- Email: [email protected]
Anxiety sensitivity and worry
Personality and Individual Differences 38 (2005) 1223–1229 www.elsevier.com/locate/paid
Anxiety sensitivity and worry Mark Floyd *, Amber Garfield, Marcus T. LaSota Psychology Department, University of Nevada, Box 455030, Las Vegas, NV 89154-5030, United States Received 10 November 2003; received in revised form 17 May 2004; accepted 4 August 2004 Available online 5 November 2004
Abstract Anxiety sensitivity (AS), the fear of oneÕs response to anxiety provoking stimuli, has been correlated with anxiety disorders and has been theorized to be a risk factor in the development of anxiety disorders. Consistent with prior research (Dugas, Gosselin, & Ladouceur, 2001) it was hypothesized that AS and varying levels of worry share an underlying feature of anxiety specifically related to a perceived lack of control over future events, with a resultant emphasis on negative potential outcomes. In this study, the association between AS, overall distress, pathological worry, and non-pathological worry was investigated in a sample of 342 undergraduate volunteers. Results indicated that AS was significantly correlated with the Penn State Worry Questionnaire, the Worry Domain Questionnaire, and the Symptom Checklist 90-Revised. AS was a significant predictor of worry, even after factoring out overall distress, although the incremental amount of explained variance was small. Thus, AS and worry have more in common than distress. We speculate that the common element is a tendency to focus on the uncertainty of the future. Ó 2004 Elsevier Ltd. All rights reserved. Keywords: Anxiety sensitivity; Worry; Anxiety; Fear; Distress
1. Introduction Over the past two decades, the concept of a sensitivity to become fearful of the sensations associated with anxiety (i.e., anxiety sensitivity; Reiss & McNally, 1985) has garnered a great *
Corresponding author. Tel.: +1 702 895 0109. E-mail address: mfl[email protected] (M. Floyd).
0191-8869/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.paid.2004.08.005
1224
M. Floyd et al. / Personality and Individual Differences 38 (2005) 1223–1229
deal of research support as a common factor among persons experiencing anxiety disorders and related conditions. Based on the expectancy model of fear (Reiss, 1991; Reiss & McNally, 1985), anxiety sensitivity (AS) is regarded as a unique characteristic of the individual that serves as a ‘‘predisposing factor in the development of anxiety-mediated disorders’’ (Reiss & McNally, 1985, p. 119). Reiss & McNallyÕs expectancy model considering ‘‘fear behavior’’ is comprised of two components: an expectancy that danger will occur (danger expectancy), which leads to avoidance behaviors, and a preemptive ‘‘fear of fear’’ that is comprised of both the expectation that anxiety will occur following exposure to an anxiety-provoking stimulus (anxiety expectancy) and the belief that the experience of anxiety will bring about feelings of illness, embarrassment, or possibly further anxiety symptoms (anxiety sensitivity). Thus, AS relates strictly towards the predisposed tendency to fear oneÕs reactions to an anxiety invoking stimulus and not the stimulus per se. Reiss and Havercamp (1996) posed an argument that, likely due to genetic and predispositional factors, some individuals are especially susceptible (i.e., those with high degrees of AS) to experience anxious symptoms as particularly unpleasant. Persons with high levels of AS will tend to worry that if forced to encounter an anxiety-inducing stressor, something unpredictable and harmful will occur (Reiss & McNally, 1985; Taylor, Koch, McNally, & Crockett, 1992b). To this extent, persons with high AS levels are likely to be highly alert to signals portending anxiety and to have exaggerated anxiety reactions (Reiss & McNally, 1985) when faced with such stimuli. In an effort to empirically assess AS, the Reiss–Epstein–Gursky Anxiety Sensitivity Scale, now better known as the Anxiety Sensitivity Index (ASI; Reiss, Peterson, Gursky, & McNally, 1986), was created. The ASI is a modified version of a scale created by Epstein (1982, as cited in Reiss et al., 1986), who demonstrated that AS is a qualitatively distinct construct from anxiety. ASI scores have been found to reliably differentiate anxious from non-anxious individuals (Reiss et al., 1986) and panic disorder from other anxiety disorders (Taylor, Koch, & Crockett, 1991). Consistent with AS and expectancy theories (Reiss, 1991; Reiss & McNally, 1985), research on the ASI has demonstrated that panic disorder tends to be most strongly associated with the endorsement of high ASI scores, and that persons diagnosed with post-traumatic stress disorder (McNally, 1990), generalized anxiety disorder (GAD), obsessive-compulsive disorder, and social phobia all had ASI scores higher than normal controls (Taylor et al., 1991; Taylor, Koch, & McNally, 1992a). Other researchers have suggested that the ASI may have applications beyond anxiety disorders. Depression has been speculated to hold a positive correlation to high AS, as two groups of researchers (Otto, Pollack, Fava, Uccello, & Rosenbaum, 1995; Taylor, Koch, Woody, & McLean, 1996) found persons with major depression to evidence ASI scores that, with the exception of panic disorder, were comparable to anxiety disorders. It is particularly curious as to why depression would relate to the construct of AS. In the current study, the concept of AS was extended to investigate the relationship between AS and worry. Although the primary feature of GAD when in excess, worry is believed to be a commonplace occurrence in the day-to-day lives of well-adjusted people (Dupuy, Beaudoin, Rhe´aume, Ladouceur, & Dugas, 2001; Joormann & Sto¨ber, 1997). The specific content of worry has not been shown to differ much between clinical worriers and non-clinical worriers (Hoyer, Becker, & Roth, 2001). However, the quality and degree of worry may be distinct, as those with GAD
