Body

Today, young people spend much of their lives online where they encounter abundant appearance-focused content. In light of the important role that social media platforms like Instagram play in young people’s attitudes towards the cosmetic surgery industry, this study investigates passive and active social media usage in relation to young adults’ perceptions of cosmetic procedures. In a cross-sectional survey study among 470 Instagram users aged 18–25 years ( M age = 21.00, SD age = 2.26), young adults indicated a low cosmetic procedure intention themselves, but overestimated the prevalence of cosmetic procedures undertaken by others. Considering passive social media use, young adults who follow influencers who had undergone cosmetic procedures reported higher intentions to have cosmetic procedures themselves, and following influencers who had not undergone procedures was related to lower intentions. In terms of active social media usage, frequency of Instagram posting was generally insignificant in relation to acceptance of and intention to undergo cosmetic procedures. Yet, those who used filters to edit pictures more frequently reported increased cosmetic procedure acceptance and intention. Overall, this paper offers a nuanced perspective on the relation between young adults’ perceptions of cosmetic procedures and their social media behaviors, highlighting the importance of specific appearance-oriented social media usage.


Introduction
The popularity of cosmetic procedures has increased over the past decades (ASPS 2020; ISAPS 2020). Moreover, there is some evidence that the number of young adults undergoing cosmetic procedures has been growing, although this trend is not unequivocal (ASPS 2020(ASPS , 2015(ASPS , 2010. In the Netherlands, Zarringam, Decates, Slijpe, and Velthuis (2020) found a particular increase in the proportion of young adults between the ages of 18 and 25 who opt for injectable treatments. The reasons for the increased popularity and even normalization of cosmetic procedures are diverse and may relate to individual, interpersonal, and societal factors.
Both passive and active use of particularly highly visual social media like Instagram, which propagates appearance-focused content, play a significant role in attitudes towards cosmetic procedures. For example, both viewing and/or liking celebrity content on Instagram and exposure to images of women believed to have undergone facial cosmetic procedures are associated with increased acceptance of and desire for cosmetic procedures (Di Gesto et.al, 2021;Walker et al., 2021). In general, appearance-related content has been shown to have a more significant impact on body imagerelated variables and cosmetic surgery intention compared to nonappearance focused social media content (Bij de Vaate, Veldhuis, & Konijn, 2020;Cohen, Newton-John, & Slater, 2017;Holland & Tiggemann, 2016). In line with traditional communication theorizing H1. : The frequency of using social media, and particularly highly visual social media, will be positively related to (a) cosmetic procedure acceptance; (b) cosmetic procedure intention; and (c) normalization.
H2. : Following appearance-focused social media content (i.e., following influencers, clinics, and cosmetics brands) will be positively related to (a) cosmetic procedure acceptance; (b) cosmetic procedure intention; and (c) normalization.
H3. : The frequency of posting on Instagram and usage of Instagram filters will be positively related to (a) cosmetic procedure acceptance; (b) cosmetic procedure intention; and (c) normalization.

Method
Using a commercial survey company, we administered a survey amongst a sample of Dutch young adults (18-25 years), representative in terms of gender, educational attainment, and place of residence. We excluded respondents younger than 18 or older than 25, those who failed both attention checks, and those who did not have an Instagram account. The final sample consisted of 470 responses. Table 1 presents all details and measures. It was particularly important to account for gender as cosmetic procedures are highly gendered; it is estimated that around 90 % of all recipients of cosmetic procedures identify as female (ASPS, 2020;ISAPS, 2020). Nevertheless, the moderating role of gender in social media research has been contested as several studies into the relation between social media usage and body image disturbances and/or attitudes towards cosmetic procedures did not find gender to be a significant moderator (cf. De Vries, Peter, Nikken, & de Graaf, 2014;Saiphoo & Vahedi, 2019).

Data analysis
We first checked the frequencies of all measures to gain insights into mean scores and percentages. To test our hypotheses, we ran stepwise regression analyses for the three dependent variables. In the first model, we only included the social media factors, in model 2 we added the personal factors, and in model 3 we added the demographic variables. Detailed results can be found in Tables 2, 3, and 4. In addition, Table 5 presents a correlation matrix of all continuous variables.

