The patient has good sleep hygiene. Fixed time, dark room, no screens before bed. The protocol is correctly executed. Still, when night arrives, something does not yield. The body is in bed. The system has not arrived.
The clinician who asks carefully finds a description the patient had not framed as clinical: night does not feel like a safe place. There is no identifiable threat, no noise, no danger, no concrete reason. There is something more diffuse: an inability to let go that the patient interprets as personality, or simply as not being good at sleep.
That description has its own clinical reading.
Sleep requires a condition rarely named in sleep-hygiene protocol: the nervous system must perceive the environment as safe enough to deactivate vigilance networks. Darkness, silence, and routine prepare external conditions. The transition to sleep occurs when internal conditions also yield, when threat-monitoring systems reduce activity enough for consolidation processes to begin.
In a system that has operated in alert mode for a long time, that transition does not happen automatically because external conditions have improved.
Kredlow and colleagues describe how the amygdala and threat-processing networks interact with the prefrontal cortex in learned fear and emotional response regulation. Under sustained activation, the balance between threat detection and prefrontal regulation shifts. The system prioritizes vigilance over recovery. That dynamic does not require a PTSD diagnosis to be clinically relevant. It can operate in any nervous system that has learned, for reasons that once made sense, that alertness is safer than rest.
Arnsten documents how sustained stress impairs prefrontal function: the capacity to regulate emotional responses, inhibit automatic reactions, and sustain deliberate calm is reduced under chronic load. In sleep, that means the system has less capacity to actively modulate vigilance networks at night. The effort to relax competes with a system whose resources for deliberate regulation are already reduced.
The result is a common clinical paradox: the patient who tries harder to sleep sleeps worse, not from lack of discipline, but because active regulatory effort engages networks sleep requires the organism to release.
The distinction the sleep-hygiene protocol does not always call forward is this: some insomnia is a problem of external conditions or circadian habits. Some insomnia is a problem of safety perception. The system does not initiate sleep transition because it has not received enough signal that the environment and the organism are safe enough for this night.
The second category does not respond like the first. CBT-I remains a first-line intervention for insomnia, especially where the core mechanisms involve conditioned arousal around the bed, reduced homeostatic pressure, circadian misalignment, or dysfunctional expectations around voluntary control of sleep. Its design is coherent with that structure. In the patient whose insomnia is organized around safety perception, the behavioral layer may be present, but it may not be primary. The autonomic state from which the patient enters sleep conditions is vigilance.
Transcendental Meditation is located at a different point of action. Orme-Johnson and Barnes document reductions in trait anxiety, with stronger effects in individuals with higher baseline activation. Trait anxiety is the dimension that reflects the tendency of the system to sustain alertness with relative independence from the immediate environment. Walton and colleagues associate regular TM practice with reductions in stress-related neuroendocrine activation, including cortisol. The practice does not instruct the system to relax. It offers conditions in which the system may reorganize without active regulatory effort.
Constitutional reading clarifies why this patient’s nocturnal activation has this texture. A profile of high variability and environmental sensitivity that has learned silence can be deceptive has a different vigilance architecture from a profile of sustained tone that never fully processes the day’s accumulation. That difference does not replace sleep treatment. It changes which layer the clinician hears speaking.
CBT-I keeps its place. This article names the layer before it: the clinician who can distinguish insomnia of conditions from insomnia of safety perception has a different question before initiating any protocol.
The clinician who recognizes one of their own patients in this article already has the right question.
It is not how to improve sleep conditions. It is whether this patient’s system has safe enough ground for sleep to land.