Friday, May 31, 2013

Vaccines, Autism, and the Scientific Method

Disclaimer: As with other articles posted in this blog, several of the links will lead you to the original research articles as they are found on the websites of their respective publishers.  I feel it is important to provide citations that direct people to the purest source of information that is available.  Unfortunately, many journal publishers cannot allow free access to the public of full articles, so the links may allow you to see only the abstracts in some cases.  Full articles can be purchased, or accessed via university networks that have a paid subscription with publishers.

It is a well-known axiom in any science that ‘correlation does not imply causation’.  Because coincidences run rampant in nature and in life, focused efforts must be made to extract the effect with its true cause from a mess of random concurrences.  This is where the scientific method comes in (Figure 1).  

The observation of simultaneous events may lead one to hypothesize that one causes the other.  But, before any solid conclusions can be drawn there is a prerequisite step: Experimentation.  This is the inconvenient step; the one that requires meticulous planning, long drawn-out trials, repetition, peer-review of the results, and funding.  

This step was bypassed by some during the early 2000’s in their zeal to hop to the conclusion that vaccines cause autism.  But 13 years later, some new results are in from a very broad study designed to probe the alleged vaccine/autism link.

Figure 1. The fundamental scientific method, shown left.  The box on the right shows how faulty theories can be drawn  from conclusions that have no experimental basis and incidentally no results to support them.

First, we go back to 1998 when the correlation was most ostensibly observed.  Andrew Wakefield was at the time a doctor and medical researcher specializing in bowel disorders at the London Royal Free Hospital School of Medicine.  

In the prestigious medical journal The Lancet, Wakefield was the lead author of a published observation that 12 children (11 boys, 1 girl; ages 3 - 10) were all diagnosed with autism-related behavior abnormalities by parents or physicians after receiving the measles-mumps-rubella (MMR) vaccine.  It was noted that 8 of these children had a confirmed onset of behavioral abnormalities within 1-14 days following vaccination.  All 12 children also were found to possess gastrointestinal disorders.  

Wakefield’s work was funded by the UK Special Trustees of Royal Free Hampstead NHS Trust and the Children's Medical Charity.

In consideration of the fact that rates of autism were on the rise (and continue to be) following the introduction of the MMR vaccine in 1988 in the UK, a hypothesis appeared to become self-evident.  As stated objectively in Wakefield’s paper:

“We identified associated gastrointestinal disease and developmental regression in a group of previously normal children, which was generally associated in time with possible environmental triggers.”

The ‘environmental triggers’ referred to here is the MMR vaccine.  So, to paraphrase the hypothesis: ‘12 kids that were once normal now have bowel problems and autism, and we think these issues might be correlated, and there is a possibility that both could be caused by the MMR vaccine’.  This was a fair observation worthy of further investigation, based on the data presented.  The authors conservatively added:

“We did not prove an association between measles, mumps, and rubella vaccine and the syndrome described. Virological studies are underway that may help to resolve this issue.”

This caution was appropriate considering that the sample size was small (12 cases) and the study involved no controls (children not exposed to the vaccine, or children not diagnosed with autism) for comparison.  Because the diagnosis of autism usually happens early in life, as do vaccinations, controls would be indispensable for proving a casue/effect relationship.  Otherwise one could just as well declare that Flinstones Vitamins are the cause of autism and/or gastrointestinal irregularities.

The publication of Wakefield’s paper caused a stir as people mulled the impossible choice between protecting their children from infectious diseases, or protecting them from autism.  

Inevitably, a few chose to abstain from vaccinating their children.  A few years later, an increase in measles cases began to emerge in the UK.  In light of this, 10 of the 12 authors on Wakefield’s 1998 paper clarified their position by issuing a formal ‘Retractionof an Interpretation’ in a 2004 publication of The Lancet:

“The main thrust of this paper was the first description of an unexpected intestinal lesion in the children reported….We wish to make it clear that in this paper no causal link was established between MMR vaccine and autism as the data were insufficient.”

