The rapid growth of the global older adult population has made frailty an emerging challenge in geriatric medicine and public health [1]. Frailty is a clinical syndrome, distinct from disability or comorbidity, characterized by reduced physiological reserves. This reduction increases vulnerability to stressors and results in functional decline in daily activities [2,3].

Frailty in community-dwelling older adults has been associated with adverse clinical outcomes, including falls, disability, hospitalization, and death [4,5]. It also negatively affects the health of individuals with cardiovascular disease, infectious diseases (eg, COVID-19, pneumonia), or cancer because it strongly predicts treatment intolerance [6], more frequent complications [7], and increased mortality [8]. These outcomes emphasize the clinical importance of timely frailty identification in older adults and suggest that biomarkers may improve early detection strategies [9,10].

Although frailty is clearly associated with adverse outcomes, routine detection remains challenging because of subtle or nonspecific presentations and time constraints during assessments, particularly in older adult care [8,11-13]. The multisystem and dynamic nature of frailty reduces the effectiveness of single-indicator approaches. This highlights the need for objective, scalable biomarkers that enable rapid and reproducible identification. In recent years, the unmet need has led to increased interest in biological markers that may offer a more accurate and less subjective approach to predicting, diagnosing, and monitoring frailty in older adults [14-16]. Several systematic reviews and meta-analyses have shown that the levels of cytokines, including interleukin-6, tumor necrosis factor-α, and C-reactive protein, are consistently elevated in frail older adults [10,17]. Beyond cytokines and inflammation, interest has increased in hormonal, metabolic, and molecular markers, as well as multimarker panels. Omics-based approaches (proteomics and transcriptomics) have been proposed to more accurately capture the heterogeneous and dynamic trajectories of frailty than any single biomarker. European consortia such as the FRAILOMIC initiative are conducting comprehensive profiling of genomic, proteomic, and clinical biomarkers to improve the prediction of frailty onset and progression, as well as the potential therapeutic response [18,19].

Despite growing interest in frailty biomarkers, most previous studies have been cross-sectional, primarily comparing biomarker levels between frail and robust individuals at a single point in time [17,20]. These explorations have highlighted inflammatory and metabolic dysregulation; however, because they rely on static measurements, they cannot capture biomarker dynamics or clarify whether these changes precede or result from frailty progression. This limitation reinforces the need for longitudinal evidence to delineate causal pathways and identify biomarkers that predict frailty trajectories.

Clinical inconsistency in frailty definitions, biomarker assays, and analytical techniques has further increased heterogeneity across studies [14,19]. For example, frailty has been operationalized using different instruments such as the Fried phenotype (which focuses on physical manifestations) and the Frailty Index (which aggregates deficits across multiple domains), leading to discordant classifications of the same individuals [14]. Similarly, inflammatory biomarkers such as interleukin-6 and C-reactive protein have been measured using assays with varying sensitivities (eg, standard enzyme-linked immunosorbent assay vs high-sensitivity assays) and inconsistent cutoff values, complicating cross-study comparisons [19]. These discrepancies have impeded the comparability and replication of the findings. The performance of biomarkers has been erratic because sensitivity and specificity have been affected by complicating factors, including comorbidity, drug use, and population variation. Large-scale longitudinal studies remain limited, which restricts our understanding of dynamic biomarker trajectories and their predictive value [21,22]. Consequently, no single biomarker panel has demonstrated sufficient robustness for routine clinical use in identifying frailty in older adults [23]. These gaps indicate the need for a systematic synthesis of longitudinal evidence to clarify consistencies, identify weaknesses, and guide future research.

