Female senescence marker and female senescence clock determination method

Abstract

The invention relates to a method for measuring a female aging marker and a female aging clock, and belongs to the technical field of medical examination. A group of female senescence markers comprises a plurality of cell factors related to female senescence, and the cell factors comprise but are not limited to a cell factor 6Ckine, a cancer antigen 125, a leukocyte differentiation antigen 14, a leukocyte differentiation antigen 99-L2, follicle stimulating hormone, a growth differentiation factor 15, a low-density lipoprotein receptor, a neutrophil activating protein-2, and a cell factor Ckine. P-hydroxybenzoate esterase and a vascular endothelial growth factor C. The method has the advantages that a group of specific senescence markers related to female senescence are confirmed, a female biological senescence clock determination model is constructed, and the biological age of a female individual is evaluated and the senescence speed is predicted through quantitative analysis of the biomarkers.

Description

Technical Field

[0001] The invention relates to a method for determining female aging markers and female aging clocks, and belongs to the technical field of medical testing. Background Art

[0002] With the increasing aging of society, research on aging biology has become a hot topic.

[0003] Aging markers refer to those biological molecules and cell characteristics that are closely related to the aging process. Common aging markers in current research include but are not limited to: telomere shortening, oxidative stress products, inflammatory markers, metabolites, protein aggregates, etc. The detection of these markers usually requires specific biochemical analysis techniques, such as ELISA, Western blot, mass spectrometry, etc. Despite this, a single marker is often difficult to fully reflect the aging status of an individual.

[0004] Research has shown that gender plays an important role in the aging process. There are significant differences between men and women in terms of physiology, hormone levels, genetic susceptibility, etc., and these differences affect the rate and characteristics of aging. Some aging markers show gender specificity. For example, a decrease in estrogen levels in women is closely related to the aging process, and changes in testosterone levels in women are also associated with aging. In addition, the risk of certain diseases and the rate of aging differ between the sexes.

[0005] In recent years, researchers have begun to focus on the role of gender in the aging process and explore gender-specific aging markers. For example, some studies have attempted to reveal the impact of gender on aging by analyzing gene expression profiles, proteome and metabolome data of different genders.

[0006] Although sex differences have received attention in aging research, current research is relatively limited in scope, focusing mainly on a few known sex-specific markers and lacking exploration of broader sex differences. Existing studies often focus on a single sex or a single marker, lacking research that integrates and analyzes aging marker data from different sexes, which limits the ability to fully understand sex differences.

[0007] The concept of the aging clock originates from the biological record of time in organisms. The aging clock refers to the biomarkers in the body that reflect age changes. Aging markers are changes at the molecular, cellular or tissue levels related to the aging process, and aging acceleration refers to the rate of change of an individual's aging rate. At present, the determination of the aging clock mainly relies on biomarkers, such as telomere length and DNA methylation levels. These markers can reflect the replication history and aging state of cells. However, existing aging clock research mostly focuses on a single marker, lacking an evaluation system that integrates multiple biomarkers.

[0008] Cytokines are small secretory proteins released by cells that play an important role in cell signaling and are crucial in the aging process, affecting inflammation, cell aging, and tissue repair mechanisms. Cytokines are messengers of intercellular communication and they play a key role in regulating biological processes such as immune responses, inflammatory processes, tissue repair, and cell proliferation. Using cytokines as indicators of the aging clock can reflect systemic biological changes, rather than just local or tissue-specific changes.

[0009] Changes in cytokine levels can provide real-time information about an individual's physiological state. By monitoring changes in cytokines, the aging process can be assessed dynamically rather than relying solely on static biomarkers. Many aging-related diseases, such as cardiovascular disease, neurodegenerative diseases, and cancer, are associated with abnormal levels of specific cytokines. Therefore, incorporating cytokines into the assessment of the aging clock will help to identify and predict the occurrence of these diseases early. Aging is associated with a chronic, low-grade inflammatory state (known as "inflammaging"). Cytokines, especially inflammatory cytokines such as interleukins (ILs), tumor necrosis factor (TNF), and interferons (IFNs), are key indicators of the inflammatory state. Monitoring the levels of these cytokines can serve as a useful tool for assessing the risk of aging- and inflammation-related diseases.

[0010] Many cytokines can be easily measured by blood tests and other biological sample detection methods. These detection methods are relatively mature and easy to standardize, making cytokines feasible and reliable biomarkers in aging clock research. Because cytokine levels are affected by individual lifestyle, genetic factors, and environmental factors, they can serve as the basis for personalized aging assessment and intervention. This helps to develop anti-aging strategies that target individual specific needs. As indicators of the aging clock, cytokines can connect multiple disciplines such as immunology, molecular biology, physiology, and pathology, promote interdisciplinary research, and thus have a more comprehensive understanding of the aging process. Cytokines not only serve as biomarkers of aging, but also as potential targets for intervention and treatment. Regulating the levels of specific cytokines or their signaling pathways may help slow the aging process or treat aging-related diseases. By analyzing the expression patterns and changing trends of cytokines, the aging rate and possible lifespan of an individual can be predicted, providing important information for personalized health management. Summary of the invention

[0011] The technical problem to be solved by the present invention is: in view of the deficiencies of existing research, a method for determining aging markers and aging clocks that comprehensively considers gender differences is proposed. The method provides female-specific aging markers and also constructs a more comprehensive female aging clock model based on cytokines, aiming to provide a more accurate scientific basis for female aging intervention.

[0012] To achieve the above objectives, the present invention adopts the following technical solutions.

[0013] A group of female aging markers, including multiple cytokines associated with female aging, including but not limited to: growth differentiation factor 15 (GDF-15), leukocyte differentiation antigen 99-L2 (CD99-L2), leukocyte differentiation antigen 14 (CD14), follicle stimulating hormone (FSH), cytokine 6Ckine (6Ckine), parahydroxybenzoate esterase (PON1), cancer antigen 125 (CA125).

