1. Introduction

Acetaminophen became widely available in the 1950s and eventually became one of the mostly frequently used medications for the treatment of pains and fevers. By the 1970s, a consensus was reached in the medical community that acetaminophen is safe and effective when used as directed in the pediatric population ^[1][2]. Following this consensus, the discovery of an association between aspirin exposure and Reye Syndrome in the early 1980s led to increased dependence on acetaminophen for treatment of pediatric populations, with both prescription and over-the-counter formulations available.

As early as 1999 it was noted that many children with ASD have difficulty metabolizing acetaminophen ^[2][3]. Although it had long been understood that difficulty with acetaminophen metabolism increases the toxicity of the drug ^[3][4], the observation that individuals with ASD had difficulty with acetaminophen metabolism did not, prior to 2008, elicit the idea that acetaminophen might be a source of injury in individuals with ASD. After decades of acetaminophen use in the pediatric population, strongly encouraged by direct-to-consumer advertising by pharmaceutical manufacturers, a case-controlled study published by Stephen Schultz in 2008 involving 81 children with autism spectrum disorder (ASD) found a 20-fold greater risk of regressive ASD when acetaminophen was used between 12 and 18 months of age compared to controls ^[4][5]. Two years later, studies in adult laboratory rats show neuronal cell death with non-lethal doses of acetaminophen ^[5][6]. Five years after the study by Schultz, the first study in neonatal laboratory animals using acetaminophen showed profound neurodevelopmental injury at doses only 2-fold higher than doses administered to human children ^[6][7]. Subsequent evaluation of safety claims found that the consensus of safety was based on the false assumption that acetaminophen is metabolized the same in babies and in children as in adults ^[1][2]. By 2022, mounting evidence that acetaminophen (paracetamol) use in susceptible babies and children is associated with the development of ASD and other neurodevelopmental disorders was overwhelming ^[7][8][9][1, 8, 9]. Based on approximately 20 independent lines of evidence, it was concluded without reasonable doubt and with no evidence to the contrary that acetaminophen administration in susceptible babies and children is a causative agent for the induction of many, if not most, cases of ASD ^[7][8][9][1, 8, 9]. The conclusion was based on (a) studies in laboratory animal models ^{[10][11][12][13][14][15]}[10-15], (b) understanding of the pharmacological mechanisms associated with acetaminophen toxicity [16], (c) connections between ASD, acetaminophen exposure, and human activities such as vaccination ^[4][5] and circumcision [17], (d) associations between acetaminophen administration and ASD during the later stages of pregnancy [18] and in early childhood ^[4][19][5, 19], and (e) associations between acetaminophen use and ASD through time ^[16][20][16, 20].

2. Summary of Evidence

Fourteen associations between acetaminophen use during early neurodevelopment and ASD were evident as of late 2023 ^[7][1]. A summary of those associations, along with additional lines of evidence available at that time ^[7][1] are as follows:

