Neurobehavioural Exploration of Breath-counting & Breath-awareness in Novice Indian Meditators: A Naturalised Ānāpānasati-based Paradigmatic Approach

Abstract

Background

Neural activity and subjective experiences indicate that breath-awareness practices, which focus on mindful observation of breath, promote tranquil calm and thoughtless awareness.

Purpose

This study explores the impact of tristage Ānāpānasati-based breath meditation on electroencephalography (EEG) oscillations and self-reported mindfulness states in novice meditators following a period of effortful cognition.

Methods

Eighty-nine novice meditators (82 males; Mean Age = 24.59 years) underwent a breath-based meditation intervention consisting of three stages: Resting State (RS), Breath Counting (BC), and Breath Focus (BF). EEG assessed neural oscillatory changes throughout the three stages while providing spectral indices for arousal and cognitive workload (CWL) stagewise. State mindfulness and breath awareness-related self-reported feedback were also collected using the Amsterdam Resting-State Questionnaire (ARSQ) post-BF stage and the curated Breath Count Feedback (BCF) post-BC stage, respectively. The internal reliability and construct validity of the standardised ARSQ and the designed BCF were satisfactorily computed within our sample. A within-subjects cross-sectional neurobehavioural examination of the breath self-regulatory novice experiences was thus conducted.

Results

The breath-based intervention significantly increased alpha power across all stages, indicating relaxation. Theta and delta powers increased during BC and BF in the prefrontal cortex (PFC), suggesting enhanced working memory and focused attention. Gamma power in meditation-associated brain regions and occipital beta oscillations showed significant positive correlations with breath counts, reflecting improved visual and attentional concentration. Lower pre-meditative arousal and smaller in-meditation CWL levels were associated with fewer distractions and increased confidence accuracy during BC.

Conclusion

The results suggest that BC may serve as a valuable tool for improving present-centric control and concentration, highlighting the importance of managing CWL and arousal levels to optimise meditation outcomes.

Introduction

Among the different types of meditation, approaches that focus on the breath are highly recognised for their effectiveness in inducing calm and improving cognitive abilities.^{1, 2} These are recognised for their positive impact on mental well-being, focus, and emotional equilibrium.^{3, 4} Although there are differences among many meditation traditions and styles, there certainly are shared elements regarding their goals, methods, and experiences; the foremost aspect of meditation practices is to attain a state of equanimity and thoughtless awareness. ⁵ This can be achieved by directing meditators to concentrate on a specific object, such as their breath or bodily sensations, and to gently redirect their attention to the present moment whenever their mind starts to wander without passing judgment. ¹ Research has demonstrated that present-centred techniques can reduce the frequency and intensity of mind-wandering. ⁶ It can be expected that meditation practices aimed at achieving a state of ‘mental emptiness’ by reducing the frequency and involvement of mind wandering result in a decrease in the interaction between the memory (retrieval and retention of information) and executive (manipulation of information) aspects of cognition.^{1, 6}

Breath-awareness practices, which focus on the mindful observation and regulation of breath, are believed to harmonise prana flow, leading to enhanced states of consciousness and mental well-being, ⁷ as indicated by measurements of brain activity and self-reported subjective experiences.^{2, 8} In our study involving novice meditators, we thus utilised a multimodal approach in a breath-awareness meditation intervention.

Breath-awareness Novice Meditation Paradigm

The breath-meditation intervention paradigm was curated to investigate the calming effects of interventions focusing on breath. This involved a thirty-minute zeroth stage at the beginning, during which the participants engaged in a structured process of doing arithmetic tasks to produce stress and cognitive workload (CWL). Cognitive tasks such as solving arithmetic problems have been found to elevate cognitive load and stress levels.^{8, 9} Arithmetic exercises are frequently employed in stress-inducing protocols to induce stress consistently. ¹⁰ Effortful cognition is thus employed, involving cognitive processes that necessitate intentional attention, concentration, and exertion. ⁶ This enabled us to examine the diverse effects of meditation after a thorough period of effortful cognition, ¹ representing the pervasive presence of stress in everyday life.

