**1. Introduction**

Obesity is a multifactorial disease that is currently considered a public health problem. Specifically, the prevalence of obesity has increased rapidly in childhood and adolescence (5–19 years old), with more than 100 million affected in 2016 [1]. Although recent literature has suggested maintenance, even slight decrease, of childhood obesity prevalence in Spain, it continues to be one of the highest prevalence of obesity and overweight in Europe [2].

These high prevalence figures are partly because there is currently no effective treatment for obesity, which can be generalized to most children and adolescents [3]. Recently, the American Association of Clinical Endocrinologists [4] and the European Association for the Study of Obesity [5] has proposed a new diagnostic term, namely "adiposity-based chronic disease" (ABCD). The phrase "adiposity-based" is justified because the disease is primarily due to abnormalities in the mass, distribution, and/or function of adipose tissue. This fact reinforces the concept of obesity as a chronic, complex, multifactorial, and incurable disease at present. For this reason, it is important to continue exploring the different roots of obesity since this is the only way to develop individualized treatments that ensure a higher success rate.

Eating behaviors stand out as crucial proximal determinants of body weight and the motivation to eat since childhood [6,7]. Eating styles represent dispositional tendencies of food intake and have been closely related to obesity in adults [8] and children and adolescents [9,10]. According to the three main theories about impaired food consumption (i.e., the external, psychosomatic, and restraint theories), three predominant eating styles have been identified [11]. External eating refers to the predominance of external environmental factors, such as smell or sight, as determinants of eating behavior, independently of the internal bodily state (i.e., feeling hungry or satiated) [12,13]. Emotional eating emphasizes the influence of emotional factors so that eating behavior is prompted in response to emotional arousal states, such as fear, anxiety, or anger, again without considering internal physiological signals of hunger and satiety [14]. Restraint eating concerns the attempt to control eating behavior through cognitive control and suppression of internal hunger signals to lose weight or avoid weight gain [13].

Previous research has shown that emotional eating can be considered an "obese" eating pattern, since it has been consistently found that higher emotional eating is related to overweight and long-term weight gain in both adults [8,15,16] and children [17], although the prevalence in childhood (7 to 12 years old) is actually low [11]. Emotional eating seems to emerge in adolescence [18,19] and is more prevalent in females than in males [20].

Restrained eating is also conceptualized as an "obese" eating style [21]. Overall, the distinctive features of restrained eaters are restricting food intake for long periods of time to lose or maintain weight and expressing dissatisfaction towards their body size and/or shape [22]. According to the restraint theory, when the cognitive control of eating is disrupted, restrained eaters tend to show a more disinhibited behavior, increasing their food intake and overeating [23]. Hence, although paradoxical, the involvement of restrained eating in obesity, as well as eating disorders, is not surprising [24].

Unlike emotional and restrained eating, some studies suggest that external eating might not be an "obese eating" pattern (i.e., an eating tendency related to a higher body mass index (BMI)) since levels of external eating seem to be similar between normal-weight and overweight individuals [8,25]. Moreover, it has been argued that external eating could represent an adaptive response that helped survive whenever food is available in periods of food shortage [26]. However, it should be noted that these findings contrast with evidence indicating that overweight and obese individuals show poor interoceptive abilities (i.e., awareness of internal bodily cues accompanying homeostatic states such as satiety) [27], which might predispose them to rely on external cues rather than internal signals of satiety and, therefore, to less adaptive eating behaviors [28]. In this line, Mata and colleagues found that external eating was positively related to the insula activation—which is considered a crucial brain hub for interoceptive processing [29]—in adolescents with excess weight, whereas this relationship was negative among healthy weight adolescents [30]. These results suggest the presence of an altered relationship between insula function and interoceptive/exteroceptive processing in adolescents with excess weight. Moreover, it is known that children with higher reward sensitivity (including sensitivity towards external cues of appetitive stimuli) are more vulnerable to become overweight [31], although this relationship is not straightforward. Specifically, external eating and food responsive behavior mediate the relationship between weight and reward sensitivity in childhood [32].

