The official journal of the Japan Atherosclerosis Society and the Asian Pacific Society of Atherosclerosis and Vascular Diseases Original Article J Atheroscler Thromb, 2019; 26: 879-889. http://doi.org/10.5551/jat.47498 Skin Autofluorescence Measurement in Subclinical Atheromatous Disease: Results from the ILERVAS Project Enric Sánchez 1, Àngels Betriu 2, Andree Yeramian 1, Elvira Fernández 2, Francesc Purroy 3, Manuel Sánchez-de-la-Torre 4, 5, Reinald Pamplona 6, Eva Miquel 7, Mohsen Kerkeni 8, Cristina Hernández 9, 10, Rafael Simó 9, 10, and Albert Lecube 1, 10 on behalf of the ILERVAS project ILERVAS Project: Marta Hernández 1, Ferran Rius 1, Dinora Polanco 4, Ferran Barbé 4, 5, Gerard Torres 4, 5, Guillermo Suárez 4, 5, Manuel Portero-Otin 6, Mariona Jové 6, Laura Colàs-Campàs 3, Ikram Benabdelhak 3, Cristina Farràs 7, Marta Ortega 7, José Manuel Valdivielso 2, Marcelino Bermúdez-López 2, Montse Martínez-Alonso 2 Enric Sánchez and Àngels Betriu contributed equally to this work. Albert Lecube and Rafael Simó are joint senior authors. 1 Endocrinology and Nutrition Department. University Hospital Arnau de Vilanova. Obesity, Diabetes and Metabolism (ODIM) research group. IRBLleida. University of Lleida. Lleida, Catalonia, Spain. 2 Unit for the Detection and Treatment of Atherothrombotic Diseases (UDETMA V&R). University Hospital Arnau de Vilanova. Vascular and Renal Translational Research Group. IRBLleida. University of Lleida. Lleida, Catalonia, Spain. 3 Stroke Unit. University Hospital Arnau de Vilanova. Clinical Neurosciences Group. IRBLleida. University of Lleida. Lleida, Catalonia, Spain. 4 Respiratory Department. University Hospital Arnau de Vilanova-Santa María. Translational Research in Respiratory Medicine. IRBLleida. University of Lleida. Lleida, Catalonia, Spain. 5 Centro de Investigación Biomédica en Red de Enfermedades Respiratorias (CIBERES), Instituto de Salud Carlos III (ISCIII), Madrid, Spain. 6 Department of Experimental Medicine. IRBLleida. University of Lleida. Lleida, Catalonia, Spain. 7 Borges Blanques Primary Health Care Unit. Lleida, Catalonia, Spain. 8 Laboratory of Biochemistry, LR12ES05, Faculty of Medicine, University of Monastir. Monastir, Tunisia. 9 Endocrinology and Nutrition Department, University Hospital Vall d’Hebron. Diabetes and Metabolism Research Unit, Vall d’Hebron Institut de Recerca (VHIR). Universitat Autònoma de Barcelona. Barcelona, Catalonia, Spain. 10 Centro de Investigación Biomédica en Red de Diabetes y Enfermedades Metabólicas Asociadas (CIBERDEM), Instituto de Salud Carlos III (ISCIII), Madrid, Spain. Aim: Advanced glycation end-products (AGEs) have been involved in the atherogenic process in the high-risk population. The goal of this study was to demonstrate that AGEs are related to subclinical atheromatous disease in subjects with low to moderate vascular risk. Methods: A cross-sectional study in which 2,568 non-diabetic subjects of both sexes without cardiovascular dis- ease were included. Subcutaneous content of AGEs was assessed by skin autofluorescence (SAF) and subclinical atheromatous disease was measured by assessing the atheromatous plaque burden in carotid and femoral regions using ultrasonography. In addition, serum pentosidine, carboxymethyl-lysine (CML) and AGE receptors (RAGE) were assessed in a nested case-control study with 41 subjects without plaque and 41 individuals subjects with generalized disease. Results: Patients with atheromatous plaque had a higher SAF than those with no plaque (1.9 [1.7 to 2.3] vs. 1.8 [1.6 to 2.1] arbitrary units (AU), p ＜ 0.001). The SAF correlated with the total number of affected regions (r = 0.171, p ＜ 0.001), increasing progressively from 1.8 [1.6 to 2.1] AU in those without atheromatous disease to 2.3 [1.9 to 2.7] AU in patients with ≥ 8 plaques (p ＜ 0.001). A correlation was also observed between SAF and the total plaque area (r = 0.113, p ＜ 0.001). The area under the Receiver Operating Characteristic curve was 0.65 (0.61 to 0.68) for identifying male subjects with atheromatous disease. The multivariable logistic regression model showed a significant and independent association between SAF and the presence of atheromatous disease. 879 Sánchez et al . However, no significant differences in serum pentosidine, CML, and RAGE were observed. Conclusions: Increased subcutaneous content of AGEs is associated with augmented atheromatous plaque bur- den. Our results suggest that SAF may provide clinically relevant information to the current strategies for the evaluation of cardiovascular risk, especially among the male population. Key words: Advanced glycation end-products, Atheromatous plaque burden, Cardiovascular risk, Skin autofluorescence AGEs and vascular disease has been previously evalu- Introduction ated with opposing results 11-13). Therefore, we also The natural progression of atherosclerosis designed a nested case-control study addressed to eval- involves a long silent period, and the advanced stages uate the serum concentration of pentosidine, carboxy- of the disease are frequently not detected until cardio- methyl-lysine (CML), and AGEs receptors (RAGE) in vascular ischemic events occur 1). For this reason, the 41 subjects without plaque and 41 patients with gen- assessment of classical determinants of atherosclerosis, eralized disease. such as high blood pressure, atherogenic dyslipidemia, smoking habit, and type 2 diabetes has well-recog- nized limitations, particularly in lower-risk groups Methods such as young people and women 2, 3). In addition, the Design of the Study and Description of the Study identification of subclinical atherosclerosis may help Population physicians to identify vulnerable subjects at increased Cross-Sectional Study risk of cardiovascular disease (CVD) and develop A total of 2,568 Caucasian subjects were enrolled strategies to arrest disease progress 4, 5). between July 2015 and December 2017 from primary The advanced glycation end-products (AGEs) health centers in an ongoing prospective study aimed constitute a complex group of compounds formed by at evaluating