A Protein Diet Score, Including Plant and Animal Protein, Investigating the Association with HbA1c and eGFR-The PREVIEW Project

nu ien s A icle A P o ein Die Sco e, Including Plan and Animal P o ein, In es iga ing he Associa ion wi h HbA1c and eGFR—The PREVIEW P ojec G i h Mølle 1,*ID , Diewe je Sluik 2ID , Ch is ian Ri z 1ID , Ve a Mikkilä 3, Olli T. Rai aka i 4,5, Nina Hu i-Kähönen 6, La s O. D ags ed 1, Thomas M. La sen 1ID , Sally D. Poppi 7, Ma a P. Sil es e 7, Edi h J.M. Feskens 2, Jennie B and-Mille 8ID and Anne Raben 1ID 1Depa men o Nu i ion, Exe cise and Spo s, Facul y o Science, Uni e si y o Copenhagen, Roligheds ej 26, 1958 F ede iksbe g C, Denma k; [email p o ec ed] (C.R.); [email p o ec ed] (L.O.D.); [email p o ec ed] (T.M.L.); [email p o ec ed] (A.R.) 2Di ision o Human Nu i ion, Wageningen Uni e si y, S ippeneng 4, 6708 WE Wageningen, The Ne he lands; Diewe je.sluik@wu .nl (D.S.); edi h. eskens@wu .nl (E.J.M.F.) 3Depa men o Food and En i onmen al Sciences, Uni e si y o Helsinki, 00014 Helsinki, Finland; [email p o ec ed] 4Resea ch Cen e o Applied and P e en i e Ca dio ascula Medicine, Kiinamyllynka u 10, Uni e si y o Tu ku, 20520 Tu ku, Finland; [email p o ec ed] 5 Depa men o Clinical Physiology and Nuclea Medicine, Tu ku Uni e si y Hospi al, 20521 Tu ku, Finland 6Depa men o Pedia ics, Tampe e Uni e si y Hospi al and Facul y o Medicine and Li e Sciences, Uni e si y o Tampe e, 33014 Tampe e, Finland; [email p o ec ed] 7Human Nu i ion Uni , School o Biological Sciences, 18 Ca ick Place, M Eden, Uni e si y o Auckland, Auckland 1024, New Zealand; [email p o ec ed] (S.D.P.); [email p o ec ed] (M.P.S.) 8School o Li e and En i onmen al Sciences & Cha les Pe kins Cen e, Uni e si y o Sydney, Campe down, NSW 2006, Aus alia; jennie.b andmille @sydney.edu.au *Co espondence: [email p o ec ed]; Tel.: +45-35-33-48-18 Recei ed: 7 June 2017; Accep ed: 11 July 2017; Published: 17 July 2017 Abs ac : Highe -p o ein die s ha e been ad oca ed o body-weigh egula ion o he pas ew decades. Howe e , he po en ial heal h isks o hese die s a e s ill unce ain. We aimed o de elop a p o ein sco e based on he quan i y and sou ce o p o ein, and o examine he associa ion o he sco e wi h glyca ed haemoglobin (HbA1c) and es ima ed glome ula il a ion a e (eGFR). Analyses we e based on h ee popula ion s udies included in he PREVIEW p ojec (PREVen ion o diabe es h ough li es yle In e en ion and popula ion s udies in Eu ope and a ound he Wo ld): NQplus, Li elines, and he Young Finns S udy. C oss-sec ional da a om ood- equency ques ionnai es (n= 76,777 subjec s) we e used o de elop a p o ein sco e consis ing o wo componen s: 1) pe cen age o ene gy om o al p o ein, and 2) plan o animal p o ein a io. An in e se associa ion be ween p o ein sco e and HbA1c (slope − 0.02 ± 0.01 mmol/mol, p< 0.001) was seen in Li elines. We ound a posi i e associa ion be ween he p o ein sco e and eGFR in Li elines (slope 0.17 ± 0.02 mL/min/1.73 m 2 ,p< 0.0001). P o ein sco ing migh be a use ul ool o assess bo h he e ec o quan i y and sou ce o p o ein on heal h pa ame e s. Fu he s udies a e needed o alida e his newly de eloped p o ein sco e. Keywo ds: p o ein die sco e; HbA1c; eGFR; heal hy subjec s; popula ion s udies 1. In oduc ion A die ich in p o ein, anging om 1.2 o 1.6 g p o ein/kg/day, may imp o e body weigh egula ion [ 1 – 3 ]. P o ein- ich die s also appea o supplemen o he s a egies, such as ene gy es ic ion and physical ac i i y, o comba he global obesi y epidemic [ 4 ]. As obesi y is an independen Nu ien s 2017,9, 763; doi:10.3390/nu9070763 www.mdpi.com/jou nal/nu ien s Nu ien s 2017,9, 763 2 o 14 isk ac o o ype 2 diabe es (T2D), highe p o ein die s o weigh egula ion and main enance migh also be bene icial o he p e en ion o T2D [ 5 ]. Al hough he o al amoun o p o ein may be impo an , he sou ce o p o ein will in luence o he componen s o he die , such as die a y ib e and mic onu ien s. The e o e, p o ein sou ce is likely o