KR102377089B1 - Diagnostic test kit for diagnosing prediabetes and a method for diagnosing prediabetes - Google Patents

Abstract

The present invention relates to a prediabetes diagnostic kit and a diagnostic method using a serum metabolite, and more specifically, to a prediabetes-associated serum metabolite marker that specifically increases in the serum of pre-diabetes, that is, pre-diabetes patients. The present invention relates to a prediabetes diagnostic kit and diagnostic method using the serum metabolite marker.

Description

Prediabetes diagnostic kit and diagnostic method {Diagnostic test kit for diagnosing prediabetes and a method for diagnosing prediabetes}

The present invention relates to a prediabetes diagnostic kit and a diagnostic method using a serum metabolite, and more specifically, to a prediabetes-associated serum metabolite marker that specifically increases in the serum of pre-diabetes, that is, pre-diabetes patients. The present invention relates to a prediabetes diagnostic kit and diagnostic method using the serum metabolite marker.

Prediabetes is a stage before the diagnosis of diabetes, which can be said to be a potential defective state of glucose metabolism. Impaired fasting glucose (IFG), impaired glucose tolerance (IGT), increased HbA1c level, In addition, it is diagnosed by diabetes risk indicators such as insulin resistance, and although it is known as a time to slow the progression of diabetes or recover to normal, there is a difficulty in diagnosis because it does not meet the criteria for diabetes mellitus.

According to the statistics of the Korea Centers for Disease Control and Prevention in 2008, the prevalence of diabetes among Koreans over the age of 30 reached 9.1%, and the prevalence increased rapidly after the age of 40, reaching 20% in those in their 50s. In Korea, type 2 diabetes patients account for about 90-95%, and diabetes is rapidly increasing with the increase in living standards. has been reported Since diabetes can lead to very serious complications such as retinopathy, renal failure, and peripheral neuropathy, it is essential to accurately diagnose and manage it early. According to the American Diabetes Association's criteria for determining diabetes, glycated hemoglobin (HbAlc), FPG (Fast plasma gluclose value after fasting for at least 8 h), 2h-PG (75 g Oral Glucose Tolerance Test (OGTT)), plasma glucose level 2 hours after ), but recently, as a method for more accurate diagnosis, studies on serum metabolites that can serve as markers for diabetes have been extensively conducted.

Furthermore, since diabetes or complications (micro/macrovascular disease) often occur at the time prediabetes is diagnosed, the shortcomings of the blood glucose measurement method can be supplemented through metabolite research, and diabetes progression can be predicted and developed more accurately. The current challenge is to discover preventable metabolite markers. Serum metabolites are small molecules that represent biochemical changes in vivo caused by gene expression and environmental influences, and provide important information for disease diagnosis and understanding mechanisms. Therefore, a serum metabolite that can serve as a marker for prediabetes may more accurately diagnose the potential for diabetes or serve as a target for the development of a diabetes treatment. Under this background, recent studies on serum metabolites show that human serum metabolites are related to insulin resistance, type 2 diabetes, and glycemic load caused by prediabetes (Non-Patent Documents 1 to 4). However, there are still insufficient reports on reliable serum metabolites related to prediabetes.

Meanwhile, patent documents related to the diagnosis of diabetes have also been reported.

For example, in the case of Patent Publication No. 10-2017-0072650 disclosed in Patent Document 1, a diabetes diagnosis kit comprising a quantification device for a serum metabolite selected from the group consisting of C16, PC ae C36:0, and combinations thereof, and the above It provides information necessary for diagnosing diabetes or Korean diabetes using serum metabolites. However, here, since metabolite markers themselves are serum metabolite markers associated with type 2 diabetes, even if they are diagnosed using them, the purpose is to diagnose already advanced diabetes, not prediabetes.

In addition, in the case of Patent Publication No. 10-2016-0087035 presented in Patent Document 2, a diabetes diagnostic kit including a quantification device for peripheral blood mononuclear cell (PBMC) metabolites such as leucine and phenylalanine, and the serum metabolites It provides information necessary to diagnose diabetes or to diagnose diabetes in Koreans. However, here too, the statistical analysis method is different from the present invention, and the metabolite marker is also different from that for prediabetes diagnosis.

Furthermore, in the case of US Patent Publication No. US2010/0197028 A1 presented as Patent Document 3, a method for evaluating the state of diabetic conditions including determining the amount of a metabolite in a sample from a body fluid or tissue of an individual is disclosed, but this method shows that it can be used, for example, to diagnose and monitor insulin resistance, prediabetes, or responsiveness to drugs that alter diabetic conditions. Here too, metabolite markers have a drawback in that they are insufficient for prediabetes as a whole.

