One More Air-Pollution Study

In the last post I confused this study (PDF) with an earlier one by the same group of researchers; I wrote about the earlier one, but linked to a post on Paul Whiteley's blog about the more recent one, which was published just last November.

(I also started describing this one, and then switched to describing the earlier one; in my last post, only one of the studies I mentioned used air-pollution data from the EPA's air-monitoring stations. The earlier study by these authors only used proximity to high-traffic roadways as their variable indicating pollution exposure.)

This paper combined the methods of the two studies I wrote about in the last post; it used the same pool of children born in California between 1997 and 2006 and drew on two sources of data on air pollution at the time and place those children were born: the EPA's air-quality data that I wrote about yesterday, and a computer model of average traffic flow, and exhaust emissions, along California's major roadways.

To some extent, you could see it as a more geographically dispersed version of the study I described yesterday that looked at prenatal exposure to air pollution in just Los Angeles County. 

(Weirdly, the LA-County-only study, though restricted to a smaller geographic area, involved way more people than the two traffic-related studies: 7,603 autistic and 75,782 control subjects, as opposed to this study's 279 autistic and 245 control subjects.)

But the relatively straightforward EPA data, which are direct measurements of the concentration of various pollutants at regular intervals, and which are only abstracted from each child's actual prenatal exposure in that 1) they do not measure what concentration of those pollutants actually got into the mothers' bodies, much less the fetuses', and 2) they were taken at sites some distance away from where the children actually live. 

So it's not a perfect data source, but it's still a lot more directly reflective of reality than this computer model seems to be:

The principal model inputs are roadway geometry, link-based traffic volumes, period-specific meteorological conditions (wind speed and direction, atmospheric stability, and mixing heights), and vehicle emission rates. Detailed roadway geometry data and annual average daily traffic counts were obtained from Tele Atlas/Geographic Data Technology in 2005. These data represent an integration of state-, county-, and city-level traffic counts collected between 1995 and 2000. Because our period of interest was from 1997 to 2008, the counts were scaled to represent individual years based on estimated growth in county average vehicle-miles-traveled data. Traffic counts were assigned to roadways based on location and street names. Traffic volumes on roadways without count data (mostly small roads) were estimated based on median volumes for similar class roads in small geographic regions. Meteorological data from 56 local monitoring stations were matched to the dates and locations of interest. Vehicle fleet average emission factors were based on the California Air Resource Board's EMFAC2007 (version 2.3) model. Annual average emission factors were calculated by year (1997-2008) for travel on freeways (65 mph), state highways (50 mph), arterials (35 mph), and collector roads (30 mph) (to convert to kilometers, multiply by 1.6). We used the CALINE4 model to estimate locally varying ambient concentrations of nitrogen oxides contributed by freeways, nonfreeways, and all roads located within 5 km of each child's home. Previously, we have used the CALINE4 model to estimate concentrations of other traffic-related pollutants, including elemental carbon and carbon monoxide, and found that they were almost perfectly correlated (around 0.99) with estimates for nitrogen oxides. Thus, our model-based concentrations should be viewed as an indicator of the traffic-related pollutant mixture rather than of any pollutant specifically.

So, to arrive at an estimate of how much of a certain category of air pollution (traffic-related air pollution) each mother and child in their study had been exposed to, they used a computer model to come up with average emissions for vehicles all over the state, traveling at various average speeds corresponding with their various categories of roads, for each year in their study. Then they entered that, along with all the other types of data mentioned above (winds, atmospheric conditions, traffic volume, road layout) into another computer model to arrive at the final answer.

I'm not criticizing their model; it actually seems like a pretty good one to my untrained eye. But my point is that there's a lot of extrapolating, averaging, assuming that what's true for location x will also be true for location y, and other things that make the model work but aren't grounded in direct observation and thus might not actually be true. 

That will be the case for any model, and this one has a few serious gaps in its data pool. They're missing eight years of traffic data from their eleven-year "period of interest," so they have to guess at what those numbers might be based on expected growth in traffic volumes. They're also missing traffic counts for some roads, so they estimate them based on the counts for other, similarly-sized roads.

It bears repeating that this model was not their only source of data on pollution exposure; they also used direct measurements taken by the EPA air-monitoring station(s) nearest to study participants' houses throughout the study period.

For traffic-related air pollution --- the type of pollution exposure they modeled rather than measured directly --- they found a difference between the highest- and lowest-exposure groups (with the former three times as likely to develop autism as the latter), but no difference between the lowest-exposure group and the two groups in the middle.

For the specific pollutants measured at EPA air-monitoring stations --- coarse and fine particulate matter, nitrogen oxides, ozone --- they found an increased likelihood of autism associated with greater exposure to particulate matter and nitrogen oxides, but not ozone. This effect was strongest during the third trimester of pregnancy. 

Unlike the other study I described that used the EPA air-quality data, this one did not find any change in the pattern when they adjusted for sociodemographic variables like child's sex, race/ethnicity, parents' educational level, mother's age, or mother's smoking during pregnancy.

Volk, H., Lurmann, F., Penfold, B., Hertz-Picciotto, I., & McConnell R. (2013). Traffic-Related Air Pollution, Particulate Matter, and AutismAir Pollution, Particulate Matter, and Autism JAMA Psychiatry, 70 (1) DOI: 10.1001/jamapsychiatry.2013.266

Bizarre Things Purported to Cause Autism: Early Exposure to Air Pollution

A quick note: my (very) occasional "Bizarre Things..." series was never intended solely as a crank roundup. No, in my mind I resolved to cover every autism hypothesis*, however well- or ill-founded, plausible or implausible, that I ever heard of and thought "well, that's weird!"

Indeed, I have tagged (though not titled) a post discussing a pretty rigorous, well-thought-out study investigating something that's been common wisdom for a long, long time with the "bizarre hypotheses" label just because its tentative conclusion --- that there may be something more to the relationship between autism diagnoses and social class than the fact that rich people can afford to get their children seen by specialists and poor people can't --- surprised me.

So sometimes bizarreness is in the eye of the beholder, I guess.

Anyway, on to the business at hand.

Paul Whiteley has written a couple of posts on the idea that exposure to air pollution might play a role in determining whether a kid develops autism; he talks specifically about these two studies working with the same two data sets: the state of California's Department of Developmental Services' records of all children diagnosed with autism in the state, and where they lived at , and also data from the federal Environmental Protection Agency's Ambient Air Monitoring Program.

