Digging into digital buffets: A systematic review of eating-related social media content and its relationship with body image and eating behaviours

Recently, researchers have started investigating the influence of eating-focused social media content on viewers ’ body image and eating behaviour. The current systematic review collates the evidence for the relationship be-tween exposure to eating-related social media content and people ’ s body image, disordered eating, food consumption, and eating-related opinions. Following PRISMA guidelines, searches were conducted on five databases (i.e., Medline, PsycINFO, Web of Science, Scopus, and ProQuest), resulting in 38 eligible studies that examined the link between eating-related content and body image, eating behaviour and/or viewers ’ perceptions. Evidence showed positive relationships between exposure to certain material on social media (e.g., fast-food advertisements, clean eating, eating videos) and body image concerns, disordered eating, and consumption of, or a willingness to consume the foods seen on social media; however, studies that focused on ‘healthy food ’ content showed no such link. Viewers expressed mixed feelings towards content such as nutrition information and mukbang videos which record hosts eating a large amount of food. Future research would benefit from a consistent measure of exposure to different eating-related content. There is also a need for more experimental research to examine the impact of watching different types of eating videos on body image, disordered eating, and food


Introduction
Over recent years, using and being exposed to social media has become prevalent, raising concerns around its links to body image and disordered eating.In 2023, approximately 60% of the world's population were using social media (Kemp, 2023).Image-based social networking sites such as Instagram and Facebook and video-based sites such as YouTube are popular platforms that people view and share content on.Compared to traditional media, social media platforms enable individuals to create their own profiles and content, search for material they are interested in, interact with others, and receive algorithm-generated posts based on their social media activity (Braghieri et al., 2022;Vandenbosch et al., 2021).Previous systematic reviews and meta-analyses have established a positive correlation between social media use (e.g., time spent on, frequency of using, and viewing photos on social networking sites) and disordered eating and body image concerns (Holland & Tiggemann, 2016;Mingoia et al., 2017;Rodgers & Melioli, 2016;Saiphoo & Vahedi, 2019).In addition, one recent systematic review focused on the effect of general social media use on food-related outcomes and suggested that viewing unhealthy food-related social media content may lead to unhealthy food consumption (Ventura et al., 2021).In particular, appearance-related social media use (exposure to appearance-focused imagery) has been shown to have a stronger relationship with negative body image than general social media use (Mingoia et al., 2017;Saiphoo & Vahedi, 2019).A longitudinal study also found that appearance-focused social media use but not communication-focused use (e.g., sending messages) was associated with body image issues such as body shame (Markey & Daniels, 2022).
More recently, a small body of research has begun to look specifically at social media content around eating or food and its relationship with viewers' body image and eating behaviour.This interest stems from previous research on traditional media that has suggested that mediarelated behaviours such as watching television may be associated with increased energy intake (Gore et al., 2003;Hetherington et al., 2006) and unhealthy eating habits (Harris & Bargh, 2009).Furthermore, exposure to food-related content such as TV programs about food and eating has been linked to increased intake of sweet snacks (Bodenlos & Wormuth, 2013;De Backer & Hudders, 2016).Social media then provides even greater reach for this type of material in combination with the emphasis on the thin ideal and dieting.As such, it is likely that eating-related social media content could influence viewers' body image and eating behaviour.For example, a cross-sectional study showed that the duration of exposure to food and beverage content on social media was linked to higher protein and energy intake (Adiba et al., 2020).
Notably, there are various types of popular eating content on social media, ranging from still food images and healthy eating advice to eating videos.Increased exposure to social media content related to clean eating (a style of eating that promotes the consumption of whole and unprocessed food) has been positively related to dietary restraint (Allen et al., 2018), disordered eating, and negative body image such as body dissatisfaction and appearance-ideal internalisation (Wu et al., 2022).Certain eating video content (i.e., mukbang videos which record hosts eating a large amount of food) on social media may also be linked to unhealthy eating behaviours (e.g., restrictive or binge eating) or body image concerns (Kang et al., 2020;Kircaburun et al., 2020;Strand & Gustafsson, 2020).As such, it is possible that eating-related behaviours may be affected by watching a range of eating-related content on social media.In addition, research on certain eating-related social media content has suggested that exposure to content such as clean eating advice (Allen et al., 2018) and food blogs (Lynch, 2010) may promote a positive attitude towards the type of content and restrained eating.As the link between eating content and body image or eating behaviours is an emerging area of research, gaining insight into people's opinions regarding such content may provide a more comprehensive understanding of how it could influence viewers.
To date, there is no systematic review that focuses specifically on eating-related social media content and its links to viewers' body image and/or eating behaviour.Given the links between social media and body image concerns, this is important to consider.In addition, eating videos have gained in popularity, reaching millions of people on social networking sites worldwide.A search of "eating" on YouTube through Google (as of September 2023) showed over 400 million results.Considering the prevalence and potential negative impact of this type of content, the present review aimed to collate the findings of research on the influence of eating-related content on social media, including video content (e.g., YouTube eating videos).Specifically, the review aimed to draw together the current body of evidence on how exposure to (i.e., viewing) eating-related social media content is linked to body image and eating behaviour (including food consumption and disordered eating), and viewers' opinions regarding such content.

