Global trend of nondrug and nonsedativehypnotic treatment forinsomnia: a bibliometric study

Abstract

OBJECTIVE:

To present a bibliometric analysis of global scientific publications on the nondrug and nonsedative hypnotic treatment of insomnia with regard to influential institutions, publications, countries, research hotspots, trends, and frontiers.

METHODS:

A literature review was conducted by searching the Web of Science Core Collection (WoSCC) and China National Knowledge Infrastructure (CNKI) databases to identify all publications related to the nondrug and nonsedative hypnotic treatment of insomnia from 2000 to 2021. Eligible publications were reviewed, including annual publication increments, citation analyses, international collaborations, and keyword analyses. The data were analysed using CiteSpace (vers5.8.R3, 6.1.R2 and 6.1.6, College of Computing and Informatics, Philadelphia, PA, USA) and virtualized by knowledge maps.

RESULTS:

In total, 9832 publications were included in this analysis. The results from the WoSCC showed that the United States of America (Count = 2268, 40.33%), Stanford University (Count = 141, 2.51%), and the United States Department of Health and Human Services were the leading country, institute, and funding agency regarding the number of publications, respectively. ‘Cognitive-behavioural therapy” was the most popular research topic generated from the cocited reference. The most frequently co-occurring keywords were insomnia, cognitive behavioural therapy, disorder, depression, quality of life, Meta-analysis, older adult, sleep, prevalence and efficacy, while keywords including clinical practice guideline, guideline, and Tai Chi remained popular after 2021. Circadian rhythm was the strongest research frontier for 2000-2021. In China, Chengdu University of Traditional Chinese Medicine (Count = 69, 4.79%) was the most productive institute in this field. The most frequently co-occurring keywords from Chinese literature were sleep disorder, sleep quality, acupuncture and moxibustion, Parkinson's disease, transcranial magnetic stimulation, health education, music therapy, chronic insomnia, quality of life, and nonmotor symptoms. Traditional Chinese medicine was the strongest research frontier for 2019-2021.

CONCLUSION:

This bibliometric study provides an exhaustive mapping encompassing pertinent institute, publications, influential articles, researchers and topics of the global trend of nondrug and nonsedative hypnotic treatment for insomnia. The results show that the research trend has shifted from primary studies on the efficacy and safety of nondrug and nonsedative hypnotic treatment for insomnia to comorbidity studies. Clinical practice guidelines will potentially become the research frontier for this field post-2021. The findings are important for researchers, clinicians, journal editors, and policy-makers working in the field of nondrug and nonsedative hypnotic treatment for insomnia to understand the strengths and potentials in the current studies and guide future clinical practice, research, and science policy.

1. INTRODUCTION

Insomnia refers to sleep disturbance characterized by difficulties falling asleep or staying asleep, leading to excessive daytime sleepiness, mood disturbances, and accidents.^1,2 As the second-most prevalent mental disorder, insomnia affects 3.9% to 22.1% of the world population, and epidemiological studies from different countries yield similar prevalence estimates.³ In China, more than 300 million residents suffer from sleep disorders, and adults show a higher incidence of insomnia (38.2%). The Chinese government has published a 10-year plan pledging improvements in national sleep health.^4,5

Insomnia is an independent risk factor for cardiovascular diseases, type 2 diabetes, suicide, brain disorders, etc. In addition, mental, neurological, and substance use problems associated with chronic insomnia can affect individuals as well as their families, friends, and coworkers.^6,7

Insomnia is also associated with a significant economic burden and estimated to have cost the United States US$150.36 to US$174.89 billion in 2016.^{8,⇓,⇓- 11} The domestic sleep industry is rapidly increasing in market size. From 2016 to 2020, the overall market size of China's sleep economy increased from 261.63 billion yuan to 377.86 billion yuan, an increase of 44.42%.^12,13 Therefore, insomnia poses severe threats to both individuals and society.

Although sedative-hypnotic medications are frequently used to treat insomnia,¹⁴ safety concerns have been raised by a broad range of reported clinical data.^15,16 Past studies suggest that sedative-hypnotic drugs may increase the likelihood of psychomotor retardation, as well as other drug-related side effects. Moreover, polypharmacy may also increase fall risks and accidents in the elderly population.^17,18 The Food and Drug Administration has added a black box warning to “sedativehypnotic drug” labels to highlight the potentially fatal adverse effects.¹⁹ Therefore, continuous advancement in nondrug and nonsedative hypnotic treatment would be a big step forward in enriching the armamentarium of treatments for insomnia. Over the years, nondrug treatments,²⁰ and nonsedative hypnotics have gained popularity due to their low side effect profiles and long-term efficacy.^21,⇓-23 Currently, clinical practice guidelines in the US, Canada and many European countries unanimously recommend nondrug and nonsedative hypnotic treatment, especially Congnitive Behavioral Therapy for Insomnia (CBT-I), as the first-line treatment for chronic insomnia (symptoms for > 3 months).^6,24,⇓-26

