Influence of Metabolic Syndrome on Risk of Breast Cancer: A Study Analyzing Nationwide Data from Korean National Health Insurance Service

Metabolic syndrome is a constellation of interrelated risk factors of metabolic origin that appear to directly promote the development of cardiovascular disease and type 2 diabetes mellitus (1). The name of metabolic syndrome (2) has also been described in various terms such as Reaven syndrome (3), syndrome X (3), deadly quartet (4), insulin resistance syndrome (5), hypertriglyceridemic waist (6), and dysmetabolic syndrome. The most widely proposed metabolic risk factors are dyslipidemia, elevated blood pressure, elevated plasma glucose, and abdominal obesity. Criteria for clinical diagnosis of metabolic syndrome have been proposed by several expert groups. The World Health Organization (WHO) proposed the criteria of metabolic syndrome in 1998 (7). Since then, the European Group for Study of Insulin Resistance (EGIR; ref. 8), National Cholesterol Education Program Adult Treatment Panel III (NCEP-ATP III; ref. 9), American Association of Clinical Endocrinologists (AACE; ref. 10), International Diabetes Foundation (IDF; ref. 11), and American Heart Association and the National Heart, Lung, and Blood Institute (AHA/NHLBI) have proposed their own criteria (1).

Metabolic syndrome has been reported to be associated with risks of various human cancers. A previous meta-analysis has shown that the presence of metabolic syndrome is associated with increased risk of liver, colorectal, and bladder cancers in men (12). It also found that increased risk of endometrial, pancreatic, colorectal, and postmenopausal breast cancer in women was associated with metabolic syndrome. Regarding breast cancer, previous studies have reported the association between metabolic syndrome and breast cancer, although results are inconsistent. Some studies have reported a positive association between metabolic syndrome and breast cancer (13–18), while other studies have reported no association (19–22). Furthermore, some studies have reported a negative association between them in premenopausal women (20). Currently, the association between metabolic syndrome and breast cancer remains largely unknown. Thus, more studies are needed to clarify this association.

The Republic of Korea (South) has a National Health Insurance system which is compulsory and required by the Korean law. The National Health Insurance Service (NHIS) is responsible for the management of National Health Insurance system. It covers all residents who are enrolled in the National Health Insurance system of Republic of Korea (South). The NHIS provides data of routine health checkups including information of metabolic syndrome and breast cancer. In this study, we investigated the association between metabolic syndrome and breast cancer risk in women using large nationwide data from the Korean NHIS.

This study is a retrospective cohort study utilizing data of routine health checkups provided by the NHIS. Target subjects of this study were women who received routine health checkups between 2009 and 2014 at the age of older than or equal to 20 years. The total number of subjects who had received routine health checkups between January 2009 and December 2014 was 27,155,170. Of these, subjects who met the exclusion criteria were sequentially excluded as follows: male subjects (n = 13,677,080), subjects younger than 20 years old (n = 32,339), subjects with incomplete information of metabolic syndrome (n = 28,723), and subjects who had been diagnosed with primary breast cancer at baseline, time of this study (n = 39,679). Primary cancers other than breast cancer were not considered at all. The final number of subjects included in this study was 13,377,349. End of follow-up of this study for all final subjects was December 31, 2015.

Different definitions of metabolic syndrome have been proposed by WHO, EGIR, NCEP-ATP III, AACE, IDF, and AHA/NHLBI. Of these, this study adopted the AHA/NHLBI criteria (1). Metabolic syndrome was defined as having any three of the five metabolic syndrome components: elevated waist circumference (≥80 cm), elevated blood pressure (≥130 mmHg systolic blood pressure or ≥85 mmHg diastolic blood pressure or on antihypertensive drug), elevated fasting glucose (≥100 mg/dL or on drug treatment for elevated glucose), elevated triglyceride (≥150 mg/dL or on drug treatment for elevated triglyceride), and reduced high-density lipoprotein cholesterol (<50 mg/dL or on drug treatment for reduced high-density lipoprotein cholesterol). Although a subject could have repetitive data of metabolic syndrome during the enrollment period, we only utilized data of metabolic syndrome from the first routine health checkup as baseline information. The other information of metabolic syndrome except baseline information was not utilized. Age was defined as the baseline age when the subject was enrolled in this study. Body mass index (BMI) was defined as the ratio of body weight in kilograms to height in square meters. Smoking was classified into no smoker, ex-smoker, and current smoker. Alcohol drinking was classified into no drinker, moderate drinker (ethanol 0–30 g/day), and heavy drinker (ethanol >30 g/day). Exercise was classified into no (<3 days of vigorous intensity and <5 days of moderate intensity) and yes (≥3 days of vigorous intensity or ≥5 days of moderate intensity). Income was classified into four quartiles with Q1 as the lowest and Q4 as the highest. Breast cancer type was classified into two groups including invasive breast cancer and in situ breast cancer by pathologic diagnosis. Information of breast cancer type was acquired from claim data of NHIS.

