Metacognition is a psychological construct that significantly contributes to recognizing and attributing mental states to oneself and others. It involves reflecting and analyzing mental states, ultimately utilizing this awareness to navigate interpersonal conflicts (Semerari et al., 2003). According to this framework, metacognition consists of five distinct sub-functions that interact and can be compromised individually: monitoring, integration, differentiation, understanding others’ mind/decentration, and mastery.

Monitoring involves identifying and defining our mental states, encompassing thoughts, emotions, desires, and motivations. Integration pertains to the general ability to reflect on various mental states, identifying internal conflicts and contradictions. Poor differentiation prevents individuals from maintaining a critical distance from their subjective mental representations. Understanding others’ minds refers to the ability to formulate plausible hypotheses about their mental states, and decentration is the use of that ability to adopt a non-egocentric perspective. Lastly, mastery involves employing psychological information about mental states to address problems of escalating complexity and is connected to activities focused on regulation and control (Semerari et al., 2007, 2012; Carcione et al., 2011).

Researchers and clinicians have theoretically suggested the presence of specific metacognitive failures in patients with eating disorders (EDs) (Olstad et al., 2015; Simonsen et al., 2020). Eating disorders are challenging to treat, and there is still a need to identify psychological mechanisms associated with disordered eating patterns that may serve as important targets for therapeutic interventions. The capacity to monitor, interpret, and regulate mental states is a crucial factor influencing emotion regulation processes related to eating behaviors (Vann et al., 2014; Leppanen et al., 2022). Moreover, research has shown that individuals with adverse childhood experiences are at increased risk for developing emotion dysregulation and eating problems (Michopoulos et al., 2015). While emotion dysregulation has been identified as a mediator between childhood trauma and eating psychopathology (Moulton et al., 2015), it remains unclear whether various patterns of metacognitive functioning are associated with emotional dysregulation and childhood trauma experiences in patients with EDs (Martin and Strodl, 2023).

It is noteworthy that the few existing studies on metacognition in patients with EDs mainly focused on individuals with anorexia nervosa (AN) and bulimia nervosa (BN) (Rothschild-Yakar et al., 2018, 2019; Sacchetti et al., 2019; Kjaersdam Telléus et al., 2023), whereas patients with binge eating disorder (BED) received little attention from researchers (Simonsen et al., 2020). Recently, two studies have investigated the role of metacognition in the severity of binge eating and as a possible treatment target in this population. The first study supported the idea that low self-monitoring metacognition and high negative urgency lead to a worsening of binge severity through the mediation of emotional dysregulation (Aloi et al., 2020); a network analysis study, on the other hand, reported that impaired self-monitoring metacognition and difficulties in impulse control were central nodes in the psychopathological network of BED, while eating symptoms appearing marginal (Aloi et al., 2021).

Although the link between metacognition and EDs has received some empirical support, there is still a need for a greater understanding of heterogeneity in their presentations. In this framework, latent profile analysis (LPA) may represent a suitable approach to better identify and describe the profile of individuals characterized by specific patterns of metacognition (Gibson, 1959; Lazarsfeld and Henry, 1968). LPA serves as a useful technique for conducting person-centered analyses, distinct from variable-centered approaches. Specifically, LPA is a statistical method involving continuous variables (i.e., indicators) to identify hidden subgroups within a population (i.e., latent profiles) using specific variables. This approach operates on the assumption that individuals can be categorized with differing probabilities into various profiles or groups (Williams and Kibowski, 2016).

To the best of our knowledge, LPA was been employed in only a study with patients with EDs to investigate metacognition patterns (Gagliardini et al., 2020). However, the authors used the Mentalization Imbalances Scale (MIS) (Gagliardini et al., 2018), a measure of mentalizing imbalances based on six subscales according to the model proposed by Fonagy and colleagues (Bateman and Fonagy, 2004, 2016; Luyten et al., 2012). In this study, four different profiles of impairments in the dimension of mentalizing were identified. These profiles were heterogeneous in EDs represented in each group and showed notable distinctions across multiple factors such as attachment style, emotion dysregulation, empathy, interpersonal reactivity, and reflective function (Gagliardini et al., 2020).

