Radiation literature survey

ARPANSA article review date

December 2024

Article publication date

February 2025

Summary

This study was a re-analysis of data from the INTEROCC case-control study and assessed the risk of glioma and meningioma from occupational exposure to radiofrequency electromagnetic energy (RF EME) using an updated job exposure matrix (JEM). The study included data from 7 countries (Australia, Canada, France, Germany, Israel, New Zealand, and the United Kingdom) collected between 2000 and 2004. Participants were aged from 30–59 years and included 1819 glioma and 1758 meningioma cases and 5227 controls. The RF exposure was estimated for both the electric and magnetic fields using a JEM that estimates the exposure for 468 different occupations. The study generally found no statistically significant associations. Overall, the authors concluded there was no risk of glioma or meningioma. 

Link to the study

Occupational exposure to radiofrequency electromagnetic fields and brain tumor risk: Application of the INTEROCC job-exposure matrix

Published in

Cancer Epidemiology

Commentary by ARPANSA

This study was a re-analysis of the INTEROCC study by Vila et al (2018 and 2022) that estimated occupational exposure based on spot measurements. This study obtained similar results, with Vila et al (2018 and 2022) who also concluded there was no clear association between occupational RF-EME exposure and glioma or meningioma. One limitation of the study was that 70% of the occupational exposures relied on data from 5 or less measurements. This reduces the reliability of the JEM and could mean that exposures in the JEM for individual occupations may not be completely represented. 

The overall conclusion of this study is consistent with the findings of the recent WHO systematic review by Karipidis et al (2024) that reported no overall increase in the risk of glioma from occupational RF exposure. The systematic review reported there was limited research on the risk of meningioma from occupational RF exposure. However, the review did report that there was no increased risk in meningioma among mobile phone users. The conclusions are also consistent with studies that investigated trends in brain tumour incidence rates over time (Elwood et al, 2022; Deltour et al, 2022), including an Australia study (Karipidis et al, 2018) that have consistently found no increase in the rates of brain tumours.  

Date of review by ARPANSA

2 December 2024

Article publication date

29 September 2024

Summary

This study examined the effect of short wavelength ultraviolet (UV) radiation on the acute skin response in mice. Hairless mice were irradiated with UVC light of wavelengths between 200-270 nanometres (nm) at varying intensity. The mice were then visually inspected for various skin damage markers like reddening and fissures at 24, 48 and 72 hours after exposure. The UV radiation dose required to produce a perceptible skin response increased greatly as the wavelength became shorter, ranging from 80.8 J/m² at 270 nm to 269000 J/m² at 215 nm. Doses for wavelengths shorter than 215 nm increased further but were excluded from later discussion due to high measurement uncertainty. 

The highest dose where no adverse skin effect was observed was estimated as the safe dose for that wavelength and compared to limits published by the International Commission on Non-Ionizing Radiation Protection (ICNIRP) and the American Conference of Governmental Industrial Hygienists (ACGIH). The comparison shows that substantially more UVC radiation (at wavelengths below 240 nm) was needed to cause skin damage in mice than the limit values described by the ICNIRP and the ACGIH. The authors further suggest that the values derived from the mouse tests will under-estimate the threshold in humans due to the thicker outer layers in human skin.

Published in

Photochemistry and Photobiology

Link to study

Extending the acute skin response spectrum to include the far-UVC

ARPANSA's commentary

Historically, interest in UV radiation with respect to health has been mostly restrained to wavelengths that are a component of solar radiation at the Earth’s surface (280-400 nm). An increase in interest for germicidal applications of far UVC light represents a need to re-evaluate the potential health effects of this type of light so that germicidal devices can be used safely. This study presents new evidence for the level at which UVC radiation can cause acute skin effects and thus how antimicrobial technology could be more effectively operated at higher power while maintaining safety. 

It should be noted that this study only investigated the acute skin effects of UVC exposure. Other well-documented effects of UV radiation like direct DNA damage (Mizutani, R. & Yokoyama, H., 2014; You, Y. et al., 2001) and the production of ozone (Claus, H. 2021) which can lead to chronic health outcomes must also be evaluated for a complete assessment of safety. The exposure limits set by the ICNIRP are designed to protect against long-term and delayed effects of UV radiation in addition to acute skin effects. Therefore, the comparison between the ICNIRP limits and the results of this study, which only considers acute skin effects, is somewhat mismatched. Further, when assessing the overall safety of far UVC light, consideration must also be given to the harmful effects to the eyes in addition to the skin.

