Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Apr 1.
Published in final edited form as: Eat Behav. 2020 Feb 13;37:101369. doi: 10.1016/j.eatbeh.2020.101369

Emotion Differentiation and Disordered Eating Behaviors: The Role of Appearance Schemas

Gail A Williams-Kerver a,b,1, Janis H Crowther a
PMCID: PMC7246154  NIHMSID: NIHMS1564221  PMID: 32087556

Abstract

Emotion differentiation, or the ability to distinguish between discrete emotions in the moment, has been linked to maladaptive behaviors, including disordered eating. Appearance schemas may impact this relationship, as it has been suggested that individuals who are preoccupied with appearance-related information in their environment have limited attentional resources to devote to other internal processes. This study sought to expand existing research by examining: 1) the relationships between emotion differentiation and self-reported eating disorder symptomatology, and 2) strength of implicit appearance schemas as a moderator of these relationships. Participants were 118 female undergraduate students who completed a self-report disordered eating symptomatology questionnaire and a word stem completion task (measuring implicit appearance schemas) at baseline. Participants then reported their daily disordered eating behaviors and emotions through ecological momentary assessment for seven days. Emotion differentiation indices were calculated from negatively-valenced (NED) and positively-valened (PED) daily affect ratings using intraclass correlation coefficients. Analyses demonstrated significant relationships between NED, severity of eating disorder symptomology, and frequency of compensatory behaviors; however, these relationships did not emerge with PED. Strength of appearance schemas was a moderator, suggesting that poor NED paired with stronger appearance schemas resulted in more severe eating disorder symptoms and more frequent engagement in compensatory behaviors. Multilevel models revealed that better NED predicted daily engagement in dietary restriction. By examining the relationship between emotion differentiation and disordered eating symptoms, this study contributes clinically significant information regarding a facet of emotional experience that may be important to our understanding of eating disorder symptomatology.

Keywords: Emotion differentiation, binge eating, compensatory behaviors, appearance schemas, ecological momentary assessment (EMA)

1. Introduction

Within the eating disorder (ED) literature, considerable research has examined the emotion regulatory functions of binge eating and compensatory weight control behaviors. The affect regulation model of eating psychopathology suggests that binge eating serves to reduce negative affect through comfort and distraction (Hawkins & Clement, 1984). Other research purports that self-induced vomiting regulates increased anxiety following binge eating among women with bulimia nervosa (BN; Leitenberg, Gross, Peterson, & Rosen, 1984; Rosen & Leitenberg, 1982). These models suggest that increasing levels of aversive affect prompt maladaptive ED behaviors and that these behaviors subsequently alter the level of affect intensity, presumably in a beneficial manner. However, research is mixed on whether these maladaptive behaviors effectively alter affect intensity (Engel et al., 2013; Goldschmidt et al., 2014; Haedt-Matt & Keel, 2015) and is further complicated by methodological and measurement issues (Haedt-Matt & Keel, 2011). Consequently, the exploration of alternative perspectives on emotional experience may contribute novel information regarding the role of binge eating and inappropriate compensatory behaviors as maladaptive coping strategies among women with varying levels of ED symptomatology.

One alternative perspective on emotional experience focuses on emotion differentiation (or “emotion granularity”; Barrett, Gross, Christensen, & Benvenuto, 2001), which represents an individual’s ability to distinguish between discrete similarly-valenced emotional states during moment-to-moment experience. Individuals high in emotion differentiation can distinguish their emotional experience using discrete emotional terms (e.g., “anger” or “fear”) whereas individuals with low emotion differentiation describe their emotional experience in more global terms (e.g., “bad”; Barrett, 2006). In accordance with Feelings as Information Theory (Schwarz & Clore, 1983), which proposes that knowing how you are feeling provides information on how to act in a given context, poor emotion differentiation serves as a disadvantage in regulating emotional experience because individuals lack information needed to make judgments about behaviors that may facilitate emotion regulation (Kashdan, Barrett, & McKnight, 2015). Indeed, research has demonstrated associations between poor differentiation of negatively-valenced emotions (i.e., negative emotion differentiation, or “NED”) and more frequent engagement in maladaptive coping behaviors, including excessive alcohol use (Emery, Simons, Clarke, & Gaher, 2014; Kashdan, Ferssizidis, Collins, & Muraven, 2010), aggression (Pond et al., 2012), and non-suicidal self-injury (Zaki, Coifman, Rafaeli, Berenson, & Downey, 2013).

Emotion differentiation is assessed by examining associations between momentary ratings of emotions, which are typically collected during the day using ecological momentary assessment (EMA; Stone & Shiffman, 1994). EMA allows participants to record aspects of experience as they occur in their natural environment, thus reducing threats to reporting bias that can occur through the use of retrospective self-report measures. This methodological benefit becomes particularly important when distinguishing the construct of emotion differentiation from the trait construct alexithymia, which is defined as the difficulty in “identifying, understanding, and expressing feelings (Bagby, Taylor, & Ryan, 1986).” Alexithymia has been the focus of significant research within the eating disorder literature (see Nowakowski, McFarlane, & Cassin, 2013 for a review). However, alexithymia is most frequently assessed via retrospective self-report questionnaires, which rely on an individual’s metacognitive ability to accurately identify their aggregate emotional experience. As such, this creates potential retrospective recall biases that have implications towards the validity of the construct (Smidt & Suvak, 2015), further arguing for more direct and reliable measurement of how one identifies their feelings as they occur in the moment (i.e., emotion differentiation) rather than relying on an individual’s ability to characterize their own emotional capabilities.

