Introduction
Breast cancer is the most frequent cancer among female both in Western countries and in China [
1,
2]. An analysis from 40 years of cancer registry data showed that the incidence of breast cancer have a tremendous increase, while the mortality rate only slightly increase in urban Chinese population [
3]. It indicates that a rising number of women with breast cancer would live with the disease for longer time. In Hong Kong, breast cancer has also aroused heavy disease burden in local female [
4]. Breast Cancer Survivorship Care Guideline established by the American Society of Clinical Oncology (ASCO) has underscored that adoption of a healthy lifestyle as an essential element of survivorship care [
5]. For such, body weight management is an important component, and ASCO recommended that primary care clinicians should counsel breast cancer survivors to achieve and maintain a healthy body weight [
5].
Weight gain averaging 1.5 to 2.1 kg during 1–2 year post-diagnosis has been very frequent in Caucasian women [
6‐
9]. In Asian women, weight changes have also been observed after breast cancer diagnosis, but the results have been inconsistent, [
10‐
13] with most of those studies being cross-sectional and involving small patient number [
10,
12,
13]. The reasons for weight gain could be multifactorial. A series of factors have been suggested to be associated with weight gain, including chemotherapy, use of medication in association with chemotherapy such as dexamethasone, treatment-related amenorrhea, “stress eating” and reduced physical activity [
13‐
16]. As weight management is an essential element in the long-term management of breast cancer survivors, there is still a need to fully investigate the pattern of weight change and explore potential risk factors associated with very substantial weight gain among Asian women with breast cancer.
The sole use of body mass index (BMI) cannot fully reflect fat distribution over body compartments [
17]; therefore, waist-to-hip ratio (WHR), have been recommended for the measurement of abdominal obesity in recent decades [
18]. Freedman et al. have reported that patients with breast cancer who had adjuvant chemotherapy would experience unfavorable changes in body composition without a significant weight change [
19]. The pattern of WHR change after breast cancer diagnosis has not been fully illustrated. As previous study showed that Chinese and South Asian displayed a greater amount of visceral adipose tissue for a given waist circumference when compared to Europeans, the measurement of WHR in Asian may need special attention [
20].
Based on longitudinal follow-up of a cohort of Chinese women with breast cancer, the present study has the following two aims: 1) to characterize the patterns of weight change from diagnosis to immediately post-diagnosis and at 18-, 36-, 60-months follow-up, and to identify potential socio-demographic, clinical and lifestyle factors associated with very substantial weight gain; 2) to describe the patterns of WHR change from immediately post-diagnosis to 18-, 36- and 60-months follow-up, and to identify potential factors associated with very substantial WHR increase.
Patients and methods
Study cohort
The Hong Kong NTEC-KWC Breast Cancer Survival Study (HKNKBCSS) was a prospective breast cancer cohort study, initiated to investigate the associations between lifestyle factors with breast cancer recurrence and mortality [
21‐
25]. The inclusion criteria included patients of any age, had histologically confirmed breast cancer with American Joint Committee on Cancer (AJCC) stage 0-III diagnosed no more than 12 months before study entry, [
26] female gender, mentally stable, Chinese ethnicity, able to read Chinese, and did not have prior history of breast or other cancers. Between January 2011 and February 2014, 1462 eligible patients provided written informed consent and participated in the study. The study was approved by the Joint CUHK-NTEC Clinical Research Ethics Committee and the KWC Research Ethics Committee of the Chinese University of Hong Kong and the Hong Kong Hospital Authority.
Consented patients were interviewed at four time-points: baseline at study entry (described as T0; conducted within 12-months after breast cancer diagnosis), 18-months follow-up (T1; conducted between 12 and 24 months after diagnosis), 36-months follow-up (T2; which was conducted between 30 and 42 months after diagnosis) and 60-months follow-up (T3; 54–66 months after diagnosis). A telephone call would be made prior to the planned interview, which would coincide with their scheduled clinic follow-up.
As of December 2017, the 60-months follow-up interview had been completed. Among 1462 patients who completed assessment at T0, 1310, 1162 and 1171 participants completed interviews at T1, T2 and T3, respectively (follow-up rate: 89.6, 79.5 and 80.1%, respectively). The present study was based on the anthropometric measurements data assessed during interviews at all four time-points.
