A new umbrella review in The BMJ tries to answer a deceptively simple question: what actually works for ADHD?

Not in theory, not in opinion—but across hundreds of randomised trials and decades of research.

The authors analysed 221 re-estimated meta-analyses covering 31 interventions across preschoolers, children, adolescents, and adults. It’s one of the most comprehensive attempts we’ve seen to bring order to an evidence base that is often contradictory, selective, or skewed by industry attention.

And it delivers a clear high-level message:

Short-term benefits of medication are well supported.
Long-term evidence is thin.
Non-pharmacological interventions show promise but sit in a methodological no man’s land.

But as always, the details matter—and this is where the story becomes interesting.

What the Study Actually Looked At

The review uses three primary outcomes:

1. Symptom Severity

Measured by clinicians, teachers, parents, or the individual.
This includes attention, hyperactivity, impulsivity, or the combined picture.

2. Acceptability

How many participants dropped out for any reason.
Dropout often reflects burden, side effects, or perceived lack of benefit.

3. Tolerability

Dropout due to side effects specifically.

Secondary outcomes included:

• quality of life
• emotional dysregulation
• executive functions
• academic or job performance
• anxiety and depression symptoms
• sleep or appetite changes

And the biggest limitation: most trials lasted about 8–12 weeks.

This short window shapes almost everything the review concludes.

The Headline Findings—Short-Term Wins

Medications

Drugs such as methylphenidate, atomoxetine, amphetamines, and alpha-2 agonists show medium to large short-term effects with moderate to high certainty.
In other words:

“We know these work in the short term, and we’re reasonably confident about it.”

Adults show similar patterns, though effect sizes are smaller.

Non-Drug Interventions

Some non-drug approaches produced large effect sizes, but with low or very low certainty.
That includes:

• CBT (especially in adults)
• behavioural parent interventions
• mindfulness
• physical training
• organisational skills training
• and yes, even acupuncture (more on that later)

Low certainty means:

“This might help, but the trials are too small, too varied, or too poorly blinded to be sure.”

The review stresses that no intervention—drug or non-drug—had high-certainty evidence for sustained benefits at 26 or 52 weeks.

And this is where the conversation needs to shift.

The Real Question: Why Don’t Non-Drug Interventions Look Stronger?

From a neurodiversity-informed perspective, the “weakness” of non-pharmacological interventions in the evidence base isn’t about inefficacy.
It’s about misalignment between mechanism and measurement.

Most behavioural and psychological interventions improve ADHD not through symptom suppression, but through:

• habituation
• skill acquisition
• environmental restructuring
• metacognitive insight
• practice-dependent neural change

These mechanisms unfold slowly. Gradually. Often non-linearly.
But the RCTs in this review only capture a 12-week snapshot—the steepest and least stable part of the learning curve.

In short:

Non-pharmacological interventions require long-term practice, but the evidence system only measures short-term outcomes.

This is not a flaw of the interventions.
It’s a flaw of the research architecture.

Let’s break that down.

CBT: Effective, but Not Designed for Quick Fixes

CBT shows moderate certainty evidence in adults and low certainty in younger groups.

But the mechanism of CBT is not acute symptom relief. It is:

• identifying patterns of avoidance
• reframing defeatist thinking
• building compensatory routines
• creating structure that eventually becomes habitual

CBT benefits tend to accumulate, not burst into existence at week eight.

Yet the trials measure:

• symptom reduction now,
  rather than:
• how impairment changes months after the skills have been practised.

It’s like measuring someone’s fitness by testing them after their first week in the gym.

Mindfulness: Notable Signals, Wrong Timescale

Mindfulness produces medium to large effects in adults—but with low certainty.

Why?

1. Programs vary widely
2. Blinding is impossible
3. Samples are small
4. Many adults with ADHD need to learn how to attend before they can benefit from attention training
5. The real impact of mindfulness often appears after months of continued practice

A 12-week trial can show early effects, but it cannot capture the transformation from:

• state-level calm → to
• trait-level attentional regulation.

Again, the measurement window is too short for the mechanism.

Behavioural Parent Training: Environment Needs Time to Adapt

For children, parent-led behavioural interventions look inconsistent in the data. Yet real-world experience tells a different story.

