Functional Health Services for Your Well Being

diabetes 3

Measuring Insulin Resistance

by Alex Boersma

In Part I of this series I described type II diabetes as the most pernicious and calamitous disease of the 21st century.  Indeed, it seems that despite modern advances in the  treatment of heart disease and cancer, the capability of our medical system to deal with blood sugar dysregulation remains pathetically inadequate.  In fact, just this week  Medscape reported on the 47th annual meeting of the European Association for the Study of Diabetes which described the diabetes epidemic as  ”on a relentlessly upward trajectory, with no signs of abating“.  From the meeting:

“Data from international studies demonstrate that the number of people withdiabetes in 2011 has reached 366 million. This year, 4.6 million deaths will be attributed to diabetes, with 1 person dying from diabetes every 7 seconds. Healthcare spending on diabetes has reached $465 billion.”

And it’s not just in Europe.  Below is a chart from a  recent analysis of diabetes prevalence in Ontario

ontario diabetes statistics

 Follow the blue or red lines into the future and perhaps you’ll reconsider that venti frappucino if you plan on becoming a sexagenarian!

In Part II of the series, I examined the role of insulin resistance as a determining factor in the progression to type II diabetes.  Primary factors in the development of insulin resistance were discussed, including cellular accumulation of excess free fatty acids (FFA), stress, inflammation and fatty liver.  It was established that these four factors were  associated with each other in a cyclic fashion, meaning an increase in one factor frequently causes increases in other factors.  All four factors factors are clearly related to excess weight, particularly when that weight is carried predominantly in the abdominal area.

If insulin resistance is such a dominant risk factor in the progression to type II diabetes, it might be useful to have an effective tool for diagnosing it.  In this section I will discuss the various options available for measuring insulin resistance and their efficacy in predicting progression to type II diabetes.

Before we begin, though, a little perspective.  Type II diabetes is clearly a disease which develops and progresses over many years.  In prospective studies, participants who eventually develop diabetes often present with altered glucose metabolism years in advance.  In the graphs below, we see that insulin resistance begins a decade before the pancreas starts losing its ability to produce insulin.  Likewise, post meal glucose intolerance increases sharply 3 or 4 years before fasting blood glucose hits the diagnostic marker (dotted line) for diabetes.

insulin resistance

Taken from the American Association of Clinical Endocrinologists

Keep in mind, the earlier you can diagnose blood sugar dysregulation, the more likely you are to prevent progression to full blown diabetes.  From the graphs above, it is obvious that a simple diagnostic tool for accurately measuring  insulin resistance would be exactly “the bomb” for pushing our impending type II diabetes epidemic into the slow lane.  If we could know 10 years ahead of time that somebody is going to become diabetic, all we would have to do is get them to change their lifestyle and they could save themselves and the medical system a great deal of cost and aggravation.  Yeah, that’s all we’d have to do! 

Unfortunately, no such simple diagnostic tool exists.  There are a number of effective tools for accurately measuring insulin resistance, but none are simple.  The “gold standard” for measuring insulin resistance is the euglycemic hyperinsulinemic clamp.  It works great if you happen to have one of these setups:


Too bad only highly sophisticated metabolic research labs our outfitted with these contraptions.  So unless you volunteer to be a guinea pig at a research facility, it is unlikely that you will have the opportunity to test your insulin resistance with this degree of accuracy.  Your doctor – in case you’re wondering –  probably doesn’t have one stuffed in his clinic closet!

What we are left with, then, is a triumvirate of blood tests – none of which give us a completely accurate picture of what our body is doing with the box of tinbits we just shovelled in.  Still, as we shall see, there is much that we can learn if we know how to interpret them correctly.  Let’s take a look.

    insulin insulin 2

A1C (also known as HbA1C) is a measure of glycated hemoglobin.  If you expose blood to sugar for long enough, some of the sugar attaches to some of the blood…this is called glycation.  Since blood cells usually live in the body for up to 3 months (after which they die and get replaced) and since there is always some sugar in the blood, a percentage of your blood is always glycated.  The more sugar you have in your blood, the more likely it is to attach to blood cells and the higher the percentage of your hemoglobin which becomes glycated.  Therefore, HbA1C provides us with a rough estimate of average blood sugar levels over a 6 to 12 week period.

