The role of depression and dissociation in the relationship between childhood trauma and bulimic symptoms among ethnically diverse female undergraduates

Child Abuse & Neglect 30 (2006) 1161–1172

The role of depression and dissociation in the relationship between childhood trauma and bulimic symptoms among ethnically diverse female undergraduates夽,夽夽 Clarice K. Gerke ∗ , Suzanne E. Mazzeo, Wendy Kliewer Department of Psychology, P.O. Box 842018, Virginia Commonwealth University, Richmond, VA 23284-2018, USA Received 8 November 2004; received in revised form 3 March 2006; accepted 31 March 2006 Available online 5 October 2006

Abstract Objective: The goals of this study were to examine the role of dissociation and depression as possible mediators of the relationship between several forms of childhood trauma and bulimic symptomatology and to explore potential ethnic differences in these relationships. Method: Four hundred seventeen female undergraduates participated in this cross-sectional study. They completed measures of dissociative, depressive, and bulimic symptoms, and childhood trauma. Experiences of multiple forms of childhood trauma were measured, including physical abuse, sexual abuse, emotional abuse, physical neglect, and emotional neglect. However, only emotional abuse was correlated with bulimic symptoms at p < .01. Therefore, other forms of trauma were excluded from the analyses to control for Type I error. Results: Dissociation was not associated with emotional abuse after controlling for depression; therefore, tests of dissociation as a mediator were discontinued. Depression was significantly associated with emotional abuse after controlling for dissociation. Emotional abuse was significantly associated with bulimia. Finally, emotional abuse and depression together were significantly associated with bulimia after controlling for dissociation. However, emotional abuse became nonsignificant when entered with depression, indicating that depression mediated the relationship between emotional abuse and bulimic symptoms. There were no ethnic differences in this relationship. 夽

This study was based on a Master’s thesis conducted by Clarice K. Gerke. A version of this paper was presented at the 2004 International Conference on Eating Disorders. 夽夽 This research was partially supported by an award from the Health Psychology Section of Division 17 of the American Psychological Association (Gerke) and a National Institutes of Health Grant #MH-068520-02 (Mazzeo). ∗ Corresponding author. 0145-2134/$ – see front matter © 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.chiabu.2006.03.010

1162

C.K. Gerke et al. / Child Abuse & Neglect 30 (2006) 1161–1172

Conclusions: It appears that among women who have experienced childhood emotional abuse, depression is more strongly associated with unhealthy eating behaviors than is dissociation. Results also suggest that emotional abuse is a form of childhood trauma particularly relevant to bulimia. © 2006 Elsevier Ltd. All rights reserved. Keywords: Disordered eating; Ethnic differences; Depression; Dissociation; Childhood trauma

Introduction Traumatic experiences tend to have long-lasting negative effects on psychological well-being (e.g., Becker-Lausen, Sanders, & Chinsky, 1995; McCallum, Lock, Kulla, Rorty, & Wetzel, 1992; Oppenheimer, Howells, Palmer, & Chaloner, 1985). In particular, several authors have noted a link between childhood trauma and bulimic behaviors. However, there are inconsistencies across studies. For example, in their review of the literature, Wonderlich, Brewerton, Jocic, Dansky, and Abbott (1997) reported that studies using the same methodology and similar samples produced contradictory results regarding the relationship between childhood sexual abuse and bulimia. Furthermore, some research has found that rates of eating disorders are equivalent to or even less than rates of other forms of psychopathology in trauma survivors (de Groot & Rodin, 1999; Wonderlich et al., 1997). These findings have led some investigators to suggest that any relationship between childhood trauma and disordered eating is complex (e.g., Everill & Waller, 1995). The present study was conducted in order to elucidate the association between childhood trauma and bulimic behaviors. Although studies using clinical samples of women with eating disorders are vital to understanding and treating the effects of childhood trauma, it is important to extend this area of research to women who have not sought treatment. The majority of women will never develop bulimia; however, the percentage of those who engage in subthreshold abnormal eating behaviors, such as bingeing and purging, is much higher, with estimates ranging from 15 to 28% (Mintz, O’Halloran, Mulholland, & Schneider, 1997; Mulholland & Mintz, 2001). Moreover, subthreshold eating disorders are associated with significant distress (Mintz & Betz, 1988). Finally, women of color with eating disorders are less likely to seek treatment than Caucasian women (Cachelin, Rebeck, Veisel, & Striegel-Moore, 2001; Cachelin, Veisel, Barzegarnazari, & Striegel-Moore, 2000). Thus, it is particularly important to study participants who have not sought treatment to clarify the association between childhood trauma and bulimic behaviors in traditionally underserved groups. Several authors have attempted to outline possible mechanisms linking childhood trauma and eating pathology. For example, it has been suggested that among survivors of childhood trauma, self-destructive behaviors and dissociative states may facilitate coping with experiences of low self-regard and depression (Chandarana & Malla, 1989). Although ultimately harmful, it is crucial to recognize that for some trauma survivors, these behaviors are an attempt to heal and preserve the self (Molinari, 2001). Zerbe (1993) suggests that dissociation and eating disorders may co-occur in a fractured self in the absence of other symbolic or verbal modes of expressing painful affect and memories. The relief that harmful behaviors and psychological symptoms provide may explain why they are frequently difficult to abandon (Abraham & Beumont, 1982; Root, 1991). These theoretical assertions provide useful ideas regarding the development of post-trauma bulimia. However, many questions about the mechanisms through which trauma influences bulimic symptoms

