Skip to main content
PMC home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2020 Oct 1.
Published in final edited form as: Appetite. 2019 Jun 4;141:104309. doi: 10.1016/j.appet.2019.06.001

Associations between Eating Expectancies and Eating Disorder Symptoms in Men and Women

Leigh C Brosof 1, Melissa A Munn-Chernoff 2, Cynthia M Bulik 2,3,4, Jessica H Baker 2
PMCID: PMC7079228  NIHMSID: NIHMS1563369  PMID: 31170435

Abstract

Eating expectancies, or learned expectations that an individual has about eating, prospectively predict eating disorder (ED) symptoms. Most studies examining eating expectancies have focused on one or two eating expectancies and their relation with bulimic symptoms. In addition, these studies have been conducted mostly in women. Thus, it is unclear whether: 1) associations between eating expectancies and ED symptoms vary between men and women, and 2) extend to ED symptoms other than bulimic symptoms. The current study (N = 197 undergraduate men and 246 undergraduate women) investigated sex variance in a model of eating expectancies and ED symptoms, including factors associated with ED symptoms (i.e., negative urgency, negative affect, alcohol use, drug use, and body mass index). Sex variance was tested using path analysis in a model including eating expectancies and associated factors, with excessive exercise, negative attitudes toward obesity, restricting, cognitive restraint, binge eating, purging, muscle building, and body dissatisfaction as dependent variables. Unconstrained (i.e., unconstrained paths across men and women) and constrained (i.e., constraining paths across men and women) models were tested. The unconstrained and constrained models differed significantly, indicating that the models varied by sex. For both sexes, eating expectancies were uniquely associated with ED symptoms. For men, Eating Manages Negative Affect was significantly associated with the most ED symptoms. In contrast, for women, Eating Leads to Feeling Out of Control was associated with the most ED symptoms. Previous findings regarding eating expectancies and ED symptoms in women may not generalize to men. Intervening on eating expectancies in a sex-specific way may help reduce specific ED symptoms.

Keywords: women, men, eating disorders, eating expectancies, college

Eating expectancies are learned expectations surrounding eating that an individual develops through positive or negative reinforcement over time (Hohlstein, Smith, & Atlas, 1998) and are related to and prospectively predict eating disorder (ED) symptoms (e.g., Pearson, Zapolski, & Smith, 2015; Smith, Simmons, Flory, Annus, & Hill, 2007). Five distinct eating expectancies have been proposed, as measured by the Eating Expectancies Inventory (EEI; Holstein et al., 1998): 1) Eating Manages Negative Affect (EEI NA), 2) Eating is Pleasurable and Useful as a Reward (EEI Reward), 3) Eating Leads to Feeling out of Control (EEI Control), 4) Eating Alleviates Boredom (EEI Boredom), and 5) Eating Enhances Cognitive Competence (EEI Cognitive Competence). A significant body of literature has investigated the associations between two eating expectancies – EEI NA and EEI Reward – and bulimic symptomatology (i.e., binge eating and purging; De Young, Zander, & Anderson, 2014; Fischer, Peterson, & McCarthy, 2013; Fischer et al., 2018). However, less research has examined the association between the remaining eating expectancies and ED symptoms. In addition, associations between eating expectancies and ED symptoms have been conducted largely in women. Understanding the symptom-specific associations between eating expectancies and ED symptoms by sex will aid in more targeted treatment for both men and women.

The robust literature on EEI NA has indicated a consistent positive association with bulimic symptoms (e.g., Combs, Pearson, & Smith, 2011; Pearson & Smith, 2015; Racine & Martin, 2017). However, findings are mixed regarding EEI Reward and bulimic symptoms, with some studies showing a positive association and others showing a negative association (e.g., Bohon, Stice, & Burton, 2009; Combs, Smith, & Simmons, 2011). The existing literature on the other three eating expectancies indicate potential associations with ED symptomatology. For instance, lower EEI Control relates to eating disorder recovery (Fitzsimmons-Craft, Keatts, & Bardone-Cone, 2013), and women with bulimia nervosa (BN) report higher EEI Boredom than healthy controls (Bruce, Mansour, & Steiger, 2009). Conversely, no relation between ED symptoms and EEI Cognitive Competence has been observed (Hearon, Utschig, Smits, Moshier, & Otto, 2013). Thus, though the literature strongly supports an association between eating expectancies and bulimic symptoms, more research is needed in understanding how eating expectancies relate to a broader range of ED symptoms.

Sex Differences in Eating Expectancies and ED Symptom Associations

In general, research examining associations between eating expectancies and ED symptoms has been conducted largely in women. However, tests of other constructs and their association with ED symptoms, such as perfectionism (e.g., Forbush, Heatherton, & Keel, 2007), have shown that findings in female-only samples are not necessarily generalizable to men and that sex-specific investigation of these related factors is sorely needed (Dakanalis, Timko, Clerici, Zanetti, & Riva, 2014). The existing research investigating sex differences in eating expectancies has yielded mixed findings. For instance, whereas Hayaki and Free (2016) and Davis, Guller, and Smith (2016a; 2016b) reported differences between men and women in how eating expectancies were associated with ED symptoms, Schaumberg and Earleywine (2014) found that this relation did not differ by sex. These contrasting findings may be due to the use of different eating expectancies (i.e., EEI NA versus a composite of eating expectancies) and the ED symptom under study (e.g., binge eating and purging versus excessive exercise). Thus, a closer examination of the relation between eating expectancies and multiple ED symptoms in men are warranted to determine if results observed in women are generalizable to men, ultimately informing sex-specific treatment of EDs.

