Associations between Specific Redox Biomarkers and Age in a Large European Cohort: The MARK-AGE Project

Resea ch A icle Associa ions be ween Speci ic Redox Bioma ke s and Age in a La ge Eu opean Coho : The MARK-AGE P ojec Daniela Webe , 1,2 Wol gang S ue z, 3 Oli ie Toussain , 4 Flo ence Debacq-Chainiaux, 4 Ma ijn E. T. Dollé, 5 Eugène Jansen, 5 E s a hios S. Gonos, 6 Claudio F anceschi, 7 Ewa Siko a, 8 An i He onen, 9 Nicolle B eusing, 3,10 Thilo Sindlinge , 11 Ma ía Mo eno-Villanue a, 11 Alexande Bü kle, 11 and Tilman G une 1,2,12,13 1 Depa men o Molecula Toxicology, Ge man Ins i u e o Human Nu i ion Po sdam-Rehb uecke (DI E), 14558 Nu he al, Ge many 2 Nu iAc -Compe ence Clus e Nu i ion Resea ch Be lin-Po sdam, 14458 Nu he al, Ge many 3 Ins i u e o Biological Chemis y and Nu i ion, Uni e si y o Hohenheim, 70599 S u ga , Ge many 4 URBC-NARILIS, Uni e si y o Namu , 5000 Namu , Belgium 5 Na ional Ins i u e o Public Heal h and he En i onmen (RIVM), 3720 BA Bil ho en, Ne he lands 6 Ins i u e o Biological Resea ch and Bio echnology, Na ional Hellenic Resea ch Founda ion (NHRF), 11635 A hens, G eece 7 Depa men o Expe imen al Pa hology, Uni e si y o Bologna, 40126 Bologna, I aly 8 Nencki Ins i u e o Expe imen al Biology, Polish Academy o Sciences, 02-093 Wa saw, Poland 9 School o Medicine, Uni e si y o Tampe e, 33014 Tampe e, Finland 10 Depa men o Applied Nu i ional Science/Die e ics, Ins i u e o Nu i ional Medicine, Uni e si y o Hohenheim, 70599 S u ga , Ge many 11 Molecula Toxicology, Depa men o Biology, Uni e si y o Kons anz, 78457 Kons anz, Ge many 12 Ge man Cen e o Diabe es Resea ch (DZD), 85764 Munich-Neuhe be g, Ge many 13 Ge man Cen e o Ca dio ascula Resea ch (DZHK), 13357 Be lin, Ge many Co espondence should be add essed o Daniela Webe ; [email p o ec ed] Recei ed 24 Feb ua y 2017; Re ised 29 May 2017; Accep ed 15 June 2017; Published 19 July 2017 Academic Edi o : Daniela Gius a ini Copy igh © 2017 Daniela Webe e al. This is an open access a icle dis ibu ed unde he C ea i e Commons A ibu ion License, which pe mi s un es ic ed use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly ci ed. Oxida i e s ess and an ioxidan s play a ole in age- ela ed diseases and in he aging p ocess. We he e p esen da a on p o ein ca bonyls, 3-ni o y osine, malondialdehyde, and cellula and plasma an ioxidan s (glu a hione, cys eine, asco bic acid, u ic acid, α- ocophe ol, and lycopene) and hei ela ion wi h age in he Eu opean mul icen e s udy MARK-AGE. To a oid con ounding, only da a om coun ies which ec ui ed subjec s om all h ee s udy g oups (fi e o eigh cen e s) and only pa icipan s aged ≥55 yea s we e selec ed esul ing in da a om 1559 pa icipan s. These included subjec s om (1) he gene al popula ion, (2) membe s om long-li ing amilies, and (3) hei spouses. In addi ion, 683 middle-aged e e ence pa icipan s (35–54 yea s) se ed as a con ol. A e adjus men o age, BMI, smoking s a us, gende , and coun y, he e we e diffe ences in p o ein ca bonyls, malondialdehyde, 3-ni o y osine, α- ocophe ol, cys eine, and glu a hione be ween he 3 s udy g oups. P o ein ca bonyls and 3-ni o y osine as well as cys eine, u ic acid, and lycopene we e iden ified as independen bioma ke s wi h he highes co ela ion wi h age. In e es ingly, om all an ioxidan s measu ed, only lycopene was lowe in all aged g oups and om he oxida i e s ess bioma ke s, only 3-ni o y osine was inc eased in he descendan s om long-li ing amilies compa ed o he middle-aged con ol g oup. We conclude ha bo h li es yle and gene ics may be impo an con ibu o s o edox bioma ke s in an aging popula ion. Hindawi Oxida i e Medicine and Cellula Longe i y Volume 2017, A icle ID 1401452, 12 pages h ps://doi.o g/10.1155/2017/1401452 1. In oduc ion Oxida i e s ess and an ioxidan s ha e been in he ocus o esea ch o decades due o hei associa ion wi h nume ous age- ela ed diseases [1–4]. In humans, ele a ed le els o oxida i e s ess ha e been epo ed in se e al medical condi- ions, including neu odegene a i e diseases [2], obesi y, diabe es melli us, and he aging p ocess i sel [3–5]. The e is some epidemiological e idence on he ole o oxida i e s ess in aging, age- ela ed diseases, and mo ali y [6–10]. Howe e , hese s udies ocused mos ly on single bioma ke s o on bioma ke s which a e no widely used, hus making i difficul o compa e esul . When a emp ing o assess edox bioma ke s, i is impo an o no e ha he e is no one single bioma ke which is conside ed a “gold s anda d.”In ac , i is ecom- mended o measu e a se o diffe en bioma ke s [11]. Thus, ou a emp was o analyze ma ke s o p o ein oxida- ion, ni a ion, lipid pe oxida ion, and cellula and plasma an ioxidan s and s udy hei ela ion wi h age in pa icipan s o he MARK-AGE p ojec . The MARK-AGE p ojec was a Eu opean mul icen e s udy, suppo ed by he Eu opean Commission, aiming o iden i y bioma ke s o human aging and o model a obus se o ma ke s o biological age and heal hy aging [12]. Fo his pu pose, women and men we e ec ui ed om he gene al popula ion om eigh Eu opean coun ies as age-s a ified subg oups, as well as subjec s belonging o a amily wi h long-li ing membe s oge he wi h hei spouses. In diffe en biological ma ices (whole blood, se um, plasma, u ine, buccal mucosa cells, and pe iphe al blood