Oral Contraceptive Used More than 5 Years is Associated with Increased Risk of Breast Cancer: A Meta-Analysis of 28,776 South East Asian Women

Background: Breast cancer associated with variety of hormonal inter-ethnic and reproductive factors. Duration of oral contraceptives use with breast cancer risk is not well understood in Southeast Asian women. Methods: A comprehensive literature search of published articles from January 2001 to June 2019 (PubMed, ProQuest, and EBSCO online article databases) and meta-analysis were performed to link the duration of oral contraceptive application with risk of breast cancer among women in Southeast Asia. We identified the reference category, being ≤5 years' duration and >5 years' duration of oral contraceptive application. We use the Fixed and random-effect models to rely pooled odds ratios (OR). Egger's and Begg's test used for publication bias was presented with funnel plots. All analyzed data in Review Manager 5.3 (RevMan 5.3) and Stata version 14.2. Results: A total of 385 studies were reviewed and 10 studies involving a total of 28,776 women were included in a meta-analysis. This study found a slight increase in breast cancer risk in Southeast Asian women with oral contraceptives application ≤5 years with OR = 1.21 (95% CI 0.96-1.52, p>0.05). A higher risk of breast cancer was found in women with oral contraceptive application >5 years with OR = 2.66 (95% CI 1.79-3.94, p<0.00001). Publication bias and heterogeneity were not found particularly in a group of Southeast Asian women with oral contraceptive application for more than 5 years. Conclusion: Oral contraceptives use more than 5 years are at a higher risk in breast cancer among women in Southeast Asia. Although other reproductive factors including age at first childbirth, menarche, menopause, and lactation might influence the risk of breast cancer.


Background
Breast cancer is the most frequent cancer and the second leading cause of cancerassociated mortality among women worldwide. A total of 2,088,849 new cases and 626,679 deaths have been related to breast cancer in 2018 [1]. Multifactorial etiology has been identified as one of the risk factors known is oral contraceptive application [2,3].
Previous studies have revealed that combined oral contraceptives reduce endometrial and ovarian cancer risks through ovulation suppression [4,5]. These studies found contrary results of contraceptive pills use with occurrence of breast cancer. An increased breast cancer risk correlated with used in oral contraceptive pills [4,5,6] meanwhile other studies [7,8,9,10] have shown inconclusive results.
Other determinants of oral contraceptive use and the occurrence of breast cancer are the length of the use, dose-response, and age of the users. Another study showed a significant elevate in the number of young women who use oral hormonal contraceptive as well as women who begin to use oral contraceptives before the age of 20 [2]. Furthermore, the studies found there is no association between breast cancer and oral contraceptives use [2,4]. Increased breast cancer risk particularly in women with oral contraception application for more than 4 years before the first pregnancy [5]. Another study reported breast cancer risk increasing in the oral contraceptives pills use for more than eight years [6].
Oral contraception has recently emerged as the third most common method for family planning and the most widely distributed across the globe compared to other contraceptive methods. It is estimated that 6.4% of Asian women use contraceptive pills [11]. Meanwhile, the proportion of oral contraceptive users in Southeast Asia is 12.8% [12]. Oral contraceptive users by country in Southeast Asia revealed that Thailand as the highest users (35%), followed by Indonesia (13.6%) and Vietnam (8.6%) [11,13]. Therefore, oral contraceptive application in Southeast Asia is relatively high yet the correlation with breast cancer risk is still under study.

Study design and research sample
In this meta-analysis study, we followed the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) Statement [14]. The samples of this research included published research articles which were published from January 2001 to June 2019 in PubMed, ProQuest, and EBSCO of online article databases. In each study we identified the reference category, being ≤5 years' duration and >5 years' duration of the application of oral contraceptive.

Operational definitions
This study comprised of independent variables, (a) ≤5 years' duration and (b) >5 duration of oral contraceptive use, and a dependent variable, i.e. breast cancer risk.

