The recruitment of participants took place between June 2022 and January 2023. Recruitment for the patient group took place in the psychotherapeutic practice of Thomas Valentin in Duesseldorf-Oberkassel and the pain unit of St. Marienhospital in Marsberg. The participants of the patient group (n = 41; 30 women and 11 men) were all currently undergoing treatment for chronic pain disorder. There were 60 different pain diagnoses according to ICD-10 for n = 41 pain patients, with F45.41 (chronic pain disorder with somatic and psychological factors, n = 31) being assigned the most. A total of 20 psychological diagnoses according to ICD-10 for n = 13 pain were assigned to chronic pain patients, whereas the most common diagnosis was F33 (recurrent depressive episode, n = 6).

The mean age and standard deviation (SD) of the clinical sample was 53.34 ± 11.91 years. Healthy control group (n = 40; 29 women and 11 men), matched in age and sex, had a mean age and standard deviation of 53.15 ± 11.97 years. There was no significant age difference between the patient and the control group (t(79) = 0.072, p =0.943). The control group was recruited in the personal environment of the experimenters. Most participants were employed (n = 49) or retired (n = 25). Five participants were house spouses, and two were unemployed.

In addition to the validated questionnaires and the dream diary, participants were asked about their age, gender, occupational status, education, and sleep disturbances. Pain-related information such as duration of chronic pain and currently ongoing pain treatment was also requested from the chronic pain patients.

To measure retrospective DRF and nightmare frequency, the MADRE was used [24]. Participants could indicate how often they could remember their dreams in the last months on a 7-point Likert scale including 0 (never) to 6 (almost every morning). Nightmare frequency is measured by an 8‑point Likert scale from 0 (never) to 7 (several times a week). The Mannheim Dream Questionnaire has shown good psychometric properties (retest reliability of DRF and nightmare frequency: r =0.756 and r =0.751) in a German-speaking sample [24].

The NDQ was used to assess nightmare distress [2]. The scale measures general nightmare distress (e.g., “Do nightmares interfere with the quality of your sleep?”), impact of nightmares on sleep (e.g., “Are you ever afraid to fall asleep for fear of having a nightmare?”), and impact on daytime reality perceptions (e.g., “Do your nightmares foretell the future?”). It consists of 13 items, most ranging on a 5-point Likert scale from 1 (never/not at all) to 5 (always/a great deal), with higher scores indicate greater distress. The NDQ has proven to be a suitable instrument in the past, with an internal consistency of α =0.80 [3].

This scale quantifies the number of critical life events. It is based on the assumption that psychosocial stress contributes to the development of diseases [14]. Participants are asked which of the 43 given common life events have occurred in the last year. Predefined stress intensity points, ranging from 11 to 100 points, are assigned to these events, which are then summed up to a total distress score. According to the authors, a score of 300 or more is considered as cutoff for disability in coping with distress, thus, having an increased probability of developing a disease in the subsequent years [14].

To record the dream content, participants kept a 14-day dream diary, which consisted of two parts. The first part asked about sleep duration and dream recall, with the following answers: 0 (no recall), 1 (yes, I dreamed last night but can’t remember), and 2 (yes). If participants recalled at least one dream, they were asked to complete the second part. First, they wrote down the narrative of their dream content. All dream contents of one night were defined as one dream and a maximum of five dream reports had to be written down in order to minimize the perceived pressure on the participants and thus optimize the dream recall frequency. If participants managed to write these five dream reports, only the first part of the dream diary had to be completed for the remaining period. All dreams per participants were aggregated by means. Then, the intensity of positive and negative emotions was recorded on two separate 4‑point scales, as following: 0 (no emotions), 1 (mild), 2 (moderate), and 3 (strong emotions). Moreover, participants indicated whether they perceived their dream as a nightmare and rated the overall intensity of the nightmare on an 11-point Likert scale, ranging from 0 (negligible) to 10 (worst imaginable dream). If the participants experienced pain during the dream, they rated if the pain was identical or similar to their chronic pain in waking-life (1 = yes, 0 = no). In addition, the average and highest pain intensity during the dream was assessed on 11-point Likert scales from 0 (no pain) to 10 (worst imaginable pain).

The dream content analysis was done with a dream manual, based on Hall & Van de Castle [13]. The number of dream elements such as negative emotions (anger, confusion, apprehension, and sadness), sport (active and passive), fortune and misfortune were counted. If at least one of these dream contents was reported in a dream report, this was coded with “1”, if not, it was coded with “0”. Moreover, it was coded if the participants mentioned body sensations related to specific body parts: head (1), arms (2), legs (3), torso (4), and back (5). The coding was done blindly, i.e., the rater had no access to the participants’ data. The interrater reliability (Cohen’s kappa) was between r = 0.562 (confusion) and r = 1.000 (sadness and sports).

After the participants had expressed their interest in the study, they were informed about the procedure and asked to provide written consent. If the participants belonged to the group of chronic pain patients, a pain diagnosis according to ICD-10 [31] was previously made by the practitioners. They were then given a paper and pencil questionnaire pack and a 2-week dream diary, both with a personal identification code and completion instructions. During 14 days, they had to record their sleep and dream recall every morning. The completion of the questionnaire pack could be split over the entire 2 weeks if needed. After these 14 days, the participants returned the dream diary. Participation was completely voluntary, and participants received 24 € as compensation. The study was approved and accepted by the Ethics Committee of Heinrich Heine University Duesseldorf (MA01_2022_01).

All analyses were conducted using SPSS 25.0 (IBM, Armonk, NY USA). The sample size was calculated using G*Power [10] with α =0.05 and a power of 1 − β = 0.85, yielding a sample size of N = 82, to detect medium effect sizes (d = 0.6), which we were able to identify in a previous study [16]. Prior to this, the data from the dream content analysis was aggregated by the mean values per participant. In order to investigate between-group effects between the chronic pain patients and the healthy control group in the life event score, in dream recall, nightmare distress, and pain in dreams, t-tests for independent samples were calculated. To compare the retrospective DRF of the Mannheim Dream Questionnaire with the prospective DRF of the dream diary, a Spearman correlation was conducted. In order to analyze the differences in dream contents (negative emotions, sports, and body sensation) between patients and controls, t-tests for independent samples were conducted.