Method
The review process was conducted in accordance with the PRISMA guidelines (Page et al., 2021).The review protocol was preregistered with the Open Science Framework (OSF; registration DOI: 10.17605/OSF.IO/N4JRM).

Search strategy
The search strategy was developed with the assistance of an academic librarian, which incorporated key terms and thesaurus terms related to social media (e.g., social network, Facebook, YouTube, Tik-Tok, online), eating content (e.g., mukbang, eating competition, eating video), body image (e.g., body dissatisfaction, drive for thinness) and eating behaviours (e.g., disordered eating, food intake).See Appendix A for the full search strategies.The databases searched were Medline, PsycINFO, Web of Science, Scopus, and ProQuest.The initial search was conducted on the 30th of June 2021, and an updated search on the 9th of February 2023.Forward and backward searching were also conducted to ensure all relevant studies were identified.

Inclusion and exclusion criteria
To be included, articles needed to be peer-reviewed and examine the link between eating-related content on social media and body image and/or eating behaviour or viewers' opinions of such content.Accordingly, studies were selected if any type of eating or food-related content on social media was included, and there was a measure of body image (e. g., body dissatisfaction, internalisation of appearance ideals, drive for thinness, drive for muscularity), eating behaviour (e.g., diet, food consumption, disordered eating, food restriction, feelings about food/diet), or opinions regarding the content.The search results were limited to peer-reviewed, primary research articles published in English since 2006 (the year Facebook was released to the public).Secondary research, book chapters and magazine articles were excluded.Studies that did not investigate the link between social media food or eating content and eating behaviour or body image were also excluded.
Two independent screeners were involved in both title and abstract screening and full-text screening.Discrepancies were resolved through discussion between the screeners until a consensus was reached.

Data extraction
The first author extracted the following data from each of the included studies: author(s), year of publication, country, study design, sample size, sample characteristics, type of eating-related social media content, measurement of exposure to eating-related content, measurement of body image and/or eating behaviour, and findings in relation to the link between exposure to eating-related social media content and body image and eating behaviour.

Quality assessment
Following Goodyear et al. (2021), three quality assessment tools were used to evaluate the risk of bias to determine the certainty of evidence for the included studies, depending on study design.The quality of cross-sectional studies was assessed using the Joanna Briggs Institute (JBI) critical appraisal checklist for analytical cross-sectional studies (Moola et al., 2020).The checklist comprises of eight domains regarding study design and data analysis (e.g., sample selection, valid and reliable measurements, confounding factors; Moola et al., 2020).The Cochrane Collaboration's tool for assessing risk of bias (Higgins et al., 2011) was used to evaluate the risk of bias of experimental studies.This tool assesses bias across seven domains: sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessors, incomplete outcome data, selective outcome reporting, and other sources of bias (Higgins et al., 2011).Finally, the Mixed Methods Appraisal Tool (MMAT; Hong et al., 2018) was used to assess the quality of mixed methods and qualitative studies.This tool assesses bias across five domains for mixed methods studies (including adequate rationale for the study design, integration of different components, interpretation of the findings, divergencies and inconsistencies, and the quality of the qualitative and quantitative components), and a different set of five domains for qualitative studies (including the appropriateness of the study design, data collection, data analysis, interpretation of the results, and coherence between data sources, collection, analysis and interpretation; Hong et al., 2018).
For each assessment tool, each domain was assigned a low, unclear, or high risk of bias.According to Higgins et al. (2011), a study is considered low risk of bias when all key domains are assigned a low risk, unclear risk when all key domains are assessed as low or unclear risk with at least one domain being unclear risk, and high risk when one or more key domains are high risk.Fifty percent of the quality assessment was conducted by two independent assessors, with disagreements discussed until a consensus was reached.

Data synthesis
Differences in study characteristics, including study design, measures of exposure to eating-related content, and outcome measures resulted in heterogeneity in the findings.Due to the heterogenous nature of the included studies, meta-analysis was not deemed appropriate.Instead, narrative synthesis was used to bring together diverse strands of evidence from multiple studies (Petticrew et al., 2013;Popay et al., 2006).The Synthesis Without Meta-analysis (SWiM; Campbell et al., 2020) guideline was followed for synthesising quantitative findings where appropriate.Studies were grouped based on different outcome measures, including the links between eating-related social media content and body image, disordered eating, food consumption, and viewers' opinions.Table 1 presents the study characteristics data, including study design and risk of bias.

