A proposed model of the maintenance and exacerbation of body image and eating concerns in the context of problematic social network use

Highlights

• •Research has framed social network use as a risk for body image and eating outcomes.
• •A model of the exacerbation of body image and eating concerns through problematic social network use is provided.
• •Appearance-motivated use of social networks and appearance-centered online interactions were examined.
• •Social media literacy and positive body image may act as protective factors.

Abstract

Previous evidence supports the existence of a relationship between greater use of highly visual social media and heightened body image and eating concerns. However, to date, the role of problematic social network use has not been articulated in relation to these outcomes and the multiple dynamics underlying these associations need to be further conceptualized. Moreover, models of the exacerbation and maintenance of these body image concerns are lacking, and little attention has been paid to state-level predictors and maintenance factors. The aim of this position paper is to propose model of the maintenance and exacerbation of body image and eating concerns in the context of problematic social network use. Specifically, it is suggested that body image and eating concerns are associated with both generalized and specific problematic social network use through appearance-motivated use of social networks (appearance improvement seeking information, suggested/recommended appearance content engagement, and homogeneous social media experience), as well as appearance-centered online interactions (selfie behaviors, appearance feedback, and compulsive social network checking), leading to state negative affect and internalization of appearance ideals. In addition, four factors are proposed as risk (appearance feedback sensitivity, and social network rumination) and protective (social media literacy and positive body image) factors, respectively. The existing evidence supporting this model as well as future research directions are discussed.

1. Introduction

Body image encompasses a person's beliefs, thoughts, feelings, and actions regarding their physical appearance (Cash & Pruzinsky, 2002). Much clinical research has focused on body image concerns (Rodgers et al., 2023) which include dissatisfaction with body appearance, including shape and weight, and may be associated with disordered eating behaviors such as restrictive or compulsive eating. Increasing work has documented the relationship between the use of highly visual social media platforms and body image and eating concerns in girls and young women (de Valle et al., 2021, Fioravanti et al., 2022). This relationship is understood to be underpinned by the photo-sharing, curating, and interactive affordances of these platforms (Rodgers, 2016, Tiggemann and Barbato, 2018). A disproportionate amount of social media content portrays idealized images that can foster unattainable body standards and unfavorable social comparison among users (Bonfanti et al., 2025, Prichard et al., 2020, Saiphoo and Vahedi, 2019). However, previous research evidence has indicated that mere time spent on social networks does not primarily influence the development of disordered eating or body dissatisfaction (Sharma & Vidal, 2023) and the multiple dynamics underlying these associations need to be further conceptualized and explored. Most importantly, the role of problematic social network use (PSNU) has been partially articulated in relation to body image outcomes and there is no theoretical framework that encompasses the characteristics of this specific negative use of social networks.

2. Problematic social network use

PSNU has been defined as the compulsive and uncontrolled use of social networks characterized by “addiction-like” symptoms (e.g. salience, tolerance, mood regulation, withdrawal, conflict, relapse) (Andreassen et al., 2015) and impairments or negative consequences in users’ everyday functioning (Casale et al., 2023, Kuss and Griffiths, 2017, Shannon et al., 2022), but the construct remains controversial in the field of behavioral addictions (Cheng et al., 2021, Moretta et al., 2022). To date, PSNU has not been officially classified as a mental disorder by international diagnostic systems, and it has been argued that PSNU could be viewed more as a coping mechanism in response to an individual’s negative emotions (Casale, 2020) in the context of the Internet use. Indeed, the Compensatory Internet Use Theory (CIUT; Kardefelt-Winther, 2014) suggests that people with social deficits or mental distress may tend to attempt to regulate their negative emotion by overusing social networks and thereby developing PSNU as a maladaptive coping mechanism. Recent studies have suggested that the core dimensions of PSNU are a lack of self-regulation, in particular related to use of social networks, difficulties in resisting the urge to use social media, the tendency to think obsessively about going online, and a preference for online communication (Svicher et al., 2021).

A number of incrementally useful models have been developed to account for problematic Internet use, that can be applied to the understanding of PSNU. The seminal cognitive behavioral model developed by Davis (2001) adopts a diathesis-stress approach in which psychopathology is a distal necessary predisposing factor for problematic Internet use. The model describes how such distal vulnerabilities may combine with proximal stressors such as lack of social support or increased localized distress, cognitive distortions, behavioral reinforcers, and together lead to the development of problematic social network use as a maladaptive effort to regulate these proximal precipitators. The model was further articulated to include two dimensions, i.e. specific and generalized Internet use. The former describes users who are goal-driven in their Internet use, reporting an uncontrollable dependency on specific content (e.g. gambling or gaming). The latter dimension reflects that not all Internet activities are goal directed and individuals may go online simply to pass time. However, the unique communication context provided by the Internet, together with pre-existing mental health distress, may result in problematic use (Davis, 2001). This model has been adapted to highlight differences between a specific and generalized dimensions of PSNU (Lee et al., 2017). In recent years, a number of additional frameworks have been developed to explain the dimensions of PSNU. However, these frameworks have been found to be poorly integrated, resulting in a lack of a comprehensive view of how PSNU develops (Sun & Zhang, 2021). The Interaction of Person-Affect-Cognition-Execution (I-PACE) model of addictive behaviors by Brand and colleagues (2019) differentiates between global vulnerability factors (e.g., mental health distress or dysfunctional personality traits), which are involved in different types of addictive behaviors, from individual motives or preferences for using specific applications or platforms excessively (e.g., the gratification of social needs could play a central role in problematic use of highly visual social networks) (Brand et al., 2019, Casale and Fioravanti, 2015, Mancinelli et al., 2019). In this way, the model builds upon the cognitive behavioral model by highlighting distal vulnerabilities but incorporates consideration of platform-specific affordances, which make it a useful framework to adapt to PSNU.

