Spliting doses of Humalog?
View Full Version : Spliting doses of Humalog?
I am interested in any information on splitting the Humalog
dose and multiple injections. If a dose was split the
surface area of the total dose would be larger. Would this
alter the time curves for an injection. Guy
Guy wrote in message ...
>
>I am interested in any information on splitting the Humalog
>dose and multiple injections. If a dose was split the
>surface area of the total dose would be larger. Would this
>alter the time curves for an injection. Guy
I think Jim Dumas has experimented with such techniques but
I don't know what he found.
Bernstein asserts that doses of insulin greater than about 7
units tend to produce erratic and unpredictable absorption.
He recommends splitting all doses of all insulin into
multiple injections of 7 units or less.
He claims that doses higher than 7 units slows the Lentes.
IIRC, he didn't mention any effects on Humalog/Novolog. I
suspect that the larger doses would be slowed.
My most common "large dose" of Humalog is 8 units so I
haven't bothered to look for any size effects.
Regards
Old Al
Guy wrote:
>
> I am interested in any information on splitting the
> Humalog dose and multiple injections. If a dose was split
> the surface area of the total dose would be larger. Would
> this alter the time curves for an injection.
OK Guy,
I decided to plot a graph of the flat 2-split Humalog
doses. But I also decided to set the peak action of each
dose to 1.5 hours and not the 2.25 hours as the maximum
action time in the Eli Lilly 1993 data. I found that the
tail of the first dose is too long for the Rayleigh half
amplitude summation to work properly. That is, the insulin
action profile is not symmetric around the peak time and
therefore the Rayleigh sum doesn't work that well. I
changed the 2.6 hour dose of the previous "back of envelope
analysis" to 75% and it gives a flatter profile with >80%
maximum from 1.6 hours to 5.4 hours. So this long action
time of 5.4 hours is too long.
So I changed the model to 3 equal compartments with 1.5
hour peak action. This change is more physiological for
Humalog, IMO. Then set the second dose to 50% of the first
dose and this is taken 2 hours later. I placed a pdf of
this on the web at:
http://j_dumas.home.mindspring.com/guy-humalog.pdf
I've mentioned the math for this a few years ago on MHD. But
can give refs again for those who are interested.
In any case, this will give you a better idea of what
to expect.
Take a look and tell me what you think,
--
Jim Dumas T1 4/86, background retinopathy, rarely
hypoglycemic: <1/mo. lispro+R+U+NPH daily, moderate
exercise, typically <6% HbA1c
On Thu, 1 Jul 2004 22:55:28 -0400, "oldal4865" <oldal4865@yahoo.com>
wrote:
>
>Guy wrote in message ...
>>
>>I am interested in any information on splitting the
>>Humalog dose and multiple injections. If a dose was split
>>the surface area of the total dose would be larger. Would
>>this alter the time curves for an injection. Guy
>
> I think Jim Dumas has experimented with such techniques
> but I don't know what he found.
>
>Bernstein asserts that doses of insulin greater than about
>7 units tend to produce erratic and unpredictable
>absorption. He recommends splitting all doses of all
>insulin into multiple injections of 7 units or less.
>
>He claims that doses higher than 7 units slows the Lentes.
>IIRC, he didn't mention any effects on Humalog/Novolog. I
>suspect that the larger doses would be slowed.
>
>My most common "large dose" of Humalog is 8 units so I
>haven't bothered to look for any size effects.
>
> Regards
> Old Al
>
It is interesting that Lantus is one daily dose. But to
splitting it seems to generate some other questions, such of
maintaining a level release. If anyone comes across any real
date on these questions. Please post it.
With no natural insulin I have to careful with any leading
insulin with food input. My thought is there might be a way
to split the dose to reduce the glucose peaks without a
hypo. risk.
With pumps is it reasonable to customize the releases to
match food absorption?. An interesting subject. Guy a
oldal4865 wrote:
>
> Guy wrote in message ...
>>
>>I am interested in any information on splitting the
>>Humalog dose and multiple injections. If a dose was split
>>the surface area of the total dose would be larger. Would
>>this alter the time curves for an injection. Guy
>
> I think Jim Dumas has experimented with such techniques
> but I don't know what he found.
>
> Bernstein asserts that doses of insulin greater than about
> 7 units tend to produce erratic and unpredictable
> absorption. He recommends splitting all doses of all
> insulin into multiple injections of 7 units or less.
