Objective This study aims to evaluate the time point and magnitude of peak effectiveness of exercise and the effects of various exercise modalities for osteoarthritis (OA) symptoms and to identify factors that significantly affect the effectiveness of exercise.Design Pharmacodynamic model-based meta-analysis (MBMA).Data sources Embase, PubMed, Cochrane Library, Web of Science and Scopus were searched for randomised controlled trials (RCTs) examining the effect of exercise for OA from inception to 20 November 2023.Eligibility criteria RCTs of exercise interventions in patients with knee, hip or hand OA, using Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC) subscales or Visual Analogue Scale (VAS) pain scores as outcome measures, were included. The minimum clinically important difference (MCID) for WOMAC total, pain, stiffness, function and VAS pain was 9.0, 1.6, 0.8, 5.4 and 0.9, respectively.Results A total of 186 studies comprising 12 735 participants with symptomatic or radiographic knee, hip or hand OA were included. The effectiveness of exercise treatments peaked at 1.6-7.2 weeks after initiation of exercise interventions. Exercise was more effective than the control, but the differences in the effects of exercise compared with control on all outcomes were only marginally different with the MCID (7.5, 1.7, 1.0, 5.4 and 1.2 units for WOMAC total, pain, stiffness, function and VAS pain, respectively). During a 12-month treatment period, local exercise (strengthening muscles and improving mobilisations of certain joints) had the best effectiveness (WOMAC pain decreasing by 42.5% at 12 weeks compared with baseline), followed by whole-body plus local exercise. Adding local water-based exercise (eg, muscle strengthening in warm water) to muscle strengthening exercise and flexibility training resulted in 7.9, 0.5, 0.7 and 8.2 greater improvements in the WOMAC total score, pain, stiffness and function, respectively. The MBMA models revealed that treatment responses were better in participants with more severe baseline symptom scores for all scales, younger participants for the WOMAC total and pain scales, and participants with obesity for the WOMAC function. Subgroup analyses revealed participants with certain characteristics, such as female sex, younger age, knee OA or more severe baseline symptoms on the WOMAC pain scale, benefited more from exercise treatment.Conclusion Exercise reaches peak effectiveness within 8 weeks and local exercise has the best effectiveness, especially if local water-based exercise is involved. Patients of female sex, younger age, obesity, knee OA or more severe baseline symptoms appear to benefit more from exercise treatment than their counterparts.

