Significance of Circulating Remnant Lipoprotein Cholesterol Levels Measured by Homogeneous Assay in Patients with Type 2 Diabetes

Remnant lipoproteins (RLs), which are typically present at high concentrations in patients with type 2 diabetes mellitus (T2DM), are associated with cardiovascular disease (CVD). Although an RL cholesterol homogeneous assay (RemL-C) is available for the measurement of RL concentrations, there have been no studies of the relationship between RemL-C and clinical parameters in T2DM. Therefore, we evaluated the relationships between RemL-C and CVD-related parameters in patients with T2DM. We performed a cross-sectional study of 169 patients with T2DM who were hospitalized at Kumamoto University Hospital. Compared with those with low RemL-C, those with higher RemL-C had higher fasting plasma glucose, homeostasis model assessment for insulin resistance (HOMA-R), total cholesterol, triglyceride, small dense low-density lipoprotein cholesterol (sdLDL-C), and urinary albumin-creatinine ratio; and lower high-density lipoprotein cholesterol, adiponectin, and ankle brachial pressure index (ABI). Multivariate logistic regression analysis showed that sdLDL-C and ABI were significantly and independently associated with high RemL-C. Although LDL-C was lower in participants with CVD, there was no difference in RemL-C between participants with or without CVD. Thus, RemL-C may represent a useful index of lipid and glucose metabolism, and that may be a marker of peripheral atherosclerotic disease (PAD) in male patients with T2DM.


Introduction
Remnant lipoproteins, which are present at high circulating concentrations when lipoprotein metabolism is impaired, are associated with the progression of atherosclerosis and coronary artery disease (CAD) [1,2]. It is well known that high serum triglyceride (TG) concentrations are an important risk factor for coronary heart disease [3][4][5]. However, it has been reported that the TG-rich remnant lipoprotein cholesterol concentration is closely associated with CAD risk in patients with normal TG concentrations [6]. Therefore, remnant lipoprotein cholesterol concentration may represent a more sensitive marker of CAD risk than TG.
High serum TG concentrations are frequently present in patients with diabetes and are thought to be important contributors to the onset of cardiovascular disease (CVD) in patients with diabetes. Indeed, the Framingham Heart Study revealed that diabetes is Table 1. Characteristics of the participants as a whole (n = 169) and of those with lower (Lo RemL-C (<0.24 mmol/L): n = 83) or higher (Hi RemL-C (≥0.24 mmol/L): n = 86) RemL-C concentrations.

Unit
All 1). Data from a total of 169 patients with T2DM (98 men and 71 women) were analyzed. The characteristics of the participants are listed in Table 1. All the participants were Japanese, and each participated in a detailed interview regarding his/her clinical condition and smoking history.   Hypertension was defined using the administration of any anti-hypertensive medication, a systolic blood pressure (SBP) ≥ 140 mmHg, or a diastolic blood pressure (DBP) ≥ 90 mmHg, according to the criteria of the Japanese Society of Hypertension [20]. Hyperlipidemia was defined as having total cholesterol (TC) > 5.7 mmol/L and/or a TG > 1.7 mmol/L or being under treatment with an antihyperlipidemic agent. CVD was defined as the presence or history of stroke, ischemic heart disease, or arteriosclerosis obliterans.

Compliance with Ethics Guidelines
All the procedures in the study involving human participants were performed in accordance with the ethics standards of the institutional and/or national research committee and the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. The study protocol was approved by the Human Ethics Review Committee of Kumamoto University (Protocol Number: 1468) and registered at UMIN-CTR (UMIN000010436). All the participants provided their written informed consent.

