Skip to main content

Some NLM-NCBI services and products are experiencing heavy traffic, which may affect performance and availability. We apologize for the inconvenience and appreciate your patience. For assistance, please contact our Help Desk at info@ncbi.nlm.nih.gov.

NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Jan 1.
Published in final edited form as: Appetite. 2021 Oct 30;168:105789. doi: 10.1016/j.appet.2021.105789

External food cue responsiveness and emotional eating in adolescents: A multimethod study

Camille R Schneider-Worthington 1, Kathryn E Smith 2, James N Roemmich 3, Sarah-Jeanne Salvy 4
PMCID: PMC8671220  NIHMSID: NIHMS1753677  PMID: 34728251

Abstract

Eating in response to external food cues (i.e., external eating) and internal emotional experiences (i.e., emotional eating) are associated with obesity. While external and emotional eating co-occur, little is known about how external food cue responsiveness may interact with internal emotional cues to influence eating episodes in adolescents. The current study examined how trait-level external food cue responsiveness modulates momentary associations between affective states and eating in adolescents. Participants were drawn from a prior study of siblings (N=78; ages 13–17) who completed an ecological momentary assessment protocol to assess eating episodes and affective states. External food cue responsiveness was determined by comparing energy consumption following presentation of an appetizing food (pizza) on one day and a control activity (reading) on another day. Generalized linear mixed models examined positive and negative affective states, cue responsiveness, and their interactions as predictors of the likelihood of eating. The relationship between affective states and likelihood of eating was stronger among adolescents with higher versus lower external food cue responsiveness. Among adolescents with higher cue responsiveness, endorsing negative affect was associated with a lower likelihood of eating, whereas endorsing positive affect was associated with a higher likelihood of eating (within-person effects). Findings suggest that high sensitivity to external food cues and greater proclivity for emotional eating may be likely to coincide such that any cue, internal or external, is likely to disrupt sensitivity to internal hunger and satiety signals. Future studies are needed to elucidate how sensitivities to internal and external cues may interact to influence obesity risk.

Keywords: adolescence, food cue, emotional eating, ecological momentary assessment

1. Introduction

Eating behaviors are important contributors of children and adolescents’ health trajectories (Boutelle, Manzano, & Eichen, 2020; Schüz, Schüz, & Ferguson, 2015). As children transition into adolescence, they experience increasing autonomy over their dietary intake and are simultaneously navigating food-rich environments with a ubiquitous presence of palatable, calorie-dense foods. Moreover, adolescence is a time of heightened emotional reactivity and increased risk for maladaptive emotion regulation, including eating to cope with emotions (Compas et al., 2017; Gouveia, Canavarro, & Moreira, 2019). Consequently, self-regulation over eating and eating in response to hunger and satiety cues becomes more complex, as youth may come to associate other internal or external cues with food and eating (Birch, Fisher, & Davison, 2003; Snoek, van Strien, Janssens, & Engels, 2007; Wardle et al., 1992). Thus, early adolescence may be a particularly opportune time for studying the relationships between eating in response to external and internal cues.

Environmental cues play an important role in food consumption (Boswell & Kober, 2016; Steenhuis & Poelman, 2017). The sight and smell of food are powerful sensory stimuli driving food consumption (Boswell & Kober, 2016), and high responsiveness to food cues (external eating) has been linked to unhealthy eating practices and weight gain (Boswell & Kober, 2016). Schacter’s externality theory of obesity (Schachter, 1971; Schachter & Rodin, 1974) states that individuals with obesity are more responsive to external food cues and less sensitive to internal hunger and satiety signals. According to this theory, the combination of over-responsiveness to external food cues and decreased responsivity to homeostatic signals would lead to overeating and weight gain in an environment rife with food cues (e.g., highly palatable food is widely available, affordable, large portions common). However, aside from external environmental cues, emotions are an internal stimulus that may interfere with one’s awareness of internal hunger and satiety cues and influence eating decisions.

Emotional eating is the focus of another popularized model of self-regulation of intake and is typically defined as eating in response to an emotional state rather than homeostatic signals (Cardi, Leppanen, & Treasure, 2015). Prior literature has largely viewed emotional eating as a symptom of eating disorders or a behavior of individuals with obesity (Leehr et al., 2015). However, others have conceptualized emotional eating as a way to regulate affect (Kaplan & Kaplan, 1957). Negative emotions, such as anxiety, sadness, anger, boredom, and stress have been linked to poorer food choices (Aguiar-Bloemer & Diez-Garcia, 2018; Devonport, Nicholls, & Fullerton, 2019; Errisuriz, Pasch, & Perry, 2016), greater energy intake (Cardi et al., 2015), and obesity (Barnhart, Braden, & Dial, 2021; Braden, Musher-Eizenman, Watford, & Emley, 2018; Braet & Van Strien, 1997; Frayn & Knäuper, 2018). However, many of these studies were conducted in cohorts of predominately female adults (Aguiar-Bloemer & Diez-Garcia, 2018; Braden et al., 2018; Cardi et al., 2015; Devonport et al., 2019; Errisuriz et al., 2016) or relied on self-reported questionnaires (Barnhart et al., 2021; Braden et al., 2018; Devonport et al., 2019; Frayn & Knäuper, 2018) or isolated, standardized laboratory assessments of negative emotional eating (Aguiar-Bloemer & Diez-Garcia, 2018; Cardi et al., 2015; Devonport et al., 2019). A systematic review of studies examining emotions and eating behavior in adults concluded that participants ate more in response to general positive affect, when compared with general negative affect, based on evidence from non-experimental associational studies, laboratory studies of induced emotions and eating, and in naturalistic studies using diaries documenting snacking and emotions as they occurred in the natural environment (Devonport et al., 2019). Although positive emotional states have also been associated with hedonic eating and increased food consumption in adults (Cardi et al., 2015; Devonport et al., 2019; Evers, Adriaanse, de Ridder, & de Witt Huberts, 2013; Reichenberger et al., 2018), these associations have received less attention in youth (Mason, Do, Wang, & Dunton, 2020). In addition, there are studies in both adolescents (Snoek et al., 2007; Wardle et al., 1992) and adults (Braden et al., 2018; Thompson & Romeo, 2015) that have reported gender differences in the tendency to eat in response to emotional states, such that females would be more prone than males to emotional eating. However, the definition and assessment of emotional eating is nuanced and varies between studies.

