Application and detection kit of RvD1 as a biomarker of response to antidepressant treatment

Abstract

This invention systematically elucidates the immune-inflammatory mechanism of depression and its response to antidepressant treatment from the perspective of multidimensional biomarkers such as "neuro-immune repair". The results show that IL-1β, S100B, and RvD1 all have independent predictive value for treatment efficacy. Receiver operating characteristic (ROC) curve analysis shows that each biomarker exhibits good clinical differentiation potential: RvD1 has the highest AUC (0.911) and 100% specificity, but relatively low sensitivity (64%); S100B has an AUC of 0.898, showing high sensitivity (94%) and moderate specificity (70%); IL-1β has an AUC of 0.875, high sensitivity (92%), but slightly lower specificity (66.67%). The combined model of these three biomarkers significantly improves predictive efficacy, with an AUC as high as 0.991, sensitivity and specificity reaching 96% and 96.67%, respectively, and the highest accuracy (96.25%) and Youden index (1.927). Internal validation and calibration assessment of the model further demonstrate that this combined biomarker model has clear clinical translational value and good clinical robustness within the framework of precision medicine for depression.

Description

Technical Field

[0001] This invention belongs to the field of medication strategies and drug screening for depression, specifically involving the application of RvD1 as a biomarker for antidepressant treatment response, and an antidepressant treatment response detection kit containing an RvD1 content detection agent. Background Technology

[0002] Depression is a common mental disorder characterized primarily by persistent low mood. Core symptoms also include slowed thinking, reduced willpower, cognitive impairment, sleep deprivation, and loss of interest. In severe cases, it can lead to extreme self-harming behaviors such as suicide. Due to the clinical heterogeneity and diversity of the disease, fully elucidating the pathogenesis of depression is challenging.

[0003] Currently, numerous hypotheses have emerged regarding the pathogenesis of depression, involving various aspects such as the central monoamine neurotransmitter system, neuroendocrine system, neurotrophic substances, neuroimmune system, and the morphological structure of the central nervous system. Besides the highly complex pathogenesis, another significant challenge is the difficulty in treating depression. Drug therapy, as the cornerstone of antidepressant treatment, plays a crucial role. Most clinical guidelines for the diagnosis and treatment of depression recommend a 4-6 week observation period for the efficacy of antidepressant medication. If there is still no relief after 4-6 weeks, the treatment plan should be changed. For patients with poor treatment response, timely switching, combination therapy, or dosage adjustment of antidepressant treatment can help rapidly relieve symptoms, reduce suffering and functional impairment, and simultaneously lower the risk of suicide.

[0004] Therefore, identifying indicators of antidepressant treatment response in the early stages of treatment is of significant clinical importance. Identifying predictive biomarkers for antidepressant treatment response is not only a key requirement for optimizing clinical treatment decisions but also possesses broad translational potential.

[0005] Finding predictive biomarkers for response to antidepressant treatment is not only a key need for optimizing clinical treatment decisions, but also has broad application and translational potential. Summary of the Invention

[0006] To address the above issues, this study explores the clinical value of pre-treatment biomarkers in two progressive steps: First, it analyzes the association between baseline biomarker levels and the severity of depressive symptoms to identify biomarkers that can quantify disease severity; second, it assesses the correlation between baseline biomarker profiles and changes in efficacy endpoints, aiming to screen baseline indicators that can predict response to antidepressant medication. The core objective of this study is to identify inflammatory biomarkers that can be used to predict response to antidepressant treatment, providing a scientific basis for the practice of precision medicine for depression.

[0007] Based on the above research, this invention provides the application of RvD1 as a biomarker for response to antidepressant treatment.

[0008] The present invention also provides a method for detecting response to antidepressant treatment, including the step of detecting serum levels of RvD1.

[0009] In one specific embodiment, the method further includes the step of detecting one or more combinations of serum levels of IL-1β, S100B, and IL-1β.

[0010] In one specific implementation, the method includes the step of detecting serum levels of RvD1, S100B, and IL-1β.

[0011] This method can be used to evaluate the response of drugs to antidepressant treatment, thereby enabling the screening of antidepressant drugs.

[0012] This invention also provides the application of the RvD1 content detection reagent in the preparation of an antidepressant drug treatment response detection reagent.

[0013] The present invention also provides a kit for detecting response to antidepressant treatment, including an RvD1 content assay kit.

[0014] In one specific embodiment, the kit further includes one or both of IL-1β content detection reagent and S100B content detection reagent.

