Warm-up exercises reduce music conservatoire students’ pain intensity when controlling for mood, sleep and physical activity: A pilot study

Abstract

Introduction

Playing-related musculoskeletal pain is highly prevalent among classical music students, affecting them physically and psychologically. Unlike athletes or dancers, musicians’ pain often goes untreated due to stigma and lack of specialised healthcare. While warm-up exercises are common practice in sports and dance, there is a lack of empirical research regarding physical warm-ups for musicians’ pain. This study aimed to investigate the effects of a 2-week daily warm-up exercise intervention on conservatoire students’ pain intensity, interference and psychological distress, at day 0 and after 2 weeks.

Methods

Two groups were considered: an exercise (n = 9) and a waitlist control (n = 7) group. Data were collected using online Qualtrics surveys at three different stages.

Results

2 × 2 mixed ANOVAs indicated no significant effects on any of the outcome variables. However, when controlling for sleep, anxiety, depression and physical activity, the analyses revealed a significantly greater decrease in pain intensity in the exercise group compared to the control. In addition, bivariate correlations indicated that pain interference, anxiety and depression delta scores were significantly correlated for the exercise but not the control group. Anxiety and pain intensity delta scores were significantly correlated in both groups.

Conclusions

These findings suggest that daily warm-up exercises have the potential to reduce musicians’ pain intensity but only if certain variables are factored in. This pilot study highlights the multifactorial nature of musicians’ pain and unveils the role that mood, sleep and physical activity level may have when assessing the efficacy of interventions based on warm-up exercises.

Introduction

Playing-related musculoskeletal pain affects a strikingly high number of both students and professional classical musicians, with research demonstrating 12-month prevalence of up to 90% and point prevalence of up to 68%.^1,2

Playing-related musculoskeletal disorders (PRMDs) contributing factors involve physical, psychosocial and environmental aspects. Pain may be due to repetitive movements or load while playing their instrument. ³ Like athletes, ⁴ the locations and nature of musicians’ pain corresponds with the physical load required for the specific task. ⁵ For most musicians, PRMDs predominantly affect the upper extremities, neck and lumbar spine due to the ergonomic set-up of classical instruments.^6,7 Intrinsic physical factors such as height, sex (female) and hypermobility are also risk factors. ⁸ Furthermore, psychosocial risk factors include psychological stress, pressure from self, social phobia and perfectionism which are all common among musicians.^9,10 Poor sleep is another possible intervening factor for musicians’ pain. ³ Indeed, the relationship between sleep and chronic musculoskeletal pain is bidirectional. ¹¹ Worse pain intensity has been shown to predict poor sleep while lack of sleep can aggravate pain symptoms by increasing pain sensitivity. ¹²

In addition to sleep issues, chronic pain is commonly associated with psychological issues. ¹³ Musicians have frequently shown to have poorer mental health and higher levels of depression and anxiety than other populations.^14,15 PRMDs are significantly correlated with depression and performance anxiety among professional classical musicians. ¹⁰ However, more research into the association between general anxiety and PRMDs is needed. ¹⁶ Considering the above confounding variables is crucial for gaining a holistic picture of musicians’ pain but this is often neglected in the literature.^17,18

The onset of pain in musicians often begins when musicians transition into their first year of conservatoire, due to a sudden increase in practicing and the myriad of aforementioned risk factors, and can impact the longevity of their career. ¹⁹ Despite this, unlike athletes and dancers, musicians’ pain often goes untreated due to negative perceptions of pain and lack of specialised healthcare knowledge.^3,5,20

Physical exercise is a less popular treatment for musicians’ despite being regularly recommended for general work-related musculoskeletal disorders. ²¹ Physical exercise releases beta-endorphins and endocannabinoids, which subsequently modulate pain perception and pain sensitivity. ²² Furthermore, regular physical exercise can improve mental health and substantially alleviate anxiety and depression. ²³ Taking part in physical activity twice a week has been shown to reduce pain among professional and student musicians.^24,25 However, these studies did not have control groups meaning causation cannot be determined. Notwithstanding, an 11-week RCT using an exercise video intervention and educational workshop component significantly decreased PRMD intensity and interference. ²⁶ However, numerous participants sought additional treatments throughout the intervention and their study neglected compounding factors like sleep and psychological problems.

