Acetaminophen use and ADHD: Should we be concerned?

According to the Centers for Disease Control, the percentage of children diagnosed with attention deficit hyperactivity disorder (ADHD) has risen significantly from 7.8% in 2003 to 9.5% in 2007 and 11% in 2011.¹ Could a common over-the-counter medication be to blame?  Acetaminophen is used by more than 50% of pregnant women in the United States for pain during pregnancy and is considered the first-line drug of choice.² Historically it is regarded as safe, although a previous study shows acetaminophen use for 4 or more weeks during pregnancy can lead to cryptorchidism.³

Researchers have recently discovered that long-term acetaminophen use may have endocrine-disrupting properties.^4,5  Their findings generated concerns about exposing fetuses to acetaminophen because it may interfere with sex or thyroid hormone function, both of which are critical for normal brain development.  Interfering with brain development during pregnancy may lead to behavioral dysfunction in the child.

These concerns may not be unfounded.  Using data from Danish National Birth Cohort study registry, researchers recently found an increased risk of behavioral problems like attention deficit hyperactivity disorder (ADHD) and hyperkinetic disorder (HKD) in children whose mothers took acetaminophen during pregnancy.⁶   The researchers specifically examine correlations between parental reports of children’s ADHD-like behaviors by age 7, hospital diagnoses of HKD, and ADHD medication use in children whose mothers took acetaminophen during pregnancy.

Recruitment for the Danish National Birth Cohort study occurred between gestational weeks 6 to 12 from 1996-2002 and this analysis included 64,322 live-born children and mothers from that cohort.  Participant mothers were excluded from this analysis if they did not complete at least 1 telephone survey (n = 28,254), the pregnancy was unsuccessful (n = 6207), the pregnancy resulted in multiple gestations (n = 2080), maternal emigration (n = 51), maternal death (n = 3), unknown birth outcomes (n = 25), or missing birth date (n = 99). The study analyzed children for ADHD-like behaviors but excluded children whose primary caregivers did not respond to follow-up questionnaires after the child’s 7^th birthday. 

Women received 3 telephone surveys (12^th and 30^th week gestation, and 6 months postpartum) regarding self-reported pain medication use, both over-the-counter and by prescription, and gestational week of medication consumption. Both acetaminophen monotherapy and combination product use were elicited. If the exact week of medication use was unknown the current trimester was recorded.

To evaluate for ADHD-like behaviors, the Strengths and Difficulties Questionnaire (SDQ) assessing 5 domains (emotional symptoms, conduct problems, hyperactivity, peer relationship, and pro-social behavior) was utilized. When the child turned 7, mothers or primary caregivers completed a 25-question survey about their child’s behavior during the previous 6 months. For SDQ scoring, a total score was determined by adding 4 subcategories (emotional symptoms, conduct problems, hyperactivity, and peer problems). Total scores ranged from 0-40, with each subcategory ranging from 0-10; higher scores indicating increased behavioral problems. Scores were dichotomized using a cut-off point for the total score of 17 or more. The mother or primary caregiver was also asked to complete a 6-question survey about their own childhood behavioral problems to assess for maternal behavioral problems or ADHD-like symptoms.

HKD diagnoses were determined by the Danish National Hospital Registry.  ADHD medication use was determined using the Danish Prescription Registry. Only children who filled 2 or more prescriptions for methylphenidate, atomoxetine, or modafinil were considered positive for ADHD medication use.

The analysis identified several confounders including child’s birth year, birth weight, and sex, as well as maternal age at child’s birth, parity, gestational age at delivery, socioeconomic status, tobacco and alcohol exposure during pregnancy, pre-pregnancy body mass index, and mother’s self-reported psychiatric illnesses. The researchers also accounted for confounders that might have affected acetaminophen use during pregnancy including muscle and joint diseases, fever, and inflammation or infection. Cox proportional hazards regression models were used to estimate hazard ratios and 95% CIs for pregnancy acetaminophen use of mothers and HKD diagnosis and ADHD medication use in children. In addition to dichotomizing SDQ scores, linear regression on continuous SDQ scores was used.

Baseline characteristics of acetaminophen users versus non-users were generally similar.  Acetaminophen use was comparable when examining mothers’ age, gestational age, and tobacco exposure. Acetaminophen users were more likely to have previous psychiatric illnesses, fever during pregnancy, muscle and joint disease during pregnancy, or infection and inflammation during pregnancy, although statistical analysis was not reported.  The mean age of children at the end of follow-up for HKD diagnoses was 10.7 years (range 8.2-13.4 years) and 11.2 years (range 8.6-13.9 years) for ADHD medication use.

