Advertisement

Skin autofluorescence, a non-invasive biomarker of advanced glycation end products, and its relation to radiographic and MRI based osteoarthritis

Open AccessPublished:September 07, 2022DOI:https://doi.org/10.1016/j.joca.2022.08.014

      Summary

      Objectives

      Accumulation of advanced glycation end products (AGEs) in articular cartilage during aging has been proposed as a mechanism involved in the development of osteoarthritis (OA). Therefore, we investigated a cross-sectional relationship between skin AGEs, a biomarker for systemic AGEs accumulation, and OA.

      Methods

      Skin AGEs were estimated with the AGE Reader™ as skin autofluorescence (SAF). Knee and hip X-rays were scored according to Kellgren and Lawrence (KL) system. KL-sum score of all four joints was calculated per participant to assess severity of overall radiographic OA (ROA) including or excluding those with prosthesis. Knee MRI of tibiofemoral joint (TFMRI) was assessed for cartilage loss. Sex-stratified regression models were performed after testing interaction with SAF.

      Results

      2,153 participants were included for this cross-sectional analysis. In women (n = 1,206) for one unit increase in SAF, the KL-sum score increased by 1.15 (95% confidence interval = 1.00–1.33) but excluding women with prosthesis, there was no KL-sum score increase [0.96 (0.83–1.11)]. SAF was associated with higher prevalence of prosthesis [Odds ratio, OR = 1.67 (1.10–2.54)] but not with ROA [OR = 0.83 (0.61–1.14)] when compared to women with no ROA. In men (n = 947), there was inconclusive association between SAF and KL sum score or prosthesis.
      For TFMRI (n = 103 women), SAF was associated with higher prevalence of cartilage loss, full-thickness [OR = 5.44 (1.27–23.38)] and partial-thickness [OR = 1.45 (0.38–5.54)], when compared to participants with no cartilage loss.

      Conclusion

      Higher SAF in women was associated with higher prosthesis prevalence and a trend towards higher cartilage loss on MRI. Our data presents inconclusive results between SAF and ROA in both sexes.

      Keywords

      Introduction

      Pathological changes in articular cartilage with older age play a pivotal role in progression of osteoarthritis (OA). The exact underlying mechanisms by which aging increases the predisposition to OA has, as of yet, not been elucidated. During the last decade, a key mediator identified in the pathogenesis of aging-related diseases, such as OA, is chronic low grade inflammation
      • Robinson W.H.
      • Lepus C.M.
      • Wang Q.
      • Raghu H.
      • Mao R.
      • Lindstrom T.M.
      • et al.
      Low-grade inflammation as a key mediator of the pathogenesis of osteoarthritis.
      ,
      • Salminen A.
      Feed-forward regulation between cellular senescence and immunosuppression promotes the aging process and age-related diseases.
      which, among other effects, is thought to trigger the formation of advanced glycation end-products (AGEs) in multiple tissues
      • Maillard-Lefebvre H.
      • Boulanger E.
      • Daroux M.
      • Gaxatte C.
      • Hudson B.I.
      • Lambert M.
      Soluble receptor for advanced glycation end products: a new biomarker in diagnosis and prognosis of chronic inflammatory diseases.
      . AGEs are a diverse cluster of compounds formed non-enzymatically on proteins when reducing sugars react to amino acid residues and undergo a series of biochemical modifications
      • Singh R.
      • Barden A.
      • Mori T.
      • Beilin L.
      Advanced glycation end-products: a review.
      . AGEs accumulation on collagen in articular cartilage alters extracellular matrix composition leading to stiffness which makes it prone to damage
      • DeGroot J.
      • Verzijl N.
      • Budde M.
      • Bijlsma J.W.
      • Lafeber F.P.
      • TeKoppele J.M.
      Accumulation of advanced glycation end products decreases collagen turnover by bovine chondrocytes.
      • DeGroot J.
      • Verzijl N.
      • Jacobs K.M.
      • Budde M.
      • Bank R.A.
      • Bijlsma J.W.
      • et al.
      Accumulation of advanced glycation endproducts reduces chondrocyte-mediated extracellular matrix turnover in human articular cartilage.
      • DeGroot J.
      • Verzijl N.
      • Wenting-van Wijk M.J.
      • Jacobs K.M.
      • Van El B.
      • Van Roermund P.M.
      • et al.
      Accumulation of advanced glycation end products as a molecular mechanism for aging as a risk factor in osteoarthritis.
      . In addition, binding of AGEs to the receptor for AGEs (RAGE) on chondrocytes has been shown to induce oxidative and inflammatory stress thereby promoting cartilage matrix degradation
      • Loeser R.F.
      • Yammani R.R.
      • Carlson C.S.
      • Chen H.
      • Cole A.
      • Im H.J.
      • et al.
      Articular chondrocytes express the receptor for advanced glycation end products: potential role in osteoarthritis.
      . Altogether, AGEs accumulation in long-lived articular cartilage proteins may be involved in OA pathogenesis.
      Previous studies have used AGEs levels in serum or urine, such as pentosidine (PEN), Methylglyoxal-derived hydroimidazolone (MG-H1) and n-6-Carboxymethyllysine (CML), to study a relationship with OA. Serum PEN levels were higher in OA (n = 38) participants when compared to controls but there was no correlation of either serum or synovial fluid PEN with radiographic severity of knee OA
      • Senolt L.
      • Braun M.
      • Olejarova M.
      • Forejtova S.
      • Gatterova J.
      • Pavelka K.
      Increased pentosidine, an advanced glycation end product, in serum and synovial fluid from patients with knee osteoarthritis and its relation with cartilage oligomeric matrix protein.
      . Urinary PEN levels did not predict progressive knee cartilage loss obtained from MRI's in 127 participants with already established OA
      • Hunter D.J.
      • Lavalley M.
      • Li J.
      • Zhang Y.
      • Bauer D.
      • Nevitt M.
      • et al.
      Urinary pentosidine does not predict cartilage loss among subjects with symptomatic knee OA: the BOKS Study.
      . Higher levels of urinary PEN were found in persons with erosive hand OA compared to non-erosive OA and in those with more hand pain and stiffness
      • Braun M.
      • Hulejova H.
      • Gatterova J.
      • Filkova M.
      • Pavelkova A.
      • Sleglova O.
      • et al.
      Pentosidine, an advanced glycation end-product, may reflect clinical and morphological features of hand osteoarthritis.
      . These studies were heterogeneous with respect to AGEs estimated in various body fluids such as serum, urine or synovial fluid and type of AGE determined such as CML, MG-H1 or PEN. Serum and urine AGE levels have been shown to fluctuate based on e.g., diet and renal clearance
      • Scheijen J.
      • Hanssen N.M.J.
      • van Greevenbroek M.M.
      • Van der Kallen C.J.
      • Feskens E.J.M.
      • Stehouwer C.D.A.
      • et al.
      Dietary intake of advanced glycation endproducts is associated with higher levels of advanced glycation endproducts in plasma and urine: the CODAM study.
      ,
      • Rabbani N.
      • Thornalley P.J.
      Advanced glycation end products in the pathogenesis of chronic kidney disease.
      which might have obscured the relationship between AGEs and OA.
      AGEs measured in tissues might represent long-term AGEs burden due to long half-life of AGEs bound to proteins in tissues
      • Jiang J.
      • Zhang Y.
      • Chen J.
      • Yang X.
      • Mei C.
      • Xiong F.
      • et al.
      Serum and tissue levels of advanced glycation end products and risk of mortality in patients on maintenance hemodialysis.
      , although detecting tissue AGEs is often non-feasible due to its invasive nature. A non-invasive and reproducible technique has been used since the last decade to estimate skin (tissue) AGEs with the help of an AGE reader™
      • Meerwaldt R.
      • Links T.
      • Graaff R.
      • Thorpe S.R.
      • Baynes J.W.
      • Hartog J.
      • et al.
      Simple noninvasive measurement of skin autofluorescence.
      . As skin AGEs are predominantly bound to skin collagen which has been estimated to have a half-life of ∼14 years
      • Verzijl N.
      • DeGroot J.
      • Thorpe S.R.
      • Bank R.A.
      • Shaw J.N.
      • Lyons T.J.
      • et al.
      Effect of collagen turnover on the accumulation of advanced glycation end products.
      , skin AGEs are assumed to remain quite stable over time. Additionally, studies using human cadavers have shown that skin and cartilage pentosidine levels per milligram of collagen are moderately correlated (R = 0.473) to each other [11]. Therefore, we used non-invasive skin AGEs measurements as a proxy of cartilage AGEs accumulation over an extended period of time in order to study its relationship with OA.
      In the present study, we aimed to investigate a potential relationship of skin AGEs with hip and knee osteoarthritis based on X-rays and with cartilage loss on knee MRI in participants from a population-based cohort.

