New Studies: Nutritional and Pathologic Clues in Autism

I've devoted quite a few entires to autism (more properly called the autism spectrum disorders).  Perhaps because the illness can be so devastating, attacks so young, and is so mysterious in its origin.  I've migrated many of the autism entries over to Psychology Today, and the specifically nutritionally-related ones can be found here:

Diet and Autism
Diet and Autism - Newer Studies and Intriguing Links
Autism, Inflammation, Speculation, and Nutrition

The first new paper I want to present today is a Letter in Nature:  Transcriptomic analysis of autistic brain reveals convergent molecular pathology.  Like most Nature papers, the prose is dense with advanced molecular biology and neuroscience, but I think I can glean the meaning of it.  In this study, the researchers examined post-mortem brains from 19 autism cases and 17 controls.  They focused on three areas previously shown to be different in the autistic brain (specifically areas of the cerebral cortex and cerebellum), finding the most differences between autistic brains and controls in the cortex.

What is interesting about these sorts of studies is that they look at gene expression rather than just genetics, meaning they find out which genes are activated differently between cases and controls.  They found that many different genes were affected (consistent with the findings in autism in general - there is no single "autism gene" causing all cases, though 1% of cases have a 15q duplication, for example).  The interesting part is that the gene expression that was different between cases and controls clustered around specific areas, and were highly related to what is called "cortical patterning." These are some of the same gene areas that were already suspected to be causative for autism spectrum disorders,  and are also suspected to be causative in schizophrenia.  Other genes that were upregulated were functioning as part of the immune and inflammatory response.

What is intriguing about this study is that the researchers felt it showed convergence of the many different genetic causes in the area of transcription and splicing of certain genes as the underlying mechanism of the disorder.  They were also able to show that the inflammation was secondary rather than primary.

All right, we made it through the tough part.  Now on to the second paper, which is far more nutritionally related (possibly - it's just an observational study, but it is interesting).  Prenatal Vitamins, One-carbon Metabolism Gene Variants and Risk for Autism.

So, the intro - Prevalence in the US is about 1 in 110 children and may be rising (a recent Yale study in South Korea study suggests autism affects 1 in 38 children there).  It is widely accepted that genetic risk underlies the disorder, but there are also definitely interactions with environmental factors.  The recent study showing increased risk of autism in short intervals between pregnancies could implicate nutritional factors.  Valproic Acid (depakote) which disrupts folate metabolism is a known cause of autism and neural tube defects.

Other literature suggest that some children with autism have altered B vitamin metabolism and a reduced ability to methylate things.   So in this brand new study, the researchers specifically looked at vitamin supplement intake.  Children with autism in California were identified as part of large, population-based, case-control CHARGE study.  Their mothers were asked in telephone interviews whether they consumed prenatal vitamins, multivitamins, fortified cereals, and other supplements in the period 3 months before conception, and then during pregnancy and breastfeeding.  If they answered yes, they were asked specifically what vitamins and brands were consumed.

Interesting results ensued - 97% of mothers of the normal controls and 96% of mothers of the children with autism reported taking prenatal vitamins.  However, mothers who took the vitamins in the 3 months before conception and in the first month of pregnancy were significantly less likely to have autistic children than those who didn't.  There was no difference between vitamin and non-vitamin takers in months 2-9 of pregnancy or later.  The findings were controlled for various confounders, such as mom's education, race, whether conception was intentional or not, year of birth, etc.

Even more interesting - use of regular multivitamins in lieu of the prenatals was not associated with decreased risk of autism compared to taking no vitamins.  The major difference between most prenatal vitamins and regular multis is a doubling of folic acid.  The government says the synthetic folic acid is the same as the folate in food - Chris Kresser notes here that conversion from synthetic folic acid to the active tetrahydrofolate is poor in humans, and while synthetic folic acid pretty definitely reduces risk of neural tube defects, it also likely increases the risk of cancer - these kinds of population risks are a tough call.  If you are trying to get pregnant and you aren't always sure you are getting a lot of folate from foods, it may well be worth the risk to take synthetic folic acid for the periconception period.  Well, that's something to chew on.

The researchers in this study also checked genetics, finding that children with methylation issues (related to the folate cycle) were also more likely to have autism.  Maternal periconceptual folic acid seems to increase methylation in specific gene regions in the child (1), which can have epigenetic effects that can affect a child throughout life.

Get nutrients from food whenever possible, but there are certain circumstances in which wise supplementation may be very important.   And for public health recommendations (assuming the SAD) - supplementation for women trying to become pregnant seems wise, and when one reviews IOM and other reports, one often sees a fair accounting of the risks involved.  In fact, that is often why IOM or other recommendations seem quite conservative.   The questions continue!