Many studies rely on questionnaire-based measures of emotional eating, which typically assess an individual’s self-reported urge to eat in response to various emotions or situations (Meule, Reichenberger, & Blechert, 2018; Tatjana van Strien, Frijters, Bergers, & Defares, 1986). However, recent evidence suggests that scores on self-reported questionnaires of the urge to eat in response to emotional states do not coincide with actual food intake in response to emotional states measured in the laboratory using standardized procedures or in naturalistic settings (Altheimer, Giles, Remedios, Kanarek, & Urry, 2021; Bongers & Jansen, 2016; Braden, Emley, Watford, Anderson, & Musher-Eizenman, 2020). Further, while most studies rely on questionnaire-based or isolated, standardized laboratory assessments of emotional eating, these methods are limited in their ability to accurately capture reactivity to positive and negative emotional states as they occur in naturalistic settings. Among adolescents in particular, less is known on the influence of negative and positive emotional states on eating in naturalistic settings, though extant evidence has yielded inconsistent evidence (Mason et al., 2020). It’s also important to note there is evidence that the propensity to eat or not eat in response to emotional states may emerge in adolescence (Snoek et al., 2007; Wardle et al., 1992). As such, further research is warranted to examine emotional eating in naturalistic settings across development, as well as the extent to which individual differences may account for inconsistent findings.

At first glance, the externality and affect regulation theories of eating appear to provide opposing hypotheses in contrasting external and internal drivers of food consumption and self-regulation. Although it may be that individuals who are more responsive to external non-homeostatic cues are most likely to also eat in response to internal emotional cues, but this has not been previously studied. Interestingly, emotional eating and food cue responsiveness have been found to co-occur in numerous studies of preadolescent and adolescent children of both genders (Braet & Van Strien, 1997; Hirsch et al., 2014; T. van Strien & Oosterveld, 2008; Wardle, Guthrie, Sanderson, & Rapoport, 2001), including one study in a multi-ethnic cohort with equal distribution of African American, Hispanic, and White children (Ledoux, Watson, Baranowski, Tepper, & Baranowski, 2011). However, these previous reports have largely relied on self-report (or parent-report) questionnaires to assess emotional eating and food cue responsiveness in youth (Braet & Van Strien, 1997; Hirsch et al., 2014; Ledoux et al., 2011; T. van Strien & Oosterveld, 2008; Wardle et al., 2001). Despite the convenience of such assessments, these methods are limited by recall problems, response biases, and low ecological validity. Laboratory-based and naturalistic data collection (e.g., ecological momentary assessment [EMA]) has potential to increase methodological rigor and elucidate how external and emotional eating tendencies relate to adolescents’ eating behaviors as they occur in naturalistic settings.

To address the current gaps in the literature, this study combined rigorous laboratory assessment of food cue responsiveness and EMA to characterize the eating and emotional states as they occur in daily life in a sample of same-sex biological siblings. Our team has previously reported that siblings in this sample shared little similarity in food cue responsiveness (Ufholz, Salvy, Feda, Epstein, & Roemmich, 2019). The present study examined how external food cue responsiveness assessed in the laboratory moderated momentary associations between affective states and eating episodes in adolescents’ everyday life. If external and internal cues operate in opposite ways, then we should expect weaker relationships between internal emotional states and eating among adolescents who are more responsive to external food cues. Conversely, if external and internal cues operate similarly in overriding sensitivity to hunger and satiety signals, then we would expect stronger relationships (positive or negative) between emotional states and eating among adolescents who are more responsive to external food cues. Based on previous studies suggesting gender differences in emotional eating (Snoek et al., 2007; Thompson & Romeo, 2015; Wardle et al., 1992), as an exploratory aim we further examined whether the relationships between emotional states and eating differ among boys and girls.

2. Methods

2.1. Participants

Forty pairs of same-sex biologic siblings (N=80; ages 13–17, no more than four years apart) were originally recruited as part of a larger study evaluating putative factors contributing to differences in energy balance behaviors and adiposity among weight-discordant siblings. Recruitment information and details regarding the full study protocol have been described previously (Salvy, Feda, Epstein, & Roemmich, 2017a, 2017b; Smith et al., 2020; Ufholz et al., 2019). The Institutional Review Board at the University of Buffalo approved all study procedures. Parents provided written informed consent and adolescents provided assent for study participation. The present study excluded two participants due to missing data (see Results section below), resulting in a final analytic sample of 78 participants.