Although exposure to UVC radiation from germicidal devices poses some hazards, exposure to intense UVA and UVB radiation from the sun remains the largest contributor to personal UV radiation exposure and risk for Australians. ARPANSA recommends following the five sun protection principles whenever the UV index is over three. 

Date of review by ARPANSA

November 2024

Article publication date

October 2024

Summary

This systematic review and meta-analysis assessed the association between artificial light-at-night (ALAN) exposure and cancer incidence in human populations. A total of 28 studies (15 cohort, 13 case-control) were included in the review and meta-analysis. The strength of the association was reported in risk ratio (RR) with a 95% confidence interval (CI). RR of cancer (breast, prostrate, and others - colorectal, pancreatic, non-Hodgkin lymphoma and thyroid cancer) were assessed. The studies had indoor ALAN exposure data assessed through self-reported questionnaires, whereas outdoor ALAN exposure data were collected from satellite data. Risk of bias and quality of the included studies were evaluated using the tool recommended by the Joanna Briggs Institute (Munn et al. 2015). The meta-analyses showed no statistically significant association for indoor ALAN exposure and breast cancer, and outdoor ALAN exposure and prostate cancer. Higher levels of outdoor ALAN exposure were associated with increased breast cancer risk (RR =1.12, 95 % CI 1.03–1.23). The qualitative synthesis of evidence indicated positive associations between ALAN exposure and risk of non-Hodgkin lymphoma, colorectal, pancreatic and thyroid cancer. Overall, the included studies had high quality scores. 

Published in

Science of the Total Environment

Link to study

Indoor and outdoor artificial light-at-night (ALAN) and cancer risk: A systematic review and meta-analysis of multiple cancer sites and with a critical appraisal of exposure assessment

ARPANSA's commentary

The review indicates an association between outdoor ALAN exposure and increased breast cancer. Importantly, this review contributes to the body of knowledge on the potential risk of cancer incidence associated with ALAN exposures. The risk estimates provided by this meta-analysis are comparable to those provided by other reviews (Urbano et al., 2021; Luo et al., 2023). Strengths of this review are that it included prostate cancer in the meta-analysis and other cancers in the qualitative review, and it critically appraised exposure assessment. Accordingly, it highlighted some methodological limitations of the included studies, such as self-reported assessment of indoor ALAN exposure that results in recall bias, exposure measurement error and exposure misclassification potentially biasing the risk estimates. Similarly, the satellite data used in estimating outdoor ALAN exposure in the included studies had poor spatial resolution and had no colour (e.g., blue light) data. Therefore, as highlighted in the review, future epidemiological research using robust GIS-based ALAN exposure data and methods, controlling for potential confounders (such as noise, pollution, green space) are needed to improve the risk estimates of cancer incidence. 

The International Commission on Non-ionizing Radiation Protection (ICNIRP) has published a statement on short wavelength light (SWL) exposure from indoor artificial source and human health. The ICNIRP acknowledges that there is no scientific consensus on whether ALAN exposure from SWL light causes health effects. It is important to note that the studies informing an association between ALAN exposure and cancer risk have limitations. Therefore, well-designed epidemiological studies with improved exposure assessment tools are required to better inform on whether long-term ALAN exposure is a human health risk. There are also other publications which provide some recommendations for visible light exposure and potential health effects (e.g. SSLC, 2024; Brown et al., 2022). 

Article review date

Oct 2024

Summary

This systematic review examined the awareness and use of the ultraviolet radiation (UV) index according to the World Health Organisation definition. The authors identified 40 publications with an outcome related to either awareness of the UV index (UVI), sun exposure or protective behaviours in association with UVI, and the impact of UVI interventions. The review also assessed the risk of bias within the included studies using the working group guidelines of Joanna Briggs Institute (Munn et al. 2015).  The review identified variation in public awareness of the UV index between countries.  The highest level of awareness was noted in Australia with over 90% of study participants reporting to have UV index awareness. Notably, the awareness in other countries was reported to be much lower (Europeans 50%, New Zealanders 43%, North Americans 34%). Despite being the high awareness of the UV index in Australia, only less than 10% of Australians use the UV index to inform their sun protection behaviour. There was a high risk of bias for all the outcomes that examined the use of the UV index for sun protective behaviours in Australia. The review recommended that further information needs to be disseminated on the advantages on using the UV index to better inform sun protection behaviours globally. 

