Impact of COVID-19 and Associated Preventive Measures on Cardiometabolic Risk Factors in South Korea: An Observational Study

Minji Sohn Seoul National University Bundang Hospital Bo Kyung Koo Seoul National University Seoul Metropolitan Government Boramae Medical Center Soo Lim (  limsoo@snu.ac.kr ) Seoul National University Bundang Hospital https://orcid.org/0000-0002-4137-1671 Ho Il Yoon Seoul National University Bundang Hospital Kyoung-Ho Song Seoul National University Bundang Hospital Eu Suk Kim Seoul National University Bundang Hospital Hong Bin Kim Seoul National University Bundang Hospital

In South Korea, the COVID-19 pandemic broke out on February 22, 2020 (Fig. 1a), and the government raised the crisis alert level to "severe," the highest, for the second time, which was rst actioned during the in uenza outbreak in 2009 (Fig. 1b) [5]. The infectious disease prevention and control measures included shutting down public facilities such as libraries and sports centers, and suspending school attendance. The public was requested to stay home, refrain from going outside, and avoid crowded environments at work [6]. Thus, public movement decreased by 38.1% during the early period of the COVID-19 outbreak compared with before the outbreak [7]. Thanks to these preventive measures and cooperation from the general public, the number of daily new domestic infections in South Korea fell dramatically [8]. People have become accustomed to distancing themselves by acknowledging its importance, which has been re ected in the increase in online shopping [9]. The daily number of new COVID-19 cases in South Korea has dropped to fewer than 100 up to the end of October, but unfortunately, it has rebounded to over 500 recently [8]. This might have been because the Korean government eased the quarantine level: people could get together and became less rigorous in following preventive principles. The Korean government has now raised the crisis alert level, which means stricter observation of social distancing, and a ban on staying not only in public places but also in indoor health facilities, cafés, bars, and any places that people can get together, and is now prohibiting large-scale gatherings again (see http://ncov.mohw.go.kr/en).
As we recently reviewed [10], old age, diabetes mellitus (DM), cardiovascular disease (CVD), hypertension, and obesity are determining factors for fatal outcomes of COVID-19. People with coronary heart disease (CHD) or DM had a higher chance of being admitted to intensive care units, needing mechanical ventilation, or of dying [11,12]. According to a case-control study in South Korea (7, cases among 219,961 patients), DM, hypertension, and chronic renal failure including end-stage renal disease were associated with increased disease severity in [13].
Because elevated glucose levels directly promote SARS-CoV-2 replication, which essentially requires glycolysis in the host [14], patients with uncontrolled DM are expected to experience a more rapid progression of COVID-19. A recent study reported that increased glucose concentrations signi cantly predicted mortality in those with COVID-19, regardless of the presence of DM [15]. Conversely, inactivity associated with social and physical distancing for COVID-19 might impair metabolic control. Thus, the COVID-19 pandemic is likely to have a negative in uence on public lifestyle and behaviors [16], which could adversely affect cardiometabolic health [17,18]. So far, the exact effects of COVID-19 prevention and control measures including social distancing, movement restriction, and limitations of gathering on the impact of such chronic diseases have not yet been evaluated. Here, we hypothesized that the COVID-19 pandemic and associated unhealthy lifestyles have produced negative in uences on metabolic parameters in individuals with cardiometabolic risk factors.

