Musculoskeletal Science a n d Practice 45 (2020) 10207 0

Contents
listsavailableatScienceDirect

Musculoskeletal Science and Practice

journalhomepage:www.els
evier.com/lo
cate/msksp

Original article

Effectiveness of deep tissue massage therapy, and supervised strengthening

and stretching exercises for subacute or persistent disabling neck pain. The
Stockholm Neck (STONE) randomized controlled trial
Eva Skillgate a,b,1, Oscar Javier Pico-Espinosa a,*,1, Pierre Cot^ �e c, Irene Jensen d, Peter Viklund
a,b
, Matteo Bottai e, Lena W. Holm a
a
Musculoskeletal and Sports Injury Epidemiology Center, Institute of Environmental Medicine, Karolinska Institutet, Stockholm, Sweden b
Naprapathogskolan € – Scandinavian College of Naprapathic Manual Medicine, Stockholm, Sweden
c
Faculty of Health Sciences and UOIT-CMCC Centre for the Study of Disability Prevention and Rehabilitation, University of Ontario Institute of
Technology, Toronto, Canada
d
Unit of Intervention and Implementation Research for Worker Health, Institute for Environmental Medicine, Karolinska Institutet, Stockholm,
Sweden e Unit of Biostatistics, Institute of Environmental Medicine, Karolinska Institutet, Stockholm, Sweden

ARTICLEINFO ABSTRACT

Keywords: Objective: To compare the effectiveness of deep tissue massage, supervised strengthening and stretching exercises, and a
Neck pain combined therapy (exercise followed by massage) (index groups), with advice to stay active (control group).
Musculoskeletal manipulations
Methods: Randomized controlled trial of 619 adults with subacute or persistent neck pain allocated to massage
Manual therapies
(n ¼145), exercise (n ¼160), combined therapy (n ¼169) or advice (n ¼147). Primary outcomes were minimal clinically
Patient education
Complementary therapies/methods important improvements in neck pain intensity and pain-related disability based on adapted questions
Treatment outcome

from the Chronic Pain Questionnaire. Secondary outcomes were perceived recovery and sickness absence. Outcomes were
measured at seven, 12, 26 and 52 weeks.
Results: We found improvement in pain intensity favouring massage and combined therapy compared to advice; at seven
weeks (RR ¼1.36; 95%CI:1.04–1.77) and 26 weeks (RR ¼1.23; 95%CI:0.97–1.56); and seven (RR ¼1.39; 95%CI:1.08–1.81) and
12 weeks (RR ¼1.28; 95%CI:1.02–1.60) respectively, but not at later follow- ups. Exercise showed higher improvement of
pain intensity at 26 weeks (RR ¼1.31; 95%CI:1.04–1.65). Perceived recovery was higher in the index groups than in the
advice group at all follow-ups. We found no consistent differences in pain related disability or sickness absence.
Conclusions: In this study, at 12-months follow-up, none of the index therapies were more effective than advice in terms of
pain intensity in the long term or in terms of pain-related disability in the short or long term. However, the index therapies
led to higher incidence of improvement in pain intensity in the short term, and higher incidence of favorable perceived
recovery in the short and in the long term than advice. Trial registration: ISRCTN01453590. Registered 3 July 2014.

1. Introduction
Neck pain is a leading cause of disability worldwide and its burden
continues to grow due to its high frequency among working population,
which translates into high costs due to sick absence and productivity loss
(Fejer et al., 2006; Global burden of disease, 2015; Hansson and
Hansson, 2005). Personal, clinical, and psychosocial factors influence its
prognosis (Carroll et al., 2008a, 2008b, 2008c). Therefore, the clinical
management of neck pain is challenging and few scientifically proven
effective treatments are available to assist clinicians in the choice of care of
the patients (Guzman et al., 2008; Wong et al., 2016). In the past, research on
clinical interventions focused on pain relief and failed to

