Occupational and

Environmental MedicineAdopted as the Journal of the Faculty of

Occupational Medicine of the Royal College ofPhysicians ofLondon

Editor: Anne Cockcroft (United Kingdom)Technical Editor: Judith HaynesEditorial Assistant: Rachel HarveyEditorial Board:T C Aw (United Kingdom) R McNamee (United Kingdom) D C Snashall (United Kingdom)F J H Van Dijk (Holland) J Myers (South Africa) 0 Svane (Denmark)J S Evans (United States) B Nemery (Belgium) G Theriault (Canada)R M Harrison (United Kingdom) T Okubo (Japan) KM Venables (United Kingdom)J Jeyaratnam (Singapore) L Rosenstock (United States) Editor, Britsh MedicalJournalF Kauffmann (France) M Sim (Australia)R L Maynard (United Kingdom) M Singal (United States)

NOTICE TO CONTRIBUTORS Occupational and EnvironmentalMedicine is intended for the publication of original contribu-tions relevant to occupational and environmental medicine,including toxicological studies of chemicals of industrial, agri-cultural, and environmental importance, and epidemiologicalstudies. As well as full papers, short papers dealing with briefor preliminary observations relevant to occupational and envi-ronmental medicine will also be considered. Case reportsshould cover substantial new ground to merit publication.Other articles, including review or position papers, will be con-sidered but should not be submitted without first approachingthe Editor to discuss their suitability for the Journal. Letters tothe Editor are always welcome.INSTRUCTIONS TO AUTHORS Three copies of all submissionsshould be sent to: The Editor, Occupational and EnvironmentalMedicine, BMJ Publishing Group, BMA House, TavistockSquare, London WC1H 9JR, UK. All authors should sign thecovering letter as evidence of consent to publication. Papersreporting results of studies on human subjects must be accom-panied by a statement that the subjects gave written, informedconsent and by evidence of approval from the appropriateethics committee. These papers should conform to theprinciples outlined in the Declaration of Helsinki (BMJ1964;ii: 177).

If requested, authors shall produce the data on which themanuscript is based, for examination by the Editor.Authors are asked to submit with their manuscript the

names and addresses of three people who they considerwould be suitable independent reviewers. They will notnecessarily be approached to review the paper.

Papers are considered on the understanding that they aresubmitted solely to this Journal and do not duplicate materialalready published elsewhere. In cases of doubt, where part ofthe material has been published elsewhere, the publishedmaterial should be included with the submitted manuscript toallow the Editor to assess the degree of duplication. TheEditor cannot enter into correspondence about papers rejectedas being unsuitable for publication, and the Editor's decisionin these matters is final.Papers should include a structured abstract of not

more than 300 words, under headings of Objectives,Methods, Results, and Conclusions. Please include up tothree keywords or key terms to assist with indexing.

Papers should follow the requirements of the InternationalCommittee of Medical Journal Editors (BMJ 1991;302:338-41). Papers and references must be typewritten in doublespacing on one side of the paper only, with wide margins. SIunits should be used.

Short reports (including case reports) should be not morethan 1500 words and do not require an abstract. They shouldcomprise sections of Introduction, Methods, Results, andDiscussion with not more than one table or figure and up to10 references. The format of case reports should beIntroduction, Case report, and Discussion.Illustrations Photographs and photomicrographs on glossypaper should be submitted unmounted. Charts and graphsshould be carefully drawn in black ink on firm white paper.Legends to figures should be typed on a separate sheet ofpaper.References References will not be checked by the editorialoffice; responsibility for the accuracy and completeness of ref-erences lies with the authors. Number references consecutivelyin the order in which they are first mentioned in the text.Identify references in texts, tables, and legends by Arabicnumerals. References cited only in tables or in legends to fig-ures should be numbered in accordance with a sequence estab-

wished by the first identification in the text of a particular tableor illustration. Include only references essential to the argu-ment being developed in the paper or to the discussion ofresults, or to describe methods which are being used when theoriginal description is too long for inclusion. Information frommanuscripts not yet in press or personal communicationsshould be cited in the text, not as formal references.Use the Vancouver style, as in this issue for instance, for a

standard journal article: authors (list all authors when seven orfewer, when eight or more, list only six and add et at), title,abbreviated title of journal as given in Index Medicus (if notin Index Medicus give in full), year of publication, volumenumber, and first and last page numbers.Proofs Contributors will receive one proof. Only minor cor-rections can be made at this stage; corrections other thanprinter's errors may be charged to the author.

