This paper will explore a theoretical model linking the chronic stress of workplace bullying with alterations in cortisol secretion, provide qualitative descriptions and definitions of bullying, and describe tools that have been developed to operationalize the construct. Further, the use of salivary cortisol assays to measure free serum cortisol will be discussed. Research that set out to link the direct personal experience of bullying among workers with cortisol response deviations will be examined.

High prevalence of bullying in nursing has been suggested, making the topic of concern to the profession. Bullying has been demonstrated to have physiological and psychological effects for its victims. It has also been shown to increase employee absenteeism, turnover, and negatively impact employee work product. Finally, recommendations for further research to more fully understand implications to health and mitigation of ill effects will be proposed.

Introduction

While many might describe stress as an unpleasant but acceptable, even necessary at times, aspect of the workplace, bullying represents a unique form of stress that may induce long term health consequences in its victims. Ramifications to the individual facing such negative interactions during the course of daily employment include anxiety and depression (Niedhammer, et al 2005). Matthiesen and Einarsen (2004) found that 3 out of 4 respondents in a study of bullied workers scored above the threshold for a diagnosis of post traumatic stress disorder (PTSD). Physical symptoms are also manifested in increased absenteeism (Murray, 2010). Fries, et al (2005) reported that low cortisol pursuant to chronic stress, as exhibited in victims of long term harassment, correlated with "enhanced stress sensitivity, fatigue and pain" and loss of cortisol mitigation of the immune system resulting in increased inflammatory responses. In fact, an entire workforce can suffer from reduced morale and loss of a cooperative work environment (Rocker, 2008). Adverse effects on well-being and job satisfaction have been reported among bystanders in the workplace (Cowie, 1999).

In nursing, and perhaps the field of health delivery in general, the prevalence of bullying behaviors may exceed that of the overall work force. While many large scale studies of worker satisfaction find self-reports of exposure to the behavior on the order of 10% of respondents, studies of nursing populations are often much higher. Simmons (2008) found that 31% of nurses responding to an investigation of prevalence reported having been bullied. In nursing, behaviors consistent with bullying have been implicated in significant increases in medication errors and other adverse patient events (Martin, 2008). Whereas nurses could be at greater risk for exposure to bullying in the workplace, and would suffer individually the psychological and physiological effects of the chronic stress it represents, their patients could also suffer.

This paper will explore a theoretical model linking the chronic stress of workplace bullying with alterations in cortisol secretion, provide qualitative descriptions and definitions of bullying, and describe tools that have been developed to operationalize the construct. Further, the use of salivary cortisol assays to measure free serum cortisol will be discussed. Research that set out to link the direct personal experience of bullying among workers with cortisol response deviations will be examined.

Literature Review

A literature review was conducted to find primary research relating bullying and cortisol levels. Web of Science was searched using keywords bullying or mobbing, and cortisol or hypohalmo-pituitary-adrenal axis as topic. The search yielded 13 matches. Articles not specific to measurement of cortisol profiles in otherwise normal adults experiencing bullying in the work place were excluded. To develop the theoretical framework and context, Web of Science and Google Scholar were searched using additional keywords: bullying or mobbing and measuring, salivary cortisol, mobbing or bullying and nursing, chronic stress and cortisol.

Definitional descriptions of bullying have been advanced and expounded upon by numerous researchers. Generally, aggressive acts that are repetitive and of duration, inflicting emotional harm on a victim are viewed as potential bullying behaviors. These include demeaning comments, threats to physical or professional well being, social isolation, relational disruption, and over or under work (Cowie et al, 2002 ). Singular or isolated instances of negative acts within the work place experience do not constitute bullying. Leymann (1996) suggests that a weekly frequency over a period of six months sets the criteria for bullying.

Einarsen (2009) accepts the six month duration from a convenience of measurement perspective, suggesting that restriction to this time period both ensures capturing duration without potential problems of subjects' memory recall or distortion. As such, the specific requirements for duration and frequency of acts is somewhat arbitrary and chosen for convenience by the social scientists. Leymann and several other European researchers prefer the term "mobbing" to identify bullying; the translational difference from the Anglo-American term "bullying" suggests the possibility of operation of more than one perpetrator acting in concert against a victim.

