Differential contributions of theobromine and caffeine on mood, psychomotor performance and blood pressure

Abstract

The combination of theobromine and caffeine, methylxanthines found in chocolate, has previously been shown to improve mood and cognition. However, it is unknown whether these molecules act synergistically. This study tested the hypothesis that a combination of caffeine and theobromine has synergistic effects on cognition, mood and blood pressure in 24 healthy female subjects. The effects of theobromine (700 mg), caffeine (120 mg) or the combination of both, or placebo were tested on mood (the Bond-Lader visual analog scale), psychomotor performance (the Digit Symbol Substitution Test (DSST)) and blood pressure before and at 1, 2 and 3 h after administration. Theobromine alone decreased self-reported calmness 3h after ingestion and lowered blood pressure relative to placebo 1 h after ingestion. Caffeine increased self-reported alertness 1, 2 and 3h after ingestion and contentedness 1 and 2 h after ingestion, and increased blood pressure relative to placebo (at 1 h). The combination of caffeine+theobromine had similar effects as caffeine alone on mood, but with no effect on blood pressure. There was no treatment effect on DSST performance. Together these results suggest that theobromine and caffeine could have differential effects on mood and blood pressure. It was tentatively concluded that caffeine may have more CNS-mediated effects on alertness, while theobromine may be acting primarily via peripheral physiological changes.

Full text

Differential contributions of theobromine and caffeine on mood, psychomotor
performance and blood pressure
E.S. Mitchell
a,
⁎, M. Slettenaar
a
, N. vd Meer
a
, C. Transler
a
, L. Jans
b
, F. Quadt
a
, M. Berry
c
a
Unilever R&D, Olivier van Noortlaan 120, Vlaardingen 3133 AT, Netherlands
b
Department of Psychology, University of Leiden, Netherlands
c
Unilever R&D, Colworth House, Sharnbrook, Bedford, UK
abstractarticle info
Article history:
Received 28 February 2011
Received in revised form 7 June 2011
Accepted 27 July 2011
Keywords:
Stimulants
Chocolate
Mood
Cognition
Methylxanthine
Cocoa
Food
The combination of theobromine and caffeine, methylxanthines found in chocolate, has previously been
shown to improve mood and cognition. However, it is unknown whether these molecules act synergistically.
This study tested the hypothesis that a combination of caffeine and theobromine has synergistic effects on
cognition, mood and blood pressure in 24 healthy female subjects. The effects of theobromine (700 mg),
caffeine (120 mg) or the combination of both, or placebo were tested on mood (the Bond–Lader visual analog
scale), psychomotor performance (the Digit Symbol Substitution Test (DSST)) and blood pressure before and
at 1, 2 and 3 h after administration. Theobromine alone decreased self-reported calmness 3 h after ingestion
and lowered blood pressure relative to placebo 1 h after ingestion. Caffeine increased self-reported alertness
1, 2 and 3 h after ingestion and contentedness 1 and 2 h after ingestion, and increased blood pressure relative
to placebo (at 1 h). The combination of caffeine +theobromine had similar effects as caffeine alone on mood,
but with no effect on blood pressure. There was no treatment effect on DSST performance. Together these
results suggest that theobromine and caffeine could have differential effects on mood and blood pressure. It
was tentatively concluded that caffeine may have more CNS-mediated effects on alertness, while
theobromine may be acting primarily via peripheral physiological changes.
© 2011 Elsevier Inc. All rights reserved.
1. Introduction
Chocolate has long been associated with happiness and improved
mood. Although in most cases the sensory aspects of chocolate are most
likely responsible for its high consumption, there is evidence that psycho-
active ingredients in chocolate may play a part as well [1–6].Ofthe
various compounds that are present in chocolate, the methylxanthines
theobromine and caffeine are known to have psychoactive effects [1].
Caffeine, a non-specific adenosine receptor antagonist, is a well-
known psychostimulant which peaks in the blood 30–40 min after
ingestion of a 72 mg dose [7]. Moderate doses of caffeine (100–150 mg)
increase subjective alertness, stimulation and vigor, improve reaction
time in psychomotor tasks and also increase blood pressure [8,9].
However,reported mood andcognition effectsfrom caffeine maybe due
to withdrawal, since subjects whose daily intake of caffeine was over
100 mg report more negative mood after prolonged deprivation [10].On
the other hand, light caffeine users, who exhibit few withdrawal effects,
display improved mood and increased alertness after consumption of
caffeine [8,9,11,12]. Enhanced subjective alertness from caffeine, the
most often reported behavioral effect, is usually perceived between 30
and 60 min for a moderate dose of 150 mg, alertness plateaus for 1–2h
and then dissipates by 4 h, usually no negative symptoms are reported
at a 150 mg dose [9,11,12].
Theobromine is a caffeine derivative and metabolite found primarily
in chocolate; it is highly fat soluble; peaking in the plasma 1–2 h after
ingestion [7]. An adenosine receptor antagonist, theobromine appears
to have equal affinity for A1 compared to A2A receptors, while caffeine
shows a slightly lower affinity for A1 receptors [13]. Theobromine has
one fifth the stimulant effectof caffeine, but with a longer half-life in the
body [7].
Very few studies have investigated the behavioral effects of
theobromine, and thus no clear conclusions can yet be made about
its psychoactive profile. Although two early reports found null mood
effects from theobromine [14,15], a more recent study by Mumford et
al. [16] showed that 5 out of 7 subjects were able to discriminate a
high dose of theobromine (560 mg) from a placebo or caffeine dose.