the presence of asymptomatic CVD in the slow non-enzymatic glycation of proteins, lipids, the province of Lleida, Spain (ILERVAS Project; Clin- and nucleic acids, of which about 20 have been identi- Trials.gov Identifier: NCT03228459) 10). fied to date 6). The AGEs have been related to the The following inclusion criteria were used: men appearance and progression of vascular changes such and women aged between 45 and 70 years old, with- as increased permeability and thickening and stiffness out any history of vascular disease (ischemic heart dis- of the arterial wall, resulting in the development of ease, stroke, or peripheral arteriopathy), and at least atheromatous plaques decreasing luminal space 7-9). one traditional cardiovascular risk factor (hyperten- However, whether or not the magnitude of AGEs sion, dyslipidemia, a body mass index (BMI) ≥ 30 kg/ deposition in vascular walls is related to the main m 2, smoking, or a first-degree family member with characteristics of atheromatous disease, such as the premature cardiovascular disease). Clinical informa- number of affected regions, plaque location, and the tion was obtained from an electronic database (Infor- total area of the atheromatous plaques remain to be mation System for the Development of Research in elucidated. To shed light on this issue, we performed a Primary Care, SIDIAP) that contains anonymized, cross-sectional study using the non-invasive assess- longitudinal data from the 31 primary care centers ment of tissue accumulation of AGEs through skin pertaining to the Catalan Health Institute in the prov- autofluorescence (SAF) in 2,568 subjects without a ince of Lleida, Spain 14). The exclusion criteria com- previous history of vascular disease who had under- prised of the presence of chronic kidney disease, active gone bilateral carotid and femoral ultrasound explora- neoplasia, a life expectancy of fewer than 18 months, tion in the setting of the ILERVAS project 10). darker skin color (Fitzpatrick classes ＞ 5), and preg- In addition, the association between circulating nant women. Subjects with type 1 and type 2 diabe- Address for correspondence: Albert Lecube, Endocrinology and Nutrition Department, University Hospital Arnau de Vilanova Obesity, Diabetes and Metabolism (ODIM) research group, IRBLleida Avda. Rovira Roure 80. 25198-Lleida, Spain E-mail: alecube@gmail.com Received: October 9, 2018 Accepted for publication: January 10, 2019 Copyright©2019 Japan Atherosclerosis Society This article is distributed under the terms of the latest version of CC BY-NC-SA defined by the Creative Commons Attribution License. 880 Skin Autofluorescence and Atheromatosis tes, major risk factors for cardiovascular disease were region), intermediate (2 to 3 regions), or generalized also excluded 15). (4 to 12 regions) atheromatous disease. The carotid For all subjects, data were available regarding plaque area was measured, and the total plaque area both SAF and vascular carotid and femoral ultrasound (cm 2) was defined as the sum of the area of each examinations. Anthropometrical data were obtained plaque within the same subject 19, 20). by standardized protocols. Smoking status (never/cur- rent/ex-smoker) was recorded. Smokers who stopped Measurement of Skin Autofluorescence smoking ≥ 1 year prior to recruitment were considered SAF was measured using the AGE Reader™ ex-smokers. device (DiagnOptics Technologies, Groningen, The The prescribed antihypertensive and lipid-lower- Netherlands); a fully automated non-invasive tool that ing treatments were extracted from prescription- and measures AGEs deposition in the forearm using an pharmacy-invoicing databases provided by the Cat- Ultraviolet-A spectrum 21). The mean value of three Salut (Catalan Health Service), which are incorporated readings (arbitrary units: AU) was recorded. Skin areas yearly into the SIDIAP database. Antihypertensive that were tattooed or colored with cosmetics, were medication agents included angiotensin-converting heavily freckled, or had vessels near to the surface of enzyme inhibitors, diuretics, aldosterone receptor the skin were avoided. A single device, maintained and antagonist type II, beta-blockers, calcium antagonists, calibrated by the manufacturer following their recom- and other antihypertensives. Lipid-lowering drugs mendations, was used for all measurements. Repeated included statins, fibrates, ezetimibe, and omega-3 fatty SAF measurements with the AGE Reader™ device acids. In addition, the use of anticoagulant or anti- taken over a single day showed an overall Altman error platelet agents was also recorded. percentage of 5.03%, and intra-individual seasonal variance showed an Altman error percentage of Nested Case-Control Study 5.87% 21). Serum concentrations of pentosidine, CML, and RAGE were assessed in a subgroup of 41 subjects with Evaluation of Lipid Profile and Glycated Hemoglobin generalized atheromatous subclinical disease (≥ 4 Total cholesterol (mg/dl) concentrations were affected territories with plaque) matched (1:1) to 41 obtained in the entire population included in the controls without plaque carefully matched by gender, study from a non-fasting dried capillary blood testing age (±4 years), the BMI (±3 kg/m 2) and total choles- (fingertip puncture) using the Reflotrons ® Plus system terol (±30 mg/dL). The inclusion and exclusion crite- (Roche). Determination of the complete lipid profile ria were the same as in the cross-sectional study. [high density lipoprotein (HDL) cholesterol, low den- sity lipoprotein (LDL) cholesterol, and triglycerides] Assessment of Atheromatous Plaque Burden was determined only in subjects in which total choles- Bilateral carotid (common artery, bifurcation, terol was ≥ 200 mg/dL and fasted for 6 hours or total internal, and external) and femoral (common and cholesterol ≥ 250 mg/dL regardless of fasting hours. superficial) areas were explored. The images were As the presence of type 2 diabetes was an exclusion obtained by trained sonographers using an ultrasonic criterion in the ILERVAS project, data of fasting