Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Dec 1.
Published in final edited form as: Eat Behav. 2020 Oct 29;39:101448. doi: 10.1016/j.eatbeh.2020.101448

Binge Eating Predicts Adherence to Digital Self-Monitoring during Behavioral Weight Loss

Mary K Martinelli 1,*, Laura D’Adamo 1, Meghan L Butryn 1
PMCID: PMC7704883  NIHMSID: NIHMS1643864  PMID: 33157520

Abstract

Objective:

Consistent self-monitoring of dietary intake, weight, and physical activity predicts better outcomes during behavioral weight loss, but the factors that influence self-monitoring adherence are not well understood. This study attempted to fill gaps in the existing literature by examining whether pre-treatment eating behaviors predict adherence to digital self-monitoring during a behavioral weight loss program.

Method:

Participants (N = 77) reported on binge eating, uncontrolled eating, and emotional eating at baseline, and were instructed to self-monitor their food intake, weight, and physical activity using digital devices (food logging app, Fitbit, and wireless “smart” scale) throughout the 12-week treatment. Adherence to self-monitoring was assessed using data captured from these devices.

Results:

Greater baseline binge eating severity predicted greater adherence to self-monitoring of weight (ρ = 0.25, p = .03) and eating (ρ = 0.25, p = .03), but not self-monitoring of physical activity. Uncontrolled eating and emotional eating did not significantly predict self-monitoring adherence.

Conclusions:

In contrast to previous research, this study found that participants with greater pre-treatment binge eating severity had better adherence to self-monitoring of eating, and for the first time established a relationship between binge eating severity and digital self-monitoring of weight in behavioral weight loss. Individuals with greater pre-treatment binge eating may exhibit characteristics, such as motivation or rigidity, that are beneficial during the initial period of weight loss. Future studies should determine if there are features of analogue versus digital self-monitoring that may explain this pattern of findings, and examine these associations longitudinally.

Keywords: obesity, binge eating, self-monitoring, behavioral weight loss, treatment adherence

1. Introduction

While consistent self-monitoring of dietary intake, weight, and physical activity predicts better outcomes during behavioral weight loss interventions (Burke, Wang, & Sevick, 2011; Butryn et al., 2019; Carels, Selensky, Rossi, Solar, & Hlavka, 2017; Harvey, Krukowski, Priest, & West, 2019), the factors that influence self-monitoring adherence are not well understood. Problematic eating behaviors, such as sub-clinical binge eating, uncontrolled eating, and emotional eating, are common in individuals with obesity seeking weight loss treatment (Konttinen, van Strien, Mannisto, Jousilahti, & Haukkala, 2019; Marcus & Wildes, 2014; Rohrer, Vickers-Douglas, & Stroebel, 2009). It is unknown if these behaviors predict self-monitoring adherence. A sense of shame or avoidance related to these problematic eating behaviors is common (Mason & Lewis, 2015; Wong & Qian, 2016), which might impede self-monitoring adherence. Alternatively, individuals who engage in problematic eating behaviors might feel more motivated to adhere to treatment strategies to reduce the distress caused by these behaviors, or to control their eating following periods of overeating or weight gain (Smith, Williamson, Bray, & Ryan, 1999).

In one previous study, Ariel & Perri (2016) found that greater baseline binge eating severity predicted poorer adherence to food logging during behavioral weight loss. In contrast, Kong et al. (2012) found that binge eating, emotional eating, and uncontrolled eating were unrelated to the number of food logs completed during weight loss treatment. These mixed findings make it difficult to draw firm conclusions about the relationship between eating behaviors and self-monitoring of eating. Extant studies also relied exclusively on written food logging. Digital food logging may be more convenient, accessible, and discreet than written records. Moreover, prior research has not examined whether problematic eating behaviors predict self-monitoring of weight or physical activity. Given that self-monitoring of eating, weight, and physical activity have been shown to independently predict weight loss outcomes (Butryn et al., 2019), identifying pretreatment predictors of self-monitoring may be helpful in promoting self-monitoring adherence. The current study examined the relationship between baseline eating behaviors and adherence to digital self-monitoring of eating, physical activity, and weight among adults enrolled in a group-based behavioral weight loss program. We hypothesized that eating behaviors would significantly predict self-monitoring adherence during the 12-week treatment program.

