The Tripartite Influence Theory posits that sociocultural pressures from media, family, and peers impact eating behaviors (Thompson et al., 1999). According to the Tripartite Influence Theory, beauty short video engagement can predict dietary restraint in young women. Moreover, previous studies also provided a basis for examining this relationship. A study has shown that exposure to magazines affects women’s eating behaviors (Tiggemann, 2003). Women who frequently consume idealized body images (idealized thinness) presented in magazines are prompted to engage in dietary control behaviors to attain these idealized body images (Slevec & Tiggemann, 2011).

In addition, some studies explored the relationship between social media, social networking, and women’s eating behaviors (Qi & Cui, 2018; Yumen et al., 2023; Thompson et al., 2023). The studies on Facebook use indicate that discussions about obesity on the platform increase women’s concern about their body image, thereby increasing the risk of eating disorders (Mabe et al., 2014; Walker et al., 2015). The research on Instagram and Pinterest use also found similar results: women who use Instagram and Pinterest frequently tend to exhibit binge eating, dieting behaviors, and eating disorders (Thompson et al., 2023). Dietary control behaviors also emerge after women watch videos featuring idealized body images (Trottier et al., 2007; Mask & Blanchard, 2011). For instance, when watching film clips featuring slim female models, restrictive eaters report greater body dissatisfaction and consume fewer snacks compared to non-restrictive eaters.

Previous studies identified that women with higher exposure to television and magazines were more likely to internalize the idealized body presented in media, making them more susceptible to eating disorders (Tiggemann, 2003; Slevec & Tiggemann, 2011). However, there is limited research on the mediating of beauty ideals internalization in the relationship between beauty short video engagement and women’s eating behaviors. Some studies have focused on the relationship between social media use and women’s beauty ideals internalization and found an association between these two (Jung et al., 2022; Delgado-Rodríguez et al., 2022; Rafati et al., 2023). For instance, Jung et al. (2022) found that the higher the intensity of social media use among women, the more likely they are to internalize the ideal thin body presented on social media. Particularly, the more women engage in appearance-related photo activities on social media, the more they internalize the ideals of body shape and appearance (Lee & Lee, 2021; Duan et al., 2022). Moreover, Nerini et al. (2024) show that exposure to retouched versus unaltered Instagram images, especially when combined with disclaimers, influences thin-ideal internalization and acceptance of cosmetic surgery.

In addition, numerous studies confirmed that the internalization of ideal thinness is the risk factor for women’s eating pathology (Stice & Whitenton, 2002; Zainal et al., 2020; Wilson et al., 2022; Malik et al., 2024). Women with severe internalization of the ideal of thinness often resort to excessive dieting and strict control of high-calorie food intake in an attempt to achieve this ideal, which can trigger severe eating disorders (Morton et al., 2020; Wilson et al., 2022). An eight-year prospective study tested the correlates of eating disorder symptoms in women and found that among 51 women exhibiting symptoms of eating disorders, 47% had experienced pressure to be thin or had internalized the ideal of thinness before the onset of their eating disorder symptoms (Stice & Van Ryzin, 2019).

According to the Objectification Theory, women influenced by a sociocultural environment of sexual objectification experience various forms of sexual objectification, which can indirectly affect eating disorders through self-objectification (Fredrickson & Roberts, 1997; Sun, 2016). For instance, the frequent consumption of beauty magazines and female-oriented television dramas can trigger women to engage in more frequent appearance monitoring (Skowronski et al., 2022; Hu & Gu, 2023). Further study has found that compared to traditional media (such as television), engagement with video games or online media results in stronger self-objectification (Karsay et al., 2018). In social media, similar results also be found that higher exposure to social media correlated with higher self-reported self-objectification scores (Fioravanti et al., 2022; Gurtala & Fardouly, 2023).

Fioravanti et al. (2022) found that social media features images of women with idealized appearances, and frequently viewing these idealized images can increase negative emotions and appearance monitoring (Fioravanti et al., 2022). Similarly, research on women’s TikTok use has found that watching short videos featuring ideal appearances on TikTok increases their dissatisfaction with their appearance, negative emotions, and self-objectification (Gurtala & Fardouly, 2023). In addition, Nerini et al. (2024) also demonstrate that public self-awareness and sociocultural influences (including media and significant others) are key predictors of body dissatisfaction and cosmetic surgery acceptance.

