The effect of message framing and the presentation of health vs. social consequences on health risk perception

With health prevention campaigns health insurance providers can increase the perception of health risks among customers in order to promote healthy behaviour. The presentation format of health information has a significant influence on the health-related behaviour of customers. We investigate in our experimental study the impact of positive vs. negative presentation (message framing) as well as the presentation of health vs. social consequences in health prevention campaigns on health risk perception. The results reveal that loss- (vs. gain-) framed messages and the presentation of health (vs. social) consequences increase health risk perception more strongly. This effect is mediated by a stronger perceived severity of health (vs. social) consequences. Women perceive health (vs. social) consequences more likely and closer in time than men. Our findings show the importance of increasing health risk perception in order to promote healthy behaviour intentions.


Introduction
Health insurance providers often use prevention campaigns to inform customers about different kinds of health risks and thereby promote healthy behaviour. Effective health communication should prevent negative changes and increase positive changes in health behaviour (Fife-Schaw and Abraham 2009). Various scholars have underlined the positive relationship between health risk perceptions and behavioural intentions for different health risks (Brewer et al. 2007;Chandran and Menon 2004;Courbage et al. 2017;Heideker and Steul-Fischer 2017;Menon et al. 2008;Murdock and Rajagopal 2017;Samper and Schwartz 2013). In this respect, it is the focus of our study to investigate methods in health communication in order to increase health risk perception and healthy behaviour intentions.
Publishers of health prevention campaigns can highlight either the health or the social consequences of healthy or unhealthy behaviour. Whereas the presentation of health consequences aims to convince recipients of the consequences a certain behaviour has on their physical health, the presentation of social consequences is intended to make recipients aware of the cosmetic consequences or the consequences of social exclusion due to a certain behaviour. Previous research mainly focused on the presentation of negative health consequences in health communication, and seldom analyzed the effects of the presentation of social consequences on health risk perception. Scholars further neglected the effects of message framing for different types of consequences (Murdock and Rajagopal 2017). Framing is the presentation of objective equivalent information in semantically different ways. The use of framing in health communication evidently creates differences in health risk perception among recipients (Chandran and Menon 2004;Menon et al. 2008;Peters et al. 2011). Message framing implies the presentation of the same information in two respective ways: gains or losses. Either option influences an individual's attitudinal and behavioural responses differently (Park 2012).
The goal of this paper is to analyze the effect of message framing on health risk perception when either health or social consequences of a certain health outcome are presented. Therefore, we aim to answer the research questions of how message framing (loss vs. gain) and the presentation of different types of consequences (social vs. health) in health prevention campaigns have an impact on health risk perception, and consequently, influence health behaviour intention. We conducted an online experimental study to investigate these effects. Due to that, this paper contributes to the marketing literature and practice by providing empirically based insights regarding the influence of message framing and the presentation of consequence types on health risk perception and health behaviour intention.
This paper is structured as follows. Based on the literature on health risk perception, message framing and the presentation of different types of consequences in health communication, we derive our hypotheses and test these hypotheses within an experimental study. After presenting the results, we discuss the theoretical and practical implications of our findings and suggest avenues for future research.

Theoretical background 2.1 Health risk perception
Health risk perception can be defined as the perception of the subjective probability of the occurrence of a negative health-related event for a person or a group of persons for a specific period of time (Menon et al. 2008). Health prevention campaigns aim to inform about health risks and influence the knowledge, attitudes, and actions of individuals and the public. By investigating and applying communication strategies, a positive influence on health risk perception and the respective behaviour of individuals can be achieved (Rothman et al. 2006). For example, health insurance companies use such campaigns to make aware of health risks, and therefore profit from healthier behaviour of their insured persons (Heideker and Steul-Fischer 2016). Thus, knowledge of the influencing factors of health behaviour is essential.
Consistent with various psychological theories, health risk perception influences health behaviour intention as well as actual behaviour (Brewer et al. 2007). One theory explaining health information processing is the Health Belief Model according to Rosenstock (1974). The model includes perceived vulnerability, severity, benefits, barriers, and cues to action. It assumes that individuals perceive a health risk when they face an outcome that contains serious consequences, and when they feel vulnerable to that outcome. The Health Belief Model further suggests that individuals outweigh the perceived benefits and costs of a certain behaviour before acting in that way. As soon as individuals perceive a serious and vulnerable threat to their health, and their perceived benefits of taking health action are greater than the perceived costs, they are most likely to undertake the recommended preventive health behaviour (Park 2012;Rosenstock 1974).
Similar to the Health Belief Model, the Protection Motivation Theory according to Rogers (1975) investigates health information processing and aims to examine the intention of individuals to maintain their own health status. This intention is a function of the amount of protective motivation aroused by two cognitive appraisal processes: threat assessment and coping assessment. Both processes are subject to a cost-benefit analysis, and can be influenced by external (e.g., fear appeals) and internal factors (e.g., previous experiences). The threat assessment weighs the perceived severity and vulnerability of an outcome (e.g., obesity) with the extrinsic (e.g., social factor of eating with family) and intrinsic reward (e.g., pleasure of eating) of unhealthy behaviour. On the other hand, the coping assessment is the balancing of the self-efficacy and effectiveness of action of an individual and the perceived costs of changing behaviour. The two cognitive appraisal processes finally influence the health behaviour intention (Maddux and Rogers 1983;Rogers 1975).
In line with the Health Belief Model and the Protection Motivation Theory, health risk perception is determined by two main components: perceived vulnerability and perceived severity. Consequently, health prevention campaigns can increase individuals' health risk perception if the consequences of health risks are considered particularly vulnerable and a high degree of severity is attributed to them (Murdock and Rajagopal 2017;Yan and Sengupta 2012). Therefore, adequate health communication strategies should be examined in order to design health messages that seem particularly vulnerable and serious, and consequently influence consumers' health behaviour intention. In this context, we first examine the theoretical background of message framing effects.