Despite the limited literature above regarding physical exercise, there is a dearth of literature that specifically focuses on the effects of regular physical warm-ups for musicians’ pain. ‘Warm-up’ is simply defined as activity done before exercise to prepare the body for exercising. ²⁷ The importance of warming-up before playing or performing is widely accepted in sports and other performing arts for injury prevention and performance enhancement. ²⁸ One study involving a prevention and education course for music students with musculoskeletal disorders included a variety of physical exercises and warm-up routines which led to improved body awareness and attitude toward prevention strategies. ²⁹ However, no inferences on their pain level could be made. ²⁹ Furthermore, a study investigating the effects of three different types of warm-ups, that is, cardiovascular, core muscle and musical (warming-up using the instrument), found that all of them reduced perceived exertion compared to the control. However, such study only addressed violinists and did not measure pain level. ²⁸ Thus, further investigation on the effects of regular physical warm-ups on musicians’ pain is needed.

Based on these premises, this study sought to investigate the effects of a daily physical warm-up intervention on conservatoire students’ pain intensity, interference and psychological distress. A 2-week intensive intervention was chosen as, to achieve optimal results, musicians need to undertake regular exercise for a minimum of two weeks. ⁶ Treatments for musicians should be informed by specialists, focussing on areas exposed to pain and overload and should not be overcomplicated. ³⁰ Therefore, this work was designed by a chartered performing arts physiotherapy specialist and involved straightforward movements and aerobic exercises based on students’ self-reported pain locations and symptoms. The daily warm-ups in this study were demonstrated via video, which could be done in private at any time, to increase accessibility and likelihood of commitment while combatting financial issues and potential stigma.

In light of the above, this study had four hypotheses. Firstly, the warm-up intervention would significantly reduce music students’ self-reported pain intensity. Secondly, the warm-up intervention would significantly reduce pain interference. Thirdly, the intervention would significantly reduce psychological distress, namely, anxiety and depression. Fourthly, based on literature regarding musculoskeletal pain and workplace interference, ³¹ it was hypothesised that psychological distress and pain interference would be significantly correlated for music students.

Methods

Participants

Music students, studying at conservatoires in the United Kingdom or Ireland, were recruited via a preliminary screening survey. The preliminary survey measured demographic information such as age, gender, musical instrument as well as PRMD symptoms, locations and intensity based on the Musculoskeletal Pain Intensity and Interference Questionnaire (MPIIQM ³² ; see Appendix 1). Those who self-reported experiencing PRMD symptoms within the last four weeks (at the time of the screening survey) met the inclusion criteria and were asked to leave their contact details should they wish to be contacted for the follow-up treatment intervention.

Responses from 148 potential participants were originally received, 102 of which were complete. A total of 32 people left their contact details for the treatment intervention and were subsequently sent a participant invitation letter. Of those 32 people, 19 participants, aged 19 – 31 (Age M = 23.74, S.D = 2.77, Female N = 15, Male N = 4) gave informed written consent to participate in the treatment intervention study. Figure 1 shows a flow-chart of the whole process. This study received ethical approval from the local ethics committee and followed the British Psychological Society’s (2021) Code of Human Research Ethics. No monetary compensation was provided.

Figure 1.

Flow-chart of participant recruitment from preliminary survey to intervention.

