The influence of shift work: A bibliometric analysis of research progress and frontiers on health effects

ABSTRACT Shift work has been found to disrupt the circadian system, leading to negative health effects. The objective of this study was to assess the progress and frontiers in research on the health-related influence of shift work. The study analyzed 3,696 data points from Web of Science, using the bibliometric software CiteSpace to visualize and analyze the field. The results showed a steady increase in annual publications, particularly in the last 5 years, with a rapid increase in publications from China. The United States contributed the most to the number of publications and worldwide collaborations. The most prolific institution and author were the Brigham and Women’s Hospital and Professor Bjorn Bjorvatn, respectively. The Journal of Chronobiology International ranked at the top and focused primarily on shift worker research. In the first decade of study, the primary focus was on the associations between shift work and cardiovascular disease and metabolic disorders. Over time, research on the health effects of shift work has expanded to include cancer and mental health, with subsequent studies investigating molecular mechanisms. This study provides a comprehensive and intuitive analysis of the negative health impacts of shift work. It highlights existing research hotspots and provides a roadmap for future studies. Further research is needed to explore the adverse health consequences and related mechanisms of shift work exposure, as well as interventions to mitigate its health effects.


Introduction
With the ongoing industrialization and economic development worldwide, along with the growing demand for flexibility and productivity in the workforce, shift work has become increasingly prevalent.According to a survey conducted by Yong et al. (2017), approximately a quarter of the workforce is involved in shift work, particularly in industries that require round-the-clock operations like healthcare, public safety, and transportation.Despite the societal importance of shift work, it has been recognized as an occupational hazard (Rivera et al. 2020).leading to significant health effects and a decline in quality of life.Several epidemiological studies have demonstrated that shift work and long work hours can result in acute and chronic health impacts, including cardiovascular and metabolic diseases, fatigue, sleep disturbances, mental disorders, and an increased risk of stroke (Cheng and Drake 2019;Pallesen et al. 2021).
As public awareness of the adverse health consequences of shift work has increased, there has been a rise in epidemiological studies on shift work-related issues (Barul et al. 2019;Kang et al. 2020).However, the large number of qualitative and quantitative studies makes it challenging for researchers to gain a comprehensive understanding of the current situation and developments from a single perspective.Moreover, there are significant inconsistencies in the findings, and a lack of quantitative research that examines the existing literature.Some scholars have attempted to address these inconsistencies through meta-analysis (Khosravipour et al. 2021;Wei et al. 2021), but the debate remains superficial and subjective.To overcome these limitations, bibliometric analysis, an emerging technique that employs statistical methods and visualization, can provide researchers with qualitative and quantitative insights into the literature, allowing them to identify current trends and frontiers in the field (Pasin and Pasin 2021).While bibliometric analysis has been widely used in various biomedical domains to explore important features and inform disease treatment and clinical guidelines (Ma et al. 2021), there have been no bibliometric analyses specifically focused on the negative health implications of shift work.Therefore, statistical approaches must be employed to examine this subject from a macro perspective.
In this context, the Web of Science database was utilized to search for relevant publications pertaining to the effects of shift work on health between 1997 to August 1, 2022.A bibliometric approach was utilized to investigate, assess, and summarize the current academic authorities and results in the field, as well as to identify hotspots and future directions in order to provide researchers with valuable recommendations.

Data sources and retrieval methods
The Science Citation Index Expanded (SCI EXPANDED) and Social Sciences Citation Index (SSCI) of the Web of Science Core Collection (WoSCC) database was consulted for information about the health impacts of shift work, which can be accessed at: https://www.webofscience.com/wos/woscc/basic-search.These editions cover the years from January 1, 1997 to August 1, 2022.The search query was built mainly on MeSH terms to make sure that the retrieved documents are comprehensive enough to cover the related literatures on work shifts to health effects.Following previous studies, the search strategy was set as the following: (TS = (health OR illness OR disease)) AND (TS = ("shift work" OR "shift working" OR "shiftwork" OR "night work" OR "night shift work" OR "rotating shift")).The detailed retrieval strategy and results was shown in supplementary table S1.Before performing a formal search, we conducted a broad search and read the titles and abstracts of a significant number of relevant publications focusing on shift work and health.English papers with articles and reviews document types were selected.As a result, we only selected articles and reviews and 3696 documents were found.We retrieved information related to the titles, authors, citations, journals, institutions, references and keywords of these publications.

