• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of amjphAmerican Journal of Public Health Web SiteAmerican Public Health Association Web SiteSubmissionsSubscriptionsAbout Us
Am J Public Health. 2002 June; 92(6): 1002–1006.
PMCID: PMC1447501

Estimates of Global and Regional Smoking Prevalence in 1995, by Age and Sex

Prabhat Jha, MD, DPhil, M. Kent Ranson, MD, MPH, Son N. Nguyen, MD, MPH, and Derek Yach, MB, ChB, MPH


Objectives. We calculated regional and sex- and age-specific smoking prevalence estimates worldwide in 1995.

Methods. Sex-specific smoking prevalence data from studies in 139 countries and age distribution data from 7 studies were analyzed.

Results. Globally, 29% of persons aged 15 years or older were regular smokers in 1995. Four fifths of the world's 1.1 billion smokers lived in low- or middle-income countries. East Asian countries accounted for a disproportionately high percentage (38%) of the world's smokers. Males accounted for four fifths of all smokers, and prevalence among males and females was highest among those aged 30 to 49 years (34%).

Conclusions. Future decades will see dramatic increases in tobacco-attributable deaths in low- and middle-income regions. Although much of this excess mortality can be prevented if smokers stop smoking, quitting remains rare in low- and middle-income countries. (Am J Public Health. 2002;92:1002–1006)

It is estimated that 4 million deaths from tobacco occurred in 1999 and that the annual number of deaths is likely to rise to 10 million by the 2030s.1 Quantification of smoking's current and future effect on global disease burden relies on local retrospective and prospective epidemiological studies2–4 or indirect estimates.5 Ideally, such studies would be supported by reliable estimates of smoking prevalence, incidence, and cessation rates. Such smoking statistics help to monitor age- and sex-specific trends and the success of control programs.

From the perspective of global tobacco control, global or regional estimates are similarly useful. In 1997, the World Health Organization6 (WHO) calculated and published global prevalence rates of smoking. Much of the data in that review relied on smoking prevalence data from the late 1980s or earlier.

We estimated the sex- and age-specific prevalence of tobacco smoking globally and regionally for 1995. These values were used in a subsequent analysis to assess the costeffectiveness of tobacco control policies.7 This article relies chiefly on smoking prevalence data from 109 countries collected as part of the Tobacco Control Country Profiles8,9 as well as earlier data from WHO6 and the literature. We used World Bank definitions of regions, whereas WHO definitions of regions were used in the Tobacco Control Country Profiles. We excluded some of the Tobacco Control Country Profiles studies because the literature or the 1997 WHO publication (itself based on earlier surveys conducted around 1990) provided more recent estimates. 1998 was chosen as the base year for the Tobacco Control Country Profiles. We used actual 1995 population data because the bulk of the smoking surveys occurred closer to 1995 than to 1998. Our analyses attempted to estimate smoking prevalence by age category, information that has not previously been assessed at a global level.


Population by Region, Sex, and Age Category

World Bank population figures by sex and age category for each of the World Bank's 7 regions (described on the World Bank's Web site, at http://www.worldbank.org/html/extdr/regions.htm) were used. Population figures for 1995 were used throughout the analysis.10 We used World Bank regions rather than WHO or Global Burden of Disease regions11 so as to provide results comparable to recent World Bank publications on tobacco.12,13 Age categories (ages 15–19, 20–29, 30–39, 40–49, 50–59, ≥ 60) were chosen to coincide with the categories most commonly used in smoking prevalence studies.

Smoking Prevalence by Sex and Region

The results of 109 studies from the Tobacco Control Country Profiles8 plus 11 from the literature14–24 and 19 from the 1997 WHO6 review were used to estimate smoking prevalence, by sex, for each of the 7 regions. The Tobacco Control Country Profiles methods have been described elsewhere.8 For 19 countries, we chose prevalence as reported in the 1997 WHO publication, because these data were more recent.6 Whenever possible, survey data that adhered to the WHO definitions of smoking6 were selected so as to increase comparability between studies. The WHO guidelines state that respondents who report smoking at the time of the survey, or “current smokers,” should be further categorized as “daily” (i.e., those who smoke at least once per day) or “occasional” smokers (i.e., those who smoke, but not every day). Although we used daily smoking as the prevalence indicator in our analyses, exclusion of 27 low- or middle-income countries that did not provide a definition of smoking only marginally changed the overall results.