Doppler Ultrasound Vivid-I (General Electrics blood glucose is not available in our study. However, Healthcare, Waukesha, WI) equipped with a linear the glycated hemoglobin (HbA1c) test was performed transducer broadband linear 12L-RS that operates at to all participants using a point-of-care device [Cobas frequencies between 5–13 MHz. Standardized and B 101 ® (Roche Diagnostics) system]. validated scanning and reading protocols were used to decrease inter-operator variability and type 2 errors 16). Laboratory Assessment of Serum AGEs To evaluate intraobserver reliability, a total of 20 indi- In addition to previous determinations, a venous viduals were measured three to five times on diverse blood sample was collected from the antecubital vein, days. A k-coefficient of one was obtained for plaque separated by centrifugation (2.000 g at 4 ℃ for 20 assessment, demonstrating excellent intraobserver reli- min) and frozen at −80 ℃ for batched storage and ability. The readers were unaware of the patients’ clini- analysis. Samples were obtained with support by cal histories. IRBLleida Biobank (B.0000682) from Plataforma Subclinical atheromatosis was defined as the Biobancos (PT13/0010/0014). Serum concentrations presence of at least one plaque in any of the 12 of pentosidine, CML, and the soluble form of RAGE assessed areas 17). A plaque was defined as a focal were determined in duplicate by an enzyme-linked intima-media thickness ≥ 1.5 mm protruding in the immunosorbent assay (ELISA) kit provided by lumen 18). Participants were classified as having focal (1 MyBioSource (catalog number MBS036744, 881 Sánchez et al . MBS721957 and MBS724049, respectively; MyBio- Source, Inc. San Diego, CA, USA), according to the Results manufacturer’s instructions. Briefly, samples were The prevalence of subclinical atheromatous dis- added to a microtiter plate that has been previously ease in the study population was 70.4%: 21.8% pre-coated with a specific antibody against each restricted to carotid regions, 33.2% restricted to femo- human AGE (RAGE, CML, or pentosidine). After ral regions, and 45.0% with plaques in both regions. incubation and several washing steps, plates were incu- The main clinical characteristics and metabolic data bated with a secondary antibody conjugated to horse- according to the presence of plaque are shown in radish peroxidase and OD450 nm was determined Table 1. Asymptomatic atherosclerosis disease was using an ELISA plate reader (Source Bioscience, Not- higher in active smokers and in men, and increased tingham, UK). Intra-assay and inter-assay precision with age, blood pressure levels, and LDL cholesterol coefficient of variation ＜ 15% for all analytes. concentration. In addition, patients with subclinical atheromatous disease showed a higher SAF than sub- Statistical Analysis jects without disease (1.9 [1.7 to 2.3] vs. 1.8 [1.6 to The normal distribution of the variables was 2.1] AU, p ＜ 0.001). SAF was also higher when both evaluated using the Shapiro–Wilk test. Given its carotid and femoral regions were affected (2.0 [1.7 to skewed distribution, quantitative data are expressed as 2.3] AU) than when the disease was restricted to the median [interquartile range]. Comparisons carotid (1.9 [1.7 to 2.2] AU, p = 0.008) or femoral (1.9 between groups were made using the Mann–Whitney [1.7 to 2.2] AU, p = 0.004) regions. Finally, SAF U test for quantitative variables, and the Pearson’s chi- increased progressively from patients without plaque squared for categorical variables. The relationship (1.8 [1.6 to 2.1] AU) to those with generalized (2.0 between continuous variables was assessed by the [1.8 to 2.4] AU) subclinical atherosclerosis (p ＜ 0.001) Spearman correlation test. (Fig. 1A). Subjects with ≥ 8 affected regions showed The accuracy of SAF as a measurement of inter- the higher SAF (2.3 [1.9 to 2.7] AU). In the univari- est in discriminating diseased subjects (patients with ≥ ate analysis, a positive correlation between SAF and 1 plaque) from cases without atheromatous disease the number of affected regions was observed was evaluated using a Receiver Operating Characteris- (r = 0.171, p ＜ 0.001). tic (ROC) curve analysis with a complete sensitivity/ The total plaque area was 0.51 [0.21 to 1.07] specificity report and calculating Youden J statistic. cm 2, ranging from 0.13 [0.09 to 0.23] in patients A multivariable logistic regression model with with only one plaque to 1.41 [0.90 to 2.12] cm 2 when the enter method was used to explore the variables the disease was generalized (Fig. 1B). A positive signif- that were independently associated with the presence icant correlation was also observed between SAF and of atheromatous plaque. The independent variables the total plaque area (r = 0.113, p ＜ 0.001). included in the analysis were SAF and well-established The ROC analysis revealed that the best cut-off cardiovascular risk factors such as gender, age, smok- point for SAF (combining sensitivity plus specificity) ing status, systolic blood pressure, lipid profile (total for identifying patients with a subclinical atheroma- cholesterol, LDL cholesterol, HDL cholesterol, and tous disease was 2.05 AU in the entire population. At triglycerides), together with HbA1c and antihyperten- this point, the area under the ROC was 0.59 (0.57 to sive or lipid-lowering treatment. The calibration and 0.61). This value increased to 0.65 (0.61 to 0.68) with discrimination of the logistic model were evaluated a sensitivity of 55.6% and a specificity of 66.9% when using the test of fit Hosmer–Lemeshow and the area the ROC analysis was restricted to men, and the cut- under the curve, respectively. All “p” values were based off point was reduced to 1.85 AU (Fig. 2). At this on a two-sided test of statistical significance. Signifi- point, the percentage of subjects with any plaque cance was accepted at the level of p value ＜ 0.05. The increased from 70.6% among those with a SAF ＜ 1.85 analyses were performed using SPSS statistical package AU to 85.9% among those with a SAF ≥ 1.85 AU (p (IBM SPSS Statistics for