Sense of Coherence as a Determinant of Psychological Well-Being Across Professional Groups of Aid Workers Exposed to War Trauma

Abstract

The present study aims to test whether sense of coherence (SOC) acts as a determinant of positive psychological functioning in aid workers directly exposed to warfare. Specifically, we performed multiple regression analyses to compare different groups of aid workers in terms of the effects of SOC and cumulative trauma on their psychological distress. Palestinian helpers, both professional and non-professional (N = 159) completed three self-reported measures: the General Health questionnaire, Sense of Coherence Scale, and Impact of Events Scale. The findings bear out the predictive power of SOC and posttraumatic stress disorder (PTSD) in relation to mental health across different professional groups. In particular, volunteers without a specific professional profile, psychiatrists, medical doctors, and less markedly counselors seemed to protect their mental health through a SOC. Clinical implications and recommendations for training and supervision are discussed.

Keywords

war PTSD vicarious trauma

Humanitarian workers are commonly exposed to chronic levels of stress caused by having to deal with injury, mutilation, and even death while working in insecure and dangerous environments (Naudè & Rothmann, 2006; Niţă & Petre, 2012). It has been demonstrated that aid workers are vulnerable to the development of posttraumatic stress symptoms (PS) as a natural reaction to the stress incurred while assisting traumatized and distressed individuals (Mitchel & Dyregov, 1993). The best-known sequelae of stressful events are burnout, psychosomatic disorders, depression, and risk-taking behaviors such as substance abuse and suicide (Veronese, Fiore, Castiglioni, el-Kawaja, & Said, 2012). Prolonged exposure to stress is a key factor contributing to the development of PS (Corneil, 1993). A study conducted by Ravenscroft (1994) at the London Ambulance Service showed that job stress was the leading cause of sickness among helpers, 15% of whom reported symptoms of posttraumatic stress disorder (PTSD). Occupational stress in extreme conditions can lead to acute or prolonged stress disorders (Mitchel & Dyregov, 1993). A number of studies have explored the psychological consequences of stressful and traumatic missions or levels of stress disorder among aid workers (Andersen, Christensen, & Petersen, 1991; Korff, Balbo, Mills, Heyse, & Wittek, 2015; Thormar et al., 2015). Prevalence rates of PTSD in rescue workers have been estimated to range from 3% to 24.5% (Lundin, 1997; Wagner, Heinrichs, & Ehlert, 1998). Some experiences are found to be almost invariably traumatic, for example, car accidents, rape, violence, and war, with victims of these kinds of trauma often developing an acute traumatic stress disorder (ASD) as a precursor of PTSD. Morbidity survey showed ASD vary from 2% to 21% depending on the nature and severity of the traumas (Fear & Wessely, 2015). These acute pathological reactions to trauma are associated with dissociative phenomena that act both in the short- and long-term reactions to trauma.

Helpers involved with victims of trauma are at risk to develop both psychological and physical symptoms (Jonsson, Segesten, & Mattsson, 2013). Selley (1991) showed that along with primary victims, health workers are also at risk of developing PTSD.

The construct “secondary trauma” emerged from Charles Figley’s (1995, 2002) work on the manifestation of PTSD among the contacts of traumatized individuals, such as family members. Secondary traumatic stress symptoms typically include the same types of distress that characterize PTSD, including avoidance, arousal, and intrusive symptom clusters (Jordan et al., 1992). Figley (1995) viewed secondary trauma as arising from an empathic relationship with the traumatized individual, whereby the victim’s close contact is also affected by the trauma and manifests symptoms very similar to those of PTSD.

Nonetheless, some aid workers, despite demanding job requirements and challenging working conditions, display no symptoms of stress and burnout (Schaufeli & Bakker, 2001). On the contrary, they appear to enjoy working hard and dealing with dangers and extreme demands, showing strong engagement in their mission and feelings of personal growth and development (Seligman & Csikszentmihalyi, 2000).

Similarly, people who have been exposed to war have also been found to unexpectedly maintain good functioning and health. The construct of sense of coherence (SOC) developed by Antonovsky (1987) is used to explain why some individuals are better able than others to cope with experiencing traumatic events, suffering, and illness. SOC is defined as an overall orientation of the individual that determines the extent to which he or she maintains a feeling of confidence in the face of adversity. Individuals with a strong SOC perceive the events in their lives as structured, predictable, and explicable and are able to find the resources required to cope with traumatic experiences. As more fully expressed by Antonovsky (2000), SOC is

a global orientation that expresses the extent to which one has a pervasive and enduring though dynamic feeling of confidence that (1) the stimuli deriving from one’s internal and external environments in the course of living are structured, predictable, and explicable; (2) the resources are available to one to meet the demands posed by the stimuli; and (3) these demands are challenges worthy of investment and engagement. (p. 41)

In line with his Salutogenic Model, Antonovsky (2000) suggested that SOC is essential to maintaining health and preventing physical and mental breakdown. People with a strong SOC display the ability to experience stressors as manageable, meaningful, and comprehensible. They thus maintain a higher level of functioning than other people experiencing the same stressful events (Antonovsky, 1987, 1996, 2000). Individuals with a strong ability to make sense of the circumstances they need to cope with will experience a stressful event as a positive challenge that merits emotional investment and commitment. The availability of generalized resistance resources (GRRs) for coping with tension and stressors such as the perception that violent episodes potentially could be controlled depends on an individual’s overall SOC (Hochwälder, 2015). The extent of available GRR is the key to determining whether a stressful situation will result in a weakening or strengthening of the person’s SOC (Antonowsky, 1997). SOC is significantly related to self-rated health—the higher an individual’s SOC, the better their self-rated health—and SOC has been proposed as an indicator of positive mental health. Furthermore, a systematic review of SOC and its correlation with quality of life (QOL), has found a stronger SOC to be associated with superior QOL, whereas longitudinal studies have found SOC to predict satisfactory QOL (Eshel & Mashbood, 2014). Although evidence from systematic reviews of the effect of SOC on individual health demonstrates that SOC is positively associated with good health (e.g., a strong SOC appears to be a determinant of satisfactory psychological health; Eriksson & Lindström, 2005), there is still a paucity of detailed theoretical knowledge on the factors influencing such relationships in the context of warfare.

