The effects of music on the sleep quality of adults with chronic insomnia using evidence from polysomnographic and self-reported analysis: A randomized control trial

International Journal of Nursing Studies 49 (2012) 921–930

Contents lists available at SciVerse ScienceDirect

International Journal of Nursing Studies journal homepage: www.elsevier.com/ijns

The effects of music on the sleep quality of adults with chronic insomnia using evidence from polysomnographic and self-reported analysis: A randomized control trial En-Ting Chang a, Hui-Ling Lai b,c,*, Pin-Wen Chen d, Yuan-Mei Hsieh e, Li-Hua Lee b a

Department of Chest Medicine, Buddhist Tzu Chi General Hospital, Hualien, Taiwan Department of Nursing, Tzu Chi University, Hualien, Taiwan Department of Medical Education, Buddhist Tzu Chi General Hospital, Hualien, Taiwan d National Cheng Kung University, Institute of Art Studies, Taiwan e Department of Music, National University of Tainan, Taiwan b c

A R T I C L E I N F O

A B S T R A C T

Article history: Received 22 July 2011 Received in revised form 19 February 2012 Accepted 24 February 2012

Background: Research-based evidence supports the therapeutic use of music to improve the sleep quality measured by self-reported questionnaires. However, scientific knowledge of the effects of music measured using standard polysomnography in chronic insomnia adults is currently insufficient. Objectives: The objective of this study was to evaluate the effect of soothing music on objective and subjective sleep quality in adults with chronic insomnia. Methods: Fifty participants were enrolled in a randomized controlled trial conducted in the sleep laboratory of a hospital, with 25 participants allocated to the music group and 25 to the control group. For four days, the experimental group was exposed to soothing music selected by the participants or researchers for 45 min at nocturnal sleep time, whereas the control group was not exposed to music. Sleep was measured using polysomnography (PSG) and self-reported questionnaires. A general estimating equation was applied to analyze the data. Results: After controlling for baseline data, the music group had significantly better scores for rested rating (p = 0.01), shortened stage 2 sleep (p = 0.03), and prolonged REM sleep (p = 0.04) compared to the control group, shown by the generalized estimating equations. However, there was no evidence of the effectiveness of music on other sleep parameters as measured by PSG. Additional findings indicate no difference in sleep quality between those who listened to their own preferred music (n = 10) and those who listened to music selected by the researchers (n = 15). Conclusion: The results contribute to knowledge of the effectiveness of music as a therapy to improve sleep quality in adults experiencing insomnia. Listening to soothing music at nocturnal sleep time improved the rested rating scores, shortened stage 2 sleep, and prolonged REM sleep, but has little effect on sleep quality as measured by polysomnography and self-reported questionnaires. ß 2012 Elsevier Ltd. All rights reserved.

Keywords: Music Insomnia Polysomnography Sleep

What is already known about the topic? * Corresponding author at: Department of Nursing, Tzu Chi University, 701 Section 3, Chung Yang Road, Hualien, 970, Taiwan. Tel.: +886 3 8561825x2263; fax: +886 3 8576278. E-mail address: [email protected] (H.-L. Lai). 0020-7489/$ – see front matter ß 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.ijnurstu.2012.02.019

 Insomnia is highly prevalent, and the side effects of conventional pharmacological medication methods are well known.

922

E.-T. Chang et al. / International Journal of Nursing Studies 49 (2012) 921–930

 Listening to soothing music at nocturnal sleep time may improve subjective sleep quality.  The effects of music on sleep quality remain a controversial issue because of inconsistent study results. What this paper adds  Listening to soothing music at nocturnal sleep time improved the self-reported rested rating scores.  Listening to soothing music for 45 min over 3 days improved sleep quality, manifesting in shorter stage 2 sleep and longer REM sleep.  The effects of music on sleep quality did not differ between those who listened to their own preferred music and those who listened to the music selected by the researchers. 1. Introduction The prevalence of insomnia in the general population is estimated to be at 29.9% (Morin et al., 2006). Prevalence rates have varied from as low as 5% to as high as 50% depending on the definitions of insomnia (Ohayon, 2002). Insomnia is recognized as a significant and growing public health concern (Bartlett et al., 2008; Nomura et al., 2010). Recent epidemiological reviews indicate that between 30 and 48% of adults complain of insomnia symptoms (Ohayon, 2002), while close to 10% suffer from insomnia syndrome (Morin et al., 2006). Insomnia is defined as difficulty in initiating and/or maintaining sleep, early morning waking, or non-restorative sleep. Daytime consequences are often associated with the complaint (Schutte-Rodin et al., 2008). The consequences of insomnia are an impaired quality of life, an increase in medical consultations, and hypnotic medication consumption (Leger et al., 1999). Pharmacotherapy is the most commonly prescribed remedy for disturbed sleep. However, pharmacological management can be contraindicated because of the accompanying unwanted side effects, including daytime residual effects, tolerance, dependence, altered sleep stages, and rebound insomnia (DeMartinis et al., 2009). Prior research has not established the safety and efficacies of pharmacotherapy for sleep problems in adults sufficiently (Buscemi et al., 2007). An increase in drug consumption increases the risk of work accidents and absenteeism (Walsh, 2004) and poor self-esteem and lower efficiency (Leger et al., 2006), resulting in decreased productivity (Leger, 2000; Walsh, 2004), daytime sleepiness (Glass et al., 2005), and impaired driving abilities (Verster et al., 2006) at a substantial cost to both the individual and society (Wade, 2011). Direct medical costs of insomnia are estimated to be US$13.9 billion annually in the United States (Martin et al., 2004). Treatment-related costs were approximately US$2 billion of the total direct costs associated with insomnia (Martin et al., 2004; Kryger, 2006). The people involved in these common adverse events were frequently reported to be using pharmacological treatment; therefore, insomnia has become a crucial issue that evidence-based nursing must address.

