- Email: [email protected]
Anxiety Sensitivity and Situation-Specific Drinking
Pergamon
Journal of Anxiety Disorders, Vol. 12, No. 5, pp. 407–419, 1998 Copyright 1998 Elsevier Science Ltd Printed in the USA. All rights reserved 0887-6185/98 $19.00 ⫹ .00
PII S0887-6185(98)00025-5
RESEARCH PAPERS
Anxiety Sensitivity and Situation-Specific Drinking Sarah Barton Samoluk, ph.d., and Sherry H. Stewart, ph.d. Dalhousie University
Abstract—To assess the typical drinking situations of high anxiety sensitive individuals, relative to low anxiety sensitive individuals, a large sample of university student drinkers (N ⫽ 396) completed the 42-item version of the Inventory of Drinking Situations (IDS-42) and the Anxiety Sensitivity Index (ASI). As hypothesized, ASI scores were significantly positively correlated with drinking frequency on the higher order IDS-42 factor of negatively reinforcing drinking situations and not significantly correlated with drinking frequency on the higher order factor of positively reinforcing drinking situations. Also, ASI scores were significantly positively correlated with drinking frequency on the lower order drinking situations factors of Conflict with Others, Unpleasant Emotions, and Physical Discomfort. ASI scores were also unexpectedly significantly correlated with more frequent drinking on the higher order factor of temptation drinking situations and with more frequent drinking on the lower order drinking situations factor of Testing Personal Control. These results have important implications for designing early prevention and intervention programs for high anxiety sensitive university student drinkers by helping them to develop alternatives to drinking in those situations (e.g., negatively reinforcing drinking situations) which place them at greatest risk for alcohol misuse. 1998 Elsevier Science Ltd
This research was part of Dr. Samoluk’s dissertation under the supervision of Dr. Stewart. This study was supported by a Medical Research Council of Canada Operating Grant (MT-12514). Dr. Samoluk was supported by a Social Sciences and Humanities Research Council of Canada doctoral fellowship. The authors thank Jennifer Bird, Brent Conrad, and Carol Thompson for their assistance with questionnaire scoring and data entry, and Monica Ghosal for her assistance in questionnaire administration to the University of Toronto student sample. Requests for reprints should be sent to Dr. Sherry H. Stewart, Department of Psychology, Dalhousie University, Halifax, Nova Scotia B3H 4J1, Canada. E-mail: [email protected] or ssamoluk @is2.dal.ca
407
408
S. B. SAMOLUK AND S. H. STEWART
Anxiety Sensitivity (AS) is a cognitive individual difference variable characterized by a fear of anxiety sensations (fear of fear) and reflects the tendency to catastrophize about the consequences of such sensations (e.g., physical illness, social embarrassment, mental incapacitation; Peterson & Reiss, 1992; Stewart, Conrod, Gignac, & Pihl, 1998; Stewart, Taylor, & Baker, 1997). Fear of fear (AS) is particularly common in individuals who suffer from panic disorder with or without agoraphobia and from post-traumatic stress disorder (Peterson & Reiss, 1992). Moreover, emerging evidence suggests an important relationship between AS and heavier alcohol use (for reviews, see McNally, 1996; Stewart, Samoluk, & MacDonald, in press). For example, higher than average AS levels have been found in clinical samples diagnosed with DSM-III-R (Diagnostic and Statistical Manual of Mental Disorders, third edition, revised; American Psychiatric Association, 1987) alcohol abuse/dependence diagnoses whether or not patients are diagnosed with a co-morbid anxiety disorder (see review by McNally, 1996; Karp, 1993). Moreover, high AS university students report drinking more alcoholic beverages on a weekly basis, and drinking to “excess” (to the point of legal intoxication; Conrod, Stewart, & Pihl, 1997) more frequently, than low AS students (Stewart, Peterson, & Pihl, 1995). The relationship between AS and heavier alcohol consumption is partially clarified by exploration of self-perceived motives for alcohol use. AS was found to be significantly correlated with the frequency of self-reported use of alcohol to “cope” (i.e., to reduce/avoid negative affect such as anxiety or depression) among a university sample (Stewart & Zeitlin, 1995) using the Drinking Motives Questionnaire—a validated measure of drinking motives (Cooper, Russell, Skinner, & Windle, 1992; Stewart, Zeitlin, & Samoluk, 1996). Also high AS students were more likely to report drinking primarily for coping motives, and less likely to report drinking primarily for social motives (i.e., to affiliate with others), compared with low AS students (see also Stewart, Karp, Pihl, & Peterson, 1997). High AS students’ increased use of alcohol to cope (i.e., for negative reinforcement) is likely related to the enhanced sensitivity to alcohol’s emotional reactivity-dampening effects found among high AS young adults compared with low AS young adults in studies using the alcohol challenge paradigm (Baker, MacDonald, Stewart, & Skinner, 1998; Conrod, Pihl, & Vassileva, 1998; Stewart & Pihl, 1994). Different drinking motives have been found to be associated with different situational antecedents and consequences of alcohol use (Cooper, 1994, Cooper et al., 1992). Individuals who drink primarily for coping-related motives are more likely to drink heavily, to drink alone, to drink at home, and to experience alcohol-related problems, compared with those who drink primarily for more normative social motives. In contrast, those who drink primarily to socialize and affiliate with others are more likely to drink at parties and with mixed-sex friends, and are at lower risk for alcohol-related problems than primarily coping-motivated drinkers (Cooper, 1994; Cooper et al., 1992).
AS AND SITUATION-SPECIFIC DRINKING
409
Behaviourally oriented research has focused on developing measures to assess situational antecedents to alcohol consumption. One such measure is the 100-item Inventory of Drinking Situations and its shorter 42-item version (IDS and IDS-42; Annis, Graham, & Davis, 1987). In a large sample of university students, the IDS-42 was found to possess eight lower order factors consistent with Marlatt’s work (e.g., Marlatt & Gordon, 1980) on the situational antecedents to relapse among alcoholics from which the eight IDS subscales were derived (Carrigan, Samoluk, & Stewart, 1998). These eight lower order factors (Conflict with Others, Unpleasant Emotions, Physical Discomfort, Pleasant Times with Others, Social Pressure to Drink, Pleasant Emotions, Testing Personal Control, and Urges and Temptations) were found to load most optimally on three correlated higher order factors (Carrigan et al., 1998). The Conflict with Others, Unpleasant Emotions, and Physical Discomfort subscales loaded on a higher order factor of negatively reinforcing drinking situations; the Pleasant Times with Others, Social Pressure to Drink, and Pleasant Emotions subscales loaded on a higher order factor of positively reinforcing drinking situations; and the Testing Personal Control and Urges and Temptations subscales loaded on a higher order factor of temptation drinking situations (Carrigan et al., 1998). Using a parallel instrument to the IDS, Turner, Annis, and Sklar (1997) similarly identified a hierarchical structure involving eight lower order factors corresponding to the eight IDS subscales, and three higher order factors of negative situations, positive situations, and temptation situations, in a sample of clients receiving addiction treatment.1 Researchers are beginning to examine whether there are relationships between individual difference factors and the situations in which drinking most frequently occurs. For example, in a sample of clients receiving treatment for alcohol dependence, Annis et al. (1987) found frequent drinking in situations involving negative personal states (specifically, relative peaks on the IDS subscales of Physical Discomfort, Unpleasant Emotions, and Conflict with Others) was reported more often by women, and by individuals who tend to drink alone, and have more years of heavy alcohol use. Among a sample of alcoholics and cocaine addicts, a significant positive association was found between levels of negative temperament (i.e., anxiety, depression, boredom, anger) and frequency of drinking on a “negative” factor of the IDS which consisted of items from the Unpleasant Emotions and Conflict with Others subscales (Cannon et al., 1992). Moreover, Turner et al. (1997) found that scores on the Unpleasant Emotions, Conflict with Others, and Physical Discomfort subscales were significantly correlated with measures of negative temperament, including the depression, interpersonal sensitivity, and somatization subscales 1 Turner, Annis, and Sklar (1997) explored situational antecedents to the use of a variety of drugs including alcohol using the Inventory of Drug-Taking Situations (IDTS). The 8 subscales and 50 items of the IDTS were derived directly from the 100-item version of the Inventory of Drinking Situations and extended to the use of drugs in addition to alcohol.
