Somebody has offered you forty sessions of brain training at a few hundred dollars each, and the testimonials are glowing, and the clinic has a study. The study probably exists. The question is who was rating the child, and whether they knew what the child had been given.

The short answer: for almost every non-drug treatment marketed for ADHD, the effect is substantial when the person rating the child knows which treatment they received, and shrinks or disappears when the rater is blinded. Across 38 randomised trials and 2,472 participants, neurofeedback's blinded effect on ADHD symptoms was a standardised mean difference of 0.04, with a confidence interval running from −0.10 to 0.18. Cognitive training improves the trained task and produces, at best, a small setting-specific effect on inattention. Restricted elimination diets do not survive blinding. Two things do: free fatty acid supplementation, at a small effect of 0.16, and artificial food colour exclusion, at 0.42 — mostly in children already selected for food sensitivity. This is not a claim that parents are imagining improvements. It is a specific, well-documented feature of how these treatments are evaluated.¹ ² ³

Why blinding is the whole story here

ADHD is measured by asking someone — usually a parent or teacher — to rate a child's behaviour. If that person knows the child has just finished a course of treatment, their rating changes. That is not dishonesty; it is how human observation works, and it is why drug trials use placebos.

The European ADHD Guidelines Group pooled 54 randomised trials of dietary and psychological treatments and reported both kinds of measurement side by side:¹

"When the outcome measure was based on ADHD assessments by raters closest to the therapeutic setting, all dietary (standardized mean differences=0.21-0.48) and psychological (standardized mean differences=0.40-0.64) 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."

And the conclusion: "Better evidence for efficacy from blinded assessments is required for behavioral interventions, neurofeedback, cognitive training, and restricted elimination diets before they can be supported as treatments for core ADHD symptoms."¹

Neurofeedback

The current best evidence is a 2025 meta-analysis of 38 randomised trials and 2,472 participants aged 5 to 40:²

  • Blinded ADHD total symptoms: standardised mean difference 0.04 (95% CI −0.10 to 0.18), across 20 trials and 1,214 participants. That interval crosses zero comfortably.
  • Restricting to trials that used established standard protocols produced a small significant effect: 0.21 (0.02 to 0.40) across 9 trials.
  • Of five neuropsychological outcomes, only processing speed improved, and only slightly: 0.35 (0.01 to 0.69).
  • The authors' summary: "Overall, neurofeedback did not appear to meaningfully benefit individuals with ADHD, clinically or neuropsychologically, at the group level."²

That replicates the earlier children-and-adolescents analysis, which found significant effects on the least-blinded ratings (total symptoms 0.35, inattention 0.36, hyperactivity/impulsivity 0.26) and concluded: "Evidence from well-controlled trials with probably blinded outcomes currently fails to support neurofeedback as an effective treatment for ADHD."³

The cleanest single test is a double-blind, sham-controlled trial — where neither the family nor the child knew whether the feedback was real. Its finding: "the study findings do not suggest that fMRI-NF of the rIFC is effective in improving clinical symptoms or cognition in boys with ADHD."⁴

What to ask a clinic selling it: "What did your trial find on blinded outcome measures, and was there a sham condition?" A clinic quoting parent-rated improvement from an unblinded study is quoting the number the field has spent a decade explaining.

Cognitive and working-memory training

Thirty-six randomised trials, analysed with blinded outcomes:⁵

  • No effect on ADHD total symptoms (SMD 0.12, 95% CI −0.01 to 0.25) or on hyperactivity/impulsivity (0.12, −0.03 to 0.28).
  • A small effect on inattention: 0.17 (0.02 to 0.31), which held with semi-active controls and doubled when measured in the setting where the training was delivered — which the authors describe as "suggesting a setting-specific effect."
  • Working memory itself improved: verbal 0.38, visual-spatial 0.49. Other neuropsychological and academic outcomes did not.
  • The conclusion: "Clinical effects were limited to small, setting specific, short-term effects on inattention symptoms."⁵

The predecessor analysis found the same shape: total ADHD symptoms 0.37 on unblinded ratings, falling to 0.20 on blinded ones, with inattention losing significance entirely — "Despite improving working memory performance, cognitive training had limited effects on ADHD symptoms according to assessments based on blinded measures."⁶

The honest summary: these programmes make children better at the trained task. That improvement does not reliably transfer to attention in a classroom.

