Brain, Consciousness & Mind

ADHD — Latest Data, Treatments, and Mechanisms

Compiled November 5, 2026, from current CDC data and the 2025–2026 literature.

1. Prevalence — the current numbers (CDC)

• Children 3–17: about 7 million, 11.7%, hold a current ADHD diagnosis (2024 National Survey of Children's Health; CDC data page last reviewed July 2026). Boys 13%, girls 7%. Roughly one million more children had ever been diagnosed in 2022 than in 2016.
• Adults: 15.5 million, 6.0%, hold a current diagnosis (2023 survey, published MMWR October 2024). 55.9% were diagnosed at age 18 or older, and 36.5% were receiving no treatment.
• Of children with current ADHD, 77.9% have at least one co-occurring disorder, 58% are rated moderate or severe, and about 30% receive no ADHD-specific treatment.
• Worldwide, pooled estimates put adult ADHD at ~6.8% symptomatic and ~2.6% persistent from childhood (Song et al., Journal of Global Health, 2021).

2. New research — fall 2026

ADHD and later dementia. JAMA Psychiatry, October 7, 2026 (University of Pittsburgh; NIH All of Us data, 187,000 adults aged 50+). 6.7% of participants with ADHD developed dementia or mild cognitive impairment vs 3.4% without — more than four times the adjusted risk. Genetic risk told the same story: the highest ADHD genetic-risk group had a 35% greater chance of dementia than the lowest, diagnosed or not. The association held when at least eight years were required between ADHD diagnosis and cognitive onset. It is an association, not proven causation; whether stimulant treatment or cardiovascular risk management lowers the risk is unknown.

A candidate EEG fingerprint. BMC Psychiatry, 2026: in 121 children aged 6–15, ADHD showed reduced occurrence of EEG microstate C and reduced C→D transition probability (effect sizes ~0.56–0.58). A microstate-only classifier reached AUC 0.744 — a genuine signal, well short of a diagnostic test.

Mechanism, in mice. Neuropsychopharmacology, September 2026: deleting the NSF protein in dopamine D2-receptor-expressing cells impaired striatal development, lowered striatal dopamine, and produced ADHD-like behavior. Methylphenidate alone did not reduce hyperactivity; methylphenidate plus a D2 agonist (quinpirole) did. Basic research only.

Digital therapeutic. Psychiatry Investigation, 2026 (South Korea, n=54): the game-based device ADAM-101 added to medication beat medication alone on ADHD rating scales after four weeks. Preliminary; needs a larger trial.

3. Latest treatment

Simtriyo (centanafadine) — FDA approved July 24, 2026. First-in-class NDSRI: a norepinephrine, dopamine, and serotonin reuptake inhibitor, once-daily extended-release (140/210/280 mg), approved for adults and children 6+ weighing at least 20 kg, on four Phase 3 trials.

• Not yet in pharmacies: the FDA label classifies it as a CNS stimulant, so DEA scheduling must be completed first; launch expected late 2026.
• Efficacy is modest: effect sizes of 0.24–0.40 across the program — the bottom of the non-stimulant range, roughly half of amphetamine's ~0.79. It was tested only against placebo, never head-to-head; a manufacturer-funded indirect comparison found it significantly worse than lisdexamfetamine.
• The "triple" mechanism is mostly norepinephrine: binding data show >6x less potency at the dopamine transporter and >14x less at the serotonin transporter than at NET — pharmacologically closer to atomoxetine than to a stimulant.
• Post hoc analyses suggest gains in executive function and emotional dysregulation in adults — unpublished and exploratory, not an established finding.

4. The neurological mechanism

The braking and steering systems of the brain run on too little dopamine and norepinephrine signal in the circuits that control them, so the signal drowns in noise.

• Circuits. The prefrontal cortex (working memory, planning, inhibition), the striatum (gating of actions and thoughts, reward value), and the cerebellum (timing) are modestly smaller, under-connected, and slow to mature in ADHD — cortical thickness peaks roughly 2–3 years late, prefrontal cortex last.
• Dopamine runs three relevant roads: nigrostriatal (movement, habit), mesolimbic (reward), and mesocortical (prefrontal focus and working memory). Low dopamine tone makes ordinary rewards register weakly — the basis of delay aversion: a reward ten minutes away is neurologically almost worthless, so the immediate one wins.
• Norepinephrine (from the locus coeruleus) sets arousal and the signal-to-noise ratio in the prefrontal cortex; too little and the cortex cannot hold a representation steady against distraction.
• Receptors and the inverted U. Prefrontal networks need D1 and α2A stimulation in a narrow band — too little signal and the network is weak and noisy; too much and it collapses the other way. ADHD sits on the low side.
• Transporters. DAT and NET clear the transmitters from the synapse; genetic and imaging findings cluster at the transporters and the D2/D4/D5 receptors — signaling that ends too fast or lands too weakly.
• Network level. The default mode network (mind-wandering) normally switches off when a task begins; in ADHD it keeps intruding, so attention is interrupted rather than gradually lost.
• Treatment follows from the mechanism. Stimulants block DAT and NET and push the prefrontal cortex up into its optimal band; atomoxetine blocks NET only; centanafadine blocks all three transporters but grips NET hardest — hence its non-stimulant-range effect.