Background: Individuals with osteoarthritis (OA) often experience significant changes in their metabolic status and have a higher risk of mortality compared to the general population. However, no study has quantified the metabolic status of OA patients. Moreover, the association between metabolic status and risk of mortality among OA patients remains unclear. Objectives: To define the metabolic status in individuals with OA and to investigate its potential correlation with mortality among OA patients. Methods: This study included a total of 44,055 individuals with OA from the UK Biobank (UKBB) study as the training set, and a total of 3,376 participants with OA from the National Health and Nutrition Examination Survey (NHANES) 1999-2018, as the validating set. In UKBB, osteoarthritis diagnosis was primarily based on clinical assessment and patient-reported data, while in NHANES, OA classification relied predominantly on questionnaire responses. To assess metabolic status, a composite score was created based on factors including, body mass index, waist-hip ratio, blood pressure, fasting plasma glucose and hemoglobin A1c levels for blood glucose assessment, levels of C-reactive protein, triacylglycerols, and lipoproteins. The primary outcome of interest was all-cause mortality, which was collected via data linkage. Restricted cubic splines were used to determine the healthy range for each indicator in relation to all-cause mortality risk. Each metabolic factor was assigned a value of 1, if it felt within the healthy range, otherwise 0. The total metabolic score ranged from 0 to 7 points, with higher scores indicating healthier metabolic status. A score of 4 or more was considered as good metabolic health. The traditional definition of metabolic syndrome established by National Cholesterol Education Adult Treatment Panel III (ATPIII criteria) was employed as a (traditional) comparison. Associations between metabolic score, metabolic health, and all-cause mortality were examined in both UKBB and NHANES datasets, using Cox proportional hazards models. Age, sex, education level, Townsend deprivation index, smoking and drinking status, regular physical activity, and glucosamine use were adjusted in multivariable regression models. The Integrated Discrimination Improvement (IDI) were performed to assess the discriminatory capacity of different metabolic health definitions in predicting all-cause mortality over ten years, while the Net Reclassification Improvement (NRI) evaluated their reclassification performance for predicting all-cause mortality over the same timeframe. Results: Among the 44,055 participants with OA in the UKBB (mean age, 60.6 years; female, 60.4%), 76% were classified as metabolically unhealthy (mean scores: 3.86), while the remaining were categorized as metabolically healthy (mean scores: 1.26). Among the 3,376 participants in the NHANES dataset (mean age, 66.3 years; female, 63.7%), only 10.5% individuals were classified as metabolically healthy. After adjusting for the covariates, it was found that, individuals classified as metabolically unhealthy had a significantly increased risk of all-cause mortality [hazard ratios (HR), 1.49; 95%CI (1.4, 1.58) in UKBB and HR, 1.36; 95%CI (1.05, 1.77) in NHANES, respectively], compared to those who were metabolically healthy. In UKBB, compared to ATP III definition, the changes in IDI and NRI for our newly-established definition were 0.2 (0.1, 0.2) and 8.8 (0.9, 16.6), respectively. Our new definition showed a stronger association with mortality based on more prominent IDI and NRI than those of the ATP III definition. Conclusion: This study had proposed a new definition to quantify the metabolic status of individuals with OA. We found that OA individuals with poorer metabolic health had a significantly higher risk of mortality. Furthermore, our newly defined metabolic status in OA patients demonstrated superior predictive capabilities compared to the previous definition. REFERENCES: NIL. Acknowledgements: We thank all the participants and staff in the UK Biobank and NHANES. Disclosure of Interests: None declared.

Purpose (the aim of the study): Osteoarthritis (OA) not only leads to joint pain and reduced mobility but also poses significant risks to cardiovascular diseases (CVD) and CVD mortality. The physical activity recommendation from the World Health Organization can reduce the risk of CVD and mortality. However, participants with knee OA face difficulty in following regular physical activity, and the ideal exercise pattern for reducing CVD and mortality risks in individuals with OA remains unclear.

Background: Both social isolation and loneliness were associated with an increased risk of all-cause mortality in general population. However, there is still limited evidence elucidating the relationship between social isolation, loneliness, and all-cause mortality in osteoarthritis patients. Objectives: To investigate the prospective associations of social isolation and loneliness with risk of all-cause mortality in osteoarthritis patients and to compare the relative importance of social isolation and loneliness with traditional risk factors. Methods: A total of 55, 497 participants diagnosed with osteoarthritis from the UK Biobank were included. A two-item scale and a three-item scale were used to assess loneliness and isolation levels, respectively. The information of all-cause mortality was obtained through linkage to registries. Cox proportional hazard model was used to elucidate the relationship between isolation or loneliness and mortality in osteoarthritis participants. The relative importance of risk factors was measured by the R2 and concordance of the models. Population attributable fractions (PAFs) for all-cause mortality and isolation/ loneliness were estimated. Results: During a median follow-up of 12.57 years, 6, 186 all-cause deaths occurred. Compared to participants with the least social isolation score, those with a higher social isolation score were associated with a higher risk of all-cause mortality among individuals with osteoarthritis (hazard ratio [HR] 1.16, 95% confidence interval [CI] 1.10-1.23 for moderately isolation; HR 1.39, 95%CI 1.29-1.50 for most isolation; p for trend < 0.001). Stronger effect sizes were observed among female, younger individuals and non-obese individuals (all p for additive interaction < 0.1). Loneliness was significantly associated with a higher risk of all-cause mortality (HR 1.15, 95%CI 1.04-1.26). Social isolation ranked higher in relative strength for predicting all-cause mortality than loneliness, and common lifestyle risk factors (except for smoking) among individuals with osteoarthritis. In the PAFs analyses, 0.93% (95% CI 0.31%, 1.55%), 6.36% (95% CI 3.82%, 8.91%), and 8.57% (95% CI 6.54%, 10.59%) of deaths were attributable to loneliness, moderately isolation, and most isolation, respectively. Conclusion: Among individuals diagnosed with osteoarthritis, social isolation and loneliness were significantly associated with an increased risk of all-cause mortality, where isolation was a stronger predictor and exhibited a higher PAF for all-cause mortality than loneliness and common lifestyle risk factors except for smoking. These findings highlight the importance of social support, which could reduce premature mortality in individuals with osteoarthritis. REFERENCES: [1] Wang F, Gao Y, Han Z, et.al. A systematic review and meta-analysis of 90 cohort studies of social isolation, loneliness and mortality. Nat Hum Behav. 2023 Aug;7(8):1307-1319. doi: 10.1038/s41562-023-01617-6.[2] Wang X, Ma H, Li X, et.al. Joint association of loneliness and traditional risk factor control and incident cardiovascular disease in diabetes patients. Eur Heart J. 2023 Jul 21;44(28):2583-2591. doi: 10.1093/eurheartj/ehad306. Acknowledgements: This study makes use of data from UK Biobank (Project ID: 67654) and we thank the UK Biobank participants and the UK Biobank team for generating an important research resource. Disclosure of Interests: None declared.

