More Diabetes Answers

The power of groups: why one plus one is wa-a-a-ay more than two!

I spent the morning with one of my favorite colleagues, Dr. Jeff Gelgisser.  Jeff is a family practice doctor in Redmond, Washington who looks after a lot of people who have diabetes. 

A few years ago Jeff went through his list of diabetic patients and found a dozen or more who didn't come in very often and who had blood glucose levels and blood pressure levels that were too high and who weren't getting their eyes and feet screened as often as they should. 

 

When he called them up he found:

• Some of these people just preferred to ignore their diabetes. 
• Others just didn't like the way they were treated when they came in to the clinic. 
• Some felt lonely and isolated as if no-one understood how hard it was for them to deal with diabetes along with everything else in their lives. 

They had stopped coming in because they knew that the doctor or nurse who saw them would lecture them about how "bad" they were at looking after themselves.

Jeff then did a rather brave thing.  He invited them to come in as a group and vent their frustrations to him and his staff.  He encouraged them to share ideas with each other and he took the time to explain things to them that they were frustrated about and didn't understand. 

 

And something magic started happening.  Many people in the group changed from feeling hopeless and angry to feeling motivated to change things about how they were living their lives.  They shared positive ideas with each other.  They set new goals for themselves. 

 

They wanted to keep meeting.  They celebrated each other's successes and supported and comforted each other when new difficulties arose. Most all of those people now have lower blood glucose and blood pressure levels, are eating healthier, exercising more, are taking their medicines better and feel much better about their overall lives.

 

Jeff and his team have held group meetings like this for several years now.  Once a year they ask me to join them to answer questions about new research findings in diabetes or anything else that is on their minds.  I am honored to attend.  I always leave feeling more positive and optimistic about my life, too.

There are lots of different types of "group visits" for people with diabetes and other chronic illnesses.  One very successful and popular program was started by Dr. Kate Lorig and colleagues at Stanford. 

It is called "Living Well with Chronic Conditions"

 

You can search for that phrase on the Internet to see if there is a group near where you live. Another great resource for learning how to set up and organize group visits is at www.improvingchroniccare.org.

Is lower always better? (Part 1 - blood glucose control)

Even though I have been studying diabetes and helping people with diabetes to live better with it for over thirty years, I still get surprised constantly when a new study challenges my firmly held assumptions!  One message that has been pushed down our throats for over twenty years at least is that the lower your blood glucose levels are the better.  Yes, we are told, it can be a nuisance pushing your blood glucose levels closer and closer to normal because you run the risk of feeling weird or even passing out from a too low blood glucose (hypoglycemia) and you may need to check your blood glucose more often and may gain weight but over the long haul you will stay healthier if your average blood glucose levels are kept as close to normal as you can.  It is a simple message to push.  Organizations like the American Diabetes Association have pushed the message hard.  All the big pharmaceutical companies who make drugs that help push your blood glucose lower and lower have a vested interest in pushing the message that "lower is always better, and even lower is better still!"  But some recent research studies have challenged this simple message.

The ACCORD trial is a large randomized controlled research trial carried out in many different medical centers around the world.  In the study 6,779 people with type 2 diabetes were randomly assigned to a "standard" approach to lowering their blood glucose where they aimed to get their HbA1c between 7-8% and keep it there or to an "intensive" approach where they tried to get their HbA1c as low as they could get it and keep it there (You can read the full study results in the journal Diabetes Care 33:721-727, 2010).  These subjects had had their diabetes for several years and many of them had had heart disease already (angina, or a heart attack) or had risk factors for heart disease (like high blood pressure or a history of smoking or heart disease in their family).  After about five years of follow up the standard therapy group achieved an average HbA1c of 7.5% whereas those in the intensive treatment group kept their HbA1c at 6.4%.

The surprising results of the study were that after 4-5 years of follow up there were more deaths in the intensive therapy group than in the standard therapy group.  This was even more noticeable in people who had had very poor HbA1c levels to begin with and in people who had worse nerve damage (neuropathy) at the start of the study.

