Screening method for people with borderline diabetes

Abstract

To provide a method for screening a borderline diabetic patient easily and more accurately than before by using a simple blood sugar level.SOLUTION: The present invention is a screening method for a borderline diabetic patient using a logistic regression analysis model with a blood sugar level engagement parameter as an explanatory variable and with the presence or absence of borderline diabetes as an objective variable, and the method includes: an analysis step of determining a partial regression coefficient of the logistic regression analysis model by using the actual result values of the explanatory variable and objective variable; and an estimation step of inputting the blood sugar level engagement parameter of a person to be screened to the explanatory variable of the logistic regression analysis model to which the determined partial regression coefficient is set, and estimating the presence or absence of borderline diabetes in the person to be screened based on output data of the objective variable. The blood sugar level engagement parameter consists of a simple fasting blood sugar level and the age and / or sex, and BMI may be included in the parameter when the parameter includes the age and sex.SELECTED DRAWING: None

Description

[Technical Field]

[0001] The present invention relates to a method for screening for borderline diabetes mellitus patients, and more particularly to a method for screening for borderline diabetes mellitus patients using a simplified blood glucose level. [Background technology]

[0002] Blood sugar levels are maintained at a fairly constant level by the pancreatic islets of Langerhans secreting insulin, a type of pancreatic hormone, in response to blood sugar levels, which then controls the uptake of sugar from the blood into cellular tissues. When the action of insulin declines and blood sugar levels remain elevated above a certain level, diabetes is diagnosed.

[0003] Diabetes is broadly classified into type I and type II depending on the cause. Type I diabetes is a disease in which insulin-secreting pancreatic beta cells are destroyed for some reason, resulting in insufficient insulin secretion, or even if insulin is secreted, it leads to a hyperglycemic state due to insulin resistance in peripheral tissues. On the other hand, type II diabetes is a disease in which hyperglycemic state occurs due to a decrease in insulin secretion and insulin action caused by a high-calorie, high-fat diet, lack of exercise, etc. Type II diabetes is well known as a lifestyle-related disease.

[0004] Diabetes has few noticeable symptoms, so it can easily progress without you even realizing it, but symptoms such as dry mouth, excessive drinking, excessive urination, and weight loss appear over time, and in severe cases, it can cause loss of consciousness, coma, etc. Diabetes can also cause complications such as arteriosclerosis, diabetic retinopathy, diabetic nephropathy, and diabetic mental disorders.

[0005] Between diabetics and healthy people, there is a category known as prediabetes. Prediabetes includes not only people at risk of developing diabetes, but also those whose diabetes has improved and healthy people whose glucose tolerance has temporarily deteriorated for some reason. Prediabetes has no noticeable symptoms, is more likely to progress to diabetes than normal people, and is at higher risk of developing arteriosclerotic complications. According to the 2019 National Health and Nutrition Survey, there are an estimated 22.5 million people in Japan who are at risk of developing diabetes. In this specification, individuals classified as prediabetes may be referred to as borderline diabetic individuals or simply as borderline individuals.

[0006] For people with borderline diabetes, HbA1c and two blood glucose levels are used to determine the following conditions:

number

[0007] HbA1c refers to the ratio of glycomoglobin, which is hemoglobin (Hb) bound to glucose. HbA1c increases when hyperglycemic conditions persist, but is not affected by temporary changes in blood glucose levels due to meals, etc., and serves as an indicator of blood glucose levels over the past one to two months. HbA1c is usually measured at clinical laboratories using HPLC, immunoassay, or enzymatic methods on blood samples.

[0008] Blood glucose level is the concentration of glucose in the blood. Because blood glucose levels fluctuate throughout the day, such as increasing after meals, it is necessary to clarify collection conditions such as fasting and glucose loading. Blood glucose levels are generally measured at clinical laboratories using plasma or serum separated after blood collection using the GOD method. Detection methods for the GOD method include colorimetric (test strip method) and electrode method. The GOD colorimetric method utilizes a color reaction with a dye produced by hydrogen peroxide generated by the reaction between glucose and GOD. The GOD electrode method transports electrons generated by the reaction between glucose and GOD to an electrode and measures the amount of current.