M. Floyd et al. / Personality and Individual Differences 38 (2005) 1223–1229
1225
tend to worry more (Craske, Rapee, Jackel, & Barlow, 1989; Dupuy et al., 2001). Furthermore, worry, when in lesser degrees, has been associated with positive attributes such as preparation for actions, decision-making, and motivation towards accomplishing tasks (Borkovec & Roemer, 1995; Davey, Hampton, Farrel, & Davidson, 1992). While research exists that connects high ASI levels with the severity of GAD worry symptoms (Gross & Eifert, 1990; Taylor et al., 1992a), no known research studies have examined the connection between worry in general, overall distress, and AS. Conceptually, the constructs of worry and anxiety sensitivity appear to have similar elements. Worry is conceptualized as a fear or preoccupation with future outcomes and anxiety sensitivity is conceptualized as a fear of oneÕs response to anxiety symptoms, presumably upon exposure to future stimuli. The connection between AS and worry may be prospectively linked to a common experience of distress over uncertain future outcomes and a related intolerance for uncertainty (Dugas et al., 2001). Therefore, we hypothesized that AS would be significantly correlated with overall distress, pathological worry, and non-pathological worry. Additionally, we hypothesized that the connection between AS and worry was more than just a common experience of distress and that AS would be a significant predictor of both pathological and non-pathological worry, even after controlling for overall distress.
2. Methods 2.1. Participants Participants were 342 undergraduate student volunteers who received a Psychology 101 research credit for their participation. The average age of the sample was 19.7 years (SD = 3.3) and there were more women (59%) than men (41%). The self-reported ethnic composition of the sample was 60% Caucasian, 19% Asian-American, 10% African-American, 9% HispanicAmerican, 1% Native American, and 1% of mixed ethnic backgrounds. 2.2. Instruments The Anxiety Sensitivity Index (ASI; Reiss et al., 1986) was used as a measure of anxiety sensitivity. The ASI is a 16-item self-report questionnaire in which each item is rated on a five-point Likert scale from 0 (very little) to 4 (very much) endorsement. Each item reflects the belief that anxious sensations are unpleasantly experienced and potentially lead to harmful consequences. The degree to which one fears the experience of anxiety symptoms is reflected in higher scores. The ASI yields a total score (0–64 range) by adding the calculated responses for all 16 items. Using the total score for the instrument, the ASI has demonstrated sound psychometric properties, which include adequate test–retest reliability (Reiss et al., 1986), high internal consistency and good validity (see review by Peterson & Plehn, 1999). The Penn State Worry Questionnaire (PSWQ; Meyer, Miller, Metzger, & Borkovec, 1990) was used as a measure of pathological worry. Items on this measure are believed to reflect the uncontrollability and excessiveness of general worries. Meyer et al. (1990) reported that this measure has good psychometric properties, which include high internal consistency and good test–retest
1226
M. Floyd et al. / Personality and Individual Differences 38 (2005) 1223–1229
reliability. This instrument has also been found to reliably distinguish persons diagnosed with GAD from other anxiety disorders (Brown, Antony, & Barlow, 1992), and has been applied in both clinical and non-clinical populations. The Worry Domain Questionnaire (WDQ; Tallis, Eysenck, & Mathews, 1992) was used as a measure of non-pathological worry. This 30-item questionnaire is comprised of six subscales of five items each: (1) relationships, (2) lack of confidence, (3) aimless future, (4) work incompetence, (5) financial, and (6) socio-political. The total score is calculated by summing only the first five subscale domains (25 items), as the authors of the instrument have found that the sociopolitical domain may reflect social desirability. Although intended for use as a measure of non-clinical levels of worry, the WDQÕs authors suggest that the five ‘‘core’’ worry domains may apply generally to assess worry in clinical settings. The WDQ has been found to have adequate test–retest reliability, good internal consistency, and adequate validity properties (see review by Kelly, 2002). The Symptom Checklist 90, Revised (SCL-90R; Derogatis, 1983) was used as a measure of overall distress and psychopathology. This 90-item self-report instrument measures distress across nine symptom subscales, which include Somatization, Obsessive–Compulsive, Interpersonal Sensitivity, Depression, Anxiety, Hostility, Phobic Anxiety, Paranoid Ideation, and Psychoticism. There are three global indices: Global Severity Index (GSI), Positive Symptom Distress Index (PSDI), and Positive Symptom Total (PST). This scale has been found to have good psychometric properties, specifically as a measure of general distress (Cyr, McKenna-Foley, & Peacock, 1985). In this study, the GSI was used as the measure of general distress. 2.3. Procedure The participants entered a large classroom. They were given a packet of questionnaires, and then the experimenter read aloud the informed consent and instructions for the instruments. Participants were instructed to complete the instruments and turn them in to the experimenter, who then gave them a ‘‘receipt’’ indicating they participated in the study.