Another study published in the Lancet this year (Currie, CJ, et al.  Lancet 375:481-489, 2010) looked at over 50,000 people with type 2 diabetes treated in general practice in the United Kingdom.  The researchers followed what happened to those people after they had their diabetes treatment intensified by having an extra blood glucose lowering pill or insulin added to their treatment.  As you would expect their HbA1c levels dropped lower after the new treatments were added.  But when they looked to see how those people did in the next few years they found that those whose HbA1c levels stayed high were more likely to die and that those with a lower HbA1c did better... but only up to a point.  It was bad to have a HbA1c of over 9.0% for example.  Those who got it down to 8.5% did better and those at 8.0% or 7.5% did better still.  However, if the new treatment pushed the HbA1c down below 7.0% the outcomes started to get worse with more people dying if their HbA1c was at 6.5% and even worse if it was 6.0%.

So what does this mean for you?  There is no doubt that high blood glucose levels (HbA1c over 9.0%) are bad for your long term health and working to bring it down is a good thing to do.  However, how low to push it is more complicated.  I think it depends on what type of diabetes you have, how long you have had it, what your average blood glucose levels (HbA1c) have been for the past few years and whether or not you have risk factors for heart disease.  For people with type 1 diabetes or people with type 2 diabetes who have always had HbA1c levels that are below 8.0% and who don't have risk factors for heart disease then pushing to keep your HbA1c level down below 7.0% is probably a good thing to prevent long term problems from your diabetes.  But if you have had heart disease (or have several risk factors for heart disease like smoking or having high blood pressure and high cholesterol) and if your HbA1c has often been above 8.0% in the past few years then these research studies suggest that a better target for you might be to keep your HbA1c around 7.5%

Should I worry about my triglycerides, too?

Most everyone worries about their cholesterol these days.  Even when you are relaxing watching a sports game on TV it seems like every other advertisement seems to mention heart disease and cholesterol levels and tells you which pills you should ask your doctor about to lower your cholesterol level.  There is good reason for this.  Cholesterol is a kind of fat that floats around in your blood and can become deposited in the lining of your arteries to increase your risk of getting heart attacks, and strokes, and other bad things.  So in addition to eating a healthy diet and getting exercise it is important for many people with diabetes to take a drug that can lower the LDL cholesterol (often called the "bad cholesterol").  The most common, powerful and effective types of drug to lower LDL cholesterol and protect you from getting heart attacks and strokes are the statindrugs (like simvastatin [Zocor], atorvastatin [Lipitor] and others).  Most people with diabetes who are over age 40 should consider taking a statin drug.

But there is another type of fat that floats around in our blood called triglycerides and that can sometimes cause problems, too.  The normal range for triglycerides in your blood is less than 150 mg/dl (1.7 mmol/L).  If the triglyceride level gets very high (above 500 mg/dl, or 5.6 mmol/L) it can cause a very dangerous inflammation of your pancreas (called pancreatitis) and so most doctors know that they should treat very high levels of triglycerides.  But what if your triglyceride is just a bit higher than normal but not really high?  Suppose your triglyceride level is between 150-350 mg/dl.  Should you take drugs to bring that down into the normal range?  Another part of the famous ACCORD trial looked at this question and came up with a very interesting results.

I have written about the ACCORD trial in several other posts in this blog. In this part of the ACCORD trial 5,518 people with type 2 diabetes (at an average age of 62), many of whom already had heart disease or two or more risk factors for heart disease were studied.  These patients were already taking simvastatin, one of the powerful statin drugs to help lower their LDL cholesterol and protect them from having heart attacks and other bad cardiovascular events in the future.  Many of these people also had triglyceride levels that were in the 150-350 mg/dl range.  The ACCORD researchers wanted to know if outcomes would be even better if they added another type of drug called fenofibrate.  Fenofibrate is a drug that can lower triglyceride levels.  So half the subjects got fenofibrate added to their drug treatment while the other half got "dummy pills" (or placebos) that looked like the fenofibrate pills but didn't contain any active drug.

After almost five years of follow up the levels of triglyceride in the blood were quite a bit lower in the group who were taking fenofibrate than in the group taking the placebo pills.  However, the number of bad cardiovascular events (like heart attacks and strokes, some of which were fatal) was notany lower in the subjects who took fenofibrate compared to those who took the dummy pills.  In fact among women in the study the overall rate of bad events was higher(9.1%) in those taking the fenofibrate compared to those taking the placebos (6.6%)!