The Acquired Preparedness Model

Eating expectancies are conceptualized as a part of the acquired preparedness (AP) model, which posits that personality vulnerabilities interact with learned behaviors (i.e., eating expectancies) to create risk for the development of EDs (Combs, Pearson, & Smith, 2011; Schaumberg & Earleywine, 2014). It has been proposed that the personality traits of negative urgency and negative affect transact with eating expectancies to result in the advent of ED symptoms (Dingemans et al., 2009; Racine & Martin, 2017). This model has received substantial support for the development of binge eating and purging (Combs, Smith, & Simmons, 2011; Davis & Smith, 2018; Pearson & Smith, 2015). Though the AP model has not yet been tested as thoroughly for ED symptoms beyond binge eating and purging, it is likely that this model applies to symptoms across the ED spectrum (Davis & Smith, 2018).

The AP model may also be useful for understanding differences in the development of ED symptoms across sex. Within the AP model, it is possible for similar personality traits (i.e., negative urgency and negative affect) to lead to different maladaptive behaviors across individuals based on differences in learning experiences. For instance, women have higher levels of disordered eating behaviors than men (Forbush, Wildes, & Hunt, 2014). Men on the other hand, are more prone to developing maladaptive alcohol and drug use than women (O’Malley & Johnston, 2002), which have also been conceptualized through the AP model (Lavender et al., 2015; Vangsness, Bry, & LaBouvie, 2005). Thus, similar personality traits may also lead to differing ED symptoms based on sex-specific associations with eating expectancies. Previous literature has shown that women may be taught to attach multiple meanings to food, whereas this may not be the case for men (Pettit, Jacobs, Page, & Porras, 2010). Such learning theories may explain why EEI Reward and EEI NA confer risk for ED symptoms in women but only EEI NA confers risk for ED symptoms in men (Hayaki and Free 2016).

Despite the burgeoning literature on eating expectancies and ED symptoms, few studies have included all theorized constructs of the AP model simultaneously. The full AP model (as described above) would include the personality variables of negative urgency and negative affect transacting with eating expectancies to result in differential ED symptoms in relation to ED symptoms in women and men; however, as drug/alcohol use has also been conceptualized through the AP model, it would be important to include them as covariates in such a model. Including all related constructs in a single model allows for examining which constructs contribute unique variance to ED symptoms, beyond the variance of all other constructs in the model, and whether this varies by sex, which is essential for the development of sex-specific precision interventions.

The Current Study

The current study examined sex variance in associations between eating expectancies and multiple ED symptoms. We also adjusted for constructs established as relevant to the AP model and ED symptoms, including negative affect, negative urgency, drug use, alcohol use. We also include body mass index (BMI) as a covariate to control for potential confounds in a college-sample (e.g., Arriaza & Mann, 2001; Rø, Reas, & Rosenvinge, 2012). We hypothesized that: 1) there would be sex variance in the model, and 2) for women, EEI NA and EEI Reward would be the eating expectancies associated with the most ED symptoms, whereas for men, EEI NA would be the eating expectancy related to the most ED symptoms. These findings will elucidate the unique associations between eating expectancies and ED symptoms in men and women, information that could ultimately inform more tailored sex-specific treatments for EDs.

Methods

Participants

Participants included undergraduate students from a large southeastern university enrolled in an introductory psychology course (N = 459) who received course credit for participation in the study. One hundred ninety-seven (42.9%) participants were men and 249 (54.2%) were women; 13 (2.8%) participants did not report their sex and were removed from these analyses. The average age of the remaining 443 participants was 19.44 years old (SD = 4.02), with nearly half of participants in their first year of college (n = 220; 47.9%). The majority of participants reported Caucasian as their race (n = 284; 61.9%); other reported races included African American (n = 49; 10.7%), Asian (n = 80; 17.4%), Pacific Islander (n = 1; 0.2%), Native American (n = 1; 0.2%), Mixed (n = 16; 3.5%), and Other (n = 16; 3.5%). Twelve participants did not report their race (2.6%). This study was approved by the local institutional review board, and all students provided online consent before participating.

Measures

Participants completed questionnaires via an online self-report assessment.

Eating expectancies

The Eating Expectancy Inventory (EEI; Hohlstein, Smith, & Atlas, 1998) was used to assess learned expectations about eating and contains five subscales: Eating Manages Negative Affect (EEI NA; e.g., When I am feeling anxious or tense, eating helps me relax); Eating is Pleasurable and Useful as a Reward (EEI Reward; e.g., When I do something good, eating is a way to reward myself); Eating Leads to Feeling Out of Control (EEI Control; e.g., Eating makes me feel out of control); Eating Alleviates Boredom (EEI Boredom; e.g., When I have nothing to do, eating helps relieve the boredom); and Eating Enhances Cognitive Competence (EEI Cognitive Competence; e.g., Eating helps me think and study better). Subscales were calculated by summing items, with higher scores indicating greater endorsement of that eating expectancy. The EEI demonstrates good internal consistency and validity in a college-aged sample (Hohlstein et al., 1998). Internal consistencies in this sample were the following: 0.66 (EEI Boredom), 0.70 (EEI Reward), 0.78 (EEI Control), 0.81 (EEI Cognitive Competence), and 0.94 (EEI NA).

