Disgusting microbes: The effect of disgust on perceptions of risks related to modifying microbiomes

Abstract

Research on perceived risks of scientific issues has largely overlooked the influence of disgust as a predictor. Here, we examine the impact of disgust on perceived risks of modifying microbiomes using a 2 (emotion) × 2 (focus) experiment embedded in an online survey. We find evidence of moderated mediation where individuals exposed to an article about microbiome research and therapies with explicit references to disgusting stimuli perceived greater risk through a mediating variable, elicited disgust. This indirect effect is moderated by the focus of the article; those who viewed a human-focused article experienced greater disgust and reported greater perceived risks. These findings have implications for assessing and addressing lay audiences’ reactions to an emerging issue that has significant societal implications.

Keywords

disgust microbiomes moderated mediation risk perceptions

Research on microbiomes has diverse implications, ranging from human health and food production to environmental sustainability. The term “microbiome” describes collections of microorganisms in natural and human-made habitats (Marchesi and Ravel, 2015) and refers to a catalog of the genes of microorganism communities as well as the microbes themselves. Recent improvements in DNA sequencing technologies have enabled scientists to better understand (Dsouza and Gilbert, 2017) and observe large microbial communities across time in a variety of environments (e.g. the Human Microbiome Project, the Earth Microbiome Project, and the Tara Oceans Project). Although much scholarship has focused on the role of the human gut microbiome in physical and mental health (e.g. Cohen, 2016; Dinan and Cryan, 2017), microbiome research extends beyond the human body (Blaser et al., 2016). For example, research has examined how microbes can be used for environmental bioremediation (Van Rossum et al., 2016) and in bio-based economies (Singh and Trivedi, 2017).

As scholarship in this area advances and technologies and potential uses develop (e.g. Franzosa et al., 2015), it is imperative to proactively address the ethical, legal, and social implications associated with microbiome science and technologies (Alivisatos et al., 2015). To do so, we must first understand public perceptions of this emerging issue. The risks associated with microbiomes range from regulatory (e.g. screening stool for fecal microbiota transplants (FMT)) to social justice concerns (e.g. ownership related to human microbiomes), and public perceptions can have significant implications for governance and regulation of microbiome therapies. Here, we focus on risk perceptions related to the modification of microbiomes as one of the more commonly known microbial therapies involves transplanting feces from a donor to treat severe Clostridium difficile infections (FMT). In doing so, we test causal hypotheses about the effects of a discrete emotion, disgust, on perceived risks of microbiome modification.

Our goals for this study are two-fold: (1) to grow the literature on media effects, risk perception, and emotions using a discrete emotion approach and focusing on the relatively overlooked role of disgust and (2) to examine the impact of disgust on perceived risks. We employ an experiment with a 2 (emotion) × 2 (focus) between participants design embedded in an online survey to test our hypotheses. In our literature review, we first conceptualize and explicate disgust. Then, we consider studies examining the impact of disgust on various outcomes, including information seeking and processing, attitudes, behavioral intentions, and risk perceptions.

1. Literature review

The emotion of disgust

Emotions are feeling states that can influence thought, learning, judgment, and decision-making (Bechara et al., 2000; Lerner et al., 2015). Despite a relatively long history of scholarship, theorists still argue about which emotions are basic (Ekman, 1999; Izard, 2010; Panksepp and Watt, 2011; Plutchik, 2001). With some exceptions (Panksepp and Watt, 2011), disgust, along with anger, fear, sadness, and enjoyment, is typically considered a basic emotion (Ekman, 1999; Izard, 1977; Sauter et al., 2010; Vytal and Hamann, 2010), though it has been relatively overlooked in media effects scholarship.

In this study, we define disgust as a response to the perception of any stimulus that is noxious, revolting, repulsive, offensive, or degrading (Nabi, 2002; Rozin et al., 2016; Tybur et al., 2013). Individuals who experience disgust are repelled by the stimulus (Rozin et al., 2016). This motivation to turn away from a disgusting object is believed to protect the integrity of the body and, in the case of repulsive behaviors, social order (Rozin et al., 2016).

There is an extensive literature debating the evolutionary origins of disgust (for a detailed discussion, see Rozin et al., 2016; Tybur et al., 2013), but scholars agree that disgust is an emotion that has evolved in part through the selection pressure of pathogens (Curtis et al., 2011; Oaten et al., 2009). The debate over its evolutionary origins extends to the typology of disgust. Theorists generally agree that there is a distinct category for sociomoral disgust, which is elicited by moral transgressions (Rozin et al., 2016; Tybur et al., 2013), but propose different types of disgust as responses to elicitors such as body products, animals, contact with strangers, and hygiene. In the traditional model of disgust (Rozin et al., 2016), food and animals elicit core disgust, while contact with strangers and hygiene evoke interpersonal and animal-nature disgust, respectively. In the functional model of disgust (Tybur et al., 2013), pathogen disgust is an adaptation that “serves as an initial line of defense against pathogen infection” (p. 68), and food, animals, contact with strangers, and hygiene are common elicitors.

Microbiomes have the potential to elicit multiple types of disgust. In the traditional model, core disgust is elicited by a sense of contamination, typically related to food, for which microbes are often to blame. Animal-nature disgust is elicited by stimuli that remind us of our animal nature. This encompasses the maintenance of hygiene, the lack of which is often associated with microorganisms. Microbes can also elicit interpersonal disgust through the idea of crossing bodily boundaries (e.g. through disease transmission or modification of one’s microbiome via FMT to treat Clostridium difficile infections). Finally, microbiome research and therapies have the potential to evoke sociomoral disgust through ethical concerns such as privacy risks associated with human microbiomes (Callaway, 2015). In the functional model, food contamination, the lack of hygiene, transmission of disease, and FMT can evoke pathogen disgust.

