WASHINGTONDensity variations in breast images may be due to a number of different variables, such as age and body weight, Norman F. Boyd, MD, DSc, said at the 10th Annual Conference of the American Institute for Cancer Research (AICR).
In general, said Dr. Boyd, of the Ontario Cancer Institute, Toronto, stroma and epithelium are denser and record as lighter areas on the mammogram, while fat allows x-rays to pass easily and appears darker on the mammogram.
An Important Hypothesis
This raises the possibility that dietary fat may influence image results, an important hypothesis since, according to Dr. Boyd, four quantitative studies and five case-control studies have shown a trend associating greater breast density with increased risk of breast cancer.
Furthermore, cross-cultural differ-ences in fat intake also account for much of the worldwide variation in breast cancer rates, he said. Fat intake accounts for 63% of the variation between countries, 80% of the variation within countries, and 70% of the variation among migrants from one country to another.
These data suggested that a randomized clinical trial to lower total dietary fat intake (to 15% of total calories) among high-risk women might also lead to a reduction in breast cancer, Dr. Boyd said. He reported on such a trial involving 4,615 subjects who will be followed for 10 years.
Women were given a baseline mammogram to identify those at high risk of breast cancer (radiologic densities in more than 50% of the breast area). These women were then randomized to intervention and control groups. Average age of the subjects was 47 years.
Women in the intervention group were assigned a personal dietitian, who provided individual recommendations for shopping and cooking. The dietitian then closely followed the subjects, monthly for the first year and quarterly thereafter. Such close follow-up permitted a wider range of dietary intakes and examination of the biologic effects of a low-fat diet. Mammography images at baseline and at 2 years for 817 women were read in random order by observers blinded as to subject status.
The results showed that a low-fat, high-carbohydrate diet did indeed decrease the area of breast density. However, Dr. Boyd said, a number of other factors might also contribute to this variation in breast density, including family history, weight, calorie intake, and cholesterol levels. In addition, density declines with age, an inverse correlation to that of age and breast cancer risk.
Effects of Menopause
Menopause adds yet another potentially confounding factor. Premenopausal women showed a significant reduction in density. Women who went from pre- to postmenopausal status during the trial showed some reduction in density, and postmenopausal women showed little change.
In short, Dr. Boyd said, this study at 2 years of follow-up showed that the connection of diet and breast density by itself provides no new clues to predicting breast cancer risk. However, he suggested that breast density may offer a clue to related patterns of cancer risk.
Longer follow-up is needed to determine whether diet-related reductions in density are associated with reduced breast cancer risk, he concluded.