mononu- clea cells), a la ge numbe o candida e bioma ke s o aging, including DNA-based ma ke s, ma ke s based on p o eins and hei modifica ions, immunological ma ke s, clinical chemis y, ho mones and ma ke s o me abolism, oxida i e s ess ma ke s, and an ioxidan mic onu ien s we e assessed. The p ojec design including de ails o he s udy popula ion and s anda d ope a ing p ocedu es ha e been published ecen ly [12–15]. The objec i es o he p esen wo k we e o assess and o compa e le els o edox bioma ke s wi hin he h ee diffe en s udy g oups o he MARK-AGE p ojec . Ou hypo heses we e (1) ha oxida i e s ess is ele a ed in highe age g oups as i has been shown in diffe en a he small popula ion- based s udies in diffe en biological ma ices and (2) ha subjec s om amilies wi h long-li ing membe s would be gene ically be e equipped o handle oxida i e s ess han he gene al popula ion. The esul s o he spouses migh show whe he (3) a sha ed li es yle may be able o influence bioma ke concen a ions. 2. Ma e ials and Me hods This s udy was conduc ed in acco dance wi h he Decla a ion o Helsinki (1964) and wi h in o med w i en consen o each pa icipan . E hical clea ance had been gi en by he e hics commi ee o each o he ec ui ing cen e s. This s udy has been egis e ed e ospec i ely a he Ge man Clinical T ials Regis e (DRKS00007713). 2.1. S udy Popula ion and Sample Collec ion. Fo he whole MARK-AGE p ojec , 3158 pa icipan s we e ec ui ed in 8 ec ui ing cen e s and he de ails o he h ee s udy g oups will be desc ibed b iefly below. The fi s s udy g oup was ec ui ed h ough a ious public pla o ms such as adio and newspape ad e ising. These pa icipan s we e included in he RASIG g oup ( ec ui ed om he age-s a ified gene al popula ion). Fo his g oup, he main inclusion c i e ia we e he abili y o gi e in o med consen and being in he age- ange om 35 o 75 yea s (bo h gende s). The second s udy g oup consis ed o 537 descendan s om long-li ing subjec s (nonagena ians; pe sons who eached he age o 90) who had been ec ui ed as a ollow- up om he GEHA s udy (Gene ics o Heal hy Aging; 2004–2009, o de ails, see [16]); his g oup is abb e ia ed as GO (GEHA offsp ing). The hi d s udy g oup consis ed o he spouses o he GEHA offsp ing (n= 311) and se ed as a li es yle con ol g oup, he so-called SGO (spouses o GO). The GO and SGO pa icipan s we e be ween he age o 55 and 75 yea s [12] and we e ec ui ed in Belgium, Finland, G eece, I aly, The Ne he lands, and Poland. Collec ion o an h opome ic da a, ques ionnai e da a, and da a on cogni i e unc ion was ca ied ou by ained nu ses o physicians be ween No embe 2008 and June 2012 a he ollowing ec ui ing cen e s: Hall in Ti ol/Inns- b uck (Aus ia), Namu (Belgium), Esslingen (Ge many), A hens, and o he nea by egions (G eece), Bologna (I aly), Wa saw (Poland), Tampe e (Finland), and Leiden (The Ne he lands). Fu he mo e, pa icipan s we e asked o comple e ques ionnai es on li es yle cha ac e is ic (nu i ion, smoking habi s), amily his o y, and li ing en i onmen . To a oid possible con ounding, we used only da a om hose coun ies which ec ui ed subjec s om all h ee s udy g oups (RASIG, GO, and SGO). This was he case in fi e o he eigh cen e s (Belgium, G eece, I aly, Poland, and Finland). Ge many and Aus ia did no p o ide GO o SGO da a while The Ne he lands did no ec ui RASIG pa icipan s; he e o e, da a om hese h ee coun ies we e excluded in he p esen analyses. Since he age anges o pa icipan s in he GO and SGO g oups we e diffe en o he RASIG g oup in gene al (55– 75 e sus 35–75 yea s, esp.), only pa icipan s who we e ≥55 yea s o age we e selec ed. This esul ed in a o al o 1559 pa icipan s in all h ee s udy g oups o be included in he p esen wo k. Addi ionally, o compa e hese s udy g oups o a younge (middle-aged) e e ence g oup, we selec ed all RASIG pa icipan s aged 35–54 (“middle aged”) om Belgium, G eece, I aly, Poland, and Finland, which esul ed in a o al o 683 e e ence pa icipan s. 2.2. De e mina ion o To al Glu a hione and To al F ee Cys eine in Whole Blood. To al glu a hione and o al ee cys eine in whole blood we e measu ed as p e iously desc ibed by Chen e al. [17] by using Ellman’s eagen (5,5′ -di hiobis-[2-ni obenzoic acid]) a e he educ ion o any disulfides p esen . The modifica ions ega ding he educ ion agen DTT (1,4-di hio h ei ol), he adap ion o whole blood samples, and HPLC condi ions we e as 2 Oxida i e Medicine and Cellula Longe i y ollows: Whole blood (100 μL) was o ex-mixed wi h DTT (12.5 mM, 100 μL) and incuba ed o 3 min; cold i- chlo oace ic acid solu ion (10% w/ , 200 μL) was added; and samples we e ho oughly mixed and cen i uged a 19,500 ×g o 5 min a 4 ° C. The clea supe na an (200 μL) was hen added o Ellman’s eagen (30 mM, 50 μL) oge he wi h di-po assium hyd ogen phospha e buffe (2 M, 100 μL), and o ex-mixed. Twen y μL was analyzed on a Shimadzu P ominence HPLC (LC-20A) equipped wi h an UV-Vis de ec o (SPD-20AV se a 326 nm). The sepa a ion o cys e- ine and glu a hione was achie ed by using a Rep osil-Pu 120 C18 AQ column (5 μm, 250 mm ×4.6 mm; D . Maisch, Ge many) se o 40 ° C and a mobile phase consis ing o me h- anol (15% / ) and ace a e buffe (0.05 M, pH 5) a a flow a e o 1 mL/min. S anda ds dilu ed o physiological concen a- ions (62.5–250 μM o cys eine