A Developmental Framework of Binge-Eating Disorder Based on Pediatric Loss of Control Eating

Abstract

Though binge-eating disorder may manifest in childhood, a significantly larger proportion of youth report episodes involving a loss of control while eating, the hallmark feature of binge-eating, that predicts excess weight gain and obesity. Adults with binge-eating disorder often report that symptoms emerged during childhood or adolescence, suggesting that a developmental perspective of binge-eating may be warranted. Thus, loss of control eating may be a marker of prodromal binge-eating disorder among certain susceptible youth. The present paper offers a broad developmental framework of binge-eating disorder and proposes areas of future research to determine which youths with loss of control eating are at-risk for persistent and exacerbated behavior that may develop into binge-eating disorder and adult obesity. To this end, this paper provides an overview of loss of control eating in childhood and adolescence, including its characterization, etiology, and clinical significance, with a particular focus on associations with metabolic risk, weight gain, and obesity. A conceptual model is proposed to further elucidate the mechanisms that may play a role in determining which youths with loss of control are at greatest risk for binge-eating disorder and obesity. Ways in which treatments for adult binge-eating disorder may inform approaches to reduce loss of control eating and prevent excess weight gain in youth are discussed.

Editor’s note.

This article is part of a special issue, “Obesity: Psychological and Behavioral Aspects of a Modern Epidemic,” published in the Xxxxxx 2020 issue of American Psychologist. David B. Sarwer and Carlos M. Grilo served as editors of the special issue, with Anne E. Kazak as advisory editor.

As with many psychiatric disorders, early signs of binge-eating disorder (BED) may manifest during childhood. While the identification of a syndrome resembling BED dates back to 1959 (Stunkard, 1959), the majority of research began three decades ago, stimulated in large part by the field trials supporting the inclusion of BED as a research category in the DSM-IV (Spitzer et al., 1993; Yanovski, 1993). Binge-eating, defined as the consumption of an unusually large quantity of food while experiencing a lack of control, is the hallmark feature of BED. Currently, BED is defined in the DSM-5 (American Psychiatric Association, 2013) by recurrent episodes (once-weekly binge-eating for three or more months) in the absence of regular compensatory behaviors (Figure 1). The DSM-5 also requires that, during the binge-eating episodes, an individual must endorse at least three out of five behavioral indicators (e.g., eating faster than usual, eating until uncomfortably full). Individuals with BED present with clinically significant distress or impairment as a result of the episodes. During the 1990s, researchers studying BED reported that binge-like behaviors, such as the feeling of losing control during eating, typically manifested in middle childhood or adolescence (Bulik, Sullivan, Carter, & Joyce, 1997; Spurrell, Wilfley, Tanofsky, & Brownell, 1997). This led to a wealth of research in pediatric samples – most focusing on the construct of ‘loss of control’ eating (LOC), defined as a lack of control over eating or inability to stop eating, irrespective of the amount of food consumed, as opposed to binge-eating or BED.

Figure 1.

Definition of key terms

Simultaneously, the high rates of obesity among youth (Ogden et al., 2016) and adults (Flegal, Kruszon-Moran, Carroll, Fryar, & Ogden, 2016) were gaining attention. With the recognition that obesity is a heterogeneous condition (Field, Camargo, & Ogino, 2013), researchers sought to illuminate potentially modifiable risk factors that promote overweight across the lifespan. Obesity is influenced by multiple factors such as sociocultural determinants and genetic and metabolic influences that collectively impact energy intake and expenditure (Keith et al., 2006). Compounding these factors, disordered eating, including binge-eating and “full-threshold” BED (American Psychiatric Association, 2013) are consistently associated with obesity in epidemiological studies (Hudson, Hiripi, Pope, & Kessler, 2007; Kessler et al., 2013; Udo & Grilo, 2018).

