Can a Test Diagnose ADHD? EEG, Brain Imaging and Genetic Testing Explained

How ADHD is actually diagnosed

ADHD is a clinical diagnosis. It is made by a specialist clinician, against the DSM-5 or ICD-11 criteria, on the basis of a structured interview, standardised rating scales and information from more than one setting and ideally more than one informant. For adults that usually means a developmental history going back to childhood, school reports or a parent or sibling account where available, and a careful look at what else could explain the symptoms.

That last step matters. Poor concentration, restlessness and disorganisation are also features of depression, anxiety, sleep disorders, thyroid disease, substance use, trauma and several other conditions. A good assessment spends as much time ruling these in or out as it does counting ADHD symptoms.

NICE guideline NG87 sets this out for the NHS. It does not recommend EEG, brain imaging or genetic testing for diagnosis, and neither do the equivalent guidelines in the United States, Canada, Australia or Europe. This is not because the NHS cannot afford them. It is because none of them has been shown to add diagnostic accuracy to a properly conducted clinical assessment. A 2024 systematic review for the American Academy of Pediatrics, covering every diagnostic tool evaluated since 1980, reached the same conclusion.

People sometimes describe this process as subjective. It is better described as structured clinical judgement. Every test discussed below was validated by comparing it against exactly this kind of assessment, which remains the reference standard.

EEG and qEEG brain mapping

What it is. An electroencephalogram (EEG) records the brain's electrical activity through around 20 electrodes placed on the scalp. It is painless, safe and has been a standard tool in neurology for decades, mainly for epilepsy and sleep disorders. Quantitative EEG (qEEG) takes the same recording and runs it through software that compares your brain-wave frequencies against a database of other people, producing a colour-coded brain map.

The theta/beta ratio story. In the 1990s researchers noticed that groups of children with ADHD tended to show more slow theta waves and fewer fast beta waves, particularly over the front of the brain. The ratio between the two, the theta/beta ratio, was proposed as an ADHD biomarker. A device based on it, NEBA, was cleared by the US FDA in 2013 as an aid to assessment in children aged 6 to 17. It was never approved as a stand-alone test, was never approved for adults, and has since largely disappeared from practice.

What the evidence now shows. The early findings have not held up.

• A meta-analysis of nine studies covering 1,253 children with ADHD and 517 controls found that the difference in theta/beta ratio had shrunk steadily over the years, driven by rising ratios in the non-ADHD groups, and concluded it cannot be considered a reliable diagnostic measure.
• A study testing several different ways of calculating the ratio found that none of them distinguished children with ADHD from controls, and none correlated consistently with symptoms.
• A pooled analysis of children across three standardised cohorts concluded that the ratio has no diagnostic value for ADHD.
• The American Academy of Neurology reviewed the evidence and issued a practice advisory warning of a risk of significant harm from misdiagnosis, because the ratio has an unacceptably high false-positive rate: many other conditions, and even ordinary drowsiness, raise frontal theta.

What about AI and machine learning? You may see headlines about EEG models that detect ADHD with 95% or 99% accuracy. These studies are real, but they are almost all in children, use small single-site research datasets, and test the model on recordings from the same dataset it learned from. None has been validated in adults, or in the real clinical situation where ADHD must be told apart from anxiety, depression or poor sleep rather than from healthy volunteers. The only systematic review of EEG in adult ADHD concluded that it is not an appropriate diagnostic tool.

qEEG brain maps and differential diagnosis. Some clinics claim a qEEG map can distinguish ADHD from autism, anxiety or trauma. There is no validated qEEG signature for any of these. Normative-database brain maps are prone to false findings because they test hundreds of comparisons at once, and the result is heavily affected by medication, caffeine, drowsiness and how clean the recording was.

Where EEG does have a role. A standard clinical EEG is useful when the history raises a specific alternative, most often absence seizures, which can look like inattention in a child. That is a question for a neurologist, and it is about excluding epilepsy, not confirming ADHD. In research, EEG remains a legitimate tool for studying brain development and treatment response. Neither of these uses supports selling EEG as part of an ADHD diagnosis.

