By William Parker, Ph.D., Duke University Medical Center
What was the study about?
In 2008, a small study by Stephen T Shultz showed literally a 20-fold increase in the risk of regressive autism in children exposed to acetaminophen (paracetamol), the most commonly used medication in children. Although the results were statistically significant, and although the results suggested that acetaminophen was a causal agent (rather than simply another effect of autism or whatever caused autism), the study was immediately and publically criticized. Eight years later, I began work with some of my colleagues at Duke and with a team at Harvard, reviewing a broad range of evidence looking at the potential role of acetaminophen in the induction of autism.
What did we find?
What we found and published after more than a year of work was extremely unnerving, suggesting that the original 2008 study by Shultz was possibly correct in all regards, and pointing strongly to a need for public education regarding what is known regarding the role of acetaminophen in the development of autism.A dozen very compelling papers looking at the adverse effects of acetaminophen on the developing human brain have been published since 2008, most in the last four years. We also found an additional dozen lines of evidence, including work in laboratory animals, pointing toward acetaminophen use as damaging to the developing brain. Importantly, no study has ever suggested that acetaminophen is safe for neurodevelopment. We found no controversy about the topic. But sadly, each time new evidence is published, the medical community has disregarded that evidence. As our review was being published, I spent time catching up with experts who have publically assured the world that acetaminophen is safe for the developing brains of babies. In every case, no expert would stand behind their claim. When presented with the currently available evidence, all experts either refused to comment (even privately) or else explained that they had been quoted as an expert based on their clinical experience, not based on any knowledge of the peer-reviewed literature regarding the dangers of acetaminophen. They were basically caught off guard, being asked about accepted clinical practice at their institution without being aware that dozens of studies published over the last 10 years are all suggesting that the standard practice is tragically misguided.
The bottom line is that acetaminophen exposure after birth, in combination with oxidative stress (a result of the body’s inability to keep up with excess inflammation) is very likely a cause of autism and perhaps other neurodevelopmental disorders such as ADHD and lowered IQ. It has been argued that signs of autism can be seen before birth, and therefore post-birth exposures cannot be responsible for the induction of autism. However, the signs of oxidative stress are probably evident before birth, and the actual damage leading to many cases of autism may be induced only after birth when the mother’s liver was not available to help with the metabolism of acetaminophen.
What should parents and clinicians know?
Parents need to be aware of some key issues. First, acetaminophen is not intended to be used for colds, and studies suggest that it does not work well, if at all, for infant pain. Second, the American Academy of Pediatrics recommends that most fevers do not need to be treated because they are not dangerous. Fevers are actually a defensive response from the immune system and help fight disease. The Academy has published some helpful information about when to treat a fever, and how to help treat a fever without using drugs. (https://www.healthychildren.org/English/health-issues/conditions/fever/Pages/Treating-a-Fever-Without-Medicine.aspx) Third, parents need to be aware that well-intended family members, medical professionals, and other caregivers, being unaware of current evidence, may give their child acetaminophen. Finally, it is possible that the nutritional supplement N-acetyl cysteine (NAC), an amino acid derivative available in nutrition stores and used to counteract acetaminophen overdoses in adults, may effectively block adverse side effects of acetaminophen if given at the same time as the acetaminophen. But this idea is speculative, and as far as is known, there is no absolutely safe way to use acetaminophen in infants and small children.
The nitty-gritty details
The key points we identified in that review are enumerated below. In addition, the list below contains some updates to the review, provided to me in large part by Dr. Ann Bauer, who began a study of this topic shortly after the original 2008 paper was published.
- Twelve peer-reviewed studies now demonstrate that acetaminophen exposure is neurotoxic to the developing brain of humans. All but one of those studies evaluated the effect on children following exposure as a fetus. Effects of exposure evaluated by various groups included lowered IQ, autism, and behavioral problems. The studies were typically controlled for confounding factors, indicating that acetaminophen use was a cause of disease and not an effect. However, the original 2008 study is still the only one evaluating the effect of exposure after birth, and that study found more than a 60-fold greater risk (literally) of developing autism following exposure after birth than studies looking at the risk of exposure before birth.
- Laboratory animals develop permanent brain damage characterized by problems with social function when they are exposed to acetaminophen during an early period of brain development that corresponds to human infancy. Damage can be induced in the developing fetus during pregnancy, but the dose required is far more than levels normally seen in humans. But after birth, the amount that it takes to cause permanent brain dysfunction in laboratory animals is less than the amounts that human infants typically receive (on a per weight per day basis).
