Cancer incidence and mortality in the vicinity of nuclear installations England and Wales 1959–1980

Cancer incidence and mortality in the vicinity of nuclear installations England and Wales 1959–1980

J Envtron Radzoacuvlrv 6 (I988) 283--286 Book Reviews Cancer incidence and mortality in the vicinity of nuclear installations England and Wales 1959...

146KB Sizes 0 Downloads 50 Views

J Envtron Radzoacuvlrv 6 (I988) 283--286

Book Reviews

Cancer incidence and mortality in the vicinity of nuclear installations England and Wales 1959-80. P. J. Cook-Mozaffan, F. L. Ashx~ood, T_ Vincent, D. F o r m a n and M Alderson Office of Population Censuses and Surveys, H M S O , 1987. This report has taken several years to prepare_ The work was supported m part bv a grant from the National Radlolog~cal Protection Board The report ~tself consists of i3 introductory, pages, 280 pages of text including tables, appendices, references, e t c , and 27 microfiche transparencies with a capacity for a further 7290 pages. Thus it seems unlikely that anxone will read it from beginning to end With a document of this size, ~t ts important that the reformation about contents and ho~ to find items of interest ~s clear and readily available. In general, the report ts reasonably good m this connection The data considered m the report are largely the standardtsed mortahtx ratios (SMR) and the analogous standardtsed registration ratios (SRR). which have been used for many years by the Office of Population Censuses and Surveys (OPCS). They are based on the causes of death entered on death certificates (SMR) and diseases which have been diagnosed and registered (SRR), so the latter is a measure of the mcxdence of a pamcular disease (Note that, in the mare tables of th~s report, the term SMR ts used for incidence as well as mortality, which might be confusing ) They are normalised to 100. so that values of less or more than 100 indicate less or more cases than the expected number Since the interpretation of the rauos must be in statistical terms, a measure of the sigmficance of the ratios ~s also gwen and, in the report, the 95% confidence interval ts used, ~hich ~arms m w~dth according to the number of cases obse~'ed, assuming that the,,' are taken from a Po~sson distribution. 283 J Environ. Ra&oactwl~.' (6) (1988)---@ Elsevier Apphed Scmnce Pubhshers Lid England 1988 Printed m Great Britain

In an examination ot the tables un the report, rather than looking at the S M R and S R R values themselves, ~t ~s better to o b s e ~ e the lower and upper hmzts of the confidence mter~al m order to gain some apprec|anon ~t ,~hether the mcndence or mortaILt,, ~s sLgmficantl~ greater or less than the expected value_ For each nuclear mstallat|on, areas of different s~zes around ~t ha~e been studmd and a control area. consLdered to be as far as posstble comparable but wLthout a nuclear installation, has also been stud~ed The incidence and mortality at different time intervals before 1980 have been studied and the data have also been sub-divided according to population age m groups (~24, 25-74 and 75 and over. Seven power station and se~en non-power-stauon nuclear installations have been separately ~tudied, together V,lth se~en coastal zones surrounding the Irish Sea. Unfortunately, m practace many of the sub-divLsuons contain such a small n u m b e r of cases that thmr confidence intervals are too large for any useful m t e r p r e t a u o n to be made. A crmcusm of the report ~s that the attempt to analyse so many sub-diwsions was pointless and only succeeded m enlarging the d o c u m e n t and dduting any message that ~t mnght possibly contain If we restrict ourselves to the sum of all the non-power-stat|on installations, we nmprove the statistncal quality of the data For example, for ages 0--24, areas with 2/3rds of the population w|thLn 6 miles, between 1961 and 1980, for all malignant s~tes, we have the S R R 95% confidence interval 105-5-125-0, with the controls 88,9--106.7. ThLs appears to be a significant effect, untnl we look at the corresponding S M R confidence intervals, whnch are 86.1-109-1 (installations) and 81.4-103-7 (controls). This ~s of no sLgnificance. It is difficult to explain in physical terms wh~ there should be an effect Ln the S R R but not sn the SMR. Th~s sutuation LStypical and seems to apply to all of the data tabled for the nuclear |nstaltations In a fanrly rapid examination of the data un Table 2, the only effects which a p p e a r e d to be significant are for Lwerpool (one of the coastal zones), the control for whnch is Manchester. For example, for age 25-74, all malignant sites, the S R R 95% confidence interval is 120.0-122-4, with the control 106.2-108.5 The corresponding S M R Lntervals are 109-9-112-5 ( L w e r p o o l ) , and 107-6-110.3 (control). Thus, according to these data, there ~s a significant excess of cancer incidence and mortality m L v e r p o o [ c o m p a r e d with Manchester and w~th the populatnon as a whole. The chances of this being due to its coastal posit|on cannot be dismissed but the ~dea that th~s is caused by radioactnvity m the Irish Sea from nuclear installations seems mcred%le, since areas closer to the installations show no significant effect. The data gxven in th~s report, whmh presumably represents most of those presently available to the OPCS for England and Wales, demonstrate that

no statl~ucallv significant evidence extsts to hnk the mortahtv from cancer to residence m the ,,~cmttv of a nuclear mstallauon There are a number of sltuattons m ~ hich the 95% confidence inte~ al appears to sho~ a s~gmficant hnk as far as incidence is concerned but this is not repeated m the mortaht~ analysis [n sub-dl~ ~dmg the data to the extent which they ha~ e, the authors appear to have given themselves a great deal of ~ o r k and made their report much larger, neither of which is justified bv the extra value of the findings obtained

N. J. Pattenden 73 B Essex S t , ,\ e~ bur~. Berkshtre R G I 4 6RA. UK

O u r Radiant World. By D. W Lllhe Iowa State Umversxty Press, Ames, Iowa, t986_ 226 pp, ISBN 0-8138-1296-8. Price. $19 95 W h a t we normally think of as "radiation' in our environment--which we do, admittedly, refer to as i o m z m g radiation--~s of course but one part of the total spectrum of electromagneuc radmtton, plus partmle radiatton. The idea behind this book is thus to consider the whole continuum of such radlaUon in both the natural and man-made world As such, ~t ~s a good ~dea and makes for some interesting reading The book is divided into four main parts radtatton, origins of radiation m today's world, effects of radiation, and radiation and you Plus an "afterword' on Chernobyl. The first part is fairly lucid but--surprise, surprise-there is not a mention of SI umts. The second part concentrates almost entirely on the nuclear industries--nuclear reactors, the nuclear fuel c',cle. nuclear weapons, nuclear m e d m m e - - p l u s a brief chapter on electromagnetic radiation. The third part d~scusses familiar ground regarding the effects of iomzmg radiation, but does dwell longer on the non-iomzing radmtlons There are, for example, mterestmg accounts of the developments of US standards regarding exposure to microwaves--which began initially as a result of the use of radar for mtlitary purposes---including controversies regardmg effects on the eye, the brain, and the n e ~ ous system m general The principal studms m this field have been those made by the U S A and the USSR. It is thus fitting that one of the b~ggest ~ssues regarding effects on man has been that arising from low-level m~crowa~e stgnals apparently entering the USA Embassy m Moscow between 1953 and 1976! There is also a discussion concerning the possible effects of elecmc and