Good A1C's are not enough--How's your SD?

Thread view:

Enable EMail Alerts

Start New Thread

Thread rating:

Uncle Enrico - 08 May 2008 14:41 GMT

Two blood sugar readings illustrate the problem: 180 and 40 average out
to 110. While two readings of 110 may be OK, the 180 and the 40 are not OK.

We have to pay attention to our blood glucose variability as measured by
standard deviation in addition to looking at averages.

I maintained A1C's from 5.1 to 5.6 with 5.3's being most typical for ten
years and ended up with a painful neuropathy after 10 years post diagnosis.

Since the neuropathy began, I've reduced my standard deviation and
maintained good A1c's and have made improvements in the neuropathy. It's
still with me, however.

Read this stuff:

http://diabetesmonitor.com/m57.htm

http://tinyurl.com/38knlc

http://tinyurl.com/36f4j8

Here's an online SD calculator:

http://invsee.asu.edu/srinivas/stdev.html

You can also use Excel to calculate your Average and your SD.

The two formulas you'll need are =AVERAGE(range) and =STDEVP(range)
for standard deviation.

An example of (range) would be the cell addresses or numbers in the
range of your readings: example =STDEVP(b12-b33) assuming your
readings run from cell b12 through b33.

Also, most meter software programs do these calculations and much more.

Accu-Check gave me their Compass software for free when I phoned and
asked them about it.

Reply to this Message

Uncle Enrico - 08 May 2008 14:44 GMT

> Two blood sugar readings illustrate the problem: 180 and 40 average out
> to 110. While two readings of 110 may be OK, the 180 and the 40 are not
[quoted text clipped - 35 lines]
> Accu-Check gave me their Compass software for free when I phoned and
> asked them about it.

Merde! That formula for standard deviation in Excel should be =STDEVPA
(range).

Reply to this Message

Uncle Enrico - 08 May 2008 14:50 GMT

Double Merde!!! That formula for standard deviation in Excel should be

=STDEVPA(range)
or
=STDEVPA(b3:b29)
assuming the cell range of your readings runs from cell b3 to b29

Don't leave a space between =STDEVPA and the (range)

and don't use a dash between the cell numbers in the range. Use a colon).

Reply to this Message

Quentin Grady - 09 May 2008 22:44 GMT

>Two blood sugar readings illustrate the problem: 180 and 40 average out
>to 110. While two readings of 110 may be OK, the 180 and the 40 are not OK.
[quoted text clipped - 4 lines]
>I maintained A1C's from 5.1 to 5.6 with 5.3's being most typical for ten
>years and ended up with a painful neuropathy after 10 years post diagnosis.

That must have been a painful realization.

>Since the neuropathy began, I've reduced my standard deviation and
>maintained good A1c's and have made improvements in the neuropathy. It's
>still with me, however.

I'm glad you've been able to get some improvement.

Rather than measuring standard deviation you could simply record the
times when the blood glucose levels went over some prefixed level eg
140 mg/dL Damage occurs when post prandial glucose levels go over
quite low figures for a couple of hours.

>Read this stuff:
>
[quoted text clipped - 7 lines]
>
>http://invsee.asu.edu/srinivas/stdev.html

Best wishes,

Signature

Quentin Grady ^ ^ /
New Zealand, >#,#< [
/ \ /\
"... and the blind dog was leading."

http://homepages.paradise.net.nz/quentin

Reply to this Message

Uncle Enrico - 10 May 2008 14:35 GMT

> Rather than measuring standard deviation you could simply record the
> times when the blood glucose levels went over some prefixed level eg
> 140 mg/dL Damage occurs when post prandial glucose levels go over
> quite low figures for a couple of hours.

True enough Quentin, especially for those not ready to get fussy. On the
other hand when looking at six months' worth of readings, it's helpful
to have a finite number like standard deviation to get a more definitive
picture of bg variability over that period. One can also compare SD
numbers pre and post complications onset to judge whether the
diet/exercise/insulin changes that one makes is working on the numbers
and the condition.

If you read the Hirsch item, this is the first DM recommendation I've
seen on SD values as a percentage of bg average with respect to
complications. I expect there will be more info on these measurements if
researchers can find enough test subjects willing to test often and keep
copious records.

All one really needs is meter software, as I recently discovered. And,
with Accu-Chek, it was free for the asking.

I did put together an Excel spreadsheet that calculates bg average, SD
and graphs the values for 15 readings. If anyone is interested, I'll
send it along as an attachment, or I can post it on alt.binaries and
identify it for those interested.

>> Read this stuff:
>>
[quoted text clipped - 9 lines]
>
> Best wishes,

Reply to this Message

Uncle Enrico - 10 May 2008 14:38 GMT

> I did put together an Excel spreadsheet that calculates bg average, SD
> and graphs the values for 15 readings. If anyone is interested, I'll
> send it along as an attachment, or I can post it on alt.binaries and
> identify it for those interested.

