Article Text

Download PDFPDF

Work in Brief

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Many years have passed since Doll and Peto famously estimated the proportion of cancer deaths attributable to occupation. In this issue, Rushton et al 1 report on an ambitious project to update estimates for Great Britain. Attributable fractions and numbers were estimated for mortality and incidence of 6 cancers related to occupations and agents classified as IARC group I and IIA carcinogens (strong or suggestive evidence of cancer at specific anatomical sites in humans). Risk estimates and estimated proportions exposed came from a literature review and national data sources. The authors estimate that some 4.9% of deaths were attributable to work-related carcinogens in 2004 and 4% of incident cancers. Asbestos contributed more than half of attributable deaths. Other leading agents include silica, diesel engine exhaust, radon and environmental tobacco smoke. Estimates for most of the cancers were greater than currently assumed in UK health and safety planning, although not inconsistent with the earlier estimates by Doll et al.


Temporary employment has been linked with job insecurity, adverse psychosocial and work circumstances, and a potential for occupational injury. Tompa et al 2 have investigated the relation of employment contract and job tenure to work-related spells of sickness absence. In a cohort of almost 115,000 person-job-months of observation, from the Canadian Survey of Labour and Income Dynamics, individuals with a short job tenure (⩽ 6 months) were substantially less likely to take a work-related sickness absence spell. After allowing for this and for prior health status, no major differences in risk of sickness absence were found in relation to a temporary contract of employment (in contrast to some earlier reports). No relation was found with size of employing company either; but work-related sickness absence spells were more likely among workers represented by a trades union.


Spells of long-term sickness absence have been linked with future mortality risk. However, the predictive capacity of this metric warrants further exploration. Kivimäki et al 3 have now conducted a prospective survey of some 18,000 workers from the National Gas and Electricity Company, France (the Gazel cohort). Doctor-certified absence was determined over a 3-year period from employers’ registers and related to death registrations. Analyses separately considered workers with various types of chronic health problem. Taking more than 28 days of sick leave in a year (vs. none) was a predictor of all-cause mortality in those with cancer, depression, chronic bronchitis or asthma and hypertension, with hazard ratios increased 1.7–5.4 fold. The contrast of taking 5 spells of >14 days absence per 10 years of observation (vs. none) showed a similar strength of association. Areas under receiver operating characteristics curves for these absence measures varied between 0.56 and 0.73, suggesting that simple metrics of sickness absence carry useful prognostic information.


This month’s issue includes a report from the WHO World Mental Health Survey Initiative, estimating the scale of attention-deficit/hyperactivity disorder among workers from 10 European countries;4 a study of mesothelioma risk and exposure to tremolite-free chrysotile;5 and a survey of incident and recurrent back injuries in a dynamic cohort of US carpenters.6