Estimation device, estimation method, and estimation marker

Abstract

An estimation device (1) is provided with a storage device (12) that stores an estimation program (121) for estimating a risk of developing a disease, and a computing device (11) that estimates the risk of developing the disease on the basis of the estimation program (121). The computing device (11) acquires first data including a measured value of advanced glycation end products of a subject, and acquires second data including at least one of the subject's age, duration of diabetes, and body mass index. The computing device (11), on the basis of the first and second data, estimates the risk of developing the disease and outputs output data indicating the risk of developing the disease.

Description

Estimation apparatus, estimation method, and marker for estimation

[0001] The present disclosure relates to an estimation apparatus, an estimation method, and a marker for estimation for estimating the onset risk of vascular complications caused by diabetes in a subject.

[0002] Advanced Glycation End products (AGEs, hereinafter also referred to as "AGEs") are mentioned as one of substances that cause aging. AGEs is a general term for a plurality of compounds, which are produced when sugars bind to proteins and react through oxidation, condensation, and dehydration. AGEs accumulate in the body mainly due to various factors such as dietary habits, exercise habits, sleep habits, inflammation in response to fever or injury, and stress. Accumulation of AGEs is considered to cause lifestyle-related diseases (e.g., diabetes, dementia, etc.) or age-related diseases. Particularly, in the case of a patient suffering from diabetes, an increase in the accumulated amount of AGEs increases the probability of developing vascular complications caused by diabetes.

[0003] In this regard, Patent Document 1 (International Publication No. WO 2022 / 185554) discloses a health management system that provides advice from a registered dietitian or a pharmacist to a subject based on the measurement value of AGEs of the subject.

[0004] International Publication No. WO 2022 / 185554

[0005] According to the health management system described in Patent Document 1, advice based on the measurement value of AGEs is provided to the subject, so the subject can review their lifestyle habits to suppress an increase in the accumulated amount of AGEs. As described above, since the accumulated amount of AGEs affects the onset of vascular complications caused by diabetes, it is necessary to review lifestyle habits according to the onset risk of vascular complications. However, a technology that can easily and highly accurately estimate the onset risk of vascular complications has not been provided yet.

[0006] The present disclosure has been made to solve such a problem, and an object of the present disclosure is to provide a technology for easily and highly accurately estimating the onset risk of vascular complications.

[0007] An estimation device according to certain aspects of this disclosure estimates the risk of developing vascular complications due to diabetes in a subject. The estimation device comprises a storage device that stores an estimation program for estimating the risk of developing complications, and a computing device that estimates the risk of developing complications based on the estimation program. The computing device acquires first data including a measurement of the subject's advanced glycation end products, and second data including at least one of the subject's age, duration of diabetes, and body mass index, and estimates the risk of developing complications based on the first and second data, and outputs output data indicating the risk of developing complications.

[0008] An estimation device, according to another aspect of this disclosure, estimates the risk of developing vascular complications due to diabetes in a subject. The estimation device comprises a memory device that stores an estimation program for estimating the risk of developing complications, and a computing device that estimates the risk of developing complications based on the estimation program. The computing device obtains the subject's age, duration of diabetes, and body mass index, estimates the risk of developing complications based on the age, duration of diabetes, and body mass index, and outputs output data indicating the risk of developing complications.

[0009] An estimation method according to another aspect of this disclosure is a method by which a computing device estimates the risk of a subject developing vascular complications due to diabetes. The estimation method includes, as a process performed by the computing device, the steps of: obtaining first data including a measurement of the subject's advanced glycation end products; obtaining second data including at least one of the subject's age, duration of diabetes, and body mass index; estimating the risk of development based on the first and second data; and outputting output data indicating the risk of development.

[0010] Estimation markers, as otherwise provided in this disclosure, are markers for estimating the risk of developing vascular complications due to diabetes in a subject, comprising a measure of the subject's advanced glycation end products and at least one of the subject's age, duration of diabetes, and body mass index.

[0011] According to this disclosure, the risk of developing vascular complications due to diabetes in the subject can be easily and accurately estimated.

[0012] This is a diagram showing the estimation system according to the embodiment. This is a diagram showing the configuration of the estimation device according to the embodiment. This is a diagram showing the evaluation results of the estimation accuracy of the risk of developing vascular complications. This is a diagram showing the evaluation results of the estimation accuracy of the risk of developing vascular complications. This is a diagram showing the evaluation results of the estimation accuracy of the risk of developing vascular complications. This is a graph showing the ROC curve for the estimation results of the risk of developing vascular complications. This is a diagram showing the evaluation results of the estimation accuracy of the risk of developing large vessel complications. This is a graph showing the ROC curve for the estimation results of the risk of developing large vessel complications. This is a diagram showing the evaluation results of the estimation accuracy of the risk of developing microvascular complications. This is a graph showing the ROC curve for the estimation results of the risk of developing microvascular complications. This is a flowchart of the estimation process performed by the estimation device according to the embodiment. This is a diagram showing an example of the display of the estimation results of the risk of developing vascular complications.

[0013] This embodiment will be described in detail with reference to the drawings. Note that identical or corresponding parts in the drawings are denoted by the same reference numerals, and their descriptions will not be repeated in principle.

[0014] [Configuration of the Estimation System] Figure 1 is a diagram showing an estimation system 100 according to an embodiment. As shown in Figure 1, the estimation system 100 comprises an estimation device 1, an AGEs measuring device 2, a BMI (Body Mass Index) measuring device 3, and a user device 4.

[0015] Estimation device 1 is an information terminal that can communicate with the AGEs measurement device 2, the BMI measurement device 3, and the user device 4 via a network, such as a desktop PC (personal computer), a laptop PC, a smartphone, a smartwatch, a wearable device, or a tablet PC. Estimation device 1 may be a server device installed in a medical institution such as a hospital or clinic, or it may be a cloud computer.

[0016] Furthermore, the estimation device 1 may be managed by a service provider that provides the services offered by the estimation system 100 (hereinafter also referred to as "information provision services"). The service provider may also be the manufacturer of the AGEs measuring device 2 that lends the AGEs measuring device 2 to the subjects who are being measured for AGEs.

[0017] The subjects include individuals suffering from diabetes. The estimation device 1 uses various data, including AGEs measured by the AGEs measuring device 2, to estimate the risk of developing vascular complications due to diabetes, and outputs output data showing the estimated risk to the user device 4.