The combination of caffeine (19 mg) and theobromine (250 mg) in
capsules increased the self-reported mood construct “energetic
arousal”, and improved cognitive function as measured with a simple
reaction time test [1] compared to placebo capsules. Furthermore, the
combination of theobromine and caffeine added to novel-flavored
drinks increased liking over time in healthy volunteers, indicating that
the psychopharmacological effects of these methylxanthines may play
Physiology & Behavior 104 (2011) 816–822
⁎Corresponding author at: Unilever R&D Vlaardingen, Olivier van Noortlaan 120,
3133 AT Vlaardingen, Netherlands. Tel.: +31 10 4605578; fax: + 31 10 4605993.
E-mail address: Siobhan.mitchell@unilever.com (E.S. Mitchell).
0031-9384/$ –see front matter © 2011 Elsevier Inc. All rights reserved.
doi:10.1016/j.physbeh.2011.07.027
Contents lists available at ScienceDirect
Physiology & Behavior
journal homepage: www.elsevier.com/locate/phb
Downloaded for Anonymous User (n/a) at University of Washington - Seattle - WSC from ClinicalKey.com by Elsevier on April 08, 2019.
For personal use only. No other uses without permission. Copyright ©2019. Elsevier Inc. All rights reserved.

a role in liking of chocolate [5]. Theobromine lowers blood pressure
[17], and is also a smooth muscle relaxant and diuretic [18].
This study investigated whether theobromine and caffeine have a
synergistic action on mood, attention, and blood pressure. There is
roughly a 1:5 ratio of caffeine to theobromine in chocolate [18], thus
100 g of dark chocolate contains 120–150 mg of caffeine and
700–800 mg theobromine, similar to the doses used in this study.
We hypothesized that theobromine combined with caffeine may
have synergistic effects on some well-known caffeine symptoms such
as increased arousal and psychomotor speed, but neutral effects on
blood pressure. To measure psychomotor effects, we used the Digit
Symbol Substitution Test (DSST), where there have been previous
reports that 150 mg doses of caffeine improved performance [11,12]
but lower doses of 100 mg did not [19,20]. We were also interested in
how the blood pressure modulation from caffeine and theobromine
influences mood or perceived physiological symptoms such as
headache incidence. Lastly, given the paucity of data on theobromine
effects on mood, we used ‘implicit’mood tests such as subject rating of
the ‘pleasantness’of pre-rated words in order to capture subtle
positive biasing effects. Tests assessing implicit mood were used for
exploratory purposes only.
2. Materials and methods
2.1. Subjects and randomization
Twenty-four healthy female subjects aged 18–70 and with a BMI
ranging from 20 to 30 were recruited via fliers sent to inhabitants of
the Rotterdam region. To be included in the study subjects had to
report a daily consumption of at least one caffeinated product (coffee,
tea, or caffeinated soda). Exclusion criteria included the presence of a
medical condition, pregnancy, breastfeeding, night shift work,
excessive exercise, alcohol consumption of more than 21 units per
week or blood pressure above 160/90 mm Hg. After screening,
subjects were randomized into treatment groups. The study and the
protocol were granted by the Medical Ethical Committee of the
Wageningen University. The study was conducted according the
Declaration of Helsinki and the Medical Research Involving Human
Subjects Act (WMO). The study was registered as a basic science
clinical trial with clinical trials.gov (study # NCT01288547).
The study was explained to the volunteers as an intervention to
test how chocolate components affect cognition, and that they were
free to withdraw from the study at any moment without the need for
explanation. Written, informed consent was obtained from all
participants in the study.
Subjects were encouraged to minimize changes in lifestyle and
composition of their habitual diet during the entire study period. They
were instructed not to exercise more or less than what they were used
to. Compliance to background diet was checked by a dietician using a
form wherein subjects had to report on the first measurement day
what they had eaten and/or drank, how much and at what time.
Subjects who completed the study received a reward of €180.
2.2. Study design
This study had a placebo controlled; double blind, randomized,
cross-over design. All subjects received 4 different treatments
contained in capsules: 700 mg theobromine, 120 mg caffeine,
700 mg theobromine+ 120 mg caffeine or placebo (cellulose). Sub-
jects abstained from caffeine and chocolate overnight and, after
breakfast, consumed only food provided to them by the study director
(lunch and snack), until the end of the last test session of each
measurement day. Between measurement days there was a 1 week
washout period to ensure there would be no effect of the previous
treatment.
On each of the measurement days, subjects needed to consume the
same standardized breakfast and standardized light lunch (less than
500 calories), to ensure similar background levels of various nutrients
which may affect cognition. These meals were provided to them and
had to be consumed at home, before coming to the testing location.
They had to consume the standardized breakfast before 8:00 am and
the standardized lunch at noon. On measurement days, they were also
provided a snack (at the end of test session 2). Furthermore, they
were free to drink water.
The day before measurement subjects were restricted to have a
maximum of two alcohol units. Additionally, the subjects were asked
to forgo all caffeine-containing or theobromine-containing foods from
the start of the measurement day (0:00–midnight) until the end of the
measurement day (again 0:00–midnight). Subjects were allowed to
consume their normal amount of caffeine on non-measurement days.
Subjects were checked for caffeine use with saliva sampling (i.e.
chewing on cotton swabs for 1 min, which were then analyzed for
caffeine content).
On every measurement day, mood and cognition tests, plus blood
pressure were taken at baseline, and then every 60 min during each of
the three test sessions (except for blood pressure which was only
taken at session 1 and 3), within strict margins after consumption of
the test product for 3 h (see Table 1 for a summary of procedures).
2.3. Treatments
The intervention of this trial consisted of four treatments, one for
each measurement day, consumed by all 24 subjects.
