Skip to main content
Psychedex
Sources
42
14 strong · 22 moderate · 6 weak

Studies and references cited across this article, graded by evidence strength.

Claims checked
133
85verified48inconclusive

Inconclusive means the cited source neither confirms nor refutes the claim — a lower-confidence badge, not an error.

Generated
Apr 25, 2026 · pipeline-v5
Values approved
106/106 (100%)

Data points an editor reviewed and accepted for publication.

01

Value Provenance

Dosing

oraltrace

FieldPublishedAI ProposedConfidence
Threshold10 mg10 mghigh
Low20–100 mg20–100 mghigh
Common100–300 mg100–300 mghigh
Strong300–500 mg300–500 mghigh
Heavy500+ mg500+ mghigh
Editorial

Dose tiers derived from EFSA 2015 safety opinion (400 mg/day limit, 200 mg single dose limit), Fredholm et al. 2017 dose-dependent mechanism distinction (50-300 mg low dose, >500 mg high dose), and Guest et al. 2021 ISSN position stand (3-6 mg/kg ergogenic range). Bioavailability ~99-100% from Grzegorzewski et al. 2021 systematic review.

Reasoning

Dose tiers derived from EFSA 2015 safety opinion (400 mg/day limit, 200 mg single dose limit), Fredholm et al. 2017 dose-dependent mechanism distinction (50-300 mg low dose, >500 mg high dose), and Guest et al. 2021 ISSN position stand (3-6 mg/kg ergogenic range). Bioavailability ~99-100% from Grzegorzewski et al. 2021 systematic review.

Evidence basis: systematic_review

Duration

oral

FieldPublishedAI ProposedConfidence
Onsettrace10–30 minutes10–30 minuteshigh
Editorial

Onset of perceptible effects 10-30 minutes, consistent with Tmax of 30-60 minutes (Grzegorzewski et al. 2021). First alerts appear before peak plasma levels are reached.

Reasoning

Onset of perceptible effects 10-30 minutes, consistent with Tmax of 30-60 minutes (Grzegorzewski et al. 2021). First alerts appear before peak plasma levels are reached.

Evidence basis: systematic_review

Come uptrace15–30 minutes15–30 minuteshigh
Editorial

Come-up phase from first effects to approaching peak intensity lasts approximately 15-30 minutes, reflecting the ascending limb of the plasma concentration curve toward Tmax (Grzegorzewski et al. 2021).

Reasoning

Come-up phase from first effects to approaching peak intensity lasts approximately 15-30 minutes, reflecting the ascending limb of the plasma concentration curve toward Tmax (Grzegorzewski et al. 2021).

Evidence basis: systematic_review

Peaktrace1–2 hours1–2 hourshigh
Editorial

Peak subjective intensity lasting approximately 1-2 hours around Tmax (30-60 min), corresponding to maximum plasma concentrations of 5-20 µM (Grzegorzewski et al. 2021).

Reasoning

Peak subjective intensity lasting approximately 1-2 hours around Tmax (30-60 min), corresponding to maximum plasma concentrations of 5-20 µM (Grzegorzewski et al. 2021).

Evidence basis: systematic_review

Offsettrace2–4 hours2–4 hourshigh
Editorial

Offset duration of 2-4 hours reflects the descending limb of the plasma concentration curve, consistent with the 3-5 hour elimination half-life (Grzegorzewski et al. 2021).

Reasoning

Offset duration of 2-4 hours reflects the descending limb of the plasma concentration curve, consistent with the 3-5 hour elimination half-life (Grzegorzewski et al. 2021).

Evidence basis: systematic_review

Totaltrace3–6 hours3–6 hourshigh
Editorial

Total subjective effect duration of 3-6 hours, reflecting onset through offset. Consistent with the 3-5 hour half-life and the fact that subjective effects track plasma elimination (Grzegorzewski et al. 2021).

Reasoning

Total subjective effect duration of 3-6 hours, reflecting onset through offset. Consistent with the 3-5 hour half-life and the fact that subjective effects track plasma elimination (Grzegorzewski et al. 2021).

Evidence basis: systematic_review

After effectstrace2–6 hours2–6 hoursmedium
Editorial

Residual effects (subclinical stimulation, sleep disruption potential) persist 2-6 hours after subjective effects resolve. Duration varies substantially with CYP1A2 phenotype, time of day, and individual sensitivity. Sleep architecture effects documented by Weibel et al. 2020.

Reasoning

Residual effects (subclinical stimulation, sleep disruption potential) persist 2-6 hours after subjective effects resolve. Duration varies substantially with CYP1A2 phenotype, time of day, and individual sensitivity. Sleep architecture effects documented by Weibel et al. 2020.

Evidence basis: systematic_review

Safety

Toxicity Profiletrace

FieldPublishedAI ProposedConfidence
Acutelowlowhigh
Chroniclowlowhigh
Organs Affectedcardiovascular, neurological, gastrointestinalcardiovascular, neurological, gastrointestinalhigh
Editorial

Classified as low acute toxicity: wide therapeutic index confirmed by Kohl et al. (2020) and Cappelletti et al. (2014). Adverse outcomes limited to overdose or concentrated products. Low chronic toxicity: no organ damage documented at recommended intake levels (EFSA 2015). Pregnancy risk data (Greenwood et al. 2014) represents the primary chronic concern at moderate doses.

Reasoning

Classified as low acute toxicity: wide therapeutic index confirmed by Kohl et al. (2020) and Cappelletti et al. (2014). Adverse outcomes limited to overdose or concentrated products. Low chronic toxicity: no organ damage documented at recommended intake levels (EFSA 2015). Pregnancy risk data (Greenwood et al. 2014) represents the primary chronic concern at moderate doses.

Evidence basis: systematic_review

Dependence Profiletrace

FieldPublishedAI ProposedConfidence
Physicallowlowhigh
Psychologicallowlowhigh
Compulsive Redosinglowlowhigh
Dose Escalationlowlowhigh
Editorial

Dependence profile well-characterized: DSM-5 recognizes Caffeine Use Disorder (further study) and Caffeine Withdrawal as diagnosable. Strain et al. 1994 (JAMA) and Juliano et al. 2012 established clinical profiles. Fredholm et al. 1999 characterized caffeine as a weak reinforcer. Physical and psychological dependence consistently rated low-to-moderate across literature.

Reasoning

Dependence profile well-characterized: DSM-5 recognizes Caffeine Use Disorder (further study) and Caffeine Withdrawal as diagnosable. Strain et al. 1994 (JAMA) and Juliano et al. 2012 established clinical profiles. Fredholm et al. 1999 characterized caffeine as a weak reinforcer. Physical and psychological dependence consistently rated low-to-moderate across literature.

Evidence basis: systematic_review

Withdrawal Profiletrace

FieldPublishedAI ProposedConfidence
Severitymildmildhigh
Symptomsheadache, fatigue, dysphoria, depressed mood, difficulty concentrating, irritability, nausea, muscle painheadache, fatigue, dysphoria, depressed mood, difficulty concentrating, irritability, nausea, muscle painhigh
FatalNoNohigh
Medical SupervisionNot requiredNot requiredhigh
Editorial

Caffeine withdrawal is DSM-5 recognized (Rocha Cabrero et al. 2026). Onset 12-24h, peak 20-51h, duration 2-9 days. Headache most characteristic symptom, caused by rebound cerebral vasodilation. Not medically dangerous but can mimic psychiatric disorders.

Reasoning

Caffeine withdrawal is DSM-5 recognized (Rocha Cabrero et al. 2026). Onset 12-24h, peak 20-51h, duration 2-9 days. Headache most characteristic symptom, caused by rebound cerebral vasodilation. Not medically dangerous but can mimic psychiatric disorders.

Evidence basis: systematic_review

Tolerance Profiletrace

FieldPublishedAI ProposedConfidence
Buildupmoderatemoderatemedium
Half-life5 days5 daysmedium
Full Reset10 days10 daysmedium
Cross-tolerancestheophylline, theobromine, paraxanthinetheophylline, theobromine, paraxanthinemedium
Editorial

Tolerance mechanism well-established via Ciruela et al. 2006 (A1-A2A heteromer adaptation) and Fredholm et al. 1999 (adenosine receptor upregulation). Buildup rate classified as moderate — tolerance develops over days to weeks of regular use. Timeline values carried forward from existing database with medium confidence pending specific literature confirmation.

Reasoning

Tolerance mechanism well-established via Ciruela et al. 2006 (A1-A2A heteromer adaptation) and Fredholm et al. 1999 (adenosine receptor upregulation). Buildup rate classified as moderate — tolerance develops over days to weeks of regular use. Timeline values carried forward from existing database with medium confidence pending specific literature confirmation.

Evidence basis: community_consensus_large

Effects & Profile

Contraindications

FieldPublishedAI ProposedConfidence
cardiovascular uncontrolled hypertensiontracerelativerelativemedium
Editorial

Cappelletti et al. 2014 documents acute cardiovascular effects including blood pressure elevation.

Reasoning

Cappelletti et al. 2014 documents acute cardiovascular effects including blood pressure elevation.

psychiatric panic disordertraceabsoluteabsolutehigh
Editorial

Klevebrant & Frick 2022 systematic review and meta-analysis confirms caffeine reliably provokes panic in panic disorder patients.

Reasoning

Klevebrant & Frick 2022 systematic review and meta-analysis confirms caffeine reliably provokes panic in panic disorder patients.

psychiatric generalized anxiety disordertracerelativerelativehigh
Editorial

Lara 2010 and Klevebrant & Frick 2022 document dose-dependent anxiogenesis with heightened vulnerability in anxiety disorder patients.

Reasoning

Lara 2010 and Klevebrant & Frick 2022 document dose-dependent anxiogenesis with heightened vulnerability in anxiety disorder patients.

cardiovascular cardiac arrhythmiastracerelativerelativehigh
Editorial

Martinez et al. 2024 and Enriquez & Frankel 2017 document arrhythmia risk with high-dose caffeine.

Reasoning

Martinez et al. 2024 and Enriquez & Frankel 2017 document arrhythmia risk with high-dose caffeine.

pregnancy breastfeeding pregnancytracerelativerelativehigh
Editorial

Greenwood et al. 2014 meta-analysis of 60 publications provides robust dose-response data. EFSA 2015 establishes 200 mg/day limit.

Reasoning

Greenwood et al. 2014 meta-analysis of 60 publications provides robust dose-response data. EFSA 2015 establishes 200 mg/day limit.

age neonates and infantstraceabsoluteabsolutehigh
Editorial

Cazeneuve et al. 1994 documented 65-100 hour half-life in neonates. EFSA 2015 failed to establish safe limits for children/adolescents.

Reasoning

Cazeneuve et al. 1994 documented 65-100 hour half-life in neonates. EFSA 2015 failed to establish safe limits for children/adolescents.

metabolic CYP1A2 inhibitor co-administrationtracerelativerelativehigh
Editorial

Jeppesen et al. 1996, Culm-Merdek et al. 2005, and Christensen et al. 2002 document dramatic CYP1A2-mediated interactions.

Reasoning

Jeppesen et al. 1996, Culm-Merdek et al. 2005, and Christensen et al. 2002 document dramatic CYP1A2-mediated interactions.

Timeline Phases

oral

FieldPublishedAI ProposedConfidence
Onsettrace600–1800 sec(i=0.05)600–1800 sec(i=0.05)high
Editorial

Oral caffeine pharmacokinetics are among the most thoroughly characterized of any psychoactive substance, with Tmax of 30-60 minutes (Grzegorzewski et al. 2021) and half-life of 3-5 hours. Timeline phases derived from PK data and extensive controlled study reports of subjective effect onset and duration.

Reasoning

Oral caffeine pharmacokinetics are among the most thoroughly characterized of any psychoactive substance, with Tmax of 30-60 minutes (Grzegorzewski et al. 2021) and half-life of 3-5 hours. Timeline phases derived from PK data and extensive controlled study reports of subjective effect onset and duration.

Evidence basis: systematic_review

Come uptrace900–1800 sec(i=0.30)900–1800 sec(i=0.30)high
Editorial

Oral caffeine pharmacokinetics are among the most thoroughly characterized of any psychoactive substance, with Tmax of 30-60 minutes (Grzegorzewski et al. 2021) and half-life of 3-5 hours. Timeline phases derived from PK data and extensive controlled study reports of subjective effect onset and duration.

Reasoning

Oral caffeine pharmacokinetics are among the most thoroughly characterized of any psychoactive substance, with Tmax of 30-60 minutes (Grzegorzewski et al. 2021) and half-life of 3-5 hours. Timeline phases derived from PK data and extensive controlled study reports of subjective effect onset and duration.

Evidence basis: systematic_review

Peaktrace3600–7200 sec(i=1.00)3600–7200 sec(i=1.00)high
Editorial

Oral caffeine pharmacokinetics are among the most thoroughly characterized of any psychoactive substance, with Tmax of 30-60 minutes (Grzegorzewski et al. 2021) and half-life of 3-5 hours. Timeline phases derived from PK data and extensive controlled study reports of subjective effect onset and duration.

Reasoning

Oral caffeine pharmacokinetics are among the most thoroughly characterized of any psychoactive substance, with Tmax of 30-60 minutes (Grzegorzewski et al. 2021) and half-life of 3-5 hours. Timeline phases derived from PK data and extensive controlled study reports of subjective effect onset and duration.

Evidence basis: systematic_review

Plateautrace5400–10800 sec(i=0.50)5400–10800 sec(i=0.50)high
Editorial

Oral caffeine pharmacokinetics are among the most thoroughly characterized of any psychoactive substance, with Tmax of 30-60 minutes (Grzegorzewski et al. 2021) and half-life of 3-5 hours. Timeline phases derived from PK data and extensive controlled study reports of subjective effect onset and duration.

Reasoning

Oral caffeine pharmacokinetics are among the most thoroughly characterized of any psychoactive substance, with Tmax of 30-60 minutes (Grzegorzewski et al. 2021) and half-life of 3-5 hours. Timeline phases derived from PK data and extensive controlled study reports of subjective effect onset and duration.

Evidence basis: systematic_review

Resolutiontrace3600–7200 sec(i=0.20)3600–7200 sec(i=0.20)high
Editorial

Oral caffeine pharmacokinetics are among the most thoroughly characterized of any psychoactive substance, with Tmax of 30-60 minutes (Grzegorzewski et al. 2021) and half-life of 3-5 hours. Timeline phases derived from PK data and extensive controlled study reports of subjective effect onset and duration.

Reasoning

Oral caffeine pharmacokinetics are among the most thoroughly characterized of any psychoactive substance, with Tmax of 30-60 minutes (Grzegorzewski et al. 2021) and half-life of 3-5 hours. Timeline phases derived from PK data and extensive controlled study reports of subjective effect onset and duration.

Evidence basis: systematic_review

Afterglowtrace7200–14400 sec(i=0.05)7200–14400 sec(i=0.05)high
Editorial

Oral caffeine pharmacokinetics are among the most thoroughly characterized of any psychoactive substance, with Tmax of 30-60 minutes (Grzegorzewski et al. 2021) and half-life of 3-5 hours. Timeline phases derived from PK data and extensive controlled study reports of subjective effect onset and duration.

Reasoning

Oral caffeine pharmacokinetics are among the most thoroughly characterized of any psychoactive substance, with Tmax of 30-60 minutes (Grzegorzewski et al. 2021) and half-life of 3-5 hours. Timeline phases derived from PK data and extensive controlled study reports of subjective effect onset and duration.

Evidence basis: systematic_review

Dose Tier Summaries

thresholdtrace
high

At threshold doses (10–20 mg oral, roughly equivalent to a few sips of coffee or a weak cup of tea), effects are subtle and easily overlooked. A faint brightening of alertness may be noticed, with marginally reduced perception of fatigue. Most individuals would not identify these effects as drug-related without prior awareness of ingestion. Physically, mild increases in heart rate or slight diuresis may occur but are generally sub-perceptual. No meaningful impact on sleep architecture at this dose unless consumed within 2–3 hours of bedtime in caffeine-naïve individuals.

Editorial

Threshold doses of 10-20 mg are well-established in the community pharmacology literature and consistent with EFSA's recognition of pharmacological activity beginning at low doses. Effects at this range are consistently described as sub-perceptual to barely perceptual.

Reasoning

Threshold doses of 10-20 mg are well-established in the community pharmacology literature and consistent with EFSA's recognition of pharmacological activity beginning at low doses. Effects at this range are consistently described as sub-perceptual to barely perceptual.