We combined country-specific data to estimate regional prevalence values by weighting country estimates by the adult population (older than 15 years) of those countries. The resulting weighted average smoking prevalence rates are assumed to apply to the entire region, including those countries for which smoking prevalence is not known. When all of the sources were combined, data on smoking prevalence from individual studies were available for more than 96% of the population of each region, except eastern and central Europe and sub-Saharan Africa. For these 2 regions, data were available for 90% and 67% of the population, respectively. The median year of data for all countries was 1996.

Prevalence of Cigarette and Bidi Smoking in South Asia

In all regions with the exception of South Asia, cigarettes constitute the major form of smoked tobacco. In the countries of South Asia, however, many people smoke bidis, a hand-rolled cigarette. All calculations for South Asia were conducted separately for cigarettes and bidis. Data from 3 studies (2 from India3,25 and 1 from Sri Lanka6) suggested that 47% to 51% of the male smokers and 52% to 95% of the female smokers smoke bidis. In this analysis, it is assumed that 50% of the male and 80% of the female smokers smoke bidis, with the remainder smoking cigarettes.

Smoking Prevalence by Age Category

We attempted to find 1 large-scale study of smoking prevalence by age category for each of the 7 regions (when possible, we used the most populous country in the region). China is used as the model country for East Asia and the Pacific,23 Russia for Europe and central Asia,15,26 Brazil for Latin America and the Caribbean,27 Saudi Arabia for the Middle East and North Africa,17 India for South Asia,25,28 Sudan for sub-Saharan Africa,24 and the United States for high income.29 We validated the chosen studies, when possible, by substituting the age weights for various other countries in the same region and found little change (e.g., substituting Poland30 and Hungary31 for Russia, substituting Argentina and Chile for Brazil,27 and substituting other high-income countries32 for the United States). Ratios of smoking prevalence among older age categories compared with the 15- to 19-year age group were calculated. These ratios were applied to the entire region, including those countries for which the age ratios of smoking prevalence are not known.


Our estimates showed variations in smoking prevalence by region, sex, and age. Globally in 1995, 29% of the population aged 15 years and older, or 1.1 billion people, smoked daily (Table 1 [triangle]). Low- and middle-income countries, whose populations constitute four fifths of the global adult population, accounted for 82% of the world's smokers. East Asia and the Pacific, which includes China, accounted for 38% (429 million) of all smokers but only 32% of the population aged 15 years and older. Overall, smoking prevalence was highest in East Asia and the Pacific and in Europe and central Asia, at 34%, and lowest in sub-Saharan Africa, at 18%.

—Estimated Smoking Prevalence, by Sex, and Number of Smokers Aged 15 Years and Older, by World Bank Region, 1995

For both males and females, wide variations in smoking prevalence were found between regions. Smoking prevalence among males was highest in East Asia and the Pacific, at 62%, and lowest in sub-Saharan Africa, at 28%. Among females, the smoking prevalence was highest in Latin America, at 22%, and lowest in South Asia, at 4% (for cigarettes and bidis combined), and in the Middle East and North Africa, at 7%. Smoking prevalence was higher overall for men (47%) than for women (11%). Globally, males account for 81% of all smokers (Table 2 [triangle]).

—Global Prevalence of Smoking and Number of Smokers, by Age and Sex, 1995

Smoking prevalence was highest for persons aged 30 to 49 years (34%, Table 2 [triangle]) and lowest among youths aged 15 to 19 years (20%). It also was relatively low among people aged 60 and older (24%). The patterns of age-specific smoking prevalence were similar for males and females, although males had about a 4- to 5-fold higher prevalence in most age groups.


