A Review of the Epidemiology of Breast Cancer in Asia: Focus on Risk Factors

Background and Aim: Breast cancer is the most prevalent cancer in women. To date, regional differences in breast cancer risk factors have not been identified. The aim of our review was to gain a better understanding of the role of risk factors in women with breast cancer in Asia. Methods: We conducted a PubMed search on 15 March 2016, for journal articles published in English between 2011 and 2016, which reported data for human subjects in Asia with a diagnosis of breast cancer. Search terms included breast neoplasm, epidemiology, Asia, prevalence, incidence, risk and cost of illness. Studies of any design were included, except for review articles and meta-analyses, which were excluded to avoid duplication of data. No exclusions were made based on breast cancer treatment. We reported the results using the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines. Results: A total of 776 abstracts were retrieved. After screening against the eligibility criteria, 562 abstracts were excluded. The remaining 214 abstracts, which were published between 2013 and 2015, were included in this review. Results were summarized and reported under three categories: incidence, prevalence or outcomes for breast cancer in Asia; modifiable risk factors; and non-modifiable risk factors. We found that the increased risk of breast cancer among participants from Asia was associated with older age, family history of breast cancer, early menarche, late menopause, high body mass index, being obese or overweight, exposure to tobacco smoke, and high dietary intake of fats or fatty foods. In contrast, intake of dietary fruits, vegetables, and plant- and soy-based products was associated with a decreased breast cancer risk. While based on limited data, when compared to women from the United States, women from Asia had a decreased risk of breast cancer. Conclusions: This review of 214 abstracts of studies in Asia, published between 2013 and 2015, confirmed the relevance of known non-modifiable and modifiable risk factors for women with breast cancer.


Introduction
The most recent data from the World Health Organization (WHO) shows that breast cancer (BC) is the most prevalent cancer in women, with an incidence of 1.7 million cases in 2012 (25 % of all female cancers). It is the fifth leading cause of death from cancer overall (International Agency for Research on Cancer, 2016). The incidence is similar in regions identified by the WHO as more-and less-developed, at 13% (788,000) and 11% (883,000) respectively (International Agency for Research on Cancer, 2016). WHO figures from 2012 estimate the international 5-year prevalence for BC at 6,232,000; by region (as defined by WHO) the rate in South-East Asia is 735,000 compared to 1,936,000 in Europe and 1,618,000 in the Americas (International Agency for Research on Cancer, 2016).
Identified risk factors for BC may be non-modifiable such as sex, age, genetic characteristics including

Materials and Methods
We performed a systematic review of the current body of evidence available through PubMed. The search was conducted on 15 March 2016. Medical subject headings (MeSH) search terms used included: breast neoplasm, epidemiology, Asia, prevalence, incidence, risk and cost of illness. The search was limited to human studies, published in English between 2011 and 2016. Studies were included in our review if the participants had BC of any age, histopathology, grade or stage, and received treatment of any type; studies needed to report information on epidemiology, prevalence, incidence, risk factors, and cost of illness. Our review included all study designs, except for review articles and meta-analyses, which were excluded to avoid duplication of data. Studies were only included if they reported data from countries within South East Asia, Far East Asia, or Western Asia excluding the Middle East. Articles were excluded if their primary objective was to examine the correlation of BC to other types of cancer or to examine surgical treatments for BC.

Results
Our search identified 783 articles; seven articles were excluded due to no abstract being available. We did not find any articles published in 2016. To reduce the volume of data identified in our search, only articles published within the previous three years (2013)(2014)(2015) were included. This resulted in 169 eligible articles being excluded based on publication date (2011 and 2012). After the exclusion of duplicates and articles that did not meet the inclusion criteria, a total of 214 remaining articles were included in this review. The flow diagram of the literature selection process is presented in Figure 1.
Results have been summarized and reported under three categories: incidence, prevalence or outcomes for breast cancer in Asia; modifiable risk factors; and non-modifiable risk factors. A summary of studies showing risk and protective factors influencing changes in BC risk is provided in Supplementary Table 2; a visual representation of risk and protective factors is provided in Figure 2 and Figure 3. Due to the large number of studies eligible for review, descriptive data without an association with altered BC risk have not been provided in the tables. Key trends are summarized in the sections below.

