Interdisciplinary Social Research Reports

Influence of Social Media on Public Support for Cancer Prevention Policies

Ming Milano Li1, Zhan Thor Tuo2, Xinshu Zhao3*

1 Department of Government and Public Administration, Faculty of Social Sciences, University of Macau; email: yc17316@connect.um.edu.mo

2 Centre for Empirical Legal Studies, Faculty of Law, University of Macau; email: yc27203@connect.um.edu.mo

3 Department of Communication, Faculty of Social Sciences, University of Macau; email: xszhao@um.edu.mo

* Corresponding Author

Research Article

Received: 12 July 2024 / Accepted: 03 October 2024 / Published online: 30 October 2024

Introduction

In an era where digital platforms profoundly influence public discourse, understanding the relationship between social media use and public health perspectives is more crucial than ever. As a leading cause of mortality, cancer remains at the forefront of public health challenges ( Ferlay et al., 2019 ), and cancer prevention remains a pivotal concern in public health. Addressing this requires medical advancements and robust public health policies ( Moore et al., 2019 ). However, the strategies for policymaking demand a deep understanding of the factors influencing public attitudes.

Various cancer preventive strategies focus on reducing tobacco use and exposure to secondhand smoke, managing alcohol consumption, encouraging active lifestyles, promoting healthy eating habits, and maintaining a healthy weight ( Espina et al., 2018 ). All National Cancer Control Plans (NCCPs) include tobacco control in their objectives, actions, or recommendations, outlining measures to control alcohol consumption and promote a healthy diet ( Atun et al., 2009 ).

Additionally, the primary prevention of cancer would be the most cost-effective, and the leading carcinogenic factors are smoking, alcohol consumption, and unhealthy diet intake ( Gelband & Sloan, 2007 ). Therefore, targeted public health interventions to raise awareness and foster positive attitudes toward cancer prevention are critical. Such interventions should focus on reducing tobacco and alcohol use and promoting a healthy diet. Moreover, evidence indicates that men are prone to hold negative attitudes toward cancer and cancer prevention ( Keeney et al., 2010 ). As a result, the public attitude to cancer prevention is the essential tunnel to predict their later preventive behaviors.

Social media is rich with cancer-related information ( Attai et al., 2016 ), and the news could alter individuals’ perceptions of cancer, which influences their views on cancer prevention and health behaviors ( Hong, 2013 ). Existing literature extensively explores the role of social media in shaping individual health perceptions ( Calvo & Ventura, 2021 ; Raggatt et al., 2018 ), and there’s a notable gap in understanding the impact of social media on specific public health policy attitudes. This study aims to bridge this gap by examining how social media use affects public support for cancer prevention policies, highlighting the regulations concerning alcohol, tobacco, and junk food ( Glasgow et al., 2022 ).

At the same time, social media not only informs individuals about cancer information ( Gage-Bouchard et al., 2018 ) but also emotionally engages its users ( Park et al., 2020 ), potentially heightening concerns about cancer risks ( Falzone et al., 2017 ) and influencing attitudes towards preventative policies ( Hsing et al., 2021 ). As a result, the present study introduces an innovative perspective by exploring cancer worry as a potential mediator in the relationship between social media use and support for cancer prevention policies.

The Theoretical Framework of S-O-R

The S-O-R (Stimulus-Organism-Response) model by Mehrabian and Russell (1974) provides a structured insight to explain how external stimuli affect emotional states and influence behaviors. In this framework, the stimulus (S) refers to external influences, such as psychological or environmental factors (e.g., social media use), that affect individuals. The organism (O) represents the internal emotional states these stimuli evoke, such as fear, anxiety, or worry. Finally, the response (R) reflects the response behavior to a certain stimulus, such as attitude related to cancer prevention policies.

The S-O-R framework has been widely applied across various disciplines and serves as a useful model for public health research. For example, scholars used the S-O-R framework to show that metacognitive beliefs and catastrophic misinterpretation have a strong positive influence on health anxiety ( Amin et al., 2022 ). Other researchers identify marketing communication channels and core values of organic food and explore how these factors influence consumers’ behavioral intentions toward organic food ( Sultan et al., 2021 ). Sitar-Taut and Mican (2023) proposed that fear and risk perception can mediate the relationship between social media exposure and attitudes toward vaccination.

