You have been handed a rating scale by a school, a prescription pad by a paediatrician, and by the internet a choice between drugging your child and failing them. None of those three is describing what the research actually found, and the research is unusually good here — one of the largest treatment trials ever run in child psychiatry has been following the same children for twenty-five years.

The short answer: about 11.7 percent of US children aged 3 to 17 have a current ADHD diagnosis, while the proportion who meet criteria on standardised assessment worldwide is around 5 percent and has not risen in three decades. Treatment guidelines split by age: behavioural parent training first for four- and five-year-olds, medication plus behavioural support for six to eleven, and medication with the adolescent's own assent after twelve. The landmark trial found medication clearly superior over fourteen months — and by three years, and again at eight, the originally assigned treatment groups no longer differed on anything. What predicted long-term outcome was the child's early symptom trajectory, not which treatment they were randomised to. That is not an argument against treatment. It is an argument for a different set of expectations than either side of the public argument offers.¹ ² ³ ⁴ ⁵

Four doors

Two numbers that are not the same number

Figure What it measures Source
11.7% of US children aged 3–17 with a current diagnosis, about 7.1 million A parent reporting that a clinician told them their child has ADHD, and that the child currently has it National Survey of Children's Health, 2024¹
~5% of children worldwide meeting diagnostic criteria Standardised assessment applied to a community sample, whether or not anyone ever diagnosed them Meta-regression of 135 studies²

The gap between them is a measure of diagnostic uptake — help-seeking, insurance, clinician thresholds, awareness — not of biology. And it is compatible with over-diagnosis, under-diagnosis and mis-diagnosis all happening at once, in different children.

What has not changed is the underlying rate. Across three decades of standardised community studies: "In the past three decades, there has been no evidence to suggest an increase in the number of children in the community who meet criteria for ADHD when standardized diagnostic procedures are followed."² Whatever is rising, it is not the condition.

Among children with a current diagnosis, 77.9 percent had at least one co-occurring disorder⁶ — which is why the evaluation guideline requires screening for other conditions rather than treating ADHD as a lone finding.

The trial everyone quotes and almost nobody reads

The Multimodal Treatment Study randomised 579 children aged 7 to 9.9 with combined-type ADHD to fourteen months of one of four things: careful medication management; intensive behavioural treatment; both; or routine community care.³

At fourteen months — the only randomised phase — medication management and combined treatment beat intensive behavioural treatment and community care on ADHD symptoms. Combined treatment did not significantly beat medication alone on core symptoms, but showed advantages on oppositional and internalising symptoms, teacher-rated social skills, parent–child relations and reading. The authors: "our carefully crafted medication management was superior to behavioral treatment and to routine community care that included medication."³

Note the last clause. The comparison group was not untreated: "two thirds of community-treated subjects received medication during the study period."³

At twenty-four months, ten months after treatment ended, the medication advantage persisted but "the effect size was reduced by 50% from the end of treatment to the first follow-up."⁷

At thirty-six months: "treatment groups did not differ significantly on any measure at 36 months." By then the behavioural group had gone from 14 to 45 percent medicated and the medication groups from 91 to 71 percent — the design had become entirely naturalistic. The authors' own explanation: "possibly due to age-related decline in ADHD symptoms, changes in medication management intensity, starting or stopping medications altogether, or other factors not yet evaluated."⁴

At six to eight years: "In nearly every analysis, the originally randomized treatment groups did not differ significantly… ADHD symptom trajectory in the first 3 years predicted 55% of the outcomes." The conclusion: "Type or intensity of 14 months of treatment for ADHD in childhood… does not predict functioning 6 to 8 years later. Rather, early ADHD symptom trajectory regardless of treatment type is prognostic."⁵

How to read all that. It does not say medication stops working — the fourteen-month randomised result stands, and the later comparisons are between groups whose treatment had long since become a matter of family choice. It does say that fourteen months of good treatment at age eight does not buy a different life at sixteen. What that argues for is treatment as ongoing management rather than as a course you complete, and for measuring whether it is still working rather than assuming it.

