State Attachment, Anxiety Sensitivity, and Drinking Motives

PAGE 1

State Attachment, Anxiety Sensitivity, and Drinking Motives: The Effect of Priming State Attachment on Anxiety Sensitivity and Drinking Motives BY Cory Kennedy Costello A Thesis Submitted to the Division of Psychology New College of Florida in partial fulfillment of the requirements for the degree Bachelor of Arts Under the sponsorship of Dr. Steve Graham Sarasota, Florida March, 2012

PAGE 2

ii Acknowledgements Thanks to my thesis sponsor, Dr. Steve Graham, whose encouragement, expertise, and direction made this thesis possibl e. Thanks to Dr. Michelle Barton for directing me in th e thesis process while Dr. Steven Graham was on research leave, and for being a part of my committee Thanks to Dr. Maribeth Clark for participating as my committee member. Thanks to Dr. Duff Cooper for all of the help running and interpreting the statisti cal analyses. Thanks to my family and friends for the support and good times.

PAGE 3

iii Table of Contents ACKNOWLEDGEMENTS ii TABLE OF CONTENTS iii LIST OF TABLES iv A BSTRACT v INTRODUCTION 1 C OMORBID ANXIETY AND ALCOHOL USE DISORDERS 1 A NXIETY SENSITIVITY 7 D RINKING MOTIVES 15 ATTACHMENT 19 ATTACHMENT AND ANXIETY SENSITIVITY 19 ATTACHMENT AND DRINKING MOTIVES 21 STATE ATTACHMENT 26 THE CURRENT STUDY 31 METHOD 33 RESULT S 36 DISCUSSION 39 REFERENCES 49 TABLES 55 APPENDICES 58

PAGE 4

iv List of Tables Table 1: Descriptive Statistics for All Dependent Measures 55 Table 2: Descriptive Statistics for All Dependent Measures by Group 56 Table 3: MANOVA with Priming Group as Independent Variable 57

PAGE 5

v State Attachment, Anxiety Sensitivity, and Drinking Motives: The Effect of Priming State Attachment on Anxiety Sensitivity and Drinking Motives Cory Costello New College of Florida, 2012 Abstract Comorbid anxiety and alcohol use disorders are a serious problem with a number of serious negative consequences, but are not totally well understood. Anxiety sensitivity (AS) is a documented risk factor for Alcohol Use Disorders (Schmidt, Buckner, & Keough 2007) and an incremental predictor of later anxiety symptoms (Schmidt, Mitchell, & Richey, 2008). Previous findings suggest that coping drinking motives are associated with anxiety disorders and alcohol use disorders ( Buckner & Heimberg 2010) There is little understood of how these risk factors (AS and drinking motives) develop. Both of these risk factors have been shown empirically to relate to attachment insecurity, which has led scholars to suggest a possible causal relationship (Weems, Berman, Silve rman, & Rodriguez, 2002; Molnar, Sadava, DeCourville, & Perrier, 2010). The present research studied the relationship between attachment schemas AS, and coping drinking motives experimentally. Participants were randomly assigned to attachment prime condit ions: secure, insecure, and control. They then completed the A nxiety S ensitivity I ndex and the Modified D rinking M otive Q uestionnaire R evised The primes did not produce differences in anxiety sensitivity or drinking motives across condition These findin gs are discussed in light of the relevant research.

PAGE 6

1 STATE ATTACHMENT, ANXIETY SENSITIVITY, AND DRINKING MOTIVES : THE EFFECT OF PRIMING STATE ATTACHMENT ON ANXIETY SENSITIVITY AND DRINKING MOTIVES In Western Culture alcohol consumption is a common, socially acceptable, and often joyous activity. Many people consume alcohol in occasions including business meetings, weddings and funerals, and last minute get togethers between coworkers and friends; in other words, alcohol consumption is common in many settings. All in all, alcohol consumption is generally harmless, but in certain cases, it becomes a maladaptive and destructive behavior that ca n A substantial number of people who abuse alcohol also suffer from extraordinary anxiety (Petrakis et al., 2002). Hypothetically, it is understandable that these would be related, and in fact, it seems quite logical that certain individuals experiencing a tremendous amount of anxiety would be more likely to abuse alcohol. However, the connection may not be that simple, and so more research is required to understand how alcohol abuse relates to anxiety, how anxiety and alco hol use disorders relate to one another, and what underlying factors and motivations are facilitating these relations. Comorbid Anxiety and Alcohol Use Disorders Many researchers have investigated the relationship between alcohol and anxiety, and have foun d that alcohol use and anxiety are related in a number of ways (Buckner & Heimberg, 2010; DeHaas, Calamari, & Bair 2002; Smith & Book 2010). Anxiety d isorders are defined broadly as a condition of experiencing abnormally high amounts of anxiety that resu mental health (APA, 2000 ). Alcohol use disorders come in two varieties: alcohol abuse

PAGE 7

2 disorder refers to individuals that use alcohol in unhealthy ways, whereas alcohol dependence dis order refers to those who are also phys ically dependent on alcohol (APA, 2000 ). Research has found that comorbid or co occurring, anxiety and alcohol use disorders are quite common, with a 12 month prevalence rate of anxiety disorder diagnosis of about 36 .9% in an alcoholic population (Petrakis et al., 2002). Receiving this diagnosis means a difficult journey to recovery: there are higher rates of alcohol dependence, higher rates of relap se, and higher rates of suicide in comorbid patients than in patients diagnosed with only an alcohol use disorder. Smith and Book (2010) investigated the occurrence of comorbid G eneral ized Anxiety Disorder (GAD) and Alcohol Use D isorders (AUD). Participants included 110 individuals who were involved in three separate 4 week intensive outpatient substance abuse treatment programs The researchers attempted to control for acute withdrawal symptoms by limiting the sample to patients who had been detoxified for between 14 and 28 days; after limiting the sample to patients diagnosed with AUD the sample was reduced to 39 participants, and all of these participants were diagnosed in structured clinical interviews. The participants were administered self report measures; the Penn State Worry Que stionnaire (PSWQ), the Anxiety Sensitivity Index (ASI), the Liebowits Social Anxiety Scale (LSAS), the Beck Depression Inventory (BDI), the Worry R eduction Alcohol Expectancy S cale, and a set of questions designed by the researchers to evaluate the impact of worry on treatment outcomes. The results suggested that 46.2% of the participants who were diagnosed with AUD were also diagnosed with current Generalized Anxiety Disorder ( GAD ), and 88.9% of dually diagnosed participants were alcohol dependent (Smith & Book, 2010). The

PAGE 8

3 comorbid patients were at higher risk for panic disorder with agoraphobia, sedative hypnotic dependency, and opioid dependence, as well as higher levels of worry. Additionally, 55.6 % had a history of attempted suicide compared to 23.8% of patients with AUD alone. Dually diagnosed patients had significantly higher levels of worry, and had significantly higher levels of treatment interference as a result of worry. This study had compelling findings that bring in to focus the difficulties f aced by individuals diagnosed with comorbid anxiety and alcohol use disorders. However the nature of the relation between anxiety and alcohol use disorders remains unclear. Swendsen, Conway, Degenhardt, Glantz, Jin, Merikangas, Sampson, and Kessler (2010) attempted to help clarify the nature of this relationship by focusing on mental disorders as a risk factor for substance use, and particularly on mental disorders on the transition from use to abuse and dependence. The study used the data taken during the 1990 1992 National Comorbidity Survey (NCS) and the 2001 2003 NCS follow up, which originally consisted of 5001 respondents. They focused on anxiety disorders, mood disorders, disruptive behavior, and other substance use disorders. The NCS assessed diagno ses of respondents, and a probability subsample was reappraised by clinicians following the structured clinical interview for DSM II R; there was generally agreement between reappraisal and NCS data. They found that behavioral disorder s and pre existing su bstance abuse conditions emerged as the strongest predictors of a later diagnosis of nicotine, alcohol, and illicit substance abuse Anxiety disorders in general were also found to be good predictors of these diagnoses ; PTSD, specific and social phobia, pa nic disorder, and separation anxiety were the most associated with receiving a SUD diagnosis

PAGE 9

4 The authors concluded that mental disorders are a legitimate risk factor for developing substance use disorders; additionally, they more strongly predict the tra nsition from use to abuse and dependence, as opposed to the onset of use. This study suggests that individuals diagnosed with anxiety disorders (and other disorders) are not necessarily more likely to use alcohol, but that they are more likely to abuse alc ohol if they already consume it Taken with findings indicating a higher rate of alcohol dependence among those also diagnosed with an anxiety disorder (Smith & Book, 2010) it appears that individuals diagnosed with anxiety disorders may be at a greater risk for de veloping alcohol use disorders However, the possibility remains that instead alcohol abuse puts one at risk for anxiety disorders, or that the relationship is the result of a sha red variance with other factors, such as anxiety sensitivity. DeHa as, Calamari, and Bair (2002) studied whether substance abuse intended for distress reduction could be explained by anxiety sensitivity (AS) in individuals with anxiety disorders and substance use problems. Previous literature has found an association betw een anxiety disorders and substance use problems; there is also evidence that AS is correlated with anxiety disorders and substance use problems. They also tested the efficacy of AS as a predictor of substance use not explained by trait anxiety or psycholo gical diagnosis. The sample consisted of 54 VA patients with no psychotic disorders; each was interviewed using the Scheduled clinical interview for DSM IV non patient version (SCID NP), and then filled out self report measures. The self report measures in cluded the Bec ks Depression Index (BDI), State Trait Anxiety Inventory (STAI), Anxiety Sensitivity Index (ASI), and the Inventory of Drug Taking S ituations