[0009] Borderline diabetes can be diagnosed by meeting the above criteria 1) to 3). Since an HbA1c of 6.5% roughly corresponds to a fasting blood glucose level of 126 mg / dL and a 2-hour OGTT value of 200 mg / dL, HbA1c can be roughly determined from fasting blood glucose levels and 2-hour OGTT values. On the other hand, there are many individuals who are classified as non-borderline based on fasting blood glucose levels but as borderline based on a 75g OGTT. Therefore, to thoroughly identify or screen individuals with borderline diabetes, it is necessary to measure fasting blood glucose levels and 2-hour 75g OGTT values ​​at medical institutions or clinical testing facilities.

[0010] Methods for easily predicting blood glucose levels and borderline diabetes have been developed. For example, Patent Document 1 discloses a "blood glucose screening device for predicting a user's blood glucose level, comprising: an acquisition unit that acquires the user's blood glucose measurement, the user's HbA1c measurement, and the user's health checkup results; a level determination unit that determines whether the user is normal, borderline, or diabetic based on the blood glucose measurement, the HbA1c measurement, and the health checkup results; and a prediction unit that predicts the user's future fasting blood glucose level using the determination result and the user's past fasting blood glucose measurement" (Claim 1). Table 1 summarizes the conditions for determining whether a person is in the borderline diabetes range by the level determination unit in the flowchart of Figure 6 of Patent Document 1. A person who satisfies the determination conditions in any step of Table 1 is determined to be in the borderline diabetes range.

[0011] [Table 1]

[0012] The discrimination conditions of the invention of Patent Document 1 are different from the above 1) to 3), but as shown in Table 1, it is necessary to measure HbA1c, fasting blood glucose level, 1-hour postprandial blood glucose level, and 2-hour postprandial blood glucose level.

[0013] Non-Patent Document 1 examines how diagnostic criteria should be handled from the perspective of medical screening institutions by analyzing glucose tolerance function tests of patients undergoing health checkups. This document suggests that screening for borderline diabetes using only fasting blood glucose levels with a cutoff value of ≥ 110 mg / dL is useful for primary screening. Furthermore, it suggests that a method using cutoff values ​​of fasting blood glucose ≥ 110 mg / dL, 1-hour OGTT ≥ 185 mg / dL, and 2-hour OGTT ≥ 150 mg / dL can efficiently and accurately screen for borderline diabetes. This screening method also requires measuring three types of blood glucose: fasting blood glucose, 1-hour OGTT, and 2-hour OGTT.

[0014] The use of a simple blood glucose meter (also known as Self Monitoring of Blood Glucose, SMBG) is known as a simple method for measuring blood glucose levels. SMBG primarily follows the GOD method, in which glucose oxidase enzyme immobilized on a measurement sensor reacts with glucose in the blood, and the blood glucose level is determined from the generated hydrogen peroxide using a colorimetric or electrode method. According to Non-Patent Document 2, there is a good correlation between simple blood glucose levels measured using fingertip whole blood with SMBG and blood glucose levels measured using venous blood plasma at a medical institution, etc. SMBG is a highly convenient meter because it allows self-measurement without the need for a medical professional and without the constraints of location or time.

[0015] Non-Patent Document 3 qualitatively analyzes the effects of utilizing self-monitoring of blood glucose in people who are on the borderline of diabetes, clarifies the significance and effects of using SMBG, and examines the effectiveness of SMBG as a health guidance method in public health activities (abstract). However, as shown in Comparative Example 1 below, when screening for people on the borderline of diabetes, a logistic regression analysis was performed using only simple blood glucose values ​​as glucose-related parameters, and the screening efficiency for borderline diabetes did not exceed 80%. [Prior art documents] [Patent documents]

[0016] [Patent Document 1] WO2017 / 073713 (Blood glucose level prediction device, blood glucose level prediction method, and computer-readable recording medium) [Non-patent literature]

[0017] [Non-Patent Document 1] Nishida Yoshiko et al., "Considerations on the handling of diagnostic criteria from the perspective of screening institutions," Diabetes, Vol. 41, (1998), A57-A58 [Non-patent document 2] Fumiko Miura et al., "Clinical evaluation of five simple blood glucose monitors - effects of blood glucose concentration, patient and measurement environment -", Diabetes, 52, (10), 865-870, 2009 [Non-patent document 3] Kawasaki, Chie et al., "Study on the significance and effectiveness of self-monitoring of blood glucose in those on the borderline diabetic spectrum: A qualitative analysis using focus group interviews," Journal of Japanese Public Health, Vol. 56, (2009), No. 12, pp. 875-882 Summary of the Invention [Problem to be solved by the invention]

[0018] It is desirable to detect people who are at risk of developing diabetes early and to improve their lifestyle through dietary therapy, exercise therapy, etc., depending on their condition, as well as administer medication. There is also a demand for monitoring and clinical trials of functional foods and pharmaceuticals targeting people at risk of developing diabetes. It is desirable to develop a simple and accurate method for screening for borderline diabetes among clinical trial participants who are suspected of having diabetes or are at risk of developing diabetes.

