The Associations Between Workplace Bullying and Physical or Psychological Negative Symptoms: Anxiety and Depression as Mediators

Workplace bullying is a critical issue for its negative consequences on victims’ health and well-being. This study aimed at examining the intermediate roles of anxiety and depression, in the relations between workplace bullying as a predictor, and physical and psychological negative symptoms as outcomes. In particular, it was hypothesized that workplace bullying would be associated with higher anxiety and depression and, through them, with higher physical and psychological negative symptoms. We sampled 151 Italian employees, who called on a workplace bullying public clinical center as victims and filled a paper-and-pencil questionnaire. Results of bootstrapped regressions showed that only anxiety mediated the association of workplace bullying with negative physical symptoms while both anxiety and depression mediated its association with negative psychological symptoms. The results have implications for the development of appropriate intervention strategies for both prevention and clinical treatment. In particular, timely diagnosing and treating anxiety and depression could prevent subsequent problems related to psychological and physical symptoms such as colitis, headache, tiredness, nervousness, etc. Organizational interventions in terms of primary prevention are also discussed. From an empirical standpoint, the study contributed to disentangling the differential roles of anxiety and depression with respect to physical and psychological symptoms; moreover, overcoming a common limit of workplace bullying research, the current study was carried out on actual victims.

variation depending on how it is measured, in fact Nielsen, Matthiesen, and Einarsen (2010) found a general estimated prevalence rate for workplace bullying ranging from 11 to 18%, this variation due to the measurement method used. Moreover, with the use of meta-analytical evidence, the same authors estimated that at least 15% of the global working population was exposed to negative workplace behaviors that can be experienced as bullying.
Exposure to workplace bullying and its associations with health and well-being have been largely examined; among the others: psychological disorders, depression, burnout, anxiety and aggression, as well as psychosomatic and musculoskeletal disorders (Björkqvist, Österman, & Hjeltbäck, 1994;Einarsen & Raknes, 1997;Niedl, 1996;Zapf, Knorz, & Kulla, 1996). However, studies examining the reciprocal relations between clusters of variables consequent to workplace bullying have been largely overlooked. Building on this, this paper aims at examining the associations between workplace bullying as a predictor, anxiety and depression as intermediate variables, and psychological and physical negative symptoms as final outcomes. Disentangling the potential mediated effects from direct effects could be useful in targeting anxiety and/or depression through appropriate treatment among workplace bullying victims in case mediated effects were proved or else focusing on alternative treatments directly focused on physical and/or psychological negative symptoms if mediation was not proved. Additionally, while most studies had been usually carried out within organizational contexts with heterogeneous samples, data for this study have been collected in an Italian clinical center dedicated to workplace bullying victims, attesting over 20 years of clinical practice in terms of prevention and treatment, thus providing evidence on actual victims of workplace bullying as well as on an under-represented country in workplace bullying research as Italy.

Theoretical Foundation
"Bullying at work means harassing, offending, or socially excluding someone or negatively affecting someone's work" (Einarsen, Hoel, Zapf, & Cooper, 2011b, p.39). A single conflict can not be called workplace bullying if the incident is an isolated event or if two parties of approximately equal strength are in conflict (Einarsen, Hoel, Zapf, & Cooper, 2003). Leymann (1990Leymann ( , 1996 suggested, as a time criterion to characterize bullying as a severe form of social stress, that such events should occur at least once a week for more than 6 months. Others have used repeated exposure to negative behaviours within a 6-month period as the proposed timeframe (Einarsen & Skogstad, 1996). Hence it is clear that while many examples of interpersonal aggression take the form of single episodes, bullying in the workplace is characterized by a systematic and prolonged exposure to repeated negative and aggressive behavior also and more of a psychological nature, including non-behavior and acts of social exclusion (Einarsen et al., 2011b;Leymann, 1996). ageable, thus leading to the feeling of being unable to face the current situation (i.e., self helplessness). Therefore, the negative effects are triggered by repeated and chronic cognitive activation of a prolonged bullying process. Cognitive activation would also lead to a prolonged physiological activation which will subsequently manifest itself through the impairment of health and well-being (Ursin & Eriksen, 2004), decreased sleep quality, increased levels of cortisol, increased heart rate and increased mortality (Meurs & Perrewe, 2011;Ursin & Eriksen, 2004). More in detail, Nielsen and Einarsen (2012) differentiated between job-related outcome variables (attitudes and behaviour) and outcome variables related to health and well-being. With regard to this latter cluster, they identified several distinct variables: mental health problems (thus differentiating between anxiety and depression), physical health problems, post-traumatic stress disorder, burnout, sleep problems, strain, etc. Mikkelsen and Einarsen (2002) argued that "exposure to bullying behaviors at work may alter the victims' emotional state and their views of themselves and others, this in turn causing higher levels of reported psychological health complaints and psychosomatic complaints" (p. 398).