Incidence, prevalence or outcomes for breast cancer in Asia
BC was reported as the leading type of cancer in females in Japan (Matsuda et al., 2014;Hori et al., 2015), and China (Zhou et al., 2015). In addition, the incidence of BC was shown to increase over time in China Zhou et al., 2015), India (Asthana et al., 2014) and Thailand (Virani et al., 2014), although it was unclear if this was due to improved identification (i.e. screening and detection) or a true reflection of increased incidence. While BC incidence is predicted to increase in Hong Kong by 2025, mortality is projected to decrease from 9.3 to 8.6 per 100,000 women, respectively (Wong et al., 2015). This may be related to Hong Kong's socio-economic development, and the improvements in survival may be due to treatment advancement and improved health service delivery (Wong et al., 2015). This is in contrast to China, where BC mortality increased over a 5-year period (annual percentage change of 6.89%) between 2006 and 2010 . In one study from Malaysia, almost all (n=1453, 99.2%) had an estimated 10-year risk of BC of less than 2% (Hassan et al., 2015). In Korea, the rate of 5-year recurrence-free survival (RFS) was 90.8%, 10-year RFS was 81.9%, and 5-year overall survival (OS) was 94.6% (Kim et al., 2014;Park et al., 2015b). However other studies from Korea have reported the 10-year OS rate at 86.9% (Park et al., 2015b) and another at 61.4% . In Hong Kong and Vietnam, 5-year OS rates were 90.5% and 74%, respectively (Lan et al., 2013;Kwong et al., 2014). The 5-year OS rate was similar in male BC patients from Hong Kong (n=132, 78.7%) (Kwong et al., 2014). Supplementary Figure 1 Flow diagram of the literature selection process

Dietary Factors
Dietary factors associated with increased BC risk included intake of total n-6 polyunsaturated fatty acids (PUFA) (Kiyabu et al., 2015), vitamin D deficiency or insufficiency (Shi et al., 2014a;Park et al., 2015a), high levels of serum cadmium (Nagata et al., 2013;Itoh et al., 2014;Ding et al., 2015), high intake of salt (Park et al., 2014b), sugar (Sulaiman et al., 2014), meat (Ko et al., 2013), and saturated fat and oils (Balasubramaniam et al., 2013;Wang et al., 2013a). In women with BC, the trace elements, cadmium, magnesium, copper, cobalt and lithium, were found in high levels. This suggests a possible association between serum levels of trace elements and breast cancer risk, however, this requires further investigation and if confirmed, the modulation of trace elements may help to reduce breast cancer risk (Ding et al., 2015).
High intake of dietary fiber was associated with decreased BC risk (Sulaiman et al., 2014) along with high levels of α-carotene, β-carotene, β-cryptoxanthin, lutein/ zeaxanthin , and soy products or isoflavones (Ko et al., 2013;Li et al., 2013a;Wada et al., 2014) (Table 2). In addition, BC patients had significantly lower levels of manganese, aluminum, iron and titanium than women who did not have BC (Ding et al., 2015). High consumption of fruit and vegetables has also been linked to a reduction in breast cancer risk in Asia. (Sangrajrang et al., 2013) In one study, the consumption of crucian carp was associated with increased BC risk, but total freshwater fish intake and consumption of black carp and silver carp were associated with decreased risk .

Location of residence
Location of residence were identified as a modifiable factor associated with increased BC risk and was identified in three papers (Shi et al., 2014b;Wu et al., 2014b;Fei et al., 2015). Results of the studies identified that living in an urban region of China (versus a rural region) led to an increased BC risk. One paper noted a higher incidence of BC was correlated to higher population density, percentage Epidemiology of Breast Cancer in Asia activity (Wang et al., 2013a) were discussed in the papers.