* Corresponding author. Musculoskeletal and Sports Injury Epidemiology Center, Institute of Environmental Medicine, Karolinska Institutet, Box 210, SE-17177, Stockholm, Sweden.
E-mail address: Oscar.pico.espinosa@ki.se (O.J. Pico-Espinosa). 1 Share
first authorship. https://doi.org/10.1016/j.msksp.2019.102070
Received 13 May 2019; Received in revised form 25 September 2019; Accepted 8 October 2019
Available online 14 October 2019
2468-7812/© 2019 The Authors. Published by Elsevier Ltd. This is an open access
article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
E. Skillgate et al.
consider the multifaceted prognosis of neck pain including patient centered
outcomes. Therefore, research supports that effective clinical interventions
must target pain relief, promote active coping and reassure patients without
promoting iatrogenic disability. Previous studies suggest that educational
interventions, exercise training, mobilization, manipulation, analgesics,
acupuncture and low-level laser may provide short-term benefits (Guzman et
al., 2008; Wong et al., 2016).
Non-specific neck pain has multifaceted etiology and is commonly
managed with massage and exercise. Various local and systemic mechanisms
of action have been proposed to describe the effect of massage in the
treatment of pain, including reorganization of muscle fibers, improvement of
circulation and influence on the immune function (Weerapong et al., 2005;
Rapaport et al., 2012). For exercise, some described mechanisms include
increased proprioception, muscle activation and increased strength (Koltyn et
al., 2014; Runhaar et al., 2015) The literature suggests that deep tissue
massage or clinical massage together with advice may be helpful for neck
pain, but the evidence supporting such intervention is weak (Guzman et al.,
2008; Wong et al., 2016). Similarly, supervised strengthening, range-of-
motion and flexibility exercises are more effective than a waiting list, but
strengthening exercises alone are not superior to home range-of-motion or
stretching exercises (Southerst et al., 2014). In summary, the long term effect
of these commonly used interventions on disabling neck pain needs to be
further evaluated.
We designed a randomized clinical trail with the aim to compare the
effectiveness of deep tissue massage, supervised strengthening exercise and
stretching, and a combined therapy (exercise followed by deep tissue
massage) versus advice to stay active in patients with subacute or persistent
neck pain. We hypothesized that deep tissue massage and/or supervised
strengthening exercise and stretching would lead to greater reduction in pain
intensity, pain-related disability and improvement in perceived recovery and
a lower risk of sickness absence.
2. Materials and methods
2.1. Study design, setting and participants
The trial was registered in the ISRCTN registry on 3 July 2014
(ISRCTN01453590). Participants enrolled in the Stockholm Neck Trial (STONE
trial) (Skillgate et al., 2015) were recruited between September 2014 and
December 2015. This was done through advertisements in a free daily
newspaper of high circulation in Stockholm and, by internal advertising
among employees in the public sector. We included individuals aged 18–70
years with subacute (30–90 days duration) or chronic 0 days duration) non-
specific disabling neck pain with or without headache and/or radiating
symptoms. Those with pain intensity <2/10 and disability <1/10 on Numerical
Rating Scales (NRS) were not included (Von Korff et al., 1992). Further, those
with a history of cancer (past five years), pregnancy, severe skin disorders,
sickness absence related to neck surgery, prolapsed disc, spondylolisthesis,
fracture, spinal stenosis, arthritis, osteoporosis, recent neck trauma (in the
past 48 h), severe night pain, steroids use, drug abuse, pain debuting after 55
years, having received treatment by a manual therapist for the current
complaint, signs of infection (Sizer et al., 2007), without access to a
smartphone with connection to internet and those not being able to
communicate in Swedish were not included.
2.2. Randomization
Prior to the study start, an independent research coordinator prepared
800 sequentially numbered sealed envelopes in blocks of 160 containing one
of the three index groups (deep tissue massage, strengthening and stretching
exercises or, combined therapy) or the reference group (40 each). Cards
indicating the treatment arm were placed in a sealed urn and randomly
selected one by one to be placed in the envelopes. Participants were
numbered according to the order in which they contacted the study
coordinator, who made a first assessment of inclusion/exclusion criteria by
phone, and filled out part one of the baseline questionnaire (baseline
questionnaire A) after participants had given their informed oral consent.
2
Musculoskeletal Science and Practice 45 (2020) 102070
Those who fulfilled this part were referred to the research clinic set up in the
Scandinavian College of Naprapathic Manual Medicine in Stockholm for an
appointment within a week. All study participants received written
information about the trial at enrolment.
At the clinic, the study participants filled out part two of the baseline
questionnaire (baseline questionnaire B). Then, they underwent a clinical
examination, were included in the trial if they met the inclusion/ exclusion
criteria based on the therapist’s clinical judgement and finally, the envelope
with the assigned number of the study participant was opened by the
therapist, and the result of the randomization was revealed and treatment
started immediately. A more detailed description of the procedures is
presented in the study protocol (Skillgate et al., 2015).
2.3. Interventions
The interventions (deep tissue massage, strengthening and stretching
exercises, combined therapy of deep tissue massage and strengthening and
stretching exercise and, advice to stay active) are described in detail in
Appendix A. The interventions were delivered by 30 therapists with
experience in massage and exercise therapy. These therapists were not part
of the research team. All therapists were either naprapathy students or
licenced naprapaths (manual medicine professionals who focus on
management of pain and disability in the musculoskeletal system). Both
massage therapy and strengthening and stretching exercises are commonly
used by naprapaths in the management of musculoskeletal pain. For students
of naprapathy to be able to participate in the provision of the interventions
within the trial, they must have had at least 3 years of experience in the field
of the interventions (for example, as masseur or personal trainer). Prior to
trial start, all therapists received two sessions of 3 h each with a standardized
training in delivering all four interventions. In addition, follow-up meetings
took place once every semester (and if needed) to ensure adherence to the
protocols. The intensity of therapy was adapted to the needs of the patient.
The treatment duration was six weeks and the number of visits was limited to
six for massage, exercise and combined therapy, and three for the advice
group.
2.4. Deep tissue massage therapy
Deep tissue massage therapy included techniques tailored to pain
symptomatology (Ernst et al., 2006). The massage targeted the painful area
(upper back and neck, and if indicated, also jaw and/or chest) and was
delivered to tolerance, so that it was perceived as beneficial without reaching
pain intensity exceeding 5/10. Techniques varied from effleurage to firm
motion involving compression and pressure release, and deep muscle/fascia
massage to areas that produced concordant signs. Sessions lasted 45 min,
including 10 min for anamnesis.
2.5. Supervised strengthening and stretching exercises therapy
This therapy consisted of the following: cranio-cervical flexion exercises
for activation of m. longus colli and m. rectus capitis; push-ups for
strengthening of m. pectoralis major, m. pectoralis minor and associated
chest musculature; lying pulldowns for strengthening of m. serratus anterior
and m. trapezius pars ascendens; isometric exercises for strengthening of
deep extensors of the neck; contraction of m. trapezius pars ascendens and;
stretching of m. pectoralis major, m. pectoralis minor, m. masseter, m.
temporalis and mm. pterygoidei. The exercises were supervised by the
therapist and delivered at three levels of intensity depending on participants’
ability, tolerance, stamina, activities of daily living and aiming to provide the
most benefits. We instructed participants to repeat the exercises at home
once or twice per week. To assist with compliance, we filmed participants and
gave comments when doing the exercises at the clinic (on their smartphone)
to use at home. Sessions lasted 45 min, including 10 min for anamnesis.
2.6. Deep tissue massage therapy and supervised exercise therapy
E. Skillgate et al. Musculoskeletal Science and Practice 45 (2020) 102070

This group received exercise therapy followed by massage according the
protocols described above but allocating 25 min to each modality instead (50
min in total). Additional 10 min were used for anamnesis.
2.7. Advice to stay active
Our control intervention was advice to stay active. Participants allocated
to this group had a motivating discussion with the therapist, based on a
booklet they received. They attended up to three visits with a therapist. The
booklet included evidence-based information about back and neck pain
(Jensen et al., 2004). The advice aimed to educate participants about the
common occurrence of back and neck pain, the psychology of the condition,
the misconceptions about back and neck pain and the importance of
returning to normal activities. The visits with the therapist aimed to
personalize the education and discussion to the participant’s condition.
Finally, participants were informed about when to seek care. Sessions lasted
30 min, including 10 min for anamnesis.
2.8. Data collection and follow-up measurements
Questionnaires were filled in at baseline including values for the primary
and secondary outcomes and an extensive number of covariates. We
measured the primary outcomes and perceived recovery using web-based
questionnaires at weeks 7, 12, 26 and 52. Information on sickness absence
was collected at baseline and at 12, 26 and 52 weeks. 2.9. Primary outcomes
Pain intensity and pain related disability in the past four weeks were
measured with an adapted version of the Chronic Pain Grade questionnaire
(CPQ) (Von Korff et al., 1992; Smith et al., 1997; Paanalahti et al., 2016;
Skillgate et al., 2007). We modified the recall period from six months to four
weeks to improve the reliability and match with the frequency of our follow-
ups. The original classification in grades of pain was therefore, not possible to
perform. In the CPQ, pain intensity is measured with three items (current,
worst and average pain) and pain-related disability is also measured with
three items (pain interference with daily activities; with the ability to
participate in recreational, social and family activities; and the ability to work,
including housework). All items were answered with a number between 0
and 10 (NRS-11, where 0 meant no pain at all and 10 maximum imaginable
pain. For pain-related disability, 0 meant No interference/no change and 10
Unable to carry on activities/extreme change). A mean score was constructed
for pain intensity and one for pain-related disability. We calculated the
difference between the mean scores at baseline and each follow-up. We used
the literature (Kovacs et al., 2008) to define the primary outcomes: minimal
clinically important improvement (MCII) as a reduction of �2/10 for pain,
and of �1/10 for pain-related disability. 2.10. Secondary outcomes
All analyses followed the intention to treat approach. To describe the
population, we calculated the mean and the standard deviation for normally
distributed continuous variables, the median and the interquartile range for
not-normally distributed continuous variables and proportions for categorical
and dichotomous variables. For the primary outcomes, results are reported
as the proportion of participants achieving MCII (�2/10 for pain, and �1/10
for disability). We used generalized estimating equations (GEE) with binomial
family and log link function to analyze the effect of the therapies. We used an
exchangeable correlation structure (Twisk, 2013) and included an interaction
term between treatment and time in the model. Thereafter, we did a linear
combination of estimators (Stata Lincom command (StataCorp, 2015)) using
the estimates of treatment, time and the interaction term. The results are
presented as relative risks (RR) with 95% confidence intervals (CI). In addition,
Number Needed to Treat (NNT) was calculated for the primary and secondary
outcomes as the inverse of the difference in proportions between each of the
groups and advice to stay active (NNT ¼1/EER-CER, where EER is
“experimental event rate” (in this case, proportion in the massage, exercises
or combined therapy group) and CER is “control event rate” (in this case,
proportion in the advice group)). The NNT indicates how many participants
one would need to treat with the index treatment (massage, exercises or
combined therapy) instead of the reference treatment (advice) to achieve
one successful case. High values indicate small differences (should be
interpreted as “many persons need to be treated with the index treatment
instead of the reference treatment in order to obtain an additional case of
recovery”). Negative values indicate lower proportion of the outcome than
advice to stay active. All analyses were conducted using STATA 14.0
(StataCorp, 2015).
The trial was approved by the Regional Ethic Committee in Stockholm
(Dnr: 2014/755-31/3).
3. Results
We screened 1514 individuals and enrolled 621 participants (Fig. 1).
Treatment group characteristics were similar at baseline (Table 1). The 52-
weeks follow-up rate was highest for massage therapy (94%) and lowest for
advice to stay active (79%). The average age of the sample was 46 years and
69% were women. Most participants reported pain duration of more than 12
months and 77% used medication for their pain (Table 1). The average
number of visits was 2.4 for advice to stay active, 5.8 for massage, 5.0 for
exercise and 5.5 for massage and exercise.
3.1. Primary outcomes

