OBJECTIVE:To assess the relationship between pincer morphology and incident radiographic hip osteoarthritis (RHOA) and study-specific subgroups. METHODS:Hips completely free of RHOA at baseline and with follow-up within 4-8 years were drawn from the World COACH consortium. The lateral centre edge angle (LCEA) was calculated uniformly on all baseline radiographs. Moderate pincer morphology was defined as an LCEA ≥40°, and severe pincer morphology as an LCEA ≥45° in sensitivity analyses. The primary outcome was incident RHOA defined by a harmonised OA score. A logistic regression model with generalised mixed effects with three levels (within-cohort, within-person and within-hip side correlation) adjusted for age, biological sex and body mass index (BMI) was employed. Descriptive statistics are reported for age, biological sex and BMI. RESULTS:18 935 hips from nine cohorts were included. 4894 hips (25.8%) had moderate pincer morphology. Within 8 years (mean 6.0±1.7 years), 352 hips (1.9%) developed RHOA. Moderate pincer morphology was not associated with RHOA (OR 1.15 (0.92-1.51)), whereas severe pincer morphology was significantly associated (OR 1.50 95% CI 1.05 to 2.15). Moderate pincer morphology in groups aged 40-50 (RR 2.67, 95% CI 1.43 to 4.95) and BMI ≥25 (RR 1.23 95% CI 0.98 to 1.71) had a higher risk compared with non-pincer hips. Women (RR 1.20 95% CI 0.93 to 1.56) with pincer morphology may be more at risk than men (RR 0.95 95% CI 0.57 to 1.58). CONCLUSION:The odds of developing RHOA within 8 years for hips with severe pincer morphology are 1.5 times higher than pincer-free hips, whereas moderate pincer morphology was not significantly associated with RHOA. Further research is necessary to uncover high risk subgroups of pincer morphology.

This article summarizes key advancements of artificial intelligence (AI) for rheumatic and musculoskeletal disease imaging in the diagnosis and classification, and predictive modeling of rheumatoid arthritis, psoriatic arthritis, spondyloarthritis, and osteoarthritis since 2020. AI applications are emerging in disease diagnosis, severity classification, and prediction of incidence and progression, with ongoing challenges related to external validation, mitigation of bias, data privacy, transparency, and clinical integration. In the near term, AI could assist clinicians with diagnostic interpretation and disease monitoring. Future applications include improved prognostic modeling and identifying candidates for targeted interventions and clinical trials.

OBJECTIVE:To assess the relationship between cam morphology and the development of radiographic hip osteoarthritis (RHOA), overall and in subgroups based on age, biological sex and body mass index (BMI). METHODS:Hips with no RHOA at baseline and with available follow-up during 4-8 years were selected from the Worldwide Collaboration on Osteoarthritis PrediCtion for the Hip (World COACH) consortium. Alpha angles were uniformly measured on anteroposterior radiographs, with a threshold of 60° used to define cam morphology. Incident RHOA was defined as the transition from an RHOA-free state at baseline to definite diagnosis of RHOA at follow-up. The association between baseline cam morphology and the development of RHOA was assessed using a three-level mixed-effects logistic regression model, accounting for hip side, individual and cohort-level variation. RESULTS:A total of 23 886 hips were included (mean age: 62.2±8.4 years; 70.6% female; BMI: 27.4±4.5; mean time to follow-up: 6.1±3.0 years). Cam morphology was associated with RHOA (OR: 1.87, 95% CI 1.36 to 2.59), as was a greater alpha angle (OR 1.02, 95% CI 1.01 to 1.03 for every degree increase). The overall relative risk of developing RHOA in hips with cam morphology was 1.62 (95%CI 1.26 to 2.07), greatest for those aged 51-60 years (2.15, 95% CI 1.55 to 2.98) and higher in males (2.50, 95% CI 1.67 to 3.73), compared with females (1.75,95% CI 1.24 to 2.48). CONCLUSION:Hips with cam morphology have higher odds of developing RHOA within 4-8 years compared with hips without cam morphology. The relative risk was highest in subgroups of participants aged 51-60 years and in males, making cam morphology a potential target for primary or secondary prevention of RHOA.

