The discovery of a number of autism clusters centred around the Los Angeles area in California, USA, has left researchers struggling for an explanation. Alistair Gee reports.
In April this year, experts on autism gathered in Los Angeles, CA, USA, to discuss a curious finding. Investigators had identified clusters of autism cases in Los Angeles County, and attendees at the meeting pondered various explanations for their existence.
“None of us had any idea”, recalls Diane Anand, executive director of Los Angeles' Frank D Lanterman Regional Center, part of the state system that provides services to people with developmental disabilities. “People are very mystified by this.”
According to the results of one study this year, children in a 20 km by 50 km zone centred on West Hollywood were at four times greater risk of autism than were children anywhere else in California. And the Lanterman Center, with a catchment area that includes West Hollywood, ranked joint highest among the 21 regional centres in California for the proportion of autistic children in its mid-2007 caseload. The centre was vetted for diagnostic biases, but none were found.
In an intriguing turn, US researchers have suggested that a range of social influences, such as the education level of parents, might be contributing to clustering in Los Angeles and elsewhere in California. The scientists are not discounting the role of genetics, toxicants, infections, maternal stress or any of the other myriad factors that possibly combine to result in autism, but they say that social processes are affecting the extent to which the disease is reported.
Autism remains an intractable problem. Before the 1990s, the prevalence of autism was estimated at one in 2000, but today autism spectrum disorders are thought to affect one in 110, with an overall autism prevalence 20 times higher than past estimates, according to the US Centers for Disease Control. “Those numbers are extraordinary as they are not levelling off”, says Thomas Insel, director of the US National Institute of Mental Health. “Whichever of many factors are driving this, they are still very much in play.”
Clustering in Los Angeles was highlighted by two separate groups of scientists. In February, a team at UC Davis identified ten clusters across California, including five in the Los Angeles area and Orange County and three in the San Francisco Bay area. Then, in May, members of an autism project at Columbia University published a paper pinpointing a primary Los Angeles cluster focused on West Hollywood, and secondary clusters nearby. The Los Angeles clusters found by the two groups did not overlap precisely.
One of the first things that stands out, both sets of researchers suggest, is that their studies make it more complicated to link autism with vaccines. If vaccines are the cause they should affect children across broad areas, not in local clusters, says Peter Bearman, head of the Columbia group. “At any rate, a theory of vaccines would have to make sense of how they interact with environments,” he says.
The Columbia researchers argue that a social influence mechanism might be bolstering the clusters they identify. They say that as parents become more aware of autism—through the exchange of information with other parents at places like parks, playgrounds, and preschools—they are more likely to recognise autism symptoms in their own children. This mechanism is said to account for 16% of the prevalence increase in California from 2000 to 2005. “Where people are talking about autism and have had individual experience with it, it really helps to create awareness, in a way that seeing a poster in a doctor's office listing the main symptoms doesn't”, says Marissa King, one of the researchers.
The social-influence theory is a convincing argument to some parents. “I believe that if I had someone in our immediate circle who had a child on the autism spectrum, I probably would have recognised it in Minna Grace earlier”, says Miriam Huntley, the mother of a 3-year-old with autism living in the Silicon Valley town of Menlo Park.
However, the UC Davis researchers draw attention to a different social influence, and propose that clusters are linked to the education level of parents—those with a college education are more likely to obtain an autism diagnosis than those who did not graduate from high school. “It's the ability to negotiate the bureaucracy, to get a doctor to see what you're seeing, all those things”, says Karla Van Meter. A higher level of education could also be an indicator of being wealthier, and therefore having better access to resources such as better preventive medical care and paediatricians, adds Anand.
The researchers acknowledge that there are broader problems with the datasets used by both groups. The studies are based on records kept by California's Department of Developmental Services (DDS), which are thought to be some of the best autism records in the USA, although a child is only added to the database once a parent applies for services, and they are estimated to include at best about 80% of autism cases in the state. It is also unclear whether the electronic DDS records contain the same information as the paper records, says Lisa Croen, director of the Autism Research Program at Kaiser Permanente Northern California, who has compared the two sets. “They're not entirely clean data. You have to go beyond what's there at face value.”
Full-size image (96K) CorbisDownload to PowerPoint
The Californian clusters are not the first to be identified in the USA—the 1990s case of Brick Township, New Jersey, was particularly well-known. In its investigation, the Centers for Disease Controls (CDC) determined that the rate of autism spectrum disorders in the town was 6·7 cases per 1000 children and compared them with the rate elsewhere. But, says Catherine Rice, an epidemiologist at the CDC's National Center on Birth Defects and Developmental Disabilities, “once we went back and established more prevalence levels in other areas of New Jersey, we found that the Brick prevalence was not necessarily higher than in other areas”.
A cluster has also been reported among Somali immigrants in Minneapolis, where between two to almost seven times the number of Somalis as non-Somalis were recorded as having an autism spectrum disorder. Although some have drawn links to the stress of being a refugee, for example, Rice says that the cluster is apparently the result of better awareness of autism in the Somali community. “When you compare the number of children identified to what you would expect in the general population, it really isn't different—it seems actually to be an issue of better identification among the Somali population from the data that are there so far.”
In California, researchers are also exploring how social influences affect autism prevalence over time, particularly because the spike in cases is so striking: the number of people with autism receiving disability services increased by 12 times between 1987 and 2007. The Columbia group suggests in a study that caseloads increased rapidly when definitions of autism were changed in state and national diagnostic manuals, and that a quarter of the children diagnosed with autism today would not have been diagnosed as such in 1993.
The broader question in many of these studies is whether the increase in autism is real or an “epidemic of discovery”, meaning that autism has always existed at its current levels, and the ballooning statistics are merely the result of new ways of diagnosing and recording the disease. The issue is the subject of sometimes bitter debate among researchers, parents, and advocacy groups. Not least, says Anand, because the idea of an epidemic of discovery as opposed to an actual epidemic might make the increased caseload seem somehow less threatening. “With shrinking public resources, policy makers may determine that funding is not necessary for the intensive early intervention services that are available now”, she says.
George Washington University anthropologist Roy Richard Grinker, for one, is sceptical that there has been an increase in real cases, and draws attention to the improved skill of doctors in diagnosing autism and parents' growing ability to obtain diagnoses. “Increases in prevalence don't necessarily mean increases in incidence”, he says. But Columbia's Bearman thinks an improved discovery process “is not the whole story”, and Insel says he will suspect a true increase of autism in the absence of evidence to the contrary.
Staff at the regional centres in California also say there has been a real jump. “Increased numbers of people are knocking on the door and I can't simply believe that we failed, that these people were always out there but somehow bumping along and functioning at a higher level than they now are”, says Anand. For now, researchers say, studies of clusters and increased prevalence are indicating just how little-understood autism really is—environmental and social factors are interacting with genetics in unknown ways to produce the disease. “There are just so many things contributing to it”, says Van Meter. “There's no smoking gun.”http://rich-biofool.blogspot.com/2010/11/discovery-of-number-of-autism-clusters.html
It also has implications for the treatment of autism since the changes in the brains of fragile X and autistic people are thought to significantly overlap 
