The Paradoxical Outcomes of Observing Others’ Exercise Behavior on Social Network Sites: Friends’ Exercise Posts, Exercise Attitudes, and Weight Concern
More than one-third of U.S. adults are obese (Ogden, Carroll, Fryar, & Flegal, 2015). As such, it is perhaps not surprising that exercise and physical activity have become salient parts of a national conversation concerning weight and health. Researchers have demonstrated that media (Harrison, Taylor, & Marske, 2006) and interpersonal relationships (Burke & Segrin, 2015) can have significant consequences for weight or exercise-related health beliefs and behaviors. In recent years, scholars have begun to document the impor- tant implications of social media, including involvement in social network sites (SNSs), upon exercise-related attitudes and behaviors as well (Carpenter & Amaravadi, 2016).
In providing a mechanism for broadly sharing informa- tion about one’s life, SNSs create the potential for users to be routinely exposed to the exercise habits and physical activity of a diverse group of others ranging from close friends to distant acquaintances (Arroyo & Brunner, 2016). One can observe when and how his or her connec- tions engage in exercise or physical activity through SNS posts, which raises important theoretical and practical ques- tions. In this study, we use social cognitive theory (Bandura, 1977) and social comparison theory (Festinger, 1954) to investigate individuals’ exposure to exercise-related SNS posts from others in their networks, and their corre- sponding exercise attitudes and weight concerns. We exam- ine the conditions under which SNS use results in a paradox of simultaneously promoting healthy attitudes toward exercise and fostering a potentially unhealthy pre- occupation with one’s weight.
SNSs and communication about weight and exercise
Researchers at the Pew Internet and American Life Project estimate that approximately 70% adult Internet users in the United States
participate in one or more SNSs (Duggan & Smith, 2013). Although people use SNSs for various reasons, maintaining relationships with friends and family members is a primary function (Smith, 2011). SNSs are defined as websites making it possible to create a profile, articulate one’s connec- tions in the network, and produce and interact with streams of content created by others (Ellison & Boyd, 2013). This final component of the definition is particularly critical. It may take the form of receiving posts that have been broadcast to a user’s connections in the network (i.e., one’s “friends”), which may consist of their routine activities, thoughts, and experiences. This affordance of visibility (Treem & Leonardi, 2012) is important because behaviors that previously may have gone unmentioned—such as exercise and physical activity—are readily documented for others in SNSs.
Researchers studying health communication recognize the implications of information contributed by others in one’s SNS network for one’s health beliefs and attitudes. Posts or status updates about exercise and weight, in particular, have received attention from several scholars. Exercise among members of one activity-tracking SNS was positively asso- ciated with the number of connections or “friends” users had on the site and the volume of exercise performed by their connections (Carpenter & Amaravadi, 2016). Other researchers examining a weight loss SNS found that the amount of weight lost by members was associated with the amount of weight lost by their connections (Meng, 2016). Yet, other researchers have reported what might be considered deleterious outcomes of exercise- and diet-related SNS use. In one study, exposure to fitness posts from one’s SNS con- nections was associated with individuals’ negative body talk and social comparison (Arroyo & Brunner, 2016). Several other scholars have also reported a connection between expo- sure to weight-related information from others on SNSs and outcomes such as lower levels of psychological well-being (Lee, Taniguchi, Modica, & Park, 2013).
In the present study, we attempt to advance previous research by examining how and why SNS use can be asso- ciated with both positive and negative health outcomes. We focus specifically on the posts in the form of status updates received from one’s SNS connections or “friends” related to exercise and physical activity. Such contributions could include a variety of things related to formal exercise activities or personal fitness more broadly: reports of recent exercise, pictures of others exercising or dressed in fitness attire, fit- ness-related accomplishments or milestones, checking into the gym, fitness inspired quotations, and other activities. Drawing from social cognitive theory (Bandura, 1977) and social com- parison theory (Festinger, 1954), we investigate the possibility that exposure to others’ exercise posts in an SNS might serve to promote positive attitudes toward exercise but also encou- rage undue weight-related concerns.
