Does Internal Preoccupation with Obsessive-Compulsive Themes Affect Externally Oriented Functioning in OCD?: Behavioral Results and Clinical Cases

Difficulties with Executive Functioning in OCD

OCD is associated with broad deficits in multiple domains of executive function – memory, attention shifting, conflict resolution, decision making, and response inhibition – although empirical findings remain inconsistent (Greisberg & McKay, 2003; Kashyap et al., 2013; Shin et al., 2014; Snyder et al., 2015). Neurobiological studies have linked these executive difficulties to altered functioning in cortico-striatal circuits (Nakao et al., 2014); abnormalities in these circuits are also associated with symptoms in OCD (Brennan & Rauch, 2017; Graybiel & Rauch, 2000; Maia et al., 2008).

An executive overload model of OCD (Abramovitch et al., 2012) conceptualizes neuropsychological impairments in OCD as a consequence of ongoing attempts to control the automatic processes associated with obsessive thoughts. Effectively, it suggests that individuals with OCD may be too distracted to attend to the task at hand. This executive overload model of OCD is well aligned with predictions of the neurobiological Default Mode Interference hypothesis, DMN-ih (Sonuga-Barke & Castellanos, 2007). DMN-ih focuses on the dynamic interactions between three large scale brain networks: the default mode network (DMN), the central executive network (CEN), and the salience network (SN) (Bressler & Menon, 2010; Deco et al., 2011). The DMN is involved in self-referential processing of internal states, and is generally deactivated during externally oriented tasks, which instead recruit the CEN (Alves et al., 2019). The SN modulates the DMN↔CEN switch during the rest-task transition (Goulden et al., 2014). DMN-ih posits that the ability of the SN to control the DMN during externally oriented activities is impaired in some individuals, leading to continuous ongoing self-referential processes, such as internally oriented obsessions or mental compulsions in OCD, that distract them from effectively processing external stimuli (Alves-Pinto et al., 2019). DMN-ih is not specific to OCD; other internally oriented processes, such as rumination in depression, may similarly affect performance on externally oriented tasks.

Both the executive overload model of OCD and DMN-ih imply that if a targeted intervention can help individuals with OCD “snap out” of their overfocus on self-referential processes (i.e., give them a “wake-up call”), then their performance on externally oriented tasks should improve. We test this prediction using the revised Attention Network Test, ANT-R (Fan et al., 2009). Next, we present two clinical vignettes that are consistent with predictions of both models. We conclude with a discussion of broader implications of our results.

Testable hypotheses

The revised Attention Network Test, ANT-R (Fan et al., 2009), builds upon the Three Networks Attentional System model put forward by Posner and Petersen (1990). This model distinguishes among alerting, orienting, and executive attentional networks. The alerting network supports overall awareness and readiness to receive information and respond; the orienting network is concerned with orienting of attention in space and time; the executive attention involves higher level cognition and supports conflict detection and resolution. The alerting network is the most basic attentional network and enables functioning of the other two systems. Importantly, the contribution of alerting to the efficiency of the other networks is not linear, but rather exhibits diminishing marginal return. Thus, when the executive network is operating inefficiently, an alerting stimulus will enhance executive performance, but when the executive network is functioning optimally, the benefit from an alerting stimulus is minimal (Figure 1).

• Download figure
• Open in new tab

Figure 1. Hypothesized effects of OCs on the efficiency of alerting and executive attention networks.

The concave grey line reflects that the positive impact of the efficiency of alerting network on the efficiency of the executive network exhibits diminishing marginal return. The higher efficiency of alerting network is associated with the higher efficiency of executive network, but with the lower additional executive benefits due to the additional alerting benefits. ANT-R (Fan et al., 2009) is calibrated for healthy individuals to be able to perform at a high level of efficiency and, consequently, with very small room for additional improvement in efficiency of the executive network. An executive overload model of OCD (Abramovitch et al., 2012) predicts that internally oriented OCs reduce individual ability to respond to external stimuli (i.e., reduce efficiency of alerting network); this leads to two testable hypotheses. H1: Responsiveness to the external stimuli (target, blue arrows, and cue, red arrows, in ANT-R) is negatively associated with severity of OCs and, thus, is lower in OCD. H2: Efficiency of conflict processing is negatively associated with severity of OCs and, thus, is lower in OCD, and additional executive benefits due to additional alerting benefits are positively associated with severity of OCs and, thus, are higher in OCD.

