query <- "meditation[TIAB] AND EEG[TIAB] AND Focus[TIAB]"
results <- 9Studies <- list()
Studies$Saggar <- rpdfclown::extractPDF("Saggar et al. - 2012 - Intensive training induces longitudinal changes in.pdf")[1] %>%
gsub("[^[:alnum:][:blank:]?&/\\-\\:\\.\\,\\'\\#]", " ", .) %>% gsub("\\s(?=Page.?.?\\d)",
"\\\n", ., perl = T) %>% gsub("(?<=Page.\\d)\\:", ": ", ., perl = T) %>% str_split(pattern = "\\n") %>%
.[1] %>% lapply(function(.) gsub("\\s{2,}", " ", .)) %>% unlist
# Combine highlights that span pages
Studies$Saggar <- Studies$Saggar[!str_detect(Studies$Saggar, "Combine")] %>% append(paste(Studies$Saggar[str_detect(Studies$Saggar,
"Combine")][1], Studies$Saggar[str_detect(Studies$Saggar, "Combine")][-1] %>%
gsub(".*\\#Combine\\d+", "", .), collapse = ""), after = str_detect(Studies$Saggar,
"Combine") %>% which %>% min)Studies$Saggar %<>% gsub("(\\#[0-9A-Za-z\\.]+)", "\\<em style\\=\\'color\\:blue\\'\\>\\1\\<\\/em\\>",
., perl = T)
Studies$Saggar %>% lapply(htmltools::HTML)[[1]] Page.1: #Content Meditative training includes methods for developing enduring psychological traits through deliberate application of awareness to the contents of subjective experience, including thoughts, sensations, intentions, and emotions.
[[2]] Page.1: #Abstract The capacity to focus one s attention for an extended period of time can be increased through training in contemplative practices. However, the cognitive processes engaged during meditation that support trait changes in cognition are not well characterized. We conducted a longitudinal wait list controlled study of intensive meditation training. Retreat participants practiced focused attention FA meditation techniques for three months during an initial retreat. Wait list participants later undertook formally identical training during a second retreat. Dense array scalp recorded electroencephalogram EEG data were collected during 6 min of mindfulness of breathing meditation at three assessment points during each retreat. Second order blind source separation, along with a novel semi automatic artifact removal tool SMART , was used for data preprocessing. We observed replicable reductions in meditative state related beta band power bilaterally over anteriocentral and posterior scalp regions. In addition, individual alpha frequency IAF decreased across both retreats and in direct relation to the amount of meditative practice. These fndings provide evidence for replicable longitudinal changes in brain oscillatory activity during meditation and increase our understanding of the cortical processes engaged during meditation that may support long term improvements in cognition.
[[3]] Page.1: #Keywords: training, attention, meditation, beta, individual alpha frequency, EEG
[[4]] Page.2: #Ref #FA These traditional descriptions of the mental processes employed during shamatha and other focused attention FA meditation techniques share considerable theoretical overlap with contemporary cognitive and neuroscientifc theories of attention Lutz et al., 2008 Slagter et al., 2011
[[5]] Page.2: #Ref #FA #Benefits Recent longitudinal studies of intensive FA meditation have demonstrated training related improvements in attentional sta bility Lutz et al., 2009 MacLean et al., 2010 and alerting Jha et al., 2007 , sustained response inhibition Sahdra et al., 2011 , and information processing effciency van Vugt and Jha, 2011
[[6]] Page.2: #EEG #Ref In a cross sectional study of meditative adepts, Brefczynski Lewis et al. 2007 reported increased activity in a broad network of attention related brain regions, including frontoparietal, temporal, and posterior occipital cortical areas during meditation.
[[7]] Page.2: #Purpose a detailed charac terization of patterns of cortical activation during meditation is critical to understanding the role of meditative states in trait level improvements.
[[8]] Page.2: #EEG One avenue for inferring process specifc train ing of attention is by examining the modulation of task specifc cortical oscillations during meditation Cahn and Polich, 2006 .
[[9]] Page.2: #Ref #EEG Intrinsic rhythmicity in ongoing electrical cortical activity is traditionally organized into standard spectral frequency bands, ranging from slow to fast wave oscillations Steriade, 2006 .
