Examining Binge Eating Disorder and Food Addiction in Adults with Overweight and Obesity

Objective

To compare four subgroups of adults with overweight/obesity: those with binge eating disorder (BED) only, food addiction (FA) only, both BED+FA, and neither.

Methods

502 individuals with overweight/obesity (BMI>25) completed a web-based survey with established measures of eating and health-related behaviors. Most were female (n=415; 83.2%) and White (n=404; 80.8%); mean age and BMI were 38.0(SD=13.1) years and 33.6(SD=6.9) kg/m^{, respectively.}

Results

Among 502 participants with overweight/obesity, 43(8.5%) met BED criteria, 84(16.6%) met FA criteria, 51(10.1%) met both BED+FA criteria, and 328(64.8%) met neither (Control). The three groups with eating pathology (BED, FA, and BED+FA) had significantly greater disturbances on most measures (eating-disorder psychopathology, impulsivity, and self-control) than the Control group, while the FA and BED+FA groups reported significantly higher depression scores relative to the Control group. The three eating groups did not differ significantly from each other.

Conclusion

In this online survey, of those with overweight/obesity, nearly one-third met criteria for BED, FA, or BED+FA and these forms of disordered eating were associated with greater pathology relative to individuals with overweight/obesity without BED and FA. Future research should examine whether the presence of BED, FA, or co-occurring BED+FA require tailored interventions in individuals with overweight or obesity.

Introduction

Obesity is prevalent (1,2) and is considered one of the leading (3) and costliest (4) causes of morbidity and mortality in the United States and worldwide (5). Obesity is a heterogeneous problem and this includes a diverse range of problematic patterns of eating (6). Research has increasingly highlighted the importance of better understanding problematic eating behaviors, and particularly binge eating disorder (BED), which is associated strongly with obesity in the United States (7) and worldwide (8).

BED is characterized by recurrent binge-eating episodes (eating an unusually large quantity of food in a discrete period of time coupled with a subjective sense of loss of control while eating), lack of compensatory behaviors, marked levels of distress, and three of five associated features (i.e., speed of eating, embarrassment while eating, eating when not physically hungry, eating until uncomfortably full, and guilt, depression, or disgust associated with binge-eating episodes) (9). The presence of BED in individuals with obesity is associated with significant increased risk for psychosocial, psychiatric, and medical problems (7,8,10). While BED is currently categorized as a formal eating disorder in the Diagnostic and Statistical Manual of Mental Disorders, 5 ^edition (DSM-5) (9), some individuals with BED describe binge eating as a form of “food addiction” (FA) (11). Indeed, behavioral markers of BED mirror defining-features of substance use disorders or traditional addictions (12,13). BED and “food addiction”, as currently conceptualized and measured, appear to overlap and share many similarities, the exact nature of which remains uncertain. For example, individuals who engage in binge-eating behavior often consume larger amounts of food than intended and experience strong cravings for food (11,12). Whether binge eating constitutes addictive eating, however, is still a matter of debate.

Emerging research suggests that binge eating and food addiction, while sharing many commonalities, might have important distinctions (14). Preliminary evidence from both specialty (15) and primary care (16) settings suggest that treatment-seeking patients with obesity and BED report high frequencies of FA, and that the co-occurrence of BED and FA may represent a more disturbed BED subgroup. Among individuals seeking treatment for BED and obesity in a specialty clinic, over half met criteria for FA, and those with FA reported greater levels of depression, negative affect, emotion dysregulation, eating disorder psychopathology, and lower self-esteem than those without FA (15). These findings parallel results from a primary care study of treatment-seeking individuals with obesity and BED although rates of FA were slightly lower in the primary care (41.5%) than the specialty clinic (57%) settings. Notably, in both studies (15,16), FA was associated with binge-eating frequency.

