A method for assessing risk of dyslipidemia using metabolomic pattern analysis

Abstract

The present invention relates to a method for assessing the risk of dyslipidemia using metabolite pattern analysis, and more specifically, to a method for assessing the risk of dyslipidemia comprising: (a) a step of screening the plasma metabolites of a subject; (b) a step of determining whether the subject's metabolic state is similar to the metabolic state of a healthy control group; (c) a step of determining whether the subject's metabolites have a significant correlation with the metabolites of a healthy control group; and (d) a step of determining the subject as having a high risk of dyslipidemia if the metabolic states of the subject and the healthy control group are not similar or if there is no significant correlation between the subject and the healthy control group's metabolites.

Description

Technology Field

[0001] The present invention relates to a method for assessing the risk of dyslipidemia using metabolite pattern analysis, and more specifically, to a method for assessing the risk of dyslipidemia comprising: (a) a step of screening the plasma metabolites of a subject; (b) a step of determining whether the subject's metabolic state is similar to the metabolic state of a healthy control group; (c) a step of determining whether the subject's metabolites have a significant correlation with the metabolites of a healthy control group; and (d) a step of determining the subject as having a high risk of dyslipidemia if the metabolic states of the subject and the healthy control group are not similar or if there is no significant correlation between the subject and the healthy control group's metabolites. Background Technology

[0003] Dyslipidemia is very important to manage because it is associated with various metabolic complications.

[0004] Dyslipidemia refers to a lipid disorder defined by an abnormal blood lipid profile, such as elevated levels of triglycerides (TG), total cholesterol (TC), or low-density lipoprotein cholesterol (LDL-c); decreased levels of high-density lipoprotein cholesterol (HDL-c); or a combination of these characteristics.

[0005] In dyslipidemia, metabolic disorders caused by circulating suboptimal lipids alter the microenvironment of adipocytes and hepatocytes, leading to insulin resistance and the secretion of pro-inflammatory cytokines. In other words, dyslipidemia can trigger the pathogenesis of other metabolic diseases such as cardiovascular disease (CVD), diabetes, and hypertension.

[0006] Currently, 40% of adults in Korea have dyslipidemia, with hypercholesterolemia showing the highest rate. In 2022, the incidence of dyslipidemia did not change significantly compared to the past, but the incidence of high LDL cholesterol increased, while low HDL cholesterol decreased. In addition, about half of all diabetic patients had elevated LDL-c (≥ 100 mg / dL), and 29% of hypertensive patients showed significantly high LDL-c (≥ 130 mg / dL).

[0007] Meanwhile, there is some evidence supporting an association between lifestyle and dyslipidemia. Among these, dietary factors have a significant impact on patients with dyslipidemia. In fact, treatment guidelines emphasize healthy eating habits as a first-line therapy for patients with dyslipidemia.

[0008] However, more research is needed to develop biomarkers, kits, and dietary guidelines that can identify the characteristics of dyslipidemia and distinguish it. Prior art literature

[0010] Korean Published Patent No. 10-2023-0101468 Korean Published Patent No. 10-2023-0001659 The problem to be solved

[0011] The present invention aims to solve the problems of the prior art and provides a method for assessing the risk of dyslipidemia that can determine the degree of risk of dyslipidemia by screening the plasma metabolites of a subject and comparing the metabolic status and metabolites of the subject and a healthy control group. means of solving the problem

[0013] To achieve the above objectives, the present invention comprises: (a) a step of screening the plasma metabolites of a subject;

[0014] (b) a step of determining whether the subject's metabolic status is similar to the metabolic status of a healthy control group;

[0015] (c) a step of determining whether the subject's metabolite is significantly related to the metabolite of a healthy control group; and

[0016] (d) a step of determining that the subject is at high risk of dyslipidemia when the metabolic status of the subject and the healthy control group are not similar or when there is no significant correlation between the metabolites of the subject and the healthy control group; the present invention provides a method for assessing the risk of dyslipidemia comprising: (d) a step of determining that the subject is at high risk of dyslipidemia when the metabolic status of the subject and the healthy control group are not similar or when there is no significant correlation between the metabolites of the subject and the healthy control group.

