As evidence of widespread vitamin D deficiency grows, some scientists are wondering whether the sunshine vitamin—once only considered important in bone health—may actually play a role in one of neurology's most vexing conditions: autism.

The idea, although not yet tested or widely held, comes out of preliminary studies in Sweden and Minnesota. Last summer, Swedish researchers published a study in Developmental Medicine and Child Neurology that found the prevalence of autism and related disorders was three to four times higher among Somali immigrants than non-Somalis in Stockholm. The study reviewed the records of 2,437 children, born between 1988 and 1998 in Stockholm, in response to parents and teachers who had raised concerns about whether children with a Somali background were overrepresented in the total group of children with autism.

In Sweden, the 15,000-strong Somali community calls autism "the Swedish disease," says Elisabeth Fernell, a researcher at the Karolinska Institute in Stockholm and a co-author of the study.

In Minnesota, where there are an estimated 60,000 Somali immigrants, the situation was quite similar: There, health officials noted reports of autism among Somali refugees, who began arriving in 1993, comparable to those found in Sweden. Within several years of arrival, dozens of the Somali families whose children were born in the U.S. found themselves grappling with autism, says Huda Farah, a Somali-born molecular biologist who works on refugee resettlement issues with Minnesota health officials. The number of Somali children in the city's autism programs jumped from zero in 1999 to 43 in 2007, says Ann Fox, director of special education programs for Minneapolis schools. The number of Somali-speaking children in the Minneapolis school district increased from 1,773 to 2,029 during the same period.

Few, if any, Somalis had ever seen anything like it. "It has shocked the community," Farah says. "We never saw such a disease in Somalia. We do not even have a word for it."

What seemed to link the two regions was the fact that Somalis were getting less sun than in their native country—and therefore less vitamin D. The vitamin is made by the skin during sun exposure, or ingested in a small number of foods. At northern latitudes in the summertime, light-skinned people produce about 1,000 international units (IUs) of vitamin D per minute, but those with darker skin synthesize it more slowly, says Adit Ginde, an assistant professor at the University of Colorado Denver School of Medicine. Ginde recommends between 1,000 to 2,000 IUs per day, calling current recommendations of 200 IUs per day outmoded.

It’s hard to definitively assess the extent to which Somali immigrant families in Sweden and Minnesota are experiencing increased rates of autism. Somalia doesn't have great records of the condition, says Rebecca Berkowitz, who works for a United Nations–affiliated NGO called Global Education Motivators. "Children in Somalia may not even be getting diagnosed with autism due to the overall lack of awareness of the disorder," Berkowitz says, in a nod to the fact that there is no Somalian word for it. And Swedish scientists have reported autism rates overall have risen since they began studying the epidemiology of the disorder in the mid-1980s—just as U.S. Centers for Disease Control officials have noted an increase.

Still, proponents of the vitamin D–autism link say there is biological plausibility to their theory. They cite a 2007 review by Allan Kalueff, a researcher now at Tulane University, in Current Opinion in Clinical Nutrition and Metabolic Care. That review—based on more than 20 studies of animals and humans—concluded that vitamin D during gestation and early infancy was essential for "normal brain functioning."

At the same time, the theory needs a lot of data to back it before others will give it much credence, given how many other potential reasons there are for a climb in autism rates. Even Kalueff says he isn’t sure how vitamin D could be related to autism, even if it is an important player in the brain: "Discussions around autism specifically may be a right step or a wrong step, but they should not distract us from a much bigger picture."

Catherine Lord, the director of the University of Michigan at Ann Arbor's Autism and Communication Disorders Center, says she finds the Swedish study intriguing. "But it is going to be really important to replicate these findings," says Lord, who has studied the disorder for 40 years and has been instrumental in developing autism diagnostic instruments used in practice and research worldwide. “We are talking about a small group of children with a lot of social factors, including that these kids are very conspicuously different from your average Swedish child, and being assessed by people who are from very different culture." There is also the issue of consanguinity, she says, as many Somalis marry cousins. "This doesn't mean the study is wrong," she says. "But we need methodical testing."

So Fernell and her colleagues are now measuring vitamin D blood levels in mothers and children with autism of both Somali and Swedish origin and comparing them with a control group of mothers and healthy children. She will not say how many subjects the study includes, describe any preliminary results nor say when it will be complete. Farah says Minneapolis researchers are now preparing to study the vitamin D levels of pregnant Somalis, other ethnic groups and Minnesotans of European stock. (That data is particularly hard to come by because Vitamin D levels are not typically screened in pregnancy in the U.S., says Stacy Brooks, a spokeswoman for the American College of Obstetricians and Gynecologists.)

The other potential reasons for a climb in autism rates: There is increased attention to the condition in the U.S., and Somalis are more likely to see a doctor after moving here. Also, genes, studies have found, may play a role; a number of papers, including a 1989 study of five Nordic countries and a 1995 British study, found that the concordance rate among identical twins was as high as 90 percent. (Then there is the much-ballyhooed but ultimately disproved link to vaccines.)

Somali refugees, in particular, faced multiple stressors as they adjusted to their new lives in Sweden and Minnesota: They had fled civil war, lost a supportive tribal culture, and replaced a diet of fruit, fresh meat and grains with processed food. Perhaps, most importantly, they had traded family compounds and regular exposure to the equatorial sun for cloistered high-rise apartments.

But some of those potential cultural reasons could also point to vitamin D. Surrounded by strangers, the predominantly Muslim women covered themselves almost continuously when outdoors, says Gregory A. Plotnikoff, medical director of the Penny George Institute for Health and Healing in Minneapolis. Plotnikoff, an internist, speaks Somali and has many Somali patients. That meant less exposure to the sun for pregnant women, who would have worn less modest dress in private areas of their own family compounds.

And there is other evidence for a vitamin D link: Last November, Cornell University researchers published a study in Archives of Pediatrics & Adolescent Medicine showing that children in rainy (and therefore more overcast) counties of Oregon, Washington and California were two times more likely to be diagnosed with autism than their counterparts in drier parts of the state. "Our research is sufficiently suggestive of an environmental trigger for autism associated with precipitation, of which vitamin D deficiency is one possibility," says study co-author Michael Waldman, a professor of management and economics at Cornell's Johnson Graduate School of Management. "Further research focused on vitamin D deficiency is clearly warranted." His research on environmental links to autism are ongoing; he plans to publish in the coming months but will not disclose any of his studies until they are accepted by a journal.

Gene Stubbs, an associate professor emeritus of psychiatry and pediatrics at Oregon Health & Science University, says the preliminary research is already intriguing. "We don't have proof, but I am certainly leaning in the direction that this hypothesis could be correct for a proportion of kids," says Stubbs, who has been studying autism for 30 years. He is launching a pilot study of 150 pregnant women who have at least one child diagnosed with the disorder. The women will receive 5,000 IUs of vitamin D3 during gestation and 7,000 IUs during lactation. "If we find that we are able to reduce the recurrence rate of autism within families substantially enough, others will want to study this in larger groups with larger controls."