Identification of Potential Blood-Based Biomarkers for Frailty by Using an Integrative Approach

We have keenly read the article by Suganuma et al. [1] published in issue 70 June 2024 of your esteemed journal. We would like to applaud the authors for their in depth analysis of potential blood-based biomarkers for frailty. The study has identified these biomarkers with an integrative approach. Physical frailty (PF) and sarcopenia are “twin” entities which are intrinsically complex and have definitional ambiguities [2]. For the diagnosis of frailty, currently the Fried criteria, Clinical Frailty Scale, and Frailty Index are used widely [3]. However, the Fried criteria lack data on neurological, cognitive status and psychosocial components of frailty. Likewise, the Frailty Index lists a long list of deficit assessments which cannot be utilized practically in clinical settings. The Clinical Frailty Scale utilizes clinical judgment of the physician for diagnosis of frailty, which may be subjective [3]. Considering these limitations, the need for developing biomarkers is of the utmost importance for diagnosis of frailty [4]. The biomarkers for PF and Sarcopenia available currently capture only single aspects of frailty and thus are not in association with outcomes which can be utilized clinically [2]. Thus, a new validated approach for identification of biomarkers moves from the age-old paradigm of “one fits all” toward a multivariate methodology [2].The definition of an ideal biomarker includes supporting the diagnosis, facilitating the tracking of the illness over time and should also enable healthcare professionals in clinical and therapeutic decision-making [2]. The study by Suganuma et al. [1] has candidate biomarkers with statistically significant co-relation with components of Japanese version of Cardiovascular Health Status (J-CHS) frailty diagnostic criteria. The Japanese version of the frailty phenotype is a translated and culturally adapted version of the original frailty phenotype criteria developed. It assesses five components: weight loss, exhaustion, physical activity, walk time, and grip strength [5]. Also, the development of risk prediction models achieved higher area under curve at the time of validation [1]. This is indeed a laudable achievement.The study by Suganuma et al. [1] mentions the detection of clinical biomarkers like skeletal muscle index by use of dual energy X-ray absorptiometry, but these imaging equipments are not immediately accessible in primary health care, being the first point of contact of most frail elderly. These imaging techniques are also expensive, making it difficult to be used at the grass-root level especially in developing and under-developed nations [2]. The clinical biomarkers like systolic and diastolic BP and heart rate have physiological changes with age. Due to this and the presence of co-morbidities in an individual, measuring the vital signs at a single point has less sensitivity, whereas the same, if done serially, can increase the sensitivity. The measurement of vital signs changes subtly because of reduced physiologic ranges, although change from an individual reference range may indicate important warning signs. Hence, individualized reference range may provide increased sensitivity in frail, older patients [6]. This study could not find any data from RNA-seq analysis due to a small sample size [7]. A review article by Dato et al. [7] emphasizes the importance of mi-RNA for diagnosis of physical and cognitive domains of frailty. They have identified 57 mi-RNA’s associated with physical phenotypes and 43 mi-RNA’s associated with cognitive frailty. The RNA sequencing analysis uses peripheral blood mononuclear cells and thus is an easily accessible and non-invasive method for the diagnosis of frailty in future [1]. A study by Murabito et al. [8] demonstrates the relation between mi-RNA and hand grip strength which declines with age and muscle disease. Similarly, the study by Denham and Prestes [9] measured the levels of mi-RNAs in the blood of athletes to determine cardio-pulmonary fitness. However, the study of miRNA as a potential biomarker for diagnosis of frailty is still in its infancy and cost-effectiveness and use in clinical practice is limited [8‒10].We would like to bring attention to another method to improve the identification of frailty, i.e., the use of artificial intelligence (AI) techniques. In a study by Ambagtsheer et al. [11], the use of AI in identifying frailty in residential aged-care is evaluated. In an early study on older Canadian population, it was found that artificial neural network outperformed self-reported Frailty Index to predict survival of the frail-aged population [11]. Machine learning-based AI can be useful in the identification the future frailty conditions, as well as the risk of re-admission in hospitals of these patients by use of both clinical and socio-economic variables than can be collected in centers for healthcare [3]. The utility of AI in clinical practice needs to be monitored by weighing the administrative burden, apprehension, and potential benefit of AI for detection of frailty [11].After a review of the existing and potential methods for diagnosis of frailty, we find that the risk prediction models proposed by Suganuma et al. [1] along with newer techniques like AI can detect frailty early and thus improve the quality of life and decrease incidence of frailty [3]. However, all these developments have to be easily accessible and cost-effective so that they can be utilized to reach the maximum population for diagnosis and further also assess the severity of frailty.The authors have no conflicts of interest to declare.This study was not supported by any sponsor or funder.Dr. Shruti Karnik, Dr. Anu Gaikwad, Dr. Harishchandra Chaudhari, and Dr. Priyanka Khopkar-Kale: manuscript review and letter preparation.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.