Within the field of the psychological and clinical sciences, research has sometimes focused on professional helpers given their ongoing exposure to the tragic and traumatic events affecting their clients. While it is known that any trauma exposure can trigger stress disorders (Jonsson, Segesten, & Mattsson, 2013), recent research on people’s reactions to differential stressful conditions indicates that the relationship between traumatic events and their effects on people may be better understood by taking into account protective factors buffering the impact of events in terms of direct (cumulative traumas) or indirect traumatization (e.g., psychological distress, anxiety, depression; Richardson & Ratner, 2005). It has been suggested that SOC is one such protective factor, which may mitigate the effects of traumatic experience on people’s health. In fact, individuals with a strong SOC tend to use personal coping strategies that enable them to react more appropriately to stressful situations (Petersen, Ladelund, Carlsson, & Nilbert, 2013; Tomotsune et al., 2009; Veronese et al., 2012).

In the present study, we aimed to test whether SOC acted as a determinant of positive psychological functioning in aid workers directly exposed to war contexts. Specifically, we started out from two well-known standpoints close to our research and practical interests. First, a strong interrelationship between experience of trauma and psychological distress has been identified in numerous different contexts and across different samples (Oumette, Wade, Prins, & Schohn, 2008; Taubman-Ben-Ari, Rabinowitz, Feldman, & Vaturi, 2001). Arguably, people living and working in contexts of low-level warfare are likely to experience trauma that in turn negatively affects their health status by increasing their psychological distress. Second, as outlined above, SOC represents a means for individuals to reduce the effects of stressful situations on their health. Consequently, on one hand, we might reasonably expect traumatic situations to increase psychological distress by forcing aid workers to deal with the consequences of war for both their own lives and those of other people. On the other hand, we might also expect a strong SOC to mitigate psychological distress by providing a framework within which aid workers can situate and attribute coherence to these traumatic situations. Our main research aim was to quantify the effects of SOC and cumulative trauma, as well as of their interaction, on psychological distress in a sample composed of four types of aid worker operating in refugee camps. Previous studies have tested the role of SOC as a determinant of psychological distress in various types of aid and emergency workers, such as ambulance drivers, social workers, and psychotherapists (Binder, Mensenholl-Strhler, Paß, & Endler, 2006; Gilbar, 1998; Jonsson et al., 2013). As far as we are aware, no studies to date have reported a comparison between workers from different helping professions in respect of their ability to attribute meaning to traumatic experiences. To address the lack of data in this area, we performed multiple regression analyses testing a range of hypotheses regarding how different groups of aid professionals might vary in terms of the levels of SOC and cumulative traumas on their psychological distress. Specifically, we set out to test the following hypotheses:

Hypothesis 1 (H1): Intrusion and avoidance would contribute to increasing psychological distress.

The most common responses to high levels of stressful inputs include intrusive memories (criterion B1 of Diagnostic and Statistical Manual of Mental Disorders [5th ed.; DSM-5; American Psychiatric Association, 2013] PTSD diagnosis) and avoidance of trauma-related thoughts and feelings (criterion C1; Gittings, Paterson, & Sharpe, 2015). In this sense, we expected that health workers operating in war affected areas show intrusion and avoidance symptoms accordingly with exposure to traumatic war events (Flynn, McCarroll, & Biggs, 2015). Moving from this perspective we hypothesized that,

Hypothesis 2 (H2): Regression intercept values would be equal across all groups of professionals.

In a recent article published by Veronese, Pepe, and Afana (2016), the authors found that SOC mediate the impact of trauma on psychological well-being. Specifically, this research pointed out that SOC mediated the relationship between levels of anxiety, social dysfunction, and loss of confidence, levels of trauma with a prevalence of avoidance symptoms. Accordingly, we tested if an analogue model is replicable in different health professionals working in war-like environments. Namely,

Hypothesis 3 (H3): SOC would mitigate levels of psychological distress.

Hypothesis 4 (H4): All groups of aid professionals would have similar regression slopes.

We should point out here that we modeled the relationships among psychological distress, SOC, and experience of cumulative traumas not just for the sake of modeling a theory, but primarily because the psychological health of aid workers operating in war context is constantly jeopardized by traumatic experiences. Advancing understanding of whether and how such variables work together to affect emergency workers’ health status should be considered one of the leading duties of the social researchers who work alongside them in the different war zones. In the concluding sections of the article, we discuss the limitations of our study as well as its implications for clinical work in emergency settings characterized by war and political violence.

The Study Context

Some background data are required to provide a glimpse of the milieu in which this study was conducted as well as a basis for evaluating the transferability of the research conclusions to other conflict situations.

Tulkarm City is a town in the West Bank with two refugee camps (Tulkarm City and Nurshams) hosting roughly 27,000 “internal refugees.” The population is extremely poor as a result of the closure of the borders after the second Intifada of 2002 (Al-Aqsa Intifada; Esposito, 2005). Helpers’ work, QOL, and well-being are compromised by the precarious living conditions: territorial discontinuity, the separation wall between Israel and the Occupied Palestinian Territories, military checkpoints, and curtailing of basic resources. Recent years have been characterized by a form of low-intensity conflict involving night-time incursions, imprisonments and targeted murders, mobile checkpoints, and surprise curfews.

Gaza City is a highly populated city in the Gaza Strip that is controlled by the political faction, Hamas. “High-intensity” conflict has come on top of chronic poverty and unemployment caused by the trade embargo and total border blockade that have been progressively enforced since 2005 in opposition to the rise to power of Hamas. Two large-scale refugee camps (Jabalia and Al-Shati Camps, with a joint population of approximately 180,000 people), and other smaller camps, make Gaza one of the areas with the highest concentration of inhabitants in the world.

Method

Participants

Participants were recruited from a number of public hospitals at five different refugee camps: Tulkarm (18,000 officially registered refugees; 39.4% of the total refugee population in the West Bank), Gaza (478,854; 25.7%), Jabalia (108,000; 6%), Khan Younis (68,000; 5.3%), and Rafah (99,000; 3.6%). Given that the research project involved physicians, psychiatrists, and counselors (as defined by the Standard Occupational Classification System, U.S. Bureau of Labor Statistics, 2010), as well as community volunteers, recruitment took place on-site and the interviews were conducted during working hours. The participants were selected to cover the range of helper roles in the Palestinian care system and to obtain representative proportions of the above-mentioned professions. Their professional experience in the field of health care ranged from 10 to 30 years. All of them declared to have witnessed during the past 3 months at least an episode of war or armed fight. The resulting sample was composed of Palestinian helpers, both professional and non-professional (N = 159).

