Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Apr 8.
Published in final edited form as: Appetite. 2018 Jul 4;129:198–206. doi: 10.1016/j.appet.2018.07.003

Using ecological momentary assessment to better understand dietary lapse types

Stephanie P Goldstein a,b,f,*, Cara Dochat c, Leah M Schumacher a,b, Stephanie M Manasse a, Ross D Crosby d,e, J Graham Thomas f, Meghan L Butryn a,b, Evan M Forman a,b
PMCID: PMC13055932  NIHMSID: NIHMS1775653  PMID: 29981361

Abstract

Frequency of lapsing from a diet predicts weight loss failure, however previous studies have only utilized one definition of dietary lapse. No study has examined different types of lapse behaviors among individuals with overweight/obesity. The current study uses ecological momentary assessment (EMA) to examine predictors of three lapse types—eating a larger portion than intended, eating an unintended type of food, and eating at an unplanned time—in adults (N = 189; MBMI = 36.93 ± 5.83 kg/m2; 82.0% female; Mage = 51.81 ± 9.76 years) enrolled in a 12-month randomized controlled trial of two behavioral weight loss treatments. Participants completed 14 days of EMA at the start of treatment during which they indicated types of lapses that occurred with time and location of the lapse. Participants also responded to questions assessing current physical (e.g., hunger, tiredness), environmental (e.g., presence of “delicious” foods), and affective (e.g., loneliness, sadness) states at each prompt. Weight change was assessed at post-treatment. Separate generalized estimating equations were used to examine whether states prospectively predicted lapse occurrence at the next survey. Results indicated that lapse types differed significantly across time and location. Momentary increases in deprivation, hunger, and boredom increased likelihood of different lapse types. Lastly, we examined the prospective association between lapse type and weight loss. Eating at an unintended time was the only lapse type that predicted worse weight loss outcomes. Results support the theory that distinct lapse types exist, and that lapse types can be predicted by both momentary conditions and individual tendencies toward certain physical and affective states. However, not all lapse types may impact weight outcomes. Future research on behaviors that constitute dietary lapse is warranted and could inform personalized weight loss treatments.

Keywords: Dietary lapse, Weight loss, Ecological momentary assessment, Eating behavior

1. Introduction

Prevalence of overweight/obesity has plateaued at approximately 68% in the United States and has been largely attributable to obesogenic environmental and social contexts (Ogden, Carroll, Kit, & Flegal, 2014; Thomas et al., 2014). Overweight and obesity are known to negatively impact physical health, mental health, and quality of life (Williams, Mesidor, Winters, Dubbert, & Wyatt, 2015), and therefore weight loss interventions are critical. An overarching goal of all treatments for obesity is to achieve a negative energy balance, thereby inducing weight loss. Programs typically promote this goal through the prescription of dietary guidelines, and individuals who are adherent experience some degree of weight loss (Klem, Wing, McGuire, Seagle, & Hill, 1997; Shick et al., 1998). However, research has indicated that participants in such interventions often are not able to maintain dietary changes and, therefore, lose much less weight than expected (Wilson, 1994). In fact, it is often the case that participants experience weight regain within a year of completing treatment (Wadden & Butryn, 2003). These treatment failures likely occur due to difficulty making and maintaining suggested changes to dietary intake (Lowe, 2003).

Violations of dietary recommendations, i.e., dietary lapses, are expected in behavioral weight loss programs (Look AHEAD Research Group, 2006), but even discrete, momentary instances of nonadherence to a dietary prescription are strongly associated with suboptimal outcomes in behavioral obesity treatment (Forman et al., 2017). Lapses could be impeding weight loss and maintenance via higher calorie in-take (Blair & Leermakers, 2002). Furthermore, a series of lapses can lead to a return to unhealthy eating habits and abandoning goals of weight control (i.e., relapse; Carels, Douglass, Cacciapaglia, & O’Brien, 2004; Marlatt, 1985). Although lapses appear to have a critical role in weight loss treatments, they remain understudied (Forman et al., 2017). One factor that could be contributing to the paucity of research in this area is that lapses are difficult to capture and study.

Previous studies have asked participants to identify lapses as either “an incident where you felt that you broke your diet (e.g., overate, ate a forbidden food),” “unplanned eating episodes,” or “overeating” (Grilo, Shiffman, & Wing, 1989, 1993; Burke et al., 2015; Carels et al., 2002; Carels et al., 2004; McKee, Ntoumanis, & Taylor, 2014). These definitions of dietary lapses are vague in that they comprise several varying behaviors that one could, but does not necessarily have to consider, a dietary lapse. Forman et al. (2017) defined lapses as “instances of eating or drinking likely to cause weight gain, and/or put weight loss/maintenance at risk.” Participants endorsing a lapse were asked to categorize their lapse into one of three types of eating behavior according to a study by Thomas, Doshi, Crosby, and Lowe (2011) regarding overeating: (1) eating a larger portion than intended, (2) eating at an unplanned time, or (3) eating a food one intended to avoid (“forbidden food”). This study by Forman and colleagues was the first to use this categorization scheme to classify dietary lapses among individuals with overweight/obesity receiving behavioral weight loss treatment. The rationale for asking participants to categorize lapses was to develop a more nuanced understanding of what behaviors constitute lapses, which could improve lapse identification and reporting. Enhancing assessment of dietary lapses is critical for informing scientific discovery and eventually developing new models of behavioral and social processes that can inform intervention (Riley et al., 2011). Therefore, a sophisticated understanding of dietary lapses could lead to targeted treatments that substantially improve weight loss outcomes.

Ecological Momentary Assessment (EMA) is a useful tool for studying dietary lapses given that they are easily influenced by context (Shiffman, Stone, & Hufford, 2008). EMA employs short surveys delivered (typically via smartphone) repeatedly over the course of the day to participants to obtain self-reports of behaviors, cognitive/emotional states, and environmental contexts (Shiffman et al., 2008; Stone & Shiffman, 1994). Because EMA occurs in one’s natural environment and assesses events of interest directly in moments that they occur, EMA data are thought to be more reliable and contextually valid than traditional methodologies (e.g., lab-based simulations, retrospective self-report). EMA protocols can use event (i.e., recording at a lapse occurrence) and/or signal-based recording (i.e., recording when a time-based prompt occurs) to collect data related to lapse time (e.g., morning, afternoon, evening), location (e.g., home, work, restaurant), and relevant physical, environmental, and affective states (e.g., mood, tempting foods, watching TV).

Only six studies have used EMA to examine conditions and predictors of lapses from a weight loss diet in adults with overweight or obesity (Carels et al., 2004; Carels et al., 2002; Forman et al., 2017; Manasse, Schumacher, et al., 2018b; McKee et al., 2014; Schumacher et al., 2017). Findings from one of these studies are of particular interest, given that participants were instructed to report on specific lapse types, as mentioned above. This study found that dietary lapses were frequent (Mlapses per week = 4.23), more lapses were associated with less weight loss over time, lapses were predicted by greater overall levels of negative affect (e.g., stress, sadness) and environmental triggers (i.e., presence of tempting foods), lapses were predicted by physical states such as hunger and deprivation, and lapses occurred most commonly at home and in the evenings (Forman et al., 2017). Initial results on specific lapse types revealed that eating an unintended food was the most common type of lapse, followed by eating at an unplanned time and eating a larger portion (Forman et al., 2017). No additional analyses of lapse type, such as distinct predictors of different lapse types, were reported by Forman et al. (2017) due to the exploratory nature of this measurement method.