The process of development and maintenance of PSNU has been also conceptualized through a fear-driven/compensation-seeking hypothesis and a reward-driven hypothesis (Wegmann & Brand, 2019). In the first hypothesis, it is proposed that poor social competence, high social anxiety and the need to compensate for low interpersonal skills can represent a core motivation for engaging intensively in social networks use, thus triggering problematic use. The second hypothesis proposed that individuals with a high need to belong, high need for popularity, and high social skills may develop high expectancies regarding the receipt of gratifications or rewarding experiences when using social networks and may therefore also be at risk of developing problematic behaviors. In both cases, reinforcement mechanisms are important for the maintenance of PSNU. In terms of predisposing characteristics for PSNU, recent research evidence supported the role of global factors such as high neuroticism, low conscientiousness, low self-esteem, depression, anxiety and insecure attachment; and the role of unmet needs for self-presentation, assertiveness and belonging as well as preference for online social interactions as behavior-specific predisposing variables for PSNU (Fioravanti et al., 2025, Musetti et al., 2022, Shannon et al., 2022, Wu et al., 2024).

It has been argued that PSNU is an umbrella term comprising problematic use of a variety of different online platforms, such as Facebook, Instagram, TikTok and Snapchat, and that it would be worthwhile to identify specific beliefs, emotions, motivations that characterize dysregulated use of different types of social networks (Cataldo et al., 2022, Lee et al., 2017). Most of the previous research on PSNU has focused on the construct of Facebook addiction (Moretta and Buodo, 2018, Ryan et al., 2014), but more recently, highly visual social networks, that is, platforms that allow for the posting and sharing of photos and videos, have gained popularity, particularly among young people (Pew Research Center, 2024). Such platforms are widely used as a source of information on eating, exercising, and on beauty standards with continuous access to advertisements and posts both from peers and content creators (Sharma & Vidal, 2023). Although some research has suggested that problematic use of highly visual social networks such as Instagram or TiKTok may be associated with body image and eating concerns in both adolescents and young adults (Casale et al., 2024, Fioravanti et al., 2024, Mohsenpour et al., 2023, O'Connor et al., 2024), the complex interplay of appearance-related uses of social networks, social network-related problematic behaviors and body image concerns has been poorly conceptualized.

To date, these relationships have been framed within several useful theoretical models that emphasize the role of idealized content on these platforms, that of machine-learning driven personalized recommended and suggested content, and that of conscious and unconscious specific patterns of engagement resulting from high body image and eating concerns (Tylka et al., 2023). These theorical models are generally designed to frame social network use as a risk for body image and eating outcomes, and focus on the development of these concerns. As such, they have been useful in guiding hypothesis-driven empirical research, and working towards the development of interventions capable of supporting positive body image and adaptive eating behaviors in the context of social network use (Gordon et al., 2020). However, these models failed to account for the role of PSNU and the way in which poor regulation of engagement with social media may drive symptomatology. This constitutes an important barrier to identifying just in time strategies as well as policies and practices that could disrupt these relationships, as well as targeting intervention with individuals with pre-existing concerns. Furthermore, these models have also focused on individual differences at the trait level and overall neglected the in-the-moment mechanisms that likely participate in the exacerbation and maintenance of body image and eating concerns through the use of social networks. In line with recent calls for an emphasis on the study of individual quality of social network use in order to focus research attention on its emotional and behavioral consequences (Cataldo et al., 2022), the aim of this position paper was to propose a theoretical model of the processes that underlie the exacerbation and maintenance of existing body image and eating concerns. Although social network use represents a plausible risk factor for the development of body dissatisfaction and disordered eating (Dane & Bhatia, 2023), no previous models have included dimensions of PSNU and the current paper may fill this gap.