Research procedure
The procedure of this research was started by collecting data through published research articles identification on the oral contraceptive application length and the risk of breast cancer in Southeast Asian women on PubMed online article databases, ProQuest, and EBSCO ( Figure 1). These following keywords are used treating as title or abstract for the literature search: ("oral contraceptive" OR "hormonal oral") AND ("breast cancer" OR "breast neoplasms"). We found 385 articles were identified by examining the article titles, abstract and full-text. Furthermore, this study classified the research articles into two groups based on the oral contraceptive application length with ≤5 years' duration and >5 years' duration.
The article will be excluded if it is: (a) not breast cancer outcome, (b) conducted by using other study designs other than case-control or cohort study (c) included insufficient data for extraction.

Data collection technique
Online searching was performed in data collection. The collected data was limited by the articles written in English. The article type was limited to original research one. The publication date of the articles was limited from January 2000 to May 2019.
The research subject was limited to humans only. Conceivably significant title unique articles were inspected, while the insignificant articles were prohibited.
Those possibly significant unique articles will be evaluated in full-content structure, while then the unessential articles were excluded. The sample inclusion criterias were researched on the duration of oral contraceptive application, including ≤5 years' duration or >5 years' duration, and breast cancer in Southeast Asian women with restrospective study and prospective study design. The exclusion criterias were (a) the inclusion criterias were unsatisfyingly fulfilled, (b) the articles were unavailable in full-text form, and/or (c) the data provided in the full-text form was not sufficient for extraction of datas. The following datas were also collected from the articles: the name of the original writter, study location, study type, and a number of several samples.
The information from all of the articles that fulfilled inclusion criterias per under a standardized protocol carefully extracted by two investigator, while contradictions were settled by three different investigators. Newcastle-Ottawa Quality Assessment Scale (NOS) was performed for evaluation quality of research articles. Articles were indicated poor (score 0-3), moderate (score 4-6) and high quality (score 7-9) [15].

Data analysis
Data analysis was conducted to get pooled and combined odd ratios of the collected articles. Odd ratio with 95% confidence intervals (CIs) was utilized to pool the results. These test presented the minimum statistical power article with limited numbers and sample sizes was indicated by I 2 >50% significantly heterogeneity.
A meta-analysis study was performed to determine the duration of oral contraceptive application and breast cancer risk in Southeast Asian women ( Figure   2). The figure presented an insignificant result of ≤5 years' duration of oral contraceptive application and breast cancer risk in Southeast Asian women with OR = 1.21 (95% CI 0.96-1.52, p >0.05). The Heterogeneity test in these studies (P heterogeneity = 0.0004; I 2 = 70%) resulted in a variety of heterogeneous results on the breast cancer risk. In contrast, oral contraceptive application for >5 years and breast cancer risk in Southeast Asian women had a significant result with OR = 2.66 (95% CI 1.79-3.94, p <0.00001). Different from the ≤5 years' application duration, heterogeneity tests in these studies (P heterogeneity = 0.14; I 2 = 45%) resulted in multifariousness of homogeneous research on breast cancer risk.
Publication bias in articles collected desribed by presenting Funnel plots on the duration of oral contraceptive application with breast cancer risk among women in Southeast Asia (Figure 3). In figure 3, results showed as significant publication bias detected in articles on oral contraceptive ≤5 years' duration of oral contraceptive and risko of breast cancer among women in Southeast Asia with Egger's test (P = 0.011) and Begg's test (P = 0.09). In contrast, in articles on > 5 years' duration of oral contraceptive use and risk of breast cancer among women in Southeast Asia ( Figure 4), our study found that there was no significant publication bias with Egger's test (p = 0.270) and Begg's test (p = 0.308).