C.K. Gerke et al. / Child Abuse & Neglect 30 (2006) 1161–1172

1163

remain. A few studies have specifically investigated eating pathology as an outcome of the relationship of trauma with dissociation or depression. Results of this research are mixed. Some studies have suggested that dissociation mediates the relationship between childhood trauma and eating disorder symptoms (Kent, Waller, & Dagnan, 1999; Lyubomirsky, Casper, & Sousa, 2001), while others did not provide support for this hypothesis (Reto, Dalenberg, & Coe, 1993; Rodriguez-Srednicki, 2001). Results regarding the role of depression in the relationship between childhood trauma and disordered eating are also inconsistent. Some investigators have found that depression does mediate the relationship between these variables (Casper & Lyubomirsky, 1997) while others have not (Kent et al., 1999; Lyubomirsky et al., 2001; Wonderlich et al., 2001). Differences in these studies (e.g., types of trauma, sample characteristics) may have contributed to their discrepant findings. Thus, although depressive and dissociative symptoms appear to be related to experiences of childhood trauma and bulimia, the nature of these relationships remains unclear. However, some evidence does suggest that dissociative and depressive symptoms may function as mediators. Therefore, this study was designed to clarify inconsistencies in the literature by testing whether dissociative or depressive symptoms mediate the relationship between childhood trauma and bulimic symptomatology. Forms of childhood trauma were considered separately, as combining disparate traumatic experiences may have contributed to past inconsistencies in the literature. It is also unclear whether the relationships among trauma, depressive, dissociative, and bulimic symptoms are similar in African Americans and Caucasians. The current study is the first, to the authors’ knowledge, to address this issue. The trend in past research has been to conclude that Caucasian women are at greater risk of bulimia than African American women; more recent investigations have, however, also indicated that eating disorders do exist among African Americans, and, thus, researchers should not ignore this population when studying eating behaviors (Mulholland & Mintz, 2001; Smith, 1995). Additionally, it is not clear whether differences in the associations among childhood trauma, dissociation, depression, and disordered eating exist in various ethnic groups. Because of the exploratory nature of this portion of this study and the inconsistency in past research, no specific hypotheses were proposed regarding ethnic differences. In sum, research has not clearly specified how childhood trauma, dissociation, and depression may be associated with bulimic symptoms. The goals of the current research were to examine the role of dissociation and depression as possible mediators of the relationship between several forms of childhood trauma and bulimic symptoms, and to explore potential ethnic differences in these relationships.

Method Participants This research was approved by the Virginia Commonwealth University institutional review board. Participants were undergraduate female volunteers from the introductory psychology course subject pool. All participants in the pool were informed about the study; 420 volunteered, and 417 provided usable data. Participants received course credit for their participation. Participants ranged in age from 16 to 53 years old (M = 19.9, SD = 4.7). The modal age was 18 years. The majority of participants (64%) were in their first year of college. Fifteen percent were in their second year, 11% were in their third year, 5% were in their fourth year, 1% were in their fifth year, and 5% did not report their year in school. Ethnicity was self-reported as Caucasian (58%), African American (27%), Asian American

