Relation between low-density lipoprotein cholesterol/apolipoprotein B ratio and triglyceride-rich lipoproteins in patients with coronary artery disease and type 2 diabetes mellitus: a cross-sectional study

This study was designed as a hospital-based cross-sectional study to investigate the relationship between the LDL-C/apoB ratio and TRL-related markers including TG, VLDL, RLP-C, apoC-II, and apo C-III, in CAD patients with type-2 diabetes. In addition, we examined the relationships among the changes in the LDL-C/apoB ratio and changes in the TRL-related markers in those cases that were still available for additional measurements 6 months later. This study is the sub-analysis of our previous study, which showed LDL-particle size and TG-metabolism disorder using cross-sectional method [10].

The study was conducted on a sample of 700 consecutive outpatients with the presence of one or more risk factors for CAD, who had undergone regular examinations for treatment of their various diseases at the Cardiovascular Center, Nihon University Surugadai Hospital between April 2009 and October 2009. This study was conducted in accordance with the ethical principles of the Declaration of Helsinki. The study protocol was approved by the Institutional Review Board at our institute and written informed consent was obtained from all the study participants.

The criterion for patient registration in the cross sectional study was the presence of one or more risk factors for CAD. The diagnostic criteria for the coronary risk factors used in this study were as follows; hypertension: systolic pressure of ≥ 140 mmHg and/or a diastolic pressure of ≥ 90 mmHg, or taking antihypertensive medication; diabetes mellitus: fasting plasma glucose ≥ 126 mg/dL and HbA1c ≥ 6.5%, or current treatment with anti-diabetic agents; lipid disorder: serum LDL-C ≥ 140 mg/dL, serum TG ≥ 150 mg/dL, and/or serum high-density lipoprotein cholesterol (HDL-C) less than 40 mg/dL, or current treatment with lipid-modifying medication; Cigarette smoking was defined as current smoking or smoking cessation within 1 year prior to the start of the study; chronic kidney disease: Estimated glomerular filtration rate (eGFR) < 60 ml/min/1.73 m²; obesity: body mass index ≥ 25 kg/m². A diagnosis of hyperuricemia was made when the serum uric acid level was 7.0 mg/dL or above, or taking medications.

Angiographically, CAD was defined as a history of documented myocardial infarction, prior coronary revascularization intervention (coronary artery bypass graft surgery or percutaneous coronary intervention), or the presence of ≥ 50% stenosis in 1 or more of the coronary arteries identified during cardiac catheterization.

Patients were not enrolled if they met any of the following exclusion criteria: hepatic dysfunction (alanine aminotransferase and aspartate aminotransferase ≥ 2 times the upper limit of the normal values), known malignant disease, or diagnosis of acute coronary syndrome within 3 months prior to the study.

Fasting blood samples were collected early in the morning after a 12-h fast. The serum total cholesterol (TC), HDL-C, and TG levels were measured by the standard methods. The serum LDL-C level was estimated by using the Friedewald formula [11]. The VLDL fraction was measured by performing polyacrylamide-gel electrophoresis using the LipoPhor system (Joko, Tokyo, Japan). The RLP-C level was measured by an immunoadsorption assay (SRL Inc., Tokyo, Japan). The serum apolipoproteins were determined by turbidimetric latex agglutination assays (SRL). The malondialdehyde-modified LDL (MDA-LDL) level was measured by an enzyme-linked immunosorbent assay (SRL). The high sensitivity C-reactive protein (hs-CRP) level was measured by a nephelometric assay (Behring Diagnostic Marburg, Germany). The eGFR was calculated by using the abbreviated MDRD (Modification of Diet in Renal Disease) study formula modified by a Japanese coefficient [12].