Social comparison theory, social cognitive theory, and SNS exercise posts
Social comparison theory is founded on the idea that people’s opinions and abilities intersect to affect their beliefs about their behaviors (Festinger, 1954). Individuals tend to use others as a referent when evaluating themselves (Goethals, 1986), which indicates that their opinions and abilities are determined, in part, by how they compare to other people. These evaluations become particularly salient as individuals compare themselves to others to whom they are similar (Festinger, 1954), particularly with regard to characteristics related to a specific dimension under evaluation (Wood, 1989).
Social comparison theory has been used in studies examin- ing the effects of media images on body dissatisfaction (Botta, 1999), and more recently, in studies examining social media observations (Perloff, 2014). Specifically, researchers have found that observing SNS posts made by one’s network mem- bers is associated with one’s weight-related behaviors and beliefs (Arroyo & Brunner, 2016). Similarly, exposure to the exercise activities of one’s friends and acquaintances in the context of an SNS may serve to prime social comparisons. Research suggests that both parents and media influence young adults’ weight concern (Field et al., 2001); a parallel effect is likely to occur among individuals’ social networks.
Weight concern reflects one’s preoccupation with body and thinness (Killen et al., 1994). Given that weight concern is associated with constant dieting (Field et al., 2001), it is possible that individuals’ comparisons with others who post about exer- cise in their SNSs may serve to provoke an unhealthy focus on one’s own body. As such, we predict the following: H1: The number of exercise-related SNS posts received from network members will be positively related to respondents’ weight concern.
It also seems possible that SNS use might have pro-social consequences related to exercise. A central component of social cognitive theory is the notion that people learn about behaviors through observing others (Bandura, 1977). When people associate observed behaviors with positive sources or outcomes, they are more likely to develop positive appraisals about and enact those behaviors themselves (Bandura, 1977)—and this extends to the context of self-improvement (Wood, 1989). Following social cognitive theory, one’s SNS connections may serve as models promoting pro-exercise attitudes. Others’ online behavior both establishes and perpe- tuates social and cultural norms (Ridout, Campbell, & Ellis, 2012). Accordingly, individuals might see their network mem- bers’ behaviors as normative, which may precipitate corre- sponding attitude changes. As an example, individuals report increased smoking and positive attitudes toward alcohol after being exposed to their friends’ SNSs posts featuring smoking (Huang, Soto, Fujimoto, & Valente, 2014) and drinking beha- viors (Miller, Prichard, Hutchinson, & Wilson, 2014). Similarly, observing others’ beliefs and behaviors should also translate to health behaviors such as exercise. Seeing the virtues of fitness from one’s network may influence one’s exercise attitudes; therefore, we predict the following: H2: The number of exercise-related SNS posts received from network members will be positively related to respondents’ exer- cise attitudes.
Beyond demonstrating the positive and negative consequences of receiving exercise posts from members of one’s SNS network, it is critical to consider when these outcomes are more and less likely. Several moderators have the potential to intensify or mitigate the impact of exercise posts made by others on an SNS user’s exercise attitudes and weight-related beliefs. Similarity is one such moderator. Similarity plays a key role in social comparison theory (Festinger, 1954) and social cognitive theory (Bandura, 1977). In both theories, the influence of others is likely to be heightened when those others are similar to a target.