I've written about the former data set before, but not the latter. The EPA measures six major outdoor air pollutants (I'm sure they do a lot more, too, but these six are the ones that affect air-quality indices): two sizes of "particulate matter" (dust, ash, soot, smog), the "coarse" particles measuring between 2.5 and 10 micrometers in diameter or length, and "fine" particles smaller than 2.5 micrometers; carbon monoxide; nitrogen oxides; sulfur dioxide; lead; and ozone. They have monitoring stations set up all over the country, particularly thick on the ground near big, sprawling cities. How often the stations record measurements varies with what kind of equipment is being used to take them; some pollutants can only be measured daily, or once every few days, though some can be measured hourly. Either way, it's a huge volume of data.

The authors of the more recent study (PDF) restricted their analysis to Los Angeles County, so I can actually tell you how many monitoring stations' data that would encompass.

Map showing the locations of all the EPA air monitoring stations in LA County --- made by me!

This PDF lists, among lots and lots of other things, all the EPA air-monitoring stations in California and where they are located. Within Los Angeles County, it looks like there are nineteen: one in Commerce, one in Azusa, one in Burbank, one in Industry, one in Compton, one in Glendora, one at the Los Angeles International Airport (LAX), two in Long Beach, two in Los Angeles, one in Pasadena, one in Pico Rivera, one in Pomona, one in Vernon, one in Reseda, one in Santa Clarita, one in Santa Fe Springs, and one at the Van Nuys Airport. So I guess that's technically five that are somehow part of LA or attached to it. Not all of them measure every one of the six pollutants, either, but you can see which of them track what in the PDF I linked earlier.

The earlier study** focused on families living near highways throughout the state of California, and while I might theoretically be able to put together a list of all the cities and towns that have highways running through them, and cross-check that with the EPA's records of where their monitoring stations are (assuming they publicize them all), it sounds like more armchair detective work*** than I want to do. So I can't tell you how many stations are contributing their data to this study, but I'd guess that it's more than the LA County-only study used.

Anyway, both studies involved matching the geographic location of each autistic child in the study with some other spatial variable: for the earlier study, this other variable was distance between where they were born and a freeway or major road; for the more recent study, it was the EPA monitoring station nearest to where they were born. In that study, the researchers looked at the measurements for each pollutant recorded nearest to each child's birthplace averaged over each trimester of gestation.

Both studies compared their autistic subjects to same-age, same-sex peers from the same general area (LA County for the one study, not specified for the other); the ratio was 1:1 in the smaller, older study and 10:1 (control:autism) in the bigger, newer one. So the idea was, I guess, to check whether kids living in the same county, city, suburb, or whatever as a kid without autism tended to have lower prenatal exposure to air pollution (new study) or live somewhat further away from the closest high-traffic road (old study), than the kid with autism.

What did they find? It's complicated; in the newer, just-LA-county study, they found that exposure to some of the six pollutants was actually associated with a slightly lower likelihood of having an autism diagnosis, at least in the (more****) raw data. The only pollutant to show a significant increase in odds ratio (a measure of how much more likely a child exposed to a given pollutant is to develop autism than an unexposed child; if it's less than 1, it means less likely, more than 1 means more likely) before logistic regression was ozone, which gave a 1.19 odds ratio. (That's also the biggest increase found anywhere in this study, regression or no).

After adjusting for a bunch of variables they had expected to correlate with pollution exposure (maternal age, maternal education, race/ethnicity, gestational age and others), the odds ratio for ozone went down while the odds ratios for the other pollutants (nitrogen oxides, carbon monoxide and particulate matter) went up. Where they had been sitting at about 0.8 or 0.9 (i.e., maybe ten, fifteen or twenty percent less likely?), they moved to about 1.05. 

I have only the faintest notion of what a logistic regression actually does, but my understanding of it is that, when researchers have a lot of variables closely intertwined with the variable they're trying to study, they use a logistic regression to "correct for" those other variables. It's like a way to try and zero in on the one strand in the snarl that you're trying to trace.

Anyway, they also did another regression, this time by maternal education only, and compared odds ratios for autism risk associated with each pollutant among three groups: mothers with less than a high-school education, mothers with a high-school education, and mothers with more than a high-school education. Except for ozone, the odds ratio for each pollutant went up slightly as maternal education increased, with the biggest differences between the least- and most-educated categories. 

Somehow, the mother's level of education affects how strongly exposure to these airborne pollutants predicts whether her children develop autism.

I have absolutely no clue what to make of that, if I'm even reading it right.

The roads study's outcome was less brain-twisting: they found a correlation between living near a freeway (i.e., a state or interstate highway) and getting a diagnosis of autism, but they found no such correlation for other high-traffic roads. Since it's probably not the case that different kinds of pollutants are being spewed out by vehicles on highways vs. other big, heavily-traveled roads*****, I'm going to follow David Gorski's lead and write off the freeway association as an artifact of data-dredging.

Becerra, T., Wilhelm, M., Olsen, J., Cockburn, M., & Ritz, B. (2012). Ambient Air Pollution and Autism in Los Angeles County, California Environmental Health Perspectives DOI: 10.1289/ehp.1205827

Volk, H., Hertz-Picciotto, I., Delwiche, L., Lurmann, F., & McConnell, R. (2010). Residential Proximity to Freeways and Autism in the CHARGE Study Environmental Health Perspectives, 119 (6), 873-877 DOI: 10.1289/ehp.1002835

*Yes, I have yet to do posts on the big ones, like mercury, thimerosal (really part of the mercury one, but some people propose mercury-based explanations that don't mention thimerosal, so they probably deserve separate posts --- anyway the thimerosal one would be long enough even if it weren't folded into a larger post on mercury compounds), MMR. It's less a matter of not knowing what I want to say than it is a matter of figuring out whether, or how, to marshal all the available evidence.

**I belatedly find out that this study is actually a precursor to the one Whiteley is writing about, not the same one. The one he writes about takes an approach that looks to me like a blend of the two I'm writing about now: looking at both air-pollution data and proximity to high-traffic roads.

***Actually, it's not even armchair detective work, as I am not in an armchair when I use the computer. It's armless-hard-wooden-chair detective work, which is somewhat more grueling than armchair detective work. 

****I'm not sure how much, if any, of the data in this paper can be considered "raw" when you consider how much tinkering they did with the air-pollution data.

*****You could perhaps make an argument that it could be the case, what with there probably being more trucks on the highways than not, and with trucks using diesel instead of gasoline. I am unsure if that would make a difference in how much particulate matter or nitrogen oxides are produced, and also quite skeptical that such a difference would show up at all in a study like this.

Cognitive Sex Differences within Autism - Part I

One hypothesis about why there are so many more boys and men on the autism spectrum than there are girls and women is that boys and men, for whatever reason, are more likely to get diagnoses of autism spectrum conditions.

One of the reasons commonly put forward by people (including me) making this argument is that the diagnostic criteria reflect a picture of autism derived disproportionately from autistic boys and men, and autistic girls and women differ sufficiently from their male counterparts in their abilities, behavior, developmental histories etc. to ensure that they will often fail to meet those criteria.

