Now I remember the last conversation we had about this. She’d be maybe three sentences into whatever reaction she was having before anything registered with her at all. Door already slammed shut. Reply already sent off. She’d tried the counting thing, knew how it worked in theory. Didn’t work though. Because by the time she remembered she should probably count, the moment the counting was meant to catch was already behind her.

“Tell me what was different this time,” I said.

And she actually thought about it, which you know she doesn’t always do. Usually with Priya it’s just, tell me what to do.

“I felt it in my chest first. Before I knew what I was about to say. There was this half-second where I noticed it as a feeling in my chest, not as a sentence already forming.”

Right. That half-second. That’s the whole thing really, isn’t it?

Richard Davidson’s been doing this work for decades now, not so much on the feeling of emotion but on the biology of it, the timing. What does it actually look like on a brain scan when someone gets knocked sideways and comes back quite quickly versus when they get knocked, and they’re down for hours? He calls all of this emotional style. Different people have different emotional styles the same way we have different personality types or whatever. How long does a low mood last, how accurately can you read what’s going on in the room, how fast does recovery happen.

And his argument is that all of this stuff sits on circuits that keep changing with use. He’s done imaging work with people who’ve been doing sustained mental practice over years and years and years. The circuits change. You can measure it. It’s not a metaphor.

So I’m sitting in clinic week after week seeing people who’ve been told for years, decades sometimes, that their emotional reactions are just them. That’s their personality; it’s fixed, it’s done.

Davidson suggests that it isn’t as simple as that. The way that your emotional system works is not set in stone; you can adapt and train it. His research has shown that, and that is so important for people who believe that they cannot react in any other way. Others such as Barkley have said for a long time that these emotional storms are part of ADHD, not separate from it.

We already know if we are going to intervene, we have to get in there early before things really cascade. Marc Brackett teaches us to use the gap between feeling something and acting on it by naming what we are feeling. That can work incredibly well for some — when the naming is available, which it isn’t always. Previously I had mentioned how Linehan was interested in what survives complete overwhelm, and what doesn’t. Dunkley wrote about something quieter beneath the anger. Whilst they come at it from different angles, they all focus on the same endpoint.

So Priya came in saying something had shifted, and honestly what she described made sense. The window hadn’t got wider, it had just moved. Earlier. Like before Gross would say the cascade fully kicks in. She was catching things before they became full reactions. And she wasn’t white-knuckling anything. It had just changed through practice.

The body registers before the story catches up. She had just had more practice.

Subscribe now

It turns out there’s a name for starting with the body instead of the word. The grounding part of CARE begins in exactly the same place:

https://care.thecontextshift.com

I do want to be careful here though. ADHD isn’t something you can just practice your way out of. Most weeks for most of my patients, the noticing still doesn’t arrive in time, which is fine; six weeks isn’t actually very long when you’re up against years of the pattern already being in place.

Another patient told me at the start of this series that his focus had fixed itself on meds, but he still blew up at his wife the same way, still got snappy with the kids the same way. He said that was just him. I said wiring changes with use, just slowly.

Half a second. That’s what we’re talking about here. And it moves.

So if something comes up this week, in that window before it becomes words, what did you feel in your body? Not what you thought about it after. The actual thing. Tell me that in the comments.

Elena’s psychologist asked the question gently, the way she always did. “What are you feeling right now?” The room smelled faintly of the coffee going cold on the side table between them. Elena’s jaw was tight enough that she could hear her own pulse in her ears, a low tick she’d normally have tuned out completely. Her hands were laced together hard enough to leave white marks across the knuckles, and underneath the surface calm her stomach had drawn in on itself, hard and small. And when she reached for the word, there was nothing to reach. Not “I don’t know” as a deflection. An actual blank, the kind where you can feel the heat rising up the back of your neck and still come up with nothing to call it.

“I know something’s going on,” she told me afterwards, in my clinic, her voice flatter than the moment she was describing. “I just can’t get to it fast enough to say what it is. By the time I’ve found a word, the moment’s already moved on to the next thing.”

There’s a model built for exactly this gap. Psychologist Marc Brackett’s RULER framework runs through five steps: Recognising an emotion, Understanding its cause, Labelling it precisely, Expressing it appropriately, and Regulating it. His Mood Meter maps feelings by energy and pleasantness, and the Meta-Moment gives a pause between trigger and response. Central to how Brackett teaches the labelling step is what he calls, citing psychologist Lisa Feldman Barrett, emotional granularity: precise naming calms the nervous system where vague naming keeps it circling.

It’s a good model, and it assumes something most emotional regulation advice quietly assumes: that the internal signal has already arrived clearly enough to attach a word to it. That’s where it comes apart for a lot of ADHD adults.

The reason “just name the feeling” fails here isn’t a vocabulary problem, it’s a signal problem. RULER’s labelling step presupposes the sensation has already been detected with enough resolution to sort. Interoception, the felt sense of what’s happening inside the body, tends to run quieter and less reliably in ADHD, so the flood often arrives as noise rather than as a nameable state, there but indistinct, nothing sharp enough to fix a word to. There’s nothing precise to attach a word to, so the label defaults to a catch-all: fine, stressed, off. Not laziness with language. A sensing gap arriving before the naming gap ever gets a chance.

What survives isn’t a bigger feelings vocabulary. It’s building sensory vocabulary first, separately, before any emotion word gets attached to it. Not “what am I feeling” but “where is this sitting, and what is it actually doing.” Tight or loose. Hot or cold. Rising or flat. Sharp at one point or spread wide across the chest. That’s a different, more answerable question, and it’s one the body can answer even when the mind can’t yet produce a feeling word.

This is the exact bridge the CARE grounding tool’s interoception chips are built to walk someone through, sensation first, label second, so the naming step finally has something underneath it to work with rather than asking for the label cold.

https://care.thecontextshift.com

Subscribe now

The experiment, run outside any flood, in the calm window where it’s actually possible: three times this week, stop and name one physical sensation only, not a feeling. Where it sits. What it’s doing. Whether it has a temperature, a texture, a weight. Nothing more than that. The label comes later and comes easier once there’s something specific underneath it to attach it to.

If you’ve had that moment, reaching for a word mid-flood and coming up with nothing, it’s worth noticing where it tends to happen. In a meeting. On the phone with a parent. Mid-argument, on the stairs, the same as Elena. The setting usually says more about what’s underneath than the blank itself does.

Last piece in this series: what changes when regulation stops being something you fail at live, and starts being something you’re actually building.

By the time Naomi heard herself shouting, she was already several sentences past the point where she could have stopped.

It had started with a comment. Mild, offhand, her partner half inside the fridge: the dishwasher hadn’t been unloaded again. And something went up through her before she’d decided anything, a heat that started somewhere behind the sternum and was in her face before she’d registered it moving. The kitchen narrowed to the back of his head. The words were already arriving, and they were sharper than anything she’d have chosen if choosing had been on offer. Something about being treated like staff in her own house. She was out of the room before he’d finished the sentence, still hot, heart going, certain in that moment that she was right.

Twenty minutes later she was sitting on the stairs and the certainty had drained out of her completely. Not angry any more. Just flat, and faintly sick, the way you feel after the adrenaline leaves. Her hands had that slight unsteadiness. And underneath the flatness was something she didn’t want to look at directly, which is usually the sign that it’s the thing worth looking at.

“I don’t even care about the dishwasher,” she told me, three sessions after it happened. She’d waited that long to mention it, which is itself worth noticing, because by then she’d built a story around it that made her sound worse than she’d been. “I don’t know where it came from. It was like watching someone else do it and not being able to reach the controls.”

I asked her what she’d felt in the few seconds before the shouting started. Not the anger. Before the anger. She had to sit with it properly, and I watched her find something she clearly hadn’t enjoyed finding.

“Caught out,” she said. “Like I’d been caught failing at something small and obvious. And ashamed, if I’m honest, because I do let things pile up, and I hate that he sees it. I hate that he’s right.”

There it was. The anger hadn’t come from nowhere. It had come from somewhere that arrived and left so fast she’d never had a clean look at it.

There’s a well-established idea in emotion work, set out most clearly by the psychologist Christine Dunkley, that anger is very often a secondary emotion. Something else comes first. Usually something more exposing, sadness, shame, fear, humiliation, and anger arrives right behind it, because anger is the one that feels survivable in front of another person. Shame makes you want to disappear. Anger makes you want to stand your ground. Given the choice, the nervous system reaches for the one that lets you stay upright.

Dunkley’s model treats regulation as more than distress tolerance, more than getting through the moment without doing damage. It asks you to validate the emotion once you’ve identified it, check whether its intensity actually fits the facts in front of you, then use something like Opposite Action, deliberately moving against the emotional pull, to bring the response back in line with what the situation warrants. The whole method rests on being able to catch the primary emotion before the secondary one takes over the room.

That’s a sound clinical approach. It also assumes a gap. A window between the shame landing and the anger answering it, in which a person can notice what’s happening and step in before the reaction chooses for them.

The reframe worth sitting with: in an ADHD flood, that gap often isn’t there.

This is the part people who don’t live it struggle to believe, and the part people who do live it recognise instantly. There is supposed to be a beat. A fraction of a second where the first feeling registers as a feeling, a small hang on, before the body commits to anything. Most people get that beat. It’s where the choosing happens. Naomi doesn’t get it, or gets it so compressed that it’s gone before it’s usable. The shame and the anger arrive close enough to overlap, and by the time either one reaches the part of her that could name it, the anger is already talking, already out of her mouth, because anger recruits the body faster and more completely than shame ever will. It moves at the speed the flood moves. Shame is slower, and by the time it would have surfaced, the shouting has already happened and given her something new to be ashamed of.

She didn’t suppress the shame and choose anger instead. She was never offered the choice. There simply wasn’t a version of that moment where the shame arrived early enough to be dealt with on its own terms.

This is why “just notice what’s underneath the anger” is sound advice and useless instruction at the same time. It asks for a pause that assumes ordinary processing speed. The flood doesn’t leave that pause on the table. Dunkley’s framework isn’t wrong. It’s describing a window that, for a lot of ADHD adults, has already closed by the time anyone could reach it.

Subscribe now

What survives isn’t catching it live. It’s going back for it afterwards, on purpose, while the memory is still warm enough to hold.

The useful question is not what made you angry. It’s what would have made you feel small, or exposed, or caught out, in the seconds just before the anger showed up. Naomi’s answer, once she went properly looking, was shame about the mess and a quieter fear underneath it, that the unloaded dishwasher confirmed something she already half-believed about herself and didn’t want confirmed out loud. Neither of those was available to her on the stairs while the adrenaline was still leaving. Both were available a day later, in a chair, if she knew what she was looking for.

This is part of why the CARE grounding tool builds the Reactor profile around exactly this pattern, the anger-first flooding type where the real work happens in the pass back through it rather than in the live moment. If any of this sounds like your own arguments, the Reactor sequence is built to run that retrospective pass properly, with structure, rather than leaving it to whatever you can reconstruct on your own.

For a partner on the receiving end, this changes what the row actually was. A shame response that arrived wearing anger’s clothes, because anger got there first and shame took longer to reach the surface. That doesn’t lift the accountability for the shouting, and Naomi would be the first to say so. What it does is move the useful conversation, out of the argument itself, where nobody can reach anything, and into the quieter time afterwards, when both people can ask what the thing under it actually was. If this is the row you keep having in some form, it might be worth sending this to the person on the other side of it, not as an excuse, but as a starting point for the conversation you keep failing to have while it’s happening.

The experiment, if the dishwasher moment is familiar, is to run that pass back deliberately the next time something small produces a reaction far too big for it. Not in the moment. The moment is lost, and trying to win it live is its own trap. Afterwards, once the flood has fully gone and your hands are steady again, ask what would have made you feel exposed or caught out in the seconds before the anger arrived. Write the answer down even if it’s unflattering. Especially then. The unflattering ones are usually the true ones.

Naomi came back the following week and told me the next argument had been shorter. Not because she’d caught herself in time. She hadn’t. But because she’d stopped being blindsided by her own reaction, and stopped needing to defend it as reasonable once it was over. Knowing what was underneath it took some of the charge out of the top of it.

Next piece in this series comes back to why naming any of this precisely, rather than reaching for “stressed” or “fine,” turns out to depend on a sense most people assume runs automatically, and which, for a lot of us, doesn’t.

If you’ve had the Naomi experience, the argument you found yourself picking apart afterwards to work out what it was actually about, I’d be interested to hear what was underneath it once you went looking.

Thomas had written the script out the night before. DEAR MAN, the way his DBT group had taught it: Describe the situation, Express how you feel, Assert what you need, Reinforce why it matters, staying Mindful, Appearing confident, Negotiating if you had to. He’d rehearsed it twice on the drive in, word for word, the way you’d rehearse lines. He was asking his manager for an extension on a project that had slipped for reasons that were partly his and partly not, and he wanted to say it without either grovelling or getting defensive. He had a plan for that. He’d built one.

His manager interrupted him after the second sentence. Not unkindly, just a clarifying question, the kind that happens in every conversation like this. And the script was gone. Not paused, not adjusted. Gone, the way a phone number you were repeating to yourself disappears the second someone asks you your name. What came out instead was three minutes of over-explaining, a mention of his ADHD he hadn’t planned to make, and an offer to work the weekend that he immediately regretted and that solved nothing.

He told me about it the way people tell me about the counting-to-ten failures. Slightly bewildered. “I knew exactly what to say. I’d practised it. It just wasn’t there when I needed it.”

Dialectical Behaviour Therapy takes emotional intensity seriously in a way that a lot of frameworks don’t, and Marsha Linehan built something genuinely comprehensive out of that seriousness. She described three states of mind: Emotion Mind, where feeling drives behaviour; Reasonable Mind, the cool logical end; and Wise Mind, the integration of both, which is where DBT wants you to operate from. Underneath all three sits mindfulness itself, the skill of observing, describing and participating in the present moment without judgement. Linehan treats this as the foundation the other modules are built on, not a single step tucked inside a wider skill. From that foundation comes a full set of modules. Distress tolerance skills for surviving a crisis without making it worse. Emotion regulation skills for reducing vulnerability to intense states in the first place. Interpersonal effectiveness skills, DEAR MAN among them, for asking for things and holding boundaries without losing the relationship. It’s a serious clinical system, and it deserves to be treated as one rather than waved through as background noise before getting to the critique.

Hesslinger and Philipsen adapted DBT skills training for adult ADHD over two decades ago, and their programme already sorts modules partly by cognitive demand. What follows is my own reading of where that adaptation meets the flooding research, not a summary of their protocol, and it’s worth being clear about that distinction before going further. The wider evidence base is worth naming too. DBT-based treatment for adult ADHD carries moderate, mixed support. Zhang and colleagues’ 2026 head-to-head trial found DBT and CBT both improved outcomes over time without DBT establishing non-inferiority, and NICE guidance recommends structured psychological support for adult ADHD without recommending DBT specifically. DBT-for-ADHD is a working adaptation rather than a settled treatment of choice, and the mechanism argument below stands on its own regardless of where that wider debate lands.

