You are here

  • Home
  • Archive
  • Volume 14, Issue 8
  • Identifying pain profiles in employees including work-related factors and pain perceptions: a cross-sectional study in Belgian companies

Article Text

Article menu
Occupational and environmental medicine
Original research
Identifying pain profiles in employees including work-related factors and pain perceptions: a cross-sectional study in Belgian companies
  1. Stijn Keyaerts1,2,
  2. Lode Godderis1,2,
  3. Vero Vanden Abeele3,
  4. Liesbeth Daenen1,2,4
  1. 1 Department of Public Health and Primary Care, KU Leuven, Leuven, Belgium
  2. 2 Knowledge Information and Research Center, Group Idewe, Leuven, Belgium
  3. 3 Department of Computer Science, KU Leuven, Leuven, Belgium
  4. 4 Department of Physiotherapy, Human Physiology and Anatomy, VU Brussels, Brussels, Belgium
  1. Correspondence to Stijn Keyaerts; stijn.keyaerts{at}idewe.be

Abstract

Objectives Studies usually investigate a limited number or a predefined combinations of risk factors for sickness absence in employees with pain. We examined frequently occurring combinations across a wide range of work-related factors and pain perceptions.

Design Cross-sectional study.

Setting Belgian companies that are under supervision of IDEWE, an external service for prevention and protection at work.

Participants In total, 249 employees experiencing pain for at least 6 weeks were included and filled out an online survey.

Outcomes Latent profile analysis was used to differentiate profiles of work-related factors (physical demands, workload, social support and autonomy) and pain perceptions (catastrophising, fear-avoidance beliefs and pain acceptance). Subsequently, profiles were compared on sociodemographics (age, gender, level of education, work arrangement, duration of complaints, multisite pain and sickness absence in the previous year) and predictors of sickness absence (behavioural intention and perceived behavioural control).

Results Four profiles were identified. Profile 1 (38.2%) had favourable scores and profile 4 (14.9%) unfavourable scores across all indicators. Profile 2 (33.3%) had relatively high physical demands, moderate autonomy levels and favourable scores on the other indicators. Profile 3 (13.7%) showed relatively low physical demands, moderate autonomy levels, but unfavourable scores on the other indicators. Predictors of profiles were age (OR 0.93 and 95% CI (0.89 to 0.98)), level of education (OR 0.28 and 95% CI (0.1 to 0.79)) and duration of sickness absence in the previous year (OR 2.29 and 95% CI (0.89 to 5.88)). Significant differences were observed in behavioural intention (χ2=8.92, p=0.030) and perceived behavioural control (χ2=12.37, p=0.006) across the four profiles.

Conclusion This study highlights the significance of considering the interplay between work-related factors and pain perceptions in employees. Unfavourable scores on a single work factor might not translate into maladaptive pain perceptions or subsequent sickness absence, if mitigating factors are in place. Special attention must be devoted to employees dealing with unfavourable working conditions along with maladaptive pain perceptions. In this context, social support emerges as an important factor influencing sickness absence.

  • OCCUPATIONAL & INDUSTRIAL MEDICINE
  • Chronic Pain
  • Health Workforce
  • Social Support

Data availability statement

Data are available upon reasonable request. The data that support the findings of this study are available from the corresponding author upon reasonable request.

http://creativecommons.org/licenses/by-nc/4.0/

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.

https://doi.org/10.1136/bmjopen-2023-082804

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    Strengths and limitations of this study

    • This study used latent profile analysis, a rigorous clustering approach, to evaluate interactions in a large set of work-related factors and pain perceptions in employees.

    • To enhance interpretability and utility of the identified profiles, they were further described by comparing sociodemographics and their risk of sickness absence.

    • The risk of sickness absence was measured subjectively, by using a survey that measures work expectations, which might not fully reflect the actual risk of sickness absence.

    • As this study focused on active employees with pain, results should not be generalised to a non-working population with pain.

    Introduction

    Pain complaints account for 60% of long-term sickness absences among European workers.1 Persisting pain is associated with disability and significant costs, exerting a substantial impact on individuals, their families, employers and society.1

    Research has established that both physical factors in the workplace (eg, handling heavy loads, awkward postures and repetitive movements) and psychosocial work-related stressors (eg, high workloads, limited social support or lack of autonomy) contribute to the development and perpetuation of pain complaints and form barriers for work.2–5

    Evidence supports the role of pain perceptions and coping mechanisms in the development and maintenance of pain complaints.2 3 6 The Fear-Avoidance model, for instance, describes how maladaptive responses to pain, characterised by pain catastrophising, fear of (re)injury and avoidance behaviour, can perpetuate complaints and increase disability.7 Consequently, maladaptive pain perceptions and coping strategies have been related to sickness absence.8 9

    However, studying risk factors for pain or pain-related absenteeism in isolation may not fully reflect real-life scenarios where risk factors often co-occur and interact. A cross-sectional study, involving 3003 workers from diverse sectors, found that the combination of unfavourable physical and psychosocial factors in the workplace significantly increases the risk of pain and absenteeism compared with single risk factors.10 Moreover, workers in physically demanding jobs were found to have a reduced risk of experiencing pain complaints when favourable psychosocial factors were present in the workplace. For instance, employees exposed to awkward grip or hand movements had higher odds of wrist complaints compared with those not exposed to any risk factors (OR 2.79 and 95% CI (1.62 to 4.79)). However, the odds of these employees experiencing wrist pain decreased when they were satisfied with the level and difficulty of their work (OR 2.19 and 95% CI (1.44 to 3.34)). Another cross-sectional study of 305 healthcare workers found that fear of movement can moderate the relationship between workplace factors and pain complaints.2 For example, high autonomy levels at work were identified as a risk for pain in employees without fear of movement, whereas high autonomy served as a protective factor for those with fear of movement. Notably, these studies evaluated predefined combinations of risk factors, which might neglect relevant interactions.2 10

    Latent profile analysis (LPA) is a valuable approach for reducing complexity across a large set of factors without neglecting relevant combinations, offering a clear advantage over using predetermined combinations or a fixed set of clusters.11 One research group, for example, identified three profiles with increasing levels of complaint severity in 381 workers with chronic low back pain.12 Workers in the high severity profile had four times greater odds of being on sick leave than those in the low severity profile. In another study, they were able to demonstrate that work-related support and perceived work abilities increased the odds of being classified in a profile characterised by less pain burden and less pain catastrophising.13 To the best of our knowledge, no previous studies included work-related factors as well as pain perceptions to identify profiles of employees with pain. However, evaluating this complex interplay in real-life situations may inform treatment options to improve employee well-being and reduce sickness absence. Therefore, the present study aimed to (1) identify pain profiles of employees based on four work-related factors and five pain perceptions and (2) compare the profiles on sociodemographics and predictors of sickness absence.

    Methods

    Study design

    Participants were recruited by IDEWE, a Belgian Occupational Service for Prevention and Protection at work, between December 2019 and October 2021. IDEWE clients (ie, employers, managers from Belgian organisations) were informed about the study through various channels, such as mailings, social media posts and brochures and were encouraged to share this information with their employees. Interested employees could contact the researchers for more information. Eligible participants were experiencing pain for at least 6 weeks and provided written or electronic informed consent. Those who did not meet these criteria were excluded. Data collection occurred via a self-reported questionnaire that was available in a smartphone application developed by IDEWE (available at Apple App Store and Google Play). This questionnaire was available in Dutch, French and English.