Implicit in the conceptualization of bullying is an asymmetrical power relationship between perpetrator(s) and victim; were it not so, the target of negative acts would feel empowered to counter the aggression. Often there is an institutional hierarchical mismatch between the subjects; however, the interpersonal relationship may be one of uneven power, even among peers. Differences in personality, social standing, and experience all represent potential fields of imbalance. The perception of power and helplessness in the subjective experience of events is the essential definitional consideration in this aspect (Cowie, et al 2002)

The perception of bullying is subjective in many respects. Measurement of the phenomenon from a practical standpoint requires assessing potential victims and the potential distortions of personal perspective. However, According to Niedl (2006, as cited in Einarsen, 2009), the potential impact of bullying on health and well-being exists precisely in the subjective interpretation of behaviors; "the definitional core of bullying at work rests on the subjective perception made by the victim that these repeated acts are hostile, humiliating and intimidating and that they are directed at him/herself." A critique of relying upon subjective perception has been that it possibly admits a circular argument into the assessment (Hoel, 2009), ie because one believes oneself to have been victimized, one is, therefore, a victim. Brodsky (1976, as cited in Einarsen, 2009) suggests that truly objective measures of bullying must be performed by third party observation.

Obviously, there are difficulties in applying the approach in retrospective analysis where a trained observer was not present during the acts in question. Coworker hesitance to provide information because of their economic dependence upon relationships at work is also cited as making third party peer nominations of bullying perpetrators and victims less than reliable (Bjorkqvist, et al 2004). Critics of the self-report method note that there is little evidence of establishment of validity by test-retest to establish the accuracy of participant's recollections and interpretations of incidents over time.

Operationalizing the phenomenon of bullying has relied upon questionnaires administered to potential victims. Heinz Leymann developed and factor analyzed the Leymann Inventory of Psychological Terror (LIPT) in 1999 to diagnose severe instances of bullying. Originally written in Swedish, the inventory has been translated and validated in German, French, and English. To remove subjectivity from the self-report, Leymann establishes a rigid set of criteria, predefining negative acts and specifying frequency and duration guidelines. In this approach, several kinds of specific incidents are described, and respondents indicate whether or not they have experienced them within the time frame described. The tool hopes to measure perceived exposure to behaviors which Leymann has determined are indicative of bullying (Notelaers et al, 2006).

Leymann (1990) identified a series of behaviors during case study research which he classifies into five categories with respect to the effects they have upon the target of bullying. These correspond to diminished ability to communicate, social isolation, diminishment of personal reputation, occupational restrictions, and physical assaults. The tool has been criticized as being overly long, with 45 items; culturally biased; and possible uses wording that is emotionally charged and likely to influence responses. Leymann acknowledges the cultural bias, advising that "...these activities mainly describe hostile actions as they are carried out in northern European countries."

While the LIPT had been structured to identify severe forms, Einarsen devised the Negative Acts Questionnaire (NAQ) to allow respondents to self-identify as victims of bullying and identifies a broader range of negative behavior related to bullying. The NAQ was originally developed with 23 items to investigate subjects' perceptions of workplace bullying. Its revised form, the NAQ - R attempts to resolve issues of cultural bias (the original NAQ was written for a Scandinavian population) and factor structure limitations. The 22 item scale has been adapted to an Anglo-American cultural setting, and is devised to test for occupational, interpersonal and physical aspects of bullying using behavioral terms (Einarsen, et al 2009). Internal consistency is established with a Cronbach's alpha of .90.

