AN EXAMINATION OF BETWEEN- AND WITHIN-SUBJECT EFFECTS OF STRESS ON EMOTIONAL EATING OVER 49 CONSECUTIVE DAYS IN WOMEN By Natasha Fowler A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirement for the degree of Psychology- Doctor of Philosophy 2021 ABSTRACT AN EXAMINATION OF BETWEEN- AND WITHIN-SUBJECT EFFECTS OF STRESS ON EMOTIONAL EATING OVER 49 CONSECUTIVE DAYS IN WOMEN By Natasha Fowler Objective: Stress is associated with emotional eating (EE) in women cross-sectionally (between- subject). However, few studies have examined stress longitudinally limiting our understanding of how within-subject variations in stress level influence risk for EE over time and whether stress is in fact a risk factor or consequence of EE (within-subject). This study used an intensive, longitudinal study design to examine between- and within-subject effects of major life stress, daily stress impact, and cortisol on EE in women. Methods: An archival sample of 477 women aged 15-30 years recruited from the Michigan State University Twin Registry provided daily ratings of EE and stress impact for 49 consecutive days, along with self-reports of major life stress in the last 12 months and hair cortisol concentration (HCC), a longitudinal measure of cortisol secretion. Mixed linear models examined main and interactive effects of each stress variable on EE. Results: Both between- and within-subject analyses showed that daily stress more strongly predicted EE than major life stress. Specifically, women engaged in higher levels of EE when they experienced higher levels of daily stress impact relative to other women (between-subject) and their own daily stress levels (within-subject). There was a tendency for lower HCC to predict increased levels of EE (between-subject). Discussion: Findings confirm longitudinal associations between daily stress impact and cortisol with EE in women. Results also highlight the importance of within-subject shifts in a woman’s stress level in her risk for EE and suggest that stress management techniques may a be useful tool for treatment. TABLE OF CONTENTS LIST OF TABLES ………………….…………………………...…………………….……..…... v LIST OF FIGURES ……………….…………………………...…………………...…....….…... vi KEY TO ABBREVIATIONS …….……………………..…...…………………………...…...... vii INTRODUCTION .……………….………………………………...……………………….…… 1 METHODS ..………………….………………………...…………………………………….… 11 Participants ..…..……………………………....…………………………………….….. 11 Procedures ……....……………………………………………...………………………. 12 Measures …....…………………………………………………………..…………….… 13 Daily Measures ….…………...………………..………………….…………..… 13 Emotional eating (EE) …………………………………...…………....... 13 Daily stress …...…………………………………………………..….…. 14 Non-Daily Measures ..…...………………………………………………....….... 14 Major life stress ......................………………………………………….. 14 Hair cortisol concentration (HCC) …...………..………………………... 15 Hair care practice questionnaire …………………………….…….…….. 16 Covariates ………………………………………………………….…… 17 Statistical Analyses ………….......………………………………………………....…… 18 General Modeling Approach …………….……….…………………………….. 18 Primary Analyses ………………………………………………………….……. 19 Between-Subject Analyses …………………………………...………… 19 Within-Subject Analyses ………………………………………….……. 19 Exploratory Analyses …………………………………………………... 21 RESULTS ..………………………………………………………………………………….….. 22 Descriptive statistics ...………………………….…………………………………....…. 22 Between-subject analyses …………………….…..…………………………………….. 22 Pearson correlations ……………………….………………………………….… 22 Mixed linear models ………………………..……….………………………….. 23 Within-subject analyses ……………………………..………………………………….. 23 Pearson correlations ………….……………….……………………………….... 23 Mixed linear models ………………………….………………………..….……. 24 Exploratory analyses – HCC …………...…………….…………………………………. 24 Descriptive statistics ………………………….……………………………….... 24 Pearson correlations ………….……………….……………………………….... 25 Mixed linear models …………………………..……………..…………….…… 25 DISCUSSION ..……………………………………………………………………………….... 27 APPENDICES ...………………………………………………………………………….…….. 37 iii APPENDIX A: Primary Analyses ……………………………………………….…..…. 38 APPENDIX B: Exploratory Analyses ..………......………..………………..…….……. 48 APPENDIX C: Supplemental Analyses ………………….……………………….…….. 52 REFERENCES ...………………………………………………………….……………………. 65 iv LIST OF TABLES Table 1 Descriptive information for the full sample (N = 477) and HCC sample (N = 234). …..…39 Table 2 Results from the between-subject MLMs examining main and interactive effects of stress variables and covariates on average levels of emotional eating (N = 477). ………..................... 41 Table 3 Results from the within-subject MLMs examining main and interactive effects of the same-day and time-lagged stress variables and covariates on daily levels of emotional eating (N = 477). …………………………………………………………………………………………….. 42 Table 4 Results from the between-subject, exploratory MLMs examining the effects of hair cortisol concentration (HCC) and covariates on average levels of emotional eating. …..……………….. 51 Supplemental Table 1 Comparing descriptive information between participants who did and did not provide a hair sample. …………………………………………………………….……...…. 53 Supplemental Table 2 Results from the between-subject MLMs examining interactive effects of the daily stress variables with major life stress over the 49-day study period and across the lifetime on daily levels of emotional eating, including the covariates (N = 477). …..……........................ 55 Supplemental Table 3 Results from the within-subject MLMs examining interactive effects of the same-day and time-lagged daily stress variables with major life stress over the 49-day study period and across the lifetime on daily levels of emotional eating, including the covariates (N = 477). ………………………………………………………………………………………………..…. 57 Supplemental Table 4 Between-subject Pearson correlations for average daily stress, major life stress, emotional eating, and covariates (N = 477). ………………………….……………...……60 Supplemental Table 5 Within-subject Pearson correlations for same-day and time-lagged daily stress impact, daily emotional eating, average major life stress, and covariates (N = 477). ……. 61 Supplemental Table 6 Pearson correlations for hair cortisol concentration (HCC), emotional eating, and covariates in the full sample of HCC participants (N = 234) and subsample of HCC participants without confounding factors for HCC (N = 220). .…………………………………. 62 Supplemental Table 7 Results from the post-hoc within-subject MLMs examining predictive effects of emotional eating from one and two days ago on daily stress impact, including the covariates (N = 477). ……………………………………………………………………………. 64 v LIST OF FIGURES Figure 1. Between-subject Pearson correlations for average daily stress, major life stress, and emotional eating (N = 477). Note: ***Bonferroni corrected p<.002. ..………..……..…….....… 45 Figure 2. Two-way interaction between average stress impact and major life stress in the last 12 months. “High” and “Low” values represent 1 SD above and below the mean, respectively. .… 46 Figure 3. Within-subject Pearson correlations for same-day and time-lagged daily stress impact, daily emotional eating, and average major life stress (N = 477). ***Bonferroni corrected p<.002. …………………………………………………………………………………………………... 47 Figure 4. Pearson correlations for hair cortisol concentration (HCC), average stress impact, major life stress in the last 12 months, and average levels of emotional eating in the full sample of HCC participants (N = 234). Note: There were no significant correlations between HCC and either of the stress or eating variables. …………………………………………………...................…….. 49 Figure 5. Pearson correlations for hair cortisol concentration (HCC), average stress impact, major life stress in the last 12 months, and average levels of emotional eating in the subsample of HCC participants without confounding factors for HCC (N = 220). Note: There were no significant correlations between HCC and either of the stress or eating variables. ……………………..…. 50 vi KEY TO ABBREVIATIONS BE = binge eating BMI = body mass index EE = emotional eating vii INTRODUCTION Binge eating (BE) is characterized by repeated, intermittent bouts of overconsumption of food (typically highly palatable food that is high in sugar and/or fat) accompanied by a subjective endorsement of loss of control (American Psychiatric Association, 2013). BE is present in several eating disorders (e.g., anorexia nervosa-binge/purge [AN-BP], bulimia nervosa [BN], binge-eating disorder [BED], many forms of other specified feeding and eating disorders (OSFEDs)) and affects approximately 5% of Americans (Hudson, Hiripi, Pope, & Kessler, 2007), with much higher rates (2-10 times) in females versus males (American Psychiatric Association, 2013; Klump, Culbert, & Sisk, 2017). BE is thought to be influenced by changes in affective states (Hawkins & Clement, 1984). Supporting this hypothesis, emotional eating (EE) (i.e., overeating in response to negative emotions; Arnow, Kenardy, & Agras, 1995) is strongly associated with BE in both clinical (Masheb & Grilo, 2006; Ricca et al., 2009) and non-clinical (Stice, Presnell, & Spangler, 2002; Tanofsky-Kraff et al., 2007; van Strien, Engels, van Leeuwe, & Snoek, 2005) populations. EE has prospectively been shown to predict BE onset (Stice et al., 2002) and is positively associated with BE severity (Ricca et al., 2009). For these reasons, EE is considered a useful dimensional construct of BE behavior (Haedt-Matt et al., 2014). EE and BE are both comorbid with other psychiatric and medical conditions, including depression, anxiety, and obesity (Braden, Musher- Eizenman, Watford, & Emley, 2018; Hudson, Hiripi, Pope, & Kessler, 2007; Lazarevich, Camacho, Velazquez-Alva, & Zepeda, 2016; Ouwens, van Strien, & van Leeuwe, 2009; van Strien et al., 2016); however, the etiology of EE and BE remain poorly understood. Given the significant negative consequences associated with BE/EE, it is important to understand their development in order to better identify at-risk individuals and develop more tailored prevention and treatment options. 1 Stress has repeatedly been implicated in the etiology of EE and BE (Pike et al., 2006; Rojo, Conesa, Bermudez, & Livianos, 2006; Smyth et al., 2007; Wolff, Crosby, Roberts, & Wittrock, 2000). While the current study focused on associations between stress and EE, the majority of studies have examined associations between stress and BE. Given this, findings from studies of stress on both EE and BE are discussed. Stress can be defined in terms of stress frequency or stress response. Stress frequency refers to the amount of any real (e.g., abuse) or perceived (e.g., fear of judgement) threat to an individual’s general well-being that disrupts their homeostasis (i.e., steady state of optimal bodily functioning; Levine, 2005), while the stress response refers to either the psychological impact of (e.g., perceived impact of a stressor; typically assessed via self-report) or physiological reaction (e.g., cortisol response to stress) to a stressor. Oftentimes stress frequency and stress response are correlated, with higher frequencies of stress being associated with a greater degree of stress impact and stronger physiological response (Rab & Admon, 2020). While increased frequency and impact of major (e.g., death of a loved one, trauma) and acute, daily stressors (e.g., heavy traffic, argument with a coworker) are both associated with increased risk for EE/BE (Becker & Grilo, 2011; Degortes et al., 2014; Hay & Williams, 2013; Loth, van den Berg, Eisenberg, & Neumark-Sztainer, 2008; Pike et al., 2006), it is well accepted that the stress response has a greater influence on an individual’s overall health and functioning than stress frequency (Hay & Williams, 2013; McEwen & Akil, 2020; Rojo et al., 2006; Woods et al., 2010). The majority of studies examining associations between stress frequency, impact and EE/BE have examined between-subject effects. Between-subject studies typically compare major life and daily stress frequency, the psychological impact of a stressor or the physiological stress response between women who report high versus low levels of EE/BE. These between-subject 2 studies can be used to identify for whom stress increases risk for EE/BE. In other words, do women who exhibit higher levels of EE/BE experience more major life or daily stressors and/or exhibit greater psychological (e.g., stress impact) and physiological (e.g., cortisol) responses to stress than women with lower levels of EE/BE? Most of these studies have utilized retrospective self-reports (i.e., assessments of past stress exposure and stress impact over a certain period of time – 1 month ago, 1 year ago, over one’s lifetime) of major life and acute stress to determine how exposure to and impact of stressors are associated with EE and BE. In general, between-subject studies have shown that increased frequency of both major (e.g., death of a loved one, trauma) and acute, daily stress (e.g., traffic, argument with a coworker) is associated with increased risk for EE and BE behaviors (Becker & Grilo, 2011; Degortes et al., 2014; Hay & Williams, 2013; Loth et al., 2008; Pike et al., 2006). Additionally, increased perceived impact of major life (Micali et al., 2017; Rojo et al., 2006; Woods, Racine, & Klump, 2010) and acute, daily stressors (Crowther, Sanftner, Bonifazi, & Shepard, 2001; Diggins, Woods-Giscombe, & Waters, 2015; Jarvela-Reijonen et al., 2016; Kwan & Gordon, 2016; Richardson, Arsenault, Cates, & Muth, 2015; Smyth et al., 2007; Thurston, Hardin, Kamody, Herbozo, & Kaufman, 2018; Tomiyama, Dallman, & Epel, 2011) are associated with increased levels of BE (Hay & Williams, 2013; Kwan & Gordon, 2016; Thurston et al., 2018; Woods et al., 2010) and EE (Diggins et al., 2015; Jarvela-Reijonen et al., 2016; Richardson et al., 2015). It should be noted that while not all studies of major and acute, daily life stress examined both stress frequency and impact, results are consistent with the general stress literature in suggesting a stronger role for stress impact versus stress frequency in risk for EE/BE (Hay & Williams, 2013; Rojo et al., 2006; Woods et al., 2010). 