Association between physical and psychosocial demands and musculoskeletal pain in health workers

Introduction The physical and psychosocial demands of work are important factors in the genesis of musculoskeletal pain. Identification of these dimensions and their interfaces with workers’ individual characteristics could improve understanding of these outcomes. Objectives To analyze the associations between the physical and psychosocial demands of work and occurrence of musculoskeletal pain in health care workers. Methods This was a cross-sectional study conducted with health care workers. The exposure variables were psychosocial aspects and physical demands, investigated using the Job Content Questionnaire, and the outcomes were musculoskeletal pain in lower limbs, upper limbs, and the back, investigated as self-report of pain. A multivariate analysis was conducted to investigate associations between exposures and outcomes. Results The factors associated with musculoskeletal pain in the three areas of the body studied were female sex, physical inactivity, and “poor” self-rated health status. Additionally, being a contract worker was associated with musculoskeletal pain in the lower limbs and back. Not participating in leisure activities and being responsible for direct provision of health care were associated with pain in the lower limbs. Being the person responsible for the housework and doing the housework were associated with pain in upper limbs. Differences between the demands of tasks, poor availability of the technical resources to perform activities, and absence of leisure activities were associated with back pain. Conclusions It was concluded that both physical demands and psychosocial demands are associated with musculoskeletal pain in health care workers.


INTRODUCTION
Sickness caused by employment activities, whether physical or emotional, has increased progressively in many occupational categories. This is true of people who work in health care, whether they are workers who provide care directly or are involved in service provision. 1 One class of physical conditions that affect workers' health is musculoskeletal disorders, 2 which includes inflammation, pain, and degeneration of muscles, tendons, joints, nerves, and cartilage, causing functional limitations. These disorders are caused by excessive use of the musculoskeletal system compounded by a lack of time for recovery 3 and can result from the action of multiple factors inherent to employment activity, especially the physical and psychosocial demands of work. 4 In addition to the excessive demand on an individual's capacity, activities that require use of extreme force can also impair circulation to the muscles, leading to tension and exhaustion and increasing the time needed for recovery after performing the task. Excessive effort combined with little or no time for recovery can provoke pain. 5 Negative psychosocial aspects of work include monotony, work overload, time pressure, little control over one's own work, lack of social support, poor relationships with colleagues, lack of autonomy to perform tasks, and disorganized working procedures and these factors can also be related to emergence of musculoskeletal pain (MSP). 6 This is because of increased muscle tension and a reduction in the workers' capacity to deal with symptoms, increasing the perception of pain. 6 Activities performed in inappropriate positions, repetitive and monotonous tasks, with no rest breaks, with inappropriate furniture and equipment, and those involving lifting and carrying loads contribute to MSP, as does work performed under high psychosocial demands, resulting in greater physical effort, excessive working hours, few breaks for rest, and insufficient changes of position. 7 In addition to the physical and psychosocial demands, personal characteristics such as age and sex, lifestyle habits, comorbidities, 8 and educational level 9 can also contribute to occurrence of musculoskeletal symptomology. 8 Although the personal characteristics cannot be changed, individual behavior and the characteristics of work can be changed in order to improve working conditions and, consequently, reduce its impact on workers' health. 10 In view of the above, the objective of this study was to analyze the association between the physical and psychosocial demands of work and occurrence of MSP in health care workers.

STUDY DESIGN
This was a cross-sectional study focused on investigating the association between physical and psychosocial demands and occurrence of MSP in health care workers from six towns in Bahia, Brazil, in 2012. The study is derived from the multicenter project "Working conditions, employment conditions, and health of health care workers in Bahia" (

