Changes in visual memory in mild cognitive impairment: a longitudinal study with CANTAB

Co esponding au ho : Ma ía Campos-Magdaleno, Depa men o De elopmen al Psychology, Rúa Xosé Ma ía Suá ez Núñez, s/n. Campus Su . 15782 San iago de Compos ela, (Galicia, Spain). Email: [email p o ec ed] 1 Ti le: Changes in isual memo y in mild cogni i e impai men . A longi udinal s udy wi h CANTAB. Au ho s: Ma ía Campos-Magdaleno1, Da id Lei a2, A u o X. Pe ei o1, C is ina Lojo- Seoane1, Sabela C. Mallo1, Da id Facal1 & Onésimo Juncos-Rabadán1 Filia ions: 1. Depa men o De elopmen al Psychology, Uni e si y o San iago de Compos ela, Galicia, Spain. 2. Depa men o Me hodology o Beha iou al Sciences, Uni e si y o Ba celona, Ca alunya, Spain Numbe o wo ds: 5658 Co esponding au ho : Ma ía Campos-Magdaleno, Depa men o De elopmen al Psychology, Rúa Xosé Ma ía Suá ez Núñez, s/n. Campus Su . 15782 San iago de Compos ela, (Galicia, Spain). Email: [email p o ec ed] 2 Abs ac Backg ound: Mild cogni i e impai men (MCI), as a s age in he cogni i e con inuum be ween no mal aging and demen ia, is mainly cha ac e ized by memo y impai men . The aims o his s udy we e o examine CANTAB measu es o empo al changes o isual memo y in MCI and o e alua e he use ulness o he baseline sco es o p edic ing changes in cogni i e s a us. Me hods: The s udy included 201 pa icipan s aged o e 50 yea s wi h subjec i e cogni i e complain s. Visual memo y was assessed wi h ou CANTAB es s (PAL, DMS, PRM and SSP) adminis e ed a baseline and on wo u he occasions, wi h a ollow-up in e al o 18-24 mon hs. Pa icipan s we e di ided in o h ee g oups acco ding o he change in hei cogni i e s a us: pa icipan s wi h subjec i e cogni i e complain s who emained s able (SCC-S able), MCI pa icipan s who emained s able (MCI-S able) and MCI pa icipan s whose cogni i e de e io a ion con inued (MCI- Wo sened). Linea Mixed Models we e used o model longi udinal changes, wi h e alua ion ime as a ixed a iable, and mul inomial eg ession models we e used o p edic changes in cogni i e s a us. Resul s: Isola ed signi ican e ec s we e ob ained o Age and G oup wi h all CANTAB es s used. In e ac ions be ween E alua ion Time and G oup we e iden i ied in he PAL and DMS es s, indica ing di e en empo al pa e ns depending on he changes in cogni i e s a us. Reg ession models also indica ed ha CANTAB sco es we e good p edic o s o changes in cogni i e s a us. Conclusions: Decline in isual memo y measu ed by PAL and DMS es s can success ully dis inguish di e en ypes o MCI, and conside ed oge he PAL, DMS, PRM and SSP can p edic changes in cogni i e s a us. Co esponding au ho : Ma ía Campos-Magdaleno, Depa men o De elopmen al Psychology, Rúa Xosé Ma ía Suá ez Núñez, s/n. Campus Su . 15782 San iago de Compos ela, (Galicia, Spain). Email: [email p o ec ed] 3 Keywo ds: isual memo y; CANTAB es s; mild cogni i e impai men ; linea mixed models; longi udinal assessmen . In oduc ion Cogni i e decline in he elde ly can be conside ed a con inuum anging om a cogni i ely unimpai ed s a e (CU), o he p esence o subjec i e cogni i e complain s (SCC) wi hou objec i e cogni i e impai men , also called Subjec i e Cogni i e Decline (SCD) (Jessen e al., 2016; Molinue o e al., 2017), ollowed by Mild Cogni i e Impai men (MCI), cha ac e ized by p esence o cogni i e complain s, objec i e cogni i e de e io a ion and p ese a ion o minimal impai men o ins umen al ac i i ies o daily li ing (Pe e sen, 2004; Pe e sen e al., 2018), and inally demen ia, which is cha ac e ized by cogni i e and beha iou al symp oms ha impai no mal unc ioning in daily li e (APA, 2013). The single and mul iple domain sub ypes o amnes ic and non-amnes ic MCI ha in ol e de e io a ion in only one o in mo e han one cogni i e domain may also ep esen di e en le els o cogni i e decline, wi h he mul iple domain sub ype being he mos ex eme clinical s a e (B amba i e al., 2009; Han e al., 2012). P og ession along he con inuum is a complex p ocess cha ac e ized by cogni i e changes, ansi ions and diagnos ic ins abili y a SCD and MCI s ages, con e sion o demen ia and eco e y o CU (Facal, Guà dia-Olmos & Juncos- Rabadán 2015; Pe e sen e al., 2018). Howe e , aking he ins abili y in o accoun , MCI and he sub ypes cha ac e ized by only memo y impai men s (amnes ic single-domain) o by impai men s in memo y and in o he cogni i e domains (amnes ic mul i-domain) a e conside ed high- isk s a es o p og ession o demen ia, mainly Alzheime ’s Disease (AD). Ea ly de ec ion o he di e en s ages o cogni i e decline and he p og ess o decline is a p essing esea ch challenge in he p e en ion and ea men o demen ia (Albe e al., 2011; Pe e sen e al., 2018). Co esponding au ho : Ma ía Campos-Magdaleno, Depa men o De elopmen al Psychology, Rúa Xosé Ma ía Suá ez Núñez, s/n. Campus Su . 15782 San iago de Compos ela, (Galicia, Spain). Email: [email p o ec ed] 4 P e ious s udies ha e shown ha isual memo y impai men can di e en ia e MCI pa ien s om cogni i ely unimpai ed con ols (Alescio-Lau ie e al., 2007; Ba beau e al. 2008; Juncos-Rabadán, Facal, Pe ei o & Lojo-Seoane, 2014a; Wes e be g e al., 2013). O he s udies ha e success ully p edic ed he p og ession om MCI o AD (De Anna e al., 2014; De ancesco e al., 2013; Didic e al, 2013; Ol a-Cuca ella e al., 2018; Reijs e al, 2017; Sax on e al., 2004) and e en comple e neu odegene a i e p og ess om he cogni i ely impai ed s a e o MCI and AD (Mis idis, K umm, Monsch, Be es & Taylo , 2015). These indings indica e he impo ance o including eliable isual memo y es s o diagnosing MCI and o s udying he cou se o decline in di e en aspec s o isual memo y in p og ession o AD. Compu e ized assessmen