Effects of pyridoxine and magnesium on autistic symptoms--initial observations.

Doctors gave 44 children with autism big doses of vitamin B6 and magnesium. First everyone got the vitamins in an open trial. Then the 15 kids who looked better entered a double-blind crossover where capsules and placebos were swapped without telling families.

The team watched for any change in autistic symptoms during each phase.

Fifteen of the 44 children looked a bit better during the open phase. When those 15 switched to placebo, 13 lost the gains and regained them only on the real vitamins again.

That pattern says the vitamins, not luck, drove the small improvements.

Tromans et al. (2018) surveyed 529 autism RCTs and flagged this 1981 trial as one of the tiny early studies that still dominate the field. The survey warns that most samples are too small to trust.

Heavey et al. (2000) ran a similar crossover with methylphenidate and also saw help for only about half the kids, plus side effects. Both pills show the same lesson: a few respond, many do not.

Lu et al. (2025) meta-analysis of gut-microbiota supplements found clearer, larger benefits for GI pain in autism. Their modern evidence now supersedes the old B6/magnesium idea; probiotics have better data and safer profiles.

If a parent asks about vitamins, you can say a 1981 study saw mild gains in about one-third of children, but later work has moved on. Today you have stronger options such as probiotic protocols or standard medications for hyperactivity. Document symptom baselines before any supplement and review at four weeks so you can stop if nothing changes.