Study selection
The search strategy generated a total of 5148 articles (see Fig. 1 for the PRISMA diagram of the identification and screening of studies).First, duplicates were removed by automation tools including Endnote and Covidence (n = 2855), leaving a total of 2293 articles for title and abstract screening.Overall, 2055 articles were excluded at title and abstract screening as they were deemed not relevant, leaving 238 fulltext articles to be assessed for eligibility.Articles without a measure of exposure to eating-related content on social media (n = 111) or a measure of body image or eating behaviour (n = 59) were excluded.
Articles in languages other than English, secondary research, book chapters, and magazine articles were also excluded (n = 29).In addition, one dissertation was potentially eligible, but it reported on the same data as a published article.To avoid duplication, only the published article was included.This selection process led to a total of 38 eligible studies.

Eating-related content and body image
Eleven studies directly assessed the link between exposure to eatingrelated content on social media and participants' body image (Adiba et al., 2020;Caner et al., 2022;Friedman et al., 2022;Kinkel-Ram et al., 2022;Lambert et al., 2019;Sowles et al., 2018;Stein & Yeo, 2021;Strand & Gustafsson, 2020;Tan et al., 2016;Teng et al., 2020;Wu et al., 2022).Tan et al. (2016) identified that the top and most important reason for eating disorder patients visiting a Facebook group related to eating disorders was to lose weight.In addition, cross-sectional studies indicated that following social media Influencers who share diet and nutrition content (Caner et al., 2022) and increased frequency of exposure to clean eating imagery on Instagram (Wu et al., 2022) was positively associated with body image issues such as social appearance anxiety and body dissatisfaction.
Qualitative studies also showed that people believed nutrition information (Lambert et al., 2019) and fast-food advertisements (Friedman et al., 2022) on social media portrayed unrealistic body ideals (e.g., thin ideals) that could lead to negative body image such as body dissatisfaction.In addition, content analyses of qualitative comments showed that individuals who engaged with eating disorder related content on Reddit expressed admiration for body parts that demonstrated thinness (Sowles et al., 2018), and viewers of healthy eating videos associated healthy eating with weight concerns in YouTube video comments (Teng et al., 2020).Viewers' comments on mukbang YouTube videos also showed envy towards hosts' not gaining weight despite the amount of food eaten, and body shaming towards the hosts' body shape (Strand & Gustafsson, 2020).
On the other hand, one study which focused on mukbang watching among German and South Korean samples found no significant relationships between mukbang watching and body esteem and current dieting (Stein & Yeo, 2021).Similarly, an experimental study indicated that there was no effect of exposure to videos of Instagram pages of low-calorie density food images compared to travel images on state body image (Kinkel-Ram et al., 2022).Another study found that women who had a negative body image did not tend to select foods based on the food and beverage content they saw on social media (Adiba et al., 2020).However, the type of food was not specified in this study.Overall, the majority of the studies included provided support for the link between certain eating-related content and negative body image, while a few showed no such association.

Eating-related content and disordered eating
Ten studies investigated the disordered eating behaviours of people exposed to eating-related content (Al-Bisher & Al-Otaibi, 2022; Allen et al., 2018;Bachner-Melman et al., 2018;Chu & Xiao, 2022;Kim et al., 2021;Kinkel-Ram et al., 2022;Kircaburun, Yurdagul, et al., 2020;Mincey & Hunnicutt Hollenbaugh, 2022;Sowles et al., 2018;Wu et al., 2022).Of these, four cross-sectional studies found that adhering to clean eating advice (Allen et al., 2018), viewing clean eating imagery on Instagram (Wu et al., 2022), engaging with eating disorder content and eating/weight/body image related content (Bachner-Melman et al., 2018), and potentially addictive mukbang watching (Kircaburun, Yurdagul, et al., 2020) were correlated with higher levels of restrained eating or disordered eating, with one study showing that this relationship was mediated by thin-and athletic-ideal internalisation (Wu et al., 2022).In addition, participants who watched eating broadcasts online (i.e., on social media platforms such as YouTube) and on TV scored higher on restrained eating behaviours than those who had never seen an eating broadcast, and higher on emotional eating and external eating behaviours than those who only watched TV broadcasts and those who never watched these (Kim et al., 2021).Furthermore, Al-Bisher and Al-Otaibi (2022) reported that women with high eating concerns (e.g., disordered eating) were less likely to resist Influencers' social media food advertisements and were more attracted to Influencers' dieting experiences than those with low eating concerns.They also found that specific types of exposure such as following dieticians for nutritional information and attraction to Influencers' food advertisements were risk factors for eating concerns.An experimental study further showed that viewing Instagram feeds of low-calorie density food (e.g., fruits, vegetables) led to greater increases in intentions to engage in disordered eating than viewing travel Instagram feeds among female university students in the Midwestern United States (Kinkel-Ram et al., 2022).
In addition, a content analysis of eating disorder related comments on Reddit provided support for pro-eating disorder behaviour and information or advice on how to continue an eating disorder (Sowles et al., 2018).Another content analysis showed that some members from a binge eating related group on Baidu Tieba (a Chinese social media discussion forum that facilitates two-way communication), reported that viewing food diaries in eating disorder related discussions triggered binge eating behaviour (Chu & Xiao, 2022).Likewise, participants who engaged with eating disorder related online media reported that certain content (e.g., thin bodies, ways to hide an eating disorder, calorie intake) on pro-eating disorder sites were considered a trigger to engage in disordered eating behaviours (Mincey & Hunnicutt Hollenbaugh, 2022).Taken together, these studies provide initial evidence to support the notion that various types of eating-related social media content could be positively related to disordered eating attitudes and behaviours, and are therefore negative for overall wellbeing.