3 Studies examining between-subject physiological indices of stress (e.g., basal cortisol levels, cortisol reactivity) have been more mixed in their findings. These studies are predicated on two well established findings in the general stress literature. Firstly, increased exposure to, and impact of, major life and acute, daily stressors alter physiological indices of stress and can cause increased basal cortisol levels (Godoy, Rossignoli, Delfino-Pereira, Garcia-Cairasco, & de Lima Umeoka, 2018) and increased risk for various psychological conditions (e.g., depression; de Kloet, Joëls, & Holsboer, 2005). Secondly, responses to stress exhibit an inverted U-shaped relationship with stress severity (Sapolsky, 2015). For example, at low or chronically high levels of stress frequency or impact, physiological responses to acute stress (e.g., cortisol reactivity) are low or blunted compared to moderate levels of stress frequency or impact. Because women who EE/BE tend to have experienced more major and acute, daily stressors (see above), it is possible that elevated basal cortisol levels and a blunted cortisol response to acute stress are associated with increased risk for EE and BE. Studies examining these hypotheses have produced mixed results. For example, despite elevated cortisol levels being associated with increased hedonic value (Adam & Epel, 2007) and consumption of palatable food (Dallman et al., 2003; Gluck, 2006; Godfrey et al., 2019; la Fleur, Akana, Manalo, & Dallman, 2004; Laugero, Falcon, & Tucker, 2011; Pecoraro, Reyes, Gomez, Bhargava, & Dallman, 2004), there have been inconsistent associations reported for cortisol levels and BE in women or female animals. Three studies found that obese (Gluck, Geliebter, Hung, & Yahav, 2004; Gluck, Geliebter, & Lorence, 2004a) and non-obese (Koo-Loeb, Costello, Light, & Girdler, 2000) women who BE exhibit significantly higher levels of 24 hour urinary cortisol (i.e., urinary cortisol collected every hour for 24 hours; Koo-Loeb et al., 2000) and serum basal cortisol levels (Gluck, Geliebter, Hung, et al., 2004; Gluck, Geliebter, & Lorence, 4 2004a) than women who do not BE. In contrast, two studies found reduced 24 hour urinary (Lavagnino et al., 2014) and salivary cortisol levels (collected 3 times per day for two consecutive days; Larsen, van Ramshorst, van Doornen, & Geenen, 2009) in obese women with BED compared to obese women without BED. Lastly, two studies found no association between salivary cortisol and BE in obese women with BED compared to obese women without BED (Coutinho, Moreira, Spagnol, & Appolinario, 2007; Schulz, Laessle, & Hellhammer, 2011). Although it is difficult to identify the factors that contributed to these mixed associations, it is possible that differences in age, type of tissue in which cortisol was assessed, or time of day time of cortisol collection may have contributed. The majority of studies that found positive associations between cortisol and BE examined women in young adulthood (Gluck, Geliebter, Hung, et al., 2004; Gluck, Geliebter, & Lorence, 2004a; Koo-Loeb et al., 2000), while studies that found negative or no association between cortisol and BE examined women in middle adulthood (Coutinho et al., 2007; Larsen et al., 2009; Lavagnino et al., 2014; Schulz et al., 2011). While cortisol levels are known to increase from birth into the first six months of life, decrease during early childhood, and increase again during late childhood and adolescence (Kamin & Kertes, 2017), it is uncertain how cortisol levels vary across young to middle adulthood. This can be a fruitful avenue for future research that may help to clarify these mixed associations. Additionally, the two studies that found no association between cortisol and BE used salivary cortisol (Coutinho et al., 2007; Schulz et al., 2011) while the two studies that examined serum cortisol found positive associations between cortisol and BE in women (Gluck, Geliebter, Hung, et al., 2004; Gluck, Geliebter, & Lorence, 2004a). Because cortisol is not uniformly secreted throughout the day, many studies recommend using salivary cortisol, rather than serum or urinary assays, to assess cortisol levels (Manetti et al., 2013; Odeniyi & Fasanmde, 2013; 5 Putignano et al., 2003; Viardot et al., 2005). However, salivary cortisol is subject to diurnal fluctuations (Parikh et al., 2018); thus, salivary cortisol only represents cortisol levels in that moment, as opposed to overall. Thus, because the two studies that found positive associations (Gluck, Geliebter, Hung, et al., 2004; Gluck, Geliebter, & Lorence, 2004a) examined cortisol at noon while the study that found no association (Schulz et al., 2011) examined cortisol in the late afternoon, diurnal fluctuations in cortisol may have contributed to the mixed associations between cortisol and BE. Clearly, additional studies are needed to clarify these methodological differences and current results; but at present, it appears that there may be positive associations between serum cortisol levels and BE in women assessed in young adulthood, but associations may vary in women assessed at other ages or using other collection (e.g., salivary) methods. Far fewer studies of stress and BE have examined these associations on a within-subject level. Within-subject studies examine stress at an individual level to determine when and how variations in a woman’s stress level influences her risk for EE/BE. These studies commonly use longitudinal measures that include daily diaries or ecological momentary assessment data that assess stress and BE multiple times per day to examine if and when stress is a risk factor for (rather than a consequence of) BE. These studies typically use “within-person” centered data that index the extent to which a women’s stress levels on a particular day are higher or lower than her average stress levels (across several days) to determine how variations from baseline stress predict variations in BE. The few within-subject studies that have been conducted suggest that stress is a risk factor for, and not correlate of, BE and that daily stress impact may more strongly influence BE than daily stress frequency. For example, both increased levels of daily stress frequency and impact have been found to precede BN behaviors (e.g., BE) in women (Goldschmidt et al., 6 2014). While there appears to be no difference in the frequency of daily stressors (e.g., traffic, argument with a coworker) on BE versus non-BE days (Wolff et al., 2000), two studies found that women perceived daily stressors as more impactful on days when they engaged in BE (Smyth et al., 2007; Wolff et al., 2000). Only one study has examined the influence of stress on one day with BE on subsequent days; and found that daily stress impact is more strongly associated with same-day, as opposed to subsequent-day BE (Freeman & Gil, 2004). To the author’s knowledge, only one within-subject study has examined associations between cortisol and BE and found that individuals with BED demonstrated a greater cortisol response to acute- stress in the morning compared to the afternoon (Carnell et al., 2018). This suggests that associations between physiological indices of stress and BE may vary by time of day. Clearly, additional studies of within-subject effects of stress frequency, impact and physiological responses are needed. While the current findings suggest that stress may precede EE/BE and is more strongly associated with same-day EE/BE (Freeman & Gil, 2004), BE is typically followed by feelings of guilt, shame, or disgust over the binge episode (American Psychiatric Association, 2013). Therefore, it is possible that EE/BE may also contribute to stress about the EE/BE behaviors that then increases negative feelings (e.g., guilt, shame) and reinforces EE/EE as a way to cope with the stress and negative feelings. The cross-sectional nature of the current study did not allow for examination of this effect; however, additional within-subject studies of stress and EE/BE can help delineate stress as a risk factor for versus consequence of EE/BE. These studies can also clarify when deviations in stress levels most strongly associate with EE/BE (e.g., same-day versus subsequent days). However, other gaps in the literature also remain. Studies of physiological responses to stress have produced mixed findings, potentially due to differences in participant age, tissue in 7 which cortisol was assessed, or time of day of cortisol collection. Additionally, because reduced hours of sleep per night (Nollet, Wisden, & Franks, 2020) has been associated with increased levels of cortisol, mixed associations between cortisol and EE/BE may also be due to sleep deprivation. Additional studies, potentially using more novel measures of cortisol secretion (e.g., hair cortisol concentrations (HCC)), that control for participant age and that include measures of sleep patterns are needed. In addition, despite between- and within-subject studies examining main effects of major life and daily stress frequency and impact, few studies have examined interactions between major life stress and daily stress frequency, or daily stress impact. Evidence suggests that between-subject effects of daily stress on BE are heightened in women who experience high levels of major life stress impact (Woods et al., 2010), indicating a potential sensitization to negative effects of acute, daily stress following high major life stress exposure. Thus, examining potential interactions between major life and daily stress frequency and impact is critically important for future work. Given the above, the proposed study extends prior research on stress and EE/BE by examining main and interactive effects of major life stress, daily stress frequency, and daily stress impact on EE in an archival sample of women assessed across 49 consecutive days. Assessments of major life stress (e.g., death of a loved one) were previously collected over 49 days as well as more distal time periods (e.g., last 12 months), and daily ratings of “hassles” (e.g., traffic, argument with a coworker) were also collected. This study capitalized on the longitudinal, daily study design and conduct both between- and within-subject analyses to determine for whom stress increases risk for EE (between-subject effects), distinguish between stress as a predictor versus consequence of EE (within-subject effects), and elucidate when deviations in a woman’s stress levels are most strongly associated with deviations in her EE 8 (same day or subsequent days) (within-subject effects). Between-subject analyses were expected to show that increased levels of major life stress, average stress frequency and impact would predict increased average levels of EE in women during the study period. In within-subject analyses, main and interactive effects of increased levels of daily stress frequency and impact were expected to prospectively predict increased levels of daily EE episodes (i.e., stress would be a risk factor for EE. Within-subject prospective effects of daily stress were expected to be stronger for same-day versus subsequent day stress-EE, as well as in women with high versus low levels of major life stress. Both between- and within-subject analyses are expected to show significant interactions between stress frequency and psychological impact, such that stress frequency is associated with increased EE only when levels of stress impact are also high. In addition to examining stress frequency and impact, this study explored whether a novel measure of cortisol response (i.e., HCC) was associated with EE. Cortisol is thought to incorporate into hair via passive diffusion from the bloodstream (Stalder & Kirschbaum, 2012). Because hair grows approximately 1 cm per month (Wennig, 2000) and cortisol can remain stable in hair for up to 6 months (Kirschbaum, Tietze, Skoluda, & Dettenborn, 2009; Noppe et al., 2014), HCC provides a retrospective measure of cortisol levels over an extended time period (Stalder & Kirschbaum, 2012). In the current study, a subsample of women provided a 1.5 cm HCC sample at the end of the study that indexed cortisol secretion over the 49-days. Thus, an additional exploratory aim of this study examined between-subject associations between HCC and EE. No previous study has examined associations between HCC and BE or EE; consequently, these analyses were exploratory, but consistent with some past studies of cortisol levels and palatable food intake and BE (Adam & Epel, 2007; Dallman et al., 2003; Gluck, 2006; Gluck, Geliebter, Hung, et al., 2004; Gluck, Geliebter, & Lorence, 2004b; Koo-Loeb et al., 2000; 9 Laugero et al., 2011; Pecoraro, Reyes, Gomez, Bhargava, & Dallman, 2004) in women in young adulthood, it was expected that higher HCC would be associated with increased average EE. 10 METHODS Participants Participants included 477 female twins (aged 15-30 years old; M = 21.8 years; SD = 3.0) from the Michigan State University Twin Registry (MSUTR; see Burt & Klump, 2013, 2019; Klump & Burt, 2006 for MSUTR description). The MSUTR is a population-based twin registry that recruits twins in collaboration with the Michigan Department of Health and Human Services (for additional information about study recruitment, see Burt & Klump, 2013, 2019; Klump & Burt, 2006) via the use of birth records. Participants for the current project were recruited from an on-going study within the MSUTR (i.e., A Twin Study of Exogenous Hormone Exposure and Binge Eating; EHE-BE) that examines effects of combined oral contraceptives (COC) on disordered eating. All participants were required to meet the following criteria: 1) at least one member of the twin pair must have been taking COC for at least 3 months prior to starting the study (82.7% of participants are taking COCs, 17.3% of participants are not taking COCs); 2) for participants who are not taking COCs, menstrual cycles must be regular (between 22-32 days); 3) no psychotropic or steroid medications within past 4 weeks; 4) no pregnancy/lactation within the past 6 months; and 5) no genetic or medical conditions known to influence hormones or appetite/weight. It is important to note that while the stress measures (described below) for this study were administered to all participants, the collection of a hair sample for HCC analysis was an optional procedure for which participants received extra compensation. In order to participate in the optional procedure, EHE-BE participants had to have hair longer than 1 inch that was free of chemical treatments to the hair roots (e.g., dying, bleaching, chemical straightening). Because of these additional criteria, the sample for the exploratory analyses (n = 234, 49% of the total 11 sample of 477 participants) is smaller than the sample for the primary analyses (n = 477). Compared to previous MSUTR studies (Burt & Klump, 2013, 2019; Klump & Burt, 2006), participants in the full sample (96.2% Non-Hispanic/Latinx, 89.3% White, 5.0% Black, 1.3% Asian, and 4.4% Multiracial) and participants who opted into the HCC study (97.4% Non- Hispanic/Latinx, 91.1% White, 3.4% Black, 0.4% Asian or Pacific Islander, and 5.5% Multiracial) had a higher percentage of participants who identified as non-Hispanic/Latinx and White. Participants who provided a hair sample were also on average, 1 year older and experienced higher levels of major life stress across their lifetime than participants who did not provide a HCC sample (p<.001; Supplemental Table 1). The two groups of participants otherwise did not differ significantly from one another on the other variables (i.e., EE, stress levels) or other key demographic variables (e.g., race/ethnicity; all p’s>.05; Supplemental Table 1). Procedures All measures and procedures were approved by the Michigan State University Institutional Review Board. Participants provided behavioral data for 49 consecutive days. Questionnaires were completed each evening (after 5:00 pm) using an online data system or pre- printed scantrons. Additionally, all participants completed three in-person assessments occurring at the beginning of the study, halfway through the study (~ day 25), and at the end of the study (~ day 49). During these in-person assessments, each participant’s eligibility was reassessed, height and weight were measured, and completed materials were collected. Hair samples for HCC were collected during the last study visit to ensure that the study captured cortisol concentrations over the 49-day study period. Between visits, staff contacted participants 1x/week to confirm continued protocol adherence and answer questions. 