[[10]] Page.2: #Ref #EEG Oscillatory activity has also been linked to local and large scale synchronization of neuronal assemblies across brain regions Varela et al., 2001 Fries, 2005 Siegel et al., 2012 Tallon Baudry, 2012 , which may facilitate processes dependent on the integration of information across distributed brain networks.
[[11]] Page.2: #Ref #EEG There is increasing evidence that attentional modulation of neuronal oscillations may serve to in uence selective sensory processing Womelsdorf and Fries, 2007 .
[[12]] Page.2: #EEG #Brainwaves Attentional modulations of ongoing oscillations in the alpha 8 12 Hz and beta bands 13 30 Hz have been functionally implicated in the percep tual processing of somatosensory information and may therefore serve as potential physiological markers of the capacity to focus attention on the breath.
[[13]] Page.2: #EEG #Ref Activity in the alpha and beta bands is inversely related to cortical excitability Tamura et al., 2005 Ploner et al., 2006 Ritter et al., 2009 , speed of visual and sen sorimotor processing Thut et al., 2006 van Ede et al., 2011 , stimulus discriminability van Dijk et al., 2008 , target detection accuracy Linkenkaer Hansen et al., 2004 Romei et al., 2010 , and attentional suppression of distracting information Snyder and Foxe, 2010 Haegens et al., 2012 .
[[14]] Page.2: #EEG Attentional orienting to upcoming tactile stimuli induces hemisphere specifc suppres sion of beta oscillations in sensorimotor cortex Dockstader et al., 2010 Jones et al., 2010 van Ede et al., 2010, 2011 , suggest ing a functional role for beta in spatially oriented attention.
[[15]] Page.3: #EEG #Brainwaves Compared to a control group, participants who completed a non intensive eight week course in mindful ness meditation demonstrated greater alpha band suppression in response to anticipatory cues requiring the spatial orienting of attention to anticipated tactile stimuli. Notably, however, no evidence was found for changes in beta band oscillations.
[[16]] Page.3: #EEG #SignalProcessing In addition, the designation of frequency bands may not be sensitive to intra and inter individual differences that may in uence the distribution of spec tral frequencies. The understanding of cortical oscillatory activity during meditation may beneft from methodological advances in signal processing and the use of individualized frequency bands for the classifcation of spectral power e.g., Klimesch, 1999 .
[[17]] Page.3: #EEG #SignalProcessing We used second order blind source separation SOBI Belouchrani et al., 1997 , along with a novel artifact removal tool Saggar, 2011 , to identify and remove signal sources of putative non neural origin. Each participant s individ ual alpha frequency IAF Klimesch, 1999 was used to defne
[[18]] Page.3: #EEG #Brainwaves #Hypothesis Specifcally, we pre dicted training related changes in areas involved in attention and somatosensory processing as evidenced by reductions in alpha and beta band power across central and parietal areas of the scalp.
[[19]] Page.3: #Ref #EEG #Brainwaves We also investigated activity across the remaining spec tral frequency bands based on previous reports of meditation state dependent activity in the delta, theta Cahn and Polich, 2006 , and gamma bands Lutz et al., 2004 Cahn et al., 2010 .
[[20]] Page.3: #EEG #Brainwaves #Alpha #Hypothesis Because several meditation studies have revealed an overall slowing in oscillations within the alpha frequency, both as a trait and as a state effect Cahn and Polich, 2006 , we predicted a similar downward shift in IAF following training.
[[21]] Page.3: #EEG #Alpha #Beta #Hypothesis Finally, we predicted that increases in cortical activity reductions in beta and alpha band power and decreases in IAF would vary in direct relation to the amount of time individuals spent engaging in FA meditation during the retreat.
[[22]] Page.3: #N Sixty participants were selected out of 142 applicants based on age, availability, physical and mental health, and previous retreat experience. These par ticipants were then assigned to either an initial retreat N 30 or wait list control N 30 group through stratifed matched assignment.
[[23]] Page.3: #Design Groups were matched on age, sex, meditation expe rience, and ethnicity see MacLean et al., 2010 Sahdra et al., 2011, for full assignment and matching criteria
[[24]] Page.3: #Design Wait list control group participants were own to the retreat center for testing at each assessment point during Retreat 1 data were collected after acclimatization for 72 96 h . Approximately three months after completion of Retreat 1, these same wait list control group participants underwent formally identical training in a second three month retreat Retreat 2 .