• Association #1: Circumcision of males, often performed using acetaminophen as an analgesic, is associated with a 2-fold increase in the risk for early-onset (infantile) ASD [17].
• Association #2: Acetaminophen-containing products used by South Korean children were repeatedly found to contain amounts of drug exceeding the package label ^[21], and an exceptionally high prevalence of ASD was identified in South Korea ^[22][23].[21], and an exceptionally high prevalence of ASD was identified in South Korea [22, 23].
• Association #3: The popularity of acetaminophen use and the prevalence of ASD was substantially higher in Denmark than in Finland in the mid-2000s (Figure 1).
• Association #4: Ultra-Orthodox Jews ^[24] and Arabs ^[24][25] in Israel have a reported prevalence of ASD less than half of that of other Israelis. Traditional circumcision practices employed by Ultra-Orthodox Jews do not utilize acetaminophen, and circumcision practices in Arab communities take place outside of the neurodevelopmental window sensitive to ASD induction (Figure 3).[24] and Arabs [24, 25] in Israel have a reported prevalence of ASD less than half of that of other Israelis. Traditional circumcision practices employed by Ultra-Orthodox Jews do not utilize acetaminophen, and circumcision practices in Arab communities take place outside of the neurodevelopmental window sensitive to ASD induction (Figure 3).
• Association #5: Analysis of 61,430 babies in the Danish National Birth Cohort found an odds ratio (OR) of 1.3 (CI 1.02-1.66) for ASD associated with postnatal acetaminophen exposure ^[19]. The approach used in the analysis is expected to dramatically underestimate the real odds ratio ^[8]. Other studies have supported this association ^[26][27].[19]. The approach used in the analysis is expected to dramatically underestimate the real odds ratio [8]. Other studies have supported this association [26, 27].
• Association #6: The ratio of regressive to infantile ASD rose at the same time as pediatric acetaminophen use rose [20] after aspirin was associated with Reye’s syndrome [16].
• Association #7: The prevalence of ASD began to increase in the early 1980s, coinciding with the increase in acetaminophen use after aspirin was associated with Reye’s syndrome [16].
• Association #8: The prevalence of ASD has steadily increased [16] as direct-to-consumer advertising [28] and perhaps other factors have driven up use of pharmaceutical products. In 1998, Forbes Magazine reported that the primary manufacturer of acetaminophen in the US spent $250 million on advertising for acetaminophen containing products in the US in one year [29].
• Association #9: Maternal use of acetaminophen during pregnancy is associated with long-term effects that include lower IQ, increased ASD, and increased ADHD in their children ^{[18][19][30][31][32][33][34][35][36][37][38][39][40][41][42]}.[18, 19, 30-42].
• Association #10: Levels of acetaminophen in cord blood are associated with ASD [34].
• Association #11: Acetaminophen given alongside the MMR vaccine but not the MMR vaccination alone was associated with ASD ^[4].[5].
• Association #12: Acetaminophen use during early childhood is associated with a dramatic increase in regressive ASD ^[4].[5].
• Association #13: Many parents believe that their children’s ASD was induced by a vaccine [43, 44]. acetaminophen is frequently used with vaccinations, although vaccinations ^[43][44]. acetaminophen is frequently used with vaccinations, although vaccinations alone do not cause ASD ^[45][46].alone do not cause ASD [45, 46].
• Association #14: Cystic fibrosis is associated with unusually efficient (effective) metabolism of acetaminophen ^[47][48], and evidence suggests that the prevalence of ASD is very low in patients with cystic fibrosis ^[16].[47, 48], and evidence suggests that the prevalence of ASD is very low in patients with cystic fibrosis [16].
• Plausible mechanism #1: Acetaminophen use in adults temporarily blunts social trust [49] and ^[49] and awareness ^[50], emotional responses to external stimuli ^[51], and the ability to identify errors ^[52], indicating that the drug targets regions of the brain affected in patients with ASD.awareness [50], emotional responses to external stimuli [51], and the ability to identify errors [52], indicating that the drug targets regions of the brain affected in patients with ASD.
• Plausible mechanism #2: Genetic and immune factors associated with an increased risk of ASD have a detrimental effect on the body’s ability to metabolize acetaminophen [3, 16, 53]. ^[2][16][53]. Difficulties in metabolizing acetaminophen have long been known increase the toxicity of the drug ^[3].Difficulties in metabolizing acetaminophen have long been known increase the toxicity of the drug [4].
• Plausible mechanism #3: Acetaminophen is known to be highly toxic in the presence of oxidative stress [54] via a ^[54] via a mechanism that involves the formation of the toxic metabolite, NAPQI ^[55][56][57] and concomitant mitochondrial damage ^[58]. Oxidative stress ^[16] and possibly mitochondrial dysfunction ^[59] also play a role in ASD.mechanism that involves the formation of the toxic metabolite, NAPQI [55-57] and concomitant mitochondrial damage [58]. Oxidative stress [16] and possibly mitochondrial dysfunction [59] also play a role in ASD.
• Laboratory animal studies #1: Early life exposure to acetaminophen at doses similar to or even less than doses received by human babies and children results in long term, profound modification of brain function in both laboratory mice and rats ^{[12][13][14][15][60]}, by definition a severe adverse event that would have precluded any clinical testing of acetaminophen in babies and small children.[12-15, 60], by definition a severe adverse event that would have precluded any clinical testing of acetaminophen in babies and small children.
• Laboratory animal studies #2: In laboratory rats, acetaminophen affects the developing male brain more than the female brain [15]. In ^[15]. In laboratory mice, males are more susceptible to acetaminophen-mediated liver injury than are females ^[61]. ASD is more prevalent in males than in females ^[62].laboratory mice, males are more susceptible to acetaminophen-mediated liver injury than are females [61]. ASD is more prevalent in males than in females [62].
• Laboratory animal studies #3: Acetaminophen causes apoptosis-mediated death of cortical neurons in adult laboratory rats at concentrations lower than it causes liver failure ^[5]. Affected cortical neurons are implicated in ASD ^[63][64], and individuals with ASD have increased levels of biomarkers for neuronal apoptosis ^[65][66][67].[6]. Affected cortical neurons are implicated in ASD [63, 64], and individuals with ASD have increased levels of biomarkers for neuronal apoptosis [65-67].
• Observation in veterinary medicine: Adult cats are susceptible to acetaminophen-mediated injury due to the lack of a robust glucuronidation-dependent capacity for metabolism [68-71]. Human ^{[68][69][70][71]} Human neonates similarly lack a robust glucuronidation-dependent pathway ^[72][73].neonates similarly lack a robust glucuronidation-dependent pathway [72, 73].
• Underlying assumptions of safety proven to be misguided: Acetaminophen use in babies and children was assumed to be safe during the 1970s because it did not cause liver damage, despite the fact that it targets brain function and was never shown to be safe for neurodevelopment ^[1].[2].

