Evolutionary trajectory and knowledge structure of wearable devices for health management: A bibliometric analysis (1997–2025)

ABSTRACT

Purpose:

To deconstruct the evolutionary trajectory and intellectual structure of wearable devices for health management.

Method:

The study executed a systematic bibliometric analysis of 2,463 high-quality publications from the Web of Science Core Collection (1997–2025).

Result:

The findings reveal that the field entered an exponential growth phase post-2018, anchored by a bipolar dominance of the USA and China. While academic institutions drive innovation, the global collaborative network remains structurally fragmented. There were 637 keywords in total, key themes include “digital health,” “wearable devices,” “physical activity,” “artificial intelligence.” Research hotspots have crystallized around three strategic pillars: high-fidelity physiological sensing, AI-driven intelligent analytics, and clinical translation for proactive health. Furthermore, frontier trends indicate a critical pivot from generalized monitoring to the precision management of specific pathologies (e.g., cardiovascular diseases) and a technological shift toward energy-autonomous systems, by elucidating the pathway from technological incubation to clinical application.

Conclusion:

This study offers empirical evidence and strategic insights to guide future interdisciplinary synergy and the construction of interoperable data ecosystems for precision health management.

Introduction

The accelerating aging of the global population, compounded by the rising prevalence of chronic diseases, places an unprecedented strain on the capacity, resource allocation, and operational models of contemporary healthcare systems. Projections indicate that the global elderly population will exceed 1.5 billion by 2050.[1,2,3] Concurrently, there is a paradigm shift in healthcare demand toward personalization, continuity, and prevention, rendering traditional, reactive management models increasingly inadequate. In this landscape, wearable devices have emerged as pivotal technological enablers. Leveraging their portability and capability for continuous longitudinal monitoring, they demonstrate transformative potential across diverse applications, ranging from early disease warning[4] and diagnostic support[5] to rehabilitation management[6] and lifecycle health monitoring.[7,8]

The convergence of the Internet of Things (IoT) and Artificial Intelligence (AI) has further catalyzed the intelligent evolution of these devices. By integrating multimodal physiological data acquisition with advanced analytics, modern wearables enable dynamic health assessments and personalized interventions. This technological synergy facilitates a transition from a “standardized, post-event response” model to a “precise, proactive prevention” paradigm.[9,10] Such an evolution not only enhances clinical efficacy but also provides critical technical scaffolding for optimizing medical resource allocation.

Despite the exponential growth in related literature, the current research landscape remains notably fragmented. Existing studies predominantly focus on isolated technologies or specific application scenarios, failing to capture the holistic evolutionary trajectory and the complex interdisciplinary collaboration networks within the field. Moreover, there is a distinct paucity of systematic visual analyses that delineate core research hotspots, emerging frontiers, and international collaborative patterns. This lack of a macrolevel bibliometric perspective hinders researchers and policymakers from effectively identifying strategic directions, technology translation pathways, and opportunities for resource integration.

To address these limitations, this study employs bibliometric analysis combined with CiteSpace knowledge mapping to systematically deconstruct the research progress on wearable devices in health management.[11] By constructing a comprehensive knowledge map, this paper aims to delineate the field’s intellectual structure, revealing research hotspots, developmental trajectories, and collaborative networks among key contributors. The findings are intended to provide empirical evidence and strategic insights to foster interdisciplinary cross-fertilization and guide future research agendas. This study comprehensively sorts out the latest research progress in the field of health wearable devices, providing theoretical support for subsequent device function optimization and technological innovation in the field. It also offers practical reference for clinical professionals in the promotion and application of related wearable technologies. For patients, enhanced awareness of the functions and value of wearable devices can contribute to the adoption and maintenance of healthy behaviors, thereby improving personal health management outcomes.

Methods

Data acquisition and search strategy

To ensure the representativeness and bibliometric quality of the dataset, the WoSCC was utilized as the primary data source. As the world’s leading citation index, WoSCC is widely recognized for its rigorous indexing standards and comprehensive coverage of high-impact peer-reviewed literature, making it the standardized choice for large-scale bibliometric analyses. The data retrieval strategy was executed using the topic search (TS) field to capture relevant publications comprehensively. The specific query string was formulated as follows:

TS = (“wearable electronic device” OR “wearable technology” OR “wearable device”) AND TS = (“health management” OR “disease management”)

The search was refined based on the following inclusion and exclusion criteria: (1) Document Type: Limited to “Article” and “Review” to prioritize original research findings and authoritative theoretical syntheses. (2) Language: Restricted to English to ensure linguistic consistency during keyword co-occurrence analysis and text mining. (3) Timeframe: The search covered all available records up to October 2025 to incorporate the most recent scholarly developments. Finally, duplicate publications, irrelevant studies, and retracted articles were manually excluded.

