2.1. Study Participants
This study was analyzed using the raw data from the 2018–2021 Korea National Health and Nutrition Examination Survey (KNHANES). The KNHANES is a nationwide survey conducted in South Korea that aims to collect annual data on various aspects such as sociodemographic, economic, and health-related conditions and behaviors for all age groups. Since 2007, the Korea Centers for Disease Control and Prevention Research Ethics Review Committee has reviewed and approved the data collected on an annual basis. The KNHANES is publicly available as secondary data for researchers to further analyze 30.
There were 30,551 participants in the 2018–2021 KNHANES. 5,616 participants under the age of 20 were excluded. Then, participants were divided into day and overnight groups for each year, and matched based on age and gender. Subsequently, datasets of 194, 190, 150, and 102 participants were generated for 2018, 2019, 2020, and 2021, respectively. Among the 636 participants, 32 were excluded due to missing values in the survey questions. Finally, a total of 604 participants (394 men and 210 women) were selected for the study.
2.2 Anthropometrics
The KNHANES collects gender and age information through self-reported questionnaires, and in this study, age was divided into 20s to 70s or older for analysis. Anthropometric measurements are taken by trained examiners using standardized procedures. Height is measured with a stadiometer while participants stand upright, without shoes, and with their heels, buttocks, and shoulders touching the device. Weight is measured with a digital scale while participants wear light clothing and no shoes. Waist circumference is measured with a non-stretchable tape measure at the midpoint between the lower border of the rib cage and the iliac crest 30. BMI is calculated by dividing weight (kg) by height squared (m2) and participants are classified according to World Health Organization (WHO) criteria as underweight, normal weight, overweight, or obese.
2.3. Socioeconomic factors
The socioeconomic factors included educational level (≤ elementary, middle, high, ≥college), occupational status (white-collar, pink-collar, blue-collar, gray-collar, not classified), and household personal income quintile (low, lower middle, normal, upper middle, high).
2.4. Measures
2.4.1. Group
The group was divided by considering worktime schedules. In response to the questionnaire, "Do you mainly work during daytime hours (6 am – 6 pm) or at another time?", the participants were organized based on their work schedules into two groups: (1) Day workers (6 am – 6 pm) and (2) Overnight workers (9 pm – 8 am).
2.4.2. Lifestyle behaviors
In the KNHANES, lifestyle was defined as the number of drinks per week (1) No drinking in the past year, (2) Less than once a month, (3) About once a month, (4) 2-4 times a month, (5) 2-3 times a week, (6) 4 or more times a week, but in this study, considering the number of people belonging to each variable, we divided it into (1) not at all in the past, (2) once a month, (3) 2–4 times a month, (4) 2 or more times a week. Smoking status was analyzed by dividing it into (1) current, (2) former, and (3) none.
Health-related factors
The health-related factors included moderate to vigorous physical activity (MVPA), number of strength training days per week, sedentary behavior time (SB time), and perception of stress. Participants were categorized based on whether their weekly MVPA met the WHO-recommended 150 (mins) threshold (1) 150<, (2) ≥150 and whether they engaged in strength training (1) participate, (2) not at all. For SB time, the difference between groups was compared by dividing the total SB time (mins) of study participants into 3 parts (1) ≤ 420, (2) 421 – 600, (3) ≥ 601. The participant’s perception of stress was categorized into three levels, which were (1) not at all, (2) a little, and (3) quite.
Blood analysis
The KNHANES team obtained written informed consent from participants after explaining the purpose and procedure of blood tests for hypertension, fasting glucose, total cholesterol, HDL cholesterol, triglycerides, and HbA1c. Participants were instructed to fast for a minimum of 8 hours prior to blood collection to ensure accurate measurements of these parameters. The results were classified into three levels—normal, borderline (warning), and dangerous—based on established criteria for each health indicator. Hypertension; (1) normal (systolic <120 mmHg and diastolic <80 mmHg), (2) prehypertension (systolic 120 – 139 mmHg or diastolic 80 – 89 mmHg), and (3) hypertension (systolic ≥140 mmHg or diastolic ≥90 mmHg), Fasting glucose; (1) normal (<126 mg/dL), and (2) impaired fasting glucose (≥126 mg/dL), Total cholesterol; (1) normal (<200 mg/dL), (2) borderline (200 – 239 mg/dL), and (3) hyperlipidemia (≥240 mg/dL), HDL cholesterol; (1) low (<40 mg/dL), (2) normal (40-59 mg/dL), and (3) high (≥60 mg/dL), Triglycerides; (1) normal (<150 mg/dL), (2) borderline (150 – 199 mg/dL), and (3) high (≥200 mg/dL), HbA1c; (1) normal (<5.7%), (2) borderline (5.7 – 6.4%), and (3) diabetes (≥6.5%).
Data analysis
All data processing and statistical analysis used SPSS 28.0 version (SPSS Inc., Chicago, IL, USA). Descriptive statistics were used to summarize the sample characteristics for the first six variables (gender, age, height, weight, waist circumference, and body mass index; BMI). Educational level, occupational status, and household personal income quintile were examined as demographic factors that may impact health outcomes. The chi-square analysis (χ2 test) was used to compare the categorical data of general characteristics between day and overnight workers. An independent t test was conducted to analyze the mean difference of continuous variables (i.e., BMI, Average hours of work per week, Average hours of sleep per day (on weekdays and weekends), Waist circumference, MVPA per week (mins), SB time per week (mins), and blood related factors) between the two groups. Moreover, multinomial regression analysis was performed to find out the associations between the factors and work schedule. Then, we generated a forest plot to visually represent the results of the multinomial logistic regression. Results are presented as odds ratios (OR) with 95% confidence intervals (95% CI) and statistical significance of all analyzes was set by p < 0.05.