1164

C.K. Gerke et al. / Child Abuse & Neglect 30 (2006) 1161–1172

(1%), Hispanic (9%), or multiracial/other (4%). Three participants did not indicate their ethnic background (.07%). These percentages are comparable to those reported by the university website for the 2004 incoming freshman class. Because introductory psychology is a course required and taken by students across many disciplines, this sample appears to be fairly representative of the students at this institution. Measures Childhood Trauma Questionnaire. Trauma history was assessed using the Childhood Trauma Questionnaire (CTQ; Bernstein et al., 1994). The CTQ is a 28-item, self-report measure that assesses a range of traumatic childhood experiences. Items are rated on a 5-point scale ranging from 1 (never true) to 5 (very often true). A total score is obtained by summing the individual item scores. The CTQ does not specify an age range for experiences that it measures. Rather, the directions instruct respondents to consider the degree to which each item was true during their childhood and adolescence. It is composed of six subscales: emotional abuse, physical abuse, sexual abuse, emotional neglect, physical neglect, and minimization/denial. The minimization/denial subscale was included to identify individuals responding in a socially desirable manner. Alpha coefficients in an undergraduate sample were .60 for physical neglect, .72 for sexual abuse, .78 for physical abuse, .89 for emotional abuse, and .92 for emotional neglect (Bernstein & Fink, 1998). Testretest coefficients in a clinical sample (gathered at a mean interval of 3.6 months, SD = 1.0) for the trauma subscales ranged from .79 for physical neglect to .81 for emotional neglect and sexual abuse (Bernstein et al., 1994). Evidence for the CTQ’s convergent validity includes its significant correlation with an interview measure, the Childhood Trauma Interview (Bernstein et al., 1994). Abuse subscale scores from an undergraduate sample were only modestly associated with social desirability and confirmatory factor analyses of the items from the five trauma subscales indicated that the five-factor model provided a good fit (Bernstein & Fink, 1998). In the current study, internal consistency estimates were .81 (emotional abuse), .61 (physical abuse), .95 (sexual abuse), .88 (emotional neglect), and .58 (physical neglect). Center for Epidemiological Studies Depression Scale. Depressive symptoms were assessed using the Center for Epidemiological Studies Depression Scale (CES-D; Radloff, 1977), a 20-item, self-report measure designed for use in the general population. Participants rated the frequency with which they have experienced each symptom in the past week on a 4-point scale, ranging from 0 (rarely) to 3 (all of the time). A total score is obtained by summing the individual item scores. Higher scores indicate greater depressive symptomatology. The CES-D displays convergent validity with other measures of depression (Radloff, 1977). Furthermore, there is evidence that the CES-D predicts depressive symptoms better than the Beck Depression Inventory (BDI) in college students (Santor, Zuroff, Ramsay, Cervantes, & Palacios, 1995). It also discriminates effectively between depressed and nondepressed individuals (Radloff, 1977). The CES-D has produced internally consistent scores in past research (Mazzeo & Espelage, 2002). In the current study the alpha coefficient was .92. Dissociative Experiences Scale-II. Dissociative symptoms were assessed using the Dissociative Experiences Scale-II (DES-II; Bernstein-Carlson & Putnam, 1993), a 28-item, self-report measure. The DES-II assesses disturbances of identity, memory, cognition, and feelings of depersonalization. Participants rate the percentage of time in their daily lives that such experiences are present, from 0 to 100%, using 10%

C.K. Gerke et al. / Child Abuse & Neglect 30 (2006) 1161–1172

1165

intervals. The total score is the mean of all 28 items and ranges from 0 to 100. Total scores above 30 indicate a high likelihood of dissociative identity disorder or post-traumatic stress disorder (Ross, Joshi, & Currie, 1990). Bernstein and Putnam (1986) demonstrated that the DES-II significantly differentiated between participants with and without dissociative disorder diagnoses. Over a 4 to 8-week interval, test-retest reliability was .84. The DES-II also has high internal consistency (alpha = .95) and strong convergent, discriminant, and criterion validity (Bernstein-Carlson & Putnam, 1993). The current study produced an internal consistency estimate (alpha) of .93. Bulimia Test-Revised. Bulimic behaviors were assessed using the Bulimia Test-Revised (BULIT-R), a 28-item, self-report measure (Thelen, Farmer, Wonderlich, & Smith, 1991). This revision of the original BULIT (Smith & Thelen, 1984) has manifested good predictive validity in nonclinical populations of college women (Thelen et al., 1991). Additionally, BULIT-R scores were highly correlated with frequencies of binging and purging among college women over 21 days of self-monitoring (Brelsford, Hummel, & Barrios, 1992). Each item is rated on a scale of 1–5, with the most symptomatic response receiving a score of 5. A total score is obtained by summing the individual item scores. The BULIT-R was validated in five stages on the basis of the scores of 2,477 women without bulimia and 93 bulimic women. A significant difference was found between the scores of participants who had bulimia and those who did not. In addition, Thelen et al. (1991) reported that the 2-month test-retest reliability of the instrument was .95, and subsequent research has produced further support for the BULIT-R’s psychometric properties (Welch, Thompson, & Hall, 1993). The alpha coefficient in the current study was .94. Procedure Participants were informed that the purpose of the study was to investigate the relationships between mood, health behavior, and life experiences. After providing consent, participants were administered the measures in a randomized sequence in small groups. Participants were given a written debriefing form following completion of the measures which explained study objectives and included campus resources that could help with any concerns they might have regarding their own life experiences, mood, body image, or eating habits.