PAGE 10

5 (IDTS); the IDTS consisted of 50 questions evaluating situations in which the individual usually use s substances. The results suggested that AS adequately predicts the frequ ency of substance use in a dually diagnosed sample in many of the situations measured by the IDTS (DeHaas et al., 2002). Anxiety sensitivity predicted substance use even after contro lling for general anxiety (trait and state level) and after controlling for disorder diagnosis suggesting that AS puts one at greater risk for an AUD than any psychological disorder ; these results applied to depressant use and not stimulant use. Anxiety s ensitivity was seen to be correlated with substance use in the negative situations subscale of the IDTS, and substance use in the context of pleasant emotions, and pleasant times with others. Surprisingly there was no evidence of a greater relation betwee n anxiety sensitiv ity and substance use in dually diagnosed patients with anxiety diso rders as the primary A major limitation of this study is that they used patients who had recently detoxified, and this could affect the self report measures as was noted by Smith and Book (2010) The studies covered thus far have not considered an aspect of behavior that is instrumental in the change and maintenance of b ehavioral patterns: motivation. Buckner and Hei mberg (2010) studied the relationship between social anxiety, alcohol con sumption, and coping as a motivating factor for drinking. Alcohol use problems often co occur with social anxiety disorders, and this relationship has often been attributed to drinkin g as a coping mechanism. This research examined this relationship in the frame of alcohol. The final sample consisted of 120 individuals (80.8% women), who all fille d out

PAGE 11

6 a number of computerized versions of self report measures using a secure on line data collection website. The self report measures included the social interaction anxiety scale (SIAS), the Liebowitz social anxiety scale (LSAS), the daily drinking que stionnaire revised (DDQ R), the Rutgers alcohol problems index (RAPI), and the drinking to cope with social anxiety scale (DCSAS). The DCSAS was created for this study and has not been previously tested, but all of the other measures have displayed high re liability (Buckner & Heimberg, 2010). The study found that individuals with higher amounts of social anxiety reported greater rates of drinking to cope with social situations, and also avoiding alcohol free social situations (Buckner & Heimberg, 2010). Th ese two behaviors mediated the relationship between social anxiety and alcohol related problems. These results suggest that the higher risk of alcohol related problems that is associated with high levels of social anxiety is facilitated by viewing drinking related coping as a viable coping behavior The authors suggest that this maladaptive coping method may prevent the development of more adaptive coping methods, because individuals that use alcohol as a coping mechanism may attribute any social success to alcohol; this incorrect belief may reinforce the drinking behavior. Drawing from the literature, it appears that drinking as a coping mechanism and anxiety sensitivity both play a crucial role in the association between anxiety and alcohol use disorders (Buckner & Heimberg, 2010; DeHaas et al., 2002). It therefore seems important to understand what exactly AS and coping drinking motives are: with what factors are they associated and how do they develop. Put simply, anxiety sensitivity is the fear of anxio us arousal caused by exaggerated beliefs of the conseq uences of

PAGE 12

7 experiencing anxious arousal (Reiss et al., 198 6). Coping drinking motives (i.e., alc ohol use as a coping mechanism) are the tendency for one to consume alcohol as a way to cope with negative affect such as anxiety (Grant et al., 2007). There is a growing body of literature focusing on anxiety sensitivity and coping drinking motives due to their apparent importance in an xiety disorders, alcohol use disorders, and comorbid anxiety and alcohol use disorders. Anxiety Sensitivity Although anxiety sensitivity is a relatively new concept, psychologists have long discussed the idea of a fear of anxiety. In a review articl e, Ste ven Reiss (1987) discussed the fear of anxiety, how it has been interpreted through different perspectives, and the need for more empirical data on this phenomenon. He discussed the psychoanalytic perspective, existential perspective, conditioning or behav ioral perspective, and finally the cognitive perspective; the cognitive perspective is the conceptualization of importance to this study. Cognitive perspectives, including the one first formulated by the author, led to a construct termed anxiety sensitivit y, which consists of irrational beliefs about the consequences of anxious arousal (e.g., high AS corresponds to believing anxiety could result in risk to physical, social, or mental health, while low AS corresponds to believing anxiety is a passing discomf ort). The review ends with a call for more empirical research on the case of AS, and on how AS affects mental processes; however, the state of the literature has changed a great deal since this review, and AS has been studied by many researchers in a varie ty of contexts.

PAGE 13

8 Reiss, Peterson, Gursky, and McNally (1986) conducted a series of studies on a cognitive construct anxiety sensiti vity (AS), its measurement the Anxiety Sensitivity I ndex (ASI) possible clinical applications of AS, and its relationship wit h fear as compared to anxiety more broadly. They conducted a series of three studies, which represents some of the earliest work on the cognitive perspective of the fear of anxiety or anxiety sensitivity (Reiss et al., 1986). T he authors tested the relati onship between AS and psychological disorders; they grouped subjects into patients with agoraphobia, patients with other anxiety disorders, and non patient subjects (Reiss et al., 1986). They hypothesized that AS would be highest in the patients with agora phobia, slightly lower but still high in the patients with other anxiety disorder, and sm allest in the non patient groups The results confirmed the hypothesis; ASI scores were highest in patients with agoraphobia ( M Female = 38.2), followed by patients wit h anxiety disorders ( M Male =25.8; M Female = 23.9), and lowest for college students ( M Male = 15.4; M Female = 20.4). In the final study, the authors measured AS and anxiety more broadly, and compared the two in terms of their relationship to fears. Regression analyses found that AS accounted for 17.3% of the variance in fearfulness above and beyond the variance accounted for by general anxiety ; this supports the claim that anxiety sensitivity is distinct from general anxiety, and that it is a better predictor of fearfulness than general anxiety (Reiss et al., 1986). Zinbarg, Barlow, and Brown (1997) investi gated the possibility that the Anxiety Sensitivity I ndex (ASI) could be calculated as a hierarchical structural mo del; i.e., that there are lower o rder factors in the ASI that combine to form a single higher order factor. In research on anxiety sensitivity, there has been a debate as to whether anxiety

PAGE 14

9 sensitivity is a singular construct or an umbrella term for multiple smaller constructs. Zinbarg an d colleagues set out to empirically test these theories using factor analysis. They tested 432 outpatient participants that had been diagnosed with an anxiety disorder, and 32 non patient participants using the 16 item anxiety sensitivity index (ASI) (Zinb arg et al., 1997). They found support for three lower order factors; these lower order factors were social concerns, physical concerns, and mental incapacity concerns (Zinbarg et al., 1997). Additionally, they found that these three factors were highly correlated, which supports there being a higher order factor (i.e., a total AS level). The findings have important clinical applications; first, they provide validity for measuri ng a total AS level and specific facets of AS; this could assist with diagnosis and treatment of anxiety disorders, result of a diagnosis. Zinbarg and colleagues pro vide convincing evidence that anxiety sensitivity can be conceptualized as a single yet multi faceted construct which lends structure to research concerning AS (Zinbarg et al., 1997). Berman, Wheaton, McGrath, and Abramowitz (2010) set out to understand how AS and experiential avoidance (EA) relate to one another, and how they individually relate to the onset of clinically significant anxiety symptoms. The sample included 42 adults (22 men and 20 women), all of wh om received a primary diagnosis of an anxi ety disorder; 12 diagnosed with obsessive compulsive disorder (OCD), 6 diagnosed with panic disorder with or without agoraphobia, 10 diagnosed with social phobia, 1 diagnosed with specific phobia, 5 diagnosed with general anxiety disorder (GAD), and 8 cove ring the remaining forms of anxiety disorders. The participants filled out a packet of self

PAGE 15

10 report measures, and were also interviewed by a trained assessor, and each was reassessed by the director of the clinic. The self report measures included the Accep tance and Action Questionnaire II, which assesses experiential avoidanc e. Also included were the Beck Anxiety Inventory (BAI), the Anxiety Sensitivity I ndex 3 (ASI 3) and the Beck Depression I nventory (BDI 2); all of these self report measures have displa yed excellent reliability in previous studies. The authors found that EA and AS were significantly positively correlated, and that both were correlated with anxiety symptoms (Berman et al., 2010). They also found that the AS was correlated with anxiety di sorders even when controlling for EA, while EA was no longer significant when controlling for AS. This led the authors to conclude that EA and anxiety are related through shared variance accounted for by AS. The authors suggest that these results have impo rtant implications for treatment because AS is part of the cognitive behavior al model of anxiety disorders. T he cognitive behavioral model holds that anxiety disorders are the result of cognitive misconceptions about the nature of anxiety, and the treatme nt associated with this model attempts to correct these misconceptions. However, the EA framework of anxiety disorders holds that the problem unwanted thoughts, and oth er negative cog nitions; it is a key factor in Acceptance and Commitment T herapy (ACT) in that ACT teaches patients to accept and endure negative affect. Schmidt, Buckner, and Keough (2007) conducted a study concerned with evaluating the unique, prospective contributions of gender and AS to the onset of substance use disorders (SUD); they hypothesized that anxiety sensitivity would uniquely

PAGE 16

11 predict SUDs after accounting for gender and trait anxiety. They also wanted to elucidate the specific interaction, if any at all, between gender and anxiety sensitivity in reg ards to predicting SUD outcomes. The sample consisted of 405 participants; each participant was interviewed using the structured clinical interview for DSM IV TR non patient edition, and then given s elf report measures. The participants were given the anxiety sensitivity index and the state trait personality inventory; all participants were followed for up to 24 months. Each participant were assigned randomly to either the treatment or control conditi on; the treatment condition consisted of an intervention designed to reduce anxiety sensitivity, while the control condition group received a control intervention that focused on health and nutrition but did not address anxiety sensitivity (Schmidt et al., 2007). The results demonstrated that high anxiety sensitivity predicted the onset of (AUD) independent of trait anxiety and a history of SUD; due to the experimental design, this study identified anxiety sensitivity as a risk factor rather than merely a c orrelate of AUD (Schmidt et al., 2007). The study suggests that trait anxiety does not uniquely predict AUD, and it may simply be a correlate. The results suggested that there is no interaction between anxiety sensitivity and gender in regards to the predi ction of SUD. Additionally, levels of anxiety sensitivity in the sample were relatively low, which may have led to lower amounts of diagnosed psychopathology (Schmidt et al., 2007). Schmidt, Mitchell, and Richey (2008) performed a study on the possibility of an incremental relationship between AS and anxiety symptoms and disorders. They hypothesized that AS would incrementally predict panic and anxiety symptoms over time even after controlling for general negative affect. The sample was made up of 404 parti cipants who were all interviewed using the structured clinical interview for DSM