A considerable amount of progress has been made in the last few decades to provide clinical data and explore the mechanism of nondrug and nonsedative hypnotic treatment for insomnia. The present study used bibliometric analysis to investigate the research pattern of this field through document cocitation analysis.²⁷ Bibliometric analysis is a standard tool in research management and science policy; it applies mathematical and statistical approaches to analyse scientific communications and information processes. The findings of a bibliometric study can satisfy theoretical interest and, more importantly, are used for evaluation and prediction. Bibliometric analysis is mainly applied to (a)evaluate scientific publications or institutions,^28,⇓-30 (b) analyse the research status of emerging or rare research topics,^31,32 and (c) assess and predict research trends in a particular field.^33,⇓-35 The terms bibliometrics and scientometrics are almost used as synonyms.³⁶

Although the popularity and publication output related to the nondrug and nonsedative hypnotic treatment of insomnia are rapidly increasing, only a few existing studies have analysed and visualized the research trends.³⁷ This study aimed to conduct a bibliometric analysis of the literature and to provide deeper insight into the existing publications and research trends. The specific aims included (a) identifying scientometric patterns in the field from 2000 to 2021, (b) revealing international collaborations, research topics, and the top cocited publications in this domain, (c) discovering the evolution and status of research topics in this field, and (d) discovering innovative research topics and research frontiers for future research.

2. METHODS

2.1. Selection criteria

All articles and reviews related to insomnia and nondrug and nonsedative hypnotic treatment were included to perform a comprehensive and objective bibliometric analysis. No restrictions were imposed on language. We removed duplicate articles, advertisements, popular science articles, interviews, and teaching materials, and only the most recent updates of the same study were included.

2.2. Search methods for the identification of studies

All scientific publications related to the nondrug and nonsedative hypnotic treatment of insomnia were retrieved from the Web of Science Core Collection (WoSCC) and China National Knowledge Infrastructure (CNKI) databases. For example, the search query for CNKI database is SU = ('insomnia' + 'bumei' +'sleep disorder' )*( 'complementary and alternative medicine' + 'alternative therapy' + 'cognitive behavioral therapy' + ' acupuncture and moxibustion' + 'tuina' + 'homeopathy' + 'foot reflexology' + 'diet based therapy' + 'energy healing therapy' + 'exercise' + 'hypnosis' + 'light exposure' + 'massage' + 'meditation' + 'melatonin' + 'music therapy' + 'herbs' + 'naturopathy' + 'Qi gong' + 'Tai chi' + 'transcranial magnetic stimulation' + 'vitamin' +'yoga'), Publication Date = 2000-01-01 to 2021-12-31.

2.3. Data collection

Two researchers cross-organized and compiled the retrieved items according to the inclusion and exclusion criteria. If there was a dispute, a third researcher made an arbitration decision.

The detailed research records were exported in WoSCC and CNKI plain text “txt” format, and the files were named “download ****” according to their download and saved orders. The “txt” files were then converted by CiteSpace to data that could be directly analysed.

2.4. Analysis tool

We used the intrinsic functions of the databases to analyse the publication output and the list of funding agencies, while the co-occurring relationships of countries, institutions, and keywords and the cocitation relationships of authors and references were explored using CiteSpace. The cocitation analysis of CiteSpace is only applicable to English literature; thus, this study only conducted cocitation analysis on data from the WoSCC database.

Created by Chen and his team in early 2004, CiteSpace is a free Java-based software that has the advantage of displaying the evolution of research in a particular field from multiple perspectives.^38,39 CiteSpace visualization knowledge maps are displayed with nodes and links. Nodes represent elements including authors, nations, keywords, institutions, etc., whereas links represent relationships between nodes such as collaborations and cocitations.^40,41 Betweenness centrality measures the extent of a node being part of a path that connects two arbitrary nodes in a network. The intensity of the purple ring around the node represents the level of centrality. Bursts are used to detect a period in which a specific topic has become popular to identify the frontiers and shifts in study focus, while the sigma (∑) value represents the innovation of a publication.⁴² CiteSpace offers two visualization patterns. The cluster view involves grouping closely related keywords and assigning each keyword a value, while the timeline view depicts relationships between clusters and the historical span of literature in a cluster.

The parameters of CiteSpace were set as follows: (a) time slicing was from January 2000 to December 2021, in years per slice; (b) all options in the term source were selected; (c) one node type was selected at a time; (d) selection criteria included the top 50 objects; (e) nopruning was performed, except in Figure 4A (pathfinder, pruning sliced networks and the merged network); (f) look-back years (LBYs = 5); (g) link retaining factors (LRFs = 3.0); (h) the percentage of nodes to be labeled (PNL = 1.0%); (i) the maximum links per node (MLPN = 100); and (j) all thresholds were (c,c, ccv:2,2,20;4,3,20;4,3,20).

Figure 4. Co-occurring keywords on publications.

A: network of keywords on publications of nondrug and nonsedative hypnotic treatment of insomnia from the Web of Science Core Collection database; B: top 10 keywords with the strongest citation bursts from the Web of Science Core Collection database.

