Growing attention on the toxicity of Chinese herbal medicine: a bibliometric analysis from 2013 to 2022

Kexin Zhu, Min Wu, Zhi-Lin Bian, Shi-Liang Han, Liming Fang, Fengfeng Ge, Xue-Zhou Wang, Shengfang XieFrontiers in Pharmacology 2024 · DOI

Abstract

Introduction: Despite the clinical value of Chinese herbal medicine (CHM), restricted comprehension of its toxicity limits the secure and efficacious application. Previous studies primarily focused on exploring specific toxicities within CHM, without providing an overview of CHM's toxicity. The absence of a quantitative assessment of focal points renders the future research trajectory ambiguous. Therefore, this study aimed to reveal research trends and areas of concern for the past decade.

Methods: A cross-sectional study was conducted on publications related to CHM and toxicity over the past decade from Web of Science Core Collection database. The characteristics of the publication included publication year, journal, institution, funding, keywords, and citation counts were recorded. Co-occurrence analysis and trend topic analysis based on bibliometric analysis were conducted on keywords and citations.

Results: A total of 3,225 publications were analyzed. Number of annal publications increased over the years, with the highest number observed in 2022 (n = 475). The Journal of Ethnopharmacology published the most publications (n = 425). The most frequently used toxicity classifications in keywords were hepatotoxicity (n = 119) or drug-induced liver injury (n = 48), and nephrotoxicity (n = 40). Co-occurrence analysis revealed relatively loose connections between CHM and toxicity, and their derivatives. Keywords emerging from trend topic analysis for the past 3 years (2019-2022) included ferroptosis, NLRP3 inflammasome, machine learning, network pharmacology, traditional uses, and pharmacology.

Conclusion: Concerns about the toxicity of CHM have increased in the past decade. However, there remains insufficient studies that directly explore the intersection of CHM and toxicity. Hepatotoxicity and nephrotoxicity, as the most concerned toxicity classifications associated with CHM, warrant more in-depth investigations. Apoptosis was the most concerned toxicological mechanism. As a recent increase in attention, exploring the mechanisms of ferroptosis in nephrotoxicity and NLRP3 inflammasome in hepatotoxicity could provide valuable insights. Machine learning and network pharmacology are potential methods for future studies.

Introduction

For centuries, Chinese Herbal Medicine (CHM) has been utilized under the assumption that "natural equals safe." However, high-profile cases of toxicity—such as the kidney-damaging effects of aristolochic acids—have shattered this misconception. The challenge for modern researchers is that CHM involves complex mixtures of compounds that interact with the body in non-linear ways. While thousands of papers exist, they are often siloed, focusing on a single herb or a specific organ. There has been a lack of a "birds-eye view" to identify which toxicities are most prevalent, which molecular pathways are trending, and where the research gaps lie. This study uses bibliometric analysis to provide a quantitative roadmap, helping researchers and regulators prioritize future safety studies.

Methods and Data Sources

The authors conducted a cross-sectional bibliometric analysis, primarily using the Web of Science (WOS) Core Collection (3,225 publications) and the CNKI database (5,880 records) for comparison. Unlike a traditional systematic review that synthesizes experimental data, this study analyzes the metadata of the research itself. Key techniques included:

  • Co-occurrence Analysis: Using VOSviewer to map how frequently specific keywords (e.g., "apoptosis" and "hepatotoxicity") appear together, revealing thematic clusters.
  • Trend Topic Analysis: Utilizing the Bibliometrix R-package to identify "burst" keywords that have gained sudden popularity in recent years.
  • Citation Burst Analysis: Using CiteSpace to identify influential papers that significantly shifted the field's trajectory.

The analysis revealed a steady and significant increase in the volume of research dedicated to CHM toxicity over the last decade. Annual publications grew from 190 in 2013 to a peak of 475 in 2022, reflecting a growing global interest in the safety profiles of traditional medicines.

Figure 1: Annual growth in publications related to CHM toxicity from 2013 to 2022.

In terms of clinical focus, hepatotoxicity (liver damage, n=119) and nephrotoxicity (kidney damage, n=40) emerged as the most frequently cited toxicity classifications. The Journal of Ethnopharmacology stands as the primary venue for this research, contributing over 13% of the total global output. Furthermore, the National Natural Science Foundation of China funded nearly half (46.17%) of the global output in this niche, underscoring the regional importance of this field.

