Correlation does not Equal Causation

The most frequent problems with the arguments about vaccines and autism involve a logical fallacy called the correlation implies causation fallacy, which is a fallacy by which two events that occur together are claimed to be related by cause and effect. For instance:

- Teenage boys eat lots of chocolate.
- Teenage boys have acne.
- Therefore, chocolate causes acne.

Chocolate could cause acne in teenage boys. It may not as well. Acne may be caused by hormones, or stress, or other factors that are not identified in combination. Just because teenage boys eat lots of chocolate and have acne does not mean the two facts are causally related, that chocolate causes acne, or conversely that acne causes boys to eat chocolate. Teenage boys might just like chocolate and have acne.

This fallacy shows up in various fashions in this debate. For instance, Generation Rescue (an anti-vaccination organization) ran a full page advertisement in multiple papers a few months back in which the primary argument was that in 1966, when children received 10 vaccinations, autism rates were 1 in 10,000; today, when children receive 36 vaccinations (though many of those shots are for the same disease), autism rates are 1 in 150; therefore, vaccinations are causing neurological injuries including autism. Without getting into all of the various flaws of the ad such as the inaccuracy in their claims about past autism rates, the worst flaw is the blatant correlation implies causation fallacy – just because apparent autism rates have increased at the same general time as vaccinations does not mean the two are causally tied. Correlation does not equal causation. And, this holds true related to many other increasing exposures in our modern world. There have been huge changes in society in the last 40 years, including dramatic increases in the number of exposures to potentially noxious substances like pesticides and antibiotics, the average number of hours that children watch TV, the grams of simple carbohydrates in the average diet, and the amount of vibration to which the average person is exposed. All of these are correlated with increases in autism prevalence, but there is no scientific support for the notion that any of them in isolation, or even in combination, cause autism at this point in time.

Another common example of this fallacy involves the temporal association between regressive autism and the first MMR vaccination that has been so hyped in particular in the UK. There are tens of thousands of parents, and likely millions of others, who have become convinced autism is caused by the MMR vaccine largely because of a reported temporal association between the two – the MMR vaccination shortly precedes the onset of regressive autism symptoms in some children. As a result of this temporal link, many parents have concluded the MMR vaccination must have caused the autism. There are all kinds of problems with this theory, such as the fact that most cases of autism are not regressive, that even regressive cases typically show symptoms before the regression and the vaccination, and that the percentage of parents who associate autism onset with the MMR vaccine is vastly overblown by the media since the parents making such claims tend to garner the lion’s share of the controversy-seeking media’s attention – the parents without such a theory tend to not blog, to avoid television crews, to quietly work towards helping their autistic child.

In addition, a glaring problem with this reasoning is the correlation does not equal causation fallacy. Just because in a limited number of children the onset of regression is close in time (in the parents’ memory) with this vaccination does not mean the two are causally tied. There are all kinds of exposures and experiences that happen in a child’s life around 18 months, from having a little brother, to moving cities, to starting school, to having an infection. Some of these are bound to occur around the time autistic regression starts. In fact, the studies discussed below show a much greater parental association of household stressors with regression that with vaccines. That doesn’t mean household stressors cause autism either. It also doesn’t rule them out.


Your Contact Information

Your Feedback