Published in:

Kaiser et al. Photochemistry and Photobiology, 2024

 Link to study

A systematic review has examined the awareness of UV index and how it is used to inform sun protection messaging globally, and in Australia 

ARPANSA commentary

The study shows that awareness of UV index considerably varied globally with Australians having the highest awareness of the UV index. This finding demonstrates the success of the Australian sun protection messaging to spread awareness of the UV index. However, the UV index has not always been considered by the Australian public while considering how to protect themselves from UV exposure. In Australia, there are policies and recommendations in place to protect people from harmful UV exposure. For example, in Victoria, 97% of early childhood services and 90% of primary schools have policies to use the UV index to inform sun protection (SunSmart 2024) in their settings. Similar data for other Australian state and territory are lacking. The use of media and app to enhance the dissemination awareness of the UV index has been highlighted in the study. One of the limitations of the review is that it is difficult to draw qualitative evidence on how UV index is used to inform sun protection behaviours. Therefore, more research is needed to understand how Australians use the UV index. 

Despite the high awareness of the UV index in Australia and the Slip, Slop, Slap, Seek and Slide messaging, Australia has some of the highest rates of melanoma and skin cancer in the world and two-thirds of Australians will receive a skin cancer diagnosis of some type in their lifetime. As such, skin cancers, including melanoma, continue to constitute a large public health burden. This indicates that further work is needed to improve Australians’ sun protection behaviours and improve awareness on how to use the UV index to inform those behaviours. The UV index helps Australians know when UV exposure is high, and they should avoid sun exposure or practice other sun protection measures. More information on UV protection can be found on the ARPANSA Sun Protection factsheet. The UV index must be disseminated with greater efficacy via the media and apps. Currently, a free SunSmart Global UV app is available to know live UV index of global cities in view of boosting public awareness of the UV index globally. 

Article review date

26 October 2024

Article publication date

18 November 2024

Summary

This systematic review and narrative synthesis examined the evidence of genotoxic effects of radiofrequency electromagnetic fields (RF-EMF) on mammalian cells in vitro. The authors identified 159 articles eligible for inclusion in the review. Results from the included studies were grouped by genotoxicity endpoint and summary statistics that describe the results qualitatively were produced. Risk of bias and certainty of evidence evaluations were conducted according to OHAT-GRADE, although certainty of evidence assessments only considered studies that showed a statistically significant effect. 

The reviewers found that, in 80% of all experiments, RF-EMF exposure did not induce statistically significant effects on the analysed endpoint. Spindle disturbances (90%), chromatin condensation (63%) and 8-OH-dG adducts (57%) were the only endpoints where the majority of experiments reported significant effects, however, the total number of experiments for these endpoints was small. The restricted certainty of evidence assessments found that there was low to very low confidence in the articles that did report a statistically significant effect. The review concludes that RF exposure probably does not increase the occurrence of genotoxic effects in vitro.

The risk of bias ratings were also used to identify an inverse relationship between study quality and the proportion of experiments reporting an effect, where low quality studies reported more statistically significant effects. Similarly, the experiments with the lowest exposure levels and durations reported the most amount of statistically significant outcomes, contrary to what is expected for an ordinary dose-response relationship.

Published in

Environment International

Link to study

Systematic review on the effects of RF-EMF on genotoxicity in mammalian cells

ARPANSA commentary

The conclusions of the review are in alignment with those of past reviews (Karipidis et al., 2021) however, these conclusions are based on a qualitative synthesis only as limitations in the body of evidence unfortunately precluded a more informative meta-analysis. The primary reason for this was that the majority of studies presented data in an unextractable graphical format only. A lack of dedicated guidelines for systematic reviews on mechanistic types of evidence, in vitro studies in this case, also hindered the review process. Due to the shortcomings of narrative synthesis, the reviewers are restricted in the level of detail they can include in their conclusions about health endpoints. The most important outcomes from this review are therefore the evaluations of the relationship between studies reporting significant effects and study quality.   