PAGE 17

12 disorders by graduate students who were extensively trained. They also completed self report measures including the anxiety sensitivity index, the panic disorder severity scal e, and the positive affectivity negative affectivity scale. The participants were randomly assigned to a control or treatment condition; the treatment condition received an intervention targeting anxiety sensitivity, while the control condition received an unrelated intervention (Schmidt et al., 2008). The findings of this study demonstrated a significant association between anxiety sensitivity and the development of panic attacks, and also the frequency of panic attacks (Schmidt et al., 2008). Anxiety sen sitivity also displayed an incremental longitudinal predictor of axis I pathology, and appeared to outperform more basic vulnerabilities of affective disorders such as negative affectivity. The data from this study suggested that anxiety sensitivity preced ed the onset of anxiety disorders, which implies that anxiety sensitivity is a risk factor for anxiety disorders. The sample was a relatively low risk nonclinical sample, and using a more at risk sample may provide more evidence. The study found that anxie ty sensitivity is an adequate incremental predictor of anxiety spectrum psychopathology, and that preventative aims to treat anxiety sensitivity may be an effective way to prevent anxiety disorders. Howell, Leyro, Hogan, Buckner, and Zvolensky (2010) condu cted a study aimed to assess the relationship between anxiety sensitivity, distress tolerance, discomfort intolerance, alcohol use motives, and alcohol use problems. The authors hypothesized that coping and conformity alcohol use motives would be significa ntly correlated with anxiety sensitivity. Additionally, they believed that anxiety sensitivity would be significantly and incrementally related to alcohol use problems, relative to distress

PAGE 18

13 tolerance and discomfort intolerance; they believed that this rela tionship would display significance above and beyond both daily smoking rate and negative affectivity. The sample consisted of 224 young adults who consumed alcoholic beverages in the past thirty day s (self reported), all of which met criteria for alcohol use disorders according to the alcohol use disorders identification test (AUDIT). In addition to the AUDIT, the study utilized the Anxiety Sensitivity Index, Discomfort Intolerance Scale, Distress Intolerance Scale, Drinking Motives Questionnaire Revised, the Smoking History Questionnaire, and Positive Affect Negative Affect scale; all of these self report measures have previously displayed high reliability. Coping motives were directly, positively correlated with distress in tolerance instead of anxiety sen sitivity, which was a surprise to the authors (Howell et al., 2010). This effect accounted for about 4% of unique variance and was apparent even after controlling for the significant variance accounted for by alcohol use, cigarette smoking, and negative af fectivity. Conformity, enhancement, and social motives were also significantly positively correlated with distress tolerance. This effect was discernible after controlling for shared variance of anxiety sensitivity and discomfort intolerance. As hypothesiz ed, anxiety sensitivity shared a significant, positive, and incrementally predictive cor relat ion with conformity motives for drinking, and this relationship was not seen with distress tolerance or discomfort intolerance; AS unique ly predicted 4% of varianc e in conformity motives (Howell et al., 2010). These results suggest that anxiety sensitivity and coping drinking motives are not significantly related when controlling for the shared variance of distress intolerance (Howell et al., 2010). However, the ve rsion of the DMQ R used in this study has a single

PAGE 19

14 coping motive variable, as opposed to separate coping with anxiety and coping with depression subscales. This may explain these results, because distress intolerance would include all forms of distress (an xiety and depression), while anxiety sensitivity would likely only play a large role in coping with anxiety motives. Therefore, the results should be considered in these terms: distress intolerance may account for greater variance in coping with all forms of distress, but anxiety sensitivity may account for greater variance in specifically coping with anxiety. Additionally, due to the non experimental nature of this study, it is possible that these relationships only appear as a result of shared variance be tween a third variable. Summary The literature on anxiety sensitivity has a number of important suggestions in terms of anxiety and alcohol use disorders. First, research suggests that anxiety sensitivity is a factor separate from general anxiety, and that it shares a greater association with fearfulness and anxiety disorder diagnosis than anxiety in general and Experiential Avoidance (Berman et al., 2010; Reiss et al., 1986). Second, it suggests that anxiety sensitivity can be conceptualized as a single construct with three more specific constructs subsumed beneath it. In other words, anxiety sensitivity is a single construct (fear of anxiety) that contains three more specific aspects of this construct, namely physical, social, and psychological concerns (Zinbarg et al., 1997). Third, pre vious studies indicate that anxiety sensitivity is not only associated with anxiety and substance use disorders, but that it is a risk factor which precedes the onset of a diagnosis (Schmidt et al., 2007; Schmidt et al., 2008). This third point is of parti cular importance because it suggests that anxiety sensitivity is a risk factor, as opposed to a correlate, which reinforces the importance of understanding how AS develops in order to treat and prevent

PAGE 20

15 it. Lastly, the research suggests that anxiety sensiti vity is related to conformity drinking motives, but not related to coping motives (Howell et al., 2010). However, the findings from Howell and colleagues study (2010) may be misleading and need additional research to support these findings, but the finding s do suggest that drinking motives are also important in the context of anxiety disorders, substance use disorders, and comorbid anxiety and substance use disorders. Drinking Motives Ham, Zamboanga, Bacon, and Garcia (2009) assessed the relationship betwee n social anxiety and alcohol related problems by exploring the role of drinking motives. The authors hypothesized that the relationship between social anxiety and problematic drinking would be a result of coping and conformity motives for drinking, while s ocial and enhancement motives would be unrelated. The sample was composed of 817 college students (73.7% wo men). The researchers used the Social Interaction A nxiety scale to I dentification Test (AUDIT) was used to measure problematic drinking, and the Drinking Motives Questionnaire Revised (DMQ R) was used to asses drinking motives (Ham et al., 2009). The results of this study revealed that social anxiety is generally negative ly correlated with problematic drinking; however, the results indicated that there was a positive indirect effect of social anxiety on some aspects (drinking consequences and dependence) of hazardous drinking, but it was mediated by coping motives (Ham et al., 2009). There was no effect of gender on drinking motives, social anxiety, and problematic drinking. Social anxiety appeared to be significantly positively correlated

PAGE 21

16 with all drinking motives, but the stronger association occurred with negative reinfo rcement (coping and conformity) as opposed to positive reinforcement (Social and Enhancement). Drinking to cope with negative affect increases as social anxiety levels increase, which is consistent with the stress response dampening theory; this theory hol ds that alcohol use is reinforced through relieving some of the physiological and psychological aspects of stress. Conformity motives did not mediate the relationship between social anxiety and hazardous drinking, which suggests that coping motives are uni que in this capacity. The authors concluded that social anxiety is negatively related to problematic drinking until it is used to alleviate symptoms, and after this it becomes reinforced into a habitual behavior. onrod (2007) conducted a series of studies testing the validity, reliability, and predictive power of a five factor model interpretation of the modified drinking motives questionnaire revised (DMQ R). Drinking motives were originally conceptualized as co nsisting of two factors (coping and social), and then conceptualized as having four factors (social, coping, conformity, and enhancement), and finally as having five factors (social, conformity, enhancement, coping anxiety, and coping depression). In the f irst study, they conducted confirmatory factor analysis on the five factor model; they hypothesized that it would have a better fit over data than the previously accepted four factor model. In the second study, they measured the test retest reliability of the modified drinking motives questionnaire revised, and its predictive power in terms of drinking behavior (Grant et al., 2007). In study one, they found that the five factor model fit the data well; the five factor model outperformed the four factor mo del in terms of fitting the data they interpreted

PAGE 22

17 (Grant et al., 2007). Specifically, the distinctiveness of coping anxiety and coping depression drinking motives was supported. In the second study, they found that the measurement had adequately high test re l p has ranging from .67 to .78). Additionally, they found that this model had powerful predictive power, and that coping anxiety and coping depression motives predicted different drinking patterns They found th at coping wit h anxiety motives we re directly related to later alcohol use problems, whereas coping with depression motives we re related indirectly to problematic alcohol use via their shared variance with amount of consumption (Grant et al., 2007). Grant and her collea gues (2007) presented substantial evidence in favor of the five factor model of drinking motives. The findings of this study elucidate the importance of motivation (in this case drinking motives) on subsequent behavior, but the question remains of whether drinking motives are associated with particular personality traits. Goldstein and Flett (2009) conducted a study analyzing this question, by examining the relationship between drinking motives, personality variables, and alcohol use. They were interested i n finding whether coping motives, enhancement motives, a combination of them, or non internally based motives were most related to frequency of alcohol consumption, alcohol related problems, and a number of personality variables. They studied 138 undergrad uate students in terms of drinking motives, neuroticism, anxiety sensitivity, sensation seeking, positive and negative affect, alcohol use (frequency and typical amount), and alcohol related problems. They found that enhancement and coping drinking motives were highly positively correlated; this supports the authors claim that these motives share an association (Goldstein & Flett, 2009). Additionally, they found that negative affect was higher in

PAGE 23

18 copin g motivated drinkers than enhancement motivated drinkers, while positive affect was higher in enhancement motivated drinkers than those motivated to consume alcohol for coping reasons. Additionally, they found that coping motivated drinkers scored higher o n the anxiety sensitivity index in general than those high in enhancement motivation or non internally motivated drinkers; also, they found that individuals with a combination of coping and enhancement drinking motives looked very similar to those with jus t coping motives (Goldstein & Flett, 2009). They also found that those with enhancement motives alone had low binge drinking rates compared to those with coping motives or coping and enhancement motives; the same pattern was found with alcohol related prob lems. This article supplies evidence that drinking motives are differentially related to personality characteristics, and are associated with different drinking patterns (Goldstein & Flett, 2009). Summary Upon examining the available research on drinking motives, it becomes apparent that they play a crucial role in the association between anxiety and alcohol use. Previous research has found that coping and conformity motives facilitate a relationship between social anxiety and problematic drinking (Ham et al., 2009). Other studies have suggested that drinking motives are differentially related to personality characteristics, including a relationship between coping drinking motives and anxiety sensitivity (Goldstein & Flett, 2009). Additionally, research sug gests that drinking motives can be conceptualized as including five factors, and specifically that coping motives can be broken down into separate coping with anxiety and coping with depression factors (Grant et al., 2007).

