Accepted Manuscript

Title: Work-related psychosocial risk factors and

musculoskeletal disorders in hospital nurses and nursing
aides: A systematic review and meta-analysis

Author: Dinora Bernal Javier Campos-Serna Aurelio Tobias

Sergio Vargas-Prada Fernando G. Benavides Consol Serra

PII: S0020-7489(14)00297-1
DOI: [Link]
Reference: NS 2471

To appear in:

Received date: 25-1-2014

Revised date: 31-10-2014
Accepted date: 4-11-2014

Please cite this article as: Bernal, D., Campos-Serna, J., Tobias, A., Vargas-
Prada, S., Benavides, F.G., Serra, C.,Work-related psychosocial risk factors and
musculoskeletal disorders in hospital nurses and nursing aides: A systematic
review and meta-analysis, International Journal of Nursing Studies (2014),
[Link]

This is a PDF file of an unedited manuscript that has been accepted for publication.
As a service to our customers we are providing this early version of the manuscript.
The manuscript will undergo copyediting, typesetting, and review of the resulting proof
before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that
apply to the journal pertain.
Work-related psychosocial risk factors and musculoskeletal disorders in hospital

nurses and nursing aides: A systematic review and meta-analysis

Dinora Bernal1,2, Javier Campos-Serna 2,3,4

, Aurelio Tobias5, Sergio Vargas-Prada2,
Fernando G. Benavides2,4, Consol Serra 2,4,6

t
ip
1. Faculty of nursing care University of Panama. Panama.

cr
2. CiSAL - Center for Research in Occupational Health, Universitat Pompeu Fabra.
Barcelona, Spain

us
3. Area of Public Health and Preventive Medicine, University of Alicante. Alicante,
Spain.

an
4. CIBER Epidemiology and Public Health (CIBERESP). Spain.

5. Institute of Environmental Assessment and Water Research (IDAEA), Spanish

M
Council for Scientific Research (CSIC)

6. Occupational Health Service, Parc Salut MAR. Barcelona, Spain.

d
p te

Corresponding author: Consol Serra, Center for Research in Occupational Health –

ce

Universitat Pompeu Fabra. Carrer Doctor Aiguader 88, 08003-Barcelona, Spain.

[[Link]@[Link]] Phone: 0034-933160875. Fax: 0034-933160410

Ac

Keywords: Musculoskeletal disorders, Nurse, Hospital, Psychosocial factors,

Workplace, Systematic review, Meta-analysis.

Word count text: 3309

1
Page 1 of 45
Work-related psychosocial risk factors and musculoskeletal disorders in hospital

nurses and nursing aides: A systematic review and meta-analysis

t
What is already known about the topic

ip
 Musculoskeletal disorders (MSD) are one of the leading causes of disability in

cr
hospital nurses and nursing aides.
 Traditionally, studies on risk factors for MSD have focused on physical
activities like manual handling, and individual characteristics, such as sex and

us
age.
 Recently, a growing body of evidence suggests that organizational factors might
play an important role in the occurrence of MSD in nurses and aides.

What this paper adds

an
 Despite the small number of longitudinal studies available, our findings provide
M
consistent evidence of an association between exposure to work-related
psychosocial factors and MSD in hospital nurses and aides.
 Interventions to reduce MSD in hospitals should take into account not only
d

ergonomics, but also the improvement of organizational aspects of the work

te

environment.
p
ce

Abstract

Objectives: To estimate the association between psychosocial risk factors in the

Ac

workplace and musculoskeletal disorders (MSD) in nurses and aides.

Design: Systematic review and meta-analysis

Data sources: An electronic search was performed using MEDLINE (Pubmed),

Psychinfo, Web of Science, Tripdatabase, Cochrane Central Controlled Trials,

NIOSHTIC and Joanna Briggs Institute of Systematic Reviews on Nursing and

Midwifery, to identify observational studies assessing the role of psychosocial risk

factors on MSD in hospital nurses and nursing aides.

2
Page 2 of 45
Review methods: Two reviewers independently assessed eligibility and extracted data.

Quality assessment was conducted independently by two reviewers using an adapted

version of the Standardized Quality Scale. Random-effects meta-analysis was

performed by subsets based on specific anatomical site and the exposure to specific

t
psychosocial risk factors. Heterogeneity for each subset of meta-analysis was assessed

ip
and meta-regressions were conducted to examine the source of heterogeneity among

cr
studies.

Results: Twenty four articles were included in the review, seventeen of which were

us
selected for meta-analysis. An association was identified between high psychosocial

an
demands-low job control with prevalent and incident low back pain (OR 1.56; 95%CI

1.22-1.99 and OR 1.52; 95%CI 1.14-2.01, respectively), prevalent shoulder pain (OR
M
1.89; 95%CI 1.53-2.34), prevalent knee pain (OR 2.21; 95%CI 1.07-4.54), and

prevalent pain at any anatomical site (OR 1.38; 95%CI 1.09-1.75). Effort-reward
d

imbalance was associated with prevalent MSD at any anatomical site (OR 6.13; 95%CI
te

5.32-7.07) and low social support with incident back pain (OR 1.82; 95%CI 1.43-2.32).

Heterogeneity was generally low for most subsets of meta-analysis.

p

Conclusion: This meta-analysis suggests that psychosocial risk factors at the workplace
ce

are associated with MSD in hospital nurses and nursing aides. Although most

preventive strategies at the workplace are focused on ergonomic risk factors, improving
Ac

the psychosocial work environment might have an impact on reducing MSDs.

Keywords: Musculoskeletal disorders, Nurse, Hospital, Psychosocial factors,

Workplace, Systematic review, Meta-analysis.

3
Page 3 of 45
INTRODUCTION

Work-related musculoskeletal disorders (MSD) are defined as symptoms caused or

aggravated by occupational risk factors, including discomfort, damage or persistent pain

t
in body structures, such as muscles, joints, tendons, ligaments, nerves, bones, and the

ip
circulatory system (Barboza et., 2008; Cherry et., 2001; Kee et al., 2007; Trinkoff et al.,

cr
2002). MSD are the most common health problem associated with work in Europe,

affecting millions of workers. It has been estimated that 25% of European workers

us
complain of back pain and 23% of muscle aches. MSD are the main cause of sickness

an
absence in western European countries (Murray et al.), and in the United States and

Canada (Punnett et al., 2004). In Europe, costs due to MSD represent approximately 2
M
per cent of their Gross Domestic Product (GDP) (Bevan et al., 2009), without

considering productivity losses and social costs (Choobineh et al., 2010; Menzel, 2007;
d

Podniece et al., 2008). Furthermore, MSD is also one of the main causes of sickness
te

absence among hospital nurses and nursing aides, although underreporting is common

(Menzel, 2008).
p

Factors associated with MSD include individual characteristics, such as age and sex,
ce

occupational risk factors and non-work related exposures. Physical risk factors that arise

from a worker’s tasks (e.g. physical demands, handling loads, repetitive movements or
Ac

vibration) are well established workplace risk factors for the occurrence of MSD.

However, there is some evidence that occupational psychosocial risk factors, such as

high psychosocial demands, low job control or low social support, could also have a

role (European Agency for Safety and Health at Work, 2007; Magnago et al., 2007).

Hospital nurses and nursing aides are occupational groups especially at risk of

developing MSD (Magnago et al., 2007; Solidaki et al., 2010). The prevalence of MSD

5
Page 4 of 45
in nursing professionals has been documented in different studies (Choobineh et al.,

2010; Smith et al., 2003) and varies across countries (Coggon et al., 2013). Karahan et

al found that hospital nurses and nursing aides had the highest prevalence of MSD

(77.1%) in a sample of Turkish health care workers (Karahan et al., 2009). In Norway

t
the prevalence of MSD in nursing aides has been found to be as high as 89% (Willy,

ip
2003), whereas in Japan it is much lower at around 37% (Matsudaira et al., 2011).

cr
Several studies have shown a high risk of developing neck and low back pain in hospital

nurses, attributed to both physical and psychosocial factors at work, such as shift work,

us
long hours at work (Magnago et al., 2007; Menzel, 2007; Trinkoff et al., 2002) and the

an
stress related to patient’s management (Solidaki et al., 2010).

Although some previous systematic reviews have reported an association between

M
psychosocial risks factors in the workplace and MSD in hospital nurses and nursing

aides, to our knowledge no meta-analysis has yet been published. Thus, the aim of our
d

study was to evaluate and quantify the association between exposure to psychosocial
te

factors in the workplace and MSD in nurses and nursing aides in hospital settings.
p
ce
Ac

6
Page 5 of 45
METHODS

Search strategy

An electronic search was carried out using MEDLINE (Pubmed), Psychinfo, Web of

t
Science, Tripdatabase, Cochrane Central Controlled Trials, NIOSHTIC and Joanna

ip
Briggs Institute of Systematic Reviews on Nursing and Midwifery. Our search strategy

cr
was applied similarly to all databases and combined four blocks of keywords intended

to capture different aspects of our review: 1) the outcome (prevalence and incidence of

us
MSD), 2) the study population (nurses and nursing aides), 3) exposure (psychosocial

an
risk factors, including high psychosocial demands/low job control, low social support

and effort-reward imbalance), and 4) occupational setting (hospital). The search terms
M
used were: for study population and occupational setting “((((("nurses"[MeSH Terms]

NOT "breast feeding"[MeSH Terms] OR nurse[Text Word]) OR ("personnel,

d

hospital"[MeSH Terms] OR hospital staff[Text Word])) OR aides[All Fields]) OR

te

("nursing staff"[MeSH Terms] OR "nurses"[MeSH Terms] OR nursing personnel[Text

Word]))”; for psychological risk factors “(((("psychosocial factors"[All Fields] OR

p

("workplace"[MeSH Terms] OR workplace[Text Word])) OR (job[All Fields] AND

ce

("sprains and strains"[MeSH Terms] OR strain[Text Word]))) OR ("social

support"[MeSH Terms] OR social support[Text Word])) OR (("work"[MeSH Terms]

