Yesterday’s study linking paracetamol use during pregnancy with behavioural problems in childhood has hit news headlines around the world, and this morning ranked first among Australian health news stories, according to Google trends.
[Clarification: paracetamol brands include Panadol and Tylenol. In the US, the chemical name for the drug is acetaminophen.]
What is the fuss about?
A study published yesterday in JAMA Paediatrics analysed data from 7796 mothers from Bristol, England, who were part of a prospective birth cohort study in 1991-2. The self-recorded data included maternal use of paracetamol during pregnancy (at 18 and 32 weeks) and five years later, and the presence of behavioural issues when the child was seven years old.
The study found an association between antenatal paracetamol use at 32 weeks gestation and behavioural problems for the 7-year-old children, as measured by a high score on a conduct disorder scale (RR 1.42, 95% CI 1.25 – 1.62) and hyperactivity scale (RR 1.31, 95% CI 1.16 – 1.49).
A Relative Risk (RR) of 1.31 means that 7-year-old children in this study were 31 per cent more likely to score highly on an ADHD scale if their mother said she used paracetamol during pregnancy. Smaller RRs were found for paracetamol use at 18 weeks gestation.
The perfect media storm?
Because paracetamol is so widely used, and particularly as it has long been considered safe during pregnancy, there was never any doubt this study headline would be a blockbuster. And indeed, any suggestion of in utero harm from such a commonly used medication certainly warrants further research.
This morning’s news coverage in the major dailies has been fairly reasoned, with most reporters balancing the authors’ quotes about their findings having “important implications on public health advice,” with cautions from other commentators warning that the size of the absolute effect was small, and explaining the most crucial point of all: association is not causation.
Some social media was not so nuanced, with some Facebook discussions already contemplating suing paracetamol manufacturers for their child’s ADHD.
Association, association, association.
The study design was never going to produce evidence strong enough to change any clinical recommendation, as it was not looking for any causative effect. There are plenty of other plausible explanations for the observed association.
Not least, the presence of infection during pregnancy could result in both paracetamol use and developmental harm. And the group of mothers who used paracetamol also used more cigarettes and alcohol, albeit not at rates high enough to alone explain the results.
The authors made some attempt to adjust for potential confounding factors, but this is difficult to do, and you would really want to see a number of similar other studies before we could even be sure there is an association, let alone looking at causation.
Other studies would also help rule out ‘cherry picking’ the results that support the hypothesis, while ignoring those that don’t. The following example from this study illustrates the point nicely:
The authors happened to collect data on the mother’s partner’s use of paracetamol when the child was five years old. This served as useful comparative data, because clearly, dad popping Panadol could not possibly cause ADHD in his five-year-old child.
Those results actually showed a 38 % increased risk (RR 1.38, 95% CI 1.02 – 1.88) of conduct disorder in the child if dad used paracetamol. It is clear to everyone that this association could not possibly be causation, because it is biologically impossible, yet the mothers’ antenatal data was actually fairly similar, and it made headlines around the world.
The geeky message is that this finding now encourages the next phase of research into causation. At the same time, we also need to replicate the original study to ensure the association wasn’t just cherry picking the headline from otherwise negative results.
The message for the ordinary citizen is to remember the general caution that medication during pregnancy should be kept to a minimum—and this includes paracetamol. This study adds a small question mark to paracetamol’s safety, but does not actually suggest it is unsafe.