Profiling trait and state-like behaviour in pain modulation systems: a role for effective connectivity within endogenous pain control circuits

Endogenous pain modulation systems can be assessed through distinct psychophysical paradigms such as conditioned pain modulation (CPM), temporal summation of pain (TSP), and offset analgesia (OA). Notably, the reliability of these measures has rarely been defined within the same participants, and measures to index the consistency of each measure across sessions and in the same participants are lacking. This study examined the test-retest reliability and intra-individual consistency of CPM, TSP, and OA and explored how CPM response status across sessions relates to neural dynamics within the descending pain modulation system. In 29 healthy participants, CPM, TSP, and OA responses were assessed across two sessions. The normalised session change index (NSCI) was introduced to evaluate consistency of each measure across sessions. Spectral dynamic causal modelling (DCM) of resting-state fMRI (rs-fMRI) investigated effective connectivity within the descending pain modulation network and its association with CPM response status. OA exhibited the highest test-retest reliability and had NSCI values closest to zero, indicating stable responses. CPM and TSP showed poor reliability and NSCI values deviating from zero, reflecting greater variability. Spectral DCM analysis revealed that effective connectivity within the descending pain modulation system explained CPM response variability across sessions. Specifically, consistent strong facilitatory or inhibitory CPM was associated with greater excitatory or inhibitory PAG-to-AI effective connectivity, respectively. These findings suggest that healthy participants either demonstrate stable (i.e. trait-like) or dynamic (i.e. state-like) endogenous pain modulation across repeated sessions which can be explained based on effective connectivity within the descending pain modulation system.