A specific question, answered specifically
ADHD Medication Doesn't Fix Everything
This isn't an anti-medication piece and it isn't a pro-medication piece. It's the honest version of a thing that gets glossed over in both directions: stimulant medication helps, sometimes a lot, and there are five specific dimensions of daily ADHD life that it leaves thinly addressed. Knowing which dimensions are which changes what you do next. Most ADHD adults who medicate well still build systems around the gap, and the ones who don't tend to spend years confused about why a treatment that worked is still leaving them stuck on the same things.
What the research actually says about efficacy
The cleanest summary of stimulant efficacy in ADHD adults remains the 2018 network meta-analysis by Samuele Cortese and colleagues, which pooled 133 randomised trials covering tens of thousands of participants and compared seven medications across both efficacy and tolerability ([PMID 27262877], Cortese et al. 2018). The headline numbers are not trivial. Standardised mean differences for amphetamines came out around 1.0 and for methylphenidate around 0.8 in adults — effect sizes that count as large in clinical psychiatry, and meaningfully larger than what most psychotropic medications produce on their target conditions.
The dimensions where medication shows the strongest effect are the core ADHD symptom domains: sustained attention, impulse control, hyperactivity, and to a slightly smaller degree, working memory. These are the dimensions the medication was developed to target. They are also the dimensions that show up most clearly in standard rating scales, which means the literature is calibrated to them.
What the same literature is weaker on is the dimensions of daily life that don't show up as cleanly on a symptom scale: how you feel at the end of a hard day, what happens to your patience under stress, how reliably you build a habit over six weeks, what an RSD spike does to a Wednesday afternoon. The medication helps the substrate. It does not always reach the surface.
Medication is a real intervention with real effect sizes. It's also one tool. The bill isn't fully paid by the prescription.
The five dimensions where systems still matter
These are the five places where medicated ADHD adults consistently report ongoing struggle, where the effect of optimal medication is partial rather than complete, and where structural compensations do most of the remaining work.
1. Emotional regulation under stress
Medicated ADHD adults often report that their reaction speed improves but the depth of the reaction itself doesn't change much. The argument that used to escalate in three minutes now takes seven. The bad email that used to derail the morning now derails ninety minutes instead of three hours. These are real improvements; they aren't a fix. Emotional dysregulation in ADHD has multiple substrates, only some of which are dopamine-mediated, and the limbic-prefrontal regulatory circuit is responsive to medication but not entirely controlled by it ([PMID 25130660], Shaw et al. 2014 review on emotion dysregulation in ADHD).
The system-side compensation is to plan for emotional volatility rather than expect to medicate around it. Pre-decided cooling windows. Permission to defer hard conversations. A trusted person who can hold a hard moment with you. The medication makes the slope shallower; the system catches the fall.
2. Habit formation across weeks and months
The classic stimulant effect is acute. The dose works for the day, the next day requires another dose. What this doesn't reliably do is build the cross-day, cross-week consistency that turns a deliberate behaviour into a habit. ADHD brains have measurable difficulty with the kind of repetition-without-novelty that habit formation depends on. Medication doesn't change that; it just gives you a better single day.
This is why many medicated ADHD adults report a frustrating pattern: the productive day on medication is real, but the day-to-day variance is large, and the cumulative output across a month is lower than the good days would predict. The compensation is structural — habit-stack to existing anchors, externalise the steps, accept that the streak frame will fail and the re-entry frame will not. Cross-link to why ADHD adults abandon planners covers the related angle: planning systems built for habit consistency fail ADHD brains for the same underlying reason.
3. Decision fatigue across the day
Stimulant peak effect is typically the morning into early afternoon for short-acting formulations and through to mid-afternoon or later for extended-release. By 6pm the dose curve is descending, and even a well-titrated medication day produces a brain at 6pm that is more fatigued than a brain at 10am. The decisions made at 6pm — what to eat, whether to start that thing, how to handle the small frustration — are made by a system that has been depleted by the day, including by the medication's eventual offset.
The system compensation is to pre-decide the late-day decisions. The dinner the day before. The "what do I do tonight" set the night before. The phone-on-charger across the room rule that doesn't require a fresh decision when the doom-feed pulls. The free decision fatigue index gives back a one-page read on where the day's small decisions are clustering, which is useful for spotting where pre-decision is most needed. Decision fatigue is one of the most underestimated costs of an otherwise productive medicated day, and it's where the evening unravels even when the morning didn't.
4. Rejection-sensitive dysphoria on a tough day
RSD — the disproportionate emotional response to perceived rejection or criticism — is a widely-reported phenomenon in ADHD adults, less well-formalised in the research literature than the core symptoms but consistent across clinical accounts. Medication softens the average RSD response somewhat by reducing emotional reactivity overall, but on a tough day — sleep-deprived, low blood sugar, hormonal, conflict-loaded — the RSD spike still arrives and still hurts. The cross-link to RSD explained covers the mechanic in full.
