Involvement of fear, incompleteness, and disgust during symptoms of pediatric obsessive–compulsive disorder

Pediatric obsessive–compulsive disorder (OCD) is a heterogeneous and disabling condition affecting 2–4% of children and adolescents [1, 2]. Cognitive behavioral therapy (CBT) and selective serotonin reuptake inhibitors (SSRIs) are first-line treatments but yield sub-optimal effects for a significant proportion of patients [3] and the prognosis for patients who retain clinically significant symptoms of OCD is often poor [4‐7]. There is little consistent information about who is most likely to benefit from these treatments and further improvements in outcomes will require a better understanding of the etiological and maintaining factors that underpin this disorder [8‐10].

While numerous psychological models have been posited to explain the onset and maintenance of OCD, among the most well-researched and influential are cognitive behavioral models of OCD (for a discussion, see [11]). Briefly, cognitive behavioral models view obsessions as transient, intrusions of fearful images or thoughts of the kind that are common in the general population (e.g., the thought that something bad might happen to a loved one while walking to school). If the individual focuses on these intrusions and appraises them such that the intrusion now means the feared outcome is more likely, and the individual is responsible for preventing the feared outcome, there is an increase in fear, distress, and an urge to neutralize this thought and prevent the feared outcome. Over time, such appraisals and the resulting fear response drive the development of a wide range of neutralizing (escape) and avoidance behaviors [11]. From this perspective, OCD and anxiety disorders are similar in that they are driven by exaggerated fears about the probability of harm and maintained by the excessive use of idiosyncratic safety-seeking behaviors.

Although many OCD patients report fear and anxiety as part of their obsessions and compulsions, a large proportion do not. This is one reason why OCD was removed from the anxiety disorders chapter in DSM-5 and placed alongside other conditions characterized not by a single dominant emotional response, but by the presence of compulsivity and repetitive behaviors [12]. Indeed, since Janet described OCD in the beginning of the twentieth century [13], clinicians and researchers have recognized that emotions (or emotion-like states) other than fear are experienced during obsessions and compulsions, and as such may be important to the etiology and treatment of OCD. Examples of such emotion-related experiences include pathological doubt, anger, disgust, and incompleteness or not-just-right experiences (NJRE) [14‐19]. This study concerns itself with the role of fear, incompleteness, and disgust in the experience of OCD symptoms in children and adolescents.

Recognizing that the ritualized and repetitive nature of compulsions, and an often irrational need for (or preoccupation with) symmetry, appears to distinguish OCD from the anxiety disorders, Summerfeldt et al. [15, 20] proposed the Core Dimensions Model of OCD. Central to this model is the notion that both fear (termed harm avoidance in the model) and a felt sense of incompleteness are the core motivational processes underpinning OCD symptoms [16]. The emotional experience of incompleteness refers to an inner sensation of things being ‘not just right’ and an emotional urge to reach a sense of completeness [21]. There is a small but growing body of evidence which suggests that a felt sense of incompleteness, both as an overall trait and during OCD symptoms, is elevated among adults [16] and youth with OCD [22], and that incompleteness, compared to fear, may be uniquely associated with OCD in both children and adults [18, 22, 23]. Furthermore, a recent study found that higher levels of incompleteness during OCD symptoms was associated with poorer OCD outcomes in youth treated for this disorder [24].

Disgust has long been recognized as a potential contributor to the severity and maintenance of contamination-related fears and avoidance behaviors that occur as part of OCD and social and specific phobias [25]. There is a growing body of evidence suggesting that a general proneness or sensitivity for disgust is associated with more severe OCD symptoms, especially those involving the theme of contamination [14, 26, 27]. This association is also found in clinical and non-clinical samples of youth, with changes in disgust sensitivity during OCD treatment being positively correlated with reductions in OCD symptoms [14, 22, 28, 29]. However, as disgust proneness/sensitivity is also correlated with symptoms of anxiety, and as anxiety symptoms also tend to decrease during OCD treatment, the extent to which disgust plays a primary or secondary (to anxiety) role in OCD remains unclear [30].