Bibliometric analysis and visualization

The complete bibliographic records exported from WoSCC, including titles, authors, institutions, countries, journals, abstracts, keywords, and references, constituted the analytical dataset for this study. Data analysis integrated descriptive statistics and knowledge mapping visualization techniques.

Firstly, Microsoft Excel 2021 was utilized for basic data organization and descriptive statistics to calculate and present annual publication trends, contributions by countries/regions and institutions, and journal distributions. Subsequently, the bibliometric and science mapping software CiteSpace (version 6.4.R1) and VOSviewer (version 1.6.20) were employed for in-depth network analysis and visualization. The year of the earliest published literature is 1997; therefore, the time span is set as 1997–2024, with a time slice of one year. The analyses performed included collaborative network analysis, keyword co-occurrence analysis, keyword clustering analysis, and burst detection. All visualization parameters within CiteSpace were optimized based on the data characteristics to ensure the clarity and interpretability of the generated maps.

Results

Annual publications output and growth trends

The initial literature retrieval from the WoSCC database yielded a total of 11,391 records. Following rigorous application of exclusion criteria (database, document type, and language) and a subsequent manual screening to remove 49 irrelevant entries, a final dataset of 2,463 articles was retained for bibliometric analysis [Figure 1].

Figure 1.

Literature search process

The temporal distribution of annual publications [Figure 2] delineates the evolutionary trajectory of the wearable health management field. This trajectory can be categorized into three distinct phases: (1) The Nascent Exploration Phase (1997–2013): This period was characterized by sluggish growth, with an annual average of fewer than 10 publications. Research during this stage was primarily theoretical and tentative, reflecting a field still in its infancy with limited systematic frameworks. (2) The Fluctuating Growth Phase (2014–2017): This interval marked a visible inflection point. Despite intermittent volatility, publication volume rose substantially, signaling a growing academic consensus on the utility of wearable devices in health contexts. (3) The Exponential Surge Phase (2018–Present): Post-2018, the field entered a period of rapid proliferation. The number of publications escalated sharply, indicating that the domain has transitioned from theoretical exploration to widespread application and technological maturation.

Figure 2.

Annual publication in wearable device for health management

Geospatial distribution and international cooperation

The USA, China, and England emerged as the top contributors, collectively accounting for 66.66% of the total publication corpus [Table 1]. The USA maintained a dominant position, securing the highest H-index (H = 79) and the second-highest average citations per article (AC = 41.12), which underscores its preeminence in both academic output and scholarly influence. Notably, while Australia ranked sixth in total volume, it registered the highest average citation rate (AC = 49.99), suggesting that its research output is characterized by high per-item impact. In contrast, China ranked second in both publication count and H-index but recorded the lowest average citation count among the top 10 nations, indicating a perceptible disparity between publication quantity and citation impact. Furthermore, the network analysis of international collaboration [Figure 3] reveals a decentralized structure. Although cross-border partnerships are frequent, the most productive countries have yet to form cohesive, highly integrated collaborative clusters.

Table 1.

The top 10 countries/regions contributed to publications

Country/region	TP	PTP (%)	H-index	TC	AC
USA	712	28.90	79	29274	41.12
China	680	27.61	64	17721	26.06
England	250	10.15	52	9495	37.98
South Korea	178	7.227	37	5342	30.01
Italy	163	6.618	39	5756	35.31
Australia	158	6.415	43	7898	49.99
Canada	140	5.684	36	5441	38.86
India	132	5.359	27	4779	36.2
Germany	86	3.492	30	3173	36.9
Spain	85	3.451	27	2333	27.45

TP=Total publications; PTP=Percentage of the total publications; TC=Times cited; AC=Average number of citations per publication

Figure 3.