GRADE certainty of evidence evaluations are typically used to evaluate evidence for health endpoints that are studied by the review. In this review’s implementation they restricted this analysis to studies that showed a statistically significant effect and assessed four of the eight categories defined in the GRADE methodology. The result is that the evidence assessments are an evaluation of the quality of studies that show a statistically significant effect, instead of an evaluation of certainty in the body of evidence for the health outcome. The finding that there is low to very low confidence in the studies showing a statistically significant effect in conjunction with the observed relationship between risk of bias and significant effects lends credibility to the overarching conclusion of the review that RF-EMF exposure probably does not increase the occurrence of genotoxic effects in vitro.

An ARPANSA review (Wood et al., 2021) has previously identified a similar negative correlation between effect size and study quality as well as between effect size and RF exposure level. This was also identified in a recent systematic review (Meyer et al., 2024) on the effects on RF-EMF on biomarkers of oxidative stress that ARPANSA has previously reviewed.

Article review date

November 2024

Article publication date

September 2024

Summary

This systematic review and meta-analysis assessed the association between radon exposure and health effects other than lung cancer in human populations. A total of 129 studies, published during 1990-2023, were included in the review and 40 studies were used in the meta-analysis. Risk estimates such as risk ratio (RR) per 100 units (Bq/m³ or Working level Month) of increased radon exposure level were calculated with a 95% confidence interval (CI). A range of health outcomes, including cancers (except lung cancer) and non-cancerous diseases were considered in the evaluation. The quality of the included studies were jointly evaluated using the New-castle Ottawa Scale (NOS) and the UNSCEAR’s approach of assessing quality of studies (including ecological studies) (UNSCEAR, 2017, p 48, table 3).The meta-analyses generally showed no statistically significant positive association for any health outcomes. Some of the estimated risk ratios were positive with the significance level near the threshold  for statistical significance including lympho-hematological cancer incidence in children (RR = 1.01; 95% CI: 1.00-1.03; p = 0.08); malignant melanoma mortality in adults in the general population (RR = 1.10; 95% CI: 0.99-1.21; p = 0.07); liver cancer mortality (RR = 1.04; 95% CI: 1.00-1.10; p = 0.06), and intestine and rectal cancer mortality combined among mine workers (RR = 1.02; 95% CI: 1.00-1.04; p = 0.06). The included studies had moderate to high quality scores. In conclusion, this review showed an absence of statistically significant associations between radon exposure and potential health outcomes other than lung cancer however, these effects cannot be ruled out. 

Published in

Frontiers in Public Health

Link to study

A recent review assessed evidence on links between radon exposure and health outcomes apart from lung cancer

ARPANSA commentary:

The overall conclusion of the review indicates no significant association between radon exposure and a range of health outcomes other than lung cancer in children and adult populations. Importantly, this review contributes to the body of knowledge on the potential health effects (both malignant and non-malignant diseases) of radon exposure in occupational and residential settings. Comparison of the findings of this review with other studies was not possible due to a lack of similar published reviews. Of the epidemiological studies included (except those on miners related to occupational radon) in this review, a majority of them related to children and adults were ecological studies, which may lead to ecological fallacy. There were several methodological limitations of the included studies, such as assessment of radon exposure and health outcomes as well as accounting for potential confounders which may alter the risk estimation. Therefore, further well-designed epidemiological research using robust radon exposure assessment tools and methods are needed to evaluate the potential risk of radon exposure for a range of population health outcomes. Some limitations of the study were: the inclusion of few studies in the meta-analysis of the exposure-risk relationships may have resulted in a lack of statistical power, significant heterogeneity was estimated in studies on occupational exposure to radon which limits the interpretation of the risk estimates and meta-regressions/in-depth sensitivity analyses was not done to evaluate the sources of heterogeneity due to the limited number of included studies. The review also did not conduct a certainty in evidence assessment which is another limitation.

According to the Australian Radiation Protection and Nuclear Safety Agency’s (ARPANSA) nationwide survey of Australian homes, the average radon exposure level in Australian homes is much lower than the global average. Consequently, residential radon exposure is not expected to be a significant contributor to any health risk other than lung cancer in Australia. ARPANSA has a fact sheet on ionising radiation exposure doses, including natural sources such as radon, and health. The International Commission on Radiological Protection has published guidance on radiological protection against radon exposure.