- A very wide range of factors, all associated with inflammation and oxidative stress, are associated with an increased risk of autism. Thus, it appears that whatever is inducing increased rates of autism probably acts in concert with inflammation and oxidative stress. This conclusion is consistent with the hypothesis that acetaminophen exposure to infants and children after birth is a leading cause of autism. As an aside here, we know that inflammation is on the rise in Western society, and that autism is an inflammation-associated disorder. We have worked with an epidemiologist, Cindy Nevison, and found very strong evidence that the prevalence of autism is truly increasing. (Increasing awareness, changing diagnostic criteria and other factors do not account for the measured increases in prevalence.)
- Margaret McCarthy, chair of Pharmacology at the University of Maryland, has elucidated the mechanism by which acetaminophen damages the developing brain, and has explained why acetaminophen is more dangerous to males than to females. This potentially accounts for the increased propensity for males to acquire autism compared to females.
- Acetaminophen substantially alters brain chemistry of human adults and temporarily impairs awareness of social issues in adult humans. Although the drug could naively be considered to be anti-inflammatory, exposure to the drug actually causes inflammation, even in healthy adults, and depletes metabolic components needed to deal with inflammation. Further, the drug is immunosuppressive, blocking an important immune function (the febrile response).
- No study has ever shown acetaminophen to be safe for the developing brain of a human or a laboratory animal. A major manufacturer of acetaminophen in the US acknowledges that the drug was never proven safe for brain development when used during pregnancy or in childhood. All safety tests that have been conducted were performed with the assumption that any side effects would be acute in nature (e.g., bleeding or organ damage soon after administration of the drug). One very recent study has suggested that acetaminophen use during pregnancy may be safe if used to treat fevers only, but use to treat fevers is not the most common use of acetaminophen during pregnancy.
- A very wide range of circumstantial evidence points toward the potential for acetaminophen to be at the root of many if not most cases of autism. Alternative explanations are lacking for much of this circumstantial evidence:
- The epidemiology of autism matches the historic use of acetaminophen, with both rising sharply in the early 1980s following the identification of Reye Syndrome as a possible side effect of aspirin use in children. (See note above regarding the view that the increased prevalence is a real increase rather than an aberration caused by issues concerning awareness and diagnosis.) This evidence alone is considered very weak, but it is necessary for this hypothesis to go forward.
- The qualitative nature of autism apparently shifted from infantile to regressive at the same time that acetaminophen use in infants and children became more popular, in the early to mid-1980s.
- A wide range of inflammatory conditions that affect the potential to tolerate oxidative stress are associated with autism. It is expected that these conditions will adversely affect the metabolism of acetaminophen. A notable exception is cystic fibrosis, a condition of oxidative stress not known to be associated with autism. Cystic fibrosis is also unusual as a condition of oxidative stress that actually enhances rather than impairs metabolism of acetaminophen.
- The number of possible environmental suspects for the induction of autism at a level that would account for the epidemic is dwindling, leaving acetaminophen use in infants and small children as one of the few remaining suspects. A good suspect is one which is newly (since 1980) and widely introduced into the population, and which should be associated with a profoundly large (10-fold or greater) increase in the risk of autism. Most suspects tested to date have shown moderate and sometimes variable (dependent on the population studied) associations with autism.
- Genetic variants associated with autism, many associated with oxidative stress, likely influence the metabolism of acetaminophen.
- Circumcision has been identified as a significant risk factor for autism, and the only reasonable explanation put forth for this observation is that acetaminophen use during the procedure is probably inducing autism.
- History demonstrates to us that it is a mistake to ignore the observations of parents when it pertains to autism. Many parents who have historically blamed vaccines for their child’s autism may have actually observed the effects of acetaminophen. In the 1990s, vaccination was advertised as the number one indication for use of acetaminophen.
- Levels of acetaminophen use potentially explain previously unexplained epidemiologic observations, such as the very low rates of autism in Cuba, the very high incidence of autism in South Korea, and the culture-dependent variation in autism among the Jewish population.
- The idea that acetaminophen use, particularly in infants and small children, is responsible for many if not most cases of autism is an attractive hypothesis, as it satisfies Occam’s Razor in being a simple explanation that explains a wide range of observations.
What’s unknown and what should be done next
Given the preponderance of evidence described above, we conclude that parents and doctors should know all available facts so that they can make informed decisions about the use of acetaminophen. At the same time, we note that the amount of damage caused by current levels of acetaminophen use in babies and children is unknown and that an incisive study addressing this issue needs to be done as quickly as possible, with all available resources at its disposal. Despite the fact that acetaminophen is the most popular drug used in infants, children, and pregnant women, neither the benefits nor the risks of using the drug in those populations have been thoroughly evaluated. The cost of an incisive study would be less than the average cost of raising and caring for three individuals with autism. Just three.