Taken together, there is increasing evidence to suggest that, in addition to fear, incompleteness and to some extent disgust are elevated among individuals with OCD, and often accompany OCD symptoms, suggesting that multiple emotion-related motivators may be present in OCD. The strength of a multi-emotion based perspective on OCD is that topographically identical symptoms (e.g., excessive handwashing or repetitive checking) may be motivated and/or driven by different emotions. For example, hand washing may be motivated by a desire to reduce the risk of becoming ill, by a strong sense of disgust, or by an urge to reach a sense of completeness in the handwashing ritual. Importantly, it has been argued that these emotion-related motivators reflect not only conceptual or phenomenological differences, but also differences in more basic psychological and neural processes, which could shed new light on the etiology and treatment of OCD [31‐33]. Hence, further studies are needed of the emotions that accompany OCD symptoms, and how these relate to the severity and persistence of the disorder, particularly in children and adolescents. In addition, the relationship between fear, incompleteness, disgust, and OCD symptoms has most often been examined using self-report measures of these emotions at an overall, trait-like level (i.e., a general proneness for fear, incompleteness, and disgust). Although useful, this is not the same as asking patients about the extent to which feelings of fear, disgust, or incompleteness are experienced during or directly involved in their OCD symptoms.

It is important to note that OCD symptoms tend to cluster within several, temporally stable symptom dimensions, with the three most commonly observed being the aggressive, symmetry, and contamination-related dimensions [34, 35]. Studies suggest that these dimensions involve different neural substrates [36, 37] and possibly different etiologies [38]. As such, these dimensions provide a useful framework in which emotion-involvement in OCD symptoms can be studied [39]. Finally, given the overlap at the conceptual, etiological, and phenomenological level between OCD and the anxiety disorders [40, 41], studies examining potential emotion-related motivators in individuals with OCD should include anxiety disordered comparison groups to help distinguish between potentially OCD-specific, anxiety-specific, and transdiagnostic emotional processes.

The present study was undertaken to address the above gaps in the current literature. Our primary aim was to examine the extent to which fear, incompleteness, and disgust accompanied current OCD symptoms in a large and diverse sample of treatment-seeking youth with OCD, as a disorder overall, and in relation to its three major symptom dimensions (aggressive, symmetry, contamination). In a preliminary fashion, and for comparison purposes, we undertook a similar examination with youth seeking treatment for social anxiety disorder (SAD) and generalized anxiety disorder (GAD), and without a diagnosis of OCD. For participants with OCD, we also investigated whether fear, incompleteness, and disgust were more or less associated with certain forms of comorbidity that have been shown to be associated with the severity of OCD, namely hoarding, tics, and autism spectrum disorder [42‐44].

To better understand the onset and maintenance of the heterogeneous symptoms of OCD, emotion-related processes beyond fear and anxiety need to be investigated [53]. The present study was carried out with this aim and addressed several gaps in the extant literature on emotional motivators in OCD, particularly in relation to children and adolescents. First, we moved beyond trait-level assessment of emotion and investigated the degree to which specific emotions occur during symptoms of OCD. Second, we drew upon two previously separate programs of research, the first focusing on fear (harm avoidance) and incompleteness [16], and the second on disgust [54], to conjointly examine these emotions in youth with OCD in comparison to youth with GAD and SAD. Third, for the OCD group, we examined the relationship between these emotions and the presence of tic disorder, autism, and hoarding symptoms.

