The Impact of Group Singing on Mood, Coping, and Perceived Pain in Chronic Pain Patients Attending a Multidisciplinary Pain Clinic

Abstract

This study explored the impact of group singing on mood, coping, and perceived pain in chronic pain patients attending a multidisciplinary pain clinic. Singers participated in nine 30-minute sessions of small group singing, while comparisons listened to music while exercising. A short form of The Profile of Mood States (POMS) was administered before and after selected singing sessions to assess whether singing produced short-term elevations in mood. Results indicated that pre to post difference scores were significantly different between singing and control groups for only one of the 15 mood variables (i.e., uneasy). To test the longer term impacts of singing the Profile of Mood States, Zung Depression Inventory, Pain Self-Efficacy Questionnaire, Pain Rating Self-Statement, and Pain Disability Questionnaire were administered immediately before and after the singing sessions. All inventories other than the POMS were re-administered 6 months later. One-way ANCOVAs indicated that participants who attended the singing sessions showed evidence of postintervention improvements in active coping, relative to those who failed to attend, when preintervention differences in active coping were controlled for. While the singing group showed marked improvements from pre to postintervention on all mood, coping, and perceived pain variables, these improvements were also observed among comparison participants. The results of this study suggest that active singing may have some benefits, in terms of enhancing active coping, though the limitations of the study and small effect sizes observed suggest that further research is required to fully explore such effects.

Full text

Singing and chronic pain 54
Journal of Music Therapy
The impact of group singing on mood, coping and perceived
pain in chronic pain patients attending a multidisciplinary
pain clinic.
Dianna T. Kenny and Gavin Faunce
Journal of Music Therapy
2004; XLI (3): 241-258
This study explored the impact of group singing on mood, coping, and perceived pain in chronic pain patients attending a multi-
disciplinary pain clinic. Singers participated in nine 30-minute sessions of small group singing, while comparisons listened to
music while exercising. A short form of The Profi le of Mood States (POMS) was administered before and after selected singing
sessions to assess whether singing produced short-term elevations in mood. Results indicated that pre to post difference scores
were signifi cantly different between singing and control groups for only one of the 15 mood variables (i.e. uneasy). To test the
longer term impacts of singing the Profi le of Mood States, Zung Depression Inventory, Pain Self-Effi cacy Questionnaire, Pain
Rating Self-Statement, and Pain Disability Questionnaire were administered immediately before and after the singing sessions.
All inventories other than the POMS were re-administered six months later. One-way ANCOVAs indicated that participants who
attended the singing sessions showed evidence of post-intervention improvements in active coping, relative to those who failed to
attend, when pre-intervention differences in active coping were controlled for. While the singing group showed marked improve-
ments from pre to post-intervention on all mood, coping, and perceived pain variables, these improvements were also observed
among comparison participants. The results of this study suggest that active singing may have some benefi ts, in terms of enhancing
active coping, though the limitations of the study and small effect sizes observed suggest that further research is required to fully
explore such effects.
Music has been shown to produce emotional and
physiological reactions in listeners, with certain
structural features of music linked to specifi c
emotional and physiological responses (Aldridge,
1993, 1994; Blood, Zatorre, Bermudez, & Evans,
1999; Krumhansl, 1997; Sloboda, 1991; Spintge,
1985). Krumhansl (1997) measured physiological
changes in response to different pieces of music and
found that music in a major key and with a rapid
tempo induced a feeling of happiness in subjects
while music in a minor key and with a slow tempo
induced a feeling of sadness. Employing a PET
scanning technique, Blood et al. (1999) found
that consonant music stimulated parts of the brain
associated with pleasure, and subjects accordingly
reported positive emotions whilst listening to
this type of music. Dissonant music, however,
stimulated parts of the limbic system associated
with unpleasant emotion, and subjects accordingly
reported negative emotions whilst listening to this
type of music.
Given these fi ndings, it is not surprising that music
has become widely used by health professionals as
a form of therapeutic intervention across a diverse
range of patient groups (see Snyder & Chlan, 1999,
for a review). Listening to music has been effective
in reducing pre-operative anxiety (Augustin &
Haines, 1996; Gaberson, 1995; Winter, Paskin, &
Baker, 1994), promoting relaxation in medically
ill patients (Chlan, 1995; Guzetta, 1989; Updike,
1990), alleviating tension headaches (Linoff &
West, 1982), reducing stress (Burns et al., 2002),
and improving depressed mood and self esteem
(Hanser & Thompson, 1994). Positive effects of
music have been observed across all ages from
newborns (Caine, 1991; Kaminski & Hall, 1996)
to the elderly (Hanser & Thompson, 1994).
An important application of music therapy has
been in the area of pain management. Listening to
music has been found to decrease pain perception
in cancer patients (Bailey, 1983; Beck, 1991; Pfaff,
Smith, & Gowan, 1989; Zimmerman, Pozehl,
Duncan, & Schmitz, 1989), burn patients suffering
severe pain (Edward, 1998; Miller, Hickman, &
Lemasters, 1992), patients undergoing surgical
procedures (Davis, 1992; Dubois, Bartter, & Pratter,
1995), post-operative patients (Zimmerman,
Nieveen, Barnason, & Schmaderer, 1996), and
people suffering from chronic pain (Rider, 1987;
Schorr, 1993; Sedei Godley, 1987). Music has also
been found to induce positive mood changes in
post-surgical patients suffering from pain (Bonny,
1983; Mullooly, Levin, & Feldman, 1988), and to
increase activity levels and positive verbalizations
in chronic pain patients (Wolfe, 1978).