Any one geographic or temporal association might be spurious, possibly unrelated to causality. For example, two independent studies, when taken together, show that the popularity of acetaminophen in two Scandinavian countries, Denmark and Finland, correlated with the prevalence of ASD in those countries (Figure 1). First, the sales of acetaminophen per unit population from 2006 through 2010 in Denmark were more than 2-fold greater than the sales of acetaminophen in Finland during the same time period [74]. Second, for children born in 2006, whose brain development might have been influenced by exposure to acetaminophen between 2006 and 2010, the prevalence of ASD in children born in Denmark was about 70% greater than the prevalence of ASD in children born in Finland [75]. Limitations to the conclusions that can be drawn from this previously unreported geographic association are evident. For example, total sales of acetaminophen do not necessarily reflect use of the drug during early development, when induction of ASD is possible, and therefore this association does not necessarily reflect a direct association between pediatric use of acetaminophen and the prevalence of ASD. Nevertheless, when considered in light of other lines of evidence [1, ^[7][8][9]8, 9], this association adds to the burden of evidence demonstrating without reasonable doubt that acetaminophen use in susceptible babies and children causes many if not most cases of ASD.

Figure 1. The sales of acetaminophen (APAP) in Denmark and Finland correlated with the prevalence of autism spectrum disorder (ASD) in those countries. The sale of acetaminophen is limited to the sale of drugs containing acetaminophen as the only active ingredient, and is measured in units of a “defined daily doses” (equal to 3 grams of active ingredient) per 1000 inhabitants per day. The sales for each year from 2006 to 2010 were averaged, and the mean plotted. Trends in sales for both Denmark and Finland showed an upward trend every year, and the upper limit of the range (data from 2010) is shown by the error bars. The prevalence of ASD is shown for all children, including males and females, born in 2006 with prevalence measured at 9 years of age. The 95% confidence interval reported by Delobel-Ayoub and colleagues is indicated by the error bars.

A summary of evidence that acetaminophen causes many if not most cases of ASD was published in mid-2022 [8]. In that summary of evidence, it was noted that

“…if paracetamol exposure is not toxic to neurodevelopment, then a number of observations remain unexplained. The tally shown in [the published manuscript] describes six unknown factors that must be invoked to account for all observations, and eight largely independent observations that must be attributed to coincidence.”

