just reflects a change in methodology, which always seems to be moving around with IDF estimates. When we look at where we’ve been – there were ~30 million people diagnosed with diabetes globally in 1985, meaning we’ve seen 14x growth in the past 32 years – and where we are going, it becomes wholly apparent that we need to halt this train in its tracks by investing all-in on prevention.

From a financial point of view, an estimated one in eight global healthcare dollars ($727 billion, of which the US is responsible for 48%) went to diabetes in 2017. This is up from $673 billion in 2015, and just $232 billion in 2006 (more than tripling in 11 years). Our rough estimates show that, taking into account indirect costs, diabetes easily costs over $1 trillion. IDF conservatively projects that spending will reach $776 billion in 2045 – just ~7% growth, compared to 48% growth in global diabetes prevalence. In making the calculations, the authors assumed that mean expenditure per person and diabetes prevalence rate would remain constant.

With regards to mortality, the Atlas suggests that a person dies every eight seconds from diabetes (four million total in a year), actually an improvement from one every six seconds in 2015 – this may be chalked up to a global decrease in all-cause mortality though we’d think that it reflects just more methodology changes overall. Still, IDF President Dr. Shaukat Sadikot pointed out in a conversation with us last week that just because the overall number is not as bad as it used to be, the disparities in mortality rates between the US/Europe and the rest of the world are big causes for concern. Despite the decrease, diabetes is still a bigger killer than HIV/AIDS, tuberculosis, and malaria combined.

Unsurprisingly, China (114 million, +4% from 2015), India (73 million, +6% from 2015), and the US (30 million, +3% since 2015) sit atop the ranking for highest diabetes prevalence. By 2045, IDF projects 61 million new diabetes cases in India (+84% from 2017), and just six million in China (+5% from 2017) and the US (+20% from 2017). Our calculations show that India will have added the population of Denmark 11-times over in people with diabetes by the time 28 years have passed – this is a true disaster in progress. As for China and the US, we wonder if the projections might be an underestimate. JAMA reported in 2014 that 493 million Chinese adults have prediabetes, and nearly 90 million Americans have prediabetes as well – how in the world are these countries going to prevent a diabetes avalanche at this rate?

While India and China are without a doubt the biggest concerns in terms of population burden, 69% of people in Africa with diabetes are undiagnosed, compared to “just”38% in North America/the Caribbean. As a probable result, diabetes accounts for 77% of all deaths before the age of 60 in Africa. Africa is also expected to be a major driver of diabetes prevalence (+156% by 2045). On the whole, 90% of people with diabetes live in low- and middle-income counties, suggesting that an investment in poverty-alleviation is an investment in diabetes prevention.

IDF estimates that 352 million people – 7.3% of adults ages 20-79 years – have prediabetes as defined by the strict WHO definition of Impaired Glucose Tolerance (IGT). We trend past Atlas estimates below, which tend to vary and probably underreport the prevalence of prediabetes (as measured by fasting glucose and A1c). While the projection for people living with diabetes in 2045 increased, that for people living with IGT went up to 532 million from 481 million in 2015.

Thanks to Lilly, Merck (MSD, to be exact), and Novo Nordisk for sponsoring this venture. A recent press release announced that Merck and IDF will be teaming up further to focus on education and awareness-raising to facilitate prevention of type 2 diabetes. So far, the partnership includes a CME accredited, IDF-certified HCP course about prevention, as well as a joint awareness campaign to support women’s health today, World Diabetes Day.

At the Cities Changing Diabetes Summit in Houston last month, stakeholders from all over healthcare and policy (and all over the world!) committed to slashing obesity by a whopping 25% in cities so that diabetes rates go no higher than one in 10 (prevalence is one in 11 globally now). The Cities Changing Diabetes movement was implicated in this IDF report, which said that by 2045, cities will be home to ~75% of people with diabetes.

This data underscores the importance of prevention – we’re on the path to a truly unsustainable future. IDF concluded the report with a well-put call to action, asking that:

  • governments and healthcare providers support diabetes prevention and management by providing health lifestyle education and fostering environments that facilitate physical activity;
  • governments and other stakeholders prioritize affordable and equitable access to insulin as an essential life-saving medicine (according to a survey in the Atlas, 0% of low-income countries have access to short or long-acting insulin, 10% have access to BGM/strips, and 13% have access to syringes/needles. Interestingly, 33% do have access to GLP-1 agonists, higher than the 17% of middle-income countries).
  • there be a multisectorial response to diabetes.