While browsing PLOS ONE I found a couple of recent-ish articles looking into what those differences might be.

The more recent study, just published this week, involved giving four largish groups of participants (one autistic and one neurotypical group within each sex, with each group having thirty-two people in it) a battery of tests designed to measure five different skills: theory of mind, emotion recognition, executive functioning, perceptual attention to detail, and manual dexterity.

What are all those things and how would you go about measuring them? Well, the first of those skills, theory of mind --- also called cognitive empathy or mind-reading --- is one I've written a lot about on this blog. I've also written a lot (usually in the same posts) about the tool psychologists seem to use most of the time to assess this skill in adults: the Reading the Mind in the Eyes Test. I don't think I need to say anything more about either the skill or the test; the test is relatively straightforward, a series of black-and-white photographs of faces, cropped to show only the eyes, paired with a choice of four emotion words. You are supposed to guess which of the words best describes what the person whose eyes are in the photograph is feeling.  

I listed theory of mind and emotion recognition separately, even though the Reading the Mind in the Eyes Test obviously involves both. There's another test the researchers used, one that they describe as a test of emotion recognition, that reads to me like it's almost the same thing as the Reading the Mind in the Eyes Test, only it uses pictures of entire faces instead of just eyes. It's called the Karolinska Directed Emotional Faces Test.

To test the participants' executive-functioning abilities, the researchers used something called the Go/No Go Task. This involves pressing buttons on a computer keyboard in response to certain cues appearing on the computer screen; in this study the cues were arrows pointing either to the right or to the left. (Right and left were each associated with a different target key: P for right and Q for left. So the people taking the test were literally minding their P's and Q's!) The test also includes extraneous cues that test takers are not supposed to respond to. So what's being tested is not just one's ability to hit a key when a light goes on; it's also one's ability to restrain one's key-hitting impulse when something close to, but not exactly like, the target stimulus appears.

There were some other things the researchers included under the umbrella of executive functioning: working memory (tested by having the participants repeat nonsense syllables), word generativity (ability to come up with lots of words beginning with the same letter in a short amount of time) and motor planning, which they tested using the Assembly subtest of the Purdue Pegboard Test. (This involves putting simple objects together in a set order using whatever hand they tell you to use.) This obviously assesses manual dexterity as well as the ability to figure out what you need to do in what order (i.e., motor planning), although they also had the participants complete the other parts of the Purdue Pegboard Test to test manual dexterity alone.

Finally, the test they used to measure attention to detail was the Embedded Figures Test, which I've also blogged about before, noting that autistic people have shown a particular aptitude for this task.

The results are mostly unsurprising: the autistic study participants did worse at both of the facial-expression-interpreting tasks, and also at pressing the right key in the Go/No Go task. On the two language-related memory tasks, they performed no differently than the control group. (This is not surprising given that all of the participants are described as "high-functioning," having average-to-above-average IQs (about 115, plus or minus about 15 points) with verbal IQ greater than or equal to performance IQ).

The emotion-recognition data get more interesting when you look more closely at them, though; apparently the autistic study participants had more trouble identifying some emotions than others. Fear was the one it took them longest to recognize; it was also the hardest one for the non-autistic people to identify, too, but the gap between the two groups' average reaction times is the biggest in this category. It took the autistic study participants, on average, somewhere between four and five seconds (closer to five) to identify the fearful faces. For the non-autistic participants, it took maybe three and a half seconds. Happy faces were the easiest for both groups to identify, and the difference between groups was only half a second compared to the 1-1.5 second gap in their response times to fearful faces. All the other reaction times are clustered within a relatively narrow range for both groups: between 2.5 and 3 seconds for the autistic group, and between 2 and 2.5 for the control group. So, even though fear was tricky for everyone, it proved especially challenging for the autistic people. The study authors guess that maybe this is because autistic people studiously avoid looking at other people's eyes, and the eyes are apparently the most important cue that someone looks frightened.

There were also a couple of novel findings, too, though. For instance, on the Assembly component of the Purdue Pegboard Test, it was only the autistic men who had any problems relative to their same-sex control group. The autistic women were, you might say, indistinguishable from their peers. 

But the most surprising thing, for me, was their results on the Embedded Figures Test. This is a pretty solidly defined Thing Autistic People Are Really Good At, yet in this study, where there was a difference between autistic and non-autistic groups, it was the non-autistic people who found the hidden shape faster. (That was also true for the men and not for the women).

In general, the authors describe a pattern of male and female autistics having more or less the same degree of impairment in what they call the "hallmark" of autism spectrum conditions, which is inability to pick up on another person's nonverbal cues to their mental or emotional states*, but with the autistic men, and not women, also having additional impairments in other domains.

The authors point out that this contradicts what had been conventional wisdom about cognitive sex differences within autism, which held that autistic women were usually more severely disabled than autistic men**.

I haven't forgotten about the other, earlier study; I'm just going to give it its own post so as not to make this one overlong.


Lai, M., Lombardo, M., Ruigrok, A., Chakrabarti, B., Wheelwright, S., Auyeung, B., Allison, C., , ., & Baron-Cohen, S. (2012). Cognition in Males and Females with Autism: Similarities and Differences PLoS ONE, 7 (10) DOI: 10.1371/journal.pone.0047198


*This elusive quality can be called theory of mind, mentalizing, "mind-reading" or empathizing. It may also involve more than one cognitive, emotional or perceptual ability, but this study doesn't really go into that.

**To be REALLY picky, this study doesn't address the issue of intellectual disability, as no one who participated in it would meet the criteria for having ID. This shows that autistic men of slightly above average IQ are more likely to have problems with motor planning and figure disembedding.

Another "Extreme Female Brain" Sighting

EXECUTIVE SUMMARY: Simon Baron-Cohen's E-S theory of autism, sex differences and whatever else he's applied it to lately states that people have one of three basic cognitive styles: Type E, the empathizer, who understands people and relationships; Type S, the systemizer, who understands abstract ; and Type B, who can do both. Prof. Baron-Cohen has identified autism as exemplifying the Type S cognitive style taken to extremes, but hasn't written much about its opposite, the extreme Type E, or Extreme Female Brain. He allows that it must exist, but he doesn't think it would be as disabling in modern society as an Extreme Male Brain --- autism --- is. Other people (Crespi and Badcock) have suggested paranoid psychosis as the condition typical of the Extreme Female Brain, as it involves being morbidly obsessed with other people. In the paper I discuss here, a pair of evolutionary psychologists make the case that the psychopathology characteristic of the Extreme Female Brain is disordered eating.