Patent Publication No. 10-2017-0072650 Patent Publication No. 10-2016-0087035 US Patent Publication US2010/0197028 A1

Zhang, X. et al., J. 9 Proteome. Res., 8(11),5188-95(2009) Huffman, K. M. et al., Diabetes Care., 32(9), 1678-83 (2009) Bain, J. R. Diabetes., 58 (11), 16 2429-43 (2009) Zhao, X. et al., J. Physiol. Endocrinol. Metab., 296(2), E384-93 (2009)

It is an object of the present invention to provide a diagnostic kit for prediabetes developed based on newly discovered metabolite markers specific for prediabetes through metabolite changes in serum.

Furthermore, it is to provide a method for providing information necessary for the diagnosis of prediabetes. The information obtained according to this method functions as information that can help early diagnosis of diabetes progress, and through this, it is possible to prevent diabetes and prevent complications such as various vascular diseases that are highly likely to develop in the pre-diabetic period. Furthermore, it can be usefully used for effective treatment of diabetes and the development of new drugs.

One aspect of the present invention is Alanine (Ala), Valine (Val), Gly (Glycine), Tyrosine (Tyr), Methionine (Met), lysoPhosphatidylcholine acyl C18:2 (lysoPC a C18:2), Hydroxysphingomyeline C22:2 (SM (OH) C22:2) Octadecenoylcarnitine (C18:1), Phosphatidylcholine diacyl C36:1 (PC aa C36:1), Phosphatidylcholine acyl-alkyl C30:0 (PC ae C30:0), Phosphatidylcholine acyl-alkyl C42:1 ( Provided is a prediabetes diagnostic kit comprising a device for quantifying serum metabolites selected from the group consisting of PC ae C42:1), Sphingomyeline C18:1 (SM C18:1), and combinations thereof.

Another aspect of the present invention provides that the kit of the present invention is a prediabetes diagnostic kit applied to preventing, predicting, or diagnosing the progression to diabetes.

Another aspect of the present invention provides that the kit of the present invention is a prediabetes diagnostic kit applied when the sample provider's body mass index is mild obesity (25kg/m ² ≤ BMI < 30 kg/m ² , AUC=0.58) do.

Another aspect of the present invention is that the kit of the present invention is a diabetes diagnosis kit further comprising a device for quantifying biochemical factors selected from triglycerides, glycated hemoglobin, fasting blood sugar, fasting insulin, and any combination thereof. to provide.

Another aspect of the present invention is diabetes in which the kit of the present invention further comprises a liquid chromatography (LC) and flow-injection analysis mass spectrometer (FIA-MS) as the quantitative device. A diagnostic kit is provided.

Further, an aspect of the present invention is

obtaining a sample comprising blood isolated from a human; and

Alanine (Ala), Valine (Val), Gly (Glycine), Tyrosine (Tyr), Methionine (Met), lysoPhosphatidylcholine acyl C18:2 (lysoPC a C18:2), Hydroxysphingomyeline C22:2 (SM (OH) in the sample C22:2) Octadecenoylcarnitine (C18:1), Phosphatidylcholine diacyl C36:1 (PC aa C36:1), Phosphatidylcholine acyl-alkyl C30:0 (PC ae C30:0), Phosphatidylcholine acyl-alkyl C42:1 (PC ae C42) :1), measuring the concentration of a serum metabolite selected from the group consisting of Sphingomyeline C18:1 (SM C18:1), and combinations thereof,

Provided is a method for providing information necessary for diagnosis of prediabetes, which is determined to be prediabetes when the measured concentration of the serum metabolite increases compared to that of the control sample.

Another aspect of the present invention is a method of providing information necessary for the diagnosis of prediabetes, wherein the method is applied when the human body mass index is mild obesity (25kg/m ² ≤ BMI < 30 kg/m ² , AUC=0.58) provide that

Another aspect of the present invention provides that the method is a method of providing information necessary for diagnosis of prediabetes, further comprising a method of providing prediabetes prediction information based on fasting blood sugar.

Another aspect of the present invention is that the method further comprises a method of providing prediabetes prediction information by a CRF model including prediabetes risk factors (age, sex, body mass index and blood lipid concentration, etc.) It provides a method for providing information necessary for the diagnosis of

Another aspect of the present invention is the method for providing prediabetes prediction information by fasting blood glucose and prediabetes risk factors ((age, sex, body mass index, blood lipid concentration, etc.) It provides a method of providing information necessary for the diagnosis of prediabetes, further comprising a method of providing diabetes prediction information.