3. RESULTS

3.1. Annual publication

In total, 6310 items were retrieved from the WoSCC database and 4223 items were retrieved from the CNKI database; items were filtered manually by using the review function of Citespace following the flowchart (Figure 1). Ultimately, 9832 publications were included in the analysis. The data showed continuous growth in the number of annual publications from 2000 to 2021. As displayed in Table 1, the English publication output prior to 2006 (n = 87, 1.516%) was stable and low and began to gradually increase from 2007 (n = 128, 2.231%) to 2017 (n = 406, 7.076%). From 2018 onwards, there was an extensive increase in publication output, which peaked in 2021 (n = 718, 12.515%). The trend suggests increasing popularity in this field, as evidenced by the average annual growth rate of 13.863%. The number of annual publications of Chinese literature also continued to rise, and the trend lagged behind that of English literature after 2013.

Figure 1. Study flow diagram.

Table 1.

Annual number of publications on the nondrug and nonsedative hypnotic treatment of insomnia between Jan 2000 to Dec 2021

Year	Number of publications per year
	China National Knowledge Infrastructure database	Web of Science Core Collection database
2000	41	47
2001	47	56
2002	53	43
2003	69	58
2004	83	52
2005	71	85
2006	96	87
2007	99	128
2008	119	113
2009	134	163
2010	159	222
2011	172	214
2012	171	234
2013	257	264
2014	257	330
2015	259	352
2016	300	377
2017	299	406
2018	347	492
2019	392	612
2020	372	684
2021	413	718

3.2. Distribution of countries and institutes

The collaboration map of the 142 countries is presented in Figure 2A (N = 142, E = 748, Density = 0.0747). The size of the nodes represents the number of publications, while the purple rings represent nodes with the high betweenness centrality. The United States was associated with the highest number of publications (Count = 2268, 40.33%), followed by the People's Republic of China (Count = 630, 11.2%), Canada (Count = 373, 6.63%), England (Count = 367, 6.53%), and Australia (Count = 330, 5.87%) (Table 2). The most important country was the United States, with the highest centrality of 0.64, followed by England, France, Switzerland, and Australia. We noticed that while the People's Republic of China (PRC) was the second leading country for the highest number of publications, it had low centrality (0.04). The strength of collaborations between any two countries is proportional to the thickness of the links; thicker links indicate stronger collaborations. The United States was shown to have collaborations with most countries and regions; its relationship with Brazil appeared to be the strongest, while its connections with Chile, Canada, France, and the PRC were weak. Moreover, the PRC appeared to have strong collaborations with Singapore, Israel, and Japan. In addition, links between England and the following regions, including New Zealand, Ireland, Scotland, South Africa, and Wales, were also prominent, indicating solid relationships. We also identified two prominent collaboration circuits in this network. The first circuit was a widespread network led by the United States, which collaborated with many countries and regions worldwide, including the PRC, France, Canada, and Spain. The second circuit was a small but strong collaboration network formed by Latin American countries, including Chile, Argentina, Colombia, Mexico, Uruguay, Panama, and Paraguay.

Figure 2. Productivity and collaboration map between countries and institutes.

A: collaboration map of countries with publications of nondrug and nonsedative hypnotic treatment of insomnia from the Web of Science Core Collection database; B: collaboration map of institutions with publications of nondrug and nonsedative hypnotic treatment of insomnia from the Web of Science Core Collection database.

Table 2.

Top 5 Count and Centrality of Countries on publications of nondrug and nonsedative hypnotic treatment of insomnia from the Web of Science Core Collection database

Rank	By count				By centrality
	Countries	Count	Proportion (%)		Countries	Centrality	Proportion (%)
1	The United States of America	2268	40.33		The United States of America	0.64	40.33
2	People's Republic of China	630	11.20		England	0.11	6.53
3	Canada	373	6.63		France	0.08	2.24
4	England	367	6.53		Switzerland	0.06	1.94
5	Australia	330	5.87		Australia	0.05	5.87

The collaboration map of institutions (Figure 2B) demonstrated the 1042 global research institutes in the related field, and the collaborations among them appeared weak (N = 1042, E = 2966, Density = 0.0055). As Table 3 shows, Stanford University had the highest number of publications (Count = 141, 2.51%), followed by the University of Pennsylvania (Count = 138, 2.45%), the University of Pittsburgh (Count = 123, 2.19%), Université Laval (Count = 100, 1.78%) and the University of Washington (Count = 98, 1.74%). Stanford University also scored highest in centrality (0.13), which made it the most impactful among all institutes.

Table 3.

Top 5 Count and Centrality of Institutions on publications of nondrug and nonsedative hypnotic treatment of insomnia from the Web of Science Core Collection database

Rank	By count			By centrality
	Institution	Count	Proportion (%)		Institution	Centrality	Proportion (%)
1	Stanford University	141	2.51	Stanford University		0.13	2.51
2	University of Pennsylvania	138	2.45	University of Pittsburgh		0.10	2.19
3	University of Pittsburgh	123	2.19	University of Pennsylvania		0.09	2.45
4	Laval University	100	1.78	University of California San Diego		0.09	1.07
5	University of Washington	98	1.74	University of Toronto		0.07	1.03

Moreover, within the Chinese literature, the institution that published the largest number of studies in Chinese was Chengdu University of Traditional Chinese Medicine (Count = 69, 4.79%) (Table 4). Domestic studies lacked cross-provincial research collaboration, and the few remaining collaboration groups comprised institutions in the same city or administration section.