[0018] The AGEs measuring device 2 measures the AGEs of a subject. The AGEs measuring device 2 comprises a measuring unit 21, a display 22, and a communication unit 23. The AGEs measuring device 2 may be configured integrally with the display 22, or it may be configured separately from the display 22.

[0019] The measurement unit 21 measures the subject's AGEs. Among the various compounds contained in AGEs, some have the property of emitting fluorescence when irradiated with specific light. The measurement unit 21 measures the subject's AGEs by utilizing the properties of such compounds.

[0020] When a subject places their fingertip in contact with the measurement unit 21, the measurement unit 21 irradiates light onto the skin from a light source (not shown). The measurement unit 21 may also be configured to irradiate light onto skin other than the subject's fingertip (for example, the arm). The light irradiated by the measurement unit 21 is, for example, excitation light having a peak in the wavelength range of 410 nm or less. The measurement unit 21 receives the fluorescence excited by the light irradiated onto the skin using a photodetector (not shown), and measures the amount of AGEs accumulated based on the intensity of the received fluorescence. The display 22 displays the AGEs measurement result obtained by the measurement unit 21. The measurement result includes, for example, the intensity of the fluorescence received by the measurement unit 21 and a value converted from the amount of AGEs accumulated to a score. The measurement result may also include a corrected value obtained by correcting the value converted from the intensity of the fluorescence received by the measurement unit 21 and the amount of AGEs accumulated to a score.

[0021] In this way, subjects can measure AGEs non-invasively simply by inserting their finger into the measurement unit 21.

[0022] The communication unit 23 transmits and receives data (information) to and from the estimation device 1 via wired or wireless communication. The communication unit 23 may be a component capable of communicating with the estimation device 1, such as a network adapter, or it may be built into the AGEs measurement device 2. Alternatively, the communication unit 23 may be an information terminal capable of communicating with the estimation device 1 via a network, such as a desktop PC, laptop PC, smartphone, smartwatch, wearable device, or tablet PC, and it may be separate from the AGEs measurement device 2.

[0023] The AGEs measuring device 2 is installed in various facilities such as pharmacies, medical institutions, nursing homes, and gyms. The AGEs measuring device 2 may be managed by a supporter who assists the subject. When a subject measures AGEs using the AGEs measuring device 2, the measured value indicating the amount of accumulated AGEs (hereinafter also referred to as the "AGEs measured value") is transmitted from the AGEs measuring device 2 to the estimation device 1. The AGEs measured value is also referred to as the AGEs score.

[0024] Although there are individual differences, AGEs measurements generally change over several weeks, so subjects should measure their AGEs at a frequency of, for example, once every two weeks.

[0025] The BMI measuring device 3 is, for example, a body composition analyzer that measures the subject's BMI. The subject can have their BMI measured non-invasively simply by stepping onto the BMI measuring device 3. When the subject measures their BMI using the BMI measuring device 3, the measured value indicating the BMI (hereinafter also referred to as the "BMI measured value") is transmitted from the BMI measuring device 3 to the estimation device 1.

[0026] User device 4 is owned or used by the user. User device 4 is an information terminal that can communicate with estimation device 1 via a network, such as a desktop PC, laptop PC, smartphone, smartwatch, wearable device, and tablet PC. The user can obtain the estimated disease risk results stored in estimation device 1 by directly or indirectly accessing estimation device 1 using user device 4.

[0027] A user is a user of the information provision service. Specifically, a user may be the subject or a supporter of the subject. A supporter is someone who supports the subject and includes staff at a care facility, a social worker at a care facility, a doctor at a hospital, clinic, or corporate clinic, a nurse at a hospital, clinic, or corporate clinic, an instructor or nutrition advisor at a fitness gym, and a pharmacist at a pharmacy. A user may also be a family member, relative, or other person related to the subject (e.g., an acquaintance) who has been granted permission by the subject or their supporter to view the subject's measurement results.

[0028] In the estimation system 100 having the configuration described above, when a subject measures AGEs using the AGEs measuring device 2, the AGEs measuring device 2 outputs the AGEs measurement value to the estimation device 1. When the estimation device 1 obtains the AGEs measurement value from the AGEs measuring device 2, it stores the obtained AGEs measurement value.

[0029] Furthermore, when a subject measures their BMI using the BMI measuring device 3, the BMI measuring device 3 outputs the BMI measurement value to the estimation device 1. When the estimation device 1 obtains the BMI measurement value from the BMI measuring device 3, it stores the obtained BMI measurement value.

[0030] Furthermore, if the subject is interviewed at a medical institution such as a hospital, the interview information showing the results of the interview is stored as an electronic medical record on a server device (not shown) installed at the medical institution. Estimation device 1 obtains the subject's interview information from the medical record information stored on the server device. If estimation device 1 is installed at a medical institution, the doctor may directly input the interview information into estimation device 1.

[0031] The medical history information includes at least one of the following: the subject's age, the subject's duration of diabetes, and the subject's BMI measurement value. As mentioned above, the estimation device 1 may obtain the BMI measurement value directly from the BMI measuring device 3, or it may obtain the BMI measurement value from the medical history information included in the medical record. Furthermore, the medical history information may also include the subject's AGEs measurement value. As mentioned above, the estimation device 1 may obtain the AGEs measurement value directly from the AGEs measuring device 2, or it may obtain the AGEs measurement value from the medical history information included in the medical record. The duration of diabetes is the period from when the subject was diagnosed with diabetes to the present.

[0032] Estimation device 1 estimates the risk of developing vascular complications due to diabetes based on data including AGEs measurements (an example of "first data") and data including at least one of the subject's age, duration of diabetes, and BMI measurements (an example of "second data"). Estimation device 1 outputs output data including image data showing the risk of developing vascular complications to user device 4. Alternatively, estimation device 1 displays an image showing the risk of developing vascular complications on display 14.

[0033] The user device 4 displays an image on the display 40 that shows the risk of developing vascular complications, according to the output data acquired from the estimation device 1. This allows the subject or a user such as a supporter to obtain the subject's risk of developing vascular complications due to diabetes, as estimated by the estimation device 1.

[0034] In this way, by using the information service provided by the estimation system 100, participants can easily and accurately estimate their own risk of developing vascular complications due to diabetes. By obtaining the estimated risk of developing vascular complications, participants can review their lifestyle habits.

[0035] [Configuration of the Estimation Device] Figure 2 is a diagram showing the configuration of the estimation device 1 according to an embodiment. As shown in Figure 2, the estimation device 1 comprises a computing device 11, a storage device 12, a communication device 13, and a display 14.