Eating Disorder Symptoms

The 45-item Eating Pathology Symptoms Inventory (EPSI; Forbush et al., 2013) assessed ED symptoms and is comprised of eight subscales: purging (i.e., compensatory behaviors used to lose weight, such as vomiting, laxative, or diuretic use); binge eating (i.e., eating a very large amount of food in a short period of time [e.g., within 2 hours] and feeling out of control); muscle building (i.e., the desire to build more muscle); excessive exercise (i.e., exercising in an extreme or compulsive manner); restricting (i.e., the behavioral, actual restriction of one’s caloric intake); negative attitudes toward obesity (i.e., holding negative beliefs regarding obesity); body dissatisfaction (i.e., discontent with one’s body), and cognitive restraint (i.e., the intent to restrict one’s caloric intake, whether or not successful). The EPSI evidences good internal consistency and construct validity in both college-aged men and women (Forbush, Wildes, & Hunt, 2014; Forbush et al., 2013). Internal consistencies for the subscales in this sample are the following: 0.77 (Cognitive Restraint subscale), 0.79 (Muscle Building subscale), 0.87 (Binge Eating subscale), 0.87 (Restricting subscale), 0.87 (Negative Attitudes Toward Obesity subscale), 0.88 (Excessive Exercise subscale), 0.88 (Purging subscale), and 0.90 (Body Dissatisfaction subscale).

Covariates

The total score from the Depression Anxiety and Stress Scales (DASS; Lovibond & Lovibond, 1995) was used as a broad measure of negative affect, with higher scores reflecting greater negative affect. The DASS has 42-items that assess depression, anxiety, and stress. The DASS evidences good internal consistency and concurrent validity in a community sample (Antony, Bieling, Cox, & Enns, 1998). Internal consistency in this sample was excellent (α = 0.96).

To assess negative urgency, the negative urgency subscale of the UPPS-P Impulsive Behavior Scale (UPPS-P; Lynam, Smith, Whiteside, & Cyders, 2006) was used. The UPPS-P is invariant among male and female college students and demonstrates good validity and test-retest reliability in these students (Cyders, 2011). Internal consistency in this sample was α = 0.76.

The Alcohol Use Disorders Identification Test (AUDIT; Babor, Higgins-Biddle, Saunders, & Monteiro, 2001) was used to assess problematic drinking. The AUDIT is a brief questionnaire with items capturing hazardous alcohol use, dependence symptoms, and harmful alcohol use. The AUDIT has good psychometric properties in a college-aged population (Fleming, Barry, & Macdonald, 1991). Internal consistency in this sample was good (α = 0.83).

Similarly, the Drug Use Disorders Identification Test (DUDIT; Berman, Bergman, Palmstierna, & Schlyter, 2003) was used to assess problematic drug use. The DUDIT evidences excellent internal consistency and good sensitivity and specificity (Hildebrand, 2015). Internal consistency in this sample was good (α = 0.84).

Finally, BMI was calculated through self-reported height and weight: weight (lb) / [height (in)]2 x 703.

Statistical Analyses

First, independent samples t-tests were conducted to test whether men and women differed significantly on eating expectancies, ED symptoms, and the covariates of interest. A Holm’s Sequential Bonferroni Procedure (Holm, 1979), which is a modified version of the Bonferroni correction that maximizes power while controlling for inflation of Type I error (Abdi, 2010), was used to correct for multiple comparisons. The family-wise error rate was 0.05.

Sex variance tests

MPlus Version 8.0 (Muthen & Muthen, 1998) was used to conduct analyses. The maximum likelihood robust standard error (MLR) estimator was used to report standardized path coefficients, which estimates missing data efficiently and is robust to non-normality in the data (Muthen & Muthen, 1998). Several fit indices were used to evaluate model fit: (a) the comparative fit index (CFI) (Bentler, 1990), (b) the Tucker-Lewis incremental fit index (TLI) (Tucker & Lewis, 1973), (c) the root mean square error of approximation (RMSEA) (Steiger & Lind, 1980), and (d) the standardized root mean square residual (SRMR) (Jöreskog & Sörbom, 1981). For the CFI and TLI, adequate values are 0.90 or above and very good values are 0.95 or above. For the RMSEA and SRMR, adequate values are 0.08 or below and very good values are 0.05 or below.

To test if sex variance was present in the model, recommendations from Bollen (1989) were followed for path analysis. First, an unconstrained model (with unconstrained paths across men and women) was tested. This model included all five eating expectancies (i.e., EEI NA, EEI Reward, EEI Control, EEI Boredom, and EEI Cognitive Competence) and the related constructs (i.e., negative affect, negative urgency, drug use, alcohol use, and BMI) as independent variables. Purging, binge eating, muscle building, excessive exercise, restricting, negative attitudes toward obesity, body dissatisfaction, and cognitive restraint served as dependent variables. Next, a constrained model (with constrained paths across men and women) using these same paths was tested. A chi-square difference test was conducted using the Satorra-Bentler scaling correction (which is necessary when using the MLR estimator) to compare the unconstrained and constrained models (Satorra & Bentler, 2010). A significant chi-square value indicates the models are significantly different, showing that the models vary by sex. Missing data were excluded listwise in subsequent analyses. Sixteen individuals had missing data and therefore were excluded, resulting in a sample size of 443.

Results

Independent Samples t-tests

Results are shown in Table 1. Mean scores for binge eating, restricting, body dissatisfaction, and cognitive restraint (ps ≤ .001) were significantly higher in women than men, whereas mean scores for muscle building and excessive exercise (ps ≤ .001) were significantly higher in men than women. For eating expectancies, the mean score for EEI Reward (p = .001) was significantly higher in women than men. For covariates, negative affect (p = .002), drug use (p < .001), and alcohol use (p < .001) were significantly different in men and women. Women had significantly higher mean scores for negative affect, whereas men had significantly higher mean scores on drug use and alcohol use.