[0014] Cytokine 6Ckine (6Ckine), also known as CCL21, is a small cytokine belonging to the CC chemokine family. It can be constitutively expressed at the beginning of lymphatic vessels and in lymphoid organs, and is present on several stromal cells and high endothelial venules (HEVs). It can bind to glycosaminoglycans (GAGs) and be fixed on the surface of endothelial cells. As a chemokine, 6Ckine can inhibit hematopoiesis and stimulate chemotaxis. It has an in vitro chemotactic effect on thymocytes and activated T cells, but has no chemotactic effect on B cells, macrophages or neutrophils. At the same time, CCL21 is also an effective T cell migration and adhesion stimulator, which can bind to the glycoprotein PSGL-1 on T cells to promote the migration of T cells to secondary lymphoid organs.

[0015] Cancer antigen 125 (CA125) is a glycoprotein that originates from the coelomic epithelium during embryonic development. It does not exist in normal ovarian tissue, so it is most commonly found in the serum of patients with epithelial ovarian tumors (serous tumors). Its diagnostic sensitivity is high, but its specificity is poor. It does not exist in mucinous ovarian tumors. 80% of patients with ovarian epithelial tumors have elevated serum CA125, but nearly half of early cases do not have elevated levels, so it is not used alone for early diagnosis of ovarian epithelial cancer. Serum CA125 is related to disease progression in 90% of patients, so it is mostly used for disease detection and efficacy evaluation.

[0016] Leukocyte differentiation antigen 14 (CD14), a surface antigen, is a member of the cell surface glycoprotein family. The protein encoded by the FSH gene is preferentially expressed in monocytes / macrophages. CD14 is a high-affinity receptor for the complex of Gram-negative bacterial endotoxin lipopolysaccharide LPS and LPS binding protein (LBP). CD14 recognizes and binds to LPS, causing cell tyrosine phosphorylation, nuclear factor NF-κB translocation, triggering cytokine release and oxygen free radical production, and plays an important role in a series of reactions caused by the body's immune and defense systems. CD14 is expressed in humans and mice with tissue specificity and cell specificity. mCD14 is mainly distributed on the cell surface of monocytes, macrophages and dendritic cells, in the membrane secretory bodies and azurophilic blue granules in the cytoplasm of neutrophils, while mCD14 is not found on the surface of endothelial cells, epithelial cells, etc.

[0017] The leukocyte differentiation antigen 99-L2 (CD99-L2) gene encodes a highly conserved, widely expressed glycosylated transmembrane protein that plays a role in the human immune system. It is mainly present on the surface of lymphocytes in the body's lymphoid tissues and spleen and other organs, and also on the surface of some cells in tissues such as the bone marrow, liver, pancreas and brain. The CD99-L2 protein is considered to be an important adhesion molecule that plays an important role in inflammatory responses, leukocyte infiltration, lymphoma, autism spectrum disorders, cerebral palsy, epilepsy, etc. It may be involved in a variety of biological processes such as cell signal transduction, cell adhesion and cell differentiation. CD99-L2 may also be associated with the occurrence and development of certain diseases, such as malignant tumors such as lymphoma and leukemia and certain autoimmune diseases.

[0018] Follicle-stimulating hormone (FSH) is a concept mentioned in the six sex hormone tests. It is a protein hormone secreted by the anterior pituitary gland and plays an important role in both female and male reproductive functions. In women, the main function of FSH is to stimulate the maturation of follicles in the ovaries and promote the release of eggs. In the early stages of the menstrual cycle, FSH levels rise, which promotes the maturation of follicles in the ovaries, so that one of the follicles finally matures and is ready for fertilization. After ovulation, FSH levels drop. In men, the main function of FSH is to stimulate sperm production in the testicles. FSH stimulates spermatogonia in the testicles to cause them to divide and differentiate into mature sperm. If the FSH level in a man is too low, it will affect sperm production and lead to reproductive problems such as infertility.

[0019] Growth differentiation factor 15 (GDF-15) is a cytokine produced by tumor cells, inflammatory cells and cardiomyocytes. GDF-15 has important biological significance in regulating cell growth, proliferation, apoptosis and inflammatory response. In recent years, studies have found that the expression level of GDF-15 in a variety of diseases is closely related to the occurrence and prognosis of the disease, such as cardiovascular disease, diabetes, chronic kidney disease, tumors, etc. In cardiovascular disease, GDF-15 is considered to be an important prognostic indicator and biomarker of disease activity, which can predict the occurrence of cardiovascular events and the risk of death of patients. In tumors, the expression of GDF-15 is closely related to the malignancy of the tumor, disease progression and prognosis, so it is considered to be a potential tumor biomarker and therapeutic target. In addition, GDF-15 is also involved in regulating physiological processes such as appetite and energy metabolism, so it also has certain biological significance in metabolic diseases such as obesity and metabolic syndrome.

[0020] Para-hydroxybenzoate esterase (PON1) is an esterase present in the blood, which is mainly synthesized by the liver and also exists in other tissues. After PON1 is synthesized in the liver, the enzyme is secreted into the circulation, binds to high-density lipoprotein (HDL) particles and hydrolyzes thiolactones and xenobiotics. The mechanisms of its protective effect against atherosclerosis mainly include reducing macrophage autooxidation and its ability to oxidize LDL, inhibiting cholesterol synthesis, and promoting cholesterol efflux. Therefore, PON1 is considered an important molecule for the prevention of cardiovascular disease. In addition, PON1 is also considered to be a molecule related to toxicant metabolism and drug clearance. For example, PON1 can metabolize and clear some organophosphorus pesticides, thereby reducing their toxic effects on the human body. Therefore, PON1 has also been widely studied for its role in toxicant metabolism and drug clearance.

[0021] Another aspect of the present invention provides a method for predicting the age of a female individual, comprising the following steps:

[0022] a. Obtain biological samples from female subjects;

[0023] b. determining the content of female aging markers in the biological sample;

[0024] c. inputting the content value of the female aging marker into a preset cytokine aging clock;

[0025] d. outputting the estimated age of the female individual based on the cytokine aging clock.

[0026] The female aging markers include but are not limited to: cytokine 6Ckine, cancer antigen 125, leukocyte differentiation antigen 14, leukocyte differentiation antigen 99-L2, follicle stimulating hormone, growth differentiation factor 15, low-density lipoprotein receptor, neutrophil activation protein-2, parahydroxybenzoate esterase and vascular endothelial growth factor C.

[0027] The biological sample is blood, serum, plasma or saliva; the determination step adopts immunoassay, mass spectrometry, biochemical analysis method or biochip technology.

