The Acute Electrocortical and Blood Pressure Effects of Chocolate

Objective: The present study investigated the effects of consuming chocolate on electroencephalograph (EEG) frequencies and localization and on blood pressure. Method: Across six conditions, 122 participants consumed either higher (60%) cacao chocolate, low (0%) cacao chocolate, higher cacao chocolate + L-theanine, high sugar water, low sugar water, or water. EEGs, blood pressure, and mood were measured before and after a 60-min digestion period. Results: Analyses indicated a decrease in frontal, parietal, and temporal theta and an increase in occipital beta EEG following the consumption of a 60% cacao confection compared with control conditions. Diastolic blood pressure increased with the consumption of higher cacao chocolate when compared to water alone and to higher cacao chocolate + L-theanine. Diastolic and systolic blood pressure decreased following consumption of higher cacao + L-theanine chocolate, averaging 4–8 mmHg. No condition-specific mood changes or gender differences were found. Conclusions: This study suggests an acute stimulating effect of cacao on the human brain and vasoconstrictive effects on peripheral vasculature, the latter of which appear to be offset by an L-theanine additive. Significance: This is the first known study to investigate acute EEG effects of consuming chocolate and suggests a potential attention-enhancing effect.


Introduction
Few food products have garnered such attention or have reached such cultural and mythological significance as chocolate.Indeed, chocolate is frequently heralded as an aphrodisiac, a broadspectrum medicinal agent, a mood-altering substance, a nutritional supplement, an antihypertensive, a stimulant, and as the most frequently craved food in the Western world (Bruinsma & Taren, 1999;di Tomaso, Beltramo, & Piomelli, 1996;Dillinger et al., 2000;Hill & Heaton-Brown, 1994;Rozin, Levine, & Stoess, 1991;Weingarten & Elston, 1991).Research over the past decades has generally been supportive of these effects, if only of minimal magnitude in some cases.Chocolate is made from the cocoa beanwhich is actually the seed of the fruit of the Theobroma cacao tree-and was originally cultivated by the Olmec, Mayan, and Aztec aristocracy in Mesoamerica and by the Inca in South America.
Ancient codices from Mesoamerica indicate that foods made from the raw, fermented, roasted, shelled, and ground cocoa bean (called cacao) were used for a variety of medicinal purposes such as relieving cough, gastrointestinal aids, improving angina and heart palpitations, and even as a sexual stimulant (Dillinger et al., 2000).Modern research has linked the ingestion of flavanols, polyphenolic compounds highly concentrated in the cocoa bean, to stimulation of nitric oxide synthase production (Fisher,Hughes,doi:10.15540/nr.2.1.3Gerhard-Herman, & Hollenberg, 2003).Nitric oxide synthase increases levels of nitric oxide in arterial endothelial cells producing peripheral vasodilation.Cocoa (processed cacao with the cocoa butter removed) and chocolate (processed cacao with cocoa butter not removed) also contain trace amounts of compounds thought to potentially alter brain activity, such as anandamide, a naturally occurring neuromodulator which can bind to cannabinoid-receptors and very mildly mimic the psychoactive properties of plant-derived cannabinoid drugs (di Tomaso et al., 1996).
By far the most heavily researched and most convincing effect of cacao and chocolate ingestion is the effect on blood pressure and vasodilation.The evidence is now highly suggestive that the flavanols in cacao products have vasodilation effects through increases in nitric oxide (NO), which may result in small decreases in blood pressure (Engler et al., 2004;Fraga et al., 2005;Grassi, Lippi, Necozione, Desideri, & Ferri, 2005;Grassi, Necozione, et al., 2005;Taubert, Roesen, Lehmann, Jung, & Schomig, 2007).
Taubert, Roesen, and Schomig (2007) conducted a meta-analysis of randomized controlled studies of the effects of cocoa use over a median period of 2 weeks on blood pressure and reported a mean decrease in systolic blood pressure (SBP) of 4.7 mmHg and in diastolic blood pressure (DBP) of 2.8 mmHg.These results translate into a 20%, 10%, and 8% risk reduction in stroke, coronary heart disease, and all-cause mortality, respectively.And, in a cross-sectional study of 470 elderly men, those who consumed the equivalent of 2.3 g of cocoa powder per day over a 5-year period had a significantly lower SBP (-3.7 mmHg) and DBP (-2.1 mmHg) relative to men with low cocoa intake, translating prospectively into a 45-50% decrease in cardiovascular and all-cause risk (Buijsse, Feskens, Kok, & Kromhout, 2006).Such decreases in blood pressure following prolonged consumption of cacao products have been strongly linked in acute effect studies to peripheral vasodilation resulting from endothelium-dependent relaxation of vascular smooth muscles induced by nitric oxide, which is thought to be increased primarily by the flavanol monomer, epicatechin, in cacao products (Fisher et al., 2003;Heiss et al., 2006;Karim, McCormick, & Kappagoda, 2000;Schroeter et al., 2006).This vasodilation has been generally observed to peak at 1-2 hr following consumption of a flavanol-rich chocolate product (Schroeter et al., 2006;Taubert, Rosen, Lehmann et al., 2007).These supportive outcomes notwithstanding, Egan, Laken, Donovan, and Woolson (2010) have pointed to important inconsistencies in outcomes, designs, type, and chemical constituents of cacao confections, dose and time-dependent effects, subject blood pressure (BP) variability, and BP measurement devices, which leave the cardiovascular effects of cacao uncertain at the present time.