In the regulation of appetite, not only the central nervous system (CNS), but the adrenal glands, the pancreas, and the gastrointestinal tract are involved, and the adipose tissue also plays a relevant role [33]. Adipose tissue is a metabolically active organ involved in multiple biological processes and communicates through the secretion of peptides and hormones, known as adipokines [34]. As mentioned above, obesity is characterized by a pathological expansion and/or unhealthy distribution of body fat, producing adipose tissue dysfunction. This dysfunction produces a disbalance in the homeostasis of children and adolescents with obesity, favoring a profile characterized by resistance to insulin action (with a secondary elevation of insulin), resistance to leptin action (which favors hyperleptinemia), and low adiponectin and ghrelin values [35]. This profile increases the risk of developing associated cardiometabolic diseases [36] but also could be involved in (maladaptive) eating behaviors in childhood obesity. However, no previous studies have examined this relation.

Ghrelin is a well-known gut hormone involved not only in food intake, but also in energy storage, stimulating adipogenesis [37,38]. Specifically, circulating ghrelin levels increase during intake restriction, leading to increased appetite [39] and then fall quickly after ingestion [39]. Thus, it is postulated that one primary role of ghrelin is to act as a meal initiator. However, to our best knowledge, research on its association with eating styles is scarce in normal-weight [40,41] and obese adults [42] and even absent in obese children and adolescents.

The main metabolic disturbance driven by obesity is insulin resistance (IR) [43]. Insulin levels rise and fall rapidly in response to feeding and starvation. These changing insulin levels orchestrate the metabolic switch between anabolism and catabolism. Glucose is the main regulator of insulin, but also other nutrients, hormones, and the autonomic nervous system, influence its serum levels [44].

As insulin, leptin, which is mostly released by adipose tissue, is well-known to reduce appetite and increase energy expenditure. In obesity, leptin resistance is produced due to the excess adiposity, and, therefore, leptin does not properly reduce food consumption, leading to increased body weight [45]. Circulating leptin levels throughout the day do not undergo large variations, showing a circadian rhythm and oscillatory pattern [46].

Adiponectin is also almost exclusively produced in adipose tissue [47]. The mechanisms of this adipokine in appetite regulation are intricate, and it could be an appetite stimulator or inhibiting factor depending on feeding status, the content of glucose in cerebrospinal fluid, and the degree of fatness [48]. In contrast to other adipokines, the circulating levels of adiponectin are inversely proportional to total fat mass [49], and also with fasting insulin concentration and plasma triglycerides, but positively with the plasma cholesterol contained in HDL [50].

Leptin and adiponectin are regulated in an opposite manner in most cases. Children and adolescents with obesity tend to have higher leptin levels and lower adiponectin levels [51]. This unfavorable leptin/adiponectin ratio has been proposed as a functional biomarker of adipose tissue inflammation and seems to be a good indicator of cardiometabolic risk associated with obesity and metabolic syndrome [52]. However, it remains unknown whether this biomarker acts as an underlying factor of eating styles in obesity.

Thus, eating behavior is modulated by both external and internal signals from the body [53]. External cues include environmental factors such as the hedonic properties of food, as well as social factors such as other people's behavior [54]. Internal signals usually refer to physiological processes underlying feelings of hunger and satiety, such as the blood concentrations of ghrelin, glucose, and leptin [55]. At this point, it becomes evident that eating behavior and, ultimately, obesity are highly complex. Disentangling the multiple factors involved and the relationships between them can help us better understand obesity and, therefore, develop more effective prevention and treatment methods, especially in the early stages of human development. As pointed out by Verbiest and colleagues [10], interdisciplinary work is needed to do so.

Despite a certain amount of research focused on the regulatory role of these hormonal biomarkers in eating behavior in obese samples, most studies have considered imminent appetite and food intake as outcome variables. However, little is known regarding the contribution of peripheral obesity-related biomarkers to eating styles, that is, to (relatively) well-established patterns of eating behaviors such as emotional, external, and restrained eating.

The aim of this study is to explore the relationship between several peripheral obesityrelated biomarkers (i.e., ghrelin, IR, and leptin/adiponectin ratio) and eating styles (i.e., external, emotional, and restrained) in a Spanish sample of obese children and adolescents. Specifically, based on preliminary research on this field, it is hypothesized that poorer obesity-related biomarkers in obese children and adolescents, namely lower ghrelin, higher IR, and higher leptin/adiponectin ratio, will be associated with higher scores on external, emotional, and restrained eating.