Windows, Version 20.0. ＜ 0.001). This data indicates more than a 2-fold Armonk, NY, USA). increased risk of the presence of an atheromatous plaque (mean difference 2.5 [95% confidence interval Ethics Statement (CI) 1.9 to 3.4]; p ＜ 0.001) in comparison with sub- All participants signed an informed consent, and jects with lower SAF values. In addition, the total the study was approved by the Arnau de Vilanova plaque area was higher among men with a SAF ≥ 1.85 University Hospital ethics committee (CEIC-1410) and AU in comparison with a lower SAF amount (0.81 was conducted according to the ethical guidelines of [0.39 to 1.50] vs. 0.58 [0.24 to 1.15] cm 2, p ＜ 0.001). the 1964 Helsinki Declaration and its later amendments. The multivariable logistic regression model 882 Skin Autofluorescence and Atheromatosis Table 1. Main clinical characteristics and metabolic data of the study population according to the presence of at least one athero- matous plaque Subjects without plaque Subjects with ≥ 1 plaque p value n 760 1808 - Women, n (%) 488 (64.2) 825 (45.6) ＜ 0.001 Age (years) 56 [52 to 61] 58 [53 to 63] ＜ 0.001 Dyslipidemia, n (%) 347 (45.7) 965 (53.5) ＜ 0.001 Total Cholesterol (mg/dL) 198 [178 to 224] 203 [180 to 229] 0.015 LDL Cholesterol 1 (mg/dL) 141 [125 to 154] 144 [129 to 163] 0.017 HDL Cholesterol 1 (mg/dL) 58 [48 to 68] 54 [45 to 66] 0.015 Triglycerides 1 (mg/dL) 124 [92 to 170] 135 [129 to 163] 0.001 Lipid-lowering agents, n (%) 98 (12.9) 349 (19.3) ＜ 0.001 Blood hypertension, n (%) 257 (33.9) 782 (43.3) ＜ 0.001 Systolic BP (mm Hg) 127 [116 to 138] 131 [121 to 143] ＜ 0.001 Diastolic BP (mm Hg) 80 [74 to 87] 82 [76 to 89] ＜ 0.001 Antihypertensives, n (%) 215 (28.3) 625 (34.6) 0.002 Obesity (BMI ≥ 30 kg/m 2), n (%) 229 (30.2) 517 (28.6) 0.437 BMI (kg/m 2) 28.7 [25.6 to 31.6] 28.6 [25.9 to 31.8] 0.563 Former smoker, n (%) 591 (27.7) 1455 (31.1) ＜ 0.001 Current smoker, n (%) 407 (19.1) 1545 (33.0) ＜ 0.001 First-degree family member with premature CVD, n (%) 81 (10.7) 202 (11.2) 0.702 Anticoagulants/Antiplatelets, n (%) 15 (2.0) 57 (3.2) 0.098 HbA1c (%) 5.5 [5.3 to 5.7] 5.5 [5.3 to 5.8] 0.006 eGFR (mL/min per 1.73 m 2) 96.6 [85.5 to 102.7] 95.2 [84.9 to 101.9] 0.063 SAF (AU) 1.8 [1.6 to 2.1] 1.9 [1.7 to 2.3] ＜ 0.001 Data are median [interquartile range] n (percentage); BMI: body mass index; LDL: low density lipoprotein; 1: determination was done in cases in which total cholesterol was ≥ 200 mg/dL and fast for 6 hours or total cholesterol ≥ 250 mg/dL regardless of fasting hours; eGFR: estimated glomer- ular filtration rate estimated according the CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration) equation; BP: blood pressure; BMI: body mass index; CVD: cardiovascular disease; HbA1c: glycated hemoglobin; SAF: skin autofluorescence. Fig. 1. Skin autofluorescence according to the affected territories with atheromatous disease (A) and the quartiles of total plaque area (B) Participants in A) were defined as having focal (1 territory), intermediate (2 to 3 territories), or generalized (4 to 12 territo- ries) atheromatous disease. SAF: skin autofluorescence; AU: arbitrary units; ＊p ＜ 0.001. Patients with focal, intermediate and generalize atheromatous disease showed a significantly higher SAF than subjects without plaque (p ＜ 0.001, for all). Participants in B) were defined by quartiles of total plaque area. Patients in the second, third and fourth quartile showed a significantly higher SAF than subjects in the first quartile (p ＜ 0.001, for all). 883 Sánchez et al . showed that older age, male gender, and current smoker followed by SAF were independent risk factors (all p ＜ 0.001) for atheromatous disease in the entire population (Table 2). In fact, the risk of having ather- oma plaque increases 1.8 times for each SAF unit. Sys- tolic blood pressure and LDL cholesterol also appeared to be independently associated with atheromatous dis- ease, but with a lower odds ratio than the previous factors. When males were evaluated alone, the risk of having any atheroma plaque increases 4.11 times for each SAF unit (Table 3). Finally, no differences in serum concentrations of pentosidine, CML, or RAGE were observed among the 82 patients with and without generalized athero- matous disease (Table 4). In addition, SAF failed to Fig. 2. ROC curve analysis to evaluate the accuracy of skin show a significant correlation with CML (r = 0.036, autofluorescence in presence of subclinical atheroma- p = 0.791), pentosidine (r = 0.140, p = 0.284) or RAGE tous disease from normal cases, together with a sensi- (r = −0.072, p = 0.769) in this small subgroup of par- tivity / specificity report for the male population ticipants. AF: autofluorescence; AUC: area under the ROC curve. The total area under the ROC curve was interpreted following the following guidelines: 0.9 to 1.0, excellent; 0.8 to 0.9, good; 0.7 to 0.8, fair; 0.6 to 0.7, poor, and 0.5 to 0.6, not useful. Table 2. The multivariable logistic regression model for presence of atheromatous dis- ease and its variables associated including the entire population of the study Atheromatous disease OR (95% CI) p value Gender Women Ref. Men 3.27 (2.15 to 4.97) ＜ 0.001 Age (years) 1.11 (1.07 to 1.15) ＜ 0.001 Smoking status Never Ref. Former 2.03 (1.37 to 3.02) ＜ 0.001 Current 4.32 (2.67 to 6.98) ＜ 0.001 Skin autofluorescence (AU) 1.87 (1.23 to 2.86) 0.004 HbA1c (%) 1.39 (0.84 to 2.33) 0.202 Systolic blood pressure (mm Hg) 1.02 (1.01 to 1.03) 0.003 Antihypertensive treatment No Ref. Yes 1.01 (0.69 to 1.49) 0.961 Total Cholesterol (mg/dL) ＜ 200 Ref. ≥ 200 1.53 (0.33 to 7.07) 0.588 LDL Cholesterol (mg/dL) 1.01 (1.00 to 1.01) 0.042 HDL Cholesterol (mg/dL) 1.01 (0.99 to 1.02) 0.239 Triglycerides (mg/dL) 1.00 (0.99 to 1.05) 0.138 Lipid-lowering agents No Ref. Yes 0.70 (0.41 to 1.18) 0.175 Test of fit Hosmer–Lemeshow 0.118 Area under de ROC curve 0.76 (0.70 to 0.81) ＜ 0.001 884 Skin Autofluorescence and Atheromatosis Table 3. Independent variables significantly associated with the presence of atheroma- tous disease in the multivariable logistic regression model according to