The paper I discuss in this blog post describes a series of four different "experiments" in which groups of college students, of varying size (as large as n = 160, as small as n = 37), are given multiple psychometric exams. They are tested for the following things: disordered eating, sensitivity to other people's judgments, Empathizing Quotient (EQ), Systemizing Quotient (SQ), performance on various tasks considered representative of either empathizing or systemizing (e.g. guessing what emotion people in photographs are feeling, or mentally rotating three-dimensional objects), and schizotypal personality disorder. Some of the data are compatible with the idea of a high-empathizing, low-systemizing Extreme Female Brain that is particularly susceptible to disordered eating, and some are not.
________________________________________________

Thought to exist, but seldom glimpsed in the wild

Today on the fascinating psychology blog BPS Research Digest, I saw that some other researchers think they've found a candidate for an Extreme Female Brain, which according to Simon Baron-Cohen's E-S theory ought to exist, probably in about the same numbers as the Extreme Male Brain (which he identifies as autism) does, but which hasn't been described in the psychological literature.

The new article, by evolutionary psychologist Gordon G. Gallup, Jr. and assistant professor of psychology Jennifer Bremser --- both affiliated with the State University of New York --- and published in the journal Evolutionary Psychology, tries to make the case that eating disorders stem from the heightened sensitivity to others' judgments that is characteristic of the Extreme Female Brain.

What they did was to give a group of undergraduate psychology students at SUNY a battery of tests: the Eating Attitudes Test (EAT), which looks for disordered attitudes about eating (sample questions: "I feel extremely guilty after eating"; "I feel that food controls my life"); the Fear of Negative Evaluation Scale, which measures how anxious you are about what other people think of you (sample questions: "I often worry that I will say or do the wrong thing"; "I feel very upset when I commit some social error"; "When I am talking to someone, I worry about what they may be thinking about me"); Simon Baron-Cohen's Empathizing and Systemizing Quotients (which I have critiqued here and here); the Schizotypal Personality Questionnaire (PDF), which tests for nine traits associated with schizotypal personality disorder: ideas of reference (which means thinking that other people, or the media, are talking specifically to or about you when they aren't really), excessive social anxiety, odd beliefs or magical thinking, unusual perceptual experiences, odd or eccentric behavior, a lack of close interpersonal relationships, odd speech, constricted affect (i.e., not being very good at showing one's feelings), and general suspiciousness; and several tests that measure skills that would be associated with either an empathizing or systemizing cognitive style: the Reading the Mind in the Eyes Test, the Intuitive Physics Test and the Redrawn Vandenburg and Kuse Mental Rotations Test (Version A). Then they looked for relationships between all of these different test scores, and between any given score and the test taker's sex.

Their hypothesis --- that an empathizing cognitive style is more characteristic of women than of men, and tends in its extreme form towards disordered eating and social anxiety --- would predict the following: women scoring higher, on average, on the EQ and Mind in the Eyes tests, and lower on the SQ, Intuitive Physics, and Mental Rotation tests; women scoring higher than men on the measures of disordered eating and fear of negative evaluation, and high scores on those measures being correlated with high scores on the EQ and Mind in the Eyes tests for both sexes. 

They found significant differences between the sexes on the measures of disordered eating and fear of negative evaluation (both were higher for women), and also on the EQ, but not on the SQ. Women were also likelier to display something called "empathizing bias", which they defined numerically as each study participant's z-score on the EQ minus their z-score on the SQ. (Z-scores are a measure of how far an individual's score on a given statistical measure is above or below the average score. A negative number means the person, or sub-group, scored below average, while a positive number means they scored above average). Thus, the empathizing bias score is supposed to represent how much better a person does at empathizing than systemizing (or not, if it's zero or a negative number).

There was also a disparity between how participants scored on the EQ and SQ versus how they did on the more "objective" measures of systemizing or empathizing skill, the Intuitive Physics, Reading the Mind in the Eyes and mental rotation tests. The sexes differed strongly on their (self-reported) EQ scores, but performed almost identically on the Reading the Mind in the Eyes test. Obviously guessing what emotion a black-and-white photograph of a stranger's eyes is meant to convey is very different from accurately gauging the emotional states of one's relatives and friends, so that still leaves open the possibility that women perform better at this task when it involves real people who are known to them, but it's just as likely that women tend to overrate, and men to underrate, their interpersonal sensitivity on self-reported measures, and that real differences between the sexes on this ability are minimal. 

The opposite pattern occurred between the SQ and one of the "objective" measures of systemizing ability: men and women scored equally on the SQ, but men outperformed women on the Intuitive Physics test. (I can't find any results on the mental-rotation test broken down by sex).

There was also a very inconsistent pattern of correlations between disordered eating, fear of negative evaluation and the various systemizing and empathizing variables. High scores on the disordered-eating measure correlated with high scores on the fear-of-negative-evaluation (FNE) measure, and also with high scores on the EQ. There was no relationship between SQ scores and either disordered eating or fear of negative evaluation, and EQ and SQ scores were positively correlated with each other. Most strangely, SQ scores were correlated positively with empathizing bias: the number that reflects the imbalance between one's empathizing and systemizing abilities!

Also, within each sex the overall pattern of disordered eating patterns correlating with greater empathic ability (whether self-reported or "objectively" measured) seems to fall apart. 

Two graphs, Figure 1 and Figure 2, show bars for average male and female performance on both the EQ and the Reading the Mind in the Eyes Test, at low and high ends of the disordered-eating spectrum. According to these graphs, women score about the same on these tests regardless of whether they show disordered eating patterns or not, and for men the results are all over the place. On one test (the Mind in the Eyes), men who score low on the Eating Attitudes Test do equally well with all women, while men who score higher do more poorly on average, but their range of results is very, very broad. The error bar is wider than the bar on the graph is tall. 

On the EQ, the men show more or less the pattern the researchers expected: men at the high end of the disordered-eating spectrum scored higher than men at the low end on average, but for the former group especially there was a lot of within-group variation.

They also found correlations between disordered eating, fear of negative evaluation and several of the traits that make up schizotypal personality disorder: disordered eating correlates positively with ideas of reference (i.e., thinking everything refers to you, specifically), magical thinking and suspicion. Fear of negative evaluation correlates positively with ideas of reference, suspicion, social anxiety and constricted affect. (This latter makes sense --- if you are afraid people are judging you, of course you are going to try and minimize the extent to which those people are able to tell what you're thinking or feeling! It's called a "poker face".)

