Dietary factors modify post-menopausal breast cancer risk: a case-control study from Turkish Cypriot population

Background: Being potentially modifiable risk factor of breast malignancy, the role of diet in the development of breast cancer (BC) is of great concern. As up to 40 % of cancers can be prevented through dietary strategies; therefore, this case-control study is conducted with the aim to investigate the effects of frequently used dietary factors and postmenopausal BC risk in Turkish Cypriot population. Material and method: Total 786 postmenopausal women including 401 histologically confirmed BC cases and 385 control, recruited from two hospitals i.e. Near East Hospital and Doctor Burhan Nalbantoglu State Hospital Nicosia, Turkish Republic of Northern Cyprus, between the month of July to December 2016. A standardized interview procedure is used and the information is collected using a structured questionnaire from participants after giving informed consent form. For data analysis, 20 is used. Age-adjusted odds ratios (OD) and 95% confidence interval (CI) were calculated by logistic regression before and after adjusting for potential confounding effects of other factors. Results: The multivariable adjusted model confirmed a 3-fold increased BC risk with daily oil use of ≥ 40 mL (OR = 3.22, 95%CI 2.01-5.17, p<0.001). And 4.1-fold increased risk was associated with 4 to 6 daily servings of sugar intake (OR = 4.19, 95%CI 1.79-9.80, p = 0.001), this risk increased to 7.5-folds (OR = 7.5, 95%CI 3.25-17.32, p<0.001) when the consumption of sugar was increased to > 6 servings per day. Daily 1 to 2-liter water intake was associated with 64% decreased BC risk (OR = 0.36, 95%CI 0.20-0.63, p = 0.001). While, no significant association were observed between consumption of full-fat dairy products (FFDP), olive oil, coffee intake and BC risk. Interestingly, daily 3 or more cups of tea intake were associated with 54% decreased risk of BC (OR = 0.46, 95% CI 0.22-0.98, p = 0.043). Conclusion: The study suggests that the risk of BC can be reduced by limiting the consumption of oil and sugar and increasing daily water intake more than a liter.


Introduction
Breast cancer is the most prevalent malignancy in Turkish Cypriot women (Pervaiz et al., 2017). There are various established risk factors for BC worldwide including, exogenous and endogenous hormonal exposure, various reproductive factors (early menarche, late menopause, late pregnancy, not breastfeed, and being non-parous) and lifestyle factors (smoking, alcohol and exercise) in addition to high penetrant gene variants i.e. BRCA1&2, ATM, PALB2, and CHEK2) (Rudolph et al., 2016).
The role of diet in the aetiology of BC is also noteworthy as large international variations exist in BC incidence rates (Horn-Ross et al., 2002). These variations might be ascribed to the antioxidant properties of certain selected nutrients that influencing DNA mutation, DNA repair, growth factors stimulations. These nutrients may also have some anti-estrogenic effects and metabolic detoxification (Potter and Potter, 1997).
Recently, a randomized control trial among women at high risk of cardiovascular disease has provided the first evidence about the reduction of BC incidence by diet intervention. Women were randomly assigned to Mediterranean diet patron which is generally rich in plant food, fish and olive oil. About half of these women were as likely to develop invasive BC as those who were only assigned to a diet with only reduce fat intake (Toledo et al., 2015).
Dietary factors have been thought to be the main modifiable risk factors for cancer and it is estimated that up to 40% of cancers could be prevented through dietary strategies (Surh, 2003 various food nutrients that increase the circulation level of oestrogens and growth factors including insulin-like growth factors I (Potter and Potter, 1997).
In TRNC population, The Mediterranean dietary patron is increasingly changing to the western dietary patron, leading to the rise of diseases incidence including cancer. As to enhance general health and reduce the risk of BC, women can alter their diet successfully. Therefore, the purpose of this case-control study is to assess the strength of association between the consumption of various commonly used diet factors including oil, sugar, water, dairy products, olive oil, alcohol, coffee and black tea and BC risk in this part of the island. To the extent of our knowledge, this is the first epidemiological investigation on BC risk and dietary factors in this population.

Study Subjects
This case-control study was carried out in connection with our previous study on BC risk factors in North Cyprus population. The analysis comprised of total 786 menopausal women; 401 BC cases and 385 control healthy women without any malignancy. Both cases and control were recruited from two hospitals of the island i.e. Near East Hospital Nicosia and Doctor Burhan Nalbantoglu State Hospital Nicosia, TRNC between the month of July to December 2016.

Data collection
A standardized interview procedure was used for both cases and control and information regarding sociodemographic factors and the various commonly used dietary factors were collected on a detailed questionnaire after giving informed consent form. From control, questions about diet intake in the previous year were asked, while from cases in the previous year before diagnosis were asked. Furthermore, age at the time of interview of control women is noted and for cases, age at the time of diagnosis is noted.
In addition to the questions about the quantity of specific dietary factors consumption i.e. oil or fat, sugar, olive oil, water, full fats dairy products (FFDP), coffee and tea intake, questions about age, weight, height, education, income status, marital status, family history of BC, age at menarche and menopause, parity, age at first full-term pregnancy (FFP), number of children, breastfeeding (at least 1 month), oral contraceptive and HRT use, smoking status and exercise were included in the questionnaire. All information was self-reported except BMI which was based on actual measurement. The exercise was considered a 30 minutes' walk or physical activity for at least four times a week. Smoking was considered one cigarette a day for at least 6 months. A gestation period of 24 weeks or more is considered pregnancy, oral contraceptive used and HRT was considered the use for at least one month. Sugar consumption was considered as anything containing added sugar i.e. jam, jelly, syrup, frozen desserts, nonfrozen desserts, candies, chocolate and soft drinks etc. for a serving size of 1 teaspoon (5-7grammes) and one glass of soft drink (250-300 grammes) was indicated. For FFDP i.e. butter, cheese, full-fat milk etc. serving size of 100 grammes is considered. Only frequency and not the quantity of olive oil use were asked in the food frequency questionnaire.