Brain imaging: MRI, fMRI, PET and SPECT

What it is. Structural MRI measures the size and shape of brain regions. Functional MRI (fMRI) tracks blood-flow changes while you rest or do a task. PET and SPECT inject a radioactive tracer to measure blood flow or, in some research studies, dopamine transporter levels. All of these are excellent research tools and some have real clinical uses in neurology. None is a test for ADHD.

What imaging research has found. ADHD is associated with small differences in brain structure and function, visible when large groups are compared. The best evidence comes from the ENIGMA consortium, which pooled MRI scans from 1,713 people with ADHD and 1,529 controls across 23 sites. It found slightly smaller volumes in the amygdala, accumbens, caudate, hippocampus and putamen, and slightly smaller total brain volume. These are genuine findings and they help us understand ADHD as a condition of delayed brain maturation.

Why that does not make a test. The differences are very small. The effect sizes ranged from 0.10 to 0.19 standard deviations, meaning the ADHD and non-ADHD groups overlap almost completely. If you scanned a person with ADHD and a person without, you could not tell from the scan which was which. The differences were also largest in children and were not detectable in adults as a group. A later ENIGMA analysis found that the same regions are also smaller in autism and OCD, so the pattern is not even specific to ADHD.

This is the gap between a group difference and a diagnostic test. Men are on average taller than women, but height does not tell you someone's sex. Brain volume in ADHD is a far weaker signal than that.

AI applied to MRI. When researchers have tried to train algorithms to classify ADHD from MRI, the results have been sobering. In the ADHD-200 global competition, which released scans from nearly 1,000 children for exactly this purpose, the best imaging-based classifiers reached only around 60% accuracy, and the highest-scoring entry used no imaging at all, just age, sex and IQ, reaching 62.5%. Later work with larger datasets has improved the numbers but not to a level that would be acceptable in a clinic.

SPECT scans and brain-imaging clinics. A small number of private clinics, most prominently in the United States, offer SPECT scans as part of ADHD assessment. SPECT involves an injection of radioactive tracer and a radiation dose. No independent, blinded study has shown that reading a SPECT scan improves the accuracy of an ADHD diagnosis, and the published work comes almost entirely from the clinics selling the service. The American Academy of Child and Adolescent Psychiatry and the American Psychiatric Association have both stated that brain imaging has no role in diagnosing ADHD.

Where imaging does have a role. An MRI is appropriate when the history or examination suggests a structural problem, for example attention difficulties that began after a head injury, with new neurological signs, or in later life where a memory service may need to exclude a tumour or stroke. Again, that is about excluding something else, not confirming ADHD.

Genetic testing

ADHD is strongly genetic. Twin and family studies put the heritability of ADHD at around 70 to 80%, which is higher than for most psychiatric conditions. If a parent or sibling has ADHD, your own likelihood is several times higher than average. This is one reason a family history forms part of a good assessment.

But there is no ADHD gene. The largest genetic study to date combined DNA from 38,691 people with ADHD and 186,843 controls. It found 27 regions of the genome associated with ADHD, each with a tiny effect, and estimated that thousands of common variants together account for the genetic risk. ADHD is polygenic in the same way height is: thousands of variants each nudge the risk slightly. No single variant is necessary or sufficient, and no variant is specific to ADHD.

Polygenic risk scores. It is possible to add up all of a person's small risk variants into a single polygenic risk score. In research, these scores work: people with ADHD have higher scores on average, and the score also predicts related traits such as body mass index and neuroticism. But it explains only a few per cent of who does and does not have ADHD. A 2024 Norwegian study tested whether adding the polygenic score to standard rating scales improved identification of adult ADHD. It added about 0.6 of a percentage point to the variance explained by the rating scales, and the authors concluded that it is not currently a clinically useful diagnostic aid. Polygenic scores are also calibrated almost entirely on people of European ancestry and perform worse in everyone else.

Direct-to-consumer DNA tests. Some online services claim to report your ADHD genes, usually naming a handful of variants in dopamine-related genes such as DRD4 or DAT1. The associations these tests rely on are weak, inconsistent and were mostly reported before large genome-wide studies showed how small each individual effect is. A result from one of these tests tells you nothing useful about whether you have ADHD.

Pharmacogenomic testing. A separate type of test looks at how your liver metabolises medicines. This has a limited but real role: people who metabolise atomoxetine slowly through the CYP2D6 enzyme may need lower doses, and this is noted in the prescribing information. No pharmacogenomic test reliably predicts who will respond to methylphenidate or amfetamine, and none should be used to decide whether someone has ADHD in the first place.

