Black Friday Weekend Sale : Save 30% or more on all our guides and exclusive bundle offers. Learn more

Quick Navigation


Caffeine is a stimulatory anti-sleep compound extracted from coffee beans. Habitual caffeine use leads to tolerance, which dulls several of caffeine’s effects.

Our evidence-based analysis on caffeine features 580 unique references to scientific papers.

Research analysis led by .
Reviewed by Team
Last Updated:

Research Breakdown on Caffeine

1Sources and Structure


Caffeine is commonly found in:

Its main sources in society are coffee or tea, which constitute upwards of 90% of standard intake in society with the remaining 10% coming from cocoa products; these numbers were before the advent of energy drinks and mainstream supplementation, however.[7][8] More recent data (2005 published, collected in the 90s) show that coffee + tea dropped to 83% to accomodate the rise of soda to 12% of intake in the United States.[9]

It is seen as the worlds most popular drug, and 92-98% of persons in North America (1984 data) consume some form of caffeine.[7] More recent data (2005) using data throughout the 90s suggest that 87% of Americans consume caffeine, and in the users of caffeine the average intake was 193mg or 1.2mg/kg bodyweight. Children consume less caffeine, at around 36-56% with Americans consuming more than Canadians on average.[10] On a world-wide scale (1995 data) Norway, Netherlands, Denmark, Germany and Sweden appear to top caffeine intake from coffee at over 300mg/person/day.[2] The general trends should be noted, although exact numbers are bound to be inaccurate. Due to variability in caffeine content of sources (to be discussed) and reverse-calculating caffeine from the sources there will be inconsistencies based on whatever standard is used; variability has been noted for caffeine intake in the US given similar data (196-423mg daily) and the UK (359-621mg daily).[11]

As for main sources of caffeine, the exact caffeine content is variable. Coffee in general can have 40-180mg caffeine per 150mL and tea 24-50mg per 150mL.[8] Soda tends to be 15-29mg per 180mL. This variability tends to undermine the accuracy of society-wide estimates mentioned previously.[2]

Cocoa powder tends to have around 0.21% caffeine by weight and finished chocolate products around 0.017-0.125% caffeine by weight; commerical hot chocolates may contain 4-5mg caffeine per serving.[12]

People love their coffee and their caffeine; there is definite basis for calling this compound the world's favorite drug. Scandinavian countries and Germany consume the most, followed by North American countries and Western Europe. Stereotypical trends of coffee in Scandinavia, tea in Western Europe, and a higher soda percentage in the Americas persist.

1.2Structure and Properties

Caffeine's chemical name is 1,3,7-trimethylxanthine. It is a xanthine compound with three methyl (CH3) groups attached at the 1,3, and 7 carbons on the xanthine backbone.[13]

If it loses a methyl group, it can be converted into similar metabolites like paraxanthine (demethylation at the 3 carbon) theobromine (1 carbon) or theophylline (7 carbon). Further demethylation yeilds either 1, 3 or 7-methylxanthine, and full demethylation yields the basic xanthine molecule.[13]

Most of the metabolism of caffeine goes towards paraxanthine, via the CYP1A1/2 enzyme. Direct conversions to theobromine and theophylline are possible but less prevalent.

Caffeine is a bitter compound, and is sometimes used in sensory research as a standard for bitter. Its perception of bitter may be in part genetically influenced, as assessed by on 0.3 heritability (1 being fully heritable) in some studies.[14]

1.3Formulations and Variants

Inorganic sources of caffeine can be detected via stable isotope analysis and specifically the δ13C, as products that are synthesized in the lab (usually the original carbon source is coal) tend to have differences in values when measured via stable isotope analysis than those that are created in plants via a photosynthetic pathway; this applies to caffeine,[15][16] taurine,[17] sugar (for honey products),[18] and alcohol products[19] amongst others. This detection is mostly just for regulations since FDA regulations say that synthetic, but not naturally occurring caffeine, much be present on the label[15] and there are no known differences in the bioactivity of caffeine depending on source.

Caffeine can be made synthetically or naturally (via plants), and while they treat the body similarly when ingested there are some labelling laws that apply to synthetic but not natural products; laboratory analysis can detect the difference between the two

Caffeine is, by far and large, the most popular ingredient in over-the-counter fat burning supplements and proprietary blends.[20][21]


2.1Oral Interactions

Caffeine's pharmacokinetic parameters can be fairly reliably measured with a saliva reading of caffeine content, given that the saliva is approximately 80% of the serum value of caffeine at any given time and the correlation is fairly high.[22][23][24]

Caffeine can possibly be measured in saliva, as it can accumulate in bodily fluids

Caffeine can be absorbed through the buccal mucosa, a term for oral absorption without needing to swallow.[25][26] This leads to chewing gum infused with caffeine having a quicker absorption of caffeine relative to coffee as it does not need to pass the stomach and intestines to get into the blood.[27]

It is possible caffeine can be absorbed through the mouth and into the blood when consumed via a chewing gum

2.2Gastric (Stomach)

Caffeine is able to increase secretion of gastric acid (HCl) per se, but is one of multiple components of coffee able to do so,[28] but at a weaker potency than these other components (evidenced by decaffeinated coffee causing significantly more gastric acid release than isolated caffeine).[29] The increase of gastrin induced by coffee is not due to caffeine.[30]|published=1976 Mar 19|authors=Börger HW, Schafmayer A, Arnold R, Becker HD, Creutzfeldt W|journal=Dtsch Med Wochenschr]

Appears to increase gastric acid secretion; which may be good for digestion in the stomach yet negative for symptoms of GERD

2.3Intestines and Absorption

When ingested, it has near perfect intestinal uptake of around 99-100%[31][32] up to acute dosages of 10mg/kg bodyweight, the highest studied in humans.[33] This absorption tends to occur almost completely within 45 minutes of ingestion.[34][32][24]

Most caffeine is taken up from the gut 45 minutes after oral ingestion[32] and reaches peak values in the blood between 15 and 120 minutes dependent on individual physiology and vehicle (liquid, capsule, gum, etc.).[33] A slightly more precise estimate may be 30-60 minutes post oral ingestion,[35][36] and 45-60 minutes until 'peak' serum levels tends to be a relatively accurate estimate used frequently.

There does not appear to be a significant splanchic first pass metabolism in regards to caffeine.[34]

Caffeine absorption is slightly delayed from soda and chocolate relative to coffee,[37][38][39] and capsules have faster absorption than does coffee.[38] Caffeine in a chewing gum format is absorbed faster than capsules, as caffeine can be absorbed through the buccal muscoa (mouth).[27] 

Caffeine, when in the stomach, can act upon gastric myenteric and submucous nerves to induce gastric emptying.[40][34] The influx of digestive metabolites into the small intestines may stimulate the gastrocolic reflex[41] and be a possible mechanism behind caffeine causing the need to defecate shortly after consumption. The ultimate result of this reflex mediated by caffeine is an increased anal sphincter contraction power and less neural input needed to induce contraction (decreased sensory threshold).[42] due to these reasons, caffeine is sometimes referred to as a 'cathartic' agent, alongside prunes. Interestingly, chlorogenic acid has been implicated in aiding the laxatative effect of prunes,[43] and is an active component of coffee.[44]

Taken up nearly completely and in under 45 minutes for almost all forms of caffeine and dosages; a slight variance is seen with different vehicles of caffeine such as chewing gum, coffee or chocolate.


As mentioned previously with the graphic, caffeine can be metabolized into dimethylxanthine derivatives (paraxanthine, theobromine, and theophylline) and further metabolized into monoxanthine derivatives and then finally a xanthine molecule. Other metabolic byproducts of caffeine include di and trimethylallantion, mono and dimethyluric acids, and various uracil derivatives.[2]

Caffeine per se appears to be metabolized mostly by the P450 enzyme system in the body, and specifically the CYP1A enzymes (CYP1A1, 1A2). The metabolism by CYP1A of caffeine follows first-order kinetics and is the rate-limiting step of plasma clearance.[45][46] This enzyme, being a rate limiting step and subject to genetic variance, is one of the loci for genetic variance in caffeine. These will be discussed further in the 'Individual differences' section, but they may account for up to 40% variance in caffeine pharmacokinetics.[47][48]

Approximately 84% of ingested caffeine is initially demethylated by CYP1A into Paraxanthine by acting on the 3-carbon.[49] 10% of this Paraxanthine can be further metabolized into 1-methylxanthine by CYP1A, but most tends to be subject to other enzymes (NAT1 can create 5-acetylamino-6-formylamino- 3-methyluracil, a uracil derivative; the rest (up to 90%) is subject to CYP2A6 and converted to 1,7-dimethylurate)[49][50]

Xanthine oxidase can convert the metabolite 1-methylxanthine into 1-methylurate.[50]

In an attempt to give approximate values to bridge oral ingestion and serum values, one study suggested that since ingestion of a single cup of coffee provides a dose of 0.4 to 2.5 mg/kg, it can be estimated that this gives a peak concentration of 0.25 to 2 mg/l or approximately 1 to 10 μM.[2]

The half-life of caffeine varies widely, due to aforementioned variations in CYP1A. One study noted a range of 2.7-9.9 hours[32] with highly similar ranges in another group of persons by the same researchers.[31]

Most metabolism of the caffeine molecule (up to 84%) occurs through the CYP1A1/2 (Aromatase) enzyme, and other enzymes (Xanthine Oxidase, CYP2A6, NAT2) are involved in the entire metabolic cascade or quite briefly on caffeine itself. Variations in CYP1A1/2, either genetic or through supplements, can greatly affect caffeine pharmacokinetics.


