The association between oral contraceptive use and the risk of metabolic syndrome in Korean women: a national population-based study
Article information
Abstract
Background
This study aimed to evaluate the association between oral contraceptive (OC) use and the incidence of metabolic syndrome (MetS) and its components in Korean women.
Methods
We conducted a cross-sectional study including 11,084 Korean women between 2012 and 2020 based on nationally representative data from the Korea National Health and Nutrition Examination Survey. Multivariate logistic regression analysis was conducted to examine the relationships between OC use, MetS, and its components.
Results
In total, 11,084 women were included in the study, of whom 1,117 (10.1%) used OCs and 8.5% had MetS. The prevalence of MetS was higher in the OC users (11.5%) than in the non-OC users (8.1%) (p<0.001). The OC users had an increased risk of MetS (odds ratio [OR], 1.480; 95% confidence interval [CI], 1.215–1.802; p<0.001). Additionally, OC users had higher risks of abdominal obesity (OR, 1.319; 95% CI, 1.119–1.555), hypertension (OR, 1.557; 95% CI, 1.302–1.863), and hypertriglyceridemia (OR, 1.521; 95% CI, 1.287–1.797).
Conclusions
The incidence of MetS was higher among OC users, with waist circumference, hypertension, and hypertriglyceridemia being significantly more prevalent components of MetS. These findings can provide valuable evidence to support the development of evidence-based health policies related to women’s health.
Introduction
Metabolic syndrome (MetS) is characterized by metabolic issues, such as obesity, hyperglycemia, dyslipidemia and hypertension (HTN), that develop in an individual due to insulin resistance [1].
Oral contraception is a widely accepted and effective method globally [2]. In addition to its contraceptive effect, oral contraceptives (OCs) offer advantage such as improving in menorrhagia and dysmenorrhea, reducing symptoms of premenstrual dysphoric disorder [3], and reducing the risk of endometrial [4], and ovarian cancer [5]. However, some studies have suggested potential adverse health effects of OC use, including cervical, breast cancer [6], and ischemic stroke [7].
To date, there has been inconsistency and limited study regarding the effects of OCs on MetS components. Therefore, this nationwide study was conducted with the purpose of examining the relationship between OC use and the risk of MetS in Korean women, utilizing data from the Korea National Health and Nutrition Examination Survey (KNHANES).
Methods
Ethical statements: This study was approved by the Institutional Review Board (IRB) of Pusan National University Yangsan Hospital (IRB No. 55-2024-141). Written informed consent was waived.
1. Study participants and design
This study was based on the raw data from the 2012 to 2020 KNHANES. The KNHANES consists of three component surveys: a health interview, examination and nutrition survey. Trained medical staff and interviewers conducted the health interview and examination.
We reviewed 35,301 women who were participated in the KNHANES. Our study included 13,454 women aged 19–50 years who participated in both the health interview and the health examination surveys. Of these 13,454 participants, 2,370 were excluded if they were either pregnant, breastfeeding, had diagnosed MetS before menarche, had begun the menopause, had diagnosed breast cancer and uterine cancer, or not answering the survey question about menstrual status. Thus, 11,084 subjects were included in the final analysis.
This study was conducted using a cross-sectional design. Information on OC use was obtained through a self-administered questionnaire: “Have you ever taken oral contraceptives for at least a month?” The data were divided into two groups (OC users and non-OC users). These surveys included questions regarding the demographic (age, residence), socioeconomic (employment status, education, and income level) information. Education levels were categorized into three categories: ≤9 years (less than high school), 10–12 years (high school), ≥13 years (college or more). Income was divided into quartiles to represent income level (1, low; 2, middle low; 3, middle high; 4, high).
2. Measurements
All health examination procedures were conducted by trained medical staff and all the equipment was regularly calibrated. Height was measured to the nearest 1 mm with a portable extensimeter (SECA225; SECA). Weight was measured with digital scale (GL-6000-20; G-tech) accurate to 0.1 kg after wearing a disposable examination gowns. Body mass index (BMI) was calculated by dividing weight (kg) by height squared (m2). Waist circumference (WC) was measured during exhalation to the nearest 0.1 cm at the narrowest circumference between the inferior margin of the rib and the superior borders of the iliac crest. Systolic and diastolic blood pressure (BP) were measured manually three times using mercury sphygmomanometer (Baumanometer Wall Unit 33 (0850); Baum) after participants were rested for 5 minutes in a sitting position.
3. Definition of MetS
MetS was defined based on the 2009 criteria of International Diabetes Federation and the American Heart Association/National Heart, Lung, and Blood Institute [8]. The Korean Society for the Study of Obesity provided criteria for abdominal obesity [9]. Participants with ≥3 of the following five variables were classified as having MetS: (1) WC ≥85 cm, (2) triglycerides ≥150 mg/dL or taking medicine for hypertriglyceridemia, (3) high-density lipoprotein cholesterol (HDL-C) <50 mg/dL or taking medicine for controlling HDL-C, (4) systolic/diastolic BP ≥130/85 mmHg or taking medicine for HTN, and (5) fasting glucose ≥100 mg/dL or taking medicine for diabetes.
