A kit for detecting metabolic markers of depression in children and adolescents, and a preparation method and application thereof

Abstract

The present application belongs to the technical field of in-vitro diagnosis of depression in children and adolescents, and particularly relates to a kit for detecting metabolic markers of depression in children and adolescents, and a preparation method and application thereof. The metabolic markers detected by the kit are composed of cyclic adenosine monophosphate, glucuronic acid, undecenoyl carnitine and undecanoyl carnitine, and the kit comprises a mixed standard solution containing the markers, a quality control solution and three stable isotope internal standards. The data analysis module comprises a binary logistic regression model P=EXP(X) / (1+EXP(X)), and the model comprises an age grouping covariate, with 14 years as the dividing point (1 for ≤14 years and 2 for >14 years), and can automatically output a depression risk probability value P. Comparative experiments show that the four-index joint model can reach an AUC of 0.921 and a Youden index of 0.711, with a sensitivity of 84.4% and a specificity of 86.7% after correcting the age covariate, which is significantly better than the three-index model without the age covariate (AUC=0.813, Youden index=0.494), with an improvement of 13.3% and 43.9%, respectively. The present application realizes rapid, accurate and high-throughput detection of four metabolic markers in plasma samples of children and adolescents, and provides an objective and quantitative basis for the auxiliary diagnosis of depression in children and adolescents.

Description

Technical Field

[0001] This invention belongs to the field of in vitro diagnostic technology for childhood and adolescent depression, specifically relating to a reagent kit for detecting metabolic biomarkers of childhood and adolescent depression, its preparation method, and its application. Background Technology

[0002] Depression is one of the most common mental disorders in children and adolescents. The "World Mental Health Report 2025" shows that more than 350 million people worldwide suffer from depression, and the affected population is trending younger. Incomplete statistics estimate that by 2025, the number of people with depression in China has exceeded 100 million. Of particular concern is the continuously rising prevalence of depression among children and adolescents; however, the treatment rate in this group is less than 20%, meaning many patients are already in the moderate to severe stage at the time of diagnosis, missing the optimal window for early intervention. Depression not only severely impacts patients' academic performance, social functioning, and mental and physical development, but also places a heavy burden on families, causing significant losses to social medical resources and the workforce, making it a major public health problem that urgently needs to be addressed.

[0003] Currently, clinical diagnosis primarily relies on the symptomatological criteria in DSM-5 or ICD-11, depending on clinician interviews and patient subjective reports, lacking objective biomarkers to aid diagnosis. This subjective diagnostic model, coupled with children and adolescents' limited ability to express themselves and their lack of awareness of the need for medical attention, leads to a large number of potential patients going undetected, further exacerbating the problem of low consultation rates and high missed diagnosis rates for depression. Therefore, developing an objective, simple, and scalable screening tool for depression suitable for children and adolescents has significant clinical and social value.

[0004] In recent years, researchers have begun to explore biomarkers for depression, mainly including neuroactive substances (such as tryptophan, serotonin, and dopamine), hypothalamic-pituitary-adrenal axis-related indicators (such as cortisol), inflammatory factors (such as IL-6 and TNF-α), amino acids and their derivatives (such as cyclic adenosine monophosphate, proline, betaine, and alanine), and energy metabolomics markers (such as succinic acid and hydroxybutyric acid). Existing detection methods mainly include enzyme-linked immunosorbent assay (ELISA) and high-performance liquid chromatography (HPLC). ELISA can only detect a single indicator at a time and carries the risk of cross-reactivity, while HPLC has limited sensitivity and cannot meet the accurate quantification requirements of low-abundance metabolites.

[0005] However, existing patents for metabolic biomarkers related to the diagnosis of depression still have the following technical shortcomings:

[0006] First, the number of detection indicators is generally large, ranging from 10 to 31 (such as patents CN120427887A, CN120446459A and CN119804732A), resulting in high reagent kit costs, long detection time, and complex data analysis. Furthermore, the multicollinearity problem among multiple indicators affects the stability of the diagnostic model.

[0007] Secondly, most of these technologies are designed for adults (such as patents CN119224186A, CN120158531A, CN119804732A, CN120009439A and CN120992968A), and their reference ranges and diagnostic models are not applicable to children and adolescents whose metabolic characteristics differ significantly.

[0008] Third, most patents only provide metabolite concentration detection values ​​and do not establish a multi-indicator joint diagnostic model. Clinicians still need to interpret the data themselves and lack intuitive auxiliary diagnostic tools (such as patents CN119224186A, CN120118990A, CN120992968A).

[0009] Fourth, the internal reference system is simple in design and often uses a single mixed internal standard, making it difficult to achieve accurate calibration for analytes with different physicochemical properties (such as patent CN119224186A).

[0010] Fifth, existing technologies mostly focus on protein biomarkers and gut microbiota biomarkers (such as patents CN117723759B, CN120174123A, CN120118990A, CN120608140A, CN120652107A, etc.), while research on small molecule metabolite biomarkers is still lacking. Summary of the Invention

[0011] To address the technical problem of the lack of simple, efficient, and accurate quantitative kits specifically for detecting metabolic biomarkers of depression in children and adolescents, the inventors of this application have identified three metabolic biomarkers—cyclic adenosine monophosphate (cAMP), glucuronic acid, and undecanoylcarnitine—in previous studies (see authorized patent CN119252459A). In subsequent studies, the inventors further discovered and identified undecanoylcarnitine (C11) and confirmed its incremental diagnostic value independent of C11:1 (see Example 1 for details). Simultaneously, based on the results of previous age-group difference analysis (see Example 6 for details), age was incorporated as a covariate into the diagnostic model. Therefore, this invention provides a kit for detecting four metabolic biomarkers—cAMP, GluA, C11:1, and C11—as well as its preparation method and applications.

