Abstract
Objective:
The relationship of dietary restraint in increasing risk for binge eating among individuals with binge-spectrum eating disorders (B-EDs) is well established. However, previous research has not yet identified whether these individuals exhibit heterogeneous profiles of dietary restraint and whether these profiles are associated with differences in eating pathology.
Method:
Individuals with B-EDs (N =290) completed the Eating Disorder Examination. Latent profile analysis was conducted on dietary restraint frequency data, including restriction of overall amount of food consumed, avoidance of eating, desire for an empty stomach, food avoidance, and dietary rules. Identified latent profiles were compared on binge eating frequency, compensatory behaviors frequency, and ED pathology using the three-step procedure.
Results:
A four-class model of dietary restraint best fit the data. Classes significantly differed in frequency of compensatory behaviors (F(3, 286)=31.01, p<.001), EDE eating concern (F(3, 286)=14.36, p<.001), EDE shape concern (F(3, 286)=7.06, p<.001), EDE weight concern (F(3, 286)=6.83, p<.001) and ED pathology (F(3, 286)=12.86, p<.001), but did not differ in frequency of objective (F(3, 286)=2.45, p=.06) or subjective binge episodes (F(3, 286)=1.87, p=.14).
Discussion:
Individuals with B-EDs exhibit distinct profiles of dietary restraint, which are associated with frequency of compensatory behaviors and severity of ED pathology.
Keywords: dietary restraint, binge eating, bulimia nervosa, binge eating disorder
Introduction
Frontline treatments for binge-spectrum eating disorders (B-EDs; eating disorders (EDs) characterized by recurrent loss of control over eating, including bulimia nervosa and binge eating disorder) yield remission in only 30–60% of individuals (Linardon et al., 2018; Peat et al., 2017). Preliminary evidence suggests that poor treatment outcomes may be due to the highly heterogeneous factors that maintain B-EDs (Levinson et al., 2022). Subtyping of individuals with B-EDs has consistently identified distinct subgroups, associated with differential clinical severity and treatment responses, further indicating that B-EDs are heterogeneous (Dounchis et al., 2021; Goldschmidt et al., 2008; Stice et al., 2001). Personalized treatment approaches, which are designed to target idiographic maintenance factors based on an individual’s symptoms, may improve treatment outcomes and efficiency, reducing the public health burden and suffering associated with B-EDs (Levinson et al., 2022). However, to implement personalized treatments, the field must understand the unique relationships between facets of eating pathology among subgroups with B-EDs.
Dietary restraint (i.e., attempted or actual restriction of caloric intake) is widely considered the primary maintaining factor for B-EDs and reducing dietary restraint is a key treatment target within enhanced cognitive behavior therapy for B-EDs (CBT-E; Fairburn et al., 2009). While a robust body of literature has linked the occurrence and severity of dietary restraint to binge eating (Goldschmidt et al., 2012; Jake Linardon, 2018) and extreme weight control behaviors (Liechty & Lee, 2013), little is known about the contributions of specific forms of dietary restraint to severity of binge eating and overall eating pathology. Different forms of dietary restraint may differentially contribute to restricted energy consumption (e.g., fasting or restriction of caloric content may yield substantially restricted energy intake, whereas avoidance of specific foods may not) and therefore to binge eating. Consistent with this hypothesis, some evidence suggests frequency of meals is negatively associated with frequency of binge episodes (Elran-Barak et al., 2015). However, no previous studies have examined the association between other specific facets of dietary restraint (e.g., food avoidance, following dietary rules, etc.) and ED symptoms.
Furthermore, some individuals with binge-spectrum EDs deny engaging in dietary restraint. The symptom profile (e.g., patterns of binge eating and other facets of eating pathology) and maintenance factors for ED pathology of these individuals is relatively unknown. Understanding how these individuals differ from individuals who endorse dietary restraint may facilitate identification of key treatment targets and inform treatment decision-making.
This study utilized latent profile analysis to determine whether there are latent groups that differ in dietary restraint and examine the proportion of individuals belonging to each group in a transdiagnostic sample of individuals with B-EDs. The present study also examined the association between latent dietary restraint profiles and severity of objective and subjective binge eating episodes, compensatory behaviors, and cognitive ED pathology.
Method
Participants and Procedures.