Pre- and post-test measures

Participants were asked to complete an online survey before and after the intervention which measured PRMDs (MPIIQM) ³² , psychological distress, average hours of sleep, practise and general physical activity. Participants were also asked to report any underlying health conditions, any diagnoses for their playing-related pain and whether they are taking any medication or receiving professional help (see Appendix 2). The MPIIQM ³² was chosen over other musculoskeletal pain questionnaires as it pertains specifically to musicians. The MPIIQM has previously been shown to have good validity and reliability and uses Likert-type scales to measure pain intensity ‘0 = no pain, 10 = pain as bad as you can imagine’, interference with mood ‘0 = does not interfere, 10 = totally interferes’ and interference with activities ‘0 = no difficulty, 10 = unable’. Self-reported anxiety was measured using the Generalised Anxiety Disorder pre- and post-test (GAD-7 ³³ ). Participants are required to self-report how often they have been bothered by any of the following problems (i.e. ‘feeling nervous, anxious or on edge’) over the past two weeks, using 4-point Likert-type scales (‘0 = not at all, 4 = nearly every day’). The 7-item questionnaire is scored from 0–21, with scores of 5,10,15 indicate mild, moderate and severe anxiety, respectively. Depression was measured using the Patient Health Questionnaire 9 (PHQ-9 ³⁴ ). Participants self-report how often they have been bothered by any of the following problems pertaining to depression (i.e. ‘feeling down, depressed or hopeless’) over the past two weeks using a 4-point Likert-type scale (‘0 = not at all, 4 = Nearly every day’). The PHQ-9 has 9 items and scores range from 0–27, with 5,10,15,20 indicating mild, moderate, moderately severe and severe depression, respectively. In the post-test survey, participants were asked to state how many days they missed and why. Participants were asked questions about their experience of taking part in the intervention and their opinions on the helpfulness of the warm-ups (see Appendix 2).

Table 1 summarises the measures recorded during both the preliminary and the pre-post survey.

Table 1.

Overview of the information gathered and questionnaires used during the preliminary screening survey (n=102) and the pre-post survey (n=16).

	Outcomes	Measure
Preliminary screening survey (see Appendix 1)	Demographics
	Age
	Gender (male, female, other, prefer not to say)
	Type of instrument played
	PRMDs	MPIIQM 32
	PRMD prevalence
	PRMD symptoms
	PRMD location
	PRMD intensity
	Eligible participants leave contact details
Pre/Post survey (see Appendix 2)	Demographics	Adapted from MPIIQM 32 for conservatoire students
	Age
	Gender
	Instrument
	Years played
	Year of study
	Lifestyle (average hours)
	Practice
	Sleep
	Physical activity
	PRMDs	MPIIQM 32
	PRMD prevalence
	PRMD symptoms
	PRMD location
	PRMD intensity
	PRMD interference
	Health
	Pain diagnosis
	Medication
	Additional support
	Anxiety	GAD 7
	Depression	PHQ 9
Post survey	Days skipped
	Participant experience
	Any further comments

bold = primary outcomes.

Musician-specific warm-up exercises

The warm-up exercises were designed by a chartered performing arts physiotherapist who designed the routine based on participants’ self-reported pain symptoms and locations from the initial screening survey. The warm-up routine included gentle body movements (predominantly targeting the upper body) and aerobic exercises, for approximately 15-min a day, designed to warm-up the body before practicing which is in line with recommendations. ³⁵ A list of the exact order, description and repetition for each exercise was provided to the participants (i.e. ‘with your feet hip distance apart, gently and slowly bend your upper body to the right and to the left × 10’).

Participants also received a video of the performing arts physiotherapist demonstrating how to do the exercises correctly and safely. The video included commentary from the physiotherapist regarding the purpose of each exercise, details about each exercise (i.e. ‘warming up the individual joints in the neck’) and educational information about the body (i.e. ‘there are 15 bones in your wrist’) (see Figure 2, for screenshots of video).

Figure 2.

Screenshots of the physiotherapist in the demonstration video.

Any comments that could have potentially biased the participants (i.e. ‘this particular exercise will help you so much if you have shoulder pain’) were cut from the video. Participants were repeatedly told to refer closely to the sheet for the exact number and repetitions.