The present invention is based on a metabolite marker specific to pre-diabetes discovered through metabolite changes in serum collected from a population cohort. A more accurate diagnosis of prediabetes is possible compared to the used diagnosis. Furthermore, since it is possible to diagnose the progression of diabetes at an early stage, it is possible to prevent diabetes and prevent complications such as various vascular diseases that are highly likely to develop in the pre-diabetic period. In addition, the present invention can be effectively used for effective treatment of diabetes and development of new drugs.

1 is a schematic diagram of participants as cohorts and experimental subjects. A follow-up cohort and a follow-up cohort through 6 years of follow-up (KARE AS6) using the Korean community base 3 (KARE AS3) as the baseline cohort. The number of each participant is indicated.
2 is a diagram showing the analysis results of metabolites according to the prediabetic phenotype in the base cohort ( p <4.07E-04).
3 is a graph showing the prediction rates for prediabetes between 12 prediabetes markers (KARE) and other predictive models.
4 is a graph showing the prediction rates for prediabetes when measured with 12 prediabetes markers (KARE) and fasting blood glucose in subgroups of age (A), gender (B), and body mass index (C).

Hereinafter, the present invention will be described in more detail.

All technical terms used in the present invention, unless otherwise defined, have the meaning as commonly understood by one of ordinary skill in the art of the present invention. In addition, although preferred methods and samples are described herein, similar or equivalent ones are also included in the scope of the present invention. The contents of all publications incorporated herein by reference are hereby incorporated by reference in their entirety.

As used herein, the term “serum metabolite” refers to a metabolite obtained from a liquid sample of blood origin. The liquid sample of blood origin is, for example, whole blood, plasma, or serum. In one embodiment, the liquid sample of blood origin is serum. The blood-derived liquid sample may be pretreated for detection of serum metabolites, and may include, for example, filtration, distillation, extraction, separation, concentration, inactivation of interfering components, addition of reagents, and the like. In addition, the serum metabolites may include substances produced by metabolism and metabolic processes, substances generated by chemical metabolism by biological enzymes and molecules, and the like.

As used herein, the term “diagnosis” refers to determining the susceptibility of an object to a specific disease or disorder, determining whether an object currently has a specific disease or disorder (eg, metabolic abnormalities or diabetes mellitus) identification), determining the prognosis of a subject afflicted with a particular disease or condition, or using therametrics (e.g., monitoring a subject's condition to provide information about the efficacy of treatment). do.

Hereinafter, the process of discovering the serum metabolite marker of the present invention will be described in detail.

The inventors of the present invention conducted various statistical analyzes ( statistical analysis) was used to discover metabolite markers associated with prediabetes in Koreans according to the stratified prediabetes phenotype (prediabetes, impaired fasting blood glucose, impaired glucose tolerance, increased glycated hemoglobin, and insulin resistance according to the Diabetes Association standards).

In addition, in the follow-up cohort (KARE cohort 6), using information on the trace metabolites of 500 people, the receiver-operating characteristic (ROC) analysis was performed, and the area under curve (AUC) was calculated. As a result, the predictive rate of prediabetes of these markers was verified through prediabetes risk factors age (age), sex (sex), body mass index (BMI), high-density lipoprotein (HDL), low-density lipoprotein (LDL), and triglyceride (TG). Prediction rates when metabolite markers previously reported in Westerners (Framingham Heart Study in US; FHS, KORA cohort study in Germany; KORA) were used as models were comparatively analyzed.

That is, statistical significance in the stratified prediabetes phenotype by performing regression analysis using age, gender, and body mass index as covariates for the participants in the base cohort (Bonferroni post-test, p <4.07E-04) 39 enriched metabolites showing Seven metabolites Alanine (Ala), Valine (Val), Gly (Glycine), Tyrosine (Tyr), Methionine (Met), lysoPhosphatidylcholine acyl C18:2 (lysoPC a C18:2), Hydroxysphingomyeline C22:2 (SM ( OH) C22:2) has been reported to be related to prediabetes in Westerners, but five metabolites, namely Octadecenoylcarnitine (C18:1), Phosphatidylcholine diacyl C36:1 (PC aa C36:1) ), Phosphatidylcholine acyl-alkyl C30:0 (PC ae C30:0), Phosphatidylcholine acyl-alkyl C42:1 (PC ae C42:1), and Sphingomyeline C18:1 (SM C18:1) are the first metabolites associated with prediabetes. Confirmed. In the follow-up cohort, it was confirmed that the predictive rate (AUC 0.71) of these metabolite markers was higher than that of the prediabetes risk factor (AUC 0.64) ( p < 0.01), and the previously reported US EPIC cohort model (AUC) in Westerners. 0.56) and Germany's KORA cohort model (AUC 0.64) showed higher prediction rates ( p <0.005). In addition, in the comparison of the predictive rate of prediabetes by metabolite markers and the predictive rate of fasting glucose in subdivision groups such as age (quartile), gender (female/male), and body mass index (normal/overweight/mild obesity), the difference according to age was However, when the body mass index was mild obesity (25 kg/m ² ≤ BMI < 30 kg/m ² , AUC=0.58), it was confirmed that the prediction rate (AUC=0.79) by the metabolite marker was higher ( p <0.001). Also, it was confirmed that the prediction rate was low when the gender was female (AUC=0.73 vs. 0.85, p<0.002) and when the body mass index was normal (BMI<23kg/m ² , AUC 0.68 vs. 0.78, p<0.01). .

