ICRF/MRC Clinical Trial Service Unit & Epidemiological
Studies Unit (CTSU) University of Oxford Radcliffe Infirmary Oxford
OX2 6HE, U.K. zhengming.chen@ctsu.ox.ac.uk
The largest study ever undertaken to examine the health effects of tobacco
finds that there are already a million deaths a year from smoking in China,
and it predicts large increases in mortality over the next few decades. This
pattern is likely to be repeated in other developing countries.
China, with 20 percent of the world's population, produces and consumes
about 30 percent of the world's cigarettes. The first nationwide study of
the effects of smoking in any developing country now shows that China already
suffers almost a million deaths a year from tobacco1,
2 (the
full reports are freely available at http://www.bmj.com).
This is more than in any other country, and the hazards are expected to increase
substantially over the next few decades as a delayed effect of the recent
rise in cigarette use. To monitor the current and future hazards of tobacco
use, the Chinese study consisted of two parts: one retrospective1
and one prospective2. The retrospective study, which assessed
current mortality from tobacco use in China, was of a new design (see legend
to Fig. 1) and involved analysis of one million deaths.
The prospective study (the early results from which confirm the retrospective
results) began by interviewing a quarter of a million adults and will continue
for decades, monitoring the long−term growth of the epidemic.
Figure 1. For the one million people who died between 1986 and 1988 in 24 cities
(filled circles on map) and in various rural areas of China (open circles),
the disease that caused death was determined and the family was interviewed
to discover what the deceased person had smoked.
The smoking habits of 0.7 million Chinese adults who had died of neoplastic,
respiratory or vascular causes were compared with the habits of the 'reference
group' of 0.2 million adults who had died of other causes. The threefold excess
of lung cancer deaths among smokers could then be inferred from the excess,
in comparison with the reference group, of smokers in the lung cancer death
category. (Similarly, the excesses of various other diseases among smokers
could be calculated.) In each city the local relative risk, the local prevalence
of smoking and the local lung cancer rate (standardized for age by averaging
the seven death rates at ages 35−9, 40−4...65−9) could then
be combined to calculate the absolute lung cancer rates among smokers and
non−smokers separately (see Fig. 2). [This figure
was reproduced from 1 with permission from
the Br. Med. J.]
Apart from HIV/AIDS, tobacco is the only major cause of death that is increasing
rapidly3. Worldwide, smoking caused about 3 million deaths in
1990, out of a total of 30 million adult deaths from all causes, and it will
cause about 10 million in 2030, out of a total of about 60 million4,
5.
Most of this projected increase in tobacco deaths will take place in Asia,
Africa and South America. Until recently, such projections had to be based
chiefly on extrapolation from studies in developed countries such as Britain
or the United States6,
7. But, most of the world's one billion
smokers live in developing countries, where the effects of smoking could well
be different, so local studies are urgently needed.
This new Chinese evidence needs to be seen in the context of Western epidemiological
evidence, which demonstrates the peculiarly long delay between cause and full
effect. In countries such as Britain and the United States, most of those
who now smoke cigarettes began to do so in early adult life, and recent UK/US
prospective studies show that about half of all persistent smokers are eventually
killed by their habit6,
7. Fifty percent of these tobacco deaths
occur in middle age (here defined as 35−69 years), and half in old age4. But only those who have smoked cigarettes since early adulthood
are at particularly high risk in middle and old age, so earlier UK/US prospective
studies—conducted after the rise in cigarette use but only halfway through
the increase in mortality—misleadingly suggested that the risks of tobacco
use were lower8. The need for prolonged smoking before the full
risks become evident means that if a country experiences a nationwide surge
in cigarette use by young adults then this will cause a large increase in
tobacco deaths about a half−century later. Until then, however, there
may be several decades during which cigarette consumption is high but mortality
from tobacco is still relatively low.
In developed countries, cigarette smoking became popular during the first
half of the twentieth century, so the main increase in tobacco deaths has
been seen during the second half of this century. For example, mean US cigarette
consumption per adult in 1910, 1930 and 1950 was 1, 4 and 10 a day, respectively,
after which it remained fairly constant for a few decades. Therefore, the
proportion of all US deaths at ages 35−69 attributed to tobacco rose
from 12 percent in 1950 to 33 percent in 1990 (4).
In many developing countries, cigarette smoking became widespread only
during the last 30 years, so the main consequences of this will emerge next
century. For example, the US pattern of an increase in cigarette smoking between
1910 and 1950 has been repeated 40 years later in China. Mean cigarette consumption
per Chinese man in 1952, 1972 and 1992 was 1, 4 and 10 per day, respectively,
after which it seems to have leveled off1. Both the retrospective
study and the early results from the prospective study in China indicate that
in 1990 tobacco caused about 12 percent of adult male deaths, and by 2030
it will probably be a cause of about one third of them. About two thirds of
young Chinese men become cigarette smokers in early adult life, and in China,
as in America, about half of those who do so will eventually be killed by
their habit. Thus, about one third of all young Chinese men will eventually,
in middle or old age, be killed by tobacco. As about 10 million a year reach
manhood in China, the annual number of tobacco deaths will rise from almost
1 million now to about 3 million in the middle of the next century. Hence,
there will be a total of about 100 million deaths caused by tobacco in China
during the first half of the next century.
