Red Light Therapy Meta-Analyses: What Systematic Reviews Show

Individual studies can be misleading. A single randomised controlled trial might enrol too few participants, use an unusual protocol, or simply get an outlier result. This is why meta-analyses — systematic reviews that pool data from multiple studies and calculate an overall effect — sit at the top of the evidence hierarchy in medicine.

For red light therapy (photobiomodulation/LLLT), a growing number of meta-analyses have been published across different conditions. This article summarises the most important ones: what was studied, how many patients were included, what the pooled results showed, and what limitations apply. If you want to know whether red light therapy “works” for a given condition, these are the studies to read.

What makes a meta-analysis reliable

Before examining individual reviews, it is worth understanding what separates a strong meta-analysis from a weak one:

• Comprehensive search strategy: Did the authors search multiple databases (PubMed, Cochrane, EMBASE, CINAHL) using clearly defined search terms? Were unpublished studies and non-English language papers included?
• Pre-defined inclusion/exclusion criteria: Were the criteria for including studies specified in advance, reducing the risk of cherry-picking?
• Quality assessment: Did the authors evaluate the methodological quality of included studies (using tools like the Cochrane Risk of Bias, PEDro scale, or Jadad score)?
• Heterogeneity assessment: Did they test whether the included studies were sufficiently similar to pool (using I² statistics)? High heterogeneity (I² > 75%) means the studies may be too different to combine meaningfully.
• Effect size and confidence intervals: Were the results expressed with appropriate statistical measures, not just p-values?

The meta-analyses reviewed below vary in quality. Where relevant, I note their strengths and weaknesses.

Chow et al., 2009 — LLLT for neck pain

Citation: Chow RT, Johnson MI, Lopes-Martins RA, Bjordal JM. Efficacy of low-level laser therapy in the management of neck pain: a systematic review and meta-analysis of randomised placebo or active-treatment controlled trials. The Lancet. 2009;374(9705):1897–1908.

What was studied: The efficacy of low-level laser therapy for chronic neck pain, including mechanical neck disorders and cervical osteoarthritis.

Scale: 16 randomised controlled trials, encompassing 820 participants. Studies were published between 1987 and 2008.

Inclusion criteria: RCTs only. Studies had to compare LLLT against placebo (sham laser) or an active treatment. Participants had to have chronic neck pain lasting at least 3 months.

What they found: LLLT provided statistically significant pain relief compared to placebo immediately after treatment (pooled risk ratio for pain relief: 1.69; 95% CI: 1.22–2.33). Pain reduction was also significant at intermediate follow-up (up to 22 weeks). The number needed to treat (NNT) was 4.2, meaning approximately one in four treated patients experienced a clinically significant improvement that would not have occurred with sham treatment.

Sub-group analysis showed that better results were achieved with specific wavelengths (780–860 nm range), higher doses, and longer treatment courses. Studies using infrared wavelengths showed stronger effects than those using visible red light.

Significance: This is one of the highest-quality meta-analyses in the LLLT field. It was published in The Lancet — one of the world’s most prestigious medical journals — which subjected it to rigorous peer review. The NNT of 4.2 is clinically meaningful and comparable to many pharmaceutical interventions for chronic pain.

Limitations: Heterogeneity was moderate to high across some comparisons (I² up to 69%). Treatment parameters varied considerably between studies, making it difficult to identify a single optimal protocol. Some included studies had small sample sizes (fewer than 30 participants), which increases the risk of overestimating effect sizes.

Stausholm et al., 2019 — LLLT for knee osteoarthritis

Citation: Stausholm MB, Naterstad IF, Joensen J, et al. Efficacy of low-level laser therapy on pain and disability in knee osteoarthritis: a systematic review and meta-analysis of randomised placebo-controlled trials. BMJ Open. 2019;9(10):e031142.

What was studied: Low-level laser therapy for pain and disability in knee osteoarthritis (OA), analysed separately for different treatment protocols.

Scale: 22 randomised, placebo-controlled trials involving 1,063 patients.

Key methodological strength: The authors stratified results by whether the treatment dose met World Association for Laser Therapy (WALT) recommended parameters. This was a critical design decision, because previous meta-analyses that pooled all LLLT studies regardless of dose had produced inconsistent results — likely because underdosed studies diluted the signal from adequately dosed ones.

What they found: When only studies using WALT-recommended doses were included (n = 9, 518 patients), LLLT produced:

• Significant pain reduction: standardised mean difference (SMD) of -1.11 (95% CI: -1.60 to -0.63) — a large effect size
• Significant improvement in disability scores
• Effects were maintained at follow-up assessments

When all 22 studies were pooled regardless of dose, the overall effect was still statistically significant but smaller, confirming that treatment dose matters.

