Caffeine expectancy showed no clear interval running benefit

  • Caffeine
  • Placebo effect
  • Running performance

Thomas Solomon, PhD

Caffeine is a familiar tool for runners, but part of its reputation may come from what athletes expect it to do. This study tested whether that expectation alone could improve a hard interval session.

Reference: Valero et al. Caffeine Expectancy Does Not Affect Interval Running Performance, Physiological Responses, or Running Kinematics in Trained Runners: A Randomized Controlled Crossover Trial. Nutrients (2026) DOI: https://doi.org/10.3390/nu18152492.

Study snapshot

A quick, practical summary for runners and coaches.

Quick answer

This randomized crossover trialA study in which a group of people is randomised to receive BOTH the treatment and the no-treatment control, and the outcome of interest is measured before and after both. The “crossover” means that all participants complete all interventions (the control and the treatment), usually with a washout period in between. tested caffeine expectancy during 2 interval sessions in 14 trained male runners. The caffeine-expectancy condition produced only a small, statistically uncertain improvement in performance. The small sample and imperfect control condition mean smaller effects cannot be ruled out.

Key takeaways

  • Believing that the placebo contained caffeine did not clearly improve interval-running performance.
  • Only 14 male runners took part, and the study could not fully isolate caffeine expectancy from simply taking a capsule.
  • This study gives runners little reason to use fake caffeine as a performance strategy.

How confident should we be?

Evidence confidence: Low

The randomizedRandomization means assigning people to different parts of a study (e.g., groups in a randomised controlled trial) by chance, not by choice. This helps make the groups similar at the start and reduces bias, so any differences you see are more likely due to the treatment, not background differences. In a crossover study, randomization usually decides the order in which each person gets the treatments (for example, Treatment A first then B, or B first then A). This way, order effects—like learning, fatigue, or simple time passing—are less likely to skew the results. crossoverCrossover means that all subjects completed all interventions (control and treatment) usually with a wash-out period in between. design and relevant running test are strengths. But the very small sample, lack of a manipulation check, and imperfect control condition leave considerable uncertainty.

Bottom line

The caffeine-expectancy condition did not produce a convincing performance benefit during this interval workout. There’s no obvious reason to change your training or supplement strategy because of this study, and it tells us very little about whether actual caffeine works.

Read the deep dive below to get a practical interpretation, the wider evidence, some actionable decisions, my thoughts, my rating of perceived scientific enjoyment, and the full study details (research question, study design, participants, methods, results, and the strengths & limitations).

The deep dive

The details behind the headline result, including the practical meaning, full findings, limitations, and my interpretation.

Practical meaning

What does this research mean for runners and coaches?

This study suggests that believing you’ve taken caffeine may change how a workout feels without necessarily changing how fast you run it. That’s useful because athletes often judge supplements partly by sensations such as alertness, energy or perceived side effects. Still, this was 1 small experiment using a specific interval session, so it shouldn’t carry too much weight on its own.

For runners

The runners completed five 1000-m intervals rather than a race, marathon session, trail run or ultra. So the findings are most relevant to trained runners doing hard, repeated intervals.

The practical message is fairly simple: this study provides little reason to expect a caffeine-labelled placebo to make that sort of workout faster.

Importantly, the researchers did not test caffeine itself. Actual caffeine has pharmacological effects as well as possible expectancy effects, so you shouldn’t interpret this study as evidence that caffeine doesn’t improve running performance.

The study also offers a useful reminder about supplement feedback. Feeling more switched on after taking something doesn’t automatically mean the stopwatch has received the memo.

For coaches

The study suggests that an athlete’s expectations may affect some subjective experiences without producing a corresponding performance change.

That matters when evaluating supplements. An athlete saying, “I definitely felt it,” is interesting feedback, but it isn’t the same as showing that performance improved.

The apparent individual differences also need caution. Although 9 runners were faster in the placebo condition, only 2 experimental sessions were completed. Ordinary day-to-day variation could easily masquerade as a personal placebo response.