Background: In studies regarding the association between alcohol consumption and mortality, controversy exists regarding the effects of light to moderate alcohol intake, and no study considered the role of alcohol metabolism related genetic variants in this relationship. Similarly, study investigating impact of alcohol consumption on the relationship between these genetic variations and mortality is also lacking. We therefore investigated the associations of alcohol consumption and alcohol metabolism related genetic variants with all-cause and disease-specific mortality, as well as the multiplicative interaction of alcohol consumption and the genetic variants on mortality.Methods: This prospective cohort study utilized data from the UK Biobank. Restricted cubic splines were used to evaluate the shapes of the associations between alcohol consumption and all-cause and disease-specific mortality. Cox proportional hazards models were used to estimate hazard ratios for associations between alcohol intake and mortality, both before and after stratifying by the genetic variants. Similarly, associations between the genetic variants and mortality were also examined before and after stratifying by alcohol intake.Findings: Alcohol consumption was J-shaped associated with all-cause and disease-specific mortality except for neurological-specific mortality. Beverage-specific alcohol intake also exhibited a J-shaped association with all-cause mortality, with varying overall hazard ratios. Rs1229984, a variant of alcohol dehydrogenase 1B (class I), beta polypeptide (ADH1B), can modify the associations between alcohol consumption and mortality, resulting in elevated risks of all-cause and disease-specific mortality for individuals without the T allele when compared to carriers. Carriers of the rs1229984 C allele exhibited an increased risk of overall mortality. In stratified analyses, the hazard ratios progressively rose from never drinkers to light drinkers, moderate drinkers, and ultimately to heavy drinkers.Interpretation: Our observation underscores the necessity of curbing excessive alcohol consumption to reduce mortality, particularly among individuals lacking the T allele of rs1229984.Funding: Guangdong Basic and Applied Basic Research, the National Natural Science Foundation of China, and Zhujiang Hospital Talent Recruitment Funds and Clinical Research Startup Program of Southern Medical University.Declaration of Interest: The authors declare no conflict of interest.