Gender distribution was 85 males (53.5 %) and 74 females (46.5 %). Mean age was 29.13 years (SD = 8.25, min–max = 18–59). The age values at the 25th, 50th, and 75th percentiles were 23, 27, and 33 years, respectively. With regard to occupational role, 25 participants were physicians (15.7%), 46 were counselors (28.9%), and 20 were psychiatrists (12.6). A further 68 participants worked as volunteers (42.8%) providing non-specific support to care professionals during health procedures. Details of participants’ socio-demographic characteristics by occupational group are reported in Table 1. There was found a small but statistically significant association, χ2(3) = 10.51, p = .015, φ = .258, between gender distribution and occupation. The group of physicians had a strong prevalence of males (80%), whereas the male:female ratio was close to equal in the other groups. Age differences among the categories were rather more substantial, F(3, 156) = 14.38, p < .001: The physicians reported a mean age value of 36.3 years, which was significantly higher than those of the other groups. Conversely, the group of volunteers was the youngest (25.63).

Table 1.

Socio-Demographic Characteristics of the Sample by Occupational Category.

	Male	Female	Age
	n (%)	n (%)	M (SD)
Physicians	20 (80)	5 (20)	36.38 (9.21)
Psychiatrists	9 (45)	11 (55)	28.30 (4.33)
Counselors	22 (47.8)	24 (52.2)	31.27 (7.82)
Volunteers	32 (47)	36 (53)	25.63 (7.04)

The research design was based on the use of standardized self-report questionnaires following a pencil-and-paper procedure. The research was conducted in line with American Psychological Association (APA, 2010) ethical principles and code of conduct, and approved by the Ethics Committee of Milano-Bicocca University.

Instruments and Procedure

General Health Questionnaire (GHQ)

The GHQ (Goldberg, 1972) is a “family” of a quick and reliable screening tools aimed at “detecting psychiatric disorders among respondents in both community and non-clinical settings” (Goldberg & Williams, 1988, p. 1). The GHQ-12 is the short version of the originally developed instrument (GHQ-30), which has been often used in the context of large-scale social surveys (e.g., by the World Health Organization) as a reliable measure of psychological distress. The current research adopted the Arabic version of the GHQ-12 (Daradkeh, Ghubash, & El-Rufaie, 2001) to assess levels of psychological distress among different groups of professional and non-professional helpers. The results of an earlier study indicated that in the Arab context, the GHQ-12 may be used as a unidimensional scale with a score ranging from 0 to 36, with higher scores reflecting lower psychological distress. In the present study, Cronbach’s alpha value and the mean of inter-item correlations were .885 and .395, respectively (Veronese & Pepe, 2013).

Impact of Event Scale

The Impact of Event Scale—Arabic Revised version (IES-R-13) is a self-report measure specifically developed for assessing PTSD in adult populations living in contexts of ongoing military violence (Veronese & Pepe, 2013). The IES-R-13 is a short version of the original Impact of Event Scale (Horowitz, Wilner, & Alvarez, 1979), which may be used as a screening tool to assess the impact of stress on individuals affected by traumatic experiences. In the Arabic context, the recommended measurement model comprises two main dimensions (avoidance and intrusion/hyperarousal). This structure has been found to be particularly useful in research focused on the theoretical aspects of the impact of stressful situations on subjective well-being (Veronese & Pepe, 2013): For both subscales, higher scores reflect higher levels of impact of trauma. In the present study, Cronbach’s alpha values for avoidance (α = .801) and intrusion/hyperarousal (α = .712) were acceptable, and the means of inter-item correlation were .308 and .372, respectively.

Sense of Coherence Scale (SOC-29)

The SOC-29 (Antonovsky, 1979) is an empirical tool that measures an individual’s orientation and internal strengths with respect to coping with stressful life events (Jakobsson, 2011). From a theoretical viewpoint, the SOC construct was developed within the salutogenic (as opposed to pathogenic) paradigm and reflects a person’s capacity to respond to stressful situations. However, from an operational point of view, SOC should be viewed as a dispositional orientation rather than a set of coping strategies (Antonovsky, 1993): More specifically, it appears to be a strong and valid determinant of effective coping strategies and, consequently, of good mental health (Eriksson & Lindström, 2005). In the current research, participants completed the Arabic version of the SOC questionnaire (O. Cohen & Savaya, 2003). This scale contains items evaluating participants’ sense that the world is comprehensible, manageable, and meaningful on a 7-point Likert-type scale. In line with the findings of earlier studies (Antonovsky, 1979; Antonovsky, Sagy, Adlet, & Visel, 1990), the scores were taken to express a unidimensional scale. Cronbach’s alpha value and the mean of inter-item correlations for the SOC-29 were .737 and .338, respectively.

To correctly classify participants into homogeneous groups of health workers, the present study adopted the Standard Occupational Classification System (U.S. Bureau of Labor Statistics, 2010), a world-wide classification system of professional occupations in different work settings, as a grouping criteria. This classification system covers all situations in which work is performed for pay or profit. Occupations are grouped according to both type of work carried out and the skills, education, and training required to carry it out with competence. On the basis of the classification system, three major categories were included in the current research design:

Physicians and surgeons (root code 29-1060)

The category includes anesthesiologists (29-1061), general practitioners (29-1062), internists (29-1063), obstetricians and gynecologists (29-1064), pediatricians (29-1065), and surgeons (29-1067). That is to say, all professionals who diagnose, provide surgical (invasive, minimally invasive or non-invasive) and non-surgical treatment, and help prevent diseases and injuries that commonly occur in the general population.

Psychiatrists (root code 29-1066)

The category includes professionals who diagnose, treat, and help prevent disorders of the mind.

Counselors (root code 21-1010)

This is a broad occupation including educational, guidance and school counselors (21-1012), family therapists (21-1013), substance abuse counselors (21-1011), mental health counselors (21-1014), and rehabilitation counselors (21-1019). Thus, this particular group of professionals counsel and advise individuals regarding a wide array of domains and issues.

Finally, in keeping with the Standard Occupational Classification system, volunteers were allocated to a fourth and separate group rather than being included in the three aforementioned categories. This last category was made up of community helpers who did not receive any economic compensation for providing non-specific support to other professionals but were constantly involved in the everyday activities of standard health care settings. The volunteers are a diverse group of paraprofessional community members, some trained as primary intervention volunteers and others that joined in response to the war.

