Mapping the evolution of liver aging research: A bibliometric analysis

Abstract

BACKGROUND

With the increasing of the global aging population, healthy aging and prevention of age-related diseases have become increasingly important. The liver, a vital organ involved in metabolism, detoxification, digestion, and immunity, holds a pivotal role in the aging process of organisms. Although extensive research on liver aging has been carried out, no bibliometric analysis has been conducted to evaluate the scientific progress in this area.

AIM

To analyze basic knowledge, development trends, and current research frontiers in the field via bibliometric methods.

METHODS

We conducted bibliometric analyses via a range of analytical tools including Python, the bibliometrix package in R, CiteSpace, and VOSviewer. We retrieved publication data on liver aging research from the Web of Science Core Collection Database. A scientific knowledge map was constructed to display the contributions from different authors, journals, countries, institutions, as well as patterns of co-occurrence keywords and co-cited references. Additionally, gene regulation pathways associated with liver aging were analyzed via the STRING database.

RESULTS

We identified 4288 articles on liver aging, authored by 24034 contributors from 4092 institutions across 85 countries. Notably, the years 1991 and 2020 presented significant bursts in publication output. The United States led in terms of publications (n = 1008, 25.1%), citations (n = 55205), and international collaborations (multiple country publications = 214). Keywords such as “lipid metabolism”, “fatty liver disease”, “inflammation”, “liver fibrosis” and “target” were prominent, highlighting the current research hotspots. Notably, the top 64 genes, each of which appeared in at least 8 articles, were involved in pathways essential for cell survival and aging, including the phosphatidylinositol 3-kinase/protein kinase B, Forkhead box O and p53 signaling pathways.

CONCLUSION

This study highlights key areas of liver aging and offers a comprehensive overview of research trends, as well as insights into potential value for collaborative pursuits and clinical implementations.

Key Words: Aging; Liver; Bibliometric; Gene regulation; VOSviewer; CiteSpace; R language

Core Tip: In organismal aging, liver aging not only increases vulnerability to liver diseases but also enhances the susceptibility of other organs due to its central role in metabolism regulation. This is the first bibliometric analysis focus on liver aging research. Publication output has significantly surged over the past 40 years. The United States led in terms of publications, citations, and international collaborations. Recent studies have increasingly focused on “gut microbiota”, “inflammation”, “fibrosis” and “nonalcoholic fatty liver disease” in liver aging research. Gene function correlated with liver aging are involved primarily in the phosphatidylinositol 3-kinase/protein kinase B, Forkhead box O, and p53 signaling pathways.

INTRODUCTION

Aging involves a gradual deterioration of homeostatic balance at various levels, including genomic, cellular, tissue, and organismal[1,2], characterized by a progressive decline in physical and mental capacities, increasing susceptibility to diseases and mortality risk. López-Otín et al[3] have summarized the aging hallmarks, including genomic instability, loss of proteostasis, cellular senescence, etc. The liver is a vital organ involved in numerous physiological processes and organismal homeostasis, such as metabolism, detoxification, synthesis, storage, secretion, and immunity[4]. As the liver ages, it undergoes morphological and functional degeneration, including reduced blood flow, decreased hepatocyte regeneration, increased polyploidy, and diminished drug metabolism[5-7]. These changes raise the risk of liver diseases such as fatty liver, cirrhosis, liver cancer, and other hepatic disorders[7,8]. Additionally, the liver serves as a critical endocrine organ, and an imbalance in liver-derived factors (hepatokines) can lead to diseases in the nervous system, bone, heart and adipose tissue[9,10].

Recent studies employing RNA-seq and single-cell sequencing have detailed cell type-specific gene expression changes along with liver aging in mice[11,12]. Additionally, a single-nucleus transcriptomic atlas comparing young and aged primate livers has revealed an imbalance in lipid metabolism and increased expression of genes associated with chronic inflammation, both of which are linked to liver aging. Moreover, sterol regulatory element-binding protein 2, a transcription factor involved in cholesterol synthesis, has been identified as a key driver of hepatocyte aging when it is upregulated[13]. Studies in rats also have reported that the metabolism of arachidonic acid, glycerophospholipids, linoleic acid, and histidine becomes unstable with increasing liver age[14]. Liver-derived glycosylphosphatidylinositol-specific phospholipase D1 can improve age-related cognitive impairment[15]. Therefore, understanding liver aging is essential for preventing liver diseases and delaying systemic aging processes.

Bibliometric analysis, which employs mathematical and statistical methods along with visualization tools, has been widely used to explore the impacts and connections of published papers, including journals, institutions, authors, keywords, references, and countries. Bibliometric studies have been systematically assessed the literature in specific fields[16-19]. However, few bibliometric analyses have focused specifically on liver aging. In this study, we screened publications on liver aging from 1984 to 2023, and obtained 4288 relevant articles. We utilized bibliometric tools such as R, python, CiteSpace and VOSviewer to statistically analyze and visualize the knowledge mapping. Our aim is to delineate the developmental trajectory of liver aging research and provide insights for both foundational and clinical research advancements.

