Received: 28 May 2018 | Revised: 6 October 2018 | Accepted: 15 October 2018

DOI: 10.1002/brb3.1159

ORIGINAL RESEARCH

Short‐term Sahaja Yoga meditation training modulates brain

structure and spontaneous activity in the executive control
network

Alessandra Dodich1,2 | Maurizio Zollo3 | Chiara Crespi1,4 | Stefano F. Cappa4,5 |

Daniella Laureiro Martinez6 | Andrea Falini1,7 | Nicola Canessa4,8

1
Division of Neuroscience, San Raffaele
Scientific Institute, Milan, Italy Abstract
2
NIMTlab, University of Geneva, Geneva, Introduction: While cross‐sectional studies have shown neural changes in long‐term
Switzerland
meditators, they might be confounded by self‐selection and potential baseline differ‐
3
Invernizzi Center for Research
ences between meditators and non meditators. Prospective longitudinal studies of
in Innovation, Organization and
Strategy, Bocconi University, Milan, Italy the effects of meditation in naïve subjects are more conclusive with respect to causal
4
NeTS Center, Scuola Universitaria Superiore inferences, but related evidence is so far limited.
IUSS, Pavia, Italy
5
Methods: Here, we assessed the effects of a 4‐week Sahaja Yoga meditation training
IRCCS San Giovanni di Dio Fatebenefratelli,
Brescia, Italy on gray matter density and spontaneous resting‐state brain activity in a group of 12
6
MTEC, ETH Zurich, Switzerland meditation‐naïve healthy adults.
Results: Compared with 30 control subjects, the participants to meditation training
7
Neuroradiology Unit, San Raffaele
Scientific Institute, Milan, Italy
8
showed increased gray matter density and changes in the coherence of intrinsic brain
Cognitive Neuroscience Laboratory, ICS
Maugeri, Pavia, Italy activity in two adjacent regions of the right inferior frontal gyrus encompassing the
anterior component of the executive control network. Both these measures corre‐
Correspondence
Alessandra Dodich, San Raffaele Scientific lated with self‐reported well‐being scores in the meditation group.
Institute, Milan, Italy. Conclusions: The significant impact of a brief meditation training on brain regions
Email: dodich.alessandra@hsr.it
associated with attention, self‐control, and self‐awareness may reflect the engage‐
Funding information
This work was partially supported by the
ment of cognitive control skills in searching for a state of mental silence, a distinctive
Cariplo Foundation Grant “Formazione feature of Sahaja Yoga meditation. The manifold implications of these findings in‐
Universitaria d’Eccellenza: applicazioni
neuro‐scientifiche per la formazione
volve both managerial and rehabilitative settings concerned with well‐being and
nella gestione dell’innovazione e della emotional state in normal and pathological conditions.
sostenibilità”.

KEYWORDS
fronto‐parietal executive control network, neural plasticity, resting‐state fMRI, Sahaja Yoga
meditation, voxel‐based morphometry

This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium,
provided the original work is properly cited.
© 2018 The Authors. Brain and Behavior published by Wiley Periodicals, Inc.

Brain and Behavior. 2018;e01159.  wileyonlinelibrary.com/journal/brb3 | 1 of 11

https://doi.org/10.1002/brb3.1159
2 of 11 | DODICH et al.

1 | I NTRO D U C TI O N & Golocheikine, 2001). SY differs from a concentrative type of med‐

itation because it involves a receptive, non judgmental disposition
Research over the past two decades has started to unveil the neural toward all experiences, irrespective of their origin (i.e., external or in‐
modifications associated with different types of meditation, resulting ternal) and affective tone (Reva et al., 2014). Previous studies reported
in beneficial effects on emotional balance and cognitive performance preliminary evidence of positive effects of SY meditation in disorders
(see Dahl, Lutz, & Davidson, 2015; Sedlmeier et al., 2012). Several such as asthma (Manocha, Marks, Kenchington, Peters, & Salome,
studies based on cross‐sectional comparisons suggest that the ef‐ 2002), epilepsy (Panjwani et al., 2000), attention‐deficit/hyperactiv‐
fects of long‐term meditation practice may result in neuro‐anatomical ity disorder (Harrison, Manocha, & Rubia, 2004), anxiety, and work
and functional changes. Despite differences in subject characteristics stress (Manocha, Black, Sarris, & Stough, 2011). The long‐term effects
and type of meditation, these studies highlighted significant changes of SY meditation on neural activity have been examined by cross‐
in fronto‐insular, anterior cingulate, and hippocampal cortex, regions sectional studies on expert meditators with electroencephalography
related to, respectively, self‐awareness, self‐regulation, and episodic (EEG) (Aftanas & Golocheikine, 2001, 2002; Aftanas & Golosheikin,
memory (see (Fox et al., 2014) for a review; (Villemure, Ceko, Cotton, 2003) and functional magnetic resonance imaging (fMRI) (Hernandez
& Bushnell, 2014)). Other studies addressed the effects of extensive et al., 2015). Their results suggest that SY meditation might promote
meditation training on intrinsic (task‐free) brain activity (Jang et al., functional changes in inferior frontal, parietal, and temporal regions
2011; Taylor et al., 2013), that is, the low‐frequency fluctuations of associated with sustained attention, self‐control, and self‐awareness
BOLD signal (<0.1 Hz) observed at rest in networks overlapping the (Hernandez et al., 2015), possibly reflecting increased awareness of
typical task‐related activation maps (Smith et al., 2009). However, the sensory stimuli and attentional control to the present moment. To the
cross‐sectional nature of these studies does not allow to establish a best of our knowledge, only one cross‐sectional study investigated
causal relationship between meditation practice and such changes. in expert meditators the long‐term effects of SY meditation on brain
To overcome this limitation, longitudinal studies assessed the ef‐ structure (Hernandez et al., 2016). Compared with novices, medita‐
fect of meditative interventions compared with a test–retest control tors displayed increased gray matter (GM) volume in several regions,
group. Their results suggest that short‐term meditation practice may mostly in the right hemisphere, associated with sustained attention,
induce significant morphologic changes via mechanisms of structural compassion, and interoceptive perception. Namely, neurostructural
plasticity (Holzel et al., 2011; Santarnecchi et al., 2014). To investigate changes in the right fronto‐insular and inferior temporal cortex were
training‐related changes in functional connectivity during mindful rest, suggestive of neural plasticity associated with regular practice of this
Kilpatrick and colleagues (Kilpatrick et al., 2011) employed a group in‐ meditation. To date, however, no prospective study has tested the
dependent component analysis (gICA), a data‐driven multivariate ap‐ causal effect of short‐term SY meditation practice on brain structure
proach highlighting subtle interindividual differences in intrinsic brain and/or function. We aimed to fill this gap with a longitudinal random‐
functioning (Koch et al., 2010). Subjects were explicitly instructed to ized controlled design assessing the effect of a 4‐week SY meditation
be mindfully aware of the MR scanner sounds, and the results showed training on GM density and intrinsic brain activity. We predicted sig‐
increased functional connectivity of the auditory cortex with brain re‐ nificant longitudinal structural changes associated with the medita‐
gions underlying attentional and self‐referential mechanisms. tive intervention in the fronto‐insular and inferior temporal regions
Based on their targeted cognitive processes, meditation prac‐ previously highlighted by cross‐sectional studies on expert SY med‐
tices can be grouped into two broad categories: focused attention itators (Hernandez et al., 2016). To avoid a priori assumptions, we
(i.e., concentrative) and open monitoring meditation (Lutz, Slagter, adopted a multimodal whole‐brain approach combining voxel‐based
Dunne, & Davidson, 2008). In concentrative practices, subjects de‐ morphometry and a blind gICA decomposition of resting‐state fMRI
velop regulative skills of selective attention, while in open monitoring data. We assessed intrinsic brain functioning in terms of spectral
meditation a greater emphasis is placed on cultivating a “reflexive” power of resting‐state networks (RSNs), a measure of the coherence
awareness. It is thus likely that these two training procedures may of intranetwork intrinsic activity (maximal for high power spectra at
exert different effects on brain activity and/or structure (Travis & low frequencies). We chose to constrain our analysis of resting‐state
Shear, 2010). evidence to power spectra because, while representing a sensitive
Among open monitoring practices, Sahaja Yoga (SY) meditation is measure of individual differences in intranetwork intrinsic brain func‐
a technique based on the attempt to obtain the state of mental si‐ tioning (Allen et al., 2011), this metric prevents the need of a priori
lence (i.e., free from unnecessary mental activity), during which all hypotheses of the regions included in the network(s).
the attention is on the present moment (Aftanas & Golocheikine,
2001; Hernandez, Suero, Barros, Gonzalez‐Mora, & Rubia, 2016;
Hernandez, Suero, Rubia, & Gonzalez‐Mora, 2015; Reva, Pavlov, 2 | M ATE R I A L S A N D M E TH O DS
Loktev, Korenyok, & Aftanas, 2014). With practice, the short‐term in‐
terruption of mental activity can evolve into an enduring absence of
2.1 | Subjects
narrative thought. This mental state of “thoughtless awareness” is as‐
sociated with the experience of joy, a subsequent sense of relaxation Forty‐five healthy, right‐handed (Oldfield, 1971) monolingual na‐
and positive mood, as well as an increase in self‐awareness (Aftanas tive Italian undergraduate students (14 females and 31 males, mean
DODICH et al. | 3 of 11

