Study population

The “Génétique et Environnement en Europe du Sud” (GENES) study is a case-control study designed to assess the role of genetic, biological and environmental determinants in the occurrence of CAD. The study protocol was conducted according to the principles of the Declaration of Helsinki. The protocol was endorsed by the Scientific Council of the Toulouse University Hospital, and was approved by the “Comité Consultatif pour la Protection des Personnes se Prêtant à la Recherche Biomédicale” (Advisory Committee for the Protection of Persons Involved in Medical Investigation, Comité Toulouse/Sud-Ouest #1) (file number 1-99-48). Biological sample collection was declared to the French Ministry of Research and to the Regional Health Agency under the number DC-2008-403 #1. Information was provided about the objectives of the study, and participants provided written informed consent.

As previously described , cases were stable male CAD patients living in the Toulouse area (South-Western France), aged between 45 and 74 years and recruited from 2001 to 2004 after admission to the Cardiology Department at Toulouse University Hospital. The patients were admitted for cardiovascular examination, in the context of evaluation and management of their CAD. Stable CAD was defined as a history of acute coronary syndrome, history of coronary artery revascularization, documented myocardial ischaemia, stable angina, or the presence of a coronary stenosis ≥ 50% on a coronary angiogram. Patients who had presented a recent acute coronary episode were considered as not stable, and were not included in the study. Stratification into decadal age groups was used to approximately match age distribution between controls and cases. Patients underwent medical examination in the same health centre, and during the same period; this included clinical and anthropometric measurements, and completion of a questionnaire.

Data collection

Age, socioeconomic variables and information on cardiovascular risk factors were collected through standardized face-to-face interviews, performed by a single physician. Dyslipidaemia was defined as treatment with drugs or fasting serum total cholesterol ≥ 2.40 g/L. Hypertension was defined as treatment with drugs or systolic blood pressure ≥ 160 mmHg or diastolic blood pressure ≥ 95 mmHg. Diabetes was defined as treatment with drugs or fasting blood glucose ≥ 7.8 mmol/L. In terms of smoking status, patients were classified as current smokers, former smokers who had not smoked tobacco for > 3 months or non-smokers. Alcohol consumption was assessed using a typical week pattern. The total amount of alcohol consumed was calculated as the sum of different types of drinks, allowing categorization into three levels:

• •
  

  abstainers;

  
  
• •
  

  moderate alcohol consumption (1–39 g/day);

  
  
• •
  

  heavy consumption (≥ 40 g/day).

Physical activity was investigated through a standardized questionnaire, and categorized into three levels:

• •
  

  no physical activity;

  
  
• •
  

  moderate physical activity for 20 minutes, no more once a week;

  
  
• •
  

  intense physical activity for ≥ 20 minutes each session, at least twice a week.

Presence of dyslipidaemia, diabetes or hypertension was assessed from the subjects’ current treatments. Inpatients’ medications at discharge were also considered. Anthropometric measurements included waist circumference, height and body weight, and body mass index was calculated (kg/m ² ). Blood pressure and resting heart rate were measured after ≥ 5 minutes of rest, using an automatic sphygmomanometer; two measurements were performed and average values were recorded. Insulin resistance was estimated by the homeostasis model assessment of insulin resistance (HOMA-IR). Ankle-brachial index was evaluated as described previously . Lower limb blood pressure was determined from the right and left posterior tibial arteries, with the patient in supine position. When the posterior tibial artery blood pressure was not measurable, the dorsalis pedis artery was used. Systolic blood pressure was detected with a hand-held Doppler probe. The ankle-brachial index was calculated for each lower limb by dividing the ankle systolic blood pressure by the average of the two measurements performed on the arm. For each patient, the lowest ankle-brachial index recorded in the two ankles was kept for further analysis. An ankle-brachial index < 0.9 was considered abnormal.

Biological measurements

Blood was collected after an overnight fast. Blood glucose, triglycerides, total cholesterol and HDL cholesterol were assayed with enzymatic reagents in an automated analyser (Hitachi 912; Roche Diagnostics, Rotkreuz, Switzerland). LDL cholesterol was calculated using the Friedwald formula. C-reactive protein and gamma-glutamyl transferase (GGT) were also analysed with an automated analyser (Hitachi 912). Insulin measurements were taken using an immunometric assay (Advia Centaur; Siemens, Munich, Germany). Apolipoprotein AI, apolipoprotein B and lipoprotein(a) were determined by first order precipitation in an automated analyser (Hitachi 912). Lipoprotein AI was also determined by electroimmunoassay (Sebia, Evry, France).