They also wanted to see if any of these psychological problems --- disordered eating, fear of negative evaluation, schizotypal personality disorder or any of its component traits --- interfered with one's ability to rotate three-dimensional objects in one's head. (Their reasoning for doing this wasn't just for giggles --- let's take a bunch of emotionally stressed people and ask them to do something both pointless and difficult! Hilarious! --- it can be hard to shake the idea that a lot of psychological experiments are done with precisely that objective --- but because they wanted to see if people who were unusually and morbidly concerned with what other people thought of them, and were thus "hyper-mentalizers" who would be reckoned extreme empathizers in Simon Baron-Cohen's typology would also show the expected impairment in systemizing tasks. 

Annoyingly, they did not compare people's scores on the mental-rotation task with their scores on other direct measures of systemizing or empathizing ability: they assumed, given earlier findings, that high scorers on the measures of disordered eating and fear of negative evaluation are also high empathizers. As I have said above, the data are less than clear in their support of this assumption. 

They found that people (they do not break it down by sex, probably because at this stage in the study they were dealing with a lot fewer participants) who scored high on their measure of disordered eating tended to do worse at mental rotations, consistent with their prediction. However, they also found a small (too small to be significant) positive correlation between mental-rotation scores and fear of negative evaluation. Of those two things, I'd consider the FNE score to be the more pure indicator of a "hypermentalizing" excess of concern for what other people think; there's a lot that's still unknown about the psychology of eating disorders, to the point that calling disordered eating a behavior born out of a fear of negative evaluation by others would be speculative.

The correlations they found between the various schizotypal personality traits and mental rotation ability surprised me, too. You'd think, if greater skill at mental rotation of three-dimensional objects indicated a logical, "systemizing" type of mind, that there would be a strong negative correlation between mental-rotation scores and those schizotypal traits that bespeak irrationality, like magical thinking or ideas of reference. (Especially magical thinking). But no, there's no correlation between either of those things and how well one does at mental-rotation tasks. The only components of schizotypal personality disorder that showed any significant relationship to mental rotation were social anxiety and constricted affect, both of which were negatively correlated. 

A negative correlation between skill at mental rotation and one's score on the social anxiety subscale of the Schizotypal Personality Questionnaire does fit with the researchers' idea of "hyper-mentalizing" at the expense of systemizing. But the lack of such a correlation with fear of negative evaluation does not fit, and instead makes one wonder if what's going on is, instead, anxious people's anxiety interfering with their ability to perform the rotations. 

In their Discussion section, the researchers do not comment on most of these discrepancies, but the one they do address --- worse performance on the Reading the Mind in the Eyes Test* by people with more severely disordered eating --- they attribute to those people's having greater, not less, emotional sensitivity**. Why, they're so sensitive they see things that aren't even there, so they do poorly on certain tests of emotional sensitivity and interpersonal acuity like the Reading the Mind in the Eyes Test.

I'm being silly, but that really is their explanation:

Comparable to the heightened sensitivity to sensory stimuli (auditory, visual and tactile) common among people with autism, individuals with the EFB [Extreme Female Brain] may be hypersensitive to social stimuli. Disordered eating may ameliorate the experience of negative evaluation anxiety that results from heightened sensitivity to social stimulation. 

Consistent with this idea, emotional processing deficits have been linked with eating disorders including the inability to recognize, label and describe emotions in detail and to link feelings with bodily correlates (Bourke, Taylor, Parker and Bagby, 1992 [link]; Eizaguirre, de Cabezon, de Alda, Olariaga and Juaniz, 2004 [PDF]; Garner, Olmsted and Polivy, 1983 [link]) Also, compared to healthy controls, women with anorexia had difficulty recognizing emotions from facial expressions and vocal tones (Jansch, Harmer and Cooper, 2009 [link]; Kucharska-Pietura, Masiak and Treasure, 2003 [link]).

Jones, Harmer, Cowen and Cooper (2008) [link] investigated emotional face processing in female undergraduates with high and low levels of disordered eating. Participants completed the Eating Attitudes Test-26 and the Facial Expression Recognition Task, a computer task in which participants view faces depicting 7 different expressions (anger, disgust, fear, happiness, sadness, surprise, and neutral) at different intensities. The participants with higher levels of disordered eating were less accurate in identifying happy and neutral faces. Among participants with high levels of disordered eating, there was a tendency to classify more happy faces as neutral, and more neutral faces as either angry or sad. In addition, there was evidence that reaction times to recognize disgust were longer, while reaction times to recognize fear were faster than participants with low levels of disordered eating. When disordered eating reaches clinical levels, the effects of hyper-mentalizing may manifest as mental state misattributions. This may be because they are using their own experience to model the experience of others, and their bias to classify emotions with a negative bias may influence their attributions. For instance, when shown pictures of women who are said to have overeaten, females with high levels of disordered eating ascribed more negative emotional states to these women than control participants do. Thus, women with higher levels of disordered eating appeared to use their own experience of overeating to describe how the other women would feel (Beebe, Holmbeck, Schober, Lane, Rosa, 1996 [link]).

It is also possible that the physiological and cognitive effects of starvation produce deficits in performance. In a study looking at performance on the Reading the Mind in the Eyes Test, individuals with anorexia performed worse compared to healthy controls (Russell, Schmidt, Doherty, Young and Tchanturia, 2009 [link]. 

An alternative explanation is that low scores on the Reading the Mind in the Eyes task does not represent a deficit of theory of mind ability, but rather an excess that reflects hypermentalizing. Abu-Akel (2003) [PDF] suggests that theory of mind dysfunctions range from the complete absence of the ability to represent other people's mental states (as shown in severe autism) to having the representational understanding of mental states, but a deficit in the ability to apply this understanding (as in Asperger's syndrome) to the abnormal or excessive attribution of mental states (as in schizotypy). It may be that this third class of dysfunction is misconstrued as a deficit, rather than an excess. Consistent with this interpretation, the RME is one of the most widely used instruments to investigate theory of mind performance in adults; however, scores on the RME depend only on accuracy. The test does not identify the nature of the errors that impede performance. For example, there can be errors of absence (failing to detect a mental state when it is present) and errors of excess (wrongly inferring a mental state in its absence). We suspect the second class of errors (errors of excess) account for most of the lower scores on this test among individuals with anorexia.

(They also seem to be arguing that people cursed with an excessive degree of awareness of other people's judgment of them might be self-medicating by starving themselves: if they are women, they will produce much less estrogen, which the authors believe "may also produce changes in empathizing and theory of mind performance." So that's a novel proposal --- people with eating disorders starve themselves to dull their razor-keen interpersonal sensitivity, in the interests of making their lives bearable. Obviously there's no evidence for this, and I've never seen anyone who has an eating disorder describe it like that, but at least it does people with eating disorders the courtesy of treating them as rational beings, which not all of the accepted explanations do.)

