Eating Behaviors 11 (2010) 7984

Contents lists available at ScienceDirect

Eating Behaviors

Psychosocial correlates of binge eating in Hispanic, African American, and Caucasian

women presenting for bariatric surgery
Leila Azarbad a,, Joyce Corsica a, Brian Hall a,b, Megan Hood a
a
b

Rush University Medical Center, Department of Behavioral Sciences, 1653 W. Congress Parkway, Chicago, IL, 60612, United States
Kent State University, Department of Psychology, Kent, Ohio, 44240, United States

a r t i c l e

i n f o

Article history:
Received 27 May 2009
Received in revised form 11 September 2009
Accepted 7 October 2009
Keywords:
Binge eating
BED
Hispanic
Ethnic differences
Bariatric surgery
Obesity

a b s t r a c t
Research suggests that that binge eating, stress, and depression are prevalent among individuals seeking
bariatric surgery. However, ethnic differences in the prevalence of binge eating and binge eating disorder
(BED) in this population remain unclear, as does the impact of depression and stress on any such
relationship. Further, no studies to date have examined the prevalence of binge eating in Hispanic women
presenting for bariatric surgery. This study sought to (a) compare the prevalence and severity of binge eating
symptomatology and BED diagnosis in Hispanic, African American, and Caucasian women presenting for
gastric bypass surgery, (b) examine the impact of depressive symptoms and stress on binge eating
symptomatology, and (c) investigate whether ethnicity moderated any relationship between depression,
stress, and binge eating. Results indicated that Hispanic women exhibited equal rates of binge eating
symptomatology, BED, and depressive symptomatology as African American and Caucasian women.
However, Caucasian women exhibited greater binge eating symptomatology than African American women,
and African American women endorsed greater levels of stress than Caucasian women. Across all ethnic
groups, depressive symptomatology, but not stress, signicantly predicted binge eating severity. These
ndings suggest that Hispanic women presenting for bariatric surgery report binge eating rates equivalent to
Caucasian and African American women, and that depressive symptoms are an important predictor of binge
eating in female bariatric surgery candidates across ethnic groups.
2009 Elsevier Ltd. All rights reserved.

1. Bariatric surgery and binge eating

Obesity has become a serious public health epidemic in the United
States. According to the National Health and Nutrition Examination
Survey (NHANES; Ogden et al., 2006), over 32% of American adults
over the age of 20 are obese. Ethnic differences in the prevalence rates
of obesity have consistently been found, with minority groups
exhibiting higher prevalence of obesity relative to Caucasians.
Among the adult population, it is estimated that 45% of African
Americans, 36% of Hispanics, and 30% of Caucasians are obese (Ogden
et al., 2006).
Behavioral and pharmacological weight loss interventions used
with obese individuals have demonstrated little long-term success
(Perri & Corsica, 2002). Bariatric surgery has emerged as an
increasingly accepted and successful treatment of moderate (with
medical complications) and severe obesity. Roux-en-Y gastric bypass
surgery (RYGBP) is considered the gold standard of bariatric surgery
procedures and produces greater weight losses (66% excess weight
Corresponding author. Tel.: +1 312 942 5934; fax: +1 312 942 4990.
E-mail addresses: Leila_Azarbad@rush.edu (L. Azarbad), Joyce_Corsica@rush.edu
(J. Corsica), Brian_Hall@rush.edu (B. Hall), Megan_Hood@rush.edu (M. Hood).
1471-0153/$ see front matter 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.eatbeh.2009.10.001

loss after 2 years, 50% excess weight loss after 10 years) than other
surgical weight loss procedures (Lati, Kellum, De Maria, & Sugerman,
2002).
Binge eating symptomatology is frequently present in individuals
seeking bariatric surgery (Kalarchian, Wilson, Brolin, & Bradley, 1998;
de Zwaan et al., 2003), with estimates ranging from 25 to 44% (Hsu
et al., 2002; Greenberg, Perna, Kaplan, & Sullivan, 2005). Estimates of
frank binge eating disorder (BED) in bariatric surgery candidates have
ranged from 1.4 to 49% (de Zwaan et al., 2003). Several studies have
found binge eating to be linked with poorer outcomes after surgery,
including decreased weight loss (Rowston et al., 1992) and higher
rates of post-surgical vomiting and complications (Busetto et al.,
1996). A national survey of bariatric surgery centers found that active
binge eating was considered a denite contraindication by 48% and a
possible contraindication by 40% of bariatric programs (Bauchowitz,
Azarbad, Day, & Gonder-Frederick, 2007). Similarly, Fabricatore,
Crerand, Wadden, Sarwer, and Krasucki (2006) found that eating
disorders, including binge eating, were the second greatest concern
(after depression) of mental health professionals evaluating surgery
candidates. Given the potential negative effects of binge eating on
post-surgical outcomes, identication of binge eating remains an
important component of the pre-surgical psychological evaluation.