o isual memo y using he Camb idge Neu opsychological Tes Au oma ed Ba e y (CANTAB; Camb idge Cogni ion L d., 2012; Sahakian e al., 1988) has been used o di e en ia e con ols, MCI and AD pa icipan s in c oss-sec ional s udies (Alladi, A nold, Mi chell, Nes o & Hodges 2006; de Ro e e al., 2011; Juncos-Rabadán e al, 2014a; Junkkila, Oja, Laine & Ka asch, 2012; Swaison e al., 2001). CANTAB includes es s ha assess isual Episodic Memo y (EM) and isual Wo king Memo y (WM). Bo h ypes o memo y ha e been shown o be impai ed ea ly on in AD (Belle ille, Syl ain-Roy, de Boysson & Ména d, 2008; Economou, Papageo giou & Ka ageo giou, 2011; an Geld op e al., 2015). De e io a ion in EM has been ound o be a pa icula ly s ong p edic o o p og ession o AD (Belle ille e al., 2008; Landau e al., 2010). Longi udinal e idence om esea ch using he CANTAB isual memo y es s emains sca ce (Cacciamani e al., 2017; Juncos-Rabadán e al., 2016; Mi chell, A nold, Dawson, Nes o & Hodges, 2009; Summe s and Saunde s, 2012). Summe s and Saunde s (2012) ound ha he decline in isual memo y pe o mance assessed wi h CANTAB measu es (Pai ed Associa es Lea ning, Spa ial Span, Spa ial Wo king Memo y) in combina ion wi h he Rey Co esponding au ho : Ma ía Campos-Magdaleno, Depa men o De elopmen al Psychology, Rúa Xosé Ma ía Suá ez Núñez, s/n. Campus Su . 15782 San iago de Compos ela, (Galicia, Spain). Email: [email p o ec ed] 5 Audi o y Ve bal Lea ning Tes iden i ied 100% o cases o MCI pa ien s who p og essed o AD a e 20 mon hs. Howe e , Cacciamani e al. (2017) epo ed imp o emen s in Spa ial Wo king Memo y, Spa ial Recogni ion Memo y and Pai ed Associa ed Lea ning a e a ollow-up pe iod o 12 mon hs in a small sample o MCI pa ien s. Fu he in es iga ion including la ge sample sizes and longe in e als be ween assessmen s mus be ca ied ou o analyze he disc iminan alue and e olu ion o hese memo y measu es. The main pu pose o he p esen s udy was o de e mine longi udinal pa e ns o pe o mance o isual memo y CANTAB es s in pa ien s diagnosed a baseline wi h MCI and assessed wice wi h a ollow-up in e al o a ound 18 mon hs o measu e s abili y o de e io a ion o he condi ion. A seconda y aim was o assess he use ulness o baseline CANTAB measu es o p edic ing changes in cogni i e s a us a he inal ollow-up s age. Me hodology Pa icipan s Pa icipan s we e selec ed om he Compos ela Aging S udy (CompAS), an ongoing longi udinal p ojec in ol ing he de ec ion and ollow-up o Mild Cogni i e Impai men in pa ien s wi h subjec i e cogni i e complain s and no p io diagnos ic o demen ia, psychia ic o neu ological diso de s a ending p ima y ca e cen es in Galicia, an au onomous egion in no hwes Spain (Juncos-Rabadán e al., 2012). We selec ed 201 pa ien s aged o e 50 yea s who had comple ed 3 isi s (a Baseline, Time 1 and Time 2) wi h a be ween- es in e al o a ound 18 mon hs. The mean in e al was 18.49 mon hs (3.64 s anda d de ia ion, SD) be ween Baseline and Time 1, 17.72 mon hs (3.81 SD) be ween Time 1 and Time 2, and 36.83 mon hs (5.17 SD) be ween Baseline and Time 2. None o he pa icipan s had p e iously been diagnosed wi h MCI o demen ia, clinical s oke, auma ic b ain inju y, mo o -senso y de ec s, alcohol o d ug abuse/dependence, o any neu ological o psychia ic Co esponding au ho : Ma ía Campos-Magdaleno, Depa men o De elopmen al Psychology, Rúa Xosé Ma ía Suá ez Núñez, s/n. Campus Su . 15782 San iago de Compos ela, (Galicia, Spain). Email: [email p o ec ed] 6 disease. A baseline, pa icipan s we e classi ied as single-domain amnes ic MCI (sda-MCI), mul iple-domain amnes ic MCI (mda-MCI), single-domain non amnes ic MCI (sdna-MCI) o mul iple-domain non amnes ic MCI (mdna-MCI), acco ding o s anda d c i e ia (Albe e al., 2011; Dubois e al., 2007; Pe e sen, 2004). The c i e ia o diagnosis o MCI included he ollowing: (a) sel - epo ed, in o man -co obo a ed conce ns abou cogni ion, assessed by a sho e sion o he subjec i e memo y complain s ques ionnai e (SMCQ; Benede and Seisdedos, 1996); (b) pe o mance o 1.5 s anda d de ia ions (SD) below age and educa ion no ms in one o mo e cogni i e domains, assessed by he subscales o he Spanish e sion o he Camb idge cogni i e examina ion, CAMCOG-R (Huppe e al., 1996; Spanish e sion: López-Pousa, 2003; Pe ei o, Ramos-Lema, Juncos-Rabadán, Facal & Lojo-Seoane, 2015), excep o memo y, assessed by he sho and long delay ee ecall om he Spanish e sion o he Cali o nia e bal lea ning es (Delis e al., 1987; Spanish e sion: Benede and Alejand e, 1998); (c) no signi ican o minimal impac on ac i i ies o daily li ing, assessed by ins umen al ac i i ies o daily li ing scale (Law on and B ody, 1969); and (d) he absence o demen ia as es ablished by he DSM-IV and NINCDS-ADRDA c i e ia. Pa icipan s pe o ming as cogni i ely no mal adul s in gene al unc ioning and speci ic domain es s, acco ding o no ms by age and yea s o educa ion, and p esen ing subjec i e cogni i e complain s (SCC), we e included in he SCC g oup. This g oup me he ollowing c i e ia: (a) a ending p ima y ca e heal h cen es wi h sel - epo ed cogni i e conce ns; and (b) con i ma ion o hese conce ns by he sho Spanish e sion o he ques ionnai e o subjec i e memo y complain s (Benede and Seisdedos, 1996) adminis e ed o pa icipan s and a amily membe . The SCC g oup was conside ed a con ol g oup. All diagnoses we e eached by consensus a a special mee ing o he esea ch eam. In each successi e ollow-up assessmen , pa icipan s we e eclassi ied as SCC, sda- MCI, mda-MCI, sdna-MCI, mdna-MCI and p obable demen ia (DSM-IV and NINCDS- Co esponding au ho : Ma ía Campos-Magdaleno, Depa men o De elopmen al Psychology, Rúa Xosé Ma ía Suá ez Núñez, s/n. Campus Su . 