    Measurements

    Indicator variables for statistical subgrouping

    Physical demands were assessed using seven out of 12 items from the Dutch Musculoskeletal Questionnaire.14 Participants estimated how much of their working time ((almost) never, sometimes, often or (almost) always) involved various physical tasks (eg, standing, sitting or moving loads). A three factor structure was retained after exploratory factor analysis (online supplemental table 1). The seven selected items had high factor loadings on the first factor, and the scale had good reliability (Cronbach’s α=0.85). Scores could range from 7 to 28, with higher scores representing more physically demanding jobs. Workload and social support were evaluated using subscales of the Short Inventory to Monitor Psychosocial Hazards (SIMPH).15 The SIMPH is a validated and reliable questionnaire that assesses major psychosocial hazards at work across 11 different subscales.15 The workload subscale, consisting of three items (scores ranging from 3 to 15), measures the perceived amount of work-related tasks and the pressure to complete those tasks within a specified timeframe. The social support subscale, consisting of four items (scores ranging from 4 to 20), measures the perceived availability and appreciation from colleagues and supervisors. Both the workload (Cronbach’s α=0.79) and social support (Cronbach’s α=0.77) subscales showed good reliability, with higher scores indicating higher workload and greater social support. Autonomy at work was measured using a subscale of the Questionnaire on the Experience and Evaluation of Work (V. 2.0), a validated questionnaire measuring psychosocial well-being and work-related stress.16 The autonomy subscale, consisting of four items (scores ranging from 4 to 20), evaluates the degree of control and discretion individuals have over various aspects of their work environment, including task planning and choice of work methods. This scale also demonstrated good reliability (Cronbach’s α=0.79), with higher scores indicating higher levels of autonomy.

    Supplemental material

    Catastrophising was measured using the validated Pain Catastrophising Scale (PCS), which consists of 13 items evaluating rumination, magnification and helplessness.17 The total score could range from 0 to 52, with higher scores indicating greater degrees of catastrophising. Fear-avoidance beliefs were assessed using the validated Fear-Avoidance Beliefs Questionnaire.18 This questionnaire includes two subscales: (1) a work subscale (Fear Avoidance Beliefs Questionnaire Work, FABQW) of seven items (scores ranging from 0 to 42) that measures attitudes about work and its relationship to physical complaints, and (2) a physical activity subscale (Fear Avoidance Beliefs Questionnaire Activity, FABQA) of four items (scores ranging from 0 to 24) that evaluates the perception towards movement and physical activities. Higher scores on each subscale indicate higher levels of fear-avoidance beliefs. The two subscales were used separately for statistical clustering. Pain acceptance was measured with the short version of the validated Chronic Pain Acceptance Questionnaire.19 This questionnaire includes two subscales of four items each, assessing: (1) activity engagement (Chronic Pain Acceptance Questionnaire Activity, CPAQA), which measures the degree of participation in daily activities, and (2) pain willingness (Chronic Pain Acceptance Questionnaire Pain, CPAQP), which measures the willingness to experience pain. Scores on each subscale range from 0 to 24, with higher scores indicating higher acceptance of pain. The two subscales were used separately for statistical clustering. The scales for pain catastrophising (Cronbach’s α=0.91), fear-avoidance beliefs regarding physical activities (Cronbach’s α=0.58) and work (Cronbach’s α=0.79), activity engagement (Cronbach’s α=0.66) and pain willingness (Cronbach’s α=0.61) showed acceptable to excellent reliability.

    Variables for description of latent profiles

    Pain complaints were evaluated using the Dutch Musculoskeletal Questionnaire, where participants reported experiencing pain or discomfort in specific body regions over the past week (ie, neck, shoulders, back, elbows, wrists/hands, hips, knees or ankles/feet).14 Multisite pain was considered if participants indicated pain or discomfort in more than one body part. Participants also provided information on the duration of complaints and the frequency of absenteeism due to pain in the previous year. Additional sociodemographic variables collected included age, gender, level of education, sector of employment and work arrangement (eg, full-time or part-time).

    The risk of sickness absence was assessed using the Behavioural Intention (BI) and Perceived Behavioural Control (PBC) subscales of a validated questionnaire based on the Theory of Planned Behaviour.20 The BI subscale measures the expectations of working or continuing to work in a paid job 3 months from now. The PBC subscale measures the estimated difficulty and level of control over working or continuing to work in a paid job 3 months from now. According to the Theory of Planned Behaviour, engaging in the target behaviour (ie, working despite pain complaints) depends jointly on motivation (ie, intention) and ability (ie, behavioural control).21 Using this questionnaire, Dunstan and colleagues distinguished between employees at work and those on sick leave.20 Both subscales consist of four items, with scores ranging from 0 to 24, and higher scores indicating higher BI and PBC. Both subscales for BI (Cronbach’s α=0.62) and PBC (Cronbach’s α=0.69) demonstrated acceptable reliability.

    Data analysis

    Univariate descriptive statistical analyses were performed using SPSS V.28.0 statistical software package (SPSS, Chicago, IL). Bivariate correlations of the indicator variables did not exceed 0.85, indicating no problems with multicollinearity (online supplemental table 2).22 All other analyses were performed using the MPlus V.8.6 statistical software package (Muthén & Muthén, Los Angeles, CA). LPA was conducted to identify profiles of employees with pain based on four work-related factors (physical demands, workload, social support and autonomy) and five pain perceptions (PCS, FABQA, FABQW, CPAQA and CPAQP). Next, profiles were compared on their scores on predictors of sickness absence (BI and PBC) and sociodemographics (age, sex, level of education, work arrangement, multisite pain, duration of complaints and sickness absence in the previous year). We followed the three-step procedure presented by Asparouhov and Muthén.23 An overview of the procedure is presented in figure 1. In the first step, sum scores on the indicator variables were used to test and compare LPA models with one to eight profiles, which is done by the MIXTURE command in MPlus. The best model was selected based on information criteria and interpretability of its profiles. Fit statistics to determine the best possible solution were the Akaike Information Criterion (AIC), the Bayesian Information Criterion (BIC) and the sample size-adjusted BIC (SSA-BIC). Lower values indicate a better fit of the model. Elbow plots of the fit indices were used to visualise considerable changes in fit. In the case of disagreement, the BIC is considered the most reliable. The Vuong–Lo–Mendell–Rubin (VLMR) and Bootstrapped Likelihood Ratio (BLR) tests were used to compare the differences in model fit between the hypothesised number of profiles, k, and a model with k−1 profiles. If it is significant, then the model with k profiles is better. We also checked Entropy, which indicates how accurately the model defines profiles. Values on entropy should be close to 1.0 (0.8 is acceptable). Finally, profiles should contain at least 5% of all participants. In the second step, participants were assigned to profiles based on their most likely class membership using the latent class posterior distribution. In the third step, the relationship between the latent profiles and auxiliary variables was investigated. BI and PBC were included as distal outcomes and the sociodemographic factors as predictors. To test the predictors, the R3STEP command in MPlus was used. This command conducts multinominal logistic regressions assessing whether an increase in the predictors is related to a higher probability of belonging to a specific profile over another. To model the distal outcomes, the DU3STEP command in MPlus was used, or the BCH command in case of convergence problems. These commands determine whether the means of the outcome variables differ across the latent profiles. Sample size requirements for LPA depend on the complexity of the model. Simulation studies suggest that samples between 200 and 500 participants are adequate.24 25 It is recommended to handle missing data on indicator variables for the LPA using full information maximum likelihood estimation.11