Pearson χ² test was used to determine statistical differences in baseline characteristics between groups. Cox proportional hazards model was used to calculate HR and 95% confidence interval (CI) of breast cancer risk. All adjusted HRs were obtained by Cox proportional hazards model after adjusting for age, smoking, alcohol drinking, exercise, and income. In this study, we did not consider any competing risks or competing events. As data of metabolic syndrome was retrieved from the first routine health checkup without utilizing the other data of routine health checkups, time-dependent Cox regression was not performed. To assess differential effects of metabolic syndrome on the breast cancer risk according to the subgroup, Cox proportional hazards regression was conducted using factors such as metabolic syndrome, each subgroup, and interaction term between metabolic syndrome and each subgroup with adjustment for covariates. P_interaction was calculated to interpret the effect of metabolic syndrome on the risk of breast cancer across subgroups, and it represents the overall significance for an interaction effect of metabolic syndrome and each subgroup. All statistical analyses were carried out using IBM SPSS Statistics, version 20.0 (IBM Corp.) and R software version 3.6.0 (R Foundation for Statistical Computing). All tests were two-sided. Statistical significance was considered when P value was less than 0.05.

The total number of subjects was 13,377,349 (64,535,186 person-year). Numbers of subjects with and without metabolic syndrome were 3,578,546 (26.8%) and 9,798,803 (73.3%), respectively (Table 1). Women with metabolic syndrome showed higher proportions of all five metabolic syndrome components. Women with metabolic syndrome also showed higher proportions of age >50 years and BMI ≥25 kg/m². Women without metabolic syndrome showed higher proportions of smoking, alcohol drinking, exercise, and low income. Total numbers of subjects diagnosed as invasive breast cancer and in situ breast cancer, regardless of metabolic syndrome, were 79,447 and 8,300, respectively. Baseline characteristics of study subjects according to metabolic syndrome and breast cancer type are summarized in Supplementary Table S1.

The presence of metabolic syndrome decreased the risk of all breast cancer types (HR, 0.954; 95% CI, 0.939–0.970; Table 2). In the subgroup with age ≤50 years, metabolic syndrome decreased the risk of all breast cancer types, with similar findings for all subject groups (HR, 0.915; 95% CI, 0.892–0.939). In the subgroup with age >50 years, metabolic syndrome increased the risk of all breast cancer types (HR, 1.146; 95% CI, 1.123–1.170). Each group with invasive breast cancer or in situ breast cancer showed similar findings to groups with all breast cancer types. Details for HR of breast cancer risk in each subgroup are described in Table 2.

For the subgroup with age ≤50 years, metabolic syndrome decreased risks of breast cancers in most subgroups according to BMI, smoking, alcohol drinking, exercise, and income (Table 3). On the contrary, metabolic syndrome increased risks of breast cancers in most subgroups for the group with age >50 years. For the all subjects group, although metabolic syndrome decreased risks of breast cancers in the age subgroup with age ≤50 years, it increased risks of breast cancers in the age subgroup with age >50 years, regardless of breast cancer types. HRs of each age subgroup with 5-year interval are described in Table 4. In age groups of more than 55 years, metabolic syndrome increased risks of breast cancers. Furthermore, HRs increased as age increased until 80 years. HRs of invasive breast cancer showed similar findings with those of all breast cancer types. Subgroups with age less than 55 years showed weak or no significance regarding HRs for each type of breast cancer.