Sleep scientists recommend testing non-pharmacological methods that promote interaction between the mind and body to support sleep (Kozasa et al., 2010). Listening to music is one of the most frequently used strategies to promote sleep (Morin et al., 2006). Additionally, music is an effective intervention for managing stress in patients (Nilsson, 2011) and improving subjective sleep in older adults (Lai and Good, 2005) and university students (Harmat et al., 2008). Using recorded music as an anxiolytic has found favor in diverse clinical settings and for a wide variety of patients (Nilsson, 2010). Soothing music reduces anxiety (Lai et al., 2008a; Nilsson, 2010, 2011), stress, and cortisol levels (Lai and Li, 2011; Ventura et al., 2012), though the consequences on sleep architecture remain unclear. Therefore, researchers have recommended further studies to verify the effects of music on sleep quality using polysomnography (Lai and Good, 2005; Lazic and Ogilvie, 2007). Recently, a nationwide survey of the population revealed that 40–50% of adults listen to music as a selfhelp strategy to improve their sleep (Furihata et al., 2011). Previous cross-sectional studies have found that the prevalence of listening to music among insomniacs is higher than among non-insomniacs (Morin et al., 2006). Several studies have been conducted to document the efficacy of music on sleep quality. However, they were primarily a small number of randomized controlled trials, had poor methodological quality, and used subjective selfreported measures as the outcome index (Lai and Good, 2005). In recent decades, studies examining the beneficial effects of music on sleep using standard polysomnographic measurements were limited; moreover, research findings were inconsistent between objective and subjective sleep measurements (Lazic and Ogilvie, 2007; Harmat et al., 2008), which is insufficient to prove that soothing music improves insomnia. Therefore, the purpose of this study was to evaluate the effect of soothing music on objective and subjective sleep quality in adults with chronic insomnia. 2. Methods 2.1. Study design A randomized controlled trial was conducted from May 2010 to June 2011 to examine the effects of music on the sleep quality of insomniac adults. 2.2. Participants Participants were invited to volunteer for the study through word-of-mouth, Internet announcements, and flyer advertisements to recruit healthy but insomniac volunteers. On the study enrolment day, subjects underwent a medical check, and written consent was obtained after further information was provided. Power analysis was used to estimate the sample size. To achieve a power of 80% with a two-tailed a of 0.05 and an effect size of 0.35 on self-reported sleep quality (Lai and Good, 2005) and correlation (r = 0.50) among repeated measures (Lai and

E.-T. Chang et al. / International Journal of Nursing Studies 49 (2012) 921–930

Good, 2005; Lai et al., 2006), the total sample size was computed to be 46. Another 10% was added for attrition; therefore, 25 subjects were required for each group. To be eligible to participate in the study, the subjects were required to (1) have experienced insomnia (PSQI > 5 at screening; Buysse et al., 1989) for at least one month, and (2) be between 20 and 60 years of age. The exclusion criteria included participants with (1) psychiatric or neurological problems; (2) a history of alcohol/drug abuse; and (3) pregnant or nursing women. Over a 13-month period, 61 volunteers were contacted, and one was disqualified after the screening interview according to the inclusion criteria. Four respondents refused to sleep in a laboratory, and 6 were too busy to participate for the four consecutive days. The remaining 50 paid volunteers completed the study assessments. 2.3. Experimental intervention The music interventions consisted of listening to sedate music. The participants were encouraged to bring their own preferred music for bedtime listening. However, if they did not have their own bedtime music, they could listen to music prepared by the researchers. A Cochrane review found that the positive effects of music were similar in studies where patients selected the type of music and where patients did not (Cepeda et al., 2006). The music selected by the researchers comprised six musical pieces: Spring Rural Field, Woman under the Moon (Chinese music), Going Home (Czech music), Destiny, Heart Lotus (Taiwanese music), and Memory. The musical pieces are all familiar and have a pleasant sound, which the researchers assumed would increase the subjects’ relaxation when participating in this study. The last piece was composed by the authors of this study (Fig. 2). Previous studies demonstrated that music has beneficial effects on agitation in older adults with dementia (Ho et al., 2011) and on stress reduction in the caregivers of cancer patients (Lai et al., 2011). The length of the musical intervention was approximately 45 min (Lai and Good, 2005). The music selected by the researchers was prerecorded onto a CD. An auto-reverse multi-laser disc player (AZ-1836, Philips, Holland) was used to listen to the music. Participants were prepared for the musical intervention and sleep study by a trained registered nurse and a sleep technician. All the tracks had tempos ranging from 60 to 85 beats/ min (slow), minor tonalities, smooth melodies, and no dramatic change in volume or rhythm to achieve a relaxing effect (Nilsson, 2010). Researchers selected the music to enable participants to achieve a relaxed psychophysiological state based on psychophysiological theory (Lai and Good, 2002). The music selections prepared by the researchers or the participants all had similar musical characteristics and qualities, with no sudden changes in volume and rhythm.

