Content uploaded by Domenic A Ciraulo
Author content
All content in this area was uploaded by Domenic A Ciraulo
Content may be subject to copyright.
THE JOURNAL OF ALTERNATIVE AND COMPLEMENTARY MEDICINE
Volume 13, Number 4, 2007, pp. 419–426
© Mary Ann Liebert, Inc.
DOI: 10.1089/acm.2007.6338
Yoga Asana Sessions Increase Brain GABA Levels:
A Pilot Study
CHRIS C. STREETER, M.D.,
1–3
J. ERIC JENSEN, Ph.D.,
2,3
RUTH M. PERLMUTTER, B.S.,
1
HOWARD J. CABRAL, Ph.D.,
4
HUA TIAN, M.S.,
1
DEVIN B. TERHUNE, M.Sc.,
1
DOMENIC A. CIRAULO, M.D.,
1,3,5
and PERRY F. RENSHAW, M.D.
1,2,3
ABSTRACT
Objectives: The aim of this study was to compare changes in brain -aminobutyric (GABA) levels associ-
ated with an acute yoga session versus a reading session. It was hypothesized that an individual yoga session
would be associated with an increase in brain GABA levels.
Design: This is a parallel-groups design.
Settings/location: Screenings, scan acquisitions, and interventions took place at medical school–affiliated
centers.
Subjects: The sample comprised 8 yoga practitioners and 11 comparison subjects.
Interventions: Yoga practitioners completed a 60-minute yoga session and comparison subjects completed
a 60-minute reading session.
Outcome measures: GABA-to-creatine ratios were measured in a 2-cm axial slab using magnetic resonance
spectroscopic imaging immediately prior to and immediately after interventions.
Results: There was a 27% increase in GABA levels in the yoga practitioner group after the yoga session
(0.20 mmol/kg) but no change in the comparison subject group after the reading session (0.001 mmol/kg)
(t 2.99, df 7.87, p 0.018).
Conclusions: These findings demonstrate that in experienced yoga practitioners, brain GABA levels increase
after a session of yoga. This suggests that the practice of yoga should be explored as a treatment for disorders
with low GABA levels such as depression and anxiety disorders. Future studies should compare yoga to other
forms of exercise to help determine whether yoga or exercise alone can alter GABA levels.
419
INTRODUCTION
T
he practice of yoga includes postures (asanas), breath-
ing methods (pranayama), chanting, and meditation
(dhyana).
1,2
As the use of these techniques increases, it is
important that possible mechanisms underlying the effects
of these practices be elucidated.
3
Yoga has shown promise
in improving symptoms associated with depression, anxiety
disorders, and epilepsy.
4–10
These disorders are associated
with low -aminobutyric acid (GABA) states and are ef-
fectively treated with pharmacologic agents that increase the
activity of the GABA system.
11–18
We hypothesize that the
ability of yoga to decrease symptoms in the same disorders
is in part mediated through the GABA system and that such
1
Division of Psychiatry, Boston University School of Medicine, Boston, MA.
2
McLean Hospital, Belmont, MA.
3
Department of Psychiatry, Harvard Medical School, Boston, MA.
4
Department of Biostatistics, Boston University School of Public Health, Boston, MA.
5
Boston V.A. Healthcare System, Boston, MA.
changes can be measured using magnetic resonance spec-
troscopy (MRS).
LOW GABA LEVELS ARE FOUND IN MOOD
AND ANXIETY DISORDERS
GABA dysfunction is a factor in mood disorders.
15
In
rats, multiple efficacious treatments for affective disorders
(e.g., valproate, lithium, tricyclic antidepressants, and elec-
troconvulsive therapy [ECT]) have been shown to increase
brain GABA levels.
15
Cerebrospinal fluid studies have
demonstrated low GABA levels in depressed subjects com-
pared to controls.
19
MRS measurements of occipital GABA
levels demonstrate that patients with major depression have
decreased GABA levels compared to controls.
11
When
treated with serotonin selective reuptake inhibitors (SSRIs)
or ECT, these same patients showed clinical improvement
and increased GABA levels.
20,21
Low GABA activity has
also been observed in anxiety disorders.
22–24
YOGA AS A TREATMENT FOR MOOD AND
ANXIETY DISORDERS
Yoga has been used to reduce symptoms associated with
depression and anxiety disorders.