STUDY PARTICIPANTS
The underlying study population comprised workers attached to primary care and medium complexity care services in the towns of Feira de Santana, Santo Antônio de Jesus, Jequié, Salvador, Itabuna, and Itaberaba who were working full time and agreed to participate in the project. A representative sample was selected using a randomization procedure employing sampling stratified by geographic area, level of care complexity, and workers' occupational categories, calculated using a formula for a population of 6,191 health care workers from the five towns, with a 79.2% vaccination prevalence, 3% error, and 95% confidence interval (95%CI). Based on these parameters, the sample size would be 763 workers. While the study was ongoing, an additional center was included, with a population of 502 workers, raising the total to 6,693 workers. However, a total sample of 3,343 people were interviewed.
The study power was calculated for combined evaluation of psychosocial aspects and MSP. Considering an overall prevalence of MSP of 53.4% among exposed workers (exposed to high physical demand and high psychosocial demand) and a 31% prevalence among unexposed workers (exposed to low physical demand and low psychosocial demand), the study power was 99.9%.

DATA COLLECTION
Data were collected in 2011 and 2012 using a questionnaire made up of sections containing questions on sociodemographic, occupational, and organizational characteristics of work; lifestyle; and health problems. After contacting the municipal health departments responsible for each participating center and the administrations of each of these health centers, the units were contacted and data collection forms were distributed to the workers who had been selected (the forms were administered to workers with secondary education). The team of interviewers was trained before data collection to standardize behavior and procedures. A pilot study was run in a town in Bahia with 30 health care workers.

OUTCOME
The study's outcome variable was MSP, investigated by self-report pain indicated on a Likert response scale (never, rarely, infrequently, frequently, and very frequently). The body areas studied were lower limbs (LL), upper limbs (UL), lower back, and upper back. The last two categories were re-categorized as back pain.
The response options "frequently" or "very frequently" were defined as cases of MSP and the options "infrequently", "rarely", and "never" were defined as absence of MSP. This approach has been employed successfully to assess presence of MSP among workers in Brazil. 11,12

PRINCIPAL EXPOSURE VARIABLES
The exposure variables were physical and psychosocial demands, investigated using the Job Content Questionnaire ( JCQ). This instrument contains questions with Likert response scales (1 = completely disagree, 2 = disagree, 3 = agree, and 4 = completely agree) for psychosocial aspects (control over one's own job and psychological demands) and physical demands. The responses were used to calculate scores for each of these scales, as described in the JCQ Manual (www.JCQCenter.org). Later, once the estimated scores had been collected, these variables were dichotomized, adopting the median of the distribution of each scale as its cutoff point. Values below the median were classified as low demand and values over the median were defined as high demand. These data were used to define groups as undemanding work (a combination of low demand and high control), passive work (low demand and low control), active work (high demand and high control) and highly demanding work (high demand and low control). 13 Participants were then allocated to one of two groups: a group with high psychosocial exposure, comprising those who were under high psychological demand, had low control over their work, and had low social support; and a group with low psychosocial exposure, comprising those who were under low psychological demand, had high control over their work, and had high social support.
After this categorization (dichotomization of the variables), workers were classified into groups according to whether they had exposure to physical and psychosocial demands, so that those who were unexposed to psychosocial aspects and unexposed to physical demands were classified as P 00 ; those who were unexposed to high physical demand, but exposed to high psychosocial demand were classified as P 01 ; those exposed to high physical demand and unexposed to high psychosocial demand were allocated to P 10 ; and those who were exposed to both psychosocial aspects and physical demand, were classified as P 11 . 14

COVARIATES
The covariates considered for analysis were organized in blocks as follows: participation in leisure activities and physical activity (yes; no); self-rated health status (good; poor).

ANALYSIS OF THE DATA
Workers' characteristics were analyzed by calculating the frequencies of the covariates in the blocks described above and the combined exposure variable. Exploratory bivariate analyses were then conducted to guide inclusion of variables in the logistic regression model. Variables with p ≤ 0.20 were selected for modeling.
The association analysis was based on a hierarchical conceptual model (Figure 1), constructed on the basis of the proximal-distal relationships between the variables in the blocks and the outcome. This type of analysis can be used to work with large numbers of variables, 15,16 which are inserted in stages, starting with the most distal belonging to the same block, simultaneously with the outcome. 16,17 In the inter-block hierarchical stage, variables with p ≤ 0.05 in the bivariate analyses and variables with theoretical relevance according to the literature were retained in the model. Variables that did not exhibit associations in the intrablock analysis were removed when entering the subsequent block.