Collaborative network among countries/regions

Institutional contributions and collaborative networks

The analysis of the top 10 most productive institutions reveals a distinct geographical concentration, with the USA securing five positions, followed by the United Kingdom (3), China (1), and Canada (1) [Table 2]. The broader institutional landscape [Figure 4] encompasses a vast array of leading academic and research organizations. Pivotal entities, such as the University of California System, Harvard University, the Chinese Academy of Sciences (CAS), and University College London (UCL), exhibit high network centrality, underscoring their substantial academic influence and pivotal roles in knowledge dissemination.

Table 2.

The top 10 institutions contributed to publications

Institution	TP	PTP (%)	H-index	TC	AC
University of California System, USA	91	3.70	33	5863	64.43
Harvard University, USA	69	2.80	24	2,616	37.91
Chinese Academy of Sciences, China	66	2.68	22	1,895	33.25
Harvard University Medical Affiliates, USA	57	2.31	22	1,895	33.25
University of London, England	53	2.15	20	1,422	26.83
Harvard Medical School, USA	49	1.99	20	2,048	41.8
Stanford University, USA	41	1.67	24	2,696	65.76
University of Oxford, England	36	1.46	20	1,596	44.33
Imperial College London, England	35	1.42	21	1,259	35.97
University of Toronto, Canada	34	1.38	15	993	29.21

TP=Total publications; PTP=Percentage of the total publications; TC=Times cited; AC=Average number of citations per publication

Figure 4.

Institutional collaborative networks (a: Global Institutions; b: Institutions in the USA; c: Institutions in China) (The size of a node directly reflects the publication count and academic contribution of the corresponding institution. A larger node indicates more research output from that institution in this field. The connecting lines between nodes reveal collaborative relationships between different institutions, with denser or thicker lines indicating stronger collaboration.)

Despite the high volume of participating institutions, the overall collaborative network appears structurally decentralized. The network topology is characterized by loose coupling rather than high integration. Two prominent collaborative modules are identifiable: a primary cluster anchored by U.S. institutions (e.g., University of California System, Harvard) and a secondary cluster centered on the Chinese Academy of Sciences. These clusters primarily reflect strong domestic cohesion within the two leading nations. However, direct cross-cluster connectivity between these two major poles remains sparse [Figure 4a]. Further subnetwork analysis reveals that while institutions within both the USA and China maintain cohesive internal ties, the U.S. collaborative network demonstrates significantly higher density and topological complexity compared to its Chinese counterpart [Figure 4b and c].

Key journals and disciplinary distribution

Research on wearable health management has been disseminated across 930 academic venues, reflecting a high degree of disciplinary diversity. The top 10 most productive journals collectively published 473 articles (19.22% of the corpus), serving as the core intellectual hubs for this domain [Table 3]. Sensors emerged as the leading outlet with 134 publications (5.44%), exhibiting a publication volume significantly superior to its peers. Notably, high-impact venues, such as Journal of Medical Internet Research (IF = 6.0) and Nano Energy (IF = 17.1), featured prominently in the top tier. Regarding citation impact, Nano Energy (AC = 35.17) and JMIR mHealth and uHealth (AC = 35.08) demonstrated the highest per-article citation rates, signaling the substantial scholarly traction of research published therein. Temporal trend analysis [Figure 5a] indicates that journals, like Sensors, IEEE Access, and JMIR mHealth and uHealth, have experienced a sustained surge in output over the last five years, highlighting the enduring academic emphasis on both engineering solutions and mobile health interventions.

Table 3.

The top 10 journals that published articles on the wearable device in health management

Journals	TP	PTP (%)	IF 2024	TC	AC
Sensors	134	5.44	3.5	3777	28.19
JMIR mHealth and uHealth	74	3.00	6.2	2596	35.08
Journal of Medical Internet Research	71	2.88	6.0	1661	23.39
IEEE Access	64	2.60	3.6	1773	27.70
Chemical Engineering Journal	29	1.18	13.2	468	16.14
Nano Energy	29	1.18	17.1	1020	35.17
Public Library of Science	28	1.14	2.6	663	23.68
Biosensors	27	1.10	4.9	793	29.37
Applied Sciences	26	1.06	2.1	170	6.54
Electronics	26	1.06	2.6	423	16.27

TP=Total publications; PTP=Percentage of the total publications; TC=Times cited; AC=Average number of citations per publication

Figure 5.

