« My Favorite Things Alaska | Main | Sorry people, I can't resist »
The Minnesota Somali autism puzzle
Somali students comprise only 6 percent of the Minneapolis school system, but one-quarter of the children in the city’s early childhood autism programs. Health officials are baffled.
Here is more. Here is a follow-up story. Oddly Somali families in Sweden call autism "the Swedish disease." There hasn't been thimerosal in vaccines for some time plus that would not explain the higher incidence of autism among these groups of Somali children. It also seems that the Somali kids have especially severe cases of autism.
So what is the environmental trigger? A combined lack of sunlight and vitamin D activation is the only real hypothesis I can find.
Mother's testosterone levels, if high, can influence a male child to extreme maturational delay. (A child's maturation rate is set at six weeks before birth.) If the mother has immigrated from an equatorial region with consistent diurnal light cycles of 30% to a relatively extreme northern climate, the influence of light on the pineal gland influencing testosterone levels can dramatically skew mothers testosterone. The question is, do the Minnesota Somali autistic children's birthday's congregate in certain seasons. If so, this hypothesis becomes more likely. See http://www.neoteny.org/?cat=7, particularly http://www.neoteny.org/?p=85.
Yet even that sounds screwy to me (at least it's testable), noting that if you pursue the links you will not find mainstream science at the end of the tunnel. But the independent appearance of the phenomenon in Sweden and Minneapolis suggests it isn't just a statistical fluke. And the numerous Somali immigrants in Virginia don't seem to have the same problems.
Here are some further readings. Here is a speculation that autism rates are higher in immigrant communities more generally. Addendum: Here is more on Sweden.
Posted by Tyler Cowen on September 1, 2008 at 07:32 AM in Medicine | Permalink
Comments
Sunlight. So, if autism rates, not just identified and labeled cases, are rising in the U.S., is it the obsessive use of sunscreen on babies and small children that is triggering the increase? Maybe I should read the article links.
Posted by: liberalarts at Sep 1, 2008 8:07:22 AM
I know there's a large Somali population in Lewiston/Auburn Maine. I wonder if they've observed the same problems? (my apologies if this is mentioned in the links).
Posted by: CarterB at Sep 1, 2008 8:14:28 AM
One of the reasons slavery never had much of a hold in the northern US is because blacks died at a high rate in places like Massachusetts as compared to the South. My vote goes for vitamin D as the cause. BTW, this just goes to show that libertarians' ideas about immigration and people merely coming here to work and pay taxes and make the country "vibrant" are fantasies: those autistic kids will be a net drain on taxpayers for the rest of their lives.
Posted by: Dennis Mangan at Sep 1, 2008 8:49:02 AM
"it the obsessive use of sunscreen on babies and small children"
"My vote goes for vitamin D as the cause"
The spirit of Galileo is strong with these two.
Posted by: DJB at Sep 1, 2008 9:18:49 AM
@Libertarian ideas about immigration,
Yeah, I predict a huge influx of Somalis to the US in order to have autistic kids so they can drain our welfare coffers. In fact, Al Qaeda may be behind it.
Posted by: Andrew at Sep 1, 2008 9:23:05 AM
What's the false positive rate for autism diagnoses?
Posted by: David at Sep 1, 2008 9:31:36 AM
Here's a site with a general (non-to-low technical) discussion of the connection between Vitamin D and autism.
And for the sake of "going there," here's the final post of a series that considered the greater incidence of Vitamin D deficiency in the black population of the U.S. as likely accounting for much if not all of the observed disparity between black and white life expectancies.
Who knew it wasn't racism?
Posted by: Ironman at Sep 1, 2008 10:23:08 AM
Throwing possible causes out here:
(1) Something endemic to the population. The appropriate comparison group is not necessarily Virginia Somalis, but Somalis from the same clans that Minnesota Somalis come from.
(2) A very local environmental factor. Minneapolis Somalis are rather concentrated in a few high-rise apartment buildings. The data probably exist to due block-by-block epidemiology, maybe even floor-by-floor (they are big buildings). To go this route, the Swedish Somalis have to be suffering from an coincidence of environment.