3. Results The averages and standard deviations for the instruments were: SCL-90R GSI (M = 0.75, SD = 0.58), ASI (M = 24.59, SD = 5.43), PSWQ (M = 43.52, SD = 10.17), and WDQ (M = 32.47, SD = 21.55). As predicted, the ASI was significantly correlated with the GSI, r = .57, p < .001; the WDQ, r = .55, p < .001; and the PSWQ, r = .41, p < .001. The GSI was also significantly correlated with the WDQ, r = .69, p < .001; and the PSWQ, r = .55, p < .001. The WDQ and PSWQ were also significantly correlated, r = .507, p < .001. Regression analyses were used to further test the strength of association of the ASI with worry. In separate regression equations, total scores on the PSWQ and WDQ were used as target variables, with the SCL-90R GSI scores and ASI scores used as predictors. As shown in Tables 1 and 2, the ASI was a significant predictor for both the WDQ and PSWQ, respectively even after controlling for overall distress (as measured by the GSI). The inclusion of the ASI in the model explained an additional 3% of the variance for the WDQ and an additional 2% of the variance for the PSWQ.
M. Floyd et al. / Personality and Individual Differences 38 (2005) 1223–1229
1227
Table 1 Summary of Regression Analyses for Anxiety Sensitivity and Non-pathological Worry Variable
B
SE B
b
p
GSI ASI
20.664 .850
1.757 .186
.564 .220
.000 .000
Note. R2 = .509. The dependent variable is the Worry Domain Questionnaire. GSI is the Global Severity Index of the Symptom Checklist 90-Revised. ASI is the total score of the Anxiety Sensitivity Index. Table 2 Summary of Regression Analyses for Anxiety Sensitivity and Pathological Worry Variable
B
SE B
b
p
GSI ASI
8.123 .294
1.005 .107
.460 .157
.000 .006
Note. R2 = .318. The dependent variable is the Penn State Worry Questionnaire. GSI is the Global Severity Index of the Symptom Checklist 90-Revised. ASI is the total score of the Anxiety Sensitivity Index.
4. Discussion As expected, the ASI was significantly correlated with both pathological (PSWQ) and nonpathological worry (WDQ). Furthermore, the ASI was significantly correlated with overall distress (GSI). The effect sizes of the associations, using typical values for effect size magnitude (Cohen & Cohen, 1983), were large for non-pathological worry and overall distress, and medium to large for pathological worry. In fact, the correlation between the ASI and the GSI was actually higher than the correlations between the ASI and either of the worry measures. Since the GSI measures overall distress, perhaps the ASI is tapping an element of general distress. Alternatively, it is possible the GSI is such a broad instrument that it includes items measuring AS and worry and the strong correlations are due in part to shared items. Another explanation is, given the inherent distress element common in depression and anxiety (Katon & Roy-Byrne, 1991), that depression and AS share a common feature of stress related to the unsettling perception of unpredictable, unknown, and possibly frightening future events (Dugas et al., 2001). The correlation between the ASI and the WDQ was considerably higher than the correlation between the ASI and the PSWQ. This could be due to using a sample of college students (instead of a clinical population) and their perception of the distress associated with worry. In comparing the items on the WDQ and PSWQ, it appears that the PSWQ focuses more on the out-of-control aspect of (pathological) worry whereas the WDQ focuses on type of worry and frequency of worry. Since the GSI and WDQ had the highest correlation of all the variables, it indicates that most participants in this study perceived worry to be somewhat distressing, even if their worry was not excessive, out of control, or functional. It also suggests the WDQ is a better measure of worry for this population, although additional research would be necessary to confirm this. The correlations clearly supported the theoretical association between AS, worry, and distress. The purpose of the regression analyses was to further examine the association between AS and worry, and test if there was something to the relationship in addition to the shared experience of distress. As hypothesized, AS was a significant predictor for both the WDQ and the PSWQ
1228
M. Floyd et al. / Personality and Individual Differences 38 (2005) 1223–1229
even after controlling for overall distress. However, the AS did not explain much additional variance. Considering the large correlation between the GSI and the two worry measures, this was a very strict test of the AS. As mentioned above, the GSI is such a broad instrument that it could be including items that are measures of both AS and worry. If this is not the case, then it suggests the majority of the association between worry and AS is general distress (or some other associated variable), and there is a relatively small element of worry that is accounted for specifically by AS. AS and worry both involve the tendency to focus on the uncertainty of the future and this could be the incremental contribution of AS to worry. Thus, it could be that AS, as a tendency to focus on the future, is one of the trait characteristics that predispose people to become worriers, even at the level of non-pathological worry. Longitudinal research would be necessary to examine the risk associated with high levels of AS. This study is limited by the lack of a clinical sample for comparison with the normal college student population. Having questions specific to GAD or other anxiety disorders would have been helpful in identifying participants within this sample that could have served as clinical analogs. Another limitation is that the survey method used in this study does not permit firm conclusions regarding the relation between anxiety sensitivity and worry. It would also be interesting to follow the participants in time to see if high levels of AS lead to pathological levels of worry. In conclusion, anxiety sensitivity has long been considered to be one of the basic fears and prior research has demonstrated its association with a number of anxiety disorders. This study demonstrated that AS is associated with worry, however the vast majority of the association appears to be due to shared elements of distress. The additional amount of worry variance explained by AS is most likely due to the tendency to focus on the uncertainty of the future.