When genetic testing is appropriate. A clinical genetics referral, usually for chromosomal microarray or exome sequencing, is warranted when ADHD symptoms occur alongside intellectual disability, developmental regression, seizures, unusual physical features or a known genetic syndrome in the family. That is to look for a rare syndrome that may need its own management. It is not a test for ADHD.

Why this matters

A test that is sold as objective but is not accurate does more harm than no test at all.

• False positives. EEG ratios and brain maps are raised by tiredness, anxiety, medication and poor sleep. A person with depression or a sleep disorder can be handed an ADHD label and a stimulant prescription while the real problem goes untreated.
• False negatives. Many people with clear-cut ADHD have a normal EEG, a normal MRI and an unremarkable polygenic score. Told that the test is negative, they may give up on an assessment that would have helped them.
• Cost. Private qEEG, SPECT and genetic panels typically add several hundred to several thousand pounds to an assessment without changing the outcome.
• Misplaced confidence. A colour-coded brain map feels more scientific than a clinical interview. It is not. A clinical interview is what every one of these tests was measured against.

The same applies to the current generation of computerised attention tests such as QbTest. NICE allows them as an aid that may support a clinician's judgement, and some clinics find them useful for monitoring medication. They are not diagnostic on their own, and a normal result does not exclude ADHD, particularly in adults and in the inattentive presentation.

The honest position is that ADHD research is working hard on biomarkers and may one day produce something clinically useful. Large multi-site studies, better handling of comorbidity and proper external validation are the route there. Buying a scan today is not.

In summary

• ADHD is diagnosed clinically. No scan, EEG or DNA test can confirm or exclude it.
• EEG theta/beta ratio and qEEG brain maps have been tested extensively and do not reliably distinguish ADHD from other conditions. Machine-learning results are promising in children but unvalidated in adults and in real clinical populations.
• MRI shows real but tiny group differences that overlap almost entirely with the general population and with autism and OCD. SPECT has no independent evidence base.
• ADHD is highly heritable but polygenic. Polygenic scores explain a few per cent of risk and add almost nothing to rating scales. Consumer DNA tests for ADHD are not valid.
• Each of these tools has a legitimate role in research, and sometimes in excluding other conditions. None has a role in diagnosing ADHD.

How we approach this at the Eton ADHD Clinic

We do not offer EEG, qEEG, brain scans or genetic tests as part of ADHD diagnosis, and we will tell you plainly if another provider has sold you one that changes nothing. What we do offer is a thorough, NICE-concordant assessment by a consultant psychiatrist: a structured diagnostic interview, validated rating scales, a developmental and family history, collateral information where you can provide it, and a careful search for the conditions that mimic or accompany ADHD, including autism. Where the history raises a specific medical or neurological question, we will say so and refer you for the right investigation. As a CQC-registered provider we are also bound to advertise only what the evidence supports; if you read a claim about ADHD testing on any clinic's website, including ours, and want to know what sits behind it, ask. We assess children and adults across England and Wales, privately and under our contract with NHS Wales, with a children's and adolescents' service in London and in Manchester at Alderley Edge.

Six reasons people choose the Eton ADHD Clinic

1. A NICE-concordant assessment by a consultant psychiatrist, which is the reference standard every proposed test is measured against.
2. No tests sold that do not change the outcome, and a plain account of the evidence if someone else has sold you one.
3. A careful search for the conditions that mimic ADHD, including sleep disorders, anxiety, depression and thyroid disease.
4. ADHD and autism assessed together, so the whole picture is seen.
5. Referral for the right investigation when the history genuinely calls for an EEG, a scan or a genetics opinion.
6. NHS and private care to the same standard, including our contract with NHS Wales and Right to Choose referrals in England.