[0012] It is important to note that while some studies have reported the diagnostic potential of undecanoylcarnitine (C11) in newly diagnosed, medication-naïve adults with depression (Liu et al., Behavioural Brain Research, 2026, 497:115864), these studies showed that the AUC of C11 combined with other acylcarnitines in the adult population was 0.831. However, there are fundamental differences in metabolic characteristics between children and adolescents and adults. Children and adolescents are in a critical period of growth and development, and their metabolic levels change dynamically with age. Their energy metabolism, fatty acid β-oxidation flux, and the homeostatic regulation mechanisms of the carnitine metabolic pool are all significantly different from those in adults. Therefore, the biomarker reference ranges and diagnostic models established in adult studies cannot be directly applied to children and adolescents. The inventors of this application have found through previous research that the concentrations of metabolites such as cAMP, GluA, and C11:1 differ significantly among different age groups (P<0.001), and that concentrations rapidly increase after age 14. Therefore, when constructing diagnostic models, age must be included as an important covariate to eliminate the confounding effects of different developmental stages on metabolite concentrations, thereby establishing a truly accurate diagnostic tool applicable to children and adolescents. The findings of this invention in children and adolescents have entirely different technical implications and clinical significance:

[0013] (1) Population differences: The diagnostic value of C11 in adult studies is based on adult metabolic homeostasis, while children and adolescents are in a stage of rapid metabolic development, and their baseline level, fluctuation range and association strength with depression are significantly different from those of adults (see the age group difference analysis in Example 6 of this invention for details).

[0014] (2) Differences in model construction strategies: Adult studies did not consider age as a key covariate, while this invention for the first time included age group (≤14 years vs >14 years) as an independent covariate in the logistic regression model, effectively eliminating the confounding effects of different developmental stages on the concentration of metabolites such as C11, so that the model has stable diagnostic efficacy in different age subgroups of children and adolescents.

[0015] (3) Difference in diagnostic efficacy: The AUC of the C11 joint model in the adult study was 0.831, while the AUC of the four-indicator joint model (including age covariate) of this invention in the children and adolescents population reached 0.921, an improvement of 10.8%, which fully demonstrates that the optimized model for children and adolescents population has significant diagnostic advantages.

[0016] The technical problem solved by this invention is achieved by the following technical solution:

[0017] In a first aspect, the present invention provides a metabolic biomarker detection kit for depression in children and adolescents, wherein the metabolic biomarkers detected by the kit are composed of cyclic adenosine monophosphate (cAMP), glucuronic acid (GluA), undecanoylcarnitine (C11) and undecenoylcarnitine (C11:1).

[0018] The kit also includes:

[0019] (1) Mixed standard solutions: including a series of concentration gradients of cyclic adenosine monophosphate (cAMP), glucuronic acid (GluA), undecanoylcarnitine (C11) and undecenoylcarnitine (C11:1).

[0020] Furthermore, in the mixed standard solution, the concentration range of cyclic adenosine monophosphate is 0~78.125 ng / mL, the concentration range of glucuronic acid is 0~195.3125 ng / mL, the concentration range of undecanoylcarnitine is 0~78.125 ng / mL, and the concentration range of undecenoylcarnitine is 0~3.90625 ng / mL.

[0021] (2) Quality control solution: including low, medium and high concentration levels of cyclic adenosine monophosphate (cAMP), glucuronic acid (GluA), undecanoylcarnitine (C11) and undecenoylcarnitine (C11:1).

[0022] Further, the quality control solution contains the following concentrations: low-concentration quality control contains 5.0 ng / mL cyclic adenosine monophosphate (cAMP), 12.0 ng / mL glucuronic acid, 0.25 ng / mL undecanoylcarnitine, and 5.0 ng / mL undecanoylcarnitine. Medium-concentration quality control contains 20 ng / mL cAMP, 50.0 ng / mL glucuronic acid, 1.0 ng / mL undecanoylcarnitine, and 20 ng / mL undecanoylcarnitine. High-concentration quality control contains 60 ng / mL cAMP, 150 ng / mL glucuronic acid, 3.0 ng / mL undecanoylcarnitine, and 60 ng / mL undecanoylcarnitine.

[0023] (3) Extraction solution: a methanol solution containing stable isotope internal standards, wherein the stable isotope internal standards are composed of cyclic adenosine monophosphate stable isotope internal standards (IS_cAMP), glucuronic acid stable isotope internal standards (IS_GluA) and lauroylcarnitine stable isotope internal standards (IS_C12);

[0024] Further, in the extract, the concentration of IS_cAMP is 0.5~5ug / mL, the concentration of IS_GluA is 10~50ug / mL, and the concentration of IS_Cl2 is 5~20ug / mL; preferably, the concentration of IS_cAMP is 2.5ug / mL, the concentration of IS_GluA is 20ug / mL, and the concentration of IS_Cl2 is 10ug / mL.

[0025] Secondly, the present invention provides a method for preparing a metabolic biomarker detection kit for depression in children and adolescents, comprising the following steps: the kit is prepared according to the following method:

[0026] (1) Prepare a series of mixed standard solutions with different concentration gradients: Dissolve and dilute the cyclic adenosine monophosphate, glucuronic acid, undecyl carnitine and undecyl carnitine standards in methanol, add internal standards IS_cAMP, IS_GluA and IS_Cl2, and prepare mixed standard solutions with 6 concentration levels.

[0027] (2) Prepare mixed quality control solutions with low, medium and high concentration levels: Using blank plasma as the matrix, add cyclic adenosine monophosphate, glucuronic acid, undecyl carnitine and undecyl carnitine standards and internal standards IS_cAMP, IS_GluA and IS_C12 to prepare quality control solutions;

[0028] (3) Preparation of extract: Dissolve internal standards IS_cAMP, IS_GluA and IS_Cl2 in methanol to prepare an extract containing internal standards. The concentration of IS_cAMP in the extract is 0.5-5 μg / mL, the concentration of IS_GluA is 10-50 μg / mL and the concentration of IS_Cl2 is 5-20 μg / mL.