12 Measures Daily Measures Emotional eating (EE) EE was assessed daily using the Emotional Eating scale of the Dutch Eating Behavior Questionnaire (DEBQ; van Strien, Frijters, Bergers, & Defares, 1986). The Emotional Eating scale assesses eating in response to negative emotions (example item: “Did you have a desire to eat when you were depressed?”); responses were made using scales from 1 (not at all) to 5 (very often). Internal consistencies for the DEBQ Emotional Eating scale were excellent in previous research (α = .93; Klump, Keel, Culbert, & Edler, 2008; Racine et al., 2012; van Strien, Frijters, Bergers, & Defares, 1986) and in the current sample (45-day average α = .90). It is important to note that eating in response to negative emotions is thought to be a core feature of BE, and the DEBQ Emotional Eating scale has demonstrated validity in differentiating among individuals with clinical and subclinical objective BE episodes (i.e., binge eating episodes characterized by an overconsumption of highly palatable food in a short period of time, accompanied by a sense of loss of control over eating; American Psychiatric Association, 2013), overweight individuals, and college students (Wardle, 1987). Furthermore, the DEBQ Emotional Eating scale is significantly and positively correlated with established measures of BE (r’s = .55–.69; Racine, Culbert, Larson, & Klump, 2009; van Strien et al., 1986) as well as with palatable food intake (i.e., ice cream) in a laboratory setting (van Strien, 2000). Similar to previous research (Klump et al., 2008), the instructions for the DEBQ emotional eating scale were modified with permission to ask about EE over the current day. 13 Daily stress The Daily Stress Inventory (Brantley, Waggoner, Jones, & Rappaport, 1987) was used to assess daily stressors over the study period. The DSI is a 60-item, self-report questionnaire that asks participants to report whether or not they have experienced daily stressors or daily hassles over the course of the day. Items include stressors such as ‘traffic difficulties’, ‘an argument with another person’, or ‘experienced bad weather’. Participants rated all events as either present or absent. For all stressors that were present, participants rated the impact of those stressors on a 7- point Likert scale that ranges from 1 = ‘occurred, but was not stressful’ to 7 = ‘caused me to panic’. The total number of events that occurred each day was summed to create a daily stress frequency score. Similarly, the total impact for events that occurred each day was summed to create a daily impact score. The DSI has demonstrated excellent construct validity in prior research, as it is highly correlated with the frequency and intensity scales of other stress measures (e.g., the Hassles Scale; Brantley et al., 1987; Kanner, Coyne, Schaefer, & Lazarus, 1981) as well as positive associations and high convergent validity with endocrine measures of stress (e.g., urinary cortisol levels; Brantley, Dietz, McKnight, Jones, & Tulley, 1988), and good internal consistency. Non-Daily Measures Major life stress The Social Readjustment Rating Schedule (SRRS; Holmes & Rahe, 1967) was used to retrospectively assess major life stress. The SRRS is a 43-item questionnaire that asks participants whether or not they have experienced any of the 43 commonly reported stressful events (e.g., death of a loved one, illness) over the course of the past year. Each item/event that occurred is then assigned a life change unit (LCU) score (ranging from 11-100) that indicates the 14 overall severity of the event in terms of the relative degree of adjustment necessary following exposure to each event. The item LCUs are then summed to create a total LCU score across all events. It should be noted that the LCU score for each item combines frequency and impact scores into one value, thus there are no independent LCUs or LCU total scores for frequency versus impact. A psychometric study of the SRRS has demonstrated acceptable internal consistency (α= .72) and a very high correlation (r = .97) between the SRSS and the Schedule of Recent Events, another measure of major life stress (Lei & Skinner, 1980). Similar to the original SRRS, participants were asked to rate the SRRS items for the past 12 months. However, in order to assess for major life stress over the same time period as the daily stress measures and HCC, participants also completed the SRRS items for stressors that occurred over the 49-day study period. In addition, because women who do and do not EE/BE tend to differ in their stress exposure and impact across their lifetime, participants were also asked to complete the SRRS items for events occurring any time over the course of their life. Hair cortisol concentration (HCC) Following standard procedures (Wright et al., 2018), hair samples were obtained by cutting the hair from the posterior vertex of the scalp, as close to the scalp as possible. Because each 1 cm of hair from the scalp corresponds to cortisol secretion over the past month (Stalder & Kirschbaum, 2012), hair samples were collected during the final assessment and the first 1.5 cm of hair most proximal to the scalp were assayed for the current study to provide a retrospective index of cortisol secretion over the 49-day study period. Hair samples were wrapped in aluminum foil for protection and stored at room temperature, as previously described (Wennig, 2000), until they were shipped to the Behavioral Immunology and Endocrinology Laboratory at the University of Colorado, Denver at the Anschutz Medical Campus for analysis. Following the 15 procedures outlined in Hoffman, D’Anna-Hernandez, Benitez, Ross, & Laudenslager (2017), upon arrival at the lab, hair was ground and cortisol levels were measured using a commercial high sensitivity EIA kit (Salimetrics, LLC, State College, PA) that was conducted according to manufacturer’s instructions as previously described in (D’Anna-Hernandez, Ross, Natvig, & Laudenslager, 2011). To calculate inter-assay coefficients of variation (CV), a pooled control of previously ground hair was extracted using the same procedures outlined above and included on each EIA plate in duplicate. Inter-assay CV for the control hair pool was 9.2% for the high hair control and 11.2% for the low hair control and intra-assay CV was 1.4%. A growing number of studies have used HCC as a measure of cumulative cortisol secretion (see reviews Sander et al., 2020; Stalder et al., 2017; Stalder & Kirschbaum, 2012) and found high test-retest associations between repeated HCC assessments (r’s between 0.68-0.79; Stalder & Kirschbaum, 2012), positive associations between HCC and 30-day average salivary cortisol levels (r’s = 0.61, p = .01; Short et al., 2016) and levels of major life stress (β = 0.21, p = .04 for stressors such as death of a close relative, serious illness, divorce; Karlén, Ludvigsson, Frostell, Theodorsson, & Faresjö, 2011) assessed over the same time period. These data support the validity of HCC as a measure of cumulative cortisol concentration over the study period. Hair care practices questionnaire During the final assessment, participants completed a brief questionnaire about key hair care practices (e.g., how often they wash, color, bleach, and/or chemically straightening their hair, whether they use any scalp medication, etc.) that could influence the reliability/validity of the HCC. 16 Covariates Because of the wide age range of my sample, and the fact that negative affect, body mass index (BMI), total hours of sleep per night, and income are associated with self-reported stress (Goldschmidt et al., 2014; Langer et al., 2018; Spinosa, Christiansen, Dickson, Lorenzetti, & Hardman, 2019; Tenk et al., 2018), cortisol levels (Cohen, Doyle, & Baum, n.d.; Faresjö et al., 2013; Manenschiin, van Kruysbergen, de Jong, Koper, & van Rossum, 2011; Nollet et al., 2020; Stalder & Kirschbaum, 2012; Wester et al., 2014), and BE and EE (Goldschmidt et al., 2014; Langer et al., 2018; Smyth et al., 2007; Spinosa et al., 2019), these variables were included as covariates in the between-subject analyses only. They were not included in the within-subject models, as these models examined how variability in a woman’s own stress levels influence variation in her own levels of EE. Daily ratings of negative affect were assessed via the Negative Affect Scale from the Positive and Negative Affect Schedule (Watson, Clark, & Tellegen, 1988). This scale consists of 10 items measuring the full range of daily negative emotions (e.g., distress, nervousness, irritability, fear). Participants rated the degree to which each emotion was experienced; responses were made using Likert scales from 1 (very slightly/not at all) to 5 (extremely). Height and weight were measured using a wall-mounted ruler and digital scale during the three in-person visits (i.e., beginning, intermediate, end of study) in order to calculate BMI (i.e., weight (kilograms)/height (meters)2). Because prior work from our lab has shown that changes in weight across a 45-day period are minimal (e.g., M = -0.20 lb change, SD = 3.39; Klump et al., 2013), the average BMI across the three study visits was calculated and used in analyses. Income was assessed via the question ‘What is the approximate average income of your parents’ with response options as ‘under $20,000’, ‘$20,000-$40,000’, ‘$40,000-$60,000’, 17 ‘$60,000-$100,000’, and ‘over $100,000’. Lastly, because sleep (Nollet et al., 2020) is associated with cortisol levels, total hours of sleep per night was included as a covariate in the exploratory HCC analyses. Total hours of sleep per night was assessed via the question ‘How many hours of sleep did you get last night?’ with response options ranging from ‘0-4 hours’ to ‘more than 13 hours’. Statistical Analyses General Modeling Approach Originally, I planned to examine between- and within-subject main and interactive effects of daily stress frequency, daily stress impact, and major life stress, including covariates on EE. However, average levels of daily stress frequency and impact were very strongly correlated (r = 0.93, p<.002), suggesting that these variables assess the same or very similar constructs in this study. Therefore, to reduce multicollinearity and because prior studies suggest that stress impact more strongly influences EE as compared to stress frequency (Hay & Williams, 2013; McEwen & Akil, 2020; Rojo et al., 2006; Woods et al., 2010), only daily stress impact was included in all between- and within-subject analyses. Prior to analyses, daily stress impact, major life stress, HCC, BMI, negative affect, and EE were log transformed prior to analyses in order to account for positive skew. Mixed linear models (MLMs) were used to examine both between- and within-subjects associations, as MLMs can control for the non-independence of the twin data as well as the repeated measures that are examined in analyses of within-subject/daily effects. Primary analyses of the major life stress variables focused on major life stress in the last 12 months to maximize variability in scores, but secondary analyses explored associations between major life stress over the 49-day study period and across the lifetime with EE. Because results were nearly identical for the primary versus 18 secondary analyses of major life stress scores, only results from the primary analyses are discussed below. Results from the secondary analyses are included in the supplemental material only (see Supplemental Tables 2 & 3). To control for multiple comparisons, a p-value of .01 and the more conservative Bonferroni corrected p-value = .002 were used. Primary Analyses Between-Subject Analyses Between-subject analyses examined whether women who reported higher levels of daily stress impact and major life stress) also reported higher levels of EE compared to women with lower stress levels. For these analyses, all daily measures (i.e., daily levels of EE, stress impact, negative affect, age) were averaged across the 49-day study period in order to obtain a mean value that indexed overall levels of each variable across the study period. Variables that were assessed at only a single-time point (i.e., major life stress, income) or were already averages (i.e., BMI) did not require this averaging, but instead were included as the total score/value. All predictor variables were standardized prior to analysis in order to reduce multi-collinearity between the main and interaction effect variables and to create a common unit of measure for each variable. Pearson correlations were calculated first to provide initial indications of associations between the stress scores and EE. Two-level, MLMs were used to test the main as well as interactive effects of each stress variable (e.g., major life stress x daily stress impact) on EE, with participants (level 1) nested within twin pairs (level 2). Within-Subject Analyses Within-subject analyses examined whether variations in a woman’s reported daily stress frequency and impact were associated with her levels of same-day and subsequent-day EE. 19 These analyses focused on the daily variables (i.e., daily stress impact, negative affect, EE) in order to examine within-subject effects within and across days. For these analyses, all daily variables were within-subject centered (i.e., each participant’s daily value of each variable was subtracted from each participant’s average daily score for each variable) and then standardized to determine how daily variation in each woman’s stress levels relative to her average levels influenced daily changes in EE. Daily stress impact from 1 day ago was calculated by lagging the within-subject centered and standardized same-day daily stress variable and daily stress impact from 2 days ago was calculated by lagging the daily stress from 1 day ago variable. Because major life stress is not a daily variable, it was not within-subject centered and standardized; instead major life stress was between-subject standardized prior to being included in analyses. Pearson correlations provided initial indications of associations between same-day and time-lagged daily stress impact and daily EE. Three-level MLM models examined how changes in daily stress frequency, daily stress impact, and covariates associated with changes in same-day and subsequent-day EE. Observations (level 1) were nested within participants (level 2), and participants were nested within twin pairs (level 3). The first set of MLMs examined predictive effects of daily stress impact including covariates on same-day EE. Then, because chronic exposure to major life stress can sensitize women to the negative effects of daily stressors (see Introduction), a second series of models examined the 2-way interaction between the between-subject major life stress variables and the within-subject daily stress impact variable, controlling for the covariates. A final set of MLMs was then conducted to examine predictive effects of daily stress impact from 1 and 2 days ago on EE. Models were conducted the same as described above (i.e., 20 separate models for main and interactive effects) except that the same-day stress variables were also included in the time-lagged models with the daily stress variables from 1 and 2 days ago. This ensured that that any predictive effects of the time-lagged stress variables were above and beyond those of the same-day stress variables. Exploratory Analyses Exploratory analyses examined HCC and its association with between-subject risk for EE in the subsample of women who provided a hair sample. Because HCC is a cumulative measure of cortisol over the 49-day study period daily within-subject effects of cortisol levels on EE could not be examined. Analyses were identical to the between-subject models described above, including data preparation methods (e.g., using average EE values), the Pearson correlations examining initial associations between HCC and EE, and the MLMs examining main effects of HCC on EE. Again, because total hours of sleep has been shown to influence cortisol levels, this variable was included as a covariate, along with BMI, negative affect, age, and income. To confirm that associations between HCC and EE were unaffected by participants’ hair care practices, all analyses were repeated in the 220 women (49% of the total sample) who did not exhibit any potentially confounding factors for HCC (e.g., chemical treatments). 21 RESULTS Descriptive statistics Descriptive statistics for the full sample are presented in Table 1. A wide range of EE was represented (average EE score range = 0-2.58 out of a possible range of 0-4), indicating good variability in dysregulated eating symptoms. Participants also varied considerably on indices of daily stress impact and major life stress. Average levels of daily stress impact ranged from 0.92-253.02, out of a possible 0-399. Average levels of major life stress in the last 12 months ranged from 0-668 out of a possible 0-2246. The mean and range of these scores are consistent with findings from previous population-based studies of EE (e.g., Hildebrandt et al., 2015; Klump, Keel, Burt, et al., 2013), major life stress (e.g., Woods et al., 2010), and daily stress impact (e.g., Brantley et al., 1987; Wolff et al., 2000; Woods et al., 2010) in young adults. Between-subject analyses Pearson correlations As shown in Figure 1 and Supplemental Table 4, average levels of EE were strongly correlated with average levels of stress impact (r = 0.56, p<.002). However, average levels of EE were not significantly correlated with levels of major life stress in the last 12 months (r = 0.07, p = .16), suggesting that EE is more strongly associated with measures of daily stress impact compared to major life stress. Average levels of EE and daily stress impact were also strongly associated with average levels of negative affect (r’s = 0.67, p<.002), reflecting the affective nature of EE and stress. Minimal associations were found between average EE levels and BMI (r = 0.05, p .35), age (r = -0.10, p = .02), and income (r = 0.10, p = .82). 