Participants were 290 individuals with B-EDs who participated in one of five studies at the Drexel Center for Weight, Eating and Lifestyle Science (80.3% female; 64.8% White, 17.2% Black, 4.1% Asian, 2.1% Multiracial, 1.7% Other). Participants were recruited from Philadelphia before COVID-19 and the United States during COVID-19. Participants completed a phone screen, then provided informed consent and completed a baseline assessment, which included the Eating Disorder Examination Interview and collection of BMI and demographic information. All participants experienced ≥ 12 binge eating episodes over the past 12 weeks and were ≥ 18 years old. Exclusion criteria across all studies included inability to speak, read, and write English, previous bariatric surgery, current or planned pregnancy, or current severe psychopathology that would inhibit engagement in study protocols (e.g., active psychosis, substance use disorder; see Supplementary Table 1 for study-specific inclusion/exclusion criteria). Approximately 51% of our sample (n=149) reported no compensatory behaviors in the past month, while 48.3% (n=140) reported engaging in at least one compensatory behavior (data were missing for one participant). Of those with compensatory behaviors, 55.3% (n=78) reported engaging in purging at least once (vomiting, laxative misuse, or diuretics misuse) while 44.0% (n=44) reported engaging in only non-purging compensatory behaviors. Procedures were approved and overseen by the Drexel University Institutional Review Board.
Measures.
Demographics.
Participants self-reported demographic information including age, sex, and ethnicity.
Eating Disorder Examination.
The Eating Disorder Examination (EDE) is an interview widely used for assessing ED symptoms (Fairburn et al., 2014). The EDE measures frequency of binge eating episodes, compensatory behaviors, and dietary restraint. The EDE yields four subscale scores (Restraint, Eating Concern, Shape Concern, and Weight Concern) and a global eating pathology score. Although the EDE global score typically includes all four subscale scores, a composite score excluding the restraint subscale was computed (via the mean of the Eating Concern, Shape Concern, and Weight Concern subscales) for this study to avoid violating the assumption of independence. The items that comprise the Restraint subscale, including limiting the overall amount of food consumed, avoidance of eating for 8+ hours, desire for an empty stomach, food avoidance, and dietary rules, measured forms of dietary restraint.
Body Mass Index.
Body mass index (BMI) was calculated from height and weight data, which were collected using research-grade stadiometers and scales during in-person baseline assessments. During COVID-19, participants self-reported their height and weighed themselves using at-home scales.
Statistical Analyses.
Statistical analyses were conducted in R version 4.0.3 and alpha was <.05. Latent profile analysis with EDE restraint items was conducted using the ‘mclust’ package via tidyLPA and one to five classes (Scrucca et al., 2016). To reduce the possibility of a local solution given the discrete nature of the EDE restraint items, a scaled singular value decomposition transformation was applied to the data prior to conducting latent profile analysis (Scrucca & Raftery, 2015). The best-fit model was selected using the Analytic Hierarchy Process, as this process has been shown to outperform the use of individual fit indices alone (see Supplemental Material for more information on the Analytic Hierarchy Process; Akogul & Erisoglu, 2017). The classes were compared on ED symptoms (objective and subjective binge episode frequency, compensatory behavior frequency, EDE subscale scores, and composite ED pathology score), age, and BMI using the three-step procedure for predicting distal variables using latent profile membership (i.e., regression accounting for classification probabilities; Asparouhov & Muthen, 2014). Post hoc pairwise comparisons using Tukey’s Honestly Significant Difference were conducted for significant regressions. Previous research suggests that sample sizes of approximately 300 independent observations are adequate for latent profile analysis (Collins & Wugalter, 1992) and simulation studies have indicated that a sample size of 200 is adequate when using 8 or fewer indicator variables (Nylund et al., 2007). Further, rates of data missingness for indicator variables were between 0.0% and 0.3% in our sample, supporting the sufficiency of a relatively modest sample size (Wolf et al., 2013). These considerations suggest our sample size of 290 participants is adequate for exploratory latent profile analysis.
