Major findings of a trauma survey with two samples of adolescents (age 11-17) are presented. One was a sample of school children from Croatia screened for PTSD at the time of warfare in former Yugoslavia (N = 172). The control was a sample of peers from Hungary (N = 282), with no personal experience of any warfare in lifetime. Croatian and Hungarian versions of a self-administered screening instrument was used to measure perceptions of trauma and twelve symptoms of PTSD. The findings call attention to the fact that despite remarkable difference between the two groups in free listing of adverse life events, the levels of distress as measured with a series of PTSD symptom scales were found not that dramatic as one would expect. The following symptom scales were found with high average self-rating in both groups: Nearly equally high mean scores were estimated for both groups in the following symptom scales: 1, Anger, 2. Revenge, 3. Meaninglessness, 4. Fear,5. Powerlessness, in this order. The largest between groups difference was found in the “Hate” subscale on the behalf of the warfare sample (t = 8.46 df = 426 p < .0001). Alternative interpretations of the results are discussed.

Background

The diagnostic history of posttraumatic stress disorder (PTSD) goes back to the 1952 edition of the DSM-I, then called, “gross stress reaction”. The case was understood as patterns of reaction of otherwise normal personality to overwhelming fear caused either by combat or “civilian catastrophe”. [1] The term first appeared in the 1980 edition of the DSM-III. [2] At that time, most research and public debates on PTSD were greatly influences by the experiences and conditions of US military veterans of the Vietnam War. It has become almost synonymous with many earlier war-time diagnoses such as shell shock, battle fatigue, combat stress reaction, or traumatic war neurosis. [3]

Although the correlation between combat and PTSD is undeniable, war related traumatic events are not the only factors in the etiology of the phenomenon. The framers of the construct implicitly understood that “traumatic” should be considered only catastrophic events, the impact of which would overwrite all and any normal human experience and coping capabilities. Such events would be war, torture, rape, Nazi Holocaust, atomic bombing, airplane crashes, traffic accidents and more. [4] Further, they believed that traumatic events are clearly different from non-traumatic stressors, which may be painful tough, but they are typically vicissitudes of life such a natural death of a family member, divorce, failure, rejection, serious illness and the like, hence called “Adjustment Disorders” rather than PTSD. The latest edition of the diagnostic system (DSM-5, 2013) lists eight evidence-based criteria for PTSD [5], with the “traumatic event” criterion on the first place. The novelty was in moving the concept of PTSD out from fear-based anxiety disorders into the cluster of more complex affective-cognitive disorders characterized by fear component, anhedonic/dysphoric presentations marked by negative cognitions and mood states as well as disruptive (e.g. angry, impulsive, reckless, self-destructive) behavioral symptoms. [6]

All until the mid- eighties, psychiatrists were skeptical about the concept of PTSD in children. British authors Garmezy and Rutter wrote in 1985: “behavioural disturbances appear to be less intense that might have been anticipated; a majority of children show moderate amount of fear and anxiety…; aggressive behaviour marked by the clinging to parents and heightened dependency on adults appears, and a moderately mild sleep disturbance persists for many months.” [7: p. 162] However, evidence drawn from direct encounters with and observations of children personally exposed to dramatic events contradicted such misbelieves. It was a come-back to Freud’s classical psychanalytic thesis that that trauma in children does exists, indeed, although its manifestation is very different from that in adults [8-11] Pynoos and Nader were among the first to distinguish three developmental patterns of trauma reactions in minors: (A.) Preschool children through Second Grade would typically respond with helplessness and passivity, (B.) younger children from Third through Fifth Grade would be preoccupied with own actions during the event by focusing on responsibility and guilt; (C.) Adolescents (Sixth Grades and Up) would typically respond like adults with detachment, shame, guilt and acting out. [12] Recent evidence adds a few details to that. [13] Very young children for instance, may manifest only generalized fears such as stranger or separation anxiety that may or may not relate to any trauma. Elementary school-aged children may not experience visual flashbacks or amnesia for aspects of trauma. Instead, they likely experience “time skew”, i.e. mis-sequencing trauma related events when recalling the memory. School-aged children also reportedly exhibit posttraumatic play or reenactment the trauma in play, drawings or verbalization. Adolescents are more likely to engage in traumatic reenactment; they more likely than younger children or adults exhibit impulsive and aggressive behaviors.

