Study type: Epidemiological study (observational study)

Childhood Leukemia and 50 Hz Magnetic Fields: Findings from the Italian SETIL Case-Control Study. epidem.

Published in: Int J Environ Res Public Health 2015; 12 (2): 2184-2204

Aim of study (acc. to author)

A case-control study was conducted in Italy to investigate the association between childhood leukemia and exposure to extremely low frequency magnetic fields (ELF-MF).

Endpoint/type of risk estimation

Type of risk estimation: (odds ratio (OR))

Exposure

Assessment

Exposure groups

Group Description
Reference group 1 magnetic flux density in child's bedroom: ≤ 0.1 µT
Group 2 magnetic flux density in child's bedroom, arithmetic mean: > 0.1 - 0.2 µT
Group 3 magnetic flux density in child's bedroom, arithmetic mean: > 0.2 µT
Group 4 magnetic flux density in child's bedroom, geometric mean: > 0.1 - 0.2 µT
Group 5 magnetic flux density in child's bedroom, geometric mean: > 0.2 µT
Group 6 magnetic flux density in child's bedroom, percentiles: > 0.1 - 0.2 µT
Group 7 magnetic flux density in child's bedroom, percentiles: > 0.2 µT
Group 8 magnetic flux density in child's bedroom, percentiles: > 0.1 - 0.2 µT
Group 9 magnetic flux density in child's bedroom, percentiles: > 0.2 - 0.3 µT
Group 10 magnetic flux density in child's bedroom, percentiles: > 0.3 µT

Population

Case group

Control group

Study size

Cases Controls
Eligible 745 1,475
Evaluable 412 587
Statistical analysis method: (adjustment: )

Conclusion (acc. to author)

The geometric mean of magnetic flux density was 0.04 µT. Overall, 3 out of 540 (0.6%) children with leukemia and 13 out of 830 (1.6%) control children were exposed to magnetic flux densities of more than 0.3 µT.
No association between childhood leukemia and residential exposure to extremely low frequency magnetic fields was observed in analyses based on continuous exposure, while analyses based on categorical variables were characterized by incoherent exposure-outcome relationships.
The authors conclude that their results may be affected by several sources of bias and they are not informative at exposure levels of more than 0.3 µT. Nonetheless, the present study may contribute to future meta-analyses or pooled analyses. Furthermore, exposure levels among population can be useful to estimate attributable risk.

Study funded by

Related articles