Introduction

Obsessive–compulsive disorder (OCD) is a mental disorder characterized by the presence of obsessions and/or compulsions and affects approximately 2% of the population (American Psychiatric Association, 2013). OCD has been shown to have a significant impact on individuals’ social, personal, and professional lives (Abramowitz & Jacoby, 2015). Indeed, the World Health Organization listed anxiety disorders (including OCD) as one of the top contributors to non-fatal health loss globally (Murray et al., 1996). While the nature of individuals’ obsessional content and compulsive behaviors is idiosyncratic, symptoms have been shown to cluster in established themes — contamination, checking, repugnant obsessions, and ordering/symmetry (e.g., Goodman et al., 1989). In addition to these more traditional symptoms domains, researchers have also devoted greater empirical attention to mental contamination (i.e., feelings of contamination and/or washing behavior that arise in the absence of contact with a contaminant) (Coughtrey et al., 2012; Rachman, 2004; Rachman et al., 2015) as a symptom category distinct from contact contamination symptoms. Much remains unknown about the etiology of OCD; however, researchers have identified various vulnerability factors implicated in the onset and maintenance of the disorder. The literature in this field has traditionally focused on cognitive factors (e.g., Obsessive Compulsive Cognitions Working Group, 2005; Rector et al., 2009), although more recently, researchers have identified affective vulnerability factors, including disgust proneness, that may interact with cognitive factors to predict the development of OCD symptoms (Knowles et al., 2018).

OCD has traditionally been characterized as an anxiety- or fear-based disorder. However, researchers have acknowledged the relevance of disgust, both as an outcome of, and vulnerability factor for, OCD symptomatology. Disgust proneness has been defined as a combination of one’s propensity to experience disgust and the degree of aversiveness associated with the experience of disgust (Van Overveld et al., 2006). As an antecedent to symptomatology, individual differences in disgust proneness appear to influence individuals’ odds of developing OCD (Knowles et al., 2018). Given the associations between disgust and dirt/germs, research on the role of disgust proneness in the context of OCD has traditionally focused on contamination-related symptomatology specifically (e.g., Olatunji et al., 2004, 2010a, b) or has investigated OCD symptoms in general, collapsed across all OCD symptom domains (e.g., Olatunji et al., 2011).

A smaller body of literature has identified specific associations between disgust proneness and non-contamination-related OCD symptoms such as repugnant obsessions (Ching et al., 2018), as well as neutralizing and ordering symptoms in OCD (Olatunji et al., 2010a, b). Disgust has traditionally been conceptualized as an emotion that serves a “disease avoidance” function, explaining the well-established link between disgust proneness and contamination-based OCD symptomatology. However, researchers have demonstrated that feelings of disgust may also serve to protect against social “contaminants” by serving a similar avoidance function in response to immorality and the misfortunes of others (Olatunji et al., 2010a, b). This concept of moral disgust may, therefore, explain the link between disgust proneness and repugnant obsessions, as individuals prone to experiencing disgust may be hypervigilant for and have a more intense emotional reaction to immoral intrusive thoughts. Along these lines, individuals prone to experiencing disgust may also be more inclined to appraise “just right” obsessions (i.e., thoughts that things are not quite right or are incomplete) as being more threatening, leading to a more intense emotional response.

Research has shown that the greater the intensity of one’s negative emotional response, the more likely one is to engage in behavior to neutralize those emotions (Macatee & Cougle, 2015; Raines et al., 2014). Indeed, those high in disgust proneness tend to experience disgust more aversively and intensely than those low in disgust proneness. Therefore, one might expect a relationship between disgust proneness and neutralizing behavior. Given the prototypical conceptualization of disgust as a response to physical contamination, the bulk of research on neutralizing behavior in this context has focused on washing and cleaning behavior and has demonstrated a strong link between these constructs. However, the established associations between disgust proneness and mental neutralizing and ordering symptoms in OCD (Olatunji et al., 2010a, 2010b) suggest that this may not be the only type of behavior individuals engage in to neutralize feelings of disgust. Indeed, many individuals may arrange objects in their environment to restore a sense of cleanliness and order in response to these feelings. Others, still, may engage in physical or mental checking for signs of contaminants that could be physical or moral in nature (e.g., checking themselves/their clothes for signs of dirt or contamination; replaying events in their head to make sure they avoided contact with physical or social contaminants).

