The nest position: two blankets, a severely dissociative sample, and the limits of one session

The tool slot in a dissociation issue usually arrives carrying another questionnaire. This one arrives carrying two blankets. A group across Munich and Erlangen took a procedure devised by their physiotherapist co-author, put it on a measurement table, and ran electro- and impedance cardiography while patients used it.

How the nest is built, and how the session runs

The procedure is written out well enough to reproduce. The patient lies on her back. Two blankets, each rolled tightly from one corner to the opposite corner, are placed around the body. Small rolled towels or further blankets surround the head and meet the soles of the feet. If the blankets work loose, towels are pushed underneath along the sides of the legs; positioning rolls and stuffing may be added. The stated criterion is that the patient should feel material touching all parts of her outer body borders. The therapist handles only the soft material and does not touch the patient, which is the invasiveness control built into the design. The authors note the position can be adapted to sitting or side-lying, and that patients can build it themselves at home.

The session ran about 90 minutes in a physiotherapy treatment room and was conducted by the first author. Two rounds: in one the nest was in place, in the other the participant lay supine on the same table with nothing around her. Order was allocated 1:1 before the first round. Each round held three baseline phases of 8 minutes with two imagination phases of 2 minutes between them, one a predefined distressing scene and one a predefined comforting scene, both prepared beforehand in keywords covering the associated cognitions, bodily reactions and emotions, following the EMDR manual. Outside the imagination phases participants listened to a neutral non-fiction audiobook chosen from five, to hold attention and reduce spontaneous dissociation during baseline. Distress and comfort were each rated 0 to 10 after every phase.

This was not a mild sample, and the anti-hype reflex should stop before it reaches these patients. Mini-SCID-D mean total score was 13.33 out of 15 against an entry threshold of 10, administered by the first author in validated German translations; PTSD was established on the SCID-PTSD. DES trait dissociation averaged 38.03% against 4.15% in controls, and the Brief Symptom Inventory Global Severity Index 2.11 against 0.20. These women had already spent an average of 5.14 months in psychosomatic and 1.42 months in psychiatric inpatient care. Affective comorbidity ran at 88.9%, and 44.4% were on antidepressants. Benzodiazepines, beta-blockers and antiarrhythmics were exclusion criteria, without which the autonomic readouts would have been uninterpretable.

What survived correction, and what the registry records

The table is the honest document here, and it is the one I followed. In patients, two of five distress comparisons and four of five comfort comparisons survived Holm-Bonferroni. Comfort at the third baseline did not: uncorrected p = 0.009, adjusted p = 0.051, g = 0.34. In the healthy controls, nothing survived correction at all. On physiology, the patients' single uncorrected hit was pre-ejection period at the second baseline (p = 0.004, adjusted p = 0.060, g = 0.19), and the authors state in their own limitations that every previously significant psychophysiological difference falls away after correction. Parasympathetic tone never moved in either group at any phase. Note one internal inconsistency: the abstract reports sympathetic tone decreasing particularly in the healthy controls, while a sentence in the Results says the mixed-model effect of condition on pre-ejection period was more pronounced in the patients. Table 3 shows four uncorrected hits in controls against one in patients, so I went with the abstract and the table.

The trial is DRKS00030669 in the German Clinical Trials Register. The register gives the actual start date as 21 May 2019 and completion as 7 November 2021, and records the entry as made on 4 October 2023, labelled Retrospective by the register itself. That is roughly four years after the study began and nearly two after it ended, and the paper says so plainly in its Method rather than burying it. Registered target sample 36, final sample 36, with no drift. The registered primary outcome was two things: subjective units of distress and parasympathetic drive as RMSSD. Half of that primary was null, and the sympathetic measures carrying the paper's physiological headline are registered as secondary. The register also lists allocation as non-randomised and blinding as none, which does not contradict the title: the randomisation was of condition order within each participant, not of group, since patient versus control was fixed by diagnosis.

Who rated what, and money. Distress and comfort are the participant's own numbers, spoken after each phase to the first author, who had also allocated the order and stayed in the room throughout. There was no blinding on either side, and none was available once the condition is a nest of blankets. Against that the authors set a real precaution: participants were not told the intervention's name or method, nor the hypotheses or the phase structure, and the authors raise demand characteristics themselves. The cardiography is device-recorded and rater-independent, and it is the part that showed nothing after correction. Funding, in its own statement, reads that the research received no specific grant from any funding agency, commercial or not-for-profit sectors, and the register lists only the sponsor's institutional budget. The declaration of interest, separately, reads "None". One interest sits outside that line and deserves naming: the nest position was developed by co-author K. Paschinger, the physiotherapist, so the developer of the procedure is an author on its first evaluation. Harms were looked for and reported: no participant had an immediate aversive reaction, discontinued the session or needed psychological intervention afterwards, and no physical harm was observed.

For your practice

What you can take to a room on Monday is cheap, reproducible and modest. For a severely dissociative woman lying on a treatment table, being enclosed lowered what she said her distress was and raised what she said her comfort was, including while she deliberately held a distressing image. That is a within-session state effect measured by self-report. It is not a treatment result, and the authors do not claim one: they call the nest position potentially promising and additional.

The most useful clinical detail is the complaint, not the effect. Afterwards, some participants described the enclosure as unfamiliar and in some cases unpleasant, precisely because it stopped them getting away from the distressing image they had been asked to hold; the paper gives no count for how many. Containment and entrapment are the same physical arrangement with two meanings, and nothing in this study predicts which one a given patient will get. So state the exit before you build it, say that stop is available at any moment, and check afterwards rather than assuming comfort. Expect the paradox the authors cite from Brand and colleagues: for someone whose regulation has run on unsafe behaviour, safety arriving can itself be distressing.

Where this evidence does not reach is equally clear. It says nothing about the phase question, and cannot be used to argue that stabilisation must precede trauma processing or that it need not. It does not license the nest as a substitute for treatment, and it does not tell you which patients to select, because no moderator was tested. What it gives you is a low-cost, explicitly no-touch procedure with a first piece of measured evidence behind it and a documented downside you can warn the patient about before the first blanket goes down.