The endothelium-dependent relaxation (EDR) effects of cacao are somewhat paradoxical, as cacao also contains a number of sympathomimetics, which have vasoconstrictive and generally stimulatory effects, most notably the biogenic amines, tyramine and phenylethylamine (PEA), and the methylxanthines, caffeine and theobromine (Bruinsma & Taren, 1999;Hurst, Martin, & Zoumas, 1982).Although these substances occur in different amounts in various confections, PEA and theobromine generally are found in larger quantities and tend to potentiate the release of catecholamines, thus causing vasoconstriction and elevations in blood pressure.It is not clear at the present time whether these vasoconstrictive effects are offset by the more prominent EDR vasodilatation effects reported above, or whether the time course of these phenomena are different, resulting in more immediate stimulation followed by a delayed vasodilatation effect.Furthermore, these stimulant effects may well explain the general increase in arousal often reported by chocolate consumers (Bruinsma & Taren, 1999;Dillinger et al., 2000).
If chocolate indeed has an immediate arousing effect, one might expect to see increased arousal in the brain.However, our review of the chocolate literature over the past decades reveals no studies of the effects of chocolate consumption on central nervous system (CNS) arousal.
Martin (1998) investigated the electroencephalographic (EEG) effects of olfactory stimulation with the aroma of chocolate, but did not examine the consumption of chocolate in his research.Small, Zatorre, Dagher, Evans, and Jones-Gotman (2001), in a positron emission tomography (PET) study of brain changes following eating chocolate, were interested only in the immediate (within 10 s) reward characteristics of consuming chocolate and did not allow a sufficient time course for cacao constituents to have an effect on cerebral blood flow following digestion.Francis, Head, Morris, and Macdonald (2006) conducted functional magnetic resonance imaging (fMRI) on 16 healthy young women following 5 days of consumption of a flavanol-rich cocoa beverage compared to a low flavanol beverage and examined blood flow changes 1.5 hr after consumption of the beverage during the performance of a cognitive switching task.Although they found no flavanolspecific effects on reaction times, error rates, or doi:10.15540/nr.2.1.3heart rate, they did observe increased fMRI blood oxygenation level-dependent (BOLD) cerebral blood flow (CBF) during the cognitive task following the flavanol-rich cocoa regimen relative to the low flavanol regimen.In a separate pilot study with 4 participants of the time course of CBF changes following an acute dose of flavanol-rich cocoa, Francis et al. reported a peak blood flow response at approximately 2 hr post-ingestion with return to baseline after approximately 6 hr.Reported fMRI images were specific to the cognitive task and not to flavanol ingestion.
Although these researchers suggest that the observed effects on CBF may be due to NO-induced EDR, they also point out that the fMRI BOLD response is a neurovascular phenomenon and may result from changes in vascular tone as well as neural activity influenced by stimulants, such as caffeine, in the cocoa product.Nonetheless, these results suggest a flavanolinduced increase in cerebral blood flow consistent with the vasodilation reported in other studies.And, more recently, Camfield et al. (2012), in a study of 61 middle-aged adults who consumed a daily chocolate beverage containing 250 mg or 500 mg of cocoa flavanols compared with a low cocoa flavanol beverage over a 30-day period, found no changes in behavioral measures of accuracy or reaction time on a spatial working memory task but did observe condition-specific amplitude and latency differences on EEG visual evoked potentials (VEP) during the same task.These authors interpreted the observed VEP changes as reflective of increased neural efficiency following the chronic ingestion of cocoa flavanols.
A recent comprehensive review of the neurobiological effects of cocoa flavanols on cognition and behavior indicates rather strong support for neuroprotective effects of long-term consumption of flavanols on age-related and disease-related cognitive decline but less support for the more immediate effects of cocoa consumption on specific brain mechanisms involved in neurogenesis and neuronal function and connectivity, particularly in humans (Sokolov, Pavlova, Klosterhalfen, & Enck, 2013).These authors encourage and offer a template for future research into effects of cacao on human cognition, mood, and behavior.

