Internalizing and Externalizing Problems in Children of War Veterans in Kosovo

The aim of this study was to explore internalizing and externalizing problems in adolescent children of veterans of the war in Kosovo (1998-1999). The results of this study are quite interesting from the perspective of the dilemma for the state of the children of veterans even 15 years after the war ended. Parents’ emotional problems affect the functioning of the family in general and children in particular. Children can react to symptoms of parents by developing different symptoms as trouble sleeping, appetite loss, emotional instability or even problems in development, according to research done on children's reactions to the problems of parents explained by interactions between environment, brain and behaviour driven by trauma. The results of this study have shown that the internalizing problems have not shown gender differences, meantime externalizing problems were found higher in male participants. An interesting finding of this study was the highest scores of emotional problems in children born before and during the war, compare to those born after the war ended. We also found that anxiety problems in children [R2= .83, p < .001] were a significant predictor of internalizing problems. The assessment of the scale of positive qualities [R2= .19, p < .001] was also found to be a significant predictor for externalizing problems.Only 0.8% of the variance of internalizing problems was explained by the income. Considering that the subject of this study were adolescent children of war veterans of the 1999 conflict in Kosovo, we must take into account that the post-traumatic stress disorder is a very frequent problem among war veterans and that its impact on their personal and family life cannot be overlooked.


Internalizing and Externalizing Problems in Children of War Veterans in Kosovo
Emotional disorders are quite widespread among persons who have faced severe traumatic events such as murder, torture, witnessing murders or the missing of family members or war combats, or missing family members for a long time. According to Dekel and Goldblatt (2008), clinical observations and empirical research have shown that the consequences of traumatic events are not limited to the person immediately exposed to the event, and that they often affect family and friends. In a meta-analysis involving 42 studies, Lambert et al. (2014) suggested the importance of considering the family context of the trauma survivors. Psychological problems of children with traumatized parents should be considered from the perspective of genetic predisposition, secondary traumatisation, difficulties in family functioning, or other challenges associated with living with a traumatized parent.
Studying the effect of veterans' exposure to war on their family members, Herzog, Everson and Whitworth (2011) found a higher presence of problems with secondary traumatic stress in family members of war veterans with a high level of PTSD. Secondary traumatisation in its narrow sense involves a transfer of nightmares, intrusive thoughts, flashbacks and other Posttraumatic Stress Disorder symptoms (now on referred as PTSD), which are typically experienced by individuals suffering from PTSD, onto their immediate surroundings (Klaric et. al., 2013). A study conducted with family members of veterans in Portugal found that children had PTSD secondary traumatic symptoms that were related only with having a father who had experienced war and was suffering PTSD, and children showed _____________________________ Iliria International Review -2015/1 © Felix-Verlag, Holzkirchen, Germany and Iliria College, Pristina, Kosovo more psychopathological symptoms, physical symptoms and adverse experiences in childhood (Pedras and Pereira, 2014). The process of the transfer of trauma symptoms from veterans to their children may occur through avoidance (Ancharoff, Munroe andFisher, 1998), 1998., dentification (Rosenheck and Fontana (1998), and may be transmitted indirectly from dysfunctional relations within the family (Galovski and Lyons, 2004).
In a study conducted in Bosnia and Herzegovina children of veterans with PTSD demonstrated significantly higher levels of fear, problems in school, and a higher level of depression, emotional problems and dysfunctional behaviour in comparison to children of veterans who did not have PTSD, which in the longer term affected children's personality (Klaric, et.al., 2008). In similar studies conducted in Croatia, USA and other countries, children of veterans reported to have been more likely to harm themselves intentionally in comparison to children of non-veterans (Franic, et.al., 2012), to have suicidal thoughts, depression, and low self-esteem (Reed, Bell and Edwards, 2011;Marsanic et.al., 2014), more psychopathology (Vaage, et.al., 2011), internalizing problem (Boricevic, et.al., 2014), emotional and behavior problems (Jordan, et.al., 1992;Selimbasic, Sinanovic and Avdibegovic, 2012), academic problems, hopelessness about the future and clinical depression (Kinzie, Boehnlein and Sack, 1998), increase in physical accidents and neglect by their parents (Dias, et.al., 2014), higher level of anxiety, greater concerns related to stress and less capacity for intimacy (Dinshtein, Dekel and Polliack, 2011), Subject of this study were adolescent children of veterans of the 1999 war in Kosovo. This study analyzed the connection between a number of variables and internalizing and externalizing problems in adolescent children of veterans.

