Not long ago I gave up an expensive apartment in Tiburon with dead-on views, on a steep hill, isolated from everything. It was not a home. It looked fine from the road, and it was costing me more than money. Letting it go took me out of survival mode, and I found I had room for the things I had always meant to do. I started writing. I could finally work on the technology I had always intended to build.
I had been in survival mode for so long that I had stopped noticing it, and what ended it was not a breakthrough or a practice. It was lowering a real demand. For me that was enough to begin. For some people reducing demand may be sufficient, and for others it is only the beginning, and this essay is about why relief and rest sometimes aren't enough and what the research suggests may help a stressed nervous system settle.
§ 1 · When the system won't stand down
In your patience possess ye your souls.
A healthy nervous system is built to spike and recover. It surges for a threat and then returns to calm when the threat has passed. Dysregulation, as the clinicians use the word, is what happens when the second half stops working. The system stays switched on after the danger has gone, idling at high revolutions with nowhere to park. It's the state people call wired but tired, an engine that can't downshift even when the pressure eases.
It shows up in two directions. Hyperarousal is racing heart, shallow breath, agitation, a mind that won't stop. Hypoarousal is the opposite: numbness, flatness, disconnection, a shutdown that arrives when mobilization fails to resolve. Many people swing between the two and rarely rest in the middle, where clear thinking and steady feeling live.
Daniel Siegel gave that middle a name, the window of tolerance: the zone of arousal in which a person can think, feel and respond flexibly. A useful way to picture chronic stress is that it narrows the window, so that smaller stressors tip you out of it. I find the picture helpful and I hold it as a picture. Siegel's term is a clinical metaphor, not a measured quantity, and nobody can tell you the width of your window in units.
§ 2 · Why it stays stuck
If dysregulation were only about the present stressor, removing it would end it. The reason it often doesn't is that the body adapts to the emergency. Bruce McEwen called the cumulative wear of a stress response that never fully stands down allostatic load: the bill that comes due across the cardiovascular, metabolic, neuroendocrine and immune systems for running in crisis mode too long. The counter-intuitive implication is that after months of load, taking the stressor away is often not enough.
A useful image is a thermostat. Chronic stress may recalibrate the system's set points, for cortisol, for inflammation, for baseline arousal, to higher defaults. A break turns down the heat in the room but not the thermostat. That would explain why a weekend off, an extra hour of sleep or a vacation so often fails to move the needle for people who've carried a load for years. I want to flag how settled the science is here. The idea of allostatic load is well established as a framework. The specific claim that cortisol's daily rhythm flattens under chronic stress, with a blunted morning peak and a raised evening level, has support in studies of chronic work stress and caregiving, but the findings are mixed and the measurement is noisy. I hold the thermostat as a working model.
Where the load surfaces differs from person to person. It may show up as a racing or irregular heart, blood pressure, broken sleep, a flaring gut, recurrent infections, a mind that can't focus, or chronic tension. What looks like a set of separate complaints can be one pressure expressed through different doors. I say can be because it's also true that many of those complaints have unrelated causes, and anyone with persistent symptoms needs a physician, not an essay.
Across these doors, one measure is widely used as a non-invasive read on how flexibly the autonomic system is working: heart rate variability, the beat-to-beat variation in heart rate. Lower variability tends to accompany stress and poorer regulation. It is a useful signal and a noisy one, and it isn't a diagnosis.
§ 3 · When closeness reads as load
Here is the cruel turn. A narrowed window has no spare capacity, so even wanted, safe closeness can register as load and not relief.
Co-regulation, the settling of one nervous system by another, is one of the most powerful routes back to calm. It also requires bandwidth to receive. When the window is already narrow, the arrival of closeness, even longed-for closeness, can exceed what the system can take in. This is my reading, drawn from clinical accounts and from my own life, and not a tested finding. It would explain why a person under heavy load may pull back from contact they consciously want. The withdrawal isn't a verdict on the bond. It is load-shedding, the system reducing input because it has none to spare.
I have done this to people I love. One of them told me afterward that I disappeared exactly when she needed me closer. I stayed in the room. I answered, I did what needed doing, and something went out of me that she could feel go. I've come to think part of that was load-shedding that neither of us understood. It doesn't excuse it. It does suggest what might have helped: closeness in a form a depleted system can tolerate, which is low-demand, predictable, unhurried and asking nothing. A dose small enough to land as safety and not as one more demand.
§ 4 · More than rest
Recovery may require more than rest: less demand, reliable sleep, supportive contact, and practices that help us settle. A system that has adapted to crisis may not simply return to baseline when the stress is removed. Cues such as a slow breath, a steady voice or a familiar face can shift a body's state, and insight alone often doesn't, though that varies from person to person.
The encouraging half is that much of this appears to be reversible. I prefer to think of chronic stress as closer to a software lock than a hardware failure, though that is an analogy, and some effects of long adversity may not fully reverse. The work looks like accumulation: each slow breath, each safe contact, each spike survived and recovered from.
§ 5 · The levers
The things that seem to help are few and unglamorous, and I would put five first: reducing a real demand where you can, regular sleep, movement, slow breathing, and supportive contact.