Ac

OR work[Text Word]) AND ("organisations"[MeSH Terms] OR organisational[Text

Word]) AND factors[All Fields])))”; and for MSD “(((((((musculoskeletal[All Fields]

OR ("upper extremity"[MeSH Terms] OR upper limbs[Text Word])) OR "wrist

injuries"[MeSH Terms]) OR ("elbow"[MeSH Terms] OR "elbow joint"[MeSH Terms]

OR elbow[Text Word])) OR ("shoulder pain"[MeSH Terms] OR shoulder pain[Text

Word])) OR ("neck pain"[MeSH Terms] OR neck pain[Text Word])) OR ("low back

7
Page 6 of 45
pain"[MeSH Terms] OR low back pain[Text Word])) OR ("back pain"[MeSH Terms]

OR back pain[Text Word])). Also, the reference lists of papers which fulfilled our

inclusion criteria were reviewed to identify additional studies not included in our

electronic search.

t
Study selection and eligibility criteria

ip
Observational studies (cohort, case-control or cross-sectional), published in English or

cr
Spanish between January 2001 and March 2014, were included if they assessed the

association between MSDs and psychosocial risk factors at the workplace in hospital

us
nurses and nursing aides. Studies were excluded if: i) they were in a different language

an
than English or Spanish, ii) the study population was nursing students, or iii) a wide

range of hospital workers and occupations were included, but data for hospital nurses or
M
nursing aides were not analysed separately.

After excluding duplicates, a total of 3202 citations were obtained from the electronic
d

search. All citations were reviewed by title, and when was necessary, by abstract.
te

Ninety-one potential publications were identified and for all of them full text were

obtained. Those studies were reviewed by two independent researchers (DB and JC).
p

The degree of agreement (kappa index) between the two reviewers was 80.2 %.
ce

Disagreements (20% of the 91 identified publications) were resolved by a third reviewer

(SVP/CS) who made the final decision. Twenty-six publications (all in English) which
Ac

met the inclusion criteria were included for quality assessment. Figure 1 shows the flow

chart of study selection.

Quality assessment and extraction

The methodological quality of the 26 studies that met the inclusion criteria was assessed

independently by two reviewers (DB and JC). As has been done in previous systematic

reviews for meta-analysis (Bongers et al., 2002; Gershon et al., 2007), we used an

8
Page 7 of 45
adapted version of the Standardized Quality Scale developed by van der Windt et al.

(van der Windt et al., 2000). This scale included 15 items grouped into 5 areas: 1) study

objective, 2) study population, 3) measurement of psychosocial exposure, 4) outcome,

and 5) data analysis and presentation. Each item was rated as "positive" (when

t
requirement was met), "negative" (when requirement was not met) or "unclear" (unsure

ip
if requirement was met). A score was obtained for each study by the sum of all positive

cr
responses (1 point each item). Studies were considered as high-quality when the score

was higher than 80% of the maximum possible score, intermediate quality when the

us
score was between 70% and 79%, and low-quality when it was below 70%. Two studies

an
were of low quality, and were excluded (Dundar et al., 2010; Fonseca and Fernandez.,

2010).
M
The following general and methodological information was obtained from each of the

24 included papers (table 1): authors’ last names, country, year of publication,
d

epidemiological design, study population, sample size, response rate, mean age, and
te

work-related physical demands. Moreover, characteristics of the exposure to

psychosocial risk factors (high psychosocial demands, low job control, low social
p

support from co-workers and supervisors, and effort-reward imbalance) and information
ce

related to the outcome (prevalence of pain at any anatomical site, prevalence and

incidence of back pain, prevalence of neck pain, prevalence of shoulder pain and
Ac

prevalence of knee pain) were collected. Also, information about adjustment variables

and epidemiologic measures of association (prevalence ratio (PR), hazard ratio (HR) or

odds ratio (OR)), and their 95% confidence interval (95% CI) were identified from each

paper.

Meta-analysis

9
Page 8 of 45
From the 24 studies included in the review, we excluded for meta-analysis one study

where the 95% CIs were not provided (Camerino et al., 2001), one cohort (Herin et al.,

2011) and four cross-sectional studies (Carugno et al., 2012; Sorour and El-Maksoud.,

2012; Surawera et al., 2013; Violante et al., 2004) where different psychosocial

t
exposures and/or outcomes were assessed; and one cohort study because hazard ratios

ip
were reported as measures of association (Smedley et al., 2003). Therefore, 17 studies

cr
were considered for meta-analysis, which was carried out using version 11 of Stata

software (StataCorp, 2009).

us
Random effects models were estimated using the method proposed by DerSimonian and

an
Laird (DerSimonian et al., 1986), and the included studies were grouped into nine

subsets according to their epidemiological design, the type of psychosocial exposure

M
(high psychosocial demands/low job control, low social support and effort-reward

imbalance) and the main outcome (prevalent pain at any anatomical site, prevalent and
d

incident back pain, prevalent neck pain, prevalent shoulder pain and prevalent knee
te

pain). Forest plots of meta-analysis were depicted for each of the nine subsets of

studies: (1) exposure to high demands/low job control with prevalence of low back pain;
p

(2) exposure to high demands/low job control with prevalence of neck pain; (3)
ce

exposure to high demands/low job control with prevalence of shoulder pain; (4)

exposure to high demands/low job control with prevalence of knee pain; (5) exposure to
Ac

high demands/low job control with prevalence of pain at any anatomical site; (6)

exposure to low social support with prevalence of pain at any anatomical site; (7)

exposure to effort reward imbalance with prevalence of pain at any anatomical site; (8)

exposure to high demands/low job control with incidence of low back pain; and (9)

exposure to low social support with incidence of back pain. The outcome “pain at any

10
Page 9 of 45
anatomical site” was considered for studies where musculoskeletal pain was reported

without distinguishing a specific anatomical site.

A pooled effect size (OR) and its 95% CI were reported for each subset. The Cochrane

Q test was used to test for heterogeneity and the I2 statistic (the percentage of the total

t
variability between studies due to heterogeneity) to quantify it (Huedo-Medina et al.,

ip
2006). The I2 takes values between 0 and 100%, and a value of 0% indicates absence of

cr
heterogeneity. I2 was interpreted based on Higgins and Thompson classification

(Huedo-Medina et al., 2006); percentages of 25%, 50% and 75% were considered as

us
low, intermediate and high heterogeneity, respectively. A cut-off of p≤0.1 was

an
considered to determine if heterogeneity was statistically significant.

Meta-regression
M
Meta-analysis regression (or meta-regression) is an extension to standard meta-analysis

that investigates the extent to which statistical heterogeneity between studies can be
d

attributed to one or more specific characteristics of the studies. A meta-regression was

te

performed for those subsets where the heterogeneity was statistically significant.

Sample size, response rate and mean age where considered as potential study
p

characteristics that might partially explain most of the observed heterogeneity.

ce

Associations between those variables and outcomes (log OR) were evaluated in

univariate meta-regression models.

Ac

11
Page 10 of 45
RESULTS

Systematic review

Of the 24 intermediate and high quality studies included in the review (table 1), 18 had

a cross-sectional design (Alexopoulos et al., 2003; Alexopoulos et al., 2009; Bos et al.,

t
2007; Carugno et al., 2012; Choobineh et al., 2010; De Souza Magnago et al., 2010;

ip
Golabadi et al., 2013; Hoe et al., 2011; Menhrdad et al., 2010; Sembajwe et al., 2013;

cr
Simon et al., 2007; Smith et al., 2006; Sorour and El-Maksoud., 2012; Stone et al.,

2007; Surawera et al 2013; Violante et all., 2004; Warming et al., 2009; Weyers et al.,

us
2006) and 6 were prospective cohort studies (Alexopoulos et al., 2006; Camerino et al.,

an
2001; Herin et al., 2011; Smedley et al., 2003; Smith et al., 2004; Yip, 2002).

Most studies (n=11) were conducted in European countries, those that were not, came
M
from Iran (3), Australia (2), China (2), Brazil (1), Egypt (1), United States (1), and Brazil

and Italy (1). The number of participants in the included studies ranged from 58 to
d
16,670, and most of them were women (90%), with an overall mean age that ranged
te

between 27 and 44 years-old. Sixteen studies focused on nurses, and 8 studies recruited

both nurses and aides. Different instruments were used to assess MSD and psychosocial
p

risk factors. Most studies (70%) used the Standardized Nordic Questionnaire to measure
ce

MSD. Psychosocial risk factors were measured by the Karasek Job Content (JCQ)

questionnaire, or an adapted version (Camerino et al., 2001; Carugno et al., 2012; Hoe
Ac

et al., 2011), the Effort Reward Imbalance (ERI) questionnaire and the Copenhagen

Psychosocial Questionnaire (COPSOQ). Furthermore, 13 studies were classed as “high

quality” (with a score above 80% of the maximum possible score), and 11 were

considered as “intermediate quality” as their score ranged between 76.9% and 78.6%.

