One reason the drug makers can make this claim is that it is difficult to disprove. That's because there is currently no way to measure the beta cell mass in a pancreas located inside a living person. Imaging fails because the size of the pancreas doesn't tell you anything about the density of beta cells. So researchers base their claim either on rodent studies--though rodents have very different blood sugar metabolisms than humans--or they estimate the beta cell mass of humans taking their drugs by measuring insulin or c-peptide levels and applying a formula, HOMA, to compute a number they believe represents the beta cell mass.
HOMA stands for "Homeostatic model assessment". The formula was created back in 1985 long before we had reliable measures of blood sugar variables or any way to measure beta cell mass. To learn the details of how HOMA is computed, read this essay:
Diabetes in Control: HOMA: Often Mentioned, Rarely Defined.
I have blogged before on how HOMA calculations fail to distinguish between insulin resistance and insulin deficiency and how this measure gave doctors the erroneous belief that Type 2s are largely insulin resistant, rather than, as turns out to be the case, insulin deficient.
I am particularly sensitive to this problem because when I plugged my numbers into the Oxford HOMA calculator some years ago, it told me I was seriously insulin resistant. It was only when I injected insulin that I discovered that I was very insulin sensitive--two or three units of fast acting insulin will cover 40-60 grams of carbohydrate for me, and I can hypo on 8 units of Lantus. So I felt vindicated when I learned that the HOMA formula labels many people "insulin resistant" who, in fact, are insulin deficient.
Now a new study takes another look at how the HOMA formula is used to estimate beta cell mass and finds it defective there, too. This is important because it is the HOMA calculation that is often used to support claims that various drugs are increasing beta cell mass.
The study is:
Functional Assessment of Pancreatic β-Cell Area in Humans. Juris J. Meier et al. Diabetes. July 2009, vol. 58, no. 7, 1595-1603. doi: 10.2337/db08-1611
The researchers here rounded up a group of unfortunate people who, because of serious medical problems, were about to have their pancreata removed. These subjects had a wide range of pancreatic function before surgery as they had different diseases some of which killed beta cells and some of which did not. The researchers gave them each a Glucose Tolerance Tests with C-peptide measurements before surgery.
After surgery their pancreases were carefully examined to find out exactly what their beta cell mass might be. This is not easily done because as soon as a pancreas is removed the enzymes it contains start to digest the pancreatic tissue. It is only over the past decade that techniques have been devised that allow researchers to examine beta cells shortly after death or after surgical excision of the pancreas.
What the researchers found was this:
β-Cell area was related to fasting glucose concentrations in an inverse linear fashion (r = −0.53, P = 0.0014) and to 120-min postchallenge glycemia in an inverse exponential fashion (r = −0.89)To give you a feeling for the difference between an "inverse linear" relationship and an "inverse exponential" relationship, let's look at a hypothetical example. In the linear relationship, if your fasting glucose was up, say, 3% then your beta cell mass would be down that same 3%. If the fasting glucose rose by 6%, your beta mass would go down 6%.
With the exponential relationship that applies to the two hour glucose tolerance test result, if the two hour reading was up 3% your beta cell mass would be down 9% but if the 2 hour reading was up 6% your beta mass would be down 36%, if it was up 9% your beta mass might be down 81%. That gives you some idea of how much more intense an exponential relationship is compared to a linear relationship.
This suggests, too, why the glucose tolerance test and/or post-meal tests are so much more important than the fasting glucose test for diagnosing diabetes. Because small changes in the two hour test result point to larger changes in your beta cell mass.
What the researchers found in this study was that the blood test measure that best predicted the actual beta cell mass that they found when they examined the pancreas, was, "the C-peptide-to-glucose ratio determined 15 minutes after the glucose drink".
This is very interesting indeed, because until now we have been told because of earlier research that the 15-minute glucose tolerance test result was not predictive of anything and that only the 30 minute result was useful. That may be true if you look only at glucose levels, but this study makes it clear that it is not true if you include C-peptide levels.
Unfortunately, because of the belief that the 15 minutes result isn't useful, you will have to look long and hard to find other studies that measure 15 minutes C-peptide. Invariably when C-peptide is measured at all it is measured, fasting, at 1/2 hour intervals, and often only at the 2 hour mark.
The researchers also found that "a fasting C-peptide–to–glucose ratio already yielded a reasonably close correlation." It wasn't as good as the oral glucose tolerance test 15 minute measure, as shown by the "r" values. The lower the "r" value the better the correlation with an inverse result. The r value for the 15-minutes result was -.72 (-1.00 would have been perfect correlation and 0 would have been no correlation.). The r value for the fasting c-peptide/glucose ratio was -.63.
But here's the kicker. The researchers add "Homeostasis model assessment (HOMA) β-cell function was unrelated to β-cell area.
This is the same HOMA calculation drug company researchers are prone to use to support claims that their drugs are increasing beta cell mass.
I have long doubted that any oral drug can make a significant improvement in anyone's beta cell mass for several reasons. Long term studies show that over several years, after an initial improvement, the blood sugar control of people taking these drugs starts to declines. It does not improve the way it would if they had more beta cells cranking out insulin. This suggests that these drugs either stimulate insulin secretion--which is true of Januvia and Byetta--or make the same amount of insulin more effective, as is the case with Actos and Avandia.
But when you take people off all these drugs, within weeks their blood sugar control goes back to where it was before they started the drug, except in the case of the people for whom Byetta causes dramatic weight loss, who may have been blood sugars because they have 100 lbs less body weight and hence are more insulin sensitive.
Were a drug to actually regenerate beta cells, you'd expect to see blood sugar control increase over the long term and you'd also expect to see sustained better blood sugar control when the person stopped taking the drug. That neither of these effects occurs makes it very unlikely that these drugs are having this effect.
One cannot but wonder if perhaps there are similar studies to this one kept locked in drug company files, studies run to "prove" that Januvia or Byetta increased beta mass, whose publication was suppressed when the result was not found. The drug companies have been pushing this claim so hard for these drugs, that you would expect to see autopsy studies by now, years after the drug has been in use. And we know for a fact that drug companies have a long history of hiding or suppressing the publication of studies whose findings might lessen sales.
In any case, if you wonder about how your own beta cell mass is doing--and who doesn't--it would be nice to get that 15 minutes oral glucose tolerance test with C-peptide measurement that might give you a good read on this, but you won't. It isn't a standard lab test and doctors aren't likely to notice this particular research study since, because it doesn't promote a drug or confirm a strongly held belief, no one is going to bring it to their attention.
But what you can do is this: You can ask for a fasting C-peptide with glucose measurement every year, compute the C-peptide/glucose ratio, and track how it is progressing over time. This should give you some sense of how your beta mass is holding up.
If your doctor insists that you take Januvia or Byetta to "regrow your beta cells" insist on getting the fasting C-peptide test with fasting glucose measurement before you start the drug and then a year later. If you don't see improvement in the C-peptide/glucose ratio, you will know that the drug company claim is false. Then, if the drug hasn't made a significant improvement in your blood sugars, you'll know there is no reason for you--or your insurance company--to keep spending almost $200 a month on the drug.
Of course, if you have a personal history of cancer or a strong family history of cancer, you should not be taking Januvia, but that's another issue. Details here: Januvia and Cancer.