Abstract
Trait self-report mindfulness scales measure one's disposition to pay nonjudgmental attention to the present moment. Concerns have been raised about the validity of trait mindfulness scales. Despite this, there is extensive literature correlating mindfulness scales with objective brain measures, with the goal of providing insight into mechanisms of mindfulness, and insight into associated positive mental health outcomes. Here, we systematically examined the neural correlates of trait mindfulness. We assessed 68 correlational studies across structural magnetic resonance imaging, task-based fMRI, resting-state fMRI, and EEG. Several consistent findings were identified, associating greater trait mindfulness with decreased amygdala reactivity to emotional stimuli, increased cortical thickness in frontal regions and insular cortex regions, and decreased connectivity within the default-mode network. These findings converged with results from intervention studies and those that included mindfulness experts. On the other hand, the connections between trait mindfulness and EEG metrics remain inconclusive, as do the associations between trait mindfulness and between-network resting-state fMRI metrics. ERP measures from EEG used to measure attentional or emotional processing may not show reliable individual variation. Research on body awareness and self-relevant processing is scarce. For a more robust correlational neuroscience of trait mindfulness, we recommend larger sample sizes, data-driven, multivariate approaches to self-report and brain measures, and careful consideration of test–retest reliability. In addition, we should leave behind simplistic explanations of mindfulness, as there are many ways to be mindful, and leave behind simplistic explanations of the brain, as distributed networks of brain areas support mindfulness.
INTRODUCTION
The Study of Trait Mindfulness
Mindfulness is often defined as an accepting, open-minded attention to the present moment (Creswell, 2017). The word mindfulness originated from Eastern contemplative traditions, specifically, as a translation of the term sati from Pali or smrti from Sanskrit, which means remembering or being aware. Mindfulness was introduced to Western medicine with the advent of mindfulness-based stress reduction (MBSR) in the 1990s (Kabat-Zinn, 2005). MBSR consists of 8 weeks of instruction in mindfulness practices like body scans and breath awareness, which emphasize paying attention to the present moment, and gently redirecting one's attention whenever it wanders. MBSR and its derivatives have been used to effectively treat anxiety (Hoge et al., 2023; Goldin, Ziv, Jazaieri, Hahn, & Gross, 2013; Hölzel et al., 2013), pain (Goyal et al., 2014), depression (Strauss et al., 2023; Segal, Williams, & Teasdale, 2018; Kuyken et al., 2008), and more (Kabat-Zinn, 2011).
In tandem with research on mindfulness training, studies have been conducted on mindfulness as a trait. Although the term mindfulness is employed in research about both traits and mindfulness practice or intervention, the relations between those two sorts of mindfulness is unclear—“mindfulness is not simply a product of meditation” (Brown & Ryan, 2004, p. 246). Instead, trait, or dispositional, mindfulness may be a stable personality-like characteristic that is a window into consciousness, emotional experience, and other constructs. Trait mindfulness is usually measured using self-report scales. Two of the most influential self-report scales are the Mindful Attention Awareness Scale (MAAS) and the Five-Facet Mindfulness Questionnaire (FFMQ). The MAAS is a single-factor scale that consists of 15 reverse-scored items that measure one's tendency to make attentional lapses throughout everyday life, for example, “I rush through activities without being really attentive to them” (Brown & Ryan, 2003). In this sense, the MAAS operationalizes trait mindfulness as receptive attention and awareness to what is happening in the present moment.
The FFMQ (Baer, Smith, Hopkins, Krietemeyer, & Toney, 2006) is a multifactorial trait mindfulness assessment based on items from previous scales. The FFMQ comprises five facets—Acting with Awareness, Nonjudging, Nonreactivity, Describing, and Observing (Baer et al., 2006; Baer, Smith, & Allen, 2004). Each facet is considered an essential element of being mindful. Acting with Awareness, like the MAAS, refers to attending to one's actions in the present moment, for example, “It seems I am running on automatic without much awareness of what I'm doing” (reverse-coded). Nonjudging relates to not evaluating or judging one's thoughts or feelings, for example, “I criticize myself for having irrational or inappropriate emotions” (reverse-coded). Nonreactivity is defined as allowing thoughts to come and go without being caught up in them, for example, “I perceive my feelings and emotions without having to react to them.” Describing refers to labeling experiences with words, for example, “I am good at finding words to describe my feelings.” Finally, Observing is defined as noticing or attending to internal experiences like sensations, for example, “I pay attention to sensations, such as the wind in my hair or sun on my face.” Factor analyses have shown that these facets are statistically dissociable (Baer et al., 2008), although the Observing facet may show limited validity (Pang & Ruch, 2019; Gu et al., 2016).
Although the FFMQ and the MAAS are most commonly used, there are other scales including the Philadelphia Mindfulness Scale (PMS; Cardaciotto, Herbert, Forman, Moitra, & Farrow, 2008), Freiburg Mindfulness Inventory (FMI; Walach, Buchheld, Buttenmüller, Kleinknecht, & Schmidt, 2006), Kentucky Inventory of Mindfulness Skills (KIMS; Baer et al., 2004), and Cognitive and Affective Mindfulness Scale–Revised (CAMS-R; Feldman, Hayes, Kumar, Greeson, & Laurenceau, 2007). In addition, there are multiple valid short-form versions of the FFMQ (Bohlmeijer, ten Klooster, Fledderus, Veehof, & Baer, 2011; Baer et al., 2006). The diversity of mindfulness scales reflects the complexity of defining mindfulness (Rau & Williams, 2016). Indeed, the ontology of mindfulness is much debated (Van Dam et al., 2018; Brown, Ryan, Loverich, Biegel, & West, 2011; Grossman, 2011). Here, we discuss (a) empirical findings that support the validity of mindfulness scales, (b) concerns about a lack of validity for those scales, and (c) proposals for how the study of trait mindfulness might move forward.
Correlational studies using the FFMQ, MAAS, and the other scales have frequently associated greater trait mindfulness with better psychological well-being, specifically with enhanced positive affect, self-compassion, empathy, and life purpose (Kong, Wang, & Zhao, 2014; Keng, Smoski, & Robins, 2011; Schutte & Malouff, 2011; Howell, Digdon, Buro, & Sheptycki, 2008; Baer et al., 2004). Negative correlations have been found with negative affect, stress, anxiety, and depression symptoms (Treves et al., 2023; Carpenter, Conroy, Gomez, Curren, & Hofmann, 2019; Tomlinson, Yousaf, Vittersø, & Jones, 2018; de Bruin, Meppelink, & Bögels, 2015; Greco, Baer, & Smith, 2011; Coffey & Hartman, 2008). Likewise, greater trait mindfulness has been associated with better cognitive performance like decreased mind-wandering (Mrazek, Smallwood, & Schooler, 2012), improved sustained attention (Shin, Black, Shonkoff, Riggs, & Pentz, 2016), and improved academic scores (Caballero et al., 2019). More broadly, trait mindfulness has been identified as a mediator in the relationship between life experiences and well-being (MAAS; Saricali, Satici, Satici, Gocet-Tekin, & Griffiths, 2022; FFMQ; Beshai, Salimuddin, Refaie, & Maierhoffer, 2022; Franquesa et al., 2017; Josefsson, Larsman, Broberg, & Lundh, 2011).
Converging with this correlational approach, many of these affective and cognitive relationships have been replicated in the study of mindfulness interventions. Trait mindfulness, as proposed, generally does increase with mindfulness practice (Quaglia, Braun, Freeman, McDaniel, & Brown, 2016). In some cases, the quality of active mindfulness practice itself (“state” mindfulness) has been linked to trait mindfulness increases (Kiken, Garland, Bluth, Palsson, & Gaylord, 2015). Trait mindfulness functions as a predictor and a mediator of changes in clinical symptoms in response to mindfulness interventions (Gu, Strauss, Bond, & Cavanagh, 2015; Brown et al., 2011). The FFMQ in particular has been extensively used in randomized controlled trials of mindfulness interventions with clinical populations (Tran et al., 2022; Baer, Gu, Cavanagh, & Strauss, 2019; Alsubaie et al., 2017). Improvements in the FFMQ are often found to mediate decreases in psychological distress or improvements in clinical symptoms (D'Antoni, Feruglio, Matiz, Cantone, & Crescentini, 2022; Goldberg et al., 2020; Kearns et al., 2016; Vøllestad, Sivertsen, & Nielsen, 2011; cf. Labelle, Campbell, Faris, & Carlson, 2015).
On the other hand, there have been challenges to the validity of trait mindfulness. Experienced meditators may not consistently rate higher on trait mindfulness than novices (de Bruin et al., 2015; Christopher & Gilbert, 2007; MacKillop & Anderson, 2007), and they may understand the questionnaires differently from novices (Christopher, Christopher, & Charoensuk, 2009; Baer et al., 2006). In addition, trait mindfulness may increase in response to non-mindfulness-specific behavioral interventions and may thus reflect general mental health improvements (Tran et al., 2022; Goldberg et al., 2016). Individuals who endorse high state mindfulness may not endorse high levels of trait mindfulness (Thompson & Waltz, 2007).
Researchers have also examined whether trait mindfulness subscales make up a cohesive, independent construct by rigorously evaluating their constituent mechanisms and relationships to established psychological questionnaires. Bednar, Voracek, and Tran (2020) created a mechanism-inspired model of trait mindfulness involving factors of body awareness, attention regulation, emotion regulation, decentering, and nonattachment. They found that questionnaires beyond the FFMQ were more prominently loaded on each factor and that the created factors predicted mental health more robustly than the FFMQ. They also found associations with mental health varied between meditators and nonmeditators. Tran, Wasserbauer, and Voracek (2020) found that the FFMQ did not predict mental health above and beyond the Big Five personality measures. This limited incremental validity was also found by Altgassen, Geiger, and Wilhelm (2024) who used advanced psychometric methods to create a new single-factor model of trait mindfulness, and still found that mindfulness did not predict real-world outcomes, life satisfaction, and health behaviors beyond the Big Five personality measures. A final study found that trait mindfulness may be subject to a jangle fallacy, wherein it is considered to be distinct from other constructs but is actually composed of preexisting constructs in domains like attention regulation and emotional awareness (Beloborodova & Brown, 2023). In a network analysis, Beloborodova and Brown (2023) found that FFMQ facets clustered with other self-reports in these domains but not together.
It is increasingly evident that defining trait mindfulness as a single entity is neither accurate nor practically useful. Instead it is important to study the skills and faculties that comprise mindfulness (Van Dam et al., 2018). Of note, this paradigm shift has not yet occurred in brain imaging; we will soon see that manifold neurobiological mechanisms are often related to one overarching definition or measure of mindfulness.
Brain imaging research on trait mindfulness typically entails correlating scores on questionnaires with objective brain measures. This approach has been proposed to reveal mechanisms of action in mindfulness interventions (Goldberg et al., 2019), illuminate mental health disorders (Zhuang et al., 2017), and identify possible targets for neuromodulation (Cain et al., 2024; Zhang et al., 2023). Despite this potential, there have been no systematic reviews of brain imaging and trait mindfulness, making it difficult to pinpoint consistent findings and progress toward these goals. In this context, our objective was to systematically evaluate the neural correlates of trait mindfulness. Our investigation specifically focused on neural correlates across structural MRI studies (including diffusion-weighted imaging [DWI]), resting-state and task-based functional MRI (fMRI) studies, and EEG studies (Box 1). We included these diverse methods to encompass a wide range of brain structure and function in different contexts and at different timescales. This aligns with the many skills and faculties that make up trait mindfulness (Altgassen et al., 2024; Van Dam et al., 2018; Baer et al., 2006). It is also hoped that the discrepancies and convergences between methods illuminate foundational assumptions of interest to neuroscience researchers and personality/trait researchers.
Structural MRI/ DWI: Structural imaging methods that can be analyzed to derive gray-matter volume and surface area (structural MRI) or white-matter tract properties (DWI).
Task-based fMRI: Consists of collecting BOLD brain responses while participants perform a task. Typically, differences between the task and a baseline condition are used to estimate the activation of brain regions, but connectivity (correlational) analyses can also be conducted (Hanson & Bunzl, 2010).
Resting-state fMRI: Consists of collecting BOLD brain responses from participants while they are resting in a task-free state. Resting-state connectivity consists of correlations between timecourses of brain responses across the brain, “static functional connectivity,” and can be used to identify networks of related brain areas (Biswal, Yetkin, Haughton, & Hyde, 1995). In addition, timecourses may be windowed for analysis of correlations in each window to measure “dynamic functional connectivity” (DFC). Windowed correlation matrices may be clustered by similarity or distance, leading to the identification of DFC states, or brain states (Allen et al., 2014).
EEG: Consists of high temporal resolution electrical signals measured at the scalp. Measurements of time-locked average responses to stimuli can be assessed and are called ERPs. Other analyses can be conducted with stimuli or without stimuli present: frequency band power and microstate analysis (dynamic configurations of signals).
Proposed Neural and Psychological Mechanisms of Mindfulness
As discussed previously, researchers have recently developed new factors of mindfulness that are not specific to meditation practice but represent more general skills and faculties (Beloborodova & Brown, 2023; Bednar et al., 2020). Here, we have categorized brain imaging tasks into four core mechanisms of self-regulation (Tang, Hölzel, & Posner, 2015; Hölzel, Lazar, et al., 2011): emotion regulation, self-awareness, attention, and body awareness. The four mindfulness mechanisms are neither fully independent nor exhaustive (e.g., Vago & Silbersweig, 2012). In keeping with this, some tasks may be mapped to multiple mechanisms. As resting-state or structural findings may be related to any of the mechanisms, we do not categorize them. A strength of these selected mechanisms is that each has been independently studied in the cognitive neuroscience literature, and involved brain areas and processes have been identified. In addition, facets of mindfulness assessed through questionnaires may be associated with these mechanisms. For example, emotion regulation may be associated with Nonjudging, Nonreactivity, and related measures. In subsequent sections, we describe fundamental neuroscience research on each mechanism.
