Method for evaluating immune state in obese people and kit thereof

Abstract

The invention relates to a method for evaluating the immune state of obese people and a kit thereof, relates to the technical field of medical data mining, and comprises an index combination for evaluating the immune state of the obese people. The index combination comprises the following indexes: hematocrit (HCT), hemoglobin content (HGB), low membrane potential percentage (Ts.MMPlow%) of killer T cells, percentage (T4Tcm%) of central memory assistant T cells, low membrane potential percentage (T8Tn.MMPlow%) of pure killer T cell mitochondria, percentage (NKT.NKG2D +%) of NKT cell expression NKG2D and mitochondrial mass (T8Tcm. MM) of central memory killer T cell mitochondria. The index combination obtained through screening has high biological interpretation and clinical correlation, and the immune state of obese people can be effectively evaluated.

Description

Technical Field

[0001] This application relates to the technical field of medical data mining, and in particular to a method and kit for assessing the immune status of obese individuals. Background Technology

[0002] Obesity and its associated immune abnormalities have become a major global public health problem. Their pathogenesis is complex, involving the interaction between metabolic disorders and immune system abnormalities. Innate immune cells, particularly natural killer T (NKT) cells and natural killer (NK) cells, play a crucial role in maintaining metabolic homeostasis and regulating inflammatory responses.

[0003] Studies have shown that the functional status of NK cells is significantly altered in obese individuals, such as CD56. bright Increased NKG2D expression frequency in NK subsets suggests a close correlation between NK cell activation and obesity severity. The role of NKT cells in the obesity process is more complex: some studies have found impaired cytotoxicity, while others suggest they can participate in inflammation amplification or tissue homeostasis maintenance through cytokine secretion. NKG2D expression on the surface of NKT cells is also considered associated with obesity development, but its specific mechanisms and functional significance remain unclear.

[0004] Recent advances in immunometabolism research have highlighted the importance of mitochondrial function as a key indicator of immune cell status. With age, the proportions of individual immune cell subsets (e.g., NK% and naïve T cell percentage) and mitochondrial functional parameters (e.g., mitochondrial mass (MM) and membrane potential levels) undergo significant changes. In obese individuals, immune cells often exhibit a state of metabolic hyperactivity characterized by increased mitochondrial mass and decreased membrane potential, suggesting an "immune excitation" that may further exacerbate metabolic inflammation and organ damage.

[0005] Currently, conventional body mass index (BMI) cannot reflect the degree of immune metabolic disorder underlying obesity, and existing immune assessment methods mostly rely on cell counting or single receptor expression, lacking a comprehensive analytical system that integrates functional metabolic indicators. Therefore, there is an urgent need in this field to develop tools for accurately assessing the immune status of obese individuals. Summary of the Invention

[0006] This application combines flow cytometry with multidimensional data analysis to overcome the limitations of existing research. It not only focuses on the expression of cell surface receptors (such as NKG2D and NKG2A) but also delves deeper into the internal metabolic state of their subpopulations (mitochondrial membrane potential and mass). It integrates routine clinical indicators and uses advanced machine learning algorithms for high-throughput screening. Ultimately, it aims to establish a novel multi-indicator assessment method and matching kit that combines NKG2D expression on the surface of NKT cells with mitochondrial functional parameters. This aims to achieve early identification and risk stratification of obesity-related immunometabolic abnormalities, providing more effective early warning and intervention evidence for clinical practice.

[0007] Firstly, this application provides a combination of indicators for detecting the immune status in obese individuals, the combination of indicators including the following: hematocrit (HCT), hemoglobin content (HGB), and percentage of cytotoxic T cells with low membrane potential (Ts.MMP). low %), percentage of central memory helper T cells (T4Tcm%), percentage of pure killer T cells with mitochondrial low membrane potential (T8Tn.MMP) low %), the percentage of NKT cells expressing NKG2D (NKT.NKG2D) + %) and mitochondrial mass of central memory killer T cells (T8Tcm.MM).

[0008] Secondly, this application provides a method for screening the combination of indicators described in the first aspect, the method comprising the following steps: (1) collecting peripheral blood samples; (2) detecting blood routine indicators, cell percentage indicators, mitochondrial quality indicators and mitochondrial low membrane potential percentage indicators of the samples in step (1); (3) performing correlation analysis and difference analysis on the indicators in step (2) to screen and obtain BMI-related indicators; (4) combining machine learning technology to analyze and screen the indicators in step (3) to obtain the combination of indicators.

[0009] Optionally, the machine learning method includes the random forest algorithm.

[0010] Thirdly, this application provides a method for assessing the immune status of obese individuals, comprising the following steps: a) Obtain the detection values ​​of the indicators described in the first aspect for the subject; b) Input the detected value into the Immune Health Related Body Mass Index (IHBMI) calculation formula, output the IHBMI value, and assess whether the subject's immune status is abnormal based on the IHBMI value.

[0011] Optionally, the IHBMI calculation formula is as follows:

[0012] Fourthly, this application provides a method for assessing the immune status of obese individuals, comprising the following steps: a) Obtain the test values ​​of the indicators described in the first aspect of the examinee; b) Input the standardized data of the test values ​​into the Immune Body Mass Index (IBMI). (PCA) The calculation formula outputs IBMI. (PCA) Value, and according to the IBMI (PCA) The value is used to assess whether the subject's immune status is abnormal.

[0013] Optionally, the IBMI (PCA) The calculation formula is: IBMI (PCA) =0.51*x1+0.51*x2-0.39*x3+0.22*x4-0.31*x5+0.24*x6+0.35*x7), Where x1 to x7 represent the standardized values ​​of the following indicators: HGB, HCT, and Ts.MMP. low %, T4Tcm%, T8Tn.MMP low %, NKT.NKG2D + % and T8Tcm.MM.