[[25]] Page.3: #Analysis EEG data from 22 participants in each group retreat and wait list control were included in the analysis.
[[26]] Page.3: #Design Initial retreat and wait list control participants included in the analyses did not differ all ps 0.05 in age retreat: M 49.5, SD 13.5, control: M 44.2, SD 15.8 , gender retreat: 12 female, control: 11 female , or lifetime meditation experi ence retreat: M 2855.6 h, SD 2994.1, control: M 2272.7 h, SD 2326.3
[[27]] Page.4: Participants practiced meditation under the guidance ofDr. B. Alan Wallace, an established Buddhist teacher, contem plative, and scholar.
[[28]] Page.4: #Design Shamatha techniques included mindfulness of breathing, in which attention is directed toward the breath observing mental events, in which attention is directed toward the whole feld of mental experience thoughts, images, sensations and observing the nature of consciousness, in which attention is directed toward the experience of being aware. Benefcial aspirations included practices that cultivated loving kindness, compassion, empathic joy, and equanimity Wallace, 2006 . Participants who received training in the frst retreat M 5.7 h per day, SD 1.5 or the second retreat M 5.4 h per day, SD 1.5 spent a similar amount of time practicing solitary FA meditation t 42 0 . 486, p 0 . 629 . In addition to engaging in solitary practices, participants met twice daily for group meditation practice and discussion guided by Dr. Wallace. Participants also met with Dr. Wallace privately once a week for individual advice, clarifcation, and guidance. Dr. Wallace was not present during any data collection procedures.
[[29]] Page.4: #Measurement At the conclusion of the second day of testing, partic ipants engaged in a 12 min period of silent, eyes closed mindful ness of breathing. The meditation began with approximately 50 s of audio instructions, recorded by Dr. Wallace: During the next 12 minutes, engage in the practice of mindful ness of breathing, focusing your attention on the tactile sensations at the apertures of your nostrils or just above your upper lip. With each inhalation arouse your attention and focus clearly on these tactile sensations. With each out breath continue to maintain your attention upon the tactile sensations, while relaxing your body and mind, releasing any involuntary thoughts that may arise. So in this way maintain an ongoing ow of mindfulness, arousing with each in breath, relaxing with each out breath. A recorded chime signaled the end of the meditation period. Continuous EEG was recorded over the entire period. However, due to errors in data acquisition, only the frst 6 min of data were recorded for some subjects.
[[30]] Page.4: #Measurement These baseline data were subsequently used to calculate the IAF values for each participant at each assessment.
[[31]] Page.4: #Materials EEG was acquired at a sampling rate of 2048 Hz using the BioSemi ActiveTwo system http://www . biosemi . com and FMS electrode caps http://www . easycap . de ftted with BioSemi elec trode holders in an 88 channel equidistant montage. Individual electrode locations were localized in three dimensional space using a Polhemus Patriot digitizer http://www . polhemus . com .
[[32]] Page.4: #SignalProcessing Second order blind source identi cation SOBI Belouchrani et al., 1997 was used to separate sources of contaminat ing signal from ongoing brain electrical activity. SOBI uses joint diagonalization of correlation matrices at multiple tempo ral delays 41 delays were used, 1:1:10, 12:2:20, 25:5:100, 120:20:300 ms, as described in Tang et al., 2005 to derive signal components that have a continuous time course and xed spa tial projections, referred to as sources.
[[33]] Page.4: #SignalProcessing SOBI has two main advantages over other methods e.g., ICA for blind source identifcation: a it uses average statistics over multiple tempo ral delays and hence is less susceptible to outliers and b it uses second order statistics such that short segments of data are suff cient for estimating components for further discussion of blind source separation methods see Joyce et al., 2004 Tang et al., 2005 Congedo et al., 2008 Tang, 2010
[[34]] Page.4: #SignalProcessing For each separate participant, assessment, and 1 min epoch, the number of generated sources was equal to the number of scalp channels 104,000 individual sources . To distinguish the neural or non neural origin of these sources, signal components must be identi ed and evaluated. Although it is possible to evaluate these sources manually using quantita tive features such as topography, time series, and power spectrum, the amount of data makes a manual approach infeasible. On the other hand, fully automatic solutions are harder to validate. Thus, a novel semi automatic artifact removal tool SMART Saggar, 2011 was constructed to maximize the likelihood that only non neural i.e., artifactual sources were rejected and that neural sources were retained.
