Method for predicting chlorogenic acid absorption amount

Abstract

To provide a method for predicting an absorption amount of chlorogenic acids in a person ingesting a chlorogenic acids-containing product.SOLUTION: A prediction method of chlorogenic acids absorption amount includes the processes 1) to 2): 1) a process of acquiring the occupancy of one or more kinds of bacteria out of 5 genera selected from the genus Megamonas, the genus Lachnospira, the genus Ruminococcus 1, the genus Sellimonas and the genus Eggerthella with respect to an intestinal bacterial flora on the basis of information of intestinal bacterial flora of a subject; and 2) a process of calculating a chlorogenic acids absorption amount in the case where the subject ingests a chlorogenic acids-containing product on the basis of the occupancy.SELECTED DRAWING: None

Description

[Technical field]

[0001] The present invention relates to a method for predicting the amount of chlorogenic acids absorbed when a subject ingests chlorogenic acids. [Background technology]

[0002] Chlorogenic acids are found in plants such as coffee beans and sunflower seeds, and have been reported to have physiological functions such as antioxidant activity, blood pressure lowering activity, and lipid burning promotion activity (for example, Patent Document 1). Beverages containing chlorogenic acids are a suitable form for continuously ingesting larger amounts of chlorogenic acids and effectively obtaining their physiological functions, and in recent years, coffee beverages containing high concentrations of chlorogenic acids have been proposed.

[0003] When chlorogenic acids are orally ingested, they are absorbed through the intestine and circulate throughout the body, exerting their effects at the target site. However, the amount of chlorogenic acids circulating throughout the body, i.e., the amount of chlorogenic acids absorbed, varies considerably from person to person, and it is known that the efficacy and safety of these acids vary greatly from person to person.

[0004] On the other hand, the intestinal tract plays an important role in the pharmacokinetics of food and medicines as the site of absorption and excretion. In particular, intestinal bacteria play a role in regulating absorption by catabolizing compounds. In conjunction with recent metagenomic analysis, the analysis of the microbiome, including the intestinal flora, has progressed, and it has become clear that intestinal bacteria are involved in the efficacy and side effects of many drugs (Non-Patent Documents 1 and 2).

[0005] It has been revealed that chlorogenic acids are also catabolized by intestinal bacteria (Non-Patent Document 3). However, it is not known whether intestinal bacteria play a role as a regulator of the amount of chlorogenic acids absorbed or the expression of their effects, and the genus of intestinal bacteria involved in this regulation is also unknown. [Prior art documents] [Patent documents]

[0006]

Patent Document 1

Non-Patent Document

[0007]

Non-Patent Document 1

Non-Patent Document 2

Non-Patent Document 3

Summary of the Invention

Problems to be Solved by the Invention

[0008] The present invention relates to a method for predicting the absorption amount of chlorogenic acids in a person who ingests a chlorogenic acid-containing product.

Means for Solving the Problems

[0009] The present inventors discovered that the absorption of chlorogenic acids depends on large intestine absorption, and when examining the relationship between the intestinal flora and the absorption amount of chlorogenic acids, they found that the occupancy rate of specific intestinal bacteria in the intestinal flora correlates with the absorption amount of chlorogenic acids, and the absorption amount of chlorogenic acids can be predicted from the occupancy rate.

[0010] That is, the present invention relates to the following. A method for predicting the absorption amount of chlorogenic acids, comprising the following steps 1) to 2): 1) A step of obtaining the occupancy rate of one or more types of bacteria selected from the five genera of Megamonas, Ruminococcus, Lachnococcus 1, Cellulomonas, and Eggerthella in the intestinal flora based on the information of the intestinal flora of a subject, 2) Calculating the amount of chlorogenic acids absorbed when the subject ingests a product containing chlorogenic acids based on the occupancy rate.

Advantages of the Invention

[0011] According to the method of the present invention, the amount of chlorogenic acids absorbed by a person who ingests a product containing chlorogenic acids can be predicted.

Brief Description of the Drawings

[0012]

Fig. 1-1

Fig. 1-2

Fig. 1-3

Fig. 1-4

Fig. 2-1

Fig. 2-2

Fig. 2-3

Fig. 2-4

Fig. 3-1

Fig. 3-2

Fig. 3-3

Fig. 3-4

Fig. 4-1

Fig. 4-2

Fig. 4-3

Fig. 4-4

Fig. 5-1

Fig. 5-2

Fig. 5-3

Fig. 5-4

Fig. 6

Fig. 7

Fig. 8

Fig. 9

Fig. 10

Fig. 11

Fig. 12

Mode for Carrying Out the Invention

[0013] In the present invention, the "absorption amount of chlorogenic acids" means the absorption amount of chlorogenic acids when a product containing chlorogenic acids is orally ingested. The absorption amount of chlorogenic acids is represented by the blood concentration measured from the area under the blood concentration-time curve after ingestion of the product containing chlorogenic acids. More specifically, it is represented by the blood concentration measured from the area under the blood concentration-time curve (hereinafter referred to as "AUC(0 - 8 hours)") from 0 to 8 hours after ingestion of the product containing chlorogenic acids, or from the area under the blood concentration-time curve (hereinafter referred to as "AUC(3 - 8 hours)") from 3 to 8 hours. Here, AUC(0 - 8 hours) is a parameter representing the absorption of the entire digestive tract, and AUC(3 - 8 hours) is a parameter indicating only the absorption of the large intestine. As shown in the reference examples described later, there are individual differences in the absorption of chlorogenic acids, and the blood concentration changes after ingestion of a product containing chlorogenic acids are divided into a unimodal group (low absorption group) having one peak and a bimodal group (high absorption group) having two peaks. In the high absorption group, in addition to the first peak from 0 to 3 hours after ingestion due to absorption from the small intestine, there is a second peak from 3 to 8 hours after ingestion due to absorption in the large intestine. The subjects in the high absorption group have significantly higher AUC(3 - 8 hours) and AUC(0 - 8 hours) (Reference Example 1). Also, AUC(0 - 8 hours) and AUC(3 - 8 hours) have a very strong correlation, and the absorption of chlorogenic acid depends on the absorption of the large intestine (Reference Example 1). Therefore, it can be said that AUC(0 - 8 hours) and AUC(3 - 8 hours) are absorption parameters capable of discriminating between the high absorption group and the low absorption group.

[0014] In the present invention, "chlorogenic acids" is a general term for monocatechoylquinic acids such as 3-caffeoylquinic acid (3-CQA), 4-caffeoylquinic acid (4-CQA), and 5-caffeoylquinic acid (5-CQA), monophenoloylquinic acids such as 3-feruloylquinic acid (3-FQA), 4-feruloylquinic acid (4-FQA), and 5-feruloylquinic acid (5-FQA), and dicaffeoylquinic acids such as 3,4-dicaffeoylquinic acid (3,4-diCQA), 3,5-dicaffeoylquinic acid (3,5-diCQA), and 4,5-dicaffeoylquinic acid (4,5-diCQA). Among them, 3-CQA, 4-CQA, and 5-CQA are structural isomers with different positions where caffeic acid is added to quinic acid. In the present invention, the sum of these is defined as monocatechoylquinic acid (CQA). 3-FQA, 4-FQA, and 5-FQA are structural isomers with different positions where ferulic acid is added to quinic acid. In the present invention, the sum of these is defined as monophenoloylquinic acid (FQA). In the present invention, in a chlorogenic acid-containing product, it is sufficient to contain at least one of the above nine types of chlorogenic acids, but it is preferably to contain CQA or FQA. As CQA, it is preferably to contain at least 5-CQA.

[0015] Chlorogenic acids may be in the form of salts. Examples of salts include alkali metal salts such as sodium and potassium, alkaline earth metal salts such as magnesium and calcium, organic amine salts such as monoethanolamine, diethanolamine, and triethanolamine, and basic amino acid salts such as arginine, lysine, histidine, and ornithine.

[0016] Chlorogenic acids can be extracted or purified from plant raw materials such as coffee beans. For example, coffee bean extract and coffee polyphenols purified therefrom are rich in chlorogenic acids.

[0017] In the present invention, examples of the chlorogenic acid-containing products include preparations containing the above-mentioned chlorogenic acids, that is, preparations containing purified chlorogenic acids, coffee bean extracts, or coffee polyphenols, and those containing these preparations.

[0018] The method for predicting the absorption amount of chlorogenic acids in the present invention is as follows: 1) to 2) below: 1) A step of analyzing the occupancy rate of one or more types of bacteria selected from five genera, namely Megamonas, Ruminospira, Lachnococcus 1, Cellulomonas, and Eggerthella, in the intestinal flora based on the information on the intestinal flora of the subject; 2) A step of calculating the absorption amount of chlorogenic acids when the subject ingests a chlorogenic acid-containing product based on the occupancy rate; includes the steps of.