Reducing demand can make the other practices easier to benefit from, though not every demand can be removed. Regular sleep, and morning light, help anchor a daily rhythm the load may have flattened. Aerobic movement tends to buffer stress, so long as it stops short of overtraining, which is itself a stressor. Slow breathing, roughly six breaths a minute with a longer exhale, is the best studied: Paul Lehrer and Richard Gevirtz's work on heart rate variability biofeedback, and meta-analyses since, find that it can raise HRV and lower anxiety, with modest effects, and it is the one dial under conscious control. And safe, attuned contact is often the most direct route. Mindfulness and the metabolic basics can help too, within the limits I described in Bare Attention.
§ 6 · Why it is slow
The hardest part is the timeline. State can shift in minutes: one good session of slow breathing visibly changes arousal. Deeper change, where it comes, seems to take weeks or months and differs a great deal between people, and I know of no reliable timetable. Recovery is also not linear.
Holding that timeline loosely is itself protective. A common way a reset fails is the thought I should be over this by now, which mistakes the slow middle of the curve for a dead end. Recovery is also not linear. Setbacks are part of the curve and not evidence of failure.
§ 7 · The two environments that set the window
No reset happens in a vacuum. The two environments a person lives in most, work and close relationships, either narrow the window or widen it. Chronic occupational stress is one of the best-documented drivers of allostatic load, but the research is equally clear that work can regulate and not just tax. Robert Karasek and Töres Theorell's job-strain model points at structure and not hours: manageable demand matched by real control over how the work is done, fairness, and genuine support. Where those hold, a heavy workload is more likely to be absorbed than stored. Where demand is high and control and support are low, the body pays the difference.
Relationships carry even more weight, because proximity to a safe other is itself a resource. Social baseline research (Coan and Beckes, and others) suggests the brain treats a trusted person's presence as shared load, lowering the cost of facing a threat. A secure bond supplies a haven to return to and a base to venture from. For a depleted system the most useful conditions look boring: consistency over intensity, low-demand presence, attunement and repair after rupture, predictability the body can come to trust.
§ 8 · Measurement, and what it cannot do
A reset is invisible and slow, which makes it easy to misjudge. A good day can look like recovery, and suppression can look like calm. It helps to keep three things apart. The window of tolerance is a metaphor, useful and unmeasured. Heart rate variability is an established, noisy marker of how flexibly the autonomic system moves between mobilization and rest, and its trends can be tracked. And my own questionnaires, the Zen Score and a Coherence Score meant to tell real regulation from something that only looks like it, are proposals. Calm versus masked shutdown is the hardest thing in this field to measure, and I don't claim that mine does it. They need independent testing before anyone should rely on them. A wearable's number is a readout and not the regulation.
§ 9 · Back within the window
The day I gave up that apartment I did not become calm. I became free to start. I started writing. I began to work on the technology I had always meant to build. That is what lowering a real demand made possible for me, and it is the part of this I trust most.
If you can put something down, put it down. Then build a small stack you can repeat: a slow breath, a walk, a regular bedtime, a person you don't have to perform for. For some people that will be enough, and for others it is the start. When you catch yourself saying I should be over this by now, treat it as a sign you are in the slow part.
Sources and further reading
Siegel, D. J. (1999). The Developing Mind. Guilford. Window of tolerance.
McEwen, B. S. (1998). Protective and damaging effects of stress mediators. New England Journal of Medicine, 338; McEwen, B. S. & Stellar, E. (1993). Stress and the individual: mechanisms leading to disease. Archives of Internal Medicine, 153.
Sapolsky, R. M. (2004). Why Zebras Don't Get Ulcers. Holt.
A note on a theory I do not use: Porges, S. W. (2011), The Polyvagal Theory (Norton), is often cited in this territory. Physiologists have published serious critiques of its claims about the vagus nerve, for example Grossman, P. & Taylor, E. W. (2007), Toward understanding respiratory sinus arrhythmia, Biological Psychology, 74. Nothing here depends on it.
Lehrer, P. M. & Gevirtz, R. (2014). Heart rate variability biofeedback: how and why does it work? Frontiers in Psychology, 5.
Goessl, V. C., Curtiss, J. E. & Hofmann, S. G. (2017). The effect of heart rate variability biofeedback training on stress and anxiety: a meta-analysis. Psychological Medicine, 47.
Laborde, S. et al. (2022). Effects of voluntary slow breathing on heart rate and heart rate variability: a systematic review and meta-analysis. Neuroscience and Biobehavioral Reviews, 138.
Karasek, R. & Theorell, T. (1990). Healthy Work. Basic Books; Siegrist, J. (1996). Adverse health effects of high-effort/low-reward conditions. Journal of Occupational Health Psychology, 1.
Beckes, L. & Coan, J. A. (2011). Social baseline theory: the role of social proximity in emotion and economy of action. Social and Personality Psychology Compass, 5.
Feeney, B. C. & Collins, N. L. (2015). A new look at social support. Personality and Social Psychology Review, 19.
van der Kolk, B. (2014). The Body Keeps the Score. Viking. Influential and contested.