2.2. Procedures

Participants attended three laboratory visits over three weeks, during which a series of measures and laboratory assessments were completed, which have been previously described (Salvy et al., 2017a, 2017b; Smith et al., 2020; Ufholz et al., 2019). At the first visit, participants were screened, demographic information was collected via questionnaire, and height and weight measurements were collected. Following this first visit, participants’ momentary affect and eating were captured using EMA. The EMA protocol took place over the course of seven consecutive days (five weekdays, two weekend days) and was administered using a study-provided cell phone that delivered text messages to participants. Participants received text messages approximately every two hours between 3:00 p.m. and 9:00 p.m. on weekdays and between 10:00 a.m. and 10:00 p.m. on weekend days. A total of seven texts were sent on weekend days, and four texts were sent on weekdays, for a total possible 34 prompts over the course of the protocol. If there were issues with a participant’s responses (e.g., missing or unclear response), study staff contacted them the next working day to clarify. Participants were generally very responsive and rarely delayed in sending their responses. After the initial study visit, participants attended two subsequent study visits spaced one week apart over which the external food cue responsiveness protocol was completed. External food cue responsiveness was measured using two conditions (a control and cue-exposure condition), each assessed at a separate study visit. The order of the conditions was counterbalanced across families.

2.3. Measures

2.3.1. Baseline participant characteristics

Participants were asked to report their demographic information including gender, age, grade-level, and race/ethnicity.

2.3.2. External food cue responsiveness

Sibling participants were measured on the same days and times in identical, separate rooms. Parents were instructed to not give their children any of the cued food the day before or on the day of testing. Sibling participants fasted for three hours before testing, as confirmed by the parents. To simulate a dinner meal, testing occurred at the families’ usual dinner time. Measurements of participant’s reaction to food cues has been previous described in detail (Ufholz et al., 2019). Briefly, prior to both conditions participants had 10 minutes of quiet time reading magazines devoid of food cues. For the cue-exposure (experimental) condition, participants were presented with a plate containing slices of warm cheese pizza and instructed to smell and think about the taste of pizza for five minutes. The experimenter modeled smelling a separate plate of pizza (D’Journo; 290 kcal, 131 g total, 10.0 g fat, 36.0 g carbohydrate, and 14.0 g fat for 1/6 of pizza). For the control condition, participants continued to read the magazines for an additional five minutes. Following the five-minute cue-exposure or control condition, participants were allowed to eat a plateful of pizza and 500 mL of water. To avoid ceiling effects, boys were presented with seven slices (1,400 kcal) and girls with five slices (1,000 kcal) of warm pizza. Participants were given 15 minutes to eat and instructed to complete a taste questionnaire for the pizza, which was used to blind participants to the purpose of the food intake session. Pizza was weighed before and after consumption. Eating in response to the external food cue (‘cue responsiveness index’) was determined by subtracting the grams of pizza consumed in the control condition from grams consumed in the experimental condition. Greater cue responsiveness indicates a greater amount of pizza consumed following exposure to the food cue.

2.3.3. BMI z-score

Participants’ height and weight were measured with an electronic scale (Model BWB-800S, Tanita, Portage, MI) and digital stadiometer (Model PE-AIM-101, Perspective Enterprises). Participants removed their shoes, outer wear, belts, and emptied their pockets prior to being weighed. Weight was measured to the nearest 0.01 kg. Height was measured twice (to the nearest 0.1 cm), and a third time if these two measurements were discrepant (differences > 0.5 cm). Means of these height measurements and measured weight were used to calculate BMI (kg/m2) and BMI z-scores using the Center for Disease Control and Prevention growth charts (Kuczmarski et al., 2000).

2.3.4. EMA-measured affect

At each EMA survey, participants were asked to report on their current behaviors and affective states. Participants responded (yes/no) indicating whether they were eating (coded as 0 [no] or 1 [yes] for each response). Participants also reported on their current affective states by responding (yes/no) to whether they were dealing with a problem, feeling happy, sad, frustrated, angry, stress, and/or bothered by something (coded as 0 [no] or 1 [yes] for each response). Responses for negative affect states (i.e., problem, sad, frustrated, angry, stress, bothered) were averaged at each signal to reflect the overall likelihood of endorsing negative affect.

2.4. Data Analyses

For the analyses, cue responsiveness index was included in models as a continuous variable. Separate generalized linear mixed models (GLMMs) for repeated measures examined EMA-measured positive and negative affective states (i.e., likelihood of endorsing happiness [positive affect] or problem, frustrated, angry, stress, and bothered [negative affect]), cue responsiveness index, and their interactions as predictors of EMA-measured likelihood of eating. Models also examined whether relationships between emotional states and eating differed by gender by including the two-way interaction between gender and affective state.

In each GLMM, siblings were nested within family, and affective states were separated into within-person (i.e., person-mean centered) and between-person (i.e., grand-mean centered) components, whereas cue responsiveness index, BMI z-score, and child age were assessed as between-person (i.e., grand-mean centered) components. That is, within-person associations indicate the degree to which changes in the independent variable (e.g., affective state), relative to an individual’s own mean affective state across the EMA protocol, are related to the dependent variable (e.g., likelihood of reporting eating), whereas between-person associations reflect the degree to which an individual’s average level of an independent variable (e.g., positive affectivity across the EMA protocol or age), relative to other individuals in the sample, is associated with the dependent variable (e.g., overall likelihood of eating). Given that the research question focused on momentary associations between affect and eating, interactions were examined for within-person but not between-person effects. Participant gender, BMI z-score, and age were included as covariates in all models.

Given the dichotomous nature of the dependent variable (likelihood of eating), binary logistic functions were used. As the sample involved multiple observations from siblings nested within families, each GLMM specified Level 1 (observation), Level 2 (person), and Level 3 (family) variables, and included a random effect of family to account for differences in dependent variables across families. GLMMs also specified an AR1 serial autocorrelation to account for dependence within the nested data. GLMM analyses were conducted using SPSS version 26. Statistically significant interaction terms were interpreted using a graphical approach. The “pick a point” approach was used to draw the graphs (Rogosa, 1980). This approach involves taking representative high (value = mean + 1 SD) and low (value = mean − 1 SD) values of between-person cue responsiveness and selecting representative high and low values (mean +/− 1 SD) of within-person negative affect (or happiness) and then estimating the effect of affective state at those values on the outcome.