3. The relationship between social media and body image and eating concerns: Trait based models of risk and the need for additional explanatory frameworks

The relationship between greater social media use, in particular highly visual social networks, and body image and eating concerns has been framed within integrated models (e.g. Rodgers, 2016), that bring together elements of traditional sociocultural theories (Thompson et al., 1999), cultivation theory (Stein et al., 2021), and uses and gratifications theory (Katz et al., 1973, Whiting and Williams, 2013). Taken together, these frameworks highlight how the preponderance of highly edited, curated, and idealized images of bodies on social media can result in the internalization of appearance ideals that centered thin, lean, toned, and muscular bodies (Rousseau et al., 2017). Moreover, exposure to idealized appearance-focused content can lead to unfavorable upwards appearance comparisons, i.e. comparing oneself to more attractive individuals (Holland and Tiggemann, 2016, Bonfanti et al., 2025), leading to body dissatisfaction and behavioral efforts to bring the body closer to the social ideal through disordered eating behaviors. In parallel, objectification theory, which focuses on the way in which physical attractiveness is overly valued within society, highlights how the tendency to adopt an external viewers’ perspective of one’s body and treat oneself as an object can lead to increased body surveillance and body shame (Fredrickson & Roberts, 1997). Given that most social network platform, especially the photo-based ones, uniquely afford engagement in photo-based activities that include curating and posting images one oneself (selfies) to be viewed by others, social networks have been described as an intrinsically objectifying environment (Bell et al., 2018).

Overall, empirical work has supported the usefulness of these models. Robust evidence has now emerged for the relationship between greater use of photo-based social networks and higher levels of body image and eating concerns. Initial meta-analytic evidence has shown that greater use of social networks was associated with significantly higher internalization of a thin ideal, and especially the use of appearance-related features (Mingoia et al., 2017). Another meta-analysis of 63 cross-sectional studies (Saiphoo & Vahedi, 2019) evidenced a small relationship between greater social network use and higher body image disturbance, with the strongest effect sizes for the cognitive (e.g., internalization of the thin-ideal, that is endorsement of appearance ideals as a personal standard) and behavioral (e.g., disordered eating) dimensions of body image. A more recent meta-analysis of 24 experimental studies reported that exposure to thin-ideal images had a moderate detrimental impact on body image compared to appearance-neutral conditions (de Valle et al., 2021). Regarding specific pathways, appearance comparison has received support as a mediating process in the relationship between social media exposure and body image and eating concerns (Bonfanti et al., 2025, Fardouly and Vartanian, 2016, Tiggemann, Hayden, Brown, & Veldhuis, 2018). In addition, higher frequency of posting objectified self-images was found to be associated with trait self-objectification among young women (Bell et al., 2018).

However, despite the documented usefulness of these explanatory frameworks and the established support for elements of these models with longitudinal data (de Valle et al., 2021, Jarman et al., 2024), important gaps still remain to be bridged. First, these models are largely focused on the increased risk of developing body image and eating concerns through social network use, which neglects the existence of specific mechanisms and processes among individuals who already display high levels of these concerns. This is an important gap as previous work has highlighted that risk pathways and maintenance ones are likely distinct (Stice et al., 2021). Furthermore, these models largely focus on trait level factors, which while important, fail to adequate capture the timescale of processes on social media where individuals may be scrolling through numerous pieces of content per minute, engaging in cognitive and emotional processing, and engaging and interacting with this content in ways that impact their social network experience and their body image and eating concerns. Finally, these models mostly neglect the affective and emotion regulation components that are well-documented in eating disorders and have been identified in social network use in the form of PSNU. For example, Problematic Instagram Use has been conceptualized as an excessive and compulsive online social networking behavior (Fioravanti et al., 2024), which shares common symptoms of behavioral addictions (Andreassen, 2015). According to the cognitive-behavioral model of PSNU (Davis, 2001, Caplan, 2010), social network use can act as a dysfunctional regulator of emotions for people who exhibit a preference for online social interactions, and difficulties in controlling one's social network use, unsuccessful efforts to resist when trying, and obsessive thoughts about going online were found to be core components of PSNU (Svicher et al., 2021). Importantly, individuals with body image and eating concerns have been shown to report higher levels of PSNU (Fioravanti et al., 2024), and it has been suggested that the compulsive engagement with photo-based platforms such as Instagram or TikTok may increasing worries about physical appearance flaws (Simon et al., 2022), and appearance management behaviors (Casale et al., 2024), again highlighting the need for models that are able to capture in-the-moment processes. Thus, the model proposed here aims to start to bridge this gap in explanatory frameworks for the exacerbation and maintenance of body image and eating concerns in the context of a problematic social network use.

4. The proposed model

4.1. Description of the proposed model

The model of exacerbation and maintenance of body image and eating concerns through social network use proposed here focuses on the implication of appearance-motivated uses of social network use, appearance-centered online interactions, dimensions of both generalized and specific problematic social network use, and their relationship to state negative affect and internalization of appearance ideals, two well-documented risk factors for body image and eating concerns (Stice et al., 2017) (see Fig. 1). As a note, we define here online interactions as engagement with others on social networks (by posting, sharing, commenting, etc), as opposed to engagement that here will refer to actions of engagement with content itself (watching, clicking, liking, hovering over, etc.). Clearly this distinction is somewhat forced, however the aim is to distinguish between actions that are mostly “heard” by users versus those that are “heard” by platform algorithms. In addition, appearance related content is conceptualized as social network content that directly (e.g. a fitness video) or indirectly (e.g. an image of a thin and toned individual in exercise clothing) centers socially constructed appearance ideals, the importance of achieving them, and the malleability of appearance through diet, exercise, and the consumption of beauty products and services.

Fig. 1.

Model of the maintenance and exacerbation of body image and eating concerns.