[0028] The calculation formula of the cytokine aging clock is: predicted age = -97.40798 + GDF-15 × 52.78893 + CD99-L2 × 12.46563 + CD14 × 9.0327 + FSH × 6.32722 + 6Ckine × 0.63787 + PON1 × -10.41995 + CA125 × -10.44494, where the cytokine values ​​are all detection concentrations in pg / mL. Or predicted age = -97.40798 + GDF-15 × 52.78893 + CD99-L2 × 12.46563 + CD14 × 9.0327 + FSH × 6.32722 + 6Ckine × 0.63787 + PON1 × -10.41995 + CA125 × -10.44494, where the cytokine values ​​are all detected concentrations in pg / mL.

[0029] The method for predicting the age of a female individual also includes calculating the aging acceleration, namely Δage.

[0030] The calculation formula of the aging acceleration Δage is:

[0031] ΔAge 初始 = Predicted age – Chronological age

[0032] Δage=Δage 初始 —loess(chronological age—Δage 初始 )

[0033] Among them, chronological age refers to the age of the subject itself; "Δage" is the indicator finally obtained by the model of the present invention that can be used to evaluate aging. A large Δage value indicates accelerated aging and possible health problems; Loess is a local polynomial regression fitting (Local Polynomial Regression Fitting), which uses the initial aging acceleration Δage minus the fitted age difference to correct the influence of age on aging acceleration.

[0034] The present invention uses cytokine antibody chip technology, and on the basis of screening 1000 cytokines, identifies 10 cytokines that are closely related to female aging, and uses Lasso regression to construct a female aging clock model. Subsequently, after verification and optimization of 715 people, a combination of 7 cytokines was finally determined as candidate cytokines for aging markers in female plasma. Through the calculation of the aging clock model, the correlation coefficient between the predicted age and the chronological age is 0.97, and the root mean square error RMSE = 4.45. The predicted age is subtracted from the chronological age to obtain the difference, and the age is corrected by LOESS regression to obtain the individual's aging acceleration. The results show that the aging acceleration and individual laboratory test indicators such as liver and kidney function, intrinsic function and somatic function have a significant correlation, and the aging acceleration can be used to use the Logistic regression method to determine whether the individual has a disease. This study established an aging clock for women, accurately predicted age, and the calculated aging acceleration can be used to assess the health status of individuals.

[0035] A system for predicting the age of a female individual, comprising:

[0036] a. Biological sample collection module, which is used to collect biological samples from female individuals, such as blood, urine, saliva or other body fluids. The module can include various sampling devices, such as needles, blood collection tubes, swabs, etc. The collected biological samples can be stored in appropriate containers for subsequent processing and analysis.

[0037] b. A polypeptide content determination module, which is used to determine the content of at least 10 female aging markers in the collected biological sample. These markers may include but are not limited to: cytokine 6Ckine, cancer antigen 125, leukocyte differentiation antigen 14, leukocyte differentiation antigen 99-L2, follicle stimulating hormone, growth differentiation factor 15, low-density lipoprotein receptor, neutrophil activation protein-2, parahydroxybenzoate esterase and vascular endothelial growth factor C.

[0038] The module can use various biochemical methods to measure the content of these markers, such as antibody array, enzyme-linked immunosorbent assay (ELISA), chemiluminescence assay (CLIA), mass spectrometry (MS), etc.

[0039] c. A cytokine aging clock module, which is used to bring the measured polypeptide content into a prediction algorithm to predict the age of the female individual. This module can use a machine learning model or a statistical model to predict age.

[0040] The prediction algorithm module is based on a machine learning model or a statistical model.

[0041] The machine learning models include but are not limited to Lasso regression, random forest, XGBoost and SVM.

[0042] Lasso Regression: A linear regression model that uses L1 regularization to select feature variables. Lasso regression is able to effectively select important features through L1 regularization, which is particularly important when dealing with high-dimensional data sets. When predicting the age of women, it can automatically remove irrelevant variables, thereby simplifying the model and improving accuracy. Because Lasso tends to compress the coefficients of certain features to zero, the results of the model are easier to interpret, allowing analysts to clearly understand which factors have the greatest impact on age prediction.

[0043] Random Forest: An ensemble learning model that uses multiple decision trees to predict outcomes. Random Forest reduces the risk of overfitting that may be caused by a single tree by integrating multiple decision trees for prediction. This allows the model to maintain a high level of generalization when faced with complex data. Random Forest provides an assessment of feature importance to help analysts identify the variables that have the greatest impact on predicting a woman's age. For example, features such as gender, occupation, or education level may be identified as key factors. Random Forest has a good tolerance for missing values ​​and can still make effective predictions even when the data is incomplete.

[0044] XGBoost: A gradient boosting model that uses decision trees to predict results. XGBoost is known for its efficient computing speed and excellent prediction performance, and is particularly suitable for processing large-scale data sets. In the prediction of female age, it can quickly iterate and optimize the model to improve prediction accuracy. XGBoost's built-in regularization mechanism can effectively prevent overfitting, making the model perform better on complex data sets. Supports multiple loss functions and evaluation indicators, allowing users to adjust the model according to specific needs to adapt to different data characteristics and business scenarios.

[0045] SVM: A support vector machine model that uses kernel functions to predict results. SVM performs classification by constructing a maximum margin hyperplane, and can also effectively classify linearly inseparable data through kernel functions. In female age prediction, SVM can handle complex nonlinear relationships, thereby improving prediction accuracy. SVM is highly robust to outliers, which allows the model to maintain stable performance in noisy data sets. SVM can still perform well when the number of samples is relatively small, which may be particularly important for some female age prediction data sets.

[0046] Statistical models may include, but are not limited to:

[0047] Linear regression: A linear model that uses the method of least squares to predict an outcome.

[0048] Logistic regression: A linear model that uses the method of least squares to predict binary outcomes.

[0049] The module can be implemented using various programming languages, such as Python, R, MATLAB, etc. Prediction algorithms can be trained and validated using public datasets, such as NCBI's Gene Expression Omnibus (GEO) database.

[0050] The system can be used to predict the age of female individuals, especially middle-aged and elderly women. The system can be used in the fields of healthcare, insurance, sports, etc. For example, healthcare providers can use the system to predict the age of patients and then develop personalized treatment plans. Insurance companies can use the system to predict the age of policyholders and then determine insurance premiums. Sports coaches can use the system to predict the age of athletes and then develop personalized training plans.