Here’s where the critique sits. DBT was built for a population whose central problem is a wall of intense emotion in front of a reasoning mind that, once it can be reached, is intact. The route into Wise Mind runs through steps that need to be held in working memory and executed in sequence. STOP asks you to Stop, Take a step back, Observe, and Proceed mindfully, four moves in order, each depending on the one before it. Check the Facts asks you to hold an interpretation and a set of actual facts side by side, then weigh how likely the feared outcome really is against a fair reading of the evidence, closer to running a small hypothesis test than a simple comparison. DEAR MAN is a script, and a script needs a working memory to hold it while a real, unscripted conversation happens on top of it. Arnsten’s work on the prefrontal cortex under acute stress explains why that working memory disappears exactly when it’s needed most: catecholamine surges rapidly weaken the same prefrontal networks that hold information in mind, while strengthening the more automatic responses running underneath them.

This isn’t only a stress-physiology story. Adults with ADHD often understand a coping plan perfectly well in session and still fail to deploy it in daily life once initiation, working memory, time blindness, stress and competing demands get in the way. That gap between knowing a skill and using it under load runs through the adult ADHD-DBT literature, and Thomas’s evening in the car, script word-perfect, gone within one clarifying question, is that gap in miniature.

Which DBT skills survive an ADHD flood? The physiological ones do, and the ones you’ve already decided on before the flood arrives. TIPP, the temperature, intense exercise, paced breathing and muscle relaxation sequence, works through the body rather than the reasoning mind, so it doesn’t need the part that’s gone offline. Opposite Action survives too, but only if the action has already been chosen while calm. Radical Acceptance belongs on the same side of the ledger. It asks for a stance rather than a construction, no comparison, no multi-step sequence, just a turning towards what is already true, which makes it closer to a posture than a procedure and easier to hold onto once reasoning capacity has gone. Anything that has to be invented, sequenced, or held in mind during the flood asks for a resource the flood has already taken. Anything physiological, or anything pre-decided, doesn’t need that resource at all.

That’s the same line Gross drew between antecedent and response-focused strategies, and DBT turns out to sit across it rather than on one side. STOP, Check the Facts, and DEAR MAN are all response-stage moves dressed up as structure, and the structure doesn’t change where in the sequence they’re being asked for. Opposite Action is different. Decided on beforehand, for a specific known trigger, it stops being an in-the-moment invention and becomes something closer to a rehearsed default, available precisely because nothing had to be built from scratch under load. Thomas’s script failed for the same reason the counting failed in the last piece. It was designed to be produced under conditions where production is the one thing you can’t do.

The physiological route, the one that survives, is where the CARE grounding tool sits as well. The temperature and breathing sequence in TIPP is close to what CARE’s grounding tool runs on, built around knowing your type and having a route through the flood that doesn’t ask anything of the part of your brain that’s already gone. If the in-the-moment flood is where you’re stuck, that’s the place to start:

https://care.thecontextshift.com/

DBT has two halves, and only one of them is built to survive the exact moment it’s usually reached for. Call this the Calm-Window and Storm split. It’s the frame worth keeping, because every skill in your toolkit, DBT or otherwise, sits on one side of it or the other. That split shows up in trial data too. Halmøy and colleagues’ 2022 multicentre RCT found DBT-based group treatment beating treatment as usual on self-reported executive function, ADHD symptoms, depression and quality of life, all genuinely calm-window gains, while showing no significant advantage on emotion dysregulation scores measured straight after treatment, the one outcome closest to what actually happens in the room during a flood. The interpersonal skills, the fact-checking, the mindful pause between stimulus and response, all of that does genuine work in the calmer hours either side of a flood, when there’s a working memory available to run it. Rehearsing DEAR MAN before a difficult conversation, when you’re calm enough to think it through properly, builds something useful. Reaching for it mid-argument asks it to do a job it was never built to do at that moment. Deployment timing decides whether a skill works, not the skill itself.

One caveat worth holding: a DEAR MAN script rehearsed to the point of overlearning starts to behave more like a physiological response than a constructed one, which is the whole point of DBT’s own theory of change. A skill’s place in either pile isn’t fixed forever. It’s where that skill sits for you right now, given how much practice it’s had.

Here’s the experiment, if you want to run one. Take one DBT skill you already know, from a group, a book, or a conversation with a therapist. Ask honestly which half it belongs to. If it needs you to observe, compare, or construct a sentence in real time, it’s a calm-window skill, and its job is to be used before the trigger, not during it. Rehearse it, script it, decide it in advance for a specific recurring situation. If it works through the body rather than the mind, temperature, breath, movement, or through a stance rather than a sequence, it’s a storm skill, and it’s the one you reach for once the window has already shut. Sort the ones you know into those two piles before you need either of them, rather than discovering under pressure which pile a given skill was always in.

Thomas needed to decide, calmly, what he was going to do the next time a conversation went somewhere he hadn’t planned for, rather than trusting that the plan he’d already written would still be there when it mattered. It rarely is. It’s what happens when a skill built for one half of a sequence gets asked to perform in the other, a timing mismatch rather than a personal or clinical failure.

People tend to know immediately which pile a given skill belongs in, once they’ve been asked to sort. Name one of each in the comments: a skill that’s always let you down mid-storm, and one that’s somehow still worked even then. If you know someone who’s watched their own DEAR MAN evaporate mid-conversation, forward this to them before their next one. The next piece in this series looks at what happens when the flood isn’t yours, when you’re the calm one trying to reach someone who’s already gone.

The Context Shift is psychoeducational writing. Nothing in it, or in my replies, replaces your own clinical care.

Priya had done everything the last clinician told her to do. When she felt the heat climbing, she was going to count to ten. She was going to take the slow breath. She was going to step back before she said the thing she would regret.

She came back and told me it had not worked once. By the time she noticed the heat, she said, she was already three sentences into saying the thing. The counting happened somewhere over the top of her own voice. The breath came after the door had slammed. She had followed the instructions exactly, and the instructions had arrived too late every single time.

Then she said the part I hear most weeks, in slightly different words each time. “So I think it just doesn’t work for me. I think this is who I am now.”

The technique had genuinely failed her. The reason sat somewhere she had not been told to look.

There is a model of how emotion regulation works in time, and it explains Priya’s experience almost exactly. James Gross has spent decades mapping this, and he describes regulation as something that happens along a sequence rather than in a single moment. An emotion does not switch on fully formed. It builds. There is the situation you are in. There is what you attend to within it. There is the meaning you make of what you noticed. And then, last, there is the response that comes out of you. Gross’s point is that you can act at any stage along that sequence, and that the stages are not equal.

The strategies that act early, before the emotion has gathered its full force, he calls antecedent-focused. Choosing which situations you walk into. Changing something about a situation while you are still calm enough to change it. Directing your attention. Reframing the meaning before it hardens. The strategy that acts late, after the emotion is already running, is response-focused, and in practice it almost always means suppression. Holding the face still. Swallowing the reply. Counting to ten while the body has already committed.

Now look at where nearly all the advice an ADHD adult receives actually sits.

Pause before you react. Take a breath. Count to ten. Walk away before you say it. Notice the feeling and let it pass. Every one of those is a response-stage move. Every one of them asks you to act at the latest possible point in the sequence, the point where the emotion has already arrived at full strength.

And here is what the last piece established, carried forward. The flood takes the prefrontal cortex offline. So the advice is asking you to perform a deliberate, effortful, top-down action at the precise moment your capacity for deliberate, effortful, top-down action has been switched off. The counting is not failing because you are doing it wrong. It is failing because it depends on a brain state the flood has already removed.

People sometimes reach for the one antecedent strategy that does get taught, which is reappraisal. Reframe the situation. See it differently. And reappraisal genuinely is an earlier move than suppression. But it runs on the same machinery. Reframing a situation in real time is one of the most executively demanding things a brain can do, and the ADHD brain mid-flood has the least of that resource available at exactly the moment the technique asks for the most. So the two things most commonly offered, suppress it or reframe it, fail for the same underlying reason, in the same brain, at the same moment.

Subscribe now

There is also a cost to the suppression itself that is worth naming. Gross’s own work found that holding an emotional response in does not reduce the internal experience of it. The feeling carries on at full strength underneath. What suppression adds is a second tax on top, a metabolic and cognitive load, so you end up paying for the storm and paying again to hide it. Anyone who has held it together through a meeting and then come apart in the car already knows this in their body.

The useful question is where, along that whole sequence, an ADHD brain still has its full capacity available. The answer is the early end. Before the trigger. Before arousal has climbed. Situation selection and situation modification (the least glamorous part of the entire model, and the most useful) happen while the prefrontal cortex is still online, because nothing has flooded yet. This is the part almost nobody is taught to use, and it is the part where the control actually sits for this neurology.

What that looks like in practice is less dramatic than counting to ten and far more effective. It is deciding not to open the difficult email last thing at night when you are depleted, and opening it in the morning with food in you instead. It is having the hard conversation on a walk rather than across a kitchen table, where the eye contact and the stillness raise the heat. It is building a deliberate delay into your replies, so the message that goes out is not the one the flood wrote. It is noticing that a particular meeting, a particular person, a particular time of day reliably lights you up, and arranging the conditions around it before you arrive, rather than trying to white-knuckle your way through once you are in it.

None of that requires you to be calm in the moment. That is the whole reason for moving upstream. You make the move while you still have the brain to make it, so that the moment, when it comes, is smaller before you ever have to manage it.

This is the second piece in a series working through the established emotional regulation frameworks one at a time, each one held against what an ADHD flood actually does to it. If you want the rest as they land, subscribing is how they reach you.

There is a name for the more deliberate version of this, rehearsing a difficult moment in advance while you are calm, so the flooded brain does not have to invent a response on the spot. I am going to come back to it properly later in this series, because it belongs to a body of work that deserves its own piece. For now it is enough to notice the shape of it. The work happens upstream of the moment, while you are still the one in charge of it.

Now, the honest limit of all this. Antecedent moves are for before. They do nothing for the times you are already in it, when the situation arrived with no warning and the storm is already on you. In that moment there is no earlier point left to act on, and the thinking brain has gone, so the only thing that still reaches you is the body. Slowing the exhale until it runs longer than the inhale. Cold water on the face. The physiological brake that works from the bottom up, because it does not need the part of the brain that has gone offline. That acute, mid-storm route is what the CARE grounding tool is built around, and if the in-the-moment flooding is your main problem, then this is where I would start: https://care.thecontextshift.com/

For most of the people I see, though, the bigger discovery is that they had been aiming all their effort at the one moment where they had the least to work with, while ignoring the hours beforehand where they had the most. Priya’s progress came from a single antecedent change. She stopped having the hard conversation at eleven at night, when she was emptied out, and moved it to a Saturday-morning walk instead. The counting never improved. It did not need to.

Here is the experiment, if you want to run one. Pick one emotional flashpoint that recurs. Not a random bad day, a predictable one. The Sunday-night dread before the working week. The specific colleague. The school run. Then, instead of resolving to handle it better when it hits, change one condition before it arrives. Move the timing. Change the setting. Eat first. Build in a delay. Take it for a week and watch one thing only: whether the flashpoint itself arrived smaller because of the condition you changed in advance. That is the measure that matters here, because the size of the storm before it lands is the part this neurology can actually move.

It will not work every time. Some storms arrive with no runway at all, and that is what the in-the-moment work is for. But run the experiment honestly and you will find a whole category of flashpoint you had been treating as a moment-management problem when it was a planning problem all along, sitting upstream, in hours you did not know were yours to use.

This is the first of the established frameworks I said I would take one at a time. Gross gives us the map of when regulation can happen, and shows that most of the advice aims at the wrong moment on it. The next one is the one most people have actually heard of, the set of skills clinicians reach for most, and it is also the one with the most to answer for when it meets an ADHD flood. Some of it survives the storm intact. Some of it slides straight off. That is where this goes next.

Almost everyone I say this to goes quiet for a second and then names their one reliable flashpoint without hesitation. The Sunday night. The name in the inbox. The handover at the front door. They always knew exactly when it was. They had just never been told that the hour before it was theirs to use.

The Context Shift is psychoeducational writing. Nothing in it, or in my replies, replaces your own clinical care.

Ralph came back to clinic about six weeks into treatment and told me the medication was working. He could read a report to the end. He had cleared a backlog that had followed him from job to job for the best part of a decade. Meetings held their shape. Then, with his coat already half on and his hand near the door, he mentioned the other thing, the way people mention the thing they actually came to say. His laptop had frozen before a call that morning and he had lost his temper over it, sharp enough that a colleague nearby had gone quiet, and the size of the reaction had stayed with him for the rest of the day. The tablets had not touched that. “The ADHD’s sorted,” he said. “That part’s just me.”

I have heard that sentence, or a close version of it, more often than almost any other in this work. The focus improves. The emotional weather does not. And somewhere in the gap between the two, the patient quietly files one half of themselves under treatment and the other half under character.

What lands in that second category is always the emotional part. The frustration that arrives faster than the thought that might have softened it. The upset that floods in over something small and leaves a residue of shame once it has cleared. The eyes that fill at the wrong moment in a meeting. The reaction that is three sizes too big for the event that set it off. People build a whole identity around this. The short fuse. The sensitive one. The one who takes everything to heart. Most have been hearing some version of it since childhood, long before anyone said the word ADHD out loud, which is exactly why it has the texture of bedrock. It feels like the truth about who they are.

When Ralph reached for the door, he was already running that audit. The short fuse, filed under character. Here is what tends to be missing from that diagnosis.

That belief is expensive. If the storms are simply your character, then every one of them becomes a moral failing, and the only response available to you is to try harder to be a better person. So people apologise for years. They run private audits of their own decency. They treat a neurological event as evidence of a flaw in who they are, and they carry the weight of it everywhere, into work, into their relationships, into the quiet hours afterwards when they lie awake replaying what they said.

I have watched this conclusion settle into people who have every reason to question it. People who are thoughtful, self-aware, clinically literate. The belief that the emotional part is just you is remarkably resistant to evidence, partly because nobody ever hands you the evidence. The diagnosis arrives wrapped in language about attention and focus and organisation. The emotional half rarely gets named at all.

The flooding types, the window of tolerance, the interoception work - those are about recognising and surviving what happens inside a flood. This is about what the established clinical frameworks for emotional regulation assume you can do, and where they quietly break down for a brain like this one.

Russell Barkley, who has spent much of a career on this, argues that emotional dysregulation is a core feature of ADHD itself. He calls it deficient emotional self-regulation. The point is mechanical. The prefrontal systems that let a person hold attention on something dull, hold back an impulse and wait a beat before acting are the same systems that hold a brake between an emotion and its expression. When those systems run short, both functions take the hit. The wandering attention and the oversized reaction come from one place. They were always the same condition.

So when people ask whether the emotional side is part of their ADHD or something separate sitting next to it, the clinical answer is that it is part of the ADHD, and central to it. The same executive systems that regulate attention also regulate the gap between feeling something and acting on it, which is why a brain that struggles to hold focus will so often struggle to hold an emotional brake under pressure. A dose tuned to attention can sharpen the focus and still leave the reactions largely standing - and for many people, it does. Which is why the part you have spent your life filing under personality has a neurological address.

There is something else worth saying here, and it tends not to get said. People manage their ADHD through different routes. For some it is medication. For others it is exercise, or the changes that come with sleep and structure and cutting the load down to something the system can carry. What they often notice is the same shape Ralph noticed. The recognisable symptoms ease - the focus, the restlessness, the scattered attention the diagnosis is written around - while the emotional dysregulation carries on much as before. That gap is the thing people misread as proof the storms are character. It is worth reading differently. When the rest of the system settles, the cognitive load that was consumed by tracking everything at once starts to lift, and the moves that emotional regulation actually requires - noticing a reaction building, pausing before it lands, choosing what comes next - become reachable in a way they were not before. The settling does not regulate the emotion. It builds the pedestal you can finally stand on to do that work.