Meta-analysis

12
Page 11 of 45
Pooled risk estimates and heterogeneity values for each subset of studies are

summarised in table 2. Statistically significant associations were found for high

demands/low job control with the prevalence of low back pain (OR 1.56; 95% CI 1.22-

1.99) and the incidence of low back pain (OR 1.52; 95% CI 1.14-2.01). Exposure to

t
high demands/low job control was also associated with the prevalence of shoulder pain

ip
(OR 1.89; 95% CI 1.53-2.34), knee pain (OR 2.21; 95% CI 1.07-4.54) and pain at any

cr
anatomical site (OR 1.38; 95%CI 1.09-1.75), respectively. Likewise, low social support

was significantly associated with the incidence of back pain (OR 1.82; 95% CI 1.43-

us
2.32), and a strong association was observed between effort-reward imbalance and the

an
prevalence of pain at any anatomical site (OR 6.13; 95% CI 5.32-7.07). In two cross-

sectional subsets of meta-analysis heterogeneity was classed as moderate, and in one

M
subset of cross-sectional studies heterogeneity was considered high (table 2). Due to the

maximum number of tables and figures allowed, only forest plots of 3 subsets of cross-
d

sectional studies (figures 2-4) and 1 subset of cohort studies (figure 5) were included in
te

the manuscript.

Meta-regression
p

The three subsets of studies where heterogeneity was high (i. high psychosocial
ce

demands/low job control with the prevalence of low back pain) or moderate (ii. high

psychosocial demands/low job control with the prevalence of neck pain; and iii. low
Ac

social support with the prevalence of pain at any anatomical site) were considered for

meta-regression. Only the variable “sample size” partially explained the high and

moderate heterogeneity found in subsets (i) and (ii) (data not shown).

13
Page 12 of 45
DISCUSSION

According to our findings, work-related psychosocial factors seem to be associated with

MSD in hospital nurses and nursing aides. Specifically, exposure to high demands/low

t
control, effort-reward imbalance and low social support were found to be associated

ip
with low back, neck, shoulder, upper extremity, knee, and/or pain at any anatomical

cr
site, either in nurses, aides or both.

To our knowledge this is the first meta-analysis that explores the association between

us
the exposure to psychosocial factors in the workplace and MSD in nurses and nursing

an
aides who work in hospitals. All included studies used validated instruments to assess

the exposure of psychosocial risk factors at work, based on well-established models

M
among the scientific community, which have been widely used in previous studies. In

addition, to assess MSD, most of the studies used the validated and widely used Nordic
d

questionnaire (Kuorinka et al., 1987). It might be argued that self-reporting of workers'

te

perception could be a source of bias. However, self-rated health is considered a good

indicator of health status (Kaplan et al., 1996; Palmer et al., 2008). The quality
p

assessment of the included studies was based on validated scales previously used in
ce

other published systematic reviews (Bongers et al., 2002;Alexopoulos et al., 2006; Yip,

2002).
Ac

Against these strengths, some limitations need to be addressed. Most of the included

studies in our meta-analysis were cross-sectional and therefore, reverse causality cannot

be ruled out. Another important limitation is that we used random effects models,

assuming that the included studies were representative of the hypothetical population of

studies, and that heterogeneity among the studies may be represented by a single

variance granting too much weight to studies with small sample size. Likewise, it is

14
Page 13 of 45
possible that studies without positive or statistically significant findings may be less

likely to be published by journals. For each subset of meta-analysis we assessed the

possibility of publication bias by using Begg’s test (Palma Pérez et al., 2006). However,

the potential risk of publication bias was quite low (data not shown). Furthermore, our

t
systematic review and meta-analysis included predominantly nurses. Thus, it is possible

ip
that findings, such as the association between job control and musculoskeletal pain,

cr
would have been different in a population largely consisting of nursing aides. Finally,

the Standardized Quality Scale used to evaluate the quality assessment of the identified

us
studies does not include the assessment of bias. This is a systematic review of

an
observational studies about psychological risk factors at work; therefore, the possibility

of selection bias cannot be dismissed.

M
The prevalence of low back pain in hospital nurses and nursing aides is very high, and it

is the leading cause of sickness absence in this occupational group (Maul et al., 2003).
d

Low back pain has been a subject of extensive research and is traditionally attributed to
te

high physical demands (Harcombe et al., 2010; Menzel, 2004). However, the benefits

from interventions (training or mechanical aides) to reduce physical demands and

p

prevent low back pain have been small and of uncertain cost-effectiveness (Verbeek et
ce

al., 2011). A systematic review suggested that other underlying occupational and

individual risk factors may contribute to the occurrence of low back pain in workers
Ac

exposed to heavy manual handling (Punnett and Wegman, 2004). Among other

potential occupational risks, work-related psychosocial factors, such as high job

demand, low job control, low social support and effort-reward imbalance might have a

role in the prevalence and incidence of low back pain in nurses and nursing aides. A

systematic and critical review of cohort studies performed by Hartvigsen et al., did not

find an association between work organisational factors (such as social support) and low

15
Page 14 of 45
back pain (Hartvigsen et al., 2004). These findings are contrary to what we found in our

systematic review. However, the Hartvigsen review included a wide variety of

occupations, and used an evidence synthesis instead of meta-analysis. Likewise, it is

possible that due to the subjective nature of musculoskeletal symptoms, the definition of

t
outcomes varied among studies. Also, the definition of psychosocial factors and the way

ip
these variables were collected varies. Therefore, it is possible that differences in the

cr
“case definition” might explain, at least partially, the lack of comparability and the

contradictory findings between both reviews. Despite the small number of cohort

us
studies included in our meta-analysis, they provided consistent evidence of an

an
association between exposure to work-related psychosocial risk factors and MSD in

hospital nurses and nursing aides. Some mechanisms might explain this. Ando et al
M
proposed that workers exposed to high time pressure to meet demands at work, lack of

social support and other organisational deficiencies, could increase the number of
d

repetitive movements and also be predisposed to unhealthy awkward postures at work

te

(Ando et al., 2000). Moreover, exposure to high time pressure could also increase

workers’ mechanical work-load, which can produce muscular strain. This, in turn, may
p

generate pain or exacerbate pre-existing pain (Ando et al., 2000). Due to the strong
ce

relationship between physical demands at work and MSD, it has been suggested that

physical demands at work must be taken into account when analysing the relationship
Ac

between work-related psychosocial risk factors and MSD (MacDonald et al., 2001). In

fact, most of the studies included in our meta-analysis considered physical demands in

their analysis as a confounding factor. Only three did not incorporate physical demands

as a confounder (Sembajwe et al., 2013; Mehrdad et al., 2010; Stone et al., 2007). We

conducted the meta-analysis with and without these three studies, and findings did not

change significantly. Our meta-analysis results suggest a strong association between

16
Page 15 of 45
work-related psychosocial risk factors and the occurrence of MSD, even after

adjustment for exposure to physical demands.

In conclusion, our results provide consistent evidence of an association between

exposure to work-related psychosocial risk factors and MSD in hospital nurses and

t
aides. Nevertheless, future studies should use longitudinal designs to undertake more

ip
accurate assessments of exposure to work-related psychosocial risk factors that might

cr
have a strong impact on workers' health. Finally, interventions are needed to evaluate

the effectiveness of preventive strategies to reduce the occurrence of MSD in hospital

us
nurses and nursing aides. These interventions should take into account not only

an
ergonomics, but also the improvement of organisational aspects of the work

environment.
M
d
p te
ce
Ac

17
Page 16 of 45
Sources of funding

This study has been funded by the fellowship project Erasmus - Eracol, University of

Panama and the Center for Research in Occupational Health (CiSAL), Universitat

t
Pompeu Fabra (Barcelona, Spain).

ip
cr
us
an
M
d
p te
ce
Ac

18
Page 17 of 45
REFERENCES

Alexopoulos, E., Burdorf, A., Kalokerinou, A., 2003. Risk factors for musculoskeletal

disorders among nursing personnel in Greek hospitals. International Archives of

Occupational and Environmental Health 76 (4), 289-294.

t
Alexopoulos, E., Burdorf, A., Kalokerinou, A., 2006. A comparative analysis on

ip
musculoskeletal disorders between Greek and Dutch nursing personnel.

cr
International Archives of Occupational and Environmental Health 79 (1), 82-88.

Alexopoulos, E., Tanagra, D., Detorakis, I., Gatsi, P., Goroyia, A., Michalopoulou, M.,

us
Jelastopulu, E., 2011. Knee and low back complaints in professional hospital

an
nurses: occurrence, chronicity, care seeking and absenteeism. Work 38 (4), 329-

335.
M
Ando, S., Ono, Y., Shimaoka, M., Hiruta, S., Hattori, Y., Hori, F., Takeuchi, Y., 2000.

Associations of self estimated workloads with musculoskeletal symptoms among

d

hospital nurses. Occupational and Environmental Medicine 57 (3), 211-216.

te

Barboza, MC., Milbrath, VM., Bielemann, VM., De Siqueira, H C., 2008. Work-related

musculoskeletal disorders and their association with occupational nursing. Revista

p

Gaucha de Enfermagem / EENFUFRGS 29 (4), 633-638.

ce

Bevan S, Quadrello T, McGee R, Mahdon M, Vavrovsky A, Barham L. Fit for work?

Musculoskeletal disorders in the European workforce. The Work Foundation, 2009.

Ac

[cited 2014 July 10] Available from:

[Link]

Documents/Fit%20for%20Work%20pan-European%[Link].

Bongers, PM., Kremer, AM., Ter Laak, J., 2002. Are psychosocial factors, risk factors for

symptoms and signs of the shoulder, elbow, or hand/wrist? A review of the

19
Page 18 of 45
 epidemiological literature. American Journal of Industrial Medicine 41 (5), 315-

342.