While a substantial body of research has
studied the effects of listening to instrumental
music on emotions and physiology, few studies
have examined the effects of listening to vocal
music. These have reported a range of emotional
(Sloboda, 1992), biochemical (Wade & Tavris,

Singing and chronic pain 55
Journal of Music Therapy
1998), and physiological (Gfeller, Asmus, &
Eckert, 1991; Lenton & Martin, 1991) effects.
Aldridge (1991), for instance, detected changes
in breathing, heart rate, fi ne motor movements,
and levels of consciousness in comatose patients
exposed to improvised wordless singing. Similarly,
Coleman, Pratt, Stoddard, Gerstmann, and Abel
(1997) found improvements in heart rate, oxygen
saturation, distress behaviours, caloric intake, and
weight gain in premature neonates exposed to
male and female singing three times daily for four
days. Further, active participation in singing may
confer additional benefi ts in terms of improving
mood. Unwin, Kenny, and Davis (2002) found
that a single half-hour session of active singing
or listening to singing resulted in immediate
improvements in mood in a volunteer community
sample.
As a community, we rely heavily on both licit and
illicit substances to maintain our health and well-
being and to control our moods. These substances
are costly, and many have unwanted and dangerous
side effects. Chronic pain patients often rely on a
formidable arsenal of drugs to manage their pain,
many of which have serious unwanted side effects.
While listening to music has been found to reduce
pain perception in chronic pain sufferers (Rider,
1987; Schorr, 1993; Sedei Godley, 1987), the role
that active singing may play in enhancing mood
and reducing perceived pain in such patients has
not been examined. Like exercise, singing may
prove to be a simple, cost effective therapeutic
intervention in vulnerable and depressed groups
such as chronic pain patients, thereby reducing
the cost of health care, while simultaneously
enhancing the quality of life.
In the present study, we proposed to test the effect
of singing on the mood, coping and perceived
pain of chronic pain patients, by comparing active
singing and passive listening-to-singing groups.
While listening to music has been found to lead
to enhanced self-reported mood and reduced pain
perception in patients suffering from chronic
pain (see Snyder & Chlan, 1999, for a review),
the fi ndings of Unwin, Kenny and Davis (2002)
suggest that the therapeutic effect of active group
participation in singing may be more effective than
passively listening to singing. Participants in the
present study were given nine 30 minute singing
sessions, as opposed to one 30-minute session in
Unwin et al., (2002) to strengthen expected gains
among patients in the singing group. A positive
effect of singing on active coping was expected
due to the ability of music to distract attention
from the perception of pain and arouse positive
emotional experiences, some of which may play
a cathartic role (Brown, Chen, & Dworkin, 1989).
It was hypothesised that both singing and listening
to singing would have a positive effect on mood,
coping, and perceived pain, but that singing would
have a greater impact than listening to singing. The
greater impact of active singing was expected due
to the heightened physical activity and changes
in breathing pattern involved in singing, changes
which are integral to most emotional reactions (von
Euler, 1986). The present study also employed a
6-month follow-up to assess the maintenance of
any observed gains on the outcome measures.
METHOD
Participants
At Royal North Shore Hospital, Sydney,
Australia, the ADAPT pain management program
offers 10 new places per fortnight to chronic
pain patients who attend the program daily from
9am–5pm for a period of three weeks. Power
calculations to determine sample size were based
on pre and posttest measures obtained on the
median change score for the study conducted by
Unwin, Kenny, and Davis (2002). The scores (and
standard deviations) for experimental and control
groups respectively were 22.73 (7.58) and 19.06
(7.56). Assuming normally distributed data, a
standard deviation of 7.57, p<.05, effect size of .6,
and power = .8, then 68 participants were required
in each group (136 in total). However, taking into
consideration the difference in intensity of the two
interventions (i.e. one session for the community
study versus nine sessions for the current study),
we conservatively estimated that the effect size
of nine sessions would be at least 1.5 times
greater than the effect size for one session. This
consideration reduced the required sample size
to 31 per group. We rounded to 30 because
participants were recruited in groups of 10, and
treated this fi gure as the ideal minimum number.
Seventy-seven participants were recruited for the
study.
Ethical approval for the study was received from
the Human Research Ethics Committee (HREC)
of the Royal North Shore Hospital and The
University of Sydney. However, they specifi ed
that the original design, in which we proposed that
the control group receive only the standard pain
management program, be modifi ed to a listening-
to-singing-while-exercising group. Since both

Singing and chronic pain 56
Journal of Music Therapy
listening to singing and exercising have both been
shown to enhance mood, the study became one
of comparison between two known mood change
agents rather than one to test the effect of singing
per se. This ethical requirement changed the
intent of the study but it was still useful to assess
whether there were any differences between the
two methods in terms of mood and coping.
Instruments
The ADAPT pain program administers a
comprehensive battery of tests pre- and post-
program and at six-month follow-up. These
include:
(i) Zung Depression Inventory (ZDI)
The Zung Depression Inventory (Zung, Richards
& Short, 1965) is a 20 item self-report inventory
measuring various psychological and somatic
symptoms of depressive illness. Half the items
are worded positively, such as “Morning is when
I feel best”, and half negatively “I feel down
hearted, blue and sad”. Patients rate items based
on a 4-point rating scale (ranging from 1=“None
of the time” to 4=“Most or all of the time”). Each
of the questions relate to 20 depressive symptoms,
for example, insomnia “I have trouble sleeping
through the night”, weight loss “I notice that I
am losing weight”, clouded reasoning “My mind
is as clear as it used to be”, and suicidal ideation
“I feel that others would be better off if I were
dead”. Raw scores are converted to a 100 point
scaled score (SDS Index Score), indicating normal
(<50), mild depression (<60), moderate (<70) or
marked major depression, or severe or extreme
major depression (>70). No reliability or validity
information is available for this inventory.