Reprints Reprints will be charged for. The number ofreprints required should be stated on the form provided withthe proofs.Copyright © 1995 Occupational and Environmental Medicine.This publication is copyright under the Berne Convention andthe International Copyright Convention. All rights reserved.Apart from any relaxations permitted under national copyrightlaws, no part of this publication may be reproduced, stored ina retrieval system, or transmitted in any form or by any meanswithout the prior permission of the copyright owners.Permission is not, however, required to copy abstracts ofpapers or of articles on condition that a full reference to thesource is shown. Multiple copying of the contents of the publi-cation without permission is always illegal.NOTICE TO ADVERTISERS Applications for advertisementspace and for rates should be addressed to the AdvertisementManager, Occupational and Environmental Medicine, BMJPublishing Group, BMA House, Tavistock Square, LondonWC1H 9JR.NOTICE TO SUBSCRIBERS Occupational and EnvironmentalMedicine is published monthly. The annual subscription rate(for 1995) is £139 (US $240). Orders should be sent to theSubscription Manager, Occupational and EnvironmentalMedicine, BMJ Publishing Group, BMA House, TavistockSquare, London WC1H 9JR Orders may also be placed withany leading subscription agent or bookseller. (For the conve-nience of readers in the USA subscription orders with or with-out payment may also be sent to Brtish Medical Journal, POBox 408, Franklin, MA 02038, USA. All inquiries, however,must be addressed to the publisher in London). All inquiriesregarding air mail rates and single copies already publishedshould be addressed to the publisher in London.Subscribers may pay for their subscriptions by Access, Visa, orAmerican Express by quoting on their order the credit orcharge card preferred together with the appropriate personalaccount number and the expiry date of the card.

Second class postage paid Rahway NJ. Postmaster: sendaddress changes to: Occupational and Environmental Medicine,c/o Mercury Airfreight International Ltd Inc, 2323 RandolphAvenue, Avenel, NJ 07001, USA.FACULTY OF OCCUPATIONAL MEDICINE The Faculty ofOccupational Medicine of the Royal College of Physicians ofLondon is a registered charity founded to promote, for thepublic benefit, the advancement of knowledge in the field ofoccupational medicine. The Faculty has offices at 6 StAndrew's Place, Regent's Park, London NW1 4LB.

ISSN 1351-0711.

Published by BMJPublishing Group andprinted in England byEyre & Spottiswoode Ltd,London and Margate

Coggon

of chlorinated solvents may be increasedimmediately next to a leaking drum on awaste disposal site-and the spatial distri-bution and depth of sampling points can havea profound influence on measured concen-trations. Also, the methods of storing andtreating samples before analysis can introducefurther variability.

In some cases it may be possible to stan-dardise sampling procedures. For example, inthe routine monitoring of drinking water that iscurrently carried out in Britain, samples arecollected according to a standard protocol,both at water treatment plants and at the tap.In other cases-for example, measurement ofpollutants in soil-the optimal sampling strat-egy will depend on the study question. Ineither situation, however, the methods ofsampling and of storing and transportingsamples should be adequately documented sothat results can be properly interpreted andcompared with those from other studies.

Quality assuranceThe third major need is for better qualityassurance of environmental exposure mea-surements. At present quality assurance inthis field is not as well developed as it is, forexample, in clinical chemistry, which couldwell serve as a model for good practice.Elements of a quality assurance programmeshould include observance of documentedprotocols and standardised procedures, notonly for sampling and storage of specimensbut also for their analysis in the laboratory;internal quality control, for example, throughthe use of "quality control samples" at regular

intervals; and external quality assessmentschemes, or if these are not available-forexample, for very specialised analyses such asassays of DNA adducts-checks on accuracywith certified reference materials.

It is intended that the working group'sreport will help the MRC to formulate itsresearch strategy on environmental pollution,and assist the research community by indicat-ing some of the priorities for future investiga-tion. Proposals for studies to examine thedeterminants of personal exposure and uptakeof pollutants, and the potential of DNAadducts as markers of uptake or proxies fordisease outcome, or both, are clearly impor-tant. Also, research proposals should pay dueattention to sampling methods and qualityassurance of assays.

Full copies of the working group's report are available onrequest from Mrs J Jones, Medical Research Council, 20 ParkCrescent, London WIN 4AL. The working group was estab-lished under the auspices of the MRC Committee on ToxicHazards in the Environment and Workplace (Chairman DameBarbara Clayton). The members of the group were Dr DCoggon (Chairman), Dr B Brown, Mr J Cherrie, Dr P BFarmer, Mr J K Fawell, Dr A Lovett, Dr D Gompertz, Dr PHarrison, Dr C W Suckling, Dr D Tennant, Dr A Wadge, DrK J Finney, Mr K Brennan, and Dr G Sarna.

1 Raw GJ, Coward SKD. Exposure to nitrogen dioxide in homesin the UK: a pilot study. Proceedings of "unhealthy housing:the public health response." Warwick: Warwick University,1991.

2 Knox EG. Leukaemia clusters in childhood: geographicalanalysis in Britain. J Epidemiol Community Health1994;48:369-76.

3 Wallace LA. The exposure of the general population tobenzene. Cell Biol Toxicol 1989;5:297-314.

4 Hemminki K, Dipple A, Shuker DEG, Kadlubar FF,Segerback D, Bartsch H. DNA adducts: identification andbiological significance. Lyon: International Agency forResearch on Cancer, 1994. (IARC Sci Publ No 125).

