A CCHR Florida analysis of “Attention-Deficit/Hyperactivity Disorder Accommodations, Behavior Therapy, and Stimulant Medication: A Randomized Clinical Trial” (Fabiano et al., JAMA Pediatrics, published online November 13, 2026)
Prepared November 16, 2026
Summary
On November 13, 2026, researchers at Florida International University (FIU) published a trial in JAMA Pediatrics that tested four common classroom accommodations for children labeled with ADHD (extra time, frequent breaks, a standing desk, and a low-distraction workspace), along with a behavioral program and the stimulant methylphenidate. The authors concluded that the accommodations did not help, and that the results “support the continued use of behavioral and pharmacological treatment for ADHD” (JAMA Pediatrics).
News coverage has focused on the message that stimulant drugs “worked.” A close reading of the paper shows something different:
- The drug was tested in one-week blocks during a summer camp, so the study says nothing about what happens to a child taking a stimulant for months or years.
- The academic “benefit” was 3.9 points on a 10-question quiz, which is less than half of one question.
- The paper contains internal errors, including a reversed headline statistic and a child age reported in months instead of years.
- Adverse effects were collected every week but the data were not published.
- The study was run by the same program that designed the behavioral intervention, and the lead author reports royalties from the publisher of his book on that intervention.
- The behavioral program, with no drug at all, cut rule-breaking by 37%. Non-drug approaches worked.
Decades of earlier research, including studies from this same FIU center, show that stimulant drugs do not improve children’s learning or long-term outcomes, and that they carry real risks to growth, the heart, and mental health. A child’s brain is still developing. A week of calmer seatwork in a summer camp is not a reason to put a child on a mind-altering drug.
Part One: Flaws in the Study
- A one-week drug test cannot support long-term drug use
Children alternated between methylphenidate and placebo every week, and the classroom conditions switched every two weeks, over a summer program of about eight weeks (seven weeks in one cohort) (JAMA Pediatrics). The registered outcomes were measured only “through study completion, maximum of eight weeks” (ClinicalTrials.gov, NCT05469386).
This design can only capture the short-term, same-day effect of a stimulant. In real life, children are typically prescribed these drugs every school day for years. The authors themselves acknowledge “potential long-term, serious adverse effects of sustained high doses of stimulant medication” (JAMA Pediatrics), yet their conclusion recommends “continued use” of the drug. Nothing in an eight-week crossover trial can justify that recommendation.
- The academic gain was tiny
The headline claim is that the drug “improved quiz scores.” The improvement was 3.9 percentage points (JAMA Pediatrics). The quizzes had only 10 multiple-choice questions (JAMA Pediatrics), so the average drugged child answered fewer than half a question more correctly than the same child on placebo.
The paper’s own effect sizes for medication on quiz scores range from about 0.14 to 0.21 standard deviations, and for seatwork productivity about 0.12 to 0.18 (JAMA Pediatrics, Table 2). By common research conventions, an effect of 0.2 is considered “small.” These are small, short-term changes in a camp classroom, not evidence of better learning.
- “Rule violations” measure compliance, not wellbeing or learning
The largest drug effect was on counts of classroom “rule violations” (JAMA Pediatrics). Fewer rule violations means a child was quieter and more compliant with staff. It does not mean the child learned more, felt better, or was healthier. Making children easier to manage in a classroom is not the same as helping them.
The rule-violation data are also highly lopsided. In the general classroom on placebo, the average was 5.74 violations with a standard deviation of 15.23 (JAMA Pediatrics, Table 2). When the spread is nearly three times the average, a small number of children are likely producing most of the counts, and percentage reductions can look dramatic while most children change very little.
- The paper contains internal errors
A peer-reviewed trial being used to guide school policy should, at minimum, report its own numbers consistently. This one does not:
- The headline drug statistic is reversed in the abstract. The abstract says medication and behavioral therapy reduced rule-breaking “by 37% … and 61% …, respectively,” which credits medication with 37% and behavior therapy with 61%. But the Results section and Table 3 say the opposite: behavioral intervention lowered rule violations by 37% and medication by 61% (JAMA Pediatrics). FIU’s own press release uses the 37% behavioral and 61% medication version (FIU News).