[0019] Therefore, an object of the present invention is to provide a method for screening for borderline diabetes using a simple blood glucose level in a simple manner with higher accuracy than conventional methods. [Means for solving the problem]

[0020] The present inventors have investigated blood glucose level-related parameters other than the fasting simple blood glucose value, and have found that when logistic regression analysis is performed adding age, sex and / or BMI to the fasting simple blood glucose value, the screening efficiency for people on the borderline of diabetes is improved, and have completed the present invention based on this finding.

[0021] That is, the present invention provides a method for screening for borderline diabetes using a logistic regression analysis model with blood glucose level-related parameters as explanatory variables and the presence or absence of borderline diabetes as a response variable, comprising: an analysis step of calculating partial regression coefficients of a logistic regression analysis model using actual values ​​of the explanatory variables and the response variable; and an estimation step of inputting data on blood glucose level-related parameters of the screening subject into explanatory variables of a logistic regression analysis model in which the calculated partial regression coefficients have been set, and estimating the presence or absence of borderline diabetes in the screening subject based on output data of the objective variable, The method for screening for those on the borderline diabetic range is provided, characterized in that the blood glucose level-related parameters consist of a fasting simple blood glucose value and age and / or sex, but if age and sex are included, BMI may also be included.

[0022] Patent Document 1 does not teach or suggest the use of simplified blood glucose levels to accurately screen for people who are on the borderline of diabetes, as in the present invention. Moreover, Patent Document 1 requires that HbA1c, fasting blood glucose levels, blood glucose levels 1 hour after a meal, and blood glucose levels 2 hours after a meal be measured in order to identify people who are on the borderline of diabetes.

[0023] The blood glucose level-related parameters preferably consist of a fasting simple blood glucose level and age, or a fasting simple blood glucose level, age, sex, and BMI.

[0024] The subjects to be screened in the method of the present invention are particularly those suspected of having borderline diabetes. Such subjects to be screened include, for example: (1) Fasting blood glucose level is 100-125 mg / dL, (2) Random blood glucose level is 140-199 mg / dL, and (3) Urine sugar is positive A random blood glucose level means a blood glucose level measured regardless of the time relationship between a meal and the time of blood sampling.

[0025] The present invention also provides a screening device for borderline diabetes using a logistic regression analysis model with blood glucose level-related parameters as explanatory variables and the presence or absence of borderline diabetes as a response variable, comprising: an analysis means for calculating partial regression coefficients of a logistic regression analysis model using actual values ​​of the explanatory variables and the response variable; and an estimation means for inputting data on blood glucose level-related parameters of the screening subject into explanatory variables of a logistic regression analysis model in which the calculated partial regression coefficients are set, and estimating the presence or absence of borderline diabetes in the screening subject based on output data of the objective variable, The screening device for those on the borderline of diabetes is characterized in that the blood glucose level-related parameters consist of a fasting simple blood glucose value and age and / or gender, but if age and gender are included, BMI may also be included.

[0026] The present invention also provides a screening program for borderline diabetes patients that enables a computer to implement a function of predicting whether a patient is in the borderline diabetes state using a logistic regression analysis model that uses blood glucose level-related parameters as explanatory variables and the presence or absence of borderline diabetes as a response variable, comprising: an analysis function for calculating partial regression coefficients of a logistic regression analysis model using actual values ​​of the explanatory variables and the objective variable; and an estimation function for inputting data on blood glucose level-related parameters of the screening subject into explanatory variables of a logistic regression analysis model in which the calculated partial regression coefficients are set, and estimating whether or not the screening subject has borderline diabetes based on output data of the objective variable, The screening program for people on the borderline of diabetes is provided, characterized in that the blood glucose level-related parameters consist of a fasting simple blood glucose value and age and / or gender, but if age and gender are included, BMI may also be included.