Cortisol Responses

The HPA axis is a major route by which external social and psychological factors elicit physiological responses (Adam & Kumari, 2009). It is well well established that an individual's cortisol levels vary in response to stress. Use of measurements from saliva as opposed to serum or urinary levels has been suggested as a means of furthering psychoneuroendocrinological understanding of interactions between stress and effects on the body (Hazler, et al, 2006). Ease of sample collection and storage, facilitating greater acceptance and compliance by subjects, would make possible large scale studies of HPA axis changes in specific populations. Polla, et al (2007) found that salivary cortisol could be a reliable indicator of total free serum cortisol. While salivary concentrations are 1000 lower than found that of serum, the levels have been found to correlate (Rantonen et al, 2000).

Operational definitions using measurements of salivary cortisol to assess HPA response would be valid. Insofar as absolute cortisol levels may not be the relevant object of study, research correlating cortisol and bullying have focused on the changes of daily cortisol levels from cortisol awakening response. Minimum protocol for determining cortisol diurnal slope, that rate of change between levels at bedtime and awakening response, is measured with at least two measurements; as many as 5 points taken throughout the day or multiple days with line of best fit modeled with linear regression provides better resolution, though it is unknown if the degree of curvilinearity of the slope is significant to any process under investigation (Adam & Kumari, 2009).

Theoretical Background

Stress and psychological distress as an outcome of a hostile work environment has been established. Leymann's clinical case studies provide a body of qualitative research making the association. Niedhammer (2007) established statistical associations between negative employment outcomes, ie, absenteeism, sickness, and work injury and employee-reported feelings of workplace intimidation. Chronic stress has been linked to a wide range of physiological and psychological diseases from clinical depression to coronary heart disease (McFarlane, 2010). Hypothalmic-pituitary-adrenocortical axis (HPA) activation resulting in heightened hormonal secretion, a normal response to stressors, was thought to increase tissue exposure to damaging levels. In victims of acute trauma, highly elevated levels of cortisol have been measured shortly after the traumatic event (Reul, 2008), a phenomenon that has been noted in laboratory studies of induced stress.

Therefore, traditional models of disease related to chronic stress have implicated cortisol as an intermediary, disrupting normal biological processes and causing tissue damage (Miller, et al 2006). The confound has been in contradictory results from studies of PTSD and sufferers of chronic trauma and stress in whom cortisol levels seemed to be depressed (Rehl & Nutt 2008).

Cortisol is a hormone that is mediated by numerous external stimuli. Social interaction, physical sensation, and immune system responses can induce changes in levels of cortisol. In return, the hormone has regulatory influences over immune system responses, sensory acuity, memory acquisition, and the "fight or flight" response.

As an evolutionary adaptation, cortisol's effects serve to preserve the organism under threat or challenge; learning and memory systems are inhibited, immune responses are suppressed while the threat is expressed. Subjected to stress, levels of cortisol increase. Miller, et al (2006) conducted a meta analysis of 119 articles in which HPA axis function was assessed in subjects having experienced past or ongoing stress of duration of a month or longer. Compared with those who had not experienced chronic stress, those having had long term exposure to stress had lower awakening concentrations of cortisol, greater afternoon levels and higher daily volume of output. The overall effect was a flattening of diurnal response.

By contrast, in control subjects, the daily cortisol profile shows a high awakening level which declines rapidly during the day. The study discovered patterns of response specific to social stress, and the subjects ability to control stressful outcomes. Specifically, situations posing threats to the social self were associated with lower morning cortisol levels, higher afternoon levels, resulting in a flattened response with overall higher total daily volume. The same pattern seen in social stress is repeated for situations in which the stressor is perceived as uncontrollable.

Clear from the research is that chronic stress is associated with alterations in how the HPA axis is activated throughout the day. Studies which drew conclusions about stress related changes in HPA functioning through a single daily sample were missing the broader picture. An assessment of cortisol output due to environmental stressors can only be reliable from a series of daily measures which describe total daily volume.

Research Linking Cortisol to Workplace Bullying

In 2004, Kudielka and Kern conducted a small-scale preliminary study of cortisol profiles in subjects of workplace bullying. The pilot study conducted in Switzerland hoped to explore whether bullying could be associated with measurable signs of HPA axis changes. The German language version of the LIPT was used to assess exposure to mobbing; no indication of subject recruitment methods or sample characteristics was given. Exclusionary criteria were good health, score on LIPT indicating bullying exposure, and in active employment. Twenty eight participants were enrolled, with an average mobbing duration of 62 months.

