Objectives. We assessed social disparities in the prevalence of overall tobacco use, smoking, and smokeless tobacco use in Mumbai, India, by examining occupation-, education-, and gender-specific patterns.

Methods. Data were derived from a cross-sectional survey conducted between 1992 and 1994 as the baseline for the Mumbai Cohort Study (n=81837).

Results. Odds ratios (ORs) for overall tobacco use according to education level (after adjustment for age and occupation) showed a strong gradient; risks were higher among illiterate participants (male OR = 7.38, female OR = 20.95) than among college educated participants. After age and education had been controlled, odds of tobacco use were also significant according to occupation; unskilled male workers (OR = 1.66), male service workers (OR = 1.32), and unemployed individuals (male OR = 1.84, female OR = 1.95) were more at risk than professionals. The steepest education- and occupation-specific gradients were observed among male bidi smokers and female smokeless tobacco users.

Conclusions. The results of this study indicate that education and occupation have important simultaneous and independent relationships with tobacco use that require attention from policymakers and researchers alike.

Tobacco use in low-income and middle-income countries is predicted to contribute to an increasing share of the global burden of disease in future decades.1 Eighty-two percent of the world’s 1.1 billion smokers now reside in low- and middle-income countries, where, in contrast to declining consumption in high-income countries, tobacco consumption is on the rise.1 Indeed, the World Health Organization’s Framework Convention on Tobacco Control underscores the importance of tobacco control efforts within developing countries as part of a worldwide strategy to reduce the health, economic, and social consequences of tobacco use.2 Addressing this growing public health problem requires attention to increasing social disparities in patterns of tobacco use. Across high-, middle-, and low-income countries, smoking rates are highest among individuals of low socioeconomic position.3

Indicators of socioeconomic position vary across studies; often education, occupation, and income level are used interchangeably to measure socioeconomic position.4 It is important, however, to examine multiple indicators of socioeconomic position simultaneously if one is to understand their combined impact and thereby provide more complete descriptions of social inequalities in tobacco use. In particular, insufficient attention has been focused on occupational disparities in tobacco use, given the role of occupation in linking education and income as well as its role as a determinant of health in its own right, through hazardous workplace exposures. Indeed, recent analyses of US data indicate that education does not represent a “stand-in” surrogate for occupation, or vice versa; rather, they reflect distinct social constructs making overlapping as well as independent contributions to patterns of tobacco use.5

In this study, we examined social disparities in tobacco use in India, where multiple forms of tobacco consumption complicate attempts to reduce its overall impact on public health. It has been estimated that 65% of men use some form of tobacco, including 35% who smoke, 22% who use smokeless tobacco, and 8% who engage in both forms of tobacco use.6,7 About one third of women use at least one form of tobacco, although rates among women vary considerably by region (from approximately 15% to approximately 65%).6,7 In general, cigarettes account for an estimated 20% of tobacco consumption; about 50% of tobacco is consumed in the form of bidis, that is, traditional, leaf-wrapped unfiltered cigarettes.8,9

In previous studies, different patterns have been observed in the educational gradient in tobacco use depending on the type of tobacco used. Whereas overall tobacco use has been shown to be highest among those with the least education, cigarette smoking rates have been shown to increase with increasing education.10 In India, because of their low cost, bidis are more commonly smoked than cigarettes by individuals of lower socioeconomic position; in turn, cigarettes are more commonly consumed among those with greater financial resources.10,11 (Bidi smoking has been shown to pose significant health hazards.1214) A similar socioeconomic gradient has been observed for the use of smokeless tobacco, including chewing tobacco, snuff, burnt tobacco, powder, and paste.7,15

In general, men in India smoke as well as chew or apply tobacco, whereas women generally chew or apply tobacco, with the exception of the few areas where prevalence rates of smoking among women are high.7,16 It is estimated that more than 150 million men and 44 million women in India use tobacco in various forms,14 and approximately 635000 deaths in India are attributed to tobacco each year. Tobacco-related cancers constitute about half of the total cancer incidence among men and about 20% among women.8

The purpose of this study was to assess educational and occupational differences in the prevalence of tobacco use, including total tobacco use, bidi and cigarette smoking, and smokeless tobacco use, in a large sample of residents of Mumbai, India. In addition, we sought to assess the joint effects of occupation and education level on tobacco use after controlling for other key determinants of use (i.e., gender and age).