New waves of wars and new forms of violence since the early 1990s keep PTSD in the focus of both research and public interest. Epidemiologists would say that PTSD is now a globalized phenomenon in modern societies. The prevalence of lifetime symptoms in the US general population aged 15 to 55 years has been approximated to 7.8%, where women (10.4%) are more than twice as likely as men (5%) impacted. [14] Authors of a meta-analysis of several thousands of children affected by war reported that the prevalence rates of PTSD may reach a level as high as 47%. [15] Research of adolescents in US would show lifetime prevalence of PTSD 4.7% on the average, with significantly higher rates among females (7.3%) than among males (2.2%). [16] Stefanović et al. investigated the relationship of early-life trauma, war-related trauma, personality traits and symptoms of PTSD among female civilian victims of the recent war in Croatia (1991-95). The study involved 293 war-traumatized adult women civilians, and 101 women without war-related trauma. The authors demonstrate that older age, exposure to early--life trauma, exposure to war-related traumatic events, high neuroticism, and low extroversion are factors associated with higher level of PTSD symptoms among women civilian victims of war. [17]

Cross-cultural (ethnocultural) research is rare in the literature. [18] This is simple because the notion of trauma per se is a social (historical) construct, its experience is subjective, and not last, its expression is culture specific. What is labeled “traumatic” in one culture or social group, may be regarded “ordinary” (casual) event in another. One first cross-cultural surveys was initiated in late 1990s by the World Health Organization (WHO) in 27 countries with the general finding that the prevalence rates of PTSD may widely across culture and countries from a low 0.3% in China, for instance, all up to 6.1% in New Zealand. [19] To our knowledge, systematic research of PTSD in children and adolescent in East European countries is virtually non-existent, except for former Yugoslavia during the warfare (1991-95), mainly thanks to UNICEF initiatives and support in trauma prevention. [20] This paper is an extension to our earlier explorations of PTSD among children and adolescents in this region, hereby comparing empirical evidence from Croatia and Hungary.

Objective

To examine the range of perceived traumatic life events and their impact on adolescents living in war conditions and compare them with peers in the same measures born and living in non-war conditions in.

Method

Subjects: Two groups of school-aged children were surveyed, both between age 11-17: One was a wartime sample of minors from non-occupied parts of Croatia at the time of homeland war (1991-95), including both residents, displayed and refugee children. from war zones with high, medium and low exposure to military operations. The sample consisted of 150 subjects, 72 boys and 78 girls with mean age 12.4 (SD ±1.22),), all were participants of a UNICEF supported public library-based trauma prevention program at the time of survey, which was implemented in major cities in non-occupied regions of Croatia (January ’95), covering public libraries in major cities, including Zagreb, Karlovac, Rijeka, Slavonski Brod, Osijek among others. Most of this target group were residents (70%). a smaller number displaced or refugees (14-14%). Regarding war exposure, 17% of subjects were from low-risk zones, 58% from medium and 25% from high-risk regions such as Vukovar. [21] The control group was a casual sample of school children from cross border public schools in Hungary, Barany County including schools at Pécs, Mohács, Siklós, Villány among others (April ‘02.), living in normal conditions [22] This was a sample of 282 children (124 boys and 138 girls) with mean age 14.4 (SD ±0.9), who except for one refugee child from Croatia have had no personal experience with warfare in their lives.

