Meditation effects on cognitive function and cerebral blood flow in subjects with memory loss: a preliminary study.

Abstract

This preliminary study determined if subjects with memory loss problems demonstrate changes in memory and cerebral blood flow (CBF) after a simple 8-week meditation program. Fourteen subjects with memory problems had an IV inserted and were injected with 250 MBq of Tc-99m ECD while listening to a neutral stimulus CD. They then underwent a pre-program baseline SPECT scan. Then subjects were guided through their first meditation session with a CD, during which they received an injection of 925 MBq ECD, and underwent a pre-program meditation scan. Subjects completed an 8-week meditation program and underwent the same scanning protocol resulting in a post-program baseline and meditation scan. A region of interest (ROI) template obtained counts in each ROI normalized to whole brain to provide a CBF ratio. Baseline and meditation scans and neuropsychological testing were compared before and after the program. The meditation program resulted in significant increases (p< 0.05) in baseline CBF ratios in the prefrontal, superior frontal, and superior parietal cortices. Scores on neuropsychological tests of verbal fluency, Trails B, and logical memory showed improvements after training. This preliminary study evaluated whether an 8-week meditation program resulted in improvements in neuropsychological function and differences in CBF in subjects with memory loss. While the findings are encouraging, there are a number of limitations that can be addressed in future studies with more participants and more detailed analyses.

Full text

Page 1

Journal of Alzheimer’s Disease 20 (2010) 517–526
DOI 10.3233/JAD-2010-1391
IOS Press
517
Meditation Effects on Cognitive Function and
Cerebral Blood Flow In Subjects with
Memory Loss: A Preliminary Study
Andrew B. Newberga,b,∗, Nancy Winteringa,b, Dharma S. Khalsab,c, Hannah Roggenkampaand
Mark R. Waldmanb
aDivision of Nuclear Medicine, Department of Radiology, University of Pennsylvania, Philadelphia, PA, USA
bCenter for Spirituality and the Mind, University of Pennsylvania, Philadelphia, PA, USA
cAlzheimer Research and Prevention Foundation, Tucson, AZ, USA
Accepted 12 January 2010
Abstract.Thispreliminary study determined ifsubjects withmemory lossproblems demonstrate changes inmemory and cerebral
blood flow (CBF) after a simple 8-week meditation program. Fourteen subjects with memory problems had an IV inserted
and were injected with 250MBq of Tc-99m ECD while listening to a neutral stimulus CD. They then underwent a pre-program
baseline SPECT scan. Then subjects were guided through their first meditation session with a CD, during which they received an
injection of 925MBq ECD, and underwent a pre-program meditation scan. Subjects completed an 8-week meditation program
and underwent the same scanning protocol resulting in a post-program baseline and meditation scan. A region of interest (ROI)
template obtained counts in each ROI normalized to whole brain to provide a CBF ratio. Baseline and meditation scans and
neuropsychological testing were compared before and afterthe program. The meditation program resulted in significant increases
(p < 0.05) in baseline CBF ratios in the prefrontal, superior frontal, and superior parietal cortices. Scores on neuropsychological
tests of verbal fluency, Trails B, and logical memory showed improvements after training. This preliminary study evaluated
whether an 8-week meditation program resulted in improvements in neuropsychological function and differences in CBF in
subjects with memory loss. While the findings are encouraging, there are a number of limitations that can be addressed in future
studies with more participants and more detailed analyses.
Keywords: Cerebral blood flow, cognitive impairment, meditation, memory, single photon emission computed tomography
INTRODUCTION
The number of older Americans continues to grow
and with it the number who are thought to suffer from
cognitive impairment and Alzheimer’s disease (AD).
Amongpeopleaged65,2–3%showsignsofAD,while
25–50% of people aged 85 have symptoms of AD [1].
∗Correspondence to: Andrew B. Newberg, M.D., Division of
Nuclear Medicine, Hospital of the University of Pennsylvania, 110
Donner Building, 3400 Spruce Street, Philadelphia, PA19104, USA.
Tel.: +1 215 662 3092; Fax: +1 215 349 5843; E-mail: Andrew.
newberg@uphs.upenn.edu.
An even greater number have some of the patholog-
ical hallmarks of the disease without the characteris-
tic symptoms. Until now, there have been few treat-
ment options for patients with early cognitive impair-
ment. Several medications and vaccine trials are un-
derway. However, a non-pharmacological approach
without side effects and without interfering with med-
ications would be very useful and cost effective in the
management of such patients without interfering with
anymedicalinterventions. Initialstudieshavesuggest-
ed that specific cognitive practice programs will help
improvememory[2,3]. Meditation has long been tout-
ed as a potential technique for improving memory and
lowering levels of stress, depression, and anxiety. This
ISSN 1387-2877/10/$27.50  2010 – IOS Press and the authors. All rights reserved