Data Analysis

To answer the research questions, a series of hierarchical multiple regressions were conducted using a five-block strategy with enter method (see Pedhazur, 1997 for further information). Multiple regression techniques are a commonly used and flexible approach to data analysis that facilitates examination of the relationships between a criterion variable (often called target or dependent variable) and one or more other variables (called determinants, predictors, or independent variables), with or without controlling for the effect of other variables (usually participants’ socio-demographic characteristics; J. Cohen, Cohen, West, & Aiken, 2003). In our own case, regression equations were used to assess whether and to what extent scores on the intrusion and avoidance and SOC scales explained outcomes on the GHQ-12 measure of psychological distress, after controlling for the effect of other variables. Given that dissimilar distribution of socio-demographic characteristics between groups (as in the present study) should be taken into account to obtain “controlled for” findings (Cavallera, Passerini, & Pepe, 2013; Pepe & Addimando, 2014), at the first step in our regression analysis, the exogenous variables age and gender were entered as a single block.

At Step 2, the avoidance and intrusion/hyperarousal scales were included to evaluate their contribution to explaining overall variance in GHQ-12 scores. At Step 3, the variable occupational group was added to test the hypothesis that psychological distress was a function of occupational category and that, as a consequence, the level of intercept would not be the same among groups. The variable was dummy-coded, with the category volunteers (n = 68) serving as the reference group (baseline). A statistically significant model at Step 3 would mean that disjoined regression equations could be adopted to explore levels of psychological distress in the four different occupational groups.

At Step 4, SOC scores were added to test whether inclusion of the construct SOC explained additional variance in psychological distress. At the fifth step, the multiplicative terms of intrusion, avoidance, and SOC × Occupations were entered to further explore between-group differences. In this case, the hypothesis was that the regression slopes of the different groups would be parallel to one another. At this point, it was possible to compute the intercept values and standardized partial regression coefficients of the solved regression equation for each of the occupational groups. The corresponding final equation for the model is

Yi = α \pm B1 X1 (intrusion / hyperarousal) \pm B2 X2

(avoidance) \pm B3X3 (sense_coherence)

\pm γ 1 D 1 (psychiatrist) \pm γ 2 D 2 (counsellors) \pm

γ 3 D 3 (physicians) \pm δ i (XiDi) + ε i,

where ± is the intercept, $B_{1 -} B_{2 -} B_{3}$ are the effects of intrusion, arousal, and SOC on psychological distress, $γ_{1 -} γ_{2 -} γ_{3}$ are the effects of occupation on levels of psychological distress, and $δ s$ are the effects of multiplicative terms on the slope of the relationships between the criterion variable and its basic determinants. Table 2 summarizes the research hypotheses in relation to the multiple regression parameters.

Table 2.

Hypotheses Tested in Relation to GHQ-12 Scores.

Effect	Null Hypothesis
Intrusion (h₁)	β₁ = 0
Avoidance (h₁)	β₂ = 0
Occupational groups with same intercepts (h₂)	γ_i = 0
SOC (h₃)	β₃ = 0
Occupational group with parallel regression slope (h₄)	δ_ij = 0

Note. SOC = sense of coherence.

At each step, the regression equations were evaluated in terms of statistically significant variation in the coefficient of determination (R²). The models were also assessed in terms of unstandardized beta weights (B) and the corresponding statistical significance for each variable. This procedure was repeated separately for each of the occupational groups in the research design to obtain distinct effects and equations. All variables included in the procedure were centered, as required to test for interactive effects and enhance the comparability of beta weights (Aiken & West, 1991; Echambadi & Hess, 2007). Regression assumptions (e.g., homoscedasticity, multivariate normality) were checked and found to be fulfilled for all variables. Finally, a Mahalanobis’ distance criterion of p < .001 was adopted to identify and skip multivariate outliers.

Results

The results are presented in two sections covering general descriptive statistics/zero-correlations and the results of the multiple regression analyses, respectively.

Table 3 reports means, standard deviations, asymmetry, and kurtosis for each of the four variables, as well as the standard Pearson correlations calculated to evaluate whether each of these variables was related to psychological distress and consequently whether it was appropriate to include it in the subsequent regression analysis.

Table 3.

Descriptive Statistics and Zero-Order Correlations.

	(A)	(B)	(C)	(D)	M	SD	Sk.	Ku.
Psychological distress (GHQ-12) (A)	—				12.42	6.76	.470	−.636
Avoidance (IER-R-13) (B)	.225**	—			30.69	6.61	.137	−.626
Intrusion/hyperarousal (IES-R-13) (C)	.415***	.540***	—		2.02	1.31	.365	−.729
SOC (D)	−.470***	−.117	−.277**	—	3.78	2.17	.324	−.447

Note. GHQ = General Health Questionnaire; IES-R-13 = Impact of Event Scale–Arabic Revised version.

p < .05. **p < .01. ***p < .001.

Correlation values were generally found to be high, statistically significant, and in the expected theoretical directions. Psychological distress was positively correlated with avoidance (r = .225) and intrusion/hyperarousal (r = .415), whereas it was negatively correlated with SOC (r = −.470). In addition, the SOC scale was negatively correlated with intrusion/hyperarousal (r = −.277). Means and standard deviations of all study measures for the individual occupational groups are reported in Table 4.

Table 4.

Means and Standard Deviations for All Measures as a Function of Occupational Category.

	Physicians		Psychiatrists		Counselors		Volunteers
	M	SD	M	SD	M	SD	M	SD
Psychological distress (GHQ-12)	12.46	6.28	7.50	4.22	12.89	7.15	13.89	6.83
Avoidance (IER-R-13)	2.67	1.13	1.14	0.96	2.35	1.35	1.89	1.28
Intrusion/hyperarousal (IES-R-13)	5.27	1.95	1.66	1.40	3.63	2.31	3.92	1.85
SOC	31.73	7.20	33.25	6.35	30.85	7.38	29.61	5.68

Note. GHQ = General Health Questionnaire; IES-R-13= Impact of Event Scale–Arabic Revised version; SOC = Sense of Coherence.

