Representative of the Director General; World Health Organisation; Geneva, Switzerland (Yach)Stop TB Department; World Health Organisation; Geneva, Switzerland (Raviglione)
Checking for direct PDF access through Ovid
BACKGROUNDSmoking cigarettes and bidis is widespread among men in India. It is not known to what extent smoking contributes to the major causes of adult death in India: vascular disease, pulmonary tuberculosis, and other respiratory disease.OBJECTIVETo determine the effects of smoking on mortality in men.SETTINGTamil Nadu, India: city of Chennai (urban), 1995 and 1997; district of Vilippuram (rural area), 1997 and 1998.METHODCase-control study.PARTICIPANTS27,043 urban and 15,998 rural men who died of disease during the time period selected, whose household could be visited and family members interviewed. Men dying of non-medical causes (unintentional injury, suicide, or homicide) were excluded. Controls were 20,162 urban and 15,128 rural men aged 25 or over, living in a household where a female member had died in the time period selected. Cases were not matched to individual controls.OUTCOMESSmoking habits and cause of death.ANALYSISCause of death was obtained from official death records, hospital records and by interviewing the families of the deceased. Two medical doctors independently coded the probable cause of death according to the International Classification of Diseases (ICD-9). Information regarding the deceased men's smoking habits, tobacco chewing habits, and educational status was obtained from their families. The same information was collected for the control group from living male relatives (preferably the spouse) of deceased women. The same control group was used for all comparisons. The term smoker refers to current smokers and ex-smokers. Due to the decreased reliability with which cause of death could be determined in the older population, results were calculated separately for those aged 25 to 69 years.MAIN RESULTSMale smokers, aged 25 to 69 years, were significantly more likely than non-smokers to have died of tuberculosis, other respiratory disease, vascular disease, and cancer (see Table 1). In these men, 58% of deaths from tuberculosis, and 23% of all cause deaths were attributable to smoking.AUTHORS' CONCLUSIONSSmoking bidis and cigarettes increases the incidence of clinical tuberculosis. About half the male tuberculosis deaths and a quarter of all male deaths in India can be attributed to smoking.Notes Authors note that due to the retrospective collection of data on deceased men's smoking habits and cause of death, recall bias may have been introduced. In addition, relying on families' recollection of the deceased men's symptoms, what type of hospital they were treated in, and treatment received in order to ascertain cause of death may have lead to misclassification of cause of death.