410
S. B. SAMOLUK AND S. H. STEWART
of the Hopkins Symptom Checklist (SCL-90; Derogatis, 1979), in a sample of clients receiving addiction treatment. The present study was designed to explore the relationship between AS (as measured using the Anxiety Sensitivity Index [ASI]; Peterson & Reiss, 1992) and situational antecedents to alcohol consumption using the IDS-42 among a nonclinical sample of university students. Given the significant positive correlation between AS levels and coping-motivated drinking (Stewart & Zeitlin, 1995), we made several hypotheses pertaining to the relationship between ASI scores and the higher and lower order factors of the IDS-42. First, we predicted a significant positive correlation between ASI scores and the frequency of self-reported drinking on the higher order factor of negatively reinforcing drinking situations. We predicted that correlations between ASI scores and the higher order positively reinforcing and temptation drinking situations factors would not be significant. We also predicted that the relationship between ASI scores and the frequency of drinking in negatively reinforcing situations would be stronger than the relationship between ASI scores and the frequency of drinking in positively reinforcing situations or temptation situations. At the lower order level, we predicted significant positive correlations between ASI scores and the lower order IDS-42 factors of Conflict with Others, Unpleasant Emotions, and Physical Discomfort. We predicted that correlations between ASI scores and the lower order factors of Pleasant Times with Others, Social Pressure to Drink, Pleasant Emotions, Testing Personal Control, and Urges and Temptations would not be significant.
METHOD Participants Participants were 473 undergraduate students (338 female, 133 male, 2 gender unspecified), whose average age was 21.72 years (SD ⫽ 4.22). All participants were enrolled in undergraduate psychology courses at Dalhousie University or the University of Toronto, Canada. Materials Inventory of Drinking Situations. The shortened, 42-item version of the Inventory of Drinking Situations (IDS-42; Annis et al., 1987) was used to assess frequency of drinking across specific situations. The IDS-42 contains 8 subscales: The Conflict with Others subscale has 12 items, the Pleasant Times with Others subscale has 6 items, and the remaining subscales have 4 items each. For the present study, the IDS-42 instructions were modified from “frequency of heavy drinking” to “frequency of drinking” in order to more aptly assess the situational antecedents of students’ customary drinking, and to allow completion of the IDS-42 by student drinkers who do not frequently engage in
AS AND SITUATION-SPECIFIC DRINKING
411
“heavy” drinking (cf. Bruce & Pihl, 1997; Carrigan et al., 1998). Participants rated their frequency of drinking over the past year in each specified situation on a scale of 1 (never drank in that situation) to 4 (always drank in that situation). The scale was then recoded to range from 0 to 3 (Annis et al., 1987). Higher scores indicate a relatively higher frequency of drinking in the particular situation. The original IDS-100 and its shortened IDS-42 has been shown to possess good psychometric properties in clinical samples of alcohol abusers. These two versions are essentially equivalent with subscales correlations ranging between .93 and .98 (Annis et al., 1987). The convergent and discriminant validity of the IDS has been demonstrated in alcoholic samples. For example, percentage of time spent drinking alone correlated significantly and positively with negative drinking situations (e.g., Unpleasant Emotions) and correlated negatively with positive drinking situations (e.g., Pleasant Times with Others; Annis et al., 1987). The present version of the IDS-42, using the modified instructional set noted above, possesses good psychometric properties in university students (Carrigan et al., 1998). For example, Carrigan et al. found that the eight lower order IDS-42 factors have excellent concurrent validity with conceptually similar subscales on the Drinking Motives Questionnaire (Cooper et al., 1992). For the present study, the data were analyzed across the three higher order IDS-42 factors of negatively reinforcing drinking situations, positively reinforcing drinking situations, and temptation drinking situations, as well as across all eight lower order factors (equivalent to the eight IDS-42 subscales). Lower order factor scores were computed as weighted means using maximum likelihood factor loadings for the eight lower order factors as weights (Carrigan et al., 1998). Higher order factor scores were likewise computed as weighted means using maximum likelihood factor loadings for the higher order factors as weights applied to the lower order factor scores (Carrigan et al., 1998).2 The negatively reinforcing drinking situations factor includes situational antecedents to alcohol use that pertain to negative affect and unpleasant associations with others which might be assuaged via alcohol consumption (Carrigan et al., 1998): “When other people didn’t seem to like me,” “When I felt that I had let myself down,” and “When my stomach felt like it was tied in knots” are representative items from the Conflict with Others, Unpleasant Emotions, and Physical Discomfort subscales, respectively, which are subsumed under the higher order negatively reinforcing drinking situations factor. The positively reinforcing drinking situations factor includes situational antecedents to alcohol use that pertain to positive affect and pleasant associations with others which might be heightened via alcohol consumption (Carrigan et al., 1998): “When I wanted to celebrate with a friend,” “When I was at 2 Maximum likelihood factor loadings can be found in Carrigan, Samoluk, and Stewart (1998).
412
S. B. SAMOLUK AND S. H. STEWART
a party and other people were drinking,” and “When something good happened and I felt like celebrating” are representative items from the Pleasant Times with Others, Social Pressure to Drink,3 and Pleasant Emotions subscales, respectively, which are subsumed under the higher order positively reinforcing drinking situations factor. The temptation drinking situations factor appears to tap a cognitive preoccupation with drinking (Carrigan et al., 1998): “When I wanted to prove to myself that I could take a few drinks without becoming drunk,” and “When I remembered how good it tasted” are representative items from the Testing Personal Control and Urges and Temptations subscales, respectively, which are subsumed under the higher order temptation drinking situations factor. Anxiety Sensitivity Index. AS levels were measured using the 16-item Anxiety Sensitivity Index (ASI; Peterson & Reiss, 1992). The ASI assesses subjective fears of anxiety sensations (e.g., “It scares me when I feel faint”) and beliefs about why those sensations are harmful (e.g., “When my stomach is upset, I worry that I might be seriously ill”) (Taylor, 1995). Participants rate the degree to which each item describes themselves on a scale of 0 (very little) to 4 (very much). The total ASI score is computed as the sum of all item scores. Higher scores indicate relatively higher levels of fear of anxiety. The ASI has been shown to possess excellent psychometric properties in clinical samples and in nonclinical university student samples (see review by Peterson & Reiss, 1992). In our laboratory, the ASI was found to have an alpha coefficient of .88 in a large sample of 551 university students (Watt, Stewart, & Cox, 1998). This is within the range of reliabilities reported by Peterson and Reiss (1992). Procedure Participants provided informed consent and basic demographic information (age and gender). Only those participants who indicated that they had consumed alcohol within the last year were identified as “drinkers” for the present study and asked to complete the IDS-42. All participants completed the demographic measures. All questionnaires were completed anonymously during class time.
3 Items from this subscale have consistently been found to load on a factor reflecting drinking in pleasant social situations (Annis, Graham, & Davis, 1987; Cannon, Leeka, Patterson, & Baker, 1990; Carrigan, Samoluk, & Stewart, 1998). This indicates drinking for positive reinforcement as opposed to drinking for negative reinforcement implied in social conformity drinking (Cooper, 1994). Thus, the label “Social Cues to Drink” (Carrigan et al., 1998) may more aptly represent the content of the items forming the IDS-42 “Social Pressure to Drink” subscale (e.g., “When I was in a restaurant and the people with me ordered drinks”).
413
AS AND SITUATION-SPECIFIC DRINKING
TABLE 1 Correlations Between ASI Scores and IDS-42 Higher and Lower Order Factor Scores
IDS-42 Higher Order Factors and Subscales
Coefficient Alpha
Correlation Coefficients
.91 .88 .83 .32 .93 .85 .86 .83 .72 .72 .63
.20* .22* .17* .16* .12 .13 .10 .10 .19* .27* .10
Negatively Reinforcing Situationsa Conflict With Others Unpleasant Emotions Physical Discomfort Positively Reinforcing Situations Pleasant Times With Others Social Pressure to Drink Pleasant Emotions Temptation Situations Testing Personal Control Urges and Temptations
Disattenuated Correlations With ASI Scores .22*a .25* .19* .29* .13ab .14 .11 .11 .23b .34* .14
Notes. The three higher order factors indicated in bold type were found among university student drinkers using confirmatory factor analysis (Carrigan et al., 1998). Correlation coefficients with the same subscripts differ significantly (for subscript “a,” p ⬍ .05, one-tailed; for subscript “b,” p ⬍ .01, two-tailed). a Probability levels for this IDS-42 higher order factor and its lower order subscales are one-tailed in keeping with the hypotheses. All remaining probability levels are two-tailed. * p ⱕ .001.