The ANT-R was designed to probe the efficiency of alerting, orienting, and executive control attention networks, as well as interactions among them, by comparing reaction times (RT, msec) during trials with correct responses, averaged across task conditions. The task was calibrated for participants from non-clinical populations to respond correctly on approximately 80%-90% of trials; faster correct responses are interpreted as indicating higher efficiency of attentional processes (Fan et al., 2009; Mackie et al., 2013).

The ANT-R evaluates how quickly individuals correctly resolve two executive/cognitive control tasks, Simon (Simon & Wolf, 1963) and Flanker (Eriksen & Eriksen, 1974) conflicts. Both tasks require subjects to indicate, with a button press, the direction the middle in a row of five arrows is pointing. On Simon trials, there is conflict between the location of the correct response button and the side of the screen where the stimulus appears (e.g., a left-pointing middle arrow requires to press the button on the left side of the keyboard but appears on the right side of the screen). On Flanker trials, there is conflict between the direction of the middle arrow and that of the adjacent ‘flanking’ arrows.

In the ANT-R, on some trials, participants receive a warning – a “wake-up call” – a few hundred milliseconds before stimulus onset (Figure 2). Participants tend to respond correctly more quickly when they received such warnings; this is interpreted as improvement in alerting due to the externally orienting cue. Furthermore, participants tend to be more efficient in resolving Simon and Flanker conflicts on trials when they receive “wake-up calls” (a positive alerting x executive interaction); this is interpreted as an impact of alerting benefits on the efficiency of executive attention. Note that this effect depends on both the alerting effect and the baseline efficiency of executive attention (see Figure 1).

• Download figure
• Open in new tab

Figure 2. Schematic presentation of components of the revised Attention Network Test (ANT-R, Fan et al., 2009), relevant to study hypotheses.

ANT-R consists of 4 sequential phases. Phase I, Cue: in the beginning of some trials, two cue boxes flash for 100ms (double cue, DC); no boxes appear during some other trials (no cue, NC). Phase II, Variable interval, VIT: a fixation cross appears in the middle of the screen for 0, 400, or 800ms. Phase III, Target: five arrows are presented on the left or the right side of the fixation cross for 500 ms. Participants are asked to indicate where the middle arrow (target) points by pressing key “A” (the left side of the keyboard) if it points to the left or key “L” (the right side of the keyboard) if it points to the right. During congruent Simon (cS) trials, the arrows and the correct response button are located on the same side; during incongruent Simon (iS) trials, the arrows and the correct response button are located on opposite sides; during congruent Flanker (cF) trials all five arrows point in the same direction; during incongruent Flanker (iF) trials, the target arrow and the surrounding arrows point in opposite directions. Overall, there are four types of trials: cF/CS, cF/iS, iF/cS, and iF/iS. Phase IV, Inter Trials Interval, ITI: A fixation cross appears in the middle of the screen for 2000, 4000, 6000, 8000, 10000, or 12000 ms (mean = 4000 ms).

Note. Picture depicts stimuli for the correct response ‘L’, a key located on the right side of the keyboard.

We predict that overfocus on internal processes is associated with a reduced ability to attend to external stimuli. In OCD, such overfocus maybe linked to the severity of OCs, particularly int-OCs. In the context of ANT-R, this suggests that both readiness to respond to the target stimuli (Figure 1, blue arrows) and the benefit provided by an externally orienting cue (Figure 1, red arrows) should be negatively associated with severity of OCs. This leads to two sets of testable hypotheses.

