A recent GOLIAT research programme conducted by scientists at INERIS found no measurable biological effects from short-term exposure to 5G radiofrequency signals in healthy adults. The studies investigated whether such exposure could influence physiological stress responses or alter the brain’s electrical activity, and reported no detectable changes in either domain. The findings were published across two peer-reviewed papers in Environmental Research and represent the first coordinated human laboratory investigations to assess the acute effects of exposure to the highest-frequency band currently used for 5G, under conditions designed to reflect real environmental levels.
Both studies examined exposure to genuine, operational 5G signals rather than simulated emissions. Participants were exposed to signals comparable to those encountered outdoors, and their responses were compared with sham (placebo) exposure sessions. Across both investigations, the researchers found no differences between real and sham exposure in physiological stress markers or in resting brain activity. These results suggest that brief exposure to environmental levels of 5G does not disrupt the body’s immediate stress regulation or the brain’s natural electrical rhythms.
Methodologically, the studies were designed to a high standard. Each used a triple-blind, randomised protocol in which participants, experimenters, and data analysts were unaware of the exposure condition. All sessions took place in an electromagnetically shielded room to prevent interference from external radiofrequency sources. Volunteers were seated 120 centimetres from a horn antenna emitting a 5G New Radio signal at 26 GHz. Field strengths reached 2 V/m at head level and 1 V/m at the torso, corresponding to the highest values measured in real outdoor environments by the French national frequency authority. Each participant completed two sessions—one with real exposure and one with sham exposure—each lasting around an hour, including 26.5 minutes of exposure.
During the sessions, researchers collected saliva samples to analyse cortisol and alpha-amylase, two widely validated biomarkers of stress and autonomic nervous system activity. At the same time, brain activity was recorded using electroencephalography, allowing the team to monitor the brain’s electrical oscillations across major frequency bands, including delta, theta, alpha, and beta. Analysis showed no differences between real and sham exposure at any time point before, during, or after exposure. Similarly, cortisol and alpha-amylase levels remained stable throughout, indicating no activation of acute stress responses.
The focus on the 26 GHz frequency band reflects its emerging role in 5G networks, where it supports faster data transmission than lower bands such as 3.5 GHz. Lisa Michelant, the first author of both studies, explains that this band has been relatively understudied in humans and differs physically from lower-frequency bands, notably in its limited penetration into the skin. Senior author Brahim Selmaoui adds that the findings are consistent with existing scientific evidence showing no acute biological effects at commonly encountered levels of radiofrequency exposure. While further research is needed to examine long-term or repeated exposures, the authors conclude that their results provide reassuring data and contribute to ongoing safety assessments by health authorities of 5G technologies.
More information: Lisa Michelant et al, Millimeter-wave high frequency 5G (26 GHz) electromagnetic fields do not modulate human brain electrical activity, Environmental Research. DOI: 10.1016/j.envres.2025.123349
Journal information: Environmental Research Provided by Barcelona Institute for Global Health (ISGlobal)
