Healthy Women, Healthy Lives A Guide to Preventing Disease, from the Landmark Nurses' Health Study

by Hankinson, Susan E.; Manson, JoAnn E.; Speizer, Frank E.; Colditz, Graham A.

• ISBN13: 9780743217743
• ISBN10: 0743217748
• Edition: Reprint
• Format: Paperback
• Copyright: 2002-07-02
• Publisher: Free Press

Summary

Since 1976, the world-famous Harvard Medical School Nurses' Health Study has followed more than 120,000 real women, leading real lives, to discover what factors contribute to improving the health of women. The most important findings are made accessible to the general public in this easy-to-understand book that will revolutionize the way women live.Healthy Women, Healthy Livesgoes beyond simply labeling preventive measures and risky behavior -- it provides practical tips and strategies from clinical experts at Harvard Medical School for making healthy lifestyle changes. Here are the best ways to lower the risk of a host of chronic diseases, as well as tips for losing weight, stopping smoking, eating healthily, and exercising regularly. With easy-to-read graphs that clarify complex information and personal stories from nurses who have contributed to the remarkable study,Healthy Women, Healthy Livesis an extraordinary health book that will prove invaluable to women everywhere.

Table of Contents

Acknowledgments	vii
Introduction	xix
Donna Shalala
Preface	xxi
Contributors	xxv
PART ONE: GETTING STARTED
An Exciting Time for Women's Health	3	(6)
Personal Health Choices	4	(3)
Health Choices in the Real World	7	(2)
What Is the Nurses' Health Study?	9	(11)
The Nurses' Health Study	9	(5)
Other Research Studies	14	(6)
Making Sense of Health Research	20	(13)
Absolute Risk	20	(1)
Relative Risk	21	(5)
Drawing Conclusions	26	(2)
The Boundaries of Research	28	(5)
PART TWO: LOWERING THE RISK OF DISEASES
Lowering the Risk of Coronary Heart Disease	33	(37)
Background	33	(4)
Factors You Cannot Control	37	(2)
Factors You Can Control	39	(24)
What It All Means	63	(2)
What I Tell My Patients About Lowering Their Risk of Coronary Heart Disease	65	(5)
Dr. Carol Bates
Lowering the Risk of Breast Cancer	70	(40)
Background	70	(5)
Factors You Cannot Control	75	(8)
Factors You Can Control	83	(11)
Reproductive Factors	94	(7)
What It All Means	101	(1)
What I Tell My Patients About Lowering Their Risk of Breast Cancer	102	(8)
Dr. Nancy Rigotti
Lowering the Risk of Lung Cancer	110	(10)
Background	110	(2)
Factors You Cannot Control	112	(2)
Factors You Can Control	114	(4)
What It All Means	118	(1)
What I Tell My Patients About Lowering Their Risk of Lung Cancer	118	(2)
Dr. Celeste Robb Nicholson
Lowering the Risk of Stroke	120	(26)
Background	120	(5)
Factors You Cannot Control	125	(2)
Factors You Can Control	127	(14)
What It All Means	141	(2)
What I Tell My Patients About Lowering Their Risk of Stroke	143	(3)
Dr. Carol Bates
Lowering the Risk of Diabetes	146	(23)
Background	146	(7)
Factors You Cannot Control	153	(1)
Factors You Can Control	154	(11)
What It All Means	165	(1)
What I Tell My Patients About Lowering Their Risk of Adult-Onset Diabetes	165	(4)
Dr. Nancy Rigotti
Lowering the Risk of Colon Cancer	169	(20)
Background	169	(3)
Factors You Cannot Control	172	(2)
Factors You Can Control	174	(10)
What It All Means	184	(1)
What I Tell My Patients About Lowering Their Risk of Colon Cancer	185	(4)
Dr. Carol Bates
Lowering the Risk of Osteoporosis	189	(18)
Background	189	(3)
Factors You Cannot Control	192	(2)
Factors You Can Control	194	(9)
What I All Means	203	(1)
What I Tell My Patients About Lowering Their Risk of Osteoporosis	203	(4)
Dr. Nancy Rigotti
Lowering the Risk of Endometrial Cancer	207	(14)
Background	207	(2)
Factors You Cannot Control	209	(1)
Factors You Can Control	210	(5)
Reproductive Factors	215	(3)
What It All Means	218	(1)
What I Tell My Patients About Lowering Their Risk of Endometrial Cancer	218	(3)
Dr. Celeste Robb Nicholson
Lowering the Risk of Ovarian Cancer	221	(15)
Background	221	(3)
Factors You Cannot Control	224	(2)
Factors You Can Control	226	(4)
Reproductive Factors	230	(3)
What It All Means	233	(1)
What I Tell My Patients About Lowering Their Risk of Ovarian Cancer	234	(2)
Dr. Celeste Robb Nicholson
Lowering the Risk of Skin Cancer	236	(18)
Background	236	(3)
Factors You Cannot Control	239	(4)
Factors You Can Control	243	(4)
What It All Means	247	(1)
What I Tell My Patients About Lowering Their Risk of Skin Cancer	248	(6)
Dr. Carol Bates
Lowering the Risk of Asthma, Arthritis, Age Related Eye Disease, and Alzheimer's Disease	254	(21)
Asthma	254	(5)
Arthritis	259	(3)
Age-Related Eye Disease	262	(5)
Alzheimer's Disease	267	(4)
What I Tell My Patients About Asthma, Arthritis, Age-Related Eye Disease, and Alzheimer's Disease	271	(4)
Dr. Carol Bates
PART THREE: CHANGING BEHAVIORS
Physical Activity	275	(18)
Background	275	(4)
The Benefits of Physical Activity	279	(8)
The Risks of Physical Activity	287	(3)
What It All Means	290	(1)
What I Tell My Patients About Physical Activity	290	(3)
Dr. Celeste Robb-Nicholson
Weight Control	293	(26)
Background	293	(4)
The Risks of Being Overweight and Gaining Weight	297	(14)
A Few Benefits of Being Overweight and Gaining Weight	311	(3)
What It All Means	314	(2)
What I Tell My Patients About Weight Control	316	(3)
Dr. Nancy Rigotti
Smoking	319	(30)
Background	319	(3)
The Risks of Smoking	322	(14)
Passive Smoking	336	(3)
What It All Means	339	(1)
What I Tell My Patients About Successfully Quitting Smoking	339	(10)
Dr. Nancy Rigotti
The Major Nutrients---Fat, Carbohydrates, Fiber, and Protein	349	(23)
Background	349	(2)
Dietary Fat	351	(6)
Carbohydrates	357	(5)
Fiber	362	(3)
Protein	365	(4)
What It All Means	369	(1)
What I Tell My Patients About the Major Nutrients	369	(3)
Dr. Celeste Robb-Nicholson
Foods	372	(18)
Background	372	(3)
Fruits and Vegetables	375	(3)
Nuts	378	(1)
Whole Grain Foods	378	(4)
Red Meat	382	(1)
Fish	383	(1)
Eggs	384	(2)
What It All Means	386	(1)
What I Tell My Patients About Foods	386	(4)
Dr. Celeste Robb-Nicholson
Alcohol	390	(10)
Background	390	(1)
The Benefits of Moderate Alcohol Consumption	391	(2)
The Risks of Moderate Alcohol Consumption	393	(3)
What It All Means	396	(1)
What I Tell My Patients About Drinking Alcohol	397	(3)
Dr. Celeste Robb-Nicholson
Vitamins and Minerals	400	(28)
Background	400	(4)
Vitamin A and Carotenoids	404	(4)
Vitamin D	408	(1)
Vitamin E	409	(2)
Vitamin C	411	(1)
Folate	412	(3)
Vitamins B6 and B12	415	(2)
Sodium (Salt)	417	(1)
Calcium	418	(3)
Magnesium	421	(1)
Potassium	422	(1)
Selenium	423	(1)
Iron	424	(1)
What It All Means---Vitamins and Minerals	425	(1)
What I Tell My Patients About Vitamins and Minerals	426	(2)
Dr. Nancy Rigotti
Postmenopausal Hormones	428	(31)
Background	428	(2)
The Benefits of Postmenopausal Hormones	430	(7)
The Risks of Postmenopausal Hormones	437	(5)
What It All Means	442	(1)
Strategies for Postmenopausal Hormone Use	443	(6)
What I Tell My Patients About Postmenopausal Hormones	449	(10)
Dr. Celeste Robb-Nicholson
Birth Control	459	(26)
Background	459	(1)
Birth Control Pills	460	(9)
Hormonal Implants and Injections	469	(1)
IUDs (Intrauterine Devices)	470	(2)
Barrier Methods	472	(2)
Tubal Ligation	474	(3)
Natural Methods	477	(1)
Strategies for Birth Control Use	478	(5)
What I Tell My Patients About Birth Control	483	(2)
Dr. Carol Bates
Aspirin and Other Nonsteroidal Anti-inflammatory Drugs (NSAIDs)	485	(16)
Background	485	(2)
Benefits of Aspirin and Other NSAIDs	487	(7)
Risks of Aspirin and Other NSAIDs	494	(2)
What It All Means	496	(1)
What I Tell My Patients About the Proper and Safe Use of Aspirin and Other NSAIDs	497	(4)
Dr. Carol Bates
Appendix A Types of Epidemiologic Studies	501	(4)
Appendix B Being an Informed Consumer of Health Information	505	(4)
Appendix C Some Useful Tables	509	(4)
Glossary	513	(8)
Selected Readings	521	(4)
Index	525

Excerpts

Chapter Two: What Is the Nurses' Health Study?

