Dear Brian,
What I mentioned in
passing was about using a fixed glucose load and differentiating between normal
and impaired glucose tolerance by the 2-hr value in individuals with different
body size. It is not about individuals
who are normal and can tolerate a very large load of glucose, nor those who are
clearly diabetic, as we tend not to use OGTT for these. Perhaps you would help me get my small
head around this.
We increase the glucose
load up to a weight of 42 kg. What is
so special about a body weight around (a mean of) 42 kg? In a healthy person who is growing in height
at 42 kg, or in different individuals with the same body composition but
different body size (not BMI**), does the ‘rate of glucose uptake per unit body
mass’ change inversely and linearly with an increase in normal body size?
For
illustration, imagine we have three brothers who are genetically identical,
except that they have a different number of copies of a gene that has resulted in
them growing to perfectly normal adult weights of 50, 60 and 70 kg. As with their other organs, their pancreatic
islet cells have developed and grown to cope with the size of their
bodies. Apart from body size, they are
phenotypically absolutely identical, have the same circulating hormone
concentrations, insulin production and sensitivity for a given plasma glucose,
have the same life style, and eat in proportion to their size to maintain their
weights. They are perfectly
healthy.
We
give them a 75-g OGTT. Would they
all have similar plasma glucose peaks?
I think it is likely that they would. Would they have similar 2-hr glucose levels? I think the answer would also be most
probably yes. Have they achieved
this by secreting exactly the same amount of insulin per whole pancreas and by
taking up exactly the same amount of glucose per unit body weight? I think the answer would be no, as the
smallest one has had to deal with more absorbed glucose per circulating volume
and unit body weight. But this is not
an issue, because each has a large spare capacity for handling glucose.
Now
imagine all three eat proportionately more, and lay down exactly the same
amount of fat relative to their size with exactly the same body fat
distribution. They continue getting fatter at exactly the same rate relative to their size, and continue to be
absolutely identical but for their size. At any given time, would their insulin
insensitivity not be exactly the same for a given plasma glucose (which is
different from absolute glucose load) and for the sugar in the food that they
are eating (relative to their body weights)?
Would they therefore not have a similar reduction in glucose uptake (in mole
glucose per litre of plasma per unit time)? Now if their increase in fat stores and insulin insensitivity
progressed, the reserve capacity for glucose uptake and handling per unit time
of which individual would be reached sooner for a given 75-g circulating
load of glucose??
This is a ridiculous
example and an oversimplified scenario, but my naive point is that is it
unreasonable to assume that different individuals with different size
(according to race for example) reaching the same 2-hr IGT threshold in a 75-g
OGTT may have different degrees of impairment of glucose tolerance? And therefore can’t we be categorising into
normal and IGT different patients that have a similar glucose intolerance by
using a fixed dose of glucose that for some may be slightly inappropriate for
their normal size? And since most of
the patients that we assess by an OGTT are those that have had borderline
fasting or random glucose levels, is it not the case that a significant number
can fall near our threshold and therefore be misclassified? Now it would be impossible to even imagine
doing tiny adjustments to either the glucose load or the 2-hr threshold
according height and surface area and body composition etc (by some magic
formula). So it seems to me that if we
are rightly so fussed about using exactly 75 g of glucose with so many
potential errors associated with giving the right dose, we should, as was my
point, perhaps also consider adding this size factor to the list of things that
are imperfect about the OGTT and the limitations in interpreting the 2-hr
results.
I have no evidence for any
of what I have said, but this is a discussion forum. So please do comment and correct me (and tell me if I have pushed
my foot deep into my mouth).
Best wishes,
Reza
**
As size increases when all aspect ratios are kept absolutely identical, BMI
does not remain static, and also increases but by cubic root of the increase in
size. So when the size (equaling volume
and virtually equaling weight) of a kid doubles while he keeps his shape
unchanged, his BMI increases by about 25%.
The BMI of our healthy 70-kg man is about 12% higher than that of his
identical, but smaller, 50-kg brother.
According to dogma, the heavier healthy man with a higher BMI, which is
only attributable to his variation of a normal size, should be more insulin-resistant
compared with his lighter brother.
Reza Morovat
Clinical Biochemist
Oxford
--------
Steady on! You've already hinted that GTT's ought to be tailored to the patient's size. Isn't that enough debunking for one day?
Brian