Emotion Regulation and Mindfulness
Emotions may be defined as coordinated changes in subjective feeling states, bodily sensations, and behaviors (Ekman & Cordaro, 2011; Izard, 2010). Emotion regulation involves initiating a goal to influence emotions, whether consciously or unconsciously (Gross, 2015). We emphasize two prevalent approaches to emotion regulation, not only because of their frequent mention in the neuroimaging literature but also because of their associations with mindfulness (Garland, Farb, Goldin, & Fredrickson, 2015). Reappraisal is a type of explicit emotion regulation wherein the context of an emotional response is strategically reconsidered to alter the response. For example, when viewing scenes, an image of a gruesome war scene may provoke an automatic startle response, which could then be modulated by reappraising the image as belonging to a scene from a famous movie. Reappraisal is a type of cognitive control that is explicit and often intentional (Troy, Shallcross, Brunner, Friedman, & Jones, 2018). Emotion regulation may also be implicit and automatic (Braunstein, Gross, & Ochsner, 2017; Gyurak, Gross, & Etkin, 2011), where the initial emotional response is modulated because of attentional factors preresponse (often referred to as reactivity). For example, an initial startle response to the war scene may vary depending on whether someone is fully attentive to the scene or their bodily feelings. Although there is an extensive theoretical basis to believe mindfulness is related to reactivity and attentional factors (Kral et al., 2018; Hölzel, Lazar, et al., 2011; Moore & Malinowski, 2009; Lutz, Slagter, Dunne, & Davidson, 2008), there is also some debate over whether mindfulness involves reappraisal (Hanley et al., 2021; Chambers, Gullone, & Allen, 2009). One perspective suggests that mindfulness is not synonymous with reappraisal; however, it may provide the mindfulness practitioner space (e.g., through distancing) to engage in reappraisal if desired. There are more granular mechanisms of emotion regulation and mindfulness than reappraisal and reactivity (Guendelman, Medeiros, & Rampes, 2017; Wheeler, Arnkoff, & Glass, 2017), but reappraisal and reactivity have been robustly distinguished in the brain.
These two types of emotion regulation involve overlapping and interacting brain systems. An essential distinction between explicit regulation (reappraisal) and implicit regulation (reactivity) is top–down versus bottom–up information processing. Reappraisal is a form of top–down processing where information from brain areas at the top of the hierarchy, including the pFC, is communicated to downstream brain regions, such as the amygdala (Goldin, Moodie, & Gross, 2019; Ochsner & Gross, 2005). This information is maintained over seconds and may involve linguistic content. On the other hand, implicit regulation is bottom–up, where responses in brain regions like the amygdala are communicated to upstream prefrontal and frontal areas. This information is maintained over hundreds of milliseconds (Koole, 2009). We used Neurosynth (https://neurosynth.org/), an online automated meta-analysis tool, to meta-analyze 247 fMRI studies associated with the keyword “emotion regulation.” This yielded brain regions associated with emotion regulation that included limbic brain structures, such as the amygdala, insula, and hippocampus, as well as cortical areas, such as the pFC, orbito-frontal cortex, and ACC (Appendix A).
In summary, we will examine the relationship of trait mindfulness to emotion regulation, which can be coarsely divided into top–down and bottom–up mechanisms.
Self-awareness and Mindfulness
Self-awareness refers to one's experience of the self. One may have a relatively fixed, narrative-oriented perspective on the self, or a more fluid, experiential perspective (Gallagher & Zahavi, 2020). For example, narrative self-processing involves identifying with statements like “I'm good at x”; “I grew up in y”; “I plan on z.” This self-process involves constructing a self out of these concepts and engaging in thoughts and judgments (evaluation) about these concepts. On the other hand, the experiencing self involves disidentifying with self-concepts and recognizing sensations and thoughts in the present moment. Mindfulness training may lead participants to shift toward the experiential self (Dahl, Lutz, & Davidson, 2015; Farb et al., 2007), and studies of long-term meditators (LTMs) provide evidence for substantial changes in self-perception, even including the complete dissolution of the narrative self (Cooper, Ventura, & Northoff, 2022; Nave et al., 2021; Olendzki, 2006).
Although it is difficult to construct objective measures of self-awareness, research has identified brain areas that may be relevant. For example, the cortical midline structures of the brain are consistently activated when participants engage in narrative self-processing (Zhou et al., 2020; Northoff et al., 2006). These areas include the posterior cingulate cortex (PCC), ventromedial pFC, anterior precuneus, and inferior parietal lobule (Hiser & Koenigs, 2018; Andrews-Hanna, 2012; Sajonz et al., 2010). Notably, these areas exhibit high interconnectivity while participants are resting, indicating a singular network (Raichle et al., 2001). This network has been called the “default-mode network” (DMN), referring to a default mode of brain processing in the absence of a task. An automated meta-analysis with the keyword “self-awareness” did not yield any studies on Neurosynth; however, when prompting the keyword “self-reference,” Neurosynth found 166 studies associated with and yielded brain structures related to the DMN (Appendix A).
Our investigation will explore the relationship of trait mindfulness to experiential and narrative frames of self-awareness, possibly involving the DMN.
Attention and Mindfulness
Attention refers to the process of prioritizing and filtering information from the environment, with mindfulness definitions emphasizing the importance of allocating attention to the present moment (e.g., Bishop et al., 2004). Mindfulness can be contrasted with mind-wandering, which occurs when one's attention is not focused on tasks or stimuli in the present moment (Smallwood & Schooler, 2015). There is evidence that mindfulness interventions promote present-moment attention and decrease mind-wandering (Xu, Purdon, Seli, & Smilek, 2017; Zanesco et al., 2016; Morrison, Goolsarran, Rogers, & Jha, 2014; Mrazek, Franklin, Phillips, Baird, & Schooler, 2013), a finding that has been replicated in individuals with high trait mindfulness (Mrazek et al., 2012). This change may involve differences in subprocesses such as sustaining attention and inhibiting distractors (Chiesa, Calati, & Serretti, 2011).
Attention is not a single action or skill but involves several processes, brain regions, and networks (Posner, 2016; Petersen & Posner, 2012; Fan, McCandliss, Fossella, Flombaum, & Posner, 2005). Brain regions often implicated in attention include the ACC, parietal cortex, and dorsolateral pFC (Appendix A; Marek & Dosenbach, 2018; Behrmann, Geng, & Shomstein, 2004). Brain networks involved in attention include the DMN, frontoparietal network (FPN), and salience network (SN). The DMN has been associated with inattention to the environment. Specifically, the DMN is involved in mind-wandering (Kucyi, Kam, Andrews-Hanna, Christoff, & Whitfield-Gabrieli, 2023; Buckner & DiNicola, 2019; Christoff, Gordon, Smallwood, Smith, & Schooler, 2009; Buckner, Andrews-Hanna, & Schacter, 2008) and is deactivated during attention-demanding tasks (Hellyer et al., 2014; Fox et al., 2005). The SN is engaged in stimulus-driven attention and consists of the insula and mid-cingulate, and the FPN is involved in externally focused, goal-directed attention and consists of lateral frontal and parietal areas (Seeley et al., 2007). Additional networks overlapping with the FPN and SN are the dorsal and ventral attention networks (Corbetta, Patel, & Shulman, 2008).
The functional relationships among these brain regions and networks and the attentional components of mindfulness have been studied using different approaches. One method includes assessing networks' fMRI-based activations during attentional tasks. Using this approach, Hasenkamp, Wilson-Mendenhall, Duncan, and Barsalou (2012) advanced a mechanistic model of mindfulness. In their protocol, participants practiced breath-focused attention in the scanner and pressed a button when they noticed mind-wandering. The researchers conducted an in-depth analysis of brain network activations across various stages of the experimental task. Notably, the FPN demonstrated heightened activity during phases where participants consistently focused on the designated object (in this case, their breathing). Conversely, during intervals when attention wandered from the focal object, the DMN exhibited increased activation. Following this, the SN became active, coinciding with the participants' awareness of their mind wandering. Finally, to realign focus toward breathing again, participants re-engaged the FPN. These findings suggest different network activations characterize different attentional states.
In conclusion, we will evaluate relationships between trait mindfulness and attentional control processes in the brain, which include sustained attention to an object, stimulus-driven shifts in attention, and internally directed, unfocused attention.
Body Awareness and Mindfulness
Mindfulness is related to body awareness, which encompasses the perception of bodily signals through processes like interoception and proprioception (Farb et al., 2015). Mindfulness interventions often emphasize body-focused practices such as breath awareness, where attention is focused on the sensations of the breath in the diaphragm, chest, and nostrils, as well as body scans, where attention moves slowly from one body part to the next. A meta-analysis of trait, longitudinal intervention, and LTM designs found that mindfulness is associated with improvements in body awareness, specifically with the objective accuracy of reporting body sensations (Treves, Tello, Davidson, & Goldberg, 2019). Furthermore, parallel changes in the subjective experience of body sensations may be observed (Datko, Lutz, et al., 2022; Mehling, 2016; Garfinkel, Seth, Barrett, Suzuki, & Critchley, 2015).
Body awareness may involve brain pathways where afferent signals travel from the vagus nerve to the brainstem, and from the brainstem to insular cortex and somatosensory cortex (Berntson & Khalsa, 2021; Critchley & Harrison, 2013; Cameron, 2001). Insular cortex has received particular attention as a possible location for integrating these afferent signals with beliefs, thoughts, and desires (Seth, 2013; Seth & Critchley, 2013). Neuroimaging and neurostimulation research highlight an anterior–posterior distinction within insular cortex (Schulz, 2016; Stephani, Fernandez-Baca Vaca, Maciunas, Koubeissi, & Lüders, 2011). The posterior insula contains representations of body parts, whereas the anterior insula does not have these representations and instead seems responsive to emotions and higher-order functions (Uddin, Nomi, Hébert-Seropian, Ghaziri, & Boucher, 2017; Craig, 2009). An automatic meta-analysis of 81 studies associated with the keyword “interoception” generated by Neurosynth yielded brain structures like the insula, ACC, PCC, and medial prefrontal cortex (mPFC; Appendix A).
In summary, we will examine the relationships between trait mindfulness and body awareness.
The Present Review
There are many reviews on neurobiological mechanisms of mindfulness, focusing mainly on mindfulness practice (Rahrig et al., 2022; Sezer, Pizzagalli, & Sacchet, 2022; Guendelman et al., 2017; Wheeler et al., 2017; Mooneyham, Mrazek, Mrazek, & Schooler, 2016; Lomas, Ivtzan, & Fu, 2015; Cahn & Polich, 2006). The current review is the first critical, systematic evaluation of the neural correlates of trait mindfulness. We investigate whether there are replicable effects for particular psychological mechanisms, for example, emotion regulation, which recent studies have proposed underly trait mindfulness (e.g., Bednar et al., 2020). In addition, a thorough treatment of the correlates of trait mindfulness may inform our understanding of the neurocognitive processes that convert state-level effects of mindfulness practice into long-term trait mindfulness. We explicitly compare the neural effects of mindfulness training to the neural correlates of trait mindfulness, in the Discussion section. Finally, as trait mindfulness is associated with mental health, neural correlates of trait mindfulness may be valuable treatment targets for neuromodulation.
METHODS
Inclusion and Exclusion Criteria
We included any study that associated a valid trait mindfulness measure with a valid neuroimaging measure. Valid trait mindfulness measures include the Adult and Adolescent Mindfulness Scale (AAMS), PMS, FFMQ, MAAS, FMI, CAMS-R, Child and Adolescent Mindfulness Measure (CAMM), SMQ, CHIME, and KIMS. Valid neuroimaging measures include resting-state fMRI, task-based fMRI, structural MRI, diffusion tensor imaging (DTI), and EEG. We excluded studies that examined PET imaging (n = 2) because of insufficient evidence for synthesis. We did not exclude studies that involved LTMs responding to questionnaires but have noted them where appropriate.
Searches
We conducted a first search of the literature in November 2022, with the aim of constructing a scoping review. This first search involved reviewing 1000 non-unique reports on trait mindfulness and brain imaging on Google Scholar and Proquest. We identified 50 studies in this search. During the review process, a reviewer suggested a systematic search of PubMed to supplement and extend this initial search. We conducted this systematic search with the search terms “trait mindfulness AND neuro*,” identifying 370 studies. Next, we screened studies to ensure they included a reported association between a valid measure of trait mindfulness and valid neuroimaging measure. We retained 18 studies for further review, which, combined with the initial searches, sums to 68 included studies. We included dissertations and posters as suggested to avoid publication biases (McLeod & Weisz, 2004). We followed PRISMA guidelines for the systematic search (Page et al., 2021; Appendices B and C).
Categorization
We constructed tables to separate the studies based on imaging modality. Across all tables, study identifier, mindfulness measure used, and sample size were tabulated. For the structural imaging (structural MRI and DWI/DTI), the structural measurement and brain areas associated with mindfulness were tabulated. For the task-based fMRI table, we tabulated the task and the measure associated with mindfulness. For the resting-state fMRI, we reported the measure associated with mindfulness and separated static functional connectivity (FC) studies from DFC and other methods. Finally, for the EEG table, we tabulated the task and the measure associated with mindfulness—distinguishing ERP studies and other studies (e.g., rest). Studies were first sorted in chronological order and then alphabetically.
