Brief Report: Factor Structure of the Childhood Illness Attitude Scales (CIAS)
1 Department of Psychology, University of Regina and 2 Faculty of Kinesiology and Health Studies, University of Regina, Regina, Saskatchewan
This paper is based on data collected as part of the master’s thesis of the first author, conducted under the primary supervision of the second author. Kristi D. Wright is now in the clinical psychology doctoral program at Dalhousie University. The reported findings have not been previously published.
All correspondence should be sent to Gordon J. G. Asmundson, Faculty of Kinesiology and Health Studies, University of Regina, Regina, Saskatchewan, Canada S4S 0A2. E-mail: gordon.asmundson{at}uregina.ca.
 |
Abstract |
|---|
 Top Abstract MethodResultsDiscussionReferences |
|---|
Objective To examine the factor structure of the Childhood Illness Attitude Scales (CIAS). The CIAS is a 35-item self-report measure based on the Illness Attitudes Scales, designed for use with school-age children. The CIAS measures fears, beliefs, and attitudes associated with health anxiety and abnormal illness behavior in childhood. Methods CIAS item responses for 201 school-age children were subjected to principal-components analysis with oblique rotation. Results The CIAS was best conceptualized as comprising four factors: fears, help seeking, treatment experience, and symptom effects. Further factor analysis supported the notion that the CIAS can also be conceptualized as having a hierarchical structure, with four lower-order factors loading onto a single higher-order factor of health anxiety. Conclusions Results suggest that the CIAS possesses good psychometric properties, including factorial validity and internal consistency, and appears to be a psychometrically sound instrument for measuring children’s health anxiety.
Key words: health anxiety; school-age children; factor analysis.
Health anxiety refers to apprehension and worry regarding one’s health. This anxiety is based on the misinterpretation of bodily sensations believed to be indicative of serious disease. Minor anxiety about health is ubiquitous and, in most situations, lessens as symptoms subside or upon reassurance from a medical professional that there is nothing physically wrong (Asmundson, Taylor, Sevgur, & Cox, 2001
). In some people, however, the anxiety can be excessive and become associated with clinically significant limitations to their functional ability. Illness phobia and hypochondriasis are the most common forms of clinically significant health anxiety.
It has been suggested that severe health anxiety begins in adulthood (American Psychiatric Association [APA], 2000). Not much more is known about the course of the phenomenon (Asmundson, Taylor, Wright, & Cox, 2001). In line with the above, the prevalence rate for severe health anxiety in children is unknown (Campo & Reich, 1999). Fritz, Fritsch, and Hagino (1997) assert that this may be related to the lack of specific diagnostic criteria designated for children and adolescents. Knowledge of this area is further hampered by the substantial lack of development of clinical assessment and screening tools designed to assess health anxiety in children and adolescents. To address this issue, Wright and Asmundson (2003) adapted the Illness Attitude Scales (IAS) (Kellner, 1987) for use with children aged 8 to 15 years. Their adaptation—the Childhood Illness Attitude Scales (CIAS)—consists of items from the IAS simplified to age-appropriate language; a simplified, 5-point to 3-point Likert scale; and the addition of seven questions to evaluate the role that parents/guardians play in facilitating medical attention or treatment.
Recent factor-analytic investigations of the IAS, utilizing both clinical and nonclinical samples, have consistently indicated a smaller number of factors (e.g., four or five lower-order factors) than the original nine factors proposed by Kellner (Hadjistavropoulos, Frombach, & Asmundson, 1999; Stewart & Watt, 2000). While the factor structure of the IAS has received considerable attention, comparable studies have not yet been conducted on the CIAS. Since children and adults often differ in their experience and presentation of psychological symptoms, and given that specific item content differs, it cannot be assumed that the factor structure of the CIAS is equivalent to that of the IAS. Examining the factor structure of the CIAS will provide additional insight into the construct of health anxiety as it relates to school-age children. This is important for several reasons. First, to the extent that each identified factor represents a distinct causal mechanism (Cattell, 1978), our results will inform our understanding of the mechanisms underlying childhood health anxiety and allow for comparison with those underlying expressions of health anxiety in adulthood. Second, an understanding of underlying mechanisms will serve to guide efforts to establish age-appropriate assessments and clinical standards for childhood expressions of excessive health anxiety. The purpose of the present study was, therefore, to examine the factor structure of the CIAS.