Objective To investigate the causal relationship between low bone mineral density (BMD) and osteoarthritis (OA) using Mendelian randomization (MR) design. Methods Two-sample bi-directional MR analyses were performed using summary-level information on OA traits from UK Biobank and arcOGEN. Sensitivity analyses including MR-Egger, simple median, weighted median, MR pleiotropy residual sum, and outlier approaches were utilized in conjunction with inverse variance weighting (IVW). Gene ontology (GO) enrichment analyses and expression quantitative trait locus (eQTL) colocalization analyses were used to investigate the potential mechanism and shared genes between osteoporosis (OP) and OA. Results The IVW method revealed that genetically predicted low femoral neck BMD was significantly linked with hip ( β = 0.105, 95% CI: 0.023–0.188) and knee OA ( β = 0.117, 95% CI: 0.049–0.184), but not with other site-specific OA. Genetically predicted low lumber spine BMD was significantly associated with OA at any sites ( β = 0.048, 95% CI: 0.011–0.085), knee OA ( β = 0.101, 95% CI: 0.045–0.156), and hip OA ( β = 0.150, 95% CI: 0.077–0.224). Only hip OA was significantly linked with genetically predicted reduced total bone BMD ( β = 0.092, 95% CI: 0.010–0.174). In the reverse MR analyses, no evidence for a causal effect of OA on BMD was found. GO enrichment analysis and eQTL analysis illustrated that DDN and SMAD-3 were the most prominent co-located genes. Conclusions These findings suggested that OP may be causally linked to an increased risk of OA, indicating that measures to raise BMD may be effective in preventing OA. More research is required to determine the underlying processes via which OP causes OA.

Objective This study aims to demonstrate the cellular composition and underlying mechanisms in subchondral bone marrow lesions (BMLs) of knee osteoarthritis (OA). Methods BMLs were assessed by MRI Osteoarthritis Knee Score (MOAKS)≥2. Bulk RNA-sequencing (bulk-seq) and BML-specific differentially expressed genes (DEGs) analysis were performed among subchondral bone samples (including OA-BML=3, paired OA-NBML=3; non-OA=3). The hub genes of BMLs were identified by verifying in independent datasets and multiple bioinformatic analyses. To further estimate cell-type composition of subchondral bone, we utilized two newly developed deconvolution algorithms (MuSiC, MCP-counter) in transcriptomic datasets, based on signatures from open-accessed single-cell RNA sequencing (scRNA-seq). Finally, competing endogenous RNA (ceRNA) and transcription factor (TF) networks were constructed through multiple predictive databases, and validated by public non-coding RNA profiles. Results A total of 86 BML-specific DEGs (up 79, down 7) were identified. IL11 and VCAN were identified as core hub genes. The “ has-miR-424-5p/lncRNA PVT1” was determined as crucial network, targeting IL11 and VCAN , respectively. More importantly, two deconvolution algorithms produced approximate estimations of cell-type composition, and the cluster of heterotopic-chondrocyte was discovered abundant in BMLs, and positively correlated with the expression of hub genes. Conclusion IL11 and VCAN were identified as the core hub genes of BMLs, and their molecular networks were determined as well. We profiled the characteristics of subchondral bone at single-cell level and determined that the heterotopic-chondrocyte was abundant in BMLs and was closely linked to IL11 and VCAN . Our study may provide new insights into the microenvironment and pathological molecular mechanism of BMLs, and could lead to novel therapeutic strategies.

Objective: To assess the causal effect of systemic iron status by using four biomarkers (serum iron; transferrin saturation; ferritin; total iron-binding capacity) on knee osteoarthritis (OA), hip OA, total knee replacement, and total hip replacement using 2-sample Mendelian randomization (MR) design. Methods: Three instrument sets were used to construct the genetic instruments for the iron status: Liberal instruments (variants associated with one of the iron biomarkers), sensitivity instruments (liberal instruments exclude variants associated with potential confounders), and conservative instruments (variants associated with all four iron biomarkers). Summary-level data for four OA phenotypes, including knee OA, hip OA, total knee replacement, and total hip replacement were obtained from the largest genome-wide meta-analysis with 826,690 individuals. Inverse-variance weighted based on the random-effect model as the main approach was conducted. Weighted median, MR-Egger, and Mendelian randomization pleiotropy residual sum and outlier methods were used as sensitivity MR approaches. Results: Based on liberal instruments, genetically predicted serum iron and transferrin saturation were significantly associated with hip OA and total hip replacement, but not with knee OA and total knee replacement. Statistical evidence of heterogeneity across the MR estimates indicated that mutation rs1800562 was the SNP significantly associated with hip OA in serum iron (odds ratio, OR = 1.48), transferrin saturation (OR = 1.57), ferritin (OR = 2.24), and total-iron binding capacity (OR = 0.79), and hip replacement in serum iron (OR = 1.45), transferrin saturation (OR = 1.25), ferritin (OR = 1.37), and total-iron binding capacity (OR = 0.80). Conclusion: Our study suggests that high iron status might be a causal factor of hip OA and total hip replacement where rs1800562 is the main contributor.