22 Mixed linear models Results from the MLMs corroborated the above correlations between daily stress impact and EE (Table 2) and showed that even after controlling for negative affect, BMI, age, and income, women who experienced higher levels of average daily stress impact over the 49-day study period engaged in higher levels of average EE than women with lower levels of average daily stress impact (β = 0.35, p<.001). While major life stress in the last 12 months did not significantly predict average levels of EE (β = -0.02, p>.05), there was a trend-level two-way interaction between average stress impact and major life stress in the last 12 months (β = 0.11, p = .03). Specifically, women who experienced high levels of major life stress in the last 12 months (defined as 1 SD above the mean) engaged in significantly higher levels of EE when they also experienced high versus low levels of average stress impact (Figure 2). In contrast, women who experienced low levels of major life stress in the last 12 months (defined as 1 SD below the mean) showed no change in average levels of EE regardless of average stress impact level (Figure 2). Within-subject analyses Pearson correlations Correlations between the within-subject variables were similar to the correlations between the between-subject variables. As shown in Figure 3 and Supplemental Table 5, levels of daily EE were moderately correlated with levels of same-day stress impact (r = 0.22, p <.002). Daily EE was less strongly correlated with time-lagged measures of daily stress impact (r’s = 0.06-0.06, p’s<.002). Similar to the between-subject correlations, daily levels of EE and stress impact were significantly associated with daily levels of negative affect (r’s = 0.18-0.46, p’s<.002). 23 Mixed linear models As shown in Table 3, increases in same-day stress impact predicted increased levels of daily EE above and beyond the effects of the covariates. Specifically, women were more likely to engage in higher levels of daily EE when they experienced higher than their average levels of daily stress impact (β = 0.15, p<.001). While shifts in a woman’s daily stress impact significantly predicted shifts in her levels of EE two days later (β = 0.03, p<.001), shifts in daily stress impact more strongly predicted shifts in EE when stress impact and EE occurred on the same day (see Table 3). Counter to hypotheses, women with higher levels of major life stress in the last 12 months did not increase their daily EE levels when they experienced higher levels of daily stress impact relative to their average daily stress impact levels, either when daily stress impact occurred on the same-day or on prior days (β’s <0.01, all p’s>.05; see Tables 3). Exploratory analyses - HCC Descriptive statistics Similar to the full sample of women, the subsample of women who provided a hair sample demonstrated a wide range of EE (average EE score range = 0-2.48, out of a possible 0- 4) and considerable variability on indices of daily stress impact and major life stress (Table 1). While there is currently no standard range for HCC, participants did appear to exhibit ample variable in HCC over the study period (mean = 10.71 pg/mg, SD = 18.86 pg/mg; range = 1.82- 191.2 pg/mg). This mean and range are consistent with HCC findings from other population- based samples in young adults and adults (e.g., Cieszyński, Jendrzejewski, Wiśniewski, Owczarzak, & Sworczak, 2019; Ferro & Gonzalez, 2020; García-León, Pérez-Mármol, Gonzalez-Pérez, García-Ríos, & Peralta-Ramírez, 2019; O’Brien, Meyer, Tronick, & Moore, 2017). 24 Pearson correlations HCC was not significantly associated with average levels of EE in the full sample of participants who provided a hair sample (r = -0.13, p = .05; Figure 4 and Supplemental Table 6) or in the subsample of women who did not exhibit any potentially confounding factors for HCC (e.g., chemical treatments) (r = -0.15, p = .03; Figure 5). Additionally, and unexpectedly, HCC was not significantly associated with average levels of daily stress impact collected over the same time period (r = 0.01-0.02, p’s>.05; Figure 4 & 5) in either sample. Mixed linear models While HCC did not significantly predict average levels of EE in either the full sample of women that provided a hair sample (β = -0.14, p = .007) or the subsample of women without confounding factors for HCC (e.g., no chemical hair treatments; β = -0.16, p = .003; Table 4) when using the more conservative Bonferroni corrected p-value = .002, results were significant at p<.01. These results suggest a tendency for lower HCC levels to predict increased levels of average EE after controlling for the covariates. Notably, these results are in contrast to the correlations described above. Because HCC can be influenced by a participant’s BMI, age, average hours of sleep, etc., discrepancies between HCC’s association with EE in the Pearson correlations versus the MLMs may be due to confounding effects of the covariates. To examine this possibility further, post-hoc partial correlations were run to examine associations between HCC and EE, controlling for the covariates. Indeed, controlling for the covariates did strengthen the association between HCC and EE in both the full sample of women (r = -0.21, p = .002) and in the subsample of women who did not exhibit any potentially confounding factors for HCC (r = -0.23, p = .001). Importantly results from the partial correlations mimic the results found in the 25 MLMs. Specifically, they show there is a tendency for low HCC to predict higher levels of EE in both the full sample (p<.01) and subsample of HCC women (p<.01). 26 DISCUSSION This is the first longitudinal study to examine between- and within-subject effects of daily stress impact, major life stress, and HCC on EE in women. My findings indicate that daily stress is a more robust predictor of EE in women than major life stress. Specifically, women were more likely to report higher levels of EE when they experienced higher than average levels of daily stress relative to other women (between-subject effects) and their own daily stress levels (within- subject effects). Additionally, shifts in daily stress impact more strongly predicted same-day EE as compared to subsequent-day EE. Lastly, there was a tendency for women with lower HCC to report higher levels of EE (between-subject effects). Overall, the current study extends prior findings of stress and EE in women by distinguishing the relative influence of daily stress impact, major life stress, and cortisol on EE, identifying for whom stress increases risk for EE, and when effects of stress on EE are strongest. Stronger predictive effects of daily stress impact compared to major life stress on EE suggest that daily stress plays a stronger role in triggering EE episodes than major life stress. This may seem surprising given the inherently more severe nature of major life stressors (e.g., death of a loved one, trauma) compared to daily stressors (e.g., traffic, negative interactions with a coworker). It is important to note, however, that major life stress may be more important for precipitating the initial onset of disordered eating behavior rather than predicting on-going EE/BE. Women are 6x more likely to develop disordered eating behaviors if they experience chronically high levels of major life stress compared to women who experience low levels of major life stress (Pike et al., 2006). Compared to women without an eating disorder, women who eventually developed an eating disorder experienced significantly higher levels of major life stress in the year preceding their disorder (Rojo et al., 2006). Of particular salience are chronic 27 and/or severe stressors, such as trauma (e.g., sexual, physical, and/or emotional abuse) (Backholm, Isomaa, & Birgegård, 2013; Palmisano, Innamorati, & Vanderlinden, 2016; Smyth, Heron, Wonderlich, Crosby, & Thompson, 2008; Zelkowtiz, Zerubavel, Zucker, & Copeland, 2021). Because I did not assess trauma experience in my participants, I was unable to compare associations between trauma versus daily stress and cortisol on EE. However, given the role of trauma in disordered eating etiology, this could be an important avenue for future research to better understand how different types of stress influence risk for EE in women. The accumulation of stress impact throughout the day may also contribute to the stronger associations between daily stress and EE in my study. In fact, a recent study by Smith and colleagues (2020) showed that individuals with higher levels of perceived stress and greater stress accumulation throughout the day engaged in higher levels of BE and reported more food cravings than individuals with lower levels of perceived stress or stress accumulation (between- subject effects). They also found that within each day, individuals were more likely to engage in higher levels of BE following moments of greater stress accumulation (within-subject effects) (Smith et al., 2020). My study suggests that women who experience higher levels of major life stress in the last 12 months may experience stronger associations between daily stress impact and EE. This is an important finding that may help to better identify women at-risk for EE. However, this finding should be interpreted with caution as the finding was only at a trend-level in the within-subject models and was not observed in studies of between-subject effects. Replication studies will be needed to verify this finding and further elucidate the relationship between major life stress and daily stress impact on EE in women. Originally, this study planned to examine how both daily stress frequency and impact are associated with EE, however, due to multicollinearity between these variables, I was only able to 28 examine associations between daily stress impact and EE. Other studies have also found multicollinearity between daily stress frequency and impact measures (Hay & Williams, 2013; McEwen & Akil, 2020; Rojo et al., 2006; Woods et al., 2010) and suggested a possible need for to assess frequency and impact via multiple measures (e.g., Perceived Stress Questionnaire, Daily Stress Inventory, Daily Hassles Scale, Daily Hassles & Uplifts Scale) to reduce the likelihood of multicollinearity between the variables and allow for better delineation of the associations between stress frequency and impact on EE. However, this option is not guaranteed to eliminate the multicollinearity between the two variables. If multicollinearity persists, it may bring into question a conceptual issue – i.e., that the two constructs are not truly separate. In other words, do women who experience a low frequency of highly impactful daily stressors engage in higher levels of EE than women who experience a high frequency of low impact daily stressors? This is a yet unexplored question in the stress and disordered eating literature. Given the inter-individual nature of the stress response, both options are possible and likely depend upon a woman’s past experience with the daily stressors, level of control over the stressor, or duration between stressors, among other factors. Future studies will be needed to answer this question. If stress frequency and impact are indeed found to be separate constructs, it will also be important to examine the nature of the association between stress frequency and impact and their role in EE (e.g., does stress impact mediate or moderate associations between stress frequency and EE in women?). Importantly, in the present study, increases in same-day stress impact more strongly predicted increases in daily levels of EE than stress impact from prior days, suggesting that stress and EE must occur in close temporal proximity in order for stress to influence EE. This finding is consistent with other daily studies showing a stronger association between increased stress and 29 same-day versus next-day disordered eating (Barker, Williams, & Galambos, 2006; Freeman & Gil, 2004; Smith et al., 2020). While the current study did not assess the temporal sequence of stress and EE within the day, other studies report that stress levels increase significantly in the hours leading up to BE and decrease following BE (Smyth et al., 2009). Increased levels of negative affect appear to mediate associations between increased stress impact and increased levels of BE (Goldschmidt et al., 2014; Srivastava, Lampe, Michael, Manasse, & Juarascio, 2021). Thus, changes in stress levels may be stronger for same-day rather than subsequent-day EE because of the immediate increase in negative affect - a strong prospective risk factor for EE (Klatzkin et al., 2019; Steinsbekk, Barker, Llewellyn, Fildes, & Wichstrøm, 2018). Alternatively, because same-day and time-lagged variables are often correlated with one another (r’s = 0.10- 0.13, p<.002 in this study), a lag structure within a model may over-specify the influence of the same-day predictors by including a sequence of lagged variables that contribute only marginally to the dependent variable, thus making it difficult to detect a significant effect of time-lagged variables above and beyond that of the same-day variables. Future studies are needed to determine if stronger associations between same-day stress and EE are due to factors such as negative affect, statistical artifacts, or both. Moving forward, it will be important to identify the mechanisms underlying daily stress impact-EE associations. Given the interplay between stress and brain reward systems, alterations within the mesocorticolimbic system is a key candidate to consider. Acute stress increases activity in mesocorticolimbic regions (e.g., anterior cingulate cortex, nucleus accumbens) while simultaneously reducing activity in brain regions associated with inhibitory control and executive functioning (e.g., prefrontal cortex) that play an important role regulating emotions (Dixon, Thiruchselvam, Todd, & Christoff, 2017) and ‘braking’ reward-related behavior 30 (Arnsten, 2015; Baik, 2020). Following chronic stress, activity in regulatory regions becomes downregulated such that they are no longer able to actively regulate emotional responses to stress (Orem et al., 2019) or reward-related behavior following stress (Arnsten, 2015). With poor regulatory control over food intake and emotion regulation, it may be more difficult to not overeat, particularly when a woman is eating highly rewarding palatable food and experiencing high levels of stress impact. Additionally, stress interacts with mesocorticolimbic dopamine signaling to influence reward-related behavior. For example, acute stress enhances reward- induced dopaminergic signaling to brain reward regions (e.g., nucleus accumbens) (Graf et al., 2013), helping to reinforce the rewarding behavior (e.g., EE). However, chronic stress suppresses reward-induced dopamine release into the nucleus accumbens (de Kloet, Joëls, & Holsboer, 2005a; Minami et al., 2017), weakens signaling to reward reinforcing pathways, and increases signaling to reward inhibitory pathways (Francis et al., 2015). Because palatable food is inherently rewarding, chronic stress may promote the overeating of palatable food to compensate for this hypo-reward state. Stress may also alter signals associated with general feeding to influence risk for EE in women. Of particular interest is the orexigenic neurotransmitter, ghrelin. Ghrelin, traditionally thought of as a pro-hunger hormone (Müller et al., 2015), also regulates stress (Stone, Harmatz, & Goosens, 2020) and reward (Stievenard et al., 2017). Following acute stress, ghrelin levels transiently increase (Stone et al., 2020) and can remain elevated for prolonged periods of time following chronic stress (Stone et al., 2020). Through its actions in the mesocorticolimbic reward system, ghrelin enhances dopamine release into the nucleus accumbens and increases dopamine turnover (Stievenard et al., 2017). In particular, ghrelin’s actions in the ventral tegmental area (the central hub for dopamine synthesis; (Han et al., 2017) increase the firing rate of 31 dopaminergic neurons (Abizaid et al., 2006) and play a key role in the motivation for and consumption of palatable food (Stievenard et al., 2017). Thus, stress-induced elevations in ghrelin levels may increase risk for EE via increased dopaminergic signaling and subsequent increased motivation for palatable food reward. While little is known about how stress and ghrelin interact to influence disordered eating, ghrelin is positively associated with EE in women (Rossi et al., 2021) and has been suggested to mediate associations between early childhood trauma and BE (Rossi et al., 2021). These associations may be due to altered ghrelin feedback mechanisms (Raspopow, Abizaid, Matheson, & Anisman, 2010, 2014) that can prolong ghrelin’s orexigenic and