A 47-year-old executive arrived in my office last spring with a clean MRI and a problem nobody could name. He had been taking Adderall for three months, prescribed by his primary care doctor after he complained that his attention had degraded since the pandemic. The first week felt sharper. The second week he noticed jitters and worse sleep. By month three, his focus was no better than baseline, and his resting heart rate had climbed by 12 beats per minute. "I'm paying $400 a month to feel anxious," he told me. "Is this just what cognitive optimization looks like in your forties?"
It isn't. And new work out of Radboud University in the Netherlands explains why it isn't.
On April 14, researchers Ruben van den Bosch and Roshan Cools published a study in Imaging Neuroscience that takes apart, at the neurochemical level, why methylphenidate (the active ingredient in Ritalin) works powerfully for some brains and does almost nothing for others. Their team scanned 85 healthy adults performing a learning task under both methylphenidate and placebo, then layered those scans with PET imaging of each participant's individual dopamine synthesis capacity.
The technique itself is new and matters here. Conventional fMRI maps which brain regions light up during a task but says nothing about which neurotransmitter is doing the work. The Radboud group used a method called REACT (Receptor Enriched Analysis of Connectivity by Targets), which combines fMRI data with high-resolution receptor maps to separate dopamine-driven activity from noradrenaline-driven activity in the same scan. PET data from the same individuals confirmed that REACT was reading the chemistry correctly, not just guessing from anatomy.
What they found was clean. Methylphenidate shifted both dopamine and noradrenaline networks, as expected. But only the dopamine effect tracked with the person's baseline dopamine synthesis. People with higher dopamine capacity got the prefrontal boost. People with lower dopamine capacity got something else entirely: a noradrenergic stress response without the cognitive payoff.
The Wrong Pill in the Wrong Brain
This is what I suspect the executive in my office was experiencing. His dopamine system was probably running on the lower end of normal, which is common in midlife men carrying chronic sleep debt and the metabolic profile of someone who has been crushing it at work for fifteen years. His doctor reached for the standard stimulant. The drug raised his catecholamines, the noradrenergic side spiked his sympathetic tone, and the prefrontal cortex received exactly none of the dopamine modulation it was supposed to receive. He felt jittery because his stress system fired. He felt no sharper because the system that handles attention and working memory was never properly recruited.
The Radboud paper does not say this in clinical terms. But it strongly implies that the standard practice of prescribing methylphenidate or amphetamine without measuring or estimating baseline dopamine is, statistically, a coin flip. Roughly half the brains in front of a prescriber will respond well. The other half will get the side effects without the benefit.
What Personalized Stimulant Response Actually Looks Like
In conventional psychiatry and primary care, stimulants are titrated by trial and error. A patient reports better focus, the dose stays. A patient reports anxiety, the dose comes down or the drug gets switched. The feedback loop runs on subjective report, which means the clinician is debugging a closed black box.
What the Radboud study suggests, and what we already practice in the Intensive Brain Health Program, is that the box can be opened. Dopamine synthesis capacity is heritable to some degree, but it is also state-dependent. Sleep architecture, exercise dose, alcohol use, chronic inflammation, and the metabolic signatures of insulin resistance all modulate it. Several of these are catchable on the right blood panel and the right cognitive testing battery before a single pill is prescribed. That kind of upfront mapping is exactly what we mean by precision neuromedicine: replacing the trial-and-error loop with measurement, then deciding.
A real cognitive optimization workup, the kind that should happen before anyone starts a stimulant, looks at metabolic markers (homocysteine, hs-CRP, fasting insulin), thyroid status, B12 and folate, testosterone in men over 40, sleep architecture if there is any clinical suspicion of apnea, and a baseline neuropsychological profile that distinguishes sustained attention from working memory from processing speed. That panel gives a clinician a real prior on whether the dopamine system is likely to be the bottleneck or whether something upstream needs to be addressed first.
The Cost of Skipping the Workup
The executive I described above stopped his Adderall, started zone-2 cardio four mornings a week, fixed his sleep with a CPAP after his apnea study turned positive, and added a targeted nutritional protocol that included NAD+ precursors and omega-3s (the kind of regimen the team at Action Potential Supplements has been formulating for high-performance brains). His subjective focus came back without any stimulant. His resting heart rate dropped 14 beats per minute over three months. His follow-up cognitive testing showed processing speed gains that no pill had produced.
That is the part the Neuroeconomy thesis gets right. Cognition is not a one-knob problem. It is a system with several inputs, and the brain you bring to your work this morning is the sum of those inputs over the previous decade. Patching the system at the level of a single neurotransmitter, without knowing whether that neurotransmitter is the actual bottleneck, is what gives stimulants their reputation for inconsistent results.
What to Ask Before Anyone Hands You a Stimulant
A few questions worth raising with whoever is prescribing. What is my baseline dopamine status, even by proxy: sleep, exercise, inflammation, metabolic markers? Have we ruled out the upstream contributors before reaching for a downstream amplifier? What is the plan if I get side effects without benefit, and how will we know we are in that bucket?
The honest answer in most primary care offices is that none of these questions have been asked. The honest answer in a precision neurology workup is that all of them have been asked before the first prescription is written.
The Radboud study is small (85 people) and confined to healthy adults, not the clinical populations that most often receive these drugs. Even so, the principle generalizes. Brain chemistry is individual. Pills assume a uniform brain. Until precision neurology becomes the standard of care, the gap between those two things is where most of the wasted prescriptions live.