What lesson can we learn from this study?  First of all it teaches us that focusing on an improvement in a lab test is much less important that focusing on outcomes that really matter to us (like preventing heart attacks,strokes and deaths).  It also teaches us that sometimes the side effects of a drug cause more harm than the condition it is trying to treat.  So if you have diabetes and are already taking a statin drug to get your LDL cholesterol under 100 mg/dl and if your triglycerides are only slightly above the upper end of the normal range then taking a drug like fenofibrate is not likely to protect you from heart attacks in the future and may even increase that risk, if you are a woman.

Can I mix my fast-acting and long-acting insulins together?

The answer, of course, is, yes you can, but I will try to convince you not to!  Obviously it is more convenient to mix your fast-acting insulin (like Regular, or lispro [Humalog], or aspart [Novolog], or glulisine [Apidra]) in the same syringe as your long-acting insulin (like NPH, or glargine [Lantus], or detemir [Levemir]) and give it as one shot instead of two separate shots.  In fact some insulin preparations already come mixed together.  There is an insulin called 70/30 that is 70 percent NPH insulin mixed up with 30 percent Regular insulin.  We are supposed to believe that the 30% of fast-acting Regular insulin will be absorbed first, giving us a peak of insulin in the first 2-3 hours after taking the shot and that the 70% of longer-acting NPH insulin will kick in later giving us a second peak about 6-10 hours after we take the shot.  The problem is that this is NOT what happens at all.  After all, what is NPH insulin?  It is Regular human insulin with "junk" added to the bottle to slow it down.  the "junk" is a fish protein called protamine and a suspension of zinc molecules.  When you draw up more Regular insulin into the same syringe as the NPH then this Regular insulin also gets "slowed down" by the cloudy mixture of protamine and zinc. Instead of getting an early peak from the Regular and a second slower peak from the NPH you get one single peak that happens half way in between when the Regular peak and the NPH peak would be expected.  Now this works fine for some people but if you really want to get a quick peak to cover the carbohydrate you are eating in a meal and then get a slower, flatter second peak from the NPH insulin it works much better to give each insulin as a separate shot.

A recent study looked at this with two of the popular insulin analogs.  Eleven people with type 1 diabetes who were taking the fast-acting insulin lispro (Humalog) and the long-acting insulin glargine (Lantus) were asked to take them as separate shots on one day and the same doses mixed together in the same syringe on a second day (This study was published recently in the journal, Diabetes Care, 33:1009-1012, 2010 by Cengiz, E, Tamborlane, WV, and colleagues).  What they found was that the rapid peak from the lispro was lost entirely when it was mixed with the glargine insulin.  This meant that the blood glucose rose higher in the first few hours after a meal (because there was no early peak from the lispro to help bring the blood glucose down as the food from the meal was absorbed).  A few hours later the blood glucose dropped lower causing hypoglycemia because the insulin levels were now too high after all the food had been absorbed.

So when you mix fast-acting and long-acting insulins together you lose the beneficial effects of rapid peaks of insulin to cover your meals and the flat, steady, basal insulin level between meals.  Even if it means having to poke yourself a few more times you get more flexibility and control of your blood glucose levels if you give your fast and long-acting insulins as separate shots.

Is lower always better? (Part 2 - blood pressure control)

In The Diabetes Answer BookI talk a lot about heart disease.  The biggest risk to your future life and health if you have diabetes is from heart disease.  If you have diabetes it is incredibly important that you do everything you can to lower your risk of heart disease.  If you smoke do everything you can to quit.  If you are over age 40 talk to your doctor about whether you should be taking a statin drug, or an ACE-inhibitor drug, or low dose aspirin.  Ask to get your cholesterol checked and try to keep the LDL cholesterol level below 100mg/dl (2.6 mmol/L).  And keep track of your blood pressure.  High blood pressure increases your risk of getting a heart attack.  Doing things to lower your blood pressure will keep you healthy and alive longer.  Nothing that I have said so far is surprising, new, or controversial.  The controversial part is to ask the question, "How low should we push the blood pressure level?"

