Can Believing You Took Caffeine Improve Athletic Performance
Caffeine is one of the most extensively researched performance-enhancing supplements available.
Whether we're talking about endurance running, cycling, strength training, or repeated high-intensity efforts, there's a substantial body of evidence showing that caffeine can improve exercise performance.
We also have a pretty good understanding of how it works.
Caffeine primarily acts as an adenosine receptor antagonist, influencing the central nervous system and helping reduce feelings of fatigue. Other physiological mechanisms may contribute to its performance benefits.
But there's another factor that doesn't get nearly as much attention.
What happens when an athlete believes they've consumed caffeine, even though they haven't?
Imagine taking a pre-workout supplement before a hard training session. You're expecting more energy, better focus, and improved performance.
You start your workout feeling ready to push harder.
But what if that supplement contained absolutely no caffeine?
Would you still perform better simply because you believed it did?
A new 2026 systematic review and meta-analysis published in the Journal of the International Society of Sports Nutrition investigated exactly that question.
The researchers analyzed 17 studies involving 257 participants and found that giving athletes an inactive substance while telling them it was caffeine produced a statistically significant improvement in athletic performance.
The overall effect was classified as small to moderate.
That's an interesting finding, particularly for anyone who regularly uses caffeine or pre-workout supplements.
But before we start claiming caffeine's benefits are all in your head, there are several important details about the research that deserve a closer look.
TLDR
Believing you've consumed caffeine can improve athletic performance. Across 17 studies involving 257 participants, an inactive substance presented as caffeine produced a small-to-moderate performance benefit (Cohen's d = 0.49; 95% CI 0.28–0.69; p < 0.001).
The strongest pooled effect was observed for time to exhaustion. The effect size was large (d = 1.07), although this analysis included only two studies and 17 participants.
Perceived exertion improved. Participants generally reported more favorable ratings of perceived exertion when they believed they'd consumed caffeine (d = 0.52).
Strength-related outcomes also improved. The researchers found significant pooled benefits for repetitions performed (d = 0.48) and total weight lifted (d = 0.73).
Not every performance measurement improved. The analyses did not find statistically significant effects on power output, movement velocity, heart rate, oxygen consumption, or blood lactate.
Expectations and study design mattered. Larger performance effects were associated with studies where researchers were blinded to treatment administration. The researchers also identified differences related to participant sex composition, although those findings require considerable caution.
The overall certainty of evidence was low. Nine studies were rated as having a high risk of bias, and the remaining eight had an unclear risk of bias.
My takeaway: Caffeine has genuine physiological effects, but an athlete's expectations may contribute to the performance benefits they experience. This study provides evidence that belief alone can influence performance, although the exact magnitude of that effect remains uncertain.
What Is the Caffeine Placebo Effect?
Most people are familiar with the term placebo effect, but it's often misunderstood.
A placebo is an inactive treatment that doesn't contain the substance or intervention being investigated.
A placebo effect occurs when psychological and physiological responses to that treatment influence an outcome.
In sports nutrition research, this becomes particularly interesting because athletes often have strong expectations about certain supplements.
Caffeine is a perfect example.
If you've used caffeine before training, you probably associate it with increased alertness, improved focus, and the ability to push harder.
Those expectations develop through personal experience, marketing, scientific information, and even conversations with other athletes.
Eventually, the act of taking caffeine may become associated with being prepared to perform.
But here's the important distinction.
A placebo effect doesn't mean an improvement is imaginary.
The performance improvement can be real, even when the substance responsible for producing the expectation is inactive.
For example, an athlete who believes they've consumed caffeine may feel less fatigued, tolerate greater discomfort, or maintain a higher level of effort.
Those changes can influence measurable performance outcomes.
The question is how consistently this happens and how large the effect might be.
That's what the researchers wanted to investigate.
Why This Meta-Analysis Is Different From Typical Caffeine Research
Most caffeine studies compare actual caffeine with a placebo.
Participants are usually informed that they might receive caffeine or an inactive substance.
Researchers then compare performance between conditions.
That design is useful for evaluating caffeine's effectiveness, but it creates a problem when we're specifically interested in placebo effects.
Participants know there's a chance they didn't receive caffeine.
That uncertainty can influence their expectations.
For this review, the researchers focused on studies using a different approach.
Participants received an inactive substance but were deliberately told it was caffeine.
This is sometimes called deceptive placebo administration.
The researchers then compared performance with a baseline condition.