[0019] The information on the intestinal flora of the subject in step 1) refers to the information on the bacterial strain composition obtained by analyzing the intestinal flora for a sample derived from the subject containing intestinal bacteria, and is not limited to the information obtained by collecting and analyzing the sample when implementing the method of the present invention, but also includes the information already obtained in the subject. The bacterial strain composition refers to the types and composition ratios of bacterial genera constituting the intestinal flora, and can be determined based on the nucleotide sequence of the 16S rRNA gene as described later. Examples of the sample derived from the subject include samples containing intestinal bacteria such as feces, gastrointestinal fluid, gastrointestinal lavage fluid, and biological samples. Among these, feces are preferred because the burden on the subject is small and the invasiveness is low. The collection of feces is not particularly limited, and examples include the rectal feces collection method (using a rectal swab, etc.), the excreted feces collection method (using a cotton swab for feces, etc.), etc. The excreted feces collection method is preferred from the perspective of invasiveness. The specimen can be used for analysis immediately after collection, but can also be stored frozen at -80°C.

[0020] The subject includes humans who can ingest chlorogenic acid-containing products, preferably humans who want to grasp in advance the absorption amount of chlorogenic acids when ingesting chlorogenic acid-containing products.

[0021] The analysis means of the intestinal flora for obtaining intestinal bacteria information is not particularly limited, but preferably, metagenomic analysis for analyzing the bacterial strain composition in the intestinal flora based on the nucleotide sequence of the 16S rRNA gene contained in the flora genomic DNA can be mentioned. That is, using the DNA extracted from the sample as a template, a DNA fragment of the V1-V2 region of the 16S rRNA gene is amplified by PCR, and then the sequence of the PCR product is analyzed by, for example, the paired-end method using a Miseq System (Illumina), etc. From the obtained nucleotide sequence, the bacterial strain composition is calculated according to the workflow of QIIME2. Specifically, first, after using Cutadapt to exclude the primer sequence, 3'-terminal base deletion, removal of bases derived from PhiX, and quality filtering are performed by DADA2 to obtain high-quality nucleotide sequences. Then, noise removal (correction of sequence errors), merging of paired-end sequences, and chimera sequence removal are performed to obtain ASV (Amplicon Sequence Variant) representative sequences. Next, the bacterial genera of these representative sequences are identified by mapping them to a database of 16S rRNA genes using a naive Bayes classifier of QIIME2 or the like. Mapping can be performed, for example, by using the sequence obtained by cutting out the V1-V2 region from the representative sequences obtained by clustering SILVA SSU Ref (Version 132) provided by SILVA at a threshold of 99% as training data. And by counting the number of attributed (mapped) nucleotide sequences, the bacterial strain composition of the intestinal flora is quantified.

[0022] In step 1), based on the intestinal bacteria information, the occupancy rate of one or more bacteria selected from the genera Megamonas, Ruminospira, Lachnococcus 1, Cellulomonas, and Eggerthella in the intestinal flora is obtained. The gut microbiota refers to the bacterial species inhabiting the interior of the intestines of animals including humans and the ecosystem of their microbial community. The occupancy rate of the above five types of bacteria in the gut microbiota refers to the ratio (relative abundance) of the number of reads of each bacterial genus to the total number of reads of sequences in the metagenomic analysis of the gut microbiota.

[0023] The genus Megamonas is a bacterium of the phylum Firmicutes classified in the class Negativicutes, and for example, Megamonas funiformis is known. The genus Lachnospira is a bacterium of the phylum Firmicutes classified in the class Clostridia, and for example, Lachnospira multipara is known. The genus Ruminococcus1 is a bacterium of the phylum Firmicutes classified in the class Clostridia, and for example, Ruminococcus callidus is known. The genus Sellimonas is a bacterium of the phylum Firmicutes classified in the class Clostridia, and for example, Sellimonas intestinalis is known. The genus Eggerthella is a bacterium of the phylum Actinobacteria classified in the class Coriobacteriia.

[0024] As shown in the examples described below, the relative abundances of each of the bacteria of the genus Megamonas, genus Lachnospira, genus Ruminococcus1, genus Sellimonas, and genus Eggerthella in the gut microbiota were found to be significantly correlated with the AUC(0-8 hours) and AUC(3-8 hours) of CQA in the subjects who ingested chlorogenic acids as follows. Figure 1-1 shows the results of the linear correlation analysis between the relative abundance of Megamonas bacteria and AUC(0-8 h), and the coefficient of determination R of the regression equation 2 is 0.34. The correlation coefficient R is 0.65 (Pearson) and 0.75 (Spearman), indicating a positive correlation. Figure 1-2 shows the results of the linear correlation analysis between the relative abundance of Megamonas bacteria and AUC(3-8 h), and the coefficient of determination R of the regression equation 2 is 0.59. The correlation coefficient R is 0.88 (Pearson) and 0.82 (Spearman), indicating a positive correlation. Figure 2-1 shows the results of the linear correlation analysis between the relative abundance of Ruminococcus bacteria and AUC(0-8 h), and the coefficient of determination R of the regression equation 2 is 0.28. The correlation coefficient R is 0.56 (Pearson) and 0.51 (Spearman), indicating a positive correlation. Figure 2-2 shows the results of the linear correlation analysis between the relative abundance of Ruminococcus bacteria and AUC(3-8 h), and the coefficient of determination R of the regression equation 2 is 0.35. The correlation coefficient R is 0.63 (Pearson) and 0.63 (Spearman), indicating a positive correlation. Figure 3-1 shows the results of the linear correlation analysis between the relative abundance of Lachnococcus 1 bacteria and AUC(0-8 h), and the coefficient of determination R of the regression equation 2 is 0.42. The correlation coefficient R is -0.70 (Pearson) and -0.79 (Spearman), indicating a negative correlation. Figure 3-2 shows the results of the linear correlation analysis between the relative abundance of Lachnococcus 1 bacteria and AUC(3-8 h), and the coefficient of determination R of the regression equation 2 is 0.39. The correlation coefficient R is -0.63 (Pearson) and -0.76 (Spearman), indicating a negative correlation. Figure 4-1 shows the results of the linear correlation analysis between the relative abundance of Cellulomonas bacteria and AUC(0-8 h), and the coefficient of determination R of the regression equation 2 is 0.45. The correlation coefficient R is 0.70 (Pearson) and 0.80 (Spearman), indicating a positive correlation. Figure 4-2 shows the results of the linear correlation analysis between the relative abundance of Cellulomonas bacteria and AUC(3-8 h), and the coefficient of determination R of the regression equation2 is 0.67. The correlation coefficient R is 0.84 (Pearson) and 0.80 (Spearman), indicating a positive correlation. Figure 5-1 shows the results of the linear correlation analysis between the relative abundance of Eggerthella bacteria and AUC (0 - 8 hours), and the coefficient of determination R of the regression equation 2 is 0.26. The correlation coefficient R is 0.51 (Pearson) and 0.68 (Spearman), indicating a positive correlation. Figure 5-2 shows the results of the linear correlation analysis between the relative abundance of Eggerthella bacteria and AUC (3 - 8 hours), and the coefficient of determination R of the regression equation 2 is 0.26. The correlation coefficient R is 0.55 (Pearson) and 0.71 (Spearman), indicating a positive correlation. Similarly, for FQA, the AUC (0 - 8 hours) and AUC (3 - 8 hours) of FQA have significant correlations with the relative abundances of Megamonas bacteria, Ruminococcus bacteria, Lachnoclostridium 1 bacteria, Cellulomonas bacteria, and Eggerthella bacteria, respectively (Examples 8 and 9).

[0025] Therefore, by obtaining the ratio (occupancy rate) of the number of reads of the nucleotide sequences of one or more bacteria selected from the genera Megamonas, Ruminococcus, Lachnoclostridium 1, Cellulomonas, and Eggerthella to the number of data points (total number of sequence reads) of the nucleotide sequences of the entire gut microbiota, the AUC (0 - 8 hours) and AUC (3 - 8 hours) of chlorogenic acids, that is, the absorption amount, of the subjects who ingested chlorogenic acids can be calculated (Step 2).

[0026] Specifically, the absorption amount of chlorogenic acids of the subject can be calculated by comparing the occupancy data of each of the above gut microbiota genera obtained from the subject with the calibration curve or prediction model of the AUC (0 - 8 hours) or AUC (3 - 8 hours) of chlorogenic acids. A prediction model for the AUC(0-8 hours) or AUC(3-8 hours) of chlorogenic acids can be constructed by performing a regression analysis using the occupancy rate of each intestinal bacterial genus as an explanatory variable and the AUC(0-8 hours) or AUC(3-8 hours) of chlorogenic acids as the objective variable. In the construction of the prediction model, the occupancy rate of one or more intestinal bacterial genera selected from among five genera including Megamonas, Ruminococcus, Lachnospiraceae bacterium NK4A136 group, Cellulomonas, and Eggerthella may be used as an explanatory variable. From the perspective of prediction accuracy, it is preferable to use two or more, three or more, four or more, or all five types.

