Harvard Public Health Review
Fall 2004




Epidemiology's Odysseus

His passive smoke discovery one step in a career journey

A few years ago, the editor of the medical journal The Lancet drew up a hypothetical "canon of western medicine," --works he thought every health professional should read. On that all-time list were Galen, Lister, and Koch … not to mention Hippocrates. And oh, that other Greek fellow, Dimitrios Trichopoulos, one of the world's leading cancer epidemiologists and a 40-year veteran of HSPH.

While best known for his seminal research linking passive or "second-hand" smoking to lung cancer, Trichopoulos has reported on a vast range of factors that influence health and disease risk, most recently informing the world how to eat well and live long. A study he published last year in the New England Journal of Medicine with his wife, Antonia Trichopoulou, concluded that close adherence to the olive-oil-rich Mediterranean diet lowers cancer and heart-disease risk and promotes longevity. Diets in Greece are changing, Trichopoulos concedes, but "I think, by and large, that the Mediterraneans 100 years ago got it about right" when it comes to diet and health.

Throughout his career, Trichopoulos too has managed to get certain things right, often plumbing his native country for clues to the mysteries of human disease, frailty, and death. In studying diseases, he uses methods described variously by peers as ingenious, creative, elegant, and opportunistic, in the very best sense of that word.

"In general, I have taken up unusual ideas. You can put it that way," Trichopoulos politely accedes. But he is quick to credit circumstance. "This may not reflect originality. This may reflect the constraints of limited resources of working in Greece. The saying is that poverty is the mother of ingenuity. Limited resources--they force you to think."

Trichopoulos learned to maximize resources early in life. Born in the Greek city of Volos in 1938, he thrived and excelled as a student, despite civil war and a ruined economy that made life hard following World War II. Encouraged by his surgeon-father to pursue medicine, he chose to study psychiatry and neurology at the University of Athens Medical School. There he met epidemiologist Brian MacMahon of the Harvard School of Public Health, who noted his facility with numbers and urged him to seek a master's degree at HSPH.

For years, Trichopoulos divided his energies between Boston and Athens. A full professor at Athens' medical school by the age of 33, he transformed that school's fledgling epidemiology department into one of the more productive in the world--without the kind of well-funded research infrastructure American researchers take for granted. Eventually he was recruited to the HSPH faculty full-time, where from 1989 to 1996 he served as the School's Epidemiology Department chair.

His peers throughout the world describe Trichopoulos as wielding Occam's razor, slicing through the clutter to the essential truths of disease causation. Hans-Olov Adami, an adjunct professor at the School who chairs the medical epidemiology and biostatistics department at the Karolinska Institute in Stockholm, speaks of Trichopoulos's deep knowledge of epidemiologic methods. Others note his drive. By any measure, Trichopoulos's output is prodigious: more than 800 publications to date. Topics range from that of his first (1967) paper, which showed that the disproportionate ratio between male and female siblings in some families has a genetic basis, to research on injury, the correlation between sun exposure and suicide, and, of course, cancer.

Cancer Connections

Epidemiologists are public health's detectives, sifting through information about people's lifestyles and environments for clues about what makes them sick and die. In the early days of Trichopoulos's career, epidemiology (the word derives from "epidemic") focused almost exclusively upon infectious disease outbreaks. In the 1950s, chronic diseases began overtaking infectious diseases as the main cause of illness and death in the developed world. Soon MacMahon began to pioneer the field of cancer epidemiology at HSPH. Trichopoulos, with a foot in both infectious and chronic diseases, would become in the 1970s the first to connect liver cancer to the hepatitis B virus using simple biological markers. Later, his work helped show a liver cancer–hepatitis C link.

Trichopoulos's single most influential study, the one linking passive smoking to lung cancer, came out of his work in Greece, in collaboration with MacMahon. In the 1970s, men in that country smoked a great deal, women hardly at all. A light went on. "You have very heavy passive smoking exposures," says Trichopoulos. "I wanted to see whether it matters."