MATERIALS AND METHODS

Database and search strategy

The Web of Science is a prominent academic database for bibliometric analysis, accommodating over 12000 journals and providing comprehensive citation records that enable researchers to delve into historical evolution, hotspots and trends within their respective fields[20]. In this study, we conducted a bibliometric analysis of liver aging research via the Web of Science Core Collection (WoSCC). All WOSCC records (full records and cited references) were downloaded as “plain text” on February 19, 2024, to minimize possible deviations caused by database updates. These files, named “download_*.txt”, were uploaded on GitHub (https://github.com/QunhuaHan/Mapping-the-evolution-of-liver-aging-research-A-bibliometric-analysis/).

Data analysis and visualization

We conducted bibliometric analysis via various software tools, including the bibliometrix R package[21], CiteSpace[22], and VOSviewer[23]. Specifically, bibliometrix package in R (version 4.0.2) was employed to perform quantitative analysis across different fields, such as countries, institutions, authors, journals, total citations, and year of publication. In the bibliometrix package, the extraction methods are as follows: Authors from the author full name field (countries from the authors country and institutions from the authors university); year of publication from the publication year field, and citations from the time cites in WoSCC. The Python and R language code used for this research is publicly available at GitHub (https://github.com/QunhuaHan/Mapping-the-evolution-of-liver-aging-research-A-bibliometric-analysis/). We used various methods to unify keywords: (1) For words with singular and plural forms, we applied natural language processing technology and the spacy package (pretraining model: En_core_web_sm) in Python to convert plural forms to singular forms, such as “mechanisms” to “mechanism”, and “livers” to “liver”; (2) For words differing only by hyphens or slashes, such as “adult rat” and “adult-rat”, and “8-hydroxy-2’-deoxyguanosine” and “8-hydroxy-2-deoxyguanosine”, we standardized them via regular expressions in Python; and (3) For words with similar meanings but difficult for the computer to process, we manually identify and merge them, such as “nafld” and “nonalcoholic fatty liver”, “ageing” and “aging”. CiteSpace can be downloaded at https://citespace.podia.com/. The website offers both test tutorials and video guides, providing comprehensive instructions for using this software effectively. Herein, CiteSpace 6.1.R6 was used to generate a dual-map overlay for journals, analyze keywords or references with significant citation bursts, and aid in keyword clustering analysis. VOSviewer can be downloaded from its host website (https://vosviewer.com/getting-stared), which provides detailed instructions on its use. Herein, VOSviewer (1.6.20) was employed to explore the collaborative network of countries, institutions, authors, and the co-occurrence network of references. Other supportive software, such as Pajek, assisted in modifying the layout of the network map. In our study, the journal impact factor was obtained from the Journal Citation Report 2022 released by Clarivate Analytics. Overall, the use of the aforementioned software for visualization and analysis provided insights into liver aging and identified research frontiers via substantial data.

Analysis of gene functions that related to liver aging

We downloaded the human genome reference sequence (GRCh38.p14) and the mouse genome reference sequence (GRCm39) from National Center for Biotechnology Information, and extracted all the gene names via Python. Nearly 100000 genes were obtained by combining and deduplicating the human and mouse gene names. After an initial screening to remove unnamed genes without annotations, and long intergenic nonprotein coding RNAs, we matched the remaining genes with literature abstracts by using the Regular Expression of Python (regex Package), counted the number of publications, and eliminated some terms that were unlikely to be gene names, such as WAS, TG, MICE, and IMPACT. Then, the STRING database (https://cn.string-db.org/) was employed to explore insights into complex protein associations of these genes[24]. Moreover, Kyoto Encyclopedia of Genes Genomes and Gene Ontology pathway enrichment analyses were conducted to highlight the enriched signaling pathways of the top 64 genes.

RESULTS

Annual publications and growth trends in liver aging research

Figure 1 illustrated the detailed search strategy for data collection and analysis. In accordance with the search strategy, a total of 4288 articles concerning liver aging were identified from the WoSCC database between 1984 and 2023, spanning 40 years. As shown in Figure 2 and Supplementary Table 1, both the annual and the cumulative number of publications on liver aging have increased dramatically over the past three decades, with 2021 marking a peak in publications (n = 305, 7.11%), approximately 3.7 times greater than the number published in 1991. The average citations count per article was 36.78. The annual average number of publications per year was 107, with an annual growth rate of 0.6%. A total of 24034 authors from 85 countries contributed to the literature, and were published across 1148 journals worldwide. Notable publication bursts occurred in 1991 and 2020.

Figure 1 Detailed flowchart of the search strategy for screening publications. KEGG: Kyoto Encyclopedia of Genes Genomes; GO: Gene Ontology; RPYS: Reference publication year spectroscopy.

Figure 2  The annual number and the cumulative publications from 1984-2023.