So I am not going to hand you an excuse here, and I do not want one either. ADHD is real and the storms are neurology, and you are still accountable for what happens in their wake. What changes is where the work goes. You can stop spending it on becoming a fundamentally different person, because that was never the thing that needed solving, and start putting it toward the kind of regulation the brake actually responds to. The apology you owe a colleague still stands. The decade spent apologising for existing does not.

That reframe is load-bearing, and it is not yet the whole structure. The emotional weather has a structure, and structure has leverage points. Once you can see them, you can find the places inside a flood where a brake is still within reach, even in a brain like this one. Over the coming pieces I will take the established frameworks for emotional regulation one at a time, give each its due, and then look honestly at the exact spot where each one quietly assumes a calm, available prefrontal cortex that an ADHD brain in full flood cannot supply.

One practical note for the meantime. When a storm is already building and the thinking brain has gone offline, the moves that still reach you are physiological. I built a small browser tool around that, around the type of flooding you tend to run and a grounding sequence for the moment it actually hits. It sits here if it is useful to you.

https://care.thecontextshift.com/

That shift tends to arrive at a specific moment, and people tend to remember which one it was. Some are still waiting for it. Either way, it belongs in the comments.

The Context Shift is psychoeducational writing. Nothing in it, or in my replies, replaces your own clinical care.

Mark came to our second appointment with a notebook.

He had been doing a mindfulness practice for six months, daily without exception. He could describe what the window of tolerance was; he had been to a psychoeducation group, read around it, and arrived with a working knowledge that was better than most. He knew about bottom-up regulation. He could explain, with reasonable accuracy, why grounding techniques worked.

In a review meeting with his line manager the previous Tuesday, he had been unable to use a single thing he knew.

“I know the theory,” he said. “I just don’t have access to it when I need it.”

He had concluded the problem was execution: he needed to practise harder, remember more reliably, try differently in the moment.

I had a different reading.

What the window of tolerance actually is

The window of tolerance is the physiological range within which incoming stimuli can be processed without triggering a survival response. Within it, the nervous system can receive, organise, and respond. Below it (the hypoarousal zone) the system withdraws or shuts down. Above it (the hyperarousal zone) it floods.

Regulation strategies, all of them, are designed for a nervous system that is somewhere within that range. Breathing exercises, grounding techniques, the deliberate pause before responding: these are maintenance tools. They act on a system that is within reach. They are not rescue equipment for a system that has already left range.

Mark knew his tools. The question I was starting to form was where the window was before the meeting started.

How ADHD changes the window

In ADHD, the window has several specific characteristics that mainstream regulation training tends not to account for.

The baseline is often narrower than in a neurotypical nervous system. The same neural systems involved in attention dysregulation (the dopaminergic and noradrenergic pathways) also affect arousal regulation: how quickly the nervous system escalates under load, how readily it returns to baseline, and how much tolerance it has before reaching a ceiling. Smaller range; less margin for ordinary variation.

The window also narrows faster under load. A standard meeting is more demanding for an ADHD nervous system than for a neurotypical one. The self-regulation running throughout (managing impulse, tracking reactions, monitoring presentation) is active, continuous work with a real depletion cost.

And the factor most often missed: masking debt narrows the baseline. An adult performing regulation rather than experiencing it over an extended period does not return to a full baseline window overnight. The window at 8am on a Tuesday, following several months of sustained performance, is already narrower than it would otherwise be. (What the previous piece in this series called performance depletion, the cost of maintaining a version of yourself all day, operates on the window as well as on the body’s energy reserves.)

The implication of these three factors together: an ADHD adult arriving at a difficult meeting has often already consumed significant window margin before anything has happened. The event that tips the system past its ceiling is frequently the last straw of a window that was already narrow.

The paradox at the heart of most regulation advice

Here is what this means for Mark, his notebook, and six months of diligent practice.

Regulation strategies require enough window to work. When the nervous system is already near the ceiling, it cannot take on the additional attentional load of deploying a technique. Initiating and sustaining a regulation practice draws on the same prefrontal resources that have been running at high output since the morning.

The tools are sound. The window was already too narrow for the tools to fit.

This is the mechanism behind one of the most demoralising experiences in ADHD: knowing exactly what to do and finding it inaccessible at the moment it matters. That experience is a structural mismatch between when the strategy is needed and the conditions under which it actually works.

Why do ADHD regulation strategies fail when I most need them? Because regulation strategies are designed for a nervous system within the window of tolerance, a range that requires a baseline level of available executive resource to access. In ADHD, the window is narrower than average, narrows faster under sustained load, and is further compressed by masking debt accumulated across a working day or week. By the time a difficult event triggers the need for a strategy, the system has often already left the range in which that strategy can be initiated. The window was the problem.

The window was already closed before he sat down.

Managing within the window versus widening it

Most ADHD regulation guidance, including what I was trained to deliver, addresses the first problem: how to stay within the window, or how to return to it once briefly exceeded.

Very little is designed for the second, different problem: how to widen the window itself, or how to prevent its progressive narrowing.

These require different interventions at different timescales.

Managing within the window requires strategies. Widening the window requires recovery: specifically, rest from performance rather than from output, as I described in The Cost of Composure. The removal of the self-monitoring requirement, so the nervous system’s baseline can restore toward its actual width.

Medication widens the window pharmacologically. Sleep and consistent physical exercise also contribute; in exercise the active mechanism is bottom-up nervous system regulation produced by physical effort, a process the executive system cannot interrupt. And the window widens under genuine safety: relationships and environments where the monitoring requirement is actually suspended.

Sustainable window widening tends to require changes to the conditions under which the narrowing occurred: working environment, relationship conditions, sustained performance requirements.

That is a structural question. Most ADHD treatment is not designed to ask it.

What helps when the window is already narrow

In the meantime, Mark still had to attend meetings.

When the window is already narrow, the strategies that work are those that act on the nervous system through routes that don’t depend on executive function: movement, leaving the room, sensory change, cold water.

These are first aid for a system that has already left its range: getting back inside, or close enough that the more considered tools become accessible again. They are what is actually available when the window has already closed.

The practical question Mark and I started working on was what the window looked like before the meeting began, and what that meant for the margin he actually had.

The experiment

On the morning of something you expect to be demanding, take thirty seconds before it begins to notice the window’s current state. That is the whole experiment: knowing where it sits before anything is asked of it.

Tightness already present. Narrowing that precedes the event. Irritability with no apparent cause, which usually means a cause that arrived earlier than the feeling did.

The window’s state going in determines the margin available. A system at 30% capacity approaching a difficult event requires different preparation from a system at 80%. Knowing the difference, before anything has happened, changes what is possible after.

Mark returned three months later. He had not acquired new strategies. He had developed the habit of checking the window’s state before demanding conditions, and adjusting what he asked of it accordingly.

“I still lose access to the tools sometimes,” he said. “But I’m starting to understand when that’s going to happen. It’s the days when I arrive already at the ceiling.”

The pattern that emerges consistently, across patients who start attending to the window’s baseline, is that the days that end badly almost always started narrow. The event that landed wrong was rarely the cause.

The tools work when the window allows it. Most ADHD treatment teaches the tools. Very few frameworks teach the window first.

The arrival state is data. It is also, for most people with ADHD, data that has never been named as such.

Readers who already track sleep, medication timing, or recovery days tend to notice this pattern within a week. Readers who don’t usually trace it only after several days have landed the same way, each one blamed on whatever happened to trigger it rather than on the state that was already narrowing before it did.

Reading the state you’re actually in

The Context step of the CARE framework is the window check: a brief, specific question about the nervous system’s current state before deciding what comes next.

If the nervous system sequence from the past month has been relevant to your experience, the CARE framework is the practical tool built from it.

https://care.thecontextshift.com

The Context Shift is psychoeducational writing. Nothing in it, or in my replies, replaces your own clinical care.

Daniel came to see me three weeks after what he described as a manageable fortnight. The meetings had gone smoothly enough. He had kept it together through two difficult conversations at work. He had replied to emails, met his deadlines, made it through each day without obvious incident.

He arrived in clinic looking exhausted.

He had snapped at his children the evening before over something trivial. He had sat with his partner afterwards unable to explain it. “Nothing went wrong,” he said. “That’s what I can’t understand.”

He had looked, from the outside, like a person who was coping. He had, by his own account, been coping. The fortnight had passed without a visible crisis.

What had not passed was the cost of passing it.

Depleted from doing, depleted from performing

There is a distinction that rarely gets named in ADHD conversations, and I find it matters considerably once patients start recognising this pattern.

Depleted from doing is ordinary tiredness. A heavy workload, a difficult day, too many decisions. The body has worked hard and needs rest. Most recovery guidance addresses this kind.

Depleted from performing is something different. It is the cost of sustained self-regulation: holding reactions back, monitoring impulse, tracking how you appear to others and adjusting in real time. The body is not just doing work. The body is maintaining a version of itself that requires continuous active upkeep (and, unlike the work on the calendar, this maintenance does not appear in anyone’s account of how the day went — including your own).

For adults with ADHD, the gap between those two states is wider than it looks. A nervous system managing ADHD does not have a neutral gear. It is either containing or it is not. Performing regulation, looking regulated, presenting as regulated. This all requires active effort regardless of whether the day was objectively demanding.

A manageable day is a day in which the cost was invisible.

Subscribe now

Where the cost goes

The interoceptive system, the body’s capacity to register its own internal states tends to run with a significant lag in ADHD adults. (If the interoceptive delay is new territory, this piece on why the body’s report arrives late is worth reading alongside this one). The body’s account of what has been happening to it arrives late. Sometimes hours late.

This has a specific consequence for the cost I am describing. The signal that the nervous system has been running at high output all day does not register in real time. By the time it reaches awareness, it has accumulated. It surfaces as flat affect, disproportionate irritability, or a sudden inability to tolerate noise, mess, or small friction in the environment where the performance is finally allowed to stop.

That environment is almost always home. The person it arrives at is almost always the partner, or a child, or the silence.

I have come to think of this collapse, the flooding and shutdown at the end of a day that looked fine — not as something coming from nowhere, but as a report that was delayed. (This is the same mechanism behind the Shutter pattern). The body was keeping score throughout the day. It just was not reporting until it was somewhere safe enough to do so.

Which creates its own problem. The place that should offer recovery becomes the place where the bill arrives.

Why do I feel exhausted when nothing went wrong?

This is the question I hear most often from patients who have been managing their ADHD reasonably well for some time. They have learned to get through the day. What they cannot account for is why getting through the day keeps leaving them empty.

The answer that fits best clinically: the ADHD nervous system does not rest while performing composure. Appearing regulated and being regulated are different physiological states, and the effort of the former accumulates a cost the latter would not.

Allostatic load describes the biological wear accumulated from sustained stress. In an ADHD adult who has spent a working day presenting as composed, be it holding eye contact, managing reactions, pacing speech or suppressing impulse; underneath, the nervous system has been running at high output even when the external picture looked unremarkable. The interoceptive delay means this cost does not register in real time. It arrives in the evening as flat affect, disproportionate irritability, or shutdown in the environments where the performance was finally allowed to stop.

At the level the body was working, the day had a cost the calendar never showed. The collapse is the report arriving late.

And the practical implication that follows: the recovery required is not rest from output. It is rest from performance. Those are different states, and treating one with what helps the other, such as less activity, an early night or a quiet evening, tends not to address what is actually depleted.

What rest from performance looks like

Rest from output means not doing things. Rest from performance means removing the requirement to monitor and adjust the self.

The tools designed for top-down regulation (mindfulness practices, breathing techniques, cognitive reframing) tend to struggle here, for reasons I examined in this piece). They ask the nervous system to regulate itself using exactly the same machinery it has been running all day. What tends to help more is bottom-up: movement, sensory change, physical rest, or simply being somewhere the performance is not required and the self-monitoring can actually stop.

Some people find that in solitude. Some in physical activity that pulls attention entirely into the body. Some in relationships where they are genuinely, not strategically, allowed to be themselves.

The shutdown and flat affect that follow a manageable day are not evidence that something is broken. They are the beginning of recovery. The problem is that they arrive at the wrong moment, in the wrong place, directed at the wrong person.

Reading the body’s signal earlier in the day changes what is possible. The cost accumulates regardless. The difference is having the information while there is still something to be done with it.

The experiment

Before the usual evening context, perhaps before the conversation about the day, before the news, before whatever the evening routine requires, give yourself 60 seconds attending to the body’s state. What is present right now: what the muscles are holding, what the chest feels like, what the actual level of activation is, separate from the story of whether the day was demanding.

The cost is set. What changes is what happens next.

Daniel came back six weeks later. He had not resolved the pattern, but he had started to name it: the managed week, the delayed report, the evening that belonged to the accumulation rather than to the small thing that arrived at the wrong moment. “I still get flooded,” he said. “But I’m less surprised by it now.” In my experience, that is usually where things start to shift.

The pattern I hear most consistently from patients who start naming this is that the collapse tends to arrive at the moment they feel safest. Which is usually the first moment all day that the performance was finally allowed to stop.

The person it arrives at tends to be the one who was never told they were receiving it.

Dr Ketan Bhatt

The Context Shift is psychoeducational writing. Nothing in it, or in my replies, replaces your own clinical care.

If the day left you feeling like this and you are looking for somewhere to start, the CARE framework’s Context step was built for exactly this gap — a way to check what is actually present in the nervous system before the story about the day gets added.

https://care.thecontextshift.com/

Rachel came in for her initial assessment describing something she had never found language for.

She had walked out of a colleague’s office in tears three weeks before. Unprompted. In the middle of what had been, by all available evidence, a routine meeting. When her husband asked what had happened, she couldn’t tell him. When she thought back over the previous two hours, she found it in retrospect: the slight tightening when the meeting overran. The edge in her voice that her colleague had gently named. The growing sense that she was being criticised, building quietly over ninety minutes, without her registering that it was building.

“I thought I was fine until I wasn’t,” she said. “There was nothing in the middle.”

That phrase. I have heard it in various forms across twenty years of practice. It describes something precise. The warning was there. It wasn’t readable as a warning.

And once you understand it, a number of things that have looked like character flaws or failures of attention start to make a different kind of sense.

What the body knows before you do

Interoception is the body's ability to sense its own internal states. What matters here are the early physiological signals that precede emotional states. The slight constriction in the chest before an argument. The change in breathing pattern in the seconds after a perceived slight. The attentional narrowing that begins in the first moments of a stress response, before any conscious sense of being stressed.

It operates largely below conscious awareness. It runs a continuous feed from organs, muscles, and nervous system to the brain, translating what is happening inside the body into something the rest of the system can act on.

When interoception works as expected, you notice you are getting tense before you are tense. That early signal is what allows any regulation strategy to be deployed in the seconds or minutes before a crisis, when there is still something to act on.

When that feed is delayed, unclear, or unreliable, the first legible signal is often the acute state itself. You are already burning.

Why ADHD disrupts this

The networks affected by ADHD (prefrontal systems, attentional control, executive function) are the same networks involved in processing interoceptive signals. Taking information from the body and integrating it into conscious awareness is not automatic. It requires the brain to attend to internal state, to translate physiological data into something legible, and to hold that information available for decision-making. All of that runs through systems that ADHD makes less reliable.

The disruption is upstream of decision-making. The body sends the signal. Between that signal and conscious awareness is where things go wrong: the translation fails, or arrives late, after the state is already acute.