[0027] The present invention also provides a computer-readable information recording medium having the above-mentioned screening program for borderline diabetes patients recorded thereon. [Effects of the Invention]

[0028] As shown in the examples described below, when simple blood glucose level, age, sex, and BMI are included as candidate explanatory variables (parameters related to blood glucose levels) according to the present invention, the screening efficiency for those in the borderline or non-borderline diabetes range is high, at 80% or more. In other words, the screening method and device of the present invention enable efficient and practical screening of those in the borderline diabetes range by using a simple blood glucose meter and certain parameters related to blood glucose levels. The screening method of the present invention is superior to conventional diabetes stratification methods such as those described in Patent Document 1 in that it omits the need for blood glucose and HbA1c measurements at medical institutions or clinical testing institutions.

[0029] According to the screening method of the present invention, it is possible to easily and accurately screen participants for borderline diabetes before monitoring functional foods or clinical trials of pharmaceuticals targeting those at risk of developing diabetes, thereby effectively reducing the number of participants and thereby significantly reducing the work and cost required for a definitive diagnosis of borderline diabetes based on blood glucose levels (plasma), post-glucose load blood glucose levels, and HbA1c.

[0030] The screening method of the present invention can also be used daily, making a significant contribution to the early detection of people who are on the borderline of diabetes. People who have been diagnosed with borderline diabetes and are undergoing lifestyle changes and medication can manage their condition on a daily basis by using a computer, scale, smartwatch, etc. equipped with the program of the present invention. [Brief explanation of the drawings]

[0031] [Figure 1] 1 is a block diagram showing the overall configuration of a screening device for borderline diabetes patients according to one embodiment of the present invention. [Figure 2] FIG. 2 is a block diagram showing functions of the computer main body of the device. [Figure 3] 10 is a flowchart illustrating an example of a screening process for people on the borderline of diabetes, which is performed by the computer main body based on the values ​​of explanatory variables input to the input section of the device. [Figure 4] 1 shows the database and screening estimation results stored in the device. DETAILED DESCRIPTION OF THE INVENTION

[0032] An embodiment of the present invention will be described with reference to the accompanying drawings. The screening method of the present invention is carried out using a screening device for borderline diabetes patients as shown in Fig. 1. Fig. 1 is a block diagram showing the overall configuration of the screening device for borderline diabetes patients, Fig. 2 is a block diagram showing the functions of the computer main body of the device, Fig. 3 is a flowchart explaining an example of a screening process for borderline diabetes patients performed by the computer main body based on the values ​​of explanatory variables input to the input section of the device, and Fig. 4 shows an example of a database and an output of estimation results.

[0033] As shown in Fig. 1, the screening device for borderline diabetes patients 1 of the present invention includes a computer main unit 10, an external storage device 20, an input unit 30, and an output unit 40. The screening device for borderline diabetes patients 1 is usually a stand-alone computer, but may also be incorporated into a simple blood glucose meter, a weight scale, a mobile terminal, a smart watch, etc.

[0034] The computer main body 10 includes a central processing unit (CPU) 11, a main memory 12, and an interface 13 for connecting to the outside.

[0035] The computer main body 10 is caused to function as an analysis means for determining partial regression coefficients of a logistic regression analysis model using actual values ​​of the explanatory variables and dependent variables, as shown in FIG. 2, by a screening program for borderline diabetes patients described later, and as an estimation means for inputting data on blood glucose level-related parameters of the screening subject into the explanatory variables of the logistic regression analysis model in which the determined partial regression coefficients have been set, and for estimating whether the screening subject has borderline diabetes based on output data of the dependent variables.

[0036] The external storage device 20 may be an HDD, an optical / magnetic disk, a flash memory (e.g., an SSD, a USB memory, or an SD memory card), etc. The external storage device 20 stores a borderline diabetes screening program, a database 21, etc. The contents stored in the external storage device 20 are read into the main memory 12 of the computer main body 10 via the interface 13 as necessary.

