Visual Hypersensitivity as a Transdiagnostic Marker of Surgical Pain Response in Arthritis and Chronic Pain Syndromes.

In total, 243 participants were included in the analysis (HC: 33, OA: 35, CPP: 63, RA: 30, PsA: 50, FM: 32). Demographic characteristics and patient-reported outcomes are shown by cohort in Table 1 . The overall sample had a mean age of 46.9 years and was predominantly female (74.9%) and non-Hispanic white (81.5%). The mean clinical pain intensity across all clinical pain groups was approximately 7.8 out of 10. An ANCOVA controlling for age and sex revealed significant differences in FM Score between cohorts (p < 0.0001). Participants with FM had the highest FM Score (mean ± standard deviation, 19.1 ± 6.25), HCs the lowest (1.36 ± 1.83), and the remaining cohorts were intermediate between the FM and HCs ( Fig. 2A and Supplementary Table 1 ). One participant with RA and three participants with PsA were identified as having outlier brightness intensity ratings and were removed via listwise whole-subject deletion. Each participants’ ratings were averaged across three pseudo-random stimulus presentations at six illumination levels. The mean ± standard deviation visual brightness rating averaged across all six illumination levels was 32.7 ± 17.8 in the HC group, 38.9 ± 15.2 in the OA group, 39.7 ± 14.7 in the CPP group, 37.1 ± 14.5 in the RA group, 45.8 ± 14.7 in the PsA group, and 46.2 ± 12.1 in the FM group. All cohorts exhibited a stimulus-dependent increase in perceived brightness as illumination intensity increased ( Fig. 2B ). Linear mixed models showed significant main effects for illuminance (p < 0.0001) and cohort (p = 0.000211) and a significant illuminance X cohort interaction (p < 0.0001). Full model details are included in Supplementary Tables 2 – 3 . Significant pairwise differences in visual brightness ratings were identified between PsA and HC (p < 0.0001), FM and HC (p = 0.00548), and PsA and RA (p = 0.0237) ( Fig. 2B and Supplementary Table 4 ). Mean brightness ratings were tested for significant bivariate correlations with patient reported outcomes ( Supplementary Table 5 ). Pearson’s r correlations revealed weak yet significant positive associations between mean brightness ratings and pain intensity (r = 0.130, p = 0.0465), and PROMIS measures for fatigue cognitive impact (r = 0.129, p = 0.0481), and depression (r = 0.151, p = 0.0208). Brightness rating was not significantly associated with FM Score (r = 0.0989, p = 0.137). Participants also rated the sensory unpleasantness (affective response) evoked by visual stimulation. Results for visual unpleasantness ratings were largely the same as those for brightness. These analyses are available in Supplementary Results ( Supplementary Figure 3 and Supplementary Table 10 ). A total of 68 participants (25 with OA and 43 with CPP) underwent surgery and completed a follow-up visit six months after surgery with a clinical pain intensity assessment. Four participants were excluded from treatment response analyses, as they provided pain intensity ratings below 3 (1 with PsA and 3 with CPP), and are not included in the prior total. Of the 68 participants, 43 (63.2%) were classified as responders to treatment and 25 (36.8%) were classified as non-responders ( Supplementary Table 6 ). Cohort specific paired plots showing baseline and follow-up pain intensity values are provided in Supplementary Fig. 1 . Baseline clinical and demographic metrics did not differ between the responder and non-responder groups ( Supplementary Table 6 ). The mean ± standard deviation visual brightness rating across all six illumination levels was 36.0 ± 13.7 in the responder group and 44.4 ± 10.8 in the non-responder group, which were significantly different (t(66) = 2.64, p = 0.0105). Visual brightness ratings across six illumination levels for responders versus non-responders are shown in Fig. 3A . A series of logistic regression models were used to analyze the relationship between mean visual brightness, FM Score, baseline pain intensity, age, sex, and treatment response status. It was found that, holding all other predictor variables constant, the odds of responding to treatment decreased by 12.7% for a one-unit increase in baseline FM Score (OR = 0.873, p = 0.0423). Additionally, holding all other predictor variables constant, the odds of responding to treatment decreased by 6.3% for a one-unit increase in mean baseline visual brightness (OR = 0.937, p = 0.00916). No significant effects were found for baseline pain, age, or sex. Full statistics are reported in Table 2 . Hierarchical logistic regression models revealed that a model including average visual brightness, FM score, baseline pain intensity, age, and sex fit significantly better than a model only including FM score, pain intensity, age, and sex (p = 0.00379; additional results shown in Supplementary Table 7 ). Additional visualizations and analyses were performed for OA and CPP cohorts individually ( Supplementary Figure 2 and Supplementary Table 8 ). A marginal effect for visual brightness was found for the OA cohort (p = 