      Methods

      Study population

      This study was embedded within the Rotterdam Study (RS), an ongoing prospective population-based cohort study of Dutch individuals living in the Ommoord district of Rotterdam in the Netherlands
      • Ikram M.A.
      • Brusselle G.
      • Ghanbari M.
      • Goedegebure A.
      • Ikram M.K.
      • Kavousi M.
      • et al.
      Objectives, design and main findings until 2020 from the Rotterdam Study.
      . In short, participants were included at three different points in time, namely in 1990, 2000 and 2006 and named RS-I, RS-II (≥55 years) and RS-III (≥45 years) sub-cohorts based on year of inclusion, respectively. After the inclusion visit, participants visited the research center every 4–6 years for follow-up examinations. The Rotterdam Study has been approved by the Medical Ethics Committee of the Erasmus MC. All participants provided written informed consent.
      For this cross-sectional analysis, we included 2,153 participants from RS-I (6th follow-up, n = 629 or 29%), RS-II (4th, follow-up, n = 959 or 45%) and RS-III (2nd baseline visit, n = 565 or 26%) unless specified otherwise. Briefly, 2,196 out of 3,001 participants with skin autofluorescence (SAF) had data on the outcome i.e., knee or hip X-rays. Afterwards, we excluded only 43 participants with missing data on covariates including effective glomerular filtration rate (eGFR), body mass index (BMI), smoking and diabetes status (Fig. 1). No imputation was performed for this small subset of participants (n = 43) with missing data on covariates.
      Fig. 1
      Fig. 1Flowchart of participant inclusion from the Rotterdam study. RS, Rotterdam study; SAF, skin autofluorescence; BMI, body mass index; MRI, magnetic resonance imaging; eGFR, effective glomerular filtration rate.

      Skin autofluorescence

      Skin AGEs were estimated using an AGE reader™ (DiagnOptics Technologies B.V., Groningen, The Netherlands) as SAF
      • Waqas K.
      • Chen J.
      • Koromani F.
      • Trajanoska K.
      • van der Eerden B.C.
      • Uitterlinden A.G.
      • et al.
      Skin autofluorescence, a noninvasive biomarker for advanced glycation end-products, is associated with prevalent vertebral and major osteoporotic fractures: the Rotterdam Study.
      . Participants were advised not to apply lotions or creams on the dominant arm for 2 days preceding SAF measurement. Briefly, AGE Reader illuminates a skin surface of ∼4 cm2 with a black light source, guarded against surrounding light. The AGE Reader uses an excitation light source between 300 and 420 nm (peak excitation ∼350 nm) for illumination and measures emission in the 420–600 nm range. The measure of autofluorescence we apply is the average light intensity per nanometer in the range between 420 and 600 nm divided by the average light intensity per nanometer in the range between 300 and 420 nm (in arbitrary units [a.u.])
      • Beisswenger P.J.
      • Howell S.
      • Mackenzie T.
      • Corstjens H.
      • Muizzuddin N.
      • Matsui M.S.
      Two fluorescent wavelengths, 440(ex)/520(em) nm and 370(ex)/440(em) nm, reflect advanced glycation and oxidation end products in human skin without diabetes.
      ,
      • Atzeni I.M.
      • van de Zande S.C.
      • Westra J.
      • Zwerver J.
      • Smit A.J.
      • Mulder D.J.
      The AGE Reader: a non-invasive method to assess long-term tissue damage.
      . We have chosen this ratio to correct for the influence on autofluorescence by light absorption due to, for example, skin pigmentation. An automated software in the AGE reader ensured the incorporation of skin reflectance values between 6 and 10% (corresponding to Fitzpatrick type V) in SAF values and exclusion of participants with skin reflectance under 6%
      • Koetsier M.
      • Nur E.
      • Chunmao H.
      • Lutgers H.L.
      • Links T.P.
      • Smit A.J.
      • et al.
      Skin color independent assessment of aging using skin autofluorescence.
      . Values were defined as outliers in SAF and excluded from the analysis if it exceed the scope of mean + 4SD; based on this 8 participants were excluded.

      Plain radiographic evaluation of osteoarthritis

      Within the participants with SAF measurements, knee and hip radiographs were available for 2,153 participants. Participants from the cohort RS-I (29%) and RS-II (45%) underwent X-rays on average 4–5 years before SAF measurements. RS-III (26%) had cross-sectional measurements of both SAF and X-rays of knee and hip joint. Knee radiographs were taken with the knee and patella in an extended and central position respectively
      • Hoeven T.A.
      • Kavousi M.
      • Clockaerts S.
      • Kerkhof H.J.M.
      • van Meurs J.B.
      • Franco O.
      • et al.
      Association of atherosclerosis with presence and progression of osteoarthritis: the Rotterdam Study.
      . Radiographs of the pelvis including hips were obtained when both feet were rotated 10° inward and the X-ray beam was centred on the umbilicus
      • Hoeven T.A.
      • Kavousi M.
      • Clockaerts S.
      • Kerkhof H.J.M.
      • van Meurs J.B.
      • Franco O.
      • et al.
      Association of atherosclerosis with presence and progression of osteoarthritis: the Rotterdam Study.
      .
      We defined OA according to the original Kellgren and Lawrence (KL) system
      • Kellgren J.H.
      • Lawrence J.S.
      Radiological assessment of osteo-arthrosis.
      ,
      • Schiphof D.
      • Boers M.
      • Bierma-Zeinstra S.M.
      Differences in descriptions of Kellgren and Lawrence grades of knee osteoarthritis.
      graded scale (0–4) based on features including joint space narrowing (JSN), osteophytes and sclerosis. Briefly, trained individuals assessed X-rays and labeled them as no, possible, definite or marked JSN and osteophytes and reported KL score per joint (see Supplementary Table 1)
      • Schiphof D.
      • Boers M.
      • Bierma-Zeinstra S.M.
      Differences in descriptions of Kellgren and Lawrence grades of knee osteoarthritis.
      . KL score of 0 or 1 was defined as absence of radiographic osteoarthritis (no ROA) and a KL score of 2–4 or a KL score of 5 (total joint replacement) as presence of radiographic osteoarthritis (ROA). KL score of 5 was included to take the most severe form of OA into account. KL scores were subsequently summed for right and left joints to form the hip or knee KL sum score (0–10) and for all four joints together Hip + Knee KL sum score (0–20) per participant.

