Examination of Witness (Questions 381-399)|
Professor Sir Richard Peto
14 FEBRUARY 2006
Chairman: Thank you very much for coming.
You know the routine pretty well, I guess, and your expertise
is going to be, I am sure, very helpful to us in our study of
these questions of risk. I am told to remind everybody who comes
to be a witness to speak up and to speak slowly and clearly so
that we get an accurate report of what you have to say. You know
something of the questions, but if I may I will ask Lord Macdonald
if he would like to start.
Q381 Lord Macdonald of Tradeston:
Sir Richard, I wanted to start by asking if you could give us
your assessment of the health risks associated with passive smoking
in the home or at work and in other public places. It would be
helpful if you could give us an indication of both absolute and
relative magnitudes of the health risks and also the degree of
uncertainty attached to the available statistical evidence.
Professor Sir Richard Peto: I am sorry, I know
that is what you would like to be given, but the point is that
these risks are small and difficult to measure directly. What
is clear is that cigarette smoke itself is far and away the most
important cause of human cancer in the worldthat is, cigarette
smoke taken in by the smokerand passive smoking, exposure
to other people's smoke, must cause some risk of death from the
same diseases. Measuring that risk reliably and directly is difficult.
You can do it indirectly by suggesting approximate proportionality
of hazard to exposure, but the assumptions become almost untestable.
The arguments that have been forward for the various thresholds,
that there is some dose below which there is absolutely no risk,
have no scientific plausibility. They have come up a lot of times
because, as you know, when there is the statement "there
is some risk" then there is political pressure to get rid
of that risk, so it would be very convenient if one could be told
that there was no risk, and so various implausible models involving
thresholds got proposed. There is going to be some risk and there
is always going to be quite a lot of uncertainty about the magnitude
of that risk, I am sorry. What is definite is that cigarette smoke
is causing about 100,000 deaths a year in this country, and a
few million deaths a year worldwide, that this number of deaths
is still increasing in some other countries, although not in this
country, and that passive exposure to cigarette smoke in various
circumstances must be producing some risk. That is definite, and
the threshold arguments are often politically motivated inventions
which do not have much scientific plausibility. I am sorry not
to be more helpful; you want numbers and I could give you numbers
by direct extrapolation, but what does one make of them? These
hazards cannot be directly measured.
Q382 Lord Macdonald of Tradeston:
If you have such a large sample of people dying of lung cancer,
would it be possible to interview them about their personal circumstances
in a way that would allow you to deduce whether they had frequented
bars or had a smoky home or whatever?
Professor Sir Richard Peto: This has been done.
People who persistently smoke cigarettes have about 20 times the
lung cancer risk of those who never smoke; that is a 2000 per
cent excess. The exposure that one would get when breathing other
people's smoke obviously depends on the circumstances, but even
heavy exposure would be something like one per cent of what a
smoker gets, maybe in other circumstances 0.1 per cent, so you
would expect if there was proportionality to get something up
to about a 20 per cent excess. That is what you see in the average
of all the studies, and people have pointed to the uncertainties
in this evidenceit could under-estimate the real hazards,
or it could over-estimate the real hazards. It is however, roughly
what you would expect from simple proportionality.
Q383 Lord Skidelsky: 2000 as opposed
Professor Sir Richard Peto: Yes. You would expect
an excess of a few per cent, or several per cent. The one thing
that is often left out of this is that these are the excesses
for lifelong non-smokers. Smokers who have stopped, who have given
up smoking, have much less lung cancer risk next year than they
would have had if they had not stopped, but they still have a
lot more risk of lung cancer than if they had never smoked. A
lot of the cells in their lungs will be altered part-way towards
cancer for the rest of their life, and if they are lucky then
they will live out the rest of their life and none of those cells
will suffer that final step. Those people would probably be at
greater risk of damage from breathing other people's smoke than
lifelong non-smokers would be, because lifelong non-smokers are
going to have so few cells that are at risk of that final step.
But, of course, when studying people who have smoked for 10 years
or 20 years and then stopped, it is very difficult to work out
exactly what their risk would be without passive smoking, so it
is difficult to do reliable epidemiological studies of them. So,
people have done epidemiological studies of lifelong non-smokers
because it is a cleaner comparison. It is however, the ex-smokers,
those who have stopped, who are probably going to suffer the greatest
absolute risk as a result of exposure to other people's smoke,
at least in terms of cancer.