Baseline data for the Mumbai Cohort Study were collected between 1992 and 1994 in Mumbai (Bombay), India.17 The overall purpose of this prospective cohort study was to assess mortality associated with tobacco use in Mumbai.

Study Population

Mumbai is a large, densely populated city whose population was approximately 12 million people in 2001.18 The city is divided into 3 sectors: the main city, the suburbs, and the extended suburbs. This study exclusively focused on the main city. The sampling frame comprised the city’s electoral rolls, which are updated via house-to-house visits before each major election. From these rolls, assumed to be relatively complete given that almost all adult residents are entitled to vote, data were derived on the name, age, gender, and address of all individuals older than 18 years. The electoral rolls were organized by geographical areas; sampling was based on the smallest unit, the “polling station,” which included 1000 to 1500 eligible voters. Selection of polling stations excluded those involving a large proportion of apartment complexes with high levels of security; results of the pilot data collection indicated the need for this exclusion owing to the difficulty of gaining access to such buildings.

At the selected polling stations, all individuals 35 years or older who were listed on the electoral rolls were eligible to be interviewed. The age cutoff of 35 years was selected as a result of the study’s overall goal of studying tobacco-attributed mortality. In selected geographical areas, lists were supplemented to include individuals who were not listed on the electoral rolls but whose residence status was confirmed by a “ration card.” These cards, issued by the Bombay Municipal Corporation, serve as a proxy for residence cards and permit access to all city and state governmental services; individuals identified in this manner represented approximately 5% of the overall sample.

Of the individuals approached and invited to participate in the study, the nonresponse rate was less than 1%. It was not possible to contact approximately 50% of the individuals included on the lists as a result of incomplete addresses, houses being demolished, changes of residence, and inaccessibility of residences (often owing to security considerations). A total of 99 598 adults (40 071 men and 59 527 women) were recruited and surveyed. In the analyses presented here, we excluded respondents who reported that they were retired (n = 15 223) or had missing data for occupation (n = 2538). The final sample comprised 81 837 respondents.

Data Collection

The survey was conducted by trained interviewers within participants’ households. Hand-held computers were used to record data at the time of the interview. Interviews were conducted in the local languages, including Hindi and Marathi. No surrogate responses were permitted.


The primary outcome in the present analyses was tobacco use, categorized as follows: (1) having no habit in either the past or present (“never user”), (2) former user (including smoking and use of smokeless tobacco), (3) current smokeless tobacco user (including betel quid, mishri, and creamy snuff), (4) current cigarette smoker, and (5) current bidi smoker (including other forms of smoked tobacco as well, e.g., chilum and hooka). Smokers who also used smokeless tobacco were classified as smokers in these analyses.

Occupation was assessed according to respondents’ self-reports. Following the standard Indian classification system, occupations were coded as follows: skilled workers, unskilled workers, traders, service workers, and professionals.19 Additional categories included unemployed and housewife. Women were considered as housewives unless they were currently employed or looking for employment. Retirees were excluded from the analyses. Education level was classified as illiterate, primary school (up to 5 years of education), middle school (6–8 years of education), secondary school (9–12 years of education), and college (including both some college and attainment of college degree). Gender and age data were also collected.

Data Analysis

Descriptive statistics were calculated for the overall population as well as for men and women separately. Logistic regression was used in conducting multivariate analyses. The response variable, tobacco use, was converted into a dichotomous variable in which current tobacco users (including users of any form of tobacco) were compared with current nonusers. Multivariate analyses of cigarette and bidi smoking were conducted only among men because of the extremely low prevalence (less than 0.5%) of smoking among women. SPSS statistical software (SPSS Inc, Chicago, Ill) was used in analyzing the data.

