The Association between measures of Diet Quality and Breast Density and Body Composition in Premenopausal Women
Open Access
- Author:
- Lindgren, Jessica Ann
- Graduate Program:
- Nutrition
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- June 06, 2013
- Committee Members:
- Terryl Johnson Hartman, Dissertation Advisor/Co-Advisor
Dr Joanne Dorgan, Committee Member
Jennifer Savage Williams, Committee Member
Donna Coffman, Special Member - Keywords:
- Breast density
Body composition
Breast cancer
Diet
Nutrition
Premenopausal women - Abstract:
- Breast cancer is the most commonly diagnosed cancer in women in the United States, with 1 in 8 women being diagnosed during their lifetime. There are several well-known risk factors for breast cancer including age, genetic factors, family history, reproductive characteristics, physical attributes (e.g. dense breast tissue) and diet and diet-related factors (e.g. alcohol consumption, weight, and body fatness). While many of these risk factors remain consistent across the pre- and postmenopausal years, it has been established that menopausal status modifies the effect of obesity on breast cancer risk. Body fatness decreases risk in premenopausal women but increases risk in postmenopausal women. However, when assessing body fat distribution, central obesity has been positively associated with breast cancer risk in both pre- and postmenopausal women. Breast density, or the ratio of dense fibroglandular area to total breast area, reflects breast tissue composition and is a strong breast cancer risk factor. Breast density is higher in premenopausal women, postmenopausal women who use hormone replacement therapy (HRT), and those, regardless of menopausal status, who have a lower BMI. In contrast, women who are older, parous, have their first birth at a young age, or are smokers often have lower breast density. These factors can directly (e.g. through altering breast morphology) or indirectly (e.g. through influencing hormones) affect breast density. Breast density may also be a marker for lifestyle exposures (e.g. diet and body size) and their effects on hormone levels. Because breast density and body fatness have both been shown to be modifiable and susceptible to dietary influences, they are often targeted as potential breast cancer prevention strategies. Many current studies examine single foods or nutrients, which often provide inconsistent results. Our studies focus on measures of overall diet (e.g. dietary energy density and dietary patterns) and its influence on measures of breast density and body fatness. Measures of overall diet have been shown to more appropriately account for naturally occurring food and nutrient interactions that are often not properly considered in single nutrient studies. Data for the following cross-sectional studies were collected from young women (25-29y) who were initially enrolled in the Dietary Intervention in Children Study (DISC) and later participated in the DISC06 follow-up study. Informed consent was obtained at all points throughout the study. A total of 663 healthy, pre-pubertal, 8-10 year old children with elevated LDL-C, including 301 girls, were recruited from six clinical centers between 1988-1990. They were randomly assigned to a behavioral dietary intervention or usual care control group. The childhood DISC data collection included anthropometry which was used to calculate the childhood BMI-Z score based on the Center for Disease Control 2000 Growth Charts. Trained study staff measured body composition, anthropometrics, breast density and blood was drawn for DISC06 participants as young adults. Three 24-hour recalls were used to collect dietary information via telephone over the course of two weeks. The objective of the first study was to evaluate the associations between dietary energy density (ED; kcal/g) and three measures of breast density. Dietary ED is a property of food and can be calculated both with and without beverages. Individual linear mixed effects models were fit by maximum likelihood with robust standard error for each breast density outcome. After multivariate adjustment, a 27.0% (95% CI: 6.3 to 51.7%, p=0.009) change in the average percent dense breast volume (%DBV) and a -18.4% (-32.2 to -1.8%, p=0.03) change in the average absolute non-dense breast volume (ANDBV) was observed in our population for each 1 kcal/g increase of dietary ED from food-alone. Dietary ED from food and caloric beverages was also significantly inversely associated with the ANDBV with a -13.9% (95% CI: -22.3 to -4.7%, p=0.004) change being observed for each 1 kcal/g unit increase of dietary ED using this method. No associations were observed with the