Parents need time to:

• change reinforcement patterns
• build consistent routines
• manage their own stress
• shift expectations and communication styles

No randomised trial can compress that learning curve into 12 weeks.

The umbrella review therefore captures the initial disruption, not the eventual stability.

Organisational Skills and Executive Function Training: Habit Requires Repetition

Executive function interventions rely on repetition.
Not inspiration.
Not insight.
But behavioural automation.

Habits form over months, but trials treat these interventions as though they produce rapid symptom change, the same way stimulants do.

They don’t—and they aren’t meant to.

Physical Training: Biology Changes Slowly, Not on Trial Timelines

Exercise has strong mechanistic plausibility:

• dopamine modulation
• reduced stress reactivity
• improved sleep
• enhanced cognitive flexibility

But neuroadaptation takes time.
No 12-week trial can assess whether exercise becomes a maintained lifestyle habit, which is where long-term benefit lives.

And Acupuncture… briefly

Large effect sizes appeared for acupuncture, but with very low certainty. This is a reminder that:

• tiny trials
• weak controls
• expectancy effects
• and statistical noise

can create “signals” that don’t survive scrutiny. It’s an artefact, not a meaningful trend.

There is no plausible mechanism for long-term habituation or trait change from acupuncture in ADHD.

That’s all it needs.

So Where Does This Leave Us?

The umbrella review answers one narrow question very well:

Which interventions reduce ADHD symptoms over the short term in controlled trials?

But it cannot answer the question many ADHD adults—and many clinicians—care about:

Which interventions create sustainable, long-term improvement in functioning, identity, and daily life?

Non-pharmacological interventions look weaker on paper because they are measured with the wrong lens.
These approaches are not designed to suppress symptoms, but to build systems, habits, and emotional scaffolding that improve life beyond the acute spotlight of an RCT.

When we evaluate them only through short-term symptom change:

• we undervalue them,
• we misunderstand their purpose,
• and we risk reinforcing a treatment hierarchy tilted toward what is easiest to measure, not what is most transformative.

A neurodiversity-informed approach demands a better question:

What produces meaningful, sustainable change in the lived experience of ADHD—
not only at week twelve, but at year one and year three?

Until trials adapt, non-pharmacological interventions will continue to be judged by criteria they were never designed to meet.

And then we have the Lived Experience!

There is something the BMJ review cannot touch—something too human, too complicated, and too real to fit inside a 12-week trial window.

It’s the lived experience of trying to change your life with a brain that resists the very processes required for that change.

Non-pharmacological interventions often look weak in the evidence because they aren’t designed to create rapid, measurable symptom drops. They work quietly, slowly. They build scaffolding. They cultivate skills. They reshape patterns in ways that unfold over months, sometimes years. And yet these are the interventions that demand consistency, repetition, and follow-through—the exact places ADHD pulls hardest.

• CBT asks you to show up for yourself every day.
• Mindfulness asks you to stay with your own mind when it wants to bolt.
• Executive-function training asks you to practise routines long before they feel natural.
• Behavioural interventions ask families to rewrite entire systems of interaction.

No trial can measure how heavy those demands feel on the inside.

No meta-analysis can quantify what it takes for someone with ADHD to try again on the days they have nothing left.

No effect size can capture the exhaustion of being told “you must stick to the program” by someone who has never lived inside your brain.

And that’s the quiet truth behind the data:

These interventions often work.
But they don’t work alone.

They work when someone feels seen, supported, and understood long enough to keep going.

For many people with ADHD, the hardest part isn’t the intervention. It’s staying with it long enough for the benefits to matter.

This is where research and reality part ways. Trials assume a straight line from session to symptom change. Life rarely cooperates. Habits slip. Routines crack. Interest wavers. Fatigue wins. And yet people with ADHD come back, again and again, because the hope of a better future-self is stronger than the pull of avoidance.

If we want to understand long-term ADHD outcomes, we can’t only study what reduces symptoms in three months.

We have to study what helps people stay connected to the process when their neurology, their history, and the world around them make that almost impossibly hard.

The real story isn’t about whether CBT or mindfulness or skills training “works.”

The real story is this:

• There is no long-term intervention without long-term support.
• There is no habit without scaffolding.
• And there is no change without compassion for the immense effort ADHD asks of the people living with it.

That’s the part the data can’t hold—but every person with ADHD knows it in their bones.