Message framing effects
Message framing effects occur when a persuasive message results in different perceptions depending on whether it highlights the positive outcome of performing a behaviour to achieve a particular goal, or the negative outcome of not performing the behaviour (Levin et al. 2002). In the present context, message framing is a format of presentation that aims to promote health behaviour among recipients through objectively equivalent health information, which is framed in terms of either gains or losses. In case of a positive presentation or description of a message (gain-frame), the advantages of engaging in a health behaviour are stressed. Otherwise, a negative presentation or description of a message (loss-frame) emphasizes the disadvantages of not engaging in a health behaviour (Bassett-Gunter et al. 2013;Gerend and Maner 2011). Drawing on Prospect Theory according to Kahneman and Tversky (1979), the effectiveness of gain-vs. loss-framed messages depend upon whether the proposed behaviour is perceived to implicate risk. Individuals act according to their perception of risk, which is associated with the outcome of performing a certain behaviour.
Studies that deal with the influence of message framing on health risk perception as a predictor of health behaviour intention are rare, whereas the direct effect of message framing on health behaviour intention has been often investigated by previous work (Gallagher and Updegraff 2012;Heideker and Steul-Fischer 2017;Rothman and Salovey 1997). Various scholars have tested the effect of message framing on health behaviour intention within the context of different health risks and unhealthy behaviour. Some examples include message framing and vaccinations (Abhyankar et al. 2008;Gerend and Shepherd 2007;Wirtz et al. 2015), cancer K , physical activity (Bassett-Gunter et al. 2013;Latimer et al. 2008), healthy eating (Gerend and Maner 2011;Wirtz and Kulpavaropas 2014), smoking (Jung and Villegas 2011), alcohol consumption (Baek et al. 2013;Gerend and Cullen 2008), and sun safety behaviour (Hwang et al. 2012). Those studies have observed contradicting results regarding the effect of message framing on health behaviour intention. Whereas some scholars postulate a stronger effect of loss-framed messages (Abhyankar et al. 2008;Bassett-Gunter et al. 2013;Dillard et al. 2012;Gerend and Shepherd 2007;Wirtz and Kulpavaropas 2014;Wirtz et al. 2015), others detected higher health behaviour intentions after presenting gain-framed messages (Baek et al. 2013;Gallagher and Updegraff 2012;Gerend and Cullen 2008;Latimer et al. 2008).
Overall, loss-framed messages are thought to be more effective than gain-framed messages when targeting health behaviour with high-risk outcomes and behaviour that detects the presence of a disease (e.g., HIV testing, mammography). By contrast, gain-framed messages are supposed to be more effective than loss-framed messages when targeting health behaviour with low-risk outcomes and behaviour that prevents the onset of a disease (e.g., physical activity) (Bassett-Gunter et al. 2013;Gallagher and Updegraff 2012;Rothman et al. 2006). Regarding this general hypothesis, health prevention campaigns should use gain-framed messages because they are usually emphasizing disease prevention behaviour (e.g., sun safety behaviour to prevent skin cancer). However, the meta-analytic review from O'Keefe and Wu (2012) contradicts this hypothesis for promoting skin cancer prevention. The authors did not report a significant difference in the persuasiveness of gain-and loss-framed appeals for encouraging sun safety behaviour in order to promote skin cancer prevention.
Because of those contradicting results in previous research, and the lack of literature on health risk perception, we propose, in line with Prospect Theory (Kahneman and Tversky 1979), that a loss-framed description of the outcomes of health risks results in a higher health risk perception than a gain-framed description. The value function of Prospect Theory implies that gains and losses are measured relative to a reference point rather than the perception of absolute values. Because of loss aversion, consumers are more sensitive to losses than to gains. Bassett-Gunter et al. (2013) add in their work that loss-framing of health information leads to a higher health risk perception than gain-framing because cognitive processing is activated more strongly.

