RESEAR CH
the bmj |
BMJ
20 22;377 :e0687 88 | doi: 10 .11 36/bmj- 2021 -0687 88 1
Mul tico mponent inter v ention to p r ev ent mobility disability in fr ail
older adul ts: rando mised co ntrolled trial (SPRINT T pr ojec t)
Roberto Bernabei, 1,2 F ranc esc o Landi, 1,2 , Ricc ardo Calv ani, 1 Matteo Cesari, 3, 4
Sus anna Del Signore, 5 Stefan D An ker , 6 Raphael Bejuit, 7 Philippe Bordes, 7 Ant onio Cherub ini, 8
Alfon so J Cruz-Jentoft, 9 Mau ro Di Bari, 10 Tim Friede, 11,12 C armen Gorostiag a Ayest arán, 13
Harmonie Goye au, 7 Pálmi V Jónsson, 14 Mak oto K ashiw a, 15 Fabrizia Latt anzio, 8
Marc ello Maggio , 16,17 Luc a Mariotti, 2 Ram R Mi ll er , 18 Leocadio R odriguez-Mañ as, 19
Regina R oller -Wirn sberger , 20 Ingrid Rýzn aro vá, 21 Joachim Scholp p, 22 Annemie M W J Sc hols, 23
C ornel C Sie ber , 24 Alan J Sinclair , 25 Anna Skal sk a, 26 Timo Strandberg, 27,28 Achill e T chal la, 29
Eva Topin ková, 30 Matteo T osato , 1 Bruno Vel las, 31 Steph an von Haehlin g, 12,32 Marc o Pahor , 33
Ronenn R oubenoff, 34 Emanuel e Marzetti, 1,2 on beh alf of the SPRINTT con sor tium
Abstr Act
Obje ctive
T o determine whether a multic omponent intervention
based on physic al activity with tec hnologic al su ppor t
and nut ritional cou nselling prevents mo bility dis ability
in older adults with ph ysic al frai lty and sarc openia.
Des ign
Ev aluator blinded, r andomised c ontrol led tri al.
set ting
16 clinic al s ites ac ross 11 E urope an cou ntries, January
2016 to 31 October 2019.
P artic iP ants
1519 commu nity dwelling men and w omen aged 70
ye ars or o lder with phy sical frai lty and s arcopenia,
operation alised as the c o-occu rrenc e of low fu nctional
stat us, defined as a s hort phys ical performanc e
battery (SPPB) sc ore of 3 to 9, low appendicu lar le an
ma ss, and ability t o independently w alk 400 m. 760
participants w ere randomi sed to a mu lticomponent
intervention and 759 rec eived educ ation on healthy
ageing (c ontrols).
interventiOns
The multic omponent intervention c omprised moderate
intensity physic al activity twic e weekly at a cent re
and u p to fou r times week ly at home. Actimet ry data
were used to tailor the intervention. Participants also
rec eiv ed person alised nutrition al c ounselling. C ontrol
participants r ecei ved education on he althy ageing
once a month. Int erventions and f ollow-u p lasted for
up to 36 months.
Main Outc OMe Measures
The primary outc ome was mob ility di sability (in ability
to independently w alk 400 m in <15 minutes).
Pers istent mob ility di sability (in ability to wal k 400
m on two c onsecutiv e occ asions) and chang es
from ba seline to 24 and 36 months in phy sical
performanc e, muscle st rength, and appendicul ar
lean m as s were an alysed as pre-planned sec ondar y
outc omes. Primary c omparisons were conducted
in participants w ith baseline S PPB scores of 3-7
(n=1205). Those with SPPB scor es of 8 or 9 (n=314)
were an alysed separately for explorat ory purpose s.
resul ts
Mean ag e of the 1519 participants (1088 women)
wa s 78.9 (st andard deviation 5.8) years. The
aver age follow-u p was 26.4 (S D 9.5) months. Among
participants w ith SPPB sc ores of 3-7, mob ility
dis ability occ urred in 283/605 (46.8%) assigned
to the multic omponent intervention and 316/600
(52.7%) cont rols (haz ard ratio 0.78, 95% c onfidence
interval 0.67 to 0.92; P=0.005). P ersistent mobility
dis ability occ urred in 127/605 (21.0%) par ticipants
as signed to the multic omponent intervention and
150/600 (25.0%) cont rols (0.79, 0.62 to 1.01;
P=0.06). The between grou p differenc e in SPPB sc ore
wa s 0.8 points (95% confidenc e inter val 0.5 t o 1.1
points; P<0.001) and 1.0 point (95% confidenc e
interval 0.5 to 1.6 points; P<0.001) in f avour of the
multic omponent intervention at 24 and 36 months,
respecti vely . The decline in handgrip s trength at
24 months w as smal ler in women as signed to the
multic omponent intervention th an to cont rol (0.9
k g, 95% confidenc e inter val 0.1 t o 1.6 k g; P=0.028).