be an impo an de e minan o heal h ou comes [ 6 ]. Die a y guidelines sugges mo ing owa ds a mo e plan -based die [ 7 ]. Plan -de i ed die s p o ide a numbe o phy ochemicals ha ha e been associa ed wi h p o ec ion agains many ch onic diseases, bu con e sely compa ed o die a y p o eins om animal sou ces, plan p o eins lack su icien amoun s o key essen ial amino acids. In mos indus ialised coun ies, animal p o ein is he main sou ce o die a y p o ein a he han plan p o ein, and is he e o e o en a di e en ially inc eased choice when o al p o ein in ake is inc eased. Howe e , he consump ion o animal p oduc s, especially ed and p ocessed mea , has been associa ed wi h an inc eased isk o diseases such as cance , T2D and ca dio ascula diseases [ 8 – 10 ]. The op imal plan o animal p o ein a io in he die has no ye been es ablished. Fi s ly, we conside ed whe he a highe p o ein in ake migh be associa ed wi h lowe glyca ed haemoglobin (HbA1c). HbA1c le els measu e a e age blood glucose concen a ion o e he p e ious six o eigh weeks; high concen a ions a e a isk ac o o T2D. A ecen me a-analysis o 32 andomised con olled ials showed a long- e m posi i e e ec o highe -p o ein die s on body weigh managemen , which in u n could lead o lowe HbA1c [ 11 ]. Se e al die a y in e en ion s udies ha e also shown ha highe p o ein die s can lowe HbA1c di ec ly, a leas among T2D pa ien s [ 12 ]. Secondly, we conside ed whe he a highe p o ein in ake may be associa ed wi h ele a ed es ima ed glome ula il a ion a e (eGFR), an ad e se indica o o enal unc ion, since ca e ully con olled die a y s udies show ha highe p o ein die s may inc ease his ma ke o kidney unc ion [ 13 ]. The cu en conce n o ad e se enal e ec s o highe p o ein die s de i es om he de imen al e ec s o induced glome ula hype il a ion [ 14 ]. Thi dly, wi h espec o p o ein quali y, he e a e se e al lines o epidemiological e idence indica ing ha an inc eased consump ion o plan p o ein may be associa ed wi h a educed isk o ca dio ascula disease, T2D [ 5 , 15 ] and in lamma ion [ 16 ], which migh be asc ibed in pa o lowe le els o HbA1c. In he pas h ee decades, s udies ha e clea ly shown ha he ela ionship be ween die a y in ake and heal h is e y complex, wi h many in e ac ions [ 17 ]. Fo all hese easons, examining composi e indices o ood and nu ien in ake can be use ul. Recen ly, a low-ca bohyd a e die sco e was de eloped by Hal on e al. [ 18 ] o classi y women in he Nu ses’ Heal h S udy acco ding o hei ela i e le els o a , p o ein—including animal and ege able p o ein—and ca bohyd a e in ake. This sco e was used o p ospec i ely examine he associa ion wi h he isk o co ona y hea disease in his coho . A sco ing ool o speci ically assess quan i y, as well as he sou ce, o p o ein in ake has o ou knowledge no p e iously been de eloped. The e o e, we aimed o de elop a p o ein die sco e based on die a y p o ein quan i y and sou ce—plan o animal—as a ool o in es iga e he ole o p o ein in T2D- ela ed ad e se me abolic heal h. We hypo hesised ha a p o ein sco e wi h a highe p o ein ene gy pe cen age (E%) wi hin he accep able mac onu ien dis ibu ion ange o p o ein [ 19 ], in combina ion wi h a highe plan o animal p o ein a io, would be associa ed wi h a lowe HbA1c le el and possibly also an inc ease in eGFR. 2. Ma e ials and Me hods 2.1. S udy Design and Popula ion This s udy included c oss-sec ional da a om wo Du ch and one Finnish obse a ional s udies, NQplus, Li elines and The Young Finns S udy, all pa o he PREVIEW p ojec (PREVen ion o diabe es h ough li es yle In e en ion and popula ion s udies in Eu ope and a ound he Wo ld) www.p e iews udy.com, The o e all aim o PREVIEW is o in es iga e he impac o a highe -p o ein, low glycemic index (GI) die and physical ac i i y egime o he p e en ion o T2D in o e weigh and obese adul s and child en a high isk o de eloping his disease [20]. Nu ien s 2017,9, 763 3 o 14 The Nu i ion Ques ionnai es plus s udy ( he NQplus s udy) is