Results
Latent profile analysis yielded a four-class best fit model with equal variances and covariances (see Table 1). Class 1 (n=100), termed “Low restraint,” was characterized by low frequency of all forms of dietary restraint and contained 34.5% of participants (see Table 2). Class 2 (n=51, 17.6%) was labeled “Frequent limitation of overall food consumed and moderate food avoidance” and demonstrated high frequency of limiting the overall amount of food consumed (Meandays=26.65 in the past 28 days), and moderate frequency of food avoidance (Meandays=13.35), with low frequency of avoidance of eating (Meandays=1.39), desire for empty stomach (Meandays=0.67), and dietary rules (Meandays=1.39). Class 3 (n=49), labeled “High restraint,” demonstrated relatively high frequency of all forms of dietary restraint and contained 16.9% of participants. Class 4 (n=90, 31.0%) was labeled “Frequent limitation of overall food consumed, food avoidance, and dietary rules” and demonstrated high frequency of limiting overall food consumption (Meandays=22.24), food avoidance (Meandays=18.45), and dietary rules (Meandays=24.17). Entropy (≥ 0.80), mean posterior probability for each class (≥ 0.80), and % of participants with posterior probability ≥ 0.80 indicated good class separation (see Table 1).
Table 1.
Fit indices for one to five latent profile analysis models
| # of Classes | Class Proportions % (N) | Log Likelihood | BIC | SABIC | AIC | cAIC | Entropy | Mean Posterior Probability | % Posterior Probability > 0.80 |
|---|---|---|---|---|---|---|---|---|---|
| 1 | Class 1: 100.0% | −5174.43 | 10462.25 | 10398.83 | 10388.85 | 10482.25 | 1.00 | Class 1: 1.00 | Class 1: 1.00 |
| 2 | Class 1: 18.6% (54) Class 2: 81.4% (236) |
−4991.73 | 10130.87 | 10048.42 | 10035.45 | 10156.87 | 0.98 | Class 1: .98 Class 2: .999 |
Class 1: 0.96 Class 2: 1.00 |
| 3 | Class 1: 16.9% (49) Class 2: 32.4% (94) Class 3: 50.7% (147) |
−4949.04 | 10079.51 | 9978.04 | 9962.08 | 10111.51 | 0.93 | Class 1: 0.99 Class 2: 0.94 Class 3: 0.98 |
Class 1: 0.98 Class 2: 0.88 Class 3: 0.96 |
| 4 | Class 1: 34.5% (100) Class 2: 17.6% (51) Class 3: 16.9% (49) Class 4: 31.0% (90) |
−4922.17 | 10059.80 | 9939.29 | 9920.35 | 10097.80 | 0.91 | Class 1: 0.95 Class 2: 0.95 Class 3: 0.99 Class 4: 0.94 |
Class 1: 0.89 Class 2: 0.90 Class 3: 0.98 Class 4: 0.88 |
| 5 | Class 1: 9.3% (27) Class 2: 31.0% (90) Class 3: 16.6% (48) Class 4: 34.8% (101) Class 5: 8.3% (24) |
−4910.92 | 10071.31 | 9931.78 | 9909.83 | 10115.31 | 0.91 | Class 1: 0.91 Class 2: 0.95 Class 3: 0.999 Class 4: 0.95 Class 5: 0.86 |
Class 1: 0.74 Class 2: 0.88 Class 3: 1.00 Class 4: 0.89 Class 5: 0.75 |
Table 2.
Comparison of latent profiles of dietary restraint on dietary restraint variables, demographic characteristics, and ED symptom severity
| Descriptive Statistics and Regression Model Statistics Comparing Latent Profiles | ||||||
|---|---|---|---|---|---|---|
| Variable | “Low Restraint” (N=100) Mean (SD) | “Frequent limitation of overall food consumed and moderate food avoidance” (N=51) Mean (SD) | “High restraint” (N=49) Mean (SD) | “Frequent limitation of overall food consumed, food avoidance, and dietary rules” (N=90) Mean (SD) | F(3, 286), p | f 2 |
| Limitation of Overall Amount of Food Consumed (days in past month) | 4.61 (5.43) | 26.65 (2.84) | 21.85 (9.64) | 22.24 (8.29) | 174.95, <.001*** | Class 1: Not estimated, reference class Class 2: 1.27 Class 3: 0.76 Class 4: 1.15 |
| Avoidance of Eating for 8+ Hours (days in past month) | 1.54 (4.67) | 1.39 (3.76) | 6.45 (7.32) | 1.60 (3.22) | 15.37, <.001*** | Class 1: Not estimated, reference class Class 2: 0.0003 Class 3: 0.13 Class 4: 0.00002 |
