RESEARCH Open Access
A he apy is associa ed wi h sus ained
imp o emen in cogni i e unc ion in he
elde ly wi h mild neu ocogni i e diso de :
indings om a pilo andomized con olled
ial o a he apy and music eminiscence
ac i i y e sus usual ca e
Ra hi Mahend an
1,2,3*
, Mihi Gandhi
4,5,6
, Rajesh Babu Moo akonda
4
, Jona han Wong
1
, Madhu Ma hi Kanchi
8
,
Johnson Fam
1,2
, I is Raw ae
7
, Alan P em Kuma
8,9
, Lei Feng
1,2
and Ee Heok Kua
1,2
Abs ac
Backg ound: Mild cogni i e impai men (MCI) is a phase in cogni i e decline when i is s ill possible o in e ene o
e e se he decline. Cogni i e s imula ion deli e ed h ough psychosocial in e en ions p o ides bo h psychological
in e en ion and social s imula ion o imp o e cogni ion. A pilo open-label pa allel-a ms andomized con olled ial
was unde aken o examine he e ec s o a he apy (AT) and music eminiscence ac i i y (MRA) compa ed o he
con ol, on he p ima y ou come o neu ocogni i e domain assessmen s in elde ly people wi h MCI.
Me hods: Communi y-li ing elde ly people wi h MCI (Pe e sen’s c i e ia), assessed o s udy eligibili y, we e
andomized using a web-based sys em wi h equal alloca ion o wo in e en ion a ms: AT (guided iewing o a
pieces and p oduc ion o isual a s) and MRA (lis ening, and ecalling memo ies ela ed o music) and a con ol a m
(s anda d ca e wi hou any in e en ion). In e en ions we e led by ained he apis s weekly o 3 mon hs, hen
o nigh ly o 6 mon hs. Neu ocogni i e domains (mean o memo y, a en ion, and isuo-spa ial abili ies s anda dized
sco es), psychological wellbeing (subsynd omal dep ession and anxie y) and elome e leng h as a biological ma ke o
cellula ageing, we e assessed by in e en ion-blinded assesso s a baseline, 3 mon hs and 9 mon hs.
Resul s: In o al, 250 people we e sc eened and 68 we e andomized and included in he analysis. In he AT a m,
neu ocogni i e domains imp o ed compa ed o he con ol a m a 3 mon hs (mean di e ence (d) = 0.40; 90% CI 0.126,
0.679) and we e sus ained a 9 mon hs (d = 0.31; 90% CI 0.068, 0.548). The e was some imp o emen in dep ession and
anxie y a 3 and 9 mon hs and in elome e leng h a 9 mon hs, bu his was no signi ican . Simila imp o emen s
we e obse ed in he MRA a m o e he con ol a m, bu hey we e no signi ican . The e we e no in e en ion- ela ed
ad e se e ec s.
(Con inued on nex page)
* Co espondence: [email p o ec ed]
1
Depa men o Psychological Medicine, Yong Loo Lin School o Medicine,
Na ional Uni e si y o Singapo e, NUHS Towe Block, Le el 9, 1E Ken Ridge
Road, Singapo e 119228, Singapo e
2
Depa men o Psychological Medicine, Na ional Uni e si y Hospi al, NUHS
Towe Block, Le el 9, 1E Ken Ridge Road, Singapo e, Singapo e
Full lis o au ho in o ma ion is a ailable a he end o he a icle
© The Au ho (s). 2018 Open Access This a icle is dis ibu ed unde he e ms o he C ea i e Commons A ibu ion 4.0
In e na ional License (h p://c ea i ecommons.o g/licenses/by/4.0/), which pe mi s un es ic ed use, dis ibu ion, and
ep oduc ion in any medium, p o ided you gi e app op ia e c edi o he o iginal au ho (s) and he sou ce, p o ide a link o
he C ea i e Commons license, and indica e i changes we e made. The C ea i e Commons Public Domain Dedica ion wai e
(h p://c ea i ecommons.o g/publicdomain/ze o/1.0/) applies o he da a made a ailable in his a icle, unless o he wise s a ed.
Mahend an e al. T ials (2018) 19:615
h ps://doi.o g/10.1186/s13063-018-2988-6
(Con inued om p e ious page)
Conclusions: A he apy deli e ed by ained s a as “a as he apy”and “a psycho he apy”may ha e been he
signi ican con ibu o o cogni i e imp o emen s. The indings suppo cogni i e s imula ion o elde ly people wi h
cogni i e decline and signal he need o la ge s udies and u he in es iga ion o ca e ully designed psycho-social
in e en ions o his g oup.
T ial egis a ion: Clinical T ials.go , NCT02854085. Regis e ed on 7 July 2016.
Keywo ds: A he apy, Music eminiscence ac i i y, Elde ly, Mild cogni i e impai men
Backg ound
Psycho-social in e en ions a e inc easingly in es iga ed
as p e en i e s a egies o elde ly people wi h cogni i e
decline. To da e, he e is li le e idence o bene i when
demen ia (o majo neu ocogni i e diso de , Diagnos ic
and s a is ical manual o men al diso de s (DSM)-5) has
se in when e en cogni i e aining and cogni i e s imu-
la ion in e en ions canno signi ican ly imp o e gene al
cogni ion [1,2]. Al hough a en ion has shi ed o
p e-demen ia o mild cogni i e impai men (MCI) (mild
neu ocogni i e diso de , DSM-5), he e is again no e i-
dence ha any in e en ion is e ec i e a his s age and
u he mo e, pha maco he apy is no ecommended o
MCI [3,4]. The elde ly wi h MCI a e un o una ely a
pa icula ly ulne able “a - isk”g oup. Almos hal will
de e io a e o demen ia and a a highe a e han hose
who a e cogni i ely no mal [5,6]. Up o 40% who may
emain a he MCI s age, will con inue o expe ience
cogni i e di icul ies and psychological sequelae [7]. In-
e en ions ha imp o e domains o memo y and ex-
ecu i e unc ion would he e o e be pa icula ly
bene icial i hey could con ibu e o a slowing o he
p og ession o e en a e e sal o cogni i e impai men .