From the point of view of psychological symptomatology, Brodsky (1976) identified, in addition to the physical symptoms, both depressive reactions and symptoms related to depression (e.g., impotence, lack of self-esteem, insomnia), and reactions related to psychological symptoms (e.g., hostility, hypersensitivity, loss of memory, feelings of victimization, nervousness, avoidance of social contact). Based on this, in the present paper anxiety and depression were differentiated with respect to physical negative symptoms (e.g., high blood pressure, colitis, back pain, headache) and psychological negative symptoms (e.g., sleep problems, disinterestedness, loss of concentration, tension). As anticipated before, our aim was to differentiate between these outcomes, furthermore examining if anxiety and depression may act as intermediate variables thus mediating the effects of workplace bullying on psychological and physical negative symptoms. Available evidence and theoretical underpinning supporting our study hypotheses will be detailed in the next section.

Study Hypotheses
Available evidence suggests that bullying at work is a powerful stressor that affects, among other things, the health and well-being of individuals. According to Nielsen and Einarsen (2012), exposure to systematic and prolonged aggressive behaviors negatively impacts on the individual causing a cognitive activation and a subsequent physiological activation that, if prolonged, will lead to the impairment of health and well-being. In particular, given that each subject can have different consequences and react differently, leading to alterations in cognitive-emotional balance (e.g., depression, anxiety, obsessions, panic attacks, emotional anesthesia) as well as in the psychosomatic balance (e.g., gastrointestinal disorders, headache, dizziness, arterial hypertension, disorders of sexuality and sleep, etc.), it appears compelling to examine the differential effects on the above mentioned outcomes.
As for the cognitive-emotional effects, being a victim of systematic and intentional psychological harm by other individuals seems to produce severe reactions such as depression and anxiety, negative effects triggered, according to Nielsen and Einarsen's Model (2012), by repeated and chronic cognitive activation of a prolonged bullying process. In fact, among the different consequences of workplace bullying, higher levels of psychological disorders, not only anxiety and depression, but also burnout and aggression, have been found (Björkqvist et al., 1994;Einarsen & Raknes, 1997;Niedl, 1996;Zapf et al., 1996). Brousse et al. (2008) have found higher levels of stress and anxiety-depression disorder developed by targets of workplace bullying at 12-month postconsultation. Bilgel, Aytaç and Bayram (2006) found that workers who had reported bullying showed lower lev-Workplace Bullying and Well-Being 810 els of job satisfaction and higher levels of work-related stress, as well as higher scores of anxiety and depression than those who had not reported bullying. According to these authors, the significant association between bullying and subsequent depression suggests that bullying is an etiological factor for subsequent mental health problems. Therefore, it appears that being a victim of repeated adverse behaviors can lead to increased anxiety and depressive symptoms, given that the working environment, and in particular social relationships within it, may be perceived as threatening and unpredictable. Based on available evidence, we expect that: H1: Exposure to workplace bullying is positively associated with an increased presence of anxiety (a) and depression (b); As mentioned above, workplace bullying at work has been linked to impaired health and well-being. With regard to the latter, that is the abovementioned psychosomatic balance, previous research has focused not only on the increase in anxiety or depressive symptoms but has also on psychological and physical negative symptoms.
