it
b,c,
gele
, US
a r t i c l e i n f o
Psychiatry Research: Neuroimaging 191 (2011) 60–67
Contents lists available at ScienceDirect
Psychiatry Researc
ev1. Introduction
We previously described a model for the neurophysiologic
activation during the complex neurocognitive task of meditation
(Newberg and Iversen, 2003). This model is based, in part, on studies
that have correlated brain function with specific behaviors, tasks,
and experience. From this model, we have constructed the main
hypotheses to be tested in this study. Specifically, this involves
changes in activity in a number of brain structures that may be
associated with specific elements of meditation practices.
It has been suggested that the prefrontal cortex and anterior
cingulate gyrus are activated during meditation since an important
element of meditation practice involves attentional focus (Newberg
several investigators have shown activation of the prefrontal cortex in
subjects performing purposely willed tasks, or tasks that required
sustained attention (Frith et al., 1991; Pardo et al., 1991). Further-
more, focused attention on a mathematical task resulted in increased
regional cerebral blood flow (rCBF) in the prefrontal cortex (Sammer
et al., 2007).
Imaging studies of meditation performed by our group also
suggest that the prefrontal cortex and anterior cingulate gyrus are
activated during meditation. We observed increased rCBF (or
metabolism) in the prefrontal cortex and anterior cingulate gyrus
during several different types of meditation practices. These practices
included Tibetan Buddhist meditation, Iyengar Yoga meditation, and
Centering Prayer (Newberg et al., 2001; Newberg et al., 2003;and Iversen, 2003; Cahn and Polich, 2006)
shown that the prefrontal cortex and anteri
are activated during attention focusing task
1990; Vogt et al., 1992). Using positron emis
⁎ Corresponding author. 110 Donner Building, Uni
Spruce Street, Philadelphia, PA, 19104, USA. Tel.:+1 215 6
E-mail address: andrew.newberg@uphs.upenn.edu (
0925-4927/$ – see front matter © 2010 Elsevier Ireland
doi:10.1016/j.pscychresns.2010.09.011different meditation practices in the same individuals and that these changes correlated with the subjective
experiences of the practitioners.
© 2010 Elsevier Ireland Ltd. All rights reserved.Article history:
Received 27 November 2009
Received in revised form 23 September 2010
Accepted 25 September 2010
Keywords:
Meditation
fMRI
Cerebral blood flow
Frontal lobe
Stress
Arterial spin labeling (ASL)a b s t r a c t
Our goal in this study was to advance the understanding of the neural pathways of meditation by addressing
the cerebral blood flow (CBF) responses associated with two different meditation practices performed by
the same individuals and how such changes related to the “stress” circuits in the brain. Ten experienced
meditators performed two types of meditation, a “focused-based” practice and a “breath-based” practice.
Subjects were scanned using perfusion functional magnetic resonance imaging (fMRI) during a baseline state,
both meditation states, and a post meditation baseline state. Using general linear model, we found that the
frontal regions, anterior cingulate, limbic system and parietal lobes were affected during meditation and that
there were different patterns of CBF between the two meditation states. We observed strong correlations
between depth of meditation and neural activity in the left inferior forebrain areas including the insula,
inferior frontal cortex, and temporal pole. There were persistent changes in the left anterior insula and the
precentral gyrus even after meditation was stopped. This study revealed changes in the brain during two. Studies have generally
or cingulate gyrus areas
s (Posner and Petersen,
sion tomography (PET),
Cohen et al., 20
fluorodeoxygluc
subjects undergo
in the frontal/o
However, the sub
the glucosemeta
study utilizing
subjects perform
versity of Pennsylvania, 3400
62 3092; fax:+1 215 349 5843.
A.B. Newberg).
