The regulation of IPD appears to be automatic, involving a delicate balance between approach-avoidance forces that safeguard our physical well-being in different situations, such as those involving potential health risks (Hayduk, 1978; Iachini et al., 2021). Here, we set out to investigate whether conspicuous facial disease cues would modulate participants’ IPD behavior in a modified IVAS task. Moreover, we also postulated that this effect would be further influenced by individual differences in perceived vulnerability to disease. Our results only partially confirmed our hypotheses, as facial disease cues were capable of modulating the distance assumed from another in social interaction. However, perceived vulnerability to disease did not appear to significantly influence participants' approach intents, not supporting our second hypothesis.
As anticipated, our findings revealed that avatars associated with facial disease cues prompted larger distances compared to those without such indicators, aligning with previous studies on approach-avoidance behaviors with disease-related stimuli (e.g., Iachini et al., 2021; Lisi et al., 2021; Rumsey et al., 1982). As such, our results provide additional support for the link between IPD and BIS, which, from an evolutionary standpoint, reflects a crucial and highly adaptable behavioral strategy of maintaining a safe physical gap from potential sources of infection to mitigate risks. Participants’ evaluations of the face stimuli also corroborate this finding by validating our initial manipulation: faces with disease cues were deemed as more disgusting, more associated with disease, and more uncomfortable to interact with than their control counterparts.
Furthermore, a more in-depth analysis revealed that faces with flu-like appearance were associated with a greater IPD compared to facial rash. One possible explanation for this result might pertain to the fact that facial rash can be perceived as less symptomatic of a contagious disease, similar to a birthmark or other non-infectious condition and, thus, viewed as a more permanent and less menacing cue, compared to flu-like appearance (Ryan et al., 2012). This interpretation is consistent with participants' ratings, which indicated that flu-like faces were more strongly associated with disease and aligns with previous research on the impact of permanent disfigurement on interpersonal distance (Houston & Bull, 1994; Rumsey et al., 1982). However, it is worth noting that our findings appear to diverge somewhat from the results reported by Ryan and colleagues (2012), who did not observe significant differences between “false alarm cues” (e.g., birthmarks) and “real disease signs” (e.g., influenza) across a number of measures. Despite these discrepancies, our study aligns with their conclusion, as both disease cues elicited significant differences in participants' IPD behavior and evaluations of the face stimuli compared to control faces, as mentioned earlier.
Therefore, we hereby reinforce the idea that all cues conveying potential disease, albeit to varying degrees, can elicit behavioral responses associated with the BIS, such as disgust and avoidance. This is especially relevant if one considers the potential negative implications (e.g., discrimination, prejudice, stigmatization; Schaller & Park, 2011) that might befall individuals displaying benign disease cues (e.g., acne, vitiligo). It implies that efforts to use educational programs aimed at reducing avoidant behaviors towards individuals with facial disease cues, regardless of their potential for contamination, may not be as effective as desired. Instead, understanding the underlying mechanisms of these social processes can help identify and mitigate their impact on people with such conditions. However, caution is needed as these are merely exploratory results, and demand a more detailed exploration by future studies.