Sample Characteristics

Men represented about one third of the sample (Table 1). More than 40% of men were employed in service positions, and one third were unskilled workers, whereas a large majority (88%) of women were classified as housewives. Women were generally less educated than men; 45% of women were illiterate, as compared with 11% of men. In addition, only 5% of women had completed secondary school or college, whereas 16% of men had done so. Overall, about a quarter of the participants were between the ages of 35 and 39 years; more than a third were between 40 and 49 years of age.

Tobacco Use Prevalence: Bivariate Analyses

Patterns of tobacco use differed dramatically according to gender (Table 1). While women were less likely than men to have ever used tobacco (26% vs 41%), they were more likely to currently use smokeless tobacco (57% vs 44%). Smoking prevalence rates were 27% among men and, as mentioned, less than 0.5% among women (thus, data on female smokers are not shown separately in Table 1 or described in subsequent analyses). Among male smokers, 12% were cigarette smokers and 15% were bidi smokers. Overall, 2% of the sample members were former tobacco users, an indicator of cessation rates.

Among men as well as women, professionals were least likely to have ever used tobacco, whereas unskilled workers and unemployed individuals were most likely to have done so. Use of smokeless tobacco was more common than smoking across all occupational categories. Rates of smokeless tobacco use among women were highest among unskilled workers, those who were unemployed, and housewives. Among men, smokeless tobacco use was especially prevalent among service and unskilled workers and unemployed individuals. Bidi smoking among men followed a similar pattern, with high prevalence rates among unemployed individuals and unskilled workers. In contrast, cigarette smoking was most common among professionals and traders. Self-reported rates of former tobacco use ranged from less than 2% to 6%.

There was a strong gradient in tobacco use according to education level. Among both men and women, the rate of smokeless tobacco was highest among the illiterate and lowest among those with a college education. Among men, the prevalence of bidi smoking was highest among those at low levels of education, but the prevalence of cigarette smoking was highest among those at the highest education levels.

Multivariate Analyses

Table 2 presents gender-specific tobacco use odds ratios comparing current tobacco users, current cigarette smokers, current bidi smokers, and current smokeless tobacco users with individuals reporting no current use of any type of tobacco. Odds ratios according to occupation and education were adjusted for age and the other relevant model variable (i.e., either occupation or education). The reference category for occupation was professional, and the reference category for education was college.

Tobacco use was inversely related to education level across all types of tobacco use. The magnitudes of the odds ratios were especially large among those with no more than a primary school education; in addition, in this subgroup, odds ratios were particularly pronounced among women who used smokeless tobacco and men who were bidi smokers. Relative to participants in the reference educational category (college), odds ratios for all forms of tobacco use were significantly higher among those in the other educational categories. After adjusting for age and education, we also observed an inverse relationship between cigarette smoking and education (see Table 2).

Although the magnitudes of the relationships were not as large, occupation continued to play an important role in patterns of tobacco use when education and age were controlled. In the case of men, odds ratios for smokeless tobacco use remained statistically significant among unskilled workers, service workers, and unemployed individuals, and the odds ratios for bidi smoking remained significant among unemployed individuals and both skilled and unskilled workers. None of the odds ratios for cigarette smoking were significant. After education level had been controlled, male traders were actually less likely to use smokeless tobacco than were professionals, suggesting an interesting interaction between education and occupation. Among women, after control for education level and age, only the odds ratios for those who were unemployed remained statistically significant.

