June 26, 2009

Clear Indicators of Diabetes Three Years Before Formal Diagnoses

This week the Lancet published an epidemiological study which asked the question, "Can we see clear cut signs that a person is on the way to developing diabetes in the years before diagnosis?"

The answer was of course, "Yes."

The abstract of this article is here:

Trajectories of glycaemia, insulin sensitivity, and insulin secretion before diagnosis of type 2 diabetes: an analysis from the Whitehall II study.
Dr Adam G Tabák et al. The Lancet The Lancet, Volume 373, Issue 9682, Pages 2215 - 2221, 27 June 2009 doi:10.1016/S0140-6736(09)60619-X

This was a prospective study--i.e. one that takes measurements and then follows the subjects for a set period of time--in this case, almost 10 years, to see what happens. The subjects were 6,538 British civil servants, 71% of them males.

Diabetes was diagnosed using the most conservative measure: subjects had to have a 2 hour glucose tolerance test result over 200 mg/dl. This standard might have missed people who were going over 200 but who were below 200 mg/dl at 2 hours post challenge. Though according to the ADA official diagnostic criteria for Diabetes Mellitus anyone who scores over 200 mg/dl more than once should be considered diabetic. But the use of the 2 hour glucose tolerance test criteria tells us the people in this study who were diagnosed as diabetic were seriously diabetic.

What the researchers found was this: In the subjects who did not become diabetic, "metabolic measures followed linear trends in the group of non-diabetics ... except for insulin secretion that did not change during follow-up.

This finding is important, so let's repeat it: The normal people in this study saw their blood sugar creep up as they got older, but the researchers found that normal people's ability to secrete insulin did not change over time.

In the people who became diabetic,
... a linear increase in fasting glucose was followed by a steep quadratic increase (from 5·79 mmol/L [104 mg/dl] to 7·40 mmol/L [133 mg/dl]) starting 3 years before diagnosis of diabetes.
"Steep quadratic increase" means that their fasting glucose did not go up gradually, but shot up in a pattern that, when graphed, made a steep upwardly curving parabola.

In addition their 2 hour glucose tolerance test result showed a rapid increase starting 3 years before diagnosis from 7·60 mmol/L [137 mg/dl) to 11·90 mmol/L [214 mg/dl]).

Since we know that lots of people have pre-diabetic blood sugars that don't deteriorate, including the overwhelming majority of insulin resistant obese people who never develop diabetes, it's important to note that it is failing insulin secretion, not insulin resistance, that appears to sort out the people who go on to become diabetic.

Though insulin resistance rose starting five years before diagnosis, in the people who became diabetic,
HOMA β-cell function increased between years 4 and 3 before diagnosis (from 85·0% to 92·6%) and then decreased until diagnosis (to 62·4%).
These findings make one thing crystal clear. If you wait for an official diabetes diagnosis--one based on either a fasting glucose test or a two hour glucose tolerance test, you will be giving high blood sugars three years in which to ravage your body.

Consider all the people in this study who were classed as "nondiabetic" because their 2 hour glucose tolerance test result was "only" 180 mg/dl. How high do you think their blood sugars were at 1 hour?

And given that we know that blood sugars that go over 180 mg/dl increase insulin resistance, you have to wonder whether the insulin resistance that was observed to increase five years before diagnosis was due to some independent disease process or whether it was a direct result of the "nondiabetic" blood sugars these people were experiencing after each meal that exposed them to hours of damaging glucose toxicity as their blood sugars remained above 140 mg/dl.

Because, of course, prolonged exposure to blood sugars over 140 mg/dl kills beta cells. Which makes you wonder if the decrease in insulin secretion observed three years before diagnosis, that decrease which happened two years after the high IR was first observed, might have been caused by the subject's blood sugars going over 180 mg/dl meal after meal. Was the decrease in insulin secretion the final result of exposure to "nondiabetic" but much higher than normal blood sugars?

The study I'd like to see would be one that asked: what happens when you find people whose 2 hour glucose tolerance test result is over 140 mg/dl--high enough to damage their organs, and treat them with a protocol that lowers their post-prandial blood sugars by having them cut down on their carbohydrate intake?

Do you see insulin resistance drop when they stop experiencing those 1 hour 180 mg/dl and higher blood sugars after each meal?

Can they preserve their ability to secrete insulin because they aren't exposing their beta cells to those blood sugars over the 140 mg/dl level that we know produces organ damage?

Doctor's aren't likely to do this study, because there are so many medical-political forces fighting against letting the public know that the cheap carbohydrates that enrich the food industry are what raise blood sugars and ruin health. So we aren't likely to get a definitive answer to the question of whether normalizing blood sugars with dietary change can stop the progress to diabetes or not.

But if you have been diagnosed as prediabetic you have nothing to lose by making your own personal experiment to see if keeping your blood sugars in the normal range will prevent your own diabetes from progressing.