0.0657) and minimal effect for the CPP cohort (p = 0.105). A marginal effect for FM Score was found for the CPP cohort (p = 0.0640). All treatment response analyses were performed in parallel for visual unpleasantness ratings. Unpleasantness was also a significant predictor of treatment responsiveness. No substantial differences were present between the unpleasantness and brightness results. Visual unpleasantness analysis results are available in Supplementary Results ( Supplementary Figure 4 and Supplementary Table 11 ). A subset of 52 participants from the OA (n = 12) and CPP (n = 40) cohorts also completed pressure-based QST testing on the thumbnail bed at baseline. Of those, 34 (65.4%) were responders and 18 (34.6%) were non-responders to surgery. Responders and non-responders exhibited a stimulus-dependent increase in evoked pain intensity as pressure increased ( Fig. 3B ). The mean ± standard deviation pain intensity rating across all six pressure levels was 24.5 ± 11.4 in the responder group and 25.0 ± 9.78 in the non-responder group, which was not significantly different. Logistic regression was used to analyze the relationship between mean evoked pain intensity at the thumbnail, FM score, baseline clinical pain intensity, age, sex, and treatment response status. No significant effects of evoked pain ratings, baseline clinical pain intensity, age, or sex were found. FM Score was no longer a significant predictor of treatment response in this subset of individuals. Full model statistics are reported in Supplementary Table 9 . Additional visualizations and analyses were performed for the OA and CPP cohorts individually. No significant effects were found among the individual cohorts.

Individuals across etiologically diverse chronic pain conditions exhibited varying degrees of visual sensitivity. Heightened visual sensitivity suggests nociplastic processes may contribute to their pain, reducing the likelihood that treatments targeting peripheral issues, like surgery, will be effective. In longitudinal observational studies across two surgical cohorts, we demonstrated that FM score and, novelly, visual sensitivity can serve as transdiagnostic predictors of poor analgesic response. FM score and visual QST together predicted more variance in responder status than either measure alone. Although we observed positive correlations between visual sensitivity and clinical features of nociplastic pain, these were weak, suggesting that visual QST captures aspects of nociplastic pain and its underlying mechanisms that are not fully represented in current clinical self-report measures. These findings underscore the potential value of visual sensitivity assessments across diagnostic categories. Many chronic pain conditions exhibit nociplastic features, which can be indexed by the FM Survey Criteria 2 , 17 , 23 – 24 . In our study, individuals with FM had the highest average FM Score, whereas HCs had the lowest. The OA, CPP, RA, and PsA cohorts reported intermediate scores in an ascending manner, consistent with prior evidence for nociplastic pain in OA, RA, and PsA 36 , 37 . Among these, the OA cohort had the lowest FM Scores, possibly reflecting a dominant peripheral nociceptive element driven by joint degeneration. However, hip OA research indicates both nociceptive and distributed nociplastic components, the latter varying widely between individuals 36 . fMRI studies suggest that nociplastic pain may be intrinsic to systemic rheumatic diseases like RA and PsA, where inflammation drives CNS changes 38 . These diseases may show higher degrees of widespread pain and nociplastic symptoms compared to OA. The CPP cohort reported intermediate FM Scores, consistent with literature on nociplastic pain in CPP 9 . Importantly, high intra-cohort FM Score variation across pain conditions suggests mechanistic heterogeneity within diagnoses. Healthy controls reported the lowest visual sensitivity, while the RA, OA, CPP, PsA, and FM cohorts showed ascending sensitivity. The difference between HC and FM replicates prior results 17 and aligns with literature on multimodal hypersensitivity in FM 13 , 17 , 39 , 22 . We also found significantly increased visual sensitivity in the PsA cohort compared to HCs. While evidence for pressure QST hypersensitivity in PsA is growing 40 , visual hypersensitivity data are limited. Although uveitis, a chronic inflammatory eye disease, often occurs in PsA 41 , it likely did not drive our findings, as only one participant had prior uveitis. The neural mechanisms underlying visual hypersensitivity in chronic pain remain unclear. Aberrant anterior insula activity 17 , 39 and rostral ventromedial medulla involvement 14 have been linked to FM-related visual sensitivity. In females at risk for CPP, visual unpleasantness has been tied to increased primary visual cortex activity moderated by bladder pain 18 . Visual sensitivity may also arise from a process akin to central sensitization from repeated somatosensory input 42 . In the experimental paradigm used here, visual stimuli were presented 24 times per session, and bright lights