Complementary Therapies in Clinical Practice 20 (2014) 219e223

Contents lists available at ScienceDirect

Complementary Therapies in Clinical Practice

journal homepage: www.elsevier.com/locate/ctcp

Neck arthritis pain is reduced and range of motion is increased by

massage therapy
Tiffany Field a, b, *, Miguel Diego a, b, Gladys Gonzalez a, C.G. Funk c
a
University of Miami School of Medicine, United States
b
Fielding Graduate University, United States
c
Massage Envy, United States

abstract
Keywords: Background: The literature on the effects of massage therapy on neck arthritis pain is mixed depending
Neck arthritis
on the dose level, and it is also based on self-report. In the present study an attempt was made to
Pain
enhance the effects of weekly massage therapy by having the participants massage themselves daily. And
Range of motion
Massage therapy
in addition to self-reports on pain, range of motion (ROM) and the associated ROM pain were assessed
before and after the first massage session and pre-post the last session one month later.
Methods: Staff and faculty members at a medical school who were eligible for the study if they had neck
arthritis pain were randomly assigned to a massage or a waitlist control group (N ¼ 24 per group). The
massage group received moderate pressure massages weekly by a massage therapist plus daily self-
massages. The waitlist control group received the same schedule massages one month after being
control subjects.
Results: The massage group showed significant short-term reductions after the first and last day mas-
sages in self-reported pain and in ROM-associated pain as well as an increase in ROM. Comparisons
between the massage group (N ¼ 23) and the control group (N ¼ 14) on the last versus the first day data
suggested significantly different changes including increased ROM and reduced ROM-associated pain
for the massage group and reduced ROM and increased ROM-associated pain for the control group.
These changes occurred specifically for flexion and right and left lateral flexion motions.
Discussion: These data highlight the importance of designing massage therapy protocols that target the
most affected neck muscle groups and then assessing range of motion and related pain before and after
the massage therapy. Comparisons with other studies also suggest that moderate pressure may
contribute to the massage effects, and the use of daily self-massages between sessions may sustain the
effects and serve as a cost-effective therapy for individuals with neck arthritis pain.
© 2014 Published by Elsevier Ltd.

The literature on neck arthritis is limited, although several

Despite the frequency of massage therapy visits for neck pain,
recent studies have been conducted on neck pain. Thirty to 50%
the research literature is not conclusive. The two reviews of mul-
of adults reportedly complain about neck pain [1]. In one study,
tiple studies basically concluded that more research was needed on
in- dividuals suffering from chronic neck pain were likely to visit
the effectiveness of massage therapy for neck pain [5,6]. In the
pri- mary care providers (72%), chiropractors (40%), physical
older review, only 6 of the studies focused on massage therapy
therapists (35%), orthopedic surgeons (32%) and massage
alone, and these studies had methodological problems including
therapists (28%) [2]. Another study cited neck pain as one of the
not offering more than one treatment session and featuring atypical
most frequent pain conditions (second to back pain) that led to
massage therapy protocols [5]. In the more recent review on studies
visits to complemen- tary/alternative therapists, most often
comparing massage with control groups and other therapies for
chiropractors and massage therapists [3] with almost 1 in 5 visits
neck pain, massage therapy was said to be effective for neck pain
to massage therapists being for neck pain [4].
immediately after treatment, but no data were given on longer
term outcomes [6].
One study attempted to correct these methodological problems
* Corresponding author. University of Miami School of Medicine, United States.
[7]. In this study, improvement was noted in those who received
E-mail address: tfield@med.miami.edu (T. Field).
massage for neck pain following 60-min massages once per week

http://dx.doi.org/10.1016/j.ctcp.2014.09.001
1744-3881/© 2014 Published by Elsevier Ltd.
22 T. Field et al. / Complementary Therapies in Clinical Practice 20 (2014) 219e223