Descriptive statistics
As can be seen in Table 6, the hypothetical intention to undergo different cosmetic procedures was quite low in our sample. Means range from 1.68 (Botox) to 3.38 (laser treatment), and thus were below the neutral point (4 = Maybe). On average, most respondents (> 67 %) indicated not wanting any of the procedures under any circumstances. Acceptance of cosmetic procedures varied: 40 % found cosmetic procedures (highly) unacceptable (scores 1-3), 51 % was neutral (scores > 3 and < 5), and only 9 % indicated that they thought procedures were (highly) acceptable (scores 5-7).
The mean estimated percentages for the procedures measured in the normalization scale (see last column in Table 6) were very high and provided evidence for the normalization of cosmetic procedures, as young adults believed these procedures to be much more common than they actually are. For example, respondents estimated that approximately a third of the Dutch population had opted for Botox and/or filler treatments, whereas Decates, De Wijs, Nijsteen, and Velthuis (2018) estimate that this figure is closer to just 3 % of Dutch adults.
The three dependent variables have significant, positive weak to moderate correlations with each other (intention -acceptance r = 0.44, p < .001; intention -normalization r = 0.19, p < .001; acceptance -normalization r = 0.14, p = .002), demonstrating the importance of the three separate constructs.

Hypothesis testing
Regarding H1, results showed that the frequency of using highly visual social media was positively and significantly related to cosmetic procedure intention (Table 3, model 3 beta = 0.12, p = .014). No significant relation with the frequency of using other social media was found. Thus, H1b can only be supported for highly visual social media, H1a and H1c were not supported.
With respect to H2, it was found that following social media influencers who are open about their cosmetic procedures was significantly and positively related to cosmetic procedure intention (Table 3, model 3 beta = 0.29, p < .001) and normalization (Table 4, model 2 beta = 0.17, p = .037; though, the latter relation did not hold when controlling for demographics; model 3 beta = 0.11, p = .189).  • All in all, I am inclined to think that I am a failure (R) • I take a positive attitude toward myself (1 = Strongly disagree, 7 = Strongly agree) Items were presented in random order, and five items were recoded (R) so a higher score on the items represents more positive self-esteem. Personal experience with cosmetic procedures Have you had the following procedures performed? (1 = Strongly disagree, 7 = Strongly agree) One item was recoded (R), so higher scores represent more acceptance.
Cronbach's alpha = 0.91; M = 3.40, SD = 1.14 Adapted from Henderson-King and Henderson King (2005) Vicarious experience with cosmetic procedures How many people do you personally know (for instance friends or family) who have undergone cosmetic procedures?
Although following clinics was positively related to cosmetic procedure acceptance (Table 2, model 1 beta = 0.20, p = .002) and intention (Table 3, model 1 beta = 0.12, p = .047), and following cosmetics brands was positively related to normalization (Table 4, model 2 beta = 0.15, p = .026), these relationships did not hold when controlling for relevant personal and demographic factors. This means that partial support for H2b was found, only with respect to following influencers with cosmetic procedures, and some support for H2a and H2c.
With regard to active social media behavior (H3), the frequency of filter use on Instagram was positively related to cosmetic procedure acceptance (Table 2, model 3 beta = 0.10, p = .043)  Note. Significant relationships are bold. b = unstandardized regression coefficient; se = standard error associated with the b coefficient; beta = standardized regression coefficient. *** p < .001, ** p < .01, * p < .05, + p < .10. and intention (Table 3, model 2 beta = 0.11, p = .023; although the latter was only significant when not controlling for demographics, model 3 beta = 0.08, p = .084). The frequency of posting on Instagram was positively related to normalization, when not controlling for relevant personal and demographic factors (Table 4, model 1 beta = 0.13, p = .013). This means H3a and H3b can be supported but only for Instagram filter usage, and there was no strong support for H3c.

Personal and demographic factors
Although we only wanted to control for personal and demographic factors, the results provided useful insights into their significance. More specifically, appearance orientation, body satisfaction, and personal and vicarious experience were found to relate to young adults' acceptance of cosmetic procedures, but acceptance did not depend on gender, education, or age. Note. Significant relationships are bold. b = unstandardized regression coefficient; se = standard error associated with the b coefficient; beta = standardized regression coefficient. *** p < .001, ** p < .01, * p < .05, + p < .10.
Specifically, the results showed that cosmetic procedure intentions were positively related to appearance orientation (beta = 0.13, p = .005); and that mostly females (beta = 0.10, p = .037), and those with previous personal experience with procedures (beta = 0.17, p < .001) were more likely to want to undergo procedures.