1.1. Current study

The potential for characterizing types of lapses to obtain a fuller understanding about the behaviors that constitute a lapse warrants further investigation. As an example, different types of lapses may be differentially associated with weight loss, which would have significant implications for which behaviors to specifically target in behavioral weight loss treatment. Perhaps eating a larger portion of food than intended would lead to greater overall caloric intake, thus threatening weight loss to a greater extent than other lapse behaviors (Nestle, 2003). Another example is that lapse types could have differential physical (e.g., feelings of hunger), affective (e.g., mood states), and environmental (e.g., time, location, presence of food) predictors, which would indicate how certain lapses could be targeted in treatment. It is possible that participants may be more likely to eat more than planned if hungrier than usual, or more likely to eat at an unplanned time in the evenings (Johnson, 2013; Moynihan et al., 2015).

Given that there is no prior research on types of lapses among in dividuals with overweight/obesity receiving weight loss treatment, we conducted a secondary analysis of lapse types from data collected from the Forman et al. (2017) trial. We expected that results would enhance current understanding of dietary lapses, generate additional data-driven hypotheses about lapse types, and provide evidence for the utility of measuring differentiated lapse types. Further, we hoped that our findings would add to extant literature by enhancing current obesity treatments and informing novel approaches to intervention. The aims of the current study were to (1) examine lapse types across time and location to identify environmental predictors of lapse, (2) identify which physical (i.e., deprivation, hunger, and fatigue), environmental (i.e., presence of food), and affective (i.e., sadness, loneliness, boredom, anger, and stress) states predict which types of lapses, and (3) examine whether the frequency of any specific lapse types at baseline predict weight loss during treatment. The lack of research in this area precludes a priori hypotheses and therefore we are considering each of these aims to be exploratory in nature.

2. Methods

2.1. Participants

Participants were weight loss-seeking adults (82.0% female; 70.9% Caucasian; Mage = 51.81 ± 9.76 years; MBody Mass Index = 36.93 ± 5.83 kg/m2) enrolled in a 12-month structured behavioral weight loss program as part of a larger study (Forman et al., 2016). Inclusion criteria for the parent study were: body mass index of 27–50 kg/m2, aged 18–70, and seeking weight loss. Exclusion criteria included: medical or psychiatric condition that posed risk to participant during weight loss, inability to engage in exercise plan of the program, initiation or titration of a weight-affecting medication within the previous 3 months, pregnancy (or planned pregnancy), having greater than 5% weight loss in the prior 6 months, and diagnosis of binge eating disorder.

2.2. Procedure

For the parent study (Forman et al., 2016), as approved by the Drexel University Institutional Review Board, participants were randomized to one of two 12-month behavioral weight loss programs that were both based on Look AHEAD and the Diabetes Prevention Program after informed consent was obtained (Diabetes Prevention Program Research Group, 2002; Look AHEAD Research Group, 2006). As a part of both treatments, participants were prescribed a calorie goal based on their starting weight (i.e., 1200–1500 calories per day if less than 250 lbs.; 1500–1800 calories per day if greater than 250 lbs.). One condition was a standard behavioral weight loss intervention and the other incorporated acceptance-based principles to assist with skills related to self-regulation. Participants in the acceptance-based group lost more weight at 12-months (13.3% ± 0.83%) than those in the standard treatment group (9.8% ± 0.87%; p = 0.005). Refer to Forman et al. (2016) for a more thorough description of the treatment conditions.

We collapsed data across conditions for the current study because treatments were identical for the first several weeks in that they both focused on self-monitoring, reducing caloric intake through reduction of high-fat/high-sugar foods intake, and beginning an exercise routine. At the outset of treatment, participants completed 14 days of EMA protocol. Participants completing the EMA protocol were loaned an Android Samsung Galaxy Player 4.0 (firmware version 2.3.5, build GINGERBREAD. UEKI8) with a custom EMA smartphone application (DrexelEMA) created for this study. At the outset of data collection, participants were given written directions and an in-person demonstration on how to use the application and how to identify dietary lapses.

Semi-random prompts were sent six times a day anchored at 9:30 a.m., 12:00 p.m., 2:30 p.m., 5:00 p.m., 7:15 p.m., and 9:30 p.m., with a standard deviation of 30 min around the anchor times to introduce uncertainty. If participants did not start the survey when initially prompted, they received reminder signals every 5 min until the survey was completed for up to 45 min, at which point the survey prompt expired. The procedure of semi-random prompting ensured a well-rounded assessment of each day and served to reduce participant re-activity by reducing anticipation of the prompts (i.e., altering behavior as a result of completing EMA). To further discourage reactivity and promote honest reporting, participants were told that their data would have no impact on their future study participation and that group leaders could not access participant data.

Participants could earn up to $42 for their participation in this portion of the study. For each missed survey, participants were deducted $1 from the $42 total compensation possible. Participants were also instructed to initiate and complete an EMA survey when a dietary lapse occurred outside of the prompted survey windows. All surveys were time stamped automatically at survey initiation. In addition to self-initiated lapse reports, participants were asked if a dietary lapse had occurred since the previous survey at each EMA prompt. Participants also reported on their current affective, physical, and environmental states at each prompt.

2.3. Measures

2.3.1. Height and Weight

The stadiometer and scale used in the current study was a Tanita PH7407121 (Serial No. 14080079). The scale was calibrated to manufacturer standards annually (per Tanita recommendation). Participants were weighed at baseline (pre-treatment) and post-treatment (12 months). Additionally, weight was measured prior to each treatment session. Participants were weighed in street clothes without shoes using the scale (accurate to 0.1 kg). Height was measured at baseline using the stadiometer (accurate to 0.5 cm) to calculate body mass index. We examined both early (within the first four weeks) and overall weight change (over 12 months), expressed in terms of percent weight loss.

2.3.2. Lapse Type and Lapse Triggers

The lapse definition used was “eating or drinking likely to cause weight gain, and/or put weight loss/maintenance at risk.” Participants endorsing a lapse were asked to specify the date and time, location (home, school, work, restaurant/café, or other), and type of lapse. Participants chose one of three lapse types: eating a larger potion than intended, eating at an unplanned time, or eating a type of food they intended to avoid. Two lapses occurring within 20 minutes of each other were considered a single lapse, and only the first response was used in data analyses.

At each survey, participants also answered questions on a 5-point Likert-type scale (1 = Not at all to 5 = Extremely) on physical (hunger, perceived deprivation, tiredness) and affective (sadness, loneliness, boredom, anger/irritation, stress) states. The affective state questions were adapted from the Negative Affect subscale of the Positive and Negative Affect Schedule (PANAS; Watson, Clark, & Tellegen, 1988) to refer to an individuals’ affect ‘right now’ (e.g., “Right now, how strongly do you feel sad?”). The physical states questions adopted a similar format. Environmental states were assessed automatically (i.e., time) and manually (i.e., self-reported location, self-reported presence of tempting food). The presence of tempting food was assessed using a yes or no question: “Since the last time you completed this survey, was delicious food or drink available that would put your calorie goal/weight control at risk?”

