Self-regulation deficits moderate treatment outcomes in a clinical trial evaluating just-in-time adaptive interventions as an augmentation to cognitive-behavioral therapy for bulimia-spectrum eating disorders

Abstract

Objective:

Just-in-time adaptive interventions (JITAIs), momentary interventions delivered at identified times of risk, may improve skill utilization during cognitive-behavioral therapy (CBT-E) for bulimia-spectrum eating disorders (BN-EDs). JITAIs may be especially helpful for individuals with self-regulation deficits, including emotion regulation deficits and elevated impulsivity.

Method:

Participants (N = 55 with BN-EDs) received 16 sessions of CBT-E with electronic self-monitoring and were randomized to receive JITAIs (JITAIs-On) or not receive JITAIs (JITAIs-Off). Baseline Difficulties in Emotion Regulation Scale (DERS) and UPPS-P Impulsive Behavior Scale (UPPS-P) total scores were examined as moderators of baseline to post-treatment change in binge episodes, compensatory behaviors, and Eating Disorder Examination (EDE) global score using repeated measures ANOVAs.

Results:

Emotion regulation difficulties significantly moderated compensatory behavior change (F[1, 51] = 4.31, p = .04, η_p² = 0.08) such that individuals with emotion regulation deficits demonstrated greater improvements in the JITAIs-On condition. Impulsivity moderated change in binge episodes (F[1, 51] = 8.94, p = .004, η_p² = 0.15) and compensatory behaviors (F[1, 51] = 7.83, p = .007, η_p² = 0.13), such that individuals with high impulsivity showed greater improvement in the JITAIs-On condition. Neither DERS nor UPPS-P scores moderated EDE global score change.

Discussion:

JITAIs appear particularly beneficial for facilitating skill use during treatment for BN-EDs for individuals with self-regulation deficits, yielding improved treatment outcomes.

Public Significance Statement:

Reminders to use therapy skills that are delivered via smartphone as an individual goes about their daily life may improve treatment response among individuals with bulimia nervosa who have difficulty coping with emotions or who tend to act impulsively. Results from this study indicate that individuals with these difficulties benefitted more from cognitive-behavioral therapy when it was accompanied by in-the-moment reminders to use therapeutic skills, which may facilitate long-term recovery.

Clinical Trials.gov Registration Number:

NCT03673540.

1 |. INTRODUCTION

Bulimia nervosa (BN) is an eating disorder (ED) characterized by recurrent binge eating and compensatory behaviors. Current treatments for BN are sub-optimal, with nearly 70% of individuals failing to achieve full remission at post-treatment (Linardon & Wade, 2018), highlighting a need to identify maintenance factors of BN that might negatively impact outcomes. Self-regulation deficits (i.e., difficulties with regulating emotions and behavior; Claes et al., 2012; Prefit, Candea, & Szentagotai-Tatar, 2019; Waxman, 2009), including emotion regulation difficulties and impulsivity, are implicated in the maintenance of BN. Literature suggests that individuals with BN have emotion regulation deficits, including difficulties in emotional awareness, clarity, and acceptance, difficulty controlling behavior when distressed, and lack of adaptive emotion regulation strategies (Lavender et al., 2015; Svaldi, Griepenstroh, Tuschen-Caffier, & Ehring, 2012; Weinbach, Sher, & Bohon, 2018). Greater deficits in emotion regulation are associated with greater eating pathology and more frequent binge eating and compensatory behaviors (Lavender et al., 2014; Trompeter et al., 2021). Individuals with BN also demonstrate elevated impulsivity (i.e., tendency to act without forethought; Fischer, Smith, & Anderson, 2003; Sysko et al., 2017; Waxman, 2009). Elevated impulsivity may contribute to the maintenance of BN as individuals may engage in binge eating or compensatory behaviors without considering long-term consequences (Bruce, Koerner, Steiger, & Young, 2003). A large body of literature substantiates that self-regulation deficits, including deficits in emotion regulation and impulsivity, contribute to the maintenance of BN.