References Borkovec, T. D., & Roemer, L. (1995). Perceived function of worry among generalized anxiety disorder subjects: Distraction from more emotionally distressing topics? Journal of Behavioral Therapy and Experimental Psychiatry, 26, 25–30. Brown, T. A., Antony, M. M., & Barlow, D. H. (1992). Psychometric properties of the Penn State Worry Questionnaire in a clinical anxiety disorders sample. Behaviour Research and Therapy, 30, 33–37. Cohen, J., & Cohen, P. (1983). Applied Multiple regression/correlation analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. Craske, M. G., Rapee, R. M., Jackel, L., & Barlow, D. H. (1989). Qualitative dimensions of worry in DSM-III-R generalized anxiety disorder subjects and nonanxious controls. Behaviour Research and Therapy, 27, 189–198. Cyr, J. J., McKenna-Foley, J. M., & Peacock, E. (1985). Factor structure of the SCL-90-R: Is there one? Journal of Personality Assessment, 49, 571–577. Davey, G. C. L., Hampton, J., Farrel, J., & Davidson, S. (1992). Some characteristics of worrying: Evidence for worrying and anxiety as separate constructs. Personality and Individual Differences, 13, 133–147. Derogatis, L. R. (1983). SCL-90R administration, scoring and procedures manual. Baltimore, MD: Johns Hopkins University Press. Dugas, M. J., Gosselin, P., & Ladouceur, R. (2001). Intolerance of uncertainty and worry: Investigating specificity in a nonclinical sample. Cognitive Therapy and Research, 25, 551–558. Dupuy, J.-B., Beaudoin, S., Rhe´aume, J., Ladouceur, R., & Dugas, M. J. (2001). Worry: Daily self-report in clinical and non-clinical populations. Behaviour Research and Therapy, 39, 1249–1255. Gross, P. R., & Eifert, G. H. (1990). Components of generalized anxiety: The role of intrusive thoughts vs worry. Behaviour Research and Therapy, 28, 421–428.
M. Floyd et al. / Personality and Individual Differences 38 (2005) 1223–1229
1229
Hoyer, J., Becker, E. S., & Roth, W. T. (2001). Characteristics of worry in GAD patients, social phobics, and controls. Depression and Anxiety, 13, 89–96. Joormann, J., & Sto¨ber, J. (1997). Measuring facets of worry: A LISREL analysis of the worry domains questionnaire. Personality and Individual Differences, 23, 827–837. Katon, W., & Roy-Byrne, P. P. (1991). Mixed anxiety and depression. Journal of Abnormal Psychology, 100, 337–345. Kelly, W. E. (2002). Some evidence for nonpathological and pathological worry as separate constructs: An investigation of worry and boredom. Personality and Individual Differences, 33, 345–354. McNally, R. J. (1990). Psychological approaches to panic disorder: A review. Psychological Bulletin, 108, 403–419. Meyer, T. J., Miller, M. L., Metzger, R. L., & Borkovec, T. D. (1990). Development of the Penn State Worry Questionnaire. Behaviour Research and Therapy, 28, 487–495. Otto, M. W., Pollack, M. H., Fava, M., Uccello, R., & Rosenbaum, J. F. (1995). Elevated Anxiety Sensitivity Index scores in patients with major depression: Correlates and changes with antidepressant treatment. Journal of Anxiety Disorders, 9, 117–123. Peterson, R. A., & Plehn, K. (1999). Measuring anxiety sensitivity. In S. Taylor (Ed.), Anxiety Sensitivity: Theory, Research, and Treatment of the Fear of Anxiety (pp. 61–81). Mahwah, NJ: Lawrence Erlbaum Associates. Reiss, S. (1991). The expectancy model of fear, anxiety, and panic. Clinical Psychology Review, 11, 141–153. Reiss, S., & Havercamp, S. H. (1996). The sensitivity theory of human motivation: Implications for psychopathology. Behaviour Research and Therapy, 34, 621–632. Reiss, S., & McNally, R. J. (1985). Expectancy model of fear. In S. Reiss & R. R. Bootzin (Eds.), Theoretical Issues in Behavior Therapy (pp. 107–121). Orlando: Academic Press, Inc. Reiss, S., Peterson, R. A., Gursky, D. M., & McNally, R. J. (1986). Anxiety sensitivity, anxiety frequency and the prediction of fearfulness. Behaviour Research and Therapy, 24, 1–8. Tallis, F., Eysenck, M., & Mathews, A. (1992). A questionnaire for the measurement of nonpathological worry. Personality and Individual Differences, 13, 161–168. Taylor, S., Koch, W. J., & Crockett, D. J. (1991). Anxiety sensitivity, trait anxiety, and the anxiety disorders. Journal of Anxiety Disorders, 5, 293–311. Taylor, S., Koch, W. J., & McNally, R. J. (1992a). How does anxiety sensitivity vary across the anxiety disorders? Journal of Anxiety Disorders, 6, 249–259. Taylor, S., Koch, W. J., McNally, R. J., & Crockett, D. J. (1992b). Conceptualizations of anxiety sensitivity. Psychological Assessment, 4, 245–250. Taylor, S., Koch, W. J., Woody, S., & McLean, P. (1996). Anxiety sensitivity and depression: How are they related? Journal of Abnormal Psychology, 105, 474–479.