First, we anticipate an improvement in RT during correct trials following “wake-up calls” (i.e., alerting benefits) in all study participants, but we anticipate that these improvements should be reduced in OCD, and negatively associated with severity of OCs. Given the significant heterogeneity of OCs, we anticipate that some of the OC dimensions maybe linked to reduced alerting benefits more than others.

H1: Alerting benefits are lower in OCD than in healthy controls.

H1a: Association of alerting benefits with severity of OCs varies across OCD symptom dimensions.

Second, recall that ANT-R is calibrated for participants from non-clinical populations to perform at a high level of efficiency during Simon/Flanker conflict processing (see Figure 1). If readiness to respond to target stimuli is negatively associated with severity of OCs, then efficiency of Simon/Flanker conflict processing is also negatively associated with severity of OCs and is expected to be lower in individuals with OCD than in healthy control group. However, alerting benefits (even if they are reduced) may lead to a larger improvement in efficacy of conflict processing in OCD participants than in healthy controls (see Figure 1). Given the significant heterogeneity of OCs, we anticipate that some of the OC dimensions maybe linked to improvement in efficacy of conflict processing more than others.

H2: Alerting benefits result in stronger improvement in efficiency of executive attention in individuals with OCD than in healthy controls.

H2a: The impact of alerting benefits on efficiency of executive attention varies across OCD symptom dimensions.

Note that hypotheses H2/H2a depend on the assumption that healthy controls operate close to maximum efficiency during ANT-R (i.e., on the flat portion of the curve in Figure 1); if this assumption is violated, then stronger alerting benefits in healthy controls could result in greater improvements in efficiency of executive attention in healthy controls than in individuals with OCD.

Potential confounds

We control for effects of two potentially important confounds, symptoms of attention deficit-hyperactivity disorder and depressive symptoms. Both are common comorbidities of OCD (Pallanti et al., 2011) that may contribute to neuropsychological deficits (Brem et al., 2014; Moritz et al., 2001). For instance, studies that used ANT-R reported that ADHD affects alerting and executive attention (Arora et al., 2020; Berger & Posner, 2000) and that depression affects executive attention (Wang et al., 2020).

Participants

All procedures were approved by the Yale University Human Investigation Committee. A total of 97 unmedicated participants were recruited through the Yale OCD Research Clinic between 2012 and 2020 by posting flyers in the local community as a part of a larger series of investigations (Ma et al., 2021; Pushkarskaya et al., 2019; Pushkarskaya et al., 2015). The sample included 43 unmedicated individuals with clinically significantly OCs (OCD) and 54 individuals with no psychiatric diagnosis (HC). All provided written informed consent before any procedure.

Diagnoses were established by the Mini International Neuropsychiatric Interview and confirmed by doctoral-level clinicians. Clinically significant OCs were defined as 16 or higher on the Yale-Brown Obsessive Compulsive Scale (Y-BOCS; Goodman et al., 1989a,b). OCD was the primary clinical diagnosis in all patients; comorbidities included depression, hoarding disorder, panic disorder, social phobia, agoraphobia, post-traumatic stress disorder, and generalized anxiety disorder.

Procedure

Participants completed a demographic questionnaire, the Kaufman Brief Intelligence Test (Kaufman & Kaufman, 2004), the revised Obsessive-Compulsive Inventory (OCI-R; (Foa et al., 1998), the Beck Depression Inventory II (BDI-II; Beck et al., 1996), the Adult ADHD Self-Report Scale (ASRS; Adler et al., 2006), and the revised Attention Network Test (ANT-R). They were compensated for their time.