Bibliometric analysis methods
This study used Microsoft Excel 2019 to analyze patterns in the number of articles published annually and to forecast publication growth trends.Bibliometrics software CiteSpace 5.8.R3 (Chen et al. 2012), a significant instrument for displaying new trends and performing cooperation network analysis in the scientific literature was utilized.CiteSpace software is used to analyze institutional cooperation networks, journals and references co-citation analysis, and keywords co-occurrence analysis.The parameters of the software were set as follows: Time slicing: January 1,1997-August 1, 2022, with 1 year per slice; Selection criteria: 50 most common items from each slice; Pruning: Pathfinder, pruning slice network, and the merged network.In addition, we utilized the literature knowledge program VOSviewer 1.6.18,which is based on the visualization of similarities technology, to show the collaborations between authors of a list of publications.Tableau Public 2021.3.Ink and Gephi are used to illustrate the global distribution of scientific publications and international collaboration networks.RStudio (2022.02.3 + 492) was used to perform the time-dependent heat map of keywords.

Visualization and analysis on trend of publication outputs and citation rate
Worldwide publication growth trends over the period 1997.01-2022.08was shown Figure 1A.There was a total of 105,727 citations to the 3,696 articles on shift-related health effects, with an average of 28.6 citations per article and an h-index of 136.From 1997 to 2007, the growth trend was consistent and sluggish; from 2008 to 2017, the number of publications climbed substantially each year, indicating that the association between work shift and harmful health impacts began to attract the attention of worldwide experts and scholars.The production accounted for 48.5% of the total number of publications from 2018 to 2022, and the number of articles published reached 462 in 2020, 66 times that in 1997.The potential growth in annual publications can be predicted by the fitting curve y = 12.32e 0.1471x , R 2 = 0.979.The expected number of publications in 2022 is approximately 380.

Visualization and analysis of countries/regions and institutions
On the basis of a country-by-country analysis of a collaborative network, the publications on the investigation of health effect by work shift came from 100 nations or regions.Figure 1B depicts the global distribution of the number of publications.The top 10 countries with the largest number of published papers are shown in Supplementary table S2.The United States, Australia, and China published the most articles in the globe, with 1040 articles (28.13%), 320 articles (8.66%) and 315 articles (8.52%) respectively.In 1997, the United States was the first country to publish an article in this topic, followed by China in 2003.Before 2017, the number of articles by Chinese was relatively low, and it started to expand rapidly in 2020.According to the centrality of published papers, the United States ranks top in terms of communication with other countries in this area of research with a centrality of 0.39.The cooperative interaction between countries is presented in Figure 1C.
3, 347 nodes and 482 linkages represent the published studies on the health impacts of shift work by the leading research institutes (Figure 2A).The node size represents the number of articles published by institution and the thickness of connection line represents the degree of collaboration between institutions.The top 3 institutions are Brigham and Women's Hospital published the most articles (127), Harvard University (87) and Harvard Medical School (81) (Supplementary table S2).In terms of the centrality, Cent Queensland University (0.44), Karolinska Institute (0.35) and Brigham and Women's Hospital (0.29), were more central and authoritative in the collaborative network.The intensity of the burst showed that these institutions had great power for a short period of time in the region, such as Harvard University (burst = 21.89),Harvard Medical School (burst = 9.51), and Chiba University (burst = 6.85), which had the strongest bursts among these institutes (Supplementary Figure S1).