Subscribe now

Why the flood arrives without warning

This is where the five types of ADHD emotional flooding I have written about over the last several weeks become clearer at a different level.

The Reactor has no gap between stimulus and response. The usual explanation is impulsivity. But there is something else underneath: if the early signals of rising emotion don’t arrive into awareness, there is no early signal on which impulse control could act. The gap never opened.

The Hidden Flooder manages external presentation while flooding internally, because the flooding isn’t legible to them as flooding while it is happening. It becomes legible in the exhaustion and irritability of the aftermath, when the body’s report arrives late and the moment is already past.

The Shutter goes quiet without registering how close the threshold was. The build toward overwhelm wasn't legible in time. By the time the withdrawal happens, there was no window left

The Anticipator’s window of tolerance narrows before the difficult conversation begins. The nervous system is already reading threat cues, already activating, already narrowing. The conscious account of the situation hasn’t caught up. Fully activated before anyone has said anything.

And the Ruminator’s 3am replay contains the emotional processing that the body began during the original event but that never reached conscious awareness on time. The replay is the system catching up.

The interoceptive deficit is a partial account. What it does explain, across all five types, is why recognition arrives after the event rather than during it. The early information simply doesn’t arrive on time.

The “nothing in the middle” account is also the most useful thing to share with someone who keeps asking why you didn’t just notice and stop yourself. The mechanism is the answer.

This is also why last week’s piece on mindfulness matters at a different level. The top-down/bottom-up mismatch explains why mindfulness fails during a flood. The interoceptive deficit explains why the flood has often been building for ninety minutes before the person knows it is happening. The window for any intervention closes earlier than most regulation training addresses.

What helps

There are two separate problems here, and they require different responses.

The acute problem (managing the flood once it has arrived) is what last week’s piece addressed. Top-down tools don’t work reliably inside a bottom-up state. Movement, cold water, leaving the room: these act on the nervous system through channels that don’t depend on executive function. None of that changes here.

The upstream problem: building earlier detection

This is a practice for low-arousal states. Brief, deliberate attention to what is happening in the body before any emotional charge is present. Noticing muscle tension at rest. Noticing the quality of breathing before a meeting. Noticing the small shifts that happen in the chest, throat, or jaw in response to ordinary events throughout a quiet day.

This kind of practice doesn’t build rapidly. It also doesn’t work if attempted during a flood, which would be trying to develop a new attentional capacity at exactly the moment attentional resources are most depleted. The practice belongs in the quiet moments, done consistently over weeks and months.

The aim is earlier detection rather than perfect interoceptive accuracy. Enough to shift from “I thought I was fine” to “something was already shifting and I can see it now.” That is a different relationship with the body’s data, and it can be built.

An experiment for this week

Once today, in a quiet moment with no emotional charge, try this: sit for sixty seconds and put attention on the body rather than the mind. No breath work, no meditation. Just notice what is present. Tension in the jaw, weight in the chest, pace of breathing. Whatever is actually there.

Then at some point this week, before something you expect to be slightly difficult (a conversation, a meeting, a commute) do the same thing for thirty seconds beforehand. Name, silently, whatever you find. Tense. Settled. Already narrowing. Whatever is accurate.

The aim is to practise the noticing before the state is acute. Detection, before anything needs to be done with it.

I’m curious what the sixty seconds shows people who haven’t deliberately attended to their body in a quiet moment before. Whether there’s a pattern in which flooding type finds the most already happening. My suspicion is that the Anticipator and the Reactor find quite different things.

The CARE framework’s Context step is built for exactly this gap. It is the first question in the sequence for a reason: what state am I actually in right now, before the story forms? Practised in low-arousal moments rather than crisis ones, it trains the body-first check that interoceptive deficit makes unreliable. The full tool (type quiz, guided Context-Aims-Risks-Experiments sequence, breathing pacer) is at here.

The Context Shift is psychoeducational writing. Nothing in it, or in my replies, replaces your own clinical care.

Next week I want to look at what happens to the nervous system when ADHD adults spend extended periods performing regulation they are not actually achieving, and what the debt that accumulates looks like.

Darren came in for his six-month follow-up looking tired in a way that was different from before.

The medication had helped. He said so plainly. Work was easier to navigate. The school run had stopped feeling like a wall to scale before 8am. By the measures we use in clinic, things were going well.

Then he said something I have heard in different forms across twenty years of practice.

“I tried the mindfulness. During an argument with my wife last month. I thought — I know what to do here. I sat on the stairs, focused on my breath, watched my thoughts go past. And it made everything worse. I got more wound up, not less. I ended up walking out of the house.”

A pause.

“I think I’m just bad at it.”

He was trying the right technique at the wrong moment.

The problem was the prescription.

Mindfulness, as most people encounter it, is a top-down practice.

That phrase needs unpacking. When the brain is operating in a top-down state, the prefrontal cortex is running the show. Attention is deliberate. You can choose where to direct it, notice when it wanders, and bring it back. Metacognition is available — the capacity to stand slightly outside your own thought stream and observe it. These are precisely the skills mindfulness asks you to use. Focus on the breath. Notice the mind moving. Return, without judgment.

All of that requires a functioning relationship between intention and attention, and that relationship runs through the prefrontal cortex.

In ADHD, that relationship is less reliable even at baseline. The prefrontal systems involved in voluntary attentional control are the same ones ADHD disrupts. The research is consistent on this: ADHD is a disorder of doing what you know, in real time, under conditions that require it. Access to deliberate attention is possible, but less dependable, more effortful, and more easily lost under pressure. I wrote last January about why emotional dysregulation is often more disabling than inattention in ADHD — this is part of why. (thecontextshift.substack.com/p/why-adhd-emotional-dysregulation)

During emotional flooding, the picture shifts considerably.

When the nervous system detects threat or overwhelm, processing changes state. Subcortical structures take precedence. Attention narrows. Reactivity accelerates. Prefrontal authority drops. The shift happens in every nervous system under sufficient pressure. In ADHD, it happens faster, tends to go further, and the window back to baseline is longer.

Here is the mismatch in one sentence: the ADHD brain during emotional flooding is operating in a bottom-up state — driven by subcortical pressure, not prefrontal direction. Mindfulness is a top-down tool. It requires the prefrontal capacity that the state has already reduced.

Darren, sitting on the stairs mid-argument, was in that state. And he asked himself to execute a top-down protocol.

Watch your thoughts. Return to the breath. Observe without judgment.

These are instructions that require the very capacity the state had already removed.

The failure Darren described — doing everything the instruction said, and feeling worse — carries a particular cost beyond the immediate distress.

When mindfulness fails during a flood, it doesn’t read as a timing problem. It reads as a personal one. He had tried the thing he’d been told would help. He’d failed at it. He couldn’t even do mindfulness correctly.

That interpretation lands in a place that’s already full of similar material. Adults with ADHD arrive at middle life with a long inventory of things they found unmanageable that other people appeared to handle without much effort. When mindfulness joins that list, it confirms something they’ve quietly suspected for years.

The confirmation is inaccurate. But the experience that generated it is real, and it makes sense that it lands the way it does.

Here is what I want to be precise about, because the precision changes what you do next.

Mindfulness has genuine evidence behind it for ADHD — in the right context. Regular practice in low-arousal baseline states, sustained over weeks and months, does appear to build something. The repeated experience of noticing the mind has wandered and returning it, done consistently in a calm state, seems to develop attentional flexibility over time. That is the training version of the practice.

The acute version — mindfulness deployed in the middle of a flood as a real-time regulation strategy — is a different prescription pointing at the same technique. And the state you’re in when you try it changes everything.

The more useful question about any regulation approach is what the tool actually requires from the person using it, and whether their current state makes that available.

Mindfulness requires deliberate attentional control. A flooded state has reduced that capacity. Reaching for a top-down tool at the moment top-down access has gone offline is not a failure of commitment. The match between prescription and state is what’s off.

Bottom-up approaches work differently. They don’t need the prefrontal cortex to direct the process. Movement. Cold water on the hands or face. Leaving the room. Physical engagement that gives the arousal somewhere to go. These work by acting on the nervous system through channels that don’t depend on executive function.

Once the system returns to something closer to baseline — and the body usually signals this: the chest loosens, the pace of thought slows, reactivity drops a few degrees — top-down tools become accessible again.

Mindfulness has its place in that sequence. It’s just not at the beginning of it, inside a flood.

Darren came back three weeks later.

He’d tried a different sequence during an argument. He’d walked to the kitchen, drunk a glass of cold water standing at the sink, and returned when his chest had stopped feeling tight.

“It’s not dramatic,” he said. “But I didn’t walk out.”

That’s usually how it goes. The change is quieter than people expect.

This distinction — state first, then tool — is what the CARE framework is built around. Context: what state am I actually in? Aims: what is realistic from this state? Risks: am I reaching for something my current state can’t run? Experiments: which approaches match where I am right now, not where I’d prefer to be?

The CARE framework works through this sequence — if you want to explore it, the full tool is right here. It starts by identifying flooding type, because which approach fits first depends on how someone arrives at dysregulation. Different entry points, different first moves.

Here is a specific thing to try this week.

Before reaching for any regulation strategy — mindfulness, breathing, grounding, whatever is in your current toolkit — name the state first. Flooded? Wound up and narrowing? Flat and closed down? One of those.

Then ask one question: does this tool require me to have a regulated brain to run it?

If the answer is yes, it needs to wait, or be replaced by something that doesn’t carry that requirement. Movement first. Physical discharge of what’s already built up. Then, once the system has settled, the top-down work.

Choosing the tool for the state you’re actually in — rather than the one you’d prefer to be in — is where most regulation work starts to hold.

Next week I want to look at why ADHD adults so often don’t notice the flood starting until it’s already arrived — which turns out to be a different problem with a different name.

The Context Shift is psychoeducational writing. Nothing in it, or in my replies, replaces your own clinical care.

Andrew came back to clinic six weeks after his initial assessment. He’d been reading the newsletter. He knew, he said, that he was a Ruminator: the argument that ends at 10pm but continues internally until 3am, the emotional backlog that doesn’t clear just because the room goes quiet.

He wasn’t confused about what had happened. That was new. What he didn’t know was what to do when it started happening again.

He’d flooded on a Wednesday evening. He’d watched it happen with something close to clinical detachment, noting the physical tightening, the point where his wife’s words stopped landing as information and started landing as evidence. He knew the pattern. He was inside it anyway.

“Does knowing help?” he asked. It’s a genuine question, and the honest answer is: sometimes, and not in the way most people expect.

Why recognition isn’t enough

Recognition does something real.

When you can name what’s happening, when “I’m flooding” replaces the automatic sense that reality has simply turned hostile, you’ve introduced a small gap between the state and the story the state generates. That gap is where most of the useful work eventually happens.

But recognition operates at the level of the story-after. You piece it together in the morning, or on the drive to work, or in a clinic room six weeks later. The flood itself doesn’t announce itself in language. It announces itself in the narrowing of the world, in the felt certainty that the other person is doing something to you, in the sudden inaccessibility of every piece of relational knowledge you’ve ever accumulated.

The prefrontal cortex (the part that holds your good intentions, your repair strategies, your understanding of your own patterns) is among the first casualties of emotional flooding. The knowledge that should help you is precisely what you can’t reach when you need it.

The five types of ADHD emotional flooding

There are five types of ADHD emotional flooding. Over the last five weeks I’ve introduced each one separately. Here, for the first time, the full picture.

The Hidden Flooder floods without knowing it. The emotional state is present and influencing behaviour, but it isn’t legible as flooding. The body knows before the mind catches up, if it catches up at all.

The Reactor floods immediately and visibly. The response is disproportionate by external standards, entirely proportionate by internal ones. It’s the type most likely to be labelled a temper problem.

The Shutter doesn’t expel. The system closes. From outside it looks like stonewalling. From inside it’s the only available form of self-protection when further engagement feels like danger.

The Ruminator carries emotional backlog. The argument ended at 10pm, but the internal processing continued until 3am, and Wednesday’s flooding is partly about something that happened in February.

The Anticipator floods before the event. The timeline diverges. The ADHD brain is already living the version where things go badly while everyone else is still in the present.

Most people recognise more than one type in themselves. Some recognise all five at different times or in different relationships. The taxonomy isn’t a diagnosis. It’s a map, and maps are only useful if you know where you are on one.

From recognition to response

What Andrew needed wasn’t a better map. He needed a tool he could actually access in the 90 seconds before he said something that would take three days to repair.

This is where most emotional regulation advice breaks down for ADHD brains. The strategies are taught outside the flood and assumed accessible inside it. Flooding forecloses the cognitive functions those strategies rely on. The advice becomes unavailable precisely when it’s needed.

What works differently is a structure rehearsed so thoroughly it doesn’t require retrieval under pressure. Something that runs on procedural memory rather than working memory. Something that doesn’t ask you to think clearly in order to start it.

The CARE framework (Context, Aims, Risks, Experiments) is built for this.

Context is a body-first check: what state am I in right now? The state, before the story. Flooding generates the interpretation. State precedes story, always. Starting with narrative skips the step that matters.

Aims identifies what you actually want from this moment, underneath what you’re about to do. Usually those two things are in direct conflict when flooding is active.

Risks maps the likely cost of acting from the current state. Brief, honest, accurate.

Experiments names one thing to try. One small test.

The sequence is the point. Each step narrows the space in which the flood operates. Done enough times outside the moment, it becomes something closer to reflex inside it.

The gap between recognising a pattern and being able to interrupt it is where most ADHD adults quietly live.

If you’ve followed the last five weeks, you now have the map. Most ADHD adults spend years without one. If someone in your life would recognise themselves in any of these five types, this piece is worth sending them.

Subscribe now

If the five flooding types gave you the map, The CARE Framework in Practice is the next step.
It is a guided tool for the moment after recognition: take the type quiz, work through Context, Aims, Risks and Experiments in real time, and use the breathing pacer to support the physiological regulation underneath the cognitive work.

It is available here for £19.

The Context Shift is psychoeducational writing. Nothing in it, or in my replies, replaces your own clinical care.

Claire didn’t come to her appointment with a story about what had happened. She came with a story about what hadn’t.

“We were supposed to go to his parents for dinner on Saturday,” she said. “Nothing had gone wrong. He hadn’t said anything. I’d been fine all week.” She paused. “I drove home from work on Friday and just started running through it in my head. By the time I got through the front door I was done. I told him I couldn’t go. He didn’t understand why. I couldn’t explain it.”

She wasn’t describing avoidance. She wasn’t describing anticipatory anxiety in the clinical sense — no generalised dread, no catastrophising about the dinner itself. She was describing something more specific: the complete exhaustion of emotional capacity before the event had arrived. The flood had happened in the car. By the time she walked in, there was nothing left.

I’ve seen this pattern often enough to recognise it as distinct. The Reactor floods in the moment and recovers quickly. The Shutter goes offline under the pressure of the event itself. The Ruminator processes hours after. The Anticipator floods in the approach. The capacity is spent before the starting gun.

I’ve started calling it the Anticipator.

The mechanism

The ADHD nervous system is, among other things, a pattern-matching machine. It reads threat cues early — often before conscious awareness has caught up. For most people, that early-warning function is useful. It narrows the window slightly in the approach to something difficult, which is appropriate. It doesn’t close it.

For the Anticipator, the window doesn’t narrow. It shuts. The threat-matching system, shaped by years of flooding experiences and their aftermath — the damage done, the things said, the repairs that took weeks — has learned to treat the approach to hard conversations as equivalent to the hard conversations themselves. The body responds accordingly. Heart rate up, cognitive load increasing, emotional regulation capacity draining, all before anyone has said a word.