Anxiety sensitivity and worry Mark Floyd *, Amber Garfield, Marcus T. LaSota Psychology Department, University of Nevada, Box 455030, Las Vegas, NV 89154-5030, United States Received 10 November 2003; received in revised form 17 May 2004; accepted 4 August 2004 Available online 5 November 2004
Abstract Anxiety sensitivity (AS), the fear of oneÕs response to anxiety provoking stimuli, has been correlated with anxiety disorders and has been theorized to be a risk factor in the development of anxiety disorders. Consistent with prior research (Dugas, Gosselin, & Ladouceur, 2001) it was hypothesized that AS and varying levels of worry share an underlying feature of anxiety specifically related to a perceived lack of control over future events, with a resultant emphasis on negative potential outcomes. In this study, the association between AS, overall distress, pathological worry, and non-pathological worry was investigated in a sample of 342 undergraduate volunteers. Results indicated that AS was significantly correlated with the Penn State Worry Questionnaire, the Worry Domain Questionnaire, and the Symptom Checklist 90-Revised. AS was a significant predictor of worry, even after factoring out overall distress, although the incremental amount of explained variance was small. Thus, AS and worry have more in common than distress. We speculate that the common element is a tendency to focus on the uncertainty of the future. Ó 2004 Elsevier Ltd. All rights reserved. Keywords: Anxiety sensitivity; Worry; Anxiety; Fear; Distress
1. Introduction Over the past two decades, the concept of a sensitivity to become fearful of the sensations associated with anxiety (i.e., anxiety sensitivity; Reiss & McNally, 1985) has garnered a great *
Corresponding author. Tel.: +1 702 895 0109. E-mail address: mfl[email protected] (M. Floyd).
0191-8869/$ - see front matter Ó 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.paid.2004.08.005
1224
M. Floyd et al. / Personality and Individual Differences 38 (2005) 1223–1229
deal of research support as a common factor among persons experiencing anxiety disorders and related conditions. Based on the expectancy model of fear (Reiss, 1991; Reiss & McNally, 1985), anxiety sensitivity (AS) is regarded as a unique characteristic of the individual that serves as a ‘‘predisposing factor in the development of anxiety-mediated disorders’’ (Reiss & McNally, 1985, p. 119). Reiss & McNallyÕs expectancy model considering ‘‘fear behavior’’ is comprised of two components: an expectancy that danger will occur (danger expectancy), which leads to avoidance behaviors, and a preemptive ‘‘fear of fear’’ that is comprised of both the expectation that anxiety will occur following exposure to an anxiety-provoking stimulus (anxiety expectancy) and the belief that the experience of anxiety will bring about feelings of illness, embarrassment, or possibly further anxiety symptoms (anxiety sensitivity). Thus, AS relates strictly towards the predisposed tendency to fear oneÕs reactions to an anxiety invoking stimulus and not the stimulus per se. Reiss and Havercamp (1996) posed an argument that, likely due to genetic and predispositional factors, some individuals are especially susceptible (i.e., those with high degrees of AS) to experience anxious symptoms as particularly unpleasant. Persons with high levels of AS will tend to worry that if forced to encounter an anxiety-inducing stressor, something unpredictable and harmful will occur (Reiss & McNally, 1985; Taylor, Koch, McNally, & Crockett, 1992b). To this extent, persons with high AS levels are likely to be highly alert to signals portending anxiety and to have exaggerated anxiety reactions (Reiss & McNally, 1985) when faced with such stimuli. In an effort to empirically assess AS, the Reiss–Epstein–Gursky Anxiety Sensitivity Scale, now better known as the Anxiety Sensitivity Index (ASI; Reiss, Peterson, Gursky, & McNally, 1986), was created. The ASI is a modified version of a scale created by Epstein (1982, as cited in Reiss et al., 1986), who demonstrated that AS is a qualitatively distinct construct from anxiety. ASI scores have been found to reliably differentiate anxious from non-anxious individuals (Reiss et al., 1986) and panic disorder from other anxiety disorders (Taylor, Koch, & Crockett, 1991). Consistent with AS and expectancy theories (Reiss, 1991; Reiss & McNally, 1985), research on the ASI has demonstrated that panic disorder tends to be most strongly associated with the endorsement of high ASI scores, and that persons diagnosed with post-traumatic stress disorder (McNally, 1990), generalized anxiety disorder (GAD), obsessive-compulsive disorder, and social phobia all had ASI scores higher than normal controls (Taylor et al., 1991; Taylor, Koch, & McNally, 1992a). Other researchers have suggested that the ASI may have applications beyond anxiety disorders. Depression has been speculated to hold a positive correlation to high AS, as two groups of researchers (Otto, Pollack, Fava, Uccello, & Rosenbaum, 1995; Taylor, Koch, Woody, & McLean, 1996) found persons with major depression to evidence ASI scores that, with the exception of panic disorder, were comparable to anxiety disorders. It is particularly curious as to why depression would relate to the construct of AS. In the current study, the concept of AS was extended to investigate the relationship between AS and worry. Although the primary feature of GAD when in excess, worry is believed to be a commonplace occurrence in the day-to-day lives of well-adjusted people (Dupuy, Beaudoin, Rhe´aume, Ladouceur, & Dugas, 2001; Joormann & Sto¨ber, 1997). The specific content of worry has not been shown to differ much between clinical worriers and non-clinical worriers (Hoyer, Becker, & Roth, 2001). However, the quality and degree of worry may be distinct, as those with GAD