Media coverage has reinforced the idea that microorganisms are disgusting (e.g. Britlab, 2015). For example, coverage of FMT in response to gastrointestinal infections in Canadian media has been found to emphasize the “ick” factor (Chuong et al., 2015). Moreover, studies show that once news stories trigger an emotion, future assessments of the information are colored by that reaction (Kühne and Schemer, 2015; Nabi, 1999). Thus, if individuals are exposed to information about microbiomes that they find disgusting, their future evaluations of and attitudes toward microbiome research and technologies are likely to be influenced by disgust. Thus, we designed our experiment to elicit a message-referent emotional response—one that is anchored in images evoked by the content of the message (Dunlop et al., 2008)—as disgust is experienced more easily when mental images are made accessible to individuals (Rozin et al., 2016).

Our experimental manipulation involves altering the language in textual information related to microbiome research and FMT. In one condition, we use language with explicit references to evoke vivid mental imagery (e.g. “fecal transplants”); in the other, we employ neutral language (e.g. “transplant therapies”). We refer to these manipulation conditions as the disgust and no disgust conditions. Given the evolutionary origins of disgust as a mechanism by which humans avoid pathogens and disease (Curtis et al., 2011; Oaten et al., 2009), we expect information about microbiome research that focuses on humans is more relatable and, thus, more likely to generate visceral reactions and mental images among viewers. This leads us to predict that individuals exposed to the disgust condition will experience the emotion more intensely relative to those in the no disgust condition. Moreover, when the information focuses on humans, it more acutely reminds us of the primal nature of our physical bodies, and we propose that this amplifies the effect of the disgust condition on experienced emotion:

H1: The effect of the disgust condition on elicited disgust will be stronger when the information focuses on humans.

Disgust as an antecedent variable

Emotion and affect have a long history of scholarship in communication (e.g. Cantril and Allport, 1935; Hovland et al., 1953). Typically, research on the effects of emotion takes one of two theoretical approaches—the dimensional perspective that characterizes emotion on a single affective continuum ranging from positive to negative (Watson and Tellegen, 1985) or the discrete perspective, which identifies specific emotions such as anger or fear (Frijda, 1986). Here, we take the discrete approach as it is the more promising perspective for media effects studies (Dillard and Peck, 2001): The “affect structure suggested by discrete emotions theories is necessary to understand the effect of affect” (p. 64). This does not imply that the dimensional approach is not useful for media effects studies of emotion. Indeed, many studies in communication have employed a dimensional approach (e.g. Keller et al., 2012; Lee and Scheufele, 2006) and the two perspectives should be viewed as complementary (Dillard and Peck, 2001).

Much of the work on the effect of affect focuses on information seeking or processing outcomes (e.g. Nabi, 1999) and disgust generally receives less scholarly attention than fear or anger, though it can be the primary emotion evoked by a message (Nabi, 1998). A fair amount of scholarship on the role of disgust in message processing has been conducted in the context of anti-tobacco and smoking advertisements (Clayton et al., 2017; Leshner et al., 2009; Netemeyer et al., 2016). In a study of smokers who participated in a 2 (presence/absence of smoking cues) × 2 (low/high ratings of disgust) experiment (Clayton et al., 2017), those who saw advertisements with smoking cues and high disgust images were more likely to engage in defensive information processing, a finding that is consistent with other research (Leshner et al., 2009). Disgust has also been examined as an emotion that motivates avoidance (Newhagen, 1998). Viewers, when asked to move computer paddles toward or away from the television screen while watching news images designed to elicit one of three emotions, tended to approach images designed to evoke anger yet moved away from scary or disgusting images. In entertainment media, one study examined how core and sociomoral disgust affected information processing, finding that core, relative to sociomoral, disgust elicited more heuristic processing (Rubenking and Lang, 2014).

Disgust has also been shown to influence people’s attitudes and behavioral intentions as both a mediating and moderating variable. As a mediator, disgust evoked by persuasive messages decreases support for animal experimentation (Nabi, 1998) and lowers intentions to patronize a fast-food restaurant (Shimp and Stuart, 2004). As a moderator, disgust enhances the effect of a persuasive message (Clayton et al., 2017; Morales et al., 2012). In a series of studies, Morales et al. (2012) compared the persuasive effects of disgust-inducing fear appeals and fear appeals alone. In one study, they used visual images in their experiment; in another, experimental conditions were presented as text. In both studies, disgust-inducing fear appeals relative to fear appeals alone led to more effective persuasion (e.g. greater intention to avoid and engage in negative (methamphetamine use) and positive behaviors (sunscreen use), respectively). Moreover, disgusting fear appeals led individuals to respond to messages more quickly compared with fear appeals alone.

Existing evidence (e.g. Finucane et al., 2000) leads us to believe that experienced disgust can influence risk perceptions. Although these studies often employ discrete indicators of emotion that include disgust, these indicators are often combined and modeled as overall negative affect. Our study is thus motivated, in part, by the paucity of scholarship on risk perception that takes a discrete emotion perspective. Some of the earliest empirical evidence taking a dimensional emotion approach can be found in Slovic et al.’s (1991) work on nuclear waste. Americans primarily had negative associations with a nuclear waste repository and these associations were correlated with attitudes and perceived risks. Subsequent research confirms this finding in a variety of issue contexts; negative affect is generally associated with greater perceived risks (Alhakami and Slovic, 1994; Finucane et al., 2000; Mou and Lin, 2014). As perceptions of risk are inherently subjective, emotion can serve as a way of orienting oneself when faced with novel scientific information (Peters and Slovic, 1996). Emotion “directs fundamental psychological processes such as attention, memory, and information processing” (Slovic, 1999: 694) that helps us navigate the world efficiently when faced with uncertain and new information.

