Clinical library · Modalities
Low-Level Laser and Photobiomodulation
The one passive modality on this site where dose appears to be the whole argument. Given at the doses the reviews specify, laser reduces pain in knee osteoarthritis and lower-limb tendinopathy. Given at other doses, or in carpal tunnel, it does not.
In one line. Red or near-infrared light applied to tissue at intensities too low to heat it, on the proposal that photons absorbed by mitochondrial chromophores alter cell metabolism. Also called photobiomodulation.
It is worth reading carefully rather than dismissing, because the pattern here differs from the other passive modalities: where the dose matches what the reviews recommend, the effects are real and outlast the treatment period.
Dose is the variable. A meta-analysis of 22 trials and 1,063 patients set out to test whether a dose-response relationship exists in knee osteoarthritis, and found one. Overall, pain fell by 14.23 mm on a visual analogue scale against placebo at end of therapy. Restricted to the recommended doses, it fell by 18.71 mm at end of therapy and by 31.87 mm at 2 to 4 weeks afterwards. The authors specify the doses: 4 to 8 J at 785–860 nm, or 1 to 3 J at 904 nm, per treatment spot. [1] A machine set differently is not delivering this evidence.
What the evidence shows
| Question | Finding | Source and quality |
|---|---|---|
| Knee osteoarthritis, all doses | Pain -14.23 mm VAS (95% CI 7.31 to 21.14) at end of therapy; -15.92 mm (6.47 to 25.37) at 1 to 12 weeks after | 22 trials, n = 1,063; risk of bias insignificant. Disability also significantly reduced. No adverse events reported [1] |
| Knee osteoarthritis, recommended doses only | -18.71 mm at end of therapy; -23.23 mm at 2 to 12 weeks; peak -31.87 mm (18.18 to 45.56) at 2 to 4 weeks | The effect grows after treatment stops, which is unusual and worth noting [1] |
| Lower-limb tendinopathy and plantar fasciitis | Pain -13.15 mm VAS at completion; -12.56 mm at 4 to 12 weeks. Disability SMD 0.39 then 0.32 | 18 trials, moderate to high quality. Long-term data were not available [2] |
| As an add-on to exercise therapy | -18.15 mm VAS versus exercise alone at completion; -15.90 mm at 4 to 9 weeks | Recommended doses, in tendinopathy and plantar fasciitis. This is the add-on comparison that most modalities fail [2] |
| Carpal tunnel syndrome | Laser plus splinting beat sham plus splinting for pain, but by -0.53 cm VAS — not clinically significant | Six RCTs, 418 patients. No superiority over splinting alone for symptom severity or function. Authors do not recommend it [4] |
| Across conditions generally | Significant effects for 12 of 35 endpoints; moderate certainty for knee osteoarthritis disability (0.65) and fibromyalgia fatigue (1.25) | 15 meta-analyses, 204 RCTs, over 9,000 participants. Most other outcomes low or very low certainty; publication bias identified in some [3] |
| High-intensity versus low-level laser | No statistical difference in pain, disability, quality of life or postural sway | 12 articles, 704 participants across six conditions [5] |
Where it misleads
1. A positive review does not license any laser machine at any setting
The knee osteoarthritis review exists specifically because the modality is not recommended in major treatment guidelines, and its finding is conditional on dose. [1] Pooling all doses gave 14.23 mm; the recommended doses gave up to 31.87 mm. A clinic using a device outside those parameters cannot cite this evidence for what it is doing.
2. Effect sizes are modest against the measurement error of the scale used
A 13 to 19 mm reduction on a 100 mm visual analogue scale is a real effect and a modest one. Read it alongside the pain scale page: the change that counts is not uniform along the scale, and depends on where the patient started.
3. It fails where it is heavily marketed
In carpal tunnel syndrome, a network meta-analysis found laser plus splinting statistically superior to sham plus splinting for pain but by a magnitude that is not clinically significant, with no superiority over splinting alone for symptom severity or function, and concluded that its use in addition to splinting is not recommended. [4] See carpal tunnel syndrome.
4. The wider literature is much weaker than the two best reviews
The umbrella review covering 204 randomised trials found significant effects for 12 of 35 endpoints, with moderate certainty for only a handful, and states that most outcomes showed low or very low certainty due to heterogeneity or small-study effects, with potential publication bias identified in some. Its conclusion calls for standardisation of protocols before widespread clinical adoption. [3]
5. Paying more for a high-intensity device is not supported
Across 12 studies and 704 participants covering tennis elbow, carpal tunnel, low back pain, knee arthritis, plantar fasciitis and subacromial impingement, there were no statistical differences between high-intensity and low-level laser in pain, electrophysiological parameters, disability, quality of life, postural sway or algometry. [5]
What the evidence supports — and what it does not
Supported
- Knee osteoarthritis at 4 to 8 J (785–860 nm) or 1 to 3 J (904 nm) per spot. [1]
- Lower-limb tendinopathy and plantar fasciitis, short and medium term. [2]
- As an add-on to exercise therapy at recommended doses. [2]
- Safety. No adverse events reported in either meta-analysis. [1][2]
Not supported
- Carpal tunnel syndrome. Not recommended in addition to splinting. [4]
- Any dose the operator prefers. The effect is dose-dependent. [1]
- High-intensity laser as superior. No difference across six conditions. [5]
- Long-term benefit. Long-term data were not available. [2]
- The wider range of claimed indications. Mostly low or very low certainty. [3]
How certain is this?
