Circulating Ceramides Predict Cardiovascular Outcomes in the Population-Based FINRISK 2002 Cohort

2424 Se um lipids, including o al choles e ol, low-densi y lipo- p o ein-choles e ol (LDL-C) and high-densi y lipop o ein- choles e ol (HDL-C), ha e all been used in isk p edic ion o a he oscle o ic hea disease, and LDL-C has been es ablished as he main lipid a ge in he managemen o pa ien s wi h co ona y a e y disease (CAD). On he basis o b oad lipido- mic analyses in di e en CAD pa ien coho s, we and o h- e s ha e epo ed ha o he ci cula ing lipid species may also hold p ognos ic po en ial in CAD.1–4 In ou ea lie s udies, ce amide molecules Ce (d18:1/16:0), Ce (d18:1/18:0), and Ce (d18:1/24:1), as well as hei a ios o Ce (d18:1/24:0), ha e eme ged as no el isk s a i ie s in pa ien s wi h es ablished CAD. These ce amides a e ana- ly ically obus and ha e been p o en o p o ide signi ican inc emen al alue o e ou inely used lipid ma ke s in a sec- onda y p e en ion se ing, pa icula ly o ha d ou comes such as ca dio ascula dea h.5 Li e a u e on ce amide biology associa es hese molecules wi h many cen al a he oscle o ic p ocesses, such as lipop o ein agg ega ion, up ake o lipop o- eins and accumula ion o choles e ol wi hin mac ophages, egula ion o ni ic oxide syn hesis, p oduc ion o supe oxide anions, and exp ession o di e en cy okines.6–11 Fu he mo e, expe imen al s udies ha e shown ha ce amides and ela ed sphingolipids a e associa ed wi h he de elopmen o a h- e oscle osis.6 Inhibi ion o glycosphingolipid biosyn hesis dec eased a he oscle osis in mice, and consequen ly, se e al enzymes o he sphingolipid syn he ic pa hway ha e been es ed as po en ial d ug a ge s.12,13 As an ex ension o obse a ions in seconda y p e en ion coho s, he p esen s udy e alua ed ce amide beha io in he la ge-scale p ospec i e FINRISK s udy o alida e he p og- nos ic alue o he ce amides a he popula ion le el. A e ioscle Th omb Vasc Biol is a ailable a h p://a b.ahajou nals.o g DOI: 10.1161/ATVBAHA.116.307497 Objec i e—Ce amides a e molecula lipids implica ed in apop osis, in lamma ion, obesi y, and insulin esis ance. An ea lie s udy epo ed ha ce amides we e associa ed wi h a al ou come among pa ien s wi h co ona y hea disease. He e, we examined whe he ce amides a e associa ed wi h majo ad e se ca dio ascula e en s (MACEs) among appa en ly heal hy indi iduals. App oach and Resul s—FINRISK 2002 is a popula ion-based isk ac o su ey, which ec ui ed men and women aged 25 o 74 yea s. The coho was ollowed up un il he end o 2014. We quan i ied 4 ci cula ing ce amides, Ce (d18:1/16:0), Ce (d18:1/18:0), Ce (d18:1/24:0), and Ce (d18:1/24:1), in 8101 se um samples by a a ge ed liquid ch oma og aphy– andem mass spec ome y assay. P ima y ou come o in e es was inciden MACE (n=813). Seconda y analyses we e pe o med o MACE dea h (n=116) wi hou p e ious non a al MACE and o ecu en MACE (n=226) among su i o s o a p e ious inciden MACE. We used Cox p opo ional haza d models adjus ed o he F amingham co a ia es o de e mine he associa ion o ce amides wi h he ou comes. O he ce amide species, Ce (d18:1/18:0) had he s onges associa ion wi h inciden MACE and he highes unadjus ed haza d a io o 1.31 (95% con idence in e al, 1.21–1.41), which emained signi ican a 1.21 (95% con idence in e al, 1.11–1.33) a e F amingham isk ac o adjus men s. The haza d a ios we e gene ally s onge o ecu en and a al e en s han o i s e en s. Clinical ne eclassi ica ion imp o emen was 7.5% (P=6.9×10−5) o Ce (d18:1/18:0). Conclusions—Dis inc se um ce amides a e associa ed wi h he isk o inciden MACE in appa en ly heal hy indi iduals. These esul s should encou age mo e de ailed analyses o ce amides in ca dio ascula pa hobiology and sugges new bioma ke s o MACE isk. (A e ioscle Th omb Vasc Biol. 2016;36:2424-2430. DOI: 10.1161/ATVBAHA.116.307497.) Key Wo ds: ca dio ascula diseases ◼ ce amides ◼ lipids ◼ lipidomics ◼ me abolomics ◼ isk ac o s Recei ed on: Ma ch 7, 2016; inal e sion accep ed on: Oc obe 3, 2016. F om he Depa men o Heal h, Na ional Ins i u e o Heal h and Wel a e, Helsinki, Finland (A.S.H., V.S.); Zo a Biosciences Oy, Espoo, Finland (M.S.