Self-perceived recovery was measured with a global perceived effect scale

by asking: “How do you feel your symptoms in the neck have changed since
you joined the study?” (Kamper et al., 2010; Dworkin et al., 2005; van der
Windt et al., 1998) Participants who reported to be significantly improved or
completely pain-free (in comparison to somewhat improved, no change,
somewhat worsened or significantly worsened) were classified as recovered.
Sickness absence was self-reported. Participants who had missed at least half
a day of work due to neck pain in the previous follow-up period were
classified as having had a sickness absence. 2.11. Adverse events

We measured adverse events at every return visit to the clinic with a

questionnaire. 2.12. Sample size

Based on results from our previous research (Skillgate et al., 2010), we

estimated a sample size of 600 subjects (150 in each group) to detect a
relative risk of 1.2–1.3 at twelve months for MCII in pain intensity and
disability with a power of 80% and an alpha level of 0.05.

2.13. Statistical analysis

3
E. Skillgate et al.
At seven and 12 weeks follow-up, participants in the massage and the
combined therapy groups had lower mean pain intensity than participants in
the advice group (Fig. 2A and B). The mean of pain related disability for
combined therapy and exercise were lower than for advice after 12 weeks.
Compared to advice, those receiving combined therapy (RR ¼1.39; 95%
CI: 1.08–1.81) or massage alone (RR ¼1.36; 95% CI: 1.04–1.77) were more
likely to report MCII in pain intensity at seven weeks (Table 3). At 12 weeks,
those in the combined therapy group were more likely than those in the
advice group to reach MCII in pain intensity (RR ¼1.28, 95% CI: 1.02–1.60). At
26 weeks, massage (RR ¼1.23; 95% CI: 0.97–1.56) and exercise (RR ¼1.31;
95% CI: 1.04–1.65) were more likely to show MCII in pain intensity than
advice. We did not find consistent differences between groups at 52 weeks in
pain intensity.
We did not find consistent differences in pain related disability between
the groups.
3.2. Secondary outcomes
The incidence of perceived recovery at 52 weeks was 20% in the advice
group, 27% in the exercise group (NNT ¼15), 35% in the massage group (NNT
¼7) and 40% in the combined therapy group (NNT ¼5) (Tables 2 and 3).
Compared to the advice group, those in the other three interventions were
more likely to report self-perceived recovery during follow-up (Table 3): at
seven weeks; massage (RR ¼3.29; 95% CI: 2.00–5.42); exercise (RR ¼1.66;
95% CI: 0.95–2.89); combined therapy (RR ¼3.01; 95% CI: 1.82–4.96). At 52
weeks: massage (RR ¼1.74; 95% CI: 1.13–2.67); exercise (RR ¼1.33; 95% CI:
0.84–2.09); combined therapy (RR ¼1.99; 95% CI: 1.32–3.00) (Table 3). The
risk of having at least one day of sick absence due to neck was similar in all
groups.
There was no difference in the number of visits to additional healthcare
providers at 52 weeks across groups, however, 23% of the participants in the
advice group visited a masseur during the first three months compared to 12,
14 and 13% in massage, exercise and combined therapy groups respectively.
3.3. Adverse events
No participants developed serious (life threatening, resulting in
hospitalizations or in a significant change of the treatment strategy) adverse
events. One person in the massage group who reported highly bothersome
dizziness discontinued treatment after the third session. A full report on
occurrence and risk of adverse events from the STONE trial will be published
separately.
4
Musculoskeletal Science and Practice 45 (2020) 102070
4. Discussion
We aimed to determine the effectiveness of deep tissue massage,
supervised strengthening and stretching exercise therapy and a combination
of both for the management of subacute and persistent neck pain compared
to advice to stay active. We did not find massage alone, exercise alone nor
the combination of both, more effective than advice to stay active in the long
term regarding the primary outcomes. This is also illustrated by the fact that
the NNT (which is a rough indicator of the direction and magnitude of the
effect) in some comparisons for the primary outcomes were high or in some
instances even negative. However, we found that massage alone or with
exercises were more effective regarding a clinically meaningful improvement
in pain intensity than advice to stay active in the short term, and that
massage and exercise alone were more effective in the mid term. Regarding
the patient centered secondary outcome perceived recovery, we found short
and long term effects of deep tissue massage, supervised strengthening and
stretching exercise therapy and a combination of both therapies.
Our finding that massage alone or with exercises was effective in reducing
pain in the short term but not in the long term, and that perceived recovery
was reported more often in these groups than in the advice to stay active
group in the short and long-term should not be viewed as conflicting. In
individuals with persistent pain, self-perceived recovery might not necessarily
imply resolution of pain. Rather, it may suggest that participants have
readjusted to pain (the person has learned how to cope with pain), redefined
what it means to be healthy (the person has redefined health) or reached an
acceptable quality of life for them (Beaton et al., 2001; Hush et al., 2009). Our
findings suggest that massage, exercise or a combination of both may favour
these outcomes. We think that the observed effects might also be due to
mechanisms beyond the immediate effects of deep tissue stimulation and the
progressive muscle strengthening. For instance, the effect of being taken care
of and the relations of empathy that took place along repeated patient-
therapist interaction may have had an effect in overall patient satisfaction,
anxiety and distress, facilitating patient enablement (Derksen et al., 2013).
4.1. Comparison with previous studies
Exercise therapy was not more effective to reduce pain intensity or pain
related disability than advice to stay active in the short or the long term. A
recent systematic review found that supervised combined exercises
(strengthening, stretching/range of motion and flexibility) twice
Table 1
Baseline characteristics of the study participants by treatment group.
Massage
n ¼ 145
Age, mean (SD) 48(14)
E. Skillgate et al. Musculoskeletal Science and Practice 45 (2020) 102070