| Desire for Empty Stomach (days in past month) | 1.05 (2.82) | 0.67 (2.87) | 25.69 (4.12) | 1.93 (3.92) | 745.58, <.001*** | Class 1: Not estimated, reference class Class 2: 0.002 Class 3: 6.35 Class 4: 0.01 |
| Food Avoidance (days in past month) | 5.49 (8.33) | 13.35 (11.90) | 19.04 (11.54) | 18.45 (10.19) | 33.41, <.001*** | Class 1: Not estimated, reference class Class 2: 0.07 Class 3: 0.22 Class 4: 0.27 |
| Dietary Rules (days in past month) | 2.92 (5.42) | 1.39 (3.47) | 19.37 (11.29) | 24.17 (5.00) | 264.45, <.001*** | Class 1: Not estimated, reference class Class 2: 0.01 Class 3: 0.83 Class 4: 1.98 |
| Age (years) | 45.79 (13.73) | 47.76 (14.42) | 36.85 (14.40) | 40.68 (15.03) | 7.09, <.001*** | Class 1: Not estimated, reference class Class 2: 0.003 Class 3: 0.05 Class 4: 0.02 |
| BMI (kg/m2) | 34.44 (7.31) | 33.14 (5.61) | 29.03 (6.35) | 30.40 (7.10) | 9.66, <.001*** | Class 1: Not estimated, reference class Class 2: 0.003 Class 3: 0.07 Class 4: 0.06 |
| Objective Binge Episodes (in past month) | 12.90 (8.80) | 12.63 (9.24) | 18.37 (17.19) | 15.66 (18.35) | 2.45, .06 | Class 1: Not estimated, reference class Class 2: 0.0002 Class 3: 0.02 Class 4: 0.004 |
| Subjective Binge Episodes (in past month) | 12.04 (11.99) | 11.98 (10.81) | 11.18 (12.89) | 8.46 (10.03) | 1.87, .14 | Class 1: Not estimated, reference class Class 2: 0.0001 Class 3: 0.001 Class 4: 0.02 |
| Compensatory Behaviors (in past month) | 4.78 (10.45) | 3.82 (8.20) | 32.00 (27.94) | 18.71 (23.94) | 31.01, <.001*** | Class 1: Not estimated, reference class Class 2: 0.0004 Class 3: 0.26 Class 4: 0.08 |
| EDE Eating Concern | 1.41 (1.20) | 1.62 (1.28) | 2.71 (1.36) | 2.15 (1.19) | 14.36, <.001*** | Class 1: Not estimated, reference class Class 2: 0.002 Class 3: 0.13 Class 4: 0.06 |
| EDE Weight Concern | 3.22 (1.12) | 3.28 (0.84) | 4.02 (1.18) | 3.37 (1.12) | 6.83, <.001*** | Class 1: Not estimated, reference class Class 2: 0.0004 Class 3: 0.07 Class 4: 0.003 |
| EDE Shape Concern | 3.53 (1.16) | 3.55 (0.94) | 4.37 (1.18) | 3.69 (1.18) | 7.06, <.001*** | Class 1: Not estimated, reference class Class 2: 0.00005 Class 3: 0.06 Class 4: 0.003 |
| Composite Eating Disorder Pathology Score | 2.72 (0.97) | 2.82 (0.82) | 3.70 (1.06) | 3.07 (0.94) | 12.86, <.001*** | Class 1: Not estimated, reference class Class 2: 0.001 Class 3: 0.12 Class 4: 0.02 |
| Post Hoc Pairwise Comparisons for Significant Regression Models | ||||||
| Variable | Class Comparisons | Mean Difference | 95% CI | Familywise Adjusted p | ||
| Age (years) | 2–1 3–1 4–1 3–2 4–2 4–3 |
2.25 −9.07 −4.74 −11.32 −6.99 4.33 |
−4.20, 8.69 −15.61, −2.54 −10.19, 0.70 −18.83, −3.81 −13.57, −0.41 −2.34, 11.00 |
.80 .002** .11 <.001*** .03 .34 |
||
| BMI (kg/m2) | 2–1 3–1 4–1 3–2 4–2 4–3 |
−1.14 −5.38 −4.08 −4.24 −2.94 1.30 |
−4.17, 1.89 −8.45, −2.31 −6.64, −1.52 −7.76, −0.72 −6.03, 0.15 −1.83, 4.42 |
.77 <.001*** <.001*** .01* .07 .71 |
||
| Compensatory Behaviors (in past month) | 2–1 3–1 4–1 3–2 4–2 4–3 |
−1.08 27.30 13.09 28.38 14.17 −14.21 |
−9.53, 7.36 18.75, 35.86 5.93, 20.24 18.58, 38.19 5.57, 22.78 −22.93, −5.50 |
.99 <.001*** <.001*** <.001*** <.001*** <.001*** |
||
| EDE Eating Concern | 2–1 3–1 4–1 3–2 4–2 4–3 |
0.18 1.29 0.73 1.11 0.55 −0.56 |
−0.37, 0.73 0.73, 1.85 0.27, 1.20 0.47, 1.75 −.01, 1.11 −1.12, .01 |
.84 <.001*** <.001*** <.001*** .05 .06 |
||
| EDE Weight Concern | 2–1 3–1 4–1 3–2 4–2 4–3 |
0.07 0.82 0.15 0.75 0.08 −0.67 |
−0.42, 0.55 0.33, 1.31 −0.26, 0.55 0.19, 1.31 −0.41, 0.57 −1.17, −0.17 |
.98 <.001*** .79 .004** .98 .003** |
||
| EDE Shape Concern | 2–1 3–1 4–1 3–2 4–2 4–3 |