a longi udinal obse a ional s udy in Du ch adul s wi hin he su oundings o Wageningen, he Ne he lands. I s main aims a e o de elop a na ional die a y assessmen e e ence da abase o he u u e de elopmen and imp o emen o die a y assessmen me hods, o alida e newly de eloped ood equency ques ionnai es (FFQs), and o s udy die a y ac o s and in e media e heal h ou comes. A o al o 2048 indi iduals aged om 20–70 yea s ook pa in he s udy [21,22]. Li elines is a la ge obse a ional popula ion-based coho s udy o bo h adul s and child en conduc ed in he no h o he Ne he lands [23]. The o e all aim o his s udy was o gain insigh in o he e iology o heal hy aging in he gene al popula ion. Li elines was ini ia ed in 2006 and baseline da a ha e been collec ed o 167,729 indi iduals, aged 6 mon hs–93 yea s. Regula ollow-up measu emen s a e planned. The Young Finns S udy is a mul i-cen e ollow-up s udy in Finland. I s main goals a e o de e mine he con ibu ion made by childhood li es yle, biological, and psychological measu es o he isk o ca dio ascula diseases in adul hood. A baseline in 1980, 3596 subjec s aged 3–18 yea s we e included, and he same subjec s ha e now been ollowed o mo e han 30 yea s [ 23 ]. Fo he pu poses o his c oss-sec ional in es iga ion, we used measu emen s om he 2007 ollow-up, when subjec s we e 30–45 yea s o age. All subjec s a baseline who had missing da a on he FFQ, o subjec s who had implausibly high (>3500 kcal) o low (<500 kcal) daily ene gy in ake on he FFQ, we e excluded [ 18 ]. Fu he mo e, subjec s wi h missing alues in ei he exposu e o ou come a iables we e omi ed. Also, subjec s wi h a his o y o diabe es, hype ension, hype choles e olemia, ca dio ascula disease, cance o kidney disease we e excluded, because hese diseases can cause al e a ions in he habi ual die . A e hese exclusions, a o al o 76,777 subjec s om Li elines (n= 75,131), NQplus (n= 492), and The Young Finns S udy (n= 1154) emained in he cu en analysis. 2.2. Assessmen o Die NQplus used a alida ed 180 i em semi-quan i a i e gene al FFQ o assess usual die a y in ake [24,25] . Answe ca ego ies o equency ques ions anged be ween ne e pe mon h o 6–7 days pe week, and po ion sizes we e es ima ed using ypical po ion size es ima es (e.g., glass, slice) and commonly used household measu es. A e age daily nu ien in akes we e calcula ed by mul iplying he equency o consump ion by po ion size and nu ien con en pe g am using he Du ch ood composi ion able o 2011 [ 26 ]. Wi h espec o he de ini ion o animal and ege able p o ein, he Du ch ood composi ion able dis inguishes be ween hese sou ces and was used o bo h NQplus and Li elines. In Li elines, a newly de eloped FFQ consis ing o 110 i ems was used o es ima e in ake o ene gy, a , ca bohyd a e, p o ein and alcohol. Responses o ood equency ques ions anged be ween ne e pe mon h o 6–7 days pe week, and po ion sizes we e es ima ed using ypical po ion size es ima es and commonly used household measu es. A e age daily nu ien in akes we e calcula ed by mul iplying equency o consump ion by po ion size and nu ien con en pe g am using he Du ch ood composi ion able o 2011 [26]. In he Young Finns S udy, a alida ed 131-i em quan i a i e FFQ was used [ 27 ]. The ood i ems we e p esen ed in 12 subg oups, e.g., dai y p oduc s, ege ables, o ui s and be ies. A e each subg oup, he e we e emp y lines o subjec s o add oods no lis ed in he ques ionnai e. The po ion sizes we e ixed and, i possible, speci ied using ypical po ion size es ima es. The nine equency ca ego ies anged om ne e o seldom o six o mo e imes a day. The ood consump ion and nu ien in akes we e calcula ed by mul iplying he equency o ood consump ion by ixed po ion sizes o ob ain he weigh o each lis ed ood i em consumed as an a e age pe day [ 28 ]. The Finnish ood composi ion da abase o 2008 was used o he nu ien calcula ions [ 29 ]. The da abase does no dis inguish be ween animal o plan p o ein as sepa a e nu ien s; he e o e, all oods we e i s Nu ien s 2017,9, 763 4 o 14 classi ied in o ei he “animal” o “plan ” oods. P o ein in akes we e calcula ed by he sou ce using hese classes. 