All articles
6 min readrunning / endurance

3 mg per kilo, 60 minutes out: the caffeine dose that works

I race 100 km, and I have taken caffeine before every one of those races on the strength of habit. The habit was close to right on the dose and vague on everything else. Controlled trials put the useful window at 3 to 6 mg per kilo of body mass, taken around 60 minutes before the start, and they settle three questions I had been guessing at: whether a daily coffee habit blunts the effect, whether a gel taken late in a race still does anything, and what an evening dose costs you in sleep.

What the research shows.

The strongest layer of evidence is an umbrella review. Jozo Grgic and colleagues pooled 21 meta-analyses in the British Journal of Sports Medicine and found caffeine ergogenic for aerobic endurance, muscular strength, muscular endurance, power, jumping and exercise speed 1. Few supplements survive that level of aggregation.

The size of the effect is modest. Kirsty Southward's meta-analysis of endurance time trials found gains of roughly 2 to 3% for doses of 3 to 6 mg/kg taken an hour before exercise, in efforts lasting about 10 to 120 minutes 2. On a 3 hour marathon, 2% is close to 4 minutes.

Runners get their own numbers. Ziyu Wang and Juan Del Coso pooled 21 randomized trials covering 254 runners at doses between 3 and 9 mg/kg. Time to exhaustion rose 16.97% on average, with a standard deviation of 14.65% and a medium pooled effect (g = 0.392) 3. The spread is wider than the mean, which is the honest summary of this literature: most runners gain, some gain nothing.

The mechanism you feel is perceptual. Michael Doherty and Paul Smith pooled 21 studies and 109 effect sizes and found caffeine lowered rating of perceived exertion during exercise by 5.6% 4. You hold the same pace at a lower reported effort, so you hold it longer.

On dose, Nanci Guest tested 101 competitive athletes over a 10 km cycling time trial at 0, 2 and 4 mg/kg. Caffeine at 4 mg/kg cut completion time by 3% against placebo 5. The International Society of Sports Nutrition position stand lands on 3 to 6 mg/kg roughly 60 minutes pre-exercise, and notes that very high doses around 9 mg/kg bring side effects without adding performance 6.

Daily coffee drinkers do not need a withdrawal week. Luana Gonçalves ran 40 trained cyclists through time trials after 6 mg/kg, placebo and no supplement, splitting them into low (58 mg/day), moderate (143 mg/day) and high (351 mg/day) habitual consumers. Caffeine beat both control conditions, and an analysis of covariance found no influence of habitual intake on the outcome 7.

Timing has more slack than the 60 minute rule suggests. Gregory Cox had cyclists ride 2 hours of steady state then a time trial. Three protocols worked about equally: 6 mg/kg one hour before (3.4% faster), the same total split into 6 doses of 1 mg/kg through the ride (3.1%), and cola taken only in the last 20 minutes of the steady state and during the trial (3.1%) 8. A dose late in a long race still buys something.

Form changes the clock. Caffeine gum releases through the cheek lining and peaks at 44.2 to 80.4 minutes, against 84 to 120 minutes for capsules 6.

The genotype test is not worth buying yet.

Guest's study also split athletes by the CYP1A2 polymorphism rs762551, which governs how fast the liver clears caffeine. Fast metabolizers (AA) improved by 4.8% at 2 mg/kg and 6.8% at 4 mg/kg. AC carriers gained nothing, and CC carriers rode slower at 4 mg/kg 5. That result launched a consumer testing market. Grgic then reviewed 17 studies on the same polymorphism and found performance gains in AA and in AC/CC groups alike, with no clean split 9. Two protocol trials of your own cost nothing and answer the question for your body.

Sleep is where you pay.

Carissa Gardiner pooled 24 studies on caffeine and subsequent sleep. Total sleep time fell by 45 minutes, sleep efficiency by 7%, time to fall asleep rose 9 minutes, and time awake after falling asleep rose 12 minutes 10. The authors converted that into cutoffs: a 250 ml coffee (107 mg) wants 8.8 hours before bed, and a 217.5 mg pre-workout serving wants 13.2 hours 10. An evening interval session with 200 mg on board is a trade, and on a training day rather than a race day it is usually a bad one.

The protocol.

