Wednesday, November 05, 2008

Is Precipitation Associated with Autism? Apparently Not.

A while back I wrote a critique of the TV hypothesis by Waldman et al. I noted the likely confound is population density, which should not be considered a "fixed effect" in Waldman's methodology (an interesting statistical methodology that is apparently used in Economics frequently). When we talk about population density as a confound, we're really using it as a proxy of other confounds that are clearly not fixed in time. These more specific confounds could be things like awareness, availability of autism specialists, etc.

In general, studies like Waldman's and Palmer's likely suffer from the fundamentally incorrect assumption that regional differences in the administrative prevalence of autism reflect a real difference in actual prevalence. But I do believe it is possible to use administrative data to draw preliminary conclusions, so long as confounding factors are accounted for.

My intention in writing this post was to walk through an analysis of publicly available data, controlling for population density, to see if the rainfall effect remained. I fully expected there to be a naive ecological association between precipitation and autism. To my surprise, the effect didn't appear to exist in the first place at the US level, and there was no need to control for confounding.

The following is a scatter graph of annual precipitation by state (1971-2000) vs. the 3-5 IDEA prevalence of autism (estimated for 2006).

There's not even a trend in the expected direction. This is quite the head-scratcher, and it left me wondering what was going on. Why is it unexpected? Let's first look at a population density map of the US.

It would be reasonable to expect that counties with a higher concentration of people will have higher rates of autism diagnoses, due to increased awareness and a greater availability of autism specialists. Let's now look at a map of precipitation in the US.

The correlation between precipitation and population density is quite clear, isn't it? Why didn't we see an association trend in the expected direction in the scatter graph then? First, it seems that a few states bring the slope down. These would be states with a low autism prevalence but high precipitation rates, like Louisiana, Alabama and Mississippi.

That's a bit of bad luck for Waldman et al. Additionally, we don't have that many data points. There's unfortunately too much variability in this US-level data, which makes it pretty inadequate. Perhaps using 6-11 IDEA prevalence would be better than 3-5 prevalence. In any case, it's doubtful statistical significance would be achieved, and even if it were, it is doubtful it could withstand controlling for population density.

I think the association needs to be revisited in a different way. But this exercise left me wondering why Waldman et al. decided to only look at counties from certain states, namely, California, Oregon and Washington (with California not showing a clear association).

I'm going to suggest cherry-picking might have occurred when it comes to Oregon and Washington. In order to argue this point, I will simply post population density and precipitation maps of each of these states. You will see that the pattern in these two states is fairly unique. Most people live in the west side of the state, and that's also where it rains.

To summarize: (1) It was not easy to confirm the reported association. (2) Analysis of any such associations should account for population density. (3) Cherry-picking might have occurred in this particular case.


  1. Look closer at the maps of Washington. Do you notice that the penisula and Pacific coast areas have lots more rain, and much less population?

    Does Forks, WA in Clallam County with at a 100 inches of rain per year have an equivalent level of autism compared to Seattle, WA in King County with about 38 inches of rain per year?

  2. Do bears crap in the woods? To be honest I have never observed that particular phenomenon personally so perhaps they do not.

  3. @HCN: Yes, there are a few places with little population and a lot of rain. But then these localities have a good proximity to areas of high density. Proximity presumably matters when it comes to diagnoses.

    The areas of very high population density tend to have a moderately high precipitation rate, but not super-high. (Too rainy to live there, I guess).

    But I think what mostly drives the correlation is that big deserted regions have very little rain.

  4. You've obviously never been to the Olympic Peninsula. Do you consider a more than five hour drive and only with good ferry scheduling close proximity? Double it if you have to drive around. There is no freeway.

    Sure, Bainbridge and Silverdale are closer... but they are in the rain shadow, and near the ferries. The towns on the coast where there is little populaiton and the heavier rain falls are much more isolated. Plus there is a fairly large mountain range between them and the ferries (that is why there is high rainfall).

    We tried to find a motel room in Forks many years ago, there were none to be had. We ended up driving, driving, driving some more to a place further up towards Port Angeles. It is a very isolated area. Lots of people don't realize what the distances are in this part of the country.

  5. Correction, it is about a 3.5 hour drive... double that when there are ferry problems.

    But it is still silly.

    What are the differences in Texas? Like wet and rainy Houston versus dryer Dallas and Ft. Worth?

  6. What I mean is that it's still realistic for someone there to go see a doctor in the city. Not so much for someone who lives in the east side of the state. But again, it doesn't have to be a perfect correlation, just statistically significant.

    I don't have data on Texas, but that would be an excellent state to look at. First, they have 254 counties. The general pattern is similar. It rains more in the east, and that's where the cities are. But they seem to have a nice gradient of precipitation going from west to east.