Cognitive models of OCD emphasize particular beliefs and appraisals as key cognitive mechanisms responsible for the experience of symptomatology (Rachman, 1997, 1998; Salkovskis, 1985). More specifically, the Obsessive Compulsive Cognitions Working Group (2005) identified three main OCD-relevant belief domains in the development of the Obsessive Beliefs Questionnaire 44-item version (OBQ-44) — perfectionism/intolerance of uncertainty, importance/control of thoughts, and responsibility/threat overestimation. These belief domains have been shown to interact with individuals’ levels of disgust proneness to predict different types of OCD symptoms. Specifically, researchers suggest that when individuals are more prone to experiencing disgust and experience disgust as a more aversive emotional state, they may be more likely to go on to develop negative beliefs and appraisals about disgust-provoking triggers. For example, those high in disgust proneness may overestimate the threat associated with disgust-evoking stimuli or may appraise themselves as more responsible for preventing negative outcomes associated with disgust-evoking stimuli. These appraisals of inflated responsibility and overestimation of threat may, in turn, lead to heightened OCD symptomatology. Along these lines, individuals high in disgust proneness and OBQ beliefs of responsibility and threat were shown to endorse more severe sexual orientation-related OCD symptoms (Ching et al., 2018).

Other contamination-related cognitive mechanisms that have been identified as relevant predictors of contact contamination symptoms in OCD are contamination thought–action fusion (CTAF) and contamination sensitivity (CS). CTAF, which captures the tendency to fuse contamination-related thoughts and events, and CS, which refers to the degree of distress caused by feelings of contamination, are two constructs that have proven to be helpful intervention targets for clinicians working with individuals with contamination-related OCD (Rachman et al., 2015; Radomsky et al., 2014). Further, CTAF appears to interact with disgust propensity and sensitivity to predict feelings of mental contamination, which in turn predict subsequent contact contamination symptoms (Inozu et al., 2021). Taken together, this suggests that in addition to the traditional OBQ-44 belief domains, CTAF and CS may also be important cognitive mechanisms at play in the relationship between disgust proneness and OCD symptoms, particularly with regard to both contact and mental contamination-related symptoms.

Links have been established between cognitive mechanisms and OCD symptoms, as well as between affective vulnerability factors and OCD symptoms. However, few studies have examined how these cognitive factors might account for the relationship between vulnerability factors such as disgust proneness and OCD symptoms. Fewer still have looked at the relative significance of these different cognitive mechanisms across different OCD symptom domains. Thus, identifying the specific vulnerability pathways that contribute to different OCD symptom domains could help to establish more specific intervention targets for individuals with different OCD symptom presentations. As such, the aim of the present study was to address the gap in our understanding of the cognitive mechanisms that explain the relationship between disgust proneness and OCD symptomatology by systematically exploring the associations between disgust proneness, cognitive processes, and individual OCD symptom domains. We hypothesized that CTAF, CS, and all OBQ-44 belief domains would be significant mediators of the relationships between disgust proneness and OCD symptoms across symptom domains. Our investigation of cognitive mediators between disgust proneness and OCD symptom domains was largely exploratory in nature given the sparsity of existing research on disgust as an affective vulnerability factor across OCD symptom domains.

Method

Participants

The sample consisted of 149 female undergraduate students from Concordia University who were recruited as part of an experimental study exploring the role of responsibility in mental contamination (results reported elsewhere; Krause & Radomsky, 2021). The overarching study from which the data was collected involved a sexual harassment imagery task. For this reason, only female participants were recruited. Participants received either course credit or an entry ballot into a cash draw as compensation for participation. To be eligible, participants had to be female, at least 18-year-old, and be able to read, write, and communicate in English. Participants’ mean age was 22.86 (SD = 4.90, range = 18 to 48) years. See Table 1 for participant demographics.

Table 1 Demographics and descriptive statistics for self-report measures (N = 149)
Full size table

Measures

Demographic Questionnaire

Participants completed a questionnaire assessing basic demographic information including age, ethnicity, marital status, and occupation.

