Could Tylenol be linked with neurodevelopmental disorder, Attention Deficit Hyperactivity Disorder?
A recent study in The Journal of The American Medical Association Pediatrics (JAMA Pediatrics) has suggested a link between ADHD in children and mothers taking Tylenol while pregnant. This has been widely covered by the news media. The study is a correlational statistical evaluation. Of course correlation is not causation. Yet the study has been widely reported as an open ended question. Should pregnant mothers taking Tylenol be worried that they are giving their child ADHD? Given what is known about ADHD I found the premise improbable. I decided to take an intensive look at this research and try to tease out the facts.
Tylenol(AKA Acetaminophen and Paracetamol) is a widely used over the counter as a pain medicine. In my opinion it is just about the safest pain reliever available. Like all medicines it has known risks. It commonly causes liver failure due to overdosing, both accidentally and intentionally( Suicide Attempt). The most common source of overdose is taking multiple over the counter medicines. People are unaware that Tylenol is often an ingredient in many cold and flu remedies. They fail to read the boxes and end up overdosing themselves. If you have a normally functioning liver and you limit yourself to 2-3 grams per day it is probably the safest medication that you can take bar none. It is given to newborns and pregnant mothers. So it is quite a surprise that there is correlation between ADHD and intra-partum tylenol dosing.
The Actual Research:
Layman’s Synopsis- It was a large Danish study. Involving 64000 children and their mothers. They had multiple lines of evaluation. Parent response questionnaires, the danish hospital registry, and ADHD medication prescriptions were all tracked. At least half the mothers took tylenol while pregnant. They found a signficant correlation with first trimester usage of tylenol, and frequent dosing throughout all three trimesters with higher rates of ADHD. They tried to control for maternal inflammatory disease, infection during pregnancy, the mother’s mental health problems, or other potential confounders.
There are problems with this study but I think they did a reasonable evaluation. One question that plagues me is, why did they look at this aspect of ADHD? In the abstract they cite concerns over Tylenol and hormone regulation. Sure there are some theoretical models that predict that problem, and some high dose animal studies that duplicate that. They are making a complicated and precarious assumption to even come up with this study. They make two complicated assumptions to look at this. Namely that hormone changes from Tylenol affect brain development and that in turn relates to ADHD. They are drawing a lot of lines together that do not have a solid foundation. In my estimation the only true link is the reverse. They looked at Tylenol because it is one of the few drugs thought to be relatively safe intra-partum(pregnant) therefore widely taken. Even that doesn’t answer my question. Out of all the neurodevelopmental issues in childhood why is ADHD the one that is correlated with Tylenol? That makes me suspicious of an agenda but it does not invalidate the work.
There are real problems with their methods. Self reporting tools are a consistent problem with ADHD or Hyper kinetic disorder studies. Although they used a standardized tool, its benefit as a diagnostic tool is weak. Also tracking prescriptions as a measure of disease incidence(frequency) is also problematic. Ritalin in particular can be over utilized and has been criticized as being used as a diagnostic tool. Meaning, I give your kid Ritalin and he/she gets better therefore he/she has ADHD. Successful medication treatment is not truly a comprehensive way to diagnose this condition. ADHD is a behavioral disorder which by its nature is a very difficult diagnose even in expert hands. Using ADHD medication prescriptions as a method for determining prevalence of ADHD is very flawed.
The study tried to control for maternal mental health issues. Their methods were self reporting, intra-partum eval, and statistical controls. This is a relatively weak control for a large confounding variable. Research clearly shows that there is a genetic/environmental link in ADHD. How much is environment and how much is genetic is unclear. One thing is clear, home parental mental health is a large variable to exclude and this study has not done that very stringently.
My final significant objection is statistical. Although using large numbers is very good to generalize results, it causes a problem. The P value to big to fail problem. Meaning, with a very large sample, the standard error becomes extremely small, so that even minuscule distances between the estimate and the null hypothesis become statistically significant. TO laypersons- The more subjects you have the more likely you will produce a statistical anomaly that falsely positive.
So am I ready to call BS on this bit of widely reported research?