Taking a discrete approach, the appraisal tendency framework serves as a useful theoretical foundation for understanding how emotions might influence risk perceptions. This framework has two underlying assumptions (Lerner and Keltner, 2000): (1) Emotions cause changes in cognition that are often persistent and guide subsequent behavior and (2) emotions are associated with specific appraisals of an event or object. Thus, emotions elicited by an event serve as predispositions that guide evaluation of future events. Using this framework, scholars have found that fear induced risk-aversion while anger produced the opposite effect (Lerner and Keltner, 2001). Anger motivates people to approach causative stimuli; fear and disgust both instigate avoidance responses (Newhagen, 1998). Here, we draw on the appraisal tendency framework and expect that disgust, a negative discrete emotion, will increase risk perceptions:

H2: Elicited disgust will positively predict perceived risks of modifying microbiomes.

Together, H1 and H2 posit a moderated mediation model in which the language manipulation impacts perceived risks through experienced or elicited disgust, the effect of which is moderated by the human focus of the information (Figure 1):

H3: The focus manipulation will moderate the effects of the language manipulation on perceived risks through elicited disgust, that is, there will be a moderated mediation effect.

Figure 1.

Conceptual (a) and statistical (b) path models in which the effect of the language manipulation (disgust vs no disgust) on perceived risks is moderated by the focus manipulation (human vs non-human). Elicited disgust serves as a mediator between the experimental manipulations and perceived risks. Path coefficient estimates are shown in Table 2. Dashed arrows, labeled e, represent errors in estimation of each consequent variable.

2. Method

Data were obtained through an experiment embedded in an online survey fielded in March 2016 using opt-in panels from Qualtrics, which randomly select respondents from its online market research panel partners (Qualtrics, 2014). We used a quota sample that matched the 2010 US Census in terms of age, gender, and geographic region. Individuals were notified via panel real-time software, email, or text and invited to participate in the survey for incentives. Some 1702 panelists started the survey and 1343 completed it, yielding a completion rate of 78.9%. Our final sample size was 1262 after removal of responses that were not fully complete. A response rate cannot be calculated as we do not know how many individuals were invited to participate. This is a result of using Qualtrics’ real-time software in addition to email or text invitations. While a non-probability sample is not optimal, it is suitable for addressing our hypotheses and research questions.

Experimental design

The experiment employed a 2 (emotion) × 2 (focus) between subjects design (Shadish et al., 2002) with an additional control group. Participants were randomly assigned to read one of five information articles about microbiome-related research. Scholars have primarily employed visual representations of repulsive images (e.g. Nabi, 1998) or physiological means (e.g. a bad smell) to evoke disgust (e.g. Schnall et al., 2008). However, as scientific issues are often communicated via text, how language shapes people’s attitudes toward science merits research. In addition, a meta-analysis of experiments that elicited emotions found that effect sizes among studies that used texts to elicit discrete emotions were comparable to those that elicited emotion through personal experiences (Lench et al., 2011). Therefore, we opted to manipulate textual information about microbiomes.

In each stimulus article, the first and last paragraphs were identical, and these comprised a control condition. The language manipulation added information to the control condition that either used neutral language (no disgust condition) or explicitly referred to disgusting stimuli (e.g. feces, fecal transplant therapies; disgust condition). To manipulate the focus of the article, the information was about research on either human or animal microbiomes. The control group was not included in this analysis as a balanced design allows for more effective testing and interpretation of interaction effects between the experimental manipulations.¹ After removal of the control group, the final sample size analyzed was 1005. Stimuli and sample sizes per condition can be found in the Supplementary Appendix.

Following exposure to the stimulus, respondents were asked questions designed to assess emotion evoked by the article and perceptions of risks and benefits. All variables used in this study were measured after exposure to the stimulus. We did not employ a manipulation check as they are unnecessary when the variations between conditions are intrinsic to the message. O’Keefe (2003) recommends this approach and we accordingly conceptualize elicited disgust as a mediating variable.

Measures

The language manipulation was dichotomously coded, with respondents exposed to the disgust condition coded high. Similarly, the focus manipulation was binary, with respondents who read about human microbiomes coded high.

Elicited disgust (M = 1.97, SD = 1.20) was measured immediately after exposure to one of the four experimental conditions. We asked respondents to report the extent to which they experienced disgust, distaste, anger, and fear while reading the information (1 = “None,” 5 = “A lot”). To examine whether elicited disgust was empirically distinguishable from other negative emotions, we ran two confirmatory factor analysis (CFA) models with the four emotion items. The first model was a single factor model, treating the four items as indicators of negative affect. This model failed to yield a satisfactory fit:² χ²(2) = 104.97 (p < .001), root mean square error approximation (RMSEA) = .226 with 90% confidence interval [.191, .264] ( $p_{R M S E A . 05} . 001$ ), comparative fit index (CFI) = .947, and standardized root mean square residual (SRMR) = .051. Modification indices suggested a strong correlation between disgust and distaste, and between anger and fear. We then ran a two-factor CFA with disgust and distaste loaded on one factor, and anger and fear on another.³ The two-factor model was a much better fit to the data: χ²(1) = 6.42 (p < .001), RMSEA = .073 with 90% confidence interval [.028, .132] ( $p_{R M S E A < . 05} = . 171$ ), CFI = .997, and SRMR = .006. The second model, compared with the first, was a significant improvement: Δχ²(1) = 98.55 (p < .001).

