The systematic review by Mevissen et al. (2025, Environment International) evaluated the evidence on the carcinogenicity of radiofrequency electromagnetic fields (RF-EMF) in laboratory animals and concluded with a high certainty of evidence (CoE) that exposure to RF-EMF increases the risk of malignant glioma and malignant schwannoma in the brain and heart, respectively. Deviating from their pre-published systematic review protocol, the authors did not perform meta-analyses. Instead, they based their assessment on whether or not statistically significant increases in tumour rates were observed in the included studies. One positive finding was deemed sufficient to conclude an adverse effect of RF-EMF exposure in a specific organ, thereby setting the target of the CoE rating for that organ. Here, we question this approach because it does not consider all the available evidence, and highlight further methodologically inconsistent decisions, while laying out a quantitative alternative based on the protocol and common guidelines for systematic reviews. Re-assessing the eligible long-term carcinogenicity experiments, we consider important studies to be sufficiently similar to be combined in a meta-analysis (MA). We calculate odds ratios as the effect measure and perform MA as well as dose-response MA. Rating the results using GRADE and OHAT guidance, we downgrade the CoE for imprecision due to the very wide confidence intervals of the pooled odds ratios, and upgrade the CoE for malignant heart schwannomas because of a positive exposure-response association, concluding moderate and low CoE for carcinogenicity of RF-EMF exposures in the heart and brain, respectively. In summary, our quantitative assessment of the evidence results in lower CoE conclusions than those of Mevissen et al.
Belenki et al. (Sun,) studied this question.