In the subgroup with age >50 years, HRs for all breast cancer types increased as the number of metabolic syndrome components increased except for the subgroup with only one metabolic syndrome component (Table 5). On the contrary, HRs for all breast cancer types decreased as the number of metabolic syndrome components increased in the subgroup with age ≤50 years, although the HR of the subgroup with five metabolic syndrome components was not lower than that of the subgroup with four components. As a whole, HRs for all breast cancer types lost their trends in all subject groups. However, subgroups with any of metabolic syndrome components showed increased risks of breast cancers except for the subgroup with four components. Invasive breast cancers showed similar patterns with all breast cancer types. HRs according to metabolic syndrome components in each breast cancer type are described in Supplementary Table S2. The presence of any metabolic syndrome component increased the risks of breast cancers in the subgroup with age >50 years in both all types of breast cancers and invasive breast cancer type. On the contrary, any metabolic syndrome component decreased the risks of breast cancers in the subgroup with age ≤50 years with several exceptions. The influence of metabolic syndrome components weakened in the all subjects group. HRs of breast cancer risk according to metabolic syndrome components in each age group and breast cancer type are further described in Supplementary Table S3. The influence pattern of metabolic syndrome components distinctly changed with the cutoff value of age at 55 years. In the age subgroups with age of more than 60 years, positive correlation was prominent between HRs and total number of metabolic syndrome components (Fig. 1A). More prominent positive correlations were observed as the age of subgroups became older. BMI was strongly associated with breast cancer risk (Fig. 1B). HRs increased as BMI increased. These findings were more prominent as the age of subgroups became older.

This study utilized nationwide data of 13,377,349 female subjects (64,535,186 person-year) older than 19 years from the Korean NHIS and investigated the influence of metabolic syndrome on the risk of breast cancer. This study showed that the presence of metabolic syndrome increased the risks of both invasive breast cancer and in situ breast cancer in the subgroup with age >50 years, but decreased the risks of both invasive and in situ breast cancers in the subgroup with age ≤50 years. As a whole, metabolic syndrome decreased the risks of both breast cancer types in all subjects.

The prevalence of metabolic syndrome has been reported to have broad ranges according to definitions of metabolic syndrome and study populations (23–25). Rochlani and colleagues (23) have performed a literature review and reported that prevalence rates of metabolic syndrome in women in the United States and Republic of Korea (South) are 21.8%–37.4% and 21.3% (NCEP-ATP III), respectively. O'Neill and O'Driscoll (24) have reported that prevalence rates of metabolic syndromes in females in the United States and Republic of Korea (South) are 38.1% (IDF criteria) and 9.1% (NCEP-ATP III), respectively. Agnoli and colleagues (13) have used the same definition as in our study (AHA/NHLBI) and reported a prevalence of 29.8% in Italian postmenopausal women. In our study, the prevalence rate of metabolic syndrome in Korean women older than 19 years was 26.8%.

For postmenopausal women, previous studies have reported the influence of metabolic syndrome on the risk of breast cancer, although results are inconsistent. Some studies have insisted no association (19–22), while others have reported a positive association (13–18). Kabat and colleagues (19) have analyzed data of 4,888 postmenopausal women [6% of subjects in the Women's Health Initiative (WHI) clinical trial and 1% of women in the observational study] during 1993–1998. They reported that the presence of the metabolic syndrome had no significant association with altered risk of postmenopausal breast cancer [HR (95% CI), 1.12 (0.78–1.62) for all breast cancers and HR (95% CI), 1.19 (0.79–1.79) for invasive breast cancer]. Bjørge and colleagues (20) have analyzed data of 287,320 women enrolled in the Me-Can study during 1974–2005. They also reported no significant association between metabolic syndrome and the risk of breast cancer in postmenopausal women [RR (95% CI), 0.95 (0.87–1.01) for age 50–59 years and RR (95% CI), 1.04 (0.97–1.12) for age ≥60 years]. Agnoli and colleagues (13) have analyzed data of 3,966 postmenopausal women enrolled in the ORDET study during 1987–1992. They reported that the presence of metabolic syndrome was significantly associated with increased postmenopausal breast cancer risk (rate ratio, 1.58; 95% CI, 1.07–2.33). Rosato and colleagues (15) have analyzed data of 3,869 postmenopausal women with breast cancer and 4,082 postmenopausal controls during 1983–2007. They reported that the presence of metabolic syndrome significantly increased the risk of postmenopausal breast cancer (OR, 1.75; 95% CI, 1.37–2.22). Esposito and colleagues (18) have analyzed results of meta-analysis using nine articles with 6,417 postmenopausal breast cancer cases and 371,545 control subjects. They also reported that the presence of metabolic syndrome increased 52% of breast cancer risk in postmenopausal women (risk ratio, 1.52; 95% CI, 1.20–1.93). This study showed that the presence of metabolic syndrome increased the risk of all breast cancer types by 14.6% in postmenopausal women. Age subgroup analysis showed that this finding was prominent in subgroups with age more than 55 years. The effect of metabolic syndrome on the risk of postmenopausal breast cancer increased as the age of subgroups became older.