You can find out exactly how to do this on this page: How To Get Your Blood Sugar Under Control

Even if it turns out that this technique cannot prevent further progression of diabetes0though it is quite possible it CAN--you have nothing to lose by trying it.

Many of us who have already been diagnosed as diabetic have found that keeping our blood sugars in the normal range prevents us from developing diabetic complications. Since the reason we don't want to have diabetes is because we don't want diabetic complications, prediabetes who adopt a rigorous approach to normalizing their blood sugar with carb restriction will come out way ahead in the future, whatever their glucose tolerance test result might be.

Remember: A diagnosis of Diabetes doesn't cause complications. Prolonged exposure to high blood sugars cause complications. Whatever your diagnosis, if you keep your blood sugars in the normal range 70 mg/dl to 120 mg/dl after meals and under 100 mg/dl fasting you can have normal health.


Anonymous said...

Even if keeping your sugars down doesn't stop progression, it does at least stop the lethargy and the mid afternoon dips.

Stargazey said...

Jenny, I don't know if this is the proper place to put my question, but here it is.

I've been following the forum Zeroing In On Health for a few months. Much of what they say seems to have merit. But a couple of days ago I realized that a typical fasting blood sugar for a zero carber is anywhere from 95 to 120. I did a Googlesearch and found that after a year of eating only meat and fat, Karsen Anderson and Vilhjalmur Stefansson had fasting blood sugars of 110 and 105 respectively. That means that their blood sugars essentially never went below 100. This link is a reference to the article.

I know that zero carbing is a unique situation, but since you have quite a bit of experience in evaluating blood sugars, can you give me some idea of what you think would be the long-term outcome of a diet consisting of only fat and meat if it meant that fasting blood sugars stayed above 100?

Jenny said...

I think a long term nothing but meat and fat diet is ill-advised.

The traditional arctic peoples eat not only meat, they eat all the stomach contents of their prey and all their organs. They are also adapted to this diet by many generations of evolution that eliminated children who could not survive on it.

Anthropologists have learned that most traditional "hunter" peoples world wide eat a large percentage of their food in the form of what is gathered by their women folk who go out daily looking for roots, tubers, edible leaves, grasses and fruits. The men hang around bragging about what they killed, but the meat is not the main food of these peoples.

The explorers who are always cited by the zero carb people were early in the 20th century so their blood sugar measurement may not be all that exact, given the older technology.

There is a lot of silliness written by doctors about imagined "paleolithic diets" which would make any trained anthropologist wince. Did I mention I have a degree in Anthropology and spent hours in my youth reading first hand reports of field studies with pre-agricultural peoples, collected in the early years of the 20th century?

Stargazey said...

Thanks for your response, Jenny. Could you give me some insight about the long-term effect of the persistent higher-than-100 blood sugars these people seem to have?

Jenny said...

Damage appears to be done to organs when blood sugars spend significant amounts of time over 140 mg/dl.

The reason rising fasting blood sugars are problematic is that they usually reflect rising post meal blood sugars. If the highest blood sugar a person ever saw was 110 mg/dl they'd probably be fine.

My blood sugar will rarely drop below 95 and was at 108 for many years due to my genetic diabetes. I don't have any traditional diabetic complications.

Stargazey said...

Thanks, Jenny. That's good to know.

Anna said...


I also have a somewhat higher FBG than one would expect, rarely lower than 95 and often 100 and up into the teens, esp if I had some dessert the night before (though that previous night's post prandial wasn't too high).

I don't eat only meat and fat. But I do eat a lot more than average. I nearly always have natural fat and high quality animal protein at nearly every meal - so if not meat, then eggs and/or cheese. Virtually no grains or beans and no wheat or soy at all anymore. I don't eat much fruit, though in summer I have a bit more (and when I do eat fruit, it tends to be moderate amounts, whole fruit, and lower sugar varieties like berries, and northern climate fruits (not high sugar tropical fruits).

I also eat a fair amount of non-starchy veggies (as salads (Iget a big CSA veggie box each week). On average my dinner plate is 2/3 veggies, 1/3 meat w/fat, rarely a very small serving of starch like rice, sweet potato, or winter squash.

The small amount of concentrated sugar I consume is in very dark chocolate, homemade ice cream and custards, and/or some coconut flour (gluten-free) baked items. My post meal BG is rarely over 115 unless I overindulge or miscalculate when eating away from home.

Peter's hyperlipid blog had some posts about the higher FBG of people who eat this way (we are such a small minority we don't register in the research).

Stargazey said...

Thanks for the Hyperlipid reference, Anna!

The discussion of physiological insulin resistance over at Peter's blog was fascinating. It was a bit hard to tease apart the insulin resistance measured on an actual glucose tolerance test (where low-carbers lack the enzymes) and the insulin resistance from muscles preferentially using fatty acids.

In any case, I wonder if the higher blood glucose might result from gluconeogenesis of excess protein. I've been doing zero carb for a few days as a personal experiment. When I cut my protein intake back by 25% (fat was unchanged), my hourly blood glucose values went down by about 10 mg/dL.