Bos, E., Krol, B., van der Star, L., Groothoff, J., 2007. Risk factors and musculoskeletal

complaints in non-specialized nurses, IC nurses, operation room nurses, and X-ray

t
technologists. International Archives of Occupational and Environmental Health 80

ip
(3), 198-206.

cr
Camerino, D., Molteni, G., Vito, GD., Evaristi, C., Latocca, R., Cesana, GC., 2001. Job

strain and musculoskeletal disorders of Italian nurses. Occupational Ergonomics 2

us
(4), 215-223.

an
Carugno, M., Pesaroti, AC., Ferrario, MM., Ferrari, AL., Da Silva, FJ., Martins, AC., Felli,

VE., Coggon, D, Bonzini, M.,2012. Physical and psychosocial risk factors for
M
musculoskeletal disorders in Brazilian and Italian nurses. Cad. Saúde Pública. Rio

de Janeiro, 28(9),1632-1642.
d

Cherry, NM., Meyer, JD., Chen, Y., Holt, DL., McDonald, JC., 2001. The reported
te

incidence of work-related musculoskeletal disease in the UK: MOSS 1997-2000.

Occupational Medicine 51 (7), 450-455.

p

Choobineh, A., Movahed, M., Tabatabaie, S H., Kumashiro, M., 2010. Perceived demands
ce

and musculoskeletal disorders in operating room nurses of Shiraz city hospitals.

Industrial Health 48(1), 74-84.

Ac

Coggon, D., Ntani, G., Palmer, KT., et al.,2013. Disabling musculoskeletal pain in

working populations: is it the job, the person or the culture? Pain 154 (6), 856-631.

De Souza Magnago, TS., Lisboa, MT., Griep, RH., Kirchhof, AL., De Azevedo Guido, L.,

2010. Psychosocial aspects of work and musculoskeletal disorders in nursing

workers. Revista Latino-americana de Enfermagem 18 (3), 429-435.

20
Page 19 of 45
DerSimonian, R., Laird, N., 1986. Metaanalysis in clinical trials. Control of Clinical Trials

7 (3), 177-188.

Dundar, PE., Ozmen, D., 2010. Low back pain and related factors in nurses in a university

hospital. Turkish Journal of Public Health 8 (2), 95-104 .

t
European Agency for Safety and Health at Work. 2007. Introduccion a los trastornos

ip
musculoesqueléticos de origen laboral. FACTS 71 (es), 1681-2085.

cr
[Link]

Fonseca, N., Fernandez. R., 2010. Factors related to musculoskeletal disorders in nursing

us
workers. Revista Latino-americana de Enfermagem 18 (6), 1076-1083.

an
Gershon, RR., Stone, PW., Zeltser, M., Faucett, J., MacDavitt, K., Chou, SS., 2007.

Organizational climate and nurse health outcomes in the United States: a systematic
M
review. Industrial Health 45 (5), 622-636.

Golabadi, M., Attarachi, M., Raeisi, S., Namvar, M., 2013. Effects of Psychosocial Strain
d

on Back Symptoms in Tehran General Hospital Nursing Personnel. Arhiv za

te

higijenu rada i toksikologiju, 64(4), 505-511.

Harcombe, H., McBride, D., Derrett, S., Gray, A., 2010. Physical and psychosocial risk
p

factors for musculoskeletal disorders in New Zealand nurses, postal workers and
ce

office workers. Injury Prevention 16 (6), 96-100.

Hartvigsen, J., Lings, S., Leboeuf-Yde, C., Bakketeig, L., 2004. Psychosocial factors at
Ac

work in relation to low back pain and consequences of low back pain; a systematic,

critical review of prospective cohort studies. Occupational and Environmental

Medicine 61 (1), e2.

Herin, F., Paris, C., Levant, A., Vignaud, MC., Sobaszek, A., Soulat, JM., 2011. Links

between nurses' organisational work environment and upper limb musculoskeletal

21
Page 20 of 45
 symptoms: independently of effort-reward imbalance! The ORSOSA study. Pain

152 (9), 2006-2015.

Hoe, V. C., Kelsall, H. L., Urquhart, D. M., Sim, M. R., 2012. Risk factors for

musculoskeletal symptoms of the neck or shoulder alone or neck and shoulder

t
among hospital nurses. Occupational and Environmental Medicine, 69 (3), 198-

ip
204.

cr
Huedo-Medina, T., Sánchez-Meca, J., Marín-Martínez , F., Botella, J., 2006. Assessing

heterogeneity in meta-analysis: Q statistics or I2 index? . Psychological Methods 11

us
(2), 193–206.

an
Kaplan, GA., Goldberg, DE., Cohen, RD., Salonen, R., Tuomilehto, J., Salonen, J.,

Everson, S A., 1996. Perceived health status and morbidity and mortality: evidence
M
from the Kuopio ischaemic heart disease risk factor study. International Journal of

Epidemiology 25 (2), 259-265.

d

Karahan, A., Kav, S., Abbasoglu, A., Dogan, N., 2009. Low back pain: prevalence and
te

associated risk factors among hospital staff. Journal of advanced nursing 65 (3),

516-524.
p

Karasek, R., Baker, D., Marxer, F., Ahlbom, A., Theorell, T., 1981. Job decision latitude,
ce

job demands and cardiovascular disease: a prospective study of Swedish men.

American Journal of Public Health 71 (7), 694-705.

Ac

Kee, D., Rim S., Sun, S., Dohyung, K., 2007. Musculoskeletal disorders among nursing

personnel in Korea. International Journal of Industrial Ergonomics 37 (3), 207-212.

Kristensen, TS., Harald, H., Høgh, A., Borg, V., 2005. The Copenhagen Psychosocial

Quiestionaire a tool for the assessment and improvement of the psychosocial work

environment. Scandinavian Journal of work, Environment and Health 31 (6), 405-

408.

22
Page 21 of 45
Kuorinka, I., Jonsson, B., Kilbom, A., Vinterberg, H., Biering-Sørensen, F., Andersson, J.,

1987. Standardised Nordic questionnaires for the analysis of musculoskeletal

symptoms. Applied Ergonomics 18 (3), 233-237.

MacDonald, LA., Karasek, R., Punnett, L., Scharf, T., 2001. Covariation between

t
workplace physical and psychosocial stressors: evidence and implications for

ip
occupational health research and prevention. Ergonomics 51 (7), 696-718.

cr
Magnago, TS., Lisboa, M., Souza, I., Moreira, M., 2007. Musculoskeletal disorders in

nursing workers: evidences associated to work conditions. Revista Brasileira De

us
Enfermagem 60 (6), 701-705.

an
Matsudaira, K., Palmer, K., Reading, I., Hirai, M., Yoshimura, N., Coggon, D., 2011.

Prevalence and correlates of regional pain and associated disability in Japanese

M
workers. Occupational and Environmental Medicine 68 (3), 191-196.

Maul, I., Laubli, T., Klipstein, A., Krueger, A., 2003. Course of low back pain among
d

nurses: a longitudinal study across eight years. Occupational and Environmental

te

Medicine 60 (7), 497-503.

Mehrdad, R., Dennerlein, JT., Haghighat, M., Aminian, O., 2010. Association between
p

psychosocial factors and musculoskeletal symptoms among Iranian nurses.

ce

American Journal of Industrial Medicine 53 (10), 1032-1039.

Menzel, N., 2004. Back pain prevalence in nursing personnel: measurement issues.
Ac

American Association of Occupational Health Nursing Journal 52 (2), 54-65.

Menzel, N., 2007. Psychosocial factors in musculoskeletal disorders. Critical care nursing

clinics of North America 19 (2), 145-153.

Menzel, N., 2008. Underreporting of musculoskeletal disorders among health care

workers: research needs. American Association of Occupational Health Nursing

Journal 56 (12), 487-494.

23
Page 22 of 45
Murray CJ, Vos T, Lozano R, Naghavi M, Flaxman AD, Michaud C, et al., 2012.

Disability-adjusted life years (DALYs) for 291 diseases and injuries in 21 regions,

1990-2010: a systematic analysis for the Global Burden of Disease Study 2010. The

Lancet. 380 (9859), 2197-2223.

t
Palma, S., Delgado, M., 2006. Consideraciones prácticas acerca de la detección del sesgo

ip
de publicación. Gaceta Sanitaria 20 (2), 10-16.

cr
Palmer, KT., Reading, I., Linaker, C., Calnan, M., Coggon, D., 2008. Population-based

cohort study of incident and persistent arm pain: role of mental health, self-rated

us
health and health beliefs. Pain 136 (1-2), 30-37.

an
Podniece, Z., Taylor, T., 2008. Work-related musculoskeletal disorders: prevention report.

Luxembourg. Office for Official Publications of the European Communities.

M
Punnett, L., Wegman, D., 2004. Work-related musculoskeletal disorders: the

epidemiologic evidence and the debate. Journal of the International Society of

d

Electrophysiological Kinesiology 14 (1), 13-23.

te

Sembajwe, G., Tveito, T. H., Hopcia, K., Kenwood, C., O'Day, E. T., Stoddard, A. M.,

Sorensen, G. 2013. Psychosocial stress and multi-site musculoskeletal pain: a cross-

p

sectional survey of patient care workers. Workplace health & safety, 61(3), 117-
ce

125.

Siegrist, J., Peter, R., Cremer, P., Seidel, D., 1997. Chronic work stress is associated with
Ac

atherogenic lipids and elevated fibrinogen in middle-aged men. Journal of Internal

Medicine 242 (2), 149-156.

Simon, M., Tackenberg, P., Nienhaus, A., Estryn-Behar, M., Conway, PM., Hasselhorn,

HM., 2008. Back or neck-pain-related disability of nursing staff in hospitals,

nursing homes and home care in seven countries--results from the European

NEXT-Study. International Journal of Nursing Studies 45 (1), 24-34.

24
Page 23 of 45
Smedley, J., Inskip, H., Trevelyan, F., Buckle, P., Cooper, C., Coggon, D., 2003. Risk

factors for incident neck and shoulder pain in hospital nurses. Occupational and

Environmental Medicine 60 (11), 864-869.