The present results demonstrate the important roles of education and occupation in tobacco use patterns in India. Research in the West has consistently documented a strong socioeconomic gradient in tobacco use, with higher rates of use among those of greater social disadvantage.4,5,2022 Indeed, Jarvis and Wardle23 concluded that, in Western countries, “any marker of disadvantage that can be envisaged and measured, whether personal, material or cultural, is likely to have an independent association with cigarette smoking.” Recent evidence documents the same socioeconomic tobacco use gradient in India; tobacco use has been found to be higher among individuals at lower levels of education,10,11,15,2427 of lower castes,15,27 and with lower standards of living.27,28 (Other research, however, has failed to reveal an association between tobacco use and socioeconomic position.29)

Education is a powerful correlate of tobacco use patterns.10 In this study, after adjustment for occupation and age, all forms of tobacco use followed an inverse linear pattern in terms of educational level; similar results have been reported by others.11,15,27 Odds ratios were alarmingly high among individuals with no more than a primary school education, particularly, as described earlier, women using smokeless tobacco and men smoking bidis. Of note, when we adjusted only for age (data not shown), the direction of the relationship between education and cigarette smoking among men was reversed relative to the bivariate relationships presented in Table 1. Unlike the use of other forms of tobacco, cigarette smoking was most prevalent among the younger groups within this sample; among male participants, age contributed significantly to both education- and occupation-specific odds of cigarette smoking. These findings underscore the importance of adjusting for age in analyses such as those described here.

Our analyses also offer evidence of the independent effects of occupation and education on tobacco use among men; even after control for education, odds ratios for occupation were statistically significant among the most disadvantaged workers in regard to bidi smoking and use of smokeless tobacco. One interesting exception in these occupation-specific results involved the odds of using smokeless tobacco among male traders; although the overall prevalence of smokeless tobacco use was somewhat higher among traders than among professionals, a lower proportion of traders than professionals in each of the various educational groups used smokeless tobacco (data not shown).

Occupation appeared to carry more weight in regard to men’s tobacco use than that of women. Because a large proportion of the women in this sample were housewives and 45% were illiterate, it is not surprising that education was a more important indicator of socioeconomic position than current occupation. The “housewife” category provided insufficient information to adequately describe socioeconomic position because it included women living in a range of social and economic circumstances. In addition, education appeared to swamp any influence of occupation among women; for example, the odds of smokeless tobacco use were more than 20 times greater among women who were illiterate than among women with a college education.

Unemployment was a particularly powerful predictor of tobacco use. In the case of all comparisons, even those taking education into account, unemployed individuals were at the highest risk of using tobacco, a relationship that has been reported in other populations as well.3034 In addition, unemployment was most strongly associated with bidi use among men (OR = 3.5). Unemployment is an indicator of increased economic disadvantage and associated stressors such as poor housing conditions, unmet needs for food, and potential lack of social connectedness.23,35 Expenditures on tobacco products have been found to represent a significant portion of the daily incomes of Indian residents in low income categories, including unemployed individuals.36

The present findings demonstrate the need, in studies assessing social disparities in tobacco use, to examine occupation and education separately as well as simultaneously. This will allow researchers to gain a more complete understanding of such disparities than might be the case when considering either indicator alone.5 Others have noted the importance of considering multiple indicators of socioeconomic position in understanding patterns of tobacco use.5,23,37 Education and occupation are likely to operate through differing pathways. Education is one of the most widely used indicators of socioeconomic position, given that it is easy to measure, applicable to individuals both inside and outside the labor force, and stable across the life course. It has consistently been shown to be a strong correlate of tobacco use, both in India and elsewhere.5,10,11,15,22,2426 Nonetheless, it may fail to capture some of the elements of socioeconomic position expressed by occupation; occupation may further indicate one’s standing in the community, reveal aspects of the normative environment prevalent within one’s occupational “culture,” and serve as a marker for the general conditions present at one’s workplace.5,37

Several caveats must be noted in interpreting our results. For example, our education and occupation data were based on self-reports. In addition, the complexities of obtaining, recording, and coding occupational data can lead to misclassification.3740 Furthermore, our occupational categories were combined into broad groupings, which could have contributed to biased estimates in terms of the gradients observed. Nonetheless, these groupings provided greater precision than those used in earlier tobacco use research in India; in these studies, occupation was grouped into even more general categories.41 We collected data at the individual level, not the household level, and thus our data on socioeconomic position may have been incomplete, particularly in the case of women.37 Future studies could include other indicators of socioeconomic position, such as caste or different standard of living measures.