2.3. Calcula ion o he P o ein Sco e Following he me hodology o Hal on e al. [ 18 ], a p o ein sco e was de eloped based on ela i e cu -o poin s i.e., he popula ion dis ibu ion. The sco ing was se acco ding o he hypo hesis ha a highe p o ein in ake, as well as a highe plan o animal p o ein a io, would be associa ed wi h imp o ed ma ke s o heal h ou comes, including a lowe HbA1c (a measu e o a e age blood glucose concen a ions in he p e ious 6–8 weeks) and highe eGFR (a measu e o good kidney unc ion). Each s udy popula ion was di ided in o 11 s a a acco ding o o al p o ein in ake (exp essed as E%), and 11 s a a acco ding o plan o animal p o ein a io. Subjec s in he highes s a um o p o ein in ake ecei ed 10 poin s, subjec s in he nex s a um ecei ed 9 poin s, and so on, down o subjec s in he lowes s a um, who ecei ed 0 poin s. In e ms o calcula ing plan o animal p o ein a io, hose wi h he highes in ake o plan o animal p o ein ecei ed 10 poin s, and hose wi h he lowes plan o animal p o ein a io ecei ed 0 poin s. The sum o he poin s o each o he wo componen s c ea ed he o e all p o ein die sco e, which anged om 0 o 20 poin s. The e o e, a highe sco e e lec s a highe ene gy pe cen age o o al p o ein, and a highe plan o animal p o ein a io, while a lowe sco e e lec s a lowe p o ein and a lowe plan o animal p o ein a io. Each componen o he sco e was also conside ed sepa a ely. 2.4. Valida ion o he P o ein Sco e in NQplus Using a U ina y Bioma ke Ni ogen (N) om 24-h u ine was de e mined in NQplus by he Kjeldahl echnique (Foss Kjel ecTM 2300 analyse , Foss Analy ical, Hille oed, Denma k). This allowed us o e alua e he alidi y o he p o ein E% componen sco e based on FFQ e sus ni ogen exc e ion in u ine, assuming N balance. Adjus men s o incomple e u ine samples we e also done by pa a-aminobenzoic acid [ 30 ]. An a enua ion ac o , usually be ween 0 and 1, was de ined as he co ela ion be ween he sel - epo ed N in akes using he FFQ and he measu ed N losses (assumed o be “ ue” in ake) using he u ina y bioma ke [30,31]. 2.5. Risk Fac o Assessmen In NQplus, as ing enous blood was collec ed, and immedia ely cen i uged and s o ed a −80 ◦C un il u he analyses. Se um c ea inine was de e mined using enzyma ic me hods ia a Dimension Vis a 1500 au oma ed analyse o Roche Modula P800 chemis y analyse (Roche, Basel Swi ze land). HbA1c was de e mined wi h a high-pe o mance liquid ch oma og aphic (HPLC) measu emen echnology using an ADAMSTM A1c HA-8160 analyse (A. Mena ini Diagnos ics, Flo ence, I aly). eGFR was es ima ed using he ch onic kidney disease epidemiology collabo a ion c ea inine equa ion (CKD-EPI) [32]. In Li elines, as ing blood samples we e p ocessed on he day o collec ion and s o ed a − 80 ◦ C un il analysis. Se um c ea inine was measu ed on a Roche Modula P chemis y analyse (G enzache s asse, Swi ze land). The HbA1c le el was measu ed using a u bidime ic inhibi ion immunoassay (COBAS INTEGRA 800 CTS analyse ; Roche Diagnos ics, Alme e, he Ne he lands), bu s anda dised agains he e e ence me hod o he In e na ional Fede a ion o Clinical Chemis y and Labo a o y Medicine (IFCC). eGFR was es ima ed using he ch onic kidney disease epidemiology collabo a ion c ea inine equa ion (CKD-EPI) [32]. In he Young Finns S udy, as ing blood samples we e collec ed and s o ed a − 70 ◦ C un il analysis. Se um c ea inine was de e mined pho ome ically (C ea inine eagen , Dublin, I eland) on an AU400 analyse (Olympus, Tokyo, Japan). eGFR was es ima ed using he Modi ica ion o Die in Renal Disease (MDRD) o mula [ 33 ]. The HbA1c ac ion in blood was measu ed by an ARCHITECT ci8200 analyse (Abbo Labo a o ies, Abbo Pa k, IL, USA). The concen a ion o HbA1c was measu ed Nu ien s 2017,9, 763 5 o 14 immuno u bidime ically wi h he mic opa icle agglu ina ion inhibi ion me hod (HbA1c eagen ; The mo Fishe Scien i ic, Wal ham, MA, USA). 