Smith, DR., Ohmura, K., Yamagata, Z., Minai, J., 2003. Musculoskeletal disorders self-

t
reported by female nursing students in central Japan: a complete cross-sectional

ip
survey. International Journal of Nursing Studies 5 (3), 725-729.

cr
Smith, DR., Wei, N., Zhao, L., Wang, RS., 2004. Musculoskeletal complaints and

psychosocial risk factors among Chinese hospital nurses. Occupational Medicine

us
(London) 54 (8), 579-582.

an
Smith, DR., Mihashi, M., Adachi, Y., Koga, H., Ishitake, T., 2006. A detailed analysis of

musculoskeletal disorder risk factors among Japanese nurses. Journal of Safety

M
Research 37 (2), 195-200.

Solidaki, E., Chatzi, L., Bitsios, P., Markatzi, I., Plana, E., Castro, F., Palmer, K., Coggon,
d

D., Kogevinas, M., 2010. Work-related and psychological determinants of multisite

te

musculoskeletal pain. Scandinavian Journal of Work Environment and Health 36

(1), 54-61.
p

Sorour, A. S., El-Maksoud, M. M. A., 2012. Relationship between musculoskeletal

ce

disorders, job demands, and burnout among emergency nurses. Advanced

emergency nursing journal, 34(3), 272-282.

Ac

StataCorp. 2009. 2009. Stata Statistical Software: Release 11. College Station, TX:

StataCorp .

Stone, PW., Du, Y., Gershon, R., 2007. Organizational climate and occupational health

outcomes in hospital nurses. Journal of Occupational and Environmental Medicine

45 (59), 50-58.

25
Page 24 of 45
Surawera, I. K., Hoe, V. C., Kelsall, H. L., Urquhart, D. M., Sim, M. R. 2012. Physical

and psychosocial factors associated with wrist or hand pain among Australian

hospital-based nurses. Injury prevention, injuryprev 19(1), 13-8.

Trinkoff, AM., Lipscomb, JA., Geiger-Brown, J., Brady, B., 2002. Musculoskeletal

t
problems of the neck, shoulder, and back and functional consequences in nurses.

ip
American Journal of Industrial Medicine 41 (3), 170-178.

cr
van der Windt, DA., Thomas, E., Pope, DP., de Winter, AF., Macfarlane, GJ., Bouter,

LM., Silman, AJ., 2000. Occupational risk factors for shoulder pain: a systematic

us
review. Occupational and Environmental Medicine 57 (7), 433-442.

an
Violante, FS., Fiori, M., Fiorentini, C., Risi, A., Garagnani, G., Bonfiglioli, R., Mattioli, S.,

2004. Associations of psychosocial and individual factors with three different

M
categories of back disorder among nursing staff. Journal of Occupational Health 46

(2), 100-108.
d

Verbeek JH, Martimo KP, Karppinen J, Kuijer PP, Viikari-Juntura E, Takala EP. 2011.
te

Manual material handling advice and assistive devices for preventing and treating

back pain in workers. Cochrane Database of Systematic Reviews. (6):CD005958.

p

Willy, E., 2003. The prevalence of musculoskeletal pain in Norwegin nurses' aides.
ce

International Archives of Occupational and Environmental Health 76 (8), 625-630.

Warming, S., Precht, DH., Suadicani, P., Ebbehoj, NE., 2009. Musculoskeletal complaints
Ac

among nurses related to patient handling tasks and psychosocial factors--based on

logbook registrations. Applied Ergonomics 40 (2), 569-576.

Weyers, S., Peter, R., Boggild, H., Jeppesen, HJ., Siegrist, J., 2006. Psychosocial work

stress is associated with poor self-rated health in Danish nurses: a test of the effort-

reward imbalance model. Scandinavian Journal of Caring Sciences 20 (1), 26-34.

26
Page 25 of 45
Yip, B, 2002. The association between psychosocial work factors and future low back pain

among nurses in Hong Kong: a prospective study. Psychology, Health & Medicine

7 (2), 223-233.

t
ip
cr
us
an
M
d
p te
ce
Ac

27
Page 26 of 45
Figure(s)

Records identified through Additional records identified

database searching through other sources
(n = 5,788 ) (n = 6)

Records after removal of duplicates

t
(n =3,102 )

ip
Excluded based on title or abstract
(n=3,011):

cr
 Not relevant (n= 2,924)
 Review (n = 87)
Full-text articles assessed for eligibility

us
(n =91)
Excluded (n = 65):
 Other occupation than nurse (n=16)
an  Nursing students (n=1)
 Not measuring association between
MSD and psychosocial factors (n=34)
Studies included in
 Setting not hospital (n= 14)
M
qualitative assessment 
(n = 26)

Excluded (n = 2):
ed

 Low quality (score <70%) (n=2)

Studies included in the review

pt

(n = 24) Excluded (n=7):

 95% CI not provided (n = 1)
ce

 Variables available for the matching

(n=5)
 Reported hazard ratio (n=1)
Ac

Studies included in quantitative

synthesis (meta-analysis)
(n = 17)

 Low-quality (n= 1)
Figure 1. Flow chart for selection of included studies.

Page 27 of 45
Figure(s)

i
cr
us
Figure 2. High demands-low control and low back pain in cross sectional studies

an
Study Year OR (95% CI)

M
Alexopoulos 2003 1.50 (0.92, 2.45)

Smith 2006 1.94 (1.32, 2.86)

ed
Bos 2007 1.09 (1.08, 1.14)

Stone 2007 1.37 (1.05, 1.80)

Choobineh
pt
2009 2.25 (1.26, 4.01)

Magnago 2010 1.36 (0.72, 2.60)

ce
Mehrdad 2010 2.86 (1.10, 7.44)

Golabadi 2013 1.73 (1.18, 2.53)

Ac

Fixed (I-squared = 75.3%, p = 0.000) 1.10 (1.07, 1.13)

Randon 1.56 (1.22, 1.99)

.1 1 10

Page 28 of 45
 i
cr
us
Figure 3. High demands-low control and shoulder pain in cross sectional studies

an
Study Year OR (95% CI)

M
Alexopoulos 2003 1.84 (1.21, 2.82)

Smith 2006 2.07 (1.35, 3.17)

ed
Choobineh 2009 1.70 (1.04, 2.73)

Magnago
pt
2010 1.97 (1.07, 3.64)

Mehrdad 2010 1.30 (0.51, 3.30)

ce
Hoe 2011 2.19 (1.25, 3.83)
Ac

Fixed (I-squared = 0.0%, p = 0.937) 1.89 (1.53, 2.34)

Randon 1.89 (1.53, 2.34)

.1 1 10

Page 29 of 45
 i
cr
us
Figure 4. High demands-low control and neck pain in cross sectional studies

an
Study Year OR (95% CI)

M
Alexopoulos 2003 1.93 (1.24, 2.99)

ed
Smith 2006 1.58 (1.02, 2.51)

Bos 2007 1.01 (0.97, 1.06)

Warming
pt
2009 1.16 (0.24, 5.54)

Magnago 2010 1.43 (0.75, 2.73)

ce
Mehrdad 2010 0.95 (0.37, 2.42)

Hoe 2011 1.51 (0.88, 2.59)

Ac

Fixed (I-squared = 59.9%, p = 0.021) 1.03 (0.98, 1.07)

Randon 1.34 (1.02, 1.78)

.1 1 10

Page 30 of 45
 i
cr
us
Figure [Link] demands-low control and back pain in cohort studies

an
Year

M
Study publication OR (95% CI)

ed
Yip 2002 1.85 (1.00, 3.42)

Smith
pt 2004 1.14 (0.68, 1.91)
ce
Alexopoulos 2006 1.66 (1.10, 2.50)

Fixed (I-squared = 0.0%, p = 0.415) 1.52 (1.14, 2.01)

Ac

Random
1.52 (1.14, 2.01)

.1 1 10

Page 31 of 45
Table(s)

t
ip
Table 1. Characteristics of included studies and association between exposure to psychosocial factors at work and musculoskeletal disorders in nursing professionals.

cr
Study ID, Design, Study Measure Outcome measured Analysis, Quality % Exposure variables Anatomic OR 95%IC
Country Follow-up participants, (psychosocial Adjustment score site

us
period, Sample size, factors and variables (QS)
Response Mean age MSD)
rate
Alexopoulos Cross Nurses and Model Karasek. Musculoskeletal Logistic regression, 11/13 84.6 High job demands Back 1.50 0.92-2.45

an
2003 sectional aides from 6 Nordic complaints of the adjusted for age, High job demands Neck 1.93 1.24-2.99
Greece November hospital Questionnaire back, neck, or gender, physical High job demands Shoulder 1.84 1.21-2.81
2000 and n=420 shoulder were demands Low supervisor Back 1.13 0.69-1.85
march 2001 37 years defined as pain in support Neck 1.68 1.08-2.60

M
84% the past 12 months Low supervisor Shoulder 1.45 0.95-2.22
support
Low supervisor
support

d
Alexopoulos Retrospective Nurses and Karasek model Back pain in the past Logistic regression, 11/14 78.6 High job demands Back 1.66 1.10-2.50
2006 cohort aides Nordic 12 months adjusted for physical
The
Netherlands
and Greece
November
2000-March
2001
n=393
38 years
te
Questionnaire demands
ep
64%
Alexopoulos Cross Nurses Karasek model Musculoskeletal Logistic regression, 10/13 76.9 High job demands Knee 4.60 1.57-13.50
2011 sectional n= 448 Nordic complaints low back adjusted for age, Low co-workers Knee 3.13 1.12-8.78
Greece September to 38 years Questionnaire. pain and knee pain gender, physical support
c

