• Thu, Nov 20 2008

About the “Cluster” of Autism Among Somali Children in Minneapolis

Back in July, it was reported that the rate of autism in Somali children in Minnapolis was notably high. According to the Minnesota Department of Education:

in the Minneapolis’ early childhood and kindergarten programs, more than 12 percent of the students with autism reported speaking Somali at home. According to Minneapolis school officials, more than 17 percent of the children in the district’s early childhood special education autism program are Somali speaking.

Almost 6 percent of the district’s total enrollment is made up of Somali-speaking students, and about 6 percent of the children in the district’s overall early childhood and kindergarten special education programs are Somali.

Speculation about what could be causing this “cluster” of autism cases in so specific a population immediately started up, especially among proponents of environmental causes of autism such as journalist David Kirby. Mike Stanton at Action for Autism gives an overview of all this, and notes how Kirby and others sought to connect the Somali autism rate—or, more precisely, the rate of autism among children born in the Minneapolis area to immigrant parents from Somalia—to vaccinations, and also to a theory that a Vitamin D deficiency can be linked to autism. Dr. Steve Novella at the Neurologica blog writes specifically about the notion of a “cluster” of autism cases being found:

apparent clusters of diseases are reported all the time. Most of the time the clusters are not real, meaning they are just statistical flukes. So the first question to answer with any apparent cluster is – does this represent a real epidemiological phenomenon.


One problem with the cluster hypothesis is that other immigrant Somali communities have not experienced increased autism rates. If there is an environmental trigger causing the two identified clusters, why are there not clusters in these other communities?

If it turns out to be true that autism rates have significantly increased in some Somali immigrant communities, above what is seen in Somalia or in non-Somali in the same communities, then we can conclude that something is going on and a potential trigger should be sought.

It also has to be noted that autism is really a collection of diseases, not a specific disease. So we may be seeing a new entity that has clinical overlap in features and symptoms with recognized forms of autism.

Dr. Novella writes that more investigation is indeed needed about the Somali “cluster” and if it is real, and what factors might be playing a role, whether genetic or environmental. He references an article from the Simons Foundation Autism Research Initiative that cites Judy Punyko, an epidemiologist at the Minnesota Department of Health. I August, Punyko formed a study group of 12 experts (including epidemiologists, physicians, school administrators and special education teachers) to study the rate of autism in Somali children in Minneapolis with “age-matched controls.”

Even then, she adds, educational data may be incomplete or inaccurate. The 13 special education categories reported to the state and federal government are intended to help provide a child with the best available educational services. If a child has two conditions, such as autism and developmental delay, they can only be assigned to one primary category; Minnesota does not require a medical diagnosis of autism to be included in the category. Finally, some schools, especially those in poorer districts, often overlook mild forms of autism.

Population data used for the analysis also comes from the 2000 census, which may be vastly different than current numbers. “The population of Somalis is a fluid number in Minnesota,” says Punyko.

Mike Stanton also notes that “in Minnesota there is no reliable epidemiological data for autism”—because, as he points out, the “administrative data for children in receipt of autism services” is based on “teacher assessment,” with a diagnosis from a “trained clinician” not required.

I know this from experience: My son was first evaluated for “delays” in Minnesota, by a Child Study Team from the St. Paul Public Schools. That was in the spring of 1999 (he wasn’t 2 years old yet) and he immediately started to receive services (speech, OT, and special ed, only a few hours each week). He wasn’t actually diagnosed with autism until July of 1999 (and the only change in the services was that he qualified for more hours of special education, which we were urged to have him receive in a school setting, rather than at home—that classroom was not appropriate for Charlie—but this is another story, and a whole ‘nother post). Also (and this is completely anecdotal), a number of Somali families lived in a high-rise apartment that was right next to the building where Charlie’s pediatrician had his office, and there were many Somali mothers with strollers and young children waiting beside us among the little tables and fish tanks in the waiting room). (And, also really anecdotally, no Somali children in Charlie’s special ed program in St. Paul, back in the summer of 1999.)

One question that (following on today’s earlier DSM-V post) needs to be addressed is how cultural factors might be at work here. And Dr. Novella writes that

Somali parents certainly believe they are experiencing something new, and some pediatric neurologists in these areas have had their suspicions also. But this is not enough to form a scientific conclusion – only to justify further research.

The true autism rate in Somalia needs to be investigated also. We should not assume that because the culture does not recognize autism it does not exist.

Besides keeping in mind the particular conditions that a child is said to be “autistic” in Minnesota—again, a child can receive autism services through assessment by a teacher (as my son did) and without receiving an official diagnosis from a “trained clinician”—we need to get a better sense of how autism is understood in Somalia, and what the numbers there are, and how these are determined.

What We're Reading:
Share This Post:
  • http://hardwonwisdom.blogspot.com Rose

    I think Bettleheim had one thing right. I do believe the children had something in common with the pow’s in Nazi concentration camps. For both, life was overwhelming…and they had no control over what happened to them.

