onsumption of tobacco kills approx six million people each year moreover, it is the measure threat of disease and death [1,2]. Tobacco is the most important oral cavity and pharyngeal cancer risk factors. Approximately 90% of people with mouth cancer are tobacco users. Some 7.5% of the world's (53.9 million) deaths were attributable to tobacco use in 1998 and if same smoking patterns continue, that number will Author ? ?: National Institute of Medical Statistics, Indian Council of Medical Research, Medical Enclave, Ansari Nagar, New Delhi, India. e-mails: [email protected], [email protected] rise to 10 million deaths annually by 2030 [3]. Tobacco can be consumed both in smoke and smokeless form.

Smokeless tobacco is tobacco that is not burned; it is also known as chewing tobacco, oral tobacco, spit or spitting tobacco, dip, chew, and snuff. Harmful health effects of smokeless tobacco include: mouth, tongue, cheek, gum, and throat cancer. Smokeless tobacco also causes nicotine addiction. South Asian people consume smokeless tobacco the most. More than one third of total tobacco consumption in this region is in the form of smokeless tobacco [4][5][6]. Smokeless tobacco consumption in south Asia is a major public health threat, in India (prevalence: 18.4%), Bangladesh (32.6%), Sri Lanka (6.9%) and Nepal (6%) by the estimation of WHO [7].

India is the second most populous country and one of the world's largest producers and consumers of tobacco. Here, tobacco is available in a variety of different types and brands e.g. bidi, gutkka, khaini, pan masala, hookah, cigarettes etc [3] and also the form of consumption varies from place to place like smoke cheroot in Odisha and Andhra Pradesh, dry snuff in western part, while creamy snuff in northeast part of India. Mostly the tobacco is consumed in smokeless form in northeast states. The prevalence of tobacco consumption in India, either smoked or smokeless tobacco, in the population aged 15 year and above was 47 per cent among men and 14 per cent among women while overall prevalence was 37 percent [8,9]. Consumption of smokeless tobacco products in India is increasing rapidly [10,11], which is showing a negative effects for both male and female. As smokeless tobacco is quite famous among women, affecting their oral morbidity and perinatal health, including premature delivery, low birth weight and shortened length [12][13][14][15][16][17] In northeast states of the India, smokeless tobacco is a part of the socio cultural [18]. They have different customs, food habits, life-style, diverse ethnic groups, type and pattern of tobacco consumption as compared to the rest of the country. Research have shown that in northeast states, betel quid (55.4%), is the most popular smokeless tobacco followed by tuibur (13.1%), gul (12.0%) and khaini (9.1%), Gul and tubur are primarily used by women and recent study shows that the prevalence of smokeless tobacco in northern, eastern and northeast states is 8.4%, 31.8% and 23.8% [18][19][20]. Betel quid is a combination of betel leaf, areca nut, and slaked lime. Like other smokeless tobacco products, betel quid and gutka are known to cause Esophageal cancer, Lip cancer, Mouth cancer, Pharynx cancer, Tongue cancer. The most harmful cancercausing substances in smokeless tobacco are tobacco-specific nitrosamines (TSNAs). TSNA levels vary by product, but the higher the level the greater the cancer risk.

Table 1 shows the prevalence of smokeless tobacco, Meghalaya, Tripura and Mizoram was higher in women while in Arunachal Pradesh, Manipur, Nagaland and Sikkim males are consuming more smokeless tobacco in GATS (2009-2010), while in DLHS (2012-2013) Meghalaya, Manipur and Tripura have the highest prevalence of smokeless tobacco among both male and female in DLHS (2012-2013). Hence, the objective of this study is to examine the prevalence of smokeless tobacco consumption among male and female in northeast state, India and to study the socioeconomic demographic characteristics correlates with tobacco use in the form of smokeless tobacco only.

characteristics (result not shown). For computation of age-adjusted prevalence rate, we use 2011 census data, RGI, Govt. of India as the standard population structure. We computed the standard age proportion by dividing the age-specific census population by the total census population number and standardizing proportion sum to 1. Then, age-adjusted factors for 6 (six) age grouping (10 year intervals each) were used for computation of age-adjusted prevalence. Binary logistic regression was applied to measure the association and to check the net effect of factors on the current consumers for males and females. Variance inflation factor (VIF) of all the variables were computed to check collinearity prior to inclusion in multivariate logistic regression problem of collinearity among independent variables not found (highest VIF, 2.36). The results of logistic regression, are presented in the form of estimated odds-ratios with 95% CI. The whole analysis was performed using STATA version 13.0 with survey commands and R software. Ethical statement: This study is based on data available in public domain, therefore no ethical issue is involved. Data for this study was taken from the fourth round of the District Level Household Survey (DLHS-4) conducted during 2012-13. DLHS-4 adopted a multistage stratified systematic sampling design. Detailed information about sampling employed in this survey can be obtained from the DLHS-4 report. All seven states, namely, Arunachal Pradesh, Manipur, Meghalaya, Mizoram, Nagaland, Tripura and Sikkim separate CAB (Clinical Anthropometric and Biochemical) data files (excluding Assam) were merged together for this study. The outcome variable included in the analysis is "personal habit of age 15 and above using smokeless tobacco". Where the response was further divided into two categories like "Current Consumption" and "No Consumption" the household members was considered current consumers if they had responded that they were consuming (smokeless tobacco) and they coded as 1, while the never consumers, ex-consumers and don't know (0.3%) are coded as 0. To measure level of prevalence and association of smokeless tobacco with factors, this study used both bivariate and multivariate analyses. Chi-square test is used to check the association of the current users with selected characteristics like age, literacy, occupation etc Table 2 presents the unweighted count of sampled respondents and population estimates classified by selected socio demographic and occupational background. The estimated population of person age 15 year and above in northeast state in DLHS (2012-2013) was 143,729 where 67,930 were male and 75,799 were female respondent were taken as unit of analysis.

(70.1%) having the highest adjusted prevalence than female (55.8%). We have also found that the wide variation in adjusted and unadjusted age prevalence especially non working male and unmarried male, female. Among states, Meghalaya have the highest prevalence followed by Mizoram.

Table 4 presents odds ratio among male and female after performing logistic regression models which examine the effect of individuals household and The prevalence of smokeless tobacco consumption among male and female in our study is 69.6% and 50.8%. Present study reveals that education is significantly associated with smokeless tobacco consumption. This is consistent with observations that those with lower level of education are more likely to consume smokeless tobacco [23,24]. In this study, the age wise prevalence of smokeless tobacco consumption is higher as the age advanced and the highest rate is found in the age group of 20-34 and 35-59 years and then declined after 60 years in both the sexes, similar finding was also reported [23,25]. Those who are married have a higher rate of smokeless tobacco consumption as compared to the unmarried respondents. This may be due to influences of the spouses consuming smokeless tobacco. Similar association between smokeless tobacco consumption and marital status was also reported [25].

In northeast states, India smokeless tobacco consumption is strongly associated with the level of education, religion, caste, marital status, occupation and place of residence. A comprehensive ban on tobacco advertising, promotion and sponsorship needs to be implemented according to the standard outlined in 'Article 13' in the WHO Framework Convention on Tobacco Control. Display and visibility of smokeless tobacco products at points of sale constitutes advertising and promotion and should therefore be banned [22]. In addition to proper enforcement of the new law, there is a need for a nationwide campaign educating people in both rural and urban areas about the law and health risks of smokeless tobacco.

The authors gratefully acknowledge members of the study field team including who were involved in mapping/listing and main survey team during data collection in northeast states. The authors also acknowledge all the respondent for their active participation in this study.