Red Light Therapy for Autism: What the Research Actually Shows

TL;DR

• Two published clinical studies using transcranial near-infrared photobiomodulation in children with ASD have found significant reductions in Childhood Autism Rating Scale (CARS) scores, with no moderate or severe adverse effects reported in either study. A 2024 randomised sham-controlled trial found a 7.23-point CARS improvement in the active group versus sham (p=0.011). A 2025 open-label study found a 7-point CARS reduction (p less than 0.0001).
• The most likely mechanisms are directly relevant to documented features of ASD: mitochondrial dysfunction is present in up to 80% of autistic children, and transcranial photobiomodulation addresses this by restoring cytochrome c oxidase function and ATP production in neural cells. Neuroinflammation, also well documented in ASD, is reduced through photobiomodulation's anti-inflammatory mechanisms.
• Both studies used 850nm NIR light at 40Hz pulse frequency delivered to specific brain areas twice weekly. The 40Hz frequency is significant - it aligns with gamma brain oscillation research showing that 40Hz stimulation supports cognitive processing and may reduce neuroinflammation in neural tissue.
• The evidence base is promising but early. Both published studies are small and the primary study was funded by the device manufacturer. Larger independent replication studies are underway. Red light therapy should be approached as a complementary support tool alongside established behavioural therapies and existing medical care, not as a primary or standalone treatment.
• Red light therapy does not treat autism. Autism is a neurodevelopmental condition, not a disease. The research examines whether photobiomodulation can support quality of life by addressing some of the biological features associated with ASD that affect daily functioning, including sleep difficulties, anxiety, sensory sensitivity, and attention.

Autism spectrum disorder (ASD) is a neurodevelopmental condition characterised by differences in social communication, sensory processing, and behaviour. It is a spectrum condition with enormous variability in presentation, strengths, and support needs across individuals. In Australia, approximately one in 40 children is diagnosed with autism, and families actively seek evidence-based complementary approaches to support daily functioning and quality of life alongside established therapies.

Transcranial photobiomodulation, the application of near-infrared light to the scalp to modulate brain function, has attracted serious research attention in the ASD space over the past five years, with two published clinical studies and a growing body of mechanistic evidence supporting its potential as a complementary support tool. This article covers what the research actually shows, the honest limitations of the current evidence base, the biological mechanisms most relevant to ASD, and how to approach photobiomodulation as a complement to existing support.

A note on language: autism is a neurodevelopmental condition, not a disease or disorder to be cured. Many autistic individuals and advocacy communities embrace neurodiversity. This article discusses photobiomodulation as a potential tool to support some of the associated biological challenges that affect daily quality of life for some autistic people, not as a treatment for autism itself.

The Biological Features of ASD Most Relevant to Photobiomodulation

Understanding why photobiomodulation has attracted research interest in ASD requires understanding which biological features of the condition align with its documented mechanisms.

Mitochondrial Dysfunction

Mitochondrial dysfunction is one of the most consistently documented biological features of ASD. Research suggests it is present in up to 80% of autistic children, significantly higher than in the general population. Mitochondria are the energy-producing structures in every cell, and their dysfunction in neural tissue produces a cascade of consequences including reduced neuronal energy availability, increased reactive oxygen species, impaired calcium regulation, neurotransmitter imbalances, and neuroinflammation. These consequences directly affect brain function, sensory processing, and behaviour.

The primary mechanism of photobiomodulation is cytochrome c oxidase activation in mitochondria, restoring ATP production and improving mitochondrial function. In neural tissue where mitochondrial dysfunction is a primary pathological feature, this mechanism is directly and specifically relevant. As the 2025 Frontiers in Child and Adolescent Psychiatry study notes, tPBM has been shown to effectively restore or enhance mitochondrial function, reduce inflammation and oxidative damage, and potentially modulate default mode network connectivity, all features directly relevant to ASD pathology.

Neuroinflammation

Chronic neuroinflammation, specifically prolonged activation of microglia (the brain's immune cells), is well documented in ASD. Reactive microglia produce sustained inflammatory mediators that contribute to loss of synaptic connections, neuronal cell death, and the long-range neuronal underconnectivity documented across multiple brain imaging studies in autistic individuals. Photobiomodulation reduces pro-inflammatory cytokines and supports the shift from pro-inflammatory to anti-inflammatory microglial phenotype through its well-documented anti-inflammatory mechanisms.