2. Method

2.1. Participants and Procedures

This study is a secondary analysis of data collected for a clinical trial (R21DK112741) focused on digital data sharing and supportive accountability following face-to-face behavioral weight loss treatment. Eligible participants were 18–70 years old with a body mass index (BMI) between 25 and 45 kg/m2, capable of engaging in physical activity, and had a smartphone. Participants who currently met diagnostic criteria for binge eating disorder (assessed via clinical interview) were excluded, as were those who had a medical or psychiatric condition that may constitute a risk during the intervention, had recent weight loss (>5%), or used weight-affecting medication.

The parent study received Institutional Review Board approval and was registered at www.clinicaltrials.gov (NCT03337139). Detailed study design, recruitment, and intervention methods are reported elsewhere (Butryn et al., 2019). Briefly, the parent study consisted of an orientation session followed by a 12-week group-based behavioral weight loss treatment. The present study analyzed data from this 12-week period only, which was considered to be the weight loss phase. At the start of treatment, participants were provided digital devices and directed to use them to self-monitor their diet, physical activity, and weight, and review and reflect on their data throughout treatment. The Fitbit app was used for dietary self-monitoring, and also synced with physical activity and weight device data, allowing participants to track progress toward goals using one application. Participants were instructed to log all dietary intake in the app as soon as possible after eating. A wrist-worn physical activity sensor (Fitbit Flex), was provided to participants to measure physical activity daily. Participants were instructed to weigh themselves on the Yunmai Smart Scale weekly during the first 10 weeks of treatment and daily during the final 2 weeks, an approach commonly used in behavioral weight loss (Burke et al., 2011).

2.2. Measures

Eating behaviors were measured at baseline. Binge eating (BE) severity was assessed using the 16-item Binge Eating Scale (BES; Gormally, Black, Daston, & Rardin, 1982). Items are rated on a 3- or 4-point scale. Total scores range from 0 to 46, with higher scores reflecting greater BE severity. The Three-Factor Eating Questionnaire (TFEQ-R18; Karlsson, Persson, Sjostrom, & Sullivan, 2000) was used to assess uncontrolled eating (UE) (excessive appetite and cravings, eating in response to external cues; 9 items) and emotional eating (EE) (eating in response to negative mood; 3 items). Items are rated on a 4-point scale and subscale scores are transformed to a 0–100 scale with higher scores indicating greater uncontrolled or emotional eating.

Adherence to self-monitoring during treatment was assessed using data collected from the digital devices. Self-monitoring data were accessed after the completion of treatment via a remote, web-based research portal that captured data from the devices. The full 12-week treatment period was used to compute percent adherence. Adherence to dietary self-monitoring was calculated as the percentage of days with ≥800 calories logged (Patel, Hopkins, Brooks, & Bennett, 2019; Shay, Seibert, Watts, Sbrocco, & Pagliara, 2009). The treatment program advised participants not to eat less than 1200 calories per day. Therefore 800 calories was chosen as the threshold for adherence because it was suspected that even if some participants truly consumed less than 1200 calories, it would be unlikely that participants would restrict below 800 calories. Thus, it was assumed that days with less than 800 calories likely represented incomplete self-monitoring records. Adherence to physical activity self-monitoring was determined by the percentage of days with ≥500 steps were tracked with the Fitbit (Korinek et al., 2018; Meyer, Wasmann, Heuten, El Ali, & Boll, 2017). Percent adherence to weight self-monitoring was calculated based on adherence to the varied self-weighing prescription (described above).

2.3. Data Analysis

Data were analyzed in R and SPSS version 25 and statistical level was set at ∝ = .05. Associations between eating behavior and age, BMI, sex, and race were examined using independent samples t-tests for categorical variables and Pearson correlation for continuous variables, and subsequent analyses controlled for characteristics that were significantly related to eating behavior. Race was dichotomized for analyses (White = 1, other racial groups = 0) given that there was a general pattern of White participants scoring higher on baseline measures of eating behavior. Additionally White participants made up the largest racial category in the sample (41%), and if race was examined with multiple dichotomized variables (e.g., a dummy coded variable for each race) this would have limited the statistical power needed to detect relationships for the primary aim. The self-monitoring adherence variables were non-normally distributed (e.g., negatively skewed), causing violations of assumptions for parametric models. Thus, non-parametric tests (Spearman correlations) were used to assess relationships between eating behavior and self-monitoring adherence. If participants discontinued treatment (n = 10), device data were retained only until the date that they discontinued treatment. Therefore, only completers (N = 77) were used in the present analyses. One participant was missing BES data and was excluded from those analyses.

