Supervised Exercise Intervention Restores Physical Activity and MultiOMIC-Derived Frailty Biomarkers in Older Adults

The molecular mechanisms underlying frailty are under intense investigation. Independent observational studies revealed various biological processes and biomarkers differentially expressed in frail individuals and after interventions. Recently, several transcriptomic approaches have described molecular patterns associated with frailty status. However, the effects of physical exercise interventions remain unexplored. Two studies (FRAILOMIC and BIOFRAIL) identified molecular signatures associated with frailty. The objective of the study was to analyse the expression of the FRAILOMIC and BIOFRAIL biomarkers following a multicomponent personalized exercise intervention in older adults.

A multi-centre study included a control (n = 24, with 14 women and 10 men, median age 80 ± 5) and an exercise intervention group (n = 44, with 38 women and 6 men, median age 79 ± 4). The latter underwent a 16-week multicomponent presential physical exercise programme; clinical, functional and frailty assessments were performed before and after the intervention in both groups. Latent class analysis identified 'responders' and 'non-responders' to the intervention. Data in a testosterone-suppressed control cohort of older men with prostatic cancer receiving androgen deprivation therapy (n = 21) were included. Molecular biomarkers were measured via qRT-PCR and ELISA.

The exercise group presented improvements in functional capacity, as shown by Short Physical Performance Battery (SPPB) (post- 10.1 vs. 9.1 in pre-intervention), grip strength (18.3 vs. 16.9 kg), gait speed (0.85 vs. 0.78 m/s), sarcopenia (2.1 vs. 3.0) and frailty status by Fried Frailty Phenotype (1.6 vs 2.1 criteria) (all ≤ 0.004). ANCOVA revealed significant improvements in SPPB (p = 0.04) and grip strength (p < 0.005) with exercise. Four of the six FRAILOMIC biomarkers (miR125b, miR194, RAGE and Troponin) and six of seven BIOFRAIL biomarkers (EGR1, CXCL8, GOS2, NSF, DDX11L1 and miR454) reversed significantly their expression after exercise training (p ≤ 0.05 for FRAIOLMIC and ≤ 0.01 for BIOFRAIL respectively). In contrast, these changes were not observed in testosterone-suppressed controls, who experienced muscle loss. Latent class analysis revealed 36 'responders' (with 32 cases from intervention group) and 29 'non-responders'. Logistic regression showed that miR454, EGR1, DDX11L1 and Troponin displayed the strongest predictive value in distinguishing responders from non-responders.

Exercise intervention improved functional status and reversed biological frailty, as evidenced by changes in multiple MultiOMIC frailty biomarkers. Among these, the BIOFRAIL trio-miR454, EGR1 and DDX11L1-showed the strongest predictive power for intervention response with AUC of 0.83. This study provides proof-of-concept for the utility of OMIC-derived frailty biomarkers as outcome measures after intervention.