Discussion
Our results offer a nuanced perspective on the relation between young adults' perceptions of cosmetic procedures and their social media behaviors, highlighting the importance of specific appearance-oriented social media usage. The findings of this study elucidate that passive and active use of highly visual social media relate to the increased acceptance and normalization of cosmetic procedures as well as the (hypothetical) intention to undergo them.
First, more frequent use of highly visual social media related to increased levels of cosmetic procedure intention. Thus, it can be argued that the more frequently young adults use social media like Instagram and TikTok, the more likely they are to consider cosmetic procedures (and vice versa). This finding aligns with the notion that highly visual social media enable and stimulate appearance-focused content, which has been shown to relate to increased positive attitudes towards cosmetic procedures through higher appearance orientation and investment (cf. De Vries et al., 2014;Sharp et al., 2014). As also shown in this study, higher appearance orientation was related to higher cosmetic procedure acceptance and intention.
Next to a general sense of the relation between social media usage and perceptions of cosmetic procedures, we differentiated social media usage into active and passive uses. With regard to passive social media use, evidence was found for the persuasive potential of social media influencers: Following influencers who had undergone cosmetic procedures to a higher extent, aligned with increased perceptions of procedures' normality as well as a higher intention to undergo cosmetic procedures. This positive association is in line with what could be expected from cultivation principles (Gerbner et al., 2002), theorizing on mimicking behavior, and the Doppelganger effect (Ki & Kim, 2019). Nevertheless, it must be noted that from our measures it cannot be concluded that people would want to mimic the influencer's cosmetically enhanced appearance exactly by getting the same procedures. Despite anecdotal evidence by cosmetic doctors and surgeons on recipients of cosmetic procedures who bring along images of the influencers they want to look like (Lukas, 2021;Strugatz, 2018), it may be that influencers' openness on (their) cosmetic procedures lowers the threshold for cosmetic procedures more generally.
Although following influencers who had undergone cosmetic procedures positively related to cosmetic procedure intention and perceived normalization, there was no significant relation with cosmetic procedure acceptance. This is slightly surprising, as, based on cultivation principles (Gerbner et al., 2002), greater exposure to cosmetic procedures content would be expected to relate to a higher sense of normality as well as acceptability. Nevertheless, the absence of a relation between following influencers and cosmetic procedure acceptance can be explained by the specific content of the acceptance-scale that directly relates undergoing cosmetic procedures to benefits for people's self-image: The intention and normalization scales contain items that purely focus on the cosmetic procedures, and refrain from further connecting these procedures to what they mean for people's self-image. Table 5 Correlation matrix. Note. *** p < .001, ** p < .01, * p < .05, + p < .10.
A.-M. Hermans, S.C. Boerman and J. Veldhuis Body Image 43 (2022) 440-449 Whereas a positive relation was found between following influencers who had undergone cosmetic procedures and young people's intention to undergo procedures, this relationship could not be generalized to following influencers more generally. On the contrary, our results revealed that following influencers who have not had cosmetic procedures related to a lower intention to undergo cosmetic procedures, which illustrates the need to differentiate influencer content. This finding is in line with the logic of mimicking behavior (Ki & Kim, 2019): After all, if the influencer does not engage in particular (appearance-based) consumption patterns or behaviors, respondents would follow this example. Moreover, it is important to emphasize that following influencers who have not undergone cosmetic procedures may be a conscious decision: It could be that people follow only influencers who have not had a cosmetic procedure so that they would not be exposed to this particular type of appearance-focused content. However, this hypothesis would need further examination as we did not examine respondents' evaluations of or motivations for (not) following particular influencers.
Besides the relations between following influencers and perceptions of cosmetic procedures, partial evidence was found to support the notion that exposure to (other) appearance-focused content positively influenced people's perceptions of these procedures (cf. Sharp et al., 2014). Interestingly, following cosmetic clinics -which could indicate an interest in cosmetic procedures -only significantly related to cosmetic procedure acceptance and intention when not accounting for personal factors or demographics. Moreover, following cosmetics brands did not significantly relate to cosmetic procedure acceptance and intention, but was associated with a higher perceived prevalence of cosmetic procedures. In conclusion, we argue that the relation between passive exposure to different types of (non) appearance-related content and perceptions of cosmetic procedures is equivocal and multifaceted. Moreover, the different types of social media content must be discerned when investigating relations with and/or impact on perceptions of cosmetic procedures.
Lastly, this study investigated the expectation that active social media usage, in terms of both the frequency of posting pictures and the use of filters to enhance these pictures, would be positively related to cosmetic procedure acceptance, intention, and normalization. Whereas the frequency of posting was not significant, active engagement with content -in the form of applying filters -positively related to cosmetic procedure acceptance and intention (cf. Chen et al., 2019;Shome et al., 2020). A potential explanation for this is that the editing process highlights the apparent unfixedness and malleability of entities, in particular the body, an idea which is inherent to the field of cosmetic medicine.
Overall, considering the active versus passive social media use paradigm in relation to cosmetic procedure perceptions, the results reveal no simple, unidirectional relationship between following particular (non-) appearance-based content, actively posting of content -with or without prior editing -and perceptions of cosmetic procedures. Nevertheless, in terms of hypothetical intention to undergo cosmetic procedures, it seems that both active and passive social media use may play a role: Frequently using highly visual social media, particularly following influencers who have undergone cosmetic procedures, as well as editing images oneself, are positively related to intention.
A final note relates to the relatively unexpected low interest in cosmetic procedures among the sample, on the one hand, but also the inflated perceived prevalence of cosmetic procedures in the Netherlands on the other hand. To explain the lack of enthusiasm for procedures, it must be acknowledged here that some of the treatments that the respondents could choose from (e.g., Botox treatments) are generally less popular among a younger demographic (ASPS, 2020). Contrary to their own relative disinterest, the young adults in this study overestimated the popularity of procedures. This misperception can be considered a clear indication of the normalization of procedures, which relates to several factors (cf. Sarwer, 2019;Swami et al., 2008), but which has been advanced and affirmed by diverse media platforms in terms of the publication of an abundance of content related to cosmetic procedures and how this content has framed procedures (Hermans, 2021).

