Method for assisting diagnosis of mental stress

Abstract

To provide information to help an accurate diagnosis of the presence or absence and the degree of mental stress in a subject or the degree of mental stress.SOLUTION: A method for assisting a diagnosis of mental stress according to the present invention includes the steps of: measuring the concentrations of plural components contained in a biological sample collected from a subject; classifying the subject based on the characteristic of the personality; and generating stress diagnosis assistance information, which is information for diagnosing the degree of mental stress of the subject, based on the measured concentrations of the components and the characteristic of the personality. The plural components include betaine, creatine, GABA, creatinine, serotonin, kynurenic acid, 3-hydroxybutyric acid, and citric acid.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a method for assisting diagnosis of mental stress. [Background technology]

[0002] In recent years, the percentage of workers experiencing mental stress, strong worries, and anxiety related to work and workplace life has been increasing, and the number of cases in which mental disorders caused by mental stress and other factors are recognized as industrial accidents is also on the rise. In this specification, "mental stress" is also referred to as "stress." Working while feeling stressed could lead to a serious accident depending on the nature of the work. In light of this situation, the Ministry of Health, Labor and Welfare in Japan has established "Guidelines for Maintaining and Promoting the Mental Health of Workers," and is promoting the implementation of measures to maintain and promote the mental health of workers in the workplace (mental health care) (Non-Patent Document 1).

[0003] The above guidelines define "mental health problems" as "not only mental disorders and suicide classified as mental and behavioral disorders, but also a wide range of mental and behavioral problems that may affect a worker's physical and mental health, social life, and quality of life, such as stress, severe distress, and anxiety," and they take measures to prevent mental health problems before they occur or to detect them early.

[0004] For example, the "Act Partially Amending the Industrial Safety and Health Act" (Act No. 82 of 2014), which was promulgated on June 25, 2014, established a "stress check system" that involves tests (stress checks) to determine the level of psychological stress and face-to-face guidance based on the results, and calls for the proactive implementation of mental health care in the workplace, including the stress check system. Here, the "stress check system" refers to the entire series of initiatives in workplaces related to Article 66-10 of the Industrial Safety and Health Act.

[0005] The main methods for stress checks are interviews and blood tests. Typical interviews used for stress checks include the Occupational Stress Questionnaire and the PHQ9 (Patient Health Questionnaire-9). Another typical blood test is to measure the concentration of metabolites (biomarkers) in the blood, which are indicators of stress, using a liquid chromatograph or mass spectrometer, and determine the level of stress from the results. [Prior art documents] [Non-patent literature]

[0006] [Non-Patent Document 1] Ministry of Health, Labor and Welfare, Japan Occupational Health and Safety Agency, "Promoting Mental Health in the Workplace: Guidelines for Maintaining and Promoting Workers' Mental Health," March 2017, [Retrieved May 17, 2023], Internet<URL:https: / / www.mhlw.go.jp / content / 000560416.pdf> Summary of the Invention [Problem to be solved by the invention]

[0007] Because most medical interviews rely on the subject's subjective complaints and attitudes, it is difficult to establish absolute standards for determining the presence or absence of stress and the level of stress from the interview results. Meanwhile, blood tests determine the presence or absence and level of stress based on objective numerical values, such as biomarker measurements (concentrations), but do not take into account the influence of personality traits, such as the subject's character and temperament, on the measurements. For example, because an active, stress-tolerant subject and a timid, sensitive subject perceive stress differently, differences in biomarker measurements may occur even in the same stressful environment. This has led to problems such as overlooking severe stress in an active, stress-tolerant subject, or overestimating the stress felt by a timid, sensitive subject.

[0008] While the explanation here has been mainly on stress checks in the workplace and professional life, similar issues arise when it comes to stress checks in other settings, such as the home, living environment, and school life.