This is where it diverges from clinical anxiety. Anxiety generalises. It spreads across contexts, attaches itself to uncertainty broadly. The Anticipator’s flooding is situation-specific — it activates around relational threat in particular, around the moments where flooding has previously cost something. The nervous system is not overreacting to everything. It is over-reading the specific terrain where flooding has hurt before.

That distinction matters for what helps. Generalised anxiety responds to anxiety treatment. The Anticipator responds to something different: understanding the mechanism, and changing the relationship to the approach itself.

What it looks like from outside

Claire’s partner saw her withdraw on Friday evening. He hadn’t done anything. The dinner plans were still in place. From his perspective, something had shifted without cause — she’d come home quiet, gone to bed early, and in the morning told him she couldn’t go.

This is one of the hardest features of the Anticipator for partners to make sense of. The flooding is invisible because it happened somewhere else — in the car, in the work corridor, in the ten minutes before the difficult phone call. By the time the partner registers that something is wrong, the process is already over. They don’t see someone flooded. They see someone already depleted.

The most common interpretation is avoidance, or coldness, or not caring enough to push through. All of those land painfully. None of them are accurate. The Anticipator is often the person who cares most about the relationship — which is precisely why the pattern-matching fires so early and so hard.

And the difficulty for Claire — and for most Anticipators I see in clinic — is that she knows this about herself and still cannot always interrupt it. Knowing the mechanism doesn’t automatically restore capacity. The window has already closed by the time the knowledge is available.

Subscribe now

The asymmetry no-one names

There’s an asymmetry at the heart of the Anticipator pattern that tends to stay invisible until someone says it plainly.

The partner experiences the withdrawal as the event. Claire didn’t come, the evening was cancelled, the relationship sustained some damage. From the partner’s timeline, that’s where the difficulty happened — Saturday, the absent seat at the table, the explanation that never quite landed.

For Claire, the event happened on Friday at quarter to six, on the A414, in traffic. The flood, the narrowing, the depletion — all of it was complete before she reached the front door. The Saturday cancellation was not a decision born of that moment. It was the consequence of something that was already over.

Two people with entirely different maps of when the hard thing happened. Both maps are accurate. Neither is sufficient on its own to understand what’s going on.

This is the territory where the Anticipator pattern does its most lasting damage — not in the event itself, but in the divergent timelines. The partner is trying to understand something that, from their perspective, happened on Saturday. Claire is trying to explain something that happened the day before. Without a shared frame, those conversations don’t resolve. They accumulate.

One experiment

I don’t offer this as a fix. The Anticipator pattern is built on years of nervous system learning, and a single insight doesn’t undo that. But there is one thing I’ve found useful to suggest, and it’s small enough to be genuinely testable.

Name the narrowing when it starts.

For Claire, that might have sounded like: “I’m running the scenarios.” Said out loud, to her partner, before she’d reached depletion. Not as a warning, not as an explanation of the cancellation to come — just as a signal that the approach had started, that her window was already narrowing, that what was happening in her nervous system was visible and nameable.

This does two things. It makes the process slightly less invisible to the partner, which changes the map problem. And it gives Claire a moment of external anchorage — saying the thing out loud interrupts the purely internal loop for long enough to sometimes change its course.

Sometimes. The Anticipator pattern doesn’t yield to willpower, and naming it won’t prevent every cancellation. But it offers a point of entry that comes earlier than most interventions — which is exactly where the Anticipator needs something to land.

The five types

This is the fifth pattern in this series. The Hidden Flooder, the Reactor, the Shutter, the Ruminator, the Anticipator. Five distinct presentations. Five different places where the flood begins, builds, and does its damage. Five different things that partners misread, often through no fault of their own.

What they share is a common problem: the gap between what emotional dysregulation looks like from outside and what is actually happening internally. That gap is where the misunderstanding lives. It is also where the work is.

I’ve been building something specifically for that gap — a structured protocol for working with emotional flooding in real time, mapped directly onto each of these five types. If you’ve recognised yourself in one of these patterns and want to understand what working with it actually looks like, this is where to start.

The CARE framework is available now right here.

Most people who recognise themselves in the Anticipator have been explaining themselves as anxious for years. The mechanisms overlap. They are not the same thing. In clinic, naming the difference tends to land as relief rather than correction — as though a more precise word changes the weight of it slightly. The trigger point is usually specific and consistent. The same stretch of road. The same moment in the working day. The same conversation that hasn’t started yet.

It is 2:47am and Jessica is not asleep.

The argument ended hours ago. She and her partner talked it through in the kitchen, agreed on something, said goodnight. He is asleep. She can hear him. She is lying in the dark running the conversation back, and what she is finding is not what happened at 10pm.

At 10pm, her partner said something about the plans for the weekend. She felt something shift when he said it, a small snag, a sensation of something catching, and then she managed it. Filed it. Continued the conversation. Said the reasonable thing. By the time they agreed to go to bed, the surface was intact.

Now it is 2:47am, and the surface is not the thing she is working with.

She is replaying his exact words. Finding new weight in them. The way he said it. Just the way. What that implies about how he sees her. Whether it connects to something he said three months ago. By 3:30am she has a fully formed position on what the argument really meant, what it reveals about their relationship, what she is going to have to think about.

By morning, her partner will make coffee and ask what she wants to do today. He will have no idea that any of this has happened.

This is the Ruminator.

When Jessica described this to me in clinic, she mentioned it as an aside, the way people name things they have stopped expecting anyone to take seriously. I recognised it immediately. The fourth of the five flooding types I have been describing in this series. Possibly the most privately exhausting of them all.

The Reactor erupts. The Shutter goes offline. Both leave a visible trace. The Ruminator leaves none. The difficult moment passes intact. The flood arrives afterwards, in private, in the dark.

What makes this neurologically distinct is what the replay is actually doing.

The ADHD nervous system does not always flood in real time. In some presentations, there is enough executive function available during the event to manage the surface, to track what is being said, to monitor tone, to produce a regulated response. The flood is deferred. The processing queue fills during the event and runs later, after the prefrontal cortex has stopped standing guard.

What Jessica likely missed during the argument, what most Ruminators miss, is the early accumulation. The interoception deficit that characterises ADHD means the body’s early warning signals are unreliable. The tightening she felt when her partner spoke was there. The load was building. But the signal was not loud enough, or clear enough, to register as something she could act on. The flood did not feel like it was coming. It just arrived.

The replay running at 3am is not running on a neutral system. It is running on a nervous system that is now flooded, activated, distressed, on alert, but no longer connected to the original event. The body is in distress. The brain, doing what it always does, generates a story to explain it. But the story is being written retrospectively, by a dysregulated system, without the moderating information that was present at 10pm: her partner’s face, his tone, the context of the room.

The 3am account of what happened is not the same account as the 10pm event. It is harder. More certain of injury. More certain of intent.

The CARE framework gives both accounts — the 10pm version and the 3am one — a structure to work from. The Context step is the first move: what is actually present right now, before the nervous system’s version of events is added.

https://care.thecontextshift.com/

I want to be careful here, because the easy version of this misses something important.

The 3am account is not simply wrong. Sometimes the managed surface at 10pm suppressed something real, a genuine hurt that did not yet have language, a pattern Jessica has noticed before and could not name in the moment. The replay occasionally surfaces something true.

From inside a flooded state, there is no reliable way to know which is which. The signal and the noise arrive together, dressed identically. Jessica cannot tell, at 3am, whether what she is feeling is a genuine perception or a dysregulated construction. She will go into the morning carrying both, unable to separate them.

This is the part that partners rarely understand, and that Ruminators rarely articulate. She genuinely cannot calibrate the account she is holding. She is holding two versions of the previous night and has no reliable instrument for telling them apart.

From the outside, the Ruminator’s pattern has a particular quality. The argument ends. Both people go to bed. And then, the next morning, or three days later, or in the middle of an otherwise unrelated conversation, something surfaces.

The partner’s interpretation is almost always the same: she cannot let things go. She agreed it was resolved. She said she was fine. And now here it is again.

The accusation of score-keeping comes from this. Of holding grudges. Of refusing to accept a resolution that was clearly mutual. None of this is how Jessica experiences it. She is not choosing to return to the argument. The argument never fully closed for her, because the processing that would have closed it ran hours after the conversation did, in conditions that did not allow for resolution.

What the partner reads as emotional score-keeping is, more accurately, emotional backlog.

Subscribe now

And the backlog accumulates.

This is what distinguishes the Ruminator from the other types in ways that matter over time. The Reactor’s damage is acute, visible, nameable, something both people can point to and address. The Shutter’s is confusing, a disappearance with an eventual return. The Ruminator’s damage is erosive. It does not happen in a single event. It happens across hundreds of small 3am experiences the partner never knew occurred, each one shaping, slightly, how Jessica reads him, their relationship, her own position in it.

There is no obvious rupture to point to. No outburst. No withdrawal. Just a slow drift in how the Ruminator experiences what is, to everyone else, a functioning relationship.

This is why the pattern is particularly hard to interrupt. It never becomes visible enough to address directly.

Jessica knows she replays things. She has been told it is exhausting. Not always by her partner. Sometimes by the part of herself that watches the 3am activity and cannot make it stop. She has tried. She has told herself the argument is over. She has left the room, picked up her phone, read until her eyes closed. The rumination does not respond to those instructions.

It does not respond because it is not a decision. The nervous system is processing on a delay, in the dark, and there is no off switch she can reach from the outside. She has ended up with a nervous system that processes emotional load late, in private, without adequate brakes, and a brain that has filled the gap with the best account it could build from available materials.

The account happens to be the one her partner never heard. The argument he thought ended at 10pm.

The Ruminator’s specific problem is that the replay generates conclusions, and those conclusions get carried into the next day as though they are facts.

The CARE Framework gives the gap a structure. A way of working with both accounts, the 10pm version and the 3am one, without the flooded system being the only thing deciding which one is true.If you recognise this pattern, one thing worth trying: the next time you catch yourself at 3am building the version of the argument your partner never heard, pause before the next conclusion, just long enough to ask one question: am I working from what was said, or from what my nervous system has decided it meant? You will not be able to answer it cleanly. That is the point. The inability to answer it is what the CARE Framework is designed to work with.

The guide is available here.

Next in the series: the Anticipator, the type who is already flooded before anything has happened.

The pattern I hear most often from Ruminators in clinic is some version of this: I know it was probably nothing. But I cannot make myself believe that at 3am.

If that is familiar, the 3 am construction, the partner’s misread, the accumulation over time, I would find it useful to know which part of this lands.

The presenting problem was the word “stonewalling.”

Their partner had been reading a relationship book — one of the Gottman ones, the kind you find on every couples therapist’s shelf — and had come to Jamie with the page open. Gottman calls stonewalling one of the Four Horsemen of relationship breakdown. He places it alongside contempt, criticism, and defensiveness. The person who shuts down, goes quiet, and becomes unreachable. The person who, when conflict arises, simply exits.

Jamie had read the page. Then read it again.

“I don’t think that’s what I’m doing,” they said. “But I can’t explain what I am doing.”

That inability to explain it — not defensiveness, not denial, just a genuine absence of language for what happens internally — is usually where the real work begins.

Gottman’s model is robust, and his research is serious. Stonewalling, in his framework, is contempt expressed through withdrawal. The person going quiet has, on some level, decided the conversation isn’t worth having, or that their partner isn’t worth engaging with. It is a posture. It carries intent. And in his longitudinal data, it is one of the strongest predictors of relationship breakdown — not because silence is inherently destructive, but because the contempt underneath it is.

That framing matters here. Because the behaviour Gottman describes and the behaviour Jamie was describing looked identical from the outside. Same result. Completely different mechanism.

When emotional input exceeds the ADHD brain’s processing capacity, the nervous system defaults to freeze — its third option after fight or flight. Unlike stonewalling, which involves a conscious decision to withdraw, ADHD shutdown is neurological: the brain goes offline before the person can identify what they’re feeling, let alone communicate it. The interoception deficit common in ADHD means the flood arrives before the feeling is even named.

The ADHD brain has three options when emotional input exceeds its processing capacity. Fight — the Reactor’s territory, loud and fast and over quickly. Flight — leaving the situation, physically or mentally, before the flood arrives. And freeze.

The Shutter freezes.

What happens in the freeze response isn’t suppression. It isn’t a decision to withhold. The emotional input reached a level that exceeded processing capacity before labelling was even possible — before the person could identify what they were feeling, let alone communicate it. The system hit a wall and shut the gates. The silence that follows isn’t the silence of someone who has decided not to engage. It is the silence of a system that has gone offline.

If this is landing — if you’ve been in that room, on either side of that silence — subscribe below. This is what The Context Shift is for.

Interoception is the brain’s capacity to read its own internal states — to notice, in real time, what’s happening physiologically and emotionally. In ADHD, this process is frequently impaired. Under ordinary conditions, that means a beat or two of delay in recognising what you’re feeling. Under emotional load, it can mean the flood arrived before the feeling was even named. They went quiet because they didn’t know what was happening inside them, not because they’d decided what to do about it.

The shutdown precedes awareness. That’s the part that’s almost impossible to explain to a partner who is watching it happen.

The CARE framework’s Context step was built for exactly this threshold problem — checking internal state before the system reaches the point where the only option left is to close down.

https://care.thecontextshift.com/

What tends to follow is entirely predictable and almost entirely counterproductive.

The partner sees the silence and reaches toward it. “What’s wrong?” Then again. Then with more urgency. Then frustration at the absence of response. Then — and this is usually where the evening tips into something that takes days to recover from — hurt or anger at the silence itself.

But every question is input. Every attempt to reach the person adds load to a system that shut down precisely because the load was too much.

The partner’s most loving instinct — not to let the person disappear, to stay present, to reach them — is what extends the shutdown. I find this genuinely difficult to explain without it sounding like blame, because it isn’t. The pursuit comes from care. And it still makes things worse. The only thing that helps is the thing that feels most like abandonment: space, quiet, the removal of input. Which, when you’re watching someone you love become unreachable, feels nearly impossible to offer.

There is an asymmetry in the Reactor that I’ve described before. The ADHD person explodes, then recovers quickly — sometimes within minutes — while the partner carries the impact for days. The emotional timelines don’t match, and neither person fully understands the other’s experience of the same event.

The Shutter asymmetry runs differently. And I think it is, quietly, more damaging.

The Shutter goes offline. Sometimes an hour. Sometimes the rest of the evening. Sometimes, the nervous system’s most complete form of load-shedding is that they fall asleep. And then they come back. Not gradually. Actually back. Ready to talk, ready to reconnect, the flood having subsided and the system having rebooted.

The partner has been in distress the entire time.

So the Shutter re-emerges saying “I’m ready to talk now” while the partner is at their own peak. And because the Shutter seems genuinely fine — because they’ve recovered fully and want to connect — the partner reaches for the only explanation that seems to fit.

“If you can be fine now, then you were choosing not to be there earlier.”

That misread does serious damage. It takes a neurological event — a system overload with a specific physiological mechanism — and converts it, retrospectively, into an act of contempt. It makes the shutdown look like a decision. And once that reading takes hold, it is very hard to dislodge, because the evidence keeps seeming to confirm it. The Shutter keeps recovering. The partner keeps watching them recover. Each recovery becomes further proof that the silence was chosen.

It wasn’t. That’s what Jamie needed language for. And it took the better part of three sessions to find it.

