*1.3. From the Health Psychologist's Point of View: Toward a More Comprehensive Psychological Approach to Food Addiction*

Contrary to the categorical approach of the medical or psychiatric practices, psychology considers that eating behaviors can be mapped onto a continuum ranging from normal to disordered eating, prompted by multiple environmental, contextual, and individual factors [89,90]. For instance, environments constantly influence unhealthy food choices and overeating through food cues—sights, sounds, and smells—associated with palatable food [91], which may undermine the self-regulatory capacity in obesogenic environments [92]. It is recognized that problematic eating behaviors—such as binge eating episodes, overeating, and (failed) cognitive restriction—are not limited to psychological disorders and tend to increase over time in the general population [93].

Studies have found that 7.2% to 13% of the population currently engage in regular binge eating episodes [94]. Another study found that their prevalence increased six-fold from 1998 (2.7%) to 2015 (13.0%) in the adult general population [95]. In addition to these binge eating episodes, eating in response to specific emotional cues was investigated in relation to weight gain [96,97], ED [98,99], and psychiatric and addictive disorders [100,101]. However, this behavioral response is common in normal-weight women, as half of the female students participating in our study reported overeating in response to anxiety in the last 28 days, and 4 in 10 in response to loneliness, sadness, and happiness [102]. These intermittent overeating episodes were used as a time-limited response to emotional states and negatively correlated with alcohol use, which suggests two distinct and somewhat exclusive ways of coping for negative emotions. The Three-Factor Eating Questionnaire Revised, 18-item (TFEQ-R18), measures the cognitive and behavioral components of eating [103], which originate from obesity research but are present in other populations. It includes three subscales: (1) Cognitive Restraint (conscious restriction of food intake in order to control body weight or to promote weight loss) comprised of six items (e.g., "I consciously hold back at meals in order not to gain weight"), (2) Uncontrolled Eating (tendency to eat more than usual due to a loss of control over intake accompanied by subjective feelings of hunger), comprised of nine items (e.g., "When I see a real delicacy, I often get so hungry that I have to eat right away"), and (3) Emotional Eating (inability to resist emotional cues), comprised of three items (e.g., "When I feel blue, I often overeat"). In our study, inability to resist emotional cues outweighed other cognitive components of eating which, again, suggests that overeating is a common tendency to cope with negative emotions. Moreover, while problematic eating behaviors were initially approached independently, they may interact and/or co-exist in complex patterns.

Inmany cases, overeatingmay be a paradoxical consequence of attempts at caloric restriction [104,105], and overlaps exist with emotional overeating and binge eating episodes, as studies showed a direct relationship between binge eating disorder (BED), stress, anxiety, and anxiety proneness [106,107]. However, outside of bulimia nervosa studies, much of the theoretical and empirical binge eating research to date has not directly addressed the role of anxiety [108]; even less has addressed the role of other emotional states such as depression, boredom, or fatigue. Personality may also have a structural albeit overlooked role in problematic eating. For instance, a recent study provided a phenotypic characterization of the FA construct by conducting a clustering analysis of FA in patients with eating disorder and obesity [109]. They found the highest FA symptoms in the "dysfunctional clusters", characterized by more dysfunctional personality traits, greater impulsivity, and more general psychopathology. Conversely, the "adaptive" cluster presented with more functional personality traits and low levels of general psychopathology, as well as the lowest levels of FA. This suggests that FA in the adaptive cluster may be the result of different factors than in other clusters, which could have important implications for treatment. Another study showed that emotional eating was strongly

positively associated with neuroticism, particularly impulsiveness and depression [110]. External eating was likewise mainly associated with the characteristics of impulsiveness (e.g., tendency to act impulsively under strong negative and positive affective experiences, to act on the spur of the moment without regard for the consequences, to enjoy activities that are exciting or novel, etc.) and lower self-discipline [111]. Restrained eating was, on the other hand, related to higher conscientiousness, extraversion and openness, and lower neuroticism. These results imply that poor self-control seen in impulsiveness and lower self-discipline was most important for eating due to negative emotions as well as in response to external food stimuli. Attempts to control food intake and body weight seen in restrained eating were associated with more character strengths and ambitions and also a more outgoing personality style with more stable emotions.