In addition, as described earlier, the present data were collected as part of the initial data collection effort in a prospective cohort study; they were not part of a surveillance study designed to assess population prevalence rates of tobacco use. The sample was not a random or representative sample of the population. In particular, we excluded individuals who resided in upper-middle-class and upper-class housing complexes that were not accessible as a result of security issues. Thus, the proportions of individuals in different occupational categories might not have been comparable to the proportions in other cities or in India as a whole. Nonetheless, our findings provide important insight into the interrelationships between education, occupation, and tobacco use. Moreover, although the proportions of different occupation types and the prevalence rates of tobacco use may not have been representative of the general population, it is highly unlikely that the interrelationships observed would have been seriously affected by our sampling methods.

Identifying occupation- and education-specific disparities in tobacco use can provide a useful “signpost” indicating inequities that need to be addressed by policymakers and the broader community through allocation of resources.42 Our results indicate that tobacco use in India follows a social gradient mirroring that reported for Western countries. If one is to shed light on patterns of disparities, it is important to consider multiple indicators of socioeconomic position, including both education and occupation, as well as gender. Additional research elucidating the differing pathways by which occupation and education may influence tobacco use can inform future policies and other interventions.

TABLE 1— Tobacco Use, by Gender, Occupation, Education, and Age: Mumbai Cohort Study
TABLE 1— Tobacco Use, by Gender, Occupation, Education, and Age: Mumbai Cohort Study
 Total, No. (%)Never Users, %Former Users, %Current Smokeless Tobacco Users, %Current Cigarette Smokers, %Current Bidi Smokers,b %Total, No. (%)Never Users, %Former Users,%Current Smokeless Tobacco Users, %
    Professional422 (1.6)48.16.426.316.42.8293 (0.5)72.01.426.6
    Trader2 620 (9.7)37.43.429.216.113.9265 (0.5)
    Service11 605 (42.8)27.61.748.912.39.61 721 (3.1)50.21.348.2
    Skilled2 000 (7.4)28.22.738.912.517.8336 (0.6)
    Unskilled8 835 (32.6)21.62.444.611.420.13 796 (6.9)35.02.361.8
    Unemployed1 659 (6.1)18.53.647.011.419.6131 (0.2)28.22.366.4
    Housewife48 154 (88.0)
    None/illiterate3 090 (11.4)11.91.747.07.232.324 678 (45.1)25.82.371.3
    Primary10 090 (37.2) 773 (36.2)
    Middle9 519 (35.1)31.81.842.814.09.67 358 (13.5)60.11.937.9
    Secondary2 765 (10.2)33.34.440.515.36.52 070 (3.8)74.42.323.1
    College1 677 (6.2)53.14.422.917.32.4817 (1.5)89.10.710.0
Age, y
    35–397 697 (28.4)30.11.542.913.312.113 920 (25.4)
    40–446 447 (23.8)28.61.942.813.813.09 530 (17.4)45.11.453.2
    45–494 962 (18.3)25.01.646.012.315.18 516 (15.6)
    50–544 036 (14.9)23.42.848.211.014.67 364 (13.5)34.62.362.6
    55–592 554 (9.4)20.83.946.610.418.35 300 (9.7)32.72.764.2
    60–64780 (2.9)19.56.840.57.925.34 783 (8.7)
    65–69384 (1.4) 609 (4.8)
    ≥ 70281 (1.0) 674 (4.9)
    Total27 141 (100.0)26.42.344.412.414.554 696 (100.0)

a This group was not categorized separately, because very few women were smokers (less than 0.5%).

b Includes all current smokers who were not exclusively current cigarette smokers.

