Professional Documents
Culture Documents
Residential Exposure To Electric Power Lines and Risk of Various Disorders: A Case-Control Study
Residential Exposure To Electric Power Lines and Risk of Various Disorders: A Case-Control Study
2
3
Background
• Reported association between residence close to
power lines and childhood leukaemia
• Wertheimer & Leeper (1979) reported increased
childhood leukaemia associated with certain
‘household wiring codes’
• Multiple studies since gave variable results
• Meta-analysis reported doubling of risk
• Subject remains highly controversial
• Public concern
4
Electric and Magnetic Fields
Electric fields (kV/m)
related to voltage
7
Methods
8
Cases & controls
• We used an existing dataset collected 1972-80
• Included all diagnosed cases of LPD & MPD in Tasmania
• 783 adult cases and 71 childhood cases
• One control per patient
– By random numbers from electoral role (registration compulsory)
– Matched for sex and age ± 5 years
• Original study sought links between diagnoses &
– Occupation
– Place of residence
– Published x 2 in Journal of the Natl Cancer Institute 1984
• Contained complete residential history 9
Risk Factors Considered
Tasmanian
All pairs
only
Children 71 52
15
Results
16
Closest distance ever lived
to a power transmission line
Patients Controls
Distance
n=854 n=854
0-50 m 19 9
51-300 m 75 55
18
Risk associated with ever having
lived
0-50 m or 51-300 m from a power
line compared with ‘never’
Distance (m) Odds Ratio 95% CI
>300 1.00 -
19
Risk associated with ever having
lived
0-50 m or 51-300 m from a power
line compared with ‘never’ -
Tasmanian pairs
Distance (m) Odds Ratio 95% CI
>300 1.00 -
20
Risk associated with every year of
residence 0-50 m or 51-300 m from a
power line compared with ‘never’
>300 1.00 -
21
Added risks relating to
line voltage, compared to
never
88 kV 33%
110 kV 44%
220 kv 45%
P= ns
22
Risk associated with having lived 0-
300 m from a power line at 0-5 years
or 6-17 years compared with ‘≥18
years or never’
26
Exposure at time of birth
Patients 8
Controls 1
27
Methodological concerns
• Problems with controls
– Controls & cases selected differently
– Only a minority of invited controls accepted
• Controls more likely to have been born outside Tasmania
– Cases interviewed, controls sent mail questionnaire
• Possible ascertainment bias of case addresses
– However we attempted to control for these concerns
• Small numbers - wide CIs
• LPD, MPD pooled
• Observational study – cannot infer causality
28
Difficulties in
interpretation
• Factors other than lines themselves or EMFs
– SE status (perhaps incompletely controlled for)
• Lack of laboratory support for a link
– However no tests on newborn animals or in utero
• Difficult to measure exposure (1/rn?)
• 300 m too far for an effect?
– But consider ionised particles
29
But our findings are
consistent
• Consistent dose-response effect
– Distance categories (p for trend 0.02)
– Years of exposure
– Line voltage (n.s.)
– Age of exposure (p for trend 0.01)
• Findings strengthened when restricted to
201 Tasmania-only pairs
• Numerous precedents for finding of greatest
risk in childhood or in utero
30
Conclusion