Therefore, according to the above study results, serum metabolites Alanine (Ala), Valine (Val), Gly (Glycine), Tyrosine (Tyr), Methionine (Met), lysoPhosphatidylcholine acyl C18:2 (lysoPC a C18:2), Hydroxysphingomyeline C22 :2(SM(OH) C22:2), Octadecenoylcarnitine(C18:1), Phosphatidylcholine diacyl C36:1(PC aa C36:1), Phosphatidylcholine acyl-alkyl C30:0(PC ae C30:0), Phosphatidylcholine acyl- It was confirmed that alkyl C42:1 (PC ae C42:1) and sphingomyeline C18:1 (SM C18:1) are specific serum metabolites whose amounts change in the serum of the diabetic group compared to the normal group.

Accordingly, the present invention in one aspect,

Alanine (Ala), Valine (Val), Gly (Glycine), Tyrosine (Tyr), Methionine (Met), lysoPhosphatidylcholine acyl C18:2 (lysoPC a C18:2), Hydroxysphingomyeline C22:2 (SM (OH) C22:2 ) Octadecenoylcarnitine (C18:1), Phosphatidylcholine diacyl C36:1 (PC aa C36:1), Phosphatidylcholine acyl-alkyl C30:0 (PC ae C30:0), Phosphatidylcholine acyl-alkyl C42:1 (PC ae C42:1) , Sphingomyeline C18:1 (SM C18:1), and to provide a diabetes diagnostic kit comprising a quantification device for a serum metabolite selected from the group consisting of a combination thereof.

As used herein, the term “prediabetes” includes a condition in which blood sugar is higher than normal, but additional information is required until confirmed as diabetes. Most people go through prediabetes before being diagnosed with type 2 diabetes. The rise in blood sugar seen in prediabetes is caused by problems with insulin resistance, but being in the prediabetic stage does not automatically progress to diabetes. However, the risk of developing diabetes is high, and prediabetes is also a risk factor for the development of heart disease. Like people with type 2 diabetes, people with prediabetes tend to be overweight, have high blood pressure and abnormal cholesterol levels.

In any aspect of the present invention, the quantitative device may include any device capable of quantifying the serum metabolite and used as a kit.

In one embodiment, the quantitative device may include a chromatograph (liquid chromatography, LC) and a mass spectrometer.

The chromatography is liquid-solid chromatography (Liquid-Solid Chromatography, LSC), paper chromatography (Paper Chromatography, PC), thin-layer chromatography (Thin-Layer Chromatography, TLC), gas-solid chromatography (Gas-Solid Chromatography) , GSC), Liquid-Liquid Chromatography, LLC, Foam Chromatography, FC, Emulsion Chromatography, EC, Gas-Liquid Chromatography, GLC ), ion chromatography (Ion Chromatography, IC), gel filtration chromatography (Gel Filtration Chromatograhy, GFC) or gel permeation chromatography (Gel Permeation Chromatography, GPC), including, but not limited to, commonly used in the related art Any quantitative chromatography can be used

The mass spectrometer includes, but is not limited to, MALDI-TOF MS, Q-TOF MS, or flow injection-mass spectrometry (FIA MS), and any qualitative mass spectrometer commonly used in the related art may be used. In one embodiment, the mass spectrometer is a flow injection-mass spectrometer. In the serum metabolite, each component is separated in liquid chromatography according to the different mobility of each component, and components can be identified through elemental composition as well as accurate molecular weight information using information obtained through mass spectrometry. there is. In a preferred embodiment, the quantitative device may include liquid chromatography (LC) and flow injection-mass spectrometry.

The kit of the present invention may be provided with an appropriate container and may be provided with instructions on how to use it.