One concern with the CFA reported above was that these models had too few degrees of freedom, which could lead to us retaining mis-specified models (Goodboy and Kline, 2017; Holbert and Stephenson, 2008). To increase the degrees of freedom, we ran another two CFA models with an added factor, “involvement,” which was measured using three items.⁴ The first CFA, therefore, was a two-factor model with four negative emotional items loaded onto one factor, and the other factor was measured by the three involvement items. This model failed to yield a satisfactory fit: χ²(13) = 154.82 (p < .001), RMSEA = .104 with 90% confidence interval [.090, .119] (p_RMSEA _< .₀₅ < .001), CFI = .968, and SRMR = .057. Among the modification indices, the two strongest suggestions were again the correlated errors between disgust and distaste, and between anger and fear. The second CFA model was a three-factor model, with disgust and distaste, anger and fear, and the three involvement items loaded onto separate factors. This model provided a great fit to the data: χ²(11) = 15.28 (p = .170), RMSEA = .020 with 90% confidence interval [.000, .041] (p_RMSEA _< .₀₅ = .993), CFI = .999, and SRMR = .010. The second model, compared with the first, was a significant improvement: Δχ²(2) = 139.54 (p < .001). These two sets of CFAs, therefore, yielded consistent and clear evidence that the two items that were designed to measure disgust in our data (“disgust” and “distaste”) indeed captured a dimension that was empirically distinguishable from the other two negative emotions (“anger” and “fear”).

We operationalized our dependent variable, perceived risks (M = 7.19, SD = 2.43), by asking respondents how risky they thought modification of (1) human microbiomes, (2) plant and animal microbiomes, and (3) microbes in the environment are for society (0 = “Not at all risky,” 10 = “Very risky”). We included multiple indicators of risk perceptions as single-item measures of this construct are often plagued by random error (Binder et al., 2012). We averaged the three indicators to create an index of perceived risk (Cronbach’s α = .96).

We controlled for perceived benefits (M = 6.87, SD = 2.59) as risk and benefit perceptions have been shown to be interdependent (Cacciatore et al., 2011), which was operationalized similarly to perceived risks. We combined three items (Cronbach’s α = .96) that asked respondents how beneficial they viewed modification of (1) human microbiomes, (2) plant and animal microbiomes, and (3) microbes in the environment on scale ranging from 0 (“Not beneficial at all”) to 10 (“Very beneficial”).

Data analysis

We used regression-based path analysis with the aid of the computational tool, PROCESS 3.0 (http://www.processmacro.org; Hayes and Matthes, 2009; Preacher et al., 2007), in IBM SPSS Statistics. In addition to testing the moderating effects of the experimental manipulations on elicited disgust, we used PROCESS to test for moderated mediation (for more on the definition of moderated mediation and mediated moderation, see Muller et al., 2005), in which a moderated effect is carried through a mediator (Figure 1). We examined the mediating role of elicited disgust despite the lack of a significant relationship between the language manipulation and perceived risks as “lack of correlation does not disprove causation” (Bollen, 1989: 52).

3. Results

We first examined bivariate correlations between the variables of interest (Table 1). We found significant correlations between the emotion and focus manipulations and elicited disgust (Pearson’s r = .240, p < .001 and Pearson’s r = .122, p < .001, respectively), perceived risks and benefits (Pearson’s r = .090, p = .004), and perceived risks and elicited disgust (Pearson’s r = .136, p < .001).

Table 1.

Inter-item correlations of study variables (N = 1,005).

	(1)	(2)	(3)	(4)
(1) Perceived benefits
(2) Language manipulation (disgust vs. no disgust)	−.025 (.434)
(3) Focus manipulation (human vs. non-human)	−.023 (.474)	.007 (.827)
(4) Elicited disgust	.002 (.943)	.240 (< .001)	.122 (< .001)
(5) Perceived risks	.090 (.004)	.006 (.859)	.007 (.833)	.136 (< .001)

Values in parentheses are p-values (two-tailed).

Our first regression path model predicting elicited disgust explained 7.7% of the variance (F(4, 1000) = 21.1, p < .001) and respondents in the disgust condition, relative to the no disgust condition, experienced greater disgust (Table 2; path $a_{1}$ ; B = .393, standard error (SE) = .102, p < .001).

Table 2.

Ordinary least squares (OLS) regression model coefficients (N = 1,005).

Predictor		Elicited disgust			Perceived risks
Predictor		B (SE)	p		B (SE)	p
Intercept		1.604 (.122)	< .001		6.078 (.276)	< .001
Perceived benefits		.004 (.014)	.775		.083 (.029)	.004
Language manipulation (disgust vs. no disgust)	a₁ →	.393 (.102)	< .001	c₁ →	−.033 (.214)	.876
Focus manipulation (human vs. non-human)	a₂ →	.108 (.103)	.292	c₂ →	.055 (.214)	.796
Elicited disgust		—	—	b →	.295 (.066)	< .001
Language × Focus manipulation	a₃ →	.362 (.145)	.013	c₃ →	−.198 (.304)	.516
Model R²		.078	< .001		.027	< .001
Interaction ΔR²		.006	.013		.000	.516

H1, which proposed this effect would be stronger when the information was focused on humans compared with animals, was supported (path $a_{3}$ ; B = .362, SE = .013, p = .013). Elicited disgust was greater among respondents who were exposed to information about human microbiomes compared with those exposed to the article with a non-human focus (Figure 2).

Figure 2.

Interaction between the emotion and focus manipulations on elicited disgust.

In the model predicting perceived risks of modifying microbiomes (R² = 2.8%, F(5, 999) = 5.70, p < .001), we find support for H2 (path $b$ ; B = .295, SE = .066, p < .001), which proposed that elicited disgust would positively predict perceived risks of modifying microbiomes. H3 concerned moderated mediation and our data support elicited disgust as a mediator in the relationship between the language manipulation and perceived risks (paths $a_{1}$ and $b$ ), with the focus manipulation moderating the relationship between the language manipulation and elicited disgust (path $a_{3}$ ). The estimated conditional effect of the language manipulation on perceived risks is the following

(a_{1} + a_{3} \cdot f o c u s m a n i p u l a t i o n) b = (0.393 + 0.362 \cdot f o c u s m a n i p u l a t i o n) 0.295

We used bootstrapping to make inferences about the indirect effect (Hayes, 2009) and graphed the point estimates of the conditional indirect effect (Figure 3). That the 95% confidence interval, based on 5000 bootstrap samples, is above zero provides evidence that elicited disgust is a mediator of the interaction between the emotion and focus manipulations on perceived risks. This indirect conditional effect increased when the article focused on humans. The index of moderated mediation was .107 with a bootstrapped 95% confidence interval of [.018, .220].