3. Results

Table 1 displays baseline participant characteristics and descriptive statistics for eating behaviors and self-monitoring adherence. Eating behaviors significantly differed by race, with White participants reporting significantly higher BE (t = 3.66, p < .001), UE (t = 2.47, p = .02), and EE (t = 3.04, p = .003) compared to participants of other racial groups. Age was negatively associated with UE (r = −.29, p < .001). Age was not related to BE (r = −.15, p = .21) or EE (r = −.19, p = .10). Eating behaviors did not differ by sex (BE: t = 3.66, p = .10; UE: t = 3.66, p = .10; EE: t = 3.66, p = .10) and were not associated with baseline BMI (BE: r = .04, p = .74; UE: r = −.0004, p = .997; EE: r = .20, p = .08). Eating behaviors were significantly intercorrelated: BE and UE (r = .68, p < .001), BE and EE (r = .68, p < .001), EE and UE (r = .67, p < .001). Associations between baseline anthropometric and demographic variables (BMI, age, sex, and race) and the self-monitoring adherence variables were examined, and no significant relationships were observed (p’s > .05).

Table 1.

Participant characteristics

M (SD) or n (%)
Demographic/anthropometric factors
BMI (kg/m2) 34.86 (4.75)
Age (years) 50.77 (13.39)
Gender (n, % female) 62 (80.52%)
Race
 American Indian/Native Alaskan 1 (1.30%)
 Asian 2 (2.60%)
 Native Hawaiian or other Pacific Islander 0 (0%)
 Black/African American 28 (36.36%)
 White/Caucasian 41 (53.25%)
 Other or more than one race 5 (6.49%)
Ethnicity
 Hispanic/Latina 3 (3.90%)
 Not Hispanic/Latina 74 (96.10%)
Eating behaviors
 Binge eating 13.19 (7.58)
 Uncontrolled eating 40.10 (18.56)
 Emotional eating 55.19 (24.05)
Self-monitoring % adherence
 Eating 82.51 (17.60)
 Weight 85.55 (18.67)
 Physical activity 93.64 (13.92)
Open in a new tab

Note. BMI = body mass index

Spearman correlations controlling for race revealed that greater BE severity at baseline predicted greater adherence to self-monitoring of weight (ρ = 0.25, p = .03), and eating (ρ = 0.25, p = .03). BE severity did not predict adherence to physical activity self-monitoring (ρ = 0.08, p = .50). Spearman correlations controlling for race and age indicated that UE at baseline did not predict adherence to self-monitoring of weight (ρ = 0.01, p = .94), eating (ρ = 0.01, p = .92), or physical activity (ρ = −0.005, p = .97). Similarly, EE at baseline, controlling for race, did not predict adherence to self-monitoring of weight (ρ = 0.15, p = .190), eating (ρ = 0.10, p = .41), or physical activity (ρ = −0.03, p = .79).

4. Discussion

The current study was designed to examine the relationship between eating behaviors and adherence to digital self-monitoring of eating, weight, and physical activity during a 12-week behavioral weight loss intervention. This is the first study with findings to suggest that greater baseline binge eating severity is associated with better adherence to eating and weight self-monitoring; of note, even the most severe binge eating behaviors were subclinical, as individuals who met diagnostic criteria were not eligible for treatment. Given that self-monitoring is a cornerstone of treatment for both obesity and binge eating, one explanation for these findings is that eating and weight self-monitoring may have been perceived as particularly helpful to those with greater eating problems. It is also possible that those with greater pretreatment eating severity differ from other individuals in meaningful ways, such as higher motivation to adhere to treatment or greater tendency to rigidly adopt weight-related behaviors initially following periods of overeating or weight gain (Smith et al., 1999). However, it is unclear if this pattern persists over time; two studies that found no relationship (Kong et al., 2012) or a negative association between binge eating and self-monitoring of eating (Ariel & Perri, 2016) were conducted over a 6-month period, compared to 3 months in the current study, raising the possibility that those with greater binge eating severity may adhere strictly to treatment initially, but are less successful at maintaining adherence as treatment progresses. Lastly, the finding that binge eating was related to eating and weight self-monitoring while uncontrolled and emotional eating were not suggests that, while the three constructs overlap, the BES may capture factors uniquely predictive of adherence. For instance, the BES includes items assessing cognitive features of binge eating, such as weight concerns and high dieting standards, which are hypothesized to motivate a pattern of strict dieting followed by periods of overeating (Gormally et al., 1982).