Table 1.

Independent samples t-tests comparing means for all study variables between men and women.

Variable Men Women
Mean (SD) Mean (SD) t-statistic df Cohen’s d p-value
Purging 0.56 (2.17) 1.23 (2.90) −2.64 434 0.26 .009
Binge Eating 5.63 (4.95) 7.42 (6.31) −3.22 434 0.32 .001
Muscle Building 5.02 (4.47) 1.59 (2.35) 10.32 434 0.96 < .001
Excessive Exercise 6.87 (5.31) 5.04 (5.30) 3.46 386.05 0.34 .001
Restricting 3.31 (4.11) 5.06 (5.23) −3.80 434 0.37 < .001
Negative Attitudes Toward Obesity 5.72 (4.64) 4.67 (4.51) 2.38 434 0.23 .018
Body Dissatisfaction 4.65 (4.67) 11.33 (7.09) −11.24 434 1.11 < .001
Cognitive Restraint 3.45 (2.71) 4.91 (3.25) −5.00 434 0.49 < .001
EEI NA 45.92 (20.98) 51.20 (22.29) −2.46 394.44 0.24 .014
EEI Reward 27.67 (6.23) 29.86 (6.77) −3.51 419.32 0.34 < .001
EEI Control 9.38 (4.75) 10.20 (5.45) −1.67 426.22 0.16 .095
EEI Boredom 14.74 (5.07) 15.67 (5.35) −1.87 414.60 0.18 .062
EEI Cognitive Competence 7.61 (3.34) 7.99 (3.28) −1.17 400.77 0.11 .244
Negative Affect 13.78 (15.43) 19.23 (18.57) −3.17 405 0.32 .002
Negative Urgency 25.68 (6.85) 25.63 (6.58) 0.07 396.08 0.01 .942
Drug Use 3.15 (5.80) 1.36 (3.43) 3.82 397 0.38 < .001
Alcohol Use 8.51 (5.61) 5.60 (4.67) 5.16 330 0.56 < .001
Current BMI 23.59 (3.39) 22.92 (4.02) 1.87 424.54 0.18 .063
Open in a new tab

The p-values that remained significant after conducting the Holm’s Sequential Bonferonni Procedure to correct for multiple comparisons are reported; the family-wise error rate was .05; variables with significant differences are bolded for clarity. SD = standard deviation; df = degrees of freedom; EEI NA = Eating Manages Negative Affect; EEI Reward = Eating is Pleasurable and Useful as a Reward; EEI Control = Eating Leads to Feeling Out of Control; EEI Boredom = Eating Alleviates Boredom; EEI Cognitive Competence = Eating Enhances Cognitive Competence; BMI = body mass index; All eating disorder symptoms are scored according to the Eating Pathology Symptoms Inventory (Forbush et al., 2013); Eating expectancies are scored according to the Eating Expectancy Inventory (Hohlstein, Smith, & Atlas, 1998); Negative affect is scored according to the Depression Anxiety and Stress Scales (Lovibond & Lovibond, 1995); Negative urgency is scored according to the UPPS-P Impulsive Behavior Scale (Lynam, Smith, Whiteside, & Cyders, 2006); Drug use is scored according to the Drug Use Disorders Identification Test (Berman, Bergman, Palmstierna, & Schlyter, 2003); Alcohol use is scored according to the Alcohol Use Disorders Identification Test (Babor, Higgins-Biddle, Saunders, & Monteiro, 2001); BMI was calculated according to self-reported height (inches) and weight (pounds).

Sex Variance Model

The unconstrained model was saturated, meaning that there are no remaining degrees of freedom in the model; when non-significant paths were removed, the model had excellent fit: CFI = .99, TLI = .97, RMSEA = .03, SRMR = .03. When comparing the unconstrained and constrained models, the Santorra-Bentler corrected chi-square statistic indicated a significant difference between the two models (X2 (54) = 115.33, p < .001), suggesting the models varied by sex. Therefore, paths were examined according to sex. Unique associations for men and women between eating expectancies and ED symptoms emerged, even after adjusting for related constructs. Table 2 presents the significant paths for the associations between eating expectancies and ED symptoms by sex. Since no significant associations emerged with EEI Cognitive Competence, these are not discussed below.

Table 2.

Significant paths for the associations between eating expectancies and eating disorder symptoms by sex.

Men Women
b*-value p-value b*-value p-value
Purging EEI NA .27 <.001 EEI NA .25 < .001
EEI Reward −.13 .008 EEI Reward −.22 < .001
Drug Use −.30 <.001 EEI Control .21 .001
Negative Urgency −.17 .002
Negative Affect .17 .043
Binge Eating EEI Control .30 < .001 EEI Control .36 < .001
Negative Affect .22 .001 Negative Affect .17 .007
Drug Use −.22 .006 EEI NA .25 .001
Alcohol Use .26 .003 EEI Boredom .26 < .001
Muscle Building Drug Use −.25 < .001 EEI Control .15 .035
EEI Boredom −.23 .006 EEI NA 19 .036
Negative Affect .24 < .001 EEI Reward −.17 .026
Excessive Exercise EEI Control .25 .007 EEI Control .44 < .001
EEI NA .25 .021 Alcohol Use .16 .042
EEI Reward .16 .043
Negative Affect .17 .029
Restricting Negative Affect .23 .009 Negative Affect .49 < .001
EEI NA .18 .043 EEI Boredom −.18 .007
EEI Reward −.12 .046
EEI Control .25 .008
BMI −.16 .004
Negative Attitudes Toward Obesity Negative Affect .25 .004 EEI NA .19 .027
EEI Control .24 .002
Alcohol Use .17 .015
Body Dissatisfaction EEI NA .22 .024 EEI NA .14 .043
EEI Control .18 .022 EEI Control .41 < .001
Negative Affect .45 < .001 Negative Affect .29 < .001
BMI .18 .015 BMI .24 .001
Alcohol Use .20 .002
Cognitive Restraint EEI Control .36 < .001 EEI Control .56 < .001
Negative Affect .33 < .001 EEI Boredom −.14 .039
Open in a new tab