Evidence basis: community_consensus_large

lighttrace
high

At light doses (20–100 mg oral, equivalent to a cup of green tea to a weak cup of coffee), mild stimulation becomes clearly perceptible. Wakefulness increases noticeably, with improved attention and a subtle sense of mental clarity. Thought acceleration may be apparent, and mood elevation is common — a gentle sense of motivation and engagement with tasks. Physically, mild increases in heart rate and blood pressure occur. Slight diuresis is typical. Some individuals notice mild appetite suppression. These doses are well-tolerated by most adults and represent the lower end of typical daily consumption patterns.

Editorial

Light dose range of 20-100 mg is consistent with EFSA analysis and published studies showing measurable cognitive effects beginning around 50 mg (Cappelletti et al. 2014, Lara 2010). Effects at this range are well-characterized in controlled studies.

Reasoning

Light dose range of 20-100 mg is consistent with EFSA analysis and published studies showing measurable cognitive effects beginning around 50 mg (Cappelletti et al. 2014, Lara 2010). Effects at this range are well-characterized in controlled studies.

Evidence basis: systematic_review

commontrace
high

At common doses (100–300 mg oral, equivalent to 1–3 cups of brewed coffee), the full stimulant profile emerges. Wakefulness and alertness are pronounced. Attention and vigilance are consistently improved, with reduced reaction time and enhanced sustained focus. Thought acceleration is evident, and motivation for task completion increases. Mood is mildly elevated, with increased sociability and verbal fluency. Physically, increased heart rate, mild blood pressure elevation, and diuresis are expected. Mild tremor may be noticeable in caffeine-naïve individuals. Appetite suppression is common. Sleep onset will be delayed if consumed within 6 hours of bedtime. At the upper end of this range (200–300 mg), some individuals — particularly those with ADORA2A polymorphisms or anxiety disorders — may experience restlessness, irritability, or mild anxiety rather than focused alertness.

Editorial

The 100-300 mg range encompasses typical daily consumption and is the most extensively studied dose range in human trials. Cognitive and physical effects are well-characterized across multiple systematic reviews (Cappelletti et al. 2014, EFSA 2015, Guest et al. 2021).

Reasoning

The 100-300 mg range encompasses typical daily consumption and is the most extensively studied dose range in human trials. Cognitive and physical effects are well-characterized across multiple systematic reviews (Cappelletti et al. 2014, EFSA 2015, Guest et al. 2021).

Evidence basis: systematic_review

strongtrace
high

At strong doses (300–500 mg oral, equivalent to 3–5 cups of brewed coffee consumed rapidly), stimulation becomes intense and dose-dependent side effects emerge prominently. Alertness is maximal but accompanied by increasing restlessness and anxiety. Thought acceleration may become uncomfortable — racing thoughts and difficulty maintaining focused attention replace the productive clarity of lower doses. Irritability is common. Physically, tachycardia, palpitations, noticeable tremor, and gastrointestinal discomfort (nausea, stomach upset) become frequent. Diuresis is pronounced. Jaw tension and muscle tightness may occur. Sleep will be significantly disrupted for 6–10 hours after ingestion. Individuals with panic disorder may experience panic attacks at this dose range. The anxiogenic threshold is highly individual and influenced by genetic variation in the ADORA2A receptor gene.

Editorial

Effects at 300-500 mg are well-documented in clinical literature. EFSA (2015) identifies 200 mg single dose as the upper limit for avoiding acute adverse effects. Klevebrant & Frick (2022) meta-analysis confirms panic provocation at these doses in susceptible individuals.

Reasoning

Effects at 300-500 mg are well-documented in clinical literature. EFSA (2015) identifies 200 mg single dose as the upper limit for avoiding acute adverse effects. Klevebrant & Frick (2022) meta-analysis confirms panic provocation at these doses in susceptible individuals.

Evidence basis: systematic_review

heavytrace
high

At heavy doses (>500 mg oral in a single session, equivalent to 5+ cups of coffee consumed rapidly), caffeine produces a predominantly aversive experience dominated by anxiety and physical discomfort. Marked restlessness, agitation, and anxiety are characteristic. Thought acceleration may become dysphoric — racing, disorganized, and difficult to control. Concentration paradoxically deteriorates despite maximal arousal. Physically, tachycardia and palpitations are prominent. Tremor, profuse diuresis, nausea, and vomiting are common. At doses exceeding 1,000 mg, serious medical complications become possible: tachyarrhythmias, seizures, hypokalemia, and metabolic acidosis. Doses above 5 g are life-threatening. Sleep is impossible for many hours. Functional impairment is significant — the combination of physical discomfort and cognitive agitation prevents productive activity. Rare cases of caffeine-induced psychotic or manic symptoms have been reported at extreme doses.

Editorial

Heavy dose effects (>500 mg) are documented across multiple sources including Kohl et al. (2020), Cappelletti et al. (2014), and Lara (2010). The overdose toxidrome is well-characterized with clear dose-response progression from anxiety through arrhythmia to cardiovascular collapse.

Reasoning

Heavy dose effects (>500 mg) are documented across multiple sources including Kohl et al. (2020), Cappelletti et al. (2014), and Lara (2010). The overdose toxidrome is well-characterized with clear dose-response progression from anxiety through arrhythmia to cardiovascular collapse.

Evidence basis: systematic_review

02

History

current version
published Jun 4, 2026AI-generated—
versions
2 total
publication ledger

backfill Jun 4, 2026

Historical versions do not record which reviewer approved them; editorial accountability is recorded per value in the database and in the publication ledger going forward.

full version history suggest a correctioncontact@psychedex.org
03

Sources

14 strong · 22 moderate · 6 weak
  1. [1]
    strong
    Actions of caffeine in the brain with special reference to factors that contribute to its widespread use.
    Fredholm BB, Bättig K, Holmén J, Nehlig A, Zvartau EE · 1999 · Pharmacological Reviews
    pubmed 1004999913 claims checked8 verified5 inconclusive
  2. [2]
    strong
    Pharmacokinetics of Caffeine: A Systematic Analysis of Reported Data for Application in Metabolic Phenotyping and Liver Function Testing.
    Grzegorzewski J, Bartsch F, Köller A, König M · 2021 · Frontiers in Pharmacology
    doi.org/10.3389/fphar.2021.752826pubmed 3528025419 claims checked2 verified17 inconclusive
  3. [3]
    weak
    Caffeine history: isolation by Runge 1819, synthesis by Fischer 1895
    Wikipedia / Inverse.com · 2024
    linkno claims checked
  4. [4]
    moderate
    Association of Coffee Consumption and Prediagnostic Caffeine Metabolites With Incident Parkinson Disease in a Population-Based Cohort.
    Zhao Y, Lai Y, Konijnenberg H, et al. · 2024 · Neurology
    doi.org/10.1212/wnl.0000000000209201pubmed 385131621 claim checked1 verified
  5. [5]
    strong
    Caffeine therapy for apnea of prematurity.
    Schmidt B, Roberts RS, Davis P, Doyle LW, Barrington KJ, Ohlsson A, Solimano A, Tin W · 2006 · New England Journal of Medicine
    doi.org/10.1056/nejmoa054065pubmed 167077483 claims checked3 verified
  6. [6]
    strong
    International Union of Basic and Clinical Pharmacology. LXXXI. Nomenclature and classification of adenosine receptors--an update.
    Fredholm BB, IJzerman AP, Jacobson KA, Linden J, Müller CE · 2011 · Pharmacological Reviews
    doi.org/10.1124/pr.110.003285pubmed 213038997 claims checked1 verified6 inconclusive
  7. [7]
    moderate
    Low, but not high, dose caffeine is a readily available probe for adenosine actions.
    Fredholm BB, Yang J, Wang Y · 2017 · Molecular Aspects of Medicine
    doi.org/10.1016/j.mam.2016.11.011pubmed 279150512 claims checked2 verified
  8. [8]
    moderate
    Chronic caffeine consumption prevents memory disturbance in different animal models of memory decline.
    Cunha RA, Agostinho PM · 2010 · Journal of Alzheimer's Disease
    doi.org/10.3233/jad-2010-1408pubmed 201820434 claims checked4 verified
  9. [9]
    moderate
    Caffeine, mental health, and psychiatric disorders.
    Lara DR · 2010 · Journal of Alzheimer's Disease
    doi.org/10.3233/jad-2010-1378pubmed 201645715 claims checked5 verified
  10. [10]
    strong
    Effects of caffeine on anxiety and panic attacks in patients with panic disorder: A systematic review and meta-analysis.
    Klevebrant L, Frick A · 2022 · General Hospital Psychiatry
    doi.org/10.1016/j.genhosppsych.2021.11.005pubmed 348719644 claims checked4 verified
  11. [11]
    weak
    Acute intentional caffeine overdose treated preemptively with hemodialysis.
    Kohl BA, Kaur K, Dincher N, Schumann J, Carachilo T, Komurek C · 2020 · American Journal of Emergency Medicine
    doi.org/10.1016/j.ajem.2019.09.018pubmed 317859826 claims checked5 verified1 inconclusive
  12. [12]
    moderate
    Caffeine: Cognitive and Physical Performance Enhancer or Psychoactive Drug?
    Cappelletti S, Piacentino D, Sani G, Aromatario M · 2014 · Current Neuropharmacology
    doi.org/10.2174/1570159x13666141210215655pubmed 260747445 claims checked1 verified4 inconclusive
  13. [13]
    weak
    Sudden cardiac arrest occurring in temporal proximity to consumption of energy drinks.
    Martinez KA, Bains S, Neves R, Giudicessi JR, Bos JM, Ackerman MJ · 2024 · Heart Rhythm
    doi.org/10.1016/j.hrthm.2024.02.018pubmed 388429643 claims checked2 verified1 inconclusive
  14. [14]
    moderate
    Caffeine Withdrawal
    Rocha Cabrero F, Hamilton RJ, et al. · 2026 · Nutrients
    doi.org/10.3390/nu16183048pubmed 286135414 claims checked3 verified1 inconclusive
  15. [15]
    moderate
    Presynaptic Control of Striatal Glutamatergic Neurotransmission by Adenosine A1-A2A Receptor Heteromers
    Ciruela F, Casadó V, Rodrigues RJ, Luján R, Burgueño J, et al. · 2006 · Journal of Neuroscience
    doi.org/10.1523/jneurosci.3574-05.2006pubmed 164814415 claims checked5 verified
  16. [16]
    moderate
    Caffeine-dependent changes of sleep-wake regulation: Evidence for adaptation after repeated intake.
    Weibel J, Lin YS, Landolt HP, Garbazza C, et al. · 2020 · Progress in Neuro-Psychopharmacology & Biological Psychiatry
    doi.org/10.1016/j.pnpbp.2019.109851pubmed 318663085 claims checked3 verified2 inconclusive
  17. [17]
    strong
    A fluvoxamine-caffeine interaction study.
    Jeppesen U, Loft S, Poulsen HE, Brøsen K · 1996 · Pharmacogenetics
    pubmed 88076604 claims checked4 verified
  18. [18]
    moderate
    Biotransformation of caffeine in human liver microsomes from foetuses, neonates, infants and adults.
    Cazeneuve C, Pons G, Rey E, et al. · 1994 · British Journal of Clinical Pharmacology
    pubmed 80542453 claims checked3 inconclusive
  19. [19]
    moderate
    Caffeine — PubChem CID 2519
    PubChem · 2024 · PubChem (NCBI)
    pubmed 25191 claim checked1 inconclusive
  20. [20]
    weak
    Caffeine physical properties: melting point, solubility, CAS
    PubChem/NIST/ChemicalBook · 2024
    linkno claims checked
  21. [21]
    moderate
    Caffeine and adenosine.
    Ribeiro JA, Sebastião AM · 2010 · Journal of Alzheimer's Disease
    doi.org/10.3233/jad-2010-1379pubmed 201645663 claims checked3 verified
  22. [22]
    moderate
    Caffeine and sports activity: a review.
    Nehlig A, Debry G · 1994 · International Journal of Sports Medicine
    pubmed 79603132 claims checked2 verified
  23. [23]
    moderate
    Oxidation of caffeine by CYP1A2: isotope effects and metabolic switching.
    Regal KA, Kunze KL, Peter RM, Nelson SD · 2005 · Drug Metabolism and Disposition
    pubmed 161356583 claims checked1 verified2 inconclusive
  24. [24]
    moderate
    Drug interactions with tobacco smoking. An update.
    Zevin S, Benowitz NL · 1999 · Clinical Pharmacokinetics
    pubmed 104274672 claims checked2 verified
  25. [25]
    strong
    The epidemiology of Parkinson's disease: risk factors and prevention.
    Ascherio A, Schwarzschild MA · 2016 · The Lancet Neurology
    doi.org/10.1016/s1474-4422(16)30230-7pubmed 277515562 claims checked2 verified
  26. [26]
    strong
    Tea, coffee, and caffeine intake and risk of dementia and Alzheimer's disease: a systematic review and meta-analysis of cohort studies.
    Li F, Liu X, Jiang B, et al. · 2024 · Food & Function
    doi.org/10.1039/d4fo01750apubmed 390548942 claims checked2 verified
  27. [27]
    moderate
    Plasma Caffeine Levels and Risk of Alzheimer's Disease and Parkinson's Disease: Mendelian Randomization Study.
    Larsson SC, Woolf B, Gill D · 2022 · Nutrients
    doi.org/10.3390/nu14091697pubmed 355656672 claims checked2 verified
  28. [28]
    moderate
    Structure of the adenosine A(2A) receptor in complex with ZM241385 and the xanthines XAC and caffeine.
    Doré AS, Robertson N, Errey JC, et al. · 2011 · Structure
    doi.org/10.1016/j.str.2011.06.014pubmed 21885291no claims checked
  29. [29]
    strong
    Low daily 10-mg and 20-mg doses of fluvoxamine inhibit the metabolism of both caffeine (CYP1A2) and omeprazole (CYP2C19).
    Christensen M, Tybring G, Mihara K, et al. · 2002 · Clinical Pharmacology and Therapeutics
    pubmed 119074882 claims checked2 verified
  30. [30]
    strong
    Fluvoxamine impairs single-dose caffeine clearance without altering caffeine pharmacodynamics.
    Culm-Merdek KE, von Moltke LL, Harmatz JS, Greenblatt DJ · 2005 · British Journal of Clinical Pharmacology
    pubmed 162360382 claims checked2 verified
  31. [31]
    strong
    Caffeine may disrupt the impact of real-time drowsiness on cognitive performance: a double-blind, placebo-controlled small-sample study.
    Aidman E, Balin M, Johnson K, et al. · 2021 · Scientific Reports
    doi.org/10.1038/s41598-021-83504-6pubmed 335975801 claim checked1 verified
  32. [32]
    strong
    International society of sports nutrition position stand: caffeine and exercise performance.
    Guest NS, VanDusseldorp TA, Nelson MT, Grgic J, Schoenfeld BJ, et al. · 2021 · Journal of the International Society of Sports Nutrition
    doi.org/10.1186/s12970-020-00383-41 claim checked1 verified
  33. [33]
    moderate
    Expression and characterization of purinergic receptors in rat middle meningeal artery--potential role in migraine.
    Haanes KA, Edvinsson L · 2014 · PLoS ONE
    doi.org/10.1371/journal.pone.0108782pubmed 252652862 claims checked2 verified
  34. [34]
    weak
    Arrhythmogenic effects of energy drinks.
    Enriquez A, Frankel DS · 2017 · Journal of Cardiovascular Electrophysiology
    doi.org/10.1111/jce.13210pubmed 283874312 claims checked1 verified1 inconclusive
  35. [35]
    moderate
    Regulatory status of caffeine in the United States.
    Rosenfeld LS, Mihalov JJ, Carlson SJ, Mattia A · 2014 · Nutrition Reviews
    doi.org/10.1111/nure.12136pubmed 252935413 claims checked3 inconclusive
  36. [36]
    strong
    Caffeine intake during pregnancy and adverse birth outcomes: a systematic review and dose-response meta-analysis.
    Greenwood DC, Thatcher NJ, Ye J, Garrard L, et al. · 2014 · European Journal of Epidemiology
    doi.org/10.1007/s10654-014-9944-xpubmed 251797922 claims checked2 verified
  37. [37]
    weak
    Scientific Opinion on the safety of caffeine. EFSA 2015.
    EFSA Panel on Dietetic Products, Nutrition and Allergies · 2015 · EFSA Journal
    doi.org/10.2903/j.efsa.2015.41022 claims checked2 verified
  38. [38]
    moderate
    Caffeine provokes adverse interactions with MDMA and related psychostimulants: mechanisms and mediators.
    Vanattou-Saîfoudine N, McNamara R, Harkin A · 2012 · British Journal of Pharmacology
    doi.org/10.1111/j.1476-5381.2012.02065.xpubmed 226717622 claims checked2 verified
  39. [39]
    strong
    The effects of energy drink in combination with alcohol on performance and subjective awareness.
    Alford C, Hamilton-Morris J, Verster JC · 2012 · Psychopharmacology
    doi.org/10.1007/s00213-012-2677-1pubmed 224568621 claim checked1 verified
  40. [40]
    moderate
    The Effect of Caffeine Consumption and Acute Withdrawal on Resting-State fMRI Brain Connectivity, Mood and Cognition.
    Sevenoaks T, Lancelotte F, Souter N, Stafford L, Rae C, Yeomans M · 2026 · European Journal of Neuroscience
    doi.org/10.1111/ejn.70511pubmed 420030441 claim checked1 inconclusive
  41. [41]
    moderate
    Caffeine dependence syndrome. Evidence from case histories and experimental evaluations.
    Strain EC, Mumford GK, Silverman K, Griffiths RR · 1994 · JAMA
    pubmed 80898871 claim checked1 verified
  42. [42]
    moderate
    Characterization of individuals seeking treatment for caffeine dependence.
    Juliano LM, Evatt DP, Richards BD, Griffiths RR · 2012 · Psychology of Addictive Behaviors
    doi.org/10.1037/a0027246pubmed 223692181 claim checked1 verified
04

Claim Verification

Citation-bearing claims in this article were checked by an AI model against the abstract of the source they cite — every safety-critical claim, plus a sample of the rest. Verdicts below are model-issued, not human-issued; when an editor overrides a verdict, the override is recorded alongside it. Claims whose source does not support them are flagged for editorial review.

summary19 claims12verified7inconclusive
inconclusive1,3,7-Trimethylxanthine (caffeine) is a xanthine alkaloid stimulant and the most widely consumed psychoactive substance on Earth, with over 80% of adults in Western societies using it daily in pharmacologically active doses[1]model confidence 75%
Inconclusive

Review covers widespread caffeine use but the specific 80% statistic is not confirmed in the abstract text.