[0051] In clinical applications, the cytokine aging clock of the present invention has important academic and application value:

[0052] First, personalized aging assessment: The clock can provide patients with personalized aging assessments by measuring specific cytokines to assess biological age. This can help doctors accurately understand the aging rate of individuals and provide more accurate health management advice, especially in preventing aging-related diseases such as cardiovascular disease and diabetes.

[0053] Second, early disease screening and intervention: Changes in cytokines are often closely related to pathological processes such as inflammation and decreased immune function. The aging clock can be used as a screening tool for early diseases, helping doctors to detect potential health problems early and conduct early intervention based on biological age changes to delay the aging process and the occurrence of diseases.

[0054] Third, dynamic health monitoring and efficacy evaluation: In clinical practice, the aging clock can be used for long-term health monitoring, by dynamically tracking changes in cytokines and observing the effects of interventions (such as drug treatment or lifestyle changes) on individual biological age. This can provide an objective biological basis for efficacy evaluation and help adjust treatment plans.

[0055] Finally, aging-related disease risk assessment: Cytokines are closely related to the risk of a variety of elderly diseases, such as Alzheimer's disease, arteriosclerosis, etc. By regularly evaluating the patient's cytokine aging clock, doctors can assess the risk of these diseases in advance and take preventive measures.

[0056] The present invention has industrial applicability: the present invention has broad application potential in personalized medicine, health management and aging mechanism research, and can significantly improve the ability to accurately monitor and intervene in the aging process, as shown in the following:

[0057] ① Accurate screening of aging markers: By screening 7 key cytokines from the female population as aging markers, the clock can accurately assess the biological age of individuals. Compared with other more generalized biomarker screenings, this clock model focuses on cytokine changes in women, enhancing the specificity and accuracy of the model.

[0058] ② Multidimensional aging monitoring tool: By combining multiple cytokines, this model can not only predict the aging process in a single dimension, but also integrate the changes in multiple biological pathways to provide multidimensional aging progress information. This makes it a powerful tool for comprehensively monitoring the aging status of individuals.

[0059] ③Personalized health management application: The cytokine aging clock can be used as an important tool for personalized health management, especially in predicting the risk of aging-related diseases and formulating health intervention measures. It can provide clinicians with a more scientific basis for individual aging assessment and promote early intervention and optimization of health management decisions.

[0060] ④ Promotion of basic research: The application of this clock model is not only valuable in health management, but also can further promote basic research in aging biology. By long-term tracking and studying the changing patterns of cytokines, we can deepen our understanding of the aging mechanism and promote the development of aging intervention therapies.

[0061] In clinical applications, the cytokine aging clock has the following key values:

[0062] ① Individualized aging assessment: The clock measures specific cytokines to assess biological age and can provide patients with personalized aging assessments. This can help doctors accurately understand the aging rate of individuals and provide more accurate health management advice, especially in preventing aging-related diseases such as cardiovascular disease and diabetes.

[0063] ② Early disease screening and intervention: Changes in cytokines are usually closely related to pathological processes such as inflammation and decreased immune function. The aging clock can be used as a screening tool for early diseases, helping doctors to detect potential health problems early and conduct early intervention based on biological age changes to delay the aging process and the occurrence of diseases.

[0064] ③ Dynamic health monitoring and efficacy evaluation: In clinical practice, the aging clock can be used for long-term health monitoring, by dynamically tracking changes in cytokines and observing the effects of interventions (such as drug treatment or lifestyle changes) on individual biological age. This can provide an objective biological basis for efficacy evaluation and help adjust treatment plans.

[0065] ④Aging-related disease risk assessment: Cytokines are closely related to the risk of various geriatric diseases, such as Alzheimer's disease, arteriosclerosis, etc. By regularly evaluating the patient's cytokine aging clock, doctors can assess the risk of these diseases in advance and take preventive measures.

[0066] The present invention successfully constructed a biological aging clock model: the constructed biological aging clock model can evaluate the biological age of an individual by quantitatively analyzing biomarkers. This method provides a new perspective for traditional age assessment and can more accurately reflect the true health status of an individual. This model can not only be used to assess current health conditions, but also provide a scientific basis for anti-aging interventions. By monitoring changes in biomarkers, researchers can evaluate the effectiveness of different interventions and thus optimize anti-aging strategies.

[0067] This invention plays an important role in the growing body of knowledge on biomarkers of aging. By systematically identifying and analyzing specific markers associated with female aging, it provides a new tool for future age-related research and clinical evaluation. This will not only help deepen the understanding of the mechanisms of aging, but also promote the development of related fields and provide support for public health strategies. BRIEF DESCRIPTION OF THE DRAWINGS

[0068] Figure 1 Diagram of the serial dilution of cytokine standards

[0069] Figure 2A It shows that among the 1000 cytokines in women, 125 are linearly positively correlated with age, and 34 are negatively correlated.

[0070] Figure 2B Shown: Bootstrap-Lasso screening of aging markers in women

[0071] Figure 2C Shown: Cross-validation curve of Lasso regression of 10 cytokines in women

[0072] Figure 2D Shown: Lasso female 10 cytokine regression coefficient path diagram

[0073] Figure 2E Shown: The 10-factor aging clock for women

[0074] Figure 3A Shown: Validation of 10 aging markers in a large female population

[0075] Figure 3B Shown: Cytokines with consistent expression trends in both cohorts

[0076] Figure 3C Shown: The newly established aging clock

[0077] Figure 3D Shown: New aging clock validated in large population

[0078] Figure 3E Shown: Lasso coefficient of each cytokine of the new aging clock

[0079] Figure 4A Shown: Predicted age for each age group for women

[0080] Figure 4B Shown: ROC curve of the self-reported disease discrimination model based on female Δ age

[0081] Figure 4C The figure shows the correlation analysis between female cytokines and laboratory test indicators. The abbreviations are PLT, platelet; RBC, red blood cell; D3-H, D3-hydroxybutyrate; GLU, blood glucose; HbA1c, glycosylated hemoglobin; IGF, insulin-like growth factor; INS, insulin; FOL, folic acid; HCY, homocysteine; hsCRP, high-sensitivity C-reactive protein; IGM, immunoglobulin M; SOD, superoxide dismutase; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol alcohol; NFFA, free fatty acids; TC, total cholesterol; TG, triglycerides; ALB, albumin; ALP, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; CHE, cholinesterase; GGT, gamma-glutamyltransferase; LDH, lactate dehydrogenase; PAB, prealbumin; TBiL, total bilirubin; CRE, creatinine; Cystatin-C, cystatin C; GFR, glomerular filtration rate; UREA, urea nitrogen.

