MMR jab has not raised autism

The MMR jab is "off the hook", The Guardian reported today, referring to controversial theories that the vaccination may be linked to autism. The newspaper said that new research shows that rates of autistic disorders are similar among adults and children, a finding that further undermines theories that the combined MMR jab is responsible for a supposed rise in cases over recent years.

The Times reported that the research has spurred the National Autistic Society to get "off the fence" over the supposed link between MMR and autism. The organisation has previously taken a neutral position in the controversial debate but has updated its position statement on autism and the MMR vaccine, adding an acknowledgement that a weight of epidemiological evidence “indicates that there is no statistically significant link between the MMR vaccine and autism”.

The report behind these stories is based on the findings from the latest Adult Psychiatric Morbidity Survey, carried out in 2007 by the National Centre for Social Research. The survey was completed by 7,400 adults living in English households, with a selection of participants also completing clinical interviews to further investigate mental illness including autistic spectrum disorders (ASD).

The researchers estimated the rates of autism in the responders and extrapolated the findings to estimate the prevalence rate of ASD for England. They conclude that 1% of the adult population is affected, a similar rate to that seen in children.

Where did the story come from?

The National Centre for Social Research, has recently published a report based on findings from the Adult Psychiatric Morbidity Survey 2007, which specifically aimed to determine the prevalence of autistic spectrum disorders (which includes autism and Asperger’s syndrome) in adults in England.

What kind of scientific study was this?

This was a survey of adults living in private households in England between October 2006 and December 2007.

In the first phase of this research, households were selected in a way that ensured they were representative of the population of England. At total of 13,171 households were eligible to receive an interview during phase one. Participants were interviewed for an average of 90 minutes, using methods that screened for mental health disorders. Interviewees were also asked about their demographics, risk factors and their use of services. In total, 7,461 adults participated in phase one, including 58 adults that had proxy respondents answer on their behalf.

In the second phase of the study a subset of those interviewed in phase one were interviewed by clinically trained interviewers. The probability of a participant being selected for the second phase of the study was weighted according to the probability that the participant had psychosis, ASD, borderline personality disorder or antisocial personality disorder. This was estimated using the participants' responses to the screening questionnaire in phase one. This meant that people with a greater likelihood of having these disorders were more likely to be selected for clinical interview in which their possible disorders were more formally assessed.

Using this process the researchers selected 849 adults for phase two interviews, which were conducted with 630 of them.

In their analysis of the data from the survey, the researchers weighted the findings, i.e. they extrapolated them to generate a prevalence rate for the population of the country as a whole.

What were the results of the study?

On the basis of the initial survey of more than 7,000 adults and their 649 in-depth clinical interviews in phase two, the researchers estimate that 1% of the adult population in this country has an autistic spectrum disorder (ASD). The rate is higher among males (1.8%) than it is in females (0.2%), a pattern similar to that seen in children.

ASD was more common in single people, those with no educational qualifications, those with greater deprivations levels and in people with a lower IQ.

What interpretations did the researchers draw from these results?

The survey has several key findings, but the one most relevant to the news coverage is that an estimated 1% of the adult population in England has an ASD.

What does the NHS Knowledge Service make of this study?

There are some important issues that should be highlighted:

• Only 19 people had a clinically confirmed ASD in phase two of the study. The researchers estimate that if everyone in phase one had been re-interviewed, a total of 72 cases would have been identified.
• The number of people confirmed to have ASD is a small sample, so further comparisons between those with and without ASD (for example, in terms of their demographics) should be interpreted cautiously. The researchers say that “great caution is required in interpreting the population distribution of ASD (particularly among women)” because of the small number of ASD cases clinically identified.
• Although an association was found between ASD and demographics (a finding that ASD is associated with lower education, greater deprivation, lower IQ and so on) this cannot be interpreted in a causal context. A study design such as this, which is cross-sectional, cannot establish causal links.
• The report itself does not mention MMR, but this issue is raised by the NHS’ Information Centre in an accompanying press release. This says, “If MMR were a factor in the development of the condition, prevalence would be expected to be higher among children and young adult age bands because MMR was introduced in 1990/91 and only those currently aged in their early twenties or younger have been routinely offered it.” This seems reasonable.

This report provides further evidence that adds to the many studies showing no link between MMR and autism. These findings will be welcomed by parents, doctors and many others who have been involved in investigating the controversial link first suggested by Dr Andrew Wakefield in 1998.