[0027] Examples of regression analysis methods include, in addition to principal component regression analysis, PLS (Partial least squares projection to latent structures) regression analysis, OPLS (Orthogonal projections to latent structures) regression analysis, and generalized linear regression analysis, as well as multivariate regression analysis methods such as bagging, support vector machines, random forests, and neural network regression analysis. Among these, it is preferable to use generalized linear regression analysis.

[0028] For example, in the simple regression model derived from the regression analysis shown in FIGS. 1-1 and 1-2, the relative abundance of Megamonas bacteria can predict the AUC(0-8 hours) of CQA in 73% of the subjects and the AUC(3-8 hours) of CQA in 63% of the subjects with an accuracy of 60-140%, which is the coefficient of variation (%CV) of 100%±AUC. In the simple regression model derived from the regression analysis shown in FIGS. 2-1 and 2-2, the relative abundance of Ruminococcus bacteria can predict the AUC(0-8 hours) of CQA in 83% of the subjects and the AUC(3-8 hours) of CQA in 61% of the subjects with an accuracy of 60-140%, which is the coefficient of variation (%CV) of 100%±AUC. In the simple regression model derived from the regression analysis shown in FIGS. 3-1 and 3-2, the relative abundance of Lachnococcus genus 1 bacteria can predict the AUC(0-8 hours) of CQA in 90% of the subjects, the AUC(3-8 hours) of CQA in 40% of the subjects, with an accuracy of 60-140% which is the coefficient of variation (%CV) of 100%±AUC. In the simple regression model derived from the regression analysis shown in FIGS. 4-1 and 4-2, the relative abundance of Cellulomonas genus bacteria can predict the AUC(0-8 hours) of CQA in 100% of the subjects, the AUC(3-8 hours) of CQA in 58% of the subjects, with an accuracy of 60-140% which is the coefficient of variation (%CV) of 100%±AUC. In the simple regression model derived from the regression analysis shown in FIGS. 5-1 and 5-2, the relative abundance of Eggerthella genus bacteria can predict the AUC(0-8 hours) of CQA in 71% of the subjects, the AUC(3-8 hours) of CQA in 47% of the subjects, with an accuracy of 60-140% which is the coefficient of variation (%CV) of 100%±AUC.

[0029] Also, in the multiple regression model derived from the linear regression analysis with the relative abundances of Megamonas genus bacteria, Ruminospira genus bacteria, Lachnococcus genus 1 bacteria, Cellulomonas genus bacteria, and Eggerthella genus bacteria as explanatory variables respectively and the AUC(3-8 hours) of CQA as the target variable (y), the AUC(3-8 hours) of CQA can be predicted in 80% of the subjects with an accuracy of 40-160% which is the coefficient of variation (%CV) of 100%±AUC. Similarly, for the AUC(0-8 hours) of CQA, by performing a linear regression analysis with the relative abundances of Ruminospira genus bacteria, Lachnococcus genus 1 bacteria, Cellulomonas genus bacteria, and Eggerthella genus bacteria as explanatory variables respectively and the AUC(0-8 hours) of CQA as the target variable (y), and using the derived multiple regression model, the AUC(0-8 hours) of CQA can be predicted in 84% of the subjects with an accuracy of 60-140% which is the coefficient of variation (%CV) of 100%±AUC.

[0030] Not only for CQA, but also prediction of FQA is possible. For example, in the multiple regression model derived from a linear regression analysis using the relative abundances of Megamonas bacteria, Ruminococcus bacteria, Lachnoclostridium 1 bacteria, Cellulomonas bacteria, and Eggerthella bacteria as explanatory variables and the AUC (3 - 8 hours) of FQA as the objective variable (y), the AUC (3 - 8 hours) of FQA can be predicted with an accuracy of 40 - 160% of 100% ± the coefficient of variation (%CV) of AUC in 72% of the subjects. Similarly, by performing a linear regression analysis using the relative abundances of Ruminococcus bacteria, Lachnoclostridium 1 bacteria, and Eggerthella bacteria as explanatory variables and the AUC (0 - 8 hours) of FQA as the objective variable (y), and using the derived multiple regression model, the AUC (0 - 8 hours) of FQA can be predicted with an accuracy of 60 - 140% of 100% ± the coefficient of variation (%CV) of AUC in 84% of the subjects.

[0031] Furthermore, prediction of individual components is also possible. For example, for 5 - CQA, by deriving a multiple regression model from a linear regression analysis using the relative abundances of Megamonas bacteria, Ruminococcus bacteria, Lachnoclostridium 1 bacteria, Cellulomonas bacteria, and Eggerthella bacteria as explanatory variables and the AUC (3 - 8 hours) of 5 - CQA as the objective variable (y), the AUC (3 - 8 hours) of 5 - CQA can be predicted with an accuracy of 40 - 160% of 100% ± the coefficient of variation (%CV) of AUC in 64% of the subjects.

[0032] In the high chlorogenic acid absorption group, when 185 g of Helsea coffee (manufactured by Kao Corporation, containing 270 mg of the total amount of the nine kinds of chlorogenic acids, 170 mg of CQA, and 42 mg of FQA) was ingested, the value of the AUC (3 - 8 hours) of CQA was 29.5 - 77.3 ng / mL*hr, and in the low chlorogenic acid absorption group, it was 4.75 - 36.8. That is, subjects predicted to have an AUC (3 - 8 hours) value of 29.5 ng / mL*hr or more, preferably 36.8 ng / mL*hr or more, are in the high chlorogenic acid absorption group, and subjects predicted to have a value less than 36.8 ng / mL*hr, preferably less than 29.5 ng / mL*hr, can be determined to be in the low chlorogenic acid absorption group.

[0033] Similarly, in the AUC (0 - 8 hours) of CQA, subjects predicted to have 69.9 ng / mL*hr or more, preferably 127.9 ng / mL*hr or more, are the high chlorogenic acid absorption group, and subjects predicted to have less than 127.9 ng / mL*hr, preferably less than 69.9 ng / mL*hr, can be determined to be the low chlorogenic acid absorption group.

[0034] Furthermore, in the AUC (3 - 8 hours) of FQA, subjects predicted to have 27.1 ng / mL*hr or more, preferably 37.4 ng / mL*hr or more, are the high chlorogenic acid absorption group, and subjects predicted to have less than 37.4 ng / mL*hr, preferably less than 27.1 ng / mL*hr, can be determined to be the low chlorogenic acid absorption group.

[0035] In the AUC (0 - 8 hours) of FQA, subjects predicted to have 64.4 ng / mL*hr or more, preferably 93.7 ng / mL*hr or more, are the high chlorogenic acid absorption group, and subjects predicted to have less than 93.7 ng / mL*hr, preferably less than 64.4 ng / mL*hr, can be determined to be the low chlorogenic acid absorption group.

Example

[0036] Reference Example 1 Measurement of Changes in Chlorogenic Acid Blood Concentration by a Single Oral Intake Test of Chlorogenic Acids in Humans and Derivation of AUC (0 - 8 hours) and AUC (3 - 8 hours) 1. Single-dose ingestion test of human chlorogenic acids This human test was conducted after obtaining approval from the company's ethics committee. The subjects were 25 adult Japanese men (35.2 ± 9.10 years old, height 171.3 ± 6.95 cm, weight 69.0 ± 10.40 kg, BMI 23.5 ± 2.63 kg / m 2) and, after explaining the test details, obtained informed consent. The subjects were asked to refrain from consuming chlorogenic acids and polyphenols since the previous day, and were only allowed to drink water after 21:00 on the previous night. On the test day, 185 g of Helcia Coffee (manufactured by Kao Corporation, containing 270 mg of the total amount of the nine types of chlorogenic acids, 170 mg of CQA, and 42 mg of FQA) was ingested, and blood samples were collected at 0, 0.5, 1, 2, 3, 4.5, 6, and 8 hours after ingestion. The collected blood samples were immediately centrifuged (3000×g, 10 minutes, 4°C) to separate the plasma. To the plasma samples, 0.4 mol / L phosphate buffer (pH 3.6: manufactured by Fujifilm Wako Pure Chemical Corporation) containing an amount equivalent to 10% of its volume of 20 w / v% ascorbic acid and 0.1 w / v% disodium ethylenediaminetetraacetate was added, and after further adding 1.5 times the volume of methanol, they were stored at -80°C. Stool excreted at any time from the night before the test until the end of the test was collected once and immediately frozen in a deep freezer (-80°C).

