What cognitive deficits are associated with OCD?
Research has been successful in characterizing the underlying neurobiology of OCD, and cognitive abnormalities associated with the disorder have been documented within the domains of basic learning processes, attention, and executive functioning, each of which arguably is requisite for intact decision making. We will begin by highlighting abnormalities in basic associative learning processes that have been observed in OCD and subsequently will extend our discussion to higher level cognitive processes.
Basic Associative Learning in OCD
The idea that pathological anxiety might develop initially via basic associative learning mechanisms was advanced largely by Mowrer (1939), who proposed a two factor theory about anxiety development. Mowrer (1939) hypothesized that during an initial learning stage, anxiety develops via classical conditioning processes. This occurs when a neutral cue becomes associated with fear due to a negative event. Subsequent re-exposure to this previously neutral cue reactivates the somatic state of anxiety. During a second stage of learning, the newly acquired anxiety is maintained and facilitated through operant reinforcement, such that the fear and anxiety aroused by exposure to the anxiety-eliciting cue is reduced by engaging in negatively reinforcing, stereotyped behaviors. The anxiety reduction obtained by engaging in these responses is rewarding which increases the likelihood that these operant responses will be repeated in the future. These stereotyped behaviors often are performed automatically and never are subjected to a reality test; thus, the individual never realizes that engaging in these behaviors is unnecessary.
Tracy, Ghose, Stecher, McFall, and Steinmetz (1999) have suggested that individuals who later develop OCD might be biologically-predisposed to form associations rapidly between neutral environmental stimuli and anxiety. Furthermore, these same individuals might learn quickly that engaging in compulsive behaviors can reduce this conditioned anxiety. Thus, an enhancement of basic associative learning processes might mediate the development and maintenance of OCD. This theory seems intuitively plausible; however, until recently, it had received little empirical attention. Converging lines of evidence now support this learning-based behavioral conceptualization of OCD.
Tracy et al. (1999) showed that individuals reporting elevated OCD characteristics acquired conditioned eyeblink responses to auditory stimuli predicting corneal air puffs more quickly than controls on a simple classical conditioning task. Interestingly, enhanced classical conditioning was observed only on tasks completed under low attentional load. Using an auditory discrimination conditioning paradigm, Mortimer (2002) subsequently found that OCD-like individuals exhibited a significantly higher percentage of conditioned responses to reinforced stimuli (CS+) over training blocks compared to healthy controls. In contrast, Seay, McFall, and Steinmetz (2005) found no evidence of facilitated discrimination conditioning in an OCD-like population when using visual stimuli. However, they did observe that although individuals in the OCD-like group exhibited normal discrimination between stimuli over the first half of training, discriminability between reinforced and non-reinforced stimuli became impaired relative to individuals in a control group midway through training. This impaired discriminability was due to a reduction in the frequency of CS+ responses. Comparable group performance was re-established during later training blocks.
It is unclear how best to reconcile these findings, given the methodological variations among studies (i.e., differences in stimulus modalities, differences in pure acquisition vs. discrimination processes). However, it appears generally that under simplified task conditions, OCD-like individuals exhibit enhanced associative learning. In more challenging contexts, however, associative learning either is unremarkable or impaired. Tracy et al.’s (1999) observation that OCD-related enhancement of basic associative processes depends partly on attentional load supports this conceptualization.
An independent line of evidence that also implicates a role for learning in OCD can be derived from the clinical treatment literature. The most effective treatment for OCD is exposure and ritual prevention cognitive behavioral therapy (EX/RP or ERP), which is based upon general extinction principles (Hiss, Foa, & Kozak, 1994). Consistent with Mowrer’s (1939) two factor theory, the theoretical basis for ERP treatment is that repeated exposure to anxiety-provoking stimuli without the opportunity to engage in operant rituals weakens the associative strength between the stimulus and its response (and ultimately between the stimulus and conditioned anxiety). Over the course of treatment, individuals with OCD learn incrementally that the anxiety associated with obsessive thoughts habituates naturally and that the compulsive acts they have been using to reduce their anxiety are, in fact, unnecessary. EX/RP treatment has been shown to be associated with a robust therapeutic response.
Taken together, converging evidence derived from basic associative learning methodologies and the clinical literature provides support for Mowrer’s (1939) two factor theory about how anxiety might develop and be maintained. These findings suggest that abnormalities in basic associative learning processes might influence the development of OCD, but in a way that appears to be mediated by attention.
Attention and Attentional Inhibition in OCD
What do we know of attentional alterations in OCD? Much empirical support for altered attentional processes in OCD can be derived from a review of the literature (see Muller & Roberts, 2005). We begin by examining evidence from several classic attentional paradigms, including the emotional Stroop, dot probe, and dichotic listening tests.
The emotional Stroop test is a neuropsychological task that is used to measure attentional biases by contrasting the amount of time required to process emotionally relevant content vs. emotionally neutral content. Lavy, van Oppen, and van den Hout (1994) demonstrated that, in comparison to controls, individuals with OCD selectively attend to disorder-specific negative cues (e.g., “filthy”) but not to disorder-specific positive cues (e.g., “clean”), generally positive cues (e.g., “love”), or generally negative cues (e.g., “hate”). Similar findings have been reported by Foa, Ilai, McCarthy, Shoyer, and Murdock (1993). Direnfeld (2003) demonstrated that this bias varies as a function of treatment by showing that Stroop interference scores for disorder-relevant negative stimuli are reduced significantly post-treatment. Moreover, patients who exhibited the greatest reduction in Stroop interference scores also demonstrated the greatest symptomatic improvement.
