I get a daily compendium of science-related headlines – yesterday one in particular caught my eye. It said:
Autism linked with rainfall in study: Children who live in the US Northwest’s wettest counties are more likely to have autism, but it is unclear why.
Now, I must confess to a potential bias here – I’ve been following with interest the vaccines-cause-autism kerfuffle in the US (since I half expect it to raise its ugly head here, eventually), & particularly admire the voice of rational thinking provided by Orac, among others. For the record, to date scientists have found no link between vaccines & autism (& this holds whether they consider vaccines in general or the thimerosal that was until the early 2000s used as a preservative). So… rainfall????
As it turns out, Orac is onto this one already. He notes that, taken at face value, the study does seem to show a correlation between rainfall & the incidence of autism – but then reminds his readers that correlation does not equal causation. And in any case, the link with rainfall is awfully tenuous. Its authors note that in a US Government study of 14 states, frequency of autism was higher in the northern states & least in the southernmost state studied (Alabama). They then made the rather startling leap to an hypothesis: that there is an environmental trigger for autism among genetically vulnerable children that is correlated with bad weather.
But why? Why would there be such a correlation? How would rainfall have this impact? What evidence did the authors have for their assumption about genetic vulnerability? (And is weather really, routinely, so much worse the further north you go?)
Well, the correlation may not really exist anyway. Why not? Because the authors don’t even present data on actual rainfall. As Orac says: Note that the authors did not correlate autism prevalence directly with raw mean precipitations but instead used a "relative precipitation variable." When I see something like that, I know right away that there was no correlation between raw mean precipitation levels and autism. If there had been, you can rest assured that the authors would not have bothered to go to the trouble to do this little bit of mathematical legerdemain–excuse me, I mean "transformation." This is a general principle of epidemiological studies: If there’s a correlation with raw data, then there’s no reason to do any sort of mathematical adjustment.
And there are other flaws. There’s no control for ‘urbanicity’ – a measure of how built-up an area is. (Alabama scores lowest on urbanicity, as well as rainfall.) Rainfall varies from place to place & from year to year (we Waikato folks were reminded of this last summer!) – but rates of autism don’t. And the study didn’t even attempt to look at ‘genetic susceptibility’ in the study population, which means that they can’t possibly draw any conclusions about links between this & the prevailing weather. To quote Orac again: To use the words "genetic susceptibility" in the conclusions and to say that this study somehow supports an interaction of genetic susceptibility and environmental factors is just plain incorrect.
But I guess ‘no link between weather & autism’ isn’t such an exciting headline…