Of the 412 peer-reviewed studies in this collection on radiofrequency exposure, 298 — 72.3% — reported biological effects. Millimeter waves are represented, but as cell-culture work at 60 GHz rather than at the 24–47 GHz bands high-band 5G deploys on, and not in people — so the distinctive part of 5G is studied at one remove rather than directly. Lower-band 5G uses frequencies already covered by decades of research, where effects are documented but dose matters: one careful study found no DNA damage in human cells until average absorption reached at least 5.0 W/kg.
Based on analysis of 415 peer-reviewed studies
5G technology has generated significant public concern about health effects. The topic has also attracted misinformation, making it difficult for people to understand what scientific research actually shows about 5G safety.
5G operates across different frequency bands—some similar to existing 4G networks, others using higher frequencies (millimeter waves) that are relatively new for widespread consumer exposure. This page focuses on what peer-reviewed research says about radiofrequency radiation at 5G frequencies.
We present the scientific evidence objectively, including both studies that raise concerns and those that find no effects, so you can make informed judgments based on actual research.
5G is not one thing. Low-band and mid-band 5G operate at frequencies overlapping what 4G, WiFi and cordless phones have used for years, and those frequencies carry decades of study. High-band 5G uses millimeter waves, which is genuinely new for consumer deployment. Of the 412 peer-reviewed studies indexed here on radiofrequency exposure, of which 298 (72.3%) reported biological effects, millimeter waves are represented — but the work sits at 60 GHz, above the 24–47 GHz range high-band 5G actually deploys on, and it is done in cell cultures rather than people.
Habauzit and colleagues (2014) exposed human keratinocytes to 60 GHz and found 665 genes differentially expressed. When the temperature rise was held constant, that change disappeared. A heat-shock control did not reproduce the exposure exactly, leaving 34 genes that suggest a small specific electromagnetic effect on top of the heating. Le Quément and colleagues (2014) exposed the same cell type at 20 mW/cm² and found no change in endoplasmic-reticulum stress markers at baseline; where exposure did alter the stress response, the controls tied it to heat.
That is the shape of the millimeter-wave evidence here: mostly thermal, with a small residual effect, in cells. For the part of 5G that resembles existing technology, there is direct evidence. For the part that is new, there is adjacent evidence at the wrong frequency and the wrong scale.
Tice and colleagues (2002) exposed cultured human blood cells to four different cellular signal technologies at 837 and 1909 MHz, including analog, TDMA, CDMA and GSM formats. Exposure for 3 or 24 hours did not produce a significant increase in DNA damage in leukocytes, and 3 hours did not increase micronucleated lymphocytes. Extended exposure at an average specific absorption rate of at least 5.0 W/kg did induce chromosomal damage. The conclusion is specific: these signals can damage chromosomes under extended exposure at that dose.
Lai and Singh (1995) exposed rats for two hours to 2450 MHz and found a dose-rate-dependent increase in DNA single-strand breaks in brain cells at whole-body absorption rates of 0.6 and 1.2 W/kg. Notably, immediately after pulsed exposure no significant effect appeared — the effect emerged with continuous-wave exposure and at four hours after pulsed exposure. Their 1996 follow-up found both single- and double-strand breaks at 1.2 W/kg.
Lai and Singh (1997) tested whether free radicals explain the damage by treating rats with melatonin or the spin-trap compound PBN before and after exposure. Both blocked the DNA strand breaks. Since both are efficient free radical scavengers, the result points at oxidative stress as the pathway rather than heating.
Leszczynski and colleagues (2002) showed a separate non-thermal response: one hour of mobile phone exposure changed the phosphorylation status of numerous proteins in human endothelial cells, including heat shock protein-27, at intensities that did not heat the cells.
The exposures producing these effects — 0.6 to 5.0 W/kg — are far above what a person absorbs from a distant 5G antenna, and in the range of a phone held against the head. Ambient exposure from network infrastructure is typically thousands of times lower than the levels in these experiments. A guide that reports the effects without the doses leaves out the part that determines whether they apply to you.