|
Method |
|---|
|
TopAbstractMethodResultsDiscussionReferences |
|---|
Participants and Procedure
Two hundred two school-age children from two Canadian public elementary schools participated. Complete data were provided by 201 of a potential 500 students (40.2% participation rate) and these data were used in statistical analysis. This sample size is adequate for factor-analytic studies (Tabachnick & Fidell, 2001
). All students over the age of 8 years who were attending the schools were eligible to participate. We did not assess reading level. The mean age of the sample was 10.9 years (SD = 1.6; range = 8–15 years), and approximately 56% (n = 112) were male. The majority of the sample was caucasian. Parental consent was obtained via a parental/proxy/legal guardian consent form that was sent home with each child and then returned to school. Participants completed a questionnaire package that included the CIAS and several other self-report measures (reported in Wright & Asmundson, 2003), as well as demographic questions. Questionnaires were completed independently in the classroom. Students who did not participate completed an alternate classroom activity with their teacher. All students were given a McDonald’s food voucher (approximate value = $1.50) following their participation.
Measure
The CIAS (Wright & Asmundson, 2003) is a 35-item self-report measure, based on the IAS (Kellner, 1987), designed for school-age children. The CIAS is intended to assess fears, beliefs, and attitudes that are associated with health anxiety and abnormal illness behavior. Most items are rated on a 3-point Likert scale (1 = none of the time, 2 = sometimes, 3 = a lot of the time). Items 29 through 31 are rated on a 3-point scale designed to evaluate the frequency of various treatment experiences (1 = zero times, 2 = one or two times, 3 = three or more times). Thirty-three of the 35 items are used in scoring, and therefore total scores can range from 33 to 99. Items 28 and 32 are open-ended and provide supplementary information but are not used in scoring. The CIAS has been found to possess good construct validity and high test-retest (10–14 days) reliability (Wright & Asmundson, 2003).
|
Results |
|---|
|
TopAbstractMethodResultsDiscussionReferences |
|---|
Factor-analytic procedures similar to those of Hadjistavropoulos et al. (1999)
were used. A principal-components analysis (PCA) with oblique (Oblimin) rotation was performed on 33 of the 35 items. Items 28 and 32 are not used in scoring and, therefore, were not used in the factor analysis. As suggested by Tabachnick and Fidell (2001), an exploratory factor analysis was used in order to generate hypotheses about underlying processes measured by the instrument. Two empirical indices regarding the sample’s correlation matrix revealed that it was suitable for factor-analytic procedures: Bartlett’s test of sphericity was significant, 
2(496) = 2218.76, p < .0001, and Kaiser’s measure of sampling adequacy was .84. In order to accurately determine the number of components to retain, parallel analysis was used (Longman, Cota, Holden, & Fekken, 1989). This procedure statistically determines the break in the scree plot and is typically performed once using mean eigenvalues and once using 95th percentile eigenvalues. Based on parallel analysis, a three- or four-factor model was suggested. Following the comparison of the simple structure of the two models, it was determined that the four-factor model had better simple structure (i.e., fewer cross-loadings and fewer items with nonsalient loadings) than the three-factor model. The four-factor solution accounted for 43.9% of the variance in the CIAS item scores. Individually, the amount of variance (after rotation) accounted for by factors 1 through 4 were (eigenvalues in parentheses) 22.9% (7.5), 9.2% (3.1), 6.8% (2.3), and 4.9% (1.6). The pattern matrix of this solution is presented in Table I. Two items (11 and 25) had equivalent, salient (i.e., >.30) cross-loadings on several factors. Item 11 loaded onto factors 1 and 3, and item 25 loaded onto factors 2 and 4. These items loaded within .15 of each other on the two different factors. There was one hyperplane item (item 15). The correlations between the factors were small, ranging from a low of .04 to a high of .34.
