Deconstructing the effects of concentration meditation practice on interference control: The roles of controlled attention and inflammatory activity

Highlights

• An intensive concentration meditation retreat improved controlled attention.

• The meditation retreat did not reduce automatic attention to distractors.

• Inflammatory activity selectively predicted controlled attention.

• Effects were consistent between- and within-groups in a waitlist-control experiment.

Abstract

Prior work has linked meditation practice to improvements in interference control. However, the mechanisms underlying these improvements are relatively unknown. In the context of meditation training, improvements in interference control could result either from increases in controlled attention to goal-relevant stimuli, or from reductions in automatic capture by goal-irrelevant stimuli. Moreover, few studies have linked training-related changes in attention to physiological processes, such as inflammatory activity, that are thought to influence cognitive function. This study addresses these gaps by examining associations between cognitive performance and cytokines in the context of an intensive meditation retreat. Participants were randomly assigned to complete 3 months of meditation training first, or to serve as waitlist controls. The waitlist-control participants then later completed a separate 3-month training intervention. We assessed participants’ interference control with a flanker task and used computational modeling to derive component processes of controlled and automatic attention. We also collected blood samples at the beginning, middle, and end of training to quantify changes in cytokine activity. Participants who completed training evidenced better controlled attention than waitlist controls during the first retreat intervention, and controls showed significant improvements in controlled attention when they completed their own, second retreat. Importantly, inflammatory activity was inversely associated with controlled attention during both interventions. Our results suggest that practice of concentration meditation influences interference control by enhancing controlled attention to goal-relevant task elements, and that inflammatory activity relates to individual differences in controlled attention.

Introduction

Emerging research indicates that training in meditation, mindfulness, and other contemplative practices can improve performance on inhibitory control tasks (Gallant, 2016, Gothe et al., 2014, MacLean et al., 2010, Mitchell et al., 2017, Sahdra et al., 2011, Zanesco et al., 2013, Zanesco et al., 2018). An outstanding question that stems from this work concerns the psychological or physiological mechanisms that may account for meditation-based changes in interference control (i.e., the cognitive ability to suppress attention to distracting thoughts or stimuli and focus on goal-relevant information). That is, in what ways, and through what routes, might training in meditation influence the typical functioning of cognitive systems? To date, most researchers have attempted to tackle this question by examining the neural correlates of meditation-related improvements in interference control (e.g., Moore et al., 2012, Moynihan et al., 2013, Slagter et al., 2007, Taren et al., 2017). In the current study, we take a different approach: We leverage multiple methods to examine how intensive meditation training influences—and inflammatory cytokine activity relates to—the component cognitive processes that underpin performance on an interference control task.

There is good reason to think that meditation training might affect biological and immunological processes in ways conducive to improvements in attentional control. Traditional accounts of Buddhist meditation emphasize the importance of a calm, serene, and equanimous bodily state for facilitating concentration and sustained attention (Gunaratana, 2002, Wallace, 2006). In addition, many contemporary forms of meditation training—including mindfulness-based stress reduction—aim to increase physical wellness and well-being, while strengthening faculties of attention, awareness, and executive function (Creswell, 2017). Although direct empirical support for these connections is lacking, indirect support can be derived from two separate research threads: the effects of meditation training on cognitive performance (e.g., Jha et al., 2007, MacLean et al., 2010), and the effects of meditation training on stress reduction, reactivity, and regulation (Conklin et al., 2019, Creswell and Lindsay, 2014).

In the present study, we examined the role of cytokine activity in relation to cognitive processes that are thought to be influenced by meditation practice. We focused on cytokine activity because previous work has shown that meditation practice can appreciably influence cytokines (Black et al., 2013, Black and Slavich, 2016, Buric et al., 2017, Creswell et al., 2016, Pace et al., 2009) and because the neuroimmune environment of the brain has been found to influence similar cognitive processes (Donzis and Tronson, 2014, McAfoose and Baune, 2009, Shields et al., 2017, Yirmiya and Goshen, 2011). In addition, we focused our investigation on concentrative styles of meditation practice—namely, shamatha practices that aim to cultivate calm, sustained attention and experiential states of meditative quiescence. Prior work in our lab has shown that intensive, full-time practice of shamatha meditation can (a) increase the ability to sustain one’s attention over time (MacLean et al., 2010, Zanesco et al., 2019), (b) improve the capacity for cognitive control (Sahdra et al., 2011, Zanesco et al., 2018), and (c) modulate biological markers associated with physical well-being (Jacobs et al., 2011).

