Type-1 Impact on the Brain

This posting is not about cure research, but is a summary of a talk about type-1's effect on the brain.

CarbDM  recently hosted a talk by Dr. Tandy Aye, who studies the impact of type-1 diabetes on developing brains.  That sounded to me like a very interesting talk, so I attended.  Below are my notes.  Please remember that these are my thoughts on the topic (not Dr. Aye's!) and that I only wrote a little during the talk, so most of this is from memory.

Historical Background

Diabetes effects on many parts of the body has been well known for decades, however study of type-1 diabetes's effects on the brain started in 1985.  Various IQ type tests have been used, and also various brain imaging technologies.  When a new brain imaging technique is developed, it is first applied to adults, and then to children, and then the next advance comes along, and the cycle repeats.

Doing brain imaging on child is more difficult than for adults, because most of these technologies require the patient to lie vary still, in a small, enclosed space, sometimes while the machine makes strange noises or vibrates.  Children are not good at this.  However, a major breakthrough was to have a monitor play movies that the kids could watch, and that would freeze if the kid moved.  This quickly teaches the kid not to move!

Brain Studies

Various brain studies have been published since 1985, mostly on adults, but a few on children.  A good summary (actually a meta-analysis) of the adult studies was published by Gaudieri in 2008.

Aye's pilot study included just under 30 children with type-1 diabetics, ran for a few years (3?), and is complete and published.   Dr. Aye had originally submitted a study proposal for a large multi-site study that would have told us a lot about how type-1 effects brain development.  That study was rejected because the reviewers thought that children would never sit still in an MRI machine long enough to get a clear image.  One of the main goals of this pilot study was to prove to the reviewers that children could do that (if properly motivated and supported).  This part of the trial was a complete success as the much larger study has now been approved.

A second, much larger study (about 200 children) that Dr. Aye is also participating in, has started, but is still underway.  Dr. Aye could only share some of the data from the patients when they started the study, but nothing about what happened during the study.

Results

There are three conclusions which I took away from the talk.  These conclusions come from Dr. Aye's completed small trial, from the initial measurements from the much larger trial, and (to a lessor degree) from the summary of all previous research, and especially the 2008 meta-analysis.

First, Type-1 diabetes does appear to effect the brain, although as a small effect.   For example, adults with type-1 have decreased grey matter density, on average, as compared to non-type-1s.  White matter (a different part of the brain) is expected to grow during childhood, but (on average) grows less or not at all in type-1 diabetics.  There were some differences in neurons, as well.  None of these changes pushed type-1 diabetics outside the normal range of human growth.  They were just statistical differences in overall populations.

Second, this effect (overall) is not good.  Gaudieri's meta analyses showed an over all decrease in IQ by about 5 points.  That's an average over all the people involved in all the different studies that were included in the meta analysis.  Verbal IQ scores seemed to be effected the most.  Some other IQ tests were not effected at all.  Another study showed that type-1s with A1c above 8.8 had a about 9% lower numbers than those with A1c below 7.4.  (My notes are unclear if this one one specific test, or all tests, or what.)

Note: I do not want to get into an argument about the accuracy or "truthfulness" of IQ tests.  If you don't believe in the accuracy or truthfulness of standard IQ tests, then you probably should not be reading this post at all.  Dr. Aye stressed that all of her tests were aimed at general IQ, not specific scholastic achievement.  Obviously, the arguments about the usefulness of these tests (compared to other tests, or no tests) are endless.

Third, type-1's effect on the brain is correlated with high BG levels, not with low BG levels.  (But see the "correlation vs. causation" discussion below.)  This focus on high BGs is a major change of view from the previous conventional wisdom.  I can remember very clearly being told (when my daughter was diagnosed, before 2008) that low BGs might effect brain development, and especially that there was nervousness that seizures specifically were bad for the developing brain.  But the data generally does not support this nervousness   Instead, it suggests that high BG numbers are correlated with whatever is hurting the brain.  (Much like high BG numbers are associated with other bad, long term side effects of type-1.)

The analysis of low BG looked at two different measures: first, the number of seizures that child had, and second, the overall-lifetime A1c.  The first is tied to "short deep" low BG numbers, and the second to "long average" low BG numbers.  But neither was associated with lower IQs.  (Side note: about 1/3 of the kids in the pilot study had at least one seizure.   I had no idea they were that common.)

The analysis of high BG numbers looked at he overall-lifetime A1c, and that showed a very clear correlation to lower Verbal IQ (not to other IQ measurements).  The graph showed generally higher verbal IQs at A1c of 6, and linearly dropping as the A1c rose to 9.

There was some discussion about "correlation vs. causation" for this data, as there should be.

One of the parents in the audience pointed out that the flow of causality might flow from IQ to low A1c, and not the other way around.  I thought this was an excellent point.  He pointed out that kids with higher IQ might remember more of their boluses.  They might be better at dosing, or they might just be more careful and thoughtful about managing their type-1 diabetes.  Therefore, high IQ might be a cause of lower A1c, and not the other way around (higher A1c causing lower IQ).  The big study is likely to provide strong data in this area, since it will measure IQ at the start and at the end of a multi-year test period.

Of course, we know that one of the long term bad side effects of type-1 diabetes is peripheral nerve damage ("Diabetic Peripheral Neuropathy") , and this is correlated with high A1c numbers.  So it is very reasonable to think that high A1c numbers might cause nerve damage in the brain as well.

So at the end of the day, we need to wait for the larger clinical trial to get data to answer the causality question.  And even then: one study is unlikely to answer all the questions, but it would be a nice start.

(An unrelated note, but I found interesting: is that people who were depressed tended to have noticeably higher A1c numbers.  I think this data came mostly from adults, but it makes sense to me, and was an important reminder that -- especially with type-1 diabetes -- a positive outlook really does help.  Also, while numbers are very important to managing type-1 diabetes, they are not the be-all and end-all of management feelings do matter.)

Thanks very much to CarbDM for organizing this event!

Joshua Levy
All the views expressed here are those of Joshua Levy, and nothing here is official JDRF, JDCA, or CarbDM news, views, policies or opinions. My blog contains a more complete non-conflict of interest statement. 
Clinical Trials Blog: http://cureresearch4type1diabetes.blogspot.com
Cured in Mice Blog: http://t1dcuredinmice.blogspot.com/