Data collection
During each follow-up interviews, patients were assessed with structured questionnaires conducted at baseline assessment, which collected socio-demographic characteristics (education level, marital status, working status and family income), reproductive history, menopausal status, active and passive smoking, alcohol use and prior medical history (self-reported comorbidities including but not limited to diabetes, hypertension, hyperlipidemia, chronic liver disease and chronic kidney disease). At T0 assessment, patients’ menopause status was classified as two groups: pre-menopausal and post-menopausal. Patients who had their last menstrual period within 1 year were regarded as pre-menopausal. Post-menopausal was defined as patients who had a cessation of menstruation for 12 months or longer. From T0 to T3 assessment, patients’ menopause status could be classified as three groups: pre-menopausal, peri-menopausal and post-menopausal. Peri-menopausal was defined as pre-menopausal patients at T0 who described a change in menopause status by T3.
Clinical information was retrieved by reviewing hospital medical records. These included patient’s age at breast cancer diagnosis, cancer characteristics [histology, AJCC stage, estrogen receptors (ER), progesterone receptors (PR) and human epidermal-growth-factor receptor 2 (HER2) status of the breast tumor] and treatment for breast cancer (type of breast surgery, details of adjuvant radiotherapy, chemotherapy and endocrine therapy).
Physical activity was measured by a validated modified Chinese Baecke questionnaire [
27]. The MET code of each sport was based on the values in the Ainsworth compendium of physical activity [
28]. According to the level of sports activity, patients were categorized into 3 groups as follow: never (0 MET-hours/week), rare/occasional (< 10 MET-hours/week) and frequent (≥10 MET-hours/week) physical activity. The cut-off point of 10 MET-hours/week was based on the recommendations for cancer survivors [
29,
30]. Habitual dietary intake was collected by a validated and interviewer-administered food frequency questionnaire (FFQ) [
31]. The average daily intake of nutrients, such as total energy, fat and other nutrients were calculated according to the Chinese Food Composition Table [
32]. Meeting dietary recommendation on vegetables and fruits intake was defined as eating at least five servings (at least 400 g) of nonstarchy vegetables and fruits every day according to the WCRF/AICR guidelines for cancer survivors [
29]. At T0 assessment, patients were asked to recall their habitual physical activity and dietary intake in the preceding 12 months before cancer diagnosis. At T1, T2 and T3 assessments, patients were asked to report these parameters over the previous 12 months.
Anthropometric measurements
Patients who have cancer recurrence or died during the first five years did not measure their body weight in the subsequent follow-ups, and those patients did not include into the present analysis. Anthropometric measurements including body weight, height, waist and hip circumference were performed based on standard protocol. Trained interviewers measured body weight, waist and hip circumference at T0, T1, T2 and T3 assessment, respectively, and the body height was only measured at T0 assessment. Body height was measured to the nearest 0.1 cm with the patients in bare feet, back against the wall, heels together and eyes looking straight ahead. Body weight was measured to the nearest 0.1 kg with the participant in light clothing and bare feet using a TANITA Body Fat Scale (Model BF-522, TANITA, Japan). Weight at diagnosis was collected from hospital medical records.
Body mass index (BMI) was calculated by weight (kg) divided by the square of height (m
2). According to BMI classification of the Asia-Pacific region, patients can be categorized into 5 groups as following: underweight < 18.5 kg/m
2, normal 18.5–22.9 kg/m
2, overweight 23–24.9 kg/m
2, obese ≥25 kg/m
2 [
33]. Compared to weight at breast cancer diagnosis, absolute weight change at T0, T1, T2 and T3 assessment was calculated (weight at T0, T1, T2 or T3 - weight at diagnosis), and then classified into five groups: substantial loss (> 5 kg), moderate loss (> 2 kg and ≤ 5 kg), stable change (within 2 kg), moderate gain (> 2 kg and ≤ 5 kg) and substantial gain (> 5 kg). The relative percent of weight change at T0, T1, T2 and T3 assessment was also calculated (absolute weight change at T0, T1 or T2/weight at diagnosis*100), and then classified those changes into six groups: substantial loss (> 5%), moderate loss (> 2% and ≤ 5%), stable change (within 2%), moderate gain (> 2 and ≤ 5%), substantial gain (5–10%) and very substantial gain (> 10%).
WHR was calculated as the ratio of waist circumference to hip circumference, which is regarded as an index of abdominal obesity [
34]. World Health Organization (WHO) expert consultation defines abdominal obesity as WHR above 0.85 for women [
35]; patients were grouped into two categories: without abdominal obesity < 0.85 and with abdominal obesity ≥0.85. As WHR at diagnosis were not available, the value obtained at T0 assessment was used as reference in the measurement of post-diagnosis WHR change. Percentage of changes between T0 assessment and T1, T2 and T3 assessment were calculated (absolute WHR change at T1, T2 or T3/WHR at T0*100), and then classified those changes into six groups: substantial decrease (> 5%), moderate decrease (> 2% and ≤ 5%), stable change (within 2%), moderate increase (> 2% and ≤ 5%), substantial increase (5–10%) and very substantial increase (> 10%).