[0009] The problem to be solved by the present invention is to provide information that assists in accurately diagnosing the presence or absence and degree of mental stress in a subject. [Means for solving the problem]

[0010] In order to solve the above problems, the present invention provides a method for assisting diagnosis of mental stress, measuring the concentrations of a plurality of components contained in a biological sample collected from a subject; classifying subjects based on personality traits; and generating auxiliary stress diagnostic information, which is information for diagnosing the degree of stress of the subject, based on the measured concentrations of the components and the classification according to personality characteristics; The multiple ingredients include betaine, creatine, GABA, creatinine, serotonin, kynurenic acid, 3-hydroxybutyric acid, and citric acid. [Effects of the Invention]

[0011] In the mental stress diagnostic support method of the present invention, stress diagnostic support information is generated based on both the concentration of a specific metabolite in a biological sample, which is an objective indicator of mental stress, and a classification based on the subject's personality characteristics.By using this information, the diagnostician can appropriately diagnose the presence or absence and degree of mental stress in the subject. [Brief explanation of the drawings]

[0012] [Figure 1] 1 is a flow chart showing an embodiment of a method for assisting diagnosis of mental stress according to the present invention. [Figure 2]A graph showing the relationship between the actual and predicted scores of the PHQ9 obtained from a stress assessment model (Model 1) constructed using interview scores and the concentrations of 11 types of test substance. [Figure 3] A graph showing the relationship between the actual and predicted scores of the PHQ9 obtained from a stress assessment model (Model 2) constructed using the interview score and the concentrations of eight components. [Figure 4] A graph showing the relationship between the actual and predicted scores of the PHQ9 obtained from a stress assessment model (Model 3) constructed using the interview score and the concentrations of 10 components. [Figure 5] A graph showing the relationship between the actual and predicted scores of the PHQ9 obtained from a stress assessment model (Model 4) constructed using the interview score and the concentrations of 10 components. [Figure 6] A graph showing the relationship between the actual and predicted scores of the PHQ9 obtained from a stress assessment model (Model 5) constructed using the interview score and the concentrations of 10 components. DETAILED DESCRIPTION OF THE INVENTION

[0013] 1 is a process diagram showing one embodiment of a mental stress diagnosis support method according to the present invention. The mental stress diagnosis support method of this embodiment includes the steps of measuring the concentrations of multiple components contained in a biological sample collected from a subject, classifying the subject based on personality characteristics, and generating stress diagnosis support information, which is information for diagnosing the level of mental stress in the subject, based on the measured component concentrations and the classification based on the personality characteristics, where the multiple components contained in the biological sample include betaine, creatine, GABA, creatinine, serotonin, kynurenic acid, 3-hydroxybutyric acid, and citric acid.

[0014] The medical interviews conventionally used to evaluate the mental stress of a subject have been largely dependent on the subject's subjective complaints and attitudes, and blood tests have had the problem of not reflecting the subject's character, temperament, etc. (personality characteristics). In contrast, the mental stress diagnosis support method of this embodiment can generate stress diagnosis support information based on both the results of classifying the subject based on personality characteristics and the concentrations of multiple predetermined components contained in a biological sample collected from the subject, and can therefore appropriately evaluate the mental stress of the subject by referring to this information.

[0015] In the mental stress diagnostic support method of this embodiment, the step of measuring the concentrations of multiple components contained in a biological sample collected from a subject (hereinafter referred to as the "measurement step") and the step of classifying the subject based on their personality characteristics (hereinafter referred to as the "classification step") can be performed in either order. Note that the mental stress diagnostic support method of this embodiment does not include a step in which a diagnoser determines the presence or absence and degree of mental stress in the subject. Therefore, it does not fall under the category of a method for diagnosing humans.

[0016] In the measurement step, the "biological sample collected from the subject" is not particularly limited as long as the concentration of a component contained in the biological sample can be measured, and examples include blood, biological tissue, feces, urine, sweat, saliva, etc., with blood being preferred. The blood may be not only blood (whole blood) collected from the subject, but also serum, plasma, and other samples obtained by processing the blood. Serum and plasma can be obtained, for example, by leaving the blood to stand or centrifuging it.

[0017] The biological sample may be used as is for measuring the concentration of a component, or may be pretreated as needed before being used for measuring the concentration of a component. Examples of pretreatment include stopping the enzyme reaction in the biological sample, removing fat-soluble substances, removing proteins, etc. These pretreatments can be performed using known methods. The biological sample may also be diluted or concentrated as appropriate before being used for measuring the concentration of a component.

[0018] The method for measuring the concentration of a component in a biological sample may be selected from known methods appropriate for the type of component. For example, the concentration of the component can be measured by selecting and using a quantification method appropriate for the component to be measured from among quantification by nuclear magnetic resonance (NMR), quantification by acid-alkali neutralization titration, quantification by an amino acid analyzer, quantification by an enzymatic method, quantification using an aptamer such as a nucleic acid aptamer or a peptide aptamer, colorimetric quantification, etc. The concentrations of the components can also be measured using commercially available quantitative kits according to the requirements. Furthermore, the concentrations of the components can be measured using, for example, capillary electrophoresis, liquid chromatography, gas chromatography, mass spectrometry, etc., either alone or in appropriate combination. These measurement methods are particularly suitable for measuring multiple components simultaneously.