>
> He claims that doses higher than 7 units slows the Lentes.
> IIRC, he didn't mention any effects on Humalog/Novolog. I
> suspect that the larger doses would be slowed.
>
> My most common "large dose" of Humalog is 8 units so I
> haven't bothered to look for any size effects.
Hi Al and Guy,
Just to say that I used 4 equi-spaced doses of Humulin R
about the umbilicus in the early 1990s. This simultaneous
dosing did have a more rapid fall in BG relative to a single
large dose of Humulin R. But when I started to use Humalog,
one single large dose was fine. Moreover, I want it to last
longer when hyperglycemic, because I usually have a BG
minimum of ~180
mg/dl (stalls here) if over a 500 mg/dl starting point.
So splitting the Humalog dose was not viable since it
was too fast already for me. I typically have a free
fall in BG of about 100 mg/dl/hour. So in 4 hours I
should be about 100 mg/dl, if starting at a 500 mg/dl.
But this usually ends at 180 mg/dl since the Humalog
wanes quickly.
I suspect the 4 equal equi-spaced Humulin R doses would be
bound and greatly delayed by my high antibody titer these
days. So this method would not be that effective now. The
antibody concentration is the rate-limiter and something we
have no control over.
Next, the 7U magic dose size will be clobbered by the
antibody concentration as well. So long-term T1s will
probably not get much bang from the 7U bolus, since antibody
titers continue to grow with long-term use of insulin. But
Eli Lilly did show, at the 1993 ADA convention, that there
is an oscillatory behavior in peak insulin action with bolus
size. It has a minimum action peak (fastest) at 7U and 21U
for R and Humalog. But it has a maximum at 14U and 28U
(slowest) for these preparations as well. Bolus sizes
smaller that 7U had longer peak action times for some
paradoxical unknown reason. I have an hypothesis that these
small bolus sizes do not have many capillary beds nearby,
relative to the 7U and 21U dose sizes. So if we placed a
single insulin hexamer at a point in the middle of
capillaries, how long for diffusion to push it up against a
capillary so it could be absorbed into the bloodstream? But
if a 7U bolus of U-100 just covers the maximum number of
capillaries per dose volume, it would be absorbed faster. If
the volume is the critical value, then we should see this
magic number of 7/2.5=3U for U-40 data. But I have yet to
look for this in the literature.
The Eli Lilly glucose-clamp data, Diabetes, vol 42, suppl 1,
1993, p.54A:
SC Peak Action Time (max glucose uptake) in
Minutes Dose Humulin R Humalog Size
3. 6U 152 105
4. 3U 145 93
5. 8U 240 135
6. 0U 210 101
7. 8U 195 130
As usual, this is in 18 normals with no previous insulin
use. So the antibody concentrations to the manufactured
insulin (as in covalently bonded dimers that have a very
long half-life in the bloodstream) are zero.
In any case, I think this 7U magic number comes from this
data. The original data was in normalized dose of U/kg body
weight. But I called up the Lilly Labs researcher and asked
for the mean body weights for this article, so I could work
back to absolute dose sizes.
Finally, if I were to split Humalog doses today, it would be
to delay the other doses to prolong insulin action. So I
would probably use the 14-15U dose on a regular basis.
Food for thought,
--
Jim Dumas T1 4/86, background retinopathy, rarely
hypoglycemic: <1/mo. lispro+R+U+NPH daily, moderate
exercise, typically <6% HbA1c
On Sat, 03 Jul 2004 04:04:33 GMT, Jim Dumas
<j-dumas@.no.SPAM!mindspring.com> wrote:
>Guy wrote:
>
>>
>> I am interested in any information on splitting the
>> Humalog dose and multiple injections. If a dose was split
>> the surface area of the total dose would be larger. Would
>> this alter the time curves for an injection.
>
>OK Guy,
>
>I decided to plot a graph of the flat 2-split Humalog
>doses. But I also decided to set the peak action of each
>dose to 1.5 hours and not the 2.25 hours as the maximum
>action time in the Eli Lilly 1993 data. I found that the
>tail of the first dose is too long for the Rayleigh half
>amplitude summation to work properly. That is, the insulin
>action profile is not symmetric around the peak time and
>therefore the Rayleigh sum doesn't work that well. I
>changed the 2.6 hour dose of the previous "back of envelope
>analysis" to 75% and it gives a flatter profile with >80%
>maximum from 1.6 hours to 5.4 hours. So this long action
>time of 5.4 hours is too long.