dopaminergic effects to increase palatable food consumption even after the stress is no longer present. Additional work using animal studies will be needed to examine these mechanisms more directly to better elucidate how stress influences EE in women. Lastly, a novel finding of my study was that lower HCC predicted higher levels of EE, potentially indicative of hypoactive hypothalamic-pituitary-adrenal (HPA)-axis functioning in women who engage in higher levels of EE. These findings are consistent with prior findings showing a blunted/hypoactive cortisol response to stress in women who EE (Het, Vocks, Wolf, Herpertz, & Wolf, 2020; Het et al., 2015; Tomiyama et al., 2011; van Strien, Roelofs, & de Weerth, 2013). Of note, these findings are also in contrast to other studies reporting either a positive (Gluck, Geliebter, Hung, & Yahav, 2004; Gluck, Geliebter, & Lorence, 2004a; Koo- Loeb, Costello, Light, & Girdler, 2000) or no association (Coutinho et al., 2007; Schulz et al., 2011) between cortisol and EE/BE in women. This inconsistency may be due to diurnal fluctuations in salivary and urinary cortisol measures (Carnell et al., 2018). Because HCC provides a cumulative, longitudinal measure of cortisol that is unaffected by diurnal fluctuations in cortisol, it may provide a useful index of cortisol-EE/BE associations in women. Importantly, 32 however, HCC significantly predicted EE only after controlling for BMI, age, total hours of sleep per night, income, and hair hygiene practices. Additionally, while hypoactive HPA-axis functioning can develop following chronic stress exposure (Lo Sauro, Ravaldi, Cabras, Faravelli, & Ricca, 2008), my findings found no significant association between HCC and daily stress impact (r = 0.02; p>.05) assessed over the same time period (Table 4). Other studies have reported a similar lack of association between HCC and self-reported indices of stress (e.g., Braig et al., 2016; O’Brien, Tronick, & Moore, 2013; Schlotz et al., 2008; Streit et al., 2016). A misalignment in time periods over which HCC and self-reported measures assessed cortisol/stress (e.g., HCC assessed cortisol over the past 3 months and self-reported measures assessed stress in the past 1 month) is one potential reason for these lack of associations; however this was controlled for in my study. Thus, other yet unknown factors must be contributing to these lack of associations. Future studies are needed to clarify this discrepancy and replicate associations between HCC and EE. Before concluding, it is important to note limitations of the current study. Firstly, while my goal was to examine how stress is associated with EE, it is possible that stress and EE exhibit a reciprocal relationship. In fact, despite studies reporting stress levels decreasing immediately after BE (Smyth et al., 2009), BE is often accompanied by feelings of shame, guilt, and/or disgust over the binge (American Psychiatric Association, 2013), suggesting that BE is distressing in the long-term. To test the possibility that EE is also distressing, post-hoc MLMs were conducted examining predictive effects of EE from one and two days ago on daily stress impact (see Supplemental Table 7). Results showed that while EE from two days ago did not predict daily stress impact (β = 0.02; p>.01), EE from one day ago did (β = 0.03; p<.001). These findings are intriguing given that the difficulty detecting time-lagged associations using MLMs. 33 Nevertheless, these findings do suggest that there may be a reciprocal relationship between EE and stress that can persist over multiple days. Because I didn’t examine when in the day stress and EE occurred relative to each other, I was unable to verify if EE predicts stress impact later that same day. While some studies suggest that bulimic symptoms, including BE, predict increased levels of daily hassles (Kwan & Gordon, 2016) and negative affect (Barker et al., 2006), additional studies will be needed to better capture the temporal dynamic between EE and stress on the same-day (e.g., ecological momentary assessments studies). Secondly, because HCC provided a cumulative measure of cortisol levels over the 49-day study period, I was only able to examine how between-subject changes in average cortisol levels were associated with changes in average levels of EE. However, in order to more fully understand how cortisol influences EE in women, within-subject studies that examine how daily variation in a woman’s cortisol levels contribute to shifts in her daily levels of EE are needed. Given individual differences in cortisol reactivity to stress (Appelhans, Pagoto, Peters, & Spring, 2010; Raspopow et al., 2010; Tomiyama et al., 2011; van Strien et al., 2013) and diurnal variation in cortisol-disordered eating associations (Carnell et al., 2018), assessing within-subject effects of cortisol levels on EE can be difficult and may require multiple measures of cortisol levels throughout the day. Additionally, because cortisol response to stress varies as a function of stress impact (Sapolsky, 2015), within-subject changes in daily cortisol levels may influence daily shifts in EE differently in women with high versus low levels of stress impact. To address these challenges and improve our understanding of the individual factors that contribute to a woman’s risk for EE, future studies that examine how within-subject changes in both daily stress impact and cortisol levels are associated with shifts in a woman’s EE are needed. 34 Lastly, my sample identified as predominantly non-Hispanic/Latinx and White and reported high parental income. Prior studies have reported that women identifying as Black, Indigenous, or a Person of Color (BIPOC) experience significantly higher levels of stress than White women, due to increased levels of discrimination, oppression, and lower socioeconomic status, among other stressors (Hatch & Dohrenwend, 2007; Lehrer, Goosby, Dubois, Laudenslager, & Steinhardt, 2020; O’Brien et al., 2017; Pickett, McCoy, & Odetola, 2020). Therefore, it will be important to replicate this study in a more racially, ethnically, and socioeconomically diverse sample of women to determine if findings from this study apply to women from other demographic groups. Nevertheless, my study had several strengths, including its intensive, longitudinal, daily study design, the inclusion of both self-reported and physiological indices of stress, and the examination of both between- and within-subject effects of stress on EE. Overall, the current study extends prior findings of stress and EE in women by distinguishing the relative influence of daily stress impact, major life stress, and cortisol on EE, identifying for whom stress increases risk for EE, and when shifts in stress levels most strongly influence a woman’s shifts in her EE. To build upon these findings and continue improving our understanding of how stress is associated with EE in women, future studies will be needed to address the following questions. Firstly, how are different types of stress associated with EE in women? While trauma (Backholm, Isomaa, & Birgegård, 2013; Palmisano, Innamorati, & Vanderlinden, 2016; Smyth, Heron, Wonderlich, Crosby, & Thompson, 2008; Zelkowtiz, Zerubavel, Zucker, & Copeland, 2021), interpersonal (Cain, Bardone-Cone, Abramson, Vohs, & Joiner, 2008; Goldschmidt et al., 2014; Monteleone et al., 2019), and psychosocial (Azarbad, Corsica, Hall, & Hood, 2010; Badrasawi & Zidan, 2019) stressors have been noted to be particularly salient forms of major life 35 and daily stress for disordered eating in women, most studies use measures that do not distinguish between the different stress types (e.g., interpersonal, psychosocial, work- and health- related stress). This limits our understanding of how certain types of stress are associated with risk for EE and if particular stress types are more likely to interact with one another to augment risk for EE in certain women compared to others. Studies that examine specific stress types would be helpful for better identifying at-risk women and can inform prevention/treatment practices to include more targeted resources for certain ‘risky’ stress types. Secondly, in order to better understand how stress frequency and impact are associated with EE, studies will need to examine women who experience low levels of high impact stressors are more likely to EE than women who experience high levels of low impact events. As mentioned above, both options are plausible given the inter-individual nature of the stress response. But until this question has been addressed, it will be difficult to understand whether stress frequency and impact are truly separate constructs or not. Lastly, more studies are needed to identify the mechanisms that underlie associations between daily stress impact and EE in women. Both human and animal studies will be necessary to answer this. Given strong associations between same-day stress impact and EE, human studies can focus on elucidating the factors that contribute to the temporal dynamics between daily stress and EE (e.g., increases in negative affect). To better understand how associations between stress, reward, and general feeding circuits influence EE, animal studies can examine how acute versus chronic stress influences dopaminergic regulation of palatable food consumption and how alterations in stress-induced ghrelin release contributes to this association. Together these studies will expand our understanding of how stress influences EE in women, help better identify at-risk women, and highlight potential mechanisms that inform prevention/treatment practices for EE. 36 APPENDICES 37 APPENDIX A Primary Analyses 38 Table 1 Descriptive information for the full sample (N = 477) and HCC sample (N = 234). FULL SAMPLE HCC SAMPLE Total Observed Observed Mean Mean Possible Range Range Variable (SD) (SD) Range Daily Variables – 49-Day Avga Avg. emotional eating 0.34 (0.42) 0-2.58 0.37 (0.47) 0-2.48 0-4 Avg. stress impact 29.53 0.92- 15.08 10.38- 0-399 (27.37) 253.02 (3.68) 32.13 Avg. stress frequency 11.08 0.67-56.11 11.81 0.91-56.11 0-57 (7.53) (7.94) Avg. negative affect 15.26 10.38- 1.37 (0.47) 1-3.48 0-80 (3.89) 42.37 Non-Daily Variables Major life stress in last 12 151.79 0-668 130.46 0-668 0-2246 monthsb (217.53) (107.38) HCC (pg/mg)b -- -- 10.71 1.82- -- (18.86) 191.20 BMI (kg/m2)c 24.63 17.06- 24.63 17.06- -- (5.38) 58.12 (5.15) 54.08 Ethnicity/Race/Income Percent Percent (N) (N) Ethnicity Hispanic or Latinx 3.8% (18) -- 2.6% (6) -- -- Non-Hispanic or Latinx 96.2% -- 97.4% -- -- (460) (229) Race White 89.3% -- 91.1% -- -- (427) (214) Black or African 5.0% (24) -- 3.4% (8) -- -- American Asian 1.3% (6) -- 0.4% (1) -- -- More than one race 4.4% (21) -- 5.1% (12) -- -- Income Under $20,000 2.1% (10) -- 2.1% (5) -- -- $20,000-$40,000 3.6% (17) -- 3.1% (7) -- -- $40,000-$60,000 10.9% -- 11.1% (25) -- -- (52) 39 Table 1 (cont’d) $60,000-$100,000 27.8% -- 27.1% (61) -- -- (133) Over $100,000 52.3% -- 56.4% -- -- (250) (127) Note: Avg. = average; BMI = body mass index; HCC = hair cortisol concentration; stress impact = daily stress impact; stress frequency = daily stress frequency. a These values are the non-standardized means and standard deviations (SDs) for each daily variable that was collected over the 49-day collection period. b These values are the non-standardized means and SDs for major life stress in the last 12 months and HCC across the 49-day study period. c This value is the non-standardized mean and SD for BMI that was measured at the beginning, middle, and end of the 49-day collection period. 40 Table 2 Results from the between-subject MLMs examining main and interactive effects of stress variables and covariates on average levels of emotional eating (N = 477). MAIN EFFECTS Variables β (SD) t (df) p Intercept -0.07 (0.34) -0.22 (206.26) .83 Avg. stress impact 0.35 (0.06) 5.77 (148.63) <.001*** Avg. negative affect 0.28 (0.06) 4.35 (218.29) <.001*** Avg. BMI 0.02 (0.04) 0.55 (313.12) .59 Age <0.01 (0.02) 0.08 (209.57) .94 Income 0.01 (0.04) 0.28 (62.67) .78 Intercept 0.05 (0.37) 0.13 (205.01) .90 Major life stress in the -0.02 (0.05) -0.33 (103.59) .75 last 12 months Avg. negative affect 0.52 (0.06) 9.40 (170.10) <.001*** Avg. BMI 0.06 (0.05) 1.24 (283.52) .22 Age <0.01 (0.02) -0.22 (208.42) .83 Income 0.01 (0.05) 0.17 (57.95) .87 INTERACTION EFFECTS β (SD) t (df) p Intercept -0.09 (0.35) -0.27 (200.31) .79 Avg. stress impact 0.39 (0.06) 6.13 (153.75) <.001*** Major life stress in the -0.11 (0.05) -2.22 (79.57) .03 last 12 months Avg. stress impact X 0.11 (0.05) 2.18 (88.64) .03 major life stress in the last 12 months Avg. negative affect 0.27 (0.06) 4.14 (202.83) <.001*** Avg. BMI 0.03 (0.04) 0.73 (286.46) .47 Age <0.01 (0.02) -0.01 (205.04) .99 Income 0.01 (0.04) 0.15 (55.31) .88 Note: BMI = body mass index; Stress impact = daily stress impact. All daily variables were averaged prior to being included in the above models. BMI was averaged across the 3 study assessments prior to being included in the above models. *** Bonferroni corrected p<.002 41 Table 3 Results from the within-subject MLMs examining main and interactive effects of the same- day and time-lagged stress variables and covariates on daily levels of emotional eating (N = 477). SAME-DAY DAILY STRESS IMPACT MAIN EFFECTS Variables β (SD) t (df) p Intercept 0.01 (0.01) 0.68 (2804.13) .50 Same-day stress impact 0.15 (0.01) 10.45 (329.19) <.001*** Negative affect 0.13 (0.01) 9.09 (415.05) <.001*** BMI <0.01 (0.03) 0.06 (112.56) .96 INTERACTION EFFECTS β (SD) t (df) p Intercept 0.01 (0.01) 0.81 .42 (2881.49) Same-day stress impact 0.15 (0.02) 9.95 (327.49) <.001*** Major life stress in the last <0.01 (0.01) 0.18 .86 12 months (2933.70) Same-day stress impact X <0.01 (0.01) -0.18 (249.34) .86 major life stress in the last 12 months Negative affect 0.13 (0.02) 8.82 (406.50) <.001*** BMI 0.01 (0.01) 0.79 .43 (4636.52) DAILY STRESS FROM ONE DAY AGO MAIN EFFECTS β (SD) t (df) p Intercept -0.02 (0.01) -2.72 .01 (3180.25) Same-day stress impact 0.13 (0.02) 8.60 (370.53) <.001*** Stress impact from one 0.04 (0.01) 2.68 (344.12) .008** day ago Negative affect 0.13 (0.01) 9.04 (422.47) <.001*** BMI <0.01 (0.01) 0.54 .59 (4664.91) INTERACTION EFFECTS β (SD) t (df) p 42 Table 3 (cont’d) Intercept -0.02 (0.01) -2.62 .01 (3088.08) Same-day stress impact 0.13 (0.02) 8.33 (369.94) <.001*** Stress impact from one 0.04 (0.01) 2.77 (340.22) .006** day ago Major life stress in the last <0.01 (0.01) 0.17 .86 12 months (3179.58) Same-day stress impact X 0.01 (0.02) 0.76 (276.65) .45 major life stress in the last 12 months Stress impact from one <0.01 (0.01) -0.33 (274.43) .74 day ago X major life stress in the last 12 months Negative affect 0.13 (0.01) 8.78 (416.59) <.001*** BMI <0.01 (0.01) 0.51 .61 (4519.89) DAILY STRESS FROM TWO DAYS AGO MAIN EFFECTS β (SD) t (df) p Intercept -0.03 (0.01) -3.49 <.001*** (3046.27) Same-day stress impact 0.12 (0.01) 8.24 (366.51) <.001*** Stress impact from two 0.03 (0.01) 3.00 (318.08) <.001*** days ago Negative affect 0.14 (0.02) 8.83 (425.05) <.001*** BMI 0.01 (0.01) 0.78 .43 (4372.88) INTERACTION EFFECTS β (SD) t (df) p Intercept -0.03 (0.01) -3.40 .001** (2946.43) Same-day stress impact 0.12 (0.02) 8.12 (370.16) <.001*** Stress impact from two 0.03 (0.01) 2.42 (323.26) .02 days ago Major life stress in the last <0.01 (0.01) 0.27 .79 12 months (3038.50) Same-day stress impact X 0.01 (0.01) 0.63 (272.52) .53 major life stress in the last 12 months 43 Table 3 (cont’d) Stress impact from two <0.01 (0.01) -0.17 (243.41) .86 days ago X major life stress in the last 12 months Negative affect 0.13 (0.02) 8.51 (417.74) <.001*** BMI 0.01 (0.01) 0.74 .46 (4223.49) Note: BMI = body mass index; Stress impact = daily stress impact. ** p<.01; *** Bonferroni corrected p<.002 44 1 0.9 Correlation with Avg. 0.8 0.7 *** 0.6 0.5 0.4 EE 0.3 0.2 0.56 0.1 0.07 0 Avg. Stress Impact Avg. Major Life Stress - Last 12 months Figure 1. Between-subject Pearson correlations for average daily stress, major life stress, and emotional eating (N = 477). Note: ***Bonferroni corrected p<.002. 45 3000 2500 Low major life stress 2000 High major life stress Av. Emotional Eating 1500 1000 500 0 Low stress impact High stress impact -500 -1000 Figure 2. Two-way interaction between average stress impact and major life stress in the last 12 months. “High” and “Low” values represent 1 SD above and below the mean, respectively. 