Diets and supplements

This is where the picture is genuinely mixed, and two things do survive blinding.

Artificial food colour exclusion: SMD 0.42, significant under blinding.¹ The largest effect of any non-drug intervention in that analysis. Its caveat, in the authors' words: the effects were "often in individuals selected for food sensitivities" — so this describes a subgroup, not every child.

Free fatty acid supplementation (omega-3): SMD 0.16, significant under blinding.¹ Small. The authors: "the clinical significance of these effects remains to be determined." Cheap and low-risk, which is the argument for trying it; small and uncertain, which is the argument against expecting much.

Restricted elimination diets: significant unblinded, "substantially attenuated to nonsignificant levels" under blinding.¹ These are demanding — weeks of a severely restricted diet, then reintroduction — and the burden falls on a family whose life is already difficult. If you do one, do it with a dietitian and with someone blinded rating the outcome, and set a stopping rule in advance.

The exception worth taking seriously: behavioural parent training

Parent training shows the same collapse on ADHD symptoms — a standardised mean difference of 0.35 on unblinded parent ratings, and no blinded evidence of symptom reduction.⁷

But it is not in the same category as neurofeedback, because two of its effects do survive blinding, and they are effects worth having:⁷

  • Parenting quality: positive parenting 0.63, negative parenting 0.43
  • Child conduct problems: 0.31

That is a genuine, blinded, replicated improvement in how a household runs. It is also exactly what the paediatric guideline recommends as first-line treatment for four- and five-year-olds and as a component of treatment from six upward.⁸ The mistake is not in doing it; it is in expecting it to move an ADHD symptom score. What parent training does →

The summary table

Intervention Unblinded raters Blinded raters Worth doing?
Neurofeedback Significant (0.26–0.36 in children)³ 0.04 (−0.10 to 0.18)² Not on this evidence, at this cost
Cognitive / working-memory training 0.37 total symptoms⁶ 0.12 total; 0.17 inattention, setting-specific⁵ Improves the trained task; limited transfer
Restricted elimination diets Significant¹ Non-significant¹ High burden, unproven; only with a dietitian
Artificial food colour exclusion Significant¹ 0.42, significant¹ Plausible, mainly in food-sensitive children
Free fatty acids (omega-3) Significant¹ 0.16, significant¹ Small, cheap, low risk
Behavioural parent training 0.35 on ADHD symptoms⁷ No blinded ADHD effect; parenting 0.63 / 0.43, conduct 0.31⁷ Yes — for what it actually does

And the two things that are not on this list

Sleep. Not a treatment for ADHD, but half an hour less sleep a night measurably worsened restlessness and emotional lability in typically developing children, rated by blinded teachers.⁹ Fixing sleep is worth doing on its own terms, whatever the diagnosis. What else looks like ADHD →

School supports. The guideline calls educational interventions and individualised instructional supports "a necessary part of any treatment plan."⁸ They are not an alternative therapy; they are part of the treatment, and they are free. Request a school assessment → · 504 plans and IEPs →

Q&A

Q: Does neurofeedback work for ADHD? A: Not on blinded measurement. Across 38 randomised trials and 2,472 participants, the blinded effect on ADHD symptoms was 0.04 (95% CI −0.10 to 0.18), and the authors concluded that neurofeedback "did not appear to meaningfully benefit individuals with ADHD, clinically or neuropsychologically, at the group level."²

Q: But we saw a real improvement. A: Very possibly you did — children change over forty weeks, expectations shift, families do other things at the same time, and unblinded ratings capture all of that. The blinded trials exist to separate the treatment from everything else happening around it, and when they do, the treatment effect is what disappears.

Q: Are brain-training apps worth it? A: They improve the skill they train. The blinded effect on ADHD symptoms is 0.12 and non-significant; the small inattention effect doubles when measured in the room where the training happened, which is the definition of a setting-specific effect.⁵

Q: Should we try cutting out food dyes? A: It is the largest blinded effect among the non-drug options at 0.42, concentrated in children selected for food sensitivity.¹ Low risk and worth a trial; worth deciding in advance what would count as it working.

Q: What about omega-3? A: A small significant blinded effect of 0.16, with the authors noting its clinical significance is undetermined.¹ Cheap, low risk, modest expectations.