[0029] (4) Assemble the above standard solution, quality control solution and extraction solution into a kit.

[0030] Thirdly, the present invention provides a detection system for metabolic biomarkers of depression in children and adolescents, comprising a sample pretreatment unit, a data acquisition module, and a data analysis module. The data acquisition module includes the aforementioned reagent kit and a liquid chromatography-tandem mass spectrometry (LC-MS / MS) unit. The sample pretreatment unit is used to perform the sample pretreatment steps to obtain the test solution. The LC-MS / MS unit is used to perform the LC-MS / MS detection steps to obtain detection data, which is then analyzed by the data analysis module.

[0031] Further sample pretreatment methods include: after collecting whole blood using standard venous blood collection methods, gently invert the anticoagulant tube to mix, avoiding hemolysis, and perform plasma separation as soon as possible: symmetrically place the anticoagulant whole blood into a centrifuge, balance it, and centrifuge at 1500~2000×g for 10~15 minutes (temperature 20~25℃). After centrifugation, carefully aspirate the supernatant plasma, avoiding contact with the white membrane layer and red blood cell layer, and transfer it to a sterile capped sample tube. If the separated plasma can be tested immediately, store it at room temperature; if it cannot be tested immediately, it should be frozen at -20℃ or -80℃, avoiding repeated freeze-thaw cycles.

[0032] Sample pretreatment unit: Frozen plasma needs to be thawed at 4℃ in advance and thoroughly mixed before use; add 25~100µL of plasma sample to a 1.5mL centrifuge tube, then add 100~400µL of extraction buffer (methanol solution containing internal standard), shake vigorously for 0.5~2 minutes to mix thoroughly, place at -10~-30℃ for 0.5~2 hours for extraction, and shake vigorously once every 25~35 minutes during the process; after extraction, centrifuge at 15000 rpm~17000 rpm for 15~20 minutes at 4℃, carefully aspirate 100~120 µL of supernatant into a sample vial with an inner tube, avoiding aspirating the precipitate, cap the vial, and immediately perform instrumental analysis or temporarily store at 4℃.

[0033] Furthermore, the liquid chromatography-tandem mass spectrometry unit includes mobile phase A, which is an aqueous solution containing 0.1–0.5% ammonia and 0.1–0.5% 3–7M ammonium acetate, and mobile phase B, which is acetonitrile-water; gradient elution program: 0–1.0 min, phase B linearly increases from 0% to 100%; 1.0–4.5 min, phase B linearly decreases from 100% to 50%; 4.5–6.5 min, maintains 50% B; 6.5–6.7 min, phase B linearly increases from 50% to 100%; 6.7–11.5 min, maintains 100% B; column temperature 20–40 °C, flow rate 0.1–0.5 mL / min, injection volume 0.5–1.5 µL.

[0034] Furthermore, the data analysis module includes a binary logistic regression model, whose expression is P=EXP(X) / (1+EXP(X)); where X=C+a∗C cAMP +b∗C GluA +c∗C C11:1 +d∗C C11 +e*AgeGroup;C cAMP、 C GluA、 C C11:1、 C C11These represent the concentrations of cyclic adenosine monophosphate (cAMP), glucuronic acid, undecenoylcarnitine, and undecanoylcarnitine collected by the data acquisition module; AgeGroup represents age grouping, with 14 years as the dividing point, where e is 1 for ages ≤ 14 and 2 for ages > 14; C is a constant term, and a, b, c, and d are the coefficients of cAMP, glucuronic acid, undecenoylcarnitine, and undecanoylcarnitine, respectively. The constant term C and the coefficients a, b, c, and d are obtained by building a model using the training set.

[0035] Compared with the prior art, the beneficial technical effects of the present invention are as follows:

[0036] 1. This invention provides a kit specifically for detecting metabolic biomarkers of depression in children and adolescents. The metabolic biomarkers detected by the kit consist of cyclic adenosine monophosphate, glucuronic acid, undecyl carnitine and undecyl carnitine. The number of biomarkers is reduced to four, which significantly reduces the detection cost and analysis complexity, and is suitable for large-scale screening scenarios in children and adolescents.

[0037] 2. The kit of the present invention uses three independent stable isotope internal standards (IS_cAMP, IS_GluA and IS_C12) to calibrate each analyte. Experimental verification shows that the spiked recovery rate is 95.0% to 100.5% and RSD ≤ 7.6%, which realizes accurate quantification of each metabolite and is significantly better than the single mixed internal standard calibration strategy in the prior art.

[0038] 3. This invention provides a one-step extraction pretreatment method specifically for this kit. Compared with traditional methods (such as CN119224186A), which often involve multi-step derivatization, solid-phase extraction, or separate preparation of internal standards followed by stepwise addition (requiring 8-10 steps), this method directly pre-dissolves three stable isotope internal standards in the methanol extract. Only three core steps are required: adding plasma, shaking, low-temperature precipitation, and centrifugation. No additional derivatization or reconstitution steps are needed, significantly simplifying the operation, shortening the time (approximately 1.5 hours in total), and reducing human error. Experiments have shown (see Example 4) that this method, combined with this kit, achieves a spike recovery rate of 95.0%–100.5% and an RSD ≤ 7.6%, greatly increasing detection throughput while ensuring excellent accuracy and precision.

[0039] 4. This invention is equipped with LC-MS / MS technology, employing a BEH Amide HILIC column and a specific mobile phase system with a high acetonitrile gradient (mobile phase A is an aqueous solution containing 0.1%–0.5% ammonia and 0.1%–0.5% 3–7M ammonium acetate). The constant addition of ammonia and ammonium acetate stabilizes the pH environment, improving peak shape and ionization efficiency. Combined with the positive and negative ion switching MRM mode, baseline separation of four target analytes and their internal standards can be achieved in a single injection, combining high selectivity and high throughput.

