ObjectiveThe essence of syndrome manifestation recognition in traditional Chinese medicine (TCM) is to infer the body’s latent pathogenesis state from clinical observational information, rather than to perform simple label matching. However, previous studies have largely modeled this task as syndrome pattern classification within a fixed label space, which does not adequately reflect the cognition process of TCM syndrome differentiation centered on pathogenesis reasoning, and is also insufficient to capture the openness, semantic variability, and cross-disease reusability of syndrome manifestation expression. This study aimed to investigate whether introducing pathogenesis reasoning chain-of-thought (PR-CoT) supervision into large language models (LLMs) could improve the quality and cognitive consistency of syndrome manifestation recognition and support cross-disease transfer. MethodsSyndrome manifestation recognition was formulated as a conditional generation task under the framework of clinical observational information (X)→pathogenesis structure (Z)→syndrome pattern output (Y), where Z serves as an explicit intermediate structural variable linking the clinical evidence and syndrome judgment. Within this framework, a PR-CoT-supervised dataset for syndrome manifestation recognition was constructed based on medical case records of spleen-stomach disorders. After preprocessing, information extraction, manual proofreading, and data cleaning, the dataset comprised 4 800 training cases, 400 development cases, and 400 test cases. Each sample was annotated with a structured PR-CoT consisting of three progressive levels: clinical information summarization, comprehensive pathogenesis analysis, and syndrome pattern output. Supervised fine-tuning was conducted on open-source LLMs, with an end-to-end model serving as the baseline. Qwen3-32B was used as the primary experimental model, and Qwen3-14B as the scale comparison model. A progressive multidimensional evaluation framework was further established, comprising a structural parsing level, a semantic similarity level, and an expert blind review level. At the structural parsing level, syndrome pattern expressions were decomposed into structural elements and evaluated using Precision, Recall, F1 score, and Jaccard similarity. At the semantic similarity level, independent LLMs scored the theoretical proximity between predicted and reference syndrome patterns. At the expert blind review level, three TCM experts independently evaluated model outputs on two dimensions: syndrome differentiation consistency and terminology standardization of syndrome patterns. In addition, zero-shot cross-disease transfer evaluation was conducted on gynecological and heart-system disorder test sets. ResultsAt the structural parsing level, PR-CoT supervision did not lead to a stable improvement in the element-wise overlap of syndrome pattern structural components. Compared with the corresponding baselines, neither Qwen3-32B nor Qwen3-14B showed consistent advantages in structural matching metrics after the introduction of PR-CoT supervision. In contrast, at the semantic similarity level, PR-CoT supervision produced stable positive gains across different model scales and evaluation systems. The average semantic score of Qwen3-32B increased from 6.425 8 in the baseline model to 6.585 0 after PR-CoT supervision, and that of Qwen3-14B increased from 5.870 0 to 5.964 2. At the expert blind review level, the overall score of Qwen3-32B (PR-CoT) was 7.026 0±0.107 7, higher than 6.416 3±0.288 9 for its baseline. In zero-shot cross-disease testing, the PR-CoT model still showed advantages in semantic evaluation and expert evaluation on both gynecological and heart-system disorder test sets, indicating a certain degree of transferability. ConclusionThe benefits of PR-CoT supervision are mainly reflected in TCM semantic consistency and clinical plausibility, rather than in improved hard matching of structural elements. These findings support understanding syndrome manifestation recognition as a process of generating and expressing latent pathogenesis structures, rather than as a classification task within a traditional fixed label space. By introducing pathogenesis reasoning as an explicit intermediate structure into the modeling process and combining it with a progressive multidimensional evaluation framework, this study provides a methodological pathway for intelligent TCM syndrome differentiation that integrates theoretical alignment, interpretability, and multi-level evaluation.