This scoping review will systematically map longitudinal evidence on biomarkers associated with frailty in older adults. Existing evidence has identified the relevance of inflammatory and molecular markers; however, most studies have focused on single biological pathways or relied on cross-sectional associations. In contrast, we aim to describe findings across a broader range of biomarker types to reflect the biological complexity of frailty. Rather than following a hypothesis-driven approach, we will present a structured framework for interpreting and classifying diverse biomarkers (eg, inflammatory markers, hormonal mediators, genetic signatures, imaging-based measures) evaluated across diverse longitudinal study designs (eg, prospective cohort studies, follow-up assessment, and nested case-control analyses) and various demographic groups (eg, sex/gender, age strata, and clinical subpopulations such as heart failure or chronic kidney disease cohorts). This synthesis aims to identify clinically promising biomarkers, describe consistent trends, reveal key knowledge gaps to inform future research, and provide evidence-based suggestions for developing personalized frailty care strategies.

We will conduct a scoping review to identify and review biomarkers with longitudinal associations with frailty, following the Joanna Briggs Institute (JBI) Manual for Evidence Synthesis guidelines [24]. We will report the review in accordance with the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses Extension for Scoping Reviews) guideline to ensure transparency and methodological rigor [25]. The review questions are as follows:

The review will include studies that meet the following criteria, structured according to population, concept, and context formats.

We will import all retrieved citations into Covidence (Veritas Health Innovation Ltd, Melbourne, Australia) for reference management, with duplicates automatically removed. Two independent reviewers (KS and YY) will screen titles and abstracts to assess eligibility according to predefined inclusion and exclusion criteria (Textbox 1). The same reviewers will then independently review the full texts of potentially relevant articles. We will resolve any disagreements through discussion; if consensus cannot be reached, a third reviewer (JWS or SH) will make the final decision. We will record reasons for excluding studies at the full-text stage. We will document and report the entire process using a PRISMA flow diagram [29].

Two independent reviewers will extract data using a standardized form developed for this review. Extracted information will include study characteristics (author, year of publication, country, and study design), participant details (age and frailty assessment tools used), types of biomarkers measured at baseline, specimen used for measurement, follow-up duration, statistical or analytical methods, and key findings on the longitudinal association between biomarkers and frailty. When annual (or other interval) assessments are available, each follow-up period will be explicitly recorded to capture how biomarker trajectories and frailty status change over time, enabling clearer interpretation of the temporal relationships reported across studies.

We will refine the data extraction form as needed during the extraction process to ensure comprehensive capture of information relevant to the review questions. Discrepancies between the two reviewers (KS and YY) will be resolved through discussion. If consensus cannot be reached, a third reviewer (JWS) will make the final decision. For unclear or missing data, we will record “not reported” in the data extraction sheet. We will not apply imputation or substitution. These gaps will be acknowledged in the narrative synthesis.

We will provide a narrative synthesis to describe how the findings align with the objectives and research questions of the review, following [30]. In the results section, we will synthesize the identified biomarkers and their longitudinal associations with frailty. We will stratify results by study population, including community-dwelling older adults and patients with metabolic disorders, musculoskeletal conditions, neurological diseases, cancer, infectious disease, and cardiovascular disease. Biomarkers will be categorized by specific lineage: inflammatory (eg, C-reactive protein, interleukin-6, tumor necrosis factor-α), metabolic (eg, glucose, hemoglobin A_1c, lipid profiles), hormonal (eg, testosterone, sex hormone–binding globulin, insulin-like growth factor I), and other groups. We will create a summary figure to present the initial and follow-up periods for each study, enabling comparison of the longitudinal effects over time.

We will conduct this review using the JBI critical appraisal tools for cohort studies [31]. In this review, we will consider biomarkers as the exposure and frailty as the outcome. We will present the results of the critical appraisal in a summary table, which will show the total number and percentage of “yes” responses for each study. As this study is a scoping review, all relevant studies will be included regardless of methodological quality. However, we will provide a summary of the number of high-quality studies in the current evidence base.

Because this study relied exclusively on publicly available published and gray literature, ethics committee approval was not required.