for 4 weeks. However, this study also had several

that can sustain the effects in between sessions. We have used
methodological problems including: 1) the massage protocol was
this paradigm with upper limb arthritis pain [10] and it has been
not standardized;
effective. In this study we recruited 42 adults with rheumatoid
2) no objective measures were used for the massage therapy effects,
arthritis in the upper limbs and randomly assigned them to a
only self-reports by mail; 3) the therapists made other self-care
moderate pressure or a light pressure massage therapy group
suggestions such as stretching which could confound the mas-
[10]. A therapist massaged the affected arm and shoulder once a
sage therapy effects; 4) significant differences were noted on one
week for a 4-week period and also taught the participants self-
self-report measure, i.e. Neck Disability Index, but not on the other,
massage to be done once daily. The moderate versus the light
i.e. the Copenhagen Neck Functional Disability scale; and 5)
pressure massage therapy group had less pain and greater
dichotomized values were used rather than mean values for com-
perceived grip strength following the first and last massage ses-
parisons between groups. Nonetheless, this research group more
sions. By the end of the one month period the moderate pressure
recently conducted a larger trial to correct these problems [8]. In
massage group had less pain, greater grip strength and greater
this randomized control study 60 min massages 2 or 3 times a week
range of motion (greater wrist and elbow flexion and greater
led to decreased neck dysfunction based on the Neck Disability
shoulder abduction).
Index and pain intensity measured on a 10-point pain scale as
In the present study we assessed the effects of moderate pres-
compared to a group who received 30 min massages 2 or 3 times a
sure massage versus being in a waitlist control group on adults who
week and a waitlist control group. They also made some interesting
have neck arthritis pain and limited range of motion. The massage
suggestions about the need to extend the time frame for the mas-
group members were given weekly massages for a one month
sage benefits, for example, by having booster sessions and using
period plus directions on applying the same massages to them-
self-massaging devices.
selves daily in between the massage therapist sessions. The par-
Turning to a different literature on specific types of massage,
ticipants were assessed before and after the first and last massages
recent studies at our research institute have compared the effects of
and were expected to show a reduction in neck arthritis pain and
moderate versus light pressure massage [9,10]. The moderate
increased range of motion. Thus, the unique features of this neck
pressure massage resulted in slower heartrate and blood pressure
pain study were: 1) assessing the effects of massage therapy on
and in EEG (theta) waves suggestive of relaxation [9] and in less
neck arthritis pain; 2) assessing its effects on neck range of motion;
pain and greater range of motion for individuals with rheumatoid
3) complementing the neck massage therapy with self-massages on
arthritis in their upper limbs [10].
the days between massage therapy sessions; and 4) using a direct
This literature highlights the need for a moderate pressure
observation of ROM-related pain versus relying strictly on self-
massage therapy study for neck pain that involves multiple ses-
report measures.
sions, objective measures, and a method such as self-massage

Fig. 1. Consort flow diagram.

1. Method
they were also taught self-massage. The participants were asked
to massage their necks daily and to keep a record of their self-
1.1. Participants
massages. Compliance in conducting the self-massages was esti-
mated to be approximately 80%. The 15-min massages (repeated
Based on our hand and upper limb arthritis studies, a power
twice during the therapist sessions) consisted of moderate
analysis was conducted and suggested that we needed 18 partici-
pressure stroking focused on each side of the neck. The massage
pants per group for 80% power at p¼.05. To account for potential
protocol was designed by the first and third authors (TF, the
attrition (approximately 20% in our previous studies), we recruited
principal investigator, and GG, the massage therapist) to be
48 medical school staff/faculty with neck arthritis pain (diagnosed
sufficiently simple to be used as a self-massage by the
by their physicians). They were then randomly assigned to the
participants (described in Box 1). The participants in the waitlist
control group were

Box 1
Massage to reduce neck arthritis pain protocol.

Right side
1. Glide knuckles vertically from below right ear to back of neck. Repeat 3×.
2. With head in center position, place right fingers under right ear. As head turns slowly to the left, your right fingers glide
un- derneath scalp to back of neck. Repeat 3×.
3. With head turned left, place right fingers under right ear. Do circular movements to back of neck. Repeat 3×.
4. With head back to center, place knuckles below the right ear. As head turns slowly to the right, glide right knuckles down
toward sternum. Repeat 3×.
5. With head back to center, place knuckles below the right ear. As head turns slowly to the left, glide right knuckles down to
right side of neck to the base of the neck. Repeat 3×.
6. With head to center, place right hand on right side at base of neck. Hold in place as head gently stretches toward the left
shoulder (left ear toward left shoulder). Hold the stretch on left side for 3 s and return head to center. Repeat 3×.
Left side
1. Glide knuckles vertically from below left ear to back of neck. Repeat 3×.
2. With head in center position, place left fingers under left ear. As head turns slowly to the right, your left fingers glide
underneath scalp to back of neck. Repeat 3×.
3. With head turned right, place left fingers under left ear. Do circular movements to back of neck. Repeat 3×.
4. With head back to center, place knuckles below the left ear. As head turns slowly to the left, glide left knuckles down
toward sternum. Repeat 3×.
5. With head back to center, place knuckles below the left ear. As head turns slowly to the right, glide left knuckles down to right
side of neck to the base of the neck. Repeat 3×.
6. With head to center, place left hand on left side at base of neck. Hold in place as head gently stretches toward the right
shoulder (right ear toward right shoulder). Hold the stretch on right side for 3 s and return head to center. Repeat 3×.
Back of neck (Do this section with both hands)
1. Place right knuckles under right ear and left knuckles under left ear. Glide knuckles vertically to back of neck. Repeat 3×.
2. Place right fingers under right ear and left fingers under left ear. Do circular movements with your fingers toward back of neck.
Repeat 3×.
3. Place right fingers and left fingers at back of neck under base of skull (fingers facing each other). Glide fingers slowly
and gently toward the ears. Continue down the back of the neck to the base of the neck. As you do this your left and
right fingers move horizontally to the sides of the neck. Repeat 3×.