[0014] Optionally, the standardized values ​​are obtained using the min-max standardization method.

[0015] Fifthly, this application provides the use of reagents for detecting the indicators described in the first aspect in the preparation of products for assessing the immune status of obese individuals, or in the preparation of kits for achieving this use.

[0016] Sixthly, this application provides a kit for assessing the immune status of obese individuals, the kit comprising: The reagents for detecting the combination of indicators described in the first aspect include specific fluorescently labeled antibodies against NKG2D and / or CD4, and fluorescent dyes for detecting mitochondrial membrane potential and mitochondrial mass. Optionally, a user manual may be included, which describes the methods described in the third and / or fourth aspects.

[0017] In summary, this application includes at least one of the following beneficial technical effects: 1. Through multi-dimensional data mining and machine learning screening, HCT, HGB, NKT, and NKG2D were identified for the first time. + %, T8Tcm.MM, T4Tcm %, T8Tn.MMP low %, Ts.MMP lowThe study included seven key indicators, including percentage; this combination has high biological interpretability and clinical relevance, and can effectively quantify "immune obesity" in obese individuals. 2. Immune Body Mass Index (IBMI) constructed based on Principal Component Analysis (PCA) (PCA) The value is significantly positively correlated with BMI. P The ROC curve area under the curve (AUC) of the normal underweight group and the obese group was over 90%, indicating that the model has high predictive accuracy and discrimination ability. It can be used to indicate poor immune status, possibly immune obesity or obesity-induced immune status or immune dysfunction, and assist clinical early intervention and personalized health management. 3. Calculate the Immune Health Related Body Mass Index (IHBMI) using a formula, and correlate it with BMI and IBMI. (PCA) Significant negative correlation ( P <0.001), and the area under the ROC curve (AUC) of the normal underweight group and the obese group is over 90%, indicating that it can be used to indicate poor immune status, possibly immune obesity or obesity-induced immune status or low immune function, and assist clinical early intervention and personalized health management. 4. Through multiple screening steps such as correlation analysis, difference analysis and random forest algorithm, 7 of the most representative indicators were selected from 110 initial indicators, which not only ensured the simplicity and interpretability of the model, but also improved its stability and generalization ability. 5. This method reveals the close relationship between the mitochondrial functional status of NKT cells and their subsets and obesity. The indicators involved can be detected by conventional experimental methods such as flow cytometry. The supporting reagents and detection systems are easy to standardize and industrialize, and are suitable for promotion and application in health check-up centers, hospitals and research institutions. Attached Figure Description

[0018] Figure 1A To reduce the precision of BMI-related means in random forest analysis from largest to smallest; Figure 1B Mean DecreaseGini coefficient for the BMI-related mean in random forest analysis; Figure 1C The distributional cross-validation error for random forest analysis; Figure 2A To demonstrate the correlation between BMI and age; Figure 2B For IBMI (PCA) Correlation analysis with BMI; Figure 2C Correlation analysis between IHBMI and BMI; Figure 2DFor IHBMI and IBMI (PCA) Correlation analysis; Figure 2E For each group IBMI (PCA) Significance analysis of differences; Figure 2F For the significance analysis of the differences in IHBMI among the groups; Figure 2G ROC curve analysis for the normal underweight group and the obese group. Detailed Implementation

[0019] To provide a clearer understanding of the technical features, objectives, and beneficial effects of this application, the technical solution of the present invention will now be described in detail with reference to the following specific embodiments and accompanying drawings. However, this should not be construed as limiting the scope of implementation of the present invention.

[0020] Example 1 Information on gender, BMI, and age was collected from 204 participants aged 5–89 years, with all data conforming to a normal distribution as closely as possible. Reference values ​​for peripheral blood T cell subsets and NK lymphocyte subsets were established. Relevant data on immune cells and mitochondrial parameters included the percentage of lymphocyte subsets, median mitochondrial fluorescence index (Mito-MFI), mitochondrial mass (MM), and percentage of mitochondrial low membrane potential (MMP). low The analysis combined blood routine data and biochemical indicators, including mean platelet volume (MPV), white blood cell count (WBC), red blood cell count (RBC), hemoglobin (HGB), hematocrit (HCT), mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), mean corpuscular hemoglobin concentration (MCHC), red blood cell distribution width (RDW), platelet count (PLT), platelet hematocrit (PCT), and platelet distribution width (PDW), with reference to the Asian obesity BMI standard. Based on BMI values, healthy donors were divided into a normal-to-underweight group (BMI < 20, n = 43), a normal control group (BMI: 20-23, n = 73), a normal-to-overweight group (BMI: 23-27.5, n = 76), and an obese group (BMI > 27.5, n = 12). For each indicator in each group, the quartile values ​​(...) p 25 , p 75 As shown in Table 1.

[0021] Mitochondrial parameter data: Mitochondrial staining with MitoDye-APC (Ubiquitin Biotechnology, Hangzhou, China) and flow cytometry (NovoCyte, Agilent, USA) were used to obtain mitochondrial parameters Mito-MFI and MMP. low % , specifically as follows: Collect 2-5 mL of venous blood using an EDTA anticoagulant tube. Immediately after collection, gently mix the blood sample to prevent clotting. Sample preparation procedure: a) Incubate 100 μL of peripheral blood sample containing EDTA anticoagulant and mixed antibody at room temperature in the dark for 15 min, then add 2 mL of hemolysin to destroy red blood cells; b) Centrifuge the sample at 300×g for 5 min; c) Discard the supernatant, resuspend the precipitate in 200 μL of PBS, transfer it to a container containing MitoDye, and incubate at 37°C in the dark for 30 min; d) Finally, transfer it to a flow cytometer, then count the labeled immune cells using NovoCyte, and perform final analysis and graphical output using NovoExpress software (Agilent Technology, USA).