Page 2

518
A.B. Newberg et al. / Meditation Enhances Cognitive Function
study was designedto explorethe effects of a very spe-
cific form of meditation in patients with cognitive im-
pairmentandto trackpotentialchangeswithfunctional
brain imaging using single photon emission computed
tomography(SPECT).
Meditation, in general, is a complex neurocognitive
task that is often associated with alterations in brain
physiology and neuropsychological measures. Over
the past 30 years, there have been a number of studies
whichhaveexploredthephysiologicalcorrelatesofdif-
ferent types of meditation. It is important to note here
that meditationrefersto a largevarietyofpracticesthat
range from purely relaxation based techniques to those
performed with the goal of attaining intense spiritual
experiences. This variation, in itself, makes the study
of such practices difficult. However, we have tried to
find similarities among these practices, and since sub-
stantial prior studies have demonstrated improvements
in a variety of cognitive functions with these practices,
it is worthwhile to continue to explore them.
This is the first study to investigate potential im-
provements in cognition in subjects with actual mem-
ory loss.
Studies utilizing positron emission tomography
(PET), SPECT, and functional magnetic resonance
imaging(fMRI)havealldemonstratedspecificchanges
in cortical and sub-cortical structures when subjects
were actively meditating [4–7]. There is a growing,
although still relatively limited, number of studies that
haveevaluatedthelong-termeffectsofmeditationprac-
tices. Specifically, these studies showed increased ac-
tivity in expert meditators in the frontoparietalregions,
cerebellar, temporal, parahippocampal, and posterior
occipital cortex during meditation. These areas have
also been implicated in a variety of memory tasks uti-
lizing fMRI [8]. Other studies have compared experi-
encedmeditatorstonon-meditatorswithregardtobrain
structure and function [9]. For example, a study by
LazarandcolleaguesusingstructuralMRIdemonstrat-
ed that long-term meditators, practicing for approxi-
mately1houraday,hadthickerprefrontalcortexesthan
non-meditators[10]. Additionally,Lutz andcollabora-
tors using electroencephalography (EEG) showed that
expert meditators were able to induce changes in their
brain activity during meditation [11]. However, these
studies did not determineif the findings were the result
of the meditation practices since these studies were not
longitudinal. Thus, one possibility is that the individu-
als had brains that were fundamentallydifferent which
predisposed them to such practices. A second possi-
bility is that the individual affected their brain over the
course of their practice. Fewer studies have actually
tried to evaluate the longitudinal effects of meditation
over time. Davidson et al. showed that there were sig-
nificant changes over time in the brain’s EEG when in-
dividuals engaged in a daily 1 hour mindfulness based
meditation practice [12]. Moreover, no previous stud-
ieshaveinvestigatedtheeffectofmeditationinpatients
with actual cognitivedecline or AD. From the perspec-
tiveofimprovingmemoryusingmeditation,itisimper-
ative to longitudinally study subjects before and after
a meditation training program to determine if there are
long term effects of such a program.
Therefore, the primary purpose of this study was
to investigate a particular type of meditation practice
called Kirtan Kriya (KK), in patients who presented
with memory problems, before and after an 8-week
program. KK meditation is a simple technique that in-
volves the repetition of four sounds – SA TA NA MA.
While the person vocalizes these sounds, they sequen-
tially touch their thumb to their index finger, middle
finger, fourth finger, and then fifth finger. This is per-
formed out loud for 2 minutes, in a whisper for 2 min-
utes, in silence for 4 minutes, followed by in a whisper
for 2 more minutes and finally out loud for the final
2 minutes. The total time is 12 minutes. Since this is
a simple and quick practice, it has the potential to be
a very practical and low cost measure to help improve
memory. This also distinguishes this practice from a
number of other meditation practices that require ex-
tendedclasssessionsandlongmeditationpracticesthat
may not be practical in an older population. This was
the purpose for the present study which was to evalu-
ate the effects of performing daily KK meditation for
8 weeks on brain function and cognition and measure
changes in cerebral blood flow (CBF) utilizing SPECT
imaging.
We hypothesized that several structures would be
particularly affected by the KK meditation program.
Specifically, attention focusing practices such as med-
itation have activated the attentional network in the
brain which includes frontal lobe structures as well as
the anteriorcingulatecortex[13,14]. We also hypothe-
sizedthatstructuressuchastheamygdalaandthalamus
would be affected since these structures have shown
changesinotherstudiesofthelongtermeffectsofmed-
itation and are also part of the network of structures
involved in the default network [15,16]. We have pre-
viously found changes in the temporal lobe associat-
ed with verbal meditation practices and changes in the
parietallobeassociatedwith alteredspatial perceptions
during meditation [6].