5 Quality of Urban Air Review Group. Urban air quality inthe United Kingdom. London: QUARG, 1993.

Vancouver styleAll manuscripts submitted to Occup EnvironMed should conform to the uniformrequirements for manuscripts submitted tobiomedical journals (known as theVancouver style.)

Occup Environ Med, together with manyother international biomedical journals, hasagreed to accept articles prepared in accor-dance with the Vancouver style. The style(described in full in the BMJ, 24 February1979, p 532) is intended to standardiserequirements for authors.

References should be numbered consec-utively in the order in which they are firstmentioned in the text by Arabic numeralsabove the line on each occasion the refer-ence is cited (Manson' confirmed otherreports2-5 . . .). In future references topapers submitted to Occup Environ Med

should include: the names of all authors ifthere are seven or less or, if there are more,the first six followed by et al; the title ofjournal articles or book chapters; the titlesof journals abbreviated according to thestyle of Index Medicus; and the first and finalpage numbers of the article or chapter.Titles not in Index Medicus should be givenin full.

Examples of common forms of refer-ences are:

1 International Steering Committee of Medical Editors,Uniform requirements for manuscripts submitted tobiomedical journals. BrMedJ 1979;1:532-5.

2 Soter NA, Wasserman SI, Austen KF. Cold urticaria:release into the circulation of histamine and eosino-phil chemotactic factor of anaphylaxis during coldchallenge. N EnglJ Med 1976;294:687-90.

3 Weinstein L, Swartz MN. Pathogenic properties ofinvading micro-organisms. In: Sodeman WA Jr,Sodeman WA, eds. Pathologic physiology, mechanismsof disease. Philadelphia: W B Saunders, 1974:457-72.

564

How have Zambian businesses reacted to the HIV epidemic?

to cope more effectively with the HIV epi-demic, make humane and workable policieson HIV testing for employees and give theirworkforces education to reduce HIV trans-mission and the stigma that hampers preven-tive measures. We are currently studying theeffect of outreach education on employees'behaviour and attitudes to HIV.

We acknowledge gratefully the help of Boniface Moonze andJanet Bennett (data entry); Maria Quigley (statistical advice);and Eric van Praag and Keith McAdam for their helpful com-ments. The study was funded by the Enid Linder Foundation.

1 Nkowane BM. Prevalence and incidence ofHIV infection inAfrica: a review of data published in 1990. AIDS1991;5(suppl 1):S7-15.

2 Cross S, Whiteside A, eds. Facing up to AIDS. London:Macmillan 1993.

3 Baggaley R, Godfrey-Faussett P. Msiska R, Chilangura D,Chitu E, Porter J, Kelly M. Impact of HIV on Zambianbusinesses. BMJ 1994;309:1549-50.

4 Murray GIL, Yang G, Qiao X. Adult mortality: levels, pat-terns and causes. In: Feachem RGA, Kjellstrom T,Murray GIL, Over M, Phillips MA, eds. The health ofadults in the developing world. New York: OxfordUniversity Press, 1992;23-1 11.

5 Buve A, Foster SD, Mbwili C, Mungo E, Tollenare N,Zeko M. Mortality among female nurses in the face ofthe AIDS epidemic: a pilot study conducted in two hos-pitals in the Southern Province of Zambia. AIDS1994;8:396-7.

6 Elliot A, Luo N, Tembo G, Halwiindi B, Steenbergen G,Machiels L, et al. The impact of HIV on tuberculosis inZambia: a cross sectional study. BMJ 1990;301:412-5.

7 Piot P. Kapita BM, Ngugi EN, Mann JM, Colebunders R.,Wabitsch R. AIDS in Africa: a manual for physicians.Geneva: World Health Organisation, 1992.

8 Hodgkinson N. The plague that never was. The SundayTimes London 3 October 1993.

Instructions to authorsThree copies of all submissions should besent to: The Editor, Occupational andEnvironmental Medicine, BMJ PublishingGroup, BMA House, Tavistock Square,London WC1H 9JR, UK. All authorsshould sign the covering letter as evidenceof consent to publication. Papers reportingresults of studies on human subjects must beaccompanied by a statement that thesubjects gave written, informed consentand by evidence of approval from theappropriate ethics committee. These papersshould conform to the principles outlinedin the Declaration of Helsinki (BMJ 1964;ii: 177).

If requested, authors shall produce thedata on which the manuscript is based, forexamination by the Editor.Authors are asked to submit with

their manuscript the names andaddresses of three people who theyconsider would be suitable independentreviewers. They will not necessarily beapproached to review the paper.Papers should include a structured

abstract of not more than 300 words,under headings of Objectives, Methods,Results, and Conclusions. Pleaseinclude up to three keywords or keyterms to assist with indexing.

569

Soutar, Harker, Seaton, Packe

11 Evans KA, Allen-Williams U. Electroantennogramresponses of the cabbage seed weevil, Ceutorhychusassimilis, to oilseed rape, Brassica napus ssp. olifera,volatiles. J Chem Ecol 1992;18:1641-59.

