bNonadditivity is accounted for by a few missing values. cIn parentheses: regular errors for quantative variables; percentages for categorical variables. In Table 2 , cases and handles are written by marginal case and control-mixed quintiles of total energy intake and meals groupings that are known main resources of flavonoids in the diet. Age-adjusted linear styles assessing the difference in the distributions of the indicated food groups and energy intake between cases and controls are also shown. Linear trends were assessed through (1995). bNonadditivity is accounted for by a few missing values. cvalues are age-adjusted and are interpretable as standard normal deviates. In Table 3 , cases and controls are distributed by the studied categories of flavonoids in the diet. There is evidence that breast cancer risk is inversely associated with flavone intake and less strong evidence for inverse associations with flavan-3-ol and flavonol intake. However, as in Table 2, the data in Table 3 are neither energy nor mutually adjusted and the patterns are not directly interpretable. Table 3 Distribution of 820 women with breast cancer and 1548 control women by marginal quintiles of flavonoid intake categories values are age-adjusted and are interpretable as standard normal deviates. In Table 4 , OR (and 95% confidence intervals) of breast cancer for 1?s.d. increase in the consumption of each of the examined major categories of flavonoids are presented. For each of the six categories of flavonoids, ORs derived from three different models are shown. In model I, the OR is adjusted for the sociodemographic, lifestyle and reproductive variables demonstrated in Desk 1, aswell for energy intake. These OR estimates could be confounded by the consumption of other flavonoid classes or by additional compounds in fruit and veggies that are inversely connected with breast malignancy risk in these data. In model II, the ORs are modified for fruit and veggie consumption, as well as the variables managed for in model I. In model III, the ORs for every group of flavonoids are modified mutually, aswell for the variables contained in model I. Therefore, OR estimates from versions II and III are much less at the mercy of confounding by additional compounds in vegetables and fruit, and additional flavonoids in the dietary plan. It is obvious that the OR for flavones is rather robust and indicates a statistically significant inverse association with breast cancer risk, even after taking into account the potential confounding effect of fruit, vegetable and other flavonoid intake. No such evidence exists for any other category of flavonoids examined. Table 4 Multiple logistic regression-derived ORs for breast cancer, per 1?s.d. increment of each of the examined flavonoid categories thead valign=”bottom” th align=”left” valign=”top” charoff=”50″ rowspan=”1″ colspan=”1″ Flavonoid category /th th align=”center” valign=”best” charoff=”50″ rowspan=”1″ colspan=”1″ Chances ratio /th th align=”middle” valign=”best” charoff=”50″ rowspan=”1″ colspan=”1″ 95% self-confidence interval /th th align=”middle” valign=”best” charoff=”50″ rowspan=”1″ colspan=”1″ em P /em -worth /th /thead em Flavanones (per 24.3?mgday /em ? em 1 /em em ) /em ????Model Ia0.920.83C1.020.09?Model IIb1.130.98C1.290.09?Model IIIc0.960.87C1.070.44 em Flavan-3-ols (per 16.2?mg?day time /em ? em 1 /em em ) /em ????Model Ia0.830.75C0.930.001?Model IIb0.970.83C1.140.71?Model IIIc0.930.78C1.110.43 em Flavonols (per 8.3?mg?day time /em ? em 1 /em em ) /em 0.810.73C0.900.001?Model Ia0.930.81C1.080.35?Model IIb0.910.78C1.060.22?Model IIIc??? em Flavones (per 0.5?mg?day time /em ? em 1 /em em ) /em 0.840.75C0.930.001?Model Ia0.860.77C0.960.01?Model IIb0.870.77C0.970.02?Model IIIc??? em Anthocyanidins (per 50.0?mg?day time /em ? em 1 /em em ) /em ????Model Ia0.860.76C0.970.01?Model IIb0.950.83C1.080.40?Model IIIc0.940.81C1.090.39 em Isoflavones (per 0.8?mg?day time /em ? em 1 /em em ) /em ????Model Ia1.070.97C1.180.15?Model IIb1.050.95C1.160.31?Model IIIc1.070.97C1.180.17 Open in another window aAdjusted for age group, host to birth, parity, age group initially pregnancy, age in menarche, menopausal position, body system mass index, total energy intake and alcoholic beverages consumption. bAdjusted because in model We, controlling also pertaining to fruit and vegetable usage. cAdjusted because in model We, and mutually between flavonoid classes. DISCUSSION We’ve found proof that flavones are inversely linked to breast cancer risk. The inverse association of flavone intake with breast cancer was only marginally affected when intake of fruits and vegetables, BIBW2992 supplier or other flavonoids was accounted for. The inverse association of flavones with breast cancer is not trivial, since it implies a 13% reduction in breast cancer risk per 1?s.d. (0.5?mg?day?1) of increase in the intake of the respective compounds. Inverse associations with breast cancer risk were also found for flavonols, flavan-3-ols and anthocyanidins. These associations were sharply attenuated and became nonsignificant, however, when