10
Two reviews of the lit-
erature, reporting studies in which yoga was used to treat
depression and anxiety, identified five randomized con-
trolled trials of depression, of which four reported signifi-
cant reduction in symptoms,
25
and six randomized studies
of anxiety, of which all reported a significant reduction in
symptoms.
25,26
GABA LEVELS, SEIZURES, AND YOGA
Increasing brain GABA levels is one mechanism through
which seizure frequency is decreased.
27
MRS has docu-
mented increases in brain GABA levels associated with the
administration of antiepileptic agents.
12,28
Additionally, the
practice of yoga has shown promise as an adjunctive treat-
ment for epilepsy.
7–9
We speculate that the antiepileptic ef-
fect of yoga, like the anti-epileptic effects of many seizure
medications, is mediated through the GABA system.
Although research in this area would benefit from more
carefully controlled studies, the available studies suggest
that yoga may be used to decrease symptoms associated with
low brain GABA states, such as depression, anxiety disor-
ders, and epilepsy. We hypothesize that one mechanism
through which yoga decreases the symptoms associated with
these conditions is an increase in brain GABA levels. MRS
is a valuable tool in this line of inquiry, because it allows
changes in brain neurochemistry, in the form of GABA lev-
els, to be correlated with a behavioral intervention, such as
the practice of yoga. Understanding the relationship between
behavior and neurochemistry would be an important step in
understanding how the practice of yoga exerts its beneficial
effects.
This study examined brain GABA levels in two groups:
established yoga practitioners (YP) and comparison subjects
(CS), to test the hypothesis that brain GABA levels increase
in YP after a 60-minute yoga session relative to CS after a
60-minute reading period. To our knowledge, this is the first
study to measure the effect of yoga on brain GABA levels.
MATERIALS AND METHODS
Study design
Subjects were recruited through advertisements. Informed
consents, approved by the Boston Medical Center and
McLean Hospital Institutional Review Boards, were ob-
tained for each subject. All subjects were evaluated using
the Structured Clinical Interview for DSM-IV Axis I Dis-
orders: Patient Edition (SCID) and the Addiction Severity
Index (ASI).
29,30
All Axis I diagnoses were established us-
ing SCID criteria. Variables related to alcohol and nicotine
use were obtained from the ASI. A yoga history, medical
history, and physical examination were performed to deter-
mine eligibility. All subjects participated in a baseline mag-
netic resonance spectroscopic imaging (MRSI) scan that
took approximately 60 minutes. Subjects then participated
in a 60-minute intervention, consisting of the practice of
yoga (YP) or reading (CS). The intervention was followed
by a second MRSI scan.
Inclusion/exclusion criteria
Subjects were 18–45-year-old men and women with no
past or current history of psychiatric illness, alcohol or sub-
stance abuse or dependence. Women were required to be us-
ing an acceptable method of birth control and have a nega-
tive urine pregnancy test prior to scanning. YP needed to
report practicing yoga at least 2 days per week for at least
4 months, whereas CS needed to have no previous history
of yoga practice. Individuals with contraindication to mag-
netic resonance evaluation (e.g., pregnancy, claustrophobia,
a cardiac pacemaker, or ferrous implant) or medical histo-
ries that could have influenced scan results (e.g., head in-
jury with loss of consciousness greater than 15 minutes, neu-
rological illness, or serious medical illness) were excluded.
No psychoactive prescription or nonprescription medica-
tions were allowed.
Interventions
The following procedures were used to facilitate a con-
sistent yoga experience across subjects. YP were instructed
to modify their usual practice to a 60-minute time period
that focused on the yoga postures (asanas). Brief quiet pe-
STREETER ET AL.
420
riods at the beginning and end of the 60-minute interven-
tion were allowed, but at least 55 minutes of the hour were
spent doing the asanas and associated breathing exercises
(pranayama). Pranayama or meditation not associated with
an asana or the quiet periods described above were not al-
lowed. All yoga sessions were observed by research staff
with yoga training. The practice guidelines resulted in the
observation of a practice containing very similar sequences
of well-known asanas (e.g., sun salutation, standing poses,
sitting poses, twists, supine and prone poses, inverted poses,
balancing poses, and backbends). The study design stressed
the practice of asanas because they are observable, as op-
posed to the internal state of meditation. The reading mate-
rial used by CS was screened by research staff to exclude
self-help or religious topics. Reading materials consisted of
periodicals in the waiting area and books of popular fiction
brought by the subjects.