Statistical analysis
Patients’ socio-demographic, clinical and lifestyle factors described as follows: continuous variables were expressed as means with standard deviation or median with range as appropriate, and categorical variables were summarized as patient number (n) and percentage (%). Univariate logistic regression was performed to identify any potential factors associated with very substantial weight gain from diagnosis to T3 assessment. The potential variables with P < 0.1 in univariate analysis were included into the multivariate logistic regression model. Similar analyses were performed to identify any potential factors associated with very substantial WHR increase from T0 to T3 assessment. All analyses were performed using SPSS 26.0; and P values < 0.05 at two-sided analysis were considered statistically significant.
Discussion
Based on a longitudinal breast cancer cohort, the present study prospectively measured anthropometric parameters during the first 5 years of survival. The results showed that weight gain was not common among Hong Kong women with breast cancer. Over the first 12-months and at 18-months post-diagnosis, the median weight change was − 0.5 kg and 0 kg, respectively. At 36-months to 60-months after diagnosis, it observed a modest weight gain with a medium value of 0.5 and 1 kg, respectively. At 60-months after diagnosis, only 14.3% of women had weight gain by > 5 kg; and the percentage of women who had weight gain by > 10% was 10.7%. Of note, nearly half of patients had abdominal obesity at study entry, and this figure gradually increased to nearly 70% at 60-months after diagnosis. Being older and having frequent sports participation were independent protective factors for very substantial weight gain in multivariate analysis. Additionally, aged 40–49 years at diagnosis, had ≥2 comorbidities and ER negative are independently associated with less likelihood of very substantial WHR increase.
The weight change pattern in the current study were inconsistent with findings from Western countries, which have generally reported weight gain following adjuvant treatment. Goodwin et al. reported the weight change from baseline to 1 year post-diagnosis among 535 women with newly diagnosed breast cancer in Canada, and showed that about 84% of women gained weight with a mean value of 1.6 kg [
6]. Another study which included 185 women in United States (US) with early stage breast cancer, showed that the mean weight gain was 1.5 kg, 2.7 kg and 2.8 kg at 1 year, 2 year and 3 year after diagnosis, respectively [
9]. Rock et al. examined the weight change from diagnosis to study enrollment (mean time since diagnosis: 26 months) among 1116 US breast cancer patients, showing that the mean weight gain was 2.7 kg and 60% of the participants reported weight gain [
37]. In contrast to data from Western women, weight gain after breast cancer diagnosis among Asian women has been relatively modest. The Shanghai Breast Cancer Survival Study (SBCSS) reported that the median weight change from diagnosis to 6, 18 and 36 months post-diagnosis were 1.0, 2.0 and 1.0 kg, respectively; about 26, 37 and 33% of gained weight by ≥5% at 6, 18 and 36 months post-diagnosis, respectively [
11]. It is noted that the magnitude of weight gain in the SBCSS was slightly greater than that obtained in the present study. Although the study design of the SBCSS was similar to the present cohort, the two studies enrolled patients diagnosed at different times; with the present study having enrolled patients who were diagnosed nearly 10 years later than the SBCSS (from 2011 to 2014). In a study of 260 Korean women with early stage breast cancer who received adjuvant treatment, the investigators reported that no weight gain was found after treatment, and the mean weight change was − 0.3 and − 0.4 kg at 1- and 2- year after treatment, respectively [
10]. Another cross-sectional study included 280 premenopausal women (median age at diagnosis was 41 year) with breast cancer after chemotherapy in Hong Kong and reported similar weight change pattern: the median weight gain from diagnosis to 5 years after diagnosis was 1.8 kg, with 63.2% of women gaining weight by > 2% [
13]. The varied results among studies in Asian women may be explained by difference in study design, variations in time interval for assessment since initial diagnosis, as well as diverse lifestyle habits.