Certainly, we will not break through without meeting all three of these criteria, and learning from the things that do work well in prevention. Our hats are off to IDF for all of this work to bring to the surface every couple of years the unrelenting havoc that diabetes is wreaking on the world. As Dr. Sadikot writes in the foreword: “Diabetes is not a health crisis; it is a global societal catastrophe.”

  • As with 2015, IDF has added confidence intervals to quantify the uncertainty around estimates of diabetes prevalence. These are fairly wide in most cases – for instance, the number of people with diabetes globally could be as low as 346 million or up to 545 million rather than 425 million. Similarly, the 2040 number could be as low as 477 million or as high as 809 million rather than 629 million. This is obviously a tremendously difficult task and we salute IDF for taking it on and admitting the intervals around its estimates. It certainly does not change the overall message.
  • In 32 years, we’ve gone from 30 million people with diabetes to 425 million (see below). The actual number of people with diabetes has surpassed the future projection ~8-11 years early, and:
    • The 1995 and 2000 projections for 2025 were passed in 2010, 15 years early
    • The 2003 estimate for 2025 was passed in 2011, 14 years early
    • The 2006 estimate for 2025 was passed in 2013, 12 years early
    • The last four future projections (2011, 2013, 2015, 2017) have given up on 2025, instead projecting out to 2030, 2035, 2040, and 2045. The growth in future projections has ranged from 0-26%, while actual diabetes growth has ranged from 12-28%.

Global Diabetes Prevalence

Figure 1: Actual (1985-2017) and Projected (2025-2045) Number of People with Diabetes Worldwide

Table 1: Growth of Actual (1985-2017) and Projected (2025-2045) Number of People with Diabetes Worldwide

Year of Estimate Future Projection


1995 2025

2000 2025

2003 2025

2006 2025

2009 2030

2011 2030

2013 2035

2015 2040



People with Diabetes (millions)











Growth from Previous Estimate











Projected People with Diabetes (millions)











Growth from Previous Projection












There are now an estimated 425 million people ages 20-79 with diabetes in the world, 10 million more (+2.4%) than in 2015. IDF estimates that there will be 629 million people living with this disease in 2045 (48% growth). On average, that equates to the diabetes load increasing by 7.3 million people each year – put that in the context of the past two years, where ~5 million people were added per year. Of course, the trend won’t actually be linear. On an ever-so-slightly encouraging note, the future prevalence estimation has actually decreased for the first time since we’ve been tracking: IDF previously guided for 642 million people with diabetes in 2040, and now they are projecting that there will be 629 million in 2045. That is the smallest of victories, and we’re not sure why, given the rising prevalence today, it went down. Perhaps epidemiologists have reason to believe that we can begin to bend the curve, though we undoubtedly have to do more, because preventing 13 million cases will do little in the grand scheme. Still, we’ll look at any decrease as favorable in today’s climate.

  • Gestational diabetes affects nearly one in six births worldwide (16.2%), or 21.3 million live births affected in 2017. Prevalence of hyperglycemia in pregnancy is 24.2# in Southeast Asia, and “just” 10.4% in Africa – 88% of cases were in low- and middle-income countries. Notably, age is a huge contributor to hyperglycemia in pregnancy, as glucose dysregulation is over 4.5x more common in mothers ages 45-49 vs. 20-24. 
    • The JDRF-supported CONCEPTT trial of CGM in pregnancy, which read out at EASD, suggests that CGM can significantly improve neonatal outcomes, if not maternal glycemia. We are under no illusion that every health system can afford to give expectant mothers CGM and support – especially because the argument of ‘investment’ (pay for CGM up front, so you don’t have to treat the infant later) breaks down when neonatal care isn’t up to par – but it is a moonshot (and a potential area for high-impact philanthropy).
  • IDF reports that 1.1 million children and adolescents under the age of 20 years, and 586,000 under the age of 15 years, has type 1 diabetes. In 2015 and prior, IDF has reported type 1 diabetes prevalence in children <14 years old (542,000 in 2015), making it difficult to make a direct comparison. That said, we can assume there was a meaningful uptick – it’s unlikely that the 44,000-person increase in the past two years can be accounted for by 14-year-olds not captured in 2015. This is further supported by the fact that there are estimated to be ~96,100 new cases of type 1 diabetes in children <15 years old per year, up from 86,000 in children <14 years old in 2015. As in previous years, the Atlas materials don’t report type 1 diabetes in adults, even though JDRF has previously reported in 2014 that 85% of people with type 1 diabetes are adults – presumably, this number continues to rise as better, safer drugs and technologies become available to wider swathes of people. The JDRF report doesn’t define what constitutes an “adult” in terms of years, but assuming that adults are 15+ years, then there could be 3.3 million adults globally with type 1 diabetes. That would total ~3.9 million people globally, or less than 1% of diabetes on the planet.
    • The US still tops the list of countries with the most people <20 years with type 1 diabetes (169,900). The US also has the most new cases per year (17,100), followed closely by India at 16,800/year. However, when normalized for population size, the Nordic countries are highly enriched at the top for incidence rates per year: #1 Finland, #3 Sweden, #5 Norway, and #10 Denmark. Kuwait and Saudi Arabia are number two and number four, respectively. If Finland were as large as the US, it would have ~185,000 new cases of type 1 diabetes every year!
    • The Atlas mentions that “there is evidence that type 2 diabetes in children and adolescents is increasing in some countries,” but concedes that reliable data is sparse. The authors conclude that more information is “needed urgently.”
  • As it stands, ~66% of people with diabetes live in cities (279 million vs. 146 million). This gap is expected to grow even more marked by 2045, when the Atlas estimates ~75 of people with diabetes will live in cities (473 million vs. 156 million). Further, the prevalence of diabetes in urban settings (10.2%) is higher than that in rural settings (6.9%). This is one of the key rationales for using the city as the nucleus for change in Cities Changing Diabetes.