Given
the long-term neuroprotective and neuromodulatory effects of cocoa consumption, the suggested stimulant characteristics of cacao, and the glaring absence of published acute CNS arousal studies, we elected to conduct a controlled EEG study of the comparative effects of consuming a higher cacao-content chocolate (with a high flavanol content) with a low-cacao content chocolate (with no flavanol content) and with balanced sugar and water controls.
Sugar controls were included in the present study to control for reported general arousal effects of glucose (Hoffman & Polich, 1998;Hoffman, Friedmann, Saltman, & Polich, 1999).Additionally, as a partial test of the hypothesized acute sympathomimetic effects of cacao, we included a third chocolate condition by the addition of L-theanine to the same higher cacao-content chocolate formulation.L-theanine, an extract of green tea, has been shown in numerous animal and human studies to counteract the stimulating effects of caffeine and stressors, apparently by its ability to bind to the glutamate receptor and to block binding of L-glutamic acid in cortical neurons (Kimura, Ozeki, Juneja, & Ohira, 2007;Mason, 2004;).L-theanine has been found to reduce blood pressure (Yokogoshi et al., 1995;Yokogoshi & Kobayashi, 1998), to elevate posterior EEG alpha activity (Kobayashi et al., 1998), to reduce the psychological and physiological response to a mental stressor (Kimura et al., 2007), and to improve learning in animal models (Juneja, Chu, Okubo, Nagato, & Yokogoshi, 1999).
We hypothesized that consumption of the higher cacao-content condition, relative to the low cacao-content, sugar, and water controls, in human volunteers would result in increased activation of the neocortex and increased blood pressure within 1 hr after ingestion and that these effects would be reversed in the higher cocoacontent plus L-theanine condition.