>
>So I changed the model to 3 equal compartments with 1.5
>hour peak action. This change is more physiological for
>Humalog, IMO. Then set the second dose to 50% of the first
>dose and this is taken 2 hours later. I placed a pdf of
>this on the web at:
>
>http://j_dumas.home.mindspring.com/guy-humalog.pdf
>
>I've mentioned the math for this a few years ago on MHD.
>But can give refs again for those who are interested.
>
>In any case, this will give you a better idea of what
>to expect.
>
>Take a look and tell me what you think,
Thanks Jim, I think I need to look more at my options. I
need to input more food without the peaks. The options
seems is to add a second test and inject. I looked at your
info on dose size. It seems that it may fit. I had allergy
problems long before I develop demonstrated diabetes. I
have never been clear on the immunity system but I think it
is a big factor.
I appreciate your efforts. Thanks. Will go through a few
strips and see the fit. Guy Once an old doc told me that
he noted that some patients with allergies later
developed diabetes.
Guy wrote in message ...
>>
>It is interesting that Lantus is one daily dose. But to
>splitting it seems to generate some other questions, such
>of maintaining a level release. If anyone comes across any
>real date on these questions. Please post it.
>
>With no natural insulin I have to careful with any leading
>insulin with food input. My thought is there might be a way
>to split the dose to reduce the glucose peaks without a
>hypo. risk.
>
>With pumps is it reasonable to customize the releases to
>match food absorption?. An interesting subject. Guy
>
The pumpers do have options intended to match insulin
input to food absorption. I think they call it the
"square wave" option. AFAIK, if a pumper asks his pump to
deliver a bolus under normal condition, it attempts to
inject insulin as fast as it can (which isn't too fast).
The "square wave" option slows it down.
I attend a Pumpers' Support Group meeting but they haven't
gotten into that yet.
I couldn't find anything specific on Lantus doses in
Bernstein. He just keeps repeating the 7 unit per
injection rule.
Regards
Old Al
Jim Dumas wrote:
> Finally, if I were to split Humalog doses today, it would
> be to delay the other doses to prolong insulin action. So
> I would probably use the 14-15U dose on a regular basis.
I did a simple "back of the envelope" analysis with a 5
equal compartment insulin transport model to estimate the
time delay for a second dose to "square-off" the insulin
action. The basic idea is to use the Rayleigh, (as in
resolving star spacing in astronomy), half intensity points
and overlap these half amplitude values. My estimate is to
delay the second bolus of 14-15U Humalog by 2.6 hours, or
156 minutes to get a flat squarewave-like insulin action.
This assumes 14-15U Humalog dose sizes with peak action at
135 minutes from the Eli Lilly 1993 data. My simple model
shows the early 50% point is at
1.2h and the late 50% point is at 3.8h. To get the half
amplitude overlap, we need 3.8-1.2=2.6 hours for the
second dose, relative to the first dose. This will give a
flat response from roughly 1-5 hours. (And don't blame me
for your hypoglycemia!)
But if I were going to do this, I'd just take a single dose
of Humulin R instead. Humulin R has a peak insulin action of
5.7 hours post-dose for my antibody binding metabolism. So
that's my pizza dose.
More food for thought,
--
Jim Dumas T1 4/86, background retinopathy, rarely
hypoglycemic: <1/mo. lispro+R+U+NPH daily, moderate
exercise, typically <6% HbA1c
Guy wrote:
> I appreciate your efforts. Thanks. Will go through a few
> strips and see the fit.
You're welcome, Guy.
I just plotted another of 3 split Humalog dosing. This is
the same with another 50% dose at 3.5 hours after the first
dose. I added some colors to the graph to separate the
first, second and third doses. I also doubled the resolution
to 401 time points over 10 hours. This is up at:
http://j_dumas.home.mindspring.com/humalog-3split.pdf
The third dose is very risky, IMO. But it was a natural
progression from the previous graph of 2 split doses. This
"Pizza Dosing" has strong action from 0.9 to 5.2 hours. I'll
probably try this on my next pizza. But will test prior to
dosing, of couse.