Procedure

Data were collected using online Qualtrics surveys at three different stages. During the initial screening survey, participants were briefed about the nature of this project. Informed consent was obtained via Qualtrics before they could proceed. People who gave their contact information were emailed an information sheet containing further details about the intervention and a new consent form. Participants returned their consent via email and were then sent the pre-test survey. Once all participants had completed the pre-test survey, they were randomised into exercise and waitlist control group. The exercise group were emailed the list of exercises and the demonstration video (via WeTransfer) and were asked to do the exercises once a day before their first practice session for the next 14 days. Participants were repeatedly given the opportunity to ask questions. The waitlist control group were told they will start the exercise programme in two weeks’ time. During the two weeks, a check-in email was sent to the exercise group to see how they were getting on, give them another opportunity to ask questions and to remind them to continue doing the exercises every day. After the two weeks, all participants were sent the post-test survey. Finally, all participants received a debrief letter and the waitlist control group received the exercise list and video.

Statistical analysis

A series of 2 × 2 mixed ANOVAs were run to investigate the effects of the exercise intervention on the four outcome variables. To ascertain whether the exercise intervention would significantly reduce music students’ self-reported pain intensity and pain interference, two separate two-way mixed ANOVAs were run considering the within-subjects independent variable (Time) and the between-subjects independent variable (Group) on Pain_Intensity and Pain Interference. To investigate the possible effects of the intervention on psychological distress, two separate two-way mixed ANOVAs were run with Time and Group on anxiety and depression. The potential correlation between psychological distress (depression and anxiety separately) and pain interference, was tested via Pearson’s r. Finally, to account for possible mediating effects of sleep, ³⁶ anxiety, ³⁷ depression ³⁸ and physical activity ³⁹ on pain, two separate one-way ANCOVAs were run, one with anxiety, sleep and physical activity, and the other one with depression, sleep and exercise as covariates. The choice to have two separate ANCOVAs was due to an expected high correlation between anxiety and depression.

Significant interactions were followed up by Bonferroni-corrected post-hoc tests. The alpha level was set at 0.05.

Results

Preliminary screening survey

After incomplete responses were removed (N = 46), there were 102 complete responses for the preliminary screening survey (Age M = 22.94, SD = 5.10, Female = 73.53%, Male = 23.52%, Non-Binary = 1.99%). 81.37% of respondents (N = 83) self-reported experiencing a PRMD at some point in their life, 76.47% (N = 78) in the last 12 months, 50.98% (N = 52) in the last week and 18.62% (N = 19) never experienced a PRMD. These findings are in line with the literature.^1,2 From here on, only data pertaining to the intervention participants is discussed and analysed to address the four hypotheses.

Warm-up intervention

19 participants completed the pre-test survey. However, three participants dropped out during the 2-week intervention (exercise (1), control (2)) so their pre-test data was not included. The final post-test sample consisted of 16 participants aged 19–31 (M = 23.63, SD = 2.99, 12 Female (75%), 4 Male (25%)). There were nine participants in the exercise group (Age M = 23.22, SD = 2.64, 7 Females (77.78%)) and seven in the control (Age M = 24.14, SD = 3.53, 5 Females (71.43%)). See Tables 2–4 for demographic summaries.

Table 3.

Exercise group demographic summary (N = 9).

	Gender	Age	Instrument	Years played	Place of study	Year of study	Pain symptoms	Pain location
1	Female	24	Violin	19	UK	UG2	tingling, ache	neck, upper back, lower back, left shoulder, left arm
2	Female	27	Cello	21	UK	PG2	stabbing, tingling, ache	upper back, left arm, right arm, left hand, right hand
3	Male	23	Flute	12	Ireland	PG1	numbness, ache	neck, left hand
4	Female	21	Cello	14	UK	UG3	weakness, ache,	neck, upper back, left shoulder, right shoulder, left arm, right arm, left hand, right hand
5	Female	22	Violin	14	UK	UG3	weakness, ache, tenderness	right shoulder, right arm, right hand
6	Female	19	Viola	15	UK	UG1	ache	upper back, right arm, right wrist
7	Female	24	Piano	20	UK	PG2	weakness, ache	neck, upper back, lower back, left shoulder, right shoulder, left arm, right arm
8	Male	27	Drum kit	11	Ireland	PG1	numbness, tingling	lower back, right arm, right hand, left knee
9	Female	22	Violin	20	UK	PG1	numbness, tingling, ache	neck, upper back, left shoulder, right arm, right hand

Table 2.