BMI and physical activity
A Thai study evaluated the relationship between BC risk and obesity and physical activity. The study found that being underweight at ages 10 and 20 years showed a reduced risk of BC in all women (OR=0.70, 95%CI 0.56-0.88 and OR=0.74, 95%CI 0.59-0.93, respectively) and undertaking regular physical activity also showed a reduced risk (OR=0.78, 95% CI 0.68-0.98). When reviewed by type of physical activity undertaken, those who had the highest levels of walking had the greatest reduction in risk (OR= 0.58, 95% CI 0.38-0.88) (Sangrajrang et al., 2013).

Reproductive history and breastfeeding
Several of the identified studies have investigated the effect of parity, abortion, age at pregnancy and history of breastfeeding on BC risk. In a Korean study, parity of five or more increased the risk of luminal A and B breast cancer (HR 1.95, 95% CI, 0.96-3.97, p=0.0055; and HR 1.12; 95% CI, 0.42-3.02, p=0.0073 respectively) and an early age of first birth (less than 20 years) was associated with increased risk of luminal A BC recurrence (HR 1.61, 95% CI, 0.62-4.26; p=0.039). However, between one and three childbirths was associated with decreased risk of luminal A and B (HR, 0.56; 95% CI, 0.34-0.91, p= 0.0055; and HR, 0.32; 95% CI, 0.17-0.61, p=0.0073 respectively) (Lee and Oh, 2014). In India, women with BC, had a higher age at last child birth, lower mean duration of breastfeeding, and a higher number of abortions compared to women who did not have BC (Bhadoria et al., 2013). In China, Wu et al., (2014a) reported reduced BC risk in parous women with 3 or more induced abortions or 3 or more induced abortions after first live birth, compared with parous women without induced abortion. Also in China, age over 25 years at first full-term pregnancy was associated with increased risk of BC, but a history of full-term pregnancy was associated with a decreased risk of BC (Pei et al., 2014). Huang et al., (2014) reported that among participants from China, of non-agricultural population, and second industry output; in contrast, incidence was negatively correlated with the percentage of population employed in primary industry (Fei et al., 2015). Other modifiable risk factors including active or passive smoking (Mizoo et al., 2013;Pimhanam et al., 2014;Tong et al., 2014;Wada et al., 2015), high body mass index (BMI) (Mizoo et al., 2013;Sangrajrang et al., 2013;Suzuki et al., 2013;Wang et al., 2013b;Anothaisintawee et al., 2014;Wada et al., 2014;Fu et al., 2015), use of oral contraceptives (Bhadoria et al., 2013;Anothaisintawee et al., 2014;Poosari et al., 2014), high perceived level of stress, and low level of physical later age at either first live birth, or at first pregnancy and last pregnancy were associated with increased BC risk (p-trend=0.002, 0.015, 0.008, respectively). Among participants from Japan, BC was increased in women who performed mixed feeding (HR 1.12, 95% CI, 0.92-1.37, p-trend=0.014) or feeding only with formula (HR 1.80, 95% CI, 1.14-2.86, p-trend=0.014) compared with women who only breastfed (Sugawara et al., 2013).

Breast cancer risk by geographic location
Several studies demonstrated differences in the prevalence and mortality associated with BC between nations in the Asian region, and significantly, differences within individual nations. Ginsburg et al., (2017) argue that the majority of women who die from BC live in low-income countries; this is in contrast to high-income countries where most women who are diagnosed with BC will survive. Malvia et al., (2017) indicate that the overall age adjusted rates for BC in India is as high as 25.8 per 100,000 women but lower than that of the United Kingdom (95 per 100,000). Indian mortality rates were reported at 12.7 per 100,000 women, which is comparable to the BC mortality rates in the UK (17.1 per 100,000). Further, disparities in BC risk and mortality can be seen within individual countries in Asia and within different regions within the same country. When Indian BC mortality rates are reported by region, some regions reported rates as high as 41 per 100,000 in Delhi and as low as 12.4 per 100,000 in Barshi rural region (Malvia et al., 2017).

