Anxiety sensitivity in traumatized Cambodian refugees: A discriminant function and factor analytic investigation

https://doi.org/10.1016/j.brat.2005.01.001Get rights and content

Abstract

We examined the psychometric properties and factor structure of a Cambodian translation of the Anxiety Sensitivity Index (ASI) and an Augmented ASI (the ASI supplemented with a 9-item addendum that assesses additional Cambodian concerns about anxiety-related sensations). Both the ASI and the Augmented ASI distinguished among three diagnostic groups: highest score, PTSD with panic disorder (PP group); next, panic disorder without PTSD (P group); and then, other disorders than PTSD or panic disorder (O group). In the discriminant function analysis using the Augmented ASI, the best classificatory predictor (PP vs. P vs. O) was an Addendum item (“It scares me when I stand up and feel dizzy”). The principal component analysis (oblimin rotation) of the ASI yielded a 3-factor solution (I, Weak Heart Concerns; II, Control Concerns; III, Social Concerns) and of the Augmented ASI, a 4-factor solution (I, Weak Heart Concerns; II, Control Concerns; III, Wind Attack Concerns; IV, Social Concerns). The item clustering within the factor solution of both the ASI and Augmented ASI illustrates the role of cultural syndromes in generating fear of mental and bodily events.

Introduction

The Anxiety Sensitivity Index (ASI) assesses the fear of anxiety-related sensations (Reiss, Peterson, Gursky, & McNally, 1986). Elevated scores on the ASI indicate increased risk for spontaneous panic attacks (Schmidt, Lerew, & Jackson, 1997) and anxious response to symptom-provocation procedures such as hyperventilation or carbon dioxide inhalation (McNally & Eke, 1996). The ASI distinguishes among diagnostic groups. For example, patients with panic disorder score higher than those with generalized anxiety disorder even when their scores on measures of trait anxiety are the same (Taylor, Koch, & McNally, 1992).

Most factor analyses of the ASI (the standard 16-item version) in Western populations have revealed a 3-factor solution: Physical Concerns, Mental Incapacitation Concerns, and Social Concerns (Zinbarg, Barlow, & Brown 1997). Taylor and Cox (1998) argue that the ASI contains too few items to assess lower-order factors; they found that a principal component analysis of a 36-item ASI (ASI-R) revealed a 4-factor solution: Fear of Respiratory Symptoms; Fear of Publicly Observable Anxiety Reactions; Fear of Cardiovascular Symptoms; and Fear of Cognitive Dyscontrol.

Some researchers have examined the ASI factor structure in US ethnic groups, demonstrating some differences in the number and item mix of the factors. A factor analysis in an American Indian (Zvolensky, McNeil, Porter, & Stewart, 2001) and a Spanish (Sandin, Chorot, & McNally, 1996) population revealed a 3-factor structure similar to that described in previous investigations of Western populations (e.g., Zinbarg et al., 1997). A study of Native Americans revealed a 1-factor solution to be optimal (Norton, De Coteau, Hope, & Anderson, 2004), and a study of African Americans, a 4-factor solution (Carter, Miller, Sbrocco, Suchday, & Lewis, 1999): Mental Incapacitation Fear; Unsteadiness Fear (e.g., “It scares me when I am faint,” and “It scares me when I am shaky”); Emotional Control Concerns (e.g., “It is important for me to stay in control of my emotions,” and “It is important for me not to appear nervous”); and Cardiovascular Concerns.

A culture may have elaborate ideas about anxiety symptoms that increase catastrophic cognitions, and hence scores on the ASI. Cambodians have a complex ethnophysiology and several culturally specific syndromes that generate great anxiety about arousal-reactive sensations; the panicogenic nature of these culturally specific catastrophic cognitions has been demonstrated in several studies, including an orthostatic challenge (e.g., Hinton et al., 2004).

Cambodians interpret many anxiety symptoms—such as palpitations—as evidence of cardiac weakness, or “weak heart” (khsaoy beh doung), and believe that “heart weakness” most often results from a depletion of bodily energy caused by excessive worry, decreased sleep, or reduced appetite. The self-diagnosis of “heart weakness” leads to many fears (Hinton, Hinton, Um, Chea, & Sak, 2002): (a) the heart will be hyperreactive to any stimulus (e.g., to a noise); (b) the act of standing may overstrain the heart, leading to palpitations, with palpitations then causing a sudden dangerous rise of blood pressure; (b) the heart may suddenly stop beating—especially during palpitations—and cause death; (c) bouts of shortness of breath, even asphyxia, may occur, for breathing itself is driven by a piston-like action of the heart; (d) the body, including the mind, is in a state of energy depletion; and (e) worry—or any state of fear—will further deplete the bodily energy supplies, thereby worsening heart weakness and enervating mental powers.

Cambodians interpret anxiety symptoms in terms of an ethnophysiology of khyâl, a wind-like substance thought to run alongside blood (Hinton, Um, & Ba (2001a), Hinton, Um, & Ba (2001b), Hinton, Um, & Ba (2001c)); Cambodians refer to a bout of anxiety as a “khyâl attack” (kaeut khyâl). The “wind attack” (“khyâl attack”) interpretation of somatic anxiety symptoms generates multiple catastrophic cognitions:

  • 1.

