ADHD

Attention-deficit/hyperactivity disorder, usually called ADHD, describes persistent difficulties with attention, activity and impulse control that interfere with everyday life. The difficulties begin during development, although they may not be recognised until adulthood. People differ greatly in their strengths, the problems they experience and the support they need. Occasional distraction is common; a diagnosis concerns a lasting pattern and its consequences across important parts of life. Assessment needs a full clinical and developmental history, including functioning and circumstances in different settings. A rating scale alone is insufficient. NICE guidance on assessing ADHD

Researchers agree that inherited differences make a substantial contribution. Some early experiences and developmental exposures also matter. They have identified differences in brain development, attention, motivation and alertness, but no single biological change explains everybody's ADHD. Knowing that something contributes to a condition is different from knowing every step through which it affects a person's life. That distinction helps us understand both the science and the limits of familiar explanations such as “too little dopamine”. Causes increase or decrease probabilities; neither inherited susceptibility nor an exposure establishes who is to blame for an individual's difficulties. Scientific overview of ADHD

Inherited susceptibility does not mean a fixed future

ADHD often runs in families. Research involving twins, families, adoption and DNA converges on a substantial inherited contribution. A review of 37 twin studies put the average estimate of heritability at about 74%. That figure concerns differences between people in the populations studied. It does not mean that 74% of one person's ADHD is genetic, that a genetic explanation can be found in 74% of diagnosed people, or that changing the environment cannot help. The remaining percentage is not a list of known environmental causes. Review of ADHD genetics

Many common DNA differences each make a small contribution. A major 2025 analysis combined symptom measurements from 70,953 people with genetic findings from diagnosed ADHD. It identified 39 associated regions of DNA, including 17 newly identified regions. This supports continuity between attention-related traits across the population and more disabling difficulties. It does not make ordinary distractibility and clinically important ADHD interchangeable. Nor does an association identify the precise biological step responsible. The 2025 genetic study

Less common genetic changes can have larger effects for the people who carry them. A study published in 2026 compared 8,895 people with ADHD with 53,780 others and implicated rare damaging changes in three genes: MAP1A, ANO8 and ANK2. The biological clues concerned the structure of nerve cells, their connections and how they process genetic instructions. Some overlap with other developmental conditions. These findings do not identify one “ADHD gene”, explain most cases or provide a routine genetic diagnostic test. A large effect in a small group can explain only a small share of ADHD overall. Research on rare genetic changes

Early development and the environment

Not all reported environmental links are equally convincing. Very early birth is among the better-supported developmental risks. In a Swedish study of more than four million people, recorded ADHD affected 12.1% of those born at 22–27 weeks, compared with 4.5% of those born at 39–41 weeks. Comparing siblings changed the association little, making shared family characteristics an incomplete explanation. Yet early birth involves several connected influences, including interrupted development and neonatal illness; the study cannot isolate one mechanism. The figures also mean that almost 88% of the extremely preterm group did not have recorded ADHD. Increased likelihood is not inevitability. Prematurity and later ADHD

Severe early deprivation can provide another route. Children adopted into UK families after prolonged deprivation in Romanian institutions had more persistent ADHD difficulties into young adulthood. This unusually informative study concerns extreme institutional conditions, a relatively small sample and several accompanying developmental problems. It cannot be used to blame ordinary differences in parenting, attachment or family organisation. The English and Romanian Adoptees study

Childhood lead exposure also has consistent links with later ADHD in studies that measured exposure before assessing symptoms. A 2024 review covered 66 studies of chemical exposures, including 22 on lead. Nevertheless, poverty, other exposures and measurement differences complicate ADHD-specific causal estimates. The studies in that review had been published by 2019: a recent review does not automatically contain recent underlying data. Review of chemical exposures and ADHD