Q384 Lord Macdonald of Tradeston:
I am just trying to see if you can break it down into categories.
Presumably, non-smoking shepherds would be much healthier than
non-smoking traffic wardens.
Professor Sir Richard Peto: There probably would
not be a very big difference. The amount of exposure from cigarette
smoke is so much greater than the amount that you would get from
the ambient air outside in the city, certainly nowadays, that
there is not so much difference between non-smokers in urban and
rural populations. There are some differences, but they are really
Q385 Lord Macdonald of Tradeston: What
I was heading towards is the suspicion that the next way this
is taking us might be the banning of smoking in the home, and
I wondered what evidence you would bring forward that would allow
that debate to take some sensible form.
Professor Sir Richard Peto: I do not want to
be cast in the role of advocating banning smoking in public places
or in private places. What I am concerned with is that enormous
risks should be taken seriously, like the extent to which smokers
kill themselvesthere is about a 50 per cent chance that
a person who smokes cigarettes and continues to do so will be
killed by tobacco, which is vastly greater than almost any other
risks around. It is the relative importance of this that I somehow
want to get across. But, we are concentrating now, because this
is your task, on the effects of breathing other people's smoke,
although the main way smokers kill people is by killing themselves,
not by killing other peoplethey are a lot better at killing
themselves than they are at killing other people.
Q386 Chairman: It is very difficult,
but the message I am getting is that when you look at the evidence
that is put forward by the tobacco companies to say there is not
any serious evidence, or when you get evidence put forward by
people who are totally opposed to smoking of any sort, then when
you come to deal with the question of passive smoking you would
be sceptical about any of the evidence that is produced on either
side, further than you have indicated.
Professor Sir Richard Peto: I think there has
got to be some risk. The extent to which active smoking is causing
cancer, heart disease and lung cancer is enormous. There is about
a 50 per cent chance that a smoker who carries on smoking will
eventually be killed by this, and that is definite. When you place
the same mix of chemicals, more or less, in to the general air,
it must cause some risks, so I think the statement that there
is some risk to non-smokers is well-founded; you would have to
use the most extraordinary, implausible metaphysical arguments
to argue for threshold doses below which there is zero risk, especially
in a world where there are a lot of ex-smokers whose lungs are
already well off any zero on the dose response relationship. The
definite statement is that some people are killed by breathing
other people's smoke, and then there is reasonable uncertainty
about the number killed. I am sorrythere is bound to be
wide uncertainty when you are trying to measure risks like this.
There is, however, a rather odd finding, which seems to come up
repeatedly, of quite a marked excess of mortality from heart disease
among those exposed to other people's smoke. This is odd because
the hazard is much bigger than you would expect from extrapolation
from the hazards faced by smokers. It is not impossible, but it
is odd, and in the case of heart disease, the apparent risks in
passively exposed non-smokers are bigger than one would have expected.
It is possible, however, that there are paradoxical dose response
relationships, with rather substantial risks at low doses. To
take an extreme example, if you plotted the probability of sneezing
against the dose of smoke, smokers on the whole do not make themselves
sneeze but they can quite easily make non-smokers sneeze. This
example shows that there can be a medical effect with a very odd
dose response relationship. It is not impossible, therefore, that
passive smoking produces a substantial risk of heart disease,
and if that were so then it would represent a really substantial
public health hazard. There is argument as to whether such a hazard
is real, and there is still reasonable disagreement about this,
but I do not think there is reasonable disagreement with the statement
that smokers do kill some non-smokers.
Q387 Lord Skidelsky: I just want
to ask a supplementary. You would be reluctant then to commit
yourself to a statement such as that in first report of the Select
Committee on Health: "It is currently estimated that second-hand
smoking causes at least 12,000 deaths each year in the United
Kingdom . . .", which was repeated by the Secretary of State
on the Today programme this morning. The way the legislation
and the case are being presented is much more definite than the
statement you have just made.
Professor Sir Richard Peto: Yes. That is the
approximate number that you would get if you take the excess of
heart disease deaths that is observed among people exposed to
other people's smoke. If you take the heart disease risk as real,
then you could get estimates like the one you quoted, and there
is argument as to whether the heart disease risk is real or not.
