Abstract
Objectives
The present study was undertaken to determine the prevalence and demographic correlates of tobacco use in the adult population.
Method
The sample selection was based on stratified random sampling technique selecting a total of 1635 persons from different parts of Gorakhpur district of Uttar Pradesh, India. The demographic details of participants along with prevalence of tobacco use were recorded.
Results
The prevalence of tobacco use among the subpopulation studied was 31.1% for males and 6.1 for females. Both smoking and chewing tobacco were found to be associated with various demographic parameters. The rural population showed a higher prevalence of tobacco use among both males and females; the male smoking prevalence was higher in the urban population (23.0%) than its rural counterpart (18.1%). Also the prevalence of tobacco use was directly proportionate to age increasing upto the age of approximately 60 years, then declined. Stepwise Regression analysis showed gender as the strongest predictor for smoking followed by area of residence, education and age. Whereas education was the most significant predictors for chewing tobacco, followed by gender and age.
Conclusion
The results of this study clearly indicate an association between tobacco prevalence and various demographic factors. To reduce the cancer burden, intervention measures to control tobacco use should specifically target those socio-demographic predictors of the subpopulation which are more susceptible to tobacco consumption.
Keywords: Tobacco use among general public, Prevalence of smoking, Predictors of tobacco use
Oral cancer in almost 80% of cases is attributable to tobacco (smoking and chewing) and excessive alcohol use and it ranks number one in terms of incidence among men and third among women.1 The World Health Organization predicts that tobacco deaths in India may exceed 1.5 million annually by 2020.2 A wide geographical variation in the incidence for oral cancer is found which is related to difference in exposure to known etiological factors such as tobacco and betel nut chewing, smoking and alcohol consumption. Also, there is great discrepancy in the pattern of tobacco consumption across various demographic groups of society.3 In India, the use of smokeless tobacco is more prevalent than other forms of tobacco use; resulting in an alarming number of people being affected by oral cancer.4 Among the smoked form of tobacco use, bidi, cigarette and hukkas are the popular modes and the smokeless tobacco is usually taken in the form of gutka, khaini, mishri, zarda, mawa, pan masala and many other forms which vary according to specific region in the country. The adverse effect of tobacco resulting in human disease and mortality is clear regardless of the form it may be consumed in.3,5
Understanding the demographic predictors of specific regions for tobacco smoking and chewing are important; if interventional measurers are to be undertaken to control tobacco use in the society. Hence, a cross-sectional study was planned on the general population of Gorakhpur district in order to determine the predictors and prevalence of tobacco use.
1. Materials and methods
1.1. Method
The target population for this study included individuals of 15 year age and above from different parts of Gorakhpur district, a north eastern part of Uttar Pradesh of India; also a region where tobacco use is apparently very high. The study was conducted during the period of one month in June 2012. The sample selection was based on stratified random sampling technique selecting a total of 1635 persons; belonging to different strata of society. The study procedure was approved by the institutional review board and its ethical committee. Appropriate consent was taken from the selected participants prior to data collection. The collected data was analyzed using SPSS software to assess the prevalence of tobacco use in both smoking and chewing forms and associated with other confounding variables.
1.2. Measure
The consenting participants were then given a self-administered questionnaire comprising of two sections, the first section recorded the demographic details including information regarding the age, sex, education, marital status, occupation and address and the second section recorded the respondent's tobacco use behavior. The data for tobacco consumption was recorded by means of only two questions ‘do you smoke tobacco or did you smoke tobacco in past 30 days’ and ‘do you chew tobacco or did you chew tobacco in past 30 days’. Tobacco smoking and chewing included bidi, cigarettes, cigar, hukka, pan masala, mishri, gutka, khaini and other local forms of tobacco use which were explained to the respondent. Respondents were categorized as ‘smokers’ and ‘tobacco chewers’ respectively if answered ‘yes’ to the mentioned questions. Responses were scored in accordance with defined rules.
1.3. Statistical analyses
The demographic details and the scores were recorded, data was analyzed with SPSS 20.0 version software for windows. The scores were treated statistically and descriptive analyses was performed using frequency for categorical variation. Bivariate analysis was carried out in terms of odds ratio and confidence interval to identify the differences in the prevalence of tobacco habits across the demographic groups based on gender, age and level of education and to identify the possible demographic factors associated with use of tobacco in both chewing and smoking form. Significance was established at p < 0.05. Multivariate analysis of data by stepwise multiple regression analysis (SWRA) was done to tease out the significant demographic predictors for both smoking and chewing tobacco.