Caffeine is hydrophobic enough to tranverse most barriers in the body and is readily distributed to all organs.[51] A steady state volume of distribution appears to occur at 500-800mL/kg with dosages below 250mg.[52][53][54] Doses above 250mg start to increase this amount[54][34] and habitual users of caffeine are associated with greater distribution.[55]

Caffeine appears to be readily distributed between serum and extra-cellular fluid, and then the cell. This appears in the mouse[56] and isolated rat cultures[57] with nearly perfect correlation, although the correlation is likely to be less in humans.[58]

The circulating levels of caffeine in the Extra-Cellular tissue of adipose (fat mass) is not significantly different than that circulating in plasma.[58] This suggests that circulating levels of caffeine correlate well with the levels exposed to fat cells, and the approximate estimate of 1mg/kg bodyweight oral ingestion inducing 5-10uM increases in serum concentration reliable when assessing in vitro studies.

2.6Individual differences in pharmacology

One review has noted that the degree of genetic inheritability for caffeine appears to be in the range of 0.36-0.58 for varying populations (with 1 being completely inheritable) by assessing twin studies, most studies conducted in Caucasians of American or European descent.[59] When looking at heavy caffeine intake (more than 625mg daily) inheritability spikes to 0.77.[60] Studies that assess both genders individually note no significant differences in heritability.[61][62] These differences appear to rise during adolescence, however, and then stabilize; prior to adolescence there is less of a heritable influence on caffeine.[63]

Genetic variations in the CYP1A enzyme that degrades caffeine can alter its ergogenic (performance increasing) effects, with the AA homozygotes outperforming the C allele carriers during endurance exercise.[1]

The AA genotype, in a sample of Caucasians, affects about 46% of persons. The C allele carriers can be either 44% (heterozygous, AC) or 10% (homozygous, CC).[64] AA Genotype is known as a 'fast caffeine metabolizer' and has a higher inducability rate, and degrades caffeine at a faster rate than AC or CC.[64] A higher 'CYP1A activity' tends to be either a genotype of AA, or lifestyle factors that increase CYP1A content.

Swedish persons have higher CYP1A activity than do Korean persons when lifestyle is controlled for,[64]

Activity of the enzyme can be upregulated by smoking, increasing caffeine metabolic rate.[65][64] Heavy coffee drinking may also increase CYP1A activity[66] although perhaps only for the AA genotype.[67]

Some differences exist with other enzymes in caffeine's metabolic pathway, although CYP1A is of most importance. Xanthine Oxidase appears to be slightly higher in Swedish women relative to men and contributed to sex rather than confounding variables, yet shows no differences when comparing ethnicities, in this study Swedes against Koreans.[68] This difference between genders is not overly potent, and some studies fail to find such a significant difference.[69][70]

Four populations have been identified as having a certain percentage of persons as 'slow' Xanthine Oxidisers, with drastically reduced Xanthine Oxidase activity. In particular; Ethiopians (4% of population[71]), Japanese (11%[72]), Caucasian (20%[73]), and Spanish (4%[74]). Even without consideration to demographics, a 2-4 fold difference can be found among tested women.[48]

NAT (N-Acetyl Transferase) appears to work faster in Koreans than it does in Swedes, on average.[68]

2.7Changes in Pharmacology with Lifestyle

In regards to caffeine itself, a 24 hour absence from caffeine does not seem to significantly alter CYP1A2 activity and its subsequent pharmacokinetic profile[75] and intake of coffee is correlated with increased aromatase activity at an extra 1.45-fold increase per 1L of coffee consumed.[76]

Exercise does not tend to influence caffeine pharmacokinetics,[77] although one study did note that a higher Cmax was seen with exercise relative to sedentary persons.[55]

Many nutraceuticals commonly found in foods, such as the Bioflavonoid class of nutrients and select ones such as Quercetin or Genistein can influence Aromatase function and thus metabolism of caffeine. Some other commonly consumed compound such as alcohol (drinking ethanol) or Green tea can influence caffeine's pharmacokinetics. For more information, visit the nutrient-nutrient interactions section or respective pages of the above compounds.

Smoking tobacco appears to upregulate the aromatase enzyme, insofar that smokers tend to have 50-70% less circulating caffeine after ingestion.[78][79][80][81]

Age does not influence Aromatase activity or variability.[82] Gender doesn't influence aromatase activity as it pertains to caffeine for the most part,[83][77] but at least some studies suggest women may have less active aromatase.[84][85]

2.8Differences in humans and animals

The half life of caffeine in rats is typically 0.7-1.2 hours whereas in humans it tends to fluctuate between 2.5-4.5 hours.[86][2] Metabolites also differ, with trimethyl derivatives accounting for 40% of total derivatives in rats but only 6% of total derivatives in man.[87][2]

According to one review, it is generally assumed that 10 mg/kg in a rat represents about 250 mg of caffeine in a human weighing 70 kg (3.5 mg/kg), and that this would correspond to about 2 to 3 cups of coffee.[2] This is with consideration to the differences in metabolic half-life.

3Longevity and Life Extension


Caffeine has been noted to induce nuclear activity of DAF-16 in C. Elegans (nematodes) and promote lifespan by 52%, and inhibiting either DAF-16 or Creb-binding protein 1 (CBP-1; involved in a high level of lifespan variance in mice[88]) abolished these benefits; tannic acid and Baicalein (from Scutellaria baicalensis) also proved to increase lifespan via DAF-16 translocation.[89] 

Caffeine has also been noted to inhibit TORC1 activation to a greater degree than TORC2 (IC50 of 0.22mM; Rapamycin as reference drug with an IC50 of 5.2nM),[90] which via Sch9 (mammalian ortholog S6K[91]) phosphorylates PKA and prevents nuclear actions of Rim15[92] (mammalian ortholog LATS[93]) which normally mediates oxidant defense and stress response genes;[94][95] bypassing TORC1 abolishes the longevity promoting effects of caffeine,[90] and the authors hypothesized this may be relevant to humans (one cup of coffee reaches plasma levels of 1-10µM which may result in 4-8% inhibition of mTORc1, with only 3% being required in yeast to promote longevity[90]).

There appears to be dependence on both mTOR inhibition and DAF-16 nuclear accumulation (usually not co-dependent) associated with the longevity promoting effects of caffeine


4.1Adenosinergic Neurotransmission

Caffeine is most well known for being an adenosine receptor antagonist. It is a competitive inhibitor of adenosine, and fits into the adenosine receptor without activating it and preventing adenosine from acting.[96] As adenosine mediates the perception of drowsiness, preventing its actions results in alertness.[97][98][99] It is non-selective (hits all isomers of adenosine receptors) although it shows slightly more affinity for the A1 receptor,[100] and has been quantified at occupying 7-44% of A1 receptors at varying IV doses of 0.5-4.3mg/kg after short abstinence in caffeine consumers; a concentration of 67uM (in serum) is needed to occupy half of the receptors.[101] There are four isomers of adenosine receptors heterogeneously expressed across the brain, the A1, A2A, A2B and A3 subsets. All subsets are G-protein coupled receptors.

The A1 subset, when activated, promotes sleep and drowsiness by attenuating brain activation and excitation.[102][103] Thus antagonism (preventing activation) serves to promote wakefulness and excitation by preventing their decline. A1 is a G-protein coupled receptor, and is coupled to Gi-1, Gi-2, Gi-3, Go1, and Go2, the pertussin toxin sensitive proteins.[2] Through these G-proteins, activation of A1 causes inhibition of Adenyl Cyclase, some voltage-sensitive calcium channels (N, Q) and activate K+ channels while activating Phospholipase C and D.[104][105]

The A2A subset, when activated, actually promotes wakefulness rather than induces sleepiness.[106] As a G-protein coupled receptor, A2A is coupled to the Gs proteins, which cause activation of Adenyl Cyclase rather than inhibition. These effects, in part, oppose the actions of A1.

A2B does not appear to be activated at circulating normal levels of adenosine (between 30uM and 300uM estimate), although it can be activated at higher concentrations.[2] This may implicate A2B at pathological levels of adenosine but may not be relevant to normal human levels.

The A3 subset also contributes to sedative effects in a similar fashion to A1, although weaker.[107] A3 has a KD approaching 80uM[104][2] and thus isn't likely to be of much practical significance relative to caffeine usage due to its low activity.

Adenosine receptors are upregulated by 20% in response to chronic caffeine intake,[108] although binding efficacy and the post-translational effects of adenosine antagonism are not altered in chronic vs. naive mice subject to caffeine.[109] The receptors may also become sensitive to adenosine agonism, as assessed by human platelet A2A.[110] Following this, adenosine levels circulating in the body may also increase following chronic caffeine ingestion in rats.[111]

The 'primary', or at least most popular, mechanism of action due to the wakefulness effect. Its effects also induce changes in Acetylcholine and Dopamine systems, and interact with the Serotonergic system. The actual effects of adenosine blockade do not disappear with tolerance, but may be diminished by opposing reactions through adenosine and new receptors.