As expected, and in line with the dominant cognitive behavioral models of OCD [11], participants with OCD reported elevated levels of fear during symptoms within the aggressive, symmetry, and contamination-related dimensions, with the presence of fear being most strongly elevated during symptoms in the aggressive dimension. This is not surprising given that aggressive symptoms are characterized by obsessions about possible harm befalling oneself or a loved one, often paired with checking and/or idiosyncratic rituals to reduce the likelihood of harm. Also consistent with expectations, youth with SAD and GAD reported that fear was the primary emotion experienced during the respective symptoms of these disorders. These results provide preliminary support for the argument that OCD is more similar than different from the anxiety disorders—at least in this age range. However, in the present study (higher) levels of incompleteness during symptoms differentiated youth with OCD (overall and for the aggressive dimension subtype) from those with SAD and GAD. Also, within the aggressive OCD symptom dimension, incompleteness was both elevated and associated with symptom severity. This is the first study to undertake a comparison between OCD and anxiety disorders assessing experience of different emotions during symptoms, as opposed to the use of trait-like measures of emotional proneness given contemporaneously to OCD symptom measures (interview or self-report). The present results provide further support that, and in contrast to the anxiety disorders, OCD is a more heterogeneous condition at the emotion level, which may help explain why OCD symptoms can be so varied in their topography. Although further studies are needed, the present results suggest that incompleteness, and the mechanisms that might underpin this emotion-like state, warrant further study in this age range, and as part of a large program of research to identify factors that may add (and improve upon) current etiological models of, and treatments for, pediatric OCD.

It is important to note that although incompleteness was elevated across all OCD symptom dimensions, the strongest and most unique association was found with symptoms from the symmetry dimension. Indeed, within this dimension, almost all participants with OCD reported moderate to extreme levels of incompleteness during symptoms, and incompleteness, but not fear or disgust, was associated with the severity of symptoms. Thus, in future studies examining the relationship between incompleteness and OCD, it is important that the OCD symptoms be assessed at both the overall (disorder) and at the symptom dimension level.

The limitations of a fear-centric model of OCD were also evident in relation to OCD symptoms within the contamination dimension, where disgust was the emotion that participants most often reported as occurring during these symptoms. Although disgust by some participants was reported during symptoms from the other OCD symptom dimensions, and to some extent during symptoms of SAD and GAD, it was most strongly related to symptoms within the contamination dimension. Previous studies have found a relatively weak relationship between disgust proneness, measured at the trait-level via self-report questionnaires, and the overall severity of OCD symptoms in clinically referred youth and adult samples [22, 30]. The difference between these and the present findings can be partly attributed to our use of an interview-based assessment of emotion involvement during OCD symptoms and of OCD at the dimensional level. The present findings, while preliminary in nature, suggest that disgust may play a clearer role in the contamination symptoms of OCD when this emotion is measured at the state-level (i.e., as an emotion that accompanies symptoms) than at the trait-level. It is also important to note that although disgust was clearly elevated within the contamination symptom dimension, the overall level of symptom variance within this dimension was weakly explained by the presence of fear, incompleteness, and disgust. These results suggest that other emotion-related motivators may play a role in contamination symptoms, or that some complex interaction between several different emotions is present.

As described above, we attempted to address variation in emotion-involvement across subtypes of OCD defined by its major symptom dimensions. Another potentially important subtype is tic-related OCD, which is found in 10–40% of youth with OCD and is associated with differences in gender distribution, onset, heritability, and a poor response to SSRIs [42, 55]. We found no differences in the severity of fear, incompleteness, and disgust in youth with versus without this subtype of OCD. It is also important to consider influences arising from other comorbid disorders. For example, there is ample evidence in the literature of high levels of comorbidity between OCD and autism [56]. Contrasting with the view that incompleteness may help explain this comorbidity [57], we found no evidence of a relationship between incompleteness and the presence versus absence of a comorbid diagnosis of autism spectrum disorder, or to the severity of the young person’s autistic traits. Finally, while hoarding is now its own disorder in DSM-5, it is still possible for hoarding to occur within the context of OCD symptoms [43]. Incompleteness and not just right experiences have been found to be associated with hoarding symptoms in adults with OCD [58] and a small but significant correlation was found also in our study, which adds to the literature about potential mechanisms involved in hoarding symptoms and how these symptoms may be linked to OCD.