are ubiquitous in both natural and built environments. Supporting this, sensitization to repeated visual stimuli was found in ganglionic cells of healthy individuals 43 , suggesting CNS sensitization may underlie ongoing visual hypersensitivity in chronic pain. Increased visual sensitivity, as measured by mean brightness ratings, was weakly linked to greater clinical pain, depression, and fatigue-related cognitive impact. These associations support our hypothesis that pain features relate to visual hypersensitivity, but the weak correlations (r’s ≤ .151) and lack of association with FM Score suggest visual sensitivity and self-reports capture distinct aspects of the pain experience. Consistent with previous work, we found that FM Score, used continuously, predicted non-response to peripherally-directed treatments 6 , 8 . Importantly, baseline visual sensitivity added predictive value beyond FM Score. No other baseline clinical features distinguished responders from non-responders. Treatment-resistant pain is often maintained by dysregulated “top-down” CNS processes, independent of peripheral input 2 , 44 . This contrasts with “bottom-up” nociplastic pain, driven by peripheral pathology and resolved by targeting the source 45 . Visual stimuli bypass peripheral nociceptors and spinal mechanisms, engaging supraspinal pathways directly. Thus, visual hypersensitivity may reflect a transdiagnostic, top-down CNS marker of nociplastic pain. This idea is consistent with the framework proposed by Bar-Shalita et al. (2019) 46 , who described shared central mechanisms underlying heightened responses to both non-painful and painful sensory input in sensory modulation disorder (SMD). In particular, the sensory over-responsivity (SOR) subtype of SMD manifests clinically as a condition in which non-painful stimuli are perceived as abnormally unpleasant, driven by cortical hyper-excitation, excitatory-inhibitory imbalance, and broader sensory modulation abnormalities. Notably, SOR has been identified as a risk factor and comorbidity of chronic non-neuropathic pain disorders 46 . Similarly, our findings suggest that visual hypersensitivity may reflect overlapping central processes of sensory amplification, providing further evidence that altered multisensory modulation contributes to the broader nociplastic pain phenotype. Pressure-based stimuli, by contrast, involve peripheral and spinal pathways before reaching supraspinal centers and may better reflect bottom-up nociplastic pain. In our data, pressure-based QST at the thumbnail did not predict surgical outcomes, possibly indicating: (1) many participants exhibited top-down phenotypes, and/or (2) limited predictive utility of peripheral QST for surgical response. Broader literature on this topic is mixed. A meta-analysis found QST correlates with central sensitization measures, strongest for pressure pain thresholds 47 . By extension, one study found that preoperative pinprick hyperalgesia predicted chronic postoperative pain up to 12 months following total knee replacement in patients with knee osteoarthritis 48 . However, another review of preoperative experimental pain assessments and postoperative outcomes concluded that many studies showed moderate to high risk of bias and that most QST variables showed no consistent correlations with pain after surgery 49 . This study has some limitations. Limitations include unbalanced responder/non-responder sizes within treatment cohorts, necessitating group combination and limiting cohort-specific analyses. The combined sample had a roughly equal responder distribution, enabling transdiagnostic comparisons. Second, there was a sex imbalance, notably in CPP (all female). We controlled for sex in analyses and stratified CPP comparisons; no sex differences in FM Score or visual sensitivity emerged. Third, fewer participants completed pressure QST, limiting pressure vs. visual sensitivity comparisons. Additionally, only one pressure modality was tested; others may yield different predictive insights. This study identified visual hypersensitivity across chronic pain conditions and linked it to treatment response likelihood. A major strength was the multisite design (University of Michigan and University of Glasgow) and parallel QST protocols. While visual sensitivity appears promising as a marker of nociplastic pain and treatment resistance, future work should integrate fMRI or other imaging to examine underlying functional components. It is unlikely that visual QST is uniquely affected. Future studies could assess other centrally-directed stimuli (e.g., auditory, olfactory) to probe multisensory sensitivity. FM studies report auditory hypersensitivity 13 , 50 , making acoustic sensitivity in rheumatic pain conditions a compelling area for further investigation. Our findings suggest that visual hypersensitivity extends beyond FM, providing a potential transdiagnostic marker of the presence of underlying nociplastic pain mechanisms. These insights may inform precision medicine by refining pain phenotypes and improving treatment stratification in chronic pain care.