ObjectiveThe essence of syndrome manifestation recognition in traditional Chinese medicine (TCM) is to infer the body’s latent pathogenesis state from clinical observational information, rather than to perform simple label matching. However, previous studies have largely modeled this task as syndrome pattern classification within a fixed label space, which does not adequately reflect the cognition process of TCM syndrome differentiation centered on pathogenesis reasoning, and is also insufficient to capture the openness, semantic variability, and cross-disease reusability of syndrome manifestation expression. This study aimed to investigate whether introducing pathogenesis reasoning chain-of-thought (PR-CoT) supervision into large language models (LLMs) could improve the quality and cognitive consistency of syndrome manifestation recognition and support cross-disease transfer. MethodsSyndrome manifestation recognition was formulated as a conditional generation task under the framework of clinical observational information (X)→pathogenesis structure (Z)→syndrome pattern output (Y), where Z serves as an explicit intermediate structural variable linking the clinical evidence and syndrome judgment. Within this framework, a PR-CoT-supervised dataset for syndrome manifestation recognition was constructed based on medical case records of spleen-stomach disorders. After preprocessing, information extraction, manual proofreading, and data cleaning, the dataset comprised 4 800 training cases, 400 development cases, and 400 test cases. Each sample was annotated with a structured PR-CoT consisting of three progressive levels: clinical information summarization, comprehensive pathogenesis analysis, and syndrome pattern output. Supervised fine-tuning was conducted on open-source LLMs, with an end-to-end model serving as the baseline. Qwen3-32B was used as the primary experimental model, and Qwen3-14B as the scale comparison model. A progressive multidimensional evaluation framework was further established, comprising a structural parsing level, a semantic similarity level, and an expert blind review level. At the structural parsing level, syndrome pattern expressions were decomposed into structural elements and evaluated using Precision, Recall, F1 score, and Jaccard similarity. At the semantic similarity level, independent LLMs scored the theoretical proximity between predicted and reference syndrome patterns. At the expert blind review level, three TCM experts independently evaluated model outputs on two dimensions: syndrome differentiation consistency and terminology standardization of syndrome patterns. In addition, zero-shot cross-disease transfer evaluation was conducted on gynecological and heart-system disorder test sets. ResultsAt the structural parsing level, PR-CoT supervision did not lead to a stable improvement in the element-wise overlap of syndrome pattern structural components. Compared with the corresponding baselines, neither Qwen3-32B nor Qwen3-14B showed consistent advantages in structural matching metrics after the introduction of PR-CoT supervision. In contrast, at the semantic similarity level, PR-CoT supervision produced stable positive gains across different model scales and evaluation systems. The average semantic score of Qwen3-32B increased from 6.425 8 in the baseline model to 6.585 0 after PR-CoT supervision, and that of Qwen3-14B increased from 5.870 0 to 5.964 2. At the expert blind review level, the overall score of Qwen3-32B (PR-CoT) was 7.026 0±0.107 7, higher than 6.416 3±0.288 9 for its baseline. In zero-shot cross-disease testing, the PR-CoT model still showed advantages in semantic evaluation and expert evaluation on both gynecological and heart-system disorder test sets, indicating a certain degree of transferability. ConclusionThe benefits of PR-CoT supervision are mainly reflected in TCM semantic consistency and clinical plausibility, rather than in improved hard matching of structural elements. These findings support understanding syndrome manifestation recognition as a process of generating and expressing latent pathogenesis structures, rather than as a classification task within a traditional fixed label space. By introducing pathogenesis reasoning as an explicit intermediate structure into the modeling process and combining it with a progressive multidimensional evaluation framework, this study provides a methodological pathway for intelligent TCM syndrome differentiation that integrates theoretical alignment, interpretability, and multi-level evaluation.