ABSTRACT OBJECTIVE To investigate the association of both individual and combined healthy lifestyle factors with the risk of all-cause mortality among patients with osteoarthritis (OA). DESIGN Prospective population-based cohort study. SETTING UK biobank and US National Health and Nutrition Examination Survey (US NHANES, 2007-2018) PARTICIPANTS 104, 142 UK participants with OA aged 39-72 years and 3, 472 US participants with OA aged 20-80 years. EXPOSURES Individual healthy lifestyle factors and a combined healthy lifestyle score were constructed from body mass index (BMI) and self-reported information on diet, sleep duration, physical activity, sedentary time, social connection, smoking and alcohol drinking. MAIN OUTCOME MEASURES All-cause mortality was the primary outcome in both studies. Secondary outcomes included cause-specific mortalities (cardiovascular, cancer, digestive and respiratory). Hazard ratios were adjusted for age, sex, economic situation, race, education and employment (UK biobank only). RESULTS UK Biobank documented 9,914 deaths during a median follow-up of 12.7 years, and US NHANES documented 463 deaths during a mean follow-up of 6.01 years. For all-cause mortality using restricted cubic spline graph (RCS) models, sleep duration had a U-shaped (with a nadir at 7 hours/day), moderate physical activity (MPA) had an L-shaped (with a turning point at 550 minutes/week), while BMI, vigorous physical activity (VPA) and sedentary time had J-shaped (with turning points at 28 kg/m 2 , 240 minutes/week and 5 hours/day, respectively) associations in the UK biobank. Similar results were observed in US NHANES. In multivariable Cox models, each healthy lifestyle factor was significantly associated with all-cause mortality (hazard ratio [HRs] range 0.49 to 0.84 for UK biobank, and 0.26 to 0.73 for US NHANES), and HRs (95% CI) for associations with combined healthy lifestyle score (scoring 6-8 vs. 0-2) were 0.38 (0.35, 0.41) in UK biobank and healthy lifestyle score (scoring 5-7 vs. 0-1) were 0.20 (0.13, 0.31) in US NHANES for all-cause mortality. The results for cause-specific mortality were largely similar and consistent across two cohorts. CONCLUSIONS The nonlinear relationships suggested patients with OA had the lowest risk of all-cause mortality when BMI was 28 kg/m 2 , sleep was 7 hours/day, VPA was 240 minutes/week, sedentary time was less than 5 hours/day, MPA was more than 550 minutes/week. The newly constructed healthy lifestyle score for OA population was associated with a significantly lower risk of all-cause mortality. WHAT IS ALREADY KNOWN ON THIS TOPIC Healthy lifestyles are thought to reduce the risk of multiple causes of mortality in the general population. People with osteoarthritis (OA) are at higher risk of mortality than the general population. However, evidence of associations between combined healthy lifestyle and risks of all-cause and cause-specific mortality among OA patients are lacking. Whether and what kind of healthy lifestyles in patients with OA could offset the risk of mortality are unknown. WHAT THIS STUDY ADDS By using two nationwide cohort studies in UK and US, a comprehensive healthy lifestyle pattern that integrates 8 or 7 lifestyle factors was established among OA individuals for the first time. People who had the highest combined healthy lifestyle score were significantly associated with 62% to 80% lower risk of all-cause mortality compared with those who had the lowest score in both the UK and US OA populations.