All paths were tested; non-significant predictors are not included for clarity. Bolded variables for men and women indicate shared eating expectancies between men and women. EEI NA = Eating Manages Negative Affect EEI Reward = Eating is Pleasurable and Useful as a Reward; EEI Control = Eating Leads to Feeling Out of Control; EEI Boredom = Eating Alleviates Boredom; BMI = body mass index.

Shared associations between men and women

EEI Control (bs ≥ .18, ps < .050) had the most shared associations between men and women, including significant positive associations with binge eating, excessive exercise, body dissatisfaction, and cognitive restraint in both sexes. EEI NA was positively associated with purging (bs ≥ .25, ps < .001) and body dissatisfaction (bs ≥ .14, ps < .050) in both men and women. In contrast, EEI Reward was significantly negatively associated with purging (bs ≤ −.13, ps < .010) in both sexes.

Associations unique to men

EEI Control (bs ≥ .18, ps < .050) had the most unique associations of any eating expectancy with ED symptoms in men, including positive associations with binge eating, excessive exercise, restricting, and body dissatisfaction. This was followed by EEI NA, which had significant positive (bs ≥ .18, ps < .050) associations with four ED symptoms (i.e., purging, excessive exercise, restricting, and body dissatisfaction). EEI Reward (bs ≥ .12, ps < .050) was uniquely positively associated with three ED symptoms (i.e., purging, excessive exercise, and restricting), whereas EEI Boredom (b* = −.23, p = .006) was only uniquely related to muscle building.

Associations unique to women

EEI Control (bs ≥ .15, ps ≤ .035) was uniquely associated with the most ED symptoms (i.e., purging, muscle building, and negative attitudes toward obesity) in women. EEI NA (bs ≥ .19, ps ≤ .043) was uniquely positively related to three ED symptoms (i.e., binge eating, muscle building, and negative attitudes toward obesity). EEI Boredom also had unique associations with three ED symptoms: it was positively associated with binge eating (b* = .26, p < .001) and negatively associated with restricting and cognitive restraint (bs ≤ −.14, ps ≤ .039). Finally, EEI Reward (b* = −.17, p = .026) was only uniquely associated with muscle building.

Discussion

In our investigation of sex variance in a model including five proposed eating expectancies in association with multiple ED symptoms, while adjusting for related constructs, we found that this model varied by sex; unique associations for men and women between eating expectancies and ED symptoms existed. For both sexes, the expectancies that eating leads to feeling out of control (EEI Control) and eating manages negative affect (EEI NA) played an important role in the majority of ED symptoms. However, unique associations between eating expectancies and ED symptoms were also present for both sexes.

Shared Associations Between Men and Women

Eating expectancies showed significant associations with all ED symptoms. In particular, the expectancies that eating manages negative affect and eating leads to feeling out of control were related to the majority of ED symptoms in both men and women. This finding is consistent with previous literature showing that eating to reduce negative affect is related to bulimic symptoms (e.g., Lavender et al., 2015; Pearson & Smith, 2015; Racine & Martin, 2017) and extends this observed relation to other ED symptoms, such as dietary restriction and body dissatisfaction. This finding also supports the application of the AP model to ED symptoms beyond binge eating and purging. In contrast, the limited research including the expectancy that eating leads to feeling out of control suggests that it is associated with ED recovery and with overeating (Fitzsimmons-Craft, Keatts, & Bardone-Cone, 2013; Hearon et al., 2013). Thus, eating to manage negative affect and eating leads to feeling out of control may represent vulnerabilities that generalize across ED symptoms for both men and women.

Unique Associations in Men and Women

Though similarities were observed between men and women, there were also unique eating expectancy-ED symptom associations by sex. In contrast to our hypothesis, in men, the expectancy that eating is rewarding was uniquely associated with purging, excessive exercise, and restricting. It appears that viewing eating as rewarding may be particularly linked with compensatory behaviors in men. It is possible that men who view eating as rewarding are more likely to feel that their eating is out of control (regardless of whether binge eating is present) and thus may be more likely to compensate for eating. This finding contrasts with previous literature showing that eating for pleasure/reward was not related to ED symptoms in men (Hayaki & Free, 2016). This difference may be explained by the fact that the previous study investigated eating expectancies in relation to a composite score of bulimic symptoms, rather than specific compensatory behaviors or other ED symptoms. In women, eating in response to boredom was positively associated with binge eating and negatively associated with restricting and cognitive restraint. Eating to relieve boredom has previously been shown to be higher in women with BN symptoms (Bruce, Mansour, & Steiger, 2009), and this finding adds support for the role eating to reduce boredom in perpetuating binge eating while extending this to restricting behaviors.