Treatment 1: placebo treatment, consisting of three capsules filled
with Avicel PH-102 (micro-crystalline cellulose), Ph
Eur (FMC)
Treatment 2: theobromine treatment, consisting of two capsules
filled with 350 mg theobromine, Ph Eur (Fagron, the
Netherlands) and one placebo capsule
Treatment 3: caffeine treatment, consisting of one capsule filled with
120 mg caffeine, Ph Eur (Fagron, the Netherlands) and 2
placebo capsules
Treatment 4: combined theobromine and caffeine treatment, con-
sisting of 2 theobromine capsules and 1 caffeine
capsule.
At each test day after the baseline measurements, every subject
consumed 1 of the 4 test treatments (i.e. 3 capsules), provided in a
random order (treatments pre-labeled with subject code and testing
day). They were instructed to consume the capsules with 250 ml of
water. The different capsules contained an active ingredient (caffeine
or theobromine) and a filler material (microcrystalline cellulose) or
microcrystalline cellulose only (placebo).
The products were prepared at Unilever Vlaardingen according to
GMP and standard HACCP prerequisites. A statistician allocated the
personal codes (pcodes) (1–24) to treatment orders as defined by a
within cohort Williams design (which consists of four Latin squares).
2.4. Behavioral measurements
All questionnaires and tasks were performed on computers using
E-Prime software.
The Bond–Lader Questionnaire was used to measure self-reported
alertness, calmness and contentment [21] using the summed scores
from 16 questions. The Bond–Lader summed score demonstrates the
intensity of a specificoverallemotionbyaskingsubjectstoratethe
intensityof an emotion on a 100 mm line anchored by a specific emotion
on one end and its antonym on the opposite end (thus a rating of 50 is
considered neutral). The alertness summary compiles data from the
intensity of these nine sensations or their antonyms: alert, strong, well-
coordinated, energetic, quick-witted, attentive, proficient, interested
817E.S. Mitchell et al. / Physiology & Behavior 104 (2011) 816–822
Downloaded for Anonymous User (n/a) at University of Washington - Seattle - WSC from ClinicalKey.com by Elsevier on April 08, 2019.
For personal use only. No other uses without permission. Copyright ©2019. Elsevier Inc. All rights reserved.

and clear-headed. Calmness was assessed by the average rating of two
sensations: calmness and relaxation. The contentedness subscale is
measured bycombining the average scores on the following sensations:
contented, tranquil, happy, amicable and gregarious.
A caffeine negative symptom questionnaire, adapted from Rogers et
al. [22] and consisting of 10 questions, was used to assess the negative
physiological sensations from caffeine consumption. Subjects were
asked to rate the intensity of a symptom using a line anchored by “Ihave
this feeling strongly”on one end and “notatall”at the opposite end,
using these descriptors: head-ache, heart pounding, jittery, shaky, light-
headed, feeling faint, dizzy, hands trembling, scared, feeling hot/sweaty
(not due to heat).
The Digit Symbol Substitution Task (DSST) is a psychomotor and
working memory task where subjects must decode a series of digits
into symbols [23]. The computer adapted version of this task required
subjects to decode as many digits as possible in 90 s using a table of
symbols and their corresponding codes at the top of the screen. Two
rows of squares are displayed where the upper row squares contain a
simple symbol (the same used in written version). In the second row,
each square contains a digit, from one to nine. Under the first set of
rows is a similar table filled with the same symbols in random order
with empty squares below each symbol. During the task, the symbols
are highlighted. Thesubject is instructed to type the digit corresponding
to the highlighted symbol, according to the template above. Only digits
from one to nine are accepted in order to proceed to the next square.
When the subject completes a table, a new table appears and this is
repeated until 90 s have passed. This computerized DSST has been
shown to have high correlation to written DSST performance [24].
Subjects were given a practice task to complete during recruitment and
before baseline testing on each measurement day. The total score for
performance of the DSST is the sum of all correct responses. Two
parameters were analyzed: mean reaction time, in milliseconds, of
correct responses and accuracy—number of correct answers minus
number of incorrect answers.
The Motivation and Workload Questionnaire and Emotive Reac-
tion Time Test (ERTT) were used as implicit mood measures.
The Motivation and Workload Questionnaire [25], was used in this
experiment to assess mood and motivation during cognitively demand-
ing activities like timed computer tests, with questions how difficult
tasks were perceived and enjoyment of tasks. It is based on a visual
analog scale with 14 items, to measure how motivated the subjects were;
and how much workload they experienced while performing the tests.
In the Emotive Reaction Time Test (ERTT) subjects must categorize a
series of words as quickly as possible [26]. Twenty words with pre-rated
positive, neutral or negative salience are presented on the screen at
random intervals (2, 4,6, 10 s). Subjects designatethe affective salience
by pressing the 5 or 4 key for positive salience, the 3 key for neutral
salience or the 2 and1 keys for negative salience. A new series of words
were used for each test (16 tests in all for each subject). The words were
validated for emotional rating in a Dutch population previously [27].
Each series consisted of the same number of negative, neutral and
positive words as validated previously, and each series contained words
with similar average emotional valence, syllable number and common
usage. The difference in each subject rating of a word from the validated
salience may indicate possible positive or negative affect. Subject rating
of a specific positive, neutral or negative word was subtracted against
previously validated emotional valence rating for the same word, and
the average changes in pleasantness rating of positive, neutral or
negative words were plotted per group. Variations of emotional rating
tasks havebeen used to demonstrate changesin emotional valencefrom
acute antidepressant administration [28,29].
Subjects performed these tests every testing session day, at the
same times each day. Please see Table 1 for exact timings of each test.
