BACKGROUND: Tobacco use and body mass are major risk factors for many cancers. Despite this, very little is known about their combined effect on cancer mortality. These relationships are virtually unexplored in populations having patterns of both tobacco use and body habitus atypical of those typically enrolled in epidemiologic studies. METHODS: A prospective cohort study of 148,173 men and women aged > or =35 years was conducted in Mumbai, India. Subjects were recruited during 1991-1997, measured for a variety of risk factors, including tobacco use and anthropometry, and then followed for approximately 5-6 years. RESULTS: During 774,129 person-years of follow up, a total of 796 cancer deaths were observed. Tobacco use, especially smoking in men, was associated with particularly high risk of death in extreme categories of body mass. At highest risk were underweight smoking males [hazard ratio (HR)=9.45, 5.87, and 5.75 for those smokers who were extremely thin (<16.0kg/m(2)), very thin (16.0 to <17.0kg/m(2)), or thin (17.0 to <18.5), respectively]. Significant effects of underweight among never and smokeless tobacco users disappeared with exclusion of individuals with < or =2 years of follow up. Extremely thin (<16.0kg/m(2)) women smokeless tobacco users had an elevation in risk, HR=2.95, that actually appeared to increase (to 3.21) with exclusion of individuals who were diagnosed within 2 years of follow up. CONCLUSION: Tobacco use and undernutrition are known to be serious problems in developing countries. The current study underlines the strikingly elevated risk of cancer when they occur together.
BACKGROUND:Tobacco use and body mass are major risk factors for many cancers. Despite this, very little is known about their combined effect on cancer mortality. These relationships are virtually unexplored in populations having patterns of both tobacco use and body habitus atypical of those typically enrolled in epidemiologic studies. METHODS: A prospective cohort study of 148,173 men and women aged > or =35 years was conducted in Mumbai, India. Subjects were recruited during 1991-1997, measured for a variety of risk factors, including tobacco use and anthropometry, and then followed for approximately 5-6 years. RESULTS: During 774,129 person-years of follow up, a total of 796 cancer deaths were observed. Tobacco use, especially smoking in men, was associated with particularly high risk of death in extreme categories of body mass. At highest risk were underweight smoking males [hazard ratio (HR)=9.45, 5.87, and 5.75 for those smokers who were extremely thin (<16.0kg/m(2)), very thin (16.0 to <17.0kg/m(2)), or thin (17.0 to <18.5), respectively]. Significant effects of underweight among never and smokeless tobacco users disappeared with exclusion of individuals with < or =2 years of follow up. Extremely thin (<16.0kg/m(2)) women smokeless tobacco users had an elevation in risk, HR=2.95, that actually appeared to increase (to 3.21) with exclusion of individuals who were diagnosed within 2 years of follow up. CONCLUSION:Tobacco use and undernutrition are known to be serious problems in developing countries. The current study underlines the strikingly elevated risk of cancer when they occur together.