Analysis of publications by country

A total of 85 countries contributed to the literature concerning liver aging. According to the number of publications, as shown in Table 1 and the country distribution map (Figure 3A), the United States (n = 1008, 25.1%) was the most prolific, followed by China (n = 820, 20.42%), Japan (n = 428, 10.66%), Italy (n = 205, 5.1%) and Spain (n = 167, 4.16%) (Table 1). The top 10 countries accounted for 76.4% of the total publications, with five from Europe, four from Asia, and one from North America. Figure 3B illustrates the extensive collaboration among the 51 most prolific countries, highlighting that greater collaboration between countries was associated with higher publication output. The United States led in terms of publications (n = 1008), citations (n = 55205), and joint publications [multiple country publications (MCP) = 214], indicating a high rate of international collaboration. Although Germany ranks seventh in total publications, it exhibited the highest proportion of international collaborations, as evidenced by an MCP ratio of 0.408. The MCP ratios of China, Japan, and India were less than 0.2, indicating limited interaction with other countries. Additionally, an analysis of the annual publication trends of the top 10 countries revealed a surge in publications on liver aging in China after 2007, whereas countries such as the United Kingdom, the United States, Japan, and Germany have consistently published papers since 1990, with steady growth (Supplementary Figure 1).

Figure 3 Analysis of publications by country and institution. A: The distribution of countries (log-transformed scale of article count); B: Network maps showing countries involved in the research on liver aging, the size of the node indicates the number of papers; C: Three-field plot of keyword, countries and journals; D: The top 10 institutions ranked by publication count; E: Network maps of 49 institutions with at least 25 articles published in liver aging research. The node size represents the number of articles.

Table 1 The top 10 most productive countries.

Rank	Country	Publications	Frequency	TC	AC	SCP	MCP	Ratio
1	United States	1008	0.251	55205	54.77	794	214	0.212
2	China	820	0.2042	16228	19.79	674	146	0.178
3	Japan	428	0.1066	14259	33.32	366	62	0.145
4	Italy	205	0.051	6551	31.96	150	55	0.268
5	Spain	167	0.0416	6774	40.56	125	42	0.251
6	Korea	153	0.0381	3826	25.01	109	44	0.288
7	Germany	152	0.0378	6511	42.84	90	62	0.408
8	United Kingdom	140	0.0349	8635	61.68	87	53	0.379
9	France	119	0.0296	9485	79.71	91	28	0.235
10	India	84	0.0209	1712	20.38	71	13	0.155

Frequency: Number of articles published per country/total number of articles published. Total number of articles published is 4016 (among the total 4288 published articles, 272 articles lack the information of countries). TC: Total citations; AC: Average citations; SCP: Single country publication; MCP: Multiple country publication; Ratio: Multiple country publication/publications.

A three-field plot was used to evaluate the contributions of various countries to liver aging research, depicting the proportions of keywords, countries, and journals with the most publications (Figure 3C). In this plot, the rectangle size represents the importance of a keyword, country or journal. According to Figure 3C, the associations between the United States, China, and Japan with the keywords listed in the left column highlight a consistent focus on liver aging. These countries have made substantial contributions to research areas such as gene expression, aging, liver function, and oxidative stress. The right column highlights influential journals that have played an essential role in advancing liver aging research. Journals such as Aging Cell, Hepatology, and Experimental Gerontology have published many papers on liver aging research.

Analysis of publications by institution

A total of 4092 institutions worldwide have conducted research related to liver aging, with the top 10 institutions contributing to 11.82% of all publications. Figure 3D and Table 2 show the top 10 institutions concerning publication number and their citation count. The University of Texas ranked first with 71 publications, followed by the Tokyo Metropolitan Institute of Gerontology with 60 publications. Among the top 10 institutions, six are located in the United States, two in China, and one in Japan.

Table 2 The top 10 institutions with the most publications.

Institution	NP	TC	AC	Country
University of Texas	71	3599	50.69	United States
Tokyo Metropolitan Institute of Gerontology	60	1868	31.13	Japan
University of Sydney	57	2693	47.25	Australia
National Institute on Aging	54	2683	49.69	United States
Harvard University	50	4133	82.66	United States
University of California, Los Angeles	44	2711	61.61	United States
University of Pittsburgh	44	2299	52.25	United States
Chinese Academy of Sciences	43	792	18.42	China
Albert Einstein College of Medicine	42	3510	83.57	United States
Shanghai Jiao Tong University	42	1220	29.05	China

NP: Number of publications; TC: Total citations; AC: Average citations.

To further explore collaboration between institutions, a co-occurrence analysis was conducted. Figure 3E shows a network of 49 institutions, with each institution publishing at least 25 papers. These institutions formed four clusters: Red (n = 18 institutions), blue (n = 13 institutions), green (n = 11 institutions) and yellow (n = 7 institutions). The red cluster predominantly comprises United States institutions led by Harvard University and the National Institute on Aging. The yellow cluster includes several Chinese research institutions, such as Shanghai Jiao Tong University, Fudan University and Chinese Academy of Sciences, indicating significant collaboration among Chinese institutions. The analysis of institutional co-occurrence revealed that collaboration between institutions is a key driver of success in liver aging research, with domestic collaboration being more prevalent than international collaboration.