    One was caused by cruel, inhumane treatment;the other by inborn hypersensitivities to daily life to a much greater degree than typical. The physiological manifestations are probably similar.

    I’m going to give an example, so you don’t think I’m too off my rocker. When Ben was about 3 years old, he refused to sit for “reading time” at school. We had his hearing tested. At the range of a high-pitched womans voice, he heard 10x more acutely than normal. The teacher’s voice must have sounded like screaming to him! It was only at this range that it was acute, and I”ve no idea why…To have sit near her would have been torturous to him.

    The physiological manifestations of severe stress
    are probably similar regardless of their source. I would think anyone under severe stress because of biological or environmental differences could show “autistic tendencies”. Then again, I could be totally wrong.

    Once, I saw a dog get run over by a car right in front of me. A person walking on the sidewalk also saw it and began hand-flapping. Were they autistic, or overwhelmed? That simple act got me to thinking about it.

  • RAJ

    Severe emotional deprivation leading to ‘autism’ can also be seen in orphans who meet diagnostic criteria using universally accepted Gold Standard diagnostic tools such as ADOS and AUTI-R all based on DSM-IV an ICD-10 criteria. ‘Gold Sandard’ is a term constantly invoked by those who have designed the various diagnostic tools.


    The real question is are these children actually ‘autistic’ or is the inability of ‘experts’ to define exactly what autism is that may have led to overdiagnosis and misdiagnosis on a global scale that is the real explanation for the ‘autism epidemic’.

    Authority doesn’t use the controversial term ‘misdiagnosis’ but rather invokes the more politically correct ‘Diagnostic Substitution’.

    Wer are headed to 1 1/10 or a 1/20 prevelance of ‘autism’. DSM-V which is considering including ADHD as being ‘on the spectrum’ is the best evidence that no one can define ‘autism.

  • alyric

    Interesting link there RAJ. Noting the quarter that lost the diagnosis was probably what you’d expect. Now, will Hannah Poling also lose her diagnosis? After all, what does a non specific mitochondrial disorder caused encephalopathy have to do with autism?

  • Joseph

    I understand Hannah Poling is right in the threshold of the ADOS.

    I realize we’ve discussed this before, but RAJ’s critiques of the autism construct don’t make sense to me.

    By his reasoning, mental retardation must be a bogus construct too. It’s not measured biologically, it doesn’t matter what causes it to diagnose it, and the IQ threshold (70) is essentially arbitrary.

  • http://hardwonwisdom.blogspot.com Rose

    RAJ…1 in 10 or 1 in 20. I see it, and not that far in the future, either.

  • Ed

    Clustering means one thing. It is not all genetics. This puts a gaping hole in the theory that the rise in autism is due to the change in diagnostics.

  • http://www.autismvox.com Kristina Chew, PhD


    ….apparent clusters of diseases are reported all the time. Most of the time the clusters are not real, meaning they are just statistical flukes. So the first question to answer with any apparent cluster is – does this represent a real epidemiological phenomenon.

    from the Neurologica blog.

  • Joseph

    Clustering means one thing. It is not all genetics. This puts a gaping hole in the theory that the rise in autism is due to the change in diagnostics.

    That there are environmental causes for conditions (and there are some possible causes for autism, like congenital Rubella) does not mean that the prevalence of the condition is not stable. Again, mental retardation is a good analogy here.

  • Ed

    There are clusters and then there are clusters. The Somalis appear to have twice the rate of the rest of the population. That is quite a rate.

    As I said before, I witnessed the change in autism rate. What I saw was a change. Classical autism went from unknown to expected in the school district. Yes there were odd kids that rightly would be classified as Aspergers, but the classic autistic behavior was unknown.

    The idea that there is no epidemic comes from the question of why the rate jumped if it “can’t be vaccines”. “It can’t be vaccines.” is not a reason for no epidemic.

  • http://www.autismvox.com Kristina Chew, PhD

    I don’t think “It can’t be vaccines” is the only statement made to dispute claims of an “epidemic of autism.”

  • Ed


    Reread chapter 6 of Paul Offit’s book. There were three experiments where the rate of autism went up when thimerosal was removed from the vaccines. The question, “If not mercury then what?” came up. It was answered without proof as being because of the diagnostics. It could just as easily been answered as being because the vaccination rate went up during the experiments. That, too, would have been speculation. Both sides have allowed speculation to morph into dogmatic faith. This does the autism community no good.

    I see autism as a medical issue. There are a number of medical aspects of autism that will never be explored because they step dangerously close to thimerosal and/or vaccines. These aspects that could be exploited to make the lives of our children better will never be exploited. I find that frustrating.

  • http://www.camelmilkusa.com Dr. Millie Hinkle

    There is a proposal that will be introduced by Dr. Millie Hinkle in April to the Dairy board and the FDA to allow camel milk to be sold in the US so that studies can be done with autistic children. We need double blind medically supervised studies done to see if the results will be as promising as those done in Isreal with camel milk and autism.