Laboratory Measurements
Blood samples were collected from all the participants in the early morning after an overnight fast. Enzymatic methods (Minaris Medical Co. Ltd., Tokyo, Japan) were used to measure the serum TC and TG concentrations. Homogeneous methods were used to measure the RemL-C, high-density lipoprotein cholesterol (HDL-C) (Minaris Medical Co. Ltd., Tokyo, Japan), and small dense LDL-C (sdLDL) (Denka Seiken Co. Ltd., Tokyo, Japan) concentrations. Immunoturbidimetric methods (Sekisui Medical Co. Ltd., Tokyo, Japan) were used to measure the Apo-AI, Apo-AII, Apo-B, Apo-CIII, and Apo-E concentrations. The enzymatic and immunoturbidimetric measurements were performed on a Hitachi 7180 Auto Analyzer (Hitachi High-Tech Corporation, Tokyo, Japan). Latex aggregation methods (Denka Seiken Co. Ltd., Tokyo, Japan) were used for the measurement of lipoprotein(a) (Lp(a)) and adiponectin concentrations on the Hitachi 7180 Auto Analyzer. An oxygen electrode method (Arkray, Inc., Kyoto, Japan) was used to measure the fasting plasma glucose (FPG) concentration on a GA1170. An ion exchange chromatography method (Tosoh Corporation, Tokyo, Japan) was used to quantify HbA1c on an HLC-732 G8. An electrochemiluminescence method (Roche Diagnostics K.K., Tokyo, Japan) was used to measure fasting plasma insulin (FPI) concentration on a Modular Analytics E module. A latex aggregation method (Denka Seiken Co. Ltd., Tokyo, Japan) was used to measure high-sensitivity C-reactive protein (hsCRP) concentration on a BM2250. The measurement of the interleukin-6 (IL-6) and tumor necrosis factor-α (TNF-α) concentrations was outsourced to SRL, Inc. (Tokyo, Japan). LDL-C concentration was measured using the Friedewald formula [21]. HbA1c (%) was calculated as National Glycohemoglobin Standardization Program (NGSP) equivalent values (%) using the formula HbA1c (%) = 1.02 × HbA1c [Japan Diabetes Society (JDS)] (%) + 0.25%, considering the HbA1c (JDS) (%) levels measured relative to the previous Japanese standard substance, the measurement method, and the HbA1c (NGSP) [22]. Estimated glomerular filtration rate (eGFR) was calculated using the age, sex, and serum creatinine concentration and the equation for Japanese people of Matsuo et al. [23]. The urinary albumin-creatinine ratio (ACR) was calculated using the urinary albumin and creatinine concentrations measured in early-morning fasting spot urine samples. Insulin resistance was assessed using the homeostasis model assessment for insulin resistance (HOMA-R), calculated using the fasting concentrations of insulin and glucose using the following formula: HOMA-R = FPI (µU/mL) × FPG (mmol/L)/22.5 [24].

Carotid Ultrasonography and the Measurement of Ankle Brachial Pressure Index (ABI) and Brachial-Ankle Pulse Wave Velocity (baPWV)
The intima-media thickness (IMT) of the common carotid artery (CCA-IMT) was measured using an ultrasound machine (Shimadzu SDU-2200, Shimadzu Co., Ltd., Kyoto, Japan) and a transducer frequency of 5-10 MHz, which provides an axial resolution of 0.30 mm. Intima Scope software (Media Cross Co. Ltd., Tokyo, Japan) was used for computer-assisted acquisition, processing, the storage of B-mode images, and the calculation of CCA-IMT. CCA-IMT was defined and measured as previously described [25]. The mean CCA-IMT and the maximum value of CCA-IMT (max CCA-IMT) were recorded.
ABI and baPWV were measured automatically using an ABI-form (BP-203RPE II; Nippon Colin, Komaki, Japan). The lower value of ABI and the higher values of baPWV recorded for the left and the right sides were used in data analysis.

Statistical Analyses
All the statistical analyses were performed using JMP software version 13.0 (SAS Institute, Cary, NC, USA). Values are presented as mean ± SEM or as numbers. Categorical data were compared using the chi-square test. Multivariate analysis of the relationships of low RemL-C (<0.24 mmol/L) and high RemL-C (≥0.24 mmol/L) with other parameters was performed using logistic regression. The relationships between RemL-C as a continuous variable and other parameters were evaluated using Pearson's correlation coefficients, and in the multivariate analysis, stepwise linear regression was performed. Because some of the variables showed skewed distributions (duration of diabetes, HOMA-R, TG, hs-CRP, RemL-C, Lp(a), adiponectin, TNF-α, IL-6, eGFR, and ACR), these data were logarithmically transformed before this analysis. p values < 0.05 were considered to be statistically significant.