For premenopausal women, less evidence for the influence of metabolic syndrome on the risk of breast cancer has been reported compared with that for postmenopausal women. The Me-Can study reported that metabolic syndrome reduced the risk of premenopausal breast cancer (RR, 0.83; 95% CI, 0.76–0.90; ref. 20). A previous study reported no association between metabolic syndrome and premenopausal breast cancer (22). This study showed similar results to the Me-Can study, revealing that metabolic syndrome decreased the risk of premenopausal breast cancer by 8.5%. For all subjects, including both premenopausal and postmenopausal women, this study showed that the presence of metabolic syndrome decreased the risk of breast cancer. The effect of metabolic syndrome was less prominent compared with that of each subgroup possibly due to the summation effect of results in premenopausal and postmenopausal women. A previous study has performed a meta-analysis using nine independent cohorts with 97,277 women aged 18 years and older (26). It reported that there was a modest positive association between metabolic syndrome and breast cancer risk (RR, 1.47; 95% CI, 1.15–1.87).

Previous studies have reported the association between each component of metabolic syndrome and the risk breast cancer. Results were highly inconsistent across studies. The WHI clinical trial reported that only elevated diastolic blood pressure showed a borderline positive association with the increased risk of breast cancer in postmenopausal women (19). The Me-Can study reported that high BMI, blood pressure, cholesterol, and triglyceride were associated with decreased risk of breast cancer in age <50 years (20). Only high BMI and glucose were associated with increased risk of breast cancer in age ≥60 years (20). The ORDET study reported that the increased risk of breast cancer was only associated with high triglyceride and low high-density lipoprotein cholesterol in postmenopausal women (13). Rosato and colleagues (15) reported that increased risk of breast cancer was significantly associated with diabetes, hypertension, BMI, and waist circumference, but not with hyperlipidemia. Esposito and colleagues (18) reported that high glucose, high blood pressure, and low high-density lipoprotein cholesterol were associated with high risk of breast cancer in postmenopausal women. This study showed that every metabolic syndrome component increased the risks of breast cancers in postmenopausal women, but decreased the risks of breast cancers in premenopausal women. Waist circumference in postmenopausal women and triglyceride in premenopausal women were the strongest risk factors.

The Me-Can study reported the influence of metabolic syndrome on breast cancer according to BMI. It reported that postmenopausal women showed no association between metabolic syndrome and risk of breast cancer in low or high BMI subgroup (20). In premenopausal women, the presence of metabolic syndrome decreased the risk of breast cancer in only high BMI subgroups (RR, 0.67; 95% CI, 0.57–0.78). However, there was no association in low BMI subgroups (20). This study showed different results from those of the Me-Can study. The presence of metabolic syndrome increased the risks of breast cancers regardless of BMI status in postmenopausal women. In premenopausal women, the presence of metabolic syndrome decreased the risk of breast cancer in only low BMI subgroups. Further studies are needed to clarify the relationship between metabolic syndrome and BMI in breast cancers.

As most previous studies have reported the influence of metabolic syndrome on invasive breast cancer, little is unveiled for in situ breast cancer. This study showed that the presence of metabolic syndrome increased the risk of in situ breast cancer in postmenopausal women, but decreased the risk in premenopausal women. As a whole, metabolic syndrome increased the risk of in situ breast cancer in all subjects. These findings were largely the same as those of invasive breast cancer. The risk reduction effect of metabolic syndrome was more prominent for in situ breast cancer compared with that for invasive breast cancer in premenopausal and all subject groups. All metabolic syndrome components decreased the risk of in situ breast cancer in premenopausal women, but only triglyceride and high-density lipoprotein significantly increased the risk of in situ breast cancer in postmenopausal women.