TA B L E 1 Characteristics of meditation
MG (n = 12) CG (n = 30) Statistics
(MG) and control (CG) groups at baseline
(mean ± standard deviation) Demographics
Age (years) 21.63 ± 2.02 22.16 ± 1.33 t(40) = 0.99, p = 0.33
Sex (F/M) 2/10 12/18 X2(1) = 2.1, p = 0.15
Temperament and character inventory (TCI‐56)
Harm avoidance 21.08 ± 6.51 22.87 ± 5.08 t(40) = 0.94, p = 0.35
Novelty seeking 22.75 ± 4.16 21.8 ± 4.2 t(40) = 0.98, p = 0.5
Reward Dependence 27.58 ± 7.14 29.3 ± 5.56 t(40) = 0.84, p = 0.4
Persistence 28 ± 3.64 28.73 ± 5.1 t(40) = 0.45, p = 0.65
Self‐directedness 28.83 ± 5.5 29.7 ± 4.9 t(40) = 0.51, p = 0.6
Cooperativeness 27.75 ± 3.47 29.93 ± 4.7 t(40) = 1.4, p = 0.15
Self‐transcendence 19.42 ± 8.7 22.57 ± 7.98 t(40) = 1.25, p = 0.27

age = 21.68, SD = 1.57, range = 19–25) were recruited at Bocconi via non parametric analyses (Wilcoxon signed‐rank test) comparing
University, Milan. Exclusion criteria were MR incompatibility, prior pre‐ and post‐training scores obtained in a questionnaire providing a
experience with meditation, history of neurological or psychiatric measure of emotional self‐assessment. This questionnaire, designed
disease, previous or current use of substances, or any psychoac‐ ad hoc based on standardized models (Crawford & Henry, 2004),
tive medications. Participants gave their written informed consent ­included 19 questions based on a 10‐level Likert scale assessing gen‐
to the experimental procedure, in accordance with the local Ethics eral well‐being, subjective experience directly related to meditation
Committee. Subjects were randomly assigned to a 4‐week SY train‐ (i.e., thoughtless awareness and vibration), as well as the presence
ing course (meditation group, MG) or to a 4‐week waiting period of negative (e.g., anger and fatigue) and positive (e.g., joy and peace)
(control group, CG). Three participants assigned to the MG dropped emotional states. Despite the existence of validated questionnaires
out due to lack of time (n = 1) or failure in attending classes (n = 2). for mood state evaluation, we opted for an ad hoc questionnaire as
The final sample thus included 12 subjects in the MG (2 females, we aimed at measuring the subjective experiences specifically re‐
mean age = 21.63, SD = 2.02) and 30 subjects in the CG (12 fe‐ lated to Sahaja Yoga meditation.
males, mean age = 22.16, SD = 1.33) (Table 1). At baseline, subjects
completed the short version of the Temperament and Character
2.3 | MRI data acquisition
Inventory (TCI‐56) (Adan, Serra‐Grabulosa, Caci, & Natale, 2009),
assessing four temperament (i.e., harm avoidance, novelty seeking, Our longitudinal study design entailed two time points for each
reward dependence, and persistence) and three character (i.e., self‐ group, in which we collected anatomical images for voxel‐based
directedness, cooperativeness, and self‐transcendence) dimensions morphometry (VBM) analyses of GM density and resting‐state
(Table 1). functional images for analyses of intrinsic brain activity. We used
a 3 Tesla Philips Achieva scanner (Philips Medical Systems, Best,
NL), equipped with an 8‐channel sense head coil (sense reduc‐
2.2 | Training protocol and behavioral
tion factor = 2), to collect anatomical T1‐weighted (150 slices,
statistical analyses
TR = 600 ms, TE = 20 ms, slice thickness = 1 mm, in‐plane reso‐
The SY meditation training consisted of four one‐hour sessions per lution = 1 × 1 mm) and functional T2*‐weighted (gradient‐echo,
week over 4 consecutive weeks, that is, a total of 16 hr. The overall echo‐planar pulse sequence, 37 continuous ascending trans‐
duration and schedule of the course was planned based on previous verse slices covering the whole brain, tilted 30° downward with
evidence on the behavioral effects of short‐term meditation train‐ respect to the bicommissural line to reduce susceptibility arti‐
ing (Chung, Brooks, Rai, Balk, & Rai, 2012; Fox et al., 2014; Holzel facts in orbitofrontal regions; TR = 2,000 ms, TE = 30 ms, flip
et al., 2011). Participants received guided meditation instructions angle = 85°, FOV = 192 × 192 mm, slice thickness = 3.7 mm, inter‐
and joined group activities designed to promote the state of mental slice gap = 0.55 mm, in‐plane resolution = 2 × 2 mm) images. The
silence, known also as thoughtless awareness, held by an instruc‐ resting‐state scan was preceded by six “dummy” functional vol‐
tor with over 25 years (approximately 2,700 hr) of teaching expe‐ umes, which were automatically discarded, covering the amount
rience. Each hour of daily training was divided into 10–15 min of of time needed to allow for T1 equilibration effects. Participants
theoretical lectures, introducing each time a new subject, followed were positioned comfortably on the scanner bed and fitted with
by 45–50 min of meditative practice experience, including various soft earplugs; foam pads were used to minimize head movement.
workshop techniques (for further details, see (Manocha, 2014)). No They were instructed to lie quietly with their eyes open and stare
other commitments, including home practice, were required. In the passively at a foveally presented gray fixation cross to facilitate
MG, the effect of SY training on subjective well‐being was assessed network delineation (Van Dijk et al., 2010).
4 of 11 | DODICH et al.

