Thursday, 25 September 2014

Temporal trends in US autism prevalence: mainly real increase

"Diagnosed autism prevalence has risen dramatically in the U.S over the last several decades 
and continued to trend upward as of birth year 2005. The increase is mainly real and has occurred mostly since the late 1980s".
"They call me Cha Cha because I'm the
best dancer at St. Bernadette's"

That was the conclusion reached in the paper by Cynthia Nevison [1] (open-access) following her analysis of temporal trends in autism diagnosis for birth years between 1970 and 2005. Based on an analysis of datasets derived from IDEA (Individuals with Disabilities Education Act) and the CDDS (California Department of Developmental Services), the author suggested that between 75-80% "of the tracked increase in autism since 1988 is due to an actual increase in the disorder rather than to changing diagnostic criteria". The question of what environmental factors might have been driving such an increase in cases is also discussed in the Nevison paper, with the author concluding "children’s exposure to most of the top ten toxic compounds has remained flat or decreased over this same time frame". That top 10 list by the way, seems to come at least partly from the Landrigan paper talked about a couple of years back (see here).

There is quite a bit of information included in the Nevison paper which I'm reluctant to write a mega-blog entry on at this time. This includes various caveats about the use of IDEA and CDDS databases and their constraints on for example, what ASDs (autism spectrum disorders) are included in the datasets. I will however summarise some of the main findings in relation to the environmental factors probed by the author bearing in mind that autism research does not appear to be her main area of scientific interest (with all due respect).


  • Air pollution... something which has cropped up with ever-increasing frequency in the peer-reviewed research on autism (see here and see here for example). Nevison concludes: "there is no obvious evidence to suggest that trends in estimated vehicular emissions or directly measured air pollution are consistent with the sharp temporal increase in U.S. autism". So no smoking tailpipe (exhaust) there then as per other recent research findings [2].
  • Mercury in vaccines... a topic guaranteed to furrow brows and raise blood pressure in some quarters (see here and see here for example). Nevison discusses the phaseout of thimerosal (thiomersal) from paediatric vaccines used in the US concluding that "the expansion of thimerosal exposure in the late 1980s and early 1990s coincides closely with the rise in autism around that time". But... "the temporal trends in autism and thimerosal following the childhood vaccine thimerosal phaseout are incompatible". Other sources of mercury exposure get a similarly 'unlikely' mark from Nevison.
  • Organophosphate (OP) pesticide exposure... an interesting area which again has been covered previously on this blog (see here and see here). Nevison points out the declining use of such pesticides in the US following "the adoption of crops genetically modified to produce Bt toxin, which repels targeted insect pests, thus reducing the need for external insecticides". Recognising that pesticides are not to be trifled with (see here) I've always been a little confused about the mechanism(s) through which OP exposure could theoretically impact on the presentation of autism. I know people have talked about PON1 and autism [3] (some autism) but I do feel as though the primary effect of OPs - acting on acetlycholinesterase -  is something in need of a lot more research with autism in mind before anyone gets too carried away. 
  • There is however a caveat to the pesticides-autism conclusion by the author following some mention of "the rapidly increasing application of glyphosate, the active ingredient in the herbicide Roundup®". She continues: "it appears that glyphosate cannot be responsible for the first autism cases in the 1930s and is unlikely to have caused the late 1980s uptick, but could be interacting in recent years with other toxins to drive up the prevalence of U.S. autism". Depending on where you look, you'll get various different messages about the pros and cons of glyphosate. I remember reading a report a few years back (see here) authored by one of the researchers involved in that 'organic food might be better for you' paper recently (see here) which painted a rather disturbing picture of the product. For balance, I'm going to also refer you to the various documents provided by Monsanto (the producer of Roundup) for their response to safety concerns. When it comes to a search of PubMed with the terms 'autism and glyphosate' the current result is zero although it has been mentioned elsewhere in the peer-reviewed domain [4]. 
  • Maternal obesity... I'm being quite careful here accepting the previous discussions in this area of autism research (see here). Nevison reports that: "the time trend in obesity among U.S. women correlates well to that of autism, suggesting maternal obesity may be a direct influence or a comorbid consequence of the dietary factors contributing to autism, or both". 

