top of page

The Patterns in which Type 2 Diabetes Develops


If you've been diagnosed with prediabetes or impaired glucose tolerance, or if your fasting blood sugar is higher than normal, or if you have a family history of diabetes, you may wonder what your real risk is of developing full-fledged diabetes. Several high quality medical studies have answered this question. These were studies that looked at what happened to large numbers of people over long periods of time and then reported about how many of them became diabetic and how their progression to diabetes unfolded.


The Baltimore Longitudinal Study of Aging: Decades of Observation

One of the most thorough studies of a large population ever conducted is the Baltimore Longitudinal Study of Aging (BLSA).This is an ongoing study that was begun by the National Institutes of Health in 1958. It's goal is to track the health experience of volunteers from Baltimore and the Washington, DC area through many decades. The people who participate were primarily white, middle- and upper-middle socioeconomic class volunteers who returned to the Gerontology Research Center in Baltimore every two years for an exam.


The study is what is called an "open cohort" design study, which means that any people who drop out of the study replaced to maintain equal numbers of subjects in each 10-year age group. About 1,000 volunteers have been examined at each study cycle.


Researchers Tested Glucose Tolerance and Fasting Glucose over at Least Eight Years


For their analysis of what happens to blood sugars over time, Meigs and his team included only subjects who had attended at least three examination and had had an Oral Glucose Tolerance Tests (OGTT) within an 8 year period. Participants were excluded if they had less than 2 OGTTs or if more than four years had elapsed between two OGTTs.


The methodology used to administer the OGTT in this study was a bit different from the usual way that glucose tolerance tests are done, in that the amount of glucose administered was adjusted to match the participant's body size. The average dose of glucose given to men was 78 g and to women 68 g.


BLSA Results: 52% of Normals Developed Abnormal Blood Sugar and 11% Developed Diabetes


When the study of their blood sugar started, participants' average age was 57. At that point, 60% of them had normal glucose tolerance as measured by the OGTT. Normal was defined using the American Diabetes Association definition, which calls considers a OGTT blood sugar test result normal if two hours after consuming glucose it is under 140 mg/dl.


The researchers reported that of the 437 people who started out with normal glucose tolerance whom they were able to track for roughly a decade, 48% remained normal. Of the rest, 52% developed abnormal blood sugars during the course of the study.


Breaking down exactly what kind of dysfunction the previously normal participants developed, the researchers found that by the end roughly a decade:


  • 31% of the original group of 437 participants had impaired glucose tolerance (defined as having a two hour OGTT test result> 140 mg/dl) though they still had normal fasting blood glucose levels under 110 mg/dl.


  • A smaller subgroup, 3% of those studied, had developed both impaired glucose tolerance and impaired fasting glucose.


  • 5% had developed impaired fasting glucose but still had normal glucose tolerance test results two hours after consuming a big slug of glucose.


  • 11% of those who had been considered normal a decade before developed full-fledged diabetes, defined as having a two hour OGTT result greater than 200 mg/dl or a fasting plasma glucose greater than 126 mg/dl.



BLSA Shows Abnormally High Post-Challenge Blood Sugar with Normal Fasting Glucose is The Most Common Pattern for Those Developing Type 2 Diabetes


"Post challenge" simply means "after taking in a big dose of glucose all at once, as happens during the OGTT. Normal high carbohydrate meals are also "challenges" as the carbohydrates digest down to glucose which usually enters the blood stream all at once. Impaired Glucose Tolerance, the term used in the BLSA study, has been replaced more recently by the term prediabetes.


The BLSA data answered the question of which was more common, developing high fasting blood sugar first or high post-challenge numbers.


In the group that went on to develop diabetes, 225 people initially developed abnormal two hour glucose tolerance test results (Impaired Glucose Tolerance) while maintaining normal Fasting Plasma Glucose.


Of those who eventually developed diabetes, only 30 participants developed abnormal Fasting Plasma Glucose while still maintaining normal blood sugars on their glucose tolerance tests.