December in the past 12 demands

2007 months
Ac

78%
Bos 2007 Cross Nurses from Copenhagen Complaints during Logistic regression 11/13 84.6 Job demands Back 1.09 1.03-1.14
The sectional 8 hospitals, Psychosocial the past year. multivariate model, Support Back - -
Netherlands January 2001 different Questionnaire Low back, neck or adjusted for physical Control Back 1.00 0.96-1.04
to December areas. Nordic shoulder demands Job demands Neck 1.01 0.97-1.06
2003 n= 3,169 Questionnaire Support Neck 1.02 0.97-1.08
63% 37 years Control Neck 1.03 0.99-1.07

Page 32 of 45
 t
ip
a
Camerino Prospective Nurses from Adaptation of Back pain Logistic regression 11/14 78.6 High demands/low Back 12.43 0.0004

cr
2001 cohort different Karasek model experienced during adjusted age, sex, decision
Italy Period not departments (MONICA the last year physical demands Low decision/ low
specified n=1159 study); demands

us
87% 35 years validated High decision/ high
Italian. demands
Ergonomics High decision/low
questionnaire demands

an
Carugno Cross Nurses from Adaptation of Pain at three or Logistic regression, 12/13 92.3 Job dissatisfaction Multi-site 1.50 0.86-2.63
2012 sectional public Karasek model more sites in past adjusted for age, (Italy)
Brazil and May 2008 hospitals in (CUPID Study month sex, physical Multi-site 2.55 0.63-10.35
Italy and March Brazil and questionnaire) demands (Brazil)

M
2010 Italy Nordic
96% (Brazil) n= 751 Questionnaire
76% (Italy) (50% <40
years)

d
Choobineh Cross Operation Job Content Musculoskeletal Logistic regression 11/13 84.6 Perceived job demands Back 2.25 1.26-4.01
2010
Iran
sectional
February to
September
room nurse
n=375
34 years te
Questionnaire
(JCQ)
Nordic
problem in different adjusted for age,
body regions in the physical demands
past month
high/low demands Shoulder 1.68 1.04-2.73
ep
2007 Questionnaire
80%
De Souza Cross Nurses and Job Content Experienced some Logistic regression 10/13 76.9 Demands-control Back 1.36 0.72-2.60
Magnago sectional aides Questionnaire pain or discomfort adjusted for age, Low demands Neck 1.43 0.75-2.73
c

2010 March- n= 491 (JCQ) during last year smoking, time on (reference category) Shoulder 1.97 1.07-1.64
Ac

Brazil September 38 years Nordic the job, physical Legs 1.51 0.83-2.76
2006 Questionnaire demands Ankles 2.05 1.05-4.02
93% (Brazilian).

Page 33 of 45
 t
ip
Golabadi Cross Nursing Job Content Pain in past 12 Logistic regression, 11/13 84.6 High demands Lower back 1.73 1.18-2.53

cr
2013 sectional professionals Questionnaire months disrupted adjusted for age, High demands Upper back 1.57 1.09-2.25
Iran 2011 from public (JCQ) their daily activities sex, physical
84.5% hospital Nordic demands

us
n=545 Questionnaire
32 years

Herin 2011 Prospective Female Effort-reward Pain or discomfort Multilevel models, 12/14 85.7 Effort-reward Upper limb 9.36 5.86-14.96

an
France cohort nurses from 7 imbalance during the last 7 logistic regression, imbalance Upper limb 1.77 1.31-2.40
2006 - 2008 hospitals (ERI), Nordic days: upper limb adjusted for age, Low support from
90% different Questionnaire (neck, shoulder, sex, physical managers
department elbows. hands, demands

M
n= 2194 wrist)
35 years
Hoe 2011 Cross Nurses Adaptation of Neck or shoulder Multinomial logistic 10/13 76.9 Low support Neck 1.22 0.64-2.33
Australia sectional working in Karasek model pain lasting for regression, adjusted supervisor/co-worker

d
Period not three public (CUPID Study more than 1 day for age, sex, physical High job strain Shoulder 2.26 1.22-4.16
specified hospitals questionnaire) during the previous demands Low support Neck 1.51 0.88-2.59
38,6% n= 1119
42 years
te
Nordic
Questionnaire
month supervisor/co-worker
High job strain
Shoulder 2.19 1.25-3.83
ep
Mehrdad Cross Nurses Psychosocial Musculoskeletal Logistic regression, 11/13 84.6 Task level category Back 2.86 1.10-7.44
2010 sectional n= 317 aspects QPS complaints or adjusted for age, included questions Neck 0.95 0.37-2.42
Iran 2006 – 2007 33 years Nordic symptoms in past 12 gender related job demands Shoulder 1.3 0.51-3.30
91% Questionnaire months and job control Upper limb 1.2 0.36-3.73
c

Knee 1.86 0.64-5.34

6.49 2.29-18.37
Ac

Sembajwe Cross Health care Job Content Pain in past 3 Binomial, 11/13 84.6 High job demands. Multi-site 1.98 1.55-2.53
2013 sectional workers (staff Questionnaire months (low back, multinomial, Low job control. Multi-site 0.98 0.76-1.25
USA October 2009 nurse) from (JCQ) shoulder, neck, and cumulative Low support co-worker Multi-site 1.27 0.43-0.78
and two large Nordic wrist or forearm, logistic regression,
February hospitals Questionnaire knee, ankle or feet) adjusted for age,
2010 n= 1,572 sex, occupation
79% 41 years

Page 34 of 45
 t
ip
Simon 2008 Cross Nurses and Effort- reward Pain in relation to Logistic regression, 10/13 76.9 High effort-reward Back 6.2 5.36-7.16

cr
Germany sectional aides imbalance days begin disabled adjusted for age, imbalance/low for high
2002-2003 n= 16770 (ERI), in the past six country, gender, disability
93% 30 years Copenhagen months and physical demands

us
Psychosocial interference with
Questionnaire daily activities.
Scale von Korff
grading the

an
severity of
chronic pain

Smedley Descriptive Nurses Whitehall II Neck or shoulder Cox regression 11/14 78.6 High demands Neck HR=0.9 0.7- 1.4

M
2002 cohort n=1,239 study pain, for at least hazard ratios (HRs), Low interest Neck HR=1.2 0.9- 1.8
United 18 months 39 years Nordic one month at adjusted for age, Low control Neck HR=1.1 0.8- 1.6
Kingdom 56% Questionnaire baseline BMI, frequently Low support Neck HR=0.9 0.6-1.3
feeling tired, Low satisfaction Neck HR=1.2 0.8-1.8

d
low/tense/under
stress, physical
demands

Smith 2004
China
Retrospective
cohort
Nurses from
5 hospitals in teKarasek model
Nordic
Pain or discomfort
during the last 12
Logistic regression,
adjusted for age,
12/14 85.7 High mental pressure
High mental pressure
Back
Neck
1.14
1.79
0.68-1.91
1.06- 3.03
ep
12 months different Questionnaire months (neck, department, High mental pressure Shoulder 1.69 0.99-2.89
92% department (Chinese) shoulder, upper physical demands High mental pressure Any site 1.65 0.94-2.89
n= 282 back or lower back) Low support Back 1.97 1.16-3.35
34 years Low support Neck 2.52 1.09-6.23
c

Low support Shoulder 2.00 0.90-4.59

Low support Any site 3.16 1.02-13.87
Ac

Smith 2006 Cross Nurses Karasek model Symptoms at certain Logistic regression 11/13 84.6 High mental pressure Back 1.94 1.32-2.86
Japan sectional n=844 Nordic body sites over the (Mantel Haenszel High mental pressure Neck 1.53 1.02-2.31
12 months 32 years Questionnaire previous 12 month method) adjusted High mental pressure Shoulder 2.07 1.35-3.17
72% (Chinese) period for age, physical High mental pressure Any site 1.42 0.83-2.38
demands Low support Back 1.16 0.77-1.74
Low support Neck 1.07 0.71-1.60
Low support Shoulder 0.68 0.44-1.06
Low support Any site 0.68 0.39-1.24

Page 35 of 45
 t
ip
Sorour 2012 Cross Nurses in the Job Content Pain and discomfort, Pearson correlation 10/13 76.9 Job demands Number of Β=0.077 0.14-0.02

cr
Egypt sectional emergency Questionnaire for the past 12 and multiple linear painful
October- department (JCQ) months and past 7 stepwise regression. anatomical
December of public Nordic days in each of body Variables excluded sites

us
2010 hospitals Questionnaire areas by model: sex, age,
Response n= 58 BMI, work duration
rate not 28 years
specified

an
Stone 2007 Cross Nurses Perceived Injury back, Multivariate models 11/13 84.6 Low opportunity for Any site 1.64 1.26-2.12
USA sectional n= 1,551 nursing work shoulder, neck, hip for each outcome, advancement Any site 1.37 1.05-1.80
2004 44 years environment of or leg in the last 4 adjusted for Low unit decision
50% critical care months demographic, making

M
nurses (PNWE). employment
Ad hoc characteristics
questionnaire

d
Surawera Cross Nurses Karasek model Wrist or hand Logistic regression, 11/13 84.6 High job strain Wrist or 1.56 1.03-2.,37
2013 sectional working at Nordic pain in the past adjusted for age, Job insecurity hand
Australia October 2009
and January
2010
39%
three
hospitals
n= 1,111
42 years
te
Questionnaire month sex, physical
demands
Wrist or
hand
1.55 1.04-2.28
ep
Violante Cross Nurses and Karasek model Back disorders acute Multinomial logistic 10/13 76.9 Work environment/ Acute low 1.19 0.89-1.59
2004 sectional aides from Ad hoc and chronic regression, adjusted Job satisfaction: back pain 1.11 0.82-1.50
Italy September university questionnaire for age, body mass Having to do many Chronic 1.15 0.86-1.53
c