Strengths, limitations, and future research
As for most studies, this study holds some specific strengths and limitations that can be considered in light of implications for future research. Regarding the study design, several points need to be made. Firstly, it must be acknowledged that by asking people whether they would undergo cosmetic procedures free of charge, we measured a hypothetical situation; however, the adoption of this measure was in line with previous literature as financial factors may otherwise have proven prohibitive (cf. Chen et al., 2019;De Vries et al., 2014;Sharp, et al., 2014;Walker et al. 2021). Nevertheless, it would be interesting to see whether removing the free nature of procedures would influence respondents' intention. Furthermore, our cross-sectional design suits the scope of this present study, yet no claims pertaining to causality can be made. Descriptive statistics of intention to undergo, experience with, and estimated percentages of population that did ten popular cosmetic procedures.
Mean intention % no chance under any circumstance (score 1) % likely to certainly willing to undergo procedure (scores 5-7) % of respondents with experience with procedure Future studies could apply longitudinal or experimental designs to further investigate the causal relationships between social media behavior and cosmetic procedure acceptance, intention, and normalization beliefs (cf. Walker et al., 2021). Furthering causality, reciprocal relationships should also be considered: It could well be that social media use instigates specific beliefs about cosmetic procedures, but reversely, one could also argue that those with specific beliefs about and intentions to undergo cosmetic procedures may be motivated to seek specific media content to further guide and reinforce their needs (cf. uses and gratifications principles; Chen et al., 2019;Perloff, 2014). Another point to be considered in future studies is the social context in which young people experience social media content which may further shape their ideas about cosmetic procedures. On social media, content is directly accompanied by peer feedback in the form of likes or online comments (cf. reasoning in Holland & Tiggemann, 2016;Perloff, 2014;Veldhuis, 2020). Such peer feedback might further reinforce specific appearance-ideals and the normalization of cosmetic procedures or, in contrast, downplay these (also cf. principles of Walther's Hyperpersonal Model, 1996). Such an assumption is derived from research by, for example, Veldhuis, Konijn, and Seidell (2014), who found that online comments reflecting on the weight status of media models could either further idealize or counteract the ideality of media models' bodies, resulting in respectively more or less body dissatisfaction in adolescent girls. The social context in which the media models were evaluated guided the impact they had on girls' body perceptions. Hence, most likely, (online) peer influences also interfere with how cosmetic procedures are perceived.