[0037] The above-mentioned screening program for borderline diabetes is a screening program for enabling a computer to perform a function of predicting whether a subject is at the borderline stage of diabetes using a logistic regression analysis model with blood glucose level-related parameters as explanatory variables and the presence or absence of borderline diabetes as a dependent variable. The program includes an analytical function of calculating partial regression coefficients of the logistic regression analysis model using actual values ​​of the explanatory variables and dependent variables, and an estimation function of inputting data on the blood glucose level-related parameters of the subject to be screened into the explanatory variables of the logistic regression analysis model in which the calculated partial regression coefficients have been set, and predicting the presence or absence of borderline diabetes in the subject based on the output data of the dependent variable. The blood glucose level-related parameters consist of a simple fasting blood glucose value and age and / or gender, although BMI may also be included if age and gender are included. The screening program for borderline diabetes may be recorded on a computer-readable information recording medium.

[0038] The database stores data that correlates relevant information (identification number, name, height, weight, chronic illnesses, medical history, medication history, other health checkup results, recording date and time, etc.) of people who have health checkup results related to diabetes, the presence or absence of borderline diabetes as the objective variable of the logistic regression analysis model, and blood glucose level-related parameters (fasting simple blood glucose level, sex, age, BMI) as explanatory variables. In other words, data on people who are in the borderline and non-borderline diabetes ranges required for logistic regression analysis is prepared using the actual values ​​of the explanatory variables and objective variables of people who have the health checkup results.

[0039] If the screening subject receives a definitive diagnosis of diabetes, the database is structured so that data can be sequentially added and stored that corresponds to the subject's relevant information (identification number, name, height, weight, chronic illnesses, medical history, medication history, other health check results, recording date and time, etc.), the presence or absence of borderline diabetes as the objective variable of the logistic regression analysis model, and blood glucose level-related parameters (fasting simple blood glucose value, gender, age, BMI) as the explanatory variables.

[0040] The input unit 30 comprises a keyboard, mouse, touchpad, etc., which can input relevant information, diagnosis results, and specific parameters of individuals who have been definitively diagnosed with borderline diabetes, as well as relevant information and specific parameters of individuals who will be screened (screening subjects). Simplified blood glucose level information may be directly input via the interface 13 wirelessly or wired from a simplified blood glucose meter (not shown).

[0041] The output unit 40 comprises a display and a printer that displays the contents of the database and the screening results of the person who has undergone screening.

[0042] The method of the present invention for easily screening borderline diabetic subjects from among screening subjects suspected of having borderline diabetes will be described with reference to the flowchart shown in FIG.

[0043] First, a database is prepared for use in a logistic regression analysis model in which blood glucose level-related parameters are used as explanatory variables and the presence or absence of borderline diabetes is used as the objective variable (S0). Specifically, a statistically significant number of individuals diagnosed with borderline diabetes (borderline diabetes individuals) and individuals not diagnosed with borderline diabetes (non-borderline diabetes individuals) are collected. In order to align these background factors, it is preferable to match individuals by gender, age, and BMI.

[0044] A database such as the one shown in Figure 4 will be prepared based on relevant information (identification number, name, height, weight, chronic illnesses, medical history, medication history, other health checkup results, recording date and time, etc.) of borderline and non-borderline diabetic individuals narrowed down through appropriate matching, the presence or absence of borderline diabetic status as the actual value of the objective variable of the logistic regression analysis model, and blood glucose level-related parameters (fasting simple blood glucose value, gender, age, BMI) as the actual value of the explanatory variables.

[0045] Next, in the analysis step, a logistic regression analysis is performed using the blood glucose level-related parameters of the screening subject as explanatory variables and the presence or absence of borderline diabetes as the objective variable (S1). In the logistic regression analysis model, the probability P of the appearance of "1: yes" and "0: no" for those in the borderline diabetes range is expressed by the following formula (1).

number

[0046] The partial regression coefficients of equation (1) are determined using general-purpose statistical analysis software (for example, product name SPSS (registered trademark), manufactured by IBM) equipped with a logistic regression analysis function (S2).

[0047] The blood glucose level-related parameters consist of the fasting simple blood glucose level and age and / or sex, but if age and sex are included, BMI may also be included. The efficiency of screening for borderline diabetes using these two parameters is significantly improved compared to the simple blood glucose level alone.

[0048] Finally, in the screening step, data on the blood glucose level-related parameters of the screening subject are input into the explanatory variables of the logistic regression analysis model in which the determined partial regression coefficients are set (S3).