I do not doubt at all that people who have this kind of hypersensitvity exist --- I know a few! What I doubt is that 1) these people are all the same, and all have this skill for the same reason; 2) these people are any likelier to develop eating disorders than the general population; and 3) these people are necessarily bad at "systemizing," which seems to include such things as logical reasoning, pattern recognition, organization and spatial cognition. Indeed, the data from this same study suggest that people who are good at one half of the empathizing/systemizing duality seem to be good at the other half, too.

You can also count me as agnostic, still, on the question of whether women are, overall, better empathizers and men better systemizers. I don't rule it out, as at least some cognitive differences between the sexes do seem to exist, and show up consistently in studies, but as yet I don't see strong empirical support for it. 

Also, they dance around this in their Discussion section, but never directly address it: disordered eating is quite common in autistic girls and women, and even non-autistic people with eating disorders, particularly anorexia, perform similarly to autistic people on tests of mentalizing ability (the Reading the Mind in the Eyes test again), mirror self-recognition, emotion recognition, body awareness, executive function and central coherence (the Embedded Figures Test).

If autism is supposed to be the Extreme Male Brain, and the Extreme Female Brain is supposed to be its opposite, doesn't that complicate things a lot? 

Bremser JA, and Gallup GG Jr (2012). From one extreme to the other: Negative evaluation anxiety and disordered eating as candidates for the extreme female brain. Evolutionary psychology : an international journal of evolutionary approaches to psychology and behavior, 10 (3), 457-86 PMID: 22947672

*In the present study, it was only the male participants who showed this pattern, but the researchers cite lots of other studies of people with eating disorders, particularly anorexia nervosa, who show the same impairment. Some of the other researchers hypothesized that insufficient nutrition made those patients less able to perform complex cognitive operations of any sort, which sounds convincing to me.

**I am having UNBELIEVABLE trouble spelling "sensitivity" today. I always seem to leave out one of the "I's".  

Signal Transduction in Autism

EXECUTIVE SUMMARY: A study published this past summer analyzed tissue extracts from 20 donated brains, half (10) of which came from autistic donors. Half (5) of those people had histories of regression --- that is, they started out developing normally, speaking and everything, but then they lost some of the skills they had gained.

The brain tissue extracts were analyzed using a technique I describe in the main body of this post, that tests for the presence of a certain enzyme (protein kinase A, here) by giving it an opportunity to react with a sort of dummy peptide that can't really do anything except sit there and let the enzyme (and only that enzyme) act on it, and then introducing antibodies that will "tag" the altered peptides with an enzyme that will change a solution's color under certain conditions. This allowed the researchers to measure the relative activity of the enzyme across subjects or across brain regions; a similar measure, but using antibodies to the enzyme itself, rather than to its product, was used to measure the amount of enzyme present in each extract.

Using this method, the researchers found differences in protein kinase A activity and expression only in the frontal lobes, and only between the autism-with-regression subgroup of the autism group and both the controls and the rest of the autism group.

Protein kinase A is involved in intracellular signaling; it's one of the signal-boosting enzymes that helps the cell react quickly to changes in its environment. It modifies other proteins, affecting their activity. Some of its targets are proteins involved in neurotransmission (signaling between brain and nerve cells) and long-term potentiation (reinforcing those connections between neurons that are frequently used). It's this latter process that the study authors think may be disrupted in regressive autism.
_____________________________________________________

SFARI News posted some time ago on a study published on August 31 in PLoS ONE, comparing the amount of a certain enzyme present in tissue extracts from different regions of the brain between deceased subjects with and without autism who had donated their brains to the National Institute of Child Heath and Development Brain and Tissue Bank for Developmental Disorders.

The enzyme in question is protein kinase A, which plays a hugely important role in the cell, helping mediate a process called signal transduction, through which the cell is able to react to its changing environment, or to signals from other cells. In signal transduction, a molecule from outside the cell (usually a hormone) attaches to a receptor outside the cell and causes the receptor to change shape, thus altering the part of the receptor that's inside the cell and triggering a chain reaction of changes in enzymatic activity within the cell.

Protein kinase A participates in one particular signaling pathway: the one involving a class of receptors called G proteins, which are actually clusters of several smaller proteins that split apart whenever something attaches to its extracellular binding site. The now-mobile subunits then go on to do other things in the cell, most importantly to activate* an enzyme responsible for turning adenosine monophosphate (AMP) into cyclic AMP, which works as a signaling molecule inside the cell.

(Here is a cartoon from Nature Publishing Group's Scitable website illustrating that splitting apart of the G protein after a signaling molecule binds to its associated receptor; I adapted the image somewhat to make it less busy)

(Cyclic AMP)

Cyclic AMP is part of a class of molecules called "second messengers," which are small molecules that can bind to, and either activate or inhibit, a wide range of enzymes. Also, the enzymes responsible for making these molecules are regulated by receptors on the surface of the cell, so that when a signaling molecule binds to the receptor, the enzyme gets switched on (in the case of adenylyl cyclase, which is what turns regular AMP into cAMP) and starts churning out second-messenger molecules, which then go on to tinker with their target enzymes. In this way --- by coupling receptor binding with synthesis of these second-messenger molecules --- the cell can amplify the signal it receives, allowing it to react more quickly to changes in its environment.(Cartoon showing signal transduction using cyclic AMP as a second messenger, taken from this community college's Anatomy & Physiology II webpage. You can see how a hormone binding to its receptor frees up the receptor-coupled G protein to exchange its GDP for GTP and then go off and --- depending on the hormone --- either activate or inhibit adenylyl cyclase, which either starts or stops churning out cAMP, which goes on to do lots of different things, like activating enzymes, telling the cell to secrete various things, opening ion channels, etc. The only thing I don't like about this cartoon is that it only shows one cAMP molecule as the output of all the running around happening in the cell membrane, when really cAMP is being continuously produced by every active adenylyl cyclase. So, what that looks like, relative to the amount of hormone coming to the cell from outside, is more like this other cartoon, down below) (See, look at the arrows coming out of that yellowish triangle. One arrow splits into five, then 25, then more than you can clearly see. This table from the Memorial University of Newfoundland's cell biology webpage lists the number of molecules affected by each step in a cAMP-dependent signaling pathway, from the one molecule changed when a single molecule of hormone binds to its receptor, to the 10,000 molecules changed by the time adenylyl cyclase starts producing cAMP).

Anyway, protein kinase A is one of the enzymes activated by cAMP binding to it, and it is also mostly a regulatory enzyme --- that is, it activates or deactivates other enzymes. Protein kinase A does that by transferring a phosphate group from ATP (a small molecule made up of a sugar, a nucleotide base and three phosphate groups) to certain amino acid residues on any of its target proteins.

What kinds of proteins does protein kinase A regulate? Well, that depends on what kind of cell all this is taking place in. Every cell in the body contains a complete human genome; the differences between cell types are differences in which genes are expressed --- i.e., which proteins are present. So each cell type is going to have a different mix of proteins whose activity needs to be coordinated.

