The Association Between Body Size and Breast Cancer in Han Women in Northern and Eastern China

In this case-control study of 2,800 Chinese Han women, BMI was associated with increased risk for breast cancer in all women. In premenopausal women, but not postmenopausal women, waist and hip circumferences also were associated with breast cancer risk.

Abstract

Introduction.

Although obesity has been reported worldwide as a risk factor for breast cancer, there are still some inconsistencies regarding the association between obesity and breast cancer. Body mass index (BMI) is used most to assess the extent of obesity; however, the association of other body size characteristics, such as waist and hip circumference, with susceptibility to breast cancer in Chinese Han women needs to be better assessed.

Patients and Methods.

Female Chinese Han patients (N = 2,800) were recruited from 21 hospitals in northern and eastern China from April 2012 to April 2013 for a case-control study. The significant differences of factors related to body size between the breast-cancer case and control groups were determined by Student’s t test and chi-square tests.

Results.

Premenopausal women with breast cancer had higher BMI and larger waist and hip circumferences (p = 2 × 10⁻⁴, <1 × 10⁻⁶, and 2 × 10⁻⁵, respectively). However, these body-size factors were not associated with postmenopausal breast cancer (p = .45, 0.32, and 0.12, respectively). BMI between 28 and 30 kg/m² or greater than 32 kg/m² was related to breast cancer incidence in the overall study population and in premenopausal women but not in the postmenopausal group.

Conclusion.

Obesity is significantly associated with breast cancer in Chinese Han premenopausal women but not in postmenopausal women. Thus, it is important to realize that weight control, as well as avoiding abdominal obesity, should be considered as one of the most effective methods of reducing breast cancer risk.

Implications for Practice:

To better understand the characteristics and risk factors for breast cancer in Han women in northern and eastern China, a case-control study of 2,800 Chinese Han women was conducted. Obesity was significantly associated with breast cancer in Chinese Han premenopausal women but not in postmenopausal women. Consequently, controlling body weight and avoiding abdominal obesity should be considered as one of the most effective methods of reducing breast cancer susceptibility. However, the diversity between this study’s finding among Chinese Han women and other data previously reported among European and American populations still needs further investigation.

Introduction

Breast cancer is one of the most common cancers diagnosed among women worldwide, and is the second leading cause of cancer-related deaths among women in United States and the sixth among Chinese women [1, 2]. In 2015, breast cancer alone was expected to account for 15% of all new cancers in Chinese women [3]. The standardized prevalence in northern and eastern China is still approximately 262.5 per 100,000 women, according to our previous study [4]. Unfortunately, the incidence of breast cancer in China is increasing twice as fast as that of the global rates, particularly in urban areas [5]. Thus, reducing the incidence of breast cancer has become an important global health issue. The exact causes of breast cancer are still not fully clarified, although many risk factors associated with breast cancer, such as family history [6], menarche [7], adiponectin levels [8], and lifestyle [9], have been identified. Discovering more risk factors and revealing a precision prediction model will benefit the effective prevention and intervention of breast cancer.

Obesity has been reported to be a risk factor for several kinds of cancers [10], especially breast cancer [11]. And it could also lead to worse prognosis [12]. Recently, we reported high body mass index (BMI) was associated with increased breast cancer risk in Chinese Han women [4] in the first large-scale and cross-province epidemiological investigation in China since the 1970s. Coincidentally, women aged 45 to 59 years in China have the highest increase rate in BMI [13] and also the highest rate of breast cancer [2, 14]. However, there are still some inconsistencies regarding the association between obesity and breast cancer [15–17], and it is necessary to study this relationship further.

To better understand the characteristics and risk factors for breast cancer in Chinese women, we conducted a large 1:1 case-control study of 2,800 women from 21 hospitals in northern and eastern China and investigated the associations of body weight, BMI, waist circumference, hip circumference, and waist-to-hip ratio (WHR) with breast cancer risk, along with other known risk factors. Analyses were further stratified by menopausal status, based on previous evidence of pre- and postmenopausal BMI diversity effects [16, 18].