Ltd. All rights reserved.gMyrna Brind Center of Integrative Medicine, Thomas Jefferson University, Philadelphia, PACerebral blood flow changes associated w
perceived depth of meditation
Danny J.J. Wanga, Hengyi Raob,c, Marc Korczykowski
Dharma Singh Khalsa f, Andrew B. Newberg g,⁎
aAlmanson–Lovelace Brain Mapping Center, University of California at Los Angeles, Los An
bDepartment of Neurology, University of Pennsylvania, Philadelphia, PA, USA
cCenter for Functional Neuroimaging, University of Pennsylvania, Philadelphia, PA, USA
dDepartment of Radiology, University of Pennsylvania, Philadelphia, PA, USA
eDepartment of Psychology, University of Pennsylvania, Philadelphia, PA, USA
fAlzheimer’s Research and Prevention Foundation, Tucson, AZ, USA
j ourna l homepage: www.e lsh different meditation practices and
Nancy Wintering d, John Plutab,e,
s, CA, USA
A
h: Neuroimaging
i e r.com/ locate /psychresns09). Herzog et al. (1990–1991) utilized PET with
ose (FDG) to measure glucose metabolism in eight
ing Yoga Meditative Relaxation. A significant increase
ccipital ratio of glucose metabolism was observed.
jects did not experience a strongmeditative state, and
bolismwas not measured in specific cortical regions. A
functional magnetic resonance imaging (fMRI) of
ing a similar yoga relaxation technique designed to

61D.J.J. Wang et al. / Psychiatry Research: Neuroimaging 191 (2011) 60–67bring about “relaxation response,” demonstrated increased CBF in the
frontal and limbic regions (Lazar et al., 2000).
In addition to frontal lobe activity, we have previously hypothe-
sized and found support for altered activity in the limbic structures
(Lazar et al., 2000; Cohen et al., 2009) which may be related to the
emotional effects of meditation tasks (Brefczynski-Lewis et al., 2007;
Lutz et al., 2008a,b; Goldin and Gross, 2010). We also hypothesized
that there would be decreased activity in the parietal lobes during
meditation. This is based in part upon several reports, including our
own, that have demonstrated decreased activity in the parietal lobes
during meditation practices (Newberg et al., 2001; Newberg and
Iversen, 2003). It is also known that the parietal regions are associated
with spatial processing (Silver et al., 2005; Zaehle et al., 2006), and
since alterations in the sense of self and spatial orientation are
subjectively changed during meditation practices, changes in the
parietal lobes might be expected (Tagini and Raffone, 2010).
In spite of the growingnumber of studies onmeditation, therewere
several important questions that have yet to be answered. For
instance, it remains unclear whether a common brain activation
pattern exists across variousmeditation tasks anddifferentmeditation
stages. On the other hand, results from existing studies may
be attributable to broader effects of attentional control rather than
being specifically related to the effects of meditation on the brain
(Brefczynski-Lewis et al., 2007; Lutz et al., 2008a,b; Raffone and
Srinivasan, 2010). In the present study, we employed a perfusion
based fMRI technique to continuously image the cerebral blood flow
(CBF) responses before, during and following two types of meditation
practice performed by experiencedmeditators. This design capitalized
on the capability of perfusion fMRI for noninvasive, quantitative and
repeatable CBFmeasurements (Aguirre et al., 2002;Wang et al., 2003).
Using this technique, we have previously demonstrated the signature
of the brain's response to psychological stress along with its enduing
effects which involved asymmetric prefrontal activity and increased
CBF in several limbic structures (Wang et al., 2005a, 2007). It would be
interesting to evaluate the meditation effect on CBF changes in the
brain's “stress network”, given the stress relieving effect of meditation
practice. The stress network more specifically refers to asymmetric
prefrontal activation, the anterior cingulate cortex (ACC) and several
limbic areas including amygdala, hippocampus and insula based on
both our past studies on stress and other studies in the literature. In
addition, studies have previously shown that meditation practices can
regulate emotions associated with altered activation in the limbic
structures (Lutz et al., 2008a,b).