[0019] The measurement step is preferably performed by liquid chromatography mass spectrometry (LC-MS) because it can obtain analytical results in a short time from a small amount of biological sample containing multiple components and has excellent metabolite separation capabilities. When performing LC-MS, for example, the biological sample is appropriately pretreated, and the resulting peptide fragments are separated by liquid chromatography (LC) according to their respective retention times, and each component is output as multiple peaks. The output from this liquid chromatography is ionized by a mass spectrometer, and separated and detected according to their mass-to-charge ratio (m / z). As the mass spectrometer, in addition to a general single-type mass spectrometer, a triple quadrupole mass spectrometer, a Q-TOF mass spectrometer, a TOF-TOF mass spectrometer, an ion trap mass spectrometer, an ion trap time-of-flight mass spectrometer, etc. can be suitably used.

[0020] The components of the biological sample to be measured in the measurement step are metabolites known as biomarkers of mental stress, and as described below, include betaine, creatine, GABA, creatinine, serotonin, kynurenic acid, 3-hydroxybutyric acid, and citric acid, which have been demonstrated to improve the diagnostic performance of mental stress when combined with classification based on personality characteristics. In addition to these eight components, at least one of three components consisting of kynurenine, tryptophan, and 3-hydroxykynurenine may also be included in the components to be measured. Whether to use eight components as biomarkers of mental stress or nine to eleven components, which are the eight components plus at least one of the three components, can be determined based on the personality characteristics of the subject classified in the classification step. In this case, the classification step is performed before the measurement step.

[0021] As classification methods based on personality characteristics, it is preferable to use TACS22, HQ25, and Tachikawa Resilience Scale (TRS), but any other known appropriate method can also be used.

[0022] The TACS22 is a scale known as the Modern Depressive Temperament Scale (TACS) for assessing modern depressive temperament and consists of 22 questions. The 22 questions are divided into three groups: a group related to "avoidance of social roles," a group related to "low self-esteem," and a group related to "complaining." Subjects are then assigned a score of 0 to 4 based on the degree to which each item applies or does not apply to them. The total scores for all 22 items and the total scores for each of the three groups are used to classify subjects based on their personality characteristics.

[0023] The HQ25 is a known scale for assessing the degree of hikikomori, and consists of 25 questions. The 25 questions are divided into five groups related to "socialization," "isolation," "lack of emotional support," and "lack of motivation." As with the TACS22, subjects are asked to rate each item from 0 to 4 based on the degree to which it applies or does not apply to them. The total scores for all 25 items and the total scores for each of the five groups are then used to classify subjects based on their personality characteristics.

[0024] The Tachikawa Resilience Scale is a known method for assessing resilience against mental stress, and consists of 10 questions. Each question is assigned a score of 1 to 7 depending on the answer, and the higher the total score for the 10 items, the higher the resilience, i.e., the higher the ability to recover from mental stress and maintain a healthy state when subjected to mental stress.

[0025] The present invention will be described in more detail below with reference to experimental examples, but the present invention is not limited to the following experimental examples.

[0026] [Experimental Example 1] [1] Subjects The subjects were 259 Japanese adults (age range: mean 34 years (16-72 years), 132 men and 127 women).

[0027] [2] Recruitment of subjects The subjects were patients who were confirmed to have depressive symptoms by a diagnostic interview and healthy individuals who were confirmed to have no depressive symptoms. After providing an explanation based on the attached explanatory document to each individual, patients and healthy individuals who gave their consent to participate in this study were registered as experimental subjects, and the following clinical information was obtained from medical records and interviews. [Clinical information] Age, sex, height, weight, medical history information (diagnosis, duration of illness, treatment history, prescription history)

[0028] The interviews were conducted by psychiatrists and psychologists for patients and healthy individuals (a structured psychiatric diagnostic interview using the SCID (Structured Clinical Interview for DSM-IV) or a brief structured psychiatric interview using the MINI (Mini-International Neuropsychiatric Interview), as well as a general psychiatric interview that includes questions about upbringing history, lifestyle history, and current illness history). In cases where a structured psychiatric diagnostic interview using the SCID or a brief structured psychiatric interview using the MINI was conducted during medical treatment, this information was obtained from the medical records.

