This protocol for a randomized controlled trial is reported in accordance with the SPIRIT statement [45, 46].

Study setting

The trial is mono-centered and is conducted at our outpatient clinic for acupuncture and Chinese Medicine at the Institute of General Practice and Interprofessional Care, University Hospital Tübingen, Germany.

Inclusion criteria for participants

Patients eligible for inclusion into the trial must be aged either 70 years or older with geriatric typical multimorbidity, or aged over 80 years. Furthermore, they need to suffer from sleeping disorders, measured by a Pittsburgh Sleep Quality Index (PSQI) global score of five or more [47]. Additionally, their German language skills and hearing ability have to be sufficient (simple ear examination).

Exclusion criteria for participants

Patients are excluded if their health does not allow to take part, e.g. with cognitive impairment or dementia (MoCA Score ≤ 18), hemorrhages that required treatment within the last 6 months or coagulopathy of clinical relevance like hemophilia, acute mental disorders (e.g. imminent suicidal tendency or psychosis), severe disease (e.g. end of life care in case of cancer), acute or severe dermatosis (e.g. infectious erysipelas), contact allergies (e.g. nickel, chrome, silicone), needle phobia, no consent of participation.

Eligibility criteria for therapists

The acupuncture intervention was developed based on internal evidence (clinical expertise of the study team) building on previous studies [48, 49], traditional teachings and acupuncture theory. The acupuncture sessions are provided by trained acupuncturists: medical doctors with specialization in acupuncture or by a medical doctoral student with comparable training and qualification (at least ≥ 200 h of training) under supervision. Music therapy is performed by qualified medical personnel with training in monochord playing. All therapists must be trained in both interventions to be able to apply the multimodal intervention.

Interventions

We conduct a confirmatory non-blinded randomized controlled trial with a two-factorial design. This leads to four groups. Table 1 shows the factorial (two between-patient factors) design.

Table 1 Study design of ELAMUS

The interventions are reported according to the STRICTA guidelines [50]. All participants receive printed, evidence-based materials on sleep disorders during their initial visit to the study site. the materials are specifically designed for elderly people, e.g. in a bigger font.

Intervention 1 (Treatment as Usual plus evidence-based information, TAUi

For intervention 1, participants will not get further treatment during the study period. However, they will receive five complimentary treatment sessions after completion of their study participation.

Intervention 2 (Acupuncture, AC)

In addition to TAUi, participants of this group will receive eight acupuncture treatments over 12 weeks. A single session will take around 30 min during which the patient is lying supine on a treatment bed. The acupuncture intervention follows a semi-standardized point selection. It encompasses three fixed acupoints (Shaohai H3 少海, Anmian HN 54 安眠, Shenting GV 24 神庭) and up to four points, which are individually chosen upon the guiding symptoms of the patients. In total, a maximum of 12 needles are used, chosen by the therapist as an individualized tailored TCM intervention. For all acupuncture sessions, sterile, silicone coated, single-use filiform acupuncture needles, with a length of 25 mm and a diameter of 0.25 mm each, are used. The manufacturer brand was not defined. Angle and depth of insertion is point specific according to acupuncture theory [51]. A de qi response is sought for the majority of the points but is not necessary, no further stimulation is applied during the needle retention time of 20 min. No other stimulation methods, such as moxibustion or electroacupuncture, are used. The well-being of the patient over the session time is observed by the therapist.

Intervention 3 (Music therapy, MU)

For intervention 3, participants will receive eight individual live receptive music therapy sessions over 12 weeks, in addition to TAUi. Each session lasts around 30 min with the patient lying on a treatment couch. For the music therapy, the patients receive a 15–20 min exposure to a I Ging monochord®, manufactured by Musikwelt Anklang, Germany [37]. The monochord is placed in an oblique row at the feet side of the treatment couch, approximately in a 1.5 m distance to the patient. The well-being of the patient over the session time is observed by the therapist, in case of agitation the monochord play can be adjusted to just one side of the instrument (personalization).

Intervention 4 (Multimodal therapy approach, AM)

Intervention 4 combines intervention 1, 2 and 3 (TAUi, AC and MU). Participants therefore receive both treatments simultaneously as described above, within 8 session over 12 weeks, each session lasting around 30 min. After placing the needles (see intervention 2), the therapist then starts playing the monochord (see Intervention 3).

For all intervention groups, reminders via telephone or e-mail are used in case a participant tends to miss their pre-scheduled appointments. Due to the health care services setting, there are no restrictions or prohibitions in concomitant care. However, participants are asked to report changes in routines or medications. Patients may withdraw from the study at any time without giving a reason.

Outcomes

The outcomes will be measured three times, during a baseline measurement (T0), after completion of the intervention period of 12 weeks (T1, primary confirmatory analysis), and after the 12 weeks follow-up (T2).