Data are expressed as the mean ± standard deviation (SD) for continuous variables and as percentages for discreet variables. Data that did not have a normal distribution were expressed as medians (interquartile range). The data for categorical variables were analyzed by the χ² test. For the subset analysis of four groups according to the presence or absence of DM and CAD, we used analysis of variance (ANOVA) followed by Bonferroni’s correction for covariates if differences were found in the patient characteristics or laboratory profile markers. In this study, an LDL-C/apoB ratio of 1.2 was deemed to correspond to an LDL particle size of 25.5 nm (the cutoff sd-LDL particle size), consistent with previous reports [7‐9], and an LDL-C/apoB ratio of < 1.2 served as a dependent variable for the multivariate logistic regression analysis described below. A detailed multi-logistic analysis of each group was performed in which the presence or absence of CAD and DM was evaluated in patients with an LDL-C/apoB ratio of less than 1.2. Accordingly, multi-logistic regression analyses were performed with no adjustments (model 1), after adjustment for age and gender (model 2), and after adjustment for traditional coronary risk factors and concomitant use of drugs with an action that increases LDL-particle size (e.g., statins, fibrates, and glitazones [13]) (model 3). These analyses were used to evaluate the association between LDL-C/apoB ratios of less than 1.2 and the prevalence of CAD or DM. Furthermore, univariate and multivariate regression analyses were performed to identify independent variables of LDL-C/apoB ratio. As the TRLs-related markers constituted mutually confounding factors, five multivariate regression models incorporating the respective variables were established to carry out the analyses. All variables correlated with LDL-C/apoB ratio at p < 0.05 in the univariate regression analysis were entered into the multivariate model. For the cases that were still available for additional measurements 6 months later, a multi-logistic regression analysis was performed to identify variables that were significantly associated with the changes in the LDL-C/apoB ratio. Increase/decrease of the LDL-C/apoB ratio from the baseline was entered as a dependent variable, and the age, gender, CAD risk factors, and the absolute changes of the TRL-related markers were entered as independent variables. We used SPSS Window ver 12.0 (Statistical Package for the Social Sciences, SPSS Ins., Chicago, IL) for all analyses.

We excluded 16 subjects from the study because of missing laboratory data. A final total of 684 subjects were included in the study. The participants consisted of 470 (59%) male and 214 (41%) female patients. The patients were classified into the four groups according to the presence or absence of CAD and/or DM. Comparison of the four groups according to the presence or absence of CAD and/or DM is shown in Tables 1, 2, and 3.

Table 3 shows the results of analysis from another viewpoint to provide a clearer understanding of the features of the lipid profiles for the four categories of patients shown in Table 2. The serum non-HDL-C level was significantly lower in the CAD (+) group than in the CAD (−) group, probably reflecting the statin treatment that is given to many patients of the CAD (+) group. The VLDL fraction was significantly higher in the CAD (+) group than in the CAD (−) group, probably reflecting a higher percentage of patients potentially having abnormal TG metabolism in the CAD (+) group, although the serum TG level did not differ significantly between the two groups. On the other hand, comparison between the DM (+) group and DM (−) group revealed a significantly greater number of patients with high levels of TRL-related markers in the DM (+) group than in the DM (−) group. Furthermore, the LDL-C/apoB ratio was significantly lower in the CAD (+) group than in the CAD (−) group, and TG/HDL-C ratio was higher in the CAD (+) group than in the CAD (−) group, although this difference was not statistically significant. Comparison of the DM (+) and DM (−) groups revealed a significantly lower LDL-C/apoB ratio and significantly higher TG/HDL-C ratio in the DM (+) group than in the DM (−) group.

The LDL-C/apoB ratios of the entire group of patients ranged from 0.622 to 1.694 (mean ± SD: 1.223 ± 0.146, and the LDL-C/apoB ratio ranges according to group were: CAD (+) DM (+) group, 0.622–1.408 (1.153 ± 0.133); CAD (+) DM (−) group, 0.868–1.458 (1.171 ± 0.129); CAD (−) DM (+) group, 0.779–1.525 (1.203 ± 0.124); and CAD (−) DM (−) group, 0.692–1.694 (1.250 ± 0.150)). There were significant differences in the LDL-C/apoB ratios among the 4 groups (p < 0.0001) (Fig. 1).