Menstrual stage as a covariate
GABA levels change over the menstrual cycle with a de-
cline during the luteal phase.
31
Increased GABA levels as-
sociated with the follicular phase have been reported using
MRS.
32
Because menstrual stage could affect brain GABA
levels, menstrual and contraceptive histories were obtained.
Cycling female subjects were classified as being in the fol-
licular stage if they were in the 14 days immediately after
the first day of the last period or the luteal stage if they were
15 or greater days after the first day of the last period on
the day of the imaging session. Women using hormonal
forms of birth control were classified in a separate category
because they were not cycling normally.
Image acquisition and analysis
Equipment. All data were collected on a 4 Tesla (T) Var-
ian, UnityINOVA, whole-body MR system running VNMR
1.1b (Varian Inc., Palo Alto, CA), using a volumetric TEM
design (Bioengineering Inc., Minneapolis, MN) RF head coil
operating at 170.3 MHz for proton imaging.
Proton-MRSI. Using sagittal images, the bottom of the 2-
cm-thick MRSI slab was aligned with the anterior commis-
sural–posterior commissural line and then rotated 20 de-
grees. Oblique T1-weighted images were then acquired,
allowing visualization of the cortex and deep gray structures
for voxel placement. The two-dimensional J-resolved mag-
netic resonance spectroscopic imaging (2D-JMRSI) acqui-
sition sequence used a slice-selective spin-echo MRSI
scheme modified to incrementally acquire spectra at each
phase-encode step with increasing echo-time to sample the
J-coupling of the coupled metabolites. This acquisition col-
lected 24 individual echo time (TE)-stepped spectra for each
of the 96 circular, sparsely-sampled k-space points, with the
TE ranging from 30 milliseconds to 490 milliseconds in 20
millisecond increments. Acquisition parameters were as fol-
lows: repetition time (TR) 1.25 seconds, sampling ma-
trix 14 14 (circular-sparse), spectral bandwidth 2
kHz, complex time-points 1024, field of view (FOV)
24 24 cm, slab thickness 2 cm, number of averages
1, nominal voxel volume 4.5 cc (effective voxel size 8
cc), total scan duration 48 minutes.
Proton-MRSI processing/analysis. The raw k-space data
were read into a zero-padded 16 16 matrix, for all 24 TE.
The TE-series (24 k-space 2D-JMRSI data sets) were zero-
filled out to 64 k-space 2D-JMRSI data sets in TE and dig-
itally apodized with a 0.5-Hz Gaussian filter prior to Fourier
transforming in the TE dimension. Each J-resolved, 2D-
JMRSI k-space set was then digitally filtered in kx and ky
with a Hanning partial k-space filter to produce J- and spa-
tially resolved spectra. For each subject, an 8 7 2D-JMRSI
matrix was shifted in the x and y dimensions in order to po-
sition the rectangular matrix of voxels inside the skull. Vox-
els anterior to the ventricles were omitted because of arti-
facts from the sinus cavity in the frontal regions. In addition,
voxels falling outside the brain were omitted, ensuring that
only voxels fully inside the brain were included.
33,34
Molar
estimates of global brain GABA levels were calculated for
each subject by correcting the TE-averaged creatine peak area
for T1 and T2-weighting, using values of 1.72 seconds, and
0.273 seconds, respectively, and acquisition parameters
(TR/TE min 1.25 seconds/30 milliseconds). The raw
GABA peak areas were then multiplied by a scaling correc-
tion factor, accounting for the differing number and intensity
of resonances in the 2D GABA spectrum at J 7.5 Hz, as
well as the estimated 12% macromolecule contribution from
the single-voxel data as determined in a prior study.
33
Each
corrected GABA/Cr ratio was then multiplied by 8.34
mmol/L, using the average of the gray/white-matter creatine
levels (9.11 mmol/L/7.58 mmol/L) reported by Dager et al.,
to derive molar estimates of global GABA concentration in
the brain.
35
All spectra were fitted using the LC Model spec-
tral analysis tool (Stephen Provencher, © 1992–2003).