When someone measures your blood pressure they give you two numbers.  The first number is higher.  This is called the systolic blood pressure and tells you how much pressure your heart generates when it is pumping. There is no doubt that having your systolic blood pressure at 150 is better for your long term health than having it at 170, for example.  Having your systolic blood pressure below 140 is better still, and we have been told for years that if you have diabetes you should try to get it below 130.  But would an even lower number be better still.  If you took enough medicines, drugs, and pills to push your systolic blood pressure below 120 would you live a longer and healthier life?

This very question has recently been asked in a large, well-designed, randomized controlled research trial called ACCORD.  In this study 4,733 people with type 2 diabetes, whose average age was 62 and who had had heart disease or at least two risk factors for heart disease were given two different approaches to treating their blood pressure.  Half of them were treated to get their systolic blood pressure below 140.  This group kept their blood pressure at an average of 133 for the next 4-5 years.  The second group were treated to try to get their blood pressure below 120.  they kept it at an average of 119 for the next 4-5 years.  In order to get their blood pressure at this lower level the people in this more intensive treatment group needed to be on more drugs and in bigger doses than the people in the less intensive treatment group.  Because of that the intensive group had more side effects from their treatment.  But the big question was, would they do better when they were followed up?  Would they have fewer heart attacks, and strokes? Would more of them stay alive and well during the 4-5 years of follow up? Surprisingly, the answer to those questions was NO!  Both groups did equally well.  About the same number died or developed a new cardiovascular disease (like a heart attack or a stroke).

So where does this leave us? Does the ACCORD trial tell us what the ideal systolic blood pressure is if you have diabetes?  Well, it does confirm that high blood pressure is not good for you if you have diabetes.  Getting the systolic below 140 is definitely a good goal to go for.  And, in fact, the people in the ACCORD trial who were trying to get their blood pressure below 140 ended up at 133, which is a lot closer to 130 than 140. So maybe a target of less than 130 (which is what the American Diabetes Association and other international group recommend right now) is the right target.  I wish the ACCORD trial had had a third group who were targeted to get their systolic blood pressure below 130, but they didn't.  What the study reminds us is that when taking drugs to treat any medical condition there is a balance between the drugs doing good and causing harm.  Sometimes adding more drugs and pushing the dose higher and higher only results in more cost and more side effects and may not always result in more benefit.

Do I really have to check my blood glucose?

In Chapters 13 and 14 of The Diabetes Answer BookI talk a lot about testing your own blood glucose levels.  Pricking your finger one or more times every day can be a "real pain."  Some people get so frustrated about it that they wonder why they even bother testing at all if they feel fine.  In fact some recent research studies have challenged the idea that blood glucose testing is necessary at all.  O'Kane and colleagues published an interesting study in the British Medical Journal a couple of years ago (O'Kane, MJ, Bunting, B, Copeland, M, et al. Efficacy of self monitoring of blood glucose in patients with newly diagnosed type 2 diabetes (ESMON study): randomised controlled trial.  BMJ 336:1174-1177, 2008).  They studies 184 people with newly diagnosed type 2 diabetes who were not requiring insulin but could manage their diabetes by improving their lifestyle and taking one or more pills.  They were all given educational advice about how to eat healthier and get more exercise but while half of the subjects were also taught to check their own blood glucose levels several times a day the other half did no self blood glucose monitoring.  The results of the study were surprising: after one year both groups had improved their overall blood glucose control to the same amount.  The average HbA1c level in both groups had fallen from 8.7% to 6.9%.  Both groups had lost about the same amount of weight and the number of episodes of hypoglycemia (low blood glucose levels) was the same in both groups.  In fact the only difference between the groups was that the subjects who had done no blood glucose testing were significantly less depressed than the group who did lots of blood glucose testing!

