Study population
The study included 9,895 participants aged 40 to 69 from the Anseong-Ansan Cohort of the Korean Genomic Epidemiology Study conducted by the Korea Disease Control and Prevention Agency. This study investigated genetic and environmental factors that contribute to the prevalence of metabolic and cardiovascular diseases. Individuals residing in rural (Anseong) and urban (Ansan) communities were enrolled from June 2001 to January 2003. Detailed protocols have been described in previous publications 23, 24.
A total of 10,030 eligible individuals who lived in Anseong City (n = 5,018) or Ansan City (n = 5,012) for at least 6 months were enrolled in the study. Participants diagnosed with AF at baseline (n = 74) and those without SRH records (n = 61) were excluded, resulting in a final number of participants of 9,895.
During the baseline visit to the tertiary hospital, numerous comprehensive physical examinations, detailed on-site interviews, and thorough clinical examinations were conducted. Six consecutive reassessments were conducted through biennially scheduled revisits until 2014, following the entire cohort protocol. All participants voluntarily enrolled in the study and provided written informed consent at the baseline assessment and each follow-up visit. This study followed the principles of the Declaration of Helsinki and was approved by the Institutional Review Board (IRB) of the Korean National Institutes of Health and Hanyang University Medical Center (IRB number: HYUH 2017-12-033).
Lifestyle and medical history assessment, physical exam, and laboratory tests
Well-trained investigators conducted comprehensive on-site interviews, collected important lifestyle and clinical data, and performed physical examinations at each visit to a tertiary hospital. A structured questionnaire was used to obtain data on smoking, alcohol intake, education level, and specific medical conditions including HTN, DM, dyslipidemia, cerebrovascular disease, CAD, and heart failure. Higher education was defined as obtaining a college degree or higher. Type and duration of physical activity was assessed using a detailed questionnaire and quantified using estimated daily metabolic equivalent task scores. Blood pressure was measured by a trained examiner using a mercury sphygmomanometer placed at heart level. Measurements were taken at least twice with participants seated, and the results were averaged. If a blood pressure difference of 5 mmHg or more was observed between the two measurements, a third measurement was taken and the last two measurements were averaged. WC was measured at the midpoint between the lowest rib and the iliac crest at the end of expiration in the standing position.
Blood samples were collected after an overnight fast and lipid profiles, hemoglobin A1c levels, white blood cell counts, and CRP levels were measured using an automated analyzer. The clinical laboratory evaluation was performed at a single core clinical laboratory that is accredited and participates in annual inspections and surveys by the Korean Clinical Laboratory Quality Assurance Association. Blood concentrations of glucose, total cholesterol, high-density lipoprotein (HDL) cholesterol, and triglycerides were measured using enzymatic methods (ADVIA 1650 and ADVIA 1800; Siemens Healthineers). Low-density lipoprotein (LDL) cholesterol levels were calculated using the Friedewald equation 25. CRP was measured using nephelometry (ADVIA 1650 and ADVIA 1800; Siemens Healthineers). HTN was defined as a physician’s diagnosis of HTN or regular use of antihypertensive drugs. DM was defined as a diagnosis of DM by a physician, regular use of antidiabetic drugs, or hemoglobin A1c level ≥6.5%. Dyslipidemia was defined as a diagnosis of dyslipidemia by a physician, regular use of statins without a history of cardiovascular disease or DM, or the presence of at least one of the following abnormal laboratory test results: total cholesterol level; ≥ 240 mg/dL, triglyceride level ≥ 150 mg/dL, or HDL cholesterol level < 45 mg/dL.
Evaluation of health status through self-assessment
Participants were asked to rate their overall health by answering the question, “How do you generally perceive your health?” Responses were initially categorized into five categories: “very poor,” “poor,” “fair,” “good,” and “very good.” Due to the relatively small number of participants who answered “very bad” (n = 411) and “very good” (n = 154), we reclassified the responses into three categories: “Poor (very bad/poor)” SRH (n = 3,380), “fair” SRH (n = 3,521), and “good (good/very good)” SRH (n = 2,994).
Standard 12-lead ECG and AF identification
A standard 12-lead ECG (GE Marquette MAC 5000®, GE Marquette Inc., Milwaukee, WI, USA) was performed on all participants at baseline and at each follow-up visit. All ECG traces were recorded at a paper speed of 25 mm/s and an amplitude of 0.1 mV/mm. Results were interpreted by a cardiologist and coded according to the Minnesota code classification system. Atrial fibrillation was identified by either the presence of atrial fibrillation on a 12-lead electrocardiogram or a history of self-reported atrial fibrillation using a physician-administered questionnaire before the baseline visit or during a follow-up visit. I did. Minnesota codes 8-3-1, 8-3-2, 8-3-3, and 8-3-4 were used to classify AF. Newly occurring AF was defined as the first confirmed AF between visits. The date of new AF onset was defined as the date when AF was first detected on an electrocardiogram or diagnosed by a physician.
statistical analysis
Baseline characteristics of participants were compared between groups. One-way analysis of variance was used to analyze continuous variables such as BMI, WC, and LDL cholesterol levels, and Pearson’s chi-square test was used to analyze categorical variables such as gender, comorbidities, and smoking history. Multiple comparisons were performed using post hoc analysis with Bonferroni correction. For continuous variables with skewed distributions, comparisons between groups were performed using the Kruskal-Wallis test. The Shapiro-Wilk test was used to assess the normality of the distribution of continuous variables.
All variables had approximately 1% missing values. The range of missing values for individual variables was 0–4.9%. Among the covariates in the multivariate regression model, missing values for individual variables ranged from 0 to 1.04% (Supplementary Figure S1). Rather than excluding cases with missing data, multiple imputation was performed using a bootstrap expectation maximization algorithm 26 . Five possible imputation datasets were generated. We used the mean value for continuous variables and the mode value for categorical variables.
The association between SRH and AF development was determined by age, gender, place of residence, education, BMI, physical activity, comorbidities (HTN, DM, heart failure, dyslipidemia, MI, non-MI CAD, asthma, chronic lung disease), Smoking status, alcohol intake, and laboratory data (including LDL cholesterol and CRP levels). Factors associated with the “bad” SRH group were analyzed using logistic regression analysis. We developed a predictive model for new-onset AF when combined with traditional risk factors with and without SRH variables using a multivariate CPH model to determine whether they provide predictive value. The goodness of fit of the predictive model was estimated using Harrell’s C index and Akaike’s information criterion (AIC), and Harrell’s C index was compared using the method proposed by Haibe-Kains et al. 27. The linear predictor identified using the multivariate CPH model was considered significant if the difference between the two AIC values was greater than 10. It is considered.
To assess the impact of reclassifying SRH responses from five to three categories based on our findings, we used Kaplan-Meier survival analysis and a multivariate CPH model using the original five categories of SRH responses. We performed a sensitivity analysis. Due to the small number of “very good” responses, “good” responses were used as a reference when calculating the coefficients of the multivariate CPH model.
All statistical analyzes were performed using the statistical software R-4.3.2 (R Core Team, R Foundation for Statistical Computing, Vienna, Austria) and its packages “tableone”, “rms”, “Amelia”, “survival”, “BiocManager”. was performed using. “survcomp” from RStudio-2023.12.1, Build 402 (RStudio Team, RStudio Inc., Boston, MA, USA). AP values of <0.05 were considered significant.
