Incorrect and misleading statements made by Karipidis et al. (2024b) in response to the Commission’s critique of their SR on RF-EMF exposure and cancer, include the following:

1) Karipidis et al. (2024b) claim “we were comprehensive and transparent in our systematic review and meta-analytic approach.”

However, their presentations of the data on cumulative call time were based on “dose-response meta-analyses” methodology published by one of the authors of this SR. Yet, the Karipidis et al. (2024a) paper only provides graphs of the predicted dose-response relationships, without showing the model and its parameters, analyses of parameter sensitivity and statistical fit of the model to the data, as well as analyses of alternative dose-response approaches. The latter point is important since numerous meta-analyses have reported statistically significant increases in brain tumor risk associated with exposure to RF-EMF emissions from mobile phones. There was no mention of whether latency, a critical aspect in cancer assessments, was factored into their analyses. In other words, Karipidis et al. (2024a) expect the reader to accept their graphical presentation of the data without providing any supportive analyses on the certainty of the model outcome. Thus, the paper was not transparent in reporting its dose-response meta-analyses methodology. 

2) Karipidis et al. (2024b) claim “there are no links between any of the authors in our systematic review and the telecommunications industry,” and that several of the letter’s (Frank et al., 2024a) co-authors have undeclared conflicts of interest.”

Widespread linkages between the authorship team of Karipidis et al. (2024a) with the telecommunications industry have been documented (Hardell and Carlberg, 2020; Frank, 2021; Frank et al., 2024b). For example, Martin Röösli, a co-author of this SR and two other SRs in this series, declares in his competing interest statement that his “research is entirely funded by public entities or not for profit foundations.” However this statement omits the fact that a key “not for profit” entity that Roosli has received extensive research funding from is Swiss Research Foundation for Electricity and Mobile Communications (FSM) which is sponsored by several wireless companies including Swisscom, Swissgrid, Sunrise, Cellnex and Ericsson.

In addition, three authors of Karipidis et al. (2024a), Karipidis, Röösli, and Baaken are members of the International Commission on Non-Ionizing Radiation Protection (ICNIRP). ICNIRP is a body long demonstrated to have under-reported potential conflicts of interest (Nyberg et al., 2023; Lin, 2025). In contrast, none of the authors of our critique have had relationships with the telecommunications industry.

3) Karipidis et al. (2024b) claim “authors of the Danish cohort study have validated the subscriber status of 1355 cohort members,” with “61% of subscribers were indeed mobile phone users, and 16% of the nonsubscribers were actually regular mobile phone users before 1996.”

However, the cited paper (Schüz and Johansen, 2007) was not a “validation” of subscriber status in the Danish cohort which included the entire Danish population and more than 358,400 subscribers. Moreover, the response by Karipidis does not address the fact that this information was available in 2011 when the IARC working group downplayed the utility of the Danish cohort and its updated studies (Frei et al., 2011; Schüz et al., 2006a,b) because they relied on subscriptions to mobile phone providers as a surrogate of mobile-phone use: the study “lacked information on level of mobile-phone use and there were several potential sources of misclassification of exposure” (IARC, 2013).  In their response letter, Karipidis et al. (2024b) acknowledge that “ever” versus “never” and “time since start of use” are “simple exposure proxies.” While those proxies were the major factors in their SR meta-analysis, they do not account for cumulative exposure, and therefore are not reliable measures of brain cancer risk. 

4) Karipidis et al. (2024b) claim “the Million Women Study [Schüz et al., 2022] had an attrition rate of 8.9%,” which “is acceptable handling of subject attrition including very little (≤ 20%) missing outcome data.”

However, the rate of 8.9% was based on attrition in the year 2001, at the end of the recruitment period (1996-2001). At the 14-year follow-up in 2011, only 57%  (429,407 ÷ 747,561) of the women with no prior brain cancer who completed the questionnaire in 2001 continued to participate in the study. The 43% attrition rate (1-57%) is not “very little. 

5) Karipidis et al. (2024b) cite studies that claim case-control studies are “particularly sensitive to recall bias and thus overestimation of the risk.”

However, Karipidis neglected to cite Momoli et al. (2017) which re-analyzed the Canadian data from the Interphone study and showed that there was no effect on the risk of glioma after adjustments were made for selection and recall biases. Odds ratios (OR) for glioma were increased significantly and to a similar extent when comparing the highest quartile of use to those who were not regular users regardless of whether adjustments for biases were made.