2.5. Caffeine saliva swab measurement
The saliva swab test was performed immediately before subjects
began the intervention. The subject chewed on a cotton swab for a
minute and then the swab was placed in a centrifuge tube and
centrifuged at 3000 rpm for 15 min. The samples werestored at −20 °C,
and subsequently measured via HPLCusing an Agilent1100 system with
fluorescence. Saliva swabs were analyzed before the final statistical
analyses of the study data. The criterion to exclude persons for the per
protocol analysis based on the swab test, was if the results of the person
are above 2 mg caffeine.
2.6. Data analysis
Sample size was determined via power analysis of data from a
previous study on mood effects from theobromine and caffeine [1].A
minimum of 24 subjects was necessary to detect a critical difference of
15% between treatment groups with a power of 0.8 and alpha= 0.0125
(1-sided).
The data was analyzed using SAS version 9.1 software. The
analyses were calculated via 2 × 2 ×4× 3 ANOVA including the fixed
factors Caffeine (present, absent), Theobromine (present, absent),
Visit and Session, and also as repeated measures ANOVAs using the
initial baseline data of each session as a covariate, and with Subjects as
a random factor. A Tukey–Kramer adjustment has been used for
multiple comparisons.
The Motivation and Workload Questionnaire was given at one
time point only and thus was only calculated using a 2× 2 × 4 ANOVA
including visit. Only significant effects are reported below.
3. Results
3.1. Subject characteristics
Twenty four non-smoking females, all Dutch natives, started and
completed the study. The mean (±SD) age was 51.1 ±12.7 years. The
Table 1
Description of measurement day.
Time
(minutes)
Activity
T=−90 min Caffeine-abstinent subjects consumed the standardized light lunch, containing less than 500 calories, at home.
T=−35 min A saliva swab test was used to check for prior caffeine consumption.
T=−25 min Baseline: subjects were given questionnaires on explicit mood (Bond–Lader visual analog scales. Subjects also completed 2 attention tasks: the Digit Symbol
Substitution Task (DSST), and the Emotive Reaction Time Test (ERTT). Baseline blood pressure was measured.
T=0 Treatment: subjects consumed one of the four treatments with a glass of water.
T=60 min Session 1: subjects conducted tests on mood and cognition. A separate test on implicit mood was given only at this time point (before all other mood tests).
Furthermore, blood pressure measures were taken following the cognition tests. The test session lasted approximately 20–25 min.
T=120 min Session 2: subjects conducted tests on mood and cognition (see above tests). After completing the test session, 2 snacks were provided to the subjects.
T =180 min Sess ion 3: subjects cond ucted tests on mood and c ognition, and blood p ressure was measured (se e above tests). A quest ionnaire on workloa d and perceived
performance was also given only at this time.
When not being tested, subjects were allowed to do quiet activities in the study lounge, e.g. reading (not work related), knitting, etc. Subjects could not use a cell phone, the internet
or watch TV. The subject was allowed to leave immediately after completing the third session.
818 E.S. Mitchell et al. / Physiology & Behavior 104 (2011) 816–822
Downloaded for Anonymous User (n/a) at University of Washington - Seattle - WSC from ClinicalKey.com by Elsevier on April 08, 2019.
For personal use only. No other uses without permission. Copyright ©2019. Elsevier Inc. All rights reserved.

subjects had a mean BMI of 24.2± 2.1 kg/m
2
. All subjects were
divided into 6 cohorts of 4 participants each. One subjects' data for one
visit was excluded for per protocol statistical analyses, because of
illness unrelated to treatment.
Mean± SD baseline salivary caffeine concentrations for placebo,
caffeine, theobromine, and theobromine + caff eine groups were:
0.448± 0.087, 0.709 ± 0.185, 0.677 ± 0.130 and 0.384±0.083 μg/ml,
respectively; there was no significant difference between groups. The
average caffeine intake of the subjects was 420 mg per day. The
average systolic and diastolic blood pressure during screening was
110 and/67 mm Hg, respectively (see Supplementary Table 1 for
further details).
3.2. Behavior
Due to the differential pharmacokinetics of caffeine and theobro-
mine we expected to observe temporal changes in mood, specifically
caffeine showing more effects at earlier time points and theobromine
modulating mood at later time points.
Caffeine significantly increased self-reported alertness (Fig.1A)
measured 1, 2, and 3 h after baseline [1 h F(1,55)= 9.73, p = 0.0029;
2 h F(1,55)= 13.46, p = 0.0006; 3 h F(1,54) =7.93, p =0.0068].
Theobromine decreased calmness (Fig. 1B) at the 3 h time point
only [F(1,54)= 4.77, p =0.0333]. Caffeine also significantly increased
contentedness (Fig.1C) measured 1 h [F(1,55) = 5.03, p = 0.0289))
and 2 h [F(1,55= 5.49,pb0.0228)] after consumption.
On the Motivation and Workload Questionnaire subjects treated with
theobromine alone or caffeinealone rated tasks as more interesting than
those subjects who were given placebo [F(1,56)= 7.28, p= 0.0092]; [F
(1,56) =6.95, p =0.0 108]; respectively; Fig. 2A). Caffeine also increased
eagerness to do tasks [F(1,56)=5.92,p= 0.0182] (Fig.2B). There were
no significant effects on any of the other items of the Motivation and
Workload Questionnaire.