2.4 | VBM data preprocessing and analyses correlation between the SPM EPI template and both the unsmoothed
and smoothed mean images computed from the series of normalized
We performed data preprocessing with the VBM8 toolbox (https://
and realigned functional volumes. An ANOVA with such correla‐
dbm.neuro.uni-jena.de/vbm/), an extension of the SPM8 software
tion index as dependent variable highlighted no significant main ef‐
(https://www.fil.ion.ucl.ac.uk/spm/), running on MATLAB v7.4
fect of group (smoothed: F[1,40] = 1.005, p = 0.322; unsmoothed:
(MathWorks, Inc., Sherborn, MA). VBM preprocessing of the T1‐
F[1,40] = 0.14, p = 0.71), time point (smoothed: F[1,40] = 0.08,
weighted images included bias correction for field‐intensity inhomo‐
p = 0.78; unsmoothed: F[1,40] = 0.23, p = 0.635), or interaction be‐
geneities, coregistration of the second image to the baseline image,
tween group and time point (smoothed: F[1,40] = 0.001, p = 0.97;
spatial normalization of the images to a standardized anatomical
unsmoothed: F[1,40] = 0.27, p = 0.605).
space, segmentation of the GM component from the normalized
images, and smoothing of the normalized GM maps with an 8‐mm
full‐width half‐maximum (FWHM) Gaussian kernel. Preliminary 2.6 | Group Independent Component Analysis
analyses based on a two‐sample t test excluded statistically signifi‐ (gICA)
cant baseline group differences. The effect of meditation training
We used multivariate spatial gICA, as implemented in the GIFT
was assessed with a repeated‐measures ANOVA (i.e., flexible facto‐
toolbox (https://icatb.sourceforge.net;(Allen et al., 2011; Calhoun,
rial model in SPM8) with time and group as within‐ and between‐­
Adali, Pearlson, & Pekar, 2001a), to extract temporally coherent and
subject factors, respectively. We tested an interaction between time
maximally independent spatial sources, that is, functional networks
and group using a statistical threshold of p < 0.05 FWE‐corrected at
or “spatial maps,” from resting‐state time courses. The independ‐
the cluster level (p < 0.001 uncorrected at the voxel level). We local‐
ent component analysis was preceded by a data‐reduction stage,
ized the clusters showing significant effects using the cytoarchitec‐
based on a principal component analysis (PCA) retaining 100 prin‐
tonical mapping implemented in the SPM Anatomy Toolbox v 2.2c
cipal components from single subjects’ time courses (Erhardt et al.,
(Eickhoff et al., 2005). To assess the relationship between GM den‐
2011). Subsequent gICA retained 75 components through a neural
sity and subjective well‐being, we performed non parametric corre‐
network algorithm (Infomax) that attempts to minimize the mutual
lations (i.e., Spearman’s ρ) between post‐treatment scores obtained
information of the network outputs to identify naturally grouping
at the emotional self‐assessment questionnaire and the average GM
and maximally independent sources (Bell & Sejnowski, 1995). ICA
values extracted from the significant clusters highlighted by interac‐
was repeated 250 times in Icasso (https://www.cis.hut.fi/projects/
tion analyses.
icasso). The resulting components were clustered to ensure the
consistency and reliability of the decomposition, which are quanti‐

2.5 | rs‐fMRI data preprocessing fied using a quality index Iq ranging from 0 to 1, reflecting the dif‐
ference between intracluster and extracluster similarity (Himberg,
Image preprocessing was performed using SPM8 (https://www.fil. Hyvarinen, & Esposito, 2004). Subject‐specific spatial maps and
ion.ucl.ac.uk/spm), implemented in MATLAB v7.4 (MathWorks, Inc., time courses were estimated with GICA3 back‐reconstruction
Sherborn, MA). The 150 volumes from each subject underwent a (Calhoun, Adali, Pearlson, & Pekar, 2001b; Erhardt et al., 2011).
standard preprocessing including slice‐timing correction with the
middle slice in time as a reference, spatial realignment to the first
2.7 | Resting‐state networks selection and
volume and unwarping, spatial normalization into the standard
identification
Montreal Neurological Institute (MNI) space (Friston et al., 1995),
and resampling in 2 × 2 × 2 mm3 voxels, as well as spatial smoothing Alongside the Iq index, we used the spectral characteristics of
with a 8‐mm full‐width half‐maximum (FWHM) isotropic Gaussian component time courses to discriminate reliable resting‐state
kernel. We then applied a procedure of “intensity normalization,” networks (RSNs) from physiological artifacts. Based on the no‐
converting the time series of each voxel to percent signal change tion that normal resting‐state time courses are dominated by slow
units, to improve the accuracy and test–retest reliability of the out‐ (i.e., low frequency) fluctuations (Cordes et al., 2000), we evaluate
put components of the subsequent independent component analy‐ “dynamic range” (i.e., the difference between peak spectral power
sis (ICA) (Allen, Erhardt, Eichele, Mayer, & Calhoun, 2010; Allen and minimum power at frequencies to the right of the peak) and
et al., 2011). We used the Motion Fingerprint toolbox. “low‐­frequency‐­to‐high‐frequency power ratio” (i.e., the ratio of
(https://www.medizin.uni-tuebingen.de/kinder/en/research/ the integral of spectral power below 0.1 Hz to the integral of power
neuroimaging/software/) to compute, for each subject, a compre‐ between 0.15 and 0.25 Hz) (Allen et al., 2011). In addition, the ag‐
hensive indicator of scan‐to‐scan head motion. An ANOVA including gregate spatial maps underwent a visual inspection by three inde‐
this measure as dependent variable highlighted no significant main ef‐ pendent raters, based on expectations that they should involve GM
fect of group (F[1,40] = 0.001, p = 0.98) or time point (F[1,40] = 2.81, rather than known ventricular, vascular, susceptibility or motion‐
p = 0.101), nor a significant interaction between group and time point related artifacts. Each rater scored spatial maps by assigning them
(F[1,40] = 0.48, p = 0.494). We also assessed the quality and consis‐ to one of three possible classes, that is, definite artifact (0), mixed
tency of spatial normalization across subjects by computing the spatial (1), or genuine resting‐state component (2). We retained only the
DODICH et al. | 5 of 11

components assigned to the latter class by all raters. The two spec‐ to convert the gICA maps showing a significant interactive effect
tral characteristics, alongside an Iq > 0.8 and the visual inspection into binary regions of interests (ROIs). Then, we used the toolbox
of the aggregate spatial maps, led to select a subset of 41 out of 75 REX (https://web.mit.edu/swg/software.htm) to extract, from the
components as genuine RSNs. We anatomically labeled the selected resulting masks, average GM density values for subsequent off‐line
RSNs based on the largest spatial correlation between the spatial statistical analyses. For each resulting map, we used a mixed ANOVA
maps of each component and the RSNs template provided with the to compare GM density in meditators vs controls before and after
GIFT toolbox. A refined labeling of the components showing a sig‐ treatment.
nificant time‐by‐group interaction (see below) was performed using All the reported results survived a statistical threshold of p < 0.05
the cytoarchitectonic maps implemented in the SPM Anatomy tool‐ corrected for multiple comparisons using false discovery rate (FDR;
box v2.2c (Eickhoff et al., 2005). (Genovese, Lazar, & Nichols, 2002)).