15782 San iago de Compos ela, (Galicia, Spain). Email: [email p o ec ed] 7 ADRDA) by applying he same c i e ia as a baseline. A he hi d e alua ion, pa icipan s we e classi ied in o h ee g oups acco ding o he changes in hei cogni i e s a us: pa icipan s wi h SCC a Baseline who emained s able a Time 2 (SCC-s able g oup, n=148, 71.49%); pa icipan s diagnosed wi h MCI a Baseline who emained s able a Time 2 (MCI- s able g oup, n= 31, 15.45%); and pa icipan s diagnosed as sda-MCI o sdna-MCI a Baseline who p og essed o mda-MCI, mdna-MCI o demen ia a Time 1 o Time 2 (MCI-wo sened g oup, n= 22, 13.04%). P obable AD o o he ypes o demen ia we e diagnosed acco ding o he DMS-IV and NINCDS-ADRDA c i e ia, and p og ession o demen ia was con i med by consul a ion o he medical his o y and eco ding he da e o neu ological diagnosis. We assumed, in acco dance wi h B amba i e al. (2009) and Campos-Magdaleno, Díaz-Bó eda, Juncos-Rabadán, Facal & Pe ei o (2016), ha he change om single-domain o mul iple- domain co esponds o cogni i e wo sening, in which mul i-domain MCI ep esen s he mos se e ely impai ed o he MCI sub ypes. All pa icipan s ga e hei w i en in o med consen p io o pa icipa ion in he s udy. The esea ch p ojec was app o ed by he Galician E hics Commi ee o Clinical Resea ch (Xun a de Galicia, Spain), and he s udy was pe o med in acco dance wi h he e hical s anda ds es ablished in he 1964 Decla a ion o Helsinki and e ised in Seoul 2008. Ma e ials and P ocedu e Fou CANTAB isual memo y es s we e adminis e ed: pai ed associa es lea ning (PAL), pa e n ecogni ion memo y (PRM), delayed ma ching o sample (DMS) and spa ial span (SSP). The PAL es assesses isuospa ial episodic memo y and lea ning (Sahakian e al., 1988). One o mo e boxes con aining a pa e n a e displayed on he sc een and a e opened in a andom o de . The pa e ns shown in he boxes a e hen displayed in he middle o he sc een, one a a ime, and pa icipan s a e asked o ouch he box in which he pa e n was o iginally loca ed. I he pa icipan makes an e o , he pa e ns a e shown again as a eminde o he loca ions. The Co esponding au ho : Ma ía Campos-Magdaleno, Depa men o De elopmen al Psychology, Rúa Xosé Ma ía Suá ez Núñez, s/n. Campus Su . 15782 San iago de Compos ela, (Galicia, Spain). Email: [email p o ec ed] 8 le el o di icul y (2, 4, 6 and 8 pa e ns) was inc eased h oughou he es s. The ou come a iable was he o al numbe o e o s adjus ed o le el 6, which ep esen s a high le el o di icul y and has been used by se e al esea che s o s udy MCI and AD (Alladi e al., 2006; Chambe lain e al., 2011; Lenehan e al., 2016; Mi chell e al., 2009; Polche e al., 2017). The PRM es assesses isual pa e n ecogni ion memo y in a wo-choice o ced disc imina ion pa adigm (Swainson e al., 2001). The pa icipan s we e p esen ed wi h wo blocks o 12 isual pa e ns, each displayed sepa a ely. In he ecogni ion phase, subjec s a e equi ed o choose be ween a pa e n hey ha e al eady seen and a no el pa e n. The ou come measu e was he pe cen age o co ec esponses, conside ed in some p e ious s udies as a speci ic episodic memo y ou come (de Jage , Milwain & Budge, 2002; Juncos-Rabadán, Pe ei o, Facal, Rebo edo & Lojo-Seoane, 2014b; Na han e al., 2017). Delayed ma ching o sample (DMS) assesses bo h simul aneous and sho - e m isual memo y (Sahakian e al., 1988; Owen e al., 1993). Pa icipan s mus selec he pa e n ha exac ly ma ches he sample om among ou abs ac choices ha include dis ac o s. In some ials, he sample and he choice pa e ns a e shown simul aneously, while in o he s he e is a delay o 0 ms, 4,000 ms o 12,000 ms. The ou come measu e was he pe cen age o co ec esponses, also conside ed an episodic memo y measu e ask (Juncos-Rabadán e al., 2014b, 2016; Sweeney, Kmiec & Kup e , 2000). Spa ial span (SSP) is a compu e ised e sion o he Co si blocks ask ha assesses isual wo king memo y capaci y (Owen e al., 1990). A pa e n o whi e squa es is shown on he sc een. Some o he squa es change colou , one a a ime, in a a iable sequence. A he end o he p esen a ion o each sequence, a one indica es ha he pa icipan should ouch each o he boxes in he same o de ha hey we e o iginally p esen ed. The numbe o boxes in he sequence is inc eased om a le el o wo a he s a o he es un il a inal le el o nine, wi h h ee sequences a each le el. The ou come a iable, he span leng h, was calcula ed o he longes Co esponding au ho : Ma ía Campos-Magdaleno, Depa men o De elopmen al Psychology, Rúa Xosé Ma ía Suá ez Núñez, s/n. Campus Su . 15782 San iago de Compos ela, (Galicia, Spain). Email: [email p o ec ed] 9 sequence success ully ecalled and was used as index o he SSP ask (Saunde s and Summe s, 2010). The ou CANTAB es s we e adminis e ed in a mo e ex ensi e coun e balanced assessmen ca ied ou by ained psychologis s. To con ol o he e ec o isual acui y on pe o mance o he CANTAB, we measu ed he isual acui y o bo h eyes wi h he Ligh house nea isual acui y es . S a is ical analysis C oss-sec ional analyses we e ca ied ou a baseline o socio-demog aphic and p incipal neu opsychological measu es, which we e modelled using non-pa ame ic es s (e.g. K uskal-Wallis and Mann-Whi ney es s) o de e mine di e ences be ween g oups, gi en he skewed empi ical dis ibu ions and he small sample size in some cases. In o de o model longi udinal changes in he CANTAB measu es, we ini ially used (gene alized) linea mixed models -(G)LMM- wi h andom in e cep s and andom slopes. We conside ed ha he in e cep s migh di e acco ding o he memo y ajec o ies o he pa icipan s and ha di e en slopes would ep esen a ious empo al pa e ns o change in memo y pe o mance. We inally disca ded andom slopes in he es ima ed