The goal was to isolate the influence of believing caffeine had been consumed, without the actual pharmacological effects of caffeine.
That distinction is essential.
The review wasn't designed to determine whether caffeine works.
We already have substantial evidence supporting caffeine's ergogenic effects.
Instead, it investigated whether expectations associated with caffeine could independently influence performance.
How the Researchers Conducted the Review
The researchers searched four scientific databases, including MEDLINE, EMBASE, Web of Science, and CINAHL, for studies published through July 2026.
They followed a predefined protocol and used systematic review procedures consistent with PRISMA reporting guidelines.
To qualify for inclusion, studies had to meet several criteria.
Participants needed to be healthy adults aged 18 or older.
They had to receive a placebo that was deceptively described as caffeine.
The studies also needed to include a baseline or no-treatment comparison and measure an exercise performance outcome.
After screening 1,292 records, the researchers identified 17 eligible studies reported across 16 publications.
Who Participated?
Across the 17 studies, 257 participants were recruited.
Most were young or middle-aged adults, and the majority were male.
Of the total participants, 210 were men, representing approximately 82% of the sample.
Twelve studies included only men, one included only women, and four included both sexes.
Twelve studies involved trained participants, while five involved individuals classified as untrained.
This matters because the results may not generalize equally to women, older adults, or athletes competing at different performance levels.
What Types of Exercise Were Tested?
The researchers included studies examining endurance and resistance exercise.
| Exercise Category | Studies |
|---|---|
| Weightlifting | 9 |
| Running | 5 |
| Cycling | 3 |
| Total | 17 |
The exercise protocols varied considerably.
Some studies measured cycling time-trial performance.
Others examined running performance over distances such as 1,000 meters and 4,000 meters.
Resistance-training studies investigated outcomes such as bench press performance, leg-extension repetitions, and total work performed.
This variety allowed the researchers to examine placebo effects across different types of exercise.
However, it also meant they were combining studies that measured very different outcomes.
To account for this, the researchers calculated standardized effect sizes rather than simply comparing raw performance changes.
What Were Participants Given?
The inactive substances included capsules, decaffeinated coffee, saline solutions, and chewing gum.
In most studies, participants were also given information intended to reinforce their belief that they'd consumed caffeine.
For example, researchers might explain that caffeine improves performance, provide scientific information, or verbally encourage participants to expect an improvement.
These details are important because placebo effects don't necessarily occur in isolation.
They can be influenced by the entire context surrounding a treatment.
Finding #1: Believing You Took Caffeine Improved Overall Performance
This was the main finding.
Across all 17 studies, participants performed better when they received an inactive substance described as caffeine compared with their baseline condition.
Cohen's d standardized effect size
95% confidence interval: 0.28–0.69
Statistically significant (p < 0.001). Effect size is standardized, not a percentage improvement.
A Cohen's d of 0.49 is generally interpreted as a small-to-moderate effect.
The 95% confidence interval ranged from 0.28 to 0.69.
The pooled effect was statistically significant (p < 0.001).
The researchers also reported relatively low statistical heterogeneity (I² = 21.03%), suggesting that differences in the observed effects across studies were not especially large by that measure.
A sensitivity analysis found that removing any single study did not eliminate the statistically significant pooled result.
That adds some support to the overall finding.
However, there's something important to understand about effect sizes.
An effect size of 0.49 does not mean performance improved by 49%.
Cohen's d expresses a difference relative to the variability in the measurements.
It allows researchers to combine outcomes measured in different units, such as watts, seconds, kilograms, and repetitions.
We therefore can't convert this result into a universal percentage improvement in running speed, cycling power, or strength.
And because the review compared placebo conditions with baseline performance, we also need to consider the possibility that factors other than expectation contributed to some of the observed improvements.
These could include familiarization with the exercise test, changes in motivation, or other study-related influences.
That becomes especially relevant when we examine the quality of the included studies.
Still, the central finding is interesting.
Across multiple types of exercise, believing caffeine had been consumed was associated with better performance.
Finding #2: The Largest Effect Was Seen in Time to Exhaustion
The researchers also examined individual performance outcomes.
One of the most interesting results involved time to exhaustion.
Time-to-exhaustion tests typically require an athlete to maintain a specified exercise intensity for as long as possible.
Unlike a time trial, where the athlete attempts to complete a fixed distance as quickly as possible, these tests measure how long someone can tolerate a particular workload.
The researchers found a large pooled effect for time to exhaustion.
