Why Do Women with Alcohol Problems Binge Eat?

Why Do Women with Alcohol Problems Binge Eat? Exploring Connections between Binge Eating and Heavy Drinking in Women Receiving Treatment for Alcohol Problems SHERRY H. STEWART, CATRINA G. BROWN, & KRISTINA DEVOULYTE Dalhousie University, Canada JENNIFER THEAKSTON University of Waterloo, Canada SARAH E. LARSEN St Mary s University, Canada ACKNOWLEDGEMENTS. This research was supported through a generous grant from the Nova Scotia Health Research Foundation to the first and second authors. The authors would like to acknowledge the assistance of Health Services Director Tom Payette (Capital District Health Authority), Program Managers Paul Helwig (Core), Colleen Phillips (Detoxification) and Jean McClelland (Matrix), as well as the service providers at each of these programs for their assistance in recruiting participants for the study. We would also like to thank K. Katina Garduno for her assistance in data entry. And of course, we extend our thanks to the women receiving services at these programs for their participation in our study. COMPETING INTERESTS: None declared. ADDRESS. Correspondence should be directed to: SHERRY H. STEWART, Department of Psychology, Dalhousie University, Life Sciences Centre, 1355 Oxford Street, Halifax, Nova Scotia, B3H 4J1, Canada. [email: sherry.stewart@dal.ca] Journal of Health Psychology Copyright 2006 SAGE Publications London, Thousand Oaks and New Delhi, www.sagepublications.com Vol 11(3) 409 425 DOI: 10.1177/1359105306063313 Abstract Questionnaires assessing heavy drinking and binge eating were administered to 58 women with alcohol problems. A sub-sample of the binge-eaters then participated in qualitative interviews about their perceptions of the connections between their two problems. Seventy-one percent self-identified as binge-eaters with most reporting severe binge eating. Binge-eaters were younger, more frequent drinkers and drank more often for emotional relief than non-binge-eaters. Binge eating and heavy drinking appeared to serve similar functions in a given client (i.e. emotional relief or reward functions). We discuss implications of the findings for the development of better treatments for women struggling with both health issues. Keywords alcohol problems binge eating co-morbidity motives women 409

JOURNAL OF HEALTH PSYCHOLOGY 11(3) EATING disorders and alcohol problems are significant health issues facing women today. Much research attests to the fact that they are commonly co-occurring problems (see Holderness, Brooks-Gunn, & Warren, 1994; Stewart & Brown, in press; Wilson, 1993a for reviews). Although the precise rates of substance use problems in eating disorder samples have varied across studies, likely due to factors such as the manner in which substance problems were assessed and the particular substance(s) studied, one consistent finding is elevated rates of alcohol problems in eating disordered samples (Wilson, 1993a). Typically, the greatest association between alcohol problems and eating disorders has been found among women with bulimia and women with the binge eating/purging subtype of anorexia, with up to 55 percent of women presenting for treatment of these eating disorders also displaying alcohol problems (Wilson, 1993a). High rates of eating disorders, particularly those involving binge eating, have likewise been reported among groups of women with alcohol problems. Approximately 30 40 percent of women with alcohol problems report a history of eating disorders (Taylor, Peveler, Hibbert, & Fairburn, 1993) a substantially higher rate than the 1 3 percent prevalence among women in the general population (APA, 1994). The co-occurrence of eating disorders and alcohol problems does not appear to be limited to clinical samples. Community-based studies also support the co-existence of eating disorders (particularly those involving binge eating) and alcohol problems in women (e.g. Kendler et al., 1991). Such findings rule out the possibility that the elevated co-occurrence of eating disorders and alcohol problems among women seeking treatment are simply an artefact of Berkson s bias (i.e. the increased tendency for those with more than one problem to seek treatment; see Stewart & Brown, in press). An important limitation of the majority of studies on the co-prevalence of eating and alcohol problems in women has been the tendency to examine the overlap at the diagnostic level (e.g. co-morbidity of bulimia nervosa and alcohol abuse) rather than at the symptom level (e.g. co-occurrence of binge eating and heavy drinking). Persons (1986) argues that research efforts to understand the nature of the processes underlying psychological phenomena such as binge eating and heavy drinking will be more successful if the phenomena themselves are studied directly than if diagnostic categories are studied, and that the symptoms approach is better suited to the development of theory. Research on the overlap of eating and alcohol problems in women would benefit substantially by extending the focus of scrutiny from establishing rates of co-occurring diagnoses, to investigating mechanisms to account for the overlap of specific features of these problems such as binge eating and heavy drinking behaviors. Theories abound that attempt to explain the co-occurrence of bulimia and alcohol problems in women (see Stewart & Brown, in press). Krahn (1991) proposes three general mechanisms for co-prevalence. First, alcohol problems may cause bulimia (e.g. the disinhibiting effects of alcohol intoxication may increase the likelihood of a woman succumbing to binge eating; Polivy & Herman, 1976). Second, bulimia may cause alcohol problems (e.g. the guilt and negative emotions that follow binge eating episodes may prompt self-medication with alcohol; Wiederman & Pryor, 1996a, 1996b). Finally, both bulimia and alcohol problems may be different expressions of the same underlying cause. Several theories have been advanced to explain how both eating and alcohol problems in women may be different expressions of the same underlying problem (see review by Stewart & Brown, in press). Two prominent theories are those claiming that binge eating and heavy drinking share similar rewarding properties among those deprived of the substance in question, and those claiming that both addictive behaviors share similar emotional relief functions. Each of these theories will be reviewed in turn. Emotional reward A good deal of research shows that women with eating disorders show a high propensity toward excessive use of weight control strategies including chronic dieting (see review by Polivy & Herman, 1985). Mitchell, Hatsukami, Eckert and Pyle (1985) argue that this dieting behavior may accentuate their predisposition not only to binge eating, but also to heavy drinking. Food deprivation, it is theorized, often leads to desire 410