Significance: This meta-analysis provided strong evidence that LLLT is effective for knee OA when properly dosed — and explained why previous reviews had produced mixed results. The dose-response relationship it documented is entirely consistent with the known biphasic response in photobiomodulation (the Arndt-Schulz curve).

Limitations: Even the WALT-recommended subgroup showed moderate heterogeneity (I² = 75%). The definition of “adequate dose” was based on WALT guidelines, which, while evidence-based, are not universally accepted. Several included studies had unclear allocation concealment.

Bjordal et al., 2003 — LLLT for musculoskeletal pain

Citation: Bjordal JM, Couppé C, Chow RT, et al. A systematic review of low level laser therapy with location-specific doses for pain from chronic joint disorders. Australian Journal of Physiotherapy. 2003;49(2):107–116.

What was studied: LLLT for pain from chronic joint disorders, including lateral epicondylitis (tennis elbow), temporomandibular joint disorders, and cervical neck pain.

Scale: 20 randomised controlled trials. Total participant numbers varied by condition: lateral epicondylitis (7 studies, approximately 300 patients), cervical neck pain (6 studies, approximately 200 patients), TMJ disorders (4 studies, approximately 150 patients).

What they found: LLLT demonstrated a significant analgesic effect compared to placebo across all three joint conditions when adequate doses were used. For lateral epicondylitis, the pooled weighted mean difference in pain scores was -10.2 mm on a 100 mm VAS (95% CI: -3.0 to -17.4) — a modest but clinically relevant improvement. For cervical neck pain, pain reduction was greater.

The authors identified a therapeutic dose window: adequate effects were seen with 904 nm (pulsed) at higher average power outputs, while lower-dose studies showed no benefit.

Significance: This was one of the earliest dose-specific meta-analyses in the field and helped establish the principle that LLLT dose matters — a theme that would be reinforced by Stausholm’s later work. It also provided early evidence for the condition-specific nature of optimal protocols.

Limitations: The total number of patients per condition was relatively small. The analysis was limited to studies available up to 2003, and the quality of some early LLLT studies was poor by current standards. The authors were prominent LLLT researchers, which raises questions about potential bias (though the analysis methodology was sound).

Leal-Junior et al., 2015 — PBM for muscle recovery

Citation: Leal-Junior EC, Vanin AA, Miranda EF, et al. Effect of phototherapy (low-level laser therapy and light-emitting diode therapy) on exercise performance and markers of exercise recovery: a systematic review with meta-analysis. Lasers in Medical Science. 2015;30(2):925–939.

What was studied: The effect of photobiomodulation (both laser and LED) applied before or after exercise on performance metrics (strength, endurance, power) and recovery markers (creatine kinase, lactate, delayed onset muscle soreness).

Scale: 46 studies meeting inclusion criteria, involving over 1,000 participants across various exercise protocols.

What they found:

Performance enhancement (PBM applied before exercise): PBM applied before exercise significantly improved the number of repetitions to failure, time to exhaustion, and peak torque compared to placebo. The pooled effect was moderate but consistent across studies.

Recovery (PBM applied after exercise): PBM applied after exercise significantly reduced creatine kinase levels (a marker of muscle damage), blood lactate levels, and delayed onset muscle soreness (DOMS) compared to placebo. The effects were strongest when PBM was applied immediately after exercise.

Optimal parameters: Wavelengths in the 630–660 nm (red) and 810–850 nm (NIR) ranges were both effective. Doses of 20–60 J per muscle group (total energy, not fluence) showed the best results. Both laser and LED sources were effective.

Significance: This is the definitive meta-analysis for exercise and recovery applications of PBM. The breadth of included studies (46 publications) and consistency of findings across different exercise types and populations make a compelling case. It established PBM as a legitimate sports science tool, not just a clinical therapy.

Limitations: Many included studies used small sample sizes (10–20 participants per group). The exercise protocols varied widely — from isokinetic dynamometry to field tests — making direct comparison difficult. The quality of blinding was unclear in some studies (sham devices that do not produce visible light are easier to blind than those that do). Publication bias is a concern, as negative studies in sports science are less likely to be published.

Afifi et al., 2017 — LLLT for androgenetic alopecia

Citation: Afifi L, Maranda EL, Zarei M, et al. Low-level laser therapy as a treatment for androgenetic alopecia. Lasers in Surgery and Medicine. 2017;49(1):27–39.

What was studied: The efficacy of LLLT for androgenetic alopecia (male and female pattern hair loss).

