select search filters
roundups & rapid reactions
factsheets & briefing notes
before the headlines
Fiona fox's blog

expert reaction to study looking at paracetamol in pregnancy and autism, ADHD and other developmental disabilities in children

Research, published in JAMA Psychiatry, reports on the risk of attention-deficit/hyperactivity disorder (ADHD), autism, and other developmental disabilities in children from taking paracetamol during pregnancy. 


Dr Grainne McAlonan, Clinical Reader in Sackler Centre for Translational Neurodevelopment and Forensic and Neurodevelopmental Sciences, King’s College London, said:

“The paper reports an association between paracetamol exposure around birth and risk of neurodevelopmental conditions such as autism spectrum disorder and ADHD.  Associations are not necessarily causal.  As ever more research needs to be done.  There are a few cautions to be aware of before accepting the observations at face value.

“Firstly, the measures of paracetamol only reflect exposure around the time of birth, not during pregnancy, and all the neonate samples investigated contained paracetamol.  This is itself strikes me as rather unusual.

“Secondly, although the authors looked at different levels of paracetamol in the babies’ cords and mothers’ blood samples, we do not know the amount of paracetamol taken by mothers.  For example, the amount originally taken may be similar but each individual’s ability to break down paracetamol may be what is variable.  Indeed it has been previously reported that the activity of the enzymes needed to break down paracetamol varies greatly between individuals (babies and adults) and is influenced by many factors including other medication, sex, alcohol, smoking, and very importantly by the age and the genetics of the individual.  So, a single dose of paracetamol might be quickly eliminated by one person and much more slowly by another.  Understanding how individual mothers and their babies breakdown paracetamol is a missing part of the picture.

“A thought experiment: let’s assume that the mothers taking paracetamol around birth take similar amounts within the prescribed limits.  It is possible that babies were exposed to a similar dose but those who ended up with developmental difficulties, especially ADHD, did not eliminate the paracetamol as efficiently.  It may be that whatever increased the risk of ADHD in these children might also have made them slower at eliminating paracetamol.  In other words an underlying factor predisposing to both neurodevelopmental difficulties and slower breakdown of paracetamol, rather than the paracetamol itself, may be behind any patterns.  (The paracetamol hanging around in the system may or may not additionally influence outcome of children predisposed to ADHD.  We just don’t know.)  This is one possible interpretation of the relationship reported in this paper.

“Thirdly, for whatever reason, more significant associations with paracetamol are observed for ADHD than autism spectrum disorder.

“Fourthly, to enter the follow-up needed for this study, participants were recruited at six months of age from BOTH primary and speciality paediatric services of the host hospital.  It is not clear to me what proportion came from the speciality services.  This may be important because if a child is already receiving specialist paediatric input at six months of age it suggests that this is not an everyday community sample.  Those needing specialist care early in life may have underlying issues and may potentially be more vulnerable to later difficulties.  Indeed the rate of ADHD in this study sample is very high – 25% – the population rate should be more like 9%.  It’s hard to imagine that this jump is substantially explained by paracetamol exposure around birth.  I’d like to know how many children came from the specialist services and what the associations between paracetamol and outcomes were in children recruited from routine primary care services.”


Prof Andrew Shennan, Professor of Obstetrics, King’s College London, said:

“Paracetamol is a recommended drug for use in pregnancy and will be commonly used.  This type of uncontrolled study does not imply paracetamol use causes autism, as the reason for taking paracetamol may be the issue rather than the drug itself, and a mechanism for it to cause harm is not clear.

“The babies with autism were significantly more likely to be preterm and their mothers were overweight, stressed and more likely to smoke – all significant risk factors for babies.  The authors have tried to account for this in their analysis but further work is needed to show if paracetamol use is dangerous.  Women should still use paracetamol if recommended by their health care professionals, but this work will provide a focus for more research.”


Prof Stephen Evans, Professor of Pharmacoepidemiology, London School of Hygiene & Tropical Medicine, said:

“This study in the US is noted by the authors to have various strengths and some limitations.  Its main strength is that it measured paracetamol concentrations in cord blood rather than asking mothers whether they took paracetamol.

“There are some unusual features which may reduce the strength of the conclusions.  Firstly, in this cohort of children, more than two thirds have some form of neurodevelopmental disorder.  It is supposed to be essentially a random sample of children but this very high prevalence of disorders both makes its generalisation to other settings difficult, and may also cast doubt on whether it has been carried out in an unbiased way.

“The paracetamol concentrations are shown in the figures with nice smooth curves, but the underlying data do not have this feature and there has been statistical processing to produce the results.

“A further unusual finding is that every mother and child pair had exposure at some level of paracetamol.  This again suggests there is something strange about the cohort.

“Although it is a study that involves follow-up through childhood and later, the time points at which the developmental deficits occurred are neither recorded nor displayed.  The methods of analysis are not the usual ones for this type of study design.  They do not present the key data that are relevant to mothers; that is if they don’t take paracetamol, what are the chances of their child having a neuro-developmental disorder (NDD), and, if they do take paracetamol, what extra chance in absolute terms will their child have of an NDD?

