The continuing legacy of DDT? Maternal insecticide levels and risk of offspring autism

I see that the paper published by Alan Brown and colleagues [1] reporting on "the first biomarker-based evidence that maternal exposure to insecticides is associated with autism among offspring" has been garnering headlines and ruffling feathers in equal measure.

It's seemingly guaranteed that whenever a study mentions words like 'toxins' and 'environment' in the context of (some) autism, some brows become furrowed, decrying any possibility of a *connection*. In equal measure, it has to be said, opposite views on such evidence being proof positive that 'autism is all environmental' and 'man-made' also emerge. To quote from an imaginary character: "Darkness rises and light to meet it!" (I'll leave you to decide which side is which).

DDT - dichlorodiphenyltrichloroethane - mentioned in the title of this post, is an insecticide categorised as a persistent organic pollutant (POP). Like many such products, it started off being used with noble intentions, to control important diseases like malaria and typhus, and even earned someone a Nobel prize as a result. It's also credited with saving millions of lives during the period of its use; as one WL Churchill apparently said: "The excellent DDT powder has been found to yield astonishing results against insects of all kinds, from lice to mosquitoes."

But things did not remain so rosy for DDT as data (see here) and a book subsequently emerged suggesting a link between exposure to DDT and various adverse animal and human outcomes. Restrictions on use and banning eventually came into effect and the glory days of DDT started to wane, albeit not with everyone being totally supportive of an outright ban. As per the POP status of DDT however, restrictions and banning did not mean that this stuff magically disappeared from the environment overnight. Coupled to a less than universal ban on the use of DDT, and residues of DDT can still be found in the environment today, albeit in lower levels than when it was at peak production and usage. And science is able to test for them fairly accurately...

Set in this context, the Brown paper did just that; the DDT 'connection' stems from researchers noting that levels of a specific metabolite of DDT - p,p'-dichlorodiphenyl dichloroethylene (p,p'-DDE) - measured in "maternal serum specimens from early pregnancy" seemed to show a 'threshold' *association* with offspring developmental outcomes such as autism (with and without intellectual disability). This, in the context that there is also some evidence of 'transplacental transfer' of such compounds [2], and where other studies have talked about the possible effects of in-utero exposure to DDT and offspring neurodevelopment [3] as part of a wider appreciation that infants and children may be uniquely sensitive to such pesticide exposures (see here). Brown et al report such results as part of the Finnish Prenatal Study of Autism (FIPS-A), an initiative that has already yielded quite a few important peer-reviewed findings (see here), some of which have been previously covered on this blog (see here for example). Study organisation and participant numbers are pretty good in the FIPS-A initiative, albeit with some methodological caveats (i.e. the use of observational study not able to prove cause-and-effect).

The finding of a 'dose-related' (threshold-related) relationship between p,p'-DDE and offspring autism risk noted by Brown is also interesting. Add in the observation that autism accompanied by intellectual (learning) disability also seemed to show an apparent relationship to p,p'-DDE levels and there is a scheme of additional work to be followed. Not least with regards to mechanism(s) - preterm birth risk has been mentioned which could 'overlap' with some autism-related data [4] - and whether for example, other risk 'compounds' might also be important. Indeed, I understand that the possibility of 'synergistic' interactions might be on the cards in future study for this group in light of previous results [5]. And going back to those findings from Eskenazi et al [3] who noted that "breastfeeding was found to be beneficial even among women with high levels of [DDT] exposure", I do wonder if this variable might also require further examination in the context of the Brown findings too [6].

Obviously one has to be slightly careful with any sweeping generalisations made as a result of the Brown findings. This research did not say that all autism was 'caused' by DDT exposure or any other insecticide or related compound, and neither was it designed to do so. The authors also mentioning that their study "has implications for the prevention of autism" is perhaps more than a little premature and seemingly at odds with some (recent) opinions.

But the Brown study did make a point of testing for DDT metabolites in mums-to-be and, set within other independent research findings *highlighting* various other pollutant exposures as potentially being relevant to the developing brain and important offspring outcomes like autism (see here and see here and see here for examples), one cannot simply brush off the Brown results and their important implications. Further research is implied minus hype.