    Supplemental material

    Figure 1
    Figure 1

    Three-step procedure for mixture modelling. Step 1, test and compare latent profile models of indicator variables. Step 2, assign participants to latent profiles using the latent class posterior distribution. Step 3, evaluate auxiliary variables as predictors or distal outcomes. AUT, autonomy; CPAQA, Chronic Pain Acceptance Questionnaire Activity Engagement subscale; CPAQP, Chronic Pain Acceptance Questionnaire Pain Willingness subscale; FABQA, Fear Avoidance Beliefs Questionnaire Activity subscale; FABQW, Fear Avoidance Beliefs Questionnaire Work subscale; PCS, Pain Catastrophising Scale; SOC, social support; WL, workload.

    Patient and public involvement

    None.

    Results

    Sociodemographic characteristics of the study population

    A total of 591 employees showed interest in the study. Of these, 342 were excluded because they failed to provide informed consent (n=225), never installed the app to complete the questionnaire (n=46), had missing data on all indicator variables (n=15) or did not experience pain for at least 6 weeks (n=56). The remaining 249 employees (72.3% females) were included in this study (table 1). The mean age was 41.9 years (SD=10.3 years), 74.7% had a bachelor’s degree or higher, 50.2% were employed in the healthcare sector and 62.7% were working full-time. Regarding pain complaints, 87.6% of the participants reported multisite pain and 86.3% had complaints persisting for at least 6 months. Additionally, 88.8% of employees did not take sick leave for more than a week due to their pain complaints in the previous year.

    View this table:
    Table 1

    Descriptive information of the total sample and latent profiles

    Profile selection and interpretation

    There were no missing data on indicator variables for the LPA. Information criteria for models ranging from one to eight profiles are presented in online supplemental table 3. The four-profile solution yielded the lowest BIC value, and elbow plots of the AIC and SSA-BIC also indicated a preference for the four-profile model (online supplemental figure 1). The BLRT demonstrated significant improvement in model fit for models with up to four profiles, while the three-profile solution presented the best model fit based on VLMRT. Entropy increased for models with up to four profiles and sample sizes in the four-profile solution were sufficient (smallest profile contained 13.7% of all subjects). After inspecting the standardised scores on indicator variables for each profile, it appeared that the three-profile solution classified all participants with higher physical demands into one profile. In contrast, the four-profile solution divided these participants into two separate profiles based on their scores on psychosocial work-related factors and pain perceptions. Therefore, the four-profile solution was chosen as the final model. A graphical overview of the standardised scores on indicator variables for each latent profile is shown in figure 2.

    Supplemental material

    Supplemental material

    Figure 2
    Figure 2

    Visualisation of the four latent profiles. The means of standardised indicator variables for each latent profile are depicted. Higher scores on SOC, AUT, CPAQA and CPAQP correspond to more favourable values, whereas higher scores on PD, WL, PCS, FABQA and FABQW correspond to less favourable values. AUT, autonomy; CPAQA, Chronic Pain Acceptance Questionnaire Activity Engagement subscale; CPAQP, Chronic Pain Acceptance Questionnaire Pain Willingness subscale; FABQA, Fear Avoidance Beliefs Questionnaire Activity subscale; FABQW, Fear Avoidance Beliefs Questionnaire Work subscale; PCS, Pain Catastrophising Scale; SOC, social support; WL, workload.

    Descriptive data and scale scores for each latent profile are presented in table 1. Profile 1, the largest cluster (n=95, 38.2%), demonstrated favourable scores across all indicators, reflecting a low risk regarding work environment and pain perceptions. Therefore, this profile was labelled ‘Easy goers’. Profile 2 (n=83, 33.3%) exhibited relatively high physical demands, moderate autonomy levels and favourable scores on workload, social support, pain catastrophising, fear-avoidance and pain acceptance. These scores reflected resilience and determination in managing pain at work. Therefore, this profile was labelled ‘Resilient workers’. Profile 3, the smallest cluster (n=34, 13.7%), displayed an opposite pattern compared with cluster 2, with relatively low physical demands, moderate autonomy levels and unfavourable scores on workload, social support, pain catastrophising, fear-avoidance and pain acceptance. Furthermore, profile 3 had the worst scores on social support, pain catastrophising and pain acceptance, reflecting support seeking and pain-related worries. Therefore, this profile was labelled ‘Support seekers’. Profile 4 (n=37, 14.9%) was characterised by unfavourable scores across all indicators, scoring the worst on physical demands, autonomy and FABQW. These scores reflected significant challenges regarding work and pain management, indicating high levels of strain and suffering. Therefore, this profile was labelled ‘Strain sufferers’.

    Sociodemographic predictors of profiles

    The results of the multinominal logistic regression of the sociodemographic factors on the latent profiles are presented in table 2. Profile 1 ‘Easy goers’ served as the reference category as it showed favourable scores on all indicators. Gender, duration of complaints and multisite pain were not statistically significantly associated with latent profiles. However, the odds of being classified in profile 4 ‘Strain sufferers’ decreased with increasing age (OR 0.93, 95% CI (0.89 to 0.98)) and level of education (OR 0.28, 95% CI (0.1 to 0.79)). Part-time work associated with an increased likelihood of belonging to profile 4 (OR 3.11 and 95% CI (1.39 to 6.96)). There was a marginally significant association between duration of sick leave in the past 12 months and profile group, with participants reporting more sick leave having greater odds of being classified in profile 4 (OR 2.29 and 95% CI (0.89 to 5.88)).

    View this table:
    Table 2

    Logistic regression analysis of predictors with profile 1 ‘Easy goers’ as the reference group

    Differences in distal outcomes

    Significant differences were observed in BI (χ2=8.92, p=0.030) and PBC (χ2=12.37, p=0.006) across the four latent profiles (table 3). Profile 1 ‘Easy goers’ scored significantly better than profiles 3 ‘Support seekers’ and 4 ‘Strain sufferers’ in terms of BI and profile 2 ‘Resilient workers’ scored significantly better than profile 3. With respect to PBC, profile 1 scored significantly better than the other profiles, and profile 2 scored significantly better than profile 3.

