Red Light Therapy for Stroke Recovery: Scientific Evidence

Improvement in thinking, pain, and hand skill followed treatment, according to a 2023 clinical trial in Passos, Brazil, that paired laser applied through the skull with electrical stimulation of the weakened limbs. People with weakness on one side after a stroke were split into a group receiving the laser switched on and a group receiving it switched off, with both groups given the electrical stimulation. A cluster of 12 diode beams at 660, 808 and 980 nm delivered 720 mW and 43.2 J across 1 minute to each of 15 regions of the head, totaling 648 J per session. The trial reported better physical and social wellbeing alongside the thinking and dexterity gains, and a rise in temperature in the treated regions.

Whether light therapy works for conditions of the brain and spinal cord was the question behind a 2021 review from a university in Shanghai, China, which examined 95 papers. It reported that near-infrared wavelengths between 785 and 850 nm produce more favorable results than red wavelengths, and that pulsing the light at 10, 40, or 100 Hz works better than a continuous beam. Doses between 0.1 and 15 J/cm² were effective in animals and between 10 and 84 J/cm² in people, above which the effect reverses and becomes inhibitory. The review concluded that the best settings have not yet been established and that larger studies are needed.

This 2020 clinical trial in São José dos Campos, Brazil, examined stiffness in the upper arm muscle of 27 people, 15 with long-standing weakness on one side after a stroke and 12 healthy volunteers. Each received 10 sessions of 808 nm low-level laser light at 100 mW, 3180 mW/cm², 159.24 J/cm², and 5 J for 50 seconds per session across 28 days, with elbow movement, turning force, and muscle electrical activity measured across 3 phases. Muscle activation and fiber recruitment in the stiff arm muscle increased significantly after treatment. The frequency measure taken from the muscle signal rose in all 3 phases without reaching significance in any of them.

Walking distance rose from 196.3 meters to 238.7 meters at 1 of the 4 doses tested, in a 2020 randomized controlled trial in São Paulo, Brazil. 10 stroke survivors each received sham, 10 J, 30 J, and 50 J per site in a triple-blind crossover with a week between doses, using 630 to 650, 865 to 885, and 904 to 906 nm laser and LED light together with a static magnetic field at 17 sites on each leg. Sessions lasted 76 seconds, 2.5 minutes, 3.8 minutes, and 6.35 minutes, across 28 days. Only the 30 J dose produced a significant improvement, both against sham and against the starting measurement, on the 6-minute walking test. The same dose also cut the time to rise and walk from 18.65 to 15.58 seconds, while 10 J and 50 J did not, and no adverse effects were reported.

After 10 consecutive daily sessions of laser light delivered into a vein, brain scans showed greater blood flow in the affected half of the cerebellum, in a 2017 observational study from Taichung, Taiwan. The patient was a 77-year-old man who had had a stroke in an artery at the front of the brain and had received only fluids and aspirin in the acute stage. Each session used 632.8 nm helium-neon laser light for 60 minutes at 1790 to 2040 and 3500 to 4000 mW/cm² and 6428.57 J/cm². The study described the imbalance in blood flow between the 2 halves of the cerebellum as a poor sign after stroke that no intervention had previously been shown to reverse at this stage.

Researchers at a medical university in Moscow, Russia, treated 75 people with shoulder joint problems following an ischemic stroke in a 2016 clinical trial. 27 received 10 sessions of 600 nm LED therapy to the affected joint lasting 10 minutes each, 28 received the same at 20 minutes per session, and 20 received no light therapy. Pain fell significantly, and the range of movement at the shoulder increased across the treated groups. Anxiety and low mood also decreased, alongside a mild thinning effect on the blood and better circulation in the weakened arm.

Testing each of 15 volunteers through 3 consecutive phases, a 2016 clinical trial in São José dos Campos examined low-level laser light applied to stiff leg muscles after a stroke. Participants averaged 51.5 years of age and were 11 to 48 months past their stroke, and each phase used the same endurance test, with a control phase, a placebo phase, and a real treatment phase using 808 nm light at 100 mW, 3180 mW/cm², and 127.39 J/cm² for 40 seconds across 14 days. Pain scores fell significantly, the time before the muscle tired rose from 25 to 34.5 seconds, and peak turning force increased, all compared with the control phase. The electrical activity of the muscles did not change significantly, and the difference between the placebo and real treatment phases on muscle fatigue was not significant either.

Pooling 2 earlier trials that had reached different conclusions, a 2013 meta-analysis from a university in Albuquerque, New Mexico, United States reassessed laser applied through the skull for acute ischemic stroke in 778 patients. 410 patients who had received 808 nm near-infrared laser light to the scalp for 2 minutes were compared against 368 who had received a sham device, with the groups balanced on their starting characteristics. The combined success rate at 90 days was significantly higher in the laser group than in the sham group. The analysis identified moderate strokes as the group most likely to respond, and said the finding would need confirming in a further trial.