Summary of descriptive statistics for all intervention participants.

		N	% (N = 16)
PRMD prevalence	Lifetime	16	100
	last 12 months	16	100
	last 4 weeks	16	100
	last 7 days	14	87.5
PRMD symptoms	Ache	14	87.5
	Stabbing	4	25
	Weakness	4	25
	Numbness	5	31.25
	Tingling	8	50
	Other: (Tightness, Tension, Soreness)	3	18.75
PRMD locations	Neck	9	56.25
	Upper back	9	56.25
	Lower back	4	25
	Left shoulder	8	50
	Right shoulder	4	25
	Left arm	5	31.25
	Right arm	6	6.25
	Left hand	4	25
	Right hand	4	25
	Other: (Index finger, right wrist, left knee, hip).
Instrument	Violin	5	31.25
	Cello	4	25
	Clarinet	1	6.25
	Double bass	1	6.25
	Drums	1	6.25
	Flute	1	6.25
	Piano	1	6.25
	Viola	1	6.25
	Voice	1	6.25

Table 4.

Waitlist control group demographic summary (N = 7).

	Gender	Age	Instrument	Years played	Place of study	Year of study	Pain symptoms	Pain location
1	Male	26	Double Bass	10	UK	PG1	stabbing, numbness, tingling	Left index finger
2	Female	31	Clarinet	20	UK	PG2	stabbing, weakness, ache	neck, upper back, left shoulder, left side abdomen
3	Female	23	Violin	16	Ireland	PG1	tingling, ache	neck, left shoulder, left arm
4	Female	23	Cello	20	UK	UG3	Stabbing, tingling, ache	neck, upper back, right shoulder
5	Male	20	Violin	16	UK	UG3	Numbness, tingling, ache, tension, soreness	neck, left hand, right hand
6	Female	24	Cello	18	Ireland	UG1	ache	lower back, left shoulder Hips and leg
7	Female	22	Vocal	1.5	Ireland	PG2	ache	upper back, left shoulder, right shoulder

Participants had been playing their instruments between 5–21 years (M = 15.69, SD = 4.53). See Table 5 for participants’ average hours of practise, sleep and physical exercise. No participants reported taking pain medication in the pre-test survey. However, at post-test, one control group participant reported using Voltarol Gel and Ibuprofen before a concert during the 2-week intervention. Two control group participants had a physiotherapy session and one exercise group participant had a sports massage during the intervention. Five participants (exercise (3), control (2)) sought professional psychological help during the 2-week intervention. One participant (control group) had Raynaud’s Syndrome, no other underlying health conditions were reported. Four participants were given a diagnosis for their pain (tennis elbow × 2, trapped nerve, repetitive strain injury). Four participants in the exercise group skipped between 1–4 days due to travelling or being sick. Due to the small sample size, participants were not excluded based on the aforementioned alterations.

Table 5.

Average hours of practice, physical activity and sleep between the two groups.

	Exercise Group (N = 9)	Control group (N = 7)
	M ± SD	M ± SD
Practice (hr/week)	18.81 ± 13.39	14.71 ± 10.95
Physical Activity(hr/week)	7.68 ± 7.31	6.50 ± 5.44
Sleep (hr/night)	7.22 ± 1.09	7.43 ± 7.31

Data from the pre- and post-test MPIIQM, GAD-7 and PHQ-9 were scored (i.e. the sum of the 4 pain intensity items). Table 6 shows means and standard deviations.

Table 6.

Means (and standard deviations) of four outcome variables.

	Group	PRE	POST
Pain_Intensity (MPIIQM)	Exercise	12.33 (8.8)	8.87 (5.6)
	Control	9.43 (7.9)	11.71 (6.8)
Pain_Interference (MPIIQM)	Exercise	21.78 (17.7)	15.56 (10.2)
	Control	17.86 (10.0)	17.86 (7.7)
Depression (PHQ-9)	Exercise	7.89 (6.9)	5.67 (4.4)
	Control	7.29 (6.8)	6.58 (5.3)
Anxiety (GAD-7)	Exercise	8.67 (7.5)	6.11 (4.4)
	Control	8.00 (5.7)	6.13 (4.0)

N = Number of participants, SD = Standard Deviation

As shown above, all four outcome variables decreased for the exercise group after taking part in the intervention. This decrease also occurred for the mood-related variables in the control group. However, Pain_Intensity appeared to increase in the control group between pre- and post-test.