I do not know, and you have a very difficult job in deciding what
rules to enact. I was asked in the papers you sent me whether
regulations on passive smoking or breathing other people's smoke
should be promulgated because they would reduce the number of
people who choose to continue smoking themselves. I do not want
to argue for or against any rule, but there does seem to be a
consensus that it would affect the number of people who choose
to smoke. If that were not the case, then the tobacco industry
would not be so concerned about it, and the strength of their
concern does indicate that their reckoning is that it would decrease
the number of people who continue to smoke. If that is the case
then one side-effect of such regulations would be to avoid quite
a number of premature deaths, but I do not want to distort the
direct science of what breathing other people's smoke does for
the non-smoker or, more importantly, the ex-smoker on account
of that. If, however, you are making regulations, then probably
you should bear in mind both the direct and the indirect effect
of those regulations; it would be irresponsible not to do so.
Q388 Lord Roper: Sir Richard, you
have from time to time used the word extrapolation as a way of
getting to these sorts of estimates of risk. I think I follow
what you mean by that, but I wonder if you could restate what
you would be extrapolating from and what the process would be.
Professor Sir Richard Peto: Extrapolation is
the common-sense idea that if you get about one per cent as much
exposure then you might get about one per cent as much risk, or
something of that order of magnitude. Obviously, the dose response
relationship does not have to be a straight line, and to the extent
that it is not a straight line then those answers might be wrong,
they could be too high or too low.
Q389 Lord Roper: But numbers like
the 12,000 are numbers which have arisen . . .
Professor Sir Richard Peto: That particular
number came from another source. The trouble is that because these
risks are small they are difficult to measure, for obvious reasons.
In many populations the main way cigarette smoke kills smokers
is by causing death from heart disease rather than causing death
from lung cancer. Studies have been done, as you suggested, on
lung cancer patients, asking what they smoke, how they livedand
those studies indicate in aggregate, roughly the sort of risk
that you might expect from extrapolation of the risks among smokers.
On heart disease, similar studies indicate risks from passive
exposure that are a lot bigger than would be expected from extrapolation
downwards from the effects of smoking on the smoker. Nobody has
really argued the studies away, yet everybody feels uncomfortable
with the conclusion, unless it could be better understood. Another
problem in studying heart disease rather than lung cancer is that
whereas lung cancer is quite a rare disease in non-smokers, heart
disease is quite common and has many different causes. If you
just try and compare people getting heart attacks with others,
then you do find an excess of people exposed to other people's
smoke, but it is very difficult to interpret this reliably. I
do not thinkand I could go into this if you wantthat
the suggestion by the tobacco industry that there is no risk is
plausible, but there is a wide range of estimates coming from
Q390 Lord Sheppard of Didgemere:
In addition to the debate that has taken place on passive smoking,
there has also been a debate at various times about such things
as traffic exhaust. Given that it is difficult to measure the
effects of passive smoking impact on non-smokers, it is probably
equally difficult to measure the impact of exhaust systems, but
do you want to comment on those two debates that have occurred
at various times in the last years?
Professor Sir Richard Peto: In terms of general
pollution it has been possible to demonstrate the hazards of persistent
exposure to coal smoke, because there have been lots of cases
where the hazard has been so extreme that it has been really clearly
demonstrable; it is more difficult with traffic exhausts because
whole populations in an area of a city are exposed to them and
the main concern about traffic exhausts is not that they are going
to cause cancer in the non-smoker, but they are going to make
the risks bigger in the smoker. That is very difficult to measure
because there are variations in the way people smoke and the intensity
with which they smoke their cigarettes can also produce differences
in risk, so the main concern about traffic exhausts is not whether
they kill non-smokers but whether they increase the risk among
smokers, as radon has recently been shown to do. Radon as a pollutant
of the domestic environment, of houses, increases the extent to
which cigarettes kill people, it multiplies up the risks of smoking,
and the concern is that some of these other sources of pollution
might do the same. In the case of radon, the hazard happened to
be measurable because technically it was relatively easy to measure
persistent differences in exposure.
Q391 Lord Sheppard of Didgemere:
Going back to passive smoking as such, does it matter if one cannot
measure the impact of a regulation? If, for example, the effect
is that it stops parents smoking at home in front of their children,
it most probably helps, does it?
Professor Sir Richard Peto: Does it matter if
one cannot measure it? I am sorry, I do not quite follow the question.