In the current study, we apply the S-O-R framework to examine the mechanism behind social media use and support for cancer prevention policies. Specifically, the use of social media (S, stimulus) may affect individuals’ emotional responses toward cancer (O, organism), such as SMU as a conduit for health information, potentially amplifying concerns of cancer risks, which in turn could influence their decision to support cancer prevention policies (R, response). In other words, cancer-related worry mediates the connection between the stimulus (social media use) and the behavioral outcome (support for cancer prevention policies). This study employs a mediation analysis using data from the Health Information National Trends Survey ( HINTS 5 Cycle 4, 2020 ) to explore mediation mechanisms.

Hypotheses

Social Media Use (SMU) is the application of the Internet to connect with others through social networks such as Facebook, LinkedIn, Twitter, and other platforms ( Aichner & Jacob, 2015 ; Primack et al., 2017 ). Previous studies have shown a positive association between media-based information interventions and alcohol policies aimed at reducing alcohol-related harm ( Stewart & Casswell, 1993 ). Exposure to Pictorial Warning Labels can increase participants’ cancer worry and awareness of alcohol-related cancer risk, suggesting that policymakers could use this measure to restrict alcohol ( Ma, 2022 ). Thus, SMU may increase an individual’s support for alcohol control policies (SACP), which entails our first hypothesis.

H1. SMU is positively associated with SACP (d path).

Information loaded from social media amplifies feelings of worry ( Liu, 2020 ). In particular, cancer worry (CW) can be exacerbated through a specific avenue of online cancer information seeking ( Chae, 2015 ). Prior research has found that health-specific media exposure is positively related to CW ( Jung, 2014 ). Moreover, cancer worry motivates stakeholders to participate in or promote preventive actions, and policymakers may be motivated to act in cancer prevention in their communities ( Chen & Yang, 2017 ). In Europe, the worsening trend of liver cancer has led to alcohol-related policies ( Pimpin et al., 2018 ). Thus, the second hypothesis related to a positive pathway from SMU to upward CW and SACP is proposed:

H2. SMU had a positive indirect effect on SACP through CW. (ab path).

Previous research shows that social networks impact tobacco use policies ( Maddox et al., 2014 ), and one example revealed how social media facilitated the advancement of tobacco control ( Elmore et al., 2017 ). This leads to the third hypothesis, which includes the direct and remainder paths that SMU may affect SCPCT.

H3. SMU is positively associated with the SCTCP (d path).

Cancer worry could exert a positive effect on the perceived benefit of quitting smoking ( Lipkus & Prokhorov, 2007 ). Because health perceptions and beliefs influence relevant behaviors ( Yoo et al., 2018 ), cancer worry positively leads to health-protective actions ( McCaul et al., 2020 ; Paalosalo‐Harris & Skirton, 2017 ), specifically arousing public concerns and triggering proposals toward tobacco regulations ( Fallin & Glantz, 2015 ). Social media exposure is positively associated with CW, and CW has a positive impact on SCTCP, resulting in our fourth hypothesis:

H4. SMU has a positive indirect effect on SCTCP through CW (ab path).

As social media platforms become food advertising injection and marketing promotion hotspots ( Freeman et al., 2014 ), an increasing number of loopholes prompt the urgency for policies to both monitor and regulate ( da Silva et al., 2022 ; Jaichuen et al., 2019 ). A systematic review investigated the links between social media exposure and diet-related outcomes, underscoring the need for policy restrictions ( Mc Carthy et al., 2022 ). Therefore, we propose the following hypothesis:

H5. SMU is positively associated with SJCP (d path).

According to the hypotheses demonstrated above, SMU is positively associated with CW. Respondents who considered weight and diet conditions were strongly connected to cancer and supported junk food control policies ( Korn et al., 2021 ). Thus, we propose the following hypotheses:

H6. SMU had a positive indirect effect on SJCP via CW. (ab path).

H1, H3, and H5 together could entail H7:

H7. SMU is positively associated with support for cancer prevention policies (SCPP) (d path).

H2, H4, and H6 compound together entail H8:

H8. SMU had a positive indirect effect on the SCPP through CW. That is, SMU is positively correlated with CW, and further positively correlated with SCPP (ab path).