(ii) Pain Self-Effi cacy Questionnaire (PSEQ)
Derived from Bandura’s (Bandura, 1977)
concept of self-effi cacy, the PSEQ was specifi cally
developed to measure patients’ beliefs that they can
engage in tasks despite being in pain (Nicholas,
1989). The psychometric properties of this scale,
such as internal consistency and stability as well
as concurrent, construct and discriminant validity
have been reported in a number of studies (Gibson
& Strong, 1996; Nicholas, 1989). The PSEQ
contains ten items, each of which asks patients to
rate how confi dent they are that they can perform
a range of activities despite pain. Items include
statements such as `I can do some form of work
despite the pain (work includes housework, paid
and unpaid work)’. Each item is rated on a 7-point
rating scale (ranging from 0=`not at all confi dent’
to 6=`completely confi dent’), and scores from the
ten items are totalled. Total scores on the PSEQ
range from 0 to 60, with higher scores indicating
stronger pain self-effi cacy beliefs (Sharp &
Nicholas, 2000). The PSEQ has been shown to
have good internal consistency (0.92) and test-
retest reliability (0.79), as well as being highly
correlated with other relevant pain measures
(Nicholas, 1989). It is sensitive to treatment
effects as shown by changes following treatment
in cognitive-behavioural programmes (Nicholas,
1989; Williams et al. 1993) and compares
favourably with other measures of self-effi cacy
(Gibson & Strong, 1996) with the added advantage
of being much shorter and easier to score than
most other questionnaires.
(iii) Pain Responses Self-Statements (PRSS)
(taken from Sharp & Nicholas, 2000).
The PRSS is an 18-item scale developed to assess
specifi c, pain-related cognitions. Two scales are
derived from the questionnaire, a Catastrophizing
Scale and an Active Coping Scale. It has been
found to have excellent internal consistency (0.92
for the catastrophizing scale and 0.88 for active
coping), stability (0.87 for catastrophizing and
0.77 for active coping), factor, construct and
discriminant validity as well as being sensitive to
change (with signifi cant differences being noted
for ‘improved’ versus ‘unimproved’ patients (Flor,
Behle & Birbaumer, 1993).
(iv) Oswestry Low Back Pain Disability
Questionnaire (PDQ)
The Oswestry Low-Back Pain Disability
Questionnaire (Fairbank, Davies, Couper &
O’Brien, 1980) is a self-report instrument
measuring both quality-of-life and pain tolerance.
It consists of 10 sections, each containing fi ve
statements, representing increasing levels of
disability on that dimension, as well as a separate
pain intensity section. The maximum score on
each section is fi ve, totalling a maximum score of
50 for the full scale. This score is then converted to
a percentage, with high percentages representing
high disability. The scale has been shown to have
good internal consistency and high test-retest
reliability, correlating 0.99 (p<.001) for a one
day test-retest interval, and 0.83 for a one week
interval (Fairbank et. al. 1980). Construct validity
has been demonstrated through high correlations
with scores on other low back pain disability scales
such as the Quebec (Kopec & Esdaile, 1995) and
the Roland Morris (Co, Eaton & Maxwell, 1993).
To this standard battery, we added the Profi le

Singing and chronic pain 57
Journal of Music Therapy
of Mood States (POMS; McNair & Droppleman,
1981) in order to allow a comparison of fi ndings
with the previous study on the effects of singing
on mood in a community sample.
Setting
Groups of 10 chronic pain patients attend the
centre for fi ve days per week from 9am-5pm
for three weeks in order to undergo an intensive
three-week cognitive-behavioural program
in conjunction with a supervised functional
restoration program and supervised reduction or
withdrawal of pain medication. The program is
intensive and participants are fully engaged each
day on various activities such as education sessions,
exercise, group activities, community based tasks
and medical and psychological consultations.
Procedure
Because participants were recruited in blocks
of up to ten every second week, recruitment
of participants to the study also occurred in
blocks. Each block was randomly assigned to
an experimental or comparison condition and all
members of each block were invited to participate
in either the experimental or comparison condition.
This was the only feasible way to manage
allocation, given the structure of the program. All
participants were thoroughly screened by the pain
management team for suitability for inclusion in
the program. Hence, any individual differences
among groups would be randomly shared between
the experimental and comparison groups, so
the heterogeneity of the sample is of no greater
importance for our study as for other studies of
clinical populations. The proposed design was
robust because all participants received exactly
the same pain management program during the
three-week intervention program and the only
factor that varied was the addition of singing (E)
or listening to singing while exercising (C) as
additional activities.
Participants assigned to the singing condition
engaged in group singing in small groups ranging
in size from two to ten, depending on attendance.
A singing protocol was devised, refi ned and tested
in two pilot studies using both younger (people
aged 16-17 years) and older adults (people aged
45-65 years). A trained singing teacher and a piano
accompanist (fi rst author) conducted the singing
sessions.