Large standardized effect
95% CI: 0.38–1.75 | p < 0.01
Based on only 2 studies and 17 participants.
That's a substantial estimated effect.
But I wouldn't put too much confidence in its exact magnitude.
The analysis included only two studies and 17 participants.
With a sample that small, effect estimates can be unstable.
The confidence interval was also wide.
And there's another consideration.
Time-to-exhaustion tests are particularly sensitive to psychological factors.
If an athlete expects to perform better, they may be willing to tolerate discomfort for longer.
That could translate into a meaningful improvement in the test.
But it doesn't necessarily mean they'll experience an equally large improvement in a real-world race.
A longer time to exhaustion isn't directly equivalent to a faster 5K, 10K, or marathon.
My interpretation: The result supports the possibility that caffeine-related expectations influence exercise tolerance. But it would be premature to suggest that simply believing you've consumed caffeine will produce a large endurance benefit in competition.
Finding #3: Perceived Exertion Improved
This may be the most practically interesting finding.
The researchers analyzed ratings of perceived exertion, commonly called RPE.
RPE is a subjective measure of how difficult an exercise task feels.
It reflects more than just what's happening in the muscles.
Perceived exertion involves the interaction of physiological signals, previous experiences, motivation, attention, and expectations.
The review included 11 studies involving 167 participants that measured RPE.
The pooled effect was statistically significant.
Moderate favorable effect
95% CI: 0.26–0.78 | p < 0.01
In other words, when participants believed they'd received caffeine, their perception of exercise difficulty generally improved.
That makes sense from a performance standpoint.
Imagine two runners maintaining the same pace.
One believes they've consumed caffeine and expects to feel more energized.
The other hasn't received that expectation.
Even if their physiological responses are similar, the runner expecting a benefit may perceive the effort differently.
If the pace feels more manageable, they may be more willing to maintain it or increase their effort.
Over time, that difference in perceived exertion could influence performance.
The researchers also found that the number of techniques used to strengthen participants' expectations was associated with the size of the RPE effect.
Studies that used more methods to convince participants they had received caffeine tended to report larger improvements in perceived exertion.
That finding is consistent with the idea that expectations can influence how exercise feels.
But it's important to distinguish a statistical association from proof of a specific mechanism.
The review doesn't establish exactly which neurological or physiological processes produced these changes.
And it doesn't demonstrate that positive expectations can eliminate the physiological limitations of exercise.
It simply suggests that perception is one part of the performance equation.
Finding #4: Repetitions and Total Weight Lifted Improved
The placebo effect wasn't limited to endurance-related outcomes.
The researchers also found statistically significant improvements in certain resistance-training measurements.
| Outcome | Studies | Effect Size (d) | p-value |
|---|---|---|---|
| Number of repetitions | 5 | 0.48 | 0.03 |
| Total weight lifted | 3 | 0.73 | 0.01 |
| Movement velocity | 4 | −0.15 | 0.51 |
| Power output | 5 | 0.05 | 0.77 |
Participants who believed they'd received caffeine generally completed more repetitions and performed a greater total volume of weightlifting.
The effect on total weight lifted was relatively large.
However, there were no significant pooled improvements in movement velocity or power output.
That's an interesting distinction.
It suggests the placebo effect may be more apparent in tasks involving sustained effort or tolerance of fatigue than in some measures of explosive performance.
For example, believing you've consumed caffeine might influence your willingness to push through the final few repetitions of a challenging set.
But that doesn't necessarily mean your muscles can generate greater instantaneous power.
That's one possible interpretation, rather than a mechanism directly established by the review.
I also wouldn't conclude that placebo caffeine improves maximal strength across the board.
The included resistance-training protocols varied, and some outcomes were examined in relatively few participants.
The results are promising, but the evidence isn't equally strong for every type of muscular performance.
Finding #5: Several Physiological Measurements Didn't Change Significantly
This is where the results become particularly interesting.
If athletes performed better simply because they believed they'd consumed caffeine, did their physiological measurements change as well?
For several outcomes, the answer was no, at least not to a statistically significant degree in the pooled analyses.
| Measurement | Effect Size (d) | p-value | Significant? |
|---|---|---|---|
| Heart rate | 0.21 | 0.23 | No |
| Power output | 0.05 | 0.77 | No |
| Movement velocity | −0.15 | 0.51 | No |
| Oxygen consumption | 0.51 | 0.13 | No |
| Blood lactate | 0.23 | 0.40 | No |
This doesn't prove that placebo effects occur without any physiological changes.