    Sensations in the limbs—as in coldness, fatigue, soreness, or numbness—result from blockage of the flow of both “wind” (khyâl) and blood; the blockage may then cause (a) “death of the limb” (ngouep day ceung) owing to decreased flow of “wind” and blood and (b) an upward ascent of “wind” and blood from the limbs toward the trunk, neck, and head.

  • 2.

    Abdominal sensations indicate blockage of the normal downward flow of “wind,” which may ascend upwards in the body toward the chest, neck, and head.

  • 3.

    Neck and head symptoms result from “wind” and blood rising upward in the body, with the “wind” and blood possibly (a) distending and rupturing neck vessels; (b) rushing out the ears, causing tinnitus (referred to as “wind exiting the ears,” or khyâl ceuny pii treujieu, with patients fearing deafness); and (c) swirling in the head, causing dizziness and syncope.

  • 4.

    Dizziness upon standing usually indicates the onset of a “wind attack,” which may cause syncope and death (a “wind attack” upon standing is given a special name: “wind overload,” or khyâl kô); so patients, especially when they have recently experienced “wind attacks” (i.e., bouts of anxiety), greatly fear standing.

If members of a culture greatly fear certain anxiety symptoms not evaluated by the standard 16-item ASI, additional items (i.e., an expanded ASI) would be needed to profile the predisposition to panic—and to attain adequate factor structure. Because of culturally specific ethnophysiological understanding and syndromes, Cambodians catastrophically cognize about certain anxiety-related sensations—cold extremities, neck tension, tinnitus, and orthostatic dizziness (Hinton, Ba, Peou, & Um, 2000; Hinton, Um, & Ba (2001a), Hinton, Um, & Ba (2001b), Hinton, Um, & Ba (2001c), Hinton et al. (2004))—not assessed in the standard ASI. At our psychiatric clinic, we developed an ASI addendum to measure these concerns. We found the ASI addendum to be more sensitive to change across CBT treatment than the standard ASI (Otto et al., 2003).

If a person has experienced massive trauma leading to the linking of arousal-reactive sensations (e.g., dizziness, palpitations, cold extremities) to trauma networks—such is the case with the Cambodian refugees (see Hinton et al., 2004)—then one would expect that the person would fear those sensations. If a patient experiences certain sensations during trauma recall, then those sensations will be feared as harbingers of trauma recall. Also, for Cambodians, PTSD-related symptoms worsen “weakness fears”; Cambodians consider PTSD-related symptoms (e.g., startle-induced palpitations) to indicate “heart weakness.” Thus, as a result of trauma associations to arousal-reactive symptoms—and fears of “weakness”—Cambodian patients with PTSD and comorbid panic disorder would be expected to have higher ASI scores than patients with just panic disorder.

To better understand the nature of anxiety sensitivity in the Cambodian refugee population, we examined the psychometric properties and factor structure of the ASI and the Augmented ASI. We had four major hypotheses: (a) the Cambodian-translated ASI and the Augmented ASI (i.e., the standard 16-item ASI supplemented with a 9-item addendum) would differentiate among three diagnostic groups (PTSD with comorbid panic disorder, panic disorder without comorbid PTSD, and other disorders than PTSD or panic disorder); (b) the Augmented ASI would better differentiate among the three diagnostic groups; (c) a discriminant function analysis with the Augmented ASI would reveal some of the addendum items to be especially able to differentiate among the three groups; and (d) the factor analysis of the ASI, and especially the Augmented ASI, would reveal culturally influenced clustering of fears.

Section snippets

Participants

Our psychiatric clinic (in Lowell, MA) serves the second largest Cambodian population in the US (35,000). Most Cambodian patients initially present to the clinic with PTSD and panic disorder. The participants for the study were 157 patients (88 women) whose mean age was 47.5 years (SD=8.3). Sixty-two participants had PTSD and panic disorder (PP group); 51, panic disorder but not PTSD (P group); and 44, a disorder other than PTSD or panic disorder—for example, dysthymia, major depression

ASI scores

Mean ASI scores did not differ significantly as a function of age or gender. For ASI scores, see Table 1. For scores on individual ASI items, see Table 2.

Factor structure of the ASI

As is usually done in the factor analysis of the ASI (e.g., Norton et al., 2004), we performed a principal component analysis with oblimin rotation. We used the Kaiser (1961) rule (i.e., an eigenvalue>1) to determine the number of factors to retain, resulting in a 3-factor solution that accounted for 55.8% of the variance (see Table 3 for a

ASI and Augmented ASI scale scores: PTSD effects

ASI scores were most elevated in those patients who had PTSD with comorbid panic disorder. The ASI assesses fear of somatic (and mental) components of anxiety. Fear of anxiety-related symptoms seemingly derives from three sources (Hinton et al., 2004): (a) catastrophic thoughts about the symptom; (b) trauma associations to the symptom; and (c) interoceptive conditioning of the symptom directly to somatic and psychological fear. As compared to Cambodians with panic disorder but without PTSD,

References (23)

  • First, M. B., Spitzer, R. L., & Gibbon, M. (1995). Structured clinical interview for DSM-IV Axis I disorders. Patient...
  • Cited by (0)

    View full text