2.8 | RSN statistical analyses 3 | R E S U LT S

In subsequent statistical analyses, we considered the power spec‐
3.1 | Behavioral results
tra of RSN time course, representing the contribution of specific
frequency bins to the slow synchronous fluctuations of the BOLD The final samples did not differ significantly in mean age (t[40] = 0.99,
signal. This metric reflects the degree of intranetwork coherent p = 0.33) or gender (X2[1] = 2.1, p = 0.15). In addition, no significant
activity, which is maximal at the typical resting‐state frequencies differences were found between MG and CG in any of the TCI‐56
below 0.1 Hz, while a relative shift toward higher frequencies has subscales (Table 1). Due to the small sample size, we performed
been proposed to reflect altered intranetwork connectivity in physi‐ preliminary analyses on behavioral variables with a parametric dis‐
ological aging (Allen et al., 2011) and neurological diseases (Caminiti tribution, which excluded the presence of outliers among subjects.
et al., 2015). We estimated spectra on the detrended subject‐spe‐ All MG subjects included in the final sample completed the 16 hr
cific time courses, after removal of the mean, slope, and period π and of scheduled training. The within‐group analysis on the scores of
2π sines and cosines over each time course. emotional self‐assessment revealed a significant increase in general
In the main analysis, we assessed the effects of group (i.e., MG well‐being (Z = 3.06, p = 0.002), as well as a significant reduction in
and CG), time (i.e., pretreatment and post‐treatment), as well as the fatigue (Z = 2.67, p = 0.007) and dissatisfaction (Z = 1.91, p = 0.05),
interaction between group and time, by using a backward multivar‐ after the meditative training (Table 2).
iate model selection strategy for the outcome variable (i.e., power
spectra). This procedure first employs a multivariate analysis of co‐
3.2 | VBM results
variance (MANCOVA) to select which factors explain variability in
the outcome measure. Then, univariate tests corrected for multiple We found no significant group difference in GM density at base‐
comparisons are carried out on a reduced design matrix (thus de‐ line. Time‐by‐group interaction analyses showed, in meditators
creasing the number of statistical tests performed), to highlight the compared with control subjects, a significant increase after train‐
direction and strength of the relationship between retained factors ing of GM density in a cluster encompassing the right inferior fron‐
and power spectra. As for VBM analyses, we first performed prelimi‐ tal gyrus (k = 336, local maxima MNI coordinates x;y;z = 40;26;−21)
nary analyses comparing power spectra metrics in MG and CG to ex‐ (Figure 1a,c). A 2 × 2 repeated‐measures ANOVA confirmed a sig‐
clude significant baseline group differences. The effects of interest nificant “time × group” interaction (F[1,40] = 24.84, p < 0.001),
in the design matrix were group, time, and their interaction, as well with no significant effects of time (F[1,40] = 0.059, p = 0.809) and
as two nuisance predictors reflecting the quality of spatial normal‐ group (F[1,40] = 2.58, p = 0.116) alone. Moreover, non parametric
ization (of smoothed normalized images) and average scan‐to‐scan correlation analyses showed a positive correlation between aver‐
head motion. As previously discussed, we also aimed to test a rela‐ age GM density in this cluster and general well‐being after training
tionship between changes in the metrics of intrinsic brain activity (Spearman’s ρ = 0.615, p = 0.03).
and subjective well‐being. To this purpose, we then used non para‐
metric statistics (i.e., Spearman’s ρ) to assess a correlation between
3.3 | RSN networks
power spectra in the frequency bins displaying a significant time‐by‐
group interaction and the scores obtained at the questionnaire of The visual inspection of the spatial maps, alongside the analy‐
emotional self‐assessment in the post‐treatment stage. sis of spectral profiles, led to select 41 components which largely
To investigate a connection between the functional (rs‐fMRI) and overlapped the RSNs previously described (e.g. (Allen et al., 2011;
structural (morphometric) levels of analysis, we also tested whether Caminiti et al., 2015)). Namely, they involved the anterior and pos‐
meditation training was associated with a significant increase of GM terior components of the default mode network (Buckner, Andrews‐
density in the network(s) showing an interactive effect of group Hanna, & Schacter, 2008), as well as the anterior and posterior
and time on resting‐state power spectra. To this purpose, we first portions of the salience network (Seeley et al., 2007), alongside the
used the SPM toolbox Marsbar (https://marsbar.sourceforge.net) dorsal attentional network (Corbetta & Shulman, 2002). Distinct
6 of 11 | DODICH et al.

TA B L E 2 Training‐related effects on
Meditation group
well‐being and emotional self‐assessment
Pretraining Post‐training Statistics

Emotional self‐assessment
General well‐being 5.33 (0.98) 7 (0.739) Z = 3.06, p = 0.002
Thoughtless awareness 6.48 (1.76) 7.49 (1.35) Z = 1.34, p = 0.18
Vibration cool/warm 7.13 (0.84) 7.54 (1.28) Z = 0.71, p = 0.48
Nervousness 2.97 (1.74) 2.11 (1.39) Z = 1.61, p = 0.18
Fatigue 4.65 (1.73) 3.18 (2.20) Z = 2.67, p = 0.007
Anger 2.88 (2.39) 1.76 (1.53) Z = 1.16, p = 0.25
Fear 1.10 (087) 1.09 (1.04) Z = 0.35, p = 0.72
Dissatisfaction 3.79 (2.56) 2.38 (1.57) Z = 1.91, p = 0.05
Guilt 2.82 (2.34) 2.90 (1.91) Z = 0.26, p = 0.79
Joy 6.39 (1.33) 6.12 (1.69) Z = 0.84, p = 0.40
Peace 5.82 (2.51) 6.74 (1.76) Z = 1.1, p = 0.27
Satisfaction 4.68 (2.39) 6.11 (1.49) Z = 1.69, p = 0.09
Compassion 4.47 (2.69) 5.44 (1.76) Z = 1.88, p = 0.06
Self‐confidence 5.51 (2.39) 6.38 (1.73) Z = 1.41, p = 0.15
Aware of attention 5.05 (2.40) 5.77 (1.94) Z = 0.18, p = 0.86
In control of attention 6.24 (2.07) 6.21 (1.63) Z = 0.039, p = 0.97
Creativity 5.01 (2.25) 5.38 (1.79) Z = 0.47, p = 0.63
Open to change 7.33 (1.17) 7.68 (1.04) Z = 1.38, p = 0.17
Openness with others 5.19 (2.32) 5.47 (1.73) Z = 0.94, p = 0.34

Note. Scores obtained at the questionnaire of emotional self‐assessment by the meditation group
before and after training. The mean and standard deviation (in brackets) are reported for each vari‐
able. Significant effects are depicted in bold font.

frontal networks included regions belonging to the fronto‐­parietal There was no significant effect of training on GM density in
executive control network (Vincent, Kahn, Snyder, Raichle, & the whole component highlighted by the above resting‐state find‐
Buckner, 2008), as well as fronto‐temporal and fronto‐limbic regions ings. However, when assessing its different subclusters separately
associated with, respectively, linguistic and affective processing. we observed that the GM increase from pre‐ to post‐training was
Finally, we identified components representing both primary and larger, in the MG than CG (t[40] = 2.22, p = 0.03), in a right inferior
higher‐level visual networks, as well as the sensorimotor and audi‐ fronto‐insular region adjacent to that highlighted by VBM analyses
tory networks. (Figure 1a).

3.4 | Group effects on power spectra

4 | D I S CU S S I O N
As for VBM analyses, we found no significant baseline group differ‐
ence in terms of resting‐state metrics. A time‐by‐group interaction Available evidence suggests that distinct meditation practices re‐
on the level of coherent activity (i.e., power spectra) was assessed flect in a differential involvement of neural systems associated
in all the retained components. We found a significant interaction with selective and sustained attention vs monitoring, vigilance,
in the frontal sector of the fronto‐parietal “executive control” net‐ and representation of internal bodily states (Lutz et al., 2008). In
work (Vincent et al., 2008), involving the bilateral dorsolateral and particular, SY meditation has received growing attention in the
dorsomedial prefrontal cortex, the dorsal anterior cingulate cortex last years due to its beneficial effects on different psycho‐physi‐
(dACC), and bilateral fronto‐insular cortex (Figure 1a). Within this ological variables (Aftanas & Golocheikine, 2001; Chung, Brooks,
component, the MG displayed compared with CG decreased power Rai, Balk, & Rai, 2012). While previous related studies have taken
at ultra‐low frequencies (<0.02 Hz) and increased power in the typi‐ a cross‐sectional approach to assess the long‐term effects of this
cal low–middle frequencies (<0.1 Hz) after training (Figure 1b). In ad‐ type of meditation (Aftanas & Golocheikine, 2001; Aftanas &
dition, higher post‐training well‐being scores were reflected in lower Golosheikin, 2003; Hernandez et al., 2015, 2016; Reva et al., 2014),
power spectra at ultra‐low frequencies (<0.02 Hz) within the same here we employed a longitudinal randomized controlled approach to
network (Table 3). investigate a causal relationship between meditation practice and
DODICH et al. | 7 of 11