It did. The study, published in the International Journal of Cancer in January 1981, reported that women with smoking spouses were 2.4 times more likely to have lung cancer than those whose spouses were nonsmokers.

"Dimitri's paper had a direct, practical effect on public health and contributed to a real change in the smoking scene worldwide," says Walter Willett, the Fredrick John Stare Professor of Epidemiology and Nutrition and chair of the School's Department of Nutrition. Here was proof that smoking posed a danger, not only to smokers, but to those around them, too.

Just how great a danger remains controversial, however. Studies have yet to link passive smoking in the workplace to lung cancer. Yet it's only logical, Trichopoulos suggests: "Tobacco smoke is like ionizing radiation. Any exposure is dangerous."

Trichopoulos's study linking stress to fatal heart attacks is another landmark piece of research, widely admired for its ingenuity. Surprisingly few people died from injuries when an earthquake measuring 6.7 on the Richter scale hit Athens on February 24, 1981. Trichopoulos seized it as an opportunity to measure the health effects of a massive dose of stress. He and his colleagues found that heart- disease-related deaths in the Greek capital had doubled in the days after the quake. "It's so simple when you think about it," says

Adami, of Stockholm. "But you must have the imagination to take a look."
It was this study that Lancet editor Richard Horton identified as one of 27 texts in his hypothetical, must-read canon of Western medicine.

More than a theory

Today, Trichopoulos's chief research interest--no, make that his research passion--is his theory that in utero exposure to hormones, including estrogen, is a principal cause of breast cancer. He first proposed the connection in an essay in Lancet 14 years ago. The theory is built on two overlapping propositions: First, that exposures in utero can cause cancer later in life; second, that exposure of a fetus to hormones increases the number of cells in the mammary glands--and, more precisely, those cells that are undifferentiated, and therefore prone during cell division to making genetic mistakes that lead to cancer. (Mammary gland mass is not the same as breast size, which is determined to a large extent by fat cells.)
"Do you know which hormones increase the risk of breast cancer?" asks Trichopoulos with animation. "Essentially all of them: Estradiol, progesterone, prolactin, IGF. Why? Because they are growth hormones. The essence is growth."

Is this more than a theory? Trichopoulos stops just shy of saying it's been proved, describing it as an "accepted idea." One obstacle to proof is that it is impractical to directly measure fetal exposure to hormones on a large, epidemiologic scale. To get around this constraint, Trichopoulos and others have resorted to investigating factors correlated with fetal hormone exposure, particularly birth weight.

A Danish study last year found that a woman's breast cancer risk is significantly higher when her birth weight is higher. Results from the Nurses' Health Study have also shown a birth weight–breast cancer connection. Observes Willett, one of this study's principal investigators, "There is quite of bit of evidence that something is going on."

Trichopoulos understands others' reluctance to embrace the importance of in utero hormone exposure. "People think small babies are a bad thing. Then you tell them that large babies are also a bad thing. They're not ready to accept that." Moreover, he notes, your mother's hormones aren't nearly as attractive a culprit as organo-cholorines (PCBs) or other noxious environmental toxins, causes that "give you somebody to fight with."

Why, then, was Trichopoulos drawn to study in utero exposures? "Because it is original, suggesting a rationale for eventual intervention," comes the reply, "and because more conventional research has not been particularly successful for preventing primary breast cancers."

No greater enemy

Trichopoulos can be something of a fatalist about the disease he has studied for more than 40 years. "Perhaps half of all cancer is going to be with us, no matter how much we advance our knowledge, because of chance mutations" as people age, he says. "It is the price we pay for being big, long-living animals."
Yet the nimble, roving mind, confronted with the limitations of human life, takes inspiration from scientific methods. "I'm optimistic about the way things are going," he says, showing a printout of mortality statistics for Europe. You could ski down the trend lines.

"Mortality is declining in every developed country," he says. "Declining amazingly so. Mortality, death--is there any greater enemy?"

Peter Wehrwein is a health and science writer in Boston.




HSPH Review Text Version Home

This page is maintained by the Harvard School of Public Health Office for Resource Development, Communications
Copyright, 2004, President and Fellows of Harvard College