Non-modifiable risk factors for breast cancer in Asia Age
In India, the mean age of BC diagnosis was 53 years (Tulsyan et al., 2014), and the median age was 45 years (Krishnatreya et al., 2014). This was similar to data from China, which indicated that the estimated risk of recurrent breast cancer increased with age (mean=46 years) . Peaks in age-specific breast cancer incidence were at 50-59 years and > 85 years . In a Thai study, incidence was projected to increase at a higher rate in women aged 50 years or over compared to those aged under 50 years (period to 2029) (Virani et al., 2014). Other studies have reported that benign breast disease was more likely to occur in women aged under 40 years, with BC more likely in women aged over 40 years (Kotepui and Chupeerach, 2013;Kotepui et al., 2014).

Family history
In China (Zhou et al., 2013;Chen et al., 2014b;Pei et al., 2014), Pakistan (Shamsi et al., 2013), Hong Kong (Tse et al., 2015, India (Bhadoria et al., 2013), and Korea (Park et al., 2014a), a family history of BC was associated with an increased risk of BC. In Hong Kong, risk of BC (in women with estrogen receptor-positive tumors) was higher in women with an affected mother (OR = 3.97, 95%CI: 1.46-10.79) when compared to women with an affected sister (OR = 2.06, 95%CI: 1.07-3.97) (Tse et al., 2015). A study in Nepal showed that family history of BC was evident in a higher proportion of women who developed BC aged less than 40 years (3%) compared with those aged 40 years or over (0.3%) (Thapa et al., 2013).

Ethnicity
Race/ethnicity and age-specific incidence rates have been shown to increase continuously until 80 years in white women from the United States (US), but plateau or decrease after 50 years in Asian women . A greater absolute risk for all BC subtypes has been found in women from the US National Cancer Institute's Surveillance, Epidemiology, and End Results 18 registries database (SEER 18) compared to those from Malaysia (Horne et al., 2015). A further study demonstrated that, following adjustment for population structure, the median age of BC diagnosis was lower in China (50-54 years) than in the United States and European Union (55-59 years) (Song et al., 2014b). The authors suggest that this is due to racial differences in genetics and lifestyle and that BC screening programs should be commenced at an earlier age for Chinese women (Song et al., 2014b).

Reproductive health
Menopausal status was associated with BC risk in studies conducted in Thailand (Anothaisintawee et al., 2014), India , Pakistan (Sohail et al., 2013), and Korea . Among participants from India, compared with controls, women with BC had a lower mean age at menarche (13.20±1.33 and 14.58±0.85 respectively, p=0.00) and a later age of menopause (49.38±5.21 and 47.89±3.96 respectively, p=0.002) (Bhadoria et al., 2013).  reported that late menopause was significantly associated with increased BC risk. In Korea, age at menarche and menopausal status or age at menopause were risk factors for BC . A second study in Korea confirmed the association with early age at menarche and BC risk, irrespective of tumor subtype (Chung et al., 2013). In Pakistan, older age at menopause was associated with increased BC risk (Shamsi et al., 2013).

Hormone receptor status
I n C h i n a , e s t r o g e n l e v e l s a n d e s t r o g e n receptor-mediated serum bioactivity were higher in BC cases than controls (Lim et al., 2014). A study conducted in Malaysia found that the majority of BC cases were hormone receptor-negative (Horne et al., 2015). Conversely in a Korean study, 69.5% of BCs were hormone receptor-positive (Song et al., 2014a) and 44.6% of BCs were estrogen receptor-positive in an Indian study (Singh et al., 2014). Among participants from Korea, when estrogen receptor-positive/progesterone receptor-positive tumors were compared to estrogen receptor-positive/progesterone receptor-negative tumors, women with the latter were 4 years older at diagnosis and more likely to be postmenopausal (Chung et al., 2013).