Data analysis The possibility that depression or dissociation mediates the relationship between childhood trauma and bulimic behaviors was tested using regression analyses conducted according to Baron and Kenny’s (1986) guidelines. Mediation analyses were conducted in three steps. To avoid potential confounding variables, in each regression the effects of the alternate mediator were controlled by entering it in a step prior to the variable of interest. In the first regression, the mediator was regressed onto childhood trauma (the independent variable). In the second regression, bulimia (the dependent variable) was regressed onto childhood trauma. Finally, in a third regression, bulimia was regressed on both the mediator and childhood trauma. The mediator and childhood trauma were entered simultaneously as suggested by Holmbeck (1997) and Baron and Kenny (1986). Mediation is established if each of the variables is significantly associated with one another and in the third regression (1) the association between childhood trauma and bulimia is reduced, but remains statistically significant (partial

1166

C.K. Gerke et al. / Child Abuse & Neglect 30 (2006) 1161–1172

Table 1 Means and standard deviations of all measures Measure

CTQ-EA CTQ-PA CTQ-SA CTQ-EN CTQ-PN CES-D DES-II BULIT-R

Total sample

Caucasians

African Americans

M

SD

M

SD

M

SD

p

8.50 6.68 6.35 8.90 6.30 18.32 16.34 49.52

3.81 2.37 3.80 3.98 2.03 11.31 11.24 18.13

8.57 6.48 6.02 8.66 6.17 18.72 15.62 51.50

3.98 2.31 3.40 3.86 1.91 11.38 10.72 19.60

8.14 7.12 7.02 9.12 6.43 17.09 17.67 44.22

3.38 2.33 4.61 4.32 2.24 10.88 12.02 13.82

.29 .02 .04 .32 .26 .21 .11 <.001

Note: CTQ-EA, emotional abuse subscale of the childhood trauma questionnaire; CTQ-PA, physical abuse subscale of the CTQ; CTQ-SA, sexual abuse subscale of the CTQ; CTQ-EN, emotional neglect subscale of the CTQ; CTQ-PN, physical neglect subscale of the CTQ; CES-D, center for epidemiological studies depression scale; DES-II, dissociative experiences scale-II; BULIT-R, bulimia test-revised.

mediation) or (2) the association between childhood trauma and bulimia becomes nonsignificant (full mediation). Regression analyses also were used to test the possibility that racial differences exist in the associations among trauma, dissociative, depressive, and bulimic symptoms. Ethnic differences were investigated only between African American and Caucasians because the sample sizes of other ethnic groups were too small to be included in the analysis. Centered predictor variables were utilized to aid with interpretation, as suggested by Cohen, Cohen, West, and Aiken (2003). Race was dummy coded and was entered in the first step, followed by trauma, and depressive and dissociative symptoms in the second and third steps, respectively. Race × depression, race × dissociation, and race × trauma were entered in the fourth step, and race × dissociation × trauma and race × depression × trauma were entered in the final step.

Results Means and standard deviations for all measures, frequencies of reported childhood trauma, and correlations among measures are presented in Tables 1–3, respectively. The frequencies of the various forms of childhood trauma, except sexual abuse, as presented in Table 2 are quite high. This is likely due to the Table 2 Frequencies of reported childhood trauma CTQ subscale

Total

Caucasians

African Americans

Emotional abuse Physical abuse Sexual abuse Emotional neglect Physical neglect

327 (78%) 223 (53%) 78 (19%) 315 (76%) 186 (45%)

186 (77%) 114 (47%) 34 (14%) 178 (73%) 99 (41%)

88 (79%) 74 (67%) 28 (25%) 85 (77%) 52 (47%)

Note: These frequencies were calculated by counting the number of participants who endorsed at least one item in each subscale.

C.K. Gerke et al. / Child Abuse & Neglect 30 (2006) 1161–1172

1167

Table 3 Correlations for the total sample Measure 1 1 2 3 4 5 6 7 8

Emotional abuse Physical abuse Sexual abuse Emotional neglect Physical neglect Depression Dissociation Bulimic symptoms *

**

–

2 .53 –

3 **

4 **

.25 .28** –

5 **

.67 .44** .23** –

6 **

.43 .32** .18** .52** –

7 **

.29 .11* .01 .25** .21** –

8 **

.20 .05 .06 .10* .13** .40** –

.17** .07 .10* .08 .05 .40** .19** –

p < .05. p < .01.

