Getting to the roots of Bad Science: Reflections on the autism/vaccine debate

hypodermic needle IMG_7418

Image by stevendepolo via Flickr

With a recent measles outbreak in my home city – and the accompanying call by the authorities to get children up-to-date with their MMR – the autism/vaccine debate has again roared to life. For a while, I get passionate and roll up my virtual sleeves. I go in to bat against misinformation, bad arguments and baseless scare-mongering. After a while I get tired and battle-weary, and take a step back; I take a break from reading the forums in particular because there’s always another conspiracy-nut ready to take the last one’s place, and make the exact same assertions that I knocked down from someone else a page earlier.

Stepping back gives me perspective too, about some very central questions – in particular as to what are the main claims and arguments I’m coming up against, and why do people cling to them so fervently in the face of opposing evidence and science. Why do they keep citing the same few poor studies, in the face of scores of larger and more robust ones? Why do they think anecdote is stronger than scientific explanation?

I’m half way through a book which is doing a fantastic job of helping me to understand those “whys”. Reading the book, and reflecting on all those debates I get involved in, have helped me to see a huge part of the reason that some people just can’t open their minds to trying to find the answers, instead of asserting the conclusions first and then grasping so tightly to any evidence they can find to support it.

It’s a lot to do with a misunderstanding of how scientific inquiry works, and an inability or unwillingness to fact-check and follow where those superscript numbers lead. Because this is such a time-consuming task, we naturally look for people of authority to shorten the process – doctors, professors, trusted people who seem to know what they’re talking about. But if we don’t check the credentials and quality of those authority figures, we’ve got the same problem. Again, checking credentials is also very time-consuming and may not always be possible.

As I read my book (“Bad Science” by the amusing Ben Goldacre), I am given reason to believe his claims about science and whether certain people are worth listening to, because he as good as makes an art-form out of this in-depth fact-checking, and teaches us to do the same.

However I am also reminded that even someone as intelligent and informed as him, can get mislead when they take steps out of their arena of expertise. He tends to make certain claims about societal ills and potential political cures, without understanding the history, evidence and theories behind such suggestions. He doesn’t provide any tidy superscripts for his societal recommendations, but they are not the point of his book – attacking bad science is. He is clearly aware that these side-points are not matters he is trained to understand, but he brings them up time and again anyway. It annoys me in an otherwise sound and impressive book (which I must repeat, I have not finished just yet – I will write a post about the entire book after I have).

The lesson in that can be taken to the general discussion of the autism/vaccine debate too. That we should be careful when taking a gastroenterologist’s claims about immunology and developmental disorders. Yet we have to avoid getting stuck in thinking we can’t trust an immunologist because they’re not an autism expert, or an autism expert because they’re not a immunologist. Yes it would be ideal to have access to a developmental pediatrician who was trained in immunology, who had both decades of field experience and a professorship at a highly ranked university. But such markers (though helpful and relevant) don’t guarantee you someone worth listening to either – like the woman I’ve mentioned previously who is Harvard trained but links autism to breastfeeding.


I am used to having to deal with conflicted authorities and irreconcilable claims. My training in both philosophy and law was filled to the brim with arguments that looked unresolved and unresolvable. Many of my peers felt overwhelmed by the conflicts and became nihilists, anarchists, or relativists. I felt the same temptation. But instead I did much along the same lines as what Ben Goldacre encourages us to do: Look deeper. For him and science, it is a matter of looking at the foundation of the claims and the studies that lead to them, as well as understanding human psychology for why we get sucked in by bad science. For me, and law and philosophy, it is a matter of looking to the arguments for the claims too – go deeper to the epistemology, the logical fallacies, the contradictions. Seek to understand the method, and you will be well-armed when faced with conflicting “authoritative” statements at higher levels of discussion.

It’s hard, no doubt. It is time-consuming and exhausting. But if the conclusions matter to you – if you really want to know whether to vaccinate your child, and what might cause your child’s autism – then it is worth the effort. Is it not?

And we will make mistakes along the way. We will sometimes be in error and we may feel embarrassed by our errors and not want to admit to them. But again, embarrassment and admitting you’ve been wrong, is the much lesser of the comparative evil of continuing to espouse incorrect claims which cause harm to people. I have made errors before, and I will again. That does not make me a lesser person, and it does not mean I should conclude there are no truths in the world. I would only be a lesser person if I refused to learn from those errors, and concluded that the search for truth was impossible or pointless.

This entry was posted in Ought to vaccine or Autism Vaccine, Resources for Parents and tagged , , , , , . Bookmark the permalink.

11 Responses to Getting to the roots of Bad Science: Reflections on the autism/vaccine debate

  1. Jack says:

    As a scientist (microbiology/biochemistry) I usually need to see strong evidence before I believe anything that is written. The minute someone says they ‘believe X caused Y’ then they are in the realm of religion.
    Just comparing the ratios of boy/girl (5.5:1) who have autism against what I would assume would be close to 1:1 for immunization, the claim that they are linked just doesn’t make sense. There has to be a genetic factor. Small studies in 1977 and 1990 showed that identical twins had a higher rate of autism than fraternal. Expanded studies are underway now.

    • MJ says:

      The confounding thing about twins studies is that you get very different results in different studies. Some put the identical twin rate at almost 100% while others put it in the low 30s. This number can vary based on the gender of the twins (strangely, girls are more likely than boys to both have autism), the exact type of autism, and the time period that you are looking at.