Underlying mechanisms that link metabolic syndrome and breast cancer risk have been under active investigation, but still are not sufficiently revealed (12, 27–29). Metabolic syndrome could be a surrogate marker for cancer risk factors such as obesity, unhealthy diet, decreased physical activity, aging process, and so on (12, 30). Obesity is supposed to lead a chronic subclinical inflammatory state inducing proinflammatory cytokine production and infiltration of immune cells, and consequently to promote a protumorigenic environment (28, 31, 32). Adipose tissue could play a key role in the pathophysiology of metabolic syndrome as an active endocrine organ to secrete various adipocytokines, which mediate systemic metabolism (28). In breast cancer, obesity plays different roles in cancer risk according to menopausal status and estrogen receptor status. Obesity is associated with decreased risk of estrogen receptor–positive breast cancer in premenopausal women, but it is closely related with increased risk of estrogen receptor–positive breast cancer in postmenopausal women (32). Metabolic syndrome could be linked with other cancer risk factors such as unhealthy diet (33), decreased physical activity (32, 34), and aging process (30). Insulin and insulin-like growth factor (IGF) system are crucial factors in pathophysiology of metabolic syndrome that are mainly characterized as insulin resistance and hyperinsulinemia (27, 28). The insulin receptor and the IGF-1 receptor are overstimulated in metabolic syndrome. Dysregulated insulin and IGF system could activate growth hormones and associated signal pathways to promote apoptosis reduction, cellular proliferation, and cell survival, which increase the risks of cancer development and progression. Previous investigations proposed that each component of the metabolic syndrome is connected with systemic alterations, such as insulin resistance, proinflammatory cytokine production, oxidative stress, and angiogenesis, which potentiate the risk of cancer (16, 28, 35, 36). Common soil hypothesis is proposed to explain the association between metabolic syndrome and cancer risk, as both of them share many common risk factors such as age, genetic factors, obesity, physical inactivity, unhealthy diet, and so on (30, 33).

This study analyzed nationwide data of 13,377,349 women (64,535,186 person-year). It had the largest number of subjects regarding studies on metabolic syndrome. Nonetheless, this study has several limitations. First, this study was a retrospective cohort study. Prospective cohort study could provide more solid evidences on the association between metabolic syndrome and the risk of breast cancers. Second, this study did not analyze the association between the duration of metabolic syndrome and the risk of breast cancers. This study did not deal with the association between metabolic syndrome and breast cancer–related mortality either. Finally, as this study utilized only baseline data of metabolic syndrome, time-dependent exposures during the follow-up period were not analyzed. Change of metabolic syndrome status or new development of metabolic syndrome components over time might influence the breast cancer risk, and it is worth to be investigated in the subsequent studies. Further studies are needed to reveal more evidences regarding the influence of metabolic syndrome on breast cancer.

In conclusion, the presence of metabolic syndrome increased the risks of breast cancers in postmenopausal women, but decreased the risks in premenopausal women. The effect of metabolic syndrome on the risks of breast cancers was more prominent as age of subgroups became older, especially in postmenopausal women. The effect also became stronger as the number of metabolic syndrome components increased. Each component of metabolic syndrome and metabolic syndrome as a whole played similar roles on the risks of breast cancers. Further studies are needed to validate the relationship between metabolic syndrome and breast cancer risk. Clinicians need to assess breast cancer risks for women with metabolic syndrome.

No potential conflicts of interest were disclosed.

K.-T. Hwang: Conceptualization, formal analysis, supervision, validation, investigation, visualization, methodology, writing–original draft, project administration, writing–review and editing. K.-D. Han: Data curation, formal analysis, investigation, methodology, writing–review and editing. S. Oh: Formal analysis, investigation, methodology, writing–review and editing. B.K. Koo: Conceptualization, investigation, methodology, writing–review and editing. S.K. Lee: Investigation, writing–review and editing. J. Kim: Formal analysis, investigation, writing–review and editing. H.J. Seo: Investigation, writing–review and editing. J. Jung: Investigation, writing–review and editing. B.H. Kim: Investigation, methodology, writing–review and editing. H. Hur: Resources, formal analysis, investigation, writing–review and editing.

This study was supported by the Korean Breast Cancer Society and the National Health Insurance Corporation. The institutional review boards approved this study (Seoul Metropolitan Government Seoul National University Boramae Medical Center, 07-2016-22). This study used National Health Insurance Database (NHIS-HealS data, NHIS-2017-4-009) provided by National Health Insurance Service.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.