[0022] Data on immune cells: NKT combined flow cytometry was used for analysis. Antibodies included: CD3 FITC, CD45PerCP-Cy5.5, CD4 PE-Cy7, CD8 APC-Cy7 (Ubiquinone Biotechnology, Hangzhou, China) and TCR γ / δ BV421, NKG2AECD, CD56 KO525, NKG2D BV610 (BioLegend, San Diego, USA).

[0023] Table 1 Indicator abbreviation Normal to low group normal group Normal to high group Obese group BMI 18.370 (16.797, 19.270) 21.700 (21.140, 22.300) 24.545 (23.748, 25.948) 28.985 (28.465, 30.473) WBC 5.400 (4.900, 6.280) 5.800 (5.000, 6.600) 5.900 (5.100, 6.900) 5.750 (5.425, 6.825) RBC 4.530 (4.320, 4.720) 4.490 (4.305, 4.790) 4.600 (4.350, 4.868) 4.835 (4.473, 4.975) HGB 132.000 (128.000, 137.000) 137.000 (130.500, 143.000) 142.500 (129.250, 148.000) 146.000 (139.000, 155.250) HCT 39.500 (38.300, 41.000) 41.200 (39.200, 43.400) 42.450 (39.025, 44.400) 43.700 (41.850, 45.475) MCV 88.600 (84.400, 92.300) 91.500 (88.800, 93.600) 91.250 (89.925, 92.600) 90.950 (88.950, 93.650) MCH 29.500 (28.200, 30.700) 30.300 (29.500, 31.250) 30.400 (29.800, 31.175) 30.550 (30.000, 31.650) MCHC 333.00 (331.00, 338.00) 334.00 (328.00, 337.50) 333.00 (329.00, 337.75) 337.00 (332.00, 342.50) RDW 13.000 (12.500, 13.600) 13.300 (13.000, 14.150) 13.450 (13.100, 14.075) 13.250 (13.125, 13.500) PLT 224.00 (179.00, 289.00) 223.00 (197.00, 251.50) 212.50 (172.00, 249.75) 214.50 (183.75, 237.50) PCT 0.208 (0.165, 0.263) 0.205 (0.176, 0.230) 0.190 (0.155, 0.217) 0.192 (0.171, 0.209) MPV 9.200 (8.500, 9.600) 9.000 (8.400, 9.600) 8.950 (8.400, 9.600) 9.000 (8.125, 9.475) PDW 16.500 (11.000, 17.000) 16.700 (16.300, 17.000) 16.700 (16.300, 17.075) 16.800 (16.525, 17.150) WBC% 0.852 (0.707, 0.915) 0.889 (0.727, 0.964) 0.913 (0.798, 0.964) 0.820 (0.728, 0.935) Lymph% 0.407 (0.325, 0.458) 0.348 (0.284, 0.401) 0.352 (0.293, 0.419) 0.414 (0.318, 0.434) T% 0.628 (0.578, 0.715) 0.627 (0.541, 0.712) 0.581 (0.510, 0.661) 0.562 (0.473, 0.684) Th% 0.452 (0.369, 0.554) 0.416 (0.347, 0.481) 0.429 (0.343, 0.544) 0.427 (0.349, 0.492) Tc% 0.393 (0.317, 0.490) 0.457 (0.379, 0.529) 0.432 (0.338, 0.517) 0.454 (0.341, 0.552) B% 0.098 (0.075, 0.143) 0.097 (0.072, 0.130) 0.096 (0.074, 0.128) 0.087 (0.079, 0.119) NK% 0.256 (0.176, 0.331) 0.244 (0.202, 0.345) 0.315 (0.237, 0.365) 0.343 (0.220, 0.423) Mono% 0.063 (0.056, 0.076) 0.073 (0.056, 0.087) 0.075 (0.066, 0.087) 0.067 (0.059, 0.083) <![CDATA[Lymph.MMP low %]]> 0.735 (0.671, 0.798) 0.659 (0.569, 0.736) 0.640 (0.584, 0.712) 0.704 (0.583, 0.762) <![CDATA[T. MMP low %]]> 0.722 (0.626, 0.792) 0.600 (0.531, 0.704) 0.593 (0.504, 0.658) 0.602 (0.502, 0.703) <![CDATA[Th. MMP low %]]> 0.573 (0.435, 0.738) 0.460 (0.353, 0.596) 0.410 (0.316, 0.566) 0.424 (0.312, 0.615) <![CDATA[Ts. MMP low %]]> 0.776 (0.743, 0.852) 0.688 (0.592, 0.786) 0.664 (0.568, 0.731) 0.672 (0.566, 0.728) <![CDATA[B. MMP low %]]> 0.563 (0.453, 0.660) 0.507 (0.419, 0.624) 0.517 (0.433, 0.605) 0.644 (0.475, 0.672) <![CDATA[NK. MMP low %]]> 0.783 (0.670, 0.859) 0.781 (0.665, 0.859) 0.761 (0.655, 0.851) 0.806 (0.743, 0.891) <![CDATA[CD3 + %]]> 0.196 (0.138, 0.265) 0.181 (0.126, 0.214) 0.173 (0.137, 0.204) 0.183 (0.152, 0.229) <![CDATA[CD4 + %]]> 0.510 (0.408, 0.584) 0.465 (0.420, 0.535) 0.512 (0.426, 0.590) 0.471 (0.421, 0.580) <![CDATA[CD8 + %]]> 0.344 (0.305, 0.425) 0.400 (0.337, 0.460) 0.364 (0.304, 0.431) 0.393 (0.318, 0.464) T4Tcm% 0.214 (0.158, 0.273) 0.242 (0.202, 0.307) 0.247 (0.195, 0.287) 0.300 (0.224, 0.361) T4Tef % 0.015 (0.011, 0.023) 0.018 (0.009, 0.025) 0.018 (0.011, 0.030) 0.013 (0.010, 0.018) T4Tem% 0.330 (0.201, 0.475) 0.366 (0.286, 0.497) 0.424 (0.348, 0.497) 0.406 (0.289, 0.487) T4Tn% 0.402 (0.285, 0.589) 0.333 (0.193, 0.423) 0.257 (0.208, 0.381) 0.236 (0.163, 0.329) T8Tcm% 0.028 (0.018, 0.044) 0.037 (0.019, 0.055) 0.041 (0.025, 0.061) 0.052 (0.038, 0.061) T8Tef % 0.191 (0.109, 0.301) 0.259 (0.161, 0.356) 0.248 (0.167, 0.367) 0.370 (0.222, 0.424) T8Tem% 0.362 (0.255, 0.492) 0.442 (0.324, 0.565) 0.455 (0.340, 0.571) 0.451 (0.354, 0.533) T8Tn% 0.404 (0.126, 0.509) 0.208 (0.098, 0.367) 0.171 (0.094, 0.328) 0.119 (0.058, 0.210) <![CDATA[T4Tcm.PD1 + %]]> 0.203 (0.147, 0.247) 0.182 (0.141, 0.237) 0.162 (0.125, 0.215) 0.186 (0.126, 0.238) <![CDATA[T4Tef.PD1 + %]]> 0.322 (0.250, 0.442) 0.333 (0.237, 0.493) 0.346 (0.232, 0.497) 0.354 (0.264, 0.550) <![CDATA[T4Tem.PD1 + %]]> 0.466 (0.379, 0.579) 0.426 (0.363, 0.532) 0.452 (0.376, 0.503) 0.448 (0.396, 0.471) <![CDATA[T8Tcm.PD1 + %]]> 0.314 (0.194, 0.363) 0.275 (0.211, 0.364) 0.309 (0.237, 0.398) 0.310 (0.277, 0.381) <![CDATA[T8Tef.PD1 + %]]> 0.189 (0.119, 0.305) 0.171 (0.088, 0.273) 0.163 (0.114, 0.261) 0.094 (0.061, 0.192) <![CDATA[T8Tem.PD1 + %]]> 0.361 (0.275, 0.493) 0.389 (0.296, 0.477) 0.437 (0.327, 0.545) 0.365 (0.248, 0.433) <![CDATA[CD3.MMP low %]]> 0.633 (0.553, 0.696) 0.562 (0.487, 0.647) 0.504 (0.445, 0.611) 0.538 (0.490, 0.590) <![CDATA[CD4.MMP low %]]> 0.541 (0.397, 0.626) 0.464 (0.374, 0.557) 0.431 (0.340, 0.571) 0.441 (0.344, 0.520) <![CDATA[CD8.MMP low %]]> 0.768 (0.650, 0.833) 0.642 (0.555, 0.737) 0.611 (0.495, 0.717) 0.602 (0.518, 0.688) <![CDATA[T4Tcm.MMP low %]]> 0.562 (0.353, 0.681) 0.447 (0.314, 0.553) 0.420 (0.298, 0.626) 0.362 (0.251, 0.471) <![CDATA[T4Tef.MMP low %]]> 0.278 (0.224, 0.392) 0.267 (0.183, 0.377) 0.279 (0.161, 0.460) 0.303 (0.168, 0.535) <![CDATA[T4Tem.MMP low %]]> 0.511 (0.346, 0.619) 0.444 (0.351, 0.557) 0.458 (0.368, 0.561) 0.442 (0.352, 0.537) <![CDATA[T4Tn.MMP low %]]> 0.532 (0.411, 0.633) 0.483 (0.362, 0.600) 0.461 (0.313, 0.617) 0.545 (0.398, 0.621) <![CDATA[T8Tcm.MMP low %]]> 0.514 (0.421, 0.694) 0.500 (0.348, 0.620) 0.432 (0.305, 0.572) 0.333 (0.291, 0.425) <![CDATA[T8Tef.MMP low %]]> 0.682 (0.540, 0.773) 0.588 (0.481, 0.704) 0.582 (0.497, 0.727) 0.577 (0.492, 0.655) <![CDATA[T8Tem.MMP low %]]> 0.667 (0.578, 0.804) 0.596 (0.497, 0.679) 0.554 (0.472, 0.667) 0.534 (0.483, 0.627) <![CDATA[T8Tn.MMP low %]]> 0.873 (0.811, 0.927) 0.803 (0.675, 0.899) 0.783 (0.683, 0.877) 0.752 (0.657, 0.886) NK% 0.248 (0.144, 0.323) 0.226 (0.150, 0.304) 0.287 (0.168, 0.379) 0.288 (0.213, 0.372) NKT% 0.022 (0.009, 0.083) 0.034 (0.019, 0.059) 0.036 (0.021, 0.077) 0.043 (0.014, 0.079) γδT% 0.055 (0.025, 0.080) 0.034 (0.022, 0.056) 0.031 (0.014, 0.058) 0.039 (0.021, 0.088) <![CDATA[CD8 + NKT%]]> 0.910 (0.780, 0.967) 0.917 (0.797, 0.967) 0.936 (0.848, 0.982) 0.940 (0.797, 0.999) <![CDATA[CD4 + %]]> 0.304 (0.237, 0.391) 0.305 (0.244, 0.365) 0.308 (0.253, 0.377) 0.274 (0.205, 0.367) <![CDATA[CD8 + %]]> 0.266 (0.214, 0.341) 0.307 (0.243, 0.345) 0.274 (0.221, 0.324) 0.265 (0.239, 0.382) <![CDATA[CD45.MMP low %]]> 0.636 (0.553, 0.700) 0.594 (0.516, 0.662) 0.597 (0.532, 0.659) 0.601 (0.553, 0.695) <![CDATA[NK. MMP low %]]> 0.881 (0.842, 0.933) 0.889 (0.817, 0.942) 0.898 (0.815, 0.945) 0.940 (0.914, 0.973) <![CDATA[NKT.MMP low %]]> 0.867 (0.786, 0.964) 0.818 (0.607, 0.932) 0.811 (0.649, 0.914) 0.777 (0.669, 0.876) <![CDATA[CD8 + NKT.MMP low %]]> 0.857 (0.778, 0.952) 0.802 (0.582, 0.926) 0.789 (0.652, 0.911) 0.840 (0.471, 0.896) <![CDATA[γδT.MMP low %]]> 0.851 (0.750, 0.924) 0.765 (0.656, 0.862) 0.833 (0.695, 0.925) 0.824 (0.763, 0.892) <![CDATA[Th. MMP low %]]> 0.389 (0.286, 0.497) 0.362 (0.282, 0.481) 0.376 (0.301, 0.447) 0.394 (0.336, 0.467) <![CDATA[Ts. MMP low %]]> 0.695 (0.589, 0.800) 0.619 (0.488, 0.736) 0.583 (0.470, 0.705) 0.597 (0.499, 0.679) <![CDATA[NK.NKG2A + %]]> 0.331 (0.227, 0.429) 0.283 (0.164, 0.413) 0.259 (0.135, 0.412) 0.344 (0.252, 0.393) <![CDATA[NKT.NKG2A + %]]> 0.941 (0.867, 0.988) 0.944 (0.874, 0.984) 0.936 (0.869, 0.981) 0.931 (0.826, 0.999) <![CDATA[γδT.NKG2A + %]]> 0.423 (0.294, 0.588) 0.332 (0.184, 0.467) 0.352 (0.162, 0.583) 0.357 (0.216, 0.560) <![CDATA[CD8 + NKT.NKG2A + %]]> 0.118 (0.046, 0.260) 0.143 (0.034, 0.203) 0.142 (0.060, 0.265) 0.148 (0.086, 0.316) <![CDATA[Th.NKG2A + %]]> 0.000 (0.000, 0.000) 0.000 (0.000, 0.001) 0.000 (0.000, 0.000) 0.000 (0.000, 0.000) <![CDATA[Ts.NKG2A + %]]> 0.022 (0.012, 0.034) 0.027 (0.010, 0.099) 0.025 (0.013, 0.091) 0.021 (0.005, 0.033) <![CDATA[NK.NKG2D + %]]> 0.965 (0.913, 0.994) 0.952 (0.916, 0.986) 0.955 (0.908, 0.975) 0.979 (0.967, 0.996) <![CDATA[NKT.NKG2D + %]]> 0.857 (0.667, 0.923) 0.912 (0.798, 0.970) 0.915 (0.822, 0.970) 0.988 (0.929, 1.000) <![CDATA[γδT.NKG2D + %]]> 0.943 (0.892, 0.979) 0.957 (0.920, 0.985) 0.958 (0.904, 0.981) 0.976 (0.954, 0.994) <![CDATA[CD8 + NKT.NKG2D + %]]> 0.875 (0.628, 0.957) 0.918 (0.822, 0.983) 0.928 (0.818, 0.985) 0.994 (0.940, 1.000) <![CDATA[Th.NKG2D + %]]> 0.070 (0.026, 0.134) 0.064 (0.022, 0.180) 0.071 (0.022, 0.187) 0.012 (0.006, 0.112) <![CDATA[Ts.NKG2D + %]]> 0.022 (0.012, 0.034) 0.027 (0.010, 0.099) 0.025 (0.013, 0.091) 0.021 (0.005, 0.033) <![CDATA[NK.NKG2A + .MMP low %]]> 0.941 (0.888, 0.982) 0.906 (0.798, 0.979) 0.915 (0.829, 0.971) 0.934 (0.907, 0.974) <![CDATA[NKT.NKG2A + .MMP low %]]> 0.750 (0.000, 1.000) 0.571 (0.000, 1.000) 0.667 (0.011, 0.930) 0.650 (0.000, 0.972) <![CDATA[γδT.NKG2A + .MMP low %]]> 0.910 (0.772, 0.969) 0.861 (0.667, 0.950) 0.881 (0.615, 0.971) 0.847 (0.749, 0.930) <![CDATA[CD8 + NKT.NKG2A + .MMP low %]]> 0.833 (0.000, 1.000) 0.667 (0.232, 0.939) 0.683 (0.400, 0.885) 0.608 (0.131, 0.880) <![CDATA[Th.NKG2A + .MMP low %]]> 0.000 (0.000, 0.000) 0.000 (0.000, 0.000) 0.000 (0.000, 0.000) 0.000 (0.000, 0.000) <![CDATA[Ts.NKG2A + .MMP low %]]> 0.832 (0.667, 1.000) 0.750 (0.571, 0.907) 0.709 (0.500, 0.866) 0.691 (0.406, 0.750) <![CDATA[NK.NKG2D + .MMP low %]]> 0.920 (0.846, 0.960) 0.907 (0.836, 0.955) 0.902 (0.829, 0.964) 0.957 (0.935, 0.983) <![CDATA[NKT.NKG2D + .MMP low %]]> 0.875 (0.818, 0.966) 0.831 (0.656, 0.936) 0.815 (0.688, 0.920) 0.859 (0.682, 0.909) <![CDATA[γδT.NKG2D + .MMP low %]]> 0.928 (0.807, 0.951) 0.872 (0.778, 0.924) 0.891 (0.787, 0.962) 0.885 (0.803, 0.921) <![CDATA[CD8 + NKT.NKG2D + .MMP low %]]> 0.889 (0.759, 0.962) 0.825 (0.614, 0.924) 0.830 (0.703, 0.917) 0.846 (0.484, 0.916) <![CDATA[Th.NKG2D + .MMP low %]]> 0.500 (0.375, 0.588) 0.462 (0.362, 0.596) 0.465 (0.378, 0.547) 0.544 (0.431, 0.667) <![CDATA[Ts.NKG2D + .MMP low %]]> 0.690 (0.590, 0.840) 0.607 (0.498, 0.747) 0.602 (0.507, 0.736) 0.603 (0.502, 0.673) WBC.MM 0.296 (0.200, 0.394) 0.349 (0.267, 0.507) 0.369 (0.284, 0.474) 0.313 (0.259, 0.618) Lymph.MM 0.140 (0.082, 0.182) 0.169 (0.108, 0.271) 0.170 (0.118, 0.238) 0.127 (0.099, 0.205) Mono.MM 0.883 (0.553, 1.145) 0.776 (0.640, 1.095) 0.874 (0.615, 1.111) 1.006 (0.673, 1.402) T.MM 0.156 (0.107, 0.181) 0.194 (0.132, 0.338) 0.217 (0.152, 0.345) 0.175 (0.119, 0.404) Th.MM 0.217 (0.155, 0.465) 0.393 (0.230, 0.595) 0.416 (0.243, 0.609) 0.392 (0.157, 0.745) Ts.MM 0.120 (0.082, 0.153) 0.164 (0.106, 0.220) 0.159 (0.115, 0.228) 0.142 (0.104, 0.203) B.MM 0.189 (0.137, 0.282) 0.225 (0.148, 0.371) 0.204 (0.142, 0.350) 0.144 (0.112, 0.227) NK.MM 0.100 (0.040, 0.168) 0.111 (0.055, 0.191) 0.107 (0.063, 0.152) 0.074 (0.059, 0.107) <![CDATA[CD3 + .MM]]> 0.202 (0.135, 0.286) 0.259 (0.179, 0.338) 0.267 (0.172, 0.355) 0.192 (0.163, 0.325) <![CDATA[CD4 + CD8 - .MM]]> 0.292 (0.212, 0.475) 0.363 (0.247, 0.588) 0.387 (0.253, 0.527) 0.351 (0.222, 0.531) <![CDATA[CD4 - CD8 + .MM]]> 0.139 (0.118, 0.188) 0.194 (0.136, 0.256) 0.189 (0.134, 0.307) 0.154 (0.119, 0.298) T4Tcm.MM 0.282 (0.184, 0.531) 0.341 (0.241, 0.629) 0.402 (0.186, 0.542) 0.424 (0.271, 0.561) T4Tef.MM 0.735 (0.575, 1.067) 0.740 (0.489, 1.045) 0.684 (0.284, 0.968) 0.506 (0.191, 0.773) T4Tem.MM 0.315 (0.217, 0.564) 0.396 (0.267, 0.650) 0.412 (0.256, 0.619) 0.403 (0.221, 0.521) T4Tn.MM 0.258 (0.198, 0.405) 0.325 (0.181, 0.453) 0.320 (0.194, 0.502) 0.293 (0.176, 0.431) T8Tcm.MM 0.063 (0.040, 0.088) 0.092 (0.054, 0.137) 0.109 (0.057, 0.153) 0.169 (0.084, 0.220) T8Tef.MM 0.170 (0.131, 0.258) 0.205 (0.141, 0.379) 0.203 (0.130, 0.330) 0.206 (0.084, 0.364) T8Tem.MM 0.147 (0.117, 0.204) 0.195 (0.128, 0.346) 0.231 (0.150, 0.374) 0.257 (0.154, 0.366) T8Tn.MM 0.128 (0.102, 0.173) 0.144 (0.103, 0.190) 0.135 (0.100, 0.192) 0.117 (0.084, 0.158)