Diabetes Prevalence by Country

Table 2: Top 10 Countries by Diabetes Prevalence


Projected 2045
(Growth from 2015)

1. China: 114 million

1. India: 134 million (+84%)

2. India: 73 million

2. China: 120 million (+5%)

3. USA: 30 million

3. USA: 36 million (+20%)

4. Brazil: 13 million

4. Mexico: 22 million (+83%)

5. Mexico: 12 million

5. Brazil: 20 million (+54%)

6. Indonesia: 10 million

6. Egypt: 17 million (+%113)

7. Russian Federation: 9 million

7. Indonesia: 17 million (+70%)

8. Egypt: 8 million

8. Pakistan: 16 million (+100%)

9. Germany: 8 million

9. Bangladesh: 14 million (+100%)

10. Pakistan: 8 million

10. Turkey: 11 million (+57%)

  • China and India continue to be the biggest areas of concern for diabetes burden globally. Combined, these two large countries account for 44% of the world’s diabetes population currently. Though the combined number will grow by 67 million in the next 28 years, per the Atlas’ estimates, the share of worldwide percentage will drop to ~40%. While China is actually projected to essentially level off over the next three decades (adding only six million people with diabetes), India is a disaster in progress. By 2045, IDF estimates that India will have added the population of Denmark – 11-times over – in people with diabetes, to reach 134 million. That accounts for 30% of the growth of the world’s diabetes population – if you had to choose one place to focus prevention efforts, India must be that place.
    • The top 10 countries are expected to remain largely the same by 2045, though Russia and Germany are replaced by Bangladesh and Turkey, and there is some shuffling around throughout. We’re very surprised to see such low growth rates coming from China and the USA, and would love to know how the researchers came up with these numbers. In the US alone, there are nearly 90 million people with prediabetes – how is it that only six million of them will progress to diabetes, given current trends? Similarly, in 2015, China was expected to reach 151 million by 2040 – what can account for the tempered projection?
    • 79% of people with diabetes live in low- and middle-income countries. This is strong evidence that an investment in poverty-alleviation is an investment in diabetes prevention. 
    • Though India and China clearly dominate global gross diabetes statistics, North America and the Caribbean has the highest diabetes prevalence (11%), followed by the Middle East and North Africa (10.8%).
    • Not all countries have estimates for diabetes prevalence, so IDF specifically called for more research in this area.
  • On the whole, it is estimated that 50% of people with diabetes are undiagnosed, but there is tremendous disparity across regions: In Africa, 69% of people are undiagnosed, 58% are undiagnosed in Southeast Asia, 54% in the Western Pacific, and “just” 38% in North America/the Caribbean. Unsurprisingly, the undiagnosed proportion is highest in low-income countries (77%), followed by middle-income (53%), and high-income (37%). Widespread diabetes screening is needed so that people can know that they have insulin resistance and/or beta cell burnout before it is too late. How can we think creatively to implement screening in a low-cost, highly scalable way, and then ensure proper follow-up?
  • Driving the projected 48% global growth in diabetes are: (i) Africa (+156%); (ii) Southeast Asia (including India; +84%); (iii) the Middle East & North Africa (+72%); and (iv) South & Central America (+62%). More moderate growth is expected in the Western Pacific (including Australia and China; +15%) and Europe (+16%).