2.4. Statistical approach

IBM SPSS Statistics version 24 was used to analyze data for the current study. Lapse types were characterized across time (i.e., morning, afternoon, and evening) and locations using total proportion (across all participants) and mean proportion per participant. Chi-square tests of independence were used to examine the association between lapse types with location and time. Physical, affective, and environmental triggers of lapse were characterized by examining means and standard deviations for each trigger across participants. Histograms, normal Q-Q plots, and bivariate correlations were used to determine that the data met the necessary assumption for generalized estimating equations (GEE). Models met assumptions of multi-collinearity, normality of residuals, homoscedasticity, and absence of influential data points. Separate GEE models based on a negative binomial distribution with a logit link function and an AR (1) covariance structure were used to examine whether each trigger predicted lapse type occurrence at the next survey. For each lapse type, lapse cases in which another lapse type was reported were removed from analyses such that the models estimate only the likelihood of each individual lapse type compared to not lapsing at all. We report the proportion of EMA responses removed for each of the models below. All between-subjects variables were grand mean centered. Within-subjects effects for continuous predictor variables were centered within person. All models included between- and within-subjects centered predictors as a method for partitioning variance (Goldstein, Browne, & Rasbash, 2002). Models also included whether a lapse was reported at the survey used for prediction of the subsequent lapse. To correct for the possibility of Type I error, only p-values below .01 were interpreted. Missing weight data were imputed using a maximum likelihood estimation. Multiple linear regressions were used to assess the relationship between lapse type frequency and early (within first four weeks) as well as overall weight losses (at post-treatment). Data were determined to meet the necessary assumptions for linear regression (i.e., linear relationship, multicollinearity, no auto-correlation, homoscedasticity).

3. Results

3.1. Descriptive information

A total of 189 participants completed EMA at baseline. A total of 15,876 prompts were planned to be delivered to participants (84 prompts per person). The mean response rate among participants to prompted surveys was 82.4% (SD = 13.3%; range = 26.0–100.0%). Based on observed compliance distributions and consistent with prior publications (Forman et al., 2017), data from participants with less than 40% compliance were excluded from analyses. Thus, the final analysis sample size was 186 (98.4% of the total sample).

A total of 13,402 baseline EMA recordings were provided by the 186 participants, representing 2470 participant days. Approximately 72% (9631) EMA recordings were made on weekdays given the 14-day protocol. An equivalent proportion of ratings were made on Saturdays and Sundays (14.0% and 14.2% respectively) to other days of the week (range = 13.7%–15.4%). Participants reported a total of 1452 lapses with an average of 7.81 lapses over the 14-day period (SD = 6.55, range = 0–32). Fourteen participants reported no lapses during the assessment period (7.52% of the sample analyzed). Subsequent analyses are limited to the 172 participants reporting at least once lapse.

3.2. Characterizing lapse types

Table 1 shows time and location frequency data related to each lapse type. As reported in Forman et al. (2017), eating an unintended food was the most commonly reported type of lapse, followed by eating a larger portion than intended and eating at an unintended time. Chi-square tests of independence revealed significant associations between lapse type and location, χ2 (10, N = 1438) = 73.08, p < .001, as well as time of day, χ2 (4, N = 1438) = 15.87, p = .003. There were no significant associations between lapse type and weekday, χ2 (2, N = 1438) = 1.41, p = .49.

Table 1.

Lapse type descriptive information.

Lapse Type
Unintended Food (n = 641) Unplanned Time (n = 436) Larger Portion (n = 362)
Time
Morning 13.10% 13.79% 9.39%
Afternoon 41.34% 34.94% 32.87%
Evening 45.55% 51.26% 57.73%
Day
Weekday 45.40% 49.43% 47.51%
Weekend 54.60% 50.57% 52.48%
Location
Home 42.09% 57.94% 48.74%
Out 20.57% 10.28% 29.97%
School 0.15% 0.24% 0%
Work 16.61% 16.59% 9.80%
Other 20.56% 14.95% 11.48%
Open in a new tab

Most participants reported at least two types of lapses, but 5.3%, 6.5%, and 7.5% of participants reported only eating an unintended food, only eating at an unplanned time, and only eating larger portions than intended, respectively. Conversely, 18.6% never reported eating an unintended food, 24.4% never reported eating at an unplanned time, and 25.6% never reported eating a larger portion than intended.

3.3. Predictors for lapse types

Table 2 presents descriptive information for physical, affective, and environmental predictors of dietary lapse. Tables 3–5 present model information for triggers for each GEE model. Report of a lapse at the current survey significantly predicted lapse occurrence at the subsequent survey in all models (ps < .001). The odds of each lapse type occurring at the next survey based on current level of each predictor is displayed in Figs. 1–3.

Table 2.

Descriptive data for lapse triggers.

Trigger Mean Response
Mean SD
Sadness 1.27 0.66
Loneliness 1.19 0.56
Boredom 1.17 0.51
Anger/Irritation 1.33 0.68
Stress 1.75 0.92
Hunger 1.86 0.99
Deprivation 1.72 0.92
Tired 1.73 0.88
Presence of Delicious Food 0.41 0.49
Open in a new tab

Note: All variables are on scale 1 = Not at all to 5 = Extremely; Presence of Delicious Food is dichotomous variable.

Table 3.

Between- and within-person effects for triggers predicting eating an unintended food.

B SE 95% CI Wald χ2 p
Between-person Effects
Stress 0.31 0.10 [0.12,0.51] 9.41 < .01
Bored 0.03 0.24 [−0.43,0.48] 0.02 0.89
Anger/Irritation 0.35 0.12 [0.11,0.59] 8.02 < .01
Sadness 0.19 0.12 [−0.06,0.43] 2.23 0.14
Lonely 0.14 0.14 [−0.14,0.42] 0.97 0.32
Deprived 0.22 0.10 [0.02,0.41] 4.86 .03
Hungry 0.23 0.14 [−0.05,0.51] 2.61 0.11
Tired 0.34 0.12 [0.11,0.57] 8.69 < .01
Delicious Food Available 0.73 0.24 [0.25,1.20] 9.07 < .01
Within-person Effects
Stress −0.05 0.02 [−0.10,−0.01] 5.02 0.03
Bored 0.08 0.03 [0.01,0.14] 5.72 0.02
Anger/Irritation −0.09 0.03 [−0.14,−0.04] 11.75 .001
Sadness −0.05 0.04 [−.13,0.03] 1.46 0.23
Lonely 0.004 0.04 [−0.08,0.08] 0.01 0.93
Deprived 0.08 0.02 [0.03,0.12] 11.68 .001
Hungry 0.07 0.01 [0.04,0.09] 21.66 < .001
Tired −0.01 0.02 [−0.06,0.03] 0.22 0.64
Delicious Food Available 0.02 0.04 [−0.06,0.10] 0.31 0.58
Open in a new tab

Table 5.

Between- and within-person effects for triggers predicting eating a larger portion.