Instruments and procedure: Croatian and Hungarian versions of a self-administered questionnaire was used, called PTSD-12 Scale, developed validated during the warfare in Croatia [20]. The questionnaire consisted of two parts. Part I. was a free-listing (open) task in asking respondents to recall and identify adverse lifetime events, which they feel “traumatic” and still impacting. [23] The rationale for using this open-ended ethnographic tool instead of customary checklists was to evade response bias and stereotyping (e.g. social desirability tendency); and on the other, to uncover context-specific life events considered “traumatic” in given settings. Part II. consisted of 24 parallel items adopted from a trauma recovery workbook for survivors of war [24], used for the first time and checked for psychometric characteristics on larger samples on children and adolescents affected by war in former Yugoslavia. [25] The items described personal experiences that may indicate the consequence of any traumatic distress, including DSM criteria among others (intrusion, avoidance, hypervigilance). Subjects were asked to approximate on a five-point scale the frequency of occurrence of described experiences during the past month or (0 = never… 4 = always). The Twelve parallel scales were used in for measuring the following constructs often quoted in trauma research and discussions: S1. Powerlessness, S2. Meaninglessness, S3. Distrust, S4. Shame, S5. Anger, S6. Anxiety, S7. Guilt, S8. Mourning, S9. Fear of death, S10. Revenge, S11. Hopelessness, S12. Hate. The rationale for measuring this range specific symptoms was in their conceptual clarity and easiness to capture and express them through arts. [25] The administration of the questionnaire in Croatia was entrusted to art therapy workshop leaders (local librarians). In Hungary, a group of social work students from University of Pécs accomplished the classroom-based administration of the Hungarian translation of the survey questionnaire in casually selected public schools

Statistical analysis: Qualitative data drawn from free listings of traumatic events were content analyzed, categorized and made available both for descriptive and multivariate analysis. Reliability statistics were estimated for symptoms scales with satisfactory results for both study samples: Cronbach’s Alphas .863 and .904, respectively). Between group differences in the PTSD symptoms scales were tested by the means of t-statistics. One-way ANOVA procedure was used to explore the association between the event lists and self-ratings in PTSD symptom scales. Multiple regression analysis was applied to explore the effect of respondent’ gender, age and level of war exposure as independents on PTSD symptom ratings as dependent, both at the level of total PTSD-12 total scale and separate symptom scales. Cutting points in thePTSD-12 Scale were estimated in dividing the case distribution into three equal parts to estimate the rates of symptom severity levels in the two surveyed group. The criteria for significance of statistical tests were set to conventional standard (p< .05). The analyses were accomplished with Microsoft Excel and SPSS-25.

Results

Traumatic events listed: The table below displays the frequency distribution of the first mentioned traumatic events in the free listing task. (Table 1)

(*) First mentioned events in the free-listing task of the PTSD-12 Questionnaire

Table 1. Categories of disclosed traumatic lifetime events*

The personally experienced events and situations disclosed by members of the wartime (Croatian) sample are mainly familiar from several other sources as well. [26-27] Here too, the death or loss of a family member in the war was the most frequently mentioned traumatic event, followed by painful memories of destroyed homes and air raids. In contrast. In contrast, Hungarian adolescents listed life events typical for non-war conditions, in general. [28-29] Death of a close relative or a peer was the most frequently listed event, followed by school stressors. Curiously enough, loss of a pet was claimed traumatic in a relatively high number of subjects (13%), whereas death of mother, father or a sibling rarely happened in this sample, i.e. 10 cases (4%). No brutal acts of interpersonal violence were listed in any of the study samples.

The event lists were cross tabulated with respondents’ gender and age groups to see statistical association according to Chi-squares criteria (p< .05). Two substantial discrepancies were found, both in the Hungarian (non-war) sample but none in the Croatian wartime sample. One was a gender effect regarding school stressors, with more girls than boys reporting them as “traumatic” (23.6% vs. 14.5%; respectively; χ2 = 3,46 p< .05). Another was an age effect regarding family stressors, which were mentioned three times more often in younger age group (age 11-13) than in older age groups (29,4% vs 8-9%; χ2 = 7,61 p< .05).