Categorization into Mechanisms
Assigning brain findings to psychological mechanisms was done only for task-based imaging (not for resting-state or structural imaging). Task-based imaging manipulates a psychological process and observes neural changes, mitigating concerns over reverse inference (Poldrack, 2006). Note that multiple processes may be involved in one task.
Summaries
In the main Results section, we have highlighted findings from each imaging modality. For resting-state studies and structural MRI studies, we have reported analyses of the same brain index (e.g., pFC volume) across three or more studies, including null results. For task-based studies, we have highlighted analyses of each psychological mechanism, including null results. Note that “more mindful individuals” refers to individuals scoring higher on trait mindfulness.
Convergence with Mindfulness Training
Given the purview of this review, we did not systematically search for mindfulness training and LTM studies. We examined influential studies and reviews in the discussion.
RESULTS
Search
Of the 68 studies identified, nine involved structural imaging (Table 1), 17 involved resting-state fMRI (Table 2), 19 involved task-based fMRI (Table 3), and 23 involved EEG (Table 4). Twenty-eight studies used the FFMQ, 33 used the MAAS, and no other surveys were used in more than five studies. Most samples were small (n < 50). Mean sample sizes for task-based fMRI and EEG increased over time, mean sample sizes for resting-state fMRI were relatively stable, and mean sample sizes of structural MRI studies decreased over time (Appendix D). Preregistration and/or clinical trial registration was rare. We categorized the task-based studies into each core neural and psychological mechanisms, with some overlap: 21 studies on emotion regulation were found, 1 on self-awareness, 12 on attention, 6 on body awareness, and 3 were related to other domains (e.g., pain; Zeidan et al., 2018).
Overview of Structural MRI Studies on Mindfulness
Study . | TM Measure . | No. of Participants . | Type of Brain Structure Measurement . | Key Brain Areas Identified . |
---|---|---|---|---|
Murakami et al. (2012) | FFMQ | 19 | Gray matter volume | Right anterior insular cortex, right parahippocampal gyrus, and right amygdala |
Taren et al. (2013) | MAAS | 155 | Gray matter volume | Right amygdala, left caudate |
Lu et al. (2014) | MAAS | 247 | Gray matter volume | Right hippocampus, right amygdala, bilateral ACC, bilateral PCC, left orbito-frontal cortex |
Friedel et al. (2015) | MAAS | 82 late adolescents | Cortical thickness | Left anterior insula |
Makovac et al. (2016) | FFMQ | 16 GAD, 16 control (majority women) | Gray matter volume | In controls, amygdala (acting with awareness). In GAD, right medial frontal cortex (nonreactivity). |
Zhuang et al. (2017) | MAAS | 150 | Cortical volume | Precuneus |
FFMQ | 158 | Cortical volume, cortical thickness, and cortical surface area | Right dorsolateral pFC, left superior pFC, and inferior parietal lobe | |
Boekel and Hsieh (2018) | MAAS | 97 elderly | DTI: White matter fractional anisotropy | Internal capsule, external capsule, and corona radiata |
Baltruschat et al. (2021) | MAAS | 139 | Brain structural networks | Bilateral parahippocampal gyri, fusiform gyri, bilateral hippocampus |
FFMQ | 144 | Brain structural networks | Bilateral parahippocampal gyri, fusiform gyri, right insula, and bilateral amygdala | |
Abram et al. (2023) | FFMQ | 54 HC, 52 SZ | Cortical thickness, surface area, subcortical volumes | Brain age in SZ patientsa |
Study . | TM Measure . | No. of Participants . | Type of Brain Structure Measurement . | Key Brain Areas Identified . |
---|---|---|---|---|
Murakami et al. (2012) | FFMQ | 19 | Gray matter volume | Right anterior insular cortex, right parahippocampal gyrus, and right amygdala |
Taren et al. (2013) | MAAS | 155 | Gray matter volume | Right amygdala, left caudate |
Lu et al. (2014) | MAAS | 247 | Gray matter volume | Right hippocampus, right amygdala, bilateral ACC, bilateral PCC, left orbito-frontal cortex |
Friedel et al. (2015) | MAAS | 82 late adolescents | Cortical thickness | Left anterior insula |
Makovac et al. (2016) | FFMQ | 16 GAD, 16 control (majority women) | Gray matter volume | In controls, amygdala (acting with awareness). In GAD, right medial frontal cortex (nonreactivity). |
Zhuang et al. (2017) | MAAS | 150 | Cortical volume | Precuneus |
FFMQ | 158 | Cortical volume, cortical thickness, and cortical surface area | Right dorsolateral pFC, left superior pFC, and inferior parietal lobe | |
Boekel and Hsieh (2018) | MAAS | 97 elderly | DTI: White matter fractional anisotropy | Internal capsule, external capsule, and corona radiata |
Baltruschat et al. (2021) | MAAS | 139 | Brain structural networks | Bilateral parahippocampal gyri, fusiform gyri, bilateral hippocampus |
FFMQ | 144 | Brain structural networks | Bilateral parahippocampal gyri, fusiform gyri, right insula, and bilateral amygdala | |
Abram et al. (2023) | FFMQ | 54 HC, 52 SZ | Cortical thickness, surface area, subcortical volumes | Brain age in SZ patientsa |
Results for Baltruschat et al. (2021) and Zhuang et al. (2017) are separated by mindfulness scale (samples are overlapping). ATN = attention; BA = body awareness; ER = emotion regulation; SA = self-awareness; SZ = schizophrenic; TM = Trait mindfulness.
Brain age is a multivariate measure of predicted chronological age from brain gray matter anatomy.
Overview of Resting-state fMRI Studies on Mindfulness
Resting-state Studies . | TM Measure . | No. of Participants . | Resting-state Findings for Greater Trait Mindfulness . |
---|---|---|---|
TYPE: Static FC | |||
Shaurya Prakash et al. (2013) | MAAS | 25 (Elderly) | Within-DMN ⇑ |
Dean, Kohno, Hellemann, and London (2014) | MAAS | 15 meth. Users | Amygdala – right HC ⇓ |
Wang et al. (2014) | MAAS | 245 | Within-DMN ⇓ (Thalamus seed) |
Doll et al. (2015) | MAAS | 26 | Within-DMN ⇓ |
FMI | 26 | DMN–SN ⇓ | |
Bilevicius et al. (2018) | MAAS | 32 | Within-SN ⇑ Within-FPN ⇑ Within-DMN ⇓ DMN–SN ⇓ DMN-FPN ⇓ |
Parkinson et al. (2019) | FFMQ | 28 | DMN–SN ⇑, DMN-FPN ⇓, within-DMN ⇓, within-SN ⇑ (observing), within-DMN ⇑ (nonjudging), ATTN-DMN ⇑ (nonjudging) |
Harrison et al. (2019) | FFMQ | 40 | Within-DMN ⇓, DMN-SMN ⇑ |
Hunt et al. (2022) | FFMQ | 98 migraine patients, 36 HC | No relationship with within-DMN, DMN (PCC seed) – cerebellum ⇑ (post hoc in healthy controls) |
TYPE: Dynamic FC | |||
Mooneyham et al. (2017) | MAAS | 38 | Dwell time greater in state with correlations SN-FPN. Anticorrelations DMN–SN, DMN-FPN |
Lim, Teng, Patanaik, Tandi, and Massar (2018) | FFMQ | 39 | Dwell time greater in state with high correlations within-DMN, within-SN, anticorrelations in DMN–SN |
Marusak et al. (2018) | CAMM | 42 (Children) | Dwell time lesser in state with high correlations DMN-FPN, FPN-FPN, anticorrelations FPN-SN, DMN–SN |
Teng et al. (2022) | FFMQ | 40 | No correlation with dwell timea in state with high within-DMN, anticorrelations in DMN-FPN |
TYPE: Other | |||
Paolini et al. (2012) | MAAS | 19 (Obese, elderly) | Higher global efficiency in precuneus. Lower global efficiency in insula/auditory. Lower insula/auditory – visual connectivity |
Kong et al. (2016) | MAAS | 270 | ReHo ⇑ OFC, PHG, insula ReHo ⇓ IFG |
Gan et al. (2022) | FFMQ | 16 | Mean ALFF in rACC and L anterior insula (total, observing) ⇓ Mean ALFF in L posterior insula (nonjudging) ⇑ ALFF coupling between L posterior insula and L anterior insula (nonjudging) ⇑ |
Li et al. (2022) | FFMQ | 89 | ALFF negatively correlates in PCG (nonreactivity), negatively correlates with PMC |
Pasquini et al. (2023) | FFMQ | 56 TRD, 28 HC | Less within-network clustering/integration of DMN, CON, LIM in TRD patients |
Resting-state Studies . | TM Measure . | No. of Participants . | Resting-state Findings for Greater Trait Mindfulness . |
---|---|---|---|
TYPE: Static FC | |||
Shaurya Prakash et al. (2013) | MAAS | 25 (Elderly) | Within-DMN ⇑ |
Dean, Kohno, Hellemann, and London (2014) | MAAS | 15 meth. Users | Amygdala – right HC ⇓ |
Wang et al. (2014) | MAAS | 245 | Within-DMN ⇓ (Thalamus seed) |
Doll et al. (2015) | MAAS | 26 | Within-DMN ⇓ |
FMI | 26 | DMN–SN ⇓ | |
Bilevicius et al. (2018) | MAAS | 32 | Within-SN ⇑ Within-FPN ⇑ Within-DMN ⇓ DMN–SN ⇓ DMN-FPN ⇓ |
Parkinson et al. (2019) | FFMQ | 28 | DMN–SN ⇑, DMN-FPN ⇓, within-DMN ⇓, within-SN ⇑ (observing), within-DMN ⇑ (nonjudging), ATTN-DMN ⇑ (nonjudging) |
Harrison et al. (2019) | FFMQ | 40 | Within-DMN ⇓, DMN-SMN ⇑ |
Hunt et al. (2022) | FFMQ | 98 migraine patients, 36 HC | No relationship with within-DMN, DMN (PCC seed) – cerebellum ⇑ (post hoc in healthy controls) |
TYPE: Dynamic FC | |||
Mooneyham et al. (2017) | MAAS | 38 | Dwell time greater in state with correlations SN-FPN. Anticorrelations DMN–SN, DMN-FPN |
Lim, Teng, Patanaik, Tandi, and Massar (2018) | FFMQ | 39 | Dwell time greater in state with high correlations within-DMN, within-SN, anticorrelations in DMN–SN |
Marusak et al. (2018) | CAMM | 42 (Children) | Dwell time lesser in state with high correlations DMN-FPN, FPN-FPN, anticorrelations FPN-SN, DMN–SN |
Teng et al. (2022) | FFMQ | 40 | No correlation with dwell timea in state with high within-DMN, anticorrelations in DMN-FPN |
TYPE: Other | |||
Paolini et al. (2012) | MAAS | 19 (Obese, elderly) | Higher global efficiency in precuneus. Lower global efficiency in insula/auditory. Lower insula/auditory – visual connectivity |
Kong et al. (2016) | MAAS | 270 | ReHo ⇑ OFC, PHG, insula ReHo ⇓ IFG |
Gan et al. (2022) | FFMQ | 16 | Mean ALFF in rACC and L anterior insula (total, observing) ⇓ Mean ALFF in L posterior insula (nonjudging) ⇑ ALFF coupling between L posterior insula and L anterior insula (nonjudging) ⇑ |
Li et al. (2022) | FFMQ | 89 | ALFF negatively correlates in PCG (nonreactivity), negatively correlates with PMC |
Pasquini et al. (2023) | FFMQ | 56 TRD, 28 HC | Less within-network clustering/integration of DMN, CON, LIM in TRD patients |
Results for Doll and colleagues (2015) are separated by mindfulness scale (samples are overlapping). ATN = attention; ATTN = attentional network including dorsal and ventral components; BA = body awareness; CON = control network; ER = emotion regulation; LIM = limbic network; PCG = posterior cingulate gyrus; rACC = right ACC; SA = self-awareness; SMN = somatomotor network; TM = trait mindfulness; TRD = treatment resistant depression.
After a stressor, more mindful individuals exhibited more dwell time in state.
