Ask a senior leader what they do when pressure rises, and you will usually hear some version of a technique.
A slow breath before the board meeting. A walk after the difficult call. A pause before replying to the email that landed wrong.
These things work. I use some of them myself. And for a long time, I assumed that was the work: helping capable leaders settle themselves more quickly once pressure had already registered in the body.
Over years of coaching, however, I have come to think the more important goal sits somewhere else.
Calming down is useful. It just starts too late.
Two kinds of work
Calming down is regulation work.
It builds resilience and begins after the response is already underway — after the jaw has tightened, thinking has narrowed, and attention has shifted towards what might go wrong.
Done well, regulation shortens the episode. It stops a difficult moment from becoming a difficult afternoon. It keeps a leader functional when they do not feel calm.
What it does not necessarily do is change why the body reacts that way the next time.
Threshold work happens earlier. Instead of asking How quickly can I settle once this situation activates me?, it asks Why does this situation still produce such a strong response at all?
The distinction often becomes visible through a question I ask early in first conversations:
When pressure hits, do you have to calm yourself — or does it simply not affect you as strongly as it used to?
Most executives realise, somewhere in the middle of answering, that they have mainly experienced the first. They have become exceptionally good at managing a response that has never actually become smaller.
That is one of the quieter patterns resilience produces. Recovery gets faster. Coping gets more sophisticated. But the point at which the body reacts stays roughly where it was.
The same board meeting still produces the same anticipatory tension. The same person still creates the same vigilance. The same kind of email still causes the same internal jolt.
Every demanding week is still paid for in full — only more efficiently.
What the research on unlearning these responses says
Some of the most useful science here does not come from leadership research. It comes from decades of clinical work on how the brain learns that certain situations are threatening, and how that learning can later be updated.
Michelle Craske and colleagues at UCLA have studied the mechanisms behind exposure-based treatment for many years. In a widely cited 2014 paper in Behaviour Research and Therapy, they described what is known as the inhibitory learning model. Three parts of it matter for anyone leading under sustained pressure.
The original learning is not simply erased
Imagine your system has learned, through experience, something like: When I am challenged unexpectedly in front of the board, I need to be ready.
That learning may have formed from one genuinely difficult experience or from many smaller ones. Once established, the brain becomes very good at recognising anything that resembles it. A certain room. A particular tone. Even the calendar invitation the night before.
Rationally, you may know none of these are dangerous. Your body can still respond as though something important is at stake.
Over time, new experience can teach something different: I can be challenged and still be respected. But the new learning does not simply delete or replace the old association. It competes with it. And at first, it is fragile and tied to the context in which it formed.
A leader may feel entirely comfortable being challenged by their own team, then notice the old response return when the same thing happens in front of the CEO. That is not irrational. The system has learned safety in one context and has not yet learned it in another.
Feeling calmer is not what creates lasting change
This is the part that is easy to miss.
In the inhibitory learning model, the important question is not whether a person becomes less activated during an experience. It is whether the brain expected one outcome and experienced another. Craske’s team calls deliberately creating that mismatch an expectancy violation.
Imagine a leader who believes, without ever putting it into words: If I pause before answering a difficult question, people will think I am not on top of the business.
So they answer quickly. Every time. And the speed itself prevents them from discovering whether the prediction is true.
If instead they allow a three-second pause, feel the discomfort of it, and discover that nothing goes wrong — perhaps the room even becomes more attentive — the system receives new information. The prediction was If I pause, I will lose authority. The experience was I paused, and my authority did not change.
That is a small moment. It is also exactly where learning can change.
The goal is not that the leader feels calm while doing it. The goal is that the old expectation meets evidence that does not confirm it.
Safety behaviours can interfere with that learning
This is where calming techniques become more complicated.
One of the strategies Craske’s team discusses for strengthening new learning is reducing reliance on safety signals and safety behaviours — the things a person comes to believe they need in order to get through a difficult situation.
The logic is simple. If you get through a difficult situation while relying heavily on something that makes you feel safe, the brain can credit the good outcome to that strategy rather than updating its view of the situation.
A leader feels intense anticipatory tension before a board meeting. They spend twenty minutes regulating beforehand, rehearse every possible answer, keep detailed notes in front of them and carefully manage their breathing whenever they are challenged.
The meeting goes well.
Consciously, they think: That was fine. Their system, however, may have learned something more conditional: That was fine because I prepared obsessively, regulated beforehand and stayed tightly controlled throughout.
The meeting itself stays coded as dangerous. The strategy gets the credit. Next quarter: the same response, the same management, the same cost.
That is the case against calming down as the goal. Not that it fails. That when the system learns calm has to be produced before a situation is safe, the strategy can quietly preserve the very response it is managing.
A brain that has learned to over-respond
A second line of evidence comes from an area that seems even further removed from leadership.
In 2022, Yoni Ashar, Tor Wager and colleagues published a randomised clinical trial in JAMA Psychiatry testing Pain Reprocessing Therapy for chronic back pain. The treatment rests on the idea that, in some forms of chronic pain, the brain has learned to treat sensations that no longer signal tissue damage as danger — and that this learned interpretation helps keep the pain active. The task is not to soothe the pain each time it appears. It is to help the brain update what the sensation means.
The results were striking. Of 151 participants, two-thirds of those in the therapy group were pain-free or nearly pain-free after treatment, compared with 20% in an open-label placebo group and 10% in usual care. The gains were largely maintained at one-year follow-up, and brain imaging showed reduced responses in regions involved in processing pain.