massage or waitlist control groups (N 24 per group) following

¼ assessed on the first day and last day of the first month and then
informed consent. After attrition, the sample size was 37 (N 23 in
¼ were given massages and practiced the self-massages on the same
the massage group and N 14 in the waitlist control group) (See
¼ schedule as the massage group during the second month.
Fig. 1). The greater attrition in the control group, although self-
reported as “scheduling difficulties” could have also related to
remission in neck pain over the waitlist period. This lesser power
1.3. Assessments
would be expected to attenuate the predicted treatment effects.
The majority of the participants were female (77%), averaged 47
The participants completed the following assessments before
years of age, were middle income and were distributed 53% His-
and after the massage sessions on the first and last days of the 4-
panic, 20% Caucasian, 12% African-American, 9% Asian and 6% week study period. The control group was assessed only once on
other. The two groups did not differ on these variables.
the first and last days of a 4-week period.
The primary outcome measures were self-report scales and
range of motion (ROM) assessments along with ratings of ROM-
1.2. Procedure
associated pain. The self-report measures included: 1) the Faces
Rating Scale, a visual analogue scale of 10 faces ranging from very
The participants in the massage group were massaged by a
happy to very unhappy for assessing immediate pain [11]; 2) the
licensed massage therapist once per week for a 4-week period, and
Profile of Mood States that measures stress [12]; and 3) the range of
motion variables measured by a goniometer including: a) neck
Table 2
flexion (touching chin to chest); b) neck extension (touching head Means for pre-post range of motion measures on first and last days massage
to back); c) right lateral flexion (touching right ear to right shoul-
group. Measure First day Last day F
der); d) left lateral flexion (touching left ear to left shoulder); e)
right lateral rotation (touching chin to right shoulder); f) left lateral PRE Post PRE Post
rotation (touching chin to left shoulder); g) nodding head yes; and Flexion 2.0 2.0 2.0 2.0
h) shaking head no. The range of motion measures were rated as Flexion pain 1.2 .4 .3 .1 7.67b
Extension 1.8 2.0 2.0 2.0
0 if unsuccessful movements, 1 if halfway to full range and 2 if
Extension pain 1.6 .9 .8 .3 7.47b
full range of motion. Each of the range of motion measures was Rt. Lat. Flex. 1.5 1.7 1.7 1.9 5.28a
also rated for associated pain on the Faces Rating Scale as 0 no Rt. Lat. Flex. Pain 2.0 1.2 1.1 .5 25.92d
pain (a happy face) to 10 (an extremely unhappy face). The Lft. Lat. Flex 1.6 1.8 1.6 1.8 6.75a
individual ROM scores and individual ROM pain scores were Lft. Lat. Flex. Pain 2.0 1.3 1.4 .5 22.80c
Rt. Lat. Rotation 1.8 2.0 1.9 2.0
added for total ROM and total pain scores. Rt. Lat. Rot. Pain .9 .6 .4 .3
Lft. Lat. Rotation 1.7 2.0 1.8 2.0 5.46a
Lft. Lat. Rot. Pain .9 .7 .7 .3
1.4. Data analyses
Nodding yes 1.9 2.0 2.0 2.0
Nodding yes pain .8 .4 .4 .1 4.04a
Repeated measures multivariate analyses of variance (MAN- Shaking no 1.9 2.0 1.9 2.0
OVAS) were performed (using SPSS) with: 1) the massage versus Shaking no pain .5 .4 .3 .0
waitlist control group as the between groups measure and the pre- a
p < .05.
session measures from the first and last day as the repeated mea- b
p < .01.
sure; and 2) the pre-post session values on the first and last day as c