Metacognition and mentalization are two constructs with some overlap, both referring to the ability of human beings to reflect and reason about their own and others’ mental states, and both attributing a multidimensional nature to these functions (Semerari et al., 2003; Bateman and Fonagy, 2016). Some differences emerge in the definition of metacognition, which includes the ability to use psychological information to cope with distress and interpersonal problems (i.e., metacognitive mastery), with dysfunctions in this area correlated with specific psychopathology profiles (Carcione et al., 2011). Furthermore, unlike mentalization, metacognition does not presuppose activation of the attachment system as the main source of disorders (Dimaggio et al., 2007). In this framework, the DSM-5 emphasizes the importance of reflective abilities. According to the alternative model of personality disorder (AMPD) of Section III, diagnosing a personality disorder requires assessing an individual’s personality functioning, which includes their ability to (1) self-reflect, thereby fostering a stable sense of self and self-direction, and (2) understand others’ perspectives to establish and maintain empathetic and healthy relationships (American Psychiatric Association, 2013).

Building on these previous findings, the present study specifically investigates the role of metacognition according to the model proposed by Semerari and colleagues (Semerari et al., 2007; Carcione et al., 2021). This model suggests that the measurement of this construct includes different and relatively independent sub-functions that could be impaired in several psychiatric disorders (Semerari et al., 2015; Carcione et al., 2019; Aloi et al., 2023).

The primary objective of this study was to explore the metacognition profiles that can be detected in patients with EDs, based on the metacognition sub-functions. The secondary aim was to investigate how these empirically derived metacognitive profiles are associated with some clinical, personality, and ED-related factors. Specifically, we examined whether patients belonging to the different metacognition profiles differ in the levels of childhood maltreatment (CM), negative affectivity, and emotion dysregulation, given the previous evidence that supported their interplay with metacognition and dysfunctional eating (Moulton et al., 2015; Martin and Strodl, 2023). Finally, we investigated which psychological characteristics are associated with the different profiles. Due to the exploratory nature of the present study and the limited existing literature, no specific hypotheses about these metacognitive profiles were formulated.

The current research aimed to explore different profiles of metacognition across the EDs spectrum using LPA. For this purpose, we conducted a study where various models were assessed using goodness-of-fit statistics, and participant classification accuracy was evaluated using standardized entropy. Subsequently, differences among the derived profiles were examined using Welch ANOVA, followed by post hoc tests, and logistic regression models were employed to explore relationships between profile memberships and relevant variables.

A 3-class model emerged for the metacognition domains: “low,” “intermediate,” and “high” functioning. Nearly half of the participants showed high metacognitive functioning, whereas 44.5% of participants showed intermediate metacognitive functioning. Only 7.2% of patients reported low metacognitive functioning. Overall, the present findings support the detection of group heterogeneity in metacognitive functioning among patients with EDs (Rothschild-Yakar et al., 2019; Aloi et al., 2021; Kjaersdam Telléus et al., 2023). Moreover, our findings suggest that the different metacognition profiles are heterogeneous concerning EDs, consistent with previous research (Gagliardini et al., 2020). Specifically, the LF class comprised mostly patients with BN (39%), whereas the IF class included mostly patients with BED (49%). The HF class was the most heterogeneous in terms of diagnostic categories (30% AN-R, 33% BN, and 29% BED). It is worth noting that a trend of increasing severity in all the psychopathological variables (i.e., eating psychopathology, depressive symptoms, emotional dysregulation, dysfunctional personality traits, and CM) was consistently reported across the different profiles of metacognition (LF > IF > HF). Thus, the current findings support the study hypothesis that metacognition may be a transdiagnostic feature across the ED spectrum.

The most striking result of the current study was the association between group membership and psychopathological variables according to the three latent profiles. Interestingly, when partialling out the variance shared by the other CTQ variables, emotional neglect was more strongly associated with the IF profile, while emotional abuse was more strongly associated with the HF profile. There is increasing evidence that childhood emotional abuse is the trauma most associated with disordered eating in adulthood (Michopoulos et al., 2015; Strodl and Wylie, 2020). Our findings add that patients who experienced emotional abuse had higher odds of high metacognitive functioning compared to other groups, when controlling for other forms of abuse and neglect. Therefore, it appears that emotional abuse has the least impact on the metacognitive functioning of individuals with EDs. Further research is necessary to examine whether abusive behaviors, such as constant swearing, yelling, criticism, unrealistic expectations, or unreasonable demands on the child, contribute to the development of mature metacognition among patients with EDs. On the other hand, emotional neglect, characterized by a failure to attend to the child’s emotional needs, may represent a risk factor for developing moderate metacognitive function, especially among patients with BED. Prior research suggested higher levels of emotional neglect than obese patients without BED (Amianto et al., 2018). Thus, further research is needed to examine the link between adverse life experiences during childhood and metacognitive functioning in patients with BED. Interestingly, sexual abuse was not associated with any of the three classes. This result is in contrast with previous literature reporting a high prevalence of sexual abuse among EDs (Afifi et al., 2017) and that a history of sexual abuse could predict poor long-term outcomes (Eielsen et al., 2024). However, the relationship between childhood sexual abuse and eating disorder among adults still requires clarification of the pathways and examination of maladaptive coping responses, with consideration for metacognition patterns. A possible explanation for these mixed findings could be related to the nature of the classes, as the three empirically derived groups were categorized on metacognition features and not according to eating psychopathology. Prior research has also highlighted that the large time gap between childhood sexual trauma and present eating behaviors can convolute total effects (Becker and Grilo, 2011).