[0036] The arithmetic unit 11 is a computing entity (computer) that performs various processes by executing various programs. The arithmetic unit 11 is composed of processors such as a CPU (Central Processing Unit), MPU (Micro Processing Unit), TPU (Tensor Processing Unit), or GPU (Graphics Processing Unit). A processor, which is an example of the arithmetic unit 11, has the function of performing various processes by executing programs, but some or all of these functions may be implemented using dedicated hardware circuits such as an ASIC (Application Specific Integrated Circuit) or FPGA (Field Programmable Gate Array). The term "processor" is not limited to processors in the narrow sense that execute processing in a stored-program manner, such as a CPU, MPU, TPU, or GPU, but may also include hardwired circuits such as ASICs or FPGAs. For this reason, a "processor," which is an example of the arithmetic unit 11, can also be read as a computing processing circuitry in which processing is predefined by computer-readable code and / or hardwired circuits. The arithmetic unit 11 may consist of one chip or multiple chips. Furthermore, the processor and associated processing circuits may consist of multiple computers interconnected by wired or wireless connections via a local area network or wireless network. The processor and associated processing circuits may also consist of a cloud computer that remotely performs calculations based on input data and outputs the calculation results to other devices located at a distance.

[0037] Furthermore, the arithmetic unit 11 may include a storage unit for storing program code or work memory when the processor executes various programs. The storage unit may be one or more non-transitory computer-readable media. The storage unit may include volatile memory such as DRAM (dynamic random access memory) and SRAM (static random access memory), or non-volatile memory such as ROM (read-only memory) and flash memory.

[0038] The storage device 12 is one or more computer-readable storage media, including HDDs (Hard Disk Drives) and SSDs (Solid State Drives). The storage device 12 stores various programs and data, such as the estimation program 121 executed by the arithmetic unit 11, and medical record information 122 including the subject's medical interview information.

[0039] The arithmetic unit 11 may also include a media reader (not shown). The arithmetic unit 11 may receive one or more computer-readable storage media, such as removable disks, via the media reader, and acquire various programs and data, such as the estimated program 121 and medical record information 122, from the removable disks.

[0040] The estimation program 121 specifies instructions for the computing device 11 to perform a process (hereinafter also referred to as the "estimation process") to estimate the risk of developing vascular complications due to diabetes in the subject.

[0041] Medical record information 122 includes the subject's age, duration of diabetes, and interview information such as BMI measurements.

[0042] The communication device 13 communicates with the AGEs measurement device 2 to receive AGEs measurement values from the AGEs measurement device 2. The communication device 13 communicates with the BMI measurement device 3 to receive BMI measurement values from the BMI measurement device 3. Furthermore, the communication device 13 communicates with the user device 4 to transmit output data indicating the onset risk of vascular complications to the user device 4.

[0043] The display 14 displays a predetermined image based on image data acquired from the arithmetic device 11.

[0044] [Estimation Accuracy of Onset Risk of Vascular Complications] Referring to FIGS. 3 to 7, the estimation accuracy of the onset risk of vascular complications of a subject by the estimation device 1 will be described. FIGS. 3 to 6 are diagrams showing evaluation results of estimation accuracy of the onset risk of vascular complications. FIG. 7 is a graph showing an ROC (Receiver operating characteristic) curve relating to the estimation result of the onset risk of vascular complications.

[0045] It should be noted that, in FIGS. 3 to 7, as the onset risk of vascular complications, the estimation accuracy of the onset risk of either macrovascular complications or microvascular complications is described. Macrovascular complications are disorders that occur in large blood vessels, and include cardiovascular diseases (for example, myocardial infarction or angina pectoris), and cerebrovascular diseases (for example, cerebral infarction, cerebral hemorrhage, or subarachnoid hemorrhage). Microvascular complications include nephropathy, retinopathy, and neuropathy.

[0046] FIGS. 3 to 6 show the evaluation results of the estimation accuracy of the onset risk of vascular complications for each type of input data used when estimating the onset risk of vascular complications (macrovascular complications, microvascular complications). When determining the input data to be input to the estimation device 1, a designer of the estimation device 1 selects, from various data of the subject, input data that can improve the estimation accuracy by the estimation device 1.

[0047] Evaluation items for estimation accuracy for each input data include AUC (Area Under the Curve), standard error, 95% confidence interval, and P value.

[0048] A subject to be estimated is a person suffering from diabetes. Among the subjects, 271 persons have developed vascular complications, and 181 persons have not developed vascular complications. That is, among the 452 subjects, 271 have developed vascular complications.

[0049] As shown in Case 1 of FIG. 3, estimation accuracy in the case of estimating the onset risk of vascular complications using four types of input data: the AGEs measurement value, age, disease duration, and BMI measurement value of a subject is described.

[0050] A designer converts the AGEs measurement value of each subject into a calculated value for estimating the onset risk. For example, when the AGEs measurement value is 0 or more and less than 0.2, the calculated value is converted to point A1. When the AGEs measurement value is 0.2 or more and less than 0.4, the calculated value is converted to point A2. When the AGEs measurement value is 0.4 or more and less than 0.6, the calculated value is converted to point A3. When the AGEs measurement value is 0.6 or more and 0.8 or less, the calculated value is converted to point A4. Note that A1 to A4 are any predetermined values.

[0051] A designer converts the age of each subject into a calculated value for estimating the onset risk. For example, when the age is less than 40 years old, the calculated value is converted to point B1. When the age is 40 years old or more and less than 50 years old, the calculated value is converted to point B2. When the age is 50 years old or more and less than 60 years old, the calculated value is converted to point B3. When the age is 60 years old or older, the calculated value is converted to point B4. Note that B1 to B4 are any predetermined values.

[0052] A designer converts the disease duration of each subject into a calculated value for estimating the onset risk. For example, when the disease duration is less than 1 year, the calculated value is converted to point C1. When the disease duration is 1 year or more and less than 2 years, the calculated value is converted to point C2. When the disease duration is 2 years or more and less than 3 years, the calculated value is converted to point C3. When the disease duration is 3 years or more, the calculated value is converted to point C4. Note that C1 to C4 are any predetermined values.