Eating-related content and food consumption
Twenty studies examined the links between eating-related social media content and food consumption or willingness to eat certain foods, including participants' preferred food groups after exposure to social media and the influence of eating-related content on food intake (Adiba et al., 2020;Aljefree & Alhothali, 2022;Allen et al., 2018;An & Ha, 2022;Bayindir-Gümüs ¸ & Keser, 2019;Coates et al., 2019;Friedman et al., 2022;Gascoyne et al., 2021;Hawkins et al., 2021;Jin, 2018;Kilb et al., 2023;Nelson & Fleming, 2019;Pollack et al., 2021;Pollack et al., 2022;Qutteina et al., 2022;Ragelien ė & Grønhøj, 2021;Strand & Gustafsson, 2020;Sumaedi and Sumardjo, 2021;Tazeoglu & Kuyulu Bozdogan, 2022;Wang et al., 2021).One study reported that cake and other desserts, meat, and fast-food were identified as the top food groups university students wanted to consume after seeing content about these foods on social media, with female students reporting cake and other desserts as their first preference, and male students preferring meat and meat products (Bayindir-Gümüs ¸ & Keser, 2019).In another study, a majority of university students reported fast-food as the top food choice after watching food videos on social media (Tazeoglu & Kuyulu Bozdogan, 2022).Similarly, among social media Twitch users who recalled craving a product after seeing it advertised on this livestreaming social media platform, most participants reported craving fast-food or food delivery brands (Pollack et al., 2021).Another study showed that Twitch users with higher levels of responsiveness to external food and food cues were more likely to report craving a food product after seeing food advertisements on Twitch (Pollack et al., 2022).
Of the 20 studies that focused on eating-related social media content and food consumption, six cross-sectional studies investigated the associations between exposure to eating-related content and consumption of, or willingness to eat certain foods (Adiba et al., 2020;Aljefree & Alhothali, 2022;Allen et al., 2018;Gascoyne et al., 2021;Qutteina et al., 2022;Sumaedi and Sumardjo, 2021).For example, women who more frequently adhered to dietary advice from clean eating sites were more likely to have an adequate intake of fruit as well as meat and alternatives that met the Australian Guide to Healthy Eating (AGHE) guidelines than those who never or rarely adhered to clean eating advice from social media (Allen et al., 2018).In addition, compared to adolescents who did not engage with food marketing posts on social media, those who liked or shared this type of posts were more likely to have a higher intake of unhealthy food and drinks (Gascoyne et al., 2021).Similarly, another study indicated that adolescents who were more frequently exposed to non-core food (e.g., fried food, chips) posts on social media were more likely to prefer and consume non-core foods (i.e., foods that are high energy-dense and low-nutritious) such as sweet and fried food (Qutteina et al., 2022).Those who were less frequently exposed to non-core food messages were more likely to demonstrate higher food literacy, which in turn was associated with higher intake of core food (high-nutritious foods from the main food groups recommended by dietary guidelines, e. g., vegetables, fruits) and lower intake of non-core food (e.g., soft drinks; Qutteina et al., 2022).
In addition, Adiba et al. (2020) found that adolescent girls exposed to food and beverage content on social media for longer periods were more likely to select food based on the content they were exposed to, and in turn tended to have higher protein and energy (but not fat and carbohydrate) intake.However, neither the frequency of exposure to food and beverage content nor the number of accounts in relation to this type of content accessed was related to food selection based on social media content (Adiba et al., 2020).Likewise, another study showed that the frequency of exposure to healthy food-related information on social media was not related to the consumption of tempeh, a traditional functional food, among an Indonesian sample (Sumaedi and Sumardjo, 2021).
One experimental study showed that exposure to Facebook postings related to fruit and vegetable intake increased participants' fruit and vegetable intake, but found no change in those exposed to book and movie postings (Kilb et al., 2023).A further three experimental studies showed that exposure to still images of unhealthy snacks on social media (Coates et al., 2019), socially endorsed images of low energy-dense food (Hawkins et al., 2021), and profile pages with various food posts (e.g., burger, pizza, ice cream) and images of the thin ideal (Jin, 2018) led to an increased intake or willingness to eat the type of food seen in the images.
Another experiment did not find any significant differences in the effect of exposure to food images (including healthy, unhealthy, and neutral food) that revealed different degrees of social media information (i.e., no Instagram information, Instagram icons visible, number of likes visible, or caption visible) on the likelihood of consuming the foods in the images (Nelson & Fleming, 2019).However, this experiment did indicate a positive relationship between perceived healthiness of food and the likelihood of consuming the food depicted in healthy (e.g., vegetables, fish) but not unhealthy or neutral food images among women (Nelson & Fleming, 2019).
Two experimental studies investigated the effects of watching mukbang on food consumption.An and Ha (2022) examined the effects of watching YouTube mukbang videos featuring a Korean Influencer overeating unhealthy food (e.g., spicy rice cakes) with or without the caption 'Paid Ad' as an ad disclosure and found no direct effect of this ad disclosure on eating intent of the food in the video.However, higher eating intent was evident among participants who felt less manipulated with the inclusion of the ad disclosure and those who had favourable attitudes toward the Influencer in the video.Similarly, Wang et al. (2021) found that in comparison to participants in other scenarios (i.e., eating in the company of two co-diners (in-person commensality), and eating when videoconferencing with two co-diners (cloud-based commensality)), those who were shown a photo of a young woman eating while watching mukbang (mukbang-based commensality) and instructed to imagine being in this scenario reported the highest increase in their perceived food intake relative to eating alone.In addition, they found that mukbang-based commensality elicited a lower likelihood of choosing healthy food than solitary eating and in-person commensality scenarios.
One netnographic analysis showed that viewers reported that watching mukbang increased their consumption of food such as snacks in the comments of mukbang videos on YouTube and in the Reddit comments of the mukbang discussion (Strand & Gustafsson, 2020).One of the only two mixed methods studies found that the frequency of food products seen advertised on social media was related to a greater willingness to consume the products, which then predicted lower vegetable intake and more snack intake among children (Ragelien ė & Grønhøj, 2021).Furthermore, one digital ethnographic study using web-based conversations showed that participants reported fast-food advertisements being predominant on social media newsfeeds which influenced them to make unhealthy food choices (Friedman et al., 2022).
In summary, the existing evidence overall supports the link between exposure to certain types of eating-related material and increased consumption or willingness to consume the foods seen in the content or certain food groups, particularly unhealthy foods.On the other hand, a smaller number of studies showed that there is no association between some types of eating-related content and food consumption.

