From Sir, With Love (of Longer Life)
By Rich McManus
On the Front Page...
There aren't many in medicine who can authoritatively offer prescriptions for the entire world, but Sir Richard Peto, who has built an internationally acclaimed career examining the big picture, from the vantage of medical statistics, may be one of them. Returning to NIH for the second time in a month (the first was to accept the 2002 Charles Mott Prize from the General Motors Cancer Research Foundation), Peto explained, before a Wednesday Afternoon Lecture audience in Masur Auditorium on June 26, that halving the rate of premature death worldwide is within the capacity of current medical expertise.
The major culprits in causing early death are smoking, preventable illnesses of childhood (tetanus, diarrhea, measles, perinatal complications), and vascular diseases for which there are solid, classical tools for diagnosis, and emerging therapies, chiefly statins, that are effective, said Peto, who is professor of medical statistics and epidemiology at the University of Oxford in England.
To hear Peto tell it, the only theoretically safe smoker is one who lights up just before facing a firing squad, beyond all hope of rescue; every other smoker is going to lose a percentage of expected lifespan every time he or she smokes. The only difference between the condemned man and the habitual smoker is how far away, in time, the firing squad stands; eventually both perish early.
Peto began with a graph of the French mortality rate, from 1900 onward, showing that roughly one-quarter of the population died in early childhood, one-quarter died in early life, one-quarter perished in middle age (which he defined as "that period between youth and old age that is variously reckoned to suit the reckoner"), and one-quarter expired in late age. "Most of these deaths [in all categories but late-age] are avoidable," he suggested. Indeed, since 1950, the rate of premature death has diminished considerably. "The present is not as bad as the past," he said.
Comparing actual death rates in the 20th century to what we might expect in the 21st, Peto said about 20 million people are thought to have perished in the flu epidemic of 1918-1919 (though that could have been the number who died in India alone, he noted; 40 million or 50 million is probably the more accurate global guess). Another 200 million died in war, or war-related famine (the worst famine on record occurred 40 years ago in China, he said, when between 20 million and 40 million people are thought to have died). And 2,000 million died of preventable childhood diseases, including cholera.
Today, about 60 million people die per year, worldwide, and about twice that number are born. The deaths, early in the 21st century, occur as follows: About 10 million die in early childhood (though over the last 50 years, there has been a threefold reduction in childhood mortality, leaving out the effects of HIV); another 10 million die in the age range 5-34 (again, excluding HIV numbers, which are expanding); and 20 million die in both middle age (35-69) and late age (70-plus). Peto thinks several of those age ranges are ripe for the halving.
The earliest age range is a good target, he said, because many childhood ailments are preventable. "But halving death at middle age is very hard due to HIV, an epidemic that is increasing fast," he said. Peto paused to condemn denial as a culprit in addressing AIDS adequately; those who don't believe HIV is the causative agent have only to examine a paper published in Nature in 1995 showing conclusively, among a population of severe hemophilia patients in Britain (half of whom were unwittingly infected by HIV when pooled blood harvested for clotting factor became tainted), that HIV is the cause of AIDS.
"HIV and tobacco are the only big causes of death still on the rise," he said. Of the 130 million people born this year worldwide, about 20 million will die before middle age, and some 40 million will die in middle age (20 million from vascular diseases Peto warned of an alarming increase in obesity in China, India and North America that will elevate diabetes and its associated complications and 10 million from cancer). "Can these be halved?" he wondered. "What can we actually do with our present knowledge? Well, we can take the known causes (of premature mortality) a bit more seriously."
Peto said that classical risk factors for heart disease blood pressure, lipid profile and history of smoking have an underestimated power as predictors of early death. A meta-analysis of 59 heart disease studies showed that each 20 mmHg difference in systolic blood pressure involved a two-fold difference in vascular mortality in middle age, he said. Both diastolic and systolic values "are very strong and very important determinants of the risk of vascular disease," Peto stated. "Both are more important than they are generally held to be."
Even modest downward tweaks in blood pressure are associated with real benefits, he added. Peto suggested that a strategy of many small interventions of a cheap and easy kind (aspirin, statins) would have a substantial cumulative effect. Of statins he joked, "One little pill can turn your cholesterol level from that of an American to a rural Chinese." With one-half of all adult deaths due to vascular disease, Peto urged a return to an emphasis on classical risk factors not high-tech, expensive and hard-to-deliver diagnostics and therapeutics.
Turning to tobacco, Peto highlighted three main messages: The risk is big half of those who smoke die by the habit; one-quarter of those die in middle age, losing many years of life; and if you quit smoking, you can salvage years that would have been lost. "This ought to be common knowledge among smokers, but it's not," he protested.
Half of all cancer mortality in the U.S. was tobacco-related throughout the 1970's and 1980's, Peto said, but the rates are gently declining. He credited changes in cigarette composition as a significant factor in reducing the cancer death rate, but quipped that NCI hates to hear him acknowledge this.
The United Kingdom has also witnessed an "extraordinary decrease in lung cancer mortality" in the past half-century, he reported, but the trend is directly the opposite in France and Hungary, where smoking in both male and female populations is rising, with predictably dire results. Peto found it morbidly funny to have heard a Hungarian scientist rail against the perils of air pollution in his country, all the while puffing madly on a cigarette.
Smoking is also disastrously common in China, where two-thirds of adults smoke, and India, where 40 percent of the population smokes, Peto said. "If we keep going the way we are, there will be 1 billion tobacco-related deaths in the 21st century [compared with 100 million deaths in the 20th century]," he predicted. Adults need to quit, and the percentage of kids who start must be reduced, he prescribed.
The death rate from cancers not attributable to smoking has been halved in the past 50 years, Peto said, adding that "this could happen for other causes of death, too." The child survival rate also is improving, but needs substantial resources worldwide, he argued. His talk, given in honor of the late NIH epidemiologist Dr. Robert S. Gordon, Jr., ended whimsically: "Still, however, only a small proportion will survive to 100 halving premature death will not give us eternal life."
Up to Top