Self-regulation may also impact treatment outcomes for BN. One study identified that elevated baseline emotion regulation deficits were associated with lower rates of post-treatment remission among outpatients with BN (Trainor et al., Under review). Lower baseline impulsivity predicted long-term recovery following individual CBT among individuals with BN (Castellini et al., 2012). Improvements in self-regulation during treatment were positively correlated with reductions in binge eating and compensatory behaviors in a partial hospitalization program for adults with BN (Brown et al., 2020). Furthermore, improvements in self-regulation during the first half of treatment for BN predicted post-treatment improvements in overall eating pathology (Peterson et al., 2017).

Individuals with deficits in self-regulation may struggle to achieve abstinence from BN symptoms. Treatments for BN often include behavioral recommendations for reducing dietary restraint (e.g., regular eating) that may be challenging for patients with self-regulation deficits. For example, regular eating may prompt fears about weight gain, which may be difficult for individuals with emotion regulation deficits to manage. Individuals with high impulsivity may struggle to follow a regular eating schedule that requires planning meals and pre-purchasing food. Given the impact of self-regulation deficits on the maintenance of BN, it is critical to consider ways to improve treatment for individuals with BN and significant deficits in self-regulation.

Just-in-time adaptive interventions (JITAIs) constitute an intervention approach that may be particularly beneficial for individuals high in emotion dysregulation and impulsivity. JITAIs are momentary interventions delivered via smartphones at identified instances of risk for maladaptive behaviors (e.g., BN symptoms; Nahum-Shani et al., 2018). JITAIs may be useful for reminding individuals with self-regulation deficits of their treatment goals in daily life and facilitating skill implementation even when doing so is difficult (e.g., when distressed). For example, JITAIs can encourage patients to practice relevant treatment skills (e.g., urge management or emotion regulation skills) for resisting compensatory behaviors after a binge episode. Preliminary evidence suggests that momentary interventions can facilitate real time use of emotion regulation skills (Ter Harmsel et al., 2021). Although JITAIs may be particularly beneficial for individuals with poor self-regulation, no study to date has tested whether the addition of JITAIs to treatments for BN improves treatment outcomes for individuals with baseline deficits in emotion regulation and impulsivity. Identifying the effect of JITAIs on treatment response for these individuals may be critical for improving their treatment outcomes.

This study aimed to evaluate whether baseline emotion regulation deficits and impulsivity moderated treatment outcomes from a randomized, controlled trial examining JITAIs as an adjunct to enhanced cognitive-behavioral therapy for BN-EDs (CBT-E). We hypothesized that individuals with greater baseline emotion dysregulation and elevated impulsivity would demonstrate greater improvements in binge eating, compensatory behaviors, and overall eating pathology in the JITAIs-On condition compared to JITAIs-Off. We also hypothesized that individuals with low emotion dysregulation and impulsivity would demonstrate equivalent improvements in binge eating, compensatory behaviors, and overall eating pathology in both conditions.

2 |. METHODS

2.1 |. Participants

Participants were 55 individuals with BN-spectrum EDs (BN-EDs) who participated in an RCT examining JITAIs as an adjunct to CBT-E. Participants were included if they: (1) met DSM-5 behavioral criteria for BN or other specified feeding and eating disorder (OSFED), BN with subjective binge episodes only, (2) were aged 18–70, (3) had BMI > 17.5, and (4) were willing and able to track eating episodes and ED behaviors on a smartphone application for 16 weeks. Exclusion criteria were: (1) inability to fluently speak, read, and write English, (2) medical complications from BN preventing safe participation in outpatient treatment, (3) severe comorbid psychopathology that would inhibit engagement in treatment (i.e., active suicidal ideation, psychotic, or substance use disorder), (4) previous completion of a full course of CBT-E for BN (i.e., a complete, manualized course of CBT-E), (5) current or planned pregnancy, or (6) bariatric surgery. These criteria were selected to ensure that participants could safely and fully engage in the manualized treatment protocol. Most participants met criteria for BN (N = 51, 92.7%), while the remaining participants (N = 4, 7.3%) met criteria for OSFED, BN.