-A., M.H., D.K., R.H., K.E., R.L.); Medical School, Uni e si y o Tampe e, Finland (R.L.); and Finnish Clinical Biobank, Uni e si y Hospi al o Tampe e, Finland (R.L.). *These au ho s con ibu ed equally o his a icle. The online-only Da a Supplemen is a ailable wi h his a icle a h p://a b.ahajou nals.o g/lookup/suppl/doi:10.1161/ATVBAHA.116.307497/-/DC1. Co espondence o Reijo Laaksonen, MD, PhD, Zo a Biosciences Oy, Biologinkuja 1, FI-02150 Espoo, Finland. E-mail [email p o ec ed] © 2016 The Au ho s. A e ioscle osis, Th ombosis, and Vascula Biology is published on behal o he Ame ican Hea Associa ion, Inc., by Wol e s Kluwe Heal h, Inc. This is an open access a icle unde he e ms o he C ea i e Commons A ibu ion License, which pe mi s use, dis ibu ion, and ep oduc ion in any medium, p o ided ha he o iginal wo k is p ope ly ci ed. Ci cula ing Ce amides P edic Ca dio ascula Ou comes in he Popula ion-Based FINRISK 2002 Coho Aki S. Ha ulinna,* Ma ko Sysi-Aho,* Mika Hil o, Dimple Kauhanen, Reini Hu me, Kim Ek oos, Veikko Salomaa,* Reijo Laaksonen* by gues on Decembe 14, 2016h p://a b.ahajou nals.o g/Downloaded om by gues on Decembe 14, 2016h p://a b.ahajou nals.o g/Downloaded om by gues on Decembe 14, 2016h p://a b.ahajou nals.o g/Downloaded om by gues on Decembe 14, 2016h p://a b.ahajou nals.o g/Downloaded om by gues on Decembe 14, 2016h p://a b.ahajou nals.o g/Downloaded om by gues on Decembe 14, 2016h p://a b.ahajou nals.o g/Downloaded om by gues on Decembe 14, 2016h p://a b.ahajou nals.o g/Downloaded om by gues on Decembe 14, 2016h p://a b.ahajou nals.o g/Downloaded om by gues on Decembe 14, 2016h p://a b.ahajou nals.o g/Downloaded om by gues on Decembe 14, 2016h p://a b.ahajou nals.o g/Downloaded om by gues on Decembe 14, 2016h p://a b.ahajou nals.o g/Downloaded om by gues on Decembe 14, 2016h p://a b.ahajou nals.o g/Downloaded om by gues on Decembe 14, 2016h p://a b.ahajou nals.o g/Downloaded om by gues on Decembe 14, 2016h p://a b.ahajou nals.o g/Downloaded om by gues on Decembe 14, 2016h p://a b.ahajou nals.o g/Downloaded om Ha ulinna e al Ci cula ing Ce amides P edic CVD in FINRISK 2002 2425 Ma e ials and Me hods Ma e ials and Me hods a e a ailable in he online-only Da a Supplemen . Resul s Ele a ed Ce amide Concen a ions A e Associa ed Wi h Inc eased Majo Ad e se Ca dio ascula E en Risk This s udy included 8101 indi iduals om he FINRISK 2002 gene al popula ion coho . In sho , median age was 48.5 yea s, and 53% o he s udy subjec s we e women. A p eex- is ing a he oscle o ic ascula disease was p esen in 3.6%, whe eas 5.5% had diabe es melli us and 25.7% o he subjec s we e smoke s a s udy baseline. The baseline cha ac e is ics o he coho a e summa ized in Table I in he online-only Da a Supplemen . Du ing a ollow-up o 13 yea s, 813 subjec s expe ienced an inciden majo ad e se ca dio ascula e en (MACE), o which 116 we e a al (anno a ed as MACE dea h h oughou his wo k). Table 1 shows se um concen a ions o o al choles e ol, LDL-C, HDL-C, iglyce ides, ce amides, and C- eac i e p o ein (CRP) o asymp oma ic subjec s and o subjec s who expe ienced an inciden MACE du ing he ollow-up. Inc eased concen a ions o dis inc ce amides we e obse ed in indi iduals who expe ienced an inciden MACE du ing he ollow-up compa ed wi h asymp oma ic indi idu- als. The se um concen a ions o he 3 culp i high- isk species ha we e p e iously shown o p edic ca dio ascula dea h in CAD pa ien s we e also highe in FINRISK MACE cases com- pa ed wi h asymp oma ic subjec s as ollows: Ce (d18:1/16:0), Ce (d18:1/18:0), and Ce (d18:1/24:1) 11.4%, 21.3%, and 17.0% (P<0.001 o all), espec i ely. Two p ede ined ce amide a ios, Ce (d18:1/18:0)/Ce (d18:1/24:0) and Ce (d18:1/24:1)/ Ce (d18:1/24:0), we e also signi ican ly highe in subjec s wi h an inciden MACE compa ed wi h asymp oma ic subjec s. Ce amide (d18:1/18:0) Associa es Wi h Inciden MACE The associa ion be ween ce amides and MACE was u he e alua ed by he Cox eg ession analyses. Figu e 1 shows he associa ions o he ce amides wi h inciden MACE, including a al and non a al MACE, and sepa a ely o MACE dea h only. The 3 high- isk ce amides we e associa ed wi h inciden MACE and MACE dea h. The Ce (d18:1/18:0) displayed he highes uni a ia e haza d a io (HR) o inciden MACE (1.31; 95% con idence in e al, 1.21–1.41), which emained signi i- can (1.21; 95% con idence in e al, 1.11–1.33) a e adjus - ing o he F amingham Risk Sco e ac o s. Also, he highes uni a ia e HR o MACE dea h (1.46; 95% con idence in e - al, 1.20–1.78) was obse ed o Ce (d18:1/18:0), bu i did no emain signi ican a e adjus men o he F amingham Risk Sco e ac o s. Subjec s wi h a known p e alen MACE we e excluded om hese analyses. Ce amide (d18:1/18:0) Imp o es Clinical NRI in he FINRISK 2002 Popula ion The inc emen al p ognos ic alue o ce amides was es ed by compa ing he F amingham sco e base model wi h a new model adding ce amides on op o he F amingham sco e. The p edic ed p obabili ies o a 10-yea isk o inciden MACE by c oss- alida ed Weibull eg ession yielded signi