Measures

The Revised Attention Network Test

The ANT-R is described in detail in Fan et al. (2009). Participants are presented with five arrows (the stimulus) on the right or the left side of the screen and are asked to press a button (on the left or on the right side of the keyboard) corresponding to the direction of the middle arrow (the target; see Figure 2). Shortly before the stimulus appears, a cue (two lighted square frames) is presented during some trials (double cue), while during some trials no cue is presented (no cue). There are also trials on which a single pre-stimulus cue is presented, on the left or right side of the screen, but they are not analyzed here. The button for the correct response is located on the same side as the stimulus on some trials (congruent Simon) and on the opposite side on other trials (incongruent Simon). The target arrow points in the same direction as the remaining 4 arrows on some trials (congruent Flanker) and in the opposite direction on others (incongruent Flanker). Participants tend to respond correctly to incongruent conditions (direction/location) more slowly than to congruent conditions (Flanker/Simon; Hübner and Töbel, 2019), since the former require more executive attention effort.

Participants practiced the task for three minutes. After they indicated they understood the task, they were invited to begin the experiment.

Behavior-based measures

ANT-R measures are computed based on reaction time (RT, ms) on trials on which participants responded correctly.

Alerting benefits (“Alerting”) are measured as:

A higher alerting value indicates greater improvement in performance on the task following a “wake-up call” (i.e., faster correct responses).

Executive benefits (“Executive”) are quantified by two measures:

A higher executive value indicates higher costs of Simon/Flanker conflict processing and implies greater inefficiency of executive attention. In this paper we do not analyze Simon/Flanker interaction.

Effects of Alerting on Executive Attention (“Alerting x Executive”) are quantified by two measures:

More positive values indicate stronger improvement in efficiency of conflict processing (Simon or Flanker) following a “wake-up call” and imply higher baseline degree of inefficiency of executive attention.

Analyses

To test H1 and H2, while controlling for potential confounds, we used a subset of our participants who (i) completed ANT-R, demographic questionnaire, Kbit, BDI-II, and ASRS, and (ii) allowed us to match OCD and HC on age, gender, and IQ (matched subsample).

To test H1a and H2a, while controlling for potential confounds, we used a subset of our participants who completed ANT-R, demographic questionnaire, Kbit, BDI-II, ASRS, and OCI-R (OCI-R subsample). We ran a series of linear regression analyses with OCI-R subscales as dependent variables and IQ, age, BDI-II, and ASRS as covariates. The standardized residuals from these regressions (OCI-R subscales stdres) strongly correlated with respective OCI-R subscales but did not significantly correlate with potentially confounding variables (Appendix B). Thus, associations of these standardized residuals with ANT-R measures are independent of the potential associations between IQ, age, BDI-II, and ASRS and ANT-R measures. Appendix E reports results from analyses that do not control for these potential confounding effects.

To test H1 and H1a, we used a series of univariate ANCOVAs with alerting as the dependent variable. For H1, the between-subject factor was diagnosis (OCD/HC), and IQ, age, BDI-II, and ASRS were covariates. For H1a, OCI-R stdres subscales (washing, obsessing, hoarding, ordering, checking, and neutralizing), IQ, age, BDI-II, and ASRS were covariates. All covariates were standardized.

To test H2 and H2a, we used a series of repeated measures ANCOVAs with Alerting x Executive as the dependent variable and type of conflict (Simon/Flanker) as the repeated measure. For H2, the between-subject factor was diagnosis (OCD/HC), and IQ, age, BDI-II, ASRS, and alerting were covariates. For H2a, OCI-R subscales stdres (washing, obsessing, hoarding, ordering, checking, and neutralizing), IQ, age, BDI-II, ASRS, and alerting were covariates. All covariates were standardized.

To ensure that assumptions of ANOVA were not violated, we examined the residuals from each ANCOVA for (near) normality (using the Shapiro Wilks test and evaluating skewness) and for the presence of outliers (using z-scores ≥ |3| criteria).

Sample and subsamples

97 unmedicated participants (43 OCD, 54 HC) completed ANT-R and the demographic questionnaire; 96 of them additionally completed Kbit; 87 of them additionally completed BDI-II and ASRS; 84 of them additionally completed OCI-R (Appendix A). Two subsamples of these participants were included in two sets of analyses, as detailed below.

