Mammograms increase breast cancer risk


Overdosing on Health Risks

Americans follow news of medical research as closely as sports or the stock market. We are particularly avid for new findings about the health effects of diet and life style, because we have come to believe that no one gets sick anymore just because of bad luck. Instead, we see health as largely a matter of doing the right things, with the corollary that illness is a failure of some sort. If we can keep up with all the new research and change our habits accordingly, we will stay well.

At The New England Journal of Medicine, where I am executive editor, we often witness the impact that research papers can have on the public. No sooner do we publish a study on diet or life style than news of its conclusions, though virtually none of its qualifying details, hits the airwaves. Within 24 hours, millions of people consider eating fewer egg yolks or more oat bran to fend off disease.

This meritocratic faith ignores one critical fact: science has hardly begun to touch the big mysteries about diet and other habits. We simply do not know much about what is risky and what isn't, and what we do know is often distorted or misinterpreted. Most Americans are not very good at distinguishing big risks from little ones, or risks based on solid evidence from those that aren't. We are likely to react to all reported risks in the same way, or choose which to respond to, on the basis of irrational fears or the prominence given them on the 6 o'clock news.

The problems in evaluating risks were very evident in the controversy over mammograms for women in their 40's. In January many people were outraged when a National Institutes of Health panel of experts failed to recommend regular mammograms for women in this age group. The panel concluded that regular mammograms between 40 and 49 would at best save the life of only 1 woman for every 1,000 screened. Offsetting that benefit would be the increased risk of cancer from the extra radiation -- perhaps three more cancers for every 10,000 women screened. In addition, because pre-menopausal women have denser breast tissue, mammograms are more difficult to interpret in women under 50. As a consequence, most women in this age group with suspicious-looking mammograms would turn out not to have cancer, but because of the misleading mammogram, they might wind up undergoing unnecessary surgery. In short, so small would be the payoff of regular mammograms at this age that the risks of driving the car to get them might well outweigh the benefits of the test.

Yet many people reacted to the panel's statement as though its members had callously sentenced large numbers of women to death. The reaction was so intense that in March a second panel reversed the recommendation. One reason for the outrage was the disproportionate fear of this disease. Even though breast cancer is the most common cause of cancer death in women in their 40's, the disease still affects fewer than 2 percent of women in this age group. After age 50, the death rate from breast cancer rises rapidly, but over a woman's lifespan, it lags behind lung-cancer deaths and never comes close to the death rate from heart disease. Heart attacks kill about six times as many women as breast cancer. Still, many women fear breast cancer far more. Because of this fear, the risk seems much larger than it is.

Some risks are, of course, a lot better grounded than others. The evidence that smoking cigarettes causes lung cancer is incontrovertible. It has been found consistently in multiple studies of various types over many years. But not many risks are established by such clear-cut evidence. Smaller risks may be very difficult to demonstrate, and often the research is inconsistent or even contradictory. For example, two years ago both The New England Journal of Medicine and The Journal of the American Medical Association published studies within four weeks of each other that came to opposite conclusions about whether taking post-menopausal estrogen increases the risk of breast cancer. What is a health-conscious American supposed to believe?

A big part of the problem lies in unrealistic public expectations. Many people want science to provide quick, unequivocal answers with immediate implications -- not just get it right, but get it right the first time around. Unfortunately, inconsistency is common in medical research, particularly epidemiological research about diet and life style. Instead of becoming frustrated or feeling betrayed by the disagreements, we should see them as a message to be more cautious about accepting the results of any one study. Inconsistency teaches scientists to be cautious, and there is no reason why the public can't learn the same lesson. In general, scientists and doctors do not embrace the conclusions of a single epidemiological study until they are supported by other studies. Only with a fairly large body of evidence are we really able to evaluate risks. When multiple studies still can't give an answer, the risk being studied is probably very small -- or nonexistent.