Foa and McNally (1986) had patients with OCD complete a dichotic listening task in which they were instructed to shadow one of two aurally-presented passages while simultaneously monitoring the audio content for particular target words (i.e., either fear-relevant or neutral). Foa and McNally (1986) found that fear-relevant targets were identified more readily than neutral targets before, but not after, treatment. Similar findings also emerged from an analysis of participants’ skin conductance responses: skin conductance responses following exposure to fear-relevant targets were greater than responses to neutral targets before, but not after, treatment. Foa and McNally (1986) concluded that OCD is characterized by elevated attention to disease-relevant cues in the environment.
Tata, Leibowitz, Prunty, Cameron, and Pickering (1996) have documented similar OCD-relevant attentional biases using a dot probe task in which participants were presented with two word stimuli simultaneously, one of which was emotionally-relevant and the second of which was neutral. After revealing these words, a target was presented in one of the two positions previously occupied by the words. Shorter response latencies to probes that appeared in positions previously occupied by emotionally-relevant words, relative to probes that replaced neutral words, were considered to be indicative of selective attention for the emotional content of the words. Tata et al. (1996) found that individuals with OCD who reported contamination-related obsessions exhibited greater contamination-related attentional biases compared to high and low trait anxiety controls. In addition to displaying hypervigilance in relation to OC-specific content, individuals with OCD exhibited a larger “interference effect” on probe trials occurring after trials containing threat words, relative to trials containing only neutral words. This implies that the presence of a threat word contributed to slower reaction times on subsequent trials. Tata et al. (1996) suggested that this interference effect might represent the disruption of information processing that might occur after exposure to emotionally-negative stimuli.
OCD also has been associated with impaired attentional inhibition on tasks of negative priming, in which stimuli previously used as distractors subsequently are used as targets. Generally, this type of switch results in increased response latencies. Individuals with OCD, however, exhibit decreased response latencies relative to controls and to individuals with panic disorder or social phobia (e.g., Enright & Beech, 1990, 1993a, 1993b; McNally, Wilhelm, Buhlmann, & Shin, 2001). Muller and Roberts (2005) have interpreted these findings as evidence that individuals with OCD are impaired at attending preferentially to relevant stimuli and ignoring (i.e., actively inhibiting) irrelevant stimuli. Tolin, Abramowitz, Przeworski, and Foa (2002) conducted an interesting series of experiments that showed that, in contrast to nonanxious and anxious controls, individuals with OCD experienced a paradoxical increase in the frequency of thoughts that they were instructed to actively suppress. Furthermore, on a lexical decision task that included non-words, non-target words, and a (suppressed) target word, individuals with OCD exhibited decreased decision latencies for suppressed words, relative to anxious and non-anxious controls. Tolin, Abramowitz, Przeworski, and Foa (2002) suggested that disrupted cognitive inhibition, as evidenced by altered performance on these tasks, might support the development of clinical obsessions.
Relatedly, Bannon, Gonsalvez, Croft, and Boyce (2002) found that individuals with OCD made significantly more errors, relative to individuals with panic disorder, on a Go/No-Go task, a task of behavioral inhibition. Using a similar task, Herrmann, Jacob, Unterecker, and Fallgatter (2003) failed to detect a significant difference in error rates between individuals with OCD and healthy controls; however, the performance of the OCD group was characterized by significantly faster reaction times. Reduced frontal activity, which has been associated with impaired inhibitory processes, was observed in the OCD group during the No-Go condition, and this activity was negatively correlated with OCD symptomatology.
Executive Processes in OCD
Globally, executive functioning refers to the process of implementing deliberate strategies to achieve a pre-established goal. However, specific executive subprocesses include set-shifting and organizational strategy, among others. Set-shifting is related to the ability to reallocate attention among multiple dimensions in the environment. It requires the suppression of a dominant response in favor of a new, learned response. Lucey et al. (1997) found that individuals with OCD made significantly more perseverative errors on the Wisconsin Card Sorting Task (WCST) and had more null-sorts than did control individuals; similar findings also have emerged from studies of subclinical compulsive checkers (Goodwin & Sher, 1992; Lawrence et al., 2006; Zohar et al., 1995). Okasha (2000) also has reported that OCD is associated with impaired set-shifting and mental control, as evidenced by the fact that individuals with the disorder complete fewer categories and make significantly more perseverative errors and more errors overall on the WCST in comparison to healthy controls. However, other researchers have been unable to replicate these findings (e.g., Abbruzzese, Ferri, & Scarone, 1995).
Organizational strategy, another executive subprocess, refers to the ability to guide and structure goal-directed activity. On the Rey-Osterrieth Complex Figure, individuals with OCD appeared to use less efficient organizational processes when copying the visual figure due to the fact that they tended to begin copying nonstructural details prior to major structural components (Savage et al., 1999). Attention remained focused on these minor details for the majority of the copy period and shifted to major structural components only later in the testing session. These individuals exhibited impaired immediate recall relative to controls, which is consistent with reduced encoding during the initial copy period. Savage et al.’s (1999) findings are compatible with those of Veale, Sahakian, Owen, and Marks (1996), who detected poor strategic processes in individuals with OCD on the Tower of London test. Once these individuals had provided an incorrect response, they were impaired in their ability to generate more adaptive strategies and instead persevered in implementing ineffective ones. Similarly, on the Delayed Alternation Task, individuals with OCD exhibited relative difficulty acquiring an ideal response strategy and made significantly more errors than controls. This finding was replicated by Cavedini, Ferri, Scarone, and Bellodi (1998).
Learn more about how cognitive deficits may affect decision-making in OCD.
Full references for this page can be found here.
Questions? Comments? Sound off below.