W omen in the multicomponent int ervention arm lost
0.24 k g and 0.49 k g less appendicu lar le an mas s than
cont rols at 24 months (95% confidence interval 0.10
to 0.39 k g; P<0.001) and 36 months (0.26 to 0.73
k g; P<0.001), respecti vely . Serious adv erse events
occ urred in 237/605 (39.2%) participants a ssigned
For num bered affiliation s see
end of the article
C orrespondence to: E Marzetti
Centre for Geriat ric Medicine
(CeMI), Fondaz ione Policlinic o
Universit ario “ Agostino Gemelli”
IRCC S, Università C attolica del
Sacro C uore, Rome, 00168,
Italy em anuele.marzetti@
policlinic ogemelli.it
(or @Emanuel00962649 on
T witter;
ORCID 0000-0001-9567-6983)
Additional material is p ubli shed
online only . T o view plea se visit
the journ al online.
cite thi s a s: BMJ 2022;377:e068788
http:/ /dx. doi.org/10.1136/
bmj-2021-068788
Acc epted: 22 March 2022
WhA t is AlreAd y knoWn on this t opic
Mobility is a primary tar get to maint ain function and fos ter activ e ageing
Lifestyle int erventions (eg, phy sic al activity alone or with nut ritional cou nselling/
supp lementation) are fe asible, s afe, and effectiv e for improving ph ysic al function
in older adults at ris k of mobility dis ability
The identification of a condition enc ompass ing reduc ed ph ys ical function and
targ et org an damage (ie, muscle f ailure) might s timulat e the development of
preventi ve interventions against disability in older people who are at ris k
WhA t this stud y Adds
Phys ical frailty and sar copeni a is a novel, obj ectively me asu rable condition th at
identifies a s ubset of the older pop ulation at risk of adverse health r elated events,
including mobi lity dis ability , whose medic al needs are currently u nmet
A multic omponent intervention ba sed on moder ate intens ity ph ysic al activity with
technologic al support and nut ritional cou nselling was assoc iated with a r eduction
in the incidenc e of mob ility di sability o ver 36 months of fo llow-u p in older adults
with ph ysic al frailty and sarc openia
copyright. on 18 November 2022 at Biblioteca di Medicina e Chirurgia. Protected by http://www.bmj.com/ BMJ: first published as 10.1136/bmj-2021-068788 on 11 May 2022. Downloaded from
RESEAR CH
2 doi: 10 .1136/bmj- 2021 -0687 88 |
BMJ
20 22;377 :e0687 88 | the bmj
to the multic omponent intervention and 216/600
(36.0%) cont rols (ris k ratio 1.09, 95% confidenc e
interval 0.94 to 1.26). In p articipants w ith S PPB
scor es of 8 or 9, mobility disabi lity oc curr ed in 46/155
(29.7%) in the multic omponent intervention and
38/159 (23.9%) cont rols (haz ard ratio 1.25, 95%
confidenc e inter val 0.79 to 1.95; P=0.34).
c Oncl usiOns
A multic omponent intervention w as a ssociated with
a reduction in the incidenc e of mobility dis ability in
older adults with ph ysic al frai lty and s arcopenia and
SPPB sc ores of 3-7. Phy sical f railty and sarc openia
may be targeted to preserve mob ility in vulnerable
older people.
trial re gistra tiOn
C linicalT rials.g ov NCT02582138.
introduction
In ad vanc ed ag e, impaired mobility is associated
with hig her risk of disability , poor quality of life,
admission to hospital, admission t o residential car e,
and death, 1 2 as well as gr eater heal thcare c osts. 3
The disablin g trajectory of mo bility limited older
adul ts mig ht be deflected by lifesty le interventions . 4
In 2013, the Innovati ve Medicines Initiati v e J oint
Undertaking , a public-pri vate partnership betw een
the Eur opean Union and the E uropean F ederation of
Pharmac eutical Industries and A ssociations, proposed
to focus on “ph ysical frailty and sarc openia ” as a
prot otypical g eriatric condition to be tar get ed for
ad vancin g the care of older people at risk of disability . 5
The condition of int erest was e xpected t o encompass
red uced ph ysical function and tar get or gan dama g e
(ie, low muscle mass), both of whic h should be
objecti v el y measurable. The design and v alidation
of ph ysical frailty and sarcopenia ar e piv otal in
providin g reg ulatory a uthorities with the framework of
a nov el nosological c ondition with clinical rele vance
and a well defined pathoph ysiolo gy , that could be
adopted t o dev elop specific ther apeutics following
the standards of dru g researc h. 6 The Sarc openia and
Ph ysical fRailty IN older people: mul ti- componenT
T reatment strategies ( SPRINT T) proj ect was therefor e
designed t o elaborate a new , objecti ve definition of
ph ysical frailty and sarcopenia—conceptualised as
a pre -disability condition with muscle failur e as its
biological substratum, 7 identify and characterise a
population with ph ysical frailty and sarcopenia at risk
of ad v erse outcomes , and test int erv entions in this
population. W e defined ph ysical frailty and sarcopenia
as the combination of low ph ysical function and low
appendicular lean mass in the a bsence of actual
mobility disability . 7
W e designed a mul ticentr e, ev aluat or blinded,
randomised controlled trial t o determine whether
a multic omponent interv ention based on phy sical
acti vity with technological support and nutritional
counsellin g would r educ e the risk of mobility disability
in older adul ts with ph ysical frailty and sarcopenia
compar ed with a health y ag eing lifestyle educational
pro gramme.
Methods
study design
The SPRINT T trial was a multic entre randomised
contr olled trial cond ucted from J anuary 2016 to
31 October 2019 at 16 sit es across 11 European
countries . 8 The Eur opean Medicines A g ency ac cept ed
the trial methodolog y and anal ytical strategy d uring
an ad hoc scientific ad vice pr ocedur e that was
complet ed in earl y 2015. A summary description
of the prot ocol is a vailable on ClinicalT rials.g ov
(NCT02582138). Details were pr ovided in a dedicated
publication 9 and are incl uded in the supplementary
appendix. The U ni versità Cattolica del Sacr o C uore in
Rome, I taly , coordinat ed trial acti vities. T rial sites are
listed in the supplementary a ppendix. As part of the
Innovati ve Medicines Initiati ve J oint Undertaking of
the EU (ww w .imi.europa. eu), member companies of
the Eur opean F eder ation of Pharmaceutical Ind ustries
and Associations g a ve in-kind support. The academic
members provided an independent interpr etation of
resul ts. An independent statistician r eplicated and
v erified the anal yses.
