OBJECTIVE: The objective of this work was to describe the relationships between educational level, tobacco chewing, and cancer mortality in south India, among middle-aged adults who never smoked tobacco or drank alcohol, to eliminate confounding by those habits. METHODS: This case-control study was conducted in two areas of Tamil Nadu state. The cases studied were 2,580 lifelong non-smoking non-drinkers who died at age 35-69 years during 1995-1998, with interviews in 1998-2000 of a spouse, neighbour, or close associate, who retrospectively provided information on the education and chewing/other habits of the deceased. Underlying neoplastic cause of death was determined by verbal autopsy. The controls were 429,306 lifelong non-smoking non-drinkers aged 35-69 from these two study areas, interviewed during 1998-2001. RESULTS: Among the controls, prevalence of current tobacco chewing was much higher in those with less education, irrespective of sex, urban/rural residence, or birth year. Compared with never chewers, ever chewers had fivefold higher mortality from mouth cancer (odds ratio 4.9, 95% confidence interval 3.5-6.8), and 1.5 to twofold higher mortality from cancers of the pharynx/larynx/oesophagus combined, stomach, and cervix. Each of these cancers had a strong, independent, inverse association with educational level. CONCLUSION: This study supports a substantial body of evidence that tobacco chewing can cause mouth cancer, and adds to evidence that chewing may increase the risk of cancer at other sites. The analysis suggests a possible link with cervical cancer, but this could have been because of residual confounding by social factors. Avoidance of tobacco chewing would avert many cancer deaths in south India, especially for people who have received relatively little formal education.

Original publication




Journal article


Cancer Causes Control

Publication Date



23 Suppl 1


91 - 98


Alcohol Drinking, Case-Control Studies, Educational Status, Female, Humans, India, Male, Middle Aged, Neoplasms, Prevalence, Retrospective Studies, Risk Factors, Rural Population, Smoking, Socioeconomic Factors, Tobacco, Smokeless