Psychosocial job strain as a mediator between physical working conditions and symptoms associated with sick building syndrome

Method

Participants. The sample for this study comprised 1505 employees in white-collar occupations (with complete data) from 20 companies that participated in the third wave of a longitudinal survey on working conditions and health (for details, see Devereux, Rydstedt, Kelly, Weston, & Buckle, 2004). About 13% of the participants held managerial positions, 31% were professionals and 41% semiprofessionals, while the remaining 15% had administrative office occupations (SOC 2000). The majority of the participants were males (57%), while 43% were females. Mean age at the start of the study was 42 years; the participants reported working on average about 40 hours per week.

Psychosocial job strain. To analyse the relations between psychosocial job strain and SBS-related symptoms, the original version of the Effort-Rewards Imbalance (ERI) Model (Siegrist, 1996; 2000) was used. The ERI model is based on sociological Equity Theory and the central assumption behind the model is that distress occurs when the worker experiences a disadvantageous imbalance between the effort spent on accomplishing work in relation to the material and social rewards received. Equity Theory predicts that a subject who perceives that his/her efforts are unfavorably rewarded will react by reducing that effort; however, this may not be possible in the contemporary labor market (Siegrist, 1996). Several empirical studies have confirmed the predictive power of the ERI model (e.g. Reinhardt, Wahrendorf, & Siegrist, 2013; Van Vegschel et al., 2005). Efforts were measured by six items (α=.75), e.g. “I have constant time pressure due to work overload”, and Rewards by 11 items (α=.86), e.g. “My job security is poor (-)”. The ERI ratio (M=0.49) was calculated in accordance with the instructions given at www.uni-duesseldorf.de/MedicalSociology.

Perceived quality of the indoor physical work environment. The participants were presented with 10 potential problems affecting the indoor work environment that have been examined as potential causal agents for SBS: noise, improper lighting conditions, indoor temperature, humidity, bad air circulation, pollutions, toxic substances, ergonomic/physical working conditions, confined/ over-crowded work spaces. The participants reported an average 2.4 of these problems at their workplaces.

Outcome variable, diffuse symptoms. A scale comprising 13 different diffuse symptoms (α=.82) that are considered when diagnosing SBS e.g. fatigue, skin rash, dryness in the mouth, dizziness, eye infections—were used as the outcome variable. Seven of these items were taken from an inventory of diffuse psychosomatic symptoms (Hurrell & McLaney, 1988), while the remaining six items were chosen for this study. The frequency during the last month of each type of symptom was reported on a Likert type scale with five possible responses (1= never- 5=very often; M=1.67).

Results

The path analysis, as shown in Figure 1 below, revealed a direct correlation between perceived problems in the physical indoor work environment and SBS symptoms (r= .285; p<.001). As also shown in Figure 1 there is a significant relation between the indoor work environment and psychosocial job strain (ERI) (.245; p<.001), which in turn is strongly related to the diffuse general SBS type symptoms (.31; p<.001). Sobel’s test of the mediating effects of psychosocial job strain was highly significant (p<.000).

Figure 1

Path analysis of the relations between the indoor physical work environment, psychosocial job strain and their relations to general symptoms of SBS type. (N=1501).

*All relations in Figure 1 are statistically significant (p<.001).

Numerical values in the parentheses are beta weights, other values are zero ordercorrelations.

Discussion

The results of this study show a significant direct relationship between physical work environment factors and general SBS type symptoms. Furthermore, the physical work environment relates significantly to perceived psychosocial job strain, which in turn also significantly relates to the level of self-reported general SBS type symptoms. The findings in this study thus suggest a possible mediating role of perceived psychosocial job strain— between the indoor physical work environment and, on the other hand, vague SBS-type symptoms.

According to the proponents of the mediating hypothesis, the seemingly strong influence psychosocial job strain has on SBS symptoms could at least to some extent be explained by the impact physical work environment factors have upon psychosocial discomfort (e.g. Hansen, Meyer, & Gyntelberg, 2008). Neither of the two most applied conceptual models for analyzing the impact of psychosocial working conditions upon wellbeing and health—the ERI model (Siegrist, 1996) and the Job Demand-Control(-Support) Model (Karasek, 1979; Karasek & Theorell, 1990)—pays any attention to the role of the physical environment. It may be reasonable to assume that at least some of the reported variation in the efforts or demands may actually be caused by unsatisfactory factors in physical work environments such as over-crowded workplaces, noise, improper ergonomic design of work stations or bad air quality. Even though these factors may affect psychosocial working conditions and work-related well-being, they cannot be directly measured or reported in any of the above mentioned models. Nonetheless, it seems likely that the impact of unsatisfactory physical work environment factors may influence these inventories, and be indirectly reported work overload or lack of control.