923

accompanied by decreased daytime functioning (American Academy of Sleep Medicine, 2005) that persists for at least four weeks (Schutte-Rodin et al., 2008). A self-administered questionnaire is one of the instruments used in the evaluation and differential diagnosis of insomnia (SchutteRodin et al., 2008). Insomnia was measured by the Pittsburgh Sleep Quality Index (PSQI) (Buysse et al., 1989). The PSQI was administered one week before the night of PSG. The PSQI consisting of 19 self-rating questions is a self-reported questionnaire that measures the sleeping habits during the previous month; these 19 items are grouped into seven component scores. The sum was a global sleep quality score ranging from 0 to 21 to yield a global PSQI score, with higher scores indicating a worse sleep quality. The PSQI > 5 points identified subjects as suffering from poor sleep quality (Buysse et al., 1989), and was used to determine eligibility for participation. Cronbach’s a for the seven components was 0.83, indicating a high degree of internal consistency. The Chinese version PSQI of greater than 5 yielded a sensitivity and specificity of 98 and 55% in primary insomniacs, respectively (Tsai et al., 2005). 2.4.2. Polysomnography (PSG) A sleep laboratory in a hospital was used to collect data. The sleep laboratory comprised three single bedrooms and a monitoring room. Two-way communication was available via microphones and speakers placed in each room. All the participants slept in dim, sound-attenuated, airconditioned, temperature- and light-controlled single bedrooms. The temperature in the bedrooms ranged from 24 8C to 27 8C depending on the participants’ requests. The background noise levels in the sleep laboratory during the nights were measured at approximately 30–35 dB by decibel meter (CRL252, Cirrus, Taiwan). PSG was performed by a laboratory-based PSG (Embla, Denver, CO, USA) using electroencephalography (EEG; including O1-A2, O2-A1, C3-A2, and C4-A1), electromyography (EMG; submental), and right and left electrooculography (EOG) during the four nights. PSG included a standard 16-channel montage for scoring sleep parameters. The international 10–20 system was applied to mark the standard electrode sites. Sleep recordings were visually scored in 30 s intervals by a qualified technician according to standard criteria (Iber et al., 2007). The records were scored blindly, without the knowledge of the participants’ data (such as age, assigned group and PSQI score). PSG data were analyzed using the Somnologica Studio 3.3.2 software package (Somnologica, Flaga hf, Medical Devices, Iceland). The following objective clinical sleep measures were defined as minutes of total sleep time (TST), sleep efficiency (SE; %; total sleep time/total recording time), sleep onset latency (SOL), wake after sleep onset (WASO), number of awakenings, percentage of time in stages 1, 2, 3, 4, and rapid eye movement (REM) sleep, and arousal index (number of electroencephalographic arousals per hour of sleep).

2.4. Variables and measures 2.4.1. Chronic insomnia In this study, chronic insomnia is defined as difficulty initiating/maintaining sleep and/or non-restorative sleep

2.4.3. Morning questionnaire During the morning participants awoke in the sleep laboratory, they completed a sleep questionnaire in which they recorded their bedtime, total sleep time, time taken to

924

E.-T. Chang et al. / International Journal of Nursing Studies 49 (2012) 921–930

fall asleep, number of awakenings, time spent awake during the night, wake up time, and a subjective evaluation of feeling rested. Subjective time in bed was calculated as the duration between bedtime and waking time. Subjective TST was calculated by subtracting sleep latency and time awake during the night from time in bed. Subjective sleep efficiency was calculated by dividing subjective TST by subjective time in bed  100. Each sleep measure was recorded for one day. The self-perceived evaluation of feeling rested was measured using a visual analogue scale (VAS). Research has indicated that when measuring subjective experiences, the VAS scale is a more sensitive and effective tool than a Likert-type scale (Gift, 1989). The VAS scale consists of a horizontal line 10 cm in length fitted with a scale; 0 cm represents ‘‘not rested,’’ while 10 cm represents ‘‘very rested.’’ 2.4.4. Anxiety Anxiety states and traits were measured using Spielberger’s State-Trait Anxiety Inventory (STAI) (Spielberger, 1983). The state and trait anxiety scale consists of 20 selfreported items, with each item having a Likert-type scale of 1–4, where a score of 20 indicates the absence of anxiety and a score of 80 indicates high anxiety. STAI shows adequate validity and reliability. In this study, the Chinese version STAI had a Cronbach’s reliability coefficient of 0.82. 2.4.5. Depression The Taiwanese Depression Scale was used to measure depression (Yu et al., 2008). This depression scale consists of 22 self-reported scales, with each Likert-type scale running from 0 to 3, where score of 0 indicates the absence of depression and a score of 66 indicates severe depression. Subjects who scored more than 37 were considered to have clinical depression. The validity and reliability of this instrument have been established (Yu et al., 2008). The criterion-related validity with the CES-D scale was 0.92 and the discriminant validity was 93.2%. The scale also had a high Cronbach’s a value of 0.87–0.91 (Yu et al., 2008). In this study, the depression scale had a Cronbach’s reliability coefficient of 0.84. 2.4.6. Music evaluation Three aspects of the participants’ subjective experience associated with the musical pieces were evaluated: likeability, relaxation, and sleepiness. The music evaluation used a visual analogue scale (VAS). The VAS scale that was used consisted of a horizontal 10 cm line fitted with a scale; the left end of the scale represented ‘‘not at all,’’ while the right end of the scale represented ‘‘very much.’’ The music was rated by the subjects in the music group in the morning for three consecutive days. The test–retest reliability of the VAS scale for music evaluation has been established (Lai et al., 2008b).

Institutional Review Board approval 1 = not eligible 4 = refused to sleep in the laboratory 6 = no time to participate

Participant screening (N = 61) PSQI > 5

Demographic data collection and baseline data (subjective and objective sleep parameters)

Randomization (n = 50)

n = 25 Music group

n = 25 Control group

Posttest (subjective and objective sleep parameters)

Fig. 1. Trial profile.

sleep laboratory at the appointed time, the procedures of the study were fully explained (Fig. 1), and informed consent was obtained from each participant. The participants were assured of confidentiality, and that participation was voluntary. Subjects were scheduled to spend four consecutive nights in the sleep laboratory. Participants were randomly assigned to either the intervention group (n = 25) or the control group (n = 25), using the drawing of lots. All lots (labels) are packed in a jar that was prepared by another person. Researchers therefore did not know beforehand which group each participant would be assigned to. Researchers responsible for statistical analysis also were not aware of which group collected data belonged to. Both approaches were aimed to decrease the error rate. The participants in both groups were paid an honorarium for participation. Additionally, participants also completed a health survey to declare any known medical problems and medications that could affect their sleep assessment; none of them reported any mitigating factors. The room temperature was maintained comfortably between at 24 and 26 8C.