      MRI evaluation for knee joint osteoarthritis

      103 women from the cohort RS-III (2nd follow-up) had both SAF measurements and knee MRI imaging
      • Schiphof D.
      • Oei E.H.
      • Hofman A.
      • Waarsing J.H.
      • Weinans H.
      • Bierma-Zeinstra S.M.
      Sensitivity and associations with pain and body weight of an MRI definition of knee osteoarthritis compared with radiographic Kellgren and Lawrence criteria: a population-based study in middle-aged females.
      . MRI was performed on the knee joint using multi-sequence MRI protocol on a 1.5-T MRI scanner (Sigma Excite 2, General Electric Healthcare, Milwaukee, Wisconsin, USA), as reported elsewhere
      • Schiphof D.
      • Oei E.H.
      • Hofman A.
      • Waarsing J.H.
      • Weinans H.
      • Bierma-Zeinstra S.M.
      Sensitivity and associations with pain and body weight of an MRI definition of knee osteoarthritis compared with radiographic Kellgren and Lawrence criteria: a population-based study in middle-aged females.
      . Briefly, MRIs were scored with the semi-quantitative MRI Osteoarthritis Knee Score (MOAKS)
      • Hunter D.J.
      • Guermazi A.
      • Lo G.H.
      • Grainger A.J.
      • Conaghan P.G.
      • Boudreau R.M.
      • et al.
      Evolution of semi-quantitative whole joint assessment of knee OA: MOAKS (MRI Osteoarthritis Knee Score).
      . Two experienced readers were extensively trained by an highly experienced musculoskeletal radiologist, as described in a short report
      • Runhaar J.
      • Schiphof D.
      • van Meer B.
      • Reijman M.
      • Bierma-Zeinstra S.M.
      • Oei E.H.
      How to define subregional osteoarthritis progression using semi-quantitative MRI osteoarthritis knee score (MOAKS).
      . In short, the knee was divided into 14 sub-regions (medial/lateral patella, medial/lateral trochlea, medial/lateral central femur, medial/lateral posterior femur and medial (anterior, central and posterior) and lateral (anterior, central and posterior) tibia for scoring articular cartilage and bone marrow lesions (BML). Articular cartilage lesions were scored as partial or full thickness loss based on: (1) size of any cartilage loss (including partial and full-thickness loss) as a percentage of surface area as related to the size of each individual region; and (2) as a percentage full-thickness cartilage loss of the region (0 = none, 1 = <10% of region of cartilage surface area, 2 = 10–75% of region of cartilage surface area, 3 = >75% of region of cartilage surface area). Bone marrow lesions (BMLs) and cysts were scored based on the size relative to each subregion (0 = none, 1 = <33%, 2 = 33–66%, and 3 = >66%) and were considered to be present when grade ≥1. Osteophytes were scored based on size (0 = none, 1 = small, 2 = medium, 3 = large) in each of the twelve locations and were classified as definite when grade ≥2. For the present study we focus on the tibiofemoral joint as assessed in the conventional radiographs and therefore left out the patella and medial and lateral trochlear facet femur. The MRI definition for tibiofemoral OA (TFOAMRI) was presence of a definite osteophyte and full thickness cartilage loss, or one of these features plus two of the following features: 1) subchondral BML or cyst 2) meniscal subluxation, maceration or degeneration 3) partial thickness cartilage loss 4) bone attrition
      • Hunter D.J.
      • Arden N.
      • Conaghan P.G.
      • Eckstein F.
      • Gold G.
      • Grainger A.
      • et al.
      Definition of osteoarthritis on MRI: results of a Delphi exercise.
      .

      Covariates

      Height and weight were recorded in standing position at the research Centre without shoes and BMI was computed as weight in kilograms divided by height in meters squared. Smoking was obtained through self-reporting during home interviews and was classified as current, past or never smokers. Type 2 Diabetes Mellitus was defined by combining the information on antidiabetic medication use, fasting blood glucose levels and diagnosis in the GP registries. Serum creatinine and fasting glucose were measured through automated enzymatic method. Effective glomerular filtration rate (eGFR) was calculated by the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation using serum creatinine concentration, age and sex data
      • Hofman A.
      • Brusselle G.G.
      • Darwish Murad S.
      • van Duijn C.M.
      • Franco O.H.
      • Goedegebure A.
      • et al.
      The Rotterdam Study: 2016 objectives and design update.
      .

      Statistical analysis

      We identified potential confounders in the relationship between exposure (SAF) and outcome (OA) based on literature evidence. We consistently adjusted for age, sex and RS cohorts in our model 1 and additionally for BMI
      • den Engelsen C.
      • van den Donk M.
      • Gorter K.J.
      • Salomé P.L.
      • Rutten G.E.
      Advanced glycation end products measured by skin autofluorescence in a population with central obesity.
      , eGFR
      • Mallipattu S.K.
      • Uribarri J.
      Advanced glycation end product accumulation: a new enemy to target in chronic kidney disease?.
      ,
      • Oleniuc M.
      • Secara I.
      • Onofriescu M.
      • Hogas S.
      • Voroneanu L.
      • Siriopol D.
      • et al.
      Consequences of advanced glycation end products accumulation in chronic kidney disease and clinical usefulness of their assessment using a non-invasive technique - skin autofluorescence.
      , smoking
      • van Waateringe R.P.
      • Mook-Kanamori M.J.
      • Slagter S.N.
      • van der Klauw M.M.
      • van Vliet-Ostaptchouk J.V.
      • Graaff R.
      • et al.
      The association between various smoking behaviors, cotinine biomarkers and skin autofluorescence, a marker for advanced glycation end product accumulation.
      and diabetes status
      • Singh V.P.
      • Bali A.
      • Singh N.
      • Jaggi A.S.
      Advanced glycation end products and diabetic complications.
      in our model 2 as covariates. During all the analyses, interaction terms were tested between SAF and diabetes status and sex in the multivariate fully adjusted models, and results were reported stratified where statistically significant (P for interaction ≤0.10 for each) interaction was found. Depending on the nature of outcome, different generalized linear models were used (1) KL sum score were not normally distributed (left-skewed with an overabundance of zeros as shown in Supplementary Fig. 1), we assessed its association with SAF using a zero-inflated negative binomial (ZINB) regression model, which can account for the non-normal distribution including overdispersion and zero-inflation (KL sum score: 0–10 for both knees or both hips and 0–20 per participant for combined hip + knee joints). We reported the results from the count model part of the ZINB in all our main analysis and from the zero-inflation model only in the supplementary section; (2) multinomial logistic regression when participants with no ROA (KL-score ≤1, coded as 0) were compared with those with ROA (KL-score ≥2 and ≤ 4, coded as 1) and those with prosthesis (coded as 2) assuming to represent the most severe OA cases; (3) Multinomial logistic regression was used to evaluate tibiofemoral knee MRI data (TFMRI), when participants who had no cartilage loss (coded as 0) on TFMRI were compared with those with partial (coded as 1) and full (coded as 2) thickness cartilage loss. Due to a small subgroup of women with MRI knee and relatively low number of events, we adjusted all the analyses on MRI data only for age and BMI in order to prevent overfitting of the model. For logistic regression, nonlinear association with OA was explored by performing quintiles analyses and adding a quadratic term for SAF to the original model. However, for all analyses, a linear model had the best fit. All underlying assumptions were checked and met for the statistical models, whenever we proceeded with any analyses. All analyses were performed through IBM SPSS statistics 25 (version 25.0).