Discussion
This study result showed that >5 years' length of oral contraceptive pills use increased breast cancer risk in Southeast Asian women with risk opportunity was more than doubled. The results were also homogeneous which means the studies will give consistent or insignificantly different if the study performs at the same time and place. Furthermore, Indonesia has been known that this country has the greatest risk opportunities for breast cancer, followed by Thailand and Malaysia. Our study findings described possibly relate to the high rates of oral contraceptive application in those country mentioned in advance [11,13,24].
The previous studies found breast cancer incidence related to oral contraceptive application in several ethnicities. Non-Hispanic Caucasians have the highest percentage of oral contraceptive application (55%) as well as the highest number of incidence of breast cancer (6%). Interestingly, the lowest proportion of oral contraception application (37.5%) yet being the second-highest in the percentage of breast cancer incidence (5.7%) in non-Hispanic/Pacific Islanders. Non-Hispanic African-Americans and Hispanics have the same percentage of oral contraceptive application (52.2%) with a relatively lower figure of breast cancer incidence with 3.8% and 4.7%, respectively [25]. Our study had found the percentage of oral contraceptive application with breast cancer incidence followed by a range of 3.7-13.9%. The figure was higher in some ethnic than others as we had described, possibly due to oral contraception application percentage is known was also relatively high in the Southeast Asian region with 12.8% [12]. Data from 39 case-control studies led from 1980 to 2006 have discovered a association of contraceptive pills with breast cancer among premenopausal [26].
Another study recommended that present utilization of contraceptive pills which substance is estrogen appears to somewhat elevate the breast cancer risk. A few factors adding to the advancement of breast disease incorporate hormones, which is estrogen itself [27][28][29]. The use of contraceptive pills which contains estrogen the breast tissue to be presented to large amounts of hormones for longer periods. That estrogen stimulates growth factors that exist in breast cancer cells resulting in tumor progression [30,31].
The previous research on the impact of age at beginning of utilization of the pill on the rate of breast cancer growth presumed that ladies that had begun utilizing the pill before 18 years old contracted disease 4 years sooner [32]. In a study, the researcher accepts prolonged use increases the risk. Another study found a relative risk of 2.2 with over 10 years of length of using [33]. The other study found the afinity between breast cancer and contraceptive pills use (OR=2.11). The investigation results show that women that utilization preventative pills have more noteworthy requirements for preventive and screening measures [34].
Based on data oral contraceptive is the most common means of contraception in the world and Southeast Asian countries, utilized by millions of women [35][36][37].
According to the duration of oral hormonal contraceptives use more than 5 years will increase breast cancer risk which will be higher compared to those who have never used hormonal contraception. If a person stops taking hormonal contraception for 5 years then she will have no risk of breast cancer. This result is supported by a theory found about the imbalance of the hormones estrogen and progesterone used in hormonal contraception [4,38,39]. This hormonal imbalance has resulted in a feedback mechanism that can physiologically control the number of hormones in the body when it is not functioning properly [40]. It was triggered by the estrogen receptors up regulation so that the number of hormones continues to increase. Exposure to sexual hormones over a period of more than 5 years will increase the proliferation of breast cells and increase breast cancer stem cell mitosis [41].
A meta-analysis data in this study has several limmitations. First, there were 2 potential articles should be included in this study but the full-text articles are inavailable. Second, there were 2 studies are also potentially included in data analysis yet the data presented were incomplete and different risk factors in the calculations were also performed.
Based on our analysis, as we found of the significant association of oral hormonal contraceptives use with breast cancer risk as well as increase for more than five years' duration of contraceptive pills use, we suggest people to select effective and efficient long-term contraceptive methods to prevent pregnancy, yet safe in accordance with infertility and breast cancer risk. Another suggestion is breast cancer screening programs are necessary for Southeast Asian countries.

Conclusions
This study confirms the significant correlation of more than five years' duration of oral hormonal contraceptives use and breast cancer risk in Southeast Asia. Therefore, we recommend fertile women who want to prevent pregnancy for a relatively long period, should use long-term contraception methods.

Consent for publication
Not applicable.

Availability of data and materials
The datasets analyzed of the present study could be obtained from the corresponding author upon reasonable request.

Competing interest
The authors declared no potential conflicts of interest.

Not applicable
Authors' contributions RDN, TA, ID, and WAH conceived and designed the study. RDN, SLA, LL collected the data and performed analysis and interpretation. RDN wrote the first draft with critical feedback from TA, SLA, TA, WAH, and ID. All authors read, reviewed and edited the draft and approved the final version of the manuscript. Abbreviation: NOS, Newcastle-Ottawa Quality Assessment Scale