Evidence grade: Moderate, and conditional on dose.
This is the highest grade any passive modality carries on this site, and it is earned by two well-conducted meta-analyses that pre-specified a dose-response question rather than pooling everything. Risk of bias was judged insignificant in the knee osteoarthritis review, [1] and study quality moderate to high in the tendinopathy review. [2]
Both reviews come from the same research group, which is a limitation worth stating: they share methods and assumptions, so their agreement is less independent than two unrelated reviews agreeing would be.
The umbrella review is the corrective. Across a much wider set of conditions it finds mostly low or very low certainty, heterogeneity, small-study effects and possible publication bias, and calls for protocol standardisation before widespread adoption. [3]
What would change the grade: independent replication of the dose-response finding, and long-term follow-up, which does not currently exist. [2]
Common questions
What dose should I use?
For knee osteoarthritis, the meta-analysis specifies 4 to 8 J per treatment spot at 785 to 860 nm, or 1 to 3 J per spot at 904 nm. [1] Those parameters are the finding — the pooled all-dose effect was 14.23 mm VAS, while the recommended doses reached 31.87 mm at 2 to 4 weeks after therapy. Check what your device actually delivers per spot rather than assuming.
Does it work for plantar fasciitis?
Yes, on moderate-to-high quality evidence, in the short and medium term: pain fell 13.15 mm at completion and 12.56 mm at 4 to 12 weeks, with disability also reduced. Notably, at recommended doses it also outperformed exercise therapy alone when added to it (-18.15 mm). Long-term data were not available. [2] See plantar fasciitis.
Should I use it for carpal tunnel syndrome?
No. A network meta-analysis of six trials and 418 patients found that although laser plus splinting beat sham plus splinting statistically, the magnitude (-0.53 cm VAS) was not clinically significant, and there was no benefit over splinting alone for symptom severity or function. The authors state its use in addition to splinting is not recommended. [4]
Is a high-intensity laser worth the money?
Not on this evidence. Across 12 studies and 704 participants there were no statistical differences between high-intensity and low-level laser in pain, disability, quality of life, postural sway or algometry. Low-level laser was superior for grip strength; high-intensity was favoured on some ultrasound-measured tissue parameters. [5] Neither pattern amounts to a clinical case.
Is it safe?
No adverse events were reported in either the knee osteoarthritis meta-analysis of 22 trials or the tendinopathy meta-analysis of 18. [1][2] Standard eye protection precautions still apply.
References
- Stausholm MB, Naterstad IF, Joensen J, et al. Efficacy of low-level laser therapy on pain and disability in knee osteoarthritis: systematic review and meta-analysis of randomised placebo-controlled trials. BMJ Open. 2019 Oct 28;9(10):e031142. doi:10.1136/bmjopen-2019-031142 PMID 31662383 Systematic review and meta-analysis
- Naterstad IF, Joensen J, Bjordal JM, et al. Efficacy of low-level laser therapy in patients with lower extremity tendinopathy or plantar fasciitis: systematic review and meta-analysis of randomised controlled trials. BMJ Open. 2022 Sep 28;12(9):e059479. doi:10.1136/bmjopen-2021-059479 PMID 36171024 Systematic review and meta-analysis
- Son Y, Lee H, Yu S, et al. Effects of photobiomodulation on multiple health outcomes: an umbrella review of randomized clinical trials. Systematic Reviews. 2025 Aug 6;14(1):160. doi:10.1186/s13643-025-02902-3 PMID 40770824 Umbrella review
- Cheung WKW, Wu IXY, Sit RWS, et al. Low-level laser therapy for carpal tunnel syndrome: systematic review and network meta-analysis. Physiotherapy. 2020 Mar;106:24–35. doi:10.1016/j.physio.2019.06.005 PMID 32026843 Systematic review and network meta-analysis
- Saleh MS, Shahien M, Mortada H, et al. High-intensity versus low-level laser in musculoskeletal disorders. Lasers in Medical Science. 2024 Jul 11;39(1):179. doi:10.1007/s10103-024-04111-1 PMID 38990213 Systematic review
About this resource
Using this in clinic
Every figure here is traceable to its source.
This page states the dose alongside every effect size, because for this modality the dose is what separates the positive trials from the negative ones. Where a value could not be verified against the paper it came from, it is not on this page, and the omission is stated rather than filled with a number from a secondary source.