There are a number of limitations to all of the research studies published to date.
The study by Xia (2011) is a case report of a single child using a pre-post-test design. The parents administered Vitamin B6 and magnesium for four months and then added dimethylglycine supplements for a further month, making it difficult to know if any changes were due to the DMG or the other supplements or the combination of all of them.
The study by Kern et al (2001) is a randomised, double-blind, placebo-controlled trial of 39 children lasting four weeks. The children who were randomly assigned to the dimethylglycine group were more severely affected than the children in the placebo group.
The study by Bolman and Richard (1999) was a randomised, double-blind, placebo-controlled, crossover trial lasting three and a half months. The authors noted that the major methodological weaknesses of the study was the low dosage of dimethylglycine (half the amount recommended by researchers such as Rimland and Edelson) and the small number of participants (only ten children, two of whom dropped out).
For a comprehensive list of potential flaws in research studies, please see ‘Why some autism research studies are flawed’