حساسیت اضطراب به عنوان پیش بینی کننده توسعه علائم هراس، حملات هراس و اختلال پانیک: یک مطالعه آینده نگر

The purpose of the study was to examine how anxiety sensitivity (AS) acts as a dispositional factor in the development of panic symptoms, panic attacks, and panic disorder. Between 1986 and 1988, data were collected from 505 undergraduates at an urban university. At Time 1, measures used were the ASI to assess AS, the trait scale of the State-Trait Anxiety Inventory (STAI-T) to measure trait anxiety, and self-report questionnaires to measure personal and family history of panic and anxiety symptoms. During the Spring of 1999, 178 of these subjects were re-contacted, and information was gathered on subjects’ subsequent development of panic symptoms, panic attacks, panic disorder, and trait anxiety (STAI-T). The ASI was the strongest predictor of the development of panic symptoms and panic attacks. After controlling for trait anxiety, the ASI was not predictive of the development of panic disorder.

Panic disorder, with and without agoraphobia, affects a substantial proportion of the population. The Diagnostic and Statistical Manual of Mental Disorders (rev. 4th ed.; DSM-IV) of the American Psychiatric Association (1994) estimates that each year 1.5–3.5% of the world’s population will experience a spectrum of symptoms in a discrete period of time constituting a panic attack. Panic is not limited to clinical populations. Surveys of non-clinical populations reveal that a significant proportion of the population experience such attacks (Norton et al., 1986 and Norton et al., 1985). Findings from Telch, Lucas, and Nelson’s (1989) study of approximately 2000 college students suggested that lifetime prevalence rates are as high as 12%. A panic attack is frequently followed by a pattern of behavior change designed to minimize the possibility of similar future attacks. This behavior change may take the form of avoidance or special precautions. Moreover, behavior changes may become progressively more elaborate as well as emotionally and financially disruptive. The chronic nature of the disorder means that these negative changes are long-lasting and exponential in their impact on an individual’s quality of life. Currently, clear predictors of the development of anxiety-spectrum difficulties are unavailable. Given the extent of suffering resulting from panic disorder, improvements in identification of markers and prevention efforts are greatly needed. One variable that has shown potential as a predictor of future panic is anxiety sensitivity (AS). AS theory is a cognitive theory examining “the individual differences in what people think will happen to them when they experience anxiety” (Peterson & Reiss, 1992). People with high AS, as measured by the Anxiety Sensitivity Index (ASI; Reiss, Peterson, Gursky, & McNally, 1986), have a greater belief that the symptoms of anxiety will be harmful or have harmful consequences beyond initial unpleasantness (Reiss, 1991; Reiss & McNally, 1985). In contrast, people with low AS endure symptoms as unpleasant but not incapacitating. Research supports the Reiss and McNally (1985) view that high AS may precede experience of a panic attack. Peterson and Sacks (1987) found that the ASI distinguished students who had experienced recent panic attacks (ASI=21.1, S.D.=8.9, N=105) from those who had not (ASI=16.3, S.D.=8.1, N=155). In a college sample of 425 students, Donnell and McNally (1990) found that not only was the proportion of subjects reporting panic in the high AS category significantly higher (32.4%) than the low and medium AS categories, but 46 of the high AS subjects (68%) had no history of panic symptoms. Therefore, the presence of high AS is not determined by previous experiences with unpredictable panic symptoms. A strong argument for the utility of AS relative to trait anxiety is provided by research demonstrating that AS accounts for a greater proportion of the variance than trait anxiety in the prediction of panic. Donnell and McNally (1990) examined the relationship of AS to results of a hyperventilation through CO2 inhalation challenge test. The ASI was more predictive than history of panic symptoms. In fact, history of panic was only contributory in individuals with high ASI scores. These findings have been supported by other hyperventilation and CO2 challenge studies. Telch, Silverman, and Schmidt (1996) established that the ASI is able to predict development of panic symptoms in CO2 challenge situations in subjects with no history of panic symptoms or psychiatric diagnoses. Similarly, Rapee and Medoro (1994) demonstrated the ability of the ASI to predict panic independent of history of panic symptoms. Furthermore, challenges with caffeine (Harrington, Schmidt, & Telch, 1996) have likewise demonstrated the ASI’s ability to predict panic above and beyond trait anxiety measures. These laboratory studies provide valuable information regarding the correlation between AS and panic symptoms created