Table 4.

Top 5 Count of Institutions on publications of nondrug and nonsedative hypnotic treatment of insomnia from the China National Knowledge Infrastructure database

Rank	Institution	Count	Proportion (%)	Centrality	Province
1	Chengdu University of TCM	69	4.79	0	Sichuan
2	Heilongjiang University Of Chinese Medicine	39	2.71	0	Heilongjiang
3	Tianjin University of TCM	37	2.57	0	Tianjin
4	Changchun University Of Chinese Medicine	27	1.87	0	Jilin
5	First Teaching Hospital of Tianjin Univrsity of TCM	24	1.67	0	Tianjin

Note: TCM: Traditional Chinese Medicine.

3.3. Analysis of cocited references

The cocited reference network, which comprised 1658 nodes and 8360 links, is presented in Figure 3A (N = 1658, E = 8360, Density = 0.0061). Each node represents a reference labelled with the first author's name and the year of publication, and the node's size represents its frequency. The link between any two corresponding documents represents their cocitation relationships. Qaseem et al ⁴³ had the highest cocited frequency of 302, followed by the American Psychiatric Association⁴⁸ (Freq: 228), Riemann et al ⁶ (Freq: 210), Trauer JM et al ⁵¹ (Freq: 187) and Zachariae et al.⁵² (Freq: 164). Mayer G et al ⁵³ was the most impactful owing to the highest centrality value (0.13), whereas Qaseem et al ⁴³ had the highest sigma value (∑: 16.28), making it the most innovative.

Figure 3. Primary keywords extracted from the cocited references.

A: network of cited references on publications of the nondrug and nonsedative hypnotic treatment of insomnia from the Web of Science Core Collection database; B: cluster map of cited references on publications of the nondrug and nonsedative hypnotic treatment of insomnia from the Web of Science Core Collection database; C: timeline network of cited references on publications of the nondrug and nonsedative hypnotic treatment of insomnia from the Web of Science Core Collection database.

The cluster map used log-likelihood tests (LLR) to assess the distribution from keywords comprising30 valid clusters where all silhouettes were greater than 0.8, suggesting a high quality of clustering. The coverage and size of each cluster are demonstrated by the multicoloured polygons in Figure 3B. The top 10 largest clusters are listed in Table 5. The largest cluster was cognitive behavioural therapy, consisting of 206 members.

Table 5.

Top 10 clusters of Cited references on nondrug and nonsedative hypnotic treatment publications on insomnia from the Web of Science Core Collection database

Cluster ID	Size	Silhouette	Mean (years)	Top Terms (log-likelihood ratio, P-level)
0	206	0.856	2018	Cognitive behavioral therapy
1	186	0.918	2008	Ramelteon
2	182	0.911	2005	Psychological
3	162	0.84	2011	Internet
4	139	0.832	2016	Menopause
5	133	0.935	2001	Melatonin
6	101	0.859	2016	Mindfulness
7	97	0.897	2014	Posttraumatic stress disorder
8	70	0.923	2019	Acupuncture
9	59	0.961	2005	Jet lag

The top 10 clusters were also placed on a timeline map (Figure 3C). The clusters were placed in descending order by size from top to bottom and ascending order by recency from left to right. Cluster #0 cognitive behavioural therapy was the newest and largest cluster of references. The cross-topic cocitation relationships appeared strong among Clusters #0, 3, 4, and 6 and Clusters #3 and 7, as demonstrated by the large number of links between intercluster members.

As the largest cluster, Cluster #0 contained 206 documents in the cocitation network, and its silhouette value was 0.856. Cluster #0 was labelled cognitive behavioural therapy and focused on basic theory and clinical practice in the nondrug and nonsedativehypnotic treatment of insomnia, such as CBT-I. The concept was used to compare its safety and efficacy with pharmacological therapies. The most representative document in this cluster was that of Qaseem et al, ⁴³ an ACP clinical guideline, which recommends CBT-I as the first-line treatment for insomnia and points out low-quality evidence on pharmacological therapy to treat patients for whom CBT-I failed. The second most representative document was the European guideline by Riemann D,⁶ which is largely in line with the ACP guideline and recommends CBT-I as the first-line treatment strategy.

Cluster #1 ramelteon was the second largest cluster with 186 members, and its silhouette value was 0.918. This cluster focused on the basic science and mechanisms of nonsedative hypnotic therapies. The concept was applied to develop alternative or supplemental management approaches with minimal side effects caused by sedation. The top two representative documents were Erman et al ⁴⁴ and Roth et al, ⁴⁵ both of which were published in Sleep Medicine. Erman et al ⁴⁴ investigated the dose‒response of ramelteon in patients with chronic primary insomnia and demonstrated ramelteon's effect in reducing latency to persistent sleep and increasing total sleep time with no apparent next-day residue effect. Roth et al ⁴⁵ conducted a randomized placebo-controlled trial to assess the safety and efficacy of ramelteon and revealed ramelteon's efficacy in reducing sleep latency for older adults with chronic insomnia, with no significant rebound insomnia or withdrawal effects.