Journalists -- and even scientists themselves -- are partly responsible for the way risks are inflated or distorted in health news. Not surprisingly, news of a big risk attracts more attention than news of a small or uncertain one. The quality of health reporting is, of course, highly variable, and some of it is excellent, but many reporters succumb to the temptation to exaggerate. Even if reporters are appropriately circumspect, their headline writers may have other ideas and blow medical stories out of proportion. (The word breakthrough is a favorite of headline writers.) The sound-bite culture of television particularly lends itself to hyperbole. For their part, scientists and their research institutions are increasingly seeking the limelight. Public relations offices distribute inflated news releases of research results, and scientists are pressured by institutions and professional societies to present their findings to the media. Under these circumstances, some researchers find it difficult to resist overstating the significance of their work.

Framed in a certain way, then, a very small risk can look big. Take the study that found postmenopausal estrogen is associated with a 30 percent increase in the risk of breast cancer. That sounds like a lot. But the same risk can be expressed in much less alarming ways. For example, since we already know that 3 or 4 percent of post-menopausal women will get breast cancer in the next 10 years, we could say that this study shows that estrogen increases that risk to 5 percent. Put it yet another way, if you are a postmenopausal woman trying to decide whether to take estrogen, this study shows that your chances of remaining free of breast cancer for 10 years would decrease from over 96 percent to about 95 percent. These ways of expressing the same finding have very different psychological effects, even though they are saying the same thing.

There is another reason for confusion about risks: without anyone explicitly admitting it, the focus in health care is shifting from the individual to whole populations. The new emphasis on groups is. partly a result of increasing concern about national health care expenditures and the cost of covering groups of beneficiaries under managed care. Payers of benefits want to know what health recommendations will make the biggest difference to the group as a whole. But risks and benefits that are important across a large population are not necessarily of much importance to an individual. For example, a research study a decade or so ago showed that using a drug to lower blood cholesterol in middle-aged men with high cholesterol levels reduced their seven-year risk of heart attacks from 8.6 to 7 percent. Although this tiny improvement might not seem worthwhile to an individual particularly if it means taking a drug with side effects, when spread over the estimated 1 to 2 million Americans with similar cholesterol levels, it could account for some 32,000 fewer heart attacks over the first seven years. Many people believe that high blood levels of cholesterol are practically a death sentence, while low cholesterol means they will be forever free of heart disease. They hear the public health message and unthinkingly apply it to themselves, making great changes in their habits and diets for little individual gain.

As a result of this shift to thinking in terms of large populations, health recommendations may not account sufficiently for individual variations. The first panel on mammography broke with the trend toward blanket recommendations by concluding that women under 50 should decide for themselves whether to get routine mammograms. To many that was evasive, and maybe in part it was. (It was also patronizing, since it implied that women of all ages do not now decide for themselves.) But it underscored an important point. Decisions about mammography should be tailored to the needs and wishes of each woman. Mammography might make more sense, for example, in a 45-year-old woman with fatty breasts, especially if she had a relative with breast cancer, than it would in a woman the same age with dense breasts and no family history of the disease. That is because breast cancer often runs in families and mammograms are more reliable when breasts are fatty. Mammography also makes more sense for a woman with a particular dread of cancer, as opposed to a woman less concerned. It is striking that some people, even as they prize their new sense of control over their own health, feel betrayed when they are not told what to do.

So what should we make of the bombardment of news about health risks? The short answer is that we should be a lot more skeptical. Unless the risk is very large, the results make sense and the change in life style would be minimal, there is no reason to change habits on the basis of one study. Instead, we should wait until the risk is confirmed by other research, and then ask ourselves how the findings apply to us as individuals. We should also remember that medical journals publish works in progress for other scientists and doctors; the papers are not meant as the final word for the public. We would all like to believe that changes in the way we eat and live can greatly improve our health, but with our present state of knowledge, the likelihood is that -- with a few exceptions like giving up smoking -- many if not most such changes will produce only small effects (and large profits for the burgeoning industries that thrive on health promotion). We need to understand better the difficulties and limitations of this type of research. There's more to life than fretting about health risks.

YouTube video

Article copyright American Council on Science and Health, Inc.


By Marcia Angell

Share this with your friends