[[35]] Page.5: #SignalProcessing classifed according to scalp voltage topography, power spectrum, autocorrelation, time series characteristics, and the impact of each source on the overall power spectrum. SMART provides the user with an html based interface of initial classifcations. The user can quickly review all the sources and, if required, reclassify the ini tial classifcations provided by SMART
[[36]] Page.5: #SignalProcessing The original montage of 88 channel data was thus recon structed using only these signal sources of presumed neural ori gin.
[[37]] Page.5: #SignalProcessing the reconstructed multichannel time series was then scanned for high amplitude transient signals, or signal gaps, that may have been included in the reconstruction because they were correlated with other neural activity.
[[38]] Page.6: #SignalProcessing These data were then transformed to a reference free estimation of scalp current density CSD to limit the effects of volume conduction and improve the spatial resolu tion depicted on the scalp surface e.g., Kayser and Tenke, 2006 . CSD was calculated using the surface Laplacian estimated as a sec ond derivative of the scalp potential with CSDToolbox Kayser and Tenke, 2006 1 10 6 .
[[39]] Page.5: #Combine1 #SignalProcessing This post SOBI reconstruction signal check was conducted on each reconstructed #SignalProcessing data le using the artifact scan tool in Brain Electrical Source Analysis BESA 5.2 www.besa.de . Amplitude and gradient epoch 1 s rejection thresholds were set individually for each partic ipant at each assessment gradient threshold: M 4.6 V/s, SD 0.5 amplitude threshold: M 104.4 V, SD 16.2 . Page.5: #Combine1 #SignalProcessing This post SOBI reconstruction signal check was conducted on each reconstructed Page.6: #Combine 3 #SignalProcessing Finally, the reconstructed 88 channel EEG data were trans formed into a standard 81 channel montage international 10 10 system using spherical spline interpolation 2 10 6 Perrin et al., 1989 as implemented in BESA 5.2. This trans formation ensured that the number of channels was consistent across participants and that channel locations were standard ized. Eight channels AF9, Fp1, Fpz, Fp2, Nz, AF10, CB1, CB2 from the 81 channel montage were excluded because data from the corresponding nearest electrode sites were not available in the original montage, yielding a nal 73 channel montage for the reconstructed EEG. Page.5: #Combine1 #SignalProcessing This post SOBI reconstruction signal check was conducted on each reconstructed #Analysis #Results Signifcant beta band clusters indicate spatiotemporal changes in beta band activity over time for participants who received training in Retreat 1 and Retreat 2, but not for the wait list controls tested during Retreat 1. To examine the direction ality of these training related changes, multivariate repeated measures analyses of variance ANOVA were used. For Retreat 1, the ANOVA included the within subjects effect of assessment pre , mid , and post testing , the between subjects effect of group retreat, control , and the interaction between the two. Because no signifcant beta band cluster was found for the con trol group in Retreat 1, data for this group consisted of the log transformed beta band power averaged across the same elec trode locations as were found for the signifcant electrode clus ter for the Retreat 1 group. The ANOVA revealed signifcant main effects of group F 1 , 42 5 . 01, p 0 . 031 and assess ment F 2 , 41 13 . 03, p 0 . 001 . Importantly, a signifcant Page.5: #Combine1 #SignalProcessing This post SOBI reconstruction signal check was conducted on each reconstructed #Analysis #Results group assessment interaction was also found F 2 , 41 7 . 11, p 0 . 01 , suggesting training related changes in beta band power. To further explore this interaction, we conducted sepa rate repeated measures ANOVAs for each group with assessment as a within subjects factor. A signifcant effect of assessment was found for the retreat group F 2 , 20 40 . 12, p 0 . 001 , and post hoc pairwise t tests all reported p values are Bonferroni cor rected for three comparisons revealed a signifcant reduction in beta band power at the mid t 21 8 . 65, p 0 . 001 and post assessments t 21 2 . 72, p . 038 . No signifcant differences were found in the control group Page.5: #Combine1 #SignalProcessing This post SOBI reconstruction signal check was conducted on each reconstructed #Discussion #Beta #MeditationTime #Lifetimehours The observed reduction in beta band power following med itative training is also consistent with a cross sectional study of highly experienced meditative adepts Brefczynski Lewis et al., 2007 . Brefczynski Lewis et al. 2007 observed increased BOLD activation in brain regions typically involved in sustained atten tion during FA meditation for expert meditators with an average of 19,000 h of practice. In contrast, experts with over 40,000 h of lifetime practice showed a decreased amount of activation in the same brain regions during FA meditation. These results Page.5: #Combine1 #SignalProcessing This post SOBI reconstruction signal check was conducted on each reconstructed #Discussion #Beta #MeditationTime #Lifetimehours suggest that cortical activation during meditative practice may follow a curvilinear trajectory such that both novices and highly experienced practitioners show less attention related activation than practitioners whose lifetime experience falls between these extremes. The average level of lifetime meditation experience among participants in our study was about 2500 h a moder ate level in comparison to the above groups. Thus, our observed reductions in beta band power, presumably indicative of an increase in cortical activation, are in line with the trajectory of training related change proposed by Brefczynski Lewis et al. 2007 .