BACKGROUND:Psychological distress is associated with suboptimal outcomes after total joint arthroplasty (TJA). This study aimed to develop and evaluate machine learning models to predict a high psychological distress phenotype using only preoperative data. METHODS:We conducted a retrospective secondary analysis of patients undergoing primary hip or knee arthroplasty at Duke University between 2018 and 2024. Phenotypes were derived using latent class analysis of the Optimal Screening for Prediction of Referral and Outcome-Yellow Flag tool. There were four models such as (1) elastic net, (2) XGBoost, (3) random forest, and (4) logistic regression trained to predict a high-distress phenotype using preoperative demographic, clinical, and patient-reported data. The dataset was split into training and testing sets (70:30). Model performance was evaluated using the area under the receiver operating characteristic curve, accuracy, Brier score, sensitivity, specificity, calibration, and decision curve analysis. RESULTS:A total of 494 patients (64% women) undergoing TJA (knee 57%, hip 43%) were included; 18% (n = 89) were classified as having high postoperative distress. Patients in the high-distress phenotype demonstrated lower patient-reported outcomes measurement information system-physical function, higher patient-reported outcomes measurement information system-pain interference, pain ratings, and a greater prevalence of high-impact chronic pain. The elastic net model performed best, with an area under the receiver operating characteristic curve of 0.75 (95% CI: 0.62 to 0.85), compared with logistic regression (0.73), random forest (0.71), and XGBoost (0.65). Key preoperative predictors of high distress included a higher preoperative Optimal Screening for Prediction of Referral and Outcome-Yellow Flag count, greater pain, preoperative depression, and higher body mass index. CONCLUSIONS:Commonly collected data from routine preoperative clinical care show promise for predicting psychological distress after TJA. The elastic net model demonstrated the strongest overall performance and interpretability. Such models could support early identification of patients at risk for postoperative psychological distress, enabling targeted behavioral health referral or psychologically informed physical therapy prior to surgery. Future work should validate these models in larger, multisite cohorts.

Psychological distress is common in individuals undergoing total joint arthroplasty (TJA). Understanding psychological phenotypes and their transitions from before to after surgery can inform risk stratification and targeted care. This study aimed to characterize psychological phenotypes, examine transitions, and compare patient outcomes across phenotypes. This retrospective study included 494 patients who underwent primary hip (43%) or knee (57%) arthroplasty at Duke University Health System (2018-2024). Latent transition analysis identified and examined transitions of psychological phenotypes preoperatively and postoperatively using the Optimal Screening for Prediction of Referral and Outcome Yellow Flag tool. Demographic characteristics, phenotype transitions, Patient-Reported Outcomes Measurement Information System (PROMIS) Pain Interference (PI), PROMIS Physical Function (PF), pain intensity, and high-impact chronic pain (HICP) were compared across phenotypes. The optimal model fit was a constrained model comprising five classes: class 1 (low self-efficacy with poor pain coping), class 2 (low distress), class 3 (poor pain coping), class 4 (high distress), and class 5 (low self-efficacy with acceptance). Most patients (n = 271, 55%) transitioned to a different phenotype. The probabilities for remaining in the same class ranged from 0.19 (poor pain coping) to 0.61 (low distress). The incidence of high distress was 6% within 12 months after TJA. High distress was associated with lower PROMIS-PF and higher PROMIS-PI scores, pain intensity, and prevalence of HICP (P < 0.001). Transitions were observed across all phenotypes, with some demonstrating greater stability and others showing more state-like variability. Identifying phenotypes with distinct trajectories and outcomes may support targeted screening and preoperative risk stratification.

Objective Pain experiences are complex, multifactorial, and subjective. Pain phenotyping can provide a holistic view of pain experiences by identifying distinct pain profiles. This study aimed to characterize chronic pain phenotypes using multiple validated osteoarthritis (OA) pain measures and assess their associations with functional and psychological outcomes. Design Participants were from the Johnston County Health Study (JoCoHS), a population-based cohort of non-Hispanic White, non-Hispanic Black, and Hispanic adults aged 35–70 years. Fourteen validated pain measures were administered. Archetypal analysis, a soft clustering technique, was applied to identify pain archetypes. Associations between pain phenotypes and varying functional and psychological characteristics were examined using Dirichlet regression for modeling archetype membership proportions. Results were validated in an independent cohort. Results Among 834 participants (33% male, 27% with symptomatic knee OA, mean age 55 ± 9.6 years, mean BMI 33 ± 8 kg/m2), five pain archetypes emerged: (1) Low Pain, (2) High Somatization and Catastrophizing, (3) Widespread Pain, Aching, and Stiffness (PAS), (4) Knee Pain, and (5) High Pain and Distress. The Low Pain archetype was associated with better physical performance, higher functional status, and lower anxiety and depression, while the High Pain and Distress and Knee Pain archetypes were associated with worse outcomes. Pain-related anxiety and depression were associated with higher membership in the High Somatization archetype. Conclusions Pain phenotyping using brief self-reported measures identified distinct profiles with varying functional and psychological characteristics. These findings highlight the potential of phenotyping for improved understanding of the pain experience.

