(The following is a copy-and-paste of an email response to a brief discussion of the article -- "Cognitive impairment seen in preschool children with epilepsy." June 2nd, 2011. http://medicalxpress.com/news/2011-06-cognitive-impairment-preschool-children-epilepsy.html )
For autisms many have noted that co-occurring epilepsy tends to go with the cases of microcephaly (or tending toward microcephaly) and with the sub-type sometimes referred to as kanner-autism involving moderate to severe impairment of general cognitive functioning (low to very low IQ). (Both macro and microcephaly are common in autisms, leading I think to a lot of blurry diagnostic problems in research and clinical practices.) (Definitions - macro and micro are usually defined as being in the top 2 or bottom 2% of the population for head size adjusted for age and sex.)
Here is a very rough outline of how I am currently thinking about some of these issues.
Humans as a species are a statistical outlier with respect to brain growth, both among primates but even among mammals in general. First there is the birth canal problem, then later the large amount of growth, as well as the neural pruning that normally must follow that growth.
In neanderthals the birth canal problem appears to have been solved somewhat differently (with a twist of the fetus during birth) than with anatomically modern humans (completely backward, 180, compared to other species, to allow the head easier passage). So it would seem the whole "large-head" thing is still a major evolutionary (adaptive solution) situation as near as what appears to be our closest evolutionary relative.
Humans have a greatly delayed infancy. It is often said of humans that we are like a species where it is normal to have a premature birth, as many of our developmental milestones don't hit "normal" for mammals until around age two. So it doesn't seem too unreasonable to hypothesize that many of the developmental transitions that humans are going through around age two are the developmental transitions that other mammals go through around age zero, or zero to six months.
Around age two to four in humans there's a whole lot of neural pruning going on. It doesn't seem too unreasonable to me at this stage in my understanding of the subject that many of the cases of macrocephaly in autisms (20-25% by 2007 estimates) is due to a developmental trajectory that involves less or much less neural pruning. (This can be conceptualized either as a "failure", or as possibly an adaptive solution of some kind to conditions of an ancestral environment much higher in diets of microminerals, omega-3s, fantastically more exercise and sunshine than modern populations, etcetera. a.k.a. the "mismatch" category of hypotheses in evolutionary medicine literature.) (Note that the 2007 numbers, 20-25, go up quite a bit if it's defined slightly more broadly such as say the top 80% of head sizes.)
And then the human brain still undergoes a lot, a lot of growth up until around five.
But I digress -- the article I'm responding to is about measures of epilepsy and IQ in young children. Specifically, why the microcephaly?
The way I am currently thinking about this is that there are developmental switches (sometimes genetically associated) that tell the brain to keep growing (..cannibinoids?), stop growing (..dopamine, opiode mu-receptors?), start pruning, etcetera. So for autism microcephaly my current thinking is there is a developmental switch that goes either "hold-off" on growing or just plain "stop", while at the same time there are not adjustments between the ratio of excitatory (principally here glutamate, though I would add dopamine in this context though it's possible the levels are moving in tandem with the glutamate levels) and inhibitory neurotransmitters (principally in the context of epilepsies - GABA). (Here I'm tapping into the standard conceptualization of epilepsies are being largely due to excess neural excitability.)
In some, perhaps nearly all, population-level analyses of head size and demographics these hypotheses may be difficult to measure as there very well may have been selection pressures for developmental switches to be relatively well-balanced through the populations. (Here I have in mind the story of Arthur Jensen about observing the 80 and below IQ kids playing on the playground -- the white kids seemed severely socially impaired, but the others not nearly as much so.)
Developmental disruption prior to, roughly say, age 18 months may be much more strongly associated with microcephaly and correspondingly after 18 months, the macro.
But there is much "general sense" in the preceding paragraphs and potential for digressions, and also I reserve the option to change my mind about much of it. And there is a balance between under-explanation leading to misinterpretation, and over-explanation leading to other kinds of confusion. So I'll end it here. One - it's increasingly bothersome to me that micro and macro, which tend to associate with quite different developmental phenotype expressions, are lumped together into the fuzzy category now called autisms. (The "granularity" of the diagnoses is much too large due to many reasons but including I think institutional and educational lags.) Two - that there is significant IQ impairment in many cases of epilepsy is no surprise -- none at all. (Though it's usually nice to have more and more confirmation of a direction of emerging thinking on a subject.)