Events impact: The subsequent table summarizes ANOVA statistics on dispersion of assessed levels of global distress by categories of events claimed “traumatic”. (Table 2)

1. Croatian wartime sample (N= 140)

ANOVA: F (7,132) = .938 Sig = .479 (p>.05)

1. Hungarian control sample (N=242)

ANOVA: F (7,234) = 2.160 Sig = .039 (p<.05)

Table 2 Groups means in PTSD-12 Scale by categories of disclosed events

The findings made evident that the diversity of events in both study samples was associated with the diversity of symptom severity. Air raids and bombing/sheltering nearby were most distressing events for teens in war conditions, whereas death of a parents or close other were most impacting for peers in non-war conditions,

Self-ratings inPTSD-12 symptom scales: The subsequent table lists group means in the series of twelve PTSD symptom scale, and estimated t -tests of on between group differences in these measures (Table 3.).

The mean plots diagram below may make easier to see both the similarities and differences between the study groups across the twelve symptom scales.

Table 3. Group means in PTSD-12 symptom scales

(*) Scales with statistically significant levels of between-group differences (p< .05)

Figure 1. Mean plots in PTSD symptom scales

As indicated, the gap between the study samples were estimated statistically significant in seven symptom measures, and in the total PTSD-12 Scale as well (p<.05). The wartime sample scored higher in six scales, in this order: 1. Hate, 2. Fear, 3. Revenge, 4. Mourning, 5. Powerlessness and 6. Shame. The only symptom scale was Anger, in which the Hungarian Hungary (control) group scored higher than the wartime peer group (MH= 3.45 vs MC= 3.00, t = 2,11 p<.05). Regardless to uneven gaps in individual symptom scale, the over-all shape of the two group profiles is similar as if “mirroring” each other (correlation Rho = .622).

Demographic predictors: Multiple regression analyses was accomplished with the following backroad variables at hand for both study samples: War exposure,[1] Age and Gender (independents) and PTSD-12 symptom scales (dependents).

The principal findings are summarized in the table below (Table 5).

Table 5. Summary of multiple regression analysis on PTSD-12 Scale

R =.302; R2 = .091

ANOVA F (3, 403) = 13.529 Sig. .000

Regression effects of gender and war exposure were found statistically significant (betas p<.05) indicating that girls and those with higher levels of war exposure were scoring higher in the PTSD-12 Scale. Contrary to expectation, isolated effect of age was found statistically non-significant (beta = .033 p>.05). The same regression model was applied to the full series of twelve individual symptom scales with similar outcomes. The exception was the Anger scale (as dependent) for which the partial regressions were estimated non-significant (p>.05). Las but not least, the partial regression effects of war exposure varied across the symptom scales. Robust exposure effects were detected in the following symptom scales, in the following order: 1. Hate, 2. Fear, 3. Mourning, 4. Powerlessness, 5. Meaninglessness, 6. Revenge, 7. Shame and 8. Distrust.

Severity levels of PTSD symptoms: The segmentations of the case distributions in the PTSD-12 Scale (total scores) in three groups of nearly equal size made possible to compare the two study groups regarding the rates of cases in different segments of PTSD-12 Scale (range 0-87points). 1. Low (0-16 points), 2. Medium (17-30 points), 3. High impact (31-87 points). The results are displayed in the subsequent table (Table 6).

(*) Scale parameters: N=432, Mean = 25.o, SD = 14.82; Median 22 Mode 14

Table 6. Rates of impact levels in PTSD-12 Scale*

The chi-square test on discrepancy between the two study samples was estimated significant (χ2 = 20.898 df=2 p< .001). The rates of high scoring subjects in the wartime sample was nearly twice as high as the rates in the Hungary (control) sample (43.8% and 24.8% respectively). Substantially more girls than boys scored high in the scale in the non-war (Hungarian) sample (32.9% vs. 13.9%), which was not the case in the warfare (Croatian) sample (42.0% vs 45,5%). Regarding age, teens with younger age (11-12) were found more frequently than older in the “’severe” category of impact with rates between 35.3% and 47.0% in the total sample.