Naming the mechanism doesn’t fix it. I want to be clear about that, because there’s a version of this kind of piece that implies understanding is enough. It isn’t. But it is the necessary first step — for the Shutter to understand what’s actually happening in themselves, and for the partner to have something more accurate than contempt to work with.

That’s the gap the CARE Framework is designed to bridge — not just naming the mechanism, but giving people something structured to work with once they’ve recognised themselves in it. Something more grounded than insight alone and more honest than a list of tips. It walks you through the flooding moment itself, with tools built specifically for each type.

The CARE Framework is available. Have a look at the preview.

If you’ve been called a stonewaller — or if you’ve used that word about someone you love — I’d like to know what that label felt like to hear. Whether it landed as accurate, or whether something about it didn’t quite fit, and you couldn’t explain why. Leave a comment below.

That gap between the word and the experience is usually where the real work begins. If you’ve recognised yourself in any of these types — the Hidden Flooder, the Reactor, the Shutter — more soon.

Adam came to his six-month follow-up looking, by most external measures, like someone who had found his footing. The medication was working. His focus had improved. His productivity at work was better than it had been in years.

He sat down and said, “I’ve been told I have an anger problem. I don’t actually feel angry most of the time.”

I believed him on both counts.

What Adam was describing, without having a name for it, is something I’ve seen in clinic often enough to recognise as a distinct pattern. Not a diagnosis. Not a character type. A specific presentation of emotional dysregulation that looks quite different depending on which side of it you’re standing on.

I’ve started calling it the Reactor.

The gap nobody names

When I ask Reactors to describe what actually happens, they tend to describe two separate events.

The first is the outburst itself: a surge, a moment of too much volume or too much feeling or both, directed at someone who happened to be present when the ignition occurred. The Reactor usually knows, in some part of themselves, that the response was disproportionate. Some of them know it while it’s still happening.

The second event is twenty minutes after. The surge dissipates. The nervous system resets. The Reactor looks around at the aftermath and experiences, in quick succession, mortification, genuine remorse, and then something that functions like genuine confusion — because the internal experience of the emotion has moved on, and the external reality has not.

“I can’t hold grudges,” one patient told me. “I literally cannot sustain anger for very long. And yet somehow everyone in my life thinks I’m always angry.”

The people in a Reactor’s life are not wrong. But they are reporting a different experience of the same event. And the distance between those two reports is where most of the relational damage actually lives.

What the Reactor is

Adam is one of five patterns I see regularly — people with ADHD whose anger arrives faster than any conscious thought, peaks at full intensity, and is genuinely over within 15 to 20 minutes. What partners describe as an explosion. What Adam calls a loss of control he cannot explain or predict. The clinical literature calls this emotional dysregulation. Forums call it ADHD rage. Neither label quite explains what is actually happening neurologically. I call this pattern the Reactor — and it is one of five I have been mapping across clinical work and the published research on ADHD emotional dysregulation. The others, briefly: the Hidden Flooder (contains everything, presents as coping); the Shutter (withdraws so completely that the people around them don’t know whether something has gone wrong); the Ruminator (loops, replays, cannot put the thing down); the Anticipator (already dysregulated before the emotional event arrives, bracing for impact that may never come). Each pattern looks different. Each needs a different response. That distinction matters because most available advice on ADHD emotional dysregulation treats it as a single problem with a single solution, which is probably why so much of that advice fails so predictably.

The Reactor’s defining characteristic is fast ignition, full intensity, and rapid fade. This isn’t volatility in the colloquial sense, though it can look like it from the outside. It is a specific timing problem: the emotional response arrives before the regulatory system can properly assess what is happening.

Why does ADHD anger come on so fast — and disappear so quickly?

This is the question I hear most often from Reactors, and from their partners. The clinical literature refers to this pattern as emotional dysregulation — what gets described in forums and online communities as ADHD rage. Both labels are pointing at the same mechanism, though neither quite explains it.

In standard emotional processing, there is a brief delay between feeling and response. Long enough to do some regulatory work: assess the situation, consider the context, calibrate the intensity of the response to the actual severity of the event. This delay is largely a prefrontal cortex function.

In ADHD, the prefrontal function is impaired. Not absent, but slower to engage. The emotional signal arrives at full intensity before the context does. By the time the rational processing has caught up with what’s happening, the response has already occurred.

In the ADHD brain, the emotional signal arrives before the regulatory system can assess it. The result is a full-intensity response with no buffering. The same executive function deficit that affects attention and task completion also affects emotional inhibition. There is no defect in the emotional experience itself. The Reactor feels everything the emotion contains. What is impaired is the pause between feeling and response. That pause is where most people do their regulatory work without noticing it’s even happening.

The CARE framework maps what can happen in the space before the Reactor pattern arrives. The Context step builds the body-first awareness that interoceptive delay removes.

https://care.thecontextshift.com/

Why does ADHD anger fade so quickly?

The Reactor’s outburst typically fades within 15 to 20 minutes — not through any act of self-regulation, but because the prefrontal cortex has finally caught up with an event it was too slow to intercept. The emotional signal arrived at full intensity before the regulatory system could engage. By the time the prefrontal cortex reaches the scene, there is nothing left to regulate. The surge is already over. This is why the brevity of the outburst is not evidence of shallow feeling. It is evidence of a different neurological clock — one where the regulatory system runs roughly a quarter of an hour behind the emotional event. The person who exploded is not performing recovery. Their nervous system has genuinely reset. The person who received the outburst has not.

The shame loop

What happens after the surge is, in many ways, the more clinically significant event.

The Reactor’s outburst ends. The nervous system resets. And then the shame arrives, often at a volume comparable to the original episode. The apology tends to be extensive. Sometimes excessive. The Reactor may explain, re-explain, return to the subject hours later, and check repeatedly whether the other person is alright.

This is not, as it might appear from the outside, remorse modulating into something more calculated. It is a second neurological event: the shame loop, running on the same overloaded system that produced the outburst, trying to complete a regulatory circuit that got cut short.

But here is what the Reactor rarely understands about that loop: it often makes things worse for the person who received the outburst. To be apologised to extensively, repeatedly, and at length can feel like being put in the position of consoling the person who hurt you. The apology becomes its own demand. The person who was on the receiving end of the surge now finds themselves managing the Reactor’s shame.

And it doesn’t resolve the underlying asymmetry.

What partners are actually living with

Adam’s partner, by his account, walked carefully around certain topics. She would pre-emptively soften messages before delivering them. She had told him, more than once, that she never knew which version of him she was going to get.

He found this baffling. From where he sat, his mood was reasonably consistent.

I hear this from couples regularly. The Reactor’s nervous system has a hair-trigger threshold and an efficient reset. That combination means the Reactor genuinely experiences themselves as someone who moves through emotional events quickly, doesn’t hold on to things, and doesn’t bear grudges. And they’re right about that.

But the people around a Reactor don’t have the same reset. They carry the last outburst into the next ordinary moment. They begin to monitor for early warning signs, to manage their own responses in ways that might reduce the chance of triggering a reaction. Over time, the anticipation of the outburst becomes its own weight, separate from and sometimes heavier than the outburst itself.

Walking on eggshells is not a response to the last explosion. It is a permanent background state, calibrated around the possibility of the next one.

The Reactor forgets. The partner doesn’t. Both are psychologically explicable. Neither is wrong about their own experience. This is not a problem of bad faith on either side. It is a structural mismatch, and it needs understanding before it can be addressed.

What actually changes this

I want to be direct here, because the conventional shape of this section involves a list of strategies and an implicit reassurance that change is available to anyone who wants it enough.

The Reactor cannot will themselves to a longer pause. The nervous system cannot be reasoned with during ignition. That is not a failure of commitment or self-awareness. It is a structural feature of the impairment, and framing it as a character problem is one of the things that keeps Reactors stuck in the shame loop rather than developing any actual capacity to work with the pattern.

What can change is what happens before and after the surge, not within it.

Before the outburst: reducing the frequency of triggers matters more than managing them in the moment. Sleep, cognitive load, sensory environment, and the quality of communication in close relationships about the pattern rather than the individual incidents — these are slow-moving variables, but they are moveable ones. The Reactor who understands their own threshold is better positioned than the Reactor who is perpetually surprised by it.

After the outburst: the twenty minutes following the surge are far more clinically accessible than the two seconds during it. What the Reactor does in that window, how they navigate the shame loop, whether they can resist the pressure to over-explain, and how they and the people close to them talk about the pattern rather than the individual episode — this is where most of the practical work actually lives.

I’ve built a structured framework for exactly this territory — not a cure for the ignition, but a systematic way of mapping what happens before and after it, and working with the pattern rather than being ambushed by it. It is called the CARE Framework, and it is available now right here.

A note on names

Adam asked me, near the end of the session, whether there was a name for what he’d been describing.

There is now.

The Reactor isn’t a fixed identity. It isn’t a diagnostic category. It is a pattern — one that makes a different kind of sense once the neurology behind it becomes visible, and one that looks quite different depending on which side of it you’re standing on.

If you’ve spent time on the Reactor side of this, this piece is for you. If you’ve been the one waiting for them to register what happened, it’s for you too. Most people who recognise this pattern immediately think of one other person who needs to read it.

Naming the pattern doesn’t close the gap. But it does mean both people are finally talking about the same thing.

The Hidden Flooder — the pattern that contains everything and presents as coping — was the subject of last week’s piece. The Shutter, the Ruminator, and the Anticipator follow in the coming weeks.

The part of this that tends to land differently depending on which side of it you’re on is the shame loop — whether you’re the one caught in it, or the one being apologised to long after you’ve moved on. I’m interested in which version of this is more familiar.

A patient I’ll call Daniel came to his follow-up appointment looking, by most measures, well. Eighteen months since diagnosis. Medication working. Job stable. Relationship intact. He’d recently been promoted, which he mentioned in the same tone as everything else — measured, considered, factual.

About twenty minutes in, he said he hadn’t been sleeping properly. When I asked how long, he paused. “A while,” he said. “Maybe six months.”

He’d been meaning to bring it up.

What Daniel was describing, in that appointment, was a man who managed — in every observable sense. What he wasn’t describing, and had never been asked about directly, was what it cost to manage like that.

In twenty years of ADHD assessments, the presentation that has most consistently been missed by clinicians, by partners, sometimes by the person themselves, is not the one that looks chaotic. It’s the one that looks fine.

I’ve started calling this the Hidden Flooder.

Not because the flooding isn’t happening. It is. The ADHD nervous system generates emotional floods at the same rate regardless of what appears on the surface. What differs in the Hidden Flooder is where the flood goes. The external presentation is contained. The internal experience is not.

This distinction matters more than it sounds, because most of the systems designed to identify when someone is struggling — clinical assessment, partner observation, self-report — are calibrated to visible distress. When there’s no visible distress, there’s no identified need. The Hidden Flooder is, in my experience, the hardest presentation to support. Not because they’re less affected. Because the presentation has been trained, over the years, to suppress exactly the signals that would prompt intervention.

There’s usually a history behind it.

For some Hidden Flooders, it’s an early environment where emotional expression wasn’t safe or welcome. For others, it’s the particular combination of high intelligence and ADHD that produces someone who has compensated so thoroughly that neither they nor anyone around them has noticed the effort involved. Sometimes it’s a professional context — medicine, law, high-stakes corporate work — where the performance of regulation is a job requirement, and the person has simply never turned it off.

What all of these have in common is a long training in not surfacing the signal.

The problem with this is neurological. The emotional flood doesn’t resolve because it isn’t expressed. It routes elsewhere — into chronic exhaustion, into late-night rumination that seems disconnected from anything specific, into a kind of low-level anxiety that gets attributed to other things.

Daniel’s six months of poor sleep wasn’t a sleep problem. It was a flood problem. The events that triggered it had, in each case, resolved on the surface within hours. Internally, they were still running.

What partners of Hidden Flooders often describe is not a calm partner. It’s a partner who is, in a specific way, unreachable.

Present physically. Contained in conversation. And then, several hours after what seemed like a fine evening, I suddenly and inexplicably found myself somewhere else entirely. Or gone quiet in the kitchen at 11 pm over something that apparently happened at work on Tuesday.

The flood doesn’t disappear at home. It surfaces there because home is where the suppression can finally relax. The consequence of this, relationally, is that partners often experience a level of emotional intensity that the ADHD adult’s colleagues or managers would find unrecognisable. This creates its own problem: the ADHD adult is told, in one context, that they are calm and functional, and then experiences themselves as dysregulated and overreactive in the context that matters most to them. Neither version feels accurate. Both are.

Of the five flooding types I see in clinic, the Hidden Flooder has the most difficult relationship with self-recognition.

The Reactor knows they react. The Ruminator knows they ruminate. The Hidden Flooder has, in many cases, spent decades being told they cope well — and has internalised that verdict to the point where their own distress has become difficult to locate. Not because it isn’t there. Because the system that was trained to suppress the external signal has also suppressed the internal one.

This is worth sitting with for a moment. The Hidden Flooder isn’t concealing distress from other people while being fully aware of it themselves. In many cases, the concealment is total. The question “how are you actually feeling?” produces a genuine blank, not a diplomatic deflection.

Daniel told me, towards the end of that appointment, that he’d assumed the sleeping problems were a medication issue. That he’d been waiting until the next review to mention it. That it hadn’t occurred to him that it might be connected to anything emotional, because he didn’t feel particularly emotional about anything.

I hear this often.

The practical challenge for the Hidden Flooder is a particular one: the first stage of any useful emotional regulation work involves building awareness of the flood — noticing it as it arrives, before it has fully organised itself into a response or a shutdown. But the Hidden Flooder’s entire system has been trained not to surface that signal. The place where other flooding types have discomfort and noise, they have quiet.

This is why generic emotional regulation techniques tend to underperform for this type. The techniques assume a signal to work with. The Hidden Flooder has, over the years, learned to route around the signal.

What I’ve found more useful, clinically, is starting not with the emotional experience but with the physical one. Where is it held in the body, before there is any feeling to name? What happens to breathing, to the jaw, to the chest, in the moments after a difficult conversation that seemed to pass without incident? Not a technique for processing the emotion. A direction.

It’s a slow starting point. But for the Hidden Flooder, it’s usually the right one.

The CARE framework was built from this pattern outward. If the Hidden Flooder is familiar, the Context step is the place to start — checking what is internally present before the external management begins.

https://care.thecontextshift.com/

If this sounds familiar — if you’ve been told, more than once, that you cope well, and found yourself not quite believing it — I’d be interested in what that has actually looked like. The comments are worth using for this one.

I’ve been mapping this within the CARE Framework — the emotional regulation approach I introduced in the Resolution Protocol piece last week — building out what it looks like specifically for the Hidden Flooder, and for each of the other four types. A practical guide to identifying your flooding type and understanding what the regulation work actually looks like from that starting point, rather than from a generic model that assumes the flood will announce itself clearly. The CARE framework is live. The full tool — flooding type quiz, guided Context-Aims-Risks-Experiments sequence, breathing pacer — is at https://care.thecontextshift.com/

For now, if any of this sounds familiar, the one experiment worth trying is not an emotional one. It’s a physical one. Notice where you hold things that have apparently resolved. Not with the intention of doing anything with it yet. Just with the intention of knowing it’s there.

Knowing it’s there is further along than most Hidden Flooders have been told they need to get.

The Hidden Flooder is one of five flooding types described in The Context Shift. The others — the Reactor, the Shutter, the Ruminator, and the Anticipator — are being explored in the coming weeks.