[0017] In one embodiment of the present invention, step (b) is characterized by confirming the similarity of the metabolic patterns of the subject and the healthy control group.

[0018] In one embodiment of the present invention, step (c) is characterized in that the metabolites are 3-hydroxybutyrylcarnitine, 2-octinal, 1,3,5-heptatriene, and 5β-cholanic acid.

[0019] In one embodiment of the present invention, step (d) is characterized by determining that the subject has a low risk of dyslipidemia when the metabolic status of the subject and the healthy control group are similar and the metabolites of the subject and the healthy control group are significantly related.

[0020] In one embodiment of the present invention, subjects judged to have a high risk of dyslipidemia are characterized by having a lower intake of dietary fiber than subjects judged to have a low risk of dyslipidemia.

[0021] In addition, the present invention comprises (a) a step of screening the plasma metabolites of a subject;

[0022] (b) a step of determining whether the subject's metabolic status is similar to the metabolic status of a healthy control group;

[0023] (c) A step of determining whether the subject's metabolite is significantly related to the metabolite of a healthy control group;

[0024] (d) a step of determining that the subject is at high risk of dyslipidemia when the metabolic status of the subject and the healthy control group is not similar or there is no significant correlation between the metabolites of the subject and the healthy control group; and

[0025] (e) a step of providing a certain amount of dietary fiber to subjects judged to be at high risk of dyslipidemia and having them consume it; a method for improving the risk of dyslipidemia is provided. Effects of the invention

[0027] The present invention can provide a method for assessing the risk of dyslipidemia by screening the plasma metabolites of a subject and comparing the metabolic status and metabolites of the subject and a healthy control group to determine the degree of risk of dyslipidemia. Brief explanation of the drawing

[0029] Figure 1 shows two groups reconstructed based on the similarity of metabolic patterns using K-means clustering. Figure 2 shows the difference in levels of four metabolites between the two groups. Figure 3 shows a heatmap visualizing the correlation of four metabolites. Figure 4 shows the difference in levels of four metabolites among the three groups. Specific details for implementing the invention

[0030] The present invention will be described in detail below based on the following examples. The terms, examples, etc. used in the present invention are merely illustrative to explain the invention more specifically and to aid the understanding of those skilled in the art, and the scope of the rights, etc., of the present invention should not be interpreted as being limited thereto.

[0031] Unless otherwise defined, technical and scientific terms used in this invention represent the meanings commonly understood by those skilled in the art to which this invention pertains.

[0033] The present invention comprises (a) a step of screening plasma metabolites of a subject;

[0034] (b) a step of determining whether the subject's metabolic status is similar to the metabolic status of a healthy control group;

[0035] (c) a step of determining whether the subject's metabolite is significantly related to the metabolite of a healthy control group; and

[0036] (d) a step of determining that the subject is at high risk of dyslipidemia when the metabolic status of the subject and the healthy control group are not similar or when there is no significant correlation between the metabolites of the subject and the healthy control group; the present invention relates to a method for assessing the risk of dyslipidemia, comprising: (d) a step of determining that the subject is at high risk of dyslipidemia when the metabolic status of the subject and the healthy control group are not similar or when there is no significant correlation between the metabolites of the subject and the healthy control

[0038] Step (b) above can confirm the similarity of metabolic patterns between the subject and the healthy control group.

[0040] Step (c) above is characterized in that the metabolites are 3-hydroxybutyrylcarnitine, 2-octinal, 1,3,5-heptatriene, and 5β-cholanic acid.

[0042] In step (d) above, if the metabolic status of the subject and the healthy control group are similar and the metabolites of the subject and the healthy control group are significantly related, the subject may be determined to have a low risk of dyslipidemia.