gender OR (95% CI) p value Atheromatous disease in male Age (years) 1.07 (1.01 to 1.14) 0.025 Smoking status Never Ref. Former 1.95 (1.01 to 3.79) 0.050 Current 3.48 (1.63 to 7.42) 0.001 Skin autofluorescence (AU) 4.11 (1.71 to 9.87) 0.002 Test of fit Hosmer–Lemeshow - 0.522 Area under de ROC curve 0.77 (0.71 to 0.82) ＜ 0.001 Atheromatous disease in female Age (years) 1.12 (1.08 to 1.17) ＜ 0.001 Smoking status Never Ref. Former 2.13 (1.27 to 3.57) 0.004 Current 5.02 (2.64 to 9.53) ＜ 0.001 Systolic blood pressure (mm Hg) 1.02 (1.00 to 1.03) 0.012 LDL Cholesterol (mg/dL) 1.01 (1.00 to 1.01) 0.029 Test of fit Hosmer–Lemeshow - 0.392 Area under de ROC curve 0.73 (0.68 to 0.77) ＜ 0.001 The independent variables included in the analysis were SAF and well-established cardiovascular risk factors such as gender, age, smoking status, systolic blood pressure, lipid profile (total choles- terol, LDL cholesterol, HDL cholesterol, and triglycerides), together with HbA1c and antihyper- tensive or lipid-lowering treatment. Table 4. Main clinical characteristics and serum concentration of advanced glycation end-products and its receptor of the nested case-control study according to the presence of generalized atheromatous disease Subjects without plaque Subjects with plaque and p value generalized atheromatous disease n 41 41 Women, n (%) 14 (34.1) 15 (36.6) 1.000 Age (years) 59 [55 to 63] 60 [55 to 65] 0.656 BMI (kg/m 2) 28.0 [25.9 to 29.9] 28.0 [26.5 to 30.5] 0.597 Total cholesterol (mg/dL) 210 [189 to 228] 202 [183 to 222] 0.358 Pentosidine (ng/mL) 25.8 [8.7 to 54.2] 21.9 [8.9 to 42.8] 0.842 CML (ng/mL) 105.0 [64.4 to 121.6] 109.8 [85.6 to 143.1] 0.070 AGEs receptor (ng/mL) 0.044 [0.018 to 0.142] 0.039 [0.023 to 0.066] 0.531 SAF (AU) 1.8 [1.6 to 2.1] 2.1 [1.9 to 2.4] ＜ 0.001 BMI: body mass index; CML: carboxymethyl-lysine; AGEs: advanced end glycation products; SAF: skin autofluorescence; AU: arbitrary units. our cohort of middle-aged asymptomatic subjects, a Discussion close relationship between SAF and the atheromatous To the best of our knowledge, this is the first plaque burden (the presence of plaque, the number of study in which SAF has been evaluated as a marker of affected territories, and the total area of atheromatous subclinical atheromatous disease in a large cohort of plaque) has been observed. subjects without a history of cardiovascular disease. In The prevalence of subclinical atheromatous dis- 885 Sánchez et al . ease in our study was 70.4% compared with the 63% nents of the metabolic syndrome, as well as with lower described in the middle-aged asymptomatic sample HDL cholesterol 34). In addition, SAF assessment has without cardiovascular disease from the Progression of been associated with coronary and peripheral artery Early Subclinical Atherosclerosis study 22). These stud- disease 35-37). Moreover, the added clinical value of SAF ies reveal that the presence of atheromatous plaque in is supported by data that reflects its measurement as a asymptomatic subjects is a common condition that strong predictor of carotid atherosclerosis, long-term deserves particular attention in the prevention of car- cardiovascular complications, and mortality in popu- diovascular outcomes. Moreover, when nearly one- lations with increased cardiovascular risk, such as third of participants from the PESA study classified at patients with type 1 diabetes and renal disease 38-42). low risk by traditional scales (i.e.,: Framingham Heart As atheromatous disease is a diffuse process, the Study and the European Systematic Coronary Risk measurement of the total plaque area seems to be a Evaluation) had intermediate or generalized atheroma- more reliable approach to the patient at risk 43). It is tous disease 22). worth mentioning that our study demonstrates a sig- The present study also shows that when the nificant and independent association between SAF screening of atheromatous disease was limited to and total area plaque. carotid arteries, we failed to detect at least 12% of The association between circulating AGEs and patients with disease in femoral regions. This finding both all-cause mortality and cardiovascular mortality is in concordance with data from the Carotid-Femoral has been previously assessed in moderate-sized epide- Ultrasound Morphology and Cardiovascular Events miological studies focused on healthy aging popula- study, which showed that scanning only carotid or tion, disabled older women, and patients with type 1 only femoral regions predicted 15% and 13% fewer diabetes 11, 44-47). In this regard, in an adult population events than examining both territories in a 10-year of men and women aged 65 and older, serum CML follow-up 23). was associated with an increased incidence of coronary Previous population studies, such as the Multi- heart disease and stroke, independent of traditional ethnic Study of Atherosclerosis, the US High Risk atherosclerosis risk factors 11). However, Hanssen et al. Plaque Study, and the Northern Manhattan Study showed that plasma AGE measurements were not dif- established robust associations between cardiovascular ferent when comparing individuals with and without events, subclinical carotid and coronary disease 4, 5, 20, 24). cardiovascular disease in participants from two Dutch Similarly, in middle-aged Spanish men from the Ara- cohort studies comprising 1,291 subjects with various gon Workers’ Health Study, subclinical carotid and degrees of impaired glucose metabolism 12). Similarly, femoral plaques were associated with coronary artery in a modest-sized cohort with chronic kidney disease calcium score 25), a well-recognized risk factor of coro- subjects, no association between CML and incident nary outcomes 26). In this setting, our data shows that CVD was detected 13). Our data also failed to find any SAF is a useful tool for helping clinicians to identify difference between circulating AGEs as well as its individuals at risk of developing cardiovascular disease. receptors with the presence of subclinical atheroma- Some characteristic effects of AGEs on plaque tous