Unique Associations Across Sex

Interestingly, muscle building and restricting did not share any eating expectancy associations between men and women when controlling for other constructs of relevance (i.e., substance use, negative affect, negative urgency, and BMI). In women, the expectancies that eating leads to feeling out of control and eating to manage negative affect were associated with muscle building, whereas in men, only eating to relieve boredom was associated with muscle building. Moreover, in women, only eating to relieve boredom was associated with restricting, whereas in men, eating to reduce negative affect, eating is rewarding, and eating leads to feeling out of control were associated with restricting. Thus, men and women may be predisposed to develop different ED symptoms from the same eating expectancies. The AP model may help explain these sex differences. The AP model proposes that personality traits transact with learning to lead to the development of ED symptoms (e.g., Combs, Pearson, & Smith, 2011). Differential learning may take place according to sex, which may predispose men and women to engage in certain ED symptoms (Pettit, Jacobs, Page, & Porras, 2010). College men are more likely to endorse muscle building than women, whereas college women are more likely to endorse restricting than men (Forbush, Wildes, & Hunt, 2014). Men may learn that eating should be compensated through muscle building behaviors, which may explain the associated between eating when bored and muscle building. Women, on the other hand, may be socially taught to compensate through restriction. Thus, in line with the AP model, the interaction between psychosocial learning and sex may inform the development of different ED symptoms.

Limitations

Limitations of the current study should be considered. First, our sample comprised of college students, which may not generalize to other ages or to a clinical sample. However, undergraduate samples are useful in identifying potential risk factors for ED symptoms, as college is a high-risk period for the development of an ED (e.g., Stice, 2016; Taylor et al., 2006). Second, all constructs were assessed via self-report questionnaires, which may not accurately capture certain behaviors, such as binge eating (e.g., Burton, Abbott, Modini, & Touyz, 2016). However, self-report questionnaires can be more valid for assessing ED symptoms than diagnostic interviews (Keel, Crow, Davis, & Mitchell, 2002). Third, the current study is cross-sectional, limiting our ability to establish temporal precedence and prohibiting causal conclusions.

The strengths of the current study should also be noted. The current study included previously explored relevant constructs in the AP model, which allowed for the examination of unique associations between eating expectancies and ED symptoms, or associations between eating expectancies and ED symptoms when partialling out the relations with all other constructs in the model. Such investigations are imperative for understanding which constructs may be important for targeted interventions. In addition, this study investigated model variance by sex, which allows for elucidating unique and shared associations for both men and women. Model variance testing is crucial for understanding how theories and models differ according to different subgroups. Thus, this study is an important step and adds to the robust literature on eating expectancies in understanding how the AP model may vary across sex.

Potential Clinical Implications

A better understanding of sex-specific and symptom-specific relations between eating expectancies and ED symptoms may aid in the development of precision interventions, which are likely to show better outcomes (Fisher et al., 2017; Insel, 2014). Eating expectancies may offer promise as targets of intervention. For instance, cognitive-behavioral approaches, such as using hypothesis testing, could be used to teach individuals that their eating expectations are violated (e.g., that eating does not help manage negative affect in the long-term). Targeting eating expectancies could thus be integrated into cognitive-behavioral treatments of EDs (e.g., Fairburn, 2008; Steinglass et al., 2011) and applied in a sex-specific manner. Future research should seek to replicate these findings in a clinical sample to better understand how clinical interventions targeting eating expectancies may apply in men versus women with EDs.

Conclusions

These results suggest that previous literature on eating expectancies and ED symptoms may not generalize to men. Specifically, the AP model suggests that men may be more prone to certain ED symptoms based on differential social learning. Future research should continue to investigate all eating expectancies across a range of ED symptoms in prospective data in men and women. Longitudinal designs that clarify the temporal relationship will aid in the development of sex-specific eating expectancy-ED symptom targeted treatments. Finally, conducting ecological momentary assessment studies on eating expectancies to understand how these expectancies affect momentary processes in the real world in men compared with women would be beneficial. Such research may aid in the development of sex-specific interventions to alleviate the suffering of those with ED symptoms.

Funding Acknowledgement

Ms. Brosof received funding for this project through the University of North Carolina Center of Excellence for Eating Disorder’s Summer Research Fellowship. Dr. Munn-Chernoff acknowledges funding from K01 AA025113 and Dr. Baker acknowledges funding from K01 MH106675.