4.2Serotonergic Neurotransmission

The protein content of corticol 5HT1 and 5HT2 receptors (responsive to serotonin) are increased by 26-30% in response to chronic caffeine intake.[108] Additionally, brain levels of serotonin itself may increase.[112] One study in rats fed 30mg/kg bodyweight caffeine, for acute tolerance, noted increased brain levels of 5-HTP (serotonin precursor) and 5-HIAA (serotonin metabolite)[113] and that the increase in serotonin may be more significant in obese mice relative to lean, despite occurring in both.[114] Tryptophan levels are also increased in the brain during caffeine usage.[112]

Serotonin appears to be increased with caffeine ingestion via adenosine receptor antagonism.[115] This is seen since caffeine is a non-selective antagonist, as selective antagonism of A2 subsets reduce serotonin levels.[115][116]

The serotonergic system may interact with the analgesic (painkilling) effects of caffeine, which primarily works via caffeine's adenosine antagonism.[117]

Cessation of caffeine is associated with a transient decline in brain serotonin levels, which is associated with temporary memory impairment.[113] This decrease in brain serotonin may be through adaptations in neurons to reduce serotonin synthesis, as cessation of caffeine is associated with reduced serotonin but not reduced tryptophan.[112] As tryptophan was not reduced but 5-HTP was, the intermediate enzyme (tryptophan hydroxylase) appears to be downregulated; this can theoretically be bypassed with 5-HTP supplementation.

Serotonin (5-HT) and its related compounds (5-HTP, 5-HIAA) all seem to be elevated during chronic caffeine usage, and a short deficiency exists after caffeine cessation due to (plausible) downregulation of tryptophan hydroxylase.

4.3Adrenergic Neurotransmission

Increases in adrenaline appear to be independent of the adenosine receptors, and instead are mediated by adrenergic receptors.[118]

Acutely, caffeine ingestion increases adrenaline and noradrenaline levels in the body of persons who consume it. This increase seems to be fairly dose dependent.[119]

Chronic caffeine ingestion is able to downregulate (reduce) the amount of beta-adrenergic receptors in the brain by up to 25% in some areas.[108] This effect in independent of adenosine antagonism.[120] Caffeine, chronically, has also been shown to decrease circulating catecholamine levels and prevent against diet-induced insulin resistance in rats[121] and prolonged caffeine usage at the same dose causes a lessening of adrenergic (adrenaline-mediated) effects in humans, not just those induced by caffeine.[122]

Seemingly opposite effects with acute and chronic caffeine usage, possibly mediated through changing receptor content.

4.4Glutaminergic Neurotransmission

Caffeine has been implicated in increasing glutamate release in the shell of the nuclear accumbens in naive rats,[123] theoretically downstream of adenosine A1 receptor antagonism and via NMDA receptors.[124]

NMDA activation does not appear to be a factor in caffeine-induced locomotion.[125]

4.5Cholinergic Neurotransmission

Caffeine injections into rats can dose dependently increases acetylcholine level in the medial pre-frontal cortex.[126] This has occurred in the hippocampus, and appears to be secondary thorugh either adenosine A1 antagonism[127] or A2A agonism[106] although blocking the A2A receptor does not abolish the effects of caffeine on acetylcholine release.[106]

Caffeine tolerance (one week of 25mg/kg) is not associated with any tolerance development in acetylcholine release[126] whereas 30 days oral ingestion of 100mg/kg has been associated with tolerance development.[128]

Caffeine appears to be able to induce acetylcholine release from the brain, and this effect is susceptable to tolerance development

Upregulation of muscarinic and nicotinic acetylcholine receptors can occur after chronic exposure to caffeine by about 40-50%.[108]

The level of acetylcholine receptors is increased in response to chronic caffeine treatment

4.6Dopaminergic Neurotransmission

Caffeine injection into non-habituated rats seems to increase dopamine levels in the medial prefrontal cortex (not the nuclear accumbens), which is about the only time caffeine can increase dopamine levels per se.[126] That being said, spikes in nuclear accumbens dopamine levels have been found in other animal studies, showing inconsistencies in the literature.[123] Caffeine tends to interact with dopamine signalling, but independent of actual spikes in dopamine.

Dopamine appears to mediate the increase in physical activity seen with naive caffeine ingestion as evidenced by a lack of apparent effect after dopamine depletion[129][130] or blocking the dopamine receptors.[131] Additionally, the dose-response curves of caffeine on spontaneous activity seem to parallel that of selective D1 and D2 agonists.[132] 

The mechanism that caffeine increases activity by is via antagonizing adenosine A2A receptors (the standard mechanism of caffeine). Adenosine normally suppresses the effects of dopamine on locomotion (via working in opposition on neuronal excitation[124]) in the striatum where A2A and dopaminergic neurons co-exist, and preventing this suppression with an antagonist increases the effects of dopamine on D2, of which include spontaneous activity and (in rats) rotational behaviour when unilateral lesions are induced in the striatum.[133][134] Non-caffeine adenosine antagonists also share this effect on locomotion, further implicating A2A antagonism and dopamine as the cause rather than a separate, unseen effect of caffeine.[135] As for why A2A is mentioned more frequently than A1 in this section, it is since A2A receptors appear to co-exist with dopamine receptors in many parts of the brain (nuclear accumbens, striatum, tuberculum olfactorium) whereas although A1 are heterogeneously expressed in the brain, there is no pattern with dopamine receptors.[2] Interactions with motor control appear to be highly relevant to the striatum, where A2A predominates.

Due to the above interactions, dopamine antagonism (blocking) gives the appearance of caffeine tolerance by preventing locomotion in rats.[136] The dopamine receptors (D1, D2) can also become less responsive to standard dopamine agonists after caffeine tolerance develops[132] although their numbers do not seem to be increased or decreased.[108] Suggesting that the dopamine receptors get desensitized. Cross-tolerance can develop with chronic caffeine usage and dopamine receptors at any dose between 1-100mg/kg bodyweight in rats taken daily, although the higher dosages are more suspect[132] although the lower end of that range, 1mg/kg, doesn't hold much cross-tolernace potential.[137]

Interestingly, these same interactions can be positive. Intermittent caffeine intake (15mg/kg bodyweight, according to this review[2] about 375mg) every other day induces sensitization of motor activity in naive rats[138] and this is through reducing overall A2A receptor count in the striatum[139] and is seen as an indirect sensitization of dopamine receptors.[140]

Dopamine normally mediates locomotion, and caffeine allows dopamine to work better; this seems like caffeine causes spontaneous activity (twitching, foot tapping, urge to dance). This effect appears to be short-lived, and does not affect chronic users.

4.7GABAergic Neurotransmission

Chronic caffeine ingestion can upregulate the GABA(A) receptor subtype by about 65% in some brain areas.[108]

4.8Opioidergic Neurotransmission

Chronic caffeine ingestion can increase the protein content of the opioid receptor (delta).[141]

4.9Memory and Learning

Caffeine is a relatively unreliable agent to increasing memory, as reports on its memory increasing abilities vary widely in the literature.

A few reports look at introversion and extroversion, and note that caffeine (in doses of around 200mg) is a cognitive booster and increases working memory only in those who are extroverts, usually judged by a question of "Are you a lively person" or something similar.[142][143] It shows related cognitive benefits in alleviating performance decline in the same manner.[144] A theory put forth recently[142] hinges on dopamine, either as modulating stress or arousal.[144] This is based off of past correlations with extroversion and dopamine function.[145]

In caffeine naive persons, ingesting 450mg of caffeine before a working memory test reduces performance. This was not seen with lower dosages, and may be due to 'too much' stimulation.[146]


Caffeine possesses psychostimulant effects in naive non-users of caffeine when ingested at 150mg or above, suppressing sedation self-report scores and elevating stimulant ratings.[146]

Mechanistically, the stimulatory effects of caffeine are mediated at first through Adenosine antagonism, specifically A2(A) receptors. This causes enhanced efficacy of dopamine transmission in the striatum where dopamine D2 and A2(A) receptors are both present in high numbers. After more effective D2 agonism, post-receptor effects include an increase in cAMP levels in the dopaminergic neuron which lead to phosphorylation of PKA and phosphorylation of an intermediate known as DARPP-32 which appears to be nearing the end of the stimulatory chain of events.[147] Phosphorylation of DARPP-32 by PKA causes it to become an inhibitor of Phosphoprotein phosphatase 1 (PP1).[148] Inhibition of PP1 enhances the overall dopaminergic signal sent through the neuron.[149]


Mood is a bit tricky, practically. The mere expectation of caffeine intake can improve mood even if no caffeine is consumed, suggesting a fair bit of overlay between caffeine and placebo in this regard.[150] Blinding is important in the research, but practically the effects seen with caffeine could merely be seen with expectation.

Regardless, 6mg/kg caffeine taken an hour before 90 minutes of cycling is able to improve mood throughout the ride in a double blind manner.[151]

4.12Cerebral Blood Flow

Caffeine administration is able to reduce cerebral blood flow[152] and this has been noted via PET scans,[153] Xenon Clearance,[154] MRI,[155] and trans-cranial Doppler.[156] This reduction of blood flow tends to manifest itself as a reduction of cerebral blood pressure.

