In Nurses' Lives, a Treasure Trove of Health Data

By Carol Kaesuk Yoon Copyright 1998 New York Times September 15, 1998

More than 20 years ago, a doctor who was curious about the effects of long-term oral contraceptive use sent questionnaires to 370,000 registered nurses. The answers he and colleagues received have turned into the largest and longest repeated follow-up study of a group of women, a fountain of information about health, in particular women's health.

The project, known as the Harvard Nurses' Health Study, has yielded more than 200 published findings, including these:

*Hormone replacement therapy can increase the risk of breast cancer and can decrease the risk of heart disease and osteoporosis.

*Margarine can increase the risk of heart disease.

*Vitamin E can protect against heart disease.

*One or two drinks a day can increase the risk of breast cancer while protecting against heart disease.

These findings, and many more, have altered how people eat and drink and live their lives. Some scientists call the study the single greatest contributor to understanding women's health.

But this huge scientific effort has only begun to pay its dividends. With many of the original participants now entering their 60's and 70's, the study promises to bequeath an even greater wealth of knowledge, much of it otherwise unattainable, about how a lifetime of choices in food, drink, exercise and medications can translate into protection from or increased risk of deadly and debilitating diseases.

"It was a brilliant idea, a unique idea," said Dr. Leslie Bernstein, an epidemiologist at the University of Southern California. "It's one of the largest studies ever to have been assembled. They were leaders, ahead of their time. Now everybody is trying to emulate what they've done."

The study began in 1976 when Dr. Frank E. Speizer, the Edward H. Kass professor of medicine at Harvard Medical School and co-director of the Channing Laboratory at Brigham and Women's Hospital, sent the first health questionnaires and received more than 120,000 responses.

"As we began to gather data it became very clear that we could expand the study," Dr. Speizer recalled. "The excellent cooperation of the nurses is the most important part."

More than 90 percent of the original respondents still answer a questionnaire every other year, telling what they eat, what medicines they take, what illnesses they have had and whether they drink, smoke, exercise or take vitamins -- among other things.

The fact that those answering the questions are nurses means their answers are reliable, the researchers say. Plus, the researchers can gather blood and other samples, simply by sending a request and a few collection supplies, something that could never be done with the public.

By knowing these women's experiences, researchers can look at the histories of those who have contracted diseases and compare them with those who have not and begin teasing out the important factors in determining a woman's risk for a particular illness.

Since the study began, Dr. Speizer has been joined by many scientists from around the world seeking to capitalize on this unique trove of data. Researchers have been able to identify risk factors for more common diseases, like breast, colon and lung cancer, diabetes and heart disease. In the future, researchers say they will be able to identify risk factors for rarer conditions, like lymphoma, kidney cancer and bladder cancer, as well as afflictions more common later in life, like stroke and senility.

"Everyone is afraid of losing their memory," said Dr. Francine Grodstein, an epidemiologist at Brigham and Women's Hospital who is leading the new effort to determine whether using aspirin, taking hormones or other life style choices prevents declines in cognitive function.

Among other things, Dr. Grodstein is expanding the study with a telephone test that researchers will repeatedly administer to participants over time. Researchers have spoken with about 2,000 of the women, asking questions like "Who is the President?" and reading a list of 10 words, which the participant is asked to repeat, all typical tests of mental function.

Researchers are also beginning to study the blood samples of 30,000 women, looking for pesticide residue, antioxidants and naturally produced hormones. Over time they will be able to discern what connection, if any, these or other substances have to the incidence of cancer. They can also use the blood samples to look for genetic factors.

One person could not predict her fate by examining the findings of the study -- the numbers produced are simply average outcomes over a range of women with different genetic backgrounds. By looking at the variation in the genes of women who contract or do not contract diseases with the wealth of data on their exposure to a variety of substances, researchers hope to begin to solve this most complex of problems.

"Hopefully, we can help women understand what life style changes are most important to them with their genetic makeup to prevent disease," said Dr. Sue Hankinson, an epidemiologist at the Harvard School of Public Health who is leading the blood work. "Ultimately we could provide a much more individual kind of recommendation."

As the women in the study age, researchers will also be able to continue to identify protective or beneficial diets, activities and habits associated with a healthier, longer life.

By 1989, the study had already proved to be such a good source of information that the team of researchers began a study of younger nurses, enrolled at an earlier age, called Nurses II. Researchers have even enlisted the children of the women in Nurses II in another study to look at the risk factors for adolescent obesity, which Dr. Graham Colditz, an epidemiologist at Harvard Medical School heading the study, described as reaching epidemic proportions.

After so many years of working together, the study has produced a bond of deep mutual respect between researchers and the nurses, whom they call their collaborators.

