ISSN:
0197-8462
Keywords:
PCA
;
principal component analysis
;
effects function
;
exposure metrics
;
exposure indices
;
Life and Medical Sciences
;
Occupational Health and Environmental Toxicology
Source:
Wiley InterScience Backfile Collection 1832-2000
Topics:
Biology
,
Physics
Notes:
Epidemiologic studies examining the risk of cancer among occupational groups exposed to electric fields (EF) and or magnetic fields (MF) have relied on traditional summaries of exposure such as the time weighted arithmetic or geometric mean exposure. Findings from animal and cellular studies support the consideration of alternative measures of exposure capable of capturing threshold and intermittent measures of field strength. The main objective of this study was to identify a series of suitable exposure metrics for an ongoing cancer incidence study in a cohort of Ontario electric utility workers. Principal components analysis (PCA) and correlational analysis were used to explore the relationships within and between series of EF and MF exposure indices. Exposure data were collected using personal monitors worn by a sample of 820 workers which yielded 4247 worker days of measurement data. For both EF and MF, the first axis of the PCA identified a series of intercorrelated indices that included the geometric mean, median and arithmetic mean. A considerable portion of the variability in EF and MF exposures were accounted for by two other principal component axes. The second axes for EF and MF exposures were representative of the standard deviation (standard deviation) and thresholds of field measures. To a lesser extent, the variability in the exposure variable was explained by time dependent indices which consisted of autocorrelations at 5 min lags and average transitions in field strength. Our results suggest that the variability in exposure data can only be accounted for by using several exposure indices, and consequently, a series of metrics should be used when exploring the risk of cancer owing to MF and EF exposure in this cohort. Furthermore, the poor correlations observed between indices of MF and EF reinforce the need to be take both fields into account when assessing the risk of cancer in this occupational group. Bioelectromagnetics 19:140-151, 1998. © 1998 Wiley-Liss, Inc.
Additional Material:
12 Tab.
Type of Medium:
Electronic Resource
Permalink