4. Statistical analysis
Since the KNHANES sample design was extracted using stratified cluster sampling, statistical analysis was performed through a composite sample data analysis method considering the weights. Continuous variables were expressed as the mean±standard deviation, while categorical variables were showed as frequency and percentage (%). The appropriate statistical tests, such as the chi-square test and independent t-test, were utilized where applicable. Multivariate analysis was performed to assess the odds ratios (ORs) of MetS and its components based on the use of OCs. Several regression models were shown to control for the potentially confounding variables. Initially model I adjusted for age, followed by the inclusion of BMI in addition to the variables from model I (model II). Finally, model III included the variables from model II as well as residence, occupation, and education. The statistical analysis was performed using R 3.5.3 (The R Foundation). In all analyses, p-values were two-tailed, and a p-value <0.05 considered statistically significant.
Results
Of the 11,084 participants included in this study, 1,117 (10.1%) were users of OCs. The general characteristics of the study participants according to the use of OCs are shown in Table 1. Residence, education, income level differed statistically between the two groups. MetS was present in 8.5% of the participants with a higher prevalence of OC users (11.5%) than in the non-OC users (8.1%; p<0.001). Among the components of MetS, OC users had significantly higher incidence of WC (17.5%), hypertriglyceridemia (17.2%), and HTN (14.5%; p<0.001).
General characteristics of participants
Table 2 presents the mean values for the MetS components between the groups. The mean number of five MetS components was 1.0 in the OC user group, and 0.8 in the non-OC user group (p<0.001). For individual MetS components, the OC users had significantly higher WC, total cholesterol, systolic and diastolic BP, HDL-C, and triglyceride levels (p<0.05).
MetS components in OC and Non-OC users
Table 3 shows the differences in risk factors between the use groups. In the OC user group, age, weight, BMI, WC, systolic and diastolic BP, total cholesterol, low-density lipoprotein cholesterol (LDL-C), triglyceride, fasting blood sugar, and hemoglobin A1c were significantly higher in the group with MetS (p<0.05). However, HDL-C was significantly lower in this group (p<0.001).
Differences of risk factors according to the presence of MetS in OC users and non-OC users
Logistic regression analysis was performed to determine the association between OC use and MetS (Table 4). OC users were associated with an increased risk of MetS (OR, 1.480; 95% confidence interval [CI], 1.215–1.802; p<0.001) in the crude model. In model I, after adjustment for age, the OR was 1.463 (95% CI, 1.196–1.788). Further, in model II (adjusted for age, and BMI), the OR was 1.419 (95% CI, 1.115–1.806). In model III, after adjustment age, BMI, place of residence, occupation, and education, the risk of MetS remained statistically significant (OR, 1.373; 95% CI, 1.077–1.750).
Association between OC use and MetS
Table 5 presents the OR for the relationship between OC use and MetS components. Compared to the non-OC users, OC users were at a higher risk of abdominal obesity (OR, 1.319; 95% CI, 1.119–1.555), HTN (OR, 1.557; 95% CI, 1.302–1.863), and hypertriglyceridemia (OR, 1.521; 95% CI, 1.287–1.797) in the crude model. After controlling for the confounding factors, the risk for HTN (OR, 1.551; 95% CI, 1.287–1.869) and hypertriglyceridemia (OR, 1.511; 95% CI, 1.277–1.789) in model I, abdominal obesity (OR, 1.369; 95% CI, 1.054–1.778), HTN (OR, 1.520; 95% CI, 1.254–1.842) and hypertriglyceridemia (OR, 1.479; 95% CI, 1.239–1.766) in model II, and abdominal obesity (OR, 1.343; 95% CI, 1.031–1.749), HTN (OR, 1.457; 95% CI, 1.201–1.769) and hypertriglyceridemia (OR, 1.441; 95% CI, 1.205–1.722) in model III were higher in the OC user group with statistical significance. However, low-HDL-C was not a risk factor in model II (OR, 0.870; 95% CI, 0.758–0.998) and model III (OR, 0.853; 95% CI, 0.742–0.980).
Relationship between OC use and MetS components
Discussion
This population-based cross-sectional study aimed to investigate the relationship between OC use and the prevalence of MetS components among Korean women, using data obtained from the KNHANES. Several key findings emerged from our study. Firstly, OC users exhibited significantly higher WC, total cholesterol, systolic and diastolic BP, HDL-C, and triglyceride. Secondly, OC users were associated with a higher risk of MetS. Thirdly, among the MetS components, OC users had a higher risk of abdominal obesity, HTN, and hypertriglyceridemia.