Stage 1: Resting State (RS)

This stage lasted five minutes, during which the participants relaxed with their eyes closed to recuperate. The RS effectively aided participants in transitioning to the final two stages of our breath-based intervention, promoting recovery from the zeroth stage’s cognitive activity. This state facilitates brain recovery by reducing cortical arousal and promoting parasympathetic activity. ¹¹ Further, studies have demonstrated that entering a state of rest and closing one’s eyes can effectively promote relaxation by increasing alpha wave activity.^{11, 12}

Stage 2: Breath Counting (BC)

The duration of this stage was five minutes, wherein participants counted their breath cycles, beginning over if they lost track. The goal was progressively guiding participants into focused breathing, mainly because the sample comprised inexperienced meditators. After the stage, participants completed Likert scale-based breath count feedback (BCF) self-reports, including their breath counts, confidence, distract count, and breath-blissfulness. The BC stage is thus an initial mindfulness exercise that aids in enhancing focus and minimising mind-wandering. ¹³ The evaluation of breath-associated measurements offers valuable information regarding a participant’s mindfulness level and capacity to maintain concentration.^{13, 14}

Stage 3: Breath Focus (BF)

The stage lasted eight to ten minutes, during which the participants concentrated exclusively on their breath without counting. Breath attention is an ancient yet simple mindfulness technique advocated by Gautama Buddha known as Ānāpānasati and has been proven to decrease stress and boost emotional well-being.^{15, 16} Posteriori BF stage, the participants were instructed to complete the Amsterdam Resting-State Questionnaire (ARSQ) ¹⁷ and provide a detailed account of their breath-focus resting-state experiences.

Objectives

The Ānāpānasati-inspired tristage paradigm is designed to intentionally create stress by increasing CWL and then lead novice meditators through a series of mindfulness exercises focused on breathing. In summary, the research intends to investigate the neurobehavioural cross-sectional effects of the described tristage intervention in novice Indian meditators. The objectives studied are stated as follows:

Examining the paradigm’s novice neural oscillatory changes in the meditation-associated brain regions across the three stages (RS, BC, BF) of the paradigm. (Objective 1)

Evaluating the variations in self-regulatory neural activity and self-reported BCF among novice meditators during the breath-counting phase. (Objective 2)

• 2.1.

Analysing the role of spectral powers during the BC stage. (Objective 2.1)

2.2

Examining the BCF intra-variable comparisons. (Objective 2.2)

2.3

To study the associations between self-reported BCF and state mindfulness (ARSQ) variables. (Objective 2.3)

2.4

Investigating the electrophysiological features of the CWL and arousal before and during the BC stage. (Objective 2.4)

Methodology

Participants

The neurobehavioural data for the tristage paradigm was gathered in a silent and secluded space in the laboratory. The participants were given specific instructions for each stage of the paradigm. The naturalised intervention underscores the stressors that occur in daily life, conducting a detailed examination of electroencephalography (EEG) meditation spectral indices, along with ARSQ- and BCF-mindfulness-self-reports, in our novice sample. The sample included 89 inexperienced meditators, 82 males and seven females, with a mean age of 24.59. The sample comprised 59 individuals from STEM backgrounds and 30 from non-STEM backgrounds. The study was approved by the Institute Ethics Committee of the Indian Institute of Technology, Delhi (IEC-IITD; Proposal No. P021/P0101).

Electroencephalography

EEG data were acquired using Recorder Software (version 1.25.0001) developed by Brain Products GmbH with a sampling rate of 500 Hz from 64 electrodes utilising an Easycap (Brain Products GmbH), following the extended International 10–20 method. ¹⁸ The data was amplified with a LiveAmp 64 and filtered using a third-order low-pass filter, with bandwidth restricted to 0.01–131.0 Hz. FCz served as the recording reference, whereas AFz served as the ground reference. Electrode impedance was maintained at 5–15 kOhm before and after each stage during the experiment. The EEG recordings were preprocessed, and spectral indices were obtained using MATLAB vR2021a and EEGLAB v2023.1. ¹⁹ The preprocessing pipeline for continuous EEG data processing utilised the Artefact Subspace Reconstruction-Independent Component Analysis (ASR-ICA) methods. ²⁰

First, channel information is included, and only the middle two-thirds of the EEG meditation segments were kept. Subsequently, the data is reduced in sampling rate from 500 Hz to 250 Hz. It is then subjected to filtering using an IIR-Butterworth bandpass filter with a range of 1–60 Hz and a Zapline notch filter to eliminate 50 Hz line noise. ²¹ Channels were excluded from consideration if they exhibit spectral power deviations beyond ±3 standard deviations. The ASR-clean data method developed by Chang et al. (2019) is employed to identify and rectify artefacts. ICA decomposition, IC labelling, and IC rejection are performed to eliminate noises with a label value beyond ‘0.5’.^{22, 23} The EEG data are reprocessed by interpolating the bad channels and the online reference channel (FCz), whereafter applying mean-mastoid rereferencing to all channels.