3. Results

The analytic sample included 78 participants, as two participants were excluded due to missing data for the cue responsiveness task. Descriptive characteristics of the final analytic sample are shown in Table 1. Each participant reported an average of 32.5 ± 3.7 (range: 7–34) EMA recordings over the course of the sampling period, with an average compliance rate of 97.8% across participants. Negative feelings were endorsed on 19.7% of EMA recordings made while eating, whereas positive affect was endorsed on 91.6% of recordings made while eating.

Table 1:

Demographics (N=78).

Mean ± SD or n (%)
z-BMI 0.8 ± 0.9
Age (years) 15.4 ± 1.4
Male (n) 44 (56.4%)
Race/Ethnicity
 White, non-Hispanic (n) 70 (89.7%)
 White, Hispanic (n) 2 (2.6%)
 Black, non-Hispanic (n) 4 (5.1%)
 Multiracial/Other (n) 2 (2.6%)
Cue Responsiveness Index (Grams) −4.9 ± 82.4
Cue Responsiveness Index (kcals) −11.4 ±192.2
Instances reporting eating 3.2 ± 2.2
Instances endorsing negative affect 7.4 ± 7.6
Instances endorsing positive affect 29.0 ± 6.4

Results of GLMM models are presented in Table 2. Across models, there were significant main effects of gender, age, and z-BMI. Specifically, girls were significantly more likely to report eating compared to boys, and participants who were older and those with greater z-BMI were less likely to report eating.

Table 2:

Results of mixed effect models (multilevel regression estimates) predicting likelihood of eating when reporting negative or positive affect.

DV:Eating
Negative Affect Model B SE 95% Cl Exp(B) 95% Cl P
Intercept −2.55 0.13 (−2.81, −2.28) 0.08 (0.06, 0.10) <.001
Gender 0.55 0.20 (0.15, 0.94) 1.72 (1.16, 2.55) 0.007
Age −0.12 0.05 (−0.22, −0.01) 0.89 (0.80, 0.99) 0.026
z-BMI −0.20 0.08 (−0.35, −0.05) 0.82 (0.70, 0.95) 0.009
Cue responsiveness −0.001 0.001 (−0.003, 0.001) 1.00 (1.00, 1.00) 0.335
Negative affect (between) −0.42 0.38 (−1.16, 0.31) 0.66 (0.31, 1.37) 0.261
Negative affect (within) −1.29 0.41 (−2.09, −0.49) 0.27 (0.12, 0.61) 0.002
Gender × Negative affect (within) 1.19 0.47 (0.27, 2.11) 3.29 (1.31, 8.25) 0.011
Cue responsiveness × Negative affect (within) −0.007 0.003 (−0.01, −0.001) 0.99 (0.99, 1.00) 0.032
Positive Affect Model
Intercept −2.49 0.13 (−2.74, −2.23) 0.08 (0.06, 0.11) <.001
Gender 0.46 0.19 (0.09, 0.83) 1.58 (1.10, 2.28) 0.014
Age −0.12 0.05 (−0.22, −0.01) 0.89 (0.80, 0.99) 0.026
z-BMI −0.19 0.08 (−0.35, −0.04) 0.82 (0.71, 0.96) 0.014
Cue responsiveness −0.001 0.001 (−0.003, 0.001) 1.00 (1.00, 1.00) 0.430
Positive affect (between) 0.72 0.58 (−0.41, 1.88) 2.08 (0.66, 6.53) 0.210
Positive affect (within) 1.08 0.52 (0.07, 2.10) 2.95 (1.07, 8.15) 0.036
Gender × Positive affect (within) −1.01 0.60 (−2.17, 0.16) 0.37 (0.11, 1.18) 0.092
Cue responsiveness × Positive affect (within) 0.007 0.004 (−0.001, 0.02) 1.01 (1.00, 1.02) 0.074

Abbreviations: DV, dependent variable; CI, confidence interval, z-BMI, body mass index z-score.

Note: Between = grand-mean centered variable; within = person-mean centered variable.

Gender was coded such that boys were the reference category. Bolded values indicate significant effects at p < .05.

There were no significant main effects of cue responsiveness index or between-person negative affect on likelihood of eating (Table 2). There was a significant main effect of within-person negative affect on likelihood of eating (B=−1.29, SE=0.41, p=0.002; Table 2); participants were less likely to report eating when they reported negative affect compared to when they reported no negative affect. There was also a significant interaction of within-person negative affect and gender predicting likelihood of eating (B=1.19, SE=0.47, p=0.011; Table 2; Figure 1a). When endorsing negative affect males were less likely to report eating compared to when no negative affect was reported, whereas the likelihood of eating for females remained similar regardless of whether negative affect was reported. Additionally, there was a significant interaction of within-person negative affect and cue responsiveness predicting likelihood of eating (B=−0.007, SE=0.003, p=0.032; Table 2; Figure 2). Specifically, when adolescents with higher cue responsiveness reported negative affect, they were less likely to report eating compared to when they reported no negative affect. Among adolescents with lower cue responsiveness, the relationship between endorsing negative affect and likelihood of eating was also negative but statistically weaker.

Figure 1.

Figure 1.

Interaction of gender and within-person a) negative affect (NA) or b) positive affect predicting likelihood of eating. Not endorsed and endorsed values reflect 1 SD below and above individual means, respectively.

Figure 2.

Figure 2.

Interaction of cue responsiveness and within-person negative affect (NA) predicting likelihood of eating. Not endorsed and endorsed values reflect 1 SD below and above individual means, respectively. Low and high cue responsiveness reflect values 1 SD below and above sample means, respectively.

