This Cup is Way Less than Half Full

Abstracts & Commentary

By Barbara A. Phillips, MD, MSPH, Professor of Medicine, University of Kentucky; Director, Sleep Disorders Center, Samaritan Hospital, Lexington. Dr. Phillips serves on the speaker's bureau of Cephalon, Boehringer Ingelheim, Merck, ResMed, and GlaxoSmithKline, and is a consultant for Boehringer Ingelheim, Wyeth-Ayerst, and ResMed.

Synopsis: Three simultaneously published reports from the Women's Health Initiative demonstrate that a low-fat diet did not reduce the risk of breast cancer, colorectal cancer, or cardiovascular disease over 8 years of follow-up.

Sources: Howard BV, et al. Low-fat dietary pattern and risk of cardiovascular disease: the Women's Health Initiative Randomized Controlled Dietary Modification Trial. JAMA. 2006;295:655-666; Prentice RL, et al. Low-fat dietary pattern and risk of invasive breast cancer: the Women's Health Initiative Randomized Controlled Dietary Modification Trial. JAMA. 2006;295:629-642; Beresford SA, et al. Low-fat dietary pattern and risk of colorectal cancer: the Women's Health Initiative Randomized Controlled Dietary Modification Trial. JAMA. 2006;295:643-654.

The Women's Health Initiative (WHI) is an ongoing, longitudinal study of 48,835 women between the ages of 50 and 79 years enrolled at 40 US centers between 1993 and 1998. This group is extremely well-characterized and closely followed, with meticulous data about diet and lifestyle collected annually. Comprehensive reports of methods and baseline characteristics have been published.1,2 For this trio of studies, the dietary intervention was designed to reduce fat intake to 20% of total calories and also to increase consumption of vegetables and fruits to at least 5 servings a day, and grains to at least 6 servings a day.

The intervention group received intensive behavioral modification which included 18 group sessions (led by a nutritionist) in the first year, then quarterly thereafter. There was personalized attention to each woman's total fat gram goal based on height. Dietary data were collected using the Frequency Food Questionnaire (FFQ), an instrument designed and validated for use in the WHI. The FFQ was administered at baseline, at one year following randomization, and then yearly to about a third of the participants in a rotating sample. Comprehensive 4-day food records were also collected from all women at baseline. All participants were contacted every 6 months, and a variety of measurements were made at these contacts, including height, weight, waist circumference, and blood pressure. A small subset of women provided additional intensive dietary monitoring, with additional 4-day food record at 1 year and 24 hour dietary recalls at 3 and 6 years. Fasting serum was collected at baseline and at 1 year for all women, and at 6 years in a subset (n = 2186) of women; this serum was analyzed for a variety of measures, including lipids, carotene and tocopherol levels. Serum hormone levels were also measured at baseline and at one year from 150 women in both the intervention and the control group.

To screen for incident cardiovascular, breast cancer, or colon cancer risk, medical update questionnaires were completed every 6 months. These questionnaires were extensive and included information about new diagnoses, procedures, and overnight hospitalizations.

Additionally, to screen for cardiovascular risk, electrocardiograms (ECGs) were digitally acquired every 3 years. Major coronary heart disease (CHD) was defined as acute myocardial infarction (MI) requiring hospitalization or silent MI determined from serial ECGs or death; composite. CHD was defined as MI, CHD death, and coronary revascularization procedures. To assess for breast cancer risk, women were expected to undergo mammography at baseline and every 2 years; tissue diagnosis was required to be included as a case in this study. Screening for colon cancer occurred based on decisions made by the participants and their physicians; there was no specific protocol for this.