Overview of Task-based fMRI Studies on Mindfulness
Task-based fMRI Studies . | TM Measure . | Mechanism Area . | No. of Participants . | Task . | Activation Findings for Greater Trait Mindfulness . |
---|---|---|---|---|---|
Creswell et al. (2007) | MAAS | ER | 39 | Affect labeling | VLPFC ⇑, DLPFC ⇑, MPFC ⇑, L insula ⇑, Amygdala ⇓ |
Paul (2009) (Thesis) | KIMS | ATN | 11 (5 with MDD) | Flanker task | FEF ⇑, ACC ⇑, TPJ ⇑, IFG ⇑ |
Way, Creswell, Eisenberger, and Lieberman (2010) | MAAS | ER | 27 | Shape-matching task | Amygdala/HC⇓, MTL⇓, left VLPFC ⇓, MPFC ⇓ precuneus ⇓, PCC ⇓, thalamus ⇓, OFG ⇑ |
Frewen et al. (2010) | KIMS | ER | 19 (female) | Emotional script listening | DMPFC ⇑ (observing), L Amygdala ⇑ (observing) |
Modinos et al. (2010) | KIMS | ER | 18 | Affective picture viewing | DMPFC ⇑ |
Dickenson et al. (2013) | MAAS | ATN, BA | 31 | Focused breathing | SPL ⇑, TPJ ⇑, DLPFC ⇑ |
Paul et al. (2013) | FFMQ | ER | 19 (male) | Emotional go/no-go | Insula ⇓ (nonreactivity) |
Lutz et al. (2014) | MAAS | ER | 46 | Affective picture viewing | DMPFC ⇓, R insula ⇓ |
Lutz (2016) (Thesis) | FFMQ | SA | 44 (meditators and nonmeditators) | Self-reference task | DMPFC ⇑ (nonreactivity) |
Stillman et al. (2016) | MAAS | Not categorized | 42 elderly | Implicit learning task | Caudate-MTL connectivity negatively correlates with mindfulness and mediates relationship with implicit learning |
Lee et al. (2017) | KIMS | ER | 18 | Emotional face viewing | L IPL ⇓ (awareness) |
Martelli et al. (2018) | MAAS | ER | 39 | Cyberball (social rejection) | VLPFC ⇓ |
Kral et al. (2018) | FFMQ | ER | 158 (meditators and nonmeditators) | Affective picture viewing | Amygdala ⇓ (nonreactivity) |
Zeidan et al. (2018) | FMI | Not categorized | 76 | Thermal stimulation | PCC ⇓ |
Stein et al. (2022) | AAMS | ATN | 83 | N-back (WM) | Right VLPFC ⇓ (attention and awareness) |
Droutman et al. (2022) | FFMQ-SF | ER | 42 (women with substance use disorder) | Emotional go/no-go | Insula ⇑, caudate ⇑, putamen ⇑, thalamus ⇑, precuneus ⇑, cingulate ⇑ |
Datko, Schuman-Olivier, et al. (2022) (Poster) | FMI | BA | 18 (patients with migraine) | Focused breathing | Insula ⇓ |
Hehr et al. (2022) | CAMM | ER, BA | 12 (child cancer survivors) | Meditation strategies including focused breathing (video viewing) | No relationship with DMN activations |
Kim et al. (2023) | MAAS | ATN | 26 consistent,a 23 inconsistent | N-back task | More FPN edges positively related to MAAS in consistent group |
Task-based fMRI Studies . | TM Measure . | Mechanism Area . | No. of Participants . | Task . | Activation Findings for Greater Trait Mindfulness . |
---|---|---|---|---|---|
Creswell et al. (2007) | MAAS | ER | 39 | Affect labeling | VLPFC ⇑, DLPFC ⇑, MPFC ⇑, L insula ⇑, Amygdala ⇓ |
Paul (2009) (Thesis) | KIMS | ATN | 11 (5 with MDD) | Flanker task | FEF ⇑, ACC ⇑, TPJ ⇑, IFG ⇑ |
Way, Creswell, Eisenberger, and Lieberman (2010) | MAAS | ER | 27 | Shape-matching task | Amygdala/HC⇓, MTL⇓, left VLPFC ⇓, MPFC ⇓ precuneus ⇓, PCC ⇓, thalamus ⇓, OFG ⇑ |
Frewen et al. (2010) | KIMS | ER | 19 (female) | Emotional script listening | DMPFC ⇑ (observing), L Amygdala ⇑ (observing) |
Modinos et al. (2010) | KIMS | ER | 18 | Affective picture viewing | DMPFC ⇑ |
Dickenson et al. (2013) | MAAS | ATN, BA | 31 | Focused breathing | SPL ⇑, TPJ ⇑, DLPFC ⇑ |
Paul et al. (2013) | FFMQ | ER | 19 (male) | Emotional go/no-go | Insula ⇓ (nonreactivity) |
Lutz et al. (2014) | MAAS | ER | 46 | Affective picture viewing | DMPFC ⇓, R insula ⇓ |
Lutz (2016) (Thesis) | FFMQ | SA | 44 (meditators and nonmeditators) | Self-reference task | DMPFC ⇑ (nonreactivity) |
Stillman et al. (2016) | MAAS | Not categorized | 42 elderly | Implicit learning task | Caudate-MTL connectivity negatively correlates with mindfulness and mediates relationship with implicit learning |
Lee et al. (2017) | KIMS | ER | 18 | Emotional face viewing | L IPL ⇓ (awareness) |
Martelli et al. (2018) | MAAS | ER | 39 | Cyberball (social rejection) | VLPFC ⇓ |
Kral et al. (2018) | FFMQ | ER | 158 (meditators and nonmeditators) | Affective picture viewing | Amygdala ⇓ (nonreactivity) |
Zeidan et al. (2018) | FMI | Not categorized | 76 | Thermal stimulation | PCC ⇓ |
Stein et al. (2022) | AAMS | ATN | 83 | N-back (WM) | Right VLPFC ⇓ (attention and awareness) |
Droutman et al. (2022) | FFMQ-SF | ER | 42 (women with substance use disorder) | Emotional go/no-go | Insula ⇑, caudate ⇑, putamen ⇑, thalamus ⇑, precuneus ⇑, cingulate ⇑ |
Datko, Schuman-Olivier, et al. (2022) (Poster) | FMI | BA | 18 (patients with migraine) | Focused breathing | Insula ⇓ |
Hehr et al. (2022) | CAMM | ER, BA | 12 (child cancer survivors) | Meditation strategies including focused breathing (video viewing) | No relationship with DMN activations |
Kim et al. (2023) | MAAS | ATN | 26 consistent,a 23 inconsistent | N-back task | More FPN edges positively related to MAAS in consistent group |
ATN = attention; BA = body awareness; DLPFC = dorsolateral prefrontal cortex; ER = emotion regulation; FEF = frontal eye fields; HC = hippocampus; IFG = inferior frontal gyrus; IPL = inferior parietal lobe; MDD = major depressive disorder; MTL = medial temporal lobe; OFG = orbito-frontal gyrus; R = right; SA = self-awareness; SF = short-form; SPL = superior parietal lobe; TM = trait mindfulness; VLPFC = ventrolateral prefrontal cortex; WM = working memory.
Showing similar MAAS patterns of responses across test dates.
Overview of EEG Studies on Mindfulness
EEG Studies . | TM Measure . | Mechanism Area . | No. of Participants . | Task . | EEG Finding for Greater Trait Mindfulness . |
---|---|---|---|---|---|
TYPE: ERP | |||||
Brown et al. (2013) | MAAS | ER | 34 | Affective picture viewing | Decreased LPP (500–900 msec) |
FFMQ | ER | 34 | " | Decreased LPP (500–900 msec) | |
Teper and Inzlicht (2014) | PMS | ER | 45 | Time estimation with feedback | Decreased difference between FRN (250 msec) to reward vs. neutral but not aversive vs. neutral (acceptance) |
Cosme and Wiens (2015) | MAAS | ER | 51 | Affective picture viewing with startles | No relationship with LPP (400–800 msec), no relationship with P300, no relationship with EPN |
FFMQ | ER | 51 | " | No relationship with LPP (400–800 msec), decreased P300 (nonreactivity), no relationship with EPN | |
Dorjee, Lally, Darrall-Rew, and Thierry (2015) | MAAS | ER | 35 | Emotional word task | Increased N400, decreased P600 |
Eddy et al. (2015) | MAAS | ER | 24 | Affective picture viewing | Increased LPP (500–1500 msec) |
FFMQ | ER | 24 | " | No relationship with LPP (500–1500 msec) | |
Ho et al. (2015) | CAMS-R | ER | 22 (Meditators) | Affective picture viewing | Increased P200, no relationship with LPP (400–700 msec) |
Lin et al. (2016) | FFMQ | ER | 68 | Affective picture viewing | Decreased LPP (600–5000 msec; acting with awareness) |
Quaglia, Goodman, et al. (2016) | MAAS | ER | 62 | Emotional go/no-go | Decreased (more negative) N100, N200, P300 interaction by stimulus type |
Egan et al. (2018) | MAAS | ER | 118 | Affective picture viewing | Decreased LPP (400–1000 msec) in men |
FFMQ | ER | 118 | " | No relationship with LPP (400–1000 msec) | |
Lin, Fisher, et al. (2019) | FFMQ | ATN | 60a | Flanker task | Decreased P300s at centro-parietal electrode to incongruent stimuli (acting with awareness only) |
Lin, Eckerle, et al. (2019) | FFMQ | ATN | 103 controls, 103 guided meditation, majority femalea | Flanker task | No relationship with error-related negativity or error positivity for total or any subscales |
Zhang et al. (2019) | MAAS | ER | 20 | Affective picture viewing | More negative N200. No relationship to P100, LPP |
Verdonk et al. (2021) | FMI | BA | 17 | Heartbeat (resting-state) | No relationship to heartbeat-evoked potentialb |
Lin et al. (2024) | FFMQ+MAAS z score | ATN | 30 | Flanker task | Larger P300 after control task, smaller P300 after meditation. Overall smaller error-related negativity, larger error positivity |
TYPE: Other | |||||
Hinterberger et al. (2011) | FMI | ATN | 26 (Meditators) | Meditation | Lower theta activity in temporal regions, lower delta globally |
Jaiswal et al. (2019) | MAASc | ATN | 56c | Stroop task | Decreased delta power in posterior temporal and occipital areas |
MAASc | ATN | 56c | Spatial WM change detection | Increased alpha power | |
Doborjeh et al. (2020) | FFMQ | ATN | 20d | Auditory oddball task | Increased frontal connection weights to both targets and distractors |
Chen, Kirk, and Dikker (2021) | MAAS | Not categorized | 379 | Social interaction | No difference in MUSE power spectral density (between 1 and 30 Hz)e |
Deng, Yang, and An (2021) | FFMQ | ER | 45 | Affective picture viewing | Higher left frontal asymmetry |
Zarka et al. (2022) | FFMQ | ATN, BA | 40 (Meditators and nonmeditators) | Meditation | Microstates: decreased time in SN state (nonreactivity) |
Krishna et al. (2022)f | MAAS | Not categorized | 24 Heartfulness meditators, 30 nonmeditators | Heartfulness meditation | Higher right frontal theta and higher right frontal alpha during meditation and postmeditation. Lower right frontal beta during transmission |
Jaiswal et al. (2023) | MAASc | Not categorized | 56c | Resting | More gamma power frontotemporal, more delta and theta with eyes closed, higher delta-gamma / theta-beta cross-frequency coupling |
Zarka et al. (2024) | FFMQ | ATN, BA | 16 MBSR completers, 16 controls | Resting-state with auditory stimuli | Microstates: decreased time in state related to auditory responses (total, nonreactivity and nonjudgment), state related to SN (observing, nonreactivity) |
EEG Studies . | TM Measure . | Mechanism Area . | No. of Participants . | Task . | EEG Finding for Greater Trait Mindfulness . |
---|---|---|---|---|---|
TYPE: ERP | |||||
Brown et al. (2013) | MAAS | ER | 34 | Affective picture viewing | Decreased LPP (500–900 msec) |
FFMQ | ER | 34 | " | Decreased LPP (500–900 msec) | |
Teper and Inzlicht (2014) | PMS | ER | 45 | Time estimation with feedback | Decreased difference between FRN (250 msec) to reward vs. neutral but not aversive vs. neutral (acceptance) |
Cosme and Wiens (2015) | MAAS | ER | 51 | Affective picture viewing with startles | No relationship with LPP (400–800 msec), no relationship with P300, no relationship with EPN |
FFMQ | ER | 51 | " | No relationship with LPP (400–800 msec), decreased P300 (nonreactivity), no relationship with EPN | |
Dorjee, Lally, Darrall-Rew, and Thierry (2015) | MAAS | ER | 35 | Emotional word task | Increased N400, decreased P600 |
Eddy et al. (2015) | MAAS | ER | 24 | Affective picture viewing | Increased LPP (500–1500 msec) |
FFMQ | ER | 24 | " | No relationship with LPP (500–1500 msec) | |
Ho et al. (2015) | CAMS-R | ER | 22 (Meditators) | Affective picture viewing | Increased P200, no relationship with LPP (400–700 msec) |
Lin et al. (2016) | FFMQ | ER | 68 | Affective picture viewing | Decreased LPP (600–5000 msec; acting with awareness) |
Quaglia, Goodman, et al. (2016) | MAAS | ER | 62 | Emotional go/no-go | Decreased (more negative) N100, N200, P300 interaction by stimulus type |
Egan et al. (2018) | MAAS | ER | 118 | Affective picture viewing | Decreased LPP (400–1000 msec) in men |
FFMQ | ER | 118 | " | No relationship with LPP (400–1000 msec) | |
Lin, Fisher, et al. (2019) | FFMQ | ATN | 60a | Flanker task | Decreased P300s at centro-parietal electrode to incongruent stimuli (acting with awareness only) |
Lin, Eckerle, et al. (2019) | FFMQ | ATN | 103 controls, 103 guided meditation, majority femalea | Flanker task | No relationship with error-related negativity or error positivity for total or any subscales |
Zhang et al. (2019) | MAAS | ER | 20 | Affective picture viewing | More negative N200. No relationship to P100, LPP |
Verdonk et al. (2021) | FMI | BA | 17 | Heartbeat (resting-state) | No relationship to heartbeat-evoked potentialb |
Lin et al. (2024) | FFMQ+MAAS z score | ATN | 30 | Flanker task | Larger P300 after control task, smaller P300 after meditation. Overall smaller error-related negativity, larger error positivity |
TYPE: Other | |||||
Hinterberger et al. (2011) | FMI | ATN | 26 (Meditators) | Meditation | Lower theta activity in temporal regions, lower delta globally |
Jaiswal et al. (2019) | MAASc | ATN | 56c | Stroop task | Decreased delta power in posterior temporal and occipital areas |
MAASc | ATN | 56c | Spatial WM change detection | Increased alpha power | |
Doborjeh et al. (2020) | FFMQ | ATN | 20d | Auditory oddball task | Increased frontal connection weights to both targets and distractors |
Chen, Kirk, and Dikker (2021) | MAAS | Not categorized | 379 | Social interaction | No difference in MUSE power spectral density (between 1 and 30 Hz)e |
Deng, Yang, and An (2021) | FFMQ | ER | 45 | Affective picture viewing | Higher left frontal asymmetry |
Zarka et al. (2022) | FFMQ | ATN, BA | 40 (Meditators and nonmeditators) | Meditation | Microstates: decreased time in SN state (nonreactivity) |
Krishna et al. (2022)f | MAAS | Not categorized | 24 Heartfulness meditators, 30 nonmeditators | Heartfulness meditation | Higher right frontal theta and higher right frontal alpha during meditation and postmeditation. Lower right frontal beta during transmission |
Jaiswal et al. (2023) | MAASc | Not categorized | 56c | Resting | More gamma power frontotemporal, more delta and theta with eyes closed, higher delta-gamma / theta-beta cross-frequency coupling |
Zarka et al. (2024) | FFMQ | ATN, BA | 16 MBSR completers, 16 controls | Resting-state with auditory stimuli | Microstates: decreased time in state related to auditory responses (total, nonreactivity and nonjudgment), state related to SN (observing, nonreactivity) |
Results for Egan and colleagues (2018), Brown and colleagues (2013), Eddy and colleagues (2015), and Cosme and Wiens (2015) are separated by mindfulness scale. Results for Jaiswal and colleagues (2019) are separated by task. ATN = attention; BA = body awareness; EPN = early posterior negativity; ER = emotion regulation; SA = selfawareness; TM = trait mindfulness; WM = working memory.