80

L. Azarbad et al. / Eating Behaviors 11 (2010) 7984

1.1. Ethnic differences in binge eating

Studies that have examined ethnic differences in the prevalence of
binge eating and BED in non-bariatric samples have yielded mixed
ndings. Some studies report similar rates of binge eating symptomatology (Crago, Shisslak, & Estes, 1996; Striegel-Moore & Smolak,
2000; Shaw, Ramirez, Trost, Randall, & Stice, 2004) and BED (Smith,
Marcus, Lewis, Fitzgibbon, & Schreiner, 1998) across ethnic groups.
Other studies report higher rates of binge eating in Caucasians relative
to African Americans (Field, Camargo, Taylor, Berkey, & Colditz, 1999;
Gray, Ford, & Kelly, 1987). Of the few studies examining binge eating
in Hispanic populations, results have also been mixed. Hrabosky and
Grilo (2007) found few differences in binge eating and other eating
disordered behaviors between Hispanic and African American
women, whereas Fitzgibbon et al. (1998) found signicantly higher
rates of BED in Hispanic women relative to African American and
Caucasian women, even after controlling for demographic variables,
BMI, depression, and ideal body image. Only one study to date has
examined binge eating in female bariatric surgery candidates and
found no differences in the rates of binge eating or BED between
Caucasian and African American women (Mazzeo, Saunders, &
Mitchell, 2005).
No studies to date have examined binge eating in Hispanic women
presenting for bariatric surgery. A recent study found that 21% of
Spanish-speaking, Hispanic women recruited from the community
reported experiencing loss of control over eating at least once per
week (Elder, Paris, Aez, & Grilo, 2008). Furthermore, given that
Hispanic women experience higher rates of obesity relative to
Caucasian women (Ogden et al., 2006), and given the positive
relationship between weight and prevalence of BED (Smith et al.,
1998), Hispanic women may be at greater risk for binge eating.
1.2. Depression, stress, and binge eating
Numerous studies have documented the relationship between
depression and binge eating (Spitzer et al., 1993; Telch & Agras, 1994;
Wadden, Foster, Letizia, & Wilk, 1993). However, the extent to which
ethnicity may moderate the relationship between depression and
binge eating remains unclear. For example, Fitzgibbon et al. (1998)
found that depressive symptoms were associated with binge eating in
Hispanic and Caucasian but not African American women, with the
strongest association between BED and depression present in
Hispanic women. However, in their study of female bariatric surgery
candidates, Mazzeo et al. (2005) found that depression was equally
associated with binge eating for both Caucasian and African American
women. No studies to date have explored the association between
depressive symptoms and binge eating in Hispanic women presenting
for weight loss surgery.
Stress has also been implicated in the development and maintenance of binge eating in both laboratory (Gluck, Geliebter, Hung, &
Yahav, 2004; Epel, Lapidus, McEwen, & Brownell, 2001) and
naturalistic studies (Newman, O'Connor, & Conner, 2007; Waters,
Hill, & Waller, 2001). In general, individuals with binge eating
symptomatology and/or BED report increases in hunger and desire
to binge in response to stressful events (Gluck, 2006). Elevations in
cortisol, a physiological measure of the stress response, have also been
found at baseline and following stressful tasks in women with binge
eating symptoms as compared to non-bingers (Gluck et al., 2004),
strongly suggesting a role for stress in binge eating.
While available evidence implicates stress as a predictor of binge
eating, very few studies have examined possible ethnic differences in
the relationship between stress and binge eating. A recent study of
normal-weight college women found a positive relationship between
stress and binge eating in both African American and Caucasian
groups, though this relationship was stronger for Caucasian women
(Harrington, Crowther, Henrickson, & Mickelson, 2008). No studies to

