Measuring Mindfulness: A Psychophysiological Approach

Front Hum Neurosci. 2018; 12: 249. Published online 2018 Jun 28. doi: 10.3389/fnhum.2018.00249 PMCID: PMC6031749 PMID: 30002624 Vladimir Bostanov,1 Lilian Ohlrogge,2 Rita Britz,2 Martin Hautzinger,2 and Boris Kotchoubey1,* 1Institute of Medical Psychology and Behavioral Neurobiology, University of Tübingen, Tübingen, Germany 2Department of Psychology, University of Tübingen, Tübingen, Germany Edited by: Shuhei Yamaguchi, Shimane University, Japan Reviewed by: Domenico De Berardis, Azienda Usl Teramo, Italy; Lin Sørensen, University of Bergen, Norway *Correspondence: Boris Kotchoubey ed.negnibeut-inu@yebuohctok.sirob Author information ▼ Article notes ► Copyright and License information ► Disclaimer Go to: Abstract Mindfulness-based interventions have proved effective in reducing various clinical symptoms and in improving general mental health and well-being. The investigation of the mechanisms of therapeutic change needs methods for assessment of mindfulness. Existing self-report measures have, however, been strongly criticized on various grounds, including distortion of the original concept, response bias, and other. We propose a psychophysiological method for the assessment of the mindfulness learned through time-limited mindfulness-based therapy by people who undergo meditation training for the first time. We use the individual pre-post-therapy changes (dERPi) in the event-related brain potentials (ERPs) recorded in a passive meditation task as a measure of increased mindfulness. dERPi is computed through multivariate assessment of individual participant's ERPs. We tested the proposed method in a group of about 70 recurrently depressed participants, randomly assigned in 1.7:1 ratio to mindfulness-based cognitive therapy (MBCT) or cognitive therapy (CT). The therapy outcome was measured by the long-term change (dDS) relative to baseline in the depression symptoms (DS) assessed weekly, for 60 weeks, by an online self-report questionnaire. We found a strong, highly significant, negative correlation (r = −0.55) between dERPi (mean = 0.4) and dDS (mean = −0.7) in the MBCT group. Compared to this result, the relationship between dDS and the other (self-report) measures of mindfulness we used was substantially weaker and not significant. So was also the relationship between dERPi and dDS in the CT group. The interpretation of dERPi as a measure of increased mindfulness was further supported by positive correlations between dERPi and the other measures of mindfulness. In this study, we also replicated a previous result, namely, the increase (dLCNV) of the late contingent negative variation (LCNV) of the ERP in the MBCT group, but not in the control group (in this case, CT). We interpreted dLCNV as a measure of increased meditative concentration. The relationship between dLCNV and dDS was, however, very week, which suggests that concentration might be relatively unimportant for the therapeutic effect of mindfulness. The proposed psychophysiological method could become an important component of a “mindfulness test battery” together with self-report questionnaires and other newly developed instruments. Keywords: mindfulness, mindfulness-based cognitive therapy, MBCT, concentration, event-related potentials, ERP, depression, t-CWT Go to: 1. Introduction Mindfulness as a general concept and mindfulness meditation as a therapeutic method have become increasingly popular in the last decades and the number of research papers on the subject has grown exponentially (Williams and Kabat-Zinn, 2011). Mindfulness-based stress reduction (MBSR, Kabat-Zinn, 1990), mindfulness-based cognitive therapy (MBCT, Segal et al., 2002), and other related interventions have proved effective in reducing stress, anxiety, depression, and other clinical symptoms and in improving general mental health and well-being (Hofmann et al., 2010; Fjorback et al., 2011; Kuyken et al., 2015). The investigation of the mechanisms by which these therapeutic changes occur requires assessment methods, and, indeed, several self-report questionnaires have been developed for the purpose of measuring mindfulness (Baer, 2011). This approach has, however, been strongly criticized on various grounds, and it has been pointed out that existing questionnaires might not provide valid measures of mindfulness as defined by Buddhist sources and adopted by MBSR/MBCT (Grossman and Van Dam, 2011). Mindfulness is a notoriously elusive concept and is hard to define. We will come back to this problem in the discussion. But for the time being, we will use a simple and fairly inclusive definition: mindfulness is what is practiced in mindfulness meditation. The circularity is not a joke—this is how mindfulness was defined in some of the most ancient Buddhist sources (Bodhi, 2011), and also, implicitly, by MBSR founder (Kabat-Zinn, 1990). In his original description of the program Kabat-Zinn (1990) did not provide any definition; recently (Kabat-Zinn, 2011), he stated that mindfulness had thus been defined by the whole book. Our definition emphasizes the experiential nature of mindfulness and the inherent difficulty of putting into words something that, ultimately, must be practiced in order to be understood. It also allows for an arbitrary length of the verbal description that would attempt to convey the meaning of the concept—originally, the Buddha's discourse on “The four establishments of mindfulness” (Bodhi, 2005; Ñanamoli and Bodhi, 2009); more recently, the MBSR and MBCT treatment protocols (Kabat-Zinn, 1990; Segal et al., 2002). It is also important to emphasize that our definition does not exclude mindfulness as practiced and developed in everyday life, without meditation. It just utilizes the fact that formal meditation provides an excellent opportunity to measure mindfulness in a controlled laboratory setting. We also use another, clinically relevant, theoretical description by MBCT cofounder Teasdale (1999a), who defined mindfulness as the only “mode of mind” (Kabat-Zinn, 1990) that facilitates emotional processing and therapeutic change. The mindful mode is marked by “metacognitive awareness” (Teasdale, 1999b), the deep, intuitive, experiential understanding (or insight) that thoughts and emotions are passing mental events, and not the reality about the self, the world and the future. Teasdale (1999a) contrasted the mindful “being mode” to the habitual “doing mode” marked by problem-solving and achievement-oriented thinking characteristic of usual everyday activity. He also pointed