Improvement after a supplement is not the same as a comparative result
Start with the comparison. A before-and-after change alone cannot isolate the contribution of an ingredient.
Caffeine
Exercise performance
Partial research summary
Studies with an unspecified organism reported a change with no confirmed comparison group in Exercise performance.
Single-ingredient studies · Population not specified · comparator not confirmed
Publications for this result: 2 (1 awaiting review included)
Other outcomes and study details
Distinct publications linked to this topic: 21
- Study conditions differ; results are shown descriptively and are not pooled.
- Some publications have not cleared source or interpretation review.
- Some papers could not be reviewed, so this summary covers only findings that could be confirmed.
What has been studied?
Distinct human comparison papers: 7
Some studies tested several ingredients together. Those results cannot show this ingredient’s separate effect or establish synergy.
anaerobic performance
- Comparison result:A better result than the comparison group was reported.General population · Single ingredient
Measurement and study conditions
- Controlled human study
- Randomized trial
- Single ingredient
- General population
- Primary outcome
- Between-group comparison
- Timepoint: acute 6 mg/kg before Wingate test
- Study comparison: acute caffeine versus placebo on Wingate anaerobic performance
average power
- Comparison result:A better result than the comparison group was reported.General population · Single ingredient
Measurement and study conditions
- Controlled human study
- Randomized trial
- Single ingredient
- General population
- Secondary outcome
- Between-group comparison
- Timepoint: Acute experimental trials separated by at least 48 hours
- Study comparison: Caffeine versus placebo.
cycling time to exhaustion
- Comparison result:No clear difference from the comparison group was found.Single ingredient
Measurement and study conditions
- Controlled human study
- Randomized trial
- Single ingredient
- Unspecified population
- Outcome role not specified
- Between-group comparison
- Timepoint: unspecified
- Study comparison: unspecified
Exercise performance
Research with unclear study subject or type
- A change was reported, but the comparison could not be confirmed.Single ingredient
Measurement and study conditions
- Review
- Review
- Single ingredient
- Unspecified population
- Outcome role not specified
- Study subject not identified
- Timepoint: unspecified
- Study comparison: unspecified
Other measured outcomes (7)
muscular strength/power
- Comparison result:A better result than the comparison group was reported.General population · Single ingredient · acute dosing before trial
Measurement and study conditions
- Controlled human study
- Randomized trial
- Single ingredient
- General population
- Secondary outcome
- Between-group comparison
- Timepoint: acute dosing before trial
- Study comparison: caffeine versus placebo at 75%, 90%, and 100% 1RM
Peak power
- Comparison result:A better result than the comparison group was reported.Single ingredient
Measurement and study conditions
- Controlled human study
- Randomized trial
- Single ingredient
- Unspecified population
- Outcome role not specified
- Between-group comparison
- Timepoint: unspecified
- Study comparison: unspecified
psychophysiological performance
Research with unclear study subject or type
- A change was reported, but the comparison could not be confirmed.Single ingredient
Measurement and study conditions
- Review
- Systematic review
- Single ingredient
- Unspecified population
- Outcome role not specified
- Study subject not identified
- Timepoint: unspecified
- Study comparison: unspecified
RT
- Comparison result:A worse result than the comparison group was reported.General population · Single ingredient
Measurement and study conditions
- Controlled human study
- Randomized trial
- Single ingredient
- General population
- Secondary outcome
- Between-group comparison
- Timepoint: 3-day caffeine intake; test after treatment
- Between-group comparison
Strength measures
- Comparison result:No clear difference from the comparison group was found.General population · Combination study · 5 days
Measurement and study conditions
- Controlled human study
- Randomized trial
- Ingredient combination
- General population
- Secondary outcome
- Between-group comparison
- Timepoint: 5 days
- Study comparison: performance across creatine/caffeine/coffee combinations versus placebo; no treatment interaction
strength resistance
Research with unclear study subject or type
- A change was reported, but the comparison could not be confirmed.Single ingredient
Measurement and study conditions
- Meta-analysis
- Meta-analysis
- Single ingredient
- Unspecified population
- Outcome role not specified
- Study subject not identified
- Timepoint: unspecified
- Study comparison: unspecified
TT performance
Other human research, including observational studies and reviews
- A change was reported before and after use. Its cause cannot be isolated.General population · Single ingredient
Measurement and study conditions
- Controlled human study
- Controlled trial
- Single ingredient
- General population
- Primary outcome
- Within-group comparison
- Timepoint: Approximately 30-minute simulated cycling time trial
- Study comparison: Caffeine versus placebo and no supplement in crossover trials; habitual intake did not modify the re
This result is held for interpretation. That does not mean research found no effect.
