Mindfulness predicts insight in obsessive-compulsive disorder over and above OC symptoms: An experience-sampling study

Obsessive-compulsive disorder (OCD) is defined by the presence of repetitive and intrusive thoughts, images, or impulses, and/or overt repetitive behaviors or mental acts that are performed in order to reduce distress or prevent perceived harm (American Psychiatric Association, 2013). Traditionally, the diagnostic criteria for OCD required insight into the unreasonable and excessive nature of obsessions and compulsions (Tolin, Abramowitz, Kozak, & Foa, 2001). Although OCD with poor or absent insight had been described much earlier (e.g. Insel & Akiskal, 1986; Robinson, Winnik, & Weiss, 1976; Solyom, DiNicola, Phil, Sookman, & Luchins, 1985), the DSM-IV field trial (Foa et al., 1995) empirically confirmed that OCD is characterized by a range of insight. Specifically, 4% of the sample reported themselves as being “completely certain”, and 26% reported themselves as being“mostly certain” that feared consequences would occur if they did not perform their compulsions. Subsequent studies have corroborated these findings (Catapano et al., 2010; Eisen et al., 2001; Matsunaga et al., 2002; Ravi Kishore, Samar, Janardhan Reddy, Chandrasekhar, & Thennarasu, 2004; Shavitt et al., 2014; Turksoy, Tukel, Ozdemir, & Karali, 2002). Therefore, in the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5, American Psychiatric Association, 2013), the “poor insight” specifier was expanded to include three insight-options; namely, good or fair insight, poor insight, and absent insight/delusional OCD beliefs. Thus, even a complete lack of insight into OCD beliefs is no longer classified as a delusional disorder.

Despite this progress, the multidimensional nature of insight (Markova, Jaafari, & Berrios, 2009), coupled with various terminological inconsistencies in the literature (Brakoulias & Starcevic, 2011), make it difficult to give a precise definition of insight in OCD. According to the DSM, insight in OCD presumably refers to the (in)accuracy of the beliefs that underlie the obsessions (e.g., the belief that touching a stain of dried blood will lead to HIV) or the rationality of compulsions (e.g., the necessity to wash hands 20 times to avoid being infected by a deadly disease) (Leckman et al., 2010). However, insight can also refer to the ability to attribute the belief, itself, to OCD (Brakoulias & Starcevic, 2011). Besides its multidimensional nature, insight is also presumed to vary across different obsessions as well as over time (Abramowitz & Jacoby, 2015). Based on clinical observations, Kozak and Foa (1994) suggest that insight in OCD is situation-bound. When confronted with the feared situation, the individual may demonstrate poorer insight than when not exposed to the feared stimulus. O'Dwyer and Marks (2000) summarize it succinctly: “Patients may logically repudiate their belief while in the safety of the therapist's office, but when in a “dangerous” situation may be 100% convinced of the fact” (p. 282). Thus, insight is presumably associated with anxiety level (Steketee & Shapiro, 1995) and should therefore be conceptualized as a mental state (Markova et al., 2009). In accordance with this idea, Shimshoni, Reuven, Dar, and Hermesh (2011) question a stable trait-like insight concept, recommending repeated assessments of insight. Surprisingly, the clinically observed temporal variations of insight have never been empirically studied in OCD. Furthermore, despite a large body of research into the clinical correlates of poor insight (e.g., Bellino, Patria, Ziero, & Bogetto, 2005; de Berardis et al., 2005; Catapano et al., 2010; Cherian et al., 2012; Jacob, Larson, & Storch, 2014; Jakubovski et al., 2011; Shavitt et al., 2014), no studies investigating the temporal correlates of insight currently exist. This is, however, of particular clinical relevance, since poor insight has been found to be associated with attenuated treatment outcome in several studies (e.g., Foa, 1979; Foa, Abramowitz, Franklin, & Kozak, 1999; Solyom et al., 1985; Tolin, Maltby, Diefenbach, Hannan, & Worhunsky, 2004). This finding might be due to either a decreased willingness or an inability to engage in exposure or to modify threat-related beliefs (Tolin et al., 2001). Shedding light on processes related to insight in OCD might help to develop effective psychotherapeutic strategies for its improvement. This, in turn, might increase patients' commitment to refrain from dysfunctional neutralization strategies, and consequently, improve therapy outcome. For this purpose, this study sought to explore, for the first time, which cognitive processes are associated with level of insight in everyday life of individuals with OCD.

