ArticlePDF Available

Dark Chocolate May Improve the Metabolic Response to Stress, Nestlé Scientists report

  • Swiss Nutrition and Health Foundation

Abstract and Figures

Lausanne, SWITZERLAND 11 November 2009 - Anxiety and stress can have considerable effects on human health, causing a variety of physical and emotional conditions, and sometimes leading to more serious health concerns. Scientists at the Nestlé Research Center in Lausanne, Switzerland, found that beneficial constituents in dark chocolate may improve the metabolic state of people that report feeling higher levels of stress. The full article is available in the Journal of Proteome Research. In the present study, 30 healthy adults consumed two portions of 20g daily of dark chocolate for fourteen consecutive days. Scientists at the Nestlé Research Center measured the subjects' global metabolic responses attributed to daily dark chocolate consumption, with particular emphasis on stress-related metabolic changes such as energy metabolism and gut microbial activities. They additionally assessed participants' anxiety characteristics using validated questionnaires. Results indicated that for individuals that reported feeling higher levels of stress, daily dark chocolate consumption had a positive impact on stress-associated metabolic activities. These findings suggest beneficial attributes of consuming dark chocolate to improve the metabolic reaction to stress. Lifestyle and genetic factors, including diet, substantially influence individuals' metabolic responses. A previous study by Nestlé scientists revealed that dietary preferences, including chocolate consumption, can significantly impact energy and microbiota metabolism. Scientists at the Nestlé Research Center continue to strengthen their position that the gut ecology and metabolic activity of healthy individuals may be modulated by the diet. "Consuming dark chocolate daily can positively impact the metabolism of people that report having high-stress levels," says Sunil Kochhar, Nestlé researcher leading the study. "These results strongly support our ongoing metabonomics research efforts to ascertain the impact of certain foods on human metabolism through the adaptation of gut microbial activities."
Content may be subject to copyright.
Metabolic Effects of Dark Chocolate Consumption on Energy, Gut
Microbiota, and Stress-Related Metabolism in Free-Living Subjects
Francois-Pierre J. Martin,
Serge Rezzi,
Emma Pere´-Trepat,
Beate Kamlage,
Sebastiano Collino,
Edgar Leibold,
Ju¨rgen Kastler,
Dietrich Rein,
Laurent B. Fay,
Sunil Kochhar*
Nestle´ Research Center, Vers-chez-les-Blanc, CH-1000 Lausanne 26, Switzerland, Metanomics GmbH, Tegeler
Weg 33, 10589 Berlin, Germany, BASF SE, 67056 Ludwigshafen, Germany, and Metanomics Health GmbH,
Tegeler Weg 33, 10589 Berlin, Germany
Received July 27, 2009
Dietary preferences influence basal human metabolism and gut microbiome activity that in turn may
have long-term health consequences. The present study reports the metabolic responses of free living
subjects to a daily consumption of 40 g of dark chocolate for up to 14 days. A clinical trial was performed
on a population of 30 human subjects, who were classified in low and high anxiety traits using validated
psychological questionnaires. Biological fluids (urine and blood plasma) were collected during 3 test
days at the beginning, midtime and at the end of a 2 week study. NMR and MS-based metabonomics
were employed to study global changes in metabolism due to the chocolate consumption. Human
subjects with higher anxiety trait showed a distinct metabolic profile indicative of a different energy
homeostasis (lactate, citrate, succinate, trans-aconitate, urea, proline), hormonal metabolism (adrenaline,
DOPA, 3-methoxy-tyrosine) and gut microbial activity (methylamines, p-cresol sulfate, hippurate). Dark
chocolate reduced the urinary excretion of the stress hormone cortisol and catecholamines and partially
normalized stress-related differences in energy metabolism (glycine, citrate, trans-aconitate, proline,
β-alanine) and gut microbial activities (hippurate and p-cresol sulfate). The study provides strong
evidence that a daily consumption of 40 g of dark chocolate during a period of 2 weeks is sufficient to
modify the metabolism of free living and healthy human subjects, as per variation of both host and
gut microbial metabolism.
Keywords: Chronic stress
Dark chocolate
Mass spectrometry
Nuclear magnetic resonance spectroscopy
Metabolic phenotype of mammals results from the combi-
nation of multiple genetic, environmental and sociocultural
In man, dietary preferences, lifestyle and
genetics influence individual metabolic phenotype, and there-
fore determine health status and the likelihood to develop
Variations in the dietary pattern affect the metabo-
lism of humans via the key entry points of gut microbiota.
Therefore, there is clearly a need to understand human
metabolism at the system level with emphasis on the expression
of both host and meta-genomes, environmental and lifestyle
factors to meet the ultimate goal of providing better health and
wellbeing with nutrition. Although predominantly cultural in
origin, dietary preferences also result from multiple biological
and behavioral processes, which integrate satiety, psychological
perception and metabolic effects of foods.
We have recently
described how dietary preferences can be associated with
specific signatures in the metabolic phenotypes of healthy
humans, with a metabolic signature based on a modulation of
host and gut microbial metabolism.
Perhaps one of the greatest challenges in modern nutrition
is to interrogate and classify the critical metabolic interactions
between the complex food matricesscontaining a wide range
of biologically active compoundssand human system metabo-
lism and to understand their role in diverse human disease
processes. The sheer complexity of a food matrix, such as dark
chocolate, may determine a large variety of effects on the
metabolism. Many studies have indeed demonstrated the
potential health implications of dark chocolate constituents,
but rarely as a whole product. For instance, cocoa is rich in
flavonoids, mainly flavan-3-ols (epicatechin, catechin and their
oligomers), which were associated with benefits on cardiovas-
cular health by maintaining low blood pressure, improving
endothelial function, and by reducing thrombotic state, oxida-
tive and inflammatory states.
Benefits of cocoa on improve-
ment of insulin sensitivity and glucose tolerance were also
* To whom correspondence should be addressed. Sunil Kochhar, Nestle´
Research Center, BioAnalytical Sciences, P.O. Box 44, Vers-chez-les-Blanc,
CH-1000 Lausanne 26, Switzerland. E-mail,;
telephone, +41 785 9336; fax, +41 (21) 785 9486.
Nestle´ Research Center.
Contributed equally to the manuscript.
Metanomics GmbH.
Metanomics Health GmbH.
5568 Journal of Proteome Research 2009, 8, 5568–5579 10.1021/pr900607v CCC: $40.75 ©2009 American Chemical Society
Published on Web 10/07/2009
Other biochemically active molecules naturally
occurring in chocolate include theobromine, a bitter alkaloid
also known to reduce blood pressure, phenylethylamine, a
monoamine alkaloid which can act as neurotransmitter, and
N-oleoyl- and N-linoleoyl-ethanolamine that slow the rate of
anandamide breakdown, a brain neurotransmitter.
fore, if there is growing evidence on the health benefits
associated with chocolate, mechanisms of action of chocolate
bioactive components at the molecular levels are poorly
understood. This is particularly the case for benefits related to
brain health and improvement of stress states where only
symptomatic data, such as brain blood flow, are available.
In the present study, we have sought to capture a global view
of the metabolic changes associated with chocolate consump-
tion in healthy and free living men and women using meta-
bonomics. Nutrimetabonomics provides a system approach to
assess systemic metabolic status of an individual, which
encapsulates information on genetic and environmental
gut microbiota activity,
and food
Here, we have used proton nuclear magnetic resonance
(1H NMR) spectroscopy and mass spectrometry (MS) as comple-
mentary analytical platforms for monitoring metabolic changes
associated with a daily intake of 40 g of dark chocolate over a
period of 2 weeks in the urine and blood plasma of 30
individuals classified according to their self-reported anxiety
trait. We describe the metabolic variations induced by dark
chocolate and discuss their association with changes in energy
homeostasis, gut microbial activity and the metabolism as-
sociated with stress.