(3) Sunlight and MN winters. If we go this route, we might expect similar epidemiology in other tropical populations that migrate to high latitudes. Seattle is further north than Minneapolis. Less cold, but also less sunny. Has similar epidemiology been reported for the Filipino and Vietnamese populations there? London is still further north, and has no reputation for sunshine, nor of immigrants from tropical ethnoi around the world.
Posted by: Robert at Sep 1, 2008 10:39:13 AM
Surely, the reason why slavery flourished in the southern US was because cotton grew there, and in particular because of an accident of history and invention: the processing of cotton was mechanized before the harvesting of it.
Posted by: at Sep 1, 2008 10:43:19 AM
In the entire population of state school children in the UK (n=7.3m) there were 43,000 children recorded as having Autism as their primary type of special educational need. This is 0.56% of the population.
The categories of ethnicity are relatively coarse. Ethnic minorities can be split into those who speaking English as a second language and those who do not. This is a proxy for being born in England. The base category is white with English as first language.
Black Africans born in England are 0.25pp more likely to be recorded as autistic, black other are 0.27pp more likely. For those who speak English as a 2nd language the figures are 0.32pp and 0.23pp less likely.
Asians are 0.10-0.24pp less likely to be recorded as autistic, the ‘effect’ is larger for those who speak English as a second language. Those of Afro-Caribbean descent are no more likely to have autism than the white population regardless of language status.
There is a month of birth bias in being recorded as autistic, those born in August compared to September are 0.07pp more likely to be recorded as autistic. The size of this bias increases monotonically from September to August (from oldest to youngest in an English school year).
Interactions of Black African with English 1st language with month of birth suggests that those born between those born in March-June are 0.1-0.4pp less likely to be autistic than those born September-December, and August & July. The figures for the interactions with “Black other” are very similar.
This suggests that fewer individuals conceived in what is laughably called the English summer, (June-September) have autism.
Posted by: nmd at Sep 1, 2008 11:48:11 AM
While it's true that the Somali population in Minneapolis/St. Paul is condensed in a small neighborhood, they do not exclusively live in one or two large apartment buildings. The population has existed there for a good amount of time (around a generation) and they have moved out to surrounding areas in Phillips and even in the suburbs of Eagan and Eden Prairie. The specific environment explanation Robert throws out is false, the families live in a wide variety of dwellings and economic situations.
While the first claim is definitely on the right idea, I'd say that the population of Somalia is a lot more homogeneous than any of us are used to and while I think it'd be good to at least establish a separation between the two populations, I'd wager that it wouldn't provide enough causation to write off the validity of a direct comparison, especially if the populations immigrated within the past 80 years (I confess I have no idea the recent history of the state's human geography).
I think the comparison between cities is apt, however populations adapted to tropical climates do not share the same genes and the specific chemical cause of autism is a mutation in chromosomes. I believe we've established that this is completely specific to the Somali population.
The vitamin D deficiency resulting from a lack of sunlight, though implausible sounding, makes some definite sense. Lack of sunshine encourages depression, why not autism?
This brings up two things. 1. It might completely validate the generic crotchety old man complaint, "Kids never leave their damn video games and go outside". Autism development rates have risen dramatically since the mid 80s.
2. Could Global Warming and the depleting ozone be seen as a cure or vaccine for autism and other vitamin-D deficiency caused disorders?
Posted by: Tyler W at Sep 1, 2008 11:52:36 AM
NMR, your stats look very helpful but I'm not following your meaning. Black Africans are more likely than whom? It might be easier if u could just list the absolute rates per population group instead of just reporting differences or if you could post a link to your source. Thanks.
Posted by: dk at Sep 1, 2008 12:08:03 PM
I'm just throwing out ideas here, but what about non-genetic or non-environmental causes? For example, who made the diagnoses? Any chance one or two people are sending lots of kids to the program? Any chance there's something cultural that leads to a diagnosis of high-functioning autism? Also, the article says Somali kids make up 6% of the school system, but how dispersed is that? Surely there are a few schools without any Somali kids at all. Any chance those schools don't get many autism specialists visiting?
I certainly think there are/could be genetic factors involved, but I would first analyze the system that led to these kids being sent to the program.