In an ea lie na u alis ic s udy, Raw ae e al.[8] ound
imp o emen s in subsynd omal anxie y and dep ession
amongs communi y-li ing elde ly who pa icipa ed in
psychosocial in e en ions, pa icula ly mind ul awa e-
ness p ac ice, a he apy (AT) and music eminiscence
ac i i y (MRA). Howe e , when he li e a u e on hese
psychosocial in e en ions was explo ed o cogni i e e -
ec s in elde ly people wi h MCI, i was e iden ha eld-
e ly popula ions wi h MCI ha e no been s udied, many
o he s udies on AT and MRA in e en ions we e no
andomized, and in e en ions we e mainly ac i i ies like
a and c a wo k o lis ening o music, a he han he -
apy. Thus, o examine he e ec s o bo h hese in e en-
ions on cogni i e unc ions in elde ly people wi h MCI,
he in e en ions we e ca e ully e iewed, designed and
s uc u ed as he apy, and a pilo andomized con olled
ial (RCT) was unde aken. The p ima y indings o he
e ec s on cogni ion, subsynd omal mood s a es and
elome e leng hs a e p esen ed. The p ima y objec i e
was he compa ison o change in neu opsychological es
sco es a 3 mon hs be ween he in e en ion and con ol
g oups; he seconda y objec i es we e compa isons a
9 mon hs.
Me hods
Pa icipan s and s udy design
An open-label, pa allel RCT wi h h ee a ms ( wo in e -
en ions and one con ol) ec ui ed communi y-li ing
elde ly people who me p e-de ined inclusion c i e ia,
(age 60–85 yea s, bo h gende s, communi y-li ing, ul ill
Pe e sen’s c i e ia o MCI) [9]. The s udy had e hics ap-
p o al om he Na ional Uni e si y o Singapo e Ins i u-
ional Re iew Boa d and w i en in o med consen was
aken om pa icipan s. Subjec s who had been in a
la ge coho s udy a he Uni e si y Depa men ’s e-
sea ch si e in he communi y, (known as he T aining
and Resea ch Academy, TaRA) and who had gi en con-
sen o be app oached o in e en ion s udies we e con-
ac ed and p o ided in o ma ion abou he s udy.
Rec ui men , sc eening and in o med consen and sub-
sequen in e en ions and assessmen s we e done a
TaRA. In e en ions we e adminis e ed weekly in he
i s 3 mon hs, hen o nigh ly o 6 mon hs. Pa ici-
pan s we e andomly alloca ed in o h ee a ms (1:1:1 al-
loca ion) using a web-based andomiza ion sys em,
p o ided by he Singapo e Clinical Resea ch Ins i u e,
and in e en ion assignmen was balanced using he pe -
mu ed block andomiza ion s a i ied by gende . A de-
ailed s udy p o ocol was p e iously published [10].
In e en ions
AT in ol ed wo componen s. A pieces we e selec ed
by cu a o s om he Na ional Galle y and he Na ional
Uni e si y o Singapo e Museum and he ac i i y was
de eloped in consul a ion wi h he s udy eam (psychia-
is s and he apis s), wi h emphasis on ele ance o he
elde ly, in e ms o hemes and e en s om he coun y’s
pas . Guided iewing and cogni i e e alua ion o a
wo ks a he espec i e si es was conduc ed as a g oup
ac i i y by ained s a and in ol ed na a ion o
hough s and inne expe iences. A second componen
in ol ed isual a p oduc ion. The physical c ea ion o
hemed a wo k was ollowed by image app ecia ion ac-
i i ies o gain insigh and discuss eelings, and ook
place a he esea ch cen e . MRA in ol ed lis ening,
Mahend an e al. T ials (2018) 19:615 Page 2 o 10
and ecalling memo ies and expe iences ela ed o he
music. The he apis p epa ed songs and used pho o-
g aphs o ideo clips o accompany he music, o discus-
sion pu poses. The MRA p omo ed sha ed eelings while
he g oup p ocess p o ided alida ion. The de ailed con-
en s o he s uc u ed in e en ions a e a ailable in he
s udy p o ocol [10]. Fu he de ails on he a wo ks used
and sample a wo ks c ea ed by he pa icipan s will be
a ailable on eques . The con ol g oup (CG) did no e-
cei e any in e en ion bu con inued li e as usual.
Ou come measu es
Neu ocogni i e domain assessmen s a baseline and a 3
and 9 mon hs epo ed he e include he Rey audi o y
e bal lea ning es (RAVLT) Lis Lea ning, Delayed Re-
call, Recogni ion T ial (Memo y), Wechsle Adul
In elligence Scale-3 d edi ion (WAIS-III) Block design
(Visuospa ial abili ies), Digi Span Fo wa d (A en ion
and Wo king Memo y), and Colo T ails Tes 2 (Execu-
i e unc ion) [11–13]. Indi idual domain sco es we e
s anda dized (z-sco es) o gene al popula ion no ms ad-
jus ed o age and educa ion le el; highe sco es indica e
be e pe o mance [14].
Psychological wellbeing was assessed a baseline and a
3 and 9 mon hs wi h he (1) Ge ia ic Dep ession Scale
(GDS), which is a 15-i em “yes/no”ques ionnai e wi h
highe o al sco es associa ed wi h highe isk o dep es-
sion [15] and (2) Ge ia ic Anxie y In en o y (GAI),
which is a 20-i em “ag ee/disag ee”ques ionnai e meas-
u ing dimensional anxie y, wi h highe o al sco es asso-
cia ed wi h anxie y symp oms [16].
Sleep quali y was assessed using a 100-poin isual
analog scale (VAS) (0, wo s sleep o 100, bes sleep).
Fo elome e leng h measu emen s, genomic DNA was
ex ac ed om whole blood samples using QIA amp DNA
blood mini ki (ca alog numbe 51104) acco ding o he
manu ac u e ’s p o ocol and s o ed a −80 °C. Telome e
leng h was measu ed by using a non- adioac i e chemilu-
minescen elome e leng h assay ki (Telo TAGGG assay
ki , Sigma Ald ich; ca alog numbe 12209136001) o
isualize he elome ic DNA epea sequence TTAAGGG
om blood samples. Telome e leng h measu emen in-
ol es he diges ion o 1 μgo DNAusingHin I/RsaIen-
zymesa 37°C o 2h,and unon0.8%aga osegel.DNA
smea s we e ans e ed on o he nylon memb ane (Ame -
sham Hybond ™-XL) o e nigh . T ans e ed DNA ag-
men s a e hyb idized o a digoxigenin (DIG)-labeled p obe
o alida e he elome ic epea s de eloped by CDP-S a
which is a digoxigenin subs a e o cap u e he imaging on
x- ay ilm.Telome eleng hwas measu ed by he loca ion
o bands based on molecula weigh s anda d. A e age elo-
me e leng h is measu ed be ween 100 base pai s o 20 k
base pai s. TeloTool, which is MATLAB so wa e, was used
o measu e he elome e leng hs o he samples in his
s udy. Image p ocessing and de ec ion o DNA smea s we e
e alua ed as indica ed in he e e ences [17,18].
