aim & hypothesis
The aim is to examine whether music listening is effective in improving sleep quality, reducing symptoms of depression and improving quality of life in patients with depression related insomnia.
The following hypothesis is tested:
The controlled use of a sound pillow (figure 1) in combination with the Music Star app (figure 2) can serve as a sleep aid in reducing depression related sleep disturbances, reducing symptoms of depression and improving quality of life.
Music listening, music intervention and music medicine refer to the use of music as an intervention without an active music therapist. In this protocol, music listening implies listening to pre-recorded music.
Insomnia, sleep disturbances and sleeping problems refer to a broad understanding of the disorder. Symptoms include difficulties initiating and maintaining sleep.
The study is named MUsic STAr For Insomnia (MUSTAFI).
A single center randomized controlled trial in a two arm parallel group design is conducted from May 2018 to December 2019 following the revised Consolidated Standards of Reporting Trials (CONSORT) guidelines . The RCT includes two groups of participants, an experimental group and a waitlist control group. Both groups receive standard treatment for depression following national guidelines.
The duration of the intervention is 4 weeks. The participants are followed for 8 weeks. Participation takes place in the home of the patient. In addition, three scheduled visits to the hospital in relation to the research project are required. Baseline assessment is performed on the starting day. Follow up measurements are performed after 4 and 8 weeks (figure 3).
The black element in the left bottom of figure 3 illustrates the intervention given as an option to the waitlist control group. The participants may take home the sound equipment for four weeks. There is no data collection involved. This has been added to the study for ethical reasons and to limit drop out from the control group . The schedule for the trial is described in the SPIRIT flow chart (figure 4) and the dimensions of the study protocol has been described adhering to the SPIRIT Checklist (Additional File 1).
Based on a previous study, we assume that the mean decrease in Pittsburgh Sleep Quality Index (PSQI) scale is 3.04 (SD 2) points for the music intervention group and 2.04 (SD 1.67) points for the waitlist control group after four weeks follow-up . When estimating the sample size, we assume that 25% of the eligible patients will be willing to participate in the project. Based on the patient flow in the outpatient clinic and the available resources from the staff recruiting the patients, we assume that a total of 120 participants can be recruited over a period of 18 months. A dropout rate of 20% is expected. Hence, a total of 100 randomized patients are expected to complete the study. Based on the Satterwaite’s two-sample t test and with a total of 100 patients (n=50 per group) the power of our study is estimated to be 76% (at a confidence level of 5%).
Setting and Participants
The study will be conducted at a single study site: the Unit for Depression in Psychiatry at Aalborg University Hospital, Denmark. The participants are adult outpatients aged 18-65 years in treatment for unipolar depression in the Northern part of Denmark.
Patients are referred to the unit for depression with different backgrounds. Some patients have not previously been associated with psychiatry and are referred directly from private practice with moderate or severe symptoms of depression. Another group of patients are referred from an inpatient unit. These patients are in recovery and have a stabilized condition in need of follow up treatment after discharge. A third group of patients has been diagnosed in previous contact or hospitalization and in need for treatment after a recurrent depression.
All patients are required to have a diagnosis of unipolar depression (ICD-10 Depressive singular episode F32 or Periodic depression F33) and sleeping problems identified by the Hamilton Depression Rating Scale (HAM-D) by a score of at least 2 of a single item or at least 3 in total for the three items (sleep items 4-6). In addition, the patients should receive treatment according to national guidelines for depression. This may include pharmacological treatment, psychotherapy, psycho education and electro convulsive therapy. Participants are eligible if within age 18-65, in stabilized pharmacological treatment and having at least four weeks of treatment.
Exclusion criteria are unipolar depression with psychotic episodes, substance or alcohol abuse or sentence to treatment by law. Patients will be excluded if they have restless legs syndrome, obstructive sleep apnea or other organic sleep disorders as well as hearing loss. In addition, patients with a dislike of music will be excluded.