One needs to bear in mind that the United States was the focus on these findings and as such the conclusions may not pertain to other parts of the globe. I probably don't need to say it but one should also bear in mind the saying 'correlation is not the same as causation' too. There are also other issues which have been discounted by Nevison as being related to the autism prevalence. Lead (Pb) is one of them; something which I personally would not be so keen to disregard given the more recent evidence on even trace amounts of this stuff not being great for the developing child (see here). This on top of what has been talked about with autism in mind (see here). 

I have a few other points to make which were perhaps not readily implied in the Nevison paper: first is the assumption that autism is some kind of universal condition. What I think many people have come to realise over the past few years is that within the significant behavioural and developmental heterogeneity grouped under the label autism, also compounded by the increased frequency of various comorbid conditions, a more plural description - the autisms - might be more pertinent. What this implies is that different weightings for concepts like genetics or environment (or epigenetics!) are likely acting across risk of different types of autism. 

Second, and related to point one, is the concept of synergy across different 'causative' factors. In the same way that autism science has started to accept that there is no one gene linked to all cases of autism, so environmental examinations can't really expect there to be one environmental factor working alone pertinent to all autism. If one is to assume that various environmental factors (see here and see here for example) may be linked to autism onset, it is more likely that combinations of factors are playing a role or possibly specific environmental factors acting on specific types of autism. Take for example the early work on air pollution and genotype being potentially associated with autism risk [5] as one example.

At the risk of being too speculative, I'm also minded to bring to your attention an emerging idea which might also be relevant: transgenerational epigenetics. The idea is that what happened in previous generations might have an influence on subsequent generations without structurally altering DNA in terms of the physical language of the genome. There has been some media interest in this concept in recent times (see here). With autism in mind, we've seen hints of this idea in the peer-reviewed research literature as per discussions on how advancing grandparental age might be linked to an increased risk of autism (see here). Jill Escher also talked about past chemical exposures potentially impacting on "vulnerable fetal germline epigenetics" (see here for her presentation). The implication being that correlating current exposure patterns with current autism diagnosis prevalence might not necessarily be the best way of looking at whether (and which) environmental factors might have a bearing on autism prevalence...

The primary message from the Nevison report is that the numbers of cases of autism do appear to be on the rise in the United States, and as per other reports from geographical neighbours (see here) the debates are not yet over about the reasons for that increase. Something another study has also talked about recently [6]...

So then, Only Love Can Break Your Heart by Saint Etienne.


[1] Nevison CD. A comparison of temporal trends in United States autism prevalence to trends in suspected environmental factors. Environmental Health 2014, 13:73

[2] Gong T. et al. Exposure to Air Pollution From Traffic and Neurodevelopmental Disorders in Swedish Twins. Twin Res Hum Genet. 2014 Sep 17:1-10.

[3] D'Amelio M. et al. Paraoxonase gene variants are associated with autism in North America, but not in Italy: possible regional specificity in gene-environment interactions. Mol Psychiatry. 2005 Nov;10(11):1006-16.

[4] Seneff S. et al. Is Encephalopathy a Mechanism to Renew Sulfate in Autism? Entropy 2013. 15; 372-406.

[5] Volk HE. et al. Autism spectrum disorder: interaction of air pollution with the MET receptor tyrosine kinase gene. Epidemiology. 2014 Jan;25(1):44-7.

[6] Dave DM. & Fernandez JM. Rising autism prevalence: real or displacing other mental disorders? Evidence from demand for auxiliary healthcare workers in California. Economic Inquiry. 2014. 25 August.

---------- Nevison, C. (2014). A comparison of temporal trends in United States autism prevalence to trends in suspected environmental factors Environmental Health, 13 (1) DOI: 10.1186/1476-069X-13-73