Material and Methods
Recruitment of Volunteers. This study was conducted by
TNO Quality of Life, Zeist (The Netherlands) in accordance with
the ethical principles of Good Clinical Practice and the Dec-
laration of Helsinki. The protocol was approved by the Medical
Ethics Committee METOPP (Medisch-Ethische Toetsing Onder-
zoek Patie¨ nten en Proefpersonen/medical ethics review of
research with patients and test subjects) on April 3, 2006 (The
Netherlands). A total of 30 subjects (11 males, 19 females) were
enrolled in the study and gave written informed consent (Table
1). The study was designed as a randomized (by age, gender,
anxiety trait), parallel, open study. The inclusion of the
volunteers was decided upon medical history, age (18-35
years), body mass index (BMI, 18-25 kg ·m-2), and blood
clinical analyses (Table 1). The exclusion criteria included
psychiatric, metabolic, endocrine, gastrointestinal and eating
behavior disorders. In addition, smoking, use of medication
that may influence appetite and/or sensory functioning, preg-
nancy, reported slimming or medically prescribed diet, reported
unexplained weight loss or gain in the month prior to the
screening, alcohol consumption superior to, respectively, 21
and 28 units per week for females and males were also
considered as exclusion criteria.
Clinical Trial. Participants were asked to avoid consumption
of chocolate or chocolate containing products during an 8-day
run-in period. The nutritional intervention lasted 2 weeks with
a daily intake of 40 g of commercially available dark chocolate
(Noir Intense, 74% cocoa solids, Nestle´ ). On day 1 (preinter-
vention) and days 8 and 15 (postintervention), fasting blood
plasma and morning spot urine samples were collected. A daily
amount of 40 g of dark chocolate was consumed twice per day
as a midmorning and a midafternoon snack (20 g each).
Participants were divided up into either high or low anxiety
trait subgroup according to the evaluation of their dispositional
stress as assessed by scoring on the anxiety trait scale of the
State-Trait Anxiety Inventory (STAI) test.
The STAI is the
definitive instrument for measuring anxiety in adults that
clearly differentiates between the temporary condition of
“anxiety state” and the more general and long-standing status
of anxiety trait. The STAI scores from 70 to 78 and from 42 to
64 described low and high anxiety trait, respectively (Table 1).
The study included 4 high and 7 low anxiety trait males, and 9
high and 10 low anxiety trait females (Table 1). Metabolic data
from subjects having reported adverse events, such as nausea,
vomiting, or diarrhea (subjects ID 6, 15, 17, and 44) were
excluded from the statistical analysis to avoid introduction of
biases in the final outcome of the study.
1H NMR Analysis of Plasma and Urine Samples. Plasma
samples (495 μL) were introduced into 5 mm NMR tubes with
55 μL of deuterium oxide (D2O) used as locking substance.
Urine samples (500 μL) were adjusted to pH 6.8 using 100 μL
of a deuterated phosphate buffer solution (KH2PO4, final
concentration of 0.2 M) containing 1 mM of sodium 3-(trim-
ethylsilyl)-[2,2,3,3-2H4]-1-propionate (TSP) intoa5mmNMR
Metabolic profiles were measured at 300 K on a Bruker
Avance II 600 MHz spectrometer equipped witha5mminverse
probe (Bruker Biospin, Rheinstetten, Germany). Three types
of 1H NMR spectra were registered for each blood plasma
sample, including a standard 1H detection with water suppres-
sion, a Carr-Purcell-Meiboom-Gill spin-echo with water
suppression and a diffusion-edited pulse sequences, as reported
The standard spectra were acquired with a relax-
ation delay of 4 s and a mixing time of 100 ms. CPMG
Table 1. Participant Information
Subject ID age [years] BMI [kg/m2]
high (H)/low
(L) anxiety trait
score gender
2 28 21.4 L 72 F
5 18 21.2 H 50 F
6 24 25.4 H 61 M
7 18 21.1 L 76 M
9 23 23.3 L 72 M
15 26 24.1 H 50 F
17 21 19.7 L 74 F
20 18 21.0 L 78 F
21 24 20.9 L 71 M
25 21 21.8 L 72 M
31 19 18.8 H 55 F
36 31 22.0 H 53 F
41 22 21.6 H 56 M
44 23 22.7 L 71 M
45 23 21.8 H 56 F
49 23 20.0 H 63 F
53 21 20.8 L 72 F
56 20 23.0 L 73 F
59 26 21.4 L 76 M
61 18 20.9 H 56 M
65 22 21.4 L 72 F
72 28 24.6 L 73 F
74 23 20.2 H 56 F
76 34 20.0 H 56 F
77 22 23.6 L 73 F
78 22 24.2 L 76 F
80 26 21.3 H 61 M
87 20 18.6 H 59 F
94 21 24.4 L 70 F
95 26 20.6 L 70 M
Metabolic Effects of Dark Chocolate Consumption research articles
Journal of Proteome Research Vol. 8, No. 12, 2009 5569
spin-echo spectra were measured using a spin-echo loop time
of 19.2 ms and a relaxation delay of 4 s. Diffusion-editing
spectra were obtained using a relaxation delay of 1 s, pulsed
field gradients set at 46.8 G ·cm-1, and a diffusion delay of 120
ms during which the molecules are allowed to diffuse. Urine
spectra were acquired using the standard sequence, with a
relaxation delay of 2.5 s and a mixing time of 100 ms.
For each sample, 32 (plasma) and 256 (urine) free induction
decays (FIDs) were collected into 64 K data points using a
spectral width of 12019.2 Hz, corresponding to an acquisition
time of 2.7 s. Prior to Fourier transformation, FIDs were
multiplied by an exponential weighting function corresponding
to a line broadening of 0.3 Hz (standard and CPMG spectra)
and of 1 Hz (diffusion-edited spectra). The acquired NMR
spectra were manually corrected for phase and baseline distor-
tions, and referenced to the chemical shift of R-glucose at δ
5.236 for plasma and of TSP at δ0.0 for urine using the
TOPSPIN (version 2.1, Bruker Biospin, Rheinstetten, Germany)
software package.
The metabolite identification was achieved using literature
and confirmed by 2D 1H NMR spectroscopy experi-
ments performed on selected samples.
Mass Spectrometric Analysis of Plasma and Urine
Samples. For mass spectrometry-based metabolite profiling
analyses, proteins were removed from plasma and urine
samples by precipitation. Subsequently, two nonpolar and two
polar fractions were separated for GC-MS and LC-MS/MS
analysis, respectively, by adding water and a mixture of ethanol
and dichloromethane. For GC-MS analysis, the nonpolar frac-
tion was treated with methanol under acidic conditions to yield
the fatty acid methyl esters. Both fractions were further
derivatized with O-methyl-hydroxyamine hydrochloride and
pyridine to convert oxo-groups to O-methyl-oximes and sub-
sequently with a silylating agent before analysis.
analysis, both fractions were reconstituted in appropriate
solvent mixtures. HPLC was performed by gradient elution
using methanol/water/formic acid on reversed phase separa-
tion columns. Mass spectrometric detection technology which
allows target and high sensitivity MRM (Multiple Reaction
Monitoring) profiling was performed in parallel to a full screen
analysis. In the case of urine analysis, a photometric creatinine
analysis according to Jaffe´ was performed prior to polar MS
analysis and samples diluted to the same creatinine concentra-
The polar fraction was applied to each of the systems.
For GC-MS and LC-MS/MS profiling, data were normalized to
the median of reference samples which were derived from a
pool of all pretreatment samples (Day 1) to account for inter-
and intrainstrumental variation. Steroids, catecholamines and
their metabolites were measured by online SPE-LC-MS/MS
(Solid phase extraction-LC-MS/MS).
In the case of urine
samples, conjugated derivatives of steroids were enzymatically
cleaved prior to analyses using a beta-glucuronidase [EC] and an arylsulfatase [EC] from Helix pomatia.
For plasma measurements, absolute quantification was per-
formed by means of stable isotope-labeled standards. The
analyses of cortisol in plasma as well as catecholamines and
steroids in urine were performed by normalization to pool
levels as described for profiling.
Chemometrics. NMR data was converted into 22 K data
points over the range of δ0.2-10.0 and imported into MATLAB
environment (The MathWorks, Inc., Natick, MA). Interpolation
of all the spectra to the same chemical shift followed by zeroing
the intensity values of the water peak from δ4.68 to 5.10 was
performed. The NMR spectra were normalized to a constant
total sum of all intensities within the specified range and
autoscaled prior to multivariate data analyses (MVA). MS data
were normalized to the median of pooled samples and au-
toscaled before MVA.
