Healing in the Dark

The Promise of Sleep and Dream Engineering for PTSD and Depression

If you (or a loved one) has ever experienced depression, you probably know the feeling of waking up in a darker mood than when you went to bed, of wanting just to pull the covers over your head and fend off the day.

What if you could send specific signals during sleep to help you wake up refreshed instead? Imagine if rather than taking a pill or facing another night of restless rumination, you don a set of headphones engineered to play music at a targeted time in your sleep cycle to infiltrate your mind with more positive content.

This is what’s possible with Targeted Memory Reactivation. A new paper advocates for developing treatments for trauma and depression based on this method. The way forward looks promising, with a few caveats.

What Is Targeted Memory Reactivation?

TMR, in its basic form, is a simple audio cue paired with a specific piece of information or experience during wakefulness. That same sound is then played softly while the person is asleep. The sleeping brain, hearing its cue, tends to reactivate the neural representation of the associated experience. Memory consolidation, which turns experiences into stable, long-term memories, can be influenced by this reactivation. In this window that gets opened by the resonant sound, something helpful can infiltrate the trauma memory, perhaps shifting its valence.

TMR has been studied for over a decade, and its effects on memory are reasonably well established. In Lewis and Abdellahi’s recent paper, the researchers are suggesting ways this technique can be directed toward therapeutic ends. Specifically they are proposing ways to shape emotional memories during sleep to reduce the suffering of people with PTSD and depression.

The authors, from Cardiff University’s Brain Research Imaging Centre, are not making extravagant claims. They describe three distinct, and viable potential mechanisms by which this might work.

Three Clinical Possibilities

The first is the most straightforward: using TMR during non-REM (NREM) sleep to consolidate the gains made in a waking therapy session. Van der Heijden and colleagues (2024) used this approach to augment eye movement desensitization and reprocessing (EMDR) therapy in patients with PTSD, with improved clinical outcomes. A separate study by Schwartz and colleagues applied the same principle to imagery rehearsal therapy for nightmares, with similar promise. Sleep is where therapy is metabolised into lasting change, and TMR may accelerate that process.

The second possible treatment avenue addresses something clinicians who work with depression will recognise immediately: the phenomenon of waking up feeling worse than you went to sleep. Lewis and Abdellahi suggest depression may involve excessive reactivation of negative memories during NREM sleep, a kind of nocturnal rumination that renews habitually negative thought patterns overnight. Their suggested intervention is to fill that same consolidation window with positive memory cues instead. Because the neural resources recruited by TMR are limited, reactivating positive memories may effectively crowd out the negative ones.

The third possibility is the one I find most clinically compelling, and it deserves more than a paragraph.

Exposure Without Fear: The Promise of REM Sleep

During REM sleep, norepinephrine, the neurotransmitter most closely associated with physiological arousal and the chemical substrate of the fear response, is almost entirely absent. The body cannot mount an alarm reaction. It is, neurochemically, one of the safest possible states in which to encounter a threatening memory.

This is why, in dramatic dreams about things like massive disasters or crazy chases, the dreamer often reports feeling surprisingly calm. There is chaos all around, but the dreamer is unruffled - the perfect avenue for a therapeutic experience.

Lewis and Abdellahi propose that negative TMR during REM sleep, playing a cue associated with a distressing memory during this norepinephrine-quiet state, could allow the brain to reprocess that memory without the accompanying terror. In their words, the memory is retrieved as if during a safe therapy session. The negative emotional charge gradually decouples from the core memory content. The memory does not disappear, but its power to destabilise diminishes.

Dreams as Nocturnal Therapy

This idea has a distinguished and relevant antecedent in the work of psychiatrist and dream researcher Ernest Hartmann, whose thinking I keep returning to when I consider what dreaming actually does. Hartmann argued that one of the primary functions of dreaming is to take emotionally charged experiences, particularly disturbing or unresolved ones, and process them within a state that is inherently safe.

The dreaming mind, Hartmann proposed, re-experiences difficult material, but does so in a context that is neurologically and somatically removed from danger. The body is still, the threat is not present, and the arousal systems that would otherwise flood the experience with fear are quieted. In this protected state, the dream can make new connections around the difficult material. Hartmann suggests that dreaming weaves nightmare material into a broader emotional context, softening its edges. A curious exploration of the dream often yields unexpected associations that shift its meaning. Hartmann described this as ‘contextualising’ emotion, and he saw it as something essential work that dreaming does naturally.