"This group of women has been so tremendously dedicated," said Dr. JoAnn E. Manson, an endocrinologist at Harvard Medical School, echoing her many colleagues' sentiments. "They have been the trailblazers in terms of helping us to get the answers."

The nurses themselves say they actually look forward to hearing from the researchers.

"It's been wonderful," said Irene Kilroy, 62, a nurse at Good Samaritan Hospice in Brighton, Mass., who said she always answers the questionnaire the day it arrives. Mrs. Kilroy, who was among the 800 nurses who gathered last year to celebrate the 20th anniversary of the study and its results, added, "We feel we've made a wonderful contribution."

The questionnaire varies with research interests, but has consistently grown, now running about six pages and taking the women from 20 minutes to more than an hour to complete. Using a form, like an S.A.T. answer sheet, that will be optically scanned, the women record information like how often they eat specific foods with choices ranging from never to six times per day or more. They also write in more detailed information like which brand of vitamin or cereal or cooking oil they use.

All this self-examination has its effects, said Barbara Dwyer, 65, a nurse from Medford, Mass., who described herself as proud to be part of the study.

"It does make you think more about your weight and smoking and exercise," she said. "I did stop taking the hormone therapy, and I really watched my weight more."

So it is with mixed feelings, at best, that the researchers wait for enough of these women to succumb to disease and die, the necessary first step before the database will be, in the euphemistic language researchers adopt, "rich enough" or "powerful enough" to offer useful information about illness and death.

Dr. Meir J. Stampfer, an epidemiologist at the Harvard School of Public Health who reviews the death records and officially assigns a cause of death to participants, describes the process as "a little creepy."

The study has become a model for many other projects around the country. The Iowa Women's Health Study, a project started in 1986 with 42,000 women who were licensed drivers in Iowa, studies body fat and diet and disease using the Harvard questionnaire. The Teachers Study in California, begun in 1996, has enlisted 135,000 women in cancer research.

At first, skeptics abounded.

Dr. Walter Willett, an epidemiologist and physician at the Harvard School of Public Health, who has been with the study since its second year, said that early on there was great skepticism over the worth of such an endeavor.

"When we started, smoking was thought not to be a risk factor for heart disease in women," Dr. Willett said, "People said there was no variation in diet and that people couldn't remember what they ate, so the whole field was hopeless. You only have to scratch the surface to see that people eat differently for many different reasons. These are things we take for granted now."

In addition, because the study is based on responses to questionnaires, it can only identify associations between risk factors and diseases, unlike a randomized, clinical trial in which participants are assigned to specific treatments, testing whether a risk factor causes a disease. As a result, some have questioned the validity, in particular, of the study's more surprising results.

"The Nurses' Health Study was basically jeered at and taken to task for the finding that there was a connection between alcohol and breast cancer," Dr. Stampfer said. "We were mocked for coming up with a seemingly silly and unsubstantiated result, but now it's turned out to be pretty well accepted."

And researchers are quick to point out that many of the things the Nurses' Health Study looks at cannot be assigned as part of a tidy, experimental design.

"There's never going to be a randomized study of alcohol," Dr. Stampfer said. "There are ethical issues, compliance issues. It can never happen, and the same is true with a lot of the dietary things that we study." The same could be said of birth-control methods and physical activity levels, also studied by the Harvard group.

In addition, a study like the nurses' study can pick up on long-term or delayed effects of diet and other habits, because it has run for so much longer than nearly any randomized trial. In real life, as opposed to a clinical trial, people often change what they do over the years, altering their diet, taking up hormone use, dropping it, taking it up again, quitting smoking and much more. "We keep track of all the changes," Dr. Willett said. "If we didn't, we'd miss the important effects."

Most of the nurses reliably return their questionnaires, but those who do not can count on help from researchers -- who may send as many as six additional questionnaires before they try telephoning. If that does not work, researchers turn to the post office, computer databases and other sources of information to track them down.

"I thought it'd be just a couple of years that I'd be doing this," said Mrs. Dwyer, recalling when she received her first questionnaire in the mail so long ago. "I'm not even sure how long it's been. One year I didn't send it back, but they kept after me until I did. They keep sending it to you until you feel almost guilty. And after awhile you think, you've done it all these years -- why stop now?"

Comments on this posting?
Click here to post a public comment on the Trash Talk Bulletin Board.
Click here to send a private comment to the Junkman.

Material presented on this home page constitutes opinion of Steven J. Milloy.
Copyright © 1998 Steven J. Milloy. All rights reserved on original material. Material copyrighted by others is used either with permission or under a claim of "fair use." Site developed and hosted by WestLake Solutions, Inc.