date have examined possible ethnic differences in the stress-binge

eating relationship among women presenting for bariatric surgery.
The rst aim of the present study was to compare demographic
and psychosocial characteristics of Hispanic, African American, and
Caucasian women seeking gastric bypass surgery, with particular
emphasis on comparing levels of binge eating symptoms and BED
across ethnic groups. Our second aim was to examine the inuence of
depressive symptoms and stress on binge eating, and our third aim
was to investigate whether ethnicity moderated any inuence of
depression and stress on binge eating.
2. Material and methods
2.1. Participants
Participants were 404 women who presented for a psychological
evaluation prior to undergoing Roux-en-Y gastric bypass surgery at an
urban academic medical center from October 2002 to January 2009. Of
these women, 224 (55.4%) were African American, 142 (35.2%) were
Caucasian, and 38 (9.4%) were Hispanic. Mean BMI was 50.6
(SD = 9.3). Mean age of the participants was 41.5 years (SD = 10.4).
Forty percent of participants were married, 35% were single (never
married), 13% were divorced, 5% were separated (but not divorced),
4% were widowed, and 3% were cohabiting. In terms of educational
attainment, 11% had completed less than a high school education, 30%
had completed high school, 40% had completed some college, 11% had
earned a 4-year baccalaureate degree, and 8% had completed postbaccalaureate studies.
2.2. Measures
2.2.1. Psychosocial interview/demographic questionnaire
Data on height, weight, psychiatric history, and assessment of past
and current BED criteria was collected via a structured psychosocial
interview (further described in Section 2.3). Data on ethnicity,
educational attainment, and marital status was collected via a demographic questionnaire.
2.2.2. Binge Eating Scale (BES; Gormally, Black, Daston, & Rardin, 1982)
The BES is a 16-item self-report measure used to assess binge
eating cognitions and behaviors in overweight and obese individuals.
Scores between 17 and 26 indicate a mild to moderate problem, and
scores greater than or equal to 27 indicate a severe problem. The
BES has demonstrated good psychometric properties (Cronbach's
alpha = .85).
2.2.3. Beck Depression Inventory-2 (BDI-2; Beck, Steer, & Brown, 1996)
The BDI-2 is a widely used 21-item self-report measure that
assesses severity of depressive symptomatology in the past two
weeks. Total scores ranging from 0 to 13 represent normal to minimal
depressive symptoms; 14 to 19 represent mild depressive symptoms;
20 to 28 represent moderate depressive symptoms, and 29 to 63
represent severe symptoms. The BDI-2 has demonstrated high
internal consistency (coefcient alphas N.90) and moderate to strong
convergent validity with other self-report measures of depression
(rs N .50; Steer & Beck, 2000).
2.2.4. Personality Assessment Inventory (PAI; Morey, 1991)
The PAI is a self-administered, 344-item inventory of adult
personality and psychopathology. The PAI is comprised of 22 nonoverlapping scales consisting of 4 validity scales, 11 clinical scales, 5
treatment scales, and 2 interpersonal scales. Average alpha coefcients for the clinical scales in clinical populations range from .82
(Boone, 1998) to .86 (Morey, 1991), with testretest reliabilities
ranging from .80 to .90 (Morey, 1991).

L. Azarbad et al. / Eating Behaviors 11 (2010) 7984

The PAI Stress scale consists of 8 items that measure the impact of
current or recent stressors in the areas of family, health, employment,
nances, and interpersonal functioning. Items assess both chronic
psychosocial stressors (Things are not going well in my family) and
major but less frequent stressors (There have been many changes in
my life recently). A t-score of 59 or below reects stability,
predictability, and lack of major stressors; a t-score of 6069 reects
a moderate degree of stress; and a t-score of 70 or greater reveals
signicant stress that is likely having a substantial impact on an
individual (Morey, 1991).
2.3. Procedure
All patients rst met with a bariatric surgeon for an initial
consultation. Patients deemed to be medically appropriate candidates
were then referred for psychological evaluation, which consisted of
(a) a psychosocial interview, including personal and educational
backgrounds, weight trajectory, weight loss history, medical history,
adherence to medical recommendations, current health behaviors,
current eating habits, evaluation of emotional and binge eating1,
history of depressive episodes and other psychopathology, and
understanding and expectations regarding surgery, and (b) written
questionnaires assessing mood, personality, eating habits and eating
disorder symptomatology. This study was approved by the Rush
University Medical Center Institutional Review Board.
3. Results
3.1. Data preparation and statistical procedure
Participants who produced invalid proles on the Personality
Assessment Inventory (PAI) based on criteria recommended by Morey
(2007) were eliminated, such that cases were excluded if a) N5% of
questions were unanswered, b) inconsistency scores were greater
than 73T, or c) infrequency scores were greater than 75T. This resulted
in the elimination of 20 cases (4.9%), for a total N of 384.
Statistical assumptions were tested prior to interpretation of the
statistical results. All variables were normally distributed (i.e., range
of skewness and kurtosis values was within 2 to +2) except
education (K = 3.68), total depression score (K = 2.38), and largest
weight loss (K = 3.41) which were platykurtotic2. Levene's test was
used to evaluate the homogeneity of variances for the ANOVA test.
BMI, years of education, and stress all violated the assumption of
homogeneity. In order to ensure that this violation would not lead to
an over- or under-estimate of the signicance test, group variances
were examined by comparing the squared standard deviation for each
variable. The ratio in each case was less than 3:1, suggesting that the
interpretation of the F statistic could proceed normally and without
concern for Type I or Type II errors (Moore, 1995). Although the ratio
is favorable, given the unequal group sizes in our analyses, we utilized
the Welch test of equality of means and the GamesHowell post-hoc
test, as these tests account for heterogeneity of variances and unequal
group sizes. For the regression analysis, all variables were linear and
homoscedastic as evaluated by scatter plots. Independent variables
were examined for possible multicollinearity using a two-step
approach. We rst examined their bivariate correlations. We next
evaluated the tolerance (i.e., the proportion of variance not accounted
1
Participants were not screened for binge eating symptomatology prior to the
psychological evaluation.
2
Education, total depression score, and largest weight loss were all transformed.
Following transformation, the univariate skewness and kurtosis values of the study
variables indicated that all variables were normally distributed. No differences were
noted in study results when transformed variables were used, therefore the results of
the non-transformed variables are presented. This is expected given that the
introduction of bias owing to kurtotic distributions is attenuated with samples
exceeding 100 cases (Tabachnick & Fidell, 2007).

