While multiplicative (log-linear and logistic) models have a firmly established place in epidemiologic methodology, additive and other more general model structures are needed also. The authors propose a parametric family of relative risk functions ranging from subadditive to supramultiplicative that is generated by varying the exponent in a power transform for the log relative risk. The choice of model is facilitated by graphic analysis of goodness-of-fit statistics computed for various values of the exponent. Intermediate quantities available as by-products of the fit are useful for checking the influence of particular observations on the estimated regression coefficients. Three examples illustrate the applications of these methods to random, stratified, and matched samples of cases and controls. Computer software is available for each of these situations. Even though different relative risk models may have markedly different implications for the multifactorial nature of the disease process, it may be difficult to distinguish between them unless the data are quite extensive.
There are epidemiological studies indicating that exposure to metal fumes is a risk factor for infectious pneumonia. Whether occupational exposure to other agents, such as inorganic dust or chemicals, also increases the risk for infectious pneumonia is not clear. The aim of the present study was to elucidate whether occupational exposure to respiratory pollutants and irritants increases the risk for infectious pneumonia.
Prospective cohort study. Setting Swedish male construction workers. Participants 320,143 male construction workers exposed to inorganic dust (asbestos, man-made mineral fibres, dust from cement, concrete and quartz), wood dust, metal fumes and chemicals (organic solvents, diisocyanates and epoxi resins) or unexposed. Main outcome measures The cohort was followed from 1971 to 2003 and the main outcome measures were mortality to infectious pneumonia, lobar pneumonia or pneumococcal pneumonia. RRs were obtained by the person-years method and from Poisson regression models, adjusting for baseline values of age and smoking habits.
Among men aged 20-64 years there was increased mortality from infectious pneumonias among construction workers exposed to metal fumes (RR 2.31, 95% CI 1.35 to 3.95), inorganic dust (RR 1.87, 95% CI 1.22 to 2.87) and chemicals (RR 1.91, 95% CI 1.37 to 3.22). The mortality was also increased from both lobar pneumonia and pneumococcal pneumonia. Among men aged 65-84 years the occupational exposure to inorganic dust and chemicals was associated with slightly increased mortality from infectious pneumonia. Among groups with mutually exclusive exposures there was increased mortality from infectious pneumonias among construction workers exposed to inorganic dust, but not among those exposed to wood dust or chemicals. There were no cases among workers exposed only to metal fumes.
Our findings indicate that exposure to inorganic dust increases the mortality from infectious pneumonias, especially lobar pneumonia and pneumococcal pneumonia. The mechanism is unclear, but the effect may be mediated through induced airways inflammation.
Among a cohort of 544 men with at least 20 years of employment in chrysotile mining and milling at Thetford Mines, Canada, 16% of the deaths were from lung cancer and 15% from asbestosis. The excess over expected deaths from these causes account for 43 of 178 deaths in the group. The risk of death of asbestosis, at equal times fron onset of exposure, is very similar in miners and millers, factory workmen and insulators. The ratio of observed to expected deaths from lung cancer is similar in the miners and millers and factory workers, but higher in insulators. The risk of death of mesothelioma in miners and millers is decidedly less than the other two groups. The exact causes of the reduced risk in this category are not yet completely clarified.
Mortality among 535 asbestos-exposed and 205 nonexposed employees of an asbestos-cement factory was investigated. In the period beyond 20 yr from first exposure, the exposed workers had standardized mortality ratios of 175 for all causes of death, 370 for all malignancies, 480 for lung cancer, 240 for gastrointestinal cancers, and 17 deaths from mesothelioma; the factory control subjects had mortality rates similar to the general population. The cell-type distribution of the lung cancers was similar to that occurring in middle-aged smokers. Cumulative fiber exposures were calculated for the production workers, and mortality rates for the asbestos-associated malignancies were found to have significant trends with exposure. Exposure-related lung cancer risks were noted, with a large margin of uncertainty, to be similar to those observed in an American study of manmade mineral fiber workers.
Practical constraints may limit both cohort selection and the degree of follow-up in longitudinal studies of occupational mortality. For example, it may be possible to study only those men employed on a certain date or to trace (through pension fund records) only men employed at or close to normal retirement age or to follow men only for a very limited period after they have left employment. In an examination of the effects of such restrictions, advantage was taken of the availability of detailed work histories and the full follow-up of a large birth cohort of chrysotile asbestos workers. Age-matched case-referent analyses were performed under each type of restriction. All three analyses revealed the clear exposure-response relationship between cumulative dust exposure and lung cancer that was found in the full analysis. Most other findings were also not inconsistent with the earlier results. Such restricted cohort studies, with case-referent approaches to analysis can be justified in logic. The findings provide some empirical support for these methodological variants and suggest that they may be useful when complete follow-up of an entire birth cohort is not feasible.