35 OCD and 52 HC completed ANT-R, the demographic questionnaire, Kbit, BDI-II, and ASRS. Reliabilities for BDI-II and ASRS were excellent (Cronbach’s α = 0.94 and α = 0.87, respectively). From these individuals, we selected a subset of 35 OCD (12 males) and 46 HC (23 males) to match the two diagnostic groups on age, gender, and IQ (matched subsample, to test H1 and H2; Table 1).

32 OCD and 51 HC completed ANT-R, the demographic questionnaire, Kbit, and all three clinical scales of interest (OCI-R subsample, to test H1a and H2a; Table 1). Reliability for OCI-R was excellent (Cronbach’s α = 0.92), and for OCI-R subscales ranged from good (Cronbach’s α = 0.75, hoarding) to excellent (Cronbach’s α = 0.93, for washing). OCI-R scores were skewed toward 0 (Shapiro Wilks ps <0.001, skewness >1.03), since many healthy individuals self-reported minimal to no OCs. On average, individuals with OCD scored higher than did HC (OCD: mean 24.84 ± 11.73; HC: mean 5.35 ± 6.00; Mann-Whitney U p < 0.001); but OCI-R distributions were broad and overlapping across diagnostic groups (Appendix A). Some individuals with OCD self-reported relatively low scores on OCI-R (OCI-R < 10), despite clinically significant OCs (Y-BOCS ≥ 16, see Participants), which could indicate a tendency to minimize their symptoms.

As expected, BDI-II and ASRS scores significantly correlated with OCI-R total scale and subscales, but not with their standardized residuals (Appendix B). Most OCI-R subscales also correlated, as did their standardized residuals (Appendix B), but these relations were not strong enough to raise multicollinearity concerns (Spearman’s ρ ≤ 0.60, Dormann et al. (2013)).

H1, H1a: Alerting Benefit and severity of OCs

In the full sample, alerting benefits were positive (except for 3 OCD and 2 HC) and normally distributed (Shapiro Wilks p = 0.29).

As predicted (Figure 1), in the matched subsample, OCD tended to respond qualitatively more slowly than HC (median for OCD: RT_{no cue} = 676±117SD, RT_{double cue} = 626±106SD; median for HC: RT_{no cue} = 651±108SD & RT_{double cue} = 618±97SD), with accuracy being high and comparable across two groups (median for OCD = 0.97±0.13SD, median for HC = 0.96±0.07SD; Mann-Whitney U p = 0.36). RT improvement from the no cue condition to the double cue condition was qualitatively smaller in OCD than in HC (mean for OCD = 38±30SD; mean for HC = 46±33SD). For details see Appendix C. However, there was significant within-group variability (see above) and, in contrast to H1, univariate ANCOVA in the matched subsample did not reveal a significant effect of OCD diagnosis on alerting benefits (Table 2). The residuals from this ANCOVA were normally distributed (Shapiro-Wilk p = 0.57).

Consistent with H1a, univariate ANCOVA in the OCI-R subsample (Table 3) revealed that, when we control for potential confounds, alerting benefits were negatively associated with neutralizing (neutralizing stdres: β = -8.77, F = 4.28 p = 0.04, η² = 0.06 medium effect size); we also observed a trend toward negative association of alerting with obsessing (obsessing stdres β = -8.18, F = 2.79 p = 0.10, η² = 0.04 small-to-medium effect size) and positive association of alerting with ordering (ordering stdres β = 10.90, F = 3.93 p = 0.052, η² = 0.05 small-to-medium effect size). The residuals from this ANCOVA were near normally distributed (Shapiro-Wilk p = 0.03, skewness = 0.37, no extreme outliers).

None of the control covariates (age, IQ, BDI, and ASRS) significantly related to alerting in either ANCOVA.

Overall, these results provide only trend-level support for H1 but are consistent with H1a and with our expectations that alerting benefits are negatively associated with severity of int-OCs. Analyzing OCD participants only (excluding controls) revealed similar patterns (Appendix C).