[0015] In one specific implementation, the RvD1 content detection reagent, IL-1β content detection reagent, and S100B content detection reagent are all serum content detection reagents.

[0016] This invention systematically elucidates the immune-inflammatory mechanism of depression and its response to antidepressant treatment from the perspective of multidimensional biomarkers such as "neuro-immune repair". The results show that IL-1β, S100B, and RvD1 all have independent predictive value for treatment efficacy. Receiver operating characteristic (ROC) curve analysis shows that each biomarker exhibits good clinical differentiation potential: RvD1 has the highest AUC (0.911) and 100% specificity, but relatively low sensitivity (64%); S100B has an AUC of 0.898, showing high sensitivity (94%) and moderate specificity (70%); IL-1β has an AUC of 0.875, high sensitivity (92%), but slightly lower specificity (66.67%). The combined model of these three biomarkers significantly improves predictive efficacy, with an AUC as high as 0.991, sensitivity and specificity reaching 96% and 96.67%, respectively, and the highest accuracy (96.25%) and Youden index (1.927). Internal validation and calibration assessment of the model further demonstrate that this combined biomarker model has clear clinical translational value and good clinical robustness within the framework of precision medicine for depression. Attached Figure Description

[0017] Figure 1 The correlation between each indicator and the baseline score before treatment.

[0018] Figure 2 The correlation between biomarker levels and changes in HAMD scale scores before and after treatment.

[0019] Figure 3 Box plot showing the difference in cytokine levels between the responsive and non-responsive groups.

[0020] Figure 4 Variable selection was performed using a LASSO binary logistic regression model. A represents the selection of the optimal parameter (λ) for the LASSO model based on 10-fold cross-validation and the minimum criterion. Partial likelihood bias was plotted against log(λ). The two vertical dashed lines represent the optimal λ values ​​determined by the minimum criterion (left) and the 1-SE criterion (right), respectively. B shows the LASSO coefficient trajectory of 14 candidate biomarkers. The coefficient trajectory was plotted against the log(λ) sequence. Three biomarkers (RvD1, S100B, and IL-1β) with non-zero coefficients at the optimal λ were included in the final prediction model.

[0021] Figure 5 Scatter plots of linear regressions for each meaningful indicator with HAMD-rate (deduction rate) and Delta-HAMD.

[0022] Figure 6 ROC curves for predicting serum RvD1, S100B, and IL-1β in the non-responder group.

[0023] Figure 7 This is a decision curve analysis graph.

[0024] Figure 8 This is for internal validation of the prediction model using calibration plots. Detailed Implementation

[0025] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. Furthermore, the technical features involved in the various embodiments of this invention described below can be combined with each other as long as they do not conflict with each other.

[0026] The meanings of the English abbreviations mentioned in this article are as follows. For the sake of brevity, the abbreviations will be used in other parts of this article when the following terms are used, instead of the full Chinese or English names.

[0027]

[0028] 1. Research Design and Research Population

[0029] This study employed a case-control design, and the sample consisted of patients admitted to the Wuhan Mental Health Center from February 2025 to March 2026. All patients meeting the following criteria were included: (1) patients aged 18 to 65 years (inclusive) with depression, male or female; (2) meeting the diagnostic criteria for depression according to the International Classification of Diseases, 10th Revision (ICD-10); (3) a Hamilton Depression Rating Scale (HAMD) score of ≥14 points on all 24 items before enrollment; (4) no use of antipsychotics or antidepressants in the past 3 days (sedative-hypnotic drugs were permitted); exclusion criteria included: (1) a history of manic or hypomanic episodes; (2) comorbid mental disorders; (3) alcohol or substance dependence or abuse; (4) hereditary or organic diseases or severe physical disorders; (5) intellectual disability, unable to complete the scale test; (6) a history of epileptic seizures or a family history of epileptic seizures; (7) a history of electroconvulsive therapy. All subjects and their families signed informed consent forms.

[0030] 2. Materials and Methods

[0031] 2.1 Treatment and Assessment

[0032] Collect patient demographic information, including gender, age, years of education, whether it is the first onset of the disease, age at first onset, course of disease, and BMI.

[0033] All patients underwent a three-day drug washout period before enrollment to rule out the influence of medication. All patients received different types and flexible doses of antidepressant treatment. Among them, 51 patients received SSRIs and 29 patients received SNRIs. In addition to antidepressants and mood stabilizers, 56 patients also received additional atypical antipsychotics as an adjuvant therapy.

