Isabelle Deltour

Smartphones track indicators of a reference signal to maintain network connectivity. These can serve as proxies for a subset of environmental radiofrequency electromagnetic field (RF‐EMF) exposure from mobile networks. In this multi‐country study, we developed conversion functions linking signal indicators collected with the ETAIN 5G‐Scientist app to measured RF‐EMF exposure. Parallel measurements were conducted across six European countries and Australia using frequency‐selective exposimeters and the app. Associations of signal strength and quality indicators with exposure were analysed using log‐linear regression models: (1) serving‐cell level, relating exposure in the serving frequency band to indicators of the connected cell (Europe n = 2960; Australia n = 747); and (2) location level, relating indicators aggregated across network providers to environmental exposure from 4G‐LTE and 5G‐NR networks (Europe n = 61; Australia n = 48). At the serving‐cell level, 4G‐LTE signal strength was positively associated with exposure, while signal quality was inversely associated. For 5G‐NR, only signal strength was associated with exposure, with weaker and less consistent relationships than for 4G‐LTE. At location level, aggregation across providers reduced model error, suggesting residual variability operates at short temporal and spatial scales. These findings support using smartphone signal indicators to estimate components of environmental RF‐EMF exposure, although further validation is needed, particularly for 5G‐NR. Bioelectromagnetics. 00:00–00, 2026. © 2026 Bioelectromagnetics Society.

ObjectivesTo examine the association between self-reported mobile phone use and brain tumour risk in the UK Biobank Cohort.MethodsA total of 467,801 participants aged 40–69 years from the UK Biobank Cohort were included (recruited 2006–2010, followed up to 2023). ICD10 codes were used to identify all brain tumours and its subtypes (glioma (glioblastoma as subgroup), meningioma, acoustic neuroma). Regular mobile phone use was defined as making or receiving calls at least once per week. Hazard ratios (HR) and 95% confidence intervals (CI) were calculated using Cox proportional hazard regression models, with age as underlying time scale.ResultsOver a median follow-up of 13 years (5,583,093 person-years), 1,607 brain tumours were identified. Longest regular mobile phone use (>8 years) at baseline compared with never regular use was not associated with overall brain tumour risk (HR: 0.93, 95% CI: 0.79 to1.08) or any subtype (glioma (0.85, 95% CI: 0.69 to 1.04), glioblastoma (0.87, 95% CI: 0.69 to 1.11), meningioma (1.04, 95% CI: 0.79 to1.38), acoustic neuroma (0.91, 95% CI: 0.43 to 1.92).ConclusionThe result from this large-scale analysis adds to the evidence supporting absence of an association between mobile phone use and overall brain tumour risk or any subtype. Limitations in the exposure assessment remain, warranting further follow-up of studies with improved RF-EMF assessment.

Approximately half of cancers are attributed to modifiable risk factors or inherited genetic mutations, thus there remains the potential for as-yet unidentified environmental risk factors to explain substantially more of the global cancer burden. Current estimates of the environmental and occupational cancer burden account for a relatively small number of exposures, because of challenges in characterizing exposure patterns and human cancer risks. Hence, the environmental cancer burden may be underestimated because of combinations of not having identified all carcinogenic agents, lack of understanding of cancer risks at very low exposure levels of even known carcinogens, and lack of understanding of joint effects. The International Agency for Research on Cancer has a long history of addressing these challenges, and of advancing our understanding of occupational and environmental causes of cancer, with primary research predominantly conducted by the Environment and Lifestyle Epidemiology Branch. The International Agency for Research on Cancer’s position within the World Health Organization has facilitated the conduct of cross-border collaborations on important and sensitive topics, ranging from asbestos, ionizing radiation, and pesticides to its current work on artisanal petroleum refining. Here, we highlight some important Environment and Lifestyle Epidemiology contributions to several areas of occupational and environmental cancer, describe recent progress in other important and emerging exposures, and finally, provide suggestions for future research. In our constantly changing world, with new products, technologies, and demands transforming our environment and workplaces, there is a continued need for high-quality human cancer investigations of potential occupational and environmental carcinogens, especially those characterizing environmental levels of exposure encountered by the general population in their daily life and in underresearched regions of the world.