Will probably use it at Thanksgiving too. Food for thought,
--
Jim Dumas T1 4/86, background retinopathy, rarely
hypoglycemic: <1/mo. lispro+R+U+NPH daily, moderate
exercise, typically <6% HbA1c
oldal4865 wrote:
> Guy wrote in message ...
>
>>It is interesting that Lantus is one daily dose. But to
>>splitting it seems to generate some other questions, such
>>of maintaining a level release. If anyone comes across any
>>real date on these questions. Please post it.
>>
>>With no natural insulin I have to careful with any leading
>>insulin with food input. My thought is there might be a
>>way to split the dose to reduce the glucose peaks without
>>a hypo. risk.
>>
>>With pumps is it reasonable to customize the releases to
>>match food absorption?. An interesting subject. Guy
>>
>
> The pumpers do have options intended to match insulin
> input to food absorption. I think they call it the
> "square wave" option. AFAIK, if a pumper asks his pump
> to deliver a bolus under normal condition, it attempts
> to inject insulin as fast as it can (which isn't too
> fast). The "square wave" option slows it down.
>
> I attend a Pumpers' Support Group meeting but they
> haven't gotten into that yet.
>
> I couldn't find anything specific on Lantus doses in
> Bernstein. He just keeps repeating the 7 unit per
> injection rule.
>
> Regards
> Old Al
>
>
minimed pumps inject at the rate of 1/10 unit every 4
seconds. SOME other brands inject too quickly for my tastes.
There are 2 schools of thought on the problems associated
with fast injections. I prefer MiniMed's philosophy of
slooow and easy.
"square wave" is one option and "dual wave" is a combination
of an immediate bolus, followed by a square wave. I call it
the "pizza bolus". :)
dave
Jim Dumas wrote:
> Jim Dumas wrote:
>
>
> To get the half amplitude overlap, we need 3.8-1.2=2.6
> hours for the second dose, relative to the first dose.
> This will give a flat response from roughly 1-5 hours.
> (And don't blame me for your hypoglycemia!)
>
> But if I were going to do this, I'd just take a single
> dose of Humulin R instead. Humulin R has a peak insulin
> action of 5.7 hours post-dose for my antibody binding
> metabolism. So that's my pizza dose.
Great information!!! Jon
Jim Dumas wrote:
> http://j_dumas.home.mindspring.com/humalog-3split.pdf
I added a few more comments to the graph and posted a
new version.
The most important point is the area under each section is
the respective dose of Humalog. This is how the peaks are
scaled, by using the area under each curve section.
The next point is we could continue to add 50% doses every
1.5 hours to get a relatively flat profile (dangerous of
course). So another 50% of initial dose could be added at 5
hours post initial dose to continue this progression. (But
don't do it!)
Happy 4th!
--
Jim Dumas T1 4/86, background retinopathy, rarely
hypoglycemic: <1/mo. lispro+R+U+NPH daily, moderate
exercise, typically <6% HbA1c
JonK wrote:
> Great information!!!
Hi Jon,
I'd been eyeing this thread of Guy's for a few days and
finally decided to jump in.
Thanks,
--
Jim Dumas T1 4/86, background retinopathy, rarely
hypoglycemic: <1/mo. lispro+R+U+NPH daily, moderate
exercise, typically <6% HbA1c
Is your dosing designed for New York pizza or Chicago
pizza? Jon
JonK wrote:
> Is your dosing designed for New York pizza or Chicago
> pizza? Jon
My MiniMed pump is designed to handle Roundtable and
Little Caeser's pizzas. :) AAMOF, because it's so warm
this afternoon, instead of a BBQ, my wife and I will pick
up a pizza...
dave
JonK wrote:
> Is your dosing designed for New York pizza or
> Chicago pizza?
Hi Jon,
I decided to try the 2-split Humalog dosing today.
The wife wanted to BBQ delmonico steaks with double size
baked potato and two scoops of regular sour cream. Some
asparagus in butter then she made a chocolate mousse with
cool whip topping. Now that's a pizza equivalent. I was at
100 mg/dl at 1 pm when I dosed 26U of Humalog (I use quite
alot of insulin). At 2 hours (~3:15pm) I was at 175 mg/dl
and dosed 10U Humalog. I just checked BG at 7:15 pm at it's
154 mg/dl. I'll check again at 9:15 pm. But it looks like I
should have taken the 50% dose of 13U at 2 hours. But I was
too chicken.