H2, H2a: Impact of Alerting Benefit on Executive Attention and severity of OCs

In the full sample, on average, RT costs to successfully process Flanker conflict were higher than RT costs to successfully process Simon conflict, suggesting that Flanker is a more difficult cognitive control task. OCD took longer than HC to process both conflicts during no cue trials (OCD: RT_{no cue, Simon cost} = 18 msec, RT_{no cue, Flanker cost} = 163 msec; HC: RT_{no cue, Simon cost} = 8 msec, RT_{no cue, Flanker cost} = 153 msec); OCD performance from no cue trials to double cue trials improved more than that of HC (OCD: Alerting x Simon = 20 msec, Alerting x Flanker: 12 msec; HC: Alerting x Simon = 4 msec, Alerting x Flanker: 2 msec).

In the matched subsample, repeated measures ANCOVA revealed that the effect of alerting on executive attention was significantly stronger in OCD (main effect of OCD: F = 12.24 p < 0.001, η² = 0.14, large effect size; Figure 3, Table 4). This effect did not differ significantly between Simon and Flanker processing (OCD x conflict type: F = 0.81, p = 0.37, η² = 0.01, small effect size), suggesting that improved readiness to respond led to broad improvement in executive attention in individuals with OCD (H2).

• Download figure
• Open in new tab

Figure 3 Effects of improvement in readiness to respond due to externally orienting cues (i.e., Alerting benefits) on efficacy of Simon (Alerting x Simon) and Flanker (Alerting x Flanker) conflict processing.

Estimated marginal means from Repeated Measures ANCOVA conducted on data from the matched subsample (43 individuals with OCD and 43 healthy individuals); see Methods for details.

Note: OCD = individuals with Obsessive-compulsive disorder, HC = healthy individuals. Significance levels *** - p < 0.001; Error bars 95% Cl.

Individual responsiveness to externally orienting cues (i.e., alerting benefit) was positively associated with improvement in Flanker conflict processing (β = 17.18, t = 3.22 p = 0.002, η² = 0.12, medium-to-large effect size) but not with improvement in Simon conflict processing (t = 0.25 p = 0.80, η² = 0.001; alerting x conflict type interaction F = 7.06 p = 0.01). This is consistent with our earlier observations that Flanker maybe a more difficult conflict to process: our participants are likely to operate at a lower level of efficiency, and, thus, changes in effectiveness of Flanker processing are more sensitive to changes in alerting level than in Simon processing (see Figure 1).

Effects of control variables, age, IQ, and severity of depressive symptoms did not have significant effects in this subsample. Severity of ADHD symptoms negatively correlated with alerting x executive at trend level (F = 3.77 p = 0.06, η² = 0.05 small-to-medium effect size), suggesting that impact of alerting on executive attention may be reduced in individuals with more severe ADHD symptoms. The residuals from this ANCOVA were normally distributed (Alerting x Simon: Shapiro-Wilk p = 0.84, Alerting x Flanker: Shapiro-Wilk p = 0.62).

In the OCI-R subsample, severity of washing symptoms positively correlated with the effect of alerting in both Simon and Flanker conflicts (washing stdres: main effect F = 8.86 p = 0.004, η² = 0.12, medium-to-large effect size; washing x conflict interaction: F = 0.26, p = 0.61, η² = 0.004; Table 5).

As in the matched subsample, individual responsiveness to externally orienting cues (i.e., alerting benefit) was positively associated with improvement in efficiency of Flanker conflict processing (β = 15.09, t = 2.50 p = 0.015, η² = 0.08, medium-to-large effect size) but not with improvement in Simon conflict processing (t = -1.02 p = 0.31, η² = 0.014 small effect size; alerting x conflict interaction F = 5.85 p = 0.02). Effects of control variables, age, IQ, and severity of depressive and ADHD symptoms did not have significant effect in this subsample.