Participants
Participants wer e men and women a g ed 70 y ears or
older with ph ysical frailty and sarcopenia, defined as
ha ving a short ph ysical performanc e battery ( SPPB) 10
scor e of 3 to 9 points (sc ores rang e from 0 to 12, with
lower sc ores indicating poorer ph ysical function),
low appendicular lean mass ac cor ding to se x specific
cut -points rec ommended by the Foundation for the
National Institutes of H ealth sarcopenia pr oject, 11
and the absenc e of mobility disability , operationalised
as being a ble to c omplete a 400 m walk t est in less
than 15 minutes without sittin g, st opping for mor e
than one minute, r ecei ving help, or usin g a walk er . 12
This operational definition of ph ysical frailty and
sarc openia was discussed and agr eed with EMA.
Main ex clusion criteria w ere self -reported w alking
disability , co gniti ve impairment (defined as a mini-
mental state e xamination 13 scor e <24/30), terminal
illness, participation in a structured ph ysical acti vity
pro gramme, contraindications to safel y enga ge in trial
acti vities as judg ed by local study doct ors, and plans to
relocat e out of the stud y area within at least tw o y ears.
interventions
The multic omponent interv ention and the health y
ag eing lifestyle ed ucational pro gramme are
ext ensi vel y described elsewhere. 9 Both interv entions
wer e administered for up to 36 months , depending
on when participants w ere r ecruited durin g the
trial. The multic omponent interv ention comprised a
combination of moderate int ensity ph ysical acti vity
with technolo gical suppor t and nutritional counsellin g.
Ph ysical acti vity included aer obic, stren gth, flexibility ,
and balance e x ercises. 14 The interv ention was di vided
into an adoption phase (w eeks 1-52) and maintenance
phase (week 53 t o end of the trial). During the
adoption phase, two c entre based ph ysical acti vity
sessions wer e conducted w eekly . These sessions were
copyright. on 18 November 2022 at Biblioteca di Medicina e Chirurgia. Protected by http://www.bmj.com/ BMJ: first published as 10.1136/bmj-2021-068788 on 11 May 2022. Downloaded from
RESEAR CH
the bmj |
BMJ
20 22;377 :e0687 88 | doi: 10 .11 36/bmj- 2021 -0687 88 3
used to initiat e the aerobic pr ogramme and safel y
introd uce participants to the str ength, str etchin g,
and balance c omponents. Centr e based sessions wer e
pro gressi vel y supplemented by home based ph ysical
acti vity sessions: once w eekl y durin g weeks 1-4, twic e
weekl y during w eeks 4-8, and up to four times weekl y
durin g weeks 9-52. The maint enance phase in v ol ved
two c entre based ph ysical acti vity sessions and up to
four home based sessions weekl y . T raining intensity
was adapt ed throu gh assessment of per cei ved ex er tion
by the Borg scale (ratings rang e from 6 to 20, with 6
repr esenting no e xertion at all and 20 r epresentin g
maximal ex ertion). 15 Par ticipants wer e asked to
walk at an intensity of 13 (some what har d). Low er
extr emity stren gthening e xer cises were performed
at an intensity of 15 or 16 (har d). A dherence w as
ascertained by re gisterin g centre att endance and
participant completed diaries on fr equency of home
based sessions. The t otal amount of ph ysical acti vity
was monitor ed for seven c onsecuti ve da ys at baseline
and ev ery six months using the acti vP AL3 actimeter
(P AL T echnologies, Glas gow , UK) worn on the thig h.
Instructors could r equest additional seven da y
actimetry rec ordin gs an ytime if indications su gg ested
participants wer e not compl ying with ph ysical acti vity
prescriptions . Instructors used the information to
provide participants with personalised feedbacks
on their performance g oals to be reached as part
of beha viour al strategies to maximise adher ence
and remov e possible disincenti ves. The nutritional
component was desi gned to support the effects of
the ph ysical acti vity programme. The interv ention
in vol ved indi vidualised nutritional assessments and
prescription of personalised dietary plans with tw o
main tar g ets: a dail y energy intak e of 25-30 kcal/
k g bodywei ght and a daily pr otein intak e of at least
1.0-1.2 g/kg bod yweig ht . 16 A three da y dietary
rec ord was c ollected at least onc e a y ear , followed
by an indi vidualised dietary interview . Adher ence
to nutritional pr escriptions was ascertained thr ou gh
reg ular contacts with stud y staff durin g which
participant feedback was collect ed and dietary plans
re view ed. Additional dietary assessments c ould be
performed at the discretion of the int erv entionist to
maximise adherenc e.
The health y ag eing lifestyle educational pr ogramme
consist ed of seminars and workshops on t opics
rele vant t o older adults (e g, v accinations , chronic
pain manag ement , gastroint estinal and urolo gical
problems , technological devic es, personal safety).
Meetings w ere offered in groups of 10-20 participants
once or twic e a month, with requir ed participation of at
least once a month. A short instructor led pr ogramme
(5-10 minutes) of upper e xtremity str etchin g exer cises
or some relaxation t echniques was offer ed at the end
of each meeting .
Outc omes
The primary outc ome was mobility disa bility ,
operationalised as the inability to c omplete the 400
m walk test in less than 15 minut es without sitting,
stoppin g for more than one minut e, requiring help ,
or using a walk er . 12 14 F or the test, participants were
asked t o complete 10 laps ar ound a 20 m course at
their usual pace without ov erex er ting themsel ves. The
400 m walk test w as administer ed after three months
of randomisation and ev ery six months from baseline.