If caffeine use matters to an athlete, repeated and reasonably standardized trials with actual caffeine will be more informative than trying to identify a “placebo responder” from 1 or 2 workouts.

Evidence in context

What does the wider evidence say?

Placebo effects can influence sports performance, but the size and consistency of those effects vary. A systematic reviewA systematic review answers a specific research question by systematically collating all known experimental evidence, which is collected according to pre-specified eligibility criteria. A systematic review helps inform decisions, guidelines, and policy. of 32 studies involving 1513 participants found small-to-moderate average placebo and nocebo effects on sports performance. (Hurst et al. 2020) Small running studies have also reported faster 1-km, 4-km and 6-minute performances when runners believed they had taken caffeine.

How this study fits: This study challenges the idea that this benefit reliably carries over to repeated interval training. It adds a different exercise setting rather than overturning the earlier evidence. Whether expectancy works may depend on the athlete, the task and whether they’re training or trying to empty the tank in a time trial.

Read the full Veohtu deep dive article

Practical decision

Should runners change anything?

No. This 1 small trial doesn’t justify changing your training or supplement strategy.

Consider this if

  • You tend to interpret feeling more energetic after a supplement as proof that it improved performance.
  • You coach athletes with strong expectations about caffeine or other performance supplements.

Do not overreact if

  • You already use caffeine successfully. The researchers gave the runners no caffeine.
  • Other placebo studies have reported benefits. The effect may depend on the type of exercise and context.

A sensible next step

If caffeine matters to your performance, test actual caffeine during comparable training sessions and look at both performance and unwanted effects, particularly sleep. Treat how you feel as useful information, just not the only information.

TIP: Never make any major changes to your training or lifestyle habits based on the findings of one study, especially if the study is small or provides low-quality evidenceA low quality of evidence means that, in general, studies in this field have several limitations. This could be due to inconsistency in effects between studies, a large range of effect sizes between studies, and/or a high risk of bias (caused by inappropriate controls, a small number of studies, small numbers of participants, poor/absent randomisation processes, missing data, inappropriate methods/statistics). When the quality of evidence is low, there is more doubt and less confidence in the overall effect of an intervention, and future studies could easily change overall conclusions. The best way to improve the quality of evidence is for scientists to conduct large, well-controlled, high-quality randomised controlled trials.. Check whether other trials confirm the findings. If there is a meta-analysisA meta-analysis quantifies the overall effect size of a treatment by compiling effect sizes from all known studies of that treatment. on the topic, look at the effect sizeA standardised measure of the magnitude of an effect of an intervention. Unlike p-values, effect sizes show the size of the effect and how meaningful it might be. Common effect size measures include standardised mean difference (SMD), Cohen’s d, Hedges’ g, eta-squared, and correlation coefficients., the variability between studies, and the quality of evidenceCertainty of evidence tells us how confident we are that the published results accurately reflect the true effect. It’s based on factors like study design, risk of bias, consistency, directness, precision, and publication bias. High certainty means that the current evidence is so strong and consistent that future studies are unlikely to change conclusions. Whereas, low certainty means more doubt and less confidence, and that future studies could easily change current conclusions..

Expert interpretation

My thoughts

Running science from Thomas Solomon at Veohtu

I rather like this study. The researchers asked a neat, practical question and tested it during an outdoor interval session that looks more like normal endurance training than another immaculate laboratory time trial.

The result is also useful because very little happened. The placebo condition was about 4 seconds faster across roughly 17 minutes of running, but the difference wasn’t statistically significantEvidence that a result is unlikely to be due to chance under a “no effect” model (or null hypothesis). Statistical significance is often judged by a p-value below 0.05 to flag that “something” is going on, but not how big or important that “something” is. One statistically significant result doesn’t mean proof; replication is needed. And, a statistically significant result doesn’t necessarily indicate clinical significance. and was small. Heart rate, perceived effort and running mechanics told much the same story.

