ABC | Volume 114, Nº3, March 2020

Original Article Kurmus et al. Discordance of lipoproteins and CAD severity Arq Bras Cardiol. 2020; 114(3):469-475 missing values for any lipid measurements, 6 had systemic inflammatory disease, renal or hepatic failure, hypo/ hyperthyrodism or malignancy, and 301 had previous history of coronary revascularization. Finally, we included the data of 574 patients in our analysis. The clinical parameters assessed were age, gender and coronary risk factors. Hypertension was defined as systolic blood pressure ≥ 140 mmHg and/or diastolic blood pressure ≥ 90 mmHg and/or current medication with antihypertensive drugs. Patients were defined as diabetic if they had been informed of this diagnosis prior to the study and had been using oral antidiabetic drugs or insulin treatment upon study admission. Body mass index (BMI) was calculated as body weight in kilograms divided by the squared height in meters (kg/m 2 ). Angiographic Evaluation Baseline diagnostic angiograms of the patients were assessed independently by two experienced interventional cardiologists who were blinded to patients’ lipid paremeters. SYNTAX score for each patient was calculated by scoring all coronary lesions producing ≥ 50% diameter stenosis in vessels ≥ 1.5 mm using the SYNTAX score algorithm, which was available on the SYNTAX website. Gensini score was calculated by assigning a severity score to each coronary narrowing on the basis of the degree of luminal stenosis and its location. 9 Decreases in luminal diameter of 25%, 50%, 75%, 90%, 99%, and total occlusion were given scores of 1, 2, 4, 8, 16 and 32, respectively. The score was then multiplied by a factor symbolizing the functional significance of the myocardial area supplied by that segment, that is, 5 for the left main artery, 2.5 for the proximal left anterior descending artery or proximal circumflex artery, 1.5 for the mid left anterior descending artery, 1 for the distal left anterior descending artery, right coronary artery and obtuse marginal artery, and 0.5 for all other areas. In cases of disagreement regarding SYNTAX or Gensini scores, an additional observer was consulted and the final decision was made by consensus. A low SYNTAX score was defined as ≤ 22, while intermediate and high SYNTAX scores were set as > 22. 10 Patients with Gensini score of ≥20 were defined as severe CAD, which was approximately equal to having a 70% or more stenosis in the proximal left anterior descending artery. 11 Laboratory Measurements Lipid measurements were performed on fasting blood samples taken before the angiography. Plasma concentrations of total cholestrol, LDL-C, HDL-C and triglycerides were measured with a Clinical Biochemistry Analyzer (Abbott Architect c 8000). The enzymatic colorimetric method was used for quantitative determination of total cholestrol. The endpoint colorimetric method was used for quantitative determination of HDL-C. LDL-C was measured by quantitative colorimetric method. The glycerol phosphate oxidase method was used for quantitative determination of triglycerides level. Non-HDL-C was calculated as total cholestrol minus HDL-C. Statistical analysis Categorical variables were defined as numbers and percentages. The distribution of continuous variables was considered as normal or not based on the Kolmogorov‑Smirnov test. Unless specified otherwise, continuous data were described as mean ± standard deviation for normal distributions, and median (interquartile range) for skewed distributions. First, we determined the medians for LDL-C and non-HDL-C to examine the discordance between them. We categorized patients into groups according to less than, greater than or equal to median levels of LDL-C and non-HDL-C. As there is no standard cutoff point for discordance, we chose the median to define discordance and to make it easier to apply to our study population. Discordance was defined as LDL-C greater than or equal to the median, and non-HDL-C as less than median; or LDL-C less than median and non-HDL-C greater than or equal to median. Concordant groups were defined as both LDL-C and non-HDL-C greater than or equal to median, or both LDL-C and non-HDL-C less than median. Differences between baseline characteristics of patients across these categories were analyzed with the chi-square test for comparing categorical variables and the One-way ANOVA for comparing means of continuous measures. The LSD test was used for binary comparisons. Pearson’s correlation analysis was used to examine the correlation between continuous variables, including LDL-C, non-HDL-C, Gensini and SYNTAX scores in the sample. Spearman’s correlation was used to examine correlations between these parameters in concordant and discordant groups. Data analyses were performed on SPSS for Windows, version 22.0 (SPSS Inc., Chicago, IL, United States). A p value < 0.05 was accepted as statistically significant. Results Themean age of the study populationwas 61.1±11.4 years, and 57.5% of the 574 patients were males. Baseline characteristics are presented in Table 1. Nearly 50% of the patients had hypertension, 30% had diabetes mellitus, 32% had past smoking history, and one third of the patients were on statin treatment. Mean LDL-C was 117.4 ± 38.3 mg/dl and non-HDL-C was 156.7 ± 46.8 mg/dl. Mean difference between non-HDL-C and LDL-C was 39.2±23.6 mg/dl. Patients with high difference between non-HDL-C and LDL-C were more commonly females, receiving less statin therapy and having more diabetes mellitus and high triglycerides levels. Mean Gensini score was 25.3 ± 39.6, and the median was 12 (0-191); mean SYNTAX score was 7.1 ± 11.2, and the median was 4 (0-53). LDL-C levels were strongly and positively correlatedwith non- HDL-C levels (r = 0.865, p < 0.001), but there was discordance between them. Discordance of LDL and non‑HDL-C was found in 15%of patients. Themagnitude of discordance and distribution of LDL-C and non-HDL-C levels according tomedians are shown in Figure 1. Non-HDL-C was correlated with triglyceride (TG) (r = 0.431, p < 0.001). Gensini score was highly correlated with SYNTAX score (r = 0.927, p < 0.001). Neither Gensini nor SYNTAX were correlated with LDL-C (p = 0.9 and p = 0.9, respectively). Also, both scores were not correlated with non‑HDL-C (p = 0.4 and p = 0.4, respectively). To further evaluate the characteristics of patients with discordance and concordance of LDL-C and non-HDL-C, we classifiedpatients into4 subgroups. Group1: LDL-C<median and 470