[0040] 5. This invention substitutes the detection concentrations of four biomarkers into a pre-constructed binary logistic regression model P=EXP(X) / (1+EXP(X)), which can automatically output the probability value P of depression risk, providing objective and quantitative auxiliary diagnostic evidence for clinical practice.

[0041] 6. Through comparative experiments, this invention demonstrates that by adding C11 as a fourth indicator to the three indicators of cAMP, GluA, and C11:1, the AUC of the combined diagnostic model significantly increased from 0.813 to 0.897 (ΔAUC=+0.084), and the Youden index significantly increased from 0.494 to 0.653 (Δ=+0.159). Furthermore, by incorporating age grouping (≤14 years vs >14 years) as a covariate into the model, the AUC of the four-indicator model further increased from 0.897 to 0.921 (Δ=+0.024), and the Youden index increased from 0.653 to 0.711 (Δ=+0.058). In summary, the four-indicator joint model (including the age covariate) of this invention improved the AUC by 0.108 (0.813→0.921) and the Youden index by 0.217 (0.494→0.711) compared with the three-indicator baseline model, representing improvements of 13.3% and 43.9%, respectively. The inclusion of the age covariate effectively eliminated the confounding effects of different developmental stages on metabolite concentrations, making the model more widely applicable in different age subgroups.

[0042] 7. Compared with existing studies on metabolic biomarkers for adult depression, this invention is the first to achieve precise detection and application of C11 in childhood and adolescent depression. Through large-sample (n=934) clinical validation, this invention confirms that the diagnostic contribution of C11 in children and adolescents (Wald χ²=18.762, P<0.001) and its interaction effect with age group are statistically significant, indicating that the diagnostic value of C11 in children and adolescents cannot be simply applied directly to children and adolescents based on adult research results. This invention fills a technological gap in the field of metabolic biomarker detection for childhood and adolescent depression.

[0043] 8. In summary, this invention constructs a complete end-to-end solution from reagent kit composition, internal standard calibration strategy, sample pretreatment, LC-MS / MS detection to data analysis, realizing rapid, accurate, and high-throughput detection of metabolic biomarkers for depression in children and adolescents, and is particularly suitable for large-scale clinical screening scenarios in children and adolescents.

[0044] The above description is merely an overview of the technical solution of the present invention. In order to better understand the technical means of the present invention, it can be implemented according to the contents of the specification. Furthermore, in order to make the above contents, objectives, features and advantages of the present invention more obvious and understandable, specific embodiments of the present invention are described below. Attached Figure Description

[0045] Figure 1-2 This is the LC-MS / MS chromatogram of the internal standard IS_cAMP in this invention.

[0046] Figure 3 This is the LC-MS / MS chromatogram of cAMP, the biomarker used in this invention.

[0047] Figure 4-5 This is the LC-MS / MS chromatogram of the internal standard IS_GluA in this invention.

[0048] Figure 6 This is the LC-MS / MS chromatogram of GluA, the marker in this invention.

[0049] Figure 7 This is the LC-MS / MS chromatogram of the internal standard IS_C12 in this invention.

[0050] Figure 8 This is the LC-MS / MS chromatogram of C11, the marker of this invention.

[0051] Figure 9 The LC-MS / MS chromatogram of the biomarker C11:1 of this invention is shown.

[0052] Figure 10 This is a standard curve diagram of cAMP, the marker of this invention.

[0053] Figure 11 This is a standard curve diagram of the C11:1 marker of this invention.

[0054] Figure 12 This is a standard curve diagram of the marker C11 of this invention.

[0055] Figure 13 This is a standard curve diagram of GluA, the marker of this invention.

[0056] Figure 14 This is a scatter plot showing the correlation between the concentrations of C11 and C11:1 in this invention.

[0057] Figure 15 This is a graph showing the ROC curve analysis of the binary logistic regression model of this invention on the healthy children and adolescents and the depression group. Detailed Implementation

[0058] The technical solution of the present invention will be further described in detail below with reference to specific embodiments. It should be understood that the following embodiments are merely illustrative and explanatory of the present invention and should not be construed as limiting the scope of protection of the present invention. All technologies implemented based on the above content of the present invention are covered within the scope of protection intended by the present invention.

[0059] In addition, unless otherwise specified, all raw materials, reagents, instruments and equipment used in this invention can be obtained by purchasing them from the market or prepared by existing methods.

[0060] Example 1: Discovery and Identification of Undecanoylcarnitine (C11)

[0061] In previous studies, the inventors of this application identified cyclic adenosine monophosphate (cAMP), glucuronic acid (GluA), and undecenoylcarnitine (C11:1) as a candidate combination for metabolic biomarkers of depression in children and adolescents (see authorized patent CN119252459A). During the clinical sample testing of these three indicators using LC-MS / MS, the inventors observed an unknown chromatographic peak near the C11:1 peak (retention time approximately 3.01 min). This unknown peak's retention time (approximately 2.88 min) was highly similar to, but not completely identical to, that of C11:1. Furthermore, the peak area of ​​this unknown peak showed a asynchronous trend compared to the C11:1 peak area in different samples, suggesting that this unknown peak might be another acylcarnitine substance independent of C11:1.

[0062] To further identify the substance of the unknown peak, the inventors compared its characteristic mass spectrometry ions (mother ion m / z 330.3 → daughter ion m / z 271.0) with a standard spectral library, and simultaneously confirmed its identity using undecanoylcarnitine (C11) standard for retention time and ion pair analysis. The results showed that the chromatographic retention time and ion pair spectrum of the unknown peak were completely consistent with those of the C11 standard, confirming that the unknown peak was undecanoylcarnitine (C11).