models due o con e gence issues. The s a is ical models included he ollowing independen a iables o p edic o s as ixed e ec s: E alua ion Time (Baseline, Time 1 and Time 2), G oup (SCC-s able, MCI-s able and MCI-wo sened), and hei in e ac ion (E alua ion Time x G oup). By speci ying G oup and E alua ion Time as ixed ac o s we can es pai wise compa isons o he es ima ed ma ginal means o he dependen a iables o each g oup and a each e alua ion ime. As all models included andom e ec s o in e cep s and he e oskedas ici y due o he g oup, he co a ia e age a baseline was s anda dized o enable in e p e a ion o he in e cep . Sepa a e models we e cons uc ed o each dependen a iable: PAL o al e o s adjus ed o 6 shapes, PRM o al pe cen co ec , DMS o al pe cen co ec , and SSP span leng h. The SCC-s able g oup was conside ed he e e ence Co esponding au ho : Ma ía Campos-Magdaleno, Depa men o De elopmen al Psychology, Rúa Xosé Ma ía Suá ez Núñez, s/n. Campus Su . 15782 San iago de Compos ela, (Galicia, Spain). Email: [email p o ec ed] 16 ou comes di e en ia ed pa icipan s who showed no cogni i e impai men (SCC-S able) and pa icipan s wi h MCI, and e en be ween MCI pa icipan s who emained s able o wo sened. The esul s indica e ha assessing isual memo y wi h CANTAB measu es may be use ul o di e en ia ing be ween di e en s ages o MCI in he cogni i e con inuum o demen ia. Es ima ed simple and mul iple mul inomial logis ic models used o assess he u ili y o CANTAB sco es a he ini ial s age o p edic cogni i e e olu ion a he end o he s udy p o ed o ha e a good o e y good p edic i e capaci y (AUCs be ween 0.71 o 0.86; see sec ion S1 o Supplemen a y Ma e ial o u he in o ma ion ega ding he model es ima es and pe o mance). In summa y, he models showed ha he highe he isual memo y sco e he lowe he isk o being classi ied in he g oup wi h he wo s cogni i e ou look. The age o pa icipan s a baseline signi ican ly in luenced pe o mance o all es s o e ime. Olde pa icipan s sco ed lowe on all measu es, ega dless o he diagnos ic g oup (SCC- s able, MCI-s able, MCI-wo sened). The in luence o age on he pe o mance in he CANTAB isual memo y es s o old adul s wi h MCI and wi hou cogni i e impai men has been documen ed in c oss sec ional s udies (Juncos-Rabadán e al., 2014a). The cu en indings add new e idence om a longi udinal design. The s udy indings also show a main e ec o G oup, wi h he MCI-wo sened g oup ob aining he wo s sco es in all CANTAB measu es used a he h ee e alua ion imes. This g oup comp ised pa icipan s wi h g ea e cogni i e impai men , who we e ound o ha e p og essed o mul iple-domain MCI o demen ia a ei he o he ollow-up e alua ions. The p o ile wi h wo s pe o mance in isual memo y es s o mul iple-domain MCI has al eady been shown in p e ious s udies (Juncos-Rabadán e al., 2014b). Ou esul s suppo he capaci y o he CANTAB isual memo y es s o show di e en pe o mance p o iles and disc imina e be ween g oups in he cogni i e con inuum om no mal aging o demen ia, and sugges he use o hese es s o ea ly diagnosis o cogni i e impai men . The indings ob ained wi h CANTAB Co esponding au ho : Ma ía Campos-Magdaleno, Depa men o De elopmen al Psychology, Rúa Xosé Ma ía Suá ez Núñez, s/n. Campus Su . 15782 San iago de Compos ela, (Galicia, Spain). Email: [email p o ec ed] 17 sco es a e consis en wi h some addi ional analyses done o e i y ha cogni i e decline is signi ican ly mo e p onounced in MCI g oups. The indings showed ha he changes di e ed signi ican ly in he h ee s udy g oups and ha he indi iduals included in he MCI-wo sened g oup showed he mos nega i e changes in he gene al cogni i e pe o mance. Rega ding he main e ec o he a iable E alua ion Time, he PAL es was he only measu e ha indica ed signi ican di e ences a he h ee e alua ion momen s in all pa icipan s. This signi ican main e ec adds new e idence o p e ious s udies on he u ili y o he PAL o assess isual memo y and lea ning in old adul s wi h and wi hou cogni i e impai men (Fowle , Saling, Conway, Semple & Louis, 2002; Junkkila e al., 2012; O'Connell e al., 2004; Polche e al., 2017). Mo eo e , ou esul s indica e ha he PAL measu e can de ec changes in longi udinal pe o mance ela ed o e olu ion along a con inuum o cogni i e decline. Taking in o accoun ha longi udinal esea ch is sca ce, his inding is an impo an con ibu ion and adds e idence o he pionee ing wo k by Blackwell and colleagues (Blackwell e al., 2004), who obse ed ha he same CANTAB measu e was signi ican ly co ela ed wi h he deg ee o subsequen cogni i e de e io a ion in he ea ly s ages o AD. The mos in e es ing indings o he p esen s udy a e he signi ican in e ac ions be ween E alua ion Time x G oup in he PAL and DMS measu es. Rega ding he PAL o al e o s adjus ed-6 shapes, he in e ac ion was signi ican o he MCI-s able and he MCI-wo sened g oups, indica ing he exis ence o speci ic longi udinal pa e ns o pe o mance o each. The ma ginal means indica e a small inc ease in e o s in he SCC-s able g oup be ween he baseline and he ollow-up e alua ions, while in bo h MCI g oups he e o s inc eased signi ican ly in he same pe iods. The inc ease was mo e impo an o he MCI-wo sened g oup. The di e ences in PAL empo al pa e ns indica e a decline in he pe o mance o e ime o all g oups; howe e , hey also enable disc imina ion be ween he leas cogni i ely impai ed g oup (SCC-s able) and he MCI g oups, as well as be ween he MCI g oup ha emain s able (MCI- Co esponding au ho : Ma ía Campos-Magdaleno, Depa men o De elopmen al Psychology, Rúa Xosé Ma ía Suá ez Núñez, s/n. Campus Su . 