The Nurses' Health Study

The Question That Got Us Started: "Do Birth Control Pills Have Long-Term Health Effects?"

More than twenty-five years ago, this question sparked the beginning of what has since become one of the largest and most comprehensive studies of women's health. From birth control pills to exercise, from alcohol to zinc, the Nurses' Health Study has helped to identify many of the factors that protect health, as well as those that contribute to disease.

The study began in the mid-1970s, when birth control pills had been in widespread use for about a decade. Pleased with their effectiveness and convenience, many women were intending to use the pills for decades to come. Yet there was considerable uncertainty about whether this practice was safe. Taking the pill had already been shown to increase the risk of blood clots, heart attacks, and stroke. And as other side effects of the pill began to emerge, researchers also suspected a link with breast cancer. It was the possibility of this last link that led us to propose a bold new study: we would survey thousands of women about their method of birth control and then track their health status over time.

Because birth control pills were quickly emerging as a public health issue, the National Institutes of Health (the federal agency primarily responsible for prioritizing and funding medical research in the United States) was willing to consider funding our study. However, they first required that we submit a grant proposal so that our scientific peers could review what we had done so far -- and what we planned to do if our study was funded. We submitted a grant proposal in 1973 and received the funding to start the study in 1974. After two years of pilot-testing the methods for the study, we began recruiting participants in 1976.

FINDING THE NURSES

Our first major challenge was to identify a large group of women -- ideally tens of thousands -- who would be willing to complete a questionnaire about their method of birth control and their health status. In addition, the information they provided had to be accurate, and we had to be able to follow the women over time to determine if the pill had any long-term effects.

We realized that nurses would be the ideal study participants -- and in fact they have been. As health professionals, they are extremely aware of the value of medical research and have thus been willing to make a long-term commitment to the study. In addition, they have been relatively easy to follow over the past twenty-five years, largely because they are prompt in notifying us of address changes. Finally, and perhaps most important, because of their knowledge and training, the nurses have been able to provide us with health information that is both accurate and reliable. This accuracy has been documented repeatedly. For example, whenever a study participant reports that she has been diagnosed with a serious illness, we ask permission to review her medical records to both confirm the diagnosis and gather additional details about the disease. The vast majority of the time we find that the nurse has provided us with accurate information.

BEGINNING THE STUDY

Having chosen nurses as our study participants, we contacted the boards of nursing in eleven U.S. states and obtained the names and addresses of 170,000 female nurses. They were all married and between the ages of thirty and fifty-five to ensure that a large number would have used birth control pills. They were also primarily white due to the makeup of the nursing population at the time.

We sent our first questionnaire to these nurses in 1976. Though fairly straightforward, this questionnaire was actually the product of much deliberation and debate. The questions were designed so that they could be answered relatively easily, while still providing us with enough information to establish scientific relationships. For example, we asked the women if they had ever used birth control pills, and if so, for how long. We also asked about various medical conditions: had they ever had breast disease or a heart condition, and if so, when were they diagnosed? Because we already knew some of the factors that were linked to these conditions, we asked about those as well: how many children did they have, had they ever smoked, and had their parents developed breast cancer or heart disease?

Despite the depth of these questions, the entire survey was only two pages. This helped minimize the amount of time and effort needed to complete the survey and also helped maximize the number of responses we received. More than 120,000 nurses responded to our first questionnaire, and in doing so, became part of one of the largest and longest-running studies of women's health in the world.

TWENTY-FIVE YEARS OF LOYALTY

For the past twenty-five years, the nurses in the study have completed a questionnaire every other year. Not surprisingly, given the duration and scope of the study, the questions included in our biennial survey have changed substantially over time. While there are certain questions that we have asked on every survey (such as current weight and smoking status), many different questions have been added or dropped throughout the years. This is for several reasons.

First, the nurses have obviously grown older during the course of the study, and certain behaviors are no longer relevant to them. For example, during the first decade, we routinely asked women to update us about their contraceptive methods and childbearing. However, as the majority of the women reached menopause, these questions became less relevant and were replaced with others.

Second, the questionnaires have changed over time to reflect the emergence of new and potentially important issues in women's health. For example, in recent years scientists have suggested that a woman's health in middle age might be determined in part by how much she weighed as a baby. To address this possibility, we added to our 1992 survey a question about birth weight.

Finally, our questionnaires have evolved as a function not only of our own interests but also those of the nurses. Study participants repeatedly told us that they were concerned about their quality of life and how it was affecting their health. In response, we added a series of questions about caregiving, work stress, and social relationships. Although our questionnaires are much longer and more detailed than they were originally, more than 90 percent of the women continue to complete them.

ADDING BIOLOGICAL SAMPLES TO THE STUDY

Over the years, many of the women in the study have provided more than questionnaire responses; they have literally given a piece of themselves. On a number of occasions, when questionnaires could not provide the information we needed, we asked the nurses to send us biological samples.

The first such request was prompted by scientific evidence that the mineral selenium might help prevent certain cancers. Though we typically do not need biological samples to study the effects of dietary components, selenium is somewhat different. For most minerals, the amount found in one vegetable is equivalent to the amount found in another vegetable of the same type. However, the same is not true for selenium. The amount of selenium in one carrot may be vastly different from that of another carrot, depending on how much selenium was in the soil where the carrot grew. Thus, even if we know how many carrots (and other selenium-rich foods) a woman eats, we still will not know how much selenium she consumes. To learn this, we needed to ask for a piece of each nurse in the study: her toenail clippings.

Toenail clippings can provide an accurate indication of how much selenium, as well as other trace materials, a woman has consumed in the past several months. In 1982 we asked 92,000 participants to mail nail clippings from each toe to our research center in Boston; more than 68,000 women did so.

In 1988, we made our second request for a biological sample. This time we asked the nurses to send us several tubes of their blood. Again, the response was overwhelming: we received about 33,000 samples of blood, some of which have been analyzed for cholesterol, hormone levels, environmental chemicals, and even genetic markers.

We have since made other requests for biological samples, including cheek cells and additional blood. All of these samples have allowed us to study scientific relationships that we would otherwise never have been able to study. For this, we are incredibly grateful to the thousands of nurses who provided their samples.

LEARNING FROM THE NURSES WHILE PROTECTING THEIR PRIVACY

Despite having amassed huge amounts of information about the women in our study, we take great care to ensure that no one but researchers can ever link that information back to the individuals who provided it. Even the researchers have limited access, with the information available only on a need-to-know basis. Each woman in the Nurses' Health Study was assigned an identification number in 1976, and her information is identifiable by that number only. Her questionnaire data are stored in one secure computer system, and her name and address in another. This helps ensure that the privacy and confidentiality of study participants are protected at all times. We also have a certificate of confidentiality from the Department of Health and Human Services to further ensure that the data provided are kept confidential.

Other Research Studies

In the last two decades, the Nurses' Health Study has made many advances in identifying the factors that contribute to the development and prevention of major illnesses affecting women. However, we have not been alone in our efforts. There are many other important research studies on women's health, and together with the Nurses' Health Study, they have created the body of scientific evidence that guides health recommendations.

Three studies in particular have provided a great deal of evidence to support the topics covered in this book: the Framingham Heart Study, the American Cancer Society Cancer Prevention Study II, and the Iowa Women's Health Study. These ongoing studies have contributed substantially to the vast advances made in women's health in the past fifty years and will continue to do so for years to come. Two additional studies, the Nurses' Health Study II and the Women's Health Initiative, will also be contributing tremendously in the near future.

THE FRAMINGHAM HEART STUDY

In the 1930s, cardiovascular disease reached epidemic proportions in the United States, and by the late 1940s, it had become the leading cause of death among American men and women. In the hope of learning how to prevent this deadly disease, researchers at the U.S. Public Health Service started a series of studies, including the Framingham Heart Study in 1948. This study has tracked participants' health for the past fifty years and has contributed greatly to our understanding of cardiovascular disease, as well as a variety of other diseases and conditions. As its name indicates, the study is made up of men and women who lived in Framingham, Massachusetts, at the time the study began. This town had a very stable population, and most residents obtained their health care from one major hospital in the area; both of these factors made it easier for researchers to follow the participants over time.

When the study began, the 5,200 participants were between the ages of thirty and sixty and were free from heart disease. Since then, surviving participants have undergone extensive medical exams every two years consisting of detailed medical histories, laboratory tests, and interviews about lifestyle. The Framingham Heart Study is notable not only for its duration and the extended commitment of its participants, but also because it was one of the first major cardiovascular studies to recruit women.

THE AMERICAN CANCER SOCIETY CANCER PREVENTION STUDY II

In September of 1982, volunteers from the American Cancer Society invited their friends, neighbors, and relatives to join the Cancer Prevention Study II, which was similar to another study conducted by the American Cancer Society in the 1960s. Within a month and a half, over a million men and women had joined the second study and completed a four-page questionnaire about their health habits and personal and family medical history. Although most participants have not provided information on their lifestyles since then, researchers have continued to track their health status over the years. They have done this by having volunteers make personal inquiries about the friends and family members they recruited and by checking national death records. This study is notable particularly for its size (it includes over 675,000 women) and for its nineteen-year duration.

