Deep dive #8 · 9 November 2026

ADHD

For health professionals General practice

ADHD is a neurodevelopmental condition, not a character flaw, and it doesn't stop at 18. It's also the diagnosis people argue about most: overdiagnosed in some children, missed in many adults, and treated with medicines that work well in trials but were hard to buy in Australia for much of 2023–2025. From 2026, GPs in most states can do more of the work. Here's the evidence to do it well.

At a glance

Criteria (DSM-5-TR): ≥6 inattentive and/or ≥6 hyperactive-impulsive symptoms, or ≥5 from age 17; ≥6 months; several present before 12; in ≥2 settings; impairing.
Prevalence: children 5.3% worldwide (Polanczyk 2007) and 7.4% of Australian 4–17-year-olds (Young Minds Matter). Adults: persistent 2.58%, symptomatic 6.76% (Song 2021).
Genes: heritability about 74% (Faraone and Larsson 2019); 27 risk loci (Demontis 2023).
Medicines (Cortese 2018, 133 RCTs): children, methylphenidate SMD −0.78; adults, amphetamines −0.79 and methylphenidate −0.49, on clinician ratings at about 12 weeks.
Outcomes: mortality rate ratio 2.07, mostly accidents (Dalsgaard 2015); life expectancy 6.8 years shorter in men and 8.6 in women with diagnosed ADHD (UK, O'Nions 2025).
Australia 2026: NSW GPs have continued scripts since 1 September 2025 and train to diagnose and initiate from March 2026. Queensland, WA, SA, Tasmania and the ACT have their own models.

Biggest take-home: diagnose carefully, treat properly, and check the youngest child in the class twice.

For patients: ADHD, ADHD in adults and ADHD in children.

Definition

DSM-5-TR

A. Symptoms: a persistent pattern of inattention and/or hyperactivity-impulsivity: ≥6 symptoms in a domain up to age 16, ≥5 from 17, for ≥6 months.
B. Onset: several symptoms before age 12.
C. Settings: present in ≥2, such as home, work or study.
D–E: clear impairment; not better explained by another condition.
Presentations: predominantly inattentive, predominantly hyperactive-impulsive, or combined.

ICD-11 (6A05)

Replaces ICD-10 hyperkinetic disorder; explicitly includes adults; onset up to age 12; allows diagnosis when symptoms appear only once demands exceed a person's compensation. No fixed severity scale.

History

1798, Alexander Crichton: described a "disease of attention" with distractibility and restlessness.
1845, Heinrich Hoffmann: "Fidgety Philip" in Struwwelpeter, a children's book, not a medical text.
1902, George Still: the Goulstonian lectures on 43 children with "an abnormal defect of moral control" (Lancet).
1937, Charles Bradley: "The behavior of children receiving Benzedrine": the first stimulant reports in children.
1944–1954, Leandro Panizzon: synthesised methylphenidate, marketed as Ritalin, named after his wife Marguerite ("Rita").
1971–1972, Virginia Douglas (McGill): argued sustained attention and impulse control, not hyperactivity, were central; this shaped DSM-III's "attention deficit disorder" (1980), renamed ADHD in 1987.
2007, Guilherme Polanczyk and Luis Rohde (Porto Alegre, Brazil): pooled 102 studies from every world region and showed prevalence varies with method, not continent.

Pathophysiology

A heritable difference in brain development and self-regulation. No biomarker diagnoses it.

Heritability: 74% from family, twin and adoption studies; about a third of it from many common variants (Faraone and Larsson 2019).
Genes: 27 genome-wide loci and 76 candidate genes in 38,691 cases and 186,843 controls (Demontis 2023).
Delayed maturation: half of cortical points reached peak thickness at 10.5 years in ADHD vs 7.5 years in controls, most delayed in prefrontal regions for attention and motor planning (Shaw 2007, 223 children per group).
Pharmacology: methylphenidate and amphetamine both inhibit dopamine and noradrenaline transporters (Faraone 2018).
Environment: prenatal paracetamol showed no association in sibling analysis (HR 0.98; Ahlqvist 2024).

ADHD from genes to diagnosis, and where stimulants act (simplified; Faraone and Larsson 2019, Shaw 2007, Faraone 2018). Download this diagram

Epidemiology

Global

Children: 5.29% pooled from 102 studies and 171,756 children; variation explained mainly by study methods (Polanczyk 2007).
Adults (Song 2021): persistent ADHD with childhood onset 2.58%, about 140 million adults; symptomatic adult ADHD 6.76%, about 366 million. Both fall with age.

Australia

Children: 7.4% of 4–17-year-olds in the 12 months before the Young Minds Matter survey.
Medicines: almost 470,000 Australians were prescribed ADHD medicines in 2022–23, four times as many as nine years earlier. From January 2024 to May 2025, dispensing rose 60% for dexamfetamine, 88% for methylphenidate and 140% for lisdexamfetamine.