Viewers' opinions
Eleven studies examined individuals' perceptions or opinions of eating-related content on social media (Allen et al., 2018;Bissonnette-Maheux et al., 2015;Friedman et al., 2022;Lambert et al., 2019;Mete et al., 2019;Ragelien ė & Grønhøj, 2021;Strand & Gustafsson, 2020;Tazeoglu & Kuyulu Bozdogan, 2022;Teng et al., 2020;Wang et al., 2022;Zemlyanskaya et al., 2022).Specifically, Allen et al.'s (2018) cross-sectional study demonstrated that seldom-and non-adherers of clean eating advice on social media were less likely to have a positive opinion and more likely to have a negative opinion about clean eating in comparison to adherers.In addition, a mixed methods study showed that some participants thought that orthorexia-related content on Instagram encouraged a positive relationship with food, whereas others reported that it promoted clean eating and led to orthorexia (Zemlyanskaya et al., 2022).These two studies demonstrate the mixed perceptions towards clean eating on social media.
The seven qualitative studies also showed mixed (both positive and negative) feelings towards content such as nutrition information or healthy eating on social media.For example, Mete et al. (2019) found that people believed that fitness posts related to healthy meals on Instagram or Facebook, and blogs containing healthy eating information, inspired and connected people with healthy eating.Female participants in another study thought that healthy eating blogs written by dietitians gave recipe ideas, improved diet, helped to learn about nutrition, but also induced guilt from not complying with the dietary recommendations provided (Bissonnette-Maheux et al., 2015).
In relation to eating videos, adolescents in Wang et al.'s (2022) study reported that food content on TikTok inspired them to consider healthy eating, try specific foods in the videos, and learn dietary regimens.In addition, Teng et al.'s (2020) examination of comments on YouTube healthy eating videos found that Malaysians expressed positive feelings towards the content.They also thought that food taste, variety, and availability were the main reasons why Malaysians cannot act on eating healthily (Teng et al., 2020).Furthermore, Strand and Gustafsson's (2020) netnographic analysis of viewers' comments on mukbang videos on YouTube and Reddit found that some felt that this content increased problematic eating behaviours while others indicated that watching mukbang helped prevent binge eating.Finally, children in Ragelien ė and Grønhøj's (2021) mixed methods study recalled seeing photos or videos related to drinks or foods on social media, and reported that they were willing to consume the foods they had seen (e.g., salad, pizza).Overall, the existing evidence demonstrates viewers' mixed perceptions towards certain eating-related content on social media, particularly healthy eating related content.