Journal publication trends and subject distribution (a: Journal publication trends; b: Subject distribution)

An analysis of Web of Science subject categories underscores the field’s inherently interdisciplinary character. Research output is predominantly clustered in three domains: “Electrical and Electronic Engineering,” “Health Care Sciences and Services,” and “Materials Science” [Figure 5b]. This distribution mirrors a complete translational value chain: ranging from foundational sensing materials and components (Materials Science), through intermediate system integration and data acquisition (Electrical and Electronic Engineering), to terminal clinical applications and service delivery (Health Care Sciences and Services). This structure reflects a synergistic pathway bridging “technology R and D, product realization, and clinical implementation.”

The top 10 most highly cited publications

The landscape of the most influential research, as defined by citation frequency, reveals the interdisciplinary and evolving intellectual core of the wearable health device field. A prominent early foundation is evident, with Pantelopoulos et al.’s (2010)[12] seminal work in IEEE Transactions on Information Technology in Biomedicine establishing a highly cited cornerstone in computer science applications. The list is subsequently dominated by publications in high-impact, broad-scope review journals, particularly Nature Reviews Cardiology and Nature Reviews Disease Primers, indicating the field’s maturation and its significant clinical translation, especially in cardiology. This contrasts with foundational papers focused on engineering, software, and AI, as seen in Software: Practice and Experience and Artificial Intelligence in Medicine.

The distribution of Journal Citation Report (JCR) categories further underscores this convergence. While computer science and engineering categories remain vital, the strong presence of clinical and biomedical categories (Cardiac and Cardiovascular Systems, Medicine/General and Internal) highlights the shift from technological development to clinical validation and application. The inclusion of a materials science-focused paper (Advanced Materials) points to the ongoing innovation at the hardware level. Collectively, this citation topology maps the field’s trajectory from its computational and engineering origins toward a dominant, application-driven paradigm in clinical medicine and public health [Table 4].

Table 4.

The top 10 cited articles on the wearable device in health management

Author, Year	Journal	GTC	JCR Category
Pantelopoulos, A, 2010	IEEE Transactions on Information Technology in Biomedicine	1502	Computer Science, Artificial Intelligence & Cybernetics & Interdisciplinary Applications
Gupta, H, 2017	Software: Practice and Experience	1004	Computer Science, Software Engineering
Evenson, KR, 2015	Int J Behav Nutr Phys Act	812	Nutrition & Dietetics; Physiology
Boulos, MNK, 2011	Biomedical Engineering Online	671	Engineering, Biomedical
Chan, M, 2012	Artificial Intelligence in Medicine	614	Computer Science, Artificial Intelligence; Engineering, Biomedical; Medical Informatics
Baird, PN, 2020	Nature Reviews Disease Primers	584	Medicine, General & Internal
Siontis, KC, 2021	Nature Reviews Cardiology	528	Cardiac & Cardiovascular Systems
Bayoumy, K, 2021	Nature Reviews Cardiology	525	Cardiac & Cardiovascular Systems
Jayathilaka, WADM, 2019	Advanced Materials	518	Chemistry, Multidisciplinary, &Physical; Nanoscience & Nanotechnology; Materials Science, Multidisciplinary; Physics, Applied; Physics, Condensed Matter
Dwivedi, AD, 2019	Sensors	468	Chemistry, Analytical; Engineering, Electrical & Electronic; Instruments & Instrumentation

GTC=global times of citation by other literatures; JCR=Journal Citation Report

Keyword analysis and intellectual structure

Keyword co-occurrence network

The keyword analysis based on CiteSpace software [Figure 6] delineates the core themes and structural relationships within the domain. “Wearable electronic devices” anchors the network as the central node, establishing its status as the pivotal research subject. Radiating from this core, high-frequency keywords form a distinctive conceptual framework. The robust linkage between “physical activity” and “management” underscores physical activity monitoring as the primary application vector. Notably, the prominence of the keyword “validation” reflects a rigorous academic emphasis on verifying data fidelity, reliability, and clinical validity—a prerequisite for medical translation. Furthermore, the deep integration of “artificial intelligence” and “machine learning” with core nodes identifies intelligent data analytics as the technological engine of the field. Concurrently, the strong association between “mobile health” and “digital health” outlines an interconnected ecosystem, wherein wearable devices serve as critical data acquisition terminals seamlessly embedded within broader digital health service architectures.