What are we to make of these results?  On one hand they shouldn't surprise us.  Thirty years ago, before the techniques for testing blood glucose levels at home had been invented, people with diabetes still managed to lower their overall blood glucose levels by eating healthier, getting more exercise and taking pills.  After all, testing your own blood glucose level isn't a treatment that improves your diabetes control.  It is simply a tool that allows you to know what your blood glucose level is an any particular time.  It is like being given a flashlight on a pitch dark night to help you get where you want to get to more safely.  You might be able to get where you want to go without the flashlight but you may trip or bump into more trees if you try to move forward in the darkness.  The key thing about the O'Kane study is that the people they enrolled in the study were not in much danger of having their blood glucose go too low.  They simply needed to lose some weight, eat healthier and exercise more.  Some people might be motivated to see their blood glucose levels dropping in response to their improved lifestyle but other people are just as happy to make the changes and not worry about what is happening to their blood glucose.

The key thing about testing your blood glucose is to learn from the information that you get.  So if you want to know what happens to your blood glucose after going for a thirty minute walk test your blood glucose before hand and then again when you are done with the walk.  If you want to know how much higher your blood glucose goes after eating a baked potato compared with a piece of grilled fish then test your blood glucose before and one hour after eating each type of food.  If you have had a really stressful day and are feeling "weird" or tired when you get home you might be interested to know what your blood glucose level is.

So do you really have to check your blood glucose?  If you won't use the information to change anything you are doing to manage your diabetes then it is probably a waste of time to check.  But if you learn something useful that gives you insight about how to manage your diabetes better then testing your blood glucose can be very helpful.  Maybe I'm just not very brave and not much of a risk-taker but if I was asked to run down the street on a pitch dark night when the power was out I would prefer to have a flashlight in my hand!

Can a bleeding ulcer cure diabetes!?

Just the other day one of my patients thought that her diabetes had been "cured" after being in hospital with a bleeding stomach ulcer.  When she came out of the hospital her HbA1c, which had previously been in the 8.0 -9.0 percent range, came back at 4.9 percent!  She was delighted until I explained the reason.  She had lost several pints of blood from her bleeding ulcer and so was quite unwell with severe anemia when she was admitted to hospital.  In order to treat her she had been treated with a transfusion of several pints of blood.  That blood had been donated by one or more non-diabetic individuals.  When this was mixed with her own blood the average amount of HbA1c dropped right into the normal range because the transfused blood had very little glucose attached to it.  Over the next 10 to 12 weeks I'm afraid that my patient's HbA1c will rise back up to the 8-0 percent range as glucose gets attached to her own hemoglobin and to the hemoglobin from the transfused blood.

To understand more about HbA1c look at page 83 in The Diabetes Answer Book or check out this post:

What is glycosylated hemoglobin (or hemoglobin A1c, HbA1c)?

http://gomeworld.typepad.com/morediabetesanswers/2010/01/what-is-glycosylated-hemoglobin-or-hemoglobin-a1c-hba1c.html

Does insulin glargine (Lantus) cause cancer?

What a great, if scary, question!  I'm afraid the answer that I give you is not as straightforward as you would like.  The answer is, "Probably not, but we don't know for sure."  You might wonder why someone even asked such a question.  I mean if you have diabetes then insulin is just the hormone that your body doesn't make enough of.  Taking insulin should be good for you, not bad for you, right?  Absolutely.  Insulin has been a miraculous advance in the treatment of diabetes and the millions of people who take insulin are much healthier because of it.  So why wonder if insulin glargine (Lantus) might cause cancer?  Let me tell you what is and is not known.

By looking at large populations of people around the world researchers have known for well over a decade that people who are very overweight and people who have type 2 diabetes have an increased risk for several types of cancer (cancer of the colon, cancer of the pancreas and cancer of the breast).  It is not known why.  One possible reason is because people who are overweight and people with type 2 diabetes have a lot more insulin in their body and insulin stimulates the growth of cells, so maybe it stimulates the growth of abnormal cancer cells.

Researchers have done experiments in laboratories showing that insulin can stimulate the growth of breast cancer cells.  Insulin glargine (made by the Pharmaceutical Company Sanofi-Aventis) is an analog of human insulin.  This means that the scientists have modified the natural human insulin molecule by removing certain parts of the molecule (amino acids) and adding extra ones.  This is done all the time to create new drugs to treat all kinds of illness.  People who take steroid inhalers for asthma or steroid pills to treat their arthritis are taking analogs of the natural steroid hormone, cortisol, that we all make in our bodies.  So insulin glargine (Lantus) is a modified kind of insulin that is a bit different from naturally occurring human insulin.  Well, it turns out that in laboratory studies insulin glargine stimulates cell growth even more strongly than natural human insulin, so it is certainly possible that people taking Lantus to treat their diabetes could have a higher risk of cancer.