Genetic risk factors
Several of the studies investigated the association between gene polymorphisms and BC risk. In addition, there were also studies investigating the association between BC risk and DNA methylation (Kuchiba et al., 2014), length of CAG repeats (of androgen receptor Epidemiology of Breast Cancer in Asia
-Average annual percentage change of 11.3%.
-Mortality rate has increased yearly since 1991, a trend predicted to continue over the following 5 years.
-After median 60 months of follow-up (range, 8-60 months), local and distant recurrence of breast cancer was observed in 23.4% of cases.
-Between 2002 and 2006, the 5-year relative survival rate for breast cancer was > 40%. In Jiangsu between 2006 and 2010: -11,013 new cases of female breast cancer and 3,068 deaths due to female breast cancer were identified.
-The annual average crude incidence and age-standardized incidence were 25.2 and 17.9 per 100,000, and mortality rates were 7.03 and 4.81 per 100,000 respectively.
-Incidence and mortality increased with annual percentage changes of 4.47% and 6.89%, respectively.

India
Asthana S, et al. (Asthana et al., 2014) -The annual percentage change in incidence for breast cancer ranged from 0.46% to 2.56%.

Malaysia
Hassan N, et al. (Hassan et al., 2015) -Of the Chinese women in this study, 99.2% had a 10-year risk of breast cancer of < 2%.

Other risk factors for breast cancer in Asia
Increased BC risk was associated with insulin use for 3 or more years (mortality risk) among participants from Thailand (Tseng, 2015); polycystic ovary syndrome ; tonsillectomy (Sun et al., 2015a); sleep apnea (Chang et al., 2014); multiple sclerosis among participants from Taiwan (Sun et al., 2014c); inflammatory bowel disease among participants from Asia (Tsai et al., 2015); women who had an organ transplant ; and diabetes among participants from China (Wang et al., 2013b). In addition, a study conducted in China found that hypertension was observed in a higher proportion of BC patients with type 2 diabetes compared with non-diabetic participants with BC (41.5 versus 26.1%, p= 0.001) ) demonstrating a link between breast cancer and other comorbidities. Another study conducted in China showed that premature ovarian failure was associated with decreased incidence of BC among postmenopausal women (Wu et al., 2014c). Higher levels of education have been linked to a reduction in breast cancer risk in Asia. (Lan et al., 2013) Other investigated risk factors included hormone therapy (Lan et al., 2013), radiotherapy (Chitapanarux et al., 2013), BC-associated depression and anxiety , environmental factors, including dichlorodiphenyltrichloroethane (DDT) exposure (Tang et al., 2014), and ratio of second to fourth finger length (Hong et al., 2014).