0.02 0.86 0.16 0.83 0.14 −0.69 |
−0.48, 0.53 0.34, 1.37 −0.26, 0.59 0.25, 1.42 −.38, 0.66 −1.22, −0.17 |
.999 <.001*** .76 .002** .90 .004** |
||
| Composite Eating Disorder Pathology Score | 2–1 3–1 4–1 3–2 4–2 4–3 |
−0.09 0.99 0.35 0.90 0.26 −0.64 |
−0.51, 0.34 0.56, 1.42 −0.01, 0.70 0.40, 1.39 −0.17, 0.69 −1.08, −0.20 |
.95 <.001*** .06 <.001*** .42 .001** |
||
Note: EDE = Eating Disorder Examination. Frequency of binge eating and compensatory behaviors are based on past-month data. Effect size thresholds are: small f2 ≥ 0.02, medium f2 ≥ 0.15, large f2 ≥ 0.35.
denotes significance at p<.05,
denotes p<.01, and
denotes p<.001
The groups significantly differed in age, as participants in the “High restraint” group were significantly younger than participants in both the “Low restraint” and “Frequent limitation of overall food consumed and moderate food avoidance” groups, and participants in the “Frequent limitation of overall food consumed, food avoidance, and dietary rules” were significantly younger than participants in the “Frequent limitation of overall amount of food consumed and moderate food avoidance” group (see Table 2). Similarly, groups significantly differed in BMI, with participants in the “High restraint” group demonstrating significantly lower BMI than participants in the “Low restraint” and “Frequent limitation of overall food consumed and moderate food avoidance” groups and participants in the “Frequent limitation of overall food consumed, food avoidance, and dietary rules” group showing significantly lower BMI than participants in the “Low restraint” group. The groups did not significantly differ in frequency of objective or subjective binge episodes in the past month. Groups significantly differed in frequency of past-month compensatory behaviors, EDE Eating Concern, EDE Weight Concern, EDE Shape Concern, and composite ED pathology scores. Post hoc analyses indicated that participants in the “High restraint” group had significantly more frequent compensatory behaviors than participants in all other groups, while participants in the “Frequent limitation of overall food consumed, food avoidance, and dietary rules” group had significantly more frequent compensatory behaviors than participants in the “Low restraint” and “Frequent limitation of overall food consumed and moderate food avoidance” groups. Participants in the “High restraint” and “Frequent limitation of overall food consumed, food avoidance, and dietary rules” classes showed significantly greater eating concern than the “Low restraint” class, while participants in the “High restraint” class showed significantly greater eating concern than the “Frequent limitation of overall food consumed and moderate food avoidance” class. The “High restraint” class showed significantly elevated shape concerns and weight concerns and composite eating pathology compared to all other classes, which did not significantly differ.
Discussion
Individuals with B-EDs demonstrate heterogeneous profiles of dietary restraint. Almost 35% of our sample demonstrated low levels of all dietary restraint variables, despite demonstrating comparable frequency of both objective and subjective binge eating episodes as other dietary restraint profiles. For individuals low in dietary restraint, binge eating may be maintained by factors other than dietary restraint, such as emotion regulation deficits (Dingemans et al., 2017; Prefit et al., 2019), impaired inhibitory control (Haynos et al., 2021; Smith et al., 2020; Svaldi et al., 2014), and/or dysregulated reward processes (Haynos et al., 2021; Kessler et al., 2016).