Less is known, however, about the overlapping and non-overlapping nature of BED and FA in persons with excess weight and this is particularly the case in non-treatment seeking samples. One community-based study (17) directly compared individuals with obesity and FA to individuals with obesity only (without FA). Of the 72 participants with obesity, 18 (25%) met criteria for FA and of those, 72.2% also met criteria for BED. The group with co-occurring FA and obesity was significantly more likely to meet criteria for severe depression, childhood ADHD, impulsivity, and addictive personality traits than the non-FA group with obesity. Notably, however, both the FA and non-FA groups with obesity included individuals with BED, with a significantly higher proportion of individuals with BED in the FA group relative to the non-FA group. Importantly, it is unclear how the co-occurrence of BED and FA influenced these results. A second study with community participants (18) examined the relationship among FA, BED, bulimia nervosa, and body mass index (BMI) across the entire weight spectrum and found that FA was related to greater BMI, binge-eating behaviors, and associated eating-disorder psychopathology. To further examine these eating issues in persons with excess weight, it would be important to compare BED without food addiction, food addiction without BED, and concurrent BED and FA relative to those with neither. Moreover, such studies would need to include measures of non-eating disordered constructs that might have clinical significance. For example, it would seem important to examine difficulties with impulsivity and self-control.

In summary, BED and food addiction appear to overlap and to be common in individuals with excess weight. The present study compared four groups of individuals with excess weight: those with BED but not FA, those with FA but not BED, those with both, and those with neither. We compared these four groups on measures of eating pathology, impulsivity, and self-control in light of recent neurological findings and correlates of BED (19–21).

Methods

Participants

Participants were 502 individuals who participated in an online study seeking adults aged 18 years or older to complete a survey of eating and health-related behaviors. Participants who met criteria for overweight/obesity (BMI > 25) were included in the present study. Participants included 84 (16.8%) males and 415 (83.2%) females (3 were missing); race/ethnicity was 80.8% (n=404) White, 5.4% (n=27) Hispanic, 8.6% (n=43) Black, 2.4% (n=12) Asian, 2.8% (n=14) “other,” and 2 missing. Mean age and BMI were 38.0 (SD=13.1) years and 33.6 (SD=6.9) kg/m^{, respectively. Of the 502 participants, 59.6% (n=299) met criteria for obesity and 40.4% (n=203) met criteria for overweight.}

Procedures and Assessments

Advertisements were placed on Craiglists Internet ads. Participants completed an anonymous online survey consisting of demographic information, self-reported height and weight, and self-report questionnaires through SurveyMonkey, a secure online data-gathering platform. The study was approved by the Yale Human Investigations Committee.

BMI was calculated using self-reported height and weight (kg/m^).

The Eating Disorder Examination-Questionnaire (EDE-Q) (22) assesses the frequency of objective binge episodes (OBEs; defined as feeling a loss of control while eating unusually large quantities of food; this definition corresponds to the DSM-5 criteria for binge eating), subjective binge episodes (SBEs; defined as feeling a loss of control while eating, but without eating unusually large quantities of food), and inappropriate weight control and purging methods over the past 28 days; it comprises four subscales (Restraint, Eating Concern, Weight Concern, and Shape Concern) and a Global total score. The EDE-Q has good test-retest reliability (23), convergence with the EDE interview (24), and good performance in community studies (25).

The Yale Food Addiction Scale (YFAS) (12) is a 25-item self-report measure of addictive eating. Items correspond to substance-dependence criteria from DSM-IV (APA) (26). The YFAS has adequate internal reliability, convergent validity, and incremental validity in predicting binge eating (12,27).

The Beck Depression Inventory (BDI) (28) assesses depressive symptoms and levels; it has strong psychometric support (29) and performs well as a marker for severity and distress (30).

The Barratt Impulsiveness Scale -11 (BIS-11) (31) consists of thirty items measuring three domains and six subdomains of impulsivity: attentional (attention and cognitive instability), motor (motor and perseverance), and non-planning (self-control and cognitive complexity) impulsivity. Higher scores are indicative of greater impulsivity.

The Brief Self-Control Scale (BSCS) (32) consists of thirteen items measuring self-control over thoughts, emotions, impulse control, performance regulation and habit breaking. Higher scores are indicative of better self-control.

Creation of study groups

Participants who met criteria for overweight or obesity, but did not meet criteria for BED or FA were categorized as the Control group. Participants who met criteria for BED based on responses to the EDE-Q per DSM-5 criteria and did not meet criteria for FA were classified as the BED group. Participants who met criteria for FA based on YFAS responses, but did not meet DSM-5 BED criteria, were categorized as the FA group. Participants who met criteria for both BED and FA were categorized as the Comorbid BED+FA group. Participants with clinically significant purging behaviors were excluded.