[0034] The effectiveness of treatment was assessed using the Hamilton Depression Scale-24 (HAMD-24). The HAMD is a widely used scale for assessing depressive disorders. The scale was administered jointly by two trained psychiatrists, typically through interviews and observation. After the assessment, both assessors scored independently. Scoring before and after treatment allows for evaluation of the severity of the condition and the effectiveness of the treatment. The HAMD-24 (24-item version of the Hamilton Depression Scale) assesses symptoms across seven dimensions of depression: anxiety / somatization, weight, cognitive impairment, diurnal variation, psychomotor retardation, sleep disturbances, and hopelessness. The scale consists of 24 items, using a 5-point scale (0-4, some items 3-point scale), based on the subject's subjective feelings and behavioral performance over the past week: 0 = none, 1 = mild, 2 = moderate, 3 = severe, 4 = very severe. The total score ranges from 0 to 76 points. The commonly used clinical grading criteria are: 0-7 points for no depression, 8-19 points for mild depression, 20-34 points for moderate depression, and 35 points and above for severe depression. This scale has a Cronbach's α coefficient >0.8, and exhibits good internal consistency and test-retest reliability, making it a classic tool for assessing the severity of depressive symptoms.

[0035] The severity of depressive symptoms was assessed by experienced psychiatrists using the HAMD-24 scale at the start of treatment and at weeks 2 and 4. At week 4, a treatment response was defined as a reduction rate ≥50%, based on the change in score before and after treatment. Remission was defined as the complete disappearance of depressive symptoms within 2 weeks, an HAMD24 score ≤7, and good recovery of social functioning.

[0036] 2.2 Blood Collection and Blood Biomarker Detection

[0037] Fasting blood sampling is performed at baseline (7:00 am). 5 ml of blood is collected from a superficial vein in the upper limb in a coagulation-promoting vacuum blood collection tube. After collection, the blood sample is inverted 5-8 times and immediately centrifuged. Centrifuge at 3000×g for 10 minutes at room temperature. Carefully remove the sample and transfer the supernatant serum to a new centrifuge tube. Aliquot the serum sample into 200 μl tubes and label them. Immediately freeze the aliquoted serum samples at -20°C for short-term storage or at -80°C for long-term storage.

[0038] 1) Cytokine analysis based on multiplex microsphere detection method

[0039] The levels of seven cytokines in serum, including interferon-gamma (IFN-γ), interleukin-2 (IL-2), interleukin-4 (IL-4), interleukin-6 (IL-6), interleukin-10 (IL-10), interleukin-17A (IL-17A), and tumor necrosis factor-α (TNF-α), were measured using the ABplex Human Cytokine 7-Phase Detection Kit (Catalog No.: RK04295, ABclonal, China). The experiment was performed on an ABplex-100 detector (ABclonal, China) according to the kit instructions. 50 μL of standard or 1:4 diluted serum sample was added to a 96-well plate coated with capture antibody microspheres and incubated at room temperature with shaking for 2 h. After washing three times to remove unbound material, a biotinylated detection antibody mixture was added and incubated for 1 h, followed by incubation with streptavidin-phycoerythrin (SA-PE) for 30 min. Fluorescence signals were collected after the final wash, and data were analyzed using ABplex analysis software. The concentrations of each cytokine were calculated using a four-parameter logical fitting (4-PL) curve.

[0040] 2) Enzyme-linked immunosorbent assay (ELISA)

[0041] Serum C-reactive protein (CRP), transforming growth factor-β (TGF-β), interleukin-1β (IL-1β), interleukin-18 (IL-18), S100 calcium-binding protein B (S100B), regressor D1 (RvD1), and mature brain-derived neurotrophic factor (mBDNF) concentrations were detected using a specific quantitative ELISA kit (China Youpeng Biotechnology Co., Ltd.). All ELISAs employed a double-antibody sandwich method. Samples and standards were added to an ELISA plate pre-coated with target-specific antibodies, followed by the addition of biotin-labeled antibody. After incubation and washing, horseradish peroxidase (HRP)-labeled avidin was added, and the reaction was terminated with tetramethylbenzidine (TMB) substrate. The optical density (OD value) was measured at 450 nm using a microplate reader. All kits showed good reproducibility with intra-assay coefficient of variation (CV) <10% and inter-assay coefficient of variation <15%. All procedures were strictly performed according to the kit instructions.

