Year	Author	Study Subjects	Controls	Type of Intervention	Study Design	Outcome Measures	Results
1978	Rimland, Callaway and Dreyfus	16 autistic children	Yes	75mg - 3000mg B6, mostly 300mg - 500mg	Double blind placebo crossover	Behaviour measured on TSCL ratings by parents and teachers	Eleven of 15 improved (p < 0.05)
1981	Lelord et al - study 1	44 children with autistic symptoms	No	600mg - 1,125mg /day B6 and 400mg - 500mg/day magnesium	Open trial to identify responders	Behaviour on Brettonneau II (18 items), HVA in urine	21 of the 44 children improved and these were put forward to the second study. The other 23 children were the controls.
1981	Lelord et al - study 2	21 children selected from above 44	23 children from above 44 not selected.	600mg - 1,125mg /day B6 and 400mg - 500mg/day magnesium, for 14 days	Double blind placebo crossover comparing responders and non-responders.	Behaviour on Brettonneau II (18 items), HVA in urine	15 of 44 improved noticeably, in 14 of 15 the improvement disappeared by 3 weeks of the cessation of the treatment (p =< 0.01). HVA levels (n=37) also improved (p < 0.01). 14% improved considerably, 33% improved, 41% no improvement, 11% worsened.
1984	Jonas et al	8 adults with autism	Yes	Received B6 and magnesium for 42 day periods 42 days apart.	Double blind crossover trial	Behaviour was measured on a rating scale. Levels of HVA, dopac, dopamine and GABA were measured in the urine.	Significant clinical improvement on scores for affective responses and communication.
1988	Martineau et al	11 autistic children aged 4 - 8	Yes	30mg/kg/day B6, 10mg/kg/day magnesium. Eight weeks of treatment followed by a no treatment period.	Open non-blind controlled study	Behaviour Summarized Evaluation (20 items), urinary HVA, AEP.	The B6 group showed a significant behavioural improvement (p < 0.02) , normalisation of evoked potential and a drop in dopamine levels. The behaviours returned to the baseline when the treatment was discontinued. Significant decrease in uDA (p < 0.02).
1993	Tolbert	15 students in a residential school with ASD and mental retardation aged 6 - 18.	Yes	200mg/70kg B6 and 100mg/70kg magnesium	Double blind placebo controlled assymetric crossover after 10 week baseline data. Ten week study.	Objective (Ritvo Freeman Real Life Rating Scale).	No effect, lower dose used to avoid toxicity. There was a slight reduction in symptoms in all students including controls and subjects.
1995	Pfeiffer et al	Twelve published studies on autism	No	Comparing and analysing the different studies that have taken place on B6 and magnesium	N/A	Methodology review and summary of outcomes.	It is considered that much of the research done in this area has errors and inconsistencies and although the indications are that B6 and magnesium will often ease the symptoms of autism, some larger scale studies need to be done.
1997	Findling et al	10 autistic children, outpatients	Yes	30mg/kg B6 and 10mg / kg magnesium, no washout period, no test of compliance	Double blind placebo crossover, 4 week trials	Children's Psychiatric Rating Scale, Clinical Global Impression Scale, NIMH Global Obsessive Compulsive Scale, Conners Teach and Parent Scale.	The authors claim that there was no benefit seen, but they were unable to produce data.
2002	Kuriyama	8 children aged 6 - 16 diagnosed with pervasive developmental delay	Yes	200mg/day B6, no magnesium	4 week randomised double blind placebo controlled study	IQ test.	The subjects on B6 showed an 11.2 verbal IQ point improvement, the placebo group showed 6 points. This was statistically significant (p = 0.01)
2004	Adams and Holloway	Twenty children with autistic spectrum disorders	Yes	Three month, moderate dose multivitamin and mineral	Randomised double blind placebo controlled study	Blood tests to ascertain levels of B6 and questionnaire.	Evaluation before found that autistic children had substantially elevated levels of B6 (p < 0.0000001). This is consistent with the recent theory about low levels of pyridoxal - 5 - phosphate and low activity pyridoxal kinase. Pyridoxal is only poorly converted to pyridoxal - 5 - phosphate and therefore explains the need for high doses of B6. Significant improvement in sleep (p = 0.03) and gastrointestinal symptoms (p = 0.03).
2005	Rimland and Edelson	5780 autistic children and adults	Autistic children taking other interventions.	The dosages of B6 and magnesium were decided by parents.	Parental questionnaire.	Parents' judgement of behaviour.	Parents rated 85 biomedical interventions as to safety and efficacy. B6 and magnesium were rated 'helpful' in 47%, 'no effect' in 49% and 'made worse' in 4%.
2006	Nye and Brice	All research on the use of B6 and magnesium with children with autism	N/A	Dosages of B6 and magnesium	Survey	N/A	Whilst the one research report that they felt contained everything necessary for a reliable study showed significant improvement, they considered it not enough to prove that this a a valid method.
2006	Adams et al	35 autistic children aged 3 - 9.	11 typical children	No supplements given	Single blind controlled study	Microbiological assay used to measure level of vitamin B6	Children with autism had a 75% higher B6 level than controls (p = 0.00002). This supports the theory of low conversion of pyridoxal and pyridoxin P5P. P5P is the active cofactor for 113 well known enzymatic reactions including for formation of many key neurotransmitters.