The analyses were relatively small, and nonsignificant results don't establish that an effect is absent.
However, the findings do reinforce the importance of psychological and perceptual factors in exercise performance.
An athlete's perception of effort can change even when conventional physiological measurements don't demonstrate a corresponding significant difference.
And that perception can influence behavior.
For endurance athletes, this is particularly relevant.
Performance isn't determined solely by VO₂max, lactate threshold, or running economy.
Those physiological characteristics are extremely important, but an athlete's willingness and ability to sustain discomfort also contribute to performance.
The same applies to resistance training.
Two athletes with similar muscular capabilities may perform differently depending on their expectations, motivation, and perceived fatigue.
That doesn't make physiology irrelevant.
It means performance involves both physiological capacity and the way an athlete responds to the demands of exercise.
Finding #6: Researcher Expectations May Influence Performance
One of the more surprising findings involved the researchers conducting the experiments.
The meta-analysis found that placebo effects tended to be larger when researchers were blinded to the treatment being administered.
The reported meta-regression coefficient was 0.52 (95% CI: 0.01–0.99; p = 0.03).
Why would that matter?
The authors proposed a possible explanation.
If a researcher knows they're giving an athlete an inactive substance, they might unconsciously communicate different expectations.
Their tone of voice, body language, instructions, or interactions could influence how the athlete responds.
When researchers are unaware of the treatment assignment, those differences might be reduced.
That's a plausible explanation, but it wasn't directly tested in this meta-analysis.
The finding does highlight an important challenge in sports nutrition research.
Expectations aren't limited to the athlete.
Researchers, coaches, and other people involved in performance testing may influence the testing environment.
This is one reason well-designed, properly blinded studies are so important.
And it's particularly relevant when investigating supplements with noticeable effects, such as caffeine.
If someone feels more alert, experiences a faster heartbeat, or recognizes other familiar sensations, they may correctly guess which treatment they've received.
That can influence their expectations and potentially affect the results.
Finding #7: The Results May Differ Between Men and Women, but the Evidence Is Weak
The researchers also examined whether the proportion of female participants influenced the observed placebo effect.
Their analysis found that studies with a lower proportion of women tended to report larger effects.
The meta-regression coefficient was −0.93 (95% CI: −1.68 to −0.19; p = 0.01).
At first glance, someone might interpret that as evidence that men respond more strongly to caffeine placebos than women.
I wouldn't make that claim based on these data.
Here's why.
Of the 257 participants, approximately 82% were men.
Twelve of the 17 studies included only male participants.
Only one study exclusively recruited women.
That creates a substantial imbalance.
The meta-regression was also based on study-level characteristics, not a direct randomized comparison of individual men and women under equivalent conditions.
Differences between studies could influence the relationship.
For example, studies recruiting more women may have differed in exercise protocols, training backgrounds, or other methodological details.
The authors acknowledged this limitation and called for greater female representation in future research.
I agree.
We need better-designed studies with adequate numbers of women before drawing conclusions about sex-specific placebo responses.
Does This Mean Caffeine's Performance Benefits Are Mostly Psychological?
No.
And I think this is the most important point to make.
This study demonstrates that believing you've consumed caffeine can influence athletic performance.
It does not establish that caffeine's physiological effects are unimportant.
Actual caffeine has well-documented pharmacological actions.
Its antagonism of adenosine receptors is a major mechanism through which it influences alertness, fatigue, and exercise performance.
Other mechanisms may contribute depending on the dose, exercise protocol, and individual.
The authors noted that their pooled placebo effect was similar in magnitude to effect sizes reported in some previous meta-analyses examining actual caffeine.
That's interesting.
But comparing effect sizes across different reviews doesn't tell us how much of caffeine's overall benefit is caused by expectation.
The studies used different populations, protocols, and research designs.
In this review, only five studies also included actual caffeine administration, and the primary analysis specifically compared deceptive placebo conditions with baseline.
It wasn't designed to directly separate the pharmacological and expectancy contributions of caffeine across all 17 studies.
We therefore can't say that 50%, 70%, or any other specific percentage of caffeine's performance benefit comes from the placebo effect.
And we certainly can't conclude that caffeine itself is unnecessary.
Actual caffeine and expectations surrounding caffeine may both contribute to performance.
The review does not establish how much of caffeine's total effect comes from either factor.
The more reasonable interpretation is that actual caffeine and the expectations surrounding caffeine may both contribute to performance.