The first three months had been, by his account, transformative. Clarity after decades of static. He’d restructured his working day, started projects he’d been circling for years. Then, somewhere around month four, he began noticing what the medication wasn’t fixing. The emotional reactivity was still there. So was the procrastination on tasks he found genuinely uninteresting. By the time I saw him, he’d moved to the opposite position entirely. The medication changed nothing, he said. It was all placebo. He was thinking of stopping.

I’ve had versions of this conversation dozens of times. I’ve had a quieter version of it with myself.

When we ran the protocol together, what he found in Step 2 changed the conversation completely. The “medication fixes everything” position was protecting hope. After years of struggling without explanation, the diagnosis carried enormous emotional weight. To acknowledge its limits felt, at some level, like returning to the version of himself that had no explanation and no help. The “medication fixes nothing” position was protecting him from that hope. Better to have no expectation than to keep measuring and finding the gap.

Both positions were reasonable. Neither was accurate. And neither could be challenged directly without first understanding what it was doing. What he needed wasn’t a different view on his medication. He needed a more accurate map that made both extremes unnecessary.

Most ADHD advice tells you to pick a side.

Be more structured. Be more flexible. Commit to the system. Stop over-planning. Start earlier. Stop catastrophising about starting. Every instruction assumes one extreme is the problem and the opposite extreme is the answer. None of them asks what both extremes are protecting, or why, if one extreme were really the answer, the other keeps coming back.

The oscillation isn’t random. (I’ve been calling this the Pendulum Principle — if this is your first time here, the two earlier pieces in this sequence are linked below.) Both ends of the swing are doing something. And until you understand what each end is protecting, you can’t find the position that neither extreme can reach on its own.

I’m calling that position the Resolution Protocol. Four steps. No colour-coded planners. No commitment to permanence. Just a structured way through the middle of a pattern you’ve probably been watching in yourself for years. This piece is the method.

Step 1: Name the swing precisely.

Not in general terms. Specifically.

Not “I struggle with structure” but “I oscillate between no system whatsoever and a colour-coded planner I abandon within four days.” Not “I have a complicated relationship with my medication” but “I swing between believing it fixes everything and believing it fixes nothing, depending on what’s gone wrong that week.”

The precision matters for two reasons. Vague descriptions of patterns don’t give you anything to work with. And the act of naming something specifically tends to reduce its charge slightly, which is useful before you try to do anything with it.

Write down the two extremes. Both of them. Don’t resolve them yet.

Step 2: Find what each extreme is protecting.

This is the step most ADHD frameworks skip entirely, and it’s where most of the actual information lives.

The impulsive end of any oscillation is not chaos for its own sake. It’s protecting something real: spontaneity, reduced cognitive load, the freedom to respond to what’s actually happening rather than what was planned. The rigid overcorrection is protecting something real too: safety, predictability, proof of competence to self and others. Neither position is the enemy. Both are protective responses that have overshot the thing they were trying to protect.

Take each extreme you named in Step 1. Ask what it would lose if it were wrong. The answer is usually closer to the real issue than anything the surface position contains.

Once you can see what both sides are protecting, you have something to work with. You’re no longer trying to defeat one end of the oscillation. You’re looking for the position that keeps what both ends were trying to preserve, without the costs of either.

Step 3: Ask the bridge question.

This comes from the strategist Roger Martin, and I’ve adapted it because it’s the most useful single question I’ve found for getting out of a binary.

The question is: what would have to be true for both of these needs to be met, without the costs of either extreme?

Not: which side is right? Not: how do I compromise between them? Compromise leaves both sides partially unsatisfied. Integration addresses what both sides were actually trying to do.

The bridge question doesn’t always produce an answer immediately. Sometimes it produces the right next question. That’s still progress. It moves you from defending a position to genuinely asking what a better position would need to contain.

Step 4: Test it for 48 hours.

Not permanently. Not as a commitment. Forty-eight hours, explicitly framed as an experiment, with permission built in to return to the drawing board.

The reason this step is 48 hours and not “try it and see” is specific to how the ADHD brain handles new approaches. Permanent commitment to a new system triggers exactly the kind of avoidance the system is meant to address. The brain that has watched itself abandon colour-coded planners will not warmly embrace a new approach framed as the final answer.

A 48-hour experiment is low enough stakes to begin. And the beginning is where most of the difficulty actually lives.

Run the integrated position for two days. Note what it does and doesn’t address. Adjust. Run it again. The protocol is iterative by design, because a single attempt at integration rarely captures everything the first time, and because the ADHD brain learns better through short repeated experiments than through sustained commitment to a fixed approach.

A worked example: the email that’s been sitting there for eleven days.

You know the one. It arrived on a Tuesday. It requires a considered response — not a quick reply, something that needs thought. You flagged it. You’d get to it when you had a clear head. By Friday, it felt slightly overdue. By the following week, it had acquired a weight entirely disproportionate to its content, and now every time you open your inbox, it sits there radiating low-grade dread.

This is not a time management problem. It’s a Pendulum problem.

Step 1: Name the swing precisely. One extreme: the hyperfocused burst. When conditions are right — deadline pressure, high interest, external accountability — you can produce something excellent at speed. The other extreme: complete non-initiation. The task sits untouched for days while you do seventeen other things, none of which are the email. The swing is between “I work brilliantly under pressure” and “I cannot start anything without immediate consequences.”

Step 2: Find what each extreme is protecting. The hyper-focused burst is protecting quality. The standard you hold for your own output is high, and the burst state is the only condition where you feel confident meeting it. The non-initiation is protecting you from starting at the wrong moment and producing something below that standard. Both positions are serving the same underlying value — doing the thing well — and both are making the thing harder to do.

The eleven-day email isn’t evidence of laziness. It’s evidence of a quality standard that has become a barrier to beginning. That’s a meaningfully different problem.

Step 3: Ask the bridge question. What would have to be true for the email to get written without requiring either a crisis or a perfect window of clarity?

For most people, the answer involves separating the starting condition from the finishing condition. The brain that can’t initiate the email can often initiate opening a blank document and writing one imperfect sentence. Not the email. A sentence about what the email needs to say. Write it badly. On purpose.

The burst capacity is still available for the edit. You’re just no longer requiring it at the threshold.

Step 4: Test it for 48 hours. One intentionally imperfect sentence as the entry point. Apply it to this email, and one other sitting in the same category.

The dread lives at the threshold, not inside the task. Most people discover that the email they’d been avoiding for eleven days takes fourteen minutes once they’re past the first sentence. That’s not a character revelation. It’s the protocol working as designed.

A few things are worth naming clearly before you try this.

Step 2 can surface things that don’t resolve with a framework. Patterns rooted in earlier experience — relational, developmental, traumatic — respond to a different kind of work. If you find yourself there, the four steps have still done something useful. They’ve shown you where the edge is. What’s on the other side of it isn’t a protocol problem.

The protocol addresses the oscillation pattern specifically. It won’t fix the underlying executive function impairment, and nothing in a newsletter should claim otherwise. It’s not a replacement for medication or other structural supports — it works alongside them.

And it’s not a one-time process. The Pendulum doesn’t stop swinging. What changes is the arc. With practice, the extremes become less extreme, the recognition comes earlier, and the integration point becomes easier to find. That’s not a cure. It’s a method where previously there was only a pattern.

The protocol applied to specific domains — the medication relationship, the work pattern, emotional regulation — goes considerably deeper than a single piece can hold. That’s territory I’ll be returning to.

There’s a version of the post-diagnosis story that goes: I got the diagnosis, I understood the pattern, I found the method, things improved.

That version is true for some people and partial for most. The more honest version tends to go: I got the diagnosis, I understood the pattern, I assumed understanding would be enough, and then I spent some time in the gap between insight and change.

The gap had a method in it.

You now have the method. What you do with it is the actual work, and that part is yours.

The CARE framework was built for exactly this gap. Context, Aims, Risks, Experiments — a structured sequence for working with the pattern rather than relying on insight to interrupt it alone.

https://care.thecontextshift.com/

If someone in your life has watched you repeat a pattern you’ve clearly understood for months, this piece is worth sending to them. Not as an explanation — as a starting point for a different conversation.

The oscillation worth starting with is usually the one you already named without being prompted. It’s probably in your inbox, or in a conversation you keep meaning to have, or in the gap between what you planned this week and what actually happened. Tell me in the comments — your answer will almost certainly help someone else who hasn’t quite named their swing yet.

A note: this is perspective, not prescription. For personal medical advice, please speak to your own clinician.

Alex came in looking better than I’d seen them in two years. Medication working, sleeping properly, holding their own in meetings for the first time they could remember. They’d started reading about ADHD — really reading, the kind of sustained engagement that would have been impossible six months ago — and they had the vocabulary for their own experience now. They knew what rejection sensitivity was. They could explain time blindness. They’d listened to three podcasts that week alone.

I want to be honest before this goes any further. When I look at someone like Alex across the desk — genuinely better, finally understanding themselves — what I feel is something close to relief. This is what the work is for. This is what a diagnosis makes possible.

What I’m about to describe isn’t a failure of that. It’s what it runs into next.

I wrote about what diagnosis does to the ADHD adult’s inner world in an earlier piece — the unmasking regression, the Goodwill Clock at work, why getting better can temporarily feel like getting worse. This is the piece I promised at the end of that one. The home version.

Things were worse at home, Alex mentioned. Almost as an aside. I’ve learned to pay attention when someone says that — not to the words exactly, but to something underneath them. A timing question the diagnosis doesn’t automatically answer.

Their partner had been at the early appointments. Had asked good questions. But when I asked how things were now, Alex paused in the way people do when they’re translating something complicated into something manageable. Fine, they said. Their partner had said they were glad Alex finally had an answer. Alex thought they meant it.

I asked a few things. The laundry from Tuesday — still in the machine? Yes, but they’d been meaning to sort it. The school trip form that needed signing by Friday — the one their daughter had mentioned three times? They’d get to it. They’d been planning ahead more, they said. A weekend away with friends in the diary — first time in years they’d actually committed to something in advance.

They answered both without defensiveness. Then said, in a tone that was more statement than question: but their partner knew things were different now. Even if it didn’t look different yet.

I asked how they knew their partner knew that.

They thought about it. Said they hadn’t really talked about it.

I didn’t say what I was thinking. Not yet.

The diagnosis had done something real for Alex. Genuinely real. They understood themselves in a way they never had before, and that understanding had produced visible, measurable change. They were thriving in the ways that were legible to them — the inner world reorganised, the self-concept rebuilt, the shame beginning, finally, to lift.

What they hadn’t quite noticed was this: they’d moved from one hyperfocus to another. The subject had changed — it was themselves now, their neurology, their patterns, their potential — but the structure was identical. Intense. Consuming. Organised entirely around their own interior landscape.

That was what I’d been sitting with. Not the laundry or the form — those were symptoms. The structure underneath: the hyperfocus had a new subject, but the same shape. And the people on the periphery of a hyperfocus always experience it the same way, regardless of what it’s pointed at.

And here’s the thing about hyperfocus. It doesn’t announce itself as exclusive. From the inside, it feels like finally doing the work — purposeful, necessary, overdue. The quiet it creates around everything else isn’t noticed from within it.

Alex wasn’t indifferent to the distance. That matters to say clearly. In my experience, the ADHD adult in this moment isn’t missing care — they’re often acutely, painfully aware that something is wrong. What’s missing is the frame that makes the partner’s experience legible.

The confusion is real. So is the hurt. And underneath both, for many people, the quiet conviction that they’ve finally done everything right and it’s still not enough — which is perhaps the cruellest feature of this particular stage.

There’s a clock that runs in every relationship affected by ADHD. It doesn’t start at diagnosis. It starts much earlier — at the first missed appointment, the first forgotten anniversary, the first time someone absorbed a consequence that wasn’t theirs to carry. It runs quietly, in the background, accumulating a kind of data that never quite gets spoken aloud because there’s always a more pressing conversation to have.

Diagnosis resets something for the ADHD adult. It doesn’t reset the clock.

What the partner has been keeping — not maliciously, not consciously, but keeping nonetheless — is a running account of the years before the label arrived. The times they covered. The things they absorbed. The explanations they made to other people on your behalf. That account doesn’t clear when the prescription does.

Medication is one lever on what turns out to be a long journey. It changes what’s possible. It doesn’t change what happens next. The ADHD adult who is medicated, self-aware, and still not moving the laundry isn’t failing their medication. They’re at the beginning of something nobody adequately prepared them for.

I’ve noticed something in these post-diagnosis appointments that took me a while to name. The ADHD adult comes in transformed, or beginning to be. The partner, when they’re mentioned at all, sounds like someone standing slightly to one side of a celebration they were instrumental in making possible. Not excluded exactly. Just waiting to see what this means for them.

Alex’s daughter didn’t go on the school trip.

They mentioned it the following month, not as a significant thing. The form had slipped. They’d been away that weekend — the one they’d planned so carefully, the first time in years they’d committed to something in advance. By the time they remembered, the deadline had passed.

What struck me was the contrast they couldn’t quite see: the weekend away had made it into the diary because it mattered to them. The school trip hadn’t, and so it existed only in the soft, intentions-based system that ADHD brains have always relied on. The one that doesn’t work.

The diagnosis hadn’t changed that system. It had given Alex a more sophisticated understanding of why it kept failing — which is genuinely valuable. But understanding a pattern and interrupting it are different kinds of work, and the second one hadn’t started yet.

Their partner knew this. Had known it longer than Alex had. What they were waiting to find out — and couldn’t quite ask without it sounding like an accusation — was whether the explanation was going to become something more than an explanation.

This is not a piece about blame. The ADHD adult who arrives at diagnosis having spent years being misunderstood, managing shame, and operating at a fraction of their capacity deserves the relief that understanding brings. That relief is real and it’s earned.

But there’s a version of that relief that stops too soon. That treats the diagnosis as a destination rather than a departure point. That turns inward so completely — toward self-understanding, self-advocacy, self-repair — that the people who kept things running during the years of not-knowing are left standing in roughly the same place, just with better terminology for what they’ve been dealing with.

What the Goodwill Clock measures isn’t patience exactly. It’s accumulated faith — the ongoing bet a partner makes that the difficulty is neurological rather than chosen, and that understanding it will eventually translate into something they can feel in the texture of daily life. Not a grand gesture. Not a conversion. Just the school trip form signed before anyone had to ask.

Small things. But the clock runs on small things.

What I’ve found, in the conversations that go well, is that the shift happens not when the ADHD adult tries harder in the old way — through effort and intention and good faith that doesn’t quite make it into action — but when the self-awareness that diagnosis unlocked extends a little further than the self.

The understanding you’ve built about your own neurology is real and it’s hard-won. What becomes possible — not as an obligation but as a natural next stage — is bringing that same quality of attention to the person who has been in the room with you all along. Not to fix anything. Not to perform change. Just to actually see what’s there.

The experiment isn’t a technique. It runs for one week and asks only this: once a day, before you move on to the next thing, hold your partner’s experience in your mind for thirty seconds. Not what they need from you. Not what you’ve missed. Just — what is actually happening for them right now, in their day, in the part of the relationship that exists when neither of you is performing anything.

You don’t have to do anything with what you find. Not yet. Most people discover it’s harder than it sounds — not because they don’t care, but because the self-awareness that diagnosis unlocks is, at first, almost entirely self-directed. That’s not a flaw. It’s where the work starts. The thirty seconds is how you begin extending it somewhere else.