intake of fruits and vegetables or other flavonoids were controlled for. We found no evidence that flavanones had a major effect on breast cancer risk and, for isoflavones, the evidence, if any, was for a positive rather than inverse association. Very few studies have examined flavonoids in relation to breast cancer risk. No association between cancer, including breast cancer, and total flavonoids was within the combined evaluation of the 16 cohorts of the Seven Countries Research (Hertog em et al /em , 1995). An inverse association between urinary excretion of phyto-oestrogens, which includes isoflavones, was discovered among Chinese ladies in Shanghai (Zheng em et al /em , 1999; Dai em et al /em , 2002) and Australian females (Ingram em et al /em , 1997; Murkies em et al /em , 2000), but no such association was obvious in an identical research in Netherlands (den Tonkelaar em et al /em , 2001). Regarding isoflavones, our data usually do not support those reported from research in China and Australia. There are many feasible explanations: intake of soya and soya items has been, but still is, not a lot of in Greece; an inverse association between isoflavones and breasts cancer risk might not be captured through a dietary intake study, but could be ascertained in research employing measurements of urinary excretion; or decreased urinary excretion of isoflavones could be a consequence instead of reason behind breast malignancy and the techniques linked to the medical diagnosis and treatment of the disease. There are no similar data in the literature regarding flavone intake with regards to breast malignancy risk therefore, at this time, our results concerning these substances is highly recommended as hypothesis producing instead of as documenting an authentic association. As flavones are largely produced from grains and vegetables (Peterson and Dwyer, 1998), and there is absolutely no evidence in the literature that grains or cereals are inversely connected with breast malignancy risk (World Malignancy Analysis Fund and American Institute for Malignancy Avoidance, 1997), our results indicate leafy vegetables and herbs as the food groups with potential beneficial properties for breast cancer risk. A source of concern is usually that vegetables were more strongly inversely associated with breast cancer risk in this Greek study (Trichopoulou em et al /em , 1995) than in other caseCcontrol and particularly cohort investigations (World Cancer Research Fund and American Institute for Cancer Prevention, 1997; Smith-Warner em et al /em , 2001; Lagiou em et al /em , 2002). However, consumption of vegetables and variability of consumption is usually higher in the Greek populace than in most other populations (Agudo em et al /em , 2002) and Greek food patterns are characterised by high consumption of wild greens that are rich in flavones (Trichopoulou em et al /em , 2001). Strengths of this study are its relatively large study size, the use of a validated food frequency questionnaire and the reliance on generally sound food composition databases (US Department of Agriculture-Iowa State University Database, 2002; US Department of BIBW2992 supplier Agriculture, 2003). Limitations of the study are the lack of a flavone-specific prior hypothesis, the emergence of findings after undertaking multiple analyses and questions concerning the applicability of US-based flavonoid food composition tables to Greek foods. A generic limitation is usually that confounding by dietary elements that have not really been measured can’t be managed for (Davey Smith and Ebrahim, 2003). Various types of flavonoids have already been reported to inhibit breast cancer cell replication, oestrone sulphatase activity and Mouse monoclonal to RBP4 mammary gland tumorigenesis in experimental analyses (So em et al /em , 1996; Huang em et al /em , 1997; Kuntz em et al /em , 1999). Nevertheless, except regarding isoflavones, there is absolutely no sufficient proof, experimental or elsewhere, linking particular flavonoid substances or types to specific activities along the way of mammary carcinogenesis. Therefore, the biological plausibility of an inverse association between flavones and breasts malignancy risk can, BIBW2992 supplier at this time, be looked at as only suggestive. In conclusion, we’ve found evidence that intake of flavones C however, not intake of flavonols, flavan-3-ols, flavanones or anthocyanidins or isoflavones C could be inversely linked to breast cancer risk. This inverse association works with with and could describe the reported inverse association of breasts cancer with intake of vegetables. Acknowledgments This study was partially supported by the University of Athens and a grant to Harvard University by the Samourkas Foundation. The task was also funded partly with Federal funds from the united states Department of Agriculture Research Service under contract number 53-3K06-01. The contents of this publication perform not necessarily reflect the views or policies of the US Department of Agriculture, nor will mention of trade names, commercial products