Methodology
This paper is part of the study of the trauma transfer through generations, and it involved the veterans of the latest war in Kosovo 1998Kosovo -1999. The participants in this study were selected from the list of war veterans provided by the Association of War Veterans. The family members of the war veterans were involved in the study through a random selection from the list, involving six regions which geographically cover the whole territory of Kosovo. The veteran response rate was 72.2 %. 568 of _____________________________ Iliria International Review -2015/1 © Felix-Verlag, Holzkirchen, Germany and Iliria College, Pristina, Kosovo those veterans had a child of age 11-18. All the questionnaires were selfadministered after the adolescents gave oral consent. The Medical University of Prishtina (Ethical Review Committee) has given the Ethical approval. The procedures proposed complied with the Helsinki Declaration of 1975, as revised in 2008.
The instruments used for the propose of this paper were a demographic questionnaire prepared for the purposes of this research and Youth Self Report(YSR) (Achenbach and Rescorla, 2001) The 2001 problem scales for the YSR were computed from the 112 problem items: three broad-band scales (Internalizing, Externalizing, and Total Problems); eight empirically based syndromes (Anxious/Depressed, Withdrawn/Depressed, Somatic Complaints, Social Problems, Thought Problems, Attention Problems, Rule-Breaking Behaviour, and Aggressive Behaviour. Cronbach's alphas Internal consistencies of the 20 YSR scales were examined using Cronbach's alphas. The internal consistency for this questionnaire had the following values for: total problems (α = 0.94), internalizing problems (α = 0.86) and externalizing problems (α = 0.84). Meanwhile, no gender based differences were found in reporting α coefficients on YSR scales. The highest α coefficients were reported on the scales of aggressive behaviour (α = 0.82), while the lowest was reported on the scale of mental problems (α = 0.64). In this sample's reporting this coefficient was from very good to excellent.
The majority of the parents of the participants had primary and secondary education (26.4% vs. 44.2%), while only three of the parents had no education at all (1.4%). 28% of the parents had higher education. Adolescents reported that 60.1% of their parents were not working and 39.8% having a job (χ 2 (1) = 0.06, p = .505). The average income of the families of the adolescents was around 200-400 Euro per month (52%), while 23.9% lived on 0-200 Euro per month.