and 500–2000 μM o glu a- hione) and ea ed as a sample we e used o quan ifica ion. 2.3. De e mina ion o Asco bic Acid and U ic Acid in Plasma. Plasma asco bic acid and u ic acid we e analyzed by RP-HPLC and UV de ec ion a e educ ion wi h is- (2-ca boxye hyl)-phosphine [18]. B iefly, plasma (100 μL) was mixed wi h is-(2-ca boxye hyl)-phosphine (20% w/w, 25 μL) and incuba ed o 5 min on ice; hen, eshly p epa ed me aphospho ic acid solu ion (10% w/w,75μL) was added and o ex-mixed; and samples we e cen i uged a 19,500 ×g and 4 ° C o 10 min. Twen y μL o clea supe na an was analyzed on a Shimadzu P ominence HPLC and using a 5μm analy ical column (Rep osil-Pu 120 C18 AQ, 250 mm ×4.6 mm; D . Maisch, Ge many) se o 40 ° C, a mobile phase consis ing o 0.05 M sodium phospha e buffe (pH 2.5) a a flow a e o 1 mL/min and UV-Vis de ec o (SPD-20AV) se o 245 nm. Pu e s anda ds dilu ed o physiological concen a ions (2.5–20 mg/L o asco bic acid and 20–100 mg/L o u ic acid) and ea ed as a sample we e used o quan ifica ion. 2.4. Malondialdehyde. Plasma malondialdehyde was de e - mined by RP-HPLC coupled wi h fluo escence de ec ion a e de i a iza ion wi h hioba bi u ic acid as desc ibed by Wong e al. [19] wi h modifica ions [20]. 2.5. Analysis o P o ein Ca bonyls and 3-Ni o y osine. The analyses o p o ein ca bonyls [21] and 3-ni o y osine in plasma by in-house ELISA ha e been desc ibed elsewhe e [20]. 2.6. Analysis o α-Tocophe ol and Lycopene. Plasma lycopene and α- ocophe ol we e analyzed by RP-HPLC coupled wi h UV- is and fluo escence de ec ion as p e iously desc ibed [22]. 2.7. S a is ical Analysis. Demog aphic cha ac e is ics a e desc ibed by using means ±s anda d de ia ion (SD) o con inuous a iables (age, weigh , and BMI) and equencies (%) o ca ego ical a iables (gende , smoking s a us, age g oups, and coun y). Diffe ences in cha ac e is ics be ween age g oups and s udy g oups we e compa ed by one-way ANOVA (con inuous a iables) wi h Tukey’s pos hoc es and Pea son’s chi-squa ed es (p e alence; o ca ego ical a iables). Da a o plasma bioma ke s we e ans o med app op ia ely o achie e no mal dis ibu ion using squa e oo (SR) o loga i hmic (LN) ans o ma ion and a e desc ibed by geome ic means wi h 95% confidence in e als (CI). Co ela ions among bioma ke s and be ween bio- ma ke s and age a e shown as Pea son p oduc -momen co ela ion o ans o med da a. Mean alues o plasma bio- ma ke s be ween s udy g oups we e compa ed using one- way ANOVA and gene al linea models wi h Fishe ’s leas significan diffe ence es . The models we e adjus ed o age, BMI, gende , smoking s a us (co a ia es), and coun y ( ac o ). In addi ion, a mul iple linea eg ession analysis wi h all bioma ke s in he ini ial model and a o wa d s ep- wise app oach was applied o iden i y independen plasma bioma ke s wi h he highes co ela ion wi h age. Diffe ences o concen a ions in bioma ke s be ween RASIG, GO, SGO, and age g oups (5-yea in e als) a e p esen ed as box plo s. All s a is ical analyses we e ca ied ou using SPSS so wa e (SPSS Inc., Chicago, IL; Ve sion 19); s a is ical significance o all es s was conside ed a P<005. 3. Resul s Cha ac e is ics o he s udy g oups a e shown in Table 1. The mean age o pa icipan s was 64.3 ±5.4 (55–75) yea s wi h no significan diffe ence be ween he ma ched s udy g oups. Men ep esen ed 47.7% o pa icipan s. The mean BMI was 27.3 ±4.5 kg/m 2 wi h no s a is ical significan diffe ence among he s udy g oups. Only 14.8% o pa icipan s we e cu en smoke s. The p e alence o smoking was diffe en among he h ee s udy g oups being 9.9% in SGO, 12.6% in GO, and 17.9% in RASIG. The numbe o pa icipan s om each s udy cen e was significan ly diffe en as shown in Table 1. A o al o 1559 pa icipan s we e dis ibu ed in o he g oups as ollows: RASIG (n= 794), GO (n= 493), and SGO (n= 272). Bioma ke concen a ions diffe ed significan ly be ween s udy g oups (Table 2 and Figu es 1, 2, and 3). I is no ewo - hy ha GO and SGO diffe ed only in u ic acid wi h signifi- can ly highe concen a ions in he SGO g oup. Cys eine was highes in GO, glu a hione in SGO, α- ocophe ol was highes in bo h GO and SGO while lycopene was lowe in GO and SGO han in he RASIG g oup. In e es ingly, GO diffe ed in all bioma ke s om he RASIG g oup excep in u ic acid and o al glu a hione (Table 2). In de ail, GO had significan ly highe concen a ions o asco bic acid, o al ee cys eine, α- ocophe ol, and 3-ni o y osine and lowe con- cen a ions o lycopene, p o ein ca bonyls, and malondialde- hyde. Fu he mo e, SGO we e diffe en om RASIG in all bioma ke s excep o u ic acid, 3-ni o y osine, and malon- dialdehyde wi h highe concen a ions o asco bic acid, o al ee cys eine, o al glu a hione, and α- ocophe ol and lowe concen a ions o p o ein ca bonyls and lycopene. We pe o med a uni a ia e gene al linea model adjus ed o age, BMI, gende , smoking s a us, and coun y o assess whe he hese diffe ences we e s ill p esen a e adjus men . This was ue o glu a hione and α- ocophe ol, as well as o all h ee oxida i e s ess bioma ke s (p o ein ca bonyls, 3- 3Oxida i e Medicine and Cellula Longe i y ni o y osine, and malondialdehyde), al hough lycopene (P<0069) and cys eine (P<0072) also eached he bo de line o significance. The compa ison o bioma ke concen a ions o ou s udy g