To date, two published studies have examined the relationship between emotion differentiation and ED symptomology. Selby and colleagues (2014) examined relationships between NED, positive emotion differentiation (PED), and weight-loss behaviors among females with anorexia nervosa (AN). While NED significantly predicted the frequency of exercise and weigh-ins, PED predicted a greater variety of ED behaviors, including frequency of vomiting, laxative use, dietary restriction, exercise, and weigh-ins. Dixon-Gordon and colleagues (2014) examined relationships between NED, PED, and momentary urges to engage in binge eating and purging among individuals with varying levels of borderline personality disorder (BPD) psychopathology. They found a significant interaction between BPD symptoms and PED when predicting urges to engage in binge eating (but not purging); however, no significant associations with NED emerged. Taken together, these results suggest that PED is related to ED symptomatology. However, findings for NED were inconsistent, which is surprising given the vast support for the role of negative emotions in the development and maintenance of disordered eating behaviors (e.g., Haedt-Matt & Keel, 2011). Because NED has been the central focus of research on non-disordered eating maladaptive behaviors, further research is needed to not only replicate the significant relationships between PED and ED symptomatology among different populations, but also further examine the extent to which NED might relate to ED symptomatology.

1.1. The Role of Appearance Schemas

Self-schemas have been defined as cognition generalizations that “organize and guide the processing of self-related information contained in an individual’s social experience (p. 64; Markus, 1977).” Body weight, shape, and/or appearance is one domain through which schemas influence the processing of information about the self (Cash & Labarge, 1996; Vitousek & Hollon, 1990). Purportedly, weight and appearance schemas influence an individual’s “choice” of ED symptomatology and maintain ED psychopathology (Vitousek & Hollon, 1990). Indeed, research using a word stem completion task administered randomly during the day found that women with ED psychopathology took longer to choose appearance-related word stems than women without ED psychopathology, demonstrating that appearance schemas may interfere with the ability to process information efficiently on a day-to-day basis (Myers, Ridolfi, & Crowther, 2015). Additionally, research supports the presence of attentional biases toward appearance-related stimuli among women with EDs (e.g., Dobson & Dozois, 2004) and women with body dissatisfaction (e.g., Rodgers & DuBois, 2016), further highlighting the relevance of appearance schemas.

One question that arises is whether individuals who devote resources toward processing information through appearance schemas may be more limited in their ability to focus on other important internal experiences. Citrin, Roberts, and Fredrickson (2004) suggest that individuals who become preoccupied with body image information in their social environment have fewer resources available to attend to their “internal, body-based cues to emotion (p. 213).” Thus, individuals may have limited resources to devote to attending and processing internal experiences. To the extent that individuals generally have difficulty identifying their momentary emotional experiences, it seems likely that poor emotion differentiation would have a stronger relationship with disordered eating behaviors when more elaborated appearance schemas are present and activated. Consequently, further exploration of the interaction between recognition of one’s emotional state and implicit appearance schemas, including how this might influence engagement in disordered eating, is needed.

1.2. Current Study

Given its relevance for emotion regulation (e.g., Kashdan et al., 2015) and previous inconsistencies in the ED literature (Dixon-Gordon et al., 2014; Selby et al., 2014), this research examined NED and PED among college-aged women, a population at risk for EDs and among whom ED symptomatology is relatively prevalent (Crowther, Armey, Luce, Dalton, & Leahey, 2008; Keel, Heatherton, Dorer, Joiner, & Zalta, 2006; Pyle, Neuman, Halvorson, & Mitchell, 1991). Our first aim was to examine associations between emotion differentiation and baseline self-report of ED symptomatology. To examine novel associations between emotion differentiation and a range of affective, behavioral, and cognitive ED symptoms, we included severity of ED symptomatology as our primary outcome. However, given previous research, we also included baseline frequency of binge eating and compensatory behaviors as outcome variables. The second aim was to examine the strength of appearance schemas as a moderator of the relationship between emotion differentiation and ED symptomatology, as individuals who have difficulty differentiating their emotions may experience ED symptomatology when they have more elaborated appearance schemas. We hypothesized that poor NED and PED would be associated with more severe ED symptomatology (hypothesis 1) and that these relationships would be strengthened at higher levels of appearance schemas (hypothesis 2). The third aim was to examine the relationship between emotion differentiation and the engagement in daily disordered eating behaviors assessed via EMA and also examine strength of appearance schemas as a moderator of these relationships. It was hypothesized that poor NED and PED would be associated with the engagement in daily disordered eating behaviors (hypothesis 3) and that baseline scores for appearance schemas would moderate these daily relationships (hypothesis 4).

2. Material and Methods

2.1. Participants

Participants were 118 female undergraduates (MAge = 19.39 years [SD = 1.56]; MBMI = 24.92 kg/m2 [SD = 6.02]; 80.5% Caucasian) from a large Midwest university. Data from seven of the original 125 participants were omitted from analyses because: English was her second language, thus presenting a potential confound with the emotion differentiation construct (n =2); ratings on several baseline measures were missing (n =1); and daily diary compliance was below the required threshold (n = 4; see Results for explanation).

2.2. Measures

2.2.1. Baseline measures.

2.2.1.1. ED symptomatology.

The Eating Disorder Diagnostic Scale (EDDS; Stice, Telch, & Rizvi, 2000) is a 22-item self-report measure used to assess ED symptomatology. An algorithm created by Stice and colleagues (2000) yields a continuous symptom composite score, representing severity of overall ED symptoms. Items assessing the frequency of binge eating (item 8) and compensatory behaviors (sum of items 15-18) were also used in analyses. The EDDS has excellent internal consistency (α = .91) and test-retest reliability (r = .87) for the symptom composite score. For this sample, Cronbach’s alpha for the symptom composite score was .82.