Figure 6.

Keyword analysis using CiteSpace

The VOSviewer keyword co-occurrence network reveals three core, color-coded research modules [Figure 7]. The green cluster represents the technological foundation, focusing on hardware like sensors, triboelectric textiles, and devices. The red cluster centers on daily health and wellness monitoring, encompassing general health management, physical activity, sleep, and nutrition. The blue cluster signifies the pivot to clinical translation, specifically the prevention, diagnosis, and management of chronic diseases such as cardiovascular diseases and diabetes. Crucially, AI, big data, and the IoT serve as cross-cutting enabling technologies, deeply integrated with all clusters and forming a vital bridge between hardware (green) and disease management (blue). This structure illustrates a synergistic ecosystem where advancements in sensing hardware fuel both daily wellness tracking and precision chronic disease management, driven by intelligent data analytics. The field is thereby evolving from generalized consumer monitoring toward a medical-grade, proactive health intervention paradigm.

Figure 7.

Keyword co-occurrence network analysis using VOSviewer

Keyword cluster analysis

Cluster analysis further deconstructs the field’s knowledge topology into distinct sub-domains [Figure 8]. The largest cluster, “#0 digital health,” serves as the overarching paradigm, defining the macrocontext of the research. Closely aligned clusters, “#1 health monitoring” and “#4 wearable sensors,” delineate the field’s functional objectives and technological foundations, emphasizing the reliance on high-precision, miniaturized sensing capabilities for continuous physiological tracking. In parallel, clusters, such as “#2 machine learning,” “#3 artificial intelligence,” and “#6 deep learning,” constitute a cohesive “Intelligent Analytics” module. This clustering signifies that leveraging advanced algorithmic models for deep data mining—ranging from disease risk prediction to decision support—has become a principal driver of innovation. These clusters are not isolated; their dense interconnections reflect an integrated value chain spanning “data acquisition (Sensors),” “intelligent processing (AI),” and “service delivery (Digital Health).”

Figure 8.

Keyword cluster analysis

Burst detection and frontier evolution

Keyword burst detection identifies and keyword mountain plot emerging hotspots and forecasts temporal evolutionary trends [Figures 9 and 10]. (1) The analysis reveals distinct phasic shifts in research focus. Foundational Phase: Early bursts of “physical activity” and “mobile health” marked the field’s inception, centering on general activity tracking and conceptual exploration. (2) Specialization Phase: The subsequent emergence of “fitness trackers” and “chronic disease” signaled a deepening focus on specific device form factors and chronic care scenarios. (3) Frontier Phase (2024–Present): Most notably, recent strong bursts, such as “energy harvesting,” “personal thermal management,” and “cardiovascular diseases,” indicate a bifurcation of the current research frontier. One direction aims to address critical hardware constraints (e.g., power autonomy and biointegration) to enhance user adherence; the other focuses on the precision management of major pathologies. This trajectory reflects the field’s maturation from general-purpose health promotion tools to medical-grade intervention instruments, driven by the dual engines of hardware innovation and clinical specificity.

Figure 9.

Keyword burst detection analysis

Figure 10.

Analysis of keyword peak map

Discussion

This study employs a comprehensive bibliometric analysis and knowledge mapping of 2,463 publications (1997–2025) to delineate the evolutionary trajectory, collaborative networks, and intellectual structure of wearable device research for health management. It identifies a triphasic growth pattern, a bipolar U.S.–China dominance within a fragmented global network, and a thematic evolution from general monitoring toward precision management of specific pathologies and energy-autonomous systems.

Prior bibliometric works in this domain possess notable limitations. The seminal study by Dehghani and Kim (2019) surveyed literature only up to 2016, preceding the field’s recent exponential growth phase driven by AI and IoT convergence.[13] Subsequent analyses have been narrowly focused, examining either specific user cohorts (e.g., the elderly, cardiovascular patients) or isolated aspects such as consumer interaction or biosensor hardware, without capturing the field’s holistic translational pathway from technology development to clinical application.