While these 10 authors attempted to distance themselves from a proposed hypothesis that was mistaken for a conclusion among a small segment of the general public, one author was conspicuously absent from this statement: Andrew Wakefield.  

The journal responded in the same issue by reporting that an investigation was carried out and focused on the methods used in the 1998 study and found no ethical violations.  However, six years later, not only was this statement of support retracted by The Lancet, but the entire 1998 paper was retracted:

“Following the judgment of the UK General Medical Council's Fitness to Practise Panel on Jan 28, 2010, it has become clear that several elements of the 1998 paper by Wakefield et al are incorrect, contrary to the findings of an earlier investigation. In particular, the claims in the original paper that children were "consecutively referred" and that investigations were "approved" by the local ethics committee have been proven to be false. Therefore we fully retract this paper from the published record." 

Statements by the editors in the media were far less diplomatic than this carefully worded retraction.  Serious aspersions were cast by journalists and others in the scientific community.  Conversely, a few publicly defended Wakefield amidst the kerfuffle that ensued.  

Andrew Wakefiled lost his medical licence, but responded by telling his side of the story in a 2010 book, and continues to stand firmly by his original hypothesis.

By 2011, the World Health Organizahtion (WHO) sternly urged Europeans to vaccinate their children in response to 26,000 recorded cases of the once rare measles.  90% of those cases had not received the vaccine.  

The Centers for Disease Control (CDC) issued similar warnings in response to much smaller outbreaks that followed in the U.S., where the disease was previously considered to be practically eradicated.  Clusters were observed in highly populated cities, as well as in Minnesota and Utah.

Going back now to 2000, an elaborate study designed to investigate vaccines as a cause for autism was initiated in by Frank DeStefano and colleagues at the US Centers for Disease Control (CDC), which funded the study.  

This study involved a much larger sample size as well as controls: 256 children on the autism spectrum, and 752 without autism were included in the study.  All children received some level of vaccination.  The control subjects were matched for comparison to autistic subjects according to the child’s age, gender, and managed care provider.  The children were 85% male.  Data on these children were compiled retroactively from the records of three separate managed health care providers.  This data included the types and amount of vaccines each child received, when they received them, and when they were professionally diagnosed with autism.  Data for 24 separate vaccines were included in the study, including MMR.  At the time, many of these vaccines contained thimerosal as a preservative – a compound that has caused concerns due to its mercury content.  The care providers examined the children for autism during the vaccination period at the points of 3 months, 7 months, and 24 months.  Comparisons were made among children with and without autism, and how much of which vaccines they received, when they received them, and when the onset of autism symptoms began.

Results and Conclusions
The data was meticulously combed through over the course of the following years, and correlations were probed.  The results were published in an April 2013 issue of the Journal of Pediatrics, with charts and graphs showing that an increase in exposure to vaccine immunogens does not correlate to an increase in autism diagnosis.  Accordingly, the following succinct conclusion was stated:

“We found no evidence indicating an association between exposure to antibody-stimulating proteins and polysaccharides contained in vaccines during the first 2 years of life and the risk of acquiring ASD, AD, or ASD with regression.”

This conclusion is corroborated by several earlier studies that have been published elsewhere in the scientific literature.

Uncertainty is a sentiment that any disciplined scientist must learn to be comfortable with.  It is the gap between questions and answers that compels research to keep moving forward.  

Autism is a question that deserves answers, and it would be comforting to believe with certainty that we have found them.  However, this does not yet appear to be the case.  Because the works of science are often communicated in a language that many don’t speak, people often must accept or reject them as an act of faith.  It is my intention and hope that such faith in scientific methods can be perfected by making the works more accessible.   

In writing this post, my bias was obviously in favor of the work that I am familiar with that was presented here.  But all conclusions are subject to dialogue.  If you are aware of peer-reviewed reliable scientific research that does demonstrate a cause/effect relationship of vaccines and autism, by all means post it in the comments.  Thank you!