36,37
Statistical analyses. Data were examined for distribution
normality and homogeneity of variance between groups. In
bivariate analyses, categorical data were analyzed using
Fisher’s exact tests. Continuous and ordinal data were ana-
lyzed using student (two-sample) t tests. Analyses of co-
variance (ANCOVA) were used to control for covariates of
a priori theoretical interest (i.e., baseline GABA level, gen-
der, and menstrual stage). Statistical significance required a
two-tailed p value 0.05. Analyses used SPSS Version 13.0
(SPSS, Inc., Chicago, IL).
RESULTS
Subject acquisition
Twenty-nine (29) subjects came to the screening interview
(14 YP/15 CS), of which 22 met entrance criteria (10 YP/12
YOGA ASANA AND GABA LEVELS
421
CS). Two (2) YP were not scanned because of illness. Twenty
(20) subjects participated in the scanning session (8 YP/12
CS). Nineteen (19) subjects (8 YP/11 CS) had complete data
sets; spectral data on 1 CS was of poor quality and could not
be used. Subjects were healthy and did not take prescription
medications, except birth control pills (YP 3, CS 3) and
antibiotics for acne (YP 1), in the month prior to the screen-
ing interview. Thirty days prior to enrollment, over-the-
counter medications were limited to nonsteroidal anti-in-
flammatory medications and acetaminophen.
Demographics and yoga practice
There were no differences between the two cohorts in
age, gender, education, ethnicity, marital status, body mass
index, overall alcohol consumption or consumption to the
point of intoxication in the last 30 days, nicotine use in the
last 30 days, and nicotine use in the last year (Table 1). All
subjects were white with the exception of one Asian in the
comparison group. All anatomical magnetic resonance
imaging scans were without brain abnormalities. Details of
individual yoga practices are reviewed in Table 2. There was
no history of yoga practice in the CS.
Analysis of baseline GABA levels
At baseline, the mean GABA levels were 0.75 0.18
mmol/kg for the YP and 0.94 0.20 mmol/kg for the CS (t
2.11, df 17, p 0.050). When menstrual stage was treated
as a covariate, the differences in the GABA baseline values
between the YP and CS groups remained significant (F[1,
16] 11.83, p 0.003); the effect of menstrual stage was also
significant (F[1, 16] 7.66, p 0.014). A post-hoc t test re-
vealed that subjects in the follicular stage (N 2), all of whom
were in the CS group, exhibited significantly greater baseline
GABA levels (1.05 0.05 mmol/kg) than subjects in the
luteal phase (N 3) (0.75 0.11 mmol/kg), all of whom were
in the YP group (t 3.40, df 3, p 0.04) (Table 3).
Analysis of changes in GABA levels
After the 60-minute interventions, the mean GABA lev-
els were 0.95 0.17 mmol/kg for the YP group and 0.94
0.21 mmol/kg for the CS group. Using a two-sample t test,
there was a significant difference between the GABA change
values (from pre- to postintervention) of 0.20 0.18
mmol/kg (range 0.49) for the YP group and 0.001
0.05 (range 0.18) for the CS group (t 2.99, df 7.87,
p 0.018). Further analysis using an ANCOVA controlling
for menstrual stage confirmed a significant difference in
GABA change values (F[1, 16] 10.19, p 0.006),
whereas menstrual stage was not found to exhibit an inde-
pendent effect (F[1, 16] 0.130, p 0.72). In addition, a
second ANCOVA controlling for baseline GABA levels
found a significant difference between the GABA change
values of the YP and CS groups (F[1, 16] 6.51, df 1,
p 0.021); baseline GABA levels did not exhibit an inde-
pendent effect (F[1, 16] 1.60, p 0.224). Because there
were a greater number of males in the control group, an AN-
COVA contrasting the groups on GABA change values
treating gender as a covariate was performed. The differ-
ence in GABA change values between the YP and CS groups
remained significant (F[1, 16] 10.36, p 0.005); the ef-
fect of gender was nonsignificant (F[1, 16] 0.171, p
0.68). Spectral data demonstrate an increase in the GABA
peak after the yoga asana session in a single YP (Fig. 1).
Individual GABA changes for all subjects are shown in
Figure 2.
DISCUSSION
This study demonstrated that a 60-minute yoga asana ses-
sion in established YP is acutely associated with a 27% in-
crease in GABA levels. When differences in GABA change
scores between groups were analyzed controlling for base-
STREETER ET AL.
