A large observational analysis described medication-associated brain patterns that often pointed in the opposite direction to ADHD-associated patterns. Its term attenuation does not demonstrate that amphetamine repaired or normalized an individual child’s brain.
The study used ABCD data, with a machine-learning stage involving 1,306 participants and mixed-effects analyses involving 8,762. It examined amphetamine, methylphenidate and nonstimulant medication classes in children aged nine to ten.
This was human observational research at one developmental time point. Medication use was not randomized, and structural MRI associations were not measurements of treatment-related change within the same child.
ADHD itself was not associated with significant structural differences in the principal reported analyses; estimated associations were small and bidirectional. Amphetamine and methylphenidate patterns often opposed those ADHD associations.
Both stimulant classes were associated with smaller surface area in selected cortical regions. Other reported regional associations differed by medication class. An opposing direction of estimates does not establish a corrective biological mechanism, especially when the underlying disorder comparison is not significant.
Lifetime medication exposure before enrollment was unavailable. Some participants classified as currently unmedicated could have received treatment previously. Dose, symptom severity and behavioral outcomes were not evaluated as moderators in this analysis.
Cross-sectional selection, prior exposure and confounding by treatment indication limit causal conclusions. The findings cannot establish benefit or harm from an individual’s MRI measurement, permanent tissue change or a concentration–impairment relationship. Their contribution is a set of medication-associated structural hypotheses for longitudinal testing.
See apparent MRI changes measured shortly after medication.
ADHD medications and preadolescent brain structure: patterns of cortical attenuation from the ABCD study. doi:10.1038/s41598-026-52107-4.