Study quality
Based on the JBI critical appraisal checklist for analytical crosssectional studies (Moola et al., 2020), three cross-sectional studies (Gascoyne et al., 2021;Kim et al., 2021;Qutteina et al., 2022) were assessed as low risk on all eight domains.However, most cross-sectional studies (Al-Bisher & Al-Otaibi, 2022;Aljefree & Alhothali, 2022;Allen et al., 2018;Bachner-Melman et al., 2018;Caner et al., 2022;Kircaburun, Yurdagul, et al., 2020;Pollack et al., 2021;Stein & Yeo, 2021;Sumaedi and Sumardjo, 2021;Tan et al., 2016;Wu et al., 2022) included in the review were assessed as unclear or high risk for the identification of, and the strategies to deal with confounding factors.While all eighteen cross-sectional studies included a description of the exposure measures, four (Adiba et al., 2020;Bayindir-Gümüs ¸ & Keser, 2019;Pollack et al., 2021;Tazeoglu & Kuyulu Bozdogan, 2022) did not clearly provide the validity or reliability of the exposure measurement.Three studies (Pollack et al., 2021;Pollack et al., 2022;Tazeoglu & Kuyulu Bozdogan, 2022) did not describe the validity or reliability of the outcome measurement clearly.Other domains such as sample selection and statistical analysis were assessed as low risk for all eighteen studies.According to the Cochrane Collaboration's tool for assessing risk of bias (Higgins et al., 2011), two of the experimental studies (Hawkins et al., 2021;Kinkel-Ram et al., 2022) were deemed to be low risk of bias across all seven domains.Four of the remaining six experimental studies (Coates et al., 2019;Jin, 2018;Kilb et al., 2023;Nelson & Fleming, 2019) did not provide clear descriptions of the measures used to blind participants and personnel.In addition, four experimental studies (An & Ha, 2022;Jin, 2018;Nelson & Fleming, 2019;Wang et al., 2021) were assessed as unclear risk for incomplete data as they did not provide clear information on attrition and exclusions.Three experimental studies (An & Ha, 2022;Jin, 2018;Kilb et al., 2023) did not provide a detailed description of the random sequence generation, and two (Coates et al., 2019;Kilb et al., 2023) did not clearly describe the measures used to blind outcome assessment.According to the MMAT (Hong et al., 2018), the ten qualitative studies and the two mixed methods study included in this review were assessed as low risk on all five domains.

Discussion
With the growing popularity of eating-related content on social media, an emerging body of research is investigating the relationship between exposure to this type of content and body image and eating behaviour.The present review aimed to evaluate the evidence for this relationship.Specifically, it investigated the links between eatingrelated social media content and viewers' body image, disordered eating, food consumption, and opinions regarding this type of content.
Overall, evidence demonstrates positive correlations between exposure to certain eating-related content on social media and consumption of, or a willingness to consume foods seen on social media (e.g., fastfood, desserts), disordered eating behaviours (e.g., restrained eating, binge eating), and body image concerns (e.g., concerns with body shape and weight, internalisation of appearance ideals).In addition, studies demonstrated both positive and negative opinions about food advertisements and healthy eating or nutrition-related content, with some revealing that thin or muscular ideals were predominant in this type of content.Furthermore, comments from viewers of eating videos on YouTube included content about disordered eating (mukbang videos) and body image concerns (both mukbang and healthy eating videos).Collectively, this raises concerns about eating-related social media content.

Eating-related content and body image
To date, there is a growing body of evidence for a link between certain types of eating-related content and body image concerns.In particular, qualitative studies, including analyses of comments or discussion in relation to different eating-related content (e.g., eating disorder forums, healthy eating videos, and mukbang videos), revealed body image issues such as weight concerns, a tendency towards idealising thinness, and potential body dissatisfaction (Friedman et al., 2022;Lambert et al., 2019;Sowles et al., 2018;Strand & Gustafsson, 2020;Teng et al., 2020).In addition, two cross-sectional studies showed a positive relationship between engaging with health-related content (i.e., nutrition content, clean eating images) and body image concerns (e.g., appearance-ideal internalisation; Caner et al., 2022;Wu et al., 2022).However, at present, the only experimental study that included a body image related outcome found no significant effect of exposure to low-calorie density food images from social media on state body image (Kinkel-Ram et al., 2022).Thus, there is a need for more cross-sectional and experimental studies to verify the relationship between viewing different types of content and its effects on viewers' body image.In particular, experimental research is needed to determine causality.In addition, future research could use standardised measures to examine the body image experiences (e.g., body dissatisfaction, appearance-ideal internalisation) of people who are exposed to a broader range of eating-related social media content.