Women 97 67% (Ludvigsson et al., 2015). Likewise, a study on participants with non-specific
Education
neck pain studying strength and endurance showed better results at reducing
12 years or less 43 30%
More than 12 years 102 70%
pain intensity and disability at one year follow-up compared to a control
Occupation category (Statistics Sweden, 2012) group of advice on stretching and aerobic exercise (Ylinen et al., 2003).
Managerial or high university degree 56 39% Discrepancies in the results might partly be due to the choice of the
University degree, administration and client-oriented 41 28% comparison group. Although participants in the exercise group were advised
Service, care, sells, construction, transportation or short education 23 16% to repeat the exercises at home, we did not explore whether
Student/retired/other 25 17%
Level of job demands (Karasek et al., 1998)
Low 34 29%
Intermediate 64 56%
High 17 15%
Level of job control (Karasek et al., 1998)
Low 85 73%
Intermediate 20 17%
High 11 10%
Duration of neck pain
1–3 months 20 14%
4–6 months 15 10%
7–12 months 18 12%
12 þ months 92 64%
Characteristics of pain onset
Sudden 32 22%
Gradual 89 61%
Unsure 24 17%
Pain intensity at baseline, mean (SD) 5.9(1.3)
Disability at baseline, mean (SD) 4.4(2.0)
Previous episodes of NP Ever 67 46%
personal trainer
No 101 70%
At least once Ever 44 30%
massage
No 16 11%
At least once 129 89%
Depressive symptoms (Zigmond and Snaith, 1983) 22 15%
Ever diagnosed with depression Use of 41 28%
medication
None 40 27%
Paracetamol 24 16%
NSAID 29 20%
Paracetamol þ NSAID 27 19%
Muscle relaxants, migraine medications 14 10%
Opioids, SSRI, anticonvulsants 11 8%
Sleep disorder 12 8%
Daily smoking 8 6%
Body mass index, mean (SD) 24(3)
Low or normal (<25) 83 57%
Overweight (25–29.9) 55 38%
Obese (�30) Self-perceived 7 5%
health
Excellent 13 9%
Very good 39 27%
Good 58 40%
Satisfactory or poor 35 24%
Expectations on recovery, median (IQR) 3(5)
0–3 78 54%
4–6 39 27%
7 - 10 28 19%

IQR¼ Inter quartile range.

a week for 12 weeks should be considered as a treatment option but that

supervised strengthening exercises alone twice a week for six weeks should
not (Cot^ �e et al., 2016). A trial found no differences in pain intensity or
disability between a comprehensive exercises program during 12 weeks
compared to advice in patients with chronic neck pain due to whiplash.
However, like in our study, they found better self-perceived recovery among Fig. 2. A) Mean pain intensity at baseline and follow-ups. B) Mean pain related disability at baseline an
those in the comprehensive exercises program up to 12-months follow-up changes in pain and pain-related disability over follow-up.
(Michaleff et al., 2014). In contrast, a trial (also in participants with whiplash)
comparing the effect of neck specific exercises for 12 weeks showed better compliance influenced the outcomes.
results in pain intensity and disability than a prescription of physical activity

5
E. Skillgate et al. Musculoskeletal Science and Practice 45 (2020) 102070

A previous trial of deep tissue massage (average of eight sessions) SECONDARY OUTCOMES
Perceived recovery
suggested that its effect is short-lived compared to a booklet (Sherman et al.,
7 weeks 3.29 2.00–5.42 1.66 0.95–2.89 3.01 1.82–4.96
2009). In our study, the effects on pain intensity were seen in the short term 12 weeks 2.27 1.44–3.57 1.61 1.00–2.61 2.01 1.27–3.18
only but the effect of a deep tissue massage on perceived recovery lasted one 26 weeks 1.98 1.27–3.09 1.74 1.10–2.74 2.12 1.38–3.27
year. A meta-analysis showed only immediate beneficial effects of massage 52 weeks 1.74 1.13–2.67 1.33 0.84–2.09 1.99 1.32–3.00
on pain intensity in comparison to inactive controls and Chinese traditional Sick absence1
medicine but no beneficial effect on disability or pain intensity at follow-ups 7 weeks - - - - - -
12 weeks 0.85 0.46–1.58 1.04 0.59–1.84 1.14 0.65–1.97
(Cheng and Huang, 2014). However, it is uncertain from most of the included
26 weeks 1.37 0.64–2.95 1.71 0.83–3.51 1.98 0.99–3.96
studies whether participants had acute or chronic pain. In addition, the most
52 weeks 1.08 0.65–1.81 1.07 0.64–1.78 1.03 0.62–1.70
common massage technique used was Chinese massage, rather than deep
4.2. Strengths and limitations
tissue massage.
Our results are in line with a previous trial that showed that adding Our study has several strengths. First, our trial is, to our knowledge, the
connective tissue massage to scapula-thoracic stabilization exercises is largest ever conducted on the effectiveness of massage for the treatment for
effective for reducing night neck pain intensity after a four-week subacute and persistent neck pain. Our sample size allowed effective
Table 2 randomization. Second, our study measured outcomes at multiple time
Proportion of participants with the outcomes minimal clinically important change (MCII) in points
pain andallowing
disability,for
number needed to
a thorough treat (NNT),ofmean
description the absolute change
treatment in pain and
effectiveness
disability, proportion of perceived recovery and proportion of persons with sick absence.
over time. Third, the risk of outcome misclassification is low because our
measurement methods were valid (Von Korff et al., 1992). Fourth, we
Massage Exercise
achieved a high follow-up rate and confirmed that there were no systematic
MCII Pain % (95% CI) NNT % (95% CI) differences in baseline characteristics between those who dropped out and
7 weeks 53 (45–62) 8 45 (37–53)
those who completed the trial. Finally, we adopted a pragmatic approach to
12 weeks 52 (44–61) 34 48 (40–57)
the design and delivery of care to provide interventions that are similar to
26 weeks 58 (49–66) 10 62 (53–70)
52 weeks 57 (48–65) 33 61 (53–70)
regular clinical practice.
MCII Disability Our trial has limitations. First, blinding was not possible due to the
7 weeks 81 (74–87) 15 74 (66–81) characteristics of the interventions, and therefore expectations of recovery
12 weeks 77 (73–87) 25 74 (67–81) may differ between the treatment arms and probably explain to some extent
26 weeks 76 (68–83) 100 77 (69–84) the differences in results between the interventions. Second, the advice
52 weeks 78 (70–85) 100 82 (75–88)
group received a maximum of three sessions of therapy compared to six in
Change in Pain mean (95% CI) mean (95% CI)
the other groups. The number of advice sessions (three) were considered
7 weeks 2.1 ( 2.3 to 1.8) 1.8 ( 2.1 to 1.5)
adequate to achieve the purpose of the intervention. It is possible that the
12 weeks 2.4 ( 2.7 to 2.1) 2.4 ( 2.7 to 2.1)
less frequent patient-therapist interaction in the advice group impacted on
26 weeks 2.3 ( 2.6 to 2.0) 2.6 ( 2.9 to 2.3)
the results. Third, despite using balanced block randomization to minimize
52 weeks 2.3 ( 2.7 to 2.0) 2.7 ( 3.0 to 2.3)
differences in the number of individuals between the groups, the number of
Change in Disability
7 weeks 2.5 ( 2.9 to 2.2) 2.1 ( 2.4 to 1.8) participants differed across the groups due to the two-stage procedure used
12 weeks 2.7 ( 3.1 to 2.4) 2.5 ( 2.8 to 2.2)
to include participants. Even though the power calculation indicated that 600
26 weeks 2.7 ( 3.0 to 2.3) 2.6 ( 2.9 to 2.3)
participants was sufficient for the primary research question, we had to
account for potential loss of participants between assessment by the study
52 weeks 2.6 ( 3.0 to 2.2) 2.7 ( 3.0 to 2.4)
coordinator (baseline questionnaire A) and, the second assessment by the
Perceived Recovery % (95% CI) % (95% CI)
therapist in the first visit to the research clinic (baseline questionnaire B) and
7 weeks 41 (33–50) 4 22 (15–29)
inclusion in the trial. We therefore prepared an allocation sequence of 800.
12 weeks 37 (29–46) 5 27 (20–35)
Since 621 participants were enrolled, we only used 621 envelopes. The
26 weeks 35 (27–44) 6 31 (24–40)
treatment groups were unequal by chance. However, this is probably not a
52 weeks 35 (27–43) 7 27 (20–35)
threat to the internal validity of the trial because our analyses showed no
Sick Absence (at least one day)
Baseline 29 (21–38) 31 (23–39) differences between the groups regarding the baseline characteristics.
7 weeks
Further, there is a risk for selection bias due to attrition since those who
12 weeks 13 (07–20) 50 15 (10–23)
dropped out were more often those in the advice group. However, we did
not observe differences in terms of baseline characteristics between them
26 weeks 13 (07–20) 25 14 (09–22)
and those who completed the trial in that group. We do not have information
52 weeks 21 (14–29) 34 20 (13–28)
on the reasons for dropping out, but if those dropping out did not recover,
NNT: Number needed to treat compared to advice to stay active.
we might have underestimated the effect of massage and/or exercise. On the
other hand, if the reason for dropping out was full recovery, we might have
Table 3 instead, overestimated the effects of the mentioned therapies.
Relative risk (RR) and 95% Confidence Intervals (95%CI) of having the primary and It might have been difficult for participants to recall with precision the
secondary outcomes over one year follow-up in the index groups compared to the number of days off work due to neck pain or how they perceived their neck
advice group.
pain (the secondary outcomes sickness absence and self- perceived
Massage Exercise Combined therapy RR 95% CI RR 95% CI RR 95% CI
PRIMARY OUTCOMES recovery). However, we have no reasons to believe that this occurred
Minimal clinical important improvement of pain differentially between the treatment arms.
7 weeks 1.36 1.04–1.77 1.14 0.86–1.51 1.39 1.08–1.81 All 30 therapists provided all the four treatment arms. It is possible that
12 weeks 1.09 0.85–1.39 1.00 0.78–1.29 1.28 1.02–1.60 despite our efforts for standardizing the procedures, some therapists might
26 weeks 1.23 0.97–1.56 1.31 1.04–1.65 1.15 0.90–1.46
52 weeks 1.03 0.83–1.29 1.11 0.90–1.37 1.10 0.89–1.35 1 Information on sick absence was not collected in the 7 weeks questionnaire.
Minimal clinical important improvement of pain-related disability
7 weeks 1.11 0.97–1.27 1.00 0.86–1.15 1.08 0.95–1.23 program, as well as for mental health and pressure pain threshold (Celenay et
12 weeks 0.96 0.85–1.10 0.94 0.83–1.07 1.04 0.93–1.17
al., 2016). Thus, supporting the recommendation that massage plus
26 weeks 1.02 0.89–1.18 1.03 0.90–1.18 1.08 0.94–1.23
supervised exercises are beneficial and could be considered among persons
52 weeks 0.98 0.86–1.11 1.03 0.92–1.17 1.06 0.94–1.19
with grade I-II NP (Guzman et al., 2008).
6
E. Skillgate et al. Musculoskeletal Science and Practice 45 (2020) 102070