81

for by other variables in the model) and variance ination factor (VIF)
for each of the independent variables within a linear regression
framework. All correlations were below .70, tolerance scores were
high (e.g., above .60), and VIF values were low, indicating the absence
of multicollinearity in these data (Tabachnick & Fidell, 2007).
Analyses of variance (ANOVA) utilizing GamesHowell post-hoc
tests were conducted to examine differences in age, years of
education, depressive symptoms, stress, and binge eating severity
across the three ethnic groups. Chi-square analyses were used to
examine ethnic differences across the following categorical variables:
history of major depression, history of psychiatric hospitalization,
history of BED diagnosis, current BED diagnosis, and presence of
severe binge eating symptoms on the BES. A hierarchical linear
regression analysis was conducted to explore the role of various
predictors on binge eating severity. Ethnicity was dummy coded to
create two separate variables for African Americans and Hispanics
with Caucasians treated as the reference category. Consequently,
regression coefcients for the dummy variables were interpreted as
changes in the dependent variable with respect to the reference
group. The total score on the BES was regressed on relevant
demographic variables, depression symptom severity and overall
stress. We also tested whether ethnicity moderated the relationship
between depression and binge eating and stress and binge eating.
Variables were entered into steps in the following order: 1) age and
BMI, 2) ethnicity, 3) depressive symptoms and stress, and 4)
interactions between ethnicity and depression/stress. Centered
independent variables were included in the analyses, and interactions
were calculated using centered variables in order to minimize
multicollinearity (Aiken & West, 1991). Missing data were handled
by listwise deletion, resulting in slightly different sample sizes in the
analyses of the different variables.
3.2. Differences across ethnic groups
Means and standard deviations for demographic and psychosocial
characteristics of the sample are displayed in Table 1.
Results of ANOVAs indicated signicant differences across ethnic
groups for age, F(2, 97.82) = 8.90, p b .001, BMI, F(2, 103.44) = 5.88,
p = .004, largest lifetime weight loss, F(2, 89.54) = 8.02, p b .001, and
self-reported stress, F(2, 82.99) = 9.73, p b .001. Results of Games
Howell post-hoc tests revealed that Caucasian women were older
than Hispanic (p = .011) and African American (p b .001) women and
reported greater lifetime weight loss than Hispanic (p = .014) and
African American (p = .001) women. African American women had
greater BMI relative to both Caucasian (p = .021) and Hispanic
(p = .014) women and endorsed higher levels of stress than Caucasian
women (p b .001). The three groups did not differ with respect to
educational attainment, F(2, 85.75) = 2.97, p = .057.
Table 1
Demographic and psychosocial characteristics of the sample.
Variable

Age (years)
Education (years)
BMI (kg/m2)
Largest WL (lb)
BDI-2
PAI Stress (t-score)
BES

Caucasian
(n = 137)

Hispanic
(n = 35)

African
American
(n = 212)

SD

SD

SD

44.6a,b
13.8
49.0a
54.3a,b
9.6
48.1a
14.3a

10.8
2.5
8.6
31.3
8.4
9.1
8.5

40.1a
13.7
51.7a,b
42.0a
10.3
52.9a
12.0a

9.9
2.1
9.7
27.6
8.4
11.1
8.4

39.3b
12.3
47.6b
38.6b
11.1
49.1
13.0

8.8
3.4
7.2
26.9
8.3
9.5
8.7

Note. BMI = body mass index; largest WL = largest lifetime weight loss in pounds;
BDI-2 = Beck Depression Inventory-2; PAI Stress = PAI Stress subscale; BES = Binge
Eating Scale. Means in a row sharing subscripts are signicantly different, p b .05, twotailed.