If the 400 m walk test w as not performed, a stepwise
proc edur e was devised for out come adjudication. A
predefined al gorithm was a pplied to a utomatically
adjudicat e mobility disability based on a 4 m g ait speed
≤0.4 m/ s or >0.4 m/ s if par ticipants needed a walking
aid other than a sing le straight cane, medical r ecor ds,
or self -reported or pro xy reported walking disa bility .
Those participants who did not perform the 400 m
walk test and c ould not be automaticall y adjudicated
wer e evaluat ed by an independent committ ee based on
clinical varia bles, functional tests, and ad verse ev ents.
Participants wer e censor ed at their last successful 400
m walk test if the mo bility disability crit erion was not
met at the end of the trial, at their first consecuti ve visit
when mor e than nine months had elapsed betw een
two c onsecutiv e successful tests , or at the date of the
randomisation visit when no post -baseline 400 m walk
test was a vaila ble. Mobility disability was c onsidered
to be pr esent at the date participants failed the 400
m walk test, wer e unable t o attempt the test, did
not attempt the t est and wer e classified as mobility
disabled thr ou gh the adjudication proc ess, did not
attempt the t est and mobility disability c ould not be
adjudicat ed, or died.
The SPRINT T trial incl udes sev er al prespecified
secondary out comes (supplementary a ppendix).
Her e we r eport the secondary outc omes of per sistent
mobility disability , operationalised as failure t o
complet e the 400 m walk test on tw o consecuti ve
occasions or ina bility to c omplete the t est followed
by death, 17 and chang es from baseline to 24 and 36
months in measures of ph ysical performance, muscle
stren gth, and appendicular lean mass .
sample siz e calculation
The size of the stud y population was determined
to addr ess the main requir ements of the Innovati ve
Medicines Initiati ve J oint Undertaking to ev aluate
whether a mul ticomponent int ervention w ould
red uce the risk of incident mobility disa bility in older
adul ts with ph ysical frailty and sarcopenia, and
to characterise the ph ysical frailty and sar copenia
condition and obtain information on int ervention
effects across its w hole SPPB rang e (scores 3-9). T o
meet the first requirement, we performed a sample
size estimation based on information retrie v ed from
the Lifestyle Int erventions and Independence for
Elders (LIFE) stud y database 17 by running survi val
anal yses for mobility disa bility acc ordin g to differ ent
baseline SPPB scor e categ ories (<8 v 8 or 9). In LIFE,
the hazard of incident mobility disa bility was observed
to be si gnificantly red uced by ph ysical acti vity only in
participants with SPPB scores <8 (hazar d r atio 0.75,
95% confidenc e interv al 0.59 to 0.94; P=0.012).
W e therefore estimat ed that a sample of 1200 older
copyright. on 18 November 2022 at Biblioteca di Medicina e Chirurgia. Protected by http://www.bmj.com/ BMJ: first published as 10.1136/bmj-2021-068788 on 11 May 2022. Downloaded from
RESEAR CH
4 doi: 10 .1136/bmj- 2021 -0687 88 |
BMJ
20 22;377 :e0687 88 | the bmj
people with an SPPB scor e of 3 to 7, enrolled ov er
12 months, w ould provide 85% power (434 mobility
disability e v ents) to detect a 25% r eduction in the
hazard of mobility disa bility over a maximum follow -
up of 36 months, c onsidering a dr opout rate of 25%
ov er two y ears and a log -rank test with a 5% two sided
α lev el. 9 T o address the second objecti v e, we c hose to
enrol an e xploratory sample of 300 participants with
low appendicular lean mass and an SPPB sc ore of 8 or
9. 9 The size of this subsample was determined based
on feasibility and resour ce a vailability . A hierarchical
testin g proc edure w as devised t o contr ol type I error
rate by testin g the primary endpoint in the whole stud y
population onl y in case of a significant r esult (P<0.05)
in participants with SPPB scores of 3 t o 7. 9
Based on a blinded interim sample size reassessment
at 11 months, w e prolong ed the accrual period by six
months. A sec ond blinded power r eassessment at
29 months re v ealed a number of mobility disability
ev ents that were low er than expected. W e extended the
follow -up by seven months t o maximise the probability
of reachin g the required number of ev ents and to allow
participants recruited d uring the last phase of accrual
to r ecei ve int ervention and be followed -up for 24
months. The maximum len gth of interv entions and
follow -up was kept at 36 months.
randomis ation and blinding
Eligible participants w ere in vited t o the stud y sites for
an in-person meeting , during which trial pr ocedures
and requir ements wer e mentioned ag ain. Participants
wer e then r andomised 1:1 to the multic omponent
interv ention or lifestyle education usin g a we b based
randomisation system with permut ed block alg orithm,
stratified by study sit e, sex, and SPPB sc ore cate g ory
(3-7 and 8 or 9). An ev aluat or blinded approac h
was used to pr eserve the trial inte grity . Ac cordin gly ,
outc ome assessors wer e una ware of gr oup assignment ,
clinic and laboratory measur ements, and interv ention
adherenc e.
safety
All stud y staff monitor ed participant safety and
reported thr ee categories of ad verse ev ents: serious
ad v erse ev ents, une xpected ad verse ev ents (those
potentiall y related to stud y proced ures or acti vities
and not listed in the informed c onsent form or stud y
prot ocol), and ad verse ev ents that occurred while the
participant was under the supervision or g uidance
of stud y staff either onsite or offsit e. Some adv erse
ev ents were further flag g ed as of special inter est if
falling int o prespecified cat eg ories (ie, abnormal
test r esults r equiring medical attention, emer genc y
department visits, fractures, outpatient sur gery , and
restrict ed acti vity possibly due t o study proc edur es).