6) Karipidis et al. (2024a) omitted consideration of the results on laterality of mobile phone use in relation to tumor localization (ipsilateral use) claiming that “there was an increased risk with ipsilateral use which was compensated by a decreased risk with contralateral use.” 

This statement is contradicted by results on glioma both by Hardell and Carlberg (2015) and Interphone (2010) that showed increased risk also for contralateral exposure, although lower than for ipsilateral exposure that would be expected in carcinogenesis.

7) Karipidis et al. (2024b) acknowledge “the limitations of time-trend simulation studies, including the inability to account for patterns in mobile phone use on the individual level,” while claiming “overall brain cancer incidence has remained largely unchanged.”

As we noted in our letter to the editor (Frank et al., 2024a), the findings from the case-control studies of increased cancer risks in the areas of the brain (temporal lobe) with the highest absorption of RF radiation emitted from a mobile phone held next to the head of the user have not been addressed in cancer time-trend analyses. Hence, those analyses do not validate the rating of certainty of brain cancer evidence by Karipidis et al. (2024a). In their letter, Karipidis et al. (2024b) claim that increases in glioblastoma multiforme when accompanied by decreases in other brain cancers is due to improvements in diagnostic techniques, while ignoring the possibility that such changes might have been due to tumor promotion by RF-EMF. Further, the statement by Karipidis et al. (2024a) on overall brain cancer incidence was obtained by exclusion of studies showing increasing incidence with time. Philips et al. (2018) reported “a sustained and highly statistically significant ASR (age–standardized incidence rates) rise in glioblastoma multiforme (GBM) across all ages.” In addition, Swedish data showing increasing incidence of brain tumors with time (Hardell and Carlberg, 2017), and the Danish Cancer registry data which shows that tumors in the brain and central nervous system have increased between 2004 and 2023 (Nye kræfttilfælde i Danmark 2023) were also not mentioned by Karipidis et al. (2024a).

8) Karipidis et al. (2024b) state that their “systematic review defines long latency as 10 years or more,” while noting that “ionizing radiation has been shown to induce brain cancer by causing DNA damage with a latency period of about 5 or more years” and that “assuming a similar latency for non-ionizing radiation as observed for ionizing radiation, one would expect that any relevant risk should already have started to emerge by now.”

However, the preamble to the IARC monographs states “experience with human cancer indicates that the period from first exposure to the development of clinical cancer is sometimes longer than 20 years; therefore, latent periods substantially shorter than 30 years cannot provide evidence for lack of carcinogenicity.”  Furthermore, even for ionizing radiation, the latency period for a significant increase in brain cancer cases among atomic bomb survivors was around 10 or more years; yet, increased risks of solid tumors have been detected more than 50 years after the bombings. There is no justification for the assumption that the latency for RF-EMF carcinogenicity is the same as that of ionizing radiation.

Lastly, in their responding letter, Karipidis et al. (2024b) claim “we have highlighted that there are limitations to the included studies and data which has some impact on the ability to draw definitive conclusions; hence, our rating of moderate certainty of evidence that near field RF-EMF exposure to the head from mobile phone used likely does not increase the risk of glioma, …” and in the SR, Karipidis et al.  (2024a) noted “there was no strong indication against the hypothesis of no summary effect of CCT [cumulative call time] on glioma risk.”

In contrast to the above statements, Karipidis has claimed in the media that “exposure to radio waves from mobile phones or wireless technologies is not associated with an increased risk of brain cancer,” and that “radio waves from wireless technologies are not a hazard to human health.”  These latter statements appear to intentionally mislead the public about potential health risks from exposure to RF-EMF.

In assessing health effects of environmental agents, public health agencies address the certainty in the evidence of an increase in risk, not in rating the certainty of evidence of no increase in risk. For example, because of limitations and variations in human exposures to RF-EMF due to factors such as intensity of phone emissions, differences in how people use and hold their phones, differences in frequency and modulation patterns of those emissions, as well as interindividual differences in human responses to these exposures, it is not possible to make any definitive conclusion on the lack of cancer risk based on available data. The concordant results from the animal carcinogenicity studies on RF-EMF and the increased risk of brain gliomas and Schwann cell tumors from the case-control studies point to a potential cancer risk that was ignored by Karipidis et al. (2024a) in their overall conclusions.

In sum, the response by Karipidis et al. (2024b) to our critique (Frank et al. 2024a) of their systematic review of the research on radiofrequency electromagnetic fields and human cancer studies fails to support their seriously flawed review (Karipidis et al., 2024a). Hence, we believe that this review should be retracted. Moreover, this review cannot be used as proof of cell phone safety.