2.2.1.2. Appearance schemas.

Participants completed a 20-item word stem completion task described as “a measure of implicit processing of appearance-related information (p. 73; Tiggemann, Hargreaves, Polivy, & McFarlane, 2004).” Each word stem (e.g., PRE__) can be completed by an appearance-related (e.g., pretty) or non-appearance related (e.g., present) word. A total score was created by summing the number of appearance-related words, with higher scores indicating stronger implicit appearance schemas. The word stem completion task is sensitive to the presence and activation of appearance-related schemas (e.g., Hargreaves & Tiggemann, 2002; Tiggemann et al., 2004). Moreover, given that undergraduate women produced, on average, 3.18 to 3.85 appearance-related words in control conditions (Tiggemann et al., 2004), it appears that appearance schemas produce weight- and appearance-related words in response to word stems independent of the presentation of specific stimuli designed to activate these schemas. As such, we utilized the word stem task as a measure of strength for implicit appearance schemas.

2.2.2. Daily diary measures.

2.2.2.1. Emotions.

During each EMA signal, participants were asked to rate emotion items on a 5-point Likert scale, ranging from 1 (“Very Slightly or Not at All”) to 5 (“Extreme”). Because there is no standardized procedure for selecting emotion items across research protocols (Brose, Schmiedek, Gerstorf, & Voelkle, 2019; Smidt & Suvak, 2015), emotion terms in this study were identified from items used in similar research (e.g., Dixon-Gordon et al., 2014; Pond et al., 2012; Selby et al., 2014; Zaki et al., 2013). Next, in accordance with contemporary models of affect (e.g., Feldman Barrett & Russell, 1998; Russell, 1980) and seminal theory and research on emotion differentiation (Barrett, 1998; Barrett et al., 2001; Feldman, 1995), the final emotion items were selected to account for varying levels of arousal across the dimension of valence, an approach consistent with other research on emotion differentiation (e.g., Pond et al., 2012; Zaki et al., 2013). The validity of these items is evident from their unique position in a multitude of two-dimensional models (e.g., Feldman Barrett & Russell, 1998; Russell, 1980; Watson, 2000). Negative emotion terms included angry, ashamed, bored, guilty, nervous, and sad. Positive emotion terms included alert, calm, enthusiastic, excited, happy, relaxed, and satisfied.

2.2.2.2. Disordered Eating Behaviors.

Upon awakening each day, participants initiated an EMA signal to indicate which of nine behaviors, adapted from a health behavior checklist (Vickers, Conway, & Hervig, 1990), they engaged in during the previous day. Included in the behaviors were items assessing loss of control (LOC) eating (“felt a loss of control while eating”) and dietary restriction (“restricted the type or quantity of food that I consumed”).

2.3. Apparatus

Participants downloaded “The Personal Analytics Companion (PACO)” application on their smartphone (Evans, 2014) to complete the EMA protocol.

2.4. Procedure

This research was approved by the university’s Institutional Review Board. Using an online recruitment system, participants were invited to participate if they were female, aged 18 or greater, and owned a smartphone. Following informed consent, participants completed the baseline measures and were trained on the EMA application.

A time- and event-based EMA design was utilized, including one self-initiated morning prompt and four semi-random prompts per day for seven days. Random prompts occurred at least 180 minutes apart during waking hours, and participants were given 30 minutes to respond to signals. Participants were instructed to avoid responding during times when it might be unsafe (e.g., while driving). Following study completion, participants returned to the laboratory to receive compensation and information on their compliance.

2.5. Data Analysis Plan

To test hypotheses 1 and 2, “study-level” aggregate indices for NED and PED were calculated for each participant by conducting a series of intraclass correlations (ICCs) with absolute agreement across all scores for negative and positive emotion terms (Kashdan et al., 2010; Selby et al., 2014; Tugade, Fredrickson, & Barrett, 2004). To test hypotheses 3 and 4, “daily-level” indices for NED and PED were calculated through a series of ICCs using the scores for negative and positive affect items in a given day. For ease of interpretation, all NED and PED indices were transformed so that higher values were indicative of greater emotion differentiation.

To examine hypothesis 2, a series of three hierarchical regressions (using the EDDS symptom composite score, frequency of binge eating, or frequency of compensatory behaviors as the dependent variables) were conducted (Baron & Kenny, 1986). Average level of affect intensity (NA and PA) were entered as covariates into each model. Moderation was tested by creating an interaction term from the centered predictor variable (PED or NED) and the centered moderator (appearance schemas). For each analysis, both NA and PA intensity were entered simultaneously into Block 1, the centered NED, PED, and appearance schema variables were entered into Block 2, and the NED by Appearance Schemas and PED by Appearance Schemas interaction terms were entered into Block 3.

To examine hypothesis 3, hierarchical linear models (HLM; Raudenbush & Bryk, 2002) with Bernoulli equations were conducted. This analytical approach to daily analyses, which controlled for affect intensity and included both negative and positive affect variables into the same model, is consistent with the approach taken by Selby and colleagues (2014) with their EMA data. Because participants indicated whether they engaged in a maladaptive behavior during the morning signal of the following day, the emotion differentiation variables were lagged so that NED, PED, and eating behaviors referenced the same day. Predictor variables were entered as group-mean centered to control for within person variation in scores across signals. For each daily disordered eating behavior (i.e., LOC eating or dietary restriction), the HLM equation was:

log[(YDaily ED behavior/1−YDaily ED behaviors)ij]t=β0j+β1j(NA intensity)(t−1)j+β2j(PA intensity)(t−1)j+β3j(NED)(t−1)j+β4j(PED)(t−1)j+rij

where Yij represents the dependent variable score for participant j’s i’th diary entry, β0j represents the intercept for the log of odds, β1-4j represents the slope for each predictor variable, and rij represents the random error.