Characteristics of the Participants
The characteristics of the participants as a whole and of those with lower or higher RemL-C concentrations are shown in Table 1. Compared with participants with low RemL-C, those with higher RemL-C had higher TC, TG, and sdLDL-C and lower HDL-C (Table 1). Besides lipid metabolism, higher RemL-C had higher FPG and HOMA-R and lower adiponectin, suggesting that RemL-C was affected by obese-related glucose metabolism (Table 1). Interestingly, higher RemL-C had higher ACR and lower ABI, suggesting the relevance between RemL-C and diabetic nephropathy and/or peripheral arterial disease in T2DM. Regarding drug therapy, high RemL-C concentrations were present in significantly fewer participants being treated with a thiazolidinedione or a statin (Table 1).

Relationships between RemL-C and Clinical Parameters in the Participants with T2DM
Next, we investigated the relationships between a continuous variable of RemL-C and clinical parameters in the participants with T2DM using Pearson's correlation coefficients. In relation to lipid metabolism, univariate analysis revealed that RemL-C positively correlated with TC, TG, LDL-C, and sdLDL-C and negatively correlated with HDL-C (Table 2). Regarding glucose metabolism, RemL-C positively correlated with FPG, FPI, HOMA-R, and HbA1c and negatively correlated with adiponectin in participants with T2DM (Table 2). To clarify whether the RemL-C measurements reflected the serum concentrations of remnant lipoproteins, the correlations between RemL-C and apolipoproteins were evaluated in participants with T2DM. Univariate analysis revealed that RemL-C positively correlated with Apo-B, Apo-CII, Apo-CIII, and Apo-E, which were the major apolipoproteins in remnant lipoprotein particles, and negatively correlated with Apo-AI (Table 3). Multivariate stepwise regression analysis revealed that Apo-AI, Apo-B, Apo-CIII, and Apo-E were significantly and independently associated with RemL-C ( Table 3), suggesting that RemL-C measurements reflected the serum concentrations of remnant lipoproteins.

Relationships between High RemL-C and Clinical Parameters in T2DM
Next, to evaluate the independent risk factors associated with high RemL-C, multivariate logistic regression analysis was performed in the participants with T2DM. For multivariate analysis, HOMA-R, sdLDL-C, eGFR, ACR, adiponectin, and ABI were selected as relevant factors from the results of univariate analysis ( Table 1). As a result, sdLDL-C and ABI were significantly and independently associated with high RemL-C concentration (Table 4). The inclusion of thiazolidine use and statin use in the multivariate logistic regression analysis had no effect on the results (data not shown).  RemL-C, remnant lipoprotein cholesterol; HOMA-R, homeostasis model assessment for insulin resistance; sdLDL-C, small dense LDL cholesterol; eGFR, estimated glomerular filtration rate; ACR, uremic albumin creatinine ratio; ABI, ankle brachial pressure index. Factors included in the logistic regression analysis for lower or higher level of RemL-C were ABI, HOMA-R, sdLDL-C, adiponectin, ACR and eGFR. Estimates are for the following log odds: high RemL-C (≥0.24 mmol/L)/low RemL-C (<0.24 mmol/L).

Effect of Sex on the Relationships between RemL-C and Various Clinical Parameters
Next, to determine the effect of sex, the relationships between RemL-C and various clinical parameters were evaluated separately in participants of each sex. In relation to lipid metabolism, univariate analysis revealed that RemL-C positively correlated with TC, TG, LDL-C, and sdLDL-C and negatively correlated with HDL-C (Table 5). Besides lipid metabolism, higher RemL-C had higher HOMA-R and lower ABI (Table 5). For multivariate logistic regression analysis, HOMA-R, sdLDL-C, age, adiponectin, and ABI were selected as relevant factors from the results of univariate analysis ( Table 5). As a result, sdLDL-C and ABI were significantly and independently associated with high RemL-C concentration in the men, as well as in the entire group of participants with T2DM (Table 5). Interestingly, in the female participants, univariate analysis revealed that age, FPG, TG, HDL-C, LDL-C, sdLDL-C, and adiponectin were associated with high RemL-C concentration ( Table 6). For multivariate logistic regression analysis, ABI, HOMA-R, sdLDL-C, adiponectin, and age were selected as relevant factors from the results of univariate analysis ( Table 6). As a result, sdLDL-C alone was significantly and independently associated with high RemL-C concentration in the women (Table 6). hsCRP, high-sensitive C reactive protein; eGFR, estimated glomerular filtration rate; ACR, uremic albumin creatinine ratio; TNF-α, tumor necrosis factor-α; IL-6, interleukin-6; baPWV, brachial ankle pulse wave velocity; ABI, ankle brachial pressure index; CCA-IMT, intima-media thickness of common carotid artery; CVD, cardiovascular disease. Logistic regression analyses included HOMA-R, age, sdLDL-C, adiponectin and ABI. Estimates are for the following log odds: high RemL-C (≥0.24 mmol/L)/low RemL-C (<0.24 mmol/L).