2.6. S a is ical Analyses All s a is ical analyses we e pe o med using R e sion 3.2.0 [ 34 ], and IBM SPSS S a is ics 22. Sepa a e bu simila analyses as de ailed below we e ca ied ou o each o he h ee s udies. Associa ions be ween o al p o ein sco e, HbA1c and eGFR we e e alua ed by means o analysis o co a iance (ANCOVA). Fu he mo e, associa ions be ween he single componen s o he p o ein sco e (plan o animal p o ein a io, animal p o ein (E%), plan p o ein (E%), and o al p o ein in ake (E%)) and HbA1c and eGFR we e in es iga ed using ANCOVA. The p o ein sco e and i s componen s we e adjus ed o o al ene gy in ake by means o he esidual me hod [ 35 ] be o e en e ing he analyses. Bo h unadjus ed and adjus ed analyses o associa ions be ween o al p o ein sco e, HbA1c and eGFR we e pe o med. The adjus men o possible con ounde s included: age (yea s), gende (male/ emale), educa ion (low, medium and high), alcohol (0 g/day, 0–6 g/day, 6–12 g/day, ≥12 g/day ), smoking s a us (ne e , o me , cu en <10 ciga e es/day, cu en ≥ 10 ciga e es/day) and exe cise. Low educa ion mean no educa ion o p ima y educa ion; medium educa ion mean lowe o p epa a o y oca ional educa ion, lowe gene al seconda y educa ion, in e media e oca ional educa ion o app en iceship, highe gene al seconda y educa ion, o p e-uni e si y seconda y educa ion; high educa ion mean highe oca ional educa ion o uni e si y). Smoking s a us was included as a con ounding ac o , because smoking has been associa ed wi h inc eased T2D isk [ 36 ] and may be an independen isk ac o in he p og ession o kidney disease [ 37 ]. Ligh in ense, mode a e in ense, and in ense physical ac i i y we e measu ed in he me abolic equi alen o he ask in minu es pe week. Po en ial e ec modi ica ion was assessed h ough s a i ied analyses o age, gende , and BMI ca ego ies. A p- alue < 0.05 was conside ed s a is ically signi ican . 3. Resul s 3.1. Cha ac e is ics o he S udy Popula ions The a e age p o ein die sco e anged om 8.0 in he lowe qua ile o 12.0 in he uppe qua ile in NQplus, Li elines and he Young Finns S udy (Table 1). In NQplus, Li elines and he Young Finns S udy, 50% o he subjec s had a p o ein sco e be ween 8.0–12.0. The mean daily p o ein in ake anges we e 13.1–15.9 E% in NQplus, 13.3–16.0 E% in Li elines, and 16.0–19.0 E% in he Young Finns S udy. The median age o 39.0 yea s in he Young Finns S udy was lowe compa ed o NQplus (53.0 yea s) and Li elines (44.0 yea s). In he Young Finns S udy, median physical ac i i y was 792 MET min/week, as compa ed wi h NQplus (2376 MET min/week) and Li elines (2205 MET min/week) . Fu he mo e, he median daily in ake o animal p o ein was la ge in he Young Finns S udy (12.3 E%), as compa ed wi h NQplus (7.7 E%) and Li elines (8.4 E%). In addi ion, he median in ake o ce eals was highe in bo h NQplus (190 g/day) and Li elines (181 g/day) han in he Young Finns S udy (124 g/day) (Table 1). In NQplus, we e alua ed he alidi y o he p o ein sco e agains ni ogen exc e ion in u ine, a e adjus ing o incomple e u ine samples. The a enua ion ac o o he p o ein sco e was 0.48, indica ing a 48% a e age conco dance be ween sel - epo ed p o ein in ake and he u ine bioma ke o p o ein in ake. Table 1. Cha ac e is ics o he s udy popula ions (n=76,777). S udy NQplus (n= 492) Li elines (n= 75,131) Young Finns S udy (n= 1154) Age-yea s 53.0 (44.0; 60.0) 42.0 (32.0; 49.0) 39.0 (33.0; 42.0) Age <44 yea s % (no.) 24.6 (121) 59.9 (44,992) 83.3 (961) Age >44 yea s % (no.) 75.4 (371) 40.1 (30,139) 16.7 (193) Males-% (no.) 61.0 (312) 38.8 (29,145) 59.7 (691) Nu ien s 2017,9, 763 6 o 14 Table 1. Con . S udy NQplus (n= 492) Li elines (n= 75,131) Young Finns S udy (n= 1154) Age <44 yea s (no.) (M/F) (265/227) (16,969/28,023) (574/387) Age >44 yea s (no.) (M/F) (47/74) (12,176/17,963) (117/76) Educa ion-% (no.) - Low 0.4 (2) 1.7 (1253) 56.7 (655) - Medium 42.3 (208) 63.9 (48,038) 22.95 (264) - High 57.3 (282) 