If you are resistant to insulin, sugar will tend to stay in your blood longer than if you are not resistant to insulin.  It seems intuitive, therefore, that elevated A1C would be indicative of insulin resistance.  Indeed, it does seem that, at least according to this paper  , HbA1c is effective at predicting insulin resistance, particularly in people with normal glucose tolerance and relatively high insulin sensitivity.  In other words, HbA1C seems to tell you if you are insulin resistant before you become “too” insulin resistant!  Just the thing we are looking for if we want to catch diabetes before it becomes diabetes!

The problem is that very little research has gone into determining exactly what A1C level is optimal for non-diabetics.  Most of the scientific debate concerns itself with optimal levels for established diabetics and does little to inform us about what our A1C levels should be for optimal health.  We know that 5.7 is good (at least compared to established diabetics) but would lower levels be even better?  And if so, how low should we go?

A recent British study published in the Annals of Internal Medicine  illuminates the issue.  The researchers followed about 10,000 people (age 45 to 79) for an average of 6 years and measured how many of them developed heart/cardiovascular disease or died during that time.  Below is a graphical interpretation I composed to describe the results.   As expected, when A1C goes beyond 6.5, people get sick and die at an alarmingly high rate.  But what about levels below 6?

age adjusted relative risk

When A1C is under 6, the results are not nearly as dramatic.  Still, for participants in this study, people withA1C below 5 had about a 70% lower risk of being dead after 6 years than people with A1C over 5.9.  The correlation for heart disease was even stronger.  Don’t tell the statinators, but dropping your A1C from 5.9 to just below 5 has a much more significant impact on heart/cardiovascular disease and all cause mortality than even the 50% reduction in LDL achieved in statin trials like the Jupiter study!

Okay, I know, this is just an epidemiological study and, as such, by no means proves that having high A1C causes people to die earlier.  It is entirely possible that some unknown variable is causing both the high A1C and the mortality.

More importantly, it should be noted that there is a high degree of variability between individuals for A1C levels.  In other words, A1C of 5.7 may be normal for you but high for me.  The only way we can know what our individual A1C levels should be is by having some kind of baseline over time.  For example, if we start early in life with an annual A1C measurement, it should remain fairly consistent as long as we maintain good glycemic control.  If we begin to develop mild insulin resistance, this would probably be reflected by a gradual increase in A1C over time.  A gradual but consistent elevation in baseline A1C would probably be an excellent indicator of early insulin resistance.  Too bad almost nobody has a baseline A1C since doctors usually only start testing for it after you have been diagnosed as a diabetic!

***One final cautionary note on A1C.  When you make significant improvements in your health through lifestyle changes, you may end up with healthier blood cells which live longer and, therefore, accumulate more glycation.  Start exercising, lose some weight, eat less junk food and voila…up goes your A1C!

As you can see, A1C has a number of issues which make it a less than ideal tool for reliably measuring insulin resistance or predicting progression to diabetes.  Having said that, I do believe that, in general, an A1C  below 5 sustained over time is predictive of good health and indicative of excellent glucose tolerance.


Fasting Plasma Glucose (FPG) is by far the most common blood sugar test done by doctors. (see here for a fascinating account of the political reasons why)  It is almost always part of a regular blood panel.  It measures the amount of sugar in your blood after a long (usually 10 – 14 hours) fast.  This being the case, FPG tells you absolutely nothing about how your blood sugar responds to the things you eat. 

 If you haven’t eaten anything for half a day, any sugar in your blood is almost certainly sugar which has been put there by your liver.  If you have high blood sugar after not eating for half a day, your liver is putting more sugar into your blood than it is supposed to.  Your liver is supposed to stop pushing out sugar if it recognizes that there is insulin in the blood.  If there is sugar in your blood, there should also be at least a basal level of insulin  in the blood. 