Coincidentally, a 'roadmap' for incorporating the 'exposome' into psychiatry research was recently published [7] and could prove quite useful in the context of further research in this area...

----------

[1] Brown AS. et al. Association of Maternal Insecticide Levels With Autism in Offspring From a National Birth Cohort. Am J Psychiatry. 2018 Aug 16:appiajp201817101129.

[2] Zhang X. et al. Transplacental transfer characteristics of organochlorine pesticides in paired maternal and cord sera, and placentas and possible influencing factors. Environ Pollut. 2018 Feb;233:446-454.

[3] Eskenazi B. et al. In utero exposure to dichlorodiphenyltrichloroethane (DDT) and dichlorodiphenyldichloroethylene (DDE) and neurodevelopment among young Mexican American children. Pediatrics. 2006 Jul;118(1):233-41.

[4] Agrawal S. et al. Prevalence of Autism Spectrum Disorder in Preterm Infants: A Meta-analysis. Pediatrics. 2018 Aug 3. pii: e20180134.

[5] Cheslack-Postava K. et al. Maternal serum persistent organic pollutants in the Finnish Prenatal Study of Autism: A pilot study. Neurotoxicol Teratol. 2013 Jul-Aug;38:1-5.

[6] Tseng PT. et al. Maternal breastfeeding and autism spectrum disorder in children: A systematic review and meta-analysis. Nutr Neurosci. 2017 Oct 18:1-9.

[7] Guloksuz S. et al. The Exposome Paradigm and the Complexities of Environmental Research in Psychiatry. JAMA Psychiatry. 2018 Jun 6.

----------

[Objective] exposure to flame retardants and social behaviours

Although a few details of the study reported by Shannon Lipscomb and colleagues [1] (open-access) interested me, I was particularly taken by their use of "a silicone passive wristband sampler [worn] around his/her wrist or ankle" to "assess the child’s exposure to flame retardants" as part of their investigation "to determine if flame retardant exposure was associated with measurable differences in social behaviors among children ages 3–5 years."

I've covered the topic of potential adverse effects associated with exposure to flame retardants such as brominated diphenyl ethers (BDE) before on this blog (see here and see here for examples). Such compounds are listed as POPs (persistent organic pollutants) because of their ability to endure in the environment, accumulate in the body and potentially [adversely] affect various biological systems. In other words, these are compounds that might well have served an important purpose at one time - flame retardants - but are now realised to have quite a risk profile attached to them. Sounds familiar doesn't it?

Anyhow, Lipscomb et al relied on other research [2] suggesting that various compounds/chemicals can be sequestered from silicone wristbands - those plastic things that many people wear for various causes - with the right equipment and under the right circumstances. To any analytical chemist, this is probably scientific music to their ears. They "extracted and analyzed for 41 different flame retardant compounds using gas chromatography mass spectrophotometry" and focused on 11 compounds "PBDE-47, PBDE-99, PBDE-153, PBDE-154, PBDE-49, PBDE28 + 33, tris(1,3-dichloro-2-propyl) phosphate], TPP [e.g. triphenylphosphate], TCPP [e.g tris(1-chloro-2-propyl) phosphate], and TCEP [e.g. tris(2-chloroethyl) phosphate" that were quite readily present in 60% or more of wristbands. For some of the compounds the authors generated a 'sum of' score; for example: "ƩPBDEs is the total amount of PBDE-47, PBDE-99, PBDE-153, PBDE-154, PBDE-49, and PBDE28." Social behaviours by the way, were scored by teachers in the preschool setting of participants using the Social Skills Improvement System - Rating Scales.

Results: 92 children were initially recruited onto the study but only 77 children returned their wristbands intact (i.e. some of them 'went through the laundry'). Further: "a final sample size of 69 children with complete data... were included in the final analyses." Then: "Bivariate analysis revealed modest correlations between flame retardant exposure and some of the social behavior subscales." What this suggests is that there may be some evidence that such compounds (including organophosphate-based flame retardants (OPFRs)) might impact on aspects of social skills development but there are constraints based on the sample size used and the reliance on one primary measure of social skills for examples.