What the guideline recommends, by age

The American Academy of Pediatrics splits this three ways, and consumer summaries routinely get it backwards.⁸

Age What is recommended first
4 to 5 "evidence-based behavioral PTBM and/or behavioral classroom interventions as the first line of treatment, if available (grade A: strong recommendation). Methylphenidate may be considered if these behavioral interventions do not provide significant improvement and there is moderate-to-severe continued disturbance"
6 to 11 FDA-approved medication along with parent training and/or behavioural classroom intervention, "preferably both" — grade A for medication and grade A for the behavioural components
12 to 17 FDA-approved medication "with the adolescent's assent" (grade A), plus training or behavioural interventions where available

So behavioural treatment is first line only for the preschool years. From six upward the guideline recommends medication and behavioural support together, and after twelve it makes the adolescent's own agreement part of the recommendation.

Educational supports are described as "a necessary part of any treatment plan" at both school ages, "often includ[ing] an IEP or a rehabilitation plan (504 plan)."⁸ Requesting a school assessment → · 504 plans and IEPs →

An oddity worth knowing: amphetamine is the only ADHD drug with a US approval below age 6 — from age 3 — but the AAP notes that authorisation "was issued at a time when approval criteria were less stringent than current requirements," and the guideline's own preference for that age band is methylphenidate, which is off-label there.⁸ Approval status and evidence quality point in opposite directions.

What the medications do, and what the evidence covers

A network meta-analysis of 133 double-blind randomised trials — 10,068 children and adolescents in the efficacy analysis — found that on clinician ratings all the drugs beat placebo, with amphetamines strongest (standardised mean difference −1.02), then methylphenidate (−0.78) and atomoxetine (−0.56). On teacher ratings, only methylphenidate and modafinil beat placebo. On tolerability, amphetamines and guanfacine were worse than placebo in children.⁹

Its conclusion for this age group: "evidence from this meta-analysis supports methylphenidate in children and adolescents, and amphetamines in adults, as preferred first-choice medications for the short-term treatment of ADHD."⁹

Two things to hold onto. That recommendation rests on the efficacy-and-tolerability trade-off, not on methylphenidate being more effective — amphetamines beat it head to head on clinician-rated efficacy. And the whole evidence base is short-term: the authors state plainly, "We did not find sufficient data for the 26-week and 52-week timepoints."⁹ The full medication answer →

What everything else does

This is where a parent's time is most often spent and where the evidence is most consistently misreported. The pattern is the same across intervention after intervention: the effect is large when the person rating the child knows what treatment they got, and shrinks or vanishes when the rater is blinded.

Across 54 trials of dietary and psychological treatments: "When the outcome measure was based on ADHD assessments by raters closest to the therapeutic setting, all dietary… and psychological… treatments produced statistically significant effects. However, when the best probably blinded assessment was employed, effects remained significant for free fatty acid supplementation (standardized mean difference=0.16) and artificial food color exclusion (standardized mean difference=0.42) but were substantially attenuated to nonsignificant levels for other treatments."¹⁰

The most recent neurofeedback meta-analysis, 38 trials and 2,472 participants: "Probably blinded reports of ADHD total symptoms showed no significant improvement with neurofeedback… SMD, 0.04; 95% CI, −0.10 to 0.18."¹¹ The full answer on the alternatives →

Behavioural parent training is the important exception, and it needs stating carefully. Under blinded measurement it does not reduce ADHD symptoms. What it does reliably improve, blinded, is parenting quality (standardised mean differences 0.63 and 0.43) and child conduct problems (0.31).¹² Those are real, worth having, and not what it is usually sold as. What parent training does →

What happens later

The "half of children grow out of it" claim is wrong. Following 558 children in the MTA across sixteen years: about 30 percent experienced full remission at some point, but 60 percent of those had a recurrence; "Only 9.1% of the sample demonstrated recovery (sustained remission) by study endpoint, and only 10.8% demonstrated stable ADHD persistence." Most — 63.8 percent — had "fluctuating periods of remission and recurrence over time," and the authors state that the findings "challenge the notion that approximately 50% of children with ADHD outgrow the disorder by adulthood."¹³

The realistic picture is neither "they'll grow out of it" nor "this is permanent." It is a condition that waxes and wanes, in most children, for a long time.