To examine hypothesis 4, Bernoulli model HLM analyses were conducted. The lagged NA intensity, PA intensity, NED, and PED variables were group-mean centered and entered as predictors on Level 1, and the appearance schema variable was grand-mean centered and entered as a predictor on Level 2. The test of moderation included a cross-level interaction between the appearance schema and NED/PED variables. The full HLM equation was:

log[(YDaily ED behavior/1−YDaily ED behaviors)i]t=β00+β10(NA intensity)(t−1)i+β20(PA intensity)(t−1)i+β30(NED)(t−1)i+β40(PED)(t−1)i+β01(appearance schema)i+β31(NEDt−1*appearance schema)i+β41(PEDt−1*appearance schema)i+r0i

3. Results

3.1. Data Quality and Baseline Correlations

Consistent with previous research (Ridolfi, Myers, Crowther, & Ciesla, 2011), participants had to complete at least 20% of their signals to be included in analyses. Subsequently, data from four participants were dropped, resulting in an average compliance rating of 79.71%1. With respect to hypothesis 1, bivariate correlations revealed that NED, but not PED, was significantly associated with the EDDS symptom composite score and frequency of compensatory behaviors (Table 1)2.

Table 1.

Descriptive information and intercorrelations between key variables.

Descriptives
 Intercorrelations
M SD 1. 2. 3. 4. 5. 6. 7. 8.
 1. NED 0.50 0.34 --
 2. PED 0.27 0.19 0.29** --
 3. NA Intensity 3.08 0.67 −0.18 0.00 --
 4. PA Intensity 3.62 0.60 −0.02 0.25** 0.00 --
 5. EDDS Symptom Composite 19.89 11.92 −0.20* −0.03 0.22* −0.08 --
 6. Binge Eating 1.36 1.86 −0.04 −0.04 0.02 −0.13 0.53*** --
 7. Compensatory Behaviors 2.04 3.50 −0.25** −0.07 0.16 −0.10 0.64*** 0.19* --
 8. Appearance Schema 3.42 1.83 0.09 −0.07 −0.10 −0.04 0.12 0.01 0.16 --
Open in a new tab
*

p < .05;

**

p < .01;

***

p <.001.

“NA” indicates negative affect. “PA” indicates positive affect. “EDDS” indicates Eating Disorder Diagnostic Scale.

3.2. Positive Emotion Differentiation (PED), Negative Emotion Differentiation (NED), and Baseline Eating Disorder Symptoms

Regarding the three hierarchical regressions used to test hypothesis 2, all results for PED showed that the main effects and interaction terms were nonsignificant when the EDDS symptom composite score, frequency of binge eating, and frequency of compensatory behaviors were dependent variables (Tables 2 and 3).

Table 2.

Hierarchical regression moderation analysis predicting overall eating disorder symptom severity.

Predictor R2 ΔR2 df ΔF B SE β
Step 1 0.53 0.53 2, 113 3.16*
   NA Intensity 3.88 1.63 0.22*
   PA Intensity −1.47 1.82 −0.07
Step 2 0.11 0.06 3, 110 2.36
   NED −7.17 3.43 −0.20*
   PED 3.80 6.15 0.06
   Appearance Schemas 1.10 0.60 0.17
Step 3 0.17 0.06 2, 108 4.09*
   NED x Appearance Schemas −5.00 1.93 −0.24*
   PED x Appearance Schemas 4.99 3.16 0.14
Open in a new tab
*

p < .05.

“NA” denotes negative affect. “PA” denotes positive affect. “NED” denotes negative emotion differentiation. “PED” denotes positive emotion differentiation.

Table 3.

Hierarchical regression moderation analyses predicting frequency of binge eating and compensatory behaviors.

Binge Eating
 Compensatory Behaviors
Predictor R2 ΔR2 df ΔF B SE β R2 ΔR2 df ΔF B SE β
Step 1 0.02 0.02 2,113 0.97 0.04 0.04 2,113 2.10
   NA Intensity 0.04 0.26 0.02 0.82 0.48 0.16
   PA Intensity −0.40 0.29 −0.13 −0.61 0.54 −0.11
Step 2 0.02 0.002 3,110 0.07 0.12 0.09 3,110 3.63*
   NED −0.24 0.56 −0.04 −2.63 1.00 −0.25*
   PED 0.02 1.01 0.002 0.84 1.79 0.05
   Appearance Schemas 0.004 0.10 0.004 0.38 0.18 0.20*
Step 3 0.02 0.001 2,108 0.06 0.18 0.05 2,108 3.52*
   NED x Appearance Schemas −0.01 0.33 −0.003 −1.50 0.56 −0.24**
   PED x Appearance Schemas 0.18 0.54 0.03 0.18 0.93 0.02
Open in a new tab
*

p < .05,

**

p < .01.

“NA” denotes negative affect. “PA” denotes positive affect. “NED” denotes negative emotion differentiation. “PED” denotes positive emotion differentiation.

For NED, the hierarchical regression using the EDDS symptom composite score as the dependent variable revealed a significant main effect for NA intensity and NED, but not for appearance schemas (Table 2). The interaction term between NED and appearance schemas was significant, accounting for an additional 6.0% of the overall variance. A decomposition of the interaction term showed that the simple slope for high appearance schemas (indicated by scores +1 SD above the mean) was significant, but the simple slope for low appearance schemas (−1 SD below the mean) was nonsignificant (Figure 1). At stronger levels of appearance schemas, poorer NED was associated with greater severity of ED symptomatology.

Figure 1.

Figure 1.

Open in a new tab

Appearance schemas moderate the relationship between NED and severity of eating disorder symptoms.

Note: “EDDS” denotes Eating Disorder Diagnostic Scale. “NED” denotes negative emotion differentiation. Simple slope results: low appearance schemas, t(117) = 0.70, p = 0.483; high appearance schemas, t(117) = −5.04, p < .001.