There were no significant main effects of caffeine or theobromine
for the DSST for the total number of correct responses, or mean
response time for correct answers. The mean number of incorrect
responses per test was less than 1, and there were no differences
between the groups. However, there was a significant interaction of
theobromine and caffeine in mean response time for correct answers
[F(1,16) = 5.9, p b0.018] 3 h after consumption (mean correct
response times, placebo: 2110± 230 ms; caffeine: 2435 ±230 ms;
theobromine: 2344 ± 120 ms; caffeine+ theobromine: 2310 ±
110 ms). Specifically, the combination of theobromine and caffeine
increased mean response time compared to theobromine alone and
decreased it compared to caffeine alone.
For the ERTT 3 h after treatment the average rating of all words [F
(1,55)= 4.06, p= 0.0487]were decreased by caffeine (data not shown).
Fig. 1. The combination of caffeine+ theobromine or caffeine alone increased feelings of
alertness in Bond–Lader subscales. Theobromine alone decreased calmness and caffeine
alone increases contentment. A. Alertness SUM. B. Calmness SUM. C. Contentment SUM.
Placebo: white bars, theobromine: light gray bars, caffeine: black bars, caffeine+
theobromine: striped bars. Results shown are means and standard errors adjusted for
baseline. *pb0.05 as compared to control.
Fig. 2. Theobromine and caffeine increased the interest in doing tasks while caffeine
also increased eagerness to do tasks. A. Interest in a task as rated on a 0–9pointscale.
B. Eagerness to do a task as rated on a 0–9 point scale. Results shown are means and
standard errors adjusted for baseline. *p b0.05, **p b0.01 as compared to control.
819E.S. Mitchell et al. / Physiology & Behavior 104 (2011) 816–822
Downloaded for Anonymous User (n/a) at University of Washington - Seattle - WSC from ClinicalKey.com by Elsevier on April 08, 2019.
For personal use only. No other uses without permission. Copyright ©2019. Elsevier Inc. All rights reserved.

There wereno effects of caffeine or theobromine onthe average ratingof
positive or neutral words. However, the average rating of negative
words was decreased by caffeine 1 h [F(1,54)=7.20, p =0.0096] and
3 h after consumption [F(1,54= 8.28, p =0.0057]; but increased by
theobromi ne 3 h afte r consumption [F(1 ,54) = 6.65, p = 0.0126]
(Fig. 3). There were no effects on the average response time for positive
words. Furthermore, the average response time for negative words was
decreased by caffeine at the 3 h time point [F(1,55)=6.92, p= 0.0110]
(data not shown).
3.3. Physiological symptoms
Negative physiological symptoms are often related to high caffeine
consumption, but not often with low caffeine consumption. Therefore
we asked if there is increased perception of physiological symptoms
when caffeine is combined with theobromine via self-report
questionnaire.
Separate ANOVAs per time point revealed that caffeine significantly
decreased head-ache rating (Fig. 4A) measured at all time points [1 h F
(1,56)= 7.75, p = 0.0093, 2 h F(1,56)= 16.96, p =0.0001, 3 h F
(1,56)= 12.82, p=0.0007], while theobromine decreased head-ache
rating at 1 h after consumption only [F(1,56)= 7.56, p=0.008] and
decreased palpitations at the 2 h time point [F(1,56)= 5.61,p = 0.0213]
(data not shown).
Additionally, there was a caffeine-mediated decrease in light-
headedness in all time points [1 h F(1,56)=9.86, p= 0.0027, 2 h F
(1,56) = 8 .26, p = 0.00 57, 3 h F (1,56) = 7.78 , p = 0.0072 ] a nd a de-
crease in weakness [1 h F(1,56)= 11.28, p= 0.0014, 2 h F(1,56)= 6.70,
Fig. 3. Emotional Reaction Time Test (ERTT): effects of caffeine, theobromine, caffeine
+theobromine, and placebo on the rating of negative words. Rating on 1–5 scale.
Shown are means and standard error. *p b0.05 significant difference compared with
control.
Fig. 4. The sum of negative symptoms and headache rating aftercaffeine, theobromineor
the combination of both. A. Caffeineand theobromine reduced head-ache symptoms on a
scale of 1–100mm. B. Caffeine and theobromine decreased overall negative physiological
symptoms rated on a 0–9 point scale. Results shown are means and standard errors
adjusted for baseline. *pb0.05 control vs. caffeine.
#
Control vs. theobromine.
Fig. 5. Effects of caffeine and theobromine on heart rate and systolic blood pressure.
A. Systolic blood pressure (mm Hg). B. Diastolic blood pressure (mm Hg). C. Heart
rate (beats per minute). Results shown are means and standard errors. *p b0.05
control vs. caffeine.
#
Control vs. theobromine.
820 E.S. Mitchell et al. / Physiology & Behavior 104 (2011) 816–822
Downloaded for Anonymous User (n/a) at University of Washington - Seattle - WSC from ClinicalKey.com by Elsevier on April 08, 2019.
For personal use only. No other uses without permission. Copyright ©2019. Elsevier Inc. All rights reserved.

p=0.0122, 3 h F(1,56)=7.91, p= 0.0068] and a decrease in trembling
1 h after consumption F(1,56=11.21, p= 0.0015] (data not shown).
There was one significant interaction between theobromine and
caffeine [F(1,56= 4.40, p=0.0405], resulting in decreased trembling
as compared to either methylxanthine alone at the 1 h time point (data
not shown). The SUM of caffeine symptoms (head-ache, jitteriness,
weakness, light-headedness, warmth, fearfulness and trembling hands)
revealed significant differences between caffeine vs. placebo during all
sessions [1 h F(1,56)= 9.07, p=0.0039, 2 h F(1,56)= 9.58, p = 0.0031,
3 h F(1,56)= 8.55, p=0.005] (Fig. 4B).