82

L. Azarbad et al. / Eating Behaviors 11 (2010) 7984

Caucasian women endorsed greater incidence of one or more

lifetime depressive episodes relative to African American women, x2
(1, N = 336) = 6.922, p = .009. Specically, 49.9% of Caucasian
women met DSM-IV diagnostic criteria for at least one prior major
depressive episode, compared to 34.5% of African American women
and 48.0% of Hispanic women. Hispanic and Caucasian women did not
differ with respect to incidence of lifetime depressive episodes, x2 (1,
N = 274) = .372, p = .542. Differences in incidence of lifetime depression between Hispanic and African American women approached, but
did not reach, statistical signicance, x2 (1, N = 363) = 3.577,
p = .059. No ethnic differences were found for current depressive
symptoms (BDI-2 score), F(2, 86.29) = .514, p = .600, or history of
one or more psychiatric hospitalizations, x2 (2, N = 234) = . 906,
p = .636.
The three ethnic groups did not differ with respect to binge eating
symptomatology on the BES, F(2, 85.53) = 2.86, p = .063. However,
post-hoc comparisons revealed that Caucasians exhibited greater
binge eating symptomatology than African Americans3 (p = .045). No
differences were observed across groups with respect to incidence of
past BED diagnosis, x2 (2, N = 341) = .434, p = .605 or current BED
diagnosis, x2 (2, N = 384) = 1.290 p = .525. Current DSM-IV diagnostic
criteria for BED were met by 15.7% of Caucasians, 11.5% of African
Americans, and 11.8% of Hispanic women. Additionally, no differences
were observed with respect to classication of severe binge eating
according to the BES (BES score greater than or equal to 27), x2 (2,
N = 384) = .712, p = .701. Specically, 9.5% of Caucasian, 7.5% of
African American, and 5.7% of Hispanic women evidenced severe
binge eating symptomatology on the BES. A point-biserial correlation
revealed a signicant positive correlation between BED diagnosis and
total BES score, (r = .563, p b .01). Women who met criteria for BED
had a mean BES score of 25.2 (SD = 6.3), whereas women who did not
meet BED criteria had a mean BES score of 11.0 (SD = 7.1).
3.3. Predictors of binge eating symptomatology
Results of the hierarchical linear regression are displayed in Table 2.
Step one of the model (age and BMI) was not statistically signicant,
F(2, 341) = 2.369, p = .095. However, age emerged as a signicant
predictor of binge eating severity on this step (p = .035). Step 2 of the
model was not statistically signicant, F(4, 339) = 2.066, p = .175,
indicating that ethnicity was not a signicant predictor of binge eating symptomatology. Step three of the model (depressive symptoms
and stress) was signicant and accounted for 32.4% of the variance in
binge eating severity, F(2, 337) = 74.681, p b .001. Depressive symptoms, but not stress, were predictive of binge eating severity (p b .001
and p = .449, respectively). None of the interactions in the nal
model step were signicant (p = .412), indicating that ethnicity did
not serve as a moderator in the relationship between depressive
symptoms and binge eating or stress and binge eating.
4. Discussion
To our knowledge, this is the rst study to examine demographic
and psychosocial correlates of binge eating pathology in a group of
Hispanic women presenting for bariatric surgery. Overall, we
observed more similarities than differences among the three ethnic
groups. Demographically, Hispanic and African American women
were younger than Caucasian women. Psychologically, Hispanic
women reported equal levels of stress, depressive symptomatology,
3
Given the overall lack of signicance of the omnibus F statistic, the post-hoc
comparison should be interpreted with caution. However, this is likely owing to the
lack of signicant differences between the Hispanic and Caucasian and Hispanic and
African American groups. Given the importance of this nding to our research
hypotheses, we conducted a follow-up independent samples t-test. The results
corroborated the post-hoc test result, t(340) = 2.405, p = .017.

Table 2
Hierarchical regression analysis for predictors of binge eating symptomatology.
Variable
Step 1
Age
BMI
Step 2
African American
Hispanic
Step 3
BDI-2
PAI Stress
Step 4
African American BDI-2
Hispanic BDI-2
African American PAI Stress
Hispanic x PAI Stress

SEB

0.09
0.03

0.04
0.05

0.11*
0.03

1.81
1.92

1.00
1.79

0.11
0.06

0.59
0.04

0.06
0.05

0.57**
0.04

0.04
0.28
0.10
0.12

0.12
0.27
0.10
0.17

0.03
0.05
0.09
0.04

R2

R2
.014

.024

.010

.324**

.300**

.332**

.008

Note. BMI = body mass index. BDI-2 = Beck Depression Inventory-2.

*p b .05, two-tailed. **p b .001, two-tailed.

and binge eating symptomatology relative to the other ethnic groups,

with self-reported symptom means falling in the non-clinical range
across ethnic groups. Further research is warranted to examine
psychosocial functioning in Hispanic bariatric candidates using a
larger sample size, as well as examining postoperative outcomes in
this population.
The nding that Hispanic women experienced similar rates of
binge eating symptomatology and BED as Caucasian and African
American women contributes to a growing body of literature suggesting that ethnic minority women are equally vulnerable to disordered
eating (Lovejoy, 2001; Thompson, 1994) but may be diagnosed and
treated at lower rates than the majority population (Pike, Dohm,
Striegel-Moore, Wiley, & Fairburn, 2001). Though Hispanic women
endorsed similar rates of binge symptoms as the other ethnic groups,
our results suggest that African American women presenting for
bariatric surgery may experience fewer symptoms of binge eating than
Caucasian candidates. However, this nding should be interpreted
with caution, given that the overall omnibus F test for the three groups
was not signicant. The non-signicance of the F test was likely
inuenced by the lack of signicant differences between the Hispanic
and Caucasian groups and the Hispanic and African American groups. A
follow-up independent samples t-test conrmed the post-hoc results
that Caucasians endorsed greater binge eating symptomatology on the
BES than African Americans. Despite this statistical difference, it is
important to note that all three groups reported average binge eating
symptoms that fell in the non-clinical range.
Historically, the wide variability in the prevalence rates of binge
eating/BED in bariatric candidates has largely been attributed to
variability in the methodology used to assess binge eating symptoms.
Prevalence of severe binge eating symptoms in our sample was lower
than most, but not all, previous studies examining binge eating in
surgical candidates. Overall, approximately 8% of our total sample
endorsed severe symptoms of binge eating on the BES (9.5% of
Caucasians, 7.5% of African American women, and 5.7% of Hispanic
women). Prevalence of DSM-IV BED diagnosis was also equal across
the three groups, with approximately 15% of Caucasian, 12% of African
American, and 12% of Hispanic women meeting BED diagnostic
criteria. This is signicantly lower than the 33% and 38% of African
American and Caucasian women, respectively, who met BED criteria
in Mazzeo et al.'s (2005) study. Differences in rates of BED between
these two studies may reect differences in the methodology used to
assess BED (clinical interview plus BES vs. Questionnaire on Eating
and Weight Patterns). It is possible that the use of a structured
interview, as opposed to a written questionnaire, may lead participants to underreport BED symptoms. It is also possible that
differences in positive impression management or the nature of the
samples may have inuenced the variability in binge eating