  7. I'm sorry, Joseph, but you are wrong. It is actually easier to get services on the east side of the state compared to the Olympic Peninsula.

    It is flatter, has actual freeways, and many more decent sized towns. Look at the population density of Washington. You can see Yakima, Wenatchee, the Tri-Cities and Spokane as wih clumps of population centers. And they are not separated by a freaking large clump of mountains and body of water.

    Check out this list of Rite Care Childhood Language centers: ... I see four in Eastern Washington (Kennewick, Yakima, Spokane and Wenatchee), and of the four in Western Washington the closest to the Peninsula is in Olympia.

    Though what you say could be said for Oregon, in that the only one east of the Cascades is in Bend, OR. Though you will notice that most of the clinics are down the I-5 corridor, which pretty much follows the Willamette Valley. Not much on the coast, where it rains more... and has those very amusing "Tsunami Evacuation Route" signs.

    But still, the area in with the most rain in the three coastal states is also the area with the least services. It has issues with industry (lumber) and access (one major two-lane highway that goes around the Olympic Mountains). It is one of those areas where it is difficult to get medical personnel to move to. It is an outlier... in more ways than one.

    Texas would be a better region to check out this hypothesis. It has population centers scattered from south to north coast to the inland, with a good highway system.

    Another place would be Vancouver Island. It is an island which does get lots of rain, but it is more evenly spread from south to north. It also has one major population center on its south end, and some others scattered up its east coast to Campbell River... with some smaller towns further norh, and on its west coast (like Tofino, which is a five hour drive from Port Alberni, BC).

    I would really like to see a list of the counties, and all the data pertaining to rainfall, population, relative wealth and access to services. Especially for the areas that get more than twice rainfall of the I-5 corridor cities.

  8. Well, they have an autism map for all 3 states in the original paper. The rate in the east side of Washington state is lower than the median for the state. Clearly, services are not the only thing that will drive special education prevalence.

  9. I saw that map in their earlier TV paper, and it smoothed over big areas. Specifically leaving out Clallam County as high autism, which has very high rainfall.

    But BINGO!!! You do have a hint of what is going on, and it is similar to what the Autism-Diva wrote about. It has to do with economics.

    The other side of the state is mostly agricultural with a lower average income and a larger migrant worker population. That is a gross generalization, but still the main industry is based in agriculture.

    Looking at their maps from there previous paper, I think there is lots of data smoothing. The over/under state autism average may not have been normed, and I am suspicious of their lack of showing much for the Tri-Cities (it is a an area with lots of nuclear research).

    The problem is that the counties in this state are BIG, and I believe they are leaving out lots of information. I still don't trust their reasoning.

  10. Note: I've posted a follow-up where I look at autism and precipitation time series from California, going from 1930 to 2000.

  11. Try overlaying maps of industry and autism and there seems to be a relationship. Is it the coctail of toxins that our evironment is mixing?
    this is a map I compared to your map

  12. That's probably simply because places with a lot of industry are also places with a lot of awareness of autism.

  13. Southeast and main Asian pandora jewelry countries have twisted rubies for centuries, cheap pandora bracelets but research as to where, and how to find more deposits is Pandora charms spare, and production has figured out how and mining companies,” Pandora beads Giuliani says, to look at exactly the right time and place.” pandora set Farther investigation of claret formation, based on tectonic scenery, cheap pandora geochemistry, fluid inclusions and isotopic ratios, allowed discount pandora Giuliani’s lineup to remodel a new prototype for the French Institute pandora 2010 of Research for Development (IRD) and the National Scientific pandora sale Center of Research, two government-sponsored knowledge Pandora Bangles and technology research institutes that aim to aid in the sustainable cheap pandora bracelets development of developing countries. Before the collision pandora bracelets prices of the Eurasian and Indian plates, lagoons or deltas sat in the regions where marble is giant, pandora bracelets and charms he says, “and there is the brains to expect that the new pandora bracelets sale thoughts should help development of the artless capital.” discount pandora bracelets Virginie Garnier, Gaston Giuliani and Daniel Pandora necklace Ohnenstetter urban the shape to do just that. They work for the garnet cheap pandora charms genesis. While studying the bedrock in Vietnam in 1998, the discount pandora charms French players found rubies, which detained traces of aluminum, chromium pandora charms sale and vanadium from universities, international corporations, governments pandora charms 2010 and why the rubies got there, and has created a paradigm Pandora beads to help these evaporites, Garnier says, when the Eurasian cheap pandora beads and Indian plates collided, raising the Himalaya Mountains.