PAGE 24

19 Attachment Theory Attachment theory is a framework used to understand behavior in terms of working cognitive models of the self and others. This conceptual lens arose out of work in developmental psychology to understand the infant mother bond (for full review, see: Ainsowr th & Bowlby, 1991). This framework was extended to explain adult relationships in terms of working models of the self and others (Hazan & Shaver, 1987). Attachment behavior and cognition is now thought to be vulnerable to context (Baldwin, Keelan, Fehr, En ns, & Koh Rangarajoo, 1996). Many studies have suggested that attachment insecurity, or negative working models of the self and others, share a robust relationship with anxiety sensitivity and certain drinking motives (Doumas et al., 2007; Kassel et al., 2 007; McNally et al., 2003; Molnar et al., 2010; Sutin & Gillath, 2009; Viana & Rabian, 2008; Vungkhanching et al., 2004; Watt et al., 2005; Weems et al., 2002). Therefore, to fully understand anxiety sensitivity and drinking motives it is necessary to unde rstand attachment and the association it has with anxiety sensitivity and certain drinking motives. Attachment and Anxiety Sensitivity Studies on anxiety sensitivity had originally looked at whether or not it was a valid cognitive construct, and what kind of behaviors AS could predict. There was little research investigating possible causes of AS. In an attempt to address this gap in the literature, Weems, Berman, Silverman, and Rodriguez (2002) conducted a study focusing on the relationship between attach ment dimensions and anxiety sensitivity in adolescence and early adulthood. They hypothesized that dismissive and secure attachment styles would have a weaker relationship with anxiety sensitivity than preoccupied and fearful styles in both samples.

PAGE 25

20 They m easured attachment with the Experience in Close Relationships scale (ECR) used the anxiety sensitivity index and its equivalent for children to assess anxiety sensitivity; they used the Symptoms Check List 90 (SLC 90) to control for anxiety and overall distress The ECR provides a measure of attachment anxiety and avoidance, which can be also used to categorize individuals into four attachment groups: preoccupied (high anxiety, low avoidance), fearful (high anxiety, high avoidance), Dismissive (low anxie ty, high avoidance), and secure (low anxiety, low avoidance) (Weems et al., 2002). The results suggested that their hypothesis was correct; preoccupied and fearful attachment styles had a significantly stronger relationship with anxiety sensitivity than d ismissive and secure styles (Weems et al., 2002). The authors argued that although causality cannot be implied, this study suggests a robust relationship between attachment and anxiety sensitivity. Additionally, the authors analyzed the relationship betwee n attachment styles and clinically significant levels of anxiety sensitivity; the cut off for clinical significance has been established as 25 for possible problems, and 30 for a high likelihood of being diagnosed with anxiety disorder (Weems et al., 2002) They found that 75% of the participants with clinically significant levels of anxiety sensitivity were classified as fearful or preoccupied. Additionally, they found that the model of self dimension was significantly related to anxiety sensitivity after controlling for SCL 90 anxiety and global distress. The authors suggest that attachment theory may provide a framework to better understand the development of anxiety sensitivity (Weems et al., 2002). In a similar study, Watt, McWilliams, and Campbell (200 5) found that different attachment representations (romantic vs. general) have different associations with AS.

PAGE 26

2 1 Additionally, Viana and Rabian (2008) provided evidence that AS actually mediates the relationship between attachment and symptoms of Generalized Anxiety Disorder (GAD). Viana and Rabian suggested, similarly to Weems and colleagues (2002), that AS may actually stem from attachment representations and early interactions with the caregiver. While Watt and colleagues (2005) provided evidence that roma ntic attachment also plays a role in AS, and that the association is distinct from that of general attachment. The findings in the literature, and the theoretical perspective provided by these authors suggest that AS may be a product of attachment represen tations. The literature presented, however, has not investigated a possible relationship between attachment representations and drinking motives. Attachment is primarily a motivational factor: it motivates us to maintain proximity to our caregiver, to expl ore the social world, and to seek comfort with our caregiver (Ainsworth & Bowlby, 1991). Due to the relevance of attachment schemas in co ping behavior, such as seeking comfort from others, it follows that attachment may play a role in being motivated to se ek comfort elsewhere, such as in alcohol use. Attachment and Drinking Motives McNally, Palfai, Levine, and Moore (2003 ) studied the relationship between attachment and alcohol related problems, and the possibility of drinking motives as a mediating variable. They hypothesized that the model of self aspect of attachment (attachment anxiety) will be related to problematic alcohol consump tion, but that the model of others aspect (attachment avoidance) would not have this association. The sample consisted of 366 college students (56% females) who had indicated that they had at consumed at least one alcoholic beverage in the past year. A ttac hment was measured using four descriptive categories; participants chose which one best exemplified their experience. Alcohol use was measured in terms of frequency using

PAGE 27

22 a self report measure, and another self report measure was used to assess drinking re lated problems (both life time and in the past year); drinking motives were assessed using the drinking motives questionnaire. They found that after controlling for frequency of consumption and gender, conception of self (attachment anxiety dimension) alo ne was the sole predictor of alcohol related problems (McNally et al., 2003). Additionally, they found that coping motives were significantly related to the model of self; the other three motives were not related to the self model. They also found that cop ing motives mediated the relationship between attachment anxiety and alcohol related consequences. The authors concluded that the results suggest that the negative model of self aspect of attachment is significantly related to difficulty regulating emotion s in healthy, adaptive ways; one maladaptive form of attempted emotional regulation appears to be using alcohol as a mechanism for coping with negative affect. This study could not imply causality, due to its non experimental, cross sectional design This research has important implications because they provide evidence that attachment insecurity has a significant relationship with coping drinking motives (McNally et al., 2003). Kassel, Wardle, and Roberts (2007) investigated the possibility of dysfunctio nal beliefs and self esteem mediating the relationship between attachment and substance use. D ysfunctional beliefs were conceptualized as unrealistic standards for interpersonal interactions and achievements. They measured both frequency of drug use and wh ether they use these drugs to cope with negative affect; the researchers focused on alcohol, marijuana, and cigarette use (Kassel, Wardle, & Roberts, 2007).

PAGE 28

23 They found that insecure attachment was only related to frequency of cigarette, alcohol, or mariju ana use if it was to reduce negative affect (Kassel et al., 2007). Additionally, they found that the relationship between attachment and alcohol and marijuana use was mediated by self esteem and dysfunctional beliefs. The attachment style questionnaire had a fairly low reliability (two of three dimensions were below .7), which may have led to less robust results. The results of this study suggest that insecure attachment is related to coping motives (distress reduction) for multiple substances including al cohol (Kassel et al., 2007). Doumas, Blasey, and Mitchell (2007) examined the relationship between adult attachment style, interpersonal problems, and emotional distress in a population of alcohol and other drug dependent individuals at an inpatient rehab center. They hypothesized that individuals with a negative view of the self would be overrepresented in this population, and that those with a negative working model of the self would report higher levels of emotional distress and interpersonal problems. The sample consisted of 46 drug/alcohol dependent patients at an inpatient rehabilitation center Distress, interpersonal problems and attachment were measured using self report methods, and distress was operationalized as levels of anxiety and depression They found that a negative working model of the self (high attachment anxiety) was overrepresented in this population; there were twice as many preoccupied and fearful individuals, and a third the number of secure individuals as compared to a normative p opulation (Doumas et al., 2007). They also found that a negative self model (high attachment anxiety) was significantly associated with interpersonal problems. Additionally, they found that preoccupied and fearful (both high in attachment anxiety)

PAGE 29

24 particip ants scored significantly higher on the depression and anxiety inventories than the secure or dismissive groups (both low in attachment anxiety). This study provided more evidence of the association between attachment and alcohol abuse; additionally, it se ems likely that these individuals were abusing alcohol in an attempt to cope with their experiences of anxiety and/or depression Vungkhanching, Sher, Jackson, and Parra (2004) investigated the relationship between family history of alcoholism, attachment styles, and alcohol use disorders. Attachment styles have previously been shown to be a ffected by alcoholic caregivers such that they have higher rates of insecure attachment styles. This study included 369 first time freshman undergraduate students; they were all screened for family history of alcoholism using the family short Michigan alcoholism screening test, and split into children of alcoholics (COA) group or non were measured using categorical definitions; the participant rated how well they matched their behavior in close relationships and which definition was most self characteristic. The participants were assessed for a lcohol use disorders using the Diagnostic Interview Schedule V ersion IV. The researchers f ound that there were higher rates of insecure attachment styles in the COA group (46%) as compared to the non COA group (28%) (Vungkhanching et al., 2004). However, they were unable to isolate a specific attachment style as being the most highly correlated with family history of alcoholism. Insecure attachment styles in general were correlated with alcohol use disorder. The researchers did not find a mediating effect of family history in the relationship between attachment style and alcohol use disorders. T hough the researchers suggest that attachment is related to alcohol use disorders, it may

PAGE 30

25 be the result of a shared variance with anxiety sensitivity and coping drinking motives (Kassel et al., 2007; Schmidt et al., 2007; Weems et al., 2002). Molnar, Sadav a, DeCourville, and Perrier (2010) examined the relationship between attachment, drinking motives, and problematic alcohol use in a clinical and non clinical (student) population. Problematic drinker was defined in terms of risky drinking behavior (frequen cy/quantity) and negative consequences as a result of alcohol consumption; the latter included aspects pertaining to relationships, family, work, law, and health. These constructs were measured with questionnaires. Drinking motives were assessed using the Drinking Motives Questionnaire; attachment was assessed using the Relationships Scale Questionnaire, which provides scores for the avoidance and anxiety dimensions of attachment. They were interested in testing a dual path model they had developed for drin king problems; the model consists of an affective path (coping motive) and a social path (conformity). They hypothesized that both of these motives would lead to alcohol problems. The results supported the dual path model; both the affective and the social paths were related to problems with alcohol consumption (Molnar et al., 2010). The dual path model was entirely mediated by attachment anxiety in the nonclinical sample; in the clinical sample, the social path did not obtain significance, while the affect ive path did obtain significance. Avoidance was significantly related to enhancement drinking motives, but only among men. The researchers concluded that high attachment anxiety sets the stage for risky alcohol consumption and alcohol related consequences. They further concluded that insecure attachment acts a risk factor for the development of problematic alcohol use. This study is solid evidence that attachment anxiety and coping

