A posture and load sampling approach to determining low-back pain risk in occupational settings

https://doi.org/10.1016/S0169-8141(00)00038-XGet rights and content

Abstract

A posture and load sampling approach to measure physical exposures was implemented within a case-control study of low-back pain reporting. The purpose of this paper was to determine how well this method was able to identify known low-back pain risk factors. Subjects, including both cyclic production and non-cyclic support workers, were studied while working in an automotive assembly facility. The study included 104 (with 20 proxies) cases, workers who reported low-back pain at work, and 129 randomly selected controls. Results indicate significant associations between low-back pain reporting and peak spinal loads (OR=2.0 for compression), shift-average spinal loading (OR=1.7 for compression), percent of time with loads in the hand (OR=1.5), maximum flexion angle (OR=2.2), and percent of time spent forward flexed beyond 45°(OR=1.3). Posture and hand load variables, considered to be intermediate exposure variables, were handled separately in multivariable regression analyses from variables of peak and average spine force which directly estimate tissue loading. The work and posture sampling approach is particularly useful for heterogeneous work situations where traditional task analysis is difficult and can provide information on work and tissue load parameters which have been directly associated with risk of reporting low-back pain.

Relevance to industry

This paper demonstrates the effectiveness of an observational method in quantifying workplace exposures to physical risk factors for low-back pain. The method works for both cyclic and non-cyclic work. Quantified risk assessment provides key information for decision makers trying to control injury rates in industrial systems.

Introduction

A wide variety of variables have been studied in an effort to understand the risk factors associated with low-back pain (Garg, 1989; Bongers et al., 1993; Hagberg et al., 1995). A recent review by the National Institute for Occupational Safety and Health (NIOSH) (Bernard, 1997) has acknowledged that there is evidence for awkward postures (pp. 6–26), there is evidence for heavy physical work (pp. 6–12), and there is strong evidence for lifting and forceful movements (pp. 6–20) as risk factors for low-back pain (LBP). Recently, Norman et al. (1998) have shown that, among the physical loading factors considered, variables tended to cluster in four independent categories: peak spinal loads, accumulated spinal loads, forces in the hands, and trunk kinematic (postural) variables. Variables contributed independently to the risk estimates between categories while within each category variables were highly inter-correlated and thus found to be mutually exclusive in multivariable regression analyses. While the risk factors identified by Norman et al. (1998) are more clearly defined and precisely measurable than the category of “heavy physical work” used by necessity in the NIOSH review, both approaches are consistent with the underlying hypothesis that an injury occurs when the body's tissues are subjected to more load than they can withstand. Since tissue tolerance cannot be measured in vivo (Van Tulder et al., 1997), injury prevention efforts must rely on the ability to measure workplace exposure to physical loading to assess possible risk. This raises the question: how can we effectively measure physical exposures in the workplace?

The purpose of this paper is to determine how well an observational work and posture sampling technique was able to identify known low-back pain risk factors. Data presented in this paper come from part of an epidemiological study of low-back pain reporting. The Ontario Universities Back Pain Study (OUBPS) was a case-control study, employing an incidence density sampling strategy, run over two years in a large automotive assembly facility. The study included detailed measurements of biomechanical, psychophysical, and psychosocial variables and has shown all three of these to be strongly and independently associated with risk of reporting low-back pain at work (Kerr et al., in press), a finding that has been separately supported by other researchers (Wickström and Pentti, 1998; Smedley et al., 1995). The biomechanical measurement battery included self-report questionnaires, detailed observer checklists, digital video analysis, detailed biomechanical modelling, electromyography, and a posture and load sampling technique. The test battery was designed to facilitate inter-method comparisons by measuring known risk factors in consistent units of measurement (Wells et al., 1997; Neumann et al., 1999). The assessment of the performance of each of the methods used in this study is a necessary step for evaluating the relative performance of each tool's ability to assess workplace exposure. The posture and load sampling method, which was developed to quantify the postures, hand loads, and spinal loading during both cyclic and non-cyclic work (Wells et al., 1995), will be examined within the context of this larger epidemiological study. Specifically, this posture and load sampling assessment method will be examined for its ability to identify risk associated with known low-back pain risk factors.

Section snippets

Risk-relationship study

The study was run in a large automobile assembly facility with a study base of over 10,000 hourly paid workers. Incident cases were identified as they reported to the plant nursing station with low-back pain. Cases were not required to have any lost time due to their LBP. Controls were selected randomly from the hourly paid employee roster. Both cases and controls were screened to have had no LBP reports in the previous 90 days. When a case was not available for a physical loading assessment, a

Results

Table 1 summarizes the results of the Student's “t” test. Peak spinal loading estimates, shift-average spinal loading estimates, trunk kinematics, and hand load variables showed significant differences between cases and controls. Exposure variables which did not show significant differences between the groups studied included median spinal load, low-level compression (as indicated by the APDF's 10 percentile), trunk flexor moments, postures near neutral, and the percent time spent twisted. Some

Discussion

The work sampling technique has confirmed, in bivariable analysis, the importance of peak spine load, cumulative spine load (as represented by the shift-long average), hand forces, and posture as risk factors for low-back pain. Peak spine load as measured in the compressive, extensor moment, and both posterior and anterior shear modes showed significant and substantial odds ratios. These are, at the group level, similar in amplitude to those reported in Norman et al. (1998) using data from