- Children’s ages are reported in months. Both the abstract and the Results state a mean age of “8.8 [1.6] months” (JAMA Pediatrics), even though participants were in kindergarten through sixth grade.
These errors passed through peer review into the published article. They raise fair questions about how carefully the rest of the paper was checked.
- The productivity measure was changed in a way that penalizes extra time
The trial registration lists seatwork completion as the “percentage of assigned seat-work items completed correctly” (ClinicalTrials.gov). The published paper instead reports problems completed correctly per minute (JAMA Pediatrics). The authors also note that their initial plan called for analysis of variance, but they switched to a different statistical model “on reflecting on the nature of the outcomes” (JAMA Pediatrics).
The change matters. Under the extended-time accommodation, children got 45 minutes instead of 30 (JAMA Pediatrics). If a child completes the same amount of correct work in more time, a per-minute rate automatically goes down. That is exactly the accommodation that the paper reports as significantly “worsening” productivity (JAMA Pediatrics, Table 3). Meanwhile, on quiz scores, extended time had the largest positive estimate of any accommodation (1.18 points), though it was not statistically significant (JAMA Pediatrics, Table 3).
- The accommodations were not tested the way schools use them
Accommodations were “randomly assigned daily,” so a child might have extra time one day, a standing desk the next, and headphones the day after (JAMA Pediatrics). In schools, accommodations in an IEP or 504 plan are chosen for an individual child and used consistently over time. The authors themselves concede that “the rapid changes in accommodation use (vs sustained use of a single approach; temporal variability) may also have impacted results,” and that individual differences were not addressed in this paper (JAMA Pediatrics). A design that rotates supports daily was not a fair test of them.
- A summer camp is not a school
The trial was conducted inside FIU’s Summer Treatment Program, with groups of 12 to 15 children, a trained teacher and aide in each class, and the rest of the day spent on sports, games, and crafts (JAMA Pediatrics). The authors state that “replication in everyday school classrooms is needed” and that “the study was conducted over the summer” (JAMA Pediatrics). Findings from a tightly controlled camp should not drive drug recommendations for ordinary classrooms.
- Adverse effects were collected but not reported
The Methods state that “adverse effects ratings were gathered weekly” (JAMA Pediatrics). The paper does not publish those data. Instead, the Discussion offers a single sentence that “there were very few instances where adverse effects resulted in the discontinuation of medication” (JAMA Pediatrics). The registry page shows no posted results (ClinicalTrials.gov).
“Few discontinuations” is not the same as “few side effects.” Sleep problems, appetite loss, irritability, and stomachaches rarely cause a parent to pull a child from a study but are still real harms. Parents and policymakers deserve to see the full adverse-effect data before anyone cites this trial as proof the drug is safe.
- The sample was screened to favor the drug
Children with “prior adverse methylphenidate reactions” were excluded (JAMA Pediatrics). Removing the children most likely to react badly to the drug makes the drug look better tolerated than it would be in the general population. Children who had been suspended or expelled in the past year were also excluded (JAMA Pediatrics).
- “Blinding” is unlikely to hold with a stimulant
The trial is described as masked, with drug and placebo hidden in opaque capsules (JAMA Pediatrics). But stimulants produce noticeable effects that staff and children can recognize. The Cochrane review of methylphenidate for children concluded that if unblinding due to typical adverse events is considered, all 212 trials it examined were at high risk of bias, and it called for an “active placebo” to address this (Cochrane). The earlier Cochrane analysis likewise found that only 6 of 185 trials appeared low risk of bias in all domains (BMJ). In this trial, the same staff who knew the classroom routines were the ones counting rule violations.