34.2 (25,729) 20.4 (235) Body mass index-kg/m224.8 (22.7; 26.9) 24.7 (22.6; 27.3) 25.0 (22.4; 27.8) Physical ac i i y-MET–min/week 2376 (1380; 2950) 2205 (1260; 3582) 792 (180.0; 1899) Smoking s a us-% (no.) - Ne e 54.7 (268) 50.3 (37,821) 53.0 (611) - Fo me 37.4 (184) 27.8 (20,895) 23.8 (275) - Cu en <10 ciga e es/day 3.4 (17) 10.5 (7881) 15.5 (179) - Cu en ≥10 ciga e es/day 4.5 (22) 11.4 (8534) 7.7 (89) HbA1c (mmol/mol) 35.5 (33.3; 37.0) 36 (34.0; 38.0) 36.0 (34.0; 38.0) eGFR (mL/min/1.73 m2)91.1 (81.7; 100.1) 100.0 (89.3; 109.5) 100.5 (90.2; 107.9) Ene gy in ake-calo ies/day 2039 (1728; 2456) 1983 (1639; 2386) 2217 (1834; 2653) P o ein die sco e 10.0 (8.0; 12.0) 10.0 (8.00; 12.0) 10.0 (9.0; 12.0) To al p o ein in ake - To al p o ein-(E%/day) 14.6 (13.1; 15.9) 14.6 (13.3; 16.0) 17.4 (16.0; 19.0) - P o ein-(g/kg body weigh /day) 0.99 (0.79; 1.17) 0.95 (0.78; 1.13) 1.3 (1.1; 1.6) Animal p o ein-(E%/day) 7.7 (6.3; 9.3) 8.4 (7.1; 9.9) 12.3 (10.8; 14.0) Plan p o ein-(E%/day) 6.6 (5.9; 7.4) 6.1 (5.5; 6.7) 3.8 (3.2; 4.3) Plan o animal p o ein a io 0.9 (0.7; 1.1) 0.72 (0.58; 0.90) 0.3 (0.2; 0.4) Red mea -(g/day) 34.8 (19.2; 45.5) 38.6 (23.4; 50.5) 79.4 (55.5; 106.3) P ocessed mea -(g/day) 16.9 (6.1; 31.1) 17.6 (8.4; 31.1) 38.1 (23.0; 64.5) Poul y-(g/day) 9.5 (5.3; 15.0) 9.6 (6.1; 15.0) 26.0 (20.3; 60.8) Fish-(g/day) 14.5 (8.2; 16.2) 10.5 (4.2; 16.4) 24.0 (16.6; 36.2) Eggs-(g/day) 8.9 (7.1; 17.9) 7.2 (4.5; 17.9) 19.3 (13.9; 25.9) Dai y-(g/day) 324.0 (211.8; 456.6) 295.7 (186.0; 426.1) 546.6 (343.4; 779.7) Ce eals-(g/day) 190.4 (140.3; 254.0) 181.2 (137.5; 234.0) 123.9 (91.8; 166.7) Legumes-(g/day) 35.7 (19.2; 67.3) 4.4 (0.0; 13.3) 8.2 (4.7; 12.5) To al ca bohyd a e - To al ca bohyd a e-(E%/day) 43.3 (39.7; 47.1) 45.2 (41.8; 48.7) 46.1 (42.4; 49.8) - To al ca bohyd a e-( g/day) 220 (178.5; 269.2) 223.3 (182.0; 271.4) 254.7 (206.7; 309.2) To al a - To al a -(E%/day) 35.6 (31.7; 39.2) 35.4 (32.1; 38.4) 32.6 (29.5; 35.6) - To al a -(g/day) 80.8 (62.9; 102.6) 77.6 (61.4; 96.4) 79.6 (64.4; 96.8) Alcohol consump ion-(g/day) % (no) - 0 g/day 3.9 (19) 2.5 (1875) 0.8 (9) - >0–6 g/day 41.7 (205) 54.4 (40,775) 56.2 (646) - 6–12 g/day 21.1 (104) 23.0 (17,243) 24.6 (283) - >12 g/day 33.3 (164) 20.3 (15,238) 18.4 (212) Glycemic index 53.1 (50.7; 55.4) 56.0 (54.1; 57.8) 51.0 (48.3; 53.8) Glycemic load 117.4 (93.1; 146.7) 125.0 (100.8; 153.2) 129.8 (104.7; 159.9) Cha ac e is ics a e shown as median and IQR: in e qua ile ange o as % (no.). eGFR: es ima ed glome ula il a ion a e; E%: ene gy pe cen age; g: g am; HbA1c: glyca ed hemoglobin; low educa ion: no educa ion o p ima y educa ion; medium educa ion: lowe o p epa a o y oca ional educa ion, lowe gene al seconda y educa ion, in e media e oca ional educa ion o app en iceship, highe gene al seconda y educa ion, o p e-uni e si y seconda y educa ion; high educa ion: highe oca ional educa ion o uni e si y; F: emales; M: males; MET: me abolic equi alen o ask. 3.2. Associa ion be ween P o ein Die Sco e and HbA1c The e we e no associa ions be ween he p o ein die sco e and HbA1c in ei he NQplus, Li elines o he Young Finns S udy. A e adjus men s, an in e se associa ion be ween he p o ein die sco e and HbA1c (slope −0.02 ±0.01 mmol/mol, p< 0.001, Table 2) was seen in Li elines. Nu ien s 2017,9, 763 7 o 14 Table 2. Associa ions o glyca ed haemoglobin (HbA1c) and es ima ed glome ula il a ion a e (eGFR) (es ima e ± SE) ac oss quin iles ene gy-adjus ed p o ein die sco e and quin iles o i s ene gy-adjus ed componen s. S udy NQplus (n= 492) Li elines (n= 75,131) ** Young Finns S udy (n= 1154) Va iable Slope ±SE p-Value R2Slope ±SE p-Value R2Slope ±SE p-Value R2 HbA1c P o ein die sco e - Unadjus ed −0.04 ±0.04 0.368 0.002 −0.0003 ±0.01 0.95 0 0.01 ±0.03 0.743 0 - Adjus ed −0.02 ±0.04 0.722 0.229 −0.02 ±0.01 <0.001 0.154 −0.02 ±0.03 0.635 0.088 To al p o ein (E%) - Unadjus ed 0.03 ±0.06 0.671 0.0003 0.07 ±0.01 <0.001 0.002 0.05 ±0.04 0.211 0.001 - Adjus ed −0.04 ±0.06 