S a is ical analysis
The sample size o he s udy was calcula ed o es ima e
he in e en ion e ec (i.e., di e ence be ween he in e -
en ion and con ol g oup o mean change in neu o-
psychological es sco e a 3 mon hs om baseline) wi h
p e-speci ied p ecision. A sample size o 22 pa icipan s
in each g oup will p o ide an es ima e wi h p ecision
+/−0.5 s anda d de ia ion (SD) wi h 90% ce ain y (i.e.,
wid h o 90% con idence in e al (uppe limi –lowe
limi ) equal o 1 SD). The p ecision le el was selec ed
conside ing an in e en ion will be wo h in es iga ing
u he in a con i ma o y ial i a he leas a medium
s anda dized e ec size (0.5 SD) is shown in compa ison
o he con ol g oup.
Mean change om baseline in neu ocogni i e z-sco es
we e es ima ed and compa ed be ween he in e en ions
and he con ol g oups using he linea mixed model
wi h pa icipan -speci ic andom e ec s and indica o
a iables o he wo in e en ions ( e e ence, con ol
g oup), indica o a iables o ime (3-mon h and
9-mon h assessmen s wi h baseline as he e e ence), in-
e ac ions be ween indica o a iables o he in e en-
ions and ime, baseline alue, and gende as ixed
e ec s, along wi h 90% and 95% CIs. No mul iplici y
co ec ions we e applied due o he explo a o y na u e
o he s udy. The same model was used o compa e he
Ge ia ic Dep ession Scale (GDS) o al sco e, Ge ia ic
Anxie y Index (GAI) o al sco e, sleep quali y VAS, and
elome e leng h be ween he in e en ions and he con-
ol g oups. All he analyses we e pe o med in he
in en ion- o- ea popula ion. SAS so wa e e sion 9.4
(SAS Ins i u e, Ca y, NC, USA) was used. A p alue
<0.05 was deemed o be s a is ically signi ican : p alues
should be in e p e ed as hypo hesis gene a ing as he
s udy was no powe ed o hypo hesis es ing.
Resul s
Pa icipan cha ac e is ics and s udy conduc
Be ween 13 June 2016 and 17 Augus 2016, 250 po en-
ial pa icipan s we e sc eened o whom 68 we e e-
c ui ed in o he s udy, andomly assigned o AT (n= 22),
MRA (n= 24), o he CG (n= 22), and included in he
in en ion- o- ea popula ion. The e we e no clinically
ele an di e ences among he h ee g oups a baseline
(Table 1), excep he CG had a la ge pe cen age o pa -
icipan s (41%) who we e cu en ly wo king ( ull ime/
pa - ime/sel -employed) compa ed o he AT (18%) and
(8%) MRA g oups. O e all, he mean age o he pa ici-
pan s was 71.1 yea s, 56 pa icipan s we e emale, and
he a e age du a ion o schooling was 5.4 yea s.
Mahend an e al. T ials (2018) 19:615 Page 3 o 10
Du ing he i s 3 mon hs pos - andomiza ion, h ee
pa icipan s om he AT g oup, one om he MRA
g oup, and ou om he CG g oup discon inued he
s udy (Fig. 1). Mean compliance (a endance a in e en-
ion he apy sessions) wi h AT and MRA was 80% and
82%, espec i ely. By he end o he s udy a 9 mon hs,
he AT g oup, MRA g oup, and CG had 18, 22, and 18
pa icipan s, espec i ely; mean compliance wi h he AT
and MRA in e en ions was 75% and 83%, espec i ely.
Each o he in e en ion sessions ( o AT and MRA)
las ed an hou inclusi e o a 5-min mind ul elaxa ion
exe cise a he s a , o se le he subjec s, and a
15-min b eak). The 40 min o ac i e engagemen was
e ec i e and did no i e he elde ly pa icipan s; hey
emained engaged h oughou he session. No subjec
le any o he sessions be o e i ended. T a elling
ime by coach o he Na ional Galle y o NUS
Museum ook abou 15 min. The he apis s, we e ad-
ised on he pa icipan s’educa ional le el and cogni-
i es a eandwe e hesame h oughou hes udy
pe iod and e e y e o was made o ensu e hey
unde s ood he in e en ion and sessions.
Changes in neu opsychological es sco es
The es ima ed mean change a 3 mon hs om baseline was
s a is ically signi ican ly highe in he AT g oup compa ed
o he CG o Lis Lea ning (di e ence (d) = 0.542; 90%
con idence in e al (CI) 0.105, 0.810; p=0.042) and Digi
Span Fo wa d (d = 0.991; 90% CI 0.251, 1.730; p=0.028)
(Fig. 2). Fu he mo e, Bene icial e ec s we e also obse ed
in he AT g oup o Delayed Recall, Recogni ion T ials,
Block Design, and Colo T ails 2; howe e , hese esul s
we e no s a is ically signi ican (each p>0.05). The mean
o memo y domains (Lis Lea ning, Delayed Recall, Recog-
ni ion T ial) and he mean o all domains (Lis Lea ning,
Delayed Recall, Recogni ion T ial, Digi Span Fo wa d,
Colo T ains 2, Block Design) we e also s a is ically signi i-
can ly highe in he AT g oup compa ed o he CG (d
(memo y domains) = 0.403; 90% CI 0.126, 0.679; p=0.017;
d (all domains) = 0.462; 90% CI 0.202, 0.722; p=0.004).