The processing pipeline for spectral power-based indices began by loading the preprocessed EEG datasets. The Fourier transformation is applied to convert the data from the time domain to the frequency domain using the EEGLAB spectrum analysis plugin, specifically version 1.2 of the EEGStats plugin. Average powers and all-channel powers are calculated for all three stages subject-wise. CWL (frontal theta power ÷ parietal alpha power) and arousal defined as ‘Beta at F3 + Beta at F4’ to ‘Alpha at F3 + Alpha at F4’^24–26 are also retrieved separately for each stage and subject.

Behavioural Self-Reports Employed (ARSQ & BCF)

BCF and ARSQ ¹⁷ were administered after stages two and three, respectively, of the paradigm. The ARSQ’s comprehensive theoretical framework is particularly valuable since it integrates cognitive psychology, neuroimaging, and studies on the default mode network (DMN).^{27, 28} The questionnaire encompasses fundamental factors connected to the RS, such as mind-wandering, task-related thoughts, and external sensory perceptions. ARSQ dimensions included ‘Discontinuity of Mind (DOM)’, ‘Theory of Mind (TOM)’, ‘Self (SLF)’, ‘Planning (PLN)’, ‘Sleepiness (SLP)’, ‘Comfort (CMF)’, and ‘Somatic Awareness (SOA)’. All the dimensions demonstrated satisfactory inter-item reliability, ²⁹ construct validity, ³⁰ confirmatory factors’ structural fit ³¹ within our sample, utilising Jamovi’s ‘Factor’ module. ³² Cronbach α and McDonald’s ω for ARSQ sub-scales were found to be DOM (α = 0.640; ω = 0.660), TOM (α = 0.661; ω = 0.675), SLF (α = 0.654; ω = 0.709), PLN (α = 0.773; ω = 0.783), SLP (α = 0.796; ω = 0.811), CMF (α = 0.813; ω = 0.824), and SOA (α = 0.596; ω = 0.608).

The BCF self-report measured the ‘number of breath counts’ with a quantitative response, ‘breath confidence’ using a reverse-scored seven-point Likert scale, ‘breath blissfulness’ through a reverse-scored seven-point Likert scale, and ‘breath distraction’ with a positively-scored seven-point Likert scale. Excluding BCF’s count numerical feedback, the remaining items were assessed with a Cronbach’s α coefficient of 0.754 and a McDonald’s ω coefficient of 0.795 within our sample.

EEG Spectral Analysis & Behavioural Assessments

Pertaining to Objective 1, the spectral electrophysiological data was statistically analysed using a ‘3 × 3 × 5’ repeated measures factorial design in R-jamovi. ³³ The factors consisted of three meditation-specific brain regions, namely the midline DMN, the Prefrontal Cortex (PFC), and the Occipital region (OCC). Additionally, there were three paradigm stages, namely RS, BC, and BF. Furthermore, factors included five different neural oscillation-band powers, specifically delta (1–4 Hz), theta (4–8 Hz), alpha (8–13 Hz), beta (13–30 Hz), and gamma (30–60 Hz), with units expressed in decibels per hertz (dB/Hz). The electrodes used for the OCC region were ‘O1’, ‘O2’, ‘Oz’, ‘PO3’, ‘PO4’, and ‘POz’. ³⁴ The PFC area was covered by the electrodes ‘FP1’, ‘FP2’, ‘AF3’, ‘AF4’, ‘AFz’, ‘F1’, ‘F2’, ‘F3’, ‘F4’, and ‘Fz’. ³⁵ For the DMN region analysis, we focused on midline and posterior channels, excluding mPFC channels, as they were already considered in the PFC area. Hence, it should be emphasised that the DMN in the study comprised of the following channels: ‘CP1’, ‘CP2’, ‘CP3’, ‘CP4’, ‘CPz’, ‘P1’, ‘P2’, ‘P3’, ‘P4’, ‘Pz’, ‘PO3’, ‘PO4’, ‘POz’, and ‘Oz’. ³⁶ Lastly, the Greenhouse-Geisser corrections ³⁷ were applied to account for any violations of sphericity in the examples. In addition, post-hoc t-tests were conducted for the significant effects using Tukey and Bonferroni corrections. ³⁸