Example

Hereinafter, the present invention will be described in more detail by way of Examples. However, these examples are only provided to help the understanding of the present invention, and the scope of the present invention is not limited thereto in any sense.

Experimental method

1. Test subject

The subjects of the present invention were metabolized by selecting (based cohort) 1,723 people based on biochemical and epidemiological data according to the American Diabetes Association diabetes diagnosis criteria among the participants of the 3rd phase community-based cohort project collected as part of the Korean Genome Epidemiology Research Project. Produced body information. Furthermore, metabolite information was produced by selecting 500 people who were still normal or who had progressed to pre-diabetes through 6 years of follow-up (community-based cohort stage 6) for normal subjects participating in the baseline cohort (tracking cohort) (Table 1 and Fig. 1).

<Clinical information of participants according to the Diabetes Association standards>

NGT; Normal glucose tolerance, PD; prediabetes, BMI; Body mass index, HDL; High-density lipoprotein cholesterol, LDL; Low-density lipoprotein cholesterol, TG; Triglyceride, HOMA-IR; Homeostatic model assessment of insulin resistance

2. Quantification of metabolites

Liquid chromatography (LC) and flow-injection analysis (FIA)-mass spectrometry method using the AbsoluteIDQp180 kit (Biocrates, Austria) for serum samples collected from participants selected in the base and follow-up cohorts (mass spectrometry, MS) using 186 metabolites {40 acylcarnitines, 21 amino acids, 19 biogeneic amines, 1 hexose, 90 glycerophospholipids; And 15 sphingolipids} were quantified.

The detailed description of the experimental process is as follows.

After centrifuging 10 uL of serum at 100 x g for 2 minutes, the supernatant was collected and API 4000 QTRAP LC/MS/MS equipment (Applied Biosystems, Foster CIty, Agilent Technologies, Santa Clara, CA) equipped with an Agilent 1200 HPLC equipment CA) was used to perform flow injection mass spectrometry for acrylcarnitine, glycerophospholipid and sphingolipid in positive ion mode and hexose in negative ion mode. In the case of amino acids and biogenic amines, liquid chromatography mass spectrometry was performed in positive ion mode. These metabolites were detected using internal standards labeled with safety isotopes such as ¹³ C and ¹⁵ N in the kid and quantified in the uM concentration range using MetVal ^TM analysis tool (Biocrates life sciences AG, Innsbruck, Austria).

- LC-MS/MS analysis conditions

Detector: API 4000 QTRAP+Agilent 1200 HPLC

Column:　Zorbax Eclipse XDB C18 (3.0 x 100mm, 3.5um, Agilent)

Guard column: Security Guard Cartridges C18 4 x 100mm ID

Column oven: 50 ℃

Mobile phase: A (0.2% formic acid/water), B (0.2% formic acid/acetonitrile)

Gradient conditions:

Step Total time (minutes) Flow rate (ul/min) A (%) B(%) 0 0.00 500 100 0 One 0.50 500 100 0 2 5.50 500 5 95 3 6.50 500 5 95 4 7.00 500 100 0 5 9.50 500 100 0

Flow rate: 0.5 ml/min

Injection volume: 10ul

- FIA-MS/MS analysis conditions

Detector: API 4000 QTRAP+Agilent 1200 HPLC

Mobile phase: B (Biocrates Solvent I diluted with 290 ml of methanol)

Isocratic conditions:

Step Total time (minutes) Flow rate (ul/min) A (%) B(%) 0 0.00 30 0 100 One 1.60 30 0 100 2 2.40 200 0 100 3 2.60 200 0 100 4 3.00 30 0 100

Flow rate: 0.03 ml/min

Injection volume: 20 ul

3. Quality control for metabolite information

Quality control was performed on the produced metabolite information under the following conditions. The condition of quality control is that the coefficient of variance (CV) of repeated measurements of the reference standard sample is less than 15%, and metabolite quantification in both the quality control sample and the experimental sample group in the kit. This 50% or more and the coefficient of variation less than 15% were used as the standard. As a result, from the base cohort, 123 {1 total hexoses, 12 acylcarnitines, 21 amino acids, 7 biological amines, 10 sphingomyelines, 32 diacyl phosphatidylcholines; Metabolites of 32 acyl-alkyl phosphatidylcholines, and 8 lysophosphatidylcholines} were screened and 131 (1 total hexose, 17 acylcarnitine, 20 amino acids, 7 biological amines) from the follow-up cohort. ), 13 sphingomyelin, 29 diacylphosphatidylcholine, 34 acyl-yl phosphatidylcholine, and 10 lysophosphatidylcholine) were selected and these were used for statistical analysis.