TABLE 2— Adjusted Odds Ratios (and 95% Confidence Intervals) for Various Forms of Tobacco Use (vs No Current Habit), by Education, Occupation, and Gender: Mumbai Cohort Study
TABLE 2— Adjusted Odds Ratios (and 95% Confidence Intervals) for Various Forms of Tobacco Use (vs No Current Habit), by Education, Occupation, and Gender: Mumbai Cohort Study
 Current Tobacco UsersCurrent Cigarette SmokersCurrent Bidi SmokersaCurrent Smokeless Tobacco Users
        Women20.95 (16.60, 26.45)21.02 (16.63, 26.56)
        Men7.38 (6.36, 8.56)1.73 (1.39, 2.16)38.64 (27.38, 54.54)7.75 (6.55, 9.18)
        Women9.12 (7.22, 11.51)9.18 (7.27, 11.60)
        Men4.48 (4.00, 5.02)1.65 (1.41, 1.93)17.31 (12.46, 24.07)5.25 (4.59, 6.01)
        Women5.45 (4.31, 6.90)5.50 (4.34, 6.97)
        Men2.42 (2.16, 2.70)1.35 (1.16, 1.57)5.92 (4.26, 8.25)2.90 (2.54, 3.31)
        Women2.69 (2.09, 3.45)2.70 (2.10, 3.48)
        Men2.00 (1.76, 2.27)1.28 (1.07, 1.53)3.12 (2.18, 4.46)2.45 (2.11, 2.85)
        Women0.93 (0.64, 1.34)0.92 (0.64, 1.34)
        Men1.26 (1.00, 1.58)1.04 (0.75, 1.43)2.51 (1.34, 4.71)1.19 (0.91, 1.56)
        Women1.26 (0.93, 1.70)1.24 (0.92, 1.68)
        Men1.66 (1.34, 2.06)1.22 (0.91, 1.63)3.29 (1.77, 6.10)1.65 (1.29, 2.12)
        Women1.05 (0.71, 1.55)1.01 (0.68, 1.50)
        Men0.85 (0.68, 1.06)1.05 (0.77, 1.42)1.53 (0.82, 2.86)0.68 (0.52, 0.88)
        Women1.00 (0.73, 1.36)0.99 (0.73, 1.35)
        Men1.32 (1.07, 1.63)1.08 (0.81, 1.44)1.39 (0.75, 2.59)1.48 (1.16, 1.90)
        Women1.95 (1.18, 3.21)1.89 (1.15, 3.12)
        Men1.84 (1.45, 2.33)1.34 (0.95, 1.87)3.48 (1.84, 6.58)1.79 (1.36, 2.36)
        Women1.04 (0.77, 1.39)1.03 (0.77, 1.38)

aIncludes all current smokers who were not exclusively current cigarette smokers.