Table 1 Baseline demog aphic and neu ocogni i e cha ac e is ics
Cha ac e is ics A he apy
(n= 22)
Music eminiscence ac i i ies
(n= 24)
Con ol
(n= 22)
Age (yea s), mean (SD) 71.1 (4.8) 71.6 (5.3) 70.6 (5.8)
Female, n(%) 18 (81.8) 20 (83.3) 18 (81.8)
Li ing wi h pa ne , n(%) 14 (63.6) 15 (62.5) 20 (90.9)
Educa ion yea s, mean (SD) 5.2 (4.1) 5.0 (3.3) 6.1 (3.4)
Wo king, n(%) 4 (18.2) 2 (8.3) 9 (40.9)
Body mass index, mean (SD) 24.9 (3.6) 23.5 (3.8) 22.8 (3.9)
Medical condi ion, n(%)
High blood p essu e 14 (63.6) 18 (75.0) 6 (27.3)
High choles e ol 11 (50.0) 15 (62.5) 7 (31.8)
Diabe es melli us 2 (9.1) 8 (33.3) 5 (22.7)
O he s 2 (9.1) 3 (12.5) 3 (13.6)
Neu ocogni i e domain z-sco e, mean (SD)
RAVLT Lis Lea ning Sum 0.04 (1.35) 0.01 (1.26) −0.01 (1.26)
RAVLT Delayed Recall −0.05 (0.98) −0.01 (1.10) 0.05 (0.87)
RAVLT Recogni ion T ial −0.05 (1.28) 0.26 (1.10) 0.14 (0.91)
Mean o memo y domains (RAVLT subse s) −0.02 (0.97) 0.09 (0.95) 0.06 (0.79)
WAIS-III Block Design −0.74 (1.16) −0.93 (1.04) −0.47 (0.85)
WAIS-III Digi Span (Fo wa d) 1.33 (1.59) 1.79 (1.65) 2.15 (1.77)
Colo T ails 2 −0.55 (1.38) −1.15 (1.46) −0.82 (1.91)
Mean o all domains −0.00 (0.42) −0.01 (0.59) 0.17 (0.56)
Numbe o domains wi h z-sco e < −1.5, mean (SD) 1.41 (0.80) 1.71 (0.81) 1.14 (0.35)
GAI o al sco e, mean (SD) 2.0 (3.9) 2.1 (4.0) 2.8 (4.1)
GAD o al sco e, mean (SD) 2.5 (2.9) 1.8 (2.2) 2.8 (2.9)
Sleep quali y VAS, mean (SD) 67.0 (26.1) 68.4 (19.8) 69.8 (16.5)
See “Me hods”sec ion o ou come de ini ions
RAVLT Rey Audi o y Ve bal Lea ning Tes , WAIS-III Wechsle Adul In elligence Scale-3 d edi ion, GAI Ge ia ic Anxie y In en o y, GAD Ge ia ic Dep ession Scale,
SD s anda d de ia ion
Mahend an e al. T ials (2018) 19:615 Page 4 o 10
Mean numbe o domains wi h z-sco e < −1.5 a 3 mon hs
was lowe in he AT g oup compa ed wi h he CG (d = −
0.314; 90% CI −0.629, 0.000; p= 0.100). The e ec o AT
on memo y domains was also sus ained a 9 mon hs (d =
0.308; 90% CI 0.068, 0.548; p= 0.035). The de ailed esul s
o memo y- ela ed and o he neu opsychological ou comes
a e epo ed in Table 2and Addi ional ile 1:TablesS1and
Table S2, espec i ely.
The es ima ed mean changes we e highe in he MRA
g oup compa ed o he CG o Delayed Recall (d = 0.172),
Assessed o eligibili y (n = 250)
Excluded (n = 182)
-Did no mee inclusion c i e ia (n = 179)
-Did no comple e assessmen (n = 3)
Analyzed (n = 22)
Discon inued in e en ion (n = 3)
-No in e es (n = 1)
-Shi ed ou o he a ea (n = 1)
-S a ed employmen (n = 1)
Alloca ed o A The apy
(n = 22)
Los o ollow-up (n = 4)
-Re used u he pa icipa ion a e
alloca ion (n = 2)
- Unable o a end due o spinal
comp ession (n = 1)
- Unable o a end due o wo k
commi men s (n = 1)
Alloca ed o Con ol G oup
(n = 22)
Analyzed (n = 22)
Analysis
Randomized (n = 68)
En ollmen
Alloca ed o Music Reminiscence
Ac i i y (n = 24)
Discon inued in e en ion (n = 1)
-Pa icipan s opped a ending
a e 1 session due o a medical
issue
Follow-Up
Analyzed (n = 24)
Alloca ion
Fig. 1 Pa icipan s disposi ion (Consolida ed S anda ds o Repo ing T ials (CONSORT) diag am)
Fig. 2 Es ima ed mean change in memo y- ela ed neu opsychological ou comes a 3 mon hs (3M) and 9 mon hs (9M) om baseline: *p< 0.05
o compa ison wi h he con ol g oup. RAVLT, Rey Audi o y Ve bal Lea ning Tes . Ou comes a e p esen ed as s anda dized z-sco e (see he
“Me hods”sec ion). Mean alues we e es ima ed using a linea mixed-e ec s model adjus ed o baseline alues and gende
Mahend an e al. T ials (2018) 19:615 Page 5 o 10

Recogni ion T ial (d = 0.088), Digi Span Fo wa d (d =
0.787), Colo T ails 2 (d = 0.033), and Block Design
(d = 0.130), bu hey we e no s a is ically signi ican
(each p>0.05) (Fig. 2). Simila ly, he means o mem-
o y domains and all domains we e highe in he
MRA g oup compa ed o he CG g oup bu we e no
s a is ically signi ican (d (memo y domains) = 0.037; d
(all domains) = 0.180; each p> 0.05). The e ec in he
MRA g oup a 9 mon hs was simila o ha a 3
mon hs (Table 2; Addi ional ile 1:TablesS1andS2).
Changes in subsynd omal dep ession and anxie y
The e was some decline (less han 2 poin s) in he GDS
and GAI o al sco es a 3 mon hs and 9 mon hs om
baseline in bo h he AT and MRA g oups. Howe e ,
hese educ ions we e no s a is ically signi ican ly di e -
en om he ones obse ed in he CG (d < 1 poin in
bo h he GDS and GAI o al sco es a each imepoin ;
p> 0.05). (Table 3and Addi ional ile 1: Table S3).
Change in sleep quali y
The e was an imp o emen o 7 poin s and 3 poin s in
he sleep quali y VAS in he AT g oup a 3 mon hs and
9 mon hs. Howe e , hese imp o emen s we e no
s a is ically signi ican ly be e han he changes
obse ed in he CG (d < 4 poin s a each imepoin ;
p> 0.05). The MRA g oup also had sleep quali y simi-
la o he CG a pos -baseline assessmen s (Table 3
and Addi ional ile 1Table S3).
Changes in elome e leng hs
The e was an inc ease in he elome e leng h in he AT
g oup a 9 mon hs (mean change = 552; p= 0.003). How-
e e , he change was no s a is ically signi ican ly di e en
om he inc ease obse ed in he CG (d = −22; p>0.05).
The MRA g oup also had an inc ease in he eleme e
leng h a 9 mon hs (mean change = 292; p= 0.076), bu
his was no s a is ically signi ican ly di e en om he
CG (d = −281; p> 0.05) (Table 4and Addi ional ile 1:
Table S4). The e we e no in e en ion- ela ed ad e se
e ec s epo ed in he ial.