So if Fasting Plasma Glucose is high, it means one of two things.  Either your pancreas is not capable of responding to basal levels of blood sugar by producing basal levels of insulin, in which case your pancreas is probably FUBAR and you are quite likely already a full blown diabetic.  Or – and this is more likely if you haven’t already been diagnosed with diabetes – your liver is not recognizing the basal levels of insulin which your still- healthy pancreas is producing.  This is known as hepatic insulin resistance…your liver has become resistant to insulin.  The graph below (taken from a very interesting Medscapearticle - providing that you are the kind of person who finds Medscape articles interesting) shows just how closely fasting blood sugar and hepatic insulin resistance are related.

 hepatic insulin resistance

As you can see, the higher the FPG, the more sugar being produced by the liver.  Clearly, fasting blood glucose tells us something important about hepatic insulin resistance.  But what does it tell us about the progression to type II diabetes?  If  your liver becomes resistant to insulin early in the progression to type II, then FPG should accurately predict that progression. 

As it turns out, the liver only becomes insulin resistant early in some people.  In others, the muscles or the fat tissue lead the charge into blood sugar dysregulation.  This paper indicates that the early stages of insulin resistance are characterized primarily by impaired glycogen synthesis in the muscles.  The muscles are less inclined to sponge glucose out of the blood and turn it into glycogen, therefore it takes more time and  insulin to achieve normal blood glucose levels.  The authors concede that in some people the liver wins the race to insulin resistance, but they do not seem to think that this is common.

The point is that Fasting Plasma Glucose levels can easily lie within the “normal” range despite problems with sugar regulation.  Even if your muscles are seriously insulin resistant, you can get away with the extra large pizza and coke for dinner because your pancreas continues to pump out mega-shots of insulin to compensate for your sky-rocketing blood sugar.  2 hours after dinner, your blood sugar might be way beyond dangerous diabetic levels, but your doctor won’t know about that.  Your doctor won’t know about it because the FPG test isn’t administered until the next morning.  And by the next morning, your iron pancreas will have pumped out enough insulin to bring blood sugar back down to below the 5.5 mark.  As long as your liver remains at least somewhat sensitive to insulin, it will not start pushing sugar into the blood and upping your FPG levels.  By the time FPG finally gets high enough for your doctor to notice, you have probably already sustained years of glucose-related vascular damage.  

If we expect to use FPG as a tool in our battle against the diabetes epidemic, it is clear that we must, at a minimum, re-evaluate the parameters by which we define “normal” blood sugar. The accepted FPG “normal” of 5.6 used by most doctors may be normal in the context of a population rife with obesity, overweight, vascular disease and blood sugar dysregulation.  But it is not likely to be optimal. It probably represents at least a minimal level of insulin resistance and suggests that vascular damage due to glycation is occurring.  Healthy young people do not tend to have FPG levels as high as 5.6.  In fact, as this graph shows optimal FPG is probably closer to 4.5 than it is to 5.5.

normal blood sugar small

What is normal?

It is difficult to read here, so click on the link above if you want to see more clearly.  The blue line represents total plasma glucose for 24 healthy young study participants.  The black vertical line shows where the first meal of the day happens.  We are interested in the part of the blue line before the first meal.   From 6:00 am to 7:30 am, these truly normal people had, on average, a blood sugar around 80mg/dl which corresponds to 4.4 mmol/l for us Canadians.  4.4 is a long way removed from 5.6!  The upper limit of their morning measurements was around 95 (5.3 in Canada) while the lower limit was around 60 (3.3 in Canada).   

Over the last 15 years, standard guidelines for “normal” FPG have been revised twice.  15 years ago anything under 6.7 was considered normal.  In 1997 the American Diabetes Association brought ”normal”  down to 6.1.  In 2003, they lowered it again to 5.6.  Granted, much of this pressure to lower “normal” has probably been initiated by pharmaceutical companies.  After all, lower standards for blood sugar means more sales of blood sugar lowering drugs like Metformin.  Still, in light of the evidence provided above, it would seem that lower is indeed better…at least down to the FPG = 4.5 level or so.  And, if caught early enough, very few people would need to rely on pharmaceuticals to get them there.

