Reproductive Factors and Breast Cancer Risk in Palestine: A Case Control Study

Background: Early age at menarche, late age at menopause, and late age at rst full-term pregnancy are linked to a modest increase in the risk of developing breast cancer (breast ca). This study aims to investigate the reproductive determinants of breast cancer among women in the West Bank of Palestine. A structured questionnaire was used to collect data in a case-control study (237 registered cases and 237 controls). A multivariate analysis model was used to adjust for the association between women’s reproductive factors and breast ca risk. This study was approved by Al Quds University Ethical Research Committee and the Ministry of Health research unit. Results: In the multivariate analysis, menarche after 13 years of age, use of oral contraceptives for more than two months, hormonal contraceptives use, and becoming pregnant at an early age ( ≤ 18 years) signicantly doubled the risk for breast ca. Women who used hormone replacement therapy (HRT) were signicantly associated with higher odds (6 times) of having breast ca versus those who did not use them (p <0.05). Similarly, nulliparous women showed 6 times the odds of breast ca compared with women with one or more children (p = 0.005). Also, parental consanguinity marriage and positive family history of the condition can be strong determinants for breast ca in this study. Conclusion: This study provides clear evidence that the use of reproductive hormones, whether as a birth control tool or for therapeutic purposes, must be rationalized worldwide and in Palestine in particular.


Introduction
Breast cancer (breast ca) is the most common form of cancer mortality among women in the world 1 .Breast cancer is a multi-factorial type of cancer. Being genetically predisposed or having a family history of a rst-degree relative with breast ca was shown to increase the cancer incidence [2][3][4] . Parental marriage to a relative was also shown to increase the risk 5 . Among women aged 40 years or more, breast ca is related to increased risk 6,7 . Modi able risk factors such as obesity, physical inactivity, sedentary behavior, and poor dietary patterns were also shown to be related to breast cancer risk 8- 10 .
The effect of reproductive factors strongly supports a hormonal role in its aetiology 11-14. Early age at menarche, late age at menopause, and late age at rst full-term pregnancy are linked to a modest increase in the risk of developing breast cancer 14,15 . However, multiple full-term pregnancies and longterm breastfeeding decrease the risk of breast cancer 16,17. Reproductive surgeries such as ovariectomy, tubal sterilization, and hysterectomy may also affect the breast cancer risk by altering hormone levels before menopause or by bringing forward the age at menopause 18, 19 . Long-term use of hormone replacement therapy (HRT) 20, but not long-term use of oral contraceptives (OC), was also related to an increased risk of breast ca 21 . Moreover, it was noted that the time elapsed since last oral contraceptive use was associated with a higher risk of breast ca than recent use 22 .
Breast cancer is the most common and widespread type of cancer in Palestine, and ranks as the third cancer that causes death. It constitutes 17% of all cancer cases. At the end of 2017, there were 503 new cases documented in the West Bank and 327 new cases recorded in the Gaza Strip. The rate was 33.1 new cases per 100,000 females annually 23 . Few studies have tackled the risk factors of breast ca in Palestine 2,24. In Gaza, a study among women aged 18 to 60 years suggested that a positive family history of breast ca, high body mass index, and some common diseases (hypertension, diabetes mellitus) maybe epigenetic factors that promote the occurrence of breast ca 2. The reproductive determinants of breast cancer among women in the southern region of the West Bank will be presented here. The study ndings may help to clarify the interaction of these factors in the development of breast cancer among Palestinian women.

Study context
The cancer burden in Palestine is expected to increase and will pose a substantial challenge for the healthcare system. The limited nancial and infrastructural resources, plus political uncertainty, exacerbate the problem 25. Cancer care, diagnosis and treatment services are provided in four West Bank hospitals. However, isotope scans like PET-CT are not available and all such cases are referred to Israeli hospitals. The shortage of specialized physicians and of drugs, chemotherapy, and radiation therapy present a challenge in providing proper care for cancer patients 26 . This study was conducted at the major governmental hospital: BeitJala hospital in the southern West Bank. BeitJala hospital has an oncology department and daycare clinic that offers daycare medical services for cancer patients in the central and southern areas of the West Bank. Therefore, this study aimed to explore the various reproductive risk factors for breast cancer in the West Bank of Palestine.