In brief, the model (see Fig. 1) describes the interrelations among appearance −motivated use of social network, appearance centered online interactions, both defined through three subdimensions including dimensions of PSNU, as well as maintenance factors, and outcomes of both body image and eating concerns and generalized PSNU. The model also includes proposed moderating factors, two of which are posited to potentiate the proposed pathways, and two of which are proposed to have a protective function.

Appearance-motivated uses of social network constitute the first element of the model. These have been shown to be higher among those with body image and eating concerns (Rodgers et al., 2021). These types of use include the use of social network for appearance-improvement information gathering, which may take the form of searching for content, following accounts that are dedicated to these themes, and using this content as trustworthy information regarding ways in which to move one’s appearance closer to appearance ideals. Using social networks in this way will inevitably lead to higher levels of associated appearance and weight related content being included in recommended and suggested content by platforms. Individuals who are seeking information are more likely to engage with this suggested and recommended content, which will further promote this as a content category in their recommended content, as well as lead to an increasing homogenization of their online environment in terms of appearance-and-weight related messaging (Harriger et al., 2022). An increasingly homogenous online space will reinforce perceptions of social norms related to appearance and weight, and likely exacerbate pre-existing concerns. Regarding appearance-centered online interactions, the second element of the model, these are behaviors that involve active contributions or behaviors interpreted by others (and platforms) as interactions. Thus, selfie behaviors for example constitute an important online behavior that among individuals with high body image and image concerns may be particularly implicated (McLean et al., 2015). It would be anticipated that individuals with higher levels of concerns might completely avoid posting images of themselves due to high levels of concerns and difficulty in finding images that they feel are acceptable. In contrast, however, individuals with high levels of concerns may also engage very effortfully in the curation of self-images for the purpose of seeking reassurance and positive appearance feedback from others, and external validation of their appearance meeting social standards. This latter group of individuals may also find themselves compulsively checking the engagement around their self-images in terms of likes and appearance-related feedback, which, depending on its nature, could substantially affect mood and body image.

As described above, building on the seminal work of Davis (2001) and Caplan (2010), the usefulness of distinguishing between specific and generalized Internet use may prove to be a valuable lens through which to conceptualize the relationships between body image and eating concerns and PSNU. Applying this model to social network use, it can be argued that in the generalized PSNU individual’s behaviors and cognitions relate to the general use and consequences of social networks, regardless of their specific content (Lee et al., 2017). Thus, social networks may foster problematic use because of their affordances, including opportunities for self-presentation in an extended communication online context, as well as pre-existing individual concerns (Caplan, 2002). As such, even when not specifically tied to social network usage related to appearance (e.g. fitspiration or thinspiration videos), PSNU may develop through the focus on appearance-related elements such as online self-presentation. In contrast, specific PSNU refers to a pattern of PSNU that centers around a particular type of content. In the context of body image and eating concerns, this content might be self-images, or other users’ content related to appearance, eating behaviors, and exercise. Notably, these two different forms of PSNU rely on different motivations, and different extents to which behavior is goal directed, versus aiming to regulate emotions (Davis, 2001).

The proposed model includes both aspects related to specific and generalized PSNU. Regarding the former, compulsive social network checking through anxiety related to appearance feedback may become problematic and uncontrollable leading to increased distress as well as increased exposure to the homogenous online appearance centering content. Second, a habituation effect may occur in the context of appearance feedback through which indices of positive engagement from others (e.g., likes and positive comments) may be perceived as insufficient if they are not increasing, leading to individuals needing ever more positive engagement for the reassuring effects of this feedback to be sustained (Flayelle et al., 2023).

The proposed model additionally includes a number of potential moderating factors, building on models of social media effects on body image and eating concerns (Tylka et al., 2023), as well as extensions of the cognitive behavioral model as outlined by Davis (2001) that situates dysfunctional cognitions as a distal factor influencing risk for PSNU. Specifically, sensitivity to appearance-related feedback and rumination are suggested to strengthen the relationship between receiving such feedback and body image and eating concerns. Indeed, individuals who are highly sensitive to feedback may experience significant mood and body image fluctuations depending on the nature of the feedback they receive and how they interpret it. In contrast, positive body image and social media literacy are positioned as protective factors (Tylka et al, 2023), in that they would help to decrease the strength of these relationships. Thus, individuals with a positive body image may be less interested in feedback about their appearance and may be less affected by comments or likes about their appearance on social networks. Similarly, higher levels of social media literacy may reduce the extent to which individuals engage with appearance-related feedback, even when they seek it, and protect them from its effects on mood and body image. While this model is theoretical, some of the relationships it proposes have started to gain research attention. Below we provide a non-exhaustive summary of some of the emerging literature to support the hypothesized relationships.