Objective:To analyze the effect of continuous positive airway pressure (CPAP) on maternal and neonatal outcomes in pregnant women with obstructive sleep apnea syndrome (OSAS), especially on the incidence of hypertensive disorder in pregnancy (HDP) in women with moderate to severe OSAS.Methods:A total of 180 pregnant women with OSAS who were diagnosed through sleep monitoring during pregnancy due to high-risk factors of OSAS and registered in Peking University People′s Hospital from January 2021 to May 2024 were selected as the study subjects. Clinical data were collected from medical records for retrospective analysis. According to whether they received standardized treatment with CPAP, they were divided into the CPAP treatment group (42 cases) and the control group (138 cases). The CPAP treatment group consisted of 9 pregnant women with moderate to severe OSAS, while the control group consisted of 34 pregnant women with moderate to severe OSAS. The maternal and neonatal outcomes, the incidence of HDP, placental weight after delivery and placental weight/neonatal birth weight ratio were compared between the two groups.Results:(1) The average gestational age of pregnant women in the CPAP treatment group was higher than that in the control group [(38.7±1.0) vs (38.0±1.4) weeks], the proportion of infants small for gestational age (SGA) in the CPAP treatment group was lower [0 (0/42) vs 12.3% (17/138)], and the birth weight of infants in the CPAP treatment group was bigger [(3 396±475) vs (3 082±710) g); the differences between the two groups were statistically significant (all P<0.05). There were no significant differences between the CPAP treatment group and the control group in terms of delivery mode, rates of postpartum hemorrhage and preterm birth, umbilical artery blood gas analysis pH<7.1, lactate≥6.0 mmol/L, base excess<-12.0 mmol/L and the incidence of gestational diabetes mellitus and HDP (all P>0.05). (2) The placental weight of the CPAP treatment group was significantly lower than that of the control group [(554.0±70.6) vs (615.7±119.1) g], the placental weight/newborn birth weight ratio of the CPAP treatment group was significantly lower than that of the control group (median: 0.17 vs 0.19), and the differences were statistically significant (all P<0.05). (3) The incidence of HDP in pregnant women with moderate to severe OSAS in the CPAP treatment group was lower than that in the control group [1/9 vs 61.8% (21/34)], and the difference was statistically significant ( P<0.05). Conclusions:CPAP treatment could prolong the gestational age in pregnant women with OSAS, reduce the incidence of SGA, increase the birth weight of infants, and reduce the incidence of HDP in pregnant women with moderate to severe OSAS, and is worth promoting in clinical practice. The improvement of neonatal outcomes by CPAP treatment is closely related to the placenta, which is worthy of further exploration.

Objective:To analyze the effect of continuous positive airway pressure (CPAP) on maternal and neonatal outcomes in pregnant women with obstructive sleep apnea syndrome (OSAS), especially on the incidence of hypertensive disorder in pregnancy (HDP) in women with moderate to severe OSAS.Methods:A total of 180 pregnant women with OSAS who were diagnosed through sleep monitoring during pregnancy due to high-risk factors of OSAS and registered in Peking University People′s Hospital from January 2021 to May 2024 were selected as the study subjects. Clinical data were collected from medical records for retrospective analysis. According to whether they received standardized treatment with CPAP, they were divided into the CPAP treatment group (42 cases) and the control group (138 cases). The CPAP treatment group consisted of 9 pregnant women with moderate to severe OSAS, while the control group consisted of 34 pregnant women with moderate to severe OSAS. The maternal and neonatal outcomes, the incidence of HDP, placental weight after delivery and placental weight/neonatal birth weight ratio were compared between the two groups.Results:(1) The average gestational age of pregnant women in the CPAP treatment group was higher than that in the control group [(38.7±1.0) vs (38.0±1.4) weeks], the proportion of infants small for gestational age (SGA) in the CPAP treatment group was lower [0 (0/42) vs 12.3% (17/138)], and the birth weight of infants in the CPAP treatment group was bigger [(3 396±475) vs (3 082±710) g); the differences between the two groups were statistically significant (all P<0.05). There were no significant differences between the CPAP treatment group and the control group in terms of delivery mode, rates of postpartum hemorrhage and preterm birth, umbilical artery blood gas analysis pH<7.1, lactate≥6.0 mmol/L, base excess<-12.0 mmol/L and the incidence of gestational diabetes mellitus and HDP (all P>0.05). (2) The placental weight of the CPAP treatment group was significantly lower than that of the control group [(554.0±70.6) vs (615.7±119.1) g], the placental weight/newborn birth weight ratio of the CPAP treatment group was significantly lower than that of the control group (median: 0.17 vs 0.19), and the differences were statistically significant (all P<0.05). (3) The incidence of HDP in pregnant women with moderate to severe OSAS in the CPAP treatment group was lower than that in the control group [1/9 vs 61.8% (21/34)], and the difference was statistically significant ( P<0.05). Conclusions:CPAP treatment could prolong the gestational age in pregnant women with OSAS, reduce the incidence of SGA, increase the birth weight of infants, and reduce the incidence of HDP in pregnant women with moderate to severe OSAS, and is worth promoting in clinical practice. The improvement of neonatal outcomes by CPAP treatment is closely related to the placenta, which is worthy of further exploration.