4. Statistical analysis

Linear regression analysis according to various stratified prediabetic phenotypes (prediabetes, impaired fasting glucose, impaired glucose tolerance, increased glycated hemoglobin, and insulin resistance according to the International Diabetes Association criteria) from the base cohort using the R program and IBM SPSS v20.0 Prediabetes-related metabolites with a statistical significance level of p <4.07E-04 (multiple comparison, α=0.05 for 123 metabolites) were selected through linear regression analysis and logistic regression analysis. did In this case, age, sex, and body mass index (BMI) were used as covariates. In addition, the above results were verified through a random forest selection method, and stepwise regression analysis was performed to create an optimal prediabetes prediction model.

5. Predictive Rate of Prediabetes Development

Selected metabolite markers from the base cohort were used as predictive models (KARE) to predict prediabetic onset from the follow-up cohort. The prediabetes predictive rate was calculated as AUC (area under curve, area within the graph) obtained through ROC (receiver-operating characteristic) analysis. Body mass index (BMI), high-density lipoprotein (HDL), low-density lipoprotein (LDL), triglyceride (TG), CRF and metabolite markers previously reported in Westerners (Framingham Heart Study in US; FHS, KORA cohort study in Germany; KORA) was used as a model and the prediction rate was compared and analyzed. In addition, in the present invention, differences in prediabetes prediction rates were compared in subdivision groups such as age, sex, and body mass index (BMI), which were used as clinical confounding variables for metabolite discovery. The statistical significance level of the prediction rate in each model was set to p < 0.05 through the DeLong test.

Experiment result

Example 1. Identification of pro-diabetic metabolite markers from the base cohort

Normal group (NGT, 924 patients, fasting blood glucose <100 mg/dL and postprandial blood glucose <140 mg/dL and glycated hemoglobin <5.7%) according to the Diabetes Association criteria (Table 1) for 123 well-controlled metabolites from the base cohort. and in the prediabetic group (PD, 799 patients, 100≤fasting glucose <126 or 140≤postprandial glucose <200 or 5.7≤glycated hemoglobin <6.5) Through logistic regression analysis using age, gender, and body mass index as covariates, it was confirmed that there were significant differences in 44 metabolites between the two groups. As a result of linear regression analysis, it was confirmed that 47 and 52 metabolites, respectively, changed significantly ( p <4.07E-04) (top left of FIG. 2 ).

In addition, the fasting glucose disorder group (IFG, 255 persons, 100≤fasting glucose<126), the impaired glucose tolerance group (IGT, 265 persons, 140≤) through the stratification of the prediabetic group according to the blood glucose level according to the difference in blood glucose measurement method (Table 4) Postprandial glucose <200), and both impaired fasting glucose and impaired glucose tolerance (IFG+IGT, 206 patients, 100≤fasting glucose<126 and 140≤postprandial glucose<200) and normal group (NGT, 997 patients) As a result of logistic regression analysis, it was confirmed that each of 34, 29, and 28 metabolites changed significantly ( p <4.07E-04) (top right of FIG. 2).

<Clinical information of test subjects according to fasting glucose (2-h Glucose) in the base cohort>

IFG: impaired fasting glucose, Impaired fasting glucose,

IGT: Impaired glucose tolerance group

In addition, the results of logistic regression analysis and glycation in the normal group (NGT, 1,105, HbA1c<5.7) and prediabetic group (PD, 618, 5.7≤HbA1c<6.5) according to the difference in glycated hemoglobin (HbA1c) (Table 5) As a result of linear regression analysis according to the increase in the hemoglobin measurement value, it was confirmed that 39 and 47 metabolites, respectively, were significantly changed ( p <4.07E-04) (bottom left of FIG. 2 ).

<Participant clinical information according to the difference in glycated hemoglobin (HbA1c) in the base cohort>

On the other hand, insulin resistance as an important indicator of diabetes progression is usually diagnosed by the standard method, the Hyperinsulinemic euglycemic glucose clamp test, but it has disadvantages in terms of cost, time, and patient pain. HOMA-IR (Homeostatic model assessment of insulin resistance) has been proposed as an alternative method to compensate for these shortcomings and measure insulin resistance from a large amount of samples. In the case of insulin resistance, since there is no index value for disease judgment, it is generally analyzed through quartile analysis. In quartile analysis according to insulin resistance (HOMA-IR), logistic regression analysis between the lowest insulin resistance quadrant (Q1, 431 patients) and the highest quadrant (Q4, 431 patients) and insulin resistance measurements (1,723 patients) As a result of linear regression analysis according to the increase, it was confirmed that 25 and 37 metabolites, respectively, were significantly changed ( p <4.07E-04) (Table 6, bottom right of FIG. 2 ).