1. Gajalakshmi CK, Jha P, Ranson K, Nguyen S. Global patterns of smoking and smoking-attributable mortality. In: Jha P, Chaloupka FJ, eds. Tobacco Control in Developing Countries. New York, NY: Oxford University Press Inc; 2000:11–39. Google Scholar
2. Framework Convention on Tobacco Control. Geneva, Switzerland: World Health Organization; 2003. Google Scholar
3. Jha P, Chaloupka FJ. Curbing the Epidemic: Governments and the Economics of Tobacco Control. Washington, DC: World Bank; 1999. Google Scholar
4. Bobak M, Jha P, Nguyen S, Jarvis M. Poverty and smoking, In: Jha P, Chaloupka FJ, eds. Tobacco Control in Developing Countries. New York, NY: Oxford University Press Inc; 2000:41–61. Google Scholar
5. Barbeau E, Krieger N, Soobader M. Working class matters: socioeconomic disadvantage, race/ethnicity, gender, and smoking in the National Health Interview Survey, 2000. Am J Public Health. In press. Google Scholar
6. Tobacco or Health: A Global Status Report. Country Profiles by Region. Geneva, Switzerland: World Health Organization; 1997. Google Scholar
7. Gupta PC. A Database on Tobacco in the SouthEast Asia Region. New Delhi, India: World Health Organization; 2003. Google Scholar
8. Sharma DC. India’s welcome to foreign tobacco giants prompts criticism. Lancet.1998;352:1204. Crossref, MedlineGoogle Scholar
9. Mudur G. India finalises tobacco control legislation. BMJ.2001;322:386. Crossref, MedlineGoogle Scholar
10. Gupta PC. Socio-demographic characteristics of tobacco use among 99,598 individuals in Bombay, India, using hand-held computers. Tob Control.1996;5:114–120. Crossref, MedlineGoogle Scholar
11. Narayan KM, Chadha SL, Hanson RL, et al. Prevalence and patterns of smoking in Delhi: cross sectional study. BMJ.1996;312:1576–1579. Crossref, MedlineGoogle Scholar
12. Dikshit RP, Kanhere S. Tobacco habits and risk of lung, oropharyngeal and oral cavity cancer: a population-based case-control study in Bhopal, India. Int J Epidemiol.2000;29:609–614. Crossref, MedlineGoogle Scholar
13. Wasnik KS, Ughade SN, Zodpey SP, Ingole DL. Tobacco consumption practices and risk of oro-pharyngeal cancer: a case-control study in Central India. Southeast Asian J Trop Med Public Health.1998;29:827–834. MedlineGoogle Scholar
14. Pais P, Fay MP, Yusuf S. Increased risk of acute myocardial infarction with beedi and cigarette smoking in Indians: final report on tobacco risks from a case-control study. Indian Heart J.2001;53:731–735. MedlineGoogle Scholar
15. Rani M, Bonu S, Jha P, Nguyen SN, 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. Crossref, MedlineGoogle Scholar
16. Sudarshan R, Mishra N. Gender and tobacco consumption in India. Asian J Womens Stud.1999;5:83–114. Google Scholar
17. Gupta P, Mehta HC. Cohort study of all-cause mortality among tobacco users in Mumbai, India. Bull World Health Organ.2000;78:877–883. MedlineGoogle Scholar
18. Census of India 2001: Series 28. Maharashtra: Provisional Population Totals. Maharashtra, India: Dept. of Census Operations; 2001. Google Scholar
19. National Classification of Occupations, NCO Divisions. New Delhi, India: Directorate General of Employment and Training, Ministry of Labour; 2004. Google Scholar
20. Giovino G, Pederson L, Trosclair A. The prevalence of selected cigarette smoking behaviors by occupation in the United States. In: Work, Smoking and Health: A NIOSH Scientific Workshop. Washington, DC: Centers for Disease Control and Prevention; 2000: 22–31. Google Scholar
21. Graham H. Promoting health against inequality: using research to identify targets for intervention. A case study of women and smoking. Health Educ J.1998;57:292–302. CrossrefGoogle Scholar
22. Cigarette smoking among adults: United States, 2001. MMWR Morb Mortal Wkly Rep.2001;52:40. Google Scholar
23. Jarvis MJ, Wardle J. Social patterning of individual health behaviours: the case of cigarette smoking. In Marmot M, Wilkinson RG, eds. Social Determinants of Health. Oxford, England: Oxford University Press Inc; 1999:240–255. Google Scholar