Research on body image in adolescent friendships suggests that similarity plays an important role in body image con- cerns, such that friendship cliques with greater body image and dieting concerns tend to reinforce these concerns and behaviors within their group (Paxton, Schutz, Wertheim, & Muir, 1999). Moreover, when individuals feel that they are outperformed by a close other, they experience greater arousal than if they are outperformed by a distant other (Tesser, Millar, & Moore, 1988). Perceiving oneself to be similar to SNS members who post about exercise could influence one’s weight concerns and exercise attitudes. Accordingly, we pre- dict the following : H3: The association between the number of exercise-related SNS posts received from network members and respondents’ (a) weight concern and (b) exercise attitudes will be moder- ated by respondents’ perceived similarity to the exercise pos- ters. These associations will become stronger as respondents perceive themselves to be more similar to members of their SNS network who make exercise-related posts.
Two additional moderators can be derived from social comparison theory involving the general tendency to make upward and downward social comparisons (Festinger, 1954). The social comparison process involves two steps: evaluating an attribute (e.g., fitness), and con- sidering the factors associated with the attribute under evaluation (e.g., others’ fitness abilities and constraints in comparison to one’s own fitness abilities and constraints; Wood, 1989). When making comparisons, people tend to look to those who possess superior or inferior levels of an attribute. This tendency to make upward or downward comparisons has important implications for the outcomes of receiving SNSs posts about exercise.
People who tend to make upward comparisons see themselves in light of others who possess superior quali- ties of an attribute. In the context of receiving SNSs posts about exercise, this could involve attending to and being influenced by posts from others in one’s network who are perceived to have superior levels of fitness. The tendency to focus on superior others should strengthen the para- doxical effects of viewing others’ exercise posts and may reinforce the health benefits of exercise. At the same time, it could also highlight the notion that one’s present level of fitness could be improved and heighten undue weight concern. Because this study focuses on exercise and fit- ness, the tendency toward making social comparison regarding appearance is examined in conjunction with others’ posts and individuals’ weight concern and exercise attitudes. We predict the following: H4: The association between the number of exercise-related SNS posts received from network members and respondents’ (a) weight concern and (b) exercise attitudes will be moder- ated by respondents’ tendency toward making upward social comparisons. These associations will become stronger as respondents have a greater tendency to engage in upward comparisons.
In contrast, people who tend to make downward compar- isons may be less impacted by others’ SNS posts about exercise. In focusing on others with inferior levels of fitness, one may discount the actions of the poster and the benefits of exercise. Additionally, viewing a less fit model should serve to bolster one’s feelings about one’s own fitness and mitigate concerns about one’s weight- related deficiencies. We predict that: H5: The association between the number of exercise-related SNS posts received from network members and respondents’ (a) weight concern and (b) exercise attitudes will be moderated by respondents’ tendency toward making downward social compar- isons. These associations will become weaker as respondents have a greater tendency to engage in downward comparisons.
Method
Recruitment procedure and participants
Following Institutional Review Board approval, participants were recruited using Amazon’s Mechanical Turk (MTurk).Individuals registered on MTurk can earn small sums of money for accomplishing tasks such as completing a ques- tionnaire (for an introduction, see Buhrmester, Kwang, & Gosling, 2011). MTurk workers have been shown to provide adequate samples for social scientific research (Berinsky, Huber, & Lenz, 2012).
Among the 394 participants who started the study, 304 com- pleted the questionnaire in its entirety (77%); data from the 90 participants who did not fully complete the questionnaire were removed from the final sample. An additional 72 participants were excluded because they indicated avoiding exercise or phy- sical activity. Given that other research on physical activity and social networks was limited to participants who engaged in exercise (Hamari & Koivisto, 2013), and that our goal in this study was to better understand exercise attitudes, it was critical to limit the sample to those people who did not actively avoid exercise. As a result, the final sample included 232 participants who reported engaging in at least some exercise.
Participants in this study included 101 males (44%) and 131 females (56%), with a mean age of 31.33 years old (SD = 10.89, range = 18 to 72 years old). The racial makeup of the participants was 1.7% American Indian/Alaskan Native, 8.2% Asian/Pacific Islander, 6.5% African American, 6% Latino/a, 75% Caucasian, and 2.6% other. In terms of educa- tion achieved, 0.04% reported less than high school, 13.8% reported completing high school, 29.3% reported completing some college, 13.4% reported having an Associate’s degree, 30.2% reported having a Bachelor’s degree, 5.6% reported completing some graduate school, and 7.3% reported com- pleting graduate school.