Financial Burden of Diabetes

Figure 2: Diabetes Healthcare Expenditure (2006-2017)

  • Diabetes is projected to cost $727 billion (direct costs for ages 20-79 years) in 2017, an 8% increase from $673 billion in 2015, accounting for 12% of global healthcare expenditure. If the age bracket is broadened to 18-99 years, expenditure is expected to come to $850 billion. Pushing out to 2045, IDF’s research team projects that costs will reach $776 billion per year for ages 20-79, and $958 billion for ages 18-99 years. If the ~7% growth in costs from the 20-79-year-old cohort sounds conservative compared to the 48% growth in prevalence, it is: The authors note that the calculations “assume that the mean expenditure per person and diabetes prevalence rate remain constant, while only demographic changes are observed. This dynamic is supported by the observation that countries with the largest health expenditure will experience a very small population growth.” If this calculation were to include indirect costs, the projected hike from 2017 to 2045 would have to be significantly greater – as far as we’re concerned, these are just as important as direct healthcare expenditures, and we would love to see this analysis form IDF moving forward.
    • Including (our) estimated indirect costs, diabetes in people ages 20-79 cost the world $1.2 trillion in 2017. A recent Lancet paper suggested that indirect diabetes costs accounted for 34.7% of the $1.3 trillion global burden for ages 20-79 in 2015. That’s not to say that cost has decreased – the IDF and Lancet methods don’t align exactly, but $1.2 trillion gives a good idea of the impact of including indirect costs (absenteeism, presenteeism, etc.).

Figure 3: Mean Expenditure per Person with Diabetes by Region

  • The figure above is very telling: (i) There are huge disparities in healthcare spending on diabetes depending on region; and (ii) Bigger circles (higher mean expenditure per person with diabetes) don’t always correlate with lower diabetes prevalence – compare Europe and the US, for example. Of course, this figure doesn’t tell everything: Not all regions were created equal with respect to genetic backgrounds and cultures, and the efficacy of the dollars spent isn’t shown (e.g., how many cases of blindness did the US prevent vs. Europe?).
  • Regarding total healthcare expenditure on diabetes by country, the US easily spends the most at $348 billion per year, >3x higher than #2 China ($110 billion) and 8x higher than #3 Germany ($42 billion). The US spends 48% of all diabetes dollars spent, though it only has 7% of the global diabetes population. This means that, on a per-person basis, the US spends $11,638 per person per year on average, while the next highest, Luxembourg, spend $8,941 per patient-year. We’re reminded of a figure from “The American Health Care Paradox: Why Spending More is Getting Us Less” (2o13), where the authors show that US healthcare spending as a percentage of GDP is much higher than that of other developed nations, but when social care is thrown into the mix, other countries either spend more or similarly to the US (all of the other examined countries spend more on social care than they do healthcare, unlike the US). Given the US’s massive healthcare spend, especially on diabetes, perhaps it is time to rethink the paradigm and invest more heartily in prevention.

Diabetes Mortality

  • Four million people ages 20-79 are estimated to die from diabetes – one death every eight seconds. In this age group, diabetes accounts for ~11% of global all-cause mortality. Four million is notably one million less than the estimate from 2015 (one death every six seconds), a fact the Atlas proposes is a result of a global decrease in all-cause mortality estimates. Among all regions examined, South and Central America is the only one with rising mortality. On the whole, this is a positive sign. However, IDF President Dr. Shaukat Sadikot pointed out to us in a conversation that mean global mortality – one death every eight seconds – is just that, a mean. When you consider that, in his theoretical estimate, there is one death every 10 minutes in the US, and one every four minutes in the EU, “that has to mean that there is one every second in China. This is unacceptable.” In a real example from the Atlas, diabetes accounts for 77% of all deaths before age 60 in Africa.
    • Diabetes still accounts for more deaths in a year than HIV/AIDS, tuberculosis, and malaria combined (3.3 million). Yet because of the infectious and acute nature of these diseases – in addition to there being a relatively simple solution – we’ve seen tremendous scientific and philanthropic interest in quelling them, while diabetes is still not a very attractive place to put money if you’re looking for a rapid, black-and-white ROI.

