Posttraumatic stress disorder and psychosocial difficulties among children living in a conflict area of the Southeastern Anatolia region of Turkey.

BACKGROUND: Traumatic events and armed conflicts can lead to many mental disorders, especially posttraumatic stress disorder (PTSD), in children. We investigated the PTSD symptoms, psychosocial difficulties, general health levels of the parents, and direct and indirect effects of trauma on psychological outcomes.
MATERIALS AND METHODS: A total of 482 children were included in the study. Two hundred and eight of those children were from Derik district, where street fights, curfews, and conflict have been experienced, and 274 of those children were from Yeşilli district, where no conflicts have been observed despite being in the same province. All children filled out the child posttraumatic stress disorder reaction index, the strengths and difficulties questionnaire (SDQ), and parents filled out the General Health Questionnaire-12 (GHQ).
RESULTS: All children showed moderate PTSD symptoms; however, no significant difference was observed between the two groups. In the exposed group, children, who had to leave their homes for a while due to conflicts, had worse PTSD symptoms, higher SDQ difficulty scores, and parents' GHQ scores. It was determined that being a female having high maternal GHQ scores and leaving home due to the conflicts significantly increase the risk of occurrence of trauma symptoms.
CONCLUSION: Our findings suggest that children are susceptible to the direct and indirect effects of trauma. It is crucial for intervention programs to be developed for the detection, prevention, and treatment of PTSD symptoms to be applied to all children, regardless of exposure type, in areas affected by conflict. Copyright:

INTRODUCTION

Over the past three decades, there was a conflict between the Turkish Military and Kurdistan Worker's Party, known as the Partiya Karkeren Kurdistan (PKK), in the Eastern and Southeastern Anatolia regions of Turkey. PKK is listed as terrorist organization by Turkey and several states and organizations. During this period, PKK was responsible for the vast majority of terrorism-related events in these regions. Due to the violence and conflict environment that has been experienced over 30 years, thousands of people have lost their lives and forced migrations have been realized in the region. As expected, the biggest victims of continuous terror and violence were children. Children were sometimes forced into crime by illegal groups, politicized and left to be gotten into the spiral of violence, and directly exposed to conflict. The conflicts that decreased in the Southeastern Anatolia region of the country between 2012 and 2015 have reincreased since June 2015. In the conflicts that lasted for about 1 year, a curfew was declared in some cities and districts, especially where Kurdish populations were high, due to terrorist actions. Serious armed conflicts took place in regions in which the curfew was declared, and people who lived in this region also witnessed the events.

Whether directly or indirectly, children are always the first to be affected by armed conflicts. Traumatic events and armed conflicts can lead to many mental disorders or symptoms, especially posttraumatic stress disorder (PTSD), in children. It is also known that children who are exposed to the negative effects of situations such as war and conflict live with feelings of fear and desperation. Regressing to previous stages of development in younger children is also common. It has been reported that young children may be a little more at risk for the psychological effects of trauma due to their continuing emotional and psychological development, limited coping skills, and dependency on their parents.[1]

Problems associated with trauma can lead to disability or loss of function in children. Traumatic stress and related problems emerge in a considerable percentage of those indirectly affected by the event and emerge more in those who face the trauma directly.[23456] Children, who are directly or indirectly exposed to warfare, experience various stressors and exhibit short- and long-term posttraumatic stress reactions.[7] Symptoms and reactions that commonly occur after a traumatic event are sadness, fear, numbness, feeling nervous or timid, moodiness or nervosity, changes in appetite, difficulty sleeping, nightmares, avoidance of situations reminding them of trauma, focusing problems, and guilt.[89] Studies have shown that there are distinct and detectable symptoms of PTSD in young children, including fear of harm to the physical integrity of the self and others as a result of traumatic experiences.[1011] Children who were repeatedly exposed to war have similarly emphasized the effect of long-term wartime experiences on well-being, emotional reactivity, and mental and physical health.[121314] In previous studies conducted in countries that were exposed to conflicts, it was stated that PTSD is observed at high rates in children. For example, the percentages of PTSD diagnosis vary from 22% to 25% among Israeli[1516] and 27% among Lebanese children,[17]41% among Palestinian children[18] from Gaza, 48% among Cambodian refugee children,[19] and 52% among children from Bosnia-Herzegovina.[20]

