Sunday, 4 November 2012

Meteorological factors and autism?

If I had a pound (inflation adjusted y'know) for every time the words 'correlation does not equal causation' were used with autism and related conditions in mind, my piggy bank would probably be rather full by now. Indeed such is the application of this term when referring to the wide (very wide) variety of associations made between this, that and t'other, that a backlash has seemingly emerged to counter its usage as per this Slate article.

Cool, clear water @ Wikipedia 
Of course that's not to say that correlation does not imply causation. Recently for example, a study appeared  in the very well respected New England Journal of Medicine (NEJM) by Franz Messerli* which concluded: "Chocolate consumption enhances cognitive function, which is a sine qua non for winning the Nobel Prize, and it closely correlates with the number of Nobel laureates in each country".

Whilst still wondering whether this is some kind of spoof article (apparently it's not) this study encapsulates everything about the notion of correlation ≠ causation given the multitude of other factors which might possibly influence citizens of a particular country winning the distinguished Nobel Prize. I am willing to concede however, should more 'direct' prospective evidence emerge, that eating chocolate might be part and parcel of Nobel Prize winning.

With this same logic I was intrigued to read the recent paper by Sophie St Hilaire and colleagues** (open-access) which based on an ecological study of autism concluded that precipitation and drinking water source might be associated with autism. Drinking water you say? Y'mean similar to the fish and pharmaceutics research and the Brick chlorination by-products paper? Well, similar at least.

In brief:

  • Based on data derived from the study by Waldman and colleagues*** which concluded a possible link between annual precipitation and autism prevalence rates in the Western United States, the source of drinking water was analysed with respect to autism rates.
  • The percentage of surface drinking water was thrown into the statistical mix as was population density, farming land, SES, risk of neurological disorder from air pollution, suicide rates and meteorological data.
  • When analysing all this data, the authors concluded that precipitation was linked to autism rates but the source of drinking water was an important factor too.
  • In among the other correlations, rates of autism also correlated with the suicide rate and negatively correlated with the unemployment rate.

There are some other interesting factoids to take from the authors findings and conclusions. Surface sources of drinking water potentially implying some role for environmental contamination was a primary association. More than that however, the authors seemed to rule out any significant association with contaminated run-off water derived for example from pesticide use and indeed urban pollution and autism rates. They leave the door open to air pollution as a possible factor and indeed pharmaceutical contamination of water sources; bearing in mind they did not test the drug residues in water hypothesis.

The negative correlation between autism rates and unemployment (as in high prevalence of autism being linked to low unemployment rates) goes back to other research in this area. That being said, I'm reminded about more recent investigation which seem to imply an opposite trend (high prevalence of autism linked to lower SES) so once again we have to be careful not to generalise too much.

There's no doubt that this is an interesting study despite the reliance on population facts and figures and our old friend correlation. Accepting another old adage on 'lies, damn lies and statistics' (see here****), there are some useful components from the St Hilaire study ripe for further testing. Not least the continuing theme that environment may very much be a part of the increase in reported cases of autism.

To finish, bird is the word (or did you not know?).


* Messerli FH. Chocolate consumption, cognitive function, and Nobel Laureates. NEJM. October 2012.

** St Hilaire S. et al. An ecological study on childhood autism. Int J Health Geographics. 2012; 11: 44.

*** Waldman M. et al. Autism prevalence and precipitation rates in California, Oregon, and Washington counties. Arch Pediatr Adolesc Med. 2008; 162: 1026-1034.

**** Bartell SM. & Lewandowski TA. Administrative censoring in ecological analyses of autism and a Bayesian solution. J Environ Public Health. 2011; 2011: 202783.

---------- Sophie St Hilaire, Victor Ezike, Henrik Stryhn, & Michael A Thomas (2012). An ecological study on childhood autism International Journal of Health Geographics