Abstract: 【Abstract】 Objective To investigate the characteristics of skin pain and the changes in mechanical pain thresholds in patients with atopic dermatitis （AD）, to analyze the relationship of pain with disease severity and pruritus, and to identify the protein profiles associated with elevated mechanical pain thresholds. Methods Between March 1, 2023, and March 31, 2026, 82 and 20 patients with AD were prospectively enrolled from outpatient departments of the Dermatology Hospital, Southern Medical University, and the First Affiliated Hospital, Sun Yat-sen University, respectively; additionally, 20 healthy controls were recruited from the First Affiliated Hospital, Sun Yat-sen University. Data on age, sex, and other characteristics of the subjects were recorded. A questionnaire survey was conducted to assess pain in 82 AD patients. Mechanical pain thresholds were measured with an electronic von Frey aesthesiometer at lesional and non-lesional skin sites in 47 patients with AD and at normal skin sites in 20 healthy controls. Pain burden and mechanical pain thresholds were compared among patients stratified by disease severity. Spearman rank correlation analysis was performed to assess the correlations of the Eczema Area and Severity Index （EASI） score, Peak Pruritus Numerical Rating Scale （PP-NRS） score, and pain Visual Analogue Scale （pVAS） score with the mechanical pain threshold. The optimal cutoff values for elevated mechanical pain thresholds were determined by receiver operating characteristic （ROC） curve analysis, and the protein profiles associated with elevated mechanical pain thresholds were identified by tape-strip proteomic analysis. Results Among the subjects who underwent mechanical pain threshold testing, 47 patients were in the AD group, including 27 males and 20 females, aged 24.0 （18.0, 33.0） years, and 20 subjects were in the healthy control group, including 8 males and 12 females, aged 25.0 （21.3, 32.0） years; there were no significant differences in gender or age between the two groups （both P > 0.05）. Mechanical pain thresholds were significantly elevated at both lesional （100.2 ± 39.8 g; t = 4.17, P < 0.001） and non-lesional sites （89.2 ± 36.9 g; t = 3.29, P = 0.002） in the AD group compared with the healthy control group （56.6 ± 37.6 g）; in the AD group, the mechanical pain threshold was significantly higher at lesional sites than at non-lesional sites （paired t = 2.74, P = 0.009）. Of the 82 patients with AD who completed the questionnaire, 58 （70.7%） reported skin pain, and 73 （89.0%） could clearly distinguish pain from pruritus. Pain intensity was significantly higher in the patients with moderate-to-severe AD than in those with mild AD （P = 0.035）; no significant differences in mechanical pain thresholds were observed between the two groups at either lesional or non-lesional sites （both P > 0.05）; however, in the moderate-to-severe AD group, the mechanical pain threshold was significantly higher at lesional sites than at non-lesional sites （paired t = 2.25，P = 0.031）. Correlation analysis showed that the EASI score was weakly positively correlated with the PP-NRS score （rs = 0.42, P = 0.030）, but was not significantly correlated with the pVAS score or the mechanical pain threshold （both P > 0.05）; there was no significant correlation between PP-NRS and pVAS scores （P > 0.05）; the mechanical pain thresholds at lesional and non-lesional sites were positively correlated （rs = 0.54, P = 0.004）. ROC curve analysis identified optimal cutoff values of 70.54 g and 73.62 g for mechanical pain thresholds at lesional and non-lesional sites, respectively, for distinguishing patients with AD from healthy controls, with corresponding areas under the curve of 0.830 （95% CI: 0.692 - 0.968） and 0.802 （95% CI: 0.650 - 0.955）. Based on these cutoff values and the availability of skin tape-strip samples, 15 samples from the healthy control group, 17 lesional samples and 16 non-lesional samples from the AD group were finally included in the tape-strip proteomic analysis. The results showed that inflammation-related proteins such as interleukin （IL）-18, IL-36A, and S100 calcium-binding protein A9 were upregulated in the epidermis of AD patients with elevated mechanical pain thresholds, whereas synapse-related proteins such as synaptogyrin 1 and synaptosome-associated protein 29 were downregulated in skin lesions of AD patients with elevated mechanical pain thresholds; the differentially expressed proteins were mainly enriched in neuromodulatory and inflammatory immune pathways. Conclusions Skin pain was common in patients with AD. Mechanical pain thresholds were elevated at both lesional and non-lesional sites in patients with AD, and mechanical pain thresholds and pain scores showed no significant correlation with pruritus scores or disease severity scores. AD patients with elevated mechanical pain thresholds exhibited a distinct expression profile of neural and inflammatory immune-related proteins.

Key words: Dermatitis, atopic, Pain, Mechanical pain threshold, Pruritus, Proteomics