    View this table:
    Table 3

    Equality tests of the means on distal outcomes across the four latent profiles

    Discussion

    Interaction between work-related factors and pain perceptions

    Profiles characterised by unfavourable scores on several work-related factors (profiles 3 ‘Support seekers’ and 4 ‘Strain sufferers’) also presented more unfavourable scores on the pain perceptions. A previous study of 373 employees on sick leave due to back and neck pain reported that higher perceived physical demands at work and lower perceived control were associated with increased fear-avoidance beliefs.26 However, although profiles 2 ‘Resilient workers’ and 4 ‘Strain sufferers’ shared high physical demands, profile 2 was not burdened with unfavourable pain perceptions. Compared with profile 4, employees in profile 2 experienced a lower workload, more social support and more autonomy. This suggests that a single adverse work-related factor might not induce maladaptive pain perceptions if mitigating buffers exist in the work environment. The potential of psychosocial work-related factors to buffer high physical demands has been demonstrated in different settings, including the development of pain complaints and early retirement.10 27

    Interestingly, profiles characterised by poor social support (profiles 3 ‘Support seekers’ and 4 ‘Strain sufferers’) always exhibited unfavourable scores on pain perceptions. This pattern aligns with research suggesting that social support serves a protective role in employees with chronic back pain of being classified in profiles characterised by increased pain catastrophising and fear-avoidance beliefs and lower levels of pain acceptance.13 28 Qualitative research findings suggest that employees with chronic pain often perceive support at work as the availability of modified duties and help from others to reduce the risk of physical strain.29 However, the feasibility of implementing such modified duties is frequently regarded by supervisors as challenging, potentially causing imbalanced workloads among colleagues, leading to a sense of injustice.30 It is conceivable that individuals in profiles 3 ‘Support seekers’ and 4 ‘Strain sufferers’ perceived a lack of workplace measures to manage their complaints, such as modified duties, contributing to poor scores on social support and pain perceptions. Additionally, research has indicated that catastrophising can function as a coping mechanism to elicit social support.31 Employees with pain often believe that the workplace has limited understanding about pain and the impact on their lives.30 If employees with pain anticipate a lack of support, they might resort to heightened catastrophising behaviour, accentuating their complaints to increase support at work. In contrast to our findings, other studies have reported increased social support among patients who catastrophised more about pain or displayed lower pain acceptance.28 31 A plausible explanation is the fact that a substantial proportion (over 86%) of participants in our study reported complaints for at least 6 months. It has been suggested that pain catastrophising is associated with supportive responses in the case of short pain duration, but negative responses when complaints persist.31 Similarly, Grant and colleagues noted that employees with pain harboured doubts regarding ongoing support from colleagues if pain complaints and modified duties continued indefinitely.30

    Risk of sickness absence among the four profiles

    The four pain profiles identified in our study presented varying susceptibility to future sickness absence. Profiles 3 ‘Support seekers’ and 4 ‘Strain sufferers’ appeared to have an increased risk of sickness absence compared with profiles 1 ‘Easy goers’ and 2 ‘Resilient workers’. Profile 3 scored significantly worse on BI and PBC compared with profiles 1 and 2, while profile 4 scored significantly worse than profile 1 on these factors. Furthermore, employees in profile 4 reported more sick leave in the past year compared with profile 1, which has also been identified as a predictor of future absenteeism.32 However, these results should be interpreted with caution since 88.8% of our sample reported almost no sick leave in the previous year.

    Profiles 3 ‘Support seekers’ and 4 ‘Strain sufferers’ were characterised by maladaptive pain perceptions, which aligns with a previous study reporting higher odds of being of work in profiles marked by unfavourable scores on pain catastrophising and fear-avoidance beliefs.12 Maladaptive pain perceptions can exacerbate pain complaints by inducing altered movement patterns, attentional bias towards aversive stimuli and amplification of the cognitive response.33 If individuals interpret their complaints as threatening and perceive work as a significant factor contributing to their pain, avoiding work by going on sick leave seems a logical strategy to prevent exacerbation.34 Previous studies have demonstrated the role of maladaptive pain perceptions in the development of chronic and multisite pain.6 7 35

    Profiles 3 ‘Support seekers’ and 4 ‘Strain sufferers’ were also characterised by unfavourable working conditions, making the decision to go on sick leave even more logical. This aligns with previous studies identifying unfavourable working conditions as predictors of sickness absence.30 36 Furthermore, a study of 3003 workers from diverse sectors reported that combining unfavourable physical and psychosocial factors in the workplace significantly increases the risk of absenteeism due to pain compared with single risk factors.10 Interestingly, while profiles 2 ‘Resilient workers’ and 4 ‘Strain sufferers’ both scored high on physical demands, only profile 4 scored significantly worse on both BI and PBC compared with profile 1 ‘Easy goers’. Compared with profile 2, employees in profile 4 experienced a higher workload, less social support and less autonomy. A supportive work environment with decision authority might help with high work demands by fostering teamwork, dividing tasks and implementing work modifications (eg, more or longer breaks, use of ergonomic aids and job rotation).2 10 27 37 However, implementing work modifications in physically demanding jobs is often challenging.30 38 Reducing work hours could be a strategy to cope with high physical demands if employees have little influence over the organisation of their work or when work modifications are not possible, potentially explaining the lower rate of full-time work in profile 4 ‘Strain sufferers’ compared with profile 1 ‘Easy goers’.39 Other reasons for part-time work could include combining work with studying or care responsibilities such as caring for young children.39 These reasons could explain the younger age of employees in profile 4. It should be noted that employees might also be pushed towards part-time work involuntarily by employers due to economic reasons.39 The characteristics of profile 4 ‘Strain sufferers’ (eg, higher workloads, less autonomy, lower levels of education and younger age) are also typically associated with manual labour jobs, which are confronted with high rates of sickness absence.36 40 41 Of the 39 employees who reported being employed in sectors characterised by manual labour jobs (ie, industry/construction or retail/logistics/repair), only 10.3% were classified in profile 4. These results should be interpreted with caution since only 15.7% of our total sample was employed in these sectors. On the other hand, over 80% of the employees in profile 4 were employed in the healthcare sector, which is frequently confronted by unfavourable physical and psychosocial working conditions.37 42 43 Interestingly, our results showed a relatively wide distribution of healthcare workers across the four profiles. Of the 125 healthcare workers in our study, 26.4% were classified in profile 1 ‘Easy goers’, 36% in profile 2 ‘Resilient workers’, 13.6% in profile 3 ‘Support seekers’ and 24% in profile 4 ‘Strain sufferers’. These results align with previous studies reporting diversity in the occurrence of physical and psychosocial risk factors among healthcare workers, which can be attributed to factors such as work-related tasks, workplace policies and team dynamics.37 42 43

    Surprisingly, multisite pain and the duration of complaints did not show significant associations with the four profiles. Given that multisite pain and the duration of complaints have previously been linked to sickness absence, their association with profiles 3 ‘Support seekers’ and 4 ‘Strain sufferers’ would have reinforced our hypothesis.36 However, these results should be interpreted with caution, as 91.9% of our participants can be classified as having chronic pain and 87.6% reported experiencing multisite pain.44

    Study contributions

    By examining the interplay between work-related factors and pain perceptions and their contribution to sickness absence, we build on prior research and formulate actionable recommendations for the management and treatment of employees struggling with pain.10 12

    First, our findings underscore the need for tailored trajectories for employees with pain. Addressing a single work-related aspect, such as high physical demands, does not seem a cost-effective strategy for reducing sickness absence. Our results imply that not all employees facing high physical demands (profile 2 ‘Resilient workers’) are at risk of developing maladaptive pain perceptions and sickness absence because mitigating buffers exist in the work environment, while other employees without high physical demands (profile 3 ‘Support seekers’) are at risk. Furthermore, solely addressing high physical demands could be insufficient to reduce sickness absence in employees experiencing multiple risk factors (profile 4 ‘Strain sufferers’). Several reviews advocate for multidimensional interventions over singular component approaches for managing pain and mitigating sickness absence.45 46 These comprehensive interventions involve various stakeholders (eg, employees with pain, supervisors, colleagues, occupational physicians, ergonomists, physiotherapists, psychologists and general practitioners) and incorporate diverse treatment modalities targeted at both the workplace and pain management strategies of individuals with pain (eg, exercise, education, behavioural treatment and work modifications).