Cortisol salivette samples were taken at waking, 30 minutes after waking, and at 8:00, 11:00, 15:00, 20:00, and 22:00. The method was used for a day in which each subject worked, and a day in which the subject did not work. A compliance rate of 57% was achieved, as 17 participants returned all samples requested. No significant difference was found between waking cortisol levels or diurnal profile of days at work or days off. A trend towards lesser decrease between morning and evening levels between work day and day off was noted.

The authors acknowledge that the small sample size was an issue and made a suggestion that a larger scale study should be done. Few conclusions can be drawn from the results of such a small sample, the characteristics of which are unknown from the article. Exclusion of those who did not test positively for exposure to mobbing deprived the study of a control group with which to compare results, which is a serious flaw. Kudielka and Kern refer to studies suggesting that flattened circadian cortisol profile may indicate risk for disease, but do not elaborate on theoretical background.

Monteleone, et al. conducted an investigation into cortisol profiles and mobbing with an emphasis on subjects' temperament. The authors cite studies linking severe stress from exposure to bullying and symptoms of PTSD, chronic fatigue among others. The aim of the study was to evaluate whether bullying stress effects HPA activity, and whether personality traits play a part in HPA responses in victims of bullying. The study was conducted in Italy. Ten subjects who tested positive for having been bullied using the LIPT inventory were gender matched with a control group who did not test positive for bullying. The average duration of exposure to workplace harassment was 24.3 months. Subjects signed informed consent, were physically healthy and drug-free for at least six weeks. Cloninger's Temperament and Character Inventory - Revised (TCI-R) was administered to assess subjects' temperament. Subjects were then asked to go to bed at their normal time, then awake and take saliva samples at 6:00, 8:00, 12:00, 16:00, and 20:00.

Results showed that bullied subjects had significantly lower cortisol concentrations at awakening as well as overall lower daily cortisol volume than the control group. Monteleone correlated a characteristic of harm avoidance with victims of bullying, and suggests that the characteristic might make them less able to cope with social challenges and impair their ability to deal with work place harassment. Two way ANOVA showed that duration of bullying explained 76% of the cortisol variations in bullied subjects; after removing this effect, the harm avoidance characteristic explained another 11% of the variation.

Despite a small number of participants, the authors noted a significant difference between cortisol profiles between the bullied and control group. It is unclear from the methodology if the sampling protocol was for a single day for each subject, or if multiple days were included in the protocol. One would expect individual daily variability in cortisol levels unrelated to the phenomena studied; sampling over multiple days might allow researchers to identify and exclude from analysis these types of variations.

The methodology may have a flaw in the request that participants go to bed at their habitual time , but awaken at 6:00. The normal circadian rhythms of the participants may not have been habituated to awakening at the prescribed time, and therefore cortisol levels may not accurately indicate normal wakening cortisol levels. The authors give no rationale for this aspect of the protocol.

The authors also included a personality inventory, and concluded that behavioral and social inhibition may account for responses of victims of bullying. The error in logic is the assumption that these are stable characteristics that existed prior to exposure to bullying. The authors were aware of studies relating PTSD-like symptoms to mobbing victims. According to the DSM-II-R, PTSD in a chronic phase is a condition that results in permanent alterations of personality. Assessing personality after onset of effected changes makes it impossible to make conclusions as to its the original state; one can only describe what it has become. As such, the inclusion of this leg of the study is of questionable value.

Recommendations

Several studies conducted to measure cortisol variations in victims of bullying as identified using the LIPT assessment tool have noted flattened diurnal profiles. Each study was retrospective in that cortisol profiles were determined after identification of victimization. The conclusion that these deviations from expected values were due to a response to external experiences cannot be made conclusively. While flattened diurnal response to chronic stress is predicted from some models of HPA activity, more investigation is required to draw more definitive causality between the two phenomena.