[0024] Example 2

[0025] 2.1 Correlation Analysis Construct a correlation analysis between different groups of data ( Spearman Analysis revealed 20 indicators with strong correlation to BMI from 110 indicators. ρ >0.21 or ρ <-0.21, p <0.005. The Mann-Whitney U test showed that these 20 indicators were significantly different between the normal underweight group and the overweight + obese group. p <0.005), the results are shown in Table 2.

[0026] Table 2. Strongly correlated indicators and significant differences

[0027] 2.2 Random Forest Analysis Subsequently, random forest analysis was performed on 20 indicators to screen for the best indicators. The number of decision trees was set to 8000, the cross-validation fold count to 10, the variable reduction factor to 1.1, and the sample size n=204. The results of the random forest analysis of the differential indicators are shown in Table 3. Through random forest analysis, 7 significantly different indicators were finally selected (HCT, HGB, NKT, NKG2D). + %, T8Tcm.MM, T4Tcm %, T8Tn.MMP low %, Ts.MMP low %, Figure 1A , Figure 1B The blue indicators (indicators) are derived from the cross-validation error curve. Figure 1C ).

[0028] Table 3

[0029] 2.3 Immune status assessment 2.3.1 Principal Component Analysis Principal component analysis (PCA) was performed on the above seven key indicators and presented in a planar format. First, the original test values ​​were standardized using the min-max data method (all data were standardized to values ​​between 0 and 1). Then, dimensionality reduction analysis was performed using SPSS software. Principal component analysis can reduce the dimensionality of the data, allowing us to view the relevant characteristics of the principal components in a two-dimensional space. The results are shown in Tables 4-6. Table 4 is the total variance explanation table for the PCA analysis, showing the seven principal components (PC1~PC7 contribution rates of 33.72%, 23.67%, 15.81%, 10.98%, 9.06%, 6.38%, and 0.39%, respectively); Table 5 is the component matrix table; and Table 6 is the component coefficient table.

[0030] Table 4

[0031] Table 5

[0032] Table 6

[0033] By examining the principal component correlation features at a two-dimensional level, seven principal components are obtained. y 1~ y The expression for 7 is as follows: y 1 =0.51*x 1 +0.51*x 2 -0.39*x3 +0.22*x 4 -0.31*x 5 +0.24*x 6 +0.35*x 7； y 2 =0.44*x 1 +0.44*x 2 +0.35*x 3 +0.12*x 4 +0.50*x 5 -0.11*x 6 -0.45*x 7； y 3 =-0.19*x 1 -0.19*x 2 +0.03*x 3 +0.65*x 4 +0.25*x 5 +0.66*x 6 -0.06*x 7； y 4 =0.06*x 1 +0.05*x 2 -0.08*x 3 -0.72*x 4 +0.23*x 5 +0.64*x 6 -0.05*x 7； y 5 =0.03*x 1 +0.08*x 2 +0.76*x 3 -0.04*x 4 -0.03*x 5 +0.10*x 6 +0.63*x 7； y 6 =-0.01*x 1 -0.06*x 2 -0.36*x 3 +0.01*x 4 +0.73*x 5 -0.25*x 6 +0.52*x 7； y 7 =0.71*x 1 -0.71*x 2 +0.04*x 3 +0.01*x 4 -0.03 * x 5 +0.01*x 6 +0.003*x 7； x 1~ x 7 represent HGB, HCT, and Ts.MMP respectively. low %, T4Tcm%, T8Tn.MMP low %, NKT.NKG2D + The min-max normalized values ​​of %, T8Tcm.MM parameters.

[0034] ROC curve analysis of the components generated by principal component analysis (PCA) revealed that the first principal component (y1) had the largest area under the curve (AUC), reaching 94.186%, indicating its optimal discriminative ability in distinguishing the immune status of individuals with different BMIs. Therefore, we define y1 as the "Immune Body Mass Index (PCA version)," abbreviated as "IBMI." (PCA) (IBMI, Immune-related body mass index), its calculation formula is: IBMI (PCA) =PC1= y 1 = 0.51*x 1 +0.51* x 2 -0.39*x 3 +0.22*x 4 -0.31*x5 +0.24*x 6 +0.35*x 7， x 1~ x All 7 values ​​are min-max standardized values.

[0035] 2.3.2 Direct Calculation Method To simplify calculations, we de-standardized the formula for calculating IBMI in PC1, using the pre-standardized test values. The resulting value is defined as the "Immune Healthy-Related Body Mass Index" (IHBMI). (T8Tn.MMP) low % and Ts.MMP low The percentage indicator represents the negative correlation contribution (as part of the numerator of the division), HGB, HCT, T4Tcm%, NKT.NKG2D + %, T8Tcm.MM represents a positive correlation contribution (as part of the denominator in division). To reduce the error (CV) in the calculation, the sum of the indices lg( x Furthermore, referring to the research conclusions of Rothman KJ. [Rothman KJ. BMI-related errors in the measurement of obesity. Int J Obes (Lond). 2008 Aug;32 Suppl 3:S56-9.] and Melissa Dolan et al. [Dolan M, Libby KA, Ringel AE, et al. Ageing, immune fitness and cancer. Nat RevCancer. 2025 Aug 14.]: BMI does not necessarily reflect changes that occur with age, but changes in immunity are age-dependent. Therefore, we also include age in the formula calculation, and finally obtained the IHBMI calculation formula: , The formula allows you to calculate IHBMI by directly substituting the test index values ​​without standardization.

[0036] 2.3.3 IHBMI or IBMI (PCA) Results of immune status assessment IBMI of some subjects obtained by principal component calculation （PCA） The values ​​and IHBMI obtained by direct calculation are shown in Table 7, and the median values ​​of the seven indicators included in the formula calculation in each group are shown in Table 8.