But even if the overall 50 percent hazard is about the same in China as
elsewhere, these new studies show that the chief diseases by which tobacco
causes death are very different in China (and, within China, between one city
and another). In the United States, for example, tobacco causes far more deaths
from heart attacks than from emphysema, whereas in China the opposite is true.
Of Chinese tobacco deaths, almost half involve emphysema, and the proportions
involving tuberculosis, esophageal cancer, stomach cancer and liver cancer
are each about as large (5−8 percent) as the proportions involving heart
disease or stroke.
Lung cancer, which was the first major disease to be reliably linked to
smoking in Western studies, is also an important hazard for Chinese smokers,
but, unexpectedly, there is a tenfold variation from one Chinese city to another
in the magnitude of this hazard (see Fig. 2). In each
city, the lung cancer rate is about three times as great among smokers as
among non−smokers, but there is a tenfold variation in the non−smoker
lung cancer rates, which in some cities, perhaps chiefly because of domestic
heating and cooking fumes, are ten times greater than the rates among US non−smokers.
There was even wider variation between one city and another in the non−smoker
emphysema death rates, which in some cities were almost 100 times greater
than those in U.S. non−smokers.
Figure 2. Death rates at ages 35−69 from lung cancer in various parts of
China: smokers versus non−smokers (1986−1988).
The lung cancer rates show wide variation, with extremely high rates in
some cities among non−smokers and, particularly, among smokers. In comparison,
the nationwide US lung cancer death rates in 1990, similarly standardized
for age, were 1.4 per 1,000 men and 0.6 per 1,000 women, and were 0.1 per
1,000 male or female US non−smokers3. [This figure was
reproduced from 1 with permission from the
Br. Med. J.]
From a public health perspective, the main finding of the Chinese study
is the alarming overall risk of death, which can be stated in various ways:
tobacco will eventually kill one third of all young men in China; it will
kill half of all persistent cigarette smokers; there will be one million tobacco
deaths a year in China during the first decade of the next century, 2 million
a year by 2025, 3 million a year by mid−century, and a total of 100
million during the first 50 years of the next century. In comparison, most
of the intriguing little details of how tobacco kills are of less immediate
importance.
The chief exception, however, concerns women. The bad news is that if women
smoke like men they die like men; their overall risks are about the same.
But the good news is that on the whole Chinese women don't smoke like men;
unexpectedly (and unexplainedly), the proportion of women who became smokers
before the age of 25 has decreased enormously over the past few decades. If
this low uptake rate of smoking by young women continues then, although the
proportion of deaths attributed to smoking in 1990 and in 2030 will increase
from 12 percent to about 33 percent for Chinese men, it will decrease from
3 percent to 1 percent for Chinese women. Tobacco would then be responsible
for most of the difference in life expectancy between men and women in China.
The only hope of substantially limiting worldwide tobacco deaths in the
first half of the next century is for many of the adults who now smoke to
stop doing so, because discouraging young people from starting will take half
a century to produce its main health benefits. Western studies show that,
even in middle age, cessation of smoking is remarkably effective, removing
most of the 50 percent risk of death from tobacco if smoking persists. Stopping
at earlier ages is even more effective6. Britain, which is now
experiencing the most rapid decrease in the world in premature deaths from
tobacco, shows that large improvements are possible: over the past 30 years,
UK cigarette sales have halved, as have UK tobacco deaths in middle age3. At present, however, such changes are chiefly limited to educated
Western smokers; Chinese smokers, for example, rarely stop until they are
too ill to continue. So, the worldwide network of epidemiological studies
of tobacco deaths in developing countries—such as India, Mexico, Cuba,
Egypt and South Africa—that are being modeled on these Chinese studies,
is likely to document a growing epidemic in these nations over the next few
decades.
Liu, B.Q. et al. Emerging tobacco hazards in China: 1. Retrospective proportional mortality study of one million deaths. Br.Med. J. 317, 1411-1422 (1998). | ISI |
Niu, S.R. et al. Emerging tobacco hazards in China: 2. Early mortality results from a prospective study. Br.Med. J.317, 1423-1424 (1998). | ISI |
Ad Hoc Committee on Health Research. Investing in health research and development. The World Health Organization, Geneva, Switzerland (1996).
Peto, R., Lopez, A.D., Boreham, J., Thun, M. & Heath, C. Jr. Mortality from smoking in developed countries 1950-2000: Indirect estimates from national vital statistics. Oxford University Press (1994).
Murray, C.J.L. & Lopez, A.D. (eds). In The global burden of disease. (Harvard School of Public Health, Boston, Massachusetts, 1996).
Doll, R., Peto, R., Wheatley, K., Gray, R. & Sutherland, I. Mortality in relation to smoking: 40 years' observations on male British doctors. Br.Med. J.309, 901-911 (1994). | ISI |
Thun, M.J. et al. Alcohol consumption and mortality among middle-aged and elderly U.S. adults. N. Engl. J. Med. 337, 1705-1714 (1997). | PubMed | ISI |
U.S. Department of Health & Human Services. Reducing the health consequences of smoking: 25 years of progress. A Report of the Surgeon General. U.S. Department of Health & Human Services, Public Health Service, Centers for Disease Control, Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health. DHHS Publication No. (CDC) 89-8411 (1989).
