homework 2
kchnda
paper_2.docxs anxiety sensitivity a predictor of PTSD in children and adolescents?
· Emine Zinnur Kılıç a , ,
· Cengiz Kılıç b ,
Show more
http://dx.doi.org.ezproxy.fau.edu/10.1016/j.jpsychores.2008.02.013
Abstract
Objective
Anxiety sensitivity (AS) is the fear of the physical symptoms of anxiety and related symptoms. Longitudinal studies support AS as a vulnerability factor for development of anxiety disorders. This study aimed to investigate AS as a vulnerability factor in the development of childhood posttraumatic stress disorder ( PTSD ) following traumatic experiences.
Methods
The study included 81 children 8–15 years of age who experienced the 1999 earthquake in Bolu, Turkey. The earthquake survivors were compared to a randomized group of age- and sex-matched controls 5 years after the earthquake. Both the subject and control groups were administered the Childhood Anxiety Sensitivity Index (CASI), State and Trait Anxiety Inventory for Children (STAI-C), and Child Depression Inventory (CDI), while the PTSD symptoms of the subjects were assessed using the Child Posttraumatic Stress Reaction Index (CPTS-RI).
Results
Subjects and controls did not differ significantly in CASI, STAI-C, or CDI scores. Multiple regression analysis showed that both trait anxiety and CASI scores predicted CPTS-RI scores of the subjects; the prediction by CASI scores was over and above the effect of trait anxiety.
Conclusion
The results of this study support the hypothesis that AS may be a constitutional factor, which might increase the risk of PTSD following traumatic experiences.
Keywords
· Anxiety sensitivity;
· Child;
· Adolescent;
· PTSD
Introduction
Several factors are held responsible for the development and chronicity of posttraumatic stress disorder ( PTSD ) in children and adolescents exposed to traumas. These include within-trauma factors such as trauma severity, and pretrauma factors including gender and age [1] . Biological vulnerability factors, such as anxiety sensitivity, have also been shown to play such a role in adult samples. This study aims to show if the same relationship holds for children and adolescents.
The concept of anxiety sensitivity
Anxiety sensitivity (AS) is a concept that was originally developed by Reiss and McNally [2] to describe the degree of discomfort and negative attributions to anxiety sensations arising from the belief that these sensations are signs of physical, psychological, or social harm. Rather than being a sign of psychopathology itself, AS is seen as a constitutional trait variable, acting as a risk factor for the development of various anxiety disorders. The measurement of AS includes the assessment of the consequences of physical, cognitive, and social harm associated with anxiety symptoms, such as body awareness and feared cognitions. The Anxiety Sensitivity Index (ASI) [3] is the principal measure used to assess AS in adults.
Studies that have used ASI in adults have generally examined panic disorder and most found a positive association between the presence of panic disorder and high ASI scores [4] and [5] . Subjects with high AS have higher rates of comorbid anxiety disorders and tend to be susceptible to marked anxiety reactions following biological challenge tests [6] and [7] . Evidence from several longitudinal investigations suggests that AS is a predictor of subsequent panic attacks in nonclinical samples [8] .
Several studies have attempted to show the mechanisms that link AS to specific anxiety disorders. Avoidance has been suggested to be a clinically significant mechanism that might play a role in maintaining anxiety disorder and to be associated with poorer diagnosis. Wilson and Hayward [9] , in their prospective study on adolescents with panic disorder, showed that AS may precede and exacerbate avoidance, which, in turn, increases anxiety. Isyanov and Calamari [10] , on the other hand, hypothesized that individuals with high levels of AS appraise many life events as much more stressful due to their reactivity to both the event and to the anxious arousal they experience. They suggest that AS might lead to increased general stress levels by altering stress appraisal. Since they appraise anxious arousal as not just unpleasant, but as dangerous, life experiences that lead only to a transitory increase in anxiety will elicit greater distress in individuals with elevated AS. While individuals with high trait anxiety will respond to stressors with more fear and distress, individuals with high AS may react to both the stressor and the associated anxiety experience. Additional evidence exists that demonstrating AS is not merely a reflection of trait anxiety, but it is shown to be related to disorders other than anxiety, namely somatoform disorders and substance use disorders [11] , [12] and [13] .