[0044] In addition, subjects judged to be at high risk of dyslipidemia are characterized by having a lower intake of dietary fiber than subjects judged to be at low risk of dyslipidemia.

[0046] In addition, the present invention comprises (a) a step of screening the plasma metabolites of a subject;

[0047] (b) a step of determining whether the subject's metabolic status is similar to the metabolic status of a healthy control group;

[0048] (c) A step of determining whether the subject's metabolite is significantly related to the metabolite of a healthy control group;

[0049] (d) a step of determining that the subject is at high risk of dyslipidemia when the metabolic status of the subject and the healthy control group is not similar or there is no significant correlation between the metabolites of the subject and the healthy control group; and

[0050] (e) a step of providing a certain amount of dietary fiber to subjects judged to be at high risk of dyslipidemia and having them consume it; the present invention relates to a method for improving the risk of dyslipidemia, comprising: (e) a step of providing a certain amount of dietary fiber to subjects judged to be at high risk of dyslipidemia and having them consume it.

[0052] The present invention will be described in detail below through examples and comparative examples. The following examples are merely illustrative for the implementation of the present invention, and the scope of the present invention is not limited by the following examples.

[0054] (Example 1) Selection of Subjects

[0055] Subjects were randomly selected from among research participants recruited by the Clinical Nutritional Genetics Laboratory at Yonsei University and the National Health Insurance Service at Ilsan Hospital in Goyang, South Korea.

[0056] The diagnostic criteria for dyslipidemia were based on the diagnostic criteria of the Korean Society of Lipid and Atherosclerosis.

[0057] Dyslipidemia was classified when TG ≥ 200 mg / dL, TC ≥ 240 mg / dL, LDL-c ≥ 160 mg / dL, HDL-c < 40 mg / dL (for men) or HDL-c < 50 mg / dL (for women), or when taking lipid-lowering agents.

[0058] Exclusion criteria included current diagnoses and a history of CVD, cancer, liver disease, kidney disease, immune diseases, and drug use.

[0059] All research participants provided written consent, and the Institutional Review Board of Hannam University approved a research protocol that complies with the Helsinki Declaration.

[0060] First, 100 subjects were randomly selected for each group based on age and gender [dyslipidemia (n = 100), healthy control group (n = 100)].

[0061] 1:1 propensity score matching was performed using age and sex, excluding specimens whose plasma status was not suitable for metabolite analysis. As a result, 80 subjects from each group [dyslipidemia (n = 80), control group (n = 80)] were included in the metabolite analysis.

[0063] (Example 2) Laboratory Evaluation

[0064] Body mass index (BMI) was calculated using body weight and height, and systolic blood pressure (SBP) and diastolic blood pressure (DBP) were measured using an automatic blood pressure monitor EASY X 800 (Jawon Medical, Gyeongsan-si, South Korea).

[0065] Venous blood samples and urine were collected through night fasting for at least 12 hours and stored at -80 and -20℃, respectively.

[0066] Glucose was measured using the hexokinase method with a glucose kit (Roche, Mannheim, Germany). Insulin was analyzed using an immunoradiological analysis kit (DIAsource ImmunoAssays SA, Louvain, Belgium).

[0067] The commercial kits used for biochemical variables are as follows.

[0068] Interleukin (IL)-1β, IL-6, and Tumor Necrosis Factor (TNF-α) are Bio-Plex Reagent Kit (Bio-Rad Laboratories, Hercules, California, USA);

[0069] TG is a commercial TG kit (Roche, Mannheim, Germany);

[0070] TC is the CHOL kit (Roche, Mannheim, Germany);

[0071] HDL-c is the HDL-c Plus Kit (Roche, Mannheim, Germany);

[0072] Urological 8-epi-prostaglandin F 2α (PGF 2α ) is a urine isoprostan ELISA kit (Oxford Biomedical Research Inc., Rochester Hills, MI, USA);

[0073] Malondialdehyde (MDA) is analyzed using the TBARS assay kit (ZeptoMetrix Co., Buffalo, NY, USA); and

[0074] Oxidized LDL is obtained using an enzyme immunoassay kit (Mercodia AB, Uppsala, Sweden).