[[40]] Page.6: #Analysis #Equation IAF was estimated using the center of gravity method for the frequency range of 7 Hz f1 14Hz f2 Klimesch, 1999 ,
[[41]] Page.6: #Analysis So as not to confound a trait measure with possible task related effects, we calculated IAF separately during the pre meditation baseline period and during meditation. IAF values obtained during the 1 min baseline period were used to anchor the frequency range defnitions of all EEG bands.
[[42]] Page.6: #Table 1 Comparison between fxed frequency bands and ranges based on IAF.
[[43]] Page.6: #Analysis For each frequency band, delta, theta, alpha, beta, and gamma , retreat Retreat 1, Retreat 2 , and group retreat, wait list control , a separate non parametric cluster based permuta tion test Maris and Oostenveld, 2007 was performed using FieldTrip Oostenveld et al., 2011 to fnd contiguous clusters of electrodes that differed in power as a function of assessment pre , mid , and post retreat testing . The minimum cluster size was set to three electrodes, with no maximum limit. Ten thousand per mutations were run to assess the signifcance of clusters, using a Monte Carlo estimation of signifcance. Signifcant clusters indi cate changes over assessments for the respective frequency band. False discovery rate FDR Benjamini and Hochberg, 1995 was used to control for individual testing of each retreat, group, and frequency band
[[44]] Page.6: #Analysis A hybrid non parametric/parametric approach was used to assess training related changes in spectral band power
[[45]] Page.7: #FIGURE 2 Groups were tested three times during each three month retreat period: at the beginning pre assessment , middle mid assessment , and end post assessment of each retreat
[[46]] Page.7: #Analysis #Results To test the effects of training on beta band power in Retreat 2, a repeated measures ANOVA was used to examine the within subjects effect of assessment pre , mid , and post in wait list participants as they underwent training during the second retreat. Log transformed beta band power was averaged across the clus ter found for Retreat 2 discussed above . A repeated measures ANOVA revealed a signifcant effect of assessment F 2 , 20 17 . 20, p 0 . 001 . Post hoc tests revealed a signifcant reduc tion in beta band power at mid t 21 4 . 43, p 0 . 001 and post assessments t 21 5 . 89, p 0 . 001 , compared to the pre assessment. Thus, the pattern and spatial topography of training related changes in beta band power was replicated in Retreat 2
[[47]] Page.8: #Results Taken together these analyses suggest that intensive FA training is associated with a reduction in beta band power during mind fulness of breathing, with effects most reliably observed bilaterally overlying medial prefrontal, central, and parietal brain regions.
[[48]] Page.8: #Results Changes in IAF during Retreat 1 were examined using multivariate ANOVA in a manner analogous to the beta band power analy ses summarized above. A 3 assessment 2 group ANOVA revealed a signifcant main effect of assessment F 2 , 41 23 . 26, p 0 . 001 , indicating that IAF values shifted across time. The main effect of group F 1 , 42 0 . 13, p 0 . 72 was not signif cant. As predicted, a signifcant group assessment F 2 , 41 6 . 40, p 0 . 01 interaction was found, suggesting training related shift in IAF across three months of meditation training.