Discussion
This s udy p o ides p elimina y e idence ha speci ic
psychosocial in e en ions ha a e ca e ully s uc u ed
and egula ly deli e ed by ained s a can e ec i ely
Table 2 Es ima ed mean change in memo y- ela ed neu opsychological ou comes a 3-mon hs and 9-mon hs
A he apy
(n= 22)
Music eminiscence ac i i ies
(n= 24)
Con ol
(n= 22)
Numbe o pa icipan s assessed a 3 mon hs 19 23 18
Numbe o pa icipan s assessed a 9 mon hs 18 22 18
RAVLT Lis Lea ning Sum z-sco e
Mean change a 3 mon hs om baseline (SE) 1.01 (0.19) 0.36 (0.17) 0.47 (0.19)
Di e ence (in e en ion –con ol) (90% CI) [p] 0.54 (0.105, 0.978) [0.042] −0.11 (−0.529, 0.314) [0.674]
Mean change a 9 mon hs om baseline (SE) 1.07 (0.17) 0.73 (0.15) 0.60 (0.16)
Di e ence (in e en ion –con ol) (90% CI) [p] 0.47 (0.089, 0.854) [0.043] 0.13 (−0.240, 0.493) [0.569]
RAVLT Delayed Recall z-sco e
Mean change a 3 mon hs om baseline (SE) 0.31 (0.14) 0.25 (0.13) 0.08 (0.14)
Di e ence (in e en ion –con ol) (90% CI) [p] 0.24 (−0.088, 0.561) [0.230] 0.17 (−0.142, 0.486) [0.366]
Mean change a 9 mon hs om baseline (SE) 0.41 (0.13) 0.41 (0.12) 0.27 (0.12)
Di e ence (in e en ion –con ol) (90% CI) [p] 0.14 (−0.159, 0.430) [0.448] 0.14 (−0.145, 0.421) [0.419]
RAVLT Recogni ion T ial z-sco e
Mean change a 3 mon hs om baseline (SE) 0.57 (0.22) 0.20 (0.20) 0.11 (0.22)
Di e ence (in e en ion –con ol) (90% CI) [p] 0.46 (−0.045, 0.969) [0.134] 0.09 (−0.402, 0.578) [0.767]
Mean change a 9 mon hs om baseline (SE) 0.50 (0.21) 0.32 (0.19) 0.18 (0.20)
Di e ence (in e en ion –con ol) (90% CI) [p] 0.32 (−0.153, 0.802) [0.262] 0.14 (−0.318, 0.600) [0.612]
Mean z-sco e o RAVLT memo y domains
Mean change a 3 mon hs om baseline (SE) 0.64 (0.12) 0.27 (0.11) 0.23 (0.12)
Di e ence (in e en ion –con ol) (90% CI) [p] 0.40 (0.126, 0.679) [0.017] 0.04 (−0.230, 0.304) [0.819]
Mean change a 9 mon hs om baseline (SE) 0.67 (0.10) 0.48 (0.10) 0.36 (0.10)
Di e ence (in e en ion –con ol) (90% CI) [p] 0.31 (0.068, 0.548) [0.035] 0.13 (−0.102, 0.358) [0.358]
Mean alues we e es ima ed using a linea mixed-e ec s model adjus ed o baseline alues and gende . See “Me hods”sec ion o ou come de ini ions
RAVLT Rey Audi o y Ve bal Lea ning Tes , SE s anda d e o , CI con idence in e als
Mahend an e al. T ials (2018) 19:615 Page 6 o 10
imp o e cogni i e unc ion in speci ic domains in elde ly
people wi h MCI [4]. Compa ed o p e ious s udies on
AT and MRA, his RCT was me hodologically igo ous
and sugges s ha AT has a g ea e bene i han MRA. AT
had mo e signi ican e ec s han MRA wi h imp o e-
men s in memo y, a en ion, isuo-spa ial abili ies and ex-
ecu i e unc ion a 3 mon hs and which was sus ained in
he memo y domain a 9 mon hs. We posi ha bo h he
cogni i e e alua ion o a wo ks and he physical c ea ion
o a pieces, ollowed by discussions ha in ol e di e en
cogni i e p ocesses, con ibu ed o he imp o emen s wi h
AT (Lee R, Wong J, Wong LS, Gandhi M, Raw ae I, Feng
L, Kua EH, Mahend an R: A he apy o he p e en ion
o cogni i e decline, in p epa a ion). AT deli e ed as a
combina ion echnique o “a -as- he apy”and “a -psy-
cho he apy”may ha e been he c i ical p ocess ha led o
he cogni i e imp o emen s.
Resea ch o da e suppo s he impac o a on he
b ain wi h se e al psychological [19,20] and physio-
logical [21] p ocesses in ol ing he in eg a ion o sen-
so y inpu , in e nal decision-making and emo ional
p ocessing, and join a en ion, wi h accompanying
neu oana omical changes in unc ional connec i i y in
he de aul mode ne wo k [22] and ac i a ion in egions
such as he o bi o on al co ex [23].