Statistical analysis

Chi-square and Analysis of Variance (ANOVA) statistics were used to compare groups on categorical and dimensional variables. When ANOVAs revealed significant group differences, Games Howell post-hoc tests (which account for unequal group sizes) were used to analyze specific group differences. ANCOVAs were performed to covary for BMI. Effect sizes, partial η^{, were calculated.}

3. Results

BED and FA Rates among Persons with Excess Weight

Of the overall participant group with overweight or obesity, 12.0% (n=60) met criteria for BED and 26.7% (n=134) met FA criteria. Of those who met BED criteria, 61.7% (n=37) also met FA criteria. Of those who met FA criteria, 27.6% (n=37) also met BED criteria (Figure 1).

Figure 1

Rates of BED, FA, and Co-Occurring BED+FA in Overweight or Obesity

BED = Binge Eating Disorder; FA = Food Addiction

Of the overall participant group with overweight or obesity, 68.7% (n=345) were classified as the Control group, 4.6% (n=23) as the BED only group, 19.3% (n=97) as the FA only group, and 7.4% (n=37) as the Comorbid BED+FA group. The four groups did not significantly differ on age, sex, or race (Table 1), but significantly differed on BMI (Table 2). Post-hoc tests revealed that the FA and Comorbid BED+FA groups had significantly higher BMI than the Control group; the three eating groups did not significantly differ from each other on BMI.

Table 1

Demographic characteristics across four study groups with excess weight

	1 Control Overweight/ Obesity n=345		2 BED n=23		3 FA n=97		4 BED+FA n=37		ANOVA	η2	Posthoc
Age, mean (SD)	37.5	(13.2)	36.8	(12.1)	38.4	(12.9)	42.0	(12.7)	F(3, 437)=1.19	.008	NS
Female, No (%)	279	(81.6%)	19	(82.6%)	84	(86.6%)	33	(89.2%)	χ2 (3, n=499)=2.40	.069	NS
White, No (%)	277	(80.3%)	19	(82.6%)	74	(76.3%)	34	(91.9%)	χ2 (3, n=502)=4.23	.092	NS

Note. N=502.

Table 2

Comparison of body mass index and eating disorder psychopathology across four study groups with excess weight

	1 Control Overweight/ Obesity n=345		2 BED n=23		3 FA n=97		4 BED+FA n=37		ANOVA			ANCOVA BMI
	M	sd	M	sd	M	sd	M	sd		η2	Posthoc	η2
Body Mass Index	32.7	(6.5)	34.4	(6.7)	36.0	(7.4)	36.0	(7.4)	8.10***	.047	3,4>1	--
EDE-Q Global	2.5	(1.1)	3.7	(0.8)	3.7	(1.0)	3.8	(0.9)	51.78***	.238	2,3,4>1	.222
Restraint	2.1	(1.4)	2.6	(1.7)	2.5	(1.5)	2.5	(1.3)	4.18**	.025	3>1	.026
Eating Concern	1.3	(1.1)	3.0	(1.0)	3.2	(1.4)	3.4	(1.1)	94.35***	.362	2,3,4>1	.345
Shape Concern	3.6	(1.5)	5.0	(0.9)	4.9	(1.1)	5.1	(0.9)	34.25***	.171	2,3,4>1	.154
Weight Concern	3.1	(1.3)	4.2	(0.9)	4.3	(1.1)	4.4	(0.9)	37.75***	.185	2,3,4>1	.165
OBE	0.8	(2.1)	8.2	(6.4)	4.4	(11.1)	10.4	(8.7)	43.19***	.206	2,3,4>1; 4>3	.204
SBE	1.5	(2.8)	7.9	(5.8)	5.4	(11.7)	11.1	(9.1)	35.91***	.178	2,3,4>1; 4>3	.173
YFAS	2.4	(1.5)	4.6	(1.7)	5.3	(1.4)	5.5	(1.4)	145.08***	.466	2,3,4>1	.441

Note. N=502. Control = No Binge Eating Disorder or Food Addiction; BED = Binge Eating Disorder (No Food Addiction); FA = Food Addiction (No BED); BED+FA = BED and FA; EDE-Q = Eating Disorder Examination - Questionnaire; OBE = objective bulimic episodes (binge eating frequency); SBE = subjective bulimic episodes; YFAS = Yale Food Addiction Scale.