The “Frequent limitation of overall food consumed, moderate food avoidance” class demonstrated comparable frequency of binge eating and composite ED pathology to the “Low restraint” class and significantly lower eating concern compared to the “High restraint” profile. These results may suggest that this profile represents more normative dietary restraint that is not associated with severe cognitive ED pathology. These two lower restraint groups also demonstrated significantly higher average BMIs than participants in the “High restraint” profile, which may be partially due to a general pattern of elevated food reward within these groups.
The two remaining classes, “High restraint” and “Frequent limitation of overall food consumed, food avoidance, and dietary rules” demonstrated significantly more frequent engagement in compensatory behaviors relative to the other two classes, suggesting these profiles may exhibit more engagement in inappropriate weight control behaviors across the board. The “High restraint” group demonstrated significantly higher composite ED pathology and greater shape and weight concerns than all other profiles. This may suggest that some forms of dietary restraint, including avoidance of eating and desire for an empty stomach, are indicative of overall ED severity, including shape and weight dissatisfaction.
The current study does not provide any information regarding whether different profiles of dietary restraint are differentially associated with treatment response, but we anticipate that the present findings could have important clinical implications. Existing research indicates that CBT-E is effective for reducing dietary restraint among individuals with high levels of baseline dietary restraint, which is associated with improvements in binge eating for these patients (Accurso et al., 2016; J. Linardon, 2018; Presseller et al., 2022). This may be due to the emphasis in CBT-E on reducing dietary restraint. However, for individuals with low dietary restraint, we hypothesize that early attention to other, non-dietary factors maintaining loss of control eating may improve treatment outcomes. Future research should explore whether adjusting the timing and relative focus of treatment for B-EDs based on an individual’s dietary restraint profile can improve treatment efficacy.
The present study had several strengths and limitations. The inclusion of individuals with transdiagnostic binge eating increases generalizability. However, the sample was primarily comprised of White women, and findings may not generalize to other samples. Similarly, the present study combined samples from several parent studies with somewhat different inclusion criteria. This sample heterogeneity may limit the generalizability of findings and is a notable limitation. Dietary restraint frequencies were self-reported by participants during interviews and therefore may be biased. Dietary restraint measured by the EDE captures both attempted dietary restraint and actual dietary restriction. Future research should investigate where there are heterogeneous profiles of actual dietary restriction among individuals with B-EDs. Furthermore, data on previous treatment experiences was not collected and therefore our sample may have had heterogeneous prior exposure to ED treatment targeting dietary restraint. Finally, although our sample is adequate for the use of latent profile analysis based on previous conventions, evidence also suggests that larger samples are associated with more replicable results from latent mixed modeling (Wurpts & Geiser, 2014). Accordingly, our results should be replicated in a larger sample. The data were cross-sectional, prohibiting causal conclusions.
Given their association with different ED symptom profiles, future research should investigate whether dietary restraint profiles are differentially associated with treatment outcomes. Further research should continue to investigate maintenance factors and antecedents for binge eating episodes among individuals without dietary restraint. Longitudinal studies should also investigate whether and how patterns of dietary restraint change over time, and whether these changes contribute to changes in other ED symptoms.
In sum, individuals with B-EDs demonstrate varying profiles of dietary restraint, which are associated with differences in severity of binge eating and cognitive ED pathology. These associations may offer insight into the diverse pathogenesis of binge eating and inform personalized treatment approaches.
Supplementary Material
Public Significance Statement:
Individuals with binge-spectrum eating disorders have different patterns of restrictive eating symptoms. These profiles of restrictive eating behaviors are associated with differences in severity of compensatory behaviors and cognitive eating disorder symptoms, like shape and weight dissatisfaction. Understanding the relationships between profiles of restrictive eating behaviors and other eating disorder symptoms may allow for personalization of treatment and improvements in treatment efficacy.
Funding Statement:
National Institutes of Health grants R01DK117072, R34MH116021, K23MH105680 and a grant from the Hilda and Preston Davis Foundation to Dr. Adrienne Juarascio supported collection of the data examined in the present article.
