The effects of radiofrequency electromagnetic field exposure on biomarkers of oxidative stress in vivo and in vitro: A systematic review of experimental studies

Background: Oxidative stress is thought to be related to many diseases. Furthermore, it is hypothesized that radiofrequency electromagnetic fields (RF-EMF) may induce excessive oxidative stress in various cell types and thereby have the potential to compromise human and animal health. The objective of this systematic review (SR) is to summarize and evaluate the literature on the relation between the exposure to RF-EMF in the frequency range from 100 kHz to 300 GHz and biomarkers of oxidative stress.

Methods: The SR framework was developed following the guidelines established in the WHO Handbook for Guideline Development and NTP/OHAT's Handbook for Conducting a Literature-Based Health Assessment. We used the latter handbook's methodology for implementing the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach for environmental health assessments. We searched the following databases up until June 30, 2023: PubMed, Embase, Web of Science Core Collection, Scopus, and the EMF-Portal. The reference lists of included studies and retrieved review articles were also manually searched. We rated Risk of Bias (RoB) using the OHAT RoB Rating Tool and assessed publication bias using funnel plots of included studies. We assessed the certainty of the evidence (high, moderate, low, or very low) for an association between RF-EMF and oxidative stress using an adapted version of the GRADE framework. Data were extracted according to a predefined set of forms developed in DistillerSR. Data were analysed after grouping them first as in vitro or in vivo and then according to outcome category, species category, and exposed tissue. We synthesized study results using a random effects meta-analysis when study characteristics were judged sufficiently similar to be combined and heterogeneity (I²) was lower than 75 %, otherwise we describe the findings narratively.

Results: Fifty-six (56) studies, 45 in vivo and 11 in vitro, in which cells (in vitro) or animals (in vivo) were exposed to frequencies in the range 800-2450 MHz, were included in the systematic review after eliminating 12,353 publications because they did not meet the criteria defined in the published protocol (Henschenmacher et al., 2022). Of 56 studies 52 studies with 169 individual results were included in the meta-analysis. Together, these studies examined six human in vitro samples and fifty animal samples, including rodents (mice, rats, hamsters, and guinea pigs, (n = 46)) and rabbits (n = 4). RF-EMF were predominantly applied as continuous wave exposures in these studies. The outcome biomarkers for modified proteins and amino acids were measured in n = 30 studies, for oxidized DNA bases in n = 26 studies, for oxidized lipids in n = 3 studies and hydrogen peroxide production in 2 studies. Outcomes were mostly measured in the brain (n = 22), liver (n = 9), cells (n = 9), blood (n = 6), and testis (n = 2). RoB in studies was high, mainly due to biases in exposure and outcome assessment.