Patients are recruited in the Unit for Depression. Nurses, psychiatrists and psychologists use the screening tool 4 weeks after beginning of treatment or after medical treatment is stabilized. Participants are recruited through posters, information by doctors and word of mouth in the unit. After an initial assessment for eligibility, the patient is informed of the option to participate. The recruitment takes place during group meetings or during individual consultation in the unit. When the patient has declared an interest to participate, the psychiatrist of the patient fill out an inclusion document based on information from the patient journal and his/her knowledge of the patient. The document is forwarded to the research leader. If the patient is eligible, the music therapist makes an appointment to give detailed information on participation in the research project.
Patients are included according to the inclusion and exclusion criteria. The patient may be discharged during participation in the research project and finalize as a private person. If the increase in depression symptoms result in hospitalization, the patient may continue participation in agreement with caretakers in the unit. If the patient is changing medication but is interested in participation, the patient will be registered on a waitlist for inclusion.
The patient may sign the informed consent after receiving oral and written information about the research project. The information explains the aim and the procedures of the research project, the use of sound equipment and accelerometer and the rights as a participant in a scientific health study. The patient has the option to ask questions and if the patient need time to consider the participation, the music therapist offer to call the patient after one week and a new meeting is set up for the signing of the informed consent. When the informed consent is obtained and the inclusion criteria are confirmed in the inclusion document, the music therapist proceeds with the randomization. An additional consent for collection and use of log file data from accelerometer and the Music Star app is signed after randomization. Documents of informed consent are kept in a securely locked place.
Randomization takes place when the inclusion document has confirmed eligibility and the informed consent has been signed following the regulations of the The North Denmark Region Committee on Health Research Ethics Randomization is carried out by the use of REDCap (Research Electronic Data Capture) hosted at Aalborg University Hospital. REDCap is a secure web-based application designed to support data capture for research studies. The randomization will be stratified according to age, i.e., under 30 years vs. 30 years and above. Randomization is performed using computer-generated block randomization (random block sizes 2-8) with 1:1 allocation between the intervention group and the waitlist control group.
The principal investigator is assigned to give information on the research project to groups and individuals by team members and coming to the unit by appointment. When offering a thorough information to obtain informed consent, the music therapist takes the participant to the music therapy clinic situated in a hospital building close by.
The music intervention consists of listening to music with the use of a sound pillow applying the music from The Music Star app with special designed playlists. The duration of music listening for 30 minutes minimum at bedtime is guided by previous studies . The principal investigator HNL gives an oral guideline to music listening and use of sound equipment at the baseline appointment including how to adjust the volume. The pillow speakers allow audio listening at very low volume. Adherence to the intervention (30 min of bedtime music) is monitored through log file data from the Music Star.
Instructions concerning the experimental intervention:
- It is required that you listen to music of your choice from the Music Star at least 30 min. at bedtime every night in four weeks. You may use The Music Star during the night or in the early morning hours if you wake up and have difficulties falling asleep again.
- You may turn off the Music Star if you wake up when the music plays and it is disturbing.
- The music will automatically stop when the playlist ends after 30 or 60 minutes
- You may use The Music Star in the daytime to assist a rest. It is important only to use The Music Star for relaxation and sleep.
- You are advised to listen to the different available music to select your preferred playlists. You may select a particular playlist to play every night. Some people find that it is helpful to let a specific piece of music be a signal for sleep initiation.
- Consider choosing music that you like and that may help you fall asleep.
- You have to listen to music at night in the way it is described for the research purpose, i.e. the sound pillow in combination with The Music Star app.
- You may listen to your own music during the daytime.
Instructions on the use of the sound equipment:
- Place the sound pillow in your bed (remove your own pillow) and place The Music Star close to the bed on a small table or chair for easy access during the night.
- Make sure that you have recharged The Music Star during the day.
- If you have technical problems with the sound equipment, you may contact the research leader by phone from 9-12 on weekdays.
Waitlist control group.
The participants in the waitlist control group are asked to continue their normal bedtime routines. They are not given any instructions concerning music listening during the eight weeks. At the baseline appointment, they are informed of the option to take home the music equipment to test for four weeks at the end of the eight week period.
The data are derived from four questionnaires (three are self-reported and one is scored by the psychiatrist, researcher or project nurse) and log file data from the accelerometer and from The Music Star app. The references concerning validation of each instrument below are selected according to their relevance to this study.