Posted by: Richard at Sep 1, 2008 1:11:47 PM
It would be a mistake to take the school figures at face value. Minnesota has open enrollment, meaning that students can cross city borders to go to school. Minneapolis happens to have excellent programs for special needs children. It's possible that parents of autistic Somali children have been seeking out this program. If parents of autistic white children in Minneapolis are more likely to afford and obtain private care, the combined effect would be to inflate the proportion of Somalis in special care classes in Minneapolis. No need to invoke biological explanations.
Posted by: Ida know at Sep 1, 2008 1:24:12 PM
St. Paul also has neighborhoods with a substantial concentration of Somalis. Minneapolis has a population of Oromo families (from Ethiopia). How do those rates compare?
Posted by: Linda at Sep 1, 2008 1:24:34 PM
I should clarify that parents of autistic Somali children FROM OUTSIDE MINNEAPOLIS may have been seeking out special care in the city.
Posted by: Ida know at Sep 1, 2008 1:25:48 PM
Dennis Mangan,
Way to confuse higher relative rates with high absolute rates.
Posted by: jb at Sep 1, 2008 2:28:54 PM
I was speaking to a mental health worker in my home town in Ireland about this recently.
She said the ultimate reason is bad diagnosis and drugs. Autism is one of the few conditions that there are no drugs available for that the medical profession trust. Most other mental illness has recommended drugs.
According to my friend most people who diagnose mental illness don't know how to do it. They therefore urge on the side of the guess that will do the least harm. That guess is almost always Autism. Those so diagnosed get therapy and the more skilled therapists further down the line find out whether they actually have Autism or something else. Diagnose someone with bipolar disorder and they get a few drugs and little else.
This neatly explains the situation here, since immigration is related to low social status and earning. And, low social status and earnings are link to mental disorders of all kinds. Whether this explanation is correct or not would need much further study.
Posted by: Current at Sep 1, 2008 3:16:34 PM
I live in Eden Prairie, a southwestern suburb of Minneapolis. We have about 5000 Somalis, and there is a lot of support for them in the community, offering legal aid, assistance with federal programs. They are friendly, but they do bring school test scores down, and having higher rates of crime (relatively, this is a safe suburb). In deference to Somali customs, we got rid of 'birthday parties' at school, and other such rituals.
But they are not all in a high rise downtown. They are growing, especially in our section 8 housing (which are townhouses).
Posted by: eric at Sep 1, 2008 3:23:19 PM
How about cousin marriage matched with a medical system looking for austism. In Somalia is probably would never have been diasnosed or the child would not have lived very long.
Posted by: superdestroyer at Sep 1, 2008 5:16:47 PM
DC is cloudier than Minneapolis, especially in winter, so the difference in sun is very small. I compared total annual solar intensity on wunderground.com and it's 259 for DC and 240 in Mpls, and season-to-season the rates are pretty similar. You could argue that Somalis in Mpls spend less time outside in the winter but I also wouldn't believe that without data.
That said, I highly recommend Vitamin D supplements for breastfed newborns. Good luck getting your newborns to swallow it though, that stuff is nasty.
Posted by: Paul N at Sep 1, 2008 5:26:46 PM
Somalis are not only black, but Somali women also tend to wear clothes that cover essentially everything except the face. That would decrease their exposure to sunlight substantially and make them more likely to get vitamin D deficiency than other black groups.
Posted by: at Sep 2, 2008 4:03:02 AM
Please stop the talking about thimerosal, even as a side reference. There has never been any scientific evidence that thimerosal had any effect on childhood development and rates of autism. Vaccines were and are safe, effective, and critical to the continued improvement in public health across the world.
Posted by: Apoo at Sep 2, 2008 11:31:31 AM
Maybe brains in industrialized societies really have evolved in ways that increase the ratio of gray to white matter. You don't get nearly as much affect stimulus in an industrialized society as you would get in a village. But then why Somalia only and not all pre-industrial cultures?
Posted by: Michael F. Martin at Sep 2, 2008 4:05:50 PM
If memory serves correctly, a sizeable percent of the population in Minneapolis is of Swedish descent. If this is correct, is it possible that there is some other factor that is unique to Swedish culture (food, for example,) that is foreign to Somalis and is causing the higher rates of Autism? Have there been any studies regarding autism rates in Somali children who live in Northern Russia, or other areas that are geographically similar but culturally distinct from Scandinavia?
Posted by: JP at Sep 2, 2008 11:27:33 PM
