نقش واسطه از اختلال در نظم احساسات و افسردگی در رابطه با مواجهه با ترومای دوران کودکی و غذا خوردن احساسی

Exposure to childhood adversity is implicated in the etiology of adverse health outcomes, including depression, posttraumatic stress disorder (PTSD), and obesity. The relationship between childhood trauma and obesity may be related to the association of childhood trauma and risk for emotional eating. One pathway between trauma exposure, psychopathology, and emotional eating may be through emotion dysregulation and depression. The current study was undertaken to characterize demographic, environmental, and psychological risk factors for emotional eating in a primarily African American, low socioeconomic status (SES), inner-city population (N=1110). Emotional eating was measured using the Dutch Eating Behavioral Questionnaire and the Emotional Dysregulation Scale was used to assess emotion regulation. The Beck Depression Inventory and the modified PTSD Symptom Scale were used to assess depression and PTSD, respectively. Higher levels of emotional eating were associated with body mass index, income, childhood and adulthood trauma exposure, depressive and PTSD symptoms, negative affect and emotion dysregulation. Childhood emotional abuse was the most associated with emotional eating in adulthood. Hierarchical linear regression and mediation analyses indicated that the association between childhood trauma exposure (and emotional abuse specifically) and emotional eating was fully mediated by depression symptoms and emotion dysregulation, with emotional dysregulation contributing more to the mediation effect. Together these findings support a model in which obesity and related adverse health outcomes in stress- and trauma-exposed populations may be directly related to self-regulatory coping strategies accompanying emotion dysregulation. Our data suggest that emotion dysregulation is a viable therapeutic target for emotional eating in at-risk populations.

Exposure to childhood adversity has shown to be robustly related with an array of adverse developmental and psychological outcomes in a dose-dependent manner (Cicchetti et al, 2010, Cicchetti, Toth, 2005 and Manly et al, 2001), including increased risk for developing psychopathology (Heim, Nemeroff, 2001 and Kessler et al, 1997) and eating disorders in adulthood (Burns et al, 2012 and Kong, Bernstein, 2009). Maltreated children go on to develop depression, posttraumatic stress disorder (PTSD), and substance use disorders at higher rates than non-maltreated children (Brown et al, 1999, Horwitz et al, 2001 and Kaplow, Widom, 2007). Depression, PTSD, and childhood maltreatment are also highly comorbid with obesity (de Wit et al, 2010, Hemmingsson et al, 2014, Pagoto et al, 2012 and Rosmond, 2004) that may result from increased emotional eating (Talbot, Maguen, Epel, Metzler, & Neylan, 2013). Therefore, understanding the factors that lead to emotional eating is critical to the development of interventional strategies that reduce the health burden imposed by obesity and its adverse consequences. One risk factor for the development of adverse physical and mental health conditions among individuals exposed to childhood adversity is a deficit in adaptive coping strategies that can be deployed in response to environmental stressors (Pollak, 2008 and Shields et al, 1994). Instead, individuals with adverse childhood experiences are at increased risk for developing maladaptive coping strategies, including stress-induced emotional eating (Evers, Marijn Stok, & de Ridder, 2010). One reason this might occur is because of impairment in emotion regulation. Emotion regulation is the critical ability to modulate and maintain feelings, behaviors, and physiological responses that constitute an emotion (Gross, 2002). Extensive research now shows that poor emotion regulation contributes to an array of psychiatric conditions, including depression and PTSD (Aldao et al, 2010 and Bradley et al, 2011), in a manner distinct from negative affect alone (Bradley, et al., 2011). Emotion regulation difficulties are often found in individuals exposed to trauma, particularly childhood abuse (Pollak, 2008 and Shields et al, 1994), and are found to have long-term detrimental effects on expression, recognition, and communication of emotion, and the use of strategies for managing strong emotions (Southam-Gerow & Kendall, 2002). While research has shown a relationship between emotion dysregulation and risk for psychological disorders including eating-related psychopathology (Moulton, Newman, Power, Swanson, & Day, 2014), it remains unclear whether emotion dysregulation is associated with emotional eating. It is also unclear whether that relationship would exist above and beyond other demographic, environmental, and psychological risk factors for emotional eating. Furthermore, research comprehensively examining the relationship between the various factors that contribute to emotional eating (e.g., demographics and psychological characteristics) is limited. Thus, the goal of the current study was to examine the relationship between childhood maltreatment, other types of trauma, symptoms of PTSD and depression, and emotion dysregulation, on emotional eating using linear regression and mediation analyses. We hypothesized that childhood maltreatment, depression, and emotion dysregulation would all be associated with increased emotional eating, and that the relationship between childhood maltreatment and emotional eating would be mediated by depressive symptoms, PTSD symptoms, and emotion dysregulation. Furthermore, we assessed the role of different types of childhood adversity (sexual, physical, and emotional abuse and neglect) on risk for emotional eating, as previous reports indicate that childhood emotional abuse is strongly associated with eating disorder pathology in adulthood (Burns et al, 2012 and Kong, Bernstein, 2009). Overall, our analyses intended to identify demographic, environmental, and psychological factors for emotional eating in a low socioeconomic status (SES), primarily African American population that is exposed to greater rates of stress and trauma than the general population and has exacerbated risk for depression, PTSD, and obesity (Gillespie et al, 2009 and Harrell, Gore, 1998).