Study design and population
This was a single-center, retrospective, observational cohort study conducted at Seoul National University Bundang Hospital (SNUBH) in South Korea. The study was approved by our independent Ethics Committee/Institutional Review Board (SNUBH: B-2008/630 − 102).
The study population was adults aged over 19 years with diagnosed cardiometabolic risk factors including impaired glucose metabolism, hypertension, dyslipidemia, or obesity who visited the outpatients' clinic at the Department of Endocrinology and Metabolism at SNUBH.
Patients who visited from September 1, 2016 to May 31, 2020 at least twice a year, before and after February, the time of South Korea's COVID-19 outbreak, were further selected. In all, 7,094 patients were identi ed to have International Classi cation of Diseases Tenth Edition (ICD-10) diagnostic codes of E10-14 for DM, I10 and I15 for hypertension, E78 for dyslipidemia, and E66 for obesity using the hospital database, clinical data warehouse (CDW) [19].
Patients who were hospitalized for a major illness or major surgery, and who received dialysis during the study period were excluded (Fig. 2). Major surgery was de ned as surgery performed for neoplasms, severe diseases in the circulatory or digestive systems, and injuries determined by ICD-10 codes starting with C, D, I, K, and S. The number of patients hospitalized in the endocrinology unit did not differ between years, but they were excluded from the study analysis because hospitalization for intensive glucose-lowering therapy might have hindered identifying the impact of pandemic preventive measures.

Collection of clinical parameters
The Korean government reinforced the national public health emergency response by emphasizing the need to maintain social distance on Patients' outpatient care information, admission information, clinical laboratory values, anthropometric measurements, and prescription information were retrieved from the CDW. Body weight, body mass index (BMI), systolic and diastolic blood pressure (SBP and DBP, respectively), and metabolic pro les such as the levels of fasting plasma glucose (FPG), glycated hemoglobin (HbA1c), total cholesterol, triglyceride (TG), high-density lipoprotein-cholesterol (HDL-c), and low-density lipoprotein-cholesterol (LDL-c) in each season were analyzed as means for each individual. Data cleaning was performed for manually inputted anthropometrics into the system at the time of care.
Thus, obviously inaccurate values, recorded as ranges or as typographic errors, physically impossible values such as height > 300 cm and DBP > SBP, were discarded. Height was converted to a 4-year mean measure for each patient. Then, the BMI was recalculated as mass (in kilograms) divided by height (in meters) squared. The use of medications for DM, hypertension, and dyslipidemia was also investigated.

Anthropometric and biochemical parameters
Anthropometric and biochemical parameters were measured in SNUBH as reported previously [20]. Height and body weight were measured using standard methods with the subjects in light clothing. The FPG concentration was measured using the glucose oxidase method (747 Clinical Chemistry Analyzer; Hitachi, Tokyo, Japan). HbA1c levels were measured using a Bio-Rad Variant II Turbo High-Performance Liquid Chromatography Analyzer (Bio-Rad, Hercules, CA, USA) in a National Glycohemoglobin Standardization Program level II certi ed laboratory.
Total cholesterol, TG, HDL-c, and LDL-c levels were measured using a 747 Clinical Chemistry Analyzer (Hitachi).
Metabolic syndrome was de ned using modi ed Adult Treatment Panel III criteria [21]. Because the waist circumference data were limited, the BMI was used instead as suggested by the World Health Organization [22]; metabolic syndrome was diagnosed as the existence of at least three abnormal components of: (i) BMI ≥ 23 kg/m 2 and/or taking anti-obesity agents; (ii) SBP ≥ 130 mmHg, DBP ≥ 85 mmHg, and/or taking antihypertensive agents; (iii) TG ≥ 150 mg/dL and/or taking lipid-lowering agents; (iv) HDL-c ≤ 40 mg/dL in men and HDL-c ≤ 50 mg/dL in women, and (v) FPG ≥ 100 mg/dL and/or taking antidiabetic agents. Patients with HbA1c ≥ 6.5% and/or taking antidiabetic agents were classi ed according to the state of DM treatment.

Patient characteristics
A total of 1,485 patients were included in this study, with a mean age of 61.8 ± 11.7 years in September 2016. The proportions of men and women were almost equal among the study participants. All of them had at least one chronic cardiometabolic impairment such as DM, hypertension, dyslipidemia, or obesity at baseline ( Table 1). The number of comorbid diseases tended to increase with time. The total use of antidiabetic agents increased in 2019 compared with 2016. The total use of antiobesity agents increased in 2019-2020 compared with the previous 3 years. The usage of recently approved antidiabetic drugs, sodium-glucose co-transporter 2 (SGLT2) inhibitors and glucagon-like peptide-1 receptor agonists (GLP-1 RAs), increased in 2019-2020.
The 10-year CHD risk was calculated using the Framingham risk score (FRS) [23]. The correlation of calculated CHD risk with actual 10-year CHD was shown to be stronger when using total cholesterol levels than when using LDL-c scoring in Korean subjects [24]. Therefore, the FRS with total cholesterol scoring was used here.