PAGE 31

26 drinking motives are related, but due to the nature of the data, it cannot support such a causal claim (Molnar et al., 2010). State Attachment In the seminal work on state attachment, Baldwin, Keelan, Fehr, Enns, and Koh Rangarajoo (1996) investigated the possibility that individuals experience a variety of attachment styles, have working models of multiple attachment styles, and that these models are subject to availability and accessibility effects. Their research intent grew ou t of social cognitive theory; specifically, the subjects of and accessibility (the ease of which a representation is retrieved and applied to a novel situation) were investigated in regards to attachment models They conducted three studies to better understand the role of accessi bility and availability in attac hment representations. The first study concerned whether or not the individual experiences multiple attachmen t styles and if his or her general attachment style relates to the sheer number of relationships fitting this style. The second study investigated the relationship between attachment style and the accessibility of examples of particular attachment styles, and the third was concerned with whether or not accessibility could be manipulated using a priming paradigm (Baldwin, Keelan, Fehr, Enns, & Koh Rangarajoo, 1996). In each study, general attachment style was measured (Baldwin et al., 1996). In the first st udy, participants were asked to list 10 attachment figures, characterizing each within the frame of a specific attachment style, and then they were asked about some general expectations in these relationships. They found that the majority of participants e xperienced a variety of attachment styles (two or three out of three styles), they found a

PAGE 32

27 significant positive relationship between general attachment style and the percentage of their relationships characterized by that pattern; although individuals gene rally experienced all of the attachment styles, they experienced the style most that corresponded with their self reported global attachment style. In the second study, they measured the ease with which relationships characterized by each style were retri eved from memory ; individuals in each category reported more ease retrieving a corresponding relationship than those in other categories (e.g., anxious ambivalent individuals reported more ease generating a relationship characterized by this style than sec ure or avoidant individuals). In the third study, individuals were randomly assigned to a priming category (anxious ambivalent, avoidant, or secure) and subsequently reported on how attracted they were to a potential dating partner with a randomly assigned attachment style. The researchers found that individuals found secure potential partners more attractive in general, but their primed attachment style corresponded to how attractive they rated potential partners with the same style (those primed for anxio us ambivalent attachment rated anxious ambivalent potential partners as more attractive than those primed with security or avoidance). The author concluded based on the results from all three studies that individuals experience a variety of attachment patt erns, that their dominant pattern corresponds to the frequency of relationships characterized as such, and that the different relational representations are subject to availability and accessibility factors (Baldwin et al., 1996). Pierce and Lydon (1998) c onducted experimental research on the relationship between interpersonal schemas (attachment dimensions) and response to stressful events. The researchers conducted two studies. In the first study, the experimenters measured the

PAGE 33

28 effect of priming attachmen t representations on affective experience of a stressful event and coping strategies. Participants were primed with either attachment security or insecurity, then participants were told to imagine that they just found out about a surprise/unwanted pregnanc y. Finally, participants completed self report measures of affect experienced as a result of the stressful event, as well as coping strategies employed. The coping strategies either fit into a growth oriented pattern or a self denigrating pattern. The rese archers controlled for the effect of chronic attachment representations. The second study assessed the possibility of the prime activating affective responses, and thus confounding affect with interpersonal expectations. The results of the first study dem onstrated that priming negative attachment representations caused increases in expected negative affect as a result of the stressful event compared to the control group (Pierce & Lydon, 1998). Additionally, priming positive interpersonal expectations led t o increased support seeking behaviors when compared to the control group. In terms of coping strategies, the only significant effect was the reduction of self denigrating coping strategies when primed with positive attachment representations as opposed to the control. The results of study suggested that state affect was not influenced by the priming stimulus, and therefore played no role in the results. This study suggests that priming attachment has noticeable effects on coping behaviors whic h is of parti cular interest in light of observed associations between coping drinking motives and attachment insecurity (McNally et al., 2003) Cassidy, Shaver, Mikulincer, and Lavy (2009) investigated the influence of attachment security in the processing of psycholo gical pain in order to understand the possible buffering effects of secure attachment on experiences of psychological pain.

PAGE 34

29 Bowlby noted that i ndividuals high in attachment avoidance often defensively disengage emotionally when a psychologically painful e xperience occurs; priming security may alter this reaction (Ainsworth & Bowlby, 1991). On the other hand, he noted individuals high in attachment anxiety often overreact to psychologically painful events; having an attachment security schema activated may reduce the intensity of these exaggerated reactions. Seventy participants completed the Experiences in Close Relationships (ECR) measurement of adult attachment style. Next, participants were asked to record a story, including some specific details, about an instance in which they were psychologically security or a control prime, and finally answered some questions about how they would react to the event they described if it were to happen in the present (Cassidy et al., 2009). The results suggest that attachment anxiety displayed a positive association with feelings of rejection and other negative emotions (Cassidy et al., 2009). However, when primed with attachment securi ty, individuals with anxious attachment had described significantly fewer feelings of rejection and other negative emotions; they experienced the psychological pain with less exaggerated emotional reactions. Attachment avoidance was positively associated w ith experiencing less intense feelings of rejection, and less intense appraisals of psychological pain. Individuals with high attachment avoidance experienced significantly more psychological pain when primed for attachment security (i.e., they were engagi ng more emotionally, and not defensively shutting down). These results suggest that emotional processes are greatly affected by attachment dimensions, both trait level dimensions and state level dimensions (Cassidy et al., 2009). Therefore,

PAGE 35

30 these dimension s may impact much of our information processing, including in terms of coping methods and beliefs related to distress. Priming state attachment representations has been found to affect individuals in a variety of ways, including agency and communion orient ations (Bartz & Lydon, 2004). Attachment primes have also demonstrated effects on feelings of security/insecurity, positive/negative emotions, sense of positive care, sense of merging, and feelings of nostalgia (Carnelly & Rowe 2010). Researchers have also observed an effect of attachment primes on tru st, forgiveness, and destiny/ growth beliefs in terms of romantic relationships (Finkel, Burnette, & Scissors, 2007). Repeated attachment primes have even been shown to induce long term changes in attachment representations (Carnelly & Rowe, 200 7). The state of the literature clearly demonstrates the viability in a wide variety of experimental investigations. Of particular interest in the current study is the robust effect of priming state attachment. Another major point of interest in the literature is that attachment primes have a demonstrable effect on the experience of psychological pain (Cassidy et al., 2009). Attachment primes also have an effect on hypothetical coping behavior, which suggests that they m ay have an effect on the likelihood of using alcohol as a coping mechanism (Pierce & Lydon, 1998) The results of the previously mentioned studies suggest that priming state attachment indeed has a measurable effect on various cognitive processes of individ uals (Baldwin et al., 1996; Pierce & Lydon, 1998; Bartz & Lydon, 2004; Carnelly & Rowe, 2007; Carnelly & Rowe, 2010;Cassidy et al., 2009; Burnette et al., 2007). The overall trend of the literature suggest s that attachment primes are a viable experimental tool to examine the effects of relationship schemas on cognition, emotion, and behavior.

PAGE 36

31 The Current Study The literature on the relationship between attachment, anxiety sensitivity, and drinking motives has an observable general trend; attachment insecu rity, mostly attachment anxiety, has shown a positive correlation with anxiety sensitivity, as well as a positive correlation with coping drinking motives (Sutin & Gillath, 2009; Viana & Rabian, 2008; Watt et al., 2005; Weems et al., 2002). Some researcher s have suggested that attachment anxiety plays a causal role in the development of anxiety sensitivity (Weems et al., 2002; Viana & Rabian, 2008) and the development of alcohol use as a coping mechanism ( Molnar et al., 2010; McNally et al., 2003) ; however, the research on this subject has been predominantly correlational in nature (Doumas et al., 2007; Gonzalez et al., 2008; Kassel et al., 2007; McNally et al., 2003; Molnar et al., 2010; Vungkhanching et al., 2004;Watt et al., 2008; Weems et al., 2002; Vian a & Rabian, 2008) While correlational research provides a glimpse into the relationships between variables, experimental research is necessary in order to infer causality with any degree of confidence. In the present study, the relationships between attac hment dimensions, anxiety sensitivity, and drinking motives will be observed in an experimental setting. Two main research questions arose from findings presented in the literature. Does attachment anxiety lead to higher endorsement of the beliefs that mak e up anxiety sensitivity than attachment security? Does anxious attachment lead to a stronger belief in the efficacy of consuming alcohol as an effective coping mechanism than secure

PAGE 37

32 attachment? In order to answer these questions in an experimental setting participants were primed with either attachment anxiety or security, and then completed self report measurements of anxiety sensitivity and drinking motives. The anxiety sensitivity index and the drinking motives questionnaire were chosen due to their hi gh test retest reliability (Grant et al., 2007; Zinbarg et al., 1997). A scrambled sentence method was used to prime different state attachment representations; a subliminal prime was chosen as to not make the manipulation overt Using this method, the pre sent study attempts to test the relationship between attachment dimensions on one hand, and anxiety sensitivity and drinking motives on the other in an experimental setting. First, it was hypothesized that priming state attachment anxiety as opposed to att achment security and the control prime would result in higher adherence to the beliefs that make up anxiety sensitivity as measured by the ASI. Second, it was hypothesized that priming state attachment anxiety versus attachment security would result in higher endorsement of consuming alcohol as a coping mechanism as measured by the DMQ R. If the results support the first hypothesis, and insecure attachment primes lead to higher ASI scores than the secure and control prime, then it may be that anxiety sensitivity is rooted in attachment anxiety; it may be that individuals who have high attachment anxiety may begin to fear distress, because it has been more of a source of trouble (as opposed to a passing discomfort). If the results support the second hypothesis, insecure attachment primes lead to h igher endorsement of coping drinking motives than the control or security prime, then it may be that adhering to coping related drinking motives may be a result of attachment anxiety; those with high attachment anxiety are not easily

PAGE 38

33 soothed by caregivers, which may lead them to seek comfort elsewhere, such as in the consumption of alcohol. Method Participants The participants were recruited using an online student forum, which is connected to student email addresses. All participants were students at a s mall public liberal arts college in southwest Florida. Eight participants dropped out before completing any part of the survey ( Initial Sample, N = 80 ; Final Sample, N = 72). The sample contained more females than males ( Male s = 15 Female = 50 Other = 4, Missing = 3 ). Participants were limited to individuals 21 and over, due to questions regarding alcohol us e (Age range: 21 26; Median = 21; Missing = 12 ). Materials Priming Stimulus. The priming stimulus was a scrambled sentence paradigm that was develop ed by Finkel et al. (2007); this stimulus demonstrated a measurable increase in the accessibility of the intended attachment schema. A sub tle prime was chosen in an attempt to hide the true intent of the study from participants. The scrambled senten ce task consisted of six target sentences, with a single word changed between the secure, anxious, and control conditions. Each word scramble contained five words, and participants were instructed to eliminate one word and make a sentence with the remainin (Appendix A) whereas the attachment anxiety condition (Appendix B)

