I don't understand the math of diabetes and pre-diabetes

I understand certain glucose and A1C readings indicate pre-diabetes vs diabetes. However I have read stats like ‘within 3-5 years, 25% of people with pre-diabetes develop diabetes’, or that after 10 years over half of people with pre-diabetes have diabetes.

But in the US about 35% of adults over 20 have pre-diabetes (almost 80 million), while only about 11% of people age 20+ have diabetes. In China the stats are something like 12% are diabetic but 50% are pre-diabetic.

I’m sure genetic factors and lifestyle play a huge role in how long before pre-diabetes becomes diabetes (if ever). But how do you get stats like 'half will develop diabetes in 10ish years) when the number of prediabetics is triple the number of diabetics? if that were true, then there would be 40 million new diabetics in 10 years, tripling the US diabetic population.

Even the ‘within 3-5 years 25% with pre-diabetes will develop diabetes’ means that there will be 20 million new diabetics within 3-5 years, doubling the number in the US. That doesn’t sound right.

Ten years ago there weren’t as many prediabetics. Ten years from now there will be half again as many diabetics as there are now. (3x prediabetics/2)

That’s assuming the trend continues.

How will there be half as many diabetics in 10 years? Did you mean prediabetics? The info I’ve read implies there will be an additional 40+ million diabetics in 10 years, tripling the number of diabetics in the US (assuming none of the 20 million current diabetics die, which isn’t likely but the majority should be alive in 10 years).

This chart has the rates of diagnosed diabetes growing drastically from about 6 million in the mid 80s, to 12 million in 2000 to 22 million today.

But there are several trends that would explain that

  1. The diagnosis of diabetes is lower now, it used to be fasting glucose of 140, now it is 126. That added millions of people.
  2. More people are diagnosed compared to in the past (due to things like awareness, cheaper/more prevalent testing, etc)
  3. The US population has grown by 30% since the 80s.
  4. We are older than we were as a nation than in the past, and diabetes gets worse with age.
  5. We are a less white, more latino/black nation than in the past and blacks/latinos get diabetes at higher rates.

All in all, if that number jumps from 20 million to 60 million in 10 years that would be completely out of the current trends, seeing how the number only went from 13 million to 21 million from 2001-2011.

(S)he said “half again.” 1.5 times. 50% more. Not half as many. Which is probably an underestimate.

Increasing numbers of new prediabetics will be (and likely has been) a leading indicator for increasing numbers of diabetics years down the road. Hence in ten years there will be an additional 40 million (minus those who currently have diabetes but have died) diabetics added to the roughly 25 million now labelled. Ten years ago there were likely fairly few prediabetics relative to today’s numbers.

Funny you do not mention that we are much fatter and less fit and at earlier ages than we were before too. The other points are valid but the tightest correlation is with obesity rates.

Also note that case finding and identification of pre-diabetics has improved as well. As has the definition. Not sure if your cited numbers use the older or the newer definition of prediabetes aka impaired fasting glucose or IFG. One study that these rates come from is this one.

A similar result in a European population using the oral glucose tolerance test as the measure for prediabetes.

Note we are also seeing increasing numbers of both prediabetes and T2 diabetes in kids.

For those playing along at home. Ignore the last sentence, since that’s what the smart dudes here are talking about…

Impaired fasting glucose (IFG), more commonly known as pre-diabetes refers to a condition in which the fasting blood glucose level is consistently elevated above what is considered normal levels; however, it is not high enough to be diagnosed as diabetes mellitus.[1] This pre-diabetic state is associated with insulin resistance and increased risk of cardiovascular pathology, although of lesser risk than impaired glucose tolerance (IGT). IFG can progress to type 2 diabetes mellitus if lifestyle changes are not made. There is a 50% risk over 10 years of progressing to overt diabetes. A recent study cited the average time for progression as less than three years.[2]

Wiki cite.

I found this article from 2004.


So if 9 years ago about 40% of people age 40-74 had prediabetes, shouldn’t half of them be diabetic by now? This is just people 40-74, not people 20-39 or 75+ which would likely add another 10-20 million people to the ranks of prediabetics in 2004. If true, we should’ve added 20-30 million diabetics by now. But the rates have gone up by about 10 million and part of that is just better diagnosis (people who have it know they have it) and population growth.

Here is another thing. I haven’t read most of the article associated with this image.


But it shows rates of normal glucose, pre-diabetes and diabetes by age bracket. I do not see the rates of new diabetes ever being anywhere near half the rates of prediabetes among people a decade younger.

Example, among 40-49 year olds the rates of pre-diabetes are 42% with diabetes rates of 9-10%, among 50-59% the rates of diabetes are 13%. So the rate of diabetes only went up about 3-4% among people 10 years older despite 42% of people being prediabetic in their 40s. They aren’t the same people, but the rates of diabetes should be higher according to that, rates of pre-diabetes are 30-50% after the 30s, but even by a person’s 70s the rates of diabetes only tops out at about 30% and is only about 15% in the 50s. It makes no sense.

A couple of comments Wesley -

  1. The flaw in the repeated cite of “that after 10 years over half of people with pre-diabetes have diabetes” is evident in the cite I gave you. It is possibly true for the older, pre-2003, definition of prediabetes, fast blood glucose of 110 to 125. For that definition the annual rate was 5.56% of prediabetics becoming diabetics. Do the math and it comes to roughly about half becoming diabetic in 10 years. But the definition changed in 2003 to also include those with a fasting glucose of 100 to 109. Suddenly lots more people are prediabetic, and indeed they are at increased risk to become diabetic, but not at as much increased risk. For all included in the new definition the annual rate of becoming diabetic was under 2%. That 2004 estimate used the new to 2003 definition which includes many in the 100 to 109 group; that 40% of the population 40 to 74 is at about a 2% annual risk, not a nearly 6% annual risk. The numbers that were prediabetic by the pre-2003 impaired fasting glucose definition were, in those 45 to 74 and overweight, in 2000, 11.9%, up to 22.6% adding in those who had abnormal oral glucose challenge with normal fasting glucose. Half of that 12% probably became diabetic over the next 10 years, other than those who changed lifestyles, or died.

  2. Those aged 40 to 74 who became diabetic are also at greater risk to have died off over the next years than those who did not and not be found in the living diabetic group.

  3. Those original “half will become diabetic” numbers also were based without intervention. With identification and with consequent changes in nutrition and exercise habits the numbers can be impacted dramatically. That is increasingly occurring.

  4. This article also may be of interest. My impression before your question inspired me to look for it was that was a true rise in prediabetes using the same definition of the term but such may actually not be the case. Which shocks me. Bottom lines:

**For the too long did not reads:

In those who meet the old pre-2003 definition of prediabetes half will likely become diabetic unless they change lifestyle or die first. Many fewer under the current broader definition and less if they do something about it.**