All study procedures were reviewed and approved by the Ethics Committee of Bao’an District Hospital of Traditional Chinese Medicine, Shenzhen, China (Approval No. KY-2023-005-02). The trial was registered with the Chinese Clinical Trial Registry (Registration No. ChiCTR2300068054). All procedures were conducted in accordance with the Declaration of Helsinki. Written informed consent was obtained from all participants prior to enrollment.

All participants were recruited from Bao’an District Hospital of Traditional Chinese Medicine, Shenzhen, China, between July 26, 2023 and July 25, 2024. Female patients aged 18 to 45 years were included if they met the diagnostic criteria for mild depression according to International Classification of Diseases, 10th revision (ICD-10), with a HAMD-17 score >7 and <17, and an SDS score ≥53. Exclusion criteria were: (1) a history of suicidal ideation, schizophrenia, bipolar disorder, or other severe psychiatric disorders; (2) comorbid conditions such as infectious, cardiovascular, or respiratory diseases, substance abuse, or alcohol dependence; (3) fear of or intolerance to acupuncture; (4) use of antidepressants or other psychological interventions within the previous six weeks; (5) any immunological or inflammatory disorders or use of anti-inflammatory or immunosuppressive agents; and (6) pregnancy, lactation, or attempting to conceive. After skin disinfection, eligible patients received standardized acupuncture treatment at Baihui (GV20), Shenting (GV24), Zhongwan (CV12), Xiawan (CV10), Qihai (CV6), Guanyuan (CV4), Hegu (LI4), and Taichong (LR3) by licensed acupuncturists with at least 5 years of experience in the clinic, with Acupuncture needles in 0.25×25 mm and 0.25×40 mm sizes(Huatuo, Suzhou, China). (acupoint selection was performed according to the standardized anatomical locations published by the World Health Organization (WHO) for the Western Pacific Region, while needle manipulation followed the technical guidelines outlined in a nationally recognized Chinese acupuncture textbook). After needle insertion, both twirling-rotating and lifting-thrusting techniques are applied with uniform intensity to elicit deqi. Each session lasted 30 minutes, three times per week for six consecutive weeks. Acupuncture therapy was performed by licensed acupuncturists with at least 5 years of experience in the clinic. All participants received no other medication or physical therapy for depression during the acupuncture intervention period of this study. Specific details and the subsequent management process of puncture-related adverse events will be accurately documented. No serious acupuncture-related adverse events were observed during the intervention period. Details of the acupuncture protocol are provided in Table 1 and Figure 2.

In parallel, age- and BMI-matched healthy female volunteers were recruited as controls. Inclusion criteria for the control group were: (1) matched baseline characteristics (e.g., age and BMI) with the acupuncture group; (2) a HAMD-17 score <8 and SDS score <53; (3) no personal or family history of psychiatric disorders; (4) no major chronic or infectious diseases; and (5) normal inflammatory markers during screening. All participants provided written informed consent before enrollment.

Peripheral venous blood (2-4mL) was collected from each participant under fasting conditions using EDTA-coated anticoagulant tubes. Immediately after collection, the tubes were gently inverted several times to prevent clotting. The blood samples were left at room temperature for 15 to 30 minutes, followed by centrifugation at 3000 rpm for 10 minutes to separate the plasma. The upper clear plasma layer was carefully aspirated, avoiding any disturbance of the cellular layer or pellet. Plasma was aliquoted into sterile cryovials (200 µL per tube), labeled with sample ID, collection date, and participant information, and immediately stored at –80°C to ensure long-term stability and analytical reliability. Blood samples were collected outside the menstrual period. Women who were pregnant, preparing for pregnancy, or breastfeeding were excluded.

Plasma samples from female healthy controls (n=10) and female patients with mild depression (n=10), collected before and after acupuncture treatment, were analyzed using the Olink^® Target 96 Inflammation panel. This platform is based on the high-specificity and multiplex PEA technology, which enables simultaneous detection of 92 protein biomarkers per sample. In brief, each target protein is recognized by a pair of matched antibodies, each conjugated to a unique DNA oligonucleotide. Upon binding to the same target, the oligonucleotides are brought into proximity and hybridize, forming a double-stranded DNA “barcode” that serves as a template for DNA polymerase–mediated extension. The resulting DNA is then amplified using real-time quantitative PCR (qPCR) performed on a microfluidic platform. Fluorescence signals corresponding to each protein are transformed into Normalized Protein Expression (NPX) values after log2 transformation. NPX is an arbitrary unit on a log2 scale, allowing relative quantification of protein expression between samples.