oups wi h a middle-aged con ol g oup is shown in Table 3. In gene al, he olde g oups e ealed significan ly Table 2: Bioma ke concen a ions in he h ee s udy popula ions 1 . All (n= 1559) RASIG (n= 794)GO(n= 493) SGO (n= 272)P An ioxidan s Asco bic acid (mg/L) 4.41 (4.26; 4.58) 4.13 (3.91; 4.36) 4.73 (4.44; 5.02) ‡ 4.70 (4.33; 5.08) ‡ 0.002 Adjus ed GLM 2 4.32 (4.13; 4.52) 4.17 (3.87; 4.49) 3.85 (3.32; 4.41) 0.258 U ic acid (mg/L) 45.9 (45.3; 46.5) 45.7 (44.9; 46.5) 45.3 (44.2; 46.3) 47.4 (46.0; 48.8) ◊ 0.047 Adjus ed GLM 2 45.4 (44.6; 46.1) 46.0 (44.7; 47.2) 47.0 (44.7; 49.3) 0.341 To al cys eine (μmol/L) 148.2 (146.6; 149.8) 144.8 (142.5; 147.1) 150.8 (147.9; 153.6) ‡ 153.3 (149.3; 157.3) ‡ <0.001 Adjus ed GLM 2 144.2 (141.8; 146.6) 149.6 (145.7; 153.4) ‡ 145.5 (138.4; 152.6) 0.072 To al glu a hione (μmol/L) 1100 (1091; 1110) 1088 (1074; 1102) 1110 (1093; 1127) 1120 (1099; 1141) ‡ 0.024 Adjus ed GLM 2 1092 (1077; 1106) 1112 (1089; 1135) 1149 (1106; 1193) ‡ 0.025 Lycopene (μmol/L) 0.574 (0.556; 0.592) 0.631 (0.604; 0.659) 0.525 (0.497; 0.554) ‡ 0.503 (0.464; 0.543) ‡ <0.001 Adjus ed GLM 2 0.608 (0.583; 0.635) 0.566 (0.526; 0.606) 0.539 (0.470; 0.613) 0.069 α-Tocophe ol (μmol/L) 28.8 (28.4; 29.2) 27.8 (27.3; 28.3) 30.1 (29.4; 30.8) ‡ 29.6 (28.6; 30.5) ‡ <0.001 Adjus ed GLM 2 27.9 (27.4; 28.5) 29.3 (28.4; 30.3) ‡ 29.9 (28.2; 31.7) ‡ 0.005 Oxida i e s ess bioma ke s P o ein Ca bonyls (nmol/mg) 0.577 (0.573; 0.582) 0.595 (0.590; 0.601) 0.558 (0.551; 0.566) ‡ 0.561 (0.552; 0.571) ‡ <0.001 Adjus ed GLM 2 0.591 (0.585; 0.597) 0.566 (0.557; 0.575) ‡ 0.564 (0.547; 0.581) ‡ <0.001 3-Ni o y osine (pmol/mg) 4.11 (3.97; 4.25) 3.94 (3.75; 4.13) 4.40 (4.14; 4.67) ‡ 4.10 (3.77; 4.44) 0.018 Adjus ed GLM 2 3.90 (3.70; 4.11) 4.38 (4.03; 4.74) ‡ 3.76 (3.18; 4.38) 0.048 Malondialdehyde (μmol/L) 0.316 (0.306; 0.326) 0.336 (0.322; 0.350) 0.284 (0.266; 0.302) ‡ 0.317 (0.290; 0.346) <0.001 Adjus ed GLM 2 0.334 (0.319; 0.348) 0.288 (0.267; 0.311) ‡ 0.314 (0.274; 0.357) 0.004 1 Geome ic mean (95% CI). 2 Adjus ed gene al linea model (GLM): uni a ia e gene al linea model adjus ed o age, BMI, smoking s a us, gende , and coun y (cen e ). ‡ S a is ically significan diffe ence o RASIG. ◊ S a is ically significan diffe ence o GO. P alues: s a is ically significan diffe ences we e de e mined by one-way ANOVA wi h Tukey’s pos hoc es and by Fishe ’s leas significan diffe ence pos hoc es in he GLM. Table 1: Cha ac e is ics o he h ee s udy g oups. All (n= 1559) RASIG (n= 794)GO(n= 493) SGO (n= 272)P Age (yea s) 64.3 ±5.4 64.5 ±5.8 64.3 ±4.9 63.9 ±4.7 0.230 55–59 yea s (n, (%)) 345 (22.1) 193 (24.3) 94 (19.1) 58 (21.3) <0.001 60–64 yea s (n, (%)) 448 (28.7) 210 (26.4) 149 (30.2) 89 (32.7) 65–69 yea s (n, (%)) 465 (29.8) 204 (25.7) 166 (33.7) 95 (34.9) 70–75 yea s (n, (%)) 301 (19.3) 187 (23.6) 84 (17.0) 30 (11.0) Gende , male (n, (%)) 743 (47.7) 386 (48.6) 213 (43.2) 144 (52.9) 0.027 Smoke , cu en (n, (%)) 231 (14.8) 142 (17.9) 63 (12.6) 27 (9.9) 0.001 BMI (kg/m 2 )27.3 ±4.5 27.6 ±4.7 26.6 ±4.4 27.3 ±4.2 0.692 <25 (n, (%)) 516 (33.1) 249 (31.4) 188 (38.1) 79 (29.0) 0.03625 o <30 (n, (%)) 673 (43.2) 343 (43.3) 199 (40.4) 131 (48.2) ≥30 (n, (%)) 369 (23.7) 201 (25.3) 106 (21.5) 62 (22.8) Coun y Belgium (n, (%)) 472 (30.3) 155 (19.5) 190 (38.5) 127 (46.7) <0.001 Finland (n, (%)) 253 (16.2) 69 (8.7) 132 (26.8) 52 (19.1) G eece (n, (%)) 209 (13.4) 187 (23.6) 18 (3.7) 4 (1.5) I aly (n, (%)) 323 (20.7) 187 (23.6) 87 (17.6) 49 (18.0) Poland (n, (%)) 302 (19.4) 196 (24.7) 66 (13.4) 40 (14.7) Values a e means ±SD; P alue: one-way ANOVA (con inuous a iables) and Pea son’s chi-squa ed es (p e alence). 4 Oxida i e Medicine and Cellula Longe i y diffe en an ioxidan concen a ions compa ed o he e e - ence g oup. The RASIG, GO, and SGO g oups had signifi- can ly highe mean u ic acid, cys eine, and α- ocophe ol and lowe lycopene compa ed o he e e ence g oup. The e we e no diffe ences in e ms o oxida i e s ess bioma ke s be ween he e e ence g oup and RASIG aged 55–75 yea s. In e es - ingly, while he GO and SGO g oups had significan ly lowe p o ein ca bonyl concen a ions compa ed o he e e ence g oup, only he GO g oup had highe 3-ni o y osine compa ed o he e e ence g oup. Table 4 shows he co ela ion coefficien s among he assessed bioma ke s in all pa icipan s aged ≥55 yea s. The highes co ela ion coefficien s we e ound be ween malondialdehyde and p o ein ca bonyls ( =0322), malon- dialdehyde and asco bic acid ( =−0240), ollowed by p o ein ca bonyls and α- ocophe ol ( =−0193). A posi i e co ela ion was obse ed be ween he an ioxidan s asco bic acid and α- ocophe ol ( =0164). When hese co ela ions we e pe o med only in he RASIG g oup, he di ec ion and s eng h o co ela ions we e simila ( esul s no shown). The co ela ions o he indi idual bioma ke s wi h age a e shown in Table 5. Significan posi i e co ela ion coefficien s we e obse ed o u ic acid, cys eine, and 3-ni o y osine while a significan in e se associa ion was only seen o lycopene. A weak significan posi i e co ela ion be ween p o ein ca bonyls and age was only obse ed in he RASIG g oup ( =0098;P<001, da a no shown). In a final mul iple eg ession model wi h a o wa d app oach, we aimed o iden i y independen bioma ke s wi h he highes co ela ion wi h age (Table 6). The