- Young children were given a stimulant for research
The trial enrolled children from kindergarten through sixth grade, and the registry lists eligible ages as 5 to 12 years (ClinicalTrials.gov). The FDA label for Concerta, a leading extended-release methylphenidate, states it “should not be used in children under six years, since safety and efficacy in this age group have not been established,” and adds that “long-term effects of methylphenidate in children have not been well established” (FDA, Concerta label). The paper does not report how many children under six received the drug. That question deserves an answer.
- Investigators evaluated their own program
The study was designed and run by FIU’s Center for Children and Families inside its own Summer Treatment Program (FIU News). The behavioral intervention used daily report cards (JAMA Pediatrics). The lead author discloses “royalties from Guilford Press during the conduct of the study,” and the paper’s own reference list includes his Guilford Press book on daily behavior report cards (JAMA Pediatrics). This does not invalidate the work, but it is a relevant interest when the paper recommends that schools adopt that approach.
- The study’s own results support non-drug approaches
Lost in the headlines: the behavioral classroom program, using praise, clear rules, daily report cards, and earned rewards, reduced rule violations by 37% with no drug at all (JAMA Pediatrics). The authors note it required “no additional cost for implementation, or additional staff time” (JAMA Pediatrics). FIU’s own news release recalls that earlier work by the center’s founder found that starting with behavioral interventions was more cost-effective than beginning with medication (FIU News).
Part Two: What Earlier Research Shows
Stimulants do not improve learning (FIU, 2022)
The same FIU center ran a triple-masked crossover trial of 173 children ages 7 to 12 in the same Summer Treatment Program. The drug increased seatwork and reduced rule violations, but “there was no detectable effect of medication on learning the material taught during instruction: Children learned the same amount of subject-area and vocabulary content whether they were taking OROS-MPH or placebo” (Journal of Consulting and Clinical Psychology, via PubMed). The senior author, the late William Pelham Jr., said: “Unfortunately, we found that medication had no impact on learning of actual curriculum content” (FIU News, 2022). The test-day bump was 1.7 points out of 100, “not enough to boost most children’s grades” (FIU News, 2022).
This is the most important context for the new trial. A small same-day quiz increase is not learning, and FIU’s own prior work shows the drug does not make children learn more.
No long-term benefit (the NIMH MTA study)
The federally funded Multimodal Treatment Study (MTA) is the largest long-term study of ADHD treatment. At the 8-year follow-up, “the originally randomized treatment groups did not differ significantly” on outcomes including grades, arrests, and psychiatric hospitalizations (Journal of the American Academy of Child and Adolescent Psychiatry, via PubMed). The National Institute of Mental Health reported that children who stopped medication “were generally functioning as well as children who were still medicated” (NIMH).
Stunted growth with no symptom benefit (MTA, 16-year follow-up)
By adulthood, “extended use of medication was associated with suppression of adult height but not with reduction of symptom severity” (Journal of Child Psychology and Psychiatry, via eScholarship). Children with consistent or inconsistent medication use were 2.55 cm shorter than those with negligible use (Journal of Child Psychology and Psychiatry, via eScholarship). The FDA label for Concerta also warns of “slowing of growth rate,” about 2 cm less height and 2.7 kg less weight over three years (FDA, Concerta label).
Worse school outcomes when drug use expanded (Quebec, 2014)
When Quebec expanded drug coverage and stimulant use rose, researchers found “no evidence that the performance of children with ADHD improved.” Instead, increased use was associated with “increases in the probability of grade repetition, lower math scores, and a deterioration in relationships with parents,” and later with more depression (Journal of Health Economics, via PMC).
Ten times the odds of performing below grade level (Western Australia, 2010)
A state government study of children in the Raine birth cohort found that “ever receiving stimulant medication was found to increase the odds of being identified as performing below age-level by a classroom teacher by a factor of 10.5 times,” and concluded there was “little long-term benefit of stimulant medication” (Government of Western Australia, Department of Health).