Default Mode Network Connectivity

The Default Mode Network (DMN) is a brain network associated with social cognition, self-referential thinking, and communication. Both underconnectivity and overconnectivity of the DMN have been documented in ASD, and DMN dysfunction has been linked to the social communication differences characteristic of the condition. Multiple studies have shown that transcranial photobiomodulation modulates brain connectivity and functional neural oscillations, with the 40Hz pulse frequency used in the ASD studies specifically targeting gamma oscillation patterns associated with cognitive processing and neuroinflammation reduction.

Sleep Difficulties

Sleep problems affect an estimated 50 to 80% of autistic children, significantly higher than the general population. Poor sleep profoundly amplifies sensory sensitivity, anxiety, irritability, and attention difficulties during waking hours. Photobiomodulation's documented effects on circadian rhythm support and sleep quality through melatonin-compatible wavelengths and mitochondrial restoration in neural tissue are directly relevant to this common challenge for autistic individuals and their families.

What the Research Shows

The 2024 Randomised Sham-Controlled Trial: Frontiers in Neurology

The most rigorous published study is a 2024 randomised, double-blind, sham-controlled clinical trial published in Frontiers in Neurology (Fradkin et al., DOI: 10.3389/fneur.2024.1221193). Thirty children aged 2 to 6 with prior ASD diagnoses were randomly assigned to active treatment or sham. Active treatment delivered pulsed near-infrared light at 850nm and 40Hz to selected brain areas (cortical nodes of the Default Mode Network, Broca and Wernicke areas, and the occipital lobe) twice weekly for 8 weeks using an investigational medical device (Cognilum, JelikaLite Corp). CARS-2 (Childhood Autism Rating Scale, 2nd Edition) was used to measure ASD symptom severity before and after treatment.

The difference in CARS-2 score change between the active and sham groups was 7.23 points (95% CI: 2.357 to 12.107, p=0.011). In the active group, 87% of participants achieved the minimum clinically important difference of 4.5 points compared to 35% in the sham group (p=0.004). Mean CARS scores in the active group fell from 43.5 to 33.7, representing a shift from severe to moderate autism on the CARS scale. EEG data showed significant theta wave changes in the active group. No moderate or severe adverse effects were reported. Mild side effects included transient overexcitement in four participants and mild headaches in two, all of which resolved. The FDA awarded the Cognilum device a Breakthrough Device Designation in December 2021 based on the results.

The 2025 Open-Label Study: Frontiers in Child and Adolescent Psychiatry

A follow-up 2025 open-label study published in Frontiers in Child and Adolescent Psychiatry (Fradkin et al., DOI: 10.3389/frcha.2025.1477839) extended the protocol to 23 participants aged 2 to 7 over 10 weeks using the same device and wavelength specifications. The study found a significant 7-point reduction in mean CARS-2 scores (t=10.23, p less than 0.0001), precisely replicating the active group findings from the 2024 RCT. EEG data showed significant increases in gamma power (p=0.047), decreases in delta power, and significant correlations between EEG changes and CARS-2 improvements. Higher gamma power is associated with improved memory and cognitive processing. The reduction in delta power, which is abnormally elevated in waking states in autistic individuals, correlated directly with improvements in CARS-2 scores. No moderate or severe adverse effects were reported.

The Honest Evidence Position

Both studies were conducted by the same research group and both were funded by JelikaLite Corp, the manufacturer of the device used in the research. This conflict of interest is disclosed in both papers and is a standard limitation of early-stage medical device research. The findings are consistent across both studies and the RCT design of the 2024 paper provides meaningful protection against bias, but independent replication by groups without commercial ties to the device manufacturer is the next necessary step before this research can be considered definitively established. Both papers explicitly call for larger, independent studies, which are currently underway at multiple research centres.

The evidence supports photobiomodulation as a promising and safe complementary approach for some autistic individuals, particularly those where mitochondrial dysfunction and neuroinflammation are significant features. It does not yet support it as a primary or standalone intervention, and the published research to date does not allow confident generalisation to all autistic individuals across all presentations.