Furthermore, the physical working arrangements often signal the worker’s rank and prestige, and reflect how he/she is valued by the organization and by his/her superiors (Briner, 2000; Clements-Croome, 2011; Vural & Balnani, 2011), which it can be assumed is reflected in the Reward dimension in the ERI model (Siegrist, 1996). In a cross-sectional study based on a large sample of Canadian female employees, Burke (2002) found a strong negative relation between occupational status and hassles in the physical work environment. In a qualitative study on attitudes and reactions towards the physical working environment in call centers, Barnes (2007) found that workers often felt office spaces were designed for the purpose of enhancing managerial control and surveillance of the employees. Barnes (2007) identified the physical work environment design that has open office areas as often being the target of resistance and discontent over the overall working conditions. Arrangements that deprived the employees of any degree of privacy were seen as a threat to the dignity of the office workers. There may thus be an uninvestigated conceptual overlap between factors in the physical work environment and psychosocial working conditions as well in socioeconomic factors. It may be reasonable to assume that consideration of the physical work environment may contribute to improved explanations of work-related health and wellbeing as well as to a better understanding of the causes of hazardous working conditions.

The diffuse symptoms of SBS and the difficulties diagnosing it, along with the consistent difficulties in validating evidence of causal links between specific factors in the indoor environment and SBS, have led other scholars to question the etiological role the physical environment plays in SBS (Burge, 2004; Escobar et al., 2002). According to Wiesmüller et al. (2003) SBS should be classified as an “environmental related problem” along with other diffuse impairments for which no known causes have been identified. As a possible, but so forth mainly untested alternative explanation, Wiesmüller et al. (2003) suggest that the hitherto unknown interactions between different forms of environmental and individual psychological factors may play a crucial role in the etiology of SBS. Extending this line of thinking to explore possible interactions between physical and psychosocial work environment factors, as suggested by Bass et al. (2002), may be a way forward in gaining a causal understanding of SBS.

In the Whitehall II study Marmot et al. (2006) found that participants who claimed to have control of the physical environment at their workstation reported significantly lower frequencies of SBS symptoms than those who lacked control over their physical work environment. Marmot and colleagues also point out that the interactive Whitehall II study did not ascertain why this type of control reduced the frequency of SBS symptoms. Two alternative, but not entirely exclusive explanations could be that by having control over the physical environment of the workstation the worker is capable of eliminating or reducing environmental hassles. Alternatively, having perceived control over the physical conditions may in itself be beneficial to health. The present study suggests that the impact of the physical work environment on SBS symptoms is to some extent mediated by the psychosocial work environment. One can also assume that psychological and psychosocial work environment factors may interact to amplify the effects of the other—e.g. that a high burden of work may exert more strain on the individual if it has to be accomplished under unfavorable conditions in the physical work environment. An interactional approach to analyzing the joint health impact of physical as well as psychosocial factors in the work environment has previously been suggested by Marmot et al. (2006) but has not been undertaken thus far. An obvious suggestion for future research would be to conduct longitudinal interactional studies on this theme.

Limitations of the study. There are several limitations to this study. The cross-sectional design of this study does not allow causal inferences to be drawn about the relationships between psychosocial working conditions, indoor building quality and SBS symptoms. Controlling for baseline levels of SBS symptoms and ruling out reversed causation would have required a longitudinal study with full panel design. Neither has any attempt been made in this study to eliminate the possible impact from third variables on the path analysis.

Furthermore, since the study was exclusively based on self-reported data it is likely that the common method variance inflates the relations between the variables. Finally, it should also be noted that we do not have information on whether any of the participants actually work in buildings that have been identified as a “sick building”. Neither do we know if any of the participants have been clinically diagnosed with SBS; we only know the extent to which they report symptoms associated with the syndrome.

Conclusions

There is strong disagreement in the scientific community regarding the causes behind sick building syndrome. Previous research has not been able to firmly validate that any distinct factors in the physical environment are causal agents of SBS symptoms. The findings in this study suggest that analyzing the interacting impact of physical as well as psychosocial hassles on this type of health outcome may offer a better understanding of the complicated causes of SBS. The inventories most applied to analyzing the work environment focus exclusively on the psychosocial aspects of working conditions and may be affected by the unaccounted effects of physical working conditions. Additional longitudinal studies on the impact of physical working conditions (e.g. crowding, office design) on psychosomatic health issues are required to expand the knowledge of the thus far not fully explained relations between physical and psychosocial working conditions.