When these groups of participants were followed over time, it was found that 37% of those with abnormal fasting glucose went on to develop abnormal post-challenge glucose tolerance while only 15% of those originally diagnosed with Impaired Glucose Tolerance (i.e. post-challenge glucose) went on to develop Impaired Fasting Glucose.


This should make it very clear how bad a bad test the fasting glucose test is for identifying people at risk of future diabetes. Yet, many doctors still rely on fasting glucose to screen for diabetes among people who have no other signs their blood sugar might be deteriorating.


Just how poor a screening test the fasting glucose test is is shown by the fact that 67% of the people who progressed to diabetes as measured by the glucose tolerance test did so without ever developing impaired fasting glucose.

The True Risk of Developing Diabetes Shown by BLSA Data


The Baltimore Longitudinal Study of Aging data data suggest that a person in their fifties who has a normal blood sugar test result has roughly a 1 in 8 chance of becoming diabetic over the next decade. Of those whose blood tests show they already have some form of impairment--prediabetic post-challenge numbers or elevated fasting glucose, the risks are higher. A person who has impaired glucose tolerance has a 4 in 10 chance of progressing to diabetes over a decade while a person with impaired fasting glucose has almost a 1 in 2 chance of progressing to diabetes.


Who Progresses?
People over 56 Years Old, Males, Overweight People

When looking at exactly who the unlucky people were who progressed to diabetes, Meigs' team found that people older than 56 years, "had substantially accelerated rates of progression to abnormal 2hPG [two hour OGTT result] compared with younger subjects" but that older and younger subjects had a similar rate of progression to abnormal FPG [Fasting plasma glucose].


They also found that "Men progressed to abnormal FPG or 2hPG more rapidly than women, as did subjects with overall or central obesity compared with lean subjects."


Contrary to expectations (and the findings of other studies), the researchers found that a family history of diabetes did not modify the rates of progression to abnormal glucose tolerance.

How Many Reverted Back to Normal after an Abnormal Test?

The researchers also examined how many subjects who tested abnormally on one test reverted back to normal on a subsequent test. They found that 30% of those showing impaired fasting plasma glucose values testing on a single test remained abnormal on a test given 2 years later. Of those testing as having impaired glucose tolerance with a two hour OGTT on a single test, 48% remained abnormal on a subsequent test. (Remember, however that all those who were eventually classed as progressing from normal to impaired or impaired to diabetic remained impaired on subsequent tests.)


Another Study also Shows that Diabetes Starting with Impaired Fasting Glucose Differs from Diabetes Starting with Impaired Glucose Tolerance

The Inter99 Study was a five year study published in 2008. in it, 3,145 subjects who started out with normal glucose tolerance but developed some form of abnormal blood sugar who were given glucose tolerance tests and several more sophisticated tests that examined how sensitive they were to insulin and how much insulin they were secreting.


The researchers conclude,


A stationary reduced insulin secretion followed by a decline in primarily hepatic insulin sensitivity characterizes the transition from N[ormal] G[lucose] T[olerance] to i-I[impaired]F[asting]G[lucose]. In contrast, low whole-body insulin sensitivity with a secondary lack of ß-cell compensation is associated with the development of i-I[mpaired]G[lucose]T[olerance]. Thereby, i-IFG and i-IGT appear to result from different underlying mechanisms, which may have implications for the prevention and treatment of the diabetes that succeeds them.


Natural History of Insulin Sensitivity and Insulin Secretion in the Progression From Normal Glucose Tolerance to Impaired Fasting Glycemia and Impaired Glucose Tolerance: The Inter99 Study. Kristine Færch, et al. Diabetes Care, 32:439-444, 2009.