Page 3

A.B. Newberg et al. / Meditation Enhances Cognitive Function
519
Fig. 1. This figure shows transaxial slices of SPECT scans (with CBF represented as red > yellow > green > blue) in the pre-program baseline
state and the post-program baseline state. The post-program baseline state shows that there is relatively increased activity in the right prefrontal
cortex (thick arrows) and anterior cingulate cortex (thin arrows) after the training program.
METHODS
Subjects and imaging acquisition
Fifteen subjects were recruited from local neurolo-
gy clinics, local medical groups, and hospital based
advertisements, who presented complaining of memo-
ry problems ranging from mild age-associated memo-
ry impairment (n = 7), to mild cognitive impairment
(MCI) (n = 5), to moderateimpairmentwith a diagno-
sis of AD (n = 3). The Mini-Mental Status Examina-
tion scores (MMSE) rangedfrom 16–30. There were 6
men and 9 women with ages rangingfrom 52–77years
with a meanageof 64± 8 years. The patientswith AD
were diagnosed on the basis of criteria established by
theNationalInstituteofNeurologicalandCommunica-
tiveDisordersandStroke/Alzheimer’sDiseaseandRe-
lated Disorders Association (NINCDS-ADRDA) cri-
teria [17]. The MCI patients were diagnosed based
upon criteria reported in Grundman et al. [18] which
includes memory complaint, abnormal memory based
upon performance on the Logical Memory II subtest
of the Wechsler Memory Scale Revised, normal gen-
eral cognitive function, no impairment in activities of
daily living, and not sufficiently impaired to meet the
NINCDS/ADRDAcriteria forAD. Theremainingsub-
jects were considered to have age-associated memory
impairment and were otherwise normal controls who
perceived their memory to be impaired. Each subject
had no significant experience with meditation or yoga.
SubjectswerestudiedontheirfirstKKtrainingdayand
then again after an 8-week self-directed training pro-
gram. We excludedthedatafromthesingleAD patient
with a MMSE of 16 due to her inability to adequately
perform the meditation.
Onthefirstdayofthestudy,afterobtaininginformed
consent (approved by the human subjects Institutional
Review Board with the study protocol),a room was set
upinthehospitaltofunctionasameditationroom. Ap-
proximately 20 minutes prior to the baseline scan, an
intravenous canula (IV) was placed in one arm so that
all injections could be performed without touching or
disturbing the subject. The subjects reported minimal