Figure 3.

Conditional indirect effect of emotion and focus manipulations on perceived risks through elicited disgust based on 5000 bootstrap samples.

4. Discussion

We extend scholarship on discrete emotions, media effects, and risk perceptions by focusing on the relatively overlooked role of disgust and the context of microbiomes. This study also responds to cross-disciplinary calls for integrated efforts to better understand the social implications of a complex scientific topic, microbiomes (Handelsman and Stuhlberg, 2016). To this end, we examined the role of disgust in shaping risk perceptions related to modification of microbiomes. Given the nature of microbiomes, language that describes disgusting stimuli, evokes more vivid images, and is attention grabbing (e.g. mentioning feces and fecal transplants) is likely to be used as a lede by journalists. Our findings show that such language, particularly when focused on humans, elicits disgust, which predicted greater perceptions of risks related to modifying microbiomes. Our study offers insights that contribute to scholarship in risk perception, emotion, and broader conversations about science communication.

We are cognizant of the limitations of the present work. First, this study is vulnerable to critiques of ecological validity that can be extended to most experiments. Real-world media coverage of microbiome therapies and research is likely to include both images and text that elicit emotional responses. Future scholarship should include both, perhaps through modification of actual news articles, to further probe the effect of discrete emotions on risk (and benefit) perceptions and attitude formation. We also avoid making general claims about US adults as we used a non-probability sample, which was sufficient to test our hypotheses. Future research should employ probability samples to allow for inferences about the general population and examination of the extent to which disgust influences the dynamics of public attitude formation.

A second limitation, related to the first, concerns our single message design. To generalize the relationship between disgust and risk perceptions, a multiple message design study is preferred (O’Keefe, 2003). However, as our goal was partly to contribute to the literature on the social aspects of microbiomes, we opted to focus on one issue. Future work should include multiple messages to allow for generalization across contexts. Finally, our study overlooked cognitive routes to the formation of risk perceptions; much scholarship has already contributed to this line of research, albeit in other issue contexts (e.g. Ho et al., 2013; Runge et al., 2018). It is likely that cognitive pathways may interact with the emotional route modeled in this study and future scholarship should attempt to elucidate these pathways and articulate their interactions.

Our findings show that disgust can be elicited by information about microbiome therapies and research. This evoked emotion is negatively associated with risk perceptions related to the modification of microbiomes. These findings underscore flaws in the science literacy model (Miller, 1998), which does not account for emotional cues and assumes that knowledge and literacy will improve publics’ attitudes to scientific issues facing society. Other research has pointed out that individual characteristics, such as deference to scientific authority, religious beliefs, and political ideology, can be used as heuristic cues to predict people’s attitudes toward emerging sciences (e.g. Cacciatore et al., 2018; Yeo et al., 2014). While our work adds to the growing body of literature that challenges the science literacy model, it makes a unique contribution by examining how an elicited emotion influences people’s risk perceptions.

People are likely to encounter information about microbiomes in media like that presented in our experiment. This is especially true of online sources where most Americans obtain science information (National Science Board, 2018). As microbial therapies, the most common of which is FMT, are often represented as inherently disgusting (Chuong et al., 2015), it is likely that individuals may view this beneficial treatment and other microbiome-related technologies as distasteful. Evoked disgust can then act as an emotional cue that influences people’s risk perceptions, which, in turn, impacts public opinion.

Microbiome research is an emerging area of scholarship that has been the focus of much scientific and public interest (Ma et al., 2018). However, few studies have examined risk perceptions associated with it. Research on microbiomes and the resulting treatments and technologies have the potential to address many challenges currently facing society, including those of agriculture, public health, and the environment (Blaser et al., 2016). And while scientific effort in the field has grown, even researchers do not concern themselves overmuch with the risks associated with microbial technologies (Betts and Sawyer, 2012). Yet, a better understanding of how people perceive and form opinions about microbiome issues, either via emotion or cognition, is necessary to proactively address public concerns about moral, social, and ethical issues that may arise.

Several moral and ethical concerns related specifically to human microbiomes have recently been articulated (Ma et al., 2018; Metselaar and Widdershoven, 2017; Shamarina et al., 2017). While pathogen disgust is clearly linked to issues such as FMT, sociomoral disgust is likely to be evoked by these moral and ethical concerns. Sociomoral disgust extends the emotion into the social domain (Rozin et al., 2016). Therefore, it would be informative to consider specific types of disgust evoked by microbiomes and the extent to which it is elicited. Here, we briefly identify two issues that are likely to elicit sociomoral disgust among lay audiences: privacy and property concerns.

Microbiome research that aims to achieve personalized medicine goals or address public health issues (e.g. obesity) requires the use of donated biological materials and personal information. These biobanks necessitate public participation, but ethical concerns about data privacy, ownership, and access to donated samples and microbiomes may discourage publics from participation if they perceive associated risks. Moreover, if an individual’s microbiome has commercial value (e.g. for use as a probiotic), the question of who profits from it remains unresolved. We know much more about microbiomes now than we did when the current policies that regulate and govern biological data were put in place. Regulatory policies such as the Genetic Information Nondiscrimination Act (GINA) need to be updated to accommodate the advances that have been made in research (Shamarina et al., 2017). Public perceptions and attitudes are likely to play significant roles in how regulatory policies evolve, and a better understanding of the mechanisms and factors that influence risk perception and public opinion formation is necessary.