e o e, all bioma ke s we e assessed as co a ia es. Confi ming he co ela ions om Table 5, u ic acid, cys eine, 3-ni o y osine, P o ein ca bonyls (nmol/mg) (A) (B) RASIG SGOGO RASIG SGOGO 0.6 0.8 0.4 0.7 0.9 0.5 55‒59 60‒64 65‒69 70‒75 ⁎⁎⁎ ⁎ (a) RASIG SGOGO 3-Ni o y osine (pmol/mg) 14 8 4 2 0 10 6 12 RASIG SGOGO (A) (B) 55‒59 60‒64 65‒69 70‒75 ⁎⁎ (b) RASIG SGOGO Malondialdehyde (mol/L) 0.8 0.4 0.2 0 1.0 0.6 1.2 RASIG SGOGO (A) (B) 55‒59 60‒64 65‒69 70‒75 ⁎⁎⁎ (c) Figu e 1: Classical bioma ke s o oxida i e s ess: p o ein ca bonyls (a), 3-ni o y osine (b), and malondialdehyde (c). Bioma ke concen a ions a e displayed acco ding o s udy g oups (A) and age g oups (B), espec i ely. RASIG (n= 794); GO (n= 493); SGO (n= 272). Ou lie s and ex eme alues a e included in he analyses bu no shown in he figu e. S a is ically significan diffe ences a e indica ed by as e isks: ∗P<005 and ∗∗∗P<0001. 5Oxida i e Medicine and Cellula Longe i y and lycopene we e significan ly associa ed wi h age, and also, p o ein ca bonyls emained in he model wi h a posi i e associa ion. 4. Discussion In gene al, he olde g oups e ealed highe an ioxidan (u ic acid, cys eine, and α- ocophe ol) and lowe lycopene concen- a ions compa ed o he e e ence g oup. Lycopene, o al cys eine, u ic acid, p o ein ca bonyls, and 3-ni o y osine we e significan ly and independen ly associa ed wi h age in he mul iple linea eg ession model among pa icipan s aged 55–75 yea s. Highe cys eine and α- ocophe ol, bu lowe lycopene in bo h GO and SGO compa ed o RASIG seem o be associa ed wi h a “beneficial”li es yle, while he signi - ican ly lowe malondialdehyde and highe 3-ni o y osine only in he GO compa ed o he RASIG g oup may indica e ha amilies wi h long-li ing membe s a e gene ically be e equipped o handle oxida i e s ess, ye he cause and impac o he highe 3-ni o y osine le els emain unclea . We assume ha (1) i he e a e diffe ences be ween RASIG and he o he wo g oups bu no diffe ences be ween GO and SGO, he eason may be li es yle- ela ed and (2) di - e ences be ween GO and SGO (i espec i e o he RASIG esul s) indica e gene ic easons. Applying hese c i e ia o he bioma ke s assessed in he p esen s udy leads o he conclusion ha a e adjus men o age, BMI, smoking s a us, gende , and coun y, he diffe - ences in p o ein ca bonyls, malondialdehyde, 3-ni o y o- sine, α- ocophe ol, cys eine, and glu a hione be ween s udy g oups seem o be li es yle- ela ed whe eas gene ics seem o play a mino ole. Fo he case whe e bo h GO and SGO diffe o he con ol g oup (p o ein ca bonyls, lycopene, and α- ocophe ol), his migh indica e he influence o li e- s yle. Thus, ou esul s show ha li es yle is an impo an con ibu o o edox bioma ke s. One migh hypo hesize ha when GO diffe o he o he wo g oups and he e e ence g oup, his diffe ence may be a ibu ed o gene ics. This was he case o 3-ni o y osine, cys eine (bo h highe han he o he g oups), and malondial- dehyde (lowe han he o he g oups). In ou s udy, he e was no significan diffe ence be ween GO and SGO g oups o hese h ee bioma ke s, ye a clea endency exis s acco ding o he da a in Tables 2 and 3. Concen a ions o malondialdehyde, 3-ni o y osine, and o al cys eine we e s ill diffe en be ween he GO and he RASIG g oups a e he GLM adjus men , which may be an indica ion o a gene ic con ibu ion in age-associa ed handling o oxida i e s ess. Di ec compa ison o he s udy g oups e ealed ha he GO g oup had significan ly lowe concen a ions o p o ein ca bonyls and malondialdehyde accompanied by highe concen a ions o cys eine, asco bic acid, and α- ocophe ol compa ed o RASIG. Con a ily, GO had lowe lycopene and highe 3-ni o y osine han he RASIG g oup. These di - e ences may be due o be e me abolic p ofiles o due o a gene ally heal hie nu ien in ake, despi e less p ocessed oma o p oduc s which a e especially ich in lycopene. I is widely accep ed ha he e is a ela ionship be ween he aging p ocess and oxida i e s ess; howe e , mos s udies leading o his heo y ha e been ca ied ou in model sys ems and only ew s udies ha e analyzed di - e en bioma ke s o oxida i e s ess in heal hy humans o a ious age g oups [3, 6]. P o ein ca bonyls a e conside ed o be ela i ely s able [23] and ea ly ma ke s [24] o oxida i e s ess. Measu ing ca bonyla ion o plasma p o eins enables e alua ing he global oxida ion s a us in plasma. P o ein ca bonyls ha e been analyzed in a mul i ude o s udies anging om cell cul- u e, animal, o human s udies and a e ela ed o he aging RASIG SGOGO Lycopene (mol/L) 1.5 0.5 0 2.0 1.0 RASIG SGOGO (a) S udy g oups (b) Age g oups 55‒59 60‒64 65‒69 70‒75 ⁎⁎⁎ ⁎⁎⁎ ⁎⁎⁎ ⁎⁎ ⁎⁎ ⁎ Figu e 3: Lycopene concen a ion by s udy g oups (a) and age g oups (b). RASIG (n= 794); GO (n= 493); SGO (n= 272). Ou lie s and ex eme alues a e included in he analyses bu no shown in he figu e. S a is ically significan diffe ences a e indica ed by as e isks: ∗P<005,∗∗P<001, and ∗∗∗P<0001. RASIG SGOGO Cys eine (mol/L) 300 200 100 50 250 150 RASIG SGOGO (a) S udy g oups (b) Age g oups 55‒59 60‒64 65‒69 70‒75 ⁎⁎⁎ ⁎⁎ ⁎⁎ ⁎⁎ Figu e 2: Cys eine concen a ion by s udy g oups (a) and age g oups (b). RASIG (n= 794); GO (n= 493); SGO (n= 272). Ou lie s and ex eme alues a e included in he analyses bu no shown in he figu e. S a is ically significan diffe ences a e indica ed by as e isks: ∗∗P<001 and ∗∗∗P<0001. 