Article publication date
11 February 2026
Article review date
20 April 2026
ARPANSA summary
This epidemiological study examined the relationship between long-term exposure to extremely low frequency electromagnetic fields (ELF-EMF) from electrical supply infrastructure and mortality from neurodegenerative diseases. Adults over 30 from the Swiss National Cohort were followed between the years 2001 and 2018. Their long-term exposure to high voltage power lines (HVPL) and railway power lines was estimated based on the proximity of participant residence to the lines. Results were reported as the hazard ratio (HR) per 1 µT of exposure with a 95% confidence interval (CI) which was computed while adjusting for sociodemographic and environmental factors.
The study computed a positive association between HVPL exposure and mortality from Alzheimer’s disease (HR per 1 µT = 1.54; 95% CI: 1.23–1.92) as well as other types of dementia (HR per 1 µT = 1.31; 95% CI: 1.13–1.52). No association was found between HVPL exposure and amyotrophic lateral sclerosis, multiple sclerosis or Parkinson’s disease. There was also no association between any neurodegenerative disease and railway power line exposure.
Published in
Environment International
Link to
ARPANSA commentary
This study follows on from previous work (Huss, A. et al., 2009), expanding the years followed and also considering exposure from railway power infrastructure in addition to HVPL. There was also large improvements in how exposure and environmental confounders were assessed in this study compared to how it was assessed in the previous work. Epidemiological studies investigating associations with electrical supply infrastructure and health endpoints have typically modelled exposure based on distance as a categorical variable (SCHEER, 2024). The approach taken in this article to model exposure as a continuous variable using a proximity model validated by measurements is an improvement to the accuracy of exposure assigned to participants.
However, this change in how exposure was assessed is also reflected in how the results have been reported which is deserving of some additional clarification. Sandoval-Diez et al. have reported hazard ratios per 1 µT of exposure despite the fact that only a very small proportion of the HVPL exposed population experience an exposure close to this magnitude (Loizeau, N. et al., 2024). These hazard ratios should therefore be interpreted carefully as some nuance is required to understand their meaning as they relate to realistic exposure scenarios.
The inconsistent results between different neurodegenerative diseases and different exposure sources reported in this study is challenging to interpret or consider in the context of a plausible mechanism. Such selective effects would need to proceed through mechanisms that are highly specific to the nature of the individual disease. This is further complicated when considering that the endpoint under investigation is neurodegenerative disease mortality. For many of these diseases, including Alzheimer’s disease and other types of dementia, their relationship to a cause of death, and accurate reporting thereof, can be complex (Dobson, A. et al., 2022; Matthews, F. et al., 2018). Studies that also investigate incidence, severity or disease progression may provide more reliable information about causality or a potential mechanism.
The International Commission on Non-Ionizing Radiation Protection is currently reviewing their low frequency exposure guidelines. Studies like this will help to inform the guidelines and are considered in the context of the overarching scientific evidence. ARPANSA has previously measured ELF-EMF exposure in Australian homes and has published this information (Karipidis, K., 2014). This measurement survey remains the largest of its kind. Similarly, ARPANSA has measured ELF-EMF exposure near electrical supply infrastructure and have published this as technical report 170. ARPANSA advises that there is no substantiated scientific evidence that exposure to residential ELF-EMF causes adverse health effects, more information is available on our factsheet.
