Study Design and Population. This study included 305 women, participants of three studies involving MD, 97 were from a cross-sectional study (21), 7 from an isoflavone intervention study (22), and 201 from a soy intervention study (23). For the latter two studies, baseline mammograms and urine samples were used for the analysis. After the approval of the Committee on Human Studies at the University ofHawaii, all subjects provided informed consent. Women for all three studies were recruited at mammography clinics in Honolulu. Eligible women had no previous history of BC and required a normal mammogram at baseline. The subjects for the intervention studies were free of serious medical conditions, had regular menstrual periods, intact uterus and ovaries, were not on oral contraceptives or other hormones, and had no intention of becoming pregnant within a year. Due to the nature of the nutritional intervention, only women who reported a dietary consumption of< 7 servings of soy food/week were eligible. All subjects completed a validated food frequency questionnaire (24), which included questions on reproductive, medical, and anthropometric factors. In the cross-sectional study, self-reported information was used to classify women's menopausal status and hormone use (HRT). The majority of subjects (n = 208) donated an overnight urine specimen, while those ofthe cross-sectional study (n = 97) only a spot urine sample.
Mammogram Density Assessment. Cranio-caudal views of the mammogram were obtained from the clinics after complete radiologic evaluation and ruled out any malignancy. The films were scanned into a PC using a Kodak LS-85 X-ray digitizer with a pixel size of260 |xm (resolution = 98 pixels/inch). One ofthe authors (GM) performed computer-assisted MD assessment using a Canadian method (25). The reader chooses a threshold value that defines the outline of the breast, and then selects the best threshold to identify the regions that represent MDs. The pixel count corresponding to the dense area is determined by the computer, as is the total area within the outline of the breast. Percent MD was calculated as the ratio of the dense area to the total area ofthe breast multiplied by 100. A random sample of 58 mammograms was read in duplicate. The intra-class correlation coefficients (26) were 0.95 (95% CI, 0.92-0.97), for the size of the dense areas, and 0.98 (95% CI, 0.97-0.99) for the total breast area, and of 0.97% density (95% CI, 0.95-0.98).
Urinary Hormone Analysis. Urine concentration of Ei, E2, T, and ADIOL was performed with slight modifications previously published (27). Briefly, 1.0-ml urine samples were hydrolyzed, purified by solid phase extraction, and HPLC. Hormone concentrations were measured by RIA on the dried extracts. All measurements were done in duplicate, including hydrolysis, solid phase extraction, HPLC, and RIA. For quality control, two control samples containing known amounts of steroids were included for all the analytical steps in each sample batch. The detection limits were: 0.02 ng/ml for Ei and E2, 0.08 ng/ml for ADIOL, and 0.02 ng/ml for T. Intra- and inter-batch coefficients of variations were 4.7 & 16%, respectively, forE, (at 3.3 ng/ml), 1.7% & 14% forE2 (at 0.32 ng/ml), 4.7% & 14% for T (at 3.3 ng/ml), and7.0%& 11% for ADIOL (at 15.5 ng/ml). 2-OH-and 16a-were measured by solid-phase enzyme immunoassays after enzymatic hydrolysis with Helix Pomatia (Estramet, Immunacare Co., Bethlehem, USA). Mean intra- and inter-batch coefficients ofvariations were 10 & 15%, respectively, for both analytes.
Statistical Analysis. The SAS statistical software package version 8.2 (SAS Institute Inc., Cary, NC) was used for data management and statistical analyses. In the questionnaire, subjects marked all ethnic backgrounds that applied to themselves and to their parents. Summary categories were assigned according to the following rules: A woman was classified as Caucasian if both parents had some Caucasian ancestry and shared no other ethnic background. Subjects with no more than three ethnic backgrounds were classified as Chinese, Japanese, or Filipino, ifboth parents were ofthe same ethnicity or ifthe mother was of the respective ethnic background and the parents shared no other ethnic background. Because of the similarity in percent MDs, the 86 Japanese, 28 Chinese, and 9 Filipino women were combined into one Asian category. In agreement with rules applied in the State of Hawaii (28), women with any Hawaiian background were classified as Native Hawaiian. Because of their mixed ancestries, the Native Hawaiian women (n = 35) were included into the other category containing Pacific Islanders, African-Americans, Latinas (n = 24), and women with mixed ethnic backgrounds that did not fit any of the above categories (n = 13).
Body Mass Index. BMI was calculated as the ratio of weight in kg divided by the square of the height in m. Non-normally distributed variables were transformed using their natural logarithm. Percent MD was classified into five categories: <10%, 10 to 24.9%, 25 to 49.9%, 50 to 74.9%, and >75%. To explore associations between MD and urinary hormone measurements, we computed Spearman correlation coefficients and included potential confounders (29). Then, we applied analysis of variance to test for associations between ethnicity, hormone levels, and mammographic characteristics with adjustment for confounding variables (30). In addition, we computed least-squares means for the urinary hormone levels by category ofpercent MD using the proc glm procedure in the SAS software package (30). Finally, we performed trend tests to investigate a possible relation between the different hormones and percent MD. We regressed the mean level of the hormones onto the mean MD of each of the five MD categories.
Was this article helpful?
I lost over 60 pounds and 4+ inches off my waist without pills, strenuous exercise, or any of the things that the diet experts tell you to do...and I did it in less than 4 months! If you have the desire, and can read through my e-book , then this is for you! I could have easily made it a lot more difficult, with stacks of information that people will never read, but why?