An independent committ ee re viewed safety data onc e
a y ear .
stati stic al an alys is
All anal yses w ere performed ac cor ding t o a predefined
statistical anal ysis plan (supplementary a ppendix).
Baseline characteristics of participants allocated in
the two int ervention arms are described as means
(standard de viations) for continuous v ariables and
absol ute numbers ( percenta ges) for cate g orical
varia bles. Anal yses of interv ention effects were based
on the intention-t o -treat principle. F or the analy sis
of the primary efficacy endpoint (time t o the first
occurr ence of mobility disa bility or death from an y
ca use) we c ompared interv ention arms using a tw o
sided 5% α lev el log -rank test proc edur e stratified by
randomisation factors of site and sex. The primary
comparison was c onducted in randomised participants
with baseline SPPB scor es of 3 to 7. W e used a Cox
proportional hazard model stratified by randomisation
factors of site and se x to estimate the hazard ratio
of mobility disability betw een intervention gr oups
and the corr esponding 95% c onfidence interv al.
The K aplan-Meier method was used t o summarise
cumulati ve incidence functions . If the findings of this
primary anal ysis w ere statisticall y significant (P<0.05),
we w ould perform an additional analysis t o include the
exploratory gr oup of participants with SPPB scores of 8
or 9 onl y if no interaction was observ ed between SPPB
categ ory and interv ention arm. W e used the Ka plan-
Meier method to compar e the cumulativ e incidence
functions for the two int ervention arms between the
two SPPB cat egories. P respecified sub group anal yses
based on Co x proportional hazard models w ere
cond ucted t o determine whether int ervention effects
wer e influenced by baseline personal, clinical, or
functional characteristics.
W e analy sed secondary efficacy endpoints in the
two SPPB cat egories separatel y . A Cox pr oportional
hazard model stratified by randomisation factors of
site and se x was used to estimat e the hazard ratio of
persistent mobility disa bility between int erv ention
groups and the c orrespondin g 95% confidence
interv al. Chang es from baseline to 24 and 36 months
in SPPB scor e, hand grip strength, and a ppendicular
lean mass wer e analysed by mix ed effect models
with repeat ed measures . Models included the fix ed
categ orical effects of interv ention arm, the planned
time point , the randomisation factors of site and
sex, the int erv ention×time point interaction, and the
continuous fix ed covariates of baseline v alue and
baseline val ue×time point interaction. F or all analyses ,
a two sided P<0.05 w as consider ed to be statisticall y
significant.
W e analy sed safety data by intervention gr oup in
the two SPPB cat egories separatel y . Risk ratio with
95% confidenc e interv al was used to estimat e the
proba bility of experiencing an ad v erse ev ent .
All anal yses w ere run usin g S AS version 9.4 ( Cary ,
NC).
Patient and pu blic invo lvement
A dialog ue and knowledg e platform was established
at the beginning of the pr oject throu gh the mapping of
stakeholders (older ad ults’ r epresentati v es, heal thcare
professionals , and experts in bioethics, data security ,
pri vacy , storag e and use, and bioinformatics), and
copyright. on 18 November 2022 at Biblioteca di Medicina e Chirurgia. Protected by http://www.bmj.com/ BMJ: first published as 10.1136/bmj-2021-068788 on 11 May 2022. Downloaded from
RESEAR CH
the bmj |
BMJ
20 22;377 :e0687 88 | doi: 10 .11 36/bmj- 2021 -0687 88 5
their in vitation to conferenc e calls and in-person
meeting s focused on the operational definition of
ph ysical frailty and sarcopenia, tr eatment protocols ,
strategies for participant recruitment and en ga gement,
health lit eracy plans, and dissemination acti vities.
The platform was subsequentl y extended to incl ude
reg ulatory experts from the EU t o reach a consensus
on the definition of the tar g et population and trial
methodolog y . After trial commenc ement , quarterl y
telec onferenc es wer e held with EMA to discuss progr ess
of the proj ect , emer ging pr oblems, safety aspects , and
other rele vant e vents in the trial.
Educational c ontents w ere pr oduc ed for older
people and their caregi v ers. In particular , to promote
health lit eracy on the topics of frailty and sar copenia,
we de veloped leaflets that wer e freely downloada ble
from the pr oject w e bsite.
Participants wer e acti vel y inv olv ed in recruitment
by ad v ertising the trial among their peers . In addition,
the dialog ue and knowledg e platform provided
rec ommendations on strategies to r each out to the tar get
population and maximise participant enga gement in
the trial acti vities. Participants were r egularl y asked to
provide feedback on the int ervention burden and other
issues that mig ht affect their motiv ation to participate
in the trial. Local stud y staff eval uated the information
collect ed and forwarded it t o the coor dinating c entre
in Rome for further ev aluation with assistanc e of
the dialog ue and knowledg e platform. No correcti v e
actions wer e required.
resul ts
Participants
Participants wer e recruit ed from J anuary 2016 to
Nov ember 2017. Randomisation began on 3 F ebruary
2016 and enrolment finished on 15 Nov ember 2017.
The final follow -up visit was on 31 October 2019.
Details on screenin g, r ecruitment strategies, and
characteristics of eli gible participants are report ed
elsewher e. 8 Of 12 358 screened candidat es, 1519
wer e eligible and agreed t o be r andomised: 760 to the
multic omponent interv ention and 759 to the lifestyle
education gr oup. Ov erall, 1205 (79.3%) participants
had an SPPB scor e of 3 to 7 and 314 (20.7%) had an
SPPB scor e of 8 or 9 (fig 1).