Where to go next

• Adult ADHD assessment and treatment: https://etonpsychiatrists.co.uk/services/adhd-adult
• Child and adolescent ADHD assessment: https://etonpsychiatrists.co.uk/services/adhd-child-adolescent
• How to choose a reliable ADHD assessment clinic: https://etonpsychiatrists.co.uk/blog/how-to-choose-a-reliable-adhd-assessment-clinic-complete-guide
• ADHD or bipolar? How to tell the difference: https://etonpsychiatrists.co.uk/blog/adhd-or-bipolar-how-to-tell-the-difference
• OCD, ADHD or autism? Telling them apart: https://etonpsychiatrists.co.uk/blog/ocd-adhd-or-autism-and-adhd-treatment

Sources

• Adamou, M., Fullen, T. & Jones, S.L. EEG for diagnosis of adult ADHD: a systematic review with narrative analysis. Frontiers in Psychiatry 11, 871 (2020). https://doi.org/10.3389/fpsyt.2020.00871
• Arns, M., Conners, C.K. & Kraemer, H.C. A decade of EEG theta/beta ratio research in ADHD: a meta-analysis. Journal of Attention Disorders 17(5), 374–383 (2013). https://doi.org/10.1177/1087054712460087
• Boedhoe, P.S.W., van Rooij, D., Hoogman, M. et al. Subcortical brain volume, regional cortical thickness, and cortical surface area across disorders: findings from the ENIGMA ADHD, ASD, and OCD working groups. American Journal of Psychiatry 177(9), 834–843 (2020). https://doi.org/10.1176/appi.ajp.2020.19030331
• Boxum, M., Voetterl, H., van Dijk, H. et al. Challenging the diagnostic value of theta/beta ratio: insights from an EEG subtyping meta-analytical approach in ADHD. Applied Psychophysiology and Biofeedback 50(4), 655–666 (2025). https://doi.org/10.1007/s10484-024-09649-y
• Brown, M.R.G., Sidhu, G.S., Greiner, R. et al. ADHD-200 Global Competition: diagnosing ADHD using personal characteristic data can outperform resting state fMRI measurements. Frontiers in Systems Neuroscience 6, 69 (2012). https://doi.org/10.3389/fnsys.2012.00069
• Demontis, D., Walters, G.B., Athanasiadis, G. et al. Genome-wide analyses of ADHD identify 27 risk loci, refine the genetic architecture and implicate several cognitive domains. Nature Genetics 55, 198–208 (2023). https://doi.org/10.1038/s41588-022-01285-8
• Du Rietz, E., Coleman, J., Glanville, K., Choi, S.W., O'Reilly, P.F. & Kuntsi, J. Association of polygenic risk for attention-deficit/hyperactivity disorder with co-occurring traits and disorders. Biological Psychiatry: Cognitive Neuroscience and Neuroimaging 3(7), 635–643 (2018). https://doi.org/10.1016/j.bpsc.2017.11.013
• Gloss, D., Varma, J.K., Pringsheim, T. & Nuwer, M.R. Practice advisory: the utility of EEG theta/beta power ratio in ADHD diagnosis. Report of the Guideline Development, Dissemination, and Implementation Subcommittee of the American Academy of Neurology. Neurology 87(22), 2375–2379 (2016). https://doi.org/10.1212/WNL.0000000000003265
• Høberg, A., Solberg, B.S., Hegvik, T.A. et al. Using polygenic scores in combination with symptom rating scales to identify attention-deficit/hyperactivity disorder. BMC Psychiatry 24, 471 (2024). https://doi.org/10.1186/s12888-024-05925-7
• Hoogman, M., Bralten, J., Hibar, D.P. et al. Subcortical brain volume differences in participants with attention deficit hyperactivity disorder in children and adults: a cross-sectional mega-analysis. The Lancet Psychiatry 4(4), 310–319 (2017). https://doi.org/10.1016/S2215-0366(17)30049-4
• National Institute for Health and Care Excellence. Attention deficit hyperactivity disorder: diagnosis and management. NICE guideline NG87 (2018, updated 2019). https://www.nice.org.uk/guidance/ng87
• Peterson, B.S., Trampush, J., Brown, M. et al. Tools for the diagnosis of ADHD in children and adolescents: a systematic review. Pediatrics 153(4), e2024065854 (2024). https://doi.org/10.1542/peds.2024-065854
• van Dijk, H., deBeus, R., Kerson, C. et al. Different spectral analysis methods for the theta/beta ratio calculate different ratios but do not distinguish ADHD from controls. Applied Psychophysiology and Biofeedback 45(3), 165–173 (2020). https://doi.org/10.1007/s10484-020-09471-2

Frequently Asked Questions