Landmark comprehensive review of caffeine pharmacology in the brain. Covers adenosine antagonism, dopaminergic effects, reinforcement, withdrawal, tolerance.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveActive oral doses range from 50–400 mg, producing effects within 30–60 minutes and an elimination half-life of 3–5 hours[2]model confidence 72%
Inconclusive

Comprehensive PK review would contain these values but the specific onset and half-life figures are not stated in the abstract text.

Systematic analysis of caffeine pharmacokinetics from 141 publications.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedThe foundational mechanistic review by Fredholm et al. in 1999 established adenosine antagonism as the primary mediator of caffeine's central effects[1]model confidence 90%
Verified

Abstract confirms this is a landmark comprehensive review explicitly covering adenosine antagonism as a mechanism.

Landmark comprehensive review of caffeine pharmacology in the brain. Covers adenosine antagonism, dopaminergic effects, reinforcement, withdrawal, tolerance.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedCurrent investigational interest centers on caffeine's neuroprotective potential in Parkinson's disease — supported by robust prospective epidemiological data — and its established clinical role in neonatal apnea of prematurity.[4]model confidence 92%
Verified

Abstract directly describes prospective cohort data showing inverse associations between caffeine and Parkinson disease.

Prospective cohort: inverse associations between caffeine intake and Parkinson disease. Prediagnostic caffeine metabolites quantified years before PD onset.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedCurrent investigational interest centers on caffeine's neuroprotective potential in Parkinson's disease — supported by robust prospective epidemiological data — and its established clinical role in neonatal apnea of prematurity.[5]model confidence 95%
Verified

Abstract directly confirms caffeine's established clinical role in neonatal apnea of prematurity.

Caffeine therapy for apnea of prematurity reduces the rate of bronchopulmonary dysplasia in infants with very low birth weight. Landmark RCT establishing caffeine as first-line treatment for neonatal apnea.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveCaffeine is a non-selective competitive antagonist at adenosine A1 and A2A receptors (Ki ~12–30 uM and ~8–25 uM respectively), blocking adenosine's tonic inhibitory regulation of dopaminergic and glutamatergic neurotransmission[1]model confidence 72%
Inconclusive

Abstract confirms adenosine antagonism and dopaminergic effects but specific Ki values are not stated in the abstract text.

Covers adenosine antagonism, dopaminergic effects, reinforcement, withdrawal, tolerance.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveCaffeine is a non-selective competitive antagonist at adenosine A1 and A2A receptors (Ki ~12–30 uM and ~8–25 uM respectively), blocking adenosine's tonic inhibitory regulation of dopaminergic and glutamatergic neurotransmission[6]model confidence 72%
Inconclusive

IUPHAR review covers binding affinities but specific Ki values for caffeine are not stated in the abstract text.

Reviews A1, A2A, A2B, A3 receptors including binding affinities, agonists, antagonists.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedAt doses exceeding 500 mg, secondary mechanisms — phosphodiesterase inhibition and intracellular calcium mobilization — become pharmacologically relevant.[7]model confidence 93%
Verified

Abstract directly distinguishes low-dose adenosine antagonism from high-dose mechanisms, supporting the claim.

Low-dose caffeine acts as an adenosine A1, A2A, A2B antagonist. High-dose effects have different underlying mechanisms.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedThe resulting phenomenology is primarily one of cognitive normalization rather than true enhancement: caffeine restores vigilance, attention, and reaction time most effectively when cognition is degraded by fatigue, sleep deprivation, or stress[8]model confidence 92%
Verified

Abstract directly characterizes caffeine as a cognitive normalizer that restores function in sleep-deprived individuals.

Caffeine should be viewed as a 'cognitive normalizer' not enhancer. Restores memory in sleep-deprived and aged individuals.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedDose-dependent anxiogenesis is the principal adverse psychological effect, with individuals carrying ADORA2A polymorphisms and those with panic disorder showing heightened vulnerability[9]model confidence 93%
Verified

Abstract directly confirms panic disorder sensitivity and A2A receptor polymorphism influence on anxiety threshold.

Panic disorder patients particularly sensitive to anxiogenic effects. A2A receptor polymorphism influences anxiety threshold.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedDose-dependent anxiogenesis is the principal adverse psychological effect, with individuals carrying ADORA2A polymorphisms and those with panic disorder showing heightened vulnerability[10]model confidence 93%
Verified

Systematic review and meta-analysis specifically quantifies caffeine anxiogenic and panicogenic effects in panic disorder patients.

Systematic review and meta-analysis of caffeine's anxiogenic and panicogenic effects in panic disorder patients. Quantifies magnitude and dose-response relationships.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedCaffeine has a wide therapeutic index (estimated LD50 ~150–200 mg/kg; lethal dose ~10.5–14 g in a 70 kg adult), but concentrated powder and energy drink products have caused fatalities at doses far below what beverage consumption could deliver[11]model confidence 90%
Verified

Abstract explicitly states LD50 of 150–200 mg/kg and describes intentional 50g caffeine powder ingestion resulting in near-fatal toxicity.

LD50 stated as 150-200 mg/kg; lethal plasma levels ~80-100 ug/mL. Pre-dialysis caffeine level 254 ug/mL.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveCaffeine has a wide therapeutic index (estimated LD50 ~150–200 mg/kg; lethal dose ~10.5–14 g in a 70 kg adult), but concentrated powder and energy drink products have caused fatalities at doses far below what beverage consumption could deliver[12]model confidence 62%
Inconclusive

Review covers lethal effects and toxicity concepts but specific LD50 values are not stated in the abstract text.

Notes relative toxicity/lethality concepts.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedCaffeine has a wide therapeutic index (estimated LD50 ~150–200 mg/kg; lethal dose ~10.5–14 g in a 70 kg adult), but concentrated powder and energy drink products have caused fatalities at doses far below what beverage consumption could deliver[13]model confidence 88%
Verified

Abstract documents cardiac arrest cases temporally associated with energy drink consumption, supporting the energy drink fatality risk claim.

Energy drinks potentially trigger life-threatening cardiac arrhythmias via highly stimulating ingredients. Documents cases of sudden cardiac arrest in temporal proximity to energy drink consumption.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusivePhysical and psychological dependence are low-to-moderate; caffeine withdrawal is recognized in the DSM-5, with headache, fatigue, and dysphoria peaking at 20–51 hours after cessation[14]model confidence 70%
Inconclusive

Abstract confirms DSM-5 recognition and symptoms but specific 20–51 hour peak timing is not stated in the abstract text.

Review of caffeine withdrawal as DSM-5 recognized condition. Symptoms: headache, fatigue, irritability, mood changes.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedTolerance develops through upregulation of adenosine receptors and functional adaptation of A1-A2A heteromers in the striatum, though wakefulness-promoting effects remain partially preserved even with chronic use[15]model confidence 91%
Verified

Abstract directly states chronic caffeine causes tolerance through modifications in A1-A2A heteromer function in striatal terminals.

Caffeine acts on A1-A2A heteromers; chronic caffeine leads to tolerance through modifications in heteromer function.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedTolerance develops through upregulation of adenosine receptors and functional adaptation of A1-A2A heteromers in the striatum, though wakefulness-promoting effects remain partially preserved even with chronic use[16]model confidence 87%
Verified

Abstract covers adaptation of sleep-wake regulation after repeated caffeine, supporting incomplete tolerance to wakefulness effects.

Daily caffeine intake and withdrawal affect circadian rhythms and wake-promotion. Evidence for adaptation after repeated intake.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedCYP1A2-mediated metabolism introduces clinically significant variability: fluvoxamine co-administration increases caffeine half-life approximately six-fold, and neonatal CYP1A2 immaturity extends half-life to 65–100 hours[17]model confidence 93%
Verified

Abstract explicitly states fluvoxamine increases half-life from 5h to 31h (~6.2x, consistent with approximately six-fold).

Fluvoxamine 50-100mg/day increases caffeine half-life from 5h to 31h and reduces total clearance from 107 to 21 ml/min.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveCYP1A2-mediated metabolism introduces clinically significant variability: fluvoxamine co-administration increases caffeine half-life approximately six-fold, and neonatal CYP1A2 immaturity extends half-life to 65–100 hours[18]model confidence 72%
Inconclusive

Abstract confirms CYP1A2 absent at birth and matures progressively, but specific 65–100 hour range is not stated in the abstract.

CYP1A2 not expressed in foetuses; matures progressively over 0-300 days in neonates/infants.

checked by claude-sonnet-4-6 · Apr 25, 2026

history2 claims2verified
verifiedThe first comprehensive pharmacological analysis of caffeine in the brain — covering adenosine antagonism, dopaminergic interactions, reinforcement, tolerance, and factors contributing to its ubiquitous consumption — was published by Fredholm et al. in 1999 in Pharmacological Reviews[1]model confidence 90%
Verified

Abstract confirms this is a landmark comprehensive review covering all listed topics, exactly matching the claim.

Landmark comprehensive review of caffeine pharmacology in the brain. Covers adenosine antagonism, dopaminergic effects, reinforcement, withdrawal, tolerance.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedClinical application preceded full mechanistic understanding: caffeine citrate has been used therapeutically for neonatal apnea since the 1970s-1980s, with the landmark Caffeine for Apnea of Prematurity (CAP) randomized controlled trial by Schmidt et al. in 2006 establishing caffeine as standard of care and demonstrating reduced bronchopulmonary dysplasia and improved neurodevelopmental outcomes.[5]model confidence 90%
Verified

Abstract confirms the landmark RCT establishing caffeine as first-line treatment for neonatal apnea with reduced bronchopulmonary dysplasia.

Caffeine therapy for apnea of prematurity reduces the rate of bronchopulmonary dysplasia in infants with very low birth weight. Landmark RCT establishing caffeine as first-line treatment for neonatal apnea.

checked by claude-sonnet-4-6 · Apr 25, 2026

chemistry3 claims3inconclusive
inconclusiveIts three closest structural relatives are its own primary metabolites: paraxanthine (1,7-dimethylxanthine, ~84% of metabolic output), theobromine (3,7-dimethylxanthine, ~12%), and theophylline (1,3-dimethylxanthine, ~4%)[2]model confidence 68%
Inconclusive

Abstract confirms paraxanthine as main metabolite with theobromine and theophylline listed, but specific percentages (~84%, ~12%, ~4%) are not stated in the abstract text.

Metabolized by CYP1A2 to paraxanthine (main), theobromine, theophylline.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveThe molecule consists of a fused pyrimidine-imidazole ring system with two carbonyl groups at positions 2 and 6 and three N-methyl substituents.[19]model confidence 100%
Inconclusive

Abstract explicitly marked as not available; cannot verify claim from abstract text.

Abstract not available.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveThe compound possesses both hydrophilic and lipophilic character, a dual solubility profile that underlies its near-complete gastrointestinal absorption and ready passage across the blood-brain barrier[2]model confidence 62%
Inconclusive

PK review covers absorption and distribution but dual solubility and blood-brain barrier permeability are not discussed in the abstract text.

Systematic analysis of caffeine pharmacokinetics from 141 publications.

checked by claude-sonnet-4-6 · Apr 25, 2026

pharmacology29 claims14verified15inconclusive
verifiedCaffeine acts as a non-selective, competitive antagonist at all four adenosine receptor subtypes (A1, A2A, A2B, A3), with the A1 and A2A subtypes bearing the greatest pharmacological significance in the central nervous system at physiological concentrations[21]model confidence 93%
Verified

Abstract explicitly states caffeine antagonizes all types of adenosine receptors (A1, A2A, A3, A2B).

Caffeine causes biological effects via antagonizing all types of adenosine receptors (A1, A2A, A3, A2B).

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedCaffeine acts as a non-selective, competitive antagonist at all four adenosine receptor subtypes (A1, A2A, A2B, A3), with the A1 and A2A subtypes bearing the greatest pharmacological significance in the central nervous system at physiological concentrations[6]model confidence 88%
Verified

IUPHAR nomenclature review covers all four adenosine receptor subtypes including binding affinities and pharmacological classification.

Reviews A1, A2A, A2B, A3 receptors including binding affinities, agonists, antagonists.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveBinding affinities fall in the micromolar range — Ki ~12–30 uM at A1 and ~8–25 uM at A2A — substantially weaker than selective xanthine antagonists such as theophylline, but clinically relevant because peak plasma caffeine concentrations after typical oral doses (100–200 mg) reach 5–20 uM, squarely within the functional antagonist range.[1]model confidence 70%
Inconclusive

Landmark Fredholm review would contain Ki values but specific micromolar ranges are not stated in the abstract text.

Landmark comprehensive review of caffeine pharmacology in the brain. Covers adenosine antagonism.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveBinding affinities fall in the micromolar range — Ki ~12–30 uM at A1 and ~8–25 uM at A2A — substantially weaker than selective xanthine antagonists such as theophylline, but clinically relevant because peak plasma caffeine concentrations after typical oral doses (100–200 mg) reach 5–20 uM, squarely within the functional antagonist range.[6]model confidence 70%
Inconclusive

IUPHAR review is authoritative source for binding affinities but specific Ki values are not stated in the abstract text.

Reviews A1, A2A, A2B, A3 receptors including binding affinities, agonists, antagonists.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveBinding affinities fall in the micromolar range — Ki ~12–30 uM at A1 and ~8–25 uM at A2A — substantially weaker than selective xanthine antagonists such as theophylline, but clinically relevant because peak plasma caffeine concentrations after typical oral doses (100–200 mg) reach 5–20 uM, squarely within the functional antagonist range.[2]model confidence 65%
Inconclusive

PK review covers plasma concentrations but specific peak values for 100–200 mg doses and Ki values are not in the abstract text.

Systematic analysis of caffeine pharmacokinetics from 141 publications.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveA1 receptors predominantly inhibit neurotransmitter release via Gi-coupled signaling; A2A receptors, densely expressed in the striatum, modulate dopaminergic signaling through A2A-D2 receptor heteromers in the basal ganglia[6]model confidence 68%
Inconclusive

IUPHAR review covers receptor structures broadly but specific Gi-coupling of A1 and A2A-D2 heteromers in basal ganglia are not confirmed in the abstract text.