4.2. Appearance-motivated uses of social media

As previously suggested, and predicted by theoretical models (Rodgers, 2016) appearance-motivated uses of social network have been shown to be associated with body image and eating concerns. Thus, for example, among adolescent girls and boys, self-reported appearance motivations for social media use (e.g. “knowing how to improve how I look”, “get my friends’ opinion on how I look”) revealed a moderate sized association with higher levels of thin-ideal internalization (Rodgers et al., 2021a). Seeking appearance-related information, and specifically how to improve appearance is an important aspect of appearance-motivated media, and social network use. In traditional media contexts, information seeking related to appearance has been shown to be associated with higher levels of appearance ideal internalization as well as body image concerns in a range of populations including female college students, young men, and female eating disorder samples (Calogero et al., 2004, Pritchard and Cramblitt, 2014, Thompson et al., 2004). On social media, preliminary longitudinal evidence has also emerged to support these relationships. For example, among adolescents, reliance on social network for appearance-related information has been shown to be associated with higher levels of thin-ideal internalization and athletic-ideal internalization over time, as well as more frequent peer conversation related to appearance (Rousseau, 2024). In addition, among preadolescents information seeking use of social network was associated with greater dysfunctional beliefs and overvaluation of appearance over time (Rousseau et al., 2018). Although not explicitly assessing the information seeking element, research has shown that young women who view fitspiration social network content (content focused on promoting the pursuit of a thin and toned body through diet and exercise) report higher levels of disordered eating and compulsive exercise compared to those who do not engage with fitspiration content (Wu et al., 2022). Moreover, daily exposure to fitspiration images was associated with higher rates of growth of appearance comparison and negative mood (Fioravanti et al., 2023). Further supporting this relationship, among a large sample of adult women, following a nutrition and fitness influencers was related to higher levels of disordered eating as compared to an entertainment influencer (Bocci Benucci et al., 2024). Thus, using social media for the purpose of seeking information related to appearance has been shown to be associated with higher body image concerns (Griffiths et al., 2018) as predicted by the proposed model.

Consistent with this interest in appearance-related information, individuals with body image and eating concerns are likely to engage with appearance-related suggested and recommended content to a greater extent as compared to their peers with lower concerns. Limited research to date has explored user levels of engagement with recommended and suggested content specifically, and the data on how this engagement may differ across levels of body image and eating concerns are even more scarce. Nevertheless, preliminary research has shown that young people’s health-related affect and preoccupations impact the ways in which they interact with recommended and suggested content (Goodyear & Armour, 2018). In their synthesis across three projects including over 1600 young people, these authors describe how body image concerns may moderate engagement in diet and exercise content online (Goodyear & Armour, 2018). Specifically, their findings suggest that among some youth body dissatisfaction might override an understanding of the fact that appearance-related recommended content is inappropriate to them, leading to appearance comparisons and the perspective that the bodies portrayed were in fact achievable (Goodyear et al., 2018).

This heightened engagement with appearance-related recommended content is likely to drive increasing predominance of this type of content in users feed, and a progressive homogenization of the content users with high body image and eating concerns are exposed to. In support of this, recent research focused on the platform TikTok revealed that users with eating disorders were recommended significantly more appearance-oriented (+146 %), dieting (+335 %), exercise (+142 %), and harmful eating disorder videos (+4343 %) as compared to a group of individuals without eating disorders (Griffiths et al., 2024). Moreover, a dose dependent relationship emerged such that greater eating disorder severity was associated with the increased bias towards these categories of recommended content. Similarly, other work focusing on exposure to TikTok videos has confirmed an association between higher eating disorder symptom severity and higher rates of appearance concerns (Bissonette Mink and Szymanski, 2022, Dondzilo et al., 2024, Pryde and Prichard, 2022).

Thus, although empirical work documenting the relationship between body image and eating concerns and online behaviors is still scarce, likely due to methodological barriers as discussed later, the emerging evidence supports the fact that individuals with higher concerns exposed to and engage with more appearance-related content. Moreover, this increasing exposure and engagement likely reinforce each other over time leading to highly homogenous social network experiences. Such a homogenous experience would be expected to result in exacerbated levels of body image and eating concerns (Rodgers, 2016). Furthermore, as predicted by our proposed model, this overexposure would also lead to increased appearance-centered online interactions. Indeed, prior research has shown that more appearance social network exposure was associated with more selfie-behavior (selfie-taking/posting and selfie-investment) via thin/athletic-ideal internalization and appearance comparison on social networks (Rousseau, 2021).