Vancouver Obsessional Compulsive Inventory (VOCI; Thordarson et al., 2004)

The VOCI is a 55-item self-report measure of OCD symptoms, comprised six subscales: contamination, checking, obsessions, hoarding, indecisiveness, and “just right” experiences. All items are rated on a five-point Likert scale ranging from 0 (“not at all”) to 4 (“very much”). The VOCI has strong psychometric properties, including excellent internal consistency (\(\alpha\) = 0.94 to 0.98), excellent retest reliability (r = 0.91), and excellent convergent and divergent validity (Radomsky et al., 2006; Thordarson et al., 2004). In the present sample, this scale demonstrated excellent internal consistency (\(\alpha\) = 0.97).

Vancouver Obsessional Compulsive Inventory: Mental Contamination Scale (VOCI-MC; Radomsky et al., 2014)

The VOCI-MC is a 20-item self-report questionnaire that assesses mental contamination. Items are rated on a five-point Likert scale ranging from 0 (“not at all”) to 4 (“very much”). The VOCI-MC has demonstrated excellent internal consistency (\(\alpha\) = 0.93 to 0.97), good convergent, and divergent validity (Radomsky et al., 2014). In our sample, the VOCI-MC showed excellent internal consistency (\(\alpha\) = 0.94).

Disgust Scale (DS; Haidt et al., 1994)

The DS is 32-item self-report questionnaire designed to assess individual proneness to experience disgust across eight domains: food, animals, body products, body envelope violations, death, sex, hygiene, and magical thinking. Sixteen items are rated as true or false, and 16-items are rated on a three-point Likert scale ranging from 1 (“not disgusting at all”) to 3 (“very disgusting”). In our sample, the DS demonstrated acceptable internal consistency (\(\alpha\) = 0.75).

Obsessive Beliefs Questionnaire (OBQ-44; OCCWG, 2005)

The OBQ-44 is a 44-item self-report measure of beliefs which are associated with OCD symptomatology. The measure is comprised of three subscales: responsibility and threat overestimation, perfectionism and intolerance for uncertainty, and importance/control of thoughts. Items are rated on a seven-point Likert scale ranging from 1 (“disagree very much”) to 7 (“agree very much). The OBQ-44 subscales have good-to-excellent internal consistency (\(\alpha\) = 0.89 to 0.93) and good criterion and convergent and divergent validity (OCCWG, 2005). Excellent internal consistency was found for this scale in the current sample (\(\alpha\) = 0.96).

Contamination Thought–Action Fusion Scale (CTAF; Radomsky et al., 2014)

The CTAF is a nine-item self-report questionnaire which assesses the relationship between thoughts about contamination and feelings and behaviors associated with contamination. Each item is rated on a five-point Likert scale ranging from 0 (“strongly disagree”) to 4 (“strongly agree”). The CTAF has excellent internal consistency (\(\alpha\) = 0.93 to 0.96) and satisfactory convergent and divergent validity (Radomsky et al., 2014). In the present sample, the CTAF demonstrated excellent internal consistency (\(\alpha\) = 0.94).

Contamination Sensitivity Scale (CSS; Radomsky et al., 2014)

The CSS is a 24-item self-report questionnaire which measures the levels of distress associated with feelings of contamination. Each item is rated on a five-point Likert scale ranging from 0 (“not at all”) to 4 (“very much”). This measure has excellent internal consistency (\(\alpha\) = 0.90 to 0.94) and good convergent and divergent validity (Radomsky et al., 2014). The CSS showed excellent internal consistency in our sample (\(\alpha\) = 0.91).

Procedure

These data were collected in the context of a larger study (Krause & Radomsky, 2021) exploring the impact of perceptions of responsibility on mental contamination.Footnote 1 Participants enrolled in the study through Concordia University’s undergraduate participation pool. Participants came into the laboratory to take part in the experimental study (Krause & Radomsky, 2021). At the end of the experimental study, participants completed the questionnaire battery, including the OBQ-44, VOCI, VOCI-MC, DS, CTAF, and CSS. Finally, participants were debriefed and received course credit for their participation.

Results

Zero-order correlations between the variables of interest were calculated (see Table 2). To control for familywise error rate, a Bonferroni correction was applied (α = 0.05/12 = 0.004; Andrade, 2019). As Table 2 indicates, almost all variables showed statistically significant positive associations with one another (ps < 0.001). The only correlations that were not significant after applying the correction were between VOCI “just right” and disgust proneness and VOCI obsessions and disgust proneness. Please refer to Table 1 for means and standard deviations for all variables of interest.