Causes and risk factors

Family history: the strongest risk factor (heritability 74%).
Relative age: in WA, children born in June (youngest in class) were more likely to be medicated than July-born: RR 1.93 for boys and 2.11 for girls aged 6–10 (Whitely 2017). That's partly immaturity being labelled.
Not causes: prenatal paracetamol once family factors are accounted for (Ahlqvist 2024).

Preventable factors

Mislabelling immaturity: account for relative age in young children before diagnosing (Whitely 2017).
Preventable harms of untreated ADHD: crashes, injuries and justice involvement are lower in medicated periods (see Prognosis). These are within-person associations, not RCTs.
Cardiovascular harm of long-term medicines: monitor blood pressure and pulse (Zhang 2024).

Diagnosis essentials

Who: the Australian guideline says diagnosis needs "a thorough assessment by an appropriately trained clinician" and that primary care should offer diagnosis and treatment to people of all ages.
History: childhood onset (school reports and a family informant help), symptoms across settings, impairment, and what's been tried.
Adults: ask about work, study, driving, money, relationships and sleep. Use the 5-symptom threshold from 17.
Differentials and co-occurring: anxiety, depression, bipolar disorder, substance use, sleep disorders, autism, learning disorders, trauma, thyroid disease. Many coexist.
Before stimulants: cardiovascular history, family history of sudden cardiac death, blood pressure, pulse, weight (and height in children).

Australian prescribing in 2026 (check your state)

NSW: trained GPs continue psychostimulants for stable patients aged 6+ with an existing diagnosis (since 1 September 2025; over 800 GPs trained). Training to diagnose and initiate began March 2026; endorsed GPs can start, switch and continue dexamfetamine, lisdexamfetamine and methylphenidate from age 6. Next training round from November 2026.
Queensland: since 1 December 2025, non-specialist doctors can initiate for ages 4–17 without prior approval (up to maximum doses); specialist GPs can treat adults.
WA: accredited GPs diagnose and prescribe from age 10. SA: trained GPs from age 8. ACT: pilot, trained GPs from age 6. Tasmania: trained GPs, children and adults. Victoria: a GP training program from age 6; Schedule 8 permits otherwise.
NT: no change: individual authorisation for each patient.

Summarised from MDA National (September 2026) and NSW Health. Rules change: confirm with your state or territory health department before prescribing.

Current treatment

Medicines: effect sizes (Cortese 2018 network meta-analysis)

Children, clinician-rated (~12 weeks): amphetamines SMD −1.02, methylphenidate −0.78, atomoxetine −0.56; all drugs studied beat placebo. Teacher-rated: only methylphenidate (−0.82) and modafinil.
Adults, clinician-rated: amphetamines −0.79, methylphenidate −0.49, bupropion −0.46, atomoxetine −0.45; modafinil no better than placebo.
Tolerability: amphetamines had more dropouts for adverse effects than placebo (OR 2.30 children, 3.26 adults).
Authors' conclusion: methylphenidate first in children, amphetamines first in adults, for short-term treatment. Data at 26 and 52 weeks were insufficient.

Australian guideline (AADPA 2022)

113 clinical recommendations (12 evidence-based), NHMRC-approved.
First line, ages 5+ and adults: methylphenidate, dexamfetamine or lisdexamfetamine when impairment is significant.
Second line: atomoxetine, guanfacine or clonidine in children and adolescents; atomoxetine or guanfacine in adults.
Non-drug: parent and family training for children; CBT for adolescents and adults; lifestyle advice; school and workplace adjustments. No recommendation for cognitive training or neurofeedback.

Safety

Cardiovascular: each extra year of use was associated with 4% higher CVD risk; hypertension AOR 1.72–1.80 after 3+ years (Zhang 2024, Sweden, case–control).
Growth: consistent long-term stimulant use in childhood was associated with 2.55 cm lower adult height, without lower adult symptom severity (MTA follow-up, observational).

Shortages

Shortages have recurred since 2023: lisdexamfetamine 40 and 60 mg until about April 2024, then a nationwide methylphenidate shortage expected to last into late 2025. Causes: manufacturing problems, US production quotas and rising demand. The TGA allowed overseas brands, some later PBS-listed. Check the TGA shortages database before switching.

Emerging treatments and trials

Viloxazine extended-release: a non-stimulant approved by the US FDA for children (2021) and adults (April 2022).
Genetics: 76 candidate genes from the 2023 GWAS point to new biology; no clinical test yet.
GP-led models: the state reforms are a natural experiment in access and safety. We found no published outcome data yet.

Theoretical frontiers

Hypotheses, not recommendations.