B SE 95% CI Wald χ2 p
Between-person Effects
Stress 0.34 0.11 [0.12, 0.56] 9.19 < .01
Bored 0.61 0.18 [0.27,0.97] 11.89 .001
Anger/Irritation 0.45 0.18 [0.10,.80] 6.47 .01
Sadness 0.24 0.19 [−0.15,0.62] 1.44 0.23
Lonely 0.41 0.16 [0.16,0.72] 6.65 .01
Deprived 0.26 0.11 [0.05,0.46] 5.75 0.02
Hungry 0.26 0.15 [0.14,0.24] 2.86 0.09
Tired 0.29 0.10 [0.09,0.50] 8.18 < .01
Delicious Food Available 0.87 0.27 [0.35,1.39] 10.70 .001
Within-person Effects
Stress −0.10 0.04 [−0.19, −0.17] 5.50 0.02
Bored 0.08 0.06 [−0.04, .19] 1.89 0.17
Anger/Irritation −0.04 0.05 [−0.13,0.05] 0.61 0.43
Sadness −0.03 0.07 [−0.17,−0.11] 0.16 0.69
Lonely −0.03 0.06 [−0.16,0.09] 0.22 0.64
Deprived 0.11 0.04 [0.03,0.19] 7.56 < .01
Hungry 0.19 0.03 [0.14,0.24] 56.94 < .001
Tired 0.08 0.04 [−0.13,0.17] 2.84 0.09
Delicious Food Available −0.11 0.19 [−0.26, 0.04] 2.08 0.14
Open in a new tab

Fig. 1.

Fig. 1.

Open in a new tab

Odds ratios of eating an unintended food based on each predictor.

Note. *p < .01. Red line indicates level at which the odds of lapsing are not influenced by trigger.

Fig. 3.

Fig. 3.

Open in a new tab

Odds ratios of eating a larger portion than expected based on each predictor.(For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Note. *p < .01. Red line indicates level at which the odds of lapsing are not influenced by trigger.

3.3.1. Eating an Unintended Food

Prior to running all GEE models to predict eating an unintended food, lapse cases representing occurrences of other lapse types were removed from analyses (n = 797 lapse cases, 5.9% of all lapse and non-lapse cases). Statistically significant between-subjects effects were observed such that participants who reported greater-than-average levels of stress, irritability, fatigue, and exposure to delicious foods across EMA surveys relative to other participants in the sample were more likely to lapse by eating an unintended food at any given survey (see Table 3). Momentary (i.e., within-subject) changes in irritability, deprivation, and hunger prospectively predicted eating an unintended food (Table 3). The odds of eating an unintended food increased by 1.08 times for each unit increase in deprivation and hunger and decreased by .91 times for each unit increase in irritability (see Fig. 1).

3.3.2. Eating at an Unplanned Time

Prior to running all GEE models to predict eating at an unplanned time, lapse cases representing occurrences of other lapse types were removed from analyses (n = 1003 lapse cases, 7.5% of all lapse and non-lapse cases). Statistically significant between-subjects effects were observed such that participants who reported greater-than-average levels of stress, boredom, loneliness, deprivation, and exposure to delicious foods across EMA surveys relative to other participants in the sample predicted occurrences in which a participant reported eating at a time they had not planned (Table 4). Momentary levels of stress and boredom prospectively predicted eating at an unintended time (Table 4). The odds of eating food at an unplanned time increased 1.2 times with each unit increase in boredom and decreased .89 times with each unit increase in stress (see Fig. 2).

Table 4.

Between- and within-person effects for triggers predicting eating at an unplanned time.

B SE 95% CI Wald χ2 p
Between-person Effects
Stress 0.27 0.11 [0.06,0.49] 6.09 0.01
Bored 0.31 0.13 [0.06,0.56] 6.04 0.01
Anger/Irritation 0.32 0.14 [0.05,0.60] 5.41 0.02
Sadness 0.26 0.13 [0.01,0.50] 4.03 0.05
Lonely 0.37 0.12 [0.14,0.61] 9.66 < .01
Deprived 0.37 0.11 [0.16,0.58] 12.22 < .001
Hungry 0.28 0.15 [−0.21,0.58] 3.33 0.07
Tired 0.23 0.12 [−0.002,0.45] 3.76 0.05
Delicious Food Available 0.68 0.27 [0.15,1.22] 6.63 .01
Within-person Effects
Stress −0.11 0.04 [−0.18,−0.03] 7.84 < .01
Bored 0.18 0.05 [0.09,0.28] 13.83 < .001
Anger/Irritation −0.01 0.04 [−0.10,0.08] 0.08 0.78
Sadness 0.04 0.07 [−0.09,0.17] 0.32 0.57
Lonely 0.07 0.06 [−0.04,0.19] 1.66 0.20
Deprived 0.06 0.03 [−0.01,0.12] 3.26 0.07
Hungry 0.04 0.02 [0.002,0.08] 4.18 0.04
Tired 0.05 0.03 [−0.01,0.12] 2.69 0.10
Delicious Food Available 0.07 0.06 [−0.04,0.18] 1.49 0.22
Open in a new tab
Fig. 2.

Fig. 2.

Open in a new tab

Odds ratios of eating at an unintended time based on each predictor.(For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Note. *p < .01. Red line indicates level at which the odds of lapsing are not influenced by trigger.

3.3.3. Eating a Larger Portion

Prior to running all GEE models to predict eating a larger portion than intended, lapse cases representing occurrences of other lapse types were removed from analyses (n = 1076 lapse cases, 8% of all lapse and non-lapse cases). Statistically significant between-subjects effects were observed such that participants who reported greater-than-average levels of stress, boredom, irritability, loneliness, fatigue, and exposure to delicious foods across EMA surveys relative to other participants in the sample were more likely to lapse by eating a larger portion at any given survey (Table 5). Momentary changes in deprivation and hunger levels prospectively predicted reports of eating a larger portion (Table 5). The odds of eating a larger portion than intended increased by approximately 1.1 times and 1.2 times for each unit increase in deprivation and hunger, respectively (see Fig. 3).

3.4. Association with weight loss

A series of multiple linear regressions were calculated to predict overall program weight losses based on early weight loss and frequency of each lapse type. Eating at an unplanned time significantly predicted both early (within first four weeks), F (1,184) = 7.58, p = .006, R2 = 0.40, and overall weight losses, F (1,184) = 8.77, p = .003, R2 = 0.46. Eating an unintended food was not a significant predictor of early, F (1,184) = 0.99, p = .32, R2 = 0.01, or overall weight losses, F (1,184) = 3.04, p = .08, R2 = 0.02. Eating a larger portion than intended was not significantly associated with early, F (1,184) = 3.26, p = .07 R2 = 0.02, or overall weight loss, F (1,184) = 0.13, p = .72, R2 = 0.001.

4. Discussion

The present study was an exploratory analysis of different lapse types (i.e., eating at an unplanned time, eating a larger portion than planned, eating an unintended food) among a sample of weight loss-seeking adults using EMA. Following data presented in Forman et al. (2017) regarding the frequency of lapse types, this was the first study to examine physical, affective, and environmental predictors of different lapse types. This was also the first study to examine the association between lapse types and weight loss. While replication and extension of the present findings is warranted given the exploratory nature of this study, our results suggest that the literature may benefit from assessing specific types of dietary lapses as lapse types appear to represent distinct eating behaviors with specific antecedents and potential to differentially impact weight losses. As such, findings can inform future research on lapse types and have the potential to optimize extant behavioral weight loss interventions.

Individuals appear to engage in multiple different types of lapses throughout early behavioral weight loss treatment as few participants endorsed only one lapse type during the assessment period. The most common lapse type, as reported here and in Forman et al. (2017), was eating an unintended food (which comprised nearly half of all reported lapses). The fact that eating an unintended food was so commonly reported is consistent with self-regulation theory of dieting (Baumeister & Heatherton, 1996). Given the obesogenic food environment, dieters are likely confronted with temptation regularly, and mindfulness of long-term goals can wane over time in the presence of short-term temptations (Herman & Polivy, 2004; Papies & Hamstra, 2010). When self-regulation fails, dieters are vulnerable to making impulsive decisions regarding food choice (Guerrieri, Nederkoorn, & Jansen, 2008), and this could result in lapsing by eating an unintended food.