It is not surprising to think that, among the factors affecting PTSD development in children, the importance of family is great. Therefore, parental and familial factors have been investigated in studies. It has been stated that the family cohesion[2122] and positive family environment[23] of children who are exposed to war/conflicts could be protective from the effect of trauma. Furthermore, the presence of PTSD in parents is also a risk factor for PTSD development in children.[2425] Factors such as environmental medium, the physical and mental status of parents, parent-child emotional interaction, and the social support of the family may be considered effective factors in determining the difficulties in posttraumatic period.

In the light of the above-mentioned information, the aim of this study is to examine the PTSD-related symptoms, psychosocial difficulties of children, and general health levels of their parents in the Derik district of Mardin province, where the conflicts have lasted for a long time and curfews due to the conflicts have been declared. In addition to this, we aimed to investigate the relationship between direct and indirect effects of trauma on psychological outcomes and PTSD. The control group was consisted of children who live in Yeşilli district, 84 km from Derik district, where no conflicts have been observed despite being in the same province. Our hypotheses were as follows: (1) children living in a conflict area exhibit PTSD symptoms such as reexperiencing, avoidance, and stimulation; (2) children who have not been directly exposed to the conflict may exhibit subthreshold PTSD symptoms due to living in the region; (3) exposed children exhibit more psychosocial difficulty than nonexposed children, and this is correlated with the PTSD symptoms; and (4) parents of the exposed group have more risk of mental illness than the nonexposed group. According to our knowledge, our study is the first to be conducted on children living in the Southeastern Anatolia region of our country, where armed conflicts are experienced.

MATERIALS AND METHODS

Participants and procedure

Our study was conducted in Mardin Province in the Southeastern Anatolia region of the country. In this province, where people from different ethnicities (Turkish, Kurdish, Arabic, and Assyrian) live, while conflicts were intense in some regions, there were no conflicts in other regions. When the study was conducted, street fights and curfews in Derik had not been realized for 6 months. The sample consisted of 482 children, 208 from Derik and 274 from Yeşilli, and as well as their mothers. A stratified random sample design was used to draw a sample. In Derik Province, we determined 12 middle schools and approximately 2000 children. We also determined eight middle schools and almost 1500 children in Yesilli Province. Five schools from exposed group and four schools from the nonexposed group were selected randomly. The design for the sample selection was based on three primary stratified variables as follows: gender, age, and grade level. The design and procedure were approved by Mardin Governorship and Ethics Committee of Gazi Yasargil Training and Research Hospital. After approval, we went to the schools in both districts the day before the study was conducted, and teachers, school administrators, and students were informed about the study. The consent forms were sent to parents through the teachers. Parent's response rate to participate in the study was 62%. Parents who agreed to participate in the study filled out General Health Questionnaire 12 (GHQ). Furthermore, all children who were included in the study filled out the Child PTSD Reaction Index, the Strengths and Difficulties Questionnaire (SDQ), and the sociodemographic data form.

Data collection tools

Child posttraumatic stress disorder reaction index

The Child Posttraumatic Stress Disorder Reaction Index (CPTSD–RI) is developed to evaluate the stress reactions occurring in children and adolescents as a result of various traumatic experiences.[26] The scale consists of 20 items. Children rate the frequency of symptoms using a 4-point Likert scale from 0 (none) to 4. It is also used to investigate the severity of PTSD symptoms. A score of 12–24 is mild, 25–39 is moderate, 40–59 is severe, and above 60 are extreme PTSD symptoms. A score of 40 and above was shown to be correlated with clinical PTSD diagnosis.[27] The Turkish validity-reliability study was conducted, and Cronbach's alpha value was determined as 0.75.[28]