What Hartmann identified in dreaming is precisely what trauma therapy tries to recreate in the consulting room: re-exposure to difficult material within a felt sense of safety, so that the body can revise its response to that material. The difference is that dreaming offers this possibility every night, for free, to anyone who sleeps. When it works, we don’t notice it; we simply wake up carrying a little less distress than we did the night before.

A Parallel from Anaesthetic Research

Pilleriin Sikka, a postdoctoral researcher at Stanford, has been exploring something that illuminates this same territory from a different angle. She has been studying the quality of dreams that occur under propofol anaesthesia, and what she has found is that anaesthesia reliably produces pleasant, low-arousal dreaming, likely because the patient is profoundly physiologically relaxed, and that this state can, under the right conditions, be used therapeutically.

One participant in her pilot study had lost her son to suicide on railway tracks and had suffered ever since with PTSD and recurring nightmares of the incident. Following a brief imagery rehearsal session and imaginal exposure work, she was given anaesthesia and later dreamed of walking alongside those same tracks, calmly, in ordinary conversation with her son. The nightmare’s grip broke, her PTSD symptoms reduced, and she was able, for the first time, to access memories of him that were not entirely coloured by the circumstances of his death.

There is a parallel with what Hartmann described. Propofol, like REM sleep, appears to suppress the arousal system sufficiently that a feared memory can be visited without the body’s alarm response overwhelming everything else. The dream can do its contextualising work. The patient finds themselves in the setting of their trauma, but held, in some deep physiological sense, in safety. This is what Hartmann meant when he described dreaming as re-experiencing difficult memories while in a safe place.

It is also, at its core, what happens in good clinical work with nightmares. Imagery rehearsal therapy or other methods of dream re-entry and rescripting can create conditions of safety and support that shifts the embodied perception of the threat in the dream - it takes the charge out of the memory. The dream changes because something underneath it has changed. Dream engineering may be a way to do the same thing, potentially with greater precision and less waking-world distress.

The Cautions Are Real

None of this is without risk, and the research to date leaves many unaswered questions. Most TMR research has been conducted in single-night laboratory sessions. The recent advent of wearable devices that can deliver TMR at home is beginning to change this, but the implications of repeated, multi-night stimulation are not yet understood. A single night of TMR has been shown to produce changes in brain structure and function that continue unfolding for at least ten days, and the potential cumulative effects of applying these techniques across weeks or months remain an open question.

There is also the problem of error. TMR does not always work as intended. It can, under some conditions, strengthen a negative memory rather than weaken it, and in people with PTSD, poorly calibrated cues could trigger nightmares and retraumatisation rather than relief. If consolidation is selectively boosted for some memories and not others over extended periods, the result could be a distorted internal model, warped expectations and skewed interpretations that the person may not even be aware of. The authors raise the possibility that this kind of disruption might induce psychosis in those vulnerable to it.

Underneath all of this sits a broader concern. Dreaming evolved for a reason, and we do not yet fully understand all that is does. The nature of consciousness and the intricate ways the dreaming mind serves the waking one remain mysterious. When we intervene in that process, even gently and with the best of intentions, we risk disturbing something that was already working, and even undo what dreaming was quietly repairing. These are not reasons to abandon this research, but they are reasons to conduct it with care.

A Middle Way Worth Taking

The proposed research direction offers help where it is most needed. The proposal is modest and specific in its aims: to interrupt a vicious cycle, in people already suffering, at the point where that cycle is most physiologically tractable.

Depression feeds itself overnight, and nightmares can maintain the architecture of fear. If a gentle sound in the dark nudges the sleeping brain toward something less corrosive, and if that nudge reduces the burden that the next morning carries forward, then we have done something genuinely useful.

I will keep watching the progress of this research as it’s hopeful. I hope the research continues to support this approach so we clinicians can have a gentle, drug-free tool to reduce nightmares that sustain trauma and rumination that fuels depression. At the same time, we should remain humble about the boundaries of what we understand.

REFERENCES

Hartmann, E. (2011). The Nature and Functions of Dreaming. Oxford University Press.

Lewis, P. A. & Abdellahi, M.E.A (2026). Could sleep engineering be used to combat PTSD and depression? PLOS Biology. DOI: 10.1371/journal.pbio.3003633

van der Heijden, A. C., van der Werf, Y. D., van den Heuvel, O. A., Talamini, L. M., & van Marle, H. J. (2024). Targeted memory reactivation to augment treatment in post-traumatic stress disorder. Current Biology, 34(16