Principal components analysis (PCA)
was first performed
to visualize the global variance of the data sets and to pinpoint
outliers. PCA is an important tool for visualizing data structures
and one of the most applied dimensionality reduction methods.
The aim of PCA is to represent the original data (X) by a set of
new orthogonal variables so-called principal components (PCs)
which are linear combinations of the original variables. Because
the extracted PCs maximize the data variance, PCA is sensitive
to the presence of outliers. The data matrix Xis decomposed
to a score matrix Uand a loading matrix V, plus an error matrix
E. The elements of the loadings give information about the
contribution of the original variables (NMR or MS) to each PC
and the elements of the scores provide information about
metabolic similarities and dissimilarities between samples.
Partial Least Squares (PLS) and Orthogonal PLS (O-PLS)
discriminant analyses (PLS-DA and O-PLS-DA) were also
applied for detailed classification purposes.
In PLS-DA, a
dependent variable yis modeled using latent variables, maxi-
mizing the covariance between X(NMR or GC/LC-MS data)
and y. Variable yis a binary vector with value 0 for one class
and value 1 for the other class under study; in this paper, yis
related to time- and anxiety trait-dependent metabolic varia-
tions after dark chocolate supplementation.
In particular, O-PLS-DA, which is a modification of PLS,
separates the systematic variation in Xinto two parts, one
linearly related to yand representing the between class
variation, and another one orthogonal to yand representing
the within class variation. In other words, it provides a way to
filter out metabolic information (NMR or GC/LC-MS data) that
is not correlated to the predefined classes (time, anxiety trait).
The robustness of statistical models was assessed using the
standard 7-fold cross validation method. Validity of the model
against overfitting was determined by computing the total
explained variance of Xand y(R2(X),R2(Y)) and the cross-
validated predictive ability (Q2(Y)) values of the models as
reported in Tables 2 and 3. Negative or very low values of the
Q2(Y) indicate that no statistically significant differences were
observed. Influential variables that are correlated to the group
separation are identified using the variable coefficients accord-
ing to a previously published method.
S-plot was used to
visualize the variable influence in the MS data models.
addition, a Student’s ttest with an alpha of 0.05 was performed
on GC/LC-MS variables and one representative NMR signal
areas representative of influential metabolites.
Chemometric analysis was performed using the SIMCA-P+
(version 12.0, Umetrics AB, Umeå, Sweden) software package
and in-house developed MATLAB routines.
NMR and MS Metabolic Profiling of Blood Plasma and
Urine. A wide range of amino and organic acids, ketone bodies,
sugars, osmolytes, saturated and unsaturated fatty acids and
triglycerides were detected using 1H NMR spectroscopy and
GC/LC-MS analysis of blood plasma. Holistic NMR plasma
profiles dominated by low molecular weight components
(CPMG spectra) and macromolecules (diffusion-edited spectra)
were complemented by GC/LC-MS semiquantitative measures
of 148 targeted metabolites. Similarly, the 1H NMR urine
research articles Martin et al.
5570 Journal of Proteome Research Vol. 8, No. 12, 2009
profiles exhibited a set of resonances arising from major low
molecular weight molecules, such as ketone bodies, organic
acids, amino acids, and aromatic metabolites were completed
with 157 targeted metabolites measured with LC-MS.
Overview of Metabolic Variations by PCA. PCA was first
applied to assess the inherent similarity of the urine and plasma
metabolic profiles using 4 PCs. For urine, PC1-4explained,
respectively, 12, 7, 6, and 5% of the total variance present in
the NMR data and 17, 8, 7, and 6% in the MS data (Supple-
mentary Figure 1A,B). The urine samples from the subject 21
were removed from the statistical analysis due to a statistically
dominant dilution effects in the NMR profile. For plasma, PC1-4
explained 23.9, 8.7, 5.3, and 4.5% of the total variance,
respectively, in the NMR data and 13.8, 9.0, 6.5, and 5.6% in
the MS data (Supplementary Figure 1C,D). PCA highlighted that
interindividual variability of the metabolic profiles of urine and
plasma tended to be greater than intraindividual variations
(Supplementary Figure 1). Analysis of NMR-derived models
revealed that interindividual differences were associated with
variations in the urinary levels of creatinine, trimethylamine-
N-oxide (TMAO), hippurate, citrate, and lactate, and plasma
composition in lipoproteins, lipids, phosphocholine and glu-
cose. Investigations of the MS-derived models indicated that
the main source of metabolic variations between subjects was
due to changes in the urinary excretion of xylose, lactate,
glycerate, lysine and 4-dihydroxyhippurate, and plasma con-
centrations of serotonine, corticosterone, 3,4-hydroxypheny-
lacetate, and homovanillate. Interpretation of the PCA scores
plot did not reveal any distribution of the samples according
to age, time, BMI and self-reported anxiety. However, a
separation trend due to gender differences could be observed
and was particularly marked in MS urinary data along the
second PC.
Additional investigations were performed using O-PLS-DA
to maximize the separation between the groups of samples
(high and low anxiety trait or time of sample collection) and
identify class-specific metabolites (Tables 2 and 3).
Daily Consumption of Dark Chocolate Induces a Speci-
fic Metabolic Signature. Supervised chemometric analyses of
the urine NMR and MS data revealed statistically significant
time-dependent changes, as noted by the positive value of the
model predictability parameter Q2(Y). Interpretation of the urine
O-PLS-DA scores plots (Figures 1 and 2) and model descriptors
(Table 2) indicated that a 1 week daily intake of dark chocolate
by free living subjects is reflected in the metabolic profiles as
assessed by NMR spectroscopy. The metabolic changes became
even more significant after 2 weeks of consumption, as
observed by the greater value of Q2(Y) value and a clearer
separation of the two groups of samples in the scores plot
(Table 2, Figures 1 and 2). Interpretation of the corresponding
O-PLS-DA coefficients plots indicated that, after 1 week,
chocolate consumption resulted in increased levels of 4-hy-
droxyphenylacetate and several unassigned metabolites giving
resonances at δ7.85 (s), 8.03 (s), 8.22 (s), 3.0 (s), 3.39 (s), 3.91
(m) and 3.53(s) (Table 4, Figure 1). These metabolic changes
were associated with downward trends in creatinine, and an
unassigned aromatic metabolite giving resonances at δ7.51 (m)
and 7.70 (m). Comparison of samples obtained at baseline and
after 2 weeks of treatment revealed greater metabolic changes
in both endogenous and gut microbial metabolism. In par-
ticular, after 2 weeks of chocolate consumption, these meta-
bolic changes were maintained and were associated with
additional decreased levels of phenylacetylglutamine and p-
cresol sulfate (Table 4).
Intriguingly, MS-based metabolic profiling showed that
chocolate-induced metabolic effects were statistically signifi-
cant only in subjects with inherent high anxiety trait (Tables 2
and 4). Interpretation of the loadings plot showed that choco-
late consumption was associated with decreased levels of
normetanephrine, adrenaline, corticosterone, noradrenaline,
progesterone, leucine, cortisol and asparagine, and an increase
Table 2. O-PLS-DA Model Summary for Discriminating Urine
and Plasma Metabolic Profiles Based on the Dark Chocolate
Consumption along the Study
day 0
vs 8
day 8
vs 15
day 0
vs 15
NMR Diffusion-edited
plasma data
plasma data
NMR Urine data Q2(Y) 0.32 0.16 0.38
R2(X) 0.20 0.17 0.14
R2(Y) 0.95 0.70 0.87
NMR Urine data
(high stress subjects)
Q2(Y) 0.26 0.16 0.26
R2(X) 0.26 0.38 0.20
R2(Y) 0.96 0.74 0.90
NMR Urine data
(low stress subjects)
Q2(Y) 0.22 0.07 0.28
R2(X) 0.16 0.19 0.15
R2(Y) 0.92 0.88 0.91
MS Plasma data Q2(Y) NS NS NS
MS Urine data Q2(Y) NS NS NS
MS Urine data
(high stress subjects)
Q2(Y) NS NS 0.17
R2(X) 0.24
R2(Y) 0.88
MS Urine data
(low stress subjects)
Key: NS, the class separation obtained through the predictive
component was not statistically significant.