No study here demonstrates harm from 5G. The millimeter-wave work in this collection is cell-culture research at 60 GHz, not at 5G's own bands and not in people, which means claims that high-band 5G is dangerous and claims that it is proven safe are both running ahead of the evidence. The lower-band research shows real non-thermal biological effects, at doses generally above ambient network exposure and comparable to handset use.
The dominant source for almost everyone is their own handset, not the network. Speaker mode or wired headphones remove most head exposure. Texting over calling reduces transmit time. Keeping the phone out of the bedroom overnight removes hours of close proximity. Using WiFi rather than cellular in weak-signal areas matters because a phone reaching for a distant tower transmits at higher power — which is also why more, closer antennas can lower handset output rather than raise total exposure.
Ikeda N, Hayashida O, Kameda H, Ito H, Matsuda T · 1994
Researchers investigated thermal damage to normal dog brain using 8 MHz radiofrequency interstitial hyperthermia. The heating limits were 42 degrees C for 45 min or 43 degrees C for 15 min. Breakdown of the BBB was observed at 43 degrees C for 60 min.
Akoev IG, Mel'nikov VM, Usachev AV, Kozhokaru AF, · 1994
Researchers studied whether exposure to low-intensity radiofrequency radiation after a lethal dose of gamma-radiation could modify the resulting injury in mice. The radiofrequency exposure lasted up to 23 hours at an intensity of 5 plus or minus 1.5 microwatts per square centimeter, covering frequency ranges of 2-8, 8-18, and 19-27 GHz with a swing frequency of 12-14 Hz. A clearly pronounced modification of acute radiation injury was obtained when the radiofrequency exposure was applied immediately after the gamma-radiation.
Haider T, Knasmueller S, Kundi M, Haider M · 1994
Researchers used the Tradescantia-micronucleus bioassay to test whether short-wave electromagnetic fields used for broadcasting (10–21 MHz) may show genotoxic effects. Plant cuttings bearing young flower buds were exposed for 30 hours at various distances from broadcasting antennas, and micronucleus frequencies were compared to laboratory controls. Higher micronucleus frequencies were found at all exposure sites near the antennas where IRPA electric field strength standards were exceeded, and results at all but one site were statistically significant.
Kunjilwar KK, Behari J · 1993
Researchers exposed developing rats to amplitude-modulated radio frequency radiation at 147 MHz and its sub-harmonics for 30 to 35 days, three hours per day. A significant decrease in acetylcholine esterase activity was found in exposed rats compared to controls. The decrease in enzyme activity was independent of the carrier wave frequencies used. Short-term exposure did not significantly affect acetylcholine esterase activity.
Ellingsrud S, Johnsson A · 1993
Researchers studied the up-and-down rhythmic movements of the lateral leaflets of Desmodium gyrans when exposed to electromagnetic radiation in the radio-frequency range. The RF radiation was applied as homogeneous 27.12 MHz fields in specially-designed exposure cells, and in some cases as non-homogeneous radiation of 27 MHz amplitude modulated by 50 Hz in front of commercial diathermy equipment. The radiation caused temporary changes in amplitude, period, and phase, and could also cause temporary or complete cessations of the rhythms.
Enin LD, Akoev GN, Potekhina IL, Oleiner VD · 1992
The study aimed to investigate the effect of extremely high-frequency electromagnetic radiation on skin sensory endings in albino rats. Researchers recorded impulse activity from solitary fibers of the tibial nerve and found that receptor units responded to mechanical stimulation of the skin. Irradiation with low-intensity millimeter-band fields at 55.61 and 73 GHz reduced receptor unit sensitivity to mechanical stimulation. Half of the receptor units stopped responding 25 minutes after irradiation, while the other half continued responding after 35 minutes but with a significantly changed response character.
Dutta SK, Das K, Ghosh B, Blackman CF · 1992
Researchers examined the impact of radiofrequency radiation exposure on acetylcholinesterase activity. Neuroblastoma cells (NG108) exposed for 30 minutes to 147-MHz radiation amplitude-modulated at 16 Hz demonstrated enhanced acetylcholinesterase activity. The authors concluded that radiofrequency radiation affects both calcium-ion release and acetylcholinesterase activity in nervous system-derived cells in culture in a common dose-dependent manner.