Climate and land use changes may significantly impact the habitat distribution of Gastrodia elata, an endangered traditional medicinal plant. Accurately predicting its future potential suitable habitats is crucial for its conservation and sustainable development. This study integrates current distribution data of G. elata with 56 environmental variables and uses the MaxEnt model to predict changes in its suitable habitats under current climate conditions and four future climate scenarios(SSP1-2.6, SSP2-4.5, SSP3-7.0, and SSP5-8.5). The results show that October precipitation and December minimum temperature are key environmental factors influencing its distribution. Under the current climate, optimal habitats for G. elata are concentrated in montane forest areas in Sichuan, Yunnan, Guizhou, and Hubei, which meet the species' requirements for understory growth. Across all future scenarios, the suitable habitat of G. elata consistently shows a stable northward shift, with a steady increase in suitable areas, extending to the middle and lower reaches of the Yangtze River and the Huang-Huai region, and even expanding into Liaoning, Jilin, and southern Heilongjiang. Land use analysis, taking into account the protection of arable land and the utilization of forest resources, indicates that by 2100, under future climate conditions, arable land in medium-to high-suitability areas is expected to increase by 30%-124%. While the conversion of non-suitable forest land into suitable habitats is projected to increase by 5%-52%, the growth of medium-to high-suitability areas within forests is relatively modest, ranging from 1% to 24%. These findings highlight the need to balance agricultural expansion with forest resource conservation to ensure the long-term sustainability of G. elata and provide scientific guidance for future suitable habitat management.
Ecosystem ; China ; Climate Change ; Gastrodia/growth & development* ; Conservation of Natural Resources ; Plants, Medicinal/growth & development*

This study aims to investigate the in vitro mechanisms underlying the beneficial effects of puerarin on hepatic insulin resistance(IR) based on the carbohydrate response element-binding protein(ChREBP)/peroxisome proliferator-activated receptor(PPAR)α/PPARγ axis involved in glucose and lipid metabolism. An IR-HepG2 cell model was established by treating cells with dexamethasone for 48 h, and the cells were then treated with 10, 20, and 40 μmol·L~(-1) puerarin for 24 h. Glucose levels and output in the extracellular fluid were measured by the glucose oxidase method, while cell viability was assessed by the cell counting kit-8(CCK-8) assay. The adenosine triphosphate(ATP) content and glycogen synthesis were evaluated through chemiluminescence and periodic acid-Schiff staining, respectively. Western blot was employed to quantify the protein levels of forkhead box protein O1(FoxO1), phosphorylated forkhead box protein O1 [p-FoxO1(Ser256)], glucagon, phosphofructokinase, liver type(PFKL), pyruvate kinase L-R(PKLR), pyruvate dehydrogenase complex 1(PDHA1), insulin receptor substrate 2(IRS2), phosphatidylinositol 3-kinase p85(PI3KR1), phosphorylated protein kinase B [p-Akt(Thr308)], glycogen synthase(GYS), glycogen phosphorylase, liver type(PYGL), adiponectin(ADPN), ChREBP, PPARα, and PPARγ. Additionally, the protein levels of acetyl-CoA carboxylase 1(ACC1), phosphorylated ATP citrate lyase [p-ACLY(Ser455)], sterol regulatory element binding protein 1c(SREBP-1c), peroxisome proliferator-activated receptor gamma coactivator 1α(PGC1α), carnitine palmitoyltransferase 1α(CPT1α), and glucagon receptor(GCGR) were also determined. Immunofluorescence was employed to visualize the expression and nuclear location of ChREBP/PPARα/PPARγ. Furthermore, quantitative PCR with the antagonists GW6471 and GW9662 was employed to assess Pparα, Pparγ, and Chrebp. The findings indicated that puerarin effectively reduced both the glucose level and glucose output in the extracellular fluid of IR-HepG2 cells without obvious effect on the cell viability, and it increased intracellular glycogen and ATP levels. Puerarin down-regulated the protein levels of FoxO1 and glucagon while up-regulating the protein levels of p-FoxO1(Ser256), PFKL, PKLR, PDHA1, IRS2, PI3KR1, p-Akt(Thr308), GYS, PYGL, ADPN, ACC1, SREBP-1c, p-ACLY(Ser455), PGC1α, CPT1α, and GCGR in IR-HepG2 cells. Furthermore, puerarin up-regulated both the mRNA and protein levels of ChREBP, PPARα, and PPARγ and promoted the translocation into the nucleus. GW6471 was observed to down-regulate the expression of Pparα while up-regulating the expression of Chrebp and Pparγ. GW9662 down-regulated the expression of Pparγ while up-regulating the expression of Pparα, with no significant effect on Chrebp. In summary, puerarin activated the hepatic ChREBP/PPARα/PPARγ axis, thereby coordinating the glucose and lipid metabolism, promoting the conversion of glucose to lipids to exert the blood glucose-lowering effect.