In this regard, the lack of mental stimulation could constitute a significant vulnerability factor for excessive eating [112] and drinking [113]. Som et al. found a higher proportion of food addiction in unemployed patients [66], and our previous work showed a greater predisposition to boredom in patients with excessive drinking [113]. One possible explanation is that some vulnerable people use these compulsive behaviors to cope with excessive lack of internal and/or external stimulation in their daily lives, which may increase the risk of addiction and jeopardize their social and professional functioning. This fits with the definition of one of the eleven diagnostic criteria of addiction in the DSM-5: giving up important social, occupational, or recreational activities because of substance use. Secondly, FA includes the negative feelings following compulsive eating, typically guilt and shame, which are also commonly reported amid overeating episodes in the general population [95]. One possible explanation is that negative feelings after overeating episodes come from the social stigma attached to weight issues [114] rather than from the overeating episode itself. This emotional response to internalized weight stigma could explain the high proportion of FA diagnosis in obese patients, although, in most cases, obesity is the result of poor dietary habits rather than compulsive eating [105,115]. Accordingly, a large part of the FA syndrome, as assessed by the YFAS 2.0, could be seen as a context-dependent pattern of problematic eating behaviors and negative feeling, existing in various forms and intensity, in the general population as well as in patients with chronic conditions, independently of any psychiatric disorders. Finally, tolerance and withdrawal are the only symptoms specific to addictive processes in FA, since they are unrelated to environmental and individual factors and therefore possibly those distinguishing FA from highly frequent problematic eating behaviors. This would be consistent with a review on FA [4], concluding that behavioral and substance-related aspects of FA appear to be intertwined, but the substance (highly palatable food) component may be more salient to the diagnostic classification of this phenomenon than the behavior (eating).

From this perspective, pharmacological criteria, namely craving towards palatable food and withdrawal symptoms, could constitute the main—if not the only—solid indicators of FA, possibly co-existing with varying patterns of problematic eating behaviors, negative feelings, and social disturbances. Investigating their relative contributions to FA, together with their interactions with social environments, unhealthy eating habits, and clinical outcomes, could contribute greatly to the understanding of FA developmental history. However, the YFAS conception study adopted confirmatory factor analysis (CFA), a theory-based approach that can only estimate the extent to which questionnaire data fit the theoretical single-factor structure derived from DSM-5 criteria. The main advantage of CFA lies in its ability to help researchers to bridge the frequent gap between theory and observation. One disadvantage of CFA is that secondary factor loadings are not part of the output [116]. This may lead to the assumptions that (1) all items belong to the same single-factor variable, by simply ignoring the other possible structural hypotheses, and (2) they are equally important for characterizing FA, even though their relative contributions (loadings) to the latent variable are quite heterogeneous in most studies. It must be noted that elevated loadings are expected in construct validation studies since summary scores (or, in the present case, symptoms scores) are computed for clinical or research purposes [117]. Most analyses were performed on the 11 binary diagnoses instead of the original 7-point Likert items, which limited the total variance to be analyzed. This approach is

clearly suboptimal in the case of a newly devised instrument, being psychometrically investigated for the first time. Consequently, the extent to which each of these behavioral, psychological, and social disturbances contribute to FA is still unclear. Unraveling these complex relationships warrants data-driven approaches that establish the data's underlying structure by addressing a wide range of candidate hypotheses, i.e., exploratory factor analyses. This could allow for a more comprehensive description of FA as a biopsychosocial construct lying on a continuum from normal to disordered eating and therefore earlier identification of high-risk profiles.