Strengths and difficulties questionnaire

The SDQ was developed to be used in screening mental problems in children and young people.[29] The SDQ includes 25 questions, and these questions are gathered in five subheadings. These headings are behavioral problems, attention deficit and hyperactivity, emotional problems, peer problems, and social behaviors. As each heading is evaluated within itself, the sum of the first four headings gives the total difficulty score. It was stated that the Turkish version of the scale is consistent and reliable, and Cronbach's alpha value was determined as 0.84–0.22 range.[30]

General health questionnaire-12

The GHQ is a scale developed, especially to examine primary care mental disorders.[31] The 12-question GHQ is commonly preferred due to being short and having a high sensitivity and specificity to distinguish cases. The probability of psychiatric disorders increases as the score increases on the scale, which was translated into Turkish and for which validity and reliability studies were conducted.[32] Sensitivity and specificity of the scale were determined as 0.74 and 0.84, respectively.[32] The highest score that can be obtained from the scale is 12, and the lowest score is 0. Those who score 4 or above on the scale are considered high, those who score 2–3 are considered moderate, and those who score <2 are considered low.

Sociodemographic data form

This form was created by the authors, based on the literature, to collect information about the sociodemographic characteristics of children. The form contains information such as the age, gender, number of siblings, and the school achievement of the child, and the age, education levels, occupational status and income levels, and marital status of the parents.

Statistical analyses

Data were evaluated using IBM SPSS statistics software version 22 (IBM Corp., Armonk, NY, United States of America). The measured variables were expressed as median ± standard deviation, and the categorical variables were defined as percentage and number. The distribution of numeric variables was evaluated using the Kolmogorov–Smirnov test and by evaluating histograms. The comparison of the normal distribution of numeric variables was evaluated using the Student's t-test, and nonnormally distributed numeric variables were evaluated using the Mann–Whitney U test. The categorical variables were evaluated using the Pearson's Chi-square test and Fisher's exact test. Pearson's test was used to identify the direction and the level of the association in numeric variables, and Spearmen's correlation test was used for nonnormally distributed numeric variables. The effect size was measured with Cohen's d. Cohen's d-values of 0.2 or below reflect a small effect size, around 0.50 reflect a moderate effect size, and 0.80 and above reflect a large effect size. Logistic regression analyses were conducted to assess the possible effects of independent variables on children's PTSD symptoms. Odds ratios and 95% confidence intervals (95% CIs) were computed. The value of statistical significance was determined as P < 0.05.

RESULTS

A total of 482 children, 94 males and 114 females in the exposed group and 137 females and 137 males in the nonexposed group, were included in the study. The evaluation of the two groups regarding the age, gender, couple status of the parents, education levels and work status of the parents, and family income is summarized in Table 1.

Table 1

Sociodemographic features

	Study group (n=208), n (%)	Control group (n=274), n (%)	P
Age (mean±SD)	13±0.3	13.2±0.4	0.1*
Sex
Male	94 (45.2)	137 (50)	0.295**
Female	114 (54.8)	137 (50)
Marital status
Together	183 (100)	238 (99.6)	0.381**
Divorced/living separately	0	1 (0.4)
Mother’s occupation
Working	25 (12.1)	20 (7.4)	0.064**
Not working	181 (87.9)	249 (92.2)
Mother
Alive	199 (97.5)	267 (100)	0.015**
Dead	5 (2.5)	0
Father’s occupation
Working	150 (72.5)	217 (80.1)	0.51**
Not working	57 (27.5)	54 (19.9)
Father
Alive	189 (95.5)	255 (98.8)	0.25**
Dead	9 (4.5)	3 (1.2)
Had to leave their home
Yes	91 (44)
No	117 (56)

*Student’s t-test **Chi-square test. SD – Standard deviation

No significant difference was observed between the exposed and nonexposed groups in PTSD scores [Table 2]. As the scores obtained from the CPTSD–RI increase, it is considered that the function field is problematic. There was also no significant difference in PTSD scores between genders in the exposed group (P > 0.05).