disease, together with a lack of correlation formation have been previously described, such as between those serum markers and SAF. This result increased oxidative stress, modifications of collagen could be owing to the fact that circulating AGEs are and other proteins that constitute the extracellular rapidly broken down to AGE peptides or free AGEs, matrix of the arterial wall and the basement mem- which are excreted through the kidney, thus having a brane, the promotion of LDL trapping in the sub- fast turnover 6, 48). By contrast, AGEs linked to proteins endothelial compartment, and the increment of vascu- from subcutaneous tissue have a slow turnover, thus lar stiffness 27-31). Moreover, the binding of AGEs to reflecting a decades-long accumulation. their receptors activates nuclear factor kappa B, among There are some potential limitations that must other pro-oxidant and pro-inflammatory pathways 32, 33). be highlighted in our study. First, as an observational Therefore, the role of AGEs favors the formation of and cross-sectional study, a causal relationship between atheromatous plaques causing widespread damage to SAF or circulating AGEs could not be established. arteries through the upregulation of inflammation, Second, SAF could be unreliable in subjects with dark oxidative stress, and the cross-linking of collagen and skin due to excessive light absorption. We overcame other proteins of the arterial wall. In addition, AGEs this limitation by selecting only Caucasian subjects also could act as triggers from stable to rupture-prone and excluding those with medium brown skin. Third, atheromatous plaques 9). the lack of statistical power of our nested study Previous studies have shown that SAF levels because of the small sample size assessed (82 subjects increase proportionally with the number of compo- from the 2,568 in whom skin autofluorescence and 886 Skin Autofluorescence and Atheromatosis bilateral carotid and femoral areas were explored). Therefore, the data are not definitive but rather References hypothesis-generating in nature, and larger studies will 1) Vilahur G, Badimon J, Bugiardini R, Badimon L: Per- be needed to confirm our findings. In addition, as we spectives: the burden of cardiovascular risk factors and enrolled subjects with generalized atheromatous dis- coronary heart disease in Europe and worldwide. Eur Heart J Supplements, 2014; 16: A7-A11 ease, the generalizability to the overall population with 2) McMahan CA, Gidding SS, Viikari JS, Juonala M, at least one atheromatous plaque is limited. Kähönen M, Hutri-Kähönen N, Jokinen E, Taittonen L, Pietikäinen M, McGill HC Jr, Raitakari OT: Association of pathobiologic determinants of atherosclerosis in youth Conclusion risk score and 15-year change in risk score with carotid There is a clear need to identify subjects at artery intima-media thickness in young adults (from the increased risk of CVD. Generally, clinicians have used Cardiovascular Risk in Young Finns Study). Am J Cardiol, global risk scores that combine multiple traditional 2007; 100: 1124-1129 3) Berry JD, Liu K, Folsom AR, Lewis CE, Carr JJ, Polak JF, cardiovascular risk factors to classify patients. Our Shea S, Sidney S, O’Leary DH, Chan C, Lloyd-Jones study shows that SAF is associated with an increased DM: Prevalence and progression of subclinical atheroscle- atherosclerotic burden especially in men, and may rosis in younger adults with low short-term but high life- provide clinically relevant information to the standard time estimated risk for cardiovascular disease: the Coro- cardiovascular risk assessment strategies. nary Artery Risk Development in Young Adults Study and Multi-Ethnic Study of Atherosclerosis. Circulation, 2009; 119: 382-389 Acknowledgments 4) Zavodni AE, Wasserman BA, McClelland RL, Gomes AS, Folsom AR, Polak JF, Lima JA, Bluemke DA: Carotid The authors would also like to thank Fundació artery plaque morphology and composition in relation to Renal Jaume Arnó, Eva Castro, Virtudes María, Marta incident cardiovascular events: the Multi-Ethnic Study of Elias, Teresa Vidal, Mª del Valle Peña, Cristina Atherosclerosis (MESA). Radiology, 2014; 271: 381-389 Dominguez, Noemi Nova, Alba Prunera, Núria Sans, 5) Gibson AO, Blaha MJ, Arnan MK, Sacco RL, Szklo M, Meritxell Soria, and the Primary Care teams from Herrington DM, Yeboah J: Coronary artery calcium and Lleida for recruiting subjects and their efforts in the incident cerebrovascular events in an asymptomatic accurate development of the ILERVAS project. cohort: the MESA study. JACC Cardiovasc Imaging, 2014; 7: 1108-1115 6) Ajith TA, Vinodkumar P: Advanced glycation end prod- Funding Source ucts: association with the pathogenesis of diseases and the current therapeutic approaches. Curr Clin Pharmacol, This work was supported by grants from the 2016; 11: 118-127 Diputació de Lleida, Instituto de Salud Carlos III 7) Vlassara H, Fuh H, Donnelly T, Cybulsky M: Advanced (Action Plan II14//00008), Fundación de la Sociedad glycation endproducts promote adhesion molecule Española de Endocrinología y Nutrición (FSEEN), (VCAM-1, ICAM-1) expression and atheroma formation in normal rabbits. Mol Med, 1995; 1: 447-456 Esteve Laboratory, Generalitat de Catalunya 8) Ishibashi Y, Matsui T, Maeda S, Higashimoto Y, Yamagi- (2017SGR696 and SLT0021600250), IRBLleida Bio- shi S: Advanced glycation end products evoke endothelial bank (B.0000682) and Plataforma Biobancos cell damage by stimulating soluble dipeptidyl peptidase-4 PT13/0010/0014. CIBER de Diabetes y Enferme- production and its interaction with mannose 6-phos- dades Metabólicas Asociadas and CIBER de Enferme- phate/insulin-like growth factor II receptor. Cardiovasc dades Respiratorias are initiatives of the Instituto de Diabetol, 2013; 12: 125 Salud Carlos III. These organizations had no role in 9) Hanssen NM, Wouters K, Huijberts MS, Gijbels MJ, study design, the collection, analysis and interpreta- Sluimer JC, Scheijen JL, Heeneman S, Biessen EA, Dae- men MJ, Brownlee M, de Kleijn DP, Stehouwer CD, Pas- tion of data, report writing, or the decision to