15782 San iago de Compos ela, (Galicia, Spain). Email: [email p o ec ed] 18 s able) and he MCI g oups in which u he de e io a ion occu s (MCI-wo sened). Ou indings add a new pe spec i e o hose epo ed by Cacciamani e al. (2017), who obse ed a ma ked imp o emen in PAL when compa ing he baseline pe o mance wi h he 6-mon h ollow-up, bu no di e ence in pe o mance be ween 6- and 12-mon h ollow-ups. This imp o emen may be he esul o a p ac ice e ec due o he sho ollow-up pe iod; howe e , he p ac ice e ec may disappea when longe ollow-up in e als be ween PAL es s a e used in longi udinal assessmen s. Rega ding he DMS, he E alua ion Time x G oup in e ac ion was only signi ican in he MCI-wo sened g oup, in which he es pe o mance declined o e ime. The pe o mance o he o he wo g oups, SSC-s able and MCI-s able, did no a y signi ican ly. The E alua ion Time x G oup in e ac ion was no signi ican o ei he he PRM o al pe cen co ec o SSP span leng h. Howe e , he es ima ed ma ginal means showed signi ican di e ences be ween SCC-s able and MCI-wo sened g oups, indica ing a clea decline in he la e g oup o e ime. The measu es in which a signi ican E alua ion Time x G oup in e ac ion was obse ed co espond o he wo CANTAB es s (PAL and DMS) mos closely ela ed o episodic memo y (de Jage e al., 2002; Juncos-Rabadán e al., 2014a, 2014b, 2016; Na han e al., 2017; Sweeney e al., 2000). PAL in ol es isuospa ial episodic memo y and lea ning, and DMS in ol es sho - e m memo y o complex isual pa e ns. Decline in episodic memo y has been desc ibed as one o he mos po en p edic o s o p og ession o Alzheime 's disease (Belle ille e al., 2008; Landau e al., 2010), and ou esul s show ha he PAL o al e o s adjus ed-6 shapes and he DMS o al pe cen co ec enable de ec ion o longi udinal changes ha may be indica i e o p og ession in he con inuum o cogni i e de e io a ion. Howe e , he measu es he PRM o al pe cen co ec and he SSP span leng h ha di e ed signi ican ly be ween g oups (G oup main e ec ) did no indica e di e ences be ween g oups o e ime (E alua ion Time x G oup in e ac ion). PRM in ol es memo y and subsequen Co esponding au ho : Ma ía Campos-Magdaleno, Depa men o De elopmen al Psychology, Rúa Xosé Ma ía Suá ez Núñez, s/n. Campus Su . 15782 San iago de Compos ela, (Galicia, Spain). Email: [email p o ec ed] 19 ecogni ion o sequences o isual pa e ns, which may be ela ed o he a en ional span capaci y, which is associa ed wi h wo king memo y. In p e ious s udies, con adic o y indings ega ding span leng h as a measu e o wo king memo y ha di e en ia es pa icipan s acco ding o diagnosis and p og ession ha e been epo ed. While a la ge numbe o s udies suppo he exis ence o impai men in span leng h p io o diagnosis o demen ia (Belle ille e al., 2017; Economou e al., 2006; Gagnon and Belle ille, 2011; an Geld op e al., 2015; Saunde s and Summe s, 2010), o he s udies ob ained con adic o y o non-meaning ul esul s (G i i h e al., 2006; Gua ch, Ma cos, Salame o, Gas ó, & Blesa, 2008; Kessels, O e beek, & Bouman 2015), ques ioning he alue o he measu e o ea ly de ec ion o cogni i e impai men . Ou indings indica e ha he PRM measu e and he SSP span canno di e en ia e longi udinal pa e ns be ween g oups. We conclude ha isual episodic memo y declines in people wi h MCI o e ime and ha his decline may be a cogni i e indica o o he p og ession in he con inuum anging om he s age cha ac e ized by p esence o cogni i e complain s wi hou objec i e cogni i e impai men o demen ia, h ough he di e en le els o se e i y o MCI. PAL o al e o s adjus ed-6 shapes ou come, and DMS pe cen co ec o al measu es di e en ia e he changes in pa icipan s in he con inuum o cogni i e de e io a ion: people wi h and wi hou objec i e de e io a ion, and people who wo sen o emain s able o e ime. In addi ion, he be ween-e alua ion in e als used in longi udinal s udies should be wide enough o p e en p ac ice e ec s. Membe ship o g oups cha ac e ized by change in cogni i e s a us de eloped a he second ollow-up s age (T 2) has p o en o be accu a e in he ligh o di e en ypes o e idence. Fi s , a di e en pa e n o change was obse ed in CANTAB measu emen s acco ding o his classi ica ion. Secondly, di e en pa e ns o change we e also obse ed in o he cogni i e sco es such as MMSE and CAMCOG-R when compa ing he g oups included in his s udy. Finally, compa ison o membe ship in g oups ob ained by he p ocedu e desc ibed in his s udy Co esponding au ho : Ma ía Campos-Magdaleno, Depa men o De elopmen al Psychology, Rúa Xosé Ma ía Suá ez Núñez, s/n. Campus Su . 15782 San iago de Compos ela, (Galicia, Spain). Email: [email p o ec ed] 20 wi h a classi ica ion ob ained by means o non-pa ame ic clus e ing o mul i a ia e ajec o ies (i.e. indi idual ajec o ies in he 4 CANTAB sco es) e ealed a simila i y index o 0.74, which indica es a good le el o ag eemen . In summa y, we demons a ed ha he isual CANTAB sco es a) a e use ul o p edic ing cogni i e e olu ion in he ime-pe iod included in his s udy, b) di e o e ime depending on he change in cogni i e s a us o indi iduals, and c) allow esea che s o classi y indi iduals consis en ly in compa ison wi h o he cogni i e ou comes (i.e. clinical assessmen a he second ollow-up). The limi a ions o he p esen s udy include he ac ha only one g oup o pa ien s wi h MCI ha wo sened o e ime was conside ed. By no ha ing a la ge numbe o pa icipan s in whom de e io a ion ended o wo sen, i was no possible o di e en ia e people who p og ess o mul iple-domain MCI om hose who p og ess o demen ia, and bo h we e included wi hin he same g oup. This hinde s in e p e a ion o he esul s, as al hough he pa icipan s p og ess in he same di ec ion o he con inuum o cogni i e de e io a ion, hey show impo an di e ences ega ding he deg ee o cogni i e impai men