Differences in Several Parameters between Participants with or without a History of CVD in T2DM
Finally, the differences in parameters in participants with T2DM who did or did not have a history of CVD were evaluated. Age, the duration of T2DM, FPI, Lp(a), mean CCA-IMT, and max CCA-IMT, which were thought to be risk factors for CVD, were significantly higher in participants with CVD than in those without CVD (Table 7). Moreover, eGFR, which was a marker of chronic kidney disease, and ABI, which was a marker of peripheral atherosclerotic disease (PAD), were significantly lower in participants with CVD (Table 7). However, TC and LDL-C, which were strongly risk factors for CVD, were oppositely and significantly lower in participants with CVD (Table 7), suggesting the effect of statin use. Moreover, there were no differences in other risk factors for CVD, such as RemL-C, sdLDL-Biomolecules 2023, 13, 468 9 of 15 C, HDL-C, TG, SBP, DBP, BMI, or HOMA-R between these groups (Table 7), suggesting the existence of residual risk for CVD. hsCRP, high-sensitive C reactive protein; eGFR, estimated glomerular filtration rate; ACR, uremic albumin creatinine ratio; TNF-α, tumor necrosis factor-α; IL-6, interleukin-6; baPWV, brachial ankle pulse wave velocity; ABI, ankle brachial pressure index; CCA-IMT, intima-media thickness of common carotid artery; CVD, cardiovascular disease. Logistic regression analyses included ABI, HOMA-R, sdLDL-C, adiponectin, and age. Estimates are for the following log odds: high RemL-C (≥0.24 mmol/L)/low RemL-C (<0.24 mmol/L).  Dur., duration; BMI, body mass index; SBP, systolic blood pressure; DBP, diastolic blood pressure; FPG, fasting plasma glucose; FPI, fasting plasma insulin; HOMA-R, homeostasis model assessment for insulin resistance; TC, total cholesterol; TG, triglyceride; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; sdLDL-C, small dense LDL cholesterol; RemL-C, remnant lipoprotein cholesterol; Lp(a), lipoprotein(a); hsCRP, high-sensitive C reactive protein; eGFR, estimated glomerular filtration rate; ACR, uremic albumin creatinine ratio; TNF-α, tumor necrosis factor-α; IL-6, interleukin-6; baPWV, brachial ankle pulse wave velocity; ABI, ankle brachial pressure index; meanCCA-IMT, mean intima-media thickness of common carotid artery; maxCCA-IMT, maximum intima-media thickness of common carotid artery; CVD, cardiovascular disease.