submit terkamp G, Schalkwijk CG: Higher levels of advanced the article for publication. The authors declare that glycation endproducts in human carotid atherosclerotic there are no relationships with industry relevant to plaques are associated with a rupture-prone phenotype. this manuscript. Eur Heart J, 2014; 35: 1137-1146 10) Betriu À, Farràs C, Abajo M, Martinez-Alonso M, Arroyo D, Barbé F, Buti M, Lecube A, Portero M, Purroy F, Tor- Conflicts of Interests res G, Valdivielso JM, Fernández E: Randomised inter- vention study to assess the prevalence of subclinical vascu- The authors declare that there is no conflict of lar disease and hidden kidney disease and its impact on interest. morbidity and mortality: The ILERVAS project. Nefrolo- gia, 2016: 36: 389-396 11) Kizer JR, Benkeser D, Arnold AM, Ix JH, Mukamal KJ, 887 Sánchez et al . Djousse L, Tracy RP, Siscovick DS, Psaty BM, Zieman SJ: 2008; 70: 1200-1207 Advanced glycation/glycoxidation endproduct carboxy- 21) Meerwaldt R, Graaff R, Oomen PHN, Links TP, Jager JJ, methyl-lysine and incidence of coronary heart disease and Alderson NL, Thorpe SR, Baynes JW, Gans ROB, Smit stroke in older adults. Atherosclerosis, 2014; 235: 116- AJ: Simple non-invasive assessment of advanced glycation 121 endproduct accumulation. Diabetologia, 2004; 47: 1324- 12) Hanssen NM, Engelen L, Ferreira I, Scheijen JL, Hui- 1330 jberts MS, van Greevenbroek MM, van der Kallen CJ, 22) Fernández-Friera L, Peñalvo JL, Fernández-Ortiz A, Iba- Dekker JM, Nijpels G, Stehouwer CD, Schalkwijk CG: ñez B, López-Melgar B, Laclaustra M, Oliva B, Mocoroa Plasma levels of advanced glycation endproducts Nε-(car- A, Mendiguren J, Martínez de Vega V, García L, Molina J, boxymethyl)lysine, Nε-(carboxyethyl)lysine, and pentosi- Sánchez-González J, Guzmán G, Alonso-Farto JC, Gual- dine are not independently associated with cardiovascular lar E, Civeira F, Sillesen H, Pocock S, Ordovás JM, Sanz disease in individuals with or without type 2 diabetes: the G, Jiménez-Borreguero LJ, Fuster V: Prevalence, vascular Hoorn and CODAM studies. J Clin Endocrinol Metab, distribution, and multiterritorial extent of subclinical ath- 2013; 98: E1369-E1373 erosclerosis in a middle-aged cohort: the PESA (Progres- 13) Busch M, Franke S, Muller A, Wolf M, Gerth J, Ott U, sion of Early Subclinical Atherosclerosis) study. Circula- Niwa T, Stein G: Potential cardiovascular risk factors in tion, 2015; 131: 2104-2113 chronic kidney disease: Ages, total homocysteine and 23) Belcaro G, Nicolaides AN, Ramaswami G, Cesarone MR, metabolites, and the c-reactive protein. Kidney Int, 2004; De Sanctis M, Incandela L, Ferrari P, Geroulakos G, Bar- 66: 338-347 sotti A, Griffin M, Dhanjil S, Sabetai M, Bucci M, Mar- 14) Bolíbar B, Fina Avilés F, Morros R, Garcia-Gil Mdel M, tines G: Carotid and femoral ultrasound morphology Hermosilla E, Ramos R, Rosell M, Rodríguez J, Medina screening and cardiovascular events in low risk subjects: a M, Calero S, Prieto-Alhambra D; Grupo SIDIAP. [SID- 10-year follow-up study (the CAFES-CAVE study). Ath- IAP database: electronic clinical records in primary care as erosclerosis, 2001; 156: 379-387 a source of information for epidemiologic research]. Med 24) Baber U, Mehran R, Sartori S, Schoos MM, Sillesen H, Clin (Barc), 2012; 138: 617-621 Muntendam P, Garcia MJ, Gregson J, Pocock S, Falk E, 15) Diabetes mellitus: a major risk factor for cardiovascular Fuster V: Prevalence, impact and predictive value of disease. A joint editorial statement by the American Dia- detecting subclinical coronary and carotid atherosclerosis betes Association; The National Heart, Lung, and Blood in asymptomatic adults: the BioImage Study. J Am Coll Institute; The Juvenile Diabetes Foundation International; Cardiol, 2015; 65: 1065-1074 The National Institute of Diabetes and Digestive and 25) Laclaustra M, Casasnovas JA, Fernández-Ortiz A, Fuster Kidney Diseases; and The American Heart Association. V, León-Latre M, Jiménez-Borreguero LJ, Pocovi M, Circulation, 1999; 100: 1132-1133 Hurtado-Roca Y, Ordovas JM, Jarauta E, Guallar E, Iba- 16) Sabetai MM, Tegos TJ, Nicolaides AN, Dhanjil S, Pare ñez B, Civeira F: Femoral and carotid subclinical athero- GJ, Stevens JM: Reproducibility of computer-quantified sclerosis association with risk factors and coronary cal- carotid plaque echogenicity: can we overcome the subjec- cium: The AWHS Study. J Am Coll Cardiol, 2016; 67: tivity? Stroke, 2000; 31: 2189-2196 1263-1274 17) Touboul PJ, Hennerici G, Meairs S, Adams H, Amarenco 26) Mitchell JD, Paisley R, Moon P, Novak E, Villines TC. P, Desvarieux M, Ebrahim S, Fatar M, Hernandez Her- Coronary artery calcium and long-term risk of death, nandez R, Kownator S, Prati P, Rundek T, Taylor A, myocardial infarction, and stroke: The Walter Reed Bornstein N, Csiba L, Vicaut E, Woo KS, Zannad F; Cohort Study. JACC Cardiovasc Imaging, 2017; S1936- Advisory Board of the 3rd Watching the Risk Symposium 878X: 30891-30894 2004, 13th European Stroke Conference: Mannheim 27) Janda K, Krzanowski M, Gajda M, Dumnicka P, Jasek E, intima-media thickness consensus. Cerebrovasc Dis, Fedak D, Pietrzycka A, Kuźniewski M, Litwin JA, 2004; 18: 346-349 Sułowicz W. Vascular effects of advanced glycation end- 18) Stein JH, Korcarz CE, Hurst RT, Lonn E, Kendall CB, products: content of immunohistochemically detected Mohler ER, Najjar SS, Rembold CM, Post WS; American AGEs in radial artery samples as a predictor for arterial Society of Echocardiography Carotid Intima-Media calcification and cardiovascular risk in asymptomatic Thickness Task Force: Use of carotid ultrasound to iden- patients with chronic kidney disease. Dis Markers, 2015; tify subclinical vascular disease and evaluate cardiovascular 2015: 153978 disease risk: a consensus statement from the American 28) van Eupen MG, Schram MT, van Sloten TT, Scheijen J, society of echocardiography carotid intima-media thick- Sep SJ, van der Kallen CJ, Dagnelie PC, Koster A, ness task force endorsed by the society for vascular medi- Schaper N, Henry RM, Kroon AA, Smit AJ, Stehouwer cine. J Am Soc Echocardiogr, 