Study design
This case-control study was conducted at Beit Jala governmental hospital in the West Bank of Palestine over the period 2016 to 2017.

Study cases and control selection
Based on hospital chart number, 237 women were selected at random as study cases from those attending the daycare oncology department or the chemotherapy unit of BeitJala hospital. These women had a pathologically con rmed breast carcinoma and were aged 40 years or more at the time of interview.
To serve as a comparable and representative control group, 237 women of the same age distribution and geographic area were randomly recruited from the screening program for breast ca. The subjects in the control group were con rmed as free from breast ca and had never been suspected of having any previous neoplastic disease or any other cancer. Their medical records were checked to include a normal (BIRADS 1) mammography. Those referred by a physician for a suspected history of breast problems were excluded. Only a very low proportion (2%) of selected women (study cases and controls) refused to participate in this study.
This study was approved by Al Quds University Ethical Review Committee. Written approval was obtained from the Ministry of Health to access the patients' records from the oncology department and cancer registry. All women provided written informed consent.

Data collection
The medical records of cancer patients were used to retrieve information related to the breast ca: date of diagnosis, stage at diagnosis, type of cancer, and therapy strategy.
Trained female interviewers administered an in-person structured questionnaire during the patient visit to the oncology department. Controls were contacted by a nurse from the mammography department and were invited to participate. If a control refused to come to the clinic, the interview was conducted via a phone call.
The questionnaire included questions on demographic and lifestyle factors; parental consanguinity marriage; contraceptive history; use of hormone therapy; menstrual history; pregnancy and breastfeeding history; medical history, including cancer and mammogram history; and family history of malignancy. Women were also asked whether they had undergone surgery to remove one or both ovaries partially or fully. Women were also asked whether they had undergone a hysterectomy or tubal sterilization, and the approximate month and year of the procedure(s).
Statistical analysis SPSS version 23 (IBM Corp., Chicago, IL, USA) was used for the data analysis. Bivariate and multivariate unconditional logistic regressions were used to assess the association of breast ca with independent variables. Crude and adjusted odds ratio (AOR) and 95% con dence intervals (CIs) were calculated to determine the precision of the estimates. The level of signi cance used was 5%. The p-value < 0.05 indicated signi cance.

Results
In total, 237 cases and 237 age-matched controls were included. The mean age of those in the study was 54.6 (SD=10.9) years and 54 (SD= 9.9) years for the control group (p>0.05).
Most of the study cases were diagnosed at stage 2 and 3 of cancer (35% and 30% respectively). Most of the study cases (83%) discovered that they had cancer after they noticed a mass and only 17% were diagnosed by screening. More than half of the study cases had ductal carcinoma (n=138, 58%); 19 women had lobular carcinoma (8%); 5 women had follicular carcinoma (2.1%); 4 had mixed type (ductal lobular 1.7%); and 30% did not have a documented type in their les. Almost all cases had undergone chemotherapy treatment (98%). About 83% of cases had undergone partial mastectomy and half of them had undergone a full mastectomy. Furthermore, 75% of cases had surgery as the rst-line treatment and did not receive neo-adjuvant therapy. Table 1 shows the socio-demographic characteristics for cases and controls. Study cases and controls had signi cantly different distributions for multiple characteristics such as educational level, home type, family size, and parental consanguinity, but not for others (Table 1). Study cases had higher levels of education than those in the control group (41%) had more than 10 years of education versus 24.5%).

Socio-demographic factors
Controls had larger families than study cases (mean 6.35, SD 2.6 versus mean 5.61, SD 2.96 respectively) but lived in smaller residences than the study cases. About 43% of study cases had married a rst-degree relative compared with 21% in the control group. The odds ratio between socio-demographic factors and breast ca are summarised in Table 2. The odds of breast ca were higher among women with more than 12 years of education versus those with less education. The odds of breast ca were 3.87 times higher among women living in separate houses compared with those living in apartments (95% CI: 2.36-6.33, p =0.00). Women with no children were 2.5 times more likely to get breast ca versus women with children. Interestingly, the odds of breast ca were 2.5 times higher among women married to a rst cousin (consanguinity marriage) compared with those whose spouse was not related or were married to a second-degree relative (95% CI: 1.60-4.08, p=000).