Another Recent Study Shows That Progression to Diabetes is Not Gradual

Another study which looked more closely at the way in which diabetes develops looked not at the statistics for the group as a whole as did the study discussed above, but instead looked at how the blood sugar of individuals in the study changed over time. (2)


This team, led by Dr. Ele Ferrannini of The University of Pisa School of Medicine in Pisa, Italy, examined a population of 2,279 low-income people aged 35 to 64 in Mexico City who were known to have a high risk of developing diabetes. The subjects were tested three times over a seven year period beginning in 1990. The researchers measured their fasting plasma glucose and their fasting insulin levels. They then sampled the participants' blood glucose and insulin levels at 2 hours after the administration of a 75 gram dose of glucose.


Ferrannini's team found that of 1074 subjects followed over the whole seven years, 90 subjects (8% of the total) converted from normal glucose tolerance to diabetes in the 3 years between one examination and the next, and 73 (7%) converted from impaired glucose tolerance to diabetes over the same 3 year period between exams.


Finding: A Swift and Unexpected Deterioration in Blood Sugar Control Precedes Diabetes


Rather than being a gradual process, Ferrannini's team found that the transition to diabetes appeared to occur very quickly within a 3 year period, and was characterized by a swift increase in fasting plasma glucose values. While the fasting plasma glucose of those who did not become diabetic increased "slightly and in an apparently linear manner" that of the people who became diabetic took a sudden step up, showing an average gain in fasting plasma glucose of 50 mg/dl between one examination and another about 3 years later.


The two hour glucose tolerance test results showed a similar pattern. The people who did not become diabetic showed a "slight increase" while those who became diabetic saw an average surge of 108 mg/dl between one exam and another 3 years later.


That this change was not gradual was highlighted by the finding that when researchers looked at the test results of the people who went from normal to diabetic between their 3 year and 7 year exams, they found that the changes in their blood sugar test results between the exams at years 1 and 3 of the study were the same as those of people who remained normal throughout the whole study.


Which Individuals Deteriorated?

The two factors that the researchers found correlated to a likelihood of going from normal to diabetic was Body Mass Index and high fasting insulin levels.


Twenty-five percent of the subjects whose BMI was above the median of the group became diabetic compared to 8% of those whose BMI was below the median.


The other factor that increased the likelihood of becoming diabetic was having a high fasting insulin level--a sign of insulin resistance. Twenty-five percent of those with high fasting insulin progressed to diabetes vs 9% of those with normal insulin levels.


The "Normal" People in This Study who Deteriorated Started with Significantly Higher Blood Sugars.

What makes this study even more interesting is that it includes some vital information that Dr. Meigs and his team left out when reporting their findings with the BLSA data. Rather than just telling us that the study subjects fell into the various ADA classifications like "normal,"IGT" or "IFG" this study gives us more detailed information about the actual fasting and 2 hour OGTT test result values of the group of people who became diabetic and of those who did not.


The difference between the "normal" blood sugar test results of these two groups of "normals" is striking.


At the outset of the study the average fasting plasma glucose of the people who remained normal was 82 mg/dl with a narrow standard deviation reaching up to 92 mg/dl. The average fasting plasma glucose of the "normal" people who went on to develop diabetes was 10% higher--at 90 mg/dl but however, the standard deviation of this group's fasting blood sugar extended all the way up to 139 mg/dl, the pre-1998 ADA cut off for "normal." (The standard deviation is a measure of how tightly all the values in a group of results cluster around the average.)


So, this tells us that the supposedly "normal" people in this study, as a group, already had significantly higher fasting blood sugars than the people who remained normal three years later.


In a similar manner at the outset of the study, the two hour OGTT values of the people who remained normal averaged 93 mg/dl and one standard deviation ranged up to 116 mg/dl. But the average two hour OGTT result of the people who went from normal to diabetic was 22% higher than the people who remained normal. Their average two hour reading was 113 mg/dl, and but as was the case with the fasting results, the standard deviation of the OGTT two hour test results of the "normal" people who became diabetic was much wider--reaching up slightly above to the top ADA cutoff for normal at 143 mg/dl.