THE IOWA WOMEN'S HEALTH STUDY

Hoping to determine which lifestyle factors might be related to the development of cancer in older women, researchers at the University of Minnesota in 1986 sent a sixteen-page questionnaire to 100,000 Iowan women between the ages of fifty-five and sixty-nine. About 41,000 women responded and completed subsequent questionnaires.

The questionnaires used in this study were similar to those used in the Nurses' Health Study, with one notable exception: they did not ask women whether they had been diagnosed with cancer. Without this pertinent information, how can researchers possibly determine a link between lifestyle factors and cancer risk? The answer is relatively simple and explains why researchers chose Iowa as their study site: whenever someone in Iowa is diagnosed with any type of cancer, it is immediately reported to the State Health Registry. Each year, the Registry provides the study's researchers with a list of participants who have been diagnosed with cancer. For the past fifteen years, researchers have successfully tracked women's health in this manner and produced dozens of reports about what leads to cancer and what lowers the risk of cancer.

THE NURSES' HEALTH STUDY II

This study, which has grown out of the ongoing Nurses' Health Study, began in 1989, thirteen years after the start of the original study. The purpose was to examine reproductive health issues that had arisen in recent years and could not be addressed by the original study. For example, birth control pills contained substantially lower doses of hormones in 1989 than they did when the women in the Nurses' Health Study were taking them. In addition, most of the women in the Nurses' Health Study had already been pregnant by the time birth control pills became available. To study the health effects of the low dose pills, particularly before a first pregnancy, we decided to start a study with younger women.

In 1989, we mailed a four-page questionnaire to tens of thousands of nurses between the ages of twenty-five and forty-two and received responses from more than 116,000. These nurses have continued to complete questionnaires every two years not only about their reproductive history but also about their health status, diet, physical activity, and medication use. For any given questionnaire, about 90 percent of the nurses respond, which is a reflection of their commitment and dedication to this study. Overall, the Nurses' Health Study II promises to provide a better understanding of how women's lifestyle choices in early childhood affect their health both then and in later years.

THE WOMEN'S HEALTH INITIATIVE

The Women's Health Initiative is a research effort that began in 1992 and eventually included forty different clinical centers across the United States. At each center, researchers invited postmenopausal women between the ages of fifty and seventy-nine to take part in a nine-year randomized controlled trial.

The Randomized Controlled Trial.This type of study is quite different from the other studies in this chapter (including the Nurses' Health Study), which are called observational studies. In observational studies, the participants live as most women do -- they choose their own diets, exercise when they want to, and take the medications their health care providers prescribe. They regularly report their activities to researchers, who then use this information to determine how lifestyle choices affect health. In a randomized controlled trial, participants are more like subjects in a traditional scientific experiment. They are asked to adhere to a certain diet or take a particular medication so that researchers can study the health effects of the assigned diet or medication. Appendix A explains the different types of epidemiologic studies further.

In the Women's Health Initiative (WHI), researchers are conducting three separate trials to determine the health effects of three specific factors: low fat diets, postmenopausal hormones, and calcium and vitamin D supplements. In the first study, which includes nearly 50,000 women, researchers have asked about half the women to adhere to a special low fat diet (which was taught through several nutritional counseling sessions) and have allowed the other half to consume their usual diet. After the women have been on their respective diets for nine years, researchers will look to see which group has lower rates of breast cancer, colon cancer, and heart disease. In this way, they can determine whether the low fat diet carries a greater health benefit than the usual diet.

In the second trial, researchers assigned a daily dose of postmenopausal hormones to one group of women and a daily placebo (or sugar pill) to the other group. Again, researchers will track the health status of the women in each group and eventually determine which group has lower rates of heart disease and fractures. In the third and final trial, researchers assigned half of the women to take daily supplements of calcium and vitamin D, and the other half to take daily placebos. From this trial, researchers hope to determine whether the supplements protect against hip fractures and colon cancer, as suggested by several smaller studies.

The Observational Study.In addition to the three randomized controlled trials in the Women's Health Initiative, there is also an observational study, which includes nearly 94,000 women who volunteered for the study but were unable or unwilling to participate in the trials described above. These women have completed annual questionnaires about their lifestyle practices and have undergone routine medical exams. With this observational study, researchers have a much broader aim than with the intervention studies: to examine the multitude of lifestyle factors that contribute to heart disease, stroke, cancer, fracture, diabetes, and other major illnesses in women.

Overall, the Women's Health Initiative promises to provide some of the most conclusive evidence yet about women's health. It is distinct from other studies in this chapter not only because of its interventional approach, but also because of its study population: nearly 20 percent of the women -- about 30,000 -- are African American, Latino, Asian American, or American Indian. This is in contrast to most other research studies, where more than 95 percent of all participants are white.

Conclusions

Although this book focuses on the Nurses' Health Study and other long-term observational studies, there are many additional studies that have also contributed to our understanding of women's health. We must rely on all of these if we are to reach firm conclusions about how women can lower their risk of chronic diseases. The most substantiated and well-accepted health information is that which has been supported by a number of different studies.

In this book, we balance findings from our own study within the context of other studies' results. For example, there are now over thirty studies on birth control pills and ovarian cancer. Taken collectively, these studies suggest that taking the pill does lower the risk of ovarian cancer. We use findings from the Nurses' Health Study to explain this relationship, while making clear that ours is just one of the many studies to have shown this. When our results contradict the larger body of evidence, we note that, too. By providing a balanced summary of what is currently known about women's health, we hope to guide women toward healthier lifestyle choices.

Copyright © 2001 by the President and Fellows of Harvard College

Chapter Five: Lowering the Risk of Breast Cancer

BACKGROUND

You likely recognize the ratio one in eight as an American woman's lifetime risk of developing breast cancer. In addition to being an alarming reminder to some women of how common breast cancer is, the one in eight ratio can also raise a host of important questions. Why is the lifetime risk so high? What causes the disease? And most important, how can I lower my risk? While the recent attention given to breast cancer by the media and various health organizations has increased general awareness of the disease, many women are still confused about the answers to these questions. And with media reports on scientific studies appearing to contradict each other on a weekly basis, it is easy to understand why. One story says eating too much fat increases the risk of breast cancer; another says the fat found in olive oil actually lowers risk; and the next week a different issue altogether goes through this type of cycle.

Despite the confusion that often seems to surround the topic, breast cancer deserves much of the attention it gets. Approximately 180,000 American women are newly diagnosed with the disease each year, and it is the leading killer of women in midlife (ages forty to fifty-five). The only cancer that kills more women overall is lung cancer, but breast cancer is actually responsible for more years of life lost because it generally afflicts younger women. On top of the physical health problems, breast cancer can also exact a large emotional and psychological toll, because it affects a part of the body that many women associate directly with their femininity. Surveys have consistently found that women perceive breast cancer to be their greatest health threat.

Over the last twenty years, substantial strides have been made in treating breast cancer and in identifying factors that contribute to its development. Today, quality of life is significantly better for women treated for breast cancer than in the past, and the rate of new cases of the disease has leveled off and even started to decline slightly in some groups of women. Much work remains, however, as overall rates of the disease remain high and not all women have benefited equally from the recent decrease in disease rates. The rate of new cases of breast cancer is actually increasing in African American women, and even though white women have higher rates of breast cancer overall, African American women are still more likely to die from the disease.

The Breast and Breast Cancer

Breasts are made up of several sections called lobes. Each lobe has smaller sections called lobules, which produce milk when a woman is breast-feeding. The lobes and lobules are linked by tubes called ducts. Ducts are the tubes that carry the milk from the lobules to the nipple. The rest of the breast is mostly made up of fat but also consists of veins, arteries, nerves, and other tissue.

In breast cancer, cells in the breast divide and grow at an abnormal rate, clump together, and form a malignant (cancerous) tumor. There are two main types of breast cancer. Most common is breast cancer that begins in the ducts and spreads to nearby tissue, called invasive ductal carcinoma. The other main type is breast cancer that begins in the lobes and spreads to nearby tissue, called invasive lobular carcinoma.

There is also a condition called carcinoma in situ, where there are abnormal cells in the breast, but they are not cancerous and have not spread to other nearby tissue. In situ means "in place." Still, carcinoma in situ is a sign that breast cancer may develop at a later time. The two main categories of carcinoma in situ are ductal carcinoma in situ (DCIS) -- where the abnormal cells originate in the milk ducts -- and lobular carcinoma in situ (LCIS) -- where the abnormal cells originate in the lobules.

The most common outward sign of breast cancer is a hard lump in the breast that is usually not movable and may or may not be painful. The skin over the lump may be thickened and dimpled (like the skin of an orange) or indented in areas where the cancer has spread. The nipple may be inverted (turned inward) or leak dark fluid.

The best chance of surviving breast cancer comes from the early detection of cancerous tumors through regular clinical breast exams and mammograms. Reducing the risk of the disease ever occurring, however, should be every woman's goal.

The Importance of Early Life in Breast Cancer Risk

Ideally the prevention of breast cancer would start in youth and young adulthood, as this period has been found to be extremely important in determining a woman's risk of developing breast cancer later in life. Breast tissue during this time (from youth up until a woman gives birth to her first child) appears to be more susceptible to elements that can cause cells to subsequently become cancerous. Exposure to radiation is a good example. The risk for adult breast cancer is very high for a girl under ten exposed to a high dose of radiation (such as exposure to the atomic bomb in 1945 Japan), moderate for a teenager, and small for a young adult.