Future research is needed to understand why eating an unintended food is more commonly endorsed than any other types of dietary lapse (i.e., eating at an unplanned time or eating a larger portion of food). It could be that, during a dieting attempt, individuals set clearer goals for foods they wanted to avoid and thus had more opportunities to violate these goals and experience this lapse type. Moreover, being exposed to forbidden foods is a common trigger for overeating among dieters and so it is not surprising that the majority of lapses were a result of eating a food that participants intended to avoid (Carels et al., 2002; Soetens, Braet, Van Vlierberghe, & Roets, 2008). Alternatively, participants may not have planned specific times for eating or particular portions to consume ahead of time, making lapses of these types less straightforward to identify and report. Given that we forced participants to classify lapses as one type (rather than falling into multiple categories), it is also possible that they were choosing the option that they felt best described the lapse, but perhaps did not characterize the eating episode in its entirety.

There were significant associations between lapse types and time and location when compared to expected levels if the relationship was due to chance. In particular, eating a forbidden food occurred less often at home, while eating at an unplanned time occurred more often at home and less often while at restaurants. Eating a larger portion occurred more often while at restaurants. These findings are consistent with research that suggests most calories are consumed in the home (Smith, Ng, & Popkin, 2013), but also that eating at restaurants can contribute to weight gain (Diliberti, Bordi, Conklin, Roe, & Rolls, 2004). With regards to time and lapse type, results indicated that evenings (when lapses are most often expected based on prior literature; Forman et al., 2017; McKee et al., 2014) were the most likely times for eating a larger portion than expected. This result is notable given that eating larger portions in the evenings is associated with overweight/obesity (Wang et al., 2014). Overall, findings related to time and location of lapse types suggest that different kinds of lapsing may have different situational risk factors, and therefore bolster evidence that lapse behaviors should be studied separately rather than together. Future research may examine what factors (e.g., social cues) place individuals at greater risk for different lapse types in certain locations and/or at certain times to tailor interventions that aim to prevent lapses.

To investigate contextual factors associated with dietary lapses, we examined whether average (i.e., between-subjects effects) or momentary (i.e., within-subjects effects) levels of several affective states, physical states, and environmental factors predicted specific lapse types. Overall, lapse types shared many of the same between-subjects predictors. In particular, individuals who reported greater average levels of stress and more common exposure to delicious foods across EMA surveys had a greater likelihood of experiencing any type of lapse at a given survey, which is consistent with prior research on the effects of stress and food environment on overeating (Epel, Lapidus, McEwen, & Brownell, 2001; Guerrieri et al., 2008; Roberts, Campbell, & Troop, 2014; Thomas et al., 2011).

When examining the impact of momentary affective and physical states, as well as environmental factors, on risk for specific lapse types, several interesting findings emerged. Specifically, when participants reported greater deprivation and hunger than was typical for them, the odds of lapsing by eating an unintended food or by eating more than planned increased (while the odds of eating at an unplanned time did not). It is perhaps unsurprising that these physical states would predict some types of lapses, as a physical drive to eat is a common precursor to eating episodes, whether “on plan” or a lapse (Lindroos et al., 1997). In fact, it is perhaps more surprising that momentary hunger and deprivation did not increase risk for eating at an unplanned time. One possible explanation is that participants felt justified in eating when hungry or deprived—even if they had not planned to eat at that time at the day’s outset—and thus did not perceive this eating occurrence as a lapse (Smith, O’neil, & Rhodes, 1999). It is also possible that the types of foods participants chose to eat when hungry or deprived were ones they were intending to avoid, and therefore these lapses were categorized as an “unintended food” rather than an “unplanned time” (Polivy, Coleman, & Herman, 2005; Urbszat, Herman, & Polivy, 2002).

The momentary triggers that did significantly predict eating at an unplanned time were boredom and stress. These data are consistent with the theory that emotions can influence eating behavior via difficulty regulating uncomfortable affective states (Evers, Marijn Stok, & de Ridder, 2010; Macht, 2008; Vohs & Heatherton, 2000). Specifically, results indicated that greater boredom increased the odds of eating at an unplanned time, while increases in stress decreased the odds of this lapse type. Interestingly, when participants reported greater irritability than was typical, their odds of lapsing by eating an unintended food decreased. This pattern of results makes sense in the context of emotion regulation research that has identified low-arousal states (such as boredom) as having disinhibiting effects on food intake and high-arousal states (such as stress and irritability) as having the opposite effect (Macht, 2008; Robbins & Fray, 1980). These findings, in combination with extant research, suggest that obesity treatments focused on coping with low-arousal states may have a positive impact on outcomes (Blaine, 2008; Elfhag & Rössner, 2005).

Lastly, it is somewhat surprising that there were a limited number of affective momentary predictors of lapse types given that emotional upset is often cited as a cause for overeating when reflecting on causes of lapses (Grilo et al., 1989). These results are also inconsistent with prior EMA studies on dietary lapse (Carels et al., 2002, 2004). Future research on the role of negative affect in lapses in general and on specific lapse types is needed. It is possible that there are associations between affective states and lapsing generally but that these emotions are not specific to any one lapse type. According to self-regulation and emotion regulation theories, there could be several important contributing factors to the relationship between affect and lapsing that were not assessed in this study, including but not limited to: impulsivity (Guerrieri, Nederkoorn, Schrooten, Martijn, & Jansen, 2009; Manasse, Crochiere, et al., 2018a), coping style (Evers et al., 2010), cognitive load (Lattimore & Maxwell, 2004; Ward & Mann, 2000), implicit attitudes (Hofmann, Rauch, & Gawronski, 2007), and other environmental factors (e.g., alcohol intake, Booth et al., 2001; Hofmann & Friese, 2008). Overall, findings support the importance of examining the “when,” “where,” and “why” of lapses so that we can efficiently tailor interventions to each individual (and perhaps each lapse type).

The only lapse type frequency that was related to weight loss at the beginning and end of treatment (i.e., at 12 months) was eating at an unplanned time. Specifically, more frequent lapses through eating at an unplanned time was related to lesser weight loss during early treatment and at 12 months. Lack of objective information about lapse size precludes our ability to conclude whether eating at an unplanned time actually leads to greater caloric intake or poorer macronutrient content, as has been shown for emotional eating (Oliver, Wardle, & Gibson, 2000; Willner et al., 1998). Therefore, future research is necessary to determine whether it is the lapse type, the unique triggers, or a combination that is associated with poorer weight loss. One possible explanation for these findings is that eating at an unplanned time is reflective of grazing behavior. Grazing is repetitious, unplanned eating that has been shown to contribute to weight gain (Colles, Dixon, & O’brien, 2008). Grazing behavior was not directly targeted in our weight loss intervention nor was it assessed in the present study, despite the possible relationship to poorer outcomes. Future work should consider assessing if “unplanned lapses” and “grazing” are distinguishable from one another.