or organisations imply endorsement by the US government. Partial support was also supplied by State of Florida, Department of Citrus. Initial support was provided by Massachusetts Department of Public Health’s Breast Cancer Research Grants Program, Boston, MA. We also thank the American Institute for Cancer Research, Washington, DC.. interpretable as standard normal deviates. In Table 3 , cases and controls are distributed by the studied categories of flavonoids in the diet. There is evidence that breast cancer risk is inversely associated with flavone intake and less strong evidence for inverse associations with flavan-3-ol and flavonol intake. However, as in Table 2, the data in Table 3 are neither energy nor mutually adjusted and the patterns are not directly interpretable. Table 3 Distribution of 820 women with breast cancer and 1548 control women by marginal quintiles of flavonoid intake categories values are age-adjusted and are interpretable as standard normal deviates. In Table 4 , OR (and 95% confidence intervals) of breast cancer for 1?s.d. increase in the consumption of each of the examined major categories of flavonoids are presented. For each of the six categories of flavonoids, ORs derived from three different models are shown. In model I, the OR is adjusted for the sociodemographic, lifestyle and reproductive variables shown in Table 1, as well as for energy intake. These OR estimates may be confounded by the intake of other flavonoid categories or by other compounds in vegetables and fruits that are inversely associated with breast cancer risk in these data. In model II, the ORs are adjusted for fruit and vegetable consumption, in addition to the variables controlled for in model I. In model III, the ORs for each category of flavonoids are adjusted mutually, as well as for the variables included in model I. Thus, OR estimates from models II and III are less subject to confounding by other compounds in fruits and vegetables, and other flavonoids in the diet. It is apparent that the OR for flavones is fairly robust and indicates a statistically significant inverse association with breast cancer risk, even after taking into account the potential confounding effect of fruit, vegetable and other flavonoid intake. No such evidence exists for any other category of flavonoids examined. Table 4 Multiple logistic regression-derived ORs for breast cancer, per 1?s.d. increment of each of the examined flavonoid categories thead valign=”bottom” th align=”left” valign=”top” charoff=”50″ rowspan=”1″ colspan=”1″ Flavonoid category /th th align=”center” valign=”top” charoff=”50″ rowspan=”1″ colspan=”1″ Odds ratio /th th align=”center” valign=”top” charoff=”50″ rowspan=”1″ colspan=”1″ 95% confidence interval /th th align=”center” valign=”top” charoff=”50″ rowspan=”1″ colspan=”1″ em P /em -value /th /thead em Flavanones (per 24.3?mgday /em ? em 1 /em em ) /em ????Model Ia0.920.83C1.020.09?Model IIb1.130.98C1.290.09?Model IIIc0.960.87C1.070.44 em Flavan-3-ols (per 16.2?mg?day /em ? em 1 /em em ) /em ????Model Ia0.830.75C0.930.001?Model IIb0.970.83C1.140.71?Model IIIc0.930.78C1.110.43 em Flavonols (per 8.3?mg?day /em ? BIBW2992 supplier em 1 /em em ) /em 0.810.73C0.900.001?Model Ia0.930.81C1.080.35?Model IIb0.910.78C1.060.22?Model IIIc??? em Flavones (per 0.5?mg?day /em ? em 1 /em em ) /em 0.840.75C0.930.001?Model Ia0.860.77C0.960.01?Model IIb0.870.77C0.970.02?Model IIIc??? em Anthocyanidins (per 50.0?mg?day /em ? em 1 /em em ) /em ????Model Ia0.860.76C0.970.01?Model IIb0.950.83C1.080.40?Model IIIc0.940.81C1.090.39 em Isoflavones (per 0.8?mg?day /em ? em 1 /em em ) /em ????Model Ia1.070.97C1.180.15?Model IIb1.050.95C1.160.31?Model IIIc1.070.97C1.180.17 Open in a separate window aAdjusted for age, place of birth, parity, age at first pregnancy, age at menarche, menopausal status, body mass index, total energy intake and alcohol consumption. bAdjusted as in model I, controlling also for fruit and vegetable consumption. cAdjusted as in model I, and mutually between flavonoid categories. DISCUSSION We have found evidence that flavones are inversely related to breast cancer risk. The inverse association of flavone intake with breast cancer was only marginally affected when intake of fruits and vegetables, or other flavonoids was accounted for. The inverse association of flavones with breast cancer is not trivial, since it implies a 13% reduction in breast cancer risk per 1?s.d. (0.5?mg?day?1) of increase in the intake of the respective compounds. Inverse associations with breast cancer risk were also found for flavonols, flavan-3-ols and anthocyanidins. These associations were sharply attenuated and became non-significant, however, when intake of fruits and vegetables or other flavonoids were controlled for. We found no evidence that flavanones had a major effect on breast cancer risk and, for isoflavones, the evidence, if any, was for a positive rather than inverse association. Very few studies have examined flavonoids in relation to breast cancer risk. No association between cancer, including breast cancer, and total flavonoids was.