The average Body Mass Index (BMI) of participants was 25.40 (SD = 5.85, range = 16.91 to 54.06), with 3.9% of participants categorized as underweight, 50.4% categorized as healthy weight, 28.9% categorized as overweight, and 16.4% categorized as obese. Among the participants who reported about receiving posts from others in their SNS net- work in the last 30 days, 44% reported receiving posts about exercise from 1 to 2 people; 24% reported receiving posts about exercise from 3 to 4 people; 14% reported receiving posts about exercise from 5 to 6 people; 3% reported receiving posts about exercise from 7 to 8 people; 4% reported receiving posts about exercise from 9 to 10 people; 10% reported receiving posts from more than 10 people; and 1% did not receive posts from others in their SNS network in the last 30 days.
Procedure
Participants were instructed to log in to their most frequently used SNS platform (e.g., Facebook, Google+, etc.), and review the status updates received during the previous 30 days. Status updates were defined as messages that are directly and simul- taneously disseminated to all of a user’s connections on the site. Participants were asked to record information specifically for the updates they received about exercise, including their perceptions of the message sender. Exercise was defined for participants as any activity involving physical exertion for the purpose of maintaining fitness and health. Going for a walk, jog, hike, bike ride, to the gym, and yoga were all listed as examples of exercise.
Measures
Number of exercise-related posts
The number of SNS posts about exercise made by others in one’s social network was measured with a single item, which asked, “How many status updates have you received in the past 30 days that were about exercise?” Response options included, “none,” “1–2,” “3–4,” “5–6,” “7–8,” “9–10,” and “more than 10.” Eighteen percent of participants reported receiving no posts, 29% reported receiving 1–2 posts, 14%
reported receiving 3–4 posts, 7% reported receiving 5–6 posts, 7% reported receiving 7–8 posts, 3% reported receiving 9–10 posts, and 22% reported receiving more than 10 posts.
Perceived similarity
Respondents chose the people from their SNSs who made the most exercise-related posts (up to 3 total) and rated their perceived similarity to each person using the following item, “How similar is this person to you?” Response options ranged from 1 = “Not at all Similar” to 7 = “Very Similar.” The similarity ratings for the three people were averaged to create a single perceived similarity measure reflecting the aggregate similarity between participants and SNS members who most frequently made exercise status updates.
Social comparison
Respondents’ tendencies to make upward and downward social comparisons were measured using the upward physical appear- ances comparisons scale (UPACS) and the downward physical appearances comparisons scale (DACS) developed by O’Brien et al. (2009). Sample items from the 8-item upward social com- parison subscale included, “I tend to compare myself to people I think look better than me,” and “When I see good-looking people I wonder how I compare to them.” Sample items from the 8-item downward social comparison subscale included, “I compare myself to people less good looking than me,” and “I tend to compare my body to those who have below average bodies.” Response options for both subscales ranged from 1 = “strongly disagree” to 7 = “strongly agree”; higher scores indicated a greater tendency to make upward or downward comparisons. Cronbach’s alpha reliability in this study was .96 for upward social comparison and .95 for downward social comparison.
Weight concern
Individuals’ weight concern was assessed using Killen et al.’s (1994) 5-item weight concern scale. The wording for the response options matched the scale items, but all were scaled such that higher scores represented higher weight concern. As an example, the response options for the item “How afraid are you of gaining weight?” ranged from 1 = “not afraid of gaining weight” to 5 = “very afraid of gaining weight.” Cronbach’s alpha for the measure was .81.