46 0.5 Correlation with Daily EE 0.4 0.3 *** 0.2 0.1 0.22 *** *** 0.06 0.05 0 Same-day stress impact Stress impact from 1 day ago Stress impact from 2 days ago Figure 3. Within-subject Pearson correlations for same-day and time-lagged daily stress impact, daily emotional eating, and average major life stress (N = 477). ***Bonferroni corrected p<.002. 47 APPENDIX B Exploratory Analyses 48 0.5 Correlation with HCC 0.4 0.3 0.2 0.1 0.07 0 0.02 -0.13 -0.1 -0.2 Avg. EE Avg. Stress Impact Figure 4. Pearson correlations for hair cortisol concentration (HCC), average stress impact, major life stress in the last 12 months, and average levels of emotional eating in the full sample of HCC participants (N = 234). Note: There were no significant correlations between HCC and either of the stress or eating variables. 49 0.5 Correlation with HCC 0.4 0.3 0.2 0.1 0 0.01 -0.15 0.03 -0.1 -0.2 Avg. EE Avg. Stress Impact Major Life Stress - Last 12 months Figure 5. Pearson correlations for hair cortisol concentration (HCC), average stress impact, major life stress in the last 12 months, and average levels of emotional eating in the subsample of HCC participants without confounding factors for HCC (N = 220). Note: There were no significant correlations between HCC and either of the stress or eating variables. 50 Table 4 Results from the between-subject, exploratory MLMs examining the effects of hair cortisol concentration (HCC) and covariates on average levels of emotional eating. Full HCC sample (N = 234) β (SD) t (df) p Intercept -0.21 (0.50) -0.42 (132.08) .68 HCC -0.14 (0.05) -2.70 (203.34) .007** Avg. negative affect 0.59 (0.07) 8.81 (93.28) <.001*** Avg. BMI 0.03 (0.05) 0.48 (179.02) .64 Age 0.01 (0.02) 0.37 (135.70) .71 Avg. hours of 0.01 (0.05) 0.10 (204.55) .92 sleep/night Income 0.03 (0.06) 0.47 (128.11) .64 Subsample of women without confounding factors for HCC (N = 220) β (SD) t (df) p Intercept -0.10 (0.50) -0.16 (124.72) .88 HCC -0.16 (0.05) -2.97(179.83) .003** Avg. negative affect 0.59 (0.07) 8.49 (81.79) <.001*** Avg. BMI 0.01 (0.06) 0.18 (28.75) .86 Age <0.01 (0.02) 0.11 (128.72) .91 Avg. hours of <0.01 (0.05) 0.07 (177.99) .95 sleep/night Income 0.03 (0.06) 0.51 (119.39) .61 Note: Avg. = average; BMI = body mass index; HCC = hair cortisol concentration. All daily variables were averaged prior to being included in the above models. BMI was averaged across the 3 study assessments prior to being included in the above models. ** p<.01; *** Bonferroni corrected p<.002 51 APPENDIX C Supplemental Analyses 52 Supplemental Table 1 Comparing descriptive information between participants who did and did not provide a hair sample. HCC sample Non-HCC (N = 234) sample t-test results p (N = 243) Variable Mean (SD) Mean (SD) t(df) = F Daily variables- 49- day Avga Avg. emotional eating 0.37 (0.47) 0.31 (0.36) -1.60(475) = 8.00 .11 Avg. stress impact 30.67 (27.01) 28.42 (27.71) -0.90(475) = 0.16 .37 Avg. stress frequency 11.81 (7.94) 10.39 (7.06) -2.07(475) = 2.50 .04 Avg. negative affect 15.08 (3.68) 15.43 (4.08) 0.99(475) = 0.13 .33 Non-Daily variables Major life stress in the 130.46 142.69 (108.88) 1.22(460) = 0.29 .23 last 12 months (107.38) Major life stress over 49 62.63 (70.08) 71.92 (73.19) 1.39(460) = 0.99 .16 days Major life stress across 109.44 194.51 (250.94) 4.28(460) = 27.94 <.001 the lifetime (168.48) BMI (kg/m2)b 24.63 (5.15) 24.63 (5.70) 0.00(399) = 0.56 1.00 Z-test of Ethnicity/Race/Income Percent (N) Percent (N) proportions Ethnicity Hispanic or 2.6% (6) 4.9% (12) .81 -0.24 Latinx Non-Hispanic or 97.4% (229) 95.1% (231) .18 1.35 Latinx Race (%, n) White 91.1% (214) 87.7% (213) 1.13 .26 Black or African 3.4% (8) 6.6% (16) -0.34 .73 American Asian 0.4% (1) 2.1% (5) -0.14 .89 More than one 5.1% (12) 3.7% (9) 0.16 .87 race Income Under $20,000 2.1% (5) 2.1% (5) 0 1.00 $20,000- 3.1% (7) 4.1% (10) -0.11 .91 $40,000 53 Supplemental Table 1 (cont’d) $40,000- 11.1% (25) 11.1% (27) 0 1.00 $60,000 $60,000- 27.1% (61) 30.4% (72) -0.37 .71 $100,000 Over $100,000 56.4% (127) 51.9% (123) 0.71 .48 Note: Avg. = average; BMI = body mass index; HCC = hair cortisol concentration. a These values are unstandardized (i.e., non-z-scored) means and standard deviations (SD) across the 49-day collection period that index the average level of study variables on any given day. b This value is the unstandardized mean and SD for BMI that was measured at the beginning, middle, and end of the 49-day collection period. 54 Supplemental Table 2 Results from the between-subject MLMs examining interactive effects of the daily stress variables with major life stress over the 49-day study period and across the lifetime on daily levels of emotional eating, including the covariates (N = 477). MAIN EFFECTS Predictors β (SD) t (df) p Intercept 0.03 (0.36) 0.10 (203.40) .93 Major life stress over 49 -0.05 (0.04) -1.29 (351.50) .20 days Avg. negative affect 0.53 (0.05) 9.75 (163.24) <.001*** Avg. BMI 0.05 (0.04) 1.14 (288.41) .26 Age <0.01 (0.02) -0.19 (206.44) .85 Income 0.01 (0.05) 0.26 (57.07) .80 Intercept 0.04 (0.36) 0.11 (205.78) .91 Major life stress across the -0.06 (0.04) -1.32 (357.70) .19 lifetime Avg. negative affect 0.53 (0.05) 9.74 (160.41) <.001*** Avg. BMI 0.05 (0.04) 1.23 (290.52) .22 Age <0.01 (0.02) -0.21 (208.84) .84 Income 0.01 (0.05) 0.28 (59.40) .78 INTERACTION EFFECTS Predictors β (SD) t (df) p Intercept 0.04 (0.35) 0.12 (201.43) .90 Avg. stress impact 0.36 (0.06) 6.17 (346.42) <.001*** Major life stress -0.08 (0.04) -1.82 (365.03) .07 Avg. stress impact X 0.08 (0.04) 1.98 (345.15) .05 major life stress over 49 days Avg. negative affect 0.33 (0.06) 5.59 (345.13) <.001*** Avg. BMI 0.06 (0.05) 1.19 (67.95) .24 Age <0.01 (0.02) -0.18 (205.34) .86 Income 0.03 (0.05) 0.68 (80.22) .50 Intercept -0.01 (0.35) -0.04 (204.77) .97 Avg. stress impact 0.35 (0.06) 5.60 (154.50) <.001*** Major life stress -0.08 (0.04) -2.07 (345.29) .04 Avg. stress impact X -0.04 (0.05) -0.91 (196.32) .37 major life stress across the lifetime 55 Supplemental Table 2 (cont’d) Avg. negative affect 0.27 (0.07) 4.11 (211.79) <.001*** Avg. BMI 0.03 (0.04) 0.70 (294.34) .48 Age <0.01 (0.02) -0.16 (208.03) .87 Income <0.01 (0.04) 0.03 (202.41) .98 Note: Avg. = average; BMI = body mass index; Stress impact = daily stress impact. All daily variables were averaged prior to being included in the above models. *** Bonferroni corrected p<.002 56 Supplemental Table 3 Results from the within-subject MLMs examining interactive effects of the same- day and time-lagged daily stress variables with major life stress over the 49-day study period and across the lifetime on daily levels of emotional eating, including the covariates (N = 477). SAME-DAY DAILY STRESS IMPACT INTERACTION EFFECTS Predictors β (SD) t (df) p Intercept 0.01 (0.01) 0.80 (2879.56) .43 Same-day stress impact 0.15 (0.01) 9.98 (320.06) <.001*** Major life stress over 49 <0.01 (<0.01) -0.04 (2870.98) .97 days Same-day stress impact X <0.01 (0.01) -0.29 (320.64) .77 major life stress over 49 days Negative affect 0.13 (0.02) 8.83 (406.46) <.001*** BMI 0.01 (0.01) 0.79 (4637.11) .43 Intercept 0.01 (0.01) 0.83 (2872.77) .41 Same-day stress impact 0.15 (0.01) 9.97 (315.94) <.001*** Major life stress across the <0.01 (0.01) 0.38 (2884.66) .70 lifetime Same-day stress impact X <0.01 (0.01) 0.27 (305.11) .79 major life stress across the lifetime Negative affect 0.13 (0.02) 8.83 (406.61) <.001*** BMI 0.01 (0.01) 0.79 (4638.27) .43 DAILY STRESS IMPACT FROM ONE DAY AGO INTERACTION EFFECTS Predictors β (SD) t (df) p Intercept -0.02 (0.01) -2.62 (3083.42) .01 Same-day stress impact 0.13 (0.02) 8.39 (361.99) <.001*** Stress impact from one day 0.04 (0.01) 2.82 (335.40) .005** ago Major life stress over 49 <0.01 (0.01) 0.29 (3075.63) .78 days 57 Supplemental Table 3 (cont’d) Same-day stress impact X 0.01 (0.02) 0.41 (346.57) .68 major life stress over 49 days Stress impact from one day -0.02 (0.01) -1.50 (357.20) .14 ago X major life stress over 49 days Negative affect 0.13 (0.01) 8.78 (416.60) <.001*** BMI <0.01 (0.01) 0.51 (4521.33) .61 Intercept -0.02 (0.01) -2.63 (3073.31) .01 Same-day stress impact 0.13 (0.02) 8.43 (356.87) <.001*** Stress impact from one day 0.04 (0.01) 2.78 (332.70) .006** ago Major life stress across the <0.01 (0.01) 0.02 (3097.52) .99 lifetime Same-day stress impact X <0.01 (0.02) 0.16 (342.98) .87 major life stress across the lifetime Stress impact from one day 0.01 (0.01) 0.70 (338.34) .48 ago X major life stress across the lifetime Negative affect 0.13 (0.01) 8.77 (342.98) .87 BMI <0.01 (0.01) 0.51 (4521.98) .61 DAILY STRESS IMPACT FROM TWO DAYS AGO INTERACTION EFFECTS Predictors β (SD) t (df) p Intercept -0.03 (0.01) -3.42 (2940.67) .001** Same-day stress impact 0.13 (0.02) 8.21 (359.49) <.001*** Stress impact from two 0.03 (0.01) 2.38 (315.26) .02 days ago Major life stress over 49 <0.01 (0.01) 0.03 (2932.01) .98 days Same-day stress impact X <0.01 (0.01) -0.13 (349.06) .90 major life stress over 49 days Stress impact from two 0.01 (0.01) 0.55 (332.59) .58 days ago X major life stress over 49 days Negative affect 0.13 (0.02) 8.50 (417.74) <.001*** BMI 0.01 (0.01) 0.74 (4223.28) .46 Intercept -0.03 (0.01) -3.41 (2930.84) .001** 58 Supplemental Table 3 (cont’d) Same-day stress impact 0.13 (0.02) 8.23 (354.24) <.001*** Stress impact from two 0.03 (0.01) 2.43 (308.35) .02 days ago Major life stress across the <0.01 (0.01) 0.05 (2955.46) .96 lifetime Same-day stress impact X <0.01 (0.02) 0.15 (344.43) .88 major life stress across the lifetime Stress impact from two 0.01 (0.01) 0.97 (312.38) .33 days ago X major life stress across the lifetime Negative affect 0.13 (0.02) 8.50 (417.67) <.001*** BMI 0.01 (0.01) 0.72 (4225.50) .47 Note: BMI = body mass index; Stress impact = daily stress impact. ** p<.01; *** Bonferroni corrected p<.002 59 Supplemental Table 4 Between-subject Pearson correlations for average daily stress, major life stress, emotional eating, and covariates (N = 477). 1. 2. 3. 4. 5. 6. 7. Daily Variables 1. Avg. emotional 1.00 0.56*** 0.54 0.07 0.05 -0.10* 0.10 eating *** 2. Avg. stress -- 1.00 0.67 0.33*** 0.03 -0.12* -0.02 impact *** 3. Avg. negative -- -- 1.00 0.28*** -0.02 -0.24*** -0.04 affect Non-Daily Variables 4. Avg. Major life -- -- -- 1.00 0.09 -0.05 -0.18*** stress in last 12 months 5. Avg. BMI -- -- -- -- 1.00 0.31*** -0.09 6. Age -- -- -- -- -- 1.00 -0.10* 7. Income -- -- -- -- -- -- 1.00 Note: Avg. = average; BMI = body mass index; Stress impact = daily stress impact All daily variables were averaged prior to being included in the above models. BMI was averaged across the 3 study assessments before being included in the above models. *p <.05; ***Bonferroni corrected p<.002 60 Supplemental Table 5 Within-subject Pearson correlations for same-day and time-lagged daily stress impact, daily emotional eating, average major life stress, and covariates (N = 477). Variables 1. 2. 3. 4. 5. Same-day daily variables 1. Emotional eating 1.00 0.22*** 0.18*** 0.06*** 0.05*** 2. Same-day stress impact -- 1.00 0.46*** 0.13*** 0.10*** 3. Negative affect -- -- 1.00 0.07*** 0.05*** Time-lagged daily stress variables 4. Stress impact 1 day ago -- -- -- 1.00 0.79*** 5. Stress impact 2 days ago -- -- -- -- 1.00 Note: All variables were averaged prior to being included in analyses. ***Bonferroni corrected p<.002. 61 Supplemental Table 6 Pearson correlations for hair cortisol concentration (HCC), emotional eating, and covariates in the full sample of HCC participants (N = 234) and subsample of HCC participants without confounding factors for HCC (N = 220). Full sample 1. 2. 3. 4. 5. 6. 7. 8. 9. (N = 234) 1. HCC 1.00 -0.13 0.02 0.07 0.06 -0.05 0.05 <0.01 0.09 2. Avg. -- 1.00 0.58 0.11 0.62 0.10 -0.12 0.03 <0.01 emotional *** *** eating 3. Avg. stress -- -- 1.00 0.32 0.75 0.13* - -0.07 -0.01 impact *** *** 0.14 * 4. Major life -- -- -- 1.00 0.22 0.16* 0.13 <0.01 0.09 stress in the ** * last 12 months 5. Avg. -- -- -- -- 1.00 0.08 - <0.01 0.01 negative 0.24 affect *** 6. Avg. BMI -- -- -- -- -- 1.00 0.28 -0.09 -0.12 *** 7. Age -- -- -- -- -- -- 1.00 - - 0.14* 0.26* ** 8. Avg. hours -- -- -- -- -- -- -- 1.00 0.11 of sleep/night 9. Income -- -- -- -- -- -- -- -- 1.00 Subsample of women without confounding 1. 2. 3. 4. 5. 6. 7. 8. 9. factors for HCC (N = 220) 62 Supplemental Table 6 (cont’d) 1. HCC 1.00 - 0.01 0.03 0.06 -0.06 0.04 <0.01 0.10 0.15* 2. Avg. -- 1.00 0.58 0.02 0.62 0.08 - 0.03 0.01 emotional *** *** 0.15 eating * 3. Avg. stress -- -- 1.00 0.26 0.76 0.13 - -0.08 -0.01 impact *** *** 0.15 * 4. Major life -- -- -- 1.00 0.10 0.12 0.05 0.13 0.05 stress in the last 12 months 5. Avg. -- -- -- -- 1.00 0.08 - <0.01 0.01 negative 0.26 affect *** 6. Avg. BMI -- -- -- -- -- 1.00 0.27 -0.09 -0.11 *** 7. Age -- -- -- -- -- -- 1.00 -0.13 - 0.24* ** 8. Avg. hours -- -- -- -- -- -- -- 1.00 -0.13 of sleep/night 9. Income -- -- -- -- -- -- -- -- 1.00 Note: Avg. = average; BMI = body mass index; HCC = hair cortisol concentration. All daily variables were averaged prior to being included in the above correlations. BMI was averaged across the 3 study assessments prior to being included in the above correlations. *p <.05; **p <.01; *** Bonferroni corrected p<.002. 63 Supplemental Table 7 Results from the post-hoc within-subject MLMs examining predictive effects of emotional eating from one and two days ago on daily stress impact, including the covariates (N = 477). EMOTIONAL EATING FROM 1 DAY AGO Predictors β (SD) t (df) p Intercept -0.02 (0.01) -2.58 (2987.48) .01 Same-day emotional eating 0.09 (0.01) 8.73 (288.74) <.001*** Emotional eating from one day ago 0.03 (0.01) 3.92 (263.40) <.001*** Avg. negative affect 0.42 (0.01) 35.22 (292.04) <.001*** Avg. BMI 0.01 (0.01) 0.83 (273.23) .41 EMOTIONAL EATING FROM 2 DAYS AGO Predictors β (SD) t (df) p Intercept -0.03 (0.01) -4.08 (2861.87) <.001*** Same-day emotional eating 0.09 (0.01) 7.96 (284.70) <.001*** Emotional eating from two days ago 0.02 (0.01) 2.39 (261.92) .02 Avg. negative affect 0.42 (0.01) 33.82 (296.87) <.001*** Avg. BMI 0.02 (0.01) 1.57 (272.89) .12 Note: Avg. = average; BMI = body mass index. All daily variables were averaged prior to being included in the above models. *** Bonferroni corrected p<.002 64 REFERENCES 65 REFERENCES Abizaid, A., Liu, Z.-W., Andrews, Z. B., Shanabrough, M., Borok, E., Elsworth, J. D., … Horvath, T. I. (2006). Ghrelin modulates the activity and synaptic input organization ofmidbrain dopamine neurons while promoting appetite. J. Clin. Invest, 116, 3229–3239. https://doi.org/10.1172/JCI29867 Adam, T. C., & Epel, E. S. (2007). Stress, eating and the reward system. Physiology and Behavior, 91(4), 449–458. American Psychiatric Association. (2013). Diagnostic and Statistical Manual of Mental Disorders (5th ed.). Arlington, VA: American Psychiatric Publishing. Appelhans, B. M., Pagoto, S. L., Peters, E. N., & Spring, B. J. (2010). HPA axis response to stress predicts short-term snack intake in obese women. Appetite, 54(1), 217–220. https://doi.org/10.1016/j.appet.2009.11.005 Arnow, B., Kenardy, J., & Agras, W. S. (1995). The Emotional Eating Scale: the development of a measure to assess coping with negative affect by eating. International Journal of Eating Disorders, 18, 79–90. Arnsten, A. F. (2015). Stress weakens prefrontal networks: molecular insults to higher cognition. Nature Neuroscience, 18(10), 1376–1385. https://doi.org/10.1038/nn.4087 Azarbad, L., Corsica, J., Hall, B., & Hood, M. (2010). Psychosocial correlates of binge eating in Hispanic, African American, and Caucasian women presenting for bariatric surgery. Eating Behaviors, 11(2), 79–84. https://doi.org/10.1016/j.eatbeh.2009.10.001 Backholm, K., Isomaa, R., & Birgegård, A. (2013). The prevalence and impact of trauma history in eating disorder patients. European Journal of Psychotraumatology, 4. https://doi.org/10.3402/ejpt.v4i0.22482 Badrasawi, M. M., & Zidan, S. J. (2019). Binge eating symptoms prevalence and relationship with psychosocial factors among female undergraduate students at Palestine Polytechnic University: a cross-sectional study. International Journal of Eating Disorders, 7, 33. https://doi.org/10.1186/s40337-019-0263-1 Baik, J. (2020). Stress and the dopaminergic reward system. Experimental & Molecular Medicine, 52(12), 1879–1890. https://doi.org/10.1038/s12276-020-00532-4 Barker, E. T., Williams, R. L., & Galambos, N. L. (2006). Daily spillover to and from binge eating in first-year university females. Eating Disorders, 14(3), 229–242. https://doi.org/10.1080/10640260600639079 66 Becker, D. F., & Grilo, C. M. (2011). Childhood maltreatment in women with binge-eating disorder: Associations with psychiatric comorbidity, psychological functioning, and eating pathology. Eating and Weight Disorders, 16(2), e113–e120. 