Anxiety sensitivity and panic attack symptomatology among low-income African–American adolescents
Introduction
Between 16 and 63% of community samples of adolescents report having suffered at least one panic attack during their lifetime and up to 6% meet diagnostic criteria for panic disorder (e.g., King et al., 1993, King et al., 1996, Lau et al., 1996, Macaulay and Kleinknecht, 1989, Warren and Zgourides, 1988; also see Ollendick, 1998 for review). Studies examining the phenomenology of panic attacks in adolescents indicate that the most common symptoms are somatic in nature and include accelerated heart rate, trembling, dizziness/feeling faint, sweating, dyspnea, and hot and cold flashes (King et al., 1996; Macaulay and Kleinknecht, 1989, Warren and Zgourides, 1988). Persistent symptoms of panic in youth are associated with a significant degree of impairment in functioning and are often co-morbid with symptoms of anxiety, depression, school refusal, and familial stress (Bradley and Hood, 1993, Hayward et al., 1995, Hayward et al., 1997, King et al., 1996, Macaulay and Kleinknecht, 1989). In light of the high prevalence rates and associated impairment of panic symptoms among adolescents, as well as retrospective reports from adults with panic disorder indicating that symptom onset often began during childhood (Thyer et al., 1985, Von Korff et al., 1985), research efforts have turned toward understanding the etiology of panic and identifying potential risk factors.
Theoretical models, as well as empirical evidence from adult studies, indicate that anxiety sensitivity (AS) may be an important construct for understanding the development and/or maintenance of panic attacks and symptoms (see Taylor & Fedoroff, 1999, for review). AS, conceptualized as a stable trait, is the tendency to believe that anxiety and fear-related symptoms lead to negative or harmful somatic, social, or psychological consequences. Reiss and coworkers (Reiss, 1991; Reiss & McNally, 1985), as part of their expectancy theory of anxiety, hypothesized that when a person with high AS experiences somatic symptoms of anxiety (e.g., feeling dizzy or light headed), she or he is likely to fear that the symptoms will be devastating (e.g., she/he will go crazy/loose control). This expectation increases anxiety levels, exacerbates the somatic symptoms, and leads to the vicious cycle that results in a panic attack.
Numerous empirical studies, using adult samples and cross-sectional methodologies, support the proposed relation between AS and panic symptoms in clinical (e.g., Bruce et al., 1995, Cox et al., 1999, Taylor et al., 1992) and non-clinical samples (e.g., Asmundson and Norton, 1993, Cox et al., 1991, Donnell and McNally, 1990). A limited number of prospective studies have examined the directional relationship between AS and panic in adults—that is, whether AS is a precursor or consequence of panic symptomatology (Maller & Reiss, 1992; Ehlers, 1995; Harrington, Schmidt, & Telch, 1996; Schmidt, Lerew, & Jackson, 1997). Findings from the majority of these prospective studies support Reiss and colleagues’ expectancy theory and suggest that AS is a risk factor for the development of panic attacks in adults (see Schmidt, 1999 for review).
In contrast to the adult literature, research on AS in youth is sparse. Investigations of AS in youth have been facilitated by the recent development of the Childhood Anxiety Sensitivity Index (CASI; Silverman, Flesig, Rabian, & Peterson, 1991) and the Anxiety Sensitivity Inventory for Children (ASIC; Laurent, Schmidt, Catanzaro, Joiner, & Kelley, 1998), both downward extensions of the Anxiety Sensitivity Index (ASI; Peterson & Reiss, 1987). Initial studies with the CASI, the most widely used measure of AS in youth, indicate that the measure has good psychometric properties for primarily Caucasian samples of both clinical and non-clinical youth (Silverman et al., 1991). Recent research with the CASI also indicates that this measure holds promise in aiding our understanding of the etiology of anxiety disorders in general and panic in particular (Chorpita et al., 1996, Rabian et al., 1999; Weems, Hammond-Laurence, Silverman, & Ginsburg, 1998).
Direct evidence supporting the relation between AS and panic in youth is limited to two published studies, both using cross-sectional designs. As part of a larger study investigating the phenomenology of panic in an outpatient clinic sample, Kearney, Albano, Eisen, Allan, and Barlow (1997) compared Caucasian youth (8–17 years old) with panic disorder (n=20) and non-panic anxiety disorders (n=20) on several self-report measures, including the CASI. Relevant findings revealed that while the two groups did not differ on measures of general anxiety symptomatology, depression, or fearfulness, youth with panic disorder scored significantly higher on the CASI compared to their non-panic counterparts.The second study to examine the relation between AS and panic symptomatology in youth was conducted by Lau et al. (1996). In this study, a predominantly Caucasian sample of high school students (n=77) ages 14–18 years (mean age=16.7 years) completed the Panic Attack Questionnaire (PAQ; Norton, Dorward, & Cox, 1986) and the CASI. Based on the PAQ, adolescents were classified as non-panickers (n=47) or panickers (n=30). Similar to Kearney et al. (1997), adolescent panickers scored significantly higher on the CASI than non-panickers.