PAGE 39

34 con dition (Appendix C) All three lists contained the same four filler sentences, which were not related to attachment. These were presented as a cognitive task, where the participants completed the task by typing the formed sentences into an online survey fo rm which had the directions and word scrambles (See Appendix A B, and C ). Anxiety Sensitivity Index. The Anxiety Sensitivity Index (ASI) is a frequently used measurement for assessing anxiety sensitivity. There are multiple versions available; the versio n chosen for this study consists of 16 items related to the fear of anxious arousal ( Reiss, et al., 1986) The questions regard the degree to which participants endorse the 16 statements, which are related to different aspects of feeling apprehension towar ds behaviors, sensations, and cognition associated with anxious arousal. An example, taking participants would then rate their agreement from very little (0) to very much (4) The This version of the ASI demonstrated high test retest reliability ( N =147; r = .71 .75) in a study conducted by Reiss et al. (1986). The ASI is scored by summing the scores on the 16 items for the total ASI level, which results in a range of 0 64. Research has suggested that anxiety sensitivity scores approach clinical significance at 25, which denotes participants at risk for possible problems with anxiety; individuals with scores above 30 are regarded as having a high likelihood of a diag nosis of an anxiety disorder (Weems et al., 2002). Zinbarg et al. (1997) demonstrated the hierarchical st ructure of ASI through factor analysis; the

PAGE 40

35 measurement consists of three lower order factors (social, cognitive, and physical ), which compri se the hig her order total anxiety sensitivity score (Zinbarg et al., 1997). This interpretation of the ASI is used in this study for data analysis, resulting in a total score (16 items, score range 0 64), as well as three sub s cores (social: 4 items, score range 0 16; Physical: 7 items, score range 0 28; Psychological: 5 items, score range 0 20) (See Appendix D) Drinking Motives Questionnaire Revised. In the present study, the five factor Modified Drinking Motives Questionnaire (Modified DMQ R) was chosen to meas ure the degree to which individuals endorse reasons to consume alcohol (Grant et al., 2007). participants are asked to rate each item in terms of how often they consume alcohol ic beverages for the given reason. Each item is rated on a 5 point likert type scale (1 = never/almost never; 5 = almost always/always); the scores for each subscale are obtained by summing the items contained within them. The five factors include social m otives (5 items, score range: 5 25), conformity motives (5 items, score range: 5 25), enhancement motives (5 items, score range: 5 25), coping anxiety motives (7 items, score range: 7 35), and coping depression motives (2 items, score range: 2 10 ). This measure has shown adequate test retest reliability and predictive validity ( Grant et al., 2007) (See Appendix E) Procedure Participants were randomly assigned to the attachment security prime group, the attachment anxiety prime group, or the contr ol prime group using a program for random

PAGE 41

36 assignment ( N Secure = 18, N Control = 28, N Insecure = 26) Participants were recruited using a college run online forum, and followed a link to the online survey. The online survey included the scrambled sentence ta sk (priming task), the anxiety sensitivity index (ASI) and the Modified Drinking Motives Questionnaire Revised (Modified DMQ R). All participants first completed the scrambled sentence task. The two questionnaires, the ASI and the Modified DMQ R, came afte r the scrambled sentence task in random order, which was accomplished using an option on the online survey. After completing these two questionnaires, demographic information was obtained which consisted of gender and age. After completing the scrambled sentence task, the ASI, the Modified DMQ R, and the demographic information, participants completed the scrambled sentence task for the secure condition. This security prime was not related to the experiment itself, but was included to undo any negative af fect induced by the initial prime. Previous research has suggested that security primes increase positive affect, decrease negative affect, influence response to psychological pain, and promote feelings of security and well being (Carnelly & Rowe, 2007; Ca rnelly & Rowe, 2010; Cassidy et al., 2009; Finkel et al., 2007). The participants were then debriefed, provided with a correspondence email to address any questions, given information about campus resources for mental health, and asked to please refrain f rom discussing the experiment with fellow students for six months.

PAGE 42

37 Results Descriptive Statistics for all Dependent Measure Descriptive statistics were computed for each dependent measure overall. The results of these descriptive analyses are presented in Table 1. Descriptive statistics were also computed for each dependent measure by each priming group. The results of these descriptive analyses are presented in Table 2. The Effect of Attachment Prime on ASI Total Score In order to test the hy pothesis that state attachment has an effect on anxiety sensitivity, a one way analysis of variance (ANOVA) was conducted with group (attachment prime) as the independent variable, and the total score of the ASI as the dependent variable. In contradiction to the hypotheses, the analysis did not find a significant effect of group on the dependent variable ASI total score ( F (2, 60) = .25; p = .7833; R 2 = .0081; See Table 3). The Effect of Attachment Prime on ASI Psychological Subscale To test whether state attachment impacted the psychological aspect of AS a one way ANOVA was conducted with attachment prime as the independent variable, and the ASI psychological subscale as the dependent variable. This analysis was not significant suggesting that there was no effect of attachment prime on ASI psychological subscale ( F (2, 60) = .23; p = .7987; R 2 = .0075 ; See Table 3 ). The Effect of Attachment Prime on ASI Physical Subscale

PAGE 43

38 In regards to the effect of state attachment primes on the physical aspect of AS, a one way ANOVA was conducted with attachment prime group as the independent variable, and the ASI physical subscale as the dependent variable. The relationship between priming condition and ASI psychological subscale was not significant ( F (2, 60) = .47; p = .6292; R 2 = .0153 ; See Table 3 ). The Effect of Attachment Prime on ASI Social Subscale T he relationship between attachment prime condition and the ASI social subscale was tested using a one way ANOVA. The analysis did not obtain significance ( F = (2, 60) = .43; p = .6522; R 2 = .0141), suggesting that attachment prime had no discernible effect on self report ASI social subscale scores (See Table 3) The Effect of Attachment Prime on DMQ Combined Coping Scale To assess the effect of attachment p rimes on coping drinking motives, a one way ANOVA was conducted with attachment prime condition as the independent variable and a composite of the two coping subscales of the DMQ as the dependent variable. There was no significant effect of attachment condition on coping drinking motives ( F (2, 60) = .70; p = .5011; R 2 = .0228 ; See Table 3 ). The Effect of Attachment Prime on DMQ Coping with Depression Subscale To test the effect of attachment prime condition on score s on the DMQ coping with depression subscale was tested using a one way ANOVA. There was no effect attributable to attachment prime group ( F (2, 60) = .38; p = .6878; R 2 = .0124 ; See Table 3 ).

PAGE 44

39 The Effect of Attachment Prime on DMQ Coping with Anxiety Subsc ale A one way ANOVA was conducted to analyze the effect of attachment prime group on the coping with anxiety drinking motives. The analysis did not find a significant effect of attachment prime on coping with anxiety drinking motives ( F (2,60) = 1.16; p = .3203; R 2 = .0372 ; See Table 3 ). MANOVA with Attachment Prime as the Independent Variable A Multiple ANOVA (MANOVA) was conducted with attachment prime condition as the independent variable, and all of the above as the dependent variable. The overall mod el was not significant, suggesting that attachment prime did not uniquely contribute to predicting scores on any of the dependent variables ( F (10, 112) = .57; p = .8382 ; See Table 3 ). Discussion The results suggested that the implicit attachment prime did not affect scores on the ASI or the modified DMQ R; there was no significant effect of attachment prime on any of the dependent measures. This may have important implications for current theories of anxiety sensitivity and drinking motives. The results may imply that AS and drinking motives are not a product of attachment insecurity; instead, it may be a directional issue in that AS and or drinking motives result in attachment insecurity, or that this relationship is mediated through another variable or variables There is also the possibility that these results correspond to differences between the effects of state attachment and the more persistent trait attachment. Of course, there is a distinct possibility that the present research suffered from limitations, and that if given perfect circumstances, these

PAGE 45

40 predictions would have been supported. Future research is necessary to make any reliable conclusions on these matters. H1: Attachment Primes an d Anxiety Sensitivity The attachment primes in no way predicted levels of anxiety sensitivity, which is to say attachment primes neither accounted for ASI total scores, nor scores on any of the three subscales (social, physical, and psychological). This ma y imply that causality lies in the opposite direction such that higher levels of AS lead one to have insecure attachment representations. There is also the possibility that the observed relationship between AS and attachment representations is mediated thr ough a third variable, or that they are both related to another variable. Although the results did not support this hypothesis, there still may be value in the results, because they suggest that the previously supported theory may not be an adequate accoun t of AS. There is a possibility that the relationship between attachment and AS is directionally reversed in refernce to the hypothesis: high levels of AS cause insecure attachment representations. Since the majority of the previous research has been non e xperimental, there is no previous support for AS being a direct cause of attachment (Weems et al., 2002; Watt et al., 2005; Viana & Rabian, 2008). The authors who observed the relationship between attachment insecurity and AS suggested that the most logica l explanation is that attachment insecurity causes AS. However, the correlation could also be explained by AS causing attachment insecurity. For instance, it is possible that an individual develops AS, is afraid of experiencing anxious arousal, which may a ffect the likelihood of the individual to be socially active (due to fear of social anxiety),

PAGE 46

41 which may in turn cause their relationships, and ultimately their attachment schemas, to suffer. While this is less logically sound tha n the opposing theory, whic h holds that AS the individual associating distress with negative reactions from others, it would be supported by the present results. However, it seems more plausib le that the association between AS and attachment insecurity is actually a result of shared variance with another variable, such as general anxiety (i.e., it is possible that AS and attachment both correlate with general anxiety, but do not directly influe nce one an other). Watt and colleagues (2005) found that attachment and AS were related after controlling for general anxiety, which contradicts this account for the results of the current study. Therefore, it seems unlikely that AS and attachment are relat ed via a shared variance with general anxiety, but it is still possible that they are related through a shared association with some other variable. It is also possible that attachment insecurity and anxiety sensitivity are only related through a shared v ariance with anxiety disorders, alcohol use disorders, and comorbid anxiety and alcohol use disorders. AS is a documented preceding risk factor for substance use disorders (Schmidt et al., 2007), as well as an incremental predictor of later anxiety disorde r symptoms (Schmidt et al., 2008). There is evidence that populations of alcohol ics in in patient treatment centers have much higher rates of insecure attachment than other populations (Doumas et al., 2007), as well as in individuals diagnosed with GAD (Via na & Rabian, 2008). Therefore, it is possible that the apparent association between attachment anxiety and AS is actually a result of shared variance between AS and alcohol use disorders on the one hand, and attachment insecurity and alcohol use