References

  1. Abdi H (2010). Holm’s sequential Bonferroni procedure In Neil Salkind (Ed.), Encyclopedia of Research Design. Thousand Oaks, CA: Sage. [Google Scholar]
  2. Arriaza CA, & Mann T (2001). Ethnic differences in eating disorder symptoms among college students: The confounding role of body mass index. Journal of American College Health, 49, 309–315. [DOI] [PubMed] [Google Scholar]
  3. Atlas JG (2004). Interpersonal sensitivity, eating disorder symptoms, and eating/thinness expectancies. Current Psychology, 22, 368–378. [Google Scholar]
  4. Bentler PM (1990). Comparative fit indexes in structural models. Psychological Bulletin, 107, 238–246. [DOI] [PubMed] [Google Scholar]
  5. Bohon C, Stice E, & Burton E (2009). Maintenance factors for persistence of bulimic pathology: A prospective natural history study. International Journal of Eating Disorders, 42, 173–178. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bollen KA (1989). A new incremental fit index for general structural equation models. Sociological Methods & Research, 17, 303–316. [Google Scholar]
  7. Bruce K, Mansour S, & Steiger H (2009). Expectancies related to thinness, dietary restriction, eating, and alcohol consumption in women with bulimia nervosa. International Journal of Eating Disorders, 42, 253–258. [DOI] [PubMed] [Google Scholar]
  8. Burton AL, Abbott MJ, Modini M, & Touyz S (2016). Psychometric evaluation of self-report measures of binge-eating symptoms and related psychopathology: A systematic review of the literature. International Journal of Eating Disorders, 49, 123–140. [DOI] [PubMed] [Google Scholar]
  9. Combs JL, Pearson CM, & Smith GT (2011). A risk model for preadolescent disordered eating. International Journal of Eating Disorders, 44, 596–604. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Cyders MA, Smith GT, Spillane NS, Fischer S, Annus AM, & Peterson C (2007). Integration of impulsivity and positive mood to predict risky behavior: Development and validation of a measure of positive urgency. Psychological Assessment, 19, 107. [DOI] [PubMed] [Google Scholar]
  11. Dakanalis A, & Riva G (2013). Current considerations for eating and body-related disorders among men. Handbook on body image: Gender differences, sociocultural influences and health implications, 195–216. [Google Scholar]
  12. Dakanalis A, Timko CA, Clerici M, Zanetti MA, & Riva G (2014). Comprehensive examination of the trans-diagnostic cognitive behavioral model of eating disorders in males. Eating Behaviors, 15, 63–67. [DOI] [PubMed] [Google Scholar]
  13. Dakanalis A, Zanetti AM, Riva G, Colmegna F, Volpato C, Madeddu F, & Clerici M (2015). Male body dissatisfaction and eating disorder symptomatology: Moderating variables among men. Journal of Health Psychology, 20, 80–90. [DOI] [PubMed] [Google Scholar]
  14. Davis HA, Guller L, & Smith GT (2016). Developmental trajectories of compensatory exercise and fasting behavior across the middle school years. Appetite, 107, 330–338. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Davis HA, Guller L, & Smith GT (2016). Developmental trajectories of boys’ driven exercise and fasting during the middle school years. Journal of Abnormal Child Psychology, 44, 1309–1319. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. De Young KP, Zander M, & Anderson DA (2014). Beliefs about the emotional consequences of eating and binge eating frequency. Eating Behaviors, 15, 31–36. [DOI] [PubMed] [Google Scholar]
  17. Dingemans AE, Martijn C, van Furth EF, & Jansen AT (2009). Expectations, mood, and eating behavior in binge eating disorder. Beware of the bright side. Appetite, 53, 166–173. [DOI] [PubMed] [Google Scholar]
  18. Garner DM, Olmstead MP, & Polivy J (1983). Development and validation of a multidimensional eating disorder inventory for anorexia nervosa and bulimia. International Journal of Eating Disorders, 2, 15–34. [Google Scholar]
  19. Fairburn CG (2008). Cognitive behavior therapy and eating disorders. Guilford Press. [Google Scholar]
  20. Fischer S, Anderson KG, & Smith GT (2004). Coping with distress by eating or drinking: role of trait urgency and expectancies. Psychology of Addictive Behaviors, 18, 269–274. [DOI] [PubMed] [Google Scholar]
  21. Fischer S, Peterson CM, & McCarthy D (2013). A prospective test of the influence of negative urgency and expectancies on binge eating and purging. Psychology of Addictive Behaviors, 27, 294–300. [DOI] [PubMed] [Google Scholar]
  22. Fischer S Wonderlich J, Breithaupt L, Byrne C, & Engel S (2018). Negative urgency and expectancies increase vulnerability to binge eating in bulimia nervosa. Eating Disorders, 26, 39–51. [DOI] [PubMed] [Google Scholar]
  23. Fisher AJ, Reeves JW, Lawyer G, Medaglia JD, & Rubel JA (2017). Exploring the idiographic dynamics of mood and anxiety via network analysis. Journal of Abnormal Psychology, 126, 1044–1056. [DOI] [PubMed] [Google Scholar]
  24. Fitzsimmons-Craft EE, Keatts DA, & Bardone-Cone AM (2013). Eating expectancies in relation to eating disorder recovery. Cognitive Therapy and Research, 37, 1041–1047. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Forbush K, Heatherton TF, & Keel PK (2007). Relationships between perfectionism and specific disordered eating behaviors. International Journal of Eating Disorders, 40, 37–41. [DOI] [PubMed] [Google Scholar]
  26. Forbush KT, Hagan KE, Kite BA, Chapa DA, Bohrer BK, & Gould SR (2017). Understanding eating disorders within internalizing psychopathology: A novel transdiagnostic, hierarchical-dimensional model. Comprehensive Psychiatry, 79, 40–52. [DOI] [PubMed] [Google Scholar]
  27. Forbush KT, Wildes JE, & Hunt TK (2014). Gender norms, psychometric properties, and validity for the Eating Pathology Symptoms Inventory. International Journal of Eating Disorders, 47, 85–91. [DOI] [PubMed] [Google Scholar]
