Supplement
Caffeine
Stimulant · Trimethylxanthine
Caffeine measurably increases endurance performance: in a randomized study with 80 participants, 3 mg per kilogram of body weight improved the 3 km time in all groups. The downside is quantified just as precisely – how strongly and how long caffeine disturbs night-time sleep has been measured in minutes by meta-analyses with polysomnography.
What caffeine is
Caffeine, chemically trimethylxanthine, occurs in coffee, green tea, matcha and guarana and is also added to energy drinks and pre-workout products. In the BK-Score it reaches 9 out of 10 points for human evidence and 10 out of 10 for track record of use.
Both numbers rate the state of knowledge, not the substance. They say that a great deal is documented – on the benefits as well as on the side effects.
How it works
Caffeine blocks adenosine receptors. Adenosine is the messenger that rises over the course of the day and signals tiredness; if its docking site is occupied, the tiredness signal is absent. The mechanism has been quantified in humans and stands at 9 out of 10 points in the BK-Score.
The half-life is about five to six hours. This leads to the core of the problem: after a cup in the afternoon, a considerable part of the dose is still in the body at bedtime.
What is well supported
- Better endurance performance in a randomized study. Khodadadi 2025 studied 80 participants in four groups over eight to nine weeks. 3 mg per kilogram of body weight improved the 3 km time and mean power in all groups (p less than 0.05), more strongly in the trained. Only inactive men were studied.
- The mode of action has been quantified in humans. Caffeine blocks the docking sites of the tiredness messenger adenosine; the half-life is about five to six hours. In the BK-Score the mechanism stands at 9 out of 10 points, and human evidence likewise at 9 out of 10.
- How long caffeine continues to act has been quantified in minutes – twice, independently. Gardiner 2023 (24 studies) and Chang 2025 (22 crossover studies with polysomnography, 956 healthy adults) point in the same direction: total sleep time −45 and −34.67 minutes respectively, sleep onset latency +9 and +8.35 minutes respectively. This yields concrete minimum intervals before bedtime: 8.8 hours for a cup of coffee with 107 mg, 13.2 hours for a pre-workout product with 217.5 mg.
- Safety limits derived by an authority. In 2015, EFSA assessed 200 mg as a single dose and 400 mg per day as safe for healthy adults, 200 mg per day for pregnant and breastfeeding women, and 3 mg per kilogram of body weight per day for children and adolescents. These are figures from an authority, not manufacturer claims.
What the studies show
How much sleep caffeine costs
Gardiner 2023 summarized 24 studies on night-time sleep in a systematic review. Total sleep time fell by 45 minutes, sleep efficiency by 7 %, deep sleep by 11.4 minutes. Sleep onset latency rose by 9 minutes, wake time after sleep onset by 12 minutes. From this, the authors derived minimum intervals before bedtime: 8.8 hours for a cup of coffee with 107 mg per 250 ml and 13.2 hours for a pre-workout product with 217.5 mg.
Confirmation with sleep laboratory measurement
Chang 2025 analyzed 22 controlled crossover studies with polysomnography – instrument-based sleep measurement in the laboratory – and 956 healthy adults, each after an acute single dose. Total sleep time fell by 34.67 minutes, sleep efficiency by 4.74 %, the deep-sleep share by 1.01 %, and sleep onset latency rose by 8.35 minutes. The REM share did not change significantly. A meta-regression found neither a significant linear dose effect nor an age effect.
Habituation lowers the performance effect
Khodadadi 2025 studied 80 inactive men in four groups over eight to nine weeks in a randomized study with 3 and 6 mg per kilogram of body weight respectively. 3 mg/kg improved the 3 km time and mean power in all groups (p less than 0.05), more strongly in the trained. The second finding is decisive: high habitual intake significantly reduced the performance-enhancing effect, both aerobic and anaerobic (p less than 0.05); a short withdrawal partly restored it.
Where the data stop
- That the sleep disturbance fades with long-term consumption. Gardiner 2023 explicitly names this question as open; Chang 2025 studied only acute single doses. The common claim that regular drinkers sleep normally despite caffeine cannot be supported by these data.
- That more caffeine disturbs sleep linearly more. The meta-regression by Chang 2025 found neither a significant linear dose effect nor an age effect. The relationship is therefore less simple than the rules of thumb suggest.
- Transferability of the performance data to women and competitive athletes. Khodadadi 2025 studied only inactive men. This study provides no finding for other groups.
Status, approval and legal
Caffeine can be freely marketed as a food ingredient and additive. In 2015, EFSA assessed the following amounts as safe for healthy adults: 200 mg as a single dose, around 3 mg per kilogram of body weight, and 400 mg over the day. For pregnant and breastfeeding women the authority gives 200 mg per day, for children and adolescents 3 mg per kilogram of body weight per day.