When comparing the ability of caffeine to alter blood flow between naive and habitual users, it was noted that caffeine administration to habitual caffeine users resulted in a reduction of cerebral blood flow while abstinence increased cerebral blood flow,[157] a later study using near infrared spectroscopy noted that while both habitual and naive users experience decreases in cerebral blood flow in response to low-dose caffeine (75mg), the naive users had an exaggerrated reduction.[158] Due to the difference in naive and habitual users in response to 75mg caffeine[158] yet fairly consistent results in both habitual users and naive with higher (200+) doses,[159][160][161] it is hypothesized that tolerance can develop to a limited degree with low doses,[158] with another study noting that one week deprivation of caffeine in habitual users was associated with a greater reduction of cerebral blood pressure in response to 200mg.[162]

Blood pressure can be reduce in the brain secondary to a reduction of blood flow, which may be slightly attenuated with tolerance (when using low doses) but may be more resiliant to tolerance than other aspects of caffeine.

Interestingly, however, 200mg caffeine reduces cerebral blood flow by around 34.5+/-2.6% in tolerant users yet it does not significantly affect blood oxygenation rates (5.2+/-6.4 average increase), suggesting the reduction of blood flow is not an impairment to neural function.[159] There tends to be a general dissociation between blood oxygenation in the brain and blood flow under the influence of caffeine.[159][160]

Although a reduction of blood flow occurs, this does not appear to be associated with less oxygen reaching the brain


Hypoxia (lack of oxygen) in neural tissue begets reoxygenation, which is associated with oxidative damage to the amygdala and anterograde amnesia; as these mechanisms of amnesia appear to be mediated by adenosine receptors which activate Caspase-1, inhibition of adenosine receptors with caffiene in rats is associated with attenuation of the amnesiac effects of hypoxia.[163] 


Some persons experience anxiety with caffeine ingestion, an effect called 'anxiogenesis'.

The seen increase in brain lactate does not appear to be causative of anxiety, as chronic users who reset their tolerance to experience the same increase in brain lactate do not experience the same spikes in anxiety.[164]

The cause seems to be more genetic. Persons with a polymorphism in the adenosine A2A receptor. The '1976T > C' genotype of A2A appears to be highly related to anxiety from caffeine relative to the CC or heterozygous genotypes.[165][166][167] The polymorphism does not seem to be associated with anxiety in and of itself, only in response to caffeine.[165]


One study conducted in a model of psychosocial stress noted that 3.5mg/kg caffeine was associated with a 233% enhanced adrenaline release and 211% enhanced cortisol release in response to stress when compared to stress under placebo conditions, which was data pooled from naive and habitual users; augmentation of these hormonesdid not appear to be related to habitual caffeine intake.[168] Chronically, one animal study suggests that caffeine can reduce the adverse health effects of stress over a prolongered period of time when the stress and daily caffeine consumption co-exist.[169]

4.16Parkinson's Disease

Parkinson's Disease is a disease characterized by a loss of dopaminergic neurons in the Substantia Nigra pars Compactum (SNc) and subsequent reduction in Dopamine levels in the brain.

Caffeine has been used as a treatment of motor-control side effects from Parkinson's disease (usually when treated with L-DOPA, or Levodopa).[170] The general mechanism of A2(A) antagonism shows effectiveness in non-human primapes for reducing motor control complications after destruction of dopaminergic neurons[171][172] and by itself in doses of 100-200mg appear to reduce motor control complications in Parkinson's Disease in a therapeutic manner.[173] Interestingly, they enhance motor activity yet do not induce dyskinesia in these animals.[171][172] When tested in humans, the A2(A) antagonist in question failed to reach statistical significance in reductions of motor control complications, but only trended towards benefit.[174] These results are different than ones using caffeine in pre-clinical, smaller trials[170] which tend to note some degree of improvement with A2(A) antagonism as an adjunct treatment rather than the sole treatment.[175][176][177]

Epidemiological studies suggest an inverse relationship between lifetime caffeine usage and Parkinson's Disease in men, suggesting a protective mechanism as well as a control mechanism.[178][179][180] It has been called one of the 12 most promising drugs for prevention of Parkinson's.[181]

Some interventions have been conducted on the above protective mechanisms. Co-administration of caffeine and MPTP (a dopaminergic neuronal toxin) reduced the overall destruction of dopaminergic neurons by MPMT.[182] Interestingly, this protective mechanism is hindered by estrogen (and the above societal correlations were seen mostly in men)[183] and does not get reduced after caffeine tolerance.[184]

A decent protective measure to reduce risk of Parkinson's over a lifetime, and could be a nice adjunct therapy for Parkinson's for people who have it; not potent enough to stand alone as therapy, however.


Caffeine ingestion has been found to have reinforcing effects when participants select capsules containing caffeine (color coded with varying caffeine levels to blind subjects[185])

Caffeine has been linked to dopamine release in the nucleus accumbens (a phenomena thought, alongside glutamate release, to correlate highly with addiction[186]) following acute injections of 10-30mg/kg in rats[123] although these results are controversial, as null effects have been reported at similar doses.[126] Ex vivo, caffeine seems to induce dopamine and glutamine release in the nucleus shell rather than extracellular space.[187][123]

Several reviews suggest that there is insufficient evidence to establish a causative role of caffeine in addition[39][188] although it does somewhat interact with reward systems; this may be due to the limited evidence that supports an increase in dopamine release being acute, while chronic studies note that caffeine tolerance is associated with no dopamine release.[189]

4.18Circadian Rhythm

A double-blind, placebo-controlled study in 5 healthy participants who abstained from caffeine for two weeks prior to the experiment found that 2.9 mg/kg of caffeine (roughly equal to a double espresso) 3 hours before bed time delayed the circadian rhythm as measured by melatonin production by about 40 minutes.[190] This delay was less than the 85 minute circadian melatonin delay induced by 3000 lux of light (equivalent to weak sunlight) exposure for 3 hours at bedtime. Combing light and caffeine led to a 105 minute delay.[190] An in vitro study reported in the same paper found that caffeine's effect may be partially accounted for by blocking the adenosine A1 receptor, although other mechanisms could not be entirely ruled out.[190]

One study has noted, in trained athletes, that caffeine supplementation in the morning reverses the lack of strength seen in mornings.[191] In both Wingate testing and resistance exercise, power output appears to peak in the early afternoon with slight suppression of maximal performance in the AM and later PM.[192][193]

Additionally, caffeine has been shown to increase the voluntary maximal load used in resistance training associated with sleep when blinded to caffeine.[194] This suggests an enhancement of cognition or confidence despite sleep deprivation.

5Cardiovascular Health


Caffeine has been implicated in reducing the distribution of fatty acids and cholesterol in rats via reducing mesenteric lymph release of fatty acids into the blood.[195] Whereas both green tea catechins and caffeine seem to reduce intestinal uptake of fatty acids and cholesterol, caffeine was implicated in reducing lymphatic release of fatty acids created in the body into the blood.[195]

Caffeine doesn't appear to reduce cholesterol levels below baseline in rats without dietary cholesterol intake.[196]

5.2Blood Pressure

Increases in blood pressure may be related to adenosine receptors, and specifically two polymorphisms.[197] However, they are short lived. Spikes in blood pressure from caffeine occur significantly in persons who do not routinely use caffeine for 1-4 days, and then subside.[198] Other factors that increase that may influence blood pressure, such as catecholamines or renin, are not increased in persons consuming moderate amounts of caffeine for longer than 4 days.[198]

Alpha(2)Adrenergic receptors are also implicated in genetic variability to blood pressure spikes from caffeine.[197] 

5.3Blood flow and clotting

Caffeine can theoretically cause an increase in blood clotting, as it would inhibit the adenosine A2A receptor on platelets. Activation of this receptor results in reduced clotting, and mice without the receptor or its effects have an increase in platelet aggregation.[199]

However, caffeine appears to also have the ability to resensitize this receptor after prolonged usage which would increase adenosine's actions when caffeine is not consumed.[200][201]

It appears that, acutely, caffeine can increase thrombocytosis (platelet count) when ingested and augments the increase seen with exercise.[202] Long term effects are not known.

Coffee itself doesn't seem to interact with clotting factors much at moderate consumption[203] and tea appears to actually reduce platelet aggregation after stress.[204]

5.4Risk for Heart Disease

A recent Meta-Analysis of over 140,000 persons from 5 large scale studies[205] found a J-shaped curve where 4 cups of coffee daily was associated with a significant decrease in heart attack risk (11% lower risk) and 1-3 cups daily had ranging values of 4-10% decreased risk compared to persons who did not consume coffee. The Relative Risk (RR) was 0.89 (4 cups daily) to 0.96 (1-2 cups daily), suggesting a minor but statistically significant reduction in risk associated with coffee; a slight increase (RR=1.03) was noted with 11 cups daily.