In addition to overt and mental rituals and avoidance behavior, past research has also addressed the role of thought control strategies in maintaining OCD (e.g., Abramowitz, Whiteside, Kalsy, & Tolin, 2003; Amir, Cashman, & Foa, 1997; Fergus & Wu, 2010; Moore & Abramowitz, 2007; Purdon, Rowa, & Antony, 2007). Thought control strategies describe various ‘techniques’ people use to control their unwanted thoughts, some of which are more functional than others. In an effort to explore which strategies individuals use to control unwanted thoughts and to measure individual differences in the use of these strategies, Wells and Davies (1994) developed the Thought Control Questionnaire (TCQ). The original item pool was derived from results of semi-structured interviews conducted in a sample of 10 healthy individuals and 10 patients with different anxiety disorders. Consecutive cluster analyses based on results of student samples revealed 5 factors, corresponding to different strategies: Distraction (e.g., I keep myself busy), social control (e.g., I talk to a friend about the thought), and reappraisal (e.g., I challenge the thought's validity), which are considered to be adaptive; and worry (e.g., I think about past worries instead) and punishment (e.g., I shout at myself for having the thought), which have been found to be especially maladaptive and are additionally related to higher scores on measures of trait anxiety (Wells & Davies, 1994). Compared to anxious and non-anxious controls, OCD patients employ dysfunctional thought control strategies more often, exhibiting an increased use of worry and self-punishment and a decreased use of distraction (Abramowitz et al., 2003). Abramowitz et al. (2003) suggest that the use of these maladaptive thought control strategies is crucial to the maintenance of OCD. This assumption is supported by the finding that worry increases intrusions (Wells & Papageorgiou, 1995) and interacts with obsessive beliefs to predict the frequency of unwanted thoughts in healthy individuals (Fergus & Wu, 2010). Abramowitz et al. (2003) posit that worry and self-punishment both preserve levels of anxiety and threat associated with intrusions. Thus, worry and self-punishment may, by maintaining a heightened sense of threat, also be related to lower insight into the unreasonableness of obsessions and the senselessness of compulsions.

Contrary to the presumably negative effects of worry and self-punishment, mindfulness is assumed to be especially helpful for OCD patients (Didonna, 2009; Fairfax, 2008; Hertenstein et al., 2012; Key, Rowa, Bieling, McCabe, & Pawluk, 2017; Külz et al., 2013, 2014; Lu et al., 2018). Kabat-Zinn (1994) defines mindfulness as intentionally attending to the present moment experience in a non-judgmental and accepting way. Some studies hint at the efficacy of mindfulness training in reducing OC symptoms (e.g., Hanstede, Gidron, & Nyklicek, 2008; Hertenstein et al., 2012; Key et al., 2017; Külz et al., 2013; however, see Strauss et al. (2018), for contrary results), and in a more recent study, the capacity to allow thoughts and feelings to come and go was found to predict symptom reduction after CBT (Hawley et al., 2017). Looking beyond symptom reduction, laboratory studies have found that mindfulness strategies reduce distress associated with intrusive thoughts (Marcks & Woods, 2005; Najmi, Riemann, & Wegner, 2009), increase willingness to experience intrusive thoughts (Marcks & Woods, 2007), and reduce post-exposure anxiety and the urge to neutralize (Wahl, Huelle, Zurowski, & Kordon, 2013). With a mindful attitude, thoughts and feelings are perceived as temporary mental states rather than as accurate reflections of reality. This change in relation to experience has been termed decentering or cognitive defusion (Fresco, Segal, Buis, & Kennedy, 2007; Hayes, Strosahl, & Wilson, 1999). It has been suggested that obsessive thoughts lose their threatening character and may stop causing distress when OCD patients regard them as transient mental events. Furthermore, Didonna (2009) assumes that the sense of non-attachment which results from decentering increases tolerance towards unpleasant inner states and thereby may also improve insight in OCD patients.

To the best of our knowledge, no study has yet examined temporal fluctuations of insight and their association with antecedent worry and self-punishment, as compared to mindfulness in OCD, and consequently, adequate assessment tools are lacking. The Brown Assessment of Beliefs Scale (Eisen et al., 1998) – the gold standard of insight assessment – assesses global insight level regarding the past week, and thus is not appropriate for assessing temporal variations of insight in daily life. Ecological Momentary Assessment Method (EMA, Stone & Shiffman, 1994), also known as Experience-Sampling Method (ESM, Csikszentmihalyi & Larson, 1987), is defined by the repeated collection of real-time data on participants’ momentary experiences in their natural environments (Stone, Shiffman, Atienza, & Nebeling, 2007). The repeated and dense sampling of momentary states provides a high temporal resolution, allowing the analysis of dynamic processes and their temporal correlates. Additionally, unlike retrospective techniques of data collection, EMA is less prone to memory distortions like recall biases and guarantees the realization of ecological validity (Shiffman, Stone, & Hufford, 2008). EMA is an important research tool which has already been successfully used in a variety of clinical studies (Trull & Ebner-Priemer, 2009). Thus, EMA might prove promising for the repeated measurement of insight.

The present study had two aims. Firstly, we sought to empirically confirm clinical observations of temporal variation of insight in OCD in order to improve insight conceptualization. In this context, our aim was also to quantify the intensity of insight fluctuation by taking the temporal order of scores into account. On this basis, we conducted an exploratory analysis of differences in the intensity of insight fluctuation between OC symptom dimensions. The scientific exploration of the dynamic patterns of insight in OCD is a completely novel line of research, which is, however, of crucial importance since insight has always challenged the nosologic boundaries between OCD and psychosis. Secondly, this study was designed to investigate whether worry and self-punishment on the one hand, or a mindfulness-based strategy on the other hand, predict the consecutive degree of insight. We hypothesized that antecedent worry and self-punishment would predict lower levels of insight, whereas a mindfulness-based strategy was expected to have the reverse effect. Furthermore, we sought to analyze if worry, self-punishment and mindfulness remain significant predictors of insight with a time-lag of one assessment point (i.e., when predicting insight measured at time point t by these strategies assessed at time point t-1). A time-lag of one assessment point was chosen following the design of a study by Hartley, Haddock, Vasconcelos e Sa, and Emsley (2014) to explore the temporal stability of the associations we sought to investigate.