[0053] The designer converts each subject's BMI measurement into a calculated value for estimating the risk of developing the disease. For example, a BMI measurement of less than 25 is converted to a D1 score. A BMI measurement of 25 to less than 30 is converted to a D2 score. A BMI measurement of 30 to less than 40 is converted to a D3 score. A BMI measurement of 40 or more is converted to a D4 score. Note that D1 to D4 are arbitrary values ​​predetermined.

[0054] The designer calculates an estimated value by multiplying or adding the calculated value derived from the AGEs measurement, the calculated value derived from age, the calculated value derived from the duration of the disease, and the calculated value derived from the BMI measurement. The designer compares the estimated value with a predetermined threshold, and if the estimated value exceeds the threshold, it is estimated that the subject has developed vascular complications.

[0055] The designers performed the above-mentioned estimations for all 452 subjects and divided each subject into two groups: one that had developed vascular complications and one that had not.

[0056] The designer compares the estimated results for each subject, based on the four types of input data (AGEs measurement value, age, duration of illness, BMI measurement value) as described above, with the actual presence or absence of vascular complications, which is the ground truth data. Based on the comparison results, the designer can calculate the true positive rate and false positive rate of the estimated results, and based on these true positive and false positive rates, can plot an ROC curve as shown in Figure 7, which will be described later. Furthermore, the designer can calculate the AUC by calculating the area under the ROC curve. The higher the estimation accuracy of the risk of developing vascular complications, the larger the AUC. The designer can also calculate the standard deviation, 95% confidence interval, and p-value for the estimated calculation values ​​for each subject.

[0057] In this way, when the designer calculates the AUC for estimation accuracy using the input data of Case 1 (AGEs measurement value, age, duration of illness, BMI measurement value), the AUC is 0.711.

[0058] As shown in Case 2 of Figure 3, the estimation accuracy when estimating the risk of developing vascular complications using three types of input data—AGEs measurement value, age, and duration of illness—will be explained. Similar to Case 1, when the designer calculates the AUC for the estimation based on the input data of Case 2 (AGEs measurement value, age, duration of illness), the AUC is 0.689.

[0059] As shown in Case 3 of Figure 3, the estimation accuracy when estimating the risk of developing vascular complications using three types of input data—AGEs measurement value, age, and BMI measurement value—of the subject will be explained. Similar to Case 1, when the designer calculates the AUC for the estimation based on the input data of Case 3 (AGEs measurement value, age, BMI measurement value), the AUC is 0.681.

[0060] As shown in Case 4 of Figure 3, the estimation accuracy when estimating the risk of developing vascular complications using three types of input data—AGEs measurement value, duration of illness, and BMI measurement value—will be explained. Similar to Case 1, when the designer calculates the AUC for the estimation based on the input data of Case 4 (AGEs measurement value, duration of illness, BMI measurement value), the AUC is 0.695.

[0061] As shown in Case 5 of Figure 4, the estimation accuracy when estimating the risk of developing vascular complications using two types of input data—AGEs measurement values ​​and age—of the subject will be explained. Similar to Case 1, when the designer calculates the AUC for the estimation based on the input data (AGEs measurement values, age) in Case 5, the AUC is 0.662.

[0062] As shown in Case 6 of Figure 4, the estimation accuracy when estimating the risk of developing vascular complications using two types of input data—AGEs measurement values ​​and duration of illness—will be explained. Similar to Case 1, when the designer calculates the AUC for the estimation based on the input data (AGEs measurement values, duration of illness) in Case 6, the AUC is 0.686.

[0063] As shown in Case 7 of Figure 4, the estimation accuracy when estimating the risk of developing vascular complications using two types of input data, AGEs measurement values ​​and BMI measurement values ​​of the subject, will be explained. Similar to Case 1, when the designer calculates the AUC for the estimation based on the input data (AGEs measurement values, BMI measurement values) in Case 7, the AUC is 0.644.

[0064] As shown in Case 8 of Figure 4, the estimation accuracy when estimating the risk of developing vascular complications using three types of input data—age, duration of illness, and BMI measurement—will be explained. Similar to Case 1, when the designer calculates the AUC for the estimation based on the input data (age, duration of illness, BMI measurement) in Case 8, the AUC is 0.688.

[0065] As shown in Case 9 of Figure 5, the estimation accuracy when estimating the risk of developing vascular complications using one type of input data, the AGEs measurement value of the subject, will be explained. Similar to Case 1, when the designer calculates the AUC for the estimation based on the input data (AGEs measurement value) of Case 9, the AUC is 0.634.

[0066] As shown in Case 10 of Figure 5, the estimation accuracy when estimating the risk of developing vascular complications using only one type of input data, the subject's age, will be explained. Similar to Case 1, when the designer calculates the AUC for the estimation based on the input data (age) in Case 10, the AUC is 0.595.

[0067] As shown in Case 11 of Figure 5, the estimation accuracy when estimating the risk of developing vascular complications using only one type of input data, the duration of illness of the subject, will be explained. Similar to Case 1, when the designer calculates the AUC for the estimation based on the input data (duration of illness) in Case 11, the AUC is 0.642.

[0068] As shown in Case 12 of Figure 5, the estimation accuracy when estimating the risk of developing vascular complications using one type of input data, the subject's BMI measurement, will be explained. Similar to Case 1, when the designer calculates the AUC for the estimation based on the input data (BMI measurement) in Case 12, the AUC is 0.594.

[0069] As shown in Case 13 of Figure 6, we will explain the estimation accuracy when estimating the risk of developing vascular complications using only one type of input data, HbA1c (hemoglobin A1c), of the subject. The HbA1c value can be tested by taking a blood sample. In other words, the HbA1c test is an invasive test that places a burden on the subject's body.

[0070] Since a higher HbA1c value indicates a higher risk of developing vascular complications, it is possible to determine whether or not a vascular complication has developed based on the HbA1c measurement of the subject. Similar to Case 1, when the designer calculates the AUC based on the input data (HbA1c) for Case 13, the AUC becomes 0.566.

[0071] As described above, the AUC for estimations in Cases 1 to 7, where at least one of the following input data (AGEs measurement, age, duration of illness, and BMI measurement) was added, was found to be larger than the AUC for estimations in Cases 9 to 13, where only one of the following input data (AGEs measurement, age, duration of illness, BMI measurement, and HbA1c measurement) was used. Furthermore, it was found that the AUC for estimations in Case 1, which used four types of input data (AGEs measurement, age, duration of illness, and BMI measurement), was the largest, exceeding 0.7.