“These statistical limitations may mean that the results are not as strong as is claimed, and the authors acknowledge that it is not possible to be sure it is really the paracetamol exposure that is the cause of the developmental problems.

“This study alone isn’t enough to suggest any change in recommendations or to give any health advice.  In general, advice has always been to avoid or limit most drugs in pregnancy if possible, but some drugs are required for a mother’s health.  Alternative painkiller drugs have not been shown to have no problems, so it is difficult to offer advice on them.  Avoiding paracetamol when it is not needed is sensible and has always been the case, but millions of women with perfectly normal children will also have taken paracetamol during pregnancy.  The results of this study should not raise anxiety in pregnant women.

“(It may be noted that paracetamol is the international name, while acetaminophen is the name used in the US for the same drug.  It is one of the few drugs for which the US has its own name rather than using international naming conventions).”


Prof Emily Simonoff, Professor of Child & Adolescent Psychiatry and Head of the Department of Child & Adolescent Psychiatry, Institute of Psychiatry, Psychology and Neuroscience, King’s College London, said:

“This study examines whether there is an association between maternal acetaminophen (paracetamol) use in pregnancy and developmental disorders, including ADHD and ASD, in offspring.  The study aims to overcome limitations of self-report studies through direct measurement of acetaminophen levels and metabolites in cord samples.  However, because of the relatively rapid metabolism of the drug, these levels, by the authors’ account, this will have captured paracetamol use during the three hours prior to sampling, i.e. during the labour and delivery, presumably.

“However, despite this very narrow window of measurement, the authors do not appear to have examined the role of, for example, other perinatal factors that may have influenced whether or not such simple pain relief was used.  The rates of ADHD and ASD are much higher than would be expected based on usual prevalence rates (ADHD 5-7% and ASD 1-2% – and the upper limits here are generous) and it is unclear why this should be the case but it raises issues of a biased sample, and it is not clear how this might affect the association reported here.

“There is no very strong hypothesis about why the association should occur for ASD and ADHD but not other developmental disabilities.  Hence, while this is an interesting study, there are significant limitations in the methods and results that mean the findings should be interpreted with caution.  More research is needed before clinical practice is altered.”


Prof Jean Golding, Emeritus Professor of Paediatric and Perinatal Epidemiology, University of Bristol, said:

“In this study the information uses the blood from the umbilical cord, collected at birth, and assayed to identify traces of paracetamol.  The authors compare the results between mothers whose children developed ADHD, those who developed an autism spectrum disorder, those with other disorders and a normally developing group.

“This is a curious study and is difficult to interpret for a number of reasons: (1) the half life of paracetamol in adults is 2-3 hours, so the medication would need to be taken relatively soon before the baby was born to be detected; (2) paracetamol was found in all the blood samples, implying that all mothers had taken paracetamol before delivery; consequently there was no control group of children not exposed to paracetamol before delivery; (3) no information was given as to the reasons for the medication being given; (4) there were differences in the blood concentrations of paracetamol metabolites between the different groups – but it is unclear as to whether such differences are biologically meaningful.  For example do they indicate that the children with higher cord blood levels before delivery actually were exposed to a larger dose, whether they have a different genetic metabolism, or whether the paracetamol was an indication of the mother having a more painful labour and delivery?

“In conclusion, I am afraid I cannot see that this study helps us to deduce whether paracetamol in pregnancy has the adverse effect on the child’s behaviour that many observational studies suggest.

“In general there is accumulating evidence from other observational studies that paracetamol in pregnancy is associated with behaviour problems in the child, especially of hyperactivity – but the information used to come to that conclusion is usually the report of the mother, and the claim is made that it may be the reason for the woman taking paracetamol, rather than the medication itself, that creates the behavioural problem.”


‘Association of cord plasma biomarkers of in utero acetaminophen exposure with risk of attention-deficit/hyperactivity disorder and autism spectrum disorder in childhood’ by Yuelong Ji et al. was published in JAMA Psychiatry at 15:00 UK time on Wednesday 30 October 2019.

DOI: doi:10.1001/jamapsychiatry.2019.3259


Declared interests

Dr Grainne McAlonan: “No relevant conflicts of interest.”

Prof Andrew Shennan: “No conflicts.”

Prof Stephen Evans: “I have no conflicts of interest to declare.”

Prof Emily Simonoff: “Conflict is: Man, K.C., Chan, E. W., Ip, P., Coghill, D., Simonoff, E., Chan, P. K. L., W. C.Y. Lau, Schuemie, M. J., Sturkenboom, M. C. J., Wong, I.K. (2018) Parental antidepressant exposure and the risk of attention-deficit hyperactivity disorder in children: A systematic review and meta-analysis.  Neuroscience & Biobehavioural Reviews 86:1-11.”

Prof Jean Golding: “I have no interests to declare.”

in this section

filter RoundUps by year

search by tag