Analysis of publications by author distribution

A total of 24034 authors published papers on liver aging, with an average of 7.15 authors contributing to each article. Table 3 lists the top 10 most prolific authors in liver aging research. Among these, five authors are from the United States, two from Australia, two from Japan, and one from China. Akihito Ishigami has published the most papers (n = 30), whereas David G Le Couteur, has the highest number of citations (n = 1463). The annual publication dynamics of the top 10 most prolific authors (Figure 4A) reveal that Akihito Ishigami, David G Le Couteur, Victoria C Cogger, and Rafael De Cabo were the most productive authors between 2008 and 2023, whereas Gianfranco Alpini, Heather Francis, Shannon Glaser, Lindsey Kennedy and Tanhao Zhou were the most productive authors between 2017 and 2023.

Figure 4 Authorship analysis in liver aging research. A: Annual publication trends of the top 10 most prolific authors; B: Co-authorship network map of 65 authors, each with a minimum of 10 articles in liver aging research. The node size indicates the number of articles.

Table 3 The 10 most prolific authors.

Rank	Author	Country	NP	TC	H-index	G-index	Year
1	Akihito Ishigami	Japan	30	590	16	23	2008
2	David G. Le Couteur	Australia	28	1463	16	28	2007
3	Victoria C. Cogger	Australia	26	1417	17	26	2007
4	Gianfranco Alpini	United States	25	541	14	23	2015
5	Heather Francis	United States	24	466	14	21	2017
6	Naoki Maruyama	Japan	21	1180	16	21	1992
7	Rafael De Cabo	United States	20	896	14	20	2007
8	Shannon Glaser	United States	20	388	12	19	2017
9	Lindsey Kennedy	United States	19	329	10	18	2017
10	Tian-Hao Zhou	China	19	345	10	18	2017

Year: The publication year of the first paper related to liver aging. NP: Number of publications; TC: Total citations.

Most articles are written by a small number of authors, a phenomenon that is consistent with Lotka’s law[25]. Supplementary Figure 2 shows the observed and theoretical frequencies for the total authors based on Lotka’s law, revealing that 82.21% of the authors published only one article, whereas the proportions of authors who published two, three and four articles were 10.31%, 3.43% and 1.63%, respectively. Furthermore, we conducted a co-occurrence analysis including 65 authors with more than 10 publications each. The largest cluster contains 14 authors, four of whom are among the top 10 authors, indicating that collaboration among researchers is key to a high output of published papers (Figure 4B).

Analysis of publication quantity and journal distribution

Our analysis revealed that 1148 journals published papers on liver aging research. Table 4 ranks the top 10 journals by publication count, with five publishers based in the United States, three in the United Kingdom, one in Switzerland, and one in the Netherlands. The journal Mechanism of Ageing and Development leads with the highest publication output (n = 192, 4.48%), succeeded by Experimental Gerontology (n = 132, 3.08%) and Hepatology (n = 82, 1.91%). Collectively, these top 10 journals have published 860 articles, comprising 20.06% of the total number of articles. Hepatology, which ranks third, has a high g-index and h-index, highlighting its significant impact and contributions to the field. The annual publication counts in Aging Cell, and Hepatology has steadily increased in recent years (Supplementary Figure 3).

Table 4 The top 10 journals with the most publications.

Rank	Journal	Country	JIF	H-index	G-index	TC	NP	Year
1	Mechanisms of Ageing and Development	Switzerland	5.3	42	65	6053	192	1984
2	Experimental Gerontology	United Kingdom	3.9	38	64	4761	132	1985
3	Hepatology	United States	13.5	43	82	6929	82	1985
4	Journals of Gerontology Series A-Biological Sciences and Medical Sciences	United States	5.1	33	52	2948	81	1995
5	Aging Cell	United Kingdom	7.8	33	65	4351	76	2002
6	Aging-US	United States	5.2	22	35	1397	65	2009
7	Biogerontology	Netherlands	4.5	25	41	1831	63	2001
8	Plos One	United States	3.7	28	40	1730	57	2008
9	Biochemical and Biophysical Research Communications	United States	3.1	24	41	1809	56	1987
10	Scientific Reports	United Kingdom	4.6	21	38	1566	56	2011

Year: The publication year of the first paper related to liver aging. JIF: Journal impact factor (2022); NP: Number of publications; TC: Total citations.

To visualize the distribution of citing and cited journals, we conducted a dual-map overlay analysis of journals via CiteSpace. The left side of the map denotes the citing journals, whereas the right side signifies the cited journals, with the hues of the lines indicating different disciplines. Figure 5 depicts the thematic distribution of journals, with the thickest stripe highlighting the three core citation trajectories. Specifically, the green trajectory signifies that papers on liver aging published in medicine/medical/clinical journals are frequently cited in molecular/biological/genetics journals. The yellow trajectory suggests that most papers in molecular/biological/immunology journals are also cited in both molecular/biological/genetics and health/nursing/medicine journals.

Figure 5 Dual-map overlay of journals in liver aging research. The left side represents the citing journals and the right side represents the cited journals, and the lines’ hues indicate the different disciplines.

Analysis of highly cited articles, co-cited and temporal distributions of references

The most highly cited articles offer valuable insights into the research impact of the field. Table 5 lists the top 5 articles with the highest citation frequencies. As depicted in Table 5, these articles were published between 1997 and 2018, and each article has garnered over 1000 citations. The most cited article is titled “Natural history of liver fibrosis progression in patients with chronic hepatitis C” (1997).

Table 5 The top 5 most cited articles.