2.2 |. Procedures

Participants completed baseline, mid-treatment (after 8 sessions), and post-treatment (after 16 sessions) assessments. The treatment was 16 weeks of CBT-E focused version (Fairburn, 2008), accompanied by electronic self-monitoring of eating episodes and ED behaviors via smartphone application. Half of the participants (n = 29) also received JITAIs to promote CBT-E skill use in daily life, delivered in response to information about eating behaviors, ED behaviors, and mood from self-monitoring records. See Juarascio et al. (2021) for description of the treatment, including development and pilot evaluation, and Juarascio, Presseller, Srivastava, Manasse, and Forman (Under review) for primary feasibility, acceptability, and target engagement outcomes from the trial. Participants made 3.01 self-monitoring entries per day on average and self-monitoring compliance was comparable across conditions (t = .17, p = .86). Participants provided informed consent and study procedures were approved by Drexel University’s Institutional Review Board.

2.3 |. Measures

2.3.1 |. Eating disorders examination (EDE)

The Eating Disorders Examination 17.0 (Fairburn, Cooper, & O’Connor, 1993), a reliable (Rizvi, Peterson, Crow, & Agras, 2000) interview, evaluated study eligibility, past month frequency of BN symptoms, and global eating pathology. The EDE yields four subscales (Restraint, Eating Concern, Shape Concern, and Weight Concern) and a global score which range from 0 (lowest pathology) to 6 (highest pathology).

2.3.2 |. Difficulties in emotion regulation scale (DERS)

The DERS (Gratz & Roemer, 2004) assessed emotion regulation deficits. The DERS assesses three domains of emotion regulation: (1) awareness and understanding of emotions (Non-Acceptance, Awareness, and Clarity subscales), (2) ability to modulate emotional arousal (Strategies subscale), and (3) effective action during intense emotions (Goals and Impulse subscales). Higher scores indicate greater difficulties in emotion regulation. Total scores range from 36 to 180. The DERS has high internal consistency and test–retest reliability and adequate construct and predictive validity (Gratz & Roemer, 2004).

2.3.3 |. UPPS-P impulsive behavior scale

The UPPS-P (Lynam, Smith, Whiteside, & Cyders, 2006) is a valid measure of impulsive behavior (Smith et al., 2007) that yields five subscales (Negative Urgency, Positive Urgency, (Lack of) Premeditation, (Lack of) Perseverance, and Sensation Seeking) and a total score. Subscale and total scores range from 1 to 4. Greater scores indicate higher levels of impulsivity.

2.4 |. Statistical analyses

Analyses were conducted using SPSS version 28.0 and p < .05. Baseline DERS and UPPS-P scores were dichotomized using median split (DERS median = 102.00, UPPS-P median = 2.29). Repeated measures ANOVAs examined whether dichotomized DERS and UPPS-P scores moderated baseline to post-treatment change in BN symptoms by treatment condition (JITAIs-On versus JITAIs-Off). Post hoc repeated measures ANOVAs were completed using DERS and UPPS-P subscales for statistically significant moderator relationships. A priori power analyses conducted in G*Power with alpha = .05 and power = .80 indicated that the study is powered to detect results of medium effect size (η_p² > 0.078; effect size thresholds are: small η_p² > 0.01, medium >0.06, large >0.14).

3 |. RESULTS

Table 1 summarizes demographics, treatment outcomes, and moderator analyses. Results indicate that baseline DERS scores significantly predicted compensatory behavior change such that in the JITAIs-On group higher DERS scores were associated with greater reduction in compensatory behaviors and in the JITAIs-Off group lower DERS scores were associated with greater reduction in compensatory behaviors (Figure 1). DERS scores did not moderate changes in total binge eating or EDE global score.