Cluster #2 contained182 documents, and its silhouette value was 0.911. It was labelled as psychological. The top two representative documents were the diagnostic manual published by the American Academy of Sleep Medicine (AASM)⁴⁶ and a conference statement published by the National Institute of Health (NIH).⁴⁷ The researchers acknowledged the efficacy of nondrug and nonsedative hypnotic therapies. The NIH proposed the need for longitudinal studies of RCTs and educational programs for health professionals and the public.

Cluster #3 contained 162 documents, and its silhouette value was0.84. It was labelled internet. The top two representative documents were published by the American Psychiatric Association (APA) ⁴⁹ and Morin et al.⁴⁸ Morin et al ⁴⁸ investigated the long-term outcome of CBT-I and hypnotic medications and suggested that the addition of medication to CBT can benefit acute therapy, but medication must be discontinued during the maintenance phase to optimize the long-term outcome.

Cluster #4 contained139 documents, and its silhouette value was 0.832. It was labelled menopause, focusing on insomnia in specific populations. The most representative documents in Cluster #4 were published by Johnson JA et al ⁵⁰ and Sateia MJ et al. ¹⁶ Johnson et al ⁵⁰ conducted a systematic review and meta-analysis, concluding that CBT-I is effective in treating insomnia in cancer survivors. Sateia et al ¹⁶ focused on the pharmacological treatment of chronic insomnia in adults and outlined the recommendations for the use of each individual agent in clinical practice guidelines.

3.4. Keyword analysis

According to the data from the Web of Science Core Collection database, the visualization of the keyword co-occurrence network consisting of 816 nodes and 1269 links is presented in Figure 4A (N = 816, E = 1269, Density = 0.0039). The most frequently co-occurring keywords were insomnia (Count = 1372), cognitive behavioural therapy (Count = 1294), disorder (Count = 630), depression (Count = 612), quality of life (Count = 582), meta-analysis (Count = 578), older adult (Count = 546), sleep (Count = 543), prevalence (Count = 533), and efficacy (Count = 517). The focus on cognitive behavioural therapy appeared the latest, with only two citations in 2002; however, it peaked in 2021 (Count = 178). In addition, several keywords that had high centrality values are listed in Table 6, including arousal, light therapy, replacement, daytime sleepiness, onset, major depression, dysfunctional belief, cognitive therapy, Alzheimer's disease and mood. The most innovative topic was psychiatric disorder (∑ = 3.89), and the most active citer was Baglioni et al.⁵⁸

Table 6.

Top 10 count, centrality and sigma (∑) of keywords on publications of nondrug and nonsedative hypnotic treatment of insomnia from the Web of Science Core Collection database

Rank	By count				By centrality				By ∑
	Keyword	Count	Year		Keyword	Centrality	Year		Keyword	∑	Year
1	Insomnia	1372	2000		Arousal	0.22	2006		Psychiatric disorder	3.89	2001
2	Cognitive behavioral therapy	1294	2002		Light therapy	0.18	2007		Onset	3.44	2000
3	Disorder	630	2000		Replacement	0.16	2000		Human	3.12	2000
4	Depression	612	2000		Daytimesleepiness	0.16	2010		Alzheimer's disease	2.45	2000
5	Quality of life	582	2001		Onset	0.15	2000		Body temperature	2.34	2006
6	Meta-analysis	578	2000		Major depression	0.14	2002		Zolpidem	2.2	2001
7	Older adult	546	2000		Dysfunctional belief	0.12	2005		Population	2.17	2000
8	Sleep	543	2000		Cognitive therapy	0.1	2001		Slow wave sleep	2.16	2000
9	Prevalence	533	2000		Alzheimer's disease	0.1	2000		Phase syndrome	2.11	2001
10	Efficacy	517	2001		Mood	0.09	2000		Older adult	2.07	2000

As many as 196 keywords obtained strong citation bursts. The top ten keywords with strong citation bursts in each corresponding period are illustrated in Figure 4B. These keywords were often used to discover research topics that have attracted an extraordinary degree of attention. Circadian rhythm had the highest burst strength (32.22), followed by double blind, zolpidem, psychiatric disorder, body temperature, human safety, rhythm, bright light and older adult. In addition, circadian rhythm, psychiatric disorder and bright light had the longest duration of 14 years.

The 33 keywords with the most recent citation bursts are shown in supplementary Figure 1. These keywords were used to identify any recent shift in study focus and the rising interest in a particular topic or keyword in the past period, which could predict future research trend and frontiers. The recent top three keywords with strong citation bursts were clinical practice guideline, guideline, and Tai chi. American college was the keyword with the most recent and longest burst duration since 2017.

The most frequently co-occurring keywords of data from the CNKI database are shown in Table 7. Sleep disorder had the highest cooccurrence frequency of 321, followed by sleep quality (Count = 180), acupuncture and moxibustion (Count = 168), Parkinson's disease (Count = 83), transcranial magnetic stimulation (Count = 81), health education (Count = 52), music therapy (Count = 50), chronic insomnia (Count = 46), quality of life (Count = 45), and nonmotor symptoms (Count = 41). Sleep disorders also scored highest in centrality (0.31), followed by acupuncture and moxibustion, sleep quality, etc. The most innovative topic of Chinese literature was health education (∑ = 1.19), and the most active citer was Wang et al. ⁵⁹

Table 7.