12 Butcher RD, MacFarlane-Smith W, Robertson GW,Griffiths DW. The identification of potential aeroaller-gens/irritant(s) from oilseed rape (Brassica napus spp.oleifera) volatile organic compounds emitted duringflowering progression. CGn Exp Alery 1994;24:1105-14.

13 Boulet L-P, Morin D, Turcotte H. Variations of airwaysresponsiveness to methacholine and exercise in asth-matic and normal subjects over a 12-month period. ClnInvestMed 1990;13:60-6.

14 Thein FCK, Mencia-Huerta J-M, Lee TH. Dietary fish

oils effects on seasonal hayfever and asthma in pollensensitive subjects. Am Rev Respir Dis 1993;147:1138-43.

15 Hyland ME, Kenyon CAP, Allen R, Howarth P. Diarykeeping in asthma: comparison of written and electronicmethods. BMJ7 1993;306:487-90.

16 Choweinczyk PJ, Parkin DH, Lawson CP, Cochrane GM.Do asthmatic patients correctly record home spirometrymeasurements? BMY 1994;309:1618.

17 Colldahl H. Rape pollen allergy. Report of a case. ActaAlergologica 1954;7:367-9.

18 Bucur I, Arner B. Rape pollen allergy. Scand J Respir Dis1978;58:222-7.

19 McGorum BC, Dixon PM. Preliminary observations oninhalation and intradermal challenges of horses withoilseed rape.VetRec 1992;131:163-7.

Correspondence and editorialsOccupational and Environmental Medicine wel-comes correspondence relating to any of thematerial appearing in the journal. Resultsfrom preliminary or small scale studies mayalso be published in the correspondencecolumn if this seems appropriate. Lettersshould be not more than 500 words in lengthand contain a minimum of references. Tablesand figures should be kept to an absolute

minimum. Letters are accepted on theunderstanding that they may be subject toeditorial revision and shortening.The journal also publishes editorials which

are normally specially commissioned. TheEditor welcomes suggestions regardingsuitable topics; those wishing to submit aneditorial, however, should do so only afterdiscussion with the Editor.

580

Vergnenegre, D'arco, Melloni, Antonini, Courat, Dupont-Cuisinier, Bonnaud

be responsible for chronic injury to airways(cattle feeding, grain milling, hay storage,inhalation of pesticides, as well as climaticconditions). Although exposure can bequantified, it does not provide a picture oftotal exposure to risks. Some studies haveattempted to assess exposure in terms of timespent in certain activities-for example,threshing or cattle feeding-although theyfailed to detect any correlation betweenchronic bronchitis and the duration of expo-sure.4 6 14

Despite the fact that observed values wererelated to the predictive values, age appearedin the analysis. It is possible that age repre-

sents an indirect measurement of the durationof exposure. Some studies have the sameresults: age may be a reflection of cumulativeexposure. 101 We found a relative risk for distalobstruction of 3-02 for the smaller farms(< 50 hectares) v the larger farms, which is inline with the findings of Saia et al.4 This riskwas still present, after stratification by age(OR = 2-66). It may be that the smallerfarms employ less mechanized equipment,thereby increasing exposure to certain risks.Analysis of characteristics of agriculturalactivities obtained from a more detailed ques-tionnaire could perhaps provide an answer tothis question.

In summary, this study shows that theprevalence of distal airway obstruction ishigher than the prevalence of chronic bronchi-tis in an agricultural population, afterstratification by smoking habits and history ofcardiac and other respiratory diseases. Ageand size of farms, which are negatively corre-

lated, emerge as the two main explanatoryvariables. These results indicate that farmingpopulations may have an early propensity todistal airway obstruction. The next steps shouldbe to define agricultural exposure by creating a

more detailed questionnaire, and to providemore evidence for the hypothesis that size offarm is a risk factor for respiratory diseases.

We thank Professor F Boutros-Toni for his technical assis-tance in statistical methods. This work was supported by a

grant from Mutualite Sociale Agricole and GroupamaInsurance, Limoges, France.

1 Schenker M, Ferguson T, Gamsky T. Respiratory risksassociated with agriculture. Occup Med 1991;6:415-28.

2 Dosman JA, Cotton DJ, Graham BL, Robert Li KA,Barnett GD. Chronic bronchitis and decreased forcedexpiratory flow rates in lifetime non-smoking grain work-ers. Am Rev RespirDis 1980;121:11-6.

3 Chan-Yeung M, Schulzer M, MacLean L, Dorken E,Grzybowski S. Epidemiologic health survey of grain ele-vator workers in British Columbia. Am Rev Respir Dis1980;121:329-38.

4 Saia B, Mastrangelo G, Marcer G, Reggio 0. Prevalenceand risk factors of chronic respiratory disease in a farmingpopulation. Med Lav 1984;75:101-9.

5 Milosevic M. The prevalence of chronic bronchitis in agri-cultural workers of Slavonia. Am 7 Ind Med 1986;10:319-22.

6 Dalphin JC, Bildstein F, Pernet D, Dubiez F, Depierre A.Prevalence of chronic bronchitis and respiratory func-tion in a group of dairy farmers in the French Doubsprovince. Chest 1989;95:1244-7.