In contrast, the principal contribution and novelty of this study lie in its comprehensive longitudinal scope and integrative analytical perspective. By synthesizing nearly three decades of data, this work provides the first macrolevel visualization of the field’s complete developmental arc and its current strategic pivot. It uniquely highlights the critical disconnect between high publication volume and underdeveloped global collaboration, while simultaneously mapping the convergence of technological innovation (e.g., energy harvesting) with clinical specialization.[14] These insights offer an evidence-based framework to guide future interdisciplinary synergy and the development of interoperable data ecosystems, addressing a significant gap in the existing bibliometric literature.

Research evolution and collaborative landscape

The bibliometric analysis presented herein, based on high-impact datasets from the WoSCC, provides a comprehensive panorama of the developmental trajectory and global landscape of wearable health management research. The evolution of this field exhibits distinct phasic characteristics. The period from 1997 to 2013 represents a nascent incubation stage, characterized by low publication volume (<10 annually) and a focus on conceptual validation.[15,16] A paradigm shift occurred post-2014, accelerating into an exponential growth phase after 2018. This surge, catalyzed by the maturation of IoT and AI technologies, marks the field’s transition from theoretical exploration to applied R and D and large-scale deployment. Given this trajectory, the academic output and influence of this domain are projected to maintain a robust upward momentum.[17,18]

However, scientific advancement depends on intellectual exchange. The network analysis reveals a dichotomy. While the USA and China dominate global output as dual cores, the overall international collaborative network remains fragmented and sparse. A truly integrated global research community has yet to coalesce. This decentralized topology is incongruent with the complexity of wearable health R and D, which demands cross-disciplinary synergy across precision sensing, algorithmic modeling, and clinical validation. Addressing global health challenges necessitates the pooling of distributed scientific capital to foster comprehensive innovation.[19,20]

Consequently, transcending geographical and institutional silos is imperative for the field’s leapfrog development. Specifically, while the USA and China exhibit comparable research volumes, research has found that China exists substantial latency for deeper collaboration.[21] Future initiatives should shift from a competitive stance to a collaborative synergy, leveraging complementary strengths—such as engineering innovation, clinical resources, and market scalability—to address shared health imperatives. Strategies, including joint funding programs, cross-border clinical data sharing, and the co-development of ethical frameworks and technical standards, are essential.[22] By collectively tackling universal bottlenecks such as device interoperability and clinical efficacy, the global research community can elevate the discipline to new heights.[23]

Research hotspots

The combined keyword and thematic cluster analyses in this study reveal that current research in the field of wearable devices for health management has formed several distinct and interconnected hotspot clusters. Deep exploration is primarily centered around three axes: multidimensional data acquisition, intelligent information processing, and proactive health intervention. Through keyword co-occurrence and cluster analysis, this study identifies three interconnected pillars driving the field’s deepening exploration: multidimensional data acquisition, intelligent information processing, and proactive health intervention.

Primarily, high-precision, multimodal physiological sensing constitutes the foundational bedrock of the field. The research focus has evolved from tracking basic metrics (e.g., heart rate, activity) via general-purpose devices to the precise, longitudinal capture of specific biomarkers. Emerging biointerfaces, such as microneedle patches and flexible electronic skins, now enable noninvasive or minimally invasive analysis of biofluids (e.g., sweat, interstitial fluid). This capability allows dynamic monitoring of glucose, lactate, and drug metabolites, opening new frontiers for metabolic disease management.[24,25] Concurrently, neuro-monitoring tools, such as EEG headbands, facilitate the objective assessment of cognitive decline and mood disorders.[26] Triboelectric textiles have recently garnered significant attention for their ability to detect and analyze body movements.[27] The continuous, noninvasive, and multidimensional data acquisition capability of wearable devices lays a solid foundation for building panoramic digital health profiles that cover daily wellness and disease management.