    Second, our study corroborates previous research highlighting that limited social support poses a significant barrier for work in employees with pain.36 47 Interpersonal conflicts within the workplace could further hinder the effectiveness of interventions aimed at reducing sickness absence.30 38 Colleagues and supervisors may question the legitimacy of work modifications for employees with pain due to lack of understanding of the problem. The absence of established procedures for workplace modifications and uncertain team coping mechanisms, particularly in conjunction with high workloads, can exacerbate interpersonal conflicts. Implementing clear disability prevention policies, fostering robust communication among stakeholders and providing education on pain-related challenges are potential strategies to enhance social support at work.48 49 A review by Cullen and colleagues underscored the importance of workplace involvement in reducing sick leave due to pain complaints, advocating for practices such as maintaining contact with sick-listed employees, educating supervisors on occupational risks and mitigating strategies and involving supervisors in formulating work modifications.50

    Our study also reinforces previous findings emphasising the critical role of evaluating pain perceptions in employees since they play an important role in sickness absence.8 9 Moreover, maladaptive pain perceptions could undermine efforts aimed at reducing sickness absence, as employees with fearful interpretations of pain and work may resist gradual return-to-work approaches.38 51 Addressing maladaptive pain perceptions is essential before effective treatment outcomes can be achieved from active approaches.52 Pain neuroscience education is gaining popularity as an approach to target these maladaptive beliefs, aiming to reduce the perceived threat of pain by educating individuals about pain processing in the nervous system and related factors.53 This approach has shown promise in reducing pain catastrophising and fear of movement.54 However, pain neuroscience education should be integrated with other modalities like physical exercise to obtain meaningful effects on other outcomes such as pain and disability.54–56 Careful coordination is crucial to avoid conflicting messages.53 For instance, a pain-contingent exercise approach (ie, terminating the exercise if pain increases) may contradict the message of pain neuroscience education that pain during exercise does not necessarily equate to tissue damage.55 56 Occupational interventions such as ergonomic advice or modified duties, including a temporarily reduction of certain activities, might inadvertently reinforce beliefs that certain activities are inherently risky.38 Therefore, alignment and adequate communication of occupational and primary care interventions is paramount. It must be noted that the pain perceptions included in this study primarily stem from the Fear-Avoidance model of pain.7 Since employees with pain might be reluctant towards modified duties to not burden colleagues, future studies should extend our model to encompass persistent behaviours.38 57 This involves supressing and ignoring pain to maintain daily activities, which has been linked to emotional distress, prolonged complaints, and increased sickness absence.57

    Finally, our findings underscore the utility of profile formulation in guiding targeted treatment prioritisation for employees most in need. Specifically, directing interventions towards employees experiencing both unfavourable working conditions and maladaptive pain perceptions (profiles 3 ‘Support seekers’ and 4 ‘Strain sufferers’) holds significant promise, given their elevated risk of sickness absence. While directly assigning new employees to statistically derived clusters may present challenges, clinical experts can leverage their expertise to interpret scores across different risk factors. Alternatively, exploring advanced techniques such as decision tree analysis or deep learning algorithms could streamline this process.58

    Limitations

    Some limitations of our study must be noted.

    First, sickness absence was not measured directly and, since this was a cross-sectional study, we could only establish associations. However, robust evidence supports the predictive value of expectations, self-efficacy and previous sick leave in predicting future absenteeism.8 20 32 Future longitudinal studies are essential to establish causal relationships between these profiles and sick leave outcomes.

    Next, the present study focused on active employees. Exploring pain profiles of work factors and pain perceptions among employees on sick leave would provide additional insights on the mechanisms underlying sickness absence due to pain. Based on our findings, we hypothesise that many employees on sick leave would be classified into profiles 3 ‘Support seekers’ and 4 ‘Strain sufferers’. However, new profiles with distinct characteristics may emerge among this population. Additionally, our sample exhibited an over-representation of workers from the healthcare sector, limiting the generalisability of our findings to other sectors and the broader Belgian working population.59 Future studies should evaluate the relevance of our profiles across diverse sectors. Moreover, the current study focused solely on work-related factors and pain perceptions to differentiate profiles of employees with pain. Subsequent research could expand these profiles by incorporating lifestyle factors and mental well-being, which have also been linked to disability and sickness absence in individuals with pain.3 36 60

    Finally, studies relying on self-report data may be vulnerable to several sources of bias.61 Social desirability bias can arise when participants are questioned about sensitive topics or when confidentiality during data collection is compromised.61 However, the scales utilised in this study have been validated in epidemiological research, and participants completed the questionnaire on their smartphones without a researcher present.14–20 Recall bias is another concern in self-report studies, where participants may inaccurately recall past events.61 One strategy to overcome recall bias is to minimise the recall period. Besides several demographic characteristics such as the duration of pain complaints, participants were questioned about their current pain perceptions and work context. Affectivity could also influence self-report responses.62 Pain catastrophising, for instance, has been associated with negative affect and neuroticism, which is characterised by a limited tolerance for and attentional bias towards aversive stimuli.31 63 It is plausible that individuals in profiles 3 ‘Support seekers’ and 4 ‘Strain sufferers’ rated their work conditions more negatively due to inherent tendencies towards catastrophising and fear-avoidance. Petersen and colleagues observed a significant positive correlation between fear-avoidance beliefs and perceived physical demands among employees with back pain64 However, they found no correlation between perceived physical demands and a more accurate estimation of physical workload, involving assessment of physical exposures by occupational medicine experts. Future studies could consider repeated measurement designs and inclusion of affective state scales to further mitigate the influence of affectivity.65

    Conclusion

    This study highlights the significance of considering the interplay between various work-related factors and pain perceptions in employees with pain and, therefore, provides valuable insights to improve employee well-being and reduce sickness absence due to pain. By using a latent variable approach, we identified four distinct pain profiles among employees. Importantly, our findings suggest that unfavourable scores on a single work-related factor might not inherently translate into maladaptive pain perceptions or subsequent sickness absence. This nuanced interplay between factors highlights the potential limitations of single-domain interventions. The formulation of profiles enables targeted treatment strategies aimed at employees most vulnerable to sickness absence. It is crucial to prioritise interventions for individuals experiencing unfavourable working conditions alongside maladaptive pain perceptions. In this context, social support emerges as an important factor influencing sickness absence. Future research should focus on developing practical methods to categorise workers into risk profiles, enabling early intervention and prevention strategies. Extending our model to include additional work-related, pain-related and lifestyle factors will provide a more comprehensive understanding of sickness absence due to pain.

    Data availability statement

    Data are available upon reasonable request. The data that support the findings of this study are available from the corresponding author upon reasonable request.