up “rumination,” a cognitive style marked by circular thinking about one's physical and emotional state (Nolen-Hoeksema, 1987, 1991), as a particularly important example of the doing mode, because it is a well-known, central risk factor for depressive relapse/recurrence (Nolen-Hoeksema and Morrow, 1991; Segal et al., 2002; Donaldson and Lam, 2004, pp. 35–36). The modes of mind description is consistent with another influential model by Bishop et al. (2004), who also defined mindfulness as a mode (rather than a trait, as assumed by most self-report instruments), and additionally emphasized acceptance as an important therapeutic component of mindfulness (Hayes et al., 1999). The Toronto Mindfulness Scale (TMS, Lau et al., 2006, see section 2.4.3 below), a questionnaire developed within the framework of the model of Bishop et al. (2004), is (presently, to the best of our knowledge) the only self-report instrument that measures mindfulness as a state (mode), rather than a trait (Baer, 2011, 2016). For this purpose it is administered immediately after meditation. It suffers, however, from the same limitations as all other questionnaires, like misunderstanding of items' meaning, response bias, and other impairments of objectivity inherent in self-report measures (Grossman and Van Dam, 2011). With the present study, we propose an alternative to the self-report assessment of mindfulness. We define mindfulness as the mode of mind established during mindfulness meditation that has been learned in a standard MBSR/MBCT training. We assume that the difference between the mindful mode and the ordinary doing mode is represented by a difference in the event-related brain potentials (ERPs) recorded during meditation before and after the training (in participants with no previous experience with meditation). Further, we assume that, since mindfulness is a very complex and multifaceted concept, it is represented by different change patterns in different participants' ERPs, but the total amount of changes in each participant's ERP reflects how well he/she has learned to be mindful during meditation. Hence, a mathematical-statistical measure of the individual ERP change should predict therapy outcome. Obviously, such a psychophysiological measure of mindfulness is guaranteed to be free of any response bias and also does not suffer from the limitations of verbal expressions that may distort and misrepresent the subtle nature of mindfulness and, even when formulated accurately, may be misunderstood by participants. In our previous studies (Bostanov et al., 2012), we used event-related brain potentials (ERPs) in an attempt to find a psychophysiological measure of meditative concentration. (For the difference between mindfulness and concentration, see the Discussion). ERPs are extracted directly from the electroencephalogram (EEG) and can reflect allocation of attentional resources in real time (Tecce, 1972; Pribram and McGuinness, 1992; Tecce and Cattanach, 1993). They can therefore be applied as direct psychophysiological measures of attention during meditation (Cahn and Polich, 2006; Ivanovski and Malhi, 2007). Moreover, ERPs can be elicited under passive conditions, i.e., without an active task (Polich, 1987; Baranov-Krylov et al., 2003), which makes them particularly valuable for measurements in the being mode of mind that can be easily disturbed by any active task. We designed a special “mindfulness ERP paradigm,” in which ERPs to neutral stimuli were recorded during meditation, after mood & rumination induction (Bostanov et al., 2012, see also sections 2.4.4, 2.4.6 below). We found that, after eight weeks of MBCT, the late contingent negative variation (LCNV) component of the ERP to an auditory test stimulus was significantly increased relative to both the pre-therapy baseline and a wait list control group. The LCNV amplitude is a direct, real-time measure of the excitation of neural pathways involved in conscious attentional processing and reflects the mobilization and allocation of attentional resources of limited capacity (Tecce, 1972; Tecce and Cattanach, 1993; Brunia and van Boxtel, 2001). Active CNV paradigms (see section 2.4.7 below) have been used to assess the concentration abilities of experienced meditators (Travis et al., 2000, 2002; Cahn and Polich, 2006). Therefore, we interpreted the pre-post-therapy change in LCNV (dLCNV) in our mindfulness paradigm as a measure of the increased concentration abilities of our participants after the training. Originally, the goal of the present study was to replicate the dLCNV effect (Bostanov et al., 2012) in an improved mindfulness ERP paradigm (sections 2.4.4, 2.4.6) and in comparison to an active control group (section 2.5), and to test whether dLCNV predicts therapy outcome. Later, however, we recognized the possibility of achieving another, potentially more important goal, namely, to construct a measure of mindfulness based on individual, multivariate ERP changes (dERPi). For this purpose, we adopted a multivariate approach proposed by Bostanov (2015a), which allowed us to quantify the difference between the pre-therapy ERP and the post-therapy ERP by representing the single-trial ERPs as points in a vector space and computing the geometric distance between the two sample means for each participant (section 2.8.3). Note the strong emphasis on specific paradigm design in this approach: since any ERP change is interpreted as increase in mindfulness, such interpretation cannot rely on established results from the ERP literature (like in the case of LCNV), but is based (according to our definition of mindfulness) mostly on the fact that dERPi is measured under very specific, carefully designed conditions, namely, during mindfulness meditation (section 2.4.6) after mood & rumination induction (section 2.4.4), in a group of participants trained in mindfulness meditation (section 2.5). Further support for the interpretation of dERPi as increase in mindfulness can be provided by relationships (in the expected directions) between dERPi and other measures of mindfulness (section 2.4.3), and between dERPi and therapy outcome (section 2.6). To summarize, the present study was not aimed at measuring the effectiveness of MBCT; its central goal was rather to develop a psychophysiological (ERP-based) measure of the mindfulness learned during the eight weeks of an MBCT course by recurrently depressed participants with no previous meditation experience. A secondary goal was to investigate further a previously found ERP correlate of meditative concentration.