Footnotes
Conflict of Interest: The authors have no conflicts of interest to disclose.
IRB Statement: All study procedures were approved and overseen by the Drexel University Institutional Review Board.
Data Availability Statement:
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.
References
- Accurso EC, Wonderlich SA, Crosby RD, Smith TL, Klein MH, Mitchell JE, Crow SJ, Berg KC, & Peterson CB (2016). Predictors and moderators of treatment outcome in a randomized clinical trial for adults with symptoms of bulimia nervosa. J Consult Clin Psychol, 84(2), 178–184. 10.1037/ccp0000073 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Akogul S, & Erisoglu M (2017). An approach for determining the number of clusters in a model-based cluster analysis. Entropy, 19(9), 452. 10.3390/e19090452 [DOI] [Google Scholar]
- Asparouhov T, & Muthen B (2014). Auxiliary variables in mixture modeling: Three-step approaches using Mplus. Structural Equation Modeling: A Multidisciplinary Journal, 21(3), 329–341. 10.1080/10705511.2014.915181 [DOI] [Google Scholar]
- Collins LM, & Wugalter SE (1992). Latent class models for stage-sequential dynamic latent variables. Multivariate Behavioral Research, 27(1), 131–157. 10.1207/s15327906mbr2701_8 [DOI] [Google Scholar]
- Dingemans A, Danner U, & Parks M (2017). Emotion Regulation in Binge Eating Disorder: A Review. Nutrients, 9(11). 10.3390/nu9111274 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Dounchis JZ, Karam AM, Stein RI, & Wilfley DE (2021). Subtyping patients with binge-eating disorder by dietary restraint and negative affect: Characteristics and treatment outcome. Journal of consulting and clinical psychology, 89(12), 1020. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Elran-Barak R, Sztainer M, Goldschmidt AB, Crow SJ, Peterson CB, Hill LL, Crosby RD, Powers P, Mitchell JE, & Le Grange D (2015). Dietary restriction behaviors and binge eating in anorexia nervosa, bulimia nervosa and binge eating disorder: trans-diagnostic examination of the restraint model. Eating behaviors, 18, 192–196. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Fairburn CG, Cooper D Phil, Zafra DP, Doll D Phil, O’Connor HA, Bohn ME Phil D, Kristin DP, Hawker DM, Wales JA, & Palmer RL (2009). Transdiagnostic cognitive-behavioral therapy for patients with eating disorders: a two-site trial with 60-week follow-up. American Journal of Psychiatry, 166(3), 311–319. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Fairburn CG, Cooper Z, & O’Connor M (2014). Eating Disorder Examination. https://www.credo-oxford.com/pdfs/EDE_17.0D.pdf
- Goldschmidt AB, Tanofsky-Kraff M, Goossens L, Eddy KT, Ringham R, Yanovski SZ, Braet C, Marcus MD, Wilfley DE, & Yanovski JA (2008). Subtyping children and adolescents with loss of control eating by negative affect and dietary restraint. Behaviour research and therapy, 46(7), 777–787. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Goldschmidt AB, Wall M, Loth KA, Le Grange D, & Neumark-Sztainer D (2012). Which dieters are at risk for the onset of binge eating? A prospective study of adolescents and young adults. Journal of Adolescent Health, 51(1), 86–92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Haynos AF, Camchong J, Pearson CM, Lavender JM, Mueller BA, Peterson CB, Specker S, Raymond N, & Lim KO (2021). Resting State Hypoconnectivity of Reward Networks in Binge Eating Disorder. Cereb Cortex, 31(5), 2494–2504. 10.1093/cercor/bhaa369 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Kessler RM, Hutson PH, Herman BK, & Potenza MN (2016). The neurobiological basis of binge-eating disorder. Neurosci Biobehav Rev, 63, 223–238. 10.1016/j.neubiorev.2016.01.013 [DOI] [PubMed] [Google Scholar]
- Levinson CA, Hunt RA, Christian C, Williams BM, Keshishian AC, Vanzhula IA, & Ralph-Nearman C (2022). Longitudinal group and individual networks of eating disorder symptoms in individuals diagnosed with an eating disorder. J Psychopathol Clin Sci, 131(1), 58–72. 10.1037/abn0000727 [DOI] [PubMed] [Google Scholar]