M. Floyd et al. / Personality and Individual Differences 38 (2005) 1223–1229
1225
tend to worry more (Craske, Rapee, Jackel, & Barlow, 1989; Dupuy et al., 2001). Furthermore, worry, when in lesser degrees, has been associated with positive attributes such as preparation for actions, decision-making, and motivation towards accomplishing tasks (Borkovec & Roemer, 1995; Davey, Hampton, Farrel, & Davidson, 1992). While research exists that connects high ASI levels with the severity of GAD worry symptoms (Gross & Eifert, 1990; Taylor et al., 1992a), no known research studies have examined the connection between worry in general, overall distress, and AS. Conceptually, the constructs of worry and anxiety sensitivity appear to have similar elements. Worry is conceptualized as a fear or preoccupation with future outcomes and anxiety sensitivity is conceptualized as a fear of oneÕs response to anxiety symptoms, presumably upon exposure to future stimuli. The connection between AS and worry may be prospectively linked to a common experience of distress over uncertain future outcomes and a related intolerance for uncertainty (Dugas et al., 2001). Therefore, we hypothesized that AS would be significantly correlated with overall distress, pathological worry, and non-pathological worry. Additionally, we hypothesized that the connection between AS and worry was more than just a common experience of distress and that AS would be a significant predictor of both pathological and non-pathological worry, even after controlling for overall distress.
2. Methods 2.1. Participants Participants were 342 undergraduate student volunteers who received a Psychology 101 research credit for their participation. The average age of the sample was 19.7 years (SD = 3.3) and there were more women (59%) than men (41%). The self-reported ethnic composition of the sample was 60% Caucasian, 19% Asian-American, 10% African-American, 9% HispanicAmerican, 1% Native American, and 1% of mixed ethnic backgrounds. 2.2. Instruments The Anxiety Sensitivity Index (ASI; Reiss et al., 1986) was used as a measure of anxiety sensitivity. The ASI is a 16-item self-report questionnaire in which each item is rated on a five-point Likert scale from 0 (very little) to 4 (very much) endorsement. Each item reflects the belief that anxious sensations are unpleasantly experienced and potentially lead to harmful consequences. The degree to which one fears the experience of anxiety symptoms is reflected in higher scores. The ASI yields a total score (0–64 range) by adding the calculated responses for all 16 items. Using the total score for the instrument, the ASI has demonstrated sound psychometric properties, which include adequate test–retest reliability (Reiss et al., 1986), high internal consistency and good validity (see review by Peterson & Plehn, 1999). The Penn State Worry Questionnaire (PSWQ; Meyer, Miller, Metzger, & Borkovec, 1990) was used as a measure of pathological worry. Items on this measure are believed to reflect the uncontrollability and excessiveness of general worries. Meyer et al. (1990) reported that this measure has good psychometric properties, which include high internal consistency and good test–retest
1226
M. Floyd et al. / Personality and Individual Differences 38 (2005) 1223–1229
reliability. This instrument has also been found to reliably distinguish persons diagnosed with GAD from other anxiety disorders (Brown, Antony, & Barlow, 1992), and has been applied in both clinical and non-clinical populations. The Worry Domain Questionnaire (WDQ; Tallis, Eysenck, & Mathews, 1992) was used as a measure of non-pathological worry. This 30-item questionnaire is comprised of six subscales of five items each: (1) relationships, (2) lack of confidence, (3) aimless future, (4) work incompetence, (5) financial, and (6) socio-political. The total score is calculated by summing only the first five subscale domains (25 items), as the authors of the instrument have found that the sociopolitical domain may reflect social desirability. Although intended for use as a measure of non-clinical levels of worry, the WDQÕs authors suggest that the five ‘‘core’’ worry domains may apply generally to assess worry in clinical settings. The WDQ has been found to have adequate test–retest reliability, good internal consistency, and adequate validity properties (see review by Kelly, 2002). The Symptom Checklist 90, Revised (SCL-90R; Derogatis, 1983) was used as a measure of overall distress and psychopathology. This 90-item self-report instrument measures distress across nine symptom subscales, which include Somatization, Obsessive–Compulsive, Interpersonal Sensitivity, Depression, Anxiety, Hostility, Phobic Anxiety, Paranoid Ideation, and Psychoticism. There are three global indices: Global Severity Index (GSI), Positive Symptom Distress Index (PSDI), and Positive Symptom Total (PST). This scale has been found to have good psychometric properties, specifically as a measure of general distress (Cyr, McKenna-Foley, & Peacock, 1985). In this study, the GSI was used as the measure of general distress. 2.3. Procedure The participants entered a large classroom. They were given a packet of questionnaires, and then the experimenter read aloud the informed consent and instructions for the instruments. Participants were instructed to complete the instruments and turn them in to the experimenter, who then gave them a ‘‘receipt’’ indicating they participated in the study.