by physiological challenges; however, their predictive ability is limited. Thus far, two prospective studies have examined the stability of AS over time and as a predictor of panic. Maller and Reiss (1992) examined the ability of the ASI to predict anxiety and panic in a sample of 151 college students. The ASI was found to be predictive of frequency and intensity of panic attacks 3 years subsequent. Furthermore, individuals with high AS scores were five times as likely to have developed panic disorder or an anxiety disorder compared to individuals with low AS scores. Unfortunately, the Maller and Reiss (1992) study did not include comparison trait measures, and did not measure the presence of panic attack and panic disorder at Time 1. Thus, prediction of later panic attacks and panic disorder may have been explained by high pre-existing levels of trait anxiety and panic attacks or the presence of panic disorder. In a second prospective study Schmidt, Lerew, and Jackson (1997) assessed 1172 United States Air Force Academy (USAFA) cadets before and after completing a rigorous and stressful basic cadet training. Although the period of time was shorter than in Maller and Reiss’ study, the stressful nature of the training exercise (including ongoing scrutiny from supervisors, increased exercise, and decreased sleep as well as general stressors associated with beginning of college such as leaving home) was felt to be optimal for the development of anxiety. After controlling for trait anxiety and history of panic, Schmidt et al. (1997) found that AS predicted development of panic symptoms. Moreover, it affirmed that the ASI can uniquely predict the development of panic symptoms over trait anxiety. The ASI was significantly associated with development of spontaneous panic, whereas trait anxiety as measured by the STAI-T (Knight, Waal-Manning, & Spears, 1983) did not predict panic occurrence. Although highly contributory to the body of knowledge regarding the predictability of the ASI in the development of panic symptoms, the study of Schmidt et al. (1997) has limitations. One concern is the fact that the participants were all cadets. The lifestyle of a cadet in training has limited usefulness in its generalizability to the average subject. Cadets are subjected to acutely intense physical as well as emotional stress, which is less often the case in the general population. Surprisingly, the mean ASI score for the group (M=19.8, S.D.=8.0; Schmidt, Lerew, & Jackson, 1998) closely reflected non-clinical college sample mean levels of AS ( Peterson & Reiss, 1992). Although the sample is representative of college students in terms of ASI scores, several selection characteristics distinguish the sample. Issues of concern are the facts that: (1) the sample was self-selected, (2) the sample was predominantly male, and (3) cadets enrolling in the military are predominantly exceptionally physically fit and free of health difficulties due to entrance requirements and the rigors of the military. Thus, results of the Schmidt et al. (1997) study may not generalize to the common population. The prospective work of Maller and Reiss (1992) and Schmidt et al. (1997) has established that AS is instrumental in development of panic symptoms, and the Maller and Reiss (1992) study suggests a relationship with the future occurrence of panic disorder. Clearly, there is a need for a long-term follow up study with a normative sample that can effectively discriminate the predictive abilities of AS above and beyond trait anxiety and history of panic. For the current study, data were collected from approximately 505 undergraduates at an urban university between 1986 and 1988. Measures used were the ASI (Reiss et al., 1986) to assess AS, the STAI (Spielberger, Gorsuch, Lushene, Vagg, & Jacobs, 1983) to measure trait anxiety, and self-report questionnaires to measure personal and family history of panic and anxiety symptoms (Norton et al., 1986). Therefore, as with the study of Schmidt et al. (1997), the relationship of AS in the development of panic symptoms and/or panic disorder can be compared to history of panic as well as trait anxiety. By using a sample originally collected in college that is currently in their early thirties, the present study increases the ability to generalize findings to the general population. Three hypotheses were developed and tested in the present study. (1) Insofar as AS is instrumental in development of anxiety disorders and related pathology, the level of AS reported during a person’s college years will predict the development of (a) panic attacks, (b) panic disorder, and (c) panic symptoms during the next decade of life. (2) Additionally, AS will be predictive in the development of panic attacks, panic disorder, and panic symptoms above and beyond level of trait anxiety and history of panic measured at college age. (3) Lastly, AS will be predictive of later trait anxiety above and beyond level of trait anxiety measured at college age.