According to Djurkovic, McCormack, and Casimir (2003), workplace bullying was positively associated with physical symptoms, negative affect, and with intention to leave the job. Their results showed that bullying led to negative affect which in turn led to physical health problems. As already mentioned, according to the Nielsen and Einarsen's Model (2012), exposure to systematic and prolonged aggressive behavior causes a physiological activation that if prolonged triggers a series of physical and psychological negative symptoms. More than 40 years ago, Brodsky (1976) identified three patterns of effects on the victims. Some subjects developed vague physical symptoms, such as pains, loss of strength, chronic fatigue, weakness and various aches. Some other subjects reacted with depression and depression-related symptoms (e.g., impotence, lack of self-esteem, and sleeplessness), while others subjects, instead, reacted with psychological symptoms, such as hostility, hypersensitivity, loss of memory, feelings of victimization, nervousness, and avoidance of social contact. As mentioned above, experiencing an unmanageable cognitive dissonance between what is expected and what actually happens is associated with persistent levels of stress and disease, such as decreased sleep quality, increased levels of cortisol, increased heart rate and increased mortality (Meurs & Perrewe, 2011;Ursin & Eriksen, 2004). Thus, it is likely that individuals, apart from more specific anxiety and depressive symptoms, will develop other more general negative symptoms related both to the physical and psychological domain. Therefore, it is expected that: H2: Exposure to workplace bullying is positively associated with an increased presence of physical (a) and psychological (b) negative symptoms; This paper aimed not only at testing direct associations between workplace bullying and the abovementioned outcome variables (i.e., depression, anxiety, physical and psychological negative symptoms), but it is proposed that anxiety and depression may act as intermediate variables between workplace bullying and physical and psychological negative symptoms. Besides anxiety and depression, other common symptoms such as musculoskeletal disorders have been reported by victims in several European countries as a result of bullying at work (Einarsen & Raknes, 1997;Niedl, 1996;O'Moore, Seigne, McGuire, & Smith, 1998;Zapf et al., 1996). Brousse et al. (2008) have found that many patients, at 12-month post-consultation, reported various persistent somatic symptoms or disorders including: appetite and digestive disorders, weight gain, palpitations, migraines, giddiness and poorly systematized arthralgia or muscular pain. Trivedi (2004) found that a high proportion of patients with depression seeking treatment in a primary care setting had only physical symptoms, in fact physical pain and depression would seem to have a biological connection deeper than the simple cause and effect (see  O'Neil, 2013). Kivimäki et al. (2003) found a higher risk of cardiovascular disease due to prolonged bullying targets and a higher risk of developing depressive symptoms. Moreover, Henningsen, Zimmermann, and Sattel (2003) found, through a meta-analytical method, a moderate and significant association between unexplained physical symptoms, anxiety and depression. Other scholars have also pointed out that experience of depression (Beekman et al., 2002) and anxiety (De Beurs, Beekman, Van Balkom, & Deeg, 1999) can lead to subsequent somatic and psychological complaints. From the abovementioned evidence, it appears that depression and anxiety on one side, and physical and psychological negative symptoms on the other are related, with the former preceding and activating the latter (Nielsen & Einarsen, 2012). Thus, it is expected that: H3: Anxiety will mediate the association between workplace bullying and negative physical (a) and psychological (b) negative symptoms; H4: Depression will mediate the association between workplace bullying and negative physical (a) and psychological (b) negative symptoms;  In order to evaluate the significance of direct and indirect effects, bootstrapping procedure with 5,000 samples with replacement from the full sample to construct bias-corrected 95% confidence intervals (hereafter 95% CI; LL = lower level of confidence interval, UL = upper level of confidence interval) was used (Preacher & Hayes, 2008). Any effect is statistically significant when zero is not included in the CI. Continuous variables were standardized before calculating regression models.
Two regressions with boostrapping replacement were carried. In the first one, physical negative symptoms was regressed on control variables, predictor and mediators. In the second one, the outcome variable was psychological negative symptoms (Table 2).

Discussion
The present paper, consistently with Nielsen and Einarsen's Model (2012), examined the intermediate role of anxiety and depression, in the relations between workplace bullying as a predictor, and physical and psychological negative symptoms as outcomes. In particular, we examined whether workplace bullying was associated with higher anxiety and depression and, through them, with higher physical and psychological negative symptoms. Results showed the severity of the effects of bullying on mental health, which lead to more or less serious conditions of psychological/psychiatric impairment even in subjects without a significant psychiatric history.
First, we hypothesized (H1) that exposure to workplace bullying was positively associated with an increased presence of anxiety and depression. Results fully supported this hypothesis: workplace bullying was positively associated with anxiety and depression. Similar findings have been reported by Einarsen (2012), Lo Presti, Pappone, &Landolfi 815 Brousse at al. (2008), Bilgel et al. (2006). So, exposure to systematic and prolonged aggressive behaviors could be an etiological factor for subsequent mental health problems, as anxiety and depression.
Second, given that many scholars have stressed the importance of bullying in the workplace as a predictor of physical and psychological negative symptoms, we hypothesized (H2) that exposure to workplace bullying was positively associated with an increased presence of physical and psychological negative symptoms. This hypothesis was fully supported showing that physical and psychological negative symptoms were positively predicted by workplace bullying, confirming the critical role of bullying in the onset of somatic/physical and psychological symptomatology. There results are in line with available evidence, therefore, exposure to systematic and prolonged aggressive behavior may lead to a physiological activation that triggers a series of physical and psychological negative symptoms, as confirmed by Nielsen and Einarsen's Model (2012). It derives that beyond the anxious and/or depressive symptomatology, other negative effects of workplace bullying may be found: vague physical symptoms (such as pain, loss of strength, chronic fatigue, weakness and various pains), impotence, lack of self-esteem, insomnia, hypersensitivity, loss of memory, feelings of victimization, nervousness and avoidance of social contact, decreased quality of sleep, increased heart rate.