Two findings that pull against each other, and both belong here. The randomised evidence says fourteen months of treatment in childhood did not change functioning six to eight years later.⁵ Large within-individual studies — comparing the same people to themselves during medicated and unmedicated months — find substantially lower rates of specific acute harms while medicated: motor vehicle crashes down 38 percent in males and 42 percent in females across 2.3 million patients,¹⁴ criminal convictions down 32 and 41 percent in a Swedish cohort,¹⁵ and lower all-cause mortality after medication initiation.¹⁶

Those answer different questions. Does a course of treatment at eight change life at sixteen? Apparently not. Is a person safer this month while treated? On the observational evidence, yes. A parent is entitled to both answers.

What to do next

  1. Get the evaluation the guideline describes, not a form. It requires "documentation of symptoms and impairment in more than 1 major setting" with information from parents, teachers and others, and a process to screen for co-occurring emotional, developmental and physical conditions.⁸ The printable prep sheet →
  2. Ask about your child's age within the school year. The relative-age effect is one of the most replicated findings in this literature. Why that matters →
  3. Ask what else was ruled out — sleep, hearing and vision, learning disorders, anxiety, adversity.
  4. If medication is proposed, agree in advance what would count as it working, and when you will review it. Titration to "maximum benefit with tolerable side effects" is itself a guideline recommendation.⁸
  5. Take the parent training for what it is good at. Better parenting confidence and fewer conduct problems is a large improvement in a household, whatever it does to a symptom score.
  6. Put the school supports in writing. Request a school assessment →

Q&A

Q: Is ADHD overdiagnosed in children? A: The diagnosed rate in the US, 11.7 percent, is roughly double the proportion meeting criteria on standardised community assessment, about 5 percent — but that gap measures diagnostic uptake, and is compatible with over- and under-diagnosis happening simultaneously in different groups.¹ ² What has not risen over three decades is the proportion of children who meet criteria when standard procedures are used.²

Q: Does my child have to take medication? A: For ages 4 to 5 the guideline's first line is behavioural parent training, with medication considered if that does not help enough. From 6 upward it recommends medication together with behavioural and educational support, and from 12 with the adolescent's assent.⁸ It is a recommendation, made with families, not a requirement.

Q: Will they need it forever? A: Unlikely to be a yes-or-no answer. In the sixteen-year follow-up, only about one in ten showed stable persistence throughout and only about one in ten a sustained recovery; most fluctuated.¹³ Which argues for reviewing rather than deciding once.

Q: Doesn't the MTA study show medication stops working? A: No. It shows that the difference between the originally assigned groups had gone by three years, in a cohort where treatment had become entirely a matter of family choice and had crossed over heavily in both directions.⁴ The randomised fourteen-month result is unchanged.

Q: Will treating it lead to substance abuse later? A: The MTA looked directly at this and found "no evidence that stimulant treatment was associated with increased or decreased risk for later frequent use of alcohol, marijuana, cigarette smoking, or other substances."¹⁷

Q: My child is bright and doing fine at school. Can it still be ADHD? A: Yes, and treatment response does not depend on IQ — a placebo-controlled crossover study in 502 children found "No significant differences… with regards to treatment response" by IQ level.¹⁸ Impairment can also be social or emotional rather than academic; the criteria ask about more than grades.

Ready to find a clinician who assesses children properly? Filter by approach, schedule and payment route → · Every adult ADHD decision on one map → · ADHD therapy: what actually works →

In crisis? Call or text 988 — free, 24/7.