ican clinical ne eclassi ica ion index (NRI) o he 3 high- isk ce amide Nons anda d Abb e ia ions and Ac onyms CAD co ona y a e y disease HDL-C high-densi y lipop o ein-choles e ol LDL-C low-densi y lipop o ein-choles e ol MACE majo ad e se ca dio ascula e en NRI ne eclassi ica ion index Table 1. Medians and In e qua ile Ranges o Es ablished Lipid Ma ke s and Ce amides in Pa icipan s Wi hou MACE Du ing he Follow-Up and Wi h Inciden MACE MACE− (n=6892) MACE+ (n=813) Signi icance Ce (d18:1/16:0) 0.28 (0.24–0.33) 0.33 (0.28–0.38) *** Ce (d18:1/18:0) 0.11 (0.09–0.15) 0.15 (0.11–0.19) *** Ce (d18:1/24:0) 2.28 (1.86–2.77) 2.60 (2.14–3.14) *** Ce (d18:1/24:1) 1.52 (1.25–1.85) 1.77 (1.47–2.19) *** Ce (d18:1/16:0)/Ce (d18:1/24:0) 0.13 (0.11–0.14) 0.13 (0.11–0.14) Ce (d18:1/18:0)/Ce (d18:1/24:0) 0.05 (0.04–0.06) 0.06 (0.05–0.07) *** Ce (d18:1/24:1)/Ce (d18:1/24:0) 0.67 (0.59–0.76) 0.70 (0.61–0.79) *** LDL-C 3.30 (2.72–3.90) 3.70 (3.09–4.34) *** HDL-C 1.46 (1.21–1.75) 1.28 (1.12–1.57) TC 5.48 (4.85–6.22) 5.94 (5.21–6.62) *** TG 1.16 (0.84–1.66) 1.50 (1.16–2.18) ** CRP 1.10 (0.51–2.44) 1.94 (0.95–4.63) * CRP indica es C- eac i e p o ein; HDL-C, high-densi y lipop o ein-choles e ol; LDL-C, low-densi y lipop o ein-choles e ol; MACE, majo ad e se ca dio ascula e en ; TC, o al choles e ol; and TG, iglyce ides. The Wilcoxon es ***P<0.001; **P<0.01 and *P<0.05. by gues on Decembe 14, 2016h p://a b.ahajou nals.o g/Downloaded om 2426 A e ioscle Th omb Vasc Biol Decembe 2016 species (Table II in he online-only Da a Supplemen ). In line wi h he abo e-men ioned associa ion esul s, Ce (d18:1/18:0) showed he la ges clinical NRI (7.5%; P=6.9×10−5), which was mainly d i en by down-classi ica ion (6.9%; P=1.2×10−5) o isk among people in he in e media e isk g oup (5%–20% 10-yea MACE isk). Howe e , C-index as a disc imina ion measu e emained nonsigni ican , as well as he NRI o he a al end poin s. Ce amide (d18:1/18:0) Adds Value o Cu en Lipid and CRP Measu emen s The signi ican associa ions be ween inciden MACE and dis inc ce amides p omp ed us o es whe he ce amides could add p ognos ic alue o he cu en ly used ou ine clini- cal labo a o y measu es, such as CRP, HDL-C, o LDL-C. Co ela ion analysis showed weak o mode a e co ela ions o measu ed ce amides and hei a ios wi h CRP and ou inely Figu e 1. Unadjus ed (UHR, blue) and adjus ed (MHR, ed) haza d a ios o inciden majo ad e se ca dio ascula e en (MACE) and a al inciden MACE o pa icipan s wi h no MACE be o e baseline (n=7705). Adjus men o F amingham isk ac o s: o al choles e ol, high-densi y lipop o ein-choles e ol (HDL-C), blood p essu e (adjus ed +15 mm Hg o an ihype ensi e medica ion), diabe es melli us, and smoking. Unadjus ed (UHR, blue) and adjus ed (MHR, ed) haza d a ios o ecu en MACE and a al ecu en MACE o pa ien s wi h p e alen MACE a baseline (n=396). Adjus men o F amingham isk ac o s: o al choles e ol, HDL-C, blood p essu e (adjus ed +15 mm Hg o blood p essu e medica ion), diabe es melli us, and smoking. Age was used as he ime scale in all models, and he models we e s a i ied by sex. The e o e, he unadjus ed models a e essen ially equi alen o age- and sex-adjus ed models. by gues on Decembe 14, 2016h p://a b.ahajou nals.o g/Downloaded om Ha ulinna e al Ci cula ing Ce amides P edic CVD in FINRISK 2002 2427 used lipids (Table III in he online-only Da a Supplemen ). O no e a e he ela i ely s ong co ela ions o Ce (d18:1/18:0) wi h CRP and lipids, excep o HDL-C. Howe e , despi e he obse ed co ela ion be ween ce amides and CRP, he asso- cia ion be ween ce amides and MACE emained p ac ically unchanged a e addi ional adjus men o CRP on op o he F amingham isk sco e a iables (Figu e I in he online- only Da a Supplemen ) compa ed wi h adjus men o he F amingham a iables only (Figu e 1). We selec ed Ce (d18:1/18:0) o mo e de ailed analy- ses because i showed he s onges associa ion wi h inci- den MACE in he analyses summa ized abo e. To assess he po en ial syne gy be ween cu en ly used clinical assays and Ce (d18:1/18:0), we di ided he samples in o concen a- ion qua iles by CRP and Ce (d18:1/18:0), by HDL-C and Ce (d18:1/18:0), and by LDL-C and Ce (d18:1/18:0), and cal- cula ed he inciden MACE a e pe cen ages in hese o e lap- ping qua iles ac oss he combina ions (Figu e 2). Figu e 2A shows ha he lowes 13-yea inciden MACE isk (2.6%) was obse ed in subjec s ha ing bo h CRP and Ce (d18:1/18:0) in he lowes concen a ion qua ile ( i s qua ile [Q1]) and he highes isk (20.9%) was ound in sub- jec s ha ing bo h CRP and Ce (d18:1/18:0) in he highes con- cen a ion qua ile ( ou h qua ile [Q4]). High >15% isk was obse ed in subjec s ha ing CRP concen a ions >0.52 mg/dL (Q2–Q4) and Ce (d18:1/18:0) in Q4. The syne gy o he CRP and Ce (d18:1/18:0) combina ion was u he e idenced by he Kaplan–Meie es ima es o inciden MACE showing he high isk in subjec s wi h op qua ile se um concen a ions o bo h CRP and Ce (d18:1/18:0) (Figu e 3). HDL-C seemed o pe o m simila ly as Ce (d18:1/18:0), bu when combined, mo e cases mapped o he highe concen- a ion Ce (d18:1/18:0) (Q4) and lowe concen a ion HDL-C (Q1) qua iles whe e he MACE isk was 18.4% (Figu e 2B). When mapped oge he wi h LDL-C, he case en ichmen was mainly e iden because o he e ec s o highe concen a ions o Ce (d18:1/18:0) ac oss Q4 (Figu e 2C). These da a a e u - he illus a ed in Figu e 4 sepa a ely o Ce (d18:1/18:0) and LDL-C by Kaplan–Meie cu es o inciden MACE pe con- cen a ion qua ile showing a s eepe slope and highe HRs o Ce (d18:1/18:0). Ce amide (d18:1/16:0) and (d18:1/24:1) Associa e Wi h Recu en MACE Ou p e ious s udies indica e ha ce amides se e as p ognos- ic ma ke s o a al ou come in pa ien s wi h CAD.5,14 Wi h his in mind, we sepa a ely analyzed subjec s who al eady had expe ienced MACE be o e he baseline measu emen (n=396) and as he end poin hose who had a ecu en MACE (n=226) o a al ecu en MACE (n=70) du ing ollow-up. Fo ecu - en MACE, Ce (d18:1/16:0) and Ce (d18:1/24:1) showed signi ican uni a ia e associa ions, wi h Ce (d18:1/16:0) ha ing he highes uni a ia e HR a 1.28 (1.10–1.49), which emained signi ican a HR=1.50 (1.22–1.83) also a e adjus - ing o he F amingham Risk Sco e ac o s (Figu e 1). Fo hose pa ien s who had expe ienced MACE be o e baseline measu emen , he associa ion wi h MACE dea h was signi ican o Ce (d18:1/16:0) and Ce (d18:1/24:1). The uni- a ia e HR was highes a 1.51 (1.15–1.99) o Ce (d18:1/16:0), and i emained signi ican a HR=1.54 (1.07–2.20) a e adjus - ing o he F amingham Risk Sco e ac o s. Figu e 2. A–C, Inciden majo ad e se ca dio ascula e en (MACE) a e (%) in di e en concen a ion qua iles o Ce (d18:1/18:0), C- eac i e p o ein (CRP), low-densi y lipop o ein-choles e ol (LDL- C), and high-densi y lipop o ein-choles e ol (HDL-C). Figu e 3. Kaplan–Meie es ima es o inciden majo ad e se ca - dio ascula e en (MACE) o subpopula ions wi h Ce (d18:1/18:0) in Q4 s no in Q4, C- eac i e p o ein (CRP) in Q4 s no in Q4, and Ce (d18:1/18:0) in Q4 and CRP in Q4 s i s complemen . by gues on Decembe 14, 2016h p://a b.ahajou nals.o g/Downloaded om 2428 A e ioscle Th omb Vasc Biol Decembe 2016 Ce amide Sco e Associa es Wi h MACE a Popula ion Le el Because he da a abo e show ha he occu ence o inciden MACE on one hand, and ecu en MACE on he o he hand, may be associa ed wi h di e en ce amides, we applied a p e i- ously de eloped ce amide isk sco e14 o e alua e he associa ion be ween ce amides and MACE in he whole s udy popula ion. On he basis o his sco e, he pa ien s we e placed in o 4 isk ca ego- ies (low, mode a e, inc eased, and high), and bo h MACE and MACE dea h isk we e inc eased along wi h he inc easing sco e (Table 2). A 2.7- and 4.3- old ela i e isk inc ease was obse ed o MACE and MACE dea h, espec i ely, when compa ing he high- wi h low- isk ca ego y. When subjec s we e so ed acco d- ing o hei LDL-C concen a ions and spli in o 4 ca ego ies in he same p opo ion as o he ce amide isk sco e, he en ich- men o high- isk pa ien s was 1.4- and 1.8- old, espec i ely. Discussion The ce amide analysis o he FINRISK coho showed ha dis inc ce amide species associa e wi h majo ca dio ascula e en s a he popula ion le el. In ou p e ious wo k,5,14 we ha e shown s ong associa ions o ce amides and CAD ou comes in seconda y p e en ion. These associa ions we e u he ali- da ed in his la ge-scale p ima y p e en ion s udy. Because o hei associa ion wi h bo h non a al inciden and a al MACE, i seems ha ce amides may associa e wi h bo h he de elop- men o he a he oscle o ic ascula disease and i s p og es- sion o ascula ad e se e en s. The ela i ely s ong uni a ia e associa ions o ce amides wi h a al e en s sugges ha ce ami- des may play a ole in he up u e o a he oscle o ic plaques. The clinical NRI calcula ions demons a ed ha Ce (d18:1/18:0) holds he po en ial o imp o ing he isk classi ica ion o e he F amingham isk sco e a a popula ion le el. The majo i y o eclassi ica ion seemed o be mainly because o a shi om he in e media e- isk g oup o he low- isk g oup. This inding is in line wi h he ea lie epo s on isk o e es ima ion when using he F amingham isk sco e.15–20 The combined analysis o Ce (d18:1/18:0) wi h CRP, HDL-C, and LDL-C was in e es ing