Secondly, the synergistic integration of AI is reshaping the value extraction pathway. Traditional analytical methods are often insufficient for the heterogeneous, time-series data generated by wearables. AI has shown excellent performance in the early identification of anxiety,[28] improving the accuracy of disease diagnosis, generating personalized treatment plans, enhancing patient compliance, and the full-course management of chronic diseases.[29] AI-driven wearable devices, through the collection of continuous data, can reduce the readmission rate of patients with chronic diseases, such as diabetes and cardiovascular diseases, by up to 30%. The integration of AI-driven wearable devices can increase patients’ self-management behaviors by 20%.[30] Personalized treatment is another core area of advantage for explainable AI. A study has shown that AI can improve the effectiveness in the health outcomes of patients with chronic diseases.[31] This intelligent layer not only empowers evidence-based personalized recommendations but also enhances clinical decision-support systems.[32] Crucially, AI “democratizes” sophisticated diagnostic capabilities to portable terminals, potentially mitigating reliance on centralized medical infrastructure. This shift offers high-quality remote management solutions for resource-constrained regions, promoting health equity and universal accessibility.[33]

Ultimately, the convergence of sensing and AI has catalyzed the “Proactive Health” paradigm. Wearable devices are now deeply embedded within the closed-loop disease management cycle. In chronic care, they provide continuous, out-of-hospital monitoring for cardiovascular and diabetic patients, significantly enhancing early warning capabilities.[29] In rehabilitation, sensors quantify subtle indicators like gait and tremor to map recovery trajectories.[34] This transition from a “treatment-centered” model to an integrated “prevention-intervention-management” ecosystem highlights the core value of wearables in optimizing healthcare efficacy and patient quality of life.[35]

Research trends

Synthesizing current bottlenecks and frontier dynamics, future development is projected to bifurcate along two strategic axes: overcoming technological barriers to achieve medical-grade reliability and fostering ecosystem integration for clinical translation.

The primary technological imperative is to enhance data fidelity and energy autonomy. Despite widespread adoption, limitations in measurement accuracy and battery life currently hinder the transition of wearables from consumer gadgets to professional medical instruments.[36] Next-generation sensing must transcend simple data integration to achieve clinical-grade precision through novel materials. Simultaneously, user adherence remains a critical challenge. Frontier research in energy harvesting—such as triboelectric nanogenerators—offers a transformative solution by converting biomechanical or thermal energy into electricity.[37] This self-powered modality fundamentally addresses the “charging anxiety” pain point, thereby enabling unobtrusive, long-term monitoring essential for elderly care and chronic disease management. However, hardware innovation alone is insufficient without a robust translational ecosystem. This study reveals a disconnect between academic output (universities) and industrial/clinical implementation. A significant trend involves constructing a “Quadruple-Helix” collaboration model spanning basic research, engineering, clinical validation, and industrialization. This requires deep dialogue to align R and D with genuine clinical needs and to establish rigorous standards for efficacy evaluation.[37]

Furthermore, the ultimate hurdle for systemic implementation is data fragmentation. Current “data silos” prevent the seamless integration of wearable data with Electronic Health Records, limiting their utility in Real-World Evidence research. Therefore, a critical trajectory involves regulatory bodies, leading to the establishment of unified interoperability standards and securing data governance frameworks. Building regional or specialty-specific platforms that bridge device manufacturers and healthcare providers is inevitable for unleashing the full potential of wearable data, driving the evolution toward evidence-based personalized medicine.[38]

Conclusion

By synergizing bibliometric analysis with CiteSpace knowledge mapping, this study systematically deconstructs the evolutionary trajectory of wearable devices for health management from 1997 to 2025. Drawing upon a curated dataset of 2,463 publications from the Web of Science Core Collection, the findings identify a pivotal exponential growth phase post-2018. The global research landscape is currently anchored by a bipolar dominance of the USA and China, with academic institutions serving as the primary engines of innovation. The field’s intellectual structure has crystallized around three strategic axes: high-fidelity physiological data acquisition, AI-driven intelligent analytics, and clinically oriented proactive health applications. Notably, frontier trends reveal a critical pivot: a shift from generalized health monitoring to the precision management of specific pathologies (e.g., cardiovascular diseases), and a technological transition from external power dependence toward energy-autonomous systems. Despite this momentum, the domain faces structural impediments. The global collaborative network remains notably fragmented, lacking deep transnational and interinstitutional synergy. Furthermore, significant translational barriers persist, specifically the insufficiency of clinical-grade validation and the prevalence of data silos. Future progression necessitates a paradigm shift toward a “Quadruple-Helix” innovation model (integrating industry, academia, research, and medicine). Establishing unified interoperability standards and shared data governance frameworks will be paramount to elevate the field from disparate technological breakthroughs to scalable, evidence-based personalized health ecosystems.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

This study was financially supported by the National Nature Science Foundation of China (No. 82404225, 81972997).