Unfortunately, by the time an individual woman begins to think concretely about how she can lower her risk of breast cancer, this critical period has likely passed. Parents, however, can help their daughters adopt healthy lifestyles that can impact some of the factors that affect risk during young adulthood, as we discuss later in this chapter. Important factors include age at menarche (the first menstrual period), alcohol consumption, and smoking. Especially important is encouraging daughters to be physically active and to maintain a healthy weight, which, along with many other health benefits, can delay age at menarche.

It has been estimated that breast cancer rates could be reduced by as much as half in the United States by focusing prevention efforts on youth and making certain social changes (for example, so that women who give birth at an early age are not handicapped professionally). While the social changes to support such efforts may be far off, this estimate highlights the fact that rates of breast cancer in the United States can be reduced and that women are far from helpless in the fight against the disease.

LESSONS FROM THE NURSES' HEALTH STUDY AND OTHER STUDIES

To many women it would seem that just about everything affects the risk of developing breast cancer. While this is not the case, the scientific community has come a long way in identifying many factors that increase the risk of the disease. These factors can range from family history to lifestyle factors to reproductive variables.

Although there are a few factors that substantially increase a woman's risk of breast cancer, most individually elevate or lower risk by only a small amount. Some of these factors are those over which a woman has control (such as alcohol consumption and use of postmenopausal hormones) and some are not (such as breast cancer in an immediate family member and age at menarche). By knowing how each of these factors affects her risk, a woman can take appropriate steps to try to protect herself from the disease.

Ultimately, it seems that most cases of breast cancer are caused by a combination of factors, some of which are known and some of which are currently unknown.

Factors You Cannot Control

There are a number of factors beyond a woman's control that affect her risk of developing breast cancer. Understanding these may lead a woman to be more vigilant about those factors under her control and, if necessary, seek more frequent check ups or screening tests.

AGE

Age is a well-established risk factor for breast cancer. In general, the older a woman is, the greater her risk of the disease. National data show that rates of breast cancer are low in women under forty, begin to increase after forty, and are highest in women over seventy. Only after the age of about eighty does the risk stop increasing with age.

In a group of 100 women who are fifty years old, 9 will develop breast cancer before they reach the age of eighty. The 1 in 8 number that many women recognize is the lifetime risk of breast cancer. This means that for all adult women, 1 in 8 (or approximately 12 percent) is at risk of developing breast cancer in her lifetime. While a 12 percent lifetime risk is high for a serious disease like breast cancer, the lifetime risk of coronary heart disease is much higher (about 32 percent), and both coronary heart disease and lung cancer kill more women overall.

FAMILY HISTORY AND GENETICS

It should come as no surprise that family history is linked to the risk of breast cancer. Most people realize that the risks of many diseases are associated with the experiences of their parents and siblings. In the Nurses' Health Study, we have found that having an immediate family member (mother or sister) who has been diagnosed with breast cancer approximately doubles the risk of the disease. The age at which the mother was diagnosed, however, influenced the level of risk as well. The younger the mother was when she was diagnosed with the disease, the higher her daughter's risk of developing breast cancer. A woman whose mother was diagnosed before age forty has a large increase in risk, while a woman whose mother was diagnosed after age seventy has only a moderate increase in risk.

The number of immediate family members with the disease also influences risk. When both a woman's mother and her sister had been diagnosed with breast cancer, we found that the risk of the disease increased to about 2.5 times that of a woman without a family history of the disease. Numerous other studies have also found a similarly strong relationship.

Genetics.Family members usually have very similar lifestyles and therefore may have similar "environmental" risk factors for breast cancer -- such as weight, activity level, and alcohol intake. While this may account for the increased risk of breast cancer experienced by some families, in some cases genetics likely plays a large role.

Much has been written in scientific journals and reported in the media recently on inherited genetic mutations that greatly increase the risk of developing breast cancer. Two of the most well-known examples are BRCA 1 and BRCA 2 (which stands for BReast CAncer gene 1 and 2). Studies have found that women who have the BRCA 1 mutation have approximately 15 times the risk of developing breast cancer compared to women without the BRCA 1 mutation. A BRCA 2 mutation appears to impart a slightly lower risk. Women found to have a breast cancer gene mutation, as well as other high risk women, have certain options that may help lower their risk of getting breast cancer or at least increase their chances of identifying the disease early when it is most treatable. (See Dr. Nancy Rigotti's advice for high risk women at the end of this chapter.)

Although genetic mutations like BRCA 1 are powerful predictors of who will get breast cancer, such mutations do not account for a large portion of breast cancer cases. It is estimated that only 5 to 10 percent percent of breast cancers can be attributed to inherited gene mutations, an approximation in line with our experience in the Nurses' Health Study.

BENIGN BREAST DISEASE

The termbenign breast diseaseis often used to describe a wide range of noncancerous conditions that can affect the breast. In the Nurses' Health Study, we have found that two specific types of benign breast disease -- hyperplasia and radial scars -- increase the risk of developing breast cancer. Other benign breast conditions, such as cysts and fibroadenomas, do not appear to influence risk.

Hyperplasia.Hyperplasia is a noncancerous condition where cells begin to multiply in the breast ducts or lobules. Supporting the results of smaller studies, we have found in the Nurses' Health Study that women with cases of hyperplasia confirmed by a biopsy have a moderate to large increase in the risk of developing breast cancer compared to women without hyperplasia. The greatest increase in risk is found in one specific type of hyperplasia, in which the multiplying cells look abnormal (atypical hyperplasia). Hyperplasia in which the multiplying cells still look normal (hyperplasia without atypia) results in a modest increase in risk.

Radial Scars.Like hyperplasia, histologic radial scars are a type of benign breast disease that is identified only on breast biopsy. They are usually microscopic in size and are not scars as most people think of them. The name refers to the type of tissue that makes up the condition and not to a scar that results from a healed cut or injury. Previous studies of radial scars have had varied findings, with some showing a link to an increased risk of breast cancer and some showing no link at all. When we examined the association of radial scars with breast cancer in the Nurses' Health Study, we found that women with radial scars had about twice the risk of developing breast cancer as women without. The size and number of radial scars also affected risk: the larger and more numerous the radial scars, the greater the risk of developing breast cancer.

BREAST DENSITY

The density of a woman's breasts has been shown in many studies to be strongly related to the risk of breast cancer. Breast density is most often assessed by mammogram and depends on the proportion of fat and tissue that make up a woman's breasts. When the X-rays used in a mammogram take an image of the breast, fat appears dark and tissue appears light. Breasts with a high density have a relatively high proportion of tissue and low proportion of fat. Breasts with a low density have a relatively low proportion of tissue and high proportion of fat. We are currently assessing this issue in the Nurses' Health Study and hope to have results soon. Of note, one very large cohort study found that women with the most dense breasts had approximately 5 times the risk of breast cancer compared to women with the least dense breasts. And as breast density increased, so did the risk of the disease.

Right now, breast density is not being regularly used to assess women's risk of breast cancer, but it may become standard practice in the future.

RADIATION EXPOSURE IN CHILDHOOD OR ADOLESCENCE

Exposure to high doses of radiation to the chest in childhood or adolescence has been consistently linked to an increased risk of breast cancer in later adulthood. Although not specifically evaluated in the Nurses' Health Study, other studies have found that women who were exposed in their youth to atomic bomb blasts (in Hiroshima or Nagasaki, Japan) or radiation therapy for disorders like Hodgkin's disease or tuberculosis had 2 to 4 times the risk of breast cancer compared to women who were not exposed to such radiation. The greatest risk is generally in those women who were very young when exposed, and the risk decreases as the age at the exposure increases. Women over forty who have been exposed to high doses of radiation to the chest have only a small increase in risk.

That breast cancer risk is highest in those women who were exposed to radiation when very young illustrates the importance that early life may play in breast cancer risk. Breast tissue during this time may be more susceptible to elements that can cause cells to become cancerous later in life.

It is still unclear how exposure to low dose radiation, such as that received from an X-ray or mammogram, influences the risk of developing breast cancer. It is possible that there is some small risk associated with such procedures, but the health benefits they provide greatly outweigh any potential increase in breast cancer risk.

PERSONAL HISTORY OF CANCER

Just as family history can affect the risk of breast cancer, so can a personal history of cancer. Women who have been previously diagnosed with cancer in one breast are at increased risk of developing cancer in the other breast. Studies other than the Nurses' Health Study that have assessed the issue have found the risk to be increased two- to fourfold for women previously diagnosed with breast cancer (including ductal carcinoma in situ) compared to those without a previous diagnosis.

Having a personal history of lobular carcinoma in situ -- which is usually not treated after diagnosis, just closely followed -- also increases the risk of cancer. Women with this condition are 7 to 10 times more likely to develop breast cancer than women without the condition.

A history of other cancers can also increase the risk of breast cancer. Studies have shown that women who have been diagnosed with colon or ovarian cancer have a slightly increased risk of developing breast cancer as well. The likely explanation for this is that the three cancers share some of the same genetic or lifestyle factors that are key to cancer development.