While some youth meet diagnostic criteria for BED, many more report the experience of LOC eating or sub-threshold binge-eating. Research among pediatric samples with LOC eating indicates that the behavior maps onto comorbid psychological and physiological symptoms in a pattern similar to adults with BED, but to a lesser degree. For example, compared to weight-matched youth without LOC eating, those with LOC report more mood symptoms (Shomaker et al., 2010); similarly, adults with obesity and BED are more likely to report mood and anxiety disorders than those with obesity alone (Grilo, White, & Masheb, 2009; Spitzer et al., 1993). Thus, identifying youth with LOC eating who are at-risk for developing BED may be an opportunity to prevent the full disorder and associated adverse outcomes (Tanofsky-Kraff et al., 2007). This paper will provide a brief review of BED and its links to obesity and psychopathology. Then, an overview of the definition, prevalence, associated demographics, and predictors of LOC eating in youth will be described. The literature supporting the associations of LOC eating with obesity, related health correlates, and salient psychosocial factors will be discussed, as will the natural course of LOC eating among youth. A general conceptual model for the development of BED and areas for future research to explore mechanisms that might account for persistent and/or exacerbated LOC eating among subsets of youth will be proposed. In this endeavor, three critical disclaimers should be noted. First, given that LOC eating is not only a defining feature of binge-eating (American Psychiatric Association, 2013) and has recently become the diagnostic criterion (i.e., without the unusually large size component) for disorders involving “binge-eating” in the International Classification of Diseases (Stein D. et al., In Press), but that it also consistently predicts the development of BED (Hilbert, et al., 2013; Tanofsky-Kraff, et al., 2011) and increased body weight (Field et al., 2003; Micali et al., 2015; Mustelin, Kaprio, & Keski-Rahkonen, 2018; Sonneville et al., 2013; Tanofsky-Kraff et al., 2006; Tanofsky-Kraff, Yanovski, et al., 2009), the paper will focus primarily on LOC eating. Despite a wealth of research on other constructs related to disinhibited eating (e.g., emotional eating, reinforcing value of food, food addiction, negative urgency), prospective data are scant, and very few have found direct predictive links with weight gain (Butte et al., 2007; Fisher & Birch, 2002; Kelly et al., 2015; Stojek et al., 2017). Yet, pediatric eating-related characteristics that interact with LOC to increase risk for adult BED will be discussed accordingly. In addition, though binge-eating is present in other disorders (e.g., anorexia nervosa: binge/purge subtype and bulimia nervosa) (APA, 2013), given the distinct etiologies and courses, other clinical presentations involving binge-eating are beyond the scope of the current paper. Second, this paper will focus on childhood and adolescence, as the overwhelming majority of the research is in these cohorts. Finally, this paper is not intended to be an exhaustive review of the literature. Rather, relevant research is provided to support a conceptual framework upon which future research may be initiated.

Binge-Eating Disorder in Adults

Developing most often in early adulthood (Hudson et al., 2007; Kessler et al., 2013; Udo & Grilo, 2018), interview-based estimates indicate that BED impacts between just under 1% (Erskine & Whiteford, 2018; Udo & Grilo, 2018) and 2.6% (Hudson et al., 2007) of the general population, with higher rates observed among adults with overweight/obesity (Smith, Marcus, Lewis, Fitzgibbon, & Schreiner, 1998). Adult BED is associated with elevated rates of psychiatric and psychosocial impairment in epidemiological (Grilo et al., 2009; Hudson et al., 2007; Udo & Grilo, 2019; Ulfvebrand, Birgegård, Norring, Högdahl, & von Hausswolff-Juhlin, 2015) and community (Javaras et al., 2008) studies, as well as across diverse clinical and treatment-seeking samples (Grilo, White, Barnes, & Masheb, 2013; Grilo et al., 2009; Ulfvebrand et al., 2015).

The link between BED and body weight has been well-documented. Community and epidemiologic studies among adults consistently find that subthreshold and threshold BED are associated with higher weight and obesity, even after adjusting for sociodemographic factors (Hudson et al., 2007; Kessler et al., 2013; Mustelin, Bulik, Kaprio, & Keski-Rahkonen, 2017; Mustelin et al., 2018; Nagata, Garber, Tabler, Murray, & Bibbins-Domingo, 2018; Rodgers, Watts, Austin, Haines, & Neumark-Sztainer, 2017; Thornton et al., 2017; Udo & Grilo, 2018). Specifically, those with a lifetime diagnosis of BED may be over four times more likely to have severe obesity (Udo & Grilo, 2018). Though more prospective studies are needed, BED may predict weight gain over time, as well as exacerbated metabolic outcomes (Hudson et al., 2010).

Prevalence of BED and Loss of Control Eating in Youth

Based on large survey studies, prevalence rates of BED in youth range from 0.7% to 2.3%, with approximately 2% of girls and under 1% of boys reporting the disorder on surveys (Ackard, Fulkerson, & Neumark-Sztainer, 2007; Allen, Byrne, Oddy, & Crosby, 2013; Smink, van Hoeken, Oldehinkel, & Hoek, 2014; Swanson, Crow, Le Grange, Swendsen, & Merikangas, 2011). Some data suggest rates may be higher among youth with overweight; using interview-based assessments, BED (Glasofer et al., 2007) and subthreshold BED (Decaluwé, Braet, & Fairburn, 2003; Lourenço et al., 2008) have been reported by up to 7% of youth with high body weight. The broad range of estimates may be due to the difficulty defining a large amount of food among boys and girls at different developmental stages (MacGuire, 2011; Shomaker, Tanofsky-Kraff, & Yanovski, 2011). For example, what might be an unambiguously large amount of food for a 9-year-old girl might be quite normative for a 16-year-old boy (Shomaker et al., 2010). Further, children may have less autonomy over food decisions compared to adolescents or adults, and may lack access to objectively large amounts of food (Shomaker et al., 2011). Thus, youth may experience a lack of control over eating while consuming ambiguously large or moderate amounts of food. This has prompted most researchers to examine LOC eating in youth in addition to, or as opposed to, classic binge-eating episodes.