Many People Who Developed Diabetes Started with Normal Fasting Blood Glucose Tests

At the start of the study, the fasting blood sugars of the patients classified as having impaired glucose tolerance who also became diabetic were not significantly higher than those of the "normals". Their average fasting plasma glucose was 96 mg/dl, only 7% higher than that of the normals who became diabetic. The real tell was their post-challenge test result: Their average 2 two hour OGTT value was 153 mg/dl.


A 2007 Study Confirms Progression Within 3 Years to Diabetes with Fasting blood sugar over 100 mg/dl (5.6 mmol/L)

Another This was a study of 5,452 members of an HMO with no prior history of diabetes. You can read the full text here: Progression From Newly Acquired Impaired Fasting Glucose to Type 2 Diabetes Gregory A. Nichols et al Diabetes Care. 2007; 30(2):228-33


This study analyzed two groups, the "original IFG" group were diagnosed as having abnormal fasting glucose by the American Diabetes Associations old diagnostic standard which defined a higher cutoff for normal fasting glucose. The "added IFG subjects" were those diagnosed using the lower cutoff introduced by the ADA in 1998. The researchers reported:


Overall, 8.1% of subjects whose initial abnormal fasting glucose was 100-109 mg/dl (added IFG [impaired fasting glucose] subjects) and 24.3% of subjects whose initial abnormal fasting glucose was 110-125 mg/dl (original IFG subjects) developed diabetes (P < 0.0001). Added IFG subjects who progressed to diabetes did so within a mean of 41.4 months, a rate of 1.34% per year. Original IFG subjects converted at a rate of 5.56% per year after an average of 29.0 months. A steeper rate of increasing fasting glucose; higher BMI, blood pressure, and triglycerides; and lower HDL cholesterol predicted diabetes development. [emphasis mine]


What this means is that if your blood sugar tests over 110 mg/dl fasting more than once, your fasting blood sugar is much more likely to go over the 125 mg/dl (7.0 mmol/L) level used to diagnose full diabetes within 3 years.


More importantly, and not addressed in this article, if your blood sugar is over 100 mg/dl fasting (diagnosed as "impaired fasting glucose" it is very likely that your post-meal blood sugar is not only prediabetic but may be much closer to reaching the diabetic range. Remember how we saw from the BLSA data that most people's post-challenge blood sugars deteriorate before their fasting blood sugars move out of the normal range?


This should not be a surprise. High post-meal blood sugars kill beta cells. If you can bring down those post-meal highs, you may be able to prevent the beta cell death that is destroying your fasting blood sugar control! It is high post-meal blood sugars, rather than high fasting blood sugars that appear to damage our organs and worsen diabetes.


Truly Normal is Clearly Lower than American Diabetes Association "Normal"

If you worry about your liklihood of developing diabetes, the data we have just examined should have convinced you that it is a mistake to wait until the doctor diagnoses you using the American Diabetes Association criteria. It is even more dangerous to let a doctor reassure you that he sees no signs of diabetes if the only test he has run is a fasting plasma glucose test. Insist that your doctor tell you what test he uses to screen you. If he tells you that he used the A1c test--which increasing numbers of doctors are doing, ask him what the actual test result was. Any A1c test result over 5.6% suggests that your blood sugar has entered the prediabetic level, though many doctors will ignore an A1c result until it is over 6.5%.


Does This Mean YOU Will Deteriorate?

By now you're probably wondering how you can tell whether any mild blood sugar abnormality you have is something to worry about. Were you one of the lucky ones who will "revert" back to normal, or are you on the way to developing diabetes?


The answer has a lot to do with how close to abnormal your "normal" results might be. The study subject who is considered by researchers as having "reverted" from abnormal glucose tolerance to normal glucose tolerance may have tested into the "impaired" category with a value of 141 mg/dl on the first glucose tolerance test and tested "normal" with a value of 139 mg/dl on the second. But as far as that person's health is concerned, the difference is insignificant. Unfortunately, if dozens of study subjects make that kind of change, it makes for a big difference in the researchers' charts and graphs. This is the problem with all-or-nothing cutoffs.

