ADHD medication is one of the most extensively studied treatments in psychiatry. The research base spans decades and includes large-scale randomised controlled trials, network meta-analyses, and long-term observational studies. Despite this, the conversation between provider and patient at the point of prescribing is often rushed and incomplete. This article covers what the evidence actually shows — about how these drugs work, how effective they are, and what you should know before, during, and after starting treatment.
The two classes: stimulants and non-stimulants
Every ADHD medication on the market falls into one of two broad categories. Stimulants — based on either methylphenidate or amphetamine — are first-line treatment in every major clinical guideline, including NICE NG87 in the UK and the APA guidelines in the United States.1 2 Non-stimulants include atomoxetine (a selective norepinephrine reuptake inhibitor), guanfacine and clonidine (alpha-2 adrenergic agonists), and viloxazine (a norepinephrine reuptake inhibitor approved in the US for children and adolescents).
Within the stimulant class, the choice is between methylphenidate-based and amphetamine-based formulations. There are over 30 branded products across both, but they all derive from these two active compounds.3
How stimulants work in the brain
Both methylphenidate and amphetamine increase dopamine and norepinephrine activity in the prefrontal cortex — the brain region responsible for executive functions like planning, working memory, and impulse control.4
The difference is in the mechanism. Methylphenidate works primarily by blocking the reuptake of dopamine and norepinephrine at the synaptic cleft. It prevents the transporter proteins from pulling these neurotransmitters back into the presynaptic neuron, which means more dopamine and norepinephrine remain available in the synapse.4 5
Amphetamine does this too, but it also enters the nerve terminal and actively promotes the release of dopamine from synaptic vesicles. It inhibits vesicular monoamine transporter 2 (VMAT-2) and monoamine oxidase, resulting in a larger increase in extracellular dopamine than methylphenidate typically produces.4 5 In practical terms, amphetamine delivers a bigger dopamine signal. This is clinically relevant: a 2019 study in the New England Journal of Medicine involving over 200,000 individuals found that amphetamine-based stimulants were associated with a twofold higher rate of new-onset psychosis compared to methylphenidate (roughly 1 in 660 overall), consistent with the stronger dopaminergic action.6
At therapeutic doses — the doses prescribed for ADHD — neither compound produces the euphoria or reinforcement patterns associated with recreational stimulant use. The prefrontal cortex operates on an inverted-U dose-response curve: too little dopamine impairs function, as does too much. Therapeutic dosing aims for the narrow window at the top of that curve.5
How non-stimulants work
Atomoxetine selectively inhibits the norepinephrine transporter, increasing norepinephrine in the prefrontal cortex. Because the prefrontal cortex lacks significant dopamine transporter density, the norepinephrine transporter also clears dopamine in this region — meaning atomoxetine indirectly raises dopamine levels there as well.7 Its onset of action is slower than stimulants, typically requiring 4–6 weeks of consistent dosing before full effects are apparent.
Guanfacine (extended-release) works differently again. It is a postsynaptic alpha-2A adrenergic agonist that strengthens prefrontal cortex network connections. It does not directly increase catecholamine levels but instead improves the signal-to-noise ratio of prefrontal neurotransmission.7 It is most commonly prescribed as an adjunct to stimulants or when stimulants are contraindicated.
Viloxazine, approved by the FDA for paediatric ADHD, is a norepinephrine reuptake inhibitor with additional serotonergic activity. It represents the newest addition to the non-stimulant class.
What the evidence shows about efficacy
The largest and most comprehensive evidence synthesis to date is the 2025 network meta-analysis by Ostinelli and colleagues, published in The Lancet Psychiatry.8 This study analysed randomised controlled trials of pharmacological, psychological, and neurostimulatory interventions for ADHD in adults, covering more than 14,500 participants.
The headline findings: stimulants and atomoxetine were the only interventions with evidence of beneficial effects in reducing core ADHD symptoms in the short term, confirmed by both clinician-rated and self-reported outcomes. However, atomoxetine was less acceptable (meaning more people stopped taking it, primarily due to side effects) than placebo, and no ADHD medication showed evidence of improving quality of life at the group level.8
For children and adolescents, the evidence is stronger. A 2018 network meta-analysis of 133 double-blind RCTs by Cortese and colleagues found all included medications superior to placebo for clinician-rated ADHD symptoms. Effect sizes were highest for amphetamines (SMD = –1.0), followed by methylphenidate (SMD = –0.78) and atomoxetine (SMD = –0.56).9 Non-stimulants work, but with roughly half the effect size of stimulants.