Some of its targets are proteins expressed in almost every cell type: these include a histone, one of a large family of proteins whose function is to condense chromosomal DNA that is not actively being transcribed or replicated; transcription factors (most notably, from the CREB family); a metabolic enzyme involved in storing energy for later use; ion channels; and other kinases (enzymes that alter the activity of other proteins by transferring phosphate groups onto them from ATP).

Although protein kinase A performs specialized functions in just about every cell type, I'm only going to talk about what it does in the brain, since that is the cell type relevant to this post. There, in addition to the stuff mentioned above, protein kinase A 1) helps regulate the synthesis of a common precursor to a variety of neurotransmitters, 2) helps form synapses by guiding the specialized proteins that allow the membranous sacs that deliver neurotransmitters from one neuron to the next toward the tip of the developing axon, and 3) with another protein kinase, regulates the ion-channel activity of the NMDA receptor, which is involved in strengthening the more frequently-used conntections between neurons. There may be more, but this is what I've been able to find.

For all that background information, the experiment I'm going to describe is actually pretty simple: like I said above, the researchers took tissue samples from five different regions of donated brains from autistic and non-autistic subjects, homogenized them (basically, ran them through a blender) and tested each sample for protein kinase A activity. The test they used is called the ELISA (for Enzyme-Linked ImmunoSorbent Assay --- see why people would rather call it Eliza?), which is a plastic plate covered with small circular wells (0.7 cm across by 1 cm deep) with, in this case, short peptides containing either serine or threonine (the two amino acids to which protein kinase A can attach a phosphate group), anchored to the bottom. (ELISA is most often used to test for the presence of antibodies --- that's how HIV testing is done --- so for that, the thing stuck to the bottom of the well would be the antigen to which whatever antibody you're testing for responds). They added their brain tissue extracts one by one to each well, along with a small amount of ATP dissolved in water (for the protein kinase to "borrow" phosphate groups from), then waited an hour and a half before emptying out the wells (the substrates, which were permanently affixed to the bottoms of the wells, would stay, along with, presumably, any phosphate groups that had been attached to them during the previous 90 minutes) and introducing an antibody specifically designed to bond with the phosphorylated form of the substrate peptide. Next, they washed the wells out thoroughly (to weed out everything that was not chemically bonded to the fixed substrates) and added a second antibody, chosen for its ability to bind to the first antibody, and which was also attached to an enzyme known for producing dramatic color changes as a side effect of its interaction with certain organic molecules. (A solution containing the molecule in question was also added, so that the wells in which the greatest proportion of the well-bottom peptides had been phosphorylated, and thus had the whole antibody rigmarole sticking off of them, would have the deepest color. There is even a way to measure color --- a device that can measure the degree to which something absorbs light at a given wavelength --- so that you don't have to rely on just your eyes to tell you whether this well or that one is a darker shade of yellow).

They used a somewhat similar technique, called Western blotting, to compare the amount of active protein kinase A between groups for each brain region. They injected their tissue samples from each of the different brain regions into a polyacrylamide gel, and ran an electric current through the gel to get the proteins to move through it. Since the gel resists having things move through it, different size proteins will travel through it at different rates. After a while, most of the proteins will separate themselves into bands along the gel, by size. Once this happened, the researchers transferred the proteins to a nitrocellulose membrane, and added antibodies specific to the catalytic (active) subunit of protein kinase A. Just like with the ELISA, there was also a secondary antibody coupled to a color-producing enzyme.

One thing that's a bit unusual in this study is that the researchers divided their brains from autistic donors into two groups, based on the developmental history of the donors. They had a "regressive autism" group, whose members started out developing typically but then lost some of the skills they'd acquired: speech was the most common skill that was lost, but some of the donors in this category also lost social skills and interest in social interaction. There was also a "non-regressive autism" group, whose members were delayed in language and social development from birth.

Subtyping autism is an increasingly popular thing for researchers to do, since "autism" is such a broad, flexible category that encompasses people with a very wide range of developmental and medical histories. It makes sense that researchers would want to subdivide this large, diverse group further to make sure they're comparing apples to apples when they look at different studies of "the autistic brain" or "the autistic immune system" or whatever.

The thing that's strange about subtyping in this study is that the number of brains being looked at is already so small. Each big group (autism, both regressive and not, and controls) had samples from ten people in it, and the researchers couldn't always get a sample from every point of interest on every brain, so sometimes the number of samples in a given category (brain region + donor neurotype) was less than ten; the smallest n for any category was 7. But that means that, with subtyping, the biggest n possible for either autism subgroup is 5, which looks more like a case study than a comparison across populations. But then, histological studies of donated brains always have to deal with smaller sample sizes, since there isn't exactly a superabundance of donated brains, and I guess if you have big differences among your subjects, you might as well sort them into subcategories, even if your subcategories are tiny.

At some point in this post I should probably mention the results of this study I've gone to such lengths to describe. The authors only found differences in protein kinase A activity in one region --- the frontal cortex --- and this difference was largest between one subgroup of the autistic group --- the autism-with-regression subgroup --- and both the non-regressive autism subgroup and the control group. The regressive autism subgroup had maybe a little less than half the PKA activity of the controls and the non-regressive autism subgroup (those two groups did not differ). Taken as a whole, the autism group had about 35% less PKA activity in the frontal-lobe samples than the control group.

The results were similar for the Western blot; the only region that showed any differences in PKA expression was the frontal lobe, and again, it was only the regressive autism subgroup that differed. Tissue extracts from that group had siginificantly less PKA in them than extracts from either the control group or the non-regressive autism subgroup; the unified autism group did not differ from the control group.

The researchers also looked for a correlation between their measure of PKA activity and various possible confounding factors, like how long each donor had been dead, the age of the donors when they died, whether they had any history of seizures, and what medications they were taking; they didn't find any relationship between any of these things and either outcome variable. Their measure of PKA expression also involved measuring how much of another protein was present in each tissue extract, both because that protein is about the same size as PKA, and thus cannot be separated from it using electrophoresis, and also to have a protein whose expression is not expected to vary across groups with which to compare relative amounts of the protein that is expected to vary.

Here is a picture of the Western blot showing both PKA (top row) and the other protein, a structural protein called beta-actin (bottom row), from all tissue samples:(Figure 2A, in Ji et al., 2011 - samples from autistic donors are on the left, and subdivided into non-regressive and regressive subtypes. Controls are on the right. You can see that, in the bottom row, the blobs are all approximately the same size, indicating expression of beta-actin is more or less the same across groups. You can also see that the blobs in the top row are a lot thinner - one space has nothing at all in it - in the regressive autism group than they are in either the non-regressive autism group or the control group. It looks like PKA expression is a bit more variable within the control group than beta-actin is, though.)