There are several reasons not to declare caffeine expectancy dead. Only 14 men took part. The runners knew when they were in the no-ingestion condition, the researchers didn’t check whether they genuinely believed the caffeine story, and the power calculationA power calculation is a way to figure out how many people or data points you need in a study so you can reliably spot a real effect if it exists. It balances four things: the size of the effect you care about, how much random variation there is, how strict you are about false alarms, and how likely you want to be to detect the effect. In plain terms: it helps you avoid running a study that’s too small to be useful or so big that it wastes time and money. assumed a relatively large effect (but previous evidence shows a small to moderate effect at best). Therefore, the small number of participants might have meant that a smaller real effect could have been missed.

Still, there’s no compelling performance signal here. “Nothing to see here; move along, move along.” At least for five 1000-m reps.

I’d like to see a larger study with a better placebo control, a proper belief check, and several types of running sessions. Does expectancy become more powerful when there’s actually something worth beating besides Tuesday’s interval splits?

My Rating of Perceived scientific Enjoyment

owlRPsE: 7/10

I experienced moderate scientific enjoyment because the randomized crossover design, detailed methods and very runner-relevant outcome make this a useful experiment, but the tiny sample, retrospective registration and incomplete isolation of caffeine expectancy take some shine off the nerdiness.

down arrow

Read on for further details about the methods, results, data integrity, strengths, and limitations.

Research question

What did the researchers ask?

The authors aimed to test whether believing that you had taken caffeine could improve performance during a hard interval-running session.

Previous running studies had reported performance benefits when athletes thought they had consumed caffeine. Those experiments mainly involved single time trials. The researchers wanted to find out whether the effect also appeared during repeated intervals, which are a common part of endurance training.

Study design

What type of study was this?

This study was a randomized, counterbalanced crossover trial.

Each runner completed both experimental conditions, so each participant effectively served as his own comparison. That reduces the influence of differences in fitness, training history and running ability between participants.

The study can therefore test whether the experimental conditions caused different responses under these specific circumstances. It cannot cleanly show that caffeine expectancy alone caused any difference because the conditions differed in more than belief: the runners swallowed a capsule in 1 condition and took nothing in the other.

Participants

Who took part?

The study included 14 male runners with an average age of about 26 years.

They had an average of 9.6 years of running experience, trained about 59 km (37 miles) per week and had an average 5-km personal best of 17:00. The runners held athletics federation licences and regularly competed.

The participants were low to mild habitual caffeine consumers. They had no physical limitations, musculoskeletal injuries, or respiratory or cardiovascular diseases that had prevented sports participation during the previous 6 months.

No female runners volunteered for the study.

Methods

What did the researchers do?

Who? 14 trained male runners
What? Placebo described as caffeine vs no ingestion
How long? 2 experimental sessions, 7 days apart

Each runner completed 3 outdoor track sessions. The first was a familiarisation session. The 2 experimental sessions came afterwards in randomized, counterbalanced order, with 7 days between sessions.

In the caffeine-expectancy condition, the runners swallowed a capsule containing 100 milligrams of corn flour. The researchers told them that it contained caffeine at a dose of 3 milligrams per kilogram of body weight and reminded them about caffeine’s potential performance benefits.

In the control condition, the runners took nothing.

After the 60-minute pre-exercise period, the runners completed a standardized warm-up followed by five 1000-m intervals with 2 minutes of recovery. They were asked to run at the maximum sustainable intensity for the whole session.

The researchers measured the time for each 1000-m repetition and each 200-m split. They also recorded heart rate, perceived effort, step frequency, ground contact time and vertical oscillation.

The runners received no information about their split times or heart rate during the tests and wore the same running shoes between sessions. They were instructed to avoid caffeine and high-intensity exercise for 48 hours beforehand and to keep their sleep, diet and fluid intake similar.

The morning after each experimental session, they completed a questionnaire about possible caffeine-related side effects.