Pregnancy studies show why association and cause must be separated. Smoking during pregnancy is often associated with later ADHD, but the association weakens markedly in studies designed to account for inherited and shared family influences. Parental characteristics can contribute to both smoking and a child's susceptibility. This does not make smoking safe in pregnancy; it limits the particular claim that smoking directly causes ADHD. Review of prenatal exposures using stronger family comparisons

The same caution applies to paracetamol. A 2026 review of 43 studies found no increased ADHD association in sibling comparisons or in analyses restricted to studies with low risk of bias. The sibling-comparison estimate was 0.95, with an uncertainty range from 0.86 to 1.05; on this relative scale, 1 means no association. This does not settle every possible dose or exposure pattern, but it does not support ordinary prenatal paracetamol exposure as an established cause of ADHD. Review of prenatal paracetamol and development

Trauma also needs careful assessment. Prospective studies link childhood maltreatment and ADHD, but the direction is not consistently established. Adversity may precede difficulties, while existing child and family difficulties can also affect subsequent adversity. ADHD and trauma can coexist and both deserve attention. The evidence does not support explaining ADHD in general as a trauma disorder. Review of maltreatment and ADHD

What brain research can and cannot tell us

Large scanning studies find small average differences between people with and without ADHD. In the ENIGMA collaboration, involving 3,242 participants, several structures beneath the brain's outer surface were slightly smaller on average in the ADHD group. Differences were more apparent in children. They were roughly one-tenth to one-fifth of a standard deviation: a measure of the spread of observations, not a percentage loss of brain tissue. The groups overlapped greatly. An individual scan cannot establish an ADHD diagnosis. Large study of brain structure

A later study brought together six datasets and 8,433 young people. It found differences in how deeper and outer brain regions were connected, with the largest reported peak difference only 0.15 standard deviations. That does not mean a 15% abnormality. For perspective, if two otherwise similar bell-shaped distributions differed by that amount, they would overlap by about 94%. This is an illustration of scale, not the study's measured diagnostic accuracy. Large study of brain connectivity

Researchers therefore examine interacting systems for attention, maintaining goals, internal thought, movement, motivation and alertness. A theory confined to one small “attention centre” is unlikely to explain the variety of ADHD. Small differences in individual measurements can still coexist with major daily difficulties, because each measurement captures only one aspect of a complex developing system. Most scans are taken after difficulties are established, making it hard to distinguish earlier causes, later consequences and accompanying influences. Head movement during scanning, previous medication, accompanying conditions and how participants were recruited can also affect results. Review of ADHD brain mechanisms

Dopamine and noradrenaline are chemical messengers involved in motivation, alertness and regulating behaviour. Effective medicines act on these systems: methylphenidate changes the removal of both messengers from the spaces between nerve cells, while atomoxetine mainly affects noradrenaline. But improving symptoms by changing a system does not prove that a shortage in that system originally caused the condition. Studies do not establish one generally low-dopamine state shared by everyone with ADHD. Results vary with age, brain region, measurement and previous medication. Review of the dopamine explanation

A recent study found stimulant-associated changes particularly in systems connected with alertness and reward. Its childhood analysis was observational, while its detailed placebo comparison involved only five healthy adults. It offers clues about medication action, not a final explanation of ADHD or proof that attention systems are irrelevant. Research on stimulant effects

Attention, motivation and changing demands

On average, ADHD groups have more difficulty holding information in mind, sustaining attention, stopping an immediate response, planning and performing consistently. A review of 34 earlier reviews found an average difference of about 0.45 standard deviations across cognitive tests, with substantial variation. This is not a 45% reduction in ability. No single test difficulty appears in everybody. Review of cognitive findings

One particularly repeatable finding is greater variation from one response to the next. A review of 319 studies found a larger difference in children and adolescents than adults: approximately 0.76 versus 0.46 standard deviations. However, the distinction was much weaker when ADHD was compared with other clinical conditions rather than typically developing groups. Inconsistent performance is informative, but not unique to ADHD. Review of response-time variability