End points in this study were incident breast or colon cancer, CHD or individual coronary heart disease events. The effects of the low-fat diet were based on time-to-event curves based on intention to treat. Hazard ratios for the end points for the control compared with dietary intervention group were stratified for age at entry, prevalent disease, and whether the participant was taking HRT. Women who had a given outcome variable at baseline were excluded from calculations for that particular outcome. In secondary analyses, the investigators compared event rates in the intervention and control group controlling for relevant known risk factors for the event; for example, for CHD, data were analyzed controlling for ethnicity, age, BMI and other health characteristics known to influence the risk of CHD.

By year 6, mean fat intake decreased by about 8.1 % (of total caloric intake) in the dietary intervention group, and there were small increases in vegetable and fruit intake. Of note, although the goal was to reduce fat intake to fewer than 20% of calories, the actual reduction (despite intensive intervention) was to 28.8% compared with the control group's 37.8%. The dietary intervention group did not have reduction in incident CHD, stroke, invasive colon cancer, or invasive breast cancer, although encouraging trends were seen. Secondary analyses suggested a lower HR of breast cancer risk for women who were more adherent to the low-fat diet, or who had higher fat intake at baseline. There was a trend toward lower CHD risk in those with lower intakes of saturated fat and trans fat or higher intakes of fruits and vegetables. Low density lipoprotein cholesterol, diastolic blood pressure, and factor VII c levels were statistically significantly reduced (though absolute reductions were quite small: 3.55 mg/dL, 0.31 mm Hg and 4.29%, respectively). High-density lipoproteins, triglycerides, glucose, and insulin levels did not change between the 2 groups.


Remember the Woody Allen movie, Sleeper in which a health food store owner wakes up hundreds of years in the future? The baffled but hungry time traveler asks for a health food salad, and evokes mirth in his futuristic friends, one of whom comments incredulously, "Imagine! He believes that hot fudge sundaes are bad for you!" to which his companion responds, "Yes. Exactly the opposite of what we now know to be true!" Once again, the WHI study (which shot down routine use of hormone replacement for postmenopausal women)3 has disproved, or at least seriously dampened enthusiasm for, a cherished clinical belief. The results of these studies challenge, in triplicate, the rationale for a low-fat diet.

This issue of JAMA includes 2 editorials4,5 that address these papers. Buzdar notes the trend toward reduction in breast cancer, points out that there was a reduction in estrogen receptor positive (progesterone receptor negative) breast cancer and reviews the rationale for an association between high fat intake and breast cancer. He suggests that a low-fat diet may be of benefit for those who have survived early stage breast cancer to prevent recurrence. Anderson and Appel suggest that the WHI trial did not take into account the different kinds of fat (eg, trans fat, saturated, polyunsaturated and monounsaturated) and associated risks, and note that reduction in fat intake in the intervention group was modest. We all wonder if longer-term follow-up will ultimately demonstrate a difference; the trends noted in 8 years of follow-up certainly suggest that there might be, particularly for breast cancer.

Where does that leave us now? We can't honestly tell our patients that a low-fat diet will affect their risk of these common and deadly killers: breast cancer, colon cancer, and cardiovascular disease. There are so many things for which lifestyle change is conclusively known to cause immediate, unequivocal benefit (seat belts and smoking abstinence come to mind, not to mention exercise) that it's difficult to drum up much enthusiasm for harping on a low-fat diet based on present evidence.


1. Women's Health Initiative Study Group. Design of the Women's Health Initiative clinical trial and observational study. Control Clin Trials. 1998;19:61-109.

2. Ritenbaugh C, et al. The Women's Health Initiative Dietary Modification trial: overview and baseline characteristics of participants. Ann Epidemiol. 2003; 13(9 Suppl):S87-S97.

3. Anderson G, et al. Effects of conjugated equine estrogen in postmenopausal women with hysterectomy: the Women's Health Initiative randomized controlled trial. JAMA. 2004;291:1701-1712.

4. Buzdar AU. Dietary modification and risk of breast cancer. JAMA. 2006;295:691-692.

5. Anderson CA, Appel LJ. Dietary modification and CVD prevention: a matter of fat. JAMA. 2006;295:693-695.