Anxiety sensitivity and childhood disorders
Children with anxiety disorders have been shown to have higher levels of AS compared to normal controls in several studies [14] and [15] . A number of studies have reported a relationship between panic attacks in children and adolescents and AS [16] , [17] and [18] . Reiss et al. [5] suggested that the level of AS at ages 7-14 years might predict the development of panic disorder in adult life. According to their theory, the development of AS is influenced by cognitive factors, in addition to genetic ones. By the time children become 7-10 years old, they will have already developed beliefs about what will happen to them when they become nervous or experience stress. These beliefs are hypothesized to significantly modify the child's inherited sensitivity to anxiety [5] . Likewise, Mattis and Ollendick [19] , in their investigation of the cognitive responses of children to panic symptoms, concluded that AS in children predicted catastrophic attributions, regardless of age, and speculated that high levels of AS and elevated internal attributions in response to negative outcomes could set the stage for the development of panic attacks and subsequent panic disorder.
AS has been investigated in various other problems of children and adolescents. Although AS was found to be related to depression [20] and worry [21] , AS predicted panic attacks in children and adolescents, even after controlling for general negative affectivity [22] . The predictive value of high AS for panic attacks was also shown in an adolescent African-American sample [23] . Watt and Stewart [11] investigated this concept further, positing that heightened AS in childhood may lead children learn to catastrophize their bodily sensations, leading to hypochondriacal symptoms in adolescence. The study of Hayward et al. [18] , which showed that AS develops before panic symptoms first appear, supports the conceptualization of AS as a risk factor for anxiety disorders, rather than being a disorder itself. AS, therefore, seems to be a constitutional factor creating the basis for the development of an anxiety disorder in the presence of an external challenge.
AS and PTSD
Although AS has been widely studied in childhood anxiety disorders, studies on AS in PTSD are scarce. In adults, the relationship of AS and PTSD has been shown in various studies [24] . After the Bam earthquake in Iran, Hagh-Shenas et al. [25] found a relationship between high AS and PTSD among a group of rescue workers. They reported that among the untrained rescue workers, students with high AS scores exhibited greater adverse psychological effects.
PTSD is common in children after disasters; many studies report incidence rates ranging from 30–60% in children and adolescents [26] . The notion that PTSD symptoms decrease with time has been challenged by long-term studies. PTSD may become chronic in a group of disaster survivors, which, in turn, may significantly affect psychological development [27] and [28] . In their study, after the 1999 earthquake in Turkey, Karakaya et al. [29] found very severe or severe degrees of posttraumatic stress symptoms in 22.2% of adolescents 3.5 years after the disaster.
Risk factors for the occurrence and persistence of PTSD symptoms in children after disasters are widely studied. Proximity to the trauma scene and severity of exposure (loss of house/relatives, injury, etc.) are commonly found to be risk factors [30] . Research has shown that parental reactions to a child's symptoms and parents' symptoms are also important predictors of childhood PTSD [31] , [32] , [33] , [34] , [35] , [36] and [37] . The risk of developing PTSD depends upon the severity of trauma, preexisting vulnerability factors, and an interaction between the two. The “stress vulnerability” hypothesis holds that pretrauma characteristics make one more susceptible to the negative effects of a traumatic experience. It is typically seen as an interaction between the predisposing factor and the traumatic stressor, so that the relationship between the predisposing factor and PTSD depends on the level of trauma. In the context of extreme trauma, host factors may diminish in importance, whereas in milder trauma vulnerability factors may be of greater importance. Foy et al. [38] and McCranie et al. [39] demonstrated this form of interaction. Silva et al. [40] have demonstrated that preexisting anxiety predicted PTSD severity in children and adolescents, while higher IQ was a protective factor. AS, which has been implicated as a constitutional factor in the development of several other anxiety disorders, may be the mediating factor that leads to the development and persistence of posttraumatic stress symptoms following psychological trauma. To our knowledge, the only study to assess the relation of AS to PTSD symptoms in children was conducted by Meiser-Stedman et al. In their study on children who had had individual traumatic experiences, although AS did not directly predict PTSD status, it mediated the relationship between subjective distress during trauma and the development of acute stress disorder [41] . It will be interesting to see if a similar relationship exists between AS and chronic PTSD. It may be hypothesized that since children with high levels of AS are more sensitive to anxiety-related sensations, they will have more difficulty in overcoming the effects of traumatic experiences.
This study aimed to investigate the relationships between AS and PTSD symptoms in a group of children that experienced a major earthquake in Bolu, Turkey. Bolu is a small town that had a pre-earthquake population of 87,000 and lies 30 km away from the epicentre. The November 12, 1999 (magnitude 7.2), earthquake killed 48 people and injured 343 in Bolu, while 2400 houses were damaged. Families had to live in tent cities for up to 2 years after the earthquake and fear of future earthquakes continued for several months. The rate of severe and very severe degrees of posttraumatic stress symptoms in children living in one tent city was 18.8%, 6 months after the earthquake [37] .
