Richard Peto is one of the world鈥檚 leading experts on why smokers die prematurely. He is also one of the most outspoken critics of tobacco companies. He thinks all tobacco advertising should be banned. And he believes campaigns aimed at preventing children from starting smoking are misplaced-we should be helping adults to stop. Fred Pearce meets the Oxford statistician trying to save lives.
Life is full of risks. What鈥檚 so special about tobacco?
It鈥檚 the magnitude of the hazard. Half of all cigarette smokers are eventually killed by tobacco unless they stop. Worldwide, there are only two major causes of death that are clearly getting bigger: tobacco and HIV.
Advertisement
How many people does tobacco kill?
About 4 million a year worldwide. We now know that if you smoke throughout your adult life, there is about a one in two chance that tobacco will kill you. I find it bizarre to think that there are 10 million smokers in Britain and half of them will die from the habit. Yet you can still advertise tobacco, and tobacco companies can still make big political donations.
You attack universities that accept money from tobacco companies. But if the companies want to give it, aren鈥檛 the universities right to accept money where they can?
British American Tobacco recently gave more than 拢3 million to Nottingham University. Now, every cigarette sold makes a profit for the company of about 3p. To get its donation money back, which we have to assume BAT wants to do, it has to sell 100 million extra cigarettes. Every million cigarettes causes roughly one death. So to break even, they鈥檝e got to sell enough cigarettes to cause about 100 deaths. The money, incidentally, was given to fund a professorship in corporate responsibility.
What would you do to fight the tobacco companies?
I鈥檇 ban tobacco advertising or promotion of tobacco goods and trademarks. When the Labour government came to power in Britain in 1997 it promised to ban tobacco advertising, but mysteriously it hasn鈥檛. The tobacco companies really care about advertising. So let鈥檚 accept their judgement that it鈥檚 important-and ban it. Half of British smokers believe that since the government allows advertising, smoking can鈥檛 be all that dangerous. That, if you think about it, is a reasonable inference. After all, we have all sorts of restrictions on things that cause a very limited number of deaths.
Would you ban smoking?
No. When we had a vote here, I voted against banning it in this building.
How odd. Why?
I felt that if I am doing studies of smoking I shouldn鈥檛 be telling people whether they should smoke or not.
But if you are so libertarian, why ban advertising?
I鈥檓 in favour of individual freedom, and that鈥檚 sometimes the opposite of corporate freedom. In this case, the freedom of the tobacco companies to advertise reduces our freedom to bring up children without tobacco companies deliberately trying to trick them into something that for many people is addictive. And yes, one of my children does smoke. I hope he gives up before it kills him.
But many governments have public health campaigns that try to prevent young people from smoking.
They are fine if they work, but there shouldn鈥檛 be so much emphasis on kids not starting that we forget about the importance of adults stopping. Also, kids don鈥檛 understand that death in middle age is real and important. I鈥檓 not sure that kids respond to fears about being killed. Adults do.
But does stopping smoking work?
Yes, even in middle age. Stopping earlier is even better. When the first studies of smoking and lung cancer were done half a century ago, there were almost no long-term ex-smokers to study. Nowadays we have lots of them. We can see that even if you are 50 and have been smoking for 30 years, stopping more than halves your risk of dying from smoking. Stopping at 30 cuts the risk by 90 per cent. Many smokers wrongly believe there is little point in stopping because most of the damage must already have been done. That just isn鈥檛 what the epidemiology shows.
Even so, Britons have been pioneers in giving up. Is it a coincidence that the first research on smoking and lung cancer was done here?
Richard Doll, who brought me into this business, showed in the 1950s that doctors who smoked were much more likely to die in middle age. Doctors read it and thought: my God, this is really serious, it doesn鈥檛 just kill patients, it kills doctors. They became the first major group in the world to give up. Since then, smoking in Britain has halved, and partly as a result we have the world鈥檚 biggest decrease in lung cancer and other tobacco-related deaths in middle-aged men.