Exercise attitudes
Feelings regarding participation in regular exercise were assessed using Courneya and Bobick’s (2000) 8-item exercise attitudes measure. Sample items included, “For me participat- ing in regular exercise is boring,” (reverse scored) and “For me, participating in regular exercise is beneficial.” Response options ranged from 1 = “strongly disagree” to 7 = “strongly agree.” Cronbach’s alpha reliability for the measure was .83.
Control variables
Participants’ gender and BMI were included as control vari- ables. Each of these variables has the potential to influence one’s body image (Foster, Wadden, & Vogt, 1997) and atti- tudes toward exercise. Including these two measures as con- trol variables ensures that the results are not an artifact of either variable.
Results
The hypotheses were examined in eight separate multiple regression models tested using SPSS. Separate models were constructed for each of the four hypotheses with weight con- cern as the outcome variable and for each of the four hypoth- eses with exercise attitudes as the outcome variable. Gender and BMI were significant controls (p < .05) in all four models examining weight concern as the outcome variable, but not in the models examining exercise attitudes as the outcome vari- able. The interactions were tested and decomposed following the guidelines proposed by Aiken and West (1991), who state that “if z is continuous. . . then the investigators are free to choose any value within the full range of z” (p. 12) for plotting interactions. Thus, the conditional values for the simple slopes were calculated at the 25th and 75th percentiles. Means and correlations among study variables are presented in Table 1. Unless otherwise noted, higher scores indicate a greater amount of a variable. Hypothesis 1 addressed the relationship between exposure to others’ exercise-related SNS posts and weight concern. The results indicated that the number of exercise-related SNS posts received from others was positively related to respondents’ weight concern (see Table 2). As the number of exercise posts received from SNS network members increased, so too did respondents’ weight concern, R2 = .04. Hypothesis 2 addressed the relationship between exposure to others’ exercise-related SNS posts and exercise attitudes. In this case, the number of exercise-related SNS posts received from others was not related to respondents’ exercise attitudes, R2 = .00. Thus, Hypothesis 1 was supported, but Hypothesis 2 was not supported. Building on these hypotheses, Hypothesis 3 examined whether respondents’ perceived similarity to the most fre- quent exercise posters in their social network moderated the association between the number of others’ exercise-related SNS posts and respondents’ (a) weight concern and (b) exercise attitudes (see Table 2). The interaction between the number of others’ exercise-related SNS posts and perceived similarity was statistically significant for weight concern, R2 = .02. The results can be observed in Figure 1. When perceived similarity was low, the relationship between number of exercise posts received and weight concern was not signifi- cant, b = −0.02, SE = 0.04, p = .65. When perceived similarity was high, however, the relationship between number of exer- cise posts received and weight concern was positive and sig- nificant, b = 0.07, SE = 0.04, p = .04. These results were consistent with H3a. The positive association between volume of exercise posts received and weight concern was stronger among respondents who perceived themselves to be more similar to the most frequent exercise posters in their SNS network. In contrast, respondents’ perceived similarity to frequent exercise posters in their SNS network did not mod- erate the association between the number of posts received and exercise attitudes, R2 = .01. H3b was not supported. Hypothesis 4 addressed whether the association between the number of exercise-related SNS posts received from others and respondents’ (a) weight concern and (b) exercise attitudes would be moderated by respondents’ tendency toward upward social comparison regarding appearance (see Table 3). Respondents’ tendency toward upward social comparison did not moderate the association between the number of others’ exercise-related SNS posts and weight concern, R2 = .01. In contrast, the interaction between the number of others’ exercise-related SNS posts and tendency toward upward social comparison regarding appearance was statistically significant for exercise attitudes, R2 = .02. As illustrated in Figure 2, the slope for low upward comparison was not significant, b = −0.01, SE = 0.04, p = .73, but the slope for high upward comparison was significant, b = 0.08, SE = 0.04, p = .03. The positive association between the volume of exercise posts received and exercise attitudes was stronger among participants who had a greater tendency to make upward comparisons regard- ing physical appearance. Thus, Hypothesis 4b was supported, but Hypothesis 4a was not supported. Hypothesis 5 