[0063] Based on this, the inventors conducted a correlation analysis on the detection values ​​of C11 and C11:1 in 934 clinical samples. Using Spearman rank correlation analysis, the results showed that C11 and C11:1 were moderately positively correlated (r=0.568, P<0.001), with a coefficient of determination r²=0.323. This indicates that approximately 32.3% of the information shared between the two, while approximately 67.7% of the variation was unique to each. In other words, C11 is not a simple companion or repetitive information of C11:1, but rather a novel biomarker with independent biological information.

[0064] Based on the above findings, the inventors incorporated C11 as the fourth detection indicator into the kit, forming a joint detection scheme for four indicators: cAMP, GluA, C11:1, and C11.

[0065] Example 2: Preparation of the reagent kit

[0066] 2.1 Preparation of standard solutions

[0067] First, weigh out 1.0 mg or less of cAMP, GluA, C11, C11:1 standards and IS_GluA, IS_cAMP and IS_C12 standards respectively, dissolve them in 1.0 mL of methanol, and prepare stock solutions with a concentration of 1.0 mg / mL.

[0068] Then, according to the preset concentration in Table 1, the above stock solution was diluted with methanol to prepare a mixed standard solution with 6 concentration levels, which was stored at -20°C for later use.

[0069] Table 1 Mixed standards at different concentration levels

[0070]

[0071] A mixed standard at seven concentration levels was used to construct a standard curve for absolute quantification of cAMP, GluA, C11, and C11:1 in the samples.

[0072] 2.2 Preparation of quality control solution

[0073] According to the preset concentration gradient in Table 2 below, the above stock solution was diluted and mixed using blank plasma matrix to prepare 3 quality control solutions, which were stored at -20℃ for later use.

[0074] Table 2 Quality Control Products

[0075]

[0076] Quality control materials are available in three concentration levels: low, medium, and high, representing the low, medium, and high concentration levels within the clinical sample testing range, respectively. These are used for intra-batch and inter-batch quality control. The measured values ​​of the quality control materials should be within ±15% of the labeled values; otherwise, the batch results are considered invalid.

[0077] 2.3 Preparation of Extraction Solution

[0078] Prepare 200µL of each sample. Add 0.625µL of IS-cAMP (1.0mg / mL), 5µL of IS-GluA (1mg / mL), and 2.5µL of IS_C12 (1.0mg / mL) to the sample. Add methanol to bring the total volume to 200µL. Shake to mix well and freeze at -20℃ for at least 30 minutes. Store for later use.

[0079] In the extract, the final concentration of IS_cAMP was 3.125 µg / mL, the final concentration of IS_GluA was 25 µg / mL, and the final concentration of IS_Cl2 was 12.5 µg / mL.

[0080] 2.4 Reagent Kit Assembly and Packaging

[0081] This kit is available in 96, 480, or 960 test kits / boxes. Each component is prepared and dispensed according to the specified ratio, and mainly includes:

[0082] (1) Standards: 6 bottles, each containing a mixed standard solution at 6 concentration levels, each bottle containing 0.5 mL, sealed in a transparent brown borosilicate glass bottle with a polypropylene (PP) cap (with an inner PTFE gasket).

[0083] (2) Quality control products: 3 bottles, which are quality control solutions with low, medium and high concentration levels. Each bottle contains 0.5 mL and is packaged in the same way as the standard products.

[0084] (3) Extraction solution: 1 bottle, which is a methanol solution containing a stable isotope internal standard, with a volume of 25 mL (96-person specification), 100 mL (480-person specification) or 200 mL (960-person specification), and the packaging method is the same as that of the standard.

[0085] Arrange the pre-packaged standards, quality control samples, and extraction solutions neatly in the pre-formed foam insert. The insert is divided into compartments to prevent the components from colliding. Affix a product label to the outside of the box, indicating the reagent kit name, specifications, storage conditions, and expiration date.

[0086] Example 3: Sample Pretreatment

[0087] This kit tests plasma. After collecting whole blood using standard venous blood collection methods, gently invert the anticoagulant tube to mix, avoiding hemolysis, and perform plasma separation as soon as possible: balun the anticoagulant whole blood into a centrifuge and centrifuge at 1500-2000×g for 10-15 minutes (20-25℃). After centrifugation, carefully aspirate the supernatant plasma, avoiding contact with the white membrane and red blood cell layer, and transfer it to a sterile, capped sample tube. If the separated plasma can be tested immediately, store it at room temperature; if it cannot be tested immediately, freeze it at -20℃ or -80℃, avoiding repeated freeze-thaw cycles.

[0088] Sample pretreatment: Frozen plasma should be thawed at 4°C in advance and thoroughly mixed before use; add 50 µL of plasma sample to a 1.5 mL centrifuge tube, then add 200 µL of extraction buffer (methanol solution containing internal standard), shake vigorously for 1 minute to mix thoroughly, and place in a -20°C freezer for 1 hour for extraction, shaking vigorously once every 30 minutes during the process; after extraction, centrifuge at 16000 rpm for 20 minutes at 4°C, carefully aspirate 100~120 µL of supernatant into a sample vial containing an inner tube, avoiding aspirating any precipitate, cap the vial, and immediately perform instrumental analysis or temporarily store at 4°C.

[0089] Compared with the cumbersome pretreatment methods commonly used in existing technologies, such as multi-step derivatization, solid-phase extraction, or separate preparation of internal standards, the sample pretreatment of this invention adopts a "one-step extraction method"—the stable isotope internal standard is directly pre-dissolved in methanol extract, and only the addition of plasma, shaking, low-temperature precipitation, and centrifugation are required to complete the process. No additional derivatization or reconstitution steps are needed, which significantly simplifies the operation, shortens the time, and reduces human error.

[0090] The "one-step extraction method" described in this embodiment has the following significant differences and technical advantages compared with the pretreatment methods commonly used in existing technologies (such as CN119224186A), including multi-step derivatization, C18 solid-phase extraction, or step-by-step addition after separate preparation of internal standards:

[0091] (1) Simplified operation process: Existing technologies usually require 8-10 steps, such as adding internal standard, vortexing, adding precipitant, vortexing, incubation, centrifugation, transfer, nitrogen blowing, and reconstitution; This method adds the extract (containing methanol as internal standard) to the plasma in one step, and only requires 3 core steps: shaking, freezing precipitation, and centrifugation.