Eating-related content and disordered eating
Previous research has established that social media use and the frequency of engaging with certain types of social media content (e.g., appearance-focused content) are positively related to disordered eating (Fardouly et al., 2017;Holland & Tiggemann, 2016).More recently, a developing body of research has examined the link between eating-related material and disordered eating and has identified types of content that could be considered deleterious.Among these, four studies showed that eating disorder related social media content was linked to disordered eating attitudes and behaviours (Bachner-Melman et al., 2018;Chu & Xiao, 2022;Mincey & Hunnicutt Hollenbaugh, 2022;Sowles et al., 2018).Two studies found that clean eating, which is largely regarded as a source of dietary information and a health and wellness focused eating style, was positively related to dietary restraint (Allen et al., 2018) and disordered eating symptomatology (Wu et al., 2022).Another study reported that nutritional information and food advertisements were positively associated with eating concerns (Al-Bisher & Al-Otaibi, 2022).Only one experimental study indicated that exposure to low-calorie density food images led to increased disordered eating intentions (Kinkel-Ram et al., 2022).In addition, previous research has shown that many healthy living blogs contain unhealthy nutrition messages (Boepple & Thompson, 2014) and many health Influencers who share health and nutrition advice on social media are poorly qualified (Lynn et al., 2020).This calls for further research into the influence of Influencers and the nutrition information or healthy eating related content they provide, such as clean eating, on disordered eating symptomatology to better understand whether this type of content may be harmful for viewers' eating behaviours.
In relation to eating videos, three studies supported that watching mukbang videos could be related to disordered eating (Kim et al., 2021;Kircaburun et al., 2020;Strand & Gustafsson, 2020).This is consistent with findings from an earlier scoping review on the psychological consequences of mukbang watching which showed that it can promote disordered eating behaviours (Kircaburun et al., 2020).Unhealthy eating behaviours such as overeating shown in popular videos may influence viewers' perceptions about these behaviours, and could have unhealthy consequences, such as increasing body image concerns or promoting binge eating.However, considering the limited number of studies, more research is needed to confirm this relationship.There is also a lack of experimental research that examines the influence of watching eating videos on disordered eating attitudes and behaviours, which calls for future studies in this area to determine potential causal effects.

Eating-related content and food consumption
Several cross-sectional and experimental studies and the single mixed methods study included in this review showed a link between exposure to different types of eating-related content on social media and increased consumption or willingness to consume the food or food groups seen in the social media content.The examined content ranged from food advertisements, eating advice, food-intake posts, and images of different types of food under certain conditions (e.g., images of YouTube Influencers with healthy/unhealthy snacks) to eating videos.Three studies identified that cakes, meat, and fast-food were the most desired food groups people wanted to consume after seeing these foods on social media (Bayindir-Gümüs ¸ & Keser, 2019; Pollack et al., 2021;Tazeoglu & Kuyulu Bozdogan, 2022).This is similar to research on traditional media content which showed that exposure to food advertisements on TV could increase snack and fast-food intake (Emond et al., 2019;Harris et al., 2009).Again, as only a small number of studies have been conducted in this space, more research is needed to verify this finding in relation to food content on social media.
Across the cross-sectional studies and the single mixed methods study that examined the relationship between eating-related social media content and food consumption (Adiba et al., 2020;Aljefree & Alhothali, 2022;Allen et al., 2018;Gascoyne et al., 2021;Qutteina et al., 2022;Ragelien ė & Grønhøj, 2021;Sumaedi and Sumardjo, 2021), a majority showed a positive relationship between exposure to such content and increased intake of certain food groups (Adiba et al., 2020;Aljefree & Alhothali, 2022;Allen et al., 2018;Gascoyne et al., 2021;Qutteina et al., 2022;Ragelien ė & Grønhøj, 2021).However, it remains unclear whether the increased intake is related to the frequency, duration, or other forms of exposure to eating-related content, possibly due to differences in the measures of content exposure and the specific types of content that were examined.The direction of the effect also remains unclear, as individuals who already have a certain style of eating may seek out similar content on social media that supports their current eating behaviour.This suggests that a range of individual difference factors (e.g., dietary restraint, food preference) may also be important to consider in this field to determine moderators of how people respond to eating-related content on social media.
Overall, most experimental studies that focused on the influence of eating-related content on food consumption found that still images of certain foods under specific conditions (e.g., images of YouTube Influencers with unhealthy snacks, socially endorsed images of low energydense food) and Facebook posts related to fruit and vegetable intake increased the consumption of the foods or food types seen (Coates et al., 2019;Hawkins et al., 2021;Jin, 2018;Kilb et al., 2023), demonstrating a causal effect.Similar to the cross-sectional studies, the experimental studies focused on different types of food images (e.g., healthy or unhealthy snacks, low or high energy-dense food) or posts and various conditions (e.g., the number of likes received, the body size next to the food images).This variety makes it difficult to conclude how and why different eating-related content would influence subsequent food consumption.As such, although the evidence overall supports the claim that exposure to eating-related content is positively related to increased consumption of certain foods, future research could benefit from greater consistency in the measures used to capture exposure to eating-related social media content and the conditions in which food images are presented.
In relation to eating videos, although the type of food was not specified, two experimental studies provided initial evidence that watching mukbang may increase food intake under certain conditions (e.g., when ad disclosure was present, and viewers showed high ad recognition; An & Ha, 2022;Wang et al., 2021).In addition, women exposed to food images along with pictures of a thin body exhibited an increased intention to consume the foods shown (Jin, 2018).It is therefore possible that viewing eating videos that also include individuals who display the ideal body may lead to a similar result.This could be explained by altered perceptions around eating and being fit/thin after seeing fit/thin-looking individuals consuming an excessive amount of food (Kircaburun, Yurdagul, et al., 2020).However, it remains unknown whether mukbang videos actually portray body ideals, whether viewing them leads to disordered eating symptomatology, and how they influence food consumption.Although mukbang videos have become increasingly popular, the implications of mukbang video watching on social media is a relatively novel area of research.Algorithms used by social media platforms may also lead to repeated exposure of the same type of content, potentially influencing viewers to a greater extent.To provide a more comprehensive understanding of how engaging with this type of content influences viewers, future research may benefit from further investigating the actual body types portrayed in the videos and the effects of watching these videos on viewers' body image, disordered eating, and food consumption.