and unc ional capaci y. Di e ences be ween bo h ypes o pa icipan s in hei CANTAB longi udinal p o iles should be conside ed in u u e s udies. On he o he hand, he in e al o hi y-six mon hs be ween baseline and he inal e alua ion may no be long enough o ull assessmen o he p og ess. We hope in he u u e o be able o collec longi udinal da a o e a longe pe iod o ime, as he cu en longi udinal esea ch is s ill ongoing. We expec o conduc a hi d ollow-up e alua ion o assess changes ha ha e occu ed in a pe iod o app oxima ely 54 mon hs (4.5 yea s) a e baseline. Financial suppo This esea ch was suppo ed h ough FEDER ounds by he Spanish Di ec o a e Gene al o Scien i ic and Technical Resea ch (P ojec Re . PSI2014-55316-C3-1-R), he Na ional Resea ch Agency (Spanish Minis y o Science, Inno a ion and Uni e si ies) (P ojec Re . Co esponding au ho : Ma ía Campos-Magdaleno, Depa men o De elopmen al Psychology, Rúa Xosé Ma ía Suá ez Núñez, s/n. Campus Su . 15782 San iago de Compos ela, (Galicia, Spain). Email: [email p o ec ed] 21 PSI2017-89389-C2-1-R) and by he Galician Go e nmen (Conselle ía de Cul u a, Educación e O denación Uni e si a ia; axudas pa a a consolidación e es u u ación de unidades de in es igación compe i i as do Sis ema Uni e si a io de Galicia; GI-1807-USC: Re . ED431- 2017/27). Con lic s o in e es . None. E hical s anda ds The au ho s asse ha all p ocedu es con ibu ing o his wo k comply wi h he e hical s anda ds o he ele an na ional and ins i u ional commi ees on human expe imen a ion and wi h he Helsinki Decla a ion o 1975, as e ised in Seoul 2008. Re e ences Albe , M. S., DeKosky, S. T., Dickson, D., Dubois, B., Feldman, H. H., Fox, N. C., … Phelps, C. H. (2011). The diagnosis o mild cogni i e impai men due o Alzheime ’s disease: ecommenda ions om he Na ional Ins i u e on Aging-Alzheime ’s Associa ion wo k g oups on diagnos ic guidelines o Alzheime ’s disease. Alzheime ’s and Demen ia: The Jou nal o he Alzheime ’s Associa ion, 7(3), 270-279. Doi: 10.1016/j.jalz.2011.03.008 Alescio-Lau ie , B., Michel, B. F., He e a, C., Elahmadi, A., Chambon, C., Touze , C., & Paban, V. (2007). Visual and isuospa ial sho - e m memo y in mild cogni i e impai men and Alzheime disease: ole o a en ion. Neu opsychologia, 45(8), 1948- 1960. Doi: 10.1016/j.neu opsychologia.2006.04.033 Co esponding au ho : Ma ía Campos-Magdaleno, Depa men o De elopmen al Psychology, Rúa Xosé Ma ía Suá ez Núñez, s/n. Campus Su . 15782 San iago de Compos ela, (Galicia, Spain). Email: [email p o ec ed] 22 Alladi, S., A nold, R., Mi chell, J., Nes o , P. J., & Hodges, J. R. (2006). Mild cogni i e impai men : applicabili y o esea ch c i e ia in a memo y clinic and cha ac e iza ion o cogni i e p o ile. Psychological Medicine, 36(4), 507-515. Doi: 10.1017/S0033291705006744 Ame ican Psychia ic Associa ion. (2013). Diagnos ic and S a is ical Manual o Men al Diso de s: DSM-5. Ame ican Psychia ic Associa ion: Washing on, DC. Ba beau, E. J., Ranje a, J. P., Didic, M., Con o -Gouny, S., Felician, O., Soulie , E., … Ponce , M. (2008). P o ile o memo y impai men and g ay ma e loss in amnes ic mild cogni i e impai men . Neu opsychologia, 46(4), 1009-1019. Doi: 10.1016/j.neu opsychologia.2007.11.019 Ba es, D., Maechle , M., Bolke , B., & Walke , S. (2015). Fi ing linea mixed-e ec s models usign lme4. Jou nal o S a is ical So wa e, 67(1), 1-48. Doi:10.18637/jss. 067.i01 Belle ille, S., Fouque , C., Hudon, C., Zomahoun, H. T. V., C o eau, J., & Conso ium o he Ea ly Iden i ica ion o Alzheime ’s disease-Que ec. (2017). Neu opsychological measu es ha p edic p og ession om mild cogni i e impai men o Alzheime 's ype demen ia in olde adul s: a sys ema ic e iew and me a-analysis. Neu opsychology Re iew, 27(4), 328-353. Doi: 10.1007/s11065-017-9361-5 Belle ille, S., Syl ain-Roy, S., de Boysson, C., & Ména d, M. C. (2008). Cha ac e izing he memo y changes in pe sons wi h mild cogni i e impai men . P og ess in B ain Resea ch, 169, 365-375. Doi: 10.1016/S0079-6123(07)00023-4 Benede , M. J., & Alejand e, M. A. (1998). TAVEC: Tes de Ap endizaje Ve bal de España- Complu ense. Mad id: TEA ediciones. Co esponding au ho : Ma ía Campos-Magdaleno, Depa men o De elopmen al Psychology, Rúa Xosé Ma ía Suá ez Núñez, s/n. Campus Su . 15782 San iago de Compos ela, (Galicia, Spain). Email: [email p o ec ed] 23 Benede , M. J., & Seisdedos, N. (1996). E aluación Clínica de las Quejas de Memo ia en la Vida Co idiana. Mad id: Edi o ial Médica Paname icana. Blackwell, A. D., Sahakian, B. J., Vesey, R., Semple, J. M., Robbins, T. W., & Hodges, J. R. (2004). De ec ing demen ia: no el neu opsychological ma ke s o p eclinical Alzheime ’s disease. Demen ia and Ge ia ic Cogni i e Diso de s, 17(1-2), 42-48. Doi: 10.1159/000074081 B amba i, S. M., Belle ille, S., Ke goa , M. J., Chaye , C., Gau hie , S., Joube , S. (2009). Single- and mul iple-domain amnes ic mild cogni i e impai men : wo sides o he same coin? Demen ia and Ge ia ic Cogni i e Diso de s, 28(6), 541–549. Doi: 10.1159/000255240 Cacciamani, F., Sal ado i, N., Eusebi, P., Lise i, V., Luche i, E., Calab esi, P., & Pa ne i, L. (2018). E idence o p ac ice e ec in CANTAB spa ial wo king memo y es in a coho o pa ien s wi h mild cogni i e impai men . Applied Neu opsychology: Adul , 25(3), 237-248. Doi:10.1080/23279095.2017.1286346 Campos-Magdaleno, M., Díaz-Bó eda, R., Juncos-Rabadán, O., Facal, D., & Pe ei o, A. X. (2016). Lea ning and se ial e ec s on e bal memo y in mild cogni i e impai men . Applied Neu opsychology: Adul , 23(4), 237-250. Doi: 10.1080/23279095.2015.1053887 Chambe lain, S. R., Blackwell, A. D., Na han, P. J., Hammond, G., Robbins, T. W., Hodges, J. R., … Sahakian, B. J. (2011). Di e en ial cogni i e de e io a ion in demen ia: a wo yea longi udinal s udy. Jou nal o Alzheime ’s Disease, 24(1), 125-136. Doi: 10.3233/JAD-2010-100450. Co esponding au ho : Ma ía Campos-Magdaleno, Depa men o De elopmen al Psychology, Rúa Xosé Ma ía Suá ez Núñez, s/n. Campus Su . 