Non-modifiable risk factors
The key risk factors for BC are non-modifiable, the strongest of which is being female (American Cancer Society, 2016). Similarly to the studies in Western population (American Cancer Society, 2016), advancing age has also been identified as a non-modifiable risk factor for developing BC in participants from Asia Kotepui et al., 2014;Virani et al., 2014). The studies identified two periods in which an increased BC risk emerged; first period at approximately 40-60 years, and a second peak at 85 years and over (Kotepui and Chupeerach, 2013;Chen et al., 2014b;Kotepui et al., 2014;Krishnatreya et al., 2014;Virani et al., 2014;Wu et al., 2014b;Varughese et al., 2015), which may be an important consideration for health service planning.
Approximately 5-10% of BC cases are thought to arise from gene mutations; the most common cause of hereditary BC is an inherited mutation in the BRCA1 and BRCA2 genes (American Cancer Society, 2016). Although the lifetime risk of BC among those with BRCA1 mutations can be as high as 80%, individuals with mutations in BRCA1 or BRCA2 generally have a 55-65% or 45% increased risk of BC, respectively (National Cancer Institute, 2017). In 2016, the Asian BRCA consortium released its first paper on the management of hereditary breast and ovarian cancer (HBOC) in 14 Asian nations. This paper highlighted the disparity in the management of HBOC screening (only approximately 4,000 cases of BC in Asia had benefited from genetic services) between nations and recommended that policy makers, healthcare sectors and researchers address the limitations in HBOC testing and management (Nakamura et al., 2016). Results of the studies included in this review which considered the role of BRCA mutation in BC risk indicated that women with BC and a BRCA1 mutation had an earlier age of disease onset, higher nuclear grade of BC, and a younger age of hospitalization (Chen et al., 2013a;Liu et al., 2013b;Pei et al., 2014;Sun et al., 2014a;Fu et al., 2015;Haryono et al., 2015;Wu et al., 2015) when compared to those without the BRCA mutation. Other gene mutations can also lead to BC, for example, CHEK2, PTEN, CGH1, STK11 and PALB2 (American Cancer Society, 2016). However, they are generally viewed as being much less common and most do not increase the risk of BC to the degree seen with the BRCA genes; other non-BRCA mutations have also been linked to increased BC. Interestingly, certain gene variations were associated with decreased BC risk (e.g. RAD52, OCT4, FASL, IGFIR, APE1, BARD1, IL4, IL21) (Kang et al., 2013;Liu et al., 2013a;You et al., 2013;Joshi et al., 2014b;Kang et al., 2014;Tulsyan et al., 2014;Xu et al., 2014;Wu et al., 2015), or both increased and decreased risk, depending on the gene variation investigated (e.g. MTHFR, ESR1, VEGF, FAS, OGG1, TGFB1) (Pooja et al., 2013;Kapahi et al., 2014;Lu et al., 2014;Luo et al., 2014;Wang et al., 2014c;Xu et al., 2014;Yu et al., 2014;Lu et al., 2015;Son et al., 2015).
Many women with BC do not have a family history of the disease. However, having a first-degree relative with BC is associated with a two-fold increase in risk of BC, and having two first-degree relatives can increase the risk three-fold (American Cancer Society, 2016). Accordingly, we found that family history of BC was associated with an increased risk in studies conducted among participants from China (Zhou et al., 2013;Chen et al., 2014b;Pei et al., 2014), Pakistan (Shamsi et al., 2013), Hong Kong (Tse et al., 2015, India (Bhadoria et al., 2013), and Korea .
Compared with other races or ethnic origins, Asian, Hispanic and Native American women may have a decreased risk of developing BC (American Cancer Society, 2016). This was supported by two studies which identified a reduced risk of BC among women from Asia and Malaysia compared with populations of women in the US (Horne et al., 2015;Sung et al., 2015).
Early menarche (before age 12 years) and late menopause (after age 55 years) are associated with increased risk of BC, likely due to a longer lifetime exposure to estrogen and progesterone (American Cancer Society, 2016). Several studies demonstrated a link between menopausal status and BC risk and was seen across participants from Thailand (Anothaisintawee et al., 2014), India , Pakistan (Song et al., 2014a), and Korea . As expected, the studies showed an increased BC risk in women who had a lower age of menarche (Bhadoria et al., 2013;Chung et al., 2013) and a later age of menopause (Bhadoria et al., 2013;Shamsi et al., 2013;. An increased risk of BC was identified in participants with pre-existing medical conditions including multiple sclerosis and polycystic ovary syndrome (Sun et al., 2014c;Shen et al., 2015).

