Laser or LED light given before exercise (photobiomodulation as preconditioning) lengthened time to exhaustion in cycling (SMD 0.35, 95% CI 0.10-0.59) and muscle endurance in single-joint exercise (SMD 0.27, 0.12-0.41), with no significant change in single-joint strength, cycling sprints, any running measure or swimming time trials, in a meta-analysis of 37 double-blind placebo-controlled crossover studies (586 healthy young adults)
Main caveat: Risk of bias, heterogeneity, the number of studies in each mode and the parameters of the light are not in the abstract.
Dutra and colleagues pooled crossover, double-blind, placebo-controlled studies that used photobiomodulation, which they define as non-thermal light from a laser or light-emitting diode source, as a preconditioning method before exercise in healthy young adults.
They searched five databases to May 2022 and included 37 studies with 586 participants and 78 performance measurements, then analysed each exercise mode on its own.
Light improved muscle endurance in single-joint exercise (SMD 0.27) and time to exhaustion in cycling (SMD 0.35), and did nothing measurable for single-joint strength, all-out cycling sprints, running time to exhaustion, time trials or repeated sprints, or swimming time trials.
The authors conclude that the benefit is limited to single-joint muscle endurance and cycling time to exhaustion.
Participants
37 studies, 586 healthy young adults; the sex split, training status and sports are not in the abstract.
Protocol
Search: PubMed, Cochrane Central Register of Controlled Trials, Embase, SPORTDiscus and PEDro, to 15 May 2022.
Inclusion: crossover, double-blind, placebo-controlled studies of photobiomodulation used as preconditioning (before exercise).
What was applied: laser or LED light before exercise; the abstract does not give the device type per study, the body sites, whether the device touched the skin, the wavelengths, doses or how long before exercise.
Outcomes, by mode: single-joint muscle endurance and strength; cycling time to exhaustion and all-out sprint; running time to exhaustion, time trial and repeated sprints; swimming time trial.
Analysis: standardised mean differences with 95% CI. The risk-of-bias assessment, the heterogeneity statistics and any dose or timing analysis are not in the abstract.
Key Results
| Mode and outcome (light vs placebo) | Result |
|---|---|
| Single-joint, muscle endurance | SMD 0.27 (95% CI 0.12-0.41), p < 0.01 |
| Single-joint, muscle strength | not significant, p = 0.92 |
| Cycling, time to exhaustion | SMD 0.35 (95% CI 0.10-0.59), p < 0.01 |
| Cycling, all-out sprint | not significant, p = 0.96 |
| Running, time to exhaustion | not significant, p = 0.10 |
| Running, time trial | not significant, p = 0.61 |
| Running, repeated sprints | not significant, p = 0.37 |
| Swimming, time trial | not significant, p = 0.81 |
Effect sizes for the null outcomes, heterogeneity and the number of studies per mode are not in the abstract.
Practical Application
Light before exercise improved two outcomes, cycling time to exhaustion (SMD 0.35) and single-joint muscle endurance (SMD 0.27); running time to exhaustion (p = 0.10), single-joint strength, sprints and running and swimming time trials were not significantly changed. A time trial is closer to racing than time to exhaustion, and no time-trial result reported in the abstract was significant.
Inference, not tested here: a small gain in a ride to exhaustion in the laboratory is not a shown gain in a race of fixed distance; an athlete asking whether to use a light device before a race can be told the time-trial evidence (running and swimming here) was null and cycling time trials are not reported in the abstract.
Inference, not tested here: the review covers light from a laser or LED device used before exercise in controlled studies; what any device sold for home use delivers, and whether it matches the trials, is not known from the abstract.
Under SYNTH-83: the answer is "tested; small effects on time to exhaustion, nothing on time trials or strength", not "it works" and not "a myth".
Limitations
Risk of bias, heterogeneity, the number of studies in each mode and the parameters of the light are not in the abstract; written from the abstract and the PubMed record, full text not read.
Crossover designs only; 586 participants across 37 studies (about 16 per study by our arithmetic), so most trials are small.
78 measurements from 37 studies: several outcomes per study, so the pooled estimates are not from independent samples unless the full text corrected for it (not stated).
Overlap (our reading; not quantified in any source we read): these trials likely overlap with the other photobiomodulation meta-analyses (REF-1764 Vanin 2018, REF-1765 Luo 2022, REF-1767 Nascimento 2024, REF-1769 Qiu 2026, REF-1774 Zhai 2026). REF-1767 includes a running trial by "Malta, et al.", a surname shared with an author here, and cites an earlier meta-analysis of 12 crossover running trials that may be this one; the reviews are not independent confirmations.
Funding and conflicts of interest are not in the abstract.
How this entry was checked and changed
Why this level
a meta-analysis of placebo-controlled crossover trials, risk of bias not read (abstract only)
Content correction 2026-11-11
2026-11-11: added from the abstract and the PubMed record; level B+ (meta-analysis of crossover trials, risk of bias not read).