15782 San iago de Compos ela, (Galicia, Spain). Email: [email p o ec ed] 24 De Anna, F., Felician, O., Ba beau, E., Mancini, J., Didic, M., & Ceccaldi, M. (2014). Cogni i e changes in mild cogni i e impai men pa ien s wi h impai ed isual ecogni ion memo y. Neu opsychology, 28(1), 98–105. Doi: 10.1037/neu0000032 De Jage , C. A., Milwain, E., & Budge, M. (2002). Ea ly de ec ion o isola ed memo y de ici s in he elde ly: he need o mo e sensi i e neu opsychological es s. Psychological Medicine, 32(3), 483-491. Doi: 10.1017/S003329170200524X De Ro e , M., Pi on i, V. A., McCabe, J. A., Acos a-Ca bone o, J., A ana, F. S., Mo ein-Zami , S., e al. (2011). Hippocampal dys unc ion in pa ien s wi h mild cogni i e impai men : a unc ional neu oimaging s udy o a isuospa ial pai ed associa es lea ning ask. Neu opsychologia, 49(7), 2060-2070. Doi: 10.1016/j.neu opsychologia.2011.03.037. Delis, D. C., K ame , J. H., Kaplan, E., & Obe , B. (1987). Cali o nia Ve bal Lea ning Tes . San An onio, TX: Psychological Co po a ion. De ancesco, M., Ma ks eine , J., Deisenhamme , E., Kemmle , G., Dju dje ic, T., & Schocke, M. (2013). Impac o whi e ma e lesions and cogni i e de ici s on con e sion om mild cogni i e impai men o Alzheime 's disease. Jou nal o Alzheime 's Disease, 34(3), 665-672. Doi: 10.3233/JAD-122095 Didic, M., Felician, O., Ba beau, E. J., Mancini, J., La ge -Flo ence, C., T amoni, E., & Ceccaldi, M. (2013). Impai ed isual ecogni ion memo y p edic s Alzheime 's disease in amnes ic mild cogni i e impai men . Demen ia and Ge ia ic Cogni i e Diso de s, 35(5-6), 291-299. Doi: 10.1159/000347203 Dubois, B., Feldman, H. H., Jaco a, C., Dekosky, S. T., Ba be ge -Ga eau, P., Cummings, J., … Schel ens, P. (2007). Resea ch c i e ia o he diagnosis o Alzheime ’s disease: e ising he NINCDS-ADRDA c i e ia. Lance Neu ology, 6(8), 734-746. Doi: 10.1016/S1474-4422(07)70178-3 Co esponding au ho : Ma ía Campos-Magdaleno, Depa men o De elopmen al Psychology, Rúa Xosé Ma ía Suá ez Núñez, s/n. Campus Su . 15782 San iago de Compos ela, (Galicia, Spain). Email: [email p o ec ed] 25 Economou, A., Papageo giou, S., & Ka ageo giou, C. (2006). Wo king-delayed memo y di e ence de ec s mild cogni i e impai men wi hou being a ec ed by age and educa ion. Jou nal o Clinical and Expe imen al Neu opsychology, 28(4), 528-535. Doi: 10.1080/13803390590949340 Facal, D., Guà dia-Olmos, J., & Juncos-Rabadán, O. (2015). Diagnos ic ansi ions in mild cogni i e impai men by use o simple Ma ko models. In e na ional Jou nal o Ge ia ic Psychia y, 30(7), 669-676. doi: 10.1002/gps.4197 Fowle , K, S., Saling, M. M., Conway, E. L., Semple, J. M., & Louis, W. (2002). Pai ed associa e pe o mance in he ea ly de ec ion o DAT. Jou nal o he In e na ional Neu opsychological Socie y, 8(1), 58-71. Doi: 10.1017/S1355617702811067 Fox, J. (2016). Applied Reg ession Analysis and Gene alized Linea Models (3 d Edi ion). Thousand Oaks, CA: SAGE. Gagnon, L. G., & Belle ille, S. (2011). Wo king memo y in mild cogni i e impai men and Alzheime ’s Disease: con ibu ion o o ge ing and p edic i e alue o complez span asks. Neu opsychology, 25(2), 226-236. Doi: 10.1037/a0020919. G i i h, R. H., Ne son, K. L., Ha ell, L. E., Zam ini, E. Y., B ocking on, J. C., Ma son, D. C. (2006). Amnes ic mild cogni i e impai men : diagnos ic ou comes and clinical p edic ion o e a wo-yea ime pe iod. Jou nal o he In e na ional Neu opsychological Socie y, 12(2), 166-175. Doi: 10.1017/S1355617706060267 Gua ch, J., Ma cos, T., Salame o, M., Gas ó, C., & Blesa, R. (2008). Mild cogni i e impai men : a isk indica o o la e demen ia, o a p eclinical phase o he disease? In e na ional Jou nal o Ge ia ic Psychia y, 23(3), 257-265. Doi: 10.1002/gps.1871 Co esponding au ho : Ma ía Campos-Magdaleno, Depa men o De elopmen al Psychology, Rúa Xosé Ma ía Suá ez Núñez, s/n. Campus Su . 15782 San iago de Compos ela, (Galicia, Spain). Email: [email p o ec ed] 32 Table 1. Mean and s anda d de ia ions (in pa en heses) o he demog aphic and neu opsychological measu es a baseline o he h ee g oups: subjec i e cogni i e complain s (SCC) ha emain s able (SCC-s able); mild cogni i e impai men ha emain s able (MCI- s able); mild cogni i e impai men ha wo sened (MCI-wo sened). No e: MMSE= MiniMen al S a e Examina ion CCI=Cha lson Como bidi y Index. SCC: Subjec i e Cogni i e Complain s (pa ien ). CAMCOG= Camb idge Cogni i e Examina ion ( o al sco e). CVLT SDFR= Cali o nia Ve bal Lea ning Tes , Sho Delay F ee Recall. CVLT LDFR= Cali o nia Ve bal Lea ning Tes , Long Delay F ee Recall. Visual Acui y = Ligh house es . *= p < .05 **= p < .01 SCC-s able G oup 1 N=149 MCI-s able G oup 2 N=32 MCI-wo sened G oup 3 N=27 K uskal Wallis χ²(gl) G oup compa ison Age 64.26 (8.83) Range:50-87 70.94 (7.54) Range: 54-83 75.44 (7.14) Range: 61-87 39.46 (2)** G3 > G2 > G1 Gende Women: 70.3% Men: 29.7% Women: 68.8% Men: 31.3% Women: 55.6% Men: 44.4% Yea s o Educa ion 10.28 (4.71) Range: 2-22 9.15 (3.40) Range: 2-17 9.30 (4.79) Range: 4-25 1.09 (2) SCC 18.84 (4.54) Range: 7-31 20.25 (4.09) Range: 10-32 18.07 (4.64) Range: 13-33 6.83 (2)* G2 > G1, G3 Law on- B ody 7.55 (.95) Range: 4-8 6.8 (1.55) Range: 3-8 6.15 (2.08) Range: 2-8 17.29 (2)** G1 > G2, G3 CCI .76 (.84) Range: 0-3 1.09 (1.02) Range: 0-4 .70 (.86) Range:0-3 3.65 (2) MMSE 28.34 (1.34) 25.13 (2.89) 24.04 (2.53) 73.39 (2)** G1 > G2, G3 CAMCOG 89.88 (6.96) 77.40 (8.99) 70.92 (10.08) 81.74 (2)** G1 > G2 > G3 CVLT- SDFR 11.01 (2.50) 4.50 (3.00) 2.37 (2.04) 113.20 (2)** G1 > G2 > G3 CVLT- LDFR 11.85 (2.57) 5.53 (3.77) 2.48 (2.43) 105.13 (2)** G1 > G2 > G3 Visual Acui y .55 (.17) Range: .20-1.00 .52 (.16) Range: .20-.80 .53 (.18) Range: .33-.80 Co esponding au ho : Ma ía Campos-Magdaleno, Depa men o De elopmen al Psychology, Rúa Xosé Ma ía Suá ez Núñez, s/n. Campus Su . 