Results for NED and binge eating showed that the overall model, main effects, and interaction term were nonsignificant (Table 3). However, with frequency of compensatory behaviors as the dependent variable, results demonstrated a significant main effect for NED and appearance schemas (Table 3). The interaction term between NED and compensatory behaviors was significant, accounting for an additional 5.0% of the overall variance. Results of a decomposition analysis showed that the simple slope for high, but not low, appearance schemas was significant (Figure 2). At stronger levels of appearance schemas, poor NED was significantly associated with greater frequency of compensatory behaviors.

Figure 2.

Figure 2.

Open in a new tab

Appearance schemas moderate the relationship between NED and frequency of compensatory behaviors.

Note: “EDDS” denotes Eating Disorder Diagnostic Scale. “NED” denotes negative emotion differentiation. Simple slope results: low appearance schemas, t(117) = 0.01, p = .991; high appearance schemas, t(117) = −3.23, p = .002.

a. Daily Disordered Eating Behaviors

Participants endorsed LOC eating during 22 morning signals (3.5%) and dietary restriction during 77 morning signals (12.1%). Results of the HLM analysis examining daily NED and PED as predictors of engagement in daily LOC eating, while controlling for daily average level of NA and PA, were not significant (see Table 4). However, with daily dietary restriction as the dependent variable, better NED significantly predicted engagement in dietary restriction after controlling for the average level of daily NA and PA intensity.

Table 4.

Results of Hierarchical Linear Models Predicting Likelihood of Engagement in Daily Disordered Eating Behaviors.

LOC Eating
 Dietary Restriction
β SE t p β SE t p
Model 1: Emotion Differentiation Effects
  NA −0.40 0.49 −0.82 .412 0.06 0.26 0.23 .822
  PA −0.22 0.39 −0.56 .573 −0.08 0.34 −0.25 .807
  NED 0.07 0.07 0.96 .339 0.10 0.05 2.08 .039*
  PED −0.06 0.12 −0.49 .628 0.01 0.09 0.11 .914
Model 2: Appearance Schema Moderation
  NA −0.40 0.51 −0.79 .429 0.03 0.27 0.09 .928
  PA −0.22 0.40 −0.56 .576 −0.12 0.35 0.34 .737
  NED 0.07 0.11 0.66 .512 0.07 0.06 1.16 .249
  PED −0.06 0.14 0.41 .679 0.02 0.09 0.25 .804
  Appearance Schema 0.08 0.12 0.68 .497 0.14 0.10 1.46 .146
  NED x Appearance Schema −0.01 0.04 0.19 .847 0.04 0.06 0.68 .497
  PED x Appearance Schema −0.04 0.08 0.50 .615 −0.01 0.05 −0.22 .824
Open in a new tab
*

p < .05.

“LOC” denotes loss of control. “NA” denotes negative affect. “PA” denotes positive affect. “NED” denotes negative emotion differentiation. “PED” denotes positive emotion differentiation.

Results of HLM analyses examining baseline appearance schemas as a moderator of the relationships between daily emotion differentiation and LOC eating or dietary restriction were nonsignificant (Table 4).

4. Discussion

4.1. Baseline Eating Disorder Symptomatology, Binge Eating and Compensatory Behaviors

In partial support of hypotheses, results indicated that poorly differentiated negative emotions, but not positive emotions, were significantly related to both severity of ED symptomatology and frequency of compensatory weight control behaviors. These NED findings are consistent with research examining the relationships between emotion differentiation and maladaptive behaviors (e.g., Emery et al., 2014; Kashdan et al., 2010; Pond et al., 2012; Selby et al., 2014; Zaki et al., 2013), providing further support that difficulties in differentiating between negative emotions may inhibit individuals’ access to information about adaptive means to regulate their emotions (Barrett et al., 2001; Kashdan et al., 2015). However, in previous research on ED symptomatology, PED has been a more consistent predictor than NED (Dixon-Gordon et al., 2014; Selby et al., 2014). There may be two explanations for the discrepancy between our PED findings and those of previous research. First, with respect to sampling, this study recruited undergraduate women while previous research used samples of individuals with AN (Selby et al., 2014) and BPD (Dixon-Gordon et al., 2014). It is possible that emotion differentiation may function differently among clinical samples with varying types and levels of symptom severity. Second, with respect to measurement, different emotion terms were used in each study. There is substantial variability within the emotion literature regarding the type and quantity of emotion terms utilized, with some papers using emotion terms spanning the dimensions of valence and arousal and others using terms based on subscales from established measures. Though it is not uncommon for emotion terms to vary across studies (e.g., Brose et al., 2019), the use of alternative negative and positive emotion terms may contribute to the discrepant findings.

Interestingly, poorer NED was significantly associated with baseline frequencies of compensatory behaviors but not binge eating. This finding argues that compensatory behaviors might function differently than binge eating with respect to their role in emotion regulation. While inconsistent with the findings regarding compensatory behaviors in AN from Selby and colleagues (2014), our finding complements previous research on trajectories of affective experience among women diagnosed with AN, suggesting possible mechanistic differences in the affective experiences surrounding binge eating and purging (Engel et al., 2013). However, no significant differences in NA trajectories surrounding binge eating, purging, and binge-purge episodes were found in a separate study among women diagnosed with BN (Berg et al., 2013). Therefore, additional research is needed to further explore these affective experiences and how they might differ as precipitants to binge eating and purging behaviors in a variety of clinical samples.

4.2. Appearance Schemas

Results also indicated that appearance schemas moderated the relationships between NED and ED symptomatology. Specifically, when appearance schema activation was high but not low, the relationships between NED and eating disorder symptomatology and NED and frequency of compensatory behaviors were particularly strong. Consistent with Citrin and colleagues (2004), thse results suggest that, when individuals with poorer negative emotion differentiation are dedicating their attentional resources to incoming appearance-related information, they experience greater ED symtomatology and are more likely to engage in compensatory behaviors (i.e., an inappropriate emotion regulation strategy; Kashdan et al., 2015). However, when appearance schema activation is low, individuals may be able to direct their attentional resources toward other internal states, including emotions, thus weakening the relationship between emotion differentiation and the engagement in disordered eating behaviors.