- Liechty JM, & Lee MJ (2013). Longitudinal predictors of dieting and disordered eating among young adults in the US. International Journal of Eating Disorders, 46(8), 790–800. [DOI] [PubMed] [Google Scholar]
- Linardon J (2018). Meta-analysis of the effects of cognitive-behavioral therapy on the core eating disorder maintaining mechanisms: implications for mechanisms of therapeutic change. Cogn Behav Ther, 47(2), 107–125. 10.1080/16506073.2018.1427785 [DOI] [PubMed] [Google Scholar]
- Linardon J (2018). The relationship between dietary restraint and binge eating: Examining eating-related self-efficacy as a moderator. Appetite, 127, 126–129. [DOI] [PubMed] [Google Scholar]
- Linardon J, Messer M, & Fuller-Tyszkiewicz M (2018). Meta-analysis of the effects of cognitive-behavioral therapy for binge-eating-type disorders on abstinence rates in nonrandomized effectiveness studies: Comparable outcomes to randomized, controlled trials? International Journal of Eating Disorders, 51(12), 1303–1311. 10.1002/eat.22986 [DOI] [PubMed] [Google Scholar]
- Nylund KL, Asparouhov T, & Muthen BO (2007). Deciding on the number of classes in latent class analysis and growth mixture modeling: A Monte Carlo simulation study. Structural Equation Modeling: A Multidisciplinary Journal, 14, 535–569. 10.1080/10705510701575396 [DOI] [Google Scholar]
- Peat CM, Berkman ND, Lohr KN, Brownley KA, Bann CM, Cullen K, Quattlebaum MJ, & Bulik CM (2017). Comparative effectiveness of treatments for binge-eating disorder: Systematic review and network meta-analysis. European Eating Disorders Review, 25(5), 317–328. 10.1002/erv.2517 [DOI] [PubMed] [Google Scholar]
- Prefit AB, Candea DM, & Szentagotai-Tatar A (2019). Emotion regulation across eating pathology: A meta-analysis. Appetite, 143, 104438. 10.1016/j.appet.2019.104438 [DOI] [PubMed] [Google Scholar]
- Presseller EK, Lampe EW, Michael ML, Trainor C, Fan SC, & Juarascio AS (2022). Latent trajectories of symptom change during cognitive-behavior therapy predict post-treatment worsening of symptoms: a preliminary examination among outpatients with bulimia-spectrum eating disorders. Eat Weight Disord, 27(6), 2257–2264. 10.1007/s40519-021-01348-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Scrucca L, Fop M, Murphy TB, & Raftery AE (2016). mclust 5: Clustering, classification, and density estimation using Gaussian finite mixture models. The R Journal, 8/1, 205–233. [PMC free article] [PubMed] [Google Scholar]
- Scrucca L, & Raftery AE (2015). Improved initialisation of model-based clustering using Gaussian hierarchical partitions. Advances in Data Analysis and Classification, 9(4), 447–460. 10.1007/s11634-015-0220-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- Smith KE, Mason TB, Schaefer LM, Juarascio A, Dvorak R, Weinbach N, Crosby RD, & Wonderlich SA (2020). Examining intra-individual variability in food-related inhibitory control and negative affect as predictors of binge eating using ecological momentary assessment. J Psychiatr Res, 120, 137–143. 10.1016/j.jpsychires.2019.10.017 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Stice E, Agras WS, Telch CF, Halmi KA, Mitchell JE, & Wilson T (2001). Subtyping binge eating‐disordered women along dieting and negative affect dimensions. International Journal of Eating Disorders, 30(1), 11–27. [DOI] [PubMed] [Google Scholar]
- Svaldi J, Naumann E, Trentowska M, & Schmitz F (2014). General and food-specific inhibitory deficits in binge eating disorder. Int J Eat Disord, 47(5), 534–542. 10.1002/eat.22260 [DOI] [PubMed] [Google Scholar]
- Wolf EJ, Harrington KM, Clark SL, & Miller MW (2013). Sample size requirements for structural equation models: An evaluation of power, bias, and solution propriety. Educational and Psychological Measurement, 73(6). 10.1177/0013164413495237 [DOI] [PMC free article] [PubMed] [Google Scholar]
- Wurpts IC, & Geiser C (2014). Is adding more indicators to a latent class analysis beneficial or detrimental? Results of a Monte-Carlo study. Frontiers in Psychology, 5, 920. 10.3389/fpsyg.2014.00920 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.