Outcome groups needing additional review (6)
- eccentric strength of the knee flexors · Distinct papers needing additional review: 1Evidence criteria check: 1
- lower body muscular endurance · Distinct papers needing additional review: 1Evidence criteria check: 1
- LPT · Distinct papers needing additional review: 1Evidence criteria check: 1
- LSPT · Distinct papers needing additional review: 1Evidence criteria check: 1
- OTPT · Distinct papers needing additional review: 1Evidence criteria check: 1
- Performance time · Distinct papers needing additional review: 1Evidence criteria check: 1
One paper can appear under multiple measured outcomes or review reasons, so these counts should not be added together.
Research topics to explore
Research topics to explore
Review findings reported in each research area.
Topics with more linked papers appear first.
- View selected result
Exercise performance
Related papers 21 · human comparison papers 7
- View results
Blood pressure and vascular health
Related papers 2 · comparison findings need further review
- SBP and DBP · Comparison details unconfirmed
- View results
Bone, joint, and mobility
Related papers 2 · comparison findings need further review
- knee extensor muscle thickness · Interpretation held
- mobility · Interpretation held
- View results
Fatigue and energy
Related papers 2 · comparison findings need further review
- markers of fatigue · Finding details are limited
- View results
Digestion and gut health
Related papers 1 · comparison findings need further review
- View results
Glucose and metabolic health
Related papers 1 · comparison findings need further review
- BMI · Comparison details unconfirmed
- fat mass · Comparison details unconfirmed
More outcome groups: 1
- View results
Men's health
Related papers 1 · comparison findings need further review
- View results
Skin, hair, and nails
Related papers 1 · comparison findings need further review
- hair growth · Interpretation held
- View results
Sleep
Related papers 1 · comparison findings need further review
- View results
Stress and mood
Related papers 1 · comparison findings need further review
- cortisol levels · Finding details are limited
- View results
Women's health
Related papers 1 · comparison findings need further review
- low birth weight · Comparison details unconfirmed
- preterm delivery · Comparison details unconfirmed
More outcome groups: 3
Data coverage and detailed study classifications
Paper counts are separated by research topic below. One paper may contribute to multiple findings.
Analysis status: Public-ready
Ingredient analysis date:
How the grade was determined: B
The 64.2 evidence score includes 23 independent human studies and 23 clinical studies in the representative benefit area. The B tier reflects evidence directness, repetition, and consistency together.
The grade is based mainly on the research area with the most evidence. It does not mean the same result was found in other areas.
Papers linked to this ingredient overall: 67
Papers linked to this topic 21 · distinct papers 21 · papers with a confirmed human comparison 7 · papers needing additional review 10
Ingredient-wide papers, topic-linked papers, and sources in outcome groups use different scopes. Adding source counts may count the same paper more than once.
- anaerobic performance: Sources represented in this measurement group: 1 · Distinct papers with human comparisons for this outcome: 1
- average power: Sources represented in this measurement group: 1 · Distinct papers with human comparisons for this outcome: 1
- cycling time to exhaustion: Sources represented in this measurement group: 1 · Distinct papers with human comparisons for this outcome: 1
- eccentric strength of the knee flexors: Sources represented in this measurement group: 1 · Distinct papers with human comparisons for this outcome: 0
- Exercise performance: Sources represented in this measurement group: 2 · Distinct papers with human comparisons for this outcome: 0
- lower body muscular endurance: Sources represented in this measurement group: 1 · Distinct papers with human comparisons for this outcome: 0
- LPT: Sources represented in this measurement group: 1 · Distinct papers with human comparisons for this outcome: 0
- LSPT: Sources represented in this measurement group: 1 · Distinct papers with human comparisons for this outcome: 0
- muscular strength/power: Sources represented in this measurement group: 1 · Distinct papers with human comparisons for this outcome: 1
- OTPT: Sources represented in this measurement group: 1 · Distinct papers with human comparisons for this outcome: 0
- Peak power: Sources represented in this measurement group: 1 · Distinct papers with human comparisons for this outcome: 1
- Performance time: Sources represented in this measurement group: 1 · Distinct papers with human comparisons for this outcome: 0
- psychophysiological performance: Sources represented in this measurement group: 1 · Distinct papers with human comparisons for this outcome: 0
- RT: Sources represented in this measurement group: 1 · Distinct papers with human comparisons for this outcome: 1
- Strength measures: Sources represented in this measurement group: 1 · Distinct papers with human comparisons for this outcome: 1
- strength resistance: Sources represented in this measurement group: 1 · Distinct papers with human comparisons for this outcome: 0
- TT performance: Sources represented in this measurement group: 1 · Distinct papers with human comparisons for this outcome: 0
Exercise performance
What the research found
cycling time to exhaustion
No clear difference in this measurement was found between the compared groups.