Further, the fact that appearance schemas not only emerged as a main effect but, more importantly, moderated the relationship between NED and frequency of compensatory behavior suggests that motivation to maintain a thin weight status may be important. Indeed, compensatory behaviors often function as a means of counteracting or protecting against weight gain (American Psychiatric Association, 2013). Thus, in light of the pressure to maintain a thin weight status, poor emotion differentiation may be most strongly associated with maladaptive eating behaviors that are believed to result in weight loss (i.e., compensatory behaviors).

4.3. Daily Behaviors

Finally, better NED was associated with dietary restriction, but not LOC eating, on a daily basis. This finding is counterintuitive given that research has supported that poor emotion differentiation leads to maladaptive behaviors. One explanation for our finding is that participants may have understood the dietary restriction question differently than originally intended. It may be that restricting the “type” (e.g., foods high in fat or carbohydrates) and/or “quantity” (e.g., portion size and/or caloric intake) of food was perceived as “healthy” eating habits rather than the extreme restriction seen in EDs. If this was the case, differentiating between negative emotions was subsequently associated with adherence to healthier eating patterns. Conversely, if participants endorsed dietary restriction because they were severely limiting the types and quantities of food they consumed, an alternative explanation may be that individuals use dietary restriction to cope with a more clearly identified negative emotional experience. Finally, it is possible that better emotion differentiation provides the motivation to engage in dietary restriction. However, given the relatively low number of episodes of LOC eating and dietary restriction in diary analyses, these findings should be regarded as preliminary.

4.4. Limitations, Future Directions, and Conclusions

One limitation of this research is that the sample was comprised of undergraduate women. Although it is important to study ED symptomatology among college women, these findings cannot be generalized to men or clinical ED populations. Future research should continue exploring the relationships between emotion differentiation, appearance schema activation, and disordered eating behaviors among individuals with eating disorders, including BN or binge eating disorder. Another limitation includes the dichotomous assessment of daily disordered eating behaviors, which limited our ability to draw conclusions regarding the relationship between emotion differentiation and the frequency and/or severity of these and other maladaptive eating behaviors. Furthermore, less than 15% of EMA signals included an endorsement of LOC eating or dietary restriction, which likely diminished our power to detect more significant results in daily analyses. Future research would benefit from the inclusion and more frequent assessment of these and other disordered eating behaviors, including binge eating and other compensatory strategies, in the EMA protocol. Finally, we did not measure appearance schema activation in our EMA protocol, which may explain why the daily moderation analyses were non-significant. Given a novel method to measure appearance schemas in the naturalistic environment (Myers et al., 2015), future research should seek to include momentary appearance schems in similar EMA protocols.

In conclusion, this research provides novel empirical data supporting the relationship between negative emotion differentiation, ED symptomatology, and appearance schemas. Future research in this area may continue to refine our understanding of how the relationship between knowing how one is feeling in the moment and maladaptive eating behaviors contribute to emotion regulation. Further, these results argue for the importance of addressing appearance schemas in both ED prevention efforts and psychotherapy treatment, as mitigating the importance of appearance to one’s self-concept may help to reduce or prevent the impact of emotional experience on behaviors related to ED symptomatology.

Highlights.

  • Poor negative emotion differentiation relates to eating disorder symptomatology

  • Stronger appearance schemas moderate these relationships

  • Better daily negative emotion differentiation predicts engagement in dietary restriction

Acknowledgements:

This research was supported by the Judie Fall Lasser Graduate Psychology Research Award from Kent State University and by the National Institutes of Mental Health [T32 MH08276].

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

1

In the interest of conducting sensitivity analyses, we imputed missing data using Maximum Likelihood estimation. We then recalculated the NED and PED indices using imputed data and ran Pearson bivariate correlations between the new emotion differentiation variables and the other variables of interest in our study. The pattern of correlational results were the same as those using the non-imputed emotion differentiation variables (see Table 1). Therefore, to be consistent with recommendations for longitudinal data (e.g., Bolger & Laurenceau, 2013) and other EMA research in the field of eating disorders (e.g., Smyth et al., 2001) all analyses were conducted using non-imputed data. Imputed data is available on request.

2

BMI was considered as a potential covariate in analyses but excluded due to its non-significant relationship with both NED (r = −.060, p = .520) and PED (r = −.066, p = .479).