15782 San iago de Compos ela, (Galicia, Spain). Email: [email p o ec ed] 33 Table 2. Summa y o models compa ed o PAL o al e o s adjus ed-6 shapes. All models include andom e ec s o in e cep s and age a baseline as co a ia e. Model 1 is he null mixed model (i.e. andom in e cep s and age co a ia e only); Model 2 is he mixed model wi h main e ec s; Model 3 is he mixed model wi h main e ec s and in e ac ions. Coe icien s and s anda d e o s (in pa en heses) a e shown on a log scale o numbe o e o s (i.e. na u al log o he esponse). No e: ***p< 0.01 Dependen a iable: PAL o al e o s adjus ed-6 shapes Model 1 Model 2 Model 3 Age a baseline 0.559*** (0.054) 0.379*** (0.051) 0.377*** (0.051) E alua ion Time 0.036*** (0.009) -0.032*** (0.012) MCI-wo sened 1.065*** (0.153) 0.888*** (0.156) MCI-s able 0.904*** (0.131) 0.780*** (0.134) E alua ion Time x MCI-wo sened 0.280*** (0.030) E alua ion Time x MCI-s able 0.137*** (0.021) In e cep 3.403*** (0.054) 3.105*** (0.056) 3.170*** (0.057) Obse a ions 624 624 624 Log Likelihood -3,005.318 -2,965.740 -2,911.160 Akaike In . C i . 6,016.637 5,943.481 5,838.320 Bayesian In . C i . 6,029.561 5,969.329 5,872.785 Co esponding au ho : Ma ía Campos-Magdaleno, Depa men o De elopmen al Psychology, Rúa Xosé Ma ía Suá ez Núñez, s/n. Campus Su . 15782 San iago de Compos ela, (Galicia, Spain). Email: [email p o ec ed] 34 Table 3. Summa y o model compa ison o PRM o al pe cen co ec . All models include andom e ec s o in e cep s and age a baseline as co a ia e. Model 1 is he null mixed model (i.e. andom in e cep s and age co a ia e only); Model 2 is he mixed model wi h main e ec s; Model 3 is he mixed model wi h main e ec s and in e ac ions. Coe icien s and s anda d e o s (in pa en heses) No e: ***p< 0.01 Dependen a iable: PRM pe cen co ec o al Model 1 Model 2 Model 3 Age a baseline -5.430*** (0.656) -3.439*** (0.599) -3.460*** (0.599) E alua ion Time 0.213 (0.379) 0.405 (0.405) MCI-wo sened -16.591*** (2.362) -16.673*** (2.823) MCI-s able -12.344*** (1.645) -10.625*** (1.967) E alua ion Time x MCI-wo sened 0.286 (2.471) E alua ion Time x MCI-s able -1.974 (1.249) In e cep 83.072*** (0.649) 85.735*** (0.731) 85.546*** (0.744) Obse a ions 560 560 560 Log Likelihood -2,055.846 -2,014.233 -2,010.012 Akaike In . C i . 4,123.691 4,046.466 4,042.024 Bayesian In . C i . 4,149.637 4,085.337 4,089.493 Co esponding au ho : Ma ía Campos-Magdaleno, Depa men o De elopmen al Psychology, Rúa Xosé Ma ía Suá ez Núñez, s/n. Campus Su . 15782 San iago de Compos ela, (Galicia, Spain). Email: [email p o ec ed] 35 Table 4. Summa y o compa ed models o DMS o al pe cen co ec . All models include andom e ec s o in e cep s and age a baseline as co a ia e. Model 1 is he null mixed model (i.e. andom in e cep s and age co a ia e only); Model 2 is he mixed model wi h main e ec s; Model 3 is he mixed model wi h main e ec s and in e ac ions. Coe icien s and s anda d e o s (in pa en heses) No e: ***p< 0.01 Dependen a iable: DMS o al pe cen co ec Model 1 Model 2 Model 3 Age a baseline -6.170*** (0.565) -4.555*** (0.554) -4.456*** (0.546) E alua ion Time 0.212 (0.387) 0.563 (0.423) MCI-wo sened -12.014*** (1.831) -7.677*** (2.038) MCI-s able -5.990*** (1.470) -6.597*** (1.738) E alua ion Time x MCI-wo sened -7.506*** (1.610) E alua ion Time x MCI-s able -0.684 (1.096) In e cep 78.252*** (0.565) 80.170*** (0.707) 79.847*** (0.720) Obse a ions 555 555 555 Log Likelihood -2,001.541 -1,975.448 -1,961.768 Akaike In . C i . 4,011.083 3,964.896 3,941.536 Bayesian In . C i . 4,028.344 3,995.065 3,980.292 Co esponding au ho : Ma ía Campos-Magdaleno, Depa men o De elopmen al Psychology, Rúa Xosé Ma ía Suá ez Núñez, s/n. Campus Su . 15782 San iago de Compos ela, (Galicia, Spain). Email: [email p o ec ed] 36 Table 5. Summa y o models compa ed o SSP span leng h. All models include andom e ec s o in e cep s and age a baseline as co a ia e. Model 1 is he null mixed model ( andom in e cep s and age co a ia e only); Model 2 is he mixed model wi h main e ec s; Model 3 is he mixed model wi h main e ec s and in e ac ions. Coe icien s and s anda d e o s (in pa en heses) a e shown on he log scale o numbe o co ec esponses (i.e. na u al log o he esponse). No e: **p<0.05; ***p< 0.01 Dependen a iable: SSP span leng h Model 1 Model 2 Model 3 Age a baseline -0.078*** (0.019) -0.053** (0.021) -0.054*** (0.021) E alua ion Time 0.006 (0.024) 0.025 (0.026) MCI-wo sened -0.248*** (0.085) -0.254** (0.113) MCI-s able -0.107 (0.058) -0.044 (0.085) E alua ion Time x MCI-wo sened 0.025 (0.113) E alua ion Time x MCI-s able -0.071 (0.072) In e cep 1.581*** (0.019) 1.610*** (0.032) 1.601*** (0.034) Obse a ions 624 624 624 Log Likelihood -1,013.299 -1,007.706 -1,007.200 Akaike In . C i . 2,032.597 2,027.413 2,030.400 Bayesian In . C i . 2,045.570 2,053.359 2,064.994 Co esponding au ho : Ma ía Campos-Magdaleno, Depa men o De elopmen al Psychology, Rúa Xosé Ma ía Suá ez Núñez, s/n. Campus Su . 15782 San iago de Compos ela, (Galicia, Spain). Email: [email p o ec ed] 37 Figu e 1. Es ima ed ma ginal means and e o s ba s om Model 1 o PAL, PRM, DMS and SSP in he h ee g oups ac oss he h ee e alua ion imes. No e: SE= S anda d E o ; BL = Baseline assessmen ; T1= Time 1 assessmen ; T2= Time 2 assessmen .