Data integrity check

Do the numbers add up?

yellow lightMinor integrity concern; closer inspection may be warranted

One number deserves a closer look. The paper reports an average heart rate of 175.8 beats per minute with a 95% confidence intervalA measure of uncertainty used in Frequentist statistics. The 95% confidence interval is a plausible range of values within which the true value (e.g., the true treatment effect) would be found 95% of the time if the data were repeatedly collected in different samples of people. If this range of values (the confidence interval) crosses zero, there is little confidence that the average value is the true effect. If the confidence interval does not cross zero, we can be confident that the average value is the true effect. of 171.2 to 80.3, which cannot be correct as written; the upper value is likely a typo, perhaps 180.3. The other main results look plausible and reasonably consistent. This appears to be a reporting mistake, not evidence of misconduct.

Main findings

What did the study find?

Interval performance Small, statistically uncertain improvement
Heart rate and perceived effort No clear condition difference
Running mechanics No clear condition difference

The caffeine-expectancy condition did not produce a statistically significant improvement in overall interval-running performance.

The five 1000-m repetitions took an average of 1021.5 seconds in the control condition and 1017.2 seconds in the placebo condition. In more familiar terms, that’s about 17 minutes 1.5 seconds versus 16 minutes 57.2 seconds.

The runners were therefore about 4.3 seconds, or 0.42%, faster when they had taken the placebo described as caffeine. The standardized difference was small and wasn’t statistically significant. So the study did not provide convincing evidence that the faster average time represented a systematic performance benefit.

Individual responses varied. 9 of the 14 runners recorded faster total times in the placebo condition. 4 were faster in the control condition, while 1 was virtually unchanged. The authors rightly cautioned that this pattern could simply reflect normal day-to-day performance variation rather than genuine “responders” and “non-responders.”

The pacing analysis was slightly more complicated. The statistical models detected some overall interactions between condition and different parts of the session, but follow-up comparisons found no significant difference between conditions at any individual 1000-m interval or 200-m split. Taken together, the authors concluded that the caffeine-expectancy condition did not meaningfully alter pacing.

Heart rate and perceived effort were also similar between conditions. Step frequency, ground contact time and vertical oscillation did not show meaningful condition-related differences either.

The subjective results were more interesting. The runners reported feeling significantly more active in the placebo condition. They also reported more perceived urine production and insomnia, even though the capsule contained no caffeine. These differences ranged from small to moderate in size and were exploratory, so they deserve caution.

The authors concluded that caffeine expectancy did not significantly affect interval-running performance, pacing, physiological responses or running mechanics, although the placebo condition altered some subjective responses.

Study strengths and limitations

What helps my confidence in the findings?

The strengths

  • The randomized crossover design meant that every runner completed both experimental conditions.
  • All 14 runners completed both conditions, so there was no participant dropout from the experimental comparison.
  • The researchers used a familiarisation session and standardized several factors that could influence running performance.
  • The five 1000-m workout took place on an outdoor athletics track and has clear relevance to endurance training.
  • The paper describes the methods and statistical analyses in considerable detail and states that reporting followed CONSORT crossover guidelines.

What limits my confidence in the findings?

The limitations

  • Only 14 runners took part, and all were male. That limits precision and makes it uncertain whether the findings apply similarly to female runners.
  • The power calculation was based on a relatively large expected placebo effect from one earlier study. The experiment may therefore have been poorly equipped to detect a smaller effect.
  • The control condition involved taking nothing. Because the placebo condition involved both swallowing a capsule and being told it contained caffeine, the study cannot completely isolate caffeine expectancy from other effects of taking a capsule.
  • The researchers did not check afterwards whether the runners genuinely believed that they had consumed caffeine. That’s a rather important missing piece in an experiment about belief.
  • The study was registered retrospectively rather than before data collection. Temperature and humidity were also not formally recorded or strictly standardized.

Funding and conflicts

Who funded the study?

The research was funded through Proyecto PLACEBO, grant PID2020-119162GB-I00, from the Spanish Ministry of Science and Innovation.

The authors declared no conflicts of interest, and the paper reports no commercial funding. There is therefore no obvious manufacturer or supplement-company financial interest in the result.

Learn to train smart, run fast, and be strong.
© 2026 Thomas Solomon. All rights reserved.
Icons from Icons8.
Follow @veohtu
Join the club on