If anything, the data suggests a very small influence of coffee and caffeine, chronically, on heart health. The RRs of the meta-analysis are quite minor, and the potency of the effects are not too clinically relevant

6Interaction with Glucose Metabolism

6.1Blood Glucose and Insulin Sensitivity

Caffeine can increase blood glucose by reducing glucose disposal in muscles, via interfering with the actions of insulin.[206][207][208] During a sedentary state, caffeine can temporarily induce insulin resistance by about 13%[207] and reduce glucose disposal by 24%.[206] This, however, is not accompanied by an increase in insulin secretion as insulin levels remain fairly constant throughout caffeine ingestion.[208]

This interference is partially negated during exercise, however.[209]

The increasing of blood glucose seen via reduction of insulin sensitivity does not appear to be due to adenosine antagonism, seen as caffeine's main mechanism of action.[207]

Ingestion of coffee, however, can increase insulin sensitivity acutely in rats,[210] possibly by reducing the rate of glucose appearance in circulation mediated by chlorogenic acid.[211][212][213]


At rest, the amount of dietary carbohydrate stored as glycogen under the influence of caffeine appears to be reduced by 23% in sedentary persons,[206] while, in glycogen-depleted persons, it can enhance the rate of glycogen accrual.[214] In the experimentally induced glycogen depletion state, 8mg/kg caffeine was able to increase glycogen resynthesis rates by 66% over 4 hours when compared to the same amount of carbohydrates alone.[214]

In humans, the replenishment of glycogen during depletion is not associated with either AMPK (able to increase glucose uptake) and likely not Akt, as assessed by expression in a group experiencing enhanced glycogen replenishment against the same amount of carbohydrates without caffeine.[214] An enhanced phosphorylation (activation) of calmodulin-dependent protein kinase (CaMKThr286) was seen in the caffeine + carbohydrate group near the end of the study frame (with no differences at baseline or 1 hour after exercise), at times where CaMKThr286 was similar, rates of glycogen replenishment were similar. It is suspected that the calcium (Ca2+) release induced from caffeine can enhance CaMKThr286 activation,[215] which causes enhanced glucose uptake into muscle cells specifically.


Coffee itself is very well correlated with a reduced risk of developing diet-induced diabetes, although whether this is due to caffeine or other compounds such as chlorogenic acid that have shown anti-diabetic effects is not known.[216]

The first study to note a correlation between higher coffee intake and lesser risk of diabetes was in 2002 in which those who drank coffee at 2 cups daily were half as likely (Relative Risk Ratio of 0.50) to develop diabetes.[217] Various surveys conducted since then establish the correlation between coffee and reduced risk for diabetes, which is independent of lifestyle and race[218][219][220][221][222] and two meta-analysis' support these results.[223][224] The latter meta-analysis suggests a 7% reduction in diabetes risk with each cup of coffee consumed daily, and show that the relation holds true for decaffeinated coffee.[224][216]

The above promise for risk reduction of diabetes may not hold true for isolated caffeine. In addition to the latter meta-analysis finding benefit with decaffeinated coffee, isolated caffeine seems to increase blood glucose and induce insulin resistance transiently after consumption.[206][207][208] If mediated through adrenaline, long-term effects on insulin sensitivity may be distinct from acute, as the adrenaline increasing effects of caffeine disappear when tolerance develops over 7 days.[198]

Coffee can reduce the risk of diabetes, not overly significant from a clinical standpoint but it appears to be reliable. This may not be due to the caffeine content, but may be due to any of the myriad of other bioactives, such as chlorogenic acid or ferulic acid

7Obesity and Fat Mass

7.1Metabolic Rate and Weight Loss

30 minutes after an oral dose of 4mg/kg caffeine (in obese women), a metabolic spike was seen in all subjects but to a widely varied degree; this spike is able to reduce body weight when in conjunction with a low-calorie regimen in all subjects though.[225] This dose, 4mg/kg bodyweight, is approximately equal to 3 cups of regular coffee (average caffeine content) for persons with a BMI in the normal range.[226]

Despite the high variance seen, average increases seem to be around 34kJ/m2/hour, which translates to 8.1 calories per hour for every meter of body surface area a person has.[226] The standard 'average' for adult humans of normal weight is 1.73m2, which translates into 14kcal per hour. Some studies note higher values (32.4kcal/h) with higher dosages of caffeine (400mg).[227]

The increase of metabolic rate may be dose dependent and be somewhat related to catecholamine release.[228][226] The increases in catecholamines seem to be correlated with serum caffeine levels[227] and increases linearly with increasing dosages up to 9mg/kg bodyweight.[119]

When looking at caffeine usage on the metabolic rate of athletes and sedentary persons, the end results conflict in studies. One study using 4mg/kg bodyweight in high level aerboic athletes (marathon runners, cross country skiers) found that caffeine significantly increased metabolic rate more than sedentary persons, achieving 80kJ/kg/min for the 120 minutes after ingestion (athete) vs. just under 60kJ/kg/min for sedentary, with no significant differences prior to testing.[229] The dose in the previous study equated to an average of 280mg, and another study using 300mg found the opposite results and that sedentary persons had an increased response.[230] Both studies measured O2 consumption at rest.

This heightening of the metabolic rate is fairly well studied, but long term interventions in humans with caffeine tend to suggest caffeine is not a good long-term fat loss agent. 24 weeks of 200mg caffeine intake does not result in fat loss[231] nor 100mg for 16 months when paired with 20g soluble fiber.[232] One study that was divided into a 4 week low-calorie weight loss phase and a 3 months maintenance phase (supplemented with green tea catechins and caffeine) noted, however, that caffeine intake was associated with greater weight loss during the acute phase but was not significantly related to long-term weight maintenance.[233]

Caffeine has the mechanisms to increase metabolic rate and fat loss, but this is a mechanism that one can become desensitized to. Best usage for caffeine as a stimulant for weight loss should be intermittently and in conjunction with caloric deficits.

7.2Food intake

Caffeine has been implicated in suppressing food intake in rats[234] although studies in humans are less promising. One study noted no significant suppression of appetite or food intake in men with 3mg/kg bodyweight caffeine[235] whereas another did note suppression of food intake in men, but not women, with 300mg caffeine.[236]

7.3Substrate Cycles (Mechanism)

'Substrate Cycles' tends to refer to cycles such as the Cori Cycle (lactate) or free fatty acid-triglyceride cycles. Caffeine has been suggested to stimulate these cycles and encourage substrate turnover. This stems from correlations in plasma between caffeine and lactate as well as caffeine and triglycerides (as well as glycerol) in which regression analysis excludes caffeine as a causative variable.[227]

The spike in lactate has been found in other studies at rest[237][237] and during exercise.[238] Studies that look at lactate before and during exercise don't note much of a difference after caffeine ingestion.[237]

7.4Phosphodiesterases (Mechanism)

Caffeine is known as a non-selective phosphodiesterase inhibitor.[239] As phosphodiesterases mediate the breakdown of cyclic AMP (cAMP) levels, their inhibition should theoretically lead to increased cAMP and increase lipolysis. That being said, it is unlikely that biological effects seen with caffeine supplementation will be due to PDE inhibition, as the dose of caffeine required to sufficiently inhibit PDE enzymes is beyond normal human consumption.[240][241][152]

8Skeletal Muscle and Physical Performance

8.1Power Output

Studies show that a dose of caffeine will improve strength performance. One mechanism may be reducing the users perception of pain [242] while increasing calcium mobilization in the muscle cells,[243] which can increase power output.[244]

Despite biological plausiblity existing, experiments run with caffeine do not tend to show increased performance in a 1 rep max test.[245][246] It has been noted[191] that these null effects may be due to the time of force contraction being relatively long and contraction velocity is normally below 0.4 m s−1;[247][248] this would make a 1RM test insensitive to detect changes in power output.

Sub-maximal power output, in the 6-12 rep range, tends to be increased with caffeine as assessed by increased work volume.[245] At around 60% 1 rep max, about 11-12% greater workload can be achieved with an oral dose of 6mg/kg bodyweight.[245][249]

It has been noted that there may be 'responders' to caffeine intake as it pertains to strength increasing.[250]

It is possible for caffeine to increase 1-rep max strength, although studies seem to show null results. Muscle cells appear to have increased power output as assessed by Wingate testing, which may have implications for weight training in trained individuals.

8.2Anaerobic Cardiovascular Exercise

The delay of muscle fatigue may be part neurological and part on the level of the skeletal muscle.

Caffeine ingestion in a population of tetrapalegics still shows anti-fatigue effects of caffeine via electrical stimulation.[251] In healthy persons, caffeine still potentiates muscle contraction as induced by electrical stimulation.[244] As electrical stimulation is independent of the central nervous system, caffeine seems to in part act on either neuro-muscular junctions or contractile apparatus in muscles.

Human studies note that intermittent sprinting is increased in experienced athletes consuming 5-6mg/kg bodyweight caffeine (most common research doses)[252][253] or in general with varying dosages in the higher range.[254][255] One study conducted in physically active men noted a 1.4% reduction in fastest sprinting speed with caffeine.[252] 

8.3Aerobic Exercise

Caffeine supplementation is able to increase endurance performance significantly. Performance (typically measured as time to exhaustion at 80% VO2 max) is increased in a dose dependent manner, and significance has been noted at doses of 2.5mg/kg body weight and 5mg/kg bodyweight, which would be 227mg and 454mg for a 200lb individual respectively. One study specifically looked at comparing these dosage ranges (used 3mg and 6mg/kg) and found no statistically significant difference between the two in trained cyclists.[256] The effects in older individuals (70+) are the same as youth.[237]

Improvements in endurance performance measured by time to exhaustion hover typically in the range of 1.2 - 1.4 fold increases, with the effect being clinically significant only in those not adapted to caffeine usage[257]. Those who self-classify as caffeine users do not have significant improvements in endurance performance.

An ingested dose of caffeine can last in the body for up to 6 hours in regards to improving endurance performance. Redosing for two-a-day events separated by a 6h time frame does not appear to be needed to maintain the increased performance as measured by studies measuring multiple daily exercise sessions[258].