Blood pressure and heart rate were taken to assess whether
treatment-induced cardiovascular changes had noticeable effects on
perceived physiological symptoms. Blood pressure is a well-charac-
terized indicator of autonomic system activation. In all intervention
groups there was a tendency to an increase in BP over time, possibly
related to the morning surge, as baseline measurements were done
early in the morning. Compared to placebo caffeine significantly
increased diastolic blood pressure 1 h [F(1,56) = 19.93, p =0.0001]
and 3 h after consumption [F(1,56)= 4.30, p = 0.0424,] (Fig. 5A).
Systolic blood pressure decreased 1 h after theobromine treatment [F
(1,56)= 7.37, p= 0.0087] and increased with caffeine treatment 1 h
later [F(1,56)= 7.01, p = 0.0104] (Fig. 5B). Conversely, at all time
points theobromine increased heart rate [1 h F(1,56) = 11.17,
p= 0.0015; 3 h F(1,56) =12.75, p = 0.0007] while caffeine decreased
it [F(1,56) =18.23, p =0.0001] 1 h after consumption (Fig. 5C).
4. Discussion
This study has demonstrated behavioral and cardiovascular effects
of two methylxanthines found in chocolate, theobromine and caffeine.
Similar to previous reports, caffeine increased alertness and blood
pressure. Contrary to our expectations, theobromine in a relatively
high dose does not have stimulating properties or caffeine-associated
physiological symptoms such as light-headedness. Theobromine has
often been described as a stimulant, with one fifth the potency of
caffeine on adenosine receptors [13]. However, theobromine had no
effect on alertness at any time, although it did decrease blood
pressure. Thus it is possible that theobromine affects peripheral
physiology, but lacks the strong CNS-activating properties of caffeine.
Furthermore, there were no interactions of caffeine and theobromine
on mood or blood pressure.
4.1. Explicit mood
As expected, caffeine increased subjective alertness in subjects,
however, theobromine did not. There was also no interaction of the
two methylxanthines on mood. These results are similar to those found in
Mumford et al. [16], where 5 out of 7 subjects reported strong energizing
effects of caffeine where only 1 subject did for theobromine. However,
since almost all subjects could discriminate doses of 580 mg theobromine,
some psychoactive effects were felt. Indeed, a later study demonstrated
that 19 mg caffeine +250 mg of theobromine increased self-reported
alertness [1]. Since few studies have shown self-reported alerting effects
at doses of caffeine lower than 30 mg the Smit findings suggested a
possible stimulating contribution of theobromine. However, the present
study shows conclusively for the first time that theobromine does not act
as a stimulant alone or in combination with caffeine.
Althoughtheobromine had no alerting effectsit did decrease feelings
of calmness in subjects during the last session (3 h after ingestion),
which indicates a tension-raising effect compared to placebo. This
contradiction may be due to the novelsensations of theobromine, which
may have stronger action on peripheral adenosine receptors and less in
the central nervous system. Theobromine has been reported to have
several peripheral activities, such as anti-tussive, diuretic, hypotensive
and muscle relaxant properties [18,30]. Animal studies have shown that
theobromine has negligible effects on cerebral blood flow and glucose
use [31],belying a lack of direct CNSactivity. However,animals give high
doses of theobromine do show increases in motor activity [32].
Caffeine increased contentedness, possibly due to alleviation of
fatigue effects. Alternatively, caffeine has been shown to modulate
dopamine signaling in animals, which may bring on feelings of
contentedness [33]. Specifically, central adenosine A2A blockade
regulates DARPP-32 phosphorylation and subsequent reward- and
motor-related behaviors [34]. A factor which undoubtedly affected
the results was the choice and timing of dose. The dose for
theobromine is five times that of caffeine dose chosen. Since
theobromine's pharmacological activity is one fifth that of caffeine
[13], the doses of theobromine and caffeine were expected to show
comparable effects on some parameters. However, theobromine had
negligible effects on mood and caffeine appeared to have similar
mood effects as the combination of methylxanthines. Given that
theobromine reaches peak blood levels 3 h after consumption while
caffeine peaks 1 h after consumption, it may be that psychoactive
effects are not reported earlier due to low brain availability, and that
psychoactive effects may be more pronounced at later time points.
Another consideration is the possibility that the explicit mood
measure Bond–Lader VAS, which is commonly used for assessing
caffeine effects [26,27] may not be as appropriate for theobromine,
which seems to have a quite different physiological and psychoactive
profile. It is also possible that an implicit mood measure may be more
sensitive to its effects.
4.2. Implicit mood
It has long been observed that many subjects have difficulty
assessing their internal mood state. Therefore it is of interest to
explore possible methods which indirectly measure mood. Such
measurement can be done via questionnaires on task engagement and
motivation, which often reflect the level of a subject's positive
emotion. For instance, subjects in good moods often rate tasks as more
interesting and pleasant.
The effects on the Motivation and Workload Questionnaire (rating
things as more interesting/less dull) and ERTT (increased rating
negative words) suggest theobromine has an effect on mood that is
not apparent via the direct mood questionnaires. The decrease in
calmness may indeed be explained by the subjects experiencing an
unfamiliar sensation that perhaps is not unpleasant, but more difficult
to describe. The positive findings of theobromine on implicit mood
measures also adds to the discussion in the above section that tests
which are sensitive to caffeine effects such as alertness may not be as
useful for detecting theobromine effects. In conclusion, implicit mood
measurement development needs to be explored further before it can
be used as a corollary to more direct mood questionnaires such as the
Bond–Lader VAS.