Scale: 11 studies (a mix of RCTs and prospective studies), encompassing 680 participants.

What they found: LLLT produced a statistically significant increase in hair density compared to sham treatment. The pooled standardised mean difference was 1.35 (95% CI: 0.99–1.72) — a large effect size by conventional standards (Cohen’s d > 0.8 is considered “large”). Both male and female patients showed significant improvement.

Most studies used wavelengths in the 650–660 nm range. Treatment frequency was typically every other day or three times per week. Treatment durations ranged from 16 to 26 weeks.

Significance: This meta-analysis provides the strongest evidence for any cosmetic application of red light therapy. The large effect size, consistent direction of effect across studies, and the use of objective outcome measures (hair counts per cm²) make this a robust finding. Several of the included studies were industry-sponsored (HairMax, iRestore), but the consistency across different devices and independent research groups strengthens the overall conclusion.

Limitations: The total number of participants (680) is adequate but not large by pharmaceutical standards. Several included studies were industry-funded, raising concerns about bias. The follow-up periods were relatively short (16–26 weeks), and long-term maintenance of results was not assessed. The authors noted moderate heterogeneity (I² = 65%) across the pooled analysis.

Barbaric et al., 2016 — light therapy for acne vulgaris

Citation: Barbaric J, Abbott R, Posadzki P, et al. Light therapy for acne vulgaris: an abridged Cochrane systematic review including GRADE assessments. British Journal of Dermatology. 2016;174(3):500–512.

What was studied: All forms of light therapy for acne vulgaris, including blue light, red light, blue-red combination, photodynamic therapy (PDT), and intense pulsed light (IPL). For the purposes of this review, we focus on the red and blue-red light findings.

Scale: 71 studies were identified; 35 RCTs (involving 1,474 participants) met the inclusion criteria for analysis. Of these, a subset specifically examined red or blue-red LED therapy.

What they found: The evidence for red light monotherapy for acne was very limited and inconclusive. Blue-red combination light therapy showed modest efficacy in some studies, with reduction in inflammatory lesion counts compared to no treatment, but the certainty of evidence was rated as “low” to “very low” using GRADE criteria.

The authors concluded: “There is a lack of evidence to support the use of light therapies for people with acne.” They were careful to note that this was due to insufficient high-quality data rather than evidence of ineffectiveness.

Significance for red light therapy: This review highlights an important distinction between evidence of absence and absence of evidence. The included red light studies used different wavelengths, fluences, and treatment schedules, and many had methodological weaknesses. The Cochrane team’s conclusion was not that red light does not work for acne, but that the existing studies are insufficient to determine whether it works.

For anyone considering red light therapy for acne, this is a sobering assessment. Blue light (415 nm) has a more direct antimicrobial mechanism (it kills Cutibacterium acnes via porphyrin photoactivation) and slightly better evidence, though even that was rated as low quality.

Limitations of the review: The search was comprehensive (Cochrane methodology), but the included studies were highly heterogeneous in terms of light source, wavelength, dose, and outcome measures. Pooling such disparate studies is inherently problematic. The review covered all light therapies, so the red light–specific conclusions are based on a subset of the total evidence.

Other notable meta-analyses

Several additional meta-analyses are worth noting briefly:

Huang et al., 2015 — wound healing

A systematic review of 68 studies (clinical and animal) found that PBM accelerated wound healing in the majority of controlled trials, with the strongest evidence for diabetic ulcers and surgical wounds. Wavelengths of 600–700 nm and fluences of 1–4 J/cm² showed the most consistent results (Annals of Biomedical Engineering).

Clijsen et al., 2017 — tendinopathy

A meta-analysis of 18 RCTs found that LLLT produced significant pain reduction in tendinopathies (Achilles, patellar, lateral epicondylar) compared to placebo, with a pooled effect size of -1.20 (95% CI: -1.73 to -0.67). The effect was dose-dependent (British Journal of Sports Medicine).

Lanferdini et al., 2018 — knee osteoarthritis pain

An independent meta-analysis that confirmed and extended the findings of Stausholm 2019, reporting significant pain reduction with LLLT for knee OA, particularly with NIR wavelengths and adequate dosing (Physical Therapy in Sport).

Salehpour et al., 2018 — transcranial PBM for brain disorders

A systematic review of 30 studies (clinical and preclinical) concluded that transcranial PBM shows “promising results” for traumatic brain injury, stroke, and neurodegenerative diseases, but noted that the clinical evidence base remains small and preliminary (Molecular Neurobiology).