Roughly 23% of youth of all weight strata report at least one LOC eating episode in the past month, and another 10% report recurrent episodes (Schlüter, Schmidt, Kittel, Tetzlaff, & Hilbert, 2015). Among youth with high weight, LOC eating is even more prevalent (Tanofsky-Kraff et al., 2004); approximately 33% report the experience of LOC (He, Cai, & Fan, 2016; Morgan et al., 2002) and, among those seeking weight loss treatment, nearly 50% report past or current LOC eating (e.g., Glasofer et al., 2007). Youth report LOC eating as early as 8 years of age (Tanofsky-Kraff, Faden, Yanovski, Wilfley, & Yanovski, 2005; Tanofsky-Kraff et al., 2004), with higher rates reported by adolescents. Adolescent girls report engaging in LOC more often than their male peers (Shomaker et al., 2011). Data on racial and ethnic differences in the prevalence of LOC eating are inconsistent; some have observed higher prevalence among racial and ethnic minorities (Field, Colditz, & Peterson, 1997; Johnson, Rohan, & Kirk, 2002; Story, French, Resnick, & Blum, 1995; Swanson et al., 2011), while others have observed similar rates across groups (Cassidy et al., 2012; Glasofer et al., 2007; Tanofsky-Kraff, Goossens, et al., 2007). Research also suggests that sexual and gender minority youths may be at greater risk for disinhibited eating behaviors (Calzo, Blashill, Brown, & Argenal, 2017), though disparities in LOC eating specifically have not yet been assessed.

Risk Factors for LOC Eating

Though data are limited, there appear to be genetic underpinnings to LOC eating. After adjusting for parent and child body mass index (BMI), mothers with binge-eating were more likely to have a child with LOC eating (Zocca et al., 2011). While such data may reflect both genetic and environmental risks, candidate genes have also been found to play a role in the development of LOC eating in youth. The FTO high-risk A allele, a polymorphism that places youth at greater risk for obesity (Frayling et al., 2007), has been linked to LOC eating. Adjusting for standardized BMI (BMIz), youths with rs9939609 FTO variant high-risk A alleles were more likely to report LOC eating as compared to their peers without the high-risk allele (Tanofsky-Kraff, Han, et al., 2009). FTO is highly expressed in the hypothalamic regions of the brain, which are important for appetite regulation, suggesting the impact of the FTO high-risk A allele on the hypothalamus may promote LOC eating that in turn leads to overweight.

There are few prospective studies on the predictors of LOC onset. These data suggest that hormones during puberty, eating-related and general psychopathology, as well as temperament and home environment may influence LOC onset and severity. Higher concentrations of estradiol, which increases after the onset of puberty in females, have been associated with the development of LOC eating (Klump et al., 2012). Estradiol is involved in the expression of genes linked to negative affect and has been shown to mediate the relationship between genetic risk and LOC eating (Klump, 2013; Vannucci et al., 2015). Moreover, early menarche may lead to a discrepancy between biological puberty and psychosocial maturation (Gluckman & Hanson, 2006), which may be associated with early onset of binge-eating (Day et al., 2011). Prospective data suggest that dieting (Field et al., 2003; Neumark-Sztainer, Wall, Haines, Story, & Eisenberg, 2007; Stice, Marti, & Durant, 2011), body dissatisfaction (Neumark-Sztainer, Paxton, Hannan, Haines, & Story, 2006; Stice & Shaw, 2002), emotional eating (Stice, Presnell, & Spangler, 2002), and eating in the absence of hunger (Balantekin, Birch, & Savage, 2017) are risk factors for the onset of LOC in adolescents. Other data suggest that temperament and individual-level traits during development, such as impulsivity and reward sensitivity, may put youth at risk for LOC eating (Cassin & von Ranson, 2005; Hilbert et al., 2014). Maladaptive family environments (Le Grange, Lock, Loeb, & Nicholls, 2010), including parental under-involvement, critical weight-related comments, and life transitions, are also predictors of LOC eating in youth (Hartmann, Czaja, Rief, & Hilbert, 2012).