The results of the hierarchical regression with dummy variables for GHQ-12 yielded the outcomes reported in Table 5. In the first model, neither age (β= −.004, p = ns) nor gender (β= −.095, p = ns) had a statistically significant effect, F(2, 138) = .614, p = ns. The regression model at Step 2, F(4, 135) = 3.66, p < .001, R² = . 12, supported the idea that the variable intrusion/hyperarousal (β = .375, p < .001) was a determinant of psychological distress with a negative direct effect: Higher levels of intrusion corresponded to higher levels of psychological distress. On the contrary, the result did not confirm a direct effect of the avoidance scale (β = .020, p = ns) on psychological distress. When the variable occupational group was added (Model 3), the model was statistically significant, F(7, 133) = 6.26, p < .001, R² = . 25, with an increase of 13% in explained variance. In particular, the intercept value for the group of psychiatrists (β = −.360, p < .001) was significantly lower than the baseline value and the values for other groups of workers, whereas physicians (β = −.101, p = ns) and counselors (β = −.086, p = ns) had similar intercept values to the baseline group. Inclusion in the model of SOC scores (Step 4) led to an additional 13% explained variance, with a global R² value of .376. The model was statistically significant, F(8, 132) = 9.94, p > .001, displaying a statistically significant increase in R² value (ΔR² = .13), F(1, 132) = 27.13, p < .0001. This result suggests that, in general, SOC (β = −.413, p < .001) influenced levels of psychological distress by acting as a functional protection “dynamic” in relation to stressful events (H3), whereas intrusion/hyperarousal might be more usefully viewed as a carrier. Across different groups, greater intrusion/hyperarousal corresponded to greater psychological distress but, at the same time, the stronger the SOC the less psychological distress. It is not surprising that when SOC was entered in the regression model, the direct effect of intrusion/hyperarousal (β = .211, p = ns) on levels of psychological distress was reduced, given that SOC mediates the effect of trauma on psychological distress (Veronese & Pepe, 2014) mitigating its negative consequences. Finally, the last regression model tested the hypothesis of parallel regression slopes (H4) across the various categories of aid worker. Hierarchical analysis, F(17, 123) = 5.87, p < .001, showed that the multiplicative effect (Determinants × Occupational group) did account for statistically significant additional variance (R² = .45, ΔR² = .077), F(9, 123) = 3.72, p < .05, supporting its inclusion in the model. Specifically, analysis of the standardized beta weights revealed that, when the relations among trauma, SOC, and psychological distress were modeled, different equation slopes were obtained for the different occupational groups (after the effects of gender and age had been controlled for), disconfirming the research hypothesis that the regression slopes would be parallel (H4).

Table 5.

Hierarchical Multiple Regression Analysis for the GHQ-12 Scores.

Steps	Variables	Model 1	Model 2	Model 3	Model 4	Model 5
		R² = .01	R² = .120	R² = .248	R² = .376	R² = .448
		R² = .01	ΔR² = .11**	ΔR² = .13**	ΔR² = .13**	ΔR² = .077*
		B	β	β	β	β
1	Intercept	10.68	11.37	12.15	12.73	12.90
	Age	−.004	.039	.021	−.021	−.072
	Gender	.095	.092	.054	.055	.091
2	Intrusion		.375***	.118	.211	.211
2	Avoidance		.020	.039	.031	.021
3	Physicians			−.155	−.049	−.031
	Counselors			−.056	−.021	.005
	Psychiatrists			−.241***	−.196**	−.306**
4	SOC				−.413***	−.453***
5	Intrusion × Physicians					.283*
	Avoidance × Physicians					.268*
	SOC × physicians					−.087
	Intrusion × Counselors					.345**
	Avoidance × Counselors					−.264*
	SOC × Counselors					.129
	Intrusion × Psychiatrists					.144
	Avoidance × Psychiatrists					.003
	SOC × Psychiatrists					.054

p < .05. **p < .01. ***p < .001.

Given that it is not straightforward to interpret what the models imply about the differences between categories by simply examining the regression coefficients, we opted to solve fitted regression equations for each group separately, in line with standard procedure. On the basis of the previously outlined coding scheme, the results obtained were as follows:

\begin{array}{l} Volunteers : Y = 12.908 + . 211 (intrusion / hyperarousal) \\ + . 031 (avoidance) - . 453 (SOC) . \end{array}

\begin{array}{l} Physicians : Y = 12.317 + . 494 (intrusion / hyperarousal) + \\ . 299 (avoidance) - . 366 (SOC) . \end{array}

\begin{array}{l} Counsellors : Y = 12.839 + 556 (intrusion / hyperarousal) - . \\ 233 (avoidance) - . 324 (SOC) \end{array}

\begin{array}{l} Psychiatrists : Y = 6.92 + . 067 (intrusion / hyperarousal) + \\ . 027 (avoidance) - . 399 (SOC) \end{array}

These findings suggest that SOC was a determinant of psychological distress mitigating its levels and this was so for all the occupational categories under study. From these equations, it may be concluded, for example, that levels of intrusion and hyperarousal were linked moderately and positively (β = .566) with psychological distress in physicians and counselors (β = .494) more so than in other aid occupations but had, on the contrary, little impact on the distress of psychiatrists (β = .067). In a similar fashion, avoidance also seemed to affect psychological distress negatively, although the dynamic was dissimilar across the various categories of helper. As in the case of intrusion and hyperarousal, avoidance had a very small single effect on the psychological distress of psychiatrists (β = .027), and displayed the greatest effect on that of physicians (β = .299), whereas it had a minimal effect on mental distress in the sub-sample of volunteers (β = .031). Surprisingly, in the group of counselors, avoidance (β = −.233) appeared to have a positive single effect on psychological distress, in a dynamic resembling the role of other “protective” factors that deserves to be further explored in the course of future research on this topic.

