Letters
Effect on mortality of switching from cigarettes to pipes or cigars
Study underestimated difference in risk Patterns of inhalation are important"Switchers" will have had higher cumulative exposure to
tobaccoAmerican study supported conclusionsAuthors' replyStudy underestimated difference in risk
EDITOR There are several possible explanations for the high proportion of
non-smokers with lung cancer in Wald and Watt's study. Firstly, this
was an elite group of subjects (business and professional men), of whom
only a fifth smoked; one would therefore expect proportionally more
cancers among non-smokers. In a previous study of professional British
male subjects (the British doctors study), however, only seven (1.6%)
of 441 deaths from lung cancer occurred in non-smokers.3
Secondly, the number of lung cancers in Wald and Watt's study was
small (102 cases), so the finding of seven cancers in non-smokers
(compared with an expected finding of one or two cancers based on the
above studies) may have occurred by chance. Thirdly, some long term
ex-smokers in Wald and Watt's study may have described themselves as
non-smokers (Capewell et al found that at least a quarter of
"non-smokers" were really ex-smokers).
Whatever the explanation, Wald and Watt seem likely to have
overestimated the risk of lung cancer in lifelong non-smokers compared
with that in larger British studies. It is therefore important to
emphasise to smokers in the general population that their risk of lung
cancer is far more than 16 times that of a non-smoker and that this
risk can be reduced greatly by stopping smoking. It is also likely that
pipe and cigar smokers (and long term ex-cigarette smokers) are at
greater relative risk of lung cancer compared with non-smokers than
suggested in the authors' study.
Patterns of inhalation are important
EDITOR Men who had switched to pipes or cigars at least 20 years before
entering the study (switchers) were observed to have a lower risk of
dying than those who continued smoking cigarettes. Cigarette smokers
generally reported deeper inhalation and had higher blood carboxyhaemoglobin concentrations than either primary pipe or cigar
smokers or switchers. The best predictor of risk was the measure of
inhaled smoke (carboxyhaemoglobin concentration) taken at entry,
and after allowance was made for this, the smoking category had no
significant predictive value. The conclusion that switching conferred a
benefit therefore hinges on the assumption that it caused a lowering of
smoke intake. But no information is presented on switchers' intakes
when they were cigarette smokers. They may or may not have had intakes
similar to those of the men who continued to smoke cigarettes. If, as
is entirely possible, they were already lighter smokers than the
continuing cigarette smokers, their observed lower risk may be
attributable to patterns of inhalation established before they switched
from cigarettes and largely maintained after switching.
It would be unfortunate if cigarette smokers were encouraged by this
report to switch to pipes or cigars. Data from the large multiple risk
factor intervention trial indicate little reduction in exposure after
switching.
2 3
Nicotine addicts are notorious for grasping
at any straw rather than give up completely. Modern small cigars are
particularly unlikely to confer any benefit, as they are designed,
packaged, and promoted to be as like cigarettes as possible.
"Switchers" will have had higher cumulative exposure to
tobacco
EDITOR The validity of their findings is weakened by incomplete data on
exposure. There were no data on duration of smoking before entry to the
study, and exposure status was categorised on the basis of a single
assessment of current smoking behaviour at the time of entry, with no
reassessment during follow up. This could introduce bias if, for
example, switchers were subsequently more or less likely to give up
smoking entirely than non-switchers. There is also a paucity of data on
possible confounding factors other than blood pressure and blood
cholesterol concentrations at entry.
Even if the principal finding is accepted as valid, we question
the explanation offered for it. We believe that a higher cumulative exposure to tobacco (a known predictor of mortality, particularly from
lung cancer2) among switchers is more likely to account for the observed differences in mortality than are minor variations in
inhaling. This is for two reasons. Firstly, switchers are by definition
former cigarette smokers, who, as the paper shows, have a higher
consumption of tobacco than cigar and pipe smokers. Secondly, switchers
are likely to have had a longer duration of exposure to tobacco since
they had all given up smoking cigarettes at least 20 years before the
health examination, and there were no reported criteria for duration of
smoking among non-switchers.
The study confirms previous findings that mortality is higher among
cigar and pipe smokers than non-smokers.3 Therefore, we
believe that healthcare workers should advise cigar and pipe smokers to
give up completely and, if the findings from this study are confirmed,
could justifiably concentrate their efforts on cigar and pipe smokers
who formerly smoked cigarettes as a particularly high risk group.
American study supported conclusions
EDITOR There were 6919 male incident cases of lung cancer and 13 458
controls,
2 3
including 573 cases and 1036 controls who smoked cigarettes and cigars or pipes and 15 cases and 56 controls who
switched from cigarettes to cigars or pipes. Previous analyses concluded that cigarette smokers who switched from non-filter to filter
cigarettes or reduced the number of cigarettes smoked per day lowered
their risk of lung cancer.
4 5
Relative risks of lung cancer were lower for former than current
smokers (table 1). In addition, relative risks for cigarette and cigar
or pipe smokers were lower than those for cigarette-only smokers but
higher than those for cigar-only, pipe-only, and cigar and pipe
smokers. Those who switched from cigarettes to cigars or pipes had
risks similar to those of cigar-only and pipe-only smokers.
Table 1
For cigarette and cigar or pipe smokers, relative risks for
former smokers declined only if subjects stopped smoking cigarettes (table 2). The relative risk was 10.9 for current cigarette and cigar
smokers, increased to 12.4 for former cigar smokers who continued to
smoke cigarettes, and fell to 5.0 for former cigarette smokers who
continued to smoke cigars. The relative risk for subjects who stopped
smoking cigarettes and cigars was 4.6. A similar pattern occurred for
cigarette and pipe smokers. Relative risks were 11.6 for current
cigarette and pipe smokers, 11.4 for former pipe smokers who continued
smoking cigarettes, 7.6 for former cigarette smokers who continued
smoking pipes, and 3.5 for subjects who stopped smoking cigarettes and
pipes.