Though most data indicate that LOC eating predicts excess weight gain (Field et al., 2003; Micali et al., 2015; Mustelin, Kaprio, & Keski-Rahkonen, 2018; Sonneville et al., 2013; Tanofsky-Kraff et al., 2006; Tanofsky-Kraff, Yanovski, et al., 2009), it is also plausible that the converse is true as well. For instance, a family history of obesity may interact with certain psychosocial risk factors to exacerbate one’s predisposition to LOC eating. In addition, weight-based teasing, reported by up to 60% of youth with high body weight (Schvey et al., 2019; Puhl et al., 2013), has been shown to predict binge-eating (Puhl et al., 2017), and thus may mediate the relationship between obesity and LOC eating. Alterations in appetitive hormones resulting from excess adiposity (Myers Jr, Leibel, Seeley, & Schwartz, 2010) may also promote LOC among youth. Further research is needed to address the potentially reciprocal relationship between obesity and LOC eating.

Loss of Control Eating in Relation to Obesity and Metabolic Outcomes

Across age strata, samples, and methodologies, studies have consistently found that pediatric LOC eating is associated with, and predictive of, greater BMI and body fat, even after adjusting for initial weight and adiposity (Field et al., 2003; Micali et al., 2015; Mustelin et al., 2018; Sonneville et al., 2013; Tanofsky-Kraff et al., 2006; Tanofsky-Kraff, Yanovski, et al., 2009). In a large study of non-treatment seeking youth (6–17 years), those who reported at least one episode of LOC eating in the past month were significantly heavier and had more body fat compared to both youth who reported overeating without LOC and to youth who reported neither overeating nor LOC (Shomaker et al., 2010). In addition, after adjusting for body fat, youth reporting LOC showed greater dysfunction in components of metabolic syndrome including higher systolic blood pressure and higher low-density lipoprotein cholesterol (Radin et al., 2015). Prospectively, compared to youth without LOC, those reporting LOC in the past month gained an additional 2.4 kg (5.3 lbs) per year (Tanofsky-Kraff, Yanovski, et al., 2009). Moreover, compared to children without LOC, those with reported LOC eating were at greater risk for elevated triglycerides and increased visceral adipose tissue, and were five times more likely to meet criteria for metabolic syndrome, the latter of which was only partially accounted for by excess weight gain (Tanofsky-Kraff et al., 2012).

Although the directionality of relationships is unclear, a potential mechanism linking LOC to weight gain and retention may be aberrations in metabolic function due to recurrent LOC episodes. Supporting this notion, studies have documented associations between LOC eating and appetitive hormones, such as serum leptin, an adipose-tissue derived hormone which regulates body weight and energy homeostasis (Miller et al., 2014). Data also indicate links between LOC and poorer metabolic functioning such as higher fasting insulin and insulin resistance (Byrne et al., 2018). Dietary intake and macronutrient consumption may also be partially responsible for the association between LOC and metabolic dysfunction. In standardized laboratory test meals, youth with LOC consume more carbohydrates, including snack- and dessert-type foods (Hilbert, Tuschen-Caffier, & Czaja, 2010; Tanofsky-Kraff, McDuffie, et al., 2009) than controls with no LOC eating. In addition to dietary intake, high-sensitivity C-reactive protein, a marker of inflammation, has demonstrated an association with LOC among youth (Shank, Tanofsky-Kraff, et al., 2017). C-reactive protein is linked with greater risk for the onset of cardio-metabolic abnormalities among youth (Semiz, Rota, Ozdemir, Ozdemir, & Kaptanoglu, 2008) and adults (Freeman et al., 2002; Sesso et al., 2003), indicating that systemic inflammation may be one pathway linking LOC eating to poor metabolic health. Of note, all of these relationships remained significant after accounting for the contribution of fat mass. Collectively, laboratory test-meal and metabolic literatures suggest that important differences in eating behaviors characterize youth with LOC relative to controls.

Naturalistic Outcomes

The course of LOC eating among youth is highly variable. For some (upwards of 50%), the presence of LOC eating is associated with the onset of partial or full-syndrome BED over time (Hilbert, Hartmann, Czaja, & Schoebi, 2013; Tanofsky-Kraff et al., 2011), suggesting that for these youths, LOC eating may indicate prodromal BED. Similarly, some findings indicate that the behavior persists (Pearson et al., 2017) or increases over the course of development, with prevalence rates of LOC eating in adolescent girls estimated at 14–25% (Croll, Neumark-Sztainer, Story, & Ireland, 2002; Goldschmidt, Wall, Zhang, Loth, & Neumark-Sztainer, 2016; Neumark-Sztainer, Wall, Story, & Standish, 2012). Those reporting childhood-onset LOC eating may also be at risk for more severe binge-eating compared to those with adult-onset LOC eating (Brewerton, Rance, Dansky, O’Neil, & Kilpatrick, 2014). Even among those who do not develop a partial or full-syndrome eating disorder, LOC eating in childhood predicts the worsening of disordered eating symptoms and anxiety and depressive symptoms (Hilbert et al., 2013; Tanofsky-Kraff et al., 2011).

