MMR and false syllogisms

By AnthonyLast updated: Wednesday, November 26, 2003 • Save & ShareLeave a Comment

One of the issues surrounding MMR is that of causality, neatly illustrated and described by the following figure in an article from Prescriber [Registration required] written by Paula McDonald (a former Consultant in Communicable Disease Control in Cheshire & Wirral).

Some of these syllogisms may be plausible to some patients

Aristotle’s logic was a great legacy for mankind. One of his ideas was the concept of syllogisms, where if certain prepositions are met, something distinct will arise from necessity. However, one can have false syllogisms (there is even a silly syllogism generator). The above figure illustrates the usual example of the horse being classified as a cat. McDonald has addressed the MMR debate as a false syllogism.

The implied argument is ‘These children have autism. These children had MMR vaccine before they developed autism. Therefore, MMR vaccine caused the autism’.

A similar argument could be used to implicate teddy bears as a cause of autism: ‘These children have autism. These children had teddy bears before they developed autism. Therefore, teddy bears caused the autism’. Clearly, this is ridiculous, and it remains ridiculous if the teddy bears are replaced with nappies, parents, watching Thomas the Tank Engine on television or any other features of childhood. However, it sounds more convincing (although still invalid) if the teddy bears are replaced with exposure to electromagnetic radiation, food additives, pollution or anything else about which there is public concern.

What is wrong with this argument? The logical fallacy contained in the argument is known as ‘Post hoc ergo propter hoc’, which translates as ‘After this, therefore because of this’. An example commonly used to illustrate this is the assumption that the sun rises in the morning because the cock crows.

In research terms, the crucial distinction is between association (two things found together) and causality (A causes B). Association may

  • occur by chance
  • be spuriously identified because of poor research methodology, such as selection bias or the failure to match cases to controls properly
  • occur because both A and B are independently linked to a third factor. For example, countries with high levels of television ownership have in the past been found to have high levels of breast cancer. This is not because television ownership causes breast cancer, but because both are independently linked to affluence.

This is an article well worth reading. Her conclusion?

We have seen that using the Royal Free study as a basis to argue that MMR causes autism is no more or less logical than arguing that teddy bears cause autism. Looking systematically at all of the evidence for causality, it is clear that the incidence of autism or bowel disorder is not increased by giving MMR.

Our task now is communicating this to parents who are choosing not to give their child MMR, or who are choosing single-antigen vaccines in an attempt to avoid a nonexistent risk.

Filed in Vaccines

Comments are closed.