7 Rylander R. Lung diseases caused by organic dusts. AmInd Med 1986;10:221-7.

8 Do Pico GA, Reddan W, Tsiatis A, Peters ME, Rankin J.Epidemiologic study of clinical and physiologic parame-ters in grain handlers of northern United States. Am RevRespir Dis 1984;130:759-65.

9 Manfreda J, Cheang M, Warren CPW. Chronic respira-tory disorders related to farming and exposure to graindust in a rural adult community. Am Ind Med1989;15:7-19.

10 Chan-Yeung M, Dimich-Ward H, Enarson DA, KennedySM. Five cross sectional studies of grain elevatorworkers. Am Epidemiol 1992;136:1269-79.

11 Dosman JA, Graham BL, Hall D, Loon PV, Bahsin P,Froh F. Respiratory symptoms and pulmonary functionin farmers. Occup Med 1987;29:48-3.

12 Heller RY, Hayward DM, Farebrother MTB. Lung func-tion in farmers in England and Wales. Thorax 1986;41:117-21.

13 Iversen M, Pedersen B. Relation between respiratorysymptoms, type of farming, and lung function disorders infarmers. Thorax 1990;45:919-23.

14 Gamsky TE, Schenker MB, McCurdy SA, Samuels SJ.Smoking, respiratory symptoms, and pulmonary func-tion among a population of hispanic farmworkers. Chest1992;101: 1361-8.

15 Pham QT, Teculescu D, Chau N, Bouchy 0, ConstantinoE, Moneret-Vautrin D. Etude respiratoire et aller-gologique dans le milieu agricole du department de laMeuse. Arch Mal Prof 1991;52:467-75.

16 Terho EO, Vohlonen I, Husman K. Prevalence and inci-dence of chronic bronchitis and farmer's lung withrespect to socioeconomic factors. Eur Respir Dis 1987;71:S29-36.

17 Malmberg P, Rask-Andersen A. Natural and adaptativeimmune reactions to inhaled microorganisms in thelungs of farmers. Scand Work Environ Health 1988;1:S68-71.

18 Rask-Andersen A, Mahmberg P, Lundhohm M. Endotoxinlevels in farming: absence of symptoms despite highexposure levels. BrJ nd Med 1989;46:412-6.

19 Revsbech P, Andersen G. Diurnal variation in peak expira-tory flow rate among grain elevator workers. Br IndMed 1989;46:566-9.

20 Prefault Ch, Tcheriatchoukine J, Guerrero AJ, Moutou H,Chardon G. Debit maximum expiratoire 25-75% dansune population de non fumeurs. Bull Eur Physiopath Resp1980;16:25-40.

21 American Thoracic Society. Lung function testing: selec-tion of reference values and interpretative strategies. AmRev RespirDis 1991;144:1202-18.

22 Organisation Mondiale de la Sante. Questionnaire sur lessymptomes respiratoires in OMS ed: le depistage precoce desmaladies professionnelles, 1st ed. Geneve: OrganisationMondiale de la Sante, 1989: 278-83.

23 Quanjer P. Standardized lung function testing. Bul EurPhysiopath Respir 1985;19:S45-51.

24 Mantel N, Haenszel W. Statistical aspects of the analysisof data from retrospective studies of diseases. Jf NatlCancer Inst 1959;22:719-48.

25 McWhorter, Polis M, Kalow R. Occurrence, predictors,and consequences of adult asthma in the NHANES-Iand follow-up survey. Am Rev Respir Dis 1989;139:721-4.

26 Pariente R, Coulaud C, Martin JP, Berthet C, BaldayrouP, Touaty E. Enquete sur les bronchopneumopathieschroniques en milieu agricole. Rev Fr Mal Respir 1979;7:633-8.

27 Dalphin JC, Debieuvre D, Pernet D, Maheu MF, PolioJC, Toson B, et al. Prevalence and risk factors forchronic bronchitis and farmer's lung in French dairyfarmers. Br Ind Med 1993;50:941-4.

28 Brackbill RM, Cameron LL, Behrens V. Prevalence ofchronic diseases and impairments among US farmers1986-1990. Am Epidemiol 1994;139:1055-65.

Rejected manuscriptsFrom February 1994, authors whose sub- be returned to them. The Journal will destroymitted articles are rejected will be advised of remaining copies of the article but corres-the decision and one copy of the article, pondence and reviewers' comments will betogether with any reviewers' comments, will kept.

586

Correspondence

tional exposures of the study cohort. Forexample, heavy metalworking industrialconcentrations are found in the UnitedStates in counties around cities such asCleveland4 or Detroit5; petroleum andpetrochemical industry is dense in certaincounties of Texas or Louisiana. These con-siderations support the use of a nationalreference population.