The primary outcome of this study is improvement in sleep quality assessed via the Pittsburgh Sleep Quality Index (PSQI) global score [47, 52, 53] at T1. The PSQI is widely used in research and clinical practice with geriatric patients and consists of 19 items across seven components that form a global score [47, 54]. In addition, the following parameters will be collected as secondary outcomes via the following measurement tools: Depression (Geriatric Depression Scale, GDS-15) [55], Health-Related Quality of Life (SF-12) [56], Neurovegetative activity (Heart Rate Variability Measurement, HRV). Medication is assessed by medication plan or patient self-report. In addition, an accompanying process evaluation is carried out. A qualitative analysis of approximately n = 9–12 interviews with patients of the acupuncture and music therapy groups (n = 3–4 each for AC, MU, AM) and n = 2 interviews with the providers will be conducted [57]. The aim of the qualitative interviews is to gain insights into the individual experiences of the patients and sleep quality. Additionally, bodily sensations are assessed by drawings on a body outline for n = 20–25 patient cases (n = 7–8 each for AC, MU, AM) and the Phenomenology of Consciousness Inventory (PCI) [58]. The sample of patients for the process evaluation are chosen according to a purposeful sampling approach by the providers [59]. Tables 2 and 3 show a summary of the study’s outcomes.

Table 2 Outcome measurements (quantitative)Table 3 Outcome measurements (process evaluation)

Figure 1 shows the participants’ timeline over the course of the study.

Fig. 1

Participant timeline. PSQI: Pittsburgh Sleep Quality Index, MoCa: Montreal Cognitive Assessment, GDS-15: Geriatric Depression Scale, SF-12: Short-Form-Health-Survey, HRV: Heart Rate Variability

Sample size

The study is designed to detect a minimal clinically important difference of 3 points on the PSQI global score [47] between groups with sufficient power. With an estimated standard deviation of 3.5 points on the PSQI global score, a minimal power of 80% and a dropout rate of up to 20%, mixed analysis of variance (mixed ANOVA) requires an estimated number of n = 100 participants, i.e. n = 25 participants per group.

Recruitment

Participants are recruited via different channels: a press release was sent to local newspapers and radio channels. Additionally, printed flyers were handed out at local meeting points, events and courses for the elderly. Furthermore, the related local family practices of our Institute were asked to present the flyer to eligible patients and to display it in their practice.

Allocation

The participants are allocated to the groups by using block randomization with randomly varying block lengths of 4, 8 or 12. The randomization list is set out by HC who is not involved in patient recruitment or data collection by using R software. The list is password protected, and only the biometrician has access to it. After patients are enrolled by KR, written consent has been received and the initial data collection has been completed, participants are centrally randomized by HC. The allocation is received via e-mail and communicated to the patients face-to-face (groups AC, MU, AM) or telephone (group TAUi).

Data collection, management, and analysis

Data for the patient-reported outcomes will be collected pseudonymized via paper and pencil questionnaires, filled out by the patients, at the three measurement time points. Adverse events will be documented at each measurement and/or intervention session. The neurovegetative activity measurements will be started after the questionnaires have been filled out: the therapist starts the recording after placing the heart rate variability recorder (CardioScout Multi ECG System, SR-Medizinelektronik, Germany) at the patients’ chest. The patients themselves take off the device at home after 24 h and send the device back (pre-paid envelope). Additionally, they are asked to fill out an activity protocol (e.g. sports, medication, etc.) during these 24 h at home and send it back with the device. The paper-based data will be stored for ten years in a secured steel locker only accessible by the study administrator.

All paper-based data will be typed into the study’s server-based database system (REDCap) by the study team and all study data are stored for ten years, using REDCap electronic data capture tools hosted at the University Hospital Tübingen [60, 61].

Via a secured VPN server, the pseudonymized data stored on the HRV devices are sent to JVG, who processes the data, e.g. data preparation for data analyses. These are then sent back via the secured VPN sever and statistically processed, analyzed and interpreted by JW, JVL, CK and KR. A Data Monitoring Committee is not set up and not required.

All outcome measures will be analyzed on the basis of the intention-to-treat principle. This means that all randomized participants are included in the analysis in the groups to which they were originally assigned, regardless of whether a complete data set is available or whether the study is conducted according to the protocol. Missing values will be multiply imputed.

Data of primary outcome PSQI and secondary outcomes GDS-15, SF-12 will be analyzed via analysis of variance (mixed between-within ANOVA). After this Omnibus testing, post-hoc t-tests for paired samples will be conducted, if the ANOVA reveals significant results. Bonferroni correction will be applied to correct for multiple testing. All variables (sociodemographic data, PSQI, GDS-15, SF-12) will be visualized and proportions will be descriptively analyzed. HRV data analysis will employ an exploratory approach due to the unknown effects of acupuncture and music therapy on the HRV of geriatric participants. This analysis will cover both frequency domain (e.g., HF and LF power) and time domain measures (e.g., RMSSD, SDNN, pNN50), which have shown to be important parameter with regard to sleep and HRV [62]. We will perform both within- and between-subject comparisons across the three measurement points and among the four intervention groups. Additionally, sensitivity analyses will be carried out to assess the robustness of our findings. Qualitative data will be analyzed exploratory using qualitative content analysis according to Kuckartz [63]. Medication and bodily sensations will be analyzed exploratory.