have personal preferences for one treatment modality and might have put
more effort in providing that specific one instead of the others. We did
efforts to minimize that risk by preparing a standardized training in the
delivery of all four interventions and, by having regular meetings and
emphasizing the importance of following the procedures. Despite the fact
there was variation in the degree of expertise in the provision of the
interventions between the therapists (some of them were students), we think
that such variation was distributed evenly between the intervention groups
and has little or no effect on the results.
We adapted the Chronic Pain Grade Questionnaire and used the means of
pain intensity and pain-related disability, rather than the usual staging
system. Although this modified version has not been validated, a panel of
experts agreed on the appropriateness of using this modified instrument for
the assessment of neck pain the past four weeks, instead of the past six
months, as well as the use of a mean value for pain intensity and a mean
value for pain-related disability. This modified approach has been used
repeatedly in previous publications (Von Korff et al., 1992; Smith et al., 1997;
Paanalahti et al., 2016; Skillgate et al., 2007).
Although we aimed to minimize contamination by asking participants to
avoid other treatments in the first three months post- randomization,
participants in the advice group used additional health services to a larger
extent during the first three months (they spent on average 290 Euros,
compared to 207–220 Euros in the other groups, p- value <0.001) which we
believe may have underestimated the difference in early effect between
advice and the other interventions.
This is a multi-arm controlled trial, designed to improve the efficiency of
the evaluation of the interventions. However, there is an increased
probability of type I error (false positive results) due to multiple testing.
Although a statistical correction would have shown more conservative
results, we did not perform such correction, considering that the different
arms corresponded to different treatment modalities, rather than different
doses of the same therapy (Wason et al., 2014).
As stated in the study protocol, we aimed to study whether the therapies
in question were at least as effective as advice to stay active,
Appendix B. Supplementary data
rather than comparing all therapies against each other. Nevertheless, the
design gives a unique opportunity to study if combining treatments provide
more benefits than the interventions on their own. Even though the
statistical power was too low for such comparisons, the results indicate that
there are no clear differences in effect between combined and single
treatments in this trial.
Our findings suggest that a short course of deep tissue massage alone or
in combination with strengthening and stretching exercises could be a
treatment option for persons with subacute and persistent neck pain. Adding
strengthening and stretching exercises to massage may help in maintaining
the effect over time. Clinicians and patients should not expect a higher
decrease in levels of pain intensity or disability by using these therapies
instead of advice to stay active, but rather an improvement in other
dimensions of the condition, better measured by perceived recovery.
5. Conclusions
In this study, the results at 12-monts follow-up were that none of the
index therapies; deep tissue massage, supervised strengthening and
stretching exercises, or a combined therapy; were more effective than advice
to stay active in terms of clinically important change in pain intensity or pain-
related disability. However, the index therapies led to higher incidence of
clinically important improvement in pain intensity in the short term and to
higher incidence of favorable perceived recovery in the short and in the long
term than advice to stay active. Clinicians may consider recommending these
modalities as treatments for persons with subacute or persistent neck pain.
Ethical approval
The trial was approved by the Regional Ethic Committee in Stockholm
(Dnr: 2014/755-31/3).
Funding/support
Swedish Research Council (VR) 52120133739, Swedish Research Council
for Health, Working Life and Welfare (FORTE) 20141483 and The Swedish
Naprapathic Association funded this study.
Additional contributions
Authors extend special thanks to Anna Peterson, for the coordination of
the data collection, data cleaning and valuable input along the conduction of
the study and interpretation of the data; and to Fredrik Johansson and Martin
Asker for their orientation during the design and implementation of the
interventions.
Data sharing
The STONE data is confidential and is not publicly available.
Declaration of competing interest Nothing to
declare.

Supplementary data to this article can be found online at https://doi.org/10.1016/j.msksp.2019.102070.