manner in which the frequencies were calculated, that is, a participant was counted as having experienced a particular form of trauma if she rated at least one item in that subscale as rarely, sometimes, often, or very often true. African Americans reported experiencing more physical abuse (p < .05) and sexual abuse (p < .05) than Caucasians. There were no racial differences on the other CTQ subscales, depressive symptoms, or dissociative symptoms. Caucasians reported more bulimic symptoms than African Americans (p < .001). Both dissociation (r = .19, p < .01) and depression (r = .40, p < .01) were significantly associated with bulimic symptoms. To reduce the probability of Type I errors in the total sample (N = 417), only those trauma subscales that were significantly correlated with bulimic symptoms at p < .01 were used in the mediation analyses. As can be seen in Table 3, emotional abuse was the only trauma subscale which met this criterion (r = .17). Additionally, emotional abuse was significantly associated with both dissociation (r = .20, p < .01) and depression (r = .29, p < .01). In each case, the adjusted R2 was used to calculate the change in R2 . We first evaluated the potential mediating role of dissociation when depression was controlled. In the first step of this procedure, emotional abuse was entered to predict dissociation. However, emotional abuse was not associated with dissociation after controlling for depression (R2 = .01, p > .05). Because the first requirement for mediation was not met, no further analyses were performed testing dissociation as a mediator. Next, we evaluated whether depression mediated the association between emotional abuse and bulimic symptomatology. In the first step of this analysis, emotional abuse was significantly associated with depression after controlling for dissociation (R2 = .04, β = .22, p < .001). In the second step, emotional abuse accounted for a statistically significant portion of the variance in bulimia (R2 = .02, β = .14, p < .01). Finally, emotional abuse and depression together were significantly associated with bulimia after controlling for dissociation (R2 = .12, p < .001). Depression was a significant predictor (β = .37, p < .001). Emotional abuse became nonsignificant (β = .06, p = .23) when it was entered simultaneously with depression. This indicates that depression fully mediates the relationship between emotional abuse and bulimic symptomatology. Finally, we investigated potential racial differences in the relationships among childhood trauma, dissociative, depressive, and bulimic symptoms. Results indicated that there was a main effect of race on bulimia such that Caucasians reported more symptoms than did African Americans (R2 = .03, β = .19,

1168

C.K. Gerke et al. / Child Abuse & Neglect 30 (2006) 1161–1172

p < .001). However, the addition of interaction terms to the equation did not increase the variance accounted for in bulimic symptoms. This indicates that race does not moderate the relationships among childhood trauma, depressive, dissociative, and bulimic symptoms for Caucasians and African Americans.

Discussion The first aim of this study was to investigate whether dissociation or depression mediates the relationship between childhood trauma and bulimic symptoms among female undergraduates. The major finding of this research is that depressive symptoms, but not dissociative symptoms, mediated the relationship between emotional abuse and bulimic symptomatology. Second, this study explored the potential influence of race on the relationships among childhood trauma, depressive, dissociative, and bulimic symptoms. Race was not a significant moderator of relationships among these variables, suggesting there are no differences between Caucasian and African American women regarding the roles of dissociation or depression as mediators of the relationship between emotional abuse and bulimia. The present research adds to the growing body of knowledge on childhood trauma and its psychological sequelae. The development of disordered eating in the context of childhood trauma appears to be a complex process. Thus, one of the major strengths of this research is the incorporation of measures of depressive and dissociative symptoms in addition to assessing bulimia. Moreover, this research investigated the impact of potential mediating variables rather than only reporting correlations between variables. Another strength of this study is that the effects of the alternate potential mediator were controlled in each set of tests for mediation. It was interesting that the relationship between emotional abuse and dissociation became nonsignificant after controlling for depression as past research has documented an association between emotional abuse and dissociation (e.g., Dorahy, Lewis, Millar, & Gee, 2003; Kent et al., 1999; Simeon, Guralnik, Schmeidler, Sirof, & Knutelska, 2001). However, the association between depression and emotional abuse remained significant even when dissociation was controlled. It is possible that the relationship between dissociation and emotional abuse may be mediated by depression. This mediating role of depression is consistent with theories which propose that eating disorder symptoms are a coping strategy used to manage the effects of trauma, such as depression (e.g., Chandarana & Malla, 1989; Molinari, 2001; Zerbe, 1993). The current findings suggest that although dissociative symptoms are positively related to bulimia, there may be other mechanisms, such as depression, that better account for the relationship between emotional abuse and bulimic symptoms. Other research is also consistent with the conclusion that depression mediates the relationship between childhood trauma and disordered eating. For example, Mazzeo and Espelage (2002) reported that non-sexual forms of childhood trauma, including emotional abuse, were indirectly associated with disordered eating through the effects of alexithymia and depression. In contrast, Kent et al. (1999) found that dissociation and anxiety, but not depression, mediated the relationship between childhood emotional abuse and eating psychopathology. The discrepancy between these findings may be due to differences between the two measures used in these studies. Further research is needed to clarify whether the choice of self-report measure affects the outcome of research on emotional abuse, dissociation, depression, and disordered eating. Other research has focused primarily on the relationship between childhood sexual abuse and eating disorders (e.g., Miller, McCluskey-Fawcett, & Irving, 1993; Oppenheimer et al., 1985; Wonderlich et al., 1997). The present study is unique in that it demonstrates the importance of considering multiple forms of childhood abuse. Child sexual and physical abuse were not the strongest predictors of either the mediating