Those effects may interact.
How much each contributes likely depends on the individual, exercise task, dose, and testing conditions.
Could Your Pre-Workout Routine Influence Performance Before the Ingredients Even Work?
I think this is one of the most interesting practical implications of the research.
Many athletes have a consistent pre-workout routine.
They take their supplement, fill their water bottle, put on their running shoes, and prepare mentally for the session.
Over time, that routine becomes associated with training.
You know what comes next.
And if you've repeatedly experienced good workouts after taking a particular supplement, you may begin expecting that same result.
Those expectations could influence your motivation, attention, or perception of effort.
This doesn't mean the ingredients in your pre-workout aren't doing anything.
It means the experience of taking the supplement may be part of a broader performance routine.
Consider an endurance athlete preparing for a hard interval session.
They've used caffeine before similar workouts and associate it with being able to maintain pace when the effort becomes uncomfortable.
When they take caffeine again, they may benefit from both the actual physiological effects and the expectation that they're prepared to perform.
The current review supports the possibility that those expectations matter.
But it doesn't directly test commercial pre-workout supplements, habitual pre-workout rituals, or the long-term development of conditioned responses.
Those are reasonable practical implications to explore, rather than outcomes established by this study.
Should Athletes Use Placebos Instead of Caffeine?
I wouldn't recommend that based on this research.
First, the certainty of evidence was low.
Second, most studies relied on deception.
Participants were deliberately told they had received caffeine when they hadn't.
That approach can be useful under appropriately approved research conditions, but it raises ethical concerns in everyday coaching or sports nutrition practice.
Athletes should know what they're consuming.
Third, this review doesn't establish that knowingly taking an inactive supplement produces the same performance benefits.
That's a different research question.
And finally, actual caffeine has a much stronger evidence base supporting its use for certain exercise performance outcomes.
For athletes who tolerate caffeine well, it remains a legitimate evidence-based option.
That said, caffeine isn't appropriate for everyone.
Some people experience sleep disruption, anxiety, gastrointestinal discomfort, or other unwanted effects.
Responses also vary based on habitual intake, individual sensitivity, dose, and timing.
For someone who prefers to avoid caffeine, this review is interesting because it highlights the potential influence of expectations and psychological preparation.
But it doesn't establish a reliable caffeine-free supplement strategy.
I wouldn't recommend buying an inactive product and expecting it to reproduce caffeine's benefits.
What About Caffeine-Free Pre-Workout Supplements?
This study also raises an interesting question about stimulant-free pre-workouts.
Could some of the perceived benefits come from expectations?
Possibly.
But the current research doesn't directly answer that question.
The studies investigated inactive substances deceptively described as caffeine.
They didn't investigate commercially available stimulant-free pre-workout formulas containing ingredients such as citrulline, betaine, nitrates, or carbohydrates.
Those ingredients have different proposed mechanisms and different levels of supporting evidence.
We shouldn't use placebo research to dismiss their potential effects.
At the same time, we shouldn't ignore the possibility that expectations influence how people evaluate supplements.
If someone takes a product expecting more energy, better focus, or improved performance, that expectation may influence their subjective experience.
This is one reason I place so much importance on controlled research.
A supplement should have evidence supporting its ingredients and doses.
Positive customer experiences are useful feedback, but they aren't a substitute for appropriately controlled studies.
The Biggest Limitations of This Meta-Analysis
Although the pooled result was statistically significant, several limitations deserve attention.
1. The Overall Certainty of Evidence Was Low
The researchers used the GRADE framework to assess confidence in the findings.
They rated the overall certainty of evidence as low.
That means future research could meaningfully change the estimated effect.
A statistically significant pooled result doesn't automatically mean the evidence is strong.
2. None of the Included Studies Had a Low Risk of Bias
Nine studies were classified as having a high risk of bias.
The remaining eight were classified as unclear.
That is a substantial limitation.
Problems included inadequate allocation concealment, incomplete blinding, and uncertainty about whether participants believed the treatment information.
These issues can influence performance outcomes, especially when the intervention specifically involves manipulating expectations.
3. The Studies Were Small
Although the meta-analysis included 257 participants, those individuals were distributed across 17 studies.
Some individual studies included fewer than 10 participants.
Small studies can produce unstable estimates, particularly when measuring outcomes that vary substantially between individuals.
The secondary analyses were even smaller.
For example, the time-to-exhaustion analysis included only 17 participants.
