The antioxidant - called N-Acetylcysteine, or NAC - lowers irritability in children with autism and also reduces their repetitive behaviour patterns.Irritability affects 60 to 70 per cent of children with autism.
Researchers at Stanford University School of Medicine and Lucile Packard Children's Hospital studied 31 children with the disorder.
Lead researcher Dr Antonio Hardan said: 'We're not talking about mild things - this is throwing, kicking, hitting, the child needing to be restrained.
'It can affect learning, vocational activities and the child's ability to participate in autism therapies.'
The study tested children with autism ages 3 to 12. They were physically healthy and were not planning any changes
In a double-blind study design, children received NAC or a placebo for 12 weeks.
The NAC used was a pharmaceutical-grade preparation donated by the neutraceutical manufacturer BioAdvantex Pharma.Subjects were evaluated before the trial began and every four weeks during the study using several standardised surveys that measure problem behaviours, social behaviours, autistic preoccupations and drug side effects.
During the 12-week trial, NAC treatment decreased irritability scores from 13.1 to 7.2 on the Aberrant Behaviour Checklist, a widely used clinical scale for assessing irritability.
The change is not as large as that seen in children taking antipsychotics.
'But this is still a potentially valuable tool to have before jumping on these big guns, Dr Hardan said.
In addition, according to two standardised measures of autism mannerisms and stereotypic behaviour, children taking NAC showed a decrease in repetitive and stereotyped behaviours.
Finding new medications to treat autism and its symptoms is a high priority for researchers. Currently, irritability, mood swings and aggression, all of which are considered associated features of autism, are treated with second-generation antipsychotics.
But these drugs cause significant side effects, including weight gain, involuntary motor movements and metabolic syndrome, which increases diabetes risk.
By contrast, the side effects of NAC are generally mild, with gastrointestinal problems such as constipation, nausea, diarrhoea and decreased appetite being the most common.
The state of drug treatments for autism's core features, such as social deficits, language impairment and repetitive behaviours, is also a major problem.
Dr Hardan said: 'Today, in 2012, we have no effective medication to treat repetitive behaviour such as hand flapping or any other core features of autism.'
The researchers believe that NAC could be the first medication available to treat repetitive behaviour in autism - if the findings hold up when scrutinised further.
'One of the reasons I wanted to do this trial was that NAC is being used by community practitioners who focus on alternative, non-traditional therapies,' Hardan said.
'But there is no strong scientific evidence to support these interventions. Somebody needs to look at them.'
Dr Hardan cautioned that the NAC for sale as a dietary supplement at chemists differs in some important respects from the individually packaged doses of pharmaceutical-grade NAC used in the study, and that the over-the-counter version may not produce the same results. He said: 'When you open the bottle from the drugstore and expose the pills to air and sunlight, it gets oxidised and becomes less effective.'
Although the study did not test how NAC works, the researchers speculated on two possible mechanisms of action.
Firstly, NAC increases the capacity of the body's main antioxidant network, which some previous studies have suggested is deficient in autism. In addition, other research has suggested that autism is related to an imbalance in excitatory and inhibitory neurotransmitters in the brain. NAC can modulate the glutamatergic family of excitatory neurotransmitters, which might be useful in autism.
The scientists are now applying for funding to conduct a large trial in which they hope to replicate their findings. Dr Hardan said: 'This was a pilot study. Final conclusions cannot be made before we do a larger trial.'
Stanford University is filing a patent for the use of NAC in autism, and one of the study authors has a financial stake in a company that makes and sells the NAC used in the trial. The researchers said that the findings must be confirmed in a larger trial before NAC can be recommended for children with autism.
The study appears in Biological Psychiatry.