|
Factor content was examined for items that had salient loadings (>.30) and did not have equivalent, salient loadings on more than one factor. Based on these criteria, all items but 11, 15, and 25 were interpreted. Factor 1, consisting primarily of items intended to assess fears of illness, death, disease, and pain, was labeled fears. Factor 2 consisted primarily of new items that were intended to assess behavior motivated by concerns regarding illness, such as seeking treatment via parents or guardians and avoidance of foods that are unhealthy, and was labeled help seeking. Items with the highest loadings on factor 3 were from the treatment experience factor identified in Hadjistavropoulos et al. (1999), and thus the same label was applied. Factor 4 comprised items designed to measure the disruptive effects of symptoms on functioning and, consistent with Hadjistavropoulos et al. (1999), was labeled symptom effects.
To determine internal consistency of the empirically derived CIAS subscales, Cronbach’s 
was computed for the total score and subscale scores derived from items with primary salient loadings on each factor (i.e., items 11, 15, and 25 excluded). The fears subscale (13 items) possessed excellent reliability, as indicated by its respective coefficient, .86. The coefficients of the help seeking (9 items) and symptom effects (4 items) subscales, .78 and .76, respectively, suggested that they had good reliability. The treatment experience subscale (4 items) had poor reliability ( = .36) but was substantially improved with deletion of item 8 ( = .70). The coefficient for the CIAS total (30 items) was excellent at .88 and improved only slightly (.89) with the deletion of item 8.
PCA was performed on the four lower-order factors identified in order to assess whether a higher-order structure existed. Bartlett’s test of sphericity, 2(6) = 115.10, p < .0001, and Kaiser’s measure of sampling adequacy (.65) revealed that the sample was suitable for factor-analytic procedures. Obtained eigenvalues with mean eigenvalues and 95th percentile eigenvalues employed in parallel analysis (Longman et al., 1989) provided support for the single higher-order factor labeled health anxiety. This factor accounted for 49.2% of the variance in the lower-order factors. The higher-order structure showed good simple structure, with each factor showing a salient loading (>.30) on the higher-order factor (i.e., loadings of .75, .76, .60, and .68 for the fears, symptom effects, treatment experience, and help seeking lower-order factors, respectively).
|
Discussion |
|---|
|
TopAbstractMethodResultsDiscussionReferences |
|---|
The purpose of the present study was to examine the factor structure of the CIAS. Results suggested that the CIAS can be conceptualized as comprising four lower-order factors (fears, help seeking, treatment experience, and symptom effects) that load hierarchically onto a single higher-order factor of health anxiety. Three of the four lower-order factors (fears, help seeking, and symptom effects) showed good to excellent internal consistency. The internal consistency for the treatment experience factor was initially poor but improved with deletion of item 8. Overall, the present factor-analytic findings are similar to those of Hadjistavropoulos et al. (1999) and Stewart and Watt (2000), both of which identified and reported a hierarchical four-factor solution as the best fitting model for the IAS.
Following Cattell’s (1978) suggestion that each factor represents a distinct set of causal mechanisms, our findings suggest that there are five mechanisms underlying childhood expressions of health anxiety—a general mechanism contributing to most symptoms (corresponding to the higher-order factor) and specific mechanisms contributing to fears, help seeking, symptom effects, and treatment experiences (corresponding to the lower-order factors). The general mechanism may be closely associated with neuroticism or negative emotionality and, together with the specific mechanisms, may determine whether or not a child develops particular manifestations of severe health anxiety (Asmundson, Taylor, Wright, et al., 2001). The consistency of our factor-analytic results with those found in adult samples implies that the CIAS, like the IAS, can be used to effectively assess both the general construct of health anxiety and its lower-order components. In research settings the CIAS may prove a useful addition to efforts designed at disentangling the influence of the general and specific mechanisms on the development, course, prognosis, and treatment outcome issues in severe cases of childhood health anxiety.