[0042] 2.3 Data Quality Control

[0043] Data entry employs a dual-person, independent dual-entry model, where two systematically trained data entry personnel separately enter the raw data into an Excel database. After entry, the results are cross-checked. If discrepancies are found, the original data is immediately reviewed and corrected. During the data cleaning phase, the entered data is systematically screened by setting reasonable data ranges and logical relationships between variables, eliminating obviously abnormal or illogical records. Simultaneously, a regular review mechanism is established, covering all aspects of the process, including questionnaire completeness and the accuracy of biological sample testing records, to comprehensively ensure the authenticity and reliability of the research data.

[0044] 2.4 Statistical Analysis

[0045] Data management and statistical analysis were performed using SAS 9.4 and R 4.2. Qualitative data were described using frequency percentages [n(%)], and comparisons between groups were performed using the chi-square test and Fisher's exact test. For quantitative data, outliers were identified using the median ± 3 times the interquartile range (IQR), and winsorization was used to handle outliers. The Shaprio-Wilk test combined with histograms was used to determine normality; for normally distributed data, the mean ± standard deviation was used. The data were described using t-tests for inter-group comparisons. For quantitative data that did not conform to a normal distribution, the median and interquartile range [M(IQR)] were used for description, and the rank-sum test was used for inter-group comparisons. Based on the data distribution characteristics, Pearson product-moment correlation coefficient or Spearman rank correlation coefficient was used for correlation analysis. For continuous variables conforming to a normal distribution, Pearson's method was used to analyze the correlation between baseline indicators and baseline scale scores and their differences. For variables that did not conform to a normal distribution or were ordinal data, Spearman rank correlation analysis was used. Correlation coefficients <0.10 were considered negligible, 0.10-0.39 were considered weakly correlated, 0.40-0.69 were considered moderately correlated, 0.70-0.89 were considered strongly correlated, and ≥0.90 were considered very strongly correlated. For variables with significant partial correlation coefficients, a heatmap was used to visualize their correlation matrix, with the color intensity of the blocks indicating the strength of the correlation, and the correlation coefficient and p-value were labeled. Core variables identified by LASSO regression analysis were incorporated into a binary logistic regression model with "treatment response" as the dependent variable. The model explored whether each cytokine was an independent predictor, and reported the odds ratio (OR) and its 95% confidence interval (95% CI). Finally, the area under the receiver operating characteristic (AUC-ROC) curve was calculated to determine whether serum cytokine levels could distinguish between the responding and non-responding groups. A cutoff value was determined based on the ROC curve to balance optimal sensitivity and specificity. Furthermore, ROC curves for the combined detection of previously screened cytokines were constructed to evaluate their value in differentiating between responding and non-responding groups. The significance level was set at α=0.05, and all tests were two-tailed; P<0.05 indicated statistical significance.

[0046] 3. Results

[0047] 3.1 Baseline characteristics of the study subjects

[0048] This study compared the differences in baseline demographic parameters between the treatment-responsive group (Group 1, n=50) and the non-responsive group (Group 2, n=30) (Table 1). The results showed no statistically significant differences between the two groups in age (41.04±16.33 years in the responsive group vs. 34.50±13.51 years in the non-responsive group, t=-1.85, p=0.069), BMI (22.43±3.68 vs. 24.02±5.81, t=1.494, p=0.139), and years of education (11.64±3.58 years vs. 11.97±4.06 years, t=0.37, p=0.709) (p>0.05), indicating that these parameters had no significant impact on treatment efficacy. Furthermore, there was no significant association between the two groups in terms of gender distribution. This suggests that the treatment response was independent of gender. Notably, there was a significant difference in age of onset between the two groups (35.30±16.68 years in the effective group vs. 27.20±13.70 years in the ineffective group, t=-2.243, p=0.028).

[0049] Table 1. Comparison of baseline characteristics between the treatment-responsive and non-responsive groups.

[0050]

[0051] 3.2 Correlation between baseline cytokine levels and HAMD scores and symptoms in each dimension

[0052] Spearman correlation analysis was used to explore the correlation between inflammatory biomarkers and baseline depressive symptom dimensions (Table 2). Figure 1 At the uncorrected level (P<0.05), IL-1β was negatively correlated with the total HAMD score (ρ=-0.237, P=0.0345), IL-17A was positively correlated with cognitive symptoms (ρ=0.340, P=0.002), S100B was positively correlated with bradykinesia symptoms (ρ=0.268, P=0.0162), CRP was positively correlated with the despair dimension (ρ=0.315, P=0.0044), IL-2 was negatively correlated with the weight dimension (ρ=-0.237, P=0.034), IFN-γ and IL-1β were positively correlated with circadian rhythm symptoms, and TNF-α was negatively correlated with sleep disorders. However, after FDR multiple comparison correction, none of the above correlations reached statistical significance (FDRq values ​​were all >0.05).