Just checked alt.binaries
It's nothing but porno.
Maybe there's another newsgroup?

Reply to this Message

Robert Miles - 10 May 2008 19:21 GMT

>> I did put together an Excel spreadsheet that calculates bg average, SD
>> and graphs the values for 15 readings. If anyone is interested, I'll send
[quoted text clipped - 4 lines]
> It's nothing but porno.
> Maybe there's another newsgroup?

There are quite a few with names starting with alt.binaries.. Does
alt.binaries.freeware look suitable?

Reply to this Message

Uncle Enrico - 11 May 2008 15:08 GMT

>>> I did put together an Excel spreadsheet that calculates bg average, SD
>>> and graphs the values for 15 readings. If anyone is interested, I'll send
[quoted text clipped - 6 lines]
> There are quite a few with names starting with alt.binaries.. Does
> alt.binaries.freeware look suitable?

Hi Robert. Thank you for the suggestion. I posted the Windows 2000 Excel
spread sheet on alt.binaries.freeware this morning May 11, 2008, at 8:57
am under the subject "Blood Glucose and Standard Deviation" and my name,
Uncle Enrico.

The spread sheet is for 15 readings. If anyone would like more or fewer
readings, you'll need to unprotect the sheet and make the changes. Or,
let me know and I'll be happy to make the changes...just let me know the
number of readings you'd prefer.

Perhaps someone knows how to devise a sheet with a flexible range of
readings? Please advise. I'm always happy to learn from others.

I will also post the spread sheet to Oleg's website.
Thank you Oleg!

Reply to this Message

Uncle Enrico - 11 May 2008 17:11 GMT

> Perhaps someone knows how to devise a sheet with a flexible range of
> readings? Please advise. I'm always happy to learn from others.

It just dawned on me how to create a more flexible range. I've just
uploaded another spread sheet that allows 50 or fewer readings. It also
identifies highest and lowest readings and includes a graph.

Reply to this Message

Oleg Lego - 12 May 2008 05:51 GMT

>> Perhaps someone knows how to devise a sheet with a flexible range of
>> readings? Please advise. I'm always happy to learn from others.
>
>It just dawned on me how to create a more flexible range. I've just
>uploaded another spread sheet that allows 50 or fewer readings. It also
>identifies highest and lowest readings and includes a graph.

Which one(s) would you like posted, Uncle E? I rather like your
latest (502). I could put them all up if you like, or just the best
one.

Signature

roses are #FF0000
violets are #0000FF
all my base
are belong to you

Reply to this Message

Uncle Enrico - 12 May 2008 17:29 GMT

>>> Perhaps someone knows how to devise a sheet with a flexible range of
>>> readings? Please advise. I'm always happy to learn from others.
[quoted text clipped - 6 lines]
> latest (502). I could put them all up if you like, or just the best
> one.

Thanks to Oleg, the spreadsheet is now posted to his website. He's also
added a helpful IF statement that avoids DIV/0 error statements when the
input column is blank.

I'll let Oleg post the link.

Thanks, Oleg.

BTW, if your meter has tracking software available, you will get many
more features than this spreadsheet can offer.

Accu-Check gave me their Compass software for free when I inquired.
Unfortunately, their infrared reading device is a serial/com device and
not USB capable unless you use their adapter. The adapter does not work
with Vista and it barely works with XP. Fortunately, it worked with my
XP desktop long enough to input 4 months worth of readings. Then, a few
weeks later, Windows sent along an update which caused the adapter to
generate Blue Screens of Death.

Reply to this Message

Oleg Lego - 12 May 2008 23:10 GMT

>>>> Perhaps someone knows how to devise a sheet with a flexible range of
>>>> readings? Please advise. I'm always happy to learn from others.
[quoted text clipped - 12 lines]
>
>I'll let Oleg post the link.

You can find Uncle Enrico's spreadsheet at:
http://furble.sasktelwebsite.net/uncleenrico.html

>Thanks, Oleg.

No problem.

Signature

roses are #FF0000
violets are #0000FF
all my base
are belong to you

Reply to this Message

Oleg Lego - 10 May 2008 22:41 GMT

>> I did put together an Excel spreadsheet that calculates bg average, SD
>> and graphs the values for 15 readings. If anyone is interested, I'll
[quoted text clipped - 4 lines]
>It's nothing but porno.
>Maybe there's another newsgroup?

If you'd like, I can put it on my web site. You can email it to me at
lar3ry at sasktel dot net

Tell me the name you want it available as, and I'll put it up right
away.

Signature

roses are #FF0000
violets are #0000FF
all my base
are belong to you

Reply to this Message