Table 2

Comparison of child posttraumatic stress disorder reaction index scores between groups

	Mean±SD		P
	Study group (n=208)	Control group (n=274)
CPTSD-RI	28.9±15.3	31.9±17.4	0.09*

*Student’s t-test. CPTSD-RI – Child Posttraumatic Stress Disorder Reaction Index; SD – Standard deviation

In the exposed group, 34.8% of children exhibited mild (CPTSD–RI score of 12–24), 34.8% exhibited moderate (CPTSD–RI score of 25–39), 19.9% exhibited severe (CPTSD–RI score of 40–59), 4.3% of children exhibited extreme PTSD symptoms, and only 6.4% of the children had no symptoms of trauma (CPTSD–RI scores <12). Furthermore, 24.3% of these children in the exposed group had 40 and above trauma scores, which indicates that this scale is correlated with PTSD. This ratio was 28.7% when considering all children in the study.

The SDQ total difficulty score was compared with the Mann–Whitney U test because it did not exhibit a normal distribution. There was no significant difference in the total difficulty score between the two groups (Mann–Whitney U test, Z = –0.603, P = 0.547; study group: mean = 12, SD = 5.1 and control group: mean = 12.6, SD = 6.4).

The GHQ total scores of the mothers of the exposed and nonexposed groups were 2.4 ± 2.9 and 1.8 ± 2.5, respectively. The GHQ scores of the exposed group mothers were significantly higher (Mann–Whitney U test, Z = –2.220, P = 0.026; study group: mean = 2.4, SD = 2.9 and control group: mean = 1.8, SD = 2.5). It was determined that the GHQ mean scores of the exposed group mothers were above the threshold of 2.

In the sociodemographic data form, the exposed group was asked whether there were children near the conflicts in the city who had to leave their homes for a while. 91 (44%) of these children in the exposed group stated that they had to leave their homes for a while. In the exposed group, according to the children's responses, the PTSD symptoms, psychosocial difficulties, and the GHQ results of the mothers who had to leave their homes for a while due to conflicts were compared with children not leaving home. In the children who had to leave their homes, the PTSD symptoms, SDQ total difficulty score, and GHQ scores of the mothers were significantly higher [Table 3]. While children who had to leave their homes exhibited moderate trauma symptoms (CPTSD-RI score of 25–39), the children who did not leave their homes exhibited mild symptoms (CPTSD-RI score of 12–24).

Table 3

The scale scores of children who had to leave their homes and did not leave in the exposed group

	Exposed group		Cohen’s d	P
	Leaving home (n=91)	Not leaving home (n=117)
CPTSD-RI	33.1±14.4	23.2±12.1	0.74	<0.001*
SDQ total difficulties score	12.8±5.2	10.0±3.8	0.61	0.001**
GHQ	2.8±3.0	1.7±3.0	0.37	0.021**

*Student’s t-test; **Mann–Whitney U-test. CPTSD-RI – Child Posttraumatic Stress Disorder Reaction Index; GHQ – General health questionnaire; SDQ – Strengths And Difficulties Questionnaire

Correlations

A correlation was observed between the trauma scores and the total difficulty score and the GHQ total scores of the parents in the exposed group. There was a significant correlation between PTSD scores and SDQ total difficulty score (P < 0.001, Spearman's rho = 0.491). Furthermore, there was a significant correlation between the PTSD scores and GHQ (P = 0.004, Spearman's rho = 0.268). Despite there was a low-moderate correlation between PTSD and GHQ and SDQ scores, our findings show that trauma affects the psychological well-being of children and mother's mental health.

Logistic regression analyses

In the exposed group, binary logistic regression analyses were performed to determine the factors affecting the PTSD. Since the mean trauma scores of the children were determined to have exceeded the threshold value, the children were divided into two groups as follows: those exhibiting severe and extreme PTSD symptoms (CPTSD-RI score ≥40), which indicates that this scale is correlated with PTSD and not exhibiting them (CPTSD-RI score <40). The PTSD symptom indicators were analyzed to determine the risk factors affecting the PTSD symptoms in these children. In the regression model, children's PTSD severity (0; CPTSD-RI score <40, and 1; CPTSD-RI score ≥40) was added as a dependent variable. SDQ scores, parental GHQ score (0; GHQ score <2, and 1; GHQ scores ≥2), and leaving home due to armed conflict and sociodemographic characteristics such as age, gender, and parental status were added to the model as independent variables. Stepwise regression analyses were used to reduce the number of predictors that influence dependent variables and to find the independent variable that is the best predictor of PTSD symptoms. As a result of the analysis, in the final model, it was determined that being a female, having high maternal GHQ scores, having to leave home due to conflict and high total SDQ scores were significantly related to increased the risk of occurrence of trauma symptoms [Table 4]. In the same evaluation, it was determined that sociodemographic characteristics such as the age of the children, whether their parents work or not, and their education and marital status did not relate to increase the risk of PTSD [Table 4].