McFadden [76] convincingly argues that Occam’s Razor, when applied properly, is a guiding principle in modern science, distinguishing between conspiracy theory and science. Occam’s Razor can be described as the principle that the simplest explanation that accounts for available observations is probably the correct explanation. When applying this Razor, the addition of undefined factors is of considerable concern when used to support an explanation, particularly when such factors need not be employed when another explanation is available. As pointed out in McFadden’s closing statement [76]:

“…science is, ultimately, the method by which we use the tools of experimentation, mathematics, and logic to find the simplest explanations of the complex phenomena of our world…”

With this guiding principle in mind, the idea that acetaminophen is not responsible for extensive neurodevelopmental injury in the world today has the hallmarks of pseudoscience. Indeed, consensus without supporting data is consensus, not science.

3. Factors Affecting Susceptibility to Acetaminophen-Mediated Injury during Neurodevelopment

Most babies and children who are exposed to acetaminophen do not sustain injuries leading to ASD, indicating that some factor or factors render some individuals particularly susceptible to acetaminophen-induced injury, while others are resistant. Fortunately, the metabolism of acetaminophen has been widely studied for over half a century, and is exceptionally well characterized. A synopsis of some of the major biochemical pathways of acetaminophen metabolism are shown in Figure 2. Unfortunately, a wide range of factors, including various genetic factors, antibiotic use, infection, temporary inability to eat, exposure to environmental toxins, and problems with folate metabolism or transport can all induce oxidative stress, which leads to susceptibility to acetaminophen-induced injury. This issue was reviewed in detail in 2017 [16]. The result of this situation is that a variety of factors that induce oxidative stress and render acetaminophen more toxic are associated with ASD, but no one factor alone accounts for susceptibility to acetaminophen or for the occurrence of acetaminophen-induced ASD [16].

Figure 2. Schematic diagram showing some of the major metabolic pathways of acetaminophen (APAP) metabolism in humans. Three pathways—glucuronidation, sulfation, and oxidation—followed by a reaction with glutathione are shown. The major pathway in babies and in children, sulfation, tends to be impaired in children with autism spectrum disorder (ASD). This is expected to shunt more of the drug through the oxidative pathway, resulting in the production of excess N-acetyl-p-benzoquinone imine (NAPQI), the toxic compound shown in the diagram. Unfortunately, children with ASD also tend to have a reduced ability to detoxify NAPQI, resulting in increased toxicity of APAP due to excess NAPQI. The therapeutically important deacetylation reaction resulting in the production of p-aminophenol, and subsequent downstream reactions in that pathway, are not shown.

4. Periods of Sensitivity to Acetaminophen during Neurodevelopment

The timing of acetaminophen-mediated induction of ASD during brain development is of considerable interest. Based on an analysis of cohort data by Liew and colleagues [18], the neurodevelopmental window sensitive to acetaminophen may begin very early during pregnancy, perhaps in the first trimester (Figure 3). However, the time of birth is evidently the most critical period in terms of sensitivity to acetaminophen-induced neurodevelopmental injury (Figure 2). The view that newborns are exquisitely sensitive to acetaminophen is consistent with the 2-fold greater prevalence of infantile ASD associated with circumcision [77] and the 3.6-fold greater prevalence of ASD in the third of children with the highest levels of acetaminophen in their cord blood compared to the third with the lowest levels [34]. As discussed previously [9], the well-established age-dependent pharmacology of acetaminophen ^[78][79][78, 79] dictates that, by unit weight, the mother/fetus dyad is far more capable of metabolizing and detoxifying acetaminophen than the newborn alone. The dramatic differences in weight-adjusted capacity to metabolize and detoxify acetaminophen between pregnant women and newborns [9] are expected to adversely affect the developing brain of susceptible newborns when the umbilical cord is cut in the presence of acetaminophen. Based on pharmacokinetic considerations, the first 10 days of life in particular should be the most sensitive to acetaminophen, when the glucuronidation process, important for the detoxification of acetaminophen [9], is not yet functional [80]. Following the perinatal period, the work by Schultz ^[4][5] discussed above and the observations from parents discussed previously [16] indicate that acetaminophen induces many, if not most, cases of regressive ASD. Based on a meta-analysis by Tan and colleagues [81], regression can occur in children as old as 4 or 5 years, but the time distribution of regression is skewed toward earlier ages, with half of all cases of regression probably occurring before about 1.5 years of age. Thus, as shown in Figure 3, the neurodevelopmental window for the induction of ASD by acetaminophen is broad, with the prevalence of induction apparently dictated by levels of drug exposure and by susceptibility.