Discussion
The reason for international variations in the incidence of breast ca remains unclear. These variations can be seen between both high and low-income countries. Many of the risk factors for breast ca have been investigated but require further examination in individual nations.
In this study, we examined a broad spectrum of risk factors for breast ca, including female reproductive factors. The reproductive risk factors for breast ca identi ed in Palestinian women are similar to those observed in other studies. This study provides clear evidence that late menarche poses an additional risk for breast ca. Early marriage and having children early in life, both popular in the Palestinian community, were shown to increase the odds of breast ca. The role of oral contraceptives and hormonal replacement therapy on women's health was also clearly shown and there should be rational use of hormones, whether as a birth control tool or for therapeutic purposes. Having children proved to be protective against breast ca but as most married women in Palestine breastfeed their children, we could not show that breastfeeding is a protective factor for breast cancer among the study group. However, we can still highlight the role of breastfeeding in breast ca protection. More in-depth investigations are needed to identify the relationship between various factors, especially the protective role of having children and breastfeeding practices on breast ca in Palestine. Special attention should be devoted to the particular social and cultural factors related to sexual and reproductive issues among women in Palestine.
Several studies have indicated that women with high socioeconomic status (SES) are at risk for breast ca with an overall estimate of 20% increased risk 27. This positive association was clearer among Hispanic and Asian women 28, and not only for breast ca but for other cancers such as colon, ovary, and melanoma cancers 29. Our study found that breast ca was more common among more educated rather than less educated women, and in women with a lower family income rather than women with a higher family income. In the north of Palestine, a previous study showed that there was a four-fold increase in the risk of breast ca among highly educated women 24; this was also reported among Egyptian women 30. In European women, a direct dose-response relationship was seen between educational level and postmenopausal breast ca incidence 31. Our results found that women living in an apartment had a signi cantly lower risk of getting breast ca compared with those living in a separate home; this was assumed to be due to a higher SES. Several studies showed that lower SES increased the risk of breast ca because women were less aware of screening techniques and diagnosis 32. Our results could be explained by the fact that women with a higher family income can afford health insurance and are more willing to spend money on their health and better medical care access. Greater awareness among educated women about mammography screening tests is very clear in Palestine. It is worth mentioning that screening in Palestine is free of charge for all women over 40 years of age. Another possible explanation is that the more a woman is educated, the later she marries, the later the age of pregnancy, the shorter the period of breastfeeding, and the lower parity is characteristic of women from higher SES. Indeed, socio-economic inequalities could affect the time of diagnosis, survival or mortality due to cancer despite improved knowledge, reduction of risk factors for cancer, early diagnosis, and treatment 33. Consanguinity is becoming a very strong factor for cancers and other genetic diseases in many countries 34,35. Our study showed that daughters of unrelated parents had a decreased breast ca risk, whereas the risk increased 2.5-fold for those with rst-degree related parents. A similar nding was reported in the United Arab of Emirates (UAE) in which having unrelated parents halved the risk (RR=0.5, 95%CI: 0.27-0.93) 35. A study among Israeli Arabs of Palestinian origin showed an increase in diabetes and duodenal ulcers 34. Consanguineous practices in populations might affect the gene frequency in these populations, which could have a major effect on the carrier rate of such genes. Therefore, in countries with high consanguinity, the incidence of diseases and syndromes should be monitored with caution.
It is believed that up to 10% of breast ca cases in Western countries were due to genetic predisposition with a threefold increase in the risk of breast ca among those with a family history of breast ca 36. In our study, women with a family history of breast cancer had a fourfold increased risk of breast ca. In Qatar, a country with high consanguinity marriage, a study showed that consanguinity was lower in breast cancer patients than in controls, but a family history of breast cancer was signi cantly more prevalent in breast cancer patients 37 . The risk of breast ca ranged from 1.5 to 3.6 in a pooled analysis depending on the relative in question, with the highest risk reported among women who had a mother or a sister with breast Early age at menarche, late age at menopause, and late age at rst full-term pregnancy are linked to a modest increase in the risk of developing breast cancer 14-16. Also, parity and age of marriage are among the most common extrinsic factors that modulate breast cancer risk. It is well documented that parity has a dual effect on breast cancer risk, with an increased risk during 5 to 10 years after pregnancy, followed by a strong and life-long protective effect 15 .
In several studies, older age at menarche was inversely associated with breast ca risk. The high-risk groups were females with menarche before the age of 11 years 4 . Around 117 studies showed that the breast ca risk increased by a factor of 1.050 for every year less at menarche 15, and a delay of two years at menarche led to a 10% reduction in breast ca worldwide 40. In our study, older age at menarche was shown to be associated with an increased risk of breast ca. The risk increased three-fold with menarche at the age of 13 or more. In the north of the West Bank, the estimated risk was 6.5 which also showed an increase the risk for breast ca 24 . The protective result of menarche at an older age was explained by the lower cumulative number of ovulatory cycles, which is negatively associated with the risk, younger age at menarche, and older age at menopause, means a female would have more cycles and an increased risk 41,42 .
According to the Palestinian Central Bureau of Statistics (2016), the mean age of rst marriage was 19.8 years in the southern region of Palestine 43 . Consequently, many women may have their rst pregnancy and rst delivery at a young age (below 18 years). Women with breast ca in our study had a mean age of marriage of 20.4 years (SD=5.44). Our multivariate results showed an inverse association between age at rst marriage and age of rst pregnancy for breast ca. We could not see any difference according to whether a woman had her rst child before the age of 18 years or after. In contrast, a study in the north of the West Bank showed that there was a 10% increase in the risk of breast ca when the rst marriage was below 20 years of age 24. Another study in the Gaza Strip showed that women who had their rst pregnancy after the age of 35 years had an 11-fold increase in breast cancer risk 2.
Our results revealed no signi cant association between full-term pregnancies and the risk of breast ca. However, the number of full-term pregnancies was negatively associated with breast ca risk in almost all studies, even in the Western world. This result was consistent for not only one type but for all subtypes of breast ca in pre-and post-menopausal women. The reduction in the risk ranged from 18% to 60% 44 . In the north of Palestine, a 50% decrease in risk was reported among women with four full-term pregnancies or more 24. One of the well-established protective factors against breast ca is breastfeeding. In our results, almost all women who had children had engaged in breastfeeding, but the protective effect in our analysis was not in the breastfeeding itself but in its duration. Previous studies found that breastfeeding itself was