In sum, research examining the relation between AS and panic symptomatology in youth is limited to two studies. No published study has examined the relation between AS and panic in youth over time. Moreover, the majority of this research (both child and adult) is based on predominantly Caucasian samples, raising questions about whether the construct of AS, and its relation to panic symptoms, is applicable to individuals of diverse racial/ethnic backgrounds. In light of these gaps in the literature, the primary aim of the present study was to examine the concurrent and prospective relation between AS and panic symptoms in a community sample of African–American adolescents.
Investigating the long-term relation between AS and panic symptoms is important for several reasons. First, prospective studies allow investigators to clarify whether AS is an antecedent or consequence of panic attacks. Second, long-term studies are capable of providing evidence to determine if AS is a stable trait in youth. To the extent that AS is stable over time, and a risk factor for the development of panic attacks, it will be possible to identify individuals at risk and initiate appropriate prevention and intervention strategies.
Examining the relation between AS and panic symptoms in African–Americans is also of critical importance. Peterson and Plehn (1999), in a recent review of the literature on measuring AS, devoted only two sentences to the empirical findings related to racial patterns of AS, concluding that there is “insufficient data … available” at this time (p. 66). Yet, recent evidence indicates that AS may be associated with race. Rabian et al. (1999) found that African–American children, relative to their Caucasian peers, were more likely to report higher levels of AS, as measured by the CASI (Rabian et al., 1999). Reviews of the limited research literature on anxiety and its disorders in African–Americans (both adults and children) also suggest that specific anxiety-related phenomena may be manifested differently in this population when compared to Caucasians (Horwath et al., 1994, Last and Perrin, 1993, Neal and Ward Brown, 1994, Neal and Turner, 1991). Finally, research on non-Caucasians is important because it has the potential to address whether AS is a common risk factor for adolescents of diverse racial/ethnic backgrounds and thereby extend current etiological theories of AS and panic.
In order to address the primary aims of this study, adolescents completed self-report measures of AS and panic on two occasions, 6 months apart. Consistent with previous research and theory, it was predicted that AS would be positively correlated with symptoms of panic and that AS, measured at Time 1, would positively predict panic symptoms measured at Time 2. Moreover, adolescent panickers were expected to report higher levels of AS compared to non-panickers. Though not a primary focus of this study, preliminary psychometric properties of the CASI (i.e., test–retest and internal consistencies) were also examined in order to contribute to the literature on AS among African–Americans.
Section snippets
Participants
Participants at Time 1 were 107 African–Americans ranging in age from 14 to 17 years (mean age 15.6 years) who were recruited from an urban high school. According to school documents, approximately 35% of the students attending this school live below the level of poverty and the majority (75%) reside in single-parent households. In the Time 1 sample, 53 students were male (mean age 15.6 years) and 54 were female (mean age 15.6 years). At Time 2, 6 months later, 66 students (mean age 15.4 years)
Psychometric properties of the CASI
Mean total scores on the CASI for girls, boys, and the total sample are presented in Table 1. Results of t-test analyses revealed that girls scored significantly higher than boys on the CASI at Time 1, t(105)=−2.16, p<.05, and Time 2, t(64)=−2.57, p<.05.
Internal consistencies, calculated using Cronbach’s coefficient alpha for the total score of the CASI, were .83 at Time 1 and .76 at Time 2 for the entire sample. Test–retest reliability for the CASI over the 6-month period was .48. Age was not
Discussion
The present study adds to a growing body of research on anxiety sensitivity in youth by examining the concurrent and prospective relation between AS and panic symptomatology among African–Americans from predominantly low-income, single-parent households. Overall, findings indicated high levels of AS were generally associated with panic attack symptoms. These findings are consistent with research using adult and Caucasian samples, suggesting that excessive levels of AS may be a risk factor for
References (47)
- et al.
Anxiety sensitivity and its relationship to spontaneous and cued panic attacks in college students
Behaviour Research & Therapy
(1993) - et al.
The Screen for Child Anxiety-Related Emotional Disorders (SCARED): scale construction and psychometric characteristics
Journal of the American Academy of Child and Adolescent Psychiatry
(1997) - et al.
Psychiatrically referred adolescents with panic attacks: presenting symptoms, stressors, and comorbidity
Journal of the American Academy of Child and Adolescent Psychiatry
(1993) A cognitive approach to panic
Behaviour Research & Therapy
(1986)- et al.
Anxiety sensitivity and nonclinical panic attacks
Behaviour Research & Therapy
(1991) - et al.
Anxiety sensitivity and panic attacks in a nonclinical population
Behaviour Research & Therapy
(1990) - et al.
Gender role orientation and fearfulness in children with anxiety disorders
Journal of Anxiety Disorders
(2000) - et al.
A reanalysis of agoraphobia
Behavior Therapy
(1978) - et al.
Assessment and phenomenology of nonclincal panic attacks in adolescent girls
Journal of Anxiety Disorders
(1997) - et al.
School refusal in young adolescent girls with nonclinical panic attacks
Journal of Anxiety Disorders
(1995)