Improvements in interference control can be explained by way of dissociable cognitive mechanisms. For example, dual-process models of interference control hold that automatic attentional activation to goal-irrelevant task elements occurs in parallel with controlled attention to goal-relevant elements (Ridderinkhof, 2002). In these models, observed behavioral responses represent the summation of these parallel processes, each of which can drive improvements in task performance. For instance, although top-down processes (e.g., controlled attention) are well-known to modulate the Stroop effect (e.g., MacLeod, 1991), bottom-up processes (e.g., automatic attentional activation) appear to contribute to its effects as well (Notebaert et al., 2006). Findings such as these suggest that the cognitive processes underlying interference control may be more varied than is often assumed, and highlight the importance of considering multiple attentional mechanisms when examining how performance on interference control tasks might be modified with training.

In studies of meditation practice, enhancements in interference control have often been attributed to improvements in controlled attention (for a review, see Gallant, 2016). But the generality of this conclusion is not clear. For instance, there is evidence to suggest that training in different styles of meditation, across a range of experience and intensity levels, may influence attentional subsystems in different ways. In one study, improvements in subcomponents of attention were distinguished according to whether people engaged in intensive or non-intensive concentrative meditation training (Jha et al., 2007). In this study, meditation-naïve participants who underwent an 8-week mindfulness-based stress reduction course demonstrated improvements in endogenous attentional orienting, while experienced meditators who engaged in a 1-month concentrative meditation retreat showed improvements in exogenous attentional orienting. In another study, three months of full-time intensive meditation produced changes in event related brain responses thought to reflect both top-down attentional and bottom-up perceptual processes (Zanesco et al., 2019). This latter study used the same participant sample and training program (shamatha practice) as our present report.

The cognitive substrates of meditation-related improvements in interference control remain opaque because few studies dissociate changes in controlled attention (i.e., attention to goal-relevant task elements) from automatic attentional activation (i.e., attention to distracting or goal-irrelevant task elements). This is largely because controlled and automatic attentional processes are hard to differentiate at the behavioral level. Computational cognitive modeling, however, offers a solution by permitting the decomposition of task performance into component processes that underlie global behavioral effects (Farrell and Lewandowsky, 2018). Nonetheless, despite the promise that computational modeling holds for clarifying the cognitive processes altered through meditation practice, it has been largely underutilized as a methodological approach in this area of research (van Vugt et al., 2019).

Drawing on theory, Ulrich et al. (2015) developed a model that fits interference control tasks by combining motor and decisional processes with superimposed controlled and automatic attentional processes. In this model, controlled attention is modeled as attention to goal-relevant information, and automatic attentional activation is modeled as attention to goal-irrelevant information. More concretely, in the flanker task—a task commonly used to assess interference control—this model considers the central stimulus (i.e., the target stimulus) to be the goal-relevant information, whereas it considers the flanking stimuli (i.e., the distracting stimuli) to be the goal-irrelevant information. The model of Ulrich et al. provides excellent fit to empirical data, and is the only model that can comprehensively account for empirical data from all types of interference control tasks (Servant et al., 2016, Ulrich et al., 2015).¹ The conceptual framework of this model, which we use in this study, is illustrated in Fig. 1.