and po en ially appli- cable o imp o ed iden i ica ion o high- isk pa ien s. In pa icula , he highes CRP and Ce (d18:1/18:0) qua iles iden i ied a subs an ial p opo ion o inciden MACE cases demons a ing a syne gy be ween CRP and Ce (d18:1/18:0) (Figu e 2). This ma ke combina ion could be use ul o he iden i ica ion o subjec s in need o p e en i e ca e because Figu e 4. A and B, Kaplan–Meie es i- ma es o inciden majo ad e se ca dio- ascula e en (MACE) Ce (d18:1/18:0) and low-densi y lipop o ein-choles e ol (LDL-C) qua iles. A haza d a io in he legend indica es he qua ile-speci ic haza d ela i e o he i s qua ile (models s a i ied o sex). Table 2. Ce amide Sco e and Risk o MACE and MACE Dea h in FINRISK 2002 Popula ion MACE MACE Dea h No MACE MACE % Rela i e Risk No Dea h Dea h % Rela i e Risk Sco e 0–2 2358 252 9.7 1.0 2579 31 1.2 1.0 3–6 2854 448 13.6 1.4 3239 63 1.9 1.6 7–9 1199 339 22.0 2.3 1479 59 3.8 3.2 10–12 481 170 26.1 2.7 618 33 5.1 4.3 LDL, mmol/L ≤2.88 2271 339 13.0 1.0 2562 48 1.8 1.0 2.88–3.83 2807 495 15.0 1.2 3229 73 2.2 1.2 3.83–4.66 1281 257 16.7 1.3 1495 43 2.8 1.5 ≥4.66 533 118 18.1 1.4 629 22 3.4 1.8 LDL indica es low-densi y lipop o ein; and MACE, majo ad e se ca dio ascula e en . by gues on Decembe 14, 2016h p://a b.ahajou nals.o g/Downloaded om Ha ulinna e al Ci cula ing Ce amides P edic CVD in FINRISK 2002 2429 o a high MACE isk. In co ela ion analyses, only ela i ely modes co ela ions among ce amides, CRP, and HDL-C we e obse ed, sugges ing ha he obse ed p edic i e e ec o Ce (d18:1/18:0) is independen and canno be explained by associa ions be ween hese molecules. Howe e , u he s ud- ies a e needed o es ablish he biological mechanisms o he ce amide e ec s and he clinical u ili y o he measu emen s. The subanalysis o subjec s who had had MACE be o e 2002 baseline sugges ed ha Ce (d18:1/16:0) and Ce (d18:1/24:1) associa ed s onge han Ce (d18:1/18:0) wi h new ca dio ascu- la e en s in pa ien s wi h p eexis ing disease, which is in ag ee- men wi h ou ea lie esul s on s able CAD and acu e co ona y synd ome pa ien s.4,5,14 Fu u e mechanis ic s udies a e needed o e alua e his di e ence in ce amide beha io be ween p i- ma y and seconda y p e en ion. We ha e p e iously de eloped a ce amide sco e ha akes in o accoun all 3 culp i ce amides and hei a ios wi h Ce (d18:1/24:0) o simpli y he clinical use o he ce amide da a. We applied he same ce amide sco e in he p esen s udy o he whole FINRISK 2002 coho and obse ed a solid sco e pe o mance a he popula ion le el. In ou p e ious s udies in pa ien s wi h es ablished CAD, Ce (d18:1/24:0) appea ed ca diop o ec i e, whe eas in he cu en s udy Ce (d18:1/24:0) was weakly associa ed wi h he inc eased isk o MACE and MACE dea h. The eason o his di e ence emains o be in es iga ed in u u e s udies. One may specula e ha he di e en beha io o Ce (d18:1/24:0) in p ima y and seconda y p e en ion s udies may be ela ed o i s associa ion wi h bo h lipop o ein pa icles and in lam- ma o y p ocesses. In ac , Ce (d18:1/24:0) beha io seems o mi o ha o LDL-C. LDL-C ypically associa es wi h ca - dio ascula isk in p ima y p e en ion, whe eas in pa ien s wi h an es ablished CAD he e is an in e se ela ionship o no associa ion. This has been sugges ed o ela e o in lamma o y con ol o hepa ic LDL ecep o egula ion.21 Thus, i is pos- sible ha Ce (d18:1/24:0) is biologically less ac i e compa ed wi h o he ce amide species, and in he isk p edic ion, i is me ely a measu e o lipop o ein me abolism. By design, ou s udy is an obse a ional coho s udy, and he e o e, i is no possible o es ablish causali y o po en ial ea men consequences. Thus, i is no known a he momen , whe he ea ing ca dio ascula isk on he basis o ce amide classi ica ion would p oduce any be e clinical esul s han he ea men s based on cu en isk es ima ion algo i hms, such as he F amingham o he SCORE equa ion (Sys ema ic Co ona y Risk E alua ion). Ano he limi a ion is he ac ha ou s udy coho consis s almos o ally o whi es. Thus, u he s ud- ies in e hnically mo e di e se popula ions a e wa an ed. The s eng hs o he s udy include he la ge FINRISK 2002 coho , which is ep esen a i e o he gene al Finnish popula ion and has a long and comp ehensi e ollow-up o ca dio ascula e en s. Ano he s eng h is he mode n and sophis ica ed mass spec ome y echnology, which enabled he measu emen o ci cula ing ce amides in 8000 indi iduals wi h good accu acy. In conclusion, ou esul s show ha dis inc ci cula ing ce amides a e associa ed wi h he isk o inciden MACE in heal hy indi iduals. These esul s add insigh o ou unde - s anding o he pa hogenesis o MACE and sugges new op ions o isk es ima ion. Acknowledgmen s We hank M s Si pa Su ela-Tuominen and M s Ri a Seppänen o expe echnical suppo and sample p epa a ion. Sou ces o Funding The esea ch leading o hese esul s has ecei ed unding om he Eu opean Union’s Se en h F amewo k P og amme FP7/2007 o 2013 unde g an ag eemen n° 305739 (RiskyCAD). V.S. was suppo ed by he Finnish Founda ion o Ca dio ascula Resea ch. Disclosu es Zo a Biosciences holds pa en s o he diagnos ic use o ce amides, and R.H. and R.L. a e sha eholde s o Zo a Biosciences. R.H., R.L., M.S.