Article review date

October 2024

Article publication date

September 2024

Summary

This study compares the radiofrequency electric and magnetic field (RF-EMF) exposures estimated for the INTEROCC RF-EMF job-exposure matrix (JEM) and the measured personal exposure data to assess whether it can be used as a surrogate measure for occupational RF-EMF exposure in epidemiological studies. First, some occupations relevant to RF-EMF exposures were selected based on exposure prevalence and levels indicated in the preliminary INTEROCC-JEM. Personal RF-EMF exposure measurements were then undertaken for some selected occupational group of workers. The workplace exposure (i.e. high-intensity RF-EMF) levels over full work shifts (8 hours) were measured with a personal exposure meter (RadMan 2XT) worn outside the participant's clothing at the left chest level. Total RF-EMF exposure measured was estimated. The level of agreement between the exposure measure (exposure intensity or level and prevalence) of the INTEROCC-JEM and that of the measured personal exposure was estimated. A total of 333 workers representing 22 occupations provided their personal exposure data.  The study found a poor agreement for both intensity and exposure prevalence between the measurements and the INTEROCC-JEM estimates. On average, the INTEROCC-JEM tended to overestimate the measured RF-EMF exposure levels by >194%. The study also found that the exposure prevalence was higher in the measurements compared to the JEM.

Published in

Annals of Work Exposures and Health

Link to study

Comparison of a radiofrequency electric and magnetic field source-based job-exposure matrix with personal radiofrequency exposure measurements

Commentary by ARPANSA

Overall, the data of personal exposure-based JEM and source-based (INTEROCC) JEM showed little agreement as the former seemed to overestimate the workplace RF-EMF exposure. Therefore, the findings of the previous epidemiological research using the INTEROCC-JEM should be interpreted accordingly in view of the overestimation by this particular JEM. Importantly, the findings also indicate the need for considering new metrics for occupational RF-EMF exposure and collecting additional personal exposure data. This will help improve job exposure matrix applicable in future epidemiological studies. The main strength of this study is that it compared the measured RF-EMF exposures from a large number of workers (Migault et al., 2019). The study limitations were: i) the high detection threshold and low sensitivity of the measurement device limited the measurements to detect low level RF-EMF exposures such as those from Wi-Fi and mobile phone telecommunication sources which generally co-exist in occupational environments; ii) the exposure measurements did not characterise source (frequency band) specific exposures. The INTEROCC group has published (Vila et al., 2016; Migault et al., 2019) an occupational RF-EMF exposure measurement database to provide some evidence for epidemiological studies.

Occupational RF-EMF exposures in Australia are generally far below the limits given in the Australian Safety Standard (e.g. RPS-S1), s. The Australian Safety Standard, developed by ARPANSA, is consistent with the 2020 ICNIRP guidelines and is based on validated international scientific evidence.

Article review date

November 2024

Article publication date

August 2024

Summary

This study reports on the results of two birth cohort studies that were constructed to study the effects of Iodine-131 that was released due to a fire at Windscale nuclear reactors at Sellafield in Cumbria, England in 1957. The study aimed to examine the incidence of thyroid cancer in young children in Cumbria. Iodine-131 concentrates in the Thyroid, hence thyroid cancer was the key outcome of the studies. The two cohorts included 56,086 births in Cumbria between 1950-1959 and 137,444 more births in Cumbria from 1959-1980, respectively. These cohorts were broken up into three groups based on regions that were modelled to have high to low levels of Iodine-131. Area 1 being the highest, Area 2 with intermediate exposure level and Area 3 having the lowest. For the combined results from Areas 1 (highest) and 2 (intermediate) there was no increased risk of thyroid cancer with an Incidence rate ratio (IRR) of 0.68 (95% confidence interval (CI): 0.24-1.56). There were no cases of thyroid cancer for those in the highest exposure group. The authors conclude that a substantially raised thyroid cancer risk, even in the highest exposure groups, can be excluded. The authors said follow-up studies should continue to monitor the exposed groups for thyroid cancer. However, noted that these studies are unlikely to result in any changes to these conclusions. 