[0037] 2. Analysis of plasma samples After stirring the plasma sample for 30 seconds, it was centrifuged (4°C, 15,000×g, 5 minutes), and the supernatant was used as the analysis sample. The quantification of CQA in human plasma samples was performed using a high-performance liquid chromatograph (Exion LC AD, manufactured by AB SCIEX) equipped with a Triplequad5500+ (manufactured by AB SCIEX), and measurements were carried out in negative mode in MRM mode. As monitoring ion pairs, for 3-CQA, 352.95→190.9; for 4-CQA, 352.91→173.0; for 5-CQA, 353.04→191.0; for 3-FQA, 366.98→193.0; for 4-FQA, 366.96→172.9; and for 5-FQA, 366.96→191.1 were detected. For the HPLC analysis column, Ascentis Express 90Å RP-Amide 15 cm×2.1 mm, 2.7 μm (manufactured by Supelco) was used, and the column temperature was set at 30°C. The injection volume was 5 μL, the autosampler temperature was 4°C, and for the mobile phase, Solution A: 250 mM aqueous acetic acid solution and Solution B: 250 mM acetic acid acetonitrile were used, and the analysis was performed according to the gradient program shown in Table 1 at a flow rate of 0.55 mL / min. CQA in plasma was quantified after creating a calibration curve using a high-purity sample (manufactured by Nagara Science) as a standard and ethyl gallate (manufactured by FUJIFILM Wako Pure Chemical Corporation) as an internal standard substance.

[0038]

Table 1

[0039] 3. Calculation of absorption parameters (AUC(0-8 h) and AUC(3-8 h)) of chlorogenic acid The blood concentration trends of CQA obtained by summing the blood concentrations of quantified 3-CQA, 4-CQA, and 5-CQA, and FQA obtained by summing the blood concentrations of 3-FQA, 4-FQA, and 5-FQA were calculated. For 25 subjects using Graphpad Prism (Graphpad Software), AUC(0 - 8 hours) and AUC(3 - 8 hours), which are the AUC values 0 - 8 hours and 3 - 8 hours after ingestion, were derived for CQA and FQA, respectively.

[0040] 4. Grouping into high-absorption and low-absorption groups for CQA In the grouping, at 3 hours after ingestion, the blood concentration increased again, and the bimodal group with absorption confirmed was defined as the high-absorption group, and the unimodal group not corresponding to it was defined as the low-absorption group. As a result, 8 subjects were identified as the high-absorption group of CQA and 17 subjects as the low-absorption group. Figure 11-A shows the blood concentration changes of the unimodal low-absorption group (single peak, 17 subjects) and the bimodal high-absorption group (double peak, 8 subjects) of CQA, respectively. Figure 11-B shows the correlation analysis results of AUC(0-8 hours) and AUC(3-8 hours) in 25 subjects, and a very strong correlation was confirmed. Figure 11-C is a box-and-whisker plot of AUC(3-8 hours) for the high-absorption group and the low-absorption group, and Figure 11-D is a box-and-whisker plot of AUC(0-8 hours), where p represents the statistic (significance level: p<0.05) when performing the Mann-Whitney U test. As a result of the test, in the high-absorption group, the AUC(0-8 hours) and AUC(3-8 hours) of CQA were significantly higher than those in the low-absorption group.

[0041] 5. Grouping into high-absorption and low-absorption groups for FQA By the method shown in the previous paragraph, the high-absorption group and the low-absorption group were also grouped for FQA. As a result, 8 subjects were identified as the high-absorption group of FQA and 17 subjects as the low-absorption group, and the high-absorption group of FQA was consistent with the high-absorption group of CQA. Figure 12-A shows the blood concentration changes of the unimodal low-absorption group (single peak, 17 subjects) and the bimodal high-absorption group (double peak, 8 subjects) of FQA, respectively. Figure 12-B shows the correlation analysis results of AUC(0-8 hours) and AUC(3-8 hours) in 25 subjects, and a very strong correlation was confirmed. Figure 12-C is a box-and-whisker plot of AUC(3-8 hours) for the high-absorption group and the low-absorption group, and Figure 12-D is a box-and-whisker plot of AUC(0-8 hours), where p represents the statistic (significance level: p<0.05) when performing the Mann-Whitney U test. As a result of the test, in the high-absorption group, the AUC(0-8 hours) and AUC(3-8 hours) of FQA were significantly higher than those in the low-absorption group.

[0042] Example 1 Prediction Results of AUC(0-8 hours) and AUC(3-8 hours) of CQA Using the Relative Abundance of Megamonas bacteria 1. Single-dose ingestion test of human chlorogenic acids The AUC(0 - 8 hours) and AUC(3 - 8 hours) of CQA obtained in the test of the reference example 1 were used for analysis.

[0043] 2. Microbiome analysis of intestinal flora in fecal specimens Using the DNA extracted from the fecal specimen obtained in the test of the reference example 1 as a template, a DNA fragment in the V1 - V2 region of the 16S rRNA gene was amplified by PCR (Tks Gflex DNA Polymerase, manufactured by Takara Bio Inc.), and subsequently, the sequence of the PCR product was analyzed by the paired - end method using the Miseq System (Illumina). From the obtained nucleotide sequences, the bacterial community composition was calculated according to the workflow of QIIME2 (https: / / qiime2.org / ), which is an open - source pipeline for microbial analysis and is detailed below. First, after excluding the primer sequences using Cutadapt (https: / / cutadapt.readthedocs.io / en / stable / guide.html), 3' end base trimming, removal of PhiX - derived nucleotide sequences, and quality filtering were performed by DADA2 (https: / / benjjneb.github.io / dada2 / index.html) to obtain high - quality nucleotide sequences. Then, noise removal (correction of sequence errors), merging of paired - end sequences, and chimera sequence removal were carried out to obtain ASV (Amplicon Sequence Variant) representative sequences. Next, the bacterial genera of these representative sequences were identified by mapping them to the 16S rRNA gene database using the naive Bayes classifier of QIIME2. In this analysis, among the representative sequences obtained by clustering the SILVA SSU Ref (Version 132) provided by SILVA at a 99% threshold, the sequences with the V1 - V2 region cut out were used as training data for mapping. And by counting the number of attributed (mapped) nucleotide sequences, the bacterial community composition of the gut microbiota was quantified.

[0044] 3. Identification of Megamonas bacteria by correlation analysis Correlation analysis was performed between the AUC(0 - 8 hours) and AUC(3 - 8 hours) of 25 subjects and the relative abundance of each genus of intestinal bacteria in fecal specimens, and an attempt was made to select the genera of bacteria involved in the absorption of CQA. Graphpad Prism (Graphpad Software) was used for the correlation analysis, and a linear correlation coefficient (r) of Spearman or Pearson of 0.4 or more and a test statistic (p) of 0.05 or less in the non - correlation test were set as the significance level. Subjects with a relative abundance of intestinal bacteria genera of 0 in the analysis target were excluded data. As a result of the analysis, the presence of Megamonas bacteria was confirmed in 11 subjects excluding those with a relative abundance of 0. As shown in Figure 1 - 1, a significant correlation was confirmed between the relative abundance of Megamonas bacteria and the AUC(0 - 8 hours) of CQA in the 11 subjects (r(Pearson)=0.65, p(Pearson)=0.029, r(Spearman)=0.75, p(Spearman)=0.011). Furthermore, as shown in Figure 1 - 2, a significant correlation was confirmed with the AUC(3 - 8 hours) of CQA (r(Pearson)=0.88, p(Pearson)=0.0003, r(Spearman)=0.82, p(Spearman)=0.003).

[0045] 4. Construction of a regression model for predicting AUC(0-8 h) and AUC(3-8 h) of CQA by Megamonas bacteria and examination of prediction accuracy Using RStudio, linear simple regression analysis was performed on the data of 11 subjects with the relative abundance of Megamonas bacteria as the explanatory variable (x) and the AUC(0 - 8 hours) or AUC(3 - 8 hours) of CQA as the objective variable (y). Cross - validation was performed 4 times, with 2 - 3 subjects' data per round as test data and 8 - 9 subjects' data as model - building data, and the prediction accuracy for the AUC(0 - 8 hours) or AUC(3 - 8 hours) of CQA of 11 subjects was examined. The prediction accuracy was calculated as the prediction result / actual measurement result. As a result, for AUC(0 - 8 hours), as shown in Figure 1 - 1, the simple regression model derived from the regression analysis was y = 365.6×x + 58.1, and the coefficient of determination (r 2) was 0.34 and the statistical test quantity (p) was 0.049, indicating a significant regression relationship. For AUC (3 - 8 hours), as shown in Fig. 1 - 2, the simple regression model derived from the regression analysis was y = 244.6×x + 7.0, and the coefficient of determination (r 2 ) was 0.59 and the statistical test quantity (p) was 0.0033, indicating a significant regression relationship. The prediction accuracy by cross - validation for AUC (0 - 8 hours) is shown in Fig. 1 - 3, Table 1 - 1, and Table 1 - 2, and for AUC (3 - 8 hours) is shown in Fig. 1 - 4, Table 1 - 1, and Table 1 - 2. Based on the relative abundance of Megamonas bacteria, the AUC (0 - 8 hours) of CQA in 73% of the subjects and the AUC (3 - 8 hours) of CQA in 63% of the subjects could be predicted with an accuracy of 60 - 140% compared to the measured values.