RESULTS The mean ASI score for the entire sample was 20.10 (SD ⫽ 9.22) which is consistent with previously established nonclinical norms on this measure (Peterson & Reiss, 1992). Of the total sample, 83.7% (n ⫽ 396; 283 female, 111 male, 2 gender unspecified) reported alcohol consumption during the past year. The drinkers (n ⫽ 395, 1 age unspecified) were slightly older than the nondrinkers (n ⫽ 77): mean age (and SD) ⫽ 21.89 (4.27) versus 20.81 (3.82) years, respectively, F(1, 470) ⫽ 4.33, p ⬍ .05. However, the drinkers (n ⫽ 394, 2 gender unspecified) did not differ significantly from the nondrinkers (n ⫽ 77) in gender composition (71.4% vs. 71.8% female, respectively), 2(1) ⫽ .005, ns. All subsequent analyses were conducted using data from the subsample of 396 self-reported drinkers. Reliability of the IDS-42 Subscales The internal consistency of each of the eight IDS-42 subscales separately and combined as higher order factors was calculated using Cronbach’s coefficient alpha (as presented in Table 1) in order to determine if reliabilities differ markedly across subscales and factors. The majority of subscales had alphas of .83 or better, indicating substantial internal consistency. The Testing Personal
414
S. B. SAMOLUK AND S. H. STEWART
Control and Urges and Temptations subscales had moderate alphas of .72 and .63, respectively. Physical Discomfort was the least reliable subscale with an alpha of .32.4 The alphas for the higher order factors varied from .72 to .93, indicating moderate to substantial internal consistency (see Table 1). Correlational Analyses All correlational analyses between ASI scores and higher order factor scores, and between ASI scores and lower order factor scores, were calculated using a Bonferroni-adjusted alpha of .005 (.05/11 comparisons) for determining statistical significance.5 Given the variability in reliabilities across subscales and factors (see Table 1), correlations are presented with and without correction for attenuation in Table 1. As predicted, ASI scores were significantly positively correlated with negatively reinforcing drinking situations factor scores (p ⬍ .001, one-tailed). Also, as predicted, ASI scores were not significantly associated with positively reinforcing drinking situations factor scores. Unexpectedly, however, ASI scores were significantly positively correlated with temptation drinking situations factor scores (p ⬍ .001, two-tailed). The pattern of results remained the same prior to correction for attenuation (see Table 1). Correlations were also calculated between ASI scores and the eight lower order factor scores (see Table 1). ASI scores showed significant positive associations with the frequency of drinking in situations involving Conflict with Others, Unpleasant Emotions, and Physical Discomfort (all ps ⱕ .001, onetailed) as predicted. Also as predicted, ASI scores were not significantly associated with the frequency of drinking in situations involving Pleasant Times with Others, Social Pressure to Drink, or Pleasant Emotions (all ps ⬎ .005, two-tailed). ASI scores were not significantly correlated with the frequency of drinking in situations involving Urges and Temptations as hypothesized; but unexpectedly ASI scores were significantly positively correlated with the frequency of drinking in situations involving Testing Personal Control (p ⬍ .001, two-tailed). The pattern of results remained the same prior to correction for attenuation (see Table 1).
4
Carrigan, Samoluk, and Stewart (1998) report that the relatively low internal consistency of the Physical Discomfort subscale should be expected in a nonclinical university student sample of drinkers who are less likely than alcoholics to use alcohol to alleviate physical discomfort. Given the brevity of the Physical Discomfort subscale (four items) and its heterogenous content, a larger number of more homogeneous items might improve the reliability of the Physical Discomfort IDS-42 Subscale for university students. 5 The pattern of significant results between ASI scores and the IDS-42 scores remained the same whether or not factor scores (weighted means) or subscale scores (unweighted means) were used.
AS AND SITUATION-SPECIFIC DRINKING
415
Tests of Difference Between Dependent Correlations The significance of the correlational differences among the negatively reinforcing, positively reinforcing, and temptation drinking situations IDS-42 higher order factors was calculated using a t equation, for nonindependent correlations, with the disattenuated coefficients (see Table 1).6 The t equation revealed that, as predicted, negatively reinforcing drinking situations factor scores were more highly correlated with ASI scores than positively reinforcing drinking situations factor scores, t(393) ⫽ 1.66, p ⬍ .05, one-tailed. However, negatively reinforcing drinking situations factor scores were not more highly correlated with ASI scores than temptation drinking situation factor scores, t(393) ⫽ ⫺0.24, ns. Temptation drinking situations factor scores were more highly correlated with ASI scores than positively reinforcing drinking situations factor scores, t(393) ⫽ 2.75, p ⬍ .01, two-tailed. In sum, these findings suggest that ASI scores are more highly predictive of drinking in negatively reinforcing and temptation situations than of drinking in positively reinforcing situations.
DISCUSSION As hypothesized, ASI scores were positively and significantly correlated with the self-reported frequency of drinking in negatively reinforcing situations. Also as predicted, ASI scores were not significantly associated with the reported frequency of drinking in positively reinforcing situations. Analysis of each lower order factor score confirmed this relationship between higher ASI scores and situation-specific drinking in negatively reinforcing situations but not in positively reinforcing situations. Namely, while ASI scores were significantly and positively associated with drinking in situations involving Conflict with Others, Unpleasant Emotions, and Physical Discomfort, ASI scores were not significantly correlated with drinking in situations involving Pleasant Times with Others, Social Pressure to Drink, or Pleasant Emotions. In the general population, as well as in problem drinkers, a tendency to drink in negative affect states is correlated with increasing levels of alcohol dependence, whereas drinking in positive affect states is associated with less risky social drinking (Cunningham, Sobell, Sobell, Gavin, & Annis, 1995). Similar to previous findings which link general negative temperament to drinking in IDS negative situations among problem drinkers (Cannon et al., 1992), the results of the present correlational analyses suggest that nonclinical drinkers with higher AS levels appear more motivated than others to drink in aversive situations. This may place them at risk for alcohol problems. These
6
The t equation was devised by Williams (1959). It consists of a ratio distributed as t with N ⫺ 3 df, and takes into account the degree to which two dependent tests are correlated. A simple explanation of the use of this formula can be found in Howell (1987).
416
S. B. SAMOLUK AND S. H. STEWART
results are consistent with other studies which found that ASI scores are significantly and positively associated with self-reported frequency of drinking to cope (cf. Stewart, Karp, et al., 1997; Stewart & Zeitlin, 1995) and with drinking in response to physical discomfort in a laboratory setting (Samoluk & Stewart, 1996). The present results are also consistent with the finding that individuals with higher AS levels are more sensitive to alcohol’s emotional reactivitydampening properties (cf. Baker et al., 1998; Conrod et al., 1998; Stewart & Pihl, 1994) than individuals with lower AS levels. ASI scores were only significantly correlated with situations that do not necessitate the presence of others (Unpleasant Emotions, Physical Discomfort, Conflict with Others, and Testing Personal Control).7 This suggests that high AS persons are likely to be relatively solitary drinkers. Solitary drinking is more highly associated with alcohol problems than is social drinking (Cooper et al., 1992). The relatively solitary drinking style of high AS young adults has recently been supported in a lab-based analogue alcohol consumption study in which high AS individuals were found to drink more than low AS individuals in a solitary, but not a social, drinking context (Samoluk, Stewart, Sweet, & MacDonald, 1998). Contrary to hypothesis, ASI scores were positively and significantly correlated with the temptation drinking situations factor scores. Indeed, the strength of the association with ASI scores was not significantly different for the negatively reinforcing drinking situations factor and the temptation drinking situations factor. Correlational tests of lower order factor scores further revealed that ASI scores were correlated with the Testing Personal Control factor scores but not the Urges and Temptations factor scores. These results suggest that individuals high in AS drink more frequently for both negative reinforcement (i.e., to attenuate negative affect and physical discomfort) and for the temptation to rationalize continued alcohol use. The cognitive preoccupation with drinking reflected in the Testing Personal Control IDS-42 factor items appears to be conceptually analogous to the concept of restrained drinking (Collins, 1993). It has been theorized that failed attempts to regulate alcohol intake among “restrained drinkers” may actually promote binge drinking when restrained drinkers blame themselves for their failure to regulate their alcohol consumption (Collins, 1993). This self-blame may lead to further excessive drinking in order to assuage the resultant low mood of restrained drinkers in a continuous cycle of restraint, violation, negative affect, and excessive drinking (Marlatt, 1985; Marlatt & Gordon, 1980). Given their apparently increased cognitive preoccupation with drinking, high AS individuals would likely score relatively higher than others on restrained drinking measures. (See Collins, 1993, for a review of measures of restrained drinking.) 7 Situations involving Conflict with Others are included in this list because solitary drinking can occur following conflict as a way to cope with the resultant negative affect.