Methods

Data Source

The Health Information National Trends Survey (HINTS) is a survey conducted by the National Cancer Institute (NCI) that aims to collect nationally representative information on health information behaviors, attitudes, and knowledge among adults in the US. Data from HINTS 5, Cycle 4 (2020) were analyzed in the current study. Listwise deletion was employed for nonvalid responses in the regression analyses with SPSS software (v29). The dataset included responses from 3,865 participants, with a response rate of 37% ( Winston, 2021 ).

Measurement

Table 1 and Table 2 presents the key variables relevant to this study.

Table 1

Descriptions of Variables (n=3,865)

Note. SMU: social media use; CW: cancer worry; SCPP: support for cancer prevention policies; SACP: support for alcohol control policies; SCTCP: support for cigarette and tobacco control policies; SJCP: support for junk-food control policy.

Table 2

Percentage Frequency Descriptions of Items Used in Dependent Variables (n=3,865)

Main Variables

Support for alcohol control policies (SACP) ( Seidenberg et al., 2022 ) was measured by three questions: “To reduce the problems associated with excessive alcohol use, to what extent would you support or oppose... (1) banning outdoor alcohol advertising, (2) requiring health warnings on alcoholic beverage containers, and (3) requiring recommended drinking guidelines on alcoholic beverage containers.” The response options ranged from strongly oppose (= 1) to strongly support (= 5). A constructed variable was computed by averaging the three items (M = 3.66, SD = .93, Cronbach’s alpha = .83).

Support for cigarette and tobacco control policies (SCTCP) ( Heley et al., 2023 ) was measured by five questions, including two about cigarettes: “To what extent would you support or oppose the following measures related to cigarettes? 1) Just like with violence and sex, movies with cigarette smoking should be rated "R" to protect children and youth from seeing cigarette smoking in movies. 2) Cigarette packs should be required to have warning labels that use both images and words to show the negative health effects of smoking.” three focused on tobacco as “To what extent would you support or oppose the following measures related to all tobacco products, including cigarettes, e-cigarettes, smokeless tobacco, hookah, and cigars?1) Stores should be required to keep tobacco products out of customers’ view at the checkout counter. 2) Stores should be required to keep advertisements for tobacco products away from cash registers and out of windows. 3) Tobacco products should not be advertised on social media.” Response options ranged from strongly oppose (= 1) to strongly support (= 5). A composite variable was computed by averaging the five items (M = 3.77, SD = .96, Cronbach’s alpha = .89).

Support for junk food control policies (SJCP) ( Korn et al., 2021 ) was measured by a single question: “To what extent would you support or oppose the following? 1) Junk food products, including candy, chips, soda, and flavored sports drinks, should not be advertised to children on social media.” Response options ranged from strongly oppose (= 1) to strongly support (= 5).

Support for cancer prevention policies (SCPP) ( Glasgow et al., 2022 ) was the conceptual framework of nine questions related to alcohol, tobacco, and junk food. A composite variable was computed by averaging the nine items (M = 3.70, SD = .80, Cronbach’s alpha = .87).

Social media use (SMU) is defined as the connection between people using the internet through social networks such as Facebook or Twitter ( Chou et al., 2021 ) [ ]. The assessment of both general and health-specific SMU referred to responses to the following four questions: “In the past 12 months, have you used the Internet for any of the following reasons? 1) To visit a social networking site, such as Facebook or LinkedIn; 2) To share health information on social networking sites, such as Facebook or Twitter; 3) To participate in an online forum or support group for people with a similar health or medical issue; or 4) To watch a health-related video on YouTube.” If respondents selected the Internet as a means of social media approach, it was re-coded as 1 (and 0 otherwise). Using the mean of the dichotomous items directly, the measurement index SMU was created (M = .31, SD = .26).

Cancer worry (CW) ( Liu et al., 2023 ; Mamudu et al., 2024 ) measured the degree of worry about someone getting cancer with the question: “How worried are you about getting cancer?” A five-point scale was used, ranging from not at all (=1) to extremely (=5) (M =2.79, SD = 1.23).