Several previous studies have tried to investigate the associations between socio-demographic and clinical factors and weight change after diagnosis. In the Health, Eating, Activity and Lifestyle (HEAL) study, the results showed that postmenopausal women at diagnosis had greater weight gain than pre-menopausal women or women who had menopausal transition after diagnosis [
8]. However, in the SBCSS, more weight gain was observed among women who were premenopausal during the first 6 months [
11]. The present study supports the SBCSS findings that women who were pre-menopausal at study entry had higher risk of substantial weight gain than those who were postmenopausal in univariate analysis. Findings from studies in Western populations reported an association between chemotherapy and weight gain [
6,
7,
9,
15]. However, this finding was not confirmed in Asian population. For example, the Korean study and the SBCSS did not find a weight gain after adjuvant chemotherapy [
10,
11]. Similarly, the present study did not show chemotherapy to be associated with the risk of very substantial weight gain. On the other hand, the present study found that older age was associated with reduced risk of weight gain even in multivariate analysis, which was supported by two previous studies in Chinese women with breast cancer [
11,
13].
Several studies have reported a significant decrease in physical activity during and after treatment, and this lifestyle change may be another reason for post-diagnosis weight gain [
7,
38,
39]. Chen et al. explored the potential predictors of weight change in SBCSS and reported that higher exercise level was marginally related to weight loss [
16]. Similar relationship was observed in the current study, higher level of physical activity was statistically associated with lower risk of very substantial weight gain in both univariate and multivariate analysis. Those findings suggested that high level of physical activity might prevent post-diagnosis weight gain. The results from SBCSS suggested that higher energy intake was related to greater weight gain [
16]; while findings from the present study and Yaw et al’ s study showed that total energy intake was not associated with weight change [
12].
A number of reports have investigated the relationship between post-diagnosis weight gain and breast cancer prognosis, and strongly supported that post-diagnosis weight gain was related to higher risk of mortality among breast cancer survivors. In 2015, Playdon et al. systematically summarized these data, including 12 studies and a total of 23,832 breast cancer patients [
40]. The meta-analysis showed that patients who gained weight by ≥5% was associated with increased risk of overall mortality compared with patients who maintained their body weight (defined as weight change < ±5%) [
40]. However, it should be noted that evidence from Chinese women with breast cancer has been limited, with only one study showing that women who gained ≥5 kg had higher mortality than those who maintained their weight [
41].
WHR was regarded as an index for the measurement of central adiposity [
35]. In general population, abdominal adipose tissue (which is positively associated with waist circumference and waist-hip ratio) is associated with a range of metabolic abnormalities, including decreased glucose tolerance, reduced insulin sensitivity and adverse lipid profiles [
35]. The present study found that nearly half of women with breast cancer had central adiposity and this phenomenon is more severe as survival time increased. To our knowledge, this is the first study to describe the change pattern of WHR. Multivariate analysis showed that patients aged 40–49 years and had ≥2 comorbidities were less likely to had very substantial WHR increase when compared to those aged < 40 years and those had no comorbidity, respectively; suggesting that those patients may be more cared about abdominal obesity. Moreover, the present study suggested that patients with ER-negative disease was associated with less likelihood of very substantial WHR increase when compared to patients with ER positive; this may be related to use of adjuvant endocrine therapy, in particular, tamoxifen.
Of interests, a few studies have investigated whether central adiposity status after breast cancer diagnosis was associated with detrimental outcomes of breast cancer [
42‐
45]. For a given BMI, it has been widely reported that Asians tend to have a higher fat percentage and a higher proportion of abdominal adiposity than western populations, [
46,
47] it highlights the need to address this issue in specific ethnic groups.
This study was based on data from a longitudinal cohort study with a large sample size, with quantitatively compared the changes of body weight and WHR from immediately post-diagnosis to 5 years of survival. However, a few limitations should be noted. First, the change pattern was not compared with cancer-free women of similar age; it is, therefore, unclear whether these observed changes can be attributed to aging perse or to breast cancer and its treatment. Second, the baseline measurements of WHR were conducted within 1 year after breast cancer diagnosis; with no data captured right after diagnosis. However, the interval between diagnosis to study entry was relatively short (with median time of 3.2 months), significant WHR changes would not be expected. Third, the multivariate analysis for potential predictive factors could only be regarded as an exploratory analysis given that no clear hypothesis had been stated at priori. Fourth, as patients may experience weight loss after cancer recurrence or before death, and those patients were not included in the present analyses. As such, excluding these women who have cancer recurrence or died during the first 5 years of survival may lead to including more women with weight gain. Fifth, as the original study primarily aimed to investigate the associations between lifestyle factors with breast cancer recurrence and mortality, data on body fat was not collected during the assessments. As such, this report is unable to describe the patterns of body fat change. Finally, as the survival data in the present cohort was not mature enough, this study did not investigate the association between weight gain and breast cancer outcomes. Prospective follow-up is needed to further explore such association in Chinese women with breast cancer.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.