Demographic characteristics and relationship to emotional eating. The demographic information for the overall sample is summarized in Table 1. Most subjects in the current study self-identified their race as African-American (Table 1). The majority of the sample was female (80.4%), unemployed (68.4%), and approximately a third had achieved only a high school level education (Table 1). The socioeconomic status of the majority of the subjects was very low, with 79.2% of the sample having a mean monthly household income of less than $2,000 and 16.6% of the sample having a mean monthly household income of less than $250 (Table 1). Mean ± SEM age (in years) and BMI in participants are also described in Table 1. Table 1. Demographic characteristics of sample in frequency (N), percent (%) and mean ± SEM. Association with emotional eating severity (DEBQ total score) is indicated in P-value column. Asterisks denote demographic characteristics that influence emotional eating (all p≤0.05). Bivariate correlation analyses indicated that BMI (r=0.13; p<0.001) and age (r=0.06; p=0.05) were significantly associated with emotional eating. ANOVA results showed that of the demographic variables examined, only employment (F1,1109=7.26; p=0.007) and monthly income (F4,1087=3.76; p=0.005) were significantly associated with emotional eating. Unemployed individuals reported greater emotional eating than employed individuals (16.6±0.52 vs. 14.1±0.77, respectively). Individuals with a monthly income less than $250 reported greater emotional eating compared to individuals making more than $250 a month (19.5±0.1.07 vs. 14.4±1.55, respectively). Environmental exposures and association with emotional eating. A summary of environmental factors and how they influence emotional eating are summarized in Table 2. Bivariate correlation analyses (Table 2) indicated that emotional eating was significantly associated with childhood sexual abuse (r=0.13; p<0.001), physical abuse (r=0.09; p=0.002), physical neglect (r=0.15; p<0.001), emotional abuse (r=0.18; p<0.001), emotional neglect (r=0.15; p<0.001), and overall childhood trauma exposure (r=0.17; p<0.001). Additionally, adult trauma exposure (experienced or witnessed, as determined by the TEI, was significantly associated with emotional eating (r=0.14; p<0.001). Table 3 describes the inter-correlations between these environmental factors and psychological factors described below. Table 2. Mean ± SEM of environmental exposure characteristics of sample. Association with emotional eating (DEBQ total score) is indicated in P-value column. Asterisks denote environmental characteristics that influence emotional eating (all p≤0.05). Table 3. Inter-correlations between childhood maltreatment (overall and subscales as determine by the CTQ), PTSD and depression symptoms, emotional dysregulation and negative affect. Asterisk (*) denote significant correlation between factors. Psychological characteristics and relationship to emotional eating. The psychopathological information for the overall sample is summarized in Table 4. Approximately a third of the sample (28.2%) met for a current PTSD diagnosis (Table 4) using the mPSS. Similarly, almost a third of the sample (31.5%) met for a current depression diagnosis as determined by the BDI (Table 4). Although quite high, these prevalence rates are consistent with prior reports from our group and others studying highly traumatized, low SES populations (Gillespie, et al., 2009). ANOVA results indicated that independent diagnoses of current PTSD (F1,1062=40.17; p<0.001) and depression (F1,1107=67.16; p<0.001) were significantly associated with increased emotional eating (Table 4). Individuals with PTSD reported greater emotional eating than individuals without PTSD (20.3±0.82 vs. 14.2±0.51, respectively). Individuals who met for a depression diagnosis engaged in more severe emotional eating compared to individuals without a depression diagnosis (20.9±0.75 vs. 13.54±0.51, respectively). Table 4. Description of depression and PTSD symptoms within the study sample in frequency (N), percent (%) and mean ± SEM. Association with emotional eating (DEBQ total score) is indicated in P-value column. Asterisks denote psychopathological characteristics that influence emotional eating (all p≤0.05). Bivariate correlation analyses also showed that current PTSD (r=0.23; p<0.001) and depression (r=0.29; p<0.001) symptoms were significantly associated with self-reported emotional eating (Table 4). Additionally, emotional eating was significantly associated with negative affect as determined by the PANAS (Table 5; p<0.001) and emotion dysregulation as assessed by the EDS (Table 5; p<0.001). Table 5. Mean ± SEM of underlying psychological characteristics of sample. Association with emotional eating (DEBQ total score) is indicated in P-value column. Asterisks denote psychological characteristics that influence emotional eating (all p≤0.05). Prediction of emotional eating via hierarchical linear regression. All demographic, environmental, and psychological factors that were significantly associated with increased