STEWART ET AL.: BINGE EATING AND HEAVY DRINKING for high calorie, sweet substances including alcohol (Mitchell et al., 1985). Based on animal studies (e.g. Rodgers, McClearn, Bennett, & Herbert, 1963), it has been asserted that this desire may lead to both binge eating and excessive alcohol use. This latter explanation is referred to as reciprocal reinforcement (see Bulik et al., 1992; see also Krahn, 1991). One symptom of eating disorders that may be relevant to this mechanism is dietary restraint the tendency to chronically restrict dietary intake in attempts to control body weight and shape (Polivy & Herman, 1993). Several researchers have found a positive relationship between degree of dietary restraint and selfreported levels of alcohol consumption (e.g. Lavik, Clausen, & Pedersen, 1991; Stewart, Angelopoulos, Baker, & Boland, 2000a; Xinaris & Boland, 1990). Stewart and Samoluk (1997) found that, relative to women low in restraint, restrained eating women were more easily attracted not only to food cues, but also to alcohol cues. However, as increased rates of alcohol problems have not been observed among women with the restricting subtype of anorexia (Wilson, 1993a), it is clear that other factors in addition to dietary restraint must also contribute to the overlap between eating disorders and alcohol problems. In contrast to anorexic behavior, which pivots on deprivation, self-denial and restraint (Brown, 1993; Orbach, 1986), bulimia reflects cyclical capitulation to desire. Bulimic women tend to struggle between regulating, controlling or disciplining their behaviors and surrendering to their needs and desires. When women struggling between discipline and desire surrender to desire or want, binge eating may occur as a way to satisfy that want and to gain emotional reward. A similar argument has been made regarding the cycle between drinking restraint and excessive use of alcohol (e.g. Collins, George, & Lapp, 1989; Stewart & Chambers, 2000). This suggests the possibility that heavy drinking may serve a similar function as binge eating, in allowing a woman to capitulate to the reward or temptation that is represented by the forbidden object of desire (food or alcohol, respectively) (see Francis, Stewart, & Hounsell, 1997; Israeli & Stewart, 2001 for reviews of the concept of forbidden foods for chronic dieters). Efforts at self-restraint and restriction of either food or alcohol (typical of a dichotomous, all-or-nothing approach) may set women up to binge on the forbidden substance through the common mechanism of the cycle of discipline and capitulation to desire and reward (see Brown, 2001). Emotional relief In addition to the notion that both eating and alcohol problems may emerge from a common mechanism of the cycle of restraint and capitulation to reward, both eating and alcohol problems have been linked in their separate literatures as ways of coping with negative emotions (i.e. providing emotional relief). For example, both eating problems and alcohol problems among women have been separately associated with depression (Forth-Finegan, 1991). One possible common mechanism in the case of concurrent binge eating and heavy drinking may be that both are efforts at dealing with depression. A related psychosocial theory that attempts to explain the interplay of co-occurring women s mental health issues, such as eating and alcohol problems, incorporates the notion of complex post-traumatic stress reactions following trauma exposure (Herman, 1992). Women with histories of sexual violence and childhood sexual abuse often develop problems with eating disorders and alcohol in adulthood (Briere, 1988; Briere & Zaidi, 1989; Herman, 1992). A common contributor to the cooccurrence of binge eating and heavy drinking in women may be symptoms of posttraumatic stress disorder or depression subsequent to trauma. Women with histories of trauma may attempt to self-medicate for the resultant emotional problems with heavy drinking or binge eating (see Stewart, 1996). Both binge eating and heavy drinking have been associated with a state of constriction that allows for the temporary management of one s internal emotional life. The binge eating and heavy drinking can produce, for instance, numbing, tension relief, calming and a containment of difficult emotions. These consequences might serve a strong relief function for a traumatized, anxious or depressed woman. Indeed, many who seek the escape of constriction, but who do not easily separate themselves from difficult feelings or experiences may use alcohol 411

JOURNAL OF HEALTH PSYCHOLOGY 11(3) or excessive eating to help them achieve this valued relief (Stewart, 1996; Stewart & Israeli, 2002). The present study In the present study, we wished to study further the possibility that binge eating and heavy drinking may be so highly co-prevalent because they reflect common mechanisms involving emotional regulation: namely that both behaviors might involve the provision of emotional rewards and/or emotional relief. Filstead, Parrella and Ebbitt (1988) examined the nature of the intimate connections between binge eating and substance abuse in 54 individuals (mainly women) receiving inpatient concurrent treatment for both substance use problems (including alcohol problems) and eating disorders by focusing on whether the two problem behaviors had similar triggers. Using questionnaires assessing the situations in which binge eating and heavy drinking occur, these authors found some evidence for common triggers. For example, how often a woman drinks when in unpleasant emotional states significantly predicted how often she binge eats when experiencing unpleasant emotions as well. Similar evidence for common triggers for binge eating and heavy drinking was found for situations involving pleasant emotional states, testing personal control and interpersonal conflict. None the less, there were several limitations to the Filstead et al. (1988) results. First, the authors did not comment on the degree of relation between conceptually distinct triggers for each behavior. Our examination of their results (Flistead et al., 1988, Table 4, p. 138) suggests the potential for binge eating and heavy drinking to serve similar functions within a given individual. For example, frequency of drinking in situations involving unpleasant emotions was correlated not only with how often an individual binge eats in unpleasant emotions situations, but also how often she binge eats in situations involving conflict with others, suggesting the common motivation of emotional relief may underlie both binge eating and heavy drinking in these kinds of situations. Second, their study focused on individuals with co-existing clinical diagnoses of bulimia and alcohol abuse/dependence, all of whom were receiving treatment for both disorders, raising questions about the degree to which Filstead et al. s (1988) results can be generalized. Third, their sample included both men and women, but the groups were too small to allow for reliable gender comparisons. Given the much higher prevalence of eating problems in women (APA, 1994), it makes sense (at least at the outset) to restrict this novel area of investigation to women to reduce variability due to gender. There were three purposes of our study. First, we wished to examine the prevalence and characteristics of binge eating behaviors among a sample of women receiving treatment for an alcohol problem. Second, we wished to determine whether we could reliably distinguish women with alcohol problems who self-identify with a history of binge eating from those women with alcohol problems who do not self-identify as binge-eaters. Such information might be useful in helping service providers identify those women with alcohol problems who are at greatest risk of co-occurring binge eating problems, and might also tell us something about the factors associated with co-prevalence that might be useful in improvements to current treatments. Finally, we wanted to examine whether there are common underlying motivations for binge eating and drinking that might help explain their co-prevalence specifically, that binge eating and drinking might represent a common motivation of capitulation to emotional reward and/or that both might serve an emotional relief function (see Filstead et al., 1988). This latter purpose was achieved through the use of both quantitative, standardized questionnaires and qualitative, semi-structured interviews with women in the sample who reported co-occurring binge eating. The use of both types of methods allowed for examination of the convergence of findings across methodologies. Method Participants Participants were 58 adult women (aged 18 years or older) with alcohol problems. All were clients receiving treatment for alcohol problems at a program within Addictions Prevention and Treatment Services, Capital District Health Authority, in the Canadian province of Nova Scotia. In particular, women were recruited from a women-specific outpatient treatment 412