I am certified in Pain Reprocessing Therapy. That training is why I take this mechanism seriously rather than treating it as an interesting metaphor. The work is not about soothing a response every time it fires. It is about changing whether it needs to fire at the same intensity at all.
Executives are not chronic pain patients. The parallel is the mechanism, not the condition: a system that once learned a situation was dangerous can keep producing the same response long after the circumstances that created that learning have changed.
In leadership, that might be the board that once caught you unprepared. The stakeholder who escalated over your head. The senior colleague whose criticism once carried real consequences.
At the time, the response made sense. Years later, you are more experienced, your authority is greater, the consequences are different. You know that intellectually.
The body can still brace as if none of it has changed.
I have seen this in others, and I have recognised it in myself: the response didn’t shrink because I got better at soothing it, but because it stopped being confirmed as a threat.
Performance holds. And the bracing is where the margin quietly disappears.
What threshold work looks like in practice
If the goal is to change what produces the response, the work looks quite different from a toolkit of calming techniques.
Make the prediction visible
Before anything can be relearned, the leader has to recognise what their system expects to happen — not only intellectually, not only afterwards, but in the moment.
A board challenge might carry: If I don’t have the answer immediately, I will lose credibility. Delegating something important might carry: If I am not closely involved, something will go wrong and I will be held responsible.
What appears on the surface as “stress” usually has an expectation underneath it. The body is preparing for what it expects will happen.
Test the prediction under graduated load
Once the prediction is visible, it can be tested. Not by throwing someone into the most activating version of the situation and telling them to cope. Incrementally.
A leader who believes they must always answer immediately allows a short pause in a lower-stakes meeting. Someone who over-prepares enters one familiar conversation with fewer notes. Someone who struggles to delegate gives another person genuine ownership of a contained decision — and resists checking.
This is close to what people mean by leaving the comfort zone. But leaving the comfort zone is not automatically growth.
For the response to update, the step needs to be large enough for the old response to show up, and small enough that the person can still notice what else is true in that moment.
Too small, and there is little to learn.
Too large, and the experience can confirm the old conclusion — this is exactly why I need to brace — and strengthen the response rather than reduce it.
So the challenge has to come with attention. While the activation is present, the person needs enough room to register the evidence their system would otherwise miss:
My heart is beating faster, and nothing bad is happening.
I don’t know exactly what to say yet, and the room is still with me.
Someone disagrees with me, and the relationship is intact.
These are observations, not reassurances. Without that recognition, someone can survive difficult experiences again and again without the system ever updating what they mean. They get better at enduring. That is not the same as becoming less activated.
Flight test teams expand an aircraft’s operating envelope the same way. They do not repeatedly take an aircraft to the edge, recover it and declare it resilient. They find where the envelope actually is — small increments, real conditions, careful observation — and over time the usable envelope grows.
Learn it where it counts
Because new learning is tied to context, it eventually has to happen in the rooms that matter: the board conversation, the escalation, the quarter that goes wrong.
A leader can understand perfectly well in a coaching conversation that disagreement does not mean something is wrong. Then someone challenges them sharply in a meeting, and the old response appears within half a second.
That does not mean the insight was wrong. It means insight and a learned response are not the same thing. The system needs evidence in the environment where the response actually occurs.
An honest caveat
None of this research was conducted on executives.
The inhibitory learning work comes from the treatment of fear and anxiety. The Pain Reprocessing Therapy trial studied chronic back pain, and it is one trial whose findings still need replication at scale.
What transfers is the mechanism, not the outcome data. There is strong evidence that learned threat responses can be updated through new experience and inhibitory learning. There is not equivalent research showing that a threshold-based coaching approach produces defined outcomes in executives. I would rather say so before a sceptic has to ask. What I can offer is what I observe in practice: an established mechanism, applied in a newer context, producing the change it predicts — responses that become smaller, not just better managed.
None of this makes regulation unimportant. If your heart is racing before the most consequential presentation of the year, breathe. If an email has made you furious, take the walk. If your thinking has narrowed and you need room to decide well, regulate.
That is useful work. It is simply downstream.
Regulation asks: How do I manage this response well?
Threshold work asks: Why is my system still producing it at this intensity?
The question worth sitting with
Most leaders I meet have already invested heavily in learning how to recover well. That investment was not wasted. It is often what kept them functioning through years of demanding work.
At senior levels, though, the more useful question may no longer be how quickly you can settle yourself once the response has started.
It may be what still makes your body respond as though something is at stake when, in reality, it no longer is.
And how much more room you would have if it didn’t.
Sources
- Craske, M. G., Treanor, M., Conway, C. C., Zbozinek, T., & Vervliet, B. (2014). Maximizing exposure therapy: An inhibitory learning approach. Behaviour Research and Therapy, 58, 10–23. https://doi.org/10.1016/j.brat.2014.04.006
- Ashar, Y. K., Gordon, A., Schubiner, H., Uipi, C., Knight, K., Anderson, Z., Carlisle, J., Polisky, L., Geuter, S., Flood, T. F., Kragel, P. A., Dimidjian, S., Lumley, M. A., & Wager, T. D. (2022). Effect of pain reprocessing therapy vs placebo and usual care for patients with chronic back pain: A randomized clinical trial. JAMA Psychiatry, 79(1), 13–23. ClinicalTrials.gov: NCT03294148