However, consistent with the spontaneous remission rates of psychiatric disturbances among adults, including personality disorders (Grilo et al., 2004; Grilo, Sanislow, et al., 2005), data suggest that LOC eating tends to remit in about 50% of youth, even in the absence of intervention (Goldschmidt et al., 2016; Hilbert et al., 2013; Tanofsky-Kraff et al., 2011). Given the potentially pernicious outcomes that have been reported, yet data also suggesting that LOC eating spontaneously remits in some, elucidating which youth are at risk for persistent or worsening LOC and adverse outcomes warrants further attention.

A Developmental Framework of Adult Binge-Eating Disorder

Distinguishing these groups will be critical in allocating resources towards those for whom LOC eating suggests prodromal BED or indicates risk for obesity in adulthood. With the recognition that multi-level factors, including but not limited to, genetics, environment, family, and culture, impact the development of adult BED and obesity, a framework for studying putative psychological predictors of BED is offered to elucidate youth with LOC who are at greatest risk. This model hypothesizes that among youth reporting LOC eating, those presenting with some interaction of negative affectivity, elevated reward responsivity for food, and alterations in executive functioning are at greatest risk for the development of BED and adult obesity (Figure 2). An important caveat is that this is a broad-based model and that the factors identified under each construct can involve varying levels of analyses that should be considered in future research.

Figure 2.

Framework depicting the mechanisms that may place youth with loss of control eating at risk for adult binge-eating disorder and obesity

Negative Affectivity.

Studies aiming to elucidate the psychological underpinnings and correlates of LOC among youth have determined that negative affect plays a critical role. Indeed, affect theory, which proposes that individuals engage in episodes of uncontrolled eating in order to escape from, or alleviate, adverse emotions and negative mood states (Kenardy, Arnow, & Agras, 1996), is well-accepted. Research has consistently observed greater anxiety, social stress, and depression among youth with LOC compared to their peers without LOC (Byrne, LeMay-Russell, & Tanofsky-Kraff, 2019; Shomaker et al., 2010). In addition, studies have found that youth with LOC report significantly greater negative mood both before and after eating (after adjusting for pre-meal affect) as compared to youth with other disinhibited eating behaviors (e.g., emotional eating (Vannucci et al., 2013)) or no LOC (Tanofsky-Kraff, McDuffie, et al., 2009). Among a sample of adolescent girls with LOC eating, pre-meal negative affect was related to greater consumption of carbohydrates, dessert, and snack-type foods (Ranzenhofer et al., 2013).

Emotional eating is also associated with LOC eating (Goossens, Braet, & Decaluwé, 2007; Tanofsky-Kraff, Theim, et al., 2007) and may help to identify youth at risk for adverse outcomes. One longitudinal study found that emotional eating was associated with increased disordered eating and adiposity at 1-year follow-up, but only among youth with LOC eating at baseline (Stojek et al., 2017), indicating that eating in response to negative affect may be a salient mechanism in the promotion of adult BED and obesity among youth with LOC. Neuroimaging research lends further support for the role of anxiety insofar as it promotes and maintains LOC eating. For instance, one study observed similarities between youth with LOC (Jarcho et al., 2015) and youth with anxiety in their neural responses to peer rejection (Guyer, McClure-Tone, Shiffrin, Pine, & Nelson, 2009). Specifically, when exposed to an interpersonal stressor, for girls with overweight and LOC eating, activation in the fusiform face area, a region involved in facial recognition that plays a role in affective processing, was linked to greater energy intake in the laboratory. Such data suggest the relevance of anxiety for the promotion and maintenance of LOC among some youths. Indeed, interactions between LOC and anxiety in relation to body weight and metabolic function support this notion. Anxiety has been associated with an exacerbated metabolic state, but only among youths with LOC (Byrne, Tanofsky-Kraff, et al., 2019; Byrne et al., 2018). Another study found the frequency of LOC episodes mediated the relationship between trait anxiety and both BMIz and adiposity among a sample of healthy youth (Grammer et al., 2018). Finally, pre-meal state anxiety (but no other mood state, including depression) mediated the association between recent social stress and highly palatable food intake among youth with LOC (Shank, Crosby, et al., 2017).