One of the theories of caffeine increasing aerobic performance, by increase free fatty acids or lowering potassium levels, does not appear to be fruitful.[259] The theory of increasing endurance performance by increasing catecholamine (adrenaline, in particular) levels also appears to be less promising, as endurance performance peaks do not correlate well with maximum adrenaline secretion from caffeine.[119]

One possible mechanism is the reduction of perceived effort, and indirectly doing more work without trying harder. Many studies note a reduced perception of effort and pain associated with caffeine ingestion.[238][242][151]

8.4Muscle Soreness

In a small group of caffeine naive persons given caffeine an hour prior to eccentric exercise (followed by a rep out set) noted that supplementation was able to reduce delayed onset muscle soreness when measured 48-72 hours after exercise, but not immediately (24 hours) nor after 4-5 days after exercise.[260]

8.5Interactions with Sleep

Caffeine is known to increase power output in both a waking and sleep deprived condition (less than six hours) in elite athletes at the oral dose of 4mg/kg,[194] but the increase in power and workload is greater in sleep deprived persons and reached a level which is not significantly different than the placebo group of those with normal sleep.[194] This increase was associated with an increase in chosen workload, which only occurred in the participants who reported feeling the effects of caffeine.[194]

The impairment in voluntarily chosen weights after a poor night of sleep appears to be alleviated with a pre-workout dosage of caffeine, but seems to only occur in those who percieve an effect of the caffeine

9Interactions with Hormones


Caffeine, through chewing gum, has been implicated in increasing the testosterone response to exercise; testosterone increased 53% as a response to exercise in both groups, and a further 12+/-14% with 240mg caffeine.[261] At 800mg before resistance training, it shows an increase of 15+/-12% above placebo in trained athletes;[262] These results have been somewhat replicated in a model of sleep deprivation with 4mg/kg caffeine, with an increase in testosterone being noted (6.5-12.9%, with a larger effect in those without sleep deprivation).[194]

All studies used salivary testosterone measures despite sample sizes of 9, 24, and 16 respectively, and did not measure free (bioavailable) testosterone.[261][262] The sleep deprivation study noted that the increases in testosterone correlated well with work volume and percieved stimulation which suggested the seen increases in testosterone were secondary to work performed,[194] and this theory works well with studies without exercise (the following on rifle marksmanship) where no statistically significant differences are seen.[263]

Caffeine appears to reliably increase testosterone when consumed prior to exercise possibly due to more work performed, but the magnitude of this increase (12%) is small and the variability (+/-15%) is quite high; an unreliable intervention for testosterone boosting

One animal study has noted increased testosterone levels with chronic cola ingestion independent of exercise, but aside from also existing with an increased in estrogen the mechanism and whether this can be attributed to caffeine per se is unknown.[264] A later study in humans using 5 cups of coffee for 4 weeks noted an increase in testosterone and a decrease in estradiol in men, but not women.[265]

Limited evidence for a testosterone boosting effect independent of exercise, but a build-up effect over time may exist

9.2Sex Hormone Binding Globulin (SHBG)

Sex Hormone Binding Globulin (SHBG) is a protein carrier in the blood that can sequester and hold both androgens (testosterone) and estrogens (estradiol, with affinity for the former.[266] High SHBG status can suppress the effects of steroid hormones despite not reducing their content in the blood.

In postmenopausal women, caffeine intake and SHBG in serum appear to be positively and reliably correlated independent of source of caffeine.[267][268][269][270][271][272] This relationship is stronger among overweight/obese women, has been calculated at approximately 13% when comparing the highest group (371mg or more) against the lowest (71mg or less), and does not appear to exist in pre-menopausal women.[269] This positive association correlates nicely with the decrease in bioavailable testosterone seen in postmenopausal women with higher caffeine intakes.[267] When tested in an intervention setting in overweight healthy adults, 5 cups of coffee daily spread out over the course of the day failed to replicate the effects seen in epidemiological research; no relationship was seen between coffee intake and SHBG.[265]

Reliable positive correlations between postmenopausal women and an increase in SHBG from caffeine, which may not apply to other demographics


In sedentary persons, 3.3mg/kg caffeine at rest in habitual users after 10h abstinence is sufficient to increase serum cortisol 30%; secondary to an increase in ACTH by 36%.[273] These changes seem to be sensitive upon time and loading, with injections during sleep (a time period when caffeine is low) causing exacerbated spikes in cortisol and ACTH[274] and repeated caffeine dosing causing the increase seen in waking hours, when caffeine tends to be higher, insignificant at rest.[275]

Caffeine can increase cortisol at rest, which is seen in habitual users even after a short (10 hour) abstinence in a resting state

High (800mg) doses of caffeine may increase cortisol by up to 52+/-44%.[262] A dose of 3.3mg/kg (in grapefruit juice) in otherwise healthy young men with a 1-5 cup of coffee a day habit has also noted an increase in cortisol when paired with exercise, but this study did not show differences between the post-caffeine group and group with caffeine and exercise;[276] an increase in cortisol relative to placebo has been noted in trained persons after cycling for 2 hours at 65% VO2 max in response to 6mg/kg.[277]

Physical Stress increases secretion of cortisol, and this appears to be augmented under the influence of caffeine in sedentary or low-active persons

In competitive level cyclists given a low dose of caffeine (250mg), a reduction in exercise-induced caffeine levels is seen.[261] After a 60h washout in competitive cyclists given 6mg/kg caffeine (a dose known to increase serum cortisol in sedentary) no apparent increase in seen before or after exercise[278] and no apparent increase exists in resistance trained men given 5mg/kg.[279]

The effects of caffeine on cortisol may be attenuated or reversed in higher level athletes given low doses

In response to mental or occupational stress, caffeine works in a similar manner with 250mg able to augment stress-induced cortisol spikes in persons with a 1-5 cup of coffee a day habit with no differences in gender,[280] with slightly higher exacerbations following 3.3mg/kg;[281][282][283] caffeine increasing cortisol appears to be dose-dependent regardless of stressor placed.[284] This increase appears to be additive rather than synergistic,[283] and lower doses (200mg) after a single day of deprivation do not always result in cortisol increases.[285]

Caffeine appears to work similar in cognitive stress as it does in physical stress, augmenting a stress-induced spike in cortisol concentrations; seems more reliable at higher doses

Conversely, one study pairing smoking with caffeine intake (in smokers with a moderate caffeine habit) noted that smoking per se increased cortisol which was independent of the nicotine content, and that 150mg caffeine failed to modify the increase in cortisol.[286]

Studies that measure urinary levels of cortisol note a 13% reduction when measured 4 hours after 300mg,[287]

10Interactions with Cancer

10.1Skin Carcinoma

Caffeine is associated with a reduced risk of skin carcinoma, although with a Relative Risk Ratio of 0.82 for women and 0.88 for men when compared to the lowest quintile; a moderately weak correlation.[288]

10.2Colon Cancer

Coffee has been associated with decreased risk of colorectal cancer,[289] but this does not seen to be attributable to caffeine due to the association with both caffeinated and decaffeinated coffee.

11Interactions with Various Organ Systems


Persons exceeding 500mg caffeine daily appear to be at a slightly greater risk of exfoliation glaucoma or exfoliation glaucoma suspect (lead determinant of Glaucoma) when compared to abstainers, and a nonsignificantly increased risk rate relative to low consumers (125-500mg); the Relative Risk (RR) for 500mg relative to abstainers was 1.66.[290]

12Tolerance, Dependence and Withdrawal

12.1Metabolic differences between Naive and Tolerant users

Neurally, those who have adapted to caffeine intake report less neural stimulation[291] and become less attentive after caffeine usage.[292] Spikes in systolic blood pressure that occur in naive users to caffeine do not occur in habitual users,[293][294][295][296] the degree appears to be 3-6mmHg systolic blood pressure. Diastolic appears to increase around 3mmHg as well, although one study using coffee and mental stress noted a 6mmHg decline.[293] That being said, not everybody develops tolerance to caffeine's influence on blood pressure spikes; about 50% of subjects do not either over 5 days[297] or 4 weeks of supplementation.[298]

Brain lactate increases quite drastically in persons unaccustomed to caffeine and exhibit minor changes in those accustomed to it, and a 1-2 month cessation from caffeine can restore the rise in brain lactate from caffeine to the levels seen with unaccustomed users.[164] This was hypothesized by the authors as possibly being due to the combination of decrease cerebral blood flow and increase glycolysis.[164]

The reduction in cerebral blood flow associated with caffeine does not seem to differ between naive and chronic users.[157] And the ergogenic effects of caffeine being able to increase intermittent sprint performance do not seem to be different between naive and chronic users.[299]

12.2Withdrawal Symptoms

Withdrawal of caffeine, defined as a short-term rise of adverse effects in response to caffeine deprivation, tends to last for approximately 3 days after tolerance develops.[300] Some effects may last slightly longer or shorter, such as headaches which range from 2-6 days.[301] Headaches are commonly seen as the most problematic side-effect of withdrawal, and are caused by a temporary increase in cerebral blood pressure.[302][303] The magnitude of headache is positively correlated from the daily dose of caffeine one 'comes off' of, with the larger dose resulting in more of a headache.[157][155]

Cognition tends to decrease, and most commonly the aspects of attention and focus (decreased) and fatigue (increased).[304][185] In a study on managers (average intake of 575mg caffeine) who were subject to a blinded reduction in caffeine, systolic blood pressure was reduced and cognition on a few measured parameters (applied initiative, decisions per hour,) was reduced as well. Performance on a highly complex managerial task was not adversely affected by caffeine deprivation, however.[305] It appears that baseline cognition is reduced but cognitive potential may be unhindered.