The 40Hz Frequency: Why It Matters

Both ASD studies used 850nm NIR light pulsed at 40Hz, and this frequency choice is not arbitrary. 40Hz corresponds to gamma brainwave oscillations, which are associated with higher cognitive processing, attention, and working memory. Research on 40Hz stimulation across multiple modalities has documented reductions in neuroinflammation and improvements in cognitive function, with 40Hz gamma entrainment being studied as a potential intervention in Alzheimer's disease and other neurological conditions. The fact that both ASD studies documented EEG changes in gamma and delta oscillations consistent with 40Hz entrainment effects adds mechanistic coherence to the clinical findings.

Sleep, Anxiety, and Sensory Sensitivity: The Quality of Life Applications

Beyond the CARS-2 primary outcomes, the secondary outcome data from both studies is relevant for families considering photobiomodulation as a complement to existing support. Parent-reported improvements in the active group of the 2024 RCT included changes in eye contact, command following, responsiveness, and sleep patterns. These practical day-to-day improvements, while not the primary measured outcomes, were the improvements that led all parents of children experiencing side effects to choose to continue with the trial despite the mild effects.

For autistic individuals and their families, the most practically meaningful photobiomodulation applications may be the sleep and anxiety-adjacent benefits rather than the CARS score changes. Consistent morning sessions may support circadian alignment that improves sleep quality over time. The anti-inflammatory and mitochondrial restoration mechanisms that reduce neuroinflammation may also contribute to reduced sensory sensitivity and anxiety over consistent treatment periods.

How to Approach Red Light Therapy as a Complementary Tool for ASD

• Discuss with your treating team before starting. Any complementary approach for an autistic child should be discussed with the child's paediatrician, developmental specialist, or GP. This ensures it is integrated appropriately into the child's overall support plan and that any existing medications or conditions are considered.
• The research protocol used 850nm NIR at 40Hz, twice weekly. If replicating the research protocol at home, the 850nm wavelength and 40Hz pulse frequency are the specific parameters used in both published studies. The 940nm wavelength used in StreamShop's cap adds additional depth but 850nm is the researched wavelength for ASD applications specifically.
• Sessions should be short, particularly for young children. The clinical protocol started with 2 minutes and gradually increased to a maximum of 6 to 12 minutes. For at-home use with children, starting conservatively at 5 minutes and assessing tolerance before increasing is the appropriate approach.
• Morning sessions are most appropriate for circadian and sleep benefits. Consistent morning timing supports circadian alignment that may benefit the sleep difficulties common in autistic children.
• Do not expect rapid changes. Both studies assessed outcomes over 8 to 10 weeks of twice-weekly sessions. Meaningful change requires consistent sustained use over weeks to months.
• Photobiomodulation complements, not replaces, established support. ABA therapy, speech therapy, occupational therapy, and other evidence-based supports should continue alongside any photobiomodulation protocol. The research subjects in both studies were continuing their existing therapies throughout the trial.
• Individual responses will vary. ASD is a highly heterogeneous condition. The biological features most relevant to photobiomodulation (mitochondrial dysfunction and neuroinflammation) vary in their significance across individuals. Some autistic individuals may respond more meaningfully than others based on their specific biological profile.

StreamShop Devices for Transcranial and Systemic Support

Red Light Therapy Cap With 940nm

StreamShop's red light therapy cap with 940nm delivers red and 940nm near-infrared light across the full scalp in a comfortable wearable format designed for transcranial applications. For ASD support, the cap is the most anatomically aligned device for replicating the transcranial delivery approach used in both published studies, positioning NIR light against the scalp to reach cortical tissue. The 40Hz pulse mode allows the same pulse frequency used in the research protocols to be applied during sessions. The comfortable wearable format allows sessions during quiet activities including watching videos or gentle play, which is relevant for young children who may find staying still challenging. Sessions of 5 to 10 minutes at 40Hz are the most directly research-aligned protocol for home use.