-A., M.H., K.E., and D.K. a e employees o Zo a Biosciences. The o he au ho s epo no con lic s. Re e ences 1. Ellims AH, Wong G, Wei JM, Lew P, Meikle PJ, Taylo AJ. Plasma lipi- domic analysis p edic s non-calci ied co ona y a e y plaque in asymp- oma ic pa ien s a in e media e isk o co ona y a e y disease. Eu Hea J Ca dio asc Imaging. 2014;15:908–916. doi:10.1093/ehjci/jeu033. 2. Fe nandez C, Sandin M, Sampaio JL, Almg en P, Na kiewicz K, Ho mann M, Hedne T, Wahls and B, Simons K, She chenko A, James P, Melande O. Plasma lipid composi ion and isk o de eloping ca dio ascula dis- ease. PLoS One. 2013;8:e71846. doi: 10.1371/jou nal.pone.0071846. 3. Meikle PJ, Wong G, Tso o es D, e al. Plasma lipidomic analysis o s a- ble and uns able co ona y a e y disease. A e ioscle Th omb Vasc Biol. 2011;31:2723–2732. doi: 10.1161/ATVBAHA.111.234096. 4. Ta aso K, Ek oos K, Suoniemi M, Kauhanen D, Syl änne T, Hu me R, Gouni-Be hold I, Be hold HK, Klebe ME, Laaksonen R, Mä z W. Molecula lipids iden i y ca dio ascula isk and a e e icien ly low- e ed by sim as a in and PCSK9 de iciency. J Clin Endoc inol Me ab. 2014;99:E45–E52. doi: 10.1210/jc.2013-2559. 5. Cheng JM, Suoniemi M, Ka dys I, e al. Plasma concen a ions o molecula lipid species in ela ion o co ona y plaque cha ac e is ics and ca dio ascu- la ou come: esul s o he ATHEROREMO-IVUS s udy. A he oscle osis. 2015;243:560–566. doi: 10.1016/j.a he oscle osis.2015.10.022. 6. Bismu h J, Lin P, Yao Q, Chen C. Ce amide: a common pa hway o a he oscle osis? A he oscle osis. 2008;196:497–504. doi: 10.1016/j. a he oscle osis.2007.09.018. 7. F eed JK, Beye AM, LoGiudice JA, Hockenbe y JC, Gu e man DD. Ce amide changes he media o o low-induced asodila ion om ni ic oxide o hyd ogen pe oxide in he human mic oci cula ion. Ci c Res. 2014;115:525–532. doi: 10.1161/CIRCRESAHA.115.303881. 8. Syed I, Jaya am B, Subasinghe W, Kowlu u A. Tiam1/Rac1 signaling pa hway media es palmi a e-induced, ce amide-sensi i e gene a ion o supe oxides and lipid pe oxides and he loss o mi ochond ial memb ane po en ial in panc ea ic be a-cells. Biochem Pha macol. 2010;80:874–883. doi: 10.1016/j.bcp.2010.05.006. 9. Maceyka M, Spiegel S. Sphingolipid me aboli es in in lamma o y disease. Na u e. 2014;510:58–67. doi: 10.1038/na u e13475. 10. Es uch M, Sánchez-Quesada JL, O dóñez-Llanos J, Bení ez S. Ce amide- en iched LDL induces cy okine elease h ough TLR4 and CD14 in mono- cy es. Simila i ies wi h elec onega i e LDL. Clin In es ig A e ioscle . 2014;26:131–137. doi: 10.1016/j.a e i.2013.12.003. 11. Halasiddappa LM, Koe ele H, Fu e man AH, He me e A. Oxidized phospholipids induce ce amide accumula ion in RAW 264.7 mac ophages: ole o ce amide syn hases. PLoS One. 2013;8:e70002. doi: 10.1371/jou - nal.pone.0070002. 12. Billich A, Baum uke T. Sphingolipid me abolizing enzymes as no el he apeu ic a ge s. Subcell Biochem. 2008;49:487–522. doi: 10.1007/978-1-4020-8831-5_19. 13. Pa k JW, Pa k WJ, Fu e man AH. Ce amide syn hases as po en ial a ge s o he apeu ic in e en ion in human diseases. Biochim Biophys Ac a. 2014;1841:671–681. doi: 10.1016/j.bbalip.2013.08.019. 14. Laaksonen R, Ek oos K, Sysi-Aho M, e al. Plasma ce amides p e- dic ca dio ascula dea h in pa ien s wi h s able co ona y a e y disease and acu e co ona y synd omes beyond LDL-choles e ol. Eu Hea J. 2016;37:1967–1976. doi: 10.1093/eu hea j/ehw148. 15. B indle P, Embe son J, Lampe F, Walke M, Whincup P, Fahey T, Eb ahim S. P edic i e accu acy o he F amingham co ona y isk sco e in B i ish by gues on Decembe 14, 2016h p://a b.ahajou nals.o g/Downloaded om 2430 A e ioscle Th omb Vasc Biol Decembe 2016 men: p ospec i e coho s udy. BMJ. 2003;327:1267. doi: 10.1136/ bmj.327.7426.1267. 16. Hense HW, Schul e H, Löwel H, Assmann G, Keil U. F amingham isk unc ion o e es ima es isk o co ona y hea disease in men and women om Ge many— esul s om he MONICA Augsbu g and he PROCAM coho s. Eu Hea J. 2003;24:937–945. doi: 10.1016/ S0195-668X(03)00081-2. 