422
T
ABLE
1. D
EMOGRAPHIC
C
HARACTERISTICS OF
Y
OGA
P
RACTITIONERS
(YP) (N 8)
AND
C
OMPARISON
S
UBJECTS
(CS) (N 11)
Characteristic YP CS t/df p
Age 25.75 (5.15) 26.55 (7.62) 0.25, 17 0.80
Gender (male) 1/8 (13%) 6/11 (55%) Exact 0.15
Education (years) 16.56 (1.59) 15.82 (2.23) 0.80, 17 0.43
Ethnicity (white) 8/8 (100%) 10/11 (91%) Exact 1.00
Alcohol consumption
a
2.75 (2.43) 3.82 (2.82) 0.86, 17 0.40
Alcohol intoxication 0.25 (0.46) 0.91 (1.58) 1.31, 12.3
c
0.21
Nicotine use
a
0.00 (0.00) 0.00 (0.00) Exact 1.00
Nicotine use
b
3/8 (38%) 1/11 (9) Exact 0.26
Marital status (never married)
d
7/8 10/11 Exact 0.86
Body–mass index 24.03 (3.56) 25.48 (5.44) 0.66, 17 0.49
a
In past 30 days, using the Addiction Severity Index.
b
In last year.
c
Unequal variance.
d
Marital status: YP: 7 never married, 1 married/CS: 10 never married, 1 divorced.
t, t-value; df, degrees of freedom.
line differences in GABA levels, menstrual stage, and gen-
der, the differences in GABA change scores between the
two groups remained significant. The acute elevation of
brain GABA levels following the practice of yoga asanas in
experienced practitioners suggests that yoga asanas may be
an efficacious treatment for low GABA states.
The CS group also had two MRSI scans separated by a
60-minute reading period; however, they did not demon-
strate an increase in GABA levels. Because the change in
GABA levels was only found in the YP group, it is unlikely
that the increase in GABA levels was related to increased
familiarity with the scanning environment or a 60-minute
YOGA ASANA AND GABA LEVELS
423
T
ABLE
2. D
ESCRIPTION OF
Y
OGA
P
RACTICE OF
Y
OGA
P
RACTITIONERS
(YP) (N 8)
Current
Type of yoga sessions/ Session Minutes/
No. training Years week length week Meditation
YP1 Ashtanga 3. 4
a
90
a
360 S
YP2 Ashtanga 2. 5
a
90
a
270 S
Vinyasa
YP3 Ashtanga 4. 5
a
105
a
525 M
Vinyasa
YP4 Ashtanga 1.5 3
a
90
a
270 S
Vinyasa
Hatha
YP5 Ashtanga 10.0 5
a
60
a
300 M & S
Vinyasa
Hatha
Power
YP6 Ashtanga 7. 6
a
, 3
b
30
a
, 150
b
630 S
Vinyasa
Iyengar
Kundalini
YP7 Hatha 2. 3
a
60
b
180 S
Kirpaulo
Iyengar
YP8 Bikram 2. 4
a
90
a
360 N
a
Yoga class.
b
Individual practice.
S, 5–10 minutes of rest (shavaana) at end of yoga practice; M, regular meditation practice (1 hour per week); N, no meditation
practice.
T
ABLE
3. I
NDIVIDUAL
M
ENSTRUAL
S
TAGE AND
GABA D
ATA OF
Y
OGA
P
RACTITIONERS
(YP) (N 8)
AND
C
OMPARISON
S
UBJECTS
(CS) (N 11)
YP CS
Menstrual GABA GABA GABA Menstrual GABA GABA GABA
ID stage scan 1 scan 2 change ID stage scan 1 scan 2 change
YP1 H 0.95 0.87 0.08 CS1 N/A 0.81 0.86 0.05
YP2 H 0.44 0.78 0.35 CS2 H 1.14 1.15 0.02
YP3 L 0.86 1.07 0.21 CS3 N/A 1.13 1.04 0.10
YP4 H 0.83 0.82 0.01 CS4 N/A 0.78 0.83 0.05
YP5 L 0.75 1.10 0.35 CS5 N/A 0.56 0.51 0.05
YP6 L 0.64 0.74 0.10 CS6 N/A 0.77 0.76 0.01
YP7 H 0.95 1.19 0.24 CS7 H 1.03 1.12 0.09
YP8 N/A 0.62 1.03 0.41 CS8 H 1.18 1.19 0.01
CS9 F 1.08 1.09 0.01
C10 N/A 0.86 0.81 0.05
C11 F 1.01 1.00 0.01
Menstrual stage: L, Luteal Stage, F, Follicular Stage; H, Hormonal Therapy; N/A, male, GABA, -aminobutyric acid. Scan 1 and
2 GABA levels in mmol/kg.
period between the two scans. After the reading interven-
tion, the mean GABA level for the CS group was 0.94
0.21 mmol/kg. Multiple subjects in both groups had GABA
levels greater than 1.0 mmol/kg, suggesting that the lack of
change in the CS was not caused by a ceiling effect.