The present study was conducted 5 years after the 1999 Bolu earthquake, at a time when life in the city had returned to normal and issues about the earthquake were largely forgotten, except perhaps for temporary anniversary reactions. We hypothesized that the severity of PTSD symptoms in earthquake-survivor children, 5 years after the earthquake, would at least be partly explained by higher AS scores.
Methods
Subjects
A sample of 87 children (43 boys and 44 girls) aged between 8 and 15 years (mean: 11.2 S.D. 2.2) were recruited from the Bolu city centre. One hundred ninety-one households were randomly selected from a list of all households at the city centre. All children living in those households between ages 8–15 were invited to participate in the study. Children who did not experience the earthquake or who were unable to complete the study measures due to a physical or mental problem were excluded from the study. Six children were excluded from the study due to incomplete data on their questionnaires. The final sample thus consisted of 81 children (40 boys and 41 girls).
Controls
The control group (n=87) was randomly selected from a larger sample of age- and gender-matched children from Ankara schools. This group contained 43 boys and 44 girls aged between 8 and 15 years (mean: 11.1, S.D. 2.3) that had not experienced the earthquake and did not report any traumatic experience in their lifetime, nor did they report any psychiatric problem.
Measures
Both the subject and control groups were administered the Childhood Anxiety Sensitivity Index (CASI), State and Trait Anxiety Inventory for Children (STAI-C), Child Depression Inventory (CDI), and a sociodemographic form, while the PTSD symptoms of the subjects were assessed using the Child Posttraumatic Stress Reaction Reactions Index (CPTS-RI).
CASI is a self-report tool developed by Silverman and colleagues that assesses AS in children [42] ; it is a modified form of the Anxiety Sensitivity Index (ASI) [3] . CASI includes 18 items and a three-point Likert-type scale ranging from 1 (not at all) to 3 (very much). The reliability study of the Turkish version was conducted by Yılmaz and Kılıç with a group of Turkish school children [43] . In that study, the scale demonstrated reasonable reliability, with a Cronbach's alpha of 0.74 and a test–retest reliability of 0.77.
CPTS-RI is a 20-item self-report scale designed to assess the posttraumatic stress reactions of school-age children and adolescents [44] . Pynoos [27] used a revised method in determining caseness, which we also adopted. They found that the measure's severe and very severe categories correctly identified 78% of subjects who met Diagnostic and Statistical Manual of Mental Disorder ( DSM ), Revised Third Editioncriteria for PTSD. The reliability and validity study of the Turkish version was conducted with primary school children that survived an explosion in Turkey [45] . Test–retest reliability was .86 and internal consistency (alpha) was .75. The scale captured 80% of DSM, Fourth Edition PTSD cases.
STAI-C is a 40-item self-report questionnaire that assesses anxiety-related symptoms in children [46] . The scale computes 2 separate total scores: state anxiety and trait anxiety. The validity and reliability study of the Turkish version was conducted with a primary school sample by Özusta [47] . Test–retest reliability of state anxiety scale was .60 and internal consistency (alpha) was .82. Test–retest reliability of the trait anxiety scale was .65 and, internal consistency, (alpha) .81. STAI-C Turkish version significantly differed children with anxiety disorders from normal controls.
CDI is a 21-item self-report questionnaire that assesses depressive symptoms in children [48] . The validity and reliability study of the Turkish version was carried out by Öy in Turkish children [49] . The reliability of the scale (alpha) was .77, and test–retest reliability was .80. The sensitivity of the scale was 60%, whereas the specificity was 95%.
Data analysis
Statistical analyses were made using SPSS v.10.0. The two study groups were compared using t tests. Pearson's correlation coefficients and linear regression were used to examine the relationship between PTSD scores and other demographic and clinical variables. Ethical approval was obtained from Ankara University Medical School Ethics Committee.
Results
No significant differences were found between the subjects and controls in terms of age or gender. Comparing subjects to controls, there were no significant differences in their mean CASI, STAI-C, and CDI scores ( Table 1 ).
Table 1.