Authors and co-cited authors
There were 2,618 authors participating in the publication of the literature on the health effects of shift work, of which 321 authors published more than 5 articles and were included in this analysis.The VOSviewer software revealed collaborative networks among principal authors, generating 17 clusters and 1065 links (Figure 2B).There were active collaborations among these authors in the same cluster, including Bjorn Bjorvatn's partnership with Stale Pallesen and Siri Waage, Siri Waage's cooperation with Stale Pallesen, and Mikko Harma's collaboration with Sampsa Puttonen.
In addition, Table 1 lists  and Celine Vetter (0.07) had the highest centrality, suggesting that their work has a significant influence on the area of health consequences of shift work.150 authors out of 64 152 were quoted more than 100 times, while seven were cited more than 300 times (Table 1).Akerstedt T (824 citations) is the most frequently cited author, and 1997 is the earliest year of publication, followed by Costa G (617 citations).Reviews (11.401).The dual-map overlay of journals show the distribution of the relationship between journals and the colored paths indicating the citation relationship.A green path shows that the papers published in Medicine/Medical/Clinical are frequently cited by Environmental/Toxicology/Nutrition in Figure 3.

Cited documents and co-cited reference
There are 100,593 articles that were visualized by CiteSpace.The details of top 100 publications ranked by total number of citations are listed in supplementary table S3.The total number of citations for the top 100 most-cited articles ranged from 153 to 747, with an average of 6.16 to 83.The United States dominated the field with 40 of the top 100 publications and 11 963 citations.Literature citation research reveals that the cited frequency range for the literature with the top ten cited frequencies is between 482 and 747 (Supplementary table S4)."Sleep Health: Can We Define It? Does It Matter?" (Buysse 2014) was cited 747 times; "Rotating Night -Shifts and Risk of Breast Cancer in Women Participating in the Nurses' Health Study" (Schernhammer et al. 2001) was cited 746 times; and "Health disorders of shift workers" (Knutsson 2003) was cited 733 times."The dark side of light at night: physiological, epidemiological, and ecological consequences" (Navara and Nelson 2007), which was cited 482 times, was at the bottom of the list.
The co-citation reference demonstrated the authority of the research in this subject and the author's contributions.The top 10 co-cited references were cited at least 63 times, with a maximum of 150 co-cited references (supplementary table S5).The most frequently co-cited reference was a review by Kecklund G, which published in the British Medical Journal (BMJ) in 2016, entitled "Health consequences of shift work and insufficient sleep" (Kecklund and Axelsson 2016), followed by an article entitled "Shift work and chronic disease: the epidemiological evidence" (Wang et al. 2011).Reference co-citation clustering with a timeline view can visualize research topics and hotspots.As shown in supplementary figure S2, a combined reference cocitation network with 1,516 nodes and 7,283 links was constructed for clustering.The top 5 clusters ranked by cluster size were "cross-sectional study," "benchmark duration," "circadian clock," "sleep complaint," and "health assessment."Supplementary figure S3 listed the