[0029] In addition, 25 ml of blood was collected for measurement of blood metabolites. All subjects gave informed consent.

[0030] [2] Classification based on personality traits All subjects were interviewed using questionnaires corresponding to the Tachikawa Resilience Scale (TRS), TACS22, and HQ25, and their personality traits were evaluated based on the results. The following eight items were used to evaluate personality traits: (1) Total score of all 10 questions on the Tachikawa Resilience Scale (2) Total score for questions related to "social role avoidance" in TACS22 (3) Total score of questions related to "Low Self-Esteem (LowSE)" in TACS22 (4) Total score of questions related to "Complaints" in TACS22 (5) Total score for questions related to "Socialization" in HQ25 (6) Total score for questions related to "Isolation" in HQ25 (7) Total score for questions related to "Lack of emotional support (Motivation)" in HQ25 (8) Total score for questions related to "Lack of Motivation (Emotional_Support)" in HQ25

[0031] [3] Preparation of plasma samples Plasma samples were collected by peripheral blood collection via venipuncture. For the extraction of water-soluble metabolites, 25 μL of plasma was mixed with 100 μL (4 Vol) of ice-cold methanol, vortexed, sonicated, and centrifuged at 14,000 × g, 4°C, for 15 minutes. The supernatant was collected in a 1.5 mL Eppendorf microtube and stored. For the amino acid extraction, 25 μL of plasma was mixed with 100 μL (4 Vol) of 0.1 M perchloric acid, vortexed, sonicated, and centrifuged at 14,000 × g, 4°C, for 15 minutes. The supernatant was collected in a 1.5 mL Eppendorf microtube.

[0032] For LC-MS measurements, the collected solution was diluted 10-fold with each mobile phase, and 5 μL of sample solution (equivalent to 0.1 μL of plasma) was applied. Note that the sample solution contained an internal standard corresponding to the metabolite to be measured.

[0033] [4] LC-MS measurement The sample solution was analyzed by LC-MS using an ultrafast triple quadrupole LC / MS / MS system, LCMS-8060 (Shimadzu Corporation). To measure a wide variety of water-soluble metabolites, the extracted solution was separated using a Shim-pack GIST PFPP (2.1 mm I.D. x 150 mmL, 3.0 μm (Shimadzu Corporation)). The mobile phase consisted of solvent A (water + 0.06% formic acid) and solvent B (acetonitrile + 0.1% formic acid). The gradient elution program was as follows: 1000 s, ... 0-2 min: 0% solvent B 5 min: 25% solvent B 10 min: 35% solvent B 11-15 min: 95% solvent B 15.1-20 min: 0% solvent B

[0034] The parameters for the positive / negative electrospray ionization mode were as follows: Ionization mode: ESI positive / negative Nebulizer gas flow rate: 3.0 L / min Heating gas flow rate: 10 L / min Drying gas flow rate: 10 L / min Interface temperature: 400℃ DL temperature: 300℃ Heat block temperature: 500°C CID gas: tuning file (270 kPa)

[0035] [5] Data analysis Multiple reaction monitoring (MRM) data were processed for peak collection and adjustment using LabSolutions software (Shimadzu Corporation). Data for the eight personality traits and plasma sample components were analyzed using Python (version 3.10.9). Data were autoscaled, and k-means clustering analysis and graph-drawing stress diagnostic model creation were performed using sklearn (a machine learning library), XGboost (a machine learning algorithm), and matplotlib (a graph-drawing library).

[0036] [6] Construction of a stress assessment model To construct the stress assessment model, we used the LC-MS measurement results (metabolite concentrations) of the sample solution described above and a dataset consisting of scores for eight items representing classifications based on personality traits as training data. The metabolite concentration data used to construct the stress assessment model were calculated by dividing the ion intensity of the metabolite obtained by LC-MS measurement by the ion intensity of the corresponding internal standard. In addition, all subjects underwent a stress check using the PHQ9 questionnaire, and the results were used as the training data. In this experimental example, the PHQ9 scores (0–27 points) were binned into 5-point intervals (0–4 points, 5–9 points, 10–14 points, 15–19 points, and 20 points or greater), and each category was assigned a score of 1 to 5 (score 1: 0–4 points, score 2: 5–9 points, score 3: 10–14 points, score 4: 15–19 points, and score 5: 20 points or greater). The breakdown of the PHQ9 scores of the 259 subjects was as follows: score 1: 109, score 2: 49, score 3: 34, score 4: 36, and score 5: 31.