As per the previous sentence, I'm not totally convinced by this data but am still really interested in the use of wristband samplers described by Lipscomb and colleagues. I can see how this kind of objective measure of exposure could really add another dimension to lots of different areas of research on environmental exposures in relation to various labels. Take for example the quite complicated area of research talking about maternal air pollution exposure and offspring autism risk (see here). Instead of just relying on postcode (zip code) in relation to mapping (estimating) pollution exposure, one could potentially adapt the chemical assay to screen for particulate matter for example, as collected on those wristbands. Certainly an easier way than lugging around a portable air monitor I would have thought. No doubt there are also other uses for such simple solutions...

Music: Europe - The Final Countdown. 80s rock hairstyles at their best and perhaps an apt song given what happened here in Blighty yesterday...

----------

[1] Lipscomb ST. et al. Cross-sectional study of social behaviors in preschool children and exposure to flame retardants. Environmental Health 2017; 16: 23.

[2] O'Connell SG. et al. Silicone Wristbands as Personal Passive Samplers. Environ. Sci. Technol. 2014; 48: 3327–3335.

----------

Lipscomb ST, McClelland MM, MacDonald M, Cardenas A, Anderson KA, & Kile ML (2017). Cross-sectional study of social behaviors in preschool children and exposure to flame retardants. Environmental health : a global access science source, 16 (1) PMID: 28274271

On (banned) organochlorine compounds and autism risk

'Chemicals banned decades ago linked to increased autism risk today' went the press release attached to the findings reported by Kristen Lyall and colleagues [1] (open-access).

Observing that "higher levels of some organochlorine compounds during pregnancy are associated with ASD [autism spectrum disorder] and ID [intellectual disability]" the Lyall results once again push environmental factors back into the research spotlight. Indeed, environmental factors that were banned decades ago.

Including a cohort of children diagnosed with autism (n=545), those diagnosed with ID (also known at learning disability) (n-=181) and general population (asymptomatic?) controls (n=418) researchers accessed archived biological samples taken from mothers during the second trimester of pregnancy. Using some pretty sophisticated chemical analysis methods - "gas chromatography isotope dilution high resolution mass spectrometry (GCIDHRMS)" - various chemical compounds considered as POPs (persistent organic pollutants) were assayed for. Most if not all of these compounds were banned in the 1970s because of their potential effects on health. Because of their chemical nature however (i.e. enjoying bathing in fats) they can and do still persist in the environment, particularly in the food chain.

Results: various PCBs (polychlorinated biphenyl ethers) and "persistent pesticides" were included in the chemical analysis of maternal samples. Some, but not all, were reported in samples across the different groups. After some statistical wizardry in terms of adjusting samples and the results for various factors ("children with ASD were approximately four times as likely to be male than female... have older parents, and mothers with higher education") authors concluded that a few of the metabolites looked at might be linked to autism risk; specifically: "that exposure to PCB congeners in utero may influence risk of ASD in offspring."

"Primary analyses highlighted PCB 138/158 and PCB 153 in association with ASD, though other correlated congeners also demonstrated associations above the null." PCB 138/158 also seemed to show some sort of connection to the risk of offspring ID too "suggesting the impact of exposure to this congener on neurodevelopment broadly." Conversely, none of the other organochlorine compounds seemed to show any (significant) connection to autism offspring risk. Something similar has been talked about before with this broad collection of compounds in mind under more direct analysis conditions (see here). The authors conclude that further research is required to both substantiate their findings and also ascertain some of the hows and whys of these compounds in relation to autism and ID. Importantly, they acknowledge that their list of compounds tested may not be the whole story in terms of the 'multiple chemicals' people are exposed to over a lifetime.

These are rather interesting results. Not least because the potential legacy of these compounds continues years and years after production of them had all but ceased following health concerns. That researchers also focused on maternal pregnancy blood samples again (see here) puts gestational 'exposure' front and centre when it comes to potential effects and mechanisms too. The idea that immune function could be a target effect of such compounds when it comes to offspring autism risk is also explored by the authors: "effects on the immune system is another particularly likely mechanism." This would also seem to tally with the growing evidence that maternal immune function during the nine months that made us might be an important part of aetiology for at least some autism and/or more general neurodevelopmental issues.