Table 3. O-PLS-DA Model Summary for Discriminating Urine
and Plasma Metabolic Profiles Based on High and Low
Anxiety Traits
anxiety trait
at day 0
anxiety trait
at day 8
anxiety trait
at day 15
NMR Diffusion-
edited plasma
plasma spectra
Q2(Y) 0.28 NS 0.23
R2(X) 0.29 0.20
R2(Y) 0.85 0.89
NMR Urine
Q2(Y) 0.18 NS NS
R2(X) 0.20
R2(Y) 0.88
MS Plasma data Q2(Y) 0.12 NS 0.22
R2(X) 0.15 0.14
R2(Y) 0.90 0.95
MS Urine data Q2(Y) 0.58 0.16 NS
R2(X) 0.25 0.20
R2(Y) 0.91 0.93
Key: NS, the class separation obtained through the predictive
component was not statistically significant.
Metabolic Effects of Dark Chocolate Consumption research articles
Journal of Proteome Research Vol. 8, No. 12, 2009 5571
of glucose-6-phosphate, cystine and threonic acid (Figure 2).
Additional analyses of the urinary NMR data indicated that the
effects of dark chocolate consumption on urinary excretion of
aromatic compounds were similar in both high and low anxiety
trait subjects (Tables 2 and 4).
Additional chemometric analyses of blood plasma metabolic
profiles did not reveal statistically significant effects of dark
chocolate over the time as assessed by NMR or MS (Table 2).
Anxiety Trait is Associated with a Specific Human Meta-
bolic Signature. Chemometric analysis of urinary NMR and
MS profiles revealed significant compositional changes associ-
ated with self-reported anxiety trait (Table 3, Figures 3 and 4).
For NMR data, analysis of the coefficients plots indicated that
subjects with a higher anxiety trait were characterized with
higher urinary excretion of hippurate, glycine, citrate, and lower
levels of methyl-succinate, trans-aconitate, and a series of
unassigned signals, most likely arising from a polyol, and a
signal at δ1.24 (Table 5, Figure 3). These metabolic changes
were also associated with trends toward higher urine levels
of succinate, lactate and urea, and trends toward lower
urinary excretion of trimethylamine (TMA), and p-cresol
sulfate (Table 5).
For MS data, high anxiety trait subjects showed higher
urinary concentrations of glycine, 3-methoxytyrosine, β-alanine,
proline, 3,4-dihydroxyphenylalanine (DOPA), adrenaline, an
upward trend of lactate, and lower levels of p-cresol sulfate,
aconitate, and a downward trend of arabitol when compared
to low anxiety trait individuals (Table 5, Figure 4). Structure
annotation of p-cresol sulfate metabolite is based on strong
analytical evidence (combinations of chromatography, MS,
chemical reactions, deuterium-labeling, database and literature
search, comparison to similar/homologue/isomeric reference
Analysis of the blood plasma NMR and MS profiles reveal
subtle but significant anxiety trait related metabolic differences
at the baseline and after 2 weeks of dietary intervention (Table
Figure 1. 1H NMR time-dependent metabolic effects of regular dark chocolate consumption O-PLS-DA scores (A) and coefficients plots
(B) for the models discriminating urine samples collected at baseline (day 0) and 1 week (day 8) of treatment derived from O-PLS-DA
of 1H NMR spectra.
research articles Martin et al.
5572 Journal of Proteome Research Vol. 8, No. 12, 2009
3). Interpretation of the NMR coefficients plots indicated that
high anxiety subjects tended to have a relative higher level of
choline, and lower concentrations of glycine and glutamine
compared to low anxiety individuals (Table 5). In addition, after
2 weeks of treatment, high anxiety trait individuals showed
increased concentrations of glutamate and choline, and de-
creased levels of acetate in plasma (Table 5). Analysis of the
S-plots derived from MS data analysis showed increased levels
of lycopene and β-carotene in high anxiety trait subjects at
baseline, and the higher plasma concentration of β-carotene
was still observed after 2 weeks of treatment with dark
chocolate (Table 5). Moreover, at 2 weeks post-treatment,
subjects with a high self-reported anxiety trait also showed
increased levels of normetanephrine (Table 5).
Effects of Dark Chocolate Consumption on Anxiety Trait
Related Metabolism. The anxiety trait-related metabolic dif-
ferences observed in urine were significantly reduced following
1 and 2 weeks intervention with dark chocolate, as noted with
the low/negative value of the Q2(Y) parameter and the loss of
statistically significant differences between the groups (Tables
3 and 5). The NMR and MS signals corresponding to the
influential metabolites identified by chemometrics were ana-
lyzed using a Student ttest (Table 5) and displayed using box-
and-whiskers plots in order to explore their changes overtime
Figure 2. MS time-dependent metabolic effects of regular dark chocolate consumption O-PLS-DA scores (A) and coefficients (B) plots
for the models discriminating urine samples collected at baseline (day 0) and 2 weeks (day 15) of treatment derived from O-PLS-DA
of MS data.
Metabolic Effects of Dark Chocolate Consumption research articles
Journal of Proteome Research Vol. 8, No. 12, 2009 5573
(Figure 5). The changes of hippurate, p-cresol sulfate, glycine,
citrate, trans-aconitate, proline, DOPA, and β-alanine showed
similar trends from high anxiety trait individuals toward low
anxiety trait subjects and further support a normalization of
the metabolic profiles (Table 5). Interestingly, anxiety trait-
related metabolic signatures in blood plasma were maintained
over the duration of the clinical trial, except for lycopene for
which the difference was not significant following the dietary
intervention. In addition, ttests were performed on the
contrasts for the second ((t1-t0)high vs (t1-t0)low) and third
time points ((t1-t0)high vs (t1-t0)low) to assess the metabolic
relationships between anxiety trait and dark chocolate con-
sumption (Supplementary Table 1). The results validated
significant relationships between anxiety trait level, chocolate
consumption and the urinary excretion of 3-methoxytyrosine,
adrenaline, glycine and trans-aconitate, as well as plasma levels
of acetate (Supplementary Table 1).
In the present study, NMR- and MS-based metabolic profil-
ing are shown as complementary techniques to provide a
comprehensive modeling of the biological response of a free
living population to a daily consumption of dark chocolate. The
overview of urine and plasma metabolite profiles revealed that
interindividual differences were greater than intraindividual
variations, which illustrates the strong influence of lifestyle and
genetics on individual metabolic phenotypes. Such metabolic
variations make the study of the metabolic effects of dark
chocolate in free-living subjects difficult when using nonsu-
pervised chemometric methods. Here, metabonomics is ap-
plied to pinpoint modulation of the host and gut microbial
metabolism in response to daily consumption of dark chocolate
with emphasis on stress-associated metabolic changes.
Self-Reported Anxiety Trait is Associated with Specific
Urine and Plasma Metabolic Signatures. In the current
experiment, individuals were classified according to their
dispositional stress as assessed by scoring on the well-validated
anxiety trait scale of the State-Trait Anxiety Inventory (STAI)
The chemometric modeling of metabolite variations in
relation to anxiety trait levels revealed different physiological
processes in the absence of a specific nutritional intervention.
Others have provided evidence that chronic and acute stress
may contribute to the disruption of metabolic homeostasis, and
subsequently to individual susceptibility to diseases.
particular, the individual response to chronic stress is tightly
connected to the hypothalamic-pituitary-adrenal metabolic axis
and the sympathoadrenal system.
Our observations described
systemic changes in hormonal metabolism of high anxiety trait
individuals when compared to low anxiety trait subjects, as
observed by MS with a higher urinary excretion of adrenaline,
DOPA and 3-methoxytyrosine, two intermediates in dopamine
The first step in the catecholamine metabolism is
the hydroxylation of the amino acid tyrosine to DOPA, by the
rate-limiting enzyme in catecholamine biosynthesis tyrosine
hydroxylase. DOPA is then decarboxylated to dopamine which
is the direct precursor to noradrenaline and adrenaline. It is
well-described that physical and mental stress simulates the
release of adrenaline via the sympathetic nervous system and
synthesis of the adrenocorticotropic hormone that enhances
the activity of specific enzymes, including tyrosine hydroxy-
Therefore, the concomitant increased urinary levels of
DOPA, 3-methoxy-tyrosine and adrenaline highlight a greater
synthesis of catecholamines in subjects stratified with high
anxiety trait, with inferred effects on energy metabolism.