But aren鈥檛 things going into reverse, with more young teenagers taking up smoking?
Smokers do start younger now, but there鈥檚 no increase in the proportion who eventually start smoking. That is encouraging.
How did you get into this work?
Absolutely by accident. At university I stumbled out of maths into statistics. I was never going to be much good as a mathematician. It was 1967, I had a big blue motorbike and the sun always seemed to be shining, even when it was raining. I applied for a job with Richard Doll, who was the world鈥檚 greatest cancer epidemiologist, without ever having heard of him. For the first few months I wasn鈥檛 at all sure I wanted a serious job, but as soon as you start to get your own research results, you get hooked. I鈥檝e been addicted ever since, really. Richard is 88 now, but still works here most days, producing good papers.
How would you describe your work on the causes of disease?
Often we seem like Basil Fawlty鈥檚 wife: we specialise in the bleeding obvious. We keep discovering that old, long-established causes of disease are even more important than previously thought, while some things like environmental pollutants and diet are probably a lot less important. I鈥檓 not saying diet is unimportant, but it turns out we know much less about it than we thought. So we are left with things like smoking, blood pressure and cholesterol. And HIV, of course. That鈥檚 a monster.
You鈥檝e done a lot of work on smoking in developing countries.
It turns out smoking is just about as good at killing poor people as it is at killing rich people. It sounds obvious, maybe. But the extent to which smoking was killing people in developing countries was enormously underestimated until we collaborated with people in those countries to do big studies. Surprisingly, however, it tends to kill them by different diseases. So while smoking in the rich world generally kills by causing heart attacks and lung cancer, in China it tends to kill by causing emphysema and liver and stomach cancers, and in India by causing TB.
How come?
We don鈥檛 really know, just as we don鈥檛 really know how smoking causes heart attacks. But smoking does seem to raise the incidence of a range of major killers. Smokers are typically three or four times as likely to die of them as non-smokers.
How many people die prematurely of any disease worldwide?
Ten million a year before the age of five, and another 5 million before 35, although that may soon double because of HIV. And 20 million die prematurely in middle age. Most deaths before old age are from diseases that are largely avoidable.
Why do you say that?
Because for each major disease there is fantastic geographical variation in death rates. Take China. The risk of dying in middle age from stomach cancer varies across the country from 20 per cent to less than 1 per cent. I don鈥檛 think genetic variation accounts for much of this geographical variation. Nor does smoking. This shows there are major causes of disease we still know nothing about.
How are you addressing this?
One thing we are doing is starting to store blood samples. We want millions of them, from Britain, Mexico, China, everywhere. The idea then is to sit back and wait for the donors to die. In about a decade, we will be able to start picking out the samples from thousands of people who died of a particular disease and find out what鈥檚 special about them.
Are there ethical problems?
Not as long as the people give you the sample freely. But there are problems when rules get drafted without being properly considered.
Can you give me an example?
Forty years ago, Richard Doll looked up the records of all women who had claimed sick leave from chicken pox, measles or rubella and who then took maternity leave less than a year later. He confirmed that rubella was causing brain damage and blindness in the babies. That led directly to a vaccination programme. Who was hurt by that research? Nobody. But it would be illegal nowadays, because it used records for a purpose they weren鈥檛 collected for. Much of what is described today as unethical seems to me highly ethical, and vice versa. The proliferation of ethical committees make for a huge bureaucratic load to no obvious purpose. Our work on the benefits of Tamoxifen treatment for breast cancer patients had to go through 300 ethical committees.
What aspects of your work here in recent years have saved most lives?
Probably the large randomised trials of Tamoxifen for breast cancer or aspirin for heart attacks which started here. In the 1980s, Rory Collins ran the ISIS-2 trial, which showed that if people who have had a heart attack are given half an aspirin a day, then about a quarter of the deaths in hospital are avoided. As a direct result, millions of patients a year are being treated and several tens of thousands of deaths a year are being prevented. These really are lives saved.