Isoflavones/pharmacology* ; Humans ; PPAR gamma/genetics* ; Hep G2 Cells ; Glucose/metabolism* ; Lipid Metabolism/drug effects* ; PPAR alpha/genetics* ; Liver/drug effects* ; Basic Helix-Loop-Helix Leucine Zipper Transcription Factors/genetics* ; Insulin Resistance

Traditional Chinese medicine(TCM) decoction pieces are a core carrier for the inheritance and innovation of TCM, and their quality and safety are critical to public health and the sustainable development of the industry. Conventional quality control models, while having established a well-developed system through long-term practice, still face challenges such as relatively long inspection cycles, insufficient objectivity in characterizing complex traits, and urgent needs for improving the efficiency of integrating multidimensional quality information when confronted with the dual demands of large-scale production and precision quality control. With the rapid development of artificial intelligence, machine learning can deeply analyze multidimensional data of the morphology, spectroscopy, and chemical fingerprints of decoction pieces by constructing high-dimensional feature space analysis models, significantly improving the standardization level and decision-making efficiency of quality evaluation. This article reviews the research progress in the application of machine learning in the processing, production, and rapid quality evaluation of TCM decoction pieces. It further analyzes current challenges in technological implementation and proposes potential solutions, offering theoretical and technical references to advance the digital and intelligent transformation of the industry.
Machine Learning ; Drugs, Chinese Herbal/standards* ; Quality Control ; Medicine, Chinese Traditional/standards* ; Humans

This paper aims to investigate the hypoglycemic effect and mechanism of berberine in improving energy metabolism based on the multi-pathway regulation of brain and muscle aromatic hydrocarbon receptor nuclear translocal protein 1(BMAL1): cyclin kaput complex of day-night spontaneous output cyclin kaput(CLOCK). The dexamethasone-induced hepatic insulin resistance(IR) HepG2 cell model was used; 0.5, 1, 5, 10, 20 μmol·L~(-1) berberine were administered at 15, 18, 21, 24, 30, 36 h. The time-dose effect of glucose content in extracellular fluid was detected by glucose oxidase method. The optimal dosage and time of berberine were determined for the follow-up study. Glucose oxidase method and chemiluminescence method were respectively performed to detect hepatic glucose output and relative content of ATP in cells; Ca~(2+), reactive oxygen species(ROS), mitochondrial structure and membrane potential were detected by fluorescent probes. Moreover, ultraviolet colorimetry method was used to detect the liver type of pyruvate kinase(L-PK) and phosphoenol pyruvate carboxykinase(PEPCK). In addition, pyruvate dehydrogenase E1 subunit α1(PDHA1), phosphate fructocrine-liver type(PFKL), forkhead box protein O1(FoxO1), peroxisome proliferator-activated receptor gamma co-activator 1α(PGC1α), glucose-6-phosphatase(G6Pase), glucagon, phosphorylated nuclear factor-red blood cell 2-related factor 2(p-Nrf2)(Ser40), heme oxygenase 1(HO-1), NAD(P)H quinone oxidoreductase 1(NQO1), fibroblast growth factor 21(FGF21), uncoupled protein(UCP) 1 and UCP2 were detected by Western blot. BMAL1:CLOCK complex was detected by immunofluorescence double-staining method, combined with small molecule inhibitor CLK8. Western blot was used to detect PDHA1, PFKL, FoxO1, PGC1α, G6Pase, glucagon, Nrf2, HO-1, NQO1, FGF21, UCP1 and UCP2 in the CLK8 group. The results showed that berberine downregulated the glucose content in extracellular fluid in IR-HepG2 cells in a time-and dose-dependent manner. Moreover, berberine inhibited hepatic glucose output and reduced intracellular Ca~(2+) and ROS whereas elevated JC-1 membrane potential and improved mitochondrial structure to enhance ATP production. In addition, berberine upregulated the rate-limiting enzymes such as PFKL, L-PK