The results obtained by NMR and MS also demonstrated a
functional relationship between anxiety trait levels and several
pathways involving the tricarboxylic acid cycle (citrate, succi-
nate, aconitate), gluconeogenetic pathways (lactate), urea cycle
(urea, proline), and protection against oxidative stress (plasma
concentrations of lycopene and β-carotene). In particular, the
anticorrelated variation of citrate and trans-acotinate suggested
additional variations in renal tubular pH and aconitase activ-
Therefore, the observed metabolic changes were consis-
tent with the stress-mediated modulation of gluconeogenesis
by catecholamines.
Moreover, NMR-based metabolic profiling
of urine showed that high anxiety trait individuals tended to
have lower urinary concentrations of polyols, including arabitol
an intermediate in the pentose and glucuronate metabolism,
which may also reflect a modulation of energy metabolism as
a function of dispositional stress.
Nowadays, there is strong evidence that life stress impacts
directly on gastrointestinal function in animals and humans
via modulation of key physiological parameters, such as
intestinal permeability and secretion and release of biological
Changes of gastrointestinal functional ecology
are intimately linked to gut microbial populations and activi-
and abnormal microbiota composition is often observed
in the development of irritable bowel syndromes.
Urine of
mammals contains many polar cometabolites resulting from
gut microbial-mammalian metabolic interactions.
metabolic monitoring of urinary excretion of many aromatic
compounds (e.g., phenolics, indoles and benzoyl derivatives),
methylamines, short chain fatty acids and their hydroxylation
products provides indirect information on the gut microbial
metabolic activities.
For instance, multivariate statistical
modeling of urine and blood plasma indicated a modulation
of choline metabolism, that is, high circulating levels of plasma
choline and low urinary excretion of trimethylamine, copro-
cessed by the gut microbiota from dietary compounds contain-
ing choline.
Moreover, differences in anxiety trait levels were
associated with differential urinary excretion of p-cresol sulfate
and hippurate. These changes reflected different gut microbial
metabolism of aromatic amino acids.
Certain aromatic
compounds, such as benzoate and phenylacetate, that can be
coprocessed by the gut microbiota are well-characterized
glycine and glutamine level reducing agents.
Both NMR-
and MS-based metabolite profiling of urine revealed relatively
higher excretion of glycine in high anxiety trait individuals, with
inferred relationships with amino acid interconversion, and
Table 4. Summary of Time-Dependent Metabolite Effects of
Dark Chocolate Consumption in Urine
day 0
vs 8
day 8
vs 15
day 0
vs 15
4-Hydroxyphenylacetate NMR 0.0000 -0.0033
Adrenaline MS --0.0360
Asparagine MS -0.0455 0.0477
Corticosterone MS --0.0144
Cortisol MS --0.0157
Cystine MS --0.0440
Glucose-6-phosphate MS --0.0488
Normetanephrine MS -0.0212 0.0040
Phenylacetylglutamine NMR --0.0354
p-Cresol sulfate NMR -0.0262 0.0470
Threonic acid MS --0.0809
Key: - designates difference not significant at 95% confidence level.
research articles Martin et al.
5574 Journal of Proteome Research Vol. 8, No. 12, 2009
benzoate metabolism. Additional blood plasma metabolic
variations at baseline and at the end of the study, that is,
reduction of circulating levels of plasma glycine and glutamine/
glutamate, may be functionally related to changes of benzoate
and phenylacetate metabolism in response to bacterial pro-
cessing of dark chocolate.
The Biological Response of Free Living Subjects to a
Daily Consumption of Dark Chocolate was Dependent on
Self-Reported Anxiety Trait. The metabolic response to choco-
late intervention in the whole cohort revealed that a daily intake
of dark chocolate resulted in subtle and cumulative metabolic
effects on the urinary excretion of gut microbial cometabolites
over a 2 weeks period. Increased levels of 4-hydroxyphenylac-
etate and decreased content of phenylacetylglutamine and
p-cresol sulfate reflected the adaptation of gut microbiota to
process dark chocolate content and its active ingredients, such
as phenylethylamine, N-oleoyl- and N-linoleoyl-ethanolamine,
theobromine, flavonoids (epicatechin, catechin and their
In particular, urinary excretion of 4-hydrox-
yphenylacetate and hippurate was previously ascribed to intake
of polyphenols-rich products such as chocolate.
These ob-
servations are therefore complementary to our preliminary
investigations of metabolic signatures associated to chocolate
dietary habits.
Our observations indicated that the metabolic impact of a
daily intake of dark chocolate was strongly dependent on the
dispositional stress state of the individuals, as noted with
statistically significant metabolic effects only in subjects with
inherent high anxiety trait. Consumption of dark chocolate
resulted in the decrease of the levels of catecholamines
(adrenaline, noradrenaline, normetanephrine), corticosterone,
and the stress hormone cortisol in the urine from subjects with
high dispositional stress. Chronic stress is correlated with
increases in stress hormones cortisol and catecholamines,
Figure 3. Stress and anxiety trait metabolic signatures in 1H NMR spectra O-PLS-DA scores (A) and coefficients (B) plots for the models
discriminating urine samples collected at baseline (day 0) according to self-reported anxiety trait derived from O-PLS-DA of 1H NMR
spectra of urine.
Metabolic Effects of Dark Chocolate Consumption research articles
Journal of Proteome Research Vol. 8, No. 12, 2009 5575
and our results suggest potential beneficial implications of dark
chocolate consumption for reduction of mental and/or physical
stress and improvement of the metabolic response to stress.
Moreover, the anxiety trait-related metabolic differences ob-
served in urine (e.g., levels of hippurate, p-cresol sulfate,
glycine, citrate, trans-aconitate, proline, DOPA, and β-alanine)
tended to be normalized toward the levels observed in low
anxiety trait subjects, whereas metabolic signatures in blood
plasma were maintained over the duration of the clinical trial.
Therefore, our observations provided additional evidence that
consumption of dark chocolate may beneficially impact on
stress-associated metabolism as observed through a partial
normalization of stress-related differences in energy metabo-
lism and gut microbial activities.
Our study in free living and healthy humans demonstrates
a link between metabolic phenotype of individuals and
specific dietary patterns. The current observations strongly
support the idea that specific foods impact on human
metabolism through the modulation of gut microbial activi-
ties. The daily consumption of dark chocolate resulted in a
significant modification of the metabolism of healthy and
free living human volunteers with potential long-term
Figure 4. Stress and anxiety trait metabolic signatures in MS profiles O-PLS-DA scores (A) and coefficients (B) plots for the models
discriminating urine samples collected at baseline (day 0) according to self-reported anxiety trait derived from O-PLS-DA of MS profiles
of urine. Key: *Structure annotation is based on strong analytical evidence (combinations of chromatography, mass spectrometry,
chemical reactions, deuterium-labeling, database and literature search, comparison to similar/homologue/isomeric reference com-
research articles Martin et al.
