Abstract
The relationships among bullying or witnessing bullying at work, self-reported health symptoms, and physiological stress reactivity were analysed in a sample of 437 employees (294 women and 143 men). Physiological stress reactivity was measured as cortisol in the saliva. Of the respondents, 5% of the women (n=15) and 5% of the men (n=7) reported bullying, whereas 9% of the women (n=25) and 11% of the men (n=15) had witnessed bullying at work. The results indicated that the bullied respondents had lower social support from coworkers and supervisors, and they reported more symptoms of somatisation, depression, anxiety, and negative affectivity (NA) than did the nonbullied respondents. Witnesses reported more symptoms of anxiety and lower support from supervisor than did the nonbullied employees. Concentrations of cortisol in the saliva were lower at awakening in bullied respondents compared with nonbullied respondents. Previous studies have reported lower diurnal concentration of cortisol for people with posttraumatic stress disorder (PTSD) and chronic fatigue. To our knowledge, this is the first full study on the associations among being subjected to bullying, health outcomes, and physiological stress response.
Introduction
Bullying at work, according to most definitions, takes place when someone, repeatedly over a longer period of time (usually 6 months), is exposed to negative acts from one or several others, in a situation where he or she for different reasons may have difficulties defending him- or herself against these actions [1], [2] and [3]. Bullying may take different forms. Direct bullying is aggressive acts that are aimed directly at the target, as, e.g., teasing, scolding, spreading rumours, and threats. Indirect bullying may take the form of social isolation or withdrawal of necessary information. Bullying may be work related (e.g., acts that make it difficult for the targets to do their work) or personal (e.g., offending teasing, rumours, slander, or sexual harassment; [4] and [5]). One of the characteristics of bullying is the inequality in power between the perpetrator and the target [1]. Either supervisors are directly involved or the bullying takes place between coworkers, where the perpetrator, for some reason or another, is stronger than the target is. Even supervisors may be bullied by subordinates, especially if the subordinates act in groups [2] and [6].
The prevalence of bullying shows great variation depending on how it is measured and whether the respondent is provided with a definition of bullying. When given a definition, the prevalence mostly varies between 2% and 17% [3]. Bullying is most frequent in workplaces with a negative and stressful working environment [7] and [8]. In addition, targets of bullying often lack social support from coworkers and supervisors [9], [10] and [11], and it has been demonstrated that perceived low social support is related to distress and burnout, for instance, in the relation between supervisors and subordinates [12] and [13]. This is unfortunate because a supportive work environment seems to be able to protect employees from some of the harmful effects of bullying [14] and [15].
Of the respondents (N=437), 22 (5%) were subjected to bullying; 15 were women, representing 5% of the female population, and 7 were men, representing 4.9% of the male population. Bullied respondents came from all companies, varying between 2% and 7%. We found no significant age difference between the bullied (mean age=49 years) and nonbullied respondents (mean age=47 years; P value=.42). A total of 40 respondents had witnessed bullying, but they were not subjected to bullying themselves (F=25; M=15). The overall response rate was 65%, varying between 40% and 85% ( Table 1).
Table 2 presents Spearman's correlations (ρ) among somatisation, depression, anxiety, NA, and social support from supervisor and coworkers for witnesses and bullied respondents (Hypotheses 1a, 1b, 2a, and 2b). All outcomes were highly correlated. All the outcome variables, including NA, were correlated with being bullied, but only anxiety and low support from supervisors were significantly correlated with witnessing bullying. Pearson correlations showed similar results. A one-way ANOVA of the health symptoms, NA, and social support from coworkers and supervisor for respondents subjected to bullying, witnesses, and nonbullied respondents showed statistically significant differences between those subjected to bullying and the two other groups. Post hoc tests showed significant differences between bullied and nonbullied respondents for all the variables. There were no significant differences between the nonbullied and witnesses of bullying in any of the health symptoms and NA ( Table 3).