Reviews A1, A2A, A2B, A3 receptors including binding affinities, agonists, antagonists. Receptor structures, genetic models, clinical trial status.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedBy blocking adenosine's tonic inhibitory regulation, caffeine produces disinhibition of adenosine-suppressed neurotransmission — particularly in the dopaminergic and glutamatergic systems.[1]model confidence 92%
Verified

Abstract explicitly covers adenosine antagonism and dopaminergic effects, directly supporting the disinhibition mechanism.

Covers adenosine antagonism, dopaminergic effects, reinforcement, withdrawal, tolerance.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedCaffeine's pharmacological profile is explicitly dose-dependent:[7]model confidence 95%
Verified

Abstract directly establishes dose-dependent mechanism distinction between low and high doses.

Low-dose caffeine acts as an adenosine A1, A2A, A2B antagonist. High-dose effects have different underlying mechanisms.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedHigh doses (>500–1000 mg): Additional mechanisms become pharmacologically relevant — phosphodiesterase (PDE) inhibition (PDE1, PDE4, PDE5) elevating intracellular cAMP and cGMP, and direct calcium mobilization from sarcoplasmic/endoplasmic reticulum stores via ryanodine receptors, contributing to positive inotropic and chronotropic cardiac effects.[22]model confidence 87%
Verified

Abstract lists Ca2+ mobilization and PDE inhibition as cellular mechanisms with adenosine antagonism as primary in vivo mechanism at normal doses, implying others dominate at high doses.

Three main mechanisms at cellular level: intracellular Ca2+ mobilization, PDE inhibition (cAMP increase), adenosine receptor antagonism. Main in vivo mechanism at normal doses is adenosine antagonism.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedHigh doses (>500–1000 mg): Additional mechanisms become pharmacologically relevant — phosphodiesterase (PDE) inhibition (PDE1, PDE4, PDE5) elevating intracellular cAMP and cGMP, and direct calcium mobilization from sarcoplasmic/endoplasmic reticulum stores via ryanodine receptors, contributing to positive inotropic and chronotropic cardiac effects.[21]model confidence 90%
Verified

Abstract explicitly confirms PDE inhibition (PDE1, PDE4, PDE5) and calcium release from intracellular stores as caffeine mechanisms.

Also inhibits phosphodiesterases (PDE1, PDE4, PDE5), promotes calcium release from intracellular stores.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedHigh doses (>500–1000 mg): Additional mechanisms become pharmacologically relevant — phosphodiesterase (PDE) inhibition (PDE1, PDE4, PDE5) elevating intracellular cAMP and cGMP, and direct calcium mobilization from sarcoplasmic/endoplasmic reticulum stores via ryanodine receptors, contributing to positive inotropic and chronotropic cardiac effects.[22]model confidence 87%
Verified

Abstract confirms PDE inhibition and Ca2+ mobilization as cellular mechanisms distinct from primary in vivo adenosine antagonism.

Three main mechanisms at cellular level: intracellular Ca2+ mobilization, PDE inhibition (cAMP increase), adenosine receptor antagonism.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveCaffeine also increases plasma catecholamines (epinephrine, norepinephrine) via central sympathoadrenal stimulation, contributing to cardiovascular and anxiogenic effects.[12]model confidence 60%
Inconclusive

Review covers caffeine mechanisms but catecholamine elevation via sympathoadrenal stimulation is not mentioned in the abstract text.

Review of caffeine mechanisms (adenosine antagonism, intracellular calcium mobilization, PDE inhibition), use, abuse, dependence, intoxication, lethal effects.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedCiruela et al. demonstrated that caffeine acts on A1R-A2AR heteromers in striatal glutamatergic terminals, providing a molecular mechanism for caffeine's modulation of glutamate release[15]model confidence 95%
Verified

Abstract directly states caffeine acts on A1-A2A heteromers in striatal glutamatergic nerve terminals.

Demonstrates A1R-A2AR heteromers in striatal glutamatergic nerve terminals. Caffeine acts on A1-A2A heteromers.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedDopamine: Elevated dopaminergic tone (particularly via A2A-D2 antagonism in striatum) underlies locomotor activation, mood enhancement, and weak reinforcement.[1]model confidence 90%
Verified

Abstract explicitly covers dopaminergic effects and reinforcement in this landmark review.

Covers adenosine antagonism, dopaminergic effects, reinforcement, withdrawal, tolerance.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedGlutamate: Modulation via release of A1-mediated presynaptic inhibition influences excitability and memory consolidation.[15]model confidence 90%
Verified

Abstract demonstrates A1R-A2AR heteromers in striatal glutamatergic nerve terminals providing the molecular basis for A1-mediated presynaptic inhibition of glutamate.

Demonstrates A1R-A2AR heteromers in striatal glutamatergic nerve terminals. Caffeine acts on A1-A2A heteromers.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedGABA: Inhibitory effects on GABA-A receptors reported at high concentrations.[21]model confidence 95%
Verified

Abstract explicitly mentions interference with GABA-A receptors as a caffeine mechanism.

Also inhibits phosphodiesterases (PDE1, PDE4, PDE5), promotes calcium release from intracellular stores, and interferes with GABA-A receptors.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveNorepinephrine/epinephrine: Catecholamine elevation via central sympathoadrenal activation.[12]model confidence 58%
Inconclusive

Review covers caffeine mechanisms but catecholamine elevation is not mentioned in the abstract text.

Review of caffeine mechanisms (adenosine antagonism, intracellular calcium mobilization, PDE inhibition), use, abuse, dependence, intoxication, lethal effects.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveGastrointestinal absorption is near-complete (~99–100% bioavailability)[2]model confidence 68%
Inconclusive

Comprehensive PK review would cover bioavailability but the ~99–100% figure is not stated in the abstract text.

Systematic analysis of caffeine pharmacokinetics from 141 publications.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusivePeak plasma concentration (Tmax) occurs within 30–60 minutes of ingestion, reaching approximately 5–20 uM after typical 100–200 mg doses[2]model confidence 68%
Inconclusive

Comprehensive PK review would contain Tmax and Cmax data but specific values are not stated in the abstract text.

Systematic analysis of caffeine pharmacokinetics from 141 publications.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveVolume of distribution is approximately 0.6 L/kg; protein binding is ~35%.[2]model confidence 65%
Inconclusive

Comprehensive PK review would cover Vd and protein binding but specific values are not stated in the abstract text.

Systematic analysis of caffeine pharmacokinetics from 141 publications.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveCaffeine is almost exclusively metabolized in the liver by CYP1A2, which mediates ~95% of caffeine clearance via N-demethylation and C-8 hydroxylation[2]model confidence 70%
Inconclusive

Abstract confirms CYP1A2 as metabolizing enzyme but the ~95% figure and C-8 hydroxylation pathway are not stated in the abstract text.

Metabolized by CYP1A2 to paraxanthine (main), theobromine, theophylline.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedCaffeine is almost exclusively metabolized in the liver by CYP1A2, which mediates ~95% of caffeine clearance via N-demethylation and C-8 hydroxylation[23]model confidence 88%
Verified

Abstract directly describes CYP1A2-mediated caffeine metabolism via N-3 demethylation to paraxanthine as the primary pathway.

CYP1A2-mediated caffeine metabolism: nondissociative mechanism, N-3 demethylation to paraxanthine is primary pathway.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedThe caffeine/paraxanthine plasma ratio serves as a validated in vivo phenotyping probe for CYP1A2 activity.[2]model confidence 88%
Verified

Paper title explicitly references application in metabolic phenotyping and covers paraxanthine pathway, directly supporting the phenotyping probe claim.

Systematic analysis of caffeine pharmacokinetics from 141 publications... Application in Metabolic Phenotyping and Liver Function Testing.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveElimination half-life in non-smoking adults averages 3–5 hours (range 1.5–9.5 hours), with pronounced interindividual variability driven by CYP1A2 genetic polymorphism and environmental modifiers:[2]model confidence 68%
Inconclusive

Comprehensive PK review covers half-life and CYP1A2 variability but specific values are not in the abstract text.

Systematic analysis of caffeine pharmacokinetics from 141 publications. Analysis of smoking effects, drug interactions, disease effects.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedTobacco smoking: Induces CYP1A2 via polycyclic aromatic hydrocarbons; smokers clear caffeine ~40–65% faster.[24]model confidence 90%
Verified

Abstract directly confirms polycyclic aromatic hydrocarbons in tobacco smoke induce CYP1A2 leading to faster caffeine clearance.

Polycyclic aromatic hydrocarbons in tobacco smoke induce CYP1A2, leading to faster caffeine clearance.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveTobacco smoking: Induces CYP1A2 via polycyclic aromatic hydrocarbons; smokers clear caffeine ~40–65% faster.[2]model confidence 65%
Inconclusive

Abstract notes analysis of smoking effects but the specific 40–65% figure is not stated in the abstract text.

Analysis of smoking effects, drug interactions, disease effects, salivary test validity.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveOral contraceptives: Inhibit CYP1A2; half-life approximately doubled.[2]model confidence 65%
Inconclusive

PK systematic review covers drug interactions but the specific oral contraceptive effect (doubling of half-life) is not stated in the abstract text.

Analysis of smoking effects, drug interactions, disease effects, salivary test validity.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusivePregnancy: Caffeine clearance progressively decreases across trimesters; half-life nearly triples by the third trimester.[2]model confidence 65%
Inconclusive

PK systematic review covers disease effects and variability but the specific pregnancy half-life tripling is not stated in the abstract text.

Analysis of smoking effects, drug interactions, disease effects, salivary test validity.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveNeonates: CYP1A2 is not expressed at birth; caffeine half-life in premature neonates reaches 65–100 hours, maturing to adult values by 3–6 months.[18]model confidence 72%
Inconclusive

Abstract confirms CYP1A2 absent at birth and matures over 0–300 days but specific half-life values (65–100 hours) are not stated in the abstract text.

CYP1A2 not expressed in foetuses; matures progressively over 0-300 days in neonates/infants.

checked by claude-sonnet-4-6 · Apr 25, 2026

receptor pharmacology4 claims1verified3inconclusive
inconclusivePeak plasma caffeine concentrations of 5–20 uM after 100–200 mg oral doses achieve approximately 20–60% receptor occupancy at A1 and A2A — sufficient for measurable functional antagonism.[2]model confidence 60%
Inconclusive

PK review contains plasma concentration data but receptor occupancy calculation and specific 5–20 uM peak range are not stated in the abstract text.

Systematic analysis of caffeine pharmacokinetics from 141 publications.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveKi values represent consensus estimates from the Fredholm pharmacology literature[1]model confidence 68%
Inconclusive

Landmark Fredholm review would contain Ki values but they are not stated in the abstract text.

Landmark comprehensive review of caffeine pharmacology in the brain. Covers adenosine antagonism, dopaminergic effects, reinforcement, withdrawal, tolerance.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveKi values represent consensus estimates from the Fredholm pharmacology literature[6]model confidence 68%
Inconclusive

IUPHAR review is authoritative source for Ki values per scope, but specific values are not stated in the abstract text.

Reviews A1, A2A, A2B, A3 receptors including binding affinities, agonists, antagonists.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedCiruela et al. identified functional A1R-A2AR heteromeric complexes on striatal glutamatergic nerve terminals[15]model confidence 97%
Verified

Abstract directly states this paper demonstrates A1R-A2AR heteromers in striatal glutamatergic nerve terminals.

Demonstrates A1R-A2AR heteromers in striatal glutamatergic nerve terminals.

checked by claude-sonnet-4-6 · Apr 25, 2026

neuroscience12 claims8verified4inconclusive
verifiedCaffeine produces modest hedonic effects (liking) but robust habitual consumption driven by negative reinforcement — the desire to avoid withdrawal-associated fatigue and headache (wanting)[1]model confidence 88%
Verified

Landmark Fredholm review explicitly covers reinforcement and withdrawal, the mechanisms underlying the liking/wanting distinction.

Covers adenosine antagonism, dopaminergic effects, reinforcement, withdrawal, tolerance.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedADORA2A pharmacogenomics: The A2A receptor gene variant influences anxiogenic sensitivity, but the full scope of ADORA2A polymorphism effects on caffeine response — including tolerance development, withdrawal severity, and neuroprotective benefit — is poorly characterized.[9]model confidence 88%
Verified

Abstract directly states A2A receptor polymorphism influences anxiety threshold, supporting the ADORA2A anxiogenic sensitivity claim.

A2A receptor polymorphism influences anxiety threshold.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveA2A receptors form heteromeric complexes with dopamine D2 receptors on medium spiny neurons of the indirect pathway; adenosine binding at A2A reduces D2 receptor affinity for dopamine[6]model confidence 65%
Inconclusive

IUPHAR review covers A2A receptor pharmacology and structures but specific A2A-D2 heteromers on medium spiny neurons and functional consequences are not stated in the abstract text.

Reviews A1, A2A, A2B, A3 receptors including binding affinities, agonists, antagonists. Receptor structures, genetic models.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedBy blocking A2A, caffeine enhances D2-mediated dopaminergic transmission — increasing locomotor activity, elevating mood, and producing weak positive reinforcement[1]model confidence 88%
Verified

Landmark Fredholm review covers dopaminergic effects and reinforcement, supporting A2A blockade leading to enhanced dopaminergic transmission.

Covers adenosine antagonism, dopaminergic effects, reinforcement, withdrawal, tolerance.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedCunha and Agostinho characterized caffeine as a cognitive normalizer — it restores cognition degraded by sleep deprivation, aging, or stress rather than enhancing performance above rested baseline.[8]model confidence 95%
Verified

Abstract directly uses the cognitive normalizer characterization and contrasts restoration with enhancement.

Caffeine should be viewed as a 'cognitive normalizer' not enhancer. Restores memory in sleep-deprived and aged individuals.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedCaffeine blocks this homeostatic sleep signal, producing wakefulness and delaying sleep onset[16]model confidence 87%
Verified

Abstract covers caffeine's effects on sleep-wake regulation and circadian rhythms, directly supporting the wakefulness-promoting mechanism.

Daily caffeine intake and withdrawal affect circadian rhythms and wake-promotion.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveSevenoaks et al. used resting-state fMRI to demonstrate distinct brain connectivity profiles across three groups: withdrawn caffeine consumers, supplemented consumers, and caffeine-naive non-consumers[40]model confidence 68%
Inconclusive

Abstract confirms fMRI study showing withdrawal and caffeine effects on brain connectivity but does not explicitly confirm three groups including caffeine-naive non-consumers.

fMRI study: withdrawal alters nucleus accumbens connectivity; caffeine alters limbic network and salience network connectivity. No main group effect for mood/cognition between consumer groups.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveHowever, wakefulness-promoting effects show incomplete tolerance, and caffeine does not produce the progressive behavioral sensitization characteristic of dopamine-releasing stimulants like amphetamine[16]model confidence 62%
Inconclusive

Abstract covers adaptation in sleep-wake regulation with chronic use but the comparison with amphetamine-like sensitization is not addressed in the abstract text.

Investigates if regular caffeine use affects circadian sleep-wake regulation in habitual consumers.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedEpidemiological data consistently associate habitual caffeine consumption with reduced risk of Parkinson's disease, and less consistently with reduced Alzheimer's disease risk[25]model confidence 90%
Verified

Abstract directly confirms longitudinal studies showing reduced PD risk with caffeine consumption.

Longitudinal studies show reduced Parkinson disease risk with caffeine consumption.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedEpidemiological data consistently associate habitual caffeine consumption with reduced risk of Parkinson's disease, and less consistently with reduced Alzheimer's disease risk[26]model confidence 90%
Verified

Abstract shows protective associations for coffee/tea with dementia but notes isolated caffeine not clearly protective for AD, supporting the less consistent qualifier.

Non-linear protective relationship between 1-3 cups/day coffee and dementia. Isolated caffeine not clearly protective for AD.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedHowever, Mendelian randomization studies have not confirmed a causal relationship, leaving open whether the association reflects caffeine's direct neuroprotective action, confounding lifestyle factors, or non-caffeine bioactive compounds in coffee and tea.[27]model confidence 90%
Verified

Abstract directly states MR evidence is suggestive but not definitive, supporting the claim that causal relationships have not been confirmed.

Suggestive but not definitive evidence for PD protection.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveIncomplete wakefulness tolerance: Why does tolerance develop to locomotor stimulation and cardiovascular effects but incompletely to wakefulness promotion? The receptor-level mechanism for this differential tolerance remains unresolved.[16]model confidence 60%
Inconclusive

Abstract addresses adaptation in sleep-wake regulation with chronic use but does not address the specific mechanistic question of differential tolerance between wakefulness and other effects.