Sources

  1. Health Resources and Services Administration, Maternal and Child Health Bureau, "Attention-Deficit/Hyperactivity Disorder (ADHD) prevalence and health care access, 2024," National Survey of Children's Health data brief — "In 2024, 11.7% of U.S. children ages 3–17, or an estimated 7.1 million children, had a current ADHD diagnosis"; "A higher percentage of boys than girls had a current ADHD diagnosis (14.5% vs. 8.8%)" — hrsa.gov; summarised at cdc.gov.
  2. Polanczyk GV, Willcutt EG, Salum GA, Kieling C, Rohde LA, "ADHD prevalence estimates across three decades: an updated systematic review and meta-regression analysis," International Journal of Epidemiology 43(2), 2014, 434–442 — 135 studies in the multivariate analysis; "In the past three decades, there has been no evidence to suggest an increase in the number of children in the community who meet criteria for ADHD when standardized diagnostic procedures are followed" — pmc.ncbi.nlm.nih.gov; Polanczyk G, de Lima MS, Horta BL, Biederman J, Rohde LA, "The worldwide prevalence of ADHD: a systematic review and metaregression analysis," American Journal of Psychiatry 164(6), 2007, 942–948 — doi.org.
  3. MTA Cooperative Group, "A 14-month randomized clinical trial of treatment strategies for attention-deficit/hyperactivity disorder," Archives of General Psychiatry 56(12), 1999, 1073–1086 — "A group of 579 children with ADHD Combined Type, aged 7 to 9.9 years"; "our carefully crafted medication management was superior to behavioral treatment and to routine community care that included medication"; "two thirds of community-treated subjects received medication during the study period" — doi.org.
  4. Jensen PS, Arnold LE, Swanson JM, et al., "3-year follow-up of the NIMH MTA study," Journal of the American Academy of Child and Adolescent Psychiatry 46(8), 2007, 989–1002 — "treatment groups did not differ significantly on any measure at 36 months"; the crossover figures; "possibly due to age-related decline in ADHD symptoms, changes in medication management intensity, starting or stopping medications altogether, or other factors not yet evaluated" — doi.org.
  5. Molina BSG, Hinshaw SP, Swanson JM, et al., "The MTA at 8 years: prospective follow-up of children treated for combined-type ADHD in a multisite study," Journal of the American Academy of Child and Adolescent Psychiatry 48(5), 2009, 484–500 — "In nearly every analysis, the originally randomized treatment groups did not differ significantly"; "ADHD symptom trajectory in the first 3 years predicted 55% of the outcomes"; "Type or intensity of 14 months of treatment for ADHD in childhood… does not predict functioning 6 to 8 years later" — pmc.ncbi.nlm.nih.gov.
  6. Danielson ML, Claussen AH, Bitsko RH, et al., "ADHD prevalence among U.S. children and adolescents in 2022: diagnosis, severity, co-occurring disorders, and treatment," Journal of Clinical Child and Adolescent Psychology 53(3), 2024, 343–360 — "77.9% had at least one co-occurring disorder" — doi.org.
  7. MTA Cooperative Group, "National Institute of Mental Health Multimodal Treatment Study of ADHD follow-up: changes in effectiveness and growth after the end of treatment," Pediatrics 113(4), 2004, 762–769 — "the effect size was reduced by 50% from the end of treatment to the first follow-up" — doi.org.
  8. Wolraich ML, Hagan JF Jr, Allan C, et al., "Clinical practice guideline for the diagnosis, evaluation, and treatment of attention-deficit/hyperactivity disorder in children and adolescents," Pediatrics 144(4), 2019, e20192528 — key action statements 1, 2, 3, 5a, 5b, 5c and 6; "amphetamine is the only medication with FDA approval for use in children younger than 6 years, this authorization was issued at a time when approval criteria were less stringent than current requirements" — pmc.ncbi.nlm.nih.gov.