Self-objectification is a risk factor for dietary restraint (Niu et al., 2020; Zhao et al., 2024). A longitudinal study of college students indicated that under the influence of online media and mass media, college students are very concerned about how others perceive their appearance, and they engage in strict monitoring of their appearance, reporting more instances of restrictive eating (Zhao et al., 2024). No statistically significant differences were detected between the genders in the study. However, a meta-analysis found that compared to men, the association between self-objectification and dietary restraint is stronger for women (Schaefer & Thompson, 2018). In other words, women are more objectified in the media, report more experiences of objectification and self-objectification, and are more frequently evaluated based solely on their appearance (Fredrickson & Roberts, 1997; Klonoff et al., 2000; Schaefer & Thompson, 2018; Zhao et al., 2024). When exposed to objectifying information, trigger their self-objectification, exhibit fear of negative evaluations of their appearance, and expect to change their appearance through dietary restraint (Guo et al., 2021; Fan & Cui, 2024).

A positive correlation between beauty ideals internalization and women’s self-objectification has been confirmed by previous studies (Huang et al., 2020). Moreover, a study on female college students’ Instagram use found that the internalization of aesthetic cultural standards mediated the relationship between Instagram use and self-objectification (Feltman & Szymanski, 2018). This suggests that women’s experiences of sexual objectification are significant triggers for internalizing ideal beauty, and these experiences can indirectly predict women’s self-objectification through beauty ideals internalization (Vandenbosch & Eggermont, 2012; Huang et al., 2020; Jankauskiene & Baceviciene, 2022; Hocking & Gondoli, 2023).

According to the Tripartite Influence Theory, the Objectification Theory, and previous studies, we propose the following hypotheses (the hypothesis figure sees Figure 1): Hypothesis 1. Beauty short video engagement positively predicts dietary restraint among young women; Hypothesis 2. Beauty ideals internalization mediates the relationship between beauty short video engagement and dietary restraint among young women; Hypothesis 3. Self-objectification mediates the relationship between beauty short video engagement and dietary restraint among young women; Hypothesis 4. Beauty ideals internalization and self-objectification have a serial mediation effect in the relationship between beauty short video engagement and dietary restraint among young women.

This survey was conducted through the online survey platform Wenjuanxing. The survey subjects were selected as women aged 18 to 35. Before the survey was carried out, the participants’ informed consent was obtained. This survey was conducted by three Chinese college students, who randomly distributed the questionnaires to women aged 18 to 35 through the Wenjuanxing online platform and collated the questionnaire survey data. In total, 346 questionnaires were collected. Then, samples with missing values were excluded. Ultimately, the study yielded 312 valid questionnaires, representing an effective rate of 90.17%. Participants’ average age was 24.37 years (standard deviation = 4.95), and their average body mass index was 20.12 (standard deviation = 2.74). The sample comprised 148 rural residents (47.44%) and 164 urban residents (52.56%). Among the respondents, 80 were only children (25.64%), and 232 were non-only children (74.36%). Educational attainment included 3 individuals (0.96%) with junior high school or lower, 14 individuals (4.49%) with high school education, 273 individuals (87.50%) with bachelor’s degrees, and 22 individuals (7.05%) with graduate degrees.

To revise the Beauty Video Usage Intensity Scale compiled by Zeng and Ye (2022) to measure beauty short video engagement. This scale is made up of six items, like “Watching beauty short video is something I do every day.” It uses a 5-point Likert scale, with 1 being “strongly disagree” and 5 being “strongly agree.” A higher score indicates greater beauty short video engagement among young women. In this study, Cronbach’ s α = 0.92. We also conducted the Cronbach’ s α if Item Deleted. The test results showed that deleting any single item would not cause the Cronbach’ s α to exceed 0.92. The results of confirmatory factor analysis showed that χ²/df = 2.70, CFI = 0.97, TLI = 0.98, RMSEA = 0.07, RMR = 0.03.