Motivation and reward also matter. ADHD groups, on average, more often choose smaller immediate rewards over larger delayed ones. This is one reason researchers consider several overlapping explanations: difficulty organising behaviour around a goal, differences in responding to rewards and delay, and difficulty maintaining an effective level of alertness. A particular person's problems may involve different combinations. Laboratory tests can help describe difficulties, but cannot independently diagnose ADHD or establish their cause. International evidence statement

The pattern can change over time. In repeated follow-up of 558 participants from a major childhood study, approximately 64% moved between periods of meeting and not meeting full diagnostic criteria. That cohort began with combined-type childhood ADHD, so its percentage is not a forecast for everyone. It does show why an interval of improvement is different from permanent resolution. Research on changing ADHD symptoms

Adult diagnosis may identify difficulties previously missed, below a diagnostic threshold or manageable with earlier support. Greater independence or new work and family responsibilities may expose them. Whether apparently new adult ADHD without developmental precursors is the same condition remains debated. Assessment needs developmental history and consideration of other explanations. Among children, being younger than classmates is associated with more diagnosis and medication, particularly in teacher-based assessments. Expectations influence recognition, but that finding cannot tell us which individual diagnoses are wrong. Adult ADHD review; review of age within the school year

What medication helps—and what remains uncertain

Medication has the clearest and most consistent evidence for reducing core ADHD symptoms over weeks or months. This is a stronger conclusion than saying it reliably improves qualifications, employment, relationships, health and quality of life over decades. Those are different outcomes and need their own evidence.

A major comparison of 133 randomised trials found that stimulants and atomoxetine improved symptoms relative to placebo. Around 12 weeks, methylphenidate's clinician-rated effect in children and adolescents was approximately 0.78 standard deviations, favouring treatment. Amphetamines produced a larger average symptom effect, but methylphenidate offered a better overall balance of benefit and acceptability as a childhood starting option. Such a score does not mean 78% improvement or that 78% of patients respond. Comparative medication review

A 2025 comparison of 113 adult trials, involving 14,887 people, found short-term benefit for stimulants and atomoxetine on both patient and clinician ratings. Patient-rated differences were about 0.39 and 0.38 standard deviations respectively; clinician-rated differences were 0.61 and 0.51. These are averages, with uncertainty, and stimulants were pooled together rather than establishing one as best for everyone. Comparisons made indirectly across separate trials also depend on how similar their participants and methods were. More people stopped atomoxetine for any reason than stopped placebo. Adult treatment comparison

Reviews disagree about how confidently to estimate the size of benefit. The December 2025 Cochrane methylphenidate review, covering 212 trials but drawing on searches completed in March 2022, found a teacher-rated effect similar to other reviews, about 0.74 standard deviations. Yet it judged the evidence very uncertain because of bias, incomplete reporting, short trials and the possibility that noticeable side effects revealed who received active medicine. A 2025 umbrella review judged some short-term findings more favourably, using a different grading method, but found no high-certainty long-term evidence. These disagreements justify caution about precise promises; they have not displaced the recommendation to offer monitored medication when ADHD causes important impairment. Cochrane methylphenidate review; review of benefits and harms across treatments

Choosing treatment according to age and need

UK guidance starts with ADHD-focused parent training and changes to the environment for children under five. Medication at that age requires additional specialist consideration. For children aged five and over, methylphenidate is the usual first medicine when medication is indicated. Lisdexamfetamine may follow an inadequate trial; atomoxetine or guanfacine are options when suitable stimulant trials are unsuccessful or poorly tolerated. Adults usually start with lisdexamfetamine or methylphenidate; atomoxetine is an alternative when stimulants are unsuitable or unsuccessful. Dexamfetamine may be considered when lisdexamfetamine helps but its longer duration cannot be tolerated. Guanfacine is not a routine adult option under NICE. These pathways guide a clinical discussion rather than requiring medication for everyone with a diagnosis. NICE treatment pathways