[0072] For example, Figure 7 shows the ROC curve for Case 1, which uses four types of input data: AGEs measurement value, age, duration of illness, and BMI measurement value, and the ROC curve for Case 9, which uses only AGEs measurement value as input data.

[0073] The horizontal axis of the graph in Figure 7 represents the false positive rate (1 - specificity). In this example, the false positive rate shows the proportion of subjects who were incorrectly diagnosed as having vascular complications out of 271 subjects who did not develop vascular complications. The vertical axis of the graph in Figure 7 represents the true positive rate (sensitivity). In this example, the true positive rate shows the proportion of subjects who were correctly diagnosed as having vascular complications out of 181 subjects who did develop vascular complications.

[0074] As shown in Figure 7, the AUC of the ROC curve in Case 1, which uses four types of input data—AGEs measurement value, age, duration of illness, and BMI measurement value—is larger than the AUC of the ROC curve in Case 9, which uses only AGEs measurement value as input data.

[0075] As a result, the estimation device 1 can estimate the risk of developing vascular complications with greater accuracy by estimating the risk of developing vascular complications based on AGEs measurements and at least one of age, duration of illness, and BMI measurements. More preferably, the estimation device 1 can estimate the risk of developing vascular complications with very high accuracy by estimating the risk of developing vascular complications based on AGEs measurements, age, duration of illness, and BMI measurements.

[0076] Furthermore, as shown in Figure 4, even when AGEs measurements were not used as input data, the AUC of the estimation in Case 8, which used three types of input data—age, duration of illness, and BMI measurement—was greater than the AUC of the estimations in Cases 9 to 13, which used only one type of input data from AGEs measurement, age, duration of illness, BMI measurement, and HbA1c measurement.

[0077] As a result, the estimation device 1 can estimate the risk of developing vascular complications with high accuracy, even when estimating the risk based on age, duration of illness, and BMI measurements.

[0078] [Estimation Accuracy of the Risk of Large Vessel Complications] The estimation accuracy of the risk of large vessel complications for subjects using estimation device 1 will be explained with reference to Figures 8 and 9. Figure 8 is a diagram showing the evaluation results of the estimation accuracy of the risk of large vessel complications. Figure 9 is a graph showing the ROC curve related to the estimation results of the risk of large vessel complications.

[0079] Figure 8 shows the evaluation results of the estimation accuracy of the risk of developing macrovascular complications for each type of input data used when estimating the risk of developing macrovascular complications. When determining the input data to be input into the estimation device 1, the designer of the estimation device 1 selects input data from various data of the subject that will improve the estimation accuracy of the estimation device 1.

[0080] As shown in Case 21 of Figure 8, the estimation accuracy when estimating the risk of developing macrovascular complications using four types of input data—AGEs measurement value, age, duration of illness, and BMI measurement value—will be explained. Similar to Case 1, when the designer calculates the AUC for the estimation based on the input data (AGEs measurement value, age, duration of illness) in Case 21, the AUC is 0.731.

[0081] As shown in Case 22 of Figure 8, the estimation accuracy when estimating the risk of developing macrovascular complications using one type of input data of AGEs measurements for the subject will be explained. Similar to Case 1, when the designer calculates the AUC for the estimation based on the input data (AGEs measurements) in Case 22, the AUC is 0.646.

[0082] As shown in Case 23 of Figure 8, we will explain the estimation accuracy when estimating the risk of developing major vascular complications using one type of input data from the results of a carotid artery ultrasound of a subject. Carotid artery ultrasound is an examination that visualizes the condition of the blood vessels (carotid arteries) in the subject's neck.

[0083] The worse the carotid artery ultrasound results, the higher the risk of developing major vascular complications. Therefore, it is possible to determine whether or not a major vascular complication has occurred based on the carotid artery ultrasound results of the subject. Similar to Case 1, when the designer calculates the AUC based on the input data (carotid artery ultrasound results) for Case 23, the AUC becomes 0.614.

[0084] As shown in Case 24 of Figure 8, the estimation accuracy when estimating the risk of developing macrovascular complications using only one type of input data, the ABI (Ankle Brachial Index), of a subject will be explained. The ABI test is a test that measures the blood pressure in the subject's ankle and upper arm to evaluate arterial stenosis or occlusion.

[0085] The worse the ABI test result, the higher the risk of developing macrovascular complications. Therefore, it is possible to determine whether or not a macrovascular complication has developed based on the subject's ABI test results. Similar to Case 1, when the designer calculates the AUC based on the input data (ABI test results) for Case 24, the AUC becomes 0.662.

[0086] As described above, the AUC for estimation in Case 21, which adds age, duration of illness, and BMI measurements to AGEs measurements, was found to be higher than the AUC for estimation in Cases 22 to 24, which used only one of the following input data: AGEs measurements, carotid artery ultrasound results, and ABI test results, with the AUC being 0.7 or higher.

[0087] For example, Figure 9 shows the ROC curve for Case 21, which uses four types of input data: AGEs measurement value, age, duration of illness, and BMI measurement value, and the ROC curve for Case 22, which uses only AGEs measurement value as input data.

[0088] The horizontal axis of the graph in Figure 9 represents the false positive rate (1 - specificity). In this example, the false positive rate represents the proportion of subjects who were incorrectly diagnosed as having major vascular complications out of 345 subjects who did not have major vascular complications. The vertical axis of the graph in Figure 9 represents the true positive rate (sensitivity). The true positive rate represents the proportion of subjects who were correctly diagnosed as having major vascular complications out of 107 subjects who did have major vascular complications.

[0089] As shown in Figure 9, the AUC of the ROC curve in Case 21, which uses four types of input data—AGEs measurement value, age, duration of illness, and BMI measurement value—is larger than the AUC of the ROC curve in Case 22, which uses only AGEs measurement value as input data.

[0090] Although not shown in the diagram, similar to cases 2 to 7, the AUC for estimation when at least one of age, duration of illness, and BMI measurement values ​​was added to the input data in addition to the AGEs measurement value was found to be larger than the AUC for estimation in cases 22 to 24, when only one of the input data (AGEs measurement value, carotid artery ultrasound results, and ABI test results) was used.

[0091] As a result, the estimation device 1 can estimate the risk of developing macrovascular complications with greater accuracy by estimating the risk based on AGEs measurements and at least one of age, duration of illness, and BMI measurements. More preferably, the estimation device 1 can estimate the risk of developing macrovascular complications with very high accuracy by estimating the risk based on AGEs measurements, age, duration of illness, and BMI measurements.

