Resting-state fMRI using hidden Markov models reveals abnormal dynamic brain functional states in asthma

Asthma involves not only airway inflammation but also aberrant central nervous system regulation. While static functional connectivity studies have revealed brain network abnormalities in asthma patients, the transient temporal dynamics of brain functional states remain largely unexplored. To investigate brain dynamic functional connectivity characteristics in asthma patients using Hidden Markov Models (HMM) and to identify potential neurobiological markers associated with clinical symptoms. Resting-state fMRI data were acquired from an initial pool of participants, with 120 age- and gender-matched individuals (60 asthma patients and 60 healthy controls) included after stringent quality control and head-motion scrubbing. HMM was applied to identify recurring brain states based on the Schaefer-142 parcellation. We compared groups on fractional occupancy (FO), mean dwell time (MDT), and transition probabilities. Exploratory correlation analyses were performed to evaluate the relationship between HMM-derived metrics and clinical scores (ACT and pulmonary function). HMM identified nine distinct functional states. Asthma patients exhibited a significantly increased MDT and FO in State 2 (characterized by somatomotor and dorsal attention network involvement) compared to healthy controls (p < 0.05). Exploratory analysis revealed a nominal positive correlation between the MDT of State 2 and Asthma Control Test (ACT) scores (r = 0.30, p < 0.05, uncorrected), suggesting a potential compensatory role of this state in symptom monitoring. Our findings reveal altered brain state dynamics in asthma, particularly the prolonged occupancy in a sensory-attention-related state. While the brain-clinical associations are exploratory, these dynamic metrics provide novel insights into the central mechanisms of asthma and may serve as preliminary neurobiological markers for symptom control.