The residuals from this ANCOVA we were (near)normally distributed (Alerting x Simon: Shapiro-Wilk p = 0.50, Alerting x Flanker: Shapiro-Wilk p = 0.07, no extreme outliers).

In exploratory analyses, we examined whether gender modulates any relations of interest in univariate and repeated measures ANCOVAs detailed above. None of the significant gender effects were detected (ps > 0.10).

Summary of the behavioral results

Overall, these behavioral results are consistent with our hypotheses and indicate that individuals with OCD tend to improve their performance on cognitive control tasks (Simon and Flanker) when they receive an externally orienting cue (a “wake-up call”). Individuals with more severe int-OCs (neutralizing and obsessing symptoms) have more difficulty “snapping out” from the internal preoccupation when prompted to attend to external stimuli. In contrast, individuals with more severe cont-OCs appear to be able to benefit the most from an externally orienting cue during cognitive control tasks.

Vignette #1: Contamination OCD

Patient A was afraid of contracting a bloodborne pathogen such as Hepatitis B/C or AIDS through contact with “microscopic blood particles” in his environment. Patient A used barriers when he needed to touch something in public if he could not wash his hands immediately after. He wore long pants/shirts year-round, even during the summertime and despite overheating, to reduce risk of exposure. If something outside of his home touched his skin, Patient A would wash the area for 30 minutes. He constantly “tracked” what others touched to determine if any contaminated items had come in contact with things he would need to touch. He required his parents to buy Clorox wipes so that he could sanitize his phone and bag when he got home. He did not allow family members in his room for fear they might contaminate his “only safe space.” His OCD impaired his ability to attend school consistently and resulted in interpersonal difficulties with his friends and family.

Treatment

Patient A found it difficult at first to remember to challenge his OCD when he started washing, especially when he was not in the clinic. To help cue Patient A to his exposure goals, Patient A and his therapist created visual cue cards using his favorite memes to catch his eye and included captions that reminded him of his exposure challenges (reducing washing time/steps and resisting tracking others). Patient A placed these cards on his bathroom mirror at eye level so he could not miss them, and on his desk next to his clock and computer. These cues helped Patient A be more consistent with his exposures and increase response prevention by redirecting his attention towards his exposure tasks.

Vignette #2: Scrupulosity OCD

Patient B’s OCD focused on religious themes, and she spent nearly all day engaged in mental compulsions. She was afraid of accidentally practicing another religion and believed this would result in her eternal damnation. She repeated prayers and phrases both in her head and under her breath and felt compelled to repeat them until she recited them perfectly. Patient B described her obsessions and compulsions as automatic and uncontrollable. She stated, “sometimes I don’t even realize I am doing it until someone asks me what I am whispering. It is like they are on a never-ending loop. I never feel like I have more than a few seconds at a time without having to do one because a thought comes into my mind.” Patient B’s intrusions were nearly constant. This, in combination with the automatic nature of her compulsions, made exposure therapy particularly difficult for her to engage in.

Treatment

For Patient B, building awareness of when she experienced obsessions and compulsions was the initial target of treatment. To help her build this awareness, Patient B’s therapist asked her to practice standing up out of her chair and saying a coping statement out loud (ex. “Maybe [that’s true/will happen], maybe not. Either way I am not going to do anything. Let’s see what happens.”) whenever she noticed an obsession. This reminded Patient B to start practicing a new pattern of thinking and acted as a competing response, because she was unable to recite prayers while saying her coping statement out loud. The purpose of standing up was to help Patient B break her obsessive-thought loop; it required several purposeful actions on Patient B’s part which helped her separate from her thoughts and focus on something external (the feeling of movement, seeing her therapist stand up with her each time) and played a function of externally orienting cue (a “wake-up call”). Patient B practiced this intervention for 10-minute intervals multiple times a day. She described her thoughts becoming “less sticky” with repeated practice. Once she could resist compulsions more readily, she was able to start new exposures that relied less on competing responses or grounding techniques and more on mindfully observing her experience.