Presentation of health and social consequences
In addition to the appropriate use of gain-or loss-framed messages, the question remains as to what information should be included in health communication in order to present health risks in a more vulnerable and serious way. Besides solely mentioning the health consequences of a healthy or unhealthy behaviour, health prevention campaigns can also refer to social consequences (Keller and Lehmann 2008;Murdock and Rajagopal 2017). The presentation of health consequences as a result of a certain health behaviour shown in health communication aims to convince recipients of the health outcomes of a health risk (e.g., "Smoking causes 9 out of 10 lung carcinomas"). Whereas the presentation of social consequences is intended to draw the recipient's attention to appearance-related outcomes or the consequences of social exclusion due to a certain health behaviour (e.g., "Smoking causes your skin to age"). Most literature on health communication focuses on the effectiveness of health consequences (e.g., skin cancer) in health messages to provoke behavioural changes instead of investigating social and appearance related consequences (e.g., wrinkles). However, recent studies have shown that the presentation of social consequences also plays an important role for the effectiveness of health communication (Cornelis et al. 2014;Martin and Kamins 2019;Murdock and Rajagopal 2017).
According to Murdock and Rajagopal (2017), the presentation of an outcome with social consequences is perceived to be closer in time and more likely, but not less severely than an outcome with health consequences. This, in turn, leads to a higher behavioural intention, attitude and response assessment. Cornelis et al. (2014) further revealed that appearance-related arguments in health communication are more effective than health-related arguments when using two-sided messages instead of one-sided messages. In addition, the presentation of health consequences is more effective for health-motivated individuals, whereas the presentation of social consequences is more effective for appearance-motivated individuals. Martin and Kamins (2019) were able to show that health messages are more effective for behavioural changes when focusing on the impact that one's death has on family and friends left behind rather than focusing on the physical death of the individual.
The Construal Level Theory, according to Trope and Liberman (2010), provides an explanation for the effectiveness of emphasizing social consequences compared to health consequences. The theory describes the context between psychological distance and mental abstraction. A high psychological distance (e.g., temporal distance) is accompanied by a high mental abstraction and vice versa. Consequently, a high temporal distance, which is often the case for health consequences, leads to a high degree of mental abstraction of the event. Because of that high temporal distance of the event, consumers believe that the onset of the health consequence is less likely, and that they have plenty of time to change their behaviour and thereby avoid the consequence (Chandran and Menon 2004). As a result, health messages should aim to reduce the perceived temporal distance between the provoked behaviour and its consequence on the recipient in order to increase the perceived likelihood and riskiness of the event. Thus, consumers will have a higher willingness to comply with the targeted behaviour. Highlighting social consequences in health messages can reduce the perceived temporal distance because social consequences are perceived as temporally closer and thus less abstract (Murdock and Rajagopal 2017;Trope and Liberman 2010). The increased proximity will make these events seem more likely to occur and more vulnerable to the individual. Although health consequences are intuitively viewed as more severe than social consequences, the less perceived temporal distance of social consequences contributes to an increased perceived severity. Thus, overall message effectiveness will be enhanced by the presentation of social consequences because they lead to a higher health risk perception, and subsequently to a higher health behaviour intention among individuals (Murdock and Rajagopal 2017;Rogers 1975;Rosenstock 1974;Trope and Liberman 2010).
K Based on the theoretical consideration of health risk perception, message framing effects, and the presentation of health and social consequences, we propose the following hypotheses:

H1
The health risk perception is higher for loss-framed compared to gain-framed messages of social consequences.

H2
The health risk perception is higher for loss-framed compared to gain-framed messages of health consequences.

H3
The health risk perception is higher for the presentation of social compared to health consequences.

Experimental study
The goal of our study is to examine whether gain-or loss-framed messages increase health risk perception when either health or social consequences are presented. Thus, H1, H2 and H3 are examined in the following study.