[0037] Next, a learning model for stress assessment (stress assessment model) was constructed by performing a predetermined machine learning process using the above-mentioned learning data. The stress assessment model constructed in this experimental example outputs a PHQ9 score as an index of the degree of mental stress when a dataset consisting of scores for eight items representing classifications based on metabolite concentrations and personality characteristics is input. In other words, a stress assessment model capable of predicting a PHQ9 score is constructed from the scores for eight items representing classifications based on the subject's metabolite concentrations and personality characteristics.

[0038] The stress assessment model was constructed using a gradient boosting decision tree (regression model) algorithm. However, the learning algorithm is not limited to this, and other algorithms such as a logistic regression learning algorithm, a support vector machine (SVM) learning algorithm, a random forest learning algorithm, or a deep learning algorithm using a neural network can also be used.

[0039] In addition, all subjects' learning data was divided into training data (80%) and test data (20%), and the training data was used to build the model and adjust the hyperparameters. Specifically, the model-building data was divided into 10 parts to build a cross-validation model, and the hyperparameters were optimized based on the accuracy rate, root mean square error (RMSE), and mean absolute error (MAE) of each model. As a result, 11 components (betaine, creatine, GABA, creatinine, serotonin, kynurenic acid, 3-hydroxybutyric acid, citric acid, kynurenine, tryptophan, and 3-hydroxykynurenine) were extracted as biomarkers effective for determining whether or not a subject is experiencing mental stress.

[0040] Therefore, a stress assessment model was constructed again using a dataset consisting of scores for eight items that represent classifications based on the concentrations of the above 11 components and personality identification.The constructed stress assessment model was evaluated using test data, and the accuracy rates of the model constructed using the 11 components were higher than those of the 10 components obtained by excluding one of kynurenine, tryptophan, and 3-hydroxykynurenine from the 11 components, and the model constructed using the eight components obtained by excluding tryptophan and 3-hydroxykynurenine.

[0041] Below, the stress assessment model constructed using 11 components (betaine, creatine, GABA, creatinine, serotonin, kynurenic acid, 3-hydroxybutyric acid, citric acid, kynurenine, tryptophan, 3-hydroxykynurenine) will be referred to as Model 1, the stress assessment model constructed using 8 components (betaine, creatine, GABA, creatinine, serotonin, kynurenic acid, 3-hydroxybutyric acid, citric acid) will be referred to as Model 2, and the stress assessment models constructed using 10 components, excluding 3-hydroxykynurenine, tryptophan, and kynurenine from the above 11 components, will be referred to as Models 3 to 5. The stress assessment accuracy of each model was evaluated.

[0042] Figures 2 to 6 show the evaluation results for models 1 to 5. Figures 2 to 6 plot the evaluation results when test data and training data were used. In Figures 2 to 6, the evaluation results for test data and training data are plotted with black circles of different shades, but in the actual figures, the evaluation results when test data were used are plotted with red circles, and the evaluation results when training data were used are plotted with blue circles. Note that the darker the circle, the larger the amount of data. In Figures 2 to 6, the horizontal axis represents the PHQ9 score (1 to 5) (actual measured value) for each learning data, and the vertical axis represents the output value (predicted value) when each learning data is input into the stress assessment model.

[0043] In addition, the determination constant (R 2 ), root mean square error (RMSE), and mean absolute error (MAE) were calculated. The results are shown in Table 1 below.

[0044] [Table 1]

[0045] As can be seen from Table 1, models 1 to 4 all have a constant of determination (R 2 ) showed a high value of 0.70 or more, and the determination constant (R 2 ) were lower than those of the training data, but all were above 0.50. In addition, Model 5 had a similar coefficient of determination (R 2 ) were 0.56 and 0.47, respectively, which were lower than the other models, but were all around 0.50. These results suggest that models 1 to 5 are all useful models for predicting PHQ9 scores using a dataset consisting of scores on eight items representing classifications based on predetermined component concentrations and personality characteristics.