DEPs from the Olink data were identified using the limma package in R. Nominal p-values were used for exploratory screening (p < 0.05), and Benjamini–Hochberg FDR-adjusted p-values were additionally calculated and reported as a reference.

The plasma expression levels of FGF-19 were further validated using a commercially available ELISA kit (Shenzhen Xinbosheng Biotechnology Co., Ltd., China, catalogue number: H241230-146a) according to the manufacturer’s instructions. A total of 120 samples were analyzed, including plasma from female healthy controls (n=40) and female patients with mild depression (n=40) before and after acupuncture treatment. Each sample was assayed in duplicate, and optical density (OD) values were measured using a microplate reader. The final concentrations were calculated based on the standard curve provided with the kit.

All data are presented as mean ± standard deviation (SD), unless otherwise specified. Statistical analyses were performed using IBM SPSS Statistics version 24.0 and R software version 4.1.3. Normality was assessed using the Shapiro–Wilk test. For normally distributed variables, differences between independent groups (HC vs MD) were assessed using independent-sample t-tests, and within-group comparisons (MD vs ACU) were evaluated using paired-sample t-tests. For non-normally distributed data, the Mann–Whitney U test was used for between-group comparisons, and the Wilcoxon signed-rank test was applied for within-group comparisons.

For Olink proteomic data, NPX values (log2 scale) were used as input. Differential protein expression was assessed using the limma package in R, reporting log fold changes (logFC) and nominal p-values. In addition, multiple-testing adjusted p-values were calculated using the Benjamini–Hochberg procedure (FDR-adjusted p-values) and provided as a reference/sensitivity assessment. Given the exploratory nature of this targeted proteomics study, candidate differentially expressed proteins (DEPs) were primarily screened using nominal p < 0.05 without applying a predefined fixed |logFC| cutoff, and key findings were further validated in an expanded cohort by ELISA.

Correlation analyses between protein levels and clinical scale scores (HAMD-17 and SDS) were performed using Pearson correlation coefficients (or Spearman correlation when normality assumptions were not met). ROC curve analyses were conducted in R to evaluate discriminative performance, and AUC values with 95% confidence intervals were reported. A two-sided p-value < 0.05 was considered statistically significant.

A total of 40 female patients with mild depression and 40 age- and sex-matched healthy controls were enrolled in the study. For each patient, peripheral plasma samples were collected both before and after a 6-week course of acupuncture treatment, resulting in a total of 120 plasma samples. The healthy control group is referred to as the HC group, the pre-treatment patient group as the MD group, and the post-treatment group as the ACU group.

Baseline characteristics were compared between the HC and MD groups. No significant differences were found in age or BMI (P>0.05), indicating that the groups were comparable. However, the MD group had significantly higher HAMD-17 and SDS scores than the HC group (both P < 0.05), confirming the presence of mild depressive symptoms at baseline. After acupuncture treatment, both HAMD-17 and SDS scores in the ACU group were significantly reduced compared to pre-treatment levels (P < 0.05), as shown in Table 2.

A total of 30 plasma samples from three groups were initially subjected to internal quality control using the Olink platform. The results demonstrated good reproducibility and no significant technical deviations among samples, meeting the requirements for downstream analysis (Figure 3). Differential expression analysis of 92 inflammation-related proteins was subsequently performed across groups. Compared with the HC group, 15 proteins were significantly downregulated in the MD group, including FGF-19, SIRT2, AXIN1, 4E-BP1, DNER, CASP-8, STAMBP, CCL11, CD5, ADA, CD244, TWEAK, NRTN, IL-17A, and IL-13 (Figures 4A, B). When comparing the MD and ACU groups, six proteins showed further downregulation following acupuncture treatment (CASP-8, CCL28, LAP TGF-beta-1, STAMBP, MCP-4, and SLAMF1), while three proteins (FGF-19, FGF-5 and IL-2) were significantly upregulated (Figures 4C, D).

At baseline, plasma FGF-19 levels were negatively correlated with HAMD-17 scores in patients with mild depression, indicating that lower FGF-19 levels were associated with greater symptom severity. After acupuncture treatment, the relationship between FGF-19 levels and HAMD-17 scores differed from baseline, which may reflect concurrent changes in biomarker levels and symptom distributions following treatment rather than a stable or direct linear association (Figures 5A–C). To further evaluate the discriminatory performance of FGF-19, ROC curve analysis was conducted. The AUC for FGF-19 in distinguishing MD from HC was 0.79 (95% CI: 0.55–0.98) (Figure 5D). Similarly, the AUC for differentiating pre- and post-acupuncture states (MD vs ACU) was 0.80 (95% CI: 0.57–0.98) (Figure 5E). These AUC values indicate a moderate discriminative ability of FGF-19 for distinguishing disease status and treatment-associated changes at the group level.

