Physical working conditions as covered in European monitoring questionnaires

Physical workloads have been found to be related to both musculoskeletal complaints [7, 8] and labour market participation [9‐11].

They can cause specific serious conditions (diseases) [12‐19]. Some of these conditions are listed as occupational diseases in Europe [20]. Regarding labour market participation, physical workloads have been found to predict sickness absence [9, 21‐31] and disability pensioning [32‐41]. The physical exposures related to sickness absence are to some extent the same as those related to disability pensioning as well [32‐41]. One can hence summarize that the plethora of results on the association of workers’ physical workloads and their health and labour market participation underlines why it is important to measure them.

In spite of attention among both the scientific and the political community, the prevalence of workers with demanding physical workloads in the European work force remains high. In 2015, the fraction of workers in the EU reporting – in at least ¼ of working hours – tiring or painful positions was 43%, carrying heavy loads 32% and vibrations 20% (The European Foundation for the Improvement of Living and Working Conditions (Eurofound) [42]). Similarly, high levels of exposure have been reported based on national monitoring data [43‐46].

Altogether, this shows that 1) there is scientific evidence on the association of physical work exposures and health effects, and 2) a high number of workers is exposed to these conditions. In spite of this, a comparison of quality and quantity of exposures across EU countries is difficult because currently, no consensus exists regarding a standardized approach to be used in surveys to measure physical work exposure. The European Working Conditions Survey (EWCS), carried out in 35 European nations and covering 44,000 workers in 2015, and containing five physical working condition items, is at present our best starting point for comparisons [42].

In this paper, our aim is to establish an overview of physical workloads as measured in European monitoring instruments. Based on this overview, we will examine 1) which domains and dimensions of the physical workloads are monitored in surveys at the national level and the EU level and 2) the degree of European consensus among these surveys regarding coverage of individual domains and dimensions.

In 2009, six member countries of the “Partnership for European Research in Occupational Safety and Health (PEROSH)”, a joint collaboration of European institutes on research and development in occupational safety and health, formed a project group and in 2014 published an article on the comparability of psychosocial working conditions measured in European monitoring questionnaires [3]. Inspired by this work, the group decided to continue its cooperation in 2015, now with a focus on physical workloads measured in the national surveys. The items from the following surveys were included in the comparative review: National Working Condition Survey (NEA) in the Netherlands, Finnish National work and health survey (FNWHS), Norway - Survey of living conditions - Working environment (LKU, Statistics Norway), Work Environment and Health in Denmark (NRCWE), German Labour Force Survey (BIBB/ BAuA) and the European Working Conditions Survey. Spain uses the EWCS questionnaire in an enlarged sample to measure status and trends of working conditions, hence, the items measuring physical exposures in Spain and in EWCS are therefore identical. All included surveys are repeated periodically. For each survey, the latest wave with an English translation was selected. Professional translators in each country translated the items from their original language into English. There are two exceptions: in Germany, professional translators only translated the older version of the survey from 2006, and two English-speaking scientists from the field of occupational safety and health translated the items that were new in the latest survey; in Denmark, only a preliminary translation of the questionnaire exists as of yet. In an additional file (see Additional file 1: Description of the EU-wide and the national surveys included in our paper), we have provided a more detailed description of the surveys, also including a reference list of epidemiological papers and presentations based on the surveys.

After consulting European experts on physical exposures at work and joint discussions, the cooperating partners agreed on a taxonomy to classify the physical working condition items of all questionnaires. This taxonomy and the sorting are based on consensus among the partners and comprises nine domains (Table 1): (i) lifting, holding & carrying of loads /pushing & pulling of loads, (ii) manual work processes /repetitive hand-arm movements, (iii) working standing/walking, (iv) working sitting, (v) awkward body postures, (vi) physical work effort, (vii) vibrations, (viii) work with computer, (ix) mixed exposure. For six of these domains, a further distinction into two to six dimensions was required to account for the specificity of the aspects covered, leading to a total of 22 dimensions. All items referring to physical workload were evaluated and sorted into one of these 22 dimensions. This assignment was done independently per country by the researchers for the respective countries (TT and CCA for Norway, SVT and LLA for Denmark, MPM for Finland, JVPG and LGB for Spain, AV and AGP for Europe, WH and IH for the Netherlands, HB and MF for Germany) and cross-checked by the other researchers. If items were assigned to different categories, the assignments were discussed and decided by consensus. In cases where an item referred to a combination of working conditions, not allowing for a clear classification into one dimension, this item was placed in the domain “mixed exposures” (Table 1).