Top 10 count, centrality and sigma (∑) of keywords on publications of nondrug and nonsedative hypnotic treatment of insomnia from the China National Knowledge Infrastructure database

Rank	By count				By centrality				By ∑
	Keyword	Count	Year		Keyword	Centrality	Year		Keyword	∑	Year
1	Sleep disorder	321	2004		Sleep disorder	0.31	2004		Health education	1.19	2008
2	Sleep quality	180	2003		Acupuncture and moxibustion	0.21	2006		Traditional chinese medicine	1.16	2003
3	Acupuncture and moxibustion	168	2006		Sleep quality	0.19	2003		Non-motor symptom	1.03	2009
4	Parkinsons disease	83	2004		Transcranial magnetic stimulation	0.05	2013		Clinical application	1.03	2018
5	Transcranial magnetic stimulation	81	2013		Quality of life	0.05	2006		Sleep disorder	1	2004
6	Health education	52	2008		Pittsburgh sleep quality index	0.05	2014		Acupuncture and moxibustion	1	2006
7	Music therapy	50	2006		Parkinsons disease	0.04	2004		Sleep quality	1	2003
8	Chronic insomnia	46	2002		Clinical observation	0.04	2007		Transcranial magnetic stimulation	1	2013
9	Quality of life	45	2006		Music therapy	0.03	2003		Quality of life	1	2006
10	Non-motor symptom	41	2009		Chronic insomnia	0.03	2006		Pittsburgh sleep quality index	1	2014

Among 68 keywords, the top ten with the strong citation bursts are displayed in supplementary Figure 2. Health education had the highest burst strength (7.23), followed by Traditional Chinese Medicine, nonmotor symptoms, clinical application, Chinese medicine, systematic review, acupuncture-moxibustion therapy, cognitive function, intractable insomnia and influencing factors.

The 13 keywords with the most recent citation bursts are shown in supplementary Figure 3, which could reveal the differences between the research frontiers of Chinese and English literature. The recent top three keywords with strong citation bursts were Traditional Chinese Medicine, cognitive function and sleep disturbance.

4. DISCUSSION

4.1. Basic information

Overall, this paper analysed 5623 qualified publications from the WoSCC database and 4209 Chinese publications from the CNKI database on nondrug and nonsedative hypnotic treatment for insomnia. We observed continuous growth in English publication output during the study period, peaking in 2021 (n = 718), which was 15 times the output at the beginning of the study period in 2000 (n = 47). There was also a rapid increase after 2017; the average growth rate from 2017 to 2021 was 15.32%, surpassing that of complementary and alternative medicine for common mental disorders (15.027%).⁶⁰ In China, the growth of research output in this field was more stable. The average annual growth rate from 2000 (41) to 2021 (413) was 11.63%. This result indicates a global increased interest in researching nondrug and nonsedative hypnotic treatment for mental disorders, especially insomnia. A bibliometric analysis showed that “insomnia” has become one of the most frequently discussed topics concerning mental health⁶¹during the Corona virus disease 2019 pandemic, with the population affected by insomnia also rapidly growing during this period.^62,63 However, we noticed a decline in English publication output between 2019 and 2021, which is likely due to the lack of research collaborations and operations secondary to lockdowns and travel restrictions during the pandemic. Future research could show the research trends or any shift of study focus in the post pandemic period.

There is a global trend towards national and international collaborations to improve patient care.^64,⇓-66 In this study, we found that the United States was the leading country in research output among the 142 counties, accounting for 43.33% of the total research output, and had the highest collaboration performance, especially with Brazil, China, Chile, Canada, and France. The institute and funding agency with top research output were Stanford University and the United States Department of Health and Human Services, respectively. As a major contributor to scientific research, the United States allocates a large budget to research and has a vast number of researchers and centres. For example, the 2022 Biden administration allocated 135 billion dollars of the government's budget to research, which is a 9% increase from that in 2021.⁶⁷

In addition to funding and research levels, cultural diversity also affects people's perceptions and behaviour towards sleep in different regions. For instance, it is customary for people to take an after-lunch nap in China, Italy, Spain, and many other Mediterranean countries, and compared to that in Oceania countries, the late bedtime in Southeast Asian countries/regions leads to a shorter sleeping time at night, especially in Hong Kong.^68,69 Although CBT-I is now recommended as the first-line treatment for insomnia,^6,14,26,43 the implementation of nondrug and nonsedative hypnotic treatment remains complex and diverse in different regions due to cultural influences. For example, Mayan traditional medicine uses custard apple, red sandalwood, capers, mimosa, cosmos, and petiveria alliacea to treat central nervous system disorders.⁷⁰ In contrast, acupuncture, herbs, and Qi gong are commonly applied in China and Southeast Asian countries to improve sleep disorders.^37,71 The results of the institute analysis of Chinese literature in this study also demonstrated that almost all the institutions providing nondrug and nonsedative hypnotic treatment for insomnia in China's mainland were Traditional Chinese Medicine colleges and hospitals. Therefore, the economy, scientific research level, and cultural background are all critical variables that affect a country's contributions to this research field.