Preliminary physiologic data also lend support for the role of stress, a closely-related construct, in LOC eating. In a pilot study, adolescent girls with LOC eating completed ecological momentary assessment while wearing continuous heart rate monitors (Ranzenhofer et al., 2016). Results indicated that heart rate was higher and heart rate variability was lower prior to LOC, as compared to non-LOC eating episodes. Importantly, high heart rate and low heart rate variability are considered robust physiologic indicators of stress (Porges, 1995); thus among adolescent girls with reported LOC, stress may be a salient predictor of LOC episodes. Collectively, negative affectivity, specifically anxiety, may be an important facet of, and precipitant to, LOC eating.

Reward Responsivity.

A drive towards food and food cues has been linked to pediatric LOC eating (Shank et al., 2015), adult BED (Schag et al., 2013), and obesity (Volkow, Wang, Tomasi, & Baler, 2013; Yokum, Ng, & Stice, 2011). Specifically, among susceptible youths, biased processing of food-related cues may result in heightened reward responsivity, which may then increase the salience and desirability of certain foods. Paired with the obesogenic environment, this may place youths with greater attentional bias to food at-risk for BED and subsequent obesity. Whereas adult studies demonstrate a link between attentional bias to food and both BED and BMI (Schmitz, Naumann, Trentowska, & Svaldi, 2014; Stojek et al., 2018), the relationship in youth is less consistent. Some studies have found that attentional bias towards food is predictive of future weight gain among adolescent females (Yokum et al., 2011), while others have found that attentional bias towards highly palatable foods is positively associated with BMI-z, but only among those reporting LOC (Shank et al., 2015). Thus, among youth with LOC eating, attentional bias towards food may indicate risk for exacerbated outcomes, though additional prospective data are required to determine the temporal nature of these relationships.

Another mechanism potentially placing youth with LOC eating at risk for adult BED and obesity may be heightened reward expectancies, or an exaggerated expectation that consumption of desired foods will be highly rewarding or pleasurable. These heightened expectations may also be maintained and exacerbated by elevated taste responsiveness, which may initiate a vicious cycle which elevates risk for disinhibited eating behaviors, such as LOC eating (Berridge, 2009). Supporting this notion are findings showing that children with LOC eating exhibit greater neural activation to highly palatable food cues compared to children without LOC eating (English et al., 2018). A number of other reward-based constructs, such as the reinforcing value of food and food addiction, warrant study in relation to LOC eating to elucidate which youth may be predisposed to develop adult BED and obesity.

Executive Functioning.

Alterations in executive function and response inhibition (top-down mental control processes) have been hypothesized to play a role in the onset and maintenance of disinhibited eating. In support of this theory, studies show impaired inhibitory control among children and adolescents with obesity compared to age-matched lean peers (Lavagnino, Arnone, Cao, Soares, & Selvaraj, 2016). Studies of executive function among youth with LOC eating are limited, though data suggest that impaired self-regulation may underpin LOC and binge-eating in youth. For instance, one study found that youth with LOC had significantly greater impulse control deficits on both performance-based and parent-report assessments than children without LOC. Further, a significantly greater percentage of children with attention deficit hyperactivity disorder (ADHD) had recurrent LOC eating (70.5%) as compared to children without ADHD (20%) (Reinblatt et al., 2015). Another study found that adolescents with BED demonstrated poorer inhibitory control as compared to normal weight peers without BED, though they did not differ from their peers with overweight (Kittel, Schmidt, & Hilbert, 2017). In an experimental study (Hartmann, Rief, & Hilbert, 2013), adolescents with LOC were compared to adolescents with ADHD prior to and following exposure to an interpersonal laboratory stressor. Compared to peers with ADHD, adolescents with LOC demonstrated greater increases in both negative mood and impulsivity from pre- to post-stressor, indicating that youth with LOC may present with alterations in inhibitory control, as well as high emotional reactivity.

Links among Negative Affectivity, Reward Responsivity, and Executive Function

Some existing research appears to link the domains proposed for further study. Emotion regulation, for example, which refers collectively to the cognitive and behavioral processes involved in identifying, understanding, and modulating one’s own behaviors and mood states, may connect negative affectivity with executive function. For some, LOC eating may serve as a self-regulatory strategy wherein overconsumption of highly palatable foods, perhaps due to heightened reward responsivity, may function as a maladaptive attempt to dampen or displace negative affect and distress. Evidence demonstrating associations between parent-reported emotion dysregulation and LOC eating among youth support this model. Specifically, among youth with LOC eating, those with high parent-reported emotion dysregulation had significantly greater fat mass and BMIz (Kelly et al., 2016). This is consistent with earlier studies finding that emotion dysregulation predicts weight gain and obesity onset among youth (Graziano, Calkins, & Keane, 2010). Neuroimaging research further supports the role of emotion dysregulation in LOC eating among youth. For instance, Goldschmidt, et al. (2018) observed increased neural activation in areas implicated in attentional processes, inhibitory control, and emotion regulation among youth with LOC compared to controls with overweight during a food restriction paradigm (Goldschmidt et al., 2018). These findings suggest that youth with LOC and overweight, for whom food may be especially rewarding, may need to expend more cognitive effort relative to their counterparts in order to regulate their mood state and subsequent energy consumption (Goldschmidt et al., 2018). By contrast, the Jarcho et al. study observed a blunting in the ventral medial prefrontal cortex only among girls with, but not without LOC eating, in response to an interpersonal laboratory stressor (Jarcho et al., 2015). These data suggest that regulatory regions play a role in LOC eating, but may interact differentially depending on whether youth are exposed to stressors that are personally salient (e.g., social rejection). Larger studies, as well as prospective data, are warranted to elucidate these underlying multifactorial mechanisms.