American guidance groups ages differently, recommending parent training or behavioural classroom work first at ages four to five, and medication alongside behavioural and educational support for school-age children. The differing boundaries do not represent a fundamental disagreement about caution in very young children. American Academy of Pediatrics guideline

Finding an effective dose means balancing benefit and unwanted effects. Preferences about appetite, sleep, treatment effort and how long daily benefit lasts are part of that decision. A 2026 review of 113 trials found different dose–response patterns by medicine and age. For some drugs, extra benefit levels off while stopping because of side effects becomes more likely. An incomplete response therefore does not justify endlessly increasing the dose. Review of medication dose and effects

Skills, family support and practical adjustments

Behavioural parent training helps caregivers use clear expectations, predictable routines, encouragement and consistent responses. It can improve behaviour, parenting confidence and family relationships as well as producing more modest symptom improvements. A review of 27 randomised studies found benefits persisting an average of 5.3 months after treatment. That supports an active treatment with useful effects, not a promise of permanent change after one course. Review of parent training

For adults, structured ADHD-focused cognitive behavioural therapy, or CBT, helps with planning, starting and completing tasks, managing distraction and responding to difficult feelings. A 2026 review of 70 studies, including 38 randomised trials, found improvements in functioning, including work-related functioning. However, much relied on self-report, mixed study designs and short follow-up; quality-of-life findings were less convincing. Another major comparison found benefit on clinician symptom ratings without a corresponding established benefit on patient ratings. CBT can be useful without assuming every rater or life domain improves equally. Review of CBT and everyday functioning; adult treatment comparison

Reducing unnecessary distraction, giving clear instructions and adjusting school or workplace demands can also reduce avoidable difficulty. Guidelines recommend such changes, although individual adjustments have less extensive trial evidence than medication. Success may mean fewer arguments, more completed tasks or better participation even when a symptom score changes little. Australian ADHD guideline

Combining treatments makes sense when they address different needs. In the original 14-month study of 579 children, carefully managed medication and combined medication–behavioural care improved core symptoms more than behavioural treatment or community care. Combination did not significantly outperform medication management in their direct comparison, though some broader comparisons favoured it. A later youth review likewise did not establish a consistent extra core-symptom benefit from adding psychosocial care: the estimate from seven studies was uncertain and compatible with no additional benefit. This does not mean additional help is wasted. Improved concentration may leave family conflict, learning difficulties, anxiety or work routines unresolved. Each additional intervention should have a defined purpose. The original multimodal treatment trial; review of youth treatments

Exercise, sleep and other approaches

Exercise supports health and may provide additional ADHD benefits. A 2026 adult review found small immediate symptom improvements after exercise, but mixed evidence for sustained exercise programmes. Sleep problems deserve assessment and treatment because they add to impairment. Better sleep is valuable without establishing that every ADHD difficulty has been treated. Review of exercise in adult ADHD; Australian guidance

Evidence does not put supplements or training products alongside established first-line care. A Cochrane review found no material overall improvement in parent-rated ADHD symptoms from omega-3 and related fatty-acid supplements. A rigorous neurofeedback review found an overall effect close to zero when ratings were probably made without knowing treatment allocation, although some protocol-specific analyses found small effects. Computerised cognitive training can improve practised abilities, especially working memory, but transfer to overall symptoms or school achievement is limited. Becoming better at a computer task is not automatically becoming better organised in daily life. Supplement review; neurofeedback review; cognitive-training review

Longer-term benefits, side effects and monitoring

Large observational studies offer encouraging evidence beyond symptom scores. A Swedish analysis of 148,581 people aged 6–64 found that starting medication was associated with lower rates of first recorded suicidal behaviour, substance misuse, transport accidents and criminality over two years, but no clear reduction in first accidental injuries. The relative rates for the favourable findings were about 12–17% lower. Researchers designed the analysis to resemble a trial, but did not randomly assign treatment; remaining differences between treated and untreated people could affect the results. Study of serious everyday outcomes

