The only thing we know is its name

Autism appears to be on the increase, but the condition's cause is still a mystery. Among the current suspects, reports Adrian Mourby, are diet, genes and vaccines.

The recent controversy over a possible connection between the measles, mumps and rubella jab and autism has again raised the question about what autism actually is. Although cancer remains difficult to cure, scientists have nevertheless managed a fair consensus about the nature of the thing they are dealing with. There is even some agreement on what may cause it. But autism is different. Some experts consider it curable, others emphatically do not. Some see it as a psychiatric disorder while others view it as an extreme form of personality.

In the opinion of Paul Whiteley from Sunderland University's Autism Research Unit autism is "a pervasive developmental disorder appearing before three years of age that primarily affects the way a person communicates and interacts with other people and their surroundings". But Bernard Rimland of the American Autism Research Institute sees it as "a severely handicapping biological brain disorder that begins at birth or within the first two-and-a-half years of life". Digby Tantam, clinical professor of psychotherapy at Sheffield University, calls it "a disorder of nonverbal communication with multiple causes".

Before its symptoms were first scientifically defined in the 1940s, many people exhibiting what would now be classed as autistic behaviour were committed to mental institutions. "Even following the description of autism, parents of people with autism were demonised and blamed for their child's condition, with Freudian psychoanalysts using terms such as 'refrigerator parents' as convenient answers to the causes of autism," Whiteley says. "Although we still do not know the causes, psychiatry is beginning to recognise what many parents have said for a long time: autism is not caused by a failure of parental bonding - it is a psychiatric condition, probably with an organic cause."

This shift in perception is significant, but there remains little agreement about what Whiteley's "organic cause" might be. What everyone does agree on is that diagnoses of autism are increasing rapidly. "For many years autism occurred in about five children per 10,000 live births," Rimland says. "Since the early 1990s, the rate of autism has increased enormously throughout the world, so that figures as high as 60 per 10,000 are being reported." Although some suggest this is due to better awareness of the condition among doctors and health visitors, many researchers believe this does not account for the numbers. "In Britain there has been an epidemic of autism," Rimland says.

As yet, it is not possible to say whether the increase is across the whole range of autistic conditions or more among "high-performing" individuals with Asperger syndrome - which might support the better-awareness hypothesis. According to Whiteley, "there are no reliable figures to suggest which groups may be increasing more rapidly".

Rimland believes over-vaccination of children could be behind the rise. "The effects of mercury on autistic children has come to light only in the past 18 months after it was realised that the amount of mercury preservative in many vaccines (although not MMR) is in excess of permissible standards. In the past year, a number of physicians in the United States have reported extremely good results in improving the health and behaviour of autistic children when mercury in the children's bodies was removed. There is an enormous amount of credible evidence that vaccines can and do cause harm."

The injurious effect of mercury on the brain is well known. Hatters were notoriously eccentric long before Lewis Carroll immortalised them, the culprit being the mercury that was used to stiffen felt. But for those who are inclined to see autism as a biological brain disorder brought about by mercury poisoning, the question remains: why do a relatively small number of children succumb and why is autism more prevalent in some families than in others?

Several academics are looking into the role genes might play in autism. At Utah State University, research into the genetics of autism has been going on for several years, based on Reid Warren's exploration of the incidence of autism and gluten intolerance in certain families. But genetics is likely to be only part of the story, and these days more researchers are investigating the link between genetics and the environment.

At the University of Sunderland, Paul Shattock believes that diet may be a crucial trigger. "The notion that dietary changes could produce amelioration in autistic symptoms may appear bizarre, but anecdotal reports of improvements following removal of dietary gluten and casein stimulated this line of research. It is known that incomplete digestion of gluten and casein will result in the production of opioid (morphine-like) peptides in the intestines. If these peptides reach the brain, they will affect or regulate transmission in all the major systems and so affect perception, cognition, behaviour and brain development. Since we all produce these peptides during digestion the key element may be the increased intestinal permeability in affected individuals. There is published evidence that this exists."

The controversial Royal Free Hospital study on MMR has also been investigating the premise that gut permeability is a factor in autism: it suggests a link between MMR, the inflammatory gut condition Crohn's disease and autism. Shattock, however, believes that intestinal permeability could be increased by environment factors such as pesticides, chronic viral infections and recent dietary changes. These, he feels, would help explain the rapid increase in diagnoses of autism in recent years.

His view has its critics, Tantam among them. "I think the role of diet is unproven," he says, although he admits that evidence for a link between gut permeability and autism is "accumulating". But he cannot see how gut permeability could account for abnormal brain development in utero.

The idea of autism as the byproduct of some unwitting poisoning is not far-fetched as it sounds. George III's madness is now considered to have had an organic cause. Researchers in Japan who have been looking into a possible relationship between autism and candida believe they have proved that candida is able to produce toxins that cause severe long-term disruption of the immune system and may also attack the brain.

"It is too early to reach a firm conclusion," Rimland says. "But based on the weight of the information gathered to date, it seems to me probable that a small but significant proportion of children diagnosed as autistic are in fact victims of a severe candida infection. In a typical case of this kind, the child appears to be a normal, reasonably healthy infant for the first 18 to 24 months. Speech has started, and the child displays the usual level of interest in his family and his surroundings. A series of ear infections occur that are routinely treated by antibiotics. Soon thereafter, ominous changes begin to occur. Speech development stops, then regresses, often to the point of muteness. Within a few weeks or months the child becomes unresponsive and loses interest in his parents and his surroundings. The concerned parents take the child to various specialists and finally come up with a diagnosis of 'late-onset autism'."

Rimland cites the work of Alan Levin, a San Francisco allergist, who in 1981 was treating a three-and-a-half-year-old boy who had been diagnosed after a course of antibiotics. "Levin knew that such antibiotics often kill the micro-organisms that compete with candida in the human body and thus allow candida to grow to overwhelming proportions." Aware of the mounting evidence that candida might be less benign than commonly believed, Levin tried nystatin, an antifungal drug that is toxic to candida but not to humans. "At first the patient got worse," says Rimland, adding that this is "a common reaction caused by the toxins released by the dying candida cells". Today, however, "he is active and greatly improved with few remaining signs of autism. His immune system is still impaired, however, and he still requires treatment."

The variety of opinions about the aetiology of autism can seem confusing, but in Tantam's experience it is no greater than the range in other psychiatric disorders. "This reflects the basic difficulty of understanding conditions in which brain dysfunction is expressed as psychological disturbance. There are just too many modulating factors for a clear picture to emerge. These intervening factors include emotional, social and biological ones."

Even the term autism is a very broad umbrella for a wide spectrum of ability. Whiteley calls it "a blanket term" for children who show some symptoms of Kanner autism, the original diagnostic category, or the full diagnostic criteria for Asperger syndrome.

For scientists such as Tantam, any attempt to define or cure such a complex problem as autism is perhaps of less use than developing coping mechanisms. "I have become increasingly interested in the value of counselling for more able people with autism, to help them to overcome their disability and to empower them to overcome the social practices that unnecessarily discriminate against them because of it," he says.