F I G U R E 1 (a) Spatial contiguity between the right inferior frontal clusters showing increased GM density (in blue) and a modulation of
coherent activity (in red) after meditation training. The overlap between morphometric and resting‐state data is shown in green. (b) Spectral
power of intrinsic activity in the executive control network, providing a measure of the contribution of each frequency bin (between 0 and
0.25 Hz) to the fluctuations of BOLD signal at rest (asterisks indicate the frequency bins displaying a significant effect in time‐by‐group
interaction). Meditators, compared with non meditators, display a reduction of power at ultra‐low frequencies, and an increase at low–
middle frequencies, after training. (c) Average GM density in the cluster resulting from VBM interaction analysis for the two time points of
both training (MG) and control (CG) groups (error bars depict standard deviations). Meditators, compared with non meditators, display a
significant increase of GM density with training in the right fronto‐insular cluster depicted in green color in panel A

TA B L E 3 Relationship between
Power spectra
well‐being and neural metrics after
training Ultra‐low frequencies (<0.02 Hz)
GM
Bin 7 Bin 8 Bin 9 Bin 10 density

General well‐being r = −0.71 r = −0.72 r = −0.54 r = −0.51 r = 0.615

p < 0.01 p < 0.01 p = 0.07 p = 0.08 p < 0.05
Low frequencies (<0.1 Hz)
Bin 48
Fatigue r = −0.56 p = 0.05

Note. Significant correlations in the meditation group between resting‐state metrics, GM density,
and scores at the questionnaire of self‐emotional assessment after training. GM: Gray matter.

cortical reorganization, as indexed by brain structure and spontane‐ after training. In addition, compared with control subjects they
ous activity. also showed a significant change in brain structure and intrinsic
Subjects participating in the meditation intervention displayed activity in inferior fronto‐insular regions associated with executive
a significant improvement in self‐perceived general well‐being control, and previously highlighted by cross‐sectional studies on
8 of 11 | DODICH et al.

the neural effects of long‐term SY meditation (Hernandez et al., have been associated with an increased awareness of momentary
2016). self‐reference. The latter result was also associated with reduced ac‐
With regard to the functional modifications observed in in‐ tivity in key nodes of the default mode network, related to extended
trinsic brain functioning, we provide novel evidence of a causal and narrative self‐reference, which may reflect the enduring absence
connection between SY practice and significant changes in the of narrative thought prompted by meditation (Mason et al., 2007).
distribution of low–middle frequencies of resting‐state activity. It is noteworthy that one of the few previous studies investigating
A significant time‐by‐group interaction on spectral power high‐ the neuro‐functional correlates of mental silence in SY meditation
lighted, in meditators vs controls, a training‐related remodeling of experts has shown a significant relationship between thoughtless
the contribution of slow frequencies in the anterior component of awareness and the activation in the middle/superior temporal and
the fronto‐parietal “executive control” network. The anterior sec‐ right fronto‐insular cortex, with the latter being directly related to the
tion of this network, comprising the dorsolateral and dorsomedial subjective depth of the mental silence experience (Hernandez et al.,
prefrontal cortex, the dACC, as well as the inferior fronto‐­insular 2015). Overall, these data consistently point to the role of the execu‐
cortex bilaterally, is thought to be involved in goal‐directed be‐ tive control network in orienting attention to the external or internal
havior (Spreng, Stevens, Chamberlain, Gilmore, & Schacter, 2010; environment as a prerequisite of mental silence (i.e., interruption of
Sutherland, McHugh, Pariyadath, & Stein, 2012; Vincent et al., unnecessary mental activity). Interestingly, the same authors have also
2008). This set of regions, which is commonly recruited by tasks provided evidence of increased GM in fronto‐insular cortex in long‐
requiring controlled information processing (Dosenbach et al., term SY meditators (Hernandez et al., 2016).
2007), has been recently proposed as a “superordinate” cognitive These findings, highlighting the effects of SY meditation on the
control network, recruited across different executive domains functional and structural properties of the executive control net‐
including flexibility, working memory, initiation, and inhibition work, fit with the present evidence of morphometric changes, after
(Niendam et al., 2012). Available evidence suggests that specific meditation training, in the same right fronto‐insular region described
nodes of this network play a key role in orienting the attentional above. While analogous morphometric changes resulting from dif‐
focus to the external vs internal environment (Spreng, Sepulcre, ferent short‐term trainings have been previously reported (e.g. after
Turner, Stevens, & Schacter, 2013). Among these regions, the a visuomotor training in (Draganski et al., 2004)), their functional
right inferior fronto‐insular cortex acts as a critical node in sup‐ meaning is still debated. Alongside the functional changes in intrin‐
pressing default activity and re‐allocating attentional resources to sic brain activity previously discussed, the increase of GM density
salient events via the engagement of dorsal attention networks in the right inferior fronto‐insular cortex with meditation training
(Sridharan, Levitin, & Menon, 2008). The role of this region in reg‐ may underpin the beneficial effects of SY on the efficiency of cogni‐
ulating the relative engagement of default mode vs dorsal atten‐ tive control and attentional allocation, key processes for monitoring
tional networks—based on the salience of stimuli with respect to the moment‐to‐moment experience and constraining narrative self‐­
behavioral goals (Menon & Uddin, 2010)—accounts for previous reference. Importantly, both these brain changes appear to parallel
evidence relating this region, in conjunction with dACC, to self‐ the beneficial effects of SY training on perceived well‐being, as indi‐
regulation (Craig, 2009; Seeley et al., 2007) as well as initiation, cated by the significant correlations with the scores obtained at the
maintenance, and adjustment of attentional control (Dosenbach questionnaire of emotional self‐assessment after training.
et al., 2007; Dosenbach, Fair, Cohen, Schlaggar, & Petersen, 2008). While such evidence highlights the potential for brain plasticity
This concept fits with the demands posed by SY meditation, aim‐ beyond the developmental stages, however, the current spatial res‐
ing to obtain a state of mental silence in which the attention is olution of MRI data does not allow to distinguish different micro‐
focused on the present moment. Changes of intranetwork coher‐ structural processes which may underlie VBM results, for example,
ent activity in such a control network may support the attempt to dendritic arborization vs axon remodeling (Mietchen & Gaser, 2009).
resist narrative mind‐wandering (Mason et al., 2007), allowing to Previous literature suggests that functional and structural plasticity
maintain in working memory the awareness of the present mo‐ might depend on the balance between processes promoting molecu‐
ment and monitor one’s body state. lar flexibility (e.g., Hebbian plasticity) and stability (e.g., homeostatic
This hypothesis fits with neuroimaging evidence on other types plasticity) (Yin & Yuan, 2014), but the interaction between these
of short‐term meditation interventions, sharing with the SY medita‐ mechanisms, the subsequent large‐scale consequences on brain
tion an open monitoring approach aiming to develop the capacity for function and structure, and the behavioral response remain to be
mindfulness (i.e., awareness of present‐moment experiences with a clarified (Hamaide, De Groof, & Van der Linden, 2016).
compassionate, non judgmental, stance). Across different studies, A limitation of this study is represented by the presence of a pas‐
such interventions resulted in stronger intensity of activation (Holzel sive control group. However, this choice was driven by concerns re‐
et al., 2011) and efficiency (Xue, Tang, & Posner, 2011) in functional garding the cognitive and behavioral processes which could, or could
networks involving the dACC, possibly reflecting higher cognitive con‐ not, be controlled in any “active” control procedure. Moreover, the lack
trol and improved suppression of distracting events. Similar studies of significant differences between active and passive control groups in
highlighted a stronger recruitment of the dorsolateral prefrontal cor‐ previous studies (Chooi & Thompson, 2012; Colom et al., 2016) led us
tex alongside the right insula after training (Farb et al., 2007), which to adopt the most conservative procedure. Moreover, the use of an ad
DODICH et al. | 9 of 11