To our knowledge, the current study was the first to examine eating behaviors as predictors of physical activity tracking during behavioral weight loss. The findings from this study do not support a relationship between these eating behaviors and physical activity self-monitoring. It is possible that physical activity tracking and eating behavior are more distally related to one another as compared to weight and dietary self-monitoring and eating behavior, or that physical activity self-monitoring does not provide information that is as emotionally salient or reinforcing as weight and dietary self-monitoring information. Another possible explanation is that physical activity self-monitoring in this study was overall very high, possibly due to the passive nature of physical activity tracking, meaning that most participants were able to maintain high adherence relatively easily throughout treatment, regardless of particular barriers or motivators for this type of self-monitoring.

This study has several strengths and limitations. The strengths of this study include rigorous measurement of self-monitoring adherence using digital devices rather than analogue methods, inclusion of weight and physical activity self-monitoring adherence, and multifaceted measurement of eating behaviors. The findings from this study were limited to the first 3 months of weight loss and self-monitoring adherence during this phase of treatment was high, and it is possible, but perhaps unlikely, that some participants may have restricted below the 800 calorie threshold for eating self-monitoring adherence. It is also possible that some participants may have exceeded the target self-weighing threshold or engaged in excessive weighing, however this was not examined in this study. It should also be noted that self-monitoring data was concealed from coaches during this treatment period to avoid potential contamination of future randomized conditions. Although general encouragement to self-monitor was provided during group sessions, this was done so without reference to participant adherence. Lastly, given the emphasis of digital self-monitoring in this study, participants who self-selected for participation may have been more comfortable with or interested in using digital self-monitoring than the general population.

The current study found that higher binge eating severity at baseline was associated with greater adherence to digital self-monitoring of weight and eating, but not physical activity, during a 12-week behavioral weight loss intervention. Future studies should determine if there are psychological mechanisms (e.g., motivation) or features of digital self-monitoring that may explain this pattern of findings. Given the link between self-monitoring adherence and treatment outcomes (Butryn et al., 2019), future work examining these associations over multiple timepoints and in relation to weight loss outcomes is also warranted.

Highlights.

  • Pretreatment binge eating severity predicts better eating self-monitoring adherence

  • Pretreatment binge eating severity predicts better adherence to self-weighing

  • Binge eating may not be a barrier to self-monitoring with digital devices

  • Binge eating is not a barrier to self-monitoring during initial weight loss phase

Funding:

This research was funded by National Institute of Diabetes and Digestive and Kidney Diseases (Grant: R21DK112741) to M.L.B.

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.

Conflict of Interest: All authors declare that they have no conflicts of interest.