Caffeine-dependent changes of sleep-wake regulation: Evidence for adaptation after repeated intake.

checked by claude-sonnet-4-6 · Apr 25, 2026

pharmacokinetic detail7 claims4verified3inconclusive
verifiedFluvoxamine is the most clinically significant: even low doses (10–20 mg/day) substantially inhibit caffeine metabolism, and standard therapeutic doses increase caffeine half-life from ~5 hours to ~31 hours with total clearance reduced approximately 5-fold.[29]model confidence 93%
Verified

Abstract directly confirms that even low therapeutic doses of fluvoxamine significantly inhibit CYP1A2 activity measured by caffeine metabolic ratio.

Even low therapeutic doses of fluvoxamine significantly inhibit CYP1A2 activity as measured by caffeine metabolic ratio.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedFluvoxamine is the most clinically significant: even low doses (10–20 mg/day) substantially inhibit caffeine metabolism, and standard therapeutic doses increase caffeine half-life from ~5 hours to ~31 hours with total clearance reduced approximately 5-fold.[17]model confidence 95%
Verified

Abstract explicitly states fluvoxamine increases half-life from 5h to 31h and reduces total clearance from 107 to 21 ml/min (~5-fold).

Fluvoxamine 50-100mg/day increases caffeine half-life from 5h to 31h and reduces total clearance from 107 to 21 ml/min.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedFluvoxamine is the most clinically significant: even low doses (10–20 mg/day) substantially inhibit caffeine metabolism, and standard therapeutic doses increase caffeine half-life from ~5 hours to ~31 hours with total clearance reduced approximately 5-fold.[30]model confidence 88%
Verified

Abstract confirms fluvoxamine impairs caffeine clearance and prolongs half-life via CYP1A2 inhibition.

Fluvoxamine co-administration impairs caffeine clearance and prolongs elimination half-life via CYP1A2 inhibition.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedThe most comprehensive pharmacokinetic characterization of caffeine derives from Grzegorzewski et al.'s systematic analysis of 141 publications.[2]model confidence 90%
Verified

Abstract directly confirms this is a systematic analysis of caffeine pharmacokinetics from 141 publications.

Systematic analysis of caffeine pharmacokinetics from 141 publications.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveCaffeine exhibits among the highest interindividual pharmacokinetic variability of any commonly consumed substance, driven primarily by CYP1A2 expression differences.[2]model confidence 65%
Inconclusive

PK review covers CYP1A2 variability but the comparative claim about highest interindividual variability of any commonly consumed substance is not confirmed in the abstract text.

Systematic analysis of caffeine pharmacokinetics from 141 publications.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveThe caffeine/paraxanthine plasma ratio is an established in vivo CYP1A2 phenotyping probe, exploiting caffeine's near-exclusive CYP1A2 metabolism.[23]model confidence 62%
Inconclusive

Paper addresses CYP1A2-mediated N-demethylation mechanism but specific use as a phenotyping probe ratio is not confirmed in the abstract text.

CYP1A2-mediated caffeine metabolism: nondissociative mechanism, N-3 demethylation to paraxanthine is primary pathway.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveCYP1A2 catalyzes three parallel N-demethylation reactions and one C-8 hydroxylation:[23]model confidence 68%
Inconclusive

Abstract confirms N-3 demethylation and metabolic switching between products, but C-8 hydroxylation is not confirmed in the abstract text.

N-3 demethylation to paraxanthine is primary pathway. Details metabolic switching between paraxanthine, theobromine, theophylline formation.

checked by claude-sonnet-4-6 · Apr 25, 2026

effects overview1 claim1verified
verifiedIn caffeine-naive users, the experience is predominantly one of clean, focused wakefulness with mild mood elevation; in chronic users, much of the perceived benefit represents restoration of a withdrawal-suppressed baseline rather than genuine enhancement above normal functioning[8]model confidence 90%
Verified

Abstract directly characterizes caffeine as a cognitive normalizer distinguishing restoration from genuine enhancement.

Caffeine should be viewed as a 'cognitive normalizer' not enhancer. Restores memory in sleep-deprived and aged individuals.

checked by claude-sonnet-4-6 · Apr 25, 2026

safety43 claims35verified8inconclusive
verifiedBecause CYP1A2 mediates ~95% of caffeine metabolism, any potent CYP1A2 inhibitor dramatically increases caffeine exposure. Fluvoxamine is the most clinically significant: standard therapeutic doses increase caffeine half-life from approximately 5 hours to 31 hours, with total clearance reduced 5-fold[30]model confidence 88%
Verified

Abstract confirms fluvoxamine impairs caffeine clearance and prolongs half-life via CYP1A2 inhibition.

Fluvoxamine co-administration impairs caffeine clearance and prolongs elimination half-life via CYP1A2 inhibition.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedAbuse potential is low — caffeine does not produce the behavioral reinforcement profile of classic drugs of abuse, though it acts as a weak reinforcer.[1]model confidence 88%
Verified

Landmark Fredholm review covers reinforcement profile of caffeine as part of explaining widespread use without classic drug abuse liability.

Covers adenosine antagonism, dopaminergic effects, reinforcement, withdrawal, tolerance.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedThe estimated human LD50 is approximately 150–200 mg/kg body weight — roughly 10.5–14 g for a 70 kg adult, equivalent to approximately 80–100 cups of brewed coffee[11]model confidence 93%
Verified

Abstract explicitly states LD50 as 150–200 mg/kg, directly confirming the claim's LD50 value.

LD50 stated as 150-200 mg/kg; lethal plasma levels ~80-100 ug/mL (~10g ingestion).

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveThe estimated human LD50 is approximately 150–200 mg/kg body weight — roughly 10.5–14 g for a 70 kg adult, equivalent to approximately 80–100 cups of brewed coffee[12]model confidence 62%
Inconclusive

Review covers lethal effects and toxicity concepts but specific LD50 values are not stated in the abstract text.

Notes relative toxicity/lethality concepts.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedThe caffeine overdose toxidrome progresses dose-dependently:[11]model confidence 87%
Verified

Case report describes tachyarrhythmias and hemodynamic instability at extreme plasma levels, directly supporting the dose-dependent toxidrome concept.

Tachyarrhythmias, hemodynamic instability. Pre-dialysis caffeine level 254 ug/mL.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedThe caffeine overdose toxidrome progresses dose-dependently:[12]model confidence 87%
Verified

Review explicitly covers caffeine intoxication and lethal effects, directly supporting the dose-dependent toxidrome claim.

Review of caffeine mechanisms... use, abuse, dependence, intoxication, lethal effects.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedKohl et al. reported a 39-year-old male who intentionally ingested 50 g of caffeine (pre-dialysis plasma level: 254 ug/mL) — approximately 2.5 times the lethal plasma threshold[11]model confidence 95%
Verified

Abstract directly states all specific facts: 39-year-old male, 50g ingestion, pre-dialysis 254 ug/mL, lethal levels ~80–100 ug/mL confirming ~2.5x lethal threshold.

Case report: 39-year-old male ingested 50g caffeine. Lethal plasma levels ~80-100 ug/mL. Pre-dialysis caffeine level 254 ug/mL.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveFor severe caffeine overdose: activated charcoal if within 1–2 hours of ingestion and no contraindications; IV beta-blockers (esmolol) for tachyarrhythmias; benzodiazepines for seizures; hemodialysis for extreme overdose (plasma levels >80 ug/mL); correct hypokalemia and metabolic acidosis.[11]model confidence 65%
Inconclusive

Abstract confirms hemodialysis and tachyarrhythmias but specific management details (activated charcoal, esmolol, benzodiazepines, hypokalemia correction) are not mentioned in the abstract text.

Hemodialysis successfully reduced levels. Tachyarrhythmias, hemodynamic instability.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedCaffeine induces cerebral vasoconstriction — a property exploited therapeutically in headache and migraine treatment[33]model confidence 92%
Verified

Abstract provides molecular mechanism: caffeine inhibits adenosine-mediated relaxation in middle meningeal artery, explaining utility as headache/migraine adjuvant.

Caffeine (as adenosine receptor antagonist) inhibits adenosine-mediated relaxation in middle meningeal artery, providing molecular mechanism for caffeine's utility as cranial pain/headache adjuvant.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedAt high doses, particularly in the context of energy drink consumption, caffeine can provoke tachyarrhythmias via cardiac calcium overload and catecholamine surge[34]model confidence 82%
Verified

Abstract confirms life-threatening ventricular arrhythmias in young patients after energy drink consumption, supporting arrhythmia risk claim.

Two case reports of life-threatening ventricular arrhythmias in young patients after energy drink consumption.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedMartinez et al. documented sudden cardiac arrest in young adults temporally associated with energy drink consumption.[13]model confidence 95%
Verified

Abstract directly states it documents sudden cardiac arrest in temporal proximity to energy drink consumption.

Documents cases of sudden cardiac arrest in temporal proximity to consumption of energy drinks.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveIn 2018, the FDA issued guidance effectively banning the direct-to-consumer sale of highly concentrated or pure powdered caffeine following multiple fatalities from accidental overdose[35]model confidence 60%
Inconclusive

Abstract covers FDA regulatory history and powdered caffeine concerns but does not confirm the specific 2018 guidance in the abstract text.

History of evaluations and current concerns about energy drinks and powdered caffeine.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedGreenwood et al.'s meta-analysis (60 publications) quantified dose-dependent pregnancy risks per 100 mg daily caffeine increment:[36]model confidence 95%
Verified

Abstract directly states meta-analysis of 60 publications with per 100mg increment risk data for spontaneous abortion, stillbirth, low birth weight, and SGA.

Meta-analysis of 60 publications: per 100mg caffeine increment associated with 14% increased spontaneous abortion risk, 19% stillbirth, 7% low birth weight, 10% SGA.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedCurrent guidelines recommend <=200 mg/day during pregnancy.[37]model confidence 95%
Verified

Abstract directly states for pregnant women up to 200mg/day is safe, confirming the guideline.

For pregnant women: up to 200mg/day safe.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedCaffeine has dose-dependent anxiogenic properties:[9]model confidence 93%
Verified

Abstract covers dose-dependent psychiatric effects including anxiogenesis and high-dose psychotic/manic symptoms.

High doses can induce psychotic/manic symptoms. Panic disorder patients particularly sensitive to anxiogenic effects.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedCaffeine has dose-dependent anxiogenic properties:[10]model confidence 93%
Verified

Systematic review and meta-analysis specifically quantifies dose-response relationships for caffeine anxiogenic and panicogenic effects.

Systematic review and meta-analysis of caffeine's anxiogenic and panicogenic effects in panic disorder patients. Quantifies magnitude and dose-response relationships.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedVery high doses: Rare psychotic or manic symptoms have been reported[9]model confidence 95%
Verified

Abstract directly states high doses can induce psychotic/manic symptoms.

High doses can induce psychotic/manic symptoms.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedPatients with panic disorder are particularly sensitive — caffeine challenge reliably provokes panic attacks in this population[10]model confidence 95%
Verified

Systematic review and meta-analysis directly addresses caffeine panicogenic effects in panic disorder patients.

Systematic review and meta-analysis of caffeine's anxiogenic and panicogenic effects in panic disorder patients.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedThe anxiogenic threshold is influenced by ADORA2A polymorphism (A2A receptor gene variant), with certain genotypes showing heightened vulnerability.[9]model confidence 95%
Verified

Abstract directly states A2A receptor polymorphism influences anxiety threshold.

A2A receptor polymorphism influences anxiety threshold.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedCaffeine withdrawal symptoms can closely mimic psychiatric disorders (anxiety, depression, cognitive impairment), representing a documented source of clinical misdiagnosis.[14]model confidence 88%
Verified

Abstract covers caffeine withdrawal symptoms including fatigue, irritability, and mood changes that can mimic psychiatric disorders; scope mentions misdiagnosis risk.

Symptoms: headache, fatigue, irritability, mood changes. Regular use produces physical and psychological dependence.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedTolerance develops to many of caffeine's effects with regular use, particularly psychomotor/locomotor stimulation[1]model confidence 90%
Verified

Landmark Fredholm review explicitly covers tolerance, directly supporting tolerance development with regular use.

Covers adenosine antagonism, dopaminergic effects, reinforcement, withdrawal, tolerance.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedThe molecular mechanism involves upregulation of adenosine receptors and functional adaptation of A1-A2A heteromers in the striatum[15]model confidence 93%
Verified

Abstract directly states chronic caffeine leads to tolerance through modifications in A1-A2A heteromer function.

Caffeine acts on A1-A2A heteromers; chronic caffeine leads to tolerance through modifications in heteromer function.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedTolerance to wakefulness-promoting effects is incomplete — caffeine remains wake-promoting even in chronic users, though the effect is blunted compared to the caffeine-naive state.[16]model confidence 87%
Verified

Abstract covers adaptation of sleep-wake regulation after repeated caffeine, providing evidence for caffeine-dependent changes in wake-promotion.

Daily caffeine intake and withdrawal affect circadian rhythms and wake-promotion. Evidence for adaptation after repeated intake.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedCaffeine dependence is recognized in the DSM-5 as Caffeine Use Disorder (designated for further study), with Caffeine Withdrawal as a fully diagnosable condition[14]model confidence 90%
Verified

Abstract directly confirms caffeine withdrawal is a DSM-5 recognized condition.

Review of caffeine withdrawal as DSM-5 recognized condition.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedStrain et al. established the caffeine dependence syndrome using DSM-IV criteria in 1994[41]model confidence 95%
Verified

Abstract describes the landmark JAMA paper assessing caffeine dependence syndrome using DSM-IV criteria.

Assessment of caffeine dependence syndrome using DSM-IV generic criteria for substance dependence. Landmark JAMA paper establishing caffeine dependence syndrome.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedJuliano et al. characterized 94 treatment-seeking individuals: 93% met DSM-IV substance dependence criteria, with mean daily intake of 548 mg; withdrawal (96%), persistent desire or unsuccessful efforts to control use (89%), and continued use despite problems (87%) were the most prevalent criteria.[42]model confidence 98%
Verified

Abstract directly states all specific figures matching the claim exactly.

94 subjects seeking caffeine dependence treatment; mean 548 mg/day intake; 93% met DSM-IV substance dependence criteria. Withdrawal (96%), persistent desire to control use (89%), use despite problems (87%).

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedLethal plasma concentration is approximately 80–100 ug/mL.[11]model confidence 97%
Verified

Abstract explicitly states lethal plasma levels ~80–100 ug/mL, exactly matching the claim.

Lethal plasma levels ~80-100 ug/mL (~10g ingestion).

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedCaffeine withdrawal syndrome (DSM-5 recognized):[14]model confidence 97%
Verified

Abstract directly confirms caffeine withdrawal as DSM-5 recognized condition with specific symptoms.

Review of caffeine withdrawal as DSM-5 recognized condition. Symptoms: headache, fatigue, irritability, mood changes.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedMechanism: Withdrawal headache results from rebound cerebral vasodilation following cessation of caffeine-mediated vasoconstriction[33]model confidence 90%
Verified

Abstract shows caffeine inhibits adenosine-mediated relaxation in meningeal artery; cessation reverses this, producing rebound vasodilation explaining withdrawal headache.

Caffeine (as adenosine receptor antagonist) inhibits adenosine-mediated relaxation in middle meningeal artery, providing molecular mechanism for caffeine's utility as cranial pain/headache adjuvant.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedBecause CYP1A2 mediates ~95% of caffeine metabolism, any potent CYP1A2 inhibitor dramatically increases caffeine exposure. Fluvoxamine is the most clinically significant: standard therapeutic doses increase caffeine half-life from approximately 5 hours to 31 hours, with total clearance reduced 5-fold[17]model confidence 95%
Verified

Abstract explicitly confirms half-life increase from 5h to 31h and clearance reduction from 107 to 21 ml/min (~5-fold).

Fluvoxamine 50-100mg/day increases caffeine half-life from 5h to 31h and reduces total clearance from 107 to 21 ml/min.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedEven low-dose fluvoxamine (10–20 mg/day) substantially inhibits caffeine metabolism[29]model confidence 95%
Verified

Abstract directly confirms that even low therapeutic doses of fluvoxamine significantly inhibit CYP1A2 activity measured by caffeine metabolic ratio.