24. Gupta P. Why we should care: at-risk populations. Paper presented at: Oslo Cancer Congress, June–July 2002, Oslo, Norway. Google Scholar
25. Sen U. Tobacco use in Kolkata. Lifeline.2002;8:7–9. Google Scholar
26. Gajalakshmi CK, Peto R. Studies on tobacco in Chennai, India. Paper presented at: 10th World Conference on Tobacco and Health, August 1997, Beijing, China. Google Scholar
27. Subramanian SV, Nandy S, Kelly M, Gordon D, Smith GD. Patterns and distribution of tobacco consumption in India: cross sectional multilevel evidence from the 1998–1999 National Family Health Survey. BMJ.2004;328:801–806. Crossref, MedlineGoogle Scholar
28. National Family Health Survey (NFHS-2), 1998–1999: India. New Delhi, India: World Health Organization, Regional Office for South-East Asia; 2002. Google Scholar
29. Singh RB, Beegom R, Mehta AS, et al. Social class, coronary risk factors and undernutrition, a double burden of diseases, in women during transition in five Indian cities. Int J Cardiol.1999;69:139–147. Crossref, MedlineGoogle Scholar
30. Lee AJ, Crombie IK, Smith WCS, Tunstall-Pedoe HD. Cigarette smoking and employment status. Soc Sci Med.1991;33:1309–1312. Crossref, MedlineGoogle Scholar
31. Bennett N, Jarvis L, Rowlands O, Singleton N, Haselden L. Living in Britain: Results From the 1994 General Household Survey. London, England: Her Majesty’s Stationery Office; 1996. Google Scholar
32. Novo M, Hammarstrom A, Janlert U. Smoking habits: a question of trend or unemployment? A comparison of young men and women between boom and recession. Public Health.2000;114:460–463. Crossref, MedlineGoogle Scholar
33. Morrell SL, Taylor RJ, Kerr CB. Jobless: unemployment and young people’s health. Med J Aust.1998;168:236–240. MedlineGoogle Scholar
34. Hammarstrom A. Health consequences of youth unemployment: review from a gender perspective. Soc Sci Med.1994;38:699–709. Crossref, MedlineGoogle Scholar
35. Kaplan GA. Where do shared pathways lead? Some reflections on a research agenda. Psychosom Med.1995;57:208–212. Crossref, MedlineGoogle Scholar
36. Efroymson D, FitzGerald S, eds. Tobacco and Poverty: Observations From India and Bangladesh. Mumbai, India: PATH Canda; 2003. Google Scholar
37. Krieger N, Williams DR, Moss NE. Measuring social class in U.S. public health research: concepts, methodologies, and guidelines. Annu Rev Public Health.1997;18:341–378. Crossref, MedlineGoogle Scholar
38. Levy BS, Wegman DH. Occupational Health: Recognizing and Preventing Work-Related Disease and Injury. Philadelphia, Pa: Williams & Wilkins; 2000. Google Scholar
39. History, Origins, and Conceptual Basis: National Statistics Socio-Economic Classification. London, England: Office for National Statistics; 2002. Google Scholar
40. Standard Occupational Classification (SOC) User Guide. Washington, DC: Bureau of Labor Statistics; 2003. Google Scholar
41. Gupta PC, Ray CS. The epidemic in India. In: Boyle P, Gray N, Henningford J, Seffrin J, Zatonski W, eds. Tobacco and Public Health: Science and Policy. Oxford, England: Oxford University Press Inc; in press. Google Scholar
42. Carter-Pokras O, Baquet C. What is a health disparity? Public Health Rep.2002;17:426–436. CrossrefGoogle Scholar


No related items




Glorian Sorensen, PhD, MPH, Prakash C. Gupta, DSc, FACE, and Mangesh S. Pednekar, MScGlorian Sorensen is with the Center for Community-Based Research, Dana-Farber Cancer Institute, the Department of Society, Human Development, and Health, Harvard School of Public Health, Boston, Mass. At the time of this study, Prakash C. Gupta was with the Tata Institute of Fundamental Research, Mumbai, India; Mangesh S. Pednekar was with the Tata Memorial Centre, Mumbai, India. “Social Disparities in Tobacco Use in Mumbai, India: The Roles of Occupation, Education, and Gender”, American Journal of Public Health 95, no. 6 (June 1, 2005): pp. 1003-1008.


PMID: 15914825