[0046]

Table 1 - 1

[0047]

Table 1 - 2

[0048] Example 2 Prediction results of AUC (0 - 8 hours) and AUC (3 - 8 hours) of CQA using the relative abundance of Ruminococcus bacteria 1. Measurement of AUC(0-8 h) and AUC(3-8 h) of CQA by single-dose ingestion test of human chlorogenic acids and microbiome analysis of intestinal flora Using the intestinal flora information of 25 subjects obtained in the above Reference Example 1 and Example 1 and the AUC (0 - 8 hours) and AUC (3 - 8 hours) of CQA when 185 g of Helsea coffee (manufactured by Kao Corporation, containing 270 mg of total chlorogenic acids, 170 mg of CQA, and 42 mg of FQA, the same as in Reference Example 1) was ingested, the following analysis was carried out.

[0049] 2. Identification of Ruminococcus bacteria by correlation analysis Correlation analysis was performed between the AUC(0 - 8 hours) and AUC(3 - 8 hours) of 25 subjects and the relative abundance of each intestinal bacterial genus in fecal specimens, and an attempt was made to select the bacterial genera involved in the absorption of CQA. Graphpad Prism (Graphpad Software) was used for the correlation analysis, and a Spearman or Pearson correlation coefficient (r) of linear correlation analysis of 0.4 or more and a test statistic (p) of 0.05 or less in the uncorrelated test were used as the significance level. Subjects with a relative abundance of 0 of the intestinal bacterial genera to be analyzed were excluded data. As a result of the analysis, the presence of Ruminococcus bacteria was confirmed in 18 subjects excluding subjects with a relative abundance of 0. As shown in Figure 2 - 1, a significant correlation was confirmed between the relative abundance of Ruminococcus bacteria and the AUC(0 - 8 hours) of CQA in the 18 subjects (r(Pearson)=0.56, p(Pearson)=0.014, r(Spearman)=0.51, p(Spearman)=0.031). Furthermore, as shown in Figure 2 - 2, a significant correlation was confirmed between the relative abundance of Ruminococcus bacteria and the AUC(3 - 8 hours) of CQA in the 18 subjects (r(Pearson)=0.63, p(Pearson)=0.005, r(Spearman)=0.63, p(Spearman)=0.004).

[0050] 3. Construction of a regression model for predicting AUC(0-8 h) and AUC(3-8 h) of CQA by Ruminococcus bacteria and examination of prediction accuracy Using RStudio, linear simple regression analysis was performed on the data of 18 subjects with the relative abundance of Ruminococcus bacteria as the explanatory variable (x) and the AUC(0 - 8 hours) or AUC(3 - 8 hours) of CQA as the objective variable (y). Cross - validation was performed 4 times with 4 - 5 data per time as test data and 13 - 14 data as model - building data, and the prediction accuracy for the AUC(0 - 8 hours) or AUC(3 - 8 hours) of CQA of a total of 18 subjects was examined. The prediction accuracy was calculated as the prediction result / actual measurement result. As a result, for AUC(0 - 8 hours), as shown in Figure 2 - 1, the simple regression model derived from the regression analysis was y = 4134.7×x + 59.6, and the coefficient of determination (r 2) was 0.28 and the statistical test quantity (p) was 0.014, indicating a significant regression relationship. For AUC (3 - 8 hours), the simple regression model derived by regression analysis as shown in Figure 2 - 2 was y = 2921.82×x + 11.70, and the coefficient of determination (r 2 ) was 0.35 and the statistical test quantity (p) was 0.0054, indicating a significant regression relationship. The prediction accuracy by cross - validation for AUC (0 - 8 hours) is shown in Figure 2 - 3, Table 2 - 1, and Table 2 - 2, and for AUC (3 - 8 hours) is shown in Figure 2 - 4, Table 2 - 1, and Table 2 - 2. With the regression model using the bacteria of the genus Ruminococcus as the explanatory variable, the AUC (0 - 8 hours) of CQA could be predicted with an accuracy of 60 - 140% compared to the measured values for 83% of the subjects, and the AUC (3 - 8 hours) of CQA could be predicted for 61% of the subjects.

[0051]

Table 2 - 1

[0052]

Table 2 - 2

[0053] Example 3 Prediction results of AUC (0 - 8 hours) and AUC (3 - 8 hours) of CQA using the relative abundance of bacteria of the genus Lachnospiraceae 1 1. Measurement of AUC(0-8 h) and AUC(3-8 h) of CQA by single-dose ingestion test of human chlorogenic acids and microbiome analysis of intestinal flora Using the intestinal flora information of 25 subjects obtained in the above Reference Example 1 and Example 1 and the AUC (0 - 8 hours) and AUC (3 - 8 hours) of CQA when 185 g of Helsea coffee (manufactured by Kao Corporation, containing 270 mg of chlorogenic acids in total, 170 mg of CQA, and 42 mg of FQA, the same as in Reference Example 1) was ingested, the following analysis was carried out.

[0054] 2. Identification of Lachnococcus 1 bacteria by correlation analysis Correlation analysis was performed between the AUC(0 - 8 hours) and AUC(3 - 8 hours) of 25 subjects and the relative abundance of each gut bacterial genus in fecal specimens to attempt to identify the bacterial genera involved in the absorption of CQA. Graphpad Prism (Graphpad Software) was used for the correlation analysis, and a Spearman or Pearson correlation coefficient (r) of linear correlation analysis of 0.4 or more and a test statistic (p) of 0.05 or less in the uncorrelated test were set as the significance level. Subjects with a relative abundance of 0 for the gut bacterial genera to be analyzed were excluded as outlier data. As a result of the analysis, the presence of bacteria belonging to the genus Lachnococcus 1 was confirmed in 10 subjects excluding those with a relative abundance of 0. As shown in Figure 3 - 1, a significant correlation was confirmed between the relative abundance of bacteria belonging to the genus Lachnococcus 1 and the AUC(0 - 8 hours) of CQA in the 10 subjects (r(Pearson)= - 0.70, p(Pearson)=0.024, r(Spearman)= - 0.79, p(Spearman)=0.013). Furthermore, as shown in Figure 3 - 2, a significant correlation was confirmed between the relative abundance of bacteria belonging to the genus Lachnococcus 1 and the AUC(3 - 8 hours) of CQA in the 10 subjects (r(Pearson)= - 0.63, p(Pearson)=0.051, r(Spearman)= - 0.76, p(Spearman)=0.015).

[0055] 3. Construction of a regression model for predicting AUC(0-8 h) and AUC(3-8 h) of CQA by Lachnococcus 1 bacteria and examination of prediction accuracy Using RStudio, linear simple regression analysis was performed on the data of 10 subjects with the relative abundance of bacteria belonging to the genus Lachnococcus 1 as the explanatory variable (x) and the AUC(0 - 8 hours) or AUC(3 - 8 hours) of CQA as the objective variable (y). Cross - validation was performed 4 times, with 2 - 3 subjects' data as test data and 7 - 8 subjects' data as model - building data each time, to examine the prediction accuracy for the AUC(0 - 8 hours) and AUC(3 - 8 hours) of CQA of a total of 10 subjects. The prediction accuracy was calculated as prediction result / actual measurement result. As a result, for AUC(0 - 8 hours), as shown in Figure 3 - 1, the simple regression model derived from the regression analysis was y = - 1126.2×x + 82.52, and the coefficient of determination (r 2) was 0.42, and the statistical test quantity (p) was 0.036, indicating a significant regression relationship. For AUC (3 - 8 hours), the simple regression model derived by regression analysis as shown in Fig. 3 - 2 was y = -806.58×x + 29.72, and the coefficient of determination (r 2 ) was 0.39, and the statistical test quantity (p) was 0.033, indicating a significant regression relationship. The prediction accuracy by cross - validation for AUC (0 - 8 hours) is shown in Fig. 3 - 3, Table 3 - 1, and Table 3 - 2, and for AUC (3 - 8 hours) is shown in Fig. 3 - 4, Table 3 - 1, and Table 3 - 2. Using the regression model with the bacteria of the genus Luminococcus 1 as the explanatory variable, the AUC (0 - 8 hours) of CQA in 90% of the subjects and the AUC (3 - 8 hours) of CQA in 40% of the subjects could be predicted with an accuracy of 60 - 140% compared to the measured values.