0.480 0.229 −0.03 ±0.01 <0.001 0.154 0.0006 ±0.02 0.988 0.088 Plan o animal p o ein a io - Unadjus ed −0.49 ±0.27 0.072 0.006 −0.19 ±0.039 <0.001 0.001 −0.44 ±0.66 0.505 0.0003 - Adjus ed −0.06 ±0.26 0.823 0.229 −0.04 ±0.03 0.142 0.154 −0.17 ±0.71 0.810 0.089 eGFR P o ein die sco e - Unadjus ed 0.53 ±0.21 0.02 0.013 −0.08 ±0.02 <0.0001 0.000 −0.16 ±0.12 0.164 0.002 - Adjus ed 0.32 ±0.18 0.074 0.447 0.17 ±0.02 <0.0001 0.391 0.08 ±0.12 0.474 0.152 To al p o ein (E%) - Unadjus ed −0.74 ±0.31 0.02 0.012 −0.601 ±0.03 <0.0001 0.007 0.12 ±0.15 0.445 0.0005 - Adjus ed −0.56 ±0.25 0.02 0.449 0.08 ±0.02 <0.0001 0.390 −0.02 ±0.15 0.904 0.152 Plan o animal p o ein a io - Unadjus ed 6.46 ±1.29 <0.0001 0.048 1.32 ±0.09 <0.0001 0.003 −2.28 ±2.52 0.367 0.0007 - Adjus ed 3.96 ±1.04 <0.001 0.460 1.14 ±0.10 <0.001 0.391 2.17 ± −2.61 0.405 0.152 Change in HbA1c (mmol/mol) and eGFR (mL/min/1.73 m 2 ) espec i ely, pe 1 uni change in p o ein sco e. ** n= 69462 due o missing alues o HbA1c model adjus ed o age, gende , educa ion (low/middle/high), alcohol (0 g/day, >0–6 g/day, 6–12 g/day, ≥ 12 g/day), smoking s a us (ne e , o me , cu en <10 ciga e es/day, cu en ≥ 10 ciga e es/day), ligh in ense, mode a e in ense, and in ense physical ac i i y (MET: minu es/week); o al a (E%), GI and BMI. Abb e ia ions: eGFR: es ima ed glome ula il a ion a e, E%: ene gy pe cen age, HbA1c: glyca ed hemoglobin. Nu ien s 2017,9, 763 8 o 14 3.3. Associa ion be ween To al E% and HbA1c The e was a posi i e associa ion be ween o al p o ein (E%) and HbA1c in Li elines (slope 0.07 ±0.01 mmol/mol ,p< 0.001, Table 2). A e adjus men s, we ound an in e se associa ion be ween he o al p o ein in ake (E%) and HbA1c in Li elines (slope − 0.03 ± 0.01 E%/mmol/mol, p< 0.001, Table 2). In NQplus and he Young Finns S udy, no associa ions we e ound, nei he in he basic model no a e adjus men s. 3.4. Associa ion o Plan o Animal P o ein Ra io and HbA1c In Li elines, we ound an in e se associa ion (slope − 0.19 ± 0.04 mmol/mol, p< 0.001, Table 2) be ween he plan o animal p o ein a io and HbA1c in he unadjus ed model. A e adjus men s, his associa ion did no emain. 3.5. Associa ion be ween P o ein Die Sco e and eGFR Fo eGFR, we ound a posi i e associa ion be ween he p o ein die sco e and eGFR in NQplus (slope 0.53 ± 0.21 mL/min/1.73 m 2 ,p= 0.02, Table 2), and an in e se associa ion be ween he p o ein die sco e and eGFR in Li elines (slope − 0.08 ± 0.02 mL/min/1.73 m 2 ,p< 0.0001, Table 2). A e adjus men s, a posi i e associa ion was only de ec ed in Li elines (slope 0.17 ± 0.02 mL/min/1.73 m 2 , p< 0.0001, Table 2). We ound no signi ican associa ion be ween he p o ein die sco e and eGFR in he Young Finns S udy. 3.6. Associa ion To al P o ein E% and eGFR Be o e adjus men s, we ound an in e se associa ion in bo h NQplus (slope − 0.74 ± 0.31 E%/mL/min/1.73 m2 ,p= 0.02, Table 2) and Li elines (slope − 0.60 ± 0.03 E%/mL/min/1.73 m 2 , p< 0.0001 , Table 2), be ween he o al p o ein in ake (E%) and eGFR. Fo Li elines, he associa ion changed in o a posi i e associa ion a e adjus men s (slope 0.08 ± 0.02 E%/mL/min/1.73 m 2 , p< 0.0001 , Table 2). Howe e , in NQplus (slope − 0.56 ± 0.25 E%/mL/min/1.73 m 2 , p= 0.02 , Table 2), he in e se associa ion pe sis ed a e u he adjus men s. 3.7. Associa ion be ween Plan o Animal P o ein Ra io and eGFR Adjus ed analyses showed a posi i e associa ion be ween he plan o animal p o ein a io and eGFR in bo h NQplus (slope: 3.96 ± 1.04 mL/min/1.73 m 2 ,p< 0.001, Table 2) and Li elines (slope: 1.14 ±0.10 mL/min/1.73 m2 ,p< 0.001, Table 2). Unadjus ed analyses also showed posi i e associa ions. Fo he adjus ed analyses, he a iabili y o R-squa ed anged om 0.00–0.23 o p o ein sco e e sus. HbA1c and be ween 0.00–0.45 o p o ein sco e e sus eGFR, espec i ely. The ela ionship be ween he p o ein die sco e and ei he HbA1c o eGFR was no quali a i ely modi ied by age, gende , o BMI (Table 3). Table 3. S a i ied analyses o he associa ions o HbA1c