Controlled study in people · Single ingredient · Unspecified population
Between-group comparison
Analysis coverage and paper counts
1 paper · 1 measured outcome
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
RT
The study reported worsening in this measurement.
Controlled study in people · Single ingredient · General population
Between-group comparison
Analysis coverage and paper counts
1 paper · 1 measured outcome
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
Strength measures
No clear difference in this measurement was found between the compared groups.
Controlled study in people · Ingredient combination · General population
Several ingredients were used together. This does not establish the effect of this ingredient alone or synergy.
Between-group comparison
Analysis coverage and paper counts
1 paper · 1 measured outcome
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
anaerobic performance
The study reported improvement in this measurement.
Controlled study in people · Single ingredient · General population
Between-group comparison
Analysis coverage and paper counts
1 paper · 1 measured outcome
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
average power
The study reported improvement in this measurement.
Controlled study in people · Single ingredient · General population
Between-group comparison
Analysis coverage and paper counts
1 paper · 1 measured outcome
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
muscular strength/power
The study reported improvement in this measurement.
Controlled study in people · Single ingredient · General population
Between-group comparison
Analysis coverage and paper counts
1 paper · 1 measured outcome
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
Peak power
The study reported improvement in this measurement.
Controlled study in people · Single ingredient · Unspecified population
Between-group comparison
Analysis coverage and paper counts
1 paper · 1 measured outcome
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
TT performance
A change was reported before and after use. Its cause cannot be isolated.
Controlled study in people · Single ingredient · General population
Within-group comparison
Analysis coverage and paper counts
1 paper · 1 measured outcome
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
Exercise performance
A change was reported, but the comparison could not be confirmed.
Review · Single ingredient · Unspecified population
Analysis coverage and paper counts
1 paper · 1 measured outcome
Review overlap with primary papers is unknown; counts stay separate.
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
psychophysiological performance
A change was reported, but the comparison could not be confirmed.
Review · Single ingredient · Unspecified population
Analysis coverage and paper counts
1 paper · 1 measured outcome
Review overlap with primary papers is unknown; counts stay separate.
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
strength resistance
A change was reported, but the comparison could not be confirmed.
Meta-analysis · Single ingredient · Unspecified population
Analysis coverage and paper counts
1 paper · 1 measured outcome
Review overlap with primary papers is unknown; counts stay separate.
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
eccentric strength of the knee flexors
The study needs further checking before its result can be interpreted.
Controlled study in people · Single ingredient · General population
Between-group comparison
Analysis coverage and paper counts
1 paper · 1 measured outcome
1 papers held from direction counts
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
exercise performance
The study needs further checking before its result can be interpreted.
Review · Single ingredient · General population
The comparison could not be confirmed.
Analysis coverage and paper counts
1 paper · 1 measured outcome
1 papers held from direction counts
Review overlap with primary papers is unknown; counts stay separate.
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
lower body muscular endurance
The study needs further checking before its result can be interpreted.
Controlled study in people · Ingredient combination · General population
Several ingredients were used together. This does not establish the effect of this ingredient alone or synergy.
Between-group comparison
Analysis coverage and paper counts
1 paper · 1 measured outcome
1 papers held from direction counts
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
LPT
The study needs further checking before its result can be interpreted.
Controlled study in people · Ingredient combination · General population
Several ingredients were used together. This does not establish the effect of this ingredient alone or synergy.
Between-group comparison
Analysis coverage and paper counts
1 paper · 1 measured outcome
1 papers held from direction counts
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
LSPT
The study needs further checking before its result can be interpreted.