[0092] (2) More accurate internal standard correction: The internal standard is evenly distributed in the methanol system before extraction and undergoes the entire precipitation and extraction process synchronously with the analyte. It can effectively correct individual differences in the extraction process, while adding the internal standard step by step cannot correct the loss of the previous operation.

[0093] (3) Better control of matrix effect: This method uses low temperature precipitation at -20℃ for a long time (1 hour) combined with high speed centrifugation (16000 rpm) to remove interfering matrices such as plasma proteins and phospholipids more thoroughly. Combined with the specific HILIC column and gradient conditions in Example 3, the four target compounds with different physicochemical properties (strongly polar cAMP / GluA and weakly polar C11 / C11:1) can all obtain ideal chromatographic peak shape and signal-to-noise ratio.

[0094] Example 4: LC-MS / MS detection conditions

[0095] Analysis was performed using liquid chromatography-tandem mass spectrometry (LC-MS / MS). A Waters ACQUITY UPLCBEH Amide column (130 Å, 1.7 µm, 2.1 mm × 100 mm) was used. Mobile phase A was an aqueous solution containing 0.3% ammonia and 0.3% 5M ammonium acetate; mobile phase B was acetonitrile-water (90:10, v / v), with the aqueous phase containing 0.3% ammonia and 0.3% 5M ammonium acetate. The gradient elution program was as follows: 0–1.0 min, phase B linearly increased from 0% to 100%; 1.0–4.5 min, phase B linearly decreased from 100% to 50%; 4.5–6.5 min, maintained at 50% B; 6.5–6.7 min, phase B linearly increased from 50% to 100%; 6.7–11.5 min, maintained at 100% B. The column temperature was 30℃, the flow rate was 0.3 mL / min, and the injection volume was 1 µL.

[0096] Mass spectrometry was performed using an electrospray ionization (ESI) source in multiple reaction monitoring (MRM) mode with switching between positive and negative ions. The ESI source parameters were: sheath gas temperature 325℃, sheath gas flow rate 11.0 L / min; capillary voltage 3000V for positive ions, -3000V for negative ions; and high-purity nitrogen as the collision gas. Ion pair parameters are shown in Table 3 below.

[0097] Table 3 Ion pair parameters

[0098]

[0099] The mobile phase system of this invention uses a BEH Amide HILIC column and a high acetonitrile gradient, which enables good retention of highly polar metabolites; the constant addition of ammonia and ammonium acetate stabilizes the pH environment and improves peak shape and ionization efficiency; combined with the positive and negative ion switching mode, baseline separation of four target substances and their internal standards can be achieved in a single injection, combining high selectivity and high throughput.

[0100] Example 5: Feasibility Verification of Internal Standard Correction

[0101] To verify the feasibility of using stable isotope-labeled internal standards to correct for cyclic adenosine monophosphate (cAMP), glucuronic acid (GluA), undecanoylcarnitine (C11), and undecenoylcarnitine (C11:1), an experiment was conducted. 50 μL of blank plasma was taken, and 200 μL of methanol extract containing the corresponding internal standards (IS_cAMP, IS_GluA, and IS_C12) was added. Low, medium, and high concentration levels of cAMP, GluA, C11, and C11:1 standards were added according to Table 4, with three replicates for each concentration level. The pretreatment method described in this patent (shaking, freezing, centrifugation) was followed, and LC-MS / MS analysis was performed to calculate the spiked recoveries and relative standard deviations (RSD).

[0102] Table 4 Internal Standard Correction Results

[0103]

[0104] Using IS_cAMP as the internal standard for cAMP correction, IS_GluA as the internal standard for GluA correction, and IS_C12 as the internal standard for Car11 and Car11_1 correction, the recoveries of the four target biomarkers at low, medium, and high concentration levels were all within the range of 85%–115%, with RSDs all less than 15%, meeting the acceptance criteria for quantitative detection of biological samples. Therefore, this internal standard correction strategy is feasible for the quantitative analysis of the four target biomarkers.

[0105] Example 6: Data Processing and Model Calculation

[0106] 6.1 Method for Establishing Standard Curves

[0107] This kit utilizes system software to automatically identify the integration intervals of each biomarker. After manual checking, confirmation, or adjustment, the peak areas of each biomarker standard are exported. Independent univariate linear standard curves are fitted for each of the four biomarkers, with their theoretical concentration as the x-axis and the ratio of the analyte peak area to the corresponding internal standard peak area as the y-axis. The correlation coefficient of the curves is not less than 0.9900. The concentration of each biomarker in the sample can be calculated using this standard curve equation.

[0108] 6.2 Establishment of a Binary Logistic Regression Model

[0109] (1) Determination of covariates for age groups

[0110] To determine whether age has a significant impact on the concentration of the detected indicators, the inventors divided the subjects into four groups based on age according to the Tanner staging assessment (<12 years, 12-14 years, 14-16 years, >16 years) and used the Kruskal-Wallis H test to compare the concentration differences of each indicator among the age groups.

[0111] The results showed that cAMP (H=17.938, P<0.001), GluA (H=20.610, P<0.001), and Car11_1 (H=8.301, P=0.040) were significantly different among the four age groups. The rank mean of both cAMP and GluA increased with age (cAMP: 307.25→423.16→490.82; GluA: 386.00→406.17→488.73→501.91). Post-hoc Bonferroni pairwise comparisons showed highly significant differences in cAMP and GluA between the 12-14 year old and 14-16 year old groups (P=0.002 and P=0.001), indicating that the period around age 14 is a key turning point in concentration changes.