The finding of lower baseline GABA levels in the YP
group was unexpected. The baseline difference in GABA
levels between groups may be explained by the distribution
of females in different stages of the menstrual cycle. The
luteal phase is associated with lower GABA levels than the
follicular stage.
31,32
Of the subjects assigned to a menstrual
stage, all subjects in the luteal stage were in the YP group,
whereas all subjects in the follicular were in the CS group,
a distribution that would favor higher baseline GABA lev-
STREETER ET AL.
424
Chemical Shift (ppm)
4.0 3.8 3.6 3.4 3.2 3.0 2.8 2.6 2.4 2.2 2.0
Chemical Shift (ppm)
4.0 3.8 3.6 3.4 3.2 3.0 2.8 2.6 2.4 2.2 2.0
Chemical Shift (ppm)
4.0 3.8 3.6 3.4 3.2 3.0 2.8 2.6 2.4 2.2 2.0
Chemical Shift (ppm)
4.0 3.8 3.6 3.4 3.2 3.0 2.8 2.6 2.4 2.2 2.0
1.4
1.2
1
0.8
0.6
0.4
pre-GABA levels
Yoga practitioners Comparison subjects
GABA (mmol/kg)
post-GABA levels pre-GABA levels post-GABA levels
0.2
0
+
+
FIG. 1. (A) The chemical-shift imaging grid depicts the multiple voxels within the brain that are totaled to estimate the -gamma
aminobutyric acid (GABA) level of the whole slab. (B) J-resolved sample spectra (J 7.5 Hz) are shown from two regions before and
after yoga practice in a yoga practitioner. Spectra are displayed with LC Model fit (Stephen Provencher, © 1992–2003) and are not fil-
tered. (C) The spectral data are removed, leaving only LC Model fit magnified 11 to show the GABA resonance area at 2.95 ppm,
which demonstrates an increase in GABA peak after yoga practice.
C
A
B
B
FIG. 2. Individual GABA levels (pre- and post-intervention) for yoga practitioners (YP) (N 8) and comparison subjects (CS) (N
11).
els in the CS group, as was found. Although baseline GABA
values may have been affected by menstrual stage, changes
in GABA levels after a yoga session seemed to be unrelated
to menstrual stage. This study indicates that menstrual stage
should be considered when measuring GABA levels.
One strength of the present study is the inclusion of a
carefully screened population that is free of psychiatric and
neurologic illness and psychoactive medications. The in-
clusion of YP from many different schools with varying de-
grees of practice can be viewed as both a strength and weak-
ness in this study. The consistency of the yoga sessions
observed allows comparison and implies that the changes in
GABA levels are related to a sequence of asanas and
pranayama and not limited to a specific school of yoga. Al-
though the study was not randomized, nor was there a
crossover where YP were scanned after a reading interven-
tion, the results of this study strongly suggest that yoga can
increase GABA levels. This pilot study did not control for
activity levels. Other forms of exercise, such as walking,
may lead to the same increases in GABA levels as observed
in the YP group. Further studies are required to understand
the relationship of GABA levels to standard exercises.
CONCLUSIONS
The World Health Organization reports that mental illness
makes up 15% of the global burden of disease.
38
Depression
and anxiety disorders both contribute to this burden, and are
attended by low GABA levels, relative to normal controls, as
measured by MRS.
11,22
These disorders are successfully
treated with pharmacologic agents known to increase the ac-
tivity of the GABA system.
15,18
The significant comorbidity
between depression and anxiety disorders and their success-
ful treatment with agents that affect the GABA system sup-
port the theory that low GABA activity is involved in the
pathology of these disorders and increased GABA activity is
associated with symptom reduction.
39
The literature suggests
that the practice of yoga is also associated with symptom re-
duction in depression and anxiety disorders.