2.5. Procedures The study was approved by the institutional review board (IRB097109) and was conducted in a sleep laboratory. Participants’ eligibility was assessed one week prior to the night of PSG. After participants arrived at the

Fig. 2. Excerpt from researchers’ composition of sedate music.

E.-T. Chang et al. / International Journal of Nursing Studies 49 (2012) 921–930

925

Table 1 Demographic characteristics. Total (N = 50)

Variables

Gender Male Female Marital status Married Single (widowed and divorce) Religion None Buddhist Taoist Protestant Christian Other Education completed High school Associate’s degrees Bachelor’s degrees Master’s degrees Employment Unemployed Employed

Mean age in years PSQI State anxiety Trait anxiety Depression

Music (n = 25)

Control (n = 25)

N

%

n

%

n

%

3 47

6 94

2 23

8 92

1 23

4 92

12 38

24 76

6 19

24 76

6 19

24 26

22 14 8 5 1

44 28 16 10 2

9 8 4 3 1

36 32 16 12 4

13 6 4 2 0

52 24 16 8 0

19 13 13 5

38 26 26 10

10 7 4 4

40 28 16 16

1 6 9 1

4 24 36 4

18 32

36 64

9 16

36 64

9 16

36 64

M

SD

M

SD

M

SD

31.82 9.66 44.62 47.94 25.16

11.10 3.02 10.15 7.97 10.73

30.88 10.04 45.8 48.4 29.32

10.89 3.03 10.67 7.82 10.23

32.76 9.28 43.36 47.64 21.00

11.45 3.03 9.66 8.26 9.27

x2

p

0.35

0.55

0.0

1

2.21

0.69

3.85

0.28

0

1

t

p 0.59 0.88 0.87 0.26 2.94

0.55 0.38 0.38 0.79 0.005*

* p < 0.05.

Participants were instructed to refrain from consuming alcohol, drugs, and caffeine during the study period; they were also instructed to arrive at the sleep laboratory 1 h before their normal bedtime for electrode montage and preparation. The participants were prepared for the sleep studies by a qualified sleep technician. After preparing for bed, recording equipment was applied and the experimental procedure commenced. All the scoring was conducted by a polysomnographic technician with no knowledge of the group assignment, experiment details, or experimental conditions. Research assistants instructed the participants to listen to the music CD when they first went to bed. Participants in the music group were then instructed to evaluate their enjoyment of the music using a VAS scale when they awoke in the morning for three consecutive days. 2.6. Data analysis Data were analyzed using PASW 18.0 for Windows (SPSS Inc., Chicago, IL, USA). Demographics were compared using a t-test and x2-test. Fisher’s skewness coefficient was

used to check the normality of the data (Pett, 1997). A generalized estimating equation (GEE) analysis was used to control changes in time and baseline values. The baseline outcome measurements were used as covariates in the data analysis (Cook and Campbell, 1979). a < 0.05 was considered statistically significant. 3. Results 3.1. Description of participants Participants’ age ranged from 22 to 58 (31.82  11.10 years), and 3 (6%) were male and 47 (94%) were female. The majority of the participants were single (n = 38, 76%). There were no differences in demographics between the two groups. The mean scores of the global PSQI were 10.04 for the music group and 9.28 for the control group (Table 1). Both groups had STAI scores >40, indicating that the participants were in states of moderate anxiety. Previous studies revealed that depression and anxiety are related to increased variability of individuals’ sleep duration and fragmentation measured via wrist actigraphy (Mezick

Table 2 Sleep parameters at the baseline by group (N = 50). Sleep parameters

PSG sleep TST (min) SOL (min) SE (%)

Total (N = 50)

Music (n = 25)

Control (n = 25)

t

Mean

SD

Mean

SD

Mean

SD

377.88 13.91 90.28

58.46 10.83 7.98

373.58 15.44 89.07

56.76 13.24 10.23

328.68 12.38 91.48

60.89 7.72 4.73

p

0.57 0.99 1.06

0.56 0.32 0.29

E.-T. Chang et al. / International Journal of Nursing Studies 49 (2012) 921–930

926 Table 2 (Continued ) Sleep parameters

Total (N = 50) Mean

WASO (min) Stage 1, % of TST Stage 2, % of TST Stage 3 + 4, % of TST Stage REM, % of TST No. of awakenings AI Subjective sleep TST (min) SOL (min) No. of awakenings Rested rating

Music (n = 25) SD

Mean

Control (n = 25) SD

Mean

t

p

SD

26.55 5.52 54.41 19.41 20.68 7.78 10.11

25.55 3.42 7.72 4.95 5.53 4.08 5.37

30.68 5.54 54.67 19.51 20.54 8.08 9.53

32.08 3.86 8.05 4.90 6.40 3.87 4.33

22.42 5.50 54.14 19.29 20.81 7.48 10.70

16.37 2.99 7.54 5.09 4.63 4.35 6.27

1.14 0.03 0.23 0.15 0.17 0.51 0.76

0.25 0.97 0.81 0.87 0.86 0.60 0.44

345.9 40.46 3.12 5.84

71.53 34.88 3.17 1.72

344.60 47.40 4.04 5.74

67.11 39.66 3.93 1.88

347.20 33.52 2.20 5.94

77.05 28.48 1.82 1.58

1.42 0.12 2.12 0.01

0.16 0.89 0.03* 0.68

Note: TST, minutes of total sleep time; SOL, sleep onset latency; SE, sleep efficiency (total sleep time/total recording time); AI, arousal index; WASO, wake time after sleep onset. * p < 0.05.