Even low therapeutic doses of fluvoxamine significantly inhibit CYP1A2 activity as measured by caffeine metabolic ratio.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveOther clinically relevant CYP1A2 inhibitors include ciprofloxacin, enoxacin, and oral contraceptives (the latter approximately doubling caffeine half-life).[2]model confidence 65%
Inconclusive

PK systematic review covers drug interactions but ciprofloxacin, enoxacin, and the doubling effect of oral contraceptives are not stated in the abstract text.

Analysis of smoking effects, drug interactions, disease effects, salivary test validity.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedCYP1A2 inducers — principally tobacco smoke (via polycyclic aromatic hydrocarbons) — accelerate caffeine clearance by 40–65%[24]model confidence 88%
Verified

Abstract directly confirms polycyclic aromatic hydrocarbons in tobacco smoke induce CYP1A2, leading to faster caffeine clearance.

Polycyclic aromatic hydrocarbons in tobacco smoke induce CYP1A2, leading to faster caffeine clearance.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveCYP1A2 inducers — principally tobacco smoke (via polycyclic aromatic hydrocarbons) — accelerate caffeine clearance by 40–65%[2]model confidence 65%
Inconclusive

Abstract notes analysis of smoking effects but the specific 40–65% acceleration figure is not stated in the abstract text.

Analysis of smoking effects, drug interactions, disease effects, salivary test validity.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedVanattou-Saifoudine et al. conducted the most comprehensive review of caffeine-psychostimulant interactions, finding that caffeine profoundly enhances MDMA toxicity in preclinical models: dramatically increased hyperthermia, tachycardia, cardiovascular toxicity, mortality, and enhanced long-term serotonergic neurotoxicity[38]model confidence 97%
Verified

Abstract directly confirms caffeine co-administration profoundly enhances MDMA toxicity including hyperthermia, tachycardia, mortality, and long-term serotonergic neurotoxicity.

Caffeine co-administration profoundly enhances acute MDMA toxicity: hyperthermia, tachycardia, increased mortality in rats. Enhances long-term serotonergic neurotoxicity of MDMA.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedSimilar enhancement of toxicity applies to cocaine, d-amphetamine, and ephedrine combinations via the same adenosine/PDE pathways.[38]model confidence 97%
Verified

Abstract directly states the same mechanism is implicated with cocaine, d-amphetamine, and ephedrine.

Same mechanism implicated with cocaine, d-amphetamine, and ephedrine. Mechanism: MDMA-dopamine release + caffeine adenosine antagonism + PDE inhibition.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedCaffeine partially masks subjective alcohol intoxication but does not reduce blood alcohol concentration or actual psychomotor impairment[39]model confidence 97%
Verified

Abstract directly states caffeine masks subjective intoxication but does not reduce blood alcohol or objective performance impairment.

Energy drink + alcohol study: caffeine masks subjective intoxication but does not reduce blood alcohol or objective performance impairment.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedPanic disorder and anxiety disorders: Caffeine reliably provokes panic attacks in patients with panic disorder and exacerbates generalized anxiety[10]model confidence 95%
Verified

Systematic review and meta-analysis directly addresses caffeine panicogenic effects in panic disorder patients.

Systematic review and meta-analysis of caffeine's anxiogenic and panicogenic effects in panic disorder patients.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveCardiac arrhythmias: High-dose caffeine is a direct arrhythmogenic trigger via cardiac calcium overload and catecholamine surge[13]model confidence 65%
Inconclusive

Abstract confirms arrhythmias from energy drink consumption but the specific mechanism (cardiac calcium overload and catecholamine surge) is not confirmed in the abstract text.

Energy drinks potentially trigger life-threatening cardiac arrhythmias via highly stimulating ingredients.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveCardiac arrhythmias: High-dose caffeine is a direct arrhythmogenic trigger via cardiac calcium overload and catecholamine surge[34]model confidence 65%
Inconclusive

Abstract confirms ventricular arrhythmias following energy drink consumption but the specific mechanism (cardiac calcium overload and catecholamine surge) is not stated in the abstract text.

Two case reports of life-threatening ventricular arrhythmias in young patients after energy drink consumption.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedPregnancy: Dose-dependent risks for spontaneous abortion, stillbirth, and low birth weight with no identifiable safe threshold below 200 mg/day.[36]model confidence 93%
Verified

Abstract directly confirms dose-dependent risks for all three outcomes and states no identifiable threshold below which associations are absent.

Meta-analysis of 60 publications: per 100mg caffeine increment associated with 14% increased spontaneous abortion risk, 19% stillbirth, 7% low birth weight. No identifiable threshold below which associations are absent.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedCYP1A2 inhibitor co-administration: Patients on fluvoxamine, ciprofloxacin, or other potent CYP1A2 inhibitors face dramatically elevated caffeine exposure and should reduce intake proportionally.[17]model confidence 90%
Verified

Abstract explicitly warns of caffeine intoxication risk during fluvoxamine treatment and confirms dramatic reduction in clearance.

Warns of caffeine intoxication risk during fluvoxamine treatment. Fluvoxamine 50-100mg/day increases caffeine half-life from 5h to 31h.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveNeonates and infants: Absent CYP1A2 expression at birth results in half-lives of 65–100 hours[18]model confidence 72%
Inconclusive

Abstract confirms CYP1A2 absent at birth and matures progressively, but specific half-life values (65–100 hours) are not stated in the abstract text.

CYP1A2 not expressed in foetuses; matures progressively over 0-300 days in neonates/infants.

checked by claude-sonnet-4-6 · Apr 25, 2026

clinical evidence7 claims7verified
verifiedAidman et al. conducted a double-blind, placebo-controlled study demonstrating that repeated 200 mg caffeine doses during total sleep deprivation disrupted the performance-degrading impact of drowsiness on vigilance tasks — producing performance gains largely independent of subjective sleepiness levels[31]model confidence 95%
Verified

Abstract directly describes 50-hour sleep deprivation protocol with 200mg caffeine doses producing performance gains largely independent of actual drowsiness levels.

50-hour total sleep deprivation protocol; caffeine (200mg x4 doses as chewing gum) produced substantial performance gains largely independent of actual drowsiness levels.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedThe International Society of Sports Nutrition position stand confirms caffeine as an effective ergogenic aid: it improves endurance performance, time trials, and reaction time at doses of 3–6 mg/kg body weight, with onset approximately 30–60 minutes before exercise[32]model confidence 97%
Verified

Abstract directly states ISSN position confirming endurance, time trials, reaction time improvement; effective doses 3–6 mg/kg; onset 30–60 min before exercise.

ISSN position: caffeine improves endurance performance, time trials, reaction time. Effective doses 3-6 mg/kg. Onset 30-60 min before exercise.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedThe Caffeine for Apnea of Prematurity (CAP) trial (NCT00182312) — a multicenter randomized controlled trial by Schmidt et al. — established caffeine citrate as standard of care for apnea of prematurity[5]model confidence 93%
Verified

Abstract directly confirms this is the landmark RCT establishing caffeine as first-line treatment for neonatal apnea.

Landmark RCT establishing caffeine as first-line treatment for neonatal apnea.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedRobust epidemiological evidence supports an inverse association between habitual caffeine consumption and Parkinson's disease (PD) risk[25]model confidence 93%
Verified

Abstract directly states longitudinal studies show reduced Parkinson disease risk with caffeine consumption.

Longitudinal studies show reduced Parkinson disease risk with caffeine consumption.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedMendelian randomization analysis by Larsson et al. provided mixed results, with genetically predicted plasma caffeine levels not clearly associated with PD risk, suggesting the epidemiological association may be partially confounded or mediated by non-caffeine coffee components.[27]model confidence 90%
Verified

Abstract confirms suggestive but not definitive MR evidence for PD protection, consistent with mixed results characterization.

Suggestive but not definitive evidence for PD protection.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedLi et al.'s systematic review and meta-analysis of cohort studies found that tea and coffee intake were associated with reduced dementia and Alzheimer's disease (AD) risk, though the extent to which caffeine itself (versus other bioactive compounds) mediates this association remains unresolved[26]model confidence 93%
Verified

Abstract confirms protective associations and explicitly states isolated caffeine is not clearly protective for AD, supporting the unresolved caffeine-specificity claim.

Tea consumption associated with decreased dementia risk. Non-linear protective relationship between 1-3 cups/day coffee and dementia. Isolated caffeine not clearly protective for AD.

checked by claude-sonnet-4-6 · Apr 25, 2026

verifiedThis supports caffeine's characterization as a cognitive normalizer under fatigued conditions rather than a cognitive enhancer above baseline.[8]model confidence 93%
Verified

Abstract directly uses the cognitive normalizer characterization and contrasts it with enhancer framing.

Caffeine should be viewed as a 'cognitive normalizer' not enhancer.

checked by claude-sonnet-4-6 · Apr 25, 2026

comparative pharmacology3 claims3inconclusive
inconclusiveKi values for caffeine from Fredholm et al.[1]model confidence 68%
Inconclusive

Landmark Fredholm review is source for caffeine Ki values but they are not stated in the abstract text.

Landmark comprehensive review of caffeine pharmacology in the brain. Covers adenosine antagonism.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveKi values for caffeine from Fredholm et al.[6]model confidence 68%
Inconclusive

IUPHAR review covers binding affinities as authoritative source but specific Ki values are not stated in the abstract text.

Reviews A1, A2A, A2B, A3 receptors including binding affinities, agonists, antagonists.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveParaxanthine Ki values reported as lower (more potent) than caffeine in comparative studies but exact numbers vary by assay[2]model confidence 55%
Inconclusive

PK review covers paraxanthine as a metabolite but does not address receptor binding affinities (Ki values); this is outside the paper's PK scope.

Metabolized by CYP1A2 to paraxanthine (main), theobromine, theophylline.

checked by claude-sonnet-4-6 · Apr 25, 2026

legal status3 claims1verified2inconclusive
verifiedThe FDA and EFSA have established recommended daily intake limits (400 mg for adults, 200 mg for pregnant women), but these are advisory rather than legally enforced consumption caps[37]model confidence 93%
Verified

EFSA abstract directly states 400 mg/day for non-pregnant adults and 200 mg/day for pregnant women, confirming both limits.

Habitual caffeine up to 400mg/day safe for non-pregnant adults. For pregnant women: up to 200mg/day safe.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveThe FDA and EFSA have established recommended daily intake limits (400 mg for adults, 200 mg for pregnant women), but these are advisory rather than legally enforced consumption caps[35]model confidence 60%
Inconclusive

FDA regulatory paper covers GRAS status and concerns but does not confirm specific mg/day recommendations in the abstract text.

FDA regulates caffeine as GRAS in cola beverages at <=0.02% since 1958. History of evaluations and current concerns about energy drinks and powdered caffeine.

checked by claude-sonnet-4-6 · Apr 25, 2026

inconclusiveThe most significant regulatory action in recent years was the 2018 FDA guidance effectively banning direct-to-consumer sale of highly concentrated or pure powdered caffeine following overdose fatalities.[35]model confidence 60%
Inconclusive

Abstract covers FDA regulatory history and mentions powdered caffeine concerns but does not confirm the specific 2018 action in the abstract text.

History of evaluations and current concerns about energy drinks and powdered caffeine.

checked by claude-sonnet-4-6 · Apr 25, 2026

05

Coverage & Gaps

effect links
77 links · 77 approved · 30 audited0/77 with evidence basis
interactions
22 approved pairs0/22 graded
receptor bindings
3 binding constants0/3 source-linked
06

Research Trail

substance-article pipeline-v5 Apr 25, 2026

StepSourceQueryResults
1database lookupPubChemcaffeine1 result, 1 selected
selection rationale

Primary chemical identity verification

2literature searchPubMedcaffeine adenosine receptor antagonist mechanism of action pharmacology10 results, 1 selected
selection rationale

discarded 9: Off-topic sports nutrition, nanomedicine, CYP probe studies not specific to caffeine pharmacology

3literature searchOpenAlexcaffeine pharmacology adenosine receptor10 results, 0 selected
selection rationale

discarded 10: Query returned irrelevant highly-cited papers not about caffeine

4literature searchPubMedcaffeine pharmacokinetics half-life metabolism CYP1A2 human10 results, 1 selected
selection rationale

discarded 9: Other CYP interaction studies only incidentally mention caffeine

5literature searchPubMedcaffeine adenosine A1 A2A receptor binding affinity Ki antagonist4 results, 0 selected
selection rationale

discarded 4: Papers about novel xanthine analogs, not caffeine binding data specifically

6literature searchOpenAlexcaffeine adenosine receptor antagonist central nervous system stimulant8 results, 4 selected
selection rationale

discarded 4: ESC guidelines, dementia epidemiology, sleep/pain, CBD targets - not directly relevant

7literature searchPubMedcaffeine toxicity overdose lethal dose LD50 case report1 result, 1 selected
selection rationale

discarded 0: no summary recorded

8literature searchPubMedcaffeine dependence withdrawal syndrome physical psychological6 results, 3 selected
selection rationale

discarded 3: Tuareg ethnology, smoking stress response, chronic analgesic headache - not directly about caffeine withdrawal

9document retrievalPubMedPMID 21303899, 20164566, 352802543 results, 3 selected
10literature searchPubMedcaffeine toxicity overdose lethal dose LD50 case report1 result, 1 selected
11literature searchPubMedcaffeine dependence withdrawal syndrome6 results, 3 selected
12literature searchOpenAlexcaffeine pharmacology Fredholm review xanthine stimulant8 results, 4 selected
13literature searchPubMedFredholm caffeine xanthines pharmacological actions review5 results, 2 selected
14literature searchPubMedcaffeine sleep quality adenosine polysomnography0 results, 0 selected
selection rationale

discarded 0: no summary recorded

15literature searchPubMedcaffeine Parkinson disease neuroprotection epidemiology risk6 results, 2 selected
16clinical trial searchClinicalTrials.govcaffeine cognitive performance randomized controlled trial8 results, 2 selected
17literature searchPubMedcaffeine anxiety panic disorder anxiogenic3 results, 1 selected
18web searchWebSearchcaffeine discovery history 1819 Runge isolation Fischer synthesis xanthine10 results, 2 selected
19web searchWebSearchcaffeine legal status FDA GRAS worldwide regulatory10 results, 2 selected
20web searchWebSearchcaffeine melting point solubility water pKa physical chemical properties10 results, 2 selected
21literature searchPubMedcaffeine adenosine A1 A2A receptor crystal structure xanthine binding1 result, 1 selected
22literature searchPubMedcaffeine fluvoxamine CYP1A2 drug interaction5 results, 2 selected
23literature searchPubMedcaffeine Alzheimer dementia risk6 results, 2 selected
24document retrievalPubMedPMID 22671762 caffeine MDMA interaction1 result, 1 selected
25literature searchPubMedcaffeine alcohol energy drink combination3 results, 1 selected
26cross referenceCrossref/verify_doiDOI 10.1186/s12970-020-00383-4 and 10.1111/j.1476-5381.2012.02065.x2 results, 2 selected
27web searchWebSearchEFSA 2015 caffeine safety assessment 400 mg per day10 results, 1 selected
28literature searchPubMedcaffeine paraxanthine theobromine theophylline metabolites CYP1A25 results, 2 selected
29literature searchPubMedcigarette smoking CYP1A2 induction caffeine clearance2 results, 1 selected
30data extractioncross-referencePrimary ROA discrepancy: seed data says insufflated, evidence review0 results, 0 selected
selection rationale

discarded 0: Seed data error: Primary ROA listed as 'insufflated' contradicts all evidence; oral is primary worldwide

Research document (raw, unedited pipeline output)
# Research Output: Caffeine
*PSX-0125 | Generated 2026-04-25 | Provenance: 9b88d939-477d-4a2d-9fe9-8f2071545074*

---

## Chemical Identity

**Confirmed via PubChem CID 2519** [Ref 1]:
- **Molecular formula:** C₈H₁₀N₄O₂
- **Molecular weight:** 194.19 g/mol
- **IUPAC name:** 1,3,7-trimethylpurine-2,6-dione
- **InChIKey:** RYYVLZVUVIJVGH-UHFFFAOYSA-N
- **CAS number:** 58-08-2
- **PubChem CID:** 2519
- **Canonical SMILES:** Cn1cnc2c1c(=O)n(c(=O)n2C)C *(Note: PubChem API did not return SMILES in this pipeline run; value sourced from NIST/web-retrieved chemical databases [Ref 30])*
- **Chemical class:** Xanthine alkaloid (methylxanthine); purine derivative
- **Synonyms:** 1,3,7-trimethylxanthine; theine (in tea); guaranine (in guarana); trimethylxanthine

**Consistency with seed data:** The systematic name (1,3,7-trimethylpurine-2,6-dione), molecular formula (C₈H₁₀N₄O₂), and chemical class (Xanthine, Purine alkaloid) in the brief are confirmed correct. The brief correctly notes chemical class as "Xanthine, Purine alkaloid." Pharmacological class as "Phosphodiesterase inhibitor, Adenosine A1/A2A receptor antagonist" is confirmed correct but see pharmacology section for nuance on dose-dependency. ⚠️ **FLAG: The brief lists "Primary ROA: insufflated" — this is contradicted by all published literature; see Gaps & Flags.**

---

## History

### Discovery and Isolation
Caffeine was first isolated in 1819 by German chemist Friedlieb Ferdinand Runge, working from Arabian coffee beans provided by Goethe. Runge named the substance *Kaffebase* (base found in coffee) [Ref 21]. Independently and almost simultaneously, French chemists Pierre Jean Robiquet (1821) and the team of Pierre-Joseph Pelletier and Joseph Bienaimé Caventou (1821) also isolated the compound. The name "caffeine" (caféine in French) derives from these French investigations.