    Ethics statements

    Patient consent for publication

    Ethics approval

    This study involves human participants. This study was approved by the Social Ethics Commission of KU Leuven (G-2019081713) and adhered to Belgian and international privacy and ethical regulations. Participants gave informed consent to participate in the study before taking part.

    References

    1. ↵
      1. EU-OSHA
      . Work-related musculoskeletal disorders: prevalence, costs and demographics in the EU. European Agency for Safety and Health at Work. 2019. Available: https://osha.europa.eu/en/publications/work-related-musculoskeletal-disorders-prevalence-costs-and-demographics-eu/view
    2. ↵
      2. Keyaerts S ,
      3. Godderis L ,
      4. Delvaux E , et al
      . The association between work-related physical and psychosocial factors and musculoskeletal disorders in healthcare workers: Moderating role of fear of movement. J Occup Health 2022;64:e12314. doi:10.1002/1348-9585.12314
    3. ↵
      2. Nieminen LK ,
      3. Pyysalo LM ,
      4. Kankaanpää MJ
      . Prognostic factors for pain chronicity in low back pain: a systematic review. Pain Rep 2021;6:e919. doi:10.1097/PR9.0000000000000919
    4. ↵
      2. Buruck G ,
      3. Tomaschek A ,
      4. Wendsche J , et al
      . Psychosocial areas of worklife and chronic low back pain: a systematic review and meta-analysis. BMC Musculoskelet Disord 2019;20:480. doi:10.1186/s12891-019-2826-3
    5. ↵
      2. Beemsterboer W ,
      3. Stewart R ,
      4. Groothoff J , et al
      . A literature review on sick leave determinants (1984-2004). Int J Occup Med Environ Health 2009;22:169–79. doi:10.2478/v10001-009-0013-8
      OpenUrlCrossRefPubMed
    6. ↵
      2. Edwards RR ,
      3. Dworkin RH ,
      4. Sullivan MD , et al
      . The Role of Psychosocial Processes in the Development and Maintenance of Chronic Pain. J Pain 2016;17:T70–92. doi:10.1016/j.jpain.2016.01.001
      OpenUrlCrossRefPubMed
    7. ↵
      2. Vlaeyen JWS ,
      3. Linton SJ
      . Fear-avoidance and its consequences in chronic musculoskeletal pain: a state of the art. Pain 2000;85:317–32. doi:10.1016/S0304-3959(99)00242-0
      OpenUrlCrossRefPubMedWeb of Science
    8. ↵
      2. Carriere JS ,
      3. Thibault P ,
      4. Milioto M , et al
      . Expectancies Mediate the Relations Among Pain Catastrophizing, Fear of Movement, and Return to Work Outcomes After Whiplash Injury. J Pain 2015;16:1280–7. doi:10.1016/j.jpain.2015.09.001
      OpenUrl
    9. ↵
      2. Macías-Toronjo I ,
      3. Rojas-Ocaña MJ ,
      4. Sánchez-Ramos JL , et al
      . Pain catastrophizing, kinesiophobia and fear-avoidance in non-specific work-related low-back pain as predictors of sickness absence. PLoS One 2020;15:e0242994. doi:10.1371/journal.pone.0242994
    10. ↵
      2. Widanarko B ,
      3. Legg S ,
      4. Devereux J , et al
      . The combined effect of physical, psychosocial/organisational and/or environmental risk factors on the presence of work-related musculoskeletal symptoms and its consequences. Appl Ergon 2014;45:1610–21. doi:10.1016/j.apergo.2014.05.018
      OpenUrlCrossRefPubMed
    11. ↵
      2. Spurk D ,
      3. Hirschi A ,
      4. Wang M , et al
      . Latent profile analysis: A review and “how to” guide of its application within vocational behavior research. J Vocat Behav 2020;120:103445. doi:10.1016/j.jvb.2020.103445
      OpenUrl
    12. ↵
      2. Carlesso LC ,
      3. Raja Rampersaud Y ,
      4. Davis AM
      . Clinical classes of injured workers with chronic low back pain: a latent class analysis with relationship to working status. Eur Spine J 2018;27:117–24. doi:10.1007/s00586-017-4966-1
      OpenUrl
    13. ↵
      2. Carlesso LC ,
      3. Tousignant-Laflamme Y ,
      4. Shaw W , et al
      . Exploring pain phenotypes in workers with chronic low back pain: Application of IMMPACT recommendations. Can J Pain 2021;5:43–55. doi:10.1080/24740527.2020.1870103
      OpenUrl
    14. ↵
      2. Hildebrandt VH ,
      3. Bongers PM ,
      4. van Dijk FJ , et al
      . Dutch Musculoskeletal Questionnaire: description and basic qualities. Ergonomics 2001;44:1038–55. doi:10.1080/00140130110087437
      OpenUrlCrossRefPubMedWeb of Science
    15. ↵
      2. Notelaers G ,
      3. De Witte H ,
      4. van Veldhoven M , et al
      . Construction and validation of the short inventory to monitor psychosocial hazards. Arbeidsgezondheidzorg En Ergonomie 2007;44:11–7. Available: https://research.tilbu rguni versi ty.edu/en/publicatio ns/const ructi on-and-validation -of-the-short-inventory-to-monitor-psy
      OpenUrl
    16. ↵
      2. Veldhoven M ,
      3. Prins J ,
      4. van der Laken P , et al
      . QEEW2.0: 42 short scales for survey research on work, well-being and performance. Amsterdam: SKB, 2015:95.
    17. ↵
      2. Sullivan MJL ,
      3. Bishop SR ,
      4. Pivik J
      . The pain catastrophizing scale: development and validation. Psychol Assess 1995;7:524–32. doi:10.1037//1040-3590.7.4.524
      OpenUrlCrossRefWeb of Science
    18. ↵
      2. Waddell G ,
      3. Newton M ,
      4. Henderson I , et al
      . A Fear-Avoidance Beliefs Questionnaire (FABQ) and the role of fear-avoidance beliefs in chronic low back pain and disability. Pain 1993;52:157–68. doi:10.1016/0304-3959(93)90127-B
      OpenUrlCrossRefPubMedWeb of Science
    19. ↵
      2. Fish RA ,
      3. McGuire B ,
      4. Hogan M , et al
      . Validation of the chronic pain acceptance questionnaire (CPAQ) in an Internet sample and development and preliminary validation of the CPAQ-8. Pain 2010;149:435–43. doi:10.1016/j.pain.2009.12.016
      OpenUrlCrossRefPubMed
    20. ↵
      2. Dunstan DA ,
      3. Covic T ,
      4. Tyson GA
      . What leads to the expectation to return to work? Insights from a Theory of Planned Behavior (TPB) model of future work outcomes. Work 2013;46:25–37. doi:10.3233/WOR-2012-1481
      OpenUrl
    21. ↵
      2. Ajzen I
      . The theory of planned behavior. Org Behav Hum Decis Process 1991;50:179–211. doi:10.1016/0749-5978(91)90020-T
      OpenUrlCrossRef
    22. ↵
      2. Weston R ,
      3. Gore PA
      . A brief guide to structural equation modeling. Couns Psychol 2006;34:719–51. doi:10.1177/0011000006286345
      OpenUrlCrossRefWeb of Science
    23. ↵
      2. Asparouhov T ,
      3. Muthén BO