Discussion

In the present study, we aimed to test whether SOC acted as a determinant of positive psychological functioning in aid workers directly exposed to war contexts (Veronese & Pepe, 2014). In line with our research hypotheses, the findings support two lines of interpretation regarding, First, the predictive power of SOC and PTSD in relation to mental health across different professional groups, and second, the weight and direction of the beta coefficients found for specific occupational categories. In keeping with the current literature, SOC appears to act as an effective resource in generally mitigating the impact of stress among individuals and groups exposed to traumatic events (Braun-Lewensohn & Sagy, 2014) and this result stands for all occupational groups. Analysis of the regression model’s coefficient of determination (R²) revealed that, overall, 45% of variance in psychological distress was explained by SOC and IES scores, with the effects of variables varying as a function of occupational category. Analysis of beta weights revealed that in general SOC seemed to be negatively and moderately associated with psychological distress (with values ranging from −.366 for physicians and surgeons to −.324 for counselors and −.399 for psychiatrists). In contrast, the intrusion/hyperarousal dimension of IES-13 appeared to be positively and strongly associated with GHQ scores, with beta weights going from .494 for physician and surgeons, to .556 for counselors. The regression analyses suggested that trauma and SOC act as non-separate variables in relation to occupations given that interactive effects were found and beta weights for the multiplicative terms were not negligible.

Generally speaking, we found that professional aid workers operating in war-like conditions used their ability to attribute meaning and coherence (SOC) to their experiences, maintaining satisfactory levels of psychological functioning and well-being despite suffering from a moderate level of stress and trauma (Veronese et al., 2012; Veronese & Pepe, 2014). In particular, medical doctors, counselors, psychiatrists and, more markedly, community volunteers without a specific professional profile, protected their mental health by drawing on a SOC. It appears that in these professional groups, the more the impact of trauma puts members at risk of impairment, the more a SOC serves as a GRR for coping with exposure to trauma and stress (Bauwens & Tosone, 2010; Jonsson et al., 2013), thereby ensuring psychological health and well-being. In fact, there is clear evidence that a weaker SOC predicts higher scores on the GHQ-12.

A different trend was found for psychiatrists: In line with other groups, SOC acted as a predictor of good mental health outcomes in members of this professional category, but psychiatrists generally displayed a lower level of trauma and more satisfactory psychological well-being than other helpers after SOC levels had been controlled for. The lower level of trauma might be a consequence of the fact that psychiatrists are less involved in emergency programs on the front-line, and thus less exposed to direct and extreme trauma although more likely to suffer from vicarious traumatization and compassion fatigue. In a similar direction, Brockhouse, Msetfi, Cohen, and Joseph (2011) found that therapists displaying a moderate to strong SOC could use this ability to adjust to stress so as to maintain the quality of their therapeutic action when under pressure, but had less opportunity to benefit from SOC for growth.

A further explanation could regard the professional background of psychiatrists in comparison with other mental health providers and helpers. The Palestinian psychiatrists in our sample had been trained abroad, were highly qualified and well-positioned within the socio-economic hierarchy (in contrast with doctors, who were often not specialized and in some cases were under-qualified). It may be that this group’s access to ecological resources in terms of well-being and QOL affected their levels of generalized (and subjective) resistance resources (Veronese et al., 2012).

Some limitations must be taken into account and discussed. First, the small sample size does not allow us to generalize from the results. The limited number of institutions from which participants were recruited is not fully representative of the overall Palestinian aid system. This, together with a lack of detailed information about our informants’ level of education and years of experience may have generated some self-selection bias. Finally, the fact that the research design did not include a war trauma checklist limits the information available about how trauma history and type of exposure may affect the efficacy of SOC as a GRR. An important future development of this study will be to identify and differentiate between types of direct and indirect trauma, education level, and lifelong training opportunities that may act as intervening variables with respect to SOC in the different professional groups. In addition, the current study had a cross-sectional design: Future research on trauma and psychological distress should assess joint variation of both constructs over time. This could provide greater insight into the conditions under which SOC functions as a useful and dynamic mental process protecting aid workers from stress and trauma.

From this point of view, an interesting development will be to work toward making the SOC instrument as sensitive and effective as possible for Middle Eastern contexts. It may be therefore that we need to move onto a more qualitative research phase to explore in greater depth the peculiar features and components of the SOC construct in specific non-Western cultures (Sveaass & Castillo, 2000; Veronese, 2013).

Assisting aid workers to reinforce their personal resources, in terms of attributing meaning to experience and remaining in control in the face of traumatic events, can help national and international agencies to pursue the goal of reducing the risk of impairment to mental and physical health, stress, withdrawal, and burnout. It is therefore recommended that aid agencies provide adequate psychological support and supervision to emergency workers to prevent distress, perhaps not only by recruiting on-site psychiatrists but also by targeting helpers’ SOC so as to promote awareness and subjective well-being (Naudè & Rothmann, 2006). Turning now to the practical implications of our study, the definition of effective and efficient specific recruitment, training, and intervention models, which could also be appropriate for professionals training and operating in other “permanent crisis” contexts, should reflect the specific socio-political context as well as the living and working conditions of Palestinian helpers (Al-Krenawi, Graham, & Sehwail, 2007; Giacaman et al., 2011). A lack of economic resources and long-term social and political conflict have created a fragmented society and this must be taken into account when identifying potential strategies for promoting competence and professional well-being in Palestinian aid workers. Accordingly, the requirement for high-quality training cannot be met by implementing exclusively “standard” and “Western-informed” models (Al-Krenawi & Graham, 2000; Giacaman et al., 2011) that are based on empirical evidence and methodological rigor but do not satisfy the specific need of Palestinian helpers to construct meaning and attribute coherence to the permanent conditions of uncertainty affecting both their clients and themselves (Al-Krenawi et al., 2007). The key implication of the current findings is the value of providing training that not only conforms to international standards but also fosters local and culturally sensitive knowledge and competencies, so as to enhance helpers’ own levels of self-esteem and self-determination (Veronese, Prati, & Castiglioni, 2011). Action research and participative training models should thus be used in developing the programs of the institutions responsible for public health and higher education (Al-Krenawi et al., 2007). In a context of social trauma, it is of paramount importance for intervention models and instruments to promote maximum participation, be inclusive, provide opportunities to construct meaning, and enhance the competence of helpers already operating in the field.

In sum, despite the various demands and negative consequences generated by contexts of violence and uncertainty, higher levels of SOC can raise emergency workers’ awareness that they have the capabilities required to manage events and help them perceive their mission as worth investing their energy in.

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Author Biographies

Guido Veronese is a researcher in the Department of Human Sciences and Education ‘Riccardo Massa’, University of Milano-Bicocca, Italy.

Alessandro Pepe is a researcher in the Department of Psychology, University of Milano-Bicocca, Italy.

References

Aiken

L. S.

West

S. G.

(1991). Multiple regression: Testing and interpreting interactions. Newbury Park, CA: Sage.