In 2012, an observational study in Easton, Pennsylvania, United States, described a 29-year-old woman who had had a brainstem stroke 2 years earlier and had already spent 21 months in 2 rehabilitation centers. She remained severely dizzy, could not use her left hand, had severe stiffness in her right hand, and could not walk. Treatment combined 660 and 850 nm red and near-infrared light at 1.4 W, 2.95 J/cm², and 2016 J with muscle, bone, joint, and soft tissue recovery techniques, once a week for 60 minutes across 56 days. The study reported positive change in every area of her difficulties by 8 weeks, with those gains held since and treatment continuing every 4 weeks.

A 2010 review from a photomedicine center in Boston, Massachusetts, United States, set out how low-level laser therapy works and where it has been applied in rehabilitation. It covered wavelengths from 632.8 to 1072 nm at 7.5 to 262.5 mW/cm² and 0.9 to 103 J/cm² over 90 to 120 days, applied to the scalp, hand, face, limbs, and spine. The review reported that light of this kind reduced neurological deficits by 32% in animal models of stroke and increased the energy available to brain cells in proportion to the dose delivered. It found no reported adverse effects from laser or light therapy in any of the work it covered.

70 people with shoulder pain and a swollen hand on the paralyzed side after a stroke took part in a 2009 clinical trial in Tuzla, Bosnia and Herzegovina. They were split into 2 groups of 35: 1 receiving 830 nm low-level laser light at 50 mW and 3 J, and the other receiving electrical stimulation and galvanic current, over 42 days, with both groups also given exercise therapy and ice massage. Shoulder pain fell significantly further in the laser group, as did the swelling in the hand, and arm disability scores improved significantly more. Both groups improved significantly in daily independence, with no difference between them on that measure.

Comparing 79 patients given laser against 41 given a sham device, a 2007 randomized controlled trial tested infrared light for ischemic stroke across international centers, led by a medical center in Holon, Israel. 120 patients with moderate to severe stroke, none of whom had received clot-dissolving drugs, had 808 nm light applied to the scalp for 2 minutes in a single session, given a median of 17 to 18 hours after the stroke began. 70% of the laser group met the trial's definition of a successful outcome at 90 days against 51% of the sham group, a statistically significant difference. Deaths and serious adverse events were no more common in the laser group than in the sham group.

2 stroke patients whose persistent hiccups had not responded to medication were treated in a 2006 observational study at a military hospital in Taipei, Taiwan. Both received polarized near-infrared light between 780 and 1400 nm at 3 acupuncture points on the forearm, leg, and back, chosen for their traditional use in settling the stomach and chest. The first patient was treated once, and his hiccups stopped 3 hours later, with no return afterward. The second was treated once a day for 8 consecutive days and improved steadily over that period.

Drawing together the biology of ischemic stroke and the mechanisms of light therapy, a 2026 review from a rehabilitation hospital in Ningbo, China, assessed light applied through the skull at 630 to 1300 nm across 1350 participants. It reported that the light acts by improving the working of the energy-producing parts of cells, increasing blood flow through the brain, and raising the levels of substances that support nerve growth. The review identified 4 factors that determine the dose reaching the brain: wavelength, power density, energy density, and the amount of pigment in the person's skin. It described the response to dose as an inverted U, rising to a peak and falling away above it.

According to a 2025 review from a sports science laboratory in Guangzhou, China, the largest improvements in movement after a stroke came when treatment targeted the brain and the muscles at the same time. The review covered 600 participants and wavelengths of 633 and 694 nm, examining both prevention and rehabilitation. The review named how deep the light can reach as the main practical difficulty, noting that this differs at each stage after a stroke. It called for agreed treatment settings and for protocols matched to the individual patient.

A 2024 review from a hospital in Beijing, China, examined light applied through the skull as a treatment for conditions of the brain and spinal cord, including ischemic stroke. It covered wavelengths between 610 and 830 nm and set out how the light is absorbed inside cells to raise energy production, reduce oxidative damage, and calm inflammation. The review reported benefits in blood flow through the brain, in nerve repair and in behavior across the conditions it covered. It called for more and larger studies before light through the skull is used more widely.

Researchers in Taichung, Taiwan, reviewed the records of 34 people admitted to a post-acute stroke program between July 2018 and June 2021, in a 2022 observational study. All had had a first ischemic stroke and needed help with daily activities on admission, with 12 receiving 632.8 nm helium-neon laser light into a vein for 60 minutes per session across 42 to 84 days alongside conventional rehabilitation and 22 receiving rehabilitation alone. Overall disability scores were significantly better in the group given the light. Daily living, walking distance, and arm function all improved more in that group without reaching significance, while balance improved more in the group given rehabilitation alone.