It may be difficult to construct a study of the breadth required to establish normal baseline cortisol profiles which are then altered in response to bullying. The commonalities with PTSD, in which cortisol alteration persists for decades after the initial trauma, suggest that cortisol variations, if caused by the experience of bullying stress, may also persist as long,. Further research could be designed to establish the latency of altered profiles; whether there are conditions under which they improve to normal.

Neither study into the relationship between cortisol response and bullying has examined cortisol measurements in direct response to bullying behavior. A study could be designed to take baseline measurements, with additional sampling during acute experiences of aggressive behavior to determine whether cortisol levels remain unresponsive during episodes of increased stress, or if levels increase.

Physiological ramifications of flattened cortisol response have yet to be fully examined. It is not known if alterations to normal rhythmic pattern of secretion is in itself a pathology, or if emphasis should be placed on measurement of total daily cortisol volume. According to current theory, it is this parameter, not absolute values nor their particular diurnal profile, which leads to symptoms.

While further research may yet more fully describe the process of bullying and the physiological and psychological effects it has on victims, little investigation has been done into reversing what seems to be a destructive process. Interventions which can reestablish normality to the circadian cortisol rhythm, and repairing the psychological damage done to victims should be developed. Improving workplace awareness of the problem and developing techniques to mitigate conflict before it escalates into bullying are also indicated.

Written by: Pamela Irvin- Lazorko

Pamela Irvin- Lazorko is a nursing doctoral student at Drexel University in Philadelphia, PA USA.