Visualization and analysis on keywords co-occurrence
Keywords are the research themes and core contents of the literature.Through keyword co-occurrence analysis, it is feasible to summarize the distribution and evolution of hot spots in the health effects resulting from shift work.The network map of the top 50 frequency keywords was shown in Figure 4A.Co-occurrence analysis of the top 30 keywords with high frequency, centrality from 1997 to 2008 was presented supplementary table S6.Shift work (frequency = 1068) first was brought to the attention of scholars in 1998, followed by health (frequency = 1057), risk (frequency = 592) and sleep (frequency = 431), indicating that these fields were the most prevalent domains in which shift work was discussed.Despite "shift work" (centrality = 0.07) and "health" (centrality = 0.08), other keywords such as "melatonin" (centrality = 0.06), "fatigue" (centrality = 0.05), and "cardiovascular disease" (centrality = 0.04) have a larger centrality than others, indicating that they play a significant role in the effects of shift work on health.Then, we conducted cluster analysis on high-frequency keywords using CiteSpace.
The log-likelihood ratio (LLR) algorithm was used to cluster the keywords into 16 specific clusters, as shown in Figure 4B.The clustering structure was considered significant (Modularity Q > 0.3) and the outcome plausible (mean silhouette > 0.5).Supplementary table S7 provides more details.The five largest clusters were identified as "depression," "sleep," "obesity," "risk," and "insomnia."Cluster #0, focused on "depression," was the primary cluster, with researchers exploring the relationship between shift work and depression.Numerous studies have found a connection between night shift employment (frequency = 351) and the incidence of endometrial cancer and breast cancer in women (frequency = 159).
Cluster #1 and #4, labeled "insomnia," encompassed sleep problems, sleep patterns, and alertness, indicating that recent research has primarily focused on the relationship between shift work and sleep patterns.Cluster #2, "obesity," examined the impact of shift work on abnormal Body Mass Index (BMI).Permanent and irregular night shift work was found to be associated with overweight or abdominal obesity.Cluster #3, "risk," explored shift work as a health risk factor.Multiple studies have shown that working shifts increases the risk of metabolic disorders, cardiovascular diseases, and mortality.Due to the fact that the space-time elements of keywords are based on the interaction and mutation relationship between keywords in a particular field, this method is useful for exploring the evolution track and stage characteristics of the research field.As depicted in Figure 4C, a heat-map was generated based on the frequency of the keywords in each year, which is helpful for examining the evolution trajectory and stage characteristics of this field.The keywords appearing at high frequency are "cardiovascular disease," "breast cancer," "sleep duration," "metabolic syndrome," "melatonin," "obesity," and "depression" in the past 5 years.This indicates that the majority of research on the harmful health impacts of shift work has been conducted in these topics.Figure 4D is a keyword-based timeline viewer that visually illustrates the evolution of trending topics over time.From 2000 to 2010, the majority of research focused on the relationship between shift work and disease and menstrual cycles, with "coronary heart disease," "myocardial infarction," "accidents," "menstrual cycle," "gestational age," and "egg" as the main keywords.From 2011 to 2022, the research is mainly concerned about the relationship between shift work and cancer, mental state, metabolic and quality.The main keywords are "breast cancer," "stress," "mental health," "fatigue risk," "sleep quality" and "sleep quality index."China has given close attention to this field of study in recent years, as evidenced by the fact that the frequency of the term "China" peaked in 2020.

Burst detection of keywords
Keywords with burst detection can determine whether a research field has become a hotspot within a specific time period.The top 25 keywords ranked by the citation burst strength were shown in supplementary Figure S4.The strongest citation burst of keyword is "alertness, strength, 14, time span 1997-2011."Other keywords with stronger citation burst, such as "breast cancer, strength 13.67, time span 2004-2014," "coronary heart disease, strength, 9.93, time span 1999-2007," and "myocardial infarction, strength, 9.84, time span 1999-2010."It is inferred that the increased risk of cardiovascular disease and cancer associated with shift work has been the focus of research.Keywords like "United States, strength 5.47, time span 2014-2018" indicated that the United States is one of the major contributors to depression research, which is consistent with the results of national, research and journal analyses.Furthermore, in the beginning of the past decade, keywords such as "circadian disruption, strength 5.93, time span 2013-2017" and "chronotype, strength 5.28, time span 2017-2022" indicated that the effects of shift work on sleep patterns have been the focus of research in recent years.We can therefore deduce that chronotype is an emerging focus of research in this field.