ETHICAL QUESTIONS
The study respected all ethical principles. The project was approved by the research ethics committee at Universidade Estadual de Feira de Santana, protocol 081/2009 and Ethics Appraisal Submission Certificate number 0086.0.059.000.09.

RESULTS
A total of 3,343 health care workers were interviewed. Of these, 23.2% were not exposed to any of the exposures (high psychosocial or physical demand), 25.5% were only exposed to high psychosocial demand, 16.4%, were only exposed to high physical demand, and 34.9% were exposed both to high physical demand and to high psychosocial demand.
According to Table 1, when other combined exposure categories were compared, it was observed that group P 11 (high physical demand and high psychosocial demand) had a predominance of female workers (80.3%), those in the age group 19 to 33 years (36.7%), those married/in a stable consensual relationship (57.3%), those with secondary education/ technical college (58.9%) and black/brown skin color (81.3%), those who do the housework (93.9%) and are the person responsible for the housework (55.6%), and those who do not participate in leisure activities (20.1%), who are physically inactive (52.7%), and who rate their own health status as poor (5.2%).
With relation to job characteristics and working environment (Table 2), participants in group P 11 were predominantly support staff (31.5%); those with 13 years or more in the job (27.4%); performing activities compatible with their jobs (1.3%); working on call (17.5%); working ≥ 40 hours per week (72.3%); and those who did not have another job (80.4%). Similarly, workers in P 11 rated the condition of tables and chairs as poor (44.9%), technical resources and equipment as poor (43.4%) the level of demand of tasks as poor, and the resources available as poor (28.4%). It was also observed that those who always stand up, rarely sit down, or always walk to perform their activities accounted, respectively, for 56.2, 42, and 61.5% of the P 11 category. Finally, those who always lift, carry, or push loads and have breaks during the work shift were, respectively, 28.9 and 56.5% of P 11 .
The prevalence of MSP among workers exposed to both high physical demand and high psychosocial demand was 38.1% in the LL, 29.8% in the UL, and 46.2% in the back.    P 00 = low physical demand and low psychosocial demand; P 01 = low physical demand and high psychosocial demand; P 10 = high physical demand and low psychosocial demand; P 11 = high physical demand and high psychosocial demand.  ; type of employment "on contract" was associated, but as a protective factor (PR: 0.66; 95%CI: 0.54-0.79) ( Table 5).