Q392 Lord Sheppard of Didgemere:
The policy can still be right even if, when you got down to it,
one might even be accused of exaggerating the impact of it?
Professor Sir Richard Peto: The main thing that
parents smoking definitely does is that it encourages the kids
to smoke. Kids who live with parents who smoke are more likely
to smoke themselves and, therefore, are more likely to get killed
by smoking; there is an association of the smoking habits of the
parents with the smoking habits of the children when the children
reach adult life. When considering the hazards for smokers, we
now know that cigarette smoke is an extraordinary mix of toxic
chemicals and what is really surprising when you look at the pharmacology
of cigarette smoke is that half of all smokers do not get killed
by it. That is really the surprising thing: there are thousands
of chemicals that have been identified in cigarette smoke, you
breathe them in and you get an increased risk of cancer of the
mouth, throat, oesophagus and lungs; they go to the edge of the
lungs and cause emphysema and then they go around the body and
you finish up with mutagens damaging all the cells of the body
that are exposed to these chemicals. Eventually they get concentrated
in the urine, in the bladder, causing a risk of cancer of the
bladder, and even after the smoker passes urine the urine is mutagenic
to cultured cells in the laboratory. It is really surprising that
more than half of all smokers are not killed by their habit. I
notice that in their evidence to this committee the tobacco manufacturers
described the cellular repair systems trying to repair the damage
done by chemicals; it is remarkable how well they work, given
this permanently mutagenic fluid in which the smokers' cells live
all the time, yet you finish up with only half of the smokers
getting killed by it.
Q393 Lord Macdonald of Tradeston:
A quick follow-up on exhaust fumes. About 30 years ago there was
public concern about lead in exhaust fumes and action was taken.
Professor Sir Richard Peto: Yes.
Q394 Lord Macdonald of Tradeston:
Has there subsequently been evidence of reductions in the related
illnesses that were causing the concern?
Professor Sir Richard Peto: I am sorry, I have
just got no useful information to answer that. I have read popular
articles suggesting that, yes, exposure was reducedthe
trouble with current levels of exposure to lead is that we are
a lot too close to the levels of lead in the blood that do have
measurable effects on human intelligence. We are within a factor
or two of levels that would actually have measurable effects.
But, although the effects on intelligence of current exposure
levels might not be measurable, this is one of the exposures where
we do not have any kind of comfortable safety margin. Nobody,
if one had a choice based purely on toxicology, would want to
be as near as we are to a level of blood exposure that has measurable
effects on intelligence.
Q395 Lord Skidelsky: This is really
the follow-up to Lord Sheppard's question: recent research, based
on evidence from America, appears to suggest that a ban on smoking
in public places might actually increase smoking in the home,
which is much the most important source of the danger of passive
smoking. Does one then, by passing legislation, run the risk of
simply transferring the habit and its effects from a less dangerous
place to a more dangerous one?
Professor Sir Richard Peto: I am sorry, I am
much more of a student of the consequences of smoking than of
the causes of smoking, and I do not know whether that statement
is true or not. This is not an area that I have studied, I am
sorry, as to how regulations on smoking in public places would
affect the extent to which people smoke at home.
Q396 Lord Skidelsky: It is a question
about consequences rather than causes because legislation may
have these unintended consequences.
Professor Sir Richard Peto: I stand informed.
If you say it does then . . .
Q397 Lord Skidelsky: I am quoting
a study. I have a supplementary to thator would you like
to say any more about that?
Professor Sir Richard Peto: No, not really.
The key thing about smoking in the home is (a) the parents, or
whoever it is in the home, have a fair chance of killing themselves
and (b) they have a fair chance of making it more likely that
the kids start to smoke.
Q398 Lord Skidelsky: The argument
is that the effect of passive smoking is much greater in the home
than in a more public place because of the proximity of contact
between the smoker and the non-smoker, and it is the continuity
of contact as well as the proximity.
Professor Sir Richard Peto: I do not know what
effect such legislation would have on where people smoke. One
thing is that people addicted to nicotine seem to need is to get
their blood levels up to a certain level, and I am not sure that
what they have been doing during the day would have very much
carry-over effect as to what they were doing in the evening, but
I do not know.
Q399 Lord Skidelsky: Thank you. If
I could ask a supplementary, it is really more of a technical
one. In the evidence we received from the Imperial Tobacco Company
Professor Sir Richard Peto: I was sent a copy
of it last week.