Discussion
This study conducted a bibliometric analysis of the health implications of shift work using the visualization software CiteSpace.The analysis included a scientific assessment of 3,696 publications, which revealed that research on the health impacts of shift work has been steadily expanding from 1997.01 to 2022.08, indicating a global concern about these effects.The study found that the United States had the highest number of publications and was the earliest to publish on this topic.Additionally, the United States, Australia, and China emerged as dominant powers in shift work research and have established partnerships with countries such as the United Kingdom, Canada, South Korea, and Japan.Professor Bjorn Bjorvatn of the University of Bergen and Professor Akerstedt T of Stockholm University were identified as the most published and influential authors in this field, making significant contributions to the discipline.Overall, extensive collaboration among key countries, institutions, and authors has facilitated the development and formation of strong partnerships.
The majority of journals have an average Impact Factor (IF) of less than 10, indicating the difficulty in publishing relevant articles in high-impact journals.However, among the top 10 journals based on co-citation, 5 are classified as Q1 journals, with the Lancet having the highest IF of 202.731.This indicates the growing significance of research on the health effects of shift work, as high-quality journals play a crucial role in communication and dissemination.Our findings suggest that highly cited papers in this field focus on important epidemiological investigations and experimental results regarding the health effects of shift work.These papers explore the relationship between shift work and the risk of breast and colon cancer, the association between shift work and metabolic syndrome, as well as the effects of shift work on sleep health (Chang et al. 2015;Karlsson et al. 2001;Schernhammer et al. 2001Schernhammer et al. , 2003)).Among the co-cited literature on the health effects of shift work, the most cited review titled "Health consequences of shift work and insufficient sleep" was published by Kecklund G et al. in 2016 in BMJ-British Medical Journal, with 150 citations (Kecklund andAxelsson 2016).This review provides a comprehensive summary of the health effects of shift work, including its association with sleep deprivation, chronic disease, and accidents.It also describes the possible mechanisms through which shift work may contribute to accidents and adverse health consequences.
In addition, this study utilized three indicators, namely "keyword frequency," "keyword centrality," and "keyword explosion," to investigate the research hotspots regarding the impact of shift work on health.High-frequency keywords were employed to better reflect the current interest in topics such as shift work, health, risk, sleep, and stress.Keywords with high centrality indicated phrases associated with shift work, health, melatonin, fatigue, cardiovascular disease, and other relevant factors, which were crucial in this field of research.The keywords that have been frequently studied in the past 5 years include "cardiovascular disease," "breast cancer," "sleep duration," "metabolic syndrome," "melatonin," "obesity," and "depression," suggesting that these areas have been the focal points of recent research on the impact of shift work on health.In a recent large cohort study with a 10-year follow-up period, it was confirmed that rotational work is linked to an increased risk of atrial fibrillation and coronary heart disease, particularly among women with longer exposure to shift work (≥10 years) (Wang et al. 2021).Over time, an accumulating body of research has consistently demonstrated the detrimental effects of shift work on health.In the past decade, research has primarily focused on examining the association between shift work and cardiovascular and metabolic diseases through the analysis of high-impact keywords.Relevant epidemiological studies have shown that long-term shift work is associated with an elevated risk of cardiovascular disease, metabolic syndrome, and diabetes in women (Pan et al. 2011;Puttonen et al. 2012).The possible underlying mechanism was experimentally verified, indicating that the circadian rhythm disorder caused by shift work plays a significant role (Scheer et al. 2009).Shift work, cancer, and mental health have been extensively studied in recent years.In 2007, the International Agency for Research on Cancer classified shift work as a "probable" carcinogen due to its contribution to circadian disorders.Epidemiological findings reveal that women who work shifts for over 20 years have a 1.4-fold increased risk of breast cancer, while shift and night shift workers have a 2-fold increased risk of prostate cancer (Behrens et al. 2017).Animal research has shown that circadian rhythm disturbances, simulating shift work, promote the development of mammary tumors (Van Dycke et al. 2015).
Although research into the underlying mechanisms is ongoing, the impacts of the sympathetic nervous system, melatonin inhibition, and clock gene desynchronization have been identified as potential causes of shift work-related cancer (National Toxicology P 2021).Chronotype, an emerging issue of interest in the field, has been linked to increased risk of prostate cancer, depression, and sleep disorders in cohort studies (Cheng et al. 2021;Hittle and Gillespie 2018;Lozano-Lorca et al. 2020).Future research should expand and focus on the specific molecular mechanisms underlying the health effects of shift work.
This study is the first to use scientometric analysis to conduct a statistical analysis of the effects of rotational work on health.Despite the fact that this can help us obtain a more comprehensive understanding of the evolution of research on the health effects of shift work, there are limitations of this study.Firstly, we only included data from the WoSCC databases, excluding research not included in WoSCC.Despite this, WoSCC is an authoritative and representative database (Zhang et al. 2021), and it is believed that this work can still provide a basic overview and trend analysis of this area.Secondly, in the process of literature retrieval, in order to ensure the reflect the knowledge structure and development of the research topic area, we adopt a broader definition subject when screened the publications according to the titles and abstracts.Therefore, some less relevant literature was included inevitably.Therefore, it may result in a lower probability of the results being completely repeated.Finally, the search query was focused on shift work and human health.The definition and scope of human health is complex and broad.Therefore, it is difficult to ensure a 100% inclusion of literature on the topics.In general, our study reflects the development trend of the research topic and maintains the reliability of the research results to a certain extent.