Factors that Cause Temporary Rises in Blood Glucose Test Results

There are a lot of factors that the researchers don't check for that can push your blood sugar up or down 10 or 20 mg/dl. For example, if your body is fighting off a cold or flu, long before you start to sniffle, your blood sugar will rise perhaps by 20 mg/dl or more. This additional glucose is used by the immune system as it tools up to fight the invader.


Another factor that researchers miss is that women often see a significant boost in their blood sugars during certain parts of the menstrual cycle. Birth control pills and hormone replacement can also push their blood sugars up or down. And these hormonal changes may cause blood sugars to rise or fall by a significant amount--10 mg/dl for fasting blood sugar and 20 or 30 mg/dl in a post-meal value. So it is possible that some of the subjects in population studies who revert to normal from impaired may have been tested during a high blood sugar phase of their menstrual cycle. Remember, a rise of only 5 to 10 mg/dl could push you from the normal category into the impaired category if you are already near the cutoff point.


Another factor that can push your blood sugars up or down, particularly if you are closer to impaired than to truly normal, is what you've been eating over the week before your test. Paradoxically, eating a very low carbohydrate diet or a very high carbohydrate diet can skew your results higher.


Medications can also push your blood sugars up and down enough to push you from one category to another. Sulfa antibiotics like Septra, for example, lower blood sugars, while cortisones and some mood-changing drugs can push it up.


But in all these situations, we're talking about a small difference, the difference between a post meal test blood sugar concentration of 145 and 135 mg/dl, not the difference between 145 and 87. And though this small difference doesn't really mean that much has changed in your health, it will move you from one category to the next when researchers use hard cutoffs.

How Can You Get A Better Idea of Where You Stand?


You are very unlikely to be offered a two hour glucose tolerance test by most doctors. The test is expensive and rarely administered nowadays except to pregnant women. But you can do a home "meal tolerance test" which will show you how your blood sugar responds to the carbohydrate you eat in your daily meals. The carbs in your food show up in your blood sugar more slowly than the glucose used in a glucose tolerance test, but you can learn a great deal about your blood sugar health by measuring how high your blood sugar is one and two hours after eating a high carbohydrate meal.


You can learn exactly how to test your blood sugar at home to determine its current status HERE.


You Can Stop Progression from Prediabetes by Cutting Way Back on your Carbohydrates

The very good news is that there is no reason why your blood sugar has to get worse if you do get a prediabetes diagnosis or observe abnormal post-meal readings. Every single subject in the studies described above was eating a standard, very high carbohydrate diet providing on average 300 grams of carbohydrate a day. By cutting back on your carbohydrate intake significantly you can often normalize your post-meal blood sugars.


When you do that, your fasting blood sugars usually come down too. Though there are some forms of diabetes that don't self-correct with simple dietary changes, the most common form of Type 2 does. More importantly, the earlier you notice that your blood sugars are abnormal the better chance you have of keeping them normal for the rest of your life. Keeping your blood sugars post-meal at normal levels does stop the progression of Type 2 diabetes. You can learn a simple way to find the diet that will normalize your blood sugars on this web page HERE.




1The Natural History of Progression From Normal Glucose Tolerance to Type 2 Diabetes in the Baltimore Longitudinal Study of Aging. James B. Meigs, Denis C. Muller, David M. Nathan, Deirdre R. Blake, and Reubin Andres; Diabetes 52: 1475-1484. 2003


2 Mode of onset of type 2 diabetes from normal or impaired glucose tolerance. Ele Ferrannini, Monica Nannipieri, Ken Williams, Clicerio Gonzales, Steve M. Haffner, Michael P. Stern.Diabetes 53:160-165, 2004


310 Year CVD Risk


4 Breast Cancer Risk Assessment Tool National Cancer Institute.

Baltimore Study of Aging Diabetes Development Patterns table
bottom of page