L. Azarbad et al. / Eating Behaviors 11 (2010) 7984

symptomatology in these two studies. Nonetheless, the wide

discrepancy in published rates of BED among bariatric surgery
candidates highlights the need for more consistent methodology to
assess binge eating symptoms.
Depressive symptomatology emerged as the most robust predictor
of binge eating. The three ethnic groups reported equal levels of
depressive symptomatology (all in the non-clinical range), and the
association between depressive symptoms and binge eating was
present across all groups. Though the present study cannot speculate
about the direction of the depressionbinge eating relationship, this
nding is consistent with past research indicating a relationship
between binge eating and affective disorders (Telch & Agras, 1994;
Pike et al., 2001). These ndings also suggest that the depression
binge eating association extends to obese Hispanic women seeking
bariatric surgery, a group that had not previously been studied.
Consistent with past research (Harrington et al., 2008; Regier et al.,
1993), African American women reported higher levels of stress than
Caucasian women. Though African American and Caucasian women in
this sample did not differ with respect to educational background,
higher self-reported stress among African American women may
reect differences in socioeconomic backgrounds or greater incidence
of stressful life experiences, including racism, differences in acculturation, and other discriminatory experiences (Thompson, 1994). Of
particular interest is the nding that Hispanic women did not report
higher levels of stress relative to Caucasians. Literature on stress in the
Hispanic population highlights the need for culturally-specic
assessment tools that are sensitive to issues that affect this
population, such as immigrant stress, English prociency, and cultural
conicts (Cervantes, Padilla, & Salgado de Snyder, 1990; Alegria et al.,
2007). Given that the PAI does not provide norms for Hispanic
individuals, it is possible that it does not adequately capture stressors
that may be unique to Hispanic populations. Also, while statistically
signicant differences emerged between self-reported stress levels of
Caucasian and African American women, it should be noted that the
PAI Stress scores for all three groups were within the average range;
therefore, these differences may not necessarily reect clinically
signicant differences.
Stress was not associated with binge eating, suggesting that
depression may play a more signicant role in the development and
maintenance of binge eating. Though the PAI Stress scale measures
global stress across multiple domains, it may not adequately capture
the discrete, circumscribed stressful incidents that have been shown
to trigger binge episodes (Epel et al., 2001; Gluck et al., 2004). Use of
stress measures that assess both chronic and immediate stressors are
warranted to better understand the relationship among stress, binge
eating, and ethnicity.
Some methodological limitations existed in the present study. First,
the relatively small sample of Hispanic women may have limited the
ability to detect modest group differences. In our large sample of 384
female participants, only 38 (9.4%) were Hispanic. It is noteworthy that
although rates of obesity are higher among Hispanics than Caucasians,
Hispanics appear to seek bariatric surgery at a much lower rate than
Caucasian and African Americans. With anticipated increases in the
rates of obesity among all Americans, including Hispanic Americans, it
remains to be seen whether bariatric surgery will gain popularity in this
group. Further research is needed to explore the psychological, cultural,
and motivational processes that may contribute to the relatively low
frequency with which this group seeks bariatric surgery. Additional
research is also warranted to examine possible differences that may
exist among Hispanic subgroups with respect to psychological functioning and eating pathology. Another study limitation lies in reliance on
self-reported psychological symptoms and height/weight data, which
raises the possibility of under- or over-reporting in efforts to be
approved for surgery. Though reliance on self-report is not ideal,
empirical studies comparing self-reported and actual weight suggests
that bariatric patients tend to be accurate within 0.7 pound in their self-

83

reported weight (White, Masheb, Burke-Martindale, Rothschild, & Grilo,

2007). Lastly, our study was limited to the examination of depressive
symptoms and stress, whereas it has been shown that other
psychological, cultural, and eating variables, such as level of acculturation, loss of control, and dietary restraint/inhibition, may also inuence
binge eating (Harrington et al., 2008; Elder et al., 2008). Further research
using a more comprehensive model for examining psychosocial and
dietary correlates of binge eating is warranted. By better understanding
ethnic differences in psychological correlates of binge eating, we can
appropriately tailor pre-surgical assessments and interventions and
better prepare patients for successful outcomes following bariatric
surgery.
Role of funding sources
No nancial support was received in the conduct of this research or preparation of
this manuscript.

Contributors
Authors Leila Azarbad and Joyce Corsica designed the study. Author Leila Azarbad
wrote the rst draft of the manuscript. Author Brian Hall conducted the statistical
analysis. Author Megan Hood contributed to the literature search and provided input
on the study design. All authors contributed to and have approved the nal manuscript.