6 Oxida i e Medicine and Cellula Longe i y p ocess. They ha e been shown o be a p edic o o mo ali y in mode a ely o se e ely disabled olde women [10]. Komosinska-Vasse e al. showed a posi i e co ela ion o p o ein ca bonyls wi h age in a s udy o 56 men and women aged ≥55 yea s ( =052) [25]. O he au ho s also ound a s ong posi i e co ela ion o p o ein ca bonyls wi h age ( =0786) in 80 pa icipan s be ween he age o 18 and 85 [6], whe eas in he p esen s udy, he co ela ion o he age ange o 55–75 was e y weak ( =0098) bu s ill s a is ically significan . Caka ay e al. demons a ed ha plasma p o ein ca bonyl le els o elde ly pa icipan s we e significan ly highe in compa ison o hose o middle-aged and young pa icipan s [26]. In he p esen s udy, p o- ein ca bonyls only co ela ed wi h age in he RASIG g oup ( =0098,P<001; da a no shown) bu no when all h ee s udy g oups we e combined. Mos in e es ingly, he RASIG g oup had s a is ically significan ly highe p o ein ca bonyl concen a ions han he GO as well as he SGO while he e was no diffe ence be ween he GO and SGO. A e adjus ing o co a ia es (including smoking s a us), his diffe ence emained. Fo 3-ni o y osine, we obse ed a weak, posi i e associa ion wi h age. In con as o p o ein ca bonyls, 3- ni o y osine was significan ly highe in GO compa ed o RASIG. 3-Ni o y osine has been desc ibed o be a s able ma ke o oxida i e/ni a i e s ess in some inflamma o y diseases [27]. I a ises om ni a ion in ol ing eac i e Table 4: Co ela ions be ween bioma ke s in all h ee s udy g oups in pa icipan s aged ≥55 yea s 1 . Asco bic acid U ic acid To al cys eine To al glu a hione Lycopene α-Tocophe ol P o ein ca bonyls 3-Ni o- y osine Malondialdehyde Asco bic acid −0.083∗∗∗ 0.140∗∗∗ 0.030 −0.103∗∗∗ 0.164∗∗∗ −0.157∗∗∗ −0.027 −0.240∗∗∗ U ic acid 0.067∗∗ −0.021 −0.099∗∗∗ 0.059∗−0.020 −0.054∗0.019 To al cys eine 0.047 −0.053∗0.103∗∗∗ 0.062∗0.043 −0.034 To al glu a hione 0.032 −0.033 −0.007 −0.015 0.030 Lycopene 0.073∗∗ 0.088∗∗∗ −0.092∗∗∗ 0.159∗∗∗ α-Tocophe ol −0.193∗∗∗ 0.062∗ −0.132∗∗∗ P o ein ca bonyls 0.001 0.322∗∗∗ 3-Ni o y osine 0.019 Malondialdehyde 1 Pea son co ela ion coefficien . S a is ically significan co ela ions a e ma ked by ∗P<005,∗∗ P<001, and ∗∗∗ P<0001.(n= 1559). Table 3: Compa ison o s udy g oups wi h a e e ence g oup. Re e ence G oup (RASIG; 35–54 yea s) (n= 683) RASIG (55–75 yea s) (n= 794) GO (55–75 yea s) (n= 493) SGO (55–75 yea s) (n= 272) An ioxidan s Asco bic acid (mg/L) 4.37 (4.12; 4.63) 4.13 (3.91; 4.36) 4.73 (4.44; 5.02) 4.70 (4.33; 5.08) U ic acid (mg/L) 42.4 (41.6; 43.3) 45.7 (44.9; 46.5)∗∗∗ 45.3 (44.2; 46.3)∗∗∗ 47.4 (46.0; 48.8)∗∗∗ To al cys eine (μmol/L) 130.3 (128.1; 132.5) 143.0 (140.7; 145.3)∗∗∗ 149.0 (146.1; 151.9)∗∗∗ 151.4 (147.4; 155.5)∗∗∗ To al glu a hione (μmol/L) 1107 (1092; 1121) 1088 (1074; 1101) 1110 (1093; 1127) 1120 (1099; 1142) Lycopene (μmol/L) 0.83 (0.80; 0.86) 0.63 (0.60; 0.66)∗∗∗ 0.53 (0.50; 0.55)∗∗∗ 0.50 (0.46; 0.54)∗∗∗ α-Tocophe ol (μmol/L) 25.7 (10.4; 10.6) 27.8 (10.9; 11.2)∗∗∗ 30.1 (11.4; 11.7)∗∗∗ 29.6 (11.2; 11.7)∗∗∗ Oxida i e s ess bioma ke s P o ein ca bonyls (nmol/mg) 0.603 (0.598; 0.609) 0.595 (0.590; 0.601) 0.558 (0.551; 0.566)∗∗∗ 0.561 (0.552; 0.571)∗∗∗ 3-Ni o y osine (pmol/mg) 3.8 (3.6; 4.0) 3.9 (3.8; 4.1) 4.4 (4.1; 4.7)∗∗ 4.1 (3.8; 4.4) Malondialdehyde (μmol/L) 0.31 (0.30; 0.32) 0.34 (0.32; 0.35) 0.28 (0.27; 0.30) 0.32 (0.29; 0.35) ∗∗∗P<0001 and ∗∗ P<001 by one-way ANOVA wi h Tukey’s pos hoc es . Table 5: Co ela ions be ween bioma ke s and age among all pa icipan s (aged ≥55 yea s) 1 . P Asco bic acid 0.026 0.297 U ic acid 0.092 <0.001 To al cys eine 0.152 <0.001 To al glu a hione −0.035 0.167 Lycopene −0.224 <0.001 α-Tocophe ol −0.005 0.847 P o ein ca bonyls 0.036 0.158 3-Ni o y osine 0.066 0.009 Malondialdehyde −0.037 0.146 1 Pea son co ela ion coefficien (n= 1559). 7Oxida i e Medicine and Cellula Longe i y ni ogen species (RNS) and pe oxidase-media ed ni i e oxida ion. Today, myelope oxidase is conside ed o be in ol ed in his pa hway. Thus, 3-ni o y osine may play a ole in inflamma ion a he han he aging p ocess i sel [28]. F ijhoffe al. ques ion whe he 3-ni o y osine is clini- cally use ul, in compa ison o he al eady es ablished ma ke s o inflamma o y p ocesses such as C- eac i e p o ein [29]. Plasma 3-ni o y osine le els in pa ien s ea ed wi h an i- inflamma o y d ugs ha e been shown o dec ease [30, 31]. This is one mo e hin why 3-ni o y osine may be a be e ma ke o inflamma ion han o he aging p ocess in gen- e al. Al hough he ni a ion o y osine esidues on p o eins can esul in a loss o unc ion [32], some au ho s ha e epo ed a gain o unc ion [33]. Pe haps si e-specific ni a- ion may ha e a p o ec i e unc ion o a ole in longe i y. In ou s udy popula ion, he e migh ha e been no d ama ic diffe ence o 3-ni o y osine in he diffe en age g oups because he pa icipan s we e gene ally heal hy. Only 4.0% o ou pa icipan s had CRP concen a ions ≥10 mg/L. Ne e heless, we checked he co ela ion be ween CRP and 3-ni o y osine bu he e was no co ela ion, nei he in all pa icipan s no in hose wi h a high CRP le el. Thus, he use ulness o 3-ni o y osine as a bioma ke o e alua e oxida i e s ess emains o be elucida ed. Few la ge epidemiological s udies ha e analyzed