The protocol for each lesson consisted of warm
up vocal exercises followed by several songs of
lively tempo, syncopated rhythm, pitch range of
no more than an octave, and simple structure and
words. Songs were accompanied on a clavinova
(a portable electric keyboard). Since optimum
posture and clear articulation enhance the singing
energy, some simple instructions on clear speech
and correct posture during singing were provided
prior to commencement. An overhead projector
encouraged the desired posture of the participants
by preventing them from looking down while
singing. The participants’ practised speaking
the words of songs prior to singing, to develop
familiarity and to encourage clarity of speech.
Each song was sung fi rst by the singing teacher
and then by the group. Songs had repetitive phrases
with minimum words for easy learning. For the
last section of the session, the singing group sang
all the songs through as if in performance. This
protocol took 30 minutes. Subsequent sessions
built on work covered in the earlier sessions and
new songs were slowly introduced in subsequent
sessions as earlier songs were mastered.
Participants were free to sit, stand, lean against
the wall or lie on the fl oor for short periods during
the singing sessions to ease any discomfort that
they may have experienced as a result of their pain
conditions.
RESULTS
Due to the unavoidable limitations of the listening
to singing while exercising group as a comparison
group (as mentioned earlier), participants who
failed to attend their singing sessions were
examined as a separate comparison group. While
not randomly assigned, the presence of a failed to
attend group allows for a direct examination of
the effect of singing over and above the standard
pain management program (which the listening to
singing while exercising comparison group does
not). Thus, there were ultimately three groups
within the design: (1) active singing, (2) listening
to singing whilst exercising, and (3) failed to
attend singing sessions.
Data screening
For each participant, a difference score was
calculated for each of the six POMS sub-
scales, the Zung depression inventory, the pain
self-effi cacy questionnaire, the pain disability
questionnaire, and the 2 PRSS sub-scales (active
coping, catastrophizing). These difference
scores were calculated by subtracting the post-
intervention score from the pre-intervention

Singing and chronic pain 58
Journal of Music Therapy
score for the anxiety, depression, confusion,
tension, and fatigue sub-scales of the POMS, and
the Zung depression inventory, pain disability
questionnaire, and catstrophizing sub-scale of the
PRSS. The difference scores were calculated by
subtracting the pre-intervention score from the
post-intervention score for the vigour sub-scale
of the POMS, the pain self-effi cacy questionnaire,
and the active coping sub-scale of the PRSS. These
calculations ensured that positive changes from
pre to post-intervention were refl ected in positive
difference scores, and negative changes from pre
to post-intervention were refl ected in negative
difference scores.
Prior to analysis, the distributions for each of
the difference scores and pre, post, and follow-
up variables were examined. Scores more than
three standard deviations from the mean on
their respective variable were considered to be
univariate outliers. Based on this criterion, four
difference scores (i.e. one for POMS anxiety, one
for POMS depression, one for pain disability, and
one for catastrophizing), one pre-intervention
score (i.e. one for POMS vigour), and three post-
intervention scores (i.e. one for POMS depression,
one for active coping, and one for catastrophizing)
were deleted as outliers. The distributions for
age and pain duration were also examined,
and two univariate outliers were subsequently
deleted from the pain duration data (based on the
deletion criterion above). Following the deletion
of univariate outliers, skewness and kurtosis
were examined for each variable. No problematic
deviations from normality were detected.
Pre-intervention group differences
Given the presence of some signifi cant pre-
intervention group differences, it was decided that
Table 1. Group Means (Standard Deviations in Parentheses) for Age, Pain Duration, and each
of the Mood, Coping, and Perceived Pain Variables
Singing
Control
Singing FTA
Pre
Post
Follow-
up
Pre
Post
Follow-
up
Pre
Post
Follow-
up
Age
42.17
(9.04)
-
-
37.87
(11.46)
-
-
42.36
(12.14)
-
-
Pain duration
73.78
(52.96)
-
-
46.04
(41.16)
-
-
38.15
(47.74)
-
-
Anger
13.00
(10.31)
7.80
(9.55)
-
15.23
(11.76)
10.38
(9.38)
-
13.93
(11.32)
10.08
(10.49)
-
Anxiety
14.67
(7.77)
9.20
(5.76)
-
17.00
(8.65)
13.05
(9.09)
-
17.15
(9.08)
12.76
(9.23)
-
Confusion
10.33
(6.04)
7.40
(6.73)
-
14.18
(6.22)
10.10
(6.26)
-
11.30
(5.43)
8.36
(6.51)
-
Depression
19.17
(13.84)
10.27
(10.51)
-
24.14
(12.84)
15.52
(13.15)
-
19.19
(14.82)
13.75
(14.85)
-
Fatigue
13.83
(5.57)
11.13
(6.19)
-
15.77
(7.78)
11.67
(7.18)
-
15.96
(7.11)
13.80
(6.79)
-
Vigour
11.17
(7.09)
15.13
(8.48)
-
9.50
(5.14)
15.00
(7.71)
-
10.54
(5.89)
16.32
(6.48)
-
Zung
64.06
(11.35)
56.35
(11.78)
57.11
(12.12)
62.30
(14.15)
53.29
(13.17)
57.89
(13.67)
61.96
(14.48)
54.86
(11.83)
53.31
(16.40)
Pain self
-effi cacy
23.39
(9.08)
38.24
(14.28)
34.78
(11.81)
23.13
(12.74)
37.61
(11.86)
32.44
(12.20)
21.00
(14.93)
34.89
(12.91)
33.77
(18.44)
Pain disability
12.78
(5.42)
9.24
(6.69)
10.33
(5.45)
13.07
(5.44)
8.18
(5.35)
7.67
(4.21)
13.36
(6.62)
10.64
(5.79)
10.08
(7.14)
Catastroph-
ising
2.73
(1.05)
1.78
(1.20)
1.71
(1.15)
3.28
(.90)
2.12
(1.06)
2.23
(1.04)
2.86
(1.22)
2.14
(1.36)
1.85
(1.63)
Active coping
2.73
(.79)
3.52
(.71)
3.18
(.53)
2.60
(.95)
3.44
(.90)
3.09
(.81)
2.61
(1.05)
3.20
(.81)
3.39
(.97)
Note. Dashes indicate data was not obtained.