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. Braig, S., Grabher, F., Ntomchukwu, C., Reister, F., Stalder, T., Kirschbaum, C., … Genuneit, J. (2016). The association of hair cortisol with self-reported chronic psychosocial stress and symptoms of anxiety and depression in women shortly after delivery. Paediatric and Perinatal Epidemiology, 30(2), 97–104. https://doi.org/10.1111/ppe.12255 Brantley, P. J., Dietz, L. S., McKnight, G. T., Jones, G. N., & Tulley, R. (1988). Convergence between the Daily Stress Inventory and endocrine measures of stress. Journal of Consulting and Clinical Psychology, 56, 549–551. Brantley, P. J., Waggoner, C. D., Jones, G. N., & Rappaport, N. B. (1987). A daily stress inventory: development, reliability, and validity. Journal of Behavioral Medicine, 10(1), 61–74. Burt, S. A., & Klump, K. L. (2013). The Michigan State University Twin Registry (MSUTR): An update. Twin Research and Human Genetics, 16(1), 334–350. https://doi.org/10.1017/thg.2012.87 Burt, S. A., & Klump, K. L. (2019). The Michigan State University Twin Registry (MSUTR): 15 Years of Twin and Family Research. Twin Research and Human Genetics, 22(6), 741–745. Cain, A. S., Bardone-Cone, A. M., Abramson, L. Y., Vohs, K. D., & Joiner, T. E. (2008). Refining the relationships of perfectionism, self-efficacy, and stress to dieting and binge eating: Examining the appearance, interpersonal, and academic domains. International Journal of Eating Disorders, 41(8), 713–721. https://doi.org/10.1002/eat.20563 Carnell, S., Grillot, C., Ungredda, T., Ellis, S., Mehta, N., Holst, J., & Geliebter, A. (2018). Morning and afternoon appetite and gut hormone responses to meal and stress challenges in obese individuals with and without binge eating disorder. International Journal of Obesity (London), 42(4), 841–849. Cieszyński, L., Jendrzejewski, J., Wiśniewski, P., Owczarzak, A., & Sworczak, K. (2019). Hair cortisol concentration in a population without hypothalamic-pituitary-adrenal axis disorders. Advances in Clinical and Experimental Medicine, 28(3), 369–373. https://doi.org/10.17219/acem/90038 Cohen, S., Doyle, W. J., & Baum, A. (n.d.). Socioeconomic status is associated with stress hormones. 2006, 68(3), 414–420. https://doi.org/10.1097/01.psy.0000221236.37158.b9 67 Coutinho, W. F., Moreira, R. O., Spagnol, C., & Appolinario, J. C. (2007). Does binge eating disorder alter cortisol secretion in obese women? Eating Behaviors, 8(1), 59–64. Crowther, J. H., Sanftner, J., Bonifazi, D. Z., & Shepard, K. L. (2001). The role of daily hassles in binge eating. International Journal of Eating Disorders, 29(4), 449–454. D’Anna-Hernandez, K. L., Ross, R. G., Natvig, C. L., & Laudenslager, M. L. (2011). Hair cortisol levels as a retrospective marker of hypothalamic-pituitary axis activity throughout pregnancy: Comparison to salivary cortisol. Physiology and Behavior. https://doi.org/10.1016/j.physbeh.2011.02.041 Dallman, M. F., Pecoraro, N., Akana, S. F., la Fleur, S. E., Gomez, F., Houshyar, H., … Manalo, S. (2003). . Chronic stress and obesity: a new view of “comfort food.” Proceedings of the National Academy of Sciences, 100(20), 11696–11701. de Kloet, E. R., Joëls, M., & Holsboer, F. (2005a). Stress and the brain: from adaptation to disease. Nature Reviews Neuroscience, 6(6), 463–475. https://doi.org/10.1038/nrn1683 de Kloet, E. R., Joëls, M., & Holsboer, F. (2005b). Stress and the brain: From adaptation to disease. Nature Reviews Neuroscience, 6(6), 463–475. https://doi.org/doi: 10.1038/nrn1683. Degortes, D., Santonastaso, P., Zanetti, T., Tenconi, E., Veronese, A., & Favaro, A. (2014). Stressful life events and binge eating disorder. European Eating Disorders Review, 5, 378– 382. Diggins, A., Woods-Giscombe, C., & Waters, S. (2015). The association of perceived stress, contextualized stress, and emotional eating with body mass index in college-aged Black women. Eating Behaviors. https://doi.org/10.1016/j.eatbeh.2015.09.006 Dixon, M. L., Thiruchselvam, R., Todd, R., & Christoff, K. (2017). Emotion and the prefrontal cortex: An integrative review. Psychological Bulletin, 143(10), 1033–1081. https://doi.org/10.1037/bul0000096 Faresjö, A., Theodorsson, E., Chatziarzenis, M., Sapouna, V., Claesson, H., Koppner, J., & Faresjö, T. (2013). Higher perceived stress but lower cortisol levels found among young greek adults living in a stressful environment in comparison with Swedish young adults. PLoS One, 8(9), e73828. Ferro, M. A., & Gonzalez, A. (2020). Hair cortisol concentration mediates the association between parent and child psychopathology. Psychoneuroendocrinology, 114(104613). https://doi.org/10.1016/j.psyneuen.2020.104613 Francis, T. C., Chandra, R., Friend, D. M., Finkel, E., Dayrit, G., Miranda, J., … Lobo, M. K. (2015). Nucleus accumbens medium spiny neuron subtypes mediate depression-related outcomes to social defeat stress. Biological Psychiatry, 77(3), 212–222. 68 https://doi.org/10.1016/j.biopsych.2014.07.021 Freeman, L. M. Y., & Gil, K. M. (2004). Daily stress, coping, and dietary restraint in binge eating. International Journal of Eating Disorders, 36(2), 204–212. García-León, M. A., Pérez-Mármol, J. M., Gonzalez-Pérez, R., García-Ríos, M. D. ., & Peralta- Ramírez, M. I. (2019). Relationship between resilience and stress: Received stress, stressful life events, HPA axis response during a stressful task and hair cortisol. Physiology & Behavior, 202, 87–93. https://doi.org/10.1016/j.physbeh.2019.02.001 Gluck, M. E. (2006). Stress response and binge eating disorder. Appetite, 46(1), 26–30. Gluck, M. E., Geliebter, A., Hung, J., & Yahav, E. (2004). Cortisol, hunger, and desire to binge eat following a cold stress test in obese women with binge eating disorder. Psychosomatic Medicine, 66(6), 876–881. Gluck, M. E., Geliebter, A., & Lorence, M. (2004a). Cortisol stress response is positively correlated with central obesity in obese women with binge eating disorder (BED) before and after cognitive-behavioral treatment. Annals of the New York Academy of Sciences, 1032, 202–207. https://doi.org/10.1196/annals.1314.021 Godfrey, J. R., Diaz, M. P., Pincus, M., Kovacs-Balint, Z., Feczko, E., Earl, E., … Michopoulos, V. (2019). Diet matters: Glucocorticoid-related neuroadaptations associated with calorie intake in female rhesus monkeys. Psychoneuroendocrinology, 91, 169–178. Godoy, L. D., Rossignoli, M. T., Delfino-Pereira, P., Garcia-Cairasco, N., & de Lima Umeoka, E. H. (2018). A Comprehensive Overview on Stress Neurobiology: Basic Concepts and Clinical Implications. Frontiers in Behavioral Neuroscience, 12, 127. https://doi.org/10.3389/fnbeh.2018.00127 Goldschmidt, A. B., Wonderlich, S. A., Crosby, R. D., Engel, S. G., Lavender, J. M., Peterson, C. B., … Mitchell, J. E. (2014). Ecological momentary assessment of stressful events and negative affect in bulimia nervosa. Journal of Consulting and Clinical Psychology, 82(1), 30–39. Graf, E. N., Wheeler, R. A., Baker, D. A., Ebben, A. L., Hill, J. E., McReynolds, J. R., … Gasser, P. J. (2013). Corticosterone acts in the nucleus accumbens to enhance dopamine signaling and potentiate reinstatement of cocaine seeking. J Neuroscience, 33(29), 11800– 11810. https://doi.org/10.1523/JNEUROSCI.1969-13.2013 Grissom, N., & Bhatnagar, S. (2009). Habituation to repeated stress: get used to it. Neurobiology of Learning and Memory, 92(2), 215–224. Haedt-Matt, A. A., Keel, P. K., Racine, S. E., Burt, S. A., Hu, J. Y., Boker, S., … Klump, K. L. (2014). Do emotional eating urges regulate affect? Concurrent and prospective associations and implications for risk models of binge eating. International Journal of Eating Disorders, 69 47(8), 874–877. Han, X., Jing, M. Y., Zhao, T. Y., Wu, N., Song, R., & Li, J. (2017). Role of dopamine projections from ventral tegmental area to nucleus accumbens and medial prefrontal cortex in reinforcement behaviors assessed using optogenetic manipulation. Metabolic Brain Disease, 32(5), 1491–1502. https://doi.org/10.1007/s11011-017-0023-3 Hatch, S. L., & Dohrenwend, B. P. (2007). Distribution of traumatic and other stressful life events by race/ethnicity, gender, SES, and age, a review of the research. American Journal of Community Psychology, 40(3–4), 313–332. https://doi.org/10.1007/s10464-007-9134-z Hawkins, R. C., & Clement, P. F. (1984). Binge eating: Measurement problems and a conceptual model. (R. C. Hawkins, W. J. Fremouw, & P. F. Clement, Eds.) (The Binge). New York, NY: Springer. Hay, P., & Williams, S. E. (2013). Exploring relationships over time between psychological distress, perceived stress, life events and immature defense style on disordered eating pathology. BMC Psychology, 1(1), 27. Het, S., Vocks, S., Wolf, J. M., Hammelstein, P., Herpertz, S., & Wolf, O. T. (2015). Blunted neuroendocrine stress reactivity in young women with eating disorders. Journal of Psychosomatic Research, 78(3), 260–267. https://doi.org/10.1016/j.jpsychores.2014.11.001 Het, S., Vocks, S., Wolf, J. M., Herpertz, S., & Wolf, O. T. (2020). Treatment-Resistant Blunted HPA Activity, but Reversible Cardiovascular Stress Reactivity in Young Women With Eating Disorders. Frontiers in Psychiatry, 11, 726. https://doi.org/10.3389/fpsyt.2020.00726 Hildebrandt, B. A., Racine, S. E., Keel, P. K., Burt, S. A., Neale, M., Boker, S., … Klump, K. L. (2015). The effects of ovarian hormones and emotional eating on changes in weight preoccupation across the menstrual cycle. International Journal of Eating Disorders, 48(5), 477–486. Hoffman, M. C., D’Anna-Hernandez, K., Benitez, P., Ross, R. G., & Laudenslager, M. L. (2017). Cortisol during human fetal life: Characterization of a method for processing small quantities of newborn hair from 26 to 42 weeks gestation. Developmental Psychobiology, 59(1), 123–127. Holmes, T. H., & Rahe, R. H. (1967). The Social Readjustment Rating Scale. Journal of Psychosomatic Research, 11(2), 213–218. Hudson, J. I., Hiripi, E., Pope, H. G., & Kessler, R. C. (2007a). The Prevalence and Correlates of Eating Disorders in the National Comorbidity Survey Replication. Biological Psychiatry. https://doi.org/10.1016/j.biopsych.2006.03.040 Hudson, J. I., Hiripi, E., Pope, H. G., & Kessler, R. C. (2007b). The Prevalence and Correlates of 70 Eating Disorders in the National Comorbidity Survey Replication. Biological Psychiatry, 61(3), 348–358. https://doi.org/10.1016/j.biopsych.2006.03.040 Jarvela-Reijonen, E., Karhunen, L., Sairanen, E., Rantala, S., Laitinen, J., Puttonen, S., … Kolehmainen, M. (2016). High perceived stress is associated with unfavorable eating behavior in overweight and obese Finns of working age. Appetite, 103, 249–258. Kamin, H. S., & Kertes, D. A. (2017). Cortisol and DHEA in development and psychopathology. Hormones and Behavior, 89, 69–85. Kanner, A. D., Coyne, J. C., Schaefer, C., & Lazarus, R. S. (1981). Comparison of Two Modes of Stress Measurement: Daily Hassles and Uplifts Versus Major Life Events. Journal of Behavioral Medicine, 4(1). Karlén, J., Ludvigsson, J., Frostell, A., Theodorsson, E., & Faresjö, T. (2011). Cortisol in hair measured in young adults - A biomarker of major life stressors? BMC Clinical Pathology. https://doi.org/10.1186/1472-6890-11-12 Kirschbaum, C., Tietze, A., Skoluda, N., & Dettenborn, L. (2009). Hair as a retrospective calendar of cortisol production-Increased cortisol incorporation into hair in the third trimester of pregnancy. Psychoneuroendocrinology, 34(1), 32–37. https://doi.org/10.1016/j.psyneuen.2008.08.024 Klatzkin, R. R., Dasani, R., Warren, M., Cattaneo, C., Nadel, T., Nikodem, C., & Kissileff, H. R. (2019). Negative affect is associated with increased stress-eating for women with high perceived life stress. Physiology & Behavior, 210, 112639. https://doi.org/10.1016/j.physbeh.2019.112639 Klump, K. L., & Burt, S. A. (2006). The Michigan State University Twin Registry (MSUTR): Genetic, environmental, and neurobiological influences on behavior across development. Twin Research and Human Genetics, 9(6), 971–977. https://doi.org/10.1375/183242706779462868 Klump, K. L., Culbert, K. M., & Sisk, C. L. (2017). Sex Differences in Binge Eating: Gonadal Hormone Effects Across Development. Annu. Rev. Clin. Psychol, 13, 183–207. https://doi.org/10.1146/annurev-clinpsy-032816 Klump, K. L., Keel, P. K., Burt, S. A., Racine, S. E., Neale, M. C., Sisk, C. L., & Boker, S. (2013). Ovarian Hormones and Emotional Eating Associations across the Menstrual Cycle: An Examination of the Potential Moderating Effects of Body Mass Index and Dietary Restraint. International Journal of Eating Disorders, 46(3), 256–263. https://doi.org/10.1002/eat.22084 Klump, K. L., Keel, P. K., Culbert, K. M., & Edler, C. (2008). Ovarian hormones and binge eating: Exploring associations in community samples. Psychological Medicine, 38, 1749– 1757. https://doi.org/10.1017/S0033291708002997 71 Klump, K. L., Keel, P. K., Racine, S. E., Burt, S. A., Neale, M., Sisk, C. L., … Hu, J. Y. (2013). The interactive effects of estrogen and progesterone on changes in emotional eating across the menstrual cycle. Journal of Abnormal Psychology, 122(1), 131–137. https://doi.org/10.1037/a0029524 Koo-Loeb, J. H., Costello, N., Light, K. C., & Girdler, S. S. (2000). Women with eating disorder tendencies display altered cardiovascular, neuroendocrine, and psychosocial profiles. Psychosomatic Medicine, 62(4), 539–548. Kwan, M. Y., & Gordon, K. H. (2016). An examination of the prospective impact of bulimic symptoms and dietary restraint on life hassles and social support. Journal of Affective Disorders, 191, 24–28. la Fleur, S. E., Akana, S. F., Manalo, S. L., & Dallman, M. F. (2004). Interaction between corticosterone and insulin in obesity: regulation of lard intake and fat stores. Endocrinology, 145(5), 2174–2185. Langer, S. L., Soltero, E. G., Beresford, S. A., McGregor, B. A., Albano, D. L., Patrick, D. L., & Bowen, D. J. (2018). Socioeconomic status differences in food consumption following a laboratory-induced stressor. Health Psychology Open, 5(2), 2055102918804664. https://doi.org/10.1177/2055102918804664 Larsen, J. K., van Ramshorst, B., van Doornen, L. J. P., & Geenen, R. (2009). Salivary cortisol and binge eating disorder in obese women after surgery for morbid obesity. International Journal of Behavioral Medicine, 16(4), 311–315. Laugero, K. D., Falcon, L. M., & Tucker, K. L. (2011). Relationship between perceived stress and dietary and activity patterns in older adults participating in the Boston Puerto Rican Health Study. Appetite, 56(1), 194–204. Lavagnino, L., Amianto, F., Caprino, M. P., Maccario, M., Arvat, E., Ghigo, E., … Fassino, S. (2014). Urinary cortisol and psychopathology in obese binge eating subjects. Appetite, 83, 112–116. Lazarevich, I., Camacho, M. E. I., Velazquez-Alva, M. D. C., & Zepeda, M. Z. (2016). Relationship among obesity, depression, and emotional eating in young adults. Appetite, 107, 639–644. Lehrer, H. M., Goosby, B. J., Dubois, S. K., Laudenslager, M. L., & Steinhardt, M. A. (2020). Race moderates the association of perceived everyday discrimination and hair cortisol concentration. Stress, 23(5), 529–537. https://doi.org/10.1080/10253890.2019.1710487 Lei, H., & Skinner, H. A. (1980). A psychometric study of life events and social readjustment. Journal of Psychosomatic Research, 24, 57–65. 