[0046] The PHQ9 score can serve as supplementary information for diagnosing the presence or absence and degree of stress (corresponding to the auxiliary information for stress diagnosis of the present invention). Therefore, it has been demonstrated that the eight metabolites (betaine, creatine, GABA, creatinine, serotonin, kynurenic acid, 3-hydroxybutyric acid, and citric acid), the three components kynurenine, tryptophan, and 3-hydroxykynurenine added to these eight components (11 components), or the 10 components excluding any one of kynurenine, tryptophan, and 3-hydroxykynurenine from the 11 components, can serve as biomarkers for mental stress when combined with classification based on personality traits. It has been found that auxiliary information for diagnosis that improves the accuracy of stress diagnosis can be provided based on the concentrations of the biomarkers and classification based on personality traits.

[0047] In the above Experimental Example 1, the input data was a dataset consisting of metabolite concentrations in blood collected from the subjects and scores on eight items representing classifications based on personality characteristics, and the training data was the results of a stress check conducted on the subjects through a medical interview using the PHQ9, but the correct data could also be the results of an occupational stress questionnaire, instead of the results of a stress check conducted through a medical interview using the PHQ9. The results of the occupational stress questionnaire, like the PHQ9 score, can be supplementary information for diagnosing the presence or absence and degree of stress.

[0048] [Aspect] It will be apparent to those skilled in the art that the above-described exemplary embodiments are examples of the following aspects.

[0049] (Item 1) A method for assisting diagnosis of mental stress according to one aspect of the present invention includes: measuring the concentrations of a plurality of components contained in a biological sample collected from a subject; classifying subjects based on personality traits; and generating auxiliary stress diagnostic information, which is information for diagnosing the degree of mental stress of the subject, based on the measured concentrations of the components and the classification according to personality characteristics; A method for assisting in the diagnosis of mental stress, wherein the plurality of components include betaine, creatine, GABA, creatinine, serotonin, kynurenic acid, 3-hydroxybutyric acid, and citric acid.

[0050] According to the mental stress diagnosis assistance method of paragraph 1, it is possible to generate information useful for diagnosing whether or not a subject is experiencing mental stress, and information useful for diagnosing the degree of mental stress if the subject is experiencing mental stress, from the concentrations of multiple components contained in a biological sample collected from the subject and the personality characteristics of the subject.Therefore, by providing this information to the diagnoser, it is possible to improve the accuracy of the diagnoser's stress diagnosis.

[0051] (Item 2) The method for assisting in the diagnosis of mental stress according to item 2 is the method for assisting in the diagnosis of mental stress according to item 1, wherein the plurality of components further include kynurenine, tryptophan, and 3-hydroxykynurenine.

[0052] According to the mental stress diagnosis assistance method of the second aspect, it is possible to further improve the accuracy of stress diagnosis by the person making the diagnosis.

[0053] (Item 3) The diagnostic support method for mental stress according to item 3 is the diagnostic support method for mental stress according to item 1 or 2, in which the concentration of components contained in the biological sample is measured by liquid chromatography mass spectrometry.

[0054] According to the mental stress diagnostic support method of the third aspect, the concentrations of multiple components contained in a small amount of biological sample can be measured in a short period of time.

[0055] (4) The method for assisting in the diagnosis of mental stress according to paragraph 4 is the method for assisting in the diagnosis of mental stress according to any one of paragraphs 1 to 3, in which the classification according to personality characteristics uses one or more selected from TACS-22, HQ-25, and Tachikawa Resilience Score.

[0056] According to the mental stress diagnostic support method of paragraph 4, it is possible to provide stress diagnostic support information that makes it possible to determine the presence or absence and degree of stress taking into account the personality and temperament of the subject.

[0057] (Item 5) The mental stress diagnostic support method according to item 5 is a mental stress diagnostic support method according to any one of items 1 to 4, further comprising a step of providing a report including the generated stress diagnostic support information and the results of an occupational stress questionnaire.

[0058] (Item 6) The mental stress diagnostic support method according to item 6 is the mental stress diagnostic support method according to any one of items 1 to 4, further comprising the step of providing a report including the generated stress diagnostic support information and a PHQ9 score.

[0059] According to the mental stress diagnosis assistance method of paragraph 5 or 6, a report including stress diagnosis assistance information and information equivalent to the results of the occupational stress questionnaire and the PHQ9 score, which are information effective for diagnosing the presence or absence and degree of mental stress in the subject's work or working life, can be provided to the diagnoser.