Exposure-based CBT is the recommended first-line treatment for pediatric OCD and the rationale for this treatment draws heavily on a fear-centric model in which symptom change is assumed to occur as a result of therapist and patient-guided exposures to fear-eliciting stimuli/situations during which the urge to engage in compulsive behaviors is resisted. It is generally assumed that such exposures (in session and as homework) bring about a gradual extinction or habituation of the fear response and an overall attenuation of OCD symptoms. Recent models have challenged the role of habituation in symptom improvement but still highlight the role of fear expectancies [59]. Given the present results, with incompleteness and to some extent disgust emerging as equally important as fear in relation to pediatric OCD, it is reasonable to ask whether the efficacy of CBT for OCD is partly dependent upon the emotions that underpin symptoms. A study carried out on adults with blood/injury phobias found that disgust appeared to decline more slowly than fear with repeated exposures to distress-provoking stimuli, and that additional exposure sessions may be required [60]. Recently, using a subset of the OCD participants in this study for whom pre- and post-treatment data were available, we found that high levels of incompleteness or disgust during OCD symptoms at pre-treatment predicted a poorer response to treatment involving exposure plus response prevention with or without concomitant pharmacotherapy [24]. We are unaware of any experimental studies examining the effects of exposure on incompleteness-driven OCD. It has been suggested that habit reversal techniques may be needed to improve outcomes [20], but to the best of our knowledge, this is still an empirically untested hypothesis. Whether responsiveness to SSRIs and other pharmacological agents is dependent upon emotion involvement in symptoms is also an area of research that could help better elucidate treatment response in pediatric OCD. More research is needed, not only across different symptom dimensions of pediatric OCD, but also in relation to the emotional processes that underpin these symptoms.

Although the study benefits from the inclusion of a large and diverse sample of youth with OCD, clinical comparison groups, and the use of structured interviews to assess emotion during symptoms, the findings need to be viewed within the context of certain limitations. First, emotions are complex phenomena involving more than a person’s verbal report, and the current findings are based solely on structured interviews wherein patients were asked to rate the degree to which these specific emotions were experienced during symptom episodes. It is possible that the emotions described during the OC-CDI interview induced participants to conceptualize the emotional content accompanying their symptoms according to these descriptions. While an individual’s subjective experience of emotion is a valid and fundamental component in clinical research [61], studies employing psychophysiological, brain imaging, symptom-provocation, and experience sampling methods should be used to build on the results reported here. Novel measures may be used alongside the expanded OC-CDI interview used here, as it shows potential as a valid measure of emotion involvement in OCD. Indeed, the present results further support its construct validity, which previously has been demonstrated in relation to self-reported trait-levels of incompleteness and harm avoidance [50]. Second, the study was carried out with clinically referred children and adolescents and these findings may not extend to non-clinical samples or adults. Furthermore, substantial proportions of youth with OCD in this study also fulfilled diagnostic criteria for SAD and GAD, which obscures inferences regarding group differences. Third, emotional-involvement in symptoms was measured retrospectively, which can introduce certain biases and limit the inferences that can be drawn [62]. Fourth, fear, disgust, and incompleteness are likely only a subset of the emotions involved in OCD. Fifth, our use of the symptom dimension model to assess OCD symptoms may have been overly restrictive. It is possible that emotion-involvement is a more solid ground on which to build a multidimensional model of OCD (see [55]), but this hypothesis could not be thoroughly tested given the design of the present study.

In summary, we found that the emotions that accompany the symptoms of pediatric OCD extended beyond fear, to include incompleteness and disgust, with the level of involvement of these emotions varying as a function of the symptom dimension that characterized the child’s OCD presentation. Importantly, our results showed that incompleteness was as central to OCD as fear, and more specific as it differentiated youth with OCD from those with SAD and GAD. Disgust was also specific to OCD but primarily in relation to symptoms in the contamination dimension. While further studies are needed, the present results suggest that researchers and clinicians alike should consider models of OCD that more clearly account for emotional motivators other than fear.