References

  1. American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: Author. [Google Scholar]
  2. Bagby M, Taylor GJ, & Ryan D (1986). Toronto Alexithymia Scale: Relationship with personality and psychopathology measures. Psychotherapy and Psychosomatics, 45(4), 207–215. 10.1159/000287950 [DOI] [PubMed] [Google Scholar]
  3. Baron RM, & Kenny DA (1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51(6), 1173–1182. 10.1037/0022-3514.51.6.1173 [DOI] [PubMed] [Google Scholar]
  4. Barrett LF (1998). Discrete emotions or dimensions? The role of valence focus and arousal focus. Cognition & Emotion, 12(4), 579–599. 10.1080/026999398379574 [DOI] [Google Scholar]
  5. Barrett LF (2006). Solving the emotion paradox: Categorization and the experience of emotion. Personality and Social Psychology Review, 10(1), 20–46. 10.1207/s15327957pspr1001_2 [DOI] [PubMed] [Google Scholar]
  6. Barrett LF, Gross J, Christensen TC, & Benvenuto M (2001). Knowing what you’re feeling and knowing what to do about it: Mapping the relation between emotion differentiation and emotion regulation. Cognition & Emotion, 15(6), 713–724. 10.1080/02699930143000239 [DOI] [Google Scholar]
  7. Berg KC, Crosby RD, Cao L, Peterson CB, Engel SG, Mitchell JE, & Wonderlich SA (2013). Facets of negative affect prior to and following binge-only, purge-only, and binge/purge events in women with bulimia nervosa. Journal of Abnormal Psychology, 122(1), 111–118. 10.1037/a0029703 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Bolger N, & Laurenceau J (2013). Intensive longitudinal methods: An introduction to diary and experience sampling research. Guilford Press. [Google Scholar]
  9. Brose A, Schmiedek F, Gerstorf D, & Voelkle MC (2019). The measurement of within-person affect variation. Emotion, Advance Online Publication. 10.1037/emo0000583 [DOI] [PubMed] [Google Scholar]
  10. Cash TF, & Labarge AS (1996). Development of the Appearance Schemas Inventory: A new cognitive body-image assessment. Cognitive Therapy and Research, 20(1), 37–50. 10.1007/BF02229242 [DOI] [Google Scholar]
  11. Citrin LB, Roberts TA, & Fredrickson BL (2004). Objectification theory and emotions: A feminist psychological perspective on gendered affect In Tiedens LZ & Leach CW (Eds.), The social life of emotions (pp. 203–223). Cambridge, UK: Cambridge University Press. [Google Scholar]
  12. Crowther JH, Armey M, Luce KH, Dalton GR, & Leahey T (2008). The point prevalence of bulimic disorders from 1990 to 2004. International Journal of Eating Disorders, 41(6), 491–497. 10.1002/eat.20537 [DOI] [PubMed] [Google Scholar]
  13. Dixon-Gordon KL, Chapman AL, Weiss NH, & Rosenthal MZ (2014). A preliminary examination of the role of emotion differentiation in the relationship between borderline bersonality and urges for maladaptive behaviors. Journal of Psychopathology and Behavioral Assessment, 36(4), 616–625. 10.1007/s10862-014-9423-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Dobson KS, & Dozois DJA (2004). Attentional biases in eating disorders: A meta-analytic review of Stroop performance. Clinical Psychology Review, 23(8), 1001–1022. 10.1016/j.cpr.2003.09.004 [DOI] [PubMed] [Google Scholar]
  15. Emery NN, Simons JS, Clarke CJ, & Gaher RM (2014). Emotion differentiation and alcohol-related problems: The mediating role of urgency. Addictive Behaviors, 39(10), 1459–1463. 10.1016/j.addbeh.2014.05.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Engel SG, Wonderlich SA, Crosby RD, Mitchell JE, Crow S, Peterson CB, … Gordon KH (2013). The role of affect in the maintenance of anorexia nervosa: Evidence from a naturalistic assessment of momentary behaviors and emotion. Journal of Abnormal Psychology, 122(3), 709–719. 10.1037/a0034010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Evans B (2014). The Personal Analytics Companion (PACO). [Mobile Smartphone Software]. Retrieved from https://www.pacoapp.com
  18. Feldman Barrett L, & Russell JA (1998). Independence and bipolarity in the structure of current affect. Journal of Personality and Social Psychology, 74(4), 967–984. 10.1037/0022-3514.74.4.967 [DOI] [Google Scholar]
  19. Feldman LA (1995). Valence focus and arousal focus: Individual differences in the structure of affective experience. Journal of Personality and Social Psychology, 69(1), 153–166. 10.1037/0022-3514.69.1.153 [DOI] [Google Scholar]
  20. Goldschmidt AB, Wonderlich S. a, Crosby RD, Engel SG, Lavender JM, Peterson CB, … Mitchell JE. (2014). Ecological momentary assessment of stressful events and negative affect in bulimia nervosa. Journal of Consulting and Clinical Psychology, 82(1), 30–39. 10.1037/a0034974 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Haedt-Matt AA, & Keel PK (2011). Revisiting the affect regulation model of binge eating: A meta-analysis of studies using ecological momentary assessment. Psychological Bulletin, 137(4), 660–681. 10.1037/a0023660 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Haedt-Matt AA, & Keel PK (2015). Affect regulation and purging: An ecological momentary assessment study in purging disorder. Journal of Abnormal Psychology, 124(2), 399–411. 10.1037/a0038815 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Hargreaves D, & Tiggemann M (2002). The effect of television commercials on mood and body dissatisfaction: The role of appearance-schema activation. Journal of Social and Clinical Psychology, 21(3), 287–308. 10.1521/jscp.21.3.287.22532 [DOI] [Google Scholar]
  24. Hawkins RC, & Clement PF (1984). Binge eating: Measurement problems and a conceptual model In Hawkins RC, Fremouw WJ, & Clement PF (Eds.), The binge purge syndrome: Diagnosis, treatment, and research. (pp. 229–251). New York, NY: Springer. [Google Scholar]
  25. Kashdan TB, Barrett LF, & McKnight PE (2015). Unpacking emotion differentiation: Transforming unpleasant experience by perceiving distinctions in negativity. Current Directions in Psychological Science, 24(1), 10–16. 10.1177/0963721414550708 [DOI] [Google Scholar]