absolute dense breast volume (ADBV). We also stratified the analysis by median childhood BMI-Z score and observed stronger associations with all three breast density measures and food-only dietary ED in women with a BMI-Z score above 0.2. Overall, this research suggests that dietary ED, particularly from food alone, is associated with measures of breast density. The objective of the second study was to examine the associations between dietary ED and measures of body fatness among young women. Whole body percent fat as well as android and gynoid fat measurements were obtained via dual-energy X-ray absorptiometry (DXA) and height (nearest 0.5 cm), weight (nearest 0.2 kg), and waist circumference (nearest 0.5 cm), were all obtained by trained study staff. Individual linear mixed effects models were fit by maximum likelihood with robust standard error for each body fatness outcome. After multivariate adjustment, each 1kcal/g unit increase in food-only dietary ED was significantly positively associated with several body fatness measures including BMI [0.59 kg/m2; β (95% CI) = 0.59 (0.04 to 1.14), p=0.03)], whole body percent fat [2.08%; β (95% CI) = 2.08 (0.63 to 3.53), p=0.005], android fat [3.38%; β (95% CI) = 3.38 (2.26 to 4.50), p=<0.0001], gynoid fat [2.00%; β (95% CI) = 2.00 (0.55 to 3.45), p=0.008], and the android:gynoid fat ratio [0.3; β (95% CI) = 0.03 (0.01 to 0.05), p=0.007]. Food-only dietary ED was also positively associated with waist circumference; however, it did not reach statistical significance. This could be due to the larger amount of error associated with waist circumference measurements. There were no associations with dietary ED calculated with either food and caloric beverages or food and all beverages. Overall, this research suggests that dietary ED from food alone is positively associated with several measures of body fatness. In our third study, the primary objective was to assess the relationship between a posteriori dietary patterns derived via finite mixture modeling (FMM) and measures of body fatness and breast density, as described above. Finite mixture modeling, unlike other dietary pattern methodologies, does not restrict an individual to a particular dietary pattern, but rather each subject has a probability of belonging to one of the derived dietary patterns. An individual’s membership to a particular pattern was based on their highest respective posterior class-membership probability. Dietary patterns were derived using complete dietary data from 203 subjects. Five dietary patterns were derived and labeled based on the model-predicted servings of the 15 food groups in each of the intake patterns: “Western” [n=21, (10.3%)], “Dietary Guidelines” [n=40, (19.7%)], “Health Conscious” [n=47, (23.2%), “Alcohol and Vegetables”[n=39, (19.2%)], and “Non-Drinkers” [n=56, 27.6%)]. Briefly, the “Dietary Guidelines” pattern was the most consistent with the current 2010 Dietary Guidelines while the “Health Conscious” pattern had mediocre diet quality (as evidenced by dietary ED), but had a low prevalence of smoking and the highest amounts of physical activity. The “Alcohol and Vegetables” pattern had high intakes of alcohol (primarily beer and wine), healthy fats, and non-starchy vegetables. The “Western” pattern had high intakes of red and processed meats, sweets, and refined grains and finally, the “Non-Drinker” pattern consumed the lowest amount of alcohol. Individual linear mixed effects models were fit by maximum likelihood with robust standard error for each body fatness or breast density outcome. After multivariate adjustment, the “Dietary Guidelines” had a significantly lower dietary ED from food alone and food and all beverages than all of the other patterns. The “Western” pattern had the most unfavorable diet and often the least favorable body fatness outcomes. There were no associations with any of the dietary patterns and breast density measures except for a 34.7% (95% CI: 18.6 to 52.8, p=0.02) and 41.7% (95% CI: 26.7 to 58.5%, p=0.002) change in the ADBV in the “Non-Drinker” and “Alcohol and Vegetables” when compared to the “Dietary Guidelines” pattern. In conclusion, both dietary ED and dietary patterns were associated with body fatness and breast density, with the least healthful diets (High dietary ED from food-alone and the “Western” dietary pattern) often having the most unfavorable body fatness and breast density outcomes. Our research examines dietary profiles that may better aid in the development of future dietary recommendations. Future research should include the examination of dietary change across childhood and adolescence and its influence on body fat and breast density in adulthood.