Conclusions
This is the rst epidemiological study in Palestine to investigate the risk for breast ca based on women's reproductive factors. Signi cant differences in breast ca were found between the study cases and control group: age at puberty, use of OCP and HRT, nullparity, early marriage, early pregnancy, and early delivery. All these factors indicated a higher risk of breast ca alongside being from a family with a history of breast cancer and married to a rst cousin. In Palestine, most women breastfeed so more in-depth investigations are needed to identify the protective role of having children and breastfeeding practices on breast ca protection. Moreover, the Palestinian community must be aware of the effect of early marriage and parental consanguinity on the risk of breast cancer. These results are very important in clinical practice and women must be aware of the results on their health of the use of OCP and HRT. The use of reproductive hormones whether as a birth control tool or for therapeutic reasons must be rationalized. We encourage more studies to be conducted on breast cancer to tackle the speci c types of breast ca in all areas of Palestine and other unknown determinants. Special attention should be given to the particular social and cultural factors related to sexual and reproductive issues among women in Palestine.

Declarations
Ethics approval and consent to participate This study was approved by Al Quds University Ethical Research Committee, which is based on the Helsinki declarations. Therefore, all study methods were performed following the Helsinki guidelines and regulations. Al Quds University ethical research regulations adhere to Helsinki regulations Written approval was obtained from the Ministry of Health to access patient records from the oncology department and cancer registry. All women provided written informed consent.
Consent for publication NA Availability of data and materials The datasets used and/or analysed during the current study are available from the corresponding author upon reasonable request.
Competing interests: The authors declare that they have no competing interests.

Funding
This study was funded by Al Quds University research funds.
Author contributions NS and IK designed the survey and developed the study tool. IK was responsible for data collection, data entry, and primary analysis. NS and IK participated in the study of advanced analysis and the development of study tables. NS was responsible for writing the manuscript. All authors read and approved the nal manuscript.