Cuff Algometry Induces Large Yet Variable Conditioned Pain Modulation Effects

Conditioned pain modulation (CPM) paradigms provide a proxy measure of activity in the descending pain modulatory system. Cuff-pressure-algometry offers a standardised CPM assessment tool although comprehensive validation in large samples is lacking. To address this, we pooled cuff-algometry CPM data from 324 healthy participants across 8 studies. CPM magnitude was calculated as pain detection (PDT) and tolerance (PTT) threshold changes, assessed on the dominant leg in the presence and absence of a painful "conditioning" cuff stimulus on the contralateral leg. CPM-effects were robust for both changes in PDT and PTT (p<0.001). Using a classification approach where a [&ge;]20% change in threshold designated a CPM responder, 69% of participants were CPM-responders for PDT and 59% for PTT. Test-retest reliability data were assessed in a subset of participants (n=72; interval 16.49{+/-}18.39days) using intraclass correlation coefficients (ICC). Test-retest reliability was poor for CPM-effects (ICC=0.25-0.37) despite moderate-to-good reliability for PDT and PTT (ICC=0.69-0.87). Responder classification showed none-to-minimal agreement across sessions (Cohens {kappa}=0.17-0.21), with 38% of participants switching classification for both PDT and PTT. Bootstrap analysis revealed that smaller samples provide highly variable ICC estimates, potentially explaining discrepancies with previous reliability reports. Despite producing large group-level CPM-effects, poor test-retest reliability of cuff algometry suggests it captures dynamic, state-dependent processes rather than a stable trait-like individual characteristic. This highlights the need to consider the temporal instability of CPM when interpreting data and considering its deployment within precision pain medicine.