Materials and Methods

Study Population

Female patients were recruited from 21 hospitals in seven provinces and two municipalities (Beijing and Tianjin) in northern and eastern China from April 2012 to April 2013. The inclusion criteria for the case and control groups were described in our previous study [8]. The inclusion criteria for breast cancer cases were as follows: (a) having newly diagnosed and histologically confirmed breast cancer, (b) being of the Han ethnic group, (c) being between 25 and 70 years old, and (d) being without neoplastic diseases at any other site or history of cancer. The criteria for the control group were as follows: having (a) negative physical examination results and (b) negative findings on breast ultrasound scans and mammographic screening; (c) being of matched age with the cases (within 3 years); (d) having been hospitalized or had a regular physical examination in the same hospital with a matched case in the same time period (within 2 months); (e) having no evidence of cancer or history of cancer; and (f) being of the Han ethnic group. All participants were enrolled after informed consent was obtained. Case patients and control subjects were matched based on age, hospital, and timing of examination in a 1:1 ratio. A total of 1,489 case-control sets met our inclusion criteria. After excluding subjects with inadequate information, especially lack of BMI data, 1,400 sets of cases and controls were finally enrolled in this study.

Ethics Statement

The study protocol and procedures were approved by the institutional review boards at the Second Hospital of Shandong University and the other hospital involved in this study. All participants provided their written informed consent.

Data Collection

Patient information was obtained primarily from in-person interviews based on a self-designed structured questionnaire, which also was described in our previous study [8]. The questionnaire was used to obtain the following information: (a) basic demographic characteristics, such as age, birthplace, height, weight, income, and education level; (b) data about body size, including BMI, waist circumference, hip circumference, and WHR; (c) female physiological and reproductive factors, including age at menarche, age at menopause, menopause status, number of births, and breast feeding; (d) family history of breast cancer, including first- and second-degree relatives; (e) lifestyle habits, such as smoking (including passive smoking), alcohol intake, dietary habits, and physical exercise; (f) information from medical records on clinical breast examination and history of breast disease and other diseases; (g) awareness of breast cancer-related knowledge; and (h) medical and chemical exposure history. The clinical and histological diagnoses of breast cancer patients were also collected from the medical record.

Statistical Analysis

To better analyze the risk associated with BMI, we used the Chinese BMI cutoff points of more than 24.0 kg/m² for overweight and more than 28.0 kg/m² for obesity [13] and established 10 levels of 2.0 kg/m² each between 16.0 and 32.0 kg/m². All data were analyzed using SPSS software (version 16.0; SPSS Inc., Chicago, IL, http://www.ibm.com) in this study. Descriptive characteristics of the group variables are expressed as mean ± SD. The significant differences of continuous variables between the case and control groups were determined by Student’s t tests and categorical variables by chi-square tests. Odds ratio (OR) with 95% confidence interval (CI) was used to assess the influence of different variables on breast cancer risk. Statistical tests were two sided, and p values less than .05 were considered significant.

Results

Overall Descriptive Characteristics

A total of 2,800 Chinese Han women were enrolled in our study; their characteristics are shown in Table 1. Overall, the mean ages ± SD at enrollment for the study population were 47.64 ± 8.70 years in the case group and 47.01 ± 8.72 years in the control group, showing no significant difference (p = .06). Although the populations in the two groups were similar in height and weight (p = .07 and 0.12, respectively), the case group had higher BMI (24.33 ± 3.44 kg/m² vs. 24.04 ± 3.11 kg/m² in the control group; p = .02) and higher proportion of obesity (BMI > 28.0 kg/m²; p = 1 × 10⁻³). Women with breast cancer also exhibited larger waist and hip circumferences (p = 2 × 10⁻⁶ and .04, respectively). The WHR was also higher in the case group (p = .02).