Prior to running the ANOVAS, data was explored to assess parametric assumptions. Explorations of histograms and p-plots indicated that the data were approximately normally distributed (p > .05). However, Shapiro–Wilk scores revealed slight positive skews for pre- and post-test Pain_Intensity [(p = .02 and p = .01 respectively) and pre-test Depression (p = .03). Shapiro–Wilk was chosen over Kolmogorov–Smirnov, as it is more powerful for small sample sizes (N < 30). ⁴⁰ ANOVA is considered to be robust to violations of normality, even with very small samples (n = 3). ⁴¹ Thus, it was decided to proceed with parametric tests.

Levene’s test indicated the between-participant variable (Group) met homogeneity of variance for each combination of the independent variables. However, pre-test Pain_Interference was significant (p = .01), meaning this assumption had been violated.

Pain intensity

Results revealed no statistically significant main effect for Time [F (1,14) = 0.30, p = .60, η²p = .02] or for Group [F (1,14) = 0.00, p = 1.0, η²p = .00] on Pain_Intensity. These findings indicate no significant difference between pre- and post-test Pain_Intensity scores and no significant difference between exercise and control group. However, results revealed a statistically significant interaction between Time and Group [F (1,14) = 6.24, p = .03, η²p = .31], accounting for 31% of the total variance in Pain_Intensity, which is considered a large effect. ⁴² In other words, exercise group Pain_Intensity decreased between pre- (M = 12.33) and post-test (M = 8.78) while control group scores increased (M = 9.43, M = 11.71, respectively) (Figure 3).

Figure 3.

Means and standard errors relative to the Pain Intensity Delta scores (left) and Pain Interference Delta scores (right) per each group. The asterisk denotes a significant difference between groups after controlling for the covariates (ANCOVA results).

Post-hoc t-tests were run to investigate this interaction further, with Bonferroni correction (p = .0125). Two paired-samples t-tests revealed no significant difference in Pain_Intensity scores between pre- and post-test among the exercise group [t (8) = 2.11, p = .07] or the control group [t (6) = −1.50, p = .18]. Two independent samples t-tests revealed no significant difference in pre-test Pain_Intensity scores between the exercise and control group [t (14) = 0.68, p = .51] and no significant difference in post-test scores between the two groups [t (14) = −0.94, p = .36]. Thus, all four post-hoc checks were not statistically significant.

Pain interference

A two-way mixed ANOVA investigating the mean differences between Group and Time on Pain_Interference revealed no significant main effect for Time [F (1,14) = 2.09, p = .17, η²p = .13] or Group [F (1,14) = 0.02, p = .89, η²p = .001]. In addition, there was no significant interaction between Time and Group on Pain_Interference [F (1,14) = 2.09, p = .17, η²p = .13]. Delta scores were used to assess the change in score between pre- and post-test (i.e. delta = post - pre). A one-way between-participants ANOVA revealed no statistically significant difference between exercise and control group Pain_Interference Delta scores [(F (1,14) = 2.10, p = .17, η²p = .13], indicating no significant change in Pain_Interference between groups.

Depression

A two-way mixed ANOVA with Depression as the outcome variable demonstrated no significant main effect for Time [F (1,14) = 1.16, p = .30, η²p = .08] or for Group [F (1,14)= 0.00, p = .96, η²p = .00]. In addition, there was no significant interaction between Time and Group [F (1,14) = 0.32, p = .59, η²p = .02]. A one-way between-participants ANOVA was run with Depression Delta score and also revealed no significant main effects (p > .05), indicating no significant differences in the change in Depression between the groups.