Causes: theories

Hypothesis: ADHD is mostly delayed, not different, development. For: 3-year delay in cortical peaks (Shaw 2007). Against: many adults remain symptomatic (2.58% persistent). Test: imaging cohorts into adulthood comparing persisters and remitters.
Hypothesis: thousands of small genetic effects explain most risk. For: 27 loci; a third of heritability from common variants. Against: most heritability still unexplained. Test: larger, ancestrally diverse GWAS with rare-variant sequencing.
Hypothesis: prenatal paracetamol causes ADHD. For: HR 1.07 without sibling control. Against: HR 0.98 within siblings, no dose response (Ahlqvist 2024). Test: negative-control and Mendelian randomisation designs.
Hypothesis: rising diagnoses reflect milder cases, not more ADHD. For: 25 studies found added cases sit at the milder end (Kazda 2021). Against: adult under-recognition is also documented. Test: population surveys with identical methods over time.

Prevention: theoretical options

Hypothesis: correcting for relative age prevents overdiagnosis. For: June-born RR 1.93–2.11 (Whitely 2017). Test: a cluster trial of age-adjusted teacher ratings.
Hypothesis: treating adults prevents crashes and deaths. For: 38–42% fewer crash ED visits in medicated months (Chang 2017); mortality MRR 2.07 (Dalsgaard 2015). Against: within-person confounding by time-varying factors. Test: a pragmatic randomised treatment-timing trial with driving outcomes.
Hypothesis: medication reduces offending. For: 32% lower conviction rate in men and 41% in women while medicated (Lichtenstein 2012). Against: observational. Test: randomised treatment trials in justice settings.
Hypothesis: GP access reduces harm without overtreatment. For: state reforms widen access. Against: Kazda 2021 shows diminishing returns in milder cases. Test: linked data evaluation of the NSW reform.

Treatment: theoretical options

Hypothesis: long-term medication changes adult outcomes. For: within-person benefits. Against: no symptom benefit with long-term use in MTA. Test: randomised discontinuation trials after 2+ years.
Hypothesis: stimulant CVD risk is modifiable. For: risk concentrated in the first 3 years (AOR 1.08 per year). Test: trials of BP-guided dosing.
Hypothesis: milder ADHD does better without medicines. For: harms may outweigh benefits for milder symptoms (Kazda 2021, 5 studies). Test: an RCT of medication vs psychosocial care stratified by severity.
Hypothesis: new non-stimulants close the gap. For: viloxazine approved in the US. Against: non-stimulants had smaller effects than amphetamines (Cortese 2018). Test: head-to-head trials in adults.

Gaps and open questions

Long-term evidence: efficacy trials mostly run about 12 weeks.
Over- and under-diagnosis at once: milder cases in children, missed persistent cases in adults.
Women and girls: after excluding conduct, oppositional and substance use disorders, excess mortality was higher in girls and women than in boys and men (MRR 2.85 vs 1.27; Dalsgaard 2015). Why?
Supply: how to keep stable stock as demand grows.

Prognosis and quality of life

Mortality: 5.85 vs 2.21 deaths per 10,000 person-years (MRR 2.07); adult-diagnosed MRR 4.25; accidents most common (Dalsgaard 2015).
Life expectancy: 6.78 years shorter for men and 8.64 for women with diagnosed ADHD in UK primary care data, 2000–2019 (O'Nions 2025).
Crashes: 38% (men) and 42% (women) lower odds of crash ED visits in medicated months; up to 22.1% of crashes potentially avoidable (Chang 2017, 2.3 million US patients).
Justice: convictions 32% lower in men and 41% in women while medicated (Lichtenstein 2012, 25,656 patients).
Persistence: symptoms persisted into adulthood vs local norms in the MTA cohort (d=1.11).

Weird trivia

Ritalin is named after Marguerite "Rita" Panizzon, the chemist's wife.
Fidgety Philip came from a children's book written by a doctor, Heinrich Hoffmann.
Still's 1902 label was "an abnormal defect of moral control".
Bradley's discovery came while treating children's headaches with Benzedrine.
Birth month matters: June-born WA children were about twice as likely to be medicated as July-born classmates.
ADD lasted as the official name for only 7 years (1980–1987).

Guidelines

Key trials

Polanczyk 2007 (Am J Psychiatry): worldwide prevalence 5.29%.
Lichtenstein 2012 (NEJM): medication and criminality.
Dalsgaard 2015 (Lancet): mortality MRR 2.07.
Chang 2017 (JAMA Psychiatry): medication and car crashes.
MTA follow-up 2017 (J Child Psychol Psychiatry): persistence and height.
Cortese 2018 (Lancet Psychiatry): network meta-analysis, 133 RCTs.
Zhang 2024 (JAMA Psychiatry): long-term CVD risk.

Sources

Information only. Check product information, PBS conditions and your state's Schedule 8 rules before prescribing. Medicine shortages change; check the TGA database.

Updated 9 November 2026. General education for health professionals; not a substitute for local protocols.

maladywise.com/deep-dives/adhd/