Singing and chronic pain 59
Journal of Music Therapy
pre-intervention scores should be employed as a
covariate in the main analyses.
The impact of group singing on mood, coping,
and perceived pain
In order to assess the effect of the singing
intervention on the various measures of mood,
coping, and perceived pain, a series of one-way
ANCOVAs were conducted on the difference
scores (i.e. pre – post or post – ) for each dependent
variable, with pre-intervention scores entered as a
covariate.
In a fi rst group of analyses, participants in the
singing group were compared to participants in the
comparison group. Separate one-way ANCOVAs
were performed on the difference scores for each
of the six POMS sub-scales, Zung depression,
active coping, catastrophizing, pain self-effi cacy,
and pain disability, with group (i.e. singing/
comparison) as the between subjects variable
and pre-intervention scores as a covariate. No
signifi cant differences were found.
In a second group of analyses, participants in the
singing group were compared to participants who
failed to attend their singing sessions. Separate one-
way ANCOVAs were performed on the difference
scores for each of the six POMS sub-scales, Zung
depression, active coping, catastrophizing, pain
self-effi cacy, and pain disability, with group (i.e.
singing/singing FTA) as the between subjects
variable and pre-intervention scores as a covariate.
No signifi cant differences were found, although
participants in the singing group tended to show
a greater increase in active coping following the
intervention than participants in the singing FTA
group, a difference which just failed to reach
conventional signifi cance, F(1,41) = 3.70, p =
.061.
These data suggest that exposure to the
singing condition can increase active coping
responses, relative to comparison groups, when
pre-intervention active coping differences are
controlled for.
Follow-up data analyses
In order to assess whether or not the effects of
the singing intervention were maintained over a
six-month follow-up period, a series of two-way
repeated measures ANCOVAs were conducted,
with group as the between subjects variable, time
(i.e. post-intervention vs follow-up) as the within
subjects variable, and pre-intervention scores as a
covariate.
In a fi rst group of analyses, participants in the
singing group were compared to participants in the
comparison group. Separate two-way ANCOVAs
were performed for Zung depression, active
coping, catastrophizing, pain self-effi cacy, and pain
disability, with group (i.e. singing/comparison) as
the between subjects, time (i.e. post-intervention/
follow-up) as the within subjects variable, and pre-
intervention scores as a covariate. A signifi cant
effect of time was found for active coping, F(1,24)
= 5.78, p = .024, and pain self-effi cacy, F(1,24) =
4.54, p = .044, but the group x time interaction only
reached signifi cance for pain disability, F(1,24) =
5.13, p = .033), with the singing group showing
an increase in pain disability from post to follow-
up (though the follow-up mean remained below
the pre-intervention mean) and the comparison
group showing a decrease in pain disability post-
intervention to follow-up.
In a second group of analyses, participants in the
singing group were compared to participants who
failed to attend their singing sessions. Separate
two-way ANCOVAs were performed for Zung
depression, active coping, catastrophizing, pain
self-effi cacy, and pain disability, with group (i.e.
singing/singing FTA) as the between subjects
variable, time (i.e.pre, post-intervention/follow-up)
as the within subjects variable, and pre-intervention
scores as a covariate. The group effect approached
signifi cance for castrophizing, F(1,19) = 3.57, p =
.074, and the group x time interaction approached
signifi cance for active coping, F(1,20) = 3.39, p =
.080, with the singing group showing a decrease in
active coping from post-intervention to follow-up
(though the follow-up mean remained above the
pre-intervention mean) and the comparison group
showing an increase in active coping from post-
intervention to follow-up.
These data suggest that the improvements
observed in active coping among the singers
immediately following the intervention were not
maintained at 6-month follow-up, though their
level of active coping at follow-up was still greater
than prior to the intervention.
Age, pain duration, and gender
A series of one-way ANOVAs (consistent with
the analytic strategy above) were conducted to
examine whether or not groups differed in terms
of their age and pain duration. No signifi cant
group differences in age were found (largest F
= 1.85). Participants in the singing group had a
signifi cantly longer pain duration than those in the
singing FTA group, F(1,43) = 5.51, p = .024, and
a similar difference approached signifi cance when

Singing and chronic pain 60
Journal of Music Therapy
the singing group was compared to the comparison
group, F(1,43) = 3.86, p = .055.
Finally, a series of one-way ANCOVAs with
gender as the between subjects variable and pre-
intervention scores as a covariate revealed that
males and females did not differ signifi cantly
on any of the mood, coping, and perceived pain
difference score variables (largest F = 3.64).
Separate one-way ANOVAs also revealed that
males and females did not differ signifi cantly in
age, F(1,74) = .71, p = .403, or pain duration,
F(1,70) = .30, p = .588.