^p<.01;

^{p<.0005. df (3, 498)}

BED and FA Rates among Persons with Excess Weight

Of the overall participant group with overweight or obesity, 12.0% (n=60) met criteria for BED and 26.7% (n=134) met FA criteria. Of those who met BED criteria, 61.7% (n=37) also met FA criteria. Of those who met FA criteria, 27.6% (n=37) also met BED criteria (Figure 1).

Figure 1

Rates of BED, FA, and Co-Occurring BED+FA in Overweight or Obesity

BED = Binge Eating Disorder; FA = Food Addiction

Of the overall participant group with overweight or obesity, 68.7% (n=345) were classified as the Control group, 4.6% (n=23) as the BED only group, 19.3% (n=97) as the FA only group, and 7.4% (n=37) as the Comorbid BED+FA group. The four groups did not significantly differ on age, sex, or race (Table 1), but significantly differed on BMI (Table 2). Post-hoc tests revealed that the FA and Comorbid BED+FA groups had significantly higher BMI than the Control group; the three eating groups did not significantly differ from each other on BMI.

Comparisons between the control and eating groups

The three eating groups differed significantly from the Control group on all clinical variables. First, the three eating groups reported elevated scores on the EDE-Q Global and three EDE-Q subscales (Eating Concern, Shape Concern, and Weight Concern), objective binge-eating episodes, subjective binge-eating episodes, and YFAS scores relative to the Control group. Only the FA group reported significantly greater EDE-Q Restraint scores than the Control group (Table 2). Second, all three eating groups reported significantly lower BSCS scores (indicating poorer self-control) than the Control group, whereas both the FA and Comorbid BED+FA groups reported significantly greater BDI scores than the Control group (Table 3). Finally, results varied based on impulsivity total and subscale scores. With respect to BIS total scores, the BED and FA groups each had significantly higher scores than the Control group; however, results differed based on higher-order impulsivity domains (Attentional, Motor, and Non-Planning) (Refer to Table 3).

Comparisons among the three eating groups (BED, FA, and BED+FA

When comparing the three eating groups, the groups significantly differed from each other on OBEs and SBEs. Specifically, the Comorbid BED+FA group had significantly more OBEs and SBEs than the FA only group. Otherwise, the three eating groups did not statistically differ on EDE-Q total and subscale scores, YFAS, BDI, BIS total or subscale scores, or BSCS scores. Partial eta squared ranged from .025 (EDE-Q Restraint) to .466 (YFAS), signifying small to large effect sizes. Co-varying for BMI did not result in substantive changes in findings or in attenuation of effect sizes, which ranged from .026 (EDE-Q Restraint) to .441 (YFAS).

Discussion

In this online convenience sample of community volunteers with overweight or obesity, BED and FA were common. A recent US-based population study of women reported that 5.8% of those surveyed met criteria for FA using a modified brief version of the YFAS (33). We note that the prevalence of FA observed in the current sample of participants with excess weight is much higher than that observed by Flint and colleagues (33), although not surprising given that prior studies have identified increasing rates of FA among groups defined by overweight or obesity. In the current study, BED and FA were associated with greater psychopathology relative to the group with excess weight without either form of disordered eating. Nearly one-third of the participant group with overweight or obesity met criteria for BED, FA, or both. Overall, FA was more common than BED; 26.7% and 12.0% met FA and BED criteria, respectively. Over sixty percent of those with BED also met FA criteria, and this finding approximates the proportion of FA in a treatment-seeking group of individuals with BED and obesity (15). The three eating groups with BED, FA, and co-occurring BED and FA reported significantly greater levels of eating disorder psychopathology, impulsivity, and self-control than their counterparts with overweight/obesity only, even after adjusting for BMI. The presence of different forms of problematic eating seem to represent important subtypes of persons with excess weight, although there was little support for the distinction of BED, FA, and co-occurring BED and FA, which differed minimally from each other. These findings of individuals with overweight/obesity are generally consistent with those reported by Gearhardt and colleagues (18), which included a study group of bulimia nervosa in a sample representing the full weight range.

As expected, all three eating groups (BED, FA, and comorbid BED+FA) reported greater levels of eating disorder psychopathology, binge eating, loss-of-control eating, and food addiction than the control group with overweight/obesity. Contrary to expectations, the three eating groups did not differ from each other on various measures of eating-related concerns and features. BED, FA, and co-occurring BED+FA had similar “food addiction” scores and did not differ of on the measures of eating disorder psychopathology (including weight and shape concerns) but did show differences on overeating behaviors. The BED+FA group reported significantly greater frequency of both objective (M=10.4) and subjective (M=11.1) binge-eating episodes than the FA group (M=4.4 and M=5.4, respectively). Finally, with respect to BMI, the FA and comorbid BED+FA groups had significantly higher BMIs than the Control group; however, BMI did not statistically differ among the three groups with disordered eating.

