Transcriptomic decoding of dynamic brain activity alterations in chronic non-specific low back pain

Investigating dynamic brain fluctuations helps reveal changes in neural activity patterns associated with disease. Chronic non-specific low back pain (CNLBP) is characterized by altered brain function, yet less is known about the dynamic changes in brain activity and their relationship with regional gene expression profiles.

Based on resting-state functional MRI data, a hidden Markov model (HMM) was applied to identify recurrent whole-brain activity patterns, referred to as HMM states, in 80 patients with CNLBP and 80 healthy controls (HCs). Additionally, partial least squares analysis was employed to examine the relationships between atypical neural dynamics characteristics of CNLBP and transcriptomic data from the Allen Human Brain Atlas. Enrichment analyses were subsequently performed to explore the potential biological pathways associated with aberrant dynamics in CNLBP.

Eight HMM states representing distinct spatiotemporal patterns of neural activity were identified. Fractional occupancy (FO), lifetime (LT), and mean dwell time (MDT) are HMM metrics that describe the time spent in each brain state, and transition probability (TP) and switching rate (SR) are metrics that capture the transition patterns between states. Compared with HCs, patients with CNLBP exhibited decreased MDT in state 3, increased MDT in state 6, and reduced FO, LT, and MDT in state 7. Additionally, the FO of state 7 was negatively correlated with clinical measures. CNLBP state 7 exhibited prominent abnormal temporal dynamics, characterized by higher functional activity in the salience and sensorimotor networks, together with lower functional activity in the default mode network. Patients with CNLBP also showed significantly lower SR and altered TP, reflecting the presence of biased transition patterns. Furthermore, the mean activation of state 7 in CNLBP was significantly associated with gene expression profiles. Enrichment analysis of the 1114 genes associated with aberrant dynamic brain state highlighted nociceptive and inflammatory pathways.

This study elucidates the complex brain dynamics in CNLBP and links the activation of HMM state 7 with atypical temporal features to gene expression levels. These findings provide novel insights into the characterization of dynamic brain activity in CNLBP and reveal an association between neural configurations and gene expression.