References:
- Adam, E., Kumari, M. (2009). Assessing salivary cortisol in large-scale epidemiological research.
Psychoneuroendocrinology. volume 34 pages 1423 -1436.
- Bjorkqvist, K., Osterman, K., Hjeit-Båck, M. (1994). Aggression among university employees. Aggressive Behavior. volume 20 pages 173 - 184
- Cowie, H., Naylor, P., Rivers, I., Smith, P., Pereira, B.(2002). Measuring workplace bullying. Aggression and Violent Behavior. volume 7 (2002) pages 33-51
- Cowie, H., Bradshaw, L., Kaipiainen, S., Smith, P., et al. (1999). Adult bullying: report of a working party chaired by Helen Cowie. Nature and Prevention of Bullying. retrieved online at
http://old.gold.ac.uk/tmr/reports/aim2_surrey1.html
- Dickerson, S., Gruenewald, S., Kemeny, T. (2009). Psychobiological responses to social self threat: functional or detrimental? Self and Identity. volume: 8, Issue: 2-3. pages: 270-285
- Einarsen, S., Hoel, H., Notelaers, G.( 2009). Measuring exposure to bullying and harassment at work: Validity, factor structure and psychometric properties of the negative acts questionnaire-revised. Work and Stress.
Volume: 23. Issue: 1. pages: 24-44
- Fries, E., Hesse, J., Hellhammer, J. , Hellhammer, D. (2005). A new view on hypocortisolism.
Psychoneuroendocrinology (2005) xx. pages1-7
Hansen, A., Hogh, A., Persson, R., Karlson, B., Garde, A., Orbaek, P. (2006). Bullying at work, health outcomes, and physiological stress response. Journal of Psychometric Research. volume 60 pages 63 - 72.
- Hazler, R., Carney, J., Granger, D. (2006). Integrating biological measures into the study of bullying. Journal of Counseling and Development. volume 84 pages 298 -307
- Hoel, H., Beale, D. (2006). Workpalce bullying, psychological perspectives and industrial relations: towards a contextualized and interdisciplinary approach. British Journal of Industrial Relations. 44:2 June 2006 0007-1080 pages 239 - 262
- Joint Commission.(2008).Issue 40: behaviors that undermine a culture of safety. Sentinel Event Alert. July 9, 2008. retrieved online from: http://www.jointcommission.org/SentinelEvents/SentinelEventAlert/sea_40.htm
- Kudielka, B., Kern, S.(2004). Cortisol day profiles in victims of mobbing (bullying at the work place): preliminary results of a first psychobiological field study. Journal of Psychosomatic Research 56 (2004). pages 149-150
- Leymann, H.(1996) The content and development of mobbing at work. European Journal of Work and Organizational Psychology, 5(2), pages 165-184
- Leymann, H.(1990). Mobbing and psychological terror at workplaces. Violence and Victims, 5. (2).
- McFarlane, A. (2010). The long-term costs of traumatic stress: intertwined physical and psychological consequences. World Psychiatry. volume: 9 issue: 1 pages: 3-10
- Martin, W. (2008). Is Your Hospital Safe? Disruptive behavior and workplace bullying. Hospital Topics: Research and Perspectives on Healthcare.Vol. 86, no. 3 Summer 2008
- Matthiesen, S., Einarsen, S. (2004). Psychiatric distress and symptoms of PTSD among victims of bullying at work. British Journal of Guidance & Counseling, volume 32, issue 3 . pages 335-356
- Miller, G., Chen, E., Zhou, E. (2006). If it goes up, must it come down? chronic stress and the hypothalamic-pituitary-adrenocortical axis in humans. Psychological Bulletin . Volume 133(1), January 2007, p 25-45
- Monteleone,P., Nolfe, G., Serritella, C. Milano, V, Di Cerbo, A et al. 2009. Hypoactivity of the hypothalamo-pituitary-adrenal axis in victims of mobbing: role of the subjects' temperament and chronicity of the work-related psychological distress. Psychotherapy and Psychosomatics. Volume: 78 Issue: 6. Pages: 381-383
- Murray, J. (2010). Workplace bullying in nursing: a problem that can't be ignored. MedSurg Nursing. Sept-Oct 2009.
- Nater, U., Rohleder, N. (2009). Salivary alpha-amylase as a non-invasive biomarker for the sympathetic nervous system: current state of research. Psychoneuroendocrinology. volume 34 pages 486 - 496.
- Niedhammer, I., David, S, Degioanni, S. 2007. Association between workplace bullying and depressive symptoms in the French working population. Journal of Psychosomatic Research 61 (2006). pages 251- 259
- Notelaers, G., Einarsen, S., De Witte, H., Vermunt, J. (2006). Measuring exposure to bullying at work: the latent class cluster approach. Work and Stress. volume 20 issue 4 pages 288 -301.
- Polla, E., Kreitschmann-Andermahra, I., Langejuergena, Y., Stanzelc, S., Gilsbacha, J., Gressnerb, A., Yagmurb, E. (2007). Saliva collection method affects predictability of serum cortisol. Clinica Chimica Acta. volume 382, issues 1-2, July 2007, pages 15-19.
- Rantonen, P., Pentilla, I., Meurman, J., Savolainen, K., Narvanen, S., Helenius, T. (2000). Growth hormone and cortisol in serum and saliva. Acta Odontologica Scandinavica. volume: 58 issue: 6 pages: 299-303
- Reul,J., Nutt, D.(2008). Glutamate and cortisol a critical confluence in PTSD? Journal of Psychopharmacology; 22. page 469.
- Rocker, C. (2008). Addressing nurse-to-nurse bullying to promote nurse retention. OJIN: The Online Journal of Issues in Nursing; Vol 13 No. 3.
- Simons, S. (2008). Workplace bullying experienced by Massachusetts registered nurses and the relationship to intention to leave the organization. Advances in Nursing Science. volume31, issue 2. pages 48-59.
- Tafet,G.(2001). Correlation between cortisol level and serotonin uptake in patients with chronic stress and depression. Cognitive, Affective, & Behavioral Neuroscience 1.4 (2001). pages 388-393
- Tracy, S.(2006). Nightmares, demons and slaves: exploring the painful metaphors of workplace bullying. Management Communication Quarterly. Volume 20 Number 2. pages 1-38
- Wessa, M., Rohleder, N., Kirschbaum, C., Flor, H.(2006). Altered cortisol awakening response in post traumatic stress disorder. Psychoneuroendocrinology. volume31. pages 209-215.