Taken together, data support the notion that, in addition to genetic and environmental factors, a combination of negative affect, reward responsivity, and difficulties with self-regulation elicit LOC eating in youth. Further, the interaction of these processes may elucidate which youth with LOC eating are at greatest risk for adult BED (Figure 2 proposes a model to guide future research). It is possible that alterations in executive functioning contribute to impaired modulation of negative affect and food-cravings, leading to further exacerbations in these aversive affective states and subsequent degradations in cognitive control. While each domain may play a significant role, it is unlikely that deficits in just one of the aforementioned areas is sufficient to place a child at-risk for full-syndrome BED or adult obesity, especially given how normative some of these target constructs are. For instance, given the current obesogenic environment, moderate attentional bias to food is likely fairly common, negative affectivity tends to increase with pubertal development (Susman, Dorn, & Chrousos, 1991), and deficits in self-regulation may be common among youth, while the use of adaptive internal strategies increases with maturation. Rather, it is possible that each of these target domains is necessary, but not sufficient, to exacerbate risk for full-syndrome BED. Specifically, the subset of youth who experience high negative affectivity, heightened reward responsivity, and deficits in executive functioning and inhibitory control are likely those at highest risk for the onset of full-syndrome BED and subsequent weight gain. If this model is supported with prospective data, then identifying these high-risk youths is critical for prevention efforts. In the interim, it is expected that greater complexity will be incorporated and continued refinements will be made based on ongoing and future research. In addition, future exploration should involve multi-level factors (e.g., genetics, environment, family, and culture) in analyses to determine the extent to which each contributes to the development of BED and adult obesity.

Proposed Psychological Treatments to Prevent BED in Youth with LOC Eating

Psychological treatments for adults with BED (Grilo, 2017) may help to inform approaches to reduce LOC in youth and prevent the onset of BED in those most vulnerable. Therapies for BED reliably produce abstinence from binge-eating in roughly 50–60% of adult patients (Grilo, Masheb, & Wilson, 2005; Grilo, Masheb, Wilson, Gueorguieva, & White, 2011; Wilfley et al., 2002; Wilson, Wilfley, Agras, & Bryson, 2010), and about half remain abstinent from binge-eating two (Wilson et al., 2010) and four years (Hilbert et al., 2012) later. Reported weight losses tend to be modest overall (Grilo, 2017; Hilbert et al., 2019; Wilfley et al., 2002; Wilson et al., 2010); however, treatment-seeking patients with BED are characterized by steep weight-gain trajectories prior to intervention (Blomquist et al., 2011). For adults, therefore, weight stabilization may be conceptualized as the successful prevention of excess gain, rather than a failure to lose weight. Although this framing may be met with limited enthusiasm by adults with obesity, for youth at-risk for excess weight gain (due to LOC and other disinhibited eating risk factors) who are still developing, this approach may be more palatable. For instance, depending on pubertal status and initial weight, even successful weight management may still necessitate some weight gain. Thus, the reduction of LOC eating may be one target for preventing the development of BED and excess weight gain (Tanofsky-Kraff, 2012; Tanofsky-Kraff et al., 2007). A few studies have piloted or tested this notion utilizing modified or less intensive treatments, with mixed outcomes (Hilbert, 2013; Jones et al., 2008; Mazzeo et al., 2013; Tanofsky-Kraff et al., 2017; Tanofsky-Kraff et al., 2014). Results suggest that targeting LOC eating in isolation may not be necessary or sufficient to predict outcomes, thus lending further support to the proposed multifactorial model.