5576 Journal of Proteome Research Vol. 8, No. 12, 2009
consequences on human health within only 2 weeks treat-
ment. This was observable through the reduction of levels
of stress-associated hormones and normalization of the
systemic stress metabolic signatures. Therefore, subtle changes
in dietary habits are likely to modulate the metabolic status
of free-living individuals that might be associated with long-
Table 5. Summary of Time-Dependent Metabolite Differences between High and Low Anxiety Trait Individuals in Urine and
metabolites/p-values measured by biofluids high/low anxiety trait at day 0 high/low anxiety trait at day 8 high/low anxiety trait at day 15
3-Methoxytyrosine MS Urine 0.0080 --
Acetate NMR Plasma --0.0056
Aconitate MS Urine 0.0289 --
Adrenaline MS Urine 0.0448 --
Choline NMR Plasma --0.0260
Citrate NMR Urine 0.0420 --
DOPA MS Urine 0.0356 --
Glutamate NMR Plasma --0.0191
Glycine NMR Urine 0.0130 --
Glycine MS Urine 0.0045 0.0082 -
Hippurate NMR Urine 0.0207 --
Lycopene MS Plasma 0.0350 --
Methylsuccinate NMR Urine 0.0205 0.0169 0.0300
Normetanephrine MS Plasma --0.0046
p-Cresol sulfate MS Urine 0.0070 0.0251 -
Proline MS Urine 0.0304 --
Trimethylamine NMR Urine --0.0160
Trans-Aconitate NMR Urine 0.0100 0.0425 -
β-alanine MS Urine 0.0253 --
β-Carotene MS Plasma 0.0444 0.0333 0.0440
Key: - designates difference not significant at 95% confidence level.
Figure 5. Time-dependent metabolic differences between high and low anxiety trait individuals Comparison of MS signals of plasma
β-carotene, urine p-cresol sulfate and glycine, and area normalized intensities of representative signals of hippurate, trans-aconitate,
and citrate from 1H NMR urine metabolic profiles displayed using box-and-whisker plots. For metabolites identified by 1H NMR
spectroscopy, data are presented as area normalized intensities (a.u.) as follows: 1 ×10-2a.u. for citrate, 1 ×10-4a.u. for trans-
aconitate and 1 ×10-1a.u. for hippurate. Median values are highlighted by dashed and solid lines. Statistical significance of differences
with time and anxiety trait levels is reported in Tables 4 and 5.
Metabolic Effects of Dark Chocolate Consumption research articles
Journal of Proteome Research Vol. 8, No. 12, 2009 5577
term health consequences, in particular via the activity of
the symbiotic bacterial partners.
Abbreviations: CPMG, Carr-Purcell-Meiboom-Gill; MS,
mass spectrometry; NMR, nuclear magnetic resonance; O-PLS-
DA, orthogonal projection to latent structure discriminant
analysis; PCA, principal component analysis; PLS-DA, projec-
tion to latent structure discriminant analysis.
Acknowledgment. We thank TNO Quality of Life,
Zeist (The Netherlands) and Juliet Farrar (Nestle´ Research
Center, Lausanne, Switzerland) for conducting the study.
The authors acknowledge the help and input of Philippe
Guy, Ivan Montoliu Roura, and Nicolas Antille of Nestle´
Research Center, Lausanne, Switzerland.
Supporting Information Available: Supplementary
Figure 1, principal component analysis of biological matrices;
Supplementary Table 1, summary of contrast test statistics to
assess metabolic relationships between anxiety trait levels and
chocolate consumption in urine and plasma. This material is
available free of charge via the Internet at
(1) Nicholson, J. K.; Holmes, E.; Wilson, I. D. Gut microorganisms,
mammalian metabolism and personalized health care. Nat. Rev.
Microbiol. 2005,3(5), 431–438.
(2) Dethlefsen, L.; Fall-Ngai, M.; Relman, D. A. An ecological and
evolutionary perspective on human-microbe mutualism and
disease. Nature 2007,449 (7164), 811–818.
(3) Rezzi, S.; Ramadan, Z.; Martin, F. P.; Fay, L. B.; Bladeren, P. V.;
Lindon, J. C.; Nicholson, J. K.; Kochhar, S. Human Metabolic
Phenotypes Link Directly to Specific Dietary Preferences in Healthy
Individuals. J. Proteome Res. 2007,6(11), 4469–4477.
(4) Blaut, M.; Clavel, T. Metabolic diversity of the intestinal microbiota:
implications for health and disease. J. Nutr. 2007,137 (3), 751S–
(5) Drewnowski, A. Taste preferences and food intake. Annu. Rev.
Nutr. 1997,17, 237–253.
(6) Scalbert, A.; Williamson, G. Dietary intake and bioavailability of
polyphenols. J. Nutr. 2000,130 (8), 2073S–2085S.
(7) Serafini, M.; Bugianesi, R.; Maiani, G.; Valtuena, S.; De, S. S.;
Crozier, A. Plasma antioxidants from chocolate. Nature 2003,424
(6952), 1013.
(8) Grassi, D.; Necozione, S.; Lippi, C.; Croce, G.; Valeri, L.; Pasqualetti,
P.; Desideri, G.; Blumberg, J. B.; Ferri, C. Cocoa reduces blood
pressure and insulin resistance and improves endothelium-
dependent vasodilation in hypertensives. Hypertension 2005,46
(2), 398–405.
(9) Taubert, D.; Roesen, R.; Lehmann, C.; Jung, N.; Schomig, E. Effects
of low habitual cocoa intake on blood pressure and bioactive nitric
oxide: a randomized controlled trial. JAMA, J. Am. Med. Assoc. 2007,
298 (1), 49–60.
(10) Parker, G.; Parker, I.; Brotchie, H. Mood state effects of chocolate.
J Affect. Disord. 2006,92 (2-3), 149–159.
(11) di Tomaso, E.; Beltramo, M.; Piomelli, D. Brain cannabinoids in
chocolate. Nature 1996,382 (6593), 677–678.
(12) McDonald, I. Gordon Holmes Lecture: Gordon Holmes and the
neurological heritage. Brain 2007,130 (1), 288–298.
(13) Holmes, E.; Loo, R. L.; Stamler, J.; Bictash, M.; Yap, I. K.; Chan, Q.;
Ebbels, T.; De, I. M.; Brown, I. J.; Veselkov, K. A.; Daviglus, M. L.;
Kesteloot, H.; Ueshima, H.; Zhao, L.; Nicholson, J. K.; Elliott, P.
Human metabolic phenotype diversity and its association with diet
and blood pressure. Nature 2008,453 (7193), 396–400.
(14) Martin, F. P.; Wang, Y.; Sprenger, N.; Yap, I. K.; Lundstedt, T.; Lek,
P.; Rezzi, S.; Ramadan, Z.; van, B. P.; Fay, L. B.; Kochhar, S.; Lindon,
J. C.; Holmes, E.; Nicholson, J. K. Probiotic modulation of symbiotic
gut microbial-host metabolic interactions in a humanized micro-
biome mouse model. Mol. Syst. Biol. 2008,4, 157.
(15) Martin, F.-P. J.; Wang, Y.; Sprenger, N.; Yap, I. K. S.; Rezzi, S.;
Ramadan, Z.; Pere-Trepat, E.; Rochat, F.; Cherbut, C.; van Bladeren,
P.; Fay, L. B.; Kochhar, S.; Lindon, J. C.; Holmes, E.; Nicholson,
J. K. Top-down systems biology integration of conditional prebiotic
transgenomic interactions in a humanized microbiome mouse
model. Mol. Syst. Biol. 2008,4, 205.
(16) Stella, C.; Beckwith-Hall, B.; Cloarec, O.; Holmes, E.; Lindon, J. C.;
Powell, J.; van der, O. F.; Bingham, S.; Cross, A. J.; Nicholson, J. K.
Susceptibility of human metabolic phenotypes to dietary modula-
tion. J. Proteome Res. 2006,5(10), 2780–2788.
(17) Spielberger, C.; Gorsuch, R.; Lushene, R. Manual for the Self-
Evaluation Questionnaire; Consulting Psychologists Press: Palo
Alto, CA, 1983.
(18) Nicholson, J. K.; Foxall, P. J.; Spraul, M.; Farrant, R. D.; Lindon,
J. C. 750 MHz 1H and 1H-13C NMR spectroscopy of human blood
plasma. Anal. Chem. 1995,67 (5), 793–811.
(19) Roessner, U.; Wagner, C.; Kopka, J.; Trethewey, R. N.; Willmitzer,
L. Technical advance: simultaneous analysis of metabolites in
potato tuber by gas chromatography-mass spectrometry. Plant J.
2000,23 (1), 131–142.
(20) Cook, J. G. Creatinine assay in the presence of protein. Clin. Chim.
Acta 1971,32 (3), 485–486.