ETHNICITY

Not all groups of women in the United States are equally impacted by breast cancer. Latina, Asian American, and American Indian women tend to have the lowest risk of breast cancer in the United States, and white women tend to have the highest risk -- followed closely by Hawaiian American and African American women. For breast cancer occurring before midlife, however, African American women are at highest risk. This is likely due, at least in large part, to differences in the ages at which white women and African American women first give birth. While having children is protective against breast cancer overall, giving birth to the first child after thirty is actually linked to a modest increase in the risk of breast cancer in the short term. Since African American women tend to give birth for the first time earlier than white women, they may experience this increase in breast cancer risk earlier in life, pushing their rates of breast cancer before midlife past those of white women. Across the entire lifetime, though, these differences in reproductive behaviors seem to benefit African Americans, whose lifetime risk of breast cancer is lower than that of whites.

Unfortunately, as overall rates of breast cancer are leveling off or slightly decreasing in white women, they continue to increase in African American women.

Jewish women are another ethnic group at increased risk of breast cancer. Particularly affected are Ashkenazi Jewish women -- Jews who immigrated from central or eastern Europe. One likely reason is that Ashkenazi Jews have an unusually high rate of BRCA 1 and 2 gene mutations compared to the national average.

HEIGHT

There is a substantial amount of evidence that height is associated with the risk of breast cancer both pre- and postmenopause. In the Nurses' Health Study, we found that risk was slightly increased in all women who were 5 feet 3 inches or taller compared to those under 5 feet 3 inches. In another large cohort study, the risk of breast cancer in postmenopausal women was found to increase steadily with height. The tallest women in the study had a little more than twice the risk of the disease compared to the shortest women.

Height is generally thought to reflect, at least in part, the quantity and quality of the diet in childhood. Well fed children generally have more rapid, bigger growth spurts than children fed poor diets. This may contribute to breast cancer risk because such rapid growth can increase the chance that the DNA in some breast cells will become permanently damaged, which can lead to cancer later in life. One other possibility is that levels of a hormone called insulin-like growth factor are higher in taller children than in shorter children. In some studies (including ours), high levels of insulin-like growth factor have been linked to an increased risk of breast cancer.

Factors You Can Control

It is important to know how all of the factors related to breast cancer make up a woman's risk. However, when given a choice, most women would prefer to focus on those things they have control over and can change for the better. These so-called modifiable factors are discussed in this section and range from alcohol intake to use of postmenopausal hormones.

Aside from a small number of factors related to diet, most of the modifiable factors are thought to affect the risk of breast cancer by influencing levels of female hormones (such as estrogen and progesterone) in the body. Estrogen is believed to be a key promoter of breast cancer development, and findings from the Nurses' Health Study have helped support this conclusion. We found that postmenopausal women with the highest blood levels of certain estrogens had almost double the risk of breast cancer compared to women with the lowest levels. Most strikingly, in a subset of women who had never used postmenopausal hormones (hence, their blood hormone levels could be better measured), those with the highest blood estrogen levels had almost four times the risk of developing breast cancer. This increase in risk is similar in size to that between blood cholesterol levels and heart disease.

DIET

When asked if the food we eat plays an important role in breast cancer development, most people would respond with a definite yes. In reality, however, very few aspects of diet have been linked to breast cancer risk.

Vitamin A, Carotenoids, and Fruits and Vegetables.A number of studies have linked a low intake of vitamin A and carotenoids, such as beta-carotene, with an increased risk of breast cancer. Found in high amounts in green and yellow vegetables and certain fruits, vitamin A is important for cell growth, and carotenoids are powerful antioxidants that can help protect cells from the damaging effects of oxygen free radicals in the body. In the Nurses' Health Study, we found that premenopausal women who had a modest to high total intake of vitamin A from food had a 10 to 30 percent lower risk of the disease compared to women with the lowest intake. This reduction in risk was even more pronounced in premenopausal women with a family history of breast cancer: a 60 percent lower risk for those women with the highest intake of vitamin A compared to those with the lowest. This finding, though, was unexpected and needs to be explored further.

Similarly, when we assessed carotenoids, we found that, compared to premenopausal women with the lowest intake of certain carotenoids (beta-carotene, lutein, and zeaxanthin), those with greater intake had a slightly lower risk of breast cancer. No link was found between either carotenoids or vitamin A and breast cancer in postmenopausal women.

Because people eat foods and not specific nutrients, we have also assessed the link between fruits and vegetables and breast cancer risk. What we found was very similar to the relationship seen with vitamin A: premenopausal women who ate modest to high amounts of fruits and vegetables (over two servings per day) had a slightly lower risk of breast cancer than women who ate the least amount of fruits and vegetables (less than one serving). Again, no link was found in postmenopausal women.

When researchers combined our study results with those of seven other large studies, they found similar results: total intake of fruits and vegetables had very little effect on the risk of breast cancer. However, specific types of fruits and vegetables, such as those rich in carotenoids, do appear to reduce risk modestly, primarily in premenopausal women.

Dietary Fat.Eating high amounts of total fat in adulthood appears to have little, if any, effect on the risk of developing breast cancer. In the Nurses' Health Study, we found that, compared to women with a moderate fat intake, women with a high fat diet did not have a greater risk of developing breast cancer. Similarly, women who ate little fat (20 percent or less of total calories) did not have a lower risk of the disease compared to women who ate a moderate amount. These findings corroborate those of an analysis that pooled together the data from a number of large cohort studies (including the Nurses' Health Study). This analysis found that neither a high fat diet nor a low fat diet was related to the risk of breast cancer.

Most Americans today get about 33 percent of their total calories from fat. In the Nurses' Health Study, the women in the cohort eat a diet ranging from approximately 20 percent of total calories from fat, to approximately 50 percent. Because very few of the women eat very low amounts of fat (10 to 15 percent of total calories), we cannot assess how this extremely low intake influences breast cancer risk. It seems, though, that even very low fat intake may not provide protection from the disease. A large case-control study performed in China found no link between fat intake and the risk of breast cancer, even though some women ate a diet containing less than 15 percent of total calories from fat. The analysis that pooled data from a number of studies also found no protection from a very low fat diet (under 15 percent of total calories from fat). Upcoming results from the Women's Health Initiative will provide additional information on the relationship between breast cancer and a low fat diet in postmenopausal women.

Despite there being no apparent link between total fat intake and breast cancer, there may be a link between the type of fat women consume and their risk of the disease. High intake of monounsaturated fat (found in olive and canola oil) has been found in some studies to lower the risk of breast cancer. In the Nurses' Health Study, we have had mixed findings on this issue, with our most recent data showing no relationship between specific types of fat and the risk of breast cancer. One possible reason that our findings contradict those of some others may have to do with the source of the monounsaturated fat. Most of the studies that have shown monounsaturated fats to protect against breast cancer have been performed in Europe and specifically evaluated olive oil intake. In the Nurses' Health Study, animal products (such as red meat and cheese) have been the primary source of monounsaturated fats. That the fat comes from such different food sources could account for the difference in findings.

Although our data from the Nurses' Health Study provide information about how fat intake in midlife relates to breast cancer risk, we cannot address how fat intake in adolescence or early adulthood relates to the risk of the disease. Other studies are currently examining these issues, including our Nurses' Health Study II.

FACTORS THAT AFFECT ESTROGEN LEVELS

Many modifiable factors can increase the level of estrogen in the body, and therefore increase the risk of breast cancer. Examples of such factors include gaining substantial weight as an adult; being overweight and physically inactive after menopause; drinking modest amounts of alcohol; or using postmenopausal hormones.

Overweight and Weight Gain.Studies that have assessed the effect of weight on the risk of developing breast cancer have found that it affects women differently at different ages.

PREMENOPAUSAL WOMEN.In the Nurses' Health Study, when we examined weight's association with breast cancer risk in premenopausal women, we found that as weight increased, the risk of breast cancer decreased. Even though being overweight premenopause lowers the risk of breast cancer, younger women should not take this as carte blanche to gain weight. First, being overweight does not provide premenopausal women with any substantial protection against dying from the disease. This may be due, in part, to the fact that overweight women are often diagnosed with the disease at a more advanced stage than leaner women, making treatment less effective. Second, almost 80 percent of breast cancer cases develop in postmenopausal women, and any weight gained in the premenopausal years will likely track into the postmenopausal years, where it can increase the risk of the disease.

POSTMENOPAUSAL WOMEN.When we assessed postmenopausal women, we found that results differed substantially from those of premenopausal women. For a subgroup of postmenopausal women -- those who had never used postmenopausal hormones -- obesity was linked to a moderate increase in breast cancer risk. For all other postmenopausal women, there was no link between weight and the risk of developing breast cancer.

Weightgain,however, was linked to an increase in risk for all postmenopausal women. Those women who had gained more than 45 pounds since age eighteen had a small increase in risk of developing breast cancer. For women who had gained more than 45 pounds but had never used postmenopausal hormones, breast cancer occurred twice as often as in those who experienced little weight gain as adults.

Being overweight is thought to increase the risk of breast cancer after menopause primarily by increasing levels of the hormone estrogen. Although a woman's ovaries stop producing estrogen after menopause, her fat tissue converts estrogen precursors to estrogen. The more weight put on after menopause, the more estrogen that is produced -- a point the Nurses' Health Study helped confirm. This relation of weight to hormone levels explains why breast cancer risk in overweight postmenopausal women appears to be more pronounced in those who have never taken postmenopausal hormones. The amount of estrogen contained in postmenopausal hormones far outweighs that produced by the fat tissue in overweight and obese women. Therefore, the full effect of weight on breast cancer risk can become masked when a woman uses postmenopausal hormones. It is not that taking postmenopausal hormones eliminates the risk of breast cancer associated with being overweight; it is that using hormones likely hides the effect of weight on risk.