and eGFR (es ima e ±SE). Va iable NQPlus (n= 492) Li elines (n= 75,131) Young Finns S udy (n= 1154) Slope ±SE Slope ±SE Slope ±SE HbA1c Age <44 yea s −0.073 ±0.104 0.149 ±0.022 * −0.013 ±0.033 ≥44 yea s −0.250 ±0.036 * 0.219 ±0.024 * 0.016 ±0.094 Gende Men −0.012 ±0.058 −0.016 ±0.008 * −0.008 ±0.039 Women 0.0004 ±0.072 −0.020 ±0.006 * 0.005 ±0.056 Nu ien s 2017,9, 763 9 o 14 Table 3. Con . Va iable NQPlus (n= 492) Li elines (n= 75,131) Young Finns S udy (n= 1154) Slope ±SE Slope ±SE Slope ±SE BMI <25 kg/m2−0.009 ±0.055 −0.012 ±0.006 −0.039 ±0.041 25–30 kg/m2−0.021 ±0.076 −0.033 ±0.008 * 0.005 ±0.048 ≥30 kg/m2−0.012 ±0.044 0.005 ±0.018 −0.012 ±0.031 eGFR Age <44 yea s 0.040 ±0.414 −0.019 ±0.006 * 0.131 ±0.129 ≥44 yea s 0.424 ±0.199 −0.012 ±0.008 −0.053 ±0.275 Gende Men 0.191 ±0.277 0.189 ±0.022 * 0.016 ±0.165 Women 0.124 ±0.319 0.137 ±0.022 * −0.098 ±0.181 BMI <25 kg/m20.227 ±0.234 0.164 ±0.022 * −0.014 ±0.167 25–30 kg/m20.252 ±0.299 0.179 ±0.03 * −0.024 ±0.193 ≥30 kg/m20.858 ±0.822 0.117 ±0.049 * 0.621 ±0.333 Change in age (yea s), gende , and BMI (kg/m 2 ) espec i ely, pe 1 uni change in p o ein sco e. Model adjus ed o age, gende (excep o gende -s a i ied models), educa ion (low/middle/high), alcohol (0 g/day, >0–6 g/day, 6–12 g/day , ≥ 12 g/day), smoking s a us (ne e , o me , cu en <10 ciga e es/day, cu en ≥ 10 ciga e es/day), ligh in ense, mode a e in ense, and in ense physical ac i i y (MET: minu es/week), Fa (en%), GI and BMI (kg/m 2 ); *p< 0.05. 4. Discussion In he cu en c oss-sec ional s udy, we aimed (1) o de elop a p o ein sco e cap u ing bo h ela i e quan i y and sou ce (plan e sus animal) o die a y p o ein and (2) o examine he associa ion be ween his sco e and ma ke s o diabe es and enal unc ion. We de eloped a p o ein sco e wi h a maximum ange om 0 o 20 based on he FFQ da a om he s udy popula ions in es iga ed. In p ac ice, he sco es anged be ween 8.0 and 12.0 ac oss all h ee popula ions. The signi ican associa ions shown in Li elines we e pa ly ep oduced by NQplus and he Young Finns S udy, al hough in hese smalle sample size s udies, ends we e mo e he e ogeneous and a ec ed by adjus men o po en ial con ounde s. In pa icula , he associa ions be ween he o al p o ein sco e and bo h isk ma ke s HbA1c and eGFR showed consis en pa e ns ac oss all coho s when compa ed o esul s o he sepa a e componen s, i.e., animal and plan p o ein le els (Table S1). A highe o al p o ein die sco e was associa ed wi h a lowe HbA1c and an inc ease in eGFR a e adjus men o he po en ial con ounde s in Li elines, bu wi h no signi ican ela ionship in NQplus o he Young Finns S udy. Die a y p o ein has p e iously been shown o be bene icial, leading o be e glycemic con ol in T2D [ 38 ]; hence, a highe p o ein die has been ad oca ed o glycemic con ol in indi iduals wi h T2D [ 39 ]. Howe e , when e alua ing he associa ion be ween a highe p o ein sco e and HbA1c, i mus also be aken in o accoun ha die a y p o ein will always subs i u e ei he ca bohyd a e and/o a . In a ecen s udy, e en a ela i ely small a ia ion in he p opo ion o a and ca bohyd a es we e signi ican ly associa ed wi h me abolic isk ac o s in pa ien s wi h T2D [ 40 ]. Al hough we adjus ed o a E%, esidual con ounding by o he subs ances ela ed o mea in ake migh be p esen . P e ious s udies ha e sugges ed se e al plausible mechanisms linking ed and p ocessed mea me aboli es, including sodium, heme i on, sa u a ed a y acids, ad anced glyca ion end p oduc s, ni i es and ni a es, o an inc eased isk o T2D [ 41 – 43 ]. In Li elines, he p o ein sco e, p o ein in ake and plan p o ein we e associa ed wi h lowe HbA1c. The op imal amoun and quali y o p o ein o p e en ion o T2D is s ill con o e sial [ 44 ]. In a ecen s udy by Vi anen e al., high p o ein in ake was no independen ly associa ed wi h isk o T2D, bu he quali y o p o ein was o impo ance, a ou ing plan o e animal p o ein in he p e en ion o T2D [ 44 ]. In he s udy o Malik e al. [ 10 ], conduc ed