Red Light Therapy Laser Mat With 1064nm

StreamShop's red light therapy laser mat with 1064nm combines LED and VCSEL laser technology across six wavelengths including 850nm, 940nm, and 1064nm at 110 mW/cm² over a 1.8m x 80cm surface. For ASD support, lying on the mat with the posterior head in contact with the 1064nm VCSEL laser delivers the deepest available transcranial penetration from an at-home device. The 40Hz pulse mode is included for the neurological protocol research alignment. Whole-body sessions on the mat simultaneously address systemic inflammation, support sleep quality through circadian-aligned use, and deliver transcranial NIR across the full posterior scalp. For older children and adults wanting the most comprehensive single-session approach, the laser mat addresses the neurological, inflammatory, and sleep mechanisms simultaneously.

SS100 Class IIa Medical Grade Panel

StreamShop's SS100 class IIa medical grade panel delivers 160 mW/cm² or above at 15cm across nine wavelengths including 810nm, 830nm, 850nm, and 940nm through a 30-degree focusing lens. For older children and adults wanting a panel-based approach, the SS100 positioned at scalp distance delivers high-irradiance NIR to the full frontal and lateral brain areas. The 40Hz pulse mode and per-wavelength dimming allow precise protocol control. As a class IIa ARTG-listed medical grade device, it meets the highest regulatory standard for at-home therapeutic devices in Australia. The SS100 is also the most practical option for parents wanting to use the device themselves alongside their child for general wellness, neurological health, and sleep support.

Frequently Asked Questions

Is Red Light Therapy Good for Autism?

Two published clinical studies have found significant reductions in autism symptom severity scores (CARS-2) using transcranial NIR photobiomodulation in children aged 2 to 7. The 2024 RCT found a 7.23-point CARS improvement versus sham (p=0.011) and the 2025 open-label study replicated a 7-point reduction (p less than 0.0001). No moderate or severe adverse effects were reported in either study. The evidence is promising but the studies are small and were funded by the device manufacturer. Independent replication studies are underway. Red light therapy should be approached as a complementary support tool alongside established therapies, not a primary treatment.

Can Red Light Therapy Help Autistic Children With Sleep Problems?

Sleep difficulties affect 50 to 80% of autistic children and are one of the most significant quality of life challenges for autistic individuals and their families. Photobiomodulation's effects on circadian rhythm support, melatonin-compatible wavelengths, and mitochondrial restoration in neural tissue are all relevant to sleep quality. Parent-reported improvements in sleep patterns were among the secondary outcomes noted in the 2024 RCT. Consistent morning sessions may support circadian alignment that benefits sleep quality over time.

What Wavelength and Frequency Is Best for ASD?

Both published clinical studies used 850nm NIR light pulsed at 40Hz delivered transcranially. The 40Hz frequency aligns with gamma brainwave research and is the specific parameter used in the research showing EEG changes and CARS-2 improvements. The 940nm wavelength used in StreamShop's cap adds depth beyond 850nm. For the most direct alignment with the research protocol, 850nm or 940nm at 40Hz for transcranial application is the most evidence-guided approach.

Is Red Light Therapy Safe for Autistic Children?

Both published studies reported no moderate or severe adverse effects across a combined 53 participants aged 2 to 7. Mild side effects in the active group of the 2024 RCT included transient overexcitement in four children and mild headaches in two, all of which resolved without medical intervention. The FDA classified the investigational device used in both studies as a non-significant risk device. Red and near-infrared light is non-ionising and has no documented cumulative harm at therapeutic doses. Standard photobiomodulation precautions apply: avoid directing light at the eyes and follow manufacturer guidelines for session duration.

How Long Does Red Light Therapy Take to Show Results for ASD?

Both studies assessed outcomes after 8 to 10 weeks of twice-weekly sessions. Meaningful changes in CARS-2 scores were documented over this period. Individual responses will vary based on the specific biological features of each person. Consistent sustained use over 8 to 12 weeks is the minimum appropriate timeframe before assessing whether photobiomodulation is producing meaningful changes for a specific individual.

Should Photobiomodulation Replace ABA or Speech Therapy?

No. In both published studies, participants were continuing their existing behavioural and speech therapies throughout the trial. Photobiomodulation is a complementary tool that may support the neurological and biological environment in which established therapies operate, not a replacement for evidence-based behavioural, communication, and developmental support. The research subjects who were receiving the most intensive existing therapy (up to 20 hours per week of ABA and speech therapy) were still showing differential improvements in the active group compared to the sham group, suggesting the photobiomodulation effect is additional to, not substitutive of, established therapies.