malon- dialdehyde. Block e al. sugges plasma malondialdehyde should be conside ed o mos epidemiologic esea ch on edox bioma ke s [34] since hey obse ed in a alida ion s udy ha his bioma ke had a good day- o-day s abili y. They p opose malondialdehyde o be an effec i e ma ke o oxida i e s ess and s a e ha he use o a single measu e o malondialdehyde esul ed in li le a enua ion [34]. In ano he s udy, he same au ho s ound ha malondialdehyde was no associa ed wi h age in 298 pa icipan s aged 18–78 yea s [35]. These esul s a e suppo ed by ou finding ha malondialdehyde was no associa ed wi h age in any o he h ee s udy g oups. Howe e , he GO g oup had significan ly lowe plasma malondialdehyde concen a ions han RASIG. This diffe ence emained e en a e adjus ing o age, BMI, smoking s a us, gende , and coun y. To coun e ac oxida i e s ess, he e exis a ne wo k o an ioxidan de ense mechanisms. One o hese an ioxi- dan s is he ipep ide glu a hione (γ-glu amylcys einylgly- cine). A decline o glu a hione wi h age has p e iously been sugges ed in humans [36], and i has been demons a ed ha he co ela ion o hiol g oups in plasma wi h age was in e se ( =−0718) [6]. Concen a ions o glu a hione and cys eine we e lowe in heal hy old (mean age 70.3 yea s) han in middle-aged (mean age 39.8 yea s) pa icipan s (n=8each) [37]. We did no obse e any co ela ion be ween glu a hione and age. One explana ion may be ha we measu ed o al glu a hione ins ead o GSH/GSSH which is conside ed a be - e ma ke o he edox s a e. Gius a ini e al. ound an in e se co ela ion o glu a hione wi h age bu no co ela ion o cys eine wi h age in 41 pa icipan s [38]. Simila ly, Jones e al. used plasma o 122 heal hy indi iduals aged 19–85 yea s o analyze hiol-based edox changes [39]. They sugges ha he capaci y o he glu a hione an ioxidan sys em is main- ained un il 45 yea s and hen declines apidly. Fo he p esen analyses, pa icipan s we e selec ed which we e ≥55 yea s, ye we we e unable o show a co ela ion be ween glu- a hione and age. In con as o p e ious findings, cys eine is highe in he highe age g oups in he RASIG and GO s udy g oups and co ela es posi i ely wi h age among all pa ici- pan s ( =0152,P<0001). I is assumed ha lowe glu a hione concen a ions occu ing du ing aging and in diffe en diseases may mo e likely esul om low cys eine concen a ions a he han due o oxida ion since cys eine is he a e-limi ing p ecu so o glu a hione. F ee cys eine is one o he main nonp o ein hiols in plasma [40, 41] and conside ed a semiessen ial amino acid, since i mus be aken up o syn hesized om he essen ial amino acid me hionine [42]. I is able o egula e nu ien me abolism, oxida i e s ess, physiologic signaling pa hways, and associa ed diseases h ough he p oduc ion o glu a hione, hyd ogen sulfide, and au ine [43]. An o al in e en ion wi h cys eine (as N-ace ylcys eine) and glycine o 14 days esul ed in significan ly inc eased glu a hione concen a ions [37]. This was also ue when only cys eine was supplemen ed, esul ing in an inc ease in hepa ic glu a hione syn hesis [43]. I is likely ha he glu a hione concen a ions in ou c oss-sec ional s udy we e simila in he diffe en age g oups because cys eine may no ha e been limi ed. P e ious esul s also sugges ha he pa icipan s in ou s udy had sufficien cys eine and hus glu a hione le els [38]. Ne e heless, he equi emen o cys eine may be ele a ed a e oxida i e e en s due o he consump ion o glu- a hione. Some u u e esea ch should cla i y he ole o cys- eine in aging and as a die a y p ecu so o glu a hione. Table 6: Associa ions o oxida i e s ess ma ke s and an ioxidan s wi h age 1 . Compound B95% CI 2 P (Cons an ) 59.13 56.79, 61.48 <0.001 Lycopene (μmol/L) −2.783 −3.440, −2.126 −0.207 0.047 <0.001 To al cys eine (mol/L) 0.021 0.013, 0.029 0.133 0.020 <0.001 U ic acid (mg/L) 0.031 0.009, 0.054 0.071 0.004 0.005 P o ein ca bonyls (nmol/mg) 3.274 0.252, 6.296 0.054 0.003 0.034 3-Ni o y osine (pmol/mg) 0.094 0.004, 0.183 0.052 0.003 0.040 1 Mul iple linea eg ession analysis wi h a o wa d app oach o iden i y independen blood bioma ke s wi h highes co ela ion o age; all bioma ke s including asco bic acid, glu a hione, malondialdehyde, and α- ocophe ol we e assessed as co a ia es in he ini ial model; pa ial and 2 ;R=0277, R2=0077 (n= 1545). 8 Oxida i e Medicine and Cellula Longe i y Fu he mo e, i should be no ed ha some au ho s measu ed glu a hione/ hiol g oups in plasma [6]. Glu a hione is known as he mos powe ul cellula an ioxidan , ha is, p esen in ed blood cells, and consequen ly, we analyzed glu a hione in whole blood. Conce ning u ic acid, ou esul s show ha he e was no diffe ence be ween RASIG and GO bu highe concen a ions in he SGO g oup han in he GO. Howe e , adjus ing o age, BMI, gende , smoking s a us, and coun y, hese diffe - ences did no emain. U ic acid is de i ed om he deg ada- ion o pu ine nucleo ides which can be o die a y o endogenous o igin. The e o e, an inc eased in ake o animal p oduc s o legumes/pulses in he SGO g oup is possible. The e is a con o e sial discussion whe he he posi i e effec s o u ic acid as an an ioxidan a e ou weighed by i s ad e se effec s conce ning gou , co ona y a e y disease, hype ension, and s oke ( o e iew, see [44]). Besides endogenous an ioxidan s such as