In theory, (with these large dose sizes), I should have
enough liver enzyme activity to achieve euglycemia after the
10U dose wanes in an hour. But we'll see.
In any case, all this fat and carbohydrate is a good test
for the Humalog split dosing. So I'm testing the theory.
But it ain't for the faint hearted,
--
Jim Dumas T1 4/86, background retinopathy, rarely
hypoglycemic: <1/mo. lispro+R+U+NPH daily, moderate
exercise, typically <6% HbA1c
Jim Dumas wrote:
> At 2 hours (~3:15pm) I was at 175 mg/dl and dosed 10U
> Humalog. I just checked BG at 7:15 pm at it's 154 mg/dl.
> I'll check again at 9:15 pm. But it looks like I should
> have taken the 50% dose of 13U at 2 hours. But I was too
> chicken.
There was another BG check at 4h post 1pm dose. It was 169
mg/dl. This was 2 hours after the 50% bumper dose of 10U.
I should also mention the wife loves dinner rolls. So I had
3 with butter. That increases the carbohydrate.
So this meal rivals a pizza but without the cheese.
--
Jim Dumas T1 4/86, background retinopathy, rarely
hypoglycemic: <1/mo. lispro+R+U+NPH daily, moderate
exercise, typically <6% HbA1c
Jim Dumas wrote:
> But it looks like I should have taken the 50% dose of 13U
> at 2 hours. But I was too chicken.
>
That's the operative problem. I'm always a little to chicken
to wack up that much insulin with no proof the food is still
operating.
Jon
Jim Dumas wrote:
>> At 2 hours (~3:15pm) I was at 175 mg/dl and dosed 10U
>> Humalog. I just checked BG at 7:15 pm at it's 154 mg/dl.
>> I'll check again at 9:15 pm. But it looks like I should
>> have taken the 50% dose of 13U at 2 hours. But I was too
>> chicken.
>
> There was another BG check at 4h post 1pm dose. It was
> 169 mg/dl.
Final BG just now (9:20 pm) was 189 mg/dl. So BG is rising
and I'll have to dose alittle more. There still seems to be
some late GI tract absorption and I haven't eaten (just 2
diet sodas) anything since the large lunch.
In any case, the 2 split Humalog dosing is interesting and
has merit. I'll try it again with a pizza this week.
If anybody else takes the plunge, I'd like to hear about it.
--
Jim Dumas T1 4/86, background retinopathy, rarely
hypoglycemic: <1/mo. lispro+R+U+NPH daily, moderate
exercise, typically <6% HbA1c
On Mon, 05 Jul 2004 01:31:53 GMT, Jim Dumas
<j-dumas@.no.SPAM!mindspring.com> wrote:
>Jim Dumas wrote:
>
>>> At 2 hours (~3:15pm) I was at 175 mg/dl and dosed 10U
>>> Humalog. I just checked BG at 7:15 pm at it's 154 mg/dl.
>>> I'll check again at 9:15 pm. But it looks like I should
>>> have taken the 50% dose of 13U at 2 hours. But I was too
>>> chicken.
>>
>> There was another BG check at 4h post 1pm dose. It was
>> 169 mg/dl.
>
>Final BG just now (9:20 pm) was 189 mg/dl. So BG is rising
>and I'll have to dose alittle more. There still seems to be
>some late GI tract absorption and I haven't eaten (just 2
>diet sodas) anything since the large lunch.
>
>In any case, the 2 split Humalog dosing is interesting and
>has merit. I'll try it again with a pizza this week.
>
>If anybody else takes the plunge, I'd like to hear
>about it.
As you know, I have been playing with test doses for years.
My responses to food is interesting but varies from what
some seem to think.
I can play these games because I have fewer variables. I
have no natural insulin and can control the glucose release.
If I had the energy I would do more. I am very interested in
your work. I want more on the immunity issue. more on
insulin activity times in the blood stream.
I see people posting as if it remained until used. Qualified
people in the past have said it only lasts for minutes.
Plenty left to cover in diabetes. Guy
Automatic Translations (Powered by
):
vBulletin, Copyright ©2000-2009, Jelsoft Enterprises Ltd.
Search Engine Friendly URLs by
vBSEO 3.3.0