AS AND SITUATION-SPECIFIC DRINKING
417
The present study did not include the assessment of higher order negative temperament constructs related to both anxiety sensitivity and alcoholism, such as measures of neuroticism, trait anxiety or dysphoria (see reviews by Lilienfeld, 1996; Peterson & Reiss, 1992; Tarter, Alterman, & Edwards, 1985). Hence, this study cannot identify the degree to which the results are due specifically to AS or to other higher order personality constructs. Given the specific fear of anxiety-related sensations among high AS individuals (Peterson & Reiss, 1992), one might speculate that AS should be most strongly associated with drinking frequency in negatively reinforcing situations involving anxious emotions and physical discomfort. In contrast, the relationship between negative temperament, in general, and drinking frequency should be less situationally specific, including significant relationships between negative temperament and drinking frequency in a variety of negatively reinforcing situations (e.g., anxiety, dysphoria, anger, etc.). Unfortunately, the structure of the IDS-42 does not permit testing of this speculation (e.g., the Unpleasant Emotions subscale encompasses anxiety, depression, and anger situations). However, in a study on drinking motives, Stewart, Karp, et al. (1997) obtained equally strong and significant relationships between AS and anxiety-related reasons for alcohol use as they did between AS and depression-related reasons for alcohol use, thus failing to support a specific association between AS and anxietymotivated drinking. Future research could clarify the degree to which the present results are due specifically to AS and/or to other higher order negative temperament constructs by including the ASI and measures of such additional constructs in a multiple regression design. The information provided by the IDS has proven to be useful in relapse prevention alcohol abuse treatment programs (Marlatt, 1985; Marlatt & Gordon, 1980; Sobell & Sobell, 1993). In a similar manner, the observed situational antecedents of more frequent drinking in relatively solitary negatively reinforcing and temptation drinking situations among high AS university students, relative to low AS university students, could become the focus of prevention and treatment. Specifically, high AS young adults could be trained to develop alternative coping strategies to alcohol consumption for dealing with these particular high-risk situations.
REFERENCES American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorders (3rd ed., rev.). Washington, DC: Author. Annis, H. M., Graham, J. M., & Davis, C. S. (1987). Inventory of Drinking Situations (IDS) user’s guide. Toronto: Alcoholism and Drug Addiction Research Foundation. Baker, J. M., MacDonald, A. B., Stewart, S. H., & Skinner, M. (1998). The effects of alcohol on response to voluntary hyperventilation among individuals high and low in anxiety sensitivity. Manuscript submitted for publication.
418
S. B. SAMOLUK AND S. H. STEWART
Bruce, K. R., & Pihl, R. O. (1997). Forget “Drinking to Forget”: Enhanced consolidation of emotionally charged memory by alcohol. Experimental and Clinical Psychopharmacology, 5, 242– 250. Cannon, D. S., Leeka, J. K., Patterson, E. T., & Baker, T. B. (1990). Principal components analysis of the Inventory of Drinking Situations: Empirical categories of drinking by alcoholics. Addictive Behaviors, 15, 265–269. Cannon, D. S., Rubin, A., Keefe, C. K., Black, J. L., Leeka, J. K., & Phillips, L. A. (1992). Affective correlates of alcohol and cocaine use. Addictive Behaviors, 17, 517–524. Carrigan, G., Samoluk, S.B., & Stewart, S. H. (1998). Examination of the short form of the Inventory of Drinking Situations (IDS-42) in a young adult university student sample. Behaviour Research & Therapy, 36, 789–807. Collins, R. L. (1993). Drinking restraint and risk for alcohol abuse. Experimental and Clinical Psychopharmacology, 1, 44–54. Conrod, P. J., Pihl, R. O., & Vassileva, J. (1998). Differential sensitivity to alcohol reinforcement in groups of men at risk for distinct alcoholic syndromes. Alcoholism: Clinical and Experimental Research, 22(3), 585–597. Conrod, P., Stewart, S. H., & Pihl, R. O. (1997). Validation of a measure of excessive drinking frequency. Substance Use and Misuse, 32, 587–607. Cooper, M. L. (1994). Motivations for alcohol use among adolescents: Development and validation of a four-factor model. Psychological Assessment, 6, 117–128. Cooper, M. L., Russell, M., Skinner, J. B., & Windle, M. (1992). Development and validation of a three-dimensional measure of drinking motives. Psychological Assessment, 4, 123–132. Cunningham, J. A., Sobell, M. B., Sobell, L. S., Gavin, D. R., & Annis, H. (1995). Heavy drinking and negative affect situations in a general population and treatment sample: Alternative explanations. Psychology of Addictive Behaviors, 9, 123–127. Derogatis, L. R. (1979). The SCL-90-R Symptom Checklist Revised. Baltimore: Johns Hopkins University Press. Howell, D. C. (1987). Statistical methods for psychology (2nd ed.). Boston: PWS-Kent. Karp, J. (1993). The interaction of alcohol expectancies, personality, and psychopathology among inpatient alcoholics [Abstract]. Dissertation Abstracts International, 53, 4375-B. Lilienfeld, S. O. (1996). Anxiety sensitivity is not distinct from trait anxiety. In R. M. Rapee (Ed.), Current controversies in the anxiety disorders (pp. 228–244). New York: Guilford. Marlatt, G. A. (1985). Cognitive factors in the relapse process. In G. A. Marlatt & J. R. Gordon (Eds.), Relapse prevention: Maintenance strategies in the treatment of addictive behaviors (pp. 128–200). New York: Guilford. Marlatt, G. A., & Gordon, J. R. (1980). Determinants of relapse: Implications for the maintenance of behavior change. In P. O. Davidson & S. M. Davidson (Eds.), Behavioral medicine: Changing health lifestyles (pp. 410–452). New York: Guilford. McNally, R. J. (1996). Anxiety sensitivity is distinguishable from trait anxiety. In R. M. Rapee (Ed.), Current controversies in the anxiety disorders (pp. 214–227). New York: Guilford. Peterson, R. A., & Reiss, S. (1992). The Anxiety Sensitivity Index manual (2nd ed.). Worthington, OH: International Diagnostic Services. Samoluk, S. B., & Stewart, S. H. (1996). Anxiety sensitivity and anticipation of a self-disclosing interview as determinants of alcohol consumption. Psychology of Addictive Behaviors, 10, 45–54. Samoluk, S. B., Stewart, S. H., Sweet, S. D., & MacDonald, A. B. (1998). Anxiety sensitivity and social affiliation as determinants of alcohol consumption. Manuscript under review for publication. Sobell, M. B., & Sobell, L. C. (1993). Problem drinkers: Guided self-change treatment. New York: Guilford. Stewart, S. H., Conrod, P. J., Gignac, M. L., & Pihl, R. O. (1998). Selective processing biases in anxiety sensitive men and women. Cognition and Emotion, 11, 105–133. Stewart, S. H., Karp, J., Pihl, R. O., & Peterson, R. (1997). Anxiety sensitivity and self-reported reasons for drug use. Journal of Substance Abuse, 9, 223–240.
AS AND SITUATION-SPECIFIC DRINKING
419
Stewart, S. H., Peterson, J. B., & Pihl, R. O. (1995). Anxiety sensitivity and self-reported alcohol consumption rates in university women. Journal of Anxiety Disorders, 9, 283–292. Stewart, S. H., & Pihl, R. O. (1994). Effects of alcohol administration on psychophysiological and subjective-emotional responses to aversive stimulation in anxiety sensitive women. Psychology of Addictive Behaviours, 8, 29–42. Stewart, S. H., Samoluk, S. B., & MacDonald, A. B. (in press). Anxiety sensitivity and substance use and abuse. In S. Taylor (Ed.), Anxiety sensitivity: Theory, research and treatment of the fear of anxiety. Mahwah, NJ: Erlbaum. Stewart, S. H., Taylor, S., & Baker, J. M. (1997). Gender differences in dimensions of anxiety sensitivity. Journal of Anxiety Disorders, 11, 179–200. Stewart, S. H., & Zeitlin, S. B. (1995). Anxiety sensitivity and alcohol use motives. Journal of Anxiety Disorders, 9, 229–240. Stewart, S. H., Zeitlin, S. B., & Samoluk, S. B. (1996). Examination of a three-dimensional drinking motives questionnaire in a young adult university student sample. Behaviour Research & Therapy, 34, 61–71. Tarter, R. E., Alterman, A. I., & Edwards, K. L. (1985). Vulnerability to alcoholism in men: A behavior-genetic perspective. Journal of Studies on Alcohol, 46, 329–356. Taylor, S. (1995). Anxiety sensitivity: Theoretical perspectives and recent findings. Behaviour Research and Therapy, 33, 243–258. Turner, N. E., Annis, H. M., & Sklar, S. M. (1997). Measurement of antecedents to drug and alcohol use: Psychometric properties of the Inventory of Drug-Taking Situations (IDTS). Behaviour Research & Therapy, 35, 465–483. Watt, M., Stewart, S. H., & Cox, B. J. (1998). A retrospective study of the origins of anxiety sensitivity. Behaviour Research & Therapy, 36, 505–525. Williams, E. J. (1959). The comparison of regression variables. Journal of the Royal Statistical Society (Series B), 21, 396–399.