What's the best way to know for sure?  By far the best way would be to find a large number of people (several thousand at least) with type 2 diabetes who need insulin treatment and ask them if they would let you randomly assign them to use insulin glargine (Lantus) or another insulin (such as human NPH insulin, perhaps).  It is likely that they could get similar improvements in their blood glucose levels and could have their diabetes controlled with either insulin for many years.  Then, we would keep track of how they were doing for 5, 10, 15 years or more.  During that time some of those thousands of people would get cancer and some of them would die from cancer or from other causes.  Because the people were randomly assigned to either Lantus or NPH at the beginning, if we found a different rate of cancer or death between the two groups of people that would be very powerful evidence.  We would know with confidence whether one type of insulin gave less risk of cancer than the other.  Unfortunately that study has not been done.

There has been a small study where 1017 patients were randomized in this way and during four years of follow-up there were 20 new cases of cancer in the people taking glargine insulin (Lantus) and 31 cases in patients taking human insulin.  In other words the patients taking Lantus insulin seemed to be LESS likely to develop cancer than those taking human insulin.  The trouble was that there were too few people in the study and they weren't followed for long enough for us to know for sure.

Well, if no one has done a larger and longer study like this (a randomized controlled trial) how else might you answer this question?  You could try looking to see what happened to the many thousands of people who have been started on different types of insulin over the past decade or so.  This is called a cohort study.  For those of you who want to learn more details about the advantages or disadvantages of cohort studies compared to randomized controlled trials (RCTs) the entry in Wikipedia is very helpful:

http://en.wikipedia.org/wiki/Cohort_study

 One of the problems with a cohort study is that you can never be sure why particular patients were started on particular types of insulin.  It certainly wasn't random!  And so you never know whether what you find is because of the type of insulin the person is taking or because of some other factor that you don't know about.

Here's a great example of why you should be very cautious about drawing conclusions from cohort studies.  After the end of World War 2 the number of households that had a radio in them went up dramatically over the next decade.  During that same decade the birthrate also went up dramatically.  If you were naive you could conclude that having a radio in your house increases the risk that you will get pregnant!  But if you had done a randomized controlled study after World War 2 where half the households had been given radios and the other half were not given radios I  seriously doubt that you would have seen a difference in the birth rates in those two groups!

When large groups of people (over 100,000) with type 2 diabetes who were taking insulin in Germany and in the United Kingdom were studied the number of people who developed cancer while taking insulin glargine were slightly lower than those taking other types of insulin.  But when the researchers divided up the patients into those who were taking a small, medium or large dose of insulin, and when they looked at those who took only one type of insulin compared with those who took more than one type of insulin they concluded that patients who took only insulin glargine, and took it in high doses, were at higher risk for developing cancer.  This has set off a flurry of angry name-calling in the medical journals with various experts criticizing each other for the methods they have used and the conclusions that they have come to.  Truthfully the only way to know for sure would be for some important regulatory agency like the Federal Drug Administration (the FDA) to demand that a large randomized controlled trial be done.

So where does this leave you, if you are a someone taking insulin glargine (Lantus)?  I think that with what we know today for sure the benefits of taking it far outweigh the risks.  If some group of researchers do decide to do a proper randomized controlled study to see whether taking insulin glargine (or any other kind of insulin) causes a lower or higher risk of cancer that would be great.  I would love to know the answer.  But for now I suggest that you continue to take the insulin you are taking and focus on things that you CAN do to lower your risk of cancer in the future.  Don't smoke.  Don't drink alcohol to excess.  Eat a healthy diet. Get regular exercise.  Do something fun every day.  Try to keep your stress level down.

Why does the average blood glucose from my meter give me a different number than the eAG?