10 people between 6 months and 5 years after a stroke, aged 45 to 60, each received 4 single doses in a 2022 randomized controlled trial in São Paulo, Brazil. The doses were sham, 10 J, 30 J, and 50 J per site, given a week apart in a triple-blind crossover using a cluster of 12 diodes combining a 905 nm laser with 875 and 640 nm LEDs and a static magnetic field, applied at 17 sites on both legs. Sessions lasted 76 seconds, 3.8 minutes, and 6.35 minutes, across 28 days. None of the doses changed walking speed or the other timing and distance measures of gait. All 3 active doses produced immediate changes in how far the hip bent and straightened during the weight-bearing phase, in both the affected and unaffected legs.

According to a 2020 observational study at a veterans' hospital in Boston, Massachusetts, United States, where the light was placed on the head decided whether it helped. 6 people with language difficulty 2 to 18 years after a stroke on the left side of the brain received 18 treatments over 42 days using 633 and 870 nm red and near-infrared LED light at 500 mW and 22.2 mW/cm² for 9.75, 12.18, 19.5, and 36.55 minutes per session, under 4 different placement patterns. Naming did not improve when the light covered both sides of the head, but improved significantly when it was confined to the damaged left side. Adding placements over 2 midline regions belonging to the brain's resting network alongside the left-side placements produced good responses and significant increases in connectivity within 3 brain networks.

10 of the 40 patients received light alone, 10 received stretching, 10 received both, and 10 received neither, in a 2017 randomized controlled trial in Takasaki, Japan. All had stiffness in the calf muscles following a stroke or similar event, and the light groups received polarized near-infrared light between 600 and 1600 nm at 8 W and 100 Hz for 5 minutes over the point where the calf muscle meets its tendon. Ankle movement increased significantly compared with the control group in the stretching and combined groups but not in the light-only group. The force needed to stretch the muscle fell significantly compared with the control in all 3 treatment groups, and combining light with stretching produced no added effect.

31 patients received laser light into a vein alongside standard treatment while 22 received standard treatment alone, in a 2014 randomized controlled trial of 53 patients with small deep strokes. The treated group had 670 nm light delivered at 2 to 3 mW. Total cholesterol fell to normal levels in the treated group after treatment, and an enzyme that protects cells from oxidative damage returned to normal after treatment. A marker of inflammation stayed above normal in both groups throughout, and the products of fat breakdown in cells remained high, which the trial read as a sign that additional antioxidant treatment is needed.

In 2022, a review from a medical university in Bucharest, Romania, assessed non-drug, non-invasive treatments for acute ischemic stroke, among them light therapy, acupuncture, hyperbaric oxygen, cooling, and magnetic stimulation. 313 articles were narrowed to 54 for full review, covering 838 participants and wavelengths from 600 to 1100 nm at 1 to 30 J/cm². The review found that none of these approaches has entered routine clinical practice, despite some possible benefits. It noted that few of the 15 clinical trials it identified were complete and that most had not reported final results.

Across a trial designed for 1000 patients, a 2014 randomized controlled trial of laser applied through the skull for acute ischemic stroke was stopped early on the recommendation of its monitoring committee. 630 patients had been randomly assigned to receive 780 and 808 nm light applied to the scalp for 2 minutes within 24 hours of the stroke or to a sham device, when a futility analysis of the first 566 completed patients found 49.6% of the laser group and 49.3% of the sham group reaching a good functional outcome. The result held across all 630 patients once they were included. The trial concluded that laser applied through the skull has no measurable protective effect on the brain when applied within 24 hours of a stroke, and the company behind the device was dissolved once its funding was withdrawn.

A scan-based design tested whether laser applied through the skull reduces the volume of dead brain tissue, in a 2013 randomized controlled trial reported from a university in Philadelphia, Pennsylvania, United States. 640 patients from an earlier trial of 780 to 1400 nm laser light had scans on the fifth day after their stroke, measured both by volume and by a standard scoring system for the affected brain regions. Laser treatment had no effect on the total volume of damage, on the overall score, or on the score for the outer brain layers. No effect appeared in any of the groups selected in advance as most likely to show one.

None of the planned analyses reached statistical significance in a 2009 randomized controlled trial of laser applied through the skull for acute ischemic stroke, run from a university in San Diego, California, United States. Researchers randomly assigned 658 patients: 331 received 808 nm laser applied to the scalp for 2 minutes within 24 hours of stroke, and 327 received a sham device. 36.3% of the laser group reached a favorable outcome at 90 days against 30.9% of the sham group, a difference that did not reach significance. A later analysis of the patients with less severe strokes did show a favorable result, and death and serious adverse event rates were similar in both groups.