The compensation is mechanical. Don't make decisions during the spike. Don't reply to the message during the spike. Pre-script a delay protocol — "I need a few hours, I'll come back to this" — that you actually use. The spike resolves over a few hours; the decisions made during it don't reverse cleanly. Medication doesn't shorten the spike; the protocol does, by stopping the cascade before it expands.
5. The masking load
Social and professional masking — the ongoing work of looking neurotypical — costs cognitive resources continuously. Medication makes the underlying executive functions cheaper, which makes the mask cheaper to maintain. It does not reduce the need to mask in the first place. A medicated ADHD adult who masks heavily at work is still spending substantial cognitive budget on the performance, just at a slightly better exchange rate. The cross-link to masking burnout signs covers what the long-term cost of that looks like.
The compensation here isn't medical; it's social and structural. One trusted unmasked relationship. One quiet accommodation in the work environment. One un-performed reaction a day. The mask gets ten to twenty percent lighter and the burnout trajectory shifts.
The medication does the substrate. The systems do the surface. Neither one is enough alone for most ADHD adults.
Why the "medication is a complete fix" framing is dangerous
Two failure modes follow from the assumption that the prescription does it all.
The first is the "I shouldn't need anything else" trap. The medication starts working, the obvious symptoms improve, the systems get dropped because they look like they were the workaround for a problem that's now solved. Three to six months later, the dimensions where medication was always thin start producing problems that look like the medication is failing. It usually isn't. The systems left and the parts they were doing aren't getting done. The fix is not a higher dose; it's the systems coming back.
The second failure mode is the inverse: when the medication is partial — as it is for many people, on most days — the lived experience of "still struggling on medication" gets read as personal failure rather than as the predictable shape of the partial coverage. People conclude that they must be the broken one because the treatment that supposedly works isn't working enough. The treatment is doing what the literature says it does. The remaining gap is real and ordinary.
What this looks like for someone choosing to medicate
If you're medicated and reading this, the implication is straightforward: don't drop the systems. The systems do work that medication doesn't fully reach, and the cost of dropping them often shows up two or three months later when the dimensions they were quietly handling start producing problems again. Treat the systems and the medication as a stack rather than as alternatives.
If you're considering medication, this article doesn't have an opinion on whether you should. The dimensions covered above will be in your life either way; the medication is a question of whether the substrate is more or less expensive to operate. Either decision is reasonable; both decisions still leave you needing to handle the same five things.
If you're un-medicated by choice or because access is hard, the same five dimensions apply. The systems do more of the work and the day is more tightly engineered. Cross-link to executive function decline with age covers a related angle — the systems matter more, not less, as the substrate naturally weakens over decades.
What about non-stimulant medications?
The non-stimulant options — atomoxetine, guanfacine, clonidine, bupropion off-label — show smaller effect sizes than stimulants in the head-to-head literature ([PMID 27262877], Cortese et al. 2018), with different trade-offs. Atomoxetine's effect builds slowly over weeks and is generally smaller; guanfacine and clonidine have specific roles particularly around emotional regulation and sleep. The general pattern still holds: smaller substrate effect, same five dimensions remaining, systems still required.
One thing to do today
Pick one of the five dimensions — the one that costs you the most this week, medicated or not — and put one structural compensation in place for it. Pre-decided dinner if it's decision fatigue. A pre-scripted delay protocol if it's RSD. One trusted unmasked conversation if it's masking. One handoff anchor if it's habit formation. Don't try to address all five. The substrate-vs-surface model only works if the surface gets at least one piece of structural support, not if you try to engineer everything at once.
If you want a sharper read on where the load is highest in your week, the free tools page has four short instruments — friction score, EF load gauge, decision fatigue index, time-blindness check — that take three minutes each and give back one-page results. The cross-link to the kits covers the structured versions of the systems for the five dimensions above.
The honest summary
Stimulant medication is a real, well-evidenced intervention with effect sizes that justify its place in ADHD treatment for the people who choose it. It is also a partial intervention. Five dimensions of daily ADHD function — emotional regulation under stress, habit formation, decision fatigue, RSD on a tough day, and the masking load — remain thinly covered even on optimal medication. Systems do most of the remaining work, and dropping the systems because the medication is working is a common and predictable mistake.
The framing that helps is "medication and systems are a stack, not alternatives." Whatever you decide on the medication question, the systems question is largely independent. Build for the brain you've got, including the parts the prescription doesn't reach.
If this lands, the ADHD Executive Function Kit is built around the five dimensions medication leaves thinly addressed — emotional regulation, habit re-entry, decision pre-deciding, RSD protocols, mask load. It's $4.99 right now in the launch sale (was $9.99–$49). Sale ends May 31. See all 5 kits →