In clinical settings the CIAS total and subscale scores can be used in pre- and posttreatment assessment of health-related anxiety. The total score provides a global measure of health anxiety, while the empirically derived subscales yield specific information with regard to its cognitive and behavioral components. Based on the present findings, we recommend that items 8, 11, 15, and 25 be excluded from total and subscale scoring. We also recommend that CIAS assessment data be interpreted with caution until such time as the present findings are replicated in children with confirmed diagnoses of disease phobias or hypochondriasis.
There are other future research directions that also warrant consideration. First, confirmation of the factor structure of the CIAS, using confirmatory factor analysis (CFA) procedures, may be one of the key steps in validating its hierarchical structure and establishing its utility as a sound measure of health anxiety in children. Future CFA investigations are warranted to confirm the applicability of the hierarchical four-factor model within a variety of populations of children (e.g., those with chronic physical handicaps, psychiatric conditions, general medical conditions). Such investigations would address, in part, a limitation of the current study—not evaluating the impact of actual physical disease on responses to the CIAS. We did not assess current physical disease and were therefore unable to partial out its effects on the identified factor structure. Second, another important step in refining the CIAS, based on examination of the simple structure and internal consistency of its factors, will be to determine the appropriateness of inclusion and exclusion of some items. Additional items may need to be added to the treatment-experience subscale to improve its reliability. Finally, the impact of confirmed physical disease and culture on health anxiety in children warrants close empirical inspection. Adult samples of patients with chronic musculoskeletal pain, for example, report more health anxiety and disability than those without (e.g., Hadjistavropoulos, Hadjistavropoulos, & Quine, 2000; Hadjistavropoulos, Owens, Hadjistavropoulos, & Asmundson, 2001). Likewise, adult samples of treatment-seeking and community-dwelling adults have yielded different estimates of the prevalence of severe health anxiety across cultures, some finding elevations (Barsky, Wyshak, Klerman, & Latham, 1990) and others not (Escobar et al., 1998; Gureje, Ustun, & Simon, 1997). These issues remain to be evaluated in children.
The Anxiety Disorders Association of America (ADAA) asserts that "despite their widespread prevalence, childhood anxiety disorders remain vastly underdiagnosed, undertreated, and understudied" (ADAA, 2000). This, in part, motivated the development of the CIAS (Wright & Asmundson, 2003) as a tool for assessing health anxiety in school-age children. Initial results suggest that the CIAS possesses good psychometric properties (Wright & Asmundson, 2003) and, coupled with the results from the present study showing factorial validity and good to excellent internal consistency, indicate that the CIAS is a psychometrically sound instrument for measuring children’s health anxiety. Our specification of the general and specific mechanisms underlying childhood health anxiety provides a sound basis from which to improve our understanding of appropriate assessment and treatment protocols. Based on data derived from future research and clinical applications of the CIAS, we are hopeful that advances will be made in (1) development of diagnostic criteria for hypochondriasis (i.e., severe health anxiety) that are more developmentally appropriate for children than the current DSM-IV-TR (APA, 2000) criteria, and (2) development of diagnostic semistructured interviews for the assessment of presence of severe health anxiety. The former may depend on improved understanding of the dimensions of childhood health anxiety and their impact on course and clinical presentation. Development of diagnostic criteria and interviews will, ultimately, allow researchers and clinicians to aid in early identification of this condition and thereby reduce the number of costly, potentially dangerous, and often unnecessary medical investigations and treatments that children with severe health anxiety can be exposed to.
Received September 15, 2003;
revision received December 30, 2003; revision received April 14, 2004;
|
References |
|---|
|
TopAbstractMethodResultsDiscussionReferences |
|---|
American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Washington, DC: Author.