Appendix A. Description of the interventions in the STONE trial

1. Massage

Effleurage technique of the whole back and neck was followed by petrissage, kneading and edging/scissoring. Additionally, dynamic stretching as a
component of fascial release technique could be offered as part of the treatment. Additionally, thorax and/or jaw musculature was treated if indicated.
A maximum of six sessions of therapies during six weeks were recommended: twice a week the first week and less often thereafter. The visits lasted 45
min and at least 35 min were dedicated to active treatment at every session. After general treatment of neck and back, the therapist focused on the most
affected/sored ones. The pressure during the massage was adjusted according to the patient’s status/willingness. The massage should be experienced properly
and beneficial without reaching more than 5/10 in a visual analogue scale in pain and the participant got the information that they could ask for adjustments in
the intensity of the massage at any given time. Good communication with the patient was encouraged.
Pressure was applied with a focus on the area that produced concordant signs. Pressure on such areas was repeated with three increments of pressure
applied at every decrease of the pain; if there was no decrease in pain, the pressure was sustained for 30 s. Fascial techniques with and without active
movement participation were combined with the techniques described above.
2. Exercise Training

A maximum of six sessions of therapies during six weeks were recommended: twice a week the first week and less often thereafter. The visit lasted 45
min and at least 35 min dedicated to active treatment at every session.

7
E. Skillgate et al. Musculoskeletal Science and Practice 45 (2020) 102070

The program focused on activation of muscles of the neck area. The patient worked out all muscles/exercises, but the exercises were adjusted depending on
the patient status/tolerance and ability to perform, one out of three levels of intensity was chosen based on two aspects. This evaluation (points 1 and 2) was
also the base to decide whether the participant could progress to the next intensity level in the exercises.

1. The patient performance:

The participant should perform the exercise correctly with minimal co-activation of other muscles/movements.
Strain: the aim was that the participant performed the specified exercises in 3 �10 repetitions if no other instruction was given.

2. Pain experienced by the patient:

The exercises should not produce pain over 5/10 in a visual analogue scale and the neck pain should not increase the next day after training with more than
2 points in the same scale.
The participant was instructed to perform the exercises at home one to two times per week doing 3 �10 with good technique. For this purpose, the
participant was filmed with their smartphone, as support for the exercises and the therapist indicated what was important to consider during the execution of
the exercises with verbal instructions. Specific description of the exercises:

1 Activation of deep neck flexors (“The owl”)

� Purpose: to activate and strengthen deep cervical flexors (M. longus colli and M. rectus capitis anterior) for increased cervical strength and/or neck
function.
� Considerations: minimize pressure of extensors, avoid compensation of global musculature and observe that breathing is maintained normally.
�Level of intensity 1–3 �10 repetitions.
In supine position, slowly retract the shin against neck and go back to the start position. Repeat.
�Level of intensity 2–3 �30 s.
In sitting or standing position, slowly retract the shin against neck, hold with light pressure against the forehead and go back to the start position.
�Level of intensity 3–3 �Max seconds.
In supine position, slowly retract the shin against the neck, lift the head 1 cm above the bench maintaining the shin retracted. 2 Training of
chest musculature (Pushups plus).
� Purpose: to strengthen chest musculature and muscles around the scapula.
� Considerations: Minimize cervical and lumbar hyperlordosis and avoid elevation of scapula. Do a pushup with straight body, push the arms forwards
once the up position is reached so that the scapula separates. Repeat. �Level of intensity 1–3 �10 repetitions against the wall or a bench �Level of
intensity 2–3 �10 repetitions against the floor on the knees.
�Level of intensity 3–3 �10 repetitions against the floor on the toes. 3
Training of scapula musculature (Lying pulldown)
� Purpose: to strengthen muscles around shoulders (M. serratus anterior, M. Trapezius pars ascendens) with simultaneous static control of the cervical
musculature.
� Considerations: minimize cervical and lumbar hyperlordosis and avoid elevation of shoulders.
� In supine position retract the shin against the neck, lift the head 1 cm above the bench while maintaining the shin in the same position and hold. Drag
the arms along the body (resembling a change from a “Y”-position to a “W”-position) and finish with contraction between the scapula. Repeat.
�Level of intensity 1–3 �10 repetitions without rubber band.
�Level of intensity 2–3 �10 repetitions with rubber band 1.
�Level of intensity 3–3 �10 repetitions with rubber band 2.
4 Training of deep extensors of the neck
� Purpose: to strengthen deep extensors of the neck (Mm. Erector spinae).
� Considerations: high extension of the neck and compensation of global musculature.
�Level of intensity 1–3 �1 min.
In prone position, drag the shin against the neck, lift the head 1 cm above the bench while maintaining the same position and hold.
�Level of intensity 2–3 �1 min.
In prone position, hold the arms along the body, rotate the arms in and out with contraction of the area between scapula.
�Level of intensity 3–3 �1 min.
In prone position, abduct and adduct the shoulder joint.
5 Training of the scapula musculature (scapulothoracic control exercise)
� Purpose: to strengthen muscles around shoulders (M. Trapezius pars descendens). Alternative to patients with high levels of pain. �
Considerations: avoid elevation of scapula.
�Level of intensity 1–3 �10 repetitions
Standing/Sitting hold the hands behind the back, drag the shoulders back and downwards while contracting the area between scapulae. Repeat.
�Level of intensity 2–3 �1 min.
Lying on one side, flex and extend the free arm while maintaining the position of the scapulae.
�Level of intensity 3–3 �1 min.
In prone position, let the head to rest, resemble a diamond shape with the arms and lift the hands by contracting the area between the scapulae.
6 Stretching of chest muscles
� Purpose: to decrease the tonus of the chest musculature (M. Pectoralis major).
� 15–20 s, 3 times each side. With flexed elbows against a wall, stretch out the chest musculature by rotating the body away from the arms.

8
E. Skillgate et al. Musculoskeletal Science and Practice 45 (2020) 102070

7Stretching – depressors of the shoulder.

� Purpose: decrease the tonus in the depressors of the shoulder (M. Pectoralis minor).
� 15–20 s, 3 times each side. In standing or supine position, elevate the shoulder and arm, rotate the body towards the opposite side and hold with
flexed knees (if supine position).
8 Stretching – jaw musculature
� Purpose: decrease the tonus in the jaw musculature (M. masseter, M. temporalis, Mm. pterygoidei).
� 15–20 s �3 times. Open the jaw wide. Thereafter, strain the mouth by pressing the fingers against the upper and lower teeth.
9 Stretching – jaw musculature (Interoceptive neuromuscular facilitation)
� Purpose: decrease the tonus in the jaw musculature (M. masseter, M. temporalis, Mm. pterygoidei).
� 3 �10 repetitions. Place a fist under the shin. Open the jaw slowly with a light resistance with the fist. Hold for 6 s. Repeat.
10 Stretching – jaw musculature (Proprioceptive neuromuscular facilitation)
� Purpose: decrease the tonus in the jaw musculature (M. masseter, M. temporalis, Mm. pterygoidei).
� 3 �10 repetitions. Open the jaw, and try to close it while resisting the movement by dragging the lower portion of the jaw with the fingers placed on
the lower teeth row.
3. Massage And Training

A maximum of six sessions of therapies during six weeks were recommended. The visit lasted 60 min and at least 25 min were dedicated to active
treatment with strengthening and stretching exercises, followed by at least 25 min of deep tissue massage, both as described above.
9
E. Skillgate et al. Musculoskeletal Science and Practice 45 (2020) 102070

4. Advice

A maximum of three visits were offered. Evidence-based advice was given based on scientific statements from SBU (Statens beredning for € medicinsk
och social utvardering: € Swedish agency for health technology assessment and assessment of social services) (SBU, 2000) and Cochrane (Karjalainen et
al., 2003; Gross et al., 2015a, 2015b; Patel et al., 2013) consisting of the following elements:

� Adequate information on the condition and reassurance to the participant that the condition is not dangerous but a tolerable strain and that the most
important according to previous experience and research is to try to self-control their own pain by being active both socially and physically.
� Advice to the participant to be active and continue daily activities including work, if possible.
� Description of over the counter medications that could be used, if necessary, to relieve pain, mentioning that it is common to take regularly, in a first stage,
paracetamol, and then NSAIDs (observing that there are contraindications and risk factors).
� Revision of which movements can be relevant according to standard recommendations (Can use the online resource Exor-Live ( ExorLive, 2019) for
maximum three exercises) observing that this should not be as detailed and adjusted as the interventions in the exercise group.