Table 1.

Characteristics of individuals included in case and control groups

Next, we examined the effect of female physiological and reproductive factors and found the case group was of older age at menarche (p = .04), had a higher proportion of postmenopausal women (p = .02), and higher number of births (p < 1 × 10⁻⁶). However, breast feeding and age at menopause were not significantly different between the two groups (p = .62 and .80, respectively). We also evaluated alcohol intake and diabetes mellitus (DM) incidence and detected no distinguishable difference in these between the two groups (p = .29 and .43, respectively). The data also suggest that the positive second-degree family history and positive overall (including first and second degree) family history were related to the risk for breast cancer (p = .01 and 2 × 10⁻⁴, respectively) (Table 1).

Stratified Descriptive Analyses by Menopause Status

Subgroup descriptive analyses stratified by menopause status were performed to explore the influences of body size and other risk factors for breast cancer in pre- and postmenopause (Table 2). There were 465 women in case group and 402 in the control group who were determined to be postmenopausal. Compared with the control group, the postmenopausal women with breast cancer were shorter and had higher WHR (p = 4 × 10⁻³ for both). BMI and waist and hip circumferences did not differ between the postmenopausal case and control groups. The 894 premenopausal women in the case group were heavier and had higher BMI, higher rate of obesity, and greater waist and hip circumferences (p = 2 × 10⁻⁴, 2 × 10⁻⁴, 2 × 10⁻⁴, <1 × 10⁻⁶, and 2 × 10⁻⁵, respectively) when compared with the paired control group. However, DM was also confirmed to be not associated with breast cancer in either pre- or postmenopausal women.

Table 2.

Characteristics of individuals included in this study by menopause status

Comparative Analyses of Risk Factors in Women With Breast Cancer Relative to Family History

In previous work, we have shown that family history was positively associated with the risk for breast cancer overall and specifically among premenopausal women [8]. To analyze the interrelationship between family history and body size, we further divided the breast cancer patients into two groups according their family history of breast cancer. The descriptive data of the two groups are shown in Table 3. There were 98 subjects with a positive first- or second-degree family history of breast cancer. Importantly, significant differences were observed between case patients with or without positive family history in factors related to body size. Among the body size indexes, patients with positive family history were of lower weight and had a lower BMI, with a higher proportion of patients with BMI less than 24 kg/m² (p = 4 × 10⁻³, 3 × 10⁻³, and 9 × 10⁻³, respectively). Nevertheless, waist and hip circumferences and WHR were similar between the compared groups.

Table 3.

Characteristics of individuals in the case group with or without family history of breast cancer

The Precise Association Between BMI and Breast Cancer by Menopausal Status

To elucidate the impact of BMI on breast cancer risk in the Han population in northern and eastern China, we investigated the odds ratios between case and control groups in ascending BMI ranges and by menopausal status. ORs with 95% CIs are shown in Figure 1; detailed data are listed in supplemental online Table 1. Odds ratios among the overall case-control pairs demonstrate a tortuous increase in BMI between the odds ratio trend lines of premenopausal and postmenopausal breast cancer. BMI between 28 and 30 kg/m² was related to an increase in breast cancer risk among both overall and premenopausal individuals (p = .01 and .03, respectively), whereas a significant association between BMI greater than 32 kg/m² and breast cancer only existed between the premenopausal case and control groups (p = .007). However, there were no other notable differences in other BMI ranges; that is, there was a similar distribution of postmenopausal case patients and control patients among different BMI levels.

Figure 1.

Association between body mass index (BMI) and odds ratio of breast cancer between case and control groups by menopause status. Odds ratios of different BMI ranges between case and control groups are shown as blue dots, and show a tortuous increase in BMI interposed between the trend lines of premenopausal (red triangles) and postmenopausal breast cancer odds ratios (purple squares). I- or T-shaped bars represent 95% confidence intervals. ∗, BMI ranges significantly associated with increasing breast cancer risk.