5. Microdosing — psychedelics

Exactly one rigorous trial exists. Mueller et al., JAMA Psychiatry, 2025 (Basel and Maastricht): phase 2A double-blind RCT, 53 adults with moderate-to-severe ADHD, LSD 20 µg twice weekly for six weeks vs placebo. Both groups improved; the LSD group improved less numerically (~7.1 vs ~8.9 AISRS points), with no significant difference. LSD was safe and well tolerated (adverse events roughly double placebo, none serious); 63% of participants guessed their allocation correctly. A companion analysis found no benefit on attention, inhibition, or working memory. For psilocybin there is no controlled ADHD trial at all — only naturalistic reports. A 2026 systematic review (Chmiel et al., International Journal of Molecular Sciences) concluded the evidence does not allow effectiveness to be confirmed, and a 2026 preregistered meta-analysis of 14 controlled microdosing studies (n=1,614) found no overall cognitive benefit and a selective reduction in cognitive control. Safety footnote: LSD and psilocybin bind 5-HT2B as potently as 5-HT2A, so valvular heart disease is a stated, unquantified concern with chronic use. No pediatric data exists.

• Boxed warning for suicidal thoughts and behaviors in children. Practical points: capsules can be sprinkled on food; low CYP interaction risk; roughly doubles caffeine exposure.
• Reviewers position it second-line: predominantly inattentive presentation, insomnia blocking stimulants, or intolerance of existing non-stimulants.

The standard of care has not moved. Stimulants remain first-line (methylphenidate first for children 5+; lisdexamfetamine or methylphenidate for adults); behavioral treatment first under age 5; AAP pairs medication with behavioral support at ages 6–11. A 2025 analysis of 113 adult trials found stimulants and atomoxetine were the only treatments improving both self- and clinician-rated symptoms — and medications did not improve quality of life.

The 2026 practice landscape. FDA stimulant shortages remain current (amphetamine salts listed since 2022; lisdexamfetamine and ER methylphenidate since 2023). Telehealth prescribing of stimulants without an in-person visit is extended only through December 31, 2026.

6. Microdosing — standard antidepressants

There are no microdosing trials of standard antidepressants, for a pharmacological reason: a "micro" dose of an SSRI is not micro. PET imaging (Meyer et al., American Journal of Psychiatry, 2004) shows serotonin transporter occupancy is hyperbolic — minimum therapeutic doses give ~80% occupancy, and citalopram 3.4 mg (a sixth of a usual starting dose) already occupies ~50% of the transporter; 5.4 mg reaches ~60%; sertraline 9.1 mg ~50%. There is no sub-perceptual basement; a quarter-dose SSRI performs most of the drug's main action with a fraction of the side-effect load. Where subtherapeutic doses are genuinely studied is deprescribing: hyperbolic tapering uses progressively lower doses to step occupancy down gradually and avoid withdrawal (2025 fluoxetine-substitution protocol, PMC).

Antidepressants trialed in ADHD at ordinary doses, by mechanism:

• Bupropion (NDRI): real RCTs — 53% responders vs 31% on placebo in the XL trial, effect size ~0.6; Cochrane confirms a moderate effect. It works because it raises norepinephrine and dopamine.
• Tricyclics (desipramine, nortriptyline): older positive trials, mild benefit, largely abandoned for side-effect and cardiac burden.
• SSRIs: no effect on core ADHD symptoms; they treat comorbid depression and anxiety.
• Atomoxetine and viloxazine began as antidepressant compounds (NRIs) and became approved ADHD drugs.

Practice note (Roen, November 5, 2026): low-dose ("microdose") standard antidepressants are something he suggests regularly in practice and has used himself for years. The trial literature has never been designed to test dosing at this level; the occupancy data suggest why it may work — small doses are not small effects.

7. Sertraline for women, escitalopram for men?

Half of this pairing is evidence-backed. Kornstein et al., American Journal of Psychiatry, 2000 (635 patients, chronic depression, sertraline vs imipramine): women responded to sertraline at 62% vs men's 45%*; men responded to imipramine at 57% vs women's 46%**. The women's advantage was concentrated in premenopausal women — postmenopausal women responded equally to both — implicating estrogen in serotonergic response. Dropout followed the same crossover.

Escitalopram for men has no equivalent trial. The male-favoring drug in the data was imipramine, a tricyclic. Later analyses softened the picture further: pooled data suggest women may respond somewhat more to SSRIs generally, inconsistently (Khan et al., 2005; GENPOD secondary analysis found no sex difference for citalopram), and a 2026 reanalysis of escitalopram found no sex differences on depression outcomes — only less sexual dysfunction reported by women on escitalopram.

One sex difference is pharmacokinetic and firm: women run higher concentration-to-dose ratios across the SSRIs, escitalopram and sertraline included — the same milligram yields more exposure. Sex-based dosing therefore has a firmer footing than sex-based drug choice. Bottom line: sertraline-for-women has one strong trial behind it (modest, mostly premenopausal); the evidence-based male counterpart in the literature is a noradrenergic/tricyclic agent, not escitalopram.