2026 · Paediatric and Perinatal Epidemiology
Maternal Acetaminophen Use and Offspring's Neurodevelopmental Outcome: A Nationwide Birth Cohort Study
Verdict
UNSCIENTIFIC
Neither the paradigm nor the study inside it passes. The verdict is unscientific.
1·Study methodology · inside the paradigm
UNSCIENTIFIC
Does not follow the scientific method within Psychiatric Model.
2·Paradigm · Psychiatric Model
UNSCIENTIFIC
Fails the tests. The worse layer decides.
How the verdict is decided
The verdict grades both layers: the paradigm this study assumes, and how the study was carried out inside it. Both must be scientific for the whole thing to be scientific. A clean method inside an unscientific paradigm is unscientific. An unscientific method inside a better paradigm is unscientific. A split on either layer is mixed. The worse layer decides.
1·Study methodology
Did this study test its claim with methods that are independent, falsifiable, and non-circular?
UNSCIENTIFIC
This study claims that children whose mothers took acetaminophen during pregnancy were diagnosed with ADHD about 22 percent more often than other children, in a Japanese birth cohort of 217,602 children. All three tests fail.
Independently Verifiable
Fail
The outcome is a code a doctor entered after deciding a child met the manual's checklist for ADHD, autism, or intellectual disability, and the study counted the codes, not the children's behavior.
Falsifiable
Fail
A same-mother sibling result of 0.86 and the bias checks pointing to zero were already explained as bias, so the study keeps a 'small increase' whichever way the numbers fall.
Non-Circular
Fail
The checklists that create the ADHD, autism, and intellectual disability labels are the same checklists whose output is counted as harm to the child's development.
Why
They pulled insurance records for 217,602 children in Japan and flagged mothers with acetaminophen recorded during pregnancy, about 40 percent. Pills bought over the counter never entered the file. The outcome was whether a doctor later entered an ADHD, autism, or intellectual disability code, which happens only after the doctor applies the manual's criteria. They matched exposed and unexposed mothers on recorded traits and got 1.22 for ADHD. Then siblings, bias simulations, and other drugs pulled the number toward or below 1. The paper still reports small increases that bias 'may partially explain.' No one observed any change in a child's development. The file holds codes, and each code is the criteria applied.
2·Paradigm · Psychiatric Model
Does the framework this study assumes pass the three tests?
UNSCIENTIFIC
Independently Verifiable
Fail
A child's restless or withdrawn behavior can be watched by anyone, but the claim that ADHD, autism, and intellectual disability are separate conditions with their own causes exists only as the checklist's reading of that behavior.
Falsifiable
Fail
A risk factor that fails to predict diagnoses is called confounding, and one that does is called a cause, while rising counts are called better detection, so no result ever touches the diagnoses themselves.
Non-Circular
Fail
The criteria were written by committees from clinicians' impressions of behavior and are confirmed by agreement with clinicians who apply those same criteria.
Why
The psychiatric model takes a checklist of behaviors and treats a passing score as a disorder inside the child, then asks which pill or exposure caused the label instead of what damaged the child. The behavior is real and visible. The disorder, its subtypes, and its cutoffs come from the manual, and the count changes when the manual is rewritten while every child behaves the same. When risk studies come out null, the field says confounding. When diagnoses rise, it says awareness. Validation is agreement with clinicians using the same criteria. Nothing outside that loop can show the labels are wrong.
From the study
We investigated the association between maternal acetaminophen use during pregnancy and offspring's neurodevelopmental outcomes using Cox proportional hazards models, with primary analyses based on 1:1 propensity score (PS) matching.
PS‐matched analyses ( N = 42,123 children per comparator) yielded hazard ratios of 1.08 (95% CI 1.00, 1.16) for composite neurodevelopmental outcomes, 1.22 (95% CI 1.09, 1.36) for ADHD, 1.06 (95% CI 0.98, 1.15) for ASD, and 1.02 (95% CI 0.90, 1.19) for ID.
Sibling comparisons ( n = 23,593) showed point estimates in the opposite direction (e.g., HR of ADHD, 0.86 95% CI 0.52, 1.44).
Although PS‐matched analyses indicated small increases in risk, sensitivity analyses suggested that unmeasured confounding, misclassification and other biases may partially explain these associations.