Geographic and Thematic Insights

Research is heavily centralized, with China producing over 86% of the analyzed papers. This geographic concentration highlights a potential need for broader international collaboration to validate findings across diverse genetic populations and regulatory environments.

Figure 2: Co-occurrence analysis of countries showing the dominance of China and the USA in CHM toxicity research.

Beyond geography, the study identified a "Loose Connection" Paradox: co-occurrence analysis showed relatively weak keyword links between "CHM" and "Toxicity." This suggests that while many papers discuss both, there is still a lack of integrated studies that directly explore the intersection of herbal chemistry and toxicological outcomes.

Thematically, the field is shifting from general to specific mechanisms. While apoptosis was long the dominant mechanism studied, recent data shows a pivot toward ferroptosis (iron-dependent cell death) and the NLRP3 inflammasome. Additionally, the rise of "In Silico" toxicology—marked by keywords like machine learning and network pharmacology—indicates a move toward using computational models to predict how complex multi-component formulas interact with biological systems.

Conclusion and Future Directions

While the study provides a comprehensive overview, it is limited by its focus on the WOS database, which may underrepresent high-quality research in non-English journals. Additionally, bibliometric analysis is descriptive and cannot explain the biological causes of toxicity.

Future research should prioritize "intersectional" studies that move beyond describing damage to predicting it. Specifically, exploring the mechanisms of ferroptosis in kidney injury and leveraging machine learning to screen for toxic components before they reach clinical application will be vital for the continued safe integration of CHM into modern healthcare.

Summary updated

Abstract

<b>Introduction:</b> Despite the clinical value of Chinese herbal medicine (CHM), restricted comprehension of its toxicity limits the secure and efficacious application. Previous studies primarily focused on exploring specific toxicities within CHM, without providing an overview of CHM's toxicity. The absence of a quantitative assessment of focal points renders the future research trajectory ambiguous. Therefore, this study aimed to reveal research trends and areas of concern for the past decade. <b>Methods:</b> A cross-sectional study was conducted on publications related to CHM and toxicity over the past decade from Web of Science Core Collection database. The characteristics of the publication included publication year, journal, institution, funding, keywords, and citation counts were recorded. Co-occurrence analysis and trend topic analysis based on bibliometric analysis were conducted on keywords and citations. <b>Results:</b> A total of 3,225 publications were analyzed. Number of annal publications increased over the years, with the highest number observed in 2022 (n = 475). The <i>Journal of Ethnopharmacology</i> published the most publications (n = 425). The most frequently used toxicity classifications in keywords were hepatotoxicity (n = 119) or drug-induced liver injury (n = 48), and nephrotoxicity (n = 40). Co-occurrence analysis revealed relatively loose connections between CHM and toxicity, and their derivatives. Keywords emerging from trend topic analysis for the past 3 years (2019-2022) included ferroptosis, NLRP3 inflammasome, machine learning, network pharmacology, traditional uses, and pharmacology. <b>Conclusion:</b> Concerns about the toxicity of CHM have increased in the past decade. However, there remains insufficient studies that directly explore the intersection of CHM and toxicity. Hepatotoxicity and nephrotoxicity, as the most concerned toxicity classifications associated with CHM, warrant more in-depth investigations. Apoptosis was the most concerned toxicological mechanism. As a recent increase in attention, exploring the mechanisms of ferroptosis in nephrotoxicity and NLRP3 inflammasome in hepatotoxicity could provide valuable insights. Machine learning and network pharmacology are potential methods for future studies.
BibTeX
@inproceedings{zhu2024growing,
  author = {Kexin Zhu and Min Wu and Zhi-Lin Bian and Shi-Liang Han and Liming Fang and Fengfeng Ge and Xue-Zhou Wang and Shengfang Xie},
  title = {{Growing attention on the toxicity of Chinese herbal medicine: a bibliometric analysis from 2013 to 2022}},
  booktitle = {Frontiers in Pharmacology},
  year = {2024},
  doi = {10.3389/fphar.2024.1293468},
}
Figure