Anxiety

A two-way mixed ANOVA revealed no significant main effect for Time [F (1,14) = 2.14, p = .17, η²p = .13] or for Group [F (1,14)= 0.02, p = .90, η²p = .00] on Anxiety. In addition, there was no significant interaction between Time and Group on Anxiety [F (1,14)= 0.05, p = .82, η²p = .00]. A one way between-participants ANOVA using Anxiety Delta score also revealed no significant main effects (p > .05), indicating no significant difference in the change in Anxiety between the two groups.

Analysis of covariance

Considering the established relationship between sleep, anxiety, depression and physical activity with pain, average hours of sleep per night, average hours of physical activity per week, levels of anxiety and depression, during the two week intervention, were tested as a covariate to partial out their potential effects on pain. Data was checked for additional assumptions of ANCOVA. Scatterplots revealed the covariates were linearly related to all four outcome variables. Data also met the assumptions of normality, homogeneity of variance (Levene’s, p > .05) and homogeneity of regression slope as the interaction between each one of the covariates and Group was not significant (p >.05). Since depression and anxiety scores were, unsurprisingly, highly correlated (see Table 7), to avoid collinearity issues ⁴³ two separate models were run, one with anxiety and the other with depression. A one-way ANCOVA with anxiety, sleep and physical activity as covariates revealed a statistically significant main effect for Group on Pain_Intensity Delta scores [F_1,11 = 5.09, p = .04, η²p = .31], with the experimental group reporting a significantly greater reduction of pain intensity (M = 3.55) compared to the controls (M = -2.28). Similarly, the one-way ANCOVA with depression, sleep and physical activity as covariates revealed a statistically significant main effect for Group on Pain_Intensity Delta scores (F_1,11 = 4.99, p = .04, η²p = .31), again with the experimental group reporting a significantly greater reduction of pain intensity compared to the controls. Conversely, the two one-way ANCOVAs on Pain_Interference Delta scores did not lead to any significant difference between groups, either considering anxiety, sleep and physical activity or depression, sleep and physical activity as covariates (respectively, F_1,11 = 1.50, p = .24 and F_1,11 = 1.44, p = .25).

Table 7.

Exercise group and control group correlations using delta scores (the change in scores between pre- and post-test).

	Exercise group (n = 9)				Control group (n = 7)
	1	2	3	4	1	2	3	4
1. The change in pain intensity (MPIIQM)
2. The change in pain interference (MPIIQM)	.68*				.45
3. The change in depression (PHQ – 9)	.51	.70*			.73	.47
4. The change in anxiety (GAD-7)	.68*	.81**	.82**		.82*	.68	.78*

*p = < .05, ** p = <.01

Correlations

Bivariate correlations were run to compare the associations between the four outcome variable delta scores. For the exercise group, results revealed a large significant positive correlation between Pain_Interference and Pain_Intensity delta scores (r = .68, p = .04). In addition, Pain_Interference Delta was significantly positively correlated with Anxiety (r = 0.81, p = .01) and Depression (r = 0.70, p= .04) delta scores in the exercise group but not in the control. Anxiety was significantly correlated with Pain_Intensity for both groups. As expected, Anxiety and Depression were strongly significantly positively correlated in both groups (see Table 7 for correlations).

Participants’ feedback

55.5% of participants reported their experience as ‘excellent’ and 44.5% ‘very good’. In addition, 44.5% reported the exercises as ‘extremely helpful’.

Discussion

This pilot study investigated the effects of a 2-week musician-specific exercise programme on conservatoire students’ Pain_Intensity, Pain_Interference, Depression and Anxiety. While all four outcome variables decreased for the exercise group at post-test, 2 × 2 ANOVAs indicated that the results were not significant. However, when controlling for sleep, physical activity, depression and anxiety, a significant difference in Pain_Intensity delta scores was demonstrated, with the experimental group reporting less pain than controls. Therefore, if we take into account the covariates, the first hypothesis that the intervention would significantly reduce Pain_Intensity was supported. The second hypothesis, that the intervention would significantly reduce Pain_Interference was not supported. Similarly, the third hypothesis, that the intervention would significantly reduce psychological distress was not supported, as the levels of self-reported Anxiety and Depression did decrease at post-test for the exercise group, but the results were not statistically significant. However, the fourth hypothesis, that psychological distress would be significantly correlated with Pain_Interference was supported, but only for the exercise group. Potential explanations and implications for the overall findings are discussed.