This finding is also notable given that eating at an unplanned time was the least frequently reported lapse type. Further, eating at an unplanned time was the only lapse type for which risk increased based on greater momentary levels of an affective trigger (i.e., boredom), while momentary hunger and deprivation did not increase its odds of occurring. These findings suggest that, while less frequent (or less frequently identified and reported), eating at an unplanned time may be the most important lapse type for intervention, and there may be clear intervention targets. While further research is needed to better understand the implications of these findings, it is possible that eating in response to affective states (rather than physical states) and/or eating at an unplanned time are indicative of a more deleterious pattern of lapsing that interferes with weight loss. If so, it may be important to more directly address eating at unplanned times in response to boredom in weight control interventions to improve weight loss outcomes. Possible intervention strategies include teaching individuals to identify boredom and purposefully engage in alternative activities at these moments.

4.1. Strengths and limitations

The present study had several strengths. The use of EMA methods allowed for measurement of individuals’ physical, affective, and environmental states and lapse episodes in real-time and in participants’ natural environments, which presumably increased the validity and accuracy of participants’ reports. The repeated sampling of participants’ experiences, including at times when they had not recently lapsed, also allowed for examination of prospective triggers of dietary lapses, which enabled us to identify causal predictors of lapses (rather than concurrent associations). The present study also benefited from a large sample and two weeks of lapse assessment, which provided a large amount of data with which to examine the proposed research questions. Finally, by requiring participants to select specific lapse types when reporting on lapse behavior versus relying on a broader definition, participants may have been better able to identify and report on lapses, resulting in a more accurate portrayal of the frequency and nature of difficulties with dietary adherence during early behavioral weight loss.

There are also several limitations to the present research. First, while the benefits of EMA are notable, this methodology still relies on participants to self-report lapse episodes and their contextual states. While the amount of training in identifying dietary lapses was adequate and consistent with literature recommendations (Shiffman et al., 2008), it should still be recognized that participants’ awareness and insight of eating behavior (especially within an overweight/obese population) is limited (Bartholome, Peterson, Raatz, & Raymond, 2013; Lichtman et al., 1992). We also took appropriate measures to minimize bias in reporting (Shiffman et al., 2008), however, participant motivation to report behaviors that they may perceive as undesirable could also confound results (Smyth et al., 2001).

Another factor that could have influenced the quality of data collected was participant compliance (Shiffman et al., 2008). Many participants were carrying the provided EMA device in addition to their personal cellular devices, which likely incurred burden. As such, logistic constraints contributed to non-response rates to prompts, including participants forgetting the device at home while out, letting the device run out of power, and not connecting the device to WiFi when in a new location. Another important logistic limitation in our assessment procedure was that while we did collect data on times participants responded to prompts, we did collect data on the times when those respective prompts were received, and consequently cannot calculate latency in response time, which would have yielded important information about lapse assessment (Liao, Skelton, Dunton, & Bruening, 2016).

A final consideration that is specific to using EMA to study lapses is that the assessment protocol (two weeks) inherently provided far fewer lapse assessments on weekend days (vs. weekdays). This point is especially salient for studying eating behavior because of the known difference in quality and quantity of intake on the weekends (Haines, Hama, Guilkey, & Popkin, 2003). Overall, the limitations discussed thus far represent more general limitations of EMA methods, which should be considered by future researchers using these methods to assess eating behavior.

We are aware that this is the first study to assess lapse types using EMA and therefore acknowledge that there may be different techniques for asking about dietary lapse types via EMA. For example, we asked participants to classify each lapse as one lapse type and it is possible that one eating instance could be classified as multiple lapse types. Best practices for assessing lapse types warrant further study. Moreover, the current analyses did not directly compare differential predictors of lapse types; instead, predictors of each lapse type were examined independently with other lapse types removed from the data pool. Consequently, we are unable to comment on the relative strength of specific predictors for different lapse types (e.g., which type of lapse average levels of stress most strongly predicted).

Other notable methodological limitations include the diversity of our sample and the risk of Type I error. Our sample also was predominantly White, female, and middle-aged. While this is a common demographic makeup in behavioral weight loss studies (Pagoto et al., 2012; West, Prewitt, Bursac, & Felix, 2008), further research on lapse types with more diverse samples (e.g., men, minority groups) and with individuals not engaged in formal treatment (e.g., with individuals engaged in self-directed weight loss attempts) is warranted. Finally, we conducted multiple statistical comparisons. Although we attempted to correct for this by reducing the alpha level used to interpret statistically significant results to .01, caution is warranted when interpreting our findings and there is need for replication.

In addition to addressing the aforementioned limitations, future research should capitalize on these preliminary data by informing novel interventions that directly manipulate known lapse triggers (i.e., boredom) or target salient lapse types (i.e., unplanned lapses) for weight loss. We can also use data from the current study to inform hypotheses about specific combinations of triggers that might increase risk for specific lapse types. For example, it is possible that affective triggers alone do not increase risk of eating an unplanned food, but experiencing certain emotions in combination with being in a high-risk location/time (e.g., being at home in the evening) does increase risk.

5. Conclusion

In conclusion, it appears that individuals with overweight/obesity undergoing behavioral weight loss treatment can report on different lapse types early in treatment. Additionally, these different lapse types are predicted by several types of triggers and have different impacts on weight loss success during treatment. In other words, it appears that not all dietary lapses are created equal in terms of their triggers and impact on weight loss outcomes. Consequently, future research on dietary lapses may benefit from employing a definition and method of assessing lapses that accounts for the different lapse types an individual may experience at any particular moment. Doing so may lead to a more sophisticated understanding of factors that interfere with successful dietary adherence during weight loss, and ultimately allow for effective prevention of dietary lapses. For example, knowing what particular triggers or combination of triggers places individuals at risk for certain types of lapses may enable these risks to be addressed during treatment, either in-person with a clinician or even in the precise moment of risk via personalized interventions delivered via technology (e.g., smartphone).

Acknowledgements

We would like to thank the participants of this study as well at its research coordinators, Andrew Frohn, Emily Wyckoff, and Gerald Martin.

Funding

This work was supported by the National Institute of Diabetes and Digestive and Kidney Diseases (grant number R01DK095069 awarded to Evan M. Forman).

Appendix A. Supplementary data

Supplementary data related to this article can be found at https://doi.org/10.1016/j.appet.2018.07.003.