[0112] Table 5. Results of the overall difference test of the three indicators

[0113]

[0114] Table 6. Sample size distribution and rank mean of each age group

[0115]

[0116] Table 7. Post-hoc comparison of the two vehicles (Bonferroni correction)

[0117]

[0118] Based on the above analysis, this application uses 14 years old as the dividing point to divide the subjects into two age groups: ≤14 years old and >14 years old. These groups are included as binary covariates (AgeGroup, ≤14 years old is 1, >14 years old is 2) in the logistic regression model to correct for the confounding effect of age on metabolite concentration.

[0119] (2) Establishment of a binary logistic regression model

[0120] This kit pre-constructs a multivariate binary logistic regression model based on the concentrations of four metabolic biomarkers. During modeling, SPSS 27.0 software was used, and the samples were randomly divided into training and test sets in a 7:3 ratio. With depression status as the dependent variable and the concentrations of the four metabolic biomarkers as independent variables, multivariate binary logistic regression analysis was performed on the training set to obtain the logistic regression equation. The predictive efficacy of the model was then validated on the test set. The data analysis module includes a binary logistic regression model: P = EXP(X) / (1 + EXP(X)), X = C + a∗C. cAMP +b∗C GluA +c∗C C11:1 +d∗C C11 +e*AgeGroup, where C cAMP C GluA C C11:1 C C11 The data represent the measured concentrations (in ng / mL) of cyclic adenosine monophosphate (cAMP), glucuronic acid, undecylcarnitine, and undecanoylcarnitine collected by the data acquisition module. AgeGroup is the age grouping variable, with 14 years as the cutoff point; e-values ​​are 1 for ages ≤ 14 and 2 for ages > 14. C is a constant term, and a, b, c, d, and e are the regression coefficients for cAMP, glucuronic acid, undecylcarnitine, undecanoylcarnitine, and age grouping, respectively. The constant term C and the coefficients a, b, c, d, and e are obtained by building a model using the training set.

[0121] In this embodiment, the specific expression of the model is: X = 2.157 - 0.783 * C cAMP +0.026∗C GluA -0.078∗C C11:1 -0.943∗C C11 +0.520*AgeGroup.

[0122] Example 7: Reagent Kit Performance Evaluation

[0123] 7.1 Linear range, limit of detection, and limit of quantitation

[0124] A series of standard solutions with varying concentration gradients (six concentration points) were prepared and analyzed by LC-MS / MS according to the methods in Examples 2 and 3. Each concentration was measured three times. A standard curve was fitted with the analyte concentration on the x-axis and the peak area on the y-axis. The linear range is expressed as the range from the lowest to the highest standard concentration, and the correlation coefficient R is given. 2 ≥0.99 is an acceptable standard.

[0125] The limits of detection (LOD) and quantitation (LOQ) were determined by the signal-to-noise ratio (S / N) method: LOD is the lowest concentration when S / N ≥ 3, and LOQ is the lowest concentration when S / N ≥ 10. The results are shown in Table 8 below.

[0126] Table 8 Linear range, limit of detection, and limit of quantitation

[0127]

[0128] The above results indicate that each analyte exhibits good linearity within the specified concentration range, and the detection limit meets the requirements for clinical sample testing.

[0129] 7.2 Precision

[0130] Three quality control samples of low, medium, and high concentrations were used to conduct intra-batch and inter-batch precision experiments. Intra-batch precision: Within the same batch, each concentration of quality control sample was measured 10 times, and the coefficient of variation (CV%) was calculated. Inter-batch precision: For three consecutive batches, each concentration of quality control sample was measured 3 times in each batch, and the total coefficient of variation (CV%) for each concentration was calculated. The results are shown in Table 9 below.

[0131] Table 9 Precision

[0132]

[0133] The intra-batch CV for all analytes was ≤ 5.3%, and the inter-batch CV was ≤ 7.5%, both less than 15%, indicating that this kit has good precision.

[0134] 7.3 Recovery rate

[0135] Take blank plasma matrix and add mixed standards at low, medium and high concentrations respectively, preparing 3 samples for each concentration. After processing according to the method in Example 2, the spiked recovery rate is calculated: recovery rate (%) = (measured concentration - background concentration) / spiked concentration × 100%, and the results are shown in Table 10 below.

[0136] Table 10 Recovery Rate

[0137]

[0138] The recoveries of each analyte at low, medium and high concentrations were all between 90% and 98%, with RSD ≤ 6.2%, meeting the acceptance criteria for quantitative detection of biological samples (85%–115%).

[0139] Comparative Example 1: Construction and Performance Evaluation of a Three-Indicator Joint Diagnostic Model

[0140] A joint diagnostic model was constructed using three indicators: cAMP, GluA, and C11_1, in a sample of 934 cases. Clinical diagnosis was used as the gold standard, and a joint diagnostic model was established through binary logistic regression. To systematically evaluate the effect of covariate adjustment on model efficacy, gender and age groups (≤14 years vs >14 years) were included as covariates in the model. The results are as follows:

[0141] Table 11 Performance Evaluation of the Three-Indicator Joint Diagnostic Model

[0142]

[0143] The results showed that after incorporating the age covariate into the three-indicator model, the AUC increased from 0.813 to 0.842 (Δ=+0.029), and the Youden index increased from 0.494 to 0.529 (Δ=+0.035). This three-indicator combination has some diagnostic value, but there is still room for improvement.