Comparison of subjects and controls, in terms of mean CASI, STAI-C, and CDI scores (t tests)
|
|
Subjects (Mean, S.D.) |
Controls (Mean, S.D.) |
t |
|
CASI |
32.4 (7.3) |
31.1 (6.2) |
1.68 (NS) |
|
Trait anxiety |
35.4 (7.3) |
34.1 (6.4) |
1.24 (NS) |
|
State anxiety |
32.7 (6.5) |
31.9 (6.5) |
0.8 (NS) |
|
CDI |
10.8 (7.2) |
9.7 (3.9) |
1.2 (NS) |
NS, nonsignificant; CASI, Childhood anxiety sensitivity index; STAI-C, State and trait anxiety inventory for children; CDI, Child Depression Inventory.
When the subjects were grouped according to the level of CPTS-RI PTSD symptoms as nonsymptomatic (<12 points), mild PTSD (12–24 points), moderate PTSD (25–39 points), severe PTSD (40–59 points), and very severe PTSD (>60 points) in accord with the original form of the scale [44] , 10% did not have PTSD, 40% had mild PTSD, 33% had moderate PTSD, 16% had severe PTSD, and 2.5% had very severe PTSD. There were 42 subjects with moderate to very severe PTSD (PTSD subgroup), whereas nonsymptomatic subjects and subjects with mild PTSD (non-PTSD subgroup) numbered 39. Table 2 shows that the PTSD and non-PTSD subgroups differed significantly in terms of depression, state and trait anxiety, and AS scores; those with more severe PTSD symptoms had higher scores on all measures than the non-PTSD subgroup.
Table 2.
Comparison of the PTSD and non-PTSD subgroups, in terms of age, CDI, CASI, and trait and state anxiety mean scores (t tests)
|
|
Moderate-Very Severe PTSD (mean, S.D.) (n=42) |
Non-PTSD (mean, S.D.) [n=39] |
T |
P |
|
Age |
11.0 (2.1) |
11.4 (2.3) |
1.0 |
.3 |
|
CDI |
13.6 (8.6) |
8.1 (3.7) |
−3.7 |
.01 |
|
CASI |
35.6 (7.7) |
29.4 (5.6) |
−4.1 |
.01 |
|
Trait anxiety |
39.1 (7.5) |
31.9 (4.6) |
−5.3 |
.01 |
|
State anxiety |
35.2 (6.7) |
30.6 (5.8) |
−3.3 |
.01 |
The correlations between CPTS-RI, CASI, STAI-C, and CDI scores in the subject group are shown in Table 3 , and Table 4 shows correlations between CASI, STAI-C, and CDI in the control group. AS had significant positive correlations with state and trait anxiety in both groups; correlations with trait anxiety were higher. The correlations with CDI did not reach significance. CPTS-RI scores had significant positive correlations with all study measures in the subject group. The correlations ranged between 0.38 and 0.71; the highest correlation was with trait anxiety.
Table 3.
Correlations between CPTS-RI, CASI, state anxiety, trait anxiety, and depression scores of the subject group (Pearson's correlations)
|
|
CASI r(P) |
State anxiety r(P) |
Trait anxiety r(P) |
CDI r(P) |
|
CPTS-RI |
0.38 (P<.01) |
0.47 (P<.01) |
0.71 (P<.01) |
0. 51 (P<.01) |
|
CASI |
– |
0.29 (P<.01) |
0.32 (P<.01) |
0.17 (P<.13) |
|
State anxiety |
– |
– |
0.58 (P<.01) |
0.60 (P<.01) |
|
Trait anxiety |
– |
– |
– |
0.65 (P<.01) |
CPTS-RI, Child Post-traumatic stress reaction index; CASI, Childhood anxiety sensitivity index; CDI, Child Depression Inventory.
Table 4.
Correlations between the CASI, state anxiety, trait anxiety, and depression scores of the control group (Pearson's correlations)
|
|
State anxiety |
Trait anxiety |
CDI |
|
CASI |
0.22 (P<.05) |
0.44 (P<.01) |
0.18 (P<.10) |
|
State anxiety |
– |
0.40 (P<.01) |
0.38 (P<.01) |
|
Trait anxiety |
– |
– |
0.28 (P<.01) |
Multiple regression analysis was performed within the subject group in order to show if AS played a unique role in determining CPTS-RI scores. The independent (explanatory) variables were age, gender (1=male, 2=female), CASI score, CDI score, state anxiety score, and trait anxiety score. The dependent variable was CPTS-RI score. The explanatory variables were simultaneously entered into the regression equation (method=enter). The result of this analysis showed that higher PTSD scores were predicted independently by both CASI and trait anxiety scores, and the prediction by trait anxiety was stronger ( Table 5 ). This finding shows that the prediction of PTSD scores by CASI scores was over and above the effect of trait anxiety.