Table 2 Correlation matrix
Full size table

Parallel mediation models, testing for indirect effects simultaneously, were carried out using the PROCESS feature (version 4) (Hayes, 2017) for SPSS. In six separate parallel mediation models, we examined if the associations between disgust proneness and OCD symptom domains (VOCI total score, VOCI contamination, VOCI mental contamination, VOCI obsessions, VOCI checking, VOCI “just right”) were mediated by cognitive processes, namely, OCD belief domains (perfectionism/intolerance for uncertainty, importance/control of thoughts, responsibility/threat overestimation), CTAF, and CS. Following Hayes’s (2017) recommendation, the sampling distribution of indirect effects was bootstrapped 5000 times to provide the lower and upper bounds of the 95% confidence interval (CI) of the indirect effects. A CI excluding the value of 0 indicates that the indirect effect is statistically significant with 95% confidence. In in the interest of interpretability, standardized values for all effects are presented.

Considering that this parallel mediation approach relies on bootstrap resampling to construct 95% confidence intervals for each test (i.e., multiple resamples with replacement are generated from a single set of observations), no corrections were applied for multiple testing (Preacher & Hayes, 2004). Mediation analyses using bootstrapping are not sensitive to deviations from normality assumptions, as the bootstrap confidence interval method to test for indirect effects does not assume a normal distribution for the indirect effects (Bollen & Stine, 1990; McKinnon, 2008). In fact, in bootstrapping, the distribution of estimates is nonparametric estimates of the sampling distributions of the indirect effects. As such, contrary to regular confidence intervals, bootstrap confidence intervals can be nonnormal (Preacher & Hayes, 2004). The main assumption in mediation analyses using bootstrapping procedures for estimating indirect effects is that the indirect effects need to be stronger than the direct effects. Therefore, the direct effects of disgust proneness on all dependent variables (VOCI total score, VOCI contamination, VOCI mental contamination, VOCI obsessions, VOCI checking, VOCI “just right”) are reported below.

The initial mediation model (Fig. 1) comprised the VOCI total score as the dependent variable. The direct effect of disgust proneness on the VOCI total score was not statistically significant (unstandardized b = 0.43, SE = 0.45, p = 0.34). The results indicated that the total effect of disgust proneness on the VOCI total score was statistically significant and of moderate magnitude (β = 0.31, SE = 0.60, p = 0.001). Standardized indirect effects are presented in Table 3. There was a positive indirect effect of disgust proneness on the VOCI total score through CS that was small but statistically significant (indirect effect β coefficient = 0.06; 95% CI = 0.001 to 0.120). In addition, results indicated a statistically significant positive indirect effect of disgust proneness on the VOCI total score through CTAF that was also small in size (indirect effect β coefficient = 0.06; 95% CI = 0.007 to 0.124). There was also a small, statistically significant positive indirect effect of disgust proneness on the VOCI total score through perfectionism/intolerance for uncertainty (indirect effect β coefficient = 0.04; 95% CI = 0.003 to 0.093). A comparison of indirect effects with pairwise contrasts indicated that the indirect effects were not significantly different from each other.

Fig. 1
figure 1

Mediation model with self-reported scores on the disgust scale (DS) as the predictor variable; self-reported OCD symptoms (VOCI total) as the outcome variable; and self-reported contamination sensitivity (CSS), contamination thought–action fusion (CTAF), and domains of obsessive–compulsive beliefs (P/U = perfectionism/intolerance of uncertainty; I/C = importance of thoughts/control of thoughts; R/T = inflated responsibility for harm/overestimation of threat) as parallel mediators. Numbers represent standardized β coefficients. *p < 0.05. **p < 0.01. ***p < 0.001

Full size image
Table 3 Standardized indirect effects of disgust sensitivity on the total OCD symptoms (N = 149)
Full size table

A second mediation model tested the effect of disgust proneness through cognitive processes on VOCI (contact) contamination (Fig. 2). The direct effect of disgust proneness on VOCI (contact) contamination was statistically significant (unstandardized b = 0.24, SE = 0.11, p = 0.024). The total effect of disgust proneness on VOCI (contact) contamination was significant (β = 0.34, SE = 0.13, p < 0.001). Standardized indirect effects are presented in Table 4. There was a statistically significant positive indirect effect of disgust proneness on VOCI contamination scores via CS (indirect effect β coefficient = 0.17; 95% CI = 0.089 to 0.259). Furthermore, results showed an indirect effect of disgust proneness on the VOCI contamination subscale scores through CTAF that was small yet statistically significant (indirect effect β coefficient = 0.06; 95% CI = 0.006 to 0.128). A pairwise comparisons between the indirect effects suggested that the indirect effect via CS was significantly larger than via the CTAF of thoughts (see Table 2).