Our primary goal in this study was therefore to make several
significant advancements in the understanding of neural pathways
mediating the meditation effect by addressing the “general” and
“specific” CBF responses associated with meditation. To that end, we
attempted to determine if 1) there were different CBF changes
associated with different practices performed by the same individual;
2) there were correlations between CBF and the subjective depth of
the meditation practices; 3) there were opposite effects of meditation
and stress on CBF variations in specific brain regions, and 4) there
were persistent effects on the brain after the meditation. We
hypothesized that there would indeed be specific CBF changes
associated with different practices when performed by the same
individuals. We primarily expected increased activity in the frontal
regions and limbic structures, but that the pattern would be distinct.
This is a significant advancement since no studies to date have
assessed the same individuals doing different types of meditation. We
also hypothesized that the subjective experience during the medita-
tion practices associated with depth of practice would correlate with
CBF changes. Additionally, we hypothesized that the stress relieving
effect of meditation was mediated through an altered CBF response
representing enhancement of positive emotions and/or suppression
of negative emotions and vigilance. Finally, we predicted that there
would be a “washout” period after meditation is completed so thatthere would be persistent changes in CBF after meditation. No studies
have evaluated the after effects of meditation.
2. Materials and methods
2.1. Subjects
Ten healthy subjects (four females and six males, mean
age=53.7 years, range 43–62) who were experienced meditators
participated in this study. They each performed meditation from the
Kundalini yoga tradition for over 30 years (approximately 20,000 h)
under the guidance of Yogi Bhajan. The particular meditation studied
consisted of two basic types, both from the same tradition. The first
was a “focused-based” practice called, Kirtan Kriya, in which subjects
repeated several phrases (SA, TA, NA, and MA) while touching their
thumb and fingers in sequence (Meditation 1). It is highly repetitive
and is reported by the practitioners to enhance their attention
capabilities, reduce stress, and increase awareness. The second
meditation called, Shabad Kriya, was a “breath-based” practice in
which the practitioner takes several deep breaths in and then repeats
several phrases (SA, TA, NA, and MA) with each exhale. The goal
of this second meditation is to induce a state of deep relaxation
through the specified breathing exercises and internal repetition of
the phrases (Meditation 2). This meditation is described by the
practitioner to result in feelings of increased awareness and reduced
stress. The primary difference is the first involves intense focusing of
the mind and the second a relaxing of the mind. For this reason, we
maintained the same order of the meditation practices. The first one
creates an initial state of meditation while the second results in a
consolidation and deepening of the practice.
All the subjects were screened for neurologic and psychiatric
conditions.Written informed consentwas obtained prior to all human
studies according to an Institutional Review Board approval from the
University of Pennsylvania.
2.2. MRI experimental procedure
The experimental protocol consisted of five perfusion fMRI scans
performed with a fixed order: Baseline 1, Control task, Meditation 1,
Meditation 2 and Baseline 2. Such design allowed us to simultaneously
study the CBF changes during meditation (Meditation vs. Control),
enduring CBF responses following meditation (Baseline 2 vs. 1), and
the effect of two different styles of meditation practice (Meditation 2
vs. 1). In our experiment, the control task always preceded the two
meditation conditions to avoid potential “carry over” effects caused by
persistent CBF responses following meditation. During the first and
second meditation scans, subjects were instructed to perform the
focus-based and breath-based meditation practices respectively. As
the control condition, subjects counted to four while touching their
thumb and fingers (but not in sequence). Thus, the control condition
was similar to the meditation sessions in that both required repeating
a phrase (in this case numbers) and moving their fingers, but the
control condition did not include the specific phrases or finger
movements of the active meditation. Self reports of stress, depth of
meditation, and the subjective “feeling of connectedness” levels (on
the scale of 1 to 10) were collected between the five perfusion fMRI
scans (total four samples). The feeling of connectedness was chosen
since it is an experience that many meditators report and we have
previously postulated that it might be associated with decreased
activity in the parietal lobes. Because these concepts are such
subjective measures, we asked the subjects to rate “5” as their usual
level of stress, their usual depth of meditation, and their usual level of
connectedness during their usual meditation. However, this element
of the meditation experience has not been previously evaluated in a
brain imaging study.