[0075] In addition, apolipoprotein A1 and B levels were evaluated using the turbidity immunoassay method, and the reaction was analyzed using Cobas 6000 C501 (Roche, Mannheim, Germany).

[0076] LDL-c levels were calculated using the Friedewald formula.

[0078] (Example 3) Evaluation of Daily Energy Intake

[0079] The subject's daily energy intake was evaluated based on data obtained using the 24-hour recall method.

[0080] Total energy intake (kcal / day), proportions of major nutrients (carbohydrate, protein, fat), cholesterol (mg / day), and estimated fiber intake (g / day) were evaluated using CAN-pro 3.0 software (Korean Nutrition Society, Seoul, South Korea).

[0082] (Example 4) Non-targeted metabolites

[0083] ■ Sample Preparation

[0084] The prepared plasma samples were precipitated with cold acetonitrile (Wako Pure Chemical Industries, Osaka, Japan) (1:4, v / v) and centrifuged for 15 minutes (13,000 rpm, 4°C). After centrifugation, the supernatant was separated, and the solvent was evaporated using nitrogen. The obtained dried residue was dissolved in 200 µl of 10% methanol.

[0085] The redissolved samples were filtered through a 0.45-µm polyvinylidene difluoride syringe filter. Finally, for the quality control (QC) samples, all plasma samples were pooled and pretreated in the same manner as above. Also, L-Leucine-1- 13 C (Sigma-Aldrich, Saint Louis, MO, USA) was used as the internal standard (ISTD).

[0087] ■ Ultra-high Performance Liquid Chromatography-Mass Spectrometry (UHPLC-127 MS) Analysis

[0088] 10 µl of pre-treated plasma samples were injected into an Acquity UPLC-BEH-C18 column (Waters, Milford, MA, USA) connected to a Thermo UHPLC system (Ultimate 3000 BioRS; Dionex, Thermo Fisher Scientific, Bremen, Germany). Additionally, QC samples were analyzed periodically to monitor sensitivity and reproducibility and to evaluate within-analytical and between-analytical accuracy.

[0089] The column temperature was maintained at 4°C, and two mobile phase gradient systems [A, LC-MS grade 0.1% formic acid in water (Thermo Fisher Scientific, Fair Lawn, NJ, USA); B, LC-MS grade 0.1% formic acid in methanol (Thermo Fisher Scientific, Fair Lawn, NJ, USA)] separated compounds from plasma for 22 minutes. For MS, a Q Exactive Plus Orbitrap (Thermo Fisher Scientific, Waltham, MA, USA) was used, and positive electrospray ionization mode (ESI+) was performed at a spray voltage of 3.0 kV, nitrogen-coated gas flow rate of 60 (arbitrary), auxiliary gas flow rate of 20 (arbitrary), capillary temperature of 370°C, S-lens radiofrequency level of 45, and auxiliary gas heater temperature of 285°C. Full scan-ddms 2 The mode detected all substances with a mass-to-charge (m / z) scan range of 50–1,000.

[0091] ■ Identification of metabolites

[0092] Compound Discoverer 3.0 software (Thermo Fisher Scientific, San Jose, CA, USA) was used for raw spectrum processing. Spectral alignment and normalization were performed with reference to QC, and each region was corrected to ISTD. Finally, the processed spectra were verified based on online databases [ChemSpider (http: / / www.chemspider.com), LIPID MAPS (https: / / www.lipidmaps.org), mzCloud (https: / / www.mzcloud.org) and Kyoto Encyclopedia of Genes and Genomes (KEGG; https: / / www.genome.jp / kegg)].