There has been progress in understanding the neural mechanisms associated with meditation-related improvements in executive function (Moore et al., 2012, Moynihan et al., 2013, Slagter et al., 2007, Taren et al., 2017). However, less is known about concomitant physiological processes, such as inflammatory activity, that may mediate these effects. Indeed, recent work suggests that inflammatory activity may exert important influences on executive functions, leading to impairments in interference control (Shields et al., 2017). Higher circulating levels of proinflammatory cytokines, such as interleukin (IL)-6, predict worse interference control (Marsland et al., 2006, Mooijaart et al., 2013, Trollor et al., 2012, Trompet et al., 2008), and reduced gray and white matter in the brain (Hinwood et al., 2013, Satizabal et al., 2012, Tu et al., 2013, Wersching et al., 2010). In contrast, anti-inflammatory cytokines, such as IL-10, predict better interference control (Van Exel et al., 2003). Due to the antagonistic—and sometimes co-activating—effects of pro- and anti-inflammatory cytokines, the ratio of IL-6 to IL-10 is often used as an indicator of overall inflammatory activity, and as a predictor of psychological outcomes (e.g., Dhabhar et al., 2009, Fredericks et al., 2010). Importantly, a higher IL-6/IL-10 ratio has been shown to predict worse interference control (Fabregue et al., 2016).

In addition to its effects on interference control, emerging work suggests that the practice of meditation techniques, including mindfulness meditation, can reduce inflammatory activity (Black and Slavich, 2016, Conklin et al., 2019, Creswell et al., 2016, Pace et al., 2009). These effects have largely been interpreted as resulting from the stress-reducing features of meditative techniques, as well as the socioenvironmental contexts (e.g., residential retreats) in which meditation is often practiced (Conklin et al., 2019; see Pascoe et al., 2017, for a meta-analysis). In light of these findings, and of the broader work linking stress processes to inflammatory activity (Slavich and Irwin, 2014), we expect that reductions in inflammatory activity are one pathway through which meditation practice might improve interference control. To date, however, no study has tested this hypothesis.

We examined the cognitive and immunological mechanisms through which meditation practice might alter interference control. Specifically, we examined the interplay of these processes in the context of an intensive, residential meditation retreat. Residential retreats are a form of meditation training designed to support extended periods of dedicated practice. While on retreat, meditators follow a rigorous schedule of formal practice. This practice is conducted alongside a community of fellow practitioners and under the guidance of experienced teachers. Importantly, retreats can afford methodological advantages for research questions bearing on psychological process or mechanism (King et al., 2019; see also Zanesco et al., 2019). In the present study, we used meditation retreats to study a high dosage of practice, extended across months of training, for the purpose of determining the sensitivity of cognitive and psychobiological outcomes to meditation training.

We conducted a longitudinal, waitlist-controlled experiment examining the effects of intensive concentration meditation. Thirty participants were randomly assigned to a 3-month meditation retreat intervention and 30 participants were assigned to serve as waitlist controls. Later, the waitlist controls completed their own formally identical 3-month retreat intervention. In both groups, interference control was assessed via flanker task performance at mid-intervention (i.e., retreat or control period), after roughly 5 weeks of intensive training or after a 5-week waiting period. Flanker performance was compared between experimental groups (i.e., retreat vs. control) in the first intervention, and within participants (i.e., from waitlist status to active training) at the second intervention. We then used computational modeling to decompose flanker performance into component cognitive processes, thereby providing estimates of controlled and automatic attention. In addition, we longitudinally assessed the ratio of serum levels of the pro-inflammatory cytokine interleukin-6 (IL-6) to the anti-inflammatory cytokine interleukin-10 (IL-10) at the beginning, middle, and end of each retreat.

We had two main predictions. First, we expected that participants in the meditation retreat group would show improved flanker performance relative to waitlist controls. We remained agnostic, however, to whether these improvements would result from relatively stronger controlled attention, relatively weaker automatic attentional activation, or some combination of the two. Our second prediction draws on models of cognitive function that implicate increased low-grade peripheral inflammatory activity in poorer cognitive function (Donzis and Tronson, 2014, Shields et al., 2017). We hypothesized that poorer controlled attention would be associated with increased low-grade peripheral inflammatory activity prior to and contemporaneous with the interference control assessment across all participants.