To explore the potential biological roles of the DEPs, GO and KEGG enrichment analyses were performed.

For the comparison between the HC and MD groups, GO analysis revealed that the DEPs (n=15) were primarily enriched in apoptotic process, inflammatory response, and innate immune response in the biological process (BP) category. In terms of cellular component (CC), these DEPs were mainly located in the plasma membrane, extracellular region, and cytosol. Regarding molecular function (MF), the proteins were significantly enriched in cytokine activity, receptor ligand activity, and TNF receptor binding (Figures 6A, B). KEGG pathway analysis indicated that these DEPs were predominantly involved in cytokine–cytokine receptor interaction, the PI3K-Akt signaling pathway, and pathways in cancer (Figures 6C, D).

In the comparison between the MD and ACU groups, GO analysis showed that the nine identified DEPs were enriched in biological processes such as positive regulation of cell population proliferation, regulation of cell migration, and positive regulation of MAPK cascade. The cellular component analysis indicated enrichment in the extracellular region, extracellular space, and cytoplasm. For molecular function, the DEPs were mainly enriched in growth factor activity, identical protein binding, and fibroblast growth factor receptor binding (Figures 6E, F). Consistently, KEGG pathway analysis demonstrated that these DEPs were also involved in cytokine–cytokine receptor interaction, PI3K-Akt signaling pathway, and pathways in cancer (Figures 6G, H).

Pearson correlation analysis was performed between the expression levels of inflammation-related DEPs and clinical depression scores (HAMD-17 and SDS). In the comparison between HC and MD groups, AXIN1, SIRT2, DNER, STAMBP, 4E-BP1, CASP-8, NRTN, IL-17A, CCL11, and IL-13 were all significantly negatively correlated with HAMD-17 scores (P < 0.05). In addition, FGF-19, SIRT2, AXIN1, CASP-8, CCL11, 4E-BP1, STAMBP, DNER, CD5, ADA, and TWEAK were significantly negatively correlated with SDS scores (P < 0.05) (Figure 7A).

In the comparison between MD and ACU groups (i.e., pre- and post-acupuncture), FGF-19 and IL-2 expression levels were significantly positively correlated with HAMD-17 scores (P < 0.05), and FGF-19 was also positively correlated with SDS scores (P < 0.05) (Figure 7B). This post-treatment correlation was observed within the post-acupuncture group and may reflect residual inter-individual variability following intervention.

To further validate the clinical relevance of FGF-19, an expanded cohort of 40 participants per group was analyzed using ELISA. The results showed that FGF-19 expression was significantly lower in the MD group compared to the HC group (P < 0.001), and was significantly increased in the ACU group following acupuncture treatment compared to the MD group (P < 0.01). These findings were consistent with those obtained from the Olink proteomic analysis (Figure 8).

Several limitations of this study should be acknowledged. First, although the sample size was expanded for ELISA validation, the overall cohort size remains relatively limited, which may restrict the generalizability of the findings. In addition, the study population was restricted to young females, which reduced heterogeneity but limits extrapolation to males or other age groups. Although age and sex restriction helped control major sources of variability, menstrual cycle phase and exogenous hormone use were not systematically recorded, and residual hormonal influences on inflammatory protein expression cannot be fully excluded.

Second, the absence of a sham acupuncture or waiting-list control group precludes definitive attribution of the observed changes in FGF-19 to acupuncture-specific effects. Accordingly, the treatment-associated alterations observed in this study should not be interpreted as evidence of causal or mechanism-specific efficacy.

Third, selection bias may have been introduced during the validation phase due to resource constraints that necessitated prioritization strategies. Candidate proteins were selected based on predefined criteria, including intergroup differences and correlations with clinical outcomes. While this approach facilitated efficient validation of core hypotheses, it may have overlooked biologically relevant proteins with smaller effect sizes or indirect clinical associations.

Finally, this study is exploratory in nature and primarily aimed at identifying candidate biomarkers through high-throughput screening. Although pathway enrichment analysis provided supportive biological context, no functional experiments were conducted to directly assess causal mechanisms. Future studies incorporating larger cohorts, sham-controlled clinical designs, and multi-level central–peripheral immune–neural analyses will be necessary to clarify the mechanistic role of FGF-19 and its relevance to acupuncture-associated effects.