4.2. Research fundamentals

The influence and citation frequency of references reflect the fundamentals of a particular research field.⁷² The top 15 cited references concerning citation frequency, centrality, and ∑values were primarily guidelines/statements, evidence-based medicine studies, and clinical trials, which play fundamental and instructive roles in research on nondrug and nonsedative therapies for insomnia. The guideline that Qaseem et al ⁴³ published in 2016 was the most innovative and cited document. It systematically reviewed all randomized controlled trials published in English by the American College of Physicians (ACP) from 2004 to 2015, which recommended CBT-I as the first-line treatment for insomnia following the NIH Consensus and State of the Science Statement in 2005.¹⁵ The most pivotal article was by Mayer et al.⁵³ This study was a multicentre, double-blind randomized controlled trial published in Sleep that reported the long-term efficacy and safety of ramelteon in primary chronic insomnia. These two highly influential documents belonged to Cluster #0 (cognitive behavioural therapy) and Cluster #2 (ramelteon). However, among the 5623 papers included, many high-quality articles did not obtain high ranks. This phenomenon is in line with the principle of cumulative advantage, also known as the "Matthew effect"; articles and authors that are already frequently cited are more likely to be widely cited.^36,73

Insomnia, one of the most common sleep disorders, is often accompanied by many neurological conditions, and disease classifications and treatment evaluations are critical. The International Classification of Sleep Disorders,⁷⁴ the Diagnostic and Statistical Manual of Mental Disorders⁴⁸ and the International Classification of Diseases⁷⁵ are the three most widely used classification systems for insomnia.⁷⁶ The evaluation of insomnia treatment included (a) daily sleep diaries; (b) sleep quality scales [including the Insomnia Severity Index, Pittsburgh sleep quality index (PSQI), Morning and Evening Questionnaire, Dysfunctional Beliefs and Attitudes about Sleep, Epworth Sleepiness Scale, Ford Insomnia response to stress Test, and Sleep Dysfunction Rating Scale]; (c) polysomnography; and (d) actigraphy.

4.3. Research hotspots

A research hotspot includes scientific topics or knowledge discussed by a relatively large number of documents during a certain period. Keyword analysis can accurately identify research hotspots using the data analysis function of Citespace.³⁹ The results from the WoSCC database showed that 321 (39.34%) keywords hada citation frequency of more than 10, indicating that many topics in this field are well supported by research. The network analysis of keyword co-occurrences sheds light on research topic selections. The results showed that the keywords older adult, CBT-I, chronic insomnia, and depression were the research hotspots for the topics population, interventions, insomnia types, and comorbidities, respectively.

4.3.1. Population

Older adult (Count = 546) was the research hotspot among 16 target patient populations. The ageing population across the globe has attracted high levels of interest in the well-being of elderly adults.⁷⁷ The most impactful document was a Meta-analysis⁵⁸ indicated that nondepressed people with insomnia have a twofold risk of developing depression compared to people with no sleep difficulties. Moreover, the effect of insomnia in predicting subsequent depression is similar in children and adolescents, working-age individuals, and older adults.

4.3.2. Interventions

Cognitive-behavioural therapy (Count = 1294) was the research hotspot of the 30 interventions that were extracted from keyword analysis. Manber et al ⁷⁸ was the most cited clinical trial concerning the intervention hotspot. It suggested that the superior results associated with the CBT-I-combined treatments proved CBT-I to be more effective for depressive disorders and insomnia than antidepressants alone. And identifying the mechanisms is an emerging area of research, as they are vital in intervention improvement.

4.3.3. Insomnia types

Chronic insomnia was identified as the hotspot among the 14 insomnia types, with the highest frequency of co-occurrence (Count = 194). Lund et al,⁷⁹ in a cross-sectional online survey, reported that over 60% of college students were categorized as poor-quality sleepers according to the PSQI, while stress was perceived as the major factor. The most cited animal experiment⁸⁰ explored novel continuous stress imposed by the perpetual avoidance of water on a wheel (PAWW). This PAWW stress model was expected to assess circadian locomotor activity during continuous exposure to a stressor.

The most impactful review of chronic insomnia reviled compelling evidence suggesting that CBT is effective in treating chronic insomnia and outlined a five-level CBT-I stepped care program as a health technology solution to deliver CBT as the first-line insomnia treatment.⁸¹

4.3.4. Comorbidities

Insomnia is associated with various symptoms and exacerbated by certain underlying medical conditions.^82,83 According to the data, depression (Count = 612) was the research hotspot among the 69 diseases/symptoms extracted. The most cited review suggested that heightened levels of emotional arousal can be the maintaining factor in insomnia by mediating the interaction between cognitive and physiological hyperarousal.⁸⁴ In reverse, dysfunction in neural circuitries regulating sleep-wake cycles leads to emotional distress.