Viewers' opinions
Studies of viewers' opinions towards eating-related content demonstrated mixed feelings around different types of content, particularly healthy eating related material or nutrition information including clean eating advice (Allen et al., 2018;Bissonnette-Maheux et al., 2015;Lambert et al., 2019;Mete et al., 2019;Teng et al., 2020;Zemlyanskaya et al., 2022).For example, qualitative studies identified both positive (e. g., improving diet) and negative thoughts (e.g., inducing guilt, portraying body ideals) regarding healthy eating or nutrition-related content on social media (Bissonnette-Maheux et al., 2015;Lambert et al., 2019).Thus, these findings show that eating-related content with positive intentions may also elicit negative feelings, making it important for health professionals to be careful about the content they share on social media to limit potential negative effects.They further highlight the importance of education around eating-related social media content and improving media literacy, which may help reduce the potential negative influence of exposure to such eating material.
The findings of this review should be considered in the context of its limitations.First, the review is subject to publication bias because it only included studies that have been published.It is possible that there are studies that are unpublished due to non-significant results.Second, the review only focused on eating behaviours and body image, while other potential psychological impacts of eating-related social media content were not included (e.g., addiction, mood).Third, the conclusions reached in the review should be taken with caution given the use of narrative synthesis due to heterogeneity in the data of the included studies, precluding quantitative statistics of the relationships between variables.Further, the variety of content examined in the studies makes it difficult to clearly distinguish between specific types of content.
Nevertheless, the findings of the present review provide several practical implications.In particular, more credible information in Y. Wu et al. relation to healthy eating provided by qualified health professionals is needed to provide social media users with more reliable sources of knowledge.However, health professionals should be mindful of the potential negative effects of the content they post on social media.Education around the negative influence of exposure to certain eating material on social media might effectively raise individuals' awareness and encourage them to limit their exposure to such content.It is also recommended that social media literacy interventions, which have been shown to reduce body image concerns (McLean et al., 2016), should incorporate the potential negative effects of eating-related material on viewers' body image and eating behaviour.
In conclusion, the present systematic review provides support for the relationship between exposure to eating-related social media content and viewers' eating behaviour and body image.Overall, evidence supports a link between viewing certain eating-related material (e.g., clean eating, food advertisements, eating videos) and increased body image concerns, disordered eating, and food consumption, with a few studies showing no such link.On the other hand, the small number of studies that focused on content such as healthy food shown on social media found no significant associations with body image or eating behaviours.Studies also showed that viewers have both positive and negative feelings towards different types of eating-related content.Consistency in the measurement of specific social media content exposure would be beneficial for future studies to provide a better understanding of eatingrelated content.More generally, it is important that future research continues to investigate how different types of social media eating content, especially eating videos whose numbers and popularity are on the rise, influence social media users to inform the development of interventions to reduce the negative consequences of viewing some potentially harmful content.

CRediT authorship contribution statement
All authors contributed to the conceptualization and design of the study and writing of the manuscript.Yu Wu was responsible for conducting the systematic review under supervision of Ivanka Prichard and Eva Kemps.Yu Wu wrote the original draft of the manuscript, with the other authors reviewing and editing subsequent drafts of the manuscript.

Declaration of Competing Interest
None.

Identification Screening Included Fig. 1. PRISMA
flow diagram for the selection process.Y.Wu et al.

Table 1
Summary of articles.
eating.Exposure to both foodrelated content and health content on social media made Low (continued on next page) Y. Wu et al.Low (continued on next page) Y. Wu et al.Unclear (continued on next page)Y.Wu et al.