1997 hospital for index (BMI), things hurriedly low back 1.32 0.98-1.79
95.2% n= 901 musculoskeleta motherhood, at the same time pain
Ac

35 years l disorders smoking, scoliosis, Gratification from

trauma/fractures of responsibility
spine, pelvis and/or
legs, other
diagnosed spine
pathologies, sport
scores

Page 36 of 45
 t
ip
Warming Cross Internal Logbook Low back, neck, Logistic regression, 10/13 76.9 Time pressure (how Back 1.17 0.27-5.01

cr
2009 sectional medicine and instrument shoulder and knee adjusted for gender, has the time pressure Neck 1.16 0.24-5.54
Denmark 12 months surgery (Gonge 2001) pain at data age, time data been today at work) Knee 0.68 0.06-7.27
92% wards nurses collection collection, physical

us
n=148 demands
33 years

Weyers Cross Nurses and Effort- reward Musculoskeletal Multivariate logistic 11/13 84.6 Effort-reward Any site 4.76 2.38-9.52

an
2006 sectional aides living in imbalance (ERI) complaints regression adjusted imbalance
Denmark 1999 the county of questionnaire anywhere on the for age, smoking,
67.7% North Jutland (Danish) body alcohol
n=367 Ad hoc consumption,

M
41 years questions on physical activity,
musculoskeleta occupational status
l complaints

d
Yip 2002 Prospective Nurses Scale 3 of New low back pain Logistic regression 11/14 78.6 Relationship with Back 1.85 1.00-3.42
China cohort n=236 MMPI in past 12 months model, adjusted for colleagues
12 months
81%
31 years

te Aberdeen’s LBP
Scale
after baseline
interview
physical demands of
work, demographic,
lifestyle factors
c ep
Ac

Page 37 of 45
Table(s)

t
ip
cr
Table 2. Work-related psychosocial risk factors and musculoskeletal disorders. Pooled estimates and heterogeneity values for each subset of
studies according to study design.

us
Subgroup Studies Overall Effect * Heterogenety test Studies ID

an
(n)
Cross sectional studies OR 95% IC I2 P value

Demand- control/Back 8 1.56 1.22-1.99 75.3 <0.001 Choobineh 2010; Bos 2007; De Souza 2010; Mehrdad 2010; Smith 2006;
Stone 2007; Alexopoulos 2003; Golabadi 2013.

M
Demand- control/Neck 7 1.34 1.02-1.78 59.9 0.02 Bos 2007; De Souza 2010; Mehrdad 2010; Smith 2006; Alexopoulos
2003; Warming 2009; Hoe 2011.

d
Demand- control/Shoulder 6 1.89 1.53-2.34 0 0.93 Choobineh 2010; De Souza 2010; Mehrdad 2010; Smith 2006;
Alexopoulos 2003; Hoe 2011.

Demand- control/Knee

Demand- control/any site te

3

3
2.21

1.38
1.07-4.54

1.09-1.75
43.1

0
0.17

0.91
Mehrdad 2010; Alexopoulos 2011; Warming 2009.

Smith 2006; Stone 2007; Sembajwe 2013.

ep
Low social support/Any site 5 1.20 0.91-1.59 70.7 0.01 Bos 2007; Smith 2006; Alexopoulos 2003; Alexopoulos 2011; Sembajwe
2013.

Effort reward imbalance/Any site 2 6.13 5.32-7.07 0 0.47 Simon 2008; Weyers, 2006.
c

Cohort studies
Ac

Demand- control/Back 3 1.52 1.14-2.01 0 0.42 Alexopoulos, 2006; Yip, 2002; Smith, 2004

Low social support/Back 3 1.82 1.43-2.32 0 0.94 Alexopoulos, 2006; Yip, 2002; Smith, 2004

I2 = 0%: no heterogeneity; I2 = around 25%: low heterogeneity; I2 = around 50%: moderate heterogeneity; I2 = around 75%: high heterogeneity.

Page 38 of 45
Table(s)

t
ip
Table 1. Characteristics of included studies and association between exposure to psychosocial factors at work and musculoskeletal disorders in nursing professionals.

cr
Study ID, Design, Study Measure Outcome measured Analysis, Quality % Exposure variables Anatomic OR 95%IC
Country Follow-up participants, (psychosocial Adjustment score site

us
period, Sample size, factors and variables (QS)
Response Mean age MSD)
rate
Alexopoulos Cross Nurses and Model Karasek. Musculoskeletal Logistic regression, 11/13 84.6 High job demands Back 1.50 0.92-2.45

an
2003 sectional aides from 6 Nordic complaints of the adjusted for age, High job demands Neck 1.93 1.24-2.99
Greece November hospital Questionnaire back, neck, or gender, physical High job demands Shoulder 1.84 1.21-2.81
2000 and n=420 shoulder were demands Low supervisor Back 1.13 0.69-1.85
march 2001 37 years defined as pain in support Neck 1.68 1.08-2.60

M
84% the past 12 months Low supervisor Shoulder 1.45 0.95-2.22
support
Low supervisor
support

d
Alexopoulos Retrospective Nurses and Karasek model Back pain in the past Logistic regression, 11/14 78.6 High job demands Back 1.66 1.10-2.50
2006 cohort aides Nordic 12 months adjusted for physical
The
Netherlands
and Greece
November
2000-March
2001
n=393
38 years
te
Questionnaire demands
ep
64%
Alexopoulos Cross Nurses Karasek model Musculoskeletal Logistic regression, 10/13 76.9 High job demands Knee 4.60 1.57-13.50
2011 sectional n= 448 Nordic complaints low back adjusted for age, Low co-workers Knee 3.13 1.12-8.78
Greece September to 38 years Questionnaire. pain and knee pain gender, physical support
c

December in the past 12 demands

2007 months
Ac

78%
Bos 2007 Cross Nurses from Copenhagen Complaints during Logistic regression 11/13 84.6 Job demands Back 1.09 1.03-1.14
The sectional 8 hospitals, Psychosocial the past year. multivariate model, Support Back - -
Netherlands January 2001 different Questionnaire Low back, neck or adjusted for physical Control Back 1.00 0.96-1.04
to December areas. Nordic shoulder demands Job demands Neck 1.01 0.97-1.06
2003 n= 3,169 Questionnaire Support Neck 1.02 0.97-1.08
63% 37 years Control Neck 1.03 0.99-1.07

Page 39 of 45
 t
ip
a
Camerino Prospective Nurses from Adaptation of Back pain Logistic regression 11/14 78.6 High demands/low Back 12.43 0.0004

cr
2001 cohort different Karasek model experienced during adjusted age, sex, decision
Italy Period not departments (MONICA the last year physical demands Low decision/ low
specified n=1159 study); demands

us
87% 35 years validated High decision/ high
Italian. demands
Ergonomics High decision/low
questionnaire demands

an
Carugno Cross Nurses from Adaptation of Pain at three or Logistic regression, 12/13 92.3 Job dissatisfaction Multi-site 1.50 0.86-2.63
2012 sectional public Karasek model more sites in past adjusted for age, (Italy)
Brazil and May 2008 hospitals in (CUPID Study month sex, physical Multi-site 2.55 0.63-10.35
Italy and March Brazil and questionnaire) demands (Brazil)

M
2010 Italy Nordic
96% (Brazil) n= 751 Questionnaire
76% (Italy) (50% <40
years)

d
Choobineh Cross Operation Job Content Musculoskeletal Logistic regression 11/13 84.6 Perceived job demands Back 2.25 1.26-4.01
2010
Iran
sectional
February to
September
room nurse
n=375
34 years te
Questionnaire
(JCQ)
Nordic
problem in different adjusted for age,
body regions in the physical demands
past month
high/low demands Shoulder 1.68 1.04-2.73
ep
2007 Questionnaire
80%
De Souza Cross Nurses and Job Content Experienced some Logistic regression 10/13 76.9 Demands-control Back 1.36 0.72-2.60
Magnago sectional aides Questionnaire pain or discomfort adjusted for age, Low demands Neck 1.43 0.75-2.73
c

2010 March- n= 491 (JCQ) during last year smoking, time on (reference category) Shoulder 1.97 1.07-1.64
Ac

Brazil September 38 years Nordic the job, physical Legs 1.51 0.83-2.76
2006 Questionnaire demands Ankles 2.05 1.05-4.02
93% (Brazilian).