Participants may overlap.
The heartbeat-evoked potential is a frontal/central EEG response 200–600 msec after the heartbeat.
Participants were classified as high-mindfulness (MAAS) low anxiety, or low-mindfulness high anxiety, same across studies.
Participants were selected based on improvements in trait mindfulness in response to intervention.
The study used MUSE portable EEG system.
Heartfulness meditation and subsequent transmission were not fully elaborated besides “relaxation and cleaning” and thus not categorized.
Structural MRI
MRI-based brain structural variation of regions has been correlated with individual differences in trait mindfulness (Table 1, Figure 1). Correlations were conducted with FFMQ and MAAS scales, with no apparent differences between the two. One study of 247 college-age adults reported that greater trait mindfulness (as measured using the MAAS) correlated with increased gray matter volume of the right amygdala, hippocampus, and ACC and decreased gray matter volume of orbito-frontal cortex (Lu et al., 2014). Other studies have found inconsistent relations between trait mindfulness and amygdala volume: Heightened mindfulness has been associated with increased (FFMQ; Murakami et al., 2012) or decreased amygdala volume (MAAS; Taren, Creswell, & Gianaros, 2013). Structural differences in insular cortex have also been associated with trait mindfulness. In one study, insular cortex thickness was examined at two points in adolescence (ages 16 and 19 years; Friedel et al., 2015). Adolescents with less thinning in the left anterior insula from ages 16 to 19 years showed higher scores on the MAAS at age 19 years. Longitudinal studies are valuable for examining the development of mindfulness as a trait, yet this was the only study identified in the current review. A small-sample correlational study found a positive association between FFMQ's Describing facet and right anterior insula volume (Murakami et al., 2012). Baltruschat and colleagues (2021) identified relationships between the structure of the insula and FFMQ using a structural network analysis (nonnegative matrix factorization) in 147 individuals.
Summary of structural findings. Regions were selected with the Harvard-Oxford cortical/subcortical atlases. Positive associations between mindfulness and a structural measure are denoted by ⇑, negative associations by ⇓. *Baltruschat 2021 used a nonnegative matrix factorization that finds relationships between spatial components and mindfulness, not isolated to volume or surface area.
Summary of structural findings. Regions were selected with the Harvard-Oxford cortical/subcortical atlases. Positive associations between mindfulness and a structural measure are denoted by ⇑, negative associations by ⇓. *Baltruschat 2021 used a nonnegative matrix factorization that finds relationships between spatial components and mindfulness, not isolated to volume or surface area.
In terms of networks, multiple studies have examined DMN-related areas. Baltruschat et al. found that the FFMQ was associated with the across-individual loadings of different DMN areas. Specific areas related to the FFMQ were the hippocampus, the cerebellum, and the precuneus. This study did not include simple metrics of gray matter volume or surface area. However, their structural network analysis provided evidence of associations between the FFMQ and brain structures associated with the DMN. Other studies identified correlations between greater trait mindfulness and brain structure in the precuneus (increased volume correlates with increased MAAS [Zhuang et al., 2017]) and PCC (decreased volume correlates with increased MAAS [Lu et al., 2014]). Thus, these studies, with some inconsistencies, generally link variation in trait mindfulness with variation in the anatomy of the DMN. Attentional control network regions may also be linked to trait mindfulness—associations were observed with dorsolateral pFC (increased surface area; Zhuang et al., 2017), ACC (increased volume; Lu et al., 2014), and medial frontal cortex (increased volume in individuals with anxiety; Makovac et al., 2016).
Resting-state fMRI
First, we review static FC results within and across brain networks (Table 2, Figure 2). In one study, 40 healthy volunteers completed the FFMQ before undergoing imaging while resting in the scanner (Harrison, Zeidan, Kitsaras, Ozcelik, & Salomons, 2019). The researchers looked at the correlations between a seed from the DMN (the precuneus) and other brain regions. Those who were higher in trait mindfulness exhibited less connectivity between the precuneus and the mPFC (another node of the DMN). A replication study by Hunt and colleagues (2022) in 36 healthy adults and 98 migraine patients found no such relationship. Despite these variations, the preponderance of evidence shows that trait mindfulness, as measured across different scales, is associated with decreased connectivity within the DMN (Pasquini et al., 2023; Li, Chen, Zheng, & Qiu, 2022; Harrison et al., 2019; Bilevicius, Smith, & Kornelsen, 2018; Doll, Hölzel, Boucard, Wohlschläger, & Sorg, 2015; Wang et al., 2014; cf. Hunt et al., 2022; Shaurya Prakash, De Leon, Klatt, Malarkey, & Patterson, 2013). Findings about interactions between networks are less consistent. Most research has been conducted on the so-called triple networks: the DMN, SN, and FPN (also called Central Executive Network [CEN]). In one study, researchers correlated resting-state network maps of the DMN, SN, and FPN in meditation-naive individuals with self-reported trait mindfulness, as measured by the MAAS (Bilevicius et al., 2018). Higher trait mindfulness was associated with decreased FC between the SN and right cuneus (a node of the DMN). Doll and colleagues (2015) observed decreases in FC between SN and DMN, whereas Parkinson, Kornelsen, and Smith (2019) found increases.
Summary of within-DMN, DMN–SN, and DMN-FPN resting-state connectivity findings. Networks are displayed using the seven-network Yeo 2011 atlas (Yeo et al., 2011). Positive associations between mindfulness and connectivity are denoted by ⇑, negative associations by ⇓.
Summary of within-DMN, DMN–SN, and DMN-FPN resting-state connectivity findings. Networks are displayed using the seven-network Yeo 2011 atlas (Yeo et al., 2011). Positive associations between mindfulness and connectivity are denoted by ⇑, negative associations by ⇓.
DFC results have examined brain states characterized by across-network correlations. In one study, researchers examined DFC in 38 college-age participants while they performed a 9-min breathing exercise (Mooneyham et al., 2017) before and after a mindfulness intervention. DFC was assessed between a priori seeds in the DMN, FPN, and SN. The authors discovered one state (one pattern of connectivity) where the DMN was anticorrelated with the FPN and SN, with more mindful individuals (as measured by MAAS) spending more time in this state (“dwell time”) during the scan. In addition, the mindfulness intervention increased dwell time in this state. Interestingly, those individuals who exhibited larger changes in trait mindfulness because of the mindfulness intervention showed the most significant increases in dwell time. The investigators interpreted this state as related to focused attention. Other studies of DFC have also identified relationships between trait mindfulness and prevalence of states characterized by anticorrelation between the DMN and other networks (positive, Lim et al., 2018; negative, Marusak et al., 2018; no relationship, Treves, Marusak, et al., 2024; Teng, Massar, & Lim, 2022). As yet, there are no consistent relationships between DFC brain states and trait mindfulness.
Other methods such as amplitude of low frequency fluctuations (ALFF; Gan et al., 2022), graph theory (Paolini et al., 2012), and regional homogeneity (ReHo; Kong, Wang, Song, & Liu, 2016) have been applied. One study with a large sample size (n = 270) found that higher MAAS scores correlated with increased ReHo (local correlations) in the insula, orbito-frontal cortex, and parahippocampal gyrus, and decreased ReHo in the inferior frontal gyrus (Kong et al., 2016). Insular cortex was also implicated in two other studies (Paolini et al., 2012) and Gan and colleagues (2022). Overall, the findings suggest the insula may be less globally connected and more locally connected in more mindful individuals.
Task-based fMRI
Task-based fMRI studies primarily focused on assessing the relationship between trait mindfulness and brain responses to emotion regulation tasks (Table 3).
Emotion Regulation
Functional neuroimaging has provided insights into specific changes in explicit and implicit emotion regulation. Several studies have investigated implicit emotion regulation by providing affective stimuli to participants without any instructions to modulate emotional responses. For example, some of these studies have used the emotional go/no-go task, where participants are asked to respond as quickly as possible to faces with varying emotional expressions while inhibiting responses to faces showing one type of emotion. These types of studies have found that participants with greater trait mindfulness (as measured using the FFMQ) have decreased amygdala and insula responses to emotional stimuli (Kral et al., 2018; Paul, Stanton, Greeson, Smoski, & Wang, 2013; cf. Droutman, Poppa, Monterosso, Black, & Amaro, 2022). Kral and colleagues (2018) examined 31 LTMs and 127 nonmeditators and found that responses to positive stimuli compared with neutral stimuli were attenuated in participants rating higher mindful Nonreactivity (the increased range of scores because of the inclusion of LTM could have influenced the result). Another study demonstrated that during the expectation of both neutral and negative stimuli, there was lower activation in left dorsomedial prefrontal cortex (DMPFC) and right insula as MAAS scores increased (Lutz et al., 2014).
Other studies explicitly prompted participants to regulate their responses to stimuli. In one study, participants were asked to label the emotions of faces (Creswell, Way, Eisenberger, & Lieberman, 2007). High mindfulness (MAAS) individuals showed enhanced activations in prefrontal regions and lesser activations in the amygdala. In addition, task-evoked connectivity between prefrontal and amygdala regions was increased in high trait-mindfulness individuals. The authors interpreted these results as suggesting more top–down regulation of emotion in individuals with greater trait mindfulness (see Torre and Lieberman [2018] for a different interpretation of affect labeling). Other studies have found similar results, including diminished amygdala activations and higher pFC activations (Lutz, 2016; Frewen et al., 2010; Modinos, Ormel, & Aleman, 2010; cf. Martelli, Chester, Brown, Eisenberger, & DeWall, 2018; Lutz et al., 2014). The two studies that did not find this pattern during emotion regulation tasks used the MAAS instead of multifactorial mindfulness scales. Taken together, findings suggest more mindful individuals may show decreased automatic reactivity to affective stimuli and increased capacity to explicitly regulate emotions, although this second finding needs more examination.
Self-awareness
Only one task-based fMRI study directly examined self-referential processing in relation to trait mindfulness. In one study with LTMs, higher FFMQ scores were correlated with the difference between brain responses to self-related feedback versus neutral feedback (Lutz, 2016). Higher FFMQ scores for Nonreactivity correlated with greater activation in the DMPFC in response to self-related feedback (part of the DMN; Buckner et al., 2008). This was interpreted as reflecting decreased cognitive self-reference and rumination, as well as an increased focus on present-moment sensations.
Attention
Individuals with higher trait mindfulness may show differences in network activations during attentional states. For example, one study examined brain activation differences between focused breathing and mind-wandering conditions (Dickenson, Berkman, Arch, & Lieberman, 2013). Greater trait mindfulness (MAAS scores) correlated with increased activation in the superior parietal lobule, TPJ, and dorsolateral pFC (part of the FPN), which was interpreted as increased recruitment of attentional networks. Another study found increased activations in the dorsal attention network and ventral attention network during an attentional control task in more mindful individuals (Paul, 2009). Task-based FPN connectivity may also positively relate to mindfulness, especially in more reliable responders (Kim et al., 2023). These studies suggest that greater attentional mindfulness is associated with greater activation/connectivity across multiple brain regions and networks associated with attention. However, a study with 80 adolescents found trait mindfulness correlated negatively with brain activations in the ventrolateral pFC during a cognitive task (Stein, Bray, MacMaster, Tomfohr-Madsen, & Kopala-Sibley, 2022).
Body Awareness
Three studies involved tasks or meditations with a focus on internal body sensations. One study asking child participants to pay attention to their breathing during the display of negative affect producing video clips found no relationship between DMN activations during the video clips and trait mindfulness (Hehr et al., 2022). Another study compared exteroceptive (sight and sound) versus interoceptive (heart and lungs) attention (Datko, Schuman-Olivier, et al., 2022). Greater trait mindfulness, measured by the FMI, negatively correlated with insular activations during interoceptive attention. The third study found that activation in attentional regions positively correlated with greater trait mindfulness during focused breathing compared with mind-wandering (Dickenson et al., 2013). These studies are too diverse in terms of tasks and ROIs to draw conclusions.