17. DeFilippis AP, Young R, Ca ubba CJ, McE oy JW, Budo MJ, Blumen hal RS, K onmal RA, McClelland RL, Nasi K, Blaha MJ. An analysis o calib a ion and disc imina ion among mul iple ca dio ascula isk sco es in a mode n mul ie hnic coho . Ann In e n Med. 2015;162:266–275. doi: 10.7326/M14-1281. 18. Ma ques-Vidal P, Rodondi N, Bochud M, Chiole o A, Pécoud A, Hayoz D, Paccaud F, Moose V, Fi mann M, Waebe G, Vollenweide P. P edic i e accu acy o o iginal and ecalib a ed F amingham isk sco e in he Swiss popula ion. In J Ca diol. 2009;133:346–353. doi: 10.1016/j. ijca d.2008.01.004. 19. Eichle K, Puhan MA, S eu e J, Bachmann LM. P edic ion o i s co o- na y e en s wi h he F amingham sco e: a sys ema ic e iew. Am Hea J. 2007;153:722–31, 731.e1. doi: 10.1016/j.ahj.2007.02.027. 20. Empana JP, Ducime iè e P, A eile D, Fe iè es J, E ans A, Ruida e s JB, Haas B, Ya nell J, Bingham A, Amouyel P, Dallonge ille J; PRIME S udy G oup. A e he F amingham and PROCAM co ona y hea disease isk unc ions applicable o di e en Eu opean popula ions? The PRIME S udy. Eu Hea J. 2003;24:1903–1911. doi: 10.1016/j. ehj.2003.09.002. 21. Gie ens H, Nauck M, Ro h M, Schinke R, Schü mann C, Scha nagl H, Neuhaus G, Wieland H, Mä z W. In e leukin-6 s imula es LDL ecep o gene exp ession ia ac i a ion o s e ol- esponsi e and Sp1 binding ele- men s. A e ioscle Th omb Vasc Biol. 2000;20:1777–1783. Highligh s • The ce amide analysis o he FINRISK coho showed ha dis inc ce amide species associa e wi h majo ca dio ascula e en s a he popula ion le el. • Ce amides may be new bioma ke s o majo ad e se ca dio ascula e en isk. • The combina ion o Ce (d18:1/18:0) and C- eac i e p o ein could be use ul o he iden i ica ion o subjec s a high majo ad e se ca dio as- cula e en isk. by gues on Decembe 14, 2016h p://a b.ahajou nals.o g/Downloaded om Veikko Salomaa and Reijo Laaksonen Aki S. Ha ulinna, Ma ko Sysi-Aho, Mika Hil o, Dimple Kauhanen, Reini Hu me, Kim Ek oos, FINRISK 2002 Coho Ci cula ing Ce amides P edic Ca dio ascula Ou comes in he Popula ion-Based P in ISSN: 1079-5642. Online ISSN: 1524-4636 Copy igh © 2016 Ame ican Hea Associa ion, Inc. All igh s ese ed. G een ille A enue, Dallas, TX 75231 is published by he Ame ican Hea Associa ion, 7272A e ioscle osis, Th ombosis, and Vascula Biology doi: 10.1161/ATVBAHA.116.307497 2016; 2016;36:2424-2430; o iginally published online Oc obe 20,A e ioscle Th omb Vasc Biol. F ee ia Open Access h p://a b.ahajou nals.o g/con en /36/12/2424 Wo ld Wide Web a : The online e sion o his a icle, along wi h upda ed in o ma ion and se ices, is loca ed on he h p://a b.ahajou nals.o g//subsc ip ions/ a : is onlineA e ioscle osis, Th ombosis, and Vascula Biology In o ma ion abou subsc ibing o Subsc ip ions: h p://www.lww.com/ ep in s In o ma ion abou ep in s can be ound online a : Rep in s: documen . Ques ion and Answe Pe missions and Righ spage unde Se ices. Fu he in o ma ion abou his p ocess is a ailable in he which pe mission is being eques ed is loca ed, click Reques Pe missions in he middle column o he Web Copy igh Clea ance Cen e , no he Edi o ial O ice. Once he online e sion o he published a icle o can be ob ained ia Righ sLink, a se ice o heA e ioscle osis, Th ombosis, and Vascula Biologyin Reques s o pe missions o ep oduce igu es, ables, o po ions o a icles o iginally publishedPe missions: by gues on Decembe 14, 2016h p://a b.ahajou nals.o g/Downloaded om h p://a b.ahajou nals.o g/con en /suppl/2016/11/28/ATVBAHA.116.307497.DC3.h ml h p://a b.ahajou nals.o g/con en /suppl/2016/10/27/ATVBAHA.116.307497.DC2.h ml h p://a b.ahajou nals.o g/con en /suppl/2016/10/27/ATVBAHA.116.307497.DC1.h ml Da a Supplemen (unedi ed) a : h p://a b.ahajou nals.o g//subsc ip ions/ a : is onlineA e ioscle osis, Th ombosis, and Vascula Biology In o ma ion abou subsc ibing o Subsc ip ions: h p://www.lww.com/ ep in s In o ma ion abou ep in s can be ound online a : Rep in s: documen . Ques ion and Answe Pe missions and Righ spage unde Se ices. Fu he in o ma ion abou his p ocess is a ailable in he which pe mission is being eques ed is loca ed, click Reques Pe missions in he middle column o he Web Copy igh Clea ance Cen e , no he Edi o ial O ice. Once he online e sion o he published a icle o can be ob ained ia Righ sLink, a se ice o heA e ioscle osis, Th ombosis, and Vascula Biologyin Reques s o pe missions o ep oduce igu es, ables, o po ions o a icles o iginally publishedPe missions: by gues on Decembe 14, 2016h p://a b.ahajou nals.o g/Downloaded om ‘su i al’ package) and all models we e adequa e. We es ima ed he clinical NRI in he 5…20% isk ca ego y 11 , i.e. NRI among hose who we e a he