(21) Yamada, H.; Yamahara, A.; Yasuda, S.; Abe, M.; Oguri, K.; Fuku-
shima, S.; Ikeda-Wada, S. Dansyl chloride derivatization of meth-
amphetamine: a method with advantages for screening and
analysis of methamphetamine in urine. J. Anal. Toxicol. 2002,26
(1), 17–22.
(22) Wold, S.; Esbensen, K.; Geladi, P. Principal component analysis.
Chemom. Intell. Lab. Syst. 1987,2(1-3), 37–52.
(23) Trygg, J.; Wold, S. Orthogonal projections to latent structures
(OPLS). J. Chemom. 2002,16 (3), 119–128.
(24) Cloarec, O.; Dumas, M. E.; Trygg, J.; Craig, A.; Barton, R. H.; Lindon,
J. C.; Nicholson, J. K.; Holmes, E. Evaluation of the orthogonal
projection on latent structure model limitations caused by chemi-
cal shift variability and improved visualization of biomarker
changes in 1H NMR spectroscopic metabonomic studies. Anal.
Chem. 2005,77 (2), 517–526.
(25) Wiklund, S.; Johansson, E.; Sjostrom, L.; Mellerowicz, E. J.; Edlund,
U.; Shockcor, J. P.; Gottfries, J.; Moritz, T.; Trygg, J. Visualization
of GC/TOF-MS-based metabolomics data for identification of
biochemically interesting compounds using OPLS class models.
Anal. Chem. 2008,80 (1), 115–122.
(26) Gunnar, M.; Quevedo, K. The neurobiology of stress and develop-
ment. Annu. Rev. Psychol. 2007,58, 145–173.
(27) Noble, R. E. Diagnosis of stress. Metabolism 2002,51 (6), 37–39.
(28) Hyland, K.; Surtees, R. A.; Rodeck, C.; Clayton, P. T. Aromatic
L-amino acid decarboxylase deficiency: clinical features, diagnosis,
and treatment of a new inborn error of neurotransmitter amine
synthesis. Neurology 1992,42 (10), 1980–1988.
(29) Eisenhofer, G.; Kopin, I. J.; Goldstein, D. S. Catecholamine
metabolism: a contemporary view with implications for physiology
and medicine. Pharmacol. Rev. 2004,56 (3), 331–349.
(30) Nicholson, J. K.; Higham, D. P.; Timbrell, J. A.; Sadler, P. J.
Quantitative high resolution 1H NMR urinalysis studies on the
biochemical effects of cadmium in the rat. Mol. Pharmacol. 1989,
36 (3), 398–404.
(31) Caso, J. R.; Leza, J. C.; Menchen, L. The effects of physical and
psychological stress on the gastro-intestinal tract: lessons from
animal models. Curr. Mol. Med. 2008,8(4), 299–312.
(32) Santos, J.; Perdue, M. H. Stress and neuroimmune regulation of
gut mucosal function. Gut 2000,47 (4), 49–51.
(33) Tannock, G. W. New perceptions of the gut microbiota: implica-
tions for future research. Gastroenterol. Clin. North Am. 2005,34
(3), 361–382.
(34) Matto, J.; Maunuksela, L.; Kajander, K.; Palva, A.; Korpela, R.;
Kassinen, A.; Saarela, M. Composition and temporal stability of
gastrointestinal microbiota in irritable bowel syndrome--a longi-
tudinal study in IBS and control subjects. FEMS Immunol. Med.
Microbiol. 2005,43 (2), 213–222.
(35) Yap, I. K.; Li, J. V.; Saric, J.; Martin, F. P.; Davies, H.; Wang, Y.;
Wilson, I. D.; Nicholson, J. K.; Utzinger, J.; Marchesi, J. R.; Holmes,
E. Metabonomic and microbiological analysis of the dynamic effect
of vancomycin-induced gut microbiota modification in the mouse.
J. Proteome Res. 2008,7(9), 3718–3728.
(36) Selmer, T.; Andrei, P. I. p-Hydroxyphenylacetate decarboxylase
from Clostridium difficile. A novel glycyl radical enzyme catalysing
the formation of p-cresol. Eur. J. Biochem. 2001,268 (5), 1363–
(37) Zeisel, S. H.; Blusztajn, J. K. Choline and human nutrition. Annu.
Rev. Nutr. 1994,14, 269–296.
(38) Yang, D.; Brunengraber, H. Glutamate, a window on liver inter-
mediary metabolism. J. Nutr. 2000,130 (4), 991S–994S.
(39) Beliveau, G. P.; Brusilow, S. W. Glycine availability limits maximum
hippurate synthesis in growing rats. J. Nutr. 1987,117 (1), 36–41.
(40) Rechner, A. R.; Kuhnle, G.; Bremner, P.; Hubbard, G. P.; Moore,
K. P.; Rice-Evans, C. A. The metabolic fate of dietary polyphenols
in humans. Free Radical Biol. Med. 2002,33 (2), 220–235.
(41) Rios, L. Y.; Gonthier, M. P.; Remesy, C.; Mila, I.; Lapierre, C.;
Lazarus, S. A.; Williamson, G.; Scalbert, A. Chocolate intake
research articles Martin et al.
5578 Journal of Proteome Research Vol. 8, No. 12, 2009
increases urinary excretion of polyphenol-derived phenolic acids
in healthy human subjects. Am. J. Clin. Nutr. 2003,77 (4), 912–
(42) Seematter, G.; Binnert, C.; Tappy, L. Stress and metabolism. Metab
Syndr. Relat. Disord. 2005,3(1), 8–13.
(43) Goldstein, D. S. Catecholamines and stress. Endocr. Regul. 2003,
37 (2), 69–80.
(44) Goldstein, D. S.; Dionne, R.; Sweet, J.; Gracely, R.; Brewer, H. B.,
Jr.; Gregg, R.; Keiser, H. R. Circulatory, plasma catecholamine,
cortisol, lipid, and psychological responses to a real-life stress
(third molar extractions): effects of diazepam sedation and of
inclusion of epinephrine with the local anesthetic. Psychosom. Med.
1982,44 (3), 259–272.
(45) Berk, L. S.; Tan, S. A.; Berk, D. Cortisol and Catecholamine stress
hormone decrease is associated with the behavior of perceptual
anticipation of mirthful laughter. FASEB J. 2008,22 (1), 946.
Metabolic Effects of Dark Chocolate Consumption research articles
Journal of Proteome Research Vol. 8, No. 12, 2009 5579
ResearchGate has not been able to resolve any citations for this publication.
Full-text available
The main dietary sources of polyphenols are reviewed, and the daily intake is calculated for a given diet containing some common fruits, vegetables and beverages. Phenolic acids account for about one third of the total intake and flavonoids account for the remaining two thirds. The most abundant flavonoids in the diet are flavanols (catechins plus proanthocyanidins), anthocyanins and their oxidation products. The main polyphenol dietary sources are fruit and beverages (fruit juice, wine, tea, coffee, chocolate and beer) and, to a lesser extent vegetables, dry legumes and cereals. The total intake is ∼1 g/d. Large uncertainties remain due to the lack of comprehensive data on the content of some of the main polyphenol classes in food. Bioavailability studies in humans are discussed. The maximum concentration in plasma rarely exceeds 1 μM after the consumption of 10–100 mg of a single phenolic compound. However, the total plasma phenol concentration is probably higher due to the presence of metabolites formed in the body's tissues or by the colonic microflora. These metabolites are still largely unknown and not accounted for. Both chemical and biochemical factors that affect the absorption and metabolism of polyphenols are reviewed, with particular emphasis on flavonoid glycosides. A better understanding of these factors is essential to explain the large variations in bioavailability observed among polyphenols and among individuals.
Full-text available
Principal component analysis of a data matrix extracts the dominant patterns in the matrix in terms of a complementary set of score and loading plots. It is the responsibility of the data analyst to formulate the scientific issue at hand in terms of PC projections, PLS regressions, etc. Ask yourself, or the investigator, why the data matrix was collected, and for what purpose the experiments and measurements were made. Specify before the analysis what kinds of patterns you would expect and what you would find exciting.The results of the analysis depend on the scaling of the matrix, which therefore must be specified. Variance scaling, where each variable is scaled to unit variance, can be recommended for general use, provided that almost constant variables are left unscaled. Combining different types of variables warrants blockscaling.In the initial analysis, look for outliers and strong groupings in the plots, indicating that the data matrix perhaps should be “polished” or whether disjoint modeling is the proper course.For plotting purposes, two or three principal components are usually sufficient, but for modeling purposes the number of significant components should be properly determined, e.g. by cross-validation.Use the resulting principal components to guide your continued investigation or chemical experimentation, not as an end in itself.