Rank	Ref.	Year	Journal	Title	JIF	TC	AC
1	Poynard et al[46]	1997	Lancet	Natural history of liver fibrosis progression in patients with chronic hepatitis C. The OBSVIRC, METAVIR, CLINIVIR, and DOSVIRC groups	168.9	2444	87.29
2	Xue et al[47]	2007	Nature	Senescence and tumour clearance is triggered by p53 restoration in murine liver carcinomas	64.8	1886	104.78
3	Yoshimoto et al[48]	2013	Nature	Obesity-induced gut microbial metabolite promotes liver cancer through senescence secretome	64.8	1469	122.33
4	Krizhanovsky et al[71]	2008	Cell	Senescence of activated stellate cells limits liver fibrosis	64.5	1375	80.88
5	Estes et al[72]	2018	Hepatology	Modeling the epidemic of nonalcoholic fatty liver disease demonstrates an exponential increase in burden of disease	14	1250	178.57

JIF: Journal impact factor (2022); TC: Total citations; AC: Average citations.

A co-cited analysis via VOSviewer identified 48 co-cited references, each of which was cited more than 45 times (Figure 6A). The citation bursts reflect periods when these references were cited significantly more frequently than other periods. Using CiteSpace, we analyzed the frequency of references, and Figure 6B presents the top 30 references with the red line marks the robust burst period. The reference with the strongest citation burst (strength = 33.32, burst period = 2018-2023) is “Cellular senescence drives age-dependent hepatic steatosis”. This study suggested that cellular senescence contributes to the progression of hepatic steatosis and that elimination of senescent cells through the administration of dasatinib and quercetin could mitigate age-related hepatic steatosis, suggesting a potential therapeutic strategy[26].

Figure 6 Citation analysis in liver aging research. A: Network map of the top 48 co-cited references, each cited more than 45 times; B: Top 30 references with the strongest citation bursts, the red segment represents the period of emergence, with burst strength proportional of the segment length; C: Reference publication year spectroscopy (RPYS) revealing three citation peaks in 1990, 2000, and 2013.

Reference publication year spectroscopy (RPYS) was utilized to analyze the temporal distribution characteristics of references, trace the historical origins of research fields and quantify their influence on current research[27]. Figure 6C depicts the raw citation frequency distribution of the cited references (black line) and the deviation of the raw citation frequency from the five-year median average (red fluctuation curve)[28]. The peak years-1990, 2000 and 2013-exhibit higher relative citation frequencies compared than neighboring years do, highlighting significant citation peaks (Figure 6C, Table 6).

Table 6 Details of the historical literature on liver aging.

Ref.	Title	Journal	Cited frequency
Ji et al[49], 1990	Alteration of antioxidant enzymes with aging in rat skeletal muscle and liver	American Journal of Physiology	197
Michael et al[50], 2000	Loss of insulin signaling in hepatocytes leads to severe insulin resistance and progressive hepatic dysfunction	Molecular Cell	971
Yoshimoto et al[48], 2013	Obesity-induced gut microbial metabolite promotes liver cancer through senescence secretome	Nature	1469

Analysis of keywords and thematic term trends

Keywords in publications encapsulate the core content. We extracted a total of 8933 keywords from 4288 papers. Using CiteSpace, we constructed clusters of keywords and identified the strongest citation bursts from 1990 to 2023 to investigate the evolution trends of keywords and potential research hotspots. Figure 7 and Supplementary Table 2 highlight the top 5 most influential keywords: “gene expression”, “liver”, “oxidative stress”, “aging”, and “mouse”. The keywords formed 11 clusters, with the order of the cluster ID numbers from small to large indicating the evolutionary trajectory of the research (Figure 8A and Supplementary Table 3). As depicted in Figure 8B, research trends on liver aging before 2010 focused on “lipid peroxidation”, “oxidative damage”, “liver”, “age-related change” and “caloric restriction”. In more recent years, the keywords with the strongest citation bursts have included “lipid metabolism”, “fatty liver disease”, “inflammation”, “fibrosis”, and “target”, which reflect current and emerging research frontiers.

Figure 7  Co-occurrence network of keywords via CiteSpace.

Figure 8 Keyword analysis in liver aging research. A: Cluster map of keywords via CiteSpace; B: Top 25 keywords with robust citation bursts, the red segment indicates the period of emergence, with the burst strength proportional to the segment length.

Multiple correspondence analysis is an exploratory multivariate method applied to the graphical and numerical analysis of multivariate categorical data[21]. Figure 9A shows a conceptual structure map of keywords analyzed via multiple correspondence analysis. The keywords are classified into three main clusters. The closer the triangles within each cluster are, the more similar their keyword distributions, indicating a greater frequency of co-occurrence in the articles. Additionally, the proximity of a keyword to the central point within each cluster signifies its prominence in the field. Figure 9B visualizes the timeline of research trends and highlights emerging topics in liver aging, such as “gut microbiota”, “inflammation”, “obesity”, “fatty liver”, and “molecular mechanism”, to explore their impact on underlying mechanisms and clinical outcomes.

Figure 9 Research trends in liver aging research. A: Conceptual structure map of keyword based on multiple correspondence analysis; B: Timeline of the research trends in liver aging. MCA: Multiple correspondence analysis.