[0060] (Item 7) The mental stress diagnostic support method according to item 7 is a mental stress diagnostic support method according to any one of items 1 to 4, in which the stress diagnostic support information is a machine learning model that uses the measured component concentrations and classifications based on personality characteristics as input data and information corresponding to the results of interviews for stress checks as output data.

[0061] According to the mental stress diagnostic assistance method of paragraph 7, it is possible to provide information that is effective for diagnosing the presence or absence and degree of mental stress in a subject based on subjective and objective indicators.

[0062] (Item 8) Another aspect of the present invention relates to a method for generating a determination model for determining the degree of mental stress of a subject, the generation method comprising: A step of acquiring a data set in which the concentrations of multiple components contained in the biological sample collected from the subject and the personality characteristics of the subject are used as input data, and information corresponding to the results of an occupational stress questionnaire or the score of PHQ9 is used as correct answer data; training the decision model by performing machine learning based on the dataset; Evaluating the suitability of the judgment model using a predetermined evaluation data set; It has the following characteristics.

[0063] In the method for generating the above-mentioned judgment model, the plurality of components include, for example, betaine, creatine, GABA, creatinine, serotonin, kynurenic acid, 3-hydroxybutyric acid, and citric acid. According to the generation method, by inputting the concentrations of the plurality of components contained in a biological sample collected from a subject and the personality characteristics of the subject, it is possible to generate a judgment model that can output information useful for diagnosing whether the subject is experiencing mental stress, and, if experiencing mental stress, information useful for diagnosing the degree of stress, and also to evaluate the suitability of the generated judgment model.

[0064] (Item 9) The method for generating a judgment model according to item 9 is the method for generating a judgment model according to item 8, The evaluation dataset may use input data that is partially or completely different from the input data included in the dataset used in the process of training the judgment model, and may use information corresponding to the results of an occupational stress questionnaire or the score of PHQ9 as correct answer data.

[0065] According to the method for generating a judgment model relating to paragraph 9, it is possible to correctly evaluate whether the judgment model obtained by machine learning is suitable as a model that provides information that is effective in diagnosing whether or not a subject is experiencing mental stress and the degree of mental stress.

Claims

1. Measuring the concentrations of a plurality of components contained in a biological sample collected from a subject; classifying subjects based on personality traits; and generating auxiliary stress diagnostic information for diagnosing the degree of mental stress of the subject based on the measured component concentrations and the classification according to personality characteristics, A method for assisting in diagnosis of mental stress, wherein the plurality of components include betaine, creatine, GABA, creatinine, serotonin, kynurenic acid, 3-hydroxybutyric acid, and citric acid.

2. The method for assisting diagnosis of mental stress according to claim 1 , wherein the plurality of components further includes kynurenine, tryptophan, and 3-hydroxykynurenine.

3. 3. The method for assisting diagnosis of mental stress according to claim 1, wherein concentrations of components contained in the biological sample are measured by liquid chromatography-mass spectrometry.

4. 3. The method for assisting diagnosis of mental stress according to claim 1 or 2, wherein the classification based on personality characteristics utilizes one or more selected from TACS-22, HQ-25, and Tachikawa Resilience Score.

5. 3. The method for assisting diagnosis of mental stress according to claim 1, further comprising the step of providing a report including the generated auxiliary stress diagnosis information and a result of an occupational stress questionnaire.

6. 3. The method for assisting diagnosis of mental stress according to claim 1, further comprising the step of providing a report including the generated auxiliary stress diagnosis information and a PHQ9 score.

7. 3. The method for assisting diagnosis of mental stress according to claim 1 or 2, wherein the stress diagnosis auxiliary information is a machine learning model that uses the measured component concentrations and classification based on personality characteristics as input data and information corresponding to the results of a medical interview for a stress check as output data.

8. 1. A method for generating a determination model for determining a level of mental stress of a subject, comprising: A step of acquiring a data set in which the concentrations of a plurality of components contained in the biological sample collected from the subject and the personality characteristics of the subject are used as input data, and information corresponding to the result of an occupational stress questionnaire or the score of PHQ9 is used as correct answer data; training the decision model by performing machine learning based on the dataset; Evaluating the suitability of the judgment model using a predetermined evaluation data set; A method for generating the same.

9. 9. The method for generating a judgment model as described in claim 8, wherein the evaluation dataset uses as input data data that is partially or completely different from the input data included in the dataset used in the process of learning the judgment model, and information corresponding to the results of an occupational stress questionnaire or a score on PHQ9 is used as correct answer data.

Images