TABLE 1.

Participant demographics, treatment outcomes by treatment condition, moderation analyses with self-regulation deficits, and post hoc subscale analyses

Participant demographics
	JITAIs-On (N = 29)	JITAIs-Off (N = 26)
Demographic variable	Mean (SD)	Mean (SD)
Age (years)	38.07 (13.59)	39.58 (15.04)
BMI (kg/m2)	29.85 (6.87)	29.28 (6.97)
	%(N)	%(N)
Gender	Female: 86.2% (25)	Female: 80.8% (21)
	Male: 13.8% (4)	Male: 15.4% (4)
	Non-binary: 0.0% (0)	Non-binary: 3.8% (1)
Race	American Indian/Alaska native: 3.4% (1)	American Indian/Alaska native: 0.0% (0)
	Asian: 10.3% (3)	Asian: 7.7% (2)
	Black/African American: 17.2% (5)	Black/African American: 7.7% (2)
	White: 62.1% (18)	White: 73.1% (19)
	Other: 13.8% (4)	Other: 7.7% (2)
	Multiracial: 3.4% (1)	Multiracial: 7.7% (2)
	Unknown or prefer not to say: 0.0% (0)	Unknown or prefer not to say: 3.8% (1)
Ethnicity	Latinx/Hispanic: 3.4% (1)	Latinx/Hispanic: 11.5% (3)
	Not Latinx/Hispanic: 89.7% (26)	Not Latinx/Hispanic: 88.5% (23)
	Not reported: 6.9% (2)	Not reported: 0.0% (0)

Treatment outcomes by treatment condition
	Baseline mean (SD)		Comparing baseline ED symptoms	Post-treatment mean (SD)
Variable	JITAIs-On	JITAIs-Off	t(p)	JITAIs-On	JITAIs-Off
Total binge episodes	21.72 (14.28)	31.27 (30.03)	−1.48 (.15)	5.48 (10.26)	4.62 (7.49)
Total compensatory behaviors	33.62 (25.00)	30.50 (25.72)	0.46 (.65)	11.41 (18.58)	6.92 (10.75)
EDE global score	3.41 (0.81)	3.22 (1.11)	0.73 (.47)	1.81 (1.25)	1.56 (0.95)

Moderation results
Moderator	BN symptom	Time F (p)	Time by treatment condition F (p)	Time by moderator F (p)	Time by treatment condition by moderator F (p)	Time by treatment condition by moderator ηp2
DERS Total	Binge eating episodes	6.73 (.012)*	1.64 (.21)	0.60 (.44)	0.69 (.41)	0.01
DERS Total	Total compensatory behaviors	13.26 (<.001)***	0.08 (.93)	0.33 (.57)	4.31 (.04)*	0.08
DERS Total	EDE global score	35.09 (<.001)***	0.35 (.71)	1.17 (.29)	1.71 (.20)	0.03
UPPS-P Total	Binge eating episodes	9.87 (.003)**	2.15 (.13)	1.37 (.25)	8.94 (.004)**	0.15
UPPS-P Total	Total compensatory behaviors	15.43 (<.001)***	0.07 (.93)	0.08 (.77)	7.83 (.007)**	0.13
UPPS-P Total	EDE global score	32.65 (<.001)***	0.17 (.85)	0.29 (.59)	0.85 (.36)	0.02