Disgust is an emotion that moralizes and censors (Rozin et al., 2016) and has been shown to influence support for policy and regulation related to microbiome research (Sun et al., 2019). As many emerging scientific issues involve ethical and moral considerations, these contexts are well suited to future studies of emotion, individual attitudes and perceptions, and public opinion. Reactions of disgust in contexts other than the one studied here (e.g. gene editing, synthetic biology) are likely to be more prevalent than we realize, particularly since several types of disgust have been identified (Rozin et al., 2016; Tybur et al., 2013), and are worthy of scholars’ attention. Research on disgust should also be extended to examining risk aversion and risk seeking. Previous work has shown that fear and anger are associated with risk-averse and risk-seeking choices, respectively (Lerner and Keltner, 2001), but little work has examined the relationship of such choices and disgust.

This study responds to the call to include social science in scholarship about microbiomes and makes a cross-disciplinary contribution to the literature on emotion and risk perceptions. Our study also extends scholarship on emotion and risk perceptions related to scientific issues by employing a discrete approach and focusing on a relatively overlooked emotion, disgust. The findings presented here offer a deeper understanding of how disgust serves as an emotional shortcut that influences people’s perceptions and is a necessary step toward proactively addressing public concerns over ethical, legal, and social implications, funding, and regulatory policies associated with emerging scientific issues facing society.

Supplemental Material

Supplements – Supplemental material for Disgusting microbes: The effect of disgust on perceptions of risks related to modifying microbiomes

Supplemental material, Supplements for Disgusting microbes: The effect of disgust on perceptions of risks related to modifying microbiomes by Sara K. Yeo, Ye Sun, Meaghan McKasy and Erika C. Shugart in Public Understanding of Science

Footnotes

Acknowledgements

The authors would like to thank Jessica Houf, Joanna Urban, and Sam Mandl for their contribution to this work.

Funding

This material is based on work supported by the American Society for Microbiology (ASM). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of the ASM.

Notes

Supplemental material

Supplemental material for this article is available online.

ORCID iD

Sara K. Yeo

Author biographies

Sara K. Yeo (PhD, University of Wisconsin–Madison) is an Assistant Professor in the Department of Communication at The University of Utah. Her research interests include science and risk communication with a focus on information seeking and processing. She is also trained as a bench and field scientist with an MS in Oceanography from the University of Hawai’i at Manoa.

Ye Sun (PhD, University of Wisconsin–Madison) is an Associate Professor in the Department of Communication at The University of Utah. Her research addresses various questions related to individual perceptions of media effects, especially in the domain of health communication.

Meaghan McKasy (MA, University of Utah) is a PhD candidate in the Department of Communication at The University of Utah. Her research focuses on environmental communication with a specific interest in energy communication.

Erika Shugart (PhD, University of Virginia) is Chief Executive Officer of the American Society for Cell Biology. Her professional career, including previous jobs at the National Academy of Sciences and the American Society for Microbiology, highlights her scientific background in biological research, skills in science communication, and leadership in marketing and strategic planning.

References

Alhakami

Slovic

(1994) A psychological study of the inverse relationship between perceived risk and perceived benefit. Risk Analysis 14(6): 1085–1096.

Alivisatos

Blaser

Brodie

Chun

Dang

Donohue

et al . (2015) A unified initiative to harness Earth’s microbiomes. Science 350(6260): 507–508.

Bechara

Damasio

(2000) Emotion, decision making and the orbitofrontal cortex. Cerebral Cortex 10(3): 295–307.

Bentler

(1990) Comparative fit indexes in structural models. Psychological Bulletin 107(2): 238–246.

Betts

Sawyer

(2012) Microbiome and risk perceptions. Emerging Science for Environmental Health Decisions Newsletter, March. Available at: http://nas-sites.org/emergingscience/files/2011/05/microbiome-nl-FINAL-080712.pdf

Binder

Cacciatore

Scheufele

Shaw

Corley

. (2012) Measuring risk/benefit perceptions of emerging technologies and their potential impact on communication of public opinion toward science. Public Understanding of Science 21(7): 830–847.

Blaser

Cardon

Cho

Dangl

Donohue

Green

et al . (2016) Toward a predictive understanding of Earth’s microbiomes to address 21st century challenges. mBio 7(3): e00714–e00716.

Bollen

(1989) Structural Equation Models with Latent Variables. New York, NY: John Wiley & Sons.

Britlab (2015). The disgusting creatures inhabiting your body. BBC News, 3 April. Available at: http://www.bbc.com/future/story/20150402-why-you-are-more-bug-than-human

10.

Cacciatore

Browning

Scheufele

Brossard

Xenos

Corley

(2018) Opposing ends of the spectrum: Exploring trust in scientific and religious authorities. Public Understanding of Science 27(1): 11–28.

11.

Cacciatore

Scheufele

Corley

(2011) From enabling technology to applications: The evolution of risk perceptions about nanotechnology. Public Understanding of Science 20(3): 385–404.

12.

Callaway

(2015) Microbiome privacy risk. Nature 521(7551): 136–136.

13.

Cantril

Allport

(1935) The Psychology of Radio. New York, NY: Harper & Brothers. Available at: http://archive.org/details/psychologyofradi00hadl

14.

Chuong

O’Doherty

Secko

(2015) Media discourse on the social acceptability of fecal transplants. Qualitative Health Research 25(10): 1359–1371.

15.

Clayton

Leshner

Bolls

Thorson

(2017) Discard the smoking cues—Keep the disgust: An investigation of tobacco smokers’ motivated processing of anti-tobacco commercials. Health Communication 32(11): 1319–1330.

16.

Cohen

(2016) Vaginal microbiome affects HIV risk. Science 353(6297): 331–331.

17.