Anxiety Disorders Association of America. (2000, February 22). Age of innocence: New report confirms childhood anxiety disorders vastly underdiagnosed and undertreated. Retrieved from http://www.adaa.org/dyna/view.cfm?ID=40
Asmundson, G. J. G., Taylor, S., Sevgur, S., & Cox, B. J. (2001). Health anxiety: Classification and clinical features. In G. J. G. Asmundson, S. Taylor, & B. J. Cox (Eds.), Health anxiety: Clinical and research perspectives on hypochondriasis and related disorders (pp. 3–21). London: Wiley.
Asmundson, G. J. G., Taylor, S., Wright, K. D., & Cox. B. J. (2001). Future directions and challenges in assessment, treatment, and investigation of health anxiety. In G. J. G. Asmundson, S. Taylor, & B. J. Cox (Eds.), Health anxiety: Clinical and research perspectives on hypochondriasis and related disorders (pp. 365–381). London: Wiley.
Barsky, A. J., Wyshak, G., Klerman, G. L., & Latham, K. S. (1990). The prevalence of hypochondriasis in medical outpatients. Social Psychiatry and Psychiatric Epidemiology, 25, 89–94.[ISI][Medline]
Campo, J. V., & Reich, M. D. (1999). Somatoform disorders. In S. D. Netherton, D. Holmes, & C. E. Walker (Eds.), Child and adolescent psychological disorders (pp. 321–343). New York: Oxford University Press.
Cattell, R. B. (1978). The scientific use of factor analysis in the behavioural and life sciences. New York: Plenum.
Escobar, J. I., Gara, M., Waitzkin, H., Silver, R. C., Holman, A., & Compton, W. (1998). DSM-IV hypochondriasis in primary care. General Hospital Psychiatry, 20, 155–159.[CrossRef][ISI][Medline]
Fritz, G. K., Fritsch, S., & Hagino, O. (1997). Somatoform disorders in children and adolescents: A review of the past 10 years. Journal of the American Academy of Child and Adolescent Psychiatry, 36, 1329–1338.[CrossRef][ISI][Medline]
Gureje, O., Ustun, T. B., & Simon, G. E. (1997). The syndrome of hypochondriasis: A cross-national study in primary care. Psychological Medicine, 27, 1001–1010.[CrossRef][Medline]
Hadjistavropoulos, H. D., Frombach, I. K., & Asmundson, G. J. G. (1999). Exploratory and confirmatory factor analytic investigations of the Illness Attitudes Scale in a nonclinical sample. Behaviour Research and Therapy, 37, 6781–6784.
Hadjistavropoulos, H. D., Hadjistavropoulos, T., & Quine, A. (2000). Health anxiety moderates the effects of distraction versus attention to pain. Behaviour Research and Therapy, 38, 425–438.[CrossRef][Medline]
Hadjistavropoulos, H. D., Owens, K. M. B., Hadjistavropoulos, T., & Asmundson, G. J. G. (2001). Hypochondriasis and health anxiety among pain patients. In G. J. G. Asmundson, S. Taylor, & B. J. Cox (Eds.), Health anxiety: Clinical and research perspectives on hypochondriasis and related disorders (pp. 298–323). London: Wiley.
Kellner, R. (1987). Abridged manual of the Illness Attitudes Scale. Department of Psychiatry, School of Medicine, University of New Mexico.
Longman, R. S., Cota, A. A., Holden, R. R., & Fekken, G. C. (1989). A regression equation for the parallel analysis criterion in principal component analysis: Mean and 95th percentile eigenvalues. Multivariate Behavioral Research, 24, 59–69.
Stewart, S. H., & Watt, M. C. (2000). Illness Attitudes Scale dimensions and their associations with anxiety-related constructs in a nonclinical sample. Behaviour Research and Therapy, 38, 83–99.[CrossRef][ISI][Medline]
Tabachnick, B. G., & Fidell, L. S. (2001). Using multivariate statistics (4th ed.). New York: Harper & Row.
Wright, K. D., & Asmundson, G. J. G. (2003). Health anxiety in children: Development and psychometric properties of the Childhood Illness Attitude Scales. Cognitive Behaviour Therapy, 34, 194–201.[CrossRef]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||