Intervention of Astragalus Membranaceus Extract in rats of spinal cord injury: a systematic review and meta-analysis

Abstract

Objective: Spinal cord injury (SCI) causes motor, sensory and autonomic dysfunction below the level of injury and its incidence is increasing every year. Astragalus Membranaceus Extract (AME) has received attention in spinal cord injury in recent years, but its specific effects in spinal cord injury are unclear.

Methods: Databases of PubMed, Embase, WOS, Cochrane Library, FMRS, Clinical trial, CNKI, VIP, and WangFang were searched from their establishment to December 1, 2024 using the following terms: "Astragalus propinquus," "Huang qi," "Astragalus mongholicus Bunge," "Spinal cord injuries," "spinal cord diseases," "spinal cord trauma." To ensure comprehensiveness, the search strategy included both traditional names (Astragalus) and scientific names (Astragalus membranaceus). Only studies published in Chinese or English were included. Cross-sectional studies, survey designs, quality improvement studies, and other study designs that did not meet the inclusion criteria were excluded.

Results: After screening, a total of 16 studies with 996 animals were included in the review. Astragalus Membranaceus Extract (AME) administration was associated with more significant functional recovery (mean difference [MD] = 3.68, 95% CI = 2.74, 4.62). Subgroup analyses showed the best functional recovery of the spinal cord when the dose exceeded 20 units and the duration of treatment was less than 14 days.

Conclusion: Our study suggests that AME has therapeutic potential for spinal cord injured rats. Further studies are needed to determine if this can be developed into a new alternative therapy through experimental and clinical studies with larger samples.

Systematic review registration: Identifier: CRD42024623721, website: https://www.crd.york.ac.uk/PROSPERO/.

Keywords: Huang qi; animal experiments; meta-analysis; phytomedicine; spinal cord injury.

Figure 1

PRISMA 2020 flow diagram of study selection process. A comprehensive search was conducted using key terms across multiple databases, initially identifying 139 studies. First, 59 duplicate records were removed. Second, 24 studies were excluded based on title and abstract screening for not meeting the inclusion criteria. Third, 4 in vitro studies and 28 studies involving non–spinal cord injury models were excluded for the same reason. Additionally, 12 studies were removed due to missing essential data in the results, and 2 studies were excluded due to inaccessibility. Ultimately, 16 studies were included in the final analysis.

Figure 2

Quality of the included studies. Selection bias includes random sequence generation, baseline characteristics, and allocation concealment; performance bias includes random housing and blinding of participants and personnel; detection bias includes random outcome assessment and blinding of outcome assessment; attrition bias: incomplete outcome data; reporting bias: selective reporting; Other biases: new animals added to the experimental group, etc. Bias analysis indicates that most of the included studies had deficiencies in baseline characteristics, allocation concealment, and randomization. Related studies need to be improved in terms of the randomization process and data integrity control. However, in terms of reporting completeness and control of other potential biases, the included studies met the relevant standards.

Figure 3

Forest plot of BBB test scores for the AME group vs. the control group. The BBB test included a total of 9 studies, with 156 rats in the AME group and 156 in the control group, totaling 312 rats. Mean Difference (MD): 3.68, 95% confidence interval (95% CI): [2.74, 4.62], and Test for overall effect: Z = 7.67, (p < 0.00001), indicating a significant statistical difference between the experimental and control groups, with the AME showing superior efficacy. The heterogeneity test showed Tau² = 1.75, Chi² = 108.92, df = 8, (p < 0.00001), and I² = 93%, indicating high heterogeneity among the included studies.

Figure 4

Forest plot of inclined plate test scores for the AME group vs. the control group. A total of 6 studies were included, with 51 cases in the AME group and 51 cases in the control group, for a total sample size of 102 cases. The standardized mean difference (SMD) was calculated as 1.67, 95% CI: [0.99, 2.34], and Test for overall effect: Z = 4.83, (p < 0.00001), indicating a significant statistical difference between the experimental and control groups, with the AME group showing superior efficacy. Tau² = 0.24, Chi² = 7.01, df = 4 (p = 0.14), and I² = 43%, indicating moderate heterogeneity among the studies.

Figure 5

Forest plot of inspection results for the AME group vs. the control group. (a): SOD, (b): MDA, (c): IL-6, (d): TNF-α, (e): IL-1β. (a) Three studies were included, with 85 rats in the AME group and 84 rats in the control group, totaling 169 rats. SMD = 2.42, 95% CI: [1.78, 3.07], Z = 7.35 (p < 0.00001), indicating that the AME group had a superior effect. Heterogeneity: Tau² = 0.19, Chi² = 4.70, df = 2 (p = 0.10), I² = 57%, indicating moderate heterogeneity. (b) Two studies were included, with 36 rats in the AME group and 36 rats in the control group, totaling 72 rats. SMD = −8.60, 95% CI: [−11.86, −5.33], Z = 5.17 (p < 0.00001), indicating that the AME group had a superior effect. Heterogeneity: Tau² = 4.26, Chi² = 4.29, df = 1 (p = 0.04), I² = 77%, indicating high heterogeneity. (c) Two studies were included, with 13 rats in the AME group and 13 rats in the control group, totaling 26 rats. SMD = −2.14, 95% CI: [−3.21, −1.07], Z = 3.93 (p < 0.0001), indicating that the AME group had a superior effect. Heterogeneity: Chi² = 0.37, df = 1 (p = 0.54), I² = 0%, indicating no heterogeneity. (d) Two studies were included, with 13 rats in the AME group and 13 rats in the control group, totaling 26 rats. SMD = −4.13, 95% CI: [−5.69, −2.57], Z = 5.19 (p < 0.00001), indicating that the AME group had a superior effect. Heterogeneity: Chi² = 0.29, df = 1 (p = 0.59), I² = 0%, indicating no heterogeneity. (e) Two studies were included, with 23 rats in the AME group and 23 rats in the control group. SMD = −3.95, 95% CI: [−5.02, −2.88], Z = 7.23 (p < 0.00001), indicating that the AME group had a superior effect. Heterogeneity: Chi² = 0.33, df = 1 (p = 0.56), I² = 0%, no heterogeneity.

Figure 6

Sensitivity analysis of BBB test scores. CI, confidence interval. Estimate: after excluding various studies, the effect sizes exhibited positional changes. However, the overall did not show extreme deviation. Lower and Upper CI Limit: following the exclusion of studies, there were no dramatic changes in the upper and lower limits. As depicted in figure, the effect sizes and confidence intervals did not undergo substantial changes following the exclusion of a single study. This suggests that the results were relatively stable.