Analysis of gene functions related to liver aging

We extracted genes mentioned in the publication abstracts and analyzed the number of articles associated with each gene. In total, we identified 1103 genes, each mentioned in a number of articles ranging from 1 to 145. Supplementary Table 4 lists the top 64 genes, each of which appeared in at least 8 articles. Figure 10A shows the top 10 genes, each mentioned in at least 32 articles. Tumor necrosis factor is the most frequently mentioned gene (n = 145 articles), followed by sirtuin 1 (n = 77 articles) and mechanistic target of rapamycin (n = 56 articles). Other notable genes include BAX, CD4, hepatocyte growth factor, signal transducer and activator of transcription 3, JUN, MYC, and peroxisome proliferator-activated receptor gamma coactivator 1. The pathways of the top 64 genes were analyzed via the STRING database. Kyoto Encyclopedia of Genes Genomes pathway enrichment analysis revealed that these genes are involved in the endocrine resistance, cellular senescence, phosphatidylinositol 3-kinase (PI3K)/protein kinase B (Akt) signaling pathway, the Forkhead box O (FoxO) signaling pathway, p53 signaling pathways, and the longevity-regulating pathway (Figure 10B). Gene Ontology enrichment analysis indicated that these genes are involved primarily in biological processes related to metabolic processes, the response to stress and stimuli, and the regulation of cell death (Figure 10C). This comprehensive analysis provides valuable insights into genetic associations in liver aging.

Figure 10  Gene analysis in liver aging research. A: Top 20 genes reported in publication abstracts; B: Kyoto Encyclopedia of Genes Genomes pathway enrichment analysis of the top 64 genes; C: Gene Ontology enrichment analysis of the top 64 genes. BP: Biological process; MF: Molecular function; CC: Cellular component; TNF: Tumor necrosis factor; SIRT1: Sirtuin 1; mTOR: Mechanistic target of rapamycin; HGF: Hepatocyte growth factor; STAT3: Signal transducer and activator of transcription 3; PGC: Peroxisome proliferator-activated receptor gamma coactivator 1; PI3K: Phosphatidylinositol 3-kinase; Akt: Protein kinase B; FoxO: Forkhead box O; PD-L1: Programmed death-ligand 1; PD-1: Programmed cell death protein-1.

DISCUSSION

The liver, an organ that undergoes structural and functional changes with age, is more susceptible to impaired responsiveness to hepatic injury and an increased incidence of liver disease in elderly individuals[29,30]. In this study, we conducted a systematic analysis and bibliometric visualization of liver aging research by screening the WOSCC database via Python, Bibliometrix R software, CiteSpace, VOSviewer, and Pejek. A total of 4288 articles published across 1148 journals from 85 countries between 1984 and 2023 were identified. As shown in Figure 2, two pivotal moments in liver aging research literature output occurred in 1991 and 2020. The launch of the Human Genome Program in the United States in 1990 may have stimulated the surge in publications in 1991[31]. Similarly, the impact of the coronavirus disease 2019 pandemic on aging organs, including the liver, may have prompted extensive research on liver aging, contributing to the second rapid growth in 2020.

Among the top ten most prolific authors, Akihito Ishigami ranks first, with 30 publications. His research focused primarily on hepatic senescence, with a particular focus on the senescence marker protein-30 (SMP30) and its role in age-related changes. For example, one study investigated the role of SMP30 in the progression of nonalcoholic fatty liver disease (NAFLD) by analyzing liver biopsy tissues and blood samples from NAFLD patients categorized by different NAFLD activity scores[32]. This study revealed a significant inverse relationship between certain serum lipoprotein fractions (very small low-density lipoprotein and large very low-density lipoprotein levels) and hepatic SMP30 levels. Another study revealed that small mothers against decapentaplegic 3 exacerbates hepatic injury by inhibiting the expression of antioxidant proteins such as SMP30, glutathione S-transferases, and selenium-binding proteins[33]. David G Le Couteur and Victoria C Cogger who ranked second and third, with 28 and 26 publications, respectively, primarily investigated the effects of organismal aging on hepatocytes[34-36], liver sinusoidal endothelial cells[37-39], and telomere length[40]. Gianfranco Alpini ranks fourth with 25 publications, which have focused mainly on the mechanisms of biliary liver diseases[41-44]. For example, he analyzed the effects of aging on liver diseases from the perspectives of bile ducts and blood vessels[41]. In summary, these scientists are actively advancing the field of liver aging research from diverse perspectives at different times.