C.K. Gerke et al. / Child Abuse & Neglect 30 (2006) 1161–1172

1169

variable or the outcome. Instead, emotional abuse was the form of trauma most strongly associated with depression, dissociation, and bulimic symptoms. This may be explained by the possibility that emotional abuse occurs more frequently on its own than do physical or sexual abuse (Mazzeo & Espelage, 2002). In addition, emotional abuse may be uniquely damaging. For example, Hart and Brassard (1987) suggest that psychological maltreatment “clarifies and unifies the dynamics that underlie the destructive power of all forms of child abuse and neglect” (p. 161). Both of these possibilities would allow for a greater influence of this particular form of trauma on disordered eating. The findings from this study also add to the current literature on child abuse and eating disorders through the investigation of potential ethnic differences. Caucasians in this study reported more eating disorder symptoms than African Americans, but race did not act as a moderator of the relationships among emotional abuse, depression, and dissociation. It appears that although rates of bulimic symptoms may differ across ethnic groups, the mechanisms through which childhood trauma affect psychological outcomes are constant. This finding is consistent with Atlas’ et al. report that although African American female undergraduates reported fewer bulimic symptoms on the BULIT-R than did Caucasians, there were no ethnic differences in the relationships among bulimic symptoms and other risk factors, such as disinhibition and restraint (Atlas, Smith, Hohlstein, McCarthy, & Kroll, 2002). These results also support Mennen’s (1995) contention that the components of traumatic experiences may be universally distressing. Despite the contributions of this research, its findings must be considered in light of certain limitations. For example, because participants were not systematically sampled from college students, it is unclear how generalizable the results are to the overall college female population. Additionally, these results cannot be applied outside the sample without first replicating this study in women of varying ethnic groups and from different socioeconomic backgrounds. Unfortunately, other ethnic groups (e.g., Hispanic and Asian American) were not represented in large enough numbers to be included in the analyses examining group differences. Additionally, this research is cross-sectional and correlational in design. A longitudinal design would have permitted a clearer understanding of the temporal associations among trauma, depressive, dissociative and bulimic symptoms. However, an attempt was made to clarify causal ambiguity by performing the analyses and interpretations in the context of theory. Additionally, all measures used in this study were self-report. Particularly because the nature of the questions being answered was sensitive, the possibility of participants not being entirely truthful exists. Furthermore, the validity of the CTQ relies on participants’ ability to remember childhood experiences of trauma. Lastly, although mediation models in this study were statistically significant, the amount of variance accounted for in bulimic symptomatology by emotional abuse and depression after controlling for dissociation was relatively small. Future research could consider other variables (e.g., family environment, social support) that may further explain the development of bulimic symptoms post-trauma. Despite these limitations, the results of this study indicate that bulimic symptoms are complex and appear to be influenced by many factors. These findings underscore the importance of not only assessing whether or not childhood abuse occurred among patients presenting with eating pathology, but also of investigating the development of depressive or dissociative symptoms in response to this trauma. Because these results suggest that depressive symptoms partially account for the relationship between emotional abuse and bulimic symptoms, treatment of depression may contribute to the successful treatment of bulimia in women with a history of childhood trauma. Although this study was retrospective, its findings are applicable to clinical treatment of children and adolescents. If a child is identified as having experienced trauma, steps should be taken to prevent depressive symptoms, which may ultimately decrease the likelihood that she will develop bulimic symptoms.

