Issue #35 — What ADHD Medication Actually Does (It's Not What You Were Told)
🎯 TL;DR
The largest brain-imaging study of ADHD medications ever conducted found that stimulants don't sharpen the attention circuit — they pre-reward the brain, making ordinarily unengaging tasks feel more interesting. This isn't a minor update. It changes how you should use medication strategically.
This week: What stimulants actually do, why the conditions around the task matter more than you thought, and how to use this to get more out of your medication (or make the same case without it).
Read time: 5 minutes
💊 What the Research Actually Found
The study was published in Cell in December 2025 and has been widely discussed in neuroscience and psychiatry circles since. It's the largest brain-imaging study of stimulant medications ever conducted: resting-state fMRI from 5,795 children enrolled in the Adolescent Brain Cognitive Development (ABCD) Study, run by Washington University School of Medicine.
The researchers expected to find that stimulants improved function in the brain networks traditionally linked to attention — the fronto-parietal control network, the default mode network, the executive control circuitry. This was the consensus model.
That's not what they found.
Stimulants produced almost no measurable change in those networks. The largest effects appeared in an entirely different set of regions: the ones governing arousal and reward.
The drugs essentially pre-reward the brain. They increase the perceived salience of whatever task you're working on, making something that would ordinarily feel unrewarding feel more engaging.
A companion finding: children without ADHD who took stimulants showed no improvement in working memory, test scores, or grades. Children who were sleep-deprived showed improvement. Children with ADHD showed the strongest improvements. The mechanism isn't "attention enhancement" — it's something closer to "arousal normalization."
🧠 Why the Old Model Was Wrong
The dominant theory for decades has been something like: ADHD = low dopamine in the prefrontal cortex → stimulants raise dopamine → attention improves.
This model isn't entirely wrong, but a January 2026 Nature feature synthesizing recent dopamine research adds critical nuance: dopamine's role in ADHD isn't about overall levels — it's about the caudate nucleus, the part of the brain that decides whether a task is worth engaging with.
Think of it as the brain's engagement gatekeeper. When a task carries sufficient perceived reward (interest, urgency, novelty, stakes), the caudate nucleus opens the gate and you get sustained focus. When it doesn't, the gate stays closed and you get avoidance, distraction, and the peculiar ADHD experience of being completely unable to start a task you know you need to do.
Stimulants don't improve the quality of the gate mechanism. They raise the baseline reward signal — making more tasks clear the threshold that triggers engagement.
This explains something important: why ADHD founders can sustain extraordinary focus on work they find genuinely interesting, and be functionally paralyzed by work they find dull — even when both tasks matter.
The engagement gate isn't broken. It's calibrated to a different threshold. Stimulants lower the threshold temporarily. But so does genuine interest.
⚡ The ADHD Angle
Here's the strategic implication that most people miss from this research:
If stimulants work primarily by raising reward salience rather than by sharpening attention circuitry, then the conditions you create around the task matter enormously.
Pairing medication with work that already carries some intrinsic interest, urgency, or novelty is more effective than using it to grind through purely aversive tasks. You're amplifying an existing signal, not creating one from nothing.
Conversely, taking medication and then doing the most miserable administrative work you've been avoiding is a less efficient use of the mechanism than many ADHD founders practice.
| Task condition | Medication effect | What to do |
|---|---|---|
| Work with genuine interest | Amplifies existing engagement | Medication most effective here |
| Work with urgency/stakes | Raises salience through deadline | Medication works well |
| Purely aversive work, no interest | Lower baseline to amplify | Medication works harder, less efficiently |
| Post-medication, low sleep | Arousal floor already low | Sleep deprivation undermines the arousal mechanism |
"You're not using medication wrong if you use it for boring tasks. But if you only use it for boring tasks, you're leaving performance on the table."
🔧 How to Restructure Your Medication Strategy
1. Front-load interesting and high-stakes work during your medication window.
If your stimulant window is 4-6 hours of peak effectiveness, use the first half for cognitively demanding work that already has some intrinsic engagement for you. Not the email backlog. Not the expense reports.
2. Create interest conditions for unavoidable aversive tasks.
Since the mechanism is reward salience, you can partially engineer it for tasks that lack intrinsic interest:
- Add a time constraint (you have 25 minutes — go)
- Add an audience (work on it while on a call with someone doing parallel work)
- Add novelty (do it somewhere you've never worked before, or with a new tool)
- Add stakes (if I don't finish this today, [real consequence])
These are not hacks. They're working with the same engagement-gate mechanism that stimulants address chemically.
3. Protect sleep like it's medication — because neurologically, it is.
The Cell study found that sleep-deprived children without ADHD showed stimulant-like benefits. The arousal mechanism that stimulants activate overlaps with the arousal mechanism that adequate sleep provides. Running on poor sleep while relying on stimulants is running two systems in parallel that partially cancel each other out.
4. If you don't use medication, these principles still apply.
The caudate nucleus engagement-gate model works the same way regardless of medication. The founders who perform best without medication have usually — whether consciously or not — built workflows that engineer task interest: urgency through accountability, novelty through environmental variation, stakes through public commitments. They're doing the same thing with environmental design that stimulants do neurochemically.
🎯 This Week's Challenge
Map your next week's tasks against the engagement-gate model.
Two categories: work that has genuine interest, urgency, or novelty built in — and work that has none of those things.
Then ask: am I scheduling my highest-cognitive-demand hours for the first category, or the second?
If you're on medication: is your medication window aligned with work that has some intrinsic interest, or are you burning it on the most aversive tasks?
One scheduling adjustment from the right framework is worth more than weeks of willpower.
See you Tuesday, L-P
P.S. — The Cell study is referenced widely across neuroscience journalism from December 2025; StudyFinds and PsyPost both have accessible summaries. The Nature dopamine feature (January 2026) is worth reading if you want the deeper model on the caudate nucleus. And if you've already figured out a medication + environment pairing that works specifically for you — reply and tell me. Real examples are always more useful than abstract frameworks.
Divergent — Strategy for brains that don't do boring.