July 3, 2008

A Massive Long Term Study Sheds More Light on Truly Normal Fasting Glucose Values

A study brought to my attention by this week's edition of Diabetes in Control newsletter makes it very clear that for people with symptoms of the metabolic syndrome a fasting glucose blood sugar test result of 95 mg/dl or higher should impel their physicians to order a glucose tolerance test.

What the study found in brief is this: Over a 9 year period, within a group of 46,578 members of Kaiser Permanente Northwest, 10% of those who started out with fasting plasma glucose values of 95 to 100 mg/dl ended up diagnosed as diabetic based on a fasting plasma glucose test result greater than 125 mg/dl.

In contrast, only about 3% of those with fasting blood sugars below either 85 mg/dl or 89 mg/dl were diagnosed with diabetes.

Within the group which became diabetic, other factors which increased the likelihood of becoming diabetic, in order of impact, were diagnosed cardiovascular disease, high blood pressure, smoking, high triglycerides, and elevated BMI.

Significantly, there was NO significant correlation between the levels of LDL or HDL cholesterol and the likelihood of becoming diabetic. ONLY the trigyclerides--which are a good reflection of the blood sugar level after meals, predicted oncoming diabetes.

The complete text of the study is free, which is a nice change. You can read it here:

Normal Fasting Plasma Glucose and Risk of Type 2 Diabetes Diagnosis
Gregory A. Nichols, Ph.D. et. a.. The American Journal of Medicine. Vol 121,issue 6, 519-524 (June 2008)

Note in particular this graph: Kaplan-Meier plot of cumulative diabetes incidence by category of normal fasting plasma glucose.

In their conclusions the authors point out that the lack of glucose tolerance test data for this population limited the value of the study and suggest that those with fasting glucose values of 95 mg/dl and higher most probably would have tested at the prediabetic level on a glucose tolerance test. They also point out that "Among those who developed diabetes by our criteria, however, the mean hemoglobin A1c at diagnosis was more than 7%, a level that strongly suggests that abnormal glucose metabolism has been maintained for several months."

My guess is that the over 7% A1cs suggest that the glucose metabolism had been abnormal for several years--probably beginning when that fasting glucose went over 94 mg/dl.

The group in this study whose fasting blood glucose was between 90 and 94 mg/dl had an incidence of diabetes that was about 5%.

What does this mean for you if your fasting glucose is in the 90s?

It means you need to get yourself a meter and to start testing your post-meal blood sugars, one and two hours after eating, to see how high your blood sugar is rising. You don't have to do this very often. Once or twice a year is all you need to do assuming you do not see post-meal blood sugars over 125 mg/dl which seems to be the peak most truly normal people attain, very briefly, before their blood sugar drops back to their fasting level.

If you see your blood sugar rising over 140 mg/dl after meals, take it as a sign that you are very likely to have prediabetes and start taking steps to improve your blood sugar health now, when it is still relatively easy to reverse any early diabetic changes in your body and preserve your beta cells from harm.

If your fasting blood sugar is in the 90s here are some steps you can take to improve your blood sugar health.

1. Cut the carbs. Carbs are what raise blood sugar and you are probably eating a lot of junk carbs that are stressing your body and pushing you towards developing diabetes. Try cutting out the following: All non-diet sodas, fries, white bread, breakfast cereals containing more than 10 grams of carbs per serving (most people eat two or three "servings" every time they fill a bowl with cereal, large muffins (6-8 oz), large servings of pasta, etc. Cutting carbs will lower your triglycerides, the only part of your cholesterol linked with developing diabetes.

2. Exercise. If you aren't the gym rat kind, start taking a 40 minute walk four or five times a week. That has been shown to be enough to make significant improvements in your fitness without causing injury.

3. Check your meds. Many commonly prescribed medications have a side effect of causing "hyperglycemia" i.e. high blood sugar. Read the official Prescribing Information for all medications you are taking and see which ones might be contributing to your rising blood sugars. Some drugs known to raise blood sugar are HCTZ, SSRI antidepressants, Zyprexa, prednisone and other corticosteroids. There are others. If you are taking a medication that raises blood sugar, talk to your doctor about whether there are alternatives which won't put you at risk of kidney failure, blindness, increasing heart disease, and amputation--which are what can happen to you over time if you allow your blood sugar to rise unchecked.


ItsTheWooo said...

I think the best way to test for diabetes is to use the A1C - the A1C can most reliably predict blood sugar metabolism problems that cause consistent hyperglycemia.

Everything else merely suggests hyperglycemia... the A1C does not discriminate against any blood sugar control problem.

When I look back at all my different health states, my A1C most reliably reflects how good my blood sugar control probably was. Triglycerides and fasting sugar did not. When I was fat but losing weight I STILL had high tryglcerides, but that was because my body was recovering from obesity. My A1C was actually really low (and this was reflected by lack of hunger and stable blood glucose with no hypoglycemia after eating).

When I was underweight, I had very low triglycerides and low fasting blood sugar... but my A1C was actually higher (being too underweight impairs glucose metabolism). No one would ever think being 108 pounds would be worse for blood sugar than being 200 lbs, but it is. I was always getting reactive hypoglycemia (in spite of low carb intake) because my sugar control was worse.

Fasting glucose and triglycerides do not predict glucose control as well as A1C... why don't physicians use the A1C to predict likelihood of diabetes?