Conict of interest
All of the authors declare that they have no conicts of interest.

Acknowledgements
The authors wish to thank Ms. Julia Ivan for her assistance with data entry.

References
Aiken, L. S., & West, S. G. (1991). Multiple regression: Testing and interpreting
interactions Newbury Park, CA: Sage.
Alegria, M., Mulvaney-Day, N., Torres, M., Polo, A., Cao, Z., & Canino, G. (2007).
Prevalence of psychiatric disorders across Latino subgroups in the United States.
American Journal of Public Health, 97, 6875.
Bauchowitz, A., Azarbad, L., Day, K., & Gonder-Frederick, L. (2007). Evaluation of
expectations and knowledge in bariatric surgery patients. Surgery for Obesity and
Related Diseases, 3, 554558.
Beck, A. T., Steer, R. A., & Brown, G. K. (1996). Beck Depression Inventory, 2nd ed San
Antonio: The Psychological Corporation.
Boone, D. (1998). Internal consistency reliability of the Personality Assessment
Inventory with psychiatric inpatients. Journal of Clinical Psychology, 54, 839843.
Busetto, L., Valente, P., Pisent, C., Segato, G., de Marchi, F., & Favretti, F. (1996). Eating
pattern in the rst year following adjustable silicone gastric banding (ASGB).
International Journal of Obesity, 20, 539546.
Cervantes, R. C., Padilla, A. M., & Salgado de Snyder, N. (1990). Reliability and validity of
the Hispanic Stress Inventory. Hispanic Journal of Behavioral Sciences, 12, 7682.
Crago, M., Shisslak, C. M., & Estes, L. S. (1996). Eating disturbances among American
minority groups: A review. International Journal of Eating Disorders, 19, 239248.
de Zwaan, M., Mitchell, J. E., Howell, L. M., Monson, N., Swan-Kremeier, L., Crosby, R. D.,
et al. (2003). Characteristics of morbidly obese patients before gastric bypass
surgery. Comprehensive Psychiatry, 44(5), 428434.
Elder, K. A., Paris, M., Aez, L. M., & Grilo, C. M. (2008). Loss of control over eating is
associated with eating disorder psychopathology in a community sample of
Latinas. Eating Behaviors, 9, 501503.
Epel, E. S., Lapidus, R., McEwen, B., & Brownell, K. (2001). Stress may add bite to
appetite in women: A laboratory study of stress-induced cortisol and eating
behavior. Psychoneuroendocrinology, 26, 3449.
Fabricatore, A. N., Crerand, C. E., Wadden, T. A., Sarwer, D. B., & Krasucki, J. L. (2006).
How do mental health professionals evaluate candidates for bariatric surgery?
Survey results. Obesity Surgery, 16, 567573.
Field, A. E., Camargo, C. A., Taylor, C. B., Berkey, C. S., & Colditz, G. A. (1999). Relation of
peer and media inuences to the development of purging behaviors among
preadolescent and adolescent girls. Archives of Pediatric Adolescent Medicine, 153,
11841189.
Fitzgibbon, M. L., Spring, B., Avellone, M. E., Blackman, L. R., Pingitore, R., & Stolley, M. R.
(1998). Correlates of binge eating in Hispanic, Black, and White women. International Journal of Eating Disorders, 24, 4352.
Gluck, M. E. (2006). Stress response and binge eating disorder. Appetite, 46, 2630.
Gluck, M. E., Geliebter, A., Hung, J., & Yahav, E. (2004). Cortisol, hunger, and desire to
binge following a cold stress test in obese women with binge eating disorder.
Psychosomatic Medicine, 66, 876881.
Gormally, J., Black, S., Daston, S., & Rardin, D. (1982). The assessment of binge eating
severity among obese persons. Addictive Behaviors, 7, 4755.
Gray, J. J., Ford, K., & Kelly, L. M. (1987). The prevalence of bulimia in a Black college
population. International Journal of Eating Disorders, 9, 501512.