considered whether the association between the number of exercise-related SNS posts received and respondents’ (a) weight concern and (b) exercise attitudes would be moderated by respondents’ tendency toward down- ward social comparison regarding appearance (see Table 3). The results indicated that respondents’ tendency toward downward social comparison did not moderate the associa- tions between the number of others’ exercise-related SNS posts and respondents’ weight concern, R2 = .01, nor respon- dents’ exercise attitudes, R2 = .00. Hypotheses 5a and 5b were not supported. Discussion This study employed social comparison theory and social cogni- tive theory to examine whether individuals’ exposure to exercise- related SNS posts was related to their weight concern and exercise attitudes. In addition, we investigated whether these associations were moderated by individuals’ perceived similarity to frequent exercise posters in their SNS networks, as well as individuals’ tendency toward making appearance-related upward and down- ward social comparisons. The results suggest that individuals’ exposure to their SNS connections’ exercise-related posts might have paradoxical implications for their weight and exercise attitudes. Specifically, the results from this study demonstrated that as the number of others’ exercise-related SNS posts increased, respon- dents reported greater weight concern. This relationship was further moderated by perceived similarity, such that the associa- tion between exercise-related SNS posts received and weight con- cern was stronger among respondents who felt more similar to the members of their SNS making the exercise-related posts. Consistent with both social comparison theory (Festinger, 1954) and social cognitive theory (Bandura, 1977), respondents’ per- ceived similarity to exercise posters on their SNSs appears to be a particularly salient feature of their weight-related beliefs. Similar others likely provide a rather obvious benchmark from which to evaluate one’s weight-related (in)adequacy. Indeed, research suggests that peers influence one another’s body image concerns (Paxton et al., 1999) and that individuals are particularly concerned with the behaviors of close others (Tesser et al., 1988). In the context of a SNS where individuals elect to connect with “friends” with whom they may identify (Livingstone, 2014), it is not surprising that individuals relate to exercise posts made by these connections. Although people might have positive intentions with regard to posting about exercise online (e.g., pro- moting accountability and motivation; Teodoro & Naaman, 2013), these posts can result in observers’ body dissatisfaction (Arroyo & Brunner, 2016). Over time, these posts may even encourage unhealthy compensatory behaviors that correspond with weight concern (e.g., disordered eating; Field et al., 2001). We hypothesized parallel associations with exercise attitudes as the outcome variable, but those associations were nonsignificant. It is possible that people’s exercise attitudes are more stable, and are not as easily influenced by perceived similarity to frequent exercise posters. Given the cross-sectional nature of this study, it is also possible that the associations flow in the opposite direction of what we expected. People with greater weight concern may perceive themselves to be similar to others who post about exercise. Such an explanation, however, would seem to be limited to individuals who exercise to manage their weight concern. In contrast with the potential deleterious consequences invol- ving weight concern, the results also provide evidence that exposure to others’ exercise behavior on SNSs may foster some beneficial outcomes. Evidence of the potential for positive influ- ence lies in the finding that, among respondents with a greater tendency toward making upward social comparisons regarding appearance, receiving exercise-related SNS posts from others was associated with more positive exercise attitudes. Although there are extreme circumstances where excessive exercise can be unhealthy, the overwhelming health benefits associated with exercise are indicative of a generally pro-health behavior. These results illustrate a pattern wherein observing exercise- related posts from SNS connections may reinforce desirable fitness behavior among people who were more inclined to make upward social comparisons, making such activity seem normative, or even rewarding. According to social cognitive theory, one’s evaluation of others’ behaviors as rewarding likely results in one repeating those behaviors (Bandura, 1977). Thus, repeated exposure to others’ exercise achievements on SNSs could motivate adoption of healthier exercise behaviors. These results are also consistent with research suggesting that social comparison plays a role in observations of others’ weight-related posts on SNSs (Arroyo & Brunner, 2016) and that SNS posts have the potential to influence observers (Miller et al., 2014). They are unique from prior research, however, in