PAGE 47

42 disorders on the other hand. However, this account does not seem to be the most parsimonious explanation for the association between insecure attachment and AS, because these associations have been documented in clinical and nonclinical samples (see: Viana & Rabian, 2008; Weems et al., 2002; Watt et al., 2005). Perhaps the most compelling explanation is that the non significant results reflect a major difference between state and trait attachment, namely, a difference in the rate of recurrence in terms of attachment schemas. Weems and colleagues commented (2002), when they found a strong association between insecure attachment and AS, that AS may stem from insecure attachment. The logic behind this claim is as follows: individuals develop insecure attachment, specific ally higher levels of attachment anxiety, when their primary caregiver responds in peculiar ways to their distress; therefore, it is possible that the individual forms a fear of distress, namely anxious arousal, because of these inappropriately negative re actions from their primary caregiver. This suggests that the connection between attachment insecurity and AS is not quite as simple as the present experimental design assumed: it may be the case that simply bringing to mind different attachment representat ions is not enough to affect relative levels of anxiety sensitivity. It is possible, and given the present results it seems plausible, that the process of attachment causing changes in levels of AS is a lengthy process, involving repeatedly applying insecu re attachment schemas when experiencing anxiety. To further illustrate this previous account, the process may proceed as follows: the individual has insecure attachment, and specifically attachment anxiety, they experience anxious arousal while this schema is active, and they interpret that anxiety as dangerous, especially socially dangerous. Over time, as the individual repeatedly applies

PAGE 48

43 this schema to experiences of anxious arousal, they begin to incorporate exaggerated levels of risk associated with anx ious arousal. Therefore, it is not simply enough to induce the key attachment schemas; the individual must repeatedly apply these schemas to the experience of anxiety, altering their conception of anxiety, and related constructs (in this case AS). This exp lanation seems like an adequate reconciliation between the observational literature, and the present results (Weems et al., 2002). H2: Attachment Prime and Coping Drinking Motives king motives as assessed by the Modified DMQ R (Zinbarg et al., 2008) showed no variance associated with state attachment It may be that attachment and coping drinking motives have a causal relationship in a different direction than predicted; i.e., devel oping coping drinking motives, or even the drinking that results from coping drinking motives, may cause attachment to suffer. However, it is also a possibility that the observed relation between attachment insecurity and coping drinking motives is the res ult of a covariate, and there is no direct causal relationship. There is another possible explanation, which is that state attachment and trait attachment differ in one primary way, the rate of recurrent activation, which may be an important distinction in terms of drinking motives. The results suggest that priming state attachment had no effect on drinking motives, but it is possible that causality exists in the opposite direction. In this case, developing coping drinking motives could damage close relati onships, resulting in insecure attachment. It is possible that seeking comfort in alcohol, rather than in others, could result in insecure attachment, either because of some property unique to using

PAGE 49

44 alcohol for comfort, or because increased frequency of al cohol consumption affecting relationships, and ultimately attachment representations. It seems unlikely that there is some unique property of using alcohol as a coping mechanism that results in attachment insecurity. However, it seems possible that increas ed consumption of alcohol leads to damaged close relationships, ultimately affecting attachment representations. Another possibility is that coping drinking motives and attachment insecurity are only related indirectly, possibly through problematic alcoho l use. Alcohol abuse may result in estranging close others, which may affect attachment representations; in this case, insecure attachment can be an indirect result of problematic alcohol use, which is associated with coping drinking motives. Problems in r elationships caused by alcohol use are a prerequisite for diagnosis with an alcohol use disorder (AUD) (APA, 2000), so there is little surprise that those diagnosed with AUD often have damaged relationships. Also, it would not be a stretch that these probl ems in close relationships would affect attachment schemas, thereby demonstrating a causal link between alcohol abuse and insecure attachment. Coping drinking motives have been found to be related to alcohol abuse, and so this could explain why previous re search has found an association between attachment insecurity and coping drinking motives (McNally et al., 2003; Kassel et al., 2007; Doumas et al., 2007). This would reconcile the observed relation between coping drinking motives and insecure attachment f ound in the previous literature with the present findings. However, this explanation does not seem the most logical because it is thought that coping drinking motives precede and act as a risk factor for alcohol abuse (Molnar et al., 2010), and because pre vious research has demonstrated that the association between

PAGE 50

45 alcohol abuse and insecure attachment is mediated by coping drinking motives (McNally et al., 2003; Kassel et al., 2007). Another possible explanation lies within the distinction between state an d trait attachment, which is that state attachment is less persistent than trait attachment. It may be that insecure attachment is only related to coping drinking motives after repeatedly encountering situations of distress, applying an insecure attachment schema, and seeking comfort in alcohol rather than significant others. A key aspect of this explanation is repetition involved; the individual repeatedly applies the insecure attachment schema in times of distress, which repeatedly discourages the individ ual from seeking comfort in others, resulting in the individual choosing a more maladaptive coping mechanism such as alcohol consumption. After regularly not seeing others as sources of comfort, and repeatedly turning to alcohol in an attempt to cope with distress, the individual may then be motivated to drink in an attempt to cope with negative affect. The experimental method employed in the current study would not have accounted for this sort of causal relationship, which could explain the lack of signifi cant differences between groups. To further illustrate this explanation, if inducing state attachment directly affected levels of endorsing coping drinking motives, it would imply that drinking coping motives are either incorporated in the attachment schema, or that these motives are acti vated by the activation of insecure attachment schemas. In this way, state attachment activation would directly cause activation of the coping alcohol use script. The results suggest that this is not the case, but the results cannot be generalized to trait attachment, and the more indirect causal connection previously explained. Therefore, it seems most likely that insecure attachment only causes the adherence to coping drinking motives after the

PAGE 51

46 individual regularly applies an insecure attachment schema to stressful situations, and is consuming alcoholic beverages. This would imply that only some individuals with insecure attachment turn to alcohol consumption as a coping tool and that this is a result of seeking comfort in something that is not another in the social environment. This explanation seems like the most complete account to resolve the discrepancy between the current findings and findings of previous research in te rms of the association between coping drinking motives and insecure attachment representations (McNally et al., 2003; Molnar et al., 2010; Kassel et al., 2007; Doumas et al., 2007). Experimental Limitations Although it is possible to interpret these result s as having theoretical implications, it is possible that experimental limitations are largely responsible for these results. First, the experiment took place online, which comes with it a certain loss of experimental control. Second, it is possible that t he prime was simply not powerful enough. Third, there may have been a better way to operationalize the dependent variable, especially in terms of coping drinking motives that would have been more susceptible to the prime. Although these limitations may hav e compromised the integrity of the results, they provide suggestions that could possibly improve the state of future research with a similar aim. First and foremost, the disadvantages of collecting data online, especially in terms of experimental methodolo gy, should be addressed. Online data collection, while convenient, provides l ittle experimental control, and so it is possible that a number of

PAGE 52

47 things could have compromised the experiment. For example, primes generally have a limited time window of effect and unlike in the laboratory, there was no way to ensure that participants completed the depe ndent measures in the time when the prime could have had an effect. Also, it is possible that participants were distracted by any number of things between the ma nipulation and the dependent measures, leaving the prime ineffective. Future research should consider collecting data in a more formal laboratory setting, where the experimenter can have greater control over confounding variables. The next major limitatio n in the research is the possibility that the type of prime used in the current study was simply not strong enough, and therefore did not have the desired effect. The priming exercis e had been previously shown to a ffect state attachment schemas (Finkel et al., 2007), but there was no manipulation check in the present experiment. Without a manipulation check it is impossible to know if the prime had any demonstrable effect. A sub tle priming exercise was chosen for this experiment in an attempt to conceal the true nature of the experiment, but a more obvious prime may have been a more effective way to elicit changes in attachment schemas, and related schemas like coping drinking motiv es and anxiety sensitivity. Future research should consider a 2 (priming type) x 3 (state attachment schema) design, to test the difference between using subliminal and supraliminal primes in terms of state attachment. Another possible limitation is the na ture of the dependent variables, specifically the Modified DMQ R. This measure asks participants to rate how often each motive is the reason for their drinking, which is in some ways a measure of past behavior. Another way of conceptualizing drinking motiv es would be to ask participants how they would cope with some stressful event. This is similar to the design employed by Pierce and

PAGE 53

48 Lydon (1998), where they primed attachment, and among other things, asked participants hypothetically how they would cope wi th an unwanted pregnancy. A similar design could be used, and one of the options could be alcohol use. This design could reveal that individuals primed with different attachment states may be more or less likely to use alcohol as a coping mechanism. Future research may benefit from operationalizing coping oriented alcohol use in this way. Summary In short, the current research was conducted in an attempt to clarify the possible role of attachment in the development of AS and coping drinking motives. The imp ortance of this pursuit lays in their association with anxiety disorders, alcohol use disorders, and comorbid anxiety and alcohol use disorders (Howell et al., 2010; Schmidt, et al., 2007; Buckner & Heimberg, 2010; Ham et al., 2009). The present findings s uggest that at least state attachment does not directly affect AS and drinking motives, but the possibility remains that trait attachment is indeed a partial cause of these factors. Future research should address this possibility, possibly by using a longi tudinal design to see if attachment insecurity truly precedes the onset of coping drinking motives and high levels of AS. A better understanding of the development of AS and coping drinking motives could improve treatment and prevention efforts for anxiety disorders, alcohol use disorders, and comorbid anxiety and alcohol use disorders.