4.3. Appearance-centered online interactions

In addition to the exposure to increasingly homogenous and appearance-centered content resulting from the appearance-motivated uses of social networks described above, users may contribute self-images and content to social networks, and selfie-related behaviors have received increasing research attention (Sung et al., 2016). Prior research has shown how viewing, posting, and editing self-images may be linked to appearance concerns, negative affect and low self-esteem among adolescents and young people, especially among girls (de Lenne et al., 2020, Vandenbosch et al., 2022, Veldhuis et al., 2020). Moreover, higher levels of photo investment (concern about the quality of their selfies and the effort spent in selecting photos to share online) has been shown to be associated with higher levels of thin-ideal internalization, body dissatisfaction, overvaluation of shape and weight, and dietary restraint among adolescent girls (McLean et al., 2015). This study also assessed photo manipulation (i.e. the editing of a photo prior to posting it online) which was found to be associated with higher thin-ideal internalization and body-related concerns (McLean et al., 2015). It is worth noting that these findings on the relation between selfie editing and body dissatisfaction have been replicated with samples of young adults (Cohen et al., 2018, Lonergan et al., 2019). Moreover, effort to choose, edit and monitor responses to selfies may negatively impact mental health (Lonergan et al., 2019). Although most of these previous findings are based on cross-sectional data, promising experimental evidence also supports that taking and editing of selfies may lead to increased state facial dissatisfaction (Tiggemann et al., 2020). In addition, higher levels of photo editing were associated with increased facial dissatisfaction, as were thoughts concerning making oneself look better. Moreover, longitudinal studies suggest that taking and editing, but not posting, selfies may be harmful to body image (Vandenbosch et al., 2022, Wang et al., 2021). Alongside this accumulating evidence that supports a relationship between selfie behaviors and higher body image concerns and internalization of appearance ideals, there is initial evidence that selfie-taking may be a new form of body checking in eating disorders. For example, it was shown that the more offline selfies individuals with eating problems took, the more body checking behaviors they reported which, in turn, was associated with more severe ED symptoms (Yellowlees et al., 2019).

Given the interactive nature of social networks, users frequently experience feedback about others’ reactions to posts through “likes”, “follows” and “comments”. The Transactional Model of Social Media and Body Image Concerns (Perloff, 2014) describes how the features of visually oriented social networks can make these platforms suitable for seeking appearance related gratifications, by allowing individuals to seek approval of, and positive feedback on, one’s self-presented appearance on social network. According to the tripartite influence model (Thompson et al. 1999), the reception of peer appearance-related feedback can impact individuals’ body image in the short term either negatively when receiving positive feedback, or positively when receiving affirming and positive feedback. However, it has also been noted that receiving positive appearance feedback can be associated with heightened appearance shame, in particular in objectifying environments (Tiggemann & Boundy, 2008), which social networks arguably are, as well as body dissatisfaction (Herbozo et al., 2017). Moreover, peer appearance-related feedback can pressure adolescents to conform to an appearance ideal that they do not meet (de Vries et al., 2016).

Consistent with these predictions, it has been shown that Instagram “likes” are related to facial dissatisfaction among women (e.g., Tiggemann, Hayden, Brown, & Veldhuis, 2018). Additionally, among female college students, reported greater importance of likes and comments was associated with greater drive for thinness and body dissatisfaction, while in male students greater importance of likes and comments was related to drive for muscularity attitudes and behaviours (Seekis et al., 2020). Similarly, an experimental study (Vendemia & DeAndrea, 2021) showed that female undergraduates who received positive appearance comments on online photos of themselves reported greater self-objectification than those receiving positive non-appearance comments. Finally, a small number of studies using ecological momentary assessment have shown that negative appearance-based comments predict greater levels of state body dissatisfaction, both in clinical (e.g., individuals with eating disorder) and non-clinical samples (Liu et al., 2022, Mason et al., 2018). In addition, momentary anxiety mediated the prospective association between appearance-related stress and eating disorder behaviors (Mason et al., 2018). These findings have been replicated to some extent among adolescents, who report frequently receiving positive comments about both appearance and non-appearance (Nesi and Prinstein, 2018, Schreurs and Vandenbosch, 2022). Indeed, receiving positive appearance comments was associated with body image concerns (Fatt & Fardouly, 2023). Interestingly, it has been suggested that adolescent females may receive more positive appearance comments on social networks than males. However, both adolescent boys and girls overall report rarely receiving negative appearance or non-appearance comments on social networks (Fatt and Fardouly, 2023, Paddock and Bell, 2024).

As previously mentioned, a growing body of literature has documented a relationship between PSNU and body image and eating concerns (Fioravanti et al., 2024, Imperatori et al., 2022, Kircaburun et al., 2020, Sfeir et al., 2023). The relationship with disordered eating has shown to be both direct (likely due to emotional regulation pathways), as well as indirect via body image concerns (Imperatori et al., 2022, Mannino et al., 2021, Kircaburun et al., 2020). To date, less work has been superficially conducted on compulsive social network checking behaviors as related to appearance-content, however, emerging findings suggest that individuals with higher body image and eating concerns may also display more problematic appearance-centered interactions. Thus, for example, among both male and female adolescents, higher level of concerns regarding one’s appearance in photos on social networks was found to be associated with greater levels of compulsive social network use (Boursier et al., 2020). Based on these initial findings, the relationship between compulsive social network checking as a feature of PSNU and ED symptoms warrants further investigation.