[0053] Table 2 Spearman correlation coefficients between serum biomarkers and clinical symptom dimensions at baseline.

[0054]

[0055] Note: Only correlations with nominal significance (raw p-value < 0.05) are shown. All results are not adjusted for FDR.

[0056] 3.3 Correlation between cytokine levels and changes in HAMD scale scores before and after treatment

[0057] Spearman correlation analysis was used to explore the correlation between inflammatory biomarkers and the differences in HAMD total score and its dimensions before and after treatment (Table 3). Figure 2After FDR adjustment, IL-17A was positively correlated with cognitive improvement (ρ=0.371, q=0.013). Improvement in retardation was negatively correlated with IL-1β (ρ=-0.546, q<0.001), S100B (ρ=-0.306, q=0.046), and RvD1 (ρ=-0.363, q=0.015), while IL-10 was positively correlated (ρ=0.302, q=0.049). Improvement in anxiety symptoms was significantly negatively correlated with IL-1β (ρ=-0.350, q=0.020) and S100B (ρ=-0.346, q=0.021). Improved sleep was also negatively correlated with IL-1β (ρ=-0.327, q=0.034) and S100B (ρ=-0.419, q=0.003). Regarding overall depression severity, the decrease in the HAMD total score was strongly negatively correlated with IL-1β (ρ=-0.579, q<0.001), S100B (ρ=-0.577, q<0.001), and RvD1 (ρ=-0.475, q<0.001).

[0058] Table 3. Correlation between biomarker levels and changes in HAMD scale scores before and after treatment

[0059]

[0060] Note: Data are expressed as Spearman correlation coefficient (ρ). P-values ​​are corrected for multiple comparisons using the false discovery rate (FDR) method. Only association results with FDR q < 0.05 are shown.

[0061] 3.4 Differences in cytokine levels between the response group and the non-response group

[0062] Nonparametric tests and independent samples t-tests showed no statistically significant differences in most cytokine levels between the treatment-responsive group (Group=1) and the non-responsive group (Group=2) (p≥0.05), suggesting that these factors may not be the main differentiating indicators of treatment response. However, RvD1 (t=8.568, P<0.001), S100B (Z=5.918, P<0.001), and IL-1β (t=7.673, P<0.001) were significantly lower in the treatment-responsive group than in the non-responsive group, while TGF-β (t=2.595, P=0.011) was significantly higher in the treatment-responsive group than in the non-responsive group (Table 4). Figure 3 To construct a streamlined model and avoid overfitting, this study employed LASSO regression analysis to screen key predictive factors from 14 candidate biomarkers. Figure 4As shown in Figure A, the optimal tuning parameter (λ) was determined through 10-fold cross-validation. A trajectory plot was drawn with partial likelihood deviation as the ordinate and log(λ) as the abscissa. Based on the minimum criterion (λ.min), three core variables with non-zero coefficients were finally identified—RvD1, S100B, and IL-1β. Figure 4 B), thus constructing the final multivariate logistic regression model. Then, a stepwise binary logistic regression model was used to explore the independent predictive role of cytokines and covariates on treatment response (Table 5). The model showed excellent overall fit (likelihood ratio χ²=84.60, p<0.001; Nagelkerke R²=0.90), indicating that the selected variables explained 90% of the variation in treatment outcome and had strong classification predictive ability (-2LL=9.63). After controlling for age, sex, BMI, baseline HAMD score, and whether antipsychotic drugs were used concurrently, three significant predictors were retained through stepwise screening. IL-1β was an independent risk factor, with each unit increase increasing the probability of treatment ineffectiveness by 80% (β=0.59, OR=1.80, 95%CI [1.130-2.87], p=0.014). RvD1 increased the probability of ineffectiveness by 16% per unit increase (β=0.15, OR=1.16, 95%CI [1.00-1.35], p=0.046). S100B showed marginal significance, with each unit increase slightly increasing the probability of ineffectiveness (β=0.004, OR=1.004, 95%CI [1.00-1.01], p=0.051).

[0063] Table 4. Differences in cytokine levels between the response group and the non-response group.