PAGE 54

49 References American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4ted., text rev.). Washington, DC: Author. Ainsworth, M. S. & Bowlby, J. (1991). An ethological approach to personality development. The American Psychologist, 46 (4), 333 341. Baldwin, M, Keelan J., Fehr, B., Enns, V. & Koh Rangarajoo, E. (1996). Social cognitive conceptualization of attachment working models: Avail ability and accessibility effects. Journal of Personality and Social Psychology, 71 (1), 94 109. Bartholomew, K., Horowitz, L. (1991). Attachment styles among young adults: A test of a four category model. Journal of Personality and Social Psychology, 61 (2), 226 244. Bartz, J. A. & Lydon, J. E. (2004). Close relationships and the working self concept: Implicit and explicit effects of priming attachment on agency and communion. Personality Social Psychology Bulletin, 30 (11), 1389 1401. DOI: 10.1177/0146167 204264245. Berman, N., Wheaton, M., McGrath, P., Abramowitz, J. (2010). Predicting anxiety: The role of experiential avoidance and anxiety sensitivity. Journal of Anxiety Disorders, 24 (1), 109 113. DOI: 10.1016/j.janxdis.2009.09.005 Buckner, J., Heimberg, R. (2010). Drinking behaviors in social situations account for alcohol related problems among socially anxious individuals. Psychology of Addictive Behaviors, 24 (4), 640 648. DOI: 10.1037/a0020968

PAGE 55

50 Carnelley, K. B. & Rowe, A. C. (2007). Repeated priming of attachment security influences later views of self and relationships. Personal Relationships, 14 (2), 307 320. DOI:10.1111/j.1475 6811.2007.00156.x Carnelley, K. B. & Rowe, A. C. (2010). Priming a sense of security: What goes through Journal of Social and Personal Relationships, 27 (2), 253 261. DOI: 10.1177/0265407509360901 Cassidy, J., Shaver, P., Mikulincer, M., & Lavy S. (2009). Experimentally induced security influences responses to psychological pain. Journal of Social and Clinical Psychology, 28 (4), 463 478. DeHaas, R. A. B., Calamari, J. E., & Bair, J. P. (2002). Anxiety sensitivity and the situational antecedents to drug and alcohol use: An evaluation of anxiety patients with substance use disorders. Cognitive Therapy and Research 26 (3), 335. DOI: 10.1023/A: 1016076911164. Doumas, D. M., Blasey, C. M., & Mitchell, S. (2007). Adult attachment, emotional distress, and interpersonal problems in alcohol and drug dependency treatment. Alcoholism Treatment Quarterly, 24 (4), 41. DOI: 10 .1300/J020v24n04_04. Finkel, E. J., Burnette, J. L., & Scissors, L. E., (2007). Vengefully ever after: Destiny beliefs, state attachment anxiety, and forgiveness. Journal of Personality and Social Psychology, 92 (5), 871 886. DOI:10.1037/0022 3514.92.5.871.

PAGE 56

51 Goldstein, A. L. & Flett, G. L. (2009). Personality, alcohol use, and drinking motives: A comparison of independent and combined internal drinking motives groups. Behavior Modification, 33 (2), 182 198. DOI: 10.1177/0145445508322920. Grant, V. V., Stewart, Psychometric evaluation of the five factor modified drinking motives questionnaire -revised in undergraduates. Addictive Behaviors, 32 (11), 2611 2632. DOI:10.1016/j.addbeh.2007.07.00 Ham, L., Zamboanga, B., Bacon, A., Garcia, T. (2009). Drinking motives as mediators of social anxiety and hazardous drinking among college students. Cognitive Behaviour Therapy, 38 (3), 133 145. DOI:10.1080/16506070802610889 Hazan, C. & Shaver, P. (1987). Romantic l ove conceptualized as an attachment process. Journal of Personality and Social Psychology, 52 (3), 511 524. Howell, A. N., Leyro, T. M., Hogan, J., Buckner, J. D., & Zvolensky, M. J. (2010). Anxiety sensitivity, distress tolerance, and discomfort intoleranc e in relation to coping and conformity motives for alcohol use and alcohol use problems among young adult drinkers. Addictive Behaviors, 35 (12), 1144 1147. DOI:10.1016/j.addbeh.2010.07.003 Kassel, J. D., Wardle, M., & Roberts, J. E. (2007). Adult attachmen t security and college student substance use. Addictive Behaviors, 32 (6), 1164 1176. DOI:10.1016/j.addbeh.2006.08.005

PAGE 57

52 McNally, A. M., Palfai, T. P., Levine, R. V., & Moore, B. M. (2003). Attachment dimensions and drinking related problems among young adult s the meditational role of coping motives. Addictive Behaviors, 28 (6), 1115 1127. DOI: 10.1016/S0306 4603(02)00224 1 Molnar, D. S., Sadava, S. W., DeCourville, N. H., Perrier, C. P.K. (2010). Attachment, motivations, and alcohol: Testing a dual path model of high risk drinking and adverse consequences in transitional clinical and student samples. Canadian Journal of Behavioural Science, 42 (1), 1 13. DOI: 10.1177/0265407599162002 Petrakis, I., Gonzalez, G., Rosenheck, R., Krystal, J. (2002). Comorbidity of a lcoholism and psychiatric disorders: An overview. Alcohol Research and Health, 26 (2), 81 89. Pierce, T. & Lydon, J. (1998). Priming relational schemas: Effects of contextually activated and chronically accessible interpersonal expectations on responses to a stressful event. Journal of Personality and Social Psychology, 75 (6), 1441 1448. 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), 1 8. DOI: 10.1016/0005 7967(86)90143 9 Reiss, S. (1987). Theoretical perspectives on the fear of anxiety. Clinical Psychology Review, 7 (6), 585 596. DOI: 10.1016/0272 7358(87)90007 9 Schmidt, N. B., Buckner, J. D., & Keough, M. E. (2007). Anxiety sensitivity as a prospective predictor of alcohol use disorders. Behavior Modification, 31 (2), 202 219. DOI: 10.1177/0145445506297019

PAGE 58

53 Schmidt, N., Mitchell, M., Richey, J. (2008). Anxiety sensitivity as an incremental predictor of later anxiety symptoms a nd syndromes. Comprehensive Psychiatry,49 (4), 407 412. DOI: 10.1016/j.comppsych.2007.12.004 Smith, J., Book, S. (2010). Comorbidity of generalized anxiety disorder and alcohol use disorders among individuals seeking outpatient substance abuse treatment. Ad dictive Behaviors, 35 (1), 42 45. DOI: 10.1016/j.addbeh.2009.07.002 Sutin, A. R. & Gillath, O. (2009). Autobiographical memory phenomenology and content mediate attachment style and psychological distress. Journal of Counseling Psychology, 56 (3), 351 364. DOI: 10.1037/a0014917 Swendsen, J., Conway, K., Degenhardt, L., Glantz, M., Jin, R., Merikangas, K., Sampson, N., Kessler, R. (2010). Mental disorders as risk factors for substance use, abuse and dependence: Results from the 10 year follow up of the nation al comorbidity survey. Addiction, 105 (6), 1117 1128. DOI: 10.1111/j.1360 0443.2010.02902.x Viana, A. G. & Rabian, B. (2008). Perceived attachment: Relations to anxiety sensitivity, worry, and GAD symptoms. Behaviour Research and Therapy, 46 (6), 737 747. DO I:10.1016/j.brat.2008.03.002 Vungkhanching, M., Sher, K., Jackson, K. M., & Parra, G. R. (2004). Relation of attachment style to family history of alcoholism and alcohol use disorders in early adulthood. Drug and Alcohol Dependence, 75 (1), 47 53. DOI:10.1 016/j.drugalcdep.2004.01.013

PAGE 59

54 Watt, M. C., McWilliams, L. A., & Campbell, A. G. (2005). Relations between anxiety sensitivity and attachment style dimensions. Journal of Psychopathology and Behavioral Assessment, 27 (3), 191 200. DOI:10.1007/s10862 005 0635 5. Weems, C. F., Berman, S. L., Silverman, W. K., & Rodriguez, E. T. (2002). The relation between anxiety sensitivity and attachment style in adolescence and early adulthood. Journal of Psychopathology and Behavioral Assessment, 24 (3), 159 168. DOI: 10.102 3/A:1016058600416. Zinbarg, R. E., Barlow, D. H., & Brown, T. A. (1997). Hierarchical structure and general factor saturation of the anxiety sensitivity index: Evidence and implications. Psychological Assessment, 9 (3), 277 284.

PAGE 60

55 Table 1 Overall Descriptive Statistics for All Dependent Measures Variable N Mean Standard Deviation ASI Psychological 61 4.33 3.72 ASI Physical 58 10.71 6.31 ASI Social 61 6.82 2.56 ASI Total 61 21.85 10.54 DMQ Combined Cope 58 21.98 10.34 DMQ Coping Depression 58 13.47 6.95 DMQ Coping Anxiety 62 8.44 3.92

PAGE 61

56 Table 2 Descriptive Statistics for All Dependent Measures by Group Secure Control Insecure Variable N M SD N M SD N Mean SD ASI Psychological 17 5.00 3.33 23 3.83 3.90 21 4.33 3.89 ASI Physical 15 11.66 5.87 22 10.55 6.14 21 10.19 6.98 ASI Scoial 17 6.47 2.87 23 6.96 2.40 21 6.95 2.56 ASI Total 17 23.24 10.27 23 21.17 10.49 21 21.48 11.19 DMQ Combined Cope 17 24.59 12.23 21 21.57 11.08 20 20.20 7.49 DMQ Coping Depression 17 14.94 8.62 21 13.14 7.16 20 12.55 5.05 DMQ Coping Anxiety 17 9.65 4.12 22 7.64 3.62 23 8.30 4.00

PAGE 62

57 Table 3 MANOVA with Priming Group as the Independent Variable df F R Squared p ASI Total 2, 60 0.25 0.0081 0.7833 ASI Psy 2, 60 0.23 0.0074 0.7987 ASI Phys 2, 60 0.47 0.0153 0.6292 ASI Soc 2, 60 0.43 0.0141 0.6522 DMQ cope anx. 2, 60 1.16 0.0372 0.3203 DMQ Cope dep. 2, 60 0.38 0.0123 0.6878 DMQ Cope 2, 60 0.7 0.0227 0.5011

PAGE 63

58 Appendix A Insecure Attachment Prime

PAGE 64

59 Appendix B Secure Attachment Prime

PAGE 65

60 Appendix C Control Prime

PAGE 66

61 Appendix D Anxiety Sensitivity Index

PAGE 67

62 Appendix E Modified Drinking Motives Questionnaire Revised

PAGE 68

63 Modified Drinking Motives Questionnaire Revised (cont.)

PAGE 69

64 Appendix F Demographic Questions