The story continues as it might with other compounds too [2] on this 'TENDR' area of research...

To close, Worf don't like the lute...

----------

[1] Lyall K. et al. Polychlorinated Biphenyl and Organochlorine Pesticide Concentrations in Maternal Mid-Pregnancy Serum Samples: Association with Autism Spectrum Disorder and Intellectual Disability. Environ Health Perspect. 2016. Aug 23.

[2] Jeddi MZ. et al. The role of phthalate esters in autism development: A systematic review. Environ Res. 2016 Aug 24;151:493-504.

----------

Lyall K, Croen LA, Sjödin A, Yoshida CK, Zerbo O, Kharrazi M, & Windham GC (2016). Polychlorinated Biphenyl and Organochlorine Pesticide Concentrations in Maternal Mid-Pregnancy Serum Samples: Association with Autism Spectrum Disorder and Intellectual Disability. Environmental health perspectives PMID: 27548254

PCB congeners and 15q11-q13 duplication autism

Contains PCBs @ Wikipedia

The science news is currently awash with talk about results from the Encyclopedia of DNA Elements (ENCODE) project (see here for a good description). The net results have been summarised by quite a few media outlets including the BBC, including the suggestion that junk DNA might not be as junk as first thought (see this post from MJ over at Autism Jabberwocky).

I was interested to see that the words DNA methylation cropped up quite a few times during the course of the various discussions, and how the switching on and off of genes might actually be pretty important to our lives. It all got me thinking back to everyone's favourite gene-environment science, epigenetics (see here).

With all this in mind, a recent paper by Mitchell and colleagues* provides some food for thought. Including a significant authorship contingent from that autism research favourite, the MIND Institute, the paper abstract makes a striking assertion: "These results demonstrate a novel paradigm by which specific POPs [persistent organic pollutants] may predispose to genetic copy number variation of 15q11-q13".

I admit that I did a bit of a double-take when I read this sentence given that it seems to suggest some involvement of POPs in what is a genetic condition.

OK let's just back it up a little and take this one step at a time. 15q11-q13 duplication autism spectrum disorder refers to an issue with chromosome 15 and in particular with something called the PWS/AS critical region; that is a genetic area thought to be linked to the presentation of Prader-Willi syndrome (PWS) and Angelman syndrome (AS) depending on whether inheritance is maternal or paternal. Autism has also been tied into the 15q11-q13 region as per studies like this one from Cook and colleagues**. There is quite a bit of other research on 15q11-q13 and autism in the scientific literature which can be browsed here.

POPs are basically those chemicals/compounds which stick around in the environment and seem not to be easily environmentally degradable. POPs include a range of compounds (see here) some (many?) of which also have the ability to bioaccumulate and potentially cause some quite serious effects on health as per what happened with Yushō disease. I don't want to get too heavily into the nitty-gritty of the health impacts of POPs simply because it is a very complicated area of environmental health sciences.

Certain POPs have been mentioned previously with autism / autistic behaviours in mind as per these archive posts on polybrominated diphenyl ethers (PBDEs) and autism and the flame-proofed mice with a Rett syndrome mutation study. This last study incidentally is also from Prof Janine LaSalle who also heads the current paper.

Mitchell et al report a few things in their study. I'm sorry that I can't post a link to the full-text paper, so you'll have to take my word for it:

• The lipid-rich brain tissue of one hundred and seven human postmortem brain samples were analysed for the presence of 8 polychlorinated biphenyl (PCB) and 7 PBDE congeners via GC-MS among other things. Samples were derived from deceased patients who were diagnosed with various neurodevelopmental disorders of 'known' origin (PWS, AS, Rett syndrome, 15q11-q13 duplication syndromes; n=32), idiopathic autism "of unknown etiology" (n=32) and asymptomatic controls (n=34). Age ranges were wide; between 4 - 61 years old at death across the groups.
• "Unexpectedly, PCB 95 was significantly (p<0.001) higher in the genetic neurodevelopmental group, but not idiopathic autism, as compared to neurotypical controls". Indeed out of 8 PCB congeners analysed, the genetic neurodevelopmental group showed the greatest mean concentration in  5 (although not significantly different levels).
• Furthermore, levels of PCB 95 seemed to be specifically tied to those diagnosed with a maternal15q11-q13 duplication (Dup15q) or deletion in Prader-Willi syndrome.
• Further analysis based on birth date pre- and post- 1976 (used because of the introduction of the 1976 Toxic Substances Control Act which is looking to be replaced by the Safe Chemicals Act of 2011) suggested that levels of PCB 95 were highest in the genetic neurodevelopmental group both pre- and post-1976. 
• A possible connection between PCB 95 and DNA methylation levels in Dup15q samples was undertaken by pyrosequencing for repetitive LINE-1 methylation levels ('similar' to this method). As would probably be expected, significant DNA hypomethylation was recorded compared with control samples (approximately 2% decrease in average methylation). That being said, year of birth was a confounder in that those controls born in the 1980s-1990s tended to show lower methylating functioning than those born in the 1960s-1970s. 
• PBDE levels showed little significant differences across the groups aside that is from one congener, PBDE 153, which showed a greater mean concentration in the control group compared to the other symptomatic groups (p<0.05).

I'm intrigued. To quote again from this study: "Our results demonstrated that 3/6 Prader-Willi syndrome and 5/6 Dup15q brain samples showed detectable levels of PCB 95 suggesting that this exposure should be investigated as a potential environmental contributor of the differing copy number variation rates in different regions".

Now I'm not saying that POPs 'cause' 15q11-q13 duplication autism spectrum disorder; the evidence just isn't there to suggest something so direct. As with every study which includes a postmortem element to it, there is always the sensitive issues of why the person died and whether comorbidity (epilepsy, learning disability?) might play an interfering role in any results obtained. It certainly is however an area that needs a lot more investigation.

I've often talked about mutations like SNPs and CNVs in relation to autism and a few other conditions (see here for example); the latest being that stark quote "no single SNP shows significant association with ASD or selected phenotypes at a genome-wide level" from the paper by Richard Anney and colleagues (see post here). Where SNPs and/or CNVs do occur however, particularly when described as de novo as in the recent paper on paternal age and de novo mutations (see here), the question has to be why do such mutations occur outside of just a generic 'randomly based on age' argument? Without tempting too much criticism, I have to admit that 'random' just doesn't wash with me; there has to be a reason. One candidate outside of just maturation has to be environment.

A few final points to make and then I'm done. This is not the first time that PCB 95 has cropped up and interestingly has been previously tied into autism and calcium signalling as per this study by Wayman and colleagues*** (full-text). The issue of DNA hypomethylation and autism has also been covered on this blog previously so I'm not going to get too involved in that at this time.

And relax.

Despite the small participant numbers included in the study, I will say once again that I am intrigued by the results presented by Mitchell and colleagues and very much look forward to hearing more from this research group and others on our seemingly very delicate relationship with the modern environment around us.

---------

* Mitchell MM. et al. Levels of select PCB and PBDE congeners in human postmortem brain reveal possible environmental involvement in 15q11-q13 duplication autism spectrum disorder. Environmental & Molecular Mutagenesis. August 2012.

** Cook EH Jr. et al. Autism or atypical autism in maternally but not paternally derived proximal 15q duplication. American Journal of Human Genetics. 1997; 60: 928-934.

*** Wayman GA. et al. PCB-95 Modulates the Calcium-Dependent Signaling Pathway Responsible for Activity-Dependent Dendritic Growth. Environmental Health Perspectives. 2012; 120: 1003-1009.

----------

Mitchell MM, Woods R, Chi LH, Schmidt RJ, Pessah IN, Kostyniak PJ, & Lasalle JM (2012). Levels of select PCB and PBDE congeners in human postmortem brain reveal possible environmental involvement in 15q11-q13 duplication autism spectrum disorder. Environmental and molecular mutagenesis PMID: 22930557