      Results

      Cohort profile based on availability of SAF and X-rays of hip and knee

      Table I shows the demographic and radiographic characteristics of all included participants and also based on sex. Men had higher SAF values (2.55 ± 0.49 A.U. vs 2.31 ± 0.46 A.U.), lower mean KL score knee (1.19 ± 1.68 vs 1.61 ± 2.03), fewer participants with prevalent knee OA (18% vs 27.5%) and hip + knee prosthesis (6% vs 11%) than women. Age, BMI, eGFR, mean KL score hip and prevalence of hip OA did not differ between the two sexes. Supplementary Tables 2 and 3 shows the results of demographic and radiographic characteristics of study population based on age and sex adjusted SAF tertiles and based on no ROA, ROA and presence of prosthesis.
      Table IGeneral characteristics of Rotterdam study cohort with both skin autofluorescence (SAF) and hip and knee joint imaging data
      AllMenWomenWomen with MRI knee
      Knee MRI's were only available for an all-female obese subgroup of the Rotterdam Study with SAF measurments (n = 103).
      Cohort participants2,1539471,206103
      Skin autofluorescence2.43 ± 0.492.55 ± 0.492.31 ± 0.46##2.17 ± 0.40∗∗
      BMI (kg/m2)27.6 ± 4.227.3 ± 3.427.5 ± 4.526.9 ± 3.9
      Age (years)74.9 ± 7.874.6 ± 7.474.5 ± 7.762.8 ± 2.48∗∗∗
      eGFR (ml/min/1.73m2)75.9 ± 14.275.7 ± 13.9876.4 ± 13.686.7 ± 10.3∗∗
      Diabetes status (y/n)297 (14%)148 (16%)149 (12%)#5 (5%)∗∗
      Physical activity (n = 1,789)42.9 (17–85)42.0 (17.5–78)43.8 (17–89)46.0 (53.7)∗
      Smoking status (y/n)#
       Never32%19%42%34%
       Past55%65%47%49%
       Current13%15%11%17%
      KL sum score hip + knee2.39 ± 2.782.20 ± 2.562.53 ± 2.93##NA
      Hip or knee OA (y/n)715 (33%)276 (29%)439 (36%)###NA
      Hip or knee prosthesis185 (9%)57 (6%)128 (11%)###NA
      Hip KL sum score0.96 ± 1.811.01 ± 1.710.93 ± 1.88NA
      Hip OA (y/n)302 (14%)133 (14%)169 (14%)NA
      Hip prosthesis112 (5%)37 (4%)75 (6%)#NA
      Knee KL sum score1.42 ± 1.901.19 ± 1.681.61 ± 2.03NA
      Knee OA (y/n)508 (23.5%)174 (18%)334 (27.5%)#28 (27%)
      Knee OA was based on Knee Osteoarthritis Scoring System (KOSS) on MRI tibiofemoral joint.
      Knee prosthesis89 (4%)25 (3%)64 (5%)#NA
      SAF, skin autofluorescence; BMI, body mass index; MRI, magnetic resonance imaging; eGFR, effective glomerular filtration rate.
      Depicted are mean ± standard deviation, median (interquartile range) and number (percentage).
      # orP < 0.05, ## or ∗∗P < 0.01 and ### or ∗∗∗P < 0.001. Significant differences between women with X-rays and MRI or men and women.
      Knee MRI's were only available for an all-female obese subgroup of the Rotterdam Study with SAF measurments (n = 103).
      Knee OA was based on Knee Osteoarthritis Scoring System (KOSS) on MRI tibiofemoral joint.

      SAF and KL sum score knee and/or hip

      First, we observed that higher KL sum score for hip + knee was associated with the higher values of SAF in all subjects [RM = 1.08 (0.97–1.20)] using ratio of means (RM) from the coefficients derived from the count model part of ZINB regression (Table II) with a P-value for interaction of 0.09 for SAF∗sex. Therefore, we performed the rest of the analyses stratified in men and women (Supplementary Fig. 2 shows the distribution of SAF in men and women). One unit increase in SAF significantly increased mean hip + knee KL sum score with 15% [RM = 1.15 (1.00–1.33)] in women but no such increase was observed in men [RM = 1.00 (0.86–1.16)] in our adjusted models. This association between SAF and hip + knee KL sum score attenuated also in women [PR = 0.96 (0.83–1.11)] when participants with a hip or knee prosthesis were excluded. When we stratified the analysis according to knee and hip separately, we observed that for every unit increase in SAF, mean hip KL sum score increased by 31% [RM = 1.31 (0.97–1.77)] and mean knee KL sum score increased by 17% [RM = 1.17 (1.00–1.36)] in women. When participants with either a hip or knee prosthesis were excluded, the association between SAF and hip [RM = 0.90 (0.67–1.19)] or knee [PR = 1.02 (0.88–1.17)] KL sum score in women disappeared and become inconclusive. In men, no significant associations were found for hip or knee KL sum score (Table II). No coefficients from the zero-inflation model part of our ZINB model showed any relationship between increasing values of SAF and excess zero's for KL sum score i.e., no effect of excess zeroes on its relationship with SAF (Supplementary Table 4).
      Table IIZero-inflated negative binomial regression reporting count model coefficients for mean KL sum score hip and/or knee change with one unit increase in SAF values adjusting for potential confounders in participants from the Rotterdam study
      KL sum scoresAllMenWomen
      Estimate (95% confidence interval)P-valueEstimate (95% confidence interval)P-valueEstimate (95% confidence interval)P-value
      Including all participants n = 2,153
      HIP and KNEE1.08 (0.97–1.20)0.181.00 (0.86–1.16)0.991.15 (1.00–1.33)0.05
      HIP ALONE1.15 (0.96–1.38)0.131.12 (0.86–1.46)0.391.31 (0.97–1.77)0.08
      KNEE ALONE1.08 (0.95–1.22)0.250.89 (0.71–1.13)0.351.17 (1.00–1.36)0.04
      Excluding participants with prosthesis n = 1,948
      HIP and KNEE0.97 (0.87–1.08)0.600.93 (0.79–1.09)0.380.96 (0.83–1.11)0.60
      HIP ALONE0.93 (0.77–1.12)0.450.93 (0.73–1.19)0.560.90 (0.67–1.19)0.46
      KNEE ALONE0.98 (0.87–1.11)0.790.86 (0.69–1.07)0.181.02 (0.88–1.17)0.83
      Analyses were sex-stratified as P-value for interaction for SAF∗sex = 0.09 for HIP and KNEE, 0.03 for HIP ALONE and 0.009 for KNEE ALONE.
      All analyses were adjusted for age, sex, RS-cohorts, eGFR, BMI, diabetes and smoking status.
      Note: Results for all zero-inflation model coefficients from ZINB did not show any relationship between increasing values of SAF and excess zeroes for KL sum score and are presented elsewhere (Supplementary Table 4).

      SAF and prevalent knee and/or hip prosthesis

      Table III shows a comparison of participants without OA to those with ROA and those with a prosthesis (THP or TKP) with increasing values of SAF. In women, SAF was associated with higher prevalence of having a knee or hip prosthesis [Odds ratio, OR = 1.67 (1.10–2.54)] but not with presence of ROA [OR = 0.83 (0.61–1.14)] compared to those without OA. We repeated the same analysis separately for knee and hip joints. SAF showed a similar trend towards higher prevalence of having a hip prosthesis [OR = 1.39 (0.84–2.31)] and knee prosthesis [OR = 2.09 (1.19–2.67)] but not with presence of hip ROA [OR = 0.88 (0.54–1.42)] and knee ROA [OR = 0.92 (0.66–1.27)] in women. In men, there was no difference in the prevalence of ROA and hip and/or knee prosthesis with increasing values of SAF. Additional adjustment for physical activity and excluding T2D participants did not change the results (data not shown/Supplementary Table 5).
      Table IIIOdds ratio of prevalent knee and/or hip prosthesis by an increase of one unit in SAF values adjusting for potential confounders in participants from the Rotterdam study
      Ternary outcome using participants with no ROA or prosthesis as referenceNMen (n = 947)Women (n = 1,206)
      Estimate (95% confidence interval)P-valueEstimate (95% confidence interval)P-value
      HIP and KNEENo ROA1,438ReferenceReference
      ROA5300.87 (0.62–1.22)0.420.83 (0.61–1.14)0.24
      THP + TKP1850.96 (0.53–1.71)0.881.67 (1.10–2.54)0.02
      HIP ALONENo ROA1,850ReferenceReference
      ROA1910.97 (0.61–1.53)0.900.88 (0.54–1.42)0.59
      THP1121.28 (0.65–2.52)0.481.39 (0.84–2.31)0.20
      KNEE ALONENo ROA1,648ReferenceReference
      ROA4160.85 (0.57–1.26)0.400.92 (0.66–1.27)0.59
      TKP890.81 (0.33–2.06)0.652.09 (1.19–2.67)0.01
      ROA, radiographic osteoarthritis; THP, total hip prosthesis; TKP, total knee prosthesis.