To study the role of spectral powers in novice meditators’ breath count features during the BC stage, we conducted a correlation analysis between the amplitude of oscillations in three brain regions (DMN, PFC, and OCC) and the dimensions of the BCF towards Objective 2.1. Thereby, accounting for the correlation of five spectral powers across three brain regions, the modified significance level (α = 0.003, i.e., 0.05 ÷ 15) was set using the Bonferroni adjustment. In the internal examination of BCF reported parameters, a Principal Component Analysis (PCA) using a factoextra package ³⁹ and a correlational analysis was performed towards Objective 2.2, elucidating the underlying structural and functional relationships of the four BCF constructs (count, confidence, distractedness, blissfulness). The adjusted significance level was set at ‘0.0083’ to account for multiple comparisons in the correlational analysis of Objective 2.2. Further, Objective 2.3 investigated the associations between the participants’ performance during the BC task (post-stage two) and their self-reported state mindfulness utilising ARSQ (post-stage three). Using the Bonferroni correction, we set the significance level to ‘α = 0.0016’ for 32 comparisons.

Objective 2.4 delved deeper into the neural underpinnings of the BC stage by comparing the CWL and arousal from the first two phases of the intervention to BCF’s self-reports. The multiple-tests adjustment modified significant levels to ‘.0125’ for both the four CWL and the four arousal comparisons.

Results

Objective 1: ‘Region-Oscillation-Stage’ Effects & Interactions Across the Tristage Paradigm

The rm-anova analysis towards Objective 1 found that the main effects of ‘region’ (F (1.36) = 41.31, P < .001, η_G² = 0.013) and ‘band powers’ (F (1.77) = 77.69, P < .001, η_G² = 0.277) were statistically significant with a small and substantial effect size, respectively. Post-hoc tests employed for the ‘region’ main effect revealed that the PFC exhibited the maximum brain activity, followed by the DMN and OCC (see Figure 1b); possibly owing to the increased frontal theta associated with the CWL during the novice breath focus/count, while the DMN activity is indicative of mind-wandering during rest.^{40, 41} Post-hoc tests for the ‘band powers’ main effect indicated that alpha power is significantly higher due to resting characteristics of the three stages, validating the paradigm’s effectiveness for a novice sample (see Figure 1c). No significant differences were found between delta and theta, despite the aperiodic 1/f neural power dependence, ⁴² wherein delta power is typically expected to exceed theta power.

Figure 1.

(a) ‘Regions x Stages x Power’ Interaction Effect, Marginal Means Plot. (b) ‘Regions’ Main Effect, Marginal Means Plot. (c) ‘Band Powers’ Main Effect, Marginal Means Plot. (d) ‘Regions x Power’ Interaction Effect, Marginal Means Plot. {y-axis is Spectral Powers in all Plots (Units: dB/Hz)}.

Besides, a weak effect size for the significant ‘region x power’ interaction was noted (F (3.07) = 28.43, P < .001, η_G² = 0.018). The computed marginal means (see Figure 1d) show that the PFC has the highest delta power (M = 4.03, SE = 0.242), followed by the DMN (M = 2.84, SE = 0.182), and the OCC (M = 2.24, SE = 0.167). Theta power is highest in the PFC (M = 3.97, SE = 0.46), followed by the DMN (M = 2.86, SE = 0.33) and occipital cortex (M = 2.03, SE = 0.209). The increased engagement of the PFC and DMN is attributed to the working memory involvement ⁴³ and memory-retrieval processes.^{44, 45} This is reflected by the higher powers of theta and delta brain waves, owing to a conscious focus on breath or its counting during the BC-BF stages. Gamma power exhibited statistically significant differences only between the DMN and the PFC (t (88) = −3.92, P_tukey < .05, P_bonferroni < .05). Lastly, there were no discernible variations in alpha and beta power between brain regions.

Notably, there was no significant main effect of ‘stage’, and no significant interaction effects incorporating ‘stage’ as a factor (see Figure 1a). The minimal variations in brain activity observed during the several stages of breath-based intervention in the inexperienced adult participants are possibly owing to their lack of proficiency in meditation. Thus, the spectral profiles of the meditative stages two and three in the chosen brain regions closely resemble their RS in stage one. Notably, the apriori zeroth effortful cognition stage could also be a plausible factor towards it.

Moreover, when the average marginal means of band powers across all brain regions were compared with those in DMN, PFC, and OCC regions, the beta and gamma powers for these three meditative areas were significantly higher than the overall brain average.