Statistical analysis
Continuous variables are summarized as the mean ± standard deviation (SD) and categorical variables are shown as the numbers and percentages of subjects. Normality of data distribution was evaluated using the Shapiro-Wilk test and by histograms [25], which showed all variables to be normally distributed with bell-shaped symmetric graphs. Student's t tests for continuous variables and chi-squared tests for categorical variables were used for comparisons. The changes of values from winter 2019 to spring 2020 (2019-2020 season) were compared with that of previous seasons, 2018-2019, 2017-2018, and 2016-2017 seasons, using paired Student's t tests. Because the patient follow-up period was up to 6 months, not all the patients had complete seasonal data. To reduce the bias of visiting patient's characteristics, the dataset was imputed using multiple imputation by chained equations for the missing values from patients who did not have complete test results [26]. The imputed data were used to analyze changes in values. Relative risk (RR) was calculated as the number of patients who showed worse metabolic syndrome components in the 2019-2020 season than those in the 2018-2019 season, expressed as RR with a 95% con dence interval (CI). Statistical signi cance was considered at a two-sided p value < 0.05. All analyses were performed using R software version 4.0.2 (R Development Core Team, Vienna, Austria) and RStudio version 1.3.1056 (RStudio, Inc., Boston, MA, USA).

Changes in cardiometabolic risk factors during the COVID-19 pandemic
The raw values of cardiometabolic risk components before and during social distancing are shown in Fig. 3.  (Fig. 4a) (Fig. 4b). Male patients were affected more severely with respect to metabolic syndrome components by the COVID-19 pandemic than were female patients (Fig. 4c)

Changes in coronary heart disease risk during the COVID-19 pandemic
The changes in 10-year CHD risk by FRS are shown in Fig. 5.