Participants
A power analysis was conducted to determine the optimal sample size required to detect the hypothesized effect of chocolate on EEG and blood pressure (Howell, 2002).A complex multivariate design was modeled employing 11 dependent variables (9 EEG frequency and 2 blood pressure variables) studied across two repeated measures (pre-and post-ingestion) for each of six treatment conditions.For an alpha level of .05 and a power of .80,sample sizes in each of the 6 treatment cells of 20 allowed a hypothesized medium effect size to be detected.
Consequently, 125 participants (10 males, 10 females in each of the six treatment cells, plus 5 extra participants to allow for possible attrition and outliers) were recruited from the Psychology Department undergraduate voluntary research pool.After exclusion of outliers (3 participants had doi:10.15540/nr.2.1.3clinically-elevated blood pressure readings and were referred to the health center), 122 participants completed the study and were analyzed.Participants were between the ages of 18 and 25 years and were excluded if they used illicit drugs, stimulant or depressant medications, or nicotine, if they had diabetes mellitus, or if they were allergic to chocolate or nuts.Women were not tested during their premenstrual or menstrual phase due to the potentially confounding effect of chocolate cravings during this time and because hormone imbalances during these menstrual phases have been shown to affect the EEG (Dusser de Barenne & Gibbs, 1942;Solis-Ortiz, Ramos, Arce, Guevara, & Corsi-Cabrera, 1994).All participants abstained from caffeine and chocolate intake 24 hr prior to the EEG study.The present study was approved by the NAU Institutional Review Board for the Protection of Human Subjects in Research.

Materials and Equipment
A standard weight scale was used to weigh each participant 1 week prior to study in order to determine the amount of chocolate to administer.Participants were weighed by a same-sex research assistant, and female participants at weigh-in were given a menstrual calendar on which to plot their predicted menstrual cycle.At the initial weigh-in, the participant was briefed as to the nature of the study and the requisite informed consent documents were completed.Blood pressure readings for the study were taken by a HoMedics Automated Blood Pressure Monitor with participants seated and left arm resting at heart level.
The three chocolate treatments were prepared by The Hershey Company, individually wrapped in 40 g squares of identical appearance and coded by contents.
The higher cacao-content chocolate contained 60% cacao with 15 mg/g of total polyphenols, and 0.37 g/g of sugar; the low cacaocontent chocolate was a white chocolate (colored with 5% Hansen Brown) that contained 0.4 mg/g of total polyphenols, and 0.56 g/g of sugar; the higher cacao-content + L-theanine chocolate contained the identical components as higher cacao-content chocolate above plus 128 mg (3.2 mg/g) of Ltheanine (L-theanine has a Generally Recognized as Safe [GRAS] designation by the FDA and, with recommended dosages of 50-200 mg/serving, the amounts used in this study were well within the recommended dosages).Table 1 presents total ingredients of the three chocolate treatments.Three control treatments were also prepared, comprised of a high sugar beverage treatment containing an equivalent amount of sugar as the low cacaocontent chocolate (23 g/40 g bar or 0.57 g/kg body weight) dissolved in 350 ml (1.5 cups) of water; a low sugar beverage treatment containing an equivalent amount of sugar as the higher cacao chocolate (14 g/40 g bar or 0.35 g/kg body weight) dissolved in 350 ml of water; and a 350 ml water treatment.For the chocolate and sugar conditions, each participant received 1 g of chocolate for each kg of body weight, with an equivalent amount of sugar for each kg of body weight for the sugar conditions.For example in standard units, a 150 lb participant would receive 2.4 ounces of either of the three chocolate treatments, approximately equivalent to a standard size chocolate bar, 1.38 ounces of sugar in 1.5 cups water, 0.84 ounces of sugar in 1.5 cups water, or 1.5 cups water.
The Positive and Negative Affect Scale (PANAS) was used as a brief measure of emotional changes following chocolate consumption (Watson, Clark, & Tellegen, 1988).The PANAS was administered at the beginning of the study prior to treatment before the EEG was attached, immediately following administration of each condition, and again 1 hr 10 min later after a digestion period and second EEG.
A Mitsar 201 24-channel EEG acquisition system was used to measure each participant's EEG frequencies (Mitsar Co. LTD, 1996).The Mitsar 201 DC amplifiers have a 500 Hz digital sampling rate and input impedance not less than 200 МΩ.EEG data were recorded and preprocessed using WinEEG software (Mitsar Co. LTD, 1996), double visually artifacted by two independent artifactors, and power spectral FFT analyzed utilizing NovaTech EEG Eureka and MHyT software (Nova Tech EEG, Inc., 2006).FFT analysis employed Hamming time domain tapering, Blackman frequency domain smoothing, an overlapping FFT windows advancement factor of 8, and a moving average smoothing filter of 3.
The International 10-20 placement system was used to attach 19 Ag/AgCl monopolar electrodes on the scalp with mathematically linked-ear references utilizing the Electro-Cap System (Electro-Cap International, Inc., 2006).Electrode impedances were adjusted to < 5 kohms and to within 1 kohm of each other.All data were recorded in a sound attenuated research suite, appointed with the requisite EEG and blood pressure monitoring equipment.Participants were seated comfortably in a recliner and were able to read magazines during the digestion phase and to sit quietly during the EEG recording phases.blind as to the nature of the liquid in the cup, with the exception that the RA could visually differentiate the liquid water/sugar water controls from the solid chocolate conditions.Participants were similarly partially blind as to the exact nature of the substance they were consuming, either an unknown chocolate substance or sugar water.
When the participant arrived at the laboratory, they were seated in the recliner and were given the first PANAS to complete; blood pressure was then recorded and the Electro-Cap and EEG equipment were attached, during which time the participant completed a brief questionnaire to substantiate lack of drug use over the past 2 days and lack of caffeine and food intake for the past 24 and 4 hr, respectively.A 10-min, eyes-closed resting baseline EEG was then recorded.Afterwards, each participant was administered their respective chocolate, sugar water, or water treatment, was given 5 min to ingest the substance, and was administered the PANAS again.Sixty min were then allowed for digestion and absorption of the chocolate or water treatments; participants were also visually monitored for alertness.
Following the 60-min period, a second 10-min, eyes-closed resting EEG was recorded, then blood pressure was taken, and a final PANAS was administered.The Electro-Cap was then removed and the participant was debriefed.Figure 1 presents the timeline for the study.
In order to investigate the secondary effect of gender, an additional independent variable was included in each analysis, making separate 2 (between-groups: gender) x 6 (between-groups: treatment condition) analyses of post-treatment minus pre-treatment EEG differences for each functional cluster.As this was an investigational study and the percent cacao in the maximal treatment condition was relatively low (60%) thus lowering the magnitude of effect, simple effects analyses of EEGs were conducted utilizing LSD post-hoc t-tests.Blood pressure and PANAS effects were assessed utilizing ANOVA with simple effects comparisons made by Tukey HSD tests.
Since surface EEG recordings are aggregates of farfield potentials generated across a 3-dimensional, quasi-spherical cortical space, we wondered what deeper cortical structures might be most impacted by the biochemical constituents of chocolate following our neutral reading task.Low-resolution brain electromagnetic tomography (LORETA) is a neuroimaging software companion to contemporary EEG analyses which allows the triangulation of these surface scalp potentials to their cortical source generators (Pascual-Marqui, 1999;Pascual-Marqui, Esslen, Kochi, & Lehmann, 2002;Pascual-Marqui, Michel, & Lehman, 1994).LORETA algorithms compute a 3-dimensional inverse solution space of cortical gray matter and hippocampi mapped onto a probabilistic Talairach atlas partitioned into 2394 7 mm 3 volumetric units, or voxels.Brodmann anatomical labels may be reported for relevant regions of interest utilizing the Montreal Neurological Institute (MNI) realistic head model (The KEY Institute for Brain-Mind Research, 1995).For the present study, LORETA mapping was utilized in a purely descriptive fashion post hoc to identify cortical regions of interest involved in obtained effects.LORETA Current Source Density (CSD) maps were generated from between-groups comparisons of the natural log transformation of FFT power spectral output for each statistically significant frequency and functional cluster.