STEWART ET AL.: BINGE EATING AND HEAVY DRINKING program ( Matrix ) in downtown Halifax, and from two other treatment programs that accept both women and men clients. The latter two were outpatient ( Core ) and inpatient ( Detoxification ) programs, respectively, located at the Nova Scotia Hospital site in Dartmouth. Although neither of the latter two programs is specifically designed to meet the particular needs of women, the Core program does have a women s issues group as part of its treatment offerings for women clients. All of these programs operate according to the principles of harm reduction (Marlatt, 1998). In order to be eligible for participation in the study, a client had to be female, currently be in treatment in one of the above programs at the time of the study and self-identify with an alcohol problem. Those with other co-occurring substance use problems (e.g. cocaine abuse, analgesic dependence) were not excluded. Women had been in treatment for various lengths of time and were at various different stages in the treatment process when they participated. Materials We administered five standardized questionnaires to each participant, each of which was chosen for its excellent psychometric properties. The first was an author-compiled measure used to gather demographic information including age, ethnicity, employment status, education achieved (see Conrod, Pihl, Stewart, & Dongier, 2000a), referral source (i.e. specific Addiction Prevention and Treatment Services program) and frequency of typical alcohol consumption (Stewart et al., 2000a; Wechsler, Davenport, Dowdall, Moeykens, & Castillo, 1994). The next measure was the 10-item Brief Michigan Alcoholism Screening Test (B-MAST; Pokorny, Miller, & Kaplan, 1972), which was included as a validation of women s self-identification as problem drinkers. This measure is a shortened version of the original 25-item Michigan Alcoholism Screening Test (MAST; Selzer, 1971). The Brief version has been shown to perform as well as the original in distinguishing those with a known alcohol use disorder from those not known to have any alcohol problems (Pokorny et al., 1972). More recently, a brief version of the MAST has been shown to possess good psychometric properties in identifying alcohol problems in women (Scifres, 2003). The third measure was the 12-item Binge Scale (Hawkins & Clement, 1980), which was used to identify binge-eaters and to assess the severity of their binge eating behaviour (defined as periods of excessive eating in a very short time ) in terms of the frequency, duration and rate of binge eating. The respondent answers items in a multiple choice format; higher scores indicate more severe binge eating behaviour. The scale has acceptable internal consistency (Cronbach s alpha =.68), high test retest reliability over one month (r =.88) and a single factor structure that accounts for 71 percent of the total variance (Hawkins & Clement, 1980). It has also been shown to have good construct validity in terms of positive correlations with restrained eating and negative correlations with self-image acceptance, particularly among women (Hawkins & Clement, 1980). The fourth measure was the 42-item Inventory of Drinking Situations (Annis, Graham, & Davis, 1987; Carrigan, Samoluk, & Stewart, 1998). It was used as the measure of typical heavy drinking situations. On this questionnaire there are three types of situations as determined through factor analysis with a previously tested sample of substance-abusing women (Stewart, Samoluk, Conrod, Pihl, & Dongier, 2000c): (1) those where heavy drinking may serve a relief function: conflict with others, unpleasant emotions and physical discomfort; (2) those where heavy drinking may serve more of a reward function: pleasant times with others, pleasant emotions and social cues to drink; and (3) those where heavy drinking may be triggered by sudden temptations : situations involving testing personal control over drinking, and those involving sudden urges and temptations to drink. These scales have been shown to possess good internal consistency (Stewart et al., 2000c). Stewart, Samoluk and MacDonald (1999) have argued that it is a useful measure for looking at the underlying motivations for heavy drinking since motivations can be inferred from the situations in which the behaviour is occurring (e.g. underlying motivation of emotional relief can be inferred when heavy drinking occurs in situations involving unpleasant emotions, conflict with others or physical discomfort) and since it does not require the respondent to have an awareness of the motivations underlying their drinking behaviour. 413

JOURNAL OF HEALTH PSYCHOLOGY 11(3) Consistent with the suggestion that drinking motives can be inferred from scores on this measure, Stewart et al. (2000c) showed that scores in the Inventory of Drinking Situations correlated in theoretically expected ways with scores on a measure specifically designed to tap motivations for drinking, in a large sample of substance-abusing women. In the present study, women completed this questionnaire in terms of heavy drinking occasions over the last year immediately prior to their entry into treatment. Ratings were made on a 1 4 relative frequency scale with respect to how often the respondent drank heavily in that particular situation (1 = never; 4 = almost always). The final measure was the Inventory of Binge Eating Situations (Baker, 1998), which is a modified version of the 42-item Inventory of Drinking Situations, and which was used as the measure of typical binge eating situations (see 100-item version by Filstead et al., 1988). The majority of Inventory of Drinking Situation items have no reference to a substance (e.g. When I felt that I had let myself down ). Thus, only a few items required rewording in construction of the Inventory of Binge Eating Situations (e.g. When I suddenly had an urge to drink was changed to When I suddenly had an urge to eat ; Baker, 1998). Previous psychometric work on this measure has revealed acceptable to high Cronbach s alpha values for the various scales, thus indicating similarly good internal consistency to the drinking measure from which it was adapted (Baker, 1998). Moreover, Baker s (1998) work included exploratory factor analyses of this new scale that provided evidence of its structural validity. Participants completed the measure with respect to past year binge eating occasions. Women who had not self-identified as binge-eaters on the Binge Scale were instructed to skip this measure (see Baker, 1998). Procedure Women seeking treatment for an alcohol use problem at one of the previously described programs at Addiction Prevention and Treatment Services were given a brief description of the study and were extended an invitation to participate in the research. Recruitment methods included personal invitations by one of the experimenters at open groups, written letters, posters, verbal referrals by service providers (who were all made aware of the study at a program staff meeting) and word of mouth from client to client. The study description stressed that participation was completely voluntary and that all responses would remain confidential. It was made clear that their ability to obtain services for their alcohol problem was not at all dependent on their participation, and that the quality of their treatment would in no way be affected by their decision regarding participation in the present study. Willing women first signed an informed consent form and then completed the questionnaires in a group setting. At the end of the questionnaires appeared an invitation for participation in a second phase of this research which was extended to all 41 women who self-identified as binge eaters on the Binge Scale. Of the 28 women who were eligible and agreed to participate in this second portion of the study, we were able to contact and arrange individual interviews with 18. Three of these 18 women also participated in a focus group following their participation in the individual interviews. The interviews and focus group focused on these women s own perceptions of whether there was a relation between their heavy drinking and binge eating, and if so, what relation(s) they perceived to exist. The qualitative interviews and focus group were included in an attempt to enrich the quantitative data collection by providing a source of information with a deeper level of meaning, and by collecting sources of information that may not have been considered when the questionnaires were designed. More detailed results of this second phase have been reported elsewhere (Stewart & Brown, in press). In the present article, we present sample quotes from the qualitative data to provide converging evidence to the conclusions drawn on the basis of the questionnaire findings. Results Demographic characteristics are provided in Table 1. The participants ranged in age from 19 to 64 with a mean age of 38.5 years. Fifty-three participants (91%) identified as Caucasian, two as Caucasian/Native Canadian (3%), one as African Canadian (2%), one as Hispanic (2%) and one as Native Canadian (2%). The majority 414

STEWART ET AL.: BINGE EATING AND HEAVY DRINKING Table 1. Characteristics of women with alcohol problems who binge eat and who do not binge eat Non-binge-eater Binge-eater t-value (n = 17) (n = 41) [ 2 ] Age (in years) 43.35 36.54 2.76** (9.33) (8.23) Education (in years) 13.06 13.10 0.05 (2.95) (2.70) Ethnicity Caucasian (n) 15 38 [0.30] Other (n) 2 3 Employment Unemployed (n) 7 27 [3.02] Employed (n) 10 14 Program Women-specific (n) 5 14 [0.12] Other (n) 12 27 Weekly drinking frequency 3.91 5.37 2.29* (2.12) (2.24) Alcohol problems 19.71 19.92 0.11 (6.08) (7.29) Relief heavy drinking 58.16 68.35 2.16* (17.66) (15.84) Reward heavy drinking 65.34 71.47 1.19 (17.76) (18.00) Temptation heavy drinking 57.35 65.32 1.37 (21.37) (19.61) Notes: Binge-eaters identified by first item of the Binge Scale (Hawkins & Clement, 1980); Relief, Reward and Temptation heavy drinking assessed with the 42-item Inventory of Drinking Situations (Annis et al., 1987); Alcohol problems assessed with the Brief Michigan Alcoholism Screening Test (Pokorny et al., 1972); Weekly drinking frequency and demographic characteristics assessed with an author-compiled measure (see Conrod et al., 2000a); Program = specific program at Addiction Prevention and Treatment services *p <.05; **p <.01 (59%) was unemployed, 24 percent were employed on a full-time basis and 17 percent worked part-time. The participants years of education ranged from 7 18 years with an average length of education of 13.1 years. All women who self-identified as problem drinkers by appearing for the study met criteria for probable alcohol use disorder on the alcoholism screening test (i.e. scores of 6 or above on the B-MAST; Pokorny et al., 1972); the average score on this screen was 19.86 with a range of 6 29. As would be expected in an alcohol treatment sample, the women were also quite frequent drinkers. On average, women in the sample reported drinking on 4.94 occasions per week, ranging from 1 14 occasions. On the basis of responses to the Binge Scale (Hawkins & Clement, 1980), greater than twothirds of the sample (41/58 = 71%) selfidentified as binge-eaters. Attesting to the significance of this self-identification, 90 percent (n = 37) of those who identified as binge-eaters showed characteristics suggesting they were severe binge-eaters (Vanderehyden & Boland, 1987). Eighty-five percent of those reporting binge eating had been binge eating for more than 3 months; the other 15 percent had been binge eating for less than 3 months. The Binge Scale also provided additional information about binge eating characteristics such as the frequency of binge eating, experience of loss of control over eating and emotional reactions following a binge (see Table 2). With respect to core features of binge eating (Wilson, 1993b), 83 percent of the self-identified binge-eaters reported some degree of feeling out of control during binge eating, and 90 percent reported feeling depressed afterward (see Table 2). Substantial proportions of these women had tried various weight control strategies: 73 percent reported fasting; 49 percent excessive exercise; 41 percent laxatives; and 22 percent 415