Objective 2.1: Spectral Power’s Role During Novice BC

The objective demonstrated a negative correlation between breath confidence and PFC-delta oscillations (r = −0.335, P_adj < .003). The latter also appears to have a marginally significant positive and negative impact on Breath-Distract (r = 0.292, P = .005) and Breath-Blissfulness (r = −0.279, P = .008), respectively. PFC-delta seems to accurately underscore the antagonistic correlation between the feeling of bliss and the distraction-ridden low confidence during counting, as reported by novice meditators. The Breath-Counts showed strong positive associations with gamma power for all three regions: DMN (r = 0.40, P_adj < .003), OCC (r = 0.41, P_adj < .003), and the PFC (r = 0.34, P_adj < .003). Further, post-hoc tests revealed a significant increase in gamma power in the OCC (r = 0.43, P < .0001) and DMN (r = 0.29, P = .0052) regions compared to the RS. Also, there was a positive correlation between occipital beta power and the self-reported breath count (r = 0.31, P_adj < .003). Furthermore, examination of the OCC-beta power compared to RS demonstrated a significant positive association with the counts (r = 0.28, P = .0068).

The results indicate that improved accuracy in BC is linked to higher gamma oscillatory activity,^{46, 47} indicating enhanced cognitive processing and attentional focus during meditation. Additionally, previous investigations^{46, 48} have shown that frontal delta activity has decreased during similar breath-based meditation practices. The findings from the beta-band analysis suggest that when individuals accurately count their breaths, there is an increase in beta oscillatory activity in the OCC, indicating improved visual and attentional processing. ⁴⁹

Objective 2.2: Intra-BC Characteristics Comparisons

Breath Confidence (r = 0.541, P_adj < .0083) was highly positively correlated with Breath Count, while Breath Distract (r = −0.501, P_adj < .0083) was negatively correlated with Breath Count. Breath Confidence demonstrated a positive link with Breath Blissfulness (r = 0.509, P_adj < .0083) and a further strong negative correlation with Breath Distract (r = −0.7, P_adj < .0083). These findings imply that higher breath counts are linked to more confidence and less distraction during meditation. Moreover, confidence is related to lower distraction and more blissfulness. In summary, breath confidence, not the count of breaths, is a primary indicator of the present-centric tranquillity felt by novice meditators’ posterior effortful cognition.

PCA elucidated the underlying structure of these four constructs in novice meditative experiences, revealing a circumplex model with two key components (see Figure 2). With an eigenvalue of more than one, the first component accounted for 58.47% of the variation. In contrast, the second component, with an eigenvalue almost approximating 1 (0.935), explained 23.38% of the variance. The PCA revealed an antagonistic link between Breath Distract and Breath Count, placing them on opposite ends, while Breath Blissfulness and Breath Confidence were found adjacently between the former two. Hence, stressing the dependence of the latter two on the former two during novice BC meditation.

Figure 2.

Breath Counting Self-reported Characteristics PCA Plot.

Objective 2.3: Associations Between BC & State Mindfulness Self-reports

During BC, sleepiness was found to have a negative correlation with confidence (r = −0.355, P_adj < .0016) and a positive correlation with blissfulness (r = −0.418, P_adj < .0016). Also, blissfulness positively correlated with comfort (r = 0.405, P_adj < .0016). While not previously recommended in the literature, we curated a variable called ‘state-mindfulness’ (SMIND), defined as the sum of somatic awareness, comfort, and the negation of sleepiness, self, planning, discontinuity of mind, and theory of mind. We observed a positive correlation (r = 0.346, P_adj < .0016) between the hypothesised resting-state variable ‘SMIND’ and the blissfulness experienced during counting in our inexperienced meditators. The limited correlation between the breath metrics during the BC stage and the state mindfulness dimensions during the final BF stage indicates that novice meditators may not yet demonstrate a strong connection between their breathless practice and subsequent state mindfulness assessments.

Objective 2.4: Arousal, Workload & BC Characteristics

Correlational analysis was performed between resting-state arousal at the F3-F4 electrodes and subsequent breath measurements during counting. The aim was to investigate how the performance in the BC stage ⁵⁰ will be affected by a more naturally relaxed novice mind measured by a smaller resting (RS) arousal. Smaller resting arousal levels seem to predict higher confidence (r = −0.362, P_adj < .0125) and a more accurate count (r = −0.294, P_adj < .0125), suggesting that a less stimulated brain state before meditation can improve focus during the practice, even for a novice. Besides, more distractedness (r = 0.28, P_adj < .0125) was associated with higher arousal levels at rest. Hence, a more active pre-meditative state could lead to difficulty in maintaining attention and a greater sensitivity to distractions. The results suggest that reduced arousal apriori meditation may improve meditative results for novices by creating a more appropriate neural readiness.