Importantly, our results further indicate that although the effect of disease cues over IPD can be observed irrespective of the sex of the stranger, the magnitude of this effect seems increased for female avatars, compared to their male counterparts. Specifically, a more prominent decrease in approach behavior from control to disease conditions was found for females, compared to male, avatars. This finding suggests that the presence of disease cues in female avatars had a stronger impact on the participants' approach behavior, hinting at a potential gender-specific effect in response to disease. This differential response might be explained by several reasons, one of which concerns the social and cultural factors associated with gender stereotypes. Indeed, females are often associated with the role of caretakers and, thus, are expected to be more nurturing and engage more easily in social bonding than males. In contrast, males are tendentially perceived as more physically imposing or threatening (Cicone & Ruble, 1978) and, consequently, less approached in social interaction. Accordingly, our results revealed that female avatars were approached more frequently overall compared to male avatars, consistent with prior literature (e.g., Iachini et al., 2016). Thus, a possible explanation for the greater decline reported for female avatars could be related to the fact that they exhibited a higher level of approach behavior in the control condition. This would mean that a larger decrease was required to reach a perceived safe distance in the disease cues condition for females, even though it was still smaller than that of male avatars. On the other hand, an equally valid explanation might pertain to the number of women in our sample. Proxemics research with dyads has shown that women seem to have a preexisting tendency to approach other females more often (e.g., Hayduk, 1983; Iachini et al., 2016; Uzzell & Horne, 2006). It is also possible that a combination of both these proposed explanations is responsible for the effects observed relating to sex. However, these are speculative explanations, and further research is needed to better disentangle this result.
Perceived vulnerability to disease was found not to influence IPD behavior, contrary to what was expected, but to support the results of Lisi and colleagues (2021). Likewise, no significant effect was found for DPSS-R or LSAS. Since individual differences are known to influence BIS and IPD responses, this lack of significant differences might simply be related to sample characteristics. Thus, it is worthwhile for future studies to keep exploring their potential grasp on these mechanisms, perhaps by employing a between-subjects design that includes distinct groups with high and low levels of these variables. This would allow for a more comprehensive examination of their influence on interpersonal distance, providing valuable insights into the nuanced dynamics at play.
Some limitations should be noted. Firstly, and as discussed above, the limited number of males in our sample may be influencing the results, as dyads are known to affect IPD regulation. In particular, female dyads (i.e., both the participant and the confederate are female) are known to maintain smaller distances than male dyads (e.g., Hayduk, 1983; Uzzell & Horne, 2006), although this is not always the case, as seen in the study by Iachini and colleagues (2016). Future studies would benefit from using a more balanced sample to weed out possible biases related to participants’ sex. Secondly, although previous research has demonstrated the reliability of the IVAS task (Iachini et al., 2016), replication of the current findings in a more ecological context is warranted to enhance generalizability. For example, future studies could take advantage of virtual reality scenarios or real in-person interactions where the true goal of the experiment is masked to help mitigate potential biases introduced by participants' awareness of being observed, thus providing a more realistic assessment of IPD adjustment. Furthermore, by allowing participants to be more immersed in the task, it may be possible to better examine the real impact of infection risk on IPD, as our task utilized only photos presented on a computer screen, which do not pose an actual threat. Additionally, exploring other types of disease stimuli (e.g., auditory) or contextual backgrounds (e.g., hospital), or manipulating the task instructions (e.g., actor/fake vs actual sickness) could provide further insights into the impact of multi-modal cues, ecological validity, and cognitive-perceptual factors on this topic. Finally, while our task used an active approach (i.e., participants were asked to approach the avatar), it would be interesting to study the reverse scenario (i.e., passive approach - the avatar approaches the participant), as previous research has shown that being approached often leads to greater IPDs compared to actively approaching (Iachini et al., 2016; but see Lee & Chen, 2021; Saporta et al., 2021). Moreover, considering the role of familiarity is important, as studies consistently demonstrate that people tend to prefer a greater distance from strangers compared to family and friends (Hayduk, 1983). Thus, future studies should explore both active and passive approaches and consider the influence of familiarity to gain a comprehensive understanding of interpersonal distance regulation by facial disease cues.
In sum, the present findings provide further support to the notion that approach-avoidance behaviors and IPD regulation appear to be sensitive to salient disease cues, and that this effect may be moderated by the sex of the to-be-approached individual. Importantly, this privileged distance from potentially sick others may play a role in important social processes such as discrimination and prejudice (Schaller & Park, 2011), particularly against males. Overall, this differential adjustment of IPD may reflect the operation of the BIS by facilitating protective approach-avoidance reactions in the presence of perceived health threats.