Demographics including age, gender (1 = female, 0 = male), education (from 1 = less than 8 years to 7 = postgraduate), and political viewpoint (from 1 = very liberal to 7 = very conservative) were also controlled to eliminate potential confounding factors.

Statistical Analysis

In the mediation models, CW was the mediator, SMU was the independent variable, and SACP, SCTCP, SJCP, and SCPP were the dependent variables, respectively in each model. For interpretation and comparison between different variables, the percentage coefficient was adopted in this research (“bp” hereafter). Every variable is transformed into a 0-1 normalization, where 0 indicates the minimum conceptualization, and 1 indicates the maximum ( Zhao et al., 2024 ). Regression analysis adopted the OLS (Ordinary Least Squares) method to analyze the effect between variables. PROCESS V4.1 was applied in the calculation of mediation effects.

Results

Preliminary Analyses

Table 3 reports the outcomes of each path of the models. Regression models and percentage coefficients (bp) were adopted to provide clear indicators and comparisons of effect sizes across different variables. For example, bp = .071 (p<.001) indicated that women reported 7.1 percentage points more support for alcohol control policies than men.

Hypotheses Testing

H1 predicted a positive direct and remaining path of SMU on the SACP. As shown in Table 3, SMU was positively associated with SACP, and the statistical test passed the threshold (bp=.039, p<.05); thus, H1 was statistically supported.

H2 indicated that SMU had a positive indirect effect on SACP through CW (ab path). The results in Table 3 show that SMU was positively associated with CW (bp=.054, p<.05), and CW was positively associated with SACP (bp=.055, p<.001).

H3 depicted a positive direct and remainder path of the SMU on SCTCP. As shown in Table 3, SMU was positively associated with SCTCP (bp=.055, p<.01). Thus, H3 was statistically acknowledged.

H4 predicted that SMU would have a positive indirect effect on SCTCP through CW. The results in Table 3 show that SMU was positively associated with CW (bp=.054, p<.05), and CW was positively associated with SCTCP (bp=.080, p<.001); thus, H4 was supported.

H5 predicted a positive direct path and the remaining path of SMU on SJCP. Because the statistical test did not pass the threshold (bp=.027, p>.05), H7 was not statistically acknowledged.

H6 hypothesized that SMU has a positive indirect effect on SJCP through CW. The results in Table 3 show that SMU was positively associated with CW (bp=.054, p<.05), and CW was positively associated with SJCP (bp=.035, p<.05).

H7 illustrated the positive direct and remainder paths of the SMU on the SCPP. As shown in Table 3, SMU was positively associated with SCPP, and the statistical test passed the threshold (bp=.046, p<.01). Thus, H7 was statistically acknowledged.

H8 predicted that SMU had a positive indirect effect on SCPP through CW (ab path). The results in Table 3 show that SMU was positively associated with CW (bp=.054, p<.05), and CW was positively associated with the SCPP (bp=.065, p<.001); therefore, H8 was supported.

Analysis of Mediation Models

The effect sizes of the mediation models for support for alcohol control policies (SACP), support for cigarettes and tobacco control policies (SCTCP), support for junk food control policy (SJCP), and support for cancer prevention policies (SCPP) were assessed.

In the alcohol regulation model, the indirect path SMU→CW→SACP pointed in the same direction and contributed 6.8% of the total effect (bp=.003, CI=.000, .006; see Figure 1), constructing a complementary mediation ( Zhao et al., 2010 ).

In the tobacco restriction model, the SMU→CW→SCTCP path pointed in the same direction and contributed 7.3% of the total effect (bp=.004, CI=.001, .009; see Figure 2), also constituting complementary mediation.

However, in the junk food control model, the SMU→CW→SJCP path pointed in the same direction and contributed approximately 7.0% of the total effect (bp=.002, CI=.000, .005; see Figure 3). The direct-and-remainder path SMU→SJCP was not significant and set up an indirect-only mediation ( Zhao et al., 2010 ).

As a total model of the cancer prevention policies, the indirect path SMU→CW→SCPP pointed in the same direction and contributed approximately 9.8% of the effect (bp=.004, CI=.001, .007; see Figure 4), and complementary mediation was constructed.

Table 3

OLS Regression Results of SMU→SACP/SCTCP/SJCP/SCPP Mediated by CW