et al., 2009). However, no significant correlations between anxiety, depression, and any sleep parameters were observed in this study. Participants also reported no difference between their sleep at home and in the sleep laboratory during the first night. Additionally, paired t-tests indicated that there was no significant difference between the baseline (day 1) and day 2 scores in any of the sleep parameters of the two groups, except for the rested rating (p = 0.003); thus, this study had no first night effect. Finally, no participants reported using alcohol or other behavioral interventions to improve sleep. 3.2. Comparability of groups During the first pretest (day 1), there were no significant differences between the music and the control groups regarding demographics, anxiety states, and anxiety traits, except for depression. The PSG and subjective sleep parameters also did not differ, except for the self-reported number of awakenings (Table 2). This study used the baseline sleep parameters and depression as covariates for data analysis (Cook and Campbell, 1979). 3.3. Music evaluation The music group comprised 25 participants; of these, 10 participants listened to soothing music which they selected themselves and 15 listened to music selected

Fig. 3. Music evaluation (n = 25).

by the researchers. The participants rated the music as greater than 7 points, and the scores were increased daily for all three items of evaluation (Fig. 3). 3.4. PSG sleep data There were no significant differences in the baseline PSG sleep data between the music and the control group (Table 2). The mean TST, SE, and the percentage of stages 1, 2, 3, and 4 in this study were similar to that of an adult population not experiencing insomnia (Hirshkowitz, 2004). The mean SOL was approximately 14 min (Table 2), longer than the 8 min for adults in the same age group not experiencing insomnia (Hirshkowitz, 2004). GEE was used to test the group effects in the posttest scores of PSG sleep parameters. While controlling for the pretest scores, significant differences in stage 2 (Wald x2 = 4.54, p = 0.03) and REM sleep between the groups (Wald x2 = 3.92, p = 0.04) (Table 3) was observed, indicating that participants listening to music at bedtime had shorter stage 2 sleep (Fig. 4) and longer REM sleep (Fig. 5) compared to the control group. The effect size for stage 2 and REM sleep was 0.17 and 0.10 respectively. Additionally, there were no significant differences in the mean TST, SOL, SE, WASO, stage 1, stage 3, and stage 4 sleep, number of awakenings, and arousal index between the two groups (Table 3).

Fig. 4. Changes in stage 2 sleep between the groups at different times. After controlling for the baseline, a significant group effect was found (p = 0.03).

E.-T. Chang et al. / International Journal of Nursing Studies 49 (2012) 921–930

927

Table 3 GEE estimating group difference (N = 50). Outcome measure PSG sleep parameters TST (min) SOL (min) SE (%) WASO (min) Stage 1, % of TST Stage 2, % of TST Stage 3 + 4, % of TST Stage REM, % of TST No. of awakenings AI Subjective sleep parameters TST (min) SOL (min) No. of awakenings Rested rating

Group

Estimate

SE

95% Wald C.I.

Wald

p

Musica Control Music Control Music Control Music Control Music Control Music Control Music Control Music Control Music Control Music Control

384.54 397.30 14.03 11.41 92.90 93.07 16.59 19.31 4.7 6.00 49.72 53.50 18.83 19.28 25.00 22.91 6.48 7.09 9.37 9.20

6.95 7.40 3.14 1.95 0.79 0.88 2.81 2.72 0.63 0.62 0.01 1.45 0.83 0.79 0.80 0.67 0.80 0.60 0.67 0.92

370.92–398.17 382.78–411.82 7.85–20.20 7.58–15.23 91.35–94.45 91.34–94.80 11.07–22.10 13.96–24.66 3.51–5.98 4.79–7.21 47.74–51.70 50.65–56.36 17.20–20.47 17.72–20.84 23.42–26.57 21.59–24.23 4.90–8.05 5.90–8.27 8.05–10.68 7.39–11.01

1.57

0.20

0.50

0.47

0.21

0.88

0.48

0.48

2.01

0.15

4.54

0.03*

0.14

0.69

3.92

0.04*

0.37

0.54

0.02

0.88

Music Control Music Control Music Control Music Control

367.13 377.60 28.55 25.16 2.07 1.95 7.69 6.72

9.29 9.85 4.32 4.12 0.28 0.25 0.32 0.23

348.9–385.35 358.28–396.92 20.08–37.02 17.07–33.25 1.52–2.62 1.44–2.45 7.06–8.32 6.26–7.17

0.59

0.44

0.32

0.57

0.09

0.75

6.08

0.01*

Note: TST, minutes of total sleep time; SOL, sleep onset latency; SE, sleep efficiency (total sleep time/total recording time); AI, arousal index; WASO, wake time after sleep onset. a Reference group. * p < 0.05.