emotional eating (Table 1, Table 2, Table 3, Table 4 and Table 5) were included in a hierarchical linear regression model to predict emotional eating. A linear regression model was performed, with demographic factors (BMI, age, employment and income) in step one, environmental factors (total childhood trauma and total adult trauma experienced and witnessed) in step two, psychopathological factors (current PTSD and depression symptoms) in step three, and underlying psychological factors (emotional dysregulation and negative affect score) in step four. As shown in Table 6, BMI (β=0.11, p<0.001) and monthly household income (β=-0.07, p=0.04) were predictive of emotional eating. When environmental risk factors were entered into the model in step two, both childhood trauma (β=0.12, p<0.001) and adult trauma (β=0.08, p=0.02) were significantly predictive of emotional eating. Upon entering psychopathological risk factors into step three of the model, childhood and adult trauma were no longer significant and depressive symptoms significantly predicted emotional eating (β=0.22, p<0.001), suggesting that depressive symptoms may mediate the relationship between overall trauma exposure and emotional eating. In the final step, which included underlying psychological components of psychopathology, both emotional dysregulation (β=0.15, p<0.001) and negative affect (β=0.12, p=0.003) were significantly predictive of emotional eating while depressive symptoms were no longer significant (p>0.05), suggesting that negative affect and/or emotion dysregulation may mediate the association between depressive symptoms and emotional eating. Table 6. Emotional eating as predicted by demographic (Step 1), environmental (Step 2), psychopathological (Step 3), and psychological (Step 4) factors. Childhood trauma exposure in Step 2 was captured by total score on the CTQ. We reran the same analyses, but this time including subscales of childhood maltreatment as determined by the CTQ (sexual abuse, physical abuse, emotional abuse, emotional neglect, and physical neglect), to assess the effects of particular kinds of childhood trauma on emotional eating (Table 7). Results were overall similar as those described in Table 5. However, childhood emotional abuse (β=0.16, p=0.003) was the only type of childhood maltreatment that was significantly predictive of emotional eating in Step 2. The addition of emotional dysregulation (β=0.14, p=0.001) and negative affect (β=0.11, p=0.005) to the model indicated that both of these factors mediated the effect of childhood emotional abuse on emotional eating, as depressive symptoms were no longer significant (p>0.05). Table 7. Emotional eating as predicted by demographic (Step 1), environmental (Step 2), psychopathological (Step 3), and psychological (Step 4) factors. Childhood trauma exposure in Step 2 is now comprised of individual sub-clusters of childhood abuse and neglect as determined by the CTQ. Multiple mediation analyses. Because of the results found in the linear regression models, we formally tested the mediation effects of depression symptoms and emotional d ysregulation on the association between overall childhood maltreatment exposure and emotional eating using a multiple mediation analysis. As expected, there was a significant full mediation effect of depression (p=0.007; bias-corrected 95% confidence interval: 0.003, 0.03) and emotional dysregulation (p<0.0001; bias-corrected 95% confidence interval: 0.01, 0.05) on the relationship between childhood trauma exposure and emotional eating (F = 15.3, p<0.0001; Figure 1A), when controlling for BMI, age, income, employment, adult trauma exposure and PTSD symptoms. Full-size image (8 K) Figure 1. Multiple Mediation Analysis of Predictors of Emotional Eating. (A) The full mediating effects of adult depression symptoms (BDI) and emotional dysregulation (EDS) on the relation between total childhood trauma exposure (CTQ) and emotional eating while controlling for BMI, age, income, adult trauma exposure (TEI), PTSD symptoms (PSS), and negative affect (PANAS negative). (B) The full mediating effects of adult depression symptoms (BDI) and emotional dysregulation (EDS) on the relation between childhood emotional abuse (CTQ) and emotional eating while controlling for BMI, age, income, adult trauma exposure (TEI), PTSD symptoms (PSS), and negative affect (PANAS negative). Figure options We also formally tested the mediation effects of depression symptoms and emotional dysregulation on the association between childhood emotional abuse and emotional eating using a multiple mediation analysis, as emotional abuse was most associated with emotional eating in the secondary linear regression analysis. There was a significant full mediation effect of depression (p=0.008; bias-corrected 95% confidence interval: 0.01, 0.12) and emotional dysregulation (p<0.0001; bias-corrected 95% confidence interval: 0.06, 0.17) on the relationship between childhood emotional abuse and emotional eating (F = 15.6, p<0.0001; Figure 1B), when controlling for BMI, age, income, employment, adult trauma exposure