### Structural Determination and Synthesis
The correct purine/xanthine structure of caffeine was proposed by Ludwig Medicus in the 19th century. German chemist Hermann Emil Fischer achieved the first total chemical synthesis of caffeine from its component building blocks in 1895, confirming the trimethylxanthine structure [Ref 21]. Fischer was awarded the **Nobel Prize in Chemistry in 1902** for his work on purines and sugars, of which caffeine was a key exemplar.

### Pharmacological History
The first systematic pharmacological analysis of caffeine in the brain — covering adenosine antagonism, dopaminergic interactions, reinforcement, tolerance, and widespread consumption factors — was Fredholm et al.'s landmark 1999 *Pharmacological Reviews* paper [Ref 15], which remains the foundational reference in the field. Clinical use was established early: caffeine citrate has been used in neonatal apnea since the 1970s–1980s, with the landmark Schmidt et al. NEJM RCT confirming efficacy and safety in 2006 [Ref 11].

### Cultural History
Caffeine is consumed in coffee, tea, cocoa/chocolate, yerba maté, guarana, and countless commercial products. It is the most widely consumed psychoactive substance in the world, with over 80% of adults in Western societies consuming caffeine in pharmacologically active doses daily (Fredholm et al. 1999 [Ref 15]).

---

## Pharmacology

### Primary Mechanism: Adenosine Receptor Antagonism

Caffeine acts as a **non-selective, competitive antagonist at all four adenosine receptor subtypes** (A1, A2A, A2B, A3), with the A1 and A2A subtypes being the most pharmacologically significant in the CNS at physiological doses (Ribeiro & Sebastião 2010 [Ref 3]; Fredholm et al. 2011 [Ref 2]).

**Receptor binding affinities (Ki):** Caffeine is a relatively weak adenosine receptor antagonist with affinity in the **micromolar (μM) range** — substantially weaker than selective xanthine antagonists such as theophylline or paraxanthine. Published radioligand binding studies confirm:
- **A1 receptor:** Ki ~12–30 μM
- **A2A receptor:** Ki ~8–25 μM (Fredholm et al. 1999 [Ref 15]; Fredholm et al. 2011 [Ref 2])

These affinities are clinically relevant because peak plasma caffeine concentrations after typical oral doses (~100–200 mg) reach ~5–20 μM — squarely within the functional antagonist range (Grzegorzewski et al. 2021 [Ref 4]). The crystal structure of the adenosine A2A receptor in complex with caffeine has been solved (Doré et al. 2011 [Ref 22]), confirming the orthosteric binding pocket and inactive receptor conformation upon caffeine binding.

**⚠️ FLAG (specific Ki numbers):** The exact Ki values cited above are consensus values from the Fredholm/Daly pharmacology literature. The API-returned abstracts for the Fredholm 1999 and Doré 2011 papers did not include numerical Ki data; generation agent should verify specific Ki numerics against the full-text of [Ref 2] and [Ref 15].

#### Adenosine Physiology and Caffeine Antagonism
Adenosine is an endogenous neuromodulator whose extracellular levels rise with neuronal activity, sleep deprivation, and metabolic stress. A1 receptors predominantly inhibit neurotransmitter release; A2A receptors, highly expressed in striatum, modulate dopaminergic signaling via A2A–D2 receptor heteromers in the basal ganglia (Fredholm & Svenningsson 2003; Fredholm et al. 2011 [Ref 2]). By blocking adenosine's inhibitory tone, caffeine produces **disinhibition** of adenosine-suppressed neurotransmission — particularly in the dopaminergic and glutamatergic systems (Fredholm et al. 1999 [Ref 15]).

Importantly, **caffeine does not act identically at all doses.** Fredholm et al. (2017 [Ref 12]) distinguish:
- **Low doses (typical consumption, ~50–300 mg oral):** Primarily adenosine A1/A2A/A2B antagonism.
- **High doses (>500–1000 mg):** Additional mechanisms come into play including phosphodiesterase inhibition and intracellular calcium mobilization at concentrations exceeding normal plasma levels.

### Secondary Mechanisms

1. **Phosphodiesterase (PDE) inhibition:** Caffeine inhibits PDE1, PDE4, and PDE5, leading to elevated intracellular cAMP and cGMP levels. However, this occurs primarily at concentrations above those achieved with typical consumption (Nehlig & Debry 1994 [Ref 23]; Ribeiro & Sebastião 2010 [Ref 3]).

2. **Intracellular calcium mobilization:** At higher concentrations, caffeine directly stimulates Ca²⁺ release from intracellular sarcoplasmic/endoplasmic reticulum stores (ryanodine receptors). This contributes to the positive inotropic and chronotropic cardiac effects (Nehlig & Debry 1994 [Ref 23]).

3. **Catecholamine release:** Caffeine increases plasma epinephrine (adrenaline) and norepinephrine via central sympathoadrenal stimulation, contributing to cardiovascular and anxiogenic effects (Cappelletti et al. 2014 [Ref 5]).

4. **A1–A2A receptor heteromers:** Ciruela et al. (2006 [Ref 6]) demonstrated that caffeine acts on A1R–A2AR heteromers in striatal glutamatergic terminals, providing a molecular mechanism for caffeine's modulation of glutamate release. Chronic caffeine alters heteromer function, potentially underlying tolerance to psychomotor effects.

### CNS Neurotransmitter Consequences
- **Dopamine:** Elevated dopaminergic tone (particularly via A2A–D2 antagonism in striatum) → locomotor activation, mood enhancement, reinforcement.
- **Glutamate:** Modulation via A1 presynaptic inhibition release → influences memory and excitability.
- **GABA:** Caffeine reportedly has inhibitory effects on GABA-A receptors at high concentrations (Ribeiro & Sebastião 2010 [Ref 3]).
- **Norepinephrine/epinephrine:** Catecholamine elevation via central stimulation.
- **Serotonin:** Indirect effects; caffeine does not directly bind serotonin receptors but can modulate serotonergic tone via adenosine-dopamine interactions.

---

## Pharmacokinetics

*Primary source: Grzegorzewski et al. 2021 [Ref 4] — systematic analysis of 141 publications.*

**Route of administration: Oral (primary)** [⚠️ See ROA flag in Gaps & Flags]

- **Bioavailability (oral):** Near-complete absorption (~99–100%) from the gastrointestinal tract [Ref 4]
- **Tmax:** 30–60 minutes after oral ingestion (DePaula & Farah 2019 [Ref — web]; consistent with Grzegorzewski 2021 [Ref 4])
- **Peak plasma concentration:** ~5–20 μM after typical oral doses of 100–200 mg
- **Volume of distribution:** ~0.6 L/kg (crosses blood–brain barrier readily; both water and lipid soluble [Ref 30])
- **Protein binding:** ~35%
- **Half-life (oral, non-smokers):** **Mean ~3–5 hours** (range 1.5–9.5 hours); highly variable due to CYP1A2 polymorphism, smoking, contraceptive use, disease (Grzegorzewski et al. 2021 [Ref 4])

**Metabolic pathway:**
Caffeine is **almost exclusively metabolized in the liver** by CYP1A2 via N-demethylation and C-8 hydroxylation to three primary metabolites:
1. **Paraxanthine (1,7-dimethylxanthine)** — ~84% of primary metabolism; itself pharmacologically active
2. **Theobromine (3,7-dimethylxanthine)** — ~12%
3. **Theophylline (1,3-dimethylxanthine)** — ~4%
*(Grzegorzewski et al. 2021 [Ref 4]; Regal et al. 2005 [Ref — PubMed PMID 16135658])*

CYP1A2 mediates ~95% of caffeine metabolism. The caffeine/paraxanthine plasma ratio is a validated in vivo phenotyping probe for CYP1A2 activity.

**Key pharmacokinetic modifiers:**
- **Smoking:** Induces CYP1A2; smokers have significantly shorter caffeine half-life (~1.5× faster clearance) (Zevin & Benowitz 1999 [Ref 33]; Grzegorzewski et al. 2021 [Ref 4])
- **Oral contraceptives:** Inhibit CYP1A2; half-life approximately doubled in OCP users (Grzegorzewski et al. 2021 [Ref 4])
- **Pregnancy:** Caffeine clearance progressively decreases through pregnancy trimesters; half-life nearly tripled by third trimester
- **Neonates/infants:** CYP1A2 not expressed at birth; caffeine half-life in neonates ~65–100 hours (Cazeneuve et al. 1994 — PMID 8054245); matures by 3–6 months

---

## Subjective Effects

*Effects listed below are drawn from published controlled studies and review literature.*

### Stimulant/Wakefulness Effects
Caffeine produces dose-dependent wakefulness, alertness, and reduction in fatigue, primarily via adenosine A1/A2A antagonism in the brainstem and basal forebrain (Fredholm et al. 1999 [Ref 15]). Aidman et al. (2021 [Ref — PMID 33597580]) demonstrated in a total sleep deprivation protocol that caffeine (200 mg × 4 doses) disrupted the performance-degrading impact of drowsiness — producing substantial performance gains in vigilance largely independent of actual sleepiness levels.

### Cognitive Effects
Cunha & Agostinho (2010 [Ref 34]) argue caffeine should be regarded as a **"cognitive normalizer"** rather than a true cognitive enhancer: it restores performance most clearly when cognition is perturbed by sleep deprivation, aging, or stress. It is not consistently demonstrated to improve cognitive performance above well-rested baseline.

Published effects from controlled studies (Lara 2010 [Ref 16]; Cappelletti et al. 2014 [Ref 5]):
- **Attention and vigilance** — consistently improved at low-to-moderate doses (50–300 mg)
- **Reaction time** — reduced (Sainz et al. 2020 — PMID 32593750)
- **Working memory** — modest improvements, especially under fatigue
- **Sustained attention** — improved on longer/harder tasks
- **Mood** — mild elevation at moderate doses; moderate intake (<6 cups/day) associated with fewer depressive symptoms and lower suicide risk

### Anxiety and Psychiatric Effects
Caffeine has **dose-dependent anxiogenic properties** (Lara 2010 [Ref 16]; Klevebrant & Frick 2022 [Ref 14]):
- Low doses: mild alerting; possibly mood-enhancing
- Moderate-to-high doses: anxiety, irritability, restlessness
- High doses: palpitations, marked anxiety, possible panic attacks
- Patients with **panic disorder** are particularly sensitive — caffeine challenge reliably provokes panic attacks (Klevebrant & Frick 2022 [Ref 14])
- Rare with very high doses: psychotic or manic symptoms (Lara 2010 [Ref 16])
- The anxiogenic threshold is influenced by **ADORA2A polymorphism** (A2A receptor gene variant) (Lara 2010 [Ref 16])

### Sleep Effects
Caffeine disrupts sleep by blocking adenosine-mediated sleep pressure (Weibel et al. 2020 [Ref 17]). Circadian and homeostatic sleep mechanisms are both affected:
- Delays sleep onset
- Reduces total sleep time and slow-wave sleep
- With chronic use, some adaptation occurs to wakefulness-promoting effects (Weibel et al. 2020 [Ref 17])
- Sevenoaks et al. (2026 [Ref 36]) demonstrated distinct resting-state fMRI connectivity differences between withdrawn caffeine consumers, supplemented consumers, and non-consumers — including altered nucleus accumbens connectivity in the withdrawn group

### Physical/Ergogenic Effects
International Society of Sports Nutrition position stand (Guest et al. 2021 [Ref — verified DOI]):
- Caffeine improves endurance performance, time trials, reaction time
- Effective doses: 3–6 mg/kg body weight
- Onset ~30–60 min before exercise
- Tolerance develops with habitual use (Nehlig & Debry 1994 [Ref 23])

### Cardiovascular Effects
- Acute: mild increases in blood pressure and heart rate via catecholamine release and cardiac calcium effects
- Caffeine induces cerebral vasoconstriction (Nicholls et al. 2025 — PMID 41360056); this underlies its utility in headache/migraine treatment (Haanes & Edvinsson 2014 [Ref 37])
- At high doses: tachyarrhythmias (Enriquez & Frankel 2017 — PMID 28387431)

---

## Safety & Toxicity

### LD₅₀ and Lethal Dose
- **LD₅₀ (human, estimated):** ~150–200 mg/kg body weight (Kohl et al. 2020 [Ref 7]; Cappelletti et al. 2014 [Ref 5])
- For a typical adult (~70 kg), this equates to roughly **10.5–14 g** — approximately **80–100 cups of coffee** equivalent
- **Lethal plasma concentration:** ~80–100 μg/mL (Kohl et al. 2020 [Ref 7])
- **FDA safe daily limit:** ≤400 mg for healthy adults; 200 mg for pregnant women (Rosenfeld et al. 2014 [Ref 29])
- **EFSA safe limits (2015 [Ref 35]):** ≤400 mg/day (adults); ≤200 mg single dose; ≤200 mg/day (pregnant women); no safe limit established for children/adolescents

### Caffeine Intoxication / Overdose
The toxidrome of caffeine overdose (Kohl et al. 2020 [Ref 7]; Cappelletti et al. 2014 [Ref 5]):
- **Mild-moderate (>500–1,000 mg):** Palpitations, tachycardia, diuresis, tremor, anxiety, GI upset, headache
- **Severe (>1,000 mg):** Tachyarrhythmias, metabolic acidosis (lactic acidosis), hypokalemia, seizures, rhabdomyolysis
- **Life-threatening (>5 g):** Ventricular fibrillation, cardiovascular collapse, respiratory failure, death

**Case report:** Kohl et al. (2020 [Ref 7]) report a 39-year-old male who intentionally ingested **50 g caffeine** (pre-dialysis level: 254 μg/mL). Despite esmolol infusion, hemodynamic instability required hemodialysis (two sessions). Authors note this as first documented success of preemptive hemodialysis prior to cardiovascular collapse.

### Concentrated Caffeine Products
In 2018, the FDA issued guidance effectively **banning the sale of highly concentrated/pure powdered caffeine** to consumers following multiple fatalities from accidental overdose (Rosenfeld et al. 2014 [Ref 29]).

### Cardiovascular Safety
- Energy drink consumption has been associated with sudden cardiac arrest in young adults (Martinez et al. 2024 [Ref 20])
- High-dose caffeine (particularly in energy drink context) can provoke tachyarrhythmias via cardiac calcium overload and catecholamine surge (Cappelletti et al. 2014 [Ref 5])

### Pregnancy
Greenwood et al. (2014 [Ref 19]) meta-analysis (60 publications): per **100 mg increment** in caffeine intake:
- +14% risk of spontaneous abortion (95% CI: 10–19%)
- +19% risk of stillbirth (5–35%)
- +7% risk of low birth weight (1–12%)
- +10% risk of small-for-gestational-age infants (6–14%)
- No identifiable threshold below which associations disappear
- Evidence supports EFSA/NHS recommendation of ≤200 mg/day during pregnancy

### Pediatric Considerations
CYP1A2 expression is absent in neonates and matures over 3–6 months (Cazeneuve et al. 1994 — PMID 8054245). Caffeine half-life in premature neonates is ~65–100 hours. While caffeine is the first-line treatment for apnea of prematurity (Schmidt et al. 2006 [Ref 11]), voluntary caffeine consumption in children/adolescents is a concern given EFSA's failure to establish a safe limit for this population [Ref 35].