Fig. 2
figure 2

Mediation model with self-reported scores on the disgust scale (DS) as the predictor variable; self-reported OCD contamination symptoms (VOCI-CTN) as the outcome variable; and self-reported contamination sensitivity (CSS), contamination thought–action fusion (CTAF), and domains of obsessive–compulsive beliefs (P/U = perfectionism/intolerance of uncertainty; I/C = importance of thoughts/control of thoughts; R/T = inflated responsibility for harm/overestimation of threat) as parallel mediators. Numbers represent standardized β coefficients. *p < 0.05. **p < 0.01. ***p < 0.001

Full size image
Table 4 Standardized indirect effects of disgust sensitivity on OCD contamination symptoms (N = 149)
Full size table

The third mediation model tested cognitive processes as putative mediators of the association between disgust proneness and the VOCI obsessions subscale (Fig. 3). The direct effect of disgust proneness on VOCI obsessions was not statistically significant (unstandardized b =  − 0.02, SE = 0.13, p = 0.87). The total effect of disgust proneness on VOCI obsessions was significant (β = 0.23, SE = 0.15, p = 0.006). Standardized indirect effects are reported in Table 5. Results showed an indirect effect of disgust proneness on the VOCI obsession subscale through the importance/control of thoughts domain (indirect effect β coefficient = 0.14; 95% CI = 0.05 to 0.25). No other mediators emerged as statistically significant.

Fig. 3
figure 3

Mediation model with self-reported scores on the disgust scale (DS) as the predictor variable; self-reported OCD repugnant obsessions symptoms (VOCI-OBS) as the outcome variable; and self-reported contamination sensitivity (CSS), contamination thought–action fusion (CTAF), and domains of obsessive–compulsive beliefs (P/U = perfectionism/intolerance of uncertainty; I/C = importance of thoughts/control of thoughts; R/T = inflated responsibility for harm/overestimation of threat) as parallel mediators. Numbers represent standardized β coefficients. *p < 0.05. **p < 0.01. ***p < 0.001

Full size image
Table 5 Standardized indirect effects of disgust sensitivity on OCD repugnant obsessions symptoms (N = 149)
Full size table

The fourth mediation model included the VOCI checking subscale as the dependent variable (Fig. 4). The direct effect of disgust proneness on VOCI checking was not statistically significant (unstandardized b = 0.13, SE = 0.11, p = 0.24). While the total effect of disgust proneness on VOCI checking was significant (β = 0.25, SE = 0.11, p = 0.002), there was no significant indirect effect of any of the cognitive processes, as CIs for each of these potential mediators included the value 0. Standardized indirect effects are reported in Table 6.

Fig. 4
figure 4

Mediation model with self-reported scores on the disgust scale (DS) as the predictor variable; self-reported OCD checking symptoms (VOCI-CH) as the outcome variable; and self-reported contamination sensitivity (CSS), contamination thought–action fusion (CTAF), and domains of obsessive–compulsive beliefs (P/U = perfectionism/intolerance of uncertainty; I/C = importance of thoughts/control of thoughts; R/T = inflated responsibility for harm/overestimation of threat) as parallel mediators. Numbers represent standardized β coefficients. *p < 0.05. **p < 0.01. ***p < 0.001

Full size image
Table 6 Standardized indirect effects of disgust sensitivity on OCD checking symptoms (N = 149)
Full size table

A fifth mediation model tested the effect of disgust proneness via cognitive processes on VOCI “just right” scores (Fig. 5). The direct effect of disgust proneness on VOCI “just right” was not statistically significant (unstandardized b = 0.02, SE = 0.12, p = 0.86). The total effect of disgust proneness on VOCI “just right” was significant (β = 0.22, SE = 0.16, p = 0.006). Standardized indirect effects are reported in Table 7. There was a positive indirect effect of disgust proneness on VOCI “just right” through perfectionism/intolerance for uncertainty that was small in magnitude yet statistically significant (indirect effect β coefficient = 0.08; 95% CI = 0.007 to 0.16). Furthermore, there was a positive indirect effect of disgust proneness on VOCI “just right” via CTAF that was small yet statistically significant (indirect effect β coefficient = 0.07; 95% CI = 0.02 to 0.14). A pairwise comparison indicated that these indirect effects were not statistically different.