Various other mood effects, such as increased irritability and sleepiness alongside decreased alertness are also common.[306][307][308] Productivity, as assessed by a combination of attentiveness and actions per minute, is also decreased.[305]

12.3Withdrawal Mechanisms

As measured by EEG, an increase in theta and and alpha absolute power are seen acutely after caffeine ingestion is stopped, during the withdrawal process[309][306] which is slightly alleviated with caffeine ingestion. Increased theta waves are correlated with drowsiness and decreased alertness, and thus this reading serves as some tangible evidence beyond subjective measures.

Withdrawal symptoms can be reduced by blocking NMDA receptors from acting, and has been shown to do so in using memantine and neramexane.[300]

It was noted earlier that adenosine and acetylcholine receptors are upregulated in response to caffeine; these changes begin to reverse after 7 days cessation.[141]

During withdrawal from caffeine, rats still appear to be sensitive to amphetamine-based compounds as it pertains to increasing spontaneous activity.[310]

12.4Tolerance Symptoms

Caffeine appears to be able to invoke cross-tolerance with compounds that act on either the D1 or D2 dopamine receptors, but do not impose cross-tolerance on drugs that invoke both receptors simultanously (amphetamines, cocaine, methylphenidate).[132]

It seems possible to consume caffeine routinely without developing tolerance, although the dose must be low. In one rat study 18.3-27mg/kg daily did not result in tolerance whereas 33-57.6mg/kg did.[311]

12.5Mechanisms of Tolerance

Caffeine tolerance, in regards to neural stimulation, has been described as an insurmountable problem; one that cannot be overcome by simply increasing the dosage.[312][311][313] Insurmountable antagonisms are depicted as altering a dose-response curve by actually flattening the curve, rather than just shifting it to a higher dosage.[314] Insurmountability is also referred to as 'non-competitve antagonism' as it does reduce the effects of agonism, but not by direct binding to a receptor.

Possibly explanations for insurmountability include receptor internalization or a changing allosteric binding site or structure of the receptor have been proposed.[314][315] However, this is receptor insurmountability. That seen with caffeine is currently the phenotype insurmountability (salient results of caffeine ingestion) while the receptor or protein that mediates this effect is not known.

It has been proposed that, due to caffeine's ability to act as a competitive inhibitor of adenosine receptors, that tolerance may be vicariously due to upregulation and proliferation of adenosine receptors. This theory was due to caffeine administration causing increased adenosine receptor count[312][316] amongst others.[317][316]

This theory was questioned and tested, as there was no precedent for receptor upregulation causing a reduction in antagonist potency and the theory did not fit with dose-effect charts.[109] One study that compared spontaneous activity of rats given caffeine either naive or in a adapted (post-tolerance) state noted that while differences did exist between groups indicating neural stimulation, that both groups' adenosine receptors were equally effective in suppressing forskolin-induced cAMP build-up, and that caffeine was equally effective in inhibiting adenosine's actions.[109] Binding efficacy of caffeine to the receptors was unchanged.[109] Additionally, the notion of insurmountability is seen after caffeine ingestion but does not seem to apply to the adenosine receptor.

It is currently not known what causes tolerance. We are aware of what tolerance exists of by assessing differences in naive and chronic users, but do not know the mechanisms that underlie tolerance.


Dependence may refer to one of two phenomena. Drug dependence is the act of continual self-administration of a psychoactive compound despite potential harm to the self or society.[318] Physical dependence is required intake of a compound in order to maintain a certain activity or state of the body.[7]

13Nutrient-Nutrient Interactions


Theanine is an amino acid that is slightly sedative, it is commonly seen as synergistic with caffeine in regards to cognition.

L-Theanine is synergistic with caffeine in regards to attention switching[319] and alertness[320][321] and reduces susceptibility to distractions (focus).[321][322] However, alertness seems to be relatively subjective and may not be a reliable increase between these two compounds,[319] and increases in mood are either present of absent.[323][319][324] This may be due to theanine being a relatively subpar nootropic in and of itself pertaining to the above parameters, but augmenting caffeine's effects; some studies do note that theanine does not affect the above parameters in and of itself.[325] Due to this, any insensitivity or habituation to caffeine would reduce the effects of the combination as L-theanine may work through caffeine.

L-Theanine does not appear to be synergistic with caffeine in regards to attention to a prolonged and monotonous task.[326] Additionally, the addition of L-Theanine to a mixture of caffeine and green tea catechins does not further increase the fat loss potential of the aforementioned pairing in rats.[327]

The mechanism of action may be in reducing caffeine's stimulatory effects (as measured by EEG) while not inhibiting other non-stimulatory caffeine mechanisms of action,[328] which would assert L-Theanine in acting as a modulatory agent.[325]

The dosage used in many of the above cognitive enhancing studies were either low doses (80-100mg L-Theanine, 50mg caffeine). In supplemental form, a 200mg:200mg combination is routinely used.

Additionally, L-Theanine may attenuate the adverse effects on sleep from caffeine.[329]

Seems to work well for selective attention and focus, and is quite variable on mood and general alertness. L-Theanine just appears the quell the peak stimulation of caffeine yet not fully prevent its alertness increasing effects.


Taurine is a non-essential sulfur containing amino acid derived from methionine (essential amino acid) that is sometimes paired with caffeine in Energy Drinks.

In one study on surgeons suffering from sleep deprivation and put through a surgical test, the combination of caffeine (150mg) and taurine (2g) seemed to improve speed of the test yet did not improve errors relative to a rested state. It was insignificantly better than caffeine in isolation and differed from placebo in time to complete the test.[330]


Caffeine is one of the three components, alongside ephedrine and aspirin, which comprise the ECA stack. Ephedrine and caffeine have potent synergism in relation to fat burning.

Many of the studies on the ephedrine page are done with EC, or an ephedrine:caffeine combination at 25:200mg Ephedrine/Caffeine taken three times daily. This combination appears to be a reliable fat loss agent and appetite suppressant, although it should be noted the combination has differing effects than the plant source Ephedra which may not share the same safety profile.

The cumulative fat burning synergism between the two compounds may be around 64% more than the expected additive value,[331] which comes out to a doubling of the effects of ephedrine in isolation (ie. caffeine in addition to the synergism is just as potent as ephedrine in isolation).[332] The combination increases adrenaline much greater than either compound alone, and possible benefits (stimulant effects and appetite suppression) or drawbacks are increased with any effect mediated by adrenaline.[333]

Highly recommended to use caffeine alongside ephedrine. Epherine is not needed to derive the most benefit from caffeine, but the opposite seems to be quite true.

13.4Methamphetamine (Meth)

A combination of caffeine and Meth in a 3:1 ratio is called 'Ya-Ba' in Thailand and the surrounding areas.[334][335] The addition of caffeine is said to increase the stimulatory effects of Meth.[336][335] The combination of the two at individually non-toxic dosages can exert toxic effects, and is theorized to be caffeine's non-toxic actions increasing the potential toxicity of Methamphetamine.[335] Specifically, Meth had a 24-hour survival rate of 83% and caffeine 100% at the dosages used (12.5mg/kg caffeine, 5mg/kg meth; four dosages injected) yet the combination had a 33% survival rate.[335]

The mechanism is hypothesized to be increase dopamine release leading to increase oxidation of neurons (dopamine quinones, hydroxyl radicals, superoxide radicals, etc.),[335] or that caffeine negates adenosine's protective effects on neurotoxicity.[337] Basically, caffeine can make dopamine transmission occur at a greater rate, but the rate that this occurs is one of neuronal toxicity.

Another amphetamine based compound, MDMA, has its monoamine inducing effects synergistically increased with caffeine[338] which sometimes results in tachycardia, hyperthermia, and subsequent death.[338][339][340] The idea of low-dose (250mg or so, extrapolating from animal models) caffeine potentiating the effects of drugs extends to cocaine and other amphetamines, and theoretically any drug that agonizes both D1 and D2 receptors simultanously.[132]

Theoretically the interactions sound good, but the potency of the street drugs MDMA and Methamphetamine are at a level where further synergism seems to just result in a reduced safety threshold and further potential harm. Not an advised combination.


Alcohol, otherwise known as ethanol or drinking alcohol, interacts with caffeine via reducing the drinker's perception of impaired judgement and actions from alcohol.[341] This decreased perception of impairment does not accompany an actual decrease in level of impairment, however.[342][343][344] The pair may also simply blunt fatigue and sedation from alcohol by preventing adenosine from acting.[345][346]

Additionally, the most common source of caffeine (energy drinks) may have another synergistic mechanism of action. The carbonation enhances the speed of alcohol absorption.[347]

It has been suggested that the combination may be a problem from a social stand-point, as the reduction of perceived impairment coupled with the anti-sedative effects may increase overall alcohol intake in general.[341]

Interestingly, the social correlates of alcohol (social validation, sexual behaviour, desire of masculinity among males, illicit drug use and seatbelt omission) are also correlated with caffeine use as well, although causation has not been established.[348][349][350]

At least one rat study noted that it could interact with anxiety when consumed with alcohol during adolescence, by synergistically reducing anxiety in males but increasing it in females.[351]

13.6Green tea catechins

Green tea catechins are in part synergistic with caffeine, or otherwise have effects which practically go well with caffeine usage.