Analysis of EEG Changes
EEG Absolute Power values for each of five primary frequencies for each electrode across each condition at Baseline, at Post-Treatment, and for Difference Scores are presented in Tables 2 through 6 in the Appendix.Values in each table are Absolute Power values for that frequency with decimals removed for ease of presentation (i.e., 608 = .0608,or .0608x 10 4 ).In the presentation of these results, a negative difference score indicates that the specified EEG power decreased following treatment (post-treatment -pre-treatment).
For the functional cluster analyses (i.e., frontal, central, parietal, occipital, and temporal), EEG regional cluster scores for individual participants which exceeded 3.29 standard deviation units from the mean for that cluster were identified as outliers and were replaced by the next lower score for that cluster (Tabachnick & Fidell, 2013).Subsequent tests for departures from normality and homogeneity of variance revealed no significant departures for the tested independent variables, with the exception of gamma parietal, which showed a significant departure for the homogeneity of variance doi:10.15540/nr.2.1.3assumption.
This latter variable was Log10 transformed, subsequently tested for normality and homogeneity of variance, and was found to meet requirements for analysis.Regional cluster scores for each frequency and condition were then entered into separate 2 x 6 ANOVAs.These results are presented in Table 7.
These ANOVA analyses revealed significant main effects for Condition for frontal theta, F(5, 110) = 3.12, p = .011,η 2 = .124;parietal theta, F(5, 110) = 2.38, p = .043,η 2 = .097;and temporal theta, F(5, 110) = 2.72, p = .024,η 2 = .110;with a trend for central theta as well, F(5, 110) = 2.00, p = .085,η 2 = .083.A significant main effect for Gender was also found for frontal theta, F(1, 110) = 5.94, p = .016,η 2 = .051,with males showing significantly greater decreases in frontal theta than females across all conditions.No interactions were found to be statistically significant.Planned comparisons revealed frontal theta decreases from Baseline to Post-Ingestion to be significantly greater for the higher cacao-content chocolate condition relative to the water (p = .006),high sugar (p = .001),low sugar (p = .005),and low cacao-content chocolate (p = .006)conditions, with these latter conditions actually showing increases in frontal theta.Figure 2 presents these effects graphically for frontal theta.(For Figures 2 through 6, the ordinate scale is set to be equivalent across all figures for ease of magnitude comparisons.)Similar effects were found for parietal theta and temporal theta with the higher cacao-content chocolate confection showing significant decreases with consumption relative to water (p = .021,.026),high sugar (p = .009,.002),and low cacao-content chocolate (p = .005,.008),which each showed increases across the conditions.Additionally, the higher cacao-content chocolate + L-theanine condition showed significantly smaller increases (p = .042)in temporal theta compared to the high sugar condition across treatment.Figures 3 and 4 present these outcomes graphically.These results indicate significant decreases in frontal, parietal, and temporal theta EEG frequencies following the consumption of a 60% cacao confection relative to increases across these cortical regions following consumption of water, high sugar, an approximately 0% cacao-content confection, and, for frontal theta, a low sugar condition.Examining narrower frequency bins within the theta band revealed corresponding changes to those reported above for the higher cacao-content chocolate in frontal, parietal, and temporal low theta (4-5.99Hz) relative to water (p = .030,.010,.011),high sugar (p = .014,.010,.004),low sugar (p = .005,.010,.007),and low cacao-content chocolate (p = .008,.003,.002).Within the high theta (6-7.99Hz) bin, only the frontal region reached statistical significance for higher cacao-content chocolate relative to water (p = .009),high sugar (p < .0001),low sugar (p = .028),and low cacao-content chocolate (p = .009).These consistent changes in high and low theta frequency bins, primarily in frontal regions, suggest a suppressant effect of cacao on these frequencies.High theta frequency in the temporal region showed very small increases with consumption of low sugar (p = .008)and higher cacao-content chocolate (p = .009)relative to larger high sugar condition increases, an apparent enhancing effect of sugar on the high theta frequency in this region since the higher cacao-content chocolate contained the same amount of sugar as the low sugar condition.
Significant main effects were also obtained for frontal alpha across treatment conditions, F(5, 110) = 2.93, p = .016,η 2 = .117.There were no gender or interaction effects for any clusters or alpha frequencies, nor did analysis of any narrow alpha frequency bins result in significant effects.Examination of simple effects for frontal alpha revealed decreases across treatment for the low sugar and higher cacao-content chocolate conditions relative to increases for high sugar (p = .003,.025)and low cacao-content chocolate (p = .006,.047)conditions and for low sugar relative to water (p = .018).Given that low sugar and higher cacao-content chocolate conditions were identical for low sugar levels and that high sugar and low cacao conditions were identical for higher sugar levels, these results suggest an effect of sugar on increasing alpha frequencies in frontal regions.These frontal alpha effects are presented graphically in Figure 5. Significant main effects were obtained for frontal beta, F(5, 110) = 2.80, p = .02,η 2 = .113;central beta, F(5, 110) = 3.59, p = .005,η 2 = .14;and occipital beta EEGs, F(5, 110) = 2.45, p = .038,η 2 = .10.No significant gender or interaction effects were obtained.Planned and post-hoc comparisons revealed that for frontal and central regions high sugar was associated with beta increases relative to decreases for water (p = .003,.006),low sugar (p = .019,.016),higher cacao-content chocolate (p = .002,.004),and higher cacao-content chocolate + L-Theanine (p = .049,.022),and that for central regions low cacao-content chocolate was associated with beta increases relative to decreases for water (p = .009),low sugar (p = .025),higher cacaocontent chocolate (p = .007),and higher cacaocontent chocolate + L-Theanine (p = .035).In occipital regions, high sugar was also associated with increases in beta relative to decreases for water (p = .002)and low sugar (p = .016).Again, these beta increases following the consumption of high sugar water in frontal, central, and occipital regions and in central regions following the consumption of a low cacao-content chocolate confection containing comparable high sugar levels, suggest a beta EEG enhancement effect of sugar in these cortical regions.
However, an additional, marginally significant (p = .05)increase in beta EEG for the higher cacao-content chocolate condition relative to a decrease for water in occipital regions, in the absence of a corresponding increase for the low sugar condition, suggests a potential specific beta enhancement effect for the higher cacao confection.This beta enhancement effect is graphically presented in Figure 6.
No significant main or interaction effects were obtained for delta or gamma EEG frequencies.The absence of these outcomes suggests no statistically significant effect of any of the six conditions on delta and gamma EEG frequencies within the cortical regions studied.