Participants were classified in three different groups:

1. Those who did not have physical activity as a habit (whom were instructed on minimal exercises mainly oriented towards good circulation).
2. Those who had physical activity as a habit (adjustments were suggested to incorporate exercises to the training habit).
3. Those who were highly active (adjustments of the exercises to the training habit were suggested with focus on neck musculature).

Finally, a booklet was given with the different approaches to manage back and neck pain and information facts about exercises.
References Jensen, I., Linton, S.J., Overmeer, T., Bergstrom,€ G., Ma€at€ tanen,€ H., 2004. In: Rygg Boken
Beaton, D.E., Tarasuk, V., Katz, J.N., Wright, J.G., 2001. “ are you better ?” A qualitative study of
the meaning of recovery. Arthritis Care Res (Hoboken) 45, 270–279.
Carroll, L.J., Hogg-Johnson, S., van der Velde, G., et al., 2008a. Course and prognostic factors for
neck pain in the general population: results of the bone and joint decade 2000-2010 task
force on neck pain and its associated disorders. Spine (Phila Pa 1976) 33 (4 Suppl. l), S75–
S82. https://doi.org/10.1097/BRS.0b013e31816445be.
Carroll, L.J., Hogg-Johnson, S., Cote, P., et al., 2008b. Course and prognostic factors for neck
pain in workers results of the bone and joint decade 2000 –2010 task force on neck pain
and its associated disorders. Spine (Phila Pa 1976) 33 (45), S93–S100.
https://doi.org/10.1007/s00586-008-0629-6.
Carroll, L.J., Holm, L., Hogg-Johnson, S., et al., 2008c. Course and prognostic factors for neck
pain in whiplash-associated disorders (WAD) results: results of the bone and joint decade
2000-2010 task force on neck pain and its associated disorders. Spine (Phila Pa 1976) 33
(4), S83–S92. S0161-4754(08)00340-0 [pii]\ 10.1016/j. jmpt.2008.11.013.
Celenay, S.T., Kaya, D.O., Akbayrak, T., 2016. Cervical and scapulothoracic stabilization
exercises with and without connective tissue massage for chronic mechanical neck pain: a
prospective, randomised controlled trial. Man. Ther. 21, 144–150. https://
doi.org/10.1016/j.math.2015.07.003 LK.
Cheng, Y.H., Huang, G.C., 2014. Efficacy of massage therapy on pain and dysfunction in patients
with neck pain: a systematic review and meta-analysis. Evid. Based Complement. Alternat.
Med. 2014, 204360. https://doi.org/10.1155/2014/204360. December 2012.
Cot^ �e, P., Wong, J.J., Sutton, D., et al., 2016. Management of neck pain and associated
disorders: a clinical practice guideline from the Ontario Protocol for Traffic Injury
Management (OPTIMa) Collaboration. Eur. Spine J. 25 (7), 2000–2022. https://doi.
org/10.1007/s00586-016-4467-7.
Derksen, F., Bensing, J., Lagro-janssen, A., 2013. Effectiveness of empathy in general practice: a
systematic review. Br. J. Gen. Pract. (January), 76–84. https://doi.org/
10.3399/bjgp13X660814.
Dworkin, R.H., Turk, D c, Farrar, J.T., et al., 2005. Core outcome measures for chronic pain
clinical trials: IMMPACT recommendations. Pain 113 (1–2), 9–19.
Ernst, E., Pittler, M., Wider, B., Boddy, K., 2006. The Desktop Guide to Complementary and
Alternative Medicine. Elsevier Mosby, Edinburgh.
ExorLive, 2019. ExorLive. https://www.exorlive.com/se. Accessed June 27, 2019.
Fejer, R., Kyvik, K.O., Hartvigsen, J., 2006. The prevalence of neck pain in the world population:
a systematic critical review of the literature. Eur. Spine J. 15 (6), 834–848.
https://doi.org/10.1007/s00586-004-0864-4.
Global burden of disease study 2013 Collaborators. Global, regional, and national incidence,
prevalence, and years lived with disability for 301 acute and chronic diseases and injuries
in 188 countries, 1990-2013: a systematic analysis for the Global Burden of Disease Study
2013. Lancet 386 (9995), 2015, 743–800.
Gross, A., Langevin, P., Burnie, S.J., et al., 2015a. Manipulation and mobilisation for neck pain
contrasted against an inactive control or another active treatment. Cochrane Database
Syst. Rev. 2015 (9), CD004249. https://doi.org/10.1002/ 14651858.CD004249.pub4 LK.
Gross, A., Tm, K., Jp, P., et al., 2015b. Exercises for mechanical neck disorders (Review)
Exercises for mechanical neck disorders. Cochrane Database Syst. Rev. 1 https://doi.
org/10.1002/14651858.CD004250. N.PAG-N.PAG, pub5.Copyright.
Guzman, J., Haldeman, S., Carroll, L.J., et al., 2008. Clinical practice implications of the bone and
joint decade 2000 –2010 task force on neck pain and its associated disorders from
concepts and findings to recommendations. Spine (Phila Pa 1976) 33
(45), S199–S213.
Hansson, E.K., Hansson, T.H., 2005. The costs for persons sick-listed more than one month
because of low back or neck problems. A two-year prospective study of Swedish patients.
Eur. Spine J. 14 (4), 337–345. https://doi.org/10.1007/s00586- 004-0731-3.
Hush, J.M., Refshauge, K., Sullivan, G., et al., 2009. Recovery : what does this mean to patients
with low back Pain ? Arthritis Rheum. 61 (1), 124–131. https://doi.org/ 10.1002/art.24162.
[The book of the back]. Stockholm.
Kamper, S.J., Ostelo, R.W.J.G., Knol, D.L., Maher, C.G., Vet, HCW De, Hancock, M.J., 2010.
Global Perceived Effect scales provided reliable assessments of health transition in people
with musculoskeletal disorders , but ratings are strongly influenced by current status. J.
Clin. Epidemiol. 63 (7), 760–766. https://doi.org/ 10.1016/j.jclinepi.2009.09.009 e1.
Karasek, R., Brisson, C., Kawakami, N., Houtman, I., Bongers, P., Amick, B., 1998. The Job
Content Questionnaire (JCQ): an instrument for internationally comparative assessments
of psychosocial job characteristics. J. Occup. Health Psychol. 3 (4), 322–355.
https://doi.org/10.1037/1076-8998.3.4.322.
Karjalainen, K., Malmivaara, A., van Tulder, M.W., et al., 2003. Multidisciplinary
biopsychosocial rehabilitation for neck and shoulder pain among working age adults
(Review). Cochrane Database Syst. Rev. (2), 2–4. https://doi.org/10.1002/
14651858.CD002194.
Koltyn, K.F., Brellenthin, A.G., Cook, D.B., Sehgal, N., Hillard, C., 2014. Mechanisms of exercise-
induced hypoalgesia. J. Pain 15 (12), 1294–1304. https://doi.org/10.1016/
j.jpain.2014.09.006.
Kovacs, F.M., Abraira, V., Royuela, A., et al., 2008. Minimum detectable and minimal clinically
important changes for pain in patients with nonspecific neck pain. BMC Muscoskelet.
Disord. 9 (Mdc), 43. https://doi.org/10.1186/1471-2474-9-43.
Ludvigsson, M.L., Peterson, G., O’Leary, S., Å, Dedering, Peolsson, A., 2015. The effect of neck-
specific exercise with, or without a behavioral approach, on pain, disability, and self-
efficacy in chronic whiplash-associated disorders: a randomized clinical
trial. Clin. J. Pain 31 (4), 294–303. https://doi.org/10.1097/ AJP.0000000000000123.
Michaleff, Z.A., Maher, C.G., Lin, C.W.C., et al., 2014. Comprehensive physiotherapy exercise
programme or advice for chronic whiplash (PROMISE): a pragmatic randomised controlled
trial. Lancet 384 (9938), 133–141. https://doi.org/10.1016/ S0140-6736(14)60457-8.
Paanalahti, K., Holm, L.W., Nordin, M., et al., 2016. Three combinations of manual therapy
techniques within naprapathy in the treatment of neck and/or back pain: a randomized
controlled trial. BMC Muscoskelet. Disord. 17 (1), 176. https://doi.org/ 10.1186/s12891-
016-1030-y.
Patel, K., Gross, A., Graham, N., et al., 2013. Massage for mechanical neck disorders.
Cochrane Database Syst. Rev. 9 https://doi.org/10.1002/14651858.CD001929. pub3.
Rapaport, M.H., Schettler, P., Bresee, C., 2012. A preliminary study of the effects of repeated
massage on hypothalamic-pituitary-adrenal and immune function in healthy individuals: a
study of mechanisms of action and dosage. J. Altern. Complement. Med. 18 (8), 789–797.
https://doi.org/10.1089/acm.2011.0071.
Runhaar, J., Luijsterburg, P., Dekker, J., Bierma-Zeinstra, S.M.A., 2015. Identifying potential
working mechanisms behind the positive effects of exercise therapy on pain
and function in osteoarthritis; a systematic review. Osteoarthr. Cartil. 23 (7), 1071–1082.
https://doi.org/10.1016/j.joca.2014.12.027.
SBU, Swedish Agency for Health Technology Assessment and Assessment of Social Services,
2000. [nt i ryggen, ont i nacken. En systematisk litteraturoversikt€ [Back and
Neck Pain. A systematic literature overview], 145/1. SBU, Stockholm.
Sherman, K.J., Cherkin, D.C., Hawkes, R.J., Diana, L., Deyo, R.A., 2009. Randomized trial of
therapeutic massage for chronic neck pain. Clin. J. Pain 25 (3), 233–238.
https://doi.org/10.1097/AJP.0b013e31818b7912.
Sizer Jr., P.S., Brismee,� J., Cook, C., 2007. Medical screening for red flags in the diagnosis and
management of musculoskeletal spine pain. Pain practice 7 (1), 53–71.
Skillgate, E., Vingård, E., Alfredsson, L., 2007. Naprapathic manual therapy or evidence- based
care for back and neck pain: a randomized, controlled trial. Clin. J. Pain 23 (5), 431–439.
https://doi.org/10.1097/AJP.0b013e31805593d8.
Skillgate, E., Bohman, T., Holm, L.W., Vingård, E., Alfredsson, L., 2010. The long-term effects of
naprapathic manual therapy on back and neck pain - results from a pragmatic randomized
controlled trial. BMC Muscoskelet. Disord. 11, 26. https:// doi.org/10.1186/1471-2474-11-
26.
Skillgate, E., Bill, A.-S., Cot^ �e, P., Viklund, P., Peterson, A., Holm, L.W., 2015. The effect of
massage therapy and/or exercise therapy on subacute or long-lasting neck pain–the

