2.3. Characteristics of the Participants by Interview and Biochemical Assays

Participant’s age, gender, income, education, and area of residence, were collected using the survey questionnaires. Their height and weight were measured wearing light clothes and bare feet, as described previously [14,15]. The appendicular skeletal muscle mass and fat mass were estimated indirectly using the prediction model generated by the XGBoost algorithm from the Ansan/Ansung cohort, where the mass was measured by the Inbody^® body composition analyzer (Cheonan, Korea) [16]. A doctor measured the participant’s blood pressure in the sitting position thrice using a blood pressure monitor, and the average diastolic blood pressure (DBP) and systolic blood pressure (SBP) were calculated for the study. The participants were queried about any previous diagnosis of inflammatory diseases, including bronchitis, asthma, arthritis, osteoporosis, allergy, periodontitis, and gastritis.

After fasting for more than 12 h, serum was separated from blood collected from the vacutainer without blood coagulant and plasma from the vacutainers containing ethylenediaminetetraacetic acid (EDTA) and heparin. Fasting plasma glucose and glycosylated hemoglobin (HbA1c) concentrations were measured using a Hitachi 7600 Automatic Analyzer (Hitachi, Tokyo, Japan) and an automatic analyzer (ZEUS 9.9; Takeda, Tokyo, Japan), respectively. Biochemical parameters in the serum were determined using a Hitachi 7600 Automatic Analyzer. The parameters were as listed as follows: total cholesterol, high-density lipoprotein cholesterol (HDL-C), triglycerides (TG), total bilirubin, and creatinine concentrations, and alanine aminotransferase (ALT) and aspartate aminotransferase (AST) activities. Serum high-sensitive C-reactive protein (hs-CRP) concentrations were measured using a high-sensitive enzyme-linked immunosorbent assay (ELISA) kit (Thermofisher, Waltham, MA, USA). White blood cells (WBC) were counted with an automated hematology analyzer (Beckman Coulter, Brea, CA, USA) using the EDTA-treated blood. Insulin resistance was predicted using a machine learning prediction model made with the homeostatic model assessment of insulin resistance (HOMA-IR) in the An-san/Ansung cohort from the Park et al. study [17].

From the health survey, the daily alcohol intake (g/day) was calculated by multiplying its consumption frequency by the amounts consumed [18]. Participants who had smoked at least 20 cigarettes in the past six months were considered current smokers, and those who smoked more than 20 cigarettes but had not smoked for at least six months were former smokers [18]. Coffee intake was estimated with a similar method of alcohol intake. Regular exercise was defined as more than 150 min of moderate physical activity per week.