[0049] Examples of the screening subjects include those who meet the following criteria through a health check or self-examination: (1) Fasting blood glucose level is 100-125 mg / dL, (2) Random blood glucose level is 140-199 mg / dL, and (3) Urine sugar is positive A person who satisfies at least one of the above.

[0050] A positive urine glucose result in a postprandial urine glucose test correlates with a random blood glucose level of 170 to 180 mg / dL or higher. Therefore, individuals in categories (1) to (3) are highly suspected of being borderline diabetic. Blood glucose levels in categories (1) and (2) can be measured using an SMPG, and the urine glucose test in category (3) can be performed using commercially available test strips or a urine glucose meter. Therefore, the screening method of the present invention can be said to be an extremely simple method that can be performed consistently, from suspicion of borderline diabetes to its diagnosis, by the individual without the intervention of a medical institution or medical analysis facility.

[0051] After inputting the explanatory variables of the screening subject, the P value of formula (1) is calculated as the output data of the objective variable of the logistic regression model (S4). From the calculation result of formula (1), the presence or absence of borderline diabetes of the screening subject is determined as follows: If P>0.5, the patient is at the borderline diabetic stage. If P<0.5, the patient is not at the borderline diabetic stage. (S5) An example of the estimated results for a screening subject is shown in Figure 4.

[0052] According to the screening method of the present invention, as shown in Examples 1 to 4 below, subjects can be screened for the presence or absence of borderline diabetes with an efficiency of 80% or more, preferably 85% or more.

[0053] The results of the screening in the above screening step can be used by medical institutions and clinical testing institutions for early detection and early treatment of borderline diabetes through definitive diagnosis using plasma glucose levels, OGTT, and HbA1c, or for monitoring functional foods targeted at people at risk of developing diabetes, or for narrowing down participants in clinical trials of pharmaceuticals. [Example]

[0054] The following examples and comparative examples will explain the embodiments of the present invention in more detail, but the present invention is not limited to the following examples.

[0055] [Examples 1 to 4 and Comparative Examples 1 to 4] (1) Preparing the database To establish a method for easily screening for borderline diabetes from a screening target population consisting of individuals who were suspected of having borderline diabetes, we first recruited a total of 1,043 individuals, including those diagnosed with borderline diabetes (borderline diabetes individuals) and those not diagnosed with borderline diabetes (non-borderline diabetes individuals). To ensure consistency in background factors, we matched these individuals by gender, age, and BMI, narrowing the sample to 98 individuals (with roughly equal proportions of borderline and non-borderline diabetes individuals). A database, as shown in Figure 4, was created based on the relevant information of the 98 individuals (identification number, name, address, medical history, record date and time, etc.), the presence or absence of borderline diabetes as the objective variable in the logistic regression analysis model, and blood glucose-related parameters (fasting simple blood glucose value, gender, age, BMI) as explanatory variables.

[0056] (2) Analysis steps Next, using the SPSS (registered trademark) analysis tool, a logistic regression analysis model was performed using simple blood glucose values, age, sex, and BMI as explanatory variables as shown in Table 2, and the dependent variable as a binary value of "yes" or "no" for those with borderline fasting venous blood glucose values.

[0057] In the logistic regression analysis model, the probability P of the appearance of "yes" or "no" for borderline diabetic patients is:

number

[0058] [Table 2]

[0059] (3) Estimation step When the information of each screening subject was substituted into equation (1), if P>0.5, the subject was presumed to be in the borderline diabetic range, and if P<0.5, the subject was presumed to be in the non-borderline diabetic range. The screening efficiency of the logistic regression analysis model was calculated by comparing the estimated results of the 98 screening subjects with the confirmed diagnosis results. The results are shown in Table 3.

[0060] [Table 3]

[0061] As shown in Table 3, the screening efficiency for borderline diabetics using only the simple blood glucose value in Comparative Example 1 was low at 77.6%. This suggests that the screening method in Comparative Example 1 overlooks people who were diagnosed as borderline diabetics because their simple blood glucose values ​​were normal but their OGTT results were high.

[0062] In Comparative Examples 2 to 4, even when BMI and age or sex were added to the parameters of Comparative Example 1, the screening efficiency for the P>0.5 group and P<0.5 group was comparable to that of Comparative Example 1. In particular, it was found that the addition of BMI in Comparative Examples 2 to 4 acted to slightly reduce the screening efficiency for non-borderline individuals in the P<0.5 group compared to Comparative Example 1, where BMI was not added.