and PTSD symptoms. Once again, the contribution of emotional dysregulation was more significant than that of depression on mediating the relationship between childhood emotional abuse and emotional eating. Discussion The current data describe demographic, environmental, and psychological risk factors for emotional eating in a primarily African American, low SES population. BMI, monthly income, child and adult trauma exposure, depression and PTSD symptoms, and negative affect and emotional dysregulation were all associated with emotional eating. Importantly, childhood emotional abuse was the most predictive of emotional eating compared to other forms of childhood maltreatment assessed. Regression and mediation analyses indicated that the positive association between childhood trauma exposure and emotional eating was fully mediated by current depression symptoms and emotion dysregulation. Overall, our comprehensive results extend previous findings describing risk factors for emotional eating by concurrently examining the relationship between demographic, environmental, and psychological factors that contribute to emotional eating. Low monthly income levels were associated with increased emotional eating, corroborating previous reports indicating that low income level is a vulnerability factor for an array of diet-dependent diseases that are common in low SES populations (Leung, Epel, Ritchie, Crawford, & Laraia, 2014), including obesity, type II diabetes and cardiovascular disease (Larson, Story, 2011 and Seligman et al, 2010). Increased intake of highly palatable foods in individuals of low income may be linked to low diet quality (Wolongevicz, et al., 2010) due to increased cost of healthy foods (Horning & Fulkerson, 2014). However, low income is also associated with food insecurity, a form of psychosocial stress that is related to increased intake of palatable foods (Leung, et al., 2014). While we did not measure food insecurity as a form of psychosocial stress directly in the current study, the role of psychosocial stressor exposure in susceptibility to emotional eating was further supported by our findings indicating that childhood and adult trauma exposure were significantly associated with emotional eating in adulthood. Studies in humans, monkeys and rodents indicate that both acute and chronic exposure to psychosocial stressors can precipitate stress-induced emotional eating (Adam, Epel, 2007 and Michopoulos et al, 2012) and lead to increases in weight (Arce et al, 2010, Bjorntorp, 2001, Dallman et al, 2005 and Rosmond, 2004). Additionally, perceived stress levels are associated with emotional eating of palatable foods in humans (R. Sims, et al., 2008). While exposure to adult trauma was associated with emotional eating in the current sample, exposure to childhood trauma was more significantly associated with emotional eating, providing a potential mechanism by which exposure to stressors during childhood leads to obesity in adulthood (Hemmingsson, et al., 2014). Childhood emotional abuse was the most associated with emotional eating, a finding that is consistent with previous reports indicating that this form of childhood abuse and neglect contributes most to psychological etiology of emotional eating (Burns et al, 2012 and Kong, Bernstein, 2009). Importantly, our analyses indicated that depressive symptoms and emotional dysregulation fully mediated the relationship between childhood trauma exposure and emotional eating. The contribution of emotional dysregulation was more significant than that of depression on mediating the relationship between overall childhood maltreatment and emotional eating. This result was expected, as the inability to regulate negative and intense emotions has been characterized as a consequence of psychosocial stress exposure, including childhood abuse (Pollak, 2008 and Shields et al, 1994), as well as a characteristic trait of depression and other psychopathology (Aldao et al, 2010 and Bradley et al, 2011). Childhood adversity also increases individual risk for both emotion dysregulation and for developing psychopathology in adulthood (Heim, Nemeroff, 2001 and Kessler et al, 1997), and depression is highly comorbid with obesity (de Wit et al, 2010, Pagoto et al, 2012 and Rosmond, 2004). Thus, our data suggest that emotional eating in individuals who have been exposure to childhood adversity, and more specifically childhood emotional abuse, is due primarily to difficulties in emotion regulation, as well as depressive symptoms. This notion is supported by recent evidence indicating that deficits in executive function, such as reduced behavioral inhibition, are associated with negative eating behaviors, including intake of palatable foods, and obesity (R. C. Sims, et al., 2014). Future studies are necessary to delineate which aspects of emotional regulation are critical for engaging in emotional eating, as they might serve as important targets for therapeutic strategies aimed at attenuating emotional eating severity in at-risk and vulnerable populations, such as those exposure to childhood emotional abuse.