Fig. 5
figure 5

Mediation model with self-reported scores on the disgust scale (DS) as the predictor variable; self-reported OCD just right symptoms (VOCI-JR) as the outcome variable; and self-reported contamination sensitivity (CSS), contamination thought–action fusion (CTAF), and domains of obsessive–compulsive beliefs (P/U = perfectionism/intolerance of uncertainty; I/C = importance of thoughts/control of thoughts; R/T = inflated responsibility for harm/overestimation of threat) as parallel mediators. Numbers represent standardized β coefficients. *p < 0.05. **p < 0.01. ***p < 0.001

Full size image
Table 7 Standardized indirect effects of disgust sensitivity on OCD just right symptoms (N = 149)
Full size table

A final mediation model tested the mediating effects of cognitive processes in the putative association between disgust proneness and VOCI-MC (Fig. 6). The direct effect of disgust proneness on VOCI mental contamination was not statistically significant (unstandardized b = 0.06, SE = 0.20, p = 0.78). The total effect of disgust proneness on VOCI-MC was significant (β = 0.30, SE = 0.25, p < 0.001). Standardized indirect effects are reported in Table 8. There was a positive indirect effect of disgust proneness on VOCI mental contamination through importance/control of thoughts that was small in magnitude yet statistically significant (indirect effect β coefficient = 0.10; 95% CI = 0.04 to 0.20). Furthermore, there was a positive indirect effect of disgust proneness on VOCI-MC via CS that was small yet statistically significant (indirect effect β coefficient = 0.11; 95% CI = 0.04 to 0.20). A pairwise comparison indicated that these indirect effects were not statistically different.

Fig. 6
figure 6

Mediation model with self-reported scores on the disgust scale (DS) as the predictor variable; self-reported OCD mental contamination symptoms (VOCI-MC) as the outcome variable; and self-reported contamination sensitivity (CSS), contamination thought–action fusion (CTAF), and domains of obsessive–compulsive beliefs (P/U = perfectionism/intolerance of uncertainty; I/C = importance of thoughts/control of thoughts; R/T = inflated responsibility for harm/overestimation of threat) as parallel mediators. Numbers represent standardized β coefficients. *p < 0.05. **p < 0.01. ***p < 0.001

Full size image
Table 8 Standardized indirect effects of disgust sensitivity on OCD mental contamination symptoms (N = 149)
Full size table

Observed Power Analysis

Considering that the present study involves data that was collected in the context of another study, a power analysis was conducted to determine the power that we had to detect the effects obtained and our risk of committing type I errors. Power estimates for each model were obtained by performing Monte Carlo simulations with 200 bootstrap samples and 1000 iterations, using the Mplus software. Power estimates for each indirect effect can be found in Table 9, where the % of significant coefficients indicate the percentage of cases for which a given indirect effect was different from zero (Thoemmes et al., 2010). Power estimates for statistically significant effects ranged from 0.37 to 0.98. As such, the results described above should be interpreted in concert with these power estimates.

Table 9 Power estimates (% of significant coefficients) obtained from a Monte Carlo simulation
Full size table

Discussion

The present study examined possible cognitive mechanisms that could explain the relationship between disgust proneness and a range of different OCD symptom domains. Our results were largely consistent with our hypotheses. Specifically, disgust proneness was found to be a significant predictor of OCD symptoms broadly, as well as of each individual OCD symptom domain. This replicates previous research that has found a strong link between disgust proneness and OCD contact contamination symptoms (e.g., Moretz & McKay, 2008; Olatunji et al., 2010a, b), mental contamination symptoms (e.g., Inozu et al., 2021), repugnant obsessions (e.g., Inozu et al., 2021), and ordering compulsions (e.g., Olatunji et al., 2010a, b). Further, it extends our understanding of the relevance of disgust proneness as an affective vulnerability factor for OCD symptoms above and beyond contamination fear and demonstrates its relevance to checking symptoms as well. Across all symptom domains, the inclusion of OBQ belief domains, CTAF, and CS as mediators emerged as significant mediators, demonstrating that existing cognitive mechanisms seem to explain a large and meaningful portion of the relationship between disgust proneness and OCD symptoms. That said, the relative explanatory power of individual mechanisms varied across symptom domains.