[0056]

Table 3 - 1

[0057]

Table 3 - 2

[0058] Example 4 Prediction Results of AUC (0 - 8 hours) and AUC (3 - 8 hours) of CQA Using the Relative Abundance Ratio of Bacteria of the Genus Cellulomonas 1. Measurement of AUC(0-8 h) and AUC(3-8 h) of CQA by single-dose ingestion test of human chlorogenic acids and microbiome analysis of intestinal flora Using the intestinal flora information of 25 subjects obtained in the above - mentioned Reference Example 1 and Example 1 and the AUC (0 - 8 hours) and AUC (3 - 8 hours) of CQA when 185 g of Helsea coffee (manufactured by Kao Corporation, containing 270 mg of total chlorogenic acids, 170 mg of internal CQA, and 42 mg of FQA, the same as in Reference Example 1) was ingested, the following analysis was carried out.

[0059] 2. Identification of Cellulomonas bacteria by correlation analysis Correlation analysis was performed between the AUC (0 - 8 hours) and AUC (3 - 8 hours) of 25 subjects and the relative abundance of each genus of intestinal bacteria in fecal samples, and an attempt was made to select the genera of bacteria involved in the absorption of CQA. Graphpad Prism (Graphpad Software) was used for the correlation analysis, and a Spearman or Pearson correlation coefficient (r) of linear correlation analysis of 0.4 or more and a test statistic (p) of 0.05 or less in the non - correlation test were set as the significance level. Subjects with a relative abundance of intestinal bacteria genera of 0 in the analysis target were excluded data. As a result of the analysis, the presence of Cellulomonas bacteria was confirmed in 12 subjects excluding those with a relative abundance of 0. As shown in Figure 4 - 1, a significant correlation was confirmed between the relative abundance of Cellulomonas bacteria and the AUC (0 - 8 hours) of CQA in the 12 subjects (r(Pearson)=0.70, p(Pearson)=0.010, r(Spearman)=0.80, p(Spearman)=0.003). As shown in Figure 4 - 2, a significant correlation was confirmed between the relative abundance of Cellulomonas bacteria and the AUC (3 - 8 hours) of CQA in the 13 subjects (r(Pearson)=0.84, p(Pearson)=0.0007, r(Spearman)=0.80, p(Spearman)=0.003).

[0060] 3. Construction of a regression model for predicting AUC(0-8 h) and AUC(3-8 h) of CQA by Cellulomonas bacteria and examination of prediction accuracy Using RStudio, linear simple regression analysis was performed on 12 subjects' data with the relative abundance of Cellulomonas bacteria as the explanatory variable (x) and the AUC (0 - 8 hours) or AUC (3 - 8 hours) of CQA as the objective variable (y). Cross - validation was performed 4 times, with 3 subjects' data as test data and 9 subjects' data as model - building data each time, and the prediction accuracy for the AUC (0 - 8 hours) and AUC (3 - 8 hours) of CQA of a total of 12 subjects was examined. The prediction accuracy was calculated as the prediction result / actual measurement result. As a result, for AUC (0 - 8 hours), as shown in Figure 4 - 1, the simple regression model derived from the regression analysis was y = 4878.38×x + 55.9, and the coefficient of determination (r 2) was 0.45, and the statistical test quantity (p) was 0.0101, indicating a significant regression relationship. For AUC (3 - 8 hours), as shown in Fig. 4 - 2, the simple regression model derived from the regression analysis was y = 3787.4×x + 11.0, and the coefficient of determination (r 2 ) was 0.67, and the statistical test quantity (p) was 0.0007, indicating a significant regression relationship. The prediction accuracy by cross - validation for AUC (0 - 8 hours) is shown in Fig. 4 - 3, Table 4 - 1, and Table 4 - 2, and for AUC (3 - 8 hours) is shown in Fig. 4 - 4, Table 4 - 1, and Table 4 - 2. Using the regression model with the bacteria of the genus Cellulomonas as the explanatory variable, the AUC (0 - 8 hours) of CQA in 100% of the subjects and the AUC (3 - 8 hours) of CQA in 58% of the subjects could be predicted with an accuracy of 60 - 140% compared to the measured values.

[0061]

Table 4 - 1

[0062]

Table 4 - 2

[0063] Example 5 Prediction results of AUC (0 - 8 hours) and AUC (3 - 8 hours) of CQA using the relative abundance ratio of bacteria of the genus Eggerthella 1. Measurement of AUC(0-8 h) and AUC(3-8 h) of CQA by single-dose ingestion test of human chlorogenic acids and microbiome analysis of intestinal flora Using the AUC (0 - 8 hours), AUC (3 - 8 hours) of CQA and the intestinal flora information when 25 subjects ingested 185 g of Helsia coffee (manufactured by Kao Corporation, containing 270 mg of total chlorogenic acids, 170 mg of CQA, and 42 mg of FQA, the same as in Reference Example 1) obtained in the above - mentioned Reference Example 1 and Example 1, the following analysis was carried out.

[0064] 2. Identification of Eggerthella bacteria by correlation analysis Correlation analysis was performed between the AUC (0 - 8 hours) and AUC (3 - 8 hours) of 25 subjects and the relative abundance of each genus of intestinal bacteria in fecal specimens, and an attempt was made to select the genera of bacteria involved in the absorption of CQA. Graphpad Prism (Graphpad Software) was used for the correlation analysis, and a Spearman or Pearson correlation coefficient (r) of linear correlation analysis of 0.4 or more and a test statistic (p) in the non - correlation test of 0.05 or less were set as the significance level. Subjects with a relative abundance of intestinal bacteria genera of 0 in the analysis target were excluded data. As a result of the analysis, the presence of Eggerthella bacteria was confirmed in 17 subjects excluding subjects with a relative abundance of 0. As shown in Figure 5 - 1, a significant correlation was confirmed between the relative abundance of Eggerthella bacteria and the AUC (0 - 8 hours) of CQA in the 17 subjects (r(Pearson)=0.51, p(Pearson)=0.036, r(Spearman)=0.68, p(Spearman)=0.004). Furthermore, as shown in Figure 5 - 2, a significant correlation was confirmed between the relative abundance of Eggerthella bacteria and the AUC (3 - 8 hours) of CQA in the 17 subjects (r(Pearson)=0.55, p(Pearson)=0.022, r(Spearman)=0.71, p(Spearman)=0.002).

[0065] 3. Construction of a regression model for predicting AUC(0-8 h) and AUC(3-8 h) of CQA by Eggerthella bacteria and examination of prediction accuracy Using RStudio, linear simple regression analysis was performed on the data of 17 subjects with the relative abundance of Eggerthella bacteria as the explanatory variable (x) and the AUC (0 - 8 hours) or AUC (3 - 8 hours) of CQA as the objective variable (y). Cross - validation was performed 4 times with 4 - 5 subjects' data as test data and 12 - 13 subjects' data as model - building data each time, and the prediction accuracy for the AUC (0 - 8 hours) or AUC (3 - 8 hours) of CQA of a total of 17 subjects was examined. The prediction accuracy was calculated as the prediction result / actual measurement result. As a result, for AUC (0 - 8 hours), as shown in Figure 5 - 1, the simple regression model derived from the regression analysis was y = 8883.7×x + 64.2, and the coefficient of determination (r 2) was 0.26, and the statistical test quantity (p) was 0.036, indicating a significant regression relationship. For AUC (3 - 8 hours), the simple regression model derived by regression analysis as shown in Fig. 5 - 2 was y = 5742.5×x + 16.84, and the coefficient of determination (r 2 ) was 0.26, and the statistical test quantity (p) was 0.022, indicating a significant regression relationship. Regarding the prediction accuracy by cross - validation for AUC (0 - 8 hours), it is shown in Fig. 5 - 3, Table 5 - 1, and Table 5 - 2. For AUC (3 - 8 hours), it is shown in Fig. 5 - 4, Table 5 - 1, and Table 5 - 2. Using the regression model with bacteria of the genus Eggerthella as the explanatory variable, the AUC (0 - 8 hours) of CQA could be predicted with an accuracy of 60 - 140% compared to the measured values in 71% of the subjects, and the AUC (3 - 8 hours) of CQA could be predicted with an accuracy of 60 - 140% compared to the measured values in 47% of the subjects.