Page 40 of 45
 t
ip
Golabadi Cross Nursing Job Content Pain in past 12 Logistic regression, 11/13 84.6 High demands Lower back 1.73 1.18-2.53

cr
2013 sectional professionals Questionnaire months disrupted adjusted for age, High demands Upper back 1.57 1.09-2.25
Iran 2011 from public (JCQ) their daily activities sex, physical
84.5% hospital Nordic demands

us
n=545 Questionnaire
32 years

Herin 2011 Prospective Female Effort-reward Pain or discomfort Multilevel models, 12/14 85.7 Effort-reward Upper limb 9.36 5.86-14.96

an
France cohort nurses from 7 imbalance during the last 7 logistic regression, imbalance Upper limb 1.77 1.31-2.40
2006 - 2008 hospitals (ERI), Nordic days: upper limb adjusted for age, Low support from
90% different Questionnaire (neck, shoulder, sex, physical managers
department elbows. hands, demands

M
n= 2194 wrist)
35 years
Hoe 2011 Cross Nurses Adaptation of Neck or shoulder Multinomial logistic 10/13 76.9 Low support Neck 1.22 0.64-2.33
Australia sectional working in Karasek model pain lasting for regression, adjusted supervisor/co-worker

d
Period not three public (CUPID Study more than 1 day for age, sex, physical High job strain Shoulder 2.26 1.22-4.16
specified hospitals questionnaire) during the previous demands Low support Neck 1.51 0.88-2.59
38,6% n= 1119
42 years
te
Nordic
Questionnaire
month supervisor/co-worker
High job strain
Shoulder 2.19 1.25-3.83
ep
Mehrdad Cross Nurses Psychosocial Musculoskeletal Logistic regression, 11/13 84.6 Task level category Back 2.86 1.10-7.44
2010 sectional n= 317 aspects QPS complaints or adjusted for age, included questions Neck 0.95 0.37-2.42
Iran 2006 – 2007 33 years Nordic symptoms in past 12 gender related job demands Shoulder 1.3 0.51-3.30
91% Questionnaire months and job control Upper limb 1.2 0.36-3.73
c

Knee 1.86 0.64-5.34

6.49 2.29-18.37
Ac

Sembajwe Cross Health care Job Content Pain in past 3 Binomial, 11/13 84.6 High job demands. Multi-site 1.98 1.55-2.53
2013 sectional workers (staff Questionnaire months (low back, multinomial, Low job control. Multi-site 0.98 0.76-1.25
USA October 2009 nurse) from (JCQ) shoulder, neck, and cumulative Low support co-worker Multi-site 1.27 0.43-0.78
and two large Nordic wrist or forearm, logistic regression,
February hospitals Questionnaire knee, ankle or feet) adjusted for age,
2010 n= 1,572 sex, occupation
79% 41 years

Page 41 of 45
 t
ip
Simon 2008 Cross Nurses and Effort- reward Pain in relation to Logistic regression, 10/13 76.9 High effort-reward Back 6.2 5.36-7.16

cr
Germany sectional aides imbalance days begin disabled adjusted for age, imbalance/low for high
2002-2003 n= 16770 (ERI), in the past six country, gender, disability
93% 30 years Copenhagen months and physical demands

us
Psychosocial interference with
Questionnaire daily activities.
Scale von Korff
grading the

an
severity of
chronic pain

Smedley Descriptive Nurses Whitehall II Neck or shoulder Cox regression 11/14 78.6 High demands Neck HR=0.9 0.7- 1.4

M
2002 cohort n=1,239 study pain, for at least hazard ratios (HRs), Low interest Neck HR=1.2 0.9- 1.8
United 18 months 39 years Nordic one month at adjusted for age, Low control Neck HR=1.1 0.8- 1.6
Kingdom 56% Questionnaire baseline BMI, frequently Low support Neck HR=0.9 0.6-1.3
feeling tired, Low satisfaction Neck HR=1.2 0.8-1.8

d
low/tense/under
stress, physical
demands

Smith 2004
China
Retrospective
cohort
Nurses from
5 hospitals in teKarasek model
Nordic
Pain or discomfort
during the last 12
Logistic regression,
adjusted for age,
12/14 85.7 High mental pressure
High mental pressure
Back
Neck
1.14
1.79
0.68-1.91
1.06- 3.03
ep
12 months different Questionnaire months (neck, department, High mental pressure Shoulder 1.69 0.99-2.89
92% department (Chinese) shoulder, upper physical demands High mental pressure Any site 1.65 0.94-2.89
n= 282 back or lower back) Low support Back 1.97 1.16-3.35
34 years Low support Neck 2.52 1.09-6.23
c

Low support Shoulder 2.00 0.90-4.59

Low support Any site 3.16 1.02-13.87
Ac

Smith 2006 Cross Nurses Karasek model Symptoms at certain Logistic regression 11/13 84.6 High mental pressure Back 1.94 1.32-2.86
Japan sectional n=844 Nordic body sites over the (Mantel Haenszel High mental pressure Neck 1.53 1.02-2.31
12 months 32 years Questionnaire previous 12 month method) adjusted High mental pressure Shoulder 2.07 1.35-3.17
72% (Chinese) period for age, physical High mental pressure Any site 1.42 0.83-2.38
demands Low support Back 1.16 0.77-1.74
Low support Neck 1.07 0.71-1.60
Low support Shoulder 0.68 0.44-1.06
Low support Any site 0.68 0.39-1.24

Page 42 of 45
 t
ip
Sorour 2012 Cross Nurses in the Job Content Pain and discomfort, Pearson correlation 10/13 76.9 Job demands Number of Β=0.077 0.14-0.02

cr
Egypt sectional emergency Questionnaire for the past 12 and multiple linear painful
October- department (JCQ) months and past 7 stepwise regression. anatomical
December of public Nordic days in each of body Variables excluded sites

us
2010 hospitals Questionnaire areas by model: sex, age,
Response n= 58 BMI, work duration
rate not 28 years
specified

an
Stone 2007 Cross Nurses Perceived Injury back, Multivariate models 11/13 84.6 Low opportunity for Any site 1.64 1.26-2.12
USA sectional n= 1,551 nursing work shoulder, neck, hip for each outcome, advancement Any site 1.37 1.05-1.80
2004 44 years environment of or leg in the last 4 adjusted for Low unit decision
50% critical care months demographic, making

M
nurses (PNWE). employment
Ad hoc characteristics
questionnaire

d
Surawera Cross Nurses Karasek model Wrist or hand Logistic regression, 11/13 84.6 High job strain Wrist or 1.56 1.03-2.,37
2013 sectional working at Nordic pain in the past adjusted for age, Job insecurity hand
Australia October 2009
and January
2010
39%
three
hospitals
n= 1,111
42 years
te
Questionnaire month sex, physical
demands
Wrist or
hand
1.55 1.04-2.28
ep
Violante Cross Nurses and Karasek model Back disorders acute Multinomial logistic 10/13 76.9 Work environment/ Acute low 1.19 0.89-1.59
2004 sectional aides from Ad hoc and chronic regression, adjusted Job satisfaction: back pain 1.11 0.82-1.50
Italy September university questionnaire for age, body mass Having to do many Chronic 1.15 0.86-1.53
c

1997 hospital for index (BMI), things hurriedly low back 1.32 0.98-1.79
95.2% n= 901 musculoskeleta motherhood, at the same time pain
Ac

35 years l disorders smoking, scoliosis, Gratification from

trauma/fractures of responsibility
spine, pelvis and/or
legs, other
diagnosed spine
pathologies, sport
scores

Page 43 of 45
 t
ip
Warming Cross Internal Logbook Low back, neck, Logistic regression, 10/13 76.9 Time pressure (how Back 1.17 0.27-5.01

cr
2009 sectional medicine and instrument shoulder and knee adjusted for gender, has the time pressure Neck 1.16 0.24-5.54
Denmark 12 months surgery (Gonge 2001) pain at data age, time data been today at work) Knee 0.68 0.06-7.27
92% wards nurses collection collection, physical

us
n=148 demands
33 years

Weyers Cross Nurses and Effort- reward Musculoskeletal Multivariate logistic 11/13 84.6 Effort-reward Any site 4.76 2.38-9.52

an
2006 sectional aides living in imbalance (ERI) complaints regression adjusted imbalance
Denmark 1999 the county of questionnaire anywhere on the for age, smoking,
67.7% North Jutland (Danish) body alcohol
n=367 Ad hoc consumption,

M
41 years questions on physical activity,
musculoskeleta occupational status
l complaints

d
Yip 2002 Prospective Nurses Scale 3 of New low back pain Logistic regression 11/14 78.6 Relationship with Back 1.85 1.00-3.42
China cohort n=236 MMPI in past 12 months model, adjusted for colleagues
12 months
81%
31 years

te Aberdeen’s LBP
Scale
after baseline
interview
physical demands of
work, demographic,
lifestyle factors
c ep
Ac

Page 44 of 45
Table(s)

t
ip
cr
Table 2. Work-related psychosocial risk factors and musculoskeletal disorders. Pooled estimates and heterogeneity values for each subset of
studies according to study design.

us
Subgroup Studies Overall Effect * Heterogenety test Studies ID

an
(n)
Cross sectional studies OR 95% IC I2 P value

Demands- control/Back 8 1.56 1.22-1.99 75.3 <0.001 Choobineh 2010; Bos 2007; De Souza 2010; Mehrdad 2010; Smith 2006;
Stone 2007; Alexopoulos 2003; Golabadi 2013.

M
Demands- control/Neck 7 1.34 1.02-1.78 59.9 0.02 Bos 2007; De Souza 2010; Mehrdad 2010; Smith 2006; Alexopoulos
2003; Warming 2009; Hoe 2011.

d
Demands- control/Shoulder 6 1.89 1.53-2.34 0 0.93 Choobineh 2010; De Souza 2010; Mehrdad 2010; Smith 2006;
Alexopoulos 2003; Hoe 2011.

Demands- control/Knee

Demands- control/any site te

3

3
2.21

1.38
1.07-4.54

1.09-1.75
43.1

0
0.17

0.91
Mehrdad 2010; Alexopoulos 2011; Warming 2009.

Smith 2006; Stone 2007; Sembajwe 2013.

ep
Low social support/Any site 5 1.20 0.91-1.59 70.7 0.01 Bos 2007; Smith 2006; Alexopoulos 2003; Alexopoulos 2011; Sembajwe
2013.

Effort reward imbalance/Any site 2 6.13 5.32-7.07 0 0.47 Simon 2008; Weyers, 2006.
c

Cohort studies
Ac

Demands- control/Back 3 1.52 1.14-2.01 0 0.42 Alexopoulos, 2006; Yip, 2002; Smith, 2004

Low social support/Back 3 1.82 1.43-2.32 0 0.94 Alexopoulos, 2006; Yip, 2002; Smith, 2004

I2 = 0%: no heterogeneity; I2 = around 25%: low heterogeneity; I2 = around 50%: moderate heterogeneity; I2 = around 75%: high heterogeneity.

Page 45 of 45