In musculoskeletal epidemiology, the exposure assessment strategy remains a huge challenge, and there is no perfect instrument for measuring all relevant dimensions of physical loads [54]. The aim in a questionnaire addressing dimensions of physical workloads is to assess frequency, level and duration of certain types of exposure [55, 56]. In epidemiological studies, however, all three dimensions are rarely collected simultaneously. Questionnaires in national surveys are designed to aim at physical workloads in general, and these self-reports can give some insight into the occurrence of tasks and activities, and the approximate proportion of time spent on each of them. At the same time, most estimates of external exposures like the working situation, the actual working method, as well as the triad postures, movements, and exerted forces, are, however, imprecise and inaccurate with too low validity and reliability [55]. Exposure data from self-administered questionnaires on manual materials handling and work postures have been validated in relation to direct measurements and systematic observations [57]. At the dichotomous level, the agreement was “acceptable” for a majority of the variables concerning work postures and for handling of loads weighing >5 kg. No variable, however, showed “acceptable” agreement when the duration or the frequency was quantified in more detail (4- to 6-point scales). A limitation in our study is the fact that it is not exhaustive regarding the existing national working condition surveys in Europe, and access to more studies would have provided a broader picture of the domains and dimensions covered on this topic. A wider scope was not possible due to the limitation to those countries that volunteered to take part in the PEROSH group. In addition, the PEROSH group decided solely to focus on monitoring instruments; hence, questionnaires from other studies with a different scope (e.g. SHARE or SILC) have not been included in the overview.

A limitation regarding the EWCS is the low number of study subjects in each country and a large proportion of non-response [58]. The national surveys may also suffer from low response, and this might limit the generalisability of the results to the targeted populations.

All the survey data included in this paper are collected by self-report, either as telephone or as personal interviews or as self-administered paper questionnaires. Self-report offers the possibility of studying a great number of persons nationwide at a modest cost while allowing the investigation of a large number of variables; it is a feasible method to assess exposures that occur with highly irregular patterns, for example exposures that change seasonally or exposure in the past [59]. Self-report measures are therefore relevant in the context of large population-based surveys that include many job titles, multiple workplaces, and a wide variety of occupational tasks and are designed to monitor general trends of exposure to important determinants of musculoskeletal disorders such as physical load over time [60]. Self-reports may also help to effectively convey relative differences in exposures of heterogeneous populations, but at the same time they are imprecise measures of the absolute levels of the exposure [54]. A review of the validity of self-reported mechanical demands for occupational epidemiologic research of musculoskeletal disorders indicated that among assessments reporting correlations as a measure of validity, studies with a better match between the self-report and the reference method, and studies conducted in more heterogeneous populations tended to report higher correlations [59]. The authors concluded that the use of self-reported mechanical demands for occupational epidemiologic research requires further, better validity testing research, and that the full potential of self-reports in occupational epidemiologic research is still to be discovered. Stock et al., in their systematic review, stated that the validity of self-reports for the assessment of mechanical exposures cannot be appropriately established with the information currently available [60]. Instead, due to the need for practical, questionnaire-based measures for epidemiologic studies, efforts should pertain to improve the design of individual items and their response scales. Stock et al. evaluated self-report items on physical work demands for both reproducibility and validity. Items that performed well included those on duration or presence of sitting and standing posture, the presence of walking, kneeling or squatting postures as well as duration or frequency of hands above shoulders. Manual handling of more than or less than 10 kg, general level of physical effort, presence and duration of whole-body vibration, and duration of the use of visual display terminals also performed well. All dimensions showing good reproducibility and validity in the review by Stock et al. are included in the majority of the surveys in our study. In a recent Norwegian study aiming at determining the criterion validity of the Norwegian survey items on physical load in comparison to objective measurements, the authors concluded that the self-report overestimated the exposure durations; the highest correlations between self-report and measurements were reported for sitting and work with hands above shoulder height [61].

In our taxonomy, work with computer is defined as a domain, but one could also say that work with computer is a mix of exposures, e.g. working sitting down or standing, repetitive work, potential awkward arm−/hand movements and other exposures or demands regarding visual demands or psychosocial influences. A study of computer workers indicated that exposure may also depend upon the usability of the computer software [62]. The authors reported that users frustrated by the software, increased their forces applied to the mouse and increased EMG activity of the wrist extensor muscles. The data also suggest that other factors, such as individual factors, may confound the results: Increased exposures were only observed among those who were critical of the usability of the software.

Another example of mixed exposures is the Norwegian item “Do you work in positions where you are leaning forward without supporting yourself on your hands or arms? If yes, how much of the time do you do this during a normal working day?”, followed by two items” When working in those positions, do you work in those positions with your back twisted?” and” When working like this, do you need to lift anything that weighs more than 10 kg?” The idea with this combination of items is to find workers exposed to both awkward positions and high muscular strain implying large energy expenditure. The number of subjects exposed to this combination of exposures in the national study (in Norway about 9000 economically active take part in the survey) is, however, too small to properly evaluate potential negative muscular health effects of this combination of exposures.

There are still gaps in knowledge concerning the latency of effects of specific physical workloads, the potential of a selection bias in the form of a “healthy worker effect”, and the threshold at which the exposure starts to produce a negative physiological effect. Consequently, prospective longitudinal scientific studies in a large number of countries are needed in order to evaluate which dimensions of the physical work environment are most important for health. The Norwegian survey has a panel design since 2006 as does the Danish survey since 2012 with those that responded in the first round (in 2014: 1/3 of the whole sample). Such a design allows for a longer follow-up. Currently, however, the lack of standardized exposure metrics across surveys limits the ability to compare findings across countries and studies. Therefore, a better harmonization of items and answering categories across studies would be desirable for comparisons across countries. However, changing items in repeated surveys prohibits the analysis of trends over the years. Moreover, the stakeholders involved in these surveys are reluctant to change tools on which agreement has been made. An alternative are studies showing how well the different items and answering categories correlate before joint analysis with other European surveys can be undertaken.