DISCUSSION
Participants who attended the singing sessions
showed evidence of post-intervention improvement
in active coping, relative to those who failed to
attend, when pre-intervention differences in active
coping were controlled for. While the results suggest
that singing may enhance active coping in some
individuals suffering from chronic pain, the effect
size is small and the differences between groups
just failed to reach conventional signifi cance. This
positive effect was not maintained at six-month
follow-up, although the level of active coping
at follow-up still exceeded the pre-intervention
level. It should also be noted that these fi ndings
are limited by the fact that participants in the failed
to attend comparison group were not randomly
assigned (for reasons discussed earlier).
The singing intervention (relative to the
listening-to-singing comparison group and singing
FTA group) did not have a signifi cant impact on
other mood, coping, or perceived pain variables.
Both singing and comparison groups showed
signifi cant improvements from pre to post-
intervention on the mood, coping, and perceived
pain variables, but there were no time by group
interaction effects. This result is consistent with
the fi ndings of Unwin et al. (2002) who found that
participants exposed to either a single session of
singing, or a listening-to-singing session, showed
signifi cantly lower levels of tension, depression,
anger, fatigue and confusion, and signifi cantly
higher levels of vigour, when tested immediately
following a single half-hour session of singing.
Indeed, positive mood changes in response to
listening to singing have been reliably reported
in the literature (Behrens & Green, 1993; Gfeller,
Asmus, & Eckert, 1991; Landreth, 1974; Wedin,
1972). Further, and contrary to expectations,
multiple singing sessions (nine in the present
study as opposed to one in the study by Unwin
et al., 2002) did not signifi cantly strengthen
improvements in mood among participants in the
singing group when compared to the listening-to-
singing comparison group (who also received nine
sessions of listening to singing).
The results of the present study revealed that
participants in the singing group were less likely
to report pre-intervention psychopathology across
almost all mood, coping, and perceived pain
variables than other participants. While these pre-
intervention scores were employed as covariates
in the main analyses, a type of fl oor effect may
have occurred such that the observation of large
reductions from pre to post-intervention on
mood, coping, and perceived pain variables was
hindered by the lower pre-intervention scores
among the singing group participants. The fact
that participants in the singing group generally
showed less pre-intervention psychopathology
than other participants is particularly interesting
given that their self-reported chronic pain duration
was signifi cantly greater than both listening-to-
singing comparisons and participants assigned
to the singing group who failed to attend the
singing sessions. It may have been the case that
the signifi cantly longer duration of chronic pain
among singing group participants increased the
diffi culty that these individuals experienced
in maintaining improvements in active coping
over the follow-up period. This fi nding may
additionally refl ect the effects of regression to the
mean due to greater increases in active coping
from pre to post-intervention being observed
among the singing group participants compared
with the other groups. Given that improvements
were generally not maintained over the follow-up
period, greater decreases in active coping from
post-intervention to follow-up may be expected
among singing group participants because they
showed greater increases in active coping from
pre to post-intervention.
It is clear from Tables 1 and 2 that all groups
showed improvements in mood, coping, and
perceived pain when tested immediately following
the intervention. This cannot be explained by the
effect of exposure to singing alone (i.e. either
participating or listening) as it was also observed
in the singing FTA group, which consisted of those
who were assigned to the singing condition but
failed to attend any singing sessions. It is more
likely that this general improvement was due to
participation in the ADAPT pain management
program from which participants were recruited.

Singing and chronic pain 61
Journal of Music Therapy
Indeed, it is possible that the effect of the singing
intervention may have been masked in part by the
powerful effect that the intensive ADAPT program
has been found to have on the factors assessed in
this study. Future research will need to attempt to
control for the effect of standard pain management
programs on variables of interest if the additional
benefi cial effects of singing interventions are to be
isolated.
The fact that participants in the singing group
reported signifi cantly greater pain duration
is also somewhat problematic. It seems that
participants who have been in pain for longer,
but are somewhat less disturbed with respect to
mood, coping, and perceived pain, were more
likely to attend the singing sessions. This potential
confound was unavoidable in the present study.
Additional incentives may be required in future
studies to improve retention rates and encourage
as many participants as possible to attend all (or
at least most) singing sessions. In the current
study, the singing sessions had to be conducted at
8am prior to the commencement of the ADAPT
program at 9am. Many potential participants were
unwilling to attend at this hour, hence the poorer
than expected attendance.
Brown et al. (1989) suggested that there were
two distinct qualities to music that may explain its
ability to enhance effective pain coping skills – its
ability to distract attention from the perception
of pain, and its ability to arouse emotional
experiences, which play a cathartic role. Distraction
and emotional arousal have also been argued to
be integral to the amelioration of pain perception
(Maslar, 1986; Melzack, Weisz, & Sprague, 1963;
Siegele, 1974), and it has been suggested that the
emotional reactions elicited by music may serve to
motivate the continuation of effective pain-coping
strategies (Staats, 1986; Staats & Eifert, 1990).
While such explanations provide a rudimentary
framework for understanding the effect of music
on mood, coping, and perceived pain, further
research is clearly required into the mechanisms,
both psychological and physiological, that underlie
such effects.
The results of the present study suggest that
singing offers promise as a component of pain
management programs. Despite the present study
being somewhat underpowered (particularly for
analyses involving follow-up data) due to smaller
than expected sample size and large intra-group
variability, the results suggest that singing can
produce improvements in active coping among
chronic pain sufferers. However, techniques to
enhance the maintenance of these improvements
over time need to be explored, and the benefi ts that
singing may have in enhancing mood and hence
receptiveness and motivation for participation in
standard pain management programs need to be
further explored.
REFERENCES
Aldridge, D. (1991). Creativity and consciousness:
Music therapy in ICU.