Method
A total of 211 participants (62.6% female, mean age = 24.3 years) participated in our study. They were randomly assigned to one of four conditions of a 2 (message framing: loss vs. gain) × 2 (consequence type of health risk: social vs. health consequence) between-subject design. Our scenario text was supposed to be a health prevention campaign that emphasizes the riskiness of ultraviolet radiation. In the loss-framed message treatments, the participants were informed about the disadvantages of unprotected exposure to ultraviolet radiation. However, participants in the gain-framed message treatments were educated about the advantages of protected exposure to ultraviolet radiation. Additionally, participants treated with social consequences read the health prevention campaign text, which pointed out the social consequences (e.g., reduced vs. increased self-confidence due to premature skin aging/beautiful skin) of the respective behaviour, whereas the participants treated with health consequences were informed about the consequences of ultraviolet radiation to their health (e.g., increased vs. reduced risk of skin cancer).
Measures Participants' perceived level of health risk perception was measured using a 10-point scale (1 = no risk to health, 10 = very high risk to health) (Heideker and Steul-Fischer 2016). In accordance with Menon et al. (2008), the participants were asked to answer questions for health behaviour intention to check their perceived need to care about ultraviolet radiation. It included questions regarding the "increase of awareness for ultraviolet radiation", "increase of interest in the consequences of ultraviolet radiation", "review of one's own behaviour towards ultraviolet radiation", "need to change one's own behaviour regarding ultraviolet radiation" and "word-ofmouth of information about ultraviolet radiation" measured on 7-point rating scales (1 = totally disagree, 7 = totally agree). In addition, the covariates "perceived vulnerability to ultraviolet radiation" and "perceived severity of ultraviolet radiation" were surveyed on 7-point rating scales (1 = totally disagree, 7 = totally agree) (Greene et al. 1996). Besides that, the variables "perceived probability" (1 = very unlikely, 7 = very likely) and the "perceived temporal distance" (1 = very close, 7 = very distant) of the occurrence of health and social consequences as a result of ultraviolet radiation were also taken into consideration (Murdock and Rajagopal 2017). Lastly, we measured the participants' knowledge about health risks in general and about ultraviolet radiation in particular on a 7-point scale (1 = very low, 7 = very high), as well as their attitude towards sunscreens by using a semantic differential scale (Heideker and Steul-Fischer 2016;Murdock and Rajagopal 2017).

Results
Manipulation check According to Gerend and Maner (2011), we asked two questions using a 7-point scale, whether the health prevention campaign emphasized the negative outcomes of unprotected vs. the positive outcomes of protected exposure to ultraviolet radiation, and whether the text highlighted the social vs. health consequences of exposure to ultraviolet radiation. As expected, participants treated with the loss-framed message agreed significantly differently that the text highlighted the negative outcomes of unprotected ultraviolet radiation compared to participants treated with the gain-framed message (Mloss = 1.36 and Mgain = 4.61; t(134, 185) = -17.648; p = 0.000). Furthermore, participants in the social consequences treatment indicated significantly differently that the text underlined the social consequences of ultraviolet radiation compared to the participants in the health consequences treatment (Msocial = 2.71 and Mhealth = 6.05; t(195, 529) = -19.071; p = 0.000). Consequently, the manipulations were successful.
Message framing and consequence type The health risk perception is significantly higher for loss-framed messages compared to gain-framed messages (Mloss = 6.93 and Mpositive = 6.21; p = 0.015). Additionally, health risk perceptions are significantly higher for the presentation of health consequences compared to social consequences in case of the health risk of ultraviolet radiation (Msocial = 6.27 and Mhealth = 6.86; p = 0.044). An ANOVA underlines the significant influence of message framing (F(1,206) = 7.632; p = 0.006) and consequence type (F(1,206) = 5.650; p = 0.018) on health risk perception. There is no significant interaction effect between the message framing and the consequence type (F(1,206) = 0.096, p > 0.10) (see Fig. 1). Based on this analysis, hypothesis H1 and H2 can be supported. Lossframed messages increase health risk perception more strongly than gain-framed messages, which holds true for the presentation of social consequences, as well as for health consequences. By contrast, our analysis indicates no support for hypothesis H3. The presentation of social consequences in health prevention campaigns does not increase health risk perception significantly more than the presentation of health consequences, as detected by Murdock and Rajagopal (2017).
Contrary to Murdock and Rajagopal (2017), our sample predominantly consisted of women. Our results reveal that women judge the health (vs. social) conse- Fig. 1 The influence of "message framing" and "consequence type" on health risk perception quences to be significantly more likely than do men (Mwomen = 3.75 and Mmen = 3.09; t(209) = -3.425; p = 0.001). In addition, women assess the occurrence of health consequences to be closer in time than did men (Mwomen = 4.42 and Mmen = 4.99; t(209) = 2.517; p = 0.013). Furthermore, three mediation analyses reveal explanations for the detected effect of health (vs. social) consequences on health risk perception (see Fig. 2). The results show that the consequence type had a significant indirect effect on health risk perception through perceived severity (estimated coefficient = 0.59; CI excludes zero [0.02, 0.31]). Thus, perceived severity mediates the effect of consequence type on health risk perception. In addition, the effect of perceived severity on health risk perception is significantly mediated through the perceived temporal distance (estimated coefficient = 0.59; CI excludes zero [0.01, 0.23]). Lastly, the effect of perceived temporal distance on health risk perception is significantly mediated through participants' knowledge about health risks in general, and about ultraviolet radiation (estimated coefficient = -0.60; CI excludes zero [-0.15, -0.01]). Health behaviour intentions A regression analysis shows an overall significantly higher health behaviour intention as a result of higher health risk perception (β = 0.134, t = 3.110, p = 0.002). Nonetheless, the total mean of health behaviour intention is relatively low (Mhealth_behaviour_intention= 3.44, SD = 1.350). The following health behaviour intention items indicate a significant influence of health risk perception: increase of awareness for ultraviolet radiation (β = 0.187, t = 3.389, p = 0.001) and review of one's own behaviour towards ultraviolet radiation (β = 0.186, t = 3.460, p = 0.001). Nevertheless, the means of these health behaviour intentions are quite low (see Table 1 with 1 = totally disagree, 7 = totally agree).