In describing powerful multivariatemodelling methods, which perform internalcomparisons while incorporating externalpopulation experience, the authors citedPoisson and Cox regression examples. Notmentioned was modelling mortality (or inci-dence) odds with logistic regression, in wellenumerated cohorts, which also insertsexpected mortality (incidence) odds froman external populationA" Like Poisson andCox regression, this approach estimatesrelative (but not absolute) rates for expo-sure effects, and has some notable advan-tages. Unlike Poisson or Cox methods,complete information is required only fordecedents (or incident cases). For example,date of birth or race, retrievable from deathcertificates or cancer registries, are often notavailable for occupational cohorts orgeneral populations. Restriction of workhistory retrieval to the decedent (or incidentcancer) subpopulation is a major advantage,compared with traditional cohort designsthat may require enormous efforts.5 Themethod is actually a case-control designwith the controls (for mortality study) con-sisting of all decedents with causes of deaththought to be unrelated to exposures understudy. The advantages and disadvantages ofthe use of deceased controls has beendebated without clear resolution.12- 5 In anincidence study with a cancer registry,controls would be all incident cancers atselected other sites.The actual computation of mortality

(incidence) odds by logistic regression isrelatively simple compared with Poisson orCox regression on the full population atrisk. It can readily accommodate complexrisk factor specifications, including cumula-tive exposures adjusted for latency, demo-graphic dependencies of the healthy workereffect, and variation in mortality odds withemployment duration independent ofexposures.12The concern of Callas et al on the appro-

priateness of multiplicative and non-linearexposure-response structures imposed bylogistic (and Poisson) regression can beassessed by goodness of fit. In severalstudies, logistic models performed well.910There is, moreover, some biological basisfor this finding. For example, multistagecarcinogenesis implies a higher than lineardependence in time (or exposure duration)that an exponential function can reasonablyapproximate over the ranges of effectsfound.

R M PARKHealth and Safety Department,

International Union UAW,Solidanty House,

8000 East Jefferson Avenue,Detroit, MI, USA.

3

4

Callas PW, Pastides H, Hosmer DW. Surveyof methods and statistical models used inthe analysis of occupational cohort studies.Occup Environ Med 1994;51:649-55.

Park RM, Maizlish NA, Punnett L, Moure-Eraso R, Silverstein MA. A comparison ofPMRs and SMRs as estimators of occupa-tional mortality. Epidemiology 1991;2:49-59.

Cole P, Delzell E, Acquavella J. Reply to letterto the editor. Epidemiology 1993;4:559-60.

Rotimi C, Austin H, Delzell E, Day C,

Maculso M, Honda Y. Retrospective followup study of foundry and engine plantworkers. Am IndMed 1993;24:485-98.

5 Eisen EA, Tolbert PE, Monson RR, Smith TJ.Mortality studies of machining fluid expo-sure in the automobile industry I: a stan-dardized mortality ratio analysis. Am J IndMed 1992;22:809-24.

6 Butler WJ, Park RM. Use of the logisticregression model for the analysis of propor-tionate mortality data. Am Y Epidemiol 1987;125:515-23.

7 Robins JM, Blevins D. Analysis of proportion-ate mortality data using logistic regressionmodels. AmJ Epidemiol 1987;125:524-35.

8 Breslow NE, Day NE. Statistical methods incancer research. Vol II-The design andanalysis of cohort studies. Oxford: Inter-national Agency for Research on Cancer(Oxford University Press), 1987:154.

9 Silverstein MA, Park RM, Marmor M,Maizlish N, Mirer F. Mortality among bear-ing plant workers exposed to metalworkingfluids and abrasives. Y Occup Med 1988;30:706-14.

10 Park RM, Silverstein MA, Green MA, MirerFE. Brain cancer mortality at a manufactur-er of aerospace electromechanical systems.Amy Ind Med 1990;17:537-52.

11 Park RM, Krebs JG, Mirer FE. Mortality atan automobile stamping and assembly com-plex. Am Y Ind Med 1994;26:449-63.

12 Gordis L. Should dead cases be matched todead controls? Am Y Epidemiol 1982;115:1-5.

13 Howe GR. Using dead controls to adjust forconfounders in case-control studies. Am YEpidemiol 1991;134:689-90.

14 McLaughlin JK, Blot WJ, Mehl ES, MandelJS. Problems in the use of dead controls incase-control studies. I. General Results. AmJ Epidemiol 1985;121:131-9.

15 McLaughlin JK, Blot WJ, Mehl ES, MandelJS. Problems in the use of dead controls incase-control studies. II. Effect of excludingcertain causes of death. Am J Epidemiol1985;122:485-94.

Offspring sex ratios as an index ofpollution hazard in residential environ-ments.