JOURNAL OF HEALTH PSYCHOLOGY 11(3) Table 2. Binge-eating characteristics of women with alcohol problems who also binge eat (n = 41) Frequency Percent Binge-eating frequency a. seldom 4 10 b. once or twice a month 17 41 c. once a week 11 27 d. almost every day 8 20 e. several times each day 1 2 Length of episode a. less than 15 minutes 8 19 b. 15 minutes to 1 hour 13 32 c. 1 to 4 hours 15 37 d. more than 4 hours 5 12 Quantity of food consumed a. enough to fill me 4 10 b. until stomach feels full 11 27 c. until stomach painfully full 9 22 d. until can t eat anymore 15 36 e. until all the food is gone 2 5 Vomiting afterward a. never 16 39 b. sometimes 16 39 c. usually 7 17 d. always 2 5 Speed of eating a. more slowly than usual 3 7 b. about the same as usual 13 32 c. very rapidly 25 61 Degree of concern a. not bothered at all 6 15 b. bothers me a little bit 7 17 c. moderately concerned 14 34 d. a major concern 14 34 Loss of control a. could control eating if chose 7 17 b. have at least some control 12 29 c. completely out of control 22 54 Self-loathing afterward a. neutral/not too concerned 7 17 b. moderately upset 15 37 c. hate self 19 46 Depressed afterward a. not depressed at all 4 10 b. mildly depressed 12 29 c. moderately depressed 12 29 d. very depressed 13 32 Note: These characteristics are taken from responses to the Binge Scale (Hawkins & Clement, 1980) diuretics. The majority (59%) reported having used at least two of these strategies for controlling their body weight. We compared women with alcohol problems who reported binge eating (n = 41) with women with alcohol problems who did not binge eat (n = 17) on the various study measures (i.e. demographics, personality and drinking behavior) using a series of independent sample t-tests and chi-square analyses (see Table 1). Women with alcohol problems who binge eat were significantly younger than women who did not report binge eating, and those who reported binge eating also reported drinking more frequently on a weekly basis than those who did not binge eat. Although scores on the alcoholism screening test (B-MAST) did not differ across groups, binge-eaters scored higher on frequency of emotional relief heavy drinking on the Inventory of Drinking Situations, but not other drinking situation scales (i.e. reward or temptation heavy drinking), relative to nonbinge-eaters. There were no significant group differences in the type of Addiction Prevention and Treatment Services program from which the participants were recruited (women-specific 416

STEWART ET AL.: BINGE EATING AND HEAVY DRINKING vs other), or in their education level or employment status. Motivations for binge eating We first wished to establish the underlying, core motivations for binge eating that were contained in our measure of binge eating situations. Thus, we conducted a factor analysis on the eight scores from the Inventory of Binge Eating Situations (i.e. urges and temptations, social cues to eat, pleasant emotions, pleasant times with others, testing personal control over eating, physical discomfort, unpleasant emotions and conflict with others) to see which were most closely associated with one another. 1 Another purpose of this analysis was to explore the similarity of the factor structure for the Inventory of Binge Eating Situations to three-factor structure of the equivalent measure for drinking situations previously established among women with substance use problems (Stewart et al., 2000c). The present factor analysis of the Inventory of Binge Eating Situations revealed two factors (see Table 3). Attesting to the importance of these two factors, they together accounted for a very large proportion (73.8%) of the variance in scores on the Inventory of Binge Eating Situations. The first factor was characterized by strong contributions from the following six types of binge eating situations: urges and temptations; social cues to eat; pleasant times with others; pleasant emotions; testing personal control over eating; and physical discomfort. Thus, the first factor was labeled Reward Binge Eating because it reflected situations where binge eating could result in the addition of desired consequences (e.g. enhancement of pleasurable emotions) (see Table 3). The second factor was characterized by strong contributions from the following two types of binge eating situations: conflict with others and unpleasant emotions. Thus, the second factor was labeled Relief Binge Eating because it reflected situations where binge eating could result in the removal of undesired states (e.g. relief from negative emotions) (see Table 3). 2 These results were largely as expected based on previous findings with the measure of heavy drinking situations on which the Inventory of Binge Eating Situations was initially based (Stewart et al., 2000c). The only exception was that no separate Temptations binge eating factors emerged. Instead, reward and temptation situations jointly contributed to the first factor. This suggests that temptation reasons for binge eating hold much in common with emotional reward reasons for binge eating, at least among women with alcohol problems. The slight differences in factor structure between the binge eating situation measure and that found for the heavy drinking situations measures in previous work (e.g. Carrigan et al., 1998; Stewart et al., 2000c) argues against redundancy between the two measures despite the similar wording of many of their items. It should also be noted that the two binge eating motivations of Table 3. Results of factor analysis on eight scales from the Inventory of Binge Eating Situations (Baker, 1998): component loadings following oblique rotation Factor I Reward Factor II Relief Communality binge eating binge eating Urges and temptations.897.135.732 Social cues to eat.896.121.737 Pleasant emotions.856.067.695 Pleasant times with others.789.239.819 Testing personal control.731.192.675 Physical discomfort.632.191.526 Unpleasant emotions.070.947.852 Conflict with others.158.861.869 Percent variance accounted For prior to rotation 57.27% 16.53% Note: Salient loadings (>.600) indicated in bold. Communality refers to the proportion of variance in a given binge eating situation score (e.g. unpleasant emotions) that is explained by the two factors in combination 417

