Folate, a B vitamin, is a tremendously important part of our diet. We need it to make and repair DNA, and to synthesize chemical messengers and nerve sheaths in the brain. When folate doesn’t pass from the blood into the brain as it’s supposed to, it can lead to extremely low muscle tone and serious neurological problems.

Could low folate also contribute to autism?

A handful of scientists say it’s possible, but solid evidence has been hard to come by. A new study, published 10 January in Molecular Psychiatry, claims that low folate levels are a problem in children with autism and offers a tenuous explanation for the low levels.

Many children with the disorder carry immune molecules in their blood that may block the passage of folate from the blood into the brain, the researchers found. The study lacks a control group of healthy children, however, putting many of its claims on shaky ground.

The biggest obstacle in folate studies is that researchers must analyze cerebrospinal fluid (CSF), which can be gleaned only from a spinal tap. Because this is an invasive procedure, little is known about the range of CSF folate levels in healthy people.

A couple of studies in the past few years have reported lower CSF folate levels in children with autism than in controls. At the Society for Neuroscience annual meeting in November, researchers reported that 5 of 67 children with autism had CSF folate levels below the normal cutoff of 40 nanomoles per liter.

Folate levels in the blood of these children, however, are normal, suggesting that the problem may lie in folate’s transport to the brain.

In the new study, researchers screened blood samples from 93 children with autism for immune molecules called folate receptor autoantibodies, or FRAs. Of these children, 56 carry a ‘blocking’ FRA, which interferes with the folate receptor, and 41 have a ‘binding’ FRA that binds to the receptor but doesn’t affect its function. Some of the children carry both kinds of FRAs.

Of the children who tested positive for at least one of these autoantibodies, 16 received spinal taps to collect spinal fluid. All of them had folate levels lower than the reported normal average of about 80 nanomoles per liter, but none showed levels below 40 nanomoles per liter.

The researchers did find an intriguing association: The lower the CSF folate level, the higher the concentration of blocking FRA in the blood.

To supplement a possible folate deficiency, they gave 44 of 70 FRA-positive children leucovorin, a pill that contains a form of folate called folinic acid, twice a day for a month or longer. Afterward, their parents reported improvements in verbal communication, language, attention and stereotypical behaviors, compared with parents of children who did not receive the treatment.

Because the study lacks a control group, it’s hard to know what to make of any of these numbers. Notably, none of the children who had their CSF tested had folate levels below what’s considered normal, which calls into question the very idea that folate deficiency contributes to the disorder. 

The autoantibody levels in the children in this study do seem high: Previous work by some of the same researchers — who hold a patent on the test — found these autoantibodies in just 7 percent of healthy women in Spain and in 13 percent of women in Ireland.

Still, I think it’s premature to use these data as the basis for a new treatment for autism. Because their study did not have a placebo group, there is no hard proof that leucovorin works.

Until we have better data on autoantibody levels in healthy people, the study presents interesting, but at best speculative, ideas about the role of folate in autism.