Page 4

520
A.B. Newberg et al. / Meditation Enhances Cognitive Function
discomfort from the IV that resolved prior to initiating
the remainder of the study. For the baseline scan, the
subject was instructed to rest in the room with their
eyes closed and listen to a general informational CD
about the effects of meditation practices for approxi-
mately 12 minutes. This CD was neutral in its con-
tent. However, we would suggest that having content
on the CD that discusses meditation rather than a com-
pletely neutral topic is appropriate to exclude any pos-
sible expectation effect that might have resulted from
both conditions. The subject was injected through the
IV with 250 MBq of99mTc-Bicisate (Bristol-Myers
Squibb Medical Imaging, N. Billerica, MA), prepared
as specified by the manufacturer. The subject contin-
ued to listen to the CD for another 5 minutes while
the tracer uptake occurred in the brain. Approximately
15 minutes following the injection, the subjects under-
went SPECT scanning for 30 minutes. This scan was
labeled the “pre-programbaseline” scan.
Projection images were obtained at three-degree
angle intervals on a 128 × 128 matrix (pixel size
3.56 mm × 3.56 mm) over 360◦by rotating each head
120◦. These SPECT images were reconstructed in the
transaxial, coronal, and sagittal planes using filtered
backprojection, followed by a low pass filter and 1st
order Chang attenuation correction (attenuation coeffi-
cient0.11cm−1. Thereconstructedslicethicknesswas
4 mm with a spatial resolution of 8–10 mm.
Following this pre-program baseline scan, the sub-
ject returned to the room for their first meditation ses-
sion. Subjects initially viewed a 10 minute video with
oneoftheinvestigators(DSK)showinghowtoperform
the Kirtan Kriya meditation. This video reviewed the
phrases, the sounds, and demonstrated the manner of
performingthemeditation. Itwas explainedtosubjects
to focuson the soundsandfinger movements. Subjects
were not asked to do anything more than perform the
task. Thus, there were no additional instructions re-
garding the state of mind that they should be in, any
preparatoryexercises,oranymindfulnessexercises. At
theendofthevideo,theprincipalinvestigatoranswered
any questions and then observed the subjects doing
the meditation to make sure that it was done correct-
ly. Subjects were instructed that they would perform
the meditation while listening to a meditation CD that
guided them through the entire practice. The CD con-
tains an individual performing the meditation practice
in its intended manner with some backgroundmusic to
aid in the rhythm of the meditation. The subjects were
then asked to perform the meditation for 12 minutes
the first time duringwhichtheywouldreceivea second
injectionofapproximately925MBqof99mTc-Bicisate
though the IV while he/she continued to meditate for
approximately another five minutes. The subject was
then scanned for 30 minutes using the same imaging
parameters as for the baseline study. This scan was
labeled the “pre-program meditation” scan.
Subjects were discharged home with the medita-
tion CD so that they could practice it at home. They
were instructed to perform the practice every day for
8 weeks. Subjects completed a log to record when
they performed the meditation practice and their sub-
jective experience of the practice and its effects. We
contacted them at 4 weeks to remind them to continue
practicing daily and to enquire as to their performance
of the meditation. We also directly interviewed sub-
jects upon completion of the program to review their
progress. Upon completion of the 8-week meditation
training program, subjects returned to the University
of Pennsylvania Nuclear Medicine Department to un-
dergo a second imaging day essentially identical to the
first. They received a “post-program baseline” scan in
which they were injected with99mTc-Bicisate while
listening again to an informational CD. After the base-
line scan, the subjects then performed the meditation
for the final time during which they were injected with
99mTc-Bicisate and underwent a “post-program med-
itation” scan. We maintained the same order in the
pre- and post-program imaging studies so that the ef-
fect of doing the meditation would not interfere with
the baseline scans.
Althoughthiswasanopenlabelstudytoassesseffect
as well as feasibility, we also recruiteda small compar-
ison group in which the KK meditation was replaced
with a “music listening” task. Five subjects, two hav-
ing MCI and three having age-associated memory im-
pairment(all women with a mean age of 65 ± 10 years
anda rangefrom56to 79andmeanMMSE of29± 1),
were asked to simply listen to two Mozart violin con-
certoseachdayforapproximately12minutes,thesame
amount of time required for the KK meditation. The
subjects were asked to focus their attention on the mu-
sic and to record their progress in a log book. Subjects
underwentthe same SPECT imagingproceduresas the
KK group with listening to the music replacing listen-
ing and performing KK meditation. The music group
listened to the same neutral content CD to produce the
same comparison state. A group listening to music for
the same amount of time might provide an adequate
comparison for the KK meditation program since sub-
jects wouldundergosimilar typesof programswith the
exceptionofnotdoingtheactivepartofthemeditation.