[0092] Furthermore, although not shown in the diagram, similar to Case 8, even when AGEs measurements were not used as input data, the AUC for estimation using three types of input data—age, duration of illness, and BMI measurements—was greater than the AUC for estimation when only one type of input data was used from among AGEs measurements, age, duration of illness, BMI measurements, carotid artery ultrasound results, and ABI test results.

[0093] As a result, the estimation device 1 can estimate the risk of developing macrovascular complications with high accuracy, even when estimating the risk based on age, duration of illness, and BMI measurements.

[0094] [Estimation Accuracy of Microvascular Complication Risk] The estimation accuracy of the microvascular complications risk of subjects by the estimation device 1 will be explained with reference to Figures 10 and 11. Figure 10 is a diagram showing the evaluation results of the estimation accuracy of the microvascular complications risk. Figure 11 is a graph showing the ROC curve related to the estimation results of the microvascular complications risk.

[0095] Figure 10 shows the evaluation results of the estimation accuracy of the risk of developing microvascular complications for each type of input data used when estimating the risk of developing microvascular complications. When determining the input data to be input into the estimation device 1, the designer selected input data from various data of the subjects that would improve the estimation accuracy of the estimation device 1.

[0096] As shown in Case 31 of Figure 10, the estimation accuracy when estimating the risk of developing microvascular complications using four types of input data—AGEs measurement value, age, duration of illness, and BMI measurement value—will be explained. Similar to Case 1, when the designer calculates the AUC for the estimation based on the input data (AGEs measurement value, age, duration of illness) in Case 31, the AUC is 0.731.

[0097] As shown in Case 32 of Figure 10, the estimation accuracy when estimating the risk of developing microvascular complications using one type of input data of AGEs measurements for the subject will be explained. Similar to Case 1, when the designer calculates the AUC for the estimation based on the input data (AGEs measurements) in Case 32, the AUC is 0.642.

[0098] As shown in Case 33 of Figure 10, the estimation accuracy when estimating the risk of developing microvascular complications using one type of input data, eGFR (Estimated Glomerular Filtration Rate), for a subject will be explained. The eGFR test is a test that measures kidney function and is performed by taking a blood sample. In other words, the eGFR test is an invasive test that places a burden on the subject's body.

[0099] The higher the eGFR, the higher the risk of developing microvascular complications. Therefore, it is possible to determine whether or not microvascular complications have developed based on the subject's eGFR. Similar to Case 1, when the designer calculates the AUC based on the input data (eGFR) for Case 33, the AUC becomes 0.582.

[0100] As shown in Case 34 of Figure 10, the estimation accuracy when estimating the risk of developing microvascular complications using only one type of input data, urinary albumin (Alb), from a subject will be explained. Urinary albumin can be tested by collecting a urine sample.

[0101] Since a higher urinary albumin level increases the risk of developing microvascular complications, it is possible to determine whether or not a subject has developed microvascular complications based on their urinary albumin level. Similar to Case 1, when the designer calculates the AUC for the estimation based on the input data (urinary albumin) in Case 33, the AUC becomes 0.796.

[0102] As described above, the AUC for estimation in Case 31, which adds age, duration of illness, and BMI measurements to AGEs measurements, was found to be larger than the AUC for estimation in Cases 32 and 33, which used only one of the input data (AGEs measurements or eGFR). Furthermore, the AUC for estimation in Case 31, which adds age, duration of illness, and BMI measurements to AGEs measurements, was 0.7 or higher, similar to the AUC for estimation in Case 34, which uses urinary albumin, which is said to have high estimation accuracy, as input data.

[0103] For example, Figure 11 shows the ROC curve for Case 31, which uses four types of input data: AGEs measurement value, age, duration of illness, and BMI measurement value, and the ROC curve for Case 32, which uses only AGEs measurement value as input data.

[0104] The horizontal axis of the graph in Figure 11 represents the false positive rate (1 - specificity). In this example, the false positive rate represents the proportion of subjects who were incorrectly diagnosed as having microvascular complications out of 175 subjects who did not have microvascular complications. The vertical axis of the graph in Figure 11 represents the true positive rate (sensitivity). The true positive rate represents the proportion of subjects who were correctly diagnosed as having microvascular complications out of 277 subjects who did have microvascular complications.

[0105] As shown in Figure 11, the AUC of the ROC curve in Case 31, which uses four types of input data—AGEs measurement value, age, duration of illness, and BMI measurement value—is larger than the AUC of the ROC curve in Case 32, which uses only AGEs measurement value as input data.

[0106] Although not shown in the diagram, similar to cases 2 to 7, the AUC for estimation when at least one of age, duration of illness, and BMI measurement values ​​was added to the input data in addition to the AGEs measurement value was found to be larger than the AUC for estimation in cases 32 and 33, when only one of AGEs measurement values ​​and eGFR was used as input data.

[0107] As a result, the estimation device 1 can estimate the risk of developing microvascular complications with greater accuracy by estimating the risk of developing microvascular complications based on AGEs measurements and at least one of age, duration of illness, and BMI measurements. More preferably, the estimation device 1 can estimate the risk of developing microvascular complications with very high accuracy by estimating the risk of developing microvascular complications based on AGEs measurements, age, duration of illness, and BMI measurements.

[0108] Furthermore, although not shown in the diagram, similar to Case 8, even when AGEs measurements were not used as input data, the AUC for estimation using three types of input data—age, duration of illness, and BMI measurements—was greater than the AUC for estimation when only one of the following types of input data (AGEs measurements, age, duration of illness, BMI measurements, and eGFR) was used.

[0109] As a result, the estimation device 1 can estimate the risk of developing microvascular complications with high accuracy, even when estimating the risk based on age, duration of illness, and BMI measurements.

[0110] [Processing of the Estimation Device] Figure 12 is a flowchart of the estimation process performed by the estimation device 1 according to the embodiment. The processing steps shown in Figure 12 (hereinafter abbreviated as "S") are realized by the arithmetic unit 11 executing the estimation program 121. In the estimation process described below, the estimation device 1 may estimate the risk of developing macrovascular complications, or the risk of developing microvascular complications, or the risk of developing either macrovascular complications or microvascular complications.

[0111] As shown in Figure 12, the estimation device 1 obtains the subject's AGEs measurement value from the AGEs measuring device 2 (S1). The estimation device 1 also obtains the subject's age, duration of diabetes, and BMI measurement value (S2). For example, the estimation device 1 obtains the subject's age, duration of diabetes, and BMI measurement value from the medical interview information contained in the medical record, obtained from a server device installed in a medical institution. Alternatively, the estimation device 1 may obtain the subject's BMI measurement value from the BMI measuring device 3.