EEG
Emotion Regulation
EEG provides superior temporal resolution for examining emotional responses relative to fMRI (Table 4). A standard ERP paradigm involves presenting participants with affective stimuli and then time-locking and averaging EEG responses (evoked response potentials or ERPs) to the presentations. The amplitude of EEG responses can be compared in different time ranges, where negativities around N100 and N200 typically correspond to bottom–up responses, and late positive potentials (LPPs, 500–1000 msec) correspond to an early top–down response (Foti & Hajcak, 2008; Olofsson, Nordin, Sequeira, & Polich, 2008). One study demonstrated that men with higher MAAS and FFMQ scores showed consistent decreases in the LPP over central and parietal midline sites in response to both negative stimuli and erotica (Brown, Goodman, & Inzlicht, 2013). This was interpreted as consistent with enhanced top–down processing of emotions in more mindful individuals. Two studies replicated this finding (Egan, Hill, & Foti, 2018; Lin, Fisher, Roberts, & Moser, 2016), but other studies have examined the LPP and found null results (Zhang, Ouyang, Tang, Chen, & Li, 2019; Cosme & Wiens, 2015; Ho, Sun, Ting, Chan, & Lee, 2015) or positive relationships (Eddy, Brunyé, Tower-Richardi, Mahoney, & Taylor, 2015). Discrepancies are not clearly linked to the type of mindfulness scale used. Many studies used mindfulness inductions—short mindfulness practices—before the presentation of the affective stimuli (Zhang et al., 2019; Lin et al., 2016; Eddy et al., 2015; Ho et al., 2015). One interpretation proposed in the studies is that higher trait mindfulness individuals may paradoxically be less affected by the inductions, thus erasing the relationship between trait mindfulness and the LPP or other responses (Lin, White, Viravan, & Braver, 2024; Lin et al., 2016).
The relationship between trait mindfulness and early bottom–up responses is also unclear. One study using an emotional go/no-go task found that early startle responses were smaller in more mindful individuals (specifically, smaller P300s correlated with higher Nonreactivity; Cosme & Wiens, 2015), whereas others found greater N200s in more mindful individuals as measured by MAAS (Zhang et al., 2019; Quaglia, Goodman, & Brown, 2016). Speculatively, the MAAS may reflect attention and awareness, and hence, heightened early negativities. Conversely, Nonreactivity may correlate with smaller early positivities. Finally, a study of over 200 individuals examined neuro-affective error processing (the error-related negativity and error positivity, which may, respectively, represent unconscious and conscious responses) and found no relationship with trait mindfulness for the total FFMQ or any subscales (Lin, Eckerle, Peng, & Moser, 2019). The researchers interpreted this as reflecting self-report limitations.
Self-awareness
No EEG studies presented individuals with self-relevant stimuli.
Attention
EEG studies of attention and mindfulness are diverse—involving meditation, cognitive tasks, and other tasks. A study of the flanker task examined the P3/P300 response to incongruent (>><>>) versus congruent stimuli (<<<<<), which reflects attentional inhibition (Lin, Fisher, & Moser, 2019). The researchers found that Acting with Awareness negatively correlated with decreased P300s to incongruent stimuli, and decreased P300s mediated differences in behavioral responses (errors and RTs). This study is valuable because it specifies a unique neural mechanism associated with a subfacet of mindfulness and empirically evaluates how it relates to behavior. A subsequent study however, Lin and colleagues (2024), did not support this mechanism, finding global (independent of flanker trial type) P300s correlated with a trait mindfulness composite based on the FFMQ. Notably, this correlation was negative after a meditation induction and positive after a control induction. It is possible that ERP correlates of FFMQ overall and FFMQ-Acting with Awareness/MAAS are distinct. Lastly, two reports examined EEG power differences during cognitive tasks and resting state, respectively, between a high-trait mindful and low anxiety group and a low-trait mindful and high anxiety group (Jaiswal et al., 2023; Jaiswal, Tsai, Juan, Muggleton, & Liang, 2019). Among other findings, the researchers found increased alpha power in prefrontal regions during the task in the high-trait mindful and low anxiety group, and decreased frontotemporal gamma power during the resting state. It is hard to disentangle whether the results are because of anxiety differences or mindfulness differences. These studies on cognitive tasks and mindfulness are hard to integrate given their variable EEG measures, use of inductions, and mindfulness measures. Findings may indicate that more mindful individuals show increased attentional recruitment or, alternatively, more efficient processing.
Studies have also examined attention in comparisons of focused attention meditation and rest. One study of EEG microstates found that higher FFMQ scores, specifically Nonreactivity, negatively correlated with time in a state associated with attention-switching mediated by SN (Zarka et al., 2022). Zarka and colleagues (2024) replicated this finding in a resting-state paradigm with interspersed auditory stimuli and, additionally, found decreased time in this state in mindfulness-trained participants. Other studies have investigated the power of various EEG frequency bands in LTMs (Krishna et al., 2022; Hinterberger, Kohls, Kamei, Feilding, & Walach, 2011)—but differences in scale interpretation as well as LTM samples make generalizations difficult.
Body Awareness
DISCUSSION
Summary
In this review, we synthesized studies that related trait mindfulness to various measures of brain structure and function. Study designs included structural MRI, resting-state fMRI, task-based fMRI, and EEG. Structural MRI studies investigated how gray matter surface area, thickness, density, and volume metrics correlated with trait mindfulness. Resting-state fMRI designs investigated how brain areas correlate in the absence of tasks and how these patterns of correlations evolve over time (DFC). Task-based fMRI designs presented participants with stimuli (e.g., emotional faces) and/or behavioral instructions (e.g., “focus on your breathing”) and examined how activations correlated with trait mindfulness. Lastly, EEG-based designs examined event-related responses to stimuli, as well as changes in power in different frequency bands. By systematically reviewing all these imaging designs, we aimed to identify consistent neural correlates of trait mindfulness.
There were several relatively consistent findings (Figure 3). First, trait mindfulness is associated with reduced bottom–up reactivity to emotional stimuli in structures like the amygdala (Kral et al., 2018; Lutz et al., 2014; Paul et al., 2013; cf. Droutman et al., 2022). It seems that nonreactivity may be the foundation of mindful emotion regulation. Mindful nonreactivity eschews the evaluation and reification of emotional experience, allowing attention to be placed on the transitory nature of the present moment and emotions. Second, a preponderance of evidence suggests that trait mindfulness is associated with reduced within-DMN connectivity (Pasquini et al., 2023; Li et al., 2022; Harrison et al., 2019; Bilevicius et al., 2018; Doll et al., 2015; Wang et al., 2014; cf. Hunt et al., 2022; Shaurya Prakash et al., 2013), which has implications for understanding self-awareness and mind-wandering in the context of mindfulness. Third, cortical areas involved in attentional control (e.g., pFC; Zhuang et al., 2017; Makovac et al., 2016; Lu et al., 2014) and body awareness (e.g., insula; Baltruschat et al., 2021; Friedel et al., 2015; Murakami et al., 2012) may be thicker in more mindful individuals. These results may be considered criterion evidence for trait mindfulness—evidence that self-reported trait mindfulness indexes real objective measures (“criteria”) relevant to emotion regulation and other psychological mechanisms.
There are many open questions, however. First, despite a large number of studies (23), EEG studies of trait mindfulness were inconclusive. Perhaps this is because of variable interactions with mindfulness inductions. Another possibility is the “reliability paradox” first identified in the study of behavioral tasks (Hedge, Powell, & Sumner, 2018). The reliability paradox refers to robust effects observed between conditions that do not produce reliable estimates of individual differences. In the context of this review, ERPs like the N200 or P300 may be selected a priori because of their robust differentiation of experimental conditions. However, by virtue of this robustness, they may constrain individual variability, leading to inconsistent relationships with trait mindfulness. Another area of ongoing debate is the interactions between attentional networks like the DMN, SN, and FPN. A couple of trait mindfulness studies indicated decreased DMN–SN connectivity correlated with increased mindfulness (Bilevicius et al., 2018; Doll et al., 2015; cf. Parkinson et al., 2019). Anticorrelations between DMN-FPN, typically found to correlate with objective cognitive performance (Hellyer et al., 2014), were not consistently found in trait mindfulness studies.
Regarding differential relationships for mindfulness operationalizations, most, but not all, studies used FFMQ and MAAS and found similar results across both scales. Early EEG responses may be greater in more attentive individuals (as measured by the MAAS; Zhang et al., 2019; Quaglia, Goodman, et al., 2016), and weaker in more nonreactive individuals (as measured by the FFMQ Nonreactivity; Cosme & Wiens, 2015). In studies employing emotion regulation tasks that may involve reappraisal, however, the MAAS was not related to pFC or amygdala responses (Martelli et al., 2018; Lutz et al., 2014), whereas multifactorial studies like the FFMQ were (Lutz, 2016; Frewen et al., 2010; Modinos et al., 2010). This could reflect the FFMQ's focus on emotional experience. The Observing subscale of the FFMQ, which asks questions about experiencing sensations in the body, was not often utilized in neuroimaging studies on body awareness.
Convergences with Mindfulness Training
Before reviewing converging evidence from mindfulness training, it may be noted that cross-sectional studies comparing experienced (LTM) and inexperienced meditators and longitudinal studies randomizing inexperienced meditators to mindfulness training have limitations. Briefly, mindfulness interventions do not always use rigorous, active-controlled designs; meditation duration, instructions, and so forth may vary study-to-study; and small sample sizes are common (Davidson & Kaszniak, 2015). Cross-sectional LTM studies may be confounded by individual-level predispositions of who chooses, and who does not choose, to practice meditation, heterogeneous meditation practices, and variable definitions of “long-term” practice (Luders & Kurth, 2019). Nevertheless, multiple studies have reported higher levels of mindfulness (as measured by MAAS and FFMQ) in LTMs (Baer et al., 2008; Brown & Ryan, 2003; cf. Christopher & Gilbert, 2007) and following mindfulness training relative to a control group (Quaglia, Braun, et al., 2016). Thus, brain differences in these studies could converge with trait mindfulness studies that are independent of meditation experience.
Convergences in Structural MRI
We found some convergent evidence that cortical structures like pFC and insula may be thicker in relation to greater trait mindfulness. Increased gray matter thickness in prefrontal attentional control areas has also been discovered in LTM (Fox et al., 2014; Hölzel et al., 2008; Lazar et al., 2005). Enhanced thickness of insular cortex has also been observed in LTM (Luders et al., 2012; Hölzel et al., 2008; Lazar et al., 2005) and in intervention studies (Siew & Yu, 2023; Santarnecchi et al., 2014; cf. Kral et al., 2022). Several studies in our review noted associations between the structure of DMN regions and trait mindfulness (Baltruschat et al., 2021; Zhuang et al., 2017; Lu et al., 2014). Multiple studies of LTMs have found increased volumes of DMN areas compared with controls (Kang et al., 2013; Luders, Toga, Lepore, & Gaser, 2009; Vestergaard-Poulsen et al., 2009; Hölzel et al., 2008). Moreover, intervention studies revealed increases in DMN gray matter density (Kurth, Luders, Wu, & Black, 2014; Pickut et al., 2013; Wells et al., 2013; Hölzel, Carmody, et al., 2011). Gray matter increases may be most prevalent in the hippocampus (Fox et al., 2014).
We did not find consistent associations with amygdala structure. An MBSR intervention study reported decreases in amygdala volume compared with controls (Hölzel et al., 2010), and a large-scale study with >3000 adults found amygdala volume was negatively associated with meditation and yoga experience (surprisingly, practitioners also reported more stress and depression symptoms; Gotink et al., 2018). Similar population-level studies should assess neural relations of trait mindfulness, perhaps using open-source, maintained cortical parcellation and subcortical segmentation analysis tools such as Freesurfer and Mindboggle (Klein et al., 2017; Fischl, 2012).
Convergences in Resting-state fMRI
A majority (six of eight) of studies reported decreases in resting-state within-DMN connectivity, perhaps underlying more experiential and less narrative self-awareness. These findings generally align with LTM and intervention studies, which evidenced both decreases in connectivity (Zhang et al., 2023; Bauer, Whitfield-Gabrieli, Díaz, Pasaye, & Barrios, 2019; Taylor et al., 2013; cf. Jang et al., 2011) and in activations (Garrison, Zeffiro, Scheinost, Constable, & Brewer, 2015; Brewer et al., 2011) of the DMN. This could also reflect reduced mind-wandering (Bauer et al., 2019).
The relation of trait mindfulness to across-network connectivity is less clear, even when examining frequently reported networks like the DMN, SN, and FPN. Two small sample studies found decreasing DMN–SN connectivity in relation to increasing trait mindfulness (Bilevicius et al., 2018; Doll et al., 2015; cf. Parkinson et al., 2019). However, a meta-analysis of 12 mindfulness intervention studies demonstrated increased DMN–SN connectivity compared with control interventions and interpreted this as supporting alertness and sustained attention (Rahrig et al., 2022). On the other hand, a study of dynamic connectivity in adults who meditated in the scanner before and after a mindfulness intervention demonstrated longitudinal increases in prevalence of a brain state characterized by DMN anticorrelations with the SN and FPN (Mooneyham et al., 2017). These results may differ between trait-like static connectivity, state-like dynamic connectivity, and between tasks and rest.
Convergences in Task-based fMRI
The most consistent finding from our review of task-based fMRI was that more mindful individuals showed diminished amygdala responses to emotional stimuli. This relationship, interpreted as diminished bottom–up reactivity, has also been found following mindfulness interventions (Kral et al., 2018; Leung et al., 2018; Desbordes et al., 2012), and for both positively and negatively valenced stimuli. We also examined the relationship of trait mindfulness to top–down regulation, for example, through reappraisal, with some indications of a positive relationship with pFC activations. A systematic review of mindfulness training (MBSR) studies suggests the same relationship—training increased pFC activations (Gotink, Meijboom, Vernooij, Smits, & Hunink, 2016). On the other hand, it is possible that only beginner meditators experience this increased frontal involvement in emotion regulation (Cooper et al., 2022; Tang et al., 2015; Taylor et al., 2011).