Another important finding highlighted in this investigation was the association between detachment PID-5 domain and emotional dysregulation with membership in the LF group. The detachment domain refers to the individual tendency to withdraw emotionally and avoid close interpersonal relations. This result is not surprising, given that few studies have investigated personality traits according to the PID-5 in EDs. Additionally, a recent study found that the facets of anhedonia and depression (traits of the Detachment domain) were strongly associated with all three dimensions of well-being (i.e., emotional, psychological, and social) in patients with EDs (de Vos et al., 2022). In this vein, several studies have demonstrated that patients with EDs reported a pattern of distancing themselves from others, often driven by a desire to protect themselves from potential emotional distress or interpersonal conflicts (Lo Coco et al., 2012; Segura-García et al., 2013; Brugnera et al., 2018; Sivyer et al., 2020; Cassioli et al., 2022; Harris et al., 2023). This interpersonal issue could stem from an impaired metacognition function, specifically the decentration/differentiation subfunction. In fact, this subfunction, which involves the ability to understand others’ thoughts, generate plausible hypotheses about their mental states, and adopt a perspective that is not centered on oneself, is often impaired in several clinical conditions (Semerari et al., 2015; MacBeth et al., 2016; Riccardi et al., 2020; Aloi et al., 2023).

Finally, theoretical frameworks on metacognition have proposed a strong connection between the capacity for metacognitive function and the regulation of emotional states (Euler et al., 2021). Metacognition is considered a crucial process in influencing and adjusting emotional regulation (Rossi et al., 2023), with the roots of this connection believed to be established in early attachment relationships. However, from a psychological perspective, some important differences exist between them. In particular, prior research provided backing for the idea that the capacity for metacognitive functioning might serve as a prerequisite for emotion regulation. This association has been previously linked to increased utilization of adaptive strategies and a reduction in the application of maladaptive emotion regulation strategies (Schwarzer et al., 2021).

Present results underscore the significance of considering a metacognitive function in the spectrum of eating disorders. The current data appear helpful both for diagnostic purposes and also for clinicians who deliver psychological interventions. Evaluating metacognition could aid clinicians and researchers in tailoring psychological interventions more precisely. This method allows for active involvement with patients who have a shared mental disorder, but varying levels of metacognitive skills. On the opposite, it also facilitates communication with patients who have different diagnoses, but who share similar challenges in their metacognitive abilities. Our study suggests that patients with EDs can be classified based on impairments in various dimensions of metacognition, regardless of their specific ED diagnosis. Further, it also recommends that clinicians should consider a multidimensional approach to metacognition when treating ED patients. Future studies should investigate whether the three different identified profiles also correspond to different treatment outcomes, given the initial promising results that metacognitive interpersonal therapy seems to show in the treatment of EDs (Fioravanti et al., 2023).

Raw data supporting the results of the present study will be made available upon reasonable request from the corresponding author.

The studies involving humans were approved by Ethical Committee of “Regione Calabria, Sezione Area Centro” (identifier: Prot. 66/15.03.2018). The studies were conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation in this study was provided by the participants' legal guardians/next of kin.

MA: Conceptualization, Data curation, Methodology, Writing – original draft, Writing – review & editing. AC: Writing – original draft, Writing – review & editing. GL: Writing – original draft, Writing – review & editing. MR: Data curation, Writing – review & editing. EC: Data curation, Writing – review & editing. RF: Data curation, Writing – review & editing. CS-G: Supervision, Writing – review & editing. ML: Methodology, Supervision, Writing – review & editing.