The analysis results of Chinese literature showed that the keywords college student (Count = 5), acupuncture and moxibustion (Count = 168), chronic insomnia (Count = 46), and Parkinson's disease (Count = 83) were the research hotspots for the following topics: population, interventions, insomnia types, and comorbidities, respectively. The differences between the results from Chinese and English literature may reveal the influence of factors such as culture, policy, ethnicity and disease incidence in the studies.

4.4. Research frontiers and trends

To explore the research trends of the nondrug and nonsedative hypnotic treatment of insomnia, many important topics with strong citation bursts were selected as guides for future research. We believe that the topics with significant attention in a short period of time represent the development of a trend in the future, and the topics with strong citation bursts in recent years will reflect the newest research trends. At the beginning of our study period, the studies from the WoSCC database mainly focused on symptoms and mechanisms, and topics such as circadian rhythm, body temperature, humans, rhythm, and bright light were of special interest. The research foci then switched to clinical practice, and we identified that the most recent topics with strong citation bursts, including clinical practice guidelines and guidelines may become research frontiers after 2021. Moreover, a large proportion of patients experience recurrence with other interventions, such as Tai chi, medicinal plants, and transcranial magnetic stimulation. Although the American and European guidelines recommend CBT-I as the first-line treatment for insomnia,^6,14,26,43 many patients experience sleep disturbance after CBT-I.⁸⁵ Therefore, evidence-based data on reliable substitutes for CBT-I and the comparison between these modalities with the currently recommended approaches are urgently needed.²³ Our data also show that the nondrug and nonsedative hypnotic management of insomnia for high-risk groups such as college students and shift workers has attracted extensive research interest since 2017. We anticipate that pragmatic and group-specific treatment strategies will also become a future research focus. Thus, identifying group-specific treatment methods should be given equal attention when exploring different types of interventions and making recommendations. Moreover, influencing factors and delivery routes such as adherence and online modalities will likely remain popular after 2021 due to their likelihood of improving treatment outcomes. Although the research addressing adherence to nondrug and nonsedative hypnotic interventions is still in its infancy, the steps to advance this research are to (a) establish definitions and standard measures of treatment outcomes; (b) assess adherence outcomes across a wide variety of clinically heterogeneous groups of insomnia patients with different delivery models; and (c) explain the relationship between adherence and outcomes with the utilization of behaviour change theories.⁸⁶ The application of digital technologies (online platforms where the content is provided with rich media interactions and algorithm-based information)^87-89 is considered effective when crossing the barriers that inhibit access to mental health programs, especially CBT-I. Therefore, the comparison of effectiveness between the ‘online’ and traditional delivery routes of CBT-I has recently gained interest.^90,91

In contrast, the results of Chinese literature showed that Chinese scholars were more focused on the clinical efficacy of Traditional Chinese Medicine, especially acupuncture therapy. In addition, acupuncture therapy for insomnia in China has been subdivided into various techniques, such as auricular therapy, abdominal acupuncture, acupoint injection, back-shu point, abdominal massage treatment, acupuncture combined with medicine and acupoint catgut embedding. Different from the global trends above, the research frontier of Chinese literature after 2021 remains the clinical efficacy of several nondrug and nonsedative hypnotic treatment methods for insomnia, including cognitive behaviour therapy, acupoint combination, Traditional Chinese Medicine, vitamin D, etc. Thus, the Chinese literature has made significant contributions in the field of acupuncture treatment for insomnia, and we suggest that Chinese scholars should refer to the global trends and focus on high-risk groups, influencing factors and delivery routes of nondrug and nonsedative hypnotic treatment for insomnia.

In conclusion, Nondrug and nonsedative hypnotic treatment for insomnia has become a very popular area of research among academics and practitioners. Great efforts have been made to validate the efficacy and enhance the application of nondrug and nonsedative hypnotic treatment in clinical settings. This study explored the emerging trends and research frontiers of nondrug and nonsedative treatment for insomnia by scientometric analysis from 2000 to 2021.

The upward trend in publication output indicates an increased research interest regarding the nondrug and nonsedative hypnotic treatment of insomnia. The USA was the major contributor in this field with regard to the most prolific country and the most productive institution. The most active article was Qaseem et al.⁴³ Topics including patient demographics (older adults), interventions (cognitive behavioural therapy), insomnia subtypes (chronic insomnia), comorbidities (depression), and study methods (Meta-analyses) were the research hotspots between 2000 and 2021. The research foci shifted from mechanisms to clinical practice during the study period. Clinical practice guidelines will become the research frontier for the nondrug and nonsedative hypnotic treatment of insomnia post-2021, alongside Tai chi, adherence, and online modalities, as potential research hotspots. Chinese literature has made significant contributions to the field of acupuncture treatment for insomnia. The results will help clinicians, researchers, journal editors, and policy-makers gain a deeper understanding of the research evolution and status of patient stratification as well as the subtypes, interventions, and research contents of the nondrug and nonsedative hypnotic treatment of insomnia and provide reliable information for guiding and inspiring future clinical practice, research, and science policy.

5. SUPPORTING INFORMATION

Supporting data to this article can be found online at http://journaltcm.cn.