The evidence base is very low quality (Cochrane)
The Cochrane review of methylphenidate in children rated “the certainty of the evidence for all outcomes” as very low, found the drug may not improve quality of life, and found increased non-serious adverse events such as sleep problems and decreased appetite (Cochrane Library, via PubMed). In drug-treated groups, 538 per 1,000 children experienced non-serious adverse events compared with 437 per 1,000 controls (Cochrane Library).
Serious harms, including heart problems and psychosis (Cochrane harms review)
Cochrane’s separate review of harms found that “methylphenidate administration might lead to serious adverse (harmful) events, including death, cardiac problems, and psychotic disorders,” with more than half of exposed patients experiencing one or more adverse events (Cochrane). The reviewers warned the true risk “might be higher than reported here” (Cochrane).
Long-term cardiovascular risk (Sweden, 2024)
A study of 278,027 people in Sweden found that “longer cumulative duration of ADHD medication use was associated with an increased risk of CVD, particularly hypertension and arterial disease,” with risk higher for stimulants, and similar patterns in children and youth (JAMA Psychiatry). The Concerta label also requires monitoring “all CONCERTA-treated patients for hypertension and tachycardia” (FDA, Concerta label).
Psychosis and mania
Among 221,846 adolescents and young adults starting prescription stimulants, “new-onset psychosis occurred in approximately 1 in 660 patients,” with roughly double the risk for amphetamine compared with methylphenidate (New England Journal of Medicine). The FDA label for Concerta warns that stimulants “at the recommended dosage, may cause psychotic or manic symptoms (e.g., hallucinations, delusional thinking, or mania) in patients without a prior history of psychotic illness or mania” (FDA, Concerta label).
FDA’s strongest warning: abuse, addiction, and overdose
In 2023, the FDA required updated Boxed Warnings, its most prominent warning, across all prescription stimulants to address “misuse, abuse, addiction, and overdose,” noting that misuse “can result in overdose and death” (FDA). The revised label language adds that “some individuals may develop addiction even when taking [the drug] as prescribed” (FDA, label change table).
Many children may simply be the youngest in their class
A New England Journal of Medicine study of more than 400,000 children found that in states with a September 1 kindergarten cutoff, children born in August had a 34% higher rate of ADHD diagnosis than those born in September (New England Journal of Medicine, via author copy). The lead author said “large numbers of kids are being overdiagnosed and overtreated for ADHD because they happen to be relatively immature compared to their older classmates” (Harvard Gazette). In the new FIU trial, ADHD status was established through rating scales and a parent interview (JAMA Pediatrics), and children as young as kindergarten were included.
Conclusion
The FIU/JAMA Pediatrics trial shows that, during a summer camp, a stimulant made children somewhat quieter and gave them a fraction of a quiz question more on the days they took it. It does not show that the drug helps children learn, does not show long-term benefit, and does not publish the side-effect data it collected. FIU’s own 2022 trial found the drug did not improve learning. The NIMH MTA study found no lasting benefit and found stunted growth. Studies from Quebec and Western Australia found worse school outcomes. Cochrane rates the evidence as very low quality and documents serious harms. The FDA carries a Boxed Warning for abuse and addiction.
The most useful finding in this new trial is the one getting the least attention: a structured, positive behavioral program, with no drug, substantially improved classroom behavior at no added cost. Children’s brains are still developing. Schools, parents, and lawmakers should not treat a short-term summer camp result as a reason to give children mind-altering drugs.
Questions for the Authors and Journal
- Will the authors publish the full weekly adverse-effect data, including sleep, appetite, mood, and cardiovascular measures?
- Which figures in the abstract are correct: 37% for medication and 61% for behavior therapy, or the reverse reported in the Results and Table 3?
- Why was the registered outcome (percentage of seatwork completed correctly) changed to a per-minute rate, and what do the results show using the registered measure?
- How many children under age six received methylphenidate, given the FDA label states safety and efficacy have not been established in that age group?
- Were staff or children asked to guess drug versus placebo, to test whether blinding held?
- Will JAMA Pediatrics issue a correction for the reported age of “8.8 months”?