<Participant clinical information according to insulin resistance (HOMA-IR) in the base cohort>

As a result, 39 reinforcements showing statistical significance (Bonferroni post hoc test, p < 4.07E-04) in the stratified prediabetic phenotype (prediabetes, impaired fasting glucose, impaired glucose tolerance, increased glycated hemoglobin, and insulin resistance according to the Diabetes Association criteria) Enriched metabolites were selected (Table 7, middle part of FIG. 2 ).

<Enriched metabolites according to the prediabetic phenotype>

That is, as shown in Table 7 above, in relation to prediabetes (PD), impaired fasting glucose (IFG), impaired glucose tolerance (IGT), increased glycated hemoglobin (HbA1c), and insulin resistance (HOMA-IR) according to the Diabetes Association criteria 39 Canine accumulated metabolites were screened (Bonferroni post hoc test, p <4.07E-04). The odds ratio (OR, odds ratio) and beta coefficient (Beta) were calculated through logistic regression analysis and linear regression analysis, respectively.

Through random forest analysis on them, 26 pro-diabetic discriminant metabolites were selected (Table 8).

<Selection of metabolites discriminated against prediabetes through random forest (RF) analysis>

In Table 8, the metabolites discriminated against prediabetes were selected through random forest (RF) analysis. For 39 enriched prediabetes metabolites, the optimal 26 prediabetes discriminant metabolites (bold text) were selected through the mean decreae accurary and mean decrease gini.

In addition, the optimal prediabetes prediction model (KARE) with 12 prediabetes-related metabolite markers was finally selected through stepwise regression analysis (Table 9).

<Selection of the optimal pre-diabetes prediction model through step-by-step logistic regression analysis>

Table 9 shows markers in the optimal prediabetes prediction model selected through stepwise logisitc regression. That is, a KARE model including 12 metabolites was selected through step-by-step logistic regression analysis for 26 metabolites discriminated against prediabetes.

Seven metabolites Alanine (Ala), Valine (Val), Gly (Glycine), Tyrosine (Tyr), Methionine (Met), lysoPhosphatidylcholine acyl C18:2 (lysoPC a C18:2), Hydroxysphingomyeline C22:2 (SM ( OH) C22:2) has been previously reported for prediabetes in Westerners, but five metabolites Octadecenoylcarnitine (C18:1), Phosphatidylcholine diacyl C36:1 (PC aa C36:1), Phosphatidylcholine acyl-alkyl C30: 0 (PC ae C30:0), Phosphatidylcholine acyl-alkyl C42:1 (PC ae C42:1), and Sphingomyeline C18:1 (SM C18:1) were first discovered through the present invention.

Example 2. Prediction of prediabetes using metabolite markers

Predictive rates of prediabetes were derived by 12 metabolite markers (KARE) from the follow-up cohort ( FIG. 3 ). According to this, the prediabetes predictive rate [AUC 0.71 (0.67, 0.76)] through the KARE model is the predictive rate [(AUC 0.64 (AUC 0.64) 0.59, 0.68)], and ( p <0.01). Moreover, the prediction rate was higher than that of the EPIC model [AUC 0.56 (0.50, 0.61)] and the KORA model [AUC 0.63 (0.58, 0.68)] previously reported in Westerners. was shown ( p < 0.005).

Although it was lower than the predictive rate [AUC 0.79 (0.75, 0.83)] by fasting blood glucose used as a standard indicator for prediabetes diagnosis [AUC 0.79 (0.75, 0.83)], when the fasting blood glucose model was added to the KARE model [AUC 0.84 (0.80, 0.88)] )] was higher than the predicted rate by fasting blood glucose.

Furthermore, when the CRF model and the fasting blood glucose model were added to the KARE model, the highest prediabetes predictive rate [AUC 0.86 (0.82, 0.89)] was shown. If so, these results mean that prediabetic markers specific to Koreans exist, and that these metabolite markers can increase the predictive rate of prediabetes by the existing prediabetes diagnostic index.