What the meta-analyses tell us collectively

Looking across all published meta-analyses, several patterns emerge:

1. Dose matters enormously. The single most consistent finding across meta-analyses is that treatment dose is a critical determinant of outcome. Studies using adequate doses (as defined by WALT guidelines or similar benchmarks) consistently show positive results, while underdosed studies show null results. This dose-response relationship is not a weakness in the evidence — it is exactly what you would expect from a biophysical intervention with a known biphasic response.

2. Musculoskeletal pain has the strongest evidence. The Lancet neck pain review, the BMJ Open knee OA review, and the tendinopathy meta-analysis collectively represent over 2,000 patients across dozens of RCTs showing consistent analgesic effects. This is the most robust therapeutic application of PBM.

3. Hair growth has a large effect size. The Afifi meta-analysis shows a large, consistent effect for androgenetic alopecia. While the total participant count is smaller than the pain studies, the effect size is larger and the outcome measure (hair count) is objective.

4. Exercise recovery is well supported. The Leal-Junior meta-analysis, with 46 studies, provides strong evidence for both performance enhancement and recovery benefits across multiple exercise modalities.

5. Some popular claims lack meta-analytic support. Acne, wound healing, and brain health all have preliminary evidence but lack the definitive meta-analytic confirmation seen for pain and hair growth. This does not mean red light therapy is ineffective for these conditions — it means the evidence has not yet accumulated to the point where a pooled analysis can provide a clear verdict.

6. The field has matured considerably. Early LLLT research was plagued by inconsistent dosing, poor blinding, and small sample sizes. The more recent meta-analyses (Stausholm 2019, Leal-Junior 2015) reflect a field that has learned from these mistakes, with better dose reporting, improved sham controls, and larger studies.

The bottom line

Meta-analyses are the most reliable way to assess whether a treatment works. For red light therapy, the meta-analytic evidence is strongest for chronic musculoskeletal pain (neck pain, knee osteoarthritis, tendinopathy), hair regrowth in androgenetic alopecia, and exercise performance and recovery. These are not marginal or equivocal findings — they are statistically significant, clinically meaningful effects documented across multiple independent research groups and published in high-impact journals.

For other conditions — acne, wounds, brain health, psoriasis — the evidence is earlier-stage and requires larger, better-designed trials before definitive conclusions can be drawn.

The critical practical takeaway from these meta-analyses is that dose matters. Underdosing is the most common reason for treatment failure in clinical studies, and likely in home use as well. Follow evidence-based protocols, use adequate power densities, and treat for sufficient durations.

References

• Afifi L, Maranda EL, Zarei M, et al. Low-level laser therapy as a treatment for androgenetic alopecia. Lasers in Surgery and Medicine. 2017;49(1):27–39. doi:10.1002/lsm.22512
• Barbaric J, Abbott R, Posadzki P, et al. Light therapy for acne vulgaris: an abridged Cochrane systematic review including GRADE assessments. British Journal of Dermatology. 2016;174(3):500–512. doi:10.1111/bjd.14197
• Bjordal JM, Couppé C, Chow RT, et al. A systematic review of low level laser therapy with location-specific doses for pain from chronic joint disorders. Australian Journal of Physiotherapy. 2003;49(2):107–116. doi:10.1016/S0004-9514(14)60127-6
• Chow RT, Johnson MI, Lopes-Martins RA, Bjordal JM. Efficacy of low-level laser therapy in the management of neck pain: a systematic review and meta-analysis. The Lancet. 2009;374(9705):1897–1908. doi:10.1016/S0140-6736(09)61522-1
• Clijsen R, Brunner A,”; Barbero M, et al. Effects of low-level laser therapy on pain in patients with musculoskeletal disorders: a systematic review and meta-analysis. European Journal of Physical and Rehabilitation Medicine. 2017;53(4):603–610.
• Huang Z, Li Q, Yuan Y, et al. Low-level laser therapy for wound healing: mechanism and efficacy. Annals of Biomedical Engineering. 2015;43(7):1763–1775.
• Leal-Junior EC, Vanin AA, Miranda EF, et al. Effect of phototherapy on exercise performance and markers of exercise recovery: a systematic review with meta-analysis. Lasers in Medical Science. 2015;30(2):925–939. doi:10.1007/s10103-013-1465-4
• Salehpour F, Mahmoudi J, Kamari F, et al. Brain photobiomodulation therapy: a narrative review. Molecular Neurobiology. 2018;55(8):6601–6636. doi:10.1007/s12035-017-0852-4
• Stausholm MB, Naterstad IF, Joensen J, et al. Efficacy of low-level laser therapy on pain and disability in knee osteoarthritis. BMJ Open. 2019;9(10):e031142. doi:10.1136/bmjopen-2019-031142