LORETA Source Localization Effects
In order to separate the effects of cacao from sugar in our study, we compared the higher cacao-content condition with the low sugar condition and examined Current Source Density for the higher cacao-content constituents free of sugar effects.Figures 7 and 8 show cortical generators for the low theta and beta frequencies respectively, with relevant neuroimaging parameters reported.The parahippocampal gyrus and sub-gyral hippocampus in the right posterior limbic lobe reflect areas of maximal difference between conditions for the obtained low theta suppressant effect.Maximal differences for the obtained beta enhancement effect involved posterior portions of the medial frontal gyrus and the paracentral lobule in the frontal lobe and the anterior cingulate gyrus in the limbic lobe.Implications of these CSD localization findings are discussed below.

Analysis of Blood Pressure Changes
Diastolic and systolic blood pressure changes from baseline to 70-min after ingestion for each treatment and for each gender were analyzed by separate 2 (genders) x 6 (treatments) ANOVAs.For DBP, a significant treatment main effect was found, F(5, 110) = 6.57, p < .0001,η 2 = .23,but no significant gender effect, F(1, 110) = 0.002, p = .96,η 2 < .0001,and no significant interaction effect, F(5, 110) = 0.74, p = .59,η 2 = .03,were obtained.Planned comparisons across treatments for DBP revealed higher cacao-content chocolate DBP to be significantly greater than the higher cacao-content chocolate + L-theanine and water conditions, and higher cacao-content chocolate + L-theanine DBP to be significantly lower than the higher cacao-content chocolate, low cacao-content chocolate, low sugar, and high sugar conditions (p < .05).These effects are presented in Table 8 and Figure 9.
Figure 10.Systolic blood pressure changes post-treatment minus pre-treatment for each condition, with higher cacao + L-theanine significantly lower than higher cacao, low cacao, low sugar, and high sugar.SE bars = +/-1 SE.

Discussion
General Findings Consistent with our hypotheses, it appears that chocolate does have an acute stimulatory effect on components of both the central (CNS) and peripheral nervous systems (PNS).In order to understand the nature and implications of these outcomes, it is important to examine the effects on CNS and PNS arousal of specific treatment conditions.The act of sitting quietly and reading neutral magazines during the water condition for our sample of healthy college students tended to diminish CNS arousal, as attested to by general decreases in posterior beta and increases in anterior theta EEG frequencies.The combination of decreased beta and increased theta is commonly seen in attentional disorders and often reflects inattention and "spacing out" (Lubar, Swartwood, Swartwood, & O'Donnell, 1995;Mann, Lubar, Zimmerman, Miller, & Muenchen, 1992), not to be unexpected among college students sitting quietly and reading uninteresting magazines.This diminution of the CNS arousal was inhibited by the consumption of a high sugar drink (increased diffuse beta) and by the consumption of a higher cacaocontent chocolate (decreased anterior theta and increased posterior beta).Interestingly, although the high sugar drink was very effective in increasing cortical beta in general, it was not effective in decreasing theta across the cortex, whereas the higher cacao-content chocolate condition was statistically equivalent to high sugar in increasing posterior beta but far more effective in decreasing more diffuse theta activity, even though the higher cacao-content condition contained less sugar (0.35 g/kg body weight) than the high sugar condition (0.57 g/kg body weight).This effect is further supported by a significantly diminished beta enhancement by the low sugar condition relative to high sugar, the former containing an equivalent amount of sugar to the higher cacao-content conditions but none of the cacao bioactive components. doi:10.15540/nr.2.1.3 Furthermore, the possibility of these effects being also due to changes in mu rhythms during page turning and lambdoid waves during saccadic eye movements while reading seems unlikely in that both the pre-ingestion and post-ingestion EEGs were recorded eyes closed before and after, respectively, the reading of magazines, with approximately 5 minutes intervening after stopping reading and before post-ingestion EEG recording while the Electro-Cap was repositioned, impedances were checked, and good clean EEG traces were obtained.These observations suggest a supplemental and differential stimulating effect of the bioactive compounds in cacao (such as biogenic amines and/or methylxanthines) over that of sugar, with sugar increasing cortical beta and cacao decreasing cortical theta.

Potential Biogenic Amines Underlying EEG Cacao Effects
Changes in biogenic amine neurotransmission are associated with distinctive patterns of EEG change.In Fischer rats for example, small declines in EEG slow wave activity have been associated with administration of dopamine agonists (Dimpfel, 2005).These declines were followed, 90 min later, by increases in theta, delta, and alpha 2 activity in the hippocampus and frontal cortex.
The administration of the highest L-Dopa dose replicated the biphasic low dose finding but predominately in the frontal cortex.It is noteworthy that increases in theta spectral power have been associated with reports of increased tiredness and sedation in humans (Dimpfel, 2008;Dimpfel & Schober, 2001;Vyazovskiy & Tobler, 2005) and also with increased attentional demands during mental work (Schober, Schellenberg, & Dimpfel, 1995;Schwarz-Ottersbach & Goldberg, 1986).Also, administration of direct D2 agonists was found to decrease alpha 2 power in the frontal cortex, hippocampus, and striatum but not in the reticular formation.
Predictably the administration of a DA2 antagonist dramatically increased alpha 2 power in the frontal cortex 3 hr after administration of the highest dose.Our data, collected on humans within an hour after the administration of varied cacao doses (with or without theanine) from frontal, temporal, and parietal sites, are consistent with these acute theta wave declines reported by Dimpfel (2008).
Given the known relationship between dopamine (D) regulation and acetylcholine (ACh) neurotransmission, these low dose agonist effects are thought to be associated with the activation of heterosynaptic presynaptic D2 receptors located on cholinergic neurons involving the cAMP inhibition of activity in the b-arrestin pathway (Dimpfel, 2008).Data presented by Zhang, Zhou, and Dani (2004) demonstrating increased dopamine release following administration of an ACh esterase inhibitor supports this conclusion.Conversely, declines in alpha 2 activity have been associated with the administration of an ACh M1 antagonist.Clearly, the availability of different receptor subtypes and their pharmacological selectivity in different neuroanatomical circuits (cortical and subcortical) help regulate both DA and ACh neurotransmission and EEG frequency pattern alterations.However, the precise subcortical neuroanatomical circuitry responsible for these changes has yet to be elucidated and behavioral data collected concurrently are limited.Moreover, these effects are dose-dependent and area specific, as well as time-, drug-, and task-dependent.To complicate matters further, norepinephrine reuptake inhibitors have been demonstrated to increase theta wave activity in the septo-hippocampal area (Hajós, Hoffman, Robinson, Yu, & Hajós-Korcsok, 2003) and administration of the antihypertensive, anxiolytic, and alpha 1 adrenergic receptor agonist clonidine has also been demonstrated to increase theta EEG activity (Dimpfel & Schober, 2001).With regard to the biogenic amines, alterations in EEG outcomes not only appear to involve differing signal activity in different cortical and subcortical neuroanatomical pathways, but also the integration of differing subcellular neuronal processes involved in complex neuronal patterns and behavior associated with cacao-induced electrocortical changes.
For example, cacao flavanols are known to cross the blood brain barrier, to increase blood circulation in brain, to exert antioxidative effects, to increase nitric oxide production, and to trigger protein-receptor synthesis via mitogen-activated protein, phosphoinositide 3-kinase, and extra-signal regulated subcellular cascades, all of which are associated with the neuromodulation of long-term potentiation integral to the formation of memories and neurocognitive function (Sokolov et al., 2013).The neurological impacts of cacao flavanols have also been reported to exert neuroprotective and neuromodulatory effects that promote synaptic connectivity, alter cognition and behavior and promote endothelium-dependent vasodilation (Sokolov et al., 2013).Flavanoids have also been demonstrated to promote neurogenesis and memory formation and to protect against neuronal cell death by increasing the expression of brain derived neurotrophic factor in the hippocampus.And, with regard to acute and chronic consumption effects, the doi:10.15540/nr.2.1.3short-term oral exposure to 100 mg/100 g body weight of cacao exerted an anxiolytic effect on elevated T-maze behavior in rats, whereas exposure to cacao for 2 weeks increased brain serotonin concentration and turnover rate but failed to alter elevated T-maze behavior (Yamada, Yamada, Okano, Terashima, & Yokogoshi, 2009).Nonetheless, in humans, the effects of cacaoderived flavanols on cognitive function and mood have not been clearly elucidated and the effects of acute vs. long-term exposure to flavanols on arousal and EEG changes relating to these processes have not been adequately investigated.