hoc questionnaire, rather than a validated one, might limit the robust‐ and the Centre for Evolutionary Learning in the management of
ness of the behavioral results. However, to the best of our knowledge, the meditation training. Finally we thank the GOLDEN Foundation
no formal measurements are currently available to assess the subjec‐ for the support.
tive experiences specifically related to SY meditation and future stud‐
ies could benefit from building upon the questionnaire here proposed.
C O N FL I C T O F I N T E R E S T
Another possible limitation is represented by the relatively small sam‐
ple size of the active group, partially due to dropout. Although prelim‐ None declared.
inary analyses confirmed the intergroup homogeneity concerning the
variables of interest (thus reducing the impact of potential sources of
ORCID
error), additional studies with larger sample size are needed to control
for the potential impact of intersubject variability. Finally, the present Alessandra Dodich http://orcid.org/0000-0001-9140-0699
evidence on meditation‐related effects on spectral power of intrinsic
intranetwork brain activity should be complemented by a system‐level
description of the associated changes in internetwork functional con‐ REFERENCES

nectivity. Such evidence will likely help refining an interpretation of the Adan, A., Serra‐Grabulosa, J. M., Caci, H., & Natale, V. (2009).
present evidence on spectral power, which still remains an open issue. A reduced Temperament and Character Inventory (TCI‐56).
Psychometric properties in a non‐clinical sample. Personality and
In conclusion, this is the first longitudinal randomized con‐
Individual. Differences, 46(7), 687–692. https://doi.org/10.1016/j.
trolled study examining a causal relationship between SY med‐ paid.2009.01.023
itation practice and cortical reorganization, as indexed by brain Aftanas, L. I., & Golocheikine, S. A. (2001). Human anterior and frontal
structure and spontaneous activity. Evidence from both brain midline theta and lower alpha reflect emotionally positive state and
internalized attention: High‐resolution EEG investigation of medita‐
morphometry and resting‐state activity consistently pointed to
tion. Neuroscience Letters, 310(1), 57–60. https://doi.org/10.1016/
the right inferior fronto‐insular sector of the executive control S0304-3940(01)02094-8
network as a crucial target of SY meditation, modulating brain ac‐ Aftanas, L. I., & Golocheikine, S. A. (2002). Non‐linear dynamic complex‐
tivity and structure in this key node of the networks underlying ity of the human EEG during meditation. Neuroscience Letters, 330(2),
attentional control. A significant increase of GM density in the 143–146. https://doi.org/10.1016/S0304-3940(02)00745-0
Aftanas, L. I., & Golosheikin, S. A. (2003). Changes in cortical activity
fronto‐insular cortex, a change in coherent activity in the whole
during altered state of consciousness: Study of meditation by high
network, as well as the correlations with well‐being measures, resolution EEG. Fiziologiia Cheloveka, 29(2), 18–27.
consistently suggest a causal link between SY practice and the Allen, E. A., Erhardt, E. B., Damaraju, E., Gruner, W., Segall, J. M.,
development of some of the core experiential correlates of SY Silva, R. F., & …Calhoun, V. D., (2011). A baseline for the multi‐
variate comparison of resting‐state networks. Frontiers in Systems
meditation, such as the individual capacity to access positive af‐
Neuroscience, 5, 2. https://doi.org/10.3389/fnsys.2011.00002
fect states and long‐term emotional well‐being, together with the Allen, E., Erhardt, E., Eichele, T., Mayer, A., & Calhoun, V. (2010).
capacity to consciously modulate internal and external attention Comparison of pre‐normalization methods on the accuracy and re‐
control mechanisms. These mechanisms may thus underlie the en‐ liability of group ICA results. Paper presented at the 16th Annual
Meeting of the Organization for Human Brain Mapping.
hanced ability to reach a state of thoughtless awareness, which can
Bell, A. J., & Sejnowski, T. J. (1995). An information‐maximization approach
be connected to reduced unnecessary mental activity and more to blind separation and blind deconvolution. Neural Computation,
efficient use of cognitive skills. Our results confirm and extend 7(6), 1129–1159. https://doi.org/10.1162/neco.1995.7.6.1129
previous cross‐sectional evidence on the structural (Hernandez et Buckner, R. L., Andrews‐Hanna, J. R., & Schacter, D. L. (2008). The brain's
default network: Anatomy, function, and relevance to disease. Annals
al., 2016) and functional (Hernandez et al., 2015) correlates of SY
of the New York Academy of Sciences, 1124, 1–38.
in expert meditators. Unlike previous studies, however, here we Calhoun, V. D., Adali, T., Pearlson, G. D., & Pekar, J. J. (2001a). A method
show a persistent effect of meditation outside the formal practice. for making group inferences from functional MRI data using inde‐
In line with previous evidence of significant differences between pendent component analysis. Human Brain Mapping, 14(3), 140–151.
Calhoun, V. D., Adali, T., Pearlson, G. D., & Pekar, J. J. (2001b). Spatial and
expert meditators and controls in the electrophysiological spec‐
temporal independent component analysis of functional MRI data
tral profile at rest (Lutz, Greischar, Rawlings, Ricard, & Davidson, containing a pair of task‐related waveforms. Human Brain Mapping,
2004), our results suggest a potential broad impact of meditative 13(1), 43–53.
practice on neural organization, in turn reflecting on other out‐ Caminiti, S. P., Canessa, N., Cerami, C., Dodich, A., Crespi, C., Iannaccone, S.,
comes related to health and well‐being (Muehsam et al., 2017). … Cappa, S. F. (2015). Affective mentalizing and brain activity at rest in
the behavioral variant of frontotemporal dementia. NeuroImage Clinical,
9, 484–497. https://doi.org/10.1016/j.nicl.2015.08.012
Chooi, W.‐T., & Thompson, L. A. (2012). Working memory training does
AC K N OW L E D G M E N T S
not improve intelligence in healthy young adults. Intelligence, 40(6),
We thank Professor Stefano Brusoni for his help in study plan‐ 531–542. https://doi.org/10.1016/j.intell.2012.07.004
Chung, S. C., Brooks, M. M., Rai, M., Balk, J. L., & Rai, S. (2012). Effect of
ning and Professor Katya Rubia for her comments on a previous
Sahaja yoga meditation on quality of life, anxiety, and blood pressure
version of the manuscript. We thank Raffaele Guido and Floriana control. Journal of Alternative and Complementary Medicine, 18(6),
Mulazzi for their help in data collection, and Michela Cavalletti 589–596.
10 of 11 | DODICH et al.