There were no significant main effects of cue responsiveness index or between-person positive affect on the likelihood of eating (Table 2). Within-person positive affect was significant (B=1.08, SE=0.52, p=0.036), such that participants were more likely to report eating when they reported positive affect compared to when they reported no positive affect. Males were more likely to report eating when positive affect was endorsed compared to when positive affect was not endorsed, while the likelihood of eating among females was similar regardless of whether positive affect was endorsed or not. However, the interaction of within-person positive affect with gender on likelihood of eating did not reach statistical significance (B=−1.01, SE=0.60, p=0.092; Table 2; Figure 1b). Additionally, the interaction of within-person positive affect and cue responsiveness was trending towards significance (B=0.007, SE=0.004, p=0.074; Table 2; Figure 3). Specifically, when adolescents with higher cue responsiveness reported positive affect, they were more likely to report eating compared to when they reported no positive affect. Among participants with lower cue responsiveness, the relationship between endorsing positive affect and likelihood of eating was also positive but statistically weaker.

Figure 3.

Figure 3.

Interaction of cue responsiveness index and within-person positive affect (PA) predicting likelihood of eating. Not endorsed and endorsed values reflect 1 SD below and above individual means, respectively. Low and high cue responsiveness reflect values 1 SD below and above sample means, respectively.

4. Discussion

To our knowledge this is the first study to investigate how objectively measured external food cue responsiveness moderates the relationship between internal emotional cues and eating episodes in adolescence. Among adolescents with higher cue responsiveness, the relationship between emotional states and likelihood of eating was stronger than among adolescents with lower cue responsiveness, although the direction of association differed by valence (i.e., positive versus negative affect). Consistent with previous studies (Bennett, Greene, & Schwartz-Barcott, 2013; Meule et al., 2018; Snoek et al., 2007; Thompson & Romeo, 2015; Wardle et al., 1992), boys in our sample were less likely to report eating when experiencing negative emotional states.

Our finding that emotional state was more strongly associated with the likelihood of eating among youth with higher cue reactivity compared to those with lower cue reactivity builds on prior questionnaire-based research that has reported a positive correlation between emotional and external eating in children and adolescents (Braet & Van Strien, 1997; Hirsch et al., 2014; Ledoux et al., 2011; T. van Strien & Oosterveld, 2008). Together, these findings suggest that youth who are more responsive to external cues may also be more responsive to internal emotional cues (irrespective of valence); conversely, youth with low cue responsiveness may be more attuned to internal hunger and satiety signals and less influenced by non-homeostatic external or internal factors. In this way, “external” or “emotional” eaters may be misnomers, as both internal and external cues can disrupt or override sensitivity to homeostatic eating drives.

Overall, we found that negative and positive affect were both associated with the likelihood of eating, but in opposite directions, and these associations appeared to be stronger among youth with higher versus low cue responsiveness. Reporting negative affect was associated with a lower likelihood of eating, whereas endorsing positive affect was associated with a higher likelihood of eating. The literature regarding stress and eating behaviors suggests that negative affect can either increase or decrease eating and that chronic stress may play a role in this association (Adam & Epel, 2007; Klatzkin, Baldassaro, & Hayden, 2018; Klatzkin et al., 2019). There is evidence that those with high chronic stress are more likely to eat in response to negative feelings (Klatzkin et al., 2018; Klatzkin et al., 2019), whereas non-chronically stressed, healthy controls may be less likely to eat in response to stress or negative emotions (Klatzkin et al., 2019; Reichenberger et al., 2018; Stone & Brownell, 1994). Given our sample was not a clinical group, they were unlikely to have high chronic stress and as such, may have been more likely to decrease rather than increase eating in response to negative emotions. The growing literature on positive affect and eating suggests that positive affect is associated with increased intake among both clinical samples (e.g., individuals with bulimia nervosa) and healthy controls (Cardi et al., 2015), which is consistent with our findings. Nevertheless, further research is needed to understand how internal and external cues interact and identify the cognitive and emotional pathways accounting for greater cue responsiveness and emotional eating.

The present study has several strengths, including the laboratory assessment of external food cue responsiveness and EMA to characterize emotional state and the incidence of eating in naturalistic settings. The focus on adolescents is relevant as this is a developmental period often overlooked despite evidence that adolescence is an important time for the emergence of external and emotional eating behaviors (Birch et al., 2003; Snoek et al., 2007; Wardle et al., 1992). Additionally, the parent study from which our sample was drawn was a weight-discordant biological sibling study. While not necessary for testing our hypothesis, discordant sibling designs account for approximately 50% of the genetic variability between siblings, as well as a portion of variance associated with shared environmental factors (e.g., having the same parents, neighborhood). Given that evidence suggests emotional functioning is influenced by environmental factors (McRae et al., 2017), this design is particularly useful to examine non-shared environmental effects among discordant siblings raised in the same family.