Our estimates have several limitations. First, in a few of the source reports, it was not possible to distinguish between daily smoking and occasional smoking. Thus, data from the studies were combined even when they may not have been strictly comparable. In our analyses, excluding or including countries with different definitions of smoking, such as regular smokers (occasional and daily) or daily smokers, did not greatly alter the regional and global estimates (results not shown). Variability in the rates of cigarette consumption, type of cigarette used and its tar levels, and duration of smoking are also key determinants of health outcomes that were not covered in our analyses. Second, although the overall and sex-specific smoking prevalence data are likely to be valid given that they were derived from direct studies in various countries, our indirect estimates of age-specific patterns, which were based on 7 country-specific studies, are of uncertain validity. However, our age ratios were broadly consistent with ratios from the Tobacco Control Country Profiles adult smoking prevalence rates9 and from a recent global survey of smoking among adolescents aged 13 to 15 years.33

These estimates are also consistent with the 1997 WHO analysis of smoking prevalence,6 despite use of different data sources to estimate smoking prevalence for 1990, and with the Tobacco Control Country Profiles results,9 despite different base years and regional categorization. The similarity across the 3 studies is consistent with agricultural data showing that overall tobacco production has remained constant through most of the 1990s.34

Some key points are noteworthy. Low- and middle-income countries have an aggregate smoking prevalence similar to that of highincome countries, but low- and middleincome countries account for the majority (82%) of the world's smokers. Males in low-income countries have a higher prevalence of daily smoking (49%) than do males in high-income countries (37%), whereas the reverse is true for females (8% in low-income countries and 21% in high-income countries).

Our analysis is consistent with predictions that total numbers and proportion of tobaccoattributable disease burden accounted for by low-income countries will increase dramatically in future decades. WHO estimated that in 1999, about 4 million tobacco deaths occurred worldwide. About half were in China, India, Latin America and other low- or middleincome countries outside Europe, and half were in high-income countries and former socialist economies of Europe.6 Current tobacco mortality largely reflects past smoking. Per capita consumption has been higher in highincome countries than in low-income countries over the last few decades.6 However, future tobacco mortality depends largely on current smoking patterns. Peto and Lopez1 estimated, on the basis of current smoking patterns, that about 450 million cumulative tobacco deaths will have occurred by 2050, mostly in lowincome countries. Similarly, Murray and Lopez,35 using econometric models, estimated that 87% of the increase in tobacco-attributable deaths between 1990 and 2020 will occur in low-income countries. The increase in tobacco deaths is the result of both increases in the susceptible population size and increases in age-specific disease rates.

Also of importance in predicting future tobacco-attributable mortality is the prevalence of ex-smokers. Many of the future deaths expected among the 1995 cohort of smokers could be avoided if adults quit smoking. Much evidence indicates that smoking cessation reduces the risk of death from tobacco-related diseases. Among physicians in the United Kingdom, those who quit smoking before the onset of major disease avoided most of the excess hazard of smoking.36 The benefits of quitting were largest in those who quit early (between ages 35 and 44) but were still significant in those who quit later (between ages 45 and 54 years). Similar analyses suggested that cessation before middle age avoids more than 90% of the lung cancer risk attributable to tobacco.37

Percentages of former smokers are the best measure of smoking cessation at a population level. In high-income countries, percentages of former smokers have increased over the past 2 to 3 decades, and today about 30% of the male population are former smokers (Table 3 [triangle]). In contrast, in recent years, male percentages of former smokers were only 10% in Vietnam,22 5% in India,3 and 2% in China.23 Even these low figures may be falsely elevated, because they include people who quit because of illness. Time trend data are not available for low- and middle-income countries.

—Prevalence of Ex-Smokers at 2 Different Points in Time in Selected Countries, Stratified by Per Capita Gross Domestic Product

We did not estimate smoking incidence, because systematic data on age-specific initiation rates are rarely available. In most highincome countries, about 8 of every 10 smokers begin smoking in their teens.38 In middle- and low-income countries for which data are available, it appears that most smokers start by their early 20s, but the trend is toward younger ages.12 For example, in China between 1984 and 1996, the number of young men (but not women) aged 15 to 19 years who began smoking increased significantly.23 Incidence data are important for predicting longer-term trends in tobaccoattributable deaths and for monitoring the effectiveness of control policies aimed at preventing initiation. Because only those who start smoking in early adult life are at high risk of death from tobacco-related disease in middle and old age, there generally will be a delay of about half a century between the time of the main increase in smoking by young adults and the time of the main increase in death from tobacco-associated illness in later life. The implication of this time lag for control policies is that efforts to increase cessation among current smokers will yield the largest reductions in mortality over the next few decades.1


Support for this analysis came from the Economics Advisory Service of the WHO and the Human Development Network of the World Bank.