3.5. Subjective sleep data Participants reported an average TST of 345.9  71.53 min on the self-reported questionnaire. The self-reported SOL value using the morning questionnaire was 40.46 min (Table 2), which was nearly triple (26.55 min longer) the PSG measure. Using GEE to test the group effects in the posttest scores of PSG sleep parameters, there were significant differences in the rested rating (Wald x2 = 6.08, p = 0.01) values while controlling for pretest scores (Table 3), indicating that participants listening to music at bedtime had a higher

Fig. 5. Changes in REM sleep between the groups at different times. After controlling for the baseline, a significant group effect was found (p = 0.04).

rested rating scores during the day compared to the control group (Fig. 6). The magnitude of effect size was 0.16. However, there were no significant differences in the mean TST, SOL, and number of awakenings (Table 3). 4. Discussion Our study is the first randomized clinical trial study exploring the effectiveness of music on sleep to carefully use both subjective and objective measurements. Our study found that the mean change scores between the two groups were demonstrated in the self-reported rested

Fig. 6. Changes in the rested rating scores between groups at different times. After controlling for the baseline, a significant group effect was found (p = 0.01).

928

E.-T. Chang et al. / International Journal of Nursing Studies 49 (2012) 921–930

rating, PSG stages N2, and REM sleep. Insomniac adults who listened to sedate music for 45 min at bedtime for 3 days had a higher rested rating score over time compared to those who did not. However, participants listening to music at bedtime had shorter stage 2 sleep and longer REM sleep than the control group did, indicating that listening to music at bedtime improves PSG sleep quality as manifested in stage 2 and REM sleep. The subjective sleep quality on the rested rating scores showed good agreement with daily improvements of PSG stage 2 and REM sleep. Studies examining the effects of listening to music at bedtime on sleep quality measured using PSG are limited. The findings of this study are not comparable with that of Lazic and Ogilvie (2007), who determined that musical intervention did not improve the sleep quality of 10 volunteer students with or without sleep disturbance. The inconsistent results may be due to the participants’ characteristics, sample size, and music selection, and their enjoyment of the music. Unfortunately, no parallel studies that can be compared and contrasted with this study exist. The lack of sufficient scientific evidence, therefore, limits the confirmation of music’s effect on insomniac adults. Though this study found that listening to music at bedtime music did not assist the participants in falling asleep faster, their PSG stage 2 sleep and REM sleep were improved. REM sleep is especially vital to psychological and emotional well-being. This study also demonstrated that the rested-rating scores significantly improved in the music group, though self-reports are subject to bias. Rested rating, a subjective concept, is not measureable by PSG. Therefore, both objective physiological measures and subjective sleep parameters are required when testing the efficacy of the therapeutic use of music. Similar non-significant results were obtained in another study that compared music, tones, and control conditions using the standard polysomnographic measures of sleep to determine the efficacy of the Delta Sleep System on the sleep quality of young female students (Lazic and Ogilvie, 2007). Therefore, this study partially supports the findings of Lazic and Ogilvie (2007); the PSG measures did not identify significant effects of music on other sleep parameters. Six types of music selected by researchers in this study were used at bedtime and were found useful for sleep. Participants in this study had the opportunity to select their own music or music selected by the researchers; a Cochrane review found that the positive effects of music were similar in studies where patients selected the type of music and where patients did not (Cepeda et al., 2006). Previous studies identified that the efficacy of therapeutic music is affected by the listeners enjoyment of the music (Lai and Good, 2005; Nilsson, 2010), with preferred music having the most beneficial effects on relaxation (Lai and Good, 2002; Lai, 2004) and stress reduction (Lai et al., 2011; Nilsson, 2010). Some of the music pieces selected by the researchers had been tested previously and were found to be effective in reducing the stress of caregivers of cancer patients (Lai et al., 2011), which may have contributed to the effectiveness of the music in improving sleep quality. Gold et al. (2009) postulated that the dosage of music therapy is the best predictor of its effects. Though

Zimmerman et al. (1996) suggested that three days was sufficient for a musical intervention on sleep quality, others only found effects after two weeks (Levin, 1998). In addition, participants in the music group reported that the music sounded sleepy and enabled them to fall asleep; this was also indicated in their daily music evaluation, where their sleepiness scores increased daily. The results can be explained according to psychophysiological theory that sleep quality can be improved by relaxing the body with sedate music, which decreases the circulation of noradrenaline (Gerra et al., 1998) related to the onset of sleep (Irwin et al., 1999). The timing of the intervention at bedtime and the use of soothing music for 45 min may facilitate relaxation while attempting to sleep (Lai and Good, 2005; Harmat et al., 2008). The participants in this study reported moderate anxiety levels. The efficacy of music on sleep improvement may be due to the music’s ability to reduce stress (Lai and Li, 2011). The autonomic nervous system increases activity during stress, resulting in changes to the body’s neuroendocrine and sympathetic nervous systems. Stress can cause the hypothalamus to secrete corticotrophin-releasing hormones. Reducing stress decreases serum cortisol (Lai and Li, 2011; Ventura et al., 2012), which promotes psychological well-being and REM sleep. 4.1. Limitations and implications A randomized controlled trial was used in this study to improve the study’s effectiveness. The design of this study included an evaluation of the effectiveness of listening to music using both objective and subjective measurements of sleep quality. The findings of this study add to the evidence-based methods that healthcare workers can use to improve the sleep quality of insomniacs; however, the findings should be interpreted with caution. The participants had anxiety scores >40. Future studies should recruit diverse patients to produce results that are more comprehensive. In this study, listening to music for 45 min affected stage 2 and REM sleep. Because the sample size calculation was based on subjective sleep quality, the lack of differences in the other PSG sleep parameters may be attributed to a lack of power to determine effect on PSG parameters, and thus the limited effects of music on PSG sleep parameters should be considered exploratory findings only. Prior studies have also identified that soothing music reduces the circulating norepinephrine (Gerra et al., 1998), a hormone associated with the onset of sleep (Irwin et al., 1999). Further studies may determine sample size based on SOL to examine the efficacy of music on sleep-onset insomnia. This study suggests that researchers continue to explore both the subjective and objective components of sleep quality to examine the efficacy of the therapeutic use of music. Additionally, we also recommend testing the listening of music regarding different durations and at various times to determine when the maximum benefits of music are produced. Though the mechanism of the effect of music is not fully understood, it should not interfere with the application of music as an intervention together with medical treatment to improve sleep quality. Finally, the