Published in

Radiation and Environmental Biophysics

Link to study

Thyroid cancer incidence in cohorts exposed in childhood to ¹³¹I released during the Windscale nuclear reactor accident at Sellafield, England, in 1957

Commentary by ARPANSA

The study examined the impact of Iodine-131 released due to an accident at a nuclear power plant. It provides strong evidence that the preventative measures instituted after the emergency, such as a restriction on the distribution of contaminated milk, were correct and may have prevented an increased risk of thyroid cancer occurring in the area. This is illustrated by the events in Chernobyl where consuming heavily contaminated milk led to tens of thousands of children receiving thyroid doses > 1 Gy and thousands of additional thyroid cancer cases (UNSCEAR 2018). This study also provides confidence in the conclusions of UNSCEAR 2022 about the impact of the Fukushima Dai-ichi nuclear reactor accident in Japan during in March 2011. UNSCEAR 2022 concluded that increased thyroid cancer risks will be small and unlikely to be detectable. This is because of the protective measures put in place after the accident that limited Iodine-131 exposure.  No adverse health effects have been reported for the residents of Fukushima that could be directly attributed to radiation exposure (UNSCEAR 2022).

ARPANSA also undertook protective measures during and after the Fukushima Dai-ichi nuclear reactor accident to ensure the safety of Australians. This included investigating the levels of radioactivity in foods imported into Australia from Japan and testing the levels of radioactivity in the ocean water, seafood and the atmosphere. Other products coming into Australia from Japan were also tested such as cars, and cargo vessels. ARPANSA can use the lessons learnt from these accidents to improve how it responds to nuclear accidents in the future and to ensure the safety of Australians. 

Article review date

October 2024

Article publication date

August 2024

Summary

This systematic review and meta-analyses examined the scientific evidence of the effect of radiofrequency electromagnetic fields (RF-EMF) on biomarkers for oxidative stress as reported in experimental animal (in vivo) and cell (in vitro) studies. A literature search found 12409 studies and of these, 56 met the inclusion criteria for further analysis. Results were analysed by individual biomarkers for oxidative stress and were further subdivided by the biological system the marker was recorded from. Results were reported as the standardised mean difference (SMD) with a 95% confidence interval (CI). Certainty of evidence and risk of bias for each result was assessed according to OHAT-GRADE. 

For many outcomes, calculating a pooled effect size was not possible as the results of the included studies were too inconsistent to make a meaningful synthesis, ranging from large decreases to large increases compared to a control. Where a pooled effect size was calculated, most outcomes reported a statistically non-significant result however, some of the analyses showed significant effects ranging from small to large. These effects were reported for oxidised DNA bases in rodent cells (SMD 2.49; 95% CI [1.30, 3.67]), plasma (SMD 2.25; 95% CI [1.27, 3.24]) and testis (SMD 1.60; 95% CI [0.62, 2.59]) as well as modified proteins in the thymus (SMD 6.16; 95% CI [3.55, 8.76]) and liver (SMD 0.55; 95% CI [0.06, 1.05]) of rodents. However, all outcomes, regardless of whether they showed an effect or not, were assessed as very-low certainty evidence. This was primarily due to high risk of bias ratings across the included studies and the large degree of inconsistency and imprecision in the measured effects. In some cases, the analysis was based on results from single studies with very small sample sizes. Although a full dose response analysis was not possible, the authors identify several counter-intuitive trends across the studies where the study with the lowest RF-EMF exposure level reported the largest effect or the study with middling exposure reported the largest effect. 

The study concludes that, due to the very low certainty in the evidence, the current scientific literature does not have any implications for policy related to RF-EMF but does highlight the need for drastic improvements in future studies.

Published in

Environmental International

Link to study

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

Commentary by ARPANSA

The present study represents the most comprehensive and latest review of the scientific literature related to the effect of RF-EMF exposure on oxidative stress. The review also identifies several challenges in this field of assessment. Evaluating the generation of reactive oxygen species or biomarkers of oxidative stress is challenging because these umbrella terms represent a broad class of molecules that participate in an extremely wide range of chemical transformations, both positive and negative. One of the challenges for biomarkers of oxidative stress is that they can also be formed through other mechanisms, giving results that may lack enough specificity. As such, it is recommended that multiple biomarkers are used to assess oxidative stress (Frijhoff et al., 2015; Murphy et al., 2022). However, singular biomarkers are overwhelmingly over-represented in many of the meta-analyses in this review, reflecting the low overall quality of the scientific evidence. Further to the challenges of measurement, the connection between oxidative stress and various diseases is also complex. While oxidative stress is frequently associated with the pathology of certain diseases, it is comparatively infrequently identified as a part of the cause and onset of a disease (Forman & Zhang, 2021). 