The Pittsburgh Sleep Quality Index (PSQI) is a commonly used questionnaire measuring self-reported sleep habits in clinical populations in research and clinical practice  . A Danish version of the questionnaire will be used . The items refer to sleep habits and disturbances within the last month. The 19 items are divided in 7 domains: subjective sleep quality, sleep latency, sleep duration, sleep efficiency, sleep disturbances, daytime dysfunctions and use of antidepressant agents. The seven components form a global score ranging from 0 to 21, each component with a range from 0 to 3. Buysse reported a score of >5 (indicating poor sleep) yielded a diagnostic sensitivity of 89.6 % and a specificity of 86.5% . The scale shows good homogeneity with an internal consistency, with α = 0.83. Acceptable measures of validity were obtained through the ability to distinguish between clinically distinct groups and comparing these with polysomnographic results . A score of > 5 is indicative of severe sleep difficulties in at least two areas, and separating participants in two categories “good” and “poor”. PSQI is validated in psychiatric populations  .
The Hamilton Depression Scale (HAM-D17). The HAM-D17 consists of 17 items . The items cover the depressive state, the unspecific stress and arousal symptoms, the suicidal thoughts and lack of insight. The respondent is asked to consider the last three days when responding.
Each item in the score is rated from 0-4 or 0-2, the higher number indication an increase in the symptoms of depression. A guide to the Hamilton rating questioning is used in combination with a rating sheet. The total score indicates the severity of depression symptoms graduating in four categories from unlikely depression to severe depression, with 8-12 indication minimal depression, 13-17 indicating light depression, 18-24 indicating moderate depression and 25-52 indicating severe depression.
The Hamilton Depression Rating Scale in the 17-item version is validated .
The World Health Organization Well-Being Index. (WHO-5). The 5-item World Health Organization Well-Being Index (WHO-5) is among the most widely used questionnaires assessing subjective psychological well-being . It measures subjective quality of life based on positive mood (good spirits, relaxation), vitality (being active and waking up fresh and rested), and general interest (being interested in things). WHO-5 only contains positively phrased items. The WHO-5 items are: (1) ‘I have felt cheerful and in good spirits', (2) ‘I have felt calm and relaxed', (3) ‘I have felt active and vigorous', (4) ‘I woke up feeling fresh and rested' and (5) ‘My daily life has been filled with things that interest me'. The respondent is asked to rate how well each of the five statements applies to him or her when considering the last 14 days. Each of the five items is scored from five (all of the time) to zero (none of the time). The score therefore theoretically ranges from zero (absence of well-being) to 25 (maximal well-being). The use of WHO-5 as a measure of severity of depression is validated [24).
The World Health Organization Quality of Life (WHOQOL-Bref). In 1991, the World Health Organization initiated a project with the aim of developing an international, cross-culturally comparable QOL assessment instrument. “It assesses the individual's perceptions in the context of their culture and value systems, and their personal goals, standards and concerns. The WHOQOL instruments were developed collaboratively in a number of centres worldwide, and have been widely field-tested.” . The WHOQOL-BREF instrument is a shorter version of the original 100 item self-report questionnaire, comprising 26 items measuring four domains: physical health, psychological health, social relationships and environment. In each question, the respondent reports his/her QOL in the four domains on a five point Likert scale. A high score reflects the subjective experience of high QOL. The use of WHO-QOL Bref is validated .
The Music Star app. An app designed for iPad in 2014. The Music Star is a self-explanatory user interface to select music from specifically designed playlists. Music therapists have selected the music for the playlists. The app has a built-in log function that registers each event when using the app. The Music Star gives information from log files of music played, time and duration .
The 16 playlists of 30-60 min duration are represented by colored triangles forming a star. The four playlists in shades of blue are the most quiet and simple (lowest stimuli), the four playlists in shades of green include some variation and dynamics (moderate stimuli), the four playlists in shades of red have more intensity (highest stimuli). The grey triangles are intended to contain playlists for specific purposes. A grey triangle contains the only non-music playlist of ‘Summer Rain’ consisting of sounds of rain and bird sounds. The rain sound adds a non-music track for the purpose of variety in sound stimuli to meet individual preferences. Two other grey triangles contain special designed playlists for sleep. The classification of music is correlated to the degree of complexity and tension of the musical stimulus based on the taxonomy of music by Wärje & Bonde . The playlists in The Music Star are all categorized in subdivisions of supportive music according to the taxonomy of music. The two sleep playlists have a 30 minutes duration. Sleep 1 is a short version of a playlist with music specially composed for relaxation and sleep including a male voice humming. Sleep 2 is a playlist with a guitar-bass jazz duo playing slow ballads. Sleep 1 and 2 are included for meeting criteria of minimum duration and variety in the music selection. More than 100 music pieces in different genres are available in the playlists.