2. Results
2.1. Demographic pattern
The final sample consisted of 1635 respondents. Smoking and chewing (smokeless) tobacco use was assessed by gender (71.9% male and 28.1% female), age (41% in age range of 15–24 years, 23.5% in 25–39 years age group, 20.6% in 40–59 years age group and 14.9% in 60 plus years group), area of residence (urban 58.8% and rural 41.2% subjects) and education (5% illiterate, primary education 5.7%, higher secondary level 23.7% and graduates 65.6%)
2.2. Prevalence of tobacco use
Table 1 depicts the prevalence of smoking and chewing (smokeless) tobacco respectively. There is a significant difference in the male and female tobacco use both in the smoking and chewing form. In contrast to 31.1% male population using tobacco; only 6.1% of female prevalence is seen (including both the rural and urban population). Both smoking and chewing tobacco is higher among the male population with a 20.9% and 23.7% respective prevalence rate in comparison to the 6.1% and 4.7% prevalence rate among females. Fewer females are habitués than males for each of these habits.
Table 1.
Tobacco users (smokers and chewers) % |
Smokers (%) |
Chewers (%) |
|||||||
---|---|---|---|---|---|---|---|---|---|
Urban | Rural | Total | Urban | Rural | Total | Urban | Rural | Total | |
Male | 29.8 | 32.9 | 31.1 | 23.0 | 18.1 | 20.9 | 20.7 | 28.0 | 23.7 |
Female | 5.4 | 7.2 | 6.1 | 3.9 | 6.1 | 4.8 | 3.2 | 6.8 | 4.7 |
2.3. Habit prevalence and demographic variation
Tables 2 and 3 show the prevalence along with unadjusted odds for smoking and chewing tobacco for different demographic variables. An increase in both the smoking and chewing form of tobacco is seen with age for the male habitués upto the age of 59, after which a significant decline is observed in prevalence of tobacco use. Male in the 40–59 age group had 2.46 times higher unadjusted odds (OR) than the 15–24 age group for chewing, which was higher in comparison to the smoking prevalence (OR 1.33). No such pattern was observed for the female habitués, although a statistically significant increase in tobacco consumption was observed in all the age groups as compared to the 15–24 age group, but the 25–39 year age group showed a maximum prevalence for both smoking (OR 7.99) and chewing (OR 7.60) tobacco.
Table 2.
Variable | Males |
Females |
||
---|---|---|---|---|
Smoking prevalence | Odds ratio (95% CI) | Smoking prevalence | Odds ratio (95% CI) | |
Age | ||||
15–24 | 21.5% | Referent | 1.90% | Referent |
25–39 | 22.4% | 1.05 (0.85–1.30) | 13.4% | 7.99 (4.89–13.02)∗∗ |
40–59 | 26.7% | 1.33 (1.08–1.63)∗ | 06.8% | 3.77 (2.25–6.31)∗∗ |
60 plus | 10.6% | 0.43 (0.34–0.56)∗∗ | 12.5% | 7.37 (4.51–12.05)∗∗ |
Residence | ||||
Rural | 18.1% | Referent | 06.1% | Referent |
Urban | 23.0% | 1.35 (1.09–1.68)∗ | 03.9% | 0.63 (0.41–0.94)∗ |
Education | ||||
Graduate | 19.5% | Referent | 02.6% | Referent |
High school | 19.6% | 1.01 (0.81–1.26) | 3.8% | 1.48 (0.89–2.45) |
Primary | 41.5% | 2.93 (2.39–3.58)∗∗ | 25% | 12.49 (8.25–18.91)∗∗ |
Illiterate | 28.6% | 1.63 (1.34–2.04)∗∗ | 20.0% | 9.37 (6.16–14.24)∗∗ |
*p < 0.05; **p < 0.001. CI, confidence intervals; OR, odds ratios.
Table 3.