Discussion
To the best of our knowledge, this is the first study to investigate the relationships between RemL-C and clinical parameters related to CVD risk in patients with T2DM. In this observational study, we found that RemL-C is associated with FPG, HOMA-R, TC, TG, HDL-C, LDL-C, sdLDL-C, ACR, adiponectin, and ABI. Logistic regression analysis showed that sdLDL-C and ABI are independently associated with a high RemL-C concentration and that the relationship with ABI is male-specific. However, we found no association between a history of CVD and RemL-C.
There are two clinically available methods for the measurement of remnant lipoprotein cholesterol concentrations. The first, for RLP-C, is an immunoaffinity separation method that is convenient because it does not require ultracentrifugation [12]. The second method, the RemL-C assay, can be performed by homogeneous method [17]. Currently, these two measurements are selected by each investigator. Yoshida et al. previously reported a significant correlation between RemL-C and RLP-C in non-diabetic individuals [26]. In the present study, RemL-C was found to significantly positively correlate with Apo-B, Apo-CII, Apo-CIII, and Apo-E, which are the major apolipoproteins in remnant lipoprotein particles. Therefore, RemL-C likely reflects the circulating remnant lipoprotein particle concentration, even in patients with T2DM.
DM is closely associated with CVD. Conventional cardiometabolic risk factors, such as hypertension, dyslipidemia, and obesity, which are common in T2DM, synergistically increase the risk of CVD. In addition, many previous studies have shown that T2DM is usually characterized by genetic predisposition, hyperglycemia, obesity, and insulin resistance. In particular, obesity and the associated insulin resistance are recognized as potent risk factors for CVD [27]. There is a tendency for hypertriglyceridemia to be present in patients with T2DM, because of the high concentrations of TG-rich lipoproteins, including TG-rich remnant lipoproteins, VLDL remnants, and chylomicron remnants. These high concentrations of TG-rich lipoproteins are thought to increase CVD risk directly and by causing a decrease in HDL and an increase in sdLDL [27]. Consistent with this, in the present study, we found that RemL-C positively correlates with sdLDL-C and negatively correlates with HDL-C. However, whereas RemL-C and RLP-C have previously been reported to be associated with CVD [14][15][16]18], the present findings show no relationship between RemL-C and a history of CVD in patients with T2DM, probably because of therapeutic interventions. Furthermore, LDL-C, a potent risk factor for CVD, showed a significant negative correlation with a history of CVD in the present cohort. However, the lack of a relationship between RemL-C and CVD history, while LDL-C did show an association, may indicate inadequate management of RemL-C. It has been reported that statin therapy significantly reduces the serum RemL-C concentration of hypercholesterolemic patients [28]. Moreover, the Diabetes Atherosclerosis Intervention Study (DAIS) revealed that treatment with fenofibrate reduces serum RLP-C in patients with T2DM [29]. Therefore, it is necessary to raise awareness of the fact that RemL-C represents a residual risk factor for CVD and to implement therapeutic interventions targeting remnant lipoproteins, the efficacy of which can easily be monitored using RemL-C.
Previous studies have shown that remnant lipoproteins are associated with several clinical parameters related to CVD. de Graaf et al. reported that RLP-C positively correlates with age, BMI, SBP, DBP, TC, TG, LDL-C, Apo-B, serum glucose, FPI, and HOMA-R and negatively correlates with HDL-C in patients without combined hyperlipidemia [30]. In addition, Saeed et al. reported that RLP-C is positively associated with TC, TG, LDL-C, and sdLDL-C and negatively associated with HDL-C and Lp(a) in middle-aged adults [31]. In the Framingham Heart offspring study, RLP-C was found to positively correlate with age, BMI, SBP, DBP, serum glucose, TC, TG, and LDL-C and negatively correlate with HDL-C in women [32]. Although there have been no previous equivalent studies using RemL-C, in the present study, we have shown that RemL-C positively correlates with FPG, FPI, HOMA-R, HbA1c, TC, TG, LDL-C, and sdLDL-C and negatively correlates with HDL-C and adiponectin in patients with T2DM. In addition, when two groups were created on the basis of RemL-C, high RemL-C was found to correlate with FPG, HOMA-R, TC, TG, HDL-C, sdLDL-C, ACR, adiponectin, and ABI in patients with T2DM. Furthermore, sdLDL-C was found to be independently associated with high RemL-C in patients with T2DM. Because the clinical parameters associated with RemL-C are nearly identical to those associated with RLP-C and because RemL-C has been reported to be present at high concentration in patients with CAD [18], RemL-C, as well as RLP-C, may be a useful index in patients with T2DM who are at high risk of CVD.
A relationship between subclinical arteriosclerosis and RemL-C has also been previously reported. Taguchi et al. reported that RemL-C concentration positively correlates with max-IMT and mean-IMT in healthy women [33]. In the present study, RemL-C tended to positively correlate with max-IMT and mean-IMT (χ 2 = 3.140, p = 0.076 and χ 2 = 2.805, p = 0.094, respectively) in women with T2DM. Interestingly, multivariate logistic regression analysis indicated that ABI, which is a marker of PAD, is significantly and independently associated with high RemL-C in men with T2DM, suggesting that there may be a sex difference in the effect of remnant lipoproteins on the progression of arteriosclerosis. Further study is needed to clarify this.