Chronic pain is exceedingly difficult to treat, in part because pain mechanisms are heterogeneous. Chronic pain can develop or be maintained by both ongoing peripheral nociception and/or by sensitization of central nervous system (CNS) pain processing. The latter, central sensitization, is a critical underlying mechanism of a pain type referred to as nociplastic pain 1 – 3 . Nociplastic pain is the predominant pain mechanism in fibromyalgia (FM) and other chronic overlapping pain conditions 2 , 4 which are characterized by widespread pain, fatigue, sleep disturbances, cognitive dysfunction, mood disturbances, and sensory hypersensitivity 2 . Nociplastic pain involves a complex interplay between peripheral “bottom-up” and central “top-down” mechanisms, which may contribute to the challenges in treatment 1 , 2 . Many individuals with identifiable peripheral pathology and nociception, such as joint damage in osteoarthritis, undergo surgery or other peripherally-directed medical treatments to relieve their pain. However, some patients continue to experience pain following these procedures 1 . While some cases of treatment resistance can be explained by local responses to surgery (such as with the influence of synovium following total knee replacements) 5 , it is increasingly recognized that continued pain perception may be preferentially maintained by top-down nociplastic mechanisms 6 – 8 . Identifying and assessing the severity of nociplastic pain remains challenging. In previous work, we demonstrated that the Fibromyalgia Survey Criteria can serve as an indirect measure of nociplastic pain 9 , 10 , 1 . The Fibromyalgia Survey Criteria is a self-report measure that evaluates the extent of widespread pain and the presence of symptoms characteristic of nociplastic mechanisms. Individuals with painful knee osteoarthritis and chronic pelvic pain (CPP) who scored high on this measure, suggesting involvement of nociplastic mechanisms, were less likely to experience pain relief following knee arthroplasty and hysterectomy, respectively 6 – 8 . Another approach to assessing nociplastic pain is through quantitative sensory testing (QST), which examines alterations in sensory sensitivity. Nociplastic pain is marked by diffuse hypersensitivity to both painful and non-painful sensory stimuli, including heightened sensitivity to common environmental triggers like sounds, lights, and odors 11 – 14 . Unlike traditional QST methods that stimulate peripheral tissues (e.g., using mechanical or thermal stimuli), visual stimuli bypass peripheral mechanisms to directly engage the CNS 15 , 16 , and may provide unique insights into nociplastic mechanisms. We demonstrated that individuals with FM rated an experimental flashing checkerboard stimulus as significantly more unpleasant compared to healthy controls 17 . Functional imaging revealed that this stimulus caused increased activation in the insula, an effect that was reduced with pregabalin. The flashing checkerboard stimulus has been further applied across several studies demonstrating its validity and utility for probing central mechanisms of nociplastic pain 18 – 20 . Other studies have similarly found heightened visual sensitivity and discomfort from light (i.e., photophobia) in fibromyalgia patients compared to individuals without pain 14 , 21 , 22 , suggesting visual hypersensitivity may be a psychophysical marker of nociplastic pain. In support of this notion, visual hypersensitivity has also been observed in other conditions with CNS involvement, including chronic migraine, complex regional pain syndrome, traumatic brain injury, and among women at risk for CPP 18 , 21 , 22 , 19 . Despite these findings, the relationship between visual sensitivity, nociplastic pain, and treatment response is still largely unexplored. In the present study, we first characterized sensitivity to experimental visual stimulation among individuals with diverse types of chronic pain conditions, traditionally considered to have significant peripheral nociceptive input: osteoarthritis (OA) of the hip, CPP with and without endometriosis, rheumatoid arthritis (RA), and psoriatic arthritis (PsA). For comparison, we also included individuals with FM, representing the extreme nociplastic pain phenotype, and healthy controls (HC). We then investigated whether visual sensitivity assessed at baseline could predict pain response to surgery six months later in the OA and CPP cohorts. We compared the ability of visual QST to predict surgical outcomes against the Fibromyalgia Survey Criteria, as per our previous research, and against pressure-based QST. We hypothesized that individuals with OA, CPP, RA, and PsA would exhibit an intermediate nociplastic phenotype, showing greater visual sensitivity than controls but less than those with FM. Additionally, we hypothesized that individuals with OA and CPP who displayed increased sensitivity to visual stimulation would be less likely to experience pain relief following surgery. Figure 1 depicts a summary of the study design.