  9. Cortese S, Adamo N, Del Giovane C, et al., "Comparative efficacy and tolerability of medications for attention-deficit hyperactivity disorder in children, adolescents, and adults: a systematic review and network meta-analysis," The Lancet Psychiatry 5(9), 2018, 727–738 — 133 trials; the standardised mean differences by drug and by rater; "evidence from this meta-analysis supports methylphenidate in children and adolescents, and amphetamines in adults, as preferred first-choice medications for the short-term treatment of ADHD"; "We did not find sufficient data for the 26-week and 52-week timepoints" — pmc.ncbi.nlm.nih.gov.
  10. Sonuga-Barke EJ, Brandeis D, Cortese S, et al., "Nonpharmacological interventions for ADHD: systematic review and meta-analyses of randomized controlled trials of dietary and psychological treatments," American Journal of Psychiatry 170(3), 2013, 275–289 — 54 records analysed; the blinded-versus-unblinded contrast quoted in full — doi.org.
  11. Westwood SJ, Aggensteiner PM, Kaiser A, et al., "Neurofeedback for attention-deficit/hyperactivity disorder: a systematic review and meta-analysis," JAMA Psychiatry 82(2), 2025, 118–129 — 38 randomised trials, 2,472 participants; "Probably blinded reports of ADHD total symptoms showed no significant improvement with neurofeedback (k = 20; n = 1214; SMD, 0.04; 95% CI, -0.10 to 0.18)" — pmc.ncbi.nlm.nih.gov.
  12. Daley D, van der Oord S, Ferrin M, et al., "Behavioral interventions in attention-deficit/hyperactivity disorder: a meta-analysis of randomized controlled trials across multiple outcome domains," Journal of the American Academy of Child and Adolescent Psychiatry 53(8), 2014, 835–847 — 32 randomised trials; "With probably blinded assessments, significant effects persisted for parenting (SMD for positive parenting 0.63; SMD for negative parenting 0.43) and conduct problems (SMD 0.31)"; "In contrast to the lack of blinded evidence of ADHD symptom decrease…" — doi.org.
  13. Sibley MH, Arnold LE, Swanson JM, et al., "Variable patterns of remission from ADHD in the Multimodal Treatment Study of ADHD," American Journal of Psychiatry 179(2), 2022, 142–151 — "Only 9.1% of the sample demonstrated recovery (sustained remission) by study endpoint, and only 10.8% demonstrated stable ADHD persistence"; "The MTA findings challenge the notion that approximately 50% of children with ADHD outgrow the disorder by adulthood" — pmc.ncbi.nlm.nih.gov.
  14. Chang Z, Quinn PD, Hur K, et al., "Association between medication use for attention-deficit/hyperactivity disorder and risk of motor vehicle crashes," JAMA Psychiatry 74(6), 2017, 597–603 — 2,319,450 patients; 38% lower risk in males and 42% in females during medicated months — pmc.ncbi.nlm.nih.gov.
  15. Lichtenstein P, Halldner L, Zetterqvist J, et al., "Medication for attention deficit-hyperactivity disorder and criminality," New England Journal of Medicine 367(21), 2012, 2006–2014 — "a significant reduction of 32% in the criminality rate for men… and 41% for women" — pmc.ncbi.nlm.nih.gov.
  16. Li L, Zhu N, Zhang L, et al., "ADHD pharmacotherapy and mortality in individuals with ADHD," JAMA 331(10), 2024, 850–860 — "ADHD medication initiation was associated with significantly lower rate of all-cause mortality (hazard ratio [HR], 0.79; 95% CI, 0.70 to 0.88)" — pmc.ncbi.nlm.nih.gov.
  17. Molina BSG, Kennedy TM, Howard AL, et al., "Association between stimulant treatment and substance use through adolescence into early adulthood," JAMA Psychiatry 80(9), 2023, 933–941 — "no evidence that stimulant treatment was associated with increased or decreased risk for later frequent use of alcohol, marijuana, cigarette smoking, or other substances" — doi.org.
  18. Grizenko N, Qi Zhang DD, Polotskaia A, Joober R, "Efficacy of methylphenidate in ADHD children across the normal and the gifted intellectual spectrum," Journal of the Canadian Academy of Child and Adolescent Psychiatry 21(4), 2012, 282–288 — 502 children, double-blind placebo-controlled crossover; "No significant differences were found with regards to treatment response" — pmc.ncbi.nlm.nih.gov.

Paid for by participating therapists. Inclusion is computed from availability data — never purchased. No ads, no data sold.