4.1. Limitations

It is also important to note limitations of the study. Affect and eating behaviors were assessed using dichotomous (yes/no) measures to minimize participant burden. As such, we were unable to characterize the intensity of emotions and their impact on more specific eating behaviors, such as the specific foods being eaten or type of eating occasion (e.g., snack, main meal). At the group level, participants in our sample were more likely to report eating while experiencing positive affect as opposed to negative affect. However, the observed moderation effects for affect were within-person and indicate the likelihood of eating when a participant with higher/lower cue responsiveness was experiencing relatively more/less positive/negative affect compared to his/her usual level. Therefore, our results are specific to the particular individual and his/her own tendencies to eat when experiencing positive/negative emotions, and overall descriptive data at the group-level are not directly applicable to these effects. Affect and eating behaviors were also assessed concurrently (i.e., at the same EMA signal), and thus micro-temporal relationships between these constructs, such as whether affective states were proximal triggers, concurrent experiences, or consequences of eating, were not assessed. This was done in part due to the significant time lag between EMA prompts (i.e., approximately two hours), which may fail to capture such micro-level effects. The EMA protocol also used a signal-contingent assessment schedule that only assessed experiences at four randomly selected times throughout the day on weekdays and seven randomly selected times on weekend days. Thus, the schedule of EMA signals may have missed some eating occasions (e.g., late-night), limiting our ability to capture all eating episodes across the seven-day EMA protocol and relevant affective states. There is also the possibility of reversed causality. Although we suggested that those with high cue responsiveness were more likely to eat in response to emotional state, it is also possible that those with high cue responsiveness were more likely to report more experiences of emotional affect throughout the EMA protocol overall not just when eating. Considering the interaction of within-person negative affect with gender was significant, it is possible that gender could moderate the observed interaction between cue responsiveness and negative affect on likelihood of eating. However, due to our modest sample size our study was not sufficiently powered to detect a significant 3-way interaction. Larger studies are needed to further examine the role of gender, if any, on the relationships between affective states, cue responsiveness and eating. Finally, the sample was relatively homogenous with respect to race, ethnicity and age; thus, the extent to which our findings may generalize to other demographic groups or younger children is unclear.

5. Conclusion

In summary, these results suggest that greater sensitivity to external food cues and greater proclivity for emotional eating may coincide such that any non-homeostatic cue, internal or external, is likely to disrupt sensitivity to internal hunger and satiety signals. In turn, adolescents with greater sensitivity to external food cues and emotional eating may be more susceptible to excess weight gain and obesity (Boutelle et al., 2020; Schüz et al., 2015). However, future studies are needed to replicate our findings in larger, more diverse samples and to further examine how sensitivities to internal and external cues, as well as gender, may interact to influence eating decisions and subsequent obesity risk. Interestingly, emotional valence appears to differentially influence the likelihood of eating, and especially among those with high external cue responsiveness. In addition to emotional valence, emotions may differentially influence eating behaviors depending on their intensity (Becker, Fischer, Smith, & Miller, 2016; Reichenberger et al., 2018; Stone & Brownell, 1994). As such, future studies should include more detailed assessments of affective states using visual-analogue scales or likert-scales, and collect data to further characterize eating behaviors including the type of eating occasion (e.g., snack, main meal) and specific foods being eaten. Moving forward, additional research is warranted to identify the underlying mechanisms accounting for greater cue responsiveness and eating in response to different emotional states, and how these mechanisms may overlap or interact to influence eating behaviors.

Funding

Research reported in this publication was supported by the Eunice Kennedy Shriver National Institute of Child Health and Human Development of the National Institutes of Health under award number R01HD064958 to JNR and the United States Department of Agriculture, Agricultural Research Service, 3062-51000-51- 00D. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.

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.

Declaration of competing interest

None.

Ethical Statement

Written informed consent was obtained from all study participants prior to enrollment. Specifically, at least one parent provided written informed consent and adolescents provided written assent. This study was approved by the Social and Behavioral Sciences Institutional Review Board at the University at Buffalo.

Contributor Information

Camille R. Schneider-Worthington, Division of Preventive Medicine, Department of Medicine, University of Alabama at Birmingham, 1720 2nd Avenue South, Birmingham, AL 35294;.

Kathryn E. Smith, Department of Psychiatry and Behavioral Sciences, University of Southern California, 2250 Alcazar St, #2200, Los Angeles, CA 90033;.

James N. Roemmich, Grand Forks Human Nutrition Research Center, Agricultural Research Service, United States Department of Agriculture, 2420 2nd Avenue N, Grand Forks, ND 58201;.

Sarah-Jeanne Salvy, Research Center for Health Equity, Cedars-Sinai Medical Center, 700 N. San Vicente Blvd., West Hollywood, CA 90069;.