Author's reply-Your correspondent sug-gests that the negative findings for exposureto generalised air pollution on the sex ratiosof births reported in our paper may concealpositive, but opposing, effects of the pollu-tion on male and female parents.' Headvances the analogy with the changes inthe sex ratio of the offspring of parents withmultiple sclerosis, when the direction of thechange of sex ratio depends on the sex ofthe parent sufferer. It is indeed possible thatsuch a cryptic scenario might result fromexposure to general air pollution and ananswer to this hypothesis might well comethrough follow up studies as he has suggest-ed. That approach, however, would requireextremely expensive ad hoc studies to gener-ate the data. The main purpose of ourinvestigation was to ascertain whether thesex ratio was a sensitive barometer of expo-sure of generalised environmental pollutionin residential communities. Whether or notthe sex ratios of offspring in exposedparents differ was really beyond the designof this investigation. Our conclusions there-fore are unaltered: the routinely availabledata on sex ratios of births do not betray thepresence of potential toxins in generalindustrial air pollution in the same way aswas apparent for air pollution from specificindustrial processes.

FLR WILLIAMSOL LLOYD

SA OGSTONNinewells Hospital and Medical School,

Dundee DDI 9SY

1 James WH. Offspring sex ratios as an index ofpollution hazard in residential environ-ments. Occup Environ Med 1995;52:556.

NOTICES

NOTICES

Fibres, particles, and the lung: Newperspectives. 11-12 September 1995.Edinburgh Conference Centre, HeriotWatt University, Edinburgh.

The British Association for Lung Research(BALR) was founded in 1981, and is thepremier UK organisation for workersresearching into medical and non-medicalaspects of the lung.The last 5 years has seen exciting newinsights into the effects of particles on thelung.Fibres-The publication of the RCC studydata; the completion of the first phase of theColt Programme; the increasing importanceof bio-persistence.Environmental particles-Association withdisease; the role of particle size.Overload-The mechanisms; impact on theinterpretation of inhalation toxicology stud-ies.Key note speakers: Dr Gunter OberdorsterRochester, New York; Dr Brooke MossmanBurlington, Vermont; Dr Tom HesterbergDenver, Colorado.Colt fibre programme reports: Dr JMGDavis; Dr A Searl; Dr A Jones; Dr KDonaldson.Open sessions: Fibres; Particles; Generallung research.Young scientist competition: this session isopen to younger BALR members. The prizewill enable the winner to travel to a majorscientific meeting of their choice.The BALR summer meeting is sponsored

by The Colt Foundation.For further information contact: Dr R

Cullen, Institute of Occupational Medicine,8 Roxburgh Place, Edinburgh EH8 9SU.Tel 0131 447 8460, Fax 0131 447 2822.

Royal Society of Health NationalConference-Caring for the workingpopulation. 10 October 1995. TheSociety of Chemical Industry, 14-15Belgrave Square, London.

Aims and objectives:* To consider the need for occupational

health services and workplace health andsurveillance systems.

* Raise awareness about occupationaldiseases, injuries and disability; theirimpact on individuals, families, busi-nesses and society; and the system forcompensating victims.

* Discuss recent initiatives in industry andthe NHS aimed at promoting the healthand the wellbeing of working people andmeeting the occupational health needsof under served working groups.

Speakers and topics include:Occupational health services in economic-ally stringent times-Dr MichaelBaxendine, Director of OccupationalHealth Services, United Biscuits.

622

Book reviews

Are occupational health services neededtoday?-Dr Ewen MacDonald, Faculty ofOccupational Medicine, University ofGlasgow.New epidemics in occupational health-Simon Pickvance, Sheffield OccupationalHealth Project.Occupational disease and injury and statecompensation-Dr Alistair Brooks, BenefitsAgency Medical Services.UK Health and safety surveillance sys-tems-Nigel Bryson, Director of Healthand Safety, GMB.Promoting healthy lifestyles-Jane Huntley,Lead Project manager, Health at Work inthe NHS.Sharing and promoting best practices-S Young, The Wellness Forum.Fit for work-Peter Brown MBE, FitnessConsultant, British Gas plc.Health at work in the NHS-The HealthEducation Authority.Occupational health services provided byNHS trusts Santosh Mehta, BusinessManager-Nursing, Hillingdon HospitalNHS Trust.Occupational health and safety projects inprimary care-Jill Barlow, LiverpoolOccupational Health Project.

All Enquiries: Anne Faichney,Conference Department, The Royal Societyof Health, RSH House, 38A St George'sDrive, London SW1V 4BH. Tel: 0171-6300121. Fax: 0171-976 6847.

First European Forum of qualityimprovement in health care, QEIIConference Centre, London. 7, 8, and9 March 1996

This first European Forum will allow theexchange of ideas on quality improvementin health care and provide education. Theforum will consist of plenary lectures,parallel seminars and workshops, anddiscussions and short educational courses.The themes of the first forum are.* The fundamentals of continuous qualityimprovement

* Achieving patient orientation* Leadership and managing organizationalchange

* Improved quality and reducing costs* The importance of measurement* Involving everybody in quality improve-ment

* Professional education for quality* The politics of quality.

For further information contact: ClareMoloney. BMA Conference Unit, BMAHouse, Tavistock Square, LondonWC1H 9.1P. Fax: 0171 383 6663;Tel: 0171 383 6478.