10
This study
demonstrates that brain GABA levels increase by 27% after
a 1-hour yoga asana practice in experienced partitioners and
suggests that the practice of yoga should be explored and com-
pared to other exercise modalities as a treatment or adjunc-
tive treatment for disorders associated with low GABA states.
The development of an inexpensive, widely available inter-
vention, with few side effects, that is effective in alleviating
the symptoms of disorders associated with low GABA states
has clear public health advantages.
ACKNOWLEDGMENTS
This study was supported, in part, by grants from the Na-
tional Institute of Drug Abuse (DA09448 and DA15116 to
P.F.R., DA50038 to D.A.C.); the National Institute on Al-
cohol Abuse and Alcoholism (K23AA13149 to C.C.S.,
AA013727 to D.A.C.); The National Center for Research
Resources (M01RR00533); and Gennaro Acampora Char-
ity Trust to the Division of Psychiatry, Boston Medical Cen-
ter. This experiment complied with the current laws of the
country in which it was performed, the United States. Insti-
tutions in which the work was performed are as follows: The
General Clinical Research Center of Boston University
School of Medicine and the Brain Imaging Center of
McLean Hospital.
REFERENCES
1. Iyengar BKS. Light on Yoga. New York: Schocken Books,
1966.
2. Lowitz L, Datta R. Sacred Sanskrit Words, for Yoga Chant
and Meditation. Berkeley, CA: Stone Bridge Press, 2005.
3. Barnes PM, Powell-Griner E, McFann K, et al. Complemen-
tary and alternative medicine use among adults: United States,
2002. Adv Data 2004;343:1–19.
4. Janakiramaiah N, Gangadhar BN, Naga Venkatesha Murthy
PJ, et al. Antidepressant efficacy of Sudarshan Kriya Yoga
(SKY) in melancholia: A randomized comparison with elec-
troconvulsive therapy (ECT) and imipramine. J Affect Disord
2000;57:255–259.
5. Girodo M. Yoga meditation and flooding in the treatment
of anxiety neurosis. J Behav Ther Exp Psychiatry 1974;5:
157–160.
6. Shannahoff-Khalsa D, Ray D, Levine S, et al. Randomized
controlled trial of yogic meditation techniques for patients with
obsessive compulsive disorders. CNS Spectrums 1999;4:
34–46.
7. Yardi N. Yoga for the control of epilepsy. Seizure 2001;10:
7–12.
8. Ramaratnam S, Sridharan K. Yoga for epilepsy. Cochrane
Database Syst Rev 2000;3:CD001524.
9. Panjwani U, Selvamurthy W, Singh S, et al. Effect of Sahaja
yoga meditation on seizure control and EEG changes in pa-
tients of epilepsy. Indian J Med Res 1996;103:165–172.
10. Khalsa S. Yoga as a therapeutic intervention: A bibliometric
analysis of published research studies. Indian J Physiol Phar-
macol 2004;48:269–285.
11. Sanacora G, Mason GF, Rothman DL, et al. Reduced cortical
gamma-aminobutyric acid levels in depressed patients deter-
mined by proton magnetic resonance spectroscopy. Arch Gen
Psychiatry 1999;56:1043–1047.
12. Petroff O, Rothman D, Behar K, et al. Human brain GABA
levels rise rapidly after initiation of vigabatrin therapy. Neu-
rology 1996;47:1567–1571.
13. Sendrowski K, Sobaniec W. New antiepileptic drugs—an
overview. Rocz Akad Med Bialymst 2005;50(suppl 1):96–98.
14. Keck JP, McElroy S. Antiepileptic Drugs. Washington, DC:
American Psychiatric Press, 1998.
15. Brambilla P, Perez J, Barale F, et al. GABAergic dysfunction
in mood disorders. Mol Psychiatry 2003;8:721–737.
16. Schatzberg AF, Nemeroff C. Textbook of Psychopharmacol-
ogy. Washington, D.C.: American Psychiatric Press, 1998.
YOGA ASANA AND GABA LEVELS
425
17. Taylor C. Treatment of Anxiety Disorders. Washington, DC:
American Psychiatric Press, 1998.
18. Breier A, Paul S. The GABAa/benzodiazepine receptor: Im-
plications for the molecular basis of anxiety. J Psychiatr Res
1990;24:91–104.
19. Kasa K, Otsuki S, Yamamoto M, et al. Cerebrospinal fluid
gamma-aminobutyric acid and homovanillic acid in depressive
disorders. Biol Psychiatry 1982;17:877–883.