      Cohort profile based on availability of SAF and MRI knee

      If AGEs accumulation is causally related to increased risk of OA, a possible mechanism might be through crosslinking of AGEs between the collagen molecules in the extracellular matrix leading to stiffness of cartilage making it prone to degeneration
      • Krishnamoorthy D.
      • Hoy R.C.
      • Natelson D.M.
      • Torre O.M.
      • Laudier D.M.
      • Iatridis J.C.
      • et al.
      Dietary advanced glycation end-product consumption leads to mechanical stiffening of murine intervertebral discs.
      ,
      • Verzijl N.
      • DeGroot J.
      • Ben Z.C.
      • Brau-Benjamin O.
      • Maroudas A.
      • Bank R.A.
      • et al.
      Crosslinking by advanced glycation end products increases the stiffness of the collagen network in human articular cartilage: a possible mechanism through which age is a risk factor for osteoarthritis.
      . Table I shows also the characteristic of 103 women with SAF and available knee MRI data. Subgroup of women with knee MRI were younger (62.8 ± 2.48 vs 74.5 ± 7.7 years), had lower SAF (2.17 ± 0.40 vs 2.31 ± 0.46 A.U.) and fewer participants with T2DM (5% vs 12%) as compared to women in total population (Supplementary Tables 6 and 7).
      We observed that one unit increase in SAF was associated with higher prevalence of full thickness cartilage loss [OR = 5.44 (1.27–23.38)], but not with an increased prevalence of partial thickness cartilage loss [OR = 1.45 (0.38–5.54)] (Table IV). We also observed a trend towards higher prevalence of definite osteophytes [OR = 2.76 (0.92–8.23)] with each unit increase in SAF and a trend towards higher prevalence of TFOAMRI [OR = 2.12 (0.70–6.43)], but the associations were not significant, probably due to modest power.
      Table IVOdds ratio of presence of individual MRI defects and tibio-femoral knee OA (TFOAMRI) by an increase of one unit in SAF values adjusting for potential confounders in female participants from the Rotterdam study (n = 103)
      n = 103 with MRI knee data on TF jointOdds ratio (95% Confidence interval)P-value
      No cartilage loss (n = 65 or 55.5%)Reference
      Partial thickness cartilage loss (n = 22 or 29%)1.45 (0.38–5.54)0.58
      Full thickness cartilage loss (n = 16 or 15.5%)5.44 (1.27–23.38)0.02
      Other MRI abnormalities
      Definite osteophytes (n = 24 or 32%)2.76 (0.92–8.23)0.07
      Tibio-femoral knee OA (TFOAMRI)2.12 (0.70–6.43)0.18
      All analyses were adjusted for age and BMI.

      Discussion

      In this paper, we investigated the relationship of knee and hip X-ray based KL sum scores and knee MRI cartilage loss with SAF. High SAF in women was associated with higher prevalence of knee and hip prosthesis and with a trend towards higher prevalence of cartilage loss on MRI. High SAF was not associated with increasing severity and prevalence of knee and/or hip ROA in either sex when participants with prosthesis were excluded.
      We observed an association between higher skin AGEs values and higher prevalence of knee or hip joint prosthesis, which is most likely equivalent to severe OA, in women only but we observed inconclusive results for SAF and severity or presence of ROA. In line with these findings, PEN, a prototype AGE, in skin biopsy of 300 human participants failed to predict progression of knee and hip OA over 5 years
      • Vos P.A.
      • Welsing P.M.
      • deGroot J.
      • Huisman A.M.
      • Oostveen J.C.
      • Reijman M.
      • et al.
      Skin pentosidine in very early hip/knee osteoarthritis (CHECK) is not a strong independent predictor of radiographic progression over 5 years follow-up.
      . Cartilage AGEs in living participants have been assessed only in those undergoing surgery due to OA (equivalent to severe OA) in the absence of controls without OA or early stage OA which prevents us from understanding the course of AGEs accumulation in early stages of OA
      • Trellu S.
      • Courties A.
      • Jaisson S.
      • Gorisse L.
      • Gillery P.
      • Kerdine-Römer S.
      • et al.
      Impairment of glyoxalase-1, an advanced glycation end-product detoxifying enzyme, induced by inflammation in age-related osteoarthritis.
      . Pentosidine, an AGE also measured through SAF, has been shown to be present nearly 5 times higher in cartilage than in skin while there was a modest correlation between cartilage and skin PEN (R = 0.473) within paired macroscopically normal samples from cadavers
      • Vos P.A.
      • DeGroot J.
      • Huisman A.M.
      • Oostveen J.C.
      • Marijnissen A.C.
      • Bijlsma J.W.
      • et al.
      Skin and urine pentosidine weakly correlate with joint damage in a cohort of patients with early signs of osteoarthritis (CHECK).
      . This higher accumulation of AGEs in cartilage could be due to extremely low turnover rate of cartilage collagen
      • Verzijl N.
      • DeGroot J.
      • Thorpe S.R.
      • Bank R.A.
      • Shaw J.N.
      • Lyons T.J.
      • et al.
      Effect of collagen turnover on the accumulation of advanced glycation end products.
      . This may partially explain the absence of relationship between SAF and OA severity as the rate of increase in cartilage AGEs might not be in proportion to increase in skin AGEs. However, our results suggest a relationship between SAF and risk of prosthesis although the possibility of reverse causation remains whereby a vicious cycle may start when chronic low grade inflammation in severe OA (prosthesis) leads to increase in AGEs formation and vice versa. Besides, low pain threshold is perhaps a risk factor for undergoing joint replacement but not for OA and whether accumulation of AGEs might be related to pain or other factors specific to OA that increase risk of prosthesis is unknown. Therefore, the relationship between SAF and presence of prosthesis could be confounded by the indication of joint replacement which has not been taken into account. Future studies measuring skin AGEs in early and late stage OA in a longitudinal fashion are needed to explore this further.
      In a small subset of 103 women with available knee MRI, increasing SAF values were associated with a higher prevalence of cartilage loss although the confidence interval was very wide pointing towards an imprecise estimate. Increasing skin AGEs have been reported to be associated with greater JSN over 4 years on X-rays in men but not in women who were both at high risk of developing knee OA
      • Eaton C.B.
      • Sayeed M.
      • Ameernaz S.
      • Roberts M.B.
      • Maynard J.D.
      • Driban J.B.
      • et al.
      Sex differences in the association of skin advanced glycation endproducts with knee osteoarthritis progression.
      . In contrast to these and our findings, no relationship was found between urinary PEN and any degree of cartilage loss on MRI knee during 30 months follow-up in 127 symptomatic OA participants at baseline
      • Hunter D.J.
      • Lavalley M.
      • Li J.
      • Zhang Y.
      • Bauer D.
      • Nevitt M.
      • et al.
      Urinary pentosidine does not predict cartilage loss among subjects with symptomatic knee OA: the BOKS Study.
      . It should be noted that urine AGEs represent a transient measurement which can fluctuate depending on diet, renal function or tissue turnover rate. In our study, increasing SAF values also showed a trend towards higher prevalence of definite osteophytes (moderate and severe) and TFOAMRI. Whether this relationship between SAF and osteophytes on MRI is a consequence of overt cartilage loss needs to be further analyzed in a well powered population to study the relationship of skin AGEs with cartilage loss and osteophytes on MRI.
      In our study, we found an association of increasing SAF with higher prevalence of hip or knee prosthesis in women but not in men and a trend towards increasing cartilage loss in a small subset of women with knee MRI. Women have been known to have a higher OA incidence especially after menopause
      • Srikanth V.K.
      • Fryer J.L.
      • Zhai G.
      • Winzenberg T.M.
      • Hosmer D.
      • Jones G.
      A meta-analysis of sex differences prevalence, incidence and severity of osteoarthritis.
      but sex differences in MRI based biomarkers and joint metabolism have rarely been studied and with varying results
      • Maleki-Fischbach M.
      • Jordan J.M.
      New developments in osteoarthritis. Sex differences in magnetic resonance imaging-based biomarkers and in those of joint metabolism.
      ,
      • Karsdal M.A.
      • Byrjalsen I.
      • Bay-Jensen A.C.
      • Henriksen K.
      • Riis B.J.
      • Christiansen C.
      Biochemical markers identify influences on bone and cartilage degradation in osteoarthritis--the effect of sex, Kellgren-Lawrence (KL) score, body mass index (BMI), oral salmon calcitonin (sCT) treatment and diurnal variation.
      . Absence of any association with SAF in men could partly be due to low power as they had significantly lower prevalence of OA and prostheses than women (Table I). Also, we had no knee MRI data on men to study cartilage loss. Our findings, however, may suggest a potential sex specific effect of AGEs on OA especially in women after menopause. This might be due to a state of chronic low-grade inflammation after menopause
      • Abu-Taha M.
      • Rius C.
      • Hermenegildo C.
      • Noguera I.
      • Cerda-Nicolas J.M.
      • Issekutz A.C.
      • et al.
      Menopause and ovariectomy cause a low grade of systemic inflammation that may be prevented by chronic treatment with low doses of estrogen or losartan.
      ,
      • Friedenreich C.M.
      • O'Reilly R.
      • Shaw E.
      • Stanczyk F.Z.
      • Yasui Y.
      • Brenner D.R.
      • et al.
      Inflammatory marker changes in postmenopausal women after a year-long exercise intervention comparing high versus moderate volumes.
      which has potential for accelerated AGEs accumulation, however, a concrete mechanism still needs to be elucidated and our findings need to be replicated.
      There are various mechanisms through which AGEs might deteriorate articular cartilage and increase predisposition to OA. In vitro studies showed both intra- and extracellular AGEs accumulation leads to endoplasmic reticulum stress and apoptosis in chondrocytes
      • Yamabe S.
      • Hirose J.
      • Uehara Y.
      • Okada T.
      • Okamoto N.
      • Oka K.
      • et al.
      Intracellular accumulation of advanced glycation end products induces apoptosis via endoplasmic reticulum stress in chondrocytes.
      ,
      • Rasheed Z.
      • Haqqi T.M.
      Endoplasmic reticulum stress induces the expression of COX-2 through activation of eIF2α, p38-MAPK and NF-κB in advanced glycation end products stimulated human chondrocytes.
      . Immunostaining on cartilage sections illustrated increased expression of RAGE on chondrocytes from OA patients as compared to chondrocytes from young adults without OA
      • Loeser R.F.
      • Yammani R.R.
      • Carlson C.S.
      • Chen H.
      • Cole A.
      • Im H.J.
      • et al.
      Articular chondrocytes express the receptor for advanced glycation end products: potential role in osteoarthritis.
      . Furthermore, young RAGE-knockout mice were shown to be partly protected from developing OA after knee destabilization surgery than controls
      • Larkin D.J.
      • Kartchner J.Z.
      • Doxey A.S.
      • Hollis W.R.
      • Rees J.L.
      • Wilhelm S.K.
      • et al.
      Inflammatory markers associated with osteoarthritis after destabilization surgery in young mice with and without Receptor for Advanced Glycation End-products (RAGE).
      . AGEs lead to non-enzymatic cross-linking of collagen fibers in cartilage extracellular matrix and decreasing collagen degradability by the matrix metalloproteinases/cathepsin K either by masking the cleavage site or by preventing unfolding of collagen molecules due to an extra AGEs cross-link
      • Panwar P.
      • Butler G.S.
      • Jamroz A.
      • Azizi P.
      • Overall C.M.
      • Brömme D.
      Aging-associated modifications of collagen affect its degradation by matrix metalloproteinases.
      ,
      • Moshtagh P.R.
      • Korthagen N.M.
      • van Rijen M.H.P.
      • Castelein R.M.
      • Zadpoor A.A.
      • Weinans H.
      Effects of non-enzymatic glycation on the micro- and nano-mechanics of articular cartilage.
      . Both mechanisms could increase the risk of OA by stiffening cartilage and making it vulnerable to damage. In summary, there are several potential mechanisms by which AGE accumulation in articular cartilage could influence osteoarthritis development. Our data on knee MRI reinforces that the influence of AGEs on OA is primarily mediated through cartilage loss.
      Our study has several strengths. It is a comprehensive study of relationship of tissue AGEs and several OA outcomes based on X-ray and MRI images in the two main joints affected by OA, the knees and hips. It is embedded in a large RS cohort study of the general population which allowed for adjustment for several potential confounders. Our study has also several limitations. As mentioned, AGE levels in skin may not adequately reflect AGEs in OA cartilage although human cadaver studies have shown that cartilage and skin PEN were correlated (R = 0.473) in macroscopically normal cartilage
      • Verzijl N.
      • DeGroot J.
      • Bank R.A.
      • Bayliss M.T.
      • Bijlsma J.W.J.
      • Lafeber F.P.J.G.
      • et al.
      Age-related accumulation of the advanced glycation endproduct pentosidine in human articular cartilage aggrecan: the use of pentosidine levels as a quantitative measure of protein turnover.
      . Although a cluster of AGE moieties are known, the AGE reader™ detects only fluorescent AGEs as SAF. Residual confounding could not be excluded despite adjusting for currently know most important risk factors for AGEs accumulation. In our study, broad confidence intervals of odds ratio between radiographic osteoarthritis and SAF in men (but also in women) were observed. This indicates that we still have uncertain knowledge about the true effect and that more information is needed. Cartilage thickness on MRI data and its relationship to SAF should be interpreted with caution due to small sample size and sparse prevalence of outcome.
      In conclusion, skin AGEs showed an association with end stage OA, represented by the presence of a prosthesis, and a trend towards an association with full thickness cartilage loss on knee MRI in a small subset of women. Based on our findings, skin AGEs do not appear to have potential to be used as a marker of severity of ROA in hip and knee joints. This may be related to a different half-life of AGEs in cartilage than skin and a moderate correlation between the two. Longitudinal data in well-powered cohorts on skin AGEs and cartilage thickness are needed to confirm and expand our preliminary findings.