Following the arousal results, the relationships between CWL and breath-counting characteristics^{24, 25} were further examined, whereby self-reports of BCF were correlated against CWL of the first two intervention stages. Increased CWL during BC is negatively correlated with count accuracy (r = −0.265, P_adj < .0125), confidence (r = −0.287, P_adj < .0125), and blissfulness (r = −0.27, P_adj < .0125) while showing a positive relation with distractedness (r = 0.275, P_adj < .0125). Participants with higher cognitive demands may find it challenging to keep focus and correctly estimate their breath counts, which would lead to detraction from the meditation experience and lower their sensation of joy.^{51, 52} Moreover, when comparing BCF with CWL observed during the RS stage, only Breath-Distract was favourably associated (r = 0.29, P_adj < .0125). On a side note, the increase in CWL during the BC stage was associated with sleepiness (SLP) during the BF stage (r = 0.30, P = .0038). These results stress the significance of managing cognitive load during meditation—especially for beginners—to enhance meditative engagement.

Discussion

The study examined the impact of breath-based meditation on the induced EEG oscillations and the subjective experiential self-reports after engaging in cognitive tasks. The paradigm’s natural- and breath-induced-resting-states have consistently shown diverse yet distinctive inferences across different brain areas and oscillatory bands, as demonstrated in the spectrum analysis conducted toward Objective 1. The intervention consistently improved alpha power across all stages, given the resting-type nature of the stages, validating the effectiveness of the intervention. In addition, the rise in theta power during the BC and BF phases suggests an enhanced utilisation of working memory and concentration attention. ⁴⁰ The increased delta and theta oscillations, particularly in the PFC, indicated heightened cognitive engagement and improved memory during meditation. ⁴⁸ Significant correlations between gamma power and breath counts in several brain areas were further revealed in Objective 2.1, pointing to enhanced concentration and cognitive processing. ⁴⁷ Finally, improved precision in counting, as measured by occipital beta power, was associated with improved visual attentional processing during meditation.

The findings in Objective 2.4 highlighted the impact of pre-meditative arousal levels on meditation outcomes. Lower arousal levels were correlated with increased confidence and more precise counts, while higher arousal levels were associated with greater distractions. Notably, Objective 2.2 elucidated that developing confidence in BC rather than increasing breath count may improve the meditation experience. Objective 2.4 underscored that managing CWL during BC was crucial. An elevated workload during BC was shown to be associated with distractions, culminating in adverse effects on meditative performance. Objective 2 thus highlights the plausibility of BC as a method for improving moment-centric concentration, inducing restful bliss in novice meditators. ³⁴

Although the present research lacked a control group, similar studies employing control conditions have shown the substantial advantages of breath-based meditation. Previously, administration of brief breath-focused meditation training was reported to decrease pain perception and improve cognition relative to a control group.^{14, 54} These findings corroborate our results, indicating that the observed effects are likely due to the breath-based meditation intervention despite the lack of a direct comparison group.

Conclusion

The current work examined the effects of naturalised breath-awareness meditation, providing neurobehavioural insights into novice meditation. Overall, this study highlights BC as a method for improving present-centric control and concentration in novice meditators, underscoring the significance of regulating CWL and arousal levels to maximise meditation results. However, the focus on novice meditators and a male-dominated sample limits the generalizability of findings. Upon a follow-up analysis in a balanced large novice meditator sample (N = 580; Males = 289, Females = 291; STEM = 273; Non-STEM = 307), neither gender nor the ‘STEM versus Non-STEM’ groups revealed any statistically significant differences for trait mindfulness, as assessed by the Five Facet Mindfulness Questionnaire (FFMQ). ⁵⁵ Besides, no group differences were observed in the behavioural state mindfulness correlates in the original ‘N = 89’ sample. Thus, recognising gender- and STEM-skewness as limitations, the findings of this article should be interpreted with caution.

Moreover, the cross-sectional design and reliance on self-reported measures may introduce bias and restrict the interpretation of causality. Future research should consider longitudinal designs to explicate the long-term effects of meditation. Lastly, utilising additional neuroimaging techniques, synchronisation- and source-based EEG estimations could further clarify the mechanisms underlying meditation’s impact on novice brain function.