[0094] ■ Statistical Analysis

[0095] All statistical analyses were performed using SPSS 26 (IBM Corp, Armonk, NY, USA) and R 4.1.3. Independent t-tests and Mann-Whitney U tests were used to evaluate differences in clinical and biochemical variables between the two groups. ANOVA was performed for the comparison of the three groups, and variable distributions were examined and skewed variables were logarithmically transformed. Chi-square tests were performed on nominal variables, and data were expressed as mean ± standard error (SE); a two-sided p-value of less than 0.05 was considered significant.

[0096] The metabolite data processed for multivariate analysis was exported from Compound Discoverer 3.0 to MetaboAnalyst 5.0 (http: / / www.metaboanalyst.ca).

[0097] After logarithmic transformation, k-means clustering (k = 2) was performed. K-means clustering is an unsupervised machine learning method that clusters a dataset into multiple groups based on similarity or distance.

[0098] In addition, the eXtreme Gradient Boosting (XGBoost) model was created to select meaningful metabolites, and XGBoost is a learning framework that can prevent overfitting and improve generalization performance.

[0099] The XGBoost model has the following parameters.

[0100] Estimator(n), 400; learning rate, 3; max depth, 3; min child weight, 6; eta, 0.05.

[0101] Pearson's correlation analysis evaluated the relationship between important variables and notable metabolites.

[0102] In addition, to determine the optimal fiber intake cutoff value for distinguishing metabolite patterns, receiver operating characteristic (ROC) curve analysis was performed and the Youden index was calculated.

[0104] Among the 80 subjects in each group [dyslipidemia (n = 80) vs. healthy control group (n = 80)], 7 subjects in the healthy control group were excluded from the final analysis because the metabolic analysis results showed outliers.

[0106] Table 1 shows a summary of the comparison of clinical and biochemical characteristics for the original classification of all subjects [dyslipidemia (n = 80) vs. control group (n = 73)].

[0107] In the lipid profile, TC, TG, LDL-c, oxidized LDL, and apolipoprotein B were significantly higher in the dyslipidemia group than in the control group (all p values ​​< 0.001), while HDL-c (p = 0.001) was lower than in the control group.

[0108] In addition, there were significant differences between the two groups in BMI (p < 0.001), DBP (p = 0.002), insulin (p < 0.001), and MDA (p = 0.003). Furthermore, these significant differences were maintained even after BMI adjustment.

[0110]

[0112] Non-targeted metabolite screening in positive mode detected 5,559 mass features and identified 354 metabolites by searching online databases.

[0113] The present invention reconstructed two groups based on the similarity of metabolic patterns across all subjects using K-means clustering (Fig. 1).

[0114] Interestingly, subjects with dyslipidemia are divided into two groups based on whether they have a metabolic state similar to that of healthy controls (n = 24) or not (n = 56). Subjects with dyslipidemia clustered with healthy controls are defined as Cluster B (n = 97). Additionally, subjects with dyslipidemia not clustered with healthy controls are classified as Cluster A (n = 56).

[0116] It can be confirmed in XGBoost that this new subset of the entire population contains four metabolites (3-hydroxybutyrylcarnitine, 2-octenal, 1,3,5-heptatriene and 5β-cholanic acid).

[0117] Figures 2 and 4 show the differences in levels of the four metabolites. A t-test was performed on the four metabolites, and all metabolites except 5β-cholanic acid (p = 0.068) showed significant differences between Cluster A and Cluster B at the p < 0.001 level.

[0118] Plasma concentrations of 3-hydroxybutyrylcarnitine, 2-octenal, and 5β-cholanic acid were significantly higher in cluster A than in cluster B, but 1,3,5-heptatriene was significantly lower in cluster A than in cluster B.

[0119] After adjusting for age and sex, significance remained at the same level, and 5β-cholanic acid also showed a significant difference (p = 0.015).

[0121] A comparison of the clinical and biochemical characteristics of the two newly created groups [Cluster A (n = 56) vs. Cluster B (n = 97)] based on the similarity of metabolite patterns is presented in Table 2.