JOURNAL OF HEALTH PSYCHOLOGY 11(3) Reward and Relief were significantly intercorrelated (r =.373, p <.05). Thus, although the two motivations for binge eating are distinct, they are also inter-related. This suggests that a given woman will tend to binge eat primarily for one of these reasons, but that the more a woman binge eats for one of these reasons the more she also tends to binge eat for the other reason as well. Relations between motivations for binge eating and motivations for heavy drinking: quantitative findings Factor scores on the two binge eating motivation factors from the Inventory of Being Eating Situations factor analysis were then used as criterion variables in multiple regression analyses to determine the unique predictors of each type of binge eating (i.e. reward and relief binge eating). Demographic variables (age, ethnicity, employment status, years of education or addictions treatment program) failed to significantly correlate with either motivation for binge eating. Thus, demographic variables were excluded from further analyses. The three scores from the Inventory of Drinking Situations (i.e. Reward, Relief and Temptation heavy drinking situations) were used to predict the two motivations for binge eating in separate multiple regression analyses. This allowed us to examine the relations between motivations for binge eating and motivations for heavy drinking. The three motivations for heavy drinking together significantly predicted Reward binge eating (F (3, 357) = 6.56, p <.005). Together, the set of three reasons for heavy drinking predicted a significant proportion (34.7%) of the variance in Reward binge eating scores. Of the three motivations for heavy drinking, Reward heavy drinking emerged as the only significant independent predictor ( =.555, p <.01). This result suggests that the more that a woman engages in heavy drinking to enhance positive emotions, the more she engages in binge eating for similar reward functions. The three motivations for heavy drinking together also significantly predicted Relief binge eating (F (3, 37) = 6.67, p <.005). Together, the set of three reasons for heavy drinking predicted a significant proportion (35.1%) of the variance in Relief binge eating scores. Of the three motivations for heavy drinking, Relief heavy drinking emerged as the only significant independent predictor ( =.697, p <.001). This finding suggests that the more a woman engages in heavy drinking to cope with negative emotions, the more she engages in binge eating for similar reasons of emotional relief. Relations between motivations for binge eating and motivations for heavy drinking: qualitative findings 3 Women s narratives provided us with an indepth look into their binge eating and heavy drinking and the relations of these two behaviors. The women s descriptions of the situations in which they engage in problematic eating and drinking, as well as the emotional antecedents and the outcomes they hope to gain, largely echoed the results from the quantitative data in that two major reasons underlying both behaviors emerged in the narratives: relief from negative emotions, or enhancement of positive emotions. For example, one of our participants to whom we will refer as Charlotte commented that both binge eating and heavy drinking serve as a... way to create a feeling or take away a feeling. Many women acknowledged quite straightforwardly that they binge eat and drink heavily for the same reasons. We organize this review of the qualitative findings according to the two categories of common reasons for these behaviors that emerged in the interviews (see Stewart & Brown, in press, for more detail): emotional relief and emotional reward. Emotional relief Both binge eating and heavy drinking were often referred to in the interviews and focus group as a means of escape from painful or difficult emotions, such as depression, anxiety, shame, guilt and anger. For example, one of our participants, Chelsea, had the following to say about her drinking: I drank to kill the pain and hurt... and to hide from the world. Chelsea similarly acknowledged that [j]ust feeling really bad about yourself... would motivate her binge eating. If I am feeling really low, depressed, then I will turn to food. Common triggers for both included issues with relationships, fear of intimacy, fear of 418

STEWART ET AL.: BINGE EATING AND HEAVY DRINKING failure, feelings of low self-esteem and inadequacy, feeling hurt by others or conflict with others. For example, one of our participants, Isabelle, noted the following common antecedent for both binge eating and heavy drinking: It was usually when somebody else did something to me, if somebody hurt me. That was usually how I coped with it, was either... eating, drinking... Whatever it was I did was all how I reacted to being hurt. In addition, the respondents reported heavy drinking and binge eating in response to feeling lonely, bored or powerless (in relation to their male partners). Women repeatedly described using heavy drinking and binge eating to shut down, to dissociate, to numb themselves and to relieve tension. For example, according to Taylor : the binge eating is sedating, like the alcohol. I numb out while I am doing it... [Binge eating and heavy drinking] do the same thing... They release the pressure. I feel like a volcano is building up, and I can t handle it, and I don t know what to do. By bingeing, by drinking excessively... it s like letting some steam escape, or letting you know some... a little bit of lava out... so it doesn t explode. It s a release. It really is a release. The qualitative data thus provided a substantially enriched picture, from the women s own perspectives, of how both binge eating and heavy drinking were forms of emotional retreat and served emotional relief functions. Emotional reward Some women reported experiencing positive emotions as a consequence of binge eating and heavy drinking. For example, women described feeling happy, confident or euphoric when binge eating or heavy drinking. Some women mentioned enjoying the sugar rush from binge eating junk food and likened this to the intoxication or high they desired from heavy drinking. For example, one of the women interviewed, Claire described both binge eating and heavy drinking as follows: I find [binge eating and heavy drinking] gives me comfort and warmth. It makes me feel fulfilled. Charlotte had the following to say about her heavy drinking: It was to create feelings of happiness and joy that I couldn t find within myself... When I first start experiencing the high from drinking, it feels just great. Charlotte similarly saw binge eating as serving emotional reward functions as illustrated in the following quote: food excites me because I had so little of it growing up, being deprived of it for so many years. Again, the qualitative data provided a substantially enriched picture, from the women s own voices, of how both binge eating and heavy drinking served emotional reward functions in meeting their needs for emotional enhancement and capitulating to their desires and temptations. Discussion Our study reveals critical information about why women with alcohol problems binge eat. Previous research using the diagnostic categories approach (Persons, 1986) has shown a substantial overlap between alcohol use disorders and bulimia nervosa (see review by Wilson, 1993a). For example, Taylor et al. (1993) found that women receiving treatment for alcohol problems were more likely to experience bulimia than women in the general population. In our study, we adopted an approach alternative to the diagnostic categories approach namely we investigated the specific psychological phenomena (Persons, 1986) of binge eating and heavy drinking which are arguably behaviors at the core of bulimia and alcohol use disorders, respectively (Wilson, 1993a). In our sample of 58 women receiving treatment for alcohol problems, rates of binge eating were very high: 71 percent of these women acknowledged a history of binge eating on the self-report Binge Scale (Hawkins & Clement, 1980). This rate of binge eating is approximately double the rate of 36 percent observed in the general population of white women between 14 and 40 years old across a large number of prevalence studies using selfreport questionnaires (see review by Fairburn & Beglin, 1990). Moreover, using Vanderehyden and Boland s (1987) scheme for categorizing scores on the Binge Scale, the very large majority (90%) of our participants who acknowledged binge eating could be classified as severe binge 419