Finally, we hypothesized (H3) that anxiety may mediate the association between bullying at work and physical (a) and psychological (b) negative symptoms and (H4) that depression will mediate the association between workplace bullying and physical symptoms (a) and negative psychological negative (b) symptoms. Results showed that only anxiety mediated the effects of workplace bullying on physical negative symptoms while, both anxiety and depression were mediators with respect to psychological negative symptoms. So, hypothesis H3 was fully confirmed: anxiety mediated the effect of bullying on both physical (H3a) and psychological (H3b) negative symptoms. Hypothesis H4 was partially confirmed given that depression mediated only the effect of bullying on psychological negative symptoms (H4b), while it did not mediate the effect of bullying on physical negative symptoms (H4a). It derives that apart from being directly associated with psychological and physical negative symptoms-at least partially-anxiety and, partly, depression represent intermediate variables through which workplace bullying potentially lead to impaired physical and psychological well-being.

Conclusions
Workplace bullying is a psychosocial risk that can significantly impair individual well-being (Eurofound, 2017), but can also negatively impact on the organization through increased absenteeism (O' Connell, Calvert, & Watson, 2007) and turnover (Hoel & Cooper, 2000), as well as decreased productivity (Hoel, Sparks, & Cooper, 2001). Additionally, it has been proven that it has significant financial costs for the welfare system (Beswick et al., 2006;Kivimaki et al., 2000). It derives that acquiring a better knowledge of this phenomen, its antecedents as well its consequences (at different levels), may be useful in order to prevent it and/or effectively target victimization processes and their effects.
This paper contributed to the literature providing additional evidence about the negative effects of workplace bullying on depression and anxiety (Bilgel et al., 2006;Brousse et al., 2008), as well as on physical and psychological negative symptoms (Djurkovic et al., 2003). Moreover, it was proved that the effects of workplace bullying on physical and psychological negative symptoms were mediated by anxiety and, partly, by depression. The high percentage of variance of both outcomes explained by predictors supports the salience of these Workplace Bullying and Well-Being 816 factors in explaining variability in physical and psychological negative symptomatology. From a practical standpoint, it derives that timely diagnosing and treating anxiety and depression could prevent subsequent problems related for instance to increased blood pressure, pain, colitis, headache, tiredness, decreased quality of sleep, hypersensitivity, loss of memory, etc. Additionally, overcoming a common limit of workplace bullying researchthat is participants not being victims of actual bullying-the participants of this study were all recruited within a public clinical center for bullying treatment and prevention, being actual victims of workplace bullying.
Our results may be useful in disentangling the intertwinement between physical and somatic symptoms, and depression and anxiety. First of all, these results may contribute to overcoming the prejudice related to the exclusivity of the "non organic" treatment; in clinical terms, it means stressing the importance of avoiding concentration, both in terms of prevention and treatment, exclusively on the organic/medical dimension but to highlight both the somatic and psychological dimension in the assessment and treatment of such disorders. Therefore, differentiating the effects on anxiety and depression from the effects on negative physical and psychological symptoms respectively, could be useful for addressing victims' anxiety and/or depression, as well as more specific physical and/or psychological symptomatology, through appropriate and specific treatments. From a clinical point of view, having detailed information regarding the symptomatology could allow differentiating or not a possible pharmacological intervention in the treatment phase.
Although it was not the focus of the present study, significant organizational implications can be drawn from our results. In particular, consistently with available literature (Salin, 2003), the association between workplace bullying and anxiety/depression and psychological/physical negative symptoms should represent a further warning to employers to promote adequate organizational conditions that could reduce the likelihood that adverse behaviors and conflict could give rise to workplace bullying. Primary preventive interventions should target organizational culture and climate, work organization and job design, workgroups' functioning and leadership effectiveness, reward systems and competition, just to mention a few among the main ones (Salin & Hoel, 2010).
This study suffers from a series of limitations that need to be acknowledged. A critical aspect concerns the use of a cross-sectional research design that does not allow for causal inferences; future longitudinal studies, for instance diary studies, may better examine the evolution of the phenomenon over time and, for instance, disentangle the direction of causality between anxiety/depression and physical/psychological negative symptoms. Furthermore, we only used self-report measures, therefore the relationship between the variables could be influenced by common method variance. Although the present study provides a significant and innovative contri-