      Data availability

      Restrictions apply to the availability of some or all data generated or analyzed during this study to preserve patient confidentiality or because they were used under license. The corresponding author will on request detail the restrictions and any conditions under which access to some data may be provided. Some or all datasets generated during and/or analyzed during the current study are not publicly available but are available from the corresponding author on reasonable request.

      Disclaimers

      No.

      Authors' roles

      KW, IKS and MCZ designed the study. IKS, CJB, DS and JBM provided essential materials. KW assessed and IKS (statistically) analyzed the data. KW, IKS and MCZ interpreted the results. KW created the figures and tables. KW, IKS, CJB and MCZ drafted the manuscript. All authors provided contribution to intellectual content, acquisition and interpretation of data and critically revising the manuscript. All authors have read and revised the manuscript and approved the final submitted version.

      Conflict of interest

      All authors declare to have no disclosures and conflicts of interest relevant to the current study.

      Funding

      The Rotterdam Study is supported by Erasmus Medical Center and Erasmus University , Rotterdam, Netherlands Organization for the Health Research and Development (ZonMw) , the Research Institute for Diseases in the Elderly (RIDE) , the Netherlands Genomics Initiative , the Ministry of Education, Culture and Science , the Ministry for Health, Welfare and Sports , the European Commission (DG XII) , and the Municipality of Rotterdam . The Jaap Schouten Foundation established in Rotterdam, The Netherlands, provided funding for the analyses of Advanced Glycation End Products related to musculoskeletal health in the Rotterdam Study. The funding sources had no role in the study design, data collection, analysis, and interpretation, writing of the report, or decision to submit the article for publication.

      Acknowledgments

      We would like to thank all the participants of the Rotterdam Study for their contribution in this population-based study, research assistants (particularly Hannie van den Boogert for acquisition of the DXA scans), the general practitioners, hospitals, and pharmacies in Rotterdam.