In my opinion, Fasting Plasma Glucose, like HbA1C, can be an effective tool for measuring insulin resistance, but only for some people and only if a baseline is established early in life.  Whatever your FPG is when you are young and healthy should be your target as you age.  If and when it begins to creep up into the mid 5’s – as it seems to for most people in middle age – this is an indication that you are probably developing some level of insulin resistance.  At this point, further testing is certainly warranted, and lifestyle interventions to bring FPG back down below 5 will probably be helpful and effective.  But if you wait until FPG rises above 5.6, it will be much more difficult to bring it down and you will be much more likely to require pharmaceutical intervention.


The Glucose Tolerance Test (GTT) was once a commonly used tool in the evaluation of blood sugar dysregulation.  It was recognized decades ago that a glucose tolerance test identified distinct populations susceptible to developing diabetes independent of fasting plasma glucose.  Unfortunately, the GTT has been effectively banished from medical practice, mostly because it is rather cumbersome from a practical perspective.  This paperdescribes some of the political and practical  reasons for the elimination of GTT.  It also discusses the reasons why it’s replacement by FPG has been detrimental in the struggle to predict early insulin resistance.

Simply put, your doctor doesn’t routinely give you a glucose tolerance test because a GTT takes time…something doctors and patients alike don’t seem to have a lot of.  Here’s how the GTT works.  Fast all night, then go visit your doctor in the morning.  Have your blood drawn before you begin the test.  Drink a 75 gram shot of awful glucose syrup.  Wait 2 hours and have your blood drawn again to see how much sugar is still in it.  If you still have more than 11 mmol of sugar in every litre of your blood 2 hours after drinking the glucose goop, congratulations …you are a diabetic – you can get in line for your metformin and insulin prescriptions.  If you have less than 7.8 mmol of sugar in every litre of your blood 2 hours after drinking the glucose goop, your doctor gives you a free pass to the candy store. 

The problem, of course, is that the cut-off numbers for GTT results are rather arbitrary.  If your glucose tolerance is impaired at 7.8, does this mean that at 7.7 you are truly free and clear to steal all the chocolate bars out of your kid’s Halloween haul?  Is everything below 7.8 truly “normal” or does a higher number in the “normal” range signify increased risk for developing glucose intolerance?

The chart below, which I developed from the results of this paper, indicates that, indeed, not all “normal” GTT results are equal.

glucose sensitivity chart GOOD

As you can see, the range of insulin sensitivity within “normal” glucose tolerance varies significantly.  If your blood sugar is back below 5 two hours after drinking your glucose goop, it would seem that you retain excellent glucose sensitivity.  If blood sugar is still up above the 7 mark, then not so much!  In fact, if blood sugar is still above 7, you have probably already lost a third of your body’s ability to sense insulin.  The truly scary part, though, is that glucose sensitivity, for most people, levels off around the 50% mark- even for severe type II diabetics!   Since doctors don’t normally do a GTT until they already suspect diabetes, this means that most of the reduction in your insulin sensitivity will happen well before you ever get your glass of glucose syrup!

Here are the conclusions of the above study:

“In summary, we showed in a large cohort of Finnish men that the impairment of peripheral insulin sensitivity started at relatively low PG (GTT) levels within the normoglycemic range. In contrast, the impairment of insulin secretion progressed substantially only in the diabetic range of fasting and 2-h glucose levels. Peripheral insulin resistance was a predominant feature of isolated IGT (impaired glucose tolerance), whereas impaired insulin secretion characterized isolated IFG (impaired fasting glucose)”

Translation:  The 2 hour glucose tolerance “normal’ cut-off of 7.9 sucks at catching insulin resistance early in the progression to type II diabetes.  It does tell you that your pancreas is about to pooch out but your doctor probably already knew that if he only ordered your GTT after you had already failed a fasting plasma glucose test.



good news bad news

First the bad news.  The bad news is that conventional tools for measuring and predicting the early stages of insulin resistance are woefully inadequate.  None of the tests described above is capable of providing us with a complete picture of our individual insulin sensitivity.  For the most part, guidelines and cutoffs used by the medical establishment set off red flags only well after you have merged onto the one-way lanes of the road to diabetes. 

The good news is that there is much we can learn from the science behind these tests.  Clearly, being a “normal” North American with “normal” blood sugar test results tells you diddly about your risk of developing diabetes.  In fact, if you look at the trends, being “normal” in North America means you have an excellent chance of developing diabetes – or at least insulin resistance - as you age.