References

  1. Bartholome LT, Peterson RE, Raatz SK, & Raymond NC (2013). A comparison of the accuracy of self-reported intake with measured intake of a laboratory overeating episode in overweight and obese women with and without binge eating disorder. European Journal of Nutrition, 52(1), 193–202. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Baumeister RF, & Heatherton TF (1996). Self-regulation failure: An overview. Psychological Inquiry, 7(1), 1–15. [Google Scholar]
  3. Blaine B (2008). Does depression cause obesity? A meta-analysis of longitudinal studies of depression and weight control. Journal of Health Psychology, 13(8), 1190–1197. [DOI] [PubMed] [Google Scholar]
  4. Blair SN, & Leermakers EA (2002). Exercise and weight management. Handbook of Obesity Treatment, 437, 283–300. [Google Scholar]
  5. Booth SL, Sallis JF, Ritenbaugh C, Hill JO, Birch LL, Frank LD, et al. (2001). Environmental and societal factors affect food choice and physical activity: Rationale, influences, and leverage points. Nutrition Reviews, 59(3), S21–S36. [DOI] [PubMed] [Google Scholar]
  6. Burke LE, Ewing LJ, Shiffman S, Siewiorek D, Smailagic A, Kriska A, et al. (2015). Abstract P117: Day to day variation in self-efficacy associated with weight loss and risk of dietary lapses. Am Heart Assoc. [Google Scholar]
  7. Carels RA, Douglass OM, Cacciapaglia HM, & O’Brien WH (2004). An ecological momentary assessment of relapse crises in dieting. Journal of Consulting and Clinical Psychology, 72(2), 341. [DOI] [PubMed] [Google Scholar]
  8. Carels RA, Hoffman J, Collins A, Raber AC, Cacciapaglia H, & O’Brien WH (2002). Ecological momentary assessment of temptation and lapse in dieting. Eating Behaviors, 2(4), 307–321. [DOI] [PubMed] [Google Scholar]
  9. Colles SL, Dixon JB, & O’brien PE (2008). Grazing and loss of control related to eating: Two high-risk factors following bariatric surgery. Obesity, 16(3), 615–622. [DOI] [PubMed] [Google Scholar]
  10. Diabetes Prevention Program Research Group (2002). The Diabetes prevention program (DPP): Description of lifestyle intervention. Diabetes Care, 25(12), 2165–2171. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Diliberti N, Bordi PL, Conklin MT, Roe LS, & Rolls BJ (2004). Increased portion size leads to increased energy intake in a restaurant meal. Obesity, 12(3), 562–568. [DOI] [PubMed] [Google Scholar]
  12. Elfhag K, & Rössner S (2005). Who succeeds in maintaining weight loss? A conceptual review of factors associated with weight loss maintenance and weight regain. Obesity Reviews, 6(1), 67–85. [DOI] [PubMed] [Google Scholar]
  13. Epel E, Lapidus R, McEwen B, & Brownell K (2001). Stress may add bite to appetite in women: A laboratory study of stress-induced cortisol and eating behavior. Psychoneuroendocrinology, 26(1), 37–49. [DOI] [PubMed] [Google Scholar]
  14. Evers C, Marijn Stok F, & de Ridder DT (2010). Feeding your feelings: Emotion regulation strategies and emotional eating. Personality and Social Psychology Bulletin, 36(6), 792–804. [DOI] [PubMed] [Google Scholar]
  15. Forman EM, Butryn ML, Manasse SM, Crosby RD, Goldstein SP, Wyckoff EP, et al. (2016). Acceptance-based versus standard behavioral treatment for obesity: Results from the mind your health randomized controlled trial. Obesity, 24(10), 2050–2056. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Forman EM, Schumacher LM, Crosby R, Manasse SM, Goldstein SP, Butryn ML, et al. (2017). Ecological momentary assessment of dietary lapses across behavioral weight loss treatment: Characteristics, predictors, and relationships with weight change. Annals of Behavioral Medicine, 51(5), 741–753. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Goldstein H, Browne W, & Rasbash J (2002). Partitioning variation in multilevel models. Understanding statistics: Statistical issues in psychology, education, and the Social Sciences, 1(4), 223–231. [Google Scholar]
  18. Grilo CM, Shiffman S, & Wing RR (1989). Relapse crises and coping among dieters. Journal of Consulting and Clinical Psychology, 57(4), 488. [DOI] [PubMed] [Google Scholar]
  19. Grilo CM, Shiffman S, & Wing RR (1993). Coping with dietary relapse crises and their aftermath. Addictive Behaviors, 18(1), 89–102. [DOI] [PubMed] [Google Scholar]
  20. Guerrieri R, Nederkoorn C, & Jansen A (2008). The interaction between impulsivity and a varied food environment: Its influence on food intake and overweight. International Journal of Obesity, 32(4), 708. [DOI] [PubMed] [Google Scholar]
  21. Guerrieri R, Nederkoorn C, Schrooten M, Martijn C, & Jansen A (2009). Inducing impulsivity leads high and low restrained eaters into overeating, whereas current dieters stick to their diet. Appetite, 53(1), 93–100. [DOI] [PubMed] [Google Scholar]
  22. Haines PS, Hama MY, Guilkey DK, & Popkin BM (2003). Weekend eating in the United States is linked with greater energy, fat, and alcohol intake. Obesity, 11(8), 945–949. [DOI] [PubMed] [Google Scholar]
  23. Herman CP, & Polivy J (2004). The self-regulation of eating: Theoretical and practical problems. [Google Scholar]
  24. Hofmann W, & Friese M (2008). Impulses got the better of me: Alcohol moderates the influence of implicit attitudes toward food cues on eating behavior. Journal of Abnormal Psychology, 117(2), 420. [DOI] [PubMed] [Google Scholar]
  25. Hofmann W, Rauch W, & Gawronski B (2007). And deplete us not into temptation: Automatic attitudes, dietary restraint, and self-regulatory resources as determinants of eating behavior. Journal of Experimental Social Psychology, 43(3), 497–504. [Google Scholar]
  26. Johnson AW (2013). Eating beyond metabolic need: How environmental cues influence feeding behavior. Trends in Neurosciences, 36(2), 101–109. [DOI] [PubMed] [Google Scholar]
  27. Klem ML, Wing RR, McGuire MT, Seagle HM, & Hill JO (1997). A descriptive study of individuals successful at long-term maintenance of substantial weight loss. American Journal of Clinical Nutrition, 66(2), 239–246. [DOI] [PubMed] [Google Scholar]
  28. Lattimore P, & Maxwell L (2004). Cognitive load, stress, and disinhibited eating. Eating Behaviors, 5(4), 315–324. [DOI] [PubMed] [Google Scholar]
  29. Liao Y, Skelton K, Dunton G, & Bruening M (2016). A systematic review of methods and procedures used in ecological momentary assessments of diet and physical activity research in youth: An adapted STROBE checklist for reporting EMA studies (CREMAS). Journal of Medical Internet Research, 18(6). [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Lichtman SW, Pisarska K, Berman ER, Pestone M, Dowling H, Offenbacher E, et al. (1992). Discrepancy between self-reported and actual caloric intake and exercise in obese subjects. New England Journal of Medicine, 327(27), 1893–1898. [DOI] [PubMed] [Google Scholar]
  31. Lindroos AK, Lissner L, Mathiassen ME, Karlsson J, Sullivan M, Bengtsson C, et al. (1997). Dietary intake in relation to restrained eating, disinhibition, and hunger in obese and nonobese Swedish women. Obesity, 5(3), 175–182. [DOI] [PubMed] [Google Scholar]
  32. Look AHEAD Research Group (2006). The Look ahead study: A description of the lifestyle intervention and the evidence supporting it. Obesity (Silver Spring, Md.), 14(5), 737. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Lowe MR (2003). Self-regulation of energy intake in the prevention and treatment of Obesity: Is it feasible? Obesity Research, 11(S10), 44S–59S. [DOI] [PubMed] [Google Scholar]