Al-Krenawi

Graham

J. R.

(2000). Culturally sensitive social work practice with Arab clients in mental health settings. Health & Social Work, 25, 9-22.

Al-Krenawi

Graham

J. R.

Sehwail

M. A.

(2007). Tomorrow’s players under occupation: An analysis of the association of political violence with psychological functioning and domestic violence, among Palestinian youth. American Journal of Orthopsychiatry, 77, 427-433.

American Psychiatric Association. (2013). Diagnostic and Statistical Manual of Mental Disorders (5th ed.). Arlington, VA: American Psychiatric Publishing.

Andersen

H. S.

Christensen

A. K.

Petersen

G. O.

(1991). Posttraumatic stress reactions among rescue workers after a major rail accident. Anxiety Research, 4, 245-251.

Antonovsky

(1979). Health, stress and coping. San Francisco, CA: Jossey-Bass.

Antonovsky

(1987). Unraveling the mystery of health: How people manage stress and stay well. San Francisco, CA: Jossey-Bass.

Antonovsky

(1993). The structure and properties of the sense of coherence scale. Social Science Medicine, 36, 725-733.

Antonovsky

(1996). The sense of coherence: An historical and future perspective. Israeli Journal of Medical Sciences, 32, 170-178.

10.

Antonovsky

(1997). Salutogenese. Zur Entmystifizierung der Gesundheit [Salutogenesis. Demystifying Health], dgvt, Tübingen.

11.

Antonovsky

(2000). Hälsansmysterium [Unraveling the mystery of health]. Köping, Sweden: NaturochKultur.

12.

Antonovsky

Sagy

Adler

Visel

(1990). Attitudes towards retirement in an Israeli chorot. International Journal of Aging and Human Development, 31, 57–77.

13.

Bauwens

Tosone

(2010). Professional posttraumatic growth after a shared traumatic experience: Manhattan clinicians’ perspectives on post-9/11 practice. Journal of Loss and Trauma: International Perspectives on Stress & Coping, 15, 498-517.

14.

Binder

H. P.

Mensenholl-Strhler

Paß

Endler

P. C.

(2006). Sense of coherence among psychotherapists in Austria, differentiated accordingly to number of individually completed training therapy sessions. TSW Holistic Health and Medicine, 1, 232-235.

15.

Braun-Lewensohn

Sagy

(2014). Community resilience and sense of coherence as protective factors in explaining stress reactions: Comparing cities and rural communities during missiles attacks. Community Mental Health Journal, 50, 229-234. doi:10.1007/s10597-013-9623-5

16.

Brockhouse

Msetfi

R. M.

Cohen

Joseph

(2011). Vicarious exposure to trauma and growth in therapists: The moderating effects of sense of coherence, organizational support, and empathy. Journal of Traumatic Stress, 24, 735-742.

17.

Cavallera

Passerini

Pepe

(2013). Personality and gender in swimmers in indoor practice at leisure level. Social Behaviour and Personality, 41, 693-704.

18.

Cohen

West

S. G.

Aiken

L. S.

(2003). Applied multiple regression/correlation analysis for the behavioral sciences (3rd ed.). Mahwah, NJ: Lawrence Erlbaum.

19.

Cohen

Savaya

(2003). Sense of coherence and adjustment to divorce among Muslim Arab citizens of Israel. European Journal of Personality, 17, 309-326.

20.

Corneil

D. W.

(1993). Prevalence of posttraumatic stress disorders in a metropolitan fire department. Baltimore, MD: The Johns Hopkins University.

21.

Daradkeh

T. K.

Ghubash

El-Rufaie

O. E. F.

(2001). Reliability, validity and factor structure of the Arabic version of the 12-item General Health Questionnaire. Psychological Report, 89, 84-94.

22.

Echambadi

Hess

J. D.

(2007). Mean-centering does not alleviate collinearity problems in moderated multiple regression models. Marketing Science, 23, 438-445.

23.

Eriksson

Lindström

(2005). Validity of Antonovsky’s sense of coherence scale: A systematic review. Journal of Epidemiology & Community Health, 59, 460-466.

24.

Eshel

Mashbood

(2014). Posttraumatic recovery to distress symptoms ratio: A mediator of the links between gender, exposure to fire, economic condition, and three indices of resilience to fire disaster. Community Mental Health Journal, 50, 997-1003. doi: 10.1007/s10597-014-9734-7.

25.

Esposito

M. K.

(2005). The Al-Aqsa Intifada: Military operations, suicide attacks, assassinations, and losses in the first four years. Journal of Palestinian Studies, 34, 85–122.

26.

Fear

N. T.

Wessely

(2015). Prevention of the psychological consequences of trauma. Lancet Psychiatry, 2, 367-369. doi: 10.1016/S2215-0366(15)00159-5

27.

Figley

C. R.

(1995). Compassion fatigue: Coping with secondary traumatic stress disorder. New York, NY: Brunner/Mazel.

28.

Figley

C. R.

(2002).Treating compassion fatigue. New York, NY: Brunner/Routledge.

29.

Flynn

B. W.

McCarroll

J. E.

Biggs

Q. M.

(2015). Stress and resilience in military mortuary workers: Care of the dead from battlefield to home. Death Studies, 39, 92-98.

30.

Giacaman

Rabaiaa

Nguyen-Gillham

Batniji

Punamäki

R. L.

Summerfield

(2011). Mental health, social distress and political oppression: The case of the occupied Palestinian territory. Global Public Health, 6, 547-559.

31.

Gilbar

(1998). Relationship between burnout and sense of coherence in health social workers. Social Work Health Care, 26, 39-49.

32.

Gittings

C. B.

Paterson

H. M.

Sharpe

(2015). How does immediate recall of a stressful event affect psychological response to it? Journal of Behavior Therapy and Experimental Psychiatry, 46, 19-26.

33.

Goldberg

D. P.

(1972). The detection of psychiatric illness by questionnaire. London, England: Oxford University Press.

34.

Goldberg

D. P.

Williams

(1988). A user’s guide to the General Health Questionnaire. London, England: NFER-Nelson.

35.

Hochwälder

(2015). Test of Antonovsky’s postulate: High sense of coherence helps people avoid negative life events. Psychological Reports, 116, 363-376.

36.

Horowitz

Wilner

Alvarez

(1979). Impact of Event Scale: A measure of subjective stress. Psychosomatic Medicine, 41, 209-218.