OBJECTIVE:Patients with osteoarthritis (OA) affecting multiple joints experience greater pain than those with single-joint disease, yet most research examines isolated joints, leaving a gap in multi-joint disease. This study aimed to describe radiographic hip (rHOA) and knee OA (rKOA) within UK Biobank (UKB), exploring interrelationships across joints and associations with joint pain, body size, race and deprivation. DESIGN:This cross-sectional study applied machine learning to hip and knee dual-energy X-ray absorptiometry scans. Radiographic OA (rOA) was defined as custom grades ≥2. Joint pain was assessed through self-reported questionnaires. Logistic regression models examined bilateral and cross-joint associations, as well as associations with joint pain. Adjustments were made for age, sex, race, height, weight and deprivation. RESULTS:Among 59,475 individuals (mean age 65 years; range 45-85; SD 7.7; 52.8% female), rHOA prevalence was 4098 (6.9%, right-side) and 4841 (8.1%, left-side). The corresponding estimates for rKOA were 3750 (6.3%) and 4220 (7.1%). Overall, increasing grades of rOA and number of joints affected were more strongly associated with concurrent joint pain (four-joints: OR 4.3 [95%CI 2.6-7.2]). Regarding joint-interrelationships, bilateral associations were stronger at the knee (26.1 [24.1-28.2]) than the hip (5.6 [5.2-6.1]). Cross-joint associations (hip-knee) were weaker. BMI was more strongly associated with rKOA (1.57 [1.52-1.61]) than rHOA (1.05 [1.02-1.09]). Greater height was positively associated with rHOA but appeared protective for rKOA. CONCLUSIONS:Radiographic hip and knee OA exhibit distinct patterns of interrelationship, associations with symptoms and risk factors, suggesting heterogeneity in disease processes and the need for joint-specific treatment.

OBJECTIVE:This study aims to develop hip morphology-based radiographic hip osteoarthritis (RHOA) risk prediction models and investigates the added predictive value of hip morphology measurements and the generalizability to different populations. METHODS:We combined data from nine prospective cohort studies participating in the Worldwide Collaboration on OsteoArthritis prediCtion for the Hip (World COACH) consortium. RHOA grades were harmonized, and incident RHOA was defined as hips without definite RHOA at baseline that developed definite RHOA within four to eight years. Baseline hip morphology was quantified with automatically and uniformly determined lateral center edge angle and alpha angle measurements on anteroposterior radiographs. Discriminative performance of generalized linear mixed model (GLMM) definitions with and without hip morphology measurements was determined with stratified cross-validation. With leave-one-cohort-out cross-validation, the generalizability to unseen populations of hip morphology-based GLMMs and random forest (RF) models was evaluated. RESULTS:From the included 35,984 hips without definite RHOA at baseline, 4.7% developed incident RHOA within four to eight years. The GLMM with cohort-specific intercept, considering baseline demographics, RHOA grade, and hip morphology measurements, showed a mean area under the receiver operating characteristic curve (AUC) of 0.80 (±0.01) in stratified cross-validation. Using a marginal intercept decreased performance by 0.1 in AUC. Similar results were found for a GLMM without hip morphology measurements. Leave-one-cohort-out cross-validation showed comparable discrimination (AUC between 0.56-0.88) and calibration performance for hip morphology-based GLMMs and RF models. CONCLUSION:In hips free of definite RHOA, our AUCs for the incident RHOA models showed good predictive performance in similar populations. However, the added predictive value of the morphology measurements was small, and model performance was heterogeneous in leave-one-cohort-out cross-validation.