No I am not. It has some very compelling evidence. It was a large study and even though there are a lot of confounding variables, they did a reasonable job trying to control them.
The most compelling evidence is in the dose dependent response. Seeing a correlation with dose and disease is fairly compelling. Although compelling in is not causation. It is a argument to do more research. It is not compelling as a basis for deciding drug safety. Especially given decades of safe use and the complete uncertainty about the causes of ADHD.
This is not a study that should be creating the type of fear mongering I have already begun to see on the naturopath/Alt.Med crank blogs and websites.
This study is interesting, and it is strong enough to recommend a more controlled and focused study. It is not impossible that Tylenol is some how affecting brain development, and that the exposure results in long term behavioral issues. It is just improbable.
It is improbable because ADHD and other Hyper Kinetic disorders are not a single source disease. There seem to be many environmental, and genetic links. It is a improbable chain of reasoning to say that Tylenol affects brain development intrapartum, and that the change is long lasting, and that it results in specific behavioral changes years later with no other discernible cognitive effects.
Improbable not impossible.
So when you have a weak study correlating(not causing) a complication from a medication used in pregnancy, you always take a look at it. You do not jump to blanket warnings or label changes, especially if the drug has a generational history of safety. I recommend you look at this using a Bayesian anaylysis (in my opinion far more accurate). You will see that this study cannot possibly outbalance the prior evidence of safety. More research is needed to define if this is even a problem.
A common argument would be to err on the side of caution, but that can have unintended negative consequences as well.
Mothers may falsely believe that Tylenol is dangerous. Causing everyone to start thinking that taking OTC NSAID’s are safer. Hysteria could easily develop and people would start avoiding Tylenol for all pediatric conditions, not just pregnant mothers. NSAID’s are not dangerous either, but clearly they have a higher risk profile for the very young compared to Tylenol.
Plus, blaming a mother’s need for a pain control medication in the past may result in maternal guilt over a child’s current problems. By assuming that Tylenol caused a child’s ADHD you are pointing at the mother, as the cause. Resulting in many negative secondary effects for a family dealing with a behavioral disorder. You are making current and future pregnant women struggle with severe pain without any medication options.
Just a few ways reactionary fear could damage people. All to “err on the side of caution.”
Rule of thumb for pregnant moms. There is a chance that everything you do may affect the future development of your child. It is not limited to medication. That is a very heavy responsibility, so take it seriously. Don’t go it alone, and don’t fall for the naturalistic/alt med fallacy. Medications are tested and researched. Yes we find the occasional proverbial “warts” on the treatment. At least we are looking. What you don’t see in the Alt. med research is any indication of any problems. They expect you to assume that everything they do is perfectly safe because it is “Natural”. Utter nonsense, no treatment is perfectly safe.
My advice for pain control. Avoid medications if you can, not just Tylenol, all medications. Try to use massage therapy, and physical therapy to deal with most general pain complaints. Check with your OB before taking any treatment. Never take any supplement, vitamin, or medicine unless it is under you OB doctor’s direct advice. Always take the smallest dose for the shortest time. Stay away from untested/regulated/and poorly researched alternative medicine. I can tell you what the risks are for your baby with proven medical care. I cannot even guess what the risks are with Alt. Med. Please don’t treat a pregnant mom like she is a selfish murderer because she took some Tylenol in front of you. Tylenol and its generic versions are still the safest pain treatment that can be offered to a pregnant mommy. Pregnant mothers, do not under any circumstances take a Non Steroidal Anti-Inflammatory like OTC ibuprofen. That medicine has proven risks to your unborn child.
Most risks for pregnant women are well known. Some are overblown fear mongering. This study is interesting, but as tool to determine cause and effect it is useless. Warnings that you may hear about Tylenol and ADHD amount to no more than Fear Mongering and are not medical science.
Disclaimer: This post is my personal opinion, it does not reflect the opinion of: my practice, my partners, hospital affiliations, Brian Dunning or my academic affiliations. It is for informational/educational purposes only. It is not intended to replace personal medical evaluation and discussion with your healthcare provider.