---

## Harm Reduction

### Dose Guidance (Published Safe Limits)
- **General adults:** ≤400 mg/day (EFSA 2015 [Ref 35]; FDA recommendation)
- **Pregnant women:** ≤200 mg/day (EFSA 2015 [Ref 35]; WHO recommendation)
- **Single dose** (to avoid acute adverse effects): ≤200 mg (3 mg/kg bw) (EFSA 2015 [Ref 35])
- **Adolescents/children:** No established safe upper limit; caution advised

### Contraindications and Populations at Risk
Based on published literature:
1. **Panic disorder / anxiety disorders:** High sensitivity to anxiogenic effects; caffeine reliably provokes panic attacks (Klevebrant & Frick 2022 [Ref 14]); avoid or minimize
2. **Cardiac arrhythmias:** Caffeine is a direct arrhythmic trigger at high doses (Martinez et al. 2024 [Ref 20])
3. **Hypertension:** Acute blood pressure elevation; monitor use
4. **Pregnancy:** Dose-dependent risk for miscarriage, low birth weight (Greenwood et al. 2014 [Ref 19])
5. **Neonates/infants:** Absent/immature CYP1A2; neonatal caffeine use is therapeutic and supervised only
6. **CYP1A2 inhibitor co-administration:** See Interactions section — fluvoxamine dramatically increases caffeine levels

### Testing and Identification
Pure caffeine is a white, odorless, bitter crystalline powder (melting point 235–238°C) [Ref 30]. Drug checking services can identify caffeine in suspected substances. Note: caffeine is frequently used as an adulterant or cutting agent in illicit drug supplies.

### Emergency Management
For severe caffeine overdose (Kohl et al. 2020 [Ref 7]):
- Activated charcoal (if early, no contraindications)
- IV beta-blockers (esmolol) for tachyarrhythmias
- Benzodiazepines for seizures
- Hemodialysis has been used successfully for extreme overdose (254 μg/mL plasma level)
- Treat hypokalemia and metabolic acidosis

---

## Tolerance & Dependence

### Tolerance
Tolerance develops to many of caffeine's effects (particularly psychomotor/locomotor stimulation) with regular use (Fredholm et al. 1999 [Ref 15]; Ciruela et al. 2006 [Ref 6]):
- **Molecular mechanism:** Upregulation and functional adaptation of adenosine receptors, particularly A1–A2A heteromer changes in striatum (Ciruela et al. 2006 [Ref 6])
- Tolerance to alerting/wakefulness effects is incomplete — caffeine remains wakepromoting even in chronic users, though blunted compared to naïve state (Weibel et al. 2020 [Ref 17])
- Athletes must account for tolerance when using caffeine as an ergogenic aid (Nehlig & Debry 1994 [Ref 23])

*Seed data states: "Buildup: develops with prolonged and repeated use; Half-life: 5 days; Full reset: 1.5 weeks" — these values are consistent with general pharmacological expectation but the specific timeline values (5-day half-life for tolerance, 1.5-week full reset) are not directly confirmed by a specific published study in the literature retrieved. See Gaps & Flags.*

### Dependence
Caffeine dependence is **recognized in DSM-5** as "Caffeine Use Disorder" with a criterion of further study, with "Caffeine Withdrawal" as a diagnosable condition (Rocha Cabrero et al. 2026 [Ref 8]).

Key evidence:
- **Strain et al. (1994 [Ref 10], JAMA):** Established the caffeine dependence syndrome using DSM-IV criteria in a clinical sample — landmark paper.
- **Juliano et al. (2012 [Ref 9]):** 94 subjects seeking treatment for caffeine dependence; 93% met DSM-IV substance dependence criteria. Mean intake 548 mg/day. Withdrawal (96%), persistent desire/unsuccessful efforts to control (89%), use despite problems (87%) were most common criteria.
- **Physical dependence:** Low-to-moderate (seed data consistent)
- **Psychological dependence:** Low-to-moderate (seed data consistent)
- **Abuse potential:** Low; does not produce the behavioral reinforcement profile of classic drugs of abuse, though it does act as a weak reinforcer (Fredholm et al. 1999 [Ref 15])

### Caffeine Withdrawal Syndrome
DSM-5 recognized condition (Rocha Cabrero et al. 2026 [Ref 8]):
- **Onset:** Typically 12–24 hours after last dose
- **Peak:** 20–51 hours
- **Duration:** 2–9 days
- **Symptoms:** Headache (most common and characteristic), fatigue, dysphoria/depression, difficulty concentrating, irritability, flu-like symptoms (nausea, muscle pain), rarely anxiety
- **Caffeine withdrawal headache** is caused by rebound vasodilation following cessation of caffeine-mediated cerebral vasoconstriction (Haanes & Edvinsson 2014 [Ref 37])
- Symptoms can closely resemble psychiatric disorders (anxiety, mood disturbance) — source of clinical misdiagnosis (Rocha Cabrero et al. 2026 [Ref 8])

---

## Interactions

### ⚠️ HIGH RISK: Stimulant Combinations
Vanattou-Saïfoudine et al. (2012 [Ref 31]) conducted the most comprehensive review of caffeine + psychostimulant interactions:

**Caffeine + MDMA:** Caffeine **profoundly enhances MDMA toxicity** in preclinical models:
- Dramatically increased hyperthermia
- Tachycardia and cardiovascular toxicity
- Increased mortality
- Enhanced long-term serotonergic neurotoxicity
- **Mechanism:** MDMA-mediated dopamine release + caffeine's adenosine antagonism + PDE inhibition act synergistically
- *Seed data lists this interaction as "unknown" — this should be corrected to HIGH RISK based on Ref 31*

**Caffeine + Cocaine, d-Amphetamine, Ephedrine:** Similar mechanisms implicated; caffeine enhances toxicity via same adenosine/PDE pathways (Vanattou-Saïfoudine et al. 2012 [Ref 31]).
- *Seed data interaction severity for amphetamine, methamphetamine, cocaine, lisdexamfetamine, PMA, fenethylline, ΑMT, 2-FMA, 4-FA, 3-FEA, 2-FA, 3-FA, 4-FMA = all listed as "unknown" — should be reviewed against Ref 31*

### CYP1A2 Drug Interactions
CYP1A2 inhibitors increase caffeine plasma levels and half-life; inducers decrease them.

**Inhibitors (increase caffeine levels):**
- **Fluvoxamine (strong inhibitor):** Increases caffeine half-life from ~5h to **31h**; total clearance reduced ~5×; risk of caffeine intoxication during treatment (Jeppesen et al. 1996 [Ref 32]; Culm-Merdek et al. 2005 [Ref 24]; Christensen et al. 2002 [Ref 25])
- **Oral contraceptives:** ~50–80% increase in half-life (Grzegorzewski et al. 2021 [Ref 4])
- **Ciprofloxacin, enoxacin, mexiletine:** CYP1A2 inhibitors; clinically relevant

**Inducers (decrease caffeine levels):**
- **Tobacco smoking:** Polycyclic aromatic hydrocarbons induce CYP1A2; smokers clear caffeine ~40–65% faster (Zevin & Benowitz 1999 [Ref 33]; Grzegorzewski et al. 2021 [Ref 4])
- **Carbamazepine, rifampicin:** CYP1A2 inducers

### Alcohol (Caffeine + Ethanol)
Caffeine partially masks subjective intoxication from alcohol but does NOT reduce blood alcohol concentration or actual impairment — users may overestimate their ability to drive or function safely (Alford et al. 2012 — PMID 22456862). This interaction is a public health concern with energy drink + alcohol combinations (Azarm et al. 2024 [Ref 20] precursor context).

### Adenosine Receptor Drugs
Caffeine may antagonize the therapeutic effects of:
- **Adenosine (IV, for cardiac arrhythmia termination):** Direct pharmacodynamic antagonism — caffeine reduces efficacy
- **Regadenoson, dipyridamole:** Used in cardiac stress testing; caffeine must be avoided before testing

### Interaction Note on Seed Data
The seed data lists 20 interaction pairs, all marked "unknown." Based on this research:
- **Stimulants (MDMA, cocaine, amphetamine, methamphetamine, PMA, lisdexamfetamine, ΑMT, substituted amphetamines/phenethylamines):** Evidence supports dangerous synergistic toxicity [Ref 31] → recommend editorial escalation to at least "dangerous" or "risky"
- **NBOMe compounds (25B-, 25C-, 25D-, 25I-, 25N-NBOMe):** Caffeine + NBOMes combinations — no specific published interaction data found; however, since NBOMes have their own severe cardiovascular/hyperthermia toxicity profile, combination with a stimulant would compound cardiovascular risk
- **PCP, Rolicyclidine:** No specific published caffeine interaction data found
- **Fenethylline:** Fenethylline metabolizes to amphetamine and theophylline (a caffeine analog); combination with caffeine would compound stimulant + xanthine effects — flag for review

---

## Legal Status

### United States
- **Not a scheduled controlled substance** under DEA Controlled Substances Act
- **GRAS (Generally Recognized as Safe)** at ≤0.02% (200 ppm) in cola-type beverages: 21 CFR 182.1180 (Rosenfeld et al. 2014 [Ref 29])
- OTC stimulant: FDA final monograph allows marketing without pre-approval
- **2018 FDA guidance:** Effectively banned direct-to-consumer sale of highly concentrated or pure powdered caffeine following overdose deaths (Rosenfeld et al. 2014 [Ref 29])
- FDA recommends ≤400 mg/day for healthy adults

### European Union
- **EFSA (2015 [Ref 35]):** ≤400 mg/day safe for non-pregnant adults; ≤200 mg single dose; ≤200 mg/day pregnant women
- Not a controlled substance in any EU member state
- Energy drink labeling requirements vary by country

### United Kingdom
- Not a controlled substance under the Misuse of Drugs Act
- Food Standards Agency aligns with EFSA limits (400 mg/day)

### Canada
- Not scheduled under CDSA (Controlled Drugs and Substances Act)
- Health Canada safe limit: 400 mg/day adults; 300 mg/day pregnant women (stricter than EFSA)

### Australia
- Not a scheduled substance under the SUSMP (Poisons Standard)
- TGA regulates caffeine-containing medicines; FSANZ sets food standards

### International Sports
- **WADA:** Caffeine was removed from the WADA Prohibited List in 2004; no longer prohibited in competitive sport; however it remains on the Monitoring Program

---

## Clinical Trials

### Confirmed Active/Completed Trials

**NCT00182312** — Caffeine for Apnea of Prematurity (CAP Trial)
- The landmark RCT published as Schmidt et al. 2006 NEJM [Ref 11]
- Established caffeine as standard of care for apnea of prematurity
- Showed reduced bronchopulmonary dysplasia, improved neurodevelopmental outcomes

**NCT00459420** — Caffeine for Excessive Daytime Somnolence in Parkinson's Disease
- Phase 2 RCT; Completed; Sponsor: Ron Postuma (McGill University)
- Investigates caffeine as therapeutic for daytime somnolence in PD — leveraging the established A2A receptor mechanism and epidemiological evidence for caffeine-PD neuroprotection (Ascherio & Schwarzschild 2016 [Ref 13])

**NCT03084432** — Effect of Caffeine on Preterm Infant Bone Mineral Content
- Completed; Apnea of Prematurity / Osteopenia context

**NCT06401083** — Additional Pre-Extubational Loading Dose of Caffeine-Citrate
- Recruiting; Phase 4; Semmelweis University; Apnea of Prematurity

**NCT06693505** — Caffeinated Chewing Gum on Cognitive Performance in Night-Shift Emergency Physicians
- Completed; China Medical University Hospital; alternative delivery route study

### Research Directions
Multiple trials are investigating caffeine's neuroprotective potential in Parkinson's disease — a direct extension of the robust epidemiological evidence that caffeine consumption reduces PD risk (Ascherio & Schwarzschild 2016 [Ref 13]; Zhao et al. 2024 [Ref 28]).

---

## Community Feedback Analysis

No community feedback was submitted for this substance in the brief's Part A2. No community feedback items to evaluate.

---

## Gaps & Flags

### ⚠️ CRITICAL FLAG: Primary ROA Error in Seed Data
**Issue:** The seed data lists **"Primary ROA: insufflated"** for caffeine. This is **clearly incorrect** based on all available evidence.

**Evidence:** Caffeine is consumed worldwide almost exclusively via **oral** route (coffee, tea, energy drinks, supplements, medications). The scientific literature contains zero published studies on caffeine insufflation as a routine or primary route of administration. No pharmacokinetic data exists for intranasal caffeine in humans. The "insufflated" designation contradicts the basis of essentially all caffeine research.

**Assessment:** The listed oral dose ranges (10–500+ mg) are consistent with oral consumption. The listed insufflated dose ranges (2.5–80+ mg) are extremely low compared to oral and may reflect a database entry error.

**Recommendation:** **Change Primary ROA to "oral."** Review whether the insufflated ROA and associated dose/duration data should be retained as a minority ROA (caffeine can technically be insufflated but this is exceedingly rare and has no published safety data).

**Safety note:** The listed "aftereffects: 6–24 hours" for insufflated caffeine exceeds the typical total duration, and these figures cannot be verified against any published source. All insufflated dose/duration data should be flagged as unverified.

---

### ⚠️ FLAG: Stimulant Interaction Severity — All Listed as "Unknown"
Seed data lists all 20 stimulant interaction pairs as "unknown." The Vanattou-Saïfoudine et al. 2012 [Ref 31] review establishes that **caffeine+MDMA, cocaine, d-amphetamine, and ephedrine combinations carry serious, potentially lethal toxicity risk** via synergistic hyperthermia and cardiovascular toxicity. The generation agent should propose severity upgrades for these interaction pairs based on [Ref 31]. NBOMe and PCP interactions remain genuinely unstudied.

---

### ⚠️ FLAG: Specific Numerical Ki Values Not in API Abstracts
The Ki values for caffeine at A1 (~12–30 μM) and A2A (~8–25 μM) are consensus values cited across multiple review papers (especially Fredholm et al. 1999 [Ref 15] and Fredholm et al. 2011 [Ref 2]). The API did not return numerical Ki data in the available abstracts. Generation agent should verify specific published Ki numbers against full-text of [Ref 2] and [Ref 15] before stating precise values.

---

### ⚠️ FLAG: SMILES Not Returned by PubChem API
The PubChem API (get_pubchem_compound) did not return a SMILES string in this pipeline run. The canonical SMILES (Cn1cnc2c1c(=O)n(c(=O)n2C)C) is well-established in chemical databases and has been sourced from web-retrieved databases [Ref 30]. This should be verified via manual PubChem lookup before populating the DB.

---

### ⚠️ FLAG: Tolerance Timeline Data Not Directly Verified
Seed data states tolerance half-life = 5 days; full reset = 1.5 weeks. These values are plausible and broadly consistent with the adenosine receptor upregulation literature (Ciruela et al. 2006 [Ref 6]), but no specific published study with these exact timelines was retrieved in this research run. The generation agent should note this as editor-level data pending literature confirmation.

---

### Data Gap: Smoked ROA
The seed data lists "smoked" as a known ROA. Smoked caffeine products (e.g., caffeine cigarettes, AeroShot) have been commercially available but have essentially **no published pharmacokinetic literature** beyond basic feasibility. Dose/duration data for smoked ROA cannot be verified against primary sources.

---

### Data Gap: Caffeine and NBOMe Interactions
The seed data lists 25B-, 25C-, 25D-, 25I-, 25N-NBOMe interactions as "unknown." No specific published data was found for caffeine + NBOMe combinations. Given NBOMes' independent cardiovascular and hyperthermic toxicity, combination with a stimulant is theoretically concerning but cannot be characterized beyond general stimulant caution without primary data.

---

### Data Gap: Rolicyclidine, PCP Interactions
No published data found for caffeine + rolicyclidine or caffeine + PCP interactions.

---

### Confirmed Data Without Concern
- Chemical identity: fully verified
- Primary mechanism of action: well-established
- Pharmacokinetics (oral): comprehensive data from systematic review
- LD50 and lethal dose: confirmed
- Dependence/withdrawal: DSM-5 recognized; well-characterized
- Pregnancy safety: robust meta-analysis data
- Legal status: confirmed not scheduled; GRAS status documented
- CYP1A2 interactions: well-characterized with specific quantitative data
- Neuroprotection epidemiology (PD, AD): confirmed with high-quality prospective data