The most highly cited publications are generally considered to have high academic impact and represent fundamental themes in the research field[45]. Citation counts serve as a valuable metric for identifying research hotspots. In our study, the 5 most frequently cited articles were published between 1997 and 2018 (Table 5). Collectively, these publications demonstrate a strong link between liver aging, cancer and immunity. The most highly cited article was published in 1997 in Lancet (n = 2444 citations). The study reported that alcohol consumption, age at infection, and male sex were strongly correlated with fibrosis progression among individuals with chronic hepatitis C[46]. The second most cited article was published in 2007 in Nature (n = 1886 citations). The paper demonstrated that loss of p53 might be an essential regulatory mechanism for the maintenance of aggressive carcinomas, and p53 can induce cellular senescence and upregulate inflammatory cytokines thus effectively limiting tumor growth[47]. The third most cited article (n = 1469 citations), which was published in Nature in 2013, demonstrated that the senescence-associated secretory phenotype contributes significantly to the promotion of obesity-related hepatocellular carcinoma development in mice. Genetic obesity or dietary factors can induce changes in the gut microbiota, leading to elevated level of deoxycholic acid, while inhibiting deoxycholic acid production or decreasing the abundance of gut bacteria was shown to efficiently prevent hepatocellular carcinoma development in obese mice[48].

Scientists such as Marx and Bornmann[27], who developed the RPYS, have noted that academic research often prioritizes citations from recent literature, with only a minimal proportion referencing important or classical papers from earlier periods. Therefore, according to RPYS, peak years preceding the emergence of a particular field are likely influenced by a single highly cited piece of literature, which can be considered the historical root of the field. Our analysis, which spans the cited literature from 1761 to 2023 (Supplementary Figure 4), identified three citation peaks within the field of liver aging between 1984 and 2023, occurring in 1990, 2000, and 2013 (Figure 6C). The paper, which was published in 1990 and titled “Alteration of antioxidant enzymes with aging in rat skeletal muscle and liver”[49], was cited 197 times. This study examined the impact of aging on antioxidant enzyme activities and lipid peroxidation in rat liver and muscle, highlighting the significance in oxidative stress and its correlation with age-related changes. In a 2000 study titled “Loss of insulin signaling in hepatocytes leads to severe insulin resistance and progressive hepatic dysfunction”[50], which has been cited 971 times, researchers explored the metabolic consequences of disrupting insulin signaling specifically in liver cells. These researchers developed liver-specific insulin receptor knockout mice, which displayed pronounced metabolic disturbances, including severe glucose intolerance and significant insulin resistance. Interestingly, the metabolic phenotype becomes less severe with age in liver-specific insulin receptor knockout mice. The 2013 paper, “Obesity-induced gut microbial metabolite promotes liver cancer through senescence secretome”[48], with 1469 citations and ranking third in overall citation frequency (Table 6), illustrates the significant influence of the gut microbiota and senescence-associated secretory phenotype on obesity-related liver cancer. These three studies provide the substantial contributions to research on topics such as “lipid peroxidation”, “hepatic metabolic imbalance”, and “gut microbiota” in the context of liver aging.

Keywords reflect the core content of articles and their frequency represents their impact and the research frontiers in a specific field. To identify new hotspots in liver aging research, we utilized CiteSpace for keyword network and cluster analysis. The most frequently appearing keywords were mainly “gene expression”, “oxidative stress”, “metabolism”, and “caloric restriction” (Figure 7, Supplementary Table 2). Dietary restriction may benefit liver function by increasing the vitamin A content in liver tissue during aging[51]. Another study showed that dietary restriction could significantly reduce the accelerated aging caused by increased oxidative stress and damage[52]. In the cluster analysis, the overall structural characteristics of the network can be assessed by the index of modularity (Q value) and the mean silhouette score (S value). Specifically, a Q value greater than 0.3 signifies a notable overall structure, whereas an S value above 0.5 denotes reasonable clustering, and a value above 0.7 implies convincing clustering[53]. As shown in Figure 8A, our analysis revealed a Q value of 0.3607, and a mean S value of 0.6649, demonstrating that the clustering structure was statistically significant, and that the analysis results were robust. The evolution path of keywords in the clusters was inferred from the order of the cluster ID numbers, with a smaller number indicating closer proximity to the core research topic[54]. The three most important clusters were #0 lipid peroxidation, #1 metabolic syndrome, and #2 cellular senescence. Among the keywords with the strongest burst citations, lipid peroxidation had a citation burst over a period of 19 years (Figure 8A). Lipid peroxidation, one of the primary results of free radical-mediated damage, directly disrupts cell membrane integrity and function and generates various secondary products that accelerate aging[55]. Lipid peroxidation can also lead to DNA damage and protein oxidation, affecting gene expression and cellular activity and ultimately contributing to cellular senescence[56]. The liver, which relies heavily on oxidative metabolism for energy, is particularly vulnerable to oxidative stress. Insufficient synthesis of antioxidant enzymes can impair the mitochondrial respiratory complex in the liver, leading to high levels of reactive oxygen species in mitochondria. This increase promotes oxidative damage to mitochondrial DNA, triggering a series of cellular damage processes, including lipid peroxidation, which can cause inflammation, necrosis, and fibrosis in fatty liver[57]. Studies in both humans and animal models have reported a close correlation between the degree of lipid peroxidation and the severity of fatty liver. MacDonald et al[58] analyzed liver biopsies obtained from patients with fatty liver disease and reported that lipid peroxidation occurred in and around fat-laden hepatocytes, particularly in acinar zone 3.