Discussion

The study adds to the piling evidence that modern wars are one of possibly most traumatic events for anyone, particularly for children and adolescents causing lethal impact on their mental, social and moral well-being. In addition to, this same study calls attention to the fact that still many adolescents who live in non-war conditions may be equally, if not more vulnerable to PTSD than those in a warfare. The difference between teens living in war zones and peers in non-war conditions can results from any of the following factors and/or their combination: the difference in social visibility of distressing events, difference in levels of psychological responding to events ranging from hidden individual to shared communal. [30-32] Viewed from this perspective, it seems problematic to dichotomize life events that adolescents experience into “traumatic” and “non-traumatic” simply because mental trauma, by definition, is a culture-specific subjective construct. [33]

It was expected that the two groups of teens surveyed in this study would produce entirely different event lists claimed “traumatic” in lifetime (Table 1). However, more salient was the observation that wartime events, in contrast to those in non-war conditions, ar typically public events, which are not only fearful for anyone, but like most disasters, are life-threatening both for the self and others. Regardless to phenomenological difference in event listings, the rather curious finding was the diversity of mean scores in the PTSD-12 symptom scales across the event lists (see Table 2). Namely, some events were liked to substantially elevated symptom levels, whereas others not. This diversity surely was not the case of direct cause-effect relationship between the two variables. Rather, the events listed were most probably only reminders (triggers) of full clusters of distressing memories, which all together influenced self-assessments in symptoms measure (cumulative effects). Another finding was the striking similarity of the two groups in symptoms assessment, regardless of varying gaps in average scoring (see Table 3 & Figure 1). Specifically, the mean plots of the two study samples across the PTSD-12 symptom measures were almost parallel, as if were sub-samples of the same population. Two theoretical frameworks can be addressed here for explanation. One would be a developmental approach in stressing the globalization of PTSD in the world’s general populations, particularly among the youth in the context of increasing violence in modern societies. [34-35]. An alternative would a cultural-anthropological approach by bringing into the focus the notion of traumatized societies and related constructs. [36-37] From this perspective, the category “traumatized societies” would apply both to Hungary and the Balkans, with national histories of collective trauma reaching back several centuries [38-39] Therefore, it should be is not surprising that our study samples of adolescents scored parallel in a fair number of PTSD indicators such as Anger, Revenge, Meaninglessness and Fear. Finally, the estimated prevalence of “sever” PTSD symptoms in both samples appears alarming (Table 6). This is because the estimates for both study samples were found well above the rates estimate in other European countries, with the average of severely impacted ca. 16%, of whome10% impacted by non-interpersonal traumatic events and 25% following interpersonal trauma. [40] A host of comparable estimates are available for Croatia, virtually none from Hungary for various reasons.

Peer reviews of former versions of the paper pointed out two major limitations of the study. One was the non-comparability of study samples both in time of the accomplished surveys, and social conditions of the two countries. Another limitation was seen in the study design, which ignored the probable changing rates of PTSD among adolescents in both countries. Last, critics stressed that the use of an unconventional, mainly unknown empirical measure of trauma symptoms (PTSD-12 scales) is of questionably construct validity, compared to generally acknowledges Impact of Events (IES) Scale widely used as frame of reference both in national and cross-cultural trauma research. [41] All these critical comments are acceptable and shall be a challenge for further research on the same or different populations of adolescents facing post-traumatic disorders.

Conclusion
Adolescents experiencing warfare would perceive and call “traumatic” life events, which are entirely different of those listed by peers living in non-war conditions. However, the pattern of subjective responding to such events is similar in these groups, which is a finding open to different theoretical interpretations and further research. [1] The variable War exposure was a UNICEF expert-based categorization of war-impacted zones in Croatia and Bosnia-Herzegovina regarding military operations and children’s safety. Scale values: 0 = No war exposure (Hungarian sample), 1 = Low exposure, 2 = Medium exposure, 3 = High exposure