E.-T. Chang et al. / International Journal of Nursing Studies 49 (2012) 921–930

large number of tests should also be considered a limitation. 4.2. Conclusions The findings of this study contribute to existing knowledge of the effectiveness of music as a therapy to improve the sleep quality of insomniac adults. Listening to soothing music at nocturnal sleep time improved the rested rating scores, shortened stage 2 sleep, and prolonged REM sleep. This study suggests that soothing music improved subjective sleep quality with limited effects on the certain aspects of sleep quality measured by PSG. Acknowledgement The authors thank Zhao-Yeu Zhong for laboratory technical support.

Conflict of interest No conflict of interest is declared by the authors. Funding This study is funded by the National Science Council, Taiwan (no. 98-2314-B-303-010-MY2). Ethical approval The study was approved by the institutional review board of Tzu Chi Hospital (IRB097109). References American Academy of Sleep Medicine, 2005. International Classification of Sleep Disorders: Diagnostic and Coding Manual, 2nd ed. American Academy of Sleep Medicine, Westchester, IL. Bartlett, D.J., Marshall, N.S., Williams, A., Grunstein, R.R., 2008. Predictors of primary medical care consultation for sleep disorders. Sleep Medicine 9, 857–864. Buscemi, N., Vandermeer, B., Friesen, C., Bialy, L., Tubman, M., Ospina, M., Klassen, T.P., Witmans, M., 2007. The efficacy and safety of drug treatments for chronic insomnia in adults: a meta-analysis of RCTs. Journal of General Internal Medicine 22 (9), 1335–1350. Buysse, D.J., Reynolds 3rd, C.F., Monk, T.H., Berman, S.R., Kupfer, D.J., 1989. The Pittsburgh sleep quality index: a new instrument for psychiatric practice and research. Psychiatry Research 28, 193–213. Cepeda, M.S., Carr, D.B., Lau, J., Alvarez, H., 2006. Music for pain relief. Cochrane Database Systematic Review (2), CD004843. Cook, T.D., Campbell, D.T., 1979. Quasi-experimentation: Design and Analysis for Field Settings. Houghton Mifflin Company, Boston. DeMartinis, N.A., Kamath, J., Winokur, A., 2009. New approaches for the treatment of sleep disorders. Advances in Pharmacology 57, 187–235. Furihata, R., Uchiyama, M., Takahashia, S., Konno, C., Suzuki, M., Osakia, K., Kaneita, Y., Ohida, T., 2011. Self-help behaviors for sleep and depression: a Japanese nationwide general population survey. Journal of Affective Disorders 130, 75–82. Gerra, G., Zaimovic, A., Franchini, D., Palladino, M., Giucastro, G., Reali, N., Maestri, D., Caccavari, R., Delsignore, R., Brambilla, F., 1998. Neuroendocrine responses of healthy volunteers to techno-music: relationships with personality traits and emotional state. International Journal of Psychophysiology 28, 99–111. Gift, A., 1989. Visual analogue scales: measurement of subjective phenomena. Nursing Research 38, 286–288. Glass, J., Lanctot, K.L., Herrmann, N., Sproule, B.A., Busto, U.E., 2005. Sedative hypnotics in older people with insomnia: meta-analysis of risks and benefits. British Medical Journal 331 (7526), 1169. Gold, C., Sollic, H.P., Kruger, V., Lie, S.A., 2009. Dose–response relationship in music therapy for people with serious mental disorders:

929

systematic review and meta-analysis. Clinical Psychology Review 29, 193–207. Harmat, L., Takacs, J., Bodizs, R., 2008. Music improves sleep quality in students. Journal of Advanced Nursing 62 (3), 327–335. Hirshkowitz, M., 2004. Normal human sleep: an overview. Medical Clinics of North America 88 (3), 551–565. Ho, S.Y., Lai, H.L., Jeng, S.Y., Tang, C.Y., Sung, H.C., Chen, P.W., 2011. The effects of non-commercial music at mealtime on agitation in nursing home residents with dementia. Archives of Psychiatric Nursing 25, e49–e55. Iber, C., Ancoli-Israel, S., Chesson, A.L., Quan, S.F., 2007. The AASM Manual for the Scoring of Sleep and Associated Events: Rules, Terminology and Technical Specifications. American Academy of Sleep Medicine, Westchester, IL. Irwin, M., Thompson, J., Miller, C., Gillin, J.C., Ziegler, M., 1999. Effects of sleep and sleep deprivation on catecholamine and interleukin-2 levels in humans: clinical implications. Journal of Clinical Endocrinology and Metabolism 84 (6), 1979–1985. Kozasa, E.H., Hachul, H., Monson, C., Pinto, L., Garcia, M.C., de Araujo Moraes Mello, L.E., Tufik, S., 2010. Mind-body interventions for the treatment of insomnia: a review. Revista Brasileira de Psiquiatria 32 (4), 437–443. Kryger, M.H., 2006. The burden of chronic insomnia on society. Managed Care 15 (9), 1–5 17. Lai, H.L., Li, Y.M., Lee, L.H., 2011. Effects of music intervention with nursing presence and recorded music on psycho-physiological indices of cancer patient caregivers. Journal of Clinical Nursing, doi:10.1111/ j.1365-2702.2011.03916.x. Lai, H.L., Li, Y.M., 2011. The effect of music on biochemical markers and self-perceived stress among first-line nurses: a randomized controlled crossover trial. Journal of Advanced Nursing 67 (11), 2414–2424. Lai, H.L., Hwang, M.J., Chen, C.J., Chang, K.F., Peng, T.C., Chang, F.M., 2008a. Randomised controlled trial of music on state anxiety and physiological indices in patients undergoing root canal treatment. Journal of Clinical Nursing 17 (19), 2654–2660. Lai, H.L., Chen, P.W., Chen, C.J., Chang, H.K., Peng, T.C., Chang, F.M., 2008b. Randomized crossover trial studying the effect of music on examination anxiety. Nurse Education Today 28 (8), 909–916. Lai, H.L., Chen, C.J., Peng, T.C., Chang, F.M., Chen, M.L., Huang, H.Y., Chang, S.C., 2006. Randomized controlled trial of music during kangaroo care on perinatal anxiety and preterm infants’ responses. International Journal of Nursing Studies 43, 139–146. Lai, H.L., Good, M., 2005. Music improves sleep quality in older adults. Journal of Advanced Nursing 49 (3), 234–244. Lai, H.L., 2004. Music preference and relaxation in Taiwanese elderly people. Journal of Geriatric Nursing 25 (5), 286–291. Lai, H.L., Good, M., 2002. An overview of music therapy. The Journal of Nursing 49 (2), 80–84. Lazic, S.E., Ogilvie, R.D., 2007. Lack of efficacy of music to improve sleep: a polysomnographic and quantitative EEG analysis. International Journal of Psychophysiology 63, 232–239. Leger, D., Massuel, M.A., Metlaine, A., SISYPHE Study Group, 2006. Professional correlates of insomnia. Sleep 29 (2), 171–178. Leger, D., 2000. Public health and insomnia: economic impact. Sleep 23 (Suppl. 3), S69–S76. Leger, D., Levy, E., Paillard, M., 1999. The direct costs of insomnia in France. Sleep 22 (Suppl. 2), S394–S401. Levin, Y.I., 1998. ‘Brain music’ in the treatment of patients with insomnia. Neuroscience and Behavioral Physiology 28 (3), 330–335. Martin, S.A., Aikens, J.E., Chervin, R.D., 2004. Toward cost-effectiveness analysis in the diagnosis and treatment of insomnia. Sleep Medicine Reviews 8 (1), 63–72. Morin, C.M., LeBlanc, M., Daley, M., Gregoire, J.P., Merette, C., 2006. Epidemiology of insomnia: prevalence, self-help treatments, consultations, and determinants of help-seeking behaviors. Sleep Medicine 7, 123–130. Mezick, E.J., Matthews, K.A., Hall, M., Kamarck, T.W., Buysse, D.J., Owens, J.F., Reis, S.E., 2009. Intra-individual variability in sleep duration and fragmentation: associations with stress. Psychoneuroendocrinology 34, 1346–1354. Nilsson, U., 2011. Music: a nursing intervention. European Journal of Cardiovascular Nursing 10 (2), 73–74. Nilsson, U., 2010. Music: a nursing intervention. European Journal of Cardiovascular Nursing 10, 73–74. Nomura, K., Yamaok, K., Nakao, M., Yano, E., 2010. Social determinants of self-reported sleep problems in South Korea and Taiwan. Journal of Psychosomatic Research 69 (5), 435–440. Ohayon, M.M., 2002. Epidemiology of insomnia: what we know and what we still need to learn. Sleep Medicine Reviews 6, 97–111.

930

E.-T. Chang et al. / International Journal of Nursing Studies 49 (2012) 921–930

Pett, M.A., 1997. Nonparametric Statistics for Health Care AQ5 Research: Statistics for Small Samples and Unusual Distributions. Sage, CA. Schutte-Rodin, S., Broch, L., Buysse, D., Dorsey, C., Sateia, M., 2008. Clinical guideline for the evaluation and management of chronic insomnia in adults. Journal of Clinical Sleep Medicine 4 (5), 487–504. Spielberger, C.D., 1983. Manual for the State-Trait Anxiety Inventory. STAI (Form Y) Consulting Psychologists Press Inc., Palo Alto, CA. Tsai, P.S., Wang, S.Y., Wang, M.Y., Su, C.T., Yang, T.T., Huang, C.J., Fang, S.C., 2005. Psychometric evaluation of the Chinese version of the Pittsburgh Sleep Quality Index (CPSQI) in primary insomnia and control subjects. Quality of Life Research 14 (8), 1943–1952. Ventura, T., Gomes, M.C., Carreira, T., 2012. Cortisol and anxiety response to a relaxing intervention on pregnant women awaiting amniocentesis. Psychoneuroendocrinology 37 (1), 148–156.

Verster, J.C., Veldhuijzen, D.S., Patat, A., Olivier, B., Volkerts, E.R., 2006. Hypnotics and driving safety: meta-analyses of randomized controlled trials applying the on-the-road driving test. Current Drug Safety 1 (1), 63–71. Wade, A.G., 2011. The societal costs of insomnia. Neuropsychiatric Disease and Treatment 7, 1–18. Walsh, J.K., 2004. Clinical and socioeconomic correlates of insomnia. Journal of Clinical Psychiatry 65 (Suppl. 8), 13–19. Yu, M.N., Liu, Y.J., Li, R.H., 2008. The practical usage of cutoff score in the Taiwanese Depression Scale. Journal of Educational Research and Development 4 (4), 231–258. Zimmerman, L., Nieveen, J., Barnason, S., Schmaderer, M., 1996. The effects of music interventions on postoperative pain and sleep in coronary artery bypass graft (CABG) patients. Scholarly Inquiry for Nursing Practice 10 (2), 153–170.