Discussion

Here, we have described a case-control study of 2,800 Chinese Han women. Factors related to body size, and reproductive and family histories were confirmed to be associated with an increased risk for breast cancer, resulting in a phased association. Body size characteristics related to breast cancer risk factors were more consistent between premenopausal women and the total study population; that is, women with breast cancer, especially those who were premenopausal, manifested higher BMI and greater waist and hip circumferences. This could be because most women in our study population were premenopausal. Further analysis of the categories of BMI revealed the higher BMI of the case group could be attributed to the higher proportion of obesity. However, BMI, and waist and hip circumferences were not associated with postmenopausal breast cancer, although WHR was significantly higher in the postmenopausal cases, as it was in the overall population of breast cancer patients in this study.

In our analysis of the association between ascending BMI ranges and breast cancer risk by menopausal status, BMI between 28 and 30 kg/m² was related to increased breast cancer risk among the overall study population and premenopausal individuals. And BMI greater than 32 kg/m² was related to breast cancer only in premenopausal women. Thus, the associations between BMI and breast cancer were more marked in premenopausal women and weaker in postmenopausal women.

Obesity is one of major public health burdens worldwide [19]. In China, a developing country with a large population, the obese population increased from 18 million to nearly 100 million during the period from 2005 to 2011, and the number of overweight and obese Chinese accounted for greater than 38% of the total Chinese population in 2011 [20]. During the past decades, there has been an increasing interest in the relationship between body size (including BMI, body weight, waist and hip circumferences, and WHR) and risk for breast cancer. Similar to our study, Wang et al. [14] reported that breast cancer patients in eastern China showed significant differences in waist circumference and BMI but not in WHR or hip circumference. However, they also found a significant difference in BMI between the postmenopausal patients and control subjects, and DM was significantly associated with breast cancer.

Both the World Cancer Research Fund and the Women's Health Initiative have definitively concluded that obesity is a risk factor for breast cancer, especially postmenopausal breast cancer [21, 22]. Contrary to our result, BMI was reported to be inversely associated with premenopausal and positively associated with postmenopausal breast cancer in some studies [18, 23–26]. In a cohort study of 5.24 million adults in the U.K. [10], BMI was positively associated with both pre- and postmenopausal breast cancers at BMI less than 22 kg/m², whereas above this value, risk for premenopausal breast cancer was reduced markedly with increasing BMI. However, in our study, BMI was significantly associated with increased breast cancer risk in overall the overall study population and in premenopausal women but not in postmenopausal women. This agrees with results of previous studies in Asian women. Yu et al. [4] found that a high BMI had a statistically significant correlation with breast cancer, as identified by multivariate logistic regression analysis. This result was also consistent with other previous studies [27, 28]. Renehan et al. [16] recorded stronger associations between increased BMI and premenopausal (p = .009) and postmenopausal (p = .06) breast cancers in Asia-Pacific populations but not in other races and regions. Further meta-analysis stratified this association by ethnicity and found a significant positive association between BMI and the risk for premenopausal breast cancer among Asian women [29]. In addition, this inverse association was also observed significantly among black people and white people.

Adipose tissue is now considered as essentially a huge gland, sending out a constant stream of biological information and instructions that coordinate control processes like cell growth, metabolism, and reproductive cycles [30]. Obesity affects many different aspects of physiology, including hormones, growth signals, and inflammation. In obesity, levels of estrogen, insulin, and inflammation knit together and lead to disastrous consequences. First, obesity leads to higher estrogen levels in women [31]. There is evidence that breast cancer is linked to increased estrogen levels caused by excess body fat, even in men [32]. Thus, losing weight is the least expensive kind of antiestrogen treatment of breast cancer, acting like tamoxifen and aromatase inhibitors. Second, a large number of studies show that insulin and insulin-related growth factors could induce cells to divide more quickly and make cancer cells harder to kill [33]. Third, obesity also increases cancer risk by inducing chronic inflammation, which is a significant risk factor for cancer and other human diseases [34], especially for breast cancer [35].