Colom, R., Hua, X., Martínez, K., Burgaleta, M., Román, F. J., Gunter, J. Hamaide, J., De Groof, G., & Van der Linden, A. (2016). Neuroplasticity
L., … Thompson, P. M. (2016). Brain structural changes following and MRI: A perfect match. NeuroImage, 131, 13–28. https://doi.
adaptive cognitive training assessed by Tensor‐Based Morphometry org/10.1016/j.neuroimage.2015.08.005
(TBM). Neuropsychologia, 91, 77–85. https://doi.org/10.1016/j. Harrison, L. J., Manocha, R., & Rubia, K. (2004). Sahaja yoga meditation
neuropsychologia.2016.07.034 as a family treatment programme for children with attention deficit‐
Corbetta, M., & Shulman, G. L. (2002). Control of goal‐directed and stim‐ hyperactivity disorder. Clinical Child Psychology and Psychiatry, 9(4),
ulus‐driven attention in the brain. Nature Reviews Neuroscience, 3(3), 479–497. https://doi.org/10.1177/1359104504046155
201–215. https://doi.org/10.1038/nrn755. Hernandez, S. E., Suero, J., Barros, A., Gonzalez‐Mora, J. L., & Rubia, K.
Cordes, D., Haughton, V. M., Arfanakis, K., Wendt, G. J., Turski, P. A., (2016). Increased Grey Matter Associated with Long‐Term Sahaja
Moritz, C. H., …Meyerand, M. E. (2000). Mapping functionally re‐ Yoga Meditation: A Voxel‐Based Morphometry Study. PLoS ONE,
lated regions of brain with functional connectivity MR imaging. AJNR 11(3), e0150757. https://doi.org/10.1371/journal.pone.0150757
American Journal of Neuroradiology, 21(9), 1636–1644. Hernandez, S. E., Suero, J., Rubia, K., & Gonzalez‐Mora, J. L. (2015).
Craig, A. D. (2009). How do you feel ‐ now? The anterior insula and human Monitoring the neural activity of the state of mental silence while
awareness. Nature Reviews Neuroscience, 10(1), 59–70. https://doi. practicing Sahaja yoga meditation. Journal of Alternative and
org/10.1038/nrn2555 Complementary Medicine, 21(3), 175–179. https://doi.org/10.1089/
Crawford, J. R., & Henry, J. D. (2004). The positive and negative af‐ acm.2013.0450.
fect schedule (PANAS): Construct validity, measurement prop‐ Himberg, J., Hyvarinen, A., & Esposito, F. (2004). Validating the indepen‐
erties and normative data in a large non‐clinical sample. British dent components of neuroimaging time series via clustering and visu‐
Journal of Clinical Psychology, 43(Pt 3), 245–265. https://doi. alization. NeuroImage, 22(3), 1214–1222. https://doi.org/10.1016/j.
org/10.1348/0144665031752934 neuroimage.2004.03.027
Dahl, C. J., Lutz, A., & Davidson, R. J. (2015). Reconstructing and de‐ Holzel, B. K., Carmody, J., Vangel, M., Congleton, C., Yerramsetti, S.
constructing the self: Cognitive mechanisms in meditation prac‐ M., Gard, T., & Lazar, S. W. (2011). Mindfulness practice leads to
tice. Trends in Cognitive Sciences, 19(9), 515–523. https://doi. increases in regional brain gray matter density. Psychiatry Research,
org/10.1016/j.tics.2015.07.001 191(1), 36–43. https://doi.org/10.1016/j.pscychresns.2010.08.006
Dosenbach, N. U., Fair, D. A., Cohen, A. L., Schlaggar, B. L., & Petersen, Jang, J. H., Jung, W. H., Kang, D. H., Byun, M. S., Kwon, S. J., Choi, C. H.,
S. E. (2008). A dual‐networks architecture of top‐down control. & Kwon, J. S. (2011). Increased default mode network connectivity
Trends in Cognitive Sciences, 12(3), 99–105. https://doi.org/10.1016/j. associated with meditation. Neuroscience Letters, 487(3), 358–362.
tics.2008.01.001 https://doi.org/10.1016/j.neulet.2010.10.056
Dosenbach, N. U., Fair, D. A., Miezin, F. M., Cohen, A. L., Wenger, K. K., Kilpatrick, L. A., Suyenobu, B. Y., Smith, S. R., Bueller, J. A., Goodman, T.,
Dosenbach, R. A., … Petersen, S. E. (2007). Distinct brain networks Creswell, J. D., … Naliboff, B. D. (2011). Impact of Mindfulness‐Based
for adaptive and stable task control in humans. Proceedings of the Stress Reduction training on intrinsic brain connectivity. NeuroImage,
National Academy of Sciences, 104(26), 11073–11078. https://doi. 56(1), 290–298. https://doi.org/10.1016/j.neuroimage.2011.02.034
org/10.1073/pnas.0704320104 Koch, W., Teipel, S., Mueller, S., Buerger, K., Bokde, A. L., Hampel,
Draganski, B., Gaser, C., Busch, V., Schuierer, G., Bogdahn, U., & May, A. H., … Meindl, T. (2010). Effects of aging on default mode net‐
(2004). Neuroplasticity: Changes in grey matter induced by training. work activity in resting state fMRI: Does the method of analysis
Nature, 427(6972), 311–312. matter? NeuroImage, 51(1), 280–287. https://doi.org/10.1016/j.
Eickhoff, S. B., Stephan, K. E., Mohlberg, H., Grefkes, C., Fink, G. R., neuroimage.2009.12.008
Amunts, K., & Zilles, K. (2005). A new SPM toolbox for combin‐ Lutz, A., Greischar, L. L., Rawlings, N. B., Ricard, M., & Davidson, R. J.
ing probabilistic cytoarchitectonic maps and functional imaging (2004). Long‐term meditators self‐induce high‐amplitude gamma
data. NeuroImage, 25(4), 1325–1335. https://doi.org/10.1016/j. synchrony during mental practice. Proceedings of the National
neuroimage.2004.12.034 Academy of Sciences, 101(46), 16369–16373. https://doi.org/10.1073/
Erhardt, E. B., Rachakonda, S., Bedrick, E. J., Allen, E. A., Adali, T., & pnas.0407401101.
Calhoun, V. D. (2011). Comparison of multi‐subject ICA methods Lutz, A., Slagter, H. A., Dunne, J. D., & Davidson, R. J. (2008). Attention
for analysis of fMRI data. Human Brain Mapping, 32(12), 2075–2095. regulation and monitoring in meditation. Trends in Cognitive Sciences,
https://doi.org/10.1002/hbm.21170. 12(4), 163–169. https://doi.org/10.1016/j.tics.2008.01.005
Farb, N. A., Segal, Z. V., Mayberg, H., Bean, J., McKeon, D., Fatima, Z., & Manocha, R., Black, D., Sarris, J., & Stough, C. (2011). A randomized, con‐
Anderson, A. K. (2007). Attending to the present: Mindfulness medi‐ trolled trial of meditation for work stress, anxiety and depressed mood
tation reveals distinct neural modes of self‐reference. Social Cognitive in full‐time workers. Evidence‐Based Complementary and Alternative
and Affective Neuroscience, 2(4), 313–322. https://doi.org/10.1093/ Medicine: Ecam, 2011, 960583. https://doi.org/10.1155/2011/960583.
scan/nsm030. Manocha, R., Marks, G. B., Kenchington, P., Peters, D., & Salome, C.
Fox, K. C., Nijeboer, S., Dixon, M. L., Floman, J. L., Ellamil, M., Rumak, M. (2002). Sahaja yoga in the management of moderate to severe
S. P., … Christoff, K. (2014). Is meditation associated with altered asthma: A randomised controlled trial. Thorax, 57(2), 110–115.
brain structure? A systematic review and meta‐analysis of mor‐ https://doi.org/10.1136/thorax.57.2.110
phometric neuroimaging in meditation practitioners. Neuroscience Manocha, R. (2014). Silence your mind: improve your happiness in just 10
and Biobehavioral Reviews, 43, 48–73. https://doi.org/10.1016/j. minutes a day with this new approach to meditation. London, UK: Orion
neubiorev.2014.03.016 Publishing Group Ltd UK.
Friston, K. J., Holmes, A. P., Poline, J. B., Grasby, P. J., Williams, S. C., Mason, M. F., Norton, M. I., Van Horn, J. D., Wegner, D. M., Grafton, S. T.,
Frackowiak, R. S., & Turner, R. (1995). Analysis of fMRI time‐se‐ & Macrae, C. N. (2007). Wandering minds: The default network and
ries revisited. NeuroImage, 2(1), 45–53. https://doi.org/10.1006/ stimulus‐independent thought. Science, 315(5810), 393–395.
nimg.1995.1007 Menon, V., & Uddin, L. Q. (2010). Saliency, switching, attention and con‐
Genovese, C. R., Lazar, N. A., & Nichols, T. (2002). Thresholding of trol: A network model of insula function. Brain Structure and Function,
statistical maps in functional neuroimaging using the false discov‐ 214(5–6), 655–667. https://doi.org/10.1007/s00429-010-0262-0.
ery rate. NeuroImage, 15(4), 870–878. https://doi.org/10.1006/ Mietchen, D., & Gaser, C. (2009). Computational morphometry for
nimg.2001.1037. detecting changes in brain structure due to development, aging,
DODICH et al. | 11 of 11