Cognitive models of OCD point to particular appraisals and beliefs as being key mechanisms at play in the onset and maintenance of the disorder. Specifically, the Obsessive Compulsive Cognitions Working Group (2005) outlined three belief domains that are associated with a variety of OCD symptoms — importance/control of thoughts, perfectionism/intolerance of uncertainty, and responsibility/overestimation of threat. One of the most robust findings from the present study was the finding that beliefs about the importance/control of thoughts were found to be a significant mediator in the context of both repugnant obsessions and mental contamination. This suggests that individuals who find the experience of disgust particularly aversive may develop beliefs about the importance of controlling thoughts associated with the experience of disgust (e.g., “I must control my thoughts at all costs in order to avoid feeling disgusted”). Holding these beliefs may increase the salience of disgusting thoughts, thus leading to increased frequency of, and distress associated with, thoughts of this nature (i.e., heightened repugnant obsessions). Alternatively, individuals prone to disgust who have intrusive disgusting thoughts in the context of strong beliefs about the importance of controlling thoughts may interpret these thoughts as a violation, leading to heightened feelings of mental contamination. These findings align with past research that suggests that for individuals high in disgust proneness, these OBQ belief domains may exacerbate particular OCD symptomatology (Wheaton et al., 2010),

Contrary to our hypotheses, beliefs about responsibility/overestimation of threat were not found to significantly mediate the relationship between disgust proneness and any OCD symptom domain. Given that this belief domain primarily captures concerns about preventing harm to others, it may tap more into the anxiety-/fear-based affective component of OCD than disgust. Therefore, these beliefs might prove to better explain the relationship between anxiety-based affective vulnerability factors (e.g., anxiety sensitivity) than disgust-based affective vulnerability factors (e.g., disgust proneness (García-Soriano et al., 2016)). While beliefs about perfectionism/intolerance of uncertainty were found to be a significant mediator in the context of “just right” symptoms, the power for this effect was quite low and should be interpreted with caution.

Contamination-related cognitive mechanisms, particularly CS, were also found to be relevant mediators in the model. CS was found to significantly mediate the relationship between disgust proneness and both contact and mental contamination-related OCD symptoms. This replicates past research that has shown its relevance to washing/cleaning behavior in OCD (Rachman et al., 2015; Radomsky et al., 2014). By contrast, CTAF was not a significant mediator in the relationship between disgust proneness and mental contamination and had minimal power as a mediator in the context of contact contamination. These findings suggest that CTAF may not add substantial explanatory power in this context after accounting for other contamination- and OCD-relevant cognitive processes (i.e., CS, importance/control of thoughts). Alternatively, in line with the model proposed by Inozu et al. (2021), CTAF might be better conceptualized as a cognitive vulnerability factor for contamination-related symptomatology than a mediator between disgust proneness and contamination symptoms. Though power was limited, CTAF was found to significantly mediate the relationship between disgust proneness and “just right” symptoms. “Just right” experiences often incorporate magical thinking. For example, a patient with OCD might report tapping their hands until it “feels right” as a way of preventing bad consequences. As such, feelings of disgust might propel “just right” symptoms through the experience of CTAF (e.g., appraisal of contamination-related thoughts as causing actual, physical contamination).

Finally, findings from the present study were the first to demonstrate that disgust proneness significantly predicts checking symptoms to the best of our knowledge. While the connection between these constructs may not be as evident as for other symptom domains (e.g., contamination), this finding may underscore the notion that checking behavior in OCD may be motivated by a wider range of cognitive processes than previously thought. Specifically, checking behavior is prototypically conceptualized as an attempt to prevent harm/a dangerous outcome (e.g., checking locks to prevent a burglary, checking stoves to prevent a fire) or to reduce anxiety. However, these findings demonstrate the overlapping nature of OCD symptom domains and reinforce the idea that people may engage in checking in response to both feelings of disgust and anxiety. For example, someone who comes in contact with a disgust-provoking stimulus, such as dog feces, may engage in checking behavior to confirm whether they got any on their shoes/clothes. Further, someone who experiences repugnant intrusive thoughts about being a pedophile might engage in checking behavior (e.g., excessive internet searches, reassurance seeking from loved ones) to reduce feelings of self-directed disgust. Thus, greater proneness to experience disgust may lead to heightened checking behavior. While these findings lend support for this relationship, none of the cognitive mechanisms examined were found to significantly mediate this association. Therefore, future research should aim to better understand what unique cognitive processes might better explain the relationship between disgust proneness and checking.

