Is the rise in diagnoses of autism the result of changing diagnostic practice, rather than any sinister environmental cause? New evidence from California points in that direction.  

Parents in the UK anxious to discover why their child is autistic have focussed on the MMR vaccine, while in the US the supposed villain is thimerosal, a mercury preservative used to prevent bacterial deterioration of childhood vaccines. Neither claim is persuasively backed by evidence.
 
But the rise in diagnoses continues. In December a US team that included experts from the Centers for Disease Control and Prevention published data showing that nearly 1 in every 100 US eight year-olds is now diagnosed with an autism spectrum disorder (ASD).
 
This represents a 57 per cent increase on four years ago, the team reported in the CDC’s Morbidity and Mortality Weekly Report on December 18. The diagnoses, which come from 11 sites across the US, also show that boys are 4.5 times more likely to suffer the condition than girls.
 
The CDC and collaborators from the University of Alabama discuss possible reasons for the rise, including a broadened definition of the disability and heightened awareness. But they do not reach a conclusion on whether it represents a true increase or simply a greater tendency to diagnose ASD.
 
A second US study, published today in Autism Research, does shed some light on the question, however. A team from the University of California at Davis looked at diagnoses of ASD in California in children up to six years old, and found 10 “clusters” where diagnoses are typically about double the average for the state.These clusters, they say, occur around regional centres for autism diagnosis and treatment, and coincide with areas where parents have higher than average educational attainment.
 
In the area served by the San Diego Regional Centre, for example, there was a cluster of cases with an incidence of 61.2 per 10,000 births, while that in the rest of the area served by the centre was 27.1 per 10,000. The study covered the period between 1996 and 2000, and all the children included had reached the age of six.
 
The children of older, white, and highly educated parents are more likely to be diagnosed as suffering autism or ASD. So the team concludes that the clusters are not the result of some environmental insult, but can be explained by the education, age and ethnicity of the parents.
 
The implication is that increased ascertainment may be the key to the rise in autism. Parents with high expectations who live near treatment centres are more likely to be given a diagnosis of autism in their children than those with lower educational attainment who do not live so close to a treatment centre.
 
The actual level of diagnoses in the California study is, however, much lower than in the CDC study. Of roughly 2.5 million children born in California between 1996 and 2000, about 10,000 were diagnosed with autism or ASD by the age of six – a rate of one in every 250. The CDC figure is one in 110. (The figures may not be directly comparable, as the CDC traced children to the age of eight, not six, but it is hard to believe this could account for such a large discrepancy.)
 
The children in the CDC cohort were born in 1998 and the 57 per cent increase is diagnoses is measured against data from children born in 1994, the year in which new, broader diagnostic criteria were introduced. Whether this alone can account for the increases – and that in future years diagnostic levels will plateau – is uncertain. But together with the cluster evidence from California, it does suggest that the rise in diagnoses is the result of greater awareness and preparedness to diagnose the condition, rather than any absolute increase in its prevalence.
 
In the past, many of these children might have been described as unsociable or obsessive, but might not have had a medical label of any sort attached to them. Today, they do.
 
The California study, its authors say, does not point to any particular environmental hazard – apart from having well-educated parents. Whether this implies a genetic link between autism and intelligence, or simply that well-educated parents are more likely to seek help, is an open question.
 
The rapid rise in diagnoses cannot be attributed to the genes, even if they are a strong contributory factor. They cannot have changed fast enough to account for the rise in diagnoses. Changes in diagnostic practice remain the likeliest explanation.