While s udies sugges ha elome e leng hs espond o
li es yle and mindse ia elome ase ac i i y, (an in acellu-
la enzyme wi h a RNA e e se ansc ip ase componen
ha leng hens elome es) [24], ecen wo k e lec s ha
oxida i e s ess and in lamma ion need o be e e sed o
leng hening o occu [25]. Sho elome e leng h is associ-
a ed wi h impai ed cogni i e pe o mance [26]. The be e
cogni i e sco es associa ed wi h elome e leng hening sug-
ges s ha he psychosocial in e en ions we e e ec i e in
Table 3 Es ima ed mean change in anxie y, dep ession, and sleep quali y ou comes a 3-mon hs and 9-mon hs
A he apy
(n= 22)
Music eminiscence ac i i ies
(n= 24)
Con ol
(n= 22)
Numbe o pa icipan s assessed a 3 mon hs 19 23 18
Numbe o pa icipan s assessed a 9 mon hs 18 22 18
Ge ia ic Anxie y In en o y o al sco e
Mean change a 3 mon hs om baseline (SE) −0.45 (0.72) −0.98 (0.66) −1.25 (0.75)
Di e ence (in e en ion –con ol) (90% CI) [p] 0.80 (−0.92, 2.52) [0.444] 0.27 (−1.39, 1.92) [0.791]
Mean change a 9 mon hs om baseline (SE) −0.17 (0.71) −0.75 (0.65) −0.05 (0.72)
Di e ence (in e en ion –con ol) (90% CI) [p]−0.12 (−1.79, 1.56) [0.909] −0.70 (−2.30, 0.91) [0.474]
Ge ia ic dep ession scale o al sco e
Mean change a 3 mon hs om baseline (SE) −0.76 (0.62) −0.75 (0.57) −1.22 (0.64)
Di e ence (in e en ion –con ol) (90% CI) [p] 0.46 (−1.02, 1.93) [0.610] 0.46 (−0.96, 1.89) [0.590]
Mean change a 9 mon hs om baseline (SE) −1.06 (0.62) −1.15 (0.57) −0.47 (0.63)
Di e ence (in e en ion –con ol) (90% CI) [p]−0.59 (−2.05, 0.88) [0.509] −0.67 (−2.08, 0.74) [0.431]
Sleep quali y isual analog scale
Mean change a 3 mon hs om baseline (SE) 6.94 (4.39) −2.09 (4.09) 3.56 (4.62)
Di e ence (in e en ion –con ol) (90% CI) [p] 3.39 (−7.15, 13.92) [0.595] −5.65 (−15.85, 4.55) [0.361]
Mean change a 9 mon hs om baseline (SE) 3.44 (4.17) 4.51 (3.82) 2.06 (4.29)
Di e ence (in e en ion –con ol) (90% CI) [p] 1.39 (−8.54, 11.32) [0.817] 2.46 (−7.08, 11.99) [0.670]
Mean alues we e es ima ed using a linea mixed-e ec s model adjus ed o baseline alues and gende . See “Me hods”sec ion o ou come de ini ions
SE s anda d e o , CI con idence in e als
Table 4 Es ima ed mean change in elome e leng h a 3 mon hs and 9 mon hs
A he apy
(n= 22)
Music eminiscence ac i i ies
(n= 24)
Con ol
(n= 22)
Numbe o pa icipan s assessed a 3 mon hs 19 23 NA
Numbe o pa icipan s assessed a 9 mon hs 18 22 18
Mean change a 3 mon hs om baseline (SE) 197 (197) −41 (186) NA
Mean change a 9 mon hs om baseline (SE) 552 (177) 292 (163) 573 (182)
Di e ence (in e en ion –con ol) (90% CI) [p]−22 (−444, 400) [0.932] −281 (−686, 124) [0.252]
Mean alues we e es ima ed using a linea mixed-e ec s model adjus ed o baseline alues and gende . See “Me hods”sec ion o ou come de ini ions
NA no applicable ( elome e leng h was no collec ed in he con ol g oup a 3 mon hs pos baseline), SE s anda d e o , CI con idence in e als
Mahend an e al. T ials (2018) 19:615 Page 7 o 10
e e sing cellula -le el in lamma o y mechanisms ha
con ibu e o cogni i e decline [27].
The s udy challenges he iew ha in e en ions a e
ine ec i e in MCI and suppo s he use o “single”psy-
chosocial in e en ions when mul i ace ed in e en ions
a e no eadily a ailable o possible [2]. The equency o
guided in e en ions a weekly in e als, was also su i-
cien ly e ec i e in add essing cogni i e decline, when a
p e ious s udy had sugges ed engagemen in s imula ing
ac i i ies a leas a wice a week o educe he isk o de-
men ia by 50% [28]. Finally, he sho ime ame in which
cogni i e imp o emen s we e no ed in he cogni i e do-
mains, especially memo y and execu i e unc ion is pa -
icula ly signi ican o pa ien s wi h MCI o ea ly
demen ia whe e ea ly in e en ions could inc ease he
chances o delaying o e en e e sing cogni i e decline.
O no e a e he imp o emen s in he CG e en hough
hey we e no o e ed any in e en ion and allowed o con-
inueli easusual.Whilei maybea gued ha suchim-
p o emen s e lec a lack o speci ici y o he in e en ions,
i mus be no ed ha 40% o he CG e u ned o employ-
men du ing he s udy du a ion, which was no disallowed
in he s udy p o ocol, and which could ha e accoun ed o
he imp o emen in cogni ion sco es. Employmen may
ha e p o ided cogni i e s imula ion, and he inc ease in
daily physical ac i i y, likely con ibu ed o he gains seen
in he CG. Bu hese gains we e associa ed wi h a nega i e
e ec as hei mood s a e was a ec ed as e lec ed in he
wo sened subsynd omal anxie y sco es in he CG.
The e a e some limi a ions in his s udy ha need o be
highligh ed. As his was a pilo s udy, he sample size was
small and no powe ed o hypo hesis es ing. The e o e,
he p alues should be in e p e ed wi h cau ion. The ma-
jo i y o s udy pa icipan s we e emale. A simila p epon-
de ance o emale pa icipan s was no ed in ou o he
psychosocial in e en ion s udies; i is because many men
a e s ill employed and unable o commi o pa icipa ion
in s udies. Double blinding, while ideal, is no possible in a
s udy o his na u e. Howe e , we ensu ed ha assesso s
we e blind o he s udy in e en ion a m and pa icipan s
we e also eminded no o discuss hei in e en ion wi h
he assesso s. Al hough we excluded pa icipan s wi h de-
p ession and anxie y, i is s ill possible ha he imp o e-
men s in subsynd omal dep ession and subsynd omal
anxie y could ha e had a small con ounding e ec on
cogni i e unc ioning. We also did no in e iew pa ici-
pan s o assess hei emo ional s a es. Fu he mo e, i
could be a gued ha he design o he AT in e en ion,
wi h he wo componen s, equi es mo e ac i e pa icipa-
ion han he MRA. Finally, we need o conside whe he
socializa ion had an impac on he esponse o he in e -
en ions and i has been sugges ed ha his is possible ia
mechanisms such as join a en ion and heo y o mind
[21,29]. Al hough his was no examined, and may no
ha e made a signi ican con ibu ion o he esul s as pa -
icipan s we e independen and communi y-li ing elde ly
wi h no psychia ic illnesses, hese a eas none heless de-
se e u he inqui y.
The easibili y o his ial was dependen on se e al
ac o s he mos impo an being he in e en ion design
and he p og am deli e y. The AT in e en ion was
mo e han jus a pa icipa i e ac i i y desc ibed in o he
s udies bu in ol ed bo h a iewing and cogni i e
e alua ion, and a c ea ion wi h discussion, analysis,
and sel -e alua ion. The well-de eloped in e en ion
unde wen se e al e iews by he cu a o s, he apis s,
and s udy in es iga o s and may no be eadily deli e ed
excep by ained a he apis s, which would be a limi -
ing ac o especially o la ge sample sizes. Addi ionally,
iden i ying elde ly people wi h MCI equi es bo h clin-
ical e alua ion and de ailed neu opsychological es ing.
