Abstract | Summary | Original Article

Prevalence of autism and parentally reported triggers in a north east London population.

Lingam R, Simmons A, Andrews N, Miller E, Stowe J, Taylor B.

Archives of Disease in Childhood. 88:666-670. 2003.

Question: In children born in north east London between 1979 and 1998, what was the rate of autism per year and what factors did parents cite as causing their child's condition?

Background: This is a prevalence study that looks at how common autism is in each of the years 1979 to 1998, amongst children in five districts in north east London. Prevalence studies are particularly important to aid in resource planning and also to look at what causes conditions. The authors also looked at causation from the point of view of parents whose children experienced regression after a period of apparently normal development. Specifically, there were looking to see if there was a difference in the reasons the parents gave before and after the publication of a paper proposing that the measles, mumps and rubella (MMR) vaccine caused regression in children who were later diagnosed with autism. If there were more reports after the paper came out, then it would suggest that the publicity surrounding the study had influenced the parents' reports.

Participants: The participants were the 567 children born between 1979 and 1998 in five districts in north east London who were diagnosed with autism spectrum disorders.

Study Design: Prevalence study using records linkage.

Method: Children with a diagnosis of autism were located through computerized records, disability registers and by contacting local pediatricians, local child psychiatrists, and special schools to help ensure that all children with autism were located. Computerized lists of children's diagnostic records, immunization records, and health history records were examined to verify the diagnosis of autism, to find any reports of regression noted in the charts, and the reasons for the regression that were cited by the children's parents. The records were linked to determine first mention of "autism," any mention of "regression" and the date of MMR immunization. After finding all the records of children in whom regression was reported, the researchers looked for the parents' opinion as to the cause of their child's condition.

Main Results: The records of 567 children with a diagnosis of autism were located, giving a prevalence rate of 14.9 per 10,000. If children with atypical autism and Asperger Syndrome were included, the rate would be 29.4 per 10,000, although the authors believe these are underestimated. The study demonstrated that there was an apparent increase in the yearly rates from the early 80s up until the early 90s, when they leveled off.

Of the 567 children, 278 had childhood autism, 195 were diagnosed with atypical autism, and 94 had Asperger Syndrome. Further, in these groups, 27% of children with childhood autism had, by parental report, experienced regression , as had 23% of the atypical autism group, and 2% of the Asperger Syndrome group. The Asperger group was removed from the analysis because their symptoms and the low level of regression made them too dissimilar from the autism and atypical autism groups. The number of children who had experienced regression as reported by their parents and for whom data was available was 117 (of a possible 118).

In 44 children with detailed records of regression, 42% had a specific trigger mentioned. In 13 children the trigger was a change in the household or an event like the birth of another child; in 12 children vaccination was mentioned; viral or bacterial infections were noted by 7; occurrence in a post-surgical period was mentioned by 2; and other causes were mentioned by 3. It was also found that after 1997 when there were reports of an association between the MMR vaccine and regressive autism, although many parents reported that they remembered that their child had regressed, their children's records showed that there were mentions of concerns about the children's development prior to the child's vaccination.

Conclusions: In five districts in north east London, the authors used rigorous methods to find the records of all children diagnosed with autism between the years 1979 and 1998. The prevalence rate of 14.9 per 10,000 they calculated is in keeping with the rates found by other researchers.^1,2 The current study authors also found a leveling off of cases by the early 90s. Their task of looking at the parents' perception of causation in the children who regressed suggested that the parents were probably influenced by the publicity given to the Wakefield theory about regressive autism and that their memory for past events could be faulty (see recall bias). The possibility that the children who experienced apparent regression may have had subtle symptoms prior to the more "dramatic" events of language or other skills loss should be considered.

The main conclusions were that because of improved recognition of the symptoms, diagnosis was being made more frequently and at an earlier age, thus increasing the number of children being diagnosed with autism. However, the leveling off occurred because in the later years of the study there were fewer children being diagnosed at the age of 4 as they had already been diagnosed at the age of two.

This study further supports the contention that there is no "new variant" autism. This means that the link between the MMR vaccine and the development of a new type of autism associated with regression after vaccination has again been refuted. It also appears that some parents' memories were influenced by media reports of a connection between MMR and regression, and thus the accuracy of these reports must be questioned.

¹ Chakrabarti, S, Fombonne E. Pervasive developmental disorders in preschool children. JAMA 2001; 285: 3093-3099.
² Baird G, Charman T, Baron-Cohen S, et al. A screening instrument for autism at 18 months of age: a 6-year follow-up study. J Am Acad Child Adolesc Psychiatry 2000; 39:694-702.