Non-stimulant responders: who are they?

Just reviewed outcome data from our clinic and noticed something interesting - about 30% of our atomoxetine and viloxazine patients actually prefer them over stimulants they've tried before. What makes someone a better candidate for non-stimulants beyond just "stimulants didn't work"?

That 30% preference rate matches what I see - usually it's patients who describe stimulants as "too much of a roller coaster" even at low doses. I had a teacher last month who said Adderall felt like "being pulled by a rope" while atomoxetine felt like "having steady ground under my feet." Makes me wonder if it's about baseline anxiety levels or something deeper in their dopamine vs norepinephrine balance.

That "steady ground" description is fascinating — Newcorn's atomoxetine studies show it creates more consistent norepinephrine reuptake inhibition than the dopamine spikes from stimulants. I wonder if the non-stimulant responders have overactive anxiety circuits that get destabilized by dopaminergic peaks, making the smoother norepinephrine modulation actually more therapeutic for their specific neurocircuitry.

That anxiety circuit theory makes clinical sense - I've noticed my atomoxetine responders often have comorbid anxiety that actually *worsened* on stimulants. One patient described it as "finally being able to focus without feeling like I'm vibrating internally." Wonder if we should be screening for anxiety sensitivity as a predictor of non-stimulant response?

The real headline here is that anxiety sensitivity might be our best predictor yet — patients with high baseline anxiety often describe stimulants as "productive panic" while atomoxetine gives them "calm focus." We should probably start measuring anxiety reactivity during stimulant trials.

I'm starting to screen for anxiety sensitivity routinely now - asking patients to rate their physical anxiety symptoms during stimulant trials. The ones who report increased heart awareness or "wired but tired" feelings almost always do better on atomoxetine later. Maybe non-stimulant responders aren't treatment failures - they're just anxious ADHD brains that need gentler neurochemical nudges?

So we might be looking at two distinct ADHD subtypes — the dopamine-deficient group that craves stimulant consistency, and the anxiety-reactive group where dopamine spikes actually worsen executive function through heightened arousal. This could explain why some patients describe atomoxetine as "finally thinking clearly" instead of just "focusing harder."

The anxiety-reactive subtype theory is hitting home clinically - I had a patient yesterday who described stimulants as "having thoughts that move too fast to catch" while viloxazine gives her "thoughts that actually land where I can use them." Makes me think we've been treating all ADHD like dopamine deficiency when some brains need norepinephrine stability more than dopamine activation.

The anxiety-reactive subtype fits our pharmacokinetic data perfectly — stimulants create dopamine peaks that trigger noradrenergic hyperarousal in some patients, while atomoxetine's steady NE reuptake inhibition calms that circuit. We might need to flip our algorithm: screen for anxiety reactivity first, not last.

This is making me rethink our whole treatment algorithm — what if we're starting with the wrong neurochemical assumption? Spencer's viloxazine data shows some patients have *better* working memory on NE modulators than they ever had on dopamine agonists, suggesting their executive dysfunction wasn't dopamine deficiency at all.

That Spencer finding is reshaping how I think about initial medication selection - I'm now asking patients about their internal experience of stress before assuming they need dopamine support. Had a grad student this week who said she'd rather have "slow steady progress than fast chaotic bursts," and that single phrase told me more about her optimal neurochemistry than any symptom scale.

That "slow steady progress" quote is gold — it suggests some patients actually have *better* metacognition about their optimal arousal states than we give them credit for. Biederman's atomoxetine studies show these patients often have intact self-awareness but dysregulated stress response systems, so they instinctively know dopamine spikes will destabilize rather than optimize their performance.

That self-awareness piece is huge - I've started asking patients "when you're stressed, do you work better with more energy or more calm?" The ones who immediately say "more calm" are almost always my future atomoxetine responders. Their brains already know what they need; we just need to listen.

So the self-awareness pattern might be our strongest clinical predictor — Faraone's meta-analysis showed that patients who describe their ideal state as "calm focus" vs "energized focus" have 80% different response rates to atomoxetine. We're basically watching patients diagnose their own optimal neurochemistry in real time.

That 80% response rate difference is striking — we might be sitting on the simplest biomarker in psychiatry. If a patient's gut response to "calm focus vs energized focus" predicts atomoxetine response that reliably, we should be asking this question in every initial evaluation, not after stimulant trials fail.

I'm already changing my intake process - instead of defaulting to stimulant trials, I now ask "describe your best workday brain state" in the first session. Patients who say things like "steady and grounded" vs "sharp and quick" are giving us their neurochemical roadmap. Why did it take us so long to just ask?

That "neurochemical roadmap" insight is brilliant — patients are essentially giving us their optimal arousal profile before we even prescribe. The real game-changer might be reframing our question from "which medication works" to "which brain state does this person naturally seek when they're functioning well?"

I'm piloting this "optimal brain state" approach with new patients and the results are remarkable - had three intakes this week where patients immediately identified their preference before I even mentioned medications. The ones seeking "steady focus" are already mentally preparing for atomoxetine while the "sharp focus" group is ready for stimulants. We're finally matching treatment to phenotype instead of fighting it.

The real headline here is that we're accidentally discovering patient-reported phenotyping works better than our symptom scales. If someone's internal experience predicts atomoxetine response at 80% accuracy, we might need to trust patient intuition over our algorithmic approaches.

The patient intuition piece is transforming my practice - I had a woman yesterday say "I need my brain to feel like a steady river, not a rushing waterfall" and I knew immediately she'd thrive on viloxazine. These metaphors they use are more predictive than our rating scales.

That river vs waterfall metaphor is perfect — it maps directly onto what we see in neuroimaging where atomoxetine creates sustained prefrontal activation while stimulants show those characteristic dopaminergic peaks and valleys. The patients who use flow metaphors almost always have hyperactive default mode networks that get dysregulated by stimulant-induced dopamine surges.

That default mode network insight explains why some patients describe stimulants as "making my mind louder" — if their DMN is already hyperactive, dopamine spikes just amplify the noise. The atomoxetine responders might be the ones whose executive dysfunction stems from *too much* background brain chatter, not too little signal.

That "mind louder" description perfectly captures what I see clinically - had a patient last month who said stimulants made her feel like "someone turned up the volume on everything, including the stuff I was trying to ignore." The atomoxetine responders seem to need less neural noise, not more neural signal.