Table 4

Final model of logistic regression analyses for predicting posttraumatic stress disorder

	B	SE	Wald	Significant	Exp(B)	95% CI for Exp(B)
						Lower	Higher
Sex	0.33	0.34	4.22	0.03*	1.93	1.24	3.72
Moderate and above GHQ	0.55	0.27	4.54	0.03*	1.71	1.10	3.24
Leaving home	0.67	0.34	3.77	0.04*	1.95	0.98	3.87
SDQ total difficulties score	0.35	0.31	15.4	0.001*	1.87	1.10	2.45
Mother’s education	−0.10	0.34	0.99	0.75	0.90	0.55	1.95
Mother’s occupation	0.12	0.55	0.50	0.82	0.82	0.15	3.37
Father’s education	0.12	1.02	0.02	0.90	1.13	0.15	8.38
Father’s occupation	0.37	0.45	0.67	0.41	1.45	0.59	3.53
Age	0.54	0.43	1.59	0.20	1.73	0.73	4.04
Constant	−6.72	5.77	2.41	0.04	0.7

*P<0.05. B is the unstandardized B regression coefficient; SE is the standard error of the regression coefficient; Wald is the Chi-square test value; Significant is the P value of the Wald Chi-square test; Exp(B) is the exponential of the regression coefficient; odds ratio. GHQ – General health questionnaire; SDQ – Strengths And Difficulties Questionnaire; CI – Confidence interval

DISCUSSION

In this study, we compared children exposed directly and indirectly to the armed conflicts after 6 months in the Southeastern Anatolia region of Turkey. The most important finding of our study was that the trauma symptoms were high in both groups. It was determined that being a female, the presence of both mother and child's psychosocial difficulties, leaving home due to conflicts and are among the factors that increase the risk of PTSD development. Contrary to expectations, trauma-related symptoms were not higher in the directly exposed children. It was determined that PTSD was observed at a rate of 30% in the children directly and indirectly exposed to trauma, and considering that our study was conducted 6 months after the conflicts, the symptoms tended to become chronic, and these children were thus thought to be at high risk for psychiatric comorbidities.

The CPTSD-RI was used to evaluate the trauma symptoms in our study. Both groups exhibited moderate PTSD symptoms. This finding suggests that children, unlike adults, are more susceptible to the indirect effects of trauma. The similar general stress levels in both groups suggest that children in the exposed and nonexposed groups were affected similarly in a psychosocial aspect. Exposure to war and political conflicts, as in other acute and chronic traumas, can lead to many mental health problems in adults and children, particularly PTSD, anxiety, and depression.[1433] In a study conducted in the Gaza Strip, it was demonstrated that children whose houses were destroyed by bombardments exhibited PTSD and phobic symptoms more often than children in nonbombardment areas.[34] In this study, it was determined in accordance with this finding that the PTSD development was more frequent in children who had to leave their homes. It is possible to think that children's leaving home, where they feel confident, accelerates the development of trauma-related symptoms. However, it should not be forgotten that families who have to leave their homes in areas where there is an intense conflict are exposed to more violent conflict than others. Therefore, it is difficult to determine whether leaving home or being exposed to a violent conflict increases the symptoms more. The fact that children living in refugee camps exhibit more PTSD symptoms than children living in rural and urban areas[3536] shows the effects of the negative changes in the routine lives and importance of home of children in the posttraumatic period.

