This exposure-assessment study uses FDTD simulations to evaluate auto-induced downlink RF-EMF exposure at 3.5 GHz when downlink energy is focused toward user equipment near the head. Exposure varied substantially by device position (ear, eyes, nose) and by the precoding technique used. The authors report that the choice of normalization strategy can produce cases where ICNIRP basic restrictions are exceeded even when reference levels appear compliant, motivating a precautionary framing for compliance assessment.

This animal study examined repeated asymmetrical head exposure to a 5G-modulated 3.5 GHz signal in adult male mice for six weeks. It reports no significant changes in locomotion, anxiety, or object-based memory performance under the tested conditions. However, it found statistically significant but limited cortical gene expression changes (<1% of expressed genes), including enrichment for glutamatergic synapse-related genes and lateralized differences involving mitochondrial genome-encoded genes. The authors caution that potential health risks from these intracortical transcriptomic modifications should not be downplayed and note uncertainties about longer exposures and other populations.