Emerging research indicates that mobile technologies, such as ecological momentary assessment (EMA), just-in-time adaptive interventions (JITAIs), biosensors, and digital phenotyping (whereby passive data are collected through regular smartphone use) may hold promise for the assessment, prevention, and treatment of LOC eating in youth (Nahum-Shani et al., 2017; Smith et al., 2019; Torous, Onnela, & Keshavan, 2017). Data collected through mobile technology may enable in-the-moment interventions tailored to each individual based on their specific risk factors or triggers. Implementing these technologies within various populations, (e.g., those at-risk for LOC, those currently reporting LOC, and those in remission) may enable the identification of factors that promote the onset, maintenance, recovery, and spontaneous remission of LOC in youth. These methodologies may also be used to evaluate the natural course of LOC eating in youth, in the absence of intervention. Collectively, these technologies may elucidate risk and protective factors for LOC eating in youth and enable targeted in-the-moment intervention, while simultaneously reducing participant burden.

To select the most appropriate intervention, a precision medicine approach may be useful to match intervention type to the deficits causing the most impairment or distress. Thus, for youth high in negative affect, interpersonal psychotherapy (IPT), which targets interpersonal problems and promotes effective communication skills to reduce negative affect and improve social functioning, may prove especially beneficial (Tanofsky-Kraff et al., 2007). Supporting this notion, despite no group differences in reductions of LOC or BMIz between IPT and a control group (Tanofsky-Kraff et al., 2014), for girls reporting high baseline social-adjustment problems or anxiety, IPT was associated with a steeper decline in BMIz and stabilization of adiposity gain at three year follow-up (Tanofsky-Kraff et al., 2017). This finding corroborates data among adults indicating that those higher in psychopathology derive the greatest benefit from IPT (Wilson et al., 2010).

Given its effectiveness in ameliorating shape and weight concerns and reducing rigid restraint among adults with BED (Grilo, 2017; Grilo, Masheb, & Crosby, 2012), cognitive behavior therapy (CBT) may be well-suited for youth with LOC (Debar et al., 2013) who experience body dissatisfaction (Gall et al., 2016) and binge-restraint cycles (Goldschmidt et al., 2008). Dialectical behavior therapy (DBT), which reduces psychopathology by improving adaptive emotion regulation skills, has also shown preliminary promise among youth with binge-eating (Kamody, Thurston, Pluhar, Han, & Burton, 2019). Given the proposed framework which posits that deficits in self-regulation may be central to the development of BED, additional research is needed to elucidate the effectiveness of DBT in improving emotion regulation and reducing LOC eating among youth.

Among youth for whom food-related expectancies and reward sensitivity are most salient, treatments focused on appetite awareness and increasing sensitivity to hunger and satiety may be well-suited (Boutelle et al., 2011). To reduce attentional biases to food, attention retraining, which involves repeated cue exposure to palatable foods in order to reduce the salience of and anticipatory response to these foods, may be appropriate. Preliminary data indicate that treatments focused on appetite awareness and reduction of attentional biases may be effective at reducing binge-eating (Boutelle, Kuckertz, Carlson, & Amir, 2014; Boutelle et al., 2011), though their effects on body weight are unclear. Though slightly less effective than CBT and IPT, behavioral weight loss (BWL) reduces binge-eating substantially, produces greater weight loss, and is more widely available than the “specialist” psychological interventions (Grilo et al., 2011; Wilson et al., 2010). Thus, for youth who report infrequent or “uncomplicated” disinhibited eating that is primarily due to environmental factors, lack of nutritional knowledge, or occasional difficulties with self-regulation, BWL may serve as a cost-effective and easily disseminable first-line treatment. Given their scalability and cost-effectiveness, stepped care interventions delivered via smartphone also show promise for the prevention of eating disturbances among youth (Kass et al., 2017). If such approaches prove useful in reducing LOC and preventing BED and adult obesity, future research should prioritize the identification of optimal assessments to match specific interventions with those youth who will derive the most benefit, and the determination of which interventions best address associated symptoms such as body dissatisfaction.

Conclusion

In accordance with the proposed model, research should prioritize identifying the subset of youth for whom LOC will persist and/or become full-syndrome BED. Given the multifactorial model which identifies negative affectivity, reward responsivity, and alterations in executive functioning as putative risk factors for BED, screening should be similarly multifaceted. Following an assessment of underlying factors (e.g., genetics, metabolic anomalies, environment), it may be useful to screen youth who report LOC eating for other disinhibited eating behaviors, such as emotional eating, as well as relevant mood and cognitive-related risk factors. Prospective data are needed to predict naturalistic outcomes in the absence of treatment. Ultimately, elements of adult BED treatments may be adapted and disseminated to decrease exacerbated eating and weight problems in youth with LOC eating.

Public Significance:

Loss of control eating, which is prevalent among youth, is prospectively associated with excess weight gain and may be a precursor to binge-eating disorder. A conceptual model suggests mechanisms for study to identify youths at greatest risk for adult binge-eating disorder and obesity. The utility of adult binge-eating disorder treatments for preventing exacerbated disordered eating and excess weight gain in youth is discussed.

Biographies