3. Results The averages and standard deviations for the instruments were: SCL-90R GSI (M = 0.75, SD = 0.58), ASI (M = 24.59, SD = 5.43), PSWQ (M = 43.52, SD = 10.17), and WDQ (M = 32.47, SD = 21.55). As predicted, the ASI was significantly correlated with the GSI, r = .57, p < .001; the WDQ, r = .55, p < .001; and the PSWQ, r = .41, p < .001. The GSI was also significantly correlated with the WDQ, r = .69, p < .001; and the PSWQ, r = .55, p < .001. The WDQ and PSWQ were also significantly correlated, r = .507, p < .001. Regression analyses were used to further test the strength of association of the ASI with worry. In separate regression equations, total scores on the PSWQ and WDQ were used as target variables, with the SCL-90R GSI scores and ASI scores used as predictors. As shown in Tables 1 and 2, the ASI was a significant predictor for both the WDQ and PSWQ, respectively even after controlling for overall distress (as measured by the GSI). The inclusion of the ASI in the model explained an additional 3% of the variance for the WDQ and an additional 2% of the variance for the PSWQ.
M. Floyd et al. / Personality and Individual Differences 38 (2005) 1223–1229
1227
Table 1 Summary of Regression Analyses for Anxiety Sensitivity and Non-pathological Worry Variable
B
SE B
b
p
GSI ASI
20.664 .850
1.757 .186
.564 .220
.000 .000
Note. R2 = .509. The dependent variable is the Worry Domain Questionnaire. GSI is the Global Severity Index of the Symptom Checklist 90-Revised. ASI is the total score of the Anxiety Sensitivity Index. Table 2 Summary of Regression Analyses for Anxiety Sensitivity and Pathological Worry Variable
B
SE B
b
p
GSI ASI
8.123 .294
1.005 .107
.460 .157
.000 .006
Note. R2 = .318. The dependent variable is the Penn State Worry Questionnaire. GSI is the Global Severity Index of the Symptom Checklist 90-Revised. ASI is the total score of the Anxiety Sensitivity Index.
4. Discussion As expected, the ASI was significantly correlated with both pathological (PSWQ) and nonpathological worry (WDQ). Furthermore, the ASI was significantly correlated with overall distress (GSI). The effect sizes of the associations, using typical values for effect size magnitude (Cohen & Cohen, 1983), were large for non-pathological worry and overall distress, and medium to large for pathological worry. In fact, the correlation between the ASI and the GSI was actually higher than the correlations between the ASI and either of the worry measures. Since the GSI measures overall distress, perhaps the ASI is tapping an element of general distress. Alternatively, it is possible the GSI is such a broad instrument that it includes items measuring AS and worry and the strong correlations are due in part to shared items. Another explanation is, given the inherent distress element common in depression and anxiety (Katon & Roy-Byrne, 1991), that depression and AS share a common feature of stress related to the unsettling perception of unpredictable, unknown, and possibly frightening future events (Dugas et al., 2001). The correlation between the ASI and the WDQ was considerably higher than the correlation between the ASI and the PSWQ. This could be due to using a sample of college students (instead of a clinical population) and their perception of the distress associated with worry. In comparing the items on the WDQ and PSWQ, it appears that the PSWQ focuses more on the out-of-control aspect of (pathological) worry whereas the WDQ focuses on type of worry and frequency of worry. Since the GSI and WDQ had the highest correlation of all the variables, it indicates that most participants in this study perceived worry to be somewhat distressing, even if their worry was not excessive, out of control, or functional. It also suggests the WDQ is a better measure of worry for this population, although additional research would be necessary to confirm this. The correlations clearly supported the theoretical association between AS, worry, and distress. The purpose of the regression analyses was to further examine the association between AS and worry, and test if there was something to the relationship in addition to the shared experience of distress. As hypothesized, AS was a significant predictor for both the WDQ and the PSWQ
1228
M. Floyd et al. / Personality and Individual Differences 38 (2005) 1223–1229
even after controlling for overall distress. However, the AS did not explain much additional variance. Considering the large correlation between the GSI and the two worry measures, this was a very strict test of the AS. As mentioned above, the GSI is such a broad instrument that it could be including items that are measures of both AS and worry. If this is not the case, then it suggests the majority of the association between worry and AS is general distress (or some other associated variable), and there is a relatively small element of worry that is accounted for specifically by AS. AS and worry both involve the tendency to focus on the uncertainty of the future and this could be the incremental contribution of AS to worry. Thus, it could be that AS, as a tendency to focus on the future, is one of the trait characteristics that predispose people to become worriers, even at the level of non-pathological worry. Longitudinal research would be necessary to examine the risk associated with high levels of AS. This study is limited by the lack of a clinical sample for comparison with the normal college student population. Having questions specific to GAD or other anxiety disorders would have been helpful in identifying participants within this sample that could have served as clinical analogs. Another limitation is that the survey method used in this study does not permit firm conclusions regarding the relation between anxiety sensitivity and worry. It would also be interesting to follow the participants in time to see if high levels of AS lead to pathological levels of worry. In conclusion, anxiety sensitivity has long been considered to be one of the basic fears and prior research has demonstrated its association with a number of anxiety disorders. This study demonstrated that AS is associated with worry, however the vast majority of the association appears to be due to shared elements of distress. The additional amount of worry variance explained by AS is most likely due to the tendency to focus on the uncertainty of the future.