We are indebted to Frank J. Chaloupka, Vendhan Gajalakshmi, Emmanuel Guindon, Alan Lopez, and Ayda Yurekli for helpful comments on the manuscript and to Marlo Corrao for providing Tobacco Control Country Profiles data.

The views represented are those of the authors and not those of the organizations sponsoring the work.


P. Jha and M. K. Ranson designed the study. P. Jha, M. K. Ranson, and S. N. Nguyen analyzed the data. All authors contributed to the writing of the article.

Peer Reviewed


1. Peto R, Lopez AD. The future worldwide health effects of current smoking patterns. In: Koop EC, Pearson CE, Schwarz PM, eds. Global Health in the 21st Century. New York, NY: Jossey-Bass; 2000:154–161.
2. Liu BQ, Peto R, Chen ZM, et al. Emerging tobacco hazards in China, 1: retrospective proportional mortality study of one million deaths. BMJ. 1998;317:1411–1422. [PMC free article] [PubMed]
3. 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. [PMC free article] [PubMed]
4. Gupta PC, Mehta HC. Cohort study of all-cause mortality among tobacco users in Mumbai, India. Bull World Health Organ. 2000;78:877–883. [PMC free article] [PubMed]
5. Peto R, Lopez AD, Boreham J, Thun M, Heath C Jr. Mortality From Smoking in Developed Countries, 1950–2000. Oxford, England: Oxford University Press; 1994.
6. World Health Organization. Tobacco or Health: A Global Status Report. Geneva, Switzerland: World Health Organization; 1997.
7. Ranson MK, Jha P, Chaloupka FJ, Nguyen S. Global and regional estimates of the effectiveness and cost-effectiveness of price increases and other tobacco control policies. Nicotine Tob Res. In press. [PubMed]
8. Corrao MA, Guindon GE, Sharma N, Fakharbadi-Shokoohi D, eds. Tobacco Control Country Profiles. Atlanta, Ga: American Cancer Society; 2000.
9. Corrao MA, Guindon GE, Cokkinides V, Sharma N. Building the evidence base for global tobacco control. Bull World Health Organ. 2000;78:884–890. [PMC free article] [PubMed]
10. The World Bank Group. World Development Indicators 2001 [database online]. Washington, DC: The World Bank; 1997. Available at: http://www.worldbank.org/data/wdi2001/. Accessed May 21, 2001.
11. Murray CJ, Lopez AD. Mortality by cause for eight regions of the world: Global Burden of Disease Study. Lancet. 1997;349:1269–1276. [PubMed]
12. Jha P, Chaloupka FJ. Curbing the Epidemic: Governments and the Economics of Tobacco Control. Washington, DC: The World Bank; 1999.
13. Jha P, Chaloupka FJ, eds. Tobacco Control in Developing Countries. Oxford, England: Oxford University Press; 2000.
14. Balabanova D, Bobak M, McKee M. Patterns of smoking in Bulgaria. Tob Control. 1998;7:383–385. [PMC free article] [PubMed]
15. McKee M, Bobak M, Rose R, Shkolnikov V, Chenet L, Leon D. Patterns of smoking in Russia. Tob Control. 1998;7:22–26. [PMC free article] [PubMed]
16. Bobak M, Pikhart H, Hertzman C, Rose R, Marmot M. Socioeconomic factors, perceived control and self-reported health in Russia: a cross-sectional survey. Soc Sci Med. 1998;47:269–279. [PubMed]
17. Jarallah JS, al-Rubeaan KA, al-Nuaim AR, al-Ruhaily AA, Kalantan KA. Prevalence and determinants of smoking in three regions of Saudi Arabia. Tob Control. 1999;8:53–56. [PMC free article] [PubMed]
18. Hill DJ, White VM, Scollo MM. Smoking behaviours of Australian adults in 1995: trends and concerns. Med J Aust. 1998;168:209–213. [PubMed]
19. Haidinger G, Waldhoer T, Vutuc C. The prevalence of smoking in Austria. Prev Med. 1998;27:50–55. [PubMed]
20. Action on Smoking and Health. Smoking statistics. Available at: http://www.ash.org.uk. Accessed May 21, 2001.
21. Alam SE. Prevalence and pattern of smoking in Pakistan. JPMA J Pak Med Assoc. 1998;48:64–66. [PubMed]