Journal of Anxiety Disorders, Vol. 12, No. 5, pp. 407–419, 1998 Copyright 1998 Elsevier Science Ltd Printed in the USA. All rights reserved 0887-6185/98 $19.00 ⫹ .00
PII S0887-6185(98)00025-5
RESEARCH PAPERS
Anxiety Sensitivity and Situation-Specific Drinking Sarah Barton Samoluk, ph.d., and Sherry H. Stewart, ph.d. Dalhousie University
Abstract—To assess the typical drinking situations of high anxiety sensitive individuals, relative to low anxiety sensitive individuals, a large sample of university student drinkers (N ⫽ 396) completed the 42-item version of the Inventory of Drinking Situations (IDS-42) and the Anxiety Sensitivity Index (ASI). As hypothesized, ASI scores were significantly positively correlated with drinking frequency on the higher order IDS-42 factor of negatively reinforcing drinking situations and not significantly correlated with drinking frequency on the higher order factor of positively reinforcing drinking situations. Also, ASI scores were significantly positively correlated with drinking frequency on the lower order drinking situations factors of Conflict with Others, Unpleasant Emotions, and Physical Discomfort. ASI scores were also unexpectedly significantly correlated with more frequent drinking on the higher order factor of temptation drinking situations and with more frequent drinking on the lower order drinking situations factor of Testing Personal Control. These results have important implications for designing early prevention and intervention programs for high anxiety sensitive university student drinkers by helping them to develop alternatives to drinking in those situations (e.g., negatively reinforcing drinking situations) which place them at greatest risk for alcohol misuse. 1998 Elsevier Science Ltd
This research was part of Dr. Samoluk’s dissertation under the supervision of Dr. Stewart. This study was supported by a Medical Research Council of Canada Operating Grant (MT-12514). Dr. Samoluk was supported by a Social Sciences and Humanities Research Council of Canada doctoral fellowship. The authors thank Jennifer Bird, Brent Conrad, and Carol Thompson for their assistance with questionnaire scoring and data entry, and Monica Ghosal for her assistance in questionnaire administration to the University of Toronto student sample. Requests for reprints should be sent to Dr. Sherry H. Stewart, Department of Psychology, Dalhousie University, Halifax, Nova Scotia B3H 4J1, Canada. E-mail: [email protected] or ssamoluk @is2.dal.ca
407
408
S. B. SAMOLUK AND S. H. STEWART
Anxiety Sensitivity (AS) is a cognitive individual difference variable characterized by a fear of anxiety sensations (fear of fear) and reflects the tendency to catastrophize about the consequences of such sensations (e.g., physical illness, social embarrassment, mental incapacitation; Peterson & Reiss, 1992; Stewart, Conrod, Gignac, & Pihl, 1998; Stewart, Taylor, & Baker, 1997). Fear of fear (AS) is particularly common in individuals who suffer from panic disorder with or without agoraphobia and from post-traumatic stress disorder (Peterson & Reiss, 1992). Moreover, emerging evidence suggests an important relationship between AS and heavier alcohol use (for reviews, see McNally, 1996; Stewart, Samoluk, & MacDonald, in press). For example, higher than average AS levels have been found in clinical samples diagnosed with DSM-III-R (Diagnostic and Statistical Manual of Mental Disorders, third edition, revised; American Psychiatric Association, 1987) alcohol abuse/dependence diagnoses whether or not patients are diagnosed with a co-morbid anxiety disorder (see review by McNally, 1996; Karp, 1993). Moreover, high AS university students report drinking more alcoholic beverages on a weekly basis, and drinking to “excess” (to the point of legal intoxication; Conrod, Stewart, & Pihl, 1997) more frequently, than low AS students (Stewart, Peterson, & Pihl, 1995). The relationship between AS and heavier alcohol consumption is partially clarified by exploration of self-perceived motives for alcohol use. AS was found to be significantly correlated with the frequency of self-reported use of alcohol to “cope” (i.e., to reduce/avoid negative affect such as anxiety or depression) among a university sample (Stewart & Zeitlin, 1995) using the Drinking Motives Questionnaire—a validated measure of drinking motives (Cooper, Russell, Skinner, & Windle, 1992; Stewart, Zeitlin, & Samoluk, 1996). Also high AS students were more likely to report drinking primarily for coping motives, and less likely to report drinking primarily for social motives (i.e., to affiliate with others), compared with low AS students (see also Stewart, Karp, Pihl, & Peterson, 1997). High AS students’ increased use of alcohol to cope (i.e., for negative reinforcement) is likely related to the enhanced sensitivity to alcohol’s emotional reactivity-dampening effects found among high AS young adults compared with low AS young adults in studies using the alcohol challenge paradigm (Baker, MacDonald, Stewart, & Skinner, 1998; Conrod, Pihl, & Vassileva, 1998; Stewart & Pihl, 1994). Different drinking motives have been found to be associated with different situational antecedents and consequences of alcohol use (Cooper, 1994, Cooper et al., 1992). Individuals who drink primarily for coping-related motives are more likely to drink heavily, to drink alone, to drink at home, and to experience alcohol-related problems, compared with those who drink primarily for more normative social motives. In contrast, those who drink primarily to socialize and affiliate with others are more likely to drink at parties and with mixed-sex friends, and are at lower risk for alcohol-related problems than primarily coping-motivated drinkers (Cooper, 1994; Cooper et al., 1992).
AS AND SITUATION-SPECIFIC DRINKING
409
Behaviourally oriented research has focused on developing measures to assess situational antecedents to alcohol consumption. One such measure is the 100-item Inventory of Drinking Situations and its shorter 42-item version (IDS and IDS-42; Annis, Graham, & Davis, 1987). In a large sample of university students, the IDS-42 was found to possess eight lower order factors consistent with Marlatt’s work (e.g., Marlatt & Gordon, 1980) on the situational antecedents to relapse among alcoholics from which the eight IDS subscales were derived (Carrigan, Samoluk, & Stewart, 1998). These eight lower order factors (Conflict with Others, Unpleasant Emotions, Physical Discomfort, Pleasant Times with Others, Social Pressure to Drink, Pleasant Emotions, Testing Personal Control, and Urges and Temptations) were found to load most optimally on three correlated higher order factors (Carrigan et al., 1998). The Conflict with Others, Unpleasant Emotions, and Physical Discomfort subscales loaded on a higher order factor of negatively reinforcing drinking situations; the Pleasant Times with Others, Social Pressure to Drink, and Pleasant Emotions subscales loaded on a higher order factor of positively reinforcing drinking situations; and the Testing Personal Control and Urges and Temptations subscales loaded on a higher order factor of temptation drinking situations (Carrigan et al., 1998). Using a parallel instrument to the IDS, Turner, Annis, and Sklar (1997) similarly identified a hierarchical structure involving eight lower order factors corresponding to the eight IDS subscales, and three higher order factors of negative situations, positive situations, and temptation situations, in a sample of clients receiving addiction treatment.1 Researchers are beginning to examine whether there are relationships between individual difference factors and the situations in which drinking most frequently occurs. For example, in a sample of clients receiving treatment for alcohol dependence, Annis et al. (1987) found frequent drinking in situations involving negative personal states (specifically, relative peaks on the IDS subscales of Physical Discomfort, Unpleasant Emotions, and Conflict with Others) was reported more often by women, and by individuals who tend to drink alone, and have more years of heavy alcohol use. Among a sample of alcoholics and cocaine addicts, a significant positive association was found between levels of negative temperament (i.e., anxiety, depression, boredom, anger) and frequency of drinking on a “negative” factor of the IDS which consisted of items from the Unpleasant Emotions and Conflict with Others subscales (Cannon et al., 1992). Moreover, Turner et al. (1997) found that scores on the Unpleasant Emotions, Conflict with Others, and Physical Discomfort subscales were significantly correlated with measures of negative temperament, including the depression, interpersonal sensitivity, and somatization subscales 1 Turner, Annis, and Sklar (1997) explored situational antecedents to the use of a variety of drugs including alcohol using the Inventory of Drug-Taking Situations (IDTS). The 8 subscales and 50 items of the IDTS were derived directly from the 100-item version of the Inventory of Drinking Situations and extended to the use of drugs in addition to alcohol.