Over half of all mothers reported taking acetaminophen during pregnancy in both the cohort examining HKD diagnoses and ADHD medication use (56%; n = 36,187) and the cohort identifying ADHD-like behaviors based on SDQ score (55%; n = 22,687). The study also showed an increased risk for ADHD-like behaviors by age 7 in children whose mothers took acetaminophen during pregnancy (total difficulties scores 17 or more, risk ratio = 1.13; 95% CI, 1.01-1.27). Increased risk was observed in women taking acetaminophen during more than 1 trimester, especially if consumed later in pregnancy.  Further, each additional week of acetaminophen use was associated with increased SDQ score (see Table 1).

Table 1: Select Endpoints of Risk Ratios for ADHD-like Behavioral Problems in Children at Age 7 Year and Maternal Acetaminophen Use⁶

*Adjusted maternal age at birth, sex of child, child’s birth year, gestational age, birth weight, parity, socioeconomic status of mother, maternal smoking and alcohol drinking during pregnancy, maternal pre-pregnancy body mass index, patient’s behavioral scores in childhood, mother’s ever having had mental health problems, and maternal diseases in muscles/joints, fever, or infection/inflammation during pregnancy

Acetaminophen use was also associated with increased risk for HKD and ADHD medication use. HKD diagnosis in children nearly doubled when mothers reported taking acetaminophen in 20 or more weeks during pregnancy (hazard ratio, 1.84; 95% CI, 1.39-2.45) and risk for needing ADHD medications increased by 50% (hazard ratio, 1.53; 95% CI 1.21-1.94).  Similar to elevated SDQ scores, risk increased with acetaminophen exposure in more than 1 trimester (see Table 2).

Table 2 Select Endpoints of Hazard Ratios for HKD Hospital Diagnosis or ADHD Medication Redemption According to Maternal Acetaminophen Use During Pregnancy⁶                           

*Adjusted for maternal age at birth, sex of child, child’s birth year, gestational age, birth weight, parity, socioeconomic status of mother, maternal smoking and alcohol drinking during pregnancy, maternal pre-pregnancy body mass index, mother’s ever having had mental health problems, and maternal disease in muscles/joints, fever, or infection/inflammation during pregnancy

One strength of this study was the ability to examine a correlation to acetaminophen use and neurobehavioral disorders using 3 different endpoints: HKD diagnosis, ADHD medication use, and ADHD-like behaviors.  Selection bias was limited since patient response was not needed to assess treatment and hospital outcomes.  Follow-up was exceptional – only 1.3% (n=813 children) were lost to follow-up.  Recall bias was limited, as mothers were interviewed throughout pregnancy, well before any signs of neurodevelopmental behaviors were evident.

The authors concluded that prenatal exposure to acetaminophen may increase the risk of HKD diagnosis, ADHD medications use, and ADHD-like behaviors in children. While an association was clearly found, like all epidemiology studies, it is difficult to establish a definite cause and effect link.  For example, acetaminophen is a common remedy for body aches and pain from infections.  Infections during pregnancy also have a documented association with childhood ADHD.^7,8  While the study did attempt to adjust for inflammation/infections leading to increased acetaminophen use, it is difficult to account for the effects confounders potentially had on the outcomes.  Also, the exact etiology of ADHD is not understood, but it probably has genetic links.  The analysis accounts for only the mother’s propensity for ADHD, not the father’s or other family members.  Although the study adjusted for alcohol and tobacco exposure during pregnancy, illegal substance use was not examined.  Finally, thyroid hormone plays a direct role in regulating fetal brain development but thyroid disorders were not accounted for in this analysis.

With regard to medication use, the investigators did not include all ADHD stimulant medications and excluded non-stimulant ADHD drugs entirely.  Thus they may have underestimated the effects of acetaminophen exposure.  Conversely, methylphenidate and modafinil have another indication – narcolepsy.  While it’s unlikely a large number of children included in the analysis had narcolepsy, it’s impact on the study’s findings is unknown. Lastly, the study did not differentiate between acetaminophen monotherapy versus its use in combination products.   Nor do the authors report the acetaminophen doses consumed, the regimen (around-the-clock dosing versus as needed), or whether the recommended maximum dose per day was exceeded.

Nevertheless, the findings from this study is consistent with another study which evaluated acetaminophen use during pregnancy.⁹  The study found acetaminophen exposure greater than 28 days was associated with poorer gross motor development, communication, externalizing and internalizing behavior, and higher activity levels at 3 years of age.⁹   Both studies found a correlation between cumulative exposure to acetaminophen and developmental disturbances.

While an association is evident, causality can not be established and several questions remain.  If blacklisted, what pain medication should be used instead? Should women endure pain un-medicated throughout pregnancy? While women likely do not need to avoid acetaminophen at all costs during pregnancy, it seems caution should be exercised and long-term use discouraged.  Patients should be educated to use the lowest dose possible for the shortest duration to treat the pain. What do you think?  Should clinicians stop recommending acetaminophen use during pregnancy?