An October 2026 review found medication use associated with about a quarter lower relative mortality over the studied follow-up. That treatment estimate came from three observational studies, not all 19 studies in the wider mortality review. It cannot prove that medication caused the reduction, establish lifelong effects or show benefits continue after stopping. Mortality review

Benefits must be weighed against appetite loss, sleep disturbance, gastrointestinal problems and other unwanted effects. Atomoxetine carries a warning about suicidal thoughts in younger patients. Guanfacine can cause sleepiness and lower blood pressure. A 2025 review of 102 trials found that several stimulant and non-stimulant medicines increased pulse or blood pressure, usually modestly on average. Its median follow-up was only seven weeks. Longer observational research linked more than five years' cumulative exposure with higher odds of cardiovascular disease, especially high blood pressure and arterial disease; the estimate was about 23% higher than no exposure, with uncertainty and possible remaining confounding. This is a safety signal, not proof of that causal increase for an individual. Paediatric medication guidance; cardiovascular trial review; longer-term cardiovascular study

Rare psychiatric effects also matter. In a study of people aged 13–25, psychosis meeting a diagnostic-and-prescription definition occurred in 0.10% receiving methylphenidate and 0.21% receiving amphetamines. That is approximately one versus two cases per thousand over variable follow-up, not an annual risk estimate or a proven causal comparison. A separate 16-year follow-up found no evidence that prescribed stimulant exposure increased or decreased later frequent substance use or substance-use disorder after measured differences were taken into account. It supports neither the claim that appropriate treatment causes later addiction nor a promise that it prevents it. Psychosis comparison; long-term substance-use study

Clinical monitoring is part of treatment. NICE recommends assessing physical and mental health, current medicines and cardiovascular history before medication, and checking pulse and blood pressure after dose changes and at least every six months. An electrocardiogram is not routinely needed without a clinical reason. Decisions about starting, adjusting or stopping medication belong in a discussion with the treating professional. NICE monitoring recommendations

What better evidence needs to show

The broad inherited and developmental account is stronger than any single explanation of brain mechanisms. Removing small scanning studies or arguments based only on medication response would not remove the main genetic evidence. Stronger research needs to connect genetic findings, development and behaviour over time; test modifiable exposures using designs with different biases; and show that proposed biological subgroups genuinely improve prediction or care.

Treatment research needs longer comparisons that measure everyday life as well as symptoms. Only 33 of 312 studies in a major youth review reported outcomes at or beyond a year. Sparse follow-up cannot establish either lasting benefit or eventual loss of benefit. Reviews often reuse the same trials, so several agreeing reviews are not necessarily several independent confirmations. Missing results, selective publication, sponsorship, unblinded ratings and dropout can affect estimates. Sponsorship warrants scrutiny without automatically invalidating a result. The cognitive overview found larger effects in drug-funded reviews, although that association alone cannot explain why. Expert consensus statements also reflect a selected body of evidence; some contributors disclose industry relationships, and agreement cannot substitute for examining the underlying studies. Cognitive research overview; international evidence statement Youth treatment review; Cochrane assessment of trial limitations

Ordinary care also differs from a trial: missed appointments, inconsistent medication use, changing demands and accompanying illnesses can alter results. Longer practical trials, linked where appropriate to health records, should compare whole treatment strategies, including harms. Direct comparisons are needed to discover who benefits most from which option; population averages cannot reliably select the best treatment for one person.

Girls and women, older adults, people with complex accompanying conditions and populations outside mainly Western and European-ancestry research need better representation. The practical response is to combine careful assessment with explicit goals and review: what is easier, what remains difficult, and whether the benefits justify the burdens. Evidence can inform those decisions without reducing a person to a test score, a genetic finding or one account of how their brain works.