[0082] Figure 4D Shown is the correlation analysis between female cytokines and laboratory test indicators above the reference interval, where the abbreviations are GLU, blood glucose; Cystatin-C, cystatin C.

[0083] Figure 4E Shown: Correlation analysis between cytokines and laboratory test indicators below the reference interval in women, where the abbreviations are NFFA, free fatty acids; GFR, glomerular filtration rate.

[0084] Figure 4F Shown: Correlation analysis between laboratory test indicators and Δage.

[0085] Figure 4G Shown: Correlation analysis between intrinsic function and physical function and Δage.

[0086] The specific implementation modes of the present invention are further described in detail below in conjunction with the accompanying drawings. DETAILED DESCRIPTION

[0087] Example 1: Screening 10 aging markers from 1000 cytokines in women

[0088] 1. Sample Collection

[0089] The age distribution of female subjects was 26-89 years old, and they were divided into four age groups: young group, young-middle-aged group, middle-aged group and elderly group, with average ages of 30.5±2.26, 44.6±1.74, 59.6±2.22 and 80.93±3.28 years, respectively, with 15 people in each group.

[0090] The sample collection time was from December 2019 to January 2020. The samples were collected at the Beijing Hospital Physical Examination Center, No. 1 Dahua Road, Dongdan, Beijing. Fasting blood samples were collected from the subjects at 8 am every day. Anticoagulant tubes were used to collect venous blood from the subjects' elbows. The operators were professionally trained nurses. After the blood samples were collected, they were centrifuged at 4°C, 3000×g for 10 minutes, and the supernatant was collected into a cryopreservation tube and stored in a -80°C refrigerator. Collector: Zeng Lutao, contact number: 010-58115080.

[0091] 2. Cytokine Detection

[0092] Cytokine aging markers were detected using a cytokine antibody chip, which was a chip kit for detecting 1,000 cytokines (QAH-CAA-X00) purchased from Ruiboao (Guangzhou) Biotechnology Co., Ltd.

[0093] Information on the 1000 cytokines used in this example is shown in Table 1.

[0094]

[0095]

[0096]

[0097]

[0098]

[0099]

[0100]

[0101]

[0102]

[0103]

[0104]

[0105]

[0106]

[0107]

[0108]

[0109]

[0110]

[0111]

[0112]

[0113]

[0114]

[0115]

[0116]

[0117]

[0118]

[0119]

[0120]

[0121]

[0122]

[0123]

[0124]

[0125]

[0126]

[0127]

[0128] 3. Specific testing steps:

[0129] 3.1 Complete drying of the slide

[0130] Take the glass slide out of the box, equilibrate at room temperature for 20-30 minutes, open the packaging bag, remove the sealing strip, and then place the chip in a vacuum dryer or dry it at room temperature for 1-2 hours.

[0131] 3.2 Configuration of standard products

[0132] 3.2.1 Schematic diagram of the gradient dilution of cytokine standards is shown in Figure 1 shown.

[0133] 3.2.2 Add 500 μL of sample diluent to the cytokine standard mixture tube to re-dissolve the standard. Before opening the tube, centrifuge quickly and gently pipette up and down to dissolve the powder. Mark this tube as Std1.

[0134] 3.2.3 Label 6 clean centrifuge tubes as Std2, Std3 to Std7 respectively, and add 200 μL of sample diluent to each tube.

[0135] 3.2.4 Take 100 μL of Std1 and add it to Std2 and mix gently. Then take 100 μL of Std2 and add it to Std3. Continue to dilute in gradient to Std7.

[0136] 3.2.5 Take 100 μL of the sample dilution into another new centrifuge tube and mark it as CNTRL as a negative control.

[0137] Note: Because the starting concentration of each cytokine is different, the concentration series of each cytokine is different after the gradient dilution from Std1 to Std7. Please refer to the kit instructions.

[0138] 3.3 Chip operation process

[0139] 3.3.1 Add 100 μL of sample diluent to each well, incubate on a shaker at room temperature for 1 hour, and block the quantitative antibody chip.

[0140] 3.3.2 Draw out the buffer in each well, add 60 μL of standard solution and sample (sample diluted 2 times, 100 μL sample loaded) to the well, and incubate at 4°C overnight.

[0141] 3.3.3 Wash the slides using the Thermo Scientific Wellwash Versa chip washer in two steps. First, wash with 1× Wash Solution I, 250 μL of 1× Wash Solution I per well, wash 10 times, shake for 10 seconds each time, select high shaking intensity, and dilute 20× Wash Solution I with deionized water. Then switch to 1× Wash Solution II for washing, 250 μL of 1× Wash Solution II per well, wash 6 times, shake for 10 seconds each time, select high shaking intensity, and dilute 20× Wash Solution II with deionized water.

[0142] 3.3.4 Incubation of detection antibody mixture

[0143] Centrifuge the detection antibody mixture tube, then add 1.4 ml of sample diluent, mix well and centrifuge again. Add 80 μL of detection antibody to each well and incubate on a shaker at room temperature for 2 hours.

[0144] 3.3.5 Cleaning, same as step 3.3.3

[0145] 3.3.6 Incubation of Cy3-streptavidin

[0146] Centrifuge the Cy3-streptavidin tube, then add 1.4 ml of sample diluent, mix well and centrifuge quickly again. Add 80 μL of Cy3-streptavidin to each well, cover the slide with aluminum foil and incubate in the dark, and incubate on a shaker at room temperature for 1 hour.