  1. Weigh the dose. Start at 3 mg per kilo. At 60 kg that is 180 mg, at 75 kg 225 mg. A 250 ml filter coffee carries about 107 mg 10, so most people need two, or one plus a gel.
  2. Take it 60 minutes before the gun, capsules or coffee. If you are using gum, take it 30 minutes out 6.
  3. Test it twice in training before you race it. Same course, same time of day, one session with the dose and one without, both timed. Compare the pace at a fixed heart rate. Two sessions will not prove much on their own, so repeat the pair across a training block.
  4. Keep your normal coffee routine. Skipping caffeine for a week before a race buys nothing measurable 7 and costs you a week of headaches.
  5. Split the dose for anything over 2 hours. Take 3 mg/kg an hour before the start, then 1 mg/kg every 45 to 60 minutes through the middle of the race, capping the total near 6 mg/kg. Cox's cyclists gained as much from the spread version as from the front-loaded one 8.
  6. Count backwards from bedtime. Take your last 200 mg no later than 13 hours before you plan to sleep, and 100 mg no later than 9 hours 10. For a 22:30 bedtime, that means nothing after 09:30 and 13:30 respectively.
  7. Stop at 6 mg/kg. Above that the side effect rate climbs while the performance curve flattens 6.
  8. Log four things after every race you fuel this way: the dose in mg, the clock time you took it, your finish time, and how you slept. Three races in, you will know whether you are one of the runners who gains from this or one of the ones who does not.

If you race this weekend, the decision is small: weigh yourself, multiply by 3, set a timer for 60 minutes before the start, and write down what happens.

Sources.

  1. Grgic, J., Grgic, I., Pickering, C., Schoenfeld, B. J., Bishop, D. J., & Pedisic, Z. (2020). Wake up and smell the coffee: caffeine supplementation and exercise performance, an umbrella review of 21 published meta-analyses. British Journal of Sports Medicine, 54(11), 681-688. doi.org/10.1136/bjsports-2018-100278
  2. Southward, K., Rutherfurd-Markwick, K. J., & Ali, A. (2018). The Effect of Acute Caffeine Ingestion on Endurance Performance: A Systematic Review and Meta-Analysis. Sports Medicine, 48(8), 1913-1928. doi.org/10.1007/s40279-018-0939-8
  3. Wang, Z., Qiu, B., Gao, J., & Del Coso, J. (2022). Effects of Caffeine Intake on Endurance Running Performance and Time to Exhaustion: A Systematic Review and Meta-Analysis. Nutrients, 15(1), 148. doi.org/10.3390/nu15010148
  4. Doherty, M., & Smith, P. M. (2005). Effects of caffeine ingestion on rating of perceived exertion during and after exercise: a meta-analysis. Scandinavian Journal of Medicine & Science in Sports, 15(2), 69-78. doi.org/10.1111/j.1600-0838.2005.00445.x
  5. Guest, N., Corey, P., Vescovi, J., & El-Sohemy, A. (2018). Caffeine, CYP1A2 Genotype, and Endurance Performance in Athletes. Medicine & Science in Sports & Exercise, 50(8), 1570-1578. doi.org/10.1249/MSS.0000000000001596
  6. Guest, N. S., VanDusseldorp, T. A., Nelson, M. T., et al. (2021). International society of sports nutrition position stand: caffeine and exercise performance. Journal of the International Society of Sports Nutrition, 18(1), 1. doi.org/10.1186/s12970-020-00383-4
  7. Gonçalves, L. S., Painelli, V. S., Yamaguchi, G., et al. (2017). Dispelling the myth that habitual caffeine consumption influences the performance response to acute caffeine supplementation. Journal of Applied Physiology, 123(1), 213-220. doi.org/10.1152/japplphysiol.00260.2017
  8. Cox, G. R., Desbrow, B., Montgomery, P. G., et al. (2002). Effect of different protocols of caffeine intake on metabolism and endurance performance. Journal of Applied Physiology, 93(3), 990-999. doi.org/10.1152/japplphysiol.00249.2002
  9. Grgic, J., Pickering, C., Bishop, D. J., Del Coso, J., Schoenfeld, B. J., Tinsley, G. M., & Pedisic, Z. (2021). CYP1A2 genotype and acute ergogenic effects of caffeine intake on exercise performance: a systematic review. European Journal of Nutrition, 60(3), 1181-1195. doi.org/10.1007/s00394-020-02427-6
  10. Gardiner, C., Weakley, J., Burke, L. M., et al. (2023). The effect of caffeine on subsequent sleep: A systematic review and meta-analysis. Sleep Medicine Reviews, 69, 101764. doi.org/10.1016/j.smrv.2023.101764

Keep reading