ADHD Medication and Brain Structure: The ABCD Study

"Does ADHD medication change the brain?" is a question that everyone who takes or is thinking of taking medication asks at some point. A new study looks for a kind of answer to it, but when you read the abstract carefully you see that the answer it gives is smaller than the question itself. First let us look at what exactly the abstract says, especially the difference between thickness, surface area and volume, and then at what it does not say.

Who the study is from and where it comes from

The authors of the study are L. Nate Overholtzer, Katherine L. Bottenhorn, Sarah L. Karalunas, Bradley S. Peterson and Megan M. Herting. The study was published in the journal Scientific Reports, in June 2026. The data come from the ABCD study (Adolescent Brain Cognitive Development) and, as far as can be understood from the title, concern preadolescent children. The abstract does not separately write the age range.

The method has two steps. First, machine learning was used on 1,306 participants to look for the brain regions targeted by the medications. Then, in 8,762 participants, the effect of ADHD status and of medication use (amphetamine, methylphenidate and nonstimulant medications) on brain measures was estimated with linear mixed-effects models (a type of regression that can handle data with more than one level). The abstract does not detail how the two groups relate to each other. As far as can be understood from the abstract, the study is observational: medication use does not appear to have been assigned at random in an experiment; the authors looked at the children's existing situations.

This post is a summary of a paper published with open access in the journal Scientific Reports. The paper was released under the Creative Commons Attribution 4.0 (CC BY 4.0) license. The findings belong to the authors of the paper; the account and the assessment here are mine. I rely only on the abstract; I do not know the full text of the paper.

What question the authors are asking

In the introduction of the abstract the authors say this: ADHD is the most common neurodevelopmental disorder in the USA, and amphetamine, methylphenidate and nonstimulant medications are among the medications most often prescribed to children. Medication treatment is described as the most effective treatment, but it is noted that the effects of the medications on the structure of the brain are still not well defined. The study looks at this gap.

The authors' stance matters here. They say that medication treatment is effective, and then they say "the effect on brain structure is not well known." So the study begins not with a suspicion but with a lack of knowledge. The article What is methylphenidate and how does it work? describes the functional side, which lies outside this gap.

Thickness, surface area and volume: three separate measures

The cortex, the outer layer of the brain, can be measured in a few different ways. The study examined four: cortical thickness, surface area, cortical volume and subcortical volume.

Thickness is the distance from the outside of the cortex to the inside, that is, how thick the layer is. Surface area is the size of the area over which the cortex spreads. Volume is roughly the combination of the two, the total space a region takes up. These three measures measure different things, and in a region one can change while another does not.

This distinction is critical for reading the abstract. The abstract says that cortical thickness was examined too, but lists as significant results only the surface area and volume findings. That is, no finding about thickness is reported in the abstract. A sentence like "ADHD medication thins the cortex" cannot come out of this abstract. The abstract says nothing on this, neither positive nor negative.

What ADHD itself showed

This is perhaps the least discussed but most important part of the abstract: ADHD status was not linked to significant differences in brain structure. The authors write that the effects are small and bidirectional. Bidirectional means that in some regions people with ADHD have higher measures and in some lower, but none is large enough to count as statistically significant.

You can read this in two ways. The first is that the brain structure of children with ADHD cannot be told apart as a group from that of children without ADHD on these measures. The second is that no difference showing up in the average of these measures does not mean that the brains of individual children are the same. The abstract does not discuss what this finding means for diagnosis.

What "attenuation" means

The main finding of the abstract is what the authors call "a consistent pattern": the effects of amphetamine and methylphenidate use on brain measures came out in the opposite direction to the ADHD associations. The authors call this "attenuation," that is, a shift toward the pattern of the control group. For nonstimulant medications this pattern is weaker.

To understand this correctly you need to keep two things in mind together. The first is that ADHD itself does not show a significant difference, so what is "in the opposite direction" is the opposite of a small, non-significant difference. The second is that the expression "shift toward the control pattern" is a description of direction, not an evaluation. The abstract does not say whether this shift is good or bad.

What was found in which regions

The significant findings are these:

• Amphetamine and methylphenidate use, both of them, were linked to a lower surface area of the right banks of the superior temporal sulcus and of the left posterior cingulate region.
• Amphetamine was additionally linked to a lower volume of the left posterior cingulate region.
• Methylphenidate was linked to a higher volume of the right temporal pole.

So the number of findings is small. Three regions, a few measures. The abstract says nothing about the function of these regions, so I too will not form sentences such as "this region controls attention" or "this region is related to memory." With methylphenidate there is a volume increase in one region and a decrease in surface area in two regions. So it is not a one-directional story of "shrinking" either.

The authors write that the findings on stimulant medications point to partly independent structural effects in the cortex, and that these effects extend beyond the ADHD phenotype. This says that the differences linked to medication use may not be merely a reversal of ADHD, and that they may reflect other things too. This "may" is a possibility, not something proven.

What this finding means for you

This study does not give a reason to take medication or not to take it. It says neither "use it" nor "stop." The abstract itself calls medication treatment the most effective treatment, but that is the opening sentence, not a finding of this study.

Why is it hard to infer cause and effect in an observational study? Because children who use medication may differ from those who do not in other ways too. For example their symptoms may be more severe, or they may have received the diagnosis earlier. The abstract does not say how it accounted for these differences. The abstract also does not detail how long the children have been using medication or at what dose. From the same data, saying "the medication changed brain structure in this way" would be wrong, and so would saying "those who use medication already had a brain like this."

Most of the real worries about medication are elsewhere anyway: ADHD medication side effects, Does ADHD medication affect height in children?, Does ADHD medication change personality?. The decision about medication is something to be talked through with a doctor.

What this study does not say

This is a finding of association. It does not show causation.

The abstract does not say "harm" or "benefit." A lower surface area or a higher volume cannot, by itself, be interpreted as good or bad.

No finding is reported in cortical thickness. So a sentence like "medication thins the cortex" or "prevents thinning" cannot come out of the abstract.

The abstract does not explain the medication dose, the duration, or the children's state before medication. I should also point out that the abstract does not state the design of the study (whether the same child was followed over time).

One study, one dataset. We do not know what other studies in this area will find.

What stays with me

What caught my attention while reading this study is how cautious the abstract's own language is: "the effects are small," "no significant difference," "associated," "suggest." The easy headline to write would be "ADHD medications change the brain." That is not what the abstract says.

I tried medication a second time at 28 and stopped. So on this subject I can say neither "medication is a must" nor "medication is bad." In my view the decision about medication should be made not by looking at small differences in a brain image but by looking at whether that medication works in everyday life, at its side effects, and at the whole picture you discuss with your doctor. This study is data that can enter that conversation, but it does not take the place of the conversation. For a wider frame you can look at the article Medication or therapy for ADHD.