Prevalence of Impaired Glucose Tolerance

Figure 4: Actual (2003-2017) and Projected (2025-2045) Number of People with Impaired Glucose Tolerance Worldwide

Table 3: Growth of Actual (2003-2017) and Projected (2025-2040) Number of People with Impaired Glucose Tolerance Worldwide

Year of Estimate

Future Projection















People with IGT (millions)








Growth from Previous Estimate








Projected People with IGT (millions)








Growth from Previous Projection








  • 352 million people – 7.3% of adults – ages 20-79 years old are estimated to have impaired glucose tolerance (IGT). Frighteningly, this number grew 11% from 2015 (318 million people), and 72% of people with IGT live in low- and middle-income countries. IDF now projects that there will be 532 billion people with IGT in 2045 (51 million more than expected for 2040 two years ago).
  • The countries with the highest numbers of people with IGT closely resemble those with the highest rates of diabetes: China has ~49 million, the US has ~37 million, Indonesia has ~28 million, and India has 24 million.
  • Overall expenditure due to prediabetes were not broken out (they are difficult to estimate), but a study suggests that the US spent $44 billion in a year.

Diabetes Complications

Cardiovascular Disease

  • People with diabetes are two-to-three times more likely to have cardiovascular disease (CVD) than people without diabetes, and 25% of diabetes inpatient spending are a consequence of CVD. As a demonstration of the interconnectedness between the two conditions, >25% of all CVD expenditure can be blamed on diabetes. The Atlas lists a host of other harrowing statistics: In low- and middle-income countries, $84 billion dollars of GDP were lost due to CVD and diabetes from 2005-2015; In the US, 20% of all diabetes-related inpatient days and 15% of physician office visits are due to CV complications; and each year, 1.4%-4.7% of people ages 50-69 in high- and middle-income countries have a CV event. In the two years since the last Atlas came out, there has arisen reason for hope in CV disease, in the form of cardio-protective anti-hyperglycemic drugs (namely GLP-1 agonists and SGLT-2 inhibitors). We’d love to see these given to people all over the world early in the treatment scheme in order to diminish the burden of one of the most costly diabetes complications.

Eye Disease

  • IDF reports that retinopathy affects 35% of people with diabetes and is the leading cause of vision loss in working-age adults. The prevalence of proliferative (vision-threatening) retinopathy is 7%. Eye disease is the most present in Southeast Asia, where it affects 41% of people with diabetes. Eye complications are not the most costly (retinopathy cost 10 million euros in Sweden in a year) but it has tremendous quality of life implications: 58% of people with diabetes-related eye disease experience limitations on performing daily activities, compared to 37% of those with diabetes but no eye disease.

Kidney Disease

  • Prevalence of end-stage renal disease (ESRD) is up to 10x higher in people with diabetes. Kidney disease is extremely costly, as clinical nephropathy boosts health expenditures by 50%(!) for people with diabetes, per the Atlas. Further, dialysis for ESRD sends annual mean healthcare costs soaring to nearly $35,000, vs. those with ESRD but no dialysis who spent $10,322. It is also quite prevalent, as among people with diabetes, ~20% of people in the UK and ~25% of people in the US will develop chronic kidney disease. A striking 2016 NEJM letter also estimated that 24 million people with diabetes have chronic kidney disease in China.


  • Every 30 seconds, a lower limb/part of a limb is amputated because of diabetes. Overall, the reported prevalence of peripheral neuropathy spans a wide range of 16%-66%, and the global prevalence of diabetest-related foot ailments is 6.4% – 3% in Oceania, 13% in North America, and higher in men than women.  In 2007, ~33% of diabetes costs were estimated to be linked to foot ulcers. Compared to those without foot ulcers, the cost of care for people with diabetes and foot ulcers is 5.4x higher in the year of the first episode (though we realize that foot ulcers could also correlate with other costly complications). We’d note that the number of products that serve as early warning systems for developing ulcers – see Wellpepper’s Sugarpod (winner of the recent Alexa Diabetes Challenge), Podimetrics, Siren Care (just started roll out of temperature-sensing socks and app), and Bonbouton (2nd place at the 2017 Novo Nordisk-Lyfebulb Innovation Summit) – is growing. We hope to see these make an impact in the coming years, though they of course only alert, and there has to be solid foot care education and treatment on the other end.

Periodontal Disease

  • Periodontal disease is present in ~93% of people with diabetes, and ~83% of people without diabetes. The Atlas cites a US study suggesting that oral healthcare could save the $39-$53 billion each year. On the high end, that could knock off ~15% of diabetes healthcare expenditure in the US (assuming that periodontal care is included in the Atlas’ calculation of healthcare expenditure).

Hyperglycemia in Pregnancy

  • Hyperglycemia in pregnancy is associated with $16,000 extra spent per pregnant women. The dollars are primarily spent on treating complications for the mother, but ~25% goes to addressing neonatal complications for the potentially macrosomic child.


-- by Brian Levine and Kelly Close