Arts in Psychotherapy
, 18,
359-362.
Aldridge, D. (1993). The music of the body: Music
therapy in medical settings.
Advances
, 9, 17-35.
Aldridge, D. (1994). Alzheimer’s Disease: Rhythm,
timing and music as therapy.
Biomedicine &
Pharmacotherapy
, 48, 275-284.
Augustin, P., & Hains, A. (1996). Effect of music on
ambulatory surgery patients’ preoperative anxiety.
AORN Journal
, 63, 750-758.
Bailey, L.M. (1983). Music therapy as an intervention in
pain management. In New York-New Jersey Cancer
Nursing Regional Committee (Eds.),
Handbook on
interventions in pain management
(pp. 39-40). New
interventions in pain management (pp. 39-40). New interventions in pain management
York: Columbia University Comprehensive Cancer
Center.
Bandura, A. (1977).
Social learning theory
. Englewood
Cliffs, NJ: Prentice Hall.
Beck, S.L. (1991). The therapeutic use of music for
cancer-related pain.
Oncology Nursing Forum
, 18,
1327-1337.
Behrens, G. A., & Green, S. (1993). Ability to identify
emotional content of solo improvisation performed
vocally and on three different instruments.
Psychology
of Music
, 21, 20-33.
Blood, A.J., Zatorre, R.J., Bermudez, P., & Evans,
A.C. (1999). Emotional responses to pleasant and
unpleasant music correlate with activity in paralimbic
brain regions.
Nature Neuroscience
, 2(4), 382-387.
Bonny, H.L. (1983).
Music Rx: An innovative program
designed for the hospital setting
. Port Townsend,
WA: ICM West.
Brown, C.J., Chen, A.C.N., & Dworkin, S.F. (1989).
Music in the control of human pain.
Music Therapy
,
8, 47-60.
Burns, J.L., Labbe, E., Arke, B., Capeless, K., Cooksey,
B., Steadman, A., & Gonzales, C. (2002). The
effects of different types of music on perceived and
physiological measures of stress.
Journal of Music
Therapy
, 39, 101-116.
Caine, J. (1991). The effects of music on selected stress
behaviors, weight, caloric and formula intake, and
length of hospital stay of premature and low birth
weight neonates in a newborn intensive care unit.

Singing and chronic pain 62
Journal of Music Therapy
Journal of Music Therapy
, 28, 180-192.
Chlan, L.L. (1995). Psychophysiological responses of
mechanically ventilated patients to music: a pilot
study.
American Journal of Critical Care
, 4, 233-
238.
Co, Y., Eaton, S., & Maxwell, M. (1993). The
relationship between the St Thomas and Oswestry
disability scores and the severity of low back pain.
Journal of Manipulative Physiology and Therapy
,
16, 14-18.
Coleman, J.M., Pratt, R.R., Stoddard, R.A., Gerstmann,
D.R., & Abel, H.H. (1997). The Effects of male and
female singing and speaking voices on selected
physiological and behavioural measures of premature
infants in the intensive care units.
International
Journal of Arts Medicine
, 5, 4-11.
Davis, C.A. (1992). The effects of music and basic
relaxation instruction on pain and anxiety of women
undergoing in-offi ce gynecological procedures.
Journal of Music Therapy
, 29, 202-216.
Dubois, J.M., Bartter, T., & Pratter, M.R. (1995). Music
improves patient comfort level during outpatient
bronchoscopy.
Chest
, 108, 129-130.
Edward, J. (1998). Music therapy for children with
severe burn injury.
Music Therapy Perspectives
, 16,
21-26.
Euler, C. von, (1986) Brain stem mechanisms for
generation and control of breathing pattern. In N.S.
Cherniack and J.G. Widdicombe (Eds.),
Handbook
of Physiology, Section 3, The Respiratory System,
Vol II, Control of Breathing
. American Physiological
Society.
Fairbank, J., Davies, J., Couper, J., & O’Brien, J.
(1980). The Oswestry Low Back Pain Disability
Questionnaire.
Physiotherapy
, 66(8), 271-273.
Flor, H., Behle, D., & Birbaumer, N. (1993). Assessment
of pain-related cognitions in chronic pain patients.
Journal of Psychosomatic Research
, 31, 251–259.
Gaberson, K. (1995). The effect of humorous and
musical distraction on preoperative anxiety.
AORN
Journal
, 62, 784-791.
Gfeller, K., Asmus, E., & Eckert, M. (1991). An
investigation of emotional responses to music and
text.
Psychology of Music
, 19(2), 128-141.
Gibson, L., & Strong, J. (1996). The reliability and
validity of a measure of perceived functional
capacity for work in chronic back pain.
Journal of
Occupational Rehabilitation
, 6, 159–175.
Guzetta, C. (1989). Effects of relaxation and music
therapy on patients in a coronary care unit with
presumptive acute myocardial infarction.
Heart &
Lung
, 18, 609-616.
Hanser, S.B., & Thompson, L.W. (1994). Effects of
a music therapy strategy on depressed older adults.
Journal of Gerontology: Psychological Sciences
, 49,
265-269.
Kaminski, J., & Hall, W. (1996). The effect of soothing
music on neonatal behavioral states in the hospital
newborn nursery.
Neonatal Network
, 15, 45-54.
Kopec, J., & Esdaile, J. (1995). Functional scales for
low back pain.
Spine
, 20, 1943-1949.
Krumhansl, C. L. (1997). An Exploratory Study of
Musical Emotions and Psychophysiology.
Canadian
Journal of Experimental Psychology
, 51, 336-352.