Table 2
Our analysis showed that cigarette smokers who switch to cigars or
pipes reduce their risk of lung cancer, thus supporting the conclusion
of Wald and Watt. We also found that mixed smokers who stop smoking
cigarettes but continue smoking cigars or pipes also lower their risk
of lung cancer, although they continue to incur a risk five times
higher than that of non-smokers.
Authors' reply
EDITOR Jarvis expresses concern that the men who switched from smoking
cigarettes to smoking pipes and cigars (switchers) may have had lower
former cigarette consumption than those who continued to smoke
cigarettes, in which case there would not necessarily be a reduction in
risk because it would be lower anyway. This is possible, although our
data suggest that, if so, it had only a small effect. In men aged 15-24 the mean cigarette consumption in switchers and continuing cigarette
smokers was the same, and in men aged 25-34 it was on average three
cigarettes a day lower among switchers. This indicates that most of the
difference in risk between switchers and continuing cigarette smokers
is likely to be a reduction in risk as a result of switching.
We agree with Edwards and Jakubovic that obtaining repeated measures of
smoking habit would improve the precision of smoking data, but it is
remarkable that a single assessment of smoking was so predictive of
mortality many years later. If there were any error, it is more likely
that it would have masked effects, not "created" them. We believe
that confounding is a material issue only with respect to heart
disease, and we adjusted for blood pressure and serum cholesterol
concentration, which are two factors that are strongly related to
ischaemic heart disease. It is unlikely that other factors would
introduce significant confounding. We acknowledge that the amount of
tobacco smoked per day may be more important than the extent of
inhaling in determining risk of smoking related death, but there is
evidence that both are involved.
Finally, we were pleased to see the corroborative results of
Lubin and Fraumeni.
In their study of lung cancer and other diseases in male pipe
and cigar smokers, Wald and Watt seem to have underestimated the
difference in the risk of lung cancer between male current smokers and
lifelong non-smokers (a 16-fold difference in their study).1 Non-smokers accounted for seven of 102 lung
cancers in the study (figures for most ex-smokers were not supplied). This is very different from the experience of Capewell et al, who
studied 3070 Scottish patients with lung cancer.2 Only 0.7% of men with lung cancer were lifelong non-smokers. My experience in Salford (three (0.8%) non-smokers and 217 current smokers among 380 men with lung cancer) is almost identical with that of Capewell et al.
On the basis of local data on the prevalence of smoking, I calculate a
62-fold increased risk of lung cancer for Salford smokers compared with
non-smokers. Capewell et al's figures would also imply a risk of
cancer of at least 50-fold for current smokers.
Hope Hospital, Salford M6 8HD
Wald and Watt reported that switching from smoking cigarettes to
pipes or cigars would roughly halve cigarette smokers' risk of dying
of smoking related disease.1 If this were true it would be
of great importance for public health. But it does not follow from the
data presented.
ICRF Health Behaviour Unit, Department of Epidemiology and
Public Health, University College London, London WC1E 6BT
Wald and Watt reported increased mortality related to smoking
among cigar and pipe smokers who previously smoked cigarettes (switchers) compared with cigar and pipe smokers who had not previously smoked cigarettes (non-switchers).1 We have some concerns
about the validity and interpretation of the findings, and would like to add to the authors' key messages.
Department of Epidemiology and Public Health, Medical School,
University of Newcastle, Newcastle upon Tyne NE2 4HH
Michael Jakubovic
County Durham Health Authority, County Durham DL1 5XZ
Wald and Watt presented results of a prospective study
indicating that cigarette smokers decrease their chance of death from
ischaemic heart disease, lung cancer, and chronic obstructive lung
disease by switching to cigars or pipes.1 The study was limited by small numbers of deaths, particularly from lung cancer, on
which changing smoking habits would be expected to have the greatest
impact; an inability to evaluate cigar and pipe smoking separately; and
the use of disease mortality rather than incidence. The findings
prompted us to re-examine data from a large case-control study of lung
cancer carried out at seven locations in Europe.
; data are on men only
Joseph F Fraumeni Jr
Division of Cancer Epidemiology and Genetics, National Cancer
Institute, National Institutes of Health, Bethesda, MD 20892, USA
The risk of lung cancer among current cigarette smokers compared
with lifelong non-smokers in our paper (a 16-fold increase) is
virtually the same as that found in the prospective study of British
physicians (a 15-fold increase).1 This confirms that our
estimate of risk is reasonably accurate. The risk of death from lung
cancer in lifelong non-smokers was 7.8 per 100 000 per year (95%
confidence interval 3.7 to 16.5) in our study, which was of men aged
35-64 at entry who were followed up for an average of 14 years and 4 months.
H C Watt
Department of Environmental and Preventive Medicine, Wolfson
Institute of Preventive Medicine, St Bartholomew's and the Royal
London School of Medicine and Dentistry, London EC1M 6BQ
© BMJ 1998
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This Article
-
Extract
- Respond to this article
- Alert me when this article is cited
- Alert me when responses are posted
- Alert me when a correction is posted
Services
- Email this article to a friend
- Find similar articles in BMJ
- Add article to my folders
- Download to citation manager
- Request Permissions
Citing Articles
Google Scholar
PubMed
Related Content
Bookmark with
What's new
Latest blogs and video
Services
Tools
Online poll
Resources
Rapid responses for this article
There are no rapid responses for this article.Most read
Print issues
Student BMJ
Asylum seekers' care
UK medical students have published unreleased government plans to restrict failed asylum seekers' access to medical care
