Sunday, 15 May 2016

Folate, autism and IMFAR: keep calm

IMFAR - The International Meeting for Autism Research - has just closed it's doors, bringing to an end one of the premier annual autism research conferences where one might expect quite a few of the presentations to eventually become (peer-reviewed) fodder for this blog.

This year (2016) has seen yet another startling array of research papers covering all manner of topics pertinent to autism; it's presumed aetiologies (plural as per the 'autisms'), diagnosis and management, all discussed and dissected. This year also saw some well needed focus on other important issues linked to autism such as various (medical and psychiatric) comorbidity linked to the label, the idea that children with autism become autistic adults (mostly) and the rather distressing idea that suicide (ideation or actual) seems to be over-represented when a diagnosis is received. Without trying to portray too negative an image of the very wide autism spectrum, autism can in some cases be a life-limiting as well as a life-changing condition (see here and see here).

As an outsider looking in on proceedings (#IMFAR2016) I was particularly interested in a few discussions that were covered during the event. This included the keynote address delivered by Irva Hertz-Picciotto on how, despite various environmental agents being traditionally correlated with some 'types' of autism (rubella, valproic acid), there are still challenges in terms of putting environmental factors on a par with genetic issues for example. Indeed, one or two tweets about her presentation kinda summed up the stark lack of knowledge and expertise in this area; specifically how 'chemical' insults (being careful with that word) potentially linked to cases of autism or at least autistic traits, already pervade the peer-reviewed literature (see here for example). It seems we need to further organise how research is done in this area, where exposure and genetic fragility are taken into account alongside the idea of synergy when it comes to the chemical soup that we all live in [1]; similar to the idea that multiple genes might be involved in multiple autisms so multiple non-genetic exposures might also show a connection (might I also suggest a greater focus on subgroups on the 'autisms' spectrum too?)

One of the other 'environmental' factors that was raised at IMFAR 2016 was that of folate (folic acid) and the possibility of a connection between pregnancy levels of this stuff and some autism as per the paper by Ramkripa Raghavan and colleagues [2]. For those who might already be well-read in autism research, the idea that maternal levels of folate during pregnancy might have a bearing on risk of offspring autism is not a new one (see here). With sentiments not a million miles away from those proposed by the late David Barker and the 'foetal programming hypothesis', the collected research on folate availability/supplementation during pregnancy impacting on the developing child is still the topic of some discussion (see here and see here). What is clear is that much like the need for folate during pregnancy to reduce the risk of neural tube defects, there may also be additional developmental requirements for suitable levels when it comes to other outcomes too.

Rather interestingly however, the data from Raghavan et al was not all one way when it came to 'risk' of offspring autism and levels of folate and a related nutrient, vitamin B12. To quote from their un-peer-reviewed paper: "In this urban low-income minority birth cohort, we observed an elevated risk of ASD [autism spectrum disorder] associated with high maternal plasma folate levels (>59 nmol/L), which far exceeds the excess cutoff suggested by the WHO (>45.3 nmol/L). Excess maternal vitamin B12 (>600 pmol/L) was also shown to be associated with greater ASD risk in offspring.  The risk of ASD was highest if mothers had both excess in folate and B12 levels." As you might imagine, the accompanying press release that followed this un-peer-reviewed paper was snapped up by various media outlets with titles like: "Taking too many vitamins during pregnancy 'can treble children's risk of being autistic'". Lo and behold, we have yet another 'scare story'...

Accepting that this was un-peer-reviewed research and that inflated press releases seem to abound in the domain of science communication (see here) I was a little less ruffled by the data reported by Raghavan and colleagues. I can see why such findings might make great headlines - "The risk was greatest for those children whose mothers had both high plasma folate (>59 nmol/L) and vitamin B12 (>600 pmol/L) (HR [adjusted hazard ratio]: 17.59; p value: <0.001)" - but this is not the first time that such sentiments have been expressed with relation to autism. Indeed, on a previous post asking whether some of the data of pregnancy folate levels/supplementation and offspring autism risk might not be just all about deficiency (see here) I discussed some rather speculative ideas (albeit peer-reviewed ideas) about how there may be a balance to be struck between potentially too little and too much of a good thing.

'Scientists urge caution over 'alarmist' claim of link between pregnancy folate and autism' was a rather more restrained headline in relation to the Raghavan report. I was much happier with this headline and coverage that put into perspective the preliminary nature of the report (yes, un-peer-reviewed) and how: “There are many epidemiologically based associations made of this sort – increasingly so in autism at the moment." Indeed there are, but unfortunately the commentator goes on to say that "Without details of the analysis, or any theory of action this looks like low-grade evidence."

Actually there are 'theories of action' and they've been discussed quite a bit in the peer-reviewed literature in this area. Not least, the idea that folate and vitamin B12 are important compounds in something called the folate cycle which intersects with another set of important metabolic process: the methylation cycle (see here) and all that DNA methylation stuff. There are a number of possible 'issues' that might be autism-relevant in these biological cycles, not least related to something called MTHFR (see here) and some emerging data on folate receptor autoantibodies (see here). Issues with these systems could very much impact on how folate is used and whether high plasma folate for example, might not be just as the result of too much supplementation. As per what we know from data from more formal medicines, drug metabolism can be quite an individual thing.

I do also want to bring in a little more data about vitamin B12 and 'some' autism as potentially being relevant. Accepting that there has been very little data on maternal vitamin B12 levels and offspring autism, there is certainly quite a bit of data out there about 'issues' with vitamin B12 being tied into specific cases of autism. This year (2016) we've seen the results of placebo-controlled study on the use of methyl B12 for aspects of autism (see here) as well as a suggestion that decreased brain levels of vitamin B12 might link cases of autism and schizophrenia (see here). That severe vitamin B12 deficiency has been linked to cases of Heller's syndrome is also potentially important (see here) given the focus on 'regression'...

Cumulatively what the Raghavan and other data point to is a potentially complicated relationship between mother's nutrition during pregnancy and offspring outcomes. I don't say this to somehow hark back to the darker days of autism theory in terms of 'blame' but rather, alongside other lines of evidence, to point out that nutrition during the nine months that made us (and perhaps earlier) plays an important role in making us who we are. The way that said nutrition is metabolised is also likely to vary from person to person. Minus sweeping generalisations and inflated media headlines, a greater research focus on how that nutrition might impact on at least some autism is very much implied again keeping in mind all that individuality in terms of how the body 'processes' nutrition. Indeed, for those mums identified in the Raghavan data as showing high levels of folate and vitamin B12, in the spirit of scientific endeavour, I'd be asking 'why?' and what does it mean for other related markers such as 'the big H' (homocysteine) for example?

For now however, keep calm and carry on with Love, love, peace, peace (song starts after 1 minute).


[1] Boggess A. et al. Mean serum-level of common organic pollutants is predictive of behavioral severity in children with autism spectrum disorders. Sci Rep. 2016 May 13;6:26185.

[2] Raghavan R. et al. Maternal Plasma Folate, Vitamin B12 Levels and Multivitamin Supplement during Pregnancy and Risk of Autism Spectrum Disorders in the Boston Birth Cohort. IMFAR 2016; 22533. [NOT PUBLISHED IN A PEER-REVIEWED JOURNAL]