Safety
The EFSA amounts refer to healthy adults and say nothing about individual tolerance. Described are a racing heart and restlessness in sensitive individuals, as well as the development of tolerance with high daily intake. In the BK-Score, safety stands at 8 out of 10 points, with a deduction for the thinner data on high doses and combination products. This text is information and does not replace medical advice.
BK-Score Well supported
| Human evidence | 9 | |
|---|---|---|
| Mechanism | 9 | |
| Safety data | 8 | |
| Hype gap | 7 | |
| Track record of use | 10 |
Adenosine antagonism has been quantified in humans; large cohorts and many RCTs exist on performance, alertness and coffee consumption. Deduction on safety for the data on high doses and combination products.
The score rates the state of knowledge, not the substance. “Safety data 9” means well studied – not harmless. “Track record of use 9” means used long and widely – that is not proof of efficacy.
Subjective assessment by Biohacking Kompakt based on published scoring rules – not a scientific rating and not a medical recommendation. Rules and all ratings (German)
Frequently asked questions about caffeine
How does caffeine work?
Caffeine blocks the docking sites for adenosine. Adenosine rises over the course of the day and signals tiredness; if its receptors are occupied, this signal is absent. The mechanism has been quantified in humans and is considered very well understood. The half-life is about five to six hours, which is why late intake continues to act into the night.
How long before going to bed should you stop taking caffeine?
A 2023 systematic review derives concrete minimum intervals from 24 studies. For a cup of coffee with 107 milligrams per 250 milliliters it gives 8.8 hours before bedtime, for a pre-workout product with 217.5 milligrams as much as 13.2 hours. These intervals are well above the common rules of thumb.
Does caffeine really disturb sleep measurably?
Yes, and in the sleep laboratory too. A meta-analysis of 22 crossover studies with polysomnography in 956 healthy adults found, after acute intake, 34.67 minutes less total sleep, a sleep efficiency 4.74 percent lower, 8.35 minutes longer to fall asleep and a deep-sleep share 1.01 percent smaller. The REM share remained unchanged.
Does the effect wear off with daily consumption?
For performance, this is established. In a randomized study with 80 inactive men, high habitual intake significantly reduced the performance-enhancing effect, both aerobic and anaerobic; a short withdrawal partly restored it. Whether the sleep disturbance also weakens with long-term consumption, by contrast, is unclear.
How much caffeine is considered safe?
In 2015, EFSA assessed 200 milligrams as a single dose and 400 milligrams over the day as safe for healthy adults, which corresponds to about three milligrams per kilogram of body weight. For pregnant and breastfeeding women it gives 200 milligrams daily, for children and adolescents three milligrams per kilogram of body weight per day.
What side effects does caffeine have?
The best documented is the disturbance of night-time sleep, measured as shorter sleep time, lower sleep efficiency and less deep sleep. Also described are a racing heart and restlessness in sensitive individuals, as well as habituation with high daily intake. The data on very high doses and on combination products are thinner. Not medical advice.
The podcast episode (in German)
Episode 63
Coffee & caffeine: the everyday booster fact-checked
The podcast by Paul Höser (Episode 63) · with Paul & Paula. Adenosine mechanics, half-life and the 8-hour deadline, the U-curve with 15–30 % lower mortality at 2–4 cups, morning vs. all-day drinkers, CYP1A2 genetics, cafestol and paper filters, L-theanine fine-tuning. Information only, no dosage or usage recommendation.
Related
- Works together withL-theanine
- Works together withCreatine monohydrate
- Also for energy and focusRhodiola rosea
- Also for energy and focusExogenous ketones / ketone esters
- Also for energy and moodVitamin D3
- Also for focus and moodL-tyrosine
- Often mentioned togetherIron
- Often mentioned togetherSleep & sleep hygiene
- Often mentioned togetherVitamin B12
- Same goal: Focus & concentrationBacopa Monnieri
- Same goal: More energyMaca
- Same goal: More energyTaurine
Sources
- Gardiner 2023, Sleep Medicine Reviews – systematic review and meta-analysis, 24 studies (PMID 36870101)
- Chang 2025, Sleep Medicine – meta-analysis, 22 crossover studies with polysomnography (PMID 41124973)
- Khodadadi 2025, Sports Health – RCT, 80 inactive men, habitual intake (PMID 39905628)
- EFSA 2015 – scientific opinion on the safety of caffeine
Open in the database – with search, filters and comparison (German app)
Information only, not medical advice and not a usage or dosage recommendation. Prescription-only and unapproved substances belong in the hands of a physician. Last updated: 2026-09-13.