COSMOS-France is the French part of the COSMOS project, an international prospective cohort study that investigates whether the use of mobile phones and other wireless technologies is associated with health effects and symptoms (cancers, cardiovascular diseases, neurologic pathologies, tinnitus, headaches, or sleep and mood disturbances). Here, we provide the first descriptive results of COSMOS-France, a cohort nested in the general population-based cohort of adults named Constances. Methods: A total of 39,284 Constances volunteers were invited to participate in the COSMOS-France study during the pilot (2017) and main recruitment phase (2019). Participants were asked to complete detailed questionnaires on their mobile phone use, health conditions, and personal characteristics. We examined the association between mobile phone use, including usage for calls and Voice over Internet Protocol (VoIP), cordless phone use, and Wi-Fi usage with age, sex, education, smoking status, body mass index (BMI), and handedness. Results: The participation rate was 48.4%, resulting in 18,502 questionnaires in the analyzed dataset. Mobile phone use was reported by 96.1% (N = 17,782). Users reported typically calling 5–29 min per week (37.1%, N = 6600), making one to four calls per day (52.9%, N = 9408), using one phone (83.9%, N = 14,921) and not sharing it (80.4% N = 14,295), mostly using the phone on the side of the head of their dominant hand (59.1%, N = 10,300), not using loudspeakers or hands-free kits, and not using VoIP (84.9% N = 15,088). Individuals’ age and sex modified this picture, sometimes markedly. Education and smoking status were associated with ever use and call duration, but neither BMI nor handedness was. Cordless phone use was reported by 66.0% of the population, and Wi-Fi use was reported by 88.4%. Conclusion: In this cross-sectional presentation of contemporary mobile phone usage in France, age and sex were important determinants of use patterns.

The Cohort Study of Mobile Phone Use and Health (COSMOS) has repeatedly collected self-reported and operator-recorded data on mobile phone use. Assessing health effects using self-reported information is prone to measurement error, but operator data were available prospectively for only part of the study population and did not cover past mobile phone use. To optimize the available data and reduce bias, we evaluated different statistical approaches for constructing mobile phone exposure histories within COSMOS. We evaluated and compared the performance of 4 regression calibration (RC) methods (simple, direct, inverse, and generalized additive model for location, shape, and scale), complete-case analysis, and multiple imputation in a simulation study with a binary health outcome. We used self-reported and operator-recorded mobile phone call data collected at baseline (2007-2012) from participants in Denmark, Finland, the Netherlands, Sweden, and the United Kingdom. Parameter estimates obtained using simple, direct, and inverse RC methods were associated with less bias and lower mean squared error than those obtained with complete-case analysis or multiple imputation. We showed that RC methods resulted in more accurate estimation of the relationship between mobile phone use and health outcomes by combining self-reported data with objective operator-recorded data available for a subset of participants.

We investigated the association of allergic diseases and epilepsy with risk of brain tumours, in Interphone, a 13-country case–control study. Data were obtained from 2693 glioma cases, 2396 meningioma cases, and 1102 acoustic neuroma cases and their 6321 controls. Conditional logistic regression models were used to estimate pooled odds ratios (ORs) and their respective 95% confidence intervals (CIs), adjusted for education and time at interview. Reduced ORs were observed for glioma in relation to physician-diagnosed asthma (OR = 0.73; CI 0.58–0.92), hay fever (OR 0.72; CI 0.61–0.86), and eczema (OR 0.78, CI 0.64–0.94), but not for meningioma or acoustic neuroma. Previous diagnosis of epilepsy was associated with an increased OR for glioma (2.94; CI 1.87–4.63) and for meningioma (2.12; CI 1.27–3.56), but not for acoustic neuroma. This large-scale case–control study adds to the growing evidence that people with allergies have a lower risk of developing glioma, but not meningioma or acoustic neuroma. It also supports clinical observations of epilepsy prior to the diagnosis of glioma and meningioma.