Through this scoping review, we expect to identify and synthesize biomarkers of frailty that reflect the sequential influence of biological factors on frailty progression in older adults. We hypothesize that multiple heterogeneous biomarkers, including inflammatory, metabolic, cardiovascular, endocrine, and neurologic markers, exist. However, we also expect substantial variability in biomarker types, frailty definitions, follow-up durations, and statistical and analytic approaches across studies.

Previous reviews suggested albumin, hemoglobin, interlukin-6, and tumor necrosis factor-α as potential biomarkers for frailty and sarcopenia and reported inflammatory markers such as leukocyte, lymphocytes, C-reactive protein, interleukin-6, interleukin-10, and tumor necrosis factor-α as a biomarker for frailty in a cross-sectional study [10,32]; however, few longitudinal studies have been conducted, leaving a knowledge gap about the potential use of biomarkers of frailty for early detection and prevention. This review will address the range of biomarkers of frailty that precede frailty incidence or progression and overview existing findings by biological process and biomarker source, as well as follow-up duration and analytical and statistical methods.

This review addresses the challenge of systematically examining the complex biological changes that precede frailty and evaluating their potential clinical relevance. By integrating current evidence without restricting the scope to predefined biomarker categories, this work captures the multifaceted biological processes underlying frailty and reflects its dynamic and systemic nature. Importantly, this approach allows for a more comprehensive understanding of frailty-related biological alterations across diverse physiological systems.

A key strength of this review lies in its inclusive exploration of heterogeneous biomarker types and analytical strategies. We examined a broad spectrum of biomarkers and analytical methodologies rather than focusing on a single class, thereby providing insight into how different approaches contribute to biomarker discovery and interpretation. This approach also enabled a critical discussion of methodological considerations, including study design, analytic frameworks, and appropriate follow-up durations needed to meaningfully capture longitudinal changes in frailty status.

Furthermore, by incorporating not only blood-based biomarkers but also neurologic markers such as fMRI, this review highlights the multidimensional nature of frailty. This broader biological perspective is particularly relevant for understanding cognitive frailty and other neurobiological dimensions of frailty that extend beyond traditional physical or inflammatory markers, thereby supporting a more holistic and personalized framework for frailty prediction and management.

Despite these strengths, several limitations should be acknowledged. As noted, most included studies were observational or cohort-based, limiting causal inference. In addition, substantial heterogeneity in biomarker assessment methods, statistical analyses, and outcome definitions precluded meta-analytic synthesis and the identification of statistically robust biomarkers. The lack of standardized protocols further limited direct comparisons across studies. Moreover, this review did not include body composition or functional biomarkers derived from measures such as bioelectrical impedance analysis (eg, InBody), which may restrict the scope of biological domains represented. Excluding such markers may underestimate the contribution of musculoskeletal and body composition-related pathways to frailty and sarcopenia, highlighting an important area for future integrative research.

Future biomarker research should adopt a more integrative approach by combining molecular biomarkers with functional and body-composition measures such as bioelectrical impedance–derived indices, physical performance metrics, and wearable-based life-log data, while integrating multi-omics data (eg, genomics, transcriptomics, proteomics, and metabolomics) within a systems biology framework to better capture the complex biological mechanisms underlying frailty.

In addition, future studies should also identify and describe biomarker evidence specific to distinct frailty subgroups, including physical, cognitive, and affective frailty, to better reflect the heterogeneous biological pathways underlying frailty. For biomarkers that are consistently reported and show strong associations with frailty across studies, meta-analytic synthesis is needed to strengthen the evidence base and support clinical translation.

The findings of this scoping review will be disseminated through publication in a peer-reviewed journal, with manuscript submission anticipated in December 2026.

This scoping review will systematically identify and examine longitudinal biomarkers that precede frailty in older adults, providing an integrated overview of the complex biological changes underlying frailty progression. By synthesizing evidence across diverse biomarker domains and analytical approaches, this review will emphasize the potential for advancing personalized frailty prediction and management while highlighting the multifaceted and dynamic nature of frailty biology.