A 2024 systematic review in JAACAP confirmed that both stimulant and non-stimulant medications improve quality of life in children, adolescents, and adults with ADHD, though with smaller effect sizes than for core symptom reduction. Notably, amphetamines, methylphenidate, and atomoxetine showed similar effects on quality of life despite their different effects on symptom severity.10
Response rates in clinical practice sit around 70% for any given stimulant. If the first stimulant does not work, switching to the other class (methylphenidate to amphetamine or vice versa) captures an additional subset of responders, bringing cumulative response rates to approximately 80–85%.2
Titration: what it is and why it matters
Titration is the process of starting at a low dose and gradually increasing until the optimal balance of symptom relief and side effects is found. This is not a one-appointment event.
NICE NG87 recommends that titration should involve regular review, typically every 1–2 weeks during the adjustment period. The target is the lowest effective dose — not a standardised amount.1 Your provider should be assessing symptom improvement, side effects, vital signs (heart rate, blood pressure), appetite, and sleep at each review.
For methylphenidate, starting doses are typically 5–10 mg for immediate-release formulations, titrated upward in increments. For amphetamine-based stimulants, starting doses are generally lower (5 mg). Extended-release formulations follow different schedules depending on the delivery mechanism.
Atomoxetine is typically started at 40 mg daily in adults, increasing to 80–100 mg after a minimum of three days, though full effects may take 4–6 weeks to appear.2
Side effects: what is common, what is rare
The most frequently reported side effects across stimulant medications are decreased appetite, insomnia, headache, dry mouth, and irritability.11 Most are dose-dependent and many diminish within the first two weeks as the body adjusts.
Cardiovascular effects deserve specific attention. Both stimulants and atomoxetine can increase heart rate (typically 3–6 bpm) and blood pressure (1–4 mmHg systolic). A 2025 network meta-analysis in The Lancet Psychiatry examining cardiovascular safety across ADHD medications in children, adolescents, and adults found that these effects are statistically significant but generally modest in magnitude for the overall population.12 NICE NG87 requires baseline cardiovascular assessment and monitoring every six months during treatment.1
Appetite suppression is the side effect most likely to persist. It can be managed through timing strategies (taking medication after breakfast, using the appetite window when medication wears off for the main meal) and dietary approaches.
Sleep disruption varies significantly between individuals. For some, stimulants paradoxically improve sleep by reducing the hyperarousal that prevents settling. For others, particularly with afternoon or evening dosing, insomnia is a significant problem. Extended-release formulations with shorter duration of action or adding a low-dose melatonin are common clinical strategies.
Emotional blunting — a reported sense of flatness or reduced emotional range — affects a subset of stimulant users. This is not a personality change but a dose-dependent side effect that typically resolves with dose adjustment or medication switching.
Rare but serious side effects include new-onset psychotic symptoms (approximately 1 in 660 for amphetamines, lower for methylphenidate), seizure threshold lowering in predisposed individuals, and serious cardiovascular events in people with pre-existing structural heart disease.6 12
What your provider should explain before prescribing
Before you start ADHD medication, you should have a clear conversation about:
The expected timeline — stimulants work on the first dose but finding the right formulation and dose takes weeks. Non-stimulants take 4–6 weeks before efficacy can be assessed.
Monitoring requirements — baseline weight, blood pressure, heart rate. For children: height. Review intervals during titration and at least six-monthly thereafter.
What "working" looks like — medication does not eliminate ADHD. It reduces the severity of core symptoms, particularly inattention and impulsivity, enough that behavioural strategies and environmental modifications become more effective.
The plan if it does not work — if the first medication does not produce adequate improvement at tolerable side-effect levels, the standard approach is to try the other stimulant class before moving to non-stimulants.
How to stop — stimulants do not require gradual tapering for most people, but abrupt discontinuation of non-stimulants (particularly atomoxetine and guanfacine) should be discussed with your prescriber.
Medication as part of a broader approach
The evidence is consistent: medication alone reduces symptoms but does not address the full impact of ADHD on functioning. The 2025 Cortese et al. review in World Psychiatry notes that while medications demonstrate short-term efficacy on core symptoms, evidence for long-term benefits and broader functional outcomes remains limited.2 Multimodal approaches — combining medication with psychoeducation, skills-based interventions (particularly CBT adapted for ADHD), and environmental modifications — produce the best outcomes.
Medication tracking is not optional. Documenting dose, timing, symptom changes, side effects, and sleep patterns provides the data your prescriber needs to optimise treatment. Without it, titration is guesswork.