Conclusion
This study utilized CiteSpace and VOSviewer to visually analyze the existing research on the health effects of shift work.The analysis revealed a consistent increase in the number of studies conducted in this field over the years.Numerous institutions and authors have directed their attention towards investigating the health implications of shift work.The United States emerged as the primary hub of research activity in this area, with global collaboration being actively pursued.Previous studies primarily focused on the impact of shift work on physical health, specifically depression, obesity, and insomnia.However, recent research has expanded to include cardiovascular disease, breast cancer, sleep duration, metabolic syndrome, and melatonin as emerging areas of interest.By adopting a more fundamental and objective approach, this study offers valuable insights into the topic, complementing previous research.The findings of this study are expected to aid researchers in identifying research gaps, determining areas of collaboration, and predicting the future impact of shift work on health.

Figure 1 .
Figure 1.The growth trend and global distribution of publication.(A) The growth trend and prediction of health effects caused by shift work trends over the period 1997.01-2022.08.(B) Global distribution map of publication from different countries/regions.(C) The chord chart of top 30 countries in field of health effects caused by shift work.Each curve segment in the figure represents a country, with the length of each curve segment representing that country's publication performance.The curve between the line segments indicates the degree of collaboration between countries, with broader segments representing closer cooperation.
the top 10 most influential authors and co-cited authors based on the output of CiteSpace.Bjorn Bjorvatn and Stale Pallesen from University of Bergen have published the most articles, with 53 (1.43%) and 48 (1.30%) publications, respectively.Bjorn Bjorvatn cooperated closely with Stale Pallesen (link strength = 51) and Siri Waage (link strength = 34).Moreover, Eva S. Schernhammer (39) and Mikko Harma (31) have published more than 30 articles.The association between shift work and sick leave, sleep patterns and psychological health, turnover intention, and night work are Bjorvatn's key research interests.In terms of centrality, among the top 10 authors, Goran Kecklund (0.08)

Figure 2 .
Figure 2. The cooperation network of main research institutes and co-cited authors.(A) The cooperation network of main research institutes.The size of each node in the graph reflects the number of publications produced by the institution.The connecting lines between institutions signify collaboration; the thicker the line, the more frequently the institutions collaborate.(B) Network visualization map of co-cited authors.Thicker lines indicate stronger collaborations.Authors represented with larger circle size or font size had relatively more collaborations.

Figure 3 .
Figure 3.The dual-map overlay of journals on health effects caused by shift work.It showed relationship distribution among journals, with citing journals on the left and cited journals on the right, and the colored paths between them suggesting the cited relationships.

Figure 4 .
Figure 4. Analyze and visualize of keywords.(A) Visualization on co-occurring keywords network.(B) keywords clustering analysis.(C) keywords distribution over time.The degree of color of each square in the figure represents the frequency of keywords per year.The frequency is correlated to the yellowness of the color.(D) timeline viewer of co-occurring keywords.

Table 1 .
Top 10 authors and co-cited authors related to health effects caused by shift work.Journals with IF > 5, with the Lancet (202.731)having the highest IF, followed by Sleep Medicine

Table 2 .
Top 10 journals and co-cited journals related to health effects caused by shift work.