410
S. B. SAMOLUK AND S. H. STEWART
of the Hopkins Symptom Checklist (SCL-90; Derogatis, 1979), in a sample of clients receiving addiction treatment. The present study was designed to explore the relationship between AS (as measured using the Anxiety Sensitivity Index [ASI]; Peterson & Reiss, 1992) and situational antecedents to alcohol consumption using the IDS-42 among a nonclinical sample of university students. Given the significant positive correlation between AS levels and coping-motivated drinking (Stewart & Zeitlin, 1995), we made several hypotheses pertaining to the relationship between ASI scores and the higher and lower order factors of the IDS-42. First, we predicted a significant positive correlation between ASI scores and the frequency of self-reported drinking on the higher order factor of negatively reinforcing drinking situations. We predicted that correlations between ASI scores and the higher order positively reinforcing and temptation drinking situations factors would not be significant. We also predicted that the relationship between ASI scores and the frequency of drinking in negatively reinforcing situations would be stronger than the relationship between ASI scores and the frequency of drinking in positively reinforcing situations or temptation situations. At the lower order level, we predicted significant positive correlations between ASI scores and the lower order IDS-42 factors of Conflict with Others, Unpleasant Emotions, and Physical Discomfort. We predicted that correlations between ASI scores and the lower order factors of Pleasant Times with Others, Social Pressure to Drink, Pleasant Emotions, Testing Personal Control, and Urges and Temptations would not be significant.
METHOD Participants Participants were 473 undergraduate students (338 female, 133 male, 2 gender unspecified), whose average age was 21.72 years (SD ⫽ 4.22). All participants were enrolled in undergraduate psychology courses at Dalhousie University or the University of Toronto, Canada. Materials Inventory of Drinking Situations. The shortened, 42-item version of the Inventory of Drinking Situations (IDS-42; Annis et al., 1987) was used to assess frequency of drinking across specific situations. The IDS-42 contains 8 subscales: The Conflict with Others subscale has 12 items, the Pleasant Times with Others subscale has 6 items, and the remaining subscales have 4 items each. For the present study, the IDS-42 instructions were modified from “frequency of heavy drinking” to “frequency of drinking” in order to more aptly assess the situational antecedents of students’ customary drinking, and to allow completion of the IDS-42 by student drinkers who do not frequently engage in
AS AND SITUATION-SPECIFIC DRINKING
411
“heavy” drinking (cf. Bruce & Pihl, 1997; Carrigan et al., 1998). Participants rated their frequency of drinking over the past year in each specified situation on a scale of 1 (never drank in that situation) to 4 (always drank in that situation). The scale was then recoded to range from 0 to 3 (Annis et al., 1987). Higher scores indicate a relatively higher frequency of drinking in the particular situation. The original IDS-100 and its shortened IDS-42 has been shown to possess good psychometric properties in clinical samples of alcohol abusers. These two versions are essentially equivalent with subscales correlations ranging between .93 and .98 (Annis et al., 1987). The convergent and discriminant validity of the IDS has been demonstrated in alcoholic samples. For example, percentage of time spent drinking alone correlated significantly and positively with negative drinking situations (e.g., Unpleasant Emotions) and correlated negatively with positive drinking situations (e.g., Pleasant Times with Others; Annis et al., 1987). The present version of the IDS-42, using the modified instructional set noted above, possesses good psychometric properties in university students (Carrigan et al., 1998). For example, Carrigan et al. found that the eight lower order IDS-42 factors have excellent concurrent validity with conceptually similar subscales on the Drinking Motives Questionnaire (Cooper et al., 1992). For the present study, the data were analyzed across the three higher order IDS-42 factors of negatively reinforcing drinking situations, positively reinforcing drinking situations, and temptation drinking situations, as well as across all eight lower order factors (equivalent to the eight IDS-42 subscales). Lower order factor scores were computed as weighted means using maximum likelihood factor loadings for the eight lower order factors as weights (Carrigan et al., 1998). Higher order factor scores were likewise computed as weighted means using maximum likelihood factor loadings for the higher order factors as weights applied to the lower order factor scores (Carrigan et al., 1998).2 The negatively reinforcing drinking situations factor includes situational antecedents to alcohol use that pertain to negative affect and unpleasant associations with others which might be assuaged via alcohol consumption (Carrigan et al., 1998): “When other people didn’t seem to like me,” “When I felt that I had let myself down,” and “When my stomach felt like it was tied in knots” are representative items from the Conflict with Others, Unpleasant Emotions, and Physical Discomfort subscales, respectively, which are subsumed under the higher order negatively reinforcing drinking situations factor. The positively reinforcing drinking situations factor includes situational antecedents to alcohol use that pertain to positive affect and pleasant associations with others which might be heightened via alcohol consumption (Carrigan et al., 1998): “When I wanted to celebrate with a friend,” “When I was at 2 Maximum likelihood factor loadings can be found in Carrigan, Samoluk, and Stewart (1998).
412
S. B. SAMOLUK AND S. H. STEWART
a party and other people were drinking,” and “When something good happened and I felt like celebrating” are representative items from the Pleasant Times with Others, Social Pressure to Drink,3 and Pleasant Emotions subscales, respectively, which are subsumed under the higher order positively reinforcing drinking situations factor. The temptation drinking situations factor appears to tap a cognitive preoccupation with drinking (Carrigan et al., 1998): “When I wanted to prove to myself that I could take a few drinks without becoming drunk,” and “When I remembered how good it tasted” are representative items from the Testing Personal Control and Urges and Temptations subscales, respectively, which are subsumed under the higher order temptation drinking situations factor. Anxiety Sensitivity Index. AS levels were measured using the 16-item Anxiety Sensitivity Index (ASI; Peterson & Reiss, 1992). The ASI assesses subjective fears of anxiety sensations (e.g., “It scares me when I feel faint”) and beliefs about why those sensations are harmful (e.g., “When my stomach is upset, I worry that I might be seriously ill”) (Taylor, 1995). Participants rate the degree to which each item describes themselves on a scale of 0 (very little) to 4 (very much). The total ASI score is computed as the sum of all item scores. Higher scores indicate relatively higher levels of fear of anxiety. The ASI has been shown to possess excellent psychometric properties in clinical samples and in nonclinical university student samples (see review by Peterson & Reiss, 1992). In our laboratory, the ASI was found to have an alpha coefficient of .88 in a large sample of 551 university students (Watt, Stewart, & Cox, 1998). This is within the range of reliabilities reported by Peterson and Reiss (1992). Procedure Participants provided informed consent and basic demographic information (age and gender). Only those participants who indicated that they had consumed alcohol within the last year were identified as “drinkers” for the present study and asked to complete the IDS-42. All participants completed the demographic measures. All questionnaires were completed anonymously during class time.
3 Items from this subscale have consistently been found to load on a factor reflecting drinking in pleasant social situations (Annis, Graham, & Davis, 1987; Cannon, Leeka, Patterson, & Baker, 1990; Carrigan, Samoluk, & Stewart, 1998). This indicates drinking for positive reinforcement as opposed to drinking for negative reinforcement implied in social conformity drinking (Cooper, 1994). Thus, the label “Social Cues to Drink” (Carrigan et al., 1998) may more aptly represent the content of the items forming the IDS-42 “Social Pressure to Drink” subscale (e.g., “When I was in a restaurant and the people with me ordered drinks”).
413
AS AND SITUATION-SPECIFIC DRINKING
TABLE 1 Correlations Between ASI Scores and IDS-42 Higher and Lower Order Factor Scores
IDS-42 Higher Order Factors and Subscales
Coefficient Alpha
Correlation Coefficients
.91 .88 .83 .32 .93 .85 .86 .83 .72 .72 .63
.20* .22* .17* .16* .12 .13 .10 .10 .19* .27* .10
Negatively Reinforcing Situationsa Conflict With Others Unpleasant Emotions Physical Discomfort Positively Reinforcing Situations Pleasant Times With Others Social Pressure to Drink Pleasant Emotions Temptation Situations Testing Personal Control Urges and Temptations
Disattenuated Correlations With ASI Scores .22*a .25* .19* .29* .13ab .14 .11 .11 .23b .34* .14
Notes. The three higher order factors indicated in bold type were found among university student drinkers using confirmatory factor analysis (Carrigan et al., 1998). Correlation coefficients with the same subscripts differ significantly (for subscript “a,” p ⬍ .05, one-tailed; for subscript “b,” p ⬍ .01, two-tailed). a Probability levels for this IDS-42 higher order factor and its lower order subscales are one-tailed in keeping with the hypotheses. All remaining probability levels are two-tailed. * p ⱕ .001.