Landreth, J. E. (1974). Effects of music on physiologcal
responses.
Psychology of Music
, 22, 4-12.
Lenton, S. R., & Martin, P. R. (1991). The contribution
of music vs instruction in the musical induction
procedure.
Behaviour Research and Therapy
, 29,
623-625.
Linoff, M.G., & West, C.M. (1982). Relaxation training
systematically combined with music: Treatment of
tension headaches in a geriatric patient.
International
Journal of Behavioral Geriatrics
, 1, 11-16.
Maslar, P.M. (1986). The effect of music on the
reduction of pain: A review of the literature.
The Arts
in Psychotherapy
, 13, 215-219.
McNair, D. M., & Droppleman, L. F. (1981).
Profi le of
Mood States
. San Diego, California: Educational and
Industrial Service.
Melzack, R., Weisz, A.Z., & Sprague, L.T. (1963).
Stratagems for controlling pain: Contributions of
auditory stimulation and suggestion.
Experimental
Neurology
, 8, 239-247.
Miller, A.C., Hickman, L.C., & Lemasters, G.K. (1992).
A distraction technique for control of burn pain.
Journal of Burn Care Rehabilitation
, 13, 576-580.
Mullooly, V.M., Levin, R.F., & Feldman, H.R. (1988).
Music for postoperative pain and anxiety.
Journal of
New York State Nurses
, 19, 4-7.
Nicholas, M. (1989).
Self-effi cacy and chronic pain
.
Paper presented at the Annual Conference of the
British Psychological Society, St. Andrews, UK.
Pfaff, V.K., Smith, K.E., & Gowan, D. (1989). The
effects of music assisted relaxation on the distress
of pediatric cancer patients undergoing bone marrow
aspirations.
Children’s Health Care
, 18, 232-236.
Rider, M.S. (1987). Treating chronic disease and
pain with music-mediated imagery.
The Arts in
Psychotherapy
, 14, 113-120.
Schorr, J.A. (1993). Music and pattern change in chronic
pain.
Advances in Nursing Science
, 15, 27-36.
Sedei Godley, C.A. (1987). The use of music therapy in
pain clinics.
Music Therapy Perspectives
, 4, 24-28.
Sharp, T. J., & Nicholas, M. K. (2000). Assessing
the signifi cant others of chronic pain patients:
the psychometric properties of signifi cant other
questionnaires.
Pain
, 88(2), 135-144.
Siegele, D.S. (1974). Gate control theory.
American
Journal of Nursing
, 74, 498-502.
Sloboda, J. (1991). Music structure and emotional

Singing and chronic pain 63
Journal of Music Therapy
response: Some empirical fi ndings.
Psychology of
Music
, 19, 110-120.
Sloboda, J. A. (1992). Empirical studies of emotional
response to music. In M. R. Jones & S. Hollerman
(Ed.),
Cognitive Bases of Musical Communication
(pp. 33-46). Washington: American Psychological
Society.
Snyder. M., & Chlan, L. (1999). Music therapy.
Annual
Review of Nursing Research
, 17 (1), 3-19.
Spintge, R. (1985). Some neuroendocrinological effects
of socalled anxiolytic music.
International Journal of
Neurology
, 19-20, 186-196.
Staats, A.W. (1986). Behaviorism with a personality:
The paradigmatic behavior assessment approach.
In R.O. Nelson & S.C. Hayes (Eds.),
Conceptual
foundations of behavioral assessment
. New York:
Guilford.
Staats, A.W., & Eifert, G.H. (1990). The paradigmatic
behaviorism theory of emotion: Basis for unifi cation.
Clinical Psychology Review
, 10, 539-566.
Unwin, M.M., Kenny, D.T., & Davis, P.J. (2002). The
effects of group singing on mood.
Psychology of
Music
, 30, 175-185.
Updike, P. (1990). Music therapy results for ICU
patients.
Dimensions of Critical Care Nursing
, 9, 39-
45.
Wade, C., & Tavris, C. (1998).
Psychology
. (5th ed.),
New York: Longman.
Wedin, L. (1972). A multidimensional study of
perceptual-emotional qualities in music.
Scandinavian
Journal of Psychology
, 13, 241-257.
Williams, A., Nicholas, M., Richardson, P., Pither,
C., Justins, D., Chamberlain, J., et al. (1993).
Evaluation of a cognitive-behavioural programme
for rehabilitating patients with chronic pain.
British
Journal of General Practice
, 43, 513–518.
Winter, M., Paskin, S., & Baker, T. (1994). Music
reduces stress and anxiety of patients in the surgical
holding area.
Journal of Post-Anesthesia Nursing
, 9,
340-343.
Wolfe, D.E. (1978). Pain rehabilitation and music
therapy.
Journal of Music Therapy
, 15, 162-178.
Zimmerman, L., Nieveen, J., Barnason, S., & Schmaderer,
M. (1996). The effects of music intervention on
postoperative pain and sleep in coronary artery bypass
graft (CABG) patients.
Scholarly Inquiry for Nursing
Practice: An International Journal
, 10, 153-170.
Zimmerman, L., Pozehl, B., Duncan, K., & Schmitz, R.
(1989). Effects of music in patients who had chronic
cancer pain.
Western Journal of Nursing Research
,
11, 289-309.
Zung, W. W., Richards, C. B., & Short, M. J. (1965).
Self-rating depression scale in an outpatient clinic.
Further validation of the SDS.
Archives of General
Psychiatry
, 13(6), 508-515.