Participants from three harmonized studies were used for analysis: 1) Mechanisms of the Centralized Pain Phenotype (MiCAPP - University of Michigan) included healthy controls and individuals with OA, RA, and FM; 2 ) Peripheral and Central Nervous System Mechanisms of Persistent Post-Hysterectomy Pain (MiHyst - University of Michigan) included females with CPP; and 3) Characterizing the Centralized Pain Phenotype in Chronic Rheumatic Disease - A Stride Towards Personalized Analgesia (CENTAUR - University of Glasgow) included individuals with PsA. MiCAPP (HUM00120181) and MiHYST (HUM00117473) protocols received ethical approval from the Medical Institution Review Board at the University of Michigan (Ann Arbor, MI, USA). CENTAUR (IRAS ID: 257990) was approved by the National Health Service Health Research Authority at the University of Glasgow (Glasgow, UK). All participants read and signed an informed consent form prior to participation. Inclusion criteria for all participants were as follow: (1) ability to read and speak English to allow for written or verbal informed consent, phenotyping, and patient reported outcomes measures, (2) patients < 76 years old (to mitigate against vascular disease leading to functional imaging abnormalities in elderly patients), (3) right hand dominant, (4) normal visual acuity or correctable to at least 20/40 for reading instructions in the visual sensitivity testing, (5) willingness to refrain from pain medications such as NSAIDs, acetaminophen, corticosteroids, and opioids on the day of QST, (6) willingness to refrain from alcohol and nicotine on day of QST, (7) willingness to refrain from physical activity or exercise that would cause muscle and/or joint soreness for 48 hours prior to testing, (8) preferred no chronic daily use of adjunctive pain medications, including tricyclic antidepressants, serotonin-norepinephrine reuptake inhibitors, and gabapentinoids (alternatively, individuals were requested to be weaned off of these medications for at least two weeks prior to being studied, or if necessary, asked to remain on a stable dose for at least two weeks prior to QST assessments), and (9) completion of the FM Survey Criteria (regardless of cohort). Additional inclusion criteria specific to cohort were as follow: HC: (1) no major diseases or diagnoses influencing pain or function; OA: (1) Kellgren/Lawrence III or IV osteoarthritis confirmed by routine radiographs, (2) pain longer than a year, (3) scheduled for primary hip arthroplasty surgery; RA: (1) diagnosed with adult onset RA as defined by the American College of Rheumatology/European League Against Rheumatism (ACR/EULAR) 2010 Criteria for the Classification of RA 52 , (2) pain longer than a year, (3) scheduled to be treated with new immunosuppressant treatment as part of standard clinical care; FM: (1) meet the 2011 FM Survey Criteria for the diagnosis of FM; CPP: (1) pelvic pain that localizes to the anatomic pelvis, and is not limited to only dysmenorrhea, dyspareunia, or dyschezia, (2) pelvic pain lasting longer than six months; (3) female scheduled for hysterectomy for benign (non-cancer) indication; and PsA: (1) Adult patients (≥18 years) attending rheumatology clinics with (2) PsA as defined by CASPAR criteria, (3) scheduled to be treated with a new immunosuppressant treatment as part of standard clinical care, (4) evidence of ongoing inflammation (synovitis on clinical examination). Exclusions criteria for all participants were as follow: (1) inability to provide written or verbal informed consent, (2) severe physical impairment (e.g., blindness, deafness, paraplegia), (3) comorbid medical conditions that may significantly impair physical functional status (e.g., history of non-skin malignancy, or autoimmune disorder [except the RA and PsA cohorts]), (4) illicit drug or unreported opioid use, (5) medical or psychiatric conditions that in the judgment of study personnel would preclude participation in this study (e.g., malignancy, psychosis, suicidal ideation; note that stable anxiety and depression were not exclusions), (6) pregnant, (7) liver failure or self-reported liver cirrhosis, (8) self-reported hepatitis, (9) uncontrolled diabetes, (9) severe cardiovascular disease that