Baseline characteristics wer e comparable betw een
interv ention groups within SPPB categ ories (table
1). The mean ag e of the study population was 78.9
(standard de viation 5.8) y ears, 1088/1519 (71.6%)
wer e women, and the a verag e bod y mass index (BMI)
was 28.6 ( SD 5.7). The a v erag e SPPB scor e was 6.7 ( SD
1.0). The mean appendicular lean mass was 21.0 ( SD
3.6) k g in men and 14.6 (SD 2.1) k g in women; mean
BMI adjusted appendicular lean mass w as 0.72 ( SD
0.07) and 0.53 ( SD 0.07), respecti vel y . Osteoarthritis
was report ed by 76.9% (1168/1519) of participants,
h ypertension by 65.9% (1001/1519), and diabetes
by 21.5% (326/1519). Ov erall, 44.6% (678/1519)
reported a fall in the pr evious year . The a v erag e length
of follow -up from randomisation was 26.4 ( SD 9.5)
months.
intervention adherenc e
After e x cludin g medical lea ve and other cir cumstances
that pre v ented participants from e xercisin g (eg, tra vel,
personal problems , transportation issues, national
holida ys), those assigned t o the multic omponent
interv ention on av erage att ended 67.0% ( SD 22.8%)
and 73.5% ( SD 36.5%) of centre based and home based
ph ysical acti vity sessions, respecti vel y . The mean number
of ex cluded centr e based and home based sessions was
47.8 and 63.9, respecti vel y . W alking acti vity , sitting
and l ying time, and standin g acti vity time, captured
throu gh a w earable actimeter , showed participants in
the multic omponent interv ention had a more acti ve
lifestyle than those in the c ontrol group, especiall y those
with SPPB scor es of 3 to 7, d uring the first two y ears of
the trial (supplementary appendix, fi g S1). Differences
in actimetry data between int ervention groups w ere no
long er evident 24 months after randomisation, when
the number of observations w as substantiall y lower .
Ov erall, 78.6% of participants completed full nutritional
assessments, incl uding dietary recor ds ov er three da ys.
Relati ve t o baseline dail y ener gy intak e (23.3 (SD 7.4)
kcal/k g/ da y), values increased by 6.8% at 24 months
(24.1 ( SD 7.1) kcal/kg/ day) and 10.7% at 36 months
(26.1 ( SD 7.5) kcal/kg/ day). A similar patt ern was
observ ed for daily prot ein intake, the v alues of which
increased fr om baseline (0.98 ( SD 0.32) g/ k g/ day) by
10.9% at 24 months (1.10 ( SD 0.32) g/ k g/ day) and by
14.8% at 36 months (1.15 ( SD 0.32) g/ k g/ day).
Participants in the lifestyle ed ucation group
attended on a verag e 65.9% ( SD 26.4%) of scheduled
meeting s, aft er medical lea ve and other circumstanc es
that pre v ented participation had been ex cluded. A
mean of 7.9 meeting s wer e ex cluded.
Primary outc ome
Post -baseline 400 m walk t ests wer e unav ailable for
36/760 (4.7%) participants in the multicomponent
interv ention group and 39/759 (5.1%) in the lifestyle
education gr oup. In participants with an SPPB scor e of
3 to 7, mobility disa bility occurred in 283/605 (46.8%)
in the multic omponent interv ention group (six deaths,
1.0%) and 316/600 (52.7%) in the lifestyle ed ucation
group (se v en deaths, 1.2%) (hazar d ratio 0.78, 95%
confidenc e interv al 0.67 to 0.92; P=0.005) (fi g 2).
Results w ere consistent w hen death was removed fr om
the primary outc ome (0.79, 0.67 to 0.93; P=0.006).
As a qualitati ve int er action between SPPB cat eg ory
and interv ention arm was found when the cumulati ve
ev ent curves for participants with SPPB scor es of 3-7
and a scor e of 8 or 9 wer e compared, w e anal ysed
those with an SPPB scor e of 8 or 9 separatel y . In this
subset , mobility disability oc curred in 46/155 (29.7%)
participants in the multicomponent int ervention
group (thr ee deaths, 1.9%) and 38/159 (23.9%) in the
lifestyle ed ucation group (two deaths, 1.3%) (hazard
ratio 1.25, 95% confidence int erval 0.79 t o 1.95;
P=0.34) (supplementary appendix, fi g S2). Subgr oup
anal yses in participants with an SPPB scor e of 3 to 7
showed that the effects of int erventions on incident
mobility disability w ere comparable across se xes,
copyright. on 18 November 2022 at Biblioteca di Medicina e Chirurgia. Protected by http://www.bmj.com/ BMJ: first published as 10.1136/bmj-2021-068788 on 11 May 2022. Downloaded from
RESEAR CH
6 doi: 10 .1136/bmj- 2021 -0687 88 |
BMJ
20 22;377 :e0687 88 | the bmj
races, a g e groups, hist ory of cardiovascular disease,
history of dia betes , and 4 m gait speed <0.8 m/ s or ≥0.8
m/ s (fig 3). The gait speed cut -point was chosen based
on pre vious finding s, w hich show ed that a walking
speed at usual pace slow er than 0.8 m/ s identifies
older adul ts at risk of ad verse out comes . 2 18
second ary outcome s