The disruption of insulin signaling can affect crucial cellular pathways involved in maintaining the energy balance and glucose homeostasis, leading to insulin resistance and the development of various metabolic disorders, including cardiovascular diseases, obesity, and type 2 diabetes [10]. Experimental studies with high-dose OCs have suggested a chronic increase in blood glucose and insulin levels, as well as of impaired glucose tolerance, potentially linked to the development of overt diabetes [11]. However, depending on the type and dose, transient diabetogenic stress reportedly increases impaired glucose tolerance in high-dose OC users [12]. The effect of OCs on carbohydrate metabolism is associated with progestin’s androgenicity [13]. However, the progesterone derivatives appear to have a lesser impact on insulin secretion. Therefore, OCs do not deteriorate insulin sensitivity [14,15]. In our study, we did not observe an association between taking OCs and risk of diabetes.
Our data showed a positive association between the use of OCs and risk of HTN. Estrogens reportedly regulate vascular tone through various mechanisms, including nitric oxide, prostacyclin, angiotensin, and the sympathetic nervous system [16,17]. Additionally, other studies have indicated that the type and dosage of progestin in OCs can influence BP [18,19]. Numerous studies have consistently reported an increase in BP following hormonal contraceptive use.
A prospective cohort study performed in the United States founded that both systolic and diastolic BP were higher in current OC users than in those who have never used OCs, with a difference of only 0.7/0.4 mmHg [20]. Similarly, a study involving British women aged 18 to 30 years showed a modest increase of 2 mmHg in systolic BP, while no significant change was observed in diastolic BP [21]. Furthermore, several prospective studies have consistently demonstrated that the effects of OCs on BP decrease rapidly upon discontinuation of OC use [22,23].
Limited studies have focused on the association between past hormonal contraceptive use and the risk of subsequent high BP. The Nurse’s Health Study (NHS) examined the use of hormonal contraception in the past and its duration, and reported no increase in the risk of cardiovascular disease, coronary artery disease, or stroke [24,25]. A secondary study (NHS II) found that the past users showed a slight increase in the risk of developing HTN, although the increase was not significant [20]. More recently, Chiu and Lind [26] conducted a study and reported no relationship between previous hormonal contraceptive use and HTN. Additionally, they found that the duration of hormonal contraceptive use was not related with the odds for HTN in any age group, compared to those who had never used OCs.
A recent meta-analysis [27] found a positive association between the duration of OCs use and the risk of HTN. In addition, a linear relationship was found between the duration of OC use and the risk of HTN. However, the relationship between the duration of taking OCs and HTN was not examined in our study.
The estrogenic component of OCs has been shown to increase triglyceride levels and the production of lipoproteins such as very LDL-C and HDL-C, while having the opposite effect on LDL-C. Progestin counterbalance these effects, and combined hormonal contraceptives maintain a balanced dosage between the two hormones [8]. Numerous studies have investigated the effects of OCs on lipid measurements, including those of total cholesterol and triglycerides. Previous research has reported increased triglyceride and HDL-C levels, as well as decreased LDL-C levels among OC users [28,29]. In our study, we observed an association between OC use and increased levels of total cholesterol, triglyceride, and HDL-C, but no association was observed with LDL-C.
This study reveals a significant relationship between the OC use and WC in women. Weight gain is often considered one of the most common side effects of OC use, potentially resulting from factors such as fluid retention, fat deposition, or increase of muscle mass [30]. However, findings on the association between weight change and OC use have been inconsistent. Conversely, the OC use was not significantly associated with long-term BMI change [31].
Our study also found that OC users had a higher prevalence of MetS. One notable finding of our study was that OC use was likely to contribute to an increased prevalence of the components of MetS. WC, HTN, and hypertriglyceridemia are positively associated with OC use.
This study should acknowledge a few limitations. First, the collected data relied on self-reporting, which introduced the possibility of recall bias. Additionally, the cross-sectional design of the KNHANES limits our ability to establish causal relationships. In Korea, most OCs contain 0.02–0.035 mg of ethinyl estradiol as the estrogen component in addition to a progentin component. However, the lack of specific information regarding the specific type of OC formulation and the duration of OC use hinders our ability to establish an association between these factors and MetS.
Notwithstanding these limitations, this study has notable strength. Firstly, we analyzed a large, nationally representative survey from the KNHANES, enhancing the generalizability of our findings. Second, our models were adjusted for confounding factors, increasing the validity of our results. These findings may serve as valuable evidence for the development of evidence-based health policies related to women’s health.
In this large population-based study, the odds of MetS were significantly higher among women using OCs. Among the components of MetS, the prevalence of WC, HTN, and hypertriglyceridemia was significantly higher in OC users. Therefore, controlled studies are necessary to investigate the casual associations between OC use and MetS.
Notes
Conflicts of interest
No potential conflict of interest relevant to this article was reported.
Funding
This study was supported by a 2023 research grant from Pusan National University Yangsan Hospital.
Author contributions
Conceptualization: TK. Data curation: HKP, TK. Formal analysis: TK. Funding acquisition: TK. Methodology: HKP, TK. Writing-original draft: HKP, TK. Writing-review & editing: TK. All authors read and approved the final manuscript.