Additionally, the KARE model and the prediabetes predictive rate by fasting blood glucose were compared in subdivision groups such as age (quartile), gender (female/male), and body mass index (normal/overweight/mild obesity) (FIG. 4). Here, there was no difference according to age, and when the body mass index was mildly obese (25kg/m ² ≤BMI < 30kg/m ² ), the prediction rate by the KARE model [AUC 0.79(0.72, 0.87)] was the predicted rate by fasting blood sugar. It was confirmed to be higher than [AUC 0.58 (0.48, 0.68)] ( p <0.001). For female gender [AUC 0.73 (0.66, 0.79) vs. 0.85(0.80, 0.89), p<0.002)] and normal body mass index [BMI<23kg/m ² , AUC 0.67(0.60, 0.74) vs. 0.78 (0.71, 0.84), p<0.01)], it was confirmed that the prediction rate was low. These results not only show a better predictive rate of prediabetes than fasting blood glucose in the course of early obesity based on body mass index, but also mean that prediabetes can be predicted at a level similar to fasting glucose according to age.

As described above, the present invention discovered a marker of a related metabolite according to a stratified prediabetic phenotype based on metabolite information, and this is a marker of a prediabetic metabolite specific to Koreans. Surprisingly, these markers showed improved prediabetes predictive rates compared to pre-diabetic risk factors or conventional metabolite markers. These results can be used for predicting and diagnosing prediabetes, and can be usefully used as a target for drug development for prevention and treatment.

As described above in detail a specific part of the content of the present invention, for those of ordinary skill in the related art, it is clear that this specific description is only a preferred embodiment, and the scope of the present invention is not limited thereby. something to do. Accordingly, the substantial scope of the present invention will be defined by the appended claims and their equivalents.

Claims (10)

Alanine (Ala), Valine (Val), Glycine (Gly), Tyrosine (Tyr), Methionine (Met), lysoPhosphatidylcholine acyl C18:2 (lysoPC a C18:2), Hydroxysphingomyeline C22:2 (SM (OH) C22:2 ), Octadecenoylcarnitine (C18:1), Phosphatidylcholine diacyl C36:1 (PC aa C36:1), Phosphatidylcholine acyl-alkyl C30:0 (PC ae C30:0), Phosphatidylcholine acyl-alkyl C42:1 (PC ae C42:1) ), and a quantification device for serum metabolites composed of Sphingomyeline C18:1 (SM C18:1), wherein the sample provider's body mass index is mild obesity (25 kg/m ² ≤ BMI < 30 kg/ m ² , AUC=0.58) as a diagnostic kit for prediabetes. The prediabetes diagnosis kit according to claim 1, wherein the kit is applied to preventing, predicting, or diagnosing progression to diabetes. delete The prediabetes diagnostic kit according to claim 1 or 2, further comprising a quantitative device for a biochemical factor selected from triglycerides, glycated hemoglobin, fasting blood sugar, fasting insulin, and any combination thereof. The prediabetes diagnostic kit according to claim 1 or 2, wherein the quantification device further comprises a liquid chromatography (LC) and a flow-injection analysis mass spectrometer (FIA-MS). obtaining a sample comprising blood isolated from a human; and
Alanine (Ala), Valine (Val), Glycine (Gly), Tyrosine (Tyr), Methionine (Met), lysoPhosphatidylcholine acyl C18:2 (lysoPC a C18:2), Hydroxysphingomyeline C22:2 (SM (OH) in the sample C22:2), Octadecenoylcarnitine (C18:1), Phosphatidylcholine diacyl C36:1 (PC aa C36:1), Phosphatidylcholine acyl-alkyl C30:0 (PC ae C30:0), Phosphatidylcholine acyl-alkyl C42:1 (PC ae C42:1), and measuring the concentration of a serum metabolite composed of Sphingomyeline C18:1 (SM C18:1), wherein the measured serum metabolite concentration is increased compared to that of the control sample. As a method of providing information necessary for the diagnosis of prediabetes, when the human body mass index is mild obesity (25kg/m ² ≤ BMI < 30 kg/m ² ) A method of providing information necessary for the diagnosis of prediabetes, in which the predictive rate (AUC=0.79) by metabolite markers is higher than the predictive rate (AUC=0.58) for prediabetes. delete The method according to claim 6, wherein the method further comprises a method of providing prediabetes prediction information based on fasting blood sugar, the method for providing information necessary for diagnosis of prediabetes. The method according to claim 6, wherein the method further comprises a method of providing prediabetes prediction information by a CRF model including prediabetes risk factors (age, sex, body mass index, blood lipid concentration, etc.) How to provide the necessary information. The method according to claim 6, wherein the method is a method of providing prediabetes prediction information by fasting blood sugar and prediabetes prediction information by a CRF model including prediabetes risk factors (age, sex, body mass index and blood lipid concentration, etc.) A method of providing information necessary for the diagnosis of prediabetes, further comprising a method of providing a.

Images