Q: Is there anything with a big effect that isn't medication? A: Not for core ADHD symptoms, on blinded evidence. What does have large blinded effects is medication — clinician-rated standardised mean differences of 0.78 for methylphenidate and 1.02 for amphetamines in children — over about twelve weeks.¹⁰ The medication decision →

Q: So is everything else a waste of time? A: No — but be clear what you are buying. Parent training genuinely improves parenting and conduct problems. School supports are part of the treatment plan. Sleep matters. What none of these reliably does is move a core ADHD symptom score, and a service selling that specific promise should be asked for its blinded data.

Ready to find a clinician who will tell you which is which? Filter by approach, schedule and payment route → · ADHD in children: the parent map → · ADHD therapy: what actually works →

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

Sources

  1. 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; "When the outcome measure was based on ADHD assessments by raters closest to the therapeutic setting, all dietary (standardized mean differences=0.21-0.48) and psychological (standardized mean differences=0.40-0.64) 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"; "Better evidence for efficacy from blinded assessments is required for behavioral interventions, neurofeedback, cognitive training, and restricted elimination diets" — doi.org.
  2. 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)"; the standard-protocol subgroup and the processing-speed finding; "Overall, neurofeedback did not appear to meaningfully benefit individuals with ADHD, clinically or neuropsychologically, at the group level" — pmc.ncbi.nlm.nih.gov.
  3. Cortese S, Ferrin M, Brandeis D, et al., "Neurofeedback for attention-deficit/hyperactivity disorder: meta-analysis of clinical and neuropsychological outcomes from randomized controlled trials," Journal of the American Academy of Child and Adolescent Psychiatry 55(6), 2016, 444–455 — 13 trials, 520 participants; "Effects were not significant when probably blinded ratings were the outcome or in trials with active/sham controls"; "Evidence from well-controlled trials with probably blinded outcomes currently fails to support neurofeedback as an effective treatment for ADHD" — doi.org.
  4. Lam SL, Criaud M, Lukito S, et al., "Double-blind, sham-controlled randomized trial testing the efficacy of fMRI neurofeedback on clinical and cognitive measures in children with ADHD," American Journal of Psychiatry 179(12), 2022, 947–958 — "the study findings do not suggest that fMRI-NF of the rIFC is effective in improving clinical symptoms or cognition in boys with ADHD" — doi.org.
  5. Westwood SJ, Parlatini V, Rubia K, Cortese S, Sonuga-Barke EJS, "Computerized cognitive training in attention-deficit/hyperactivity disorder (ADHD): a meta-analysis of randomized controlled trials with blinded and objective outcomes," Molecular Psychiatry 28(4), 2023, 1402–1414 — 36 randomised trials; the blinded effects on total symptoms, hyperactivity/impulsivity and inattention, the setting-specific doubling, and the working-memory findings; "Clinical effects were limited to small, setting specific, short-term effects on inattention symptoms" — doi.org.
  6. Cortese S, Ferrin M, Brandeis D, et al., "Cognitive training for attention-deficit/hyperactivity disorder: meta-analysis of clinical and neuropsychological outcomes from randomized controlled trials," Journal of the American Academy of Child and Adolescent Psychiatry 54(3), 2015, 164–174 — 16 records, 759 children; unblinded total ADHD 0.37 and inattention 0.47, falling to 0.20 and a non-significant 0.32 under probably blinded rating; "Despite improving working memory performance, cognitive training had limited effects on ADHD symptoms according to assessments based on blinded measures" — doi.org.
  7. 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; unblinded child ADHD 0.35; "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)" — 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 5a, 5b and 5c, and the description of educational interventions as "a necessary part of any treatment plan" — pmc.ncbi.nlm.nih.gov.
  9. Gruber R, Cassoff J, Frenette S, Wiebe S, Carrier J, "Impact of sleep extension and restriction on children's emotional lability and impulsivity," Pediatrics 130(5), 2012, e1155–e1161 — 34 typically developing children with teachers blinded to sleep condition — doi.org.
  10. 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 — clinician-rated standardised mean differences of −0.78 for methylphenidate and −1.02 for amphetamines in children and adolescents, over approximately 12 weeks — pmc.ncbi.nlm.nih.gov.

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