Continuous glucose monitoring, once a technology confined to the management of established diabetes, has become one of the fastest-growing research topics in prediabetes, according to a new bibliometric analysis that maps the entire scientific literature linking the two fields. The study, published in Frontiers in Clinical Diabetes and Healthcare, identified 176 publications spanning more than three decades and found that output accelerated sharply after 2020, reaching 44 publications in 2025 alone. The authors, Jaime Arturo Ruiz Escamilla and Luis Enrique Colmenares-Guillen of Benemerita Universidad Autonoma de Puebla and the Mexican Institute of Social Security, describe a literature that has grown from scattered early reports into a dense conceptual network centered on glucose dynamics, glycemic variability, and digital health.

The analysis is explicitly a mapping exercise rather than a clinical trial or systematic review of effectiveness. The researchers searched PubMed and Scopus with a cutoff reported as May 20, 2026, combining CGM-related terms such as continuous glucose monitoring, flash glucose monitoring, and interstitial glucose monitoring with prediabetes-related terminology including impaired fasting glucose, impaired glucose tolerance, and intermediate hyperglycemia. After removing 319 duplicate records from 931 initial database hits, 612 unique records underwent title and abstract screening, 400 were excluded, and 212 reports proceeded to full-text assessment. Two reviewers independently evaluated every full text, reaching identical decisions in all cases, and one record was later reclassified as excluded because it involved an exclusively pediatric population. The final corpus contained 176 publications.

Eligible studies were peer-reviewed original research or review articles that substantively addressed CGM or related interstitial glucose sensor data in relation to prediabetes or intermediate hyperglycemic states in adults. The authors restricted inclusion to English- and Spanish-language publications because those were the languages in which both reviewers could reliably perform independent assessment, a decision they acknowledge as a potential source of language and geographic bias. No restrictions were applied at the database level for year, country, population, or study design, and the search strategy was deliberately sensitivity-oriented, with broad CGM terms always paired with the prediabetes concept block.

The corpus tells a clear story of accelerating interest. Publications remained limited through the early years of the period studied, which stretches from 1994 to 2026, and increased progressively over the last decade. Nine publications appeared in 2020, followed by 11 in 2021, 12 in 2022, 19 in 2023, 29 in 2024, and 44 in 2025. Fourteen publications were identified for 2026, but the authors caution that this represents an incomplete publication year at the search cutoff and should not be read as a decline in activity. The 176 included publications were distributed across 95 scientific sources and involved 1,221 authors, with only four publications single-authored.

Where the work is published is as revealing as how much of it exists. The leading sources were the Journal of Diabetes Science and Technology with 22 publications and Diabetes Technology & Therapeutics with 13, followed by Diabetes Research and Clinical Practice with 9, Nutrients with 6, and the Journal of Clinical Endocrinology and Metabolism with 5. The authors note that this concentration in diabetes-technology and metabolic journals characterizes where research activity has accumulated but does not by itself establish the causes of that concentration. Citation-dependent indicators were calculated separately for the 169 publications indexed in Scopus, which contained 6,858 cited references and averaged 18.34 citations per document at the May 19, 2026 snapshot.

The geographic distribution of the literature is markedly uneven. In the Scopus subset, the corresponding author was based in the United States for 63 publications, Japan for 15, China for 10, and Germany for 9. International co-authorship stood at 22.49 percent. The authors suggest that differences in research infrastructure, funding, device availability, and digital-health capacity are plausible contextual explanations for this pattern, but they emphasize that these mechanisms were not measured and cannot be established by a bibliometric design. The language restriction may further reduce representation from settings where relevant research appears in other languages.