RESULTS The mean ASI score for the entire sample was 20.10 (SD ⫽ 9.22) which is consistent with previously established nonclinical norms on this measure (Peterson & Reiss, 1992). Of the total sample, 83.7% (n ⫽ 396; 283 female, 111 male, 2 gender unspecified) reported alcohol consumption during the past year. The drinkers (n ⫽ 395, 1 age unspecified) were slightly older than the nondrinkers (n ⫽ 77): mean age (and SD) ⫽ 21.89 (4.27) versus 20.81 (3.82) years, respectively, F(1, 470) ⫽ 4.33, p ⬍ .05. However, the drinkers (n ⫽ 394, 2 gender unspecified) did not differ significantly from the nondrinkers (n ⫽ 77) in gender composition (71.4% vs. 71.8% female, respectively), 2(1) ⫽ .005, ns. All subsequent analyses were conducted using data from the subsample of 396 self-reported drinkers. Reliability of the IDS-42 Subscales The internal consistency of each of the eight IDS-42 subscales separately and combined as higher order factors was calculated using Cronbach’s coefficient alpha (as presented in Table 1) in order to determine if reliabilities differ markedly across subscales and factors. The majority of subscales had alphas of .83 or better, indicating substantial internal consistency. The Testing Personal
414
S. B. SAMOLUK AND S. H. STEWART
Control and Urges and Temptations subscales had moderate alphas of .72 and .63, respectively. Physical Discomfort was the least reliable subscale with an alpha of .32.4 The alphas for the higher order factors varied from .72 to .93, indicating moderate to substantial internal consistency (see Table 1). Correlational Analyses All correlational analyses between ASI scores and higher order factor scores, and between ASI scores and lower order factor scores, were calculated using a Bonferroni-adjusted alpha of .005 (.05/11 comparisons) for determining statistical significance.5 Given the variability in reliabilities across subscales and factors (see Table 1), correlations are presented with and without correction for attenuation in Table 1. As predicted, ASI scores were significantly positively correlated with negatively reinforcing drinking situations factor scores (p ⬍ .001, one-tailed). Also, as predicted, ASI scores were not significantly associated with positively reinforcing drinking situations factor scores. Unexpectedly, however, ASI scores were significantly positively correlated with temptation drinking situations factor scores (p ⬍ .001, two-tailed). The pattern of results remained the same prior to correction for attenuation (see Table 1). Correlations were also calculated between ASI scores and the eight lower order factor scores (see Table 1). ASI scores showed significant positive associations with the frequency of drinking in situations involving Conflict with Others, Unpleasant Emotions, and Physical Discomfort (all ps ⱕ .001, onetailed) as predicted. Also as predicted, ASI scores were not significantly associated with the frequency of drinking in situations involving Pleasant Times with Others, Social Pressure to Drink, or Pleasant Emotions (all ps ⬎ .005, two-tailed). ASI scores were not significantly correlated with the frequency of drinking in situations involving Urges and Temptations as hypothesized; but unexpectedly ASI scores were significantly positively correlated with the frequency of drinking in situations involving Testing Personal Control (p ⬍ .001, two-tailed). The pattern of results remained the same prior to correction for attenuation (see Table 1).
4
Carrigan, Samoluk, and Stewart (1998) report that the relatively low internal consistency of the Physical Discomfort subscale should be expected in a nonclinical university student sample of drinkers who are less likely than alcoholics to use alcohol to alleviate physical discomfort. Given the brevity of the Physical Discomfort subscale (four items) and its heterogenous content, a larger number of more homogeneous items might improve the reliability of the Physical Discomfort IDS-42 Subscale for university students. 5 The pattern of significant results between ASI scores and the IDS-42 scores remained the same whether or not factor scores (weighted means) or subscale scores (unweighted means) were used.
AS AND SITUATION-SPECIFIC DRINKING
415
Tests of Difference Between Dependent Correlations The significance of the correlational differences among the negatively reinforcing, positively reinforcing, and temptation drinking situations IDS-42 higher order factors was calculated using a t equation, for nonindependent correlations, with the disattenuated coefficients (see Table 1).6 The t equation revealed that, as predicted, negatively reinforcing drinking situations factor scores were more highly correlated with ASI scores than positively reinforcing drinking situations factor scores, t(393) ⫽ 1.66, p ⬍ .05, one-tailed. However, negatively reinforcing drinking situations factor scores were not more highly correlated with ASI scores than temptation drinking situation factor scores, t(393) ⫽ ⫺0.24, ns. Temptation drinking situations factor scores were more highly correlated with ASI scores than positively reinforcing drinking situations factor scores, t(393) ⫽ 2.75, p ⬍ .01, two-tailed. In sum, these findings suggest that ASI scores are more highly predictive of drinking in negatively reinforcing and temptation situations than of drinking in positively reinforcing situations.
DISCUSSION As hypothesized, ASI scores were positively and significantly correlated with the self-reported frequency of drinking in negatively reinforcing situations. Also as predicted, ASI scores were not significantly associated with the reported frequency of drinking in positively reinforcing situations. Analysis of each lower order factor score confirmed this relationship between higher ASI scores and situation-specific drinking in negatively reinforcing situations but not in positively reinforcing situations. Namely, while ASI scores were significantly and positively associated with drinking in situations involving Conflict with Others, Unpleasant Emotions, and Physical Discomfort, ASI scores were not significantly correlated with drinking in situations involving Pleasant Times with Others, Social Pressure to Drink, or Pleasant Emotions. In the general population, as well as in problem drinkers, a tendency to drink in negative affect states is correlated with increasing levels of alcohol dependence, whereas drinking in positive affect states is associated with less risky social drinking (Cunningham, Sobell, Sobell, Gavin, & Annis, 1995). Similar to previous findings which link general negative temperament to drinking in IDS negative situations among problem drinkers (Cannon et al., 1992), the results of the present correlational analyses suggest that nonclinical drinkers with higher AS levels appear more motivated than others to drink in aversive situations. This may place them at risk for alcohol problems. These
6
The t equation was devised by Williams (1959). It consists of a ratio distributed as t with N ⫺ 3 df, and takes into account the degree to which two dependent tests are correlated. A simple explanation of the use of this formula can be found in Howell (1987).
416
S. B. SAMOLUK AND S. H. STEWART
results are consistent with other studies which found that ASI scores are significantly and positively associated with self-reported frequency of drinking to cope (cf. Stewart, Karp, et al., 1997; Stewart & Zeitlin, 1995) and with drinking in response to physical discomfort in a laboratory setting (Samoluk & Stewart, 1996). The present results are also consistent with the finding that individuals with higher AS levels are more sensitive to alcohol’s emotional reactivitydampening properties (cf. Baker et al., 1998; Conrod et al., 1998; Stewart & Pihl, 1994) than individuals with lower AS levels. ASI scores were only significantly correlated with situations that do not necessitate the presence of others (Unpleasant Emotions, Physical Discomfort, Conflict with Others, and Testing Personal Control).7 This suggests that high AS persons are likely to be relatively solitary drinkers. Solitary drinking is more highly associated with alcohol problems than is social drinking (Cooper et al., 1992). The relatively solitary drinking style of high AS young adults has recently been supported in a lab-based analogue alcohol consumption study in which high AS individuals were found to drink more than low AS individuals in a solitary, but not a social, drinking context (Samoluk, Stewart, Sweet, & MacDonald, 1998). Contrary to hypothesis, ASI scores were positively and significantly correlated with the temptation drinking situations factor scores. Indeed, the strength of the association with ASI scores was not significantly different for the negatively reinforcing drinking situations factor and the temptation drinking situations factor. Correlational tests of lower order factor scores further revealed that ASI scores were correlated with the Testing Personal Control factor scores but not the Urges and Temptations factor scores. These results suggest that individuals high in AS drink more frequently for both negative reinforcement (i.e., to attenuate negative affect and physical discomfort) and for the temptation to rationalize continued alcohol use. The cognitive preoccupation with drinking reflected in the Testing Personal Control IDS-42 factor items appears to be conceptually analogous to the concept of restrained drinking (Collins, 1993). It has been theorized that failed attempts to regulate alcohol intake among “restrained drinkers” may actually promote binge drinking when restrained drinkers blame themselves for their failure to regulate their alcohol consumption (Collins, 1993). This self-blame may lead to further excessive drinking in order to assuage the resultant low mood of restrained drinkers in a continuous cycle of restraint, violation, negative affect, and excessive drinking (Marlatt, 1985; Marlatt & Gordon, 1980). Given their apparently increased cognitive preoccupation with drinking, high AS individuals would likely score relatively higher than others on restrained drinking measures. (See Collins, 1993, for a review of measures of restrained drinking.) 7 Situations involving Conflict with Others are included in this list because solitary drinking can occur following conflict as a way to cope with the resultant negative affect.