in e media e (5…20%) isk g oup when p edic ed by he e e ence model wi hou ce amides. The calcula ion o ce amide sco e has been es ablished ecen ly 12 . In b ie , o each indi idual all h ee ce amide a ios and each concen a ion, excep o Ce (d18:1/24:0), we e compa ed o he whole s udy popula ion. I he a iable belonged o he 3 d qua ile, he indi idual ecei ed +1 poin , and i o he 4 h qua ile, +2 poin s. Based on he sco es, he subjec s we e spli in o ou isk ca ego ies (low=0-2, mode a e=3-6, inc eased=7-9 and high=10-12). When compa ing he esul wi h LDL-C, he subjec s we e so ed acco ding o hei LDL-C concen a ions and spli in o ou ca ego ies in he same p opo ion as o he ce amide isk sco e. R e sion 3.2.2 (R Co e Team (2015), Vienna, Aus ia) was used o all s a is ical analyses. p<0.05 was conside ed as s a is ically signi ican and all es s we e wo-sided. Re e ences 1. Bo odulin K, Va iainen E, Pel onen M, Jousilah i P, Juole i A, Laa ikainen T, Männis ö S, Salomaa V, Sund all J, Puska P. Fo y-yea ends in ca dio ascula isk ac o s in Finland. Eu J Public Heal h. 2015;25(3):539-546. doi:10.1093/eu pub/cku174. 2. Leh o H-R, Leh o S, Ha ulinna AS, Salomaa V. Does he clinical spec um o inciden ca dio ascula disease di e be ween men and women? Eu J P e Ca diol. 2014;21(8):964-971. doi:10.1177/2047487313482284. 3. Tolonen H, Salomaa V, To ppa J, Si enius J, Immonen-Räihä P, Leh onen A. The alida ion o he Finnish Hospi al Discha ge Regis e and Causes o Dea h Regis e da a on s oke diagnoses. Eu J Ca dio asc P e Rehabil. 2007;14(3):380-385. doi:10.1097/01.hj .0000239466.26132. 2. 4. Pajunen P, Koukkunen H, Ke onen M, Je kkola T, Immonen-Raiha P, Ka ja- Koskenka i P, Mahonen M, Niemela M, Kuulasmaa K, Palomaki P, Mus onen J, Leh onen A, A s ila M, Vuo enmaa T, Leh o S, Mie inen H, To ppa J, Tuomileh o J, Kesaniemi YA, Pyo ala K, Salomaa V. The alidi y o he Finnish Hospi al Discha ge Regis e and Causes o Dea h Regis e da a on co ona y hea disease. Eu J Ca dio asc P e Rehabil. 2005;12(2):132-137. doi:10.1097/01.hj .0000140718.09768.ab. 5. Mähönen M, Jula A, Ha ald K, An ikainen R, Tuomileh o J, Zelle T, Blankenbe g S, Salomaa V. The alidi y o hea ailu e diagnoses ob ained om adminis a i e egis e s. Eu J P e Ca diol. 2013;20(2):254-259. doi:10.1177/2047487312438979. 6. Mähönen M, Salomaa V, B ommels M, Mola ius A, Mie inen H, Pyö älä K, Tuomileh o J, A s ila M, Kaa salo E, Ke onen M, Kuulasmaa K, Leh o S, Mus aniemi H, Niemelä M, Palomäki P, To ppa J, Vuo enmaa T. The alidi y o hospi al discha ge egis e da a on co ona y hea disease in Finland. Eu J Epidemiol. 13(4):403-415. doi:10.1023/A:1007306110822. 7. Kauhanen D, Sysi-Aho M, Kois inen KM, Laaksonen R, Sinisalo J, Ek oos K. De elopmen and alida ion o a high- h oughpu LC-MS/MS assay o ou ine measu emen o molecula ce amides. Anal Bioanal Chem. 2016;408(13):3475- 3483. doi:10.1007/s00216-016-9425-z. 8. D’Agos ino RB, Vasan RS, Pencina MJ, Wol PA, Cobain M, Massa o JM, Kannel WB. Gene al ca dio ascula isk p o ile o use in p ima y ca e: he F amingham Hea S udy. Ci cula ion. 2008;117(6):743-753. doi:10.1161/CIRCULATIONAHA.107.699579. 9. Tobin MD, Sheehan NA, Scu ah KJ, Bu on PR. Adjus ing o ea men e ec s in s udies o quan i a i e ai s: an ihype ensi e he apy and sys olic blood p essu e. S a Med. 2005;24(19):2911-2935. doi:10.1002/sim.2165. 10. Cook NR, Ridke PM. Ad ances in measu ing he e ec o indi idual p edic o s o ca dio ascula isk: he ole o eclassi ica ion measu es. Ann In e n Med. 2009;150(11):795-802. 11. Cook NR. Commen s on “E alua ing he added p edic i e abili y o a new ma ke : F om a ea unde he ROC cu e o eclassi ica ion and beyond” by M. J. Pencina e al., S a is ics in Medicine (DOI: 10.1002/sim.2929). S a Med. 2008;27(2):191- 195. doi:10.1002/sim.2987. 12. Laaksonen R, Ek oos K, Sysi-Aho M, Hil o M, Vihe aa a T, Kauhanen D, Suoniemi M, Hu me R, Mä z W, Scha nagl H, S ojako ic T, Vlachopoulou E, Lokki M-L, Nieminen MS, Klingenbe g R, Ma e CM, Ho nemann T, Jüni P, Rodondi N, Räbe L, Windecke S, Gence B, Pede sen ER, Tell GS, Nygå d O, Mach F, Sinisalo J, Lüsche TF. Plasma ce amides p edic ca dio ascula dea h in pa ien s wi h s able co ona y a e y disease and acu e co ona y synd omes beyond LDL-choles e ol. Eu Hea J. 2016;in p ess. doi:10.1093/eu hea j/ehw148. Thea iclebyHa ulinnae al(A e ioscle Th ombVascBiol.2016;36:2424‐2430.DOI: 10.1161/ATVBAHA.116.307497),whichpublishedonlineaheado p in onOc obe 20,2016,was changed obeanOpenAccessa iclewi haCC‐BYlicense. Thechangeis e lec edin heDecembe 2016issue,a ailablea  h p://a b.ahajou nals.o g/con en /36/12/2424.