Belyaev IYa, Alipov YD, Shcheglov VS, Lystsov VN · 1992
Researchers studied the effect of low-intensity microwaves on the conformational state of the genome of X-irradiated E. coli cells using the method of viscosity anomalous time dependencies. The frequency dependence of the observed effect had a resonance nature within the ranges of 51.62-51.84 GHz and 41.25-41.50 GHz, with a resonance half-width of the order of 100 MHz. A power density of 1 microW/cm2 was sufficient to suppress radiation-induced repair of the genome conformational state, across a tested range of 0.1-200 microW/cm2.
Khramov RN, Sosunov EA, Koltun SV, Ilyasova EN, Lednev VV · 1991
Researchers studied the effects of super high frequency microwaves (34-78 GHz) on spontaneous firing rates of slowly adapting stretch-receptor neurons in crayfish. Irradiation at power densities up to 250 mW/cm2 caused a transient decrease in firing rate, and after the field was turned off the rate increased and stabilized at pre-exposure levels. Firing rates also increased when the receptor muscle was stretched and were inversely correlated with small temperature increases of about 1.5°C. The response did not depend on frequency when temperature was held constant, and no resonant peaks were found across the millimeter-wave range.
Garson OM, McRobert TL, Campbell LJ, Hocking BA, Gordon I. · 1991
Researchers studied chromosome damage in 38 radio-linemen exposed to radio-frequency radiation and 38 matched controls. They analyzed Two hundred metaphases per person, scored blind, and found the rate of aberrant cells was the same in both groups. There were no statistically significant differences in the types of chromosome aberrations.
Ciaravino V, Meltz ML, Erwin DN · 1991
Researchers exposed Chinese hamster ovary cells to adriamycin and pulsed RFR at 2,450 MHz and 33.8 W/Kg for 2 h. The maximal temperature in the tissue-culture medium was 39.7 +/- 0.2 degrees C. The experiments were controlled for chemical and RFR exposures as well as for temperature. Verified statistically, the data indicate that the RFR did not affect changes in cell progression caused by adriamycin and did not change the number of sister chromatid exchanges induced by adriamycin.
Balcer-Kubiczek EK, Harrison GH. · 1991
Researchers exposed actively growing cultures of C3H/10T1/2 cells for 24 hours to 2.45-GHz microwaves pulse-modulated at 120 Hz. The electromagnetic fields alone had no effect on cell survival or induction of neoplastic transformation. When combined with the tumor promoter TPA, the enhancement of transformation was highly significant and ranged up to a level equivalent to that produced by 1.5 Gy of X rays. The frequency of neoplastic transformation was dependent on the level of electromagnetic exposure and was additive with doses of X rays given as a cocarcinogen.
Kerbacher JJ, Meltz ML, Erwin DN, · 1990
Cleary SF, Liu LM, Merchant RE · 1990
Researchers exposed whole human blood in vitro for 2 hours to 27 or 2,450 MHz radio-frequency electromagnetic radiation under isothermal conditions at 37 degrees C. At SARs below 50 W/kg, exposure produced a dose-dependent, statistically significant increase in lymphocyte proliferation, while exposure at 50 W/kg or higher suppressed proliferation relative to sham-exposed cells. The researchers concluded that these biphasic, dose-dependent effects on lymphocyte proliferation were not dependent on heating. There were no detectable effects of RF radiation on lymphocyte morphology or viability.
Parker JE, Kiel JL, Winters WD · 1988
Researchers exposed four rodent cell lines to 2450 MHz microwave radiation at a Specific Absorption Rate of 103.5 +/- 4.2 W/kg at temperatures of 37, 40, 42, and 45 degrees C. The researchers extracted mRNA from exposed and sham-exposed cells and used DNA probes of oncogenes, heat shock protein, or long terminal repeat sequences to look for differences in mRNA expression. With the cell lines and probes used, no significant differences in mRNA expression were observed after microwave exposure.
Further Reading:
Reduce Your Exposure:
Reduce your 5G exposure with shielding that deflects up to 99% of EMFFor a comprehensive exploration of EMF health effects and practical protection strategies, explore these books by R Blank and Dr. Martin Blank.