General discussion
Health prevention campaigns are often used by health insurance companies to inform their customers about health risks. Our study offers important insights on how those health messages should be designed in order to increase health risk perception, and consequently health behaviour intention among individuals. In our study, we examined the effects of message framing and the different presentation of consequence types on health risk perception. Besides analyzing whether K gain-or loss-framing in health messages increases health risk perception, we explored whether the presentation of health or social consequences in health messages increases health risk perception. We further examined the impact of health risk perception on health behaviour intention. Our findings reveal, in accordance to Prospect Theory (Kahneman and Tversky 1979), that loss-framed messages lead to significantly higher health risk perceptions than gain-framed messages. Furthermore, the presentation of health consequences contributes to significantly higher health risk perceptions than the presentation of social consequences. The impact of higher health risk perception due to the presentation of health (vs. social) consequences may depend on the participants' more strongly perceived severity for health outcomes, higher perceived temporal distance of social consequences and their overall high degree of knowledge about health risks and ultraviolet radiation. We observed a highly significant effect of perceived vulnerability and severity of ultraviolet radiation on health risk perception. Thus, our results are in line with the Health Belief Model according to Rosenstock (1974), and the Protection Motivation Theory according to Rogers (1975). Health behaviour intentions increase significantly with higher health risk perception, even though the intentions are quite low. Women show a significantly higher health risk perception in general, as well as a higher perceived likelihood and lower perceived temporal distance of the occurrence of health (vs. social) consequences compared to men. Our results present theoretical and managerial implications that are important to any publishers of health information (e.g., health insurance companies).

Managerial implications
Health insurance companies can derive practical recommendations for action for the design of health prevention campaigns regarding sun safety behavior. Our empirical evidence suggests the use of loss-framed messages and the presentation of health consequences in health prevention campaigns to increase health risk perception and to promote healthy behaviour. Gender differences also should be taken into account because men and women differ in health risk perception. In particular, men need to be persuaded to increase their knowledge of ultraviolet radiation, and consequently their attitude towards sunscreen products. Due to the cost-reducing effect of behavioural changes, the strategic design of health messages, which induce vulnerable and serious health risk perceptions, becomes worthwhile for insurance companies looking to reduce health care spending.
Limitations and future research directions Even though our study is not without its limitations, it is noteworthy that variables such as health consciousness, personal relevance, attitudes, motivation, and others can influence the health risk perception of individuals. We cannot provide implications for actual or future health behaviour because we measured health behaviour intention. Further studies should examine the reasons for low intentions to change one's own health behaviour. They should investigate why social consequences are perceived to be more distant in time than are health consequences and how this effect can be inverted (e.g., testing the effectiveness of temporal framing). In summary, our variables seem to be suitable for investigating health risk perception, but not exhaustively. We propose to replicate these results for other health risks and to focus on other consequences and other ways of presenting messages in order to increase health risk perception. The mentioned issues and variables should be considered in future research. Further studies could reduce the problems in designing health communication because of their presentation, and could support the research area of the influence of different types of consequences on health risk perception (Murdock and Rajagopal 2017).
Funding Open Access funding enabled and organized by Projekt DEAL.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4. 0/.