Curtis

de Barra

Aunger

(2011) Disgust as an adaptive system for disease avoidance behaviour. Philosophical Transactions of the Royal Society B: Biological Sciences 366(1563): 389–401.

18.

Dillard

Peck

(2001) Persuasion and the structure of affect. Human Communication Research 27(1): 38–68.

19.

Dinan

Cryan

(2017) Gut–brain axis in 2016: Brain–Gut–Microbiota axis—Mood, metabolism and behaviour. Nature Reviews Gastroenterology & Hepatology 14: 69–70.

20.

Dsouza

Gilbert

(2017) Microbiome: Beyond the gut. Microbiology Today 44(2): 66–69.

21.

Dunlop

Wakefield

Kashima

(2008) Can you feel it? Negative emotion, risk, and narrative in health communication. Media Psychology 11(1): 52–75.

22.

Ekman

(1999) Basic emotions. In: Robinson

Harmon-Jones

(eds) Handbook of Cognition and Emotion. Chichester: John Wiley & Sons, pp. 45–60.

23.

Finucane

Alhakami

Slovic

Johnson

(2000) The affect heuristic in judgments of risks and benefits. Journal of Behavioral Decision Making 13(1): 1–17.

24.

Franzosa

Huang

Meadow

Gevers

Lemon

Bohannan

et al . (2015) Identifying personal microbiomes using metagenomic codes. Proceedings of the National Academy of Sciences of the United States of America 112(22): E2930–E2938.

25.

Frijda

(1986) The Emotions. London: Cambridge University Press.

26.

Goodboy

Kline

(2017) Statistical and practical concerns with published communication research featuring structural equation modeling. Communication Research Reports 34(1): 68–77.

27.

Handelsman

Stuhlberg

(2016) Announcing the national microbiome initiative, 13 May. Available at: https://www.whitehouse.gov/blog/2016/05/13/announcing-national-microbiome-initiative (accessed 1 June 2016).

28.

Hayes

(2009) Beyond Baron and Kenny: Statistical mediation analysis in the new millennium. Communication Monographs 76(4): 408–420.

29.

Hayes

Matthes

(2009) Computational procedures for probing interactions in OLS and logistic regression: SPSS and SAS implementations. Behavior Research Methods 41(3): 924–936.

30.

Scheufele

Corley

(2013) Factors influencing public risk-benefit considerations of nanotechnology: Assessing the effects of mass media, interpersonal communication, and elaborative processing. Public Understanding of Science 22(5): 606–623.

31.

Holbert

Stephenson

(2008) Commentary on the uses and misuses of structural equation modeling in communication research. In: Hayes

Slater

Snyder

(eds) The SAGE Sourcebook of Advanced Data Analysis Methods for Communication Research. Thousand Oaks, CA: SAGE, pp. 185–218.

32.

Hovland

Janis

Kelley

(1953) Communications and Persuasion: Psychological Studies of Opinion Change. New Haven, CT: Yale University Press.

33.

L-T

Bentler

(1999) Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling: A Multidisciplinary Journal 6(1): 1–55.

34.

Izard

(1977) Human Emotions. New York, NY: Springer. Available at: www.springer.com/us/book/9780306309861

35.

Izard

(2010) The many meanings/aspects of emotion: Definitions, functions, activation, and regulation. Emotion Review 2(4): 363–370.

36.

Keller

Visschers

Siegrist

(2012) Affective imagery and acceptance of replacing nuclear power plants. Risk Analysis 32(3): 464–477.

37.

Kline

(2011) Principles and Practice of Structural Equation Modeling (3rd ed.). New York, NY: Guilford Publications, Inc.

38.

Kühne

Schemer

(2015) The emotional effects of news frames on information processing and opinion formation. Communication Research 42(3): 387–407.

39.

Lee

C-J

Scheufele

(2006) The influence of knowledge and deference toward scientific authority: A media effects model for public attitudes toward nanotechnology. Journalism & Mass Communication Quarterly 83(4): 819–834.

40.

Lench

Flores

Bench

(2011) Discrete emotions predict changes in cognition, judgment, experience, behavior, and physiology: A meta-analysis of experimental emotion elicitations. Psychological Bulletin 137(5): 834–855.

41.

Lerner

Keltner

(2000) Beyond valence: Toward a model of emotion-specific influences on judgement and choice. Cognition & Emotion 14(4): 473–493.

42.

Lerner

Keltner

(2001) Fear, anger, and risk. Journal of Personality and Social Psychology 81(1): 146–159.

43.

Lerner

Valdesolo

Kassam

(2015) Emotion and decision making. Annual Review of Psychology 66(1): 799–823.

44.

Leshner

Bolls

Thomas

(2009) Scare ’em or disgust ’em: The effects of graphic health promotion messages. Health Communication 24(5): 447–458.

45.

Chen

Lan

Ren

(2018) Help, hope and hype: Ethical considerations of human microbiome research and applications. Protein & Cell 9(5): 404–415.

46.

Marchesi

Ravel

(2015) The vocabulary of microbiome research: A proposal. Microbiome 3: 31.

47.

Metselaar

Widdershoven

(2017) Ethical issues in fecal microbiota transplantion: Taking into account identity and family relations. The American Journal of Bioethics 17(5): 53–55.

48.

Miller

(1998) The measurement of civic scientific literacy. Public Understanding of Science 7(3): 203–223.

49.

Morales

Fitzsimons

(2012) How disgust enhances the effectiveness of fear appeals. Journal of Marketing Research 49(3): 383–393.

50.

Mou

Lin

(2014) Communicating food safety via the social media: The role of knowledge and emotions on risk perception and prevention. Science Communication 36(5): 593–616.

51.

Muller

Judd

Yzerbyt

(2005) When moderation is mediated and mediation is moderated. Journal of Personality and Social Psychology 89(6): 852–863.