Key Sources
- Fabiano GA, et al. ADHD Accommodations, Behavior Therapy, and Stimulant Medication: A Randomized Clinical Trial. JAMA Pediatrics, 2026. https://jamanetwork.com/journals/jamapediatrics/fullarticle/2856649
- gov registration NCT05469386. https://clinicaltrials.gov/study/NCT05469386
- FIU News release, November 13, 2026. https://news.fiu.edu/2026/common-adhd-classroom-accommodations-may-not-help-students-succeed-fiu-study-finds
- Pelham WE III, et al. The effect of stimulant medication on the learning of academic curricula in children with ADHD. Journal of Consulting and Clinical Psychology, 2022. https://pubmed.ncbi.nlm.nih.gov/35604744/
- Molina BSG, et al. The MTA at 8 years. JAACAP, 2009. https://pubmed.ncbi.nlm.nih.gov/19318991/
- NIMH summary of MTA 8-year results. https://www.nimh.nih.gov/news/science-news/2009/short-term-intensive-treatment-not-likely-to-improve-long-term-outcomes-for-children-with-adhd
- Swanson JM, et al. Young adult outcomes in the MTA follow-up: height suppression. JCPP, 2017. https://escholarship.org/uc/item/0j2621pr
- Currie J, Stabile M, Jones L. Do stimulant medications improve educational and behavioral outcomes for children with ADHD? Journal of Health Economics, 2014. https://pmc.ncbi.nlm.nih.gov/articles/PMC4815037/
- Raine ADHD Study, Government of Western Australia, 2010. https://www.health.wa.gov.au/~/media/Files/Corporate/Reports-and-publications/PDF/MICADHD_Raine_ADHD_Study_report_022010.pdf
- Storebø OJ, et al. Methylphenidate for children and adolescents with ADHD. Cochrane, 2023. https://pubmed.ncbi.nlm.nih.gov/36971690/
- Cochrane plain-language summary on unblinding risk. https://www.cochrane.org/evidence/CD009885_methylphenidate-effective-treatment-children-and-adolescents-attention-deficit-hyperactivity
- Storebø OJ, et al. Cochrane review in BMJ, 2015. https://www.bmj.com/content/351/bmj.h5203
- Cochrane review of methylphenidate harms. https://www.cochrane.org/evidence/CD012069_tjwyz-mtylfnydyt-bray-akhtlal-nqs-twjh-w-byshfaly-adhd-dr-kwdkan-w-nwjwanan-arzyaby-athrat-zyanawr
- Zhang L, et al. ADHD Medications and Long-Term Risk of Cardiovascular Diseases. JAMA Psychiatry, 2024. https://jamanetwork.com/journals/jamapsychiatry/fullarticle/2811812
- Moran LV, et al. Psychosis with Methylphenidate or Amphetamine in Patients with ADHD. NEJM, 2019. https://www.nejm.org/doi/full/10.1056/NEJMoa1813751
- FDA Concerta prescribing information, 2023. https://www.accessdata.fda.gov/drugsatfda_docs/label/2023/021121s049lbl.pdf
- FDA Drug Safety Communication on stimulant Boxed Warnings, 2023. https://www.fda.gov/drugs/drug-safety-communications/fda-updating-warnings-improve-safe-use-prescription-stimulants-used-treat-adhd-and-other-conditions
- FDA table of stimulant label changes. https://www.fda.gov/media/168050/download
- Layton TJ, et al. ADHD and Month of School Enrollment. NEJM, 2018. http://urisohn.com/sohn_files/wp/wordpress/wp-content/uploads/6013-Layton-Barnett-Hicks-Jena-NEJM-2018-ADHD-and-month-of-school-enrollment.pdf
- Harvard Gazette on the Layton study, 2018. https://news.harvard.edu/gazette/story/2018/11/when-starting-school-younger-children-are-more-likely-to-be-diagnosed-with-adhd-study-says/



0 Comments