L-theanine EEG Effects
L-theanine has a documented EEG alpha enhancement effect in the research literature (Juneja et al., 1999;Kimura et al., 2007;Kobayashi et al., 1998).While we did not find such an effect with our higher cacao-content + L-theanine confection, we theorize this to be due to the sympathomimetic ingredients in chocolate suppressing slow wave (alpha and theta) and enhancing fast wave (beta) activity.The absence of a significant alpha suppression effect with the higher cacao-content condition is noteworthy.EEG alpha has been historically associated with quiet rest and relaxation.Activation of the brain with increased beta and decreased alpha could actually be perceived by participants as an increase in anxiety and agitation.
The finding of increased beta, decreased theta, and a stable alpha frequency with a moderate cacao-content confection suggests that our participants were neurologically activated but without the agitation that might have been perceived had alpha actually been suppressed as well.These CNS arousal without anxiety effects are supported by the absence of significant changes in PANAS mood scores, particularly those related to anxiety/agitation.
As noted above, our results suggest that a high sugar beverage can actually increase alpha over our 60% cacao confection, perhaps having implications for a combination of these two substances.

Cortical Source Generators
LORETA cortical source localizations of the surface potentials generated by cacao consumption, independent of sugar effects, suggest some intriguing implications for the impact of cacao on the human brain.Although theta rhythms have been associated with visual imagery, problem solving, perceptual processing, attentive performance in cognitive tasks, creativity, and dissociative states (Stevens et al., 2004), low theta during quiet waking predicts the subsequent development of sleep slow-wave activity and an increase in sleep propensity (Makeig, Jung, & Sejnowski, 2000;Vyazovskiy & Tobler, 2005).Therefore, the suppression of low theta following the consumption of cacao in the present study indicates a counteracting of natural drowsiness induced by an hour of quiet reading of neutral magazines.Since the structures activated are primarily involved with the encoding and recognition of scenes such as landscapes, cityscapes, etc., the content of many of the magazines available during the digestion phase, and with episodic memory (Orrison, 2008), these results suggest an activation of task-related processes in the brain following the consumption of the higher cacao-content confection independent of sugar.Similarly, the enhancement of posterior frontal and anterior cingulate beta frequencies indicates an activation of such executive functions as the recognition of similarities and differences, retention of long term memories, learning, problem-solving, and mental conflict resolution (Orrison, 2008).Taken together, the localization of cortical source generators of the observed surface potentials suggests an enhancement of task-related activities following consumption of cacao in our study.Furthermore, the nature of these brainwave changes directly counteracts those specific frequencies seen during diminished attention.