So, for a couple of reasons --- the extreme smallness of sample size, and also the degree of variation in PKA expression within the control group --- I am a bit skeptical as to whether this finding will hold up. It definitely needs to be tested a few more times, with bigger donor pools.

Leaving that aside, though --- what are the implications of this finding, should it be substantiated? The study authors refer to earlier literature that describes a role for cAMP signaling pathways in both brain development (obviously germane to a study about developmental disability) and long-term memory formation and learning (relevant to the question of how people can lose skills they once had). But it's not clear yet exactly what that role is; if you search for "protein kinase a brain" on BioNOT (a database of negative experimental results), you find an article claiming to find no difference in PKA activity between tissue samples taken from donors with Alzheimer's disease and those taken from healthy donors. So that complicates things a bit, as Alzheimer's is, even more than regressive autism, characterized by a loss of learned skills and memories.

Sources:
Ji, L., Chauhan, V., Flory, M., & Chauhan, A. (2011). Brain Region–Specific Decrease in the Activity and Expression of Protein Kinase A in the Frontal Cortex of Regressive Autism PLoS ONE, 6 (8) DOI: 10.1371/journal.pone.0023751

*What does it mean to activate an enzyme? Well, an enzyme is a kind of protein, and like all proteins, it has a range of three-dimensional configurations** it can assume, and only some of these possible shapes leave the binding site for the molecule the enzyme acts upon freely accessible. So when an enzyme is in one of those arrangements, and molecules of its particular substrate can just drift along and come into contact with the binding site(s), that's when the enzyme can be considered active. Binding of a phosphate group or some other small molecule at a different binding site will usually trigger a shape change; that is how enzymes can be activated or deactivated by other enzymes.

**I have this idea that proteins are called proteins just because of this shape-changing ability they have, in which they resemble the mythical Proteus.

More Autistic Strengths: Symmetry-Spotting

EXECUTIVE SUMMARY: A recent study has added to the list of cognitive strengths peculiar to autism: in this study, a group of autistic teens/young adults and a group of age-, IQ-, sex- and eyesight-matched control subjects were shown a series of paired images, all of them different arrangements of lots and lots of tiny black-and-white dots, and determine which of the two images has some of the dots arranged in a symmetric pattern. Consistently, the autistic young people were able to pick out the symmetrical images at lower signal-to-noise ratios (i.e., with smaller proportions of the dots possessing mirror images) than their non-autistic peers.
______________________________________________________

Michelle Dawson and Laurent Mottron have done lots of research on perception and cognition in autism --- particularly visual processing. (Morton Ann Gernsbacher is another frequent collaborator, but she didn't participate in the research I'm about to describe).

Their research has identified several cognitive strengths* specific to autism: enhanced sensitivity to pitch; enhanced sensitivity to, and recall of, details (without any corresponding loss of ability to see the big picture); ability to switch between different strategies (big-picture vs. small details) as needed; . Autistic people also do a lot better on one particular IQ test, Raven's Progressive Matrices, than you would predict based on their scores on other IQ tests (e.g., various Wechsler tests).

A new skill has just been added to this constellation: the ability to quickly determine whether a complex pattern is symmetrical or not.

In a study published this past spring in PLoS ONE, a group of Canadian researchers --- Mottron and Dawson, along with three others: Audrey Perreault, Rick Gurnsey and Armando Bertone --- had participants look at very complicated, visually "busy" patterns of small dots arranged on a video screen and determine, in the very short time the pattern remained onscreen (250 milliseconds), whether it was symmetrical or not. (They were shown two different patterns, one symmetrical and one not, and they had to identify the symmetrical one.)

Here's an example of the kind of image they would have to categorize:
(If it looks obvious to you, remember they only got a fraction of a second to look at it!)

The images were all just black and white, except for the one colored dot in the center, where the participants were told to focus their attention. In the above image, which is 100% symmetrical, each dot has a twin, the same size and color, placed so that they would lie one on top of the other if you printed the image out and folded it along its axis of symmetry. In that image, you can see that the vertical axis is the axis of symmetry; some images are symmetrical along the horizontal axis, and others are symmetrical along an oblique axis, the line y = x in a Cartesian coordinate plane with the colored dot at the origin.

These shapes are symmetrical about the horizontal (x) axis:
These shapes are symmetrical about the vertical (y) axis:

The pink and green curves are symmetrical to each other about the line y = x (blue)

Some of the images were also only partially made up of symmetrically-paired dots; the study participants were supposed to identify which of the two images shown to them had any degree of symmetry at all. (It was always just one; I guess you could design an experiment where both of the images had some degree of symmetry and the participants had to determine in which the degree of symmetry was greater, but that would be harder than just picking out which one had any degree of symmetry at all.)

The two groups whose performance was compared in this study were a group of 14** autistic young people (ages 14 to 35) and 15 typically-developing young men matched with the autistic subjects for age, IQ and visual acuity.

The criterion used to compare the two groups was "symmetry detection threshold", or the proportion of dots in a symmetrical design that had to have mirror images before a given person could identify the symmetrical design 75% of the time. Average detection thresholds were compared across groups, and also across what type of symmetry the image displayed. Both groups did best at spotting symmetry along a vertical axis, and both groups did the worst at spotting it across the line y = x.

But for all of these conditions, the autistic people had lower detection thresholds --- they correctly found symmetry more often in patterns that had less of it, relative to background noise --- than their non-autistic peers.

The study authors see this as indicative of our (autistic people's) ability to look at things more than one way simultaneously. (Another recent study, not referenced in this one, also found something suggestive of that: autistic people were better able to reproduce "ambiguous figures," or line drawings that look like they could be one of two things, depending on how you look at them). They also see their results as incompatible with the "weak central coherence" theory of autism, which explains our relatively keen collective eye for detail as a deficit in big-picture thinking. But this symmetry-spotting task requires both processes at once --- local-level, small detail perception for checking individual dot pairs to see if they really are exactly symmetrical, and also larger-scale, "gestalt" perception of whole shapes created by all the dots together.

Perreault, A., Gurnsey, R., Dawson, M., Mottron, L., & Bertone, A. (2011). Increased Sensitivity to Mirror Symmetry in Autism PLoS ONE, 6 (4) DOI: 10.1371/journal.pone.0019519

*Other, earlier research has also identified autistic strengths: as early as 1983, Amitta Shah and Uta Frith discovered that autistic children did especially well at disembedding figures; those two researchers were also the ones who identified the other really well-known "islet of ability", in the Block Designsection on various IQ tests.

**There were originally 17 people in that group, but three of them couldn't do the experimental task, so they did not contribute any data.