Modifiable risk factors
Being overweight or obese is associated with an increased risk of BC, especially in women who have reached menopause. This is due to body fat, rather than the ovaries, becoming the primary source of estrogen in menopause (American Cancer Society, 2016). Internationally, several epidemiological studies have found that postmenopausal women with an increased BMI are at an increased BC risk of between 10-60% (Tahergorabi et al., 2016). In addition, being overweight or obese can increase insulin levels, and higher insulin levels have also been associated with increased risk of BC (American Cancer Society, 2016). Several of the studies in this review showed a correlation between increased BC risk and a high BMI, being overweight or obese, and low level of physical activity (Mizoo et al., 2013;Sangrajrang et al., 2013;Suzuki et al., 2013;Wang et al., 2013a;Wang et al., 2013b;Anothaisintawee et al., 2014;Wada et al., 2014;Fu et al., 2015). In addition one study showed that underweight BMI at a young age (at ages 10 and 20 years) were associated with a decreased risk of BC (Sangrajrang et al., 2013).
In parallel with these observations, physical activity (walking) was associated with decreased risk of BC (Sangrajrang et al., 2013), although the exact amount of exercise needed to induce a protective effect remains unclear (American Cancer Society, 2016).
The connection between BC, diabetes and obesity, was considered in a number of the identified studies. Breast cancer mortality risk was increased in women who had used insulin for more than 3 years (Tseng, 2015); additionally BC risk was associated with the presence of diabetes (Wang et al., 2013b). when compared to nontamoxifen users (Sun et al., 2014b). Hypertension was observed in a higher proportion of BC patients with type 2 diabetes when compared to non-diabetic participants with BC .
Women who have not had children or who had their first child after age 30 may have a slightly higher BC risk, whereas having many pregnancies and becoming pregnant at an early age may reduce the risk of BC (American Cancer Society, 2016). This was supported by some of the identified studies, in which BC risk increased when first full-term pregnancy was at an age greater than 25 years. Some studies also found that a later age at first live birth, first pregnancy and last pregnancy/birth also increased BC risk (Bhadoria et al., 2013;Huang et al., 2014;Pei et al., 2014). BC risk was found to decrease in women with history of full-term pregnancy (Pei et al., 2014). However, one study noted that increased parity and early age at first birth were associated with increased risk of BC recurrence (Lee and Oh, 2014).
Diet and the use of vitamins are modifiable risk factors with an unclear effect on BC. High consumption of fruits and vegetables was associated with decreased BC risk (Sangrajrang et al., 2013), and high intake of fatty foods was associated with increased BC risk (Balasubramaniam et al., 2013;Wang et al., 2013a). Data suggested that an intake of vitamin D, dietary fiber and intake of plant-and soy-based products may be protective for BC (Ko et al., 2013;Li et al., 2013a;Wada et al., 2013;Shi et al., 2014a;Sulaiman et al., 2014;Wang et al., 2014a;Park et al., 2015a), whereas high intake of dietary salt, sugar and meat may increase the risk of BC (Ko et al., 2013;Park et al., 2014b;Sulaiman et al., 2014).
There is increasing evidence which supports the link between BC risk and heavy smoking over a long period of time (American Cancer Society, 2016). This was confirmed in several articles demonstrating a clear association between BC risk and exposure, either active or passive, to tobacco smoke (Mizoo et al., 2013;Pimhanam et al., 2014;Tong et al., 2014;Wada et al., 2015).

Limitations
There are several limitations to this study. Firstly, due to the large number of articles identified, the results included are based only on the content of article abstracts, as assessed by a single reviewer. Therefore, the results analysis is limited due to some missing data on comparators, effect sizes and other statistical analyses. The missing data also make it difficult to prioritize results according to the strength or quality of included studies and/or data. The strength of the article may also be limited due to the absence of methods to assess risk of bias within individual studies and across studies, and the absence of additional analyses of the extracted data.
In conclusions, this literature review has underscored the importance and relevance of known non-modifiable and modifiable risk factors for Asian women with BC. Furthermore, it has highlighted the differences in epidemiology of BC in Asian nations when compared to other regions. Only a few studies directly compared BC risk between Asian and non-Asian populations; in the ones that did, there was a decreased risk of BC among women from Asia compared with women from other countries such as the US.