[0147] 3.3.7 Cleaning, same as step 3.3.3

[0148] 3.3.8 Fluorescence detection

[0149] Scan the signal using a laser scanner such as InnoScan 300, using Cy3 or green channel (excitation frequency = 532nm)

[0150] 1) Instrument model: InnoScan 300 Microarray Scanner

[0151] 2) Manufacturer: Innopsys

[0152] 3) Origin: Parc d'Activités Activestre; 31390 Carbonne–France

[0153] 4) Scanning parameters: Wavelength: 532nm; Resolution: 10μm

[0154] 3.4 Use data analysis software to perform data analysis.

[0155] The GAL file dedicated to the array and the microarray analysis software GenePix were used for data extraction, and the data analysis software Q-Analyzer of QAH-CAA-X00 was used for data analysis.

[0156] The linear correlation of each factor on age was analyzed using linear regression using the R lm() function according to the following formula:

[0157] Cytokine level~α+β1 age+β2sex+ε

[0158] Gender was used as a covariate, and a Q value < 0.05 after Bonferroni correction was considered to be linearly correlated.

[0159] To determine the relative proportion of variation explained by age and sex, we calculated PartialEta2 using the following formula:

[0160]

[0161] The results are as follows Figure 2A As shown, among the 1000 cytokines in women, 125 were linearly positively correlated with age, and 34 were negatively correlated.

[0162] The cytokines with linear correlation screened out by linear regression were used to perform 1000 Boostrap self-sampling using the R sample() function. Each time, 1200 samples were randomly selected from the original data with replacement, and then the 1200 samples were subjected to Lasso regression for variable screening (α=1) using the R glmnet package. The cv.glmnet function was used for 10× cross validation to find the minimum lambda value (λ). After obtaining the regression coefficients of all cytokines, the cytokines with a coefficient of 0 were removed, and the remaining cytokines were aging markers. The regularization coefficient λ was adjusted, and finally 10 cytokines were selected from the cytokines of women as potential aging markers. That is, 6Ckine, CA125, CD14, CD99-L2, FSH, GDF-15, LDL R, NAP-2, PON1 and VEGF-C.

[0163] Table 2: Composition of female aging markers

[0164]

[0165]

[0166] 4. Establishing an Aging Clock

[0167] Figure 2BThe figure shows the initial screening of aging markers by Bootstrap-Lasso for women. As shown in the figure, Bootstrap-Lasso regression analysis was performed on the cytokines of women, and 1000 Boostrap self-service samplings were performed. Each sampling was performed with replacement and 1200 samples were randomly selected from the original data, and then Lasso regression was performed on these 1200 samples for variable screening. In the 1000 samplings, 31 cytokines were screened out 1000 times, so these 31 cytokines were used as preliminary aging markers.

[0168] Figure 2C The cross validation curve of Lasso regression is shown, showing the effect of different regularization parameters λ on the mean square error (MSE). As shown in the figure, when the number of features is 10, the mean square error is the smallest.

[0169] Figure 2D This figure shows how the various regression coefficients change as the regularization parameter λ changes. This figure shows how Lasso regression gradually shrinks some unimportant feature coefficients to zero by imposing penalties as λ increases, thereby performing feature selection.

[0170] Using the above 10 aging markers to establish an aging clock, the correlation coefficient between the predicted age and the chronological age is 0.9713, and the root mean square error is 4.4498, which is very accurate. Figure 2E shown.

[0171] We also used other machine learning methods to build models based on these 10 cytokines, including random forest, XGBoost, and support vector machine (SVM). The correlation coefficient between predicted age and chronological age was used to evaluate the prediction effect of the model.

[0172] The random forest model was built using the R language "randomForest" package. Random forest is an ensemble learning method for classification and regression tasks. It works by building multiple decision trees during training and outputting the category mode (classification) or average prediction value (regression) of each decision tree.

[0173] The XGBoost model was built using the R package “xgboost”. XGBoost is a powerful machine learning algorithm for regression, classification, and ranking tasks, known for its high performance in predictive models.

[0174] The R language "e1071" software package was used to construct a linear (kernel = linear) support vector machine (SVM) model. Support vector machine (SVM) is a powerful classifier that divides data into different categories by finding the optimal hyperplane, and is effective even in high-dimensional space.

[0175] Table 3. Predictive effects of different machine learning methods in establishing aging clocks

[0176]

[0177] As shown in the table above, Lasso regression has the best prediction effect.

[0178] Example 2: Validation of the aging clock in a large population

[0179] 1. Sample Collection

[0180] There were 715 female validation populations (hereinafter referred to as the "adult population"), with an age distribution of 25-93 years old and an average age of 55.03±15.94 years old. The samples were collected in two batches. The first batch was collected from December 2019 to January 2020 at No. 1, Dongdan Dahua Road, Dongcheng District, Beijing, and the Beijing Hospital Physical Examination Center (No. 1, Dongdan Dahua Road, Dongcheng District, Beijing). The second batch was collected in October 2021 at the Central Blood Station of Tongzhou District, Beijing (No. 21 Tonghu Street, Tongzhou District, Beijing). Fasting blood samples from subjects were collected at 8 am every day, and anticoagulant tubes were used to collect venous blood from the antecubital fossa of the subjects. The operators were professionally trained nurses. After the blood samples were collected, they were centrifuged at 4°C, 3000×g for 10 minutes, and the supernatant was collected into a cryopreservation tube and stored in a -80°C refrigerator.

[0181] 2. Cytokine Detection

[0182] Cytokine antibody chip detection was performed using the Raybiotech QAH-CUST customized kit (purchased from Raybio (Guangzhou) Biotechnology Co., Ltd.).

[0183] This kit detects the 10 cytokines screened above, namely 6Ckine, CA125, CD14, CD99-L2, FSH, GDF-15, LDL R, NAP-2, PON1 and VEGF-C in women.

[0184] 3. Detection: The detection method is the same as in Example 1.

[0185] Figure 3A To verify the 10 cytokines screened above in a large female population (715 people).

[0186] Comparing the expression trends of the two tests in Example 1 and Example 2, it was found that the seven cytokines showed consistent expression trends with aging, as shown in Figure 3B .

[0187] Therefore, we use the above 7 factors to re-establish the aging clock in the population of Example 1. The correlation coefficient between the predicted age and the chronological age is 0.933, and the mean square error is 5.621. The prediction is still very accurate. Figure 3C For the newly established female cytokine aging clock.