52.

Nabi

(1998) The effect of disgust-eliciting visuals on attitudes toward animal experimentation. Communication Quarterly 46(4): 472–484.

53.

Nabi

(1999) A cognitive-functional model for the effects of discrete negative emotions on information processing, attitude change, and recall. Communication Theory 9(3): 292–320.

54.

Nabi

(2002) The theoretical versus the lay meaning of disgust: Implications for emotion research. Cognition & Emotion 16(5): 695–703.

55.

National Science Board (2018) Science and Engineering Indicators 2018. Arlington, VA: National Science Foundation. Available at: https://www.nsf.gov/statistics/2018/nsb20181/assets/nsb20181.pdf

56.

Netemeyer

Burton

Andrews

Kees

(2016) Graphic health warnings on cigarette packages: The role of emotions in affecting adolescent smoking consideration and secondhand smoke beliefs. Journal of Public Policy & Marketing 35(1): 124–143.

57.

Newhagen

(1998) TV news images that induce anger, fear, and disgust: Effects on approach-avoidance and memory. Journal of Broadcasting & Electronic Media 42(2): 265–276.

58.

Oaten

Stevenson

Case

(2009) Disgust as a disease-avoidance mechanism. Psychological Bulletin 135(2): 303–321.

59.

O’Keefe

(2003) Message properties, mediating states, and manipulation checks: Claims, evidence, and data analysis in experimental persuasive message effects research. Communication Theory 13(3): 251–274.

60.

Panksepp

Watt

(2011) What is basic about basic emotions? Lasting lessons from affective neuroscience. Emotion Review 3(4): 387–396.

61.

Peters

Slovic

(1996) The role of affect and worldviews as orienting dispositions in the perception and acceptance of nuclear power. Journal of Applied Social Psychology 26(16): 1427–1453.

62.

Plutchik

(2001) The nature of emotions. American Scientist 89(4): 344–350.

63.

Preacher

Rucker

Hayes

(2007) Addressing moderated mediation hypotheses: Theory, methods, and prescriptions. Multivariate Behavioral Research 42(1): 185–227.

64.

Qualtrics (2014) ESOMAR 28: 28 questions to help research buyers of online samples. Available at: https://amerispeak.norc.org/Documents/FeatureDocuments/NORC_AmeriSpeak_ESOMAR_28.pdf

65.

Rozin

Haidt

McCauley

(2016) Disgust. In: Barrett

Lewis

Haviland-Jones

(eds) Handbook of Emotions, 4th edn. New York, NY: Guilford Press, pp. 815–834.

66.

Rubenking

Lang

(2014) Captivated and grossed out: An examination of processing core and sociomoral disgusts in entertainment media. Journal of Communication 64(3): 543–565.

67.

Runge

Chung

LY-F

Brossard

Scheufelea

(2018) Pink slimed: Media framing of novel food technologies and risk related to ground beef and processed foods in the U.S. Meat Science 143: 242–251.

68.

Sauter

Eisner

Ekman

Scott

(2010) Cross-cultural recognition of basic emotions through nonverbal emotional vocalizations. Proceedings of the National Academy of Sciences of the United States of America 107(6): 2408–2412.

69.

Schnall

Haidt

Clore

Jordan

(2008) Disgust as embodied moral judgment. Personality and Social Psychology Bulletin 34(8): 1096–1109.

70.

Shadish

Cook

Campbell

(2002) Experimental and Quasi-experimental Designs for Generalized Causal Inference. Belmont, CA: Wadsworth.

71.

Shamarina

Stoyantcheva

Mason

Bibby

Elhaik

(2017) Communicating the promise, risks, and ethics of large-scale, open space microbiome and metagenome research. Microbiome 5: 132.

72.

Shimp

Stuart

(2004) The role of disgust as an emotional mediator of advertising effects. Journal of Advertising 33(1): 43–53.

73.

Singh

Trivedi

(2017) Microbiome and the future for food and nutrient security. Microbial Biotechnology 10(1): 50–53.

74.

Slovic

(1999) Trust, emotion, sex, politics, and science: Surveying the risk-assessment battlefield (Reprinted from Environment, Ethics, and Behavior, pp. 277–313, 1997). Risk Analysis 19(4): 689–701.

75.

Slovic

Flynn

Layman

(1991) Perceived risk, trust, and the politics of nuclear waste. Science 254(5038): 1603–1607.

76.

Steiger

(1990) Structural model evaluation and modification: An interval estimation approach. Multivariate Behavioral Research 25(2): 173–180.

77.

Sun

Yeo

McKasy

Shugart

(2019) Disgust, need for affect, and responses to microbiome research. Mass Communication and Society. Available at: https://doi.org/10.1080/15205436.2019.1565786

78.

Tybur

Lieberman

Kurzban

DeScioli

(2013) Disgust: Evolved function and structure. Psychological Review 120(1): 65–84.

79.

Van Rossum

Pylatuk

Osachoff

Griffiths

Quach

et al . (2016) Microbiome analysis across a natural copper gradient at a proposed northern Canadian mine site. Microbiotechnology, Ecotoxicology and Bioremediation. Available at: https://www.frontiersin.org/articles/10.3389/fenvs.2015.00084/full

80.

Vytal

Hamann

(2010) Neuroimaging support for discrete neural correlates of basic emotions: A voxel-based meta-analysis. Journal of Cognitive Neuroscience 22(12): 2864–2885.

81.

Watson

Tellegen

(1985) Toward a consensual structure of mood. Psychological Bulletin 98(2): 219–235.

82.

Yeo

Cacciatore

Brossard

Scheufelea

Rungea

et al . (2014) Partisan amplification of risk: American perceptions of nuclear energy risk in the wake of the Fukushima Daiichi disaster. Energy Policy 67: 727–736.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.01 MB