10
E. Skillgate et al. Musculoskeletal Science and Practice 45 (2020) 102070

Stockholm neck trial (STONE): study protocol for a randomized controlled trial. Trials 16,
414. https://doi.org/10.1186/s13063-015-0926-4.
Smith, B., Penny, K., Purves, A., et al., 1997. The Chronic Pain Grade questionnaire: validation
and reliability in postal research. Pain 71 (2), 141–147.
Southerst, D., Nordin, M.C., Cot^ �e, P., et al., 2014. Is exercise effective for the management
of neck pain and associated disorders or whiplash-associated disorders?
A systematic review by the Ontario Protocol for Traffic Injury Management
(OPTIMa) Collaboration. Spine J. https://doi.org/10.1016/j.spinee.2014.02.014.
StataCorp, 2015. Stata Statistical Software: Release, 14.
Statistics Sweden, 2012. SSYK Standard For€ Svensk Yrkesklassificering [Swedish Standard
Classification of Occupations]. Statistics Sweden. MIS 2012:1.
Twisk, J.W.R., 2013. Applied Longitudinal Data Analysis for Epidemiologists, second ed.
Cambridge, Cambridge.
van der Windt, D.A.W.M., Koes, B.W., Boeke, A.J.P., Jong, BA De, Bouter, L.M., 1998.
Effectiveness of corticosteroid injections versus physiotherapy for treatment of painful
stiff shoulder in primary care: randomised trial. BMJ 317 (November),
1292–1296.
Von Korff, M., Ormel, J., Keefe, F., Dworkin, S., 1992. Grading the severity of chronic pain. Pain
50 (September), 133–149. https://doi.org/10.1016/0304-3959(92) 90154-4.
Wason, J.M.S., Stecher, L., Mander, A.P., 2014. Correcting for multiple-testing in multi- arm
trials: is it necessary and is it done? Trials 15, 1–7.
Weerapong, P., Hume, P.A., Kolt, G.S., 2005. The mechanisms of massage and effect on
performance, muscle recovery and injury prevention. Sport. Med. 35 (3), 235–256.
Wong, J.J., Shearer, H.M., Mior, S., et al., 2016. Are Manual Therapies, Passive Physical
Modalities, or Acupuncture Effective for the Management of Patients with Whiplash-
Associated Disorders or Neck Pain and Associated Disorders? An Update of the Bone and
Joint Decade Task Force on Neck Pain and its Ass, vol. 16. Elsevier Inc. https://
doi.org/10.1016/j.spinee.2015.08.024
Ylinen, J., Takala, E.-P., Nykanen, € M., et al., 2003. Active neck muscle training in the
treatment of chronic neck pain in women. Jama 289 (19), 2509. https://doi.org/
10.1001/jama.289.19.2509.
Zigmond, a S., Snaith, R.P., 1983. The hospital anxiety and depression scale. Acta Psychiatr.
Scand. 67 (6), 361–370. https://doi.org/10.1111/j.1600-0447.1983. tb09716.x.

11