References Borkovec, T. D., & Roemer, L. (1995). Perceived function of worry among generalized anxiety disorder subjects: Distraction from more emotionally distressing topics? Journal of Behavioral Therapy and Experimental Psychiatry, 26, 25–30. Brown, T. A., Antony, M. M., & Barlow, D. H. (1992). Psychometric properties of the Penn State Worry Questionnaire in a clinical anxiety disorders sample. Behaviour Research and Therapy, 30, 33–37. Cohen, J., & Cohen, P. (1983). Applied Multiple regression/correlation analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum Associates, Inc. Craske, M. G., Rapee, R. M., Jackel, L., & Barlow, D. H. (1989). Qualitative dimensions of worry in DSM-III-R generalized anxiety disorder subjects and nonanxious controls. Behaviour Research and Therapy, 27, 189–198. Cyr, J. J., McKenna-Foley, J. M., & Peacock, E. (1985). Factor structure of the SCL-90-R: Is there one? Journal of Personality Assessment, 49, 571–577. Davey, G. C. L., Hampton, J., Farrel, J., & Davidson, S. (1992). Some characteristics of worrying: Evidence for worrying and anxiety as separate constructs. Personality and Individual Differences, 13, 133–147. Derogatis, L. R. (1983). SCL-90R administration, scoring and procedures manual. Baltimore, MD: Johns Hopkins University Press. Dugas, M. J., Gosselin, P., & Ladouceur, R. (2001). Intolerance of uncertainty and worry: Investigating specificity in a nonclinical sample. Cognitive Therapy and Research, 25, 551–558. Dupuy, J.-B., Beaudoin, S., Rhe´aume, J., Ladouceur, R., & Dugas, M. J. (2001). Worry: Daily self-report in clinical and non-clinical populations. Behaviour Research and Therapy, 39, 1249–1255. Gross, P. R., & Eifert, G. H. (1990). Components of generalized anxiety: The role of intrusive thoughts vs worry. Behaviour Research and Therapy, 28, 421–428.
M. Floyd et al. / Personality and Individual Differences 38 (2005) 1223–1229
1229
Hoyer, J., Becker, E. S., & Roth, W. T. (2001). Characteristics of worry in GAD patients, social phobics, and controls. Depression and Anxiety, 13, 89–96. Joormann, J., & Sto¨ber, J. (1997). Measuring facets of worry: A LISREL analysis of the worry domains questionnaire. Personality and Individual Differences, 23, 827–837. Katon, W., & Roy-Byrne, P. P. (1991). Mixed anxiety and depression. Journal of Abnormal Psychology, 100, 337–345. Kelly, W. E. (2002). Some evidence for nonpathological and pathological worry as separate constructs: An investigation of worry and boredom. Personality and Individual Differences, 33, 345–354. McNally, R. J. (1990). Psychological approaches to panic disorder: A review. Psychological Bulletin, 108, 403–419. Meyer, T. J., Miller, M. L., Metzger, R. L., & Borkovec, T. D. (1990). Development of the Penn State Worry Questionnaire. Behaviour Research and Therapy, 28, 487–495. Otto, M. W., Pollack, M. H., Fava, M., Uccello, R., & Rosenbaum, J. F. (1995). Elevated Anxiety Sensitivity Index scores in patients with major depression: Correlates and changes with antidepressant treatment. Journal of Anxiety Disorders, 9, 117–123. Peterson, R. A., & Plehn, K. (1999). Measuring anxiety sensitivity. In S. Taylor (Ed.), Anxiety Sensitivity: Theory, Research, and Treatment of the Fear of Anxiety (pp. 61–81). Mahwah, NJ: Lawrence Erlbaum Associates. Reiss, S. (1991). The expectancy model of fear, anxiety, and panic. Clinical Psychology Review, 11, 141–153. Reiss, S., & Havercamp, S. H. (1996). The sensitivity theory of human motivation: Implications for psychopathology. Behaviour Research and Therapy, 34, 621–632. Reiss, S., & McNally, R. J. (1985). Expectancy model of fear. In S. Reiss & R. R. Bootzin (Eds.), Theoretical Issues in Behavior Therapy (pp. 107–121). Orlando: Academic Press, Inc. Reiss, S., Peterson, R. A., Gursky, D. M., & McNally, R. J. (1986). Anxiety sensitivity, anxiety frequency and the prediction of fearfulness. Behaviour Research and Therapy, 24, 1–8. Tallis, F., Eysenck, M., & Mathews, A. (1992). A questionnaire for the measurement of nonpathological worry. Personality and Individual Differences, 13, 161–168. Taylor, S., Koch, W. J., & Crockett, D. J. (1991). Anxiety sensitivity, trait anxiety, and the anxiety disorders. Journal of Anxiety Disorders, 5, 293–311. Taylor, S., Koch, W. J., & McNally, R. J. (1992a). How does anxiety sensitivity vary across the anxiety disorders? Journal of Anxiety Disorders, 6, 249–259. Taylor, S., Koch, W. J., McNally, R. J., & Crockett, D. J. (1992b). Conceptualizations of anxiety sensitivity. Psychological Assessment, 4, 245–250. Taylor, S., Koch, W. J., Woody, S., & McLean, P. (1996). Anxiety sensitivity and depression: How are they related? Journal of Abnormal Psychology, 105, 474–479.