Study population
The study population was from the coronary artery disease database of West China Hospital, Sichuan University. In brief, West China Hospital CAD database is a large prospective registry study designed to explore risk factors, early warnings, risk stratification and management of patients with CAD (ChiCTR-OOC-17010433). The database was created in July 2008 and is ongoing to enroll all patients who have undergone invasive coronary angiography at West China Hospital. The information collected includes demographic data, cardiovascular risk factors, laboratory data, ultrasound indicators, angiographic results, medication, revascularization and clinical outcomes. Informed consent was given to all enrolled patients in this study, and ethical approval was obtained from the local institution. The primary inclusion criteria for this study were angiographically proven of stenosis greater than 50 percent in at least one major epicardial coronary arteries. The exclusion criteria included patients who had contraindications to coronary angiography, patients who did not receive echocardiographic examination during hospitalization, patients who were lost during follow-up, patients who died within the first month of discharge, and patients who were followed up for less than 1 month.
Measurement of major exposure factors
The main exposure factors of interest in this study were percentage of body fat and left ventricular mass index. A formula based on body mass index, sex and age (Clínica Universidad de Navarra-Body Adiposity Estimator) was used to calculate body fat percentage [8]. Height and weight are measured by a trained nurse at admission. BMI was the ratio of body weight to the square of height. The study defined higher body fat as the percentage of body fat greater than the 75th percentile of included male or female patients. During the hospital stay, the patient underwent a comprehensive transthoracic echocardiographic examination in accordance with guideline recommendations. From the parasternal long axis view, the left ventricular diameter was measured using M-type or 2D guidance. Left ventricular volume and left ventricular ejection fraction were measured using the two-plane Simpsons method. The left ventricular myocardial mass was calculated using the formula recommended by the American Society of Echocardiography, and it was normalized using the height of 2.7 to obtain the left ventricular myocardial mass index [9]. According to the guideline definition, LVMI greater than 44 in women and 48 in men is defined as left ventricular hypertrophy. According to body fat percentage and LVMI, patients were divided into four groups: group 1, lower body fat and no LVH; group 2, lower body fat and LVH; group 3, higher body fat and no LVH; group 4, higher body fat and LVH.
Data on potential confounders and modifiers
Demographic characteristics, cardiovascular risk factors, comorbidities, and medical history were collected by a questionnaire interview at admission or search in medical record. Data of blood pressure, heart rate, laboratory data, angiographic results, medications, and revascularization therapy were obtained from medical records.
Follow-up and study endpoint
The endpoint of the study was all-cause death. In the follow-up of this study, telephone follow-up, review of medical records, and outpatient visits were mainly adopted. The events were confirmed by death certificates or close relatives. Each patient was followed up from the discharge until either death occurrence or the last follow-up, depending on which came first.
Statistical analysis
Continous data were presented as mean ±SD or median with interquartile range, and categorical data were presented as number and percentage. One-way analysis of variance and x2 test were conducted to compare the baseline characteristics of study participants. We used Cox proportional risk models to examine the associations between body fat percentage, left ventricular hypertrophy and all-cause mortality. Different models were used to look at the effect of covariates on prognosis. Model 1 is the result of no correction. Model 2 is adjusted for age and sex. Model 3 was adjusted for age, sex, LVEF, and creatinine. Model 4 corrects age, sex, LVEF, creatinine, hypertension, diabetes, current smoking, cholestrol, and left main or three-vessel disease. All analyses were performed using SPSS. P value < 0.05 on both sides indicated statistical difference.