[0037] Table 7

[0038] Table 8

[0039] We then analyzed IHBMI and IBMI. (PCA) The correlation coefficients with BMI are similar. ρ =-0.45 vs. The correlation between age and BMI was 0.46, which was significantly higher than that between age and BMI (BMI). vs. IHBMI: 0.37 vs. -0.45; BMI vs. IBMI (PCA) 0.37 vs. 0.46), the result is as follows Figures 2A-2C As shown. The IHBMI calculated by this formula was then compared with IBMMI. (PCA) The correlation, such as Figure 2D As shown, ρ = -0.87, indicating a very strong negative correlation between the two. The IHBMI or IBMI values ​​were also compared between the groups of subjects. (PCA) The results of the significance of the differences are as follows: Figure 2E and Figure 2F As shown, IHBMI or IBMI (PCA) In the analysis of differences among the groups, except for the normal group and the normal-to-overweight group, there was no significant difference. P >0.05), and all other groups showed significant differences ( P <0.05); the IHBMI score ranges for each group are: underweight group (2.757~2.789), normal group (2.738~2.766), overweight group (2.733~2.756), and obese group (2.729~2.744); the IBMI scores for each group are as follows: (PCA) The score ranges are as follows: Underweight group (-0.010~0.328), Normal group (0.216~0.649), Overweight group (0.309~0.763), and Obese group (0.575~0.924). Additionally, ROC curves were plotted for the underweight and obese groups, and the results are as follows. Figure 2G As shown, IHBMI and IBMI (PCA) They have similar ROC curve areas (0.94186). vs. 0.94477). IHBMI or IBMI (PCA)The change was gradual, progressing from the underweight group to the normal group, then to the overweight group, and finally to the obese group, exhibiting a step-like pattern. This indicator can be used as a quantitative method for BMI-related indicators in immunity.

[0040] 2.3.3 IBMI (PCA) Adjustment As shown in Table 7 and Figure 2B As shown, IBMI (PCA) The value is positively correlated with BMI, meaning that as BMI increases, IBMI also increases. (PCA) The BMI also increases accordingly. However, this trend is inconsistent with the actual biological significance of immune function. From an immunological perspective, chronic low-grade inflammation and metabolic disorders associated with obesity usually lead to a decline in immune regulation, meaning that individuals with high BMI are more prone to immunosuppression or immune imbalance. Therefore, the ideal immune assessment index should decrease with increasing BMI to reflect a negative change in immune function. To more accurately reflect the negative correlation between immune status and BMI, we will use IBMI... (PCA) Taking a negative number, it is also defined as the "Immune Health Related Body Mass Index", abbreviated as IHBMI. (PCA) Its relationship with IBM (PCA) The relationship is linearly dependent, and the calculation formula is: IHBMI≈IHBMI (PCA) = a×IBMI (PCA) + b, where a is -0.0599 and b is 2.7793.

[0041] After this adjustment, IHBMI (PCA) With BMI and IBMI (PCA) The negative correlation is consistent with the biological reality that immune function may be weakened in people with high BMI, making it more suitable for the assessment and risk stratification of obesity-related immune status and applicable to the quantification of BMI-related indicators in immunity.

[0042] Of course, the above description is only a specific embodiment of this application and is not intended to limit the scope of the invention. All equivalent changes or modifications made in accordance with the features and principles described in the claims of this invention should be included in the scope of the claims of this invention.

Claims

1. A combination of indicators for assessing immune status in an obese population, characterized in that, The combination of indicators includes the following indicators: HCT, HGB, Ts.MMP low , T4 Tcm, T8 Tn.MMP low , NK T.NKG2D + and T8 Tcm.MM.

2. A method of screening for a combination of markers according to claim 1, characterized in that, The method comprises the following steps: (1) collecting a peripheral blood sample; (2) detecting the blood routine index, the cell percentage index, the mitochondrial mass index and the mitochondrial low membrane potential percentage index of the sample in step (1); (3) performing correlation analysis and difference analysis screening on the indexes in step (2) to obtain BMI-related indexes; (4) performing analysis screening on the indexes in step (3) by combining machine learning technology to obtain the index combination.

3. The method of claim 2, wherein, The machine learning method comprises a random forest algorithm.

4. A method of assessing immune status in an obese population, characterized in that, The method comprises the following steps: a) obtaining the detection value of the index of the subject as claimed in claim 1; b) inputting the detection value into the IHBMI calculation formula to obtain an IHBMI value, and evaluating whether the immune state of the subject is abnormal according to the IHBMI value.

5. The evaluation method according to claim 4, characterized in that The The IHBMI calculation formula is: 。 6. A method of assessing immune status in an obese population, characterized in that, The method comprises the following steps: a) obtaining the detection value of the index of the subject as claimed in claim 1; b) inputting the data of the normalized detection value into IBM I (PCA) In the calculation formula, output IBM I (PCA) value, and evaluating whether the immune state of the subject is abnormal according to the IBM I (PCA) value.

7. The evaluation method according to claim 6, characterized in that, The IBM I (PCA) The calculation formula is: IBM I (PCA) = 0.51*x 1 +0.51*x 2 -0.39*x 3 +0.22*x 4 -0.31*x 5 +0.24*x 6 +0.35*x 7 , wherein, x 1 to x 7 represent the following index detection values after standardization: HGB, HCT, Ts.MMP low %, T4Tcm%, T8Tn.MMP low %, NKT.NKG2D + %, and T8Tcm.MM.

8. The evaluation method according to claim 6, characterized in that The standardized numerical value is obtained by using a min-max standardization method.

9. Use of a reagent for detecting the index as claimed in claim 1 in the preparation of a product for evaluating the immune state of an obese population, or in the preparation of a kit for achieving the use.

10. A kit for assessing the immune status of an obese population, characterized in that, The kit comprises: a reagent for detecting the index combination as claimed in claim 1, the reagent comprising a specific fluorescently labeled antibody for NKG2D and / or CD4, and a fluorescent dye for detecting mitochondrial membrane potential and mitochondrial mass; Optionally, a user instruction manual is included, which records the method as claimed in any one of claims 4-8.

Images