[0123]

[0125] In the comparison between dyslipidemia subjects in Cluster A (n = 56) and Cluster B (n = 24), age (p = 0.005), sex (p = 0.001), TNF-α (p = 0.014), TC (p = 0.007), LDL-c (p = 0.004), apolipoprotein A1 (p = 0.039), apolipoprotein B (p < 0.001), and 8-epi-PGF 2α (p < 0.001) showed a significant difference.

[0126] Clusters A (n = 56) and B (n = 97) were sex (p = 0.025) and insulin levels (p = 0.018), TNF-α (p = 0.034), TC (p < 0.001), TG (p < 0.001), LDL-c (p < 0.001), apolipoprotein B (p = 0.047), and 8-epi-PGF 2α A significant difference was observed at (p = 0.001).

[0127] In Cluster B, the dyslipidemia (n = 24) group and the control group (n = 73) showed significant differences in age (p = 0.038), sex (p = 0.008), BMI (p < 0.001), SBP (p = 0.008), DBP (p = 0.001), insulin (p < 0.001), TC (p < 0.001), TG (p < 0.001), HDL-c (p = 0.002), LDL-c (p < 0.001), apolipoprotein A1 (p = 0.006), apolipoprotein B (p < 0.001), MDA (p < 0.001), and oxidized LDL (p < 0.001).

[0129] Among these variables, TNF-α and 8-epi-PGF 2α It showed the same trend as the metabolic pattern obtained in the present invention.

[0130] There was a significant difference between subjects with dyslipidemia in Cluster A and subjects with dyslipidemia in Cluster B, but there was no significant difference between subjects with dyslipidemia in Cluster B and the control group.

[0131] Unexpectedly, TNF-α and 8-epi-PGF 2α Significantly higher levels were observed in dyslipidemia subjects in Cluster B than in dyslipidemia subjects in Cluster A.

[0133] Table 3 shows the results of a comparison of expected energy and nutrient intake.

[0134] No significant differences were observed in the estimated intake of total energy, carbohydrates, protein, fat, and cholesterol between the two groups.

[0135] The only significant dietary component was dietary fiber. Higher dietary fiber intake was observed in dyslipidemia subjects in Cluster B than in Cluster A (p = 0.002). However, there was no significant difference in dietary fiber intake between dyslipidemia subjects in Cluster B and the control group.

[0137]

[0139] The present invention relates to TNF-α, 8-epi-PGF 2α Significant correlations between fiber intake and four metabolites (3-hydroxybutyrylcarnitine, 2-octenal, 1,3,5-heptatriene, and 5β-cholanic acid) were investigated.

[0140] TNF-α, 8-epi-PGF 2α Dietary fiber intake was included in the correlation analysis with four metabolites because there was a difference between the dyslipidemia subjects in Cluster A and the dyslipidemia subjects in Cluster B, but no difference between the dyslipidemia subjects in Cluster B and the control subjects in Cluster B.

[0141] Their relationship was visualized as a heatmap in Figure 3.

[0142] A positive correlation (r = 0.287, p = 0.029) was found between TNF-α and 1,3,5-heptatriene, and a negative correlation (r = -0.268, p = 0.042) was found between TNF-α and 2-octenal.

[0143] 8-epi-PGF 2α It showed a positive correlation with 1,3,5-heptatriene (r = 0.389, p = 0.002), but a negative correlation with 2-octenal (r = -0.449, p < 0.001), 3-hydroxybutyrylcarnitine (r = -0.419, p < 0.001), and 5β-cholanic acid (r = -0.300, p = 0.018).

[0144] In addition, dietary fiber intake showed a positive correlation with 1,3,5-heptatriene (r = 0.318, p = 0.004) and a negative correlation with 2-octenal (r = -0.295, p = 0.008), 3-hydroxybutyryl carnitine (r = -0.361, p = 0.001), and 5β-cholanic acid (r = -0.308, p = 0.005).