BACKGROUND:Large adulthood body size was associated with increased risk of osteoarthritis. We aimed to examine the association between body size trajectories from childhood to adulthood and potential interactions with genetic susceptibility on osteoarthritis risk. METHODS:We included participants from the UK Biobank aged 38-73 years in 2006-10. Childhood body size information was collected by questionnaire. Adulthood BMI was assessed and transformed into three categories (<25 kg/m2 for normal, 25-29·9 kg/m2 for overweight, and >30 kg/m2 for obesity). A Cox proportional hazards regression model was applied to assess the association between body size trajectories and osteoarthritis incidence. Osteoarthritis-related polygenic risk score (PRS) was constructed to evaluate its interactions with body size trajectories on osteoarthritis risk. FINDINGS:For the 466 292 participants included, we identified nine body size trajectories [thinner to normal (11·6%), overweight (17·2%), or obesity (26·9%); average to normal (11·8%), overweight (16·2%), or obesity (23·7%); and plumper to normal (12·3%), overweight (16·2%), or obesity (23·6%)]. Compared with individuals in the average-to-normal group, all other trajectory groups had higher risks of osteoarthritis, after adjustment for demographic, social-economic and lifestyle covariates (hazard ratios [HRs] 1·05-2·41; all p<0·01). Among them, thinner-to-obesity (HR 2·41; 95% CI 2·23-2·49) had the most prominent association with increased osteoarthritis risk. A high PRS was significantly associated with an increased risk of osteoarthritis (1·14; 1·11-1·16), whereas no interaction between childhood-to-adulthood body size trajectories and PRS on osteoarthritis risks was observed. The population attributable fraction suggested that body size towards normal in adulthood could eliminate osteoarthritis cases by 18·67% for thinner-to-overweight to 38·74% for plumper-to-obesity. INTERPRETATION:Average-to-normal body size seems to be the healthiest childhood-to-adulthood trajectory for osteoarthritis risk, whereas a trajectory of increased body size from thinner to obesity has the highest risk for osteoarthritis. These associations are independent of osteoarthritis genetic susceptibility. FUNDING:The National Natural Science Foundation of China (32000925) and Guangzhou Science and Technology Program (202002030481).

Mesenchymal stem cells (MSCs) therapy shows the potential benefits to relieve clinical symptoms of osteoarthritis (OA), but it is uncertain if it can repair articular cartilage lesions - the main pathology of OA. Here, we prepared biomimetic cupper sulfide@phosphatidylcholine (CuS@PC) nanoparticles (NPs) loaded with plasmid DNA (pDNA) encoding transforming growth factor-beta 1 (TGF-β1) to engineer MSCs for enhanced OA therapy via cartilage regeneration. We found that the NPs not only promoted cell proliferation and migration, but also presented a higher pDNA transfection efficiency relative to commercial transfection reagent lipofectamine 3000. The resultant CuS/TGF-β1@PC NP-engineered MSCs (termed CTP-MSCs) were better than pure MSCs in terms of chondrogenic gene expression, glycosaminoglycan deposition and type II collagen formation, favoring cartilage repair. Further, CTP-MSCs inhibited extracellular matrix degradation in interleukin-1β-induced chondrocytes. Consequently, intraarticular administration of CTP-MSCs significantly enhanced the repair of damaged cartilage, whereas pure MSCs exhibited very limited effects on cartilage regeneration in destabilization of the medial meniscus (DMM) surgical instability mice. Hence, this work provides a new strategy to overcome the limitation of current stem cell therapy in OA treatment through developing more effective nanoengineered MSCs.

Objectives: We aimed to examine whether metformin (MET) use is associated with a reduced risk of total knee arthroplasty (TKA) and low severity of knee pain in patients with knee osteoarthritis (OA) and diabetes and/or obesity. Methods: Participants diagnosed with knee OA and diabetes and/or obesity from June 2000 to July 2019 were selected from the information system of a local hospital. Regular MET users were defined as those with recorded prescriptions of MET or self-reported regular MET use for at least 6 months. TKA information was extracted from patients’ surgical records. Knee pain was assessed using the numeric rating scale. Log-binomial regression, linear regression, and propensity score weighting (PSW) were performed for statistical analyses. Results: A total of 862 participants were included in the analyses. After excluding missing data, there were 346 MET non-users and 362 MET users. MET use was significantly associated with a reduced risk of TKA (prevalence ratio: 0.26, 95% CI: 0.15 to 0.45, p < 0.001), after adjustment for age, gender, body mass index, various analgesics, and insurance status. MET use was significantly associated with a reduced degree of knee pain after being adjusted for the above covariates (β: −0.48, 95% CI: −0.91 to −0.05, p = 0.029). There was a significantly accumulative effect of MET use on the reduced risk of TKA. Conclusion: MET can be a potential therapeutic option for OA. Further clinical trials are needed to determine if MET can reduce the risk of TKA and the severity of knee pain in metabolic-associated OA patients.