1170

C.K. Gerke et al. / Child Abuse & Neglect 30 (2006) 1161–1172

These findings suggest several avenues for further research addressing the psychological problems associated with childhood trauma. Further research should build on the present findings by including other types of psychopathology, such as anxiety and post-traumatic stress disorder, as well as protective factors, such as social support. Potential ethnic differences in the relationship between childhood trauma and disordered eating have been largely ignored. Although this study’s findings begin to add to knowledge in this area, further research should consider the impact that ethnicity may have on post-traumatic symptom development. Child abuse has long been recognized as a contributing factor to the development of eating problems in some individuals, but it is less clear why this relationship exists or what variables, if any, account for it. The findings of this study suggest that although dissociation is related to childhood trauma and disordered eating, it does not mediate this relationship. Depression not only correlates more strongly with both emotional abuse and bulimia, but it also appears to mediate this relationship, even when controlling for the effects of dissociation. Finally, these results are consistent with past research suggesting that bulimia is more common among Caucasians than African Americans. However, it does not appear that differences between these two groups exist in regards to the relationships among emotional abuse, dissociation, depression, and bulimic symptoms. Acknowledgment We would like to thank Humberto Fabelo for his helpful comments on an earlier version of this article. References Abraham, S. F., & Beumont, P. J. V. (1982). How patients describe bulimia or binge eating. Psychological Medicine, 12, 625–635. Atlas, J. G., Smith, G. T., Hohlstein, L. A., McCarthy, D. M., & Kroll, L. S. (2002). Similarities and differences between Caucasian and African American college women on eating and dieting expectancies, bulimic symptoms, dietary restraint, and disinhibition. International Journal of Eating Disorders, 32, 326–334. Baron, R. M., & Kenny, D. A. (1986). The moderator-mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51, 1173–1182. Becker-Lausen, E., Sanders, B., & Chinsky, J. M. (1995). Mediation of abusive childhood experiences: Depression, dissociation, and negative life outcomes. American Journal of Orthopsychiatry, 65, 560–573. Bernstein, D. P., & Fink, L. (1998). Childhood trauma questionnaire: A retrospectiveself-report: Manual. San Antonio, TX: The Psychological Corporation. Bernstein, D. P., Fink, L., Handelsman, L., Foote, K., Lovejoy, M., Wenzel, K., Sapareto, E., & Ruggiero, J. (1994). Initial reliability and validity of a new retrospective measure of child abuse and neglect. American Journal of Psychiatry, 151, 1132–1136. Bernstein, E. M., & Putnam, F. W. (1986). Development, reliability, and validity of a dissociation scale. Journal of Nervous and Mental Disease, 174, 727–735. Bernstein-Carlson, E., & Putnam, F. W. (1993). An update on the dissociative experiences scale. Dissociation, 6, 16–27. Brelsford, T. N., Hummel, R. M., & Barrios, B. A. (1992). The bulimia test-revised: A psychometric investigation. Psychological Assessment, 4, 399–401. Cachelin, F. M., Rebeck, R., Veisel, C., & Striegel-Moore, R. H. (2001). Barriers to treatment for eating disorders among ethnically diverse women. International Journal of Eating Disorders, 30, 269–278. Cachelin, F. M., Veisel, C., Barzegarnazari, E., & Striegel-Moore, R. H. (2000). Disordered eating, acculturation, and treatmentseeking in a community sample of Hispanic, Asian, Black, and White women. Psychology of Women Quarterly, 24, 244–253.