learning, disease and evolution. Frontiers in Neuroinformatics, 3, 25. Sridharan, D., Levitin, D. J., & Menon, V. (2008). A critical role for the
https://doi.org/10.3389/neuro.11.025.2009. right fronto‐insular cortex in switching between central‐executive
Muehsam, D., Lutgendorf, S., Mills, P. J., Rickhi, B., Chevalier, G., Bat, and default‐mode networks. Proceedings of the National Academy
N., … Gurfein, B. (2017). The embodied mind: A review on func‐ of Sciences, 105(34), 12569–12574. https://doi.org/10.1073/
tional genomic and neurological correlates of mind‐body therapies. pnas.0800005105
Neuroscience and Biobehavioral Reviews, 73, 165–181. https://doi. Sutherland, M. T., McHugh, M. J., Pariyadath, V., & Stein, E. A. (2012).
org/10.1016/j.neubiorev.2016.12.027. Resting state functional connectivity in addiction: Lessons learned
Niendam, T. A., Laird, A. R., Ray, K. L., Dean, Y. M., Glahn, D. C., & Carter, and a road ahead. NeuroImage, 62(4), 2281–2295. https://doi.
C. S. (2012). Meta‐analytic evidence for a superordinate cognitive org/10.1016/j.neuroimage.2012.01.117
control network subserving diverse executive functions. Cognitive, Taylor, V. A., Daneault, V., Grant, J., Scavone, G., Breton, E., Roffe‐Vidal,
Affective, and Behavioural Neuroscience, 12(2), 241–268. https://doi. S., … Beauregard, M. (2013). Impact of meditation training on the
org/10.3758/s13415-011-0083-5. default mode network during a restful state. Social Cognitive and
Oldfield, R. C. (1971). The assessment and analysis of handedness: The Affective Neuroscience, 8(1), 4–14. https://doi.org/10.1093/scan/
Edinburgh inventory. Neuropsychologia, 9(1), 97–113. https://doi. nsr087
org/10.1016/0028-3932(71)90067-4 Travis, F., & Shear, J. (2010). Focused attention, open monitoring
Panjwani, U., Selvamurthy, W., Singh, S. H., Gupta, H. L., Mukhopadhyay, and automatic self‐transcending: Categories to organize medita‐
S., & Thakur, L. (2000). Effect of Sahaja yoga meditation on auditory tions from Vedic. Buddhist and Chinese Traditions. Consciousness
evoked potentials (AEP) and visual contrast sensitivity (VCS) in epi‐ and Cognition, 19(4), 1110–1118. https://doi.org/10.1016/j.
leptics. Applied Psychophysiology and Biofeedback, 25(1), 1–12. concog.2010.01.007.
Reva, N. V., Pavlov, S. V., Loktev, K. V., Korenyok, V. V., & Aftanas, L. I. Van Dijk, K. R. A., Hedden, T., Venkataraman, A., Evans, K. C., Lazar, S.
(2014). Influence of long‐term Sahaja Yoga meditation practice on W., & Buckner, R. L. (2010). Intrinsic functional connectivity as a
emotional processing in the brain: An ERP study. Neuroscience, 281C, tool for human connectomics: theory, properties, and optimization.
195–201. https://doi.org/10.1016/j.neuroscience.2014.09.053 Journal of Neurophysiology, 103(1), 297–321. https://doi.org/10.1152/
Santarnecchi, E., D’Arista, S., Egiziano, E., Gardi, C., Petrosino, R., Vatti, jn.00783.2009.
G., … Rossi, A. (2014). Interaction between neuroanatomical and Villemure, C., Ceko, M., Cotton, V. A., & Bushnell, M. C. (2014). Insular
psychological changes after mindfulness‐based training. PLoS ONE, cortex mediates increased pain tolerance in yoga practitioners.
9(10), e108359. https://doi.org/10.1371/journal.pone.0108359. Cerebral Cortex, 24(10), 2732–2740. https://doi.org/10.1093/cercor/
Sedlmeier, P., Eberth, J., Schwarz, M., Zimmermann, D., Haarig, F., Jaeger, bht124
S., & Kunze, S. (2012). The psychological effects of meditation: A Vincent, J. L., Kahn, I., Snyder, A. Z., Raichle, M. E., & Buckner, R. L.
meta‐analysis. Psychological Bulletin, 138(6), 1139–1171. https://doi. (2008). Evidence for a frontoparietal control system revealed by
org/10.1037/a0028168 intrinsic functional connectivity. Journal of Neurophysiology, 100(6),
Seeley, W. W., Menon, V., Schatzberg, A. F., Keller, J., Glover, G. H., 3328–3342. https://doi.org/10.1152/jn.90355.2008
Kenna, H., … Greicius, M. D. (2007). Dissociable intrinsic connec‐ Xue, S., Tang, Y. Y., & Posner, M. I. (2011). Short‐term meditation
tivity networks for salience processing and executive control. increases network efficiency of the anterior cingulate cor‐
Journal of Neuroscience, 27(9), 2349–2356. https://doi.org/10.1523/ tex. NeuroReport, 22(12), 570–574. https://doi.org/10.1097/
JNEUROSCI.5587-06.2007 WNR.0b013e328348c750.
Smith, S. M., Fox, P. T., Miller, K. l., Glahn, D. C., Fox, P. M., Mackay, C. Yin, J., & Yuan, Q. (2014). Structural homeostasis in the nervous system:
E., … Beckmann, C. F. (2009). Correspondence of the brain's func‐ A balancing act for wiring plasticity and stability. Frontiers in Cellular
tional architecture during activation and rest. Proceedings of the Neuroscience, 8, 439. https://doi.org/10.3389/fncel.2014.00439.
National Academy of Sciences, 106(31), 13040–13045. https://doi.
org/10.1073/pnas.0905267106
Spreng, R. N., Sepulcre, J., Turner, G. R., Stevens, W. D., & Schacter, D.
How to cite this article: Dodich A, Zollo M, Crespi C, et al.
L. (2013). Intrinsic architecture underlying the relations among the
Short‐term Sahaja Yoga meditation training modulates brain
default, dorsal attention, and frontoparietal control networks of the
human brain. Journal of Cognitive Neuroscience, 25(1), 74–86. https:// structure and spontaneous activity in the executive control
doi.org/10.1162/jocn_a_00281. network. Brain Behav. 2018;e01159. https://doi.
Spreng, R. N., Stevens, W. D., Chamberlain, J. P., Gilmore, A. W., & org/10.1002/brb3.1159
Schacter, D. L. (2010). Default network activity, coupled with
the frontoparietal control network, supports goal‐directed cog‐
nition. NeuroImage, 53(1), 303–317. https://doi.org/10.1016/j.
neuroimage.2010.06.016