[0066]

Table 5 - 1

[0067]

Table 5 - 2

[0068] Example 6 Prediction Results of AUC (3 - 8 hours) of CQA Using the Relative Abundances of Bacteria of the Genus Megamonas, Ruminococcus, Lachnospiraceae NK4A136 group, Cellulomonas, and Eggerthella 1. Development of a regression model for predicting the AUC (3 - 8 hours) of CQA based on the relative abundances of Megamonas bacteria, Ruminospira bacteria, Lachnoclostridium 1 bacteria, Cellulomonas bacteria, and Eggerthella bacteria, and examination of the prediction accuracy Using the relative abundances of bacteria of the genus Megamonas, Ruminococcus, Lachnospiraceae NK4A136 group, Cellulomonas, and Eggerthella identified by the methods shown in Examples 1 - 5 above as explanatory variables a, b, c, d, and e respectively, and the AUC (3 - 8 hours) of CQA as the target variable (y), a linear regression analysis was performed on the data of 25 subjects. Subjects with a relative abundance of 0 for each bacterial genus were not excluded and were treated as explanatory variable 0. First, the coefficient of determination (r 2) The combination of bacteria with the highest was examined. Subsequently, for the best combination, cross-validation was performed 4 times with 6-7 data per test and 18-19 data per model creation, and the prediction accuracy for the AUC (3-8 hours) of CQA for a total of 25 subjects was examined. The prediction accuracy was calculated as the prediction result / actual measurement result. As a result, in the regression analysis, the coefficient of determination was highest when all 5 bacteria were used as explanatory variables (r 2 = 0.44). The multiple regression model derived from the regression analysis was y = 47×a + 2393.08×b - 839.57×c - 1215.2×d + 3588.27×e + 15.104. The statistical test quantity (p) was 0.42 for Megamonas bacteria, 0.0024 for Ruminospira bacteria, 0.015 for Lachnococcus 1 bacteria, 0.32 for Cellulomonas bacteria, and 0.033 for Eggerthella bacteria, and there was a significant regression relationship in some bacterial genera. The prediction accuracy by cross-validation is shown in Fig. 6, Tables 6-1, and 6-2. With the regression model using the above 5 bacterial genera as explanatory variables, the AUC (3-8 hours) of CQA could be predicted within 40-160% of the actual measurement value for 80% of the subjects. Note that 40-160% is the coefficient of variation of 100% ± AUC (3-8 hours).

[0069]

Table 6-1

[0070]

Table 6-2

[0071] Example 7 Prediction results of AUC (0-8 hours) of CQA using the relative abundance of Ruminospira bacteria, Lachnococcus 1 bacteria, Cellulomonas bacteria, and Eggerthella bacteria 1. Development of a regression model for predicting the AUC (0 - 8 hours) of CQA based on the relative abundances of Ruminospira bacteria, Lachnoclostridium 1 bacteria, Cellulomonas bacteria, and Eggerthella bacteria, and examination of the prediction accuracy Using the methods shown in Examples 1-5 above, the relative abundance ratios of Megamonas bacteria, Ruminospira bacteria, Lachnoclostridium 1 bacteria, Cellulomonas bacteria, and Eggerthella bacteria were set as explanatory variables a, b, c, d, and e respectively, and linear regression analysis was performed on the data of 25 subjects with the AUC (0-8 hours) of CQA as the target variable (y). Subjects with a relative abundance ratio of 0 for each bacterial genus were not excluded and were treated as explanatory variable 0. First, the combination of bacteria with the highest coefficient of determination (r 2 ) in the regression analysis was examined. Then, for the best combination, cross-validation was performed 4 times with 6-7 data per round as test data and 18-19 data as model creation data, and the prediction accuracy for the AUC (0-8 hours) of CQA of a total of 25 subjects was examined. The prediction accuracy was calculated as the prediction result / measured result. As a result, the combination of bacteria with the highest coefficient of determination in the regression analysis was the 4 bacteria excluding Megamonas bacteria (r 2 = 0.50). The multiple regression model derived from the regression analysis was y = 4076.09×b – 1793.00×c – 4865.4×d + 3133.10×e + 80.125. The statistical test quantity (p) was 0.0041 for Ruminospira bacteria, 0.0035 for Lachnoclostridium 1 bacteria, 0.0277 for Cellulomonas bacteria, and 0.0248 for Eggerthella bacteria, indicating a significant regression relationship. The prediction accuracy by cross-validation is shown in FIG. 7, Tables 7-1, and 7-2. With the regression model using the above four bacterial genera as explanatory variables, the AUC (0-8 hours) of CQA could be predicted within 60-140% of the measured value for 84% of the subjects. Note that 60-140% is the coefficient of variation of 100% ± AUC (0-8 hours).

[0072]

Table 7-1

[0073]

Table 7-2

[0074] Example 8 Prediction results of AUC (3 - 8 hours) of FQA using the relative abundance ratios of Megamonas bacteria, Ruminococcus bacteria, Lachnoclostridium 1 bacteria, Cellulomonas bacteria, and Eggerthella bacteria 1. Measurement of the AUC (3 - 8 hours) of FQA and microbiome analysis of the gut microbiota by a single - dose human chlorogenic acids ingestion test Using the intestinal flora information of 25 subjects obtained in Reference Example 1 and Example 1 and the AUC (3 - 8 hours) of FQA when 185 g of Helsea coffee (manufactured by Kao Corporation, containing 270 mg of total chlorogenic acids, 170 mg of CQA, and 42 mg of FQA, the same as in Reference Example 1), the following analysis was performed.

[0075] 2. Identification of Megamonas bacteria, Ruminospira bacteria, Lachnoclostridium 1 bacteria, Cellulomonas bacteria, and Eggerthella bacteria by correlation analysis A correlation analysis was performed between the AUC (3 - 8 hours) of FQA of 25 subjects and the relative abundance ratios of each intestinal bacterial genus in fecal specimens to attempt to select bacterial genera related to the absorption of FQA. Graphpad Prism (Graphpad Software) was used for the correlation analysis, and a Spearman or Pearson correlation coefficient (r) of linear correlation analysis of 0.4 or more and a test statistic (p) of 0.05 or less in the non - correlation test were set as the significance level. Data with a relative abundance ratio of 0 for the intestinal bacterial genera to be analyzed were excluded. As a result of the analysis, a significant correlation was confirmed between the AUC (3 - 8 hours) of FQA and Megamonas bacteria (r(Pearson)=0.60, p(Pearson)=0.041, r(Spearman)=0.66, p(Spearman)=0.022)). A significant correlation was confirmed between the relative abundance ratio of Ruminococcus bacteria (r(Pearson)=0.46, p(Pearson)=0.042, r(Spearman)=0.42, p(Spearman)=0.062)). A significant correlation was confirmed between the Lachnococcus 1 bacteria (r(Pearson)= - 0.63, p(Pearson)=0.051, r(Spearman)= - 0.75, p(Spearman)=0.015)). A significant correlation was confirmed between the Cellulomonas bacteria (r(Pearson)=0.75, p(Pearson)=0.003, r(Spearman)=0.85, p(Spearman)=0.0001)). Furthermore, a significant correlation was confirmed between the Eggerthella bacteria (r(Pearson)=0.50, p(Pearson)=0.029, r(Spearman)=0.67, p(Spearman)=0.002)).

[0076] 3. Development of a regression model for predicting the AUC (3 - 8 hours) of FQA based on the relative abundances of Megamonas bacteria, Ruminospira bacteria, Lachnoclostridium 1 bacteria, Cellulomonas bacteria, and Eggerthella bacteria, and examination of the prediction accuracy Regarding the relative abundance ratios of Megamonas bacteria, Ruminococcus bacteria, Lachnococcus 1 bacteria, Cellulomonas bacteria, and Eggerthella bacteria as explanatory variables a, b, c, d, and e respectively, and the AUC (3 - 8 hours) of FQA as the target variable (y), a linear regression analysis was performed on the data of 25 subjects. Subjects with a relative abundance ratio of 0 for each bacterial genus were not excluded and were treated as explanatory variable 0. First, the combination of bacteria with the highest coefficient of determination (r 2 ) was examined. Then, for the best combination, cross - validation was performed 4 times with 6 - 7 data as test data and 18 - 19 data as model - building data each time, and the prediction accuracy for the AUC (3 - 8 hours) of FQA of a total of 25 subjects was examined. The prediction accuracy was calculated as the prediction result / actual measurement result. As a result, in the regression analysis, when all of the above 5 bacteria were used as explanatory variables, the highest coefficient of determination was shown (r 2(=0.40). The multiple regression model derived by regression analysis was y = 40.6×a + 1659.2×b - 642.3×c + 480.0×d + 3349.5×e + 10.55. The statistical test quantity (p) was 0.46 for Megamonas bacteria, 0.019 for Ruminococcus bacteria, 0.043 for Lachnococcus 1 bacteria, 0.68 for Cellulomonas bacteria, and 0.035 for Eggerthella bacteria. There was a significant regression relationship in some bacterial genera. The prediction accuracy by cross-validation is shown in Fig. 8, Tables 8-1, and 8-2. With the regression model using the five bacterial genera as explanatory variables, the AUC (3-8 hours) of FQA could be predicted with an accuracy of 40-160% relative to the measured values in 72% of the subjects. Note that 40-160% is the coefficient of variation of 100% ± AUC (3-8 hours).