Coronary angiography and assessment of left ventricular function (LVF)

Coronary angiography was performed in all patients at the time of inclusion. The severity of coronary atherosclerosis was assessed by five different scores:

• •
  

  number of significant narrowed vessels (≥ 50% of luminal narrowing in 0, 1, 2 or 3 vessels);

  
  
• •
  

  Gensini score ;

  
  
• •
  

  Duke Jeopardy score ;

  
  
• •
  

  severity score;

  
  
• •
  

  spreading score .

Left ventricular ejection fraction (LVEF) was assessed by contrast ventriculography or by the isotopic method and echocardiography.

HDL subclass analysis

Serum samples ( n = 403) were stored at-80 °C after centrifugation, and were analysed immediately after thawing. HDL subfractions were analysed by an electrophoretic method on non-denaturing, linear, polyacrylamide gel, using the Lipoprint ^® HDL system (Quantimetrix Corp., Redondo Beach, CA, USA), according to the manufacturer’s instructions. Briefly, 25 μL of serum sample were added to 3% polyacrylamide gel tubes, along with 300 μL of a loading gel solution containing Sudan Black as a lipophilic dye. After 30 minutes of photopolymerization at room temperature, electrophoresis was performed for 50 minutes at 3 mA/gel tube. Each electrophoresis chamber included a quality control, which was provided by the manufacturer (Liposure ^® Serum Lipoprotein Control; Quantimetrix Corp.). For quantification, scanning was performed with an ArtixScan M1 digital scanner (Microtek International Inc., Hsinchu, Taiwan). Stained HDL subfractions were identified by the electropheretic mobility of each band. The LDL/very-low-density lipoprotein (VLDL) band was the starting reference point, and albumin was the ending reference point. The percentage of area under the curve was calculated with Lipoware computer software (Quantimetrix Corp.). Ten HDL subfractions were distributed between the LDL/VLDL and albumin bands: HDL1-HDL3 were defined as large HDLs; HDL4-HDL7 were defined as intermediate HDLs; and HDL8-HDL10 comprised the small HDLs. The cholesterol concentration of each HDL subclass was determined by multiplying the relative area under the curve of each subfraction by the HDL cholesterol concentration of the sample.

Follow-up procedure

The closing date was 31 December 2013 for all patients. Follow-up was 100% complete, and was recorded annually for assessment of overall and cardiac mortality; other outcomes were not recorded. Mean follow-up was 9.8 ± 3.1 years.

Statistical analyses

In descriptive tables, data are presented as percentages for qualitative variables and means with standard deviations for quantitative variables. The χ ² test was used to compare the distribution of categorical variables (or Fisher’s exact test, when necessary). The mean values of continuous variables were compared by Student’s t test. The Shapiro-Wilks and Levene tests were used to test the normality of distribution of residuals and the homogeneity of variances, respectively. When basic assumptions of Student’s t test were not satisfied, data were transformed logarithmically or subjected to a Wilcoxon-Mann-Whitney test. Cumulative survival of patients according to variables of interest was determined by the Kaplan-Meier method, and compared using the log-rank test for the individual endpoint of death. The relationship between baseline variables and outcome criteria was assessed using Cox proportional hazards regression analysis. Firstly, a basic multivariable model excluding lipid variables was built. All variables associated with a P value < 0.05 in univariate analysis were introduced into a multivariable Cox model. A backward procedure was applied to assess variables that were significantly and independently associated with mortality. Secondly, relationships between blood lipids and mortality were tested in a univariate analysis, and then in a multivariable analysis, by adjusting for variables of the basic model. We tested the proportionality assumption using cumulative sums of martingale-based residuals. Interactions were tested in the final model, and goodness-of-fit of the final model was assessed using martingale residuals. Analyses were two-tailed and P < 0.05 was considered to be significant. Computation was carried out with SAS software, version 9.4 (SAS Institute, Cary, IL, USA) and Stata ^® 11.2 software (StataCorp, College Station, TX, USA).