References

  1. Adam TC, & Epel ES (2007). Stress, eating and the reward system. Physiol Behav, 91(4), 449–458. doi: 10.1016/j.physbeh.2007.04.011 [DOI] [PubMed] [Google Scholar]
  2. Aguiar-Bloemer AC, & Diez-Garcia RW (2018). Influence of emotions evoked by life events on food choice. Eat Weight Disord, 23(1), 45–53. doi: 10.1007/s40519-017-0468-8 [DOI] [PubMed] [Google Scholar]
  3. Altheimer G, Giles GE, Remedios JD, Kanarek RB, & Urry HL (2021). Do emotions predict eating? The role of previous experiences in emotional eating in the lab and in daily life. Appetite, 158, 105016. doi: 10.1016/j.appet.2020.105016 [DOI] [PubMed] [Google Scholar]
  4. Barnhart WR, Braden AL, & Dial LA (2021). Emotion Regulation Difficulties Strengthen Relationships Between Perceived Parental Feeding Practices and Emotional Eating: Findings from a Cross-Sectional Study. Int J Behav Med. doi: 10.1007/s12529-021-09959-z [DOI] [PubMed] [Google Scholar]
  5. Becker KD, Fischer S, Smith GT, & Miller JD (2016). The influence of negative urgency, attentional bias, and emotional dimensions on palatable food consumption. Appetite, 100, 236–243. doi: 10.1016/j.appet.2016.02.019 [DOI] [PubMed] [Google Scholar]
  6. Bennett J, Greene G, & Schwartz-Barcott D (2013). Perceptions of emotional eating behavior. A qualitative study of college students. Appetite, 60(1), 187–192. doi: 10.1016/j.appet.2012.09.023 [DOI] [PubMed] [Google Scholar]
  7. Birch LL, Fisher JO, & Davison KK (2003). Learning to overeat: maternal use of restrictive feeding practices promotes girls’ eating in the absence of hunger. Am J Clin Nutr, 78(2), 215–220. doi: 10.1093/ajcn/78.2.215 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Bongers P, & Jansen A (2016). Emotional Eating Is Not What You Think It Is and Emotional Eating Scales Do Not Measure What You Think They Measure. Front Psychol, 7, 1932. doi: 10.3389/fpsyg.2016.01932 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Boswell RG, & Kober H (2016). Food cue reactivity and craving predict eating and weight gain: a meta-analytic review. Obes Rev, 17(2), 159–177. doi: 10.1111/obr.12354 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Boutelle KN, Manzano MA, & Eichen DM (2020). Appetitive traits as targets for weight loss: The role of food cue responsiveness and satiety responsiveness. Physiol Behav, 224, 113018. doi: 10.1016/j.physbeh.2020.113018 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Braden A, Emley E, Watford T, Anderson L, & Musher-Eizenman D (2020). Self-reported emotional eating is not related to greater food intake: results from two laboratory studies. Psychol Health, 35(4), 500–517. doi: 10.1080/08870446.2019.1649406 [DOI] [PubMed] [Google Scholar]
  12. Braden A, Musher-Eizenman D, Watford T, & Emley E (2018). Eating when depressed, anxious, bored, or happy: Are emotional eating types associated with unique psychological and physical health correlates? Appetite, 125, 410–417. doi: 10.1016/j.appet.2018.02.022 [DOI] [PubMed] [Google Scholar]
  13. Braet C, & Van Strien T (1997). Assessment of emotional, externally induced and restrained eating behaviour in nine to twelve-year-old obese and non-obese children. Behaviour research and therapy, 35(9), 863–873. doi: 10.1016/s0005-7967(97)00045-4 [DOI] [PubMed] [Google Scholar]
  14. Cardi V, Leppanen J, & Treasure J (2015). The effects of negative and positive mood induction on eating behaviour: A meta-analysis of laboratory studies in the healthy population and eating and weight disorders. Neurosci Biobehav Rev, 57, 299–309. doi: 10.1016/j.neubiorev.2015.08.011 [DOI] [PubMed] [Google Scholar]
  15. Compas BE, Jaser SS, Bettis AH, Watson KH, Gruhn MA, Dunbar JP, … Thigpen JC (2017). Coping, emotion regulation, and psychopathology in childhood and adolescence: A meta-analysis and narrative review. Psychol Bull, 143(9), 939–991. doi: 10.1037/bul0000110 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Devonport TJ, Nicholls W, & Fullerton C (2019). A systematic review of the association between emotions and eating behaviour in normal and overweight adult populations. Journal of health psychology, 24(1), 3–24. doi: 10.1177/1359105317697813 [DOI] [PubMed] [Google Scholar]
  17. Errisuriz VL, Pasch KE, & Perry CL (2016). Perceived stress and dietary choices: The moderating role of stress management. Eating behaviors, 22, 211–216. doi: 10.1016/j.eatbeh.2016.06.008 [DOI] [PubMed] [Google Scholar]
  18. Evers C, Adriaanse M, de Ridder DT, & de Witt Huberts JC (2013). Good mood food. Positive emotion as a neglected trigger for food intake. Appetite, 68, 1–7. doi: 10.1016/j.appet.2013.04.007 [DOI] [PubMed] [Google Scholar]
  19. Frayn M, & Knäuper B (2018). Emotional Eating and Weight in Adults: a Review. Current Psychology, 37(4), 924–933. doi: 10.1007/s12144-017-9577-9 [DOI] [Google Scholar]
  20. Gouveia MJ, Canavarro MC, & Moreira H (2019). Associations between Mindfulness, Self-Compassion, Difficulties in Emotion Regulation, and Emotional Eating among Adolescents with Overweight/Obesity. Journal of Child and Family Studies, 28(1), 273–285. doi: 10.1007/s10826-018-1239-5 [DOI] [Google Scholar]
  21. Hirsch O, Kluckner VJ, Brandt S, Moss A, Weck M, Florath I, … Christiansen H (2014). Restrained and external-emotional eating patterns in young overweight children-results of the Ulm Birth Cohort Study. PLoS One, 9(8), e105303. doi: 10.1371/journal.pone.0105303 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Kaplan HI, & Kaplan HS (1957). The psychosomatic concept of obesity. J Nerv Ment Dis, 125(2), 181–201. doi: 10.1097/00005053-195704000-00004 [DOI] [PubMed] [Google Scholar]
  23. Klatzkin RR, Baldassaro A, & Hayden E (2018). The impact of chronic stress on the predictors of acute stress-induced eating in women. Appetite, 123, 343–351. doi: 10.1016/j.appet.2018.01.007 [DOI] [PubMed] [Google Scholar]