72 Levine, S. (2005). Stress: An historical perspective. Techniques in the Behavioral and Neural Sciences, 15, 3–23. https://doi.org/10.1016/S0921-0709(05)80004-5 Lo Sauro, C., Ravaldi, C., Cabras, P. L., Faravelli, C., & Ricca, V. (2008). Stress, hypothalamic- pituitary-adrenal axis and eating disorders. Neuropsychobiology, 57, 95–115. https://doi.org/10.1159/000138912 Loth, K., van den Berg, P., Eisenberg, M. E., & Neumark-Sztainer, D. (2008). Stressful life events and disordered eating behaviors: findings from Project EAT. Journal of Adolescent Health, 43(5), 514–516. Manenschiin, L., van Kruysbergen, R. G., de Jong, F. H., Koper, J. W., & van Rossum, E. F. (2011). Shift work at young age is associated with elevated long-term cortisol levels and body mass index. Journal of Clinical Endocrinology and Metabolism, 96(11), E1862-5. Manetti, L., Rossi, G., Grasso, L., Raffaelli, V., Scattina, I., Del Sarto, S., … Martino, E. (2013). Usefulness of salivary cortisol in the diagnosis of hypercortisolism: comparison with serum and urinary cortisol. European Journal of Endocrinology, 168(3), 315–321. Masheb, R. M., & Grilo, C. M. (2006). Emotional overeating and its associations with eating disorder psychopathology among overweight patients with binge eating disorder. International Journal of Eating Disorders, 39(2), 141–146. McEwen, B. S., & Akil, H. (2020). Revisiting the stress concept: Implications for affective disorders. Journal of Neuroscience, 40(1), 12–21. Micali, D., Martini, M. G., Thomas, J. J., Eddy, K. T., Kothari, R., Russell, E., … Treasure, J. (2017). Lifetime and 12-month prevalence of eating disorders amongst women in mid-life: a population-based study of diagnoses and risk factors. BMC Med, 15(1), 12. Minami, S., Satoyoshi, H., Ide, S., Inoue, T., Yoshioka, M., & Minami, M. (2017). Suppression of reward-induced dopamine release in the nucleus accumbens in animal models of depression: Differential responses to drug treatment. Neuroscience Letters, 650, 72–76. https://doi.org/10.1016/j.neulet.2017.04.028 Monteleone, A. M., Ruzzi, V., Pellegrino, F., Patriciello, G., Cascino, G., Del Giorno, C., … Maj, M. (2019). The vulnerability to interpersonal stress in eating disorders: The role of insecure attachment in the emotional and cortisol responses to the trier social stress test. Psychoneuroendocrinology, 101, 278–285. https://doi.org/10.1016/j.psyneuen.2018.12.232 Müller, T. D., Nogueiras, R., Andermann, M. L., Andrews, Z. B., Anker, S. D., Argenta, J., … Tschöp, M. T. (2015). Ghrelin. Molecular Metabolism, 4(6), 437–460. https://doi.org/10.1016/j.molmet.2015.03.005 Nollet, M., Wisden, W., & Franks, N. (2020). Sleep deprivation and stress: a reciprocal relationship. Interface Focus, 6(10), 20190092. 73 Noppe, G., Van Rossum, E. F., Koper, J. W., Manenschiin, L., Bruining, G. J., de Rijke, Y. B., & van den Akker, E. L. (2014). Validation and reference ranges of hair cortisol measurement in healthy children. Hormones Research in Paediatrics, 82, 97–102. O’Brien, K. M., Meyer, J., Tronick, E., & Moore, C. L. (2017). Hair cortisol and lifetime discrimination: Moderation by subjective social status. Health Psychology Open, 4(1), 2055102917695176. https://doi.org/10.1177/2055102917695176 O’Brien, K. M., Tronick, E. Z., & Moore, C. L. (2013). Relationship between hair cortisol and perceived chronic stress in a diverse sample. Stress and Health, 29(4), 337–344. https://doi.org/10.1002/smi.2475 Odeniyi, I. A., & Fasanmde, O. A. (2013). Urinary free cortisol in the diagnosis of Cushing’s syndrome: how useful? Nigerian Journal of Clinical Practice, 16(3), 269–272. Orem, T. R., Wheelock, M. D., Goodman, A. M., Harnett, N. G., Wood, K. H., Gossett, E. W., … Knight, D. C. (2019). Amygdala and prefrontal cortex activity varies with individual differences in the emotional response to psychosocial stress. Behavioral Neuroscience, 133(2), 203–211. https://doi.org/10.1037/bne0000305 Ouwens, M. A., van Strien, T., & van Leeuwe, J. F. L. (2009). Possible pathways between depression, emotional and external eating: A structural equation model. Appetite, 53(2), 245–248. Palmisano, G. L., Innamorati, M., & Vanderlinden, J. (2016). Life adverse experiences in relation with obesity and binge eating disorder: A systematic review. Journal of Behavioral Addictions, 5(1), 11–31. https://doi.org/10.1556/2006.5.2016.018 Parikh, T. P., Stolze, B., Ozarda, Y., Jonklaas, J., Welsh, K., Masika, L., … Soldin, S. J. (2018). Diurnal variation of steroid hormones and their reference intervals using mass spectrometric analysis. Endocrine Connections, 12, 1354–1361. Pecoraro, N., Reyes, F., Gomez, F., Bhargava A., & Dallman, M. F. (2004). Chronic stress promotes palatable feeding, which reduces signs of stress: feedforward and feedback effects of chronic stress. Endocrinology, 145, 3754–3762. Pecoraro, N., Reyes, F., Gomez, F., Bhargava, A., & Dallman, M. F. (2004). Chronic stress promotes palatable feeding, which reduces signs of stress: feedforward and feedback effects of chronic stress. Endocrinology, 145, 3754–3762. Piazza, P. V., Rouge-Pont, F., Deroche, V., Maccari, S., Simon, H., & Le Moal, M. (1996). Glucocorticoids have state-dependent stimulant effects on the mesencephalic dopaminergic transmission. Proceedings of the National Academy of Sciences, 93(16), 8716–8720. Pickett, S., McCoy, T. P., & Odetola, L. (2020). The Influence of Chronic Stress and Emotions 74 on Eating Behavior Patterns and Weight among Young African American Women. Western Journal of Nursing Research, 42(11), 894–902. https://doi.org/10.1177/0193945919897541 Pike, K. M., Wilfley, D., Hilbert, A., Fairburn, C. G., Dohm, F. A., & Striegel-Moore, R. H. (2006a). Antecedent life events of binge-eating disorder. Psychiatry Research, 142(1), 19– 29. https://doi.org/10.1016/j.psychres.2005.10.006 Pike, K. M., Wilfley, D., Hilbert, A., Fairburn, C. G., Dohm, F. A., & Striegel-Moore, R. H. (2006b). Antecedent life events of binge-eating disorder. Psychiatry Research, 142, 19–29. https://doi.org/10.1016/j.psychres.2005.10.006 Putignano, P., Toja, P., Dubini, A., Giraldi, P., Corsello, S. M., & Cavagnini, F. (2003). Midnight salivary cortisol versus urinary free and midnight serum cortisol as screening tests for Cushing’s syndrome. Journal of Clinical Endocrinology and Metabolism, 88(9), 4153– 4157. Rab, S. L., & Admon, R. (2020). Parsing inter- and intra-individual variability in key nervous system mechanisms of stress responsivity and across functional domains. Neuroscience and Biobehavioral Reviews, S0149-7634(20), 30564–30569. Racine, S. E., Culbert, K. M., Keel, P. K., Sisk, C. L., Burt, S. A., & Klump, K. L. (2012). Differential associations between ovarian hormones and disordered eating symptoms across the menstrual cycle in women. International Journal of Eating Disorders, 45(3), 333–344. https://doi.org/10.1002/eat.20941 Racine, S. E., Culbert, K. M., Larson, C. L., & Klump, K. L. (2009). The possible influence of impulsivity and dietary restraint on associations between serotonin genes and binge eating. Journal of Psychiatric Research, 43(16), 1278–1286. https://doi.org/10.1016/j.jpsychires.2009.05.002 Raspopow, K., Abizaid, A., Matheson, K., & Anisman, H. (2010). Psychosocial stressor effects on cortisol and ghrelin in emotional eaters: Influence of anger and shame. Hormones & Behavior, 58(4), 677–684. Raspopow, K., Abizaid, A., Matheson, K., & Anisman, H. (2014). Anticipation of a psychosocial stressor differentially influences ghrelin, cortisol and food intake among emotional and non- emotional eaters. Appetite, 74, 35–43. https://doi.org/10.1016/j.appet.2013.11.018 Ricca, V., Castellini, G., Lo Sauro, C., Ravaldi, C., Lapi, F., Mannucci, E., … Faravelli, C. (2009). Correlations between binge eating and emotional eating in a sample of overweight subjects. Appetite, 53(3), 418–421. Richardson, A. S., Arsenault, J. E., Cates, S. C., & Muth, M. K. (2015). Perceived stress, unhealthy eating behaviors, and severe obesity in low-income women. Nutrition Journal, 14, 122. 75 Rojo, L., Conesa, L., Bermudez, O., & Livianos, L. (2006). Influence of stress in the onset of eating disorders: data from a two-stage epidemiologic controlled study. Psychosomatic Medicine, 68(4), 628–635. Rossi, E., Cassioli, E., Gironi, V., Idrizaj, E., Garella, R., Squecco, R., … Castellini, G. (2021). Ghrelin as a possible biomarker and maintaining factor in patients with eating disorders reporting childhood traumatic experiences. European Eating Disorders Review, 29(4), 588– 599. https://doi.org/10.1002/erv.2831 Sander, M., Lois, G., Streit, F., Zeier, P., Kirsch, P., Wüst, S., & Wessa, M. (2020). Investigating individual stress reactivity: High hair cortisol predicts lower acute stress responses. Psychoneuroendocrinology, 118, 104660. Sapolsky, R. M. (2015). Stress and the brain: individual variability and the inverted-U. Nature Neuroscience, 18(10), 1344–1346. Schlotz, W., Kumsta, R., Layes, I., Entringer, S., Jones, A., & Wüst, S. (2008). Covariance between psychological and endocrine responses to pharmacological challenge and psychosocial stress: A question of timing. Psychosomatic Medicine, 70(7), 787–796. https://doi.org/10.1097/PSY.0b013e3181810658 Schulz, S., Laessle, R., & Hellhammer, D. (2011). No evidence of increased cortisol stress response in obese women with binge eating disorder. Eating and Weight Disorders, 16(3), e209-11. Short, S. J., Stalder, T., Marceau, K., Entringer, S., Moog, N. K., Shirtcliff, E. A., … Buss, C. (2016). Correspondence between hair cortisol concentrations and 30-day integrated daily salivary and weekly urinary cortisol measures. Psychoneuroendocrinology, 71, 12–18. https://doi.org/10.1016/j.psyneuen.2016.05.007 Smith, K. E., Mason, T. B., Schaefer, L., Anderson, L. M., Critchley, K., Crosby, R. D., … Peterson, C. B. (2020). Dynamic Stress Responses and Real Symptoms in Binge-Eating Disorder. Annals of Behavioral Medicine, Online ahead of print. https://doi.org/10.1093/abm/kaaa061 Smyth, J. M., Heron, K. E., Wonderlich, S. A., Crosby, R. D., & Thompson, K. M. (2008). The influence of reported trauma and adverse events on eating disturbance in young adults. International Journal of Eating Disorders, 41(3), 195–202. https://doi.org/10.1002/eat.20490 Smyth, J. M., Wonderlich, S. A., Heron, K. E., Sliwinski, M. J., Crosby, R. D., Mitchell, J. E., & Engel, S. G. (2007). Daily and momentary mood and stress are associated with binge eating and vomiting in bulimia nervosa in the natural environment. J Consult Clin Psychol, 75(4), 629–638. Smyth, J. M., Wonderlich, S. A., Sliwinski, M. J., Crosby, R. D., Engel, S. G., Mitchell, J. E., & 76 Calogero, R. M. (2009). Ecological momentary assessment of affect, stress, and binge- purge behaviors: Day of week and time of day effects in the natural environment. International Journal of Eating Disorders, 42(5), 429–436. https://doi.org/10.1002/eat.20623 Spinosa, J., Christiansen, P., Dickson, J. M., Lorenzetti, V., & Hardman, C. A. (2019). From Socioeconomic Disadvantage to Obesity: The Mediating Role of Psychological Distress and Emotional Eating. Obesity, 27(4), 559–564. https://doi.org/10.1002/oby.22402 Srivastava, P., Lampe, E. W., Michael, M. L., Manasse, S., & Juarascio, A. S. (2021). Stress appraisal prospectively predicts binge eating through increases in negative affect. Eating and Weight Disorders, Online ahead of print. https://doi.org/10.1007/s40519-020-01082-4 Stalder, T., & Kirschbaum, C. (2012). Analysis of cortisol in hair - State of the art and future directions. Brain, Behavior, and Immunity, 26(7), 1019–1029. https://doi.org/10.1016/j.bbi.2012.02.002 Stalder, T., Steudte-Schmiedgen, S., Alexander, N., Klucken, T., Vater, A., Wichmann, S., … Miller, R. (2017). Stress-related and basic determinants of hair cortisol in humans: A meta- analysis. Psychoneuroendocrinology, 77, 261–274. Steinsbekk, S., Barker, E. D., Llewellyn, C., Fildes, A., & Wichstrøm, L. (2018). Emotional Feeding and Emotional Eating: Reciprocal Processes and the Influence of Negative Affectivity. Child Development, 89(4), 1234–1246. https://doi.org/10.1111/cdev.12756 Stice, E., Presnell, K., & Spangler, D. (2002). Risk factors for binge eating onset in adolescent girls: a 2-year prospective investigation. Health Psychology, 21(2), 131–138. Stievenard, A., Méquinion, M., Andrews, Z. B., Destée, A., Chartier-Harlina, M., Viltart, O., & Vanbesien-Mailliota, C. C. (2017). Is there a role for ghrelin in central dopaminergic systems? Focus on nigrostriatal and mesocorticolimbic pathways. Neurosicence and Biobehavioral Reviews, 73, 255–275. https://doi.org/10.1016/j.neubiorev.2016.11.021 Stone, L. A., Harmatz, E. S., & Goosens, K. A. (2020). Ghrelin as a stress hormone: Implications for psychiatric illness. Biological Psychiatry, 88, 531–540. https://doi.org/10.1016/j.biopsych.2020.05.013 Streit, F., Memic, A., Hasandedić, L., Rietschel, L., Frank, J., Lang, M., … Rietschel, M. (2016). Perceived stress and hair cortisol: Differences in bipolar disorder and schizophrenia. Psychoneuroendocrinology, 69, 26–34. https://doi.org/10.1016/j.psyneuen.2016.03.010 Tanofsky-Kraff, M., Thiem, K. R., Yanovski, S. Z., Bassett, A. M., Burns, N. P., Ranzenhofer, L. M., … Yanovski, J. A. (2007). Validation of the emotional eating scale adapted for use in children and adolescents (EES-C). International Journal of Eating Disorders, 40(3), 232– 240. 77 Tenk, J., Mátrai, P., Hegyi, P., Rostás, I., Garami, A., Szabó, I., … Balaskó, M. (2018). Perceived stress correlates with visceral obesity and lipid parameters of the metabolic syndrome: A systematic review and meta-analysis. Psychoneuroendocrinology, 95, 63–73. Thurston, I. B., Hardin, R., Kamody, R. C., Herbozo, S., & Kaufman, C. (2018). The moderating role of resilience on the relationship between perceived stress and binge eating symptoms among young adult women. Eating Behaviors, 29, 114–119. Tomiyama, A. J., Dallman, M. F., & Epel, E. S. (2011). Comfort food is most comforting to those most stressed: Evidence of the chronic stress response network in high stress women. Psychoneuroendocrinology, 36(10), 1513–1519. van Strien, T. (2000). Ice-cream consumption, tendency toward overeating, and personality. International Journal of Eating Disorders, 28(4), 460–464. van Strien, T., Engels, R. C., van Leeuwe, J., & Snoek, H. M. (2005). The Stice model of overeating: tests in clinical and non-clinical samples. Appetite, 45(3), 205–213. van Strien, T., Frihters, J. E., Bergers, G. P., & Defares, P. B. (1986). The Dutch Eating Behavior Questionnaire (DEBQ) for assessment of restrained, emotional, and external eating behavior. International Journal of Eating Disorders, 5, 295–315. van Strien, T., Roelofs, K., & de Weerth, C. (2013). Cortisol reactivity and distress-induced emotional eating. Psychoneuroendocrinology, 38(5), 677–684. https://doi.org/10.1016/j.psyneuen.2012.08.008 van Strien, T., Winkens, L., Toft, M. B., Pedersen, S., Brouwer, I., Visser, M., & Lahteenmaki, L. (2016). The mediation effect of emotional eating between depression and body mass index in the two European countries Denmark and Spain. Appetite, 105, 500–508. Viardot, A., Huber, P., Puder, J. J., Zulewski, H., Keller, U., & Müller, B. (2005). Reproducibility of nighttime salivary cortisol and its use in the diagnosis of hypercortisolism compared with urinary free cortisol and overnight dexamethasone suppression test. Journal of Clinical Endocrinology and Metabolism, 90(10), 5730–5736. Wardle, J. (1987). Eating style: A validation study of the Dutch Eating Behaviour Questionnaire in normal subjects and women with eating disorders. Journal of Psychosomatic Research, 31(2), 161–169. Watson, D., Clark, L. A., & Tellegen, A. (1988). Development and validation of brief measures of positive and negative affect: The PANAS scales. Journal of Personality and Social Psychology, 5, 1063–1070. Wennig, R. (2000). Potential problems with the interpretation of hair analysis results. Forensic Sci Int., 107, 5–12. 78 Wester, V. L., Staufenbiel, S. M., Veldhorst, M. A. B., Visser, J. A., Manenschijn, L., Koper, J. W., … van Rossum, E. F. C. (2014). Long-term cortisol levels measured in scalp hair of obese patients. Obesity (Silver Spring), 22(9), 1956–1958. Wolff, G. E., Crosby, R. D., Roberts, J. A., & Wittrock, D. A. (2000). Differences in daily stress, mood, coping, and eating behavior in binge eating and nonbinge eating college women. Addictive Behaviors, 25(2), 205–216. https://doi.org/10.1016/s0306-4603(99)00049-0 Woods, A. M., Racine, S. E., & Klump, K. L. (2010). Examining the relationship between dietary restraint and binge eating: Differential effects of major and minor stressors. Eating Behaviors, 11, 276–280. https://doi.org/10.1016/j.eatbeh.2010.08.001 Wright, K. D., Ford, J. L., Perazzo, J., Jones, L. M., Mahari, S., Sullenbarger, B. A., & Laudenslager, M. L. (2018). Collecting Hair Samples for Hair Cortisol Analysis in African Americans. Journal of Visualized Experiments, (136), 57288. Zelkowtiz, R. L., Zerubavel, N., Zucker, N. L., & Copeland, W. E. (2021). Longitudinal associations of trauma exposure with disordered eating: lessons from the great smoky mountain study. Eating Disorders, 19, 1–18. https://doi.org/10.1080/10640266.2021.1921326 79