Altered Sensory-Pain Transitions in Sickle Cell Disease Reflect Central Sensitization
Shidhartho Roy, Joel Disu, Nahom Mossazghi, Lara Abdelmohsen, Sossena WoodAbstract
Background
Adults with sickle cell disease (SCD) frequently exhibit a clinical pain phenotype characterized by earlier pain onset and a reduced separation between sensory detection thresholds and pain tolerances, consistent with features commonly attributed to central sensitization (CS) [1]. The neural mechanisms underlying this altered detection-to-pain transition remain unclear. We tested the hypothesis that increasing CS in SCD is associated with a narrowed thermal detection-to-pain dynamic range and reduced differentiation of cortical responses to detection versus pain measured with electroencephalography (EEG).
Methods
Four pain-free adults and seven adults with SCD (3 low CS, 4 high CS; categorized according to [1]) completed quantitative sensory testing (QST) measuring cold and warm detection thresholds and cold and heat pain tolerance. Participants underwent a 5-minute resting EEG baseline followed by controlled thermal stimulation of the dorsal forearm; temperature was ramped at 1 °C/s from detection thresholds to pain tolerance while EEG was recorded. To link EEG activity to the detection-to-pain transition during the temperature ramp, we quantified how multichannel EEG patterns changed as stimuli progressed from detection to pain and evaluated the spatial extent and frequency-band specificity of reproducible pain–detection differences.
Results
At rest, all individuals with SCD showed increased frontal-temporal slow-frequency power relative to controls, consistent with prior reports [2, 3]. On QST, group differences were most clearly expressed in the temperature separation between pain tolerance and detection thresholds (ΔT = T_pain – T_detect). Pain-free controls exhibited the largest separation (median ΔT = 19.1 °C), whereas individuals with SCD showed markedly reduced separation in both low CS (median ΔT = 7.2 °C) and high CS (median ΔT = 6.6 °C), indicating a diminished detection-to-pain dynamic range in SCD (Figure 1A). During thermal stimulation, controls exhibited focal and condition-specific cortical engagement, with pain tolerance trials eliciting spatially distinct responses relative to detection thresholds across frequency bands; this separation was most prominent in slower frequencies, where pain-related activity was concentrated in sensorimotor scalp regions and showed coherent vote-consensus contours, indicating reproducible pain–detection differences across participants and analytic methods. In contrast, as shown in Figure 1B, 1C, separation between detection- and pain-related brain states declined progressively with increasing CS. Low-CS SCD patients showed reduced and less spatially focused differentiation, suggesting partial preservation of state-specific processing. In the high-CS group, differentiation was minimal across bands, with sparse or absent vote-consensus contours; the proportion of scalp showing a reproducible pain–detection difference was markedly reduced, particularly in delta, theta, and alpha bands.
Conclusions
Reduced detection-to-pain dynamic range in SCD identified with QST is mirrored by a progressive loss of neural separation between innocuous and painful input as CS increases. This pattern suggests increasing overlap in cortical processing of sensory and painful stimulation with greater CS. These findings support EEG-based markers of CS as objective tools to stratify pain phenotypes and potentially monitor disease burden and treatment response in SCD.