  28. Forbush KT, Wildes JE, Pollack LO, Dunbar D, Luo J, Patterson K, … & Bright A (2013). Development and validation of the Eating Pathology Symptoms Inventory (EPSI). Psychological Assessment, 25, 859–878. [DOI] [PubMed] [Google Scholar]
  29. Hayaki J (2009). Negative reinforcement eating expectancies, emotion dysregulation, and symptoms of bulimia nervosa. International Journal of Eating Disorders, 42, 552–556. [DOI] [PubMed] [Google Scholar]
  30. Hayaki J, & Free S (2016). Positive and negative eating expectancies in disordered eating among women and men. Eating Behaviors, 22, 22–26. [DOI] [PubMed] [Google Scholar]
  31. Hohlstein LA, Smith GT, & Atlas JG (1998). An application of expectancy theory to eating disorders: Development and validation of measures of eating and dieting expectancies. Psychological Assessment, 10, 49–58. [Google Scholar]
  32. Holm S (1979). A simple sequentially rejective multiple test procedure. Scandinavian Journal of Statistics, 6, 65–70. [Google Scholar]
  33. Hu L, & Bentler PM (1999). Cutoff criteria for fit indices in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1–55. [Google Scholar]
  34. Insel TR (2014). The NIMH research domain criteria (RDoC) project: precision medicine for psychiatry. American Journal of Psychiatry, 171, 395–397. [DOI] [PubMed] [Google Scholar]
  35. Jöreskog KG, & Sörbom D (1981). LISREL V: Analysis of linear structural relationships by maximum likelihood and least squares methods. University of Uppsala, Department of Statistics. [Google Scholar]
  36. Keel PK, Crow S, Davis TL, & Mitchell JE (2002). Assessment of eating disorders: Comparison of interview and questionnaire data from a long-term follow-up study of bulimia nervosa. Journal of Psychosomatic Research, 53, 1043–1047. [DOI] [PubMed] [Google Scholar]
  37. Kober H, & Boswell RG (2018). Potential psychological & neural mechanisms in binge eating disorder: Implications for treatment. Clinical Psychology Review, 60, 32–44. [DOI] [PubMed] [Google Scholar]
  38. Lavender JM, Happel K, Anestis MD, Tull MT, & Gratz KL (2015). The interactive role of distress tolerance and eating expectancies in bulimic symptoms among substance abusers. Eating Behaviors, 16, 88–91. [DOI] [PubMed] [Google Scholar]
  39. McManus F, & Waller G (1995). A functional analysis of binge-eating. Clinical Psychology Review, 15, 845–863. [Google Scholar]
  40. Muthen LK, & Muthen BO (1998–2014). Mplus user’s guide (7thed). Los Angeles, CA: Muthen & Muthen. [Google Scholar]
  41. O’Dea JA, & Abraham S (2002). Eating and exercise disorders in young college men. Journal of American College Health, 50, 273–278. [DOI] [PubMed] [Google Scholar]
  42. O’Malley PM, & Johnston LD (2002). Epidemiology of alcohol and other drug use among American college students. Journal of Studies on Alcohol, s14, 23–39. [DOI] [PubMed] [Google Scholar]
  43. Pearson CM, & Smith GT (2015). Bulimic symptom onset in young girls: A longitudinal trajectory analysis. Journal of Abnormal Psychology, 124, 1003–1013. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Pearson CM, Wonderlich SA, & Smith GT (2015). A risk and maintenance model for bulimia nervosa: From impulsive action to compulsive behavior. Psychological Review, 122, 516–535. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Pearson CM, Zapolski TC, & Smith GT (2015). A longitudinal test of impulsivity and depression pathways to early binge eating onset. International Journal of Eating Disorders, 48, 230–237. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Racine SE, & Martin SJ (2017). Integrating eating disorder-specific risk factors into the acquired preparedness model of dysregulated eating: A moderated mediation analysis. Eating Behaviors, 24, 54–60. [DOI] [PubMed] [Google Scholar]
  47. Rø Ø, Reas DL, & Rosenvinge J (2012). The impact of age and BMI on Eating Disorder Examination Questionnaire (EDE-Q) scores in a community sample. Eating Behaviors, 13, 158–161. [DOI] [PubMed] [Google Scholar]
  48. Satorra A, & Bentler PM (2010). Ensuring positiveness of the scaled difference chi-square test statistic. Psychometrika, 75, 243–248. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Smith GT, Simmons JR, Flory K, Annus AM, & Hill KK (2007). Thinness and eating expectancies predict subsequent binge-eating and purging behavior among adolescent girls. Journal of Abnormal Psychology, 116, 188–197. [DOI] [PubMed] [Google Scholar]
  50. Steiger JH, & Lind JC (1980, May). Statistically-based tests for the number of factors. Paper presented at the annual spring meeting of the psychometric society; Iowa City: Iowa. [Google Scholar]
  51. Steinglass JE, Sysko R, Glasofer D, Albano AM, Simpson HB, & Walsh BT (2011). Rationale for the application of exposure and response prevention to the treatment of anorexia nervosa. International Journal of Eating Disorders, 44, 134–141. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Stice E (2016). Interactive and mediational etiologic models of eating disorder onset: Evidence from prospective studies. Annual Review of Clinical Psychology, 12, 359–381. [DOI] [PubMed] [Google Scholar]
  53. Taylor CB, Bryson S, Luce KH, Cunning D, Doyle AC, Abascal LB,…Wilfley DE. (2006). Prevention of eating disorders in at-risk college-age women. Archives of General Psychiatry, 63, 881e888. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Tucker LR, & Lewis C (1973). A reliability coefficient for maximum likelihood factor analysis. Psychometrika, 38, 1e10. [Google Scholar]
  55. Vangsness L, Bry BH, & LaBouvie EW (2005). Impulsivity, negative expectancies, and marijuana use: A test of the acquired preparedness model. Addictive Behaviors, 30, 1071–1076. [DOI] [PubMed] [Google Scholar]
  56. Whiteside SP, & Lynam DR (2001). The five factor model and impulsivity: Using a structural model of personality to understand impulsivity. Personality and Individual Differences, 30, 669–689. [Google Scholar]

RESOURCES