      Appendix A. Supplementary data

      The following is the Supplementary data to this article:

      References

        • Robinson W.H.
        • Lepus C.M.
        • Wang Q.
        • Raghu H.
        • Mao R.
        • Lindstrom T.M.
        • et al.
        Low-grade inflammation as a key mediator of the pathogenesis of osteoarthritis.
        Nat Rev Rheumatol. 2016; 12: 580-592
        • Salminen A.
        Feed-forward regulation between cellular senescence and immunosuppression promotes the aging process and age-related diseases.
        Ageing Res Rev. 2021; 67101280
        • Maillard-Lefebvre H.
        • Boulanger E.
        • Daroux M.
        • Gaxatte C.
        • Hudson B.I.
        • Lambert M.
        Soluble receptor for advanced glycation end products: a new biomarker in diagnosis and prognosis of chronic inflammatory diseases.
        Rheumatology (Oxford). 2009; 48: 1190-1196
        • Singh R.
        • Barden A.
        • Mori T.
        • Beilin L.
        Advanced glycation end-products: a review.
        Diabetologia. 2001; 44: 129-146
        • DeGroot J.
        • Verzijl N.
        • Budde M.
        • Bijlsma J.W.
        • Lafeber F.P.
        • TeKoppele J.M.
        Accumulation of advanced glycation end products decreases collagen turnover by bovine chondrocytes.
        Exp Cell Res. 2001; 266: 303-310
        • DeGroot J.
        • Verzijl N.
        • Jacobs K.M.
        • Budde M.
        • Bank R.A.
        • Bijlsma J.W.
        • et al.
        Accumulation of advanced glycation endproducts reduces chondrocyte-mediated extracellular matrix turnover in human articular cartilage.
        Osteoarthritis Cartilage. 2001; 9: 720-726
        • DeGroot J.
        • Verzijl N.
        • Wenting-van Wijk M.J.
        • Jacobs K.M.
        • Van El B.
        • Van Roermund P.M.
        • et al.
        Accumulation of advanced glycation end products as a molecular mechanism for aging as a risk factor in osteoarthritis.
        Arthritis Rheum. 2004; 50: 1207-1215
        • Loeser R.F.
        • Yammani R.R.
        • Carlson C.S.
        • Chen H.
        • Cole A.
        • Im H.J.
        • et al.
        Articular chondrocytes express the receptor for advanced glycation end products: potential role in osteoarthritis.
        Arthritis Rheum. 2005; 52: 2376-2385
        • Senolt L.
        • Braun M.
        • Olejarova M.
        • Forejtova S.
        • Gatterova J.
        • Pavelka K.
        Increased pentosidine, an advanced glycation end product, in serum and synovial fluid from patients with knee osteoarthritis and its relation with cartilage oligomeric matrix protein.
        Ann Rheum Dis. 2005; 64: 886-890
        • Hunter D.J.
        • Lavalley M.
        • Li J.
        • Zhang Y.
        • Bauer D.
        • Nevitt M.
        • et al.
        Urinary pentosidine does not predict cartilage loss among subjects with symptomatic knee OA: the BOKS Study.
        Osteoarthritis Cartilage. 2007; 15: 93-97
        • Braun M.
        • Hulejova H.
        • Gatterova J.
        • Filkova M.
        • Pavelkova A.
        • Sleglova O.
        • et al.
        Pentosidine, an advanced glycation end-product, may reflect clinical and morphological features of hand osteoarthritis.
        Open Rheumatol J. 2012; 6: 64-69
        • Scheijen J.
        • Hanssen N.M.J.
        • van Greevenbroek M.M.
        • Van der Kallen C.J.
        • Feskens E.J.M.
        • Stehouwer C.D.A.
        • et al.
        Dietary intake of advanced glycation endproducts is associated with higher levels of advanced glycation endproducts in plasma and urine: the CODAM study.
        Clin Nutr. 2018; 37: 919-925
        • Rabbani N.
        • Thornalley P.J.
        Advanced glycation end products in the pathogenesis of chronic kidney disease.
        Kidney Int. 2018; 93: 803-813
        • Jiang J.
        • Zhang Y.
        • Chen J.
        • Yang X.
        • Mei C.
        • Xiong F.
        • et al.
        Serum and tissue levels of advanced glycation end products and risk of mortality in patients on maintenance hemodialysis.
        Am J Nephrol. 2021; 52: 8-16
        • Meerwaldt R.
        • Links T.
        • Graaff R.
        • Thorpe S.R.
        • Baynes J.W.
        • Hartog J.
        • et al.
        Simple noninvasive measurement of skin autofluorescence.
        Ann N Y Acad Sci. 2005; 1043: 290-298
        • Verzijl N.
        • DeGroot J.
        • Thorpe S.R.
        • Bank R.A.
        • Shaw J.N.
        • Lyons T.J.
        • et al.
        Effect of collagen turnover on the accumulation of advanced glycation end products.
        J Biol Chem. 2000; 275: 39027-39031
        • Ikram M.A.
        • Brusselle G.
        • Ghanbari M.
        • Goedegebure A.
        • Ikram M.K.
        • Kavousi M.
        • et al.
        Objectives, design and main findings until 2020 from the Rotterdam Study.
        Eur J Epidemiol. 2020; 35: 483-517https://doi.org/10.1007/s10654-020-00640-5
        • Waqas K.
        • Chen J.
        • Koromani F.
        • Trajanoska K.
        • van der Eerden B.C.
        • Uitterlinden A.G.
        • et al.
        Skin autofluorescence, a noninvasive biomarker for advanced glycation end-products, is associated with prevalent vertebral and major osteoporotic fractures: the Rotterdam Study.
        J Bone Miner Res. 2020; 35: 1904-1913
        • Beisswenger P.J.
        • Howell S.
        • Mackenzie T.
        • Corstjens H.
        • Muizzuddin N.
        • Matsui M.S.
        Two fluorescent wavelengths, 440(ex)/520(em) nm and 370(ex)/440(em) nm, reflect advanced glycation and oxidation end products in human skin without diabetes.
        Diabetes Technol Ther. 2012; 14: 285-292
        • Atzeni I.M.
        • van de Zande S.C.
        • Westra J.
        • Zwerver J.
        • Smit A.J.
        • Mulder D.J.
        The AGE Reader: a non-invasive method to assess long-term tissue damage.
        Methods. 2021; 203: 533-541https://doi.org/10.1016/j.ymeth.2021.02.016
        • Koetsier M.
        • Nur E.
        • Chunmao H.
        • Lutgers H.L.
        • Links T.P.
        • Smit A.J.
        • et al.
        Skin color independent assessment of aging using skin autofluorescence.
        Opt Express. 2010; 18: 14416-14429
        • Hoeven T.A.
        • Kavousi M.
        • Clockaerts S.
        • Kerkhof H.J.M.
        • van Meurs J.B.
        • Franco O.
        • et al.
        Association of atherosclerosis with presence and progression of osteoarthritis: the Rotterdam Study.
        Ann Rheum Dis. 2013; 72: 646-651
        • Kellgren J.H.
        • Lawrence J.S.
        Radiological assessment of osteo-arthrosis.
        Ann Rheum Dis. 1957; 16: 494-502
        • Schiphof D.
        • Boers M.
        • Bierma-Zeinstra S.M.
        Differences in descriptions of Kellgren and Lawrence grades of knee osteoarthritis.
        Ann Rheum Dis. 2008; 67: 1034-1036
        • Schiphof D.
        • Oei E.H.
        • Hofman A.
        • Waarsing J.H.
        • Weinans H.
        • Bierma-Zeinstra S.M.
        Sensitivity and associations with pain and body weight of an MRI definition of knee osteoarthritis compared with radiographic Kellgren and Lawrence criteria: a population-based study in middle-aged females.
        Osteoarthritis Cartilage. 2014; 22: 440-446
        • Hunter D.J.
        • Guermazi A.
        • Lo G.H.
        • Grainger A.J.
        • Conaghan P.G.
        • Boudreau R.M.
        • et al.
        Evolution of semi-quantitative whole joint assessment of knee OA: MOAKS (MRI Osteoarthritis Knee Score).
        Osteoarthritis Cartilage. 2011; 19: 990-1002
        • Runhaar J.
        • Schiphof D.
        • van Meer B.
        • Reijman M.
        • Bierma-Zeinstra S.M.
        • Oei E.H.
        How to define subregional osteoarthritis progression using semi-quantitative MRI osteoarthritis knee score (MOAKS).