Controlled study in people · Ingredient combination · General population
Several ingredients were used together. This does not establish the effect of this ingredient alone or synergy.
Between-group comparison
Analysis coverage and paper counts
1 paper · 1 measured outcome
1 papers held from direction counts
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
OTPT
The study needs further checking before its result can be interpreted.
Controlled study in people · Ingredient combination · General population
Several ingredients were used together. This does not establish the effect of this ingredient alone or synergy.
Between-group comparison
Analysis coverage and paper counts
1 paper · 1 measured outcome
1 papers held from direction counts
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
Performance time
The study needs further checking before its result can be interpreted.
Controlled study in people · Single ingredient · Unspecified population
Between-group comparison
Analysis coverage and paper counts
1 paper · 1 measured outcome
1 papers held from direction counts
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
Unspecified measured outcome
The study needs further checking before its result can be interpreted.
Controlled study in people · Single ingredient · Unspecified population
Between-group comparison
Analysis coverage and paper counts
4 papers · 4 measured outcomes
4 papers held from direction counts
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
Unspecified measured outcome
The study needs further checking before its result can be interpreted.
Review · Single ingredient · Unspecified population
The comparison could not be confirmed.
Analysis coverage and paper counts
1 paper · 1 measured outcome
1 papers held from direction counts
Review overlap with primary papers is unknown; counts stay separate.
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
Blood pressure and vascular health
What the research found
SBP and DBP
A change was reported, but the comparison could not be confirmed.
Meta-analysis · Single ingredient · General population
Analysis coverage and paper counts
1 paper · 1 measured outcome
Review overlap with primary papers is unknown; counts stay separate.
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
Unspecified measured outcome
The study needs further checking before its result can be interpreted.
Controlled study in people · Single ingredient · Unspecified population
Between-group comparison
Analysis coverage and paper counts
1 paper · 1 measured outcome
1 papers held from direction counts
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
Bone, joint, and mobility
What the research found
knee extensor muscle thickness
The study needs further checking before its result can be interpreted.
Controlled study in people · Separable contribution · General population
Between-group comparison
Analysis coverage and paper counts
1 paper · 1 measured outcome
1 papers held from direction counts
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
mobility
The study needs further checking before its result can be interpreted.
Controlled study in people · Ingredient combination · General population
Several ingredients were used together. This does not establish the effect of this ingredient alone or synergy.
The comparison could not be confirmed.
Analysis coverage and paper counts
1 paper · 1 measured outcome
1 papers held from direction counts
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
Fatigue and energy
What the research found
markers of fatigue
A change was reported before and after use. Its cause cannot be isolated.
Controlled study in people · Single ingredient · General population
Within-group comparison
Analysis coverage and paper counts
1 paper · 1 measured outcome
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
Unspecified measured outcome
The study needs further checking before its result can be interpreted.
Controlled study in people · Ingredient combination · Unspecified population
Several ingredients were used together. This does not establish the effect of this ingredient alone or synergy.
Between-group comparison
Analysis coverage and paper counts
1 paper · 1 measured outcome
1 papers held from direction counts
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
Digestion and gut health
What the research found
Unspecified measured outcome
The study needs further checking before its result can be interpreted.
Meta-analysis · Single ingredient · Unspecified population
The comparison could not be confirmed.
Analysis coverage and paper counts
1 paper · 1 measured outcome
1 papers held from direction counts
Review overlap with primary papers is unknown; counts stay separate.
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
Glucose and metabolic health
What the research found
BMI
A change was reported, but the comparison could not be confirmed.
Meta-analysis · Single ingredient · General population
Analysis coverage and paper counts
1 paper · 1 measured outcome
Review overlap with primary papers is unknown; counts stay separate.
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
fat mass
A change was reported, but the comparison could not be confirmed.
Meta-analysis · Single ingredient · General population
Analysis coverage and paper counts
1 paper · 1 measured outcome
Review overlap with primary papers is unknown; counts stay separate.
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
weight
A change was reported, but the comparison could not be confirmed.
Meta-analysis · Single ingredient · General population
Analysis coverage and paper counts
1 paper · 1 measured outcome
Review overlap with primary papers is unknown; counts stay separate.
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
Men's health
What the research found
Unspecified measured outcome
The study needs further checking before its result can be interpreted.