[0146] Based on the results of the logistic regression analysis, an ROC curve was generated, and the cutoff for dietary fiber intake was calculated using Youden's index.

[0147] The optimal cutoff with the highest accuracy for classifying metabolic patterns similar to the control group was 17.28 g / day for dyslipidemia. That is, subjects with dyslipidemia who consume more than 17.28 g / day of dietary fiber may exhibit metabolic patterns similar to those of healthy subjects.

[0149] The subjects of the dyslipidemia in this invention exhibited typical characteristics of dyslipidemia, and this invention applied metabolomics technology to explore the metabolic patterns of dyslipidemia patients.

[0150] As a result, two clusters were observed based on different metabolic patterns, and four metabolites (3-hydroxybutyrylcarnitine, 2-octenal, 1,3,5-heptatriene, and 5β-cholanic acid) contributed significantly to distinguishing the groups.

[0151] In addition, TNF-α, 8-epi-PGF among the newly formed dyslipidemia subgroups 2α There was a significant difference in dietary fiber intake.

[0152] Therefore, the present invention can be used as a marker and supporting data for metabolic guidelines for dyslipidemia. In particular, based on the cutoff obtained from regression analysis, a dietary fiber intake of 17.28 g / day or more can be recommended.

[0154] Individuals with dyslipidemia who consume more than 17.28g of dietary fiber per day may maintain a metabolic pattern similar to that of healthy people and may also experience significant changes in the levels of four metabolites (3-hydroxybutyrylcarnitine, 2-octenal, 1,3,5-heptatriene, 5β-cholanic acid).

Claims

Claim 1 (a) a step of screening the subject's plasma metabolites; (b) a step of determining whether the subject's metabolic status is similar to that of a healthy control group; (c) a step of determining whether the subject's metabolites differ significantly from those of a healthy control group; A method for assessing the risk of dyslipidemia comprising: (d) a step of determining that a subject has a high risk of dyslipidemia when the metabolic status of the subject and the healthy control group are not similar or there is a significant difference in the metabolites of the subject and the healthy control group; wherein step (b) divides into two groups based on the similarity of metabolic patterns using K-means clustering, and determines that the subject included in the group to which the healthy control group belongs has a metabolic status similar to the metabolic status of the healthy control group; step (c) the metabolites are 3-hydroxybutyrylcarnitine, 2-octenal, 1,3,5-heptatriene, and 1,3,5-heptatriene; and step (c) compares the levels of said metabolites of the subject and the healthy control group, and determines that there is a significant difference if the p-value of said levels is less than 0.

05. Claim 2 delete Claim 3 delete Claim 4 A method for assessing the risk of dyslipidemia according to claim 1, wherein step (d) is characterized by determining that the subject has a low risk of dyslipidemia when the metabolic status of the subject and the healthy control group are similar and there is no significant difference in the metabolites of the subject and the healthy control group. Claim 5 (a) a step of screening the subject's plasma metabolites; (b) a step of determining whether the subject's metabolic status is similar to that of a healthy control group; (c) a step of determining whether the subject's metabolites differ significantly from those of a healthy control group; (d) a step of determining the subject as having a high risk of dyslipidemia if the metabolic status of the subject and the healthy control group is not similar or if there is a significant difference between the subject and the healthy control group's metabolites; A method for improving the risk of dyslipidemia comprising: (e) a step of providing a certain amount of dietary fiber to subjects judged to have a high risk of dyslipidemia and having them consume it; wherein step (b) divides subjects into two groups based on similarity of metabolic patterns using K-means clustering, and determines that subjects included in the group to which the healthy control group belongs have a metabolic state similar to the metabolic state of the healthy control group; step (c) is characterized in that the metabolites are 3-hydroxybutyrylcarnitine, 2-octenal, 1,3,5-heptatriene, and 1,3,5-heptatriene, and step (c) compares the levels of the metabolites of the subjects and the healthy control group, and determines that there is a significant difference if the p-value of the level is less than 0.05.

Images