To understand the conceptual structure of the field, the researchers harmonized author keywords using a documented synonym dictionary, merging only clear lexical or acronym equivalents such as CGM variants into continuous glucose monitoring and T2D or T2DM into type 2 diabetes, while keeping clinically related but non-equivalent concepts like prediabetes and impaired glucose tolerance separate. The most frequent harmonized keywords were continuous glucose monitoring with 90 publications, prediabetes with 67, diabetes with 25, glycemic variability with 23, type 2 diabetes with 21, HbA1c with 20, and obesity with 19. The resulting keyword co-occurrence network contained 46 connected terms organized into four clusters, with the largest, comprising 26 terms, centered on continuous glucose monitoring and prediabetes and extending to glycemic variability, time in range, machine learning, wearables, insulin resistance, and cardiovascular risk.

A second nine-term cluster grouped diabetes, glucose, hyperglycemia, impaired glucose tolerance, diet, nutrition, insulin, and postprandial glucose, while a third clustered HbA1c, obesity, oral glucose tolerance test, overweight, weight loss, fructosamine, and glycated albumin. A fourth cluster consisted of dietary intervention and weight management. A separate thematic map built from 250 candidate words contained 64 terms in nine clusters, with the dominant cluster labeled continuous glucose monitoring and including prediabetes, glycemic variability, HbA1c, and related metabolic concepts. The authors interpret quadrant positions in this map as bibliometric properties of density and centrality rather than evidence that any topic is clinically established or clinically emerging.

Temporal keyword patterns reinforce the picture of a field in rapid evolution. Continuous glucose monitoring appeared in 25 of 58 publications during 1994 through 2021, 29 of 60 during 2022 through 2024, and 36 of 58 during 2025 and 2026. Prediabetes rose from 12 to 24 to 31 publications across the same periods, and type 2 diabetes terminology jumped from 4 to 3 to 14. Technology-oriented terms such as machine learning and wearables appeared predominantly in later periods, with machine learning absent before 2022 and wearables appearing in only one publication in the middle period before reaching three in the most recent one. The authors stress that these counts describe changes in keyword representation and do not measure clinical effectiveness or maturity.

The study’s central caution is that mapping research activity is not the same as demonstrating clinical value. The authors state plainly that the analysis does not establish whether CGM should be used routinely in prediabetes, whether it prevents type 2 diabetes, whether it improves behavior or risk stratification, or whether it is cost-effective. Those questions, they write, require appropriately designed prospective clinical trials, comparative studies, and health-economic evaluations. They also note that observations about psychological and behavioral effects of glucose feedback, including increased nutritional awareness alongside data overload and emotional burden reported in a 2026 review of people with type 2 diabetes, arise from a different clinical population and cannot be generalized to prediabetes or attributed causally to CGM. The bibliometric study does not assess risk of bias, clinical effectiveness, behavioral effectiveness, or cost-effectiveness, and it depends on the completeness and consistency of database metadata. What the analysis does establish is where scientific attention has concentrated: a literature that has expanded substantially since 2020, organized around a core linking CGM, prediabetes, glycemic variability, and related metabolic and digital-health topics, with clear geographic concentration and an explicit agenda of unanswered clinical questions for future trials to address.

Subject of Research: A bibliometric analysis mapping the scientific literature on continuous glucose monitoring in prediabetes and intermediate hyperglycemic states

Article Title: Continuous glucose monitoring in prediabetes: a bibliometric analysis of research trends, thematic mapping, and evidence gaps

Article References: Ruiz Escamilla, J. A., & Colmenares-Guillen, L. E. (2026). Continuous glucose monitoring in prediabetes: a bibliometric analysis of research trends, thematic mapping, and evidence gaps. Frontiers in Clinical Diabetes and Healthcare, 7, Article 1953753. https://doi.org/10.3389/fcdhc.2026.1953753

Image Credits: AI Generated

DOI: 10.3389/fcdhc.2026.1953753

Keywords: continuous glucose monitoring, prediabetes, bibliometric analysis, glycemic variability, type 2 diabetes, digital health, wearables, machine learning, impaired glucose tolerance, time in range, research trends, evidence gaps