The counter-intuitive negative correlation between effect size and RF-EMF level has also been shown in a previous analysis conducted by ARPANSA of in vivo and in vitro studies investigating the bioeffects of RF fields used by 5G telecommunications (Wood et al, 2021). The ARPANSA review also reported a negative correlation between effect size and study quality, meaning that poor quality studies showed RF bioeffects whereas good quality studies did not.

These challenges are reflected in the current body of evidence and consequently the very low certainty ratings given to all meta-analyses in this review. The most important outcomes from the review are the flaws identified in the body of evidence and the recommendations made for improvements. Further experimental studies that follow these recommendations are required before more robust conclusions about the relationship between RF-EMF and biomarkers of oxidative stress can be made.

This review is part of the World Health Organisation’s ongoing project assessing the health effects of RF-EMF (SR9 – Effect of exposure to RF on biomarkers of oxidative stress). ARPANSA is supporting this WHO review process. 

Literature review date

September 2024

Article publication date

July 2024

Summary

This systematic review and meta-analysis examined the effect of radiofrequency electromagnetic energy (RF EME) on cognitive performance in human experimental studies. The review included 50 studies in the meta-analysis (2,433 participants) from 15 countries that were conducted between 2001 and 2021. It assessed cognitive performance of five (of the 7) cognitive domains (Orientation and attention, Perception, Memory, Construction and motor performance, and Concept formation and reasoning), including 19 (of 35) cognitive domain subclasses as defined by Lezak et al (2004). Each subclass was assessed based on the speed and accuracy of the participant performing tasks associated with that particular subclass (attentional capacity, concentration, etc). Most of the included studies used 2G-era mobile phone RF EME as the exposure source. Certainty of evidence and risk of bias were also evaluated according to OHAT-GRADE approach. 

The authors reported no statistically significant results, indicating that RF-EME exposure has not impact on cognitive performance on human in experimental studies for all of the 19 subclasses tested. However, the certainty in these results ranged from high to low. Three of the 19 subclasses were not able to be tested for both speed and accuracy. Of 35 meta-analyses performed on the subclasses, a majority (20) of them provided high certainty of evidence that RF EME exposure results in little to no difference in cognitive outcome. The evidence on the effect of RF EME on the remaining 16 subclasses of the cognitive domains was uncertain. This uncertainty was due to limitations in the study quality, inconsistency in the results or imprecision of the combined effect size estimate. Overall, the review reported no evidence that RF EME exposure impacts cognitive performance. 

Link to

The effect of exposure to radiofrequency electromagnetic fields on cognitive performance in human experimental studies: Systematic review and meta-analyses

Published in

Environment International 

ARPANSA commentary

The overall conclusion of the review was that RF EME exposure does not impact cognitive performance in human experimental studies. This systematic review is the most comprehensive that has been published on this topic as it considered all cognitive domains and people of all ages. Other recent reviews and meta-analyses have had similar outcomes (Valentini et al, 2010; Barth et al ,2011; Zubko et al, 2017) reporting no impact of mobile phone frequencies on short term cognitive performance. However, these studies only examined a few cognitive domains including, attention, processing speed and working memory. Further, these reviews did not assess the quality or risk of bias of their included studies in the same manner as this systematic review. For example, Barth et al (2011) performed no quality assessment; Valentini et al (2010) used a simple assessment that assessed the blinding, participant selection, exposure assessment and statistics used. Zubko et al (2017) did use the risk of bias tool developed by OHAT, however, it did not assess the certainty of the evidence according to the GRADE approach. 

This review is part of the World Health Organization’s (WHO) ongoing project assessing the health effects of RF-EMF. ARPANSA is supporting this WHO review process. One of the WHO systematic reviews has examined the long-term impacts on cognition of RF EME exposure (Benke et al 2024). This review was conducted in Australia and was co-led by ARPANSA scientists. This systematic review also reported that RF EME had no consistent impact on cognition.

Overall, the results from this systematic review are consistent with previous research and ARPANSA’s scientific understanding that exposure to RF EME below the limits in the ARPANSA RF safety standard do not cause health impacts, including cognitive outcomes.