Relationship between radiofrequency-electromagnetic radiation from cellular phones and brain tumor: meta-analyses using various proxies for RF-EMR exposure-outcome assessment.

The debate about whether there is an association between radiofrequency electromagnetic fields (RF-EMF) from mobile phones and the incidence of brain tumors has been ongoing since the early 2000s. Many researchers have conducted meta- and subgroup analyses to address this question. However, these studies have not found a clear positive association. More recently, with the introduction of 3G phones, user behavior has changed dramatically. The focus has shifted from voice calls to multimedia use, such as YouTube, TikTok, and other social media platforms. In addition, the use of WPAN technologies such as Bluetooth devices has become widespread. These changes increase the duration of mobile phone exposure and complicate the comparability of exposure based on individual use characteristics. In light of these developments, current metrics for assessing mobile phone exposure are imprecise and inadequate for a valid assessment of tumor incidence. In theory, an accurate assessment of mobile phone exposure should be based on i) site-specific, ii) time-integrated, and iii) specific absorption rate (SAR) parameters. Common metrics such as years of mobile phone use, cumulative call duration, and number of calls per week are rough indicators at best. The iii) SAR parameter is roughly made up of two components: the duration of use and the phone's output power. However, output power varies significantly depending on factors such as phone model, network used, and location of use. Regarding the ii) time-integrated parameter, an appropriate approximation would be the cumulative duration of use, weighted by the power at each time point, differentiated between sides of the head, and considering Bluetooth hands-free devices or headphones. This review addresses the potential association between mobile phone use and brain tumor incidence. Specifically, the authors conducted a series of meta-analyses and subgroup analyses using multiple exposure assessment categories, ranging from broad to more precise categorizations.

The authors analyzed original research articles on mobile phone use (1G, 2G, and 3G) and brain tumor risk from the PubMed, EMBASE, and Cochrane Library databases, published through July 2024. Review articles, conference abstracts, and similar publications were excluded. Cohort studies were included but analyzed separately. Data from 19 case-control studies and 5 cohort studies were extracted and included in the meta-analysis. Instead of the usual risk of bias (RoB) assessment for each study, the authors assessed selection and recall bias for total mobile phone use and misclassification and recall bias for ipsilateral/contralateral use. (Ipsilateral means on the same side of the body, contralateral means on the opposite side. For ipsilateral use, the mobile phone is held on the same side of the head where the tumor is located; for contralateral use, the phone is held on the opposite side, editor's note.) The primary meta- and subgroup analyses of the case-control studies were categorized into four groups: 1) regular vs. irregular mobile phone use; 2) laterality (ipsilateral and contralateral use vs. irregular use); 3) duration of use > 10 years or < 10 years; 4) analyses of the first three categories differentiated by tumor type (glioma, meningioma, acoustic neuroma, pituitary tumor, and malignant tumor). A separate meta-analysis was performed for participants with more than 896 hours of cumulative use. For cohort studies, the categories were: a) "ever use" vs. "never use" and b) "over 10 years use" vs. "never use"; c) within these categories, meningioma, acoustic neuroma, and glioma were distinguished. All meta-analyses were statistically evaluated with pooled point estimates (odds ratio, OR) and 95% confidence intervals (CI).

First, the results of the case-control studies are summarized. Comparing regular and infrequent mobile phone users (category 1), no statistically significant differences in tumor incidence were observed. Considering laterality (category 2), a statistically significant 40% increase in the overall risk of brain tumors was found for ipsilateral mobile phone use. In contrast, the risk was not increased for contralateral use. For use over 10 years (category 3), a statistically significant 27% increase in risk was documented. Shorter periods of use of less than 10 years were not associated with an increased incidence of brain tumors. When differentiating by tumor type in category 1, a significant 16% reduced risk of meningioma was observed for regular users. For ipsilateral use, the risk increased by 20% for meningioma, 45% for glioma, and 93% for malignant tumors. Participants with more than 10 years of use had a 32% increased risk of developing glioma. All other combinations of subcategories and tumor types did not yield statistically significant data. Cumulative use of more than 896 hours was associated with a 59% increased risk of brain tumors, regardless of tumor type. The meta-analysis of cohort studies did not provide statistically significant evidence of an increased tumor risk from mobile phone use.

In this meta-analysis, the pooled point estimates (ORs) showed increasing and statistically significant values with more precise subgrouping. (Double-digit percentage risk increases are considered significant, editor's note.) According to the authors, the number of cohort studies included was too small to draw valid conclusions. In addition, the assessment of exposure conditions in terms of i) site-specific, ii) time-integrated, and iii) specific absorption rate (SAR) parameters was too imprecise. The authors call for future cohort studies to refine exposure subcategories. It is also critical to account for changing use patterns, including WPAN technologies. Furthermore, there is evidence of potential underestimation in the risk assessments of previous studies, leading to adjustments in their risk of bias. The relatively short observation periods, considering the latency of tumor development, and the age at exposure onset are factors that, if not properly accounted for, could lead to risk underestimation. According to the authors, future studies should aim to address these biases in their study designs. (RH)

"For ipsilateral use, the risk increased by 20% for meningioma, 45% for glioma, and 93% for malignant tumors."