22. Jenkins CN, Dai PX, Ngoc DH, et al. Tobacco use in Vietnam: prevalence, predictors, and the role of the transnational tobacco corporations. JAMA. 1997;277:1726–1731. [PubMed]
23. Chinese Academy of Preventive Medicine. Smoking in China: 1996 National Prevalence Survey of Smoking Pattern. Beijing: China Science and Technology Press; 1997.
24. Idris AM, Ibrahim YE, Warnakulasuriya KAAS, Cooper DJ, Johnson NW, Nilsen R. Toombak use and cigarette smoking in the Sudan: estimates of prevalence in the Nile State. Prev Med. 1998;27:597–603. [PubMed]
25. Venkat Narayan KM, Chadha SL, Hanson RL, et al. Prevalence and patterns of smoking in Delhi: cross sectional study. BMJ. 1996;312:1576–1579. [PMC free article] [PubMed]
26. Russia Longitudinal Mortality Study. Available at: http://www.cpc.unc.edu/projects/rlms/rlms_home.html. Accessed May 21, 2001.
27. Carolina Population Center, Russian Longitudinal Mortality Project, University of North Carolina at Chapel Hill, Available at: http://www.cpc.unc.edu/projects/rlms/rlms_home.html. Accessed May 21, 2001.
28. Goa Cancer Society. Assessment of the Efficacy of Anti-Tobacco Community Education Program. Dona Paula, Goa: Goa Cancer Society; 1992.
29. Cigarette smoking among adults—United States, 1997. MMWR Morb Mortal Wkly Rep. 1999;48:993–996. [PubMed]
30. Zatonski W. Evolution of Health in Poland Since 1988. Warsaw, Poland: Maria Sklodowska-Curie Cancer Center and Institute of Oncology; 1996.
31. Hungary Health Behavior Survey. Budapest: Hungary Central Statistics Office; 1994.
32. Nicolaides-Bouman A, Wald N, Forey B, Lee P. International Smoking Statistics: A Collection of Historical Data From 22 Economically Developed Countries. Oxford, England: Oxford Medical Publications; 1993.
33. Warren CW, Riley L, Asma S, et al. Tobacco use by youth: a surveillance report from the Global Youth Tobacco Survey project. Bull World Health Organ. 2000;78:868–876. [PMC free article] [PubMed]
34. USDA Economics and Statistics System Web Site. Available at: http://usda.mannlib.cornell.edu. Accessed May 21, 2001.
35. Murray CJ, Lopez AD. Alternative projections of mortality and disability by cause 1990–2020: Global Burden of Disease Study. Lancet. 1997;349:1498–1504. [PubMed]
36. Doll R, Peto R, Wheatley K, Gray R, Sutherland I. Mortality in relation to smoking: 40 years' observations on male British doctors. BMJ. 1994;309:901–911. [PMC free article] [PubMed]
37. Peto R, Darby S, Deo H, Silcocks P, Whitley E, Doll R. Smoking, smoking cessation, and lung cancer in the UK since 1950: combination of national statistics with two case-control studies. BMJ. 2000;321:323–329. [PMC free article] [PubMed]
38. Giovino GA, Schooley MW, Zhu BP, et al. Surveillance for selected tobacco-use behaviors—United States, 1900–1994. MMWR Morb Mortal Wkly Rep. 1994;43:1–43. [PubMed]
39. Hill DJ. Australian patterns of tobacco smoking in 1986. Med J Aust. 1988;149:6–10. [PubMed]
40. Hill DJ, White VM. Australian adult smoking prevalence in 1992. Aust J Public Health. 1995;19:305–308. [PubMed]
41. La Vecchia C, Pagano R, Decarli A, Ferraroni M. Smoking in Italy, 1990–1991. Tumori. 1994;80:175–180. [PubMed]
42. Pagano R, La Vecchia C, Decarli A. Smoking in Italy, 1995. Tumori. 1998;84:456–459. [PubMed]
43. Wersall JP, Eklund G. The decline of smoking among Swedish men. Int J Epidemiol. 1998;27:20–26. [PubMed]
44. del Rio MC, Alvarez FJ. Patterns of smoking in Spain: results from a regional general population survey. Eur J Epidemiol. 1994;10:595–598. [PubMed]

Articles from American Journal of Public Health are provided here courtesy of American Public Health Association
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...