That’s not a programme. It’s an experiment — which is the only kind of change that actually works for ADHD brains. In the framework I use with patients, I call this the Experiments stage: small, reversible moves that generate information rather than demanding transformation. Context understood, aims identified, risks named. What’s left is a single experiment. One moment of attention aimed somewhere new.

When you’re ready — and you’ll know — you can tell your partner what you noticed. Not as a demonstration. Just as evidence, quiet and specific, that the clock is beginning to run on a different time.

Alex asked me, toward the end of that second appointment, what they should actually do.

I noticed something shift when they asked it. Not the question itself — the fact that they were asking it outward, toward someone else’s experience, rather than back toward their own.

If you’re finding yourself in Alex’s position — genuinely better in yourself, but sensing the distance hasn’t closed — I’d like to hear the specific version of it. Not the general shape, but the detail. The school trip equivalent in your own life. The specific thing — not the pattern, the actual thing. The form, the call, the moment. Reply to this email or leave a comment. This is the part of the ADHD journey most people navigate without a map, and the map gets built from what people actually tell me. I’m working on one.

And if this piece named something your partner hasn’t quite been able to say — send it to them. Not as an explanation. Just as evidence that someone else has been in the room with both of you.

The Goodwill Clock runs on a different fuel than most people expect. Not the diagnosis. Not the vocabulary you’ve built, or the books on your shelf, or the self-awareness that finally feels within reach. It runs on the small, daily evidence that the understanding has found its way into ordinary life. The form signed before anyone had to ask. The thirty seconds you paused before moving on to the next thing.

Small things. But that’s always been what the clock runs on.

He knew he procrastinated until crisis — not because he was lazy, he said, but because the crisis created the urgency his brain couldn’t manufacture on its own. He knew he overspent when he was dysregulated. He knew the exact sequence: the difficult meeting, the low-grade shame that followed, the purchase he didn’t need, the regret that arrived before the delivery confirmation. He could map it in real time. He sometimes narrated it to himself as it was happening — a running commentary on his own behaviour, accurate and completely powerless to stop it.

Then he looked at me and said something I’ve heard, in different forms, from more patients than I can count.

“So why can’t I just stop?”

It’s a reasonable question. It’s also the wrong one. Not because it’s naive, but because it assumes that insight and action run on the same track.

In the ADHD brain, they don’t.

I spent nearly seventeen years diagnosing and treating ADHD before I recognised it in myself.

In that time, I’d explained the neurology hundreds of times. I knew, clinically, accurately, how the ADHD brain processes information differently, how executive function operates, how the dopamine system shapes motivation and initiation. I had the full picture. I could draw the diagram.

And then, after my own diagnosis, I sat in a supermarket car park having just bought something I didn’t need, for a reason I’d already named and understood, and watched the receipt appear on my phone.

I’d known, while I was doing it, exactly what I was doing and why. I’d watched it happen in something close to real time. The recognition was perfect and completely useless.

It wasn’t that I lacked understanding. It was that understanding and changing are not the same neurological operation, and I’d spent seventeen years assuming they were — not just for myself, but for every patient I’d sat opposite. The map was genuinely excellent. I was its author. The car still wouldn’t start.

That experience changed how I think about the post-diagnosis period. Not because it’s unusual — I’ve come to believe it’s nearly universal — but because nobody prepares you for it.

Here’s what’s actually happening, and why the standard framing gets it wrong.

Metacognition — the ability to observe your own thinking, to recognise patterns in your own behaviour — can be remarkably intact in the ADHD brain. Sometimes sharper than intact. Years of trying to compensate for executive difficulties tends to produce unusually acute self-monitoring. When nothing has ever been automatic, you learn to watch yourself carefully. Over time, that vigilance becomes a kind of competence. It’s exhausting, and it’s real.

Executive function is something else entirely. Initiation, inhibition, the ability to convert intention into action — these run on different rails, powered by different mechanisms. And in ADHD, it’s specifically this system that’s impaired. Not the observing. Not the knowing. The doing.

Think of it this way: metacognition is the commentary booth. Executive function is the stadium. You can have the most accurate, articulate commentary in the world. If the stadium’s electrical system isn’t working, the game doesn’t happen.

The assumption most people carry into diagnosis is that awareness will do some of the work. That once you understand the pattern, you’ll be better placed to interrupt it. That knowing why you do something is at least partway to doing it differently. It’s a logical assumption. It also isn’t how it works.

From the outside, this looks like stubbornness. Or denial. Or a strange kind of self-sabotage that’s difficult to make sense of.

If you live with, work alongside, or care about someone with ADHD, you may have watched this exact thing happen. You’ve had the conversation. They agreed. They understood completely. And then it happened again, and you found yourself wondering whether they actually took it in, or whether the understanding was somehow performative.

It wasn’t. The understanding was real. What’s missing isn’t insight — it’s the bridge between insight and execution, and that bridge doesn’t get built by having more conversations about the pattern. This distinction matters because the responses that follow from “they don’t really get it” and “they get it but need a different kind of support” are completely different. One leads to repetition and rising frustration on both sides. The other leads somewhere more useful — and it starts with a distinction this piece is worth sharing with them.

The specific suffering this creates is, I think, one of the most underacknowledged experiences in adult ADHD.

Not the executive difficulties themselves — those get talked about. But the particular quality of distress that arrives when you understand yourself completely, and the understanding changes nothing. That’s its own thing. And it tends to be experienced as evidence of something shameful.

“I know exactly what I’m doing wrong, and I’m still doing it” reads internally as a character indictment. As proof of weakness, or lack of commitment, or some fundamental failure of will. It arrives with a specific humiliation. The humiliation of having the explanation and still not being able to act on it.

What it actually is, though, is perfectly intact metacognition in a brain where the execution system is impaired. You’re not failing to use your insight. Your insight is working exactly as designed. It’s running on a system that can’t automatically translate awareness into action.

Seeing clearly and being able to change are not the same neurological operation. Diagnosis gives you the first. It doesn’t automatically upgrade the second.

I’ve stopped thinking of insight as the goal. I think of it now as the prerequisite.

A prerequisite is something you need before the real work can begin, not something that does the work for you. If you’ve spent months or years understanding your patterns clearly and finding that the understanding alone hasn’t shifted much, you haven’t failed at the insight stage. You’ve completed it.

That matters. For most people in this position, the story they’re carrying is that they’ve failed at something others manage easily. That if they really understood themselves, they’d be able to change. That the problem is insufficient awareness, or insufficient will.

But you can’t willpower your way out of an executive function impairment. And insight, however accurate, isn’t a substitute for a method.

The question worth asking — the one that actually opens something — isn’t “why can’t I just stop?” It’s this: given that I can see this clearly, what would actually help me move?

That’s a different question. And it has a different kind of answer. I’m calling the method the Resolution Protocol — four steps, no colour-coded planners, no commitment to permanence — and next week’s piece walks through it in full, including a worked example of the pattern most readers will recognise immediately.

One question before you go: when did you first notice the gap between understanding your pattern and being able to change it? The moment you can name it is usually more specific than you’d expect. Tell me in the comments — I read every one.

Subscribe now

This content is for educational purposes and does not constitute personalised medical advice. If you’re concerned about your mental health, please speak with a qualified healthcare professional.

By Sunday night, she was still in it.

Not still upset about what she’d said. Not still worried, he was angry. He’d been clear, and she trusted that. She was replaying the moment itself: the exact words, the look on his face, the half-second before he responded, on a loop she couldn’t interrupt, couldn’t reason her way out of, and couldn’t explain to the people around her who kept saying some version of the same thing: you’ve apologised, he forgave you, just move on.

“I know it’s done,” she said. “I know he’s fine. So why am I still doing this?”

I’ve heard versions of that question more times than I can count. And for a long time, I didn’t have a satisfying answer. I’d describe it as part of the emotional intensity that comes with ADHD, suggest some strategies, and leave people with the vague impression that they needed to be a bit kinder to themselves. Which is true, and I’ve found it rarely helps in practice.

What I understand now, and what I wish I’d been able to explain much earlier, is that what Rachel was experiencing wasn’t a continuation of the original episode. It was a second neurological event. Different mechanism, different duration, different intervention required. And almost nobody talks about it as such.

What actually happened (and why it kept happening)

When Rachel was sharp with her colleague, something specific happened in her nervous system. The ADHD brain, with its heightened amygdala response and compromised prefrontal cortex regulation, reacted to the social moment more intensely than the situation warranted. That’s the part that most people who’ve read about ADHD emotional dysregulation will recognise: the disproportionate reaction, the flood, the response that bypasses the rational brain before the rational brain has a chance to intervene.

But here’s what the standard explanation leaves out.

After that episode, the brain doesn’t simply reset. In ADHD, the Default Mode Network (the brain’s internal narrative system, active during rest and self-referential thought) fails to deactivate properly. For most brains, completing a social event allows the network to quiet. In ADHD, it keeps running. And because the DMN connects directly to the hippocampus and amygdala, it keeps feeding emotionally charged memories back into the threat-detection system, which responds as if the threat is still present. The person is not choosing to replay the Friday meeting. The brain’s architecture is doing it for them, and the architecture is working exactly as it does in ADHD.

Because rejection sensitivity is still online, still scanning for social threat, the memory of the episode keeps re-triggering the same threat response that caused the original flooding. The apology resolved the interpersonal situation. It hadn’t resolved the neurological one. Those are different things, and treating them as the same, which is what “just move on” implicitly does, misses what’s actually happening.

If this is the first piece of mine you’ve read: I write about the ADHD patterns that don’t appear in diagnostic questionnaires, the neurology that was missing from my own training. Subscribe below to get the next piece when it’s out. Free.

Subscribe now

How the cultural response makes it worse

The standard responses to this kind of aftermath are, almost without exception, neurologically misaligned.

“You’ve already apologised. What more do you want?” Implies the loop should have closed when the social repair was made. For an ADHD brain, that’s not how it works.

“You’re being too hard on yourself.” Accurate. Addresses the symptom, not the mechanism. Doesn’t help the loop stop.

“Just let it go.” The letting-go mechanism is exactly what’s not working. This is roughly equivalent to telling someone with a broken wrist to just grip harder.

“Everyone makes mistakes.” Generic reassurance that bypasses the specific neurological experience entirely.

I’m not suggesting the people who say these things are wrong to care. They’re trying to help. But they’re responding to what they see: someone who seems to be choosing to ruminate, rather than what’s happening, which is someone whose brain is generating rumination without much voluntary participation.

The effect of these responses, typically, is to add a layer of shame onto the existing loop. Now the person is replaying the original episode and aware that they’re failing to “just move on” in the way a normal person apparently would. The second neurological event gets heavier.

Rachel had spent the weekend apologising to herself, in a private loop, for not being better at ending the loop. That’s what this looks like from the inside.

What the aftermath actually requires

The strategies that help with emotional flooding in the moment (physical movement, space, time) are partially relevant here, but the aftermath is a different phase and needs a slightly different approach.

Three things that I’ve seen make a consistent difference, specifically in that window after an episode, when the repair has been made but the loop hasn’t closed:

Externalising the narrative. Rumination is, in part, an internal process that feeds on itself. The act of narrating what happened, aloud or in writing or to a trusted person, in simple sequential terms, creates just enough distance from the loop to interrupt it. Not therapeutic processing. Not a forensic analysis of what went wrong. Just: “I was short with someone. I apologised. He was fine about it. That’s what happened.” The narration doesn’t resolve the neurological event, but it gives the Default Mode Network a different task, and sometimes that’s enough to shift the gear.

One specific closing action. Apologies are often experienced by the ADHD brain as insufficient to close the loop, even when the other person has accepted them. What sometimes works better is a single concrete action that the brain can register as completion. Not another apology. A practical thing. Sending a relevant article to the colleague the following week. Buying a coffee. Something small and tangible that the brain can file as “loop closed.” This sounds almost too simple to be useful. In practice, it works often enough that I now mention it routinely.

Physical reset, specifically for this phase. The nervous system in the aftermath of an emotional episode is still partly activated. Movement helps complete that activation cycle, for the same reasons it helps during flooding, but the intensity required is lower. A walk, specifically. Not vigorous exercise, which can extend the activated state. A walk, which appears to do something in particular for the rumination loop that other forms of movement don’t quite replicate. I don’t have a clean neurological explanation for why this seems to be true, but the pattern is consistent enough across enough people that I’ve stopped dismissing it.

Most of what exists about ADHD emotional dysregulation focuses on the moment of flooding. Almost nothing addresses what comes after it. That gap is worth naming.

A different kind of scaffold

What I’ve been working on over the past year, and will be sharing in full in the coming weeks, is a framework specifically designed for the two phases of ADHD emotional dysregulation that standard guidance doesn’t address well: the moment itself, and the aftermath.

The aftermath piece is what most people need most urgently, because it’s the part that goes on longest and does the most damage to relationships and self-concept. The flooding is often for minutes. The shame spiral can be days.

The framework, four words: Context, Aims, Risks, Experiments, is designed to be activated under cognitive load, which means it has to be simple enough to use when the brain isn’t working at its best. For the aftermath specifically, it looks something like this.

Context: name the phase, not the episode. “I’m in the aftermath. The episode is over. The loop is still running.” One sentence. Not analysis. Just locating yourself in the right place on the map.

Aims: what matters in the next hour. Not the next year, not the relationship in the abstract. The next hour. For Rachel on Sunday night: “Get through this evening without making the loop worse.” That’s sufficient. Anything more ambitious asks too much of an activated nervous system.

Risks: what the aftermath typically makes you do that extends rather than closes it. For most people, this includes things like: checking whether the other person seems different with you, re-reading the exchange, sending another apology when one has already been received, and catastrophising about what it says about you as a person. Naming the specific risk is the intervention. You don’t have to stop doing it. You just have to see it coming.

Experiments: one small action. The narrative externalisation. The walk. The concrete closing gesture. One thing, not a strategy. The distinction matters because strategy implies judgment about which approach fits, and judgment is in short supply in the aftermath.

What Rachel said at the end of the appointment

We spent most of the session on this: the mechanism, what was actually happening across that weekend, the difference between the interpersonal repair (done) and the neurological one (still running).

At some point, she said something that I’ve thought about since: “I’ve been apologising to myself for three days for something he forgave me for in thirty seconds.”

That’s a precise description of what this actually is.

Not weakness. Not excessive sensitivity. Not a personality trait to be managed through willpower. A brain that can’t close the loop through the usual mechanism, running the threat-detection system on a social event that, by every external measure, is already resolved.

Understanding that distinction doesn’t stop the loop immediately. But it changes the relationship to it. You stop spending the loop’s energy on self-recrimination, and you start spending it on doing the one small thing that might help it close.

That’s a meaningful shift. Not a cure. A shift.

If this named something you haven’t been able to explain to someone else, a partner, a colleague, a friend who keeps saying “just let it go,” it might be worth sending it to them. Not as an explanation of yourself, necessarily. Just as something that might shift what they reach for next time.

The aftermath, the hours after an emotional episode, when the repair has been made, but the brain hasn’t caught up: does this land? Tell me what that window looks like for you. I read every reply.

For weekly clinical observations about the patterns of ADHD that rarely make it into standard guidance. Subscribe below. Free, and the next piece will cover why the same emotional dysregulation that shapes the flooding also shapes how ADHD adults communicate in the relationships that matter most.

All clinical details have been changed to protect confidentiality. This newsletter is educational and draws on clinical and personal experience. It does not replace individual medical advice.