37.

Jakobsson

(2011). Testing construct validity of the 13-item Sense of Coherence Scale in a sample of older people. The Open Geriatric Medicine Journal, 4, 6-13.

38.

Jonsson

Segesten

Mattsson

(2013). Post-traumatic stress among Swedish ambulance personnel. Emergency Medicine Journal, 20, 79-84.

39.

Jordan

B. K.

Marmar

C. R.

Fairbank

J. A.

Schlenger

W. E.

Kulka

R. A.

Hough

R. L.

Weiss

D. S.

(1992). Problems in families of male Vietnam veterans with posttraumatic stress disorder. Journal of Consulting and Clinical Psychology, 60, 916-926.

40.

Korff

V. P.

Balbo

Mills

Heyse

Wittek

(2015). The impact of humanitarian context conditions and individual characteristics on aid worker retention. Disasters, 39, 522-545. doi: 10.1111/disa.12119

41.

Lundin

(1997). Posttraumatic stress reactions among 19 rescue men from the Swedish navy after the Estonia ferry disaster. In The loss of M/S Estonia in the Baltic on 28th September 1994 (pp. 158-162). Stockholm, Sweden: Swedish Disaster Medicine Study Organization.

42.

Mitchell

J. T.

Dyregrov

(1993). Traumatic stress in disaster workers and emergency personnel. In Wilson

J. P.

Raphael

(Eds.), The International handbook of traumatic stress syndromes (pp. 305-314). New York, NY: Plenum.

43.

Naudè

J. L. P.

Rothmann

(2006). Work-related well-being of emergency workers in Gauteng. South African Journal of Psychology, 36, 63-81.

44.

Niţă

Z. D.

Petre

(2012). Risk factors for psychological distress in humanitarian and disaster work. Paradigm Social Science Journal, 5. Retrieved from http://eras.socio.uvt.ro/?journal=eras&page=article&op=view&path%5B%5D=48&path%5B%5D=21.

45.

Oumette

Wade

Prins

Schohn

(2008). Identifying PTSD in primary care: Comparison of the Primary Care-PTSD screen (PC-PTSD) and the General Health Questionnaire-12 (GHQ). Journal of Anxiety disorders, 22, 337-343.

46.

Pedhazur

E. J.

(1997). Multiple regression in behavioral research (2nd ed.). Fort Worth, TX: Harcourt Brace.

47.

Pepe

Addimando

(2014). Teacher–parent relationships: Influence of gender and education on organizational parents’ counterproductive behaviors. European Journal of Psychology of Education, 29, 503-519.

48.

Petersen

Ladelund

Carlsson

Nilbert

(2013). Sense of coherence and self-concept in Lynch syndrome. Hereditary Cancer in Clinical Practice, 11, 1-7.

49.

Ravenscroft

(1994). Going critical: GMB/ Apex and T&G Unions 1994 survey of occupational stress factors in accident and emergency staff in the London ambulance service. London, England: GMB/Apex/and T&G Unions.

50.

Richardson

C. G.

Ratner

P. A.

(2005). A confirmatory factor analysis of the Fagerstrom Test for Nicotine Dependence. Addictive Behaviour, 30, 697-709.

51.

Schaufeli

W. B.

Bakker

A. B.

(2001). Werkenwelbevinden: Naareenpositieve benadering in de Arbeids-enGezondheidspsychologie [Work and well-being: Towards a positive occupational health psychology]. Gedrag & Organisatie, 14(15), 229-253.

52.

Seligman

Csikszentmihalyi

(2000). Positive psychology: An introduction. American Psychologist, 55, 5-14.

53.

Selley

(1991). Post-traumatic stress disorder. Practitioner, 235, 635-639.

54.

Sveaass

Castillo

(2000). From war hero to cripple: An interview study on psychological intervention and social reconstruction in Nicaragua. Peace and conflict. Journal of Peace Psychology, 6, 113-133.

55.

Taubman-Ben-Ari

Rabinowitz

Feldman

Vaturi

(2001). Post-traumatic stress disorder in primary care settings: Prevalence and physician’s detection. Psychological Medicine, 31, 555-560.

56.

Thormar

S. B.

Gersons

B. P. R.

Juen

Djakababa

M. N.

Karlsson

Olff

(2015). The impact of disaster work on community volunteers: The role of peri-traumatic distress, level of personal affectedness, sleep quality and resource loss, on post-traumatic stress disorder symptoms and subjective health. Journal of Anxiety Disorders, 28, 971-977.

57.

Tomotsune

Sasahara

Umeda

Hayashi

Usami

Yoshino

. . .Matsuzaki

(2009). The association of sense of coherence and coping profile with stress among research park city workers in Japan. Industrial Health, 47, 664-672.

58.

U.S. Bureau of Labor Statistics. (2010). Standard occupational classification policy. Washington, DC: Government Printing Office. Retrieved from http://www.bls.gov/soc/soc_2010_user_guide.pdf

59.

Veronese

(2013). Self-perceptions of well-being in professional helpers and volunteers operating in war contexts. Journal of health Psychology, 18, 911-925.

60.

Veronese

Fiore

Castiglioni

el-Kawaja

Said

(2012). Can sense of coherence moderate traumatic reactions? A cross-sectional study of Palestinian helpers operating in war contexts. British Journal of Social Work, 43, 651-666.

61.

Veronese

Pepe

(2013). Psychometric proprieties of the Impact of Event Scale (short version) in contexts of military violence. Research on Social Work Practice, 23, 710-718.

62.

Veronese

Pepe

(2014). Sense of coherence mediates the effect of trauma on the social and emotional functioning of Palestinian health providers. American Journal of Orthopsychiatry, 84, 597-606.

63.

Veronese

Pepe

Afana

(2016). Conceptualizing the well-being of helpers living and working in war-like conditions: A mixed-method approach. International Social Work, 59, 938-952. doi: 10.1177/0020872814537855

64.

Veronese

Prati

Castiglioni

(2011). Postcolonial perspectives on aid systems in multicultural contexts: Palestine and Uganda. Proceedia: Social and Behavioral Sciences, 15, 545-551.

65.

Wagner

Heinrichs

M. S.

Ehlert

(1998). Prevalence of symptoms of posttraumatic stress disorder in German professional firefighters. American Journal of Psychiatry, 155, 1727-1732.