The primary green tea catechin, EGCG, is able to attenuate (reduce) the spike in catecholamines (adrenaline, noradrenaline) in the blood that caffeine induces.[352] Due to this, the expected rises in blood pressure and heart rate that may be a side-effect of caffeine are potentially negated[352] and this interaction may also explain attentuation in anxiety.[353] Additionally, EGCG may be effective at reducing the effects of dopamine agonism from a wide variety of compounds including caffeine and meth[354] although circulating levels of dopamine are not reduced.[352] These studies investigated EGCG and caffeine at 15 and 30mg/kg bodyweight and 25mg/kg bodyweight, respectively, via intraperitoneal injection.

The above is a bit odd, as EGCG is a known inhibitor of catechol-o-methyltransferase;[355] this enzyme degrades catecholamines, and inhibiting it should increase catecholamine levels. That being said, EGCG is also implicated in reducing catecholamine secretion from the adrenal glands induced by cholinergic neurons.[356][357] This latter mechanism is consistent with the observation that EGCG and caffeine results in less adrenaline and noradrenaline but not less dopamine, which does not originate from the adrenals.[352]

That being said, the dose of EGCG used in all the above experiments was quite high. Green tea catechins have a low bioavailability overall, and the above dosages (15 or 30mg/kg bodyweight) may be unpractical.

Green tea catechins and caffeine have complex interactions, but the general notion is that they attenuate stimulation (a smoother peak) and may be slightly synergistic in regards to fat burning


Propranolol is a member of the beta-blocker class of pharmaceuticals that is used to control blood pressure and heart disease risk in the obese and metabolically ill. Its 'beta' refers to the beta-adrenergic receptors, and its blocking action makes it the opposite of ephedrine.

Caffeine and propanolol appear to antagonize each other in various ways. Propanolol can reduce sperm motility which is countered by caffeine[358] at low to moderate dosages.[359]

Additionally, propanolol has been implicated in reducing the thermogenic response to a fat burning supplement which contains caffeine (alongside green tea catechins, L-tyrosine, and capsaicin).[360]


Danshen, otherwise known as Salvia miltiorrhiza, is a Chinese medication that is commonly seen as heart healthy. The active compounds, the tanshinones, can inhibit the CYP1A enzymes that degrade caffeine to paraxanthine and increase the time caffeine spends in the blood by 11-25% and increase half-life by 12-16%.[361][362] This effect occurs at 100-200mg/kg daily in rats, a human equivalent dose of 16-32mg/kg bodyweight. The effects over a period of 7 days appear to be the same as a single bolus of both.[362]


Genistein is one of the two main isoflavones found in Soy, the other being Daidzein. In a trial using 1g Genistein daily for 14 days in healthy female volunteers, it was shown that Genistein consumption decreased metabolism of caffeine through CYP1A2 (28-51% reduction) as well as through Xanthine Oxidase (24-32% reduction) and shuttled caffeine to be metabolized through CYP2A6 (29-66% increase) with no effect on NAT2.[49] These four enzymes are the only ways caffeine can be metabolized, and Genistein causes a shift in metabolism.

14Safety and Toxicity

14.1General Toxicology

The serum concentration of caffeine that is usually seen as toxic is in the range of 200uM.[34][363] If we assume the estimations that 1mg/kg bodyweight oral ingestion equates to a 5-10uM increase in caffeine approximately an hour after ingestion[51][34] then a toxic dose of caffeine is in the range of 20-40mg/kg bodyweight, or 1.8-3.6g for a 200lb person.

14.2Caffeine and pregnancy

Various phenomena exist to suggest that caffeine may be harmful for unborn infants. Already established was a very effective intestinal uptake in humans (so it can be assumed most caffeine ingested is in the blood), however there does not appear to be a placental barrier to caffeine[364] nor does there appear to be a blood-brain barrier in the fetus.[365] Caffeine half-life is also increased from 2.5-4.5 hours in adult humans to a varying value between 75 and 100 hours in newborn or premature infants, with greater values seen in the younger.[366][367]

There is insufficient evidence at the moment as to a final stance on caffeine during pregnancy due to the problems of extrapolating rat studies onto humans.[368][369] Although it may not be needed to eliminate caffeine usage, it may be prudent to reduce caffeine intake, especially in the first trimester as the fetus is more sensitive at this time.

14.3Side Effects

In one study with healthy non-coffee drinkers, a single dose of 250 mg of caffeine increased mean blood pressure by 14/10 mmHg one hour after caffeine ingestion; however, the blood pressure effect resolved within 4 hours.[370] In hypertensive or hypertensive-prone subjects the pressor response to caffeine seems to be more pronounced than in normotensive subjects.[371] One study revealed that caffeine ingestion of 3 mg/kg in the hypertensive group showed persistent elevation in diastolic BP for 3 hours, whereas the increment of diastolic BP became smaller in the normotensive group 90 minutes after caffeine ingestion.[372] 

Caffeine intake of 250 mg or greater can acutely raise blood pressure.

One study has shown that regular coffee consumption for four weeks in healthy subjects, who were not naive coffee drinkers, does not appear to increase the risk of the development of hypertension[373]; however, the hemodynamic effects of chronic coffee and caffeine consumption has been largely debated.[371] At peak plasma caffeine levels 60 minutes after ingestion, the pressor effect of acute caffeine intake seems to be stronger in individuals who do not normally consume caffeine than in habitual users of caffeine (SBP 116 ± 5 in nonusers vs. 121 ± 6 in caffeine users; DBP 71 ± 4 in nonusers vs. 77 ± 4 in caffeine users), but there were no significant differences in heart rate between the groups.[374]

A meta-analysis of 16 studies found that when coffee trials (median intake: 725 mL/day) and caffeine trials (median dose: 410 mg/day) were analysed separately, BP elevations appeared to be larger in the caffeine group (median SBP 4.16 mmHg; median DBP 2.41 mmHg) than in the coffee group (median SBP 1.22 mmHg and median DBP 0.49 mmHg). However, it is noted that a possible reason that ingestion of caffeine tablets is more harmful to BP than coffee drinking is because it is not associated with favorable effects such as physical or mental relaxation, and caffeine tablets lack substances that could possibly exert a beneficial effect in the cardiovascular system. [375]

Over the medium term, pure caffeine supplementation tends to elevate blood pressure more than coffee. Current limited evidence suggests that coffee does not appear to increase the risk of developing hypertension, although there is some controversy surrounding this.

An observational study on cardiovascular mortality found no excessive risk of death associated with caffeine through caffeinated beverages in hypertensive individuals [376] Furthermore, caffeinated coffee consumption was associated with lower risk of coronary heart disease mortality and heart valve disease development or progression in older Framingham subjects without moderate or severe hypertension.[377] 

One prospective study randomly assigned a one time caffeine dose of 5 mg/kg or placebo to patients with symptomatic supraventricular tachycardia (SVT) one hour before starting radiofrequency ablation to determine if caffeine had any significant electrophysiology effects on cardiac refractoriness and conduction. Investigators found that caffeine was associated with a significant increase in systolic BP (median of 143 mmHg in the caffeine group vs. 132 mmHg in the placebo group) and diastolic BP (median of 83 mmHg in the caffeine group vs. 72 mmHg in the placebo group), however baseline measurements weren’t reported. There was no association with heart rate, cardiac refractoriness, cardiac conduction, tachycardia inducibility, or rates of induced tachycardia. Furthermore, no effect of caffeine on SVT induction or more rapid rates of induced tachycardias was found.[378]

Caffeine may have an effect on intraocular pressure (IOP). One study examined patients with open-angle glaucoma who consumed greater than or equal to 200 mg of caffeine per day and found that the group had a significantly higher mean IOP than patients who consumed less than 200 mg of caffeine per day (19.47 vs. 17.11 mmHg, p = 0.03).[379] A meta-analysis concluded that for normal individuals, IOP was not changed by ingestion of caffeine, while for patients with glaucoma or ocular hypertension, IOP increased at each measurement point (0.347 for normal patients, 2.395 for patients with glaucoma, and 1.998 for patients with ocular hypertension), which was statistically significant.[380]

A clinical trial evaluated the effect of caffeine on lower esophageal sphincter (LES) and esophageal peristaltic contractions in healthy adults who regularly consumed more than three cups of coffee per week and found that a single dose of caffeine 3.5 mg/kg with 100 mL water vs. the control drink of 100 mL water resulted in a decrease in basal LES pressure and distal esophageal contraction, which is known to promote the reflux of gastric contents up into the esophagus.[381]


Caffeine withdrawal syndrome is recognized as an official diagnosis in ICD-10 (World Health Organization).[382] The most commonly reported withdrawal symptoms are headache, fatigue, sleepiness/drowsiness, difficulty concentrating, irritability, less motivation to work, yawning, anxiety, flu-like symptoms i.e. nausea/vomiting and muscle stiffness, and impairment in psychomotor, vigilance and cognitive performances.[383] Onset of withdrawal symptoms usually occurs 12 to 24 hours after abstinence, with peak intensity generally occurring 20 to 51 hours after abstinence, and the duration of withdrawal most frequently ranging between two days to nine days.[304]


  1. ^