Variable | Males |
Females |
||
---|---|---|---|---|
Chewing prevalence | Odds ratio (95% CI) | Chewing prevalence | Odds ratio (95% CI) | |
Age | ||||
15–24 | 19.0% | Referent | 02.2% | Referent |
25–39 | 26.7% | 1.55 (1.26–1.91)∗∗ | 14.6% | 7.60 (4.81–12.01)∗∗ |
40–59 | 36.6% | 2.46 (2.01–3.02)∗∗ | 04.5% | 2.09 (1.25–3.51)∗ |
60 plus | 11.0% | 0.53 (0.41–0.68)∗∗ | 12.5% | 6.35 (4.00–10.08)∗∗ |
Residence | ||||
Rural | 28.0% | Referent | 07.2% | Referent |
Urban | 20.6% | 0.67 (0.54–0.82)∗∗ | 03.2% | 0.43 (0.28–0.65)∗∗ |
Education | ||||
Graduate | 19.4% | Referent | 03.2% | Referent |
High School | 20.7% | 1.08 (0.87–1.35) | 03.8% | 1.19 (0.74–1.93) |
Primary | 58.5% | 5.86 (4.79–7.16)∗∗ | 14.3% | 5.05 (3.40–7.49)∗∗ |
Illiterate | 57.1% | 5.53 (4.52–6.76)∗∗ | 24.1% | 9.61 (6.56–14.06)∗∗ |
*p < 0.05; **p < 0.001. CI, confidence intervals; OR, odds ratios.
Smoking prevalence for males in the urban population was found to be significantly higher with 1.35 times higher unadjusted odd compared to their rural counterpart. In contrast tobacco use prevalence for females was observed to be lesser in the urban population (OR 0.63 for smoking and 0.43 for chewing). Chewing prevalence was significantly higher for males belonging to the rural sector in comparison to the urban population (OR 0.67).
The number of years of education seemed to have a detrimental effect on tobacco use among the male and female population; men with only primary education had the highest prevalence for both smoking and chewing tobacco with unadjusted odds 2.93 and 5.86 times more respectively than among male with graduate level of education. Although males with no education at all had a slightly lower prevalence than those with primary education, but smoking and chewing prevalence both were significantly higher (OR 1.63 and 5.53) than for male graduates, showing a strong gradient between tobacco use and education level. Whereas among the female population the unadjusted odds were 12 times and 9.37 times higher respectively for those with only primary education and no education for smoking and 5.05 and 9.61 times for chewing.
2.4. Multivariate analysis
It is clear from the regression results (Tables 4 and 5) that smoking is strongly predicted by gender (β=0.137, R2 = 0.019) followed by area of residence, education and age. Gender alone accounted for 1.9% of the variance and with all of the predictors 3.8% of the variance was accounted for.
Table 4.
Criterion variables for smoking |
|||||
---|---|---|---|---|---|
Predictor variable | R | R2 | R2 change | Beta | F |
Gender | 0.137 | 0.019 | 0.019 | 0.137 | 31.264∗∗ |
Residence | 0.155 | 0.024 | 0.005 | 0.139 | 20.024∗∗ |
Education | 0.173 | 0.030 | 0.006 | 0.130 | 16.850∗∗ |
Age | 0.196 | 0.038 | 0.008 | 0.163 | 16.306∗∗ |
**p < 0.001.
Table 5.
Predictor variables | Criterion variable for chewing |
||||
---|---|---|---|---|---|
R | R2 | R2 change | Beta | F | |
Education | 0.198 | 0.039 | 0.039 | 0.198 | 66.403∗∗ |
Gender | 0.214 | 0.046 | 0.007 | 0.188 | 39.003∗∗ |
Age | 0.225 | 0.051 | 0.005 | 0.218 | 29.089∗∗ |
**p < 0.001.
Similarly the regression results for tobacco chewing show that education is the single best predictor followed by gender and age. More specifically education contributes positively (β = 0.198, R2 = 0.039). Moreover independently education has predicted maximum positivity with 3.9% of the variance and the combined contribution of variance is 5.1% of all 3 factors.
3. Discussion
3.1. Prevalence of habits
This study reports tobacco use in a small local population, however the estimates of tobacco consumption are comparable to previous studies. Some variation seen in prevalence may be related to difference in exposure to different chewing and smoking forms of tobacco prevalent in this specific region.
The present study clearly shows that the prevalence of tobacco use is much higher in the rural areas (31.1%) as compared to urban areas (29.8%) with chewing form of tobacco being more prevalent (23.7%) as compared to smoking (20.9%). Area of residence (urban/rural) came out as a strong predictor for smoking as seen by multivariate regression analysis. Earlier studies report a generally higher prevalence of smoking over smokeless tobacco consumption in the western regions of globe and more use of smokeless tobacco in regions of Southern Asia6,7 which was also seen in our study. A wide variation in the global prevalence of habits is seen mainly due to difference in cultural, ethnic, regional preferences and geographic factors.