It is known that the emotional reactions of parents are important but not the only factor in the emergence of posttraumatic reactions of children. Parent reactions can be affected by previous traumas they have experienced,[37] and parents may lead children to be adversely affected by changes in parenting skills or family functioning.[8] The increase in the GHQ scores of parents in our study supports this finding, which is among the predictors increasing the PTSD symptoms. It can be said that children whose parents in a better physical and mental condition may be less at risk for PTSD and other mental disorders in the posttraumatic period.

Many factors can affect PTSD development in children exposed to conflict. The age and gender of the child, maternal psychopathology, social support of the family, child's attachment pattern, and previous traumas experienced can be considered among these factors. The risk factors affecting PTSD development in our study were analyzed by logistic regression analysis, and it was determined that being a female, poor maternal and child's psychosocial health, and leaving home increased the risk of developing PTSD symptoms in children. Furthermore, the lower maternal GHQ scores in the nonexposed group were thought to suggest that parents may be more resistant to indirect trauma. In other words, children are more susceptible to developing PTSD when they are exposed to traumatic events, whether the trauma is direct or indirect. Increased parental GHQ scores increase PTSD symptoms in children indicates that the well-being and strength of parents shape the symptoms of children in the posttraumatic period. Similar to our study, the fact that distressed mothers are shown to be inadequate in responding to the emotional and physical needs of their children and helping their children to cope with emotional reactions to traumatic events[38] emphasizes the importance of the well-being of parents in the reactions of children to traumatic events. Since maternal PTSD symptoms were not examined in our study, the relationship between parental PTSD and PTSD of children was not examined.

Although among the previous studies conducted on children living in war-torn areas, there are studies showing that females are more susceptible to PTSD[203940] and there are also studies suggesting that PTSD symptoms are more common in males.[41] In our sample, even though PTSD symptoms were not greater in females, we found that females are more prone to developing PTSD. It could not be fully explained in the studies whether the gender differences in PTSD symptoms depend on the susceptibility of females to developing emotional symptoms or the previous emotional experiences they had or some other reason. As can be expected, the stress levels of children and the psychosocial difficulties of mothers were higher in the families who had to leave their home due to conflicts in the exposed group. In addition to this, there were more trauma symptoms in these children.

The fact that the PTSD symptoms were greater in children not affected directly by the conflicts is thought to be due to the media or parents' reactions to violence and conflict. It is also possible to think the effect of ethnicity in these results. The direct effect of the media and ethnicity were not examined in our study; however, it is thought that the media may play a role in the continuance of PTSD symptoms and that there is a need for more studies on this issue.

There are some limitations in this study. These include the facts that only the self-report scales were used, there were no structured psychiatric evaluation interviews, the factors among the sociodemographic characteristics predicting PTSD were not investigated further, and disorders except for PTSD, such as anxiety and depression, were not evaluated. Finally, the partial response rate of the parent may have an impact on outcomes by affecting sample homogeneity. Further studies are needed on the effect of continuing conflict and terrorist activities on the mental health, behavioral regulation, and developmental steps of children, and on intervention programs.

CONCLUSION

Our study results have contributed to the literature by mentioning the significant long-term reactions of children in the posttraumatic period in an armed conflict area. It drew attention to the importance of parental psychosocial health and changes in children's lives in the posttraumatic period. It is thought to be crucial for intervention programs to be developed, especially in school, which is an important place in children's lives, for the detection, prevention, and treatment of PTSD symptoms to be applied to all children regardless of direct and indirect exposure in the areas affected by conflicts. Otherwise, trauma-related conditions that have not been coped with appropriately during childhood may pose a risk of psychopathologies in adolescence and adulthood.

Following our study, an intervention program was created in the region where the study was conducted. For this purpose, informative training was provided to all family physicians and guidance teachers living in Mardin Province by child and adolescent psychiatrists about the usual reactions that children could have after childhood traumas and the attitudes that the family should exhibit in response to these reactions, as well as the problems that may occur in relation to the trauma, especially PTSD. Some 480 guidance teachers and 220 family physicians participated. In this training, which lasted for 16 days in total and comprised two 120-minute sessions each day, attempted to reach the professionals who personally interact primarily with children and their families.

Financial support and sponsorship

Nil.

Conflicts of interest

There are no conflicts of interest.