Résumé :
|
[BDSP. Notice produite par INIST-CNRS 8HGR0x89. Diffusion soumise à autorisation]. Background Reduction in pulmonary function, as estimated by forced expiratory volume in Is (FEV 1), has been found to predict all-cause mortality in developed-country populations. This study was designed to examine the association between FEV1 and mortality in an urban developing-country population. Methods Data from the large, well-characterized Mumbai Cohort Study (Maharashtra, India) were used to compute hazard ratios (HRs ; deaths/100-ml FEV1) and 95% confidence intervals (CIs) from Cox proportional hazards regression models in which age, tobacco use, education, height and relative body weight were controlled. Results A total of 13 261 deaths occurred in this cohort of 148173 individuals. After controlling for important covariates, there was a 1.7% reduction in risk of overall death in women for each 100-ml increment in FEV1 (HR=0.983 ; 95% CI=0.980-0.986) and a 1.5% reduction in men (HR=0.985 ; 95% CI=0.984-0.986). There was a 1.6% reduction in cancer deaths in women (HR=0.984 ; 95% CI=0.973-0.996) and a 0.8% reduction in men (HR=0.992 ; 95% CI=0.987-0.997). The largest reductions in women were observed in tuberculosis deaths (3.7%/100-ml increment in FEV1), and in men in respiratory system deaths (3.2%). Conclusions In a densely populated urban Indian population, FEV1 predicted overall and cancer mortality. Effects were larger in women and were not attenuated by exclusion of smokers or restricting analyses to subjects dying>2 years from recruitment. Because FEV1 may be affected by air pollution, which is worsening in urban areas of most developing countries, further research is recommended to deepen understanding of these factors in relation to mortality.
|