[[49]] Page.9: #Results In order to test changes in IAF during Retreat 2, a repeated measures ANOVA was used to examine the within subjects effect of assessment. A signifcant main effect was found F 2 , 20 35 . 44, p 0 . 001 . Similar to the observed pattern for the ini tial retreat group during Retreat 1, a decrease in IAF was found at both the mid t 21 7 . 30, p 0 . 001 and post assessments t 21 6 . 00, p 0 . 001 , as compared to the pre assessment Figure 4 . Again, no reductions in IAF were found between the mid and post assessments.
[[50]] Page.9: #Results The overall pattern of results suggests training related decreases in IAF in both retreats. Although these data demon strate reliable training related reductions in IAF across three months of meditation training, reductions in IAF were also observed between pre and mid assessments for the wait list con trols in Retreat 1. However, the effect size for pre to mid change in IAF was nearly three times as large for the training groups in both Retreat 1 Cohen s d 1.40 and Retreat 2 d 1 . 56 than in wait list controls d 0 . 56 .
[[51]] Page.9: #Hypothesis #MeditationTime Hierarchical multiple regression analysis was used to examine whether decreases in beta band power and/or shifts in IAF were related to the amount of average self reported daily FA medita tion. Because the pattern of training related change in beta and IAF was similar across both retreats, data from both groups were combined.
[[52]] Page.9: #Results #MeditationTime #Beta As expected, pre assessment beta signifcantly predicted post assessment beta R 2 0 . 804, F 1 , 42 171 . 79, p 0 . 001 . The second step included the average daily amount of FA medi ation practiced by each participant in order to examine the unique variance explained by daily practice in post assessment beta independent of pre assessment beta. In step 2, the addi tion of average daily FA hours did not add signifcantly to the explained variance of the model R 2 0 . 003, F 1 , 41 0 . 57, p 0 . 46 . Thus, collapsed across retreats, changes in beta were not predicted by the amount of participants daily FA meditation practice.
[[53]] Page.9: #Results #MeditationTime #Alpha In a similar step wise manner, we used a hierarchical regres sion model to examine whether the amount of FA meditation predicted changes in IAF. In the rst step, pre assessment IAF sig ni cantly predicted post assessment IAF R 2 0 . 870, F 1 , 42 280 . 134, p 0 . 001 see Table 2 . In step 2, the addition of average daily FA meditation hours added signi cantly to the explained variance of the model R 2 0 . 020, F 1 , 41 7 . 45, p 0 . 009 . This relation was negative 0 . 142 , indicating that the more the participants engaged in FA meditation the more IAF decreased. In contrast to changes in beta band power, these results suggest that reductions in IAF are signi cantly predicted by the amount of daily FA meditation engaged in over the course of meditation training.
[[54]] Page.10:#Results #Altitude it is possible that traveling to and living at that higher alti tude could have in uenced changes in spectral power and/or IAF over time Kaufman et al., 1993 Guger et al., 2005 . To rule out this explanation, we used hierarchical multiple regres sion analysis to examine whether the reductions in beta band power and IAF could be predicted by the difference in ele vation between the testing location and the participant s city of residence. All participants except one resided at a lower elevation than the retreat center N 43 Melevation difference 1770 . 12 m, SDelevation difference 767 . 21 m the single participant who resided at a slightly higher elevation than the retreat cen ter was excluded from analysis. As in the analysis of daily FA meditation, the frst step of each regression model included the pre assessment level of beta or IAF, respectively. The addition of altitude change in step 2 did not add signifcantly to the explained variance of the model for either beta band power R 2 0 . 01, F 1 , 41 2 . 26, p 0 . 14 or IAF R 2 0 . 003, F 1 , 41 0 . 85, p 0 . 36 . These analyses suggest that changes in IAF and beta band power were unrelated to changes in altitude.
[[55]] Page.10:#Discussion The robustness of these fndings, replicated across separate training periods, provide evidence of specifc longitu dinal changes in characteristic brain oscillatory activity obtained during mindfulness of breathing.