Jenny said...

The A1c is actually very poor at predicting blood sugar performance for an individual. It is influenced by how long your red blood cells live. If you are anemic it can be very low when your blood sugar is very high. If your blood cells live longer than usual, your A1c will be much higher than expected based on measuring blood sugar.

The A1c does not predict neuropathy, where 2 hour glucose tolerance test results do.

Even the ADA warns doctors in their official document on the criteria used to diagnose diabetes NOT to use the A1c to diagnose it, because it is too often innaccurate.

It is a useful measurement in large population studies where averages tend to work out well and cheap to administer. But if you rely on your A1c to know how your blood sugar is doing, rather than measuring blood sugar after eating, you may get a very nasty surprise.

ItsTheWooo said...

Oh, that is a good point I did not consider - the variation in red blood cells over time and in individuals.

In my history, though, it seems better because my fasting blood sugar was always low, but "stability" seems to be predicted by changes in the a1c and fructosamine tests. That is when I feel more often I'm getting hypoglycemia, these things tend to rise even though triglycerides and fasting blood sugar may be decreased.

Anonymous said...

Include me out as another for whom the A1c is worse than useless, with Reactive Hypoglycemia and Impaired Glucose Tolerance my FBG and preprandial numbers still are and always have been normal but my postprandials have been all over the place.

Before I gained control my A1c was 5.3, after getting control it was still 5.3, after significantly *improving* my control it was 5.6

In my case the most significant changes were in my BP and especially my lipids, I now have lipids to die for rather than to die from, and the trigs and HDL especially have parallelled the improvement in my postprandial BG.

A1c probably correlates well with control when your FBG has also gone south but as it gets lower so a significant proportion of it relates to postprandial numbers. Maybe with RH the postprandial lows and highs don't quite cancel out. Or maybe my blood is just weird

ItsTheWooo said...

trinkwasser - That's interesting...

I guess it all depends on what causes the hypoglycemia... mine seems to be triggered by hard and fast increases in blood sugar, rather than by constant excessive production from a primary insulin resistance. After 50g dextrose, my sugar went from 140 at 1 hr (and who knows what before that) to 45 at 2 hrs. Severe hypo after a hard spike.
For me an increasing A1C correlates with hypoglycemia because my hypos are always caused by previously occurring hyperglycemic episodes.

I suppose, other people might have hypoglycemia simply because they always make too much insulin and hyperglycemia (the cause of a high a1c) might not necessarily ever occur. I would expect A1C to be lower when they have more hypoglycemia, and higher when they have control.

Anonymous said...

I think the problem is that "Type 2" consists of a whole bunch of similar but different conditions.

In my case there's a familial insulin resistance syndrome which particularly attacks skinny people (like me) and even athletes (like a cousin) rather than the classic overweight people (who often have depressingly good numbers)

Most of my lows follow highs, I can still put out plenty of insulin given enough time, just not at a high rate, so excess carbs will set off a high followed by a reactive low as the pancreas struggles vainly to keep up then fails to shut down properly.

My postprandials have run between 60 and 210 but by monitoring and carefully adjusting my diet vs. exercise I can keep mostly between 70 - 90 with my postprandials maxing out at 120.

Getting rid of the carb-induced highs appears to be what has improved the lipids and BP (which were at their worst on the classic high carb low fat diet - that was also responsible for making me gain weight for the first time in my life) - and getting rid of the (mostly but not always reactive) lows appears to be why my A1c doesn't parallel the other improvements.

You seem to have a similar but more extreme version of the same thing, are we related? (grins)

I could easily be high at one hour, normal again at two hours and low at three hours so it was hard to catch what was occurring without doing a lot of my own monitoring, fortunately after only fifty years a clueful doctor came along.

Jenny said...


Actually our situations are very different because I have normal insulin resistance but do not produce insulin.

Doctors did not realize how insulin sensitive I was until they gave me a "tiny" dose of Lantus and I was hypoing.

I produce some fasting insulin and that is it. My C-peptide post high carb meal is identical to my C-peptide fasting.

That is exactly what would be expected with a secretory defect that prevents secretion in response to rising blood sugar, which appears to be what I have. I seem to have been compensating much of my life with a very robust basal insulin secretion that is finally burning out now that I'm older.

Anonymous said...

Yes sorry, that comment was directed to Itsthewooo (and reading some of her blog suggests to me she has a much more complex endocrine disruption going on).

When i first "met" you (Jenny) there were some obvious similarities (skinny and otherwise fit with familial instances of diabetes and in my case cardiovascular disease, metabolic syndrome, insulin resistance etc.)

It turns out that we have almost completely opposite reasons for similar symptoms.

That's what I find so disappointing about much medical research, they are trying to find similarities where there are none, treat a population as a whole rather than a discrete set of genetic subtypes, and impose one set of treatments on all of them.

Dropping the carbs to a level we can cope with works for the both of us, but other than that there's very little in common. Some of my symptoms are superficially similar to MODY HFN4a and 1a but the rest are completely different - those genes have also been implicated in other forms of Type 2, along with several gazillion others, so maybe future generations will benefit from having the picture clarified.

I stood a good chance of dying of cardiovascular disease with normal A1c and normal FBG, had no-one bothered to check my postprandials and relate them to my ridiculous lipids, helped on my way by the diet I was given. I think I shall buy my current GP a copy of your book as a thank you.