        Osteoarthritis Cartilage. 2014; 22: 1533-1536
        • Hunter D.J.
        • Arden N.
        • Conaghan P.G.
        • Eckstein F.
        • Gold G.
        • Grainger A.
        • et al.
        Definition of osteoarthritis on MRI: results of a Delphi exercise.
        Osteoarthritis Cartilage. 2011; 19: 963-969
        • Hofman A.
        • Brusselle G.G.
        • Darwish Murad S.
        • van Duijn C.M.
        • Franco O.H.
        • Goedegebure A.
        • et al.
        The Rotterdam Study: 2016 objectives and design update.
        Eur J Epidemiol. 2015; 30: 661-708
        • den Engelsen C.
        • van den Donk M.
        • Gorter K.J.
        • Salomé P.L.
        • Rutten G.E.
        Advanced glycation end products measured by skin autofluorescence in a population with central obesity.
        Dermatoendocrinol. 2012; 4: 33-38
        • Mallipattu S.K.
        • Uribarri J.
        Advanced glycation end product accumulation: a new enemy to target in chronic kidney disease?.
        Curr Opin Nephrol Hypertens. 2014; 23: 547-554
        • Oleniuc M.
        • Secara I.
        • Onofriescu M.
        • Hogas S.
        • Voroneanu L.
        • Siriopol D.
        • et al.
        Consequences of advanced glycation end products accumulation in chronic kidney disease and clinical usefulness of their assessment using a non-invasive technique - skin autofluorescence.
        Maedica (Bucur). 2011; 6: 298-307
        • van Waateringe R.P.
        • Mook-Kanamori M.J.
        • Slagter S.N.
        • van der Klauw M.M.
        • van Vliet-Ostaptchouk J.V.
        • Graaff R.
        • et al.
        The association between various smoking behaviors, cotinine biomarkers and skin autofluorescence, a marker for advanced glycation end product accumulation.
        PLoS One. 2017; 12e0179330
        • Singh V.P.
        • Bali A.
        • Singh N.
        • Jaggi A.S.
        Advanced glycation end products and diabetic complications.
        Korean J Physiol Pharmacol. 2014; 18: 1-14
        • Krishnamoorthy D.
        • Hoy R.C.
        • Natelson D.M.
        • Torre O.M.
        • Laudier D.M.
        • Iatridis J.C.
        • et al.
        Dietary advanced glycation end-product consumption leads to mechanical stiffening of murine intervertebral discs.
        Dis Model Mech. 2018; 11
        • Verzijl N.
        • DeGroot J.
        • Ben Z.C.
        • Brau-Benjamin O.
        • Maroudas A.
        • Bank R.A.
        • et al.
        Crosslinking by advanced glycation end products increases the stiffness of the collagen network in human articular cartilage: a possible mechanism through which age is a risk factor for osteoarthritis.
        Arthritis Rheum. 2002; 46: 114-123
        • Vos P.A.
        • Welsing P.M.
        • deGroot J.
        • Huisman A.M.
        • Oostveen J.C.
        • Reijman M.
        • et al.
        Skin pentosidine in very early hip/knee osteoarthritis (CHECK) is not a strong independent predictor of radiographic progression over 5 years follow-up.
        Osteoarthritis Cartilage. 2013; 21: 823-830
        • Trellu S.
        • Courties A.
        • Jaisson S.
        • Gorisse L.
        • Gillery P.
        • Kerdine-Römer S.
        • et al.
        Impairment of glyoxalase-1, an advanced glycation end-product detoxifying enzyme, induced by inflammation in age-related osteoarthritis.
        Arthritis Res Ther. 2019; 21: 18
        • Vos P.A.
        • DeGroot J.
        • Huisman A.M.
        • Oostveen J.C.
        • Marijnissen A.C.
        • Bijlsma J.W.
        • et al.
        Skin and urine pentosidine weakly correlate with joint damage in a cohort of patients with early signs of osteoarthritis (CHECK).
        Osteoarthritis Cartilage. 2010; 18: 1329-1336
        • Eaton C.B.
        • Sayeed M.
        • Ameernaz S.
        • Roberts M.B.
        • Maynard J.D.
        • Driban J.B.
        • et al.
        Sex differences in the association of skin advanced glycation endproducts with knee osteoarthritis progression.
        Arthritis Res Ther. 2017; 19: 36
        • Srikanth V.K.
        • Fryer J.L.
        • Zhai G.
        • Winzenberg T.M.
        • Hosmer D.
        • Jones G.
        A meta-analysis of sex differences prevalence, incidence and severity of osteoarthritis.
        Osteoarthritis Cartilage. 2005; 13: 769-781
        • Maleki-Fischbach M.
        • Jordan J.M.
        New developments in osteoarthritis. Sex differences in magnetic resonance imaging-based biomarkers and in those of joint metabolism.
        Arthritis Res Ther. 2010; 12: 212
        • Karsdal M.A.
        • Byrjalsen I.
        • Bay-Jensen A.C.
        • Henriksen K.
        • Riis B.J.
        • Christiansen C.
        Biochemical markers identify influences on bone and cartilage degradation in osteoarthritis--the effect of sex, Kellgren-Lawrence (KL) score, body mass index (BMI), oral salmon calcitonin (sCT) treatment and diurnal variation.
        BMC Musculoskelet Disord. 2010; 11: 125
        • Abu-Taha M.
        • Rius C.
        • Hermenegildo C.
        • Noguera I.
        • Cerda-Nicolas J.M.
        • Issekutz A.C.
        • et al.
        Menopause and ovariectomy cause a low grade of systemic inflammation that may be prevented by chronic treatment with low doses of estrogen or losartan.
        J Immunol. 2009; 183: 1393-1402
        • Friedenreich C.M.
        • O'Reilly R.
        • Shaw E.
        • Stanczyk F.Z.
        • Yasui Y.
        • Brenner D.R.
        • et al.
        Inflammatory marker changes in postmenopausal women after a year-long exercise intervention comparing high versus moderate volumes.
        Cancer Prev Res (Phila). 2016; 9: 196-203
        • Yamabe S.
        • Hirose J.
        • Uehara Y.
        • Okada T.
        • Okamoto N.
        • Oka K.
        • et al.
        Intracellular accumulation of advanced glycation end products induces apoptosis via endoplasmic reticulum stress in chondrocytes.
        FEBS J. 2013; 280: 1617-1629
        • Rasheed Z.
        • Haqqi T.M.
        Endoplasmic reticulum stress induces the expression of COX-2 through activation of eIF2α, p38-MAPK and NF-κB in advanced glycation end products stimulated human chondrocytes.
        Biochim Biophys Acta Mol Cell Res. 2012; 1823: 2179-2189
        • Larkin D.J.
        • Kartchner J.Z.
        • Doxey A.S.
        • Hollis W.R.
        • Rees J.L.
        • Wilhelm S.K.
        • et al.
        Inflammatory markers associated with osteoarthritis after destabilization surgery in young mice with and without Receptor for Advanced Glycation End-products (RAGE).
        Front Physiol. 2013 May; 4 (Larkin D.J.; Kartchner J.Z.; Doxey A.S.; Hollis W.R.; Rees J.L.; Wilhelm S.K.; Draper C.S.; Peterson D.M.; Jackson G.G.; Ingersoll C.; Haynie S.S.; Chavez E.; Reynolds P.R.; Kooyman D.L., [email protected]) Department of Physiology and Developmental Biology, Brigham Young University, Provo, UT, United States): 121https://doi.org/10.3389/fphys.2013.00121
        • Panwar P.
        • Butler G.S.
        • Jamroz A.
        • Azizi P.
        • Overall C.M.
        • Brömme D.
        Aging-associated modifications of collagen affect its degradation by matrix metalloproteinases.
        Matrix Biol. 2018; 65: 30-44
        • Moshtagh P.R.
        • Korthagen N.M.
        • van Rijen M.H.P.
        • Castelein R.M.
        • Zadpoor A.A.
        • Weinans H.
        Effects of non-enzymatic glycation on the micro- and nano-mechanics of articular cartilage.
        J Mech Behav Biomed Mater. 2018; 77: 551-556
        • Verzijl N.
        • DeGroot J.
        • Bank R.A.
        • Bayliss M.T.
        • Bijlsma J.W.J.
        • Lafeber F.P.J.G.
        • et al.
        Age-related accumulation of the advanced glycation endproduct pentosidine in human articular cartilage aggrecan: the use of pentosidine levels as a quantitative measure of protein turnover.
        Matrix Biol. 2001; 20: 409-417