Multi-logistic regression analyses with no adjustments (model 1), after adjustments for age and gender (model 2), and after adjustments for coronary risk factors and concomitant use of drugs (model 3) were performed to evaluate the association between an LDL-C/apoB ratio of < 1.2 and the prevalence of CAD or DM. The analysis with adjustments for traditional coronary risk factors and concomitant drug use revealed that the CAD (+) DM (+) group was the only group exhibiting a significant and independent variable for a LDL-C/apoB ratio of less than 1.2, both in the overall cohort (Fig. 2), and in the subgroup of patients with serum LDL-C levels of less than 100 mg/dL (data not shown).

All of the variables that correlated with the LDL-C/apoB ratios at p < 0.05 in the univariate regression analysis were entered into the 5 multivariate models. The results of the analyses of all of the multivariate regression models showed that serum TRLs-related markers were significant variables that were independent of LDL-C/apoB ratios. Next, similar analyses were performed in patients with serum LDL-C levels of < 100 mg/dL, patients with serum LDL-C levels of < 100 mg/dL, and patients taking/not taking lipid-modifying treatments. All the univariate and multivariate regression analyses showed that the high levels of TRL-related markers were independent determinants of a low LDL-C/apoB ratio (Table 4). Table 5 shows the correlations between LDL-C/apoB ratio and TRL-related markers among the 4 groups. LDL-C/apoB ratio and serum RLP-C level in the CAD (+) DM (−) group was the only correlation not showing statistical significant; on the other hand, statistically significantly negative correlations were noted between the LDL-C/apoB ratio and all TRL-related markers was noted.

In this cross-sectional study, we confirmed that increased levels of TRL-related markers were associated with a decrease of the LDL-C/apoB ratio. Therefore, we investigated, using the longitudinal method, the relationship between the absolute changes (∆) in the serum TG levels and the ∆ LDL-C/apoB ratio, in order to examine the causal relationship. During a follow up period of at least 6 months, multivariable logistic regression analysis conducted in the 445 patients who were followed up for at least 6 months after adjustments for age, gender and risk factors for CAD revealed that higher ∆ serum TG was an independent predictor of a decreased LDL-C/apoB ratio. Next, similar analyses were performed in patients with serum LDL-C levels of < 100 mg/dL, and patients taking/not taking lipid-modifying treatments. Statistical analyses revealed similar findings (Table 6). Similarly, all the multi-logistic regression analyses showed that higher ∆ values of other TRL-related markers were independent determinants of a decreased ∆ LDL-C/apoB ratio (data not shown).

First, the relationships bewteen the LDL-C/apoB ratio and TRL-related markers were analyzed by dividing the patients according to the history (positive/negative) of intake of lipid-modifying drug treatment and good serum LDL-C control, because lipid modifying drugs have an effect of improving the LDL and TG metabolism. The possibility of the effects of lipid modifying drugs influencing the results of the present study cannot be excluded. It had also been reported that the influence on the LDL-particle size varies among the different types of statins [23]. Furthermore, the duration of treatment involving such drugs could not be ascertained in the present study. Second, in theory the LDL-C/apoB ratio is a marker of a patient’s mean LDL-particle diameter, but it does not indicate the exact LDL-particle diameter, which is measured using density gradient ultracentrifugation and nuclear magnetic resonance spectroscopy. Moreover, the significance of calculating the absolute LDL-C/apoB ratio cut-off value for CAD risk has not been determined. Third, no patients who were taking dipeptidyl peptidase (DPP)-4 inhibitors or sodium-glucose transporter (SGLT) 2 inhibitors were included among the subjects of this study. DPP-4 inhibitors and SGLT2 inhibitors have triglyceride lowing actions, and it would be very interesting to evaluate these actions comparatively [24, 25]. Fourth, because determination of the presence of CAD in this study population relied on the findings of coronary angiography, the existence of subjects in the study population of undetected cases of asymptomatic CAD which can be diagnosed primarily by an exercise stress test or non-invasive tests such as coronary artery computed tomography cannot be ruled out. Diabetic patients often have asymptomatic myocardial ischemia, and a particularly high prevalence of asymptomatic myocardial ischemia has been reported in diabetic patients with CAD and abnormal TG metabolism [26]. Finally, in the future, an interventional study to investigate the causal relationship is needed.