Body odor disgust sensitivity is positively related to social, but not economic conservatism, & to implicit bias toward an outgroup; social attitudes may be linked to basic chemosensory processes (pathogen cues)

An Overprotective Nose? Implicit Bias Is Positively Related to Individual Differences in Body Odor Disgust Sensitivity. Marta Zuzanna Zakrzewska et al. Front. Psychol., February 28 2020. https://doi.org/10.3389/fpsyg.2020.00301

Abstract: Body odors are universal elicitors of disgust, a core emotion that plays a key role in the behavioral immune system (BIS) – a set of psychological functions working to avoid disease. Recent studies showed that body odor disgust sensitivity (BODS) is associated with explicit xenophobia and authoritarianism. In the current experimental pre-registered study (https://osf.io/6jkp2/), we investigated the association between olfactory pathogen cues, BODS and implicit bias toward an outgroup (tested by an implicit association test). Results show that BODS is positively related to implicit bias toward an outgroup, suggesting that social attitudes may be linked to basic chemosensory processes. These attitudes were not influenced by background odors. Additionally, BODS was related to social, but not economic conservatism. This study extends the BIS framework to an experimental context by focusing on the role of disgust and body odors in shaping implicit bias.

Discussion

In the current preregistered study, we found that disgust sensitivity to body odors (BODS) is related to implicit bias toward an outgroup. This result corroborates and extends previous findings about the relationship between BODS, a BIS-related measure (Liuzza et al., 2017), and attitudes toward outgroups (Zakrzewska et al., 2019). These findings strengthen the view that some individuals may have a more sensitive BIS which makes them prefer behaviors and attitudes that limit contact with out-groups.

Contrary to our hypothesis, implicit bias was not influenced by a background body-like unpleasant odor. We need to point out two possible drawbacks of the paradigm used in our study. First, the odors were presented constantly over the duration of an experimental block. This might have resulted in habituation (Smeets and Dijksterhuis, 2014) to the smell, and suppression of its effect, although similar studies in our lab has shown no evidence of habituation (Syrjänen et al., 2017, 2019). To confirm the null effect of odors (or to find evidence for its existence), further research is needed where odor cues are brief and paired with each target stimulus (as they are in a priming task) rather than serving as background odors. Research shows that congruency effects can be observed on a single trial level through priming, also in the olfactory modality (e.g. Olofsson et al., 2014; Kastner et al., 2016). Second, we did not use actual body odors but odors that resemble them. However, valeric acid is present in body odor, especially in disease (Pandey and Kim, 2011; Shirasu and Touhara, 2011), and the substance has been widely used as sweat-like odor in other studies (e.g. Anderson et al., 2003; Jacob et al., 2003). What may be of more importance is that the smell was not presented in a way suggesting that it came from another person. Again, we believe that a cue – target priming design could help reveal the influence of odor on group biases. In hindsight, although we did not find any odor effects, we think that similar studies would benefit from having not only a body-odor related unpleasant odor, but also another, non-bodily yet unpleasant odor, thus allowing to talk more directly about body odor disgust.

Even though the theoretical framework emphasizes the role body odors and body-odor related disgust we would like to acknowledge a potential limitation of our study in drawing conclusions about the effect of body odors. Namely, BODS is the only measure of disgust we used. While comparing BODS with other disgust measure (such as DS-R or TDDS) we could have potentially shown an olfaction-specific link between disgust sensitivity and implicit bias. However, previous study (Liuzza et al., 2017) showed that BODS is correlated with both DS-R and TDDS (0.37 > = r < = 0.65) while at the same time being more strongly related to PVD than to the other measures, pointing to the relevance of body odor disgust sensitivity (BODS) for pathogen avoidance. Thus, we believe that although including another measure of disgust could have strengthened our claims, we would not have gained much more valuable information.

Faulkner et al. (2004) showed elevated BIS activation and IAT bias for a task evoking danger (rather than general unpleasantness), yet we did not find this effect for our disease-related IAT. In fact, the two versions of the IAT were highly correlated in the current sample. Since health/illness-related words are also strongly valenced, the lack of task version effect might be explained by a difficulty to differentiate between the strength of the target group to a health/illness concept, vs. the target group to a positive/negative concept associations, especially given the relatively small sample.

Lastly, our results should be viewed in the light of the dual process framework. We could not replicate our previous findings on US samples (Liuzza et al., 2018), where we found a stable, small-to-medium relationship between RWA and BODS. It is possible that our sample, including Swedish college students, is not representative of all ranges of authoritarian attitudes (only two observations fell above the theoretical midpoint of the RWA scale). However, it should be noted that, when comparing nationally representative samples and convenience samples, Aarøe et al. (2017) found almost identical correlation coefficients for the association between individual differences in disgust sensitivity and social attitudes. By including these new data to update previous observations from Liuzza et al. (2018), we found that the previous positive relationship between BODS and RWA is still credible. Additionally, current results show a relationship between BODS and another measure of social conservatism, namely the SDS. These relationships can be explained in the light that RWA, SDS and BODS can be related to avoidance of possible pathogen threats, in the world perceived as a dangerous place. Similarly to Liuzza et al. (2018) we found no relationship between BODS and a measure of social dominance (SDO), which refers more to the perception of the world as a competitive jungle and is thus connected to other sources of prejudice, not related to disease avoidance.

Taken together, this study extends current knowledge about odor disgust and BIS with evidence suggesting that social attitudes may be linked to basic chemosensory processes.