[0112] Estimation device 1 calculates a calculated value for estimating the risk of developing vascular complications based on the input data obtained in S1 and S2, including the subject's AGEs measurement value, age, duration of illness, and BMI measurement value (S3). For example, estimation device 1 calculates multiple calculated values ​​corresponding to each of the AGEs measurement value, age, duration of illness, and BMI measurement value. The calculated value calculated by estimation device 1 based on the AGEs measurement value is an example of a "first calculated value." The calculated values ​​calculated by estimation device 1 based on age, duration of illness, and BMI measurement value are examples of "second calculated values."

[0113] Estimation device 1 estimates the risk of developing vascular complications based on the calculated values ​​(S4). For example, estimation device 1 calculates an estimated calculated value by multiplying or adding the multiple calculated values ​​calculated in S3. Estimation device 1 compares the estimated calculated value with a predetermined threshold, and estimates that the subject has developed vascular complications if the estimated calculated value exceeds the threshold. Alternatively, estimation device 1 may convert the estimated calculated value into a probability of developing vascular complications. That is, the risk of developing vascular complications estimated by estimation device 1 may include not only information indicating whether or not vascular complications will develop, but also information indicating the probability of developing vascular complications.

[0114] Estimation device 1 outputs output data (S5) showing the risk of developing vascular complications estimated in S4. For example, estimation device 1 outputs output data showing the estimated risk of developing complications to user device 4. Alternatively, estimation device 1 displays an image showing the estimated risk of developing complications on display 14. After that, estimation device 1 terminates the processing of this flowchart.

[0115] Figure 13 shows an example of how the estimated risk of developing vascular complications is displayed. In Figure 13, the risk of developing vascular complications is shown on the display 14 of the estimation device 1 or the display 40 of the user device 4.

[0116] As shown in Figure 13, displays 14 and 40 show images representing the probability of developing vascular complications estimated by the estimation device 1, as the risk of developing vascular complications. Note that the risk of developing vascular complications shown in Figure 13 is just one example, and the risk of developing vascular complications may be displayed on displays 14 and 40 in other ways.

[0117] As described above, the estimation device 1 according to the embodiment estimates the risk of developing vascular complications due to diabetes in a subject based on the subject's AGEs measurement value, age, duration of illness, and BMI measurement value. Therefore, it can estimate the risk of developing vascular complications with greater accuracy than, for example, estimating the risk of developing vascular complications based solely on AGEs measurement value.

[0118] Specifically, as shown in Case 1 of Figure 3, Case 21 of Figure 8, and Case 31 of Figure 10, when estimating the risk of developing vascular complications based on the subject's AGEs measurement value, age, duration of illness, and BMI measurement value, the estimation accuracy is such that the AUC of the ROC curve, based on the true positive rate and false positive rate of the estimated risk of developing complications, is 0.7 or higher.

[0119] Furthermore, by having AGEs measured non-invasively using the AGEs measuring device 2, BMI measured non-invasively using the BMI measuring device 3, and answering questions about age and duration of diabetes during a medical interview with a physician, the subject can obtain the estimated risk of developing vascular complications from the estimation device 1. In this way, the estimation device 1 can easily estimate the risk of developing vascular complications based on the AGEs measurement values ​​obtained from the AGEs measuring device 2, the BMI measurement values ​​obtained from the BMI measuring device 3, and the age and duration of diabetes obtained from the medical interview information contained in the subject's medical record.

[0120] In the flowchart of Figure 12 described above, the estimation device 1 estimated the risk of developing vascular complications using the subject's AGEs measurement value, age, duration of illness, and BMI measurement value. However, the types of input data are not limited to these. For example, as shown in Cases 2 to 7 of Figures 3 and 4, the estimation device 1 may estimate the risk of developing vascular complications based on the AGEs measurement value and at least one of the age, duration of illness, and BMI measurement value. Even with such estimation, the estimation device 1 can estimate the risk of developing vascular complications with greater accuracy than, for example, estimating the risk of developing vascular complications based solely on the AGEs measurement value.

[0121] As described above, the estimation device 1 uses the subject's AGEs measurement, the subject's age, duration of illness, and at least one of the BMI measurement to estimate the subject's risk of developing vascular complications. Therefore, the subject's AGEs measurement, the subject's age, duration of illness, and at least one of the BMI measurement function as markers for estimating the subject's risk of developing vascular complications.

[0122] Furthermore, as shown in Case 8 of Figure 4, the estimation device 1 may estimate the risk of developing vascular complications based on age, duration of illness, and BMI measurements. Even with such estimation, the estimation device 1 can estimate the risk of developing vascular complications with greater accuracy than, for example, estimating the risk of developing vascular complications based solely on AGEs measurements.

[0123] As described above, the estimation device 1 uses the subject's age, duration of illness, and BMI measurement to estimate the subject's risk of developing vascular complications. Therefore, the subject's age, duration of illness, and BMI measurement function as markers for estimating the subject's risk of developing vascular complications.

[0124] <Modes> The above-described exemplary embodiments will be understood by those skilled in the art to be specific examples of the following embodiments.

[0125] (Section 1) An estimation device according to one embodiment estimates the risk of developing vascular complications due to diabetes in a subject. The estimation device comprises a storage device that stores an estimation program for estimating the risk of developing complications, and a computing device that estimates the risk of developing complications based on the estimation program. The computing device acquires first data including the measured value of advanced glycation end products of the subject, acquires second data including at least one of the subject's age, duration of diabetes, and body mass index, estimates the risk of developing complications based on the first and second data, and outputs output data indicating the risk of developing complications.

[0126] According to the estimation device described in paragraph 1, the risk of developing vascular complications in a subject can be estimated based on the subject's AGEs measurement value, age, duration of illness, and at least one of the subject's BMI measurement value. Therefore, the risk of developing vascular complications due to diabetes in a subject can be easily and accurately estimated.

[0127] (Paragraph 2) In the estimation device described in Paragraph 1, the estimation accuracy of the risk of developing the disease based on the first data and the second data is higher than the estimation accuracy of the risk of developing the disease based on the first data alone.

[0128] According to the estimation device described in paragraph 2, the risk of developing vascular complications can be estimated with greater accuracy than estimating the risk of developing vascular complications based solely on AGEs measurements.