is self-reported by patient or by medical record, or (10) diagnosed peripheral neuropathy. Additional exclusion criteria specific to cohort were as follow: HC: (1) no existing chronic pain condition; OA: (1) other non-osteoarthritis diagnoses for total hip arthroplasty (e.g., dysplasia); RA: active extra-articular involvement from RA, (2) any type of viral, bacterial, fungal or parasitic infection, (3) a diagnosis of any significant systemic inflammatory condition other than RA or secondary Sjogren’s Syndrome, (4) severe Raynaud’s phenomenon for which the patient receives treatment; FM: (1) no underlying autoimmune, inflammatory, infectious or metabolic disorder that might contribute to the development of FM; CPP: (1) pregnant and/or lactating; and PsA: no additional exclusion criteria. To avoid possible adverse events in vulnerable patients, participants with a history of visual-evoked seizures or migraines were excluded from all cohorts. Participants completed the 2011 Fibromyalgia Survey Criteria 23 , 24 at baseline. This scale consists of two subscales: a Widespread Pain Index measuring pain or tenderness over the previous week in various body regions (scored 0–19), and a Symptom Severity Scale measuring severity of associated symptoms including fatigue, subjective cognitive problems, and headaches (scored 0–12). Subscales are summed to give a total Fibromyalgia Score (0–31). Participants completed the painDETECT questionnaire 25 , from which a specific item on worst pain intensity (sub-item 2) was extracted for analysis. Participants rated “How strong was the STRONGEST pain during the last 4 weeks?” on a scale 0–10 at baseline and at a 6-month follow-up visit. Multiple PROMIS (Patient-Reported Outcomes Measurement Information System) measures 26 were completed at baseline, including PROMIS Pain Interference (Short Form 8a), PROMIS Sleep-Related Impairment (Short Form 8a), PROMIS Fatigue FM Profile (Short Form 16; measuring a) fatigue experience and impact of fatigue on b) social, c) cognitive, and d) motivation domains 27 ), PROMIS Depression (Short Form 8a), and PROMIS Anxiety (Short Form 8a). Standardized PROMIS T-scores were used for analysis. The Michigan Visual Aversion Stress Test (M-VAST) was used to assess visual sensitivity as previously described 17 – 20 . Testing was conducted via PsychoPy presentation software (version 2.7.7) 28 in light and sound attenuated exam rooms. All QST procedures were standardized across cohorts to ensure methodological consistency. The MiCAPP and MiHyst studies were conducted in the same laboratory at the University of Michigan, while the CENTAUR study at the University of Glasgow followed identical protocols after on-site training by the Michigan research team. This harmonized approach ensured consistency in stimulus delivery, equipment calibration, and data collection across sites. Participants sat with their eyes perpendicularly aligned 20 inches away from the center of a 15-inch (diagonal) calibrated LED monitor (EIZO RadiForce MX215), acclimated to the dark for 5-minutes prior to test start. The task began with a flashing (8 Hz) annulus blue and yellow checkerboard (duration: 10 seconds). Each task-block was alternated with a black screen off-block (duration: 10 seconds). The visual stimulus was first presented at six levels of illumination intensity (2.1, 6.6, 23.8, 63.2, 122.6, and 203.9 lux, measured with a CEM DT-1309 light chronometer) in ascending order for participant familiarization and to assess task tolerance. Each illumination intensity level was then presented in triplicate in pseudo-random order following the Method of Constant Stimuli. Participants were asked to rate each stimulus on digital numerical rating scales of evoked sensory affect (0 = not unpleasant; 100 = most unpleasant imaginable) and sensory intensity (0 = not bright; 100 = brightest image imaginable). Mean ratings of sensory unpleasantness and intensity at each stimulus intensity level from the pseudo-random series were used for analysis. A subset of participants with OA and CPP also completed baseline pressure pain testing and had a 6-month follow-up visit. Pressure pain sensitivity was assessed at the dominant thumbnail bed using the Multimodal