JOURNAL OF HEALTH PSYCHOLOGY 11(3) eaters. These alarming statistics suggest that women in treatment for alcohol problems at community-based service agencies for alcohol problems are at very high risk for displaying coexistent and severe problems with binge eating. Also consistent with the previous literature that restraint and discipline around food, eating, body weight and shape, contributes to binge eating (Polivy & Herman, 1985), we found very high rates of reports of involvement in restrictive/disciplined activities around controlling the body. For example, 73 percent of those acknowledging binge eating in our study reported fasting as compared to 29 percent in the general population of women; 41 percent reported use of laxatives as compared to 6 percent in the general population; and 61 percent reported vomiting as compared to 8 percent in the general population (Fairburn & Beglin, 1990). In fact, the majority of our self-identified bingeeaters (59%) reported having used at least two of these strategies for controlling body weight. We did find that women with alcohol problems who binge eat do differ from those who do not binge eat in a few ways. Specifically, the binge-eaters were found to be younger, generally more frequent drinkers and in particular drink more often for emotional relief purposes. We can only speculate about why these particular characteristics discriminate binge-eaters from non-binge-eaters. With respect to age, it could be that younger women have not had as much life experience to allow them to develop more adaptive ways of coping, making the younger women more susceptible to using binge eating as a coping strategy. Younger women may also have more severe issues that they have to cope with (e.g. dealing with young children; uncertainty about life goals and pressure to make those choices) in addition to fewer skills to deal with these issues in comparison to the older women. Alternatively, since this is not a study that follows the same women over time, it could be that these same young women will show the same high rates of binge eating when they are older. This could be true if the younger women today are more susceptible than today s older women to cultural messages regarding the importance of control of body weight/shape messages which have been identified as contributing to bulimia (Polivy & Herman, 1993). The nature of relations of binge eating with the patterns of alcohol use was informative. The finding that women who binge eat drink more frequently and drink for relief purposes more often is consistent with the idea that the coexistence of binge eating with alcohol problems represents a more severe clinical picture than alcohol problems alone (Singer, Nutter, White, & Song, 1993). In fact, emotional relief drinking has been argued to reflect a more problematic motivation for drinking than reward or temptation drinking patterns (e.g. Annis et al., 1987). Inconsistent with this argument, however, binge-eaters and non-binge-eaters did not differ in terms of alcohol problems on the B-MAST (Pokorny et al., 1972). 4 We also observed relationships between reasons for binge eating and reasons for heavy drinking among those women with alcohol problems that binge eat, suggesting underlying common mechanisms for these behaviors. According to results obtained with standardized questionnaires, those women who frequently binge eat for emotional relief purposes (e.g. when experiencing unpleasant emotions or interpersonal conflict) also tend to engage in frequent heavy drinking for similar reasons of emotional relief. Those who binge eat frequently for reasons of capitulating to desire and emotional reward (e.g. when experiencing urges and temptations to eat, social cues to eat or pleasurable emotions) also tend to engage in frequent heavy drinking for similar reasons pertaining to pursuing emotional rewards. These results suggest possible common mechanisms involving emotional self-regulation (e.g. control of negative emotions; capitulation to desire and temptation) for binge eating and heavy drinking among separate groups of women with cooccurring problems involving alcohol and binge eating. These results are very similar to those obtained by Filstead et al. (1988) among inpatients receiving simultaneous treatment for an eating and substance use disorder providing evidence that these relations extend to those women with alcohol problems whose eating difficulties may not fit traditional diagnostic categories. Our study also improves on the previous research by Filstead et al. (1988) in that: we focused on the relations between core underlying motivations for binge eating and heavy drinking (i.e. reward and relief) rather 420

STEWART ET AL.: BINGE EATING AND HEAVY DRINKING than examining relations between binge eating and heavy drinking in eight separate situations; we focused on motivations for the heavy use of alcohol in particular (reducing the confusion involved in examining motivations for use of different substances within a single study); and we limited our investigation to women. 5 The findings from the quantitative portion of our study regarding the potential common motivations for engaging in binge eating and heavy drinking are subject to the criticism that the findings may be secondary to common method variance particularly since the Inventory of Binge Eating Situations (our measure of binge eating motivations) was modeled closely after the 42-item Inventory of Drinking Situations (our measure of heavy drinking motivations). However, we have independent data from the qualitative interviews and focus groups that further attest to the common motivations that women themselves see for the two behaviors. Specifically, the qualitative data also revealed two main reasons for engaging in either behavior (namely, emotional reward and emotional relief) and many women directly acknowledged their perceptions that the two behaviors were serving similar functions for them. Thus, the qualitative data supported, and substantially enriched, the quantitative findings that binge eating and heavy drinking were associated with similar triggers and served similar purposes for many women. Several potential limitations to our study should be acknowledged. First, we used a questionnaire to establish whether or not women were binge eating; this methodology has been criticized for over-estimating the prevalence of binge eating (Fairburn & Beglin, 1990; Wilson, 1993b). An advantage of our choice of method for assessing binge eating was that it left the decision of what qualified as a binge to the client herself, thus taking women s own experience and perspective into account. And, although we allowed for the inclusion of women whose binge eating was not of clinical severity, the very large majority (90%) of self-identified binge eaters reported severe levels of binge eating behavior (Vanderehyden & Boland, 1987). Moreover, the alternative perspective to over-reporting must be considered as well. Some women in the sample may have chosen to not self-identify as binge-eaters, believing that relative to their severe problem with alcohol, their eating difficulties are not important enough to warrant reporting. A second possible limitation pertains to the potential for some sort of recruitment bias since participation in either portion of the study (quantitative or qualitative) was not mandatory. Because the study pertains to the relation of eating and alcohol problems, it is possible that more women with eating problems agreed to take part in the first portion than women without eating problems. We attempted to minimize this possibility by making it clear in our recruitment methods that we were interested in the perspectives of all women receiving treatment for alcohol problems regardless of whether or not they have any eating problems. It is also possible that the women who participated in the qualitative interviews and focus group differed from the other women reporting binge eating in some important way(s). We attempted to minimize this latter possibility by ensuring that the invitation to participate in the second portion of the study was inclusive of all women reporting binge eating (regardless of motivations or severity) and we made very active efforts to contact all women who indicated willingness to schedule them for an interview. A third possible limitation of the present article is its focus on what the two behaviors hold in common in terms of common underlying motivations, to the exclusion of what distinguishes these two behaviors (e.g. specific contexts in which each behavior occurs might differ substantially; see Birch, Stewart, & Brown, 2005). And finally, we failed to have women report on precisely when the heavy drinking and binge eating were at their greatest severity in the quantitative portion of the data collection. Thus, we cannot rule out the possibility that the high prevalence of binge eating observed might be secondary to reduced alcohol consumption during treatment (i.e. symptom substitution ). Indeed, the fact that 15 percent of the 41 binge-eaters reported, on the Binge Scale, that they had been binge eating for less than 3 months is consistent with the possibility of symptom substitution for at least some women. In fact, several of the interviews contained information that was consistent with the possibility of symptom substitution, as we describe elsewhere (Stewart & Brown, in press). 421