glu a hione, cys eine, and u ic acid, exogenous an ioxidan s a e also equi ed o coun e ac oxida i e s ess. Some o he mos powe ul an ioxidan s a e o die a y o igin such as asco bic acid, α- ocophe ol, lycopene, and o he ca o enoids. I is known ha a high in ake o ui s and ege ables is associa ed wi h a high plasma concen a ion o asco bic acid [45]. Simul aneously, a die ich in ui s and ege ables is associ- a ed wi h a educed isk o some diseases such as CVD and cance s [46]. α-Tocophe ol and asco bic acid ac syne gis i- cally in coun e ac ing ee adicals. α-Tocophe ol is able o quench ee adicals in a hyd ophobic en i onmen , o example, o e mina e lipid pe oxida ion and he esul ing ocophe ol adical is hen ecycled by asco bic acid [47]. We ound an in e se co ela ion be ween α- ocophe ol and malondialdehyde, an in e se associa ion be ween asco bic acid and malondialdehyde (all P<0001), and a posi i e co ela ion be ween α- ocophe ol and asco bic acid. These esul s a e in acco dance wi h he assump ion ha hese an ioxidan s ac syne gis ically. Since hese a e die -de i ed an ioxidan s and we did no adjus ou models o die a y in ake o ui and ege ables, we canno exclude ha he die o he season had an effec on he plasma concen a ion o hese mic onu ien s. In a ecen s udy, we ha e shown o e he whole age- ange o he MARK-AGE p ojec om 35 o 75 (RASIG g oup) ha lycopene was lowe in highe age g oups and his effec was independen o season [22]. I emains unclea i hese diffe - ences obse ed in he olde age g oups esul om a educed in ake, educed abso p ion, inc eased s o age in adipose issue, o ele a ed deg ada ion o lycopene. The s eng hs o he p esen s udy include he la ge sample size and ha all bioma ke s desc ibed he e we e measu ed in one single labo a o y in blinded o m (samples om s udy g oups and coun ies we e in mixed andom o de by he biobank). Fo lipid-soluble mic onu ien anal- ysis, we ha e ecen ly published he in e ba ch coefficien s o a ia ion which we e 7.6% o lycopene and 6.3% o α- ocophe ol in he MARK-AGE coho [22]. Conce ning he alidi y o he me hods used o 3-ni o y osine analysis, we ha e p e iously epo ed good specifici y, ep oducibili y, and accu acy o his in-house ELISA [48]. The ELISA me hod used he e o analyze p o ein ca bonyls has been al- ida ed in a mul icen e ing s udy [49]. I was shown ha ca bonyl concen a ions om h ee ou o ou labo a o ies pa icipa ing in he ing s udy ell wi hin 95% confidence in e als. Addi ionally, p o ein concen a ion was measu ed be o e bo h ELISAs and plasma samples we e dilu ed o he same p o ein concen a ion. S anda ds o p o ein ca bonyl ELISA we e p epa ed acco ding o Buss e al. [21] and un on e e y 96-well pla e. In e ms o compa abili y o o he s udies, he e is an excellen e iew om Gius a ini e al. [50]. As he au ho s demons a e, he e exis nume ous analy ical me hods o analyze p o ein ca bonyls, malondialdehyde, GSH, ocoph- e ols, and lycopene, among o he s. An o e iew is gi en o diffe en de i a iza ion and de ec ion me hods, heal h condi- ions, and uni s, demons a ing huge a ie y. Fo ins ance o malondialdehyde, Gius a ini e al. showed ha au ho s using me hods compa able o ou s (TBA de i a iza ion and HPLC sepa a ion) published concen a ions be ween 0.44 and 6.8 μmol/L o plasma samples [50]. One eason o his a - ia ion may be ha hepa in plasma was used in some cases. Especially o malondialdehyde, i is impo an o measu e his ma ke in EDTA plasma since lipid pe oxida ion can con inue in se um and hepa in plasma, hus a ificially con- ibu ing o ele a ed malondialdehyde concen a ions. EDTA and ci a e can complex i on hus p e en ing Fen on eac ion leading o lipid pe oxida ion. Fo GSH and cys eine, he au ho s also epo diffe en me hods (e.g., enzyma ic, HPLC coupled wi h UV o fluo es- cence de ec ion) and mean concen a ions in whole blood, plasma, and e y h ocy es o heal hy and diseased indi iduals, which span o e one o wo o de s o magni ude wi hin he kind o used specimen; howe e , o al GSH le els in he p es- en s udy a e in he uppe ange o p e iously epo ed mean whole blood alues o heal hy indi iduals measu ed by HPLC o ecycling spec opho ome y (using Ellman’s eagen ). This s udy has some limi a ions which mus be men- ioned. Since his s udy was obse a ional, we canno make s a emen s on he changes o hese bioma ke s wi h age; he e o e s udies wi h epea ed measu emen s and/o ollow-up a e needed. Fu he mo e ou esul s may be spe- cific o Eu opean/Caucasian subjec s and hus no ans e - able o o he coun ies. The inclusion o h ee diffe en s udy g oups wi h a la ge sample size (n= 1559) is one ea u e ha dis inguishes his s udy om o he s. Samples we e collec ed and dis ibu ed in a blinded o m o gua an ee unbiased measu emen and in e p e a ion, and all analyses o bioma ke s desc ibed he e we e ca ied ou in one single labo a o y, which significan ly educes in e labo a o y a ia ions. The assessmen o diffe - en specific cellula and plasma bioma ke s, ha is, ma ke s o oxida i e damage oge he wi h an ioxidan s ha a e no analyzed by comme cial ki s bu by in-house me hods is ce ainly an impo an s eng h. 5. Conclusion He e, we ha e p o ided an o e iew o he le els o he diffe - en edox bioma ke s in human plasma and whole blood in 9Oxida i e Medicine and Cellula Longe i y