Analysis
The difference between socio-demographic characteristics, dietary factors and cases and control were first assessed by cross-tabulation and chi-square test. In order to assess the degree of association of potential risk from dietary factors and BC, unconditional logistic regression model before and after adjusting for potential confounders are used. The fit of the model is assessed on the basis of Pearson chi-square or Hosmer-Lemeshow goodness-of-fit statistics. Age group is not used as a potential risk for BC but is used as a confounder in the uni-and multivariable regression models. The statistical analysis was performed using SPSS statistical software version 20.

Results
The analysis confirmed that more cases than control were obese (BMI ≥ 30), single, with family history of BC, with earlier menarche, late menopause, with no parity, with no or ≤ 2 children, never breastfeed, used HRT, were smokers, physically inactive, and consumed more fatty food, more sugar and less water, less FFDP, and more likely to use alcohol. But, no significant difference was reported for education, income status, age at FFP, oral contraceptive used, olive oil used, daily coffee and black tea intakes in cases and control ( Table 1).

Discussion
The key purpose of the present study was to evaluate the strength of association of various commonly used dietary factors and BC risk in TRNC. Only postmenopausal cases and control women were included in the study since diet has been indicated to have a different influence on pre and post-menopausal BC (Psaltopoulou et al., 2011). It is to be noted that all the factors used here are selfreported excluding BMI. Of the studied dietary factors, we found that consumption of excess oil/fats and sugar were associated with increased BC risk. Also, daily water intake of more than one liter was associated with reduced BC risk.
Our results are in consistence with other studies. For instance, a prospective study on a large heterogeneous population of European women has shown the high fats diet to increase BC risk, particularly high saturated fat intake increases the risk of receptor-positive BC (Sieri et al., 2014). However prospective observational studies on association between fat consumption and BC risk have inconsistent findings, but many studies have confirmed that this association may be due to unspecific quantification of fat intake (Prentice et al., 2013) despite the fact that, other comprehensive study did not confirm this (Key et al., 2011), the cause of inconsistencies may be due to the fact that specific type of fat (not of the total fat) is linked only to some BC types. In nested case-control study within the French EPIC cohort, it is shown that trans fatty acids increase the risk of BC (Chajès et al., 2008). Evidence supports that the high fats intake increases the concentration of bio-available serum sex hormones (Parry et al., 2011), which is the main risk factor for BC. It also enhancing reabsorption in the intestine and increasing blood fatty acid level that may increase free estrogenic concentration in blood serum (Rock et al., 2004). The role of dietary fats in cancer tumor formation are given in a recent study in Nature. The study has shown the effects of high fats diet on the intestinal stem cells lineage and has established a mechanism by which progenitor cell when exposed to high fats diet become more stem cells like and prone to oncogenic transformation (Beyaz et al., 2016). Same is the case of dilatory sugar intake. Recently, in a multiple mouse model study, the impacts of dietary sugar on mammary gland tumor development and the mechanism involve were investigated, and found that sucrose intake comparable to the amount of western diet led to increasing tumor growth and metastasis by inducing 12-lipoxygenase signaling, when compared with nonsugar starch diet (Jiang et al., 2016). Furthermore, our study showed that daily 1-2 liter water intake decreased the risk up to 64%, an only small increase of 1% were observed with further increased water intake. Malignancies mostly studied in connection with water include bladder and colorectal cancers and only rarely with BC (Michaud et al., 1999). Conversely, a hospital-based study on diet and beverage consumption and BC risk by Stookey et al observed that water intake is significantly associated with 69% reduced BC risk (adjusted odds ratio of 0.31, 95%CI, 0.13-0.72) (Stookey et al., 1997).
No significant association between BC risk and consumption of FFDP and olive oil were reported. In contrast, the majority of case-control studies carried out in Mediterranean countries consistently concluded an inverse association of BC risk and olive oil consumption (Sieri et al., 2014). It is reported that the hydrocarbon squalene compound in olive oil, exerts a beneficial effect on oxidative stress and DNA damage in mammary epithelial cells (Warleta et al., 2010), polyphenols from olive oil may have a possible role in the prevention of BC (Casaburi et al., 2013). Being a part of Mediterranean island, olive oil and dairy products consumption are frequent in TRNC. Therefore, large follow-up study may Although coffee and tea consumption showed no significant association with BC risk in this study, it is reported that a daily 3 or more cups of black tea were found to associated significantly with decreased BC risk in the adjusted model. This association was not significant in the uni-variable model. As one of the most commonly proposed pathways leading to carcinogenesis is oxidative DNA damage which is strongly determined by body iron storage (Toyokuni, 2009), coffee and tea are supposed to inhibit iron absorption in the small intestine, and subsequently decreasing oxidative stress through reducing stored iron in the body (Morck et al., 1983). Recently it has been confirmed that coffee and not tea was associated with lower level of oxidative DNA damage and low body iron storage in women (Hori et al., 2014). There is inconsistency in the association of BC risk and intake of coffee and tea, in the published literature (Harris et al., 2012). Furthermore, a large meta-analysis indicated no association of black tea intake and BC risk (Nie et al., 2014).

Conclusion
It is concluded that there is a strong association between consumption of fats, sugar and BC risk. Water intake has beneficial effects on the primary prevention of BC. The most appropriate approach against cancer is the preventive strategies. To the best of our knowledge, this is the first report to validate an association of BC risk and dietary factors in this part of the island. Hopefully, these findings will give new insights in BC epidemiology. Nevertheless, these results need confirmation by long-term prospective studies.