Page 5

A.B. Newberg et al. / Meditation Enhances Cognitive Function
521
Table 1
Baseline characteristics of the KK and Music comparison groups
Baseline characteristics
Age
MMSE
Category Fluency (Animals)
Trails A
Trails B
Digit Symbols
Logical Memory Delayed
POMS
KK group
64.0 ± 8.0
28.1 ± 0.7
21.1 ± 7.9
30.5 ± 12.2
105.5 ± 52.8
63.7 ± 25.3
10.6 ± 5.2
52.2 ± 12.9
Music group
65.0 ± 9.9
29.0 ± 1.0
21.5 ± 5.0
37.0 ± 11.7
132.5 ± 58.6
67.6 ± 21.7
12.3 ± 6.5
47.5 ± 17.2
p
N.S.
N.S.
N.S.
N.S.
N.S.
N.S.
N.S.
N.S.
Since the music group received an intervention (i.e.,
the music), we felt that the results would help reduce
effects related to placeboand practice effects related to
repeated performance of the neuropsychological tests.
Furthermore, since some studies have suggested that
music might also improve cognition and mood, this
comparison group might actually be more stringent in
terms of helping to observe an effect specifically from
the KK meditation.
Subjects in both groups were also evaluated on the
first imaging day with a brief neuropsychological test
battery (see Table 1) that was adapted from the bat-
tery currently used by our Memory Disorders Clinic
at the University of Pennsylvania and comprised of
a Category Fluency task in which subjects named as
many animals as possible in a 60 second time period,
the Wechsler Adult Intelligence Scale (WAIS) Digit
Symbol Substitution Test, a Logical Memory task, and
Trails A and B. These tests also were selected based
upon other studies in which neuropsychological tests
were used to evaluate changes in cognition associated
with mental task interventions [19,20]. These same
tests were repeated on the 8-week follow up session.
Image analysis and statistics
The images of the pre- and post-program baseline
and meditation scans were reconstructed and resliced,
using an oblique reformatting program, according to
the anterior-posterior commissure line so that the fi-
nal two sets were aligned for analysis. A previously
validated template methodology consisting of regions
of interest (ROI) corresponding to the major cortical
and subcortical structures was placed over the baseline
scan [21]. For the purposes of this study, we exam-
ined the CBF as measured in only a selected number
of ROIs which was hypothesis driven. The ROIs ex-
amined included the inferior frontal, superior frontal,
dorsolateralprefrontal,orbitofrontal,inferiortemporal,
superior temporal, inferior parietal, superior parietal,
and sensorimotor areas, as well as the precuneus, tha-
lamus, amygdala, and cingulate gyrus since these are
areas that have been found to be previously affected
during meditation tasks and also because these struc-
tures subserve a number of cognitive processes. The
locationforeachROI was determinedbased uponMRI
anatomy such that they could then be placed directly
on functional SPECT scans [20]. Furthermore, each
ROI fits within each specified region which helps to
ensure proper placement and to avoid problems with
partial voluming. The ROIs were placed on the ini-
tial scan and then copied directly onto all subsequent
scans. This was possible because the images were al-
readyresliced into the same planes as describedabove.
The count values for the baseline and meditation scans
were obtained by determining the number of counts
in each ROI on the meditation scan and normalizing
those counts to the whole brain activity. This provides
a CBF ratio for each ROI comparedto the whole brain.
Since two SPECT scans were performed on the same
day, the second scan had the decay corrected counts
from the first scan subtracted out prior to analysis. We
have previously validated this technique and show that
there is a high test-retest correlation with less than 6%
variability [22].
A percentage change between the meditation and
baseline scans (for both the pre- and post-programses-
sions) was calculated using the equation:
%Change =(Meditation − Baseline)
(Baseline)
Scan results were statistically evaluated using paired
t-tests comparing the pre- and post-program baseline
scans, and also the change in activation between the
baselineandmeditationscansforboththepre-andpost-
program condition. Similarly, neuropsychological test
scores were compared using paired t-tests. We cor-
rected the CBF data analysis for multiple comparisons
using the False Discovery Rate method [23]. A lim-
ited number of Pearson correlations between changes
× 100