84

L. Azarbad et al. / Eating Behaviors 11 (2010) 7984

Greenberg, I., Perna, F., Kaplan, M., & Sullivan, M. (2005). Behavioral and psychological
factors in the assessment and treatment of obesity surgery patients. Obesity
Research, 13, 244249.
Harrington, E. F., Crowther, J. H., Henrickson, H. C. P., & Mickelson, K. D. (2008). The
relationships among trauma, stress, ethnicity, and binge eating. Cultural Diversity
and Ethnic Minority Psychology, 12(2), 212229.
Hrabosky, J., & Grilo, C. M. (2007). Body image and eating disordered behaviour in a
community sample of Black and Hispanic women. Eating Behaviors, 8, 106114.
Hsu, L. K. G., Mulliken, B., McDonagh, B., Das, S. K., Rand, W., Fairburn, C. G., et al. (2002).
Binge eating disorder in extreme obesity. International Journal of Obesity, 26,
13981403.
Kalarchian, M. A., Wilson, G. T., Brolin, R. E., & Bradley, L. (1998). Binge eating in
bariatric surgery patients. International Journal of Eating Disorders, 23, 8992.
Lati, R., Kellum, J. M., De Maria, E. J., & Sugerman, H. (2002). Surgical treatment of
obesity. In T. A. Wadden & A. J. Stunkard's (Eds.), Handbook of obesity treatment
(pp. 339356). New York: Guilford Press.
Lovejoy, M. (2001). Disturbances in the social body: Differences in body image and
eating problems among African American and white women. Gender and Society,
15, 239261.
Mazzeo, S. E., Saunders, R., & Mitchell, K. (2005). Binge eating among African American
and Caucasian bariatric surgery candidates. Eating Behaviors, 6, 189196.
Moore, D. S. (1995). The basic practice of statistics NY: Freeman and Co.
Morey, L. C. (1991). Personality Assessment Inventory Professional Manual. Odessa, FL:
Psychological Assessment Resources, Inc.
Morey, L. C. (2007). Personality Assessment Inventory Professional Manual, 2nd Edition
Odessa, FL: Psychological Assessment Resources, Inc.
Newman, E., O'Connor, D. B., & Conner, M. (2007). Daily hassles and eating behaviour:
The role of cortisol reactivity status. Psychoneuroendocrinology, 32, 125132.
Ogden, C. L., Carroll, M. D., Curtin, L. R., McDowell, M. A., Tabak, C. J., & Flegal, K. M.
(2006). Prevalence of overweight and obesity in the United States, 19992004.
Journal of the American Medical Association, 295(13), 15491555.
Perri, M. G., & Corsica, J. A. (2002). Improving the maintenance of weight lost in
behavioral treatment of obesity. In T. A. Wadden & A. J. Stunkard's (Eds.), Handbook
of obesity treatment (pp. 357382). New York: Guilford Press.
Pike, K. M., Dohm, F. A., Striegel-Moore, R. H., Wiley, D. E., & Fairburn, C. G. (2001). A
comparison of Black and White women with binge eating disorder. American
Journal of Psychiatry, 158, 14551460.

Regier, D. A., Farmer, M. E., Rae, D. S., Myers, J. K., Kramer, M., Robins, L. N., et al. (1993).
One-month prevalence of mental disorders in the United States and sociodemographic characteristics: The Epidemiologic Catchment Area Study. Acta
Psychiatrica Scandinavica, 88, 3547.
Rowston, W. M., McCluskey, S. E., Gazet, J. C., Lacey, J. H., Franks, G., & Lynch, D. (1992).
Eating behaviour, physical symptoms and psychological factors associated with
weight reduction following the Scopinaro operation as modied by Gazet. Obesity
Surgery, 2, 355360.
Shaw, H., Ramirez, L., Trost, A., Randall, P., & Stice, E. (2004). Body image and eating
disturbances across ethnic groups: More similarities than differences. Psychology of
Addictive Behaviors, 18(1), 1218.
Smith, D. E., Marcus, M. D., Lewis, C., Fitzgibbon, M., & Schreiner, P. (1998). Prevalence
of binge eating disorder, obesity, and depression in a biracial cohort of young
adults. Annals of Behavioral Medicine, 20, 227232.
Spitzer, R., Yanovski, S., Wadden, T., Wing, R., Marcus, M., Stunkard, A., et al. (1993).
Binge eating disorder: Its further validation in a multisite study. International
Journal of Eating Disorders, 13, 137153.
Steer, R. A., & Beck, A. T. (2000). The Beck Depression Inventory-II. In W. E. Craighead &
C. B. Nemeroff (Eds.), 3rd ed. The Corsini encyclopedia of psychology and behavioral
science, Vol. 1. (pp. 178179)New York: Wiley.
Striegel-Moore, R. H., & Smolak, L. (2000). The inuence of ethnicity on eating disorders
in women. In R. Eisler & M. Herson (Eds.), Handbook of gender, culture, and health
(pp. 227254). Mahwah, NJ: Erlbaum.
Tabachnick, B., & Fidell, L. S. (2007). Using multivariate statistics, 5th ed. Boston: Pearson.
Telch, C. F., & Agras, W. S. (1994). Obesity, binge eating, and psychopathology: Are they
related? International Journal of Eating Disorders, 15, 5361.
Thompson, B. (1994). Food, bodies, and growing up female: Childhood lessons about
culture, race, and class. In P. Fallon, M. A. Katzman, & S. C. Wooley (Eds.), Feminist
perspectives on eating disorders (pp. 355378). New York: Guilford Press.
Wadden, T. A., Foster, G. D., Letizia, K. A., & Wilk, J. E. (1993). Metabolic, anthropometric,
and psychological characteristics of obese binge eaters. International Journal of
Eating Disorders, 14, 125.
Waters, A., Hill, A., & Waller, G. (2001). Bulimics' response to food cravings: Is binge eating a
product of hunger or emotional state? Behaviour Research and Therapy, 39, 877886.
White, M. A., Masheb, R. M., Burke-Martindale, C., Rothschild, B., & Grilo, C. M. (2007).
Accuracy of self-reported weight among bariatric surgery candidates: The inuence
of race and weight cycling. Obesity, 15, 27612768.