Figure 3. Schematic diagram showing prevalence of acetaminophen (APAP)-induced ASD as a function of age. The neurodevelopmental window of sensitivity is broad, possibly bounded by the first trimester of pregnancy and the end of the 4^th or 5^th year of life. The shape of the curve is apparently determined by (a) levels of exposure to acetaminophen, and (b) sensitivity to acetaminophen, a function of oxidative stress and ability to metabolize acetaminophen. ASD, autism spectrum disorder; APAP, acetaminophen (paracetamol).

5. Regulatory Action and Public Awareness

The initial report of a strong association between acetaminophen and ASD published in 2008 ^[4][5] was immediately questioned and largely forgotten, although subsequent analysis demonstrated that published study criticisms were not valid [9]. Likewise, numerous studies in animal models were ignored despite compelling results using a variety of study designs ^{[10][11][12][13][14][15]}[10-15]. Rather, the consensus, based on false assumptions ^[1][2], that acetaminophen is safe for pediatric use prevails in both the medical community and the public consciousness as of early 2024. Progress toward public awareness was likely hampered because classic multivariable regression analysis used to assess the connection between acetaminophen and ASD is widely trusted, but was demonstrated to be invalid [8] as a result of (a) extremely widespread use of acetaminophen, and (b) a variety of factors that make acetaminophen more toxic and act as cofactors for injury induction, but which were treated as confounding factors in the analysis.

Given that the therapeutic target of acetaminophen involves brain function, pediatric use of acetaminophen should have been preceded by extensive tests using animal models for neurodevelopment [8]. Preclinical toxicity screens using laboratory animals often fail to detect drug toxicity in humans [82]. However, neurodevelopment is a conserved process across mammalian species, and laboratory animals provide a very good model for examining brain sensitivity in the perinatal period [83]. Indeed, the toxicity of acetaminophen for the developing brain is readily detected using perinatal laboratory animal models ^{[8][10][11][12][13][14][15]}[8, 10-15]. Given the severe adverse effects of acetaminophen on neurodevelopment observed in laboratory animals by several independent laboratories using a variety of experimental designs, Parker and colleagues conclude that the standard safeguards involving preclinical testing that protect the population from exposure to drugs with a poor benefit-to-risk ratio were not employed when acetaminophen was introduced into widespread pediatric use in the 1980s ^[7][1].

Parker and colleagues have encouraged regulatory and policy-making bodies to restrict the pediatric use of acetaminophen, and they stress the importance of consumer education to inform individuals of current knowledge regarding the impact of acetaminophen on the developing brain, particularly in the perinatal period ^[7][1]. The fact that consumers lack education regarding the dangers of acetaminophen for neurodevelopment is evident from numerous studies showing widespread frivolous use of the drug for lowering body temperatures that are not actually high enough to be classified as a fever ^{[84][85][86][87][88]}[84-88], and misuse of the drug either by giving too high of a dose or administering the drug too frequently ^{[86][87][88][89][90][91][92][93][94][95]}[86-95]. Of particular concern is the use of acetaminophen immediately before and after birth, both of which affect the neonate, when susceptibility to injury is greatest (Figure 3). It is within this time frame that the benefits of acetaminophen have not been proven ^[7][1], and when 50% or even more of all cases of ASD may be induced based on the estimate by Parker and colleagues ^[7][1].