      This is somewhat beyond the scope of this post, but the general trend seems to be that all twins (MZ/DZ) have a higher chance of both having autism than typical siblings. Typical sibling is something like 10%, DZ twins are like 20%-40%, and MZ twins are 80%-90%.

      Most people look at the MZ rate and see a genetic factor, but few people look at the DZ rate and see the environmental one.

      • Jack says:

        Environmental factors could part of the picture, but definitely not the only one. Too many studies show gender bias for males and none for socio-economic differences. If it was only environmental (and by that I assume you mean vaccines) you would see the same chance for both DZ and MZ. I would find it strange that a parent would only vaccinate one twin!

        Strange, I always hear that vaccination rates are decreasing but autism diagnosis is increasing. Something doesn’t add up.

  2. Jack says:

    That’s the problem with small studies. I think they only had a sample of ~20 individuals. It’s impossible to claim anything with that number as an absolute fact. Hence the need for expanded studies, looking at many hundreds of individuals if not more.

    The same problem was with he Wakefield study, small samples, subjective data (as in when the parent noticed autism symptoms) and other data not misreported. Until there is a biochemical or genetic marker identified most studies are not going to prove anything either way. If we have a marker then we can conduct proper studies from ~13 weeks on.

    We can look at small sets of data in isolation and suggest correlations, but when you expand out into he general population again those claims are usually found to be false or just bring up more questions.

  3. MJ says:

    “If it was only environmental (and by that I assume you mean vaccines) you would see the same chance for both DZ and MZ.”

    No, I actually just meant non-genetic influences, not vaccines specifically. When you look at the differences between DZ twins and other siblings there are two main differences – a shared prenatal environment and a very similar environment for the young years. MZ twins take it a step further and typically share more of the prenatal environment and have a common genetic code.

    “That’s the problem with small studies. I think they only had a sample of ~20 individuals.”

    Not sure if that was directed at the twin studies or not, but there are recent twins studies in autism that had more than a hundred twins.

    “Until there is a biochemical or genetic marker identified most studies are not going to prove anything either way.”

    I think one of the problems is that is that there isn’t just one form of autism. There are many hints that there are different forms of autism, or autisms if you prefer. Each subtype could have its own, distinct, biomarker.

    Although I agree that even nailing down one or two different markers would be a huge benefit.

    “Strange, I always hear that vaccination rates are decreasing but autism diagnosis is increasing. Something doesn’t add up.”

    In the US that isn’t really true. Vaccination rates have held steady or risen over the past ten or twenty years.

    • Jack says:

      Interesting points. I just realized that I should have done a bit more research before posting comments. Had some spare time last night so I did a quick very focus literature review using ‘Autism’ and ‘MZ and DZ’ as the search terms. Just went back 10 years. Summaries below

      1. Qin et al 2010 A twin study about genetic effects on mental health development of children.

      102 twins (50 MZ 43DZ) in China
      autistic symptoms (MZ group r=0.680, P<0.01; DZ group r=0.372, P<0.01)

      2. Rosenburg et al 2009 Characteristics and concordance of autism spectrum disorders among 277 twin pairs.

      210 DZ 67 MZ in US
      ASD concordance was 31% for DZ and 88% for MZ twins
      Female 100% and male 86% MZ twins were concordant

      3. Bohm and Stewart 2009 Brief report: On the concordance percentages for autistic spectrum disorder of Twins.

      Suggests that looking at whether the MZ twin is monochoronic or dichoronic could provide more information about factors affecting autism in the 1st trimester. The ratio of MZ-MC/MZ-DC is ~2:1 (60%) about the rate for MZ autism concordance. Suggest looking at biomarkers such as testosterone in the amniotic fluid for differences with MC and DC or DZ.

      4. Greenburg et al 2001 Excess of twins among affected sibling pairs with autism: Implications for the etiology of autism.

      166 siblings
      4–5-fold increase for DZ twins and much more than a 10-fold increase for MZ twins. These increases were statistically significant.
      Also 6/172 families had a non-twin sibling with ASD

      • MJ says:

        There was also what looks to be a good review of the various twin studies done over the years published just last month.

        Ronald, Angelica, and Rosa a Hoekstra. 2011. “Autism spectrum disorders and autistic traits: A decade of new twin studies.” American journal of medical genetics. Part B, Neuropsychiatric genetics : the official publication of the International Society of Psychiatric Genetics.

  4. Jack says:

    Will have a look at it when I get to work. The Greenburg paper also spoke about the population sampling issue. Parents with ASD kids usually seek out (especially if you have twins!) studies to try and get answers. Are we really seeing a genetic/environmental affect so pronounce or is it just skewed samples that are confusing the issue.
    Not sure why they are talking about confusing ASD with a neutrotypical. From my point of view it is quite obvious. But will have to read the paper to see what they are on about.

    • MJ says:

      “Parents with ASD kids usually seek out (especially if you have twins!) studies to try and get answers.”

      I guess I forgot to mention that I have twin MZ daughters who both have autism? The thing that is fascinating about them is how alike that are in the core deficits of autism and biological disruptions while at the same time being very different in the actual behaviors of their autism.

      Even more interesting is that, even though they are MZ (proven through genetic testing), they have different CNVs.

  5. MJ says:

    They are Dichorionic-Diamniotic twins.

Share your thoughts:

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s