Other large cohort studies assessing the link between weight and breast cancer in pre- and postmenopausal women have had results similar to ours.

Body Shape.Body shape also seems to influence the risk of breast cancer in postmenopausal women. Women who are apple-shaped -- those who tend to carry extra weight around the waist -- seem to be at higher risk of breast cancer than women who are pear-shaped -- those who tend to carry extra weight in their hips and thighs. Similar to the results of other studies, we found that postmenopausal women with the largest waist sizes have about a 35 percent greater risk of breast cancer compared to those with the smallest waist size. As with overall weight, the full effect of waist size on breast cancer risk may be obscured by the use of postmenopausal hormones. When we included only women who had never used postmenopausal hormones, the increase in risk was even greater: women with the largest waist sizes (36 to 55 inches) had about a 90 percent greater risk of breast cancer compared to those with the smallest waist size (15 to 28 inches).

Physical Activity.Numerous studies have examined the relationship between physical activity and breast cancer risk, but results to date have been somewhat inconsistent. This variability in findings likely relates to the different approaches researchers use to assess a woman's level of physical activity, which can be difficult to measure accurately.

In the Nurses' Health Study, we found that the effect of physical activity seems to depend on a woman's stage of life. Postmenopausal women who engaged in at least one hour of physical activity a day were 15 to 20 percent less likely to develop breast cancer than women who were sedentary. When we studied premenopausal women, we saw no clear association between amount of physical activity and risk of the disease. However, most other studies with premenopausal women have seen an association.

Physical activity may help lower the risk of breast cancer in postmenopausal women by helping curb weight gain, which in turn helps keep estrogen levels in check. In young girls, physical activity may have the added benefit of putting off the age at which a girl first has her period, therefore reducing lifetime exposure to estrogen.

Alcohol.Although news that moderate alcohol intake can lower the risk of coronary heart disease has been welcomed with enthusiasm by many, drinking alcohol does not come without accompanying risks as well -- especially for women. The results of many studies have consistently shown that alcohol intake can increase the risk of breast cancer. An analysis evaluating data from five large cohort studies (including the Nurses' Health Study) showed that women who drank two or more drinks a day -- whether beer, wine, or hard liquor -- were 40 percent more likely to develop breast cancer than women who did not drink any alcohol. In our own analysis of the Nurses' Health Study data, we found that even fewer than 2 drinks per day could increase risk. Compared to women who did not drink, those who drank about half a drink to one drink a day had a small increase in the risk of breast cancer, and those who drank more than one drink a day had a moderate increase in risk.

There are a number of ways that alcohol may affect breast cancer risk. In addition to possibly lowering the level of vitamin A in the body, it has been demonstrated in some studies (including the Nurses' Health Study) that alcohol consumption may be linked to increased blood levels of estrogen. Alcohol may also hamper the body's ability to use folate, a vitamin that may help protect against tumor growth. Interestingly, we found in the Nurses' Health Study that the risk of breast cancer associated with one drink or more a day was lower in those women with a high intake of folate compared to those with a low intake. This finding, however, still needs to be confirmed.

For more information on the risks and benefits of alcohol, see Chapter 20.

Postmenopausal Hormones.A large number of U.S. women take postmenopausal hormone therapy to alleviate the short-term symptoms and longer-term health effects of menopause. While such hormone therapy appears to be beneficial in many ways -- lowering the risk of osteoporosis and possibly coronary heart disease and colon cancer -- it also increases the risk of breast cancer in certain users.

After menopause, when a woman's ovaries stop producing estrogen, postmenopausal hormone therapy is a key source of estrogen in those women choosing to use it. After examining many different aspects of the link between breast cancer risk and postmenopausal hormones, we have found in the Nurses' Health Study that risk is substantially increased only in those women who are currently using postmenopausal hormones and have been using them long term (over 5 years). Women who are long-term current users have an approximately 50 percent greater risk of developing breast cancer than women who did not use postmenopausal hormones, and, in general, the longer women use hormones, the greater the risk.

Because postmenopausal hormones that contain only estrogen have been shown to increase the risk of cancer of the uterus, more and more women are choosing to take hormones that contain both estrogen and progestin. It was initially thought this combined therapy might reduce the excess risk of breast cancer linked to postmenopausal hormones. However, we in the Nurses' Health Study -- along with others -- have found that the risk of breast cancer was actually higher in those women using estrogen and progestin than in those using estrogen alone.

The findings from the Nurses' Health Study assessing postmenopausal hormones and breast cancer risk are very similar to those of numerous other reports, including a very large analysis based on over fifty international studies. Though the details are complicated, the bottom line message about postmenopausal hormone therapy and breast cancer is fairly simple.

• Women who use postmenopausal hormones for less than five years do not seem to be at significantly increased risk of breast cancer.

• Those who use hormones for more than five years are at an increased riskwhilethey are on the hormones, and their risk increases the longer they use them. Once they stop, however, their risk returns to that of someone who has never used hormones. Formulations that contain estrogen and progestin increase risk more than formulations that contain estrogen alone.

Issues other than breast cancer, however, also need to be considered when deciding whether or not to use hormones after menopause. Several studies suggest that the risk of osteoporosis, coronary artery disease, and colon cancer can all be reduced with postmenopausal hormones. And some very preliminary data suggest that hormone therapy may possibly even lower the risk of Alzheimer's. Ultimately, a woman and her health care provider need to consider her risk of the various disorders before deciding what approach is best for her. Chapter 22 provides a detailed discussion of the potential risk and benefits of postmenopausal hormone use.

Birth Control Pills.The use of birth control pills moderately increases the risk of breast cancer in those women who are currently using them. In the Nurses' Health Study, we found that women who were currently taking the pill had a 50 percent increase in risk compared to women who had never used the pill. Women who had used the pill in the past (but were not doing so currently) did not show any increase in breast cancer risk. Moreover, how long a woman had been on the pill did not seem to change the results.

Although a number of other studies have found no association between use of the pill and an increased risk of breast cancer, a very large analysis that combined over 50 international studies corroborated our results, finding that current users of birth control pills had a small increase in breast cancer risk and that this excess risk slowly returned to zero ten years after stopping taking the pill.

As with postmenopausal hormone therapy, there is a lot to consider when deciding whether or not to use the pill. Although the pill seems to slightly increase breast cancer risk in current users, it also has many benefits. In addition to preventing unwanted pregnancy, it lowers the risk of ovarian and endometrial cancers. Also, when women use the pill, it is during a time in life when their absolute risk of breast cancer is low, so even though a 50 percent increase in risk is substantial, it will actually result in only a few extra women developing breast cancer who otherwise would not have. Assessing such individual risks and benefits and the impact they have on a woman's lifestyle is key to determining whether or not she uses birth control pills. Chapter 23 discusses these issues in greater detail.

Reproductive Factors

Reproductive factors -- such as age at first menstrual period (menarche), age at menopause, and age at first giving birth -- have been linked to the risk of breast cancer in a wide range of studies. Much of the effect these factors have on breast cancer risk is thought to be related to their influence on levels of female hormones, such as estrogen, as well as on the maturation of the breast.

AGE AT FIRST MENSTRUAL PERIOD (MENARCHE)

Girls in the United States are having their first menstrual periods (menarche) at an increasingly early age -- with an accompanying elevation in breast cancer risk later in life. For many years, young age at menarche has been well established as a factor that increases the risk of breast cancer. In the Nurses' Health Study, we have found that women who were over thirteen when their periods began had a 35 percent lower chance of developing breast cancer than women whose periods started at age twelve or younger. A late age at menarche is thought to lower breast cancer risk by creating a shorter interval between the time a woman's period starts and the time she gives birth to her first child, reducing her exposure to the female hormones (such as estrogen and progestin) released during the menstrual cycle.

AGE AT FIRST GIVING BIRTH AND NUMBER OF CHILDREN

For a number of reasons, more and more women are putting off starting a family. Unfortunately, beginning a family at a later age increases the risk of breast cancer. Numerous studies have found that the older women are when they give birth to their first child, the higher their risk of breast cancer. Our data from the Nurses' Health Study also show that women who have given birth have a lower risk overall than women who have not, and women who have more children have a lower risk than women who have fewer children. In addition, we have found that when women have more than one child, the spacing of births can influence risk. Women whose births were spaced closer in time have a slightly lower risk of breast cancer than women whose births were spaced further apart.

Our study and others have found that the risk of breast cancer associated with giving birth to a first child varies over a woman's life. Although the first birth lowers the risk of breast cancer in the long term, there is actually a short-term increase in risk for ten or more years immediately following the event. After this time, the risk begins to drop.

In the short term, a first full-term pregnancy may increase the risk of breast cancer because of the high levels of hormones associated with fetal development that can promote a preexisting breast cancer. But pregnancy also confers long-term protection from the disease, possibly by causing changes to breast cells that make them less susceptible to becoming cancerous as well as by permanently lowering levels of estrogen in the body.

BREAST-FEEDING

Breast-feeding has many benefits for mother and child, and it has been postulated for many years that it may reduce women's risk of breast cancer. Although there is still debate on the issue, a majority of studies have found that breast-feeding can lower women's risk of the disease, particularly for breast cancer that develops before menopause. In the Nurses' Health Study, we did not find a difference in breast cancer risk for premenopausal or postmenopausal women who had breast-fed. However, we were limited in our ability to assess the long-term effects of breast-feeding.