Accelerometer. One tri-axial accelerometer with on board memory (Axivity Ax3, United Kingdom) placed in a wrist bracelet will be used to measure arm movement during the night. The measurement range was set to ±8G with a sampling frequency of 25 Hz. A sleep analysis function using a generic algorithm will provide data on sleep duration. Assessing sleep duration with accelerometer is validated .
Primary outcome measure
The primary endpoint is defined as change of subjective sleep quality from baseline at four weeks. Treatment benefit is achieved when PSQI score is decreased with statistical significant change after the four week intervention period.
Secondary outcome measures
The secondary endpoints are defined and listed below:
- Subjective sleep quality (PSQI): Change of sleep quality from baseline at eight weeks
- Objective sleep quality (Actigraphy): Change of sleep quality from baseline at four weeks
- Objective sleep quality (Actigraphy): Change of sleep quality from baseline at eight weeks
- Symptoms of depression (HAM-D17): Change of depression level from baseline at four weeks
- Symptoms of depression (HAM-D17): Change of depression level from baseline at eight weeks
- Quality of life (WHO-5): Change of quality of life from baseline at four weeks
- Quality of life (WHO-5): Change of quality of life from baseline at eight weeks
- Quality of life (WHOQOL-BREF): Change of quality of life from baseline at four weeks
- Quality of life (WHOQOL-BREF): Change of quality of life from baseline at eight weeks
All outcome measures are collected at baseline, after four weeks experimental/wait-list control and again after four weeks follow up. In addition, registration of pharmacological treatment during the period of participation is included in the data collection.
At baseline, demographic information including age, sex, ethnicity, and partnership, children living at home, handedness, diagnosis code for present depression, first diagnosis of depression and first psychiatric diagnosis will be gathered. In addition, the participant may add self-reported other disease and/or diagnosis. Work status and education, alcohol use and habitual use of music for sleep will be registered.
We initially perform a descriptive analysis on age, gender, medication, illness duration etc. The two groups are compared within groups and between groups using F-tests and chi-squared tests. Additionally we compare the groups on outcome measures at baseline.
Data analysis for the RCT will be performed using mixed effects linear regression with subject specific random effects. The intervention group will be initially compared with the control group on change of all outcome measures after four and eight weeks. Age and gender will be added as covariates in all regression analyses. Repeated-measures analysis of variance (RM-ANOVA) will be used to compare results from baseline follow up one and follow up two.
Finally, a descriptive analysis of the playlists is included in the study, i.e. analysis of playlist preferences (style, duration, time of day) in the group of patients benefitting most from treatment compared to the group of patients benefitting the least.
Blinding of the researchers performing the scorings before and after the treatment as well as the subjects of the study will not be possible. Patients will be anonymized when analyzing data. All main analyses will be carried out based on the intention-to-treat principle. Information on adherence is obtained from The Music Star log file data. Compliance rate will be part of the primary study publication.
In case drop-out rate in the two treatment arms are associated with exposure, we will conduct a secondary analysis in which we adjust for appropriate risk of drop out.
Data collection and management
Data collection consists of several elements. The PSQI, WHO-5 and WHOQOL-BREF are questionnaires filled out exclusively by the participants. The HAM-D rating is performed by the researcher (first author). Data management is carried out by the use of REDCap. Double data entry is performed by research assistants.
The study is approved by the local ethics committee (N-20170055) and registered by the Danish Data Protection Agency (ID 2017-236). A yearly report of adverse events and other unintended effects of the trial is obtained by the local ethics committee.
Results of the study are published in open-access peer reviewed scientific journals and
presented at international scientific conferences regardless of the conclusion. Broadcast of
findings in Danish radio and TV media will be planned through the study.