T able 2 and table 3 show the r esults for sec ondary
outc omes. In participants with an SPPB scor e of 3 to
7, persistent mobility disa bility occurr ed in 127/605
(21.0%) in the multic omponent interv ention group
(sev en deaths, 1.2%) and 150/600 (25.0%) in the
lifestyle ed ucation group (four deaths, 0.7%) (hazard
ratio 0.79, 95% confidence int erval 0.62 t o 1.01;
P=0.06). The SPPB scor e increased mor e in the
multic omponent interv ention group than lifestyle
education gr oup at both 24 months (least squares mean
differenc e 0.8 points, 95% c onfidence interv al 0.5 to
1.1 points; P<0.001) and 36 months (1.0 point, 0.5 to
Assessed for eligibility
Ineligible*
SPPB out of range
Medical exclusions
Normal appendicular lean mass at DEXA
Refused further screening
Physically disabled or unable to walk 400 m
Too physically active
Plans to relocate
Behavioural exclusions
Nursing home residence
Household member enrolled in SPRINTT
Safety concerns during functional testing
Poor cognition
Age <70 years
Sensory impairments
Other exclusions
2471
1902
1612
1543
1288
1136
285
275
261
222
136
88
64
43
110
Allocated to multicomponent intervention
SPPB score 3-7
SPPB score 8 or 9
Received intervention
Did not receive intervention
Refused
605
155
715
45
45
Eligible
1567
Randomised
1519
760
Allocated to lifestyle education
SPPB score 3-7
SPPB score 8 or 9
Received intervention
Did not receive intervention
Refused
600
159
704
55
759
Did not complete 24 months of follow-up
Participant’s decision
Adverse event
Lost to follow-up
88
9
4
Death
Other
Missing
25
55
1
182
Included in primary outcome analysis
SPPB score 3-7 605 SPPB score 8 or 9 155
760
Included in primary outcome analysis
SPPB score 3-7
600 SPPB score 8 or 9 159
759
Did not complete 24 months of follow-up
Participant’s decision
Adverse event
Lost to follow-up
97
8
2
Death
Other
25
45
177
12 358
10 791
Not randomised
Did not attend randomisation visit 48
48
55
Fig1 | Flow of participants through study . *sum of indiv idual items is higher than number of ineligi ble participants becau se screening w as not
alway s st opped at the fir st u nmet eligibility c riterion. some entries are different from those previous ly pub lished 8 because of data updates after
databa se cleaning. DeXa=dual energy x ray absorptiometry; sPPb=s hort phys ical per formanc e battery; s Printt=sarc openia and Phys ical frailty in
older people: multi-c omponent t reatment s tr ategies
copyright. on 18 November 2022 at Biblioteca di Medicina e Chirurgia. Protected by http://www.bmj.com/ BMJ: first published as 10.1136/bmj-2021-068788 on 11 May 2022. Downloaded from
RESEAR CH
the bmj |
BMJ
20 22;377 :e0687 88 | doi: 10 .11 36/bmj- 2021 -0687 88 7
1.6 points; P<0.001). The decline in hand grip strength
at 24 months was smaller in w omen assigned t o the
multic omponent interv ention than those assigned to
lifestyle ed ucation (0.9 kg , 95% confidenc e interv al
0.1 to 1.6 k g; P=0.028). N o significant betw een
group differ ences w ere observ ed in men. W omen
in the multic omponent interv ention group lost less
appendicular lean mass than w omen in the lifestyle
education gr oup at both time points (24 months 0.24
k g, 0.10 to 0.39 k g; P<0.001; 36 months 0.49 k g, 0.26
to 0.73 k g; P<0.001). N o significant betw een group
differenc es wer e observ ed in men.
In participants with an SPPB score of 8 or 9,
persistent mobility disa bility occurr ed in 16/155
(10.3%) in the multic omponent interv ention group
(one death, 0.6%) and 16/159 (10.1%) in the lifestyle
education gr oup (one death, 0.6%) (hazard ratio
1.14, 95% confidenc e interv al 0.55 to 2.36; P=0.72).
A 0.5 point differenc e in the SPPB scor e in fa vour of
the multic omponent interv ention was observed at 24
months (95% confidenc e interv al 0.1 to 1.0 points;
P=0.027). No si gnificant between gr oup differenc es
wer e observed for hand grip strength at an y time point
in either men or women. A t 36 months, w omen in the
multic omponent interv ention lost less appendicular
lean mass than women in the lifesty le education
group (0.60 k g, 95% confidenc e interval 0.30 to 0.90;
P<0.001).
safety
T able 4 shows the resul ts for safety . In participants
with an SPPB scor e of 3 to 7, 337/605 (55.7%) in the
multic omponent interv ention group and 297/600
(49.5%) in the lifestyle ed ucation group experienced at
least one ad v erse ev ent durin g the trial (risk ratio 1.13,
95% confidenc e interv al 1.01 to 1.25). Serious ad v erse
t able1 | b aseline ch aracteri stics of st udy participants ac cording t o short phy sic al performanc e battery (sPPb) scor e categ or y and grou p alloc ation.