The gut microbiota has emerged as a research hotspot in recent years based on the timeline of research trends (Figure 9B). Nobel Laureate Elie Metchnikoff, as early as 1907, proposed that aging reflects the toxicity of products from the body’s gut microbiota and theorized that modulating the gut microbiota could extend life[59]. The liver and the intestine are integral components of the digestive system and are characterized by mutual interactions involving both organs. This intricate communication and interaction between two organs collectively form what is known as the gut-liver axis[60]. Dysfunctions such as intestinal barrier impairment or bacterial translocation due to pathogens and their metabolites can disrupt this axis, triggering hepatic inflammation, which in turn accelerates the onset and progression of liver diseases[61]. Studies have reported dysbiosis of the gut microbiota in age-related metabolic liver diseases, such as type 2 diabetes and NAFLD[62]. Furthermore, current research has focused on inflammation and fatty liver disease. Liver aging is correlated with an elevated necroptosis rate in hepatocytes and macrophages, a process that drives chronic inflammation and subsequent liver fibrosis[63]. Additionally, metabolic imbalances in an aging liver lead to reduced liver function and heightened vulnerability to chronic liver diseases, including fatty liver, obesity, and liver cancer.

Through pathway analysis of genes, we found that among the top 64 genes ranked by the number of publications, the enriched pathways included the PI3K/Akt and the FoxO signaling pathway (Figure 10). These two pathways are interconnected and play important roles in cell survival and aging. The PI3K/Akt signaling pathway responds to extracellular signals, triggering a cascade of intracellular reactions, participating in cell proliferation, metabolic regulation, epigenetic alterations, and mitochondrial function. It also plays a significant role in age-related liver diseases[64]. The PI3K/Akt signaling pathway is engaged in regulating the balance of glucose and lipid metabolism in the liver. In the aged liver, lipid accumulation increases, which is closely associated with insulin resistance and the development of NAFLD[65]. In patients with NAFLD, excessive free fatty acids result in elevated hepatic glucose production and de novo lipogenesis in the liver, which impair pancreatic β-cell function and insulin secretion. This impairs the PI3K/Akt pathway, causing insulin resistance. Insulin resistance, in turn, exacerbates damage to the PI3K/Akt signaling pathway, creating a vicious cycle that further accelerates the progression of NAFLD[66]. FOXO is an important downstream target of the PI3K/Akt signaling pathway, and activation of the pathway inhibits the FOXO pathway. The FoxO pathway is involved in glucose metabolism, cell cycle arrest, apoptosis, oxidative stress resistance, and longevity[67,68]. Liver-specific deletion of FoxO1 in mice can reduce hepatic glucose production[69]. In aged mice, FoxO1 activity significantly increases in the liver. Inhibiting FoxO1 can significantly alleviate glucose intolerance, hepatic steatosis, and systemic inflammation in aging mice[70].

Limitations

This bibliometric study has several limitations. First, the analysis is confined to the literature sourced exclusively from the WoSCC database, and may omit valuable studies available in other databases such as PubMed, Scopus, Embase, Dimensions, and Ovid. Second, the study was restricted to English-language literature, thereby introducing a potential language bias by excluding studies published in other languages. Finally, the objective of this study is to delineate the research progress in liver aging research. While bibliometric analysis based on the available literature reflects current research trends and focal points concerning liver aging to some extent, a more extensive dataset is needed for a comprehensive future analysis. Nonetheless, this study is expected to assist researchers in gaining a deeper understanding of the developmental trends and forefront issues within the field of liver aging.

CONCLUSION

Our study provides the first bibliometric analysis of liver aging research, spanning the last four decades. We identified significant contributions from various countries, authors, journals, institutions, keywords, and references, and analyzed gene regulation pathways that related to liver aging. In addition, we synthesized a comprehensive model of liver aging based on the data in this study (Figure 11), which integrates various aspects of liver aging. In summary, this comprehensive overview provides valuable insights into the trends and frontiers of liver aging research, offering a valuable resource for researchers and clinicians in the field.

Figure 11  Holistic model in liver aging research. The figure illustrates the stimuli of liver aging, cellular, structural and functional changes caused by liver aging, and interventions and targets for delaying liver aging. PQQ: Pyrroloquinoline quinone; NMN: Nicotinamide mononucleotide; FABP4: Fatty acid binding protein 4; RAC1, Ras-related C3 botulinum substrate 1; IGF2: Insulin-like growth factor 2; FOXO1: Forkhead box O1; CISD2: CDGSH iron–sulfur domain-containing protein 2; PCSK9: Proprotein convertase subtilisin/kexin type 9; MANF: Mesencephalic-astrocyte-derived neurotrophic factor; SIRT6: Sirtuin 6; SLC13A5: Solute carrier family 13 member 5; HO-1: Heme oxygenase 1; ICAM-1: Intercellular cell adhesion molecule-1; LSECs: Liver sinusoidal endothelial cells; HSCs: Hepatic stellate cells; α-SMA: Alpha-smooth muscle; ROS: Reactive oxygen species; KCs: Kupffer cells; DNMT1: DNA (cytosine-5-)- methyltransferase 1; D + Q: Dasatinib (D) and quercetin (Q); CD32b: Fc gamma receptor IIb; PPARα: Peroxisome proliferator activated receptor alpha; SASP: Senescence-associated secretory phenotype; IL: Interleukin.