One possible reason that some studies show a benefit of breast-feeding and other do not may relate to the fact that women use breast-feeding in many different ways. Some use it as the only source of food for their infants, and some use it to supplement other modes of feeding. And over time, women may use a combination of these two approaches. The result is that two women who report that they have both breast-fed their child for six months may have, in fact, spent very different amounts of time actively breast-feeding. Better delineating the exact approaches women have used during their periods of breast-feeding -- as we are currently doing in the Nurses' Health Study -- may help clarify the relationship between breast-feeding and the risk of breast cancer. Aside from possibly protecting against breast cancer, breast-feeding also means fewer childhood infections, less work time lost to care for a sick child, and a quicker return to pre-pregnancy weight.

AGE AT MENOPAUSE

Studies have long shown that women who go through early menopause, whether naturally or through surgical removal of the ovaries (oophorectomy), have a reduced risk of breast cancer. As with late age at menarche, early age at menopause is thought to decrease risk by shortening the lifelong exposure to the hormones released during the menstrual cycle. Our data from the Nurses' Health Study show that for every one-year increase in age at natural menopause, a woman's risk for breast cancer increases by 3 percent. This translates to an approximately 35 percent increase in risk for a woman going through menopause at age fifty-five or older compared to one going through it at forty-five.

Unproven Factors and Factors Found to Have No Effect on Breast Cancer

Many factors discussed in the media and certain health circles are said to affect the risk of breast cancer but have not actually been proven to do so. Some of these may in the future be linked to breast cancer. Some, though, have been conclusively proven to have no influence on the risk of the disease. Here are some examples of these factors, many of which we have assessed in the Nurses' Health Study.

ISOFLAVONOIDS AND SOY ESTROGEN

Isoflavonoids are specific types of estrogen that are found abundantly in soy products. In laboratory studies, isoflavonoids seem to protect breast cells from becoming cancerous by blocking the cancer initiating action of other estrogens. To date in the Nurses' Health Study, we have not studied this issue, primarily because exposure to isoflavonoids has historically been low in the United States. Since women have greatly increased their intake in recent years, we will be able to examine this thoroughly in the future. Results from the few studies that have looked at the issue have been mixed, with some showing a benefit and others showing no relationship. Before any conclusions can be drawn about the link (if any) between isoflavonoids, soy, and breast cancer, there must be more human studies on the topic.

SELENIUM

Selenium is a substance found in food and nutrition supplements that has been shown in laboratory experiments to lower the risk of certain kinds of cancer in animals. When we assessed the link between selenium and breast cancer risk in the Nurses' Health Study, we found no association with the risk of disease in either premenopausal or postmenopausal women.

ORGANOCHLORINE CHEMICALS

Organochlorine chemicals, found in certain pesticides and industrial chemicals, have often been discussed as potential risk factors for breast cancer because they have qualities similar to estrogen. Early studies suggested a link with breast cancer, but now it seems that organochlorines may have no effect on risk. When we examined a sub-sample of the Nurses' Health Study cohort, we found that blood levels of two specific organochlorine chemicals -- DDE and PCBs -- were not linked to breast cancer risk.

HETEROCYCLIC AMINES

Heterocyclic amines are substances produced by cooking meat at high temperatures; they have been demonstrated to increase risk of cancer in laboratory experiments with animals. Heterocyclic amines appear in the greatest amount in meat that has been charred or flame-broiled. Though some studies have found a link between heterocyclic amines and breast cancer in humans -- including the Iowa Women's Health Study -- in an initial assessment in the Nurses' Health Study we found no link between the two.

GINSENG

As with isoflavonoids, it is thought that the herb ginseng may help protect against breast cancer by blocking the cancer-initiating action of estrogen. Some laboratory studies have shown positive results, but few studies have been conducted in humans. We have yet to assess this issue in the Nurses' Health Study but hope to do so in the future.

SMOKING

Because smoking lowers the age at which a woman enters menopause as well as affects the metabolism of estrogen, it has been thought that it may actually reduce the risk of breast cancer. However, we found in the Nurses' Health Study that smoking in adulthood had no influence on the risk of breast cancer. Smoking early in youth may modestly increase the risk of breast cancer, as some studies have suggested. However, we have not yet evaluated this in the Nurses' Health Study.

HAIR DYES

Permanent hair dye contains substances called aromatic amines that have been shown in laboratory studies to increase the risk of breast cancer in some animals. In the Nurses' Health Study, we have repeatedly found no association between the use of permanent hair dyes and the risk of breast cancer.

ABORTION

The association between abortion and breast cancer has been controversial in the past, but it now seems clear that abortion has no influence on the risk of the disease. As with other studies on the topic, our Nurses' Health Study II found no link between the two.

OTHERS

Other factors that have not been proven to increase the risk of breast cancer include electromagnetic fields, antiperspirants/deodorants, and the use of bras. For each of these, there is no clear way biologically that they could increase the risk of the disease.

WHAT IT ALL MEANS

As is apparent from the lengthy list of factors discussed above, breast cancer is a very complex disease. But the messages are relatively simple when it comes to the steps a woman can take to try to reduce the risk of breast cancer, and the good news is that these same steps also help prevent coronary heart disease, diabetes, and other chronic disorders.

•Maintain a healthy weight and avoid substantial weight gain (more than about fifteen to twenty pounds) during adulthood.Maintaining a healthy weight at every age protects women against breast cancer, even if they are not physically active. A healthy weight in childhood increases the age at which the first menstrual period occurs and slightly decreases the age at which menopause occurs: together, these two factors lead to a lower lifetime exposure to estrogen. Maintaining a healthy weight also appears to have some protective effects over and above its effects on the age at menarche and menopause.

•Lead a physically active lifestyle.Physical activity at every age protects women against breast cancer. When young girls are physically active, the age at which they experience their first menstrual period is delayed. This lowers their lifetime exposure to estrogen, and this in turn lowers their risk of breast cancer. Being physically active throughout life may directly reduce the risk of breast cancer after menopause, even in women who are not overweight. When physical activity leads to weight loss or prevents weight gain, that also protects you.

•Eat a diet rich in fruits and vegetables.Fruits and vegetables contain a number of substances that may help the body fight cancer, including vitamin A and carotenoids.

•Drink less than one alcoholic drink a day on average.A drink a day or more has been linked to an increase in breast cancer risk. There may be no safe level of alcohol consumption in relation to breast cancer risk, but moderate alcohol intake has been linked to a decreased risk of coronary heart disease.

Avoiding long-term use of postmenopausal hormones can also lower the risk of breast cancer, but because hormone use carries both risks and benefits, it is important for women to discuss this issue individually with their health care providers.

WHAT I TELL MY PATIENTS ABOUT LOWERING THEIR RISK OF BREAST CANCER

DR. NANCY RIGOTTI

Ask one of my patients which disease she fears most, and chances are that she will answer breast cancer. Most women do. Breast cancer cuts right to the heart of our identity as women, and the constant stream of stories, research findings, and advice about breast cancer keeps the disease at the forefront of women's concerns. It can, in fact, be difficult for women not to fear breast cancer.

There are two and possibly three ways in which you can help to protect yourself from breast cancer:

• Adopt a healthy lifestyle.

• Get breast cancer screening tests, when indicated.

• If you are at unusually high risk of getting breast cancer, consider whether to use certain preventive treatments (such as taking the drug tamoxifin).

Adopt a Healthy Lifestyle

We know less about how to prevent breast cancer than we do about preventing some other diseases like heart disease, but fortunately we do have effective tests that can find the disease early, when treatment has the best chance of cure. And although breast cancer is influenced by many factors that you cannot control, it is also influenced by factors you can control. Diet and physical activity, the main components of a healthy lifestyle, may also lower the risk of getting breast cancer.

START EXERCISING

Most of my patients know that regular physical activity is good for their heart and bones. But most of them don't know that it also reduces their risk of developing breast cancer after they reach menopause. It's yet another reason to get moving. The Surgeon General recommends thirty minutes of moderate physical activity on most days of the week. You don't have to go to a gym to become physically active! You can do things like take a brisk walk, garden, bicycle with your children, or swim with a friend. If you feel you don't have time to fit another thirty minutes into your hectic day, then aim for three ten-minute periods of activity a day.

KEEP YOUR WEIGHT UNDER CONTROL

Your weight affects your risk of developing breast cancer after you reach menopause. If you have gained more than forty-five pounds since you were eighteen, you are at increased risk of developing breast cancer. To reduce your risk, try not to gain more than fifteen to twenty pounds during adulthood. You can do this by eating a healthy diet and remaining physically active.

EAT PLENTY OF FRUITS AND VEGETABLES

Although you may be worrying about whether pesticides on the food you eat can cause breast cancer, you should really be thinking about what you're not eating. Americans are notorious for not eating enough fruits and vegetables. That's a shame, because fruits and vegetables contain substances that prevent normal cells from turning cancerous. The American Cancer Society recommends a diet that includes five servings of fruits and vegetables a day. So eat as many fruits and vegetables as you can manage every day. Put fruit on your cereal in the morning and lettuce on your sandwiches. Pack carrots, an apple, or a banana for a snack. Substitute a glass of orange juice for a cup of coffee at work. Try to eat vegetables or a salad every night with dinner. These small steps add up.