In addition to examining weight and eating disorder psychopathology, this study sought to better understand non-eating specific clinical domains that have been found to be associated with disordered eating including depression, impulsivity, and self-control. With respect to depression, the FA and BED+FA groups reported significantly greater depressive levels (scores in the borderline to moderate depression range) than the control group with overweight/obesity (scores in the mild mood disturbance range). While depressive symptoms in the BED group did not significantly differ from the control group with overweight or obesity, BED depressive scores matched depressive scores in a treatment-seeking sample of individuals with BED and obesity (16).

Within the impulsivity domain, the general patterning suggested the eating groups had higher scores on various impulsivity domains compared to the control group with overweight/obesity. Again, the three eating groups, however, did not significantly differ from each other on overall or subdomains of impulsivity. Relative to the control group with overweight or obesity, all eating groups reported higher levels of attentional impulsivity, which is indicative of difficulty focusing or concentrating on tasks, while the BED and FA groups reported significantly higher levels of motor impulsivity (tendency to act on the spur of the moment without thinking) and non-planning impulsivity (less careful thinking/planning or lack of future orientation). Our findings for BED replicate and extend previous literature suggesting greater motor impulsivity among individuals with binge eating and obesity when using a laboratory test meal design (34) and fMRI technology (35). In addition, higher BED scores on the cognitive complexity subdomain are consistent with an emerging literature suggesting that changes in cognitive processing may underlie the development of BED (19).

In contrast to two previous clinical studies with treatment-seekers (15,16), which found that the presence of FA in BED represented a more disturbed subgroup than BED without FA, our present findings with a community sample do not show elevated clinical disturbance in the comorbid BED+FA group. The discrepancy may be due to different assessment methods (self-report versus interview for BED) and different samples (clinical treatment-seekers versus non-clinical), although the BDI scores were similar across the studies.

These findings should be interpreted in light of the study strengths and limitations. A strength of our study was the relatively large sample size, which allowed for group comparisons and fine grained analyses. However, the sample size for some groups may have not allowed for the detection of small between-group effects. Our findings are cross-sectional and were gathered from community respondents interested in research. Future studies should use prospective and experimental designs to further tease apart these questions. In addition, our study relied on self-report measures which may be biased; alternatively, anonymity may help individuals disclose private or embarrassing behaviors, particularly related to disordered-eating behaviors (36). BMI was also calculated using self-reported height and weight. Although individuals tend to underestimate weight and overestimate height (37), overall measured and self-reported weight are highly correlated, and the magnitude of actual differences tend to be small (38). Moreover, research with persons with disordered eating has found that they are generally accurate reporters of weight/height (39) and that the generally small magnitude of reporting errors is not systematically related to eating disorder psychopathology (40).

Conclusion

This study of an online convenience sample of community volunteers with overweight/obesity, found that BED and FA are common, that BED and FA frequently co-occur with each other, and that the presence of BED, FA, or both is associated with significantly greater psychopathology. These findings add to the well-established evidence base suggesting BED is a distinct clinical entity from obesity, and importantly, provide new evidence suggesting that FA is also distinct from obesity. Disordered eating in the form of FA or BED signals elevated distress and these groups may require additional targeted intervention. Given that FA is not a formally recognized diagnosis, our findings suggest that a significant proportion of individuals with overweight/obesity with such concerns may go undetected in clinical and research settings. Future research should examine whether tailored treatment approaches for these distinct obesity subtypes are warranted.

What is already known about this subject?

• Obesity is associated with binge eating disorder (BED) and food addiction (FA)

• BED and FA share similar features but may represent unique constructs

• Comorbid BED and FA may represent a more pathological BED subgroup

What does your study add?

• FA was more common than BED among persons with overweight/obesity in this community sample

• Over 60% with BED also met FA criteria whereas 28% with FA also met BED criteria

• Presence of BED, FA, or both BED+FA signals greater psychopathology than overweight/obesity, but co-occurring BED+FA does not signal greater disturbance than BED or FA alone

Acknowledgments