4.3. Cognition
The DSST is a measure of psychomotor speed and working
memory. While some labs have shown a performance-improving
effect from caffeine [23,24] others have reported null effects [19,20],
and these differences may be related to dosage, subject selection
criteria and environmental manipulations. In this study caffeine had
no effect on the DSST performance, nor did theobromine. However the
combination of these methylxanthines decreased reaction time for
correct responses as compare to caffeine alone and increased it as
compared to theobromine alone. Originally we hypothesized that
120 mg of caffeine may act as a threshold dose for DSST performance
enhancement and that theobromine may act synergistically with
caffeine to lift performance above threshold. The results suggest that
theobromine and caffeine tended to have opposite effects on response
time. However, since the DSST is more commonly scored as the
821E.S. Mitchell et al. / Physiology & Behavior 104 (2011) 816–822
Downloaded for Anonymous User (n/a) at University of Washington - Seattle - WSC from ClinicalKey.com by Elsevier on April 08, 2019.
For personal use only. No other uses without permission. Copyright ©2019. Elsevier Inc. All rights reserved.

number of symbols decoded rather than response time, caution
should be used in interpreting such results.
4.4. Blood pressure and physiological symptoms
In this study, the natural increase of BP in the morning (morning
surge) seemed to be slowed down, showing significantly lower BP
values 1 h after theobromine consumption compared to placebo,
possibly due to its vasodilatory properties [17,30]. Conversely, caffeine
exerted the opposite effect with higher BP compared to placebo at this
time point only. It is not clear why these two methylxanthines with
comparable inhibitory activity on adenosine receptors would produce
opposite results on blood pressure. One possible explanation is that
caffeine is more bioavailable to the brain, and CNS-mediated blood
pressure control. Interestingly, when theobromine and caffeine were
combinedthe opposing effectson blood pressure werecanceled out. The
similar mood effects of the combination of theobromine and caffeine
compared to caffeine alone, were thus not reflected in similar
physiological effects. However, it must be pointed out that these
changes were from a single treatment and it still unclear how chronic
consumption of theobromine and/or caffeine influences blood pressure.
Habitual consumption of caffeine decreases adenosine A2A receptors in
rodent brains [13], though there have been no reports on changes of
peripheral adenosine receptors in animals or humans. It may be useful
to comparetheobromine'seffects on blood pressure in subjects withlow
daily caffeine intake or in non-withdrawn subjects.
Finally, in contrast to our expectations negative caffeine symptoms
such as trembling decreased in response to caffeine, even while it
raised blood pressure. However, overall negative physiological
sensations are more often reported with higher doses of caffeine
and these subjects were heavy coffee and tea consumers, who have
adapted responses to caffeine. However, even though chronic caffeine
users report less physiological symptoms they nonetheless exhibit
significant caffeine-mediated pressor effects. Meanwhile, theobromi-
ne's blood pressure-lowering effect did not appear to correlate with
noticeable physiological symptoms.
5. Conclusions
In summary, this study has shown strong alerting effects of
caffeine with negligible contribution from theobromine. Caffeine also
at early time points produced contentment effects, while theobro-
mine decreased calmness at a later time point. Theobromine
decreased blood pressure significantly while caffeine increased it in
normotensive subjects. Thus it may be concluded that these
methylxanthines were not operating in a synergistic manner with
any of the measurements used in this study. We hypothesize that
caffeine may have more CNS-mediated effects on alertness, while
theobromine may be acting primarily via peripheral physiology.
Supplementary materials related to this article can be found online
at doi:10.1016/j.physbeh.2011.07.027.
Acknowledgments
We would like to thank Renate Ganzevles, Suzanne Einother, Guus
Duchataeu, Martin Jakel, Mireille Jambroes, Vanessa Ringelberg,
Miranda Slotboom, Jeroen Sterken and Vi Dinh for their assistance
in this study.
References
[1] Smit HJ, Gaffan EA, Rogers PJ. Methylxanthines are the psycho-pharmacologically
active constituents of chocolate. Psychopharmacology 2004;176:412–9.
[2] Michener W, Rozin P. Pharmacological versus sensory factors in the satiation of
chocolate craving. Physiol Behav 1994;56:419–22.
[3] Macht M, Dettmer D. Everyday mood and emotions after eating a chocolate bar or
an apple. Appetite 2006;46:332–6.
[4] Herraiz T. Tetrahydro-beta-carbolines, potential neuroactive alkaloids, in choco-
late and cocoa. J Agric Food Chem 2000;48:4900–4.
[5] Smit HJ, Blackburn RJ. Reinforcing effects of caffeine and theobromine as found in
chocolate. Psychopharmacology 2005;181:101–6.
[6] di Tomaso E, Beltramo M, Piomelli D. Brain cannabinoids in chocolate. Nature
1996;382:677–8.
[7] Mumford GK, Benowitz NL, Evans SM, Kaminski BJ, Preston KL, Sannerud CA, et al.
Absorption rate of methylxanthines following capsules, cola and chocolate. Eur J
Clin Pharmacol 1996;51:319–25.
[8] Haskell CF, Kennedy DO, Wesnes KA, Scholey AB. Cognitive and mood
improvements of caffeine in habitual consumers and habitual non-consumers of
caffeine. Psychopharmacology 2005;179:813–25.
[9] Childs E, de Wit H. Subjective, behavioral, and physiological effects of acute caffeine
in light, nondependent caffeine users. Psychopharmacology 2006;185:514–23.
[10] Juliano LM, Griffiths RR. A critical review of caffeine withdrawal: empirical
validation of symptoms and signs, incidence, severity, and associated features.
Psychopharmacology 2004;176:1–29.
[11] Alsene K, Deckert J, Sand P, de Wit H. Association between A2A receptor gene
polymorphism and caffeine-induced anxiety. Neuropsychopharmacology
2003;28:694–1702.
