gms | German Medical Science

50. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie (gmds)
12. Jahrestagung der Deutschen Arbeitsgemeinschaft für Epidemiologie (dae)

Deutsche Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie
Deutsche Arbeitsgemeinschaft für Epidemiologie

12. bis 15.09.2005, Freiburg im Breisgau

Smoking and cancer in Africa: Estimation of lung cancer deaths in selected countries based on smoking prevalence data.

Meeting Abstract

Suche in Medline nach

  • Volker Winkler - Ruprecht-Karls-Universität Heidelberg, Heidelberg
  • A. Gbangou - Ruprecht-Karls-Universität Heidelberg, Heidelberg
  • B. Kouyate - Ruprecht-Karls-Universität Heidelberg, Heidelberg
  • H. Becher - Ruprecht-Karls-Universität Heidelberg, Heidelberg

Deutsche Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie. Deutsche Arbeitsgemeinschaft für Epidemiologie. 50. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie (gmds), 12. Jahrestagung der Deutschen Arbeitsgemeinschaft für Epidemiologie. Freiburg im Breisgau, 12.-15.09.2005. Düsseldorf, Köln: German Medical Science; 2005. Doc05gmds155

Die elektronische Version dieses Artikels ist vollständig und ist verfügbar unter:

Veröffentlicht: 8. September 2005

© 2005 Winkler et al.
Dieser Artikel ist ein Open Access-Artikel und steht unter den Creative Commons Lizenzbedingungen ( Er darf vervielfältigt, verbreitet und öffentlich zugänglich gemacht werden, vorausgesetzt dass Autor und Quelle genannt werden.




Tobacco is the major cause of cancer and a relevant risk factor for several other chronic diseases. Due to the epidemiological transition in developing countries (increasing life expectancy) there is an increasing relevance of chronic diseases to the total burden of diseases. Furthermore, there are indications for a rise in the consumption of tobacco products, particularly cigarettes, in developing countries. Since in Africa data on smoking prevalence are scarce, and cancer diagnosis is often imprecise or missing, it is difficult to estimate the current and future number of cancer cases attributable to tobacco smoking. In this paper, we present an approach to estimate the effects of smoking on lung cancer in selected developing countries in Africa.

Material and Methods

A survey on smoking behaviour in a household survey at the Nouna Demographic Surveillance Area in Burkina Faso was carried out. To obtain estimates on yearly number of lung cancer cases for rural Burkina Faso, data from this survey were combined with (i) estimates on age-specific lung cancer rates in smokers and non-smokers from industrialized countries and (ii) data on age- and sex distribution in Burkina Faso and other African countries. More specific, let λjk be the age-specific lung cancer death rate in age group j and country k. Assuming a simple dichotomization of the population into smokers and non-smokers with proportions psk and 1-psk = p0k , respectively, let λjk0 and λjks be the age-specific death rate among smokers and non-smokers, we get:

λ jk = λ jk0 × p 0k + λ jks × p sk

Assuming a relative risk R=10 of smoking, we get:

λ jk = λ jk0 × p 0k + λ jk0 × R × p sk = λ jk0 × (1-p sk ) + λ jk0 × R × p s

and thus

λ jk0 = λ jk / [1- (10-1) p sk ]

These rates can be estimated for countries where both smoking prevalence data and lung cancer rates are available.

We use these to estimate the respective rates in those countries k for which we have sufficiently data on age-specific smoking prevalence only, but not on lung cancer rates. In a final step, we estimate the absolute number of lung cancer cases death in country κ per year, nκ, as obtained by

Equation 1

when Njκ is the absolute population number in age group j in country κ. These numbers were obtained from the World factbook [1] (The index κ denotes countries where lung cancer rates are not available; the index for sex is omitted for ease of presentation).


We obtained data on lung cancer rates and smoking prevalence from Germany, Australia, UK, USA and Japan [2]. The estimated lung cancer mortality rates in non-smokers based on data from different industrialised countries are nearly identical (with the exception of Japan) and are in general for males about twice as high as for females. Possible explanation for this is the higher exposure to other lung cancer carcinogens, mainly from occupational exposure. The estimated age-specific lung cancer mortality in non-smokers ranges from about 10 per 100000 at age 50 up to 150 at age 80 ( 8 to 70 in females, respectively) (see figure 1 [Fig. 1]).

The results of the smoking survey in Burkina Faso show that smoking is uncommon in the female population, and up to 33% of males, depending on age, report regular smoking. If the prevalence in Burkina Faso will remain on this level, we estimate age-specific lung cancer rates similar to those in Germany of over 200/100.000 per year for age groups over 60, given the current age distribution, and a yearly number of about 1000 male lung cancer deaths per year. Similar results (relative to total population size) are obtained from three other African countries, Chad, Tanzania and Zimbabwe (see table 1 [Tab. 1]). Under the further simplifying assumption that these smoking prevalences are in a similar range in developing countries overall, and using the same population estimates as in Ezzati & Lopez (2003) [3] of 1.32 and 1.36 billion in males and females, respectively, we estimate the yearly number of lung cancer deaths to be 200000 in males and 50000 in females. This is in a similar order of magnitude as previous estimates based on a different methodology by Ezzati & Lopez [3].


Despite the relatively small proportion of adults aged 50 and more which is the age when most cancer occur, there is an appreciable number of deaths from lung cancer that could be prevented when smoking prevalence could be reduced. It is both reassuring and alarming that the results are largely in line with other estimates based on an indirect method from national health statistics as suggested by Peto et al. [4] and applied by Ezzati and Lopez [3]. Efforts for smoking prevalence reduction are much more needed in the developing world.


The World Factbook; accessed 03-28-04
World Health Organisation (WHO), accessed 03-23-04
Ezzati M., Lopez A.D. (2003): Estimates of global mortality attributable to smoking in 2000. Lancet. 362:847-52.
Peto R., Lopez A.D., Boreham J., Thun M., Heath C. (1992): Mortality from tobacco in developed countries: indirect estimation from national vital statistics. Lancet. 339:1268-78.