So if there is anything we can learn from the science, it is that we should not be “normal”.  We should not take for granted that fasting plasma glucose gradually goes up as we age.  We should not assume that “normal” blood sugar levels implicate normal sensitivity to insulin.  And we should not sit idly as early warning signs for insulin resistance pass us by.

What are we to do?  Well, it really is quite simple: 

  1. Moniter FPG levels.  Since these are tested automatically with most regular blood work, they are the easiest to monitor.  Any levels above 5 mmol/l should warrant further investigation. That doesn’t mean you’ll have to start mainlining insulin or popping metformin just yet.  It just means that maybe you should take a closer look at the way your body responds to that box of tinbits. 
  2. If in doubt, do your own glucose tolerance test.  In my opinion, one of the best investments anybody can make in their own health is a glucose meter.  For less than $100 you get to figure out for yourself whether or not the ingestion of carbohydrates is an issue for you. You get to know what foods and what amounts of those foods wreak havoc with your blood sugar.  And you get to accumulate this wealth of physiological information without ever having to drink even a drop of glucose syrup!


How to be your own blood sugar control board!

 Monitoring your blood sugar response to food is actually quite simple.  Start by purchasing a glucose meter and strips from your local pharmacy.  The glucose meter is usually free if you buy more than 100 strips (usually for less than $100).

Measure your blood sugar first thing in the morning for a few days.  Doing so will give you a bit of practice using the meter and also establish a baseline which you can use to compare future readings.  You are now ready to begin monitoring your blood sugar response to food.

Before testing your response to any meal, be sure to measure your fasted (relatively speaking) blood sugar.  This measure should be fairly close (within about 10%) to your morning fasted blood glucose.  If it is high before you start eating, then the results of your testing will probably be inaccurate.

The first few times you test a meal, it is useful to test every 1/2 hour after you finish eating.  This will give you an idea whether you experience any abnormal blood sugar swings either early or late in the absorptive period.  As you can tell from the “What is normal?” graph above, blood sugar normally peaks somewhere between 1/2 hour and 1 hour after a meal.  By 2 hours it should be back down fairly close to fasting level.  If your blood sugar does something significantly different than this, you probably have some issues.

We know that having a 2 hour blood glucose of more than 7.8 indicates some level of vascular damage.  We also know, according to This study, that having a 1 hour plasma glucose level above 8.5 mmol/l  predicts diabetes better than just about anything else.  It also seems to correlate very well with various measurements of insulin resistance.  Clearly, we want to keep our own numbers lower than these.

But remember, 8.5 at 1 hour and 7.8 at 2 hours are probably not optimal numbers.  If, again, we look to the “What is normal?”graph above, we see that average normal 1 hour blood glucose comes in around 115 mg/dl, which translates to about 6.4 mmol/l.  The extremes in this group were about 3.9 on the low end and 8.3 on the high end.  Moreover, another study shows that normal people spend less than 6 minutes a day above 7.8.  So keeping your blood sugar between 7.8 and 8.5 for as much as 2 hours after every meal is clearly not normal!

The goal, then, at a minimum, would be to keep 1 hour blood sugar below 8.5 and 2 hour blood sugar below 7.8.  That much is obvious.  Beyond that, it seems to me that lower is generally better – to a point.  I would guess that 1 hour results around 6 – 6.5 would be closer to optimal.  Likewise, 2 hour results around 5 -5.5 are also probably closer to optimal. 

Knowledge is Power

The evidence suggests that knowing whether or not you are resistant to insulin can tell you much about the status of your overall health.  As we have seen, traditional measures of blood sugar are not particularly effective at predicting insulin resistance.  But do not despair.  Thankfully, glucose meters are cheap and simple to use.  Thankfully, good research is available to tell us whether the numbers coming out of our glucose meters reflect optimal or sub-optimal health.  And thankfully, as we shall see in the next and final article of this series, altering our diet to change those numbers is a relatively simple (but by no means easy) thing to do.

To book an appointment online, please go to or call 519 939 9034
Copyright © 2009 Total Body Health Studio, All rights reserved.