[0064]

[0065] Table 5 Binary Logistic Regression Analysis

[0066]

[0067] 3.5 Linear Regression Analysis of Each Biomarker

[0068] Linear regression models were established with each detection indicator as the independent variable and the difference as the dependent variable. Empty models (without adjusting for confounding) and adjusted models (adjusted for Age, BMI, Education_Year, and Sex) were established separately. Results showed that among the detected inflammatory biomarkers, IL-1β, S100B, RvD1, and TGF-β1 were significantly associated with antidepressant treatment outcomes (Table 6). Figure 5Higher baseline levels of IL-1β, S100B, and RvD1 were significantly associated with lower HAMD score reduction rates and smaller decreases in total HAMD score, remaining statistically significant in both the uncorrected and corrected confounding models (adjusted P < 0.001). TGF-β was positively correlated with HAMD score reduction rates in both the uncorrected and corrected models (adjusted P = 0.026), but its association with absolute HAMD score reduction was not statistically significant after correction. In contrast, IFN-γ, IL-2, IL-4, IL-6, IL-10, IL-17A, TNF-α, CRP, IL-18, and mBDNF did not show significant associations with treatment outcomes in either the uncorrected or corrected analyses.

[0069] Table 6. Estimated Effects of Each Indicator on the Difference Reduction Rate

[0070]

[0071] 3.6 Predictive value of serum RvD1, S100B, and IL-1β for treatment response

[0072] The area under the receiver operating characteristic curve (AUC-ROC) was calculated to further explore the predictive value of RvD1, S100B, and IL-1β for treatment response. Figure 6 (Table 7). The results showed that all three single biomarkers had good discriminative ability (all P < 0.001). Among the single indicators, RvD1 had the highest AUC (0.911) and 100% specificity, but relatively low sensitivity (64%). S100B (AUC = 0.898) showed high sensitivity (94%) and moderate specificity (70%). IL-1β had an AUC of 0.875, high sensitivity (92%), but relatively low specificity (66.67%). The joint model significantly improved predictive efficacy, with an AUC of 0.991, and maintained a high balance between sensitivity (96%) and specificity (96.67%), while also having the highest accuracy (96.25%) and Youden index (1.927).

[0073] Table 7 Diagnostic efficacy of single biomarkers and combined models

[0074]

[0075] 3.7 Decision Curve Analysis of RvD1, S100B, IL-1β and Joint Model

[0076] To evaluate the clinical application value of the model, we conducted a decision curve analysis. Figure 7The results showed that, over a wide range of threshold probabilities, the predictive model (prob_trio), which incorporated RvD1 in combination with IL-1β and S100B, achieved superior net benefits across an extremely broad range of threshold probabilities. Furthermore, the net benefit generated by the combined model was significantly higher than that of single biomarkers and the "full intervention" or "no intervention" strategies.

[0077] 3.8 Internal Validation and Calibration Evaluation of the Predictive Model

[0078] To further evaluate the generalization ability of the joint prediction model and quantify the risk of overfitting, the study used the Bootstrap sampling method (repeated 1000 times) for internal validation (Table 8). Figure 8 The results show that the model exhibits excellent discriminative power and stability. The original Dxy index was 0.9827, which remained at 0.9768 after Bootstrap correction, with an optimism of only 0.0059. Correspondingly, the original AUC was 0.9913, which decreased to 0.9884 after correction, with an optimism of 0.003, indicating a very low risk of overfitting. In terms of explanatory power, the corrected coefficient of determination (R²) was 0.8735, showing a high goodness of fit. Calibration performance analysis showed that the corrected calibration slope was 0.7501 and the intercept was 0.0234, indicating slight overfitting but good overall calibration.

[0079] Table 8 Internal validation of the joint prediction model based on Bootstrap resampling (1000 repetitions)

[0080]

[0081] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. Application of RvD1 as a biomarker for response to antidepressant treatment.

2. A method for detecting response to antidepressant treatment, characterized in that, This includes steps for detecting serum RvD1 levels.

3. The method according to claim 2, characterized in that, It also includes the step of detecting one or more combinations of serum levels of IL-1β, S100B, and TGF-β1.

4. The method according to claim 3, characterized in that, This includes steps for detecting serum levels of RvD1, S100B, and IL-1β.

5. Application of RvD1 content detection reagent in the preparation of antidepressant drug treatment response detection reagent.

6. A kit for detecting response to antidepressant treatment, characterized in that, Including RvD1 content detection reagent.

7. The reagent kit according to claim 6, characterized in that, It also includes one or both of the IL-1β content detection reagent and the S100B content detection reagent.

8. The reagent kit according to claim 7, characterized in that, The RvD1, IL-1β, and S100B content detection reagents mentioned are all serum content detection reagents.

Images