AS AND SITUATION-SPECIFIC DRINKING
417
The present study did not include the assessment of higher order negative temperament constructs related to both anxiety sensitivity and alcoholism, such as measures of neuroticism, trait anxiety or dysphoria (see reviews by Lilienfeld, 1996; Peterson & Reiss, 1992; Tarter, Alterman, & Edwards, 1985). Hence, this study cannot identify the degree to which the results are due specifically to AS or to other higher order personality constructs. Given the specific fear of anxiety-related sensations among high AS individuals (Peterson & Reiss, 1992), one might speculate that AS should be most strongly associated with drinking frequency in negatively reinforcing situations involving anxious emotions and physical discomfort. In contrast, the relationship between negative temperament, in general, and drinking frequency should be less situationally specific, including significant relationships between negative temperament and drinking frequency in a variety of negatively reinforcing situations (e.g., anxiety, dysphoria, anger, etc.). Unfortunately, the structure of the IDS-42 does not permit testing of this speculation (e.g., the Unpleasant Emotions subscale encompasses anxiety, depression, and anger situations). However, in a study on drinking motives, Stewart, Karp, et al. (1997) obtained equally strong and significant relationships between AS and anxiety-related reasons for alcohol use as they did between AS and depression-related reasons for alcohol use, thus failing to support a specific association between AS and anxietymotivated drinking. Future research could clarify the degree to which the present results are due specifically to AS and/or to other higher order negative temperament constructs by including the ASI and measures of such additional constructs in a multiple regression design. The information provided by the IDS has proven to be useful in relapse prevention alcohol abuse treatment programs (Marlatt, 1985; Marlatt & Gordon, 1980; Sobell & Sobell, 1993). In a similar manner, the observed situational antecedents of more frequent drinking in relatively solitary negatively reinforcing and temptation drinking situations among high AS university students, relative to low AS university students, could become the focus of prevention and treatment. Specifically, high AS young adults could be trained to develop alternative coping strategies to alcohol consumption for dealing with these particular high-risk situations.
REFERENCES American Psychiatric Association. (1987). Diagnostic and statistical manual of mental disorders (3rd ed., rev.). Washington, DC: Author. Annis, H. M., Graham, J. M., & Davis, C. S. (1987). Inventory of Drinking Situations (IDS) user’s guide. Toronto: Alcoholism and Drug Addiction Research Foundation. Baker, J. M., MacDonald, A. B., Stewart, S. H., & Skinner, M. (1998). The effects of alcohol on response to voluntary hyperventilation among individuals high and low in anxiety sensitivity. Manuscript submitted for publication.
418
S. B. SAMOLUK AND S. H. STEWART
Bruce, K. R., & Pihl, R. O. (1997). Forget “Drinking to Forget”: Enhanced consolidation of emotionally charged memory by alcohol. Experimental and Clinical Psychopharmacology, 5, 242– 250. Cannon, D. S., Leeka, J. K., Patterson, E. T., & Baker, T. B. (1990). Principal components analysis of the Inventory of Drinking Situations: Empirical categories of drinking by alcoholics. Addictive Behaviors, 15, 265–269. Cannon, D. S., Rubin, A., Keefe, C. K., Black, J. L., Leeka, J. K., & Phillips, L. A. (1992). Affective correlates of alcohol and cocaine use. Addictive Behaviors, 17, 517–524. Carrigan, G., Samoluk, S.B., & Stewart, S. H. (1998). Examination of the short form of the Inventory of Drinking Situations (IDS-42) in a young adult university student sample. Behaviour Research & Therapy, 36, 789–807. Collins, R. L. (1993). Drinking restraint and risk for alcohol abuse. Experimental and Clinical Psychopharmacology, 1, 44–54. Conrod, P. J., Pihl, R. O., & Vassileva, J. (1998). Differential sensitivity to alcohol reinforcement in groups of men at risk for distinct alcoholic syndromes. Alcoholism: Clinical and Experimental Research, 22(3), 585–597. Conrod, P., Stewart, S. H., & Pihl, R. O. (1997). Validation of a measure of excessive drinking frequency. Substance Use and Misuse, 32, 587–607. Cooper, M. L. (1994). Motivations for alcohol use among adolescents: Development and validation of a four-factor model. Psychological Assessment, 6, 117–128. Cooper, M. L., Russell, M., Skinner, J. B., & Windle, M. (1992). Development and validation of a three-dimensional measure of drinking motives. Psychological Assessment, 4, 123–132. Cunningham, J. A., Sobell, M. B., Sobell, L. S., Gavin, D. R., & Annis, H. (1995). Heavy drinking and negative affect situations in a general population and treatment sample: Alternative explanations. Psychology of Addictive Behaviors, 9, 123–127. Derogatis, L. R. (1979). The SCL-90-R Symptom Checklist Revised. Baltimore: Johns Hopkins University Press. Howell, D. C. (1987). Statistical methods for psychology (2nd ed.). Boston: PWS-Kent. Karp, J. (1993). The interaction of alcohol expectancies, personality, and psychopathology among inpatient alcoholics [Abstract]. Dissertation Abstracts International, 53, 4375-B. Lilienfeld, S. O. (1996). Anxiety sensitivity is not distinct from trait anxiety. In R. M. Rapee (Ed.), Current controversies in the anxiety disorders (pp. 228–244). New York: Guilford. Marlatt, G. A. (1985). Cognitive factors in the relapse process. In G. A. Marlatt & J. R. Gordon (Eds.), Relapse prevention: Maintenance strategies in the treatment of addictive behaviors (pp. 128–200). New York: Guilford. Marlatt, G. A., & Gordon, J. R. (1980). Determinants of relapse: Implications for the maintenance of behavior change. In P. O. Davidson & S. M. Davidson (Eds.), Behavioral medicine: Changing health lifestyles (pp. 410–452). New York: Guilford. McNally, R. J. (1996). Anxiety sensitivity is distinguishable from trait anxiety. In R. M. Rapee (Ed.), Current controversies in the anxiety disorders (pp. 214–227). New York: Guilford. Peterson, R. A., & Reiss, S. (1992). The Anxiety Sensitivity Index manual (2nd ed.). Worthington, OH: International Diagnostic Services. Samoluk, S. B., & Stewart, S. H. (1996). Anxiety sensitivity and anticipation of a self-disclosing interview as determinants of alcohol consumption. Psychology of Addictive Behaviors, 10, 45–54. Samoluk, S. B., Stewart, S. H., Sweet, S. D., & MacDonald, A. B. (1998). Anxiety sensitivity and social affiliation as determinants of alcohol consumption. Manuscript under review for publication. Sobell, M. B., & Sobell, L. C. (1993). Problem drinkers: Guided self-change treatment. New York: Guilford. Stewart, S. H., Conrod, P. J., Gignac, M. L., & Pihl, R. O. (1998). Selective processing biases in anxiety sensitive men and women. Cognition and Emotion, 11, 105–133. Stewart, S. H., Karp, J., Pihl, R. O., & Peterson, R. (1997). Anxiety sensitivity and self-reported reasons for drug use. Journal of Substance Abuse, 9, 223–240.
AS AND SITUATION-SPECIFIC DRINKING
419
Stewart, S. H., Peterson, J. B., & Pihl, R. O. (1995). Anxiety sensitivity and self-reported alcohol consumption rates in university women. Journal of Anxiety Disorders, 9, 283–292. Stewart, S. H., & Pihl, R. O. (1994). Effects of alcohol administration on psychophysiological and subjective-emotional responses to aversive stimulation in anxiety sensitive women. Psychology of Addictive Behaviours, 8, 29–42. Stewart, S. H., Samoluk, S. B., & MacDonald, A. B. (in press). Anxiety sensitivity and substance use and abuse. In S. Taylor (Ed.), Anxiety sensitivity: Theory, research and treatment of the fear of anxiety. Mahwah, NJ: Erlbaum. Stewart, S. H., Taylor, S., & Baker, J. M. (1997). Gender differences in dimensions of anxiety sensitivity. Journal of Anxiety Disorders, 11, 179–200. Stewart, S. H., & Zeitlin, S. B. (1995). Anxiety sensitivity and alcohol use motives. Journal of Anxiety Disorders, 9, 229–240. Stewart, S. H., Zeitlin, S. B., & Samoluk, S. B. (1996). Examination of a three-dimensional drinking motives questionnaire in a young adult university student sample. Behaviour Research & Therapy, 34, 61–71. Tarter, R. E., Alterman, A. I., & Edwards, K. L. (1985). Vulnerability to alcoholism in men: A behavior-genetic perspective. Journal of Studies on Alcohol, 46, 329–356. Taylor, S. (1995). Anxiety sensitivity: Theoretical perspectives and recent findings. Behaviour Research and Therapy, 33, 243–258. Turner, N. E., Annis, H. M., & Sklar, S. M. (1997). Measurement of antecedents to drug and alcohol use: Psychometric properties of the Inventory of Drug-Taking Situations (IDTS). Behaviour Research & Therapy, 35, 465–483. Watt, M., Stewart, S. H., & Cox, B. J. (1998). A retrospective study of the origins of anxiety sensitivity. Behaviour Research & Therapy, 36, 505–525. Williams, E. J. (1959). The comparison of regression variables. Journal of the Royal Statistical Society (Series B), 21, 396–399.