4.4. Generalized and specific problematic social network use

As described above, the proposed model includes both generalized and specific problematic social network use. Generalized, or content aspecific problematic social network use can be conceptualized as including aspects related to the salience of social network as well as aspects related to tolerance and withdrawal symptoms (see Fig. 1). Consistent with Caplan's (2002) model, it can be hypothesized that the unique communication context provided by social networks, together with pre-existing eating or body image concerns, may facilitate the development of problematic use. Social network allows users to interact with other creators by following their profiles, and to repeatedly consume user-generated content by liking, sharing, or commenting (Perloff, 2014, Vandenbosch et al., 2022). It is also likely that PSNU is exacerbated by algorithms that provide viewers with personalized content designed to keep them engaged for longer periods of time (Harriger et al., 2022). In the context of body image and eating concerns, the salience of social networks, regardless of preferred and recommended content, may be associated with eating concerns as well as with negative mood. Indeed, body image and eating concerns and PSNU have been shown to co-occur (Hinojo-Lucena et al., 2019), perhaps due to underlying emotional and behavioral regulation difficulties, and thus, eating concerns may be associated with increased generalized problematic social network use, which in turn might drive higher negative affect. Previous research has documented a relationship between PSNU and body image concerns (Simon et al., 2022), as well as an indirect pathway via higher thin-ideal internalization (Delgado-Rodríguez et al., 2022). Moreover, compulsive internet use has been shown to be related to appearance management behaviors (Casale et al., 2024).

Appearance-specific PSNU has yet to be investigated, however, we propose that such a framework could be adopted to conceptualize some of the factors understood to be implicated in the relationships between social network use and body image and eating concerns. For example, individuals with eating and body image concerns may be goal-directed in their social network use and develop an addiction to specific content, such as watching dieting or fitspiration videos. Moreover, compulsive checking of social networks and the internalization of the thin-ideal likely may increase salience of social media, leading to its increased use. Similarly, the need for appearance feedback and seeking of reassurance would likely engage processes similar to those of tolerance and withdrawal. To the best of our knowledge, no empirical work has yet investigated PSNU specific to appearance related content from this lens, and this would be an important and novel contribution.

4.5. Moderating factors

The proposed model also includes several moderators, namely appearance feedback sensitivity, rumination, positive body image and social media literacy. These factors as proposed to act as potentiating risk factors (appearance feedback sensitivity and rumination) and protective factors (positive body image and social media literacy). While to date little empirical work has examined these dimensions as moderators of the proposed relationships in the model, other theoretical work has identified them as being potential moderators to consider in the context of the effects of social media use on body image (Tylka et al., 2023). Moreover, correlational work has started to establish their importance as relevant dimensions in the relationships between social network use and body image and eating concerns. Thus, for example, sensitivity to appearance related feedback has been shown to be shown to be associated with higher body image and eating concerns, and with greater intensity of social network use among adolescents and young adults (Hawes et al., 2020). Among adolescents and young adults, rejection sensitivity has also been shown to be associated with increased online appearance preoccupation over time (Zimmer-Gembeck et al., 2024). Similarly, rumination related to social network content, that is the tendency to repetitively think about one’s social network posts, related situational factors, and consequences of those posts (Lannin et al., 2021) has been shown to be associated with higher levels of socioemotional distress in the context of social network use among adolescents (Lannin et al., 2021). Empirical work has also suggested that co-rumination in the context of peer relationships is associated with higher levels of anxiety and depression among adolescents (McCauly Ohannessian et al., 2021). In the context of body image and eating concerns, rumination is well documented as a concurrently and prospective associated with higher levels of symptomatology (Smith et al., 2018). Correlationally, higher levels of rumination have been shown to be associated with higher social network use and higher appearance comparisons (Teo and Collinson, 2019). Similarly, in an experimental study, social network related rumination strengthened the relationship between exposure to diet related videos and state levels of restriction. These findings together support the fact that sensitivity to appearance-related feedback and social-network rumination might be potentiating factors in the proposed model.

Previous work has emphasized the importance of social media literacy in the relationship between social media use and body image and eating concerns (Paxton et al., 2022, Rodgers et al., 2022, Rodgers et al., 2024, Tylka et al., 2023). Contemporary models of social media literacy move beyond the transposition of critical skills as applied to traditional media. In this way, updated models of social media literacy include increased understanding of the unrealistic nature of images posted to social media; decreased perceived proximity to individuals portrayed on social media and an understanding of the intent of the poster; protections against searching for and exposure to appearance-focused content; and the contribution of online content in ways that are not focused around appearance (Rodgers et al., 2024). Empirical data regarding the potential protective role of social media come from qualitative and experimental studies, but overall the evidence is still somewhat weak (Paxton et al., 2022). Thus, for example, two experimental studies have found that self-reported social media literacy moderated the effects of exposure to experimental stimuli on body image with an overall protective effect (Rodgers et al., 2021, Tamplin et al., 2018). However, intervention work aiming to modify social media literacy to protect body image and eating behaviors in the context of social network use has met with limited success (e.g. Gordon et al., 2020).

Positive body (Tylka & Wood-Barcalow, 2015) may be considered as a dimension of body image, and therefore an outcome. However, previous work has conceptualized positive body image as a protective factor against body image and eating concerns (Levine & Smolak, 2016). In the context of social networks, body appreciation, one of the main constructs associated with positive body image, has been examined but not supported as a moderator of the effects of exposure to body positive social media content on body image and mood (Cohen et al., 2019, Giorgianni et al., 2020). Thus, while theoretical work supports the positioning of social media literacy and positive body image as protective factors in the model, the extant empirical support is more limited.