A significant gender difference was seen for tobacco use with fewer females being habitués of tobacco as compared to males in our study and a similar prevalence pattern is seen almost globally. Gender was highly significant predictor for both smoking and chewing tobacco as analyzed by stepwise regression also. Smoking among females was found to be more prevalent in the urban and chewing prevalence more in rural population, which may be due to modernization and changing social habits resulting in more smoking in urban areas and increase in tobacco chewing in rural areas.8,9
The results also show that the prevalence of smoking and chewing tobacco is directly proportional to age upto a certain age (59 years, in this study); after which a decline in the habits is seen. These prevalence rates were consistent with WHO report (1997)5 where age was seen to be directly associated with tobacco use prevalence. The prevalence for both chewing and smoking was highest for the 40–59 age group for males and 25–39 age group for females.
Lesser prevalence of tobacco consumption was seen in respondents with higher level of education for both smoking and chewing tobacco. Similar results with an inverse relationship between the levels of education and the use of tobacco are seen globally.10–12 Education was the single best predictor of chewing and significant independent factor predictive of smoking as indicated by the results of regression analysis.
In conclusion this study confirms a demographic gradient for the prevalence of tobacco use; with the male, rural and lesser educated population being more susceptible to tobacco use. Tobacco control policies and awareness programs should specially target these predictors for reducing the disease burden and health care cost of the nation.
The limitation of this study is that the data completely relies on self-reporting, and some, if not many respondents are hesitant to acknowledge tobacco use, so the actual habitués might be higher than reported. Another limitation is that the tobacco prevalence data was collected via only two questions (if the respondent smoked/chewed or not), it did not distinguish between frequency and the type of product used. Also socioeconomic factors as household income and caste, religion were not taken into account which significantly contribute to the predictors of tobacco use.3,13
Conflicts of interest
All authors have none to declare.
References
- 1.Warnakulasuriya S. Global epidemiology of oral and oropharyngeal cancer. Oral Oncol. 2009;45(4–5):309–316. doi: 10.1016/j.oraloncology.2008.06.002. [DOI] [PubMed] [Google Scholar]
- 2.Rani M., Bonu S., Jha P., Nguyen S.N., Jamjoum L. Tobacco use in India: prevalence and predictors of smoking and chewing in a national cross sectional household survey. Tob Control. 2003;12:e4. doi: 10.1136/tc.12.4.e4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Subramanian S.V., Nandy S., Kelly M., Gordon D., Smith G.D. Patterns and distribution of tobacco consumption in India: a cross sectional multilevel evidence from the 1998-9 national family health survey. BMJ. 2004;328:801–806. doi: 10.1136/bmj.328.7443.801. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Gupta P.C. Tobacco control in India. Indian J Med Res. 2006;123:579–582. [PubMed] [Google Scholar]
- 5.World Health Organization . WHO; Geneva: 1997. Tobacco or Health: A Global Status Report. [Google Scholar]
- 6.Gupta P.C., Ray C.S. Smokeless tobacco and health in India and South Asia. Respirology. 2003;8:419–443. doi: 10.1046/j.1440-1843.2003.00507.x. [DOI] [PubMed] [Google Scholar]
- 7.Giovino G.A., Henningfield J.E., Tomar S.L. Epidemiology of tobacco use and dependence. Epidemiol Rev. 1995;17:48–65. doi: 10.1093/oxfordjournals.epirev.a036185. [DOI] [PubMed] [Google Scholar]
- 8.Lopez A.D., Collishaw N.E., Piha T. A descriptive model of the cigarette epidemic in industrialized countries. Tob Control. 1994;3:242–247. [Google Scholar]
- 9.Reddy K.S., Gupta P.C., editors. Report on Tobacco Control in India. Ministry of Health & Family Welfare, Government of India; Mumbai: 2004. [Google Scholar]
- 10.Jha P., Ranson M.K., Nguyen S.N., Yach D. Estimates of global and regional smoking prevalence in 1995, by age and sex. Am J Public Health. 2002;92:1002–1006. doi: 10.2105/ajph.92.6.1002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Prabhakar B., Narake S.S., Pednekar M.S. Social disparities in tobacco use in India: the roles of occupation, education and gender. Indian J Cancer. 2012;49:41–47. doi: 10.4103/0019-509X.107747. [DOI] [PubMed] [Google Scholar]
- 12.Murti P.R., Bhonsle R.B., Gupta P.C. vol. 1. Tobacco Control SAARC Edition; 1994. p. 37. (Tobacco control activities in Kerala, India). [Google Scholar]
- 13.Davey Smith G., Gordon D., Kelly M., Nandy S., Subramanian S.V. 2004. Inequalities in Health in India: The Methodological Construction of Indices and Measures: Report Submitted to UK Department for International Development. [Google Scholar]