  24. Klatzkin RR, Dasani R, Warren M, Cattaneo C, Nadel T, Nikodem C, & Kissileff HR (2019). Negative affect is associated with increased stress-eating for women with high perceived life stress. Physiol Behav, 210, 112639. doi: 10.1016/j.physbeh.2019.112639 [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Kuczmarski RJ, Ogden CL, Grummer-Strawn LM, Flegal KM, Guo SS, Wei R, … Johnson CL (2000). CDC growth charts: United States. Adv Data(314), 1–27. [PubMed] [Google Scholar]
  26. Ledoux T, Watson K, Baranowski J, Tepper BJ, & Baranowski T (2011). Overeating styles and adiposity among multiethnic youth. Appetite, 56(1), 71–77. doi: 10.1016/j.appet.2010.11.145 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Leehr EJ, Krohmer K, Schag K, Dresler T, Zipfel S, & Giel KE (2015). Emotion regulation model in binge eating disorder and obesity--a systematic review. Neurosci Biobehav Rev, 49, 125–134. doi: 10.1016/j.neubiorev.2014.12.008 [DOI] [PubMed] [Google Scholar]
  28. Mason TB, Do B, Wang S, & Dunton GF (2020). Ecological momentary assessment of eating and dietary intake behaviors in children and adolescents: A systematic review of the literature. Appetite, 144, 104465. doi: 10.1016/j.appet.2019.104465 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. McRae K, Rhee SH, Gatt JM, Godinez D, Williams LM, & Gross JJ (2017). Genetic and environmental influences on emotion regulation: A twin study of cognitive reappraisal and expressive suppression. Emotion, 17(5), 772–777. doi: 10.1037/emo0000300 [DOI] [PubMed] [Google Scholar]
  30. Meule A, Reichenberger J, & Blechert J (2018). Development and Preliminary Validation of the Salzburg Emotional Eating Scale. Front Psychol, 9, 88. doi: 10.3389/fpsyg.2018.00088 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Reichenberger J, Kuppens P, Liedlgruber M, Wilhelm FH, Tiefengrabner M, Ginzinger S, & Blechert J (2018). No haste, more taste: An EMA study of the effects of stress, negative and positive emotions on eating behavior. Biol Psychol, 131, 54–62. doi: 10.1016/j.biopsycho.2016.09.002 [DOI] [PubMed] [Google Scholar]
  32. Rogosa D (1980). Comparing nonparallel regression lines. Psychological Bulletin, 88(2), 307. [Google Scholar]
  33. Salvy SJ, Feda DM, Epstein LH, & Roemmich JN (2017a). Friends and social contexts as unshared environments: a discordant sibling analysis of obesity- and health-related behaviors in young adolescents. Int J Obes (Lond), 41(4), 569–575. doi: 10.1038/ijo.2016.213 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Salvy SJ, Feda DM, Epstein LH, & Roemmich JN (2017b). The social context moderates the relationship between neighborhood safety and adolescents’ activities. Prev Med Rep, 6, 355–360. doi: 10.1016/j.pmedr.2017.04.009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Schachter S (1971). Some extraordinary facts about obese humans and rats. Am Psychol, 26(2), 129–144. doi: 10.1037/h0030817 [DOI] [PubMed] [Google Scholar]
  36. Schachter S, & Rodin J (1974). Obese humans and rats. Oxford, England: Lawrence Erlbaum. [Google Scholar]
  37. Schüz B, Schüz N, & Ferguson SG (2015). It’s the power of food: individual differences in food cue responsiveness and snacking in everyday life. Int J Behav Nutr Phys Act, 12, 149. doi: 10.1186/s12966-015-0312-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Smith KE, Mason TB, O’Connor SM, Wang S, Dzubur E, Crosby RD, … Roemmich JN (2020). Bi-Directional Associations Between Real-Time Affect and Physical Activity in Weight-Discordant Siblings. J Pediatr Psychol. doi: 10.1093/jpepsy/jsaa121 [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Snoek HM, van Strien T, Janssens JM, & Engels RC (2007). Emotional, external, restrained eating and overweight in Dutch adolescents. Scand J Psychol, 48(1), 23–32. doi: 10.1111/j.1467-9450.2006.00568.x [DOI] [PubMed] [Google Scholar]
  40. Steenhuis I, & Poelman M (2017). Portion Size: Latest Developments and Interventions. Curr Obes Rep, 6(1), 10–17. doi: 10.1007/s13679-017-0239-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Stone AA, & Brownell KD (1994). The stress-eating paradox: Multiple daily measurements in adult males and females. Psychology & Health, 9(6), 425–436. doi: 10.1080/08870449408407469 [DOI] [Google Scholar]
  42. Thompson SH, & Romeo S (2015). Gender and Racial Differences in Emotional Eating, Food Addiction Symptoms, and Body Weight Satisfaction among Undergraduates. J Diabetes Obes, 2(2), 93–98. doi: 10.15436/2376-0494.15.035 [DOI] [Google Scholar]
  43. Ufholz K, Salvy SJ, Feda DM, Epstein LH, & Roemmich JN (2019). Eating Responses to External Food Cues in Weight Discordant Siblings. J Adolesc Health, 65(1), 155–160. doi: 10.1016/j.jadohealth.2018.12.024 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. van Strien T, Frijters JER, Bergers GPA, & Defares PB (1986). The Dutch Eating Behavior Questionnaire (DEBQ) for assessment of restrained, emotional, and external eating behavior. International Journal of Eating Disorders, 5(2), 295–315. doi: [DOI] [Google Scholar]
  45. van Strien T, & Oosterveld P (2008). The children’s DEBQ for assessment of restrained, emotional, and external eating in 7- to 12-year-old children. Int J Eat Disord, 41(1), 72–81. doi: 10.1002/eat.20424 [DOI] [PubMed] [Google Scholar]
  46. Wardle J, Guthrie CA, Sanderson S, & Rapoport L (2001). Development of the Children’s Eating Behaviour Questionnaire. Journal of child psychology and psychiatry, and allied disciplines, 42(7), 963–970. doi: 10.1111/1469-7610.00792 [DOI] [PubMed] [Google Scholar]
  47. Wardle J, Marsland L, Sheikh Y, Quinn M, Fedoroff I, & Ogden J (1992). Eating style and eating behaviour in adolescents. Appetite, 18(3), 167–183. doi: 10.1016/0195-6663(92)90195-c [DOI] [PubMed] [Google Scholar]

RESOURCES