[0188] Substituting the data of the large population into the model calculation, the prediction accuracy can also be maintained at a high level. The correlation coefficient between the predicted age and the time series age is 0.540, and the mean square error is 11.354. Figure 3D . Figure 3E Table 4 shows the Lasso coefficient of each cytokine of the new aging clock.

[0189] Table 4. Lasso coefficients of each cytokine in the new aging clock

[0190]

[0191] The calculation formula for the predicted age based on the seven cytokines is as follows (the cytokines are all detected in pg / mL):

[0192] Predicted age = -97.40798 + GDF-15 × 52.78893 + CD99-L2 × 12.46563 + CD14 × 9.0327 + FSH × 6.32722 + 6Ckine × 0.63787 + PON1 × -10.41995 + CA125 × -10.44494

[0193] Or, predicted age = -97.41 + GDF-15 × 52.79 + CD99-L2 × 12.47 + CD14 × 9.03 + FSH × 6.33 + 6Ckine × 0.64 + PON1 × -10.42 + CA125 × -10.44, where the cytokine values ​​are all detected concentrations pg / mLf

[0194] Example 3: Application of Aging Clock

[0195] 1. Calculate the acceleration of aging

[0196] The aging acceleration Δage is calculated using the following formula:

[0197] ΔAge 初始 = Predicted age – Chronological age

[0198] Δage=Δage 初始 —loess(chronological age—Δage 初始 )

[0199] Loess is a local polynomial regression fitting, which uses the initial aging acceleration Δage minus the fitted age difference to correct the effect of age on aging acceleration. People with Δage greater than the 75% quantile are defined as the accelerated aging population, those with Δage less than the 25% quantile are defined as the decelerated aging population, and those in between are defined as the normal aging population.

[0200] The female adult population was divided into a healthy combination and a self-reported disease group (self-reported history of heart disease, diabetes, stroke and cerebrovascular disease, chronic nephritis, Parkinson's disease, dementia, cancer and other major life-threatening diseases), and their Δ ages were compared.

[0201] Table 5. Proportion of aging status in the healthy group of women with self-reported disease combinations *

[0202]

[0203] *Chi-square test: p<0.001

[0204] 2. ΔAge-based self-reported disease

[0205] Figure 4A The predicted age of women in each age group is shown. As shown in the figure, among women over 40 years old, the median predicted age of the self-reported disease group is higher than that of the healthy group (there are too few people in the disease group among people under 40 years old), considering that the self-reported disease group ages faster than the healthy group.

[0206] The R language "pROC" package was used to make the ROC curve using Δage, and the area under the curve AUC = 0.765, see Figure 4B Logistic regression analysis was performed using the R language glm function, with self-reported disease defined as 1 and no self-reported disease defined as 0 as the dependent variable, and aging acceleration Δage as the independent variable. The odds ratio (OR) was 1.09, and the p value was less than 0.001.

[0207] 3. Correlation between cytokines, Δage and laboratory test indicators

[0208] Cholinesterase (CHE), albumin (ALB), prealbumin (PAB), γ-glutamyl transpeptidase (GGT), alkaline phosphatase (ALP), serum lactate dehydrogenase (LDH), homocysteine ​​(HCY), folate (FOL), GLU, D-3-hydroxybutyrate (D-3H), high-sensitivity C-reactive protein (hsCRP), superoxide dismutase (SOD), serum creatinine (Scr), blood urea nitrogen (UREA), uric acid (URIC), cystatin C (Cystatin C), triglycerides (TG), total cholesterol (TC), high-density lipoprotein cholesterol (HDL-C), low-density lipoprotein cholesterol (LDL-C), free fatty acids (NFFA), alanine aminotransferase (ALT), aspartate aminotransferase (AST), total bilirubin (Tbil), and total bile acid (TBA) were measured using an automatic biochemical analyzer (Hitachi LABOSPECT008AS; Hitachi, Tokyo, Japan). The levels of red blood cells (RBC), white blood cells (WBC), hemoglobin (HGB), and platelets (PLT) were determined by the cyanide hemoglobin method using a fully automated modular blood and body fluid analyzer (Sysmex XN-20 [A1], Sysmex Corporation, Kobe, Japan). HbA1c was measured by high-performance liquid chromatography (Tosoh G8 high-performance liquid chromatography analyzer, Tosoh Corporation, Tokyo, Japan). Immunoglobulin M (IgM) was measured using Mindray BS-2800M (Mindray, Shenzhen, China). Insulin-like growth factor (IGF) and insulin (INS) were detected by chemiluminescence. Plasma samples were obtained after centrifugation at 4°C and 1500 × g and stored at −80°C before testing. The estimated glomerular filtration rate (eGFR) was calculated using the CKD-EPI creatinine equation (2021).

[0209] Figure 4C This is a correlation analysis between cytokines and laboratory test indicators in women.

[0210] Figure 4D This is a correlation analysis between female cytokines and laboratory test indicators above the reference range.

[0211] Figure 4E Correlation analysis between cytokines and laboratory test indicators below the reference interval in women.

[0212] FIG. 4C to FIG. 4EIt is shown that when the correlation analysis of female cytokines and laboratory test indicators is performed, it can be seen that GDF-15 has a significant correlation with most laboratory test indicators, and FSH, CA125 and 6ckine are also correlated with most indicators. Female fasting blood glucose above the reference range is positively correlated with 6ckine, CD14, CD99-L2, LDL R, NAP-2, and PON1, and glomerular filtration rate below the reference range is negatively correlated with FSH and GDF-15, and positively correlated with PON1 and VEGF-C. Figure 4F This is the correlation analysis between laboratory test indicators and Δage.

[0213] Figure 4F The results showed that multiple laboratory test indicators were significantly correlated with Δage, and were consistent with the predicted trends.

[0214] 4. Correlation between intrinsic function, physical function and Δ age

[0215] We constructed an intrinsic function assessment system using five dimensions, namely sensory (hearing), motor (gait speed), vitality (BMI and grip strength), psychological score (DASS-21 scale, self-assessment of life and self-assessment of health) and cognitive score (Montreal Cognitive Assessment (MoCA)).

[0216] Figure 4G This is the correlation analysis between intrinsic function and physical function and Δage. Figure 4G It shows that each dimension of intrinsic function is also strongly correlated with cytokines. The Δ age of women is negatively correlated with intrinsic function, that is, the older the predicted age, the lower the intrinsic function score and the worse the function. Moreover, Δ age is significantly correlated with multiple physical function indicators.