In a recent post I explained how it is possible to tell you what your average blood glucose level has been for the past 2-3 months by knowing your HbA1c.  Although this is meant to be helpful it can sometimes be confusing.  Let's say you test your blood glucose 3 or 4 times a day.  If you have a memory meter it will probably tell you what your average has been in the last 14 or 30 days.  Let's say that average is 183 mg/dL (or 10.1 mmol/L).  If you look at the table comparing HbA1C to the estimated average glucose (eAG) you would expect that when your HbA1c is tested it will be 8.0%.  but what if your HbA1c comes back at 9.8%, almost two whole percentage points higher than you meter average would have predicted.  What is going on?  The problem is probably not with your meter or with the accuracy of the lab test.  If your HbA1c is 9.8% this means that for the past 2-3 months your average blood glucose level really has been somewhere between 226-240 mg/dL (12.6-13.4 mmol/L).  The average that your meter gives you is simply the average of all the blood glucose tests that you did.  If you only test before meals or when you are feeling low, for example, and are not testing after meals or when you feel that your blood glucose is high then your "meter average" will be an underestimate of what your actual average is; it is not adding in the higher numbers to the average.  On the other hand if your "meter average" is higher than the eAG from your HbA1c test this suggests that there must be times during the day when your blood glucose is low when you are not testing.

So you can really make good use of this eAG number.  If it suggests that your blood glucose levels are higher than you thought from the tests you are doing every day then you should consider testing one or two hours after your main meals to see if it goes high at that time.  If the eAG is lower than you expected then it is likely that your blood glucose level is going low at times when you are unaware of it.  This can often happen during the night when you are asleep so it might be a good idea to test then to see if that is happening.

It is certainly a good idea to discuss your eAG with you doctor and health care team if the results don't make sense to you.

What is the Estimated Average Glucose (eAG)?

One of the most important numbers that you should track if you have diabetes is your HbA1c number (sometimes called glycosylated hemoglobin or hemoglobin A1c or simply A1c).  the HbA1c number tells you what percentage of the hemoglobin protein in your red blood cells has got glucose attached to it.  If the number is high this tells you that your average blood glucose level for the past two or three months has been high.  I explain this in more detail in Chapter 7 of The Diabetes Answer Book.  For people who don't have diabetes the HbA1c will be between four and six percent.  If your diabetes is really out of control more than 10 percent of your hemoglobin will have glucose stuck onto it.  For most people with diabetes the HbA1c is between 6-10% (and if you can safely keep it closer to 6% than to 10% that is great).

Because most people are more familiar with thinking of their blood glucose levels the American Diabetes Association is pushing for laboratories to not only give people their HbA1c result but to tell them what their average blood glucose level must have been for the past 2-3 months to get that HbA1c number.  The reason that this is possible is because of a very well-designed study that was published in the Journal Diabetes Care in August 2008.  Dr. Nathan and his colleagues worked with 507 subjects (268 had type 1 diabetes, 159 had type 2 diabetes, and 80 did not have diabetes at all) at 10 international research centers.  Over a six month period each subject checked their own blood glucose 7 or 8 times a day at least three days a week (before and90 minutes after each meal, before bedtime and once in the middle of the night).  In addition they wore a continuous glucose monitoring system (CGMS) on eight days during the six months.  So each subject had about 2,700 blood glucose tests done during the six month period as well as having their HbA1c checked several times during the same six months.  By taking an average of all these blood glucose tests and comparing them to the HbA1c taken during the same period the researchers found a very strong and straight relationship between the two.  They were able to construct a table which shows the average blood glucose level for each HbA1c.  Ihave shown this for a variety of HbA1c levels in the table below.  So the next time you get told what your HbA1c test result is you can look at this table and see what your average blood glucose level has been for the past 2-3 months.  The middle column is in "American" units of mg/dL and the right hand column give the estimated average glucose in International units of mmol/L.

HbA1c	eAG (mg/dL)	eAG (mmol/L)
6.0	126	7.0
6.5	140	7.8
7.0	154	8.6
7.5	169	9.4
8.0	183	10.1
8.5	197	10.9
9.0	212	11.8
9.5	226	12.6
10.0	240	13.4
10.5	255	14.1
11.0	269	14.9