Why your comfort zone is bigger than you think and how to grow it gently

There is a particular kind of frustration that comes with setting a step you think is too small to count. You agree with yourself that you will go to the event. You will not stay long. You do not have to talk to anyone. Just be there. And then, on the way home, the inner voice arrives. That barely counts. Other people would have at least introduced themselves. You should have done more.

This is one of the most common patterns I see in clinical work with anxiety. Not the avoidance itself. The judgement of the step that was actually taken. The decision that the small step was not the work, but the warm-up. That the real work was somewhere harder, somewhere later, somewhere that did not happen.

What follows is a different way of thinking about the comfort zone, drawing on the current research on how the anxious brain actually changes. The shift is small but it matters. The small step is not the warm-up. The small step is the work. The brain needs the safety signal first, before it can do anything else.

The comfort zone is often pictured as a circle you are stuck inside. The research describes something more useful. Three states, not two. There is the zone where the body is settled and thinking is clear. There is the zone where the body is activated but still able to learn. And there is the zone where the body is overwhelmed and learning shuts down.

The middle zone is the one that matters. Clinicians often call it the window of tolerance (Siegel) or, in autonomic terms, the ventral vagal range. Porges (2024), in his most recent synthesis of polyvagal theory, describes it as the state in which the social engagement system can come online: heart rate steady, breath relaxed, attention available, the system able to register cues of safety from the environment and from other people. This is the zone in which the brain learns. It is not the same as comfort. It includes activation. It just keeps the activation inside a range the system can work with.

The third zone is where most people get into trouble. Above the window, the body shifts into hyperarousal: the same physiology that handles real danger. Below the window, into hypoarousal: numbness, shutdown, withdrawal. In both, the parts of the brain involved in flexible thinking and new learning go offline. The system is doing what it was built to do. It is just not in a state where new learning can land.

This matters because most popular advice about the comfort zone collapses these three states into two. Inside or outside. Safe or brave. The lived experience is more granular. There is a difference between a step that lands you in the stretch zone, where new learning can occur, and a step that puts you in overwhelm, where it cannot.

There is a popular idea that anxiety is overcome by pushing through it. Confront the fear. Make yourself do the thing. The body will catch up. Some of this comes from an older model of exposure therapy, in which the goal was to stay in the feared situation long enough for the fear response to subside. The current evidence has moved on from that view.

Hofmann, Kasch and Reis (2025), in a meta-analysis published in Clinical Psychology Review covering 49 randomised trials and 3,645 participants, found that cognitive behaviour therapy is associated with moderate effects across anxiety disorders compared with control conditions. The treatment that produces these effects is not flooding. It is graduated. The steps are small enough that the system can stay in the stretch zone, not be thrown into overwhelm.

When the system is pushed too far too fast, two things happen. The first is that the experience is encoded as further evidence the situation is dangerous. The fear memory is reinforced rather than updated. The second is more subtle. Kausche, Carsten, Sobania and Riesel (2025), in an updated meta-analysis in Neuroscience and Biobehavioral Reviews, found that anxiety disorders involve a specific deficit. People with anxiety are not just more reactive to threat. They have a harder time registering safety. Cues that should signal that things are okay do not always land. When the system is overwhelmed, that difficulty becomes more pronounced, not less.

The clinical implication is straightforward. The advice to push through is, for many people, advice that pushes them out of the very state in which their brain could change. The instruction is well-intentioned. The mechanism does not match the brain it is trying to help.

What does help, then? The current synthesis comes from Craske, Treanor, Zbozinek and Vervliet (2022) in Behaviour Research and Therapy. They describe an inhibitory learning model of how exposure-based approaches actually work. The old model assumed the fear memory was erased. The new model assumes it is not. The fear memory stays. What changes is that a new memory is laid down alongside it. A safety memory. A memory in which the same situation occurred, and the feared thing did not happen, and the system is still here.

The two memories then compete for retrieval. Whichever one is more accessible in the moment is the one that drives the response. This is why old fears can return after a stressful day, in a new context, or after a long gap: the safety memory has not been erased, but the fear memory has won the retrieval contest.

What this means in practice is that exposure is not really about tolerating distress. It is about giving the brain enough exposures, in varied enough contexts, with the system inside its window of tolerance, for the safety memory to grow strong enough to compete with the fear memory. The work is not heroic. It is repetitive. And it requires the system to be in a state where new learning can occur.

This is where the size of the step becomes the whole game. A step that is too big does not produce a safety memory. It produces another fear memory. A step that is the right size, repeated, produces the safety memory the system needs. The job is not to find the bravest step. The job is to find the step that produces learning.

What follows is a simple structure for working with a step you would like to take but have been struggling with. It is not a treatment plan. It is a way of right-sizing the step so that the system has a chance to do the actual learning.

Take a worked example. Someone with social anxiety who would like to feel less daunted by gatherings. The step they often set themselves is: go to the event, talk to people, make conversation, stay for the whole thing. That step has at least four moves bundled into it. Each one is its own piece of work. Stacking them together turns a workable step into one the system has almost no chance of meeting.

The pattern people most often miss is the second move. They right-size the step, then add something back. Just go to the event and talk to one person. Just walk in the door and stay until the end. The and is what tips a workable step into an unworkable one. The instruction to oneself is usually well-meant. The system reads it as the original overwhelm, dressed up.

Wootton and colleagues (2025), in a systematic review and meta-analysis in Australian Psychologist covering 37 studies and 3,234 participants, found that cognitive behaviour therapy for social anxiety is associated with large effect sizes when delivered individually or remotely. The structured, graduated approach is part of what makes it work. The step is small. It is repeated. The progression is the point, not the size of any single move.

The barrier I encounter most often in this work is not the step itself. It is the judgement of the step. People arrive convinced that what they managed last week did not really count, because it was not the version of the step that someone less anxious would have taken. The internal comparison is to a person who is not in their nervous system. The comparison they should be making is to themselves yesterday.

Self-criticism, in this context, is not a neutral observer. It is another threat input. Anthes and Dreisoerner (2026), in a systematic review of 85 studies in Mindfulness, found that self-compassion supports change through better emotion regulation and a settled rather than mobilised body. When the body is settled, the system can take the next step. When the body is being criticised, the system is in defence. The criticism is not motivating the next move. It is making the next move harder to take.

The honest message is patient. If you took a step today that felt smaller than the step you wanted to take, you did the work. The work was the step that landed inside your window, repeated. The version of the step that someone less anxious would have taken was not available to you today. That is not a moral failing. It is information about where your edge currently is.

Two practical notes to close. First, this kind of graded work is what evidence-based therapy for anxiety is built around. If you have been trying to grow the comfort zone alone and the system is not settling, that is the point at which professional support helps. The structure is genuinely easier to hold with another person. Second, the work is slower than the culture suggests it should be. The comfort zone does not grow in leaps. It grows in increments small enough that the brain can register safety. That increment is bigger than you think.