BOOK REVIEWS

Book review editor: R L Maynard

If you wish to order, or require furtherinformation regarding the titles reviewedhere, please write or telephone the BMJBookshop, PO Box 295, London WX1H9TE. Tel: 0171 383 6244. Fax: 0171 3836662. Books are supplied post free in theUK and for British Forces Posted Overseasaddresses. Overseas customers should add15% for postage and packing. Payment can

be made by cheque in sterling drawn on a

UK bank, or by credit card (MasterCard,VISA, or American Express) stating cardnumber, expiry date, and your full name.(The price and availability are occasionallysubject to revision by the Publishers.)

Occupational Health DecennialSupplement. Edited by FRANCES DREVER.

(price C29). London: HMSO. ISBN 0-11-691618-4.

For over 150 years, data on occupationalmortality have been published in theRegistrar Generals' decennial supplements.These publications have proved to be a richsource of information for occupationalhealth researchers and a starting point formany hypotheses about work and health.They have, however, been rather drydocuments with lots of tables, and more

recently, supplemented by the option ofacquiring computer disks for further data.This is hardly the stuff of general interest tothe occupational health professional.The latest publication is a different

matter altogether. Firstly, it is a joint effortbetween the Office of Population, Census,and Survey (OPCS) and the Health andSafety Executive (HSE). Secondly, it nowincludes many more data sources than justmortality including the census, cancer regis-trations, the labour force survey, generalhousehold survey as well as, HSE andDepartment of Social Security informationon occupational illness and sicknessabsence. Thirdly, it is written up as chapterson various topics with some prominencegiven to the MRC Unit in Southamptonand, finally the whole has been extensivelyreviewed by others before final publication.The result is a highly readable whole with a

broad scope to cover diseases and occupa-tions but in the context of the changingstructure of United Kingdom industry. Thisis no longer a publication of interest to a

limited number of epidemiologists. This isnow a volume which no self respectingoccupational health professional can affordnot to have viewed.The volume extends to nearly 400 pages

and is divided into 13 chapters. As well as

the "traditional" sections on male occupa-tional mortality by disease and occupation,an interesting chapter deals specifically withfemale mortality. Cancer incidence issimilarly described and there are specificchapters devoted to asbestos relateddiseases, occupational injuries, the influenceof smoking and drinking on occupation,occupational sickness absence, fertility, as

well as a final chapter bravely attempting to

seek international comparisons. This last

chapter shows clearly the inadequacies ofcoverage in many other developed countriesin an area where Britain still seems to leadthe way.

It would be otiose in such a review to tryand list all the interesting-even fascinat-ing-findings from such a comprehensiveaccount of the "Occupational Health of theNation" but a few might serve to whet theprospective reader's appetite. Dramaticshifts in the demographic, employment, andoccupational patterns have occurred in thepast 20 years with greater service and lessheavy industrial sectors. Part time work isincreasing with more women entering thelabour force. The effects of these changeson the pattern of occupational mortalityand morbidity will inevitably follow and theimplications of this for occupational healthservices will need to be closely watched.

For mortality, accidental injury, asbestosand coal mine dust remain the most impor-tant factors. For coal mining, the geographi-cal pattern for excess risks of chronicbronchitis and emphysema are remarkablyuniform, which is in contrast with thepatterns for pneumoconiosis. Previoussupplements had noted an excess of pul-monary deaths among welders but this sup-plement has found that this excess can benarrowed to pneumococcal and unspecifiedlobar pneumonia. Another new finding isthe excess of oesophageal cancer in farmersin Hereford, Worcester, and Kent. Is this alink with cider consumption and does thiscorrelate with similar excesses for thistumour in the Calvados region of France?Suicide rates are also high for farmers and asimilar excess is found in their wives.

For cancer incidence, new findingsinclude increased risks of bladder cancer inplastic goods makers, prostate cancer indental technicians, and acute myeloidleukaemia in biological scientists. Forasbestos related disease, the analysis hadfound the recently well publicised risingexcess for mesothelioma with the greatestrisk apparently associated with the cohortborn in the early 1940s.

Another area of interest will be the chap-ter on occupational mortality studies. Itmay be a surprise to some, but OPCS havefacilitated over 130 hypothesis generatingand testing exercises in recent years in stud-ies ranging from small groups such as 150people at Liverpool University to thousandsof workers in the rubber industry orengaged in occupations involved with ionis-ing radiation.The writing throughout is clear and suc-

cinct. Each chapter is a good informativeread whether or not you intend to engage inepidemiological research yourself. Withoutdoubt, this volume is the most comprehen-sive survey of work and health ever to bepublished in the country. It is a pity that itspublication was not accompanied by greaterpublicity as it contains much of generalinterest to the public at large. The socialtrends documents always receive goodmedia coverage yet this fascinating volumehas been largely overlooked by television,radio, and the national newspapers. Anopportunity to show the general populationthe importance of occupational health hasbeen missed.

Nevertheless, for the readers of theOccupatwinal and Environmental Medicine,this is one book that cannot afford to beoverlooked, and, at £29, we should allseriously consider owning a copy.

J M HARRINGTON

623