[12] Childs E, Hohoff C, Deckert J, Xu K, Badner J, de Wit H. Association between
ADORA2A and DRD2 polymorphisms and caffeine-induced anxiety. Neuropsy-
chopharmacology 2008;33:2791–800.
[13] Svenningsson P, Nomikos GG, Fredholm BB. The stimulatory action and the
development of tolerance to caffeine is associated with alterations in gene
expression in specific brain regions. J Neurosci 1993;19:4011–22.
[14] Brunk SF, Ferguson RK, Toubes DB, Leaverton PE, Nordschow CD, Wilson WR. A
teaching format in clinical pharmacology. Comparison of two xanthines and a
placebo. J Clin Pharmacol New Drugs 1973;13:121–6.
[15] Dorfman LJ, Jarvik ME. Comparative stimulant and diuretic actions of caffeine and
theobromine in man. Clin Pharmacol Ther 1970;11:869–72.
[16] Mumford GK, Evans SM, Kaminski BJ, Preston KL, Sannerud CA, Silverman K, et al.
Discriminative stimulus and subjective effects of theobromine and caffeine in
humans. Psychopharmacology 1994;115:1–8.
[17] van den Bogaard B, Draijer R, Westerhof BE, van den Meiracker AH, van Montfrans
GA, van den Born BJ. Effects on peripheral and central blood pressur e of cocoa with
natural or high-dose theobromine: a randomized, double-blind crossover trial.
Hypertension 2010;56:839–46.
[18] Bruinsma K , Taren DL. Chocolat e: food or drug? J Am Die t Assoc 1999;10:
1249–56.
[19] Wyatt JK, Cajochen C, Cecco AR, Czeisler CA, Dijk DJ. Low-dose repeated caffeine
administration for circadian-phase-dependent performance degradation during
extended wakefulness. Sleep 2004;27:374–82.
[20] Mackay M, Tiplady B, Scholey AB. Interactions between alcohol and caffeine in
relation to psychomotor speed and accuracy. Hum Psychopharmacol 2002;17:
151–6.
[21] Bond A, Lader M. The use of analogue scales in rating subjective feelings. Br J
Psychol 1974;47:211–8.
[22] Rogers PJ, Smith JE, Heatherley SV, Pleydell-Pearce CW. Time for tea: mood, blood
pressure and cognitive performance effects of caffeine and theanine administered
alone and together. Psychopharmacology 2007;195:569–77.
[23] Kaplan GB, Greenblatt DJ, Ehrenberg BL, Goddard JE, Cotreau MM, Harmatz JS,
et al. Dose-dependent pharmacokinetics and psychomotor effects of caffeine in
humans. J Clin Pharmacol 1997;37:693–703.
[24] Baker EL, Letz RE, Fidler AT, Shalat S, Plantamura D, Lyndon M. A computer-based
neurobehavioral evaluation system for occupational and environmental epide-
miology: methodology and validation studies. Neurobehav Toxicol Teratol Jul.–
Aug. 1985;7(4):369–77.
[25] Temple JG, Warm JS, Dember WN, Jones KS, LaGrange CM, Mathews G, et al. The
effects of signal salience and caffeine on performance, workload, and stress in an
abbreviated vigilance task. Hum Factors 2002;42:183–94.
[26] Yeomans MR, Ripley T, Davies LH, Rusted JM, Rogers PJ. Effects of caffeine on
performance and mood depend on the level of caffeine abstinence. Psychophar-
macology 2002;164:241–9.
[27] Siegle GJ, Granholm E, Ingram RE, Matt GE. Pupillary and reaction time measures
of sustained processing of negative information in depression. Biol Psychiatry
2001;49:624–36.
[28] Hermans D, De Houwer J. Affective and subjective familiarity ratings of 740 Dutch
words. Psychologica Belgica 1994;34:115–39.
[29] Harmer CJ, Heinzen J, O'Sullivan U, Ayres RA, Cowen PJ. Dissociable effects of acute
antidepressant drug administration on subjective and emotional processing
measures in healthy volunteers. Psychopharmacology 2008;199:495–502.
[30] Usmani OS, Belvisi MG, Patel HJ, Crispino N, Birrell MA, Korbonitz M, et al.
Theobromine inhibits sensory nerve activation and cough. FASEB J 2005;19:231–3.
[31] Grome JJ, Stefanovich V. Differential effects of methylxanthines on local cerebral
blood flow and glucose utilization in the conscious rat. Naunyn Schmiedebergs
Arch Pharmacol 1986;333:172–7.
[32] Kuribara H, Tadokoro S. Behavioral effects of cocoa and its main active compound
theobromine: evaluation by ambulatory activity and discrete avoidance in mice.
Arukoru Kenkyuto Yakubutsu Ison 1992;27:168–79.
[33] Garrett BE, Griffiths RR. The role of dopamine in the behavioral effects of caffeine
in animals and humans. Pharmacol Biochem Behav 1997;57:533–41.
[34] Hsu CW, Chen CY, Wang CS, Chiu TH. Caffeine and a selective adenosine A2A
receptor antagonist induce reward and sensitization behavior associated with
increased phospho-Thr75-DARPP-32 in mice. Psychopharmacology 2009;204:
313–25.
822 E.S. Mitchell et al. / Physiology & Behavior 104 (2011) 816–822
Downloaded for Anonymous User (n/a) at University of Washington - Seattle - WSC from ClinicalKey.com by Elsevier on April 08, 2019.
For personal use only. No other uses without permission. Copyright ©2019. Elsevier Inc. All rights reserved.