We also reviewed studies of fMRI activations during cognitive tasks and found discrepant results. It is unclear whether more mindful individuals show diminished, more efficient attentional processing, or increased recruitment. It is possible that there are nonlinearities, for example, U-curves. In one study, LTMs showed increased activation of areas involved in sustained attention compared with naive controls, but very experienced LTMs (>44,000 hr) demonstrated decreased activations compared with controls (Brefczynski-Lewis, Lutz, Schaefer, Levinson, & Davidson, 2007).
Convergences in EEG
EEG findings were inconsistent in relation to trait mindfulness. Findings from LTM and intervention studies are also mixed. One study found decreased LPPs to negative stimuli in meditators (Sobolewski, Holt, Kublik, & Wróbel, 2011), another study found increased LPPs to rewarding stimuli after a mindfulness intervention (Garland, Froeliger, & Howard, 2015), and a final study in meditators found that the difference between LPPs to unpleasant and pleasant stimuli was attenuated compared with controls (Katyal, Hajcak, Flora, Bartlett, & Goldin, 2020).
Summary of Convergent Validity
It is proposed that long-term practice leads to trait-level neural changes (Sezer et al., 2022; Bauer et al., 2019), but these are not necessarily the same as those assessed by self-report (Lutz, 2016). Our results suggest some correspondences between trait mindfulness and short-term and long-term training. Increased cortical thickness in the insula is repeatedly linked to mindfulness meditation (both in LTMs and interventions). DMN activation decreases are found during mindfulness meditation as well as connectivity decreases. Amygdala reactivity to emotional stimuli is decreased after interventions. There are indications that the inconsistency in relationships between EEG responses and trait mindfulness parallel variable EEG findings in LTMs and interventions, although a full review of these latter designs was not feasible here.
Future Directions
Methodological Paradigms
Paradigm shifts may be necessary in correlational studies. Sample sizes may need to increase, but to what extent? Despite numerous studies (23), some with sample sizes between 100–200 (e.g., Lin, Eckerle, et al., 2019), EEG findings remained inconsistent. Relatedly, one study found that to detect replicable associations between individual differences in fMRI and structural MRI and behavioral measures, thousands of participants are needed (Marek et al., 2022). None of the articles included here were conducted on such a population level. This could explain discrepancies in the trait mindfulness literature and the trait literature more broadly (anxiety; Montag, Reuter, Jurkiewicz, Markett, & Panksepp, 2013; well-being; de Vries, van de Weijer, & Bartels, 2023). Even in the study of the well-established Big Five personality metrics, there are few consistent brain associations (Chen & Canli, 2022; Lai et al., 2019). Besides recruiting more participants, there are several ways forward.
Innovations might be possible on the side of self-reporting. One approach is to consider only participants with reliable mindfulness scores over time (Kim et al., 2023). These participants may have relatively stable self-narratives/self-concepts, or alternatively, they simply find the mindfulness questionnaires easy to interpret. Either way, variation in this subset of participants may be more meaningfully related to brain measures. Another approach is discarding the factor structure of mindfulness questionnaires and performing data-driven analyses (e.g., canonical correlation analysis) on the item level. It is possible that different combinations of self-report items are related to brain outcomes than those that have been validated psychometrically. This approach is increasingly adopted with personality measures (Stachl et al., 2020) and with behavioral cognitive measures (e.g., Eisenberg et al., 2019).
Innovations in neuroimaging analysis should also be considered. In some cases, experimenter degrees of freedom are not adequately addressed. For example, EEG ERP measures differ in latency, amplitudes, and spatial location (Šoškić, Styles, Kappenman, & Ković, 2024; Luck & Gaspelin, 2017). Exploring degrees of freedom in one data set and testing the model in an independent data set may address such concerns. This out-of-sample prediction may provide more accurate measurements of trait mindfulness–brain associations instead of correlational methods (Treves, Kucyi, et al., 2024; Rosenberg & Finn, 2022). Researchers may also consider eschewing simple ROI/network-behavior analysis and adopt multivariate approaches, for example, learning weights over a set of brain connectivity or activation features (Gratton, Nelson, & Gordon, 2022). Lastly, reliability modeling of brain measures (Treves, Marusak, et al., 2024) could provide more robust measures.
Novel Questions
Large gaps in the literature are present in the fields of self-awareness and trait mindfulness, and body awareness and trait mindfulness. Self-awareness is often discussed in the context of emotion regulation (e.g., decentering leads to differences in regulation, Hayes & Feldman, 2004), and DMN activations are often related to self-awareness based on reverse inference (e.g., Hehr et al., 2022). However, rarely in this review were participants asked to specifically respond to self-relevant adjectives. Tasks like the self-referential encoding task could be used in future studies. Studying body awareness may require more innovation. Mindfulness practices often entail focusing on sensations throughout the body, and repeated practice may lead to changes in body awareness (Treves et al., 2019; Farb et al., 2015). Individuals reporting high body awareness may also report high trait mindfulness (Hanley, Mehling, & Garland, 2017). Despite these relationships, trait mindfulness questionnaires do not focus on body awareness, preventing researchers from operationalizing this dimension and thus studying it with brain imaging. We suggest that for further clarity, researchers should explore items on the Observing subscale of the FFMQ or mindfulness-specific items on the multidimensional assessment of interoceptive awareness (Mehling et al., 2012) and conduct imaging concurrently with a variety of body awareness tasks (for a review, see Treves et al., 2019).
In general, differential knowledge of the specific subcomponents of self-report mindfulness is scarce. One exception is a study where structural correlates of two mindfulness scales were assessed (Zhuang et al., 2017). The MAAS was related to brain structures involved in self-awareness (e.g., precuneus), and the FFMQ was related to brain structures involved in emotion regulation (e.g., superior pFC). Furthermore, the MAAS mediated the relationship between brain structures and depression symptoms, whereas the FFMQ mediated the relationship between brain structures and cognitive reappraisal. The authors interpreted this as confirmatory evidence that the MAAS is related to attention control and self-awareness, and the FFMQ is related to emotion regulation.
Our ability to study the brain is limited by the quality of our operationalizations of mindfulness. We need deeper theoretical insight into the meditative traditions that inspire mindfulness questionnaires (Wright, Sanguinetti, Young, & Sacchet, 2023; Sparby & Sacchet, 2022). To this end, we propose the study of advanced meditation, that is, states and stages of practice beyond the development of basic mindfulness skills and including meditative endpoints of practice (Sacchet, Fava, & Garland, 2024; Yang, Sparby, Wright, Kim, & Sacchet, 2024). In this context, instead of studying a monolithic construct of trait mindfulness that is thought to account for change across all levels of training, to rather move toward a developmental model that highlights the process of unfolding of the states and stages of advanced meditation that might unfold in time and with mastery (Galante et al., 2023).
Limitations
We identified 68 studies through our systematic search. However, given the breadth of the research question, we may not have identified every study. This means that our claims about consistent brain correlates of trait mindfulness should be taken with caution. In addition, many studies incorporate brain–mindfulness associations as exploratory or secondary to central analyses on mindfulness meditation. It may be that correlational findings that do not align with dominant mindfulness theories are not reported. Heterogeneity may be another key limitation. Studies here were included without regard to age range, meditation experience, or clinical condition. As trait mindfulness scales may be differently interpreted based on these factors (Goodman, Madni, & Semple, 2017; Curtiss & Klemanski, 2014; Christopher et al., 2009), this could limit the consistency of our neural findings.
Other limitations relate to the brain imaging designs. No studies related structural differences associated with trait mindfulness to functional differences. For example, do more mindful individuals with thicker insulas exhibit greater activations during an emotion regulation task? As it stands, we must rely on reverse inferences to interpret many of the structural results. Reverse inferences involve using literature that links tasks or behaviors to brain activations (e.g., fearful faces show increased amygdala activations [Morris et al., 1998]) and then interpreting differences in brain activations or structure as relating to those behaviors (e.g., increased amygdala sizes reflect increased fearful behavior). This approach can be misleading if the brain area in question is associated with multiple behaviors (Poldrack, 2006). Even the link between brain activations and brain structure is fraught; for example, smaller amygdalae may result in greater activations (Gianaros et al., 2008). This could explain some counterintuitive results; trait mindfulness has been associated with greater surface area and volume of DMN areas (Zhuang et al., 2017) but also decreased activations and/or connectivity of those areas (e.g., Harrison et al., 2019). In summary, our efforts to interpret brain differences as relating to psychological functions should be taken cautiously.
A final limitation is that we could not identify the incremental validity of trait mindfulness. There is some evidence that mindfulness scales may not predict well-being outcomes above and beyond Big Five traits like neuroticism (Altgassen et al., 2024). It is possible that the neural correlates we have identified here are not unique to mindfulness. Future studies should control for related traits in their models.
Implications for Clinical Outcomes
As we have just described, there are hurdles (e.g., sample sizes) to a robust neuroscience of trait mindfulness. Nevertheless, this work is needed to illuminate relations between trait mindfulness and mental health. Higher trait mindfulness is associated with improved clinical outcomes, for example, reduced symptoms of anxiety and depression (Carpenter et al., 2019; Tomlinson et al., 2018; Arch & Craske, 2010). For this reason, trait mindfulness is often called a “promotive” factor (predictor of a positive outcome, regardless of the level of risk; Masten, 2013). How does mindfulness promote positive mental health outcomes? Longitudinal studies are needed to establish precedence (e.g., individuals with lower mindfulness at Timepoint 1 exhibit higher depression symptoms at Timepoint 2). However, mediational analyses can test whether brain–behavior links are consistent with causal models.
In one study included in this review, the MAAS was found to mediate relations between precuneus volume and scores on the Beck Depression Inventory (Zhuang et al., 2017). In this case, the mediation modeling suggested that precuneus volume influenced trait mindfulness, which then influenced depression symptoms. The authors interpreted this effect as pertaining to self-awareness. Of course, mediation models may vary based on the type of brain measure. In a resting-state study, ReHo of activations in orbito-frontal cortex mediated the relationship between trait mindfulness and positive affect (Kong et al., 2016). Their mediation model implicitly assumes that trait mindfulness' stable, intrinsic component (which could be because of brain structure) influences brain function (ReHo). This mediation outcome indicates that more mindful individuals show higher ReHo in the orbito-frontal cortex, leading to more positive affect.
In most cases, correlations between trait mindfulness and mental health will not be attributable to a specific causal mechanism. Regardless, correlational results could contribute to the development of brain-based biomarkers. A brain-based biomarker is typically an individual difference measure of brain structure or function associated with clinical outcomes (Gabrieli, Ghosh, & Whitfield-Gabrieli, 2015). As mindfulness is associated with positive mental health outcomes, so might individual brain differences related to mindfulness. For example, this review identified a positive relationship between gray matter thickness in the insula and trait mindfulness. Thus, insula gray matter thickness may be a useful biomarker for predicting mental health outcomes. Another example is DMN hyperconnectivity, which is inversely associated with trait mindfulness. Large-scale resting state or structural imaging efforts could assess the population distribution of these brain biomarkers. With this information, individuals who are outliers in the distribution could receive clinical assessment. It should be noted that this does not replace the utility of the trait measures, but it complements it (biomarkers can be assessed regardless of barriers to self-report such as introspective ability, language, etc.). It should be noted that cutting-edge multimodal biomarkers of clinical conditions like depression still perform far below clinical assessments (Winter et al., 2024).
Lastly, it remains unknown which specific aspects of trait mindfulness promote positive clinical outcomes. It could be fruitful to divide the multifaceted construct of mindfulness into components relevant to mental health. This work should consider the neural mechanisms we have identified here, aligning with the Research Domain Criteria framework (Cuthbert & Insel, 2013). The Research Domain Criteria framework aims to characterize individuals not on symptom clusters but in terms of neuropsychological domains—consisting of a behavioral pattern (e.g., reward learning) and a corresponding neural circuit (e.g., corticostriatal connections). The neuropsychological subdomains of mindfulness could be used to provide precise and tailored mindfulness interventions to the needs of individuals. For example, if an individual were to present with low levels of body awareness (which often co-occurs with depression [Smith et al., 2021; Khalsa et al., 2018]), it may be the case that body-focused mindfulness practices like breath counting and body scans may be particularly helpful. Another individual with relatively higher levels of body awareness and higher levels of catastrophizing (or “sensibility” [Garfinkel et al., 2015]) may instead benefit from mindfulness practices that focus on nonjudgment and acceptance. Mindfulness interventions are often tailored to specific clinical populations, for example, mindfulness-based cognitive therapy for depression (Segal et al., 2018) or MYmind for parents and their children with autism (de Bruin, Blom, Smit, van Steensel, & Bögels, 2015). These important efforts could be bolstered by trait mindfulness research.
APPENDIX A: NEUROSYNTH FIGURES
Emotion regulation brain regions from Neurosynth. Brain areas implicated during emotional processes generated via an automatic meta-analysis of 247 studies available from Neurosynth (https://neurosynth.org/). The figure depicts the uniformity mask corresponding to the term “emotion regulation” thresholded at a z value > 3.1. x, y, and z Correspond to coordinates in Montreal Neurological Institute space. Bilateral amygdala clusters are shown in the axial (rightmost) plot. A = anterior; L = left; P = posterior; R = right.
Emotion regulation brain regions from Neurosynth. Brain areas implicated during emotional processes generated via an automatic meta-analysis of 247 studies available from Neurosynth (https://neurosynth.org/). The figure depicts the uniformity mask corresponding to the term “emotion regulation” thresholded at a z value > 3.1. x, y, and z Correspond to coordinates in Montreal Neurological Institute space. Bilateral amygdala clusters are shown in the axial (rightmost) plot. A = anterior; L = left; P = posterior; R = right.