[0063] On the other hand, in Examples 1 to 3, by adding only age and / or gender to the fasting simple blood glucose value, the screening efficiency for borderline individuals in the P>0.5 group improved to 83.7 to 85.7%, and the screening efficiency for non-borderline individuals in the P<0.5 group also improved to 81.6 to 83.7%.

[0064] In Example 4, by adding BMI to age and gender, the screening efficiency for borderline individuals in the P>0.5 group was increased to 85.7%, while the screening efficiency for non-borderline individuals in the P<0.5 group was maintained at 81.6%, above 80%.

[0065] From the above results, when the blood glucose level-related parameters consist of the fasting simple blood glucose value and age and / or sex, or consist of age, sex and BMI according to the present invention, it is possible to simultaneously achieve high screening efficiency for borderline subjects in the P>0.5 group and high screening efficiency for non-borderline subjects in the P<0.5 group.

[0066] When emphasis is placed on high screening efficiency for borderline individuals in the P>0.5 group, it is preferable to use simple blood glucose level and age, or blood glucose level, age, sex and BMI, as in Examples 2 and 4. [Explanation of symbols]

[0067] 1: Screening device for people on the borderline of diabetes 10: Computer main body 11: Central processing unit 12: Main memory 13: Interface 20: External storage device 21: Database 30: Input unit 40: Output unit

Claims

1. A method for screening for borderline diabetes using a logistic regression analysis model with blood glucose level-related parameters as explanatory variables and the presence or absence of borderline diabetes as a response variable, comprising: an analysis step of calculating partial regression coefficients of a logistic regression analysis model using actual values ​​of the explanatory variables and the response variable; and an estimation step of inputting data on blood glucose level-related parameters of the screening subject into explanatory variables of a logistic regression analysis model in which the determined partial regression coefficients are set, and estimating the presence or absence of borderline diabetes in the screening subject based on output data of the objective variable, The method for screening for people on the borderline of diabetes, characterized in that the blood glucose level-related parameters consist of a fasting simple blood glucose value and age and / or gender, but if age and gender are included, BMI may also be included.

2. 2. The method for screening for borderline diabetes patients according to claim 1, wherein the blood glucose level-related parameters consist of a fasting simple blood glucose level and age, or a fasting simple blood glucose level, age, sex, and BMI.

3. The screening subject meets the following conditions: (1) Fasting blood glucose level is 100-125 mg / dL, (2) Random blood glucose level is 140-199 mg / dL, and (3) Urine sugar is positive 2. The method for screening for borderline diabetic subjects according to claim 1, wherein the subject is a subject exhibiting at least one of the following conditions:

4. A screening device for people in the borderline diabetic range using a logistic regression analysis model with blood glucose level-related parameters as explanatory variables and the presence or absence of borderline diabetes as a response variable, an analysis means for calculating partial regression coefficients of a logistic regression analysis model using actual values ​​of the explanatory variables and the response variable; and an estimation means for inputting data on blood glucose level-related parameters of the screening subject into explanatory variables of a logistic regression analysis model in which the determined partial regression coefficients are set, and estimating the presence or absence of borderline diabetes in the screening subject based on output data of the objective variable, The screening device for those on the borderline of diabetes, characterized in that the blood glucose level-related parameters consist of a fasting simple blood glucose value and age and / or gender, but if age and gender are included, BMI may also be included.

5. A screening program for borderline diabetes that uses a logistic regression analysis model with blood glucose level-related parameters as explanatory variables and the presence or absence of borderline diabetes as a response variable to implement in a computer a function for estimating whether a person is borderline diabetes, an analysis function for calculating partial regression coefficients of a logistic regression analysis model using actual values ​​of the explanatory variables and the response variable; and an estimation function for inputting data on blood glucose level-related parameters of the screening subject into explanatory variables of a logistic regression analysis model in which the calculated partial regression coefficients are set, and estimating whether or not the screening subject has borderline diabetes based on output data of the objective variable, The screening program for people at the borderline of diabetes, characterized in that the blood glucose level-related parameters consist of a fasting simple blood glucose value and age and / or gender, but if age and gender are included, BMI may also be included.

6. 6. A computer-readable information recording medium on which the screening program for borderline diabetes patients according to claim 5 is recorded.

Images