Here’s a study that got a lot of attention yesterday.
I am sure the authors were proud of themselves. However they might have expected a big splash, such a popular diagnosis, Attention Deficit Disorder, Hyperactivity Disorder. everyone knows it, everyone is conscious of it.
And of course Tylenol is ubiquitous, used, as recommended by the Medical people in pregnancy to avoid any concerns about Non steroidals, that are a concern in later phases of pregnancy.
So the perfect storm, all you have to do is grind the numbers, show a small association between Tylenol (acetaminophen) and take a victory lap after you establish that acetaminophen in pregnancy creates a risk of Hyperactivity, Attention Deficit, Behavior disorders in children. Ignore the obvious problem that these are soft diagnostic entities, subject to a lot of debate and criticism.
So JAMA Pediatrics gets in the hands of the press. The audience is guaranteed–it’s about mothers and babies. THE CHILDREN.
And here come the media.
So this article caught the attention of our chief resident at the place where I work, and he circulated it for our residency programs to review.
We do research so we have to know how to analyze research and design good research.
Here’s the important part of the report in the Journal JAMA Pediatrics
RESULTS More than half of all mothers reported acetaminophen use while pregnant. Children
whose mothers used acetaminophen during pregnancy were at higher risk for receiving a
hospital diagnosis of HKD (hazard ratio = 1.37; 95%CI, 1.19-1.59), use of ADHD medications
(hazard ratio = 1.29; 95%CI, 1.15-1.44), or having ADHD-like behaviors at age 7 years (risk
ratio = 1.13; 95%CI, 1.01-1.27). Stronger associations were observed with use in more than
1 trimester during pregnancy, and exposure response trends were found with increasing
frequency of acetaminophen use during gestation for all outcomes (ie, HKD diagnosis, ADHD
medication use, and ADHD-like behaviors; P trend < .001). Results did not appear to be
confounded by maternal inflammation, infection during pregnancy, the mother’s mental
health problems, or other potential confounders we evaluated.
CONCLUSIONS AND RELEVANCE Maternal acetaminophen use during pregnancy is associated
with a higher risk for HKDs and ADHD-like behaviors in children. Because the exposure and
outcome are frequent, these results are of public health relevance but further investigations
Hazard Ratio is similar to Relative Risk or increase in the rate of the problem that is identified as the adverse effect and toxic end point.
In this case endpoint is diagnosis of ADHD or ADHD type behaviors at age 7.
That’s called a soft endpoint, since ADHD is a set of signs and symptoms subject to judgement.
The Hazard ratios were below 2, which is inadequate to establish proof of causation. The Hazard Ratio for behavior barely escaped 1.0 at 1.01 which would have meant that the confidence interval included ZERO. That means that one of the endpoints, behavior is so weak a correlation as to be a joke–get that. As for the other endpoint I can think of one major confounder not commented on, Tylenol use as a marker for more active medical user–higher level of medical awareness and hyper alertness, even, dare i mention it–anxiety and neurosis–now imagine that such a thing might impact ever so slightly the inclination of a mother to take a kid to the pediatrician and press for a diagnosis like ADHD. Maybe?
Their list of confounders doesn’t come close to describing all the confounders in a study like this. Endpoints, exposures, likelihood of Tylenol use by selection and the likelihood of diagnosis of ADHD as marked by increased rate of maternal Tylenol use in pregnancy.
I don’t know, but I do know that a Hazard Ratio of 1.13 for behavior is a small association and 1.37 for Hyper kinetic disorder is small and, if there is no double counting, the 1.29 Hazard Ratio for ADHD meds is small too. The confidence intervals are not too bad, but that is derivative of the size of the study, as is the p value, which goes with the confidence interval (CI).
You betcha the authors are right when thy suggest further study, even though these observational studies are always mostly just hypothesis generating, particularly in the range of small associations.
If the rate of ADHD was found to be 150 percent higher as compared to 29% or Hyperkinetic disorder was 150% or 200 % higher instead of 37 % then, in this kind of study the strength of evidence would be more impressive, even if not dispositive on the question.