Automated Sensory Testing (MAST) system (Arbor Medical Innovations, Saline, MI) 17 , 29 , 30 . The thumbnail was chosen as an asymptomatic site, remote from the location of surgery and primary area of pain compliant. Hyperalgesia in asymptomatic body areas is suggestive of nociplastic pain 31 . Before data collection, participants underwent a familiarization procedure on their nondominant thumbnail bed to reduce test anxiety. First, the MAST system automated the delivery of a series of discrete 5-s pressures in ascending order using an electro-mechanical 1-cm 2 rubber tipped probe, starting at a pressure intensity of 0.50 kgf/cm 2 and increasing in 0.50 kgf/cm 2 steps for subsequent stimuli (with a minimum 20-s interstimulus interval) until reaching each participant’s individual pain tolerance or 10.00 kgf/cm 2 . Patients rated the pain intensity evoked at each stimulus level on a 0 to 100 numerical rating scale, with endpoints “no pain” and “worst pain imaginable.” Participants then received a series of pressures delivered twice in pseudo-random order (Method of Constant Stimuli) at each of six intensity levels: 1, 1.5, 2, 2.5, 3.5, and 4.5 kgf/cm 2 . Mean ratings of pain intensity across all stimulus intensity levels from the pseudo-random series were used for analysis. Participants with OA underwent hip arthroplasty, while those with CPP underwent hysterectomy, both aimed at reducing pain as part of their standard clinical care ( Fig. 1 ). Participants were labeled as “responders” to surgical treatment if they experienced a reduction in pain intensity of 50% or more from their baseline levels at six months post-surgery. Those who did not meet this threshold were classified as “non-responders.” Participants whose baseline pain intensity was below 3 were excluded from follow-up analyses. The OA and CPP surgical cohorts were combined to increase sample size for the responder analysis. Statistical analyses were performed in R (version 1.4.1717) 32 and GraphPad Prism (version 10.0.2) 33 with alpha set at 0.05. FM Scores and visual unpleasantness/brightness ratings in response to the visual stimulus were screened for outliers via the generalized ESD test (extreme Studentized deviate; also known as Grubbs’ Test) 34 , which is used to detect one or more outliers in a univariate data set that follows an approximately normal distribution. Differences in FM Scores were assessed via one-way analysis of covariance (ANCOVA) controlling for the influence of age and sex. Differences between other patient-reported outcomes and age were assessed via one-way analysis of variance (ANOVA). ANCOVA and ANOVA pairwise comparisons were adjusted via Tukey’s HSD. Differences in sex composition between cohorts was assessed via chi-squared analysis. Pairwise comparisons performed between cohorts in sex composition were corrected for multiple comparisons via Bonferonni’s correction (alpha = 0.0071). Within and between-group differences in baseline visual brightness and unpleasantness ratings were assessed using separate linear mixed models (LMM) using the “lme4” package in R (version 1.4.1717) 32 . These models included cohort (categorical factor with 6 levels: HC, OA, CPP, RA, PsA, FM), illuminance level (continuous), sex (categorical factor with 2 levels), and age (continuous) as fixed effects and controlled for within-subject response variation by including a random effect for participant. Null models were used to assess statistically significant differences between cohorts, where each cohort was compared to each other cohort via stepwise model comparison. Group, illuminance, and interaction effects were analyzed using R’s “glht” package with Tukey’s method for adjusting p-values. Associations between average visual stimulus ratings and patient reported outcomes were examined via Pearson’s correlations across all cohorts combined. Baseline clinical and demographic characteristics of the “responder” and “non-responder” groups were compared via unpaired t-tests. Logistic regressions were used to examine associations between baseline FM Score, mean visual and pressure QST outcomes, and surgical response status in a combined sample of OA and CPP participants with complete baseline and 6-month outcome data.