Conclusions

This paper has demonstrated the ability of an observational posture and load sampling method, with biomechanical post-processing, to quantify physical exposure in the workplace. The method has identified risk factors for reporting low-back pain of peak spinal loading, accumulated spinal loading, hand loads, and trunk postural factors. The technique used here can be readily applied to non-cyclic jobs which are difficult to analyse with task-based assessment methods. The results of the

Acknowledgements

This work was funded by the Institute for Work & Health whose core funding is provided by the Workplace Safety & Insurance Board of Ontario, Canada. The authors would like to acknowledge all of the members of the Ontario Universities Back Pain Study (OUBPS) working group: Beaton D.E., Bombardier C., Ferrier S., Hogg-Johnson S., Mondloch M., Peloso P., Smith J., Stansfeld S.A., Tarasuk V., Dobbyn M., Edmondstone M.A., Ingelman J.P., Jeans B., McRobbie H., Moore A., Mylett J., Outerbridge G., Woo

References (27)

  • Bernard, M.D. (Ed.), 1997. Musculoskeletal disorders and workplace factors: a critical review of epidemiologic evidence...
  • P.M. Bongers et al.

    Psychosocial factors and musculo-skeletal disease

    Scandinavian Journal of Work and Environmental Health

    (1993)
  • Canadian Standardized Test of Fitness Operations Manual, 3rd Edition. Fitness and Amature Sport Canada, Ottawa,...
  • Edmondstone, M.A., Neumann, P., Wells, R., Norman, R., 1996. Inter-observer reliability in working sampling: a...
  • Fransson-Hall, C., Gloria, R., Kilbom, A., Winkel, J., Karlqvist, L. Wiktorin, C., The Stockholm-Music Study Group,...
  • Garg, A., 1989. Epidemiological basis for manual lifting guidelines. Report # 937384. US Department of Health and Human...
  • M. Hagberg

    Exposure variables in ergonomic epidemiology

    American Journal of Industrial Medicine

    (1992)
  • Hagberg, M., Silverstein, B., Wells, R., Smith, R., Carayon, Hendrick, H.P., Perusse, M. Kuorinka, I., Forcier, L....
  • B. Jonsson

    Measurement and evaluation of local muscular strain in the shoulder during constrained work

    Journal of Human Ergology

    (1982)
  • Kerr, M.S., 1998. A case-control study of biomechanical and psychosocial risk factors for low-back pain reported in an...
  • Kerr, M.S., Frank, J.W., Shannon, H.S., Norman, R.W.K., Wells, R.P., Neumann, W.P., Bombardier, C., and the Ontario...
  • S. Kumar

    Cumulative load as a risk factor for back pain

    Spine

    (1990)
  • W.S. Marras et al.

    The role of dynamic three-dimensional trunk motion in occupational-related low-back disordersthe effects of workplace factors trunk position and trunk motion characteristics on risk of injury

    Spine

    (1993)
  • Cited by (49)

    • Machine learning-based identification and classification of physical fatigue levels: A novel method based on a wearable insole device

      2023, International Journal of Industrial Ergonomics
      Citation Excerpt :

      Another way to quantify workers' physical fatigue is to use biomechanical models, which examine human motions and determine joint loadings (Seo et al., 2016). Extant research studies have indicated that electromyography-assisted biomechanical models and remote-sensing techniques may be used to evaluate workers’ movements and internal joint loadings (Seo et al., 2016; Sparto and Parnianpour, 1999; Neumann et al., 2001). A study by Seo et al. (2016) presents a simulation-based framework based on biomechanical analyses to estimate physical demands and fatigue during construction operations.

    • An automatic and non-invasive physical fatigue assessment method for construction workers

      2019, Automation in Construction
      Citation Excerpt :

      However, as analyzed in the following part, biomechanical calculation methods can only be used in virtual or indoor environments, limiting the application on construction sites [32,33]. Manual observation has been widely used to collect motion data in previous studies [34,35]. However, manual observation may have the following challenges if applied on construction fields.

    • An integrated cost and worker fatigue evaluation model of a packaging process

      2019, International Journal of Production Economics
      Citation Excerpt :

      The biomechanical analysis is completed according to the entered data by estimating peak loads on the lumbar spine and other body joints such as shoulder or forearm. The software also calculates the cumulative loads on the lumbar spine and shoulders and provides a summary of the moment of force for different body joints according to the input data (Norman et al., 1998; Neumann et al., 2001). We use the biomechanical results of 4D Watbak (see Section 5.1) to estimate the expected fatigue of a worker when he/she completes the five steps of the packaging process included in one job cycle (see Fig. 1).

    • Fuzzy inference model for assessing occupational risks in construction sites

      2016, International Journal of Industrial Ergonomics
    • A comparison of low back kinetic estimates obtained through posture matching, rigid link modeling and an EMG-assisted model

      2011, Applied Ergonomics
      Citation Excerpt :

      Additionally, video can be preserved indefinitely and analyzed at the researcher’s or ergonomists’s convenience. These advantages have been reflected in the use of video in both industrial and laboratory settings (Callaghan et al., 2001; Neumann et al., 2001; Daynard et al., 2001; Jager et al., 2000; Norman et al., 1998). A recent survey of the existing literature has found video based posture analysis to be the most commonly employed tool in assessments of cumulative loading (Waters et al., 2006).

    View all citing articles on Scopus
    View full text