The Minnesota Department of Health has confirmed what Minneapolis school officials have long suspected: that there's a "significantly higher" proportion of Somalis than others in the city's preschool programs for autism.

But it still doesn't know why.

In a report released Tuesday, the Health Department said that the findings raise more questions than answers and that they are "not proof that Somali children have more autism than other children."

The investigators said they were not able to determine whether the numbers are merely a statistical blip or reflect a deeper problem.

"We recognize this study is only the first step on a very long journey," said Dr. Sanne Magnan, the commissioner of health, in a letter accompanying the report. She said the study showed that the concerns raised by parents "were important and legitimate," and that much more research will be needed to find the answers.

The Health Department launched its investigation last fall, after Minneapolis school officials and parents raised concerns about the disproportionate number of Somali children in the city's early-childhood autism program. Those reports made headlines around the country, in part because such a cluster could help scientists shed light on the causes of autism.

In this case, health investigators said they could not compare autism rates among Somali and non-Somali children, because no precise medical data exist. So they focused instead on the school data used to identify children for special education.

The investigation found that, among 3- and 4-year-olds in Minneapolis, Somali children were two to seven times likelier than others to be in special classes for autism, a brain disorder linked to behavior problems.

The report estimated that 0.9 to 1.5 percent of Somali children in that age group were in autism-related classes, compared with 0.2 to 0.7 percent of non-Somalis, between 2005 and 2008. The gap between the two groups narrowed, however, over the three years studied.

The report points out that the classroom numbers may not reflect a true picture of all children with autism. They include only children who are identified early and seek help from the Minneapolis schools. That could miss a lot of children, making true comparisons difficult, said Judy Punyko, an epidemiologist who headed the study.

The report found that two groups, Asians and Native Americans, had "strikingly low" participation rates in the preschool autism classes. But the reasons were unknown as well.

"Is it possible that these children have actually lower prevalence [of autism]? We don't know," said Punyko. She said studies suggest that cultural or other differences can affect how families view behavioral conditions such as autism, especially in young children.

State may study further

Analyzing the numbers turned out to be trickier than expected. The investigators noted that Minnesota does not have a formal system to track autism and that children don't need a medical diagnosis to enroll in autism special education programs. In fact, medical experts told the Health Department that a fourth or more of the children placed in autism programs at school turned out to have some other disorder.

Magnan, the health commissioner, said it would take a much broader study to find out whether autism rates really differ among ethnic groups. That's one of the steps under consideration, she said, but there are no definite plans for such a study.

Idil Abdull, a Somali-born woman who lives with her autistic son in Burnsville, said she's grateful for the report, even if it left many questions unanswered.

"Autism in itself is a mystery," said Abdull, who called for an investigation last year. "I wasn't counting on the Health Department of Minnesota to solve that mystery. I just wanted them to acknowledge that there's a problem."

Maura Lerner • 612-673-7384