As mentioned, the decrease in Pain_Intensity after the intervention was significantly greater for the exercise group compared to the control group when physical activity, sleep and either anxiety or depression were factored in. This suggests that, when controlling for these factors, 15 min of daily warm-up exercises demonstrate the potential to reduce music students’ self-reported pain intensity. The reduction in pain intensity is consistent with research regarding the benefits of exercise interventions on musicians’ pain^24,25 and suggests that warm-ups are a specific type of exercise which may reduce pain intensity. The results also highlight the multifactorial nature of pain, supporting the assertion that a biopsychosocial approach is essential for gaining a holistic picture of musicians’ PRMDs.^17,44,45 The finding that the intervention had no significant effect on Anxiety or Depression may slightly contradict previous literature regarding the mood enhancing effects of general exercise. ²² However, a greater sample and longer-term exercise intervention may have led to a significant effect of the exercises suggested in the present study on these psychological variables. Also, it should be restated that the current study considered only a particular type of exercise, namely, warm-ups. Thus, while greater amounts of physical exercise may improve mood, ³⁹ warm-ups may not have the same effect on psychological factors.

The strong positive correlations between Anxiety, Depression and Pain_Interference delta scores support both the final hypothesis and research regarding musculoskeletal pain interference with work and psychological distress. ³¹ Although causation cannot be determined, the change in Pain_Interference was strongly correlated with the change in Anxiety and Depression in the exercise group, indicating that when Pain_Interference decreased, Anxiety and Depression were likely to decrease as well. This may suggest that musicians who experience pain which interferes with their playing are more likely to experience psychological distress. This corroborates qualitative research findings regarding distress among musicians when pain interferes with one’s ability to perform. ⁴⁶ Further investigation into psychological distress and pain interference among musicians could lead to potential psychological treatments focused on pain acceptance and catastrophising which have shown to improve physical functioning ⁴⁷ and reduce pain interference, anxiety and depression in general chronic pain populations. ⁴⁸ Such treatments could help musicians manage their pain while continuing to play, which may mitigate the aforementioned financial and emotional difficulties of being injured. ⁴⁶ Lastly, our intervention proved to be highly implementable and was very well received by participants.

The results should be interpreted in light of some limitations and their subsequent implications. Besides the limited sample size, due to drop-outs, the group numbers and genders were not balanced. Furthermore, some participants sought medical help (physiotherapy session: n = 2 control group) during the two weeks which has occurred in a previous RCT in the literature. ²⁶ In addition, participants were not monitored, so it is not possible to know whether participants fully complied with the exercise instructions. In an attempt to measure this, participants were asked to self-report how many days they had skipped and why. However, this could be subject to issues with honesty and social desirability bias. While it would be easier to monitor participants in person, at-home exercises are favourable and more accessible, particularly for every day warm-ups. Future research could ask participants to document their exercises on an online calendar or video-record themselves to prove their compliance. Finally, the exercises were not instrument specific. Thus, future research could investigate the effects of instrument specific warm-up exercises on musicians’ pain intensity, interference and psychological distress.

Conclusion

This was the first known pilot to investigate the effects of a 2-week long daily warm-up exercise intervention on music conservatoire students pain intensity, pain interference and psychological distress. This study addressed important pain-related aspects such as psychological distress and sleep which elicited useful discussions about the multifactorial nature of musicians’ pain. Indeed, when controlling for sleep, general physical activity, anxiety and depression, pain intensity decreased significantly for those who did the warm-up exercises compared to those who did not. This indicates the potential benefits that warm-up exercises have for reducing musicians’ pain. However, our results also emphasise that researchers and clinicians should always account for interacting factors when assessing the efficacy of their interventions for musicians’ pain.

ORCID iD

Matteo Martini https://orcid.org/0000-0002-3505-1284