      . Auxiliary variables in mixture modeling: three-step approaches using Mplus. Struct Equ Modeling 2014;21:329–41. doi:10.1080/10705511.2014.915181
      OpenUrlCrossRef
    24. ↵
      2. Nylund KL ,
      3. Asparouhov T ,
      4. Muthén BO
      . Deciding on the number of classes in latent class analysis and growth mixture modeling: a Monte Carlo simulation study. Struct Equ Modeling 2007;14:535–69. doi:10.1080/10705510701575396
      OpenUrlCrossRefWeb of Science
    25. ↵
      2. Swanson SA ,
      3. Lindenberg K ,
      4. Bauer S , et al
      . A Monte Carlo investigation of factors influencing latent class analysis: an application to eating disorder research. Int J Eat Disord 2012;45:677–84. doi:10.1002/eat.20958
      OpenUrlPubMed
    26. ↵
      2. Myhre K ,
      3. Røe C ,
      4. Marchand GH , et al
      . Fear-avoidance beliefs associated with perceived psychological and social factors at work among patients with neck and back pain: a cross-sectional multicentre study. BMC Musculoskelet Disord 2013;14:329. doi:10.1186/1471-2474-14-329
    27. ↵
      2. Stengård J ,
      3. Leineweber C ,
      4. Wang H-X
      . Prolonging working life among blue-collar workers: The buffering effect of psychosocial job resources on the association between physically demanding and hazardous work and retirement timing. SSM Popul Health 2023;22:101372. doi:10.1016/j.ssmph.2023.101372
    28. ↵
      2. Rovner G ,
      3. Vowles KE ,
      4. Gerdle B , et al
      . Latent Class Analysis of the Short and Long Forms of the Chronic Pain Acceptance Questionnaire: Further Examination of Patient Subgroups. J Pain 2015;16:1095–105. doi:10.1016/j.jpain.2015.07.007
      OpenUrl
    29. ↵
      2. Palstam A ,
      3. Mannerkorpi K
      . Work Ability in Fibromyalgia: An Update in the 21st Century. Curr Rheumatol Rev 2017;13:180–7. doi:10.2174/1573397113666170502152955
      OpenUrlPubMed
    30. ↵
      2. Grant M ,
      3. O-Beirne-Elliman J ,
      4. Froud R , et al
      . The work of return to work. Challenges of returning to work when you have chronic pain: a meta-ethnography. BMJ Open 2019;9:e025743. doi:10.1136/bmjopen-2018-025743
    31. ↵
      2. Quartana PJ ,
      3. Campbell CM ,
      4. Edwards RR
      . Pain catastrophizing: a critical review. Expert Rev Neurother 2009;9:745–58. doi:10.1586/ern.09.34
      OpenUrlCrossRefPubMed
    32. ↵
      2. Rysstad T ,
      3. Grotle M ,
      4. Aasdahl L , et al
      . Identification and Characterisation of Trajectories of Sickness Absence Due to Musculoskeletal Pain: A 1-Year Population-based Study. J Occup Rehabil 2023;33:277–87. doi:10.1007/s10926-022-10070-7
      OpenUrl
    33. ↵
      2. Garland EL
      . Pain processing in the human nervous system: a selective review of nociceptive and biobehavioral pathways. Prim Care 2012;39:561–71. doi:10.1016/j.pop.2012.06.013
      OpenUrlCrossRefPubMed
    34. ↵
      2. Bunzli S ,
      3. Smith A ,
      4. Schütze R , et al
      . Making Sense of Low Back Pain and Pain-Related Fear. J Orthop Sports Phys Ther 2017;47:628–36. doi:10.2519/jospt.2017.7434
      OpenUrlPubMed
    35. ↵
      2. Niederstrasser NG ,
      3. Slepian PM ,
      4. Mankovsky-Arnold T , et al
      . An experimental approach to examining psychological contributions to multisite musculoskeletal pain. J Pain 2014;15:1156–65. doi:10.1016/j.jpain.2014.08.007
      OpenUrlCrossRefPubMed
    36. ↵
      2. Hallman DM ,
      3. Holtermann A ,
      4. Björklund M , et al
      . Sick leave due to musculoskeletal pain: determinants of distinct trajectories over 1year. Int Arch Occup Environ Health 2019;92:1099–108. doi:10.1007/s00420-019-01447-y
      OpenUrlCrossRefPubMed
    37. ↵
      2. Andersen LL ,
      3. Vinstrup J ,
      4. Villadsen E , et al
      . Physical and Psychosocial Work Environmental Risk Factors for Back Injury among Healthcare Workers: Prospective Cohort Study. Int J Environ Res Public Health 2019;16:4528. doi:10.3390/ijerph16224528
    38. ↵
      2. Coole C ,
      3. Watson PJ ,
      4. Drummond A
      . Low back pain patients’ experiences of work modifications; a qualitative study. BMC Musculoskelet Disord 2010;11:277. doi:10.1186/1471-2474-11-277
    39. ↵
      2. Mäkinen N ,
      3. Tanskanen J ,
      4. Ojala S , et al
      . Part-time workers’ employment trajectories by length of hours and reason for working part-time: an 8-year follow-up study. Sage Open 2023;13. doi:10.1177/21582440231210690
    40. ↵
      2. Andersen LL ,
      3. Pedersen J ,
      4. Sundstrup E , et al
      . High physical work demands have worse consequences for older workers: prospective study of long-term sickness absence among 69 117 employees. Occup Environ Med 2021;78:829–34. doi:10.1136/oemed-2020-107281
      OpenUrlAbstract/FREE Full Text
    41. ↵
      2. Januario LB ,
      3. Batistao MV ,
      4. Coury HJ , et al
      . Psychosocial Risk Factors and Musculoskeletal Symptoms among White and Blue-collar Workers at Private and Public Sectors. Ann Occup Environ Med 2014;26:20. doi:10.1186/s40557-014-0020-5
      OpenUrl
    42. ↵
      2. Soylar P ,
      3. Ozer A
      . Evaluation of the prevalence of musculoskeletal disorders in nurses: a systematic review. Med-Sci 2018;7:1. doi:10.5455/medscience.2017.06.8747
      OpenUrl
    43. ↵
      2. Heijden B der ,
      3. Estryn-Béhar M ,
      4. Heerkens H
      . Prevalence of, and risk factors for, physical disability among nurses in Europe. JSS 2019;07:147–73. doi:10.4236/jss.2019.711012
      OpenUrl
    44. ↵
      2. Oliveira CB ,
      3. Maher CG ,
      4. Pinto RZ , et al
      . Clinical practice guidelines for the management of non-specific low back pain in primary care: an updated overview. Eur Spine J 2018;27:2791–803. doi:10.1007/s00586-018-5673-2
      OpenUrlCrossRefPubMed
    45. ↵
      2. Stock SR ,
      3. Nicolakakis N ,
      4. Vézina N , et al
      . Are work organization interventions effective in preventing or reducing work-related musculoskeletaldisorders? A systematic review of the literature. Scand J Work Environ Health 2018;44:113–33. doi:10.5271/sjweh.3696
      OpenUrl
    46. ↵
      2. Wegrzynek PA ,
      3. Wainwright E ,
      4. Ravalier J
      . Return to work interventions for chronic pain: a systematic review. Occup Med (Lond) 2020;70:268–77. doi:10.1093/occmed/kqaa066
      OpenUrl
    47. ↵
      2. Etuknwa A ,
      3. Daniels K ,