the repeated measure for the massage group. These were followed
by ANOVAs for each of the measures and Bonferroni t tests for
multiple pairwise-comparisons.
As can be seen in Table 1 on pre versus post massage session
values, the following statistically significant immediate changes
occurred for the massage group from pre to post massage on the
first and last days of the study including (see means, standard de-
viations, F values and p levels for massage group in Table 1): 1)
range of motion total scores increased; 2) range of motion-
associated pain total scores decreased; 3) Faces Rating Scale
scores decreased; and 4) Profile of Mood States (stress) scores
decreased.
Table 2 on the more specific range of motion scores pre versus
post massage sessions illustrates the statistically significant im-
mediate changes in the more specific range of motion and range of
motion-associated pain scores for the massage group from pre to
post massage on the first and last days of the study including (see
means, F values and p levels in Table 2): 1) flexion pain decreased;
2) extension pain decreased; 3) right lateral flexion increased; 4)
right lateral flexion pain decreased; 5) left lateral flexion increased;
6) left lateral flexion pain decreased; 7) left lateral rotation
increased; and 8) nodding yes pain decreased.
As can be seen in Table 3 on comparisons between the massage
and control groups, statistically significant group by time interac-
tion effects (last day versus first day) were revealed suggesting that
the massage versus the control group experienced the following
changes (see Table 3 for means, standard deviations, F values and p
levels for the massage versus control group comparisons): 1) an
increased range of motion for the massage group versus a decrease
for the control group; and 2) a decreased range of motion-
associated pain versus an increase for the control group.
d
p < .005.
p < .001.
Table 4 shows statistically significant changes in the longer term
more specific range of motion and range of motion-associated pain
scores for the massage versus the control group suggesting the
following (see Table 4 for means, F values and p levels): 1) a
decrease in flexion pain; 2) a decrease in extension pain; 3) an
increase in right lateral flexion; 4) a decrease in right lateral flexion
pain; and 5) a decrease in left lateral flexion pain.
2. Discussion
The immediate and long-term positive effects of massage ther-
apy on neck ROM and pain in this study are consistent with those
we previously reported for changes in ROM and pain following
moderate pressure massage for arthritis in the upper limbs [10].
Others have noted a reduction in neck pain following massage, but
by self-report, not direct observation of ROM-related pain [8].
Their results were inconsistent with ours in that they only observed
reduced pain after 5 weeks of 60-min massages two or three times
weekly (as opposed to their lower dose group receiving 30-
min massages two or three times weekly) [8]. One possible expla-
nation for the positive effects following shorter and less frequent
massages in our study (30 min weekly massage sessions for 4
weeks) is our use of moderate pressure massage [9], although it is
not clear what pressure was used in their study [8]. Another po-
tential explanation is that the use of daily self-massages in our
study added significantly to the much lower dose massage therapy
protocol. However, the use of self-report pain scales in their study