[0144] Example 8. Construction and Performance Evaluation of a Four-Indicator Joint Diagnostic Model

[0145] Based on cAMP, GluA, and C11_1, C11 was added as a fourth indicator. The same clinical sample set (n=934) and modeling method as in Comparative Example 1 were used. To systematically evaluate the incremental effect of C11 and the contribution of covariate correction, different schemes were constructed in the model, and the results are as follows:

[0146] Table 12 Performance Evaluation of the Four-Indicator Joint Diagnostic Model

[0147]

[0148] The above results indicate that:

[0149] (1) Incremental effect of C11: Without including any covariates, the AUC of the four-indicator model was 0.897, an increase of 0.084 from the 0.813 of the three-indicator model, representing an increase of 10.3%; the Youden index increased from 0.494 to 0.653 (+0.159). This confirms that C11 provides independent and highly significant diagnostic information that is distinct from C11:1;

[0150] (2) Optimization effect of age covariate: After including age group (≤14 years vs >14 years), the AUC of the four-indicator model further increased from 0.897 to 0.921 (Δ=+0.024), and the Youden index increased from 0.653 to 0.711 (Δ=+0.058), confirming that the age covariate has a stable optimization effect on the model performance;

[0151] (3) The superior efficacy of the optimal solution of the present invention: The AUC of the four-indicator joint model with age covariate reached 0.921 (95% CI: 0.904–0.938), the Youden index reached 0.711, the sensitivity reached 84.4%, and the specificity reached 86.7%. Compared with the three-indicator baseline model (AUC=0.813, Youden index=0.494), the improvement was 13.3% and 43.9% respectively, indicating that the model has extremely high diagnostic accuracy and can be used as an auxiliary diagnostic tool for depression in children and adolescents.

[0152] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0153] The embodiments of the present invention have been described above with reference to the accompanying drawings. However, the present invention is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms under the guidance of the present invention without departing from the spirit and scope of the claims. All of these forms are within the protection scope of the present invention.

Claims

1. A reagent kit for detecting metabolic biomarkers of depression in children and adolescents, characterized in that, The metabolic markers detected by the kit consist of cyclic adenosine monophosphate, glucuronic acid, undecenoylcarnitine, and undecanoylcarnitine.

2. The kit according to claim 1, characterized in that: The kit also includes a mixed standard solution containing the marker, a quality control solution, and an extraction solution; the extraction solution is a methanol solution containing a stable isotope internal standard.

3. The kit according to claim 2, characterized in that: The stable isotope internal standard consists of a cyclic adenosine monophosphate stable isotope internal standard with a concentration of 0.5–5 μg / mL, a glucuronic acid stable isotope internal standard with a concentration of 10–50 μg / mL, and a lauroylcarnitine stable isotope internal standard with a concentration of 5–20 μg / mL.

4. The method for preparing the reagent kit according to any one of claims 1-3, characterized in that: include: (1) Preparation of mixed standard solutions with a series of concentration gradients: Dissolve and dilute the cyclic adenosine monophosphate, glucuronic acid, undecyl carnitine and undecyl carnitine standards with methanol, add the cyclic adenosine monophosphate stable isotope internal standard, glucuronic acid stable isotope internal standard and lauroyl carnitine stable isotope internal standard to prepare mixed standard solutions with 6 concentration levels. (2) Prepare mixed quality control solutions with low, medium and high concentration levels: Using blank plasma as a matrix, add cyclic adenosine monophosphate, glucuronic acid, undecyl carnitine and undecyl carnitine standards, as well as cyclic adenosine monophosphate stable isotope internal standard, glucuronic acid stable isotope internal standard and lauroyl carnitine stable isotope internal standard to prepare quality control solutions; (3) Preparation of extract: Dissolve the stable isotope internal standard of cyclic adenosine monophosphate, the stable isotope internal standard of glucuronic acid and the stable isotope internal standard of lauroylcarnitine in methanol to prepare an extract containing internal standards. (4) Assemble the above standard solution, quality control solution and extraction solution into a kit.

5. An application of a detection system for metabolic biomarkers of depression in children and adolescents constructed using the kit described in any one of claims 1-3, characterized in that, The detection system includes a sample pretreatment unit, a data acquisition module, and a data analysis module, wherein the data acquisition module includes the reagent kit.

6. The application as described in claim 5, characterized in that, The sample pretreatment unit includes: adding plasma sample and extraction solution to a centrifuge tube, shaking to mix, placing it at -10℃ to -30℃ for 0.5 to 2 hours for extraction, shaking once every 25 to 35 minutes during the extraction, centrifuging after extraction, and collecting the supernatant to obtain the solution to be tested.

7. The application as described in claim 6, characterized in that: The centrifugation conditions are: centrifugation at 15,000–17,000 rpm for 15–25 minutes at 4°C.

8. The application as described in claim 5, characterized in that, The data acquisition module also includes a liquid chromatography-tandem mass spectrometry (LC-MS / MS) detection unit, which comprises: mobile phase A, which is an aqueous solution containing 0.1%–0.5% ammonia and 0.1%–0.5% 3–7M ammonium acetate; mobile phase B, which is acetonitrile-water; the gradient elution program is as follows: 0–1.0 min, phase B linearly increases from 0% to 100%; 1.0–4.5 min, phase B linearly decreases from 100% to 50%; 4.5–6.5 min, maintains 50% B; 6.5–6.7 min, phase B linearly increases from 50% to 100%; 6.7–11.5 min, maintains 100% B; column temperature 20–40℃, flow rate 0.1–0.5 mL / min, and injection volume 0.5–1.5 µL.

9. The application as described in claim 5, characterized in that: The data analysis module includes a binary logistic regression model: P = EXP(X) / (1 + EXP(X)), X = C + a * C cAMP +b∗C GluA +c∗C C11:1 +d∗C C11 +e*AgeGroup; Among them, C cAMP C GluA C C11:1 C C11 The data acquisition module collected the measured concentrations of cyclic adenosine monophosphate (cAMP), glucuronic acid, undecylcarnitine, and undecanoylcarnitine, respectively. AgeGroup is the age grouping variable, with 14 years old as the cutoff point. When the age is ≤14 years old, the e value is 1, and when the age is >14 years old, the e value is 2. C is a constant term, and a, b, c, d, and e are the regression coefficients of cAMP, glucuronic acid, undecylcarnitine, undecanoylcarnitine, and age group, respectively. The constant term C and the coefficient terms a, b, c, d, and e are obtained by building a model using the training set.

Images