[0077]

Table 8-1

[0078]

Table 8-2

[0079] Example 9 Prediction results of AUC (0-8 hours) of FQA using the relative abundances of Ruminococcus bacteria, Lachnococcus 1 bacteria, and Eggerthella bacteria 1. Measurement of the AUC (0 - 8 hours) of FQA and microbiome analysis of the gut microbiota by a single - dose human chlorogenic acids ingestion test Using the intestinal microbiota information of 25 subjects obtained in Reference Example 1 and Example 1 and the AUC (0-8 hours) of FQA when 185 g of Helsea coffee (manufactured by Kao Corporation, containing 270 mg of total chlorogenic acids, 170 mg of CQA, and 42 mg of FQA, the same as in Reference Example 1) was ingested, the following analysis was performed.

[0080] 2. Identification of Megamonas bacteria, Ruminospira bacteria, Lachnoclostridium 1 bacteria, Cellulomonas bacteria, and Eggerthella bacteria by correlation analysis Correlation analysis was performed between the AUC (0 - 8 hours) of FQA in 25 subjects and the relative abundance of each genus of intestinal bacteria in fecal specimens, and an attempt was made to select the bacterial genera involved in the absorption of FQA. Graphpad Prism (Graphpad Software) was used for the correlation analysis, and the significance level was set such that the Spearman or Pearson correlation coefficient (r) of the linear correlation analysis was 0.4 or more, and the test statistic (p) in the non - correlation test was 0.05 or less. Subjects with a relative abundance of 0 for the intestinal bacterial genera to be analyzed were excluded as outlier data. As a result of the analysis, a significant correlation was confirmed between the AUC (0 - 8 hours) of FQA and Megamonas bacteria (r(Pearson)=0.46, p(Pearson)=0.089, r(Spearman)=0.51, p(Spearman)=0.049). A significant correlation was confirmed between the relative abundance of Ruminococcus bacteria (r(Pearson)=0.49, p(Pearson)=0.040, r(Spearman)=0.46, p(Spearman)=0.046). A significant correlation was confirmed between the relative abundance of Lachnococcus 1 bacteria (r(Pearson)= - 0.77, p(Pearson)=0.043, r(Spearman)= - 0.46, p(Spearman)=0.30). A significant correlation was confirmed between the relative abundance of Cellulomonas bacteria (r(Pearson)=0.75, p(Pearson)=0.007, r(Spearman)=0.84, p(Spearman)=0.001). Furthermore, a significant correlation was confirmed between the relative abundance of Eggerthella bacteria (r(Pearson)=0.41, p(Pearson)=0.099, r(Spearman)=0.60, p(Spearman)=0.011).

[0081] 3. Development of a regression model for predicting the AUC (0 - 8 hours) of FQA based on the relative abundances of Ruminospira bacteria, Lachnoclostridium 1 bacteria, and Eggerthella bacteria, and examination of the prediction accuracy Linear regression analysis was performed on the data of 25 subjects, with the relative abundances of Megamonas bacteria, Ruminococcus bacteria, Lachnococcus 1 bacteria, Cellulomonas bacteria, and Eggerthella bacteria as explanatory variables a, b, c, d, and e respectively, and the AUC (0 - 8 hours) of FQA as the target variable (y). Subjects with a relative abundance of 0 for each bacterial genus were not excluded and were treated as explanatory variable 0. First, the coefficient of determination (r 2) examined the combination of bacteria with the highest Subsequently, for the best combination, cross-validation was performed four times with 6 - 7 data per test and 18 - 19 data per model building, and the prediction accuracy for the AUC (0 - 8 hours) of FQA of a total of 25 subjects was examined. The prediction accuracy was calculated as prediction result / actual measurement result. 2 As a result, in the regression analysis, when bacteria of the genus Ruminospira, Lachnoclostridium 1, and Eggerthella were used as explanatory variables, the highest coefficient of determination was shown (r = 0.33). The multiple regression model derived from the regression analysis was y = 2015.9×b - 998.8×c + 3717.6×e + 55.95. The statistical test quantity (p) was 0.031 for the genus Ruminospira, 0.030 for the genus Lachnoclostridium 1, and 0.091 for the genus Eggerthella, and there was a significant regression relationship in some genera of bacteria. The prediction accuracy by cross-validation is shown in Fig. 9, Tables 9 - 1, and 9 - 2.

[0082]

Table 9 - 1

[0083]

Table 9 - 2

[0084] Example 10 Prediction results of AUC (3 - 8 hours) of 5 - CQA using the relative abundances of Megamonas, Ruminospira, Lachnoclostridium 1, Cellulomonas, and Eggerthella 1. Measurement of the AUC (3 - 8 hours) of 5 - CQA and microbiome analysis of the gut microbiota by a single - dose human chlorogenic acids ingestion test Using the intestinal microbiota information of 25 subjects obtained in the above Reference Example 1 and Example 1 and the AUC (3 - 8 hours) of 5-CQA when 185 g of Helsea Coffee (manufactured by Kao Corporation, containing 270 mg of chlorogenic acids in total, 170 mg of CQA and 42 mg of FQA, the same as in Reference Example 1) was ingested, the following analysis was performed.

[0085] 2. Development of a regression model for predicting the AUC (3 - 8 hours) of 5 - CQA based on the relative abundances of Megamonas bacteria, Ruminospira bacteria, Lachnoclostridium 1 bacteria, Cellulomonas bacteria, and Eggerthella bacteria, and examination of the prediction accuracy The relative abundances of Megamonas bacteria, Ruminospira bacteria, Lachnococcus 1 bacteria, Cellulomonas bacteria, and Eggerthella bacteria were used as explanatory variables a, b, c, d, and e, respectively, and a linear regression analysis was performed on the data of 25 subjects with the AUC (3 - 8 hours) of 5-CQA as the objective variable (y). Subjects with a relative abundance of 0 for each bacterial genus were not excluded and were treated as explanatory variable 0. First, the combination of bacteria with the highest coefficient of determination (r 2 ) was examined. Then, for the best combination, cross-validation was performed 4 times with 6 - 7 data per time as test data and 18 - 19 data as model creation data, and the prediction accuracy for the AUC (3 - 8 hours) of 5-CQA of a total of 25 subjects was examined. The prediction accuracy was calculated as prediction result / measured result. As a result, in the regression analysis, when all of the above 5 bacteria were used as explanatory variables, the highest coefficient of determination was shown (r 2 = 0.33). The multiple regression model derived from the regression analysis was y = 19.21×a + 616.8×b - 195.6×c - 527.2×d + 913.7×e + 2.17. The statistical test quantity (p) was 0.332 for Megamonas bacteria, 0.019 for Ruminospira bacteria, 0.050 for Lachnococcus 1 bacteria, 0.163 for Cellulomonas bacteria, and 0.092 for Eggerthella bacteria, and there was a significant regression relationship for some bacterial genera. The prediction accuracy by cross-validation is shown in FIG. 10, Tables 10-1, and 10-2. With the regression model using the above 5 bacterial genera as explanatory variables, the AUC (3 - 8 hours) of 5-CQA could be predicted within 40 - 160% of the measured value for 40% of the subjects. Note that 40 - 160% is 100% ± the coefficient of variation of the AUC (3 - 8 hours).

[0086]

Table 10-1

[0087]

Table 10-2

Claims

1. A method for predicting the absorption amount of chlorogenic acids, comprising the following steps 1) to 2): 1) A step of obtaining the occupancy rate of one or more types of bacteria selected from five genera, namely Megamonas, Ruminospira, Lachnococcus 1, Cellulomonas, and Eggerthella, in the intestinal flora based on the information of the intestinal flora of the subject; 2) A step of calculating the absorption amount of chlorogenic acids when the subject ingests a chlorogenic acid-containing product based on the occupancy rate.

2. The method according to Claim 1, wherein the information of the intestinal flora is the information of the bacterial strain composition analyzed using a sample containing intestinal bacteria collected from the subject.

3. The method according to Claim 1 or 2, wherein the chlorogenic acid-containing product is a composition containing one or more selected from 3-caffeoylquinic acid, 4-caffeoylquinic acid, 5-caffeoylquinic acid, 3-feruloylquinic acid, 4-feruloylquinic acid, 5-feruloylquinic acid, 3,4-dicaffeoylquinic acid, 4,5-dicaffeoylquinic acid, and 3,5-dicaffeoylquinic acid.

4. The method according to Claim 1 or 2, wherein the absorption amount of the chlorogenic acids is the absorption amount of monocaffeoylquinic acid, 5-caffeoylquinic acid, or monopheruloylquinic acid.

5. The method according to Claim 4, wherein the absorption amount of the chlorogenic acids is the absorption amount of 5-caffeoylquinic acid.

6. The method according to Claim 1 or 2, wherein the absorption amount of the chlorogenic acids is the blood concentration measured from the area under the blood concentration-time curve of monocaffeoylquinic acid, 5-caffeoylquinic acid, or monopheruloylquinic acid at 0 to 8 hours or 3 to 8 hours after ingestion of the chlorogenic acid-containing product.

Images