Aims To investigate whether the associations between cartilage defects and cartilage volumes with changes in knee symptoms were mediated by osteophytes. Methods Data from the Vitamin D Effects on Osteoarthritis (VIDEO) study were analyzed as a cohort. The Western Ontario and McMaster Universities Osteoarthritis Index was used to assess knee symptoms at baseline and follow-up. Osteophytes, cartilage defects, and cartilage volumes were measured using magnetic resonance imaging at baseline. Associations between cartilage morphology and changes in knee symptoms were assessed using linear regression models, and mediation analysis was used to test whether these associations were mediated by osteophytes. Results A total of 334 participants (aged 50 to 79 years) with symptomatic knee osteoarthritis were included in the analysis. Cartilage defects were significantly associated with change in total knee pain, change in weight-bearing pain, and change in non-weight-bearing pain after adjustment for age, sex, body mass index, and intervention. Cartilage volume was significantly associated with change in weight-bearing pain and change in physical dysfunction after adjustment. Lateral tibiofemoral and patellar osteophyte mediated the associations of cartilage defects with change in total knee pain (49-55%) and change in weight-bearing pain (61-62%) and the association of cartilage volume with change in weight-bearing pain (27-30%) and dysfunction (24-25%). Both cartilage defects and cartilage volume had no direct effects on change in knee symptoms. Conclusions The significant associations between cartilage morphology and changes in knee symptoms were indirect and were partly mediated by osteophytes.

Background Periodontitis is a chronic immuno-inflammatory disease characterized by inflammatory destruction of tooth-supporting tissues. Its pathogenesis involves a dysregulated local host immune response that is ineffective in combating microbial challenges. An integrated investigation of genes involved in mediating immune response suppression in periodontitis, based on multiple studies, can reveal genes pivotal to periodontitis pathogenesis. Here, we aimed to apply a deep learning (DL)-based autoencoder (AE) for predicting immunosuppression genes involved in periodontitis by integrating multiples omics datasets. Methods Two periodontitis-related GEO transcriptomic datasets (GSE16134 and GSE10334) and immunosuppression genes identified from DisGeNET and HisgAtlas were included. Immunosuppression genes related to periodontitis in GSE16134 were used as input to build an AE, to identify the top disease-representative immunosuppression gene features. Using K-means clustering and ANOVA, immune subtype labels were assigned to disease samples and a support vector machine (SVM) classifier was constructed. This classifier was applied to a validation set (Immunosuppression genes related to periodontitis in GSE10334) for predicting sample labels, evaluating the accuracy of the AE. In addition, differentially expressed genes (DEGs), signaling pathways, and transcription factors (TFs) involved in immunosuppression and periodontitis were determined with an array of bioinformatics analysis. Shared DEGs common to DEGs differentiating periodontitis from controls and those differentiating the immune subtypes were considered as the key immunosuppression genes in periodontitis. Results We produced representative molecular features and identified two immune subtypes in periodontitis using an AE. Two subtypes were also predicted in the validation set with the SVM classifier. Three “master” immunosuppression genes, PECAM1, FCGR3A, and FOS were identified as candidates pivotal to immunosuppressive mechanisms in periodontitis. Six transcription factors, NFKB1, FOS, JUN, HIF1A, STAT5B, and STAT4, were identified as central to the TFs-DEGs interaction network. The two immune subtypes were distinct in terms of their regulating pathways. Conclusion This study applied a DL-based AE for the first time to identify immune subtypes of periodontitis and pivotal immunosuppression genes that discriminated periodontitis from the healthy. Key signaling pathways and TF-target DEGs that putatively mediate immune suppression in periodontitis were identified. PECAM1, FCGR3A, and FOS emerged as high-value biomarkers and candidate therapeutic targets for periodontitis.