[0129] (Clause 3) In the estimation device described in paragraph 1 or 2, if the second data includes age, duration of illness, and body mass index, the estimation accuracy is such that the AUC of the ROC curve, based on the true positive rate and false positive rate of the estimated risk of developing the disease, is 0.7 or higher.

[0130] According to the estimation device described in paragraph 3, the risk of developing vascular complications can be estimated with an estimation accuracy of 0.7 or higher, where AUC is 0.7 or higher.

[0131] (Article 4) In the estimation device described in any one of paragraphs 1 to 3, the calculation device calculates a first calculated value for estimating the risk of developing the disease based on the first data, calculates a second calculated value for estimating the risk of developing the disease based on the second data, and estimates the risk of developing the disease based on the first calculated value and the second calculated value.

[0132] According to the estimation device described in paragraph 4, the risk of developing vascular complications can be estimated by calculating a calculated value for estimating the risk of developing the condition.

[0133] (Article 5) In the estimation device described in any one of paragraphs 1 to 4, the calculation device acquires second data from the subject's medical record.

[0134] According to the estimation device described in paragraph 5, at least one of the subject's age, duration of illness, and BMI measurement can be obtained from the subject's medical record, making it easy to estimate the risk of developing vascular complications.

[0135] (Paragraph 6) The estimation device described in any one of paragraphs 1 to 5 further comprises a display. The computing device outputs image data indicating the risk of developing the disease to the display as output data. The display shows an image indicating the risk of developing the disease.

[0136] According to the estimation device described in paragraph 6, the subject or a user such as a supporter can obtain information on the risk of developing vascular complications by viewing images displayed on the screen.

[0137] (Clause 7) An estimation device according to one embodiment estimates the risk of developing vascular complications due to diabetes in a subject. The estimation device comprises a storage device that stores an estimation program for estimating the risk of developing complications, and a computing device that estimates the risk of developing complications based on the estimation program. The computing device acquires the subject's age, duration of diabetes, and body mass index, estimates the risk of developing complications based on the age, duration of diabetes, and body mass index, and outputs output data indicating the risk of developing complications.

[0138] According to the estimation device described in paragraph 7, the risk of developing vascular complications in a subject can be estimated based on the subject's age, duration of illness, and BMI measurement, thus enabling a simple and highly accurate estimation of the subject's risk of developing vascular complications due to diabetes.

[0139] (Clause 8) An estimation method relating to one embodiment is a method by which a computing device estimates the risk of developing vascular complications due to diabetes in a subject. The estimation method includes, as a process performed by the computing device, the steps of: acquiring first data including a measurement of the subject's advanced glycation end products; acquiring second data including at least one of the subject's age, duration of diabetes, and body mass index; estimating the risk of developing complications based on the first and second data; and outputting output data indicating the risk of developing complications.

[0140] According to the estimation method described in paragraph 8, the computing device can estimate the risk of developing vascular complications in a subject based on the subject's AGEs measurement value and at least one of the subject's age, duration of illness, and BMI measurement value, thus enabling a simple and highly accurate estimation of the subject's risk of developing vascular complications due to diabetes.

[0141] (Clause 9) An estimation marker according to one embodiment is a marker for estimating the risk of developing vascular complications due to diabetes in a subject, comprising the measured value of advanced glycation end products in the subject and at least one of the subject's age, duration of diabetes, and body mass index.

[0142] According to the estimation markers described in paragraph 9, subjects can easily obtain an estimate of their risk of developing vascular complications due to diabetes by non-invasively measuring AGEs and BMI, and by answering questions from a physician about their age and the duration of their diabetes.

[0143] 1 Estimation device, 2 AGEs measurement device, 3 BMI measurement device, 4 User device, 11 Calculation unit, 12 Storage device, 13 Communication device, 14, 22, 40 Display, 21 Measurement unit, 23 Communication unit, 100 Estimation system, 121 Estimation program, 122 Medical record information.

Claims

1. An estimation device for estimating the risk of developing vascular complications due to diabetes in a subject, comprising: a storage device for storing an estimation program for estimating the risk of developing the complications; and a calculation device for estimating the risk of developing the complications based on the estimation program, wherein the calculation device acquires first data including a measurement value of advanced glycation end products in the subject; acquires second data including at least one of the subject's age, duration of diabetes, and body mass index; estimates the risk of developing the complications based on the first data and the second data; and outputs output data indicating the risk of developing the complications.

2. The estimation device according to claim 1, wherein the estimation accuracy of the risk of developing the disease based on the first data and the second data is higher than the estimation accuracy of the risk of developing the disease based on the first data alone.

3. The estimation device according to claim 2, wherein, if the second data includes the age, the duration of illness, and the body mass index, the estimation accuracy is such that the AUC of the ROC curve based on the true positive rate and false positive rate of the estimated risk of onset is 0.7 or higher.

4. The estimation device according to claim 1, wherein the calculation device calculates a first calculated value for estimating the risk of developing the disease based on the first data, calculates a second calculated value for estimating the risk of developing the disease based on the second data, and estimates the risk of developing the disease based on the first calculated value and the second calculated value.

5. The estimation device according to claim 1, wherein the calculation device acquires the second data from the subject's medical record.

6. The estimation apparatus according to claim 1, further comprising a display, wherein the computing device outputs image data indicating the risk of developing the disease to the display as output data, and the display displays the image indicating the risk of developing the disease.

7. An estimation device for estimating the risk of developing vascular complications due to diabetes in a subject, comprising: a storage device for storing an estimation program for estimating the risk of developing the complications; and a computing device for estimating the risk of developing the complications based on the estimation program, wherein the computing device obtains the subject's age, duration of diabetes, and body mass index; estimates the risk of developing the complications based on the age, duration of diabetes, and body mass index; and outputs output data indicating the risk of developing the complications.

8. An estimation method for a calculation device to estimate the risk of developing vascular complications due to diabetes in a subject, the calculation device comprising the steps of: acquiring first data including a measurement of the subject's advanced glycation end products; acquiring second data including at least one of the subject's age, duration of diabetes, and body mass index; estimating the risk of developing the complications based on the first data and the second data; and outputting output data indicating the risk of developing the complications.

9. Estimation markers for estimating the risk of developing vascular complications due to diabetes in a subject, comprising the subject's measured value of advanced glycation end products and at least one of the subject's age, duration of diabetes, and body mass index.

Images