A growing body of research has demonstrated the relevance of disgust proneness as an affective vulnerability factor implicated in the onset and maintenance of OCD (Knowles et al., 2018; Olatunji et al., 2011). Accordingly, cognitive strategies directly targeting disgust-related appraisals and beliefs have been shown to bolster the effectiveness of exposure and response prevention (ERP) interventions for contamination-related OCD (Ludvik et al., 2015) and reduce contamination-related avoidance (Wong et al., 2021). The findings from the present study provide additional support for the relevance of disgust proneness across all OCD symptom domains. Notably, several OCD-relevant belief domains were found to mediate this relationship: CS in contamination symptoms; importance of/need to control thoughts in repugnant obsessions; CTAF in “just right” symptoms; and CS and importance of/need to control thoughts in mental contamination. Therefore, to maximize the benefit of interventions for OCD, clinicians might benefit from selecting cognitive strategies that directly target these belief domains depending on the specific symptom profile of their client. For example, incorporating case formulation-driven psychoeducation about and behavioral experiments aimed at modifying these beliefs into interventions may prove to be more effective than traditional ERP on its own.

Limitations and Future Directions

The main limitation with the present study is our small sample. In fact, while half of our significant indirect effects demonstrated robust statistical power (≥ 0.80) to refute the null hypothesis, the other half of our significant indirect effects showed weaker power, such that there is a greater likelihood that these latter effects were found due to chance. Therefore, the current results warrant replication in a larger sample. While previous research has demonstrated the effectiveness of using analogue samples to study OCD phenomena (Abramowitz et al., 2014; Gagné et al., 2018), findings from the present study should be replicated in a clinical sample. The cross-sectional, non-experimental, and self-report nature of the design of the present study also limits our ability to draw definitive conclusions about causality and directionality of these effects. In fact, while cross-sectional mediation sheds light on the associations between different constructs, a prospective design would be better suited to understand the directionality of the effects of disgust proneness and cognitive processes on OCD symptoms. As such, future research would benefit from examining the role of these mediators in an experimental design and incorporating behavioral measures of OCD symptoms. Further, the present study’s sample was comprised entirely of female undergraduate students. Before these results can be generalized to the population at large, they should first be replicated in a more gender-diverse sample. Further, the experience of disgust in OCD is complex, as it entails an affective experience that goes above and beyond a revulsive reaction to contaminants. In fact, one individual with OCD might report feeling disgusted by their intrusive thoughts, while another might say that their hands are “filthy” with a curse that they could spread. Sociomoral disgust, which refers to disgust in the face of moral violations, might be a fruitful avenue for future research, given the ubiquity of morality themes in OCD (Abramovitch et al., 2013). Future research should further explore the experience of disgust in OCD, using a mixed-methods approach involving self-report measures, qualitative analysis, and experimental stimuli, in order to more thoroughly capture disgust as experienced by those with the disorder.

Conclusion

Overall, findings from the present study highlight the importance of disgust proneness as an affective vulnerability factor across OCD symptom domains. Further, they emphasize the role of various cognitive mechanisms in the maintenance of OCD symptoms. Specifically, CS appears to be an important cognitive mediator in the relationship between disgust proneness and both contact and mental contamination symptoms. Similarly, beliefs about the importance/control over thoughts appears particularly relevant in the relationship between disgust proneness and mental contamination and repugnant obsessions. Given that even the most effective treatments for OCD do not work for many individuals with the disorder (Fisher & Wells, 2005), there is a need to improve the potency of our interventions for this population. A better understanding of the role of cognitive mechanisms in the relationship between disgust proneness and specific OCD symptom domains may help us to move toward achieving this goal. Specifically, the findings from the present study lend support to the idea of incorporating interventions that directly target CS and beliefs about importance/control over thoughts when working with individuals who present with contamination concerns and/or repugnant obsessions.