Full-text available
Gut microbiome-host metabolic interactions affect human health and can be modified by probiotic and prebiotic supplementation. Here, we have assessed the effects of consumption of a combination of probiotics (Lactobacillus paracasei or L. rhamnosus) and two galactosyl-oligosaccharide prebiotics on the symbiotic microbiome-mammalian supersystem using integrative metabolic profiling and modeling of multiple compartments in germ-free mice inoculated with a model of human baby microbiota. We have shown specific impacts of two prebiotics on the microbial populations of HBM mice when co-administered with two probiotics. We observed an increase in the populations of Bifidobacterium longum and B. breve, and a reduction in Clostridium perfringens, which were more marked when combining prebiotics with L. rhamnosus. In turn, these microbial effects were associated with modulation of a range of host metabolic pathways observed via changes in lipid profiles, gluconeogenesis, and amino-acid and methylamine metabolism associated to fermentation of carbohydrates by different bacterial strains. These results provide evidence for the potential use of prebiotics for beneficially modifying the gut microbial balance as well as host energy and lipid homeostasis.
A generic preprocessing method for multivariate data, called orthogonal projections to latent structures (O-PLS), is described. O-PLS removes variation from X (descriptor variables) that is not correlated to Y (property variables, e.g. yield, cost or toxicity). In mathematical terms this is equivalent to removing systematic variation in X that is orthogonal to Y. In an earlier paper, Wold et al. (Chemometrics Intell. Lab. Syst. 1998; 44: 175–185) described orthogonal signal correction (OSC). In this paper a method with the same objective but with different means is described. The proposed O-PLS method analyzes the variation explained in each PLS component. The non-correlated systematic variation in X is removed, making interpretation of the resulting PLS model easier and with the additional benefit that the non-correlated variation itself can be analyzed further. As an example, near-infrared (NIR) reflectance spectra of wood chips were analyzed. Applying O-PLS resulted in reduced model complexity with preserved prediction ability, effective removal of non-correlated variation in X and, not least, improved interpretational ability of both correlated and non-correlated variation in the NIR spectra. Copyright © 2002 John Wiley & Sons, Ltd.
The effects of the antibiotic vancomycin (2 x 100 mg/kg/day) on the gut microbiota of female mice (outbred NMRI strain) were studied, in order to assess the relative contribution of the gut microbiome to host metabolism. The host's metabolic phenotype was characterized using (1)H NMR spectroscopy of urine and fecal extract samples. Time-course changes in the gut microbiotal community after administration of vancomycin were monitored using 16S rRNA gene PCR and denaturing gradient gel electrophoresis (PCR-DGGE) analysis and showed a strong effect on several species, mostly within the Firmicutes. Vancomycin treatment was associated with fecal excretion of uracil, amino acids and short chain fatty acids (SCFAs), highlighting the contribution of the gut microbiota to the production and metabolism of these dietary compounds. Clear differences in gut microbial communities between control and antibiotic-treated mice were observed in the current study. Reduced urinary excretion of gut microbial co-metabolites phenylacetylglycine and hippurate was also observed. Regression of urinary hippurate and phenylacetylglycine concentrations against the fecal metabolite profile showed a strong association between these urinary metabolites and a wide range of fecal metabolites, including amino acids and SCFAs. Fecal choline was inversely correlated with urinary hippurate. Metabolic profiling, coupled with the metagenomic study of this antibiotic model, illustrates the close inter-relationship between the host and microbial "metabotypes", and will provide a basis for further experiments probing the understanding of the microbial-mammalian metabolic axis.
We report the clinical features, biochemical details, and treatment of the first detected cases of an inborn error of aromatic L-amino acid decarboxylase. Male monozygotic twins presented with extreme hypotonia and oculogyric crises. Concentrations of biogenic amines and their metabolites were reduced considerably both centrally and peripherally. Pterin and phenylalanine metabolism were normal. Activity of aromatic L-amino acid decarboxylase was virtually absent in a liver biopsy sample and greatly reduced in plasma. Concentrations of L-dopa, 3-methoxytyrosine, and 5-hydroxytryptophan were elevated in CSF, plasma, and urine. CSF S-adenosylmethionine concentrations were reduced. Pyridoxine treatment had no clinical effect but led to a fall in CSF L-dopa and 3-methoxytyrosine and a rise in S-adenosylmethionine. Treatment with either bromocriptine or tranylcypromine stopped the abnormal eye movements; tranylcypromine treatment also improved muscle tone and led to a rise in plasma norepinephrine and whole blood serotonin. Combined treatment with pyridoxine, bromocriptine, and tranylcypromine produced sustained improvement in tone and voluntary movements. The twins' parents were asymptomatic but had reduced plasma aromatic L-amino acid decarboxylase activity, consistent with heterozygosity. We monitored a subsequent pregnancy through biochemical analyses of a fetal liver biopsy sample and of amniotic fluid. We predicted an unaffected fetus, which was confirmed clinically and biochemically after birth.
Quantitative changes in the urinary excretion patterns of low molecular weight compounds were followed for up to 30 days after dosing of adult Sprague-Dawley rats with single intraperitoneal injections of CdCl2 (6-24 mumol/kg), using high resolution 1H NMR multicomponent urinalysis. There was a marked reduction in the rate of urinary excretion of citrate, 2-oxoglutarate, and succinate within 4.5 hr of the administration of 24 mumol/kg Cd2+. This continued for up to 4 days after dosing in male rats and was consistent with a renal tubular acidosis, caused by inhibition of carbonic anhydrase. Histological examination of the kidneys showed no evidence of structural abnormalities at any Cd2+ dose level. Creatinine excretion was not affected by Cd2+ treatment at any dose level but hippurate excretion was significantly reduced. Severe testicular damage was noted within 24 hr of Cd2+ treatment at doses of greater than 9 mumol/kg and the degree of damage appeared to be correlated with the presence of large amounts of creatine (up to 20 mM) in the urine. Analysis of homogenates of healthy testicular material indicated the presence of high concentrations of free creatine. Cadmium-induced creatinuria appears to result from direct release of creatine from the necrotic cells of the seminiferous tubules and, hence, the measurement of creatine excretion rates may provide a useful noninvasive indicator of testicular necrosis. Because NMR is nonselective in terms of metabolite detection, this work has shed new light on the changes in urinary composition arising from Cd toxicity. As such, the technique is potentially very valuable in the search for new metabolic markers of toxicity and organ dysfunction.
The rate-limiting step for hippurate synthesis from sodium benzoate (NaB) was investigated in growing rats. Rats fed a glycine-and serine-free L-amino acid diet were injected daily with saline or varying doses of NaB. Growth was monitored for 4 d, after which time rats were killed and livers were assayed for glycine, serine, benzoyl-CoA, benzoyl-CoA ligase and glycine benzoyl-CoA transferase. In control animals, liver glycine and serine concentrations were 2.18 and 1.63 mumol/g wet wt, respectively; benzoyl-CoA was not detectable. In rats injected with 600 mg NaB/kg body wt per day, liver glycine and serine concentrations decreased to 68 and 78% of control values, respectively, and benzoyl-CoA accumulated (0.52 mumol/g wet wt). As the dose of NaB was increased, liver benzoyl-CoA concentration increased, and growth of the rats was markedly impaired. The activities of benzoyl-CoA ligase and glycine benzoyl-CoA transferase were unchanged after NaB treatment. In a second experiment, rats were treated with growth-impairing doses of NaB. When the diet was supplemented with serine and glycine, growth was normalized, liver glycine and serine concentrations returned to control levels, and benzoyl-CoA accumulation was markedly diminished. These results suggest that glycine availability limits maximum hippurate synthesis in vivo.