20. Sanacora G, Mason GF, Rothman DL, et al. Increased occip-
ital cortex GABA concentrations in depressed patients after
therapy with selective serotonin reuptake inhibitors. Am J Psy-
chiatry 2002;159(4):663–665.
21. Sanacora G, Mason GF, Rothman DL, et al. Increased corti-
cal GABA concentrations in depressed patients receiving ECT.
Am J Psychiatry 2003;160:577–579.
22. Goddard AW, Mason GF, Almai A, et al. Reductions in oc-
cipital cortex GABA levels in panic disorder detected with 1h-
magnetic resonance spectroscopy. Arch Gen Psychiatry 2001;
58:556–561.
23. Lydiard RB. The role of GABA in anxiety disorders. J Clin
Psychiatry 2003;64(suppl 3):21–27.
24. Goddard A, Mason G, Appel M, et al. Impaired GABA neu-
ronal response to acute benzodiazepine administration in panic
disorder. Am J Psychiatry 2004;161:2186–2193.
25. Kirkwood G, Rampes H, Tuffrey V, et al. Yoga for anxiety:
A systematic review of the research evidence. Br J Sports Med
2005;39:884–891.
26. Pilkington K, Kirkwood G, Rampes H, et al. Yoga for de-
pression: The research evidence. J Affect Disord 2005;89:
13–24.
27. Meldrum S. GABAergic mechanisms in the pathogenesis and
treatment of epilepsy. Br J Clin Pharmacol 1989;27:3S–11S.
28. Petroff OA, Hyder F, Rothman DL, et al. Topiramate rapidly
raises brain GABA in epilepsy patients. Epilepsia 2001;42:
543–548.
29. First MB, Spitzer R, Gibbon M, et al. Structured Clinical In-
terview for DSM-IV Axis I Disorders: Patient Edition (SCID-
I/P, Version 2.0). New York: Biometric Research Department,
New York State University, 1995.
30. McLellan A, Kushner H, Metzger D, et al. The fifth edition
of the Addiction Severity Index. J Subst Abuse Treat
1992;9:119–213.
31. Griffiths J, Lovick T. Withdrawal from progesterone increases
expression of alpha4, beta1 and delta GABA(A) receptor sub-
units in neurons in the periaqueductal gray matter in female
Wistar rats. J Comp Neurol 2005;486:89–97.
32. Epperson CN, Haga K, Mason GF, et al. Cortical gamma-
aminobutyric acid levels across the menstrual cycle in healthy
women and those with premenstrual dysphoric disorder: A pro-
ton magnetic resonance spectroscopy study. Arch Gen Psy-
chiatry 2002;59:851–858.
33. Jensen J, Frederick B, Brown J, et al. Two-dimensional, J-re-
solved spectroscopic imaging of GABA at 4 Tesla in the hu-
man brain. Magnetic Resonance in Medicine 2005;54:
783–788.
34. Jensen J, Frederick B, Renshaw P. Grey and white matter
GABA level differences in the human brain using two-di-
mensional, J-resolved spectroscopic imaging. NMR Biomed
2005;18:570–576.
35. Dager SR, Friedman SD, Parow A, et al. Brain metabolic al-
terations in medication-free patients with bipolar disorder.
Arch Gen Psychiatry 2004;61:450–458.
36. Provencher SW. Estimation of metabolite concentrations from
localized in vivo proton NMR spectra. Magn Reson Med
1993;30:672–679.
37. Provencher SW. Automatic quantitation of localized in vivo
1H spectra with LCModel. NMR Biomed 2001;14:260–264.
38. Murray C, Lopez A. The Global Burden of Disease: A Com-
prehensive Assessment of Mortality and Disability from Dis-
eases, Injuries and Risk Factors in 1990 and Projected to 2020.
Cambridge, MA: Harvard University Press, 1996.
39. Barbee JG. Mixed symptoms and syndromes of anxiety and
depression: Diagnostic, prognostic, and etiologic issues. Ann
Clin Psychiatry 1998;10:15–29.
Address reprint requests to:
Chris C. Streeter, M.D.
Division of Psychiatry
Boston University School of Medicine
85 East Newton Street M912E
Boston, MA 02118
E-mail: streeter@bu.edu
STREETER ET AL.
426
