Re: insulin calc using glucose equivalent (LONG)

[Guest guest]

This is a mind-bogglingly detailed post by Ron Sebol of the low-carbing

diabetics' group. You really have to appreciate the attention these

number-crunching " diabetes nerds " put into these calculations! We all can

benefit from their incredible dedication!

Susie

*******************

(note: GI...glycemic index)

My earlier postings said take the carb grams and multiply by the GI and

that was wrong. That will not produce a correct glucose equivalent.

I would like to thank those of you who noticed that earlier error.

Though I would like to thank you, unfortunately I can't because nobody

noticed.

Tch tch gang. Unfortunately, the process is more involved than I had

indicated. My error was the result of my forgetting to address how GI

table entry portions were determined. Specifically, that GI test portion

sizes were such that there was a net 50 gram carb yield of each item

tested to produce the GI list. About a year earlier I posted a correct

version and somehow forgot in the nonce.

A few times recently I have talked about working up an insulin dose with

a self calibration derived from pure glucose (using the glucose tablets)

and I have alluded to (or given wrong instructions concerning)

calculations that convert food portions in a meal to their respective

equivalants in pure glucose. That is a meaningful calculation because

the glucose equivalent for a mixed meal is the sum of the glucose

equivalents of each item in the meal. It is not necessary to do a unique

mixed meal

experiment for every conceivable different mix of menu items because the

cumulative effect of the menu is the sum of the effects of the menu

items and the interractions are negligable as long as huge amounts of

fat or fiber are not thrown in. There are formal published studies

showing that it is legitimate to simply add contributions in this

manner. The precision of glucose tabs is the justification for using them to

set up

our insulin self calibration and, thereafter, we only need to express a

meal as being equivalent to some amount of glucose to know how to do a

sliding scale dose with exquisiste accuracy. Below, I will go through

the steps required to scope out a food item we wish to add to our quick

reference 3 X 5 cards (or spread sheet or whatever).

Using a piece of white bread as an example of the calculation process,

here is how the numbers work. To be able to follow the logic, you will need

to remember that the tested portions of all the items on the GI were sized

to contain a net 50 grams of carbohydrate. For additional GI lore, Rick's

site is recommended.

See http://www.diabetesdigest.com/april_glycemic_intro.htm

Looking in a food data base, such as

http://www.nal.usda.gov/fnic/cgi-bin/nut_search.pl

we learn that one slice of commercially prepared white bread, having a

gross weight of 25 grams, has a carb content of 12.375 grams.

Since the GI test portions must each NET 50 grams of carb (by definition

of how GI was measured), and since the net carb fraction of bread is

12.375/25 = .495 (the net carb grams divided by the gross portion

weight) then there were 101 GROSS weight grams of bread required to

produce that 50 net grams of carb. That last number of 101 comes from

50/.495 = 101. The result is checked noting .495 X 101 = 50

We also know from Rick Mendosa's GI list at

http://www.mendosa.com/gilists.htm

that the GI of bread is 101 and that the GI of pure glucose is 137.

Putting that together we conclude that it would take a gross weight of

1.37 X 101 = 138.4 gross weight

grams of bread to do the same thing to BG as 50 gross weight grams of

glucose. Of course the gross weight portion of glucose in the GI table

was also 50 grams, the same as its net weight, because it is 100%

glucose.

One useful bread number, therefore, is 50/138.4 = .36127 gm glucose

equivalent per each gross gram of bread. We would round that to .36

Therefore, one 25 gram slice of bread, with its 12.375 grams of carb,

had a glucose equivalent of 25 X .36 = 9 gms of glucose.

After doing such a calculation for bread, what we would retain is that

there are .36 gms equiv glucose in each gross gram of bread.

Or we could say, noting that 9.124/12.375= 73%, that if we know the net

carb grams in our sliding scale bread portion then 73% of the net carb

grams is the glucose equivalent grams of that portion.

In general, using the net 50 carb gram definition constraint on GI

entries and the net carb grams per gross gram of item from a food data

base, find the gross portion weight that went into the GI table item.

Multiply that gross portion weight by 137 and then divide by the GI of

the item. That will be the gross portion weight that would yield the

same BG result as 50gms of glucose. Once that is known, a realistic

portion can be extracted by a simple proportional scaling up or down

from the GI portion.

If you calibrate using goucose tabs, also consider each cooked ounce of

meat as yielding the equivalent of 4.2 gm of glucose. That comes from 7

gms protein per oz meat (for most meats) and from the fact that 60% of

meat protein converts to glucose. The 4.2 is 60% of 7. In milk proteins,

the conversion is slightly under 50%.

I apologize for the oversimplifications of the earlier postings. This is

much more likely to be right. I hope I am not asked to boil this down

more. This distillation is as boiled as I know how to render it. As to

why bother? Thisnumber crunching is THE most precise way of determining an

insulin

dose. It is the which, than which, there is no whicher.

Regretably, many ordinary items are not seen in the GI tables. I suggest

that using the nearest similar item will be close enough. That will

still be far far better than the 200% dosing errors one can end up with

if cozened to except the inexact ADA food exchange tables.

Of course we only need to plod through the entire process shown above

once per food item of interest. Thereafter we would get the salient info

from a 3 X 5 card with our favorite foods listed along with their

glucose equiv. per gross gram or per carb gram or both. We might even

want to do it on the basis of cups (food volume) rather than weight. The

calculations would be very similar to those shown.

Please give me some feedback on this posting. What I am interested in is

how many of you understood the process of calculation described above.

If you understood, would you bother to use this method or is it too much

trouble to set up the system?

Ron