Self-awareness brain regions from Neurosynth. Brain areas implicated during self-awareness or self-referential processes generated via an automatic meta-analysis of 166 studies available from Neurosynth (https://neurosynth.org/). The figure depicts the uniformity mask corresponding to the term “self-reference” thresholded at a z value > 3.1. x, y, and z Correspond to coordinates in Montreal Neurological Institute space. A = anterior; L = left; P = posterior; R = right.
Self-awareness brain regions from Neurosynth. Brain areas implicated during self-awareness or self-referential processes generated via an automatic meta-analysis of 166 studies available from Neurosynth (https://neurosynth.org/). The figure depicts the uniformity mask corresponding to the term “self-reference” thresholded at a z value > 3.1. x, y, and z Correspond to coordinates in Montreal Neurological Institute space. A = anterior; L = left; P = posterior; R = right.
Attentional brain regions from Neurosynth. Brain areas implicated during attentional processes generated via an automatic meta-analysis of 1831 studies available from Neurosynth (https://neurosynth.org/). The figure depicts the uniformity mask corresponding to the term “attention” thresholded at a z value > 3.1. x, y, and z Correspond to coordinates in Montreal Neurological Institute space. A = anterior; L = left; P = posterior; R = right.
Attentional brain regions from Neurosynth. Brain areas implicated during attentional processes generated via an automatic meta-analysis of 1831 studies available from Neurosynth (https://neurosynth.org/). The figure depicts the uniformity mask corresponding to the term “attention” thresholded at a z value > 3.1. x, y, and z Correspond to coordinates in Montreal Neurological Institute space. A = anterior; L = left; P = posterior; R = right.
Body awareness brain regions from Neurosynth. Brain areas implicated during body awareness processes generated via an automatic meta-analysis of 81 studies available from Neurosynth (https://neurosynth.org/). The figure depicts the uniformity mask corresponding to the term “interoception” (perception of internal bodily sensations) thresholded at a z value > 3.1. x, y, and z Correspond to coordinates in Montreal Neurological Institute space. A = anterior; L = left; P = posterior; R = right.
Body awareness brain regions from Neurosynth. Brain areas implicated during body awareness processes generated via an automatic meta-analysis of 81 studies available from Neurosynth (https://neurosynth.org/). The figure depicts the uniformity mask corresponding to the term “interoception” (perception of internal bodily sensations) thresholded at a z value > 3.1. x, y, and z Correspond to coordinates in Montreal Neurological Institute space. A = anterior; L = left; P = posterior; R = right.
APPENDIX B: STUDY SELECTION FLOWCHART
PRISMA flow diagram depicting number of identified and evaluated articles for concurrent mindfulness and fMRI or EEG neurofeedback procedures. From: Page et al. (2021).
PRISMA flow diagram depicting number of identified and evaluated articles for concurrent mindfulness and fMRI or EEG neurofeedback procedures. From: Page et al. (2021).
APPENDIX C: PRISMA 2020 CHECKLIST
PRISMA Reporting Checklist
Section and Topic . | Item No. . | Checklist Item . | Location Where Item Is Reported . |
---|---|---|---|
TITLE | |||
Title | 1 | Identify the report as a literature review. | pg. 1 |
ABSTRACT | |||
Abstract | 2 | Provide a structured summary including, as applicable: background; objectives; data sources; study eligibility criteria, participants, and interventions; study appraisal and synthesis methods; results; limitations; conclusions and implications of key findings. See the PRISMA 2020 for Abstracts checklist for the complete list. | pg. 1 (structured abstract not allowed) |
INTRODUCTION | |||
Rationale | 3 | Describe the rationale for the review in the context of existing knowledge, i.e., what is already known about your topic. | pg. 3 par. 3, pg. 6 (The present review) |
Objectives | 4 | Provide an explicit statement of the objective(s) or question(s) the review addresses with reference to participants, interventions, comparisons, outcomes, and study design (PICOS). | pg. 3 par. 3, pg. 6 (The present review) |
METHODS | |||
Eligibility criteria | 5 | Specify the inclusion and exclusion criteria for the review and how studies were grouped for the syntheses with study characteristics (e.g., PICOS, length of follow-up) and report characteristics (e.g., years considered, language, publication status) used as criteria for eligibility, giving rationale. | pg. 6 par. 3 |
Information sources | 6 | Specify all databases, registers, websites, organizations, reference lists and other sources searched or consulted to identify studies. Specify the date when each source was last searched or consulted. | pg. 6 par. 4 |
Search strategy | 7 | Present the full search strategies for all databases, registers, and websites, including any filters and limits used. | pg. 6 par. 4 |
Selection process | 8 | State the process for selecting studies (i.e., screening, eligibility). Specify the methods used to decide whether a study met the inclusion criteria of the review, including how many reviewers screened each record and each report retrieved, whether they worked independently, and if applicable, details of automation tools used in the process. | pg. 6 par. 4 |
Study risk of bias assessment | 11 | Specify the methods used to assess risk of bias in the included studies, including details of the tool(s) used, how many reviewers assessed each study and whether they worked independently, and if applicable, details of automation tools used in the process. | N/A (not RCTs or manipulations) |
RESULTS | |||
Study selection | 16a | Describe the results of the search and selection process, from the number of records identified in the search to the number of studies included in the review, ideally using a flow diagram. | Appendix B1 |
16b | Cite studies that might appear to meet the inclusion criteria, but which were excluded, and explain why they were excluded. | N/A | |
Study characteristics | 17 | Cite each included study and present its characteristics (e.g., study size, PICOS, follow-up period). | Tables 1–4 |
Risk of bias in studies | 18 | Present assessments of risk of bias for each included study. | N/A (see Item 11) |
Results of individual studies | 19 | For all outcomes, present, for each study: (a) summary statistics for each group (where appropriate) and (b) an effect estimate and its precision (e.g., confidence/credible interval), ideally using structured tables or plots. | N/A (effect sizes are not comparable) |
DISCUSSION | |||
Discussion | 23a | Provide a general interpretation of the results in the context of other evidence. | pg. 18 |
23b | Discuss any limitations of the evidence included in the review. | pg. 21, Limitations | |
23c | Discuss any limitations of the review processes used. | pg. 21, Limitations | |
23d | Discuss implications of the results for practice, policy, and future research. | Pg. 20, Future directions, pg. 21 Implications for Clinical Research | |
OTHER INFORMATION | |||
Registration and protocol | 24a | Provide registration information for the review, including register name and registration number, or state that the review was not registered. | N/A |
24b | Indicate where the review protocol can be accessed, or state that a protocol was not prepared. | N/A | |
24c | Describe and explain any amendments to information provided at registration or in the protocol. | N/A | |
Support | 25 | Describe sources of financial or nonfinancial support for the review, and the role of the funders or sponsors in the review. | pg. 28 |
Competing interests | 26 | Declare any competing interests of review authors. | pg. 28 |
Availability of data, code, and other materials | 27 | Report which of the following are publicly available and where they can be found: template data collection forms; data extracted from included studies; data used for all analyses; analytic code; any other materials used in the review. | N/A |
Section and Topic . | Item No. . | Checklist Item . | Location Where Item Is Reported . |
---|---|---|---|
TITLE | |||
Title | 1 | Identify the report as a literature review. | pg. 1 |
ABSTRACT | |||
Abstract | 2 | Provide a structured summary including, as applicable: background; objectives; data sources; study eligibility criteria, participants, and interventions; study appraisal and synthesis methods; results; limitations; conclusions and implications of key findings. See the PRISMA 2020 for Abstracts checklist for the complete list. | pg. 1 (structured abstract not allowed) |
INTRODUCTION | |||
Rationale | 3 | Describe the rationale for the review in the context of existing knowledge, i.e., what is already known about your topic. | pg. 3 par. 3, pg. 6 (The present review) |
Objectives | 4 | Provide an explicit statement of the objective(s) or question(s) the review addresses with reference to participants, interventions, comparisons, outcomes, and study design (PICOS). | pg. 3 par. 3, pg. 6 (The present review) |
METHODS | |||
Eligibility criteria | 5 | Specify the inclusion and exclusion criteria for the review and how studies were grouped for the syntheses with study characteristics (e.g., PICOS, length of follow-up) and report characteristics (e.g., years considered, language, publication status) used as criteria for eligibility, giving rationale. | pg. 6 par. 3 |
Information sources | 6 | Specify all databases, registers, websites, organizations, reference lists and other sources searched or consulted to identify studies. Specify the date when each source was last searched or consulted. | pg. 6 par. 4 |
Search strategy | 7 | Present the full search strategies for all databases, registers, and websites, including any filters and limits used. | pg. 6 par. 4 |
Selection process | 8 | State the process for selecting studies (i.e., screening, eligibility). Specify the methods used to decide whether a study met the inclusion criteria of the review, including how many reviewers screened each record and each report retrieved, whether they worked independently, and if applicable, details of automation tools used in the process. | pg. 6 par. 4 |
Study risk of bias assessment | 11 | Specify the methods used to assess risk of bias in the included studies, including details of the tool(s) used, how many reviewers assessed each study and whether they worked independently, and if applicable, details of automation tools used in the process. | N/A (not RCTs or manipulations) |
RESULTS | |||
Study selection | 16a | Describe the results of the search and selection process, from the number of records identified in the search to the number of studies included in the review, ideally using a flow diagram. | Appendix B1 |
16b | Cite studies that might appear to meet the inclusion criteria, but which were excluded, and explain why they were excluded. | N/A | |
Study characteristics | 17 | Cite each included study and present its characteristics (e.g., study size, PICOS, follow-up period). | Tables 1–4 |
Risk of bias in studies | 18 | Present assessments of risk of bias for each included study. | N/A (see Item 11) |
Results of individual studies | 19 | For all outcomes, present, for each study: (a) summary statistics for each group (where appropriate) and (b) an effect estimate and its precision (e.g., confidence/credible interval), ideally using structured tables or plots. | N/A (effect sizes are not comparable) |
DISCUSSION | |||
Discussion | 23a | Provide a general interpretation of the results in the context of other evidence. | pg. 18 |
23b | Discuss any limitations of the evidence included in the review. | pg. 21, Limitations | |
23c | Discuss any limitations of the review processes used. | pg. 21, Limitations | |
23d | Discuss implications of the results for practice, policy, and future research. | Pg. 20, Future directions, pg. 21 Implications for Clinical Research | |
OTHER INFORMATION | |||
Registration and protocol | 24a | Provide registration information for the review, including register name and registration number, or state that the review was not registered. | N/A |
24b | Indicate where the review protocol can be accessed, or state that a protocol was not prepared. | N/A | |
24c | Describe and explain any amendments to information provided at registration or in the protocol. | N/A | |
Support | 25 | Describe sources of financial or nonfinancial support for the review, and the role of the funders or sponsors in the review. | pg. 28 |
Competing interests | 26 | Declare any competing interests of review authors. | pg. 28 |
Availability of data, code, and other materials | 27 | Report which of the following are publicly available and where they can be found: template data collection forms; data extracted from included studies; data used for all analyses; analytic code; any other materials used in the review. | N/A |
Adapted from: Page et al. (2021).
For more information, visit: https://www.prisma-statement.org/.
APPENDIX D: MEAN SAMPLE SIZES OVER TIME FOR IMAGING MODALITIES
Mean Sample Sizes over Brain Imaging Modalities Based on Time of Publication
. | 2000–2014 . | 2015–Present . |
---|---|---|
EEG | 35 | 69.65 |
task-based fMRI | 26.25 | 52.91 |
rsfMRI | 76 | 67.54 |
sMRI | 140.33 | 103.17 |
. | 2000–2014 . | 2015–Present . |
---|---|---|
EEG | 35 | 69.65 |
task-based fMRI | 26.25 | 52.91 |
rsfMRI | 76 | 67.54 |
sMRI | 140.33 | 103.17 |
Means were calculated for all studies in a reported timeframe, ignoring repeated studies. rsfMRI = resting-state fMRI; sMRI = structural MRI.
Corresponding author: Isaac N. Treves, Massachusetts Institute of Technology, 46-4037 Vassar Street, Cambridge, MA 02139, or via e-mail: [email protected].
Author Contributions
Isaac N. Treves: Conceptualization; Investigation; Project administration; Supervision; Visualization; Writing—Original draft; Writing—Review & editing. Kannammai Pichappan: Investigation; Writing—Review & editing. Jude Hammoud: Writing—Review & editing. Clemens C. C. Bauer: Visualization; Writing—Review & editing. Sebastian Ehmann: Writing—Review & editing. Matthew D. Sacchet: Writing—Review & editing. John D. E. Gabrieli: Resources; Supervision; Writing—Review & editing.
Funding Information
This work was supported by the Tan-Yang Center at Massachusetts Institute of Technology.
Diversity in Citation Practices
Retrospective analysis of the citations in every article published in this journal from 2010 to 2021 reveals a persistent pattern of gender imbalance: Although the proportions of authorship teams (categorized by estimated gender identification of first author/last author) publishing in the Journal of Cognitive Neuroscience (JoCN) during this period were M(an)/M = .407, W(oman)/M = .32, M/W = .115, and W/W = .159, the comparable proportions for the articles that these authorship teams cited were M/M = .549, W/M = .257, M/W = .109, and W/W = .085 (Postle and Fulvio, JoCN, 34:1, pp. 1–3). Consequently, JoCN encourages all authors to consider gender balance explicitly when selecting which articles to cite and gives them the opportunity to report their article's gender citation balance. The authors of this article report its proportions of citations by gender category to be: M/M = .502; W/M = .216; M/W = .112; W/W = .170.