JOURNAL OF HEALTH PSYCHOLOGY 11(3) Despite these potential limitations, our results raise the possibility of distinct groups of women with co-existing binge eating and heavy drinking who vary according to the specific emotional self-regulation function these behaviors are serving. Our findings suggest that there are some women who both binge eat and drink heavily in order to achieve emotional relief. Our findings further suggest that there are also women who both binge eat and drink heavily primarily due to desires to achieve emotional rewards. Separate harm reduction programs should be developed to meet the needs of each type of woman (see Conrod et al., 2000b). And, given the overlap of these distinct motivations for binge eating and heavy drinking (Stewart et al., 2000b), we need to recognize that some women may benefit from both of these kinds of programs. Current programs for treating alcohol problems in the Canadian province of Nova Scotia do not explicitly recognize the high rates of binge eating among their women clients. Beyond a single session on healthy diet, even women-focused treatments like Matrix do not include specific programming around this coexisting mental health issue. This is at least partly due to the fact that addictions counselors rarely receive training in the treatment of eating disorders, and to the separation of addictions and mental health services in this province. Although the high rates and severity of binge eating observed in the present sample of women receiving treatment for alcohol problems in these programs are disturbing, the finding of common motivational mechanisms underlying binge eating and heavy drinking is encouraging from a treatment perspective. It suggests that current harm reduction approaches being used for alcohol problems might be fairly readily adapted to include focus on co-existing binge eating which is serving a similar motivational function. Treatment for alcohol problems at Addiction Prevention and Treatment Services in the Capital District Health Authority is based on a harm reduction model (Marlatt, 1998) as opposed to the traditional all-or-nothing abstinence-based methods of treatment. Solutions to bingeing, whether the binges involve food or alcohol, that focus on abstinence fail to address the basic emotional regulation functions that these behaviors are serving. In harm reduction approaches for alcohol problems, like Relapse Prevention, high-risk situations for relapse to heavy drinking are identified and the individual is assisted in developing more helpful strategies for dealing with these situations (Marlatt, 1998). This type of approach might be readily adapted to focus on both binge eating and heavy drinking according to the emotional self-regulation functions these behaviors are serving. For example, for those women where binge eating and heavy drinking involves capitulation to temptation and reward, harm reduction treatment could focus on problem solving skills to prevent impulsive responding in response to temptations and triggers for reward, development of healthier methods of recognizing, accepting and meeting emotional needs and helping the woman overcome the discipline/ desire cycle. Programming for emotional relief bingers could focus on helping these women to develop healthier methods of managing unpleasant emotions. Given the high rates of post-traumatic stress disorder secondary to trauma in women with alcohol problems who drink for emotional relief (Stewart, Conrod, Samoluk, Pihl, & Dongier, 2000c), the latter type of harm reduction treatment may need to focus on recovery from trauma and treatment of post-traumatic stress disorder if future research more firmly establishes this potential link with relief binge eating as well. It would be important once such harm reduction treatments are developed that their relative efficacy to treatment-as-usual for alcohol problems be assessed (see Conrod et al., 2000b). Notes 1. The method of factor analysis used in this study was principal components analysis. Since prior research suggests inter-correlations between binge-eating situation subscales (Baker, 1998; Filstead et al., 1988), we chose oblique rotation ( direct oblimin ) to allow for inter-correlations between the factors extracted. We determined the number of factors to extract based on the number of eigenvalues greater than 1.00. 2. In interpreting the meaning of the factors, we defined salient loadings as those greater than or equal to 0.600. There were several reasons to consider this two-factor solution a good structure for explaining the underlying, core components of 422

STEWART ET AL.: BINGE EATING AND HEAVY DRINKING binge eating motivations in our sample of women with alcohol problems. First, each of the binge eating situations loaded on at least one of the factors identified and none loaded on more than one factor. Second, each factor showed loadings from more than one binge eating situation. Finally, a substantial proportion of the variance in each of the eight binge eating situations was explained by the two factors. 3. To maintain women s confidentiality, we have used pseudonyms, chosen by the interviewed women themselves, to identify quotes from specific participants. 4. It would be useful in future research to employ a measure more sensitive to gradations of severity of alcohol problems than the B-MAST, which was actually developed as a screen for alcohol problems, rather than as an index of alcohol problem severity (Pokorny et al., 1972). 5. Our methodology also improves on that used by Filstead et al. (1988) in that we used a smaller number of binge eating and heavy drinking situation scales in our analyses with a similarly large number of participants making our results more likely to be replicated in future. References American Psychiatric Association (APA). (1994). Diagnostic and statistical manual of mental disorders, 4th edn. Washington, DC: American Psychiatric Association. Annis, H. M., Graham, J. M., & Davis, C. S. (1987). Inventory of Drinking Situations (IDS) user s guide. Toronto, Canada: Addiction Research Foundation. Baker, J. M. (1998). Binge eating and binge drinking among university women. Unpublished Master s Thesis, Department of Psychology, Queen s University, Kingston, Ontario, Canada. Birch, C. D., Stewart, S. H., & Brown, C. G. (2005). Exploring differential patterns of situational risk for binge eating and heavy drinking. Manuscript submitted for publication. Briere, J. (1988). Long-term clinical correlates of childhood sexual victimization. Annals of the New York Academy of Sciences, 528, 327 334. Briere, J., & Zaidi, L. (1989). Sexual abuse histories and sequelae in female psychiatric emergency room patients. American Journal of Psychiatry, 146, 1602 1606. Brown, C. G. (1993). The continuum: Anorexia, bulimia, and weight preoccupation. In C. Brown & K. Jasper (Eds.), Consuming passions: Feminist approaches to weight preoccupation and eating disorders (pp. 53 68). Toronto: Second Story Press. Brown, C. G. (2001). Talking body talk: An analysis of feminist therapy epistemology. Unpublished doctoral dissertation, University of Toronto, Toronto. Bulik, C. M., Sullivan, P. F., Epstein, L. H., McKee, M., Kaye, W. H., Dalh, R. E. et al. (1992). Drug use in women with anorexia and bulimia nervosa. International Journal of Eating Disorders, 11, 213 225. Carrigan, G., Samoluk, S. B., & Stewart, S. H. (1998). Examination of the short form of the Inventory of Drinking Situations (IDS-42) in a young adult university student sample. Behaviour Research and Therapy, 36, 789 807. Collins, R. L., George, W. H., & Lapp, W. M. (1989). Drinking restraint: Refinement of a construct and prediction of alcohol consumption. Cognitive Therapy and Research, 13, 423 440. Conrod, P. J., Pihl, R. O., Stewart, S. H., & Dongier, M. (2000a). Validation of a system of classifying female substance abusers on the basis of personality and motivational risk factors for substance abuse. Psychology of Addictive Behaviors, 14, 243 256. Conrod, P. J., Stewart, S. H., Pihl, R. O., Côté, S., Fontaine, V., & Dongier, M. (2000b). Efficacy of brief coping skills interventions that match different personality profiles of female substance abusers. Psychology of Addictive Behaviors, 14, 231 242. Fairburn, C. G., & Beglin, S. J. (1990). Studies of the epidemiology of bulimia nervosa. American Journal of Psychiatry, 147, 401 408. Filstead,W. J., Parrella, D. P., & Ebbitt, J. (1988). Highrisk situations for engaging in substance abuse and binge eating behaviors. Journal of Studies on Alcohol, 49, 136 141. Forth-Finegan, J. (1991). Sugar and spice and everything nice: Gender socialization and women s addiction a literature review. In C. Bepko (Ed.), Feminism and addiction (pp. 19 28). New York: Haworth Press. Francis, J., Stewart, S. H., & Hounsell, S. (1997). Dietary restraint and the selective processing forbidden and nonforbidden food words. Cognitive Therapy and Research, 21, 633 646. Hawkins, R. C., & Clement, P. F. (1980). Development and construct validation of a self-report measure of binge eating tendencies. Addictive Behaviors, 5, 219 226. Herman, J. (1992). Trauma and recovery: The aftermath of violence from domestic abuse to political terror. New York: Basic Books. Holderness, G. C., Brooks-Gunn, J., & Warren, M. P. (1994). Co-morbidity of eating disorders and substance abuse: Review of the literature. International Journal of Eating Disorders, 16, 1 34. Israeli,A. L., & Stewart, S. H. (2001). Memory bias for forbidden food cues in restrained eaters. Cognitive Therapy and Research, 25, 37 47. Kendler, K. S., MacLean, C., Neale, M., Kessler, R., Heath, A., & Eaves, L. (1991). The genetic 423