Table 3
Table 1 Means (standard deviations in parentheses) for massage and control group first day
Means (standard deviations in parentheses) for pre-post measures on first and last versus last day measures.
days for massage group.
Measure Group F
Measure First day Last day F
Massage Control
PRE Post PRE Post
First day Last day First day Last day
Range of motion 14.5 (1.9) 15.3 (1.3) 15.0 (1.4) 16.6 (1.8) 14.10b
ROM pain 11.0 (8.6) 5.6 5.5 (5.8) 2.3 (1.2) 28.15b Range of motion 14.5 (1.9) 15.0 (1.4) 14.5 (1.8) 14.0 (1.9) 7.35a
(4.9) ROM pain 11.0 (8.6) 5.5 (4.8) 8.3 (5.9) 11.8 (7.6) 10.99b
b
Faces 8.3 (2.0) 6.2 (5.1) 2.0 (2.0) .61 (1.1) 56.99 Faces 3.3 (2.0) 2.0 (2.0) 3.4 (1.4) 3.6 (1.4) 3.60
POMS 3.2 (2.6) 2.3 (2.2) 2.2 (2.1) 1.0 (.2) 10.40a POMS 3.2 (2.6) 2.2 (1.1) 2.7 (1.8) 2.9 (1.7) .85
a
p < .005. a
p < .01.
b
p < .001. b
p < .005.
Table 4
Means for massage and control group range of motion and ROM related pain massage is accompanied by decreased heartrate, suggesting a
measures for first and last days. relaxed state [9]. We have suggested elsewhere that stimulation of
Measure Group F
pressure receptors and the resulting increase in vagal activity and
serotonin levels (the body's natural pain suppressor) is one po-
Massage Control
tential underlying mechanism for pain relief [10]. A related mech-
First day Last day First day Last day
anism we have suggested is based on substance p decreases
Flexion 2.0 2.0 1.9 1.9 [substance p causing pain] following massage [10]. Further research
Flex pain 1.2 .3 .9 1.3 5.34a
Exten 1.8 2.0 1.7 1.8
is clearly needed to explore the potential underlying mechanisms
Exten pain 1.6 .8 .9 1.2 4.63a for the relief of neck pain and other pain syndromes following
Rlat flex 1.5 1.9 1.7 1.4 13.18b massage therapy
Rlat flex pain 2.7 1.1 2.2 3.1 15.21b
Llat flex 1.6 1.6 1.7 1.4
Conflict of interest statement
Llat flex pain 2.3 1.4 1.7 2.9 5.51a
Rlat rot 1.9 1.9 1.8 1.9 None declared.
Rlat rot pain .9 .4 .9 1.1
Llat rot 1.8 1.8 1.7 1.6 Acknowledgments
Llat rot pain 1.0 .7 1.5 1.2
No dyes 1.9 2.0 2.0 2.0
No dyes pain .8 .4 .4 .7
We wish to thank the participants in this study and acknowl-
Shake no 1.9 2.0 1.9 2.0 edge the funding by Massage Envy to support this study.
Shake no pain .5 .3 .0 .2
a
p < .05. References
b
p < .001.
[1] Hogg-Johnson S, van der Vewlde G, Carroll LJ, Holm LW, Cassidy JD, Guzman
J, et al. The burden and determinants of neck pain in the general population:
results of the bone and joint decade 2000e2010 task force on neck pain and
[8] and the more direct measure of pain responses to ROM mea- its associated disorders. Spine 2008;33:539e55.
sures in our study makes cross-study comparisons difficult. None- [2] Goode AP, Freburger J, Carey T. Prevalence, practice patterns and evidence for chronic
theless, the paradigm of combining therapist-delivered massage neck pain. Arthritis Care Res 2010;62:1594e601.
[3] Barnes PM, Bloom B, Nahin RL. Complementary and alternative medicine use
with daily self-massages has been effective now in at least 2 studies among adults and children: U.S. In: Complementary and alternative medicine
and suggests that this may be a more costeeffective therapy for use among adults and children. National Center for Health Statistics; 2008.
individuals with arthritis pain and one that may have more sus- [4] Cherkin DC, Deyo RA, Sherman KJ, Hart LG, Street JH, Hrbek A, et al. Charac-
teristics of visits to acupuncturists, chiropractors, massage therapists and naturopathic
tainability [10]. physicians. J Am Board Fam Pract 2002;15:463e72.
The assessment of the different neck range of motion [5] Haraldsson BG, Gross AR, Myers CD, Ezzo JM, Morien A, Goldsmith C, et al.
measures and the associated pain measures revealed that flexion Massage for mechanical neck disorders. Cochrane Database Syst Rev 2006;3:
CD004871.
and espe- cially left and right lateral flexion pain were most
[6] Brosseau L, Wells GA, Tugwell PLK, Casimiro L, Novikov M, Loew L, et al.
prevalent at baseline, suggesting that cell phone use may be Ottawa panel evidence-based clinical practice guidelines on therapeutic
contributing to neck arthritis pain. Anecdotally we noted that massage for neck pain. J Bodyw Mov Ther 2012;16:300e25.
the individuals reporting the most severe pain were librarians, [7] Sherman KJ, Cherkin DC, Hawkes RJ, Miglioretti DL, Deyo RA. Randomized trial
of therapeutic massage for chronic neck pain. Clin J Pain 2008;15:233e8.
possibly deriving from their excessive computer use. The flexion [8] Sherman KJ, Cook AJ, Kahn JR, Wellman RD, Hawkes RJ, Kahn JR, et al. Five-
and the left and right lateral flexion and flexion-associated pain week outcomes from a dosing trial of Therapeutic massage for chronic neck
were also the most positively affected by the massage therapy, pain. Ann Fam Med 2014;12:112e20.
[9] Diego M, Field T, Sanders C, Hernandez-Reif M. Massage therapy of moderate
highlighting the importance of designing massage therapy and light pressure and vibrator effects on EEG and heart rate. Int J Neurosci
protocols that target muscle groups affected by those motions 2004;114:31e44.
and then assessing those specific range of motion measures and [10] Field T, Diego M, Delgado J, Garcia D, Funk CG. Rheumatoid arthritis in upper
limbs benefits from moderate pressure massage therapy. Complement Ther Clin
the associated pain.
Pract 2013;19:101e2.
The underlying mechanism for the relief of neck arthritis pain is [11] Wong-Baker Faces Pain Rating Scale Foundation. 2009.
not clear. We have reported elsewhere that moderate pressure [12] McNair D, Lorr M, Droppleman L. POMS-San Diego. 1971.