C.K. Gerke et al. / Child Abuse & Neglect 30 (2006) 1161–1172

1171

Casper, R. C., & Lyubomirsky, S. (1997). Individual psychopathology relative to reports of unwanted sexual experiences as predictor of a bulimic eating pattern. International Journal of Eating Disorders, 21, 229–236. Chandarana, P., & Malla, A. (1989). Bulimia and dissociative states: A case report. Canadian Journal of Psychiatry, 34, 137–139. Cohen, J., Cohen, P., West, S. G., & Aiken, L. S. (2003). Applied multiple regression/correlation analysis for the behavioral sciences. Mahwah, NJ: Lawrence Erlbaum Associates Inc. de Groot, J., & Rodin, G. M. (1999). The relationship between eating disorders and childhood trauma. Psychiatric Annals, 29, 225–229. Dorahy, M. J., Lewis, C. A., Millar, R. G., & Gee, T. L. (2003). Predictors of nonpathological dissociation in Northern Ireland: The effects of trauma and exposure to political violence. Journal of Traumatic Stress, 16, 611–615. Everill, J. T., & Waller, G. (1995). Reported sexual abuse and eating psychopathology: A review of the evidence for a causal link. International Journal of Eating Disorders, 18, 1–11. Hart, S. N., & Brassard, M. R. (1987). A major threat to children’s mental health: Psychological maltreatment. American Psychologist, 42, 160–165. Holmbeck, G. N. (1997). Toward terminological, conceptual, and statistical clarity of mediators and moderators: Examples for the child-clinical and pediatric literature. Journal of Consulting and Clinical Psychology, 65, 599–610. Kent, A., Waller, G., & Dagnan, D. (1999). A greater role of emotional than physical or sexual abuse in predicting disordered eating attitudes: The role of mediating variables. International Journal of Eating Disorders, 25, 159–167. Lyubomirsky, S., Casper, R. C., & Sousa, L. (2001). What triggers abnormal eating in bulimic and nonbulimic women? The role of dissociative experiences, negative affect, and pychopathology. Psychology of Women Quarterly, 25, 223–232. Mazzeo, S. E., & Espelage, D. L. (2002). Association between childhood physical and emotional abuse and disordered eating behaviors in female undergraduates: An investigation of the mediating role of alexithymia and depression. Journal of Counseling Psychology, 49, 86–100. McCallum, K. E., Lock, J., Kulla, M., Rorty, M., & Wetzel, R. D. (1992). Dissociative symptoms and disorders in patients with eating disorders. Dissociation, 5, 227–235. Mennen, F. E. (1995). The relationship of race/ethnicity to symptoms in childhood sexual abuse. Child Abuse & Neglect, 19, 115–124. Miller, D. A. F., McCluskey-Fawcett, K., & Irving, L. M. (1993). The relationship between hildhood sexual abuse and subsequent onset of bulimia nervosa. Child Abuse & Neglect, 17, 305–314. Mintz, L. B., & Betz, N. E. (1988). Prevalence and correlates of eating disordered behaviors among undergraduate women. Journal of Counseling Psychology, 35, 463–471. Mintz, L. B., O’Halloran, M. S., Mulholland, A. M., & Schneider, P. A. (1997). Questionnaire for eating disorder diagnoses: Reliability and validity of operationalizing DSM-IV criteria into a self-report format. Journal of Counseling Psychology, 44, 63–79. Molinari, E. (2001). Eating disorders and sexual abuse. Eating & Weight Disorders, 6, 68–80. Mulholland, A. M., & Mintz, L. B. (2001). Prevalence of eating disorders among African-American women. Journal of Counseling Psychology, 48, 111–116. Oppenheimer, R., Howells, K., Palmer, R. L., & Chaloner, D. A. (1985). Adverse sexual experiences in childhood and clinical eating disorders: A preliminary description. Journal of Psychiatric Research, 19, 357–361. Radloff, L. S. (1977). The CES-D scale: A self-report depression scale for research in the general population. Applied Psychological Measurement, 1, 385–401. Reto, C. S., Dalenberg, C. J., & Coe, M. T. (1993). Dissociation and physical abuse as predictors of bulimic symptomatology and impulse dysregulation. Eating Disorders, 1, 226–239. Rodriguez-Srednicki, O. (2001). Childhood sexual abuse, dissociation, and adult self-destructive behavior. Journal of Child Sexual Abuse, 10, 75–90. Root, M. P. P. (1991). Persistent, disordered eating as a gender-specific, post-traumatic stress response to sexual assault. Psychotherapy, 28, 96–102. Ross, C. A., Joshi, S., & Currie, R. (1990). Dissociative experiences in the general population. American Journal of Psychiatry, 147, 1547–1552. Santor, D. A., Zuroff, D. C., Ramsay, J. O., Cervantes, P., & Palacios, J. (1995). Examining scale discriminability in the BDI and CES-D as a function of depressive severity. Psychological Assessment, 7, 131–139.

1172

C.K. Gerke et al. / Child Abuse & Neglect 30 (2006) 1161–1172

Simeon, D., Guralnik, O., Schmeidler, J., Sirof, B., & Knutelska, M. (2001). The role of childhood interpersonal trauma in depersonalization disorder. American Journal of Psychiatry, 158, 1027–1033. Smith, D. E. (1995). Binge eating in ethnic minority groups. Addictive Behaviors, 20, 695–703. Smith, M. C., & Thelen, M. H. (1984). Development and validation of a test for bulimia. Journal of Consulting and Clinical Psychology, 52, 863–872. Thelen, M. H., Farmer, J., Wonderlich, S., & Smith, M. (1991). A revision of the bulimia test: The BULIT-R. Psychological Assessment, 3, 119–124. Welch, G., Thompson, L., & Hall, A. (1993). The BULIT-R: Its reliability and clinical validity as a screening tool for DSM-III-R bulimia nervosa in a female tertiary education population. International Journal of Eating Disorders, 14, 95–105. Wonderlich, S. A., Brewerton, T. D., Jocic, Z., Dansky, B. S., & Abbott, D. W. (1997). Relationship of childhood sexual abuse and eating disorders. Journal of the American Academy of Child and Adolescent Psychiatry, 36, 1107–1115. Wonderlich, S., Crosby, R., Mitchell, J., Thompson, K., Redlin, J., Demuth, G., Smyth, J., & Haseltine, B. (2001). Pathways mediating sexual abuse and eating disturbance in children. International Journal of Eating Disorders, 29, 270–279. Zerbe, K. J. (1993). Selves that starve and suffocate: The continuum of eating disorders and dissociative phenomena. Bulletin of the Menninger Clinic, 57, 319–327.