BACKGROUND:Chronic inflammation in older adults is a key contributor to functional decline and mortality. Although anti-inflammatory medications have shown limited success in improving physical function, emerging targeted approaches offer new promise. We applied machine learning to identify clusters of older adults with shared patterns of inflammatory and cardiometabolic dysregulation and evaluated their responses to specific interventions. METHODS:We conducted a secondary analysis of the Enabling Reduction of low-grade Inflammation in Seniors (ENRGISE) multicenter, double-blind, placebo-controlled, factorial trial. This trial assessed the effects of losartan, omega-3, combination therapy, or placebo on interleukin-6 (IL-6) levels and 400-m walking speed. Key variables were selected using least absolute shrinkage and selection operator (LASSO), followed by linear regression to identify those significantly affecting the outcome slope. The optimal intervention was defined as the one that maximized the estimated slope improvement. RESULTS:We included 287 participants (47.4% female; mean age 77.6 ± 5.4 years) with a baseline IL-6 of 4.81 pg/mL. If all participants had received the recommended interventions, the estimated mean IL-6 slope would be -0.70 pg/mL/year (95% CI, -3.71, 1.41), compared to -0.51 pg/mL/year (-1.47, 0.35) among those randomized. The estimated improvement in walking speed was +0.0017 m/s/year (-0.0336, 0.0407) for the recommended interventions versus +0.0015 m/s/year (-0.0145, 0.0154) observed in the trial. For grip strength, the slope was -1.02 kg/year (-2.63, 0.57) for the recommended group and -1.02 kg/year (-1.79, -0.45) for the trial group. CONCLUSION:Although results were not significant, our findings suggest that tailored interventions based on individuals' unique profiles may yield more favorable effects compared to non-tailored approaches. However, further powered studies should continue to explore precision medicine analytics and their potential to help identify more effective and personalized interventions.

OBJECTIVE:To quantify general population prevalence, and differences by sex and BMI, in ultrasound (US) features of knee osteoarthritis (KOA), patient-reported knee symptoms (Ksx), radiographic KOA (rKOA), and symptomatic KOA (sxKOA). METHODS:Participants from the Johnston County Health Study (2019-2024; n=902) provided demographic, clinical, and imaging data. Ksx was defined as self-reported pain, aching or stiffness on most days of any one month in the past 12 months. rKOA was defined as Kellgren-Lawrence grade ≥2 (or total knee replacement (TKR)), severe rKOA as grades 3-4 (or TKR), and sxKOA as rKOA and Ksx in the same knee. Standardized US scoring was performed as previously reported. Weighted prevalence estimates and 95% confidence intervals (CI) for Ksx, rKOA, sxKOA, severe rKOA, and US features were calculated overall and by sex and BMI category. RESULTS:Among 902 participants (67% female; 66% non-Hispanic white, mean age 55 years, and BMI 33 kg/m²), the weighted prevalence of Ksx, rKOA, sxKOA, and severe rKOA were 61%, 36%, 27%, and 21%, respectively, all higher (based on non-overlapping CIs) than baseline estimates from the Johnston County Osteoarthritis Project and other cohorts. Ksx, rKOA, sxKOA, and severe rKOA were higher among females and those with obesity. Compared to females with obesity, those with severe obesity had higher prevalence of all radiographic outcomes at moderate and severe grades with non-overlapping CI. US effusion and synovitis were more prevalent in males, whereas medial and lateral cartilage damage were more common in females. With increasing obesity, osteophytes and medial cartilage damage were more prevalent, while gryescale effusion was actually less frequent. CONCLUSION:KOA prevalence continues to increase in the population. Sex and BMI differences in KOA prevalence and US features were most pronounced among females with severe obesity. Distinguishing between sexes and by obesity levels may improve understanding of KOA mechanisms and guide tailored prevention and treatment strategies.

PurposeTo investigate the longitudinal relationships between serum biomarkers of joint metabolism, knee injury, and Knee Injury and Osteoarthritis Outcome Score (KOOS) using novel methodologies.MethodsData were collected from military officers who enrolled as cadets between 2004-2009, with follow-up conducted between 2015-2017. Analyses included 234 officers who had no history of knee ligament/meniscal injury at the time of military academy matriculation, had serum biomarker measurements at matriculation and graduation, demographic data, and KOOS assessment at follow-up. Biomarkers included Collagen Type II (C2C) and Type I and II (C1,2C) collagenase-generated cleavage epitopes, C-terminal propeptide of Type II collagen (CPII), and C- and N-terminal telopeptides of type I collagen (CTX and NTX). Angle-based Joint and Individual Variation Explained (AJIVE) was used to determine demographic determinants of biomarker levels and individual modes of variation specific to biomarker levels at matriculation and graduation, stratified by sex.ResultsWe confirmed known associations of joint metabolism biomarkers with age in both sexes and with smoking in males. Matriculation biomarker data in males suggested a protective biomarker profile characterized by high cartilage synthesis and low cleavage of type I and II collagen in association with healthy KOOS scores at follow-up. CPII measured at matriculation was negatively associated with incident injuries after adjustment for smoking status (p = 0.03, logistic regression), confirming results from AJIVE.ConclusionThese exploratory analyses suggest that CPII alone, or in combination with other joint metabolism biomarkers, may help identify individual risk of knee injury.