2008; 21: 93-111 CD, Schalkwijk CG: Skin autofluorescence and pentosi- 19) Tiozzo E, Gardener H, Hudson BI, Dong C, Della-Morte dine are associated with aortic stiffening: The Maastricht D, Crisby M, Goldberg RB, Elkind MS, Cheung YK, study. Hypertension, 2016; 68: 956-963 Wright CB, Sacco RL, Rundek T: High-density lipopro- 29) Saremi A, Howell S, Schwenke DC, Bahn G, Beisswenger tein subfractions and carotid plaque: The Northern Man- PJ, Reaven PD.; VADT Investigators: Advanced Glyca- hattan Study. Atherosclerosis, 2014; 237: 163-168 tion End Products, Oxidation Products, and the Extent of 20) Rundek T, Arif H, Boden-Albala B, Elkind MS, Paik MC, Atherosclerosis During the VA Diabetes Trial and Follow- Sacco RL: Carotid plaque, a subclinical precursor of vas- up Study. Diabetes Care, 2017; 40: 591-598 cular events: the Northern Manhattan Study. Neurology, 30) Xu L, Wang YR, Li PC, Feng B: Advanced glycation end 888 Skin Autofluorescence and Atheromatosis products increase lipids accumulation in macrophages 39) Kimura H, Tanaka K, Kanno M, Watanabe K, Hayashi Y, through upregulation of receptor of advanced glycation Asahi K, Suzuki H, Sato K, Sakaue M, Terawaki H, end products: increasing uptake, esterification and Nakayama M, Miyata T, Watanabe T: Skin autofluores- decreasing efflux of cholesterol. Lipids Health Dis, 2016; cence predicts cardiovascular mortality in patients on 15: 161 chronic hemodialysis. Ther Apher Dial, 2014; 18: 461- 31) Ninomiya H, Katakami N, Sato I, Osawa S, Yamamoto Y, 467 Takahara M, Kawamori D, Matsuoka TA, Shimomura I: 40) Velayoudom-Cephise FL, Rajaobelina K, Helmer C, Nov Association between subclinical atherosclerosis markers S, Pupier E, Blanco L, Hugo M, Farges B, Astrugue C, and the level of accumulated advanced glycation end- Gin H, Rigalleau V: Skin autofluorescence predicts car- products in the skin of patients with diabetes. J Athero- dio-renal outcome in type 1 diabetes: a longitudinal study. scler Thromb, 2018; 25: 1274-1284 Cardiovasc Diabetol, 2016; 15: 127 32) Schmidt AM, Hasu M, Popov D, Zhang JH, Chen J, Yan 41) Sánchez E, Betriu À, Arroyo D, López C, Hernández M, SD, Brett J, Cao R, Kuwabara K, Costache G: Receptor Rius F, Fernández E, Lecube A: Skin Autofluorescence for advanced glycation end products (AGEs) has a central and Subclinical Atherosclerosis in Mild to Moderate role in vessel wall interactions and gene activation in Chronic Kidney Disease: A Case-Control Study. PLoS response to circulating AGE proteins. Proc Natl Acad Sci One, 2017; 12: e0170778 USA, 1994: 91: 8807-8811 42) Osawa S, Katakami N, Kuroda A, Takahara M, Sakamoto 33) Basta G: Receptor for advanced glycation endproducts F, Kawamori D, Matsuoka T, Matsuhisa M, Shimomura I: and atherosclerosis: From basic mechanisms to clinical Skin autofluorescence is associated with early stage ath- implications. Atherosclerosis, 2008; 196: 9-21 erosclerosis in patients with Type 1 Diabetes. J Atheroscler 34) van Waateringe RP, Slagter SN, van Beek AP, van der Thromb, 2017; 24: 312-326 Klauw MM, van Vliet-Ostaptchouk JV, Graaff R, Pater- 43) Barnett PA, Spence JD, Manuck SB, Jennings JR: Psycho- son AD, Lutgers HL, Wolffenbuttel BHR: Skin autofluo- logical stress and the progression of carotid artery disease. rescence, a non-invasive biomarker for advanced glycation J Hypertens, 1997; 15: 49-55 end products, is associated with the metabolic syndrome 44) Semba RD, Bandinelli S, Sun K, Guralnik JM, Ferrucci L: and its individual components. Diabetol Metab Syndr, Plasma carboxymethyl-lysine, an advanced glycation end 2017; 9: 42 product, and all-cause and cardiovascular disease mortal- 35) Mulder DJ, van Haelst PL, Gross S, de Leeuw K, Bijzet J, ity in older community-dwelling adults. J Am Geriatr, Graaff R, Gans RO, Zijlstra F, Smit AJ: Skin autofluores- 2009; 57: 1874-1880 cence is elevated in patients with stable coronary artery 45) Semba RD, Ferrucci L, Sun K, Beck J, Dalal M, Varadhan disease and is associated with serum levels of neopterin R, Walston J, Guralnik JM, Fried LP: Advanced glycation and the soluble receptor for advanced glycation end prod- end products and their circulating receptors predict car- ucts. Atherosclerosis, 2008; 197: 217-223 diovascular disease mortality in older community-dwell- 36) de Vos LC, Noordzij MJ, Mulder DJ, Smit AJ, Lutgers ing women. Aging Clin Exp Res, 2009; 21: 182-190 HL, Dullaart RP, Kamphuisen PW, Zeebregts CJ, 46) Kilhovd BK, Juutilainen A, Lehto S, et al.: Rönnemaa T, Lefrandt JD: Skin autofluorescence as a measure of Torjesen PA, Hanssen KF, Laakso M serum levels of advanced glycation end products deposition is elevated in advanced glycation endproducts predict total, cardiovas- peripheral artery disease. Arterioscler Thromb Vasc Biol, cular and coronary mortality in women with type 2 dia- 2013; 33: 131-138 betes: A population-based 18 year follow-up study. Diabe- 37) Hangai M, Takebe N, Honma H, Sasaki A, Chida A, tologia, 2007; 50: 1409-1417 Nakano R, Togashi H, Nakagawa R, Oda T, Matsui M, 47) Nin JW, Jorsal A, Ferreira I, Schalkwijk CG, Prins MH, Yashiro S, Nagasawa K, Kajiwara T, Takahashi K, Taka- Parving HH, Tarnow L, Rossing P, Stehouwer CD: hashi Y, Satoh J, Ishigaki Y: Association of advanced gly- Higher plasma levels of advanced glycation end products cation end products with coronary artery calcification in are associated with incident cardiovascular disease and all- Japanese subjects with Type 2 Diabetes as assessed by skin cause mortality in type 1 diabetes: A 12-year follow-up autofluorescence. J Atheroscler Thromb, 2016; 23: 1178- study. Diabetes Care, 2011; 34: 442-447 1187 48) van Waateringe RP, Slagter SN, van der Klauw MM, van 38) Meerwaldt R, Lutgers HL, Links TP, raaff R, Baynes JW, Vliet-Ostaptchouk JV, Graaff R, Paterson AD, Lutgers Gans RO, Smit AJ: Skin autofluorescence is a strong pre- HL, Wolffenbuttel BH: Lifestyle and clinical determi- dictor of cardiac mortality in diabetes. Diabetes Care, nants of skin autofluorescence in a population-based 2007; 30: 107-112 cohort study. Eur J Clin Invest, 2016; 46: 481-490 889