JOURNAL OF HEALTH PSYCHOLOGY 11(3) epidemiology of bulimia nervosa. American Journal of Psychiatry, 148, 1627 1637. Krahn, D. (1991). The relationship between eating disorders and substance abuse. Journal of Substance Abuse, 3, 239 253. Lavik, N. J., Clausen, S. E., & Pedersen, W. (1991). Eating behavior, drug use, psychopathology and parental bonding in adolescents in Norway. Acta Psychiatrica Scandinavia, 84, 387 390. Marlatt, G. A. (1998). Harm reduction: Programmatic strategies for managing high-risk behaviors. New York: Guilford Press. Mitchell, J., Hatsukami, D., Eckert, E., & Pyle, R. (1985). Characteristics of 275 patients with bulimia. American Journal of Psychiatry, 142, 482 485. Orbach, S. (1986). Hunger strike: The anorectic struggle as a metaphor for our age. New York: W.W. Norton. Persons, J. B. (1986). The advantages of studying psychological phenomena rather than psychiatric diagnoses. American Psychologist, 41, 1252 1260. Pokorny, A. D., Miller, B. A., & Kaplan, H. B. (1972). The brief MAST: A shortened version of the Michigan Alcoholism Screening Test. American Journal of Psychiatry, 129, 342 345. Polivy, J., & Herman, C. P. (1976). Effects of alcohol on eating behavior: Influence of mood and perceived intoxication. Journal of Abnormal Psychology, 85, 601 606. Polivy, J., & Herman, C. P. (1985). Dieting and binging: A causal analysis. American Psychologist, 40, 193 201. Polivy, J., & Herman, C. P. (1993). Etiology of binge eating: Psychological mechanisms. In C. G. Fairburn & G. T. Wilson (Eds.), Binge eating: Nature, assessment, and treatment (pp. 173 205). New York: Guilford. Rodgers, D. A., McClearn, G. E., Bennett, E. L., & Herbert, M. (1963). Alcohol preference as a function of its caloric utility in mice. Journal of Comparative Physiological Psychology, 56, 666 672. Scifres, S. L. (2003). Screening for problematic drinking in women: Validity of the Michigan Alcoholism Screening Test (MAST) [Summary]. Dissertation Abstracts International: Section B: The Sciences and Engineering, 63(8-B), 3937. Selzer, M. L. (1971). The Michigan Alcoholism Screening Test: The quest for a new diagnostic instrument. American Journal of Psychiatry, 127, 1653 1658. Singer, M., Nutter, L., White, W., & Song, L. (1993). Problem behaviors, substance abuse, and sexual abuse in psychiatrically hospitalized adolescents with bulimia nervosa. Child and Adolescent Social Work Journal, 10, 207 223. Stewart, S. H. (1996). Alcohol abuse in individuals exposed to trauma: A critical review. Psychological Bulletin, 120, 83 112. Stewart, S. H., Angelopoulos, M., Baker, J. M., & Boland, F. (2000a). Relations between dietary restraint and patterns of alcohol use in young adult women. Psychology of Addictive Behaviors, 14, 77 82. Stewart, S. H., & Brown, C. G. (in press). Relationship between eating and substance use problems among women: A critical review. In L. Greaves, N. Poole, & J. Greenbaum (Eds.), Women and substance use: Current Canadian perspectives. Toronto: Centre for Addiction and Mental Health. Stewart, S. H., & Chambers, L. (2000). Relationships between drinking motives and drinking restraint. Addictive Behaviors, 25, 269 274. Stewart, S. H., Conrod, P. J., Samoluk, S. B., Pihl, R. O., & Dongier, M. (2000b). Posttraumatic stress disorder symptoms and situation-specific drinking in women substance abusers. Alcoholism Treatment Quarterly, 18, 31 47. Stewart, S. H., & Israeli, A. L. (2002). Substance abuse and co-occurring psychiatric disorders in victims of intimate violence. In C. Wekerle & A.-M. Wall (Eds.), The violence and addiction equation: Theoretical and clinical issues in substance abuse and relationship violence (pp. 98 122). New York: Brunner-Routledge. Stewart, S. H., & Samoluk, S. B. (1997). Effects of short-term food deprivation and chronic dietary restraint on the selective processing of appetitiverelated cues. International Journal of Eating Disorders, 21, 129 135. Stewart, S. H., Samoluk, S. B., Conrod, P. J., Pihl, R. O., & Dongier, M. (2000c). Psychometric evaluation of the short form Inventory of Drinking Situations (IDS-42) in a community-recruited sample of substance abusing women. Journal of Substance Abuse, 11, 305 321. Stewart, S. H., Samoluk, S. B., & MacDonald, A. B. (1999). Anxiety sensitivity and substance use and abuse. In S. Taylor (Ed.), Anxiety sensitivity: Theory, research, and treatment of the fear of anxiety (pp. 287 319). Mahwah, NJ: Erlbaum. Taylor,A. V., Peveler, R. C., Hibbert, G. A., & Fairburn, C. G. (1993). Eating disorders among women receiving treatment for an alcohol problem. International Journal of Eating Disorders, 14, 147 151. Vanderehyden, D. A., & Boland, F. J. (1987). A comparison of normals, mild, moderate, and severe binge eaters, and binge vomiters using discriminant function analysis. International Journal of Eating Disorders, 6, 331 337. Wechsler, H., Davenport, A., Dowdall, G., Moeykens, B., & Castillo, S. (1994). Health and behavioral consequences of binge drinking in college. Journal of the American Medical Association, 272, 1672 1677. Wierderman, M. W., & Pryor, T. (1996a). Multiimpulsivity among women with bulimia nervosa. 424

STEWART ET AL.: BINGE EATING AND HEAVY DRINKING International Journal of Eating Disorders, 20, 359 365. Wierderman, M. W., & Pryor, T. (1996b). Substance use among women with eating disorders. International Journal of Eating Disorders, 20, 163 168. Wilson, G. T. (1993a). Binge eating and addictive disorders. In C. G. Fairburn & G. T. Wilson (Eds.), Binge eating: Nature, assessment, and treatment (pp. 97 122). New York: Guilford. Wilson, G. T. (1993b). Assessment of binge eating. In C. G. Fairburn & G. T. Wilson (Eds.), Binge eating: Nature, assessment, and treatment (pp. 227 249). New York: Guilford. Xinaris, S., & Boland, F. J. (1990). Disordered eating in relation to tobacco use, alcohol consumption, selfcontrol and sex-role ideology. International Journal of Eating Disorders, 9, 425 443. Author biographies SHERRY H. STEWART, PhD is Killam Professor in the Departments of Psychiatry and Psychology at Dalhousie University, and holds a prestigious Investigator Award from the Canadian Institutes of Health Research. She is well known for her research on the co-occurrence of addictive and mental health disorders. CATRINA G. BROWN, PhD is Associate Professor in the School of Social Work at Dalhousie University, and her research interests are in eating disorders, addictions and depression in women. She currently holds funding from the Nova Scotia Health Research Foundation and the Social Sciences and Humanities Research Council of Canada. JENNIFER THEAKSTON was the research assistant for the quantitative portion of this study and was involved in administering the questionnaires to the women participants. She is currently completing her graduate studies in industrial/organizational psychology at the University of Waterloo. SARAH E. LARSEN was the research assistant for the qualitative portion of this study and was involved in conducting the qualitative interviews with the sub-sample of binge-eaters. She continues to work with Drs Brown and Stewart on research pertaining to mental health and addictive disorder co-morbidity in women. KRISTINA DEVOULYTE was involved in the quantitative portion of this study as part of a comprehensive requirement for her graduate studies. She is completing her doctoral degree in clinical psychology at Dalhousie University and is currently a pre-doctoral intern at the Queen Elizabeth II Health Sciences Centre in Halifax, Nova Scotia. 425