DISCUSSION
The present study investigated associations between physical and psychosocial demands and occurrence of MSP in health care workers using hierarchical analysis. The following factors were associated with MSP in all three body areas assessed: female sex, lack of physical activity, and "poor" self-rated health status. Being a contract worker was a protective factor for MSP in the LL and back.
It was observed that high percentages of workers were exposed to physical and psychosocial demands. These factors were associated with musculoskeletal disorders. 18,19 Although this study was restricted to health care workers, this finding is similar to findings from research conducted with other categories of workers in Brazil 4,20-22 and worldwide. 7 Absence of leisure activity was associated with MSP in two of the body areas studied. Exposure to physical demands at work may affect leisure activities and physical activities, 22 since, for many workers, limited free time outside of work may mean that they use their available free time to rest and recover from the physical effort they have exerted at work, 20 contributing to reduce their engagement in these practices (even though they could function to promote mental and physical health).
In line with the results of the present study, Barbosa et al. 10 also found associations between a lack of physical activity and musculoskeletal complaints in health care workers. Resistance and muscle strength acquired during physical activity protect workers from MSP and reduce its impacts on health. This is irrespective of the conditions in which physical work is performed. 20 "Poor" self-rated health status was associated with MSP. The high prevalence rates and powerful associations between these variables may reflect a preexisting MSP condition, considering that this is an influence on living conditions at work and capable of affecting the worker's opinion.
Direct provision of health care was another variable that was associated with MSP, which confirms the findings of other investigations that show that the job categories dentist, dental auxiliary, and community health worker were also associated with this condition. 10 The physical and psychological organizational demands to which health professionals are subjected contribute to this situation. 23 Repetition of movements and remaining in a static position for long periods while working increase risk of MSP, 24 because they cause compression of musculoskeletal structures. Repetition and static positions are both characteristic of the activities performed by health care professionals. 25 The findings of the present study demonstrate that workers who are on contracts had lower prevalence of MSP in the LL and back. According to published data, better health status would traditionally be expected in people with stable employment. It is therefore necessary to analyze this unexpected study result in greater detail. One likely explanation is related to the profile of health care employment. Recently, employment of more qualified workers (professionals with higher education qualifications in medicine, nursing, and dentistry) has been contracted on a temporary basis in Brazil, in particular because the standard public competitions for vacancies have not been held. In contrast, other occupations at the middle levels, such community health workers, are subject to legislation that makes public competitions mandatory. As a result, the advantages of a permanent job may not be sufficient to guarantee working conditions with less exposure. Additionally, it is also possible that workers with permanent jobs do not perform roles involving as much direct health care provision (which was one of the variables associated with pain in the body areas studied) as other jobs. It is therefore plausible to suppose that they may be less exposed to high physical and psychosocial demands. However, the prevalence rates of MSP in contract workers were not very different between those with jobs that do and do not provide direct health care, at 20.6 and 17.2% in LL and 28.1 and 24.3% in the back, respectively, although they were higher in those who do provide direct care.
Female sex was associated with MSP in all three areas of the body studied, confirming other studies that have investigated the subject. 9,10,22,26,27 Factors such as a double work load (which reduces the time available for physical activity or relaxation, thus limiting activities that protect from injuries and help to prevent them 10,28 ) and the anthropometric differences between the sexes and between the muscle fibers of men and women (who are weaker and less resistant) can also contribute to emergence of musculoskeletal injuries. 29 The double work load of women is clear from findings observed over the course of this study. Majorities of them do the housework and/or are responsible for the housework, are married, and have children. As such, the differences between the sexes should be taken into account in fair division of tasks in the home, bearing in mind the double work load put on women, and their physiological characteristics. 22 Other important factors that could contribute to this finding, particularly in relationship to spinal complaints, are pregnancy and postpartum, since the changes to the body and hormones that occur during pregnancy lead to greater flexibility in the spine and hips, causing changes to these structures. 9 Doing the housework and being the principal person responsible for it, which are both factors that contribute to emergence of painful symptomology in women, were both associated with back pain. These findings show that women who reported a painful complaint in this region were also responsible for doing the housework (93.3%).
The present study is subject to limitations. One of these is the cross-sectional study design, which cannot establish cause and effect relationships, suggesting a need for longitudinal studies of the subject. The healthy worker bias should also be acknowledged, since workers who were off sick or had left the profession did not take part in the study.

CONCLUSIONS
High prevalence rates of exposure to physical and psychosocial demands were observed in the present study. These were associated with MSP in the LL, in the UL, and in the back in analyses with the hierarchical model, with a particular emphasis on the associations when both exposures and high physical demand were present, reinforcing the hypothesis that these characteristics are important factors for occurrence of the outcome.
The results of the hierarchical analysis show the importance of considering the multiple exposures to which workers are exposed and the relationships that are established between the different exposure factors, offering more in-depth understanding and the chance to seek measures that could reduce the impacts of the physical and psychosocial demands of work and, consequently, the rate of MSP in workers. Such measures could be implemented in the form of actions that lead to restructuring of organizational aspects, minimization of work demands, and improvement of working conditions. Additionally, the associations between female sex and MSP reveal a need to reflect on strategies for prevention in this group.
Author contributions GSVD and JPC were both responsible for the study conceptualization, data curation, formal analysis of data, and writing -original draft and revision and editing of the final text. TMA was responsible for the study conceptualization, funding acquisition, investigation, and writing -review & editing of the final text All authors have read and approved the final version submitted and take public responsibility for all aspects of the work.