Post hoc analyses
Moderator	BN symptom	Time F (p)	Time by treatment condition F (p)	Time by moderator F (p)	Time by treatment condition by moderator F (p)	Time by treatment condition by moderator ηp2
DERS non-acceptance	Total compensatory behaviors	12.82 (<.001)***	0.09 (.92)	1.65 (.21)	2.94 (.09)	0.06
DERS goals	Total compensatory behaviors	13.09 (<.001)***	0.24 (.79)	0.32 (.57)	5.30 (.03)*	0.09
DERS impulse	Total compensatory behaviors	12.35 (<.001)***	0.07 (.93)	0.65 (.43)	1.48 (.23)	0.03
DERS awareness	Total compensatory behaviors	12.21 (<.001)***	0.07 (.93)	0.68 (.41)	1.05 (.31)	0.02
DERS strategies	Total compensatory behaviors	13.75 (<.001)***	0.08 (.93)	0.79 (.38)	5.81 (.02)*	0.10
DERS clarity	Total compensatory behaviors	11.75 (.001)**	0.01 (.99)	1.86 (.18)	2.40 (.13)	0.05
UPPS-P negative urgency	Binge eating episodes	7.40 (.009)**	1.23 (.30)	1.35 (.25)	2.40 (.13)	0.05
UPPS-P lack of premeditation	Binge eating episodes	10.46 (.002)**	2.36 (.11)	2.60 (.11)	9.67 (.003)**	0.16
UPPS-P lack of perseverance	Binge eating episodes	8.22 (.006)**	2.69 (.08)	1.87 (.18)	5.91 (.02)*	0.10
UPPS-P sensation seeking	Binge eating episodes	7.05 (.01)*	1.50 (.23)	0.05 (.82)	1.72 (.20)	0.03
UPPS-P positive urgency	Binge eating episodes	9.25 (.004)**	1.96 (.15)	2.20 (.15)	6.72 (.01)*	0.12
UPPS-P negative urgency	Total compensatory behaviors	12.40 (<.001)***	0.06 (.95)	0.55 (.46)	5.82 (.02)*	0.10
UPPS-P lack of premeditation	Total compensatory behaviors	14.61 (<.001)***	0.13 (.88)	0.88 (.35)	3.84 (.06)	0.07
UPPS-P lack of perseverance	Total compensatory behaviors	13.73 (<.001)***	0.17 (.84)	0.38 (.54)	4.10 (.048)*	0.07
UPPS-P sensation seeking	Total compensatory behaviors	12.19 (.001)**	0.04 (.96)	0.17 (.68)	0.01 (.94)	0.00
UPPS-P positive urgency	Total compensatory behaviors	16.55 (<.001)***	0.02 (.98)	1.23 (.27)	15.48 (<.001)***	0.23

Note: Analyses based on past-month frequency of binge eating episodes and total compensatory behaviors.

Abbreviations: BMI, body mass index; DERS, Difficulties in Emotion Regulation Scale; EDE, Eating Disorder Examination; JITAIs-Off, just-in-time adaptive interventions off treatment condition; JITAIs-On, just-in-time adaptive interventions on treatment condition; SD, standard deviation; UPPS-P, UPPS-P Impulsive Behavior Scale.

^*Designates p < .05;

^**Designates p < .01;

^***Designates p < .001.

FIGURE 1.

Change in binge eating and compensatory behaviors from pre- to post-treatment moderated by self-regulation deficits.

Note: On = just-in-time adaptive interventions on treatment condition; JITAI-Off = just-in-time adaptive interventions off treatment condition; DERS = Difficulties in Emotion Regulation Scale; UPPS-P = UPPS-P Impulsive Behavior Scale

Further, baseline UPPS-P scores moderated the association between treatment condition and total binge episodes (objective and subjective binge eating episodes). For individuals in JITAIs-On, greater impulsivity was associated with greater reduction in binge eating whereas, for those in JITAIs-Off, lower impulsivity was associated with greater improvement. Baseline UPPS-P moderated the association between treatment condition and compensatory behaviors in the same direction as the previous analysis but did not significantly moderate EDE global score (Figure 1).