References

  1. Ariel AH, & Perri MG (2016). Effect of dose of behavioral treatment for obesity on binge eating severity. Eating Behaviors, 22, 55–61. doi: 10.1016/j.eatbeh.2016.03.032 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Burke LE, Wang J, & Sevick MA (2011). Self-monitoring in weight loss: a systematic review of the literature. Journal of the American Dietetic Association, 111(1), 92–102. doi: 10.1016/j.jada.2010.10.008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Butryn ML, Godfrey KM, Martinelli MK, Roberts SR, Forman EM, & Zhang F (2019). Digital self-monitoring: Does adherence or association with outcomes differ by self-monitoring target? Obesity Science & Practice. doi: 10.1002/osp4.391 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Carels RA, Selensky JC, Rossi J, Solar C, & Hlavka R (2017). A novel stepped-care approach to weight loss: The role of self-monitoring and health literacy in treatment outcomes. Eating Behaviors, 26, 76–82. doi: 10.1016/j.eatbeh.2017.01.009 [DOI] [PubMed] [Google Scholar]
  5. Gormally J, Black S, Daston S, & Rardin D (1982). The assessment of binge eating severity among obese persons. Addictive Behaviors, 7(1), 47–55. doi: 10.1016/0306-4603(82)90024-7 [DOI] [PubMed] [Google Scholar]
  6. Harvey J, Krukowski R, Priest J, & West D (2019). Log Often, Lose More: Electronic Dietary Self-Monitoring for Weight Loss. Obesity (Silver Spring), 27(3), 380–384. doi: 10.1002/oby.22382 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Karlsson J, Persson LO, Sjostrom L, & Sullivan M (2000). Psychometric properties and factor structure of the Three-Factor Eating Questionnaire (TFEQ) in obese men and women. Results from the Swedish Obese Subjects (SOS) study. International Journal of Obesity and Related Metabolic Disorders, 24(12), 1715–1725. doi: 10.1038/sj.ijo.0801442 [DOI] [PubMed] [Google Scholar]
  8. Kong A, Beresford SA, Imayama I, Duggan C, Alfano CM, Foster-Schubert KE, … McTiernan A (2012). Adoption of diet-related self-monitoring behaviors varies by race/ethnicity, education, and baseline binge eating score among overweight-to-obese postmenopausal women in a 12-month dietary weight loss intervention. Nutrition Research, 32(4), 260–265. doi: 10.1016/j.nutres.2012.03.001 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Konttinen H, van Strien T, Mannisto S, Jousilahti P, & Haukkala A (2019). Depression, emotional eating and long-term weight changes: a population-based prospective study. The International Journal of Behavioral Nutrition and Physical Activity, 16(1), 28. doi: 10.1186/s12966-019-0791-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Korinek EV, Phatak SS, Martin CA, Freigoun MT, Rivera DE, Adams MA, … Hekler EB (2018). Adaptive step goals and rewards: a longitudinal growth model of daily steps for a smartphone-based walking intervention. Journal of Behavioral Medicine, 41(1), 74–86. doi: 10.1007/s10865-017-9878-3 [DOI] [PubMed] [Google Scholar]
  11. Marcus MD, & Wildes JE (2014). Disordered eating in obese individuals. Curr Opin Psychiatry, 27(6), 443–447. doi: 10.1097/YCO.0000000000000103 [DOI] [PubMed] [Google Scholar]
  12. Mason TB, & Lewis RJ (2015). Assessing the roles of impulsivity, food-related cognitions, BMI, and demographics in the dual pathway model of binge eating among men and women. Eating Behaviors, 18, 151–155. doi: 10.1016/j.eatbeh.2015.05.015 [DOI] [PubMed] [Google Scholar]
  13. Meyer J, Wasmann M, Heuten W, El Ali A, & Boll SC (2017). Identification and classification of usage patterns in long-term activity tracking. Paper presented at the Proceedings of the 2017 CHI Conference on Human Factors in Computing Systems. [Google Scholar]
  14. Patel ML, Hopkins CM, Brooks TL, & Bennett GG (2019). Comparing Self-Monitoring Strategies for Weight Loss in a Smartphone App: Randomized Controlled Trial. JMIR Mhealth Uhealth, 7(2), e12209. doi: 10.2196/12209 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Rohrer JE, Vickers-Douglas KS, & Stroebel RJ (2009). Uncontrolled eating and obesity in adult primary care patients. Obesity Research & Clinical Practice, 3(2), I–II. doi: 10.1016/j.orcp.2009.01.004 [DOI] [PubMed] [Google Scholar]
  16. Shay LE, Seibert D, Watts D, Sbrocco T, & Pagliara C (2009). Adherence and weight loss outcomes associated with food-exercise diary preference in a military weight management program. Eating Behaviors, 10(4), 220–227. doi: 10.1016/j.eatbeh.2009.07.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Smith CF, Williamson DA, Bray GA, & Ryan DH (1999). Flexible vs. Rigid Dieting Strategies: Relationship with Adverse Behavioral Outcomes. Appetite, 32(3), 295–305. doi: 10.1006/appe.1998.0204 [DOI] [PubMed] [Google Scholar]
  18. Wong M, & Qian M (2016). The role of shame in emotional eating. Eating Behaviors, 23, 41–47. doi: 10.1016/j.eatbeh.2016.07.004 [DOI] [PubMed] [Google Scholar]

RESOURCES