4.6. Summary of support for the model to date

In sum, although no tests of the integrated model have been conducted, emerging evidence exists to support the proposed pathways, including the tendency for individuals with greater body image and eating disorders to seek, be exposed to, and receive recommendations for more appearance-related content, and for this content to be engaged with to a greater extent leading to an increasingly homogenous social media experience. In addition, greater engagement with appearance-related content, and online appearance-related interactions have also been shown to be associated with heightened body image and eating concerns as well as internalization of appearance ideals and negative affect. Although evidence has supported to co-occurrence of generalized problematic social network use and body image and eating concerns, to date no attempts to conceptualize PSNU specific to appearance content have been formulated. Emerging evidence exists regarding the potentiating role of appearance feedback sensitivity and social network rumination, but the support for the proposed protective factors, namely social media literacy and positive body image is scant.

5. Future directions

In light of the prevalence of social networking and the pervasive challenges surrounding body image and eating concerns, the refinement of models elucidating the exacerbation and perpetuation of these concerns within the ambit of social networks constitutes a pivotal undertaking. Although not systematic, the overview above highlights important gaps in the empirical literature assessing the relationships described in the proposed model, as well as the integrated model itself. Below several important directions for future research are outlined.

5.1. Measurement issues

A first important direction is the developing of appropriate and sensitive measures of social network use. As highlighted elsewhere (Jarman et al., 2022), our understanding of the relationships between social network use and body image and eating concerns is limited by an overreliance on self-report measures that are frequently ill suited to assessing the different ways in which individuals engage with social networks in terms of their interaction styles, the types of content they engage with, the appropriate timescales, and subtle behaviors through which they provide information regarding their engagement. Moreover, measures to address specific problematic social network use related to appearance and modifications of weight and shape are lacking and would also be useful to further this area of research. Moving forward with a view to testing and refining the proposed model, additional measurements of social network use, and problematic social network use, will be needed (Moretta et al., 2022).

5.2. Model testing: Hypothesis driven research in targeted groups

As noted above, while a now robust body of literature has documented the cross-sectional, experimental and longitudinal evidence for the relationship between higher use and investment in visual social networks and higher body image and eating concerns (de Valle et al., 2021). However, the relations included in the proposed model, that specify associations between specific aspects of appearance-motivated uses of social network, appearance-centered online interactions, and levels of thin-ideal internalization, negative affect and body image and disordered eating have received comparatively little attention. Further work is required in this field, encompassing qualitative, experimental, and longitudinal studies, to achieve a more profound understanding of the complex relationships under scrutiny. Moreover, much of the extant work has been conducted among convenience or general samples. Examining these processes in high-risk samples, and exploring the existence of specific pathways according to level of risk will be important in future research. In addition, designing studies that are able to capture processes in real time, and the likely bidirectional nature of these relationships will be important.

Much of the previous work related to the associations between social network use and body image and eating has focused on girls and young women (de Valle et al., 2021), partly due to the underlying sociocultural models being more robustly documented in these groups. However, recent meta-analytic reviews have shown that women and men have a similar risk of developing problematic social network use (Casale et al., 2023, Cheng et al., 2021, Meng et al., 2022), and research has suggested that presentations and correlates may differ across gender (Wang et al., 2022). Given this, models such as the proposed one, may be better suited to capturing factors implicated in the relationships among social network use and body image and eating concerns in men, and additional research increasing our understand of these relationships in men is needed.

5.3. Ecological validity

The majority of the existing empirical research on the relationships between social network use and body image and eating concerns comes from correlational or experimental studies which, while useful, is limited in ecological validity. Moving forward, additional efforts should be made to translate research designs into more ecologically valid contexts that include a range of stimuli, multitasking, and other aspects that occur in real-life social network use. Further, additional research on specific and more recent online platforms is also needed. For example, TikTok has gained popularity among youth (Zhang & Liu, 2021) and it is likely more engaging than other platforms due to its short videos, and its ability to quickly adapt to the user's mood or desires. TikTok algorithms underlying the personalized suggested content may make individuals who selectively attend towards appearance/eating-related content more at risk for receiving higher proportions of appearance/eating-related suggested content, in turn contributing towards ED symptoms. With the exception of a small number of examples already discussed here (e.g. Dondzilo et al., 2024), research on this platform is lacking.

6. Conclusions

To conclude, we echo the urgent call for the development of new theories that can capture the unique influences of social media platforms, rather than relying solely on existing models that were developed for traditional media (Harriger et al., 2023, Vandenbosch et al., 2022). The current paper presents a new proposed model of the exacerbation and maintenance of body image and eating concerns on social networks that highlights the importance of generalized and specific PSNU, appearance-motivated uses of social network, appearance-centered online interaction, and putative moderators, and considers dimensions of PSNU that may contribute. Preliminary empirical support exists for the relationships proposed within this model, however overall many of these pathways require further examination. Future work to test and refine this model, including appropriately scaled longitudinal work among high-risk populations is warranted.

CRediT authorship contribution statement

Gianluca Lo Coco: Writing – review & editing, Writing – original draft, Conceptualization. Rachel F. Rodgers: Writing – review & editing, Writing – original draft, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

No data was used for the research described in the article.