Controlled study in people · Single ingredient · Unspecified population
Between-group comparison
Analysis coverage and paper counts
1 paper · 1 measured outcome
1 papers held from direction counts
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
Skin, hair, and nails
What the research found
hair growth
The study needs further checking before its result can be interpreted.
Review · Single ingredient · Patient group
The comparison could not be confirmed.
Analysis coverage and paper counts
1 paper · 1 measured outcome
1 papers held from direction counts
Review overlap with primary papers is unknown; counts stay separate.
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
Sleep
What the research found
Unspecified measured outcome
The study needs further checking before its result can be interpreted.
Controlled study in people · Single ingredient · Unspecified population
Between-group comparison
Analysis coverage and paper counts
1 paper · 1 measured outcome
1 papers held from direction counts
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
Stress and mood
What the research found
cortisol levels
A change was reported before and after use. Its cause cannot be isolated.
Controlled study in people · Single ingredient · General population
Within-group comparison
Analysis coverage and paper counts
1 paper · 1 measured outcome
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
Women's health
What the research found
low birth weight
A change was reported, but the comparison could not be confirmed.
Meta-analysis · Unknown intervention · General population
Analysis coverage and paper counts
1 paper · 1 measured outcome
Review overlap with primary papers is unknown; counts stay separate.
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
preterm delivery
A change was reported, but the comparison could not be confirmed.
Meta-analysis · Unknown intervention · General population
Analysis coverage and paper counts
1 paper · 1 measured outcome
Review overlap with primary papers is unknown; counts stay separate.
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
SGA
A change was reported, but the comparison could not be confirmed.
Meta-analysis · Unknown intervention · General population
Analysis coverage and paper counts
1 paper · 1 measured outcome
Review overlap with primary papers is unknown; counts stay separate.
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
spontaneous abortion
A change was reported, but the comparison could not be confirmed.
Meta-analysis · Unknown intervention · General population
Analysis coverage and paper counts
1 paper · 1 measured outcome
Review overlap with primary papers is unknown; counts stay separate.
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
stillbirth
A change was reported, but the comparison could not be confirmed.
Meta-analysis · Unknown intervention · General population
Analysis coverage and paper counts
1 paper · 1 measured outcome
Review overlap with primary papers is unknown; counts stay separate.
Direction counts are distinct papers within each row. Outcome contributions count endpoint results, so one paper may contribute more than once.
How to read research findings
How should I read research findings?
Research guidesNo clear difference and an uninterpretable result are different
A reported null comparison contains a result. An abstract without an interpretable comparison leaves a question unresolved.
Paper counts and study counts are different
Count original research separately from research that reanalyzes it. Publication volume is neither replication nor a quality rating.
Side effects and combination findings
Findings from this ingredient alone are shown separately from those involving another supplement or medication.
Caution findings
1
Findings from individual and combined use, plus reviewed cautions
Key precautions
No reviewed caution rules are available yet. Reported caution findings are shown below.
Review 3 combination findings and 0 findings that outperformed an ingredient alone.
These findings were reported in studies. They do not predict an individual's experience.
Findings to review with care
Side effects reported for the ingredient alone are separated from findings involving another supplement or medication.
Side effects reported with this ingredient alone
Symptoms or adverse events reported in studies without another active ingredient.
QRS tachycardia1 papers
A 32‑year‑old woman experienced severe caffeine toxicity after ingesting 5000 mg, leading to presyncope, QRS tachycardia, metabolic acidosis, and hypokalemia, requiring ED treatment.Human studies · Observational studyReported improvements with combinations
Combination findings are grouped by how the study compared them. This is not a recommendation to combine ingredients.
Combinations with reported improvement
The combination showed improvement, but each ingredient's contribution is unclear.
비트루트 추출물In a randomized crossover trial of 20 healthy males, the combination of caffeine and beetroot juice significantly improved 1,000-m running performance compared to placebo (p < 0.01).
L-테아닌L-theanine combined with caffeine significantly reduced caffeine-induced tachycardia (17% vs 92%) and state anxiety compared to caffeine alone.
로디올라A randomized, double-blind, placebo-controlled trial in 48 male volleyball players found that combined Rhodiola rosea and caffeine supplementation significantly improved jump height, reduced power decline during high-frequency jumps, and lowered perceived exertion compared to placebo and single-agent groups.