The purpose of this study was to examine models by which internalizing symptoms of depression and anxiety influence food intake in overweight/obese women. Our results indicate that BED women endorse significantly more symptoms of depression and anxiety. Additionally, linear regression indicated that BED diagnosis and internalizing symptoms accounted for 30% of the variance in kcal intake. Furthermore, results from path analysis imply that BED mediates the influence of internalizing symptoms on total kcal-intake, which suggests the associations between internalizing symptoms and food intake are best described as operating indirectly through a BED diagnosis.
The present study found that overweight/obese women with BED endorsed more symptoms of depression and anxiety than non-BED weight-matched controls. Mean scores for the BDI and the BAI indicated mild depression and anxiety in the BED group but normal levels in the control group. Other studies have found elevated depression and anxiety scores in BED individuals [5
]. For example, in a study by Fandiño et al., depression and anxiety scores were significantly greater in the BED group than those in the obese control group as assessed by the Symptom Checklist 90 and the BDI [37
]. The lifetime prevalence of MDD in the BED and control groups was 46.7% and 29.4%, respectively. These rates are similar to previous reports in BED [2
] and non-BED obese groups [15
]. Lifetime prevalence of anxiety disorders was similar to rates of depression in the BED group (46.7%) but was much lower in the control group (11.8%). It is possible that the lower rates of anxiety disorders in the control group were due to the inclusion of overweight women or was an artifact of the limited sample size.
Laboratory studies have demonstrated that those with BED have greater total food intake than obese controls when instructed to overeat [5
]. Two such studies have reported on both food intake and depression symptoms [5
]. In a sample of 10 obese BED women and 9 obese controls, Yanovski et al. found that the BED group consumed significantly more kcals (2962 versus 2017) and had significantly greater depression scores as measured by the BDI (18.9 versus 5.4) than controls. Additionally, they observed significant positive correlations between kcal intake and BDI score (r2
= 0.41) and between binge meal energy intake and BDI (r2
= 0.28). Geliebter and colleagues compared consumption of a liquid test-meal for 30 obese BED individuals (18 women) and 55 obese controls (43 women). The BED group consumed significantly more grams (1,032 versus 737) of the liquid test meal and endorsed significantly higher depression scores assessed by the Zung Depression Scale. However, a significant correlation between test meal intake and depression score was not found. The discrepancy could be due to several study design differences, including proportion of BED and control participants, inclusion of men, and type of food intake (solid versus liquid meal).
Furthermore, results from linear regression indicated that BED diagnosis and symptoms of depression and anxiety accounted for a significant amount (~30%) of the variance in caloric intake. However, examination of the three alternative models revealed important mechanistic differences in the relationship between BED, symptoms of depression, and anxiety and subsequent energy intake. The model that best fit our data indicated that BED mediated the influence of depression and anxiety symptoms on total kcal intake (). Specifically, our results suggest that the associations found between symptoms of depression and anxiety and food intake are best described as operating indirectly through a BED diagnosis, that is, symptoms of depression and anxiety influence whether one engages in pathological binge eating, which, in turn, influences caloric intake. Our findings did not support model b (BED predicted symptoms of depression and anxiety which, in turn, influence kcal intake), or model c (a significant interaction between symptoms of depression and anxiety and BED as being predictive of kcal intake).
These results highlight the importance of mood in relation to a BED diagnosis and subsequent caloric intake. Other research has also implicated mood in BED. Telch et al. interviewed 60 obese women with BED regarding their definition of binge eating, and 33% reported it as eating to regulate negative affect [38
]. With the advent of the ecological momentary assessment procedures (EMA), prospective data on precursors to binge eating in the natural environment have been collected [39
]. A study by Stein et al. found in 33 obese women with BED that negative mood was significantly greater at prebinge times than at nonbinge times and that participants attributed binge eating to mood more frequently than hunger or violation of extreme dietary restraint (abstinence violation) [41
]. Additionally, a study by Hilbert and Tuschen-Caffier found that mood preceding a binge eating episode was more negative than mood prior to regular eating or at random assessments in a sample of 20 obese women with BED [42
]. Furthermore, in a meta-analysis of 36 EMA studies of BED and BN, negative affect was significantly greater preceding binge eating relative to average affect and affect before regular eating [43
]. A growing body of literature implicates negative affect as a precursor to binge eating in BED.
The implications of the present study are potentially relevant to the clinical treatment of BED and obesity. Research has indicated that mood and eating disorder diagnoses affect weight loss, and other treatment efforts. For example, Pagoto et al. reported that both BED and depression were associated with less weight loss and depression was associated with study attrition [44
]. Furthermore, in BED treatment, depression symptoms have been associated with both attrition from cognitive-behavioral therapy and severity of eating disorder psychopathology [45
]. The current results suggest that targeting mood may be useful in the treatment of BED and accentuate the importance of considering mood and BED status in weight management.
Among the major strengths of this study were utilizing path analysis to test relationships between BED, symptoms of depression and anxiety, and kcal intake as well as using permutation procedures for model validation and statistical support. EMA studies have consistently demonstrated negative affect as a precursor to binge eating [43
] in BED. However, these studies have relied on self-report of food intake. Research indicates that obese and BED populations tend to underreport their food intake [13
]. Therefore, a further strength of this work was the inclusion of laboratory-measured food intake to avoid inaccuracies often associated with self-report of dietary intake. Potential limitations include limited sample size and age range, exclusion of male participants, and use of self-report questionnaires to measure symptoms of depression and anxiety. Future research is warranted to confirm our findings and should seek to compare energy intake and depression and anxiety in both women and men. Greater understanding of the mechanisms underlying the associations of depression and anxiety symptoms, binge eating, and caloric intake will facilitate the development of more effective prevention and treatment strategies for both BED and obesity.