that the outcomes of social comparison appear to be relatively pro-health. Taken as a whole, the results suggest that witnessing the health behavior of members of one’s SNS network could be consequential. Previous researchers have documented detri- mental consequences related to smoking (Huang et al., 2014) and alcohol consumption (Miller et al., 2014). Although our results also demonstrate some potentially negative outcomes related to undue weight concern, they also suggest that the implications of SNSs are more nuanced. Among respondents with a greater tendency to make upward appearance-related social comparisons, viewing their SNS connections’ exercise behavior was associated with pro-exercise attitudes. The potential for positive health-related outcomes stemming from SNS use warrants greater attention from scholars. Such efforts are critical to uncover the potentially sophisticated ways in which SNS use can benefit and undermine health. This research might also be valuable in the realm of social media intervention development. If developers understand how to trigger positive exercise attitudes and behaviors, it could facilitate the design of content algorithms or post pro- motion that could result in healthier exercise attitudes and behaviors. In addition to the significant findings, it is noteworthy that the propensity to make downward appearance-related compar- isons did not moderate the impact of receiving exercise related posts from others. It was expected that people who tend to make downward social comparisons would discount the exercise posts received from network members. It may be that the tendency to make downward comparisons has a direct impact on weight concern and exercise attitudes. Indeed, the zero-order correla- tions reported in Table 1 indicate that downward comparison tendency was positively associated with weight concern and negatively associated with exercise attitudes—although the latter relationship did not meet the conventional criterion for statis- tical significance. The tendency to make downward appearance- related social comparisons may reflect a broader underlying negative self-perception that is more influential than contextual factors such as information from others in one’s social network. Limitations and future directions Despite this study’s contributions, there are some limitations that should be noted. First, this population included fewer obese indi- viduals than the U.S. population in general, with about half of the participants’ self-reported height and weight classifying them as having healthy weight or underweight BMIs. In future research, it would be helpful to examine the study variables in a sample of participants whose weight is more representative of the U.S. popu- lation. Consistent with the established literature on social compar- ison theory, we assessed participants’ tendency toward making upward and downward social comparisons in general; however, it might be instructive to understand the tendency to make social comparisons to specific social network members. Although participants reported on the number of exercise- related SNS posts received and compared themselves to the most frequent exercise posters on their SNSs, we did not ask partici- pants to report the content or nature of these exercise-related posts. It could be that some post characteristics are motivational, thereby promoting healthy behavior change, whereas other post features result in body dissatisfaction and unhealthy compensa- tory behaviors. Future research should more directly explore the content of exercise and weight-related SNSs posts and ascertain whether particular post content differentially impacts observers’ beliefs and behaviors. Finally, the data presented here are cross- sectional. As such, it would be valuable to examine the implica- tions of exercise related posts experimentally or across different time points to further understand their impact on individuals’ weight- and exercise-related attitudes and behaviors. Conclusion Given the epidemic of obese U.S. adults, the comorbid health conditions associated with obesity, and the prevalence of SNS involvement, it is important for researchers to examine SNSs as a potential source of intervention and health promotion. In this study, we demonstrated that individuals’ exposure to exercise- related SNS posts from others in their networks was related to greater weight concern. This relationship was stronger to the extent that individuals reported greater similarity to the most frequent exercise posters in their SNS network. In contrast, the relationship between others’ exercise-related SNS posts and indi- viduals’ exercise attitudes was stronger to the extent that these individuals had a tendency to make upward social comparisons regarding appearance. Together, these results demonstrate that others’ SNS posts can influence people’s health-related beliefs, and that this potential influence might be contingent upon peo- ple’s comparison of themselves ZK53 to others within their social network.