  26. Kashdan TB, Ferssizidis P, Collins RL, & Muraven M (2010). Emotion differentiation as resilience against excessive alcohol use: An ecological momentary assessment in underage social drinkers. Psychological Science, 21(9), 1341–1347. 10.1177/0956797610379863 [DOI] [PubMed] [Google Scholar]
  27. Keel PK, Heatherton TF, Dorer DJ, Joiner TE, & Zalta AK (2006). Point prevalence of bulimia nervosa in 1982, 1992, and 2002. Psychological Medicine, 36(1), 119–127. 10.1017/S0033291705006148 [DOI] [PubMed] [Google Scholar]
  28. Leitenberg H, Gross J, Peterson J, & Rosen JC (1984). Analysis of an anxiety model and the process of change during exposure plus response prevention treatment of bulimia nervosa. Behavior Therapy, 15(1), 3–20. 10.1016/S0005-7894(84)80038-6 [DOI] [Google Scholar]
  29. Markus H (1977). Self-schemata and processing information about the self. Journal of Personality and Social Psychology, 35(2), 63–78. 10.1037/0022-3514.35.2.63 [DOI] [Google Scholar]
  30. Myers TA, Ridolfi DR, & Crowther JH (2015). Reaction times to appearance-related or non-appearance-related word choice among women with and without eating psychopathology. Cognitive Therapy and Research, 39(2), 204–214. 10.1007/s10608-014-9653-5 [DOI] [Google Scholar]
  31. Nowakowski ME, McFarlane T, & Cassin S (2013). Alexithymia and eating disorders: a critical review of the literature. Journal of Eating Disorders, 1(1), 21 10.1186/2050-2974-1-21 [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Pond RS, Kashdan TB, DeWall CN, Savostyanova A, Lambert NM, & Fincham FD (2012). Emotion differentiation moderates aggressive tendencies in angry people: A daily diary analysis. Emotion, 12(2), 326–337. 10.1037/a0025762 [DOI] [PubMed] [Google Scholar]
  33. Pyle RL, Neuman PA, Halvorson PA, & Mitchell JE (1991). An ongoing cross-sectional study of the prevalence of eating disorders in freshman college students. International Journal of Eating Disorders, 10(6), 667–677. [DOI] [Google Scholar]
  34. Raudenbush SW, & Bryk AS (2002). Hierarchical linear models : applications and data analysis methods. Sage Publications. [Google Scholar]
  35. Ridolfi DR, Myers TA, Crowther JH, & Ciesla JA (2011). Do appearance focused cognitive distortions moderate the relationship between social comparisons to peers and media images and body image disturbance? Sex Roles, 65(7), 491–505. 10.1007/s11199-011-9961-0 [DOI] [Google Scholar]
  36. Rodgers RF, & DuBois RH (2016). Cognitive biases to appearance-related stimuli in body dissatisfaction: A systematic review. Clinical Psychology Review, 46, 1–11. 10.1016/J.CPR.2016.04.006 [DOI] [PubMed] [Google Scholar]
  37. Rosen JC, & Leitenberg H (1982). Bulimia nervosa: Treatment with exposure and response prevention. Behavior Therapy, 13(1), 117–124. 10.1016/S0005-7894(82)80055-5 [DOI] [Google Scholar]
  38. Russell JA (1980). A circumplex model of affect. Journal of Personality and Social Psychology, 39(6), 1161–1178. 10.1037/h0077714 [DOI] [Google Scholar]
  39. Schwarz N, & Clore GL (1983). Mood, misattribution, and judgments of well-being: Informative and directive functions of affective states. Journal of Personality and Social Psychology, 45(3), 513–523. 10.1037/0022-3514.45.3.513 [DOI] [Google Scholar]
  40. Selby EA, Wonderlich SA, Crosby RD, Engel SG, Panza E, Mitchell JE, … Le Grange D (2014). Nothing tastes as good as thin feels: Low positive emotion differentiation and weight-loss activities in anorexia nervosa. Clinical Psychological Science, 2(4), 514–531. 10.1177/2167702613512794 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Smidt KE, & Suvak MK (2015). A brief, but nuanced, review of emotional granularity and emotion differentiation research. Current Opinion in Psychology, 3, 48–51. 10.1016/j.copsyc.2015.02.007 [DOI] [Google Scholar]
  42. Smyth J, Wonderlich S, Crosby R, Miltenberger R, Mitchell J, & Rorty M (2001). The use of ecological momentary assessment approaches in eating disorder research. International Journal of Eating Disorders, 30(1), 83–95. 10.1002/eat.1057 [DOI] [PubMed] [Google Scholar]
  43. Stice E, Telch CF, & Rizvi SL (2000). Development and validation of the Eating Disorder Diagnostic Scale: A brief self-report measure of anorexia, bulimia, and binge-eating disorder. Psychological Assessment, 12(2), 123–131. 10.1037//1040-3590.12.2.123 [DOI] [PubMed] [Google Scholar]
  44. Stone AA, & Shiffman S (1994). Ecological momentary assessment (EMA) in behavioral medicine. Annals of Behavioral Medicine, 16(3), 199–202. [Google Scholar]
  45. Tiggemann M, Hargreaves D, Polivy J, & McFarlane T (2004). A word-stem completion task to assess implicit processing of appearance-related information. Journal of Psychosomatic Research, 57(1), 73–78. 10.1016/S0022-3999(03)00565-8 [DOI] [PubMed] [Google Scholar]
  46. Tugade M, Fredrickson BL, & Barrett LF (2004). Psychological resilience and positive emotional granularity: Examining the benefits of positive emotions on coping and health, 72(6), 1161–1190. 10.1126/scisignal.2001449.Engineering [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Vickers RR, Conway TL, & Hervig LK (1990). Demonstration of replicabile dimensions of health behaviors. Preventive Medicine, 19(4), 377–401. 10.1016/0091-7435(90)90037-K [DOI] [PubMed] [Google Scholar]
  48. Vitousek KB, & Hollon SD (1990). The investigation of schematic content and processing in eating disorders. Cognitive Therapy and Research, 14(2), 191–214. 10.1007/BF01176209 [DOI] [Google Scholar]
  49. Watson D (2000). Mood and temperament. New York, NY: The Guilford Press. [Google Scholar]
  50. Zaki LF, Coifman KG, Rafaeli E, Berenson KR, & Downey G (2013). Emotion differentiation as a protective factor against nonsuicidal self-injury in borderline personality disorder. Behavior Therapy, 44(3), 529–540. 10.1016/j.beth.2013.04.008 [DOI] [PubMed] [Google Scholar]

RESOURCES