Table 5.
Predictors of PTDS scores (multiple linear regression)
|
CPTSD-RI scores
|
||
|
Full regression model (r=0.74; adjusted r2=0.52; F(15,104)=15.1; P<.001)
|
||
|
Significant predictors |
β |
P |
|
Trait anxiety |
0.55 |
.001 |
|
Anxiety sensitivity |
0.21 |
.05 |
|
Gender |
−.011 |
.185 |
|
Age |
−.03 |
.715 |
|
CDI |
.09 |
.40 |
|
State anxiety |
.07 |
.52 |
Entered variables: gender, age, state anxiety, trait anxiety, depression, and AS.
Discussion
To the best of our knowledge, this is the first study to examine the relationship of AS to chronic PTSD symptoms in children. We compared a group of children and adolescents who had experienced the 1999 Bolu earthquake to a control group with no history of traumatic experiences, in terms of depression , anxiety, and AS scores. These two groups did not differ in terms of their mean AS, depression, or anxiety symptom scores. Although we did not find any difference between the subject and control groups in terms of AS scores, the subjects with more severe PTSD symptoms had higher AS, depression, and anxiety scores than the subjects with less severe PTSD symptoms. Linear regression analyses, taking CDI, CASI, and STAI-C as independent variables, showed that AS and trait anxiety scores, but not state anxiety or depression scores, predicted PTSD. These results support the hypothesis that AS might be a constitutional trait with a normal distribution among the general population (traumatized or not), which may act as a vulnerability factor leading to psychopathology during times of stress.
The strong correlation we found between trait anxiety and AS has been reported in other studies, which led some researchers to conclude that AS is, in fact, trait anxiety. This idea has been questioned by McNally [50] , who suggests that while trait anxiety predicts a general propensity to respond anxiously to threatening stimuli, AS predicts a propensity specific to the symptoms of anxiety. Reiss [51] also argued that, in trait anxiety, the feared stimulus is regarded as dangerous, while in AS, an uncontrollable reaction is feared. Since various studies have made this distinction, Reiss et al. [5] conclude that AS is theoretically and empirically distinct from Spielberger's measure of trait anxiety. The fact that the two variables independently predicted PTSD symptoms supports the above notion that AS and trait anxiety are actually distinct constructs.
Not all individuals develop PTSD after traumas. Various factors lead to the development and persistence of PTSD symptoms. Individuals with high levels of AS are shown to appraise many life events as much more stressful due to their reactivity to both the event and to the anxious arousal they experience. They are therefore more vulnerable to the effects of stressful life events and to trauma reminders. Life experiences that lead only to a transitory increase in anxiety will elicit greater distress and catastrophic attributions in such individuals, leading to the persistence of trauma-related stress symptoms.
Many studies demonstrate that the most persistent symptoms of PTSD belong to the group of avoidance [52] and [53] . Avoidance is also known to relate to higher decrease in quality of life in trauma studies. Since individuals with high AS are more likely to respond with greater distress to stressful to life events, it is reasonable to assume that they will tend more to avoid events or situations that provoke anxiety. Avoidance may also prevent improvement by removing the chance of exposure. This avoidant attitude in children with high AS, therefore, may be responsible for the persistence of PTSD symptoms.
The findings of our study may have implications on both detection and management of high-risk groups. Trauma survivors with high AS may be less motivated to volunteer for anxiety-provoking treatments such as cognitive behavior therapy ( CBT ), and they may avoid coming into contact with treatment services or drop out of treatment prematurely. Clinicians with an awareness of the significance of AS will, therefore, have an advantage in engaging high AS trauma survivors into treatment.
Some limitations of our study should be mentioned. Although our study sample was representative of a town at the epicentre of the November 1999 Bolu earthquake, the small sample size did not allow for factor analytic studies of CASI, which would have provided additional information contributing to an understanding of how AS is related to PTSD. Future studies with larger samples could overcome this limitation. A second limitation was the cross-sectional nature of data collection. Even if it is possible to control for the contribution of other variables on outcome, the link of causality cannot be drawn in cross-sectional studies. It is, therefore, possible that higher AS may be the result of long-lasting PTSD symptoms in the affected individuals, instead of causing it. Longitudinal studies are therefore needed to clarify the role of AS in the development of PTSD following psychological trauma.