The e we e also se e al aspec s o he s udy logis ics ha
mus be no ed. A pieces could no be emo ed om
he Na ional Galle y o NUS Museum and so pa ici-
pan s had o go o hese enues o he a iewing ses-
sions. To ensu e ha he elde ly would a i e sa ely and
on ime, anspo a ion was a anged and s a accom-
panied hem om he communi y Resea ch Cen e
which added o s udy cos s. Buy-in and suppo om
he Na ional Galle y and NUS Museum we e c ucial in
unde aking he s udy [30,31].
Conclusions
While he e a e inhe en issues in he s udy design, i
would be e y di icul o con ol o hese in unde aking
a s udy o his na u e. Al hough hese ac o s may limi
he gene alizabili y o his RCT, his pilo RCT suppo s
he e ec i eness o cogni i e s imula ion p og ams, in pa -
icula a he apy o elde ly peopoe wi h mild cogni i e
impai men and highligh s he need o u he e alua ion
o hese in e en ions and eplica ion on a la ge scale.
Addi ional ile
Addi ional ile 1: Table S1. Es ima ed mean and change om baseline
(95% con idence in e al) in memo y- ela ed neu opsychological
ou comes a 3 mon hs and 9 mon hs. Table S2. Es ima ed mean and
change om he baseline (90% and 95% con idence in e als) in
neu opsychological ou comes (o he han memo y- ela ed) a 3 mon hs
and 9 mon hs. Table S3. Es ima ed mean and mean change om
baseline (95% con idence in e al) in anxie y, dep ession and sleep
quali y ou comes a 3 mon hs and 9 mon hs. Table S4. Es ima ed mean
change in elome e leng h a 3 mon hs and 9 mon hs. (DOCX 26 kb)
Abb e ia ions
AT: A he apy; CG: Con ol g oup; CI: Con idence in e al; GDS: Ge ia ic
Dep ession Scale; GIA: Ge ia ic Anxie y Index; MCI: Mild cogni i e
impai men ; MRA: Music eminiscence ac i i y; p:P alue; RAVLT: Rey
Audi o y Ve bal Lea ning Tes ; RCT: Randomized con olled ial; SD: S anda d
de ia ion; TaRA: T aining and Resea ch Academy; WAIS: Wechsle Adul
In elligence Scale
Mahend an e al. T ials (2018) 19:615 Page 8 o 10
Acknowledgemen s
Na ional Galle y and he Na ional Uni e si y o Singapo e Museum.
Funding
Funding was by Resea ch Dona ions om Kwan Imm Thong Hood Cho
Temple and Lee Kim Tah Holdings P e L d.
A ailabili y o da a and ma e ials
The da ase s suppo ing he conclusions o his a icle a e a ailable on
eques o he co esponding au ho . The da a will be sha ed o u he
esea ch subjec o app o al om he ins i u ion e hics boa d.
Au ho s’con ibu ions
RM is he P incipal In es iga o o he s udy and was in ol ed in designing
he s udy, ca ying i ou , and w i ing he manusc ip . MG was in ol ed in
he s udy design, analysis planning, analyzing da a, and w i ing he
in e p e a ion o he esul s. RBM analyzed he da a. JW e iewed he
li e a u e and ca ied ou he s udy. JF was in ol ed in designing he s udy
and in he neu oimaging. IR was in ol ed in designing he s udy and in
manusc ip e iew. APK was in ol ed in designing he s udy, and wi h MMK,
he elome e wo k. LF was in ol ed in designing he s udy and in da a
e iew. EHK o mula ed he esea ch ques ion, and was in ol ed in
designing he s udy and e iewing he manusc ip . All au ho s ead and
app o ed he inal manusc ip .
E hics app o al and consen o pa icipa e
The s udy p o ocol was e iewed by he E hics Boa d o he Na ional Uni e si y
o Singapo e Ins i u ional Re iew Boa d (NUS-IRB Re Code: B-16-095) and he
app o al was da ed 6 June 2016. W i en in o med consen was ob ained om
e e y pa icipan and con iden iali y was s ic ly main ained. All da a we e
anonymized and main ained acco ding o E hics Boa d guidelines.
Consen o publica ion
No applicable.
Compe ing in e es s
The au ho s decla e ha hey ha e no compe ing in e es s.
Publishe ’sNo e
Sp inge Na u e emains neu al wi h ega d o ju isdic ional claims in
published maps and ins i u ional a ilia ions.
Au ho de ails
1
Depa men o Psychological Medicine, Yong Loo Lin School o Medicine,
Na ional Uni e si y o Singapo e, NUHS Towe Block, Le el 9, 1E Ken Ridge
Road, Singapo e 119228, Singapo e.
2
Depa men o Psychological Medicine,
Na ional Uni e si y Hospi al, NUHS Towe Block, Le el 9, 1E Ken Ridge Road,
Singapo e, Singapo e.
3
Academic De elopmen Depa men , Duke-NUS
Medical School, 8 College Road, Singapo e, Singapo e.
4
Depa men o
Bios a is ics, Singapo e Clinical Resea ch Ins i u e, 31 Biopolis Way, Singapo e,
Singapo e.
5
Cen e o Quan i a i e Medicine, Duke-NUS Medical School, 8
College Road, Singapo e, Singapo e.
6
Tampe e Cen e o Child Heal h
Resea ch, Uni e si y o Tampe e and Tampe e Uni e si y Hospi al, Tampe e,
Finland.
7
Depa men o Psychia y, Sengkang Gene al Hospi al, 110
Sengkang Eas Way, Singapo e, Singapo e.
8
Cance Science Ins i u e o
Singapo e, Na ional Uni e si y o Singapo e, 14 Medical D i e, Singapo e,
Singapo e.
9
Medical Science Clus e , Yong Loo Lin School o Medicine,
Na ional Uni e si y o Singapo e, Singapo e, Singapo e.
Recei ed: 13 Ap il 2018 Accep ed: 15 Oc obe 2018
Re e ences
1. Ame ican Psychia ic Associa ion. Diagnos ic and s a is ical manual o men al
diso de s, Fi h Edi ion. A ling on: Ame ican Psychia ic Associa ion; 2013.
2. Hun ley JD, Gould RL, Liu K, Smi h M, Howa d RJ. Do cogni i e in e en ions
imp o e gene al cogni ion in demen ia? A me a-analysis and me a-
eg ession. BMJ Open. 2017;7. Published Online Fi s : 30 Aug 2017. h ps://
doi.o g/10.1136/bmjopen-2015-005247co 1.