The influence of socio-demographic variables on YSR scales
The highest mean was recorded for the scale of aggressive behaviour M = 5.1, (SD = 4.5); and the scale of anxiety M = 4.8,(SD = 3.5), while the lowest mean was reported on the scale of somatic problems M = 2.2, (SD = 2.4). The mean value reported for total problems was M = 46.5 (SD = 24.8), for internalizing problems M = 11.7 (SD = 7.7), and for externalizing problems M = 10.9 (SD = 6.9),(table 1). T test was conducted to find the differences in reporting the means on the three main YSR scales. The results show that significant differences were found for the scale of externalizing problems t (566) = 1.5, p =.016, where males showed a higher mean of those problems (M = 11.6, SD = 7.3) than females (M = 10.2, SD = 6.5). No significant differences were found in internalizing problems t (566) = 0.054, p =.263 and total problems t (566) = 0.042, p =.484.  One-way ANOVA was used to see whether the age of adolescents affected the variance of the three main YSR scales. No significant differences were found in the variance of total problems, internalizing and externalizing problems in relation to age. Post-hoc analysis was used to identify the influence of the age and it was found that age 16 had significant differences with age 11 on the scale of internalizing problems (I-J = 3.4, p = .028) and with age 13 on the scale of externalizing problems (I-J = 3.3, p = .021). Meanwhile, on the scale of total problems there were significant differences between age 16 and age 11 (I-J = 10.1, p = .042) and age 13 (I-J = 12.6, p = .014).
The T-test data are reported in Table 3. T-Test was performed in order to compare the mean of participant reporting on eight empirical scales between the group of female participants and the group of male participants.
The results of the test show that the difference between those two groups was significant only on the scale of aggression (t (566) = 3.1, p = .003). The male participants had a higher mean on the scale of Rulebreaking behaviour (M=5.3, DS=4.7), in comparison to female participants (M=4.7, DS = 4.2).  ANOVA results on the influence of age on reporting the variance of the eight empirical scales showed that there were no significant differences in any of them. The Post-hoc analysis found significant differences in the variance of anxiety disorder between age 11 and age 15 (I-J -1.9, p = .008) and age 16 (I-J -1.5, p = .028), and also between age 15 and age 17 (I-J 1.3, p = .048). On the scale of withdrawal, significant differences were found between age 16 and age 11 (I-J 1.3, p = .011) and age 13 (I-J 1.5, p = .007). While on the scale of somatic problems significant difference was found between age 14 and age 15 (I-J -1.04, p = .03) and on the scale of mental problems significant difference was found between age 11 and age 16 (I-J -1.2, p = .019) _____________________________ Iliria International Review -2015/1 When we dichotomized age into the age group 11-14 years (1), namely children who were born immediately after the war and the age group 15-18 years (2), children born at the beginning or during the war, we found that age group (2) had a higher reporting mean in all the fields, however those differences reach significance in reporting somatic problems (F (566)= 3.8, p .03), with the children born before the beginning and during the war having a higher mean (M = 2.3, SD = 2.5) in comparison to those born immediately after the war (M = 1.8, SD = 2.2). Source: Authors' own work A correlation was performed for the main YSR scales (total problems, internalizing problems and externalizing problems) with the income, parents' education background and the number of family members living _____________________________ Iliria International Review -2015/1 together. A significant negative, but not strong correlation was found only between income and internalizing problems (r= -.087, p = .03). Negative, but not significant correlation was found between the three scales and education background. We performed a regression analysis in order to find how much of a predictor income is for the internalizing problems R=.008. Only 0.8% of the variance of internalizing problems was explained by the income. If we add the child's gender in the model, we see that 1.7% of the variance of internalizing problems is explained by child's gender and income. Gender is a significant predictor (p =.002). We also found that anxiety problems in children [R 2 = .83, p < .001] were a significant predictor of internalizing problems. The assessment of the scale of positive qualities [R 2 = .19, p < .001] was also found to be a significant predictor for externalizing problems.

Conclusions
The aim of this study was to explore internalizing and externalizing problems in adolescent children of veterans of Kosovo war (1998)(1999). The results of this study are quite interesting from the perspective of the dilemma of the state of veterans' children of even 15 years after the war ended. In general, the results indicate that children born before and during the war have more problems than children born immediately after the war. Those results are also consistent with the results found in the community (Shahini et al 2014).
This study did not find that internalizing problems were more manifested among female children and that externalizing problems were more manifested in male children. Income and education background were not found to be predictors for internalizing and externalizing problems. One interesting data was that positivity was a significant predictor for internalizing and externalizing problems.
The Kosovo society has not treated veterans properly, leaving them in a very unfavourable situation. The results show that more than half of them live under the optimum of living conditions. They seem to have organized themselves within the family circle, disabling themselves to express emotional problems. This speculation may be supported by the fact that the study has found that adolescents of age 15-18 have more somatic problems than those of younger age. This fact, coupled with the lack of specific services for veterans and their families, makes one part of adolescents more prone to experiencing anxiety or more predisposed to developing emotional and behaviour problems.
Previous studies on children of veterans have given inconsistent results. Children of veterans according to these studies reported higher levels of conflict in their families but no significant differences on measures of psychological distress from control groups (Westerink and Giarratano 1999). Findings from Ruscio, Weathers, King, and King (2002) showed that emotional numbing was the only aspect of PTSD uniquely associated with veterans' perceived relationships with their children.
Considering that the subject of this study were adolescent children of veterans of Kosovo war in 1999 we must take into account that the posttraumatic stress disorder is a very frequent problem among war veterans and that its impact on their personal and family life cannot be overlooked. Due to the physiognomy of this disorder and the complexity of its interaction with the environment and the cultural context of the individual, this seems to be a field with abundant need for research, particularly of its impact on the second generation.