[[56]] Page.10:#Discussion #Beta This suggests a func tional role for beta band activity within tasks involving attention to tactile stimuli. Specifcally, anticipatory modulations in beta band power have been associated with spatiotemporal orienting of attention Dockstader et al., 2010 Jones et al., 2010 van Ede et al., 2010, 2011 and conscious detection of subtle tactile stim uli Linkenkaer Hansen et al., 2004 Schubert et al., 2009 . For example, when attention is cued to a lateralized tactile stimulus, beta band power over parietal cortex is suppressed contralat eral and increased ipsilateral to the attended stimulus van Ede et al., 2011 . The degree of prestimulus suppression is associ ated with both faster responding and enhanced stimulus detection Linkenkaer Hansen et al., 2004 van Ede et al., 2011 .
[[57]] Page.10:#Discussion #Beta Training related suppression of beta band power in the present study is in line with the functional role proposed for oscillatory activity in facilitating sensory pro cessing of the attended breath stimulus.
[[58]] Page.10:#Discussion #Beta #Meditation Thus, reduc tions in beta band power over the course of training may re ect increased cortical activation of sensory related attentional net works and an increased capacity to focus attention on the breath during mindfulness of breathing meditation. Beta suppression may also re ect increased perceptual discrimination and con scious perception of these tactile sensations. In support of this idea, Schubert et al. 2009 found that the absolute magnitude of beta suppression across individuals was associated with a greater ability to perceive target tactile stimuli within a context of sim ilar distracters. In the present cohort, training was previously reported to increase perceptual discrimination of subtle visual stimuli MacLean et al., 2010 . In a similar manner, we speculate that intensive meditative practice may result in increased levels of sensory processing of ongoing tactile stimuli.
[[59]] Page.11:#Discussion #Meditation #Attention Finally, the observed pattern of training related modulation of oscillatory activity may be associated with improvements in behavioral measures of attentional performance. In this same cohort, we previously reported improvements in sustained atten tion, response inhibition, and perceptual discrimination follow ing training MacLean et al., 2010 Sahdra et al., 2011 . Repeated engagement of attention networks during practice may allow for more effcient resource allocation during demanding external psychophysical tasks. For example, changes in neural signatures of attentional stability have been found following three months of intensive Vipassana meditation, which includes FA meditation as one component of training Lutz et al., 2009 .
[[60]] Page.11:#FutureDirections This link should be examined in future work by directly relating measures of cor tical activation during practice and subsequent training related behavioral outcomes. Furthermore, theoretical approaches, such as computational modeling, can be employed to test targeted hypotheses regarding the underlying cortical dynamics involved in these processes
[[61]] Page.11:#Discussion #Alpha we believe that diminished IAF may represent an overall reduction in cognitive effort during medi tation. Peak alpha frequency has been shown to increase with elevated cognitive load in visuospatial working memory Moran et al., 2010 and may re ect a greater allocation of resources to the maintenance of information in memory. After intensive training, increased attentional stability may reduce the atten tional resources required to sustain attention on the sensations of the breath. Additionally, meditation training may promote greater effciency in the reorienting of attention to a given stimu lus.
[[62]] Page.11:#Limitations #Gamma #SignalProcessing Previous studies have reported increases in gamma band activity during meditation in experienced practitioners Lutz et al., 2004 Cahn et al., 2010 . Analysis of gamma band power in humans is notoriously challenging due to the contribution of non neural sources. Scalp recorded muscle activity gener ates broadband myogenic electrical noise that overlaps sub stantially with the gamma band, which may also in uence alpha and beta bands McMenamin et al., 2010, 2011 . In the present study, we utilized novel signal processing methods to remove such putatively non neural signal sources from the ongo ing EEG. This may have contributed to the lack of changes in gamma band activity with training. Furthermore, in con trast to previous studies that used standard ranges for each frequency band, we defned the range of each spectral band according to each participant s IAF during rest. This approach accounts for individual differences in the frequency band bound aries and therefore may provide a more accurate measure of activity.
[[63]] Page.12:#FutureDirections Future research should therefore explore the potential effects of posture on state related meditation effects.
[[64]] Page.12:#Limitations Finally, motivational levels may not have been exactly matched across groups. Participants receiving training were not blind to their group assignment and therefore our results may have been susceptible to demand char acteristics resulting from varying levels of in uence from teacher expectations and commitment to a general Buddhist worldview.
[[65]] Page.12:#Conclusion In summary, we found replicable and robust reductions in beta band power over central parietal regions and decreased IAF following three months of intensive FA meditation.