3. Ma in M, Cla e L, Al gassen AM, Came on MH, Zehnde F. Cogni ion-based
in e en ions o heal hy olde people and people wi h mild cogni i e
impai men (Re iew). Coch ane Da abase Sys Re . 2011;(1):CD006220.
h ps://doi.o g/10.1002/14651858.CD006220.pub2
4. Coope C, Li R, Lyke sos C, Li ings on G. A sys ema ic e iew o ea men s
o mild cogni i e impai men . B J Psychia y. 2013;203(3):255–64.
5. Coope C, Somme land A, Lyke sos CG, Li ings on G. Modi iable p edic o s
o demen ia in mild cogni i e impai men : a sys ema ic e iew and me a-
analysis. Am J Psychia y. 2015;17(4):323–34.
6. Robe s R, Knopman DS. Classi ica ion and epidemiology o MCI. Clin Ge ia
Med. 2013;29(4). h ps://doi.o g/10.1016/j.cge .2013.07.003.
7. Manly JJ, Tang MX, Schup N, S e n Y, Vonsa el JP, Mayeux R. F equency
and cou se o mild cogni i e impai men in a mul ie hnic communi y. Ann
Neu ol. 2008;63(4):494–506. h ps://doi.o g/10.1002/ana.21326.
8. Raw ae I, Mahend an R, Yu JH, Fam J, Feng L, Kua EH. Psychosocial
in e en ions wi h a , music, Tai Chi and mind ulness o subsynd omal
dep ession and anxie y in olde adul s: a na u alis ic s udy in Singapo e.
Asia Paci ic J Psychia y. 2015;7(3):240–50.
9. Pe e sen RC. Mild cogni i e impai men as a diagnos ic en i y. J In Med.
2004;256(3):183–94.
10. Mahend an R, Raw ae I, Fam J, Wong J, Kuma AP, Gandhi M, Jing KX, Feng
L, Kua EH. A The apy and Music Reminiscence Ac i i y in he P e en ion o
Cogni i e Decline: s udy p o ocol o a andomized con olled ial. BMC
T ials. 2017;18:324. h ps://doi.o g/10.1186/s13063-017-2080-7
11. Ande sson C, Lindau M, Almk is O, Eng eld P, Johansson SE, E iksdo e
Jonhagen M. Iden i ying pa ien s a high and low isk o cogni i e decline
using Rey Audi o y Ve bal Lea ning Tes among middle-aged memo y clinic
ou pa ien s. Demen Ge ia Cogn Diso d. 2006;21:251–9.
12. Yin S, Zhu X, Huang X, Li J. Visuospa ial cha ac e is ics o an ea ly Chinese
popula ion: esul s om he WAIS-R block design es . F on Aging Neu osci.
2015;7(17). h ps://doi.o g/10.3389/ nagi.2015.00017.
13. Tombaugh TN. T ail making Tes A and B: no ma i e da a s a i ied by age
and educa ion. A ch Clin Neu opsychol. 2004;19(2):203–4. h ps://doi.o g/10.
1016/S0887-6177(03)00039-8.
14. Lee CK, Collinson SL, Feng L, Ng TZ. P elimina y no ma i e
neu opsychological da a o an elde ly Chinese popula ion. Clin
Neu opsychol. 2012. h ps://doi.o g/10.1080/13854046.2011.652180.
15. Nyun MSZ, Fones C, Ni i M, Ng TP. C i e ion-based alidi y and eliabili y o
he ge ia ic Dep ession Sc eening Scale (GDS-15) in a la ge alida ion
sample o communi y-li ing Asian olde adul s. Aging Men Heal h. 2009;
13(3):376–82.
16. Pachana NA, By ne GJ, Siddle H, Koloski N, Ha ley E, A nold E. De elopmen
and alida ion o he Ge ia ic Anxie y In en o y. In Psychoge ia ics. 2007;
19(1):103–14.
17. Rane G, Koh WP, Kanchi MM, Wang R, Yuan JM, Wang X. Associa ion
be ween leukocy e elome e leng h and plasma homocys eine in a
Singapo e Chinese popula ion. Reju ena ion Res. 2015;18(3):203–10.
18. Goh ing J, Fulche N, Jacak J, Riha K. TeloTool: a new ool o elome e
leng h measu emen om e minal es ic ion agmen analysis wi h
imp o ed p obe in ensi y co ec ion. Nucleic Acids Res. 2014;42(3):e21.
19. Kon son KL, Megjhani M, B an ley JA, C uz-Ga za JG, Nakagome S, Roble o
D, Whi e M, Ci illico E, Con e as-Vidal JL. You b ain on a : eme gen
co ical dynamics du ing aes he ic expe iences. F on Hum Neu osci. 2015;9:
626. h ps://doi.o g/10.3389/ nhum.2015.00626.
20. Scai e M, B une JS. The capaci y o join isual a en ion in he in an .
Na u e. 1975;253:265–6.
21. Bullo NJ, Rebe R. The a ul mind mee s a his o y: owa d a psycho-
his o ical amewo k o he science o a app ecia ion. Beha B ian Sci.
2013;36:123–37.
22. Bolwe k A, Mack-And ick J, Lang FR, Do le A, Malho ne C. How a
changes you b ain: di e en ial e ec s o isual a p oduc ion and
cogni i e a e alua ion on Func ional b ain connec i i y. PLoS One. 2014;
9(7):e101035. h ps://doi.o g/10.1371/jou nal.pone.0101035.
23. B own S, Gao X, Tisdelle L, Eickho SB, Lio i M. Na u alizing aes he ics: b ain
a eas o aes he ic app aisal ac oss senso y modali ies. Neu oimage2011; 58:
250–258.
24. Epel E. How “ e e sible”is elome ic aging? Cance P e Res. 2012;5(10):1163–8.
25. Kiecol -Glase J, Epel E, Belbu y M, And idge R, Lin J, Glase R, Mala key WB,
Hwang BS, Blackbu n E. Omega-3 a y acids, oxida i e s ess, and leukocy e
elome e leng h: a andomized con olled ial. B ain Beha Immun. 2013;28:
16–24.
26. Hagg S, Zhan Y, Ka lson R, Ge i sen L, Plone A, an de Lee SJ, B oe L,
Deelen J, ( o he ENGAGE Conso ium), e al. Sho elome e leng h is
Mahend an e al. T ials (2018) 19:615 Page 9 o 10