COOK WITH MONOUNSATURATED FATS

For years people worried that a high fat diet might increase the risk of breast cancer. Further studies have failed to confirm this. In fact, there is growing evidence that cooking with oils that are high in monounsaturated fats (found in olive oil and canola oil) may actually lower the risk of breast cancer. These oils have many other health benefits as well.

HAVE LESS THAN ONE ALCOHOLIC DRINK PER DAY

Who hasn't heard by now that there may be some health benefits from drinking alcohol? There are some benefits to moderate drinking, which has been linked to a reduced risk of coronary heart disease. The problem is that many of these studies have been done in men, who metabolize alcohol differently than women. So the best advice is that, if you want to drink alcohol, drink less than one drink per day (or less than seven drinks per week). Any more and you may increase your risk of breast cancer.

THINK CAREFULLY ABOUT POSTMENOPAUSAL HORMONES

The decision about whether to take postmenopausal hormones when you reach menopause is a complicated one. But when it comes to the impact on breast cancer risk, the issue is not so much whether to take hormones, but if you do, how long to continue it. Women who take hormones for less than five years following menopause do not appear to increase their risk of breast cancer. But taking hormones longer than five years may increase your risk. The picture is complicated by other issues, such as your individual risk of breast cancer based on family history and other factors. This is definitely an issue you should discuss with your health care provider.

Get Regular Breast Cancer Screening Tests

In addition to taking steps to try to prevent the disease, all women, once they reach a certain age, should be screened regularly for breast cancer. Although screening tests, like mammograms and clinical breast exams, cannot help prevent the disease from developing, they can identify it early when it is most treatable. The earlier a cancer is found, the more likely it is that a woman can be treated successfully.

MAMMOGRAMS

Mammograms are special X-rays of the breast. Virtually everyone agrees that you should get a yearly mammogram once you turn fifty and continue until you turn sixty-nine. After that, you should get a mammogram every one to two years. This is hands-down the best way to catch breast cancer early.

Almost everyone also agrees that if you have an unusually high risk of breast cancer, such as from a strong history of breast cancer in your mother or sisters, regular mammograms are probably valuable starting at a younger age. If my patient's mother or sister has had breast cancer, I'll advise annual mammograms for her starting at age forty, or ten years before the age at which the close relative was diagnosed with breast cancer, whichever age comes first.

The evidence to support having regular mammograms before age fifty is not nearly as good as the evidence that mammograms starting at age fifty save lives. There are a couple of reason why this is so. First, the risk of breast cancer in this age group is much smaller, and second, the test appears to be less accurate because younger women tend to have denser breasts (more tissue and less fat) than older women. This makes reading mammograms and detecting abnormalities more difficult: a negative or normal result may be reported in a woman who actually has breast cancer. Both of these factors make it more difficult to prove a benefit to screening mammograms for women under age fifty. Furthermore, more mammograms in women aged forty to forty-nine lead to positive or abnormal results that result in a biopsy showing there was no cancer, after all. Having an abnormal mammogram result is scary for a woman and her family, even if the biopsy shows that no cancer was present. The wisest course of action is to discuss the decision about mammography with your health care provider, who can help you to factor in your individual risk and arrive at the best decision for you.

For women ages seventy and over, there is currently not enough evidence for me or most health care providers to make any concrete recommendation about mammograms. However, because breast density decreases with age, there appears to be no reason that mammography should be less effective in women as they advance into this age group, and the risk of breast cancer continues to rise over age seventy. A woman aged seventy or older should discuss the decision about whether to get regular mammograms with her health care provider. I generally recommend it for my patients who are otherwise in good health and who could withstand breast cancer treatment if it were necessary.

CLINICAL BREAST EXAMS

A clinical breast exam is a physical examination of the breasts by a health care provider (most often a physician, but a nurse practitioner or nurse may also perform it). Clinical breast exams are most effective in women aged fifty to sixty-nine, but I advise my patients to have them every year, as part of their regular physical. As with other kinds of screening tests, experts disagree on the age at which you should start, or the frequency with which the exam should be done.

BREAST SELF-EXAM

You have probably seen laminated breast self-exam cards made to hang in the shower that outline how to properly do a breast self-exam. The main advantage of this technique is that it enables you to become more familiar with your breast tissue so that you may notice any changes or even detect a lump. Breast self-exam is vigorously promoted by many health organizations as an effective way for women to find breast cancers that develop during the period between regular mammograms and clinical breast exams. Unfortunately, very little data actually support this claim. Neither the U.S. Preventive Services Task Force nor the National Cancer Institute recommend that women of any age regularly perform breast self-exam. I do not discourage my patients from doing breast self-exam, but I stress that it should never take the place of getting regular mammograms and clinical breast exams.

When My Patient Has a High Risk of Getting Breast Cancer

ESTIMATING THE RISK

Fortunately, most women are at average or low risk for the disease and do not need to take extra precautions -- aside from leading a healthy lifestyle and getting regular screening tests. However, some women are at high risk and do need to consider taking special steps to protect themselves. We now have pretty good information on factors that increase your risk of getting breast cancer, even though the details are still being worked out.

You are clearly at a higher risk of getting breast cancer if you haveany oneof the following risk factors:

• A mutation in the BRCA 1 or BRCA 2 genes

• A mother who developed breast cancer before she turned sixty, or a sister or daughter who developed breast cancer

• Breast abnormalities (such as atypical hyperplasia) in the past

• Repeated exposure as a child or adolescent to high dose radiation

You also are at a higher risk of getting breast cancer if you have several of the following risk factors:

• Your first menstrual period was at a relatively young age (twelve or younger).

• You reached menopause at a relatively late age (fifty-five or older).

• You first gave birth later in life (age thirty or older).

• You have high estrogen levels in your blood.

For patients of mine who are at a relatively higher risk of getting breast cancer, I generally recommend that they consider getting regular mammograms and clinical breast exams starting at age forty, even though the value of this has not yet been proven conclusively.

TREATMENT OPTIONS FOR UNUSUALLY HIGH RISK WOMEN

Some of my patients are at unusually high risk of breast cancer. Those who I tend to put in this category have a very strong family history or have been diagnosed with lobular carcinoma in situ or ductal carcinoma in situ. If you are at unusually high risk of breast cancer, you basically have two options to try to lower your risk: chemoprevention or prophylactic mastectomy.

Chemoprevention.This is the use of drugs to prevent breast cancer, and it is a very active area of research. Tamoxifen and raloxifene are two of the most promising chemoprevention drugs that are in a class called selective estrogen receptor modulators (SERMs). SERMs act by blocking estrogen from attaching to breast cells.

In a National Cancer Institute-funded clinical trial of tamoxifen that included over 13,000 women at high risk of breast cancer, women who took tamoxifen over a five-year period had about half the risk of developing invasive cancer as women who did not take the drug. Though their risk of breast cancer was lower, the women on tamoxifen also experienced increased rates of some serious side effects, including endometrial cancer and blood clots in the lung (pulmonary embolism) and large veins of the legs (deep venous thrombosis). Raloxifene works in a similar fashion to tamoxifen, and early reports on its effectiveness show that it reduces the risk of breast cancer by more than half. It also seems to have an added benefit over tamoxifen in that women taking raloxifene seem to suffer fewer serious side effects.

High risk women should certainly talk with their health care providers about the risks and benefits associated with taking a chemoprevention drug. Drugs like tamoxifen and raloxifene seem to have a substantial benefit for some high risk women over a relatively short time period. However, the long-term benefits (and risks) related to these drugs have yet to be determined.

Prophylactic mastectomy.This is a drastic procedure that involves the surgical removal of both breasts in order to reduce the chances that breast cancer will develop. In my judgment, it is rarely indicated. The one exception may be for women who have the BRCA 1 or BRCA 2 genes that increase the risk of breast cancer (and ovarian cancer). Their risk is sufficiently high that -- as radical as it seems -- removing the breasts may be the right thing to do. One study has found that a thirty-five-year-old woman with BRCA 1 or BRCA 2 mutations can gain three to five years of life by having the procedure. Still, you never can know whether this radical approach was justified in the individual, because you will never know if she would have developed breast cancer or not.

And although the procedure does afford substantial protection, it does not guarantee complete protection. Breast cancer can still develop in the small amount of tissue that remains after surgery. Because it is such a disfiguring procedure, if you are at high risk I think you should consider prophylactic mastectomy only after a careful discussion with your health care provider about all of the associated risks and benefits.

Yes. The thought of breast cancer is frightening. But all women should feel they have the power to take action against the disease. The lifestyle factors I've outlined may help lower the risk of breast cancer and regular screening can find it when it's most curable. Perhaps most heartening is the great progress we've made over the past twenty-five years in our understanding of the disease, and as research continues to advance at an astounding pace, each new year will bring with it further understanding of the causes of breast cancer as well as the ways my patients and I together can combat it.

Copyright © 2001 by the President and Fellows of Harvard College

---

Excerpted from Healthy Women, Healthy Lives: A Guide to Preventing Disease, from the Landmark Nurses' Health Study by Susan E. Hankinson, Graham Colditz, Joann E. Manson, Frank Speizer
All rights reserved by the original copyright owners. Excerpts are provided for display purposes only and may not be reproduced, reprinted or distributed without the written permission of the publisher.