In vivo studies: Brain: The effect on biomarkers for oxidized DNA bases in the rodent brain (five studies, n = 98) had an inconsistent effect, varying from a large decrease with a standardized mean difference (SMD) of -3.40 (95 % CI [-5.15, -1.64]) to a large increase with an SMD of 2.2 (95 % CI [0.78, 3.62]). In the brain of rabbits (two studies, n = 44), the effect sizes also varied, from an SMD of -1.06 (95 % CI [-2.13, 0.00]) to an SMD of 5.94 (95 % CI [3.14, 8.73]). The effect on biomarkers for modified proteins and amino acids in the rodent brain (15 studies, n = 328) also varied from a large decrease with an SMD of -6.11 (95 % CI [-8.16, -4.06]) to a large increase with an SMD of 5.33 (95 % CI [2.49, 8.17]). The effect on biomarkers for oxidized lipids in the brain of rodents (one study, n = 56) also varied from a large decrease with SMD = -4.10 (95 % CI [-5.48, -2.73]) to SMD = 1.27 (95 % CI [0.45, 2.10]). Liver: The effect on biomarkers for oxidized DNA bases in the rodent liver (two studies, n = 26) was inconsistent with effect sizes in both directions: SMD = -0.71 (95 % CI [-1.80, 0.38]) and SMD = 1.56 (95 % CI [0.19, 2.92]). The effect on biomarkers for oxidized DNA bases in the rabbits' liver (two studies, n = 60) was medium with a pooled SMD of 0.39 (95 % CI [-0.79, 1.56]). Biomarkers for modified proteins and amino acids in the liver of rodents (six studies, n = 159) increased with a pooled SMD of 0.55 (95 % CI [0.06, 1.05]). Blood: The effect of RF-EMF on biomarkers for oxidized DNA bases in rodent blood (four studies, n = 104) was inconsistent, with SMDs ranging from -1.14 (95 % CI [-2.23, -0.06]) to 1.71 (95 % CI [-0.10, 3.53]). RF-EMF had no effect on biomarkers for modified proteins and amino acids in rodent blood (three studies, n = 40), with a pooled SMD of -0.08 (95 % CI [-1.32, 1.16]). There was a large increase in biomarkers for oxidized DNA bases in rodent plasma (two studies, n = 38) with a pooled SMD of 2.25 (95 % CI [1.27, 3.24]). Gonads: There was an increase in biomarkers for oxidized DNA bases in the rodent testis (two studies, n = 24) with a pooled SMD of 1.60 (95 % CI [0.62, 2.59]). The effect of RF-EMF on biomarkers for modified proteins and amino acids in the ovary of rodents (two studies, n = 52) was inconsistent with a medium effect, SMD = 0.24 (95 % CI [-0.74, 1.23])) and a large effect (SMD = 2.08 (95 % CI [1.22, 2.94])). Thymus: RF-EMF increased biomarkers for modified proteins and amino acids in the thymus of rodents (one study, n = 42) considerably with a pooled SMD of 6.16 (95 % CI [3.55, 8.76]). Cells: RF-EMF increased oxidized DNA bases in rodent cells with SMD of 2.49 (95 % CI [1.30, 3.67]) (one study, n = 27). There was a medium effect in oxidized lipids (one study, n = 18) but not statistically significant with SMD = 0.34 (95 % CI [-0.62, 1.29]).

In vitro studies: In in vitro studies in human cells (three studies, n = 110), there were inconsistent increases in biomarkers for oxidized DNA bases, where the SMDs varied between 0.01 (95 % CI [-0.59, 0.62]) and 7.12 (95% CI [0.06, 14.18]) in 4 results (2 of them statistically significant). In rodent cells (three studies, n = 24), there was a not statistically significant large effect in biomarkers for oxidized DNA bases with SMD = 2.07 (95 % CI [-1.38, 5.52]). The RF-EMF biomarkers for modified proteins and amino acids in human cells (one study, n = 18) showed a large effect with SMD = 1.07 (95 % CI [-0.05, 2.19]). In rodent cells (two studies, n = 24) a medium effect of SMD = 0.56 (95 % CI [-0.29, 1.41]) was observed.

Discussion: The evidence on the relation between the exposure to RF-EMF and biomarkers of oxidative stress was of very low certainty, because a majority of the included studies were rated with a high RoB level and provided high heterogeneity. This is due to inaccurate measurements of exposure and/or of measurement of oxidative stress biomarkers and missing information on the blinding of research personnel to exposure conditions or outcome measurements. There may be no or an inconsistent effect of RF-EMF on biomarkers of oxidative stress in the brain, liver, blood, plasma and serum, and in the female reproductive system in animal experiments but the evidence is of very low certainty. There may be an increase in biomarkers of oxidative stress in testes, serum and thymus of rodents but the evidence is of very low certainty. Future studies should improve experimental designs and characterization of exposure systems as well as the use of validated biomarker measurements with positive controls. Other: This review was partially funded by the World Health Organization. The protocol for this review is registered in PROSPERO (https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42021235573) and published in Environment International (https://doi.org/10.1016/j.envint.2021.106932) (Henschenmacher et al., 2022).