  34. Macht M (2008). How emotions affect eating: A five-way model. Appetite, 50(1), 1–11. [DOI] [PubMed] [Google Scholar]
  35. Manasse SM, Crochiere RJ, Dallal DH, Lieber EW, Schumacher LM, Crosby RD, et al. (2018a). A multimodal investigation of impulsivity as a moderator of the relation between momentary elevations in negative internal states and subsequent dietary lapses. Appetite, 127, 52–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Manasse SM, Schumacher LM, Goldstein SP, Martin GJ, Crosby RD, Juarascio AS, et al. (2018b). Are individuals with loss-of-control eating more prone to dietary lapse in behavioural weight loss treatment? An ecological momentary assessment study. European Eating Disorders Review, 26(3), 259–264. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Marlatt GA (1985). Relapse prevention: Theoretical rationale and overview of the model. Relapse Prevention, 3–70. [Google Scholar]
  38. McKee HC, Ntoumanis N, & Taylor IM (2014). An ecological momentary assessment of lapse occurrences in dieters. Annals of Behavioral Medicine, 48(3), 300–310. [DOI] [PubMed] [Google Scholar]
  39. Moynihan AB, Van Tilburg WA, Igou ER, Wisman A, Donnelly AE, & Mulcaire JB (2015). Eaten up by boredom: Consuming food to escape awareness of the bored self. Frontiers in Psychology, 6, 369. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Nestle M (2003). Increasing portion sizes in american diets: More calories, more obesity. Journal of the American Dietetic Association, 103(1), 39–40. [DOI] [PubMed] [Google Scholar]
  41. Ogden CL, Carroll MD, Kit BK, & Flegal KM (2014). Prevalence of childhood and adult obesity in the United States, 2011–2012. Jama, 311(8), 806–814. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Oliver G, Wardle J, & Gibson EL (2000). Stress and food choice: A laboratory study. Psychosomatic Medicine, 62(6), 853–865. [DOI] [PubMed] [Google Scholar]
  43. Pagoto SL, Schneider KL, Oleski JL, Luciani JM, Bodenlos JS, & Whited MC (2012). Male inclusion in randomized controlled trials of lifestyle weight loss interventions. Obesity, 20(6), 1234–1239. [DOI] [PubMed] [Google Scholar]
  44. Papies EK, & Hamstra P (2010). Goal priming and eating behavior: Enhancing self-regulation by environmental cues. Health Psychology, 29(4), 384. [DOI] [PubMed] [Google Scholar]
  45. Polivy J, Coleman J, & Herman CP (2005). The effect of deprivation on food cravings and eating behavior in restrained and unrestrained eaters. International Journal of Eating Disorders, 38(4), 301–309. [DOI] [PubMed] [Google Scholar]
  46. Riley WT, Rivera DE, Atienza AA, Nilsen W, Allison SM, & Mermelstein R (2011). Health behavior models in the age of mobile interventions: Are our theories up to the task? Translational Behavioral Medicine, 1(1), 53–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Robbins T, & Fray P (1980). Stress-induced eating: Fact, fiction or misunderstanding? Appetite, 1(2), 103–133. [Google Scholar]
  48. Roberts CJ, Campbell IC, & Troop N (2014). Increases in weight during chronic stress are partially associated with a switch in food choice towards increased carbohydrate and saturated fat intake. European Eating Disorders Review, 22(1), 77–82. [DOI] [PubMed] [Google Scholar]
  49. Schumacher LM, Martin GJ, Goldstein SP, Manasse SM, Crosby RD, Butryn ML, et al. (2017). Brief Report: Ecological momentary assessment of cognitive-affective responses to dietary lapses. Health psychology : Official Journal of the Division of Health Psychology, American Psychological Association, 37(2), 148–152. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Shick SM, Wing RR, Klem ML, McGUIRE MT, Hill JO, & Seagle H (1998). Persons successful at long-term weight loss and maintenance continue to consume a low-energy, low-fat diet. Journal of the American Dietetic Association, 98(4), 408–413. [DOI] [PubMed] [Google Scholar]
  51. Shiffman S, Stone AA, & Hufford MR (2008). Ecological momentary assessment. Annual Review of Clinical Psychology, 4, 1–32. [DOI] [PubMed] [Google Scholar]
  52. Smith LP, Ng SW, & Popkin BM (2013). Trends in US home food preparation and consumption: Analysis of national nutrition surveys and time use studies from 1965–1966 to 2007–2008. Nutrition Journal, 12(1), 45. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Smith C, O’neil P, & Rhodes S (1999). Cognitive appraisals of dietary transgressions by obese women: Associations with self-reported eating behavior, depression, and actual weight loss. International Journal of Obesity, 23(3), 231. [DOI] [PubMed] [Google Scholar]
  54. 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. [DOI] [PubMed] [Google Scholar]
  55. Soetens B, Braet C, Van Vlierberghe L, & Roets A (2008). Resisting temptation: Effects of exposure to a forbidden food on eating behaviour. Appetite, 51(1), 202–205. [DOI] [PubMed] [Google Scholar]
  56. Stone AA, & Shiffman S (1994). Ecological momentary assessment (EMA) in behavioral medicine. Annals of Behavioral Medicine, 16(3), 199–202. [Google Scholar]
  57. Thomas JG, Doshi S, Crosby RD, & Lowe MR (2011). Ecological momentary assessment of obesogenic eating Behavior: Combining person-specific and environmental predictors. Obesity, 19(8), 1574–1579. [DOI] [PubMed] [Google Scholar]
  58. Thomas DM, Weedermann M, Fuemmeler BF, Martin CK, Dhurandhar NV, Bredlau C, et al. (2014). Dynamic model predicting overweight, obesity, and extreme obesity prevalence trends. Obesity, 22(2), 590–597. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Urbszat D, Herman CP, & Polivy J (2002). Eat, drink, and be merry, for tomorrow we diet: Effects of anticipated deprivation on food intake in restrained and unrestrained eaters. Journal of Abnormal Psychology, 111(2), 396. [DOI] [PubMed] [Google Scholar]
  60. Vohs KD, & Heatherton TF (2000). Self-regulatory failure: A resource-depletion approach. Psychological Science, 11(3), 249–254. [DOI] [PubMed] [Google Scholar]
  61. Wadden TA, & Butryn ML (2003). Behavioral treatment of obesity. Endocrinology and Metabolism Clinics, 32(4), 981–1003. [DOI] [PubMed] [Google Scholar]
  62. Wang J, Patterson R, Ang A, Emond J, Shetty N, & Arab L (2014). Timing of energy intake during the day is associated with the risk of obesity in adults. Journal of Human Nutrition and Dietetics, 27(s2), 255–262. [DOI] [PubMed] [Google Scholar]
  63. Ward A, & Mann T (2000). Don’t mind if I do: Disinhibited eating under cognitive load. Journal of Personality and Social Psychology, 78(4), 753. [DOI] [PubMed] [Google Scholar]
  64. Watson D, Clark LA, & Tellegen A (1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology, 54(6), 1063. [DOI] [PubMed] [Google Scholar]
  65. West DS, Prewitt TE, Bursac Z, & Felix HC (2008). Weight loss of black, white, and Hispanic men and women in the Diabetes Prevention Program. Obesity, 16(6), 1413–1420. [DOI] [PubMed] [Google Scholar]
  66. Williams EP, Mesidor M, Winters K, Dubbert PM, & Wyatt SB (2015). Overweight and obesity: Prevalence, consequences, and causes of a growing public health problem. Current Obesity Reports, 4(3), 363–370. [DOI] [PubMed] [Google Scholar]
  67. Willner P, Benton D, Brown E, Cheeta S, Davies G, Morgan J, et al. (1998). “Depression” increases “craving” for sweet rewards in animal and human models of depression and craving. Psychopharmacology, 136(3), 272–283. [DOI] [PubMed] [Google Scholar]
  68. Wilson GT (1994). Behavioral treatment of obesity: Thirty years and counting. Advances in Behaviour Research and Therapy, 16(1), 31–75. [Google Scholar]

RESOURCES