Discussion
Here, we found that cardiometabolic risk factors deteriorated signi cantly in subjects with metabolic impairment in South Korea during the COVID-19 pandemic. In this critical 2019-2020 season, the proportion of subjects with metabolic syndrome increased signi cantly by 21% compared with the 2018-2019 season. The 10-year CHD risk also increased by 1.0 ± 6.2% compared with the past three years. We also found that not only the body weight or BMI but also blood pressures, lipid pro les, and HbA1c changed in an unfavorable direction during the COVID-19 pandemic.
Social distancing policy, to tackle COVID-19, naturally reduces people's physical activities. In many countries, trips to all major destinations except to personal residences dropped signi cantly by 50-80% in early March when COVID-19 was declared a pandemic (https://kojects.com/2020/06/01/mobility-korea-covid-19/). A recent self-reporting survey showed that people spent more time at home and actually gained weight during the COVID-19 pandemic [27]. It was reported that acutely reduced physical activities during the COVID-19 pandemic might help increase insulin resistance and gain fat mass [28]. Various public health interventions including staying at home, refraining from nonessential social activities, and school closures limit access to healthy food options [29]. Moreover, people are consuming home-delivered foods frequently, which are more obesogenic than homemade food [30]. The COVID-19 pandemic is also in uencing mental health [31]. Scared of getting an infection or dying, many people are psychologically distressed, which might lead to systemic in ammation [18]. Stress stimulates elevations in blood pressure and blood glucose levels by releasing cortisol through the hypothalamic-pituitaryadrenal axis [32].
We found that the 10-year CHD risk of patients has increased during the COVID-19 pandemic in South Korea. Notably, patients aged over 65 years increased their 10-year CHD risk score the most (1.2 ± 7.1%) and this-potentially-could contribute to a high mortality rate from COVID-19 in the elderly. Body weight, blood pressure, and lipid levels decreased in spring before the COVID-19 pandemic, similar to the results from previous studies [33][34][35] However, these cardiometabolic risk parameters increased signi cantly in the same period in the pandemic season. This opposite trend might re ect the impact of the pandemic on cardiometabolic risk parameters in most of our patients.
Several mechanisms have been suggested to explain the causality of the COVID-19 pandemic and the increased risk of metabolic disorders and CVDs. The sympathetic system is activated with increased levels of catecholamines after catastrophic events, which in uences the heart and blood vessels negatively [36]. In metabolic dysregulated status, the renin-angiotensin system is activated inappropriately, which also leads to increased production of angiotensinogen (up to 30% of circulating angiotensinogen) and to elevated plasma renin activity, which in turn contributes to increasing blood pressure and deteriorating glucose metabolism [37,38]. Although the effects of this pandemic may not be seen in the short term, its long-term impacts on cardiometabolic risk cannot be ignored given the stressful socioeconomic conditions [17].
In this study, both males and patients under 65 years showed a signi cantly increased risk for metabolic syndrome during the COVID-19 pandemic. In general, middle-aged men are more involved in economic activity than women or elderly populations [39]. Based on this, current preventive measures might have a greater impact on physical activity in these subgroups. Such insu cient physical activity is likely to deteriorate cardiometabolic health eventually [40].
In this analysis, compared with other cardiometabolic parameters, there was no increase in the HbA1c levels during the COVID-19 pandemic.
The results might be because of increased usage of potent novel antidiabetic agents such as SGLT2 inhibitors and GLP-1 RAs. Indeed, the patients who started SGLT2 inhibitors after September 2019 showed reductions in HbA1c levels in the spring of 2020 (data not shown).
Importantly, SGLT2 inhibitors should be avoided for severely ill patients because this agent can cause ketoacidosis and acute kidney injury [41]. The use of liraglutide also increased more than vefold in the 2019-2020 season compared with previous seasons. Given that the bene cial roles of GLP-1 RAs for preventing cardiovascular and kidney diseases have been well established [42], these can be an ideal option for the treatment of patients with type 2 DM at such risk even during the COVID-19 pandemic [43].
During this pandemic, speci c schemes are required to curtail a potential vicious cycle because patients with dysregulated metabolism have a worse prognosis when infected. Metabolic syndrome represents a state of chronic low-grade in ammation and the elevated release of cytokines in metabolic syndrome status is likely to provoke a "cytokine storm" in those individuals infected with SARS-CoV-2, which may lead to multiorgan failure [44,45]. Considering the deterioration in cardiometabolic pro les during the COVID-19 pandemic, physicians should focus on patients with metabolic impairments to prevent future adverse cardiovascular events. Governments and medical institutions must promote physical activity, healthy eating, and mental health care during such pandemics. Social media or web-based programs can provide convenient tools to guide such patients to have healthy lifestyles. Active counseling to help people with metabolic dysregulation cope with barriers against healthier lifestyles would be helpful in this critical situation [46].
Our research had advantages in that we exclusively included regularly attending outpatients, who were followed up for four years to reduce bias. Because SNUBH is a tertiary hospital receiving patients from all over the country, our results might be representative of our population at large. Nonetheless, some limitations need to be mentioned. We did not investigate changes in physical activity or dietary habits. Moreover, it was not possible to observe the actual occurrence of CHD given the short observation period. Instead, we used the 10-year CHD risk estimated from the FRS, but this is a well-established tool that has been used widely for this purpose [23].

Conclusions
We found that the COVID-19 pandemic had a negative in uence on cardiometabolic pro les in subjects with metabolic impairments. This     The changes in metabolic syndrome components before (i.e., fall and winter, 2019) and during (i.e., spring, 2020) the COVID-19 pandemic in South Korea: a body weight, b BMI, c SBP, d DBP, e HbA1c, f FPG, g triglycerides, h HDL-c. Key: SBP, systolic blood pressure; DBP, diastolic blood pressure; BMI, body mass index; HbA1c, glycated hemoglobin; FPG, fasting plasma glucose; HDL-c, high-density lipoproteincholesterol.