      4. Eib C
      . Sustainable Return to Work: A Systematic Review Focusing on Personal and Social Factors. J Occup Rehabil 2019;29:679–700. doi:10.1007/s10926-019-09832-7
      OpenUrlPubMed
    48. ↵
      2. Fassier J-B ,
      3. Durand M-J ,
      4. Caillard J-F , et al
      . Results of a feasibility study: barriers and facilitators in implementing the Sherbrooke model in France. Scand J Work Environ Health 2015;41:223–33. doi:10.5271/sjweh.3489
      OpenUrlPubMed
    49. ↵
      2. Main CJ ,
      3. Shaw WS ,
      4. Nicholas MK , et al
      . System-level efforts to address pain-related workplace challenges. PAIN 2022;163:1425–31. doi:10.1097/j.pain.0000000000002548
      OpenUrl
    50. ↵
      2. Cullen KL ,
      3. Irvin E ,
      4. Collie A , et al
      . Effectiveness of Workplace Interventions in Return-to-Work for Musculoskeletal, Pain-Related and Mental Health Conditions: An Update of the Evidence and Messages for Practitioners. J Occup Rehabil 2018;28:1–15. doi:10.1007/s10926-016-9690-x
      OpenUrlCrossRefPubMed
    51. ↵
      2. Wertli MM ,
      3. Rasmussen-Barr E ,
      4. Held U , et al
      . Fear-avoidance beliefs-a moderator of treatment efficacy in patients with low back pain: a systematic review. Spine J 2014;14:2658–78. doi:10.1016/j.spinee.2014.02.033
      OpenUrlCrossRefPubMed
    52. ↵
      2. Meeus M ,
      3. Nijs J ,
      4. van Wilgen P , et al
      . Moving on to movement in patients with chronic joint pain. Pain Clin Updates 2016;24:1–8.
      OpenUrl
    53. ↵
      2. Nijs J ,
      3. Paul van Wilgen C ,
      4. Van Oosterwijck J , et al
      . How to explain central sensitization to patients with “unexplained” chronic musculoskeletal pain: practice guidelines. Man Ther 2011;16:413–8. doi:10.1016/j.math.2011.04.005
      OpenUrlCrossRefPubMed
    54. ↵
      2. Watson JA ,
      3. Ryan CG ,
      4. Cooper L , et al
      . Pain Neuroscience Education for Adults With Chronic Musculoskeletal Pain: A Mixed-Methods Systematic Review and Meta-Analysis. J Pain 2019;20:1140. doi:10.1016/j.jpain.2019.02.011
      OpenUrl
    55. ↵
      2. Blickenstaff C ,
      3. Pearson N
      . Reconciling movement and exercise with pain neuroscience education: A case for consistent education. Physiother Theory Pract 2016;32:396–407. doi:10.1080/09593985.2016.1194653
      OpenUrl
    56. ↵
      2. Ryan CG ,
      3. Gray HG ,
      4. Newton M , et al
      . Pain biology education and exercise classes compared to pain biology education alone for individuals with chronic low back pain: a pilot randomised controlled trial. Man Ther 2010;15:382–7. doi:10.1016/j.math.2010.03.003
      OpenUrlCrossRefPubMed
    57. ↵
      2. Hasenbring MI ,
      3. Hallner D ,
      4. Klasen B , et al
      . Pain-related avoidance versus endurance in primary care patients with subacute back pain: psychological characteristics and outcome at a 6-month follow-up. Pain 2012;153:211–7. doi:10.1016/j.pain.2011.10.019
      OpenUrlCrossRefPubMedWeb of Science
    58. ↵
      2. Awaysheh A ,
      3. Wilcke J ,
      4. Elvinger F , et al
      . Review of Medical Decision Support and Machine-Learning Methods. Vet Pathol 2019;56:512–25. doi:10.1177/0300985819829524
      OpenUrlCrossRef
    59. ↵
      1. Vlaamse Overheid
      . Tewerkstelling per sector. 2024. Available: https://www.vlaanderen.be/statistiek-vlaanderen/arbeid/tewerkstelling-per-sector
    60. ↵
      2. Alhowimel AS ,
      3. Alotaibi MA ,
      4. Alenazi AM , et al
      . Psychosocial Predictors of Pain and Disability Outcomes in People with Chronic Low Back Pain Treated Conservatively by Guideline-Based Intervention: A Systematic Review. J Multidiscip Healthc 2021;14:3549–59. doi:10.2147/JMDH.S343494
      OpenUrl
    61. ↵
      2. Althubaiti A
      . Information bias in health research: definition, pitfalls, and adjustment methods. J Multidiscip Healthc 2016;9:211–7. doi:10.2147/JMDH.S104807
      OpenUrlCrossRefPubMed
    62. ↵
      2. Podsakoff PM ,
      3. MacKenzie SB ,
      4. Lee JY , et al
      . Common method biases in behavioral research: a critical review of the literature and recommended remedies. J Appl Psychol 2003;88:879–903. doi:10.1037/0021-9010.88.5.879
      OpenUrlCrossRefPubMedWeb of Science
    63. ↵
      2. Kadimpati S ,
      3. Zale EL ,
      4. Hooten MW , et al
      . Associations between Neuroticism and Depression in Relation to Catastrophizing and Pain-Related Anxiety in Chronic Pain Patients. PLoS One 2015;10:e0126351. doi:10.1371/journal.pone.0126351
    64. ↵
      2. Petersen J ,
      3. Kirkeskov L ,
      4. Hansen BB , et al
      . Physical demand at work and sick leave due to low back pain: a cross-sectional study. BMJ Open 2019;9:e026917. doi:10.1136/bmjopen-2018-026917
    65. ↵
      2. Brose A ,
      3. Lindenberger U ,
      4. Schmiedek F
      . Affective states contribute to trait reports of affective well-being. Emotion 2013;13:940–8. doi:10.1037/a0032401
      OpenUrlCrossRefPubMed

    Supplementary materials

    Footnotes

    • Contributors SK: concept and design of the study, data collection, statistical analyses, data interpretation and drafting the manuscript. LD: concept and design of the study, data interpretation, critical review of the manuscript. LG and VVA: data interpretation and critical review of the manuscript. SK will be responsible for the overall content as a guarantor. All authors have approved the final version of the manuscript to be published.

    • Funding This cross-sectional study is a part of a broader research project, entitled 'Personal Health Empowerment', which aimed to develop and evaluate innovative monitoring and coaching solutions for employees with pain complaints. This project is funded by ITEA (grant number 16040), an European financier of innovative projects that have social added value and make use of technological innovations, and the Flemish Agency for Innovation and Entrepreneurship (grant number HBC.2018.2012), who provides grants to companies for the development or strengthening of their research and development activities.

    • Competing interests None declared.

    • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting or dissemination plans of this research.

    • Provenance and peer review Not commissioned; externally peer reviewed.

    • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.