and PDHA1 to promote glycolysis and aerobic oxidation but also downregulated PGC1α, FoxO1, G6Pase, PEPCK and glucagon to inhibit hepatic gluconeogenesis. Berberine not only upregulated p-Nrf2(Ser40), HO-1 and NQO1 to enhance antioxidant capacity but also upregulated FGF21, UCP1 and UCP2 to promote energy metabolism. Moreover, berberine increased BMAL1, CLOCK and nuclear BMAL1:CLOCK complex whereas CLK8 reduced the nuclear BMAL1:CLOCK complex. Finally, CLK8 decreased PDHA1, PFKL, Nrf2, HO-1, NQO1, FGF21, UCP1, UCP2 and increased FoxO1, PGC1α, G6Pase and glucagon compared with the 20 μmol·L~(-1) berberine group. BMAL1:CLOCK complex inhibited gluconeogenesis, promoted glycolysis and glucose aerobic oxidation pathways, improved the reduction status within mitochondria, protected mitochondrial structure and function, increased ATP energy storage and promoted energy consumption in IR-HepG2 cells. These results suggested that berberine mediated BMAL1:CLOCK complex to coordinate the regulation of hepatic IR cells to improve energy metabolism in vitro.
Humans ; Berberine/pharmacology* ; Gluconeogenesis/drug effects* ; Hep G2 Cells ; Glucose/metabolism* ; Liver/drug effects* ; Energy Metabolism/drug effects* ; Hypoglycemic Agents/pharmacology* ; ARNTL Transcription Factors/genetics* ; Glycolysis/drug effects* ; Oxidation-Reduction/drug effects*

BACKGROUND: Delayed platelet engraftment is a common complication after umbilical cord blood transplantation (UCBT), and there is no standard therapy. Avatrombopag (AVA) is a second-generation thrombopoietin (TPO) receptor agonist (TPO-RA) that has shown efficacy in immune thrombocytopenia (ITP). However, few reports have focused on its efficacy in patients diagnosed with thrombocytopenia after allogeneic hematopoietic stem cell transplantation (allo-HSCT). METHODS: We conducted a retrospective study at the First Affiliated Hospital of the University of Science and Technology of China to evaluate the efficacy of AVA as a first-line TPO-RA in 65 patients after UCBT; these patients were compared with 118 historical controls. Response rates, platelet counts, megakaryocyte counts in bone marrow, bleeding events, adverse events and survival rates were evaluated in this study. Platelet reconstitution differences were compared between different medication groups. Multivariable analysis was used to explore the independent beneficial factors for platelet implantation. RESULTS: Fifty-two patients were given AVA within 30 days post-UCBT, and the treatment was continued for more than 7 days to promote platelet engraftment (AVA group); the other 13 patients were given AVA for secondary failure of platelet recovery (SFPR group). The median time to platelet engraftment was shorter in the AVA group than in the historical control group (32.5 days vs . 38.0 days, Z  = 2.095, P  = 0.036). Among the 52 patients in the AVA group, 46 achieved an overall response (OR) (88.5%), and the cumulative incidence of OR was 91.9%. Patients treated with AVA only had a greater 60-day cumulative incidence of platelet engraftment than patients treated with recombinant human thrombopoietin (rhTPO) only or rhTPO combined with AVA (95.2% vs . 84.5% vs . 80.6%, P  <0.001). Patients suffering from SFPR had a slightly better cumulative incidence of OR (100%, P  = 0.104). Patients who initiated AVA treatment within 14 days post-UCBT had a better 60-day cumulative incidence of platelet engraftment than did those who received AVA after 14 days post-UCBT (96.6% vs . 73.9%, P  = 0.003). CONCLUSION Compared with those in the historical control group, our results indicate that AVA could effectively promote platelet engraftment and recovery after UCBT, especially when used in the early period (≤14 days post-UCBT).
Humans ; Female ; Male ; Thrombocytopenia/etiology* ; Adult ; Retrospective Studies ; Cord Blood Stem Cell Transplantation/adverse effects* ; Middle Aged ; Adolescent ; Young Adult ; Thiazoles/adverse effects* ; Platelet Count ; Receptors, Thrombopoietin/agonists* ; Child ; Thiophenes