Acute Blood Pressure Effects
Peripherally, the acute effects (1 hr after consumption) of higher cacao-content and of higher cacao-content + L-theanine ingestion on blood pressure were rather remarkable, with BP changes on the order of 3-5 mmHg.It is noteworthy that while the higher cacao-content condition significantly increased DBP relative to the water condition, it did not significantly do so when compared with the low cacao-content condition.These higher and lower cacao conditions differed considerably in the presence of the psychoactive biogenic amines, tyramine and, particularly, PEA, and the methylxanthines, caffeine and, particularly, theobromine, and would be expected to have a stronger differential effect on blood pressure.Although not statistically significant, these differences were in the predicted direction and were on the order of 3 mmHg different.As mentioned below, it is likely that the only moderate levels of cacao used in this study contributed to this small effect.It is also possible that more prominent vasoconstrictive effects of these sympathomimetics were counteracted by beginning vasodilatation effects of the epicatechin polyphenols in the higher cacao-content product, thus diminishing treatment differences.One can only speculate at this point doi:10.15540/nr.2.1.3what these differences would have been for a higher cacao-content chocolate containing more of these psychoactive compounds.
While the more immediate effect of consuming a higher cacao-content confection was an increase in diastolic BP of 4.7 mmHg on average, the higher cacao-content + L-theanine confection actually counteracted this effect by lowering diastolic BP 3.65 mmHg on average and systolic BP 5.15 mmHg.The potential antihypertensive effect of lowering diastolic blood pressure from the 4.7 mmHg increase seen with higher cacao-content to the 3.65 mmHg decrease seen with higher cacao-content + L-theanine represents an 8.4 mmHg decrease in diastolic blood pressure.These blood pressure lowering outcomes following a single recommended dose of the L-theanine additive represent approximately one-third to one-half the effects of sustained use of standard antihypertensive medications, and without documented side effects (Mason, 2004;Wu et al., 2005).Given the apparent ability of L-theanine to inhibit the more immediate sympathomimetic effects of cacao and to acutely lower blood pressure, combined with the documented longer term antihypertensive effects of polyphenols in cacao, there is clearly the possibility of an application of this combination of L-theanine and cacao in the treatment of hypertension.This exciting possibility is certainly speculative at the present time and awaits further directive research into the longer term consequences of cacao + Ltheanine use, particularly for higher doses of both constituents.

Limitations
Overall, these CNS arousal effects suggest that the constituents in cacao (polyphenols, biogenic amines and/or methylxanthines) can inhibit naturally occurring deactivation of the brain during mundane and less interesting tasks.
The relative enhancement of beta and suppression of theta frequencies found in this study indicate that higher cacao-content chocolate may have an impact upon electrocortical processes implicated in diminished attention, a common complaint among college students attending lectures and reading academic material.A limitation of this study was that we did not directly measure attentional behavior.Given our findings, it would be of interest in future studies of the effects of chocolate to do so.It is also important to note that this same combination of suppressed beta and enhanced theta has been reported in diagnosed Attention Deficit Disorder (ADD; Lubar, et al., 1995;Mann, et al., 1992).While our participants were not ADD patients, it would be interesting to replicate this study and to observe not only EEG changes but also measures of attentional performance with such a clinical sample.
For reasons of palatability and availability, our study utilized a dark chocolate confection containing only moderate amounts (60%) of cacao.This choice of chocolate confections was a major limitation of this study and quite likely resulted in the small effect sizes (< .25;See Cohen, 1988) found in our analyses, even though the results reported were statistically significant.There are quite palatable chocolate preparations publicly available containing up to 90% cacao.Certainly this study should be replicated with a palatable chocolate confection containing higher percentages of cacao or increased concentrations of cacao bioactives to increase the magnitude of effect and to better understand which cacao constituents are predominantly responsible for these effects.Also of interest for future research would be to examine these enhanced effects for more individualized frequency bins, as suggested by Klimesch (1999).

Figure 1 .
Figure 1.Timeline for the study in minutes.

Figure 2 .
Figure 2. Frontal theta EEG absolute power changes post-treatment minus pre-treatment for each condition, showing significant decreases for higher cacao relative to water, high sugar, low sugar, and low cacao.SE bars = +/-1 SE.

Figure 3 .
Figure 3. Parietal theta EEG absolute power changes post-treatment minus pre-treatment for each condition, showing significant decreases for higher cacao relative to water, high sugar, and low cacao.SE bars = +/-1 SE.

Figure 4 .
Figure 4. Temporal theta EEG absolute power changes post-treatment minus pre-treatment for each condition, showing significant decreases for higher cacao relative to water, high sugar, and low cacao.SE bars = +/-1 SE.

Figure 5 .
Figure 5. Frontal alpha EEG absolute power changes post-treatment minus pre-treatment for each condition, showing significant decreases for low sugar and higher cacao relative to increases for high sugar and low cacao and for low sugar relative to water.SE bars = +/-1 SE.

Figure 6 .
Figure 6.Occipital beta EEG absolute power changes post-treatment minus pre-treatment for each condition, showing significant increases for higher cacao relative to decreases for water.SE bars = +/-1 SE.

Table 1
Ingredients of the three chocolate confections used in the present study Note: Nutrition information calculated using Genesis® R&D SQL nutritional analysis and labeling system(ESHA Research,  Salem, OR 2007).DMAC = 4-dimethylaminocinnamaldehyde total flavanol content; ORAC = Oxygen Radical Absorbance Capacity general antioxidant activity; PACs = proanthocyanidin flavanol polymers.

Table 7
EEG Results for Each Frequency and Regional Electrode Cluster by Condition

Table 8
Systolic and Diastolic Blood Pressure Changes (Post-Ingestion -Pre-Ingestion)

Table 3
Absolute Theta Power for each electrode site for each condition atBaseline, after Treatment, and  Differences (Post Treatment -Baseline) NOTE:Values in the table are Absolute EEG Power uV 2 x 10 4 for ease of presentation.

Table 4
Absolute Alpha Power for each electrode site for each condition atBaseline, after Treatment, and  Differences (Post Treatment -Baseline) NOTE:Values in the table are Absolute EEG Power uV 2 x 10 4 for ease of presentation.

Table 5
Absolute Beta Power for each electrode site for each condition atBaseline, after Treatment, and Differences  (Post Treatment -Baseline)