Apo-AI, which is a major protein component of HDL-C, is thought to closely and inversely correlate with the risk of atherosclerotic disease [34]. Karagiannidis et al. reported that higher levels of Apo-AI are independent predictors of lower CAD complexity in T2DM [35]. In the present study, we have shown that a high Apo-AI concentration is an independent predictor of low RemL-C. These findings may indicate that RemL-C is linked to the complexity of CAD. In addition, we have shown that a high Apo-B concentration is an independent predictor of high RemL-C. Given that an association between Apo-B and CVD has previously been reported (35), a relationship between RemL-C and CVD may be revealed by future prospective studies.
In recent years, the relationship between TG-rich remnant lipoprotein and chronic kidney disease has been attracting attention. Sonoda et al. reported that the RemL-C concentration positively correlates with the ACR and inversely correlates with eGFR [36]. In contrast, we found that RemL-C did not significantly correlate with eGFR (χ 2 = 0.20, p = 0.887). The reason for the discrepancy is unknown, but it may be explained by differences in patient background. RemL-C concentration was found to significantly positively correlate with ACR (χ 2 = 4.311, p = 0.038) in the present study, as well as the previous one. Relevant to this, Jun et al. demonstrated in a meta-analysis that fibrates reduce the risk of progression of albuminuria by 14% (p = 0.028) [37]. Moreover, Tsunoda et al. reported that fenofibrate reduces the concentration of RLP-C in patients with T2DM [29]. Thus, it is possible that RemL-C is associated with albuminuria and that a therapeutic intervention aimed at reducing RemL-C would help to prevent the progression of diabetic nephropathy in patients with T2DM.
Diabetes is associated not only with lipid metabolic disorders but also with higher concentrations of modified lipoproteins that are not routinely measured in clinical practice. Recently, the role of nitrated lipoproteins in the development of cardiovascular dysfunction in patients with diabetes has been discussed [38]. Nitrotyrosine (NT-Tyr) is a product of tyrosine modification by peroxynitrite, a potent prooxidant produced by the interaction of superoxide anions with nitric oxide [38]. The NT-Tyr concentration has been shown to be significantly higher in patients with T2DM than in non-diabetic individuals, and histopathological studies carried out to date on sections of arterial wall have shown that the NT-Tyr concentration is higher in patients with poor cardiovascular status [38]. Lipoproteins can undergo myeloperoxidase-catalyzed enzymatic nitration, and the reaction involves the apolipoproteins Apo-AI in HDL particles and Apo-B in LDL particles [38]. Because Apo-B is also present in remnant lipoprotein particles, the relationship between nitrated remnant lipoproteins and CVD risk is being clarified, and the establishment of a method for the measurement of the nitrated remnant lipoprotein concentration may be expected in the future.
The present study had several limitations. First, it was a single-center cross-sectional study with a relatively small number of participants. Thus, further large-scale prospective studies are needed to confirm the associations between RemL-C and clinical parameters in patients with T2DM. Second, because a direct comparison between RemL-C and RLP-C was not possible because of a lack of RLP-C measurements, it is difficult to discuss the factors affecting the relationships of RemL-C and RLP-C with CVD. However, Yoshida et al. reported that there is generally a significant correlation between RemL-C and RLP-C in individuals without diabetes or hypertension and not taking medications for hyperlipidemia, diabetes, or hypertension (r = 0.853, p < 0.0001) [26]. Moreover, they demonstrated that the RemL-C assay is likely to reflect the intermediate-density lipoprotein concentration more closely than RLP-C [26]. Therefore, the benefits of RemL-C measurement should be investigated further. Third, in the present study, other factors related to CVD, such as dyspnea classification and the heart failure status, were not available. It is a major limitation of the study. Further studies should address these issues in the future.

Conclusions
In the present study, we have demonstrated in T2DM that (i) RemL-C is associated with apolipoproteins related with remnant lipoproteins, (ii) high RemL-C concentration is associated with several factors related to lipid and glucose metabolism, (iii) sdLDL-C are independently associated with a high RemL-C concentration in both female and male subjects, (iv) ABI are independently associated with a high RemL-C concentration in male subjects, and (v) there was no difference in the RemL-C concentration between individuals with or without a history of CVD, possibly due to statin use. Therefore, RemL-C concentration may be a useful index of lipid and glucose metabolism, and that may be a marker of PAD in male subjects with T2DM. Regarding RemL-C as a risk factor for CVD, further prospective observational studies are needed in the future.