The Beauty Ideals Internalization Scale was adapted from the Internalization Subscale of the Sociocultural Attitudes Towards Appearance Questionnaire developed by Heinberg et al. (1995), with the item wording modified to better suit the context of short videos. The adapted scale consists of five items (e.g., “The people featured in short videos represent the appearance I aspire to”). With scores ranging from 1 to 5, it employs a 5-point Likert scale that goes from “completely disagree” to “completely agree.” A higher score indicates a greater severity of beauty ideals internalization among young women. In this study, Cronbach’ s α = 0.90.

The self-objectification of young women was measured using the Surveillance Subscale from the Objectified Body Consciousness Scale developed by Lindberg et al. (2006). The scale consists of four items (e.g., “I often compare how I look with how other people look”). It implements a seven-point Likert scale that spans from “strongly disagree” to “strongly agree,” with scores assigned from 1 to 7. A higher score indicates a greater level of self-objectification among young women. In this study, Cronbach’ s α = 0.90.

The dietary restraint of young women were measured using the Cognitive Restraint Subscale from the Three-Factor Eating Questionnaire revised by Karlsson et al. (2000). The Cognitive Restraint Scale is designed to measure dietary restraint, that is, control over food intake in order to influence body weight and body shape. The scale consists of six items (e.g., “I deliberately take small helpings as a means of controlling my weight”). Items 1–3 are scored from 1 to 4, ranging from “definitely false” to “definitely true”; item 4 is scored from 1 to 4, ranging from “almost never” to “almost always”; item 5 is scored from 1 to 4, ranging from “unlikely” to “very likely”; and item 6 uses a scoring range of 1 to 8. A higher score indicates more severe dietary restraint among young women. In this study, Cronbach’ s α = 0.82.

This study also collects the age, height, weight, hukou, and education level of participants. According to previous studies, age and BMI are correlated with eating disorder (Tzoneva et al., 2015; Chu et al., 2021). Therefore, the study added age and BMI into the model as control variables.

The questionnaire data were entered and preprocessed using SPSS 27.0. BMI was calculated using the formula weight/height². All variables underwent descriptive statistical analysis, with frequencies and percentages used for categorical variables, while means and standard deviations were applied to continuous variables. Pearson correlation analysis was conducted to explore the correlations between the study variables, with statistical significance set at p < 0.05. Mediation analysis was conducted using Model 6 in SPSS PROCESS 4.0. A statistically significant mediation effect is established when the point estimate falls within the confidence interval that excludes zero.

The Table 1 indicate that beauty short video engagement is correlated with beauty ideals internalization (r = 0.76, p < 0.05), self-objectification (r = 0.68, p < 0.05), and dietary restraint (r = 0.53, p < 0.05). Additionally, beauty ideals internalization is correlated with self-objectification (r = 0.74, p < 0.05) and dietary restraint (r = 0.57, p < 0.05). Self-objectification is also correlated with dietary restraint (r = 0.55, p < 0.05).

As shown in Table 2 and Figure 2, before the age and BMI were added, there was a positive association between beauty short video engagement and beauty ideals internalization (β = 0.76, p < 0.05). Beauty short video engagement was positively correlated with self-objectification (β = 0.28, p < 0.001); beauty ideals internalization was positively correlated with self-objectification (β = 0.52, p < 0.001). Moreover, Beauty short video engagement was positively correlated with dietary (β = 0.17, p < 0.05); Beauty ideals internalization was positively correlated with dietary (β = 0.27, p < 0.001). Self-objectification was positively correlated with dietary (β = 0.24, p < 0.001). After adding the age and BMI, those positive associations still exist. Additionally, to exclude the risk of collinearity, conduct a collinearity test. The VIF values of all variables in the regression results were less than 5, which means there was no multicollinearity.

As shown in Table 3, the relationship between beauty short video engagement and dietary restraint is mediated by beauty ideals internalization (indirect effect = 0.17, 95%CI: [.05, 0.29]) and self-objectification (indirect effect = 0.06, 95%CI: [0.01, 0.11]). Moreover, self-objectification and beauty ideals internalization act as serial mediators in the relationship between beauty short video engagement and dietary restraint, with a indirect effect of 0.08 (95%CI: [0.02, 0.15]).