We conclude that cancer risk is influenced by intrinsic and extrinsic factors. Advances in genetic studies have identified many genes and variants associated with susceptibility to breast cancer [36, 37] and, more interestingly, BMI exhibits interactions with some variants [38]. However, Chen et al. [39] tested 31 common variants for DM and obesity in a case-control study of 1,915 breast cancer patients and 2,884 control subjects to investigate a genetic basis, but found no significant association. On the other hand, a substantial proportion of breast cancers are known to be associated with environmental risk factors [40], which might explain the higher incidence of breast cancer seen in immigrants moving from countries with lower breast cancer incidence to countries with higher rates [41]. This may also explain the inconsistent incidence reported between the Han population in northern and eastern China and other populations [16, 18, 21–26, 29].

Although some notable discoveries have been revealed, there are also several limitations in our study. First, this study shows only a glance at the association between breast cancer and risk factors, including body size, and reproductive and family histories, in the Han population in northern and eastern China. It remains to be demonstrated in large-scale prospective studies if there is a more authentic relationship. Second, there are other important factors that were not included in our study, such as in vitro fertilization treatment [42], breast density [43], and, most importantly, genetic background [44]. Third, the disease history was collected by subjects’ self-reports, thus some women with DM may have been missed. Fourth, although most of the individuals in the control group were healthy women who had a regular breast cancer screening in the same hospital, there were a few individuals who were hospitalized for some benign diseases, which might have led to lower body weight and introduced potential bias. Thus, these issues need to be targeted in the future studies.

Conclusion

Our work presents more comprehensive evidence that risk factors related to body size are associated with breast cancer in the Han population in northern and eastern China, especially in premenopausal women. However, there are some inconsistencies between our finding in this population and findings reported from other populations, and these might have resulted from the substantial environmental and genetic differences between ethnicity and geography. Thus, further investigation is needed to give us some idea about how obesity contributes to breast cancer risk. Moreover, it is important to realize that weight control and avoiding abdominal obesity should be considered as among the most effective methods to reduce breast cancer susceptibility and as the least expensive kind of antiestrogen prevention for breast cancer.

See http://www.TheOncologist.com for supplemental material available online.

This article is available for continuing medical education credit at CME.TheOncologist.com.

Author Contributions

Conception/Design: Xin Wang, Liang Li, Jidong Gao, Jiaqi Liu, Mingming Guo, Liyuan Liu, Zhigang Yu, Xiang Wang

Provision of study material or patients: Xin Wang, Liang Li, Jidong Gao, Jiaqi Liu, Mingming Guo, Liyuan Liu, Wenyan Wang, Jie Wang, Zeyu Xing, Zhigang Yu, Xiang Wang

Collection and/or assembly of data: Xin Wang, Liang Li, Jidong Gao, Jiaqi Liu, Mingming Guo, Liyuan Liu, Wenyan Wang, Jie Wang, Zeyu Xing, Zhigang Yu, Xiang Wang

Data analysis and interpretation: Xin Wang, Liang Li, Jidong Gao, Jiaqi Liu, Mingming Guo, Liyuan Liu, Wenyan Wang, Jie Wang, Zhigang Yu, Xiang Wang

Manuscript writing: Xin Wang, Liang Li, Jidong Gao, Jiaqi Liu, Mingming Guo, Zhigang Yu, Xiang Wang

Final approval of manuscript: Xin Wang, Liang Li, Jidong Gao, Jiaqi Liu, Mingming Guo, Liyuan Liu, Wenyan Wang, Jie Wang, Zeyu Xing, Zhigang Yu, Xiang Wang

Disclosures

The authors indicated no financial relationships.