4. Most Participants Were Men
Approximately 82% of participants were male.
This limits our confidence in applying the findings equally to women.
The researchers' own analysis suggested that sex composition might influence effect estimates, but the evidence was too imbalanced to establish a reliable sex difference.
5. Exercise Protocols Varied Considerably
The studies included running, cycling, and resistance training.
They also measured different outcomes.
A 4,000-meter running time trial isn't physiologically or psychologically identical to performing leg extensions to failure.
Standardized effect sizes allow those outcomes to be combined statistically, but they don't eliminate all the differences between the tasks.
6. Some Improvements May Reflect Factors Other Than Placebo Effects
The primary comparison was between the placebo-caffeine condition and baseline.
The researchers did not use a separate no-treatment control as the basis for the pooled comparisons.
That matters because changes between baseline and subsequent testing can occur for reasons unrelated to the placebo intervention.
Practice, familiarization, motivation, and changes in testing conditions may contribute.
The deceptive placebo design strengthens the case for an expectancy effect, but it doesn't completely eliminate these alternative explanations.
7. The Findings Don't Establish Real-World Race Benefits
Several studies used laboratory exercise tests.
Those results may not translate directly into competition.
Athletes face different psychological and physiological demands during races, including pacing decisions, environmental conditions, and competitive pressure.
The review doesn't establish how much an athlete's 5K time, marathon performance, or race-day power output would improve from believing they'd consumed caffeine.
8. The Researchers Couldn't Fully Establish Which Mechanisms Were Responsible
The review identified performance improvements and associations with certain study characteristics.
But it didn't directly measure a consistent neurological mechanism across the included studies.
We can't conclude that changes in dopamine, motivation, pain tolerance, or any single physiological pathway explain the findings.
Those remain potential mechanisms requiring further investigation.
9. Publication and Search Limitations Remain
The researchers found no statistical evidence of publication bias in the main analysis.
That's reassuring, but it doesn't prove that publication bias was absent.
The review was also limited to English-language, peer-reviewed studies.
Unpublished or non-English research may have been missed.
What I'd Like to See in Future Research
I think the next step is to move beyond simply demonstrating that expectations can influence performance.
We need to understand when those effects are most meaningful.
I'd particularly like to see larger studies that directly compare actual caffeine, deceptive placebo caffeine, and appropriate control conditions.
That would help clarify the relative contributions of pharmacology and expectation.
I'd also like to see more research involving trained endurance athletes under realistic race conditions.
Laboratory time-to-exhaustion tests are useful, but they don't always reflect competitive performance.
Studies examining repeated sprint ability, running economy, cycling time trials, and pacing behavior could provide a clearer picture.
And we need substantially more research involving women.
Finally, I'd be interested in studies examining whether positive expectations can be developed ethically, without misleading athletes about what they're consuming.
That might have practical applications for coaching, pre-performance routines, and sports psychology.
The Bottom Line
This meta-analysis provides an interesting reminder that athletic performance involves more than physiology alone.
Across 17 studies involving 257 participants, athletes who received an inactive substance described as caffeine demonstrated a statistically significant improvement in overall performance.
The pooled effect was small to moderate.
The researchers also identified favorable effects on perceived exertion, repetitions performed, total weight lifted, and time to exhaustion.
Those findings suggest that believing you've consumed caffeine can influence how hard exercise feels and, in some circumstances, how well you perform.
But the evidence has important limitations.
None of the included studies was rated as having a low risk of bias, most participants were men, and several secondary analyses involved very small samples.
The overall certainty of evidence was low.
Most importantly, this study doesn't demonstrate that caffeine's benefits are entirely psychological.
Caffeine has genuine pharmacological effects supported by a much larger body of research.
What this review suggests is that expectations may contribute to the overall performance response.
And I think that's worth remembering.
When an athlete takes caffeine before training, they're not just consuming a substance that affects the central nervous system.
They're also bringing their previous experiences, expectations, and beliefs into that workout.
Those factors may influence how they respond.
For me, the practical lesson isn't to abandon caffeine or start experimenting with placebos.
It's to recognize that good performance preparation involves both evidence-based nutrition and psychological readiness.
The ingredients matter. So does the athlete taking them.
References
Hurst P, Hanssen J, Marticorena F, Schiphof-Godart L, Souza HLR, Marocolo M, Saunders B. Placebo effects of caffeine on sport performance: a systematic review and meta-analysis. Journal of the International Society of Sports Nutrition. 2026;23(1):2747142.