Post hoc analyses indicated that the DERS Goals and Strategies subscales significantly moderated the association between JITAIs and compensatory behaviors, with higher scores on these subscales predicting greater reduction in compensatory behaviors in JITAIs-On and lower scores predicting greater reductions in JITAIs-Off. Several UPPS-P subscales emerged as significant moderators of the association between treatment condition and BN symptom improvement. Lack of Premeditation, Lack of Perseverance, and Positive Urgency significantly moderated the association between condition and improvement in binge episodes, with higher scores predicting greater reductions in JITAIs-On whereas lower scores predicted greater reductions in JITAIs-Off. Negative Urgency, Lack of Perseverance, and Positive Urgency moderated the relationship between condition and compensatory behavior outcome, with higher scores associated with greater reductions in JITAIs-On and lower scores associated with greater reductions in JITAIs-Off.

4 |. DISCUSSION

This study examined whether baseline deficits in self-regulation moderated treatment outcomes from a clinical trial comparing CBT-E with JITAIs to CBT-E alone. Findings robustly support the role of self-regulation deficits in moderating the impact of JITAIs on treatment outcomes, with baseline emotion regulation deficits and elevated impulsivity predicting greater pre- to post-treatment symptom improvement in JITAIs-On for binge eating (impulsivity) and compensatory behaviors (emotion regulation and impulsivity) compared to JITAIs-Off. The findings also suggest that for individuals without self-regulation deficits, the addition of JITAIs to CBT-E may result in reduced treatment gains.

The moderating effect of emotion regulation deficits on change in compensatory behaviors from the JITAIs-On condition was driven by two subscales: Goals, representing difficulty engaging in goal-directed behavior when distressed, and Strategies, representing lack of access to emotion regulation strategies. JITAIs may be particularly helpful for individuals with difficulties in these domains as their delivery at times of distress could promote the identification of appropriate coping methods and remind individuals to engage in treatment recommendations despite distress. For instance, while elevated anxiety following a binge episode may typically prompt an individual with BN to engage in self-induced vomiting to alleviate anxiety, a JITAI suggesting an alternative activity triggered by self-reported urges to vomit may facilitate abstinence from vomiting and adaptive emotion regulation.

The moderating effects of impulsivity on change in binge eating and compensatory behaviors in the JITAIs-On condition were driven by Lack of Premeditation, Lack of Perseverance, and Negative Urgency (binge eating) and Negative Urgency, Lack of Perseverance, and Positive Urgency (compensatory behaviors). Individuals high in Lack of Premeditation (tendency to act without thinking), Negative Urgency, or Positive Urgency (tendency to act rashly when experiencing negative or positive emotions, respectively) may have difficulty remembering treatment goals and recommendations, especially during intense emotions. JITAIs may facilitate goal salience and provide actionable treatment-consistent steps that these individuals can take during moments of risk for BN symptoms. For individuals high in Lack of Perseverance (difficulty remaining focused on a task), JITAIs may improve treatment outcomes by prompting participants to persist with treatment-consistent strategies like urge surfing or emotion regulation strategies.

The present study had several limitations. The sample was modest and predominantly comprised of White women, limiting the capacity to generalize findings to more diverse samples. Emotion regulation and impulsivity were self-reported and may be biased by social desirability or limited insight. Future research should attempt to replicate these findings in more diverse samples (e.g., samples with greater racial/ethnic diversity, higher proportions of men and gender diverse individuals), in other EDs with marked emotion regulation deficits and/or impulsivity (e.g., binge eating disorder), and using behavioral measures of self-regulation. The study did not directly measure whether JITAIs improve emotion regulation and impulsivity, which may explain their moderating effect. Future research should investigate whether JITAIs yield greater improvements in emotion regulation and impulsivity than CBT-E alone. This preliminary study indicates that JITAIs may improve treatment outcomes for individuals with BN and deficits in emotion regulation and impulsivity.

DATA AVAILABILITY STATEMENT

Data are available upon reasonable request to corresponding author.