[0217] The above formula is used to calculate the aging acceleration (Δage). Its significant correlation with self-reported diseases, laboratory test indicators and intrinsic function suggests that it can be used as an early warning indicator of aging and health status and can be used to assess aging status. The larger the Δage, the greater the individual's aging acceleration, the greater the probability of suffering from disease, and the lower the intrinsic function score. Specifically:

[0218] (1) Association with self-reported diseases:

[0219] ΔOlder individuals are more likely to report having various chronic diseases, such as cardiovascular disease, diabetes, or arthritis.

[0220] This association suggests that Δage may reflect undetected pathological processes and help to detect potential health risks early.

[0221] (2) Correlation with laboratory test indicators:

[0222] ΔAge was significantly correlated with multiple biochemical indices, such as inflammatory markers (such as C-reactive protein), blood glucose levels, and blood lipid profiles.

[0223] This suggests that Δage may capture physiological changes at a subclinical level, providing opportunities for preclinical intervention.

[0224] (3) Relationship with intrinsic functions:

[0225] ΔOlder individuals tend to have lower scores in cognitive function, physical status, and daily living abilities.

[0226] This association reveals that Δage may indicate functional decline, which could help to take preventive measures early.

[0227] (4) Advantages of Δage as a comprehensive indicator:

[0228] Compared with a single indicator, Δage integrates information from multiple biomarkers and may provide a more comprehensive assessment of health status.

[0229] It provides new tools for personalized health management and precision medicine.

[0230] (5) Predictive value of Δage:

[0231] The older the age, the worse the current health condition may be, and the higher the risk of disease and functional decline may be in the future.

[0232] This makes Δage a valuable predictive tool that can be used to guide preventive interventions.

[0233] (6) Application value of Δage:

[0234] In clinical practice, Δage can be used to screen high-risk populations and guide personalized health management plans.

[0235] In the research field, it can be used as a new endpoint indicator to evaluate the effectiveness of anti-aging interventions.

[0236] As a comprehensive indicator, Δage not only reflects the current state of aging, but may also predict future health trajectories. Its application is expected to promote the development of preventive medicine and promote the paradigm shift from disease treatment to health management. However, further prospective studies are needed to verify its long-term predictive ability and applicability in different populations.

[0237] The series of detailed descriptions listed above are only specific descriptions of feasible implementation methods of the present invention. They are not intended to limit the scope of protection of the present invention. Those skilled in the art can design many other modifications and implementation methods, which will fall within the scope and spirit of the principles disclosed in this application.

Claims

1. A set of female aging markers, characterized in that: It includes multiple cytokines associated with female aging, including but not limited to: growth differentiation factor 15 (GDF-15), leukocyte differentiation antigen 99-L2 (CD99-L2), leukocyte differentiation antigen 14 (CD14), follicle stimulating hormone (FSH), cytokine 6Ckine (6Ckine), parahydroxybenzoate esterase (PON1), and cancer antigen 125 (CA125).

2. A method for predicting the age of a female individual, characterized in that: The method comprises the following steps: a. Obtain biological samples from female subjects; b. determining the content of female aging markers in the biological sample; c. inputting the content value of the female aging marker into a preset cytokine aging clock; d. outputting the estimated age of the female individual based on the cytokine aging clock.

3. A method for predicting the age of a female individual according to claim 2, characterized in that: The female aging markers include but are not limited to: growth differentiation factor 15 (GDF-15), leukocyte differentiation antigen 99-L2 (CD99-L2), leukocyte differentiation antigen 14 (CD14), follicle stimulating hormone (FSH), cytokine 6Ckine (6Ckine), parahydroxybenzoate esterase (PON1), and cancer antigen 125 (CA125).

4. A method for predicting the age of a female individual according to claim 2, characterized in that: The biological sample is blood, serum, plasma or saliva; the determination step adopts immunoassay, mass spectrometry, biochemical analysis method or biochip technology.

5. A method for predicting the age of a female individual according to claim 4, characterized in that: The calculation formula of the cytokine aging clock is: predicted age = -97.41 + GDF-15 × 52.79 + CD99-L2 × 12.47 + CD14 × 9.03 + FSH × 6.33 + 6Ckine × 0.64 + PON1 × -10.42 + CA125 × -10.44, wherein the cytokine values ​​are all detection concentrations pg / mL; or predicted age = -97.40798 + GDF-15 × 52.78893 + CD99-L2 × 12.46563 + CD14 × 9.0327 + FSH × 6.32722 + 6Ckine × 0.63787 + PON1 × -10.41995 + CA125 × -10.44494, wherein the cytokine values ​​are all detection concentrations pg / mL.

6. A method for predicting the age of a female individual according to claim 4, characterized in that: It also includes the calculation of the acceleration of aging, namely Δage.

7. A method for predicting the age of a female individual according to claim 6, characterized in that: The calculation formula of the aging acceleration (Δage) is: ΔAge 初始 = Predicted age – Chronological age Δage=Δage 初始 —loess(chronological age—Δage 初始 ) Wherein, Loess is a local polynomial regression fitting, which uses the initial aging acceleration Δage minus the fitted age difference to correct the effect of age on aging acceleration.

8. A system for predicting the age of a female individual, characterized in that: include: a. A biological sample collection module, used to collect biological samples from female individuals; b. A polypeptide content determination module, used to determine the content of at least 7 female aging markers in the collected biological sample, wherein the 7 female aging markers are growth differentiation factor 15 (GDF-15), leukocyte differentiation antigen 99-L2 (CD99-L2), leukocyte differentiation antigen 14 (CD14), follicle stimulating hormone (FSH), cytokine 6Ckine (6Ckine), parahydroxybenzoate esterase (PON1), and cancer antigen 125 (CA125); c. A cytokine aging clock module, used to input the measured polypeptide content into a prediction algorithm to predict the age of the female individual.

9. A method for predicting the age of a female individual according to claim 8, characterized in that: The prediction algorithm module is based on a machine learning model or a statistical model.

10. A method for predicting the age of a female individual according to claim 9, characterized in that: The machine learning models include but are not limited to Lasso regression, random forest, XGBoost and SVM.

Images