v alues ar e number (per cent ages) unles s stated otherwise
ch aracteri stics
sPPb sc ore 3-7 sPPb sc ore 8 or 9
Multic omponent
intervention
(n=605)
lifestyle educ ation
(n=600) all (n=1205)
Multic omponent
intervention
(n=155)
lifestyle
education
(n=159) all (n=314)
Person al ch aracteri stics
Mean (S D) age (years) 79.3 (5.9) 79.2 (5.8) 79.2 (5.8) 78.3 (5.7) 77.1 (5.4) 77.7 (5.6)
W omen 434 (71.7) 425 (70.8) 859 (71.3) 113 (72.9) 116 (73.0) 229 (72.9)
Ethnicity:
White 535 (88.4) 526 (87.7) 1061 (88.0) 136 (87.7) 138 (86.8) 274 (87.3)
Others 7 (1.2) 8 (1.3) 15 (1.2) 3 (1.9) 2 (1.3) 5 (1.6)
Not a vailabl e 63 (10.4) 66 (11.0) 129 (10.7) 16 (10.3) 19 (11.9) 35 (11.1)
Mean (S D) BMI 28.7 (5.4) 28.7 (5.9) 28.7 (5.7) 28.2 (5.6) 28.3 (6.1) 28.2 (5.9)
Phys ical frailty and sarcopeni a defining criteri a
Mean (S D) S PPB summary score 6.2 (1.1) 6.2 (1.1) 6.2 (1.1) 8.6 (0.5) 8.6 (0.5) 8.6 (0.5)
Mean (S D) appendicul ar lean m ass (kg):
Men 20.94 (3.55) 20.88 (3.52) 20.91 (3.53) 21.18 (3.17) 22.06 (3.99) 21.62 (3.61)
Women 14.61 (2.00) 14.74 (2.20) 14.68 (2.10) 14.54 (1.85) 14.47 (2.02) 14.50 (1.93)
Mean (S D) appendicul ar lean m ass/BMI:
Men 0.72 (0.08) 0.72 (0.07) 0.72 (0.07) 0.74 (0.09) 0.72 (0.07) 0.73 (0.08)
Women 0.52 (0.07) 0.52 (0.07) 0.52 (0.07) 0.54 (0.08) 0.54 (0.08) 0.54 (0.08)
C ognition and physic al performance
Mean (S D) MMSE score 27.9 (1.8) 27.8 (1.8) 27.9 (1.8) 28.1 (1.8) 28.4 (1.8) 28.2 (1.8)
Mean (S D) time to wal k 400 m (min) 8.99 (2.51) 9.06 (2.50) 9.02 (2.51) 7.72 (2.11) 7.27 (1.59) 7.49 (1.87)
Mean (S D) 400 m walk speed (m/s) 0.80 (0.21) 0.79 (0.21) 0.79 (0.21) 0.92 (0.22) 0.95 (0.17) 0.94 (0.20)
Mean (S D) handgrip s trength (kg):
Men 28.3 (8.7) 28.7 (9.6) 28.5 (9.1) 29.3 (9.6) 29.9 (9.9) 29.6 (9.7)
W omen 16.6 (5.5) 17.0 (5.9) 16.8 (5.7) 17.9 (4.9) 16.9 (5.0) 17.4 (4.9)
Clinic al characteri stics
Osteoarthritis 466 (77.0) 463 (77.2) 929 (77.1) 122 (78.7) 117 (73.6) 239 (76.1)
Any c ardiovascul ar medical hist ory 443 (73.2) 423 (70.5) 866 (71.9) 114 (73.5) 100 (62.9) 214 (68.2)
Hypertension 413 (68.3) 392 (65.3) 805 (66.8) 105 (67.7) 91 (57.2) 196 (62.4)
Myoc ardial infarction 46 (7.6) 50 (8.3) 96 (8.0) 16 (10.3) 16 (10.1) 32 (10.2)
Con gestive he ar t failu re 42 (6.9) 45 (7.5) 87 (7.2) 4 (2.6) 9 (5.7) 13 (4.1)
Chr onic lun g disea se 99 (16.4) 88 (14.7) 187 (15.5) 20 (12.9) 26 (16.4) 46 (14.6)
Stroke or br ain haemorrhage 46 (7.6) 41 (6.8) 87 (7.2) 8 (5.2) 6 (3.8) 14 (4.5)
Diabetes mel litus 131 (21.7) 139 (23.2) 270 (22.4) 26 (16.8) 30 (18.9) 56 (17.8)
Canc er (excluding minor sk in cancer) 79 (13.1) 82 (13.7) 161 (13.4) 25 (16.1) 25 (15.7) 50 (15.9)
Falls i n past yea r 284 (46.9) 270 (45.0) 554 (46.0) 62 (40.0) 62 (39.0) 124 (39.5)
Injurious fal ls in pa st year 102 (35.9) 86 (31.9) 188 (33.9) 25 (41.0) 20 (32.3) 45 (36.6)
Previous hip fract ure 35 (5.8) 34 (5.7) 69 (5.7) 12 (7.7) 10 (6.3) 22 (7.0)
Previous non-femoral f ractu re 198 (32.7) 191 (31.8) 389 (32.3) 48 (31.0) 53 (33.3) 101 (32.2)
Emotional, nervous, ps ychiatric problems 130 (21.5) 128 (21.3) 258 (21.4) 39 (25.2) 40 (25.2) 79 (25.2)
At le ast one drug at time of s creening 578 (95.5) 578 (96.3) 1156 (95.9) 147 (94.8) 150 (94.3) 297 (94.6)
≥5 drugs at time of sc reening 358 (59.2) 340 (56.7) 698 (57.9) 76 (49.0) 80 (50.3) 156 (49.7)
BMI=body ma ss inde x; MMSE=mini-ment al st ate examination.
copyright. on 18 November 2022 at Biblioteca di Medicina e Chirurgia. Protected by http://www.bmj.com/ BMJ: first published as 10.1136/bmj-2021-068788 on 11 May 2022. Downloaded from
RESEAR CH
8 doi: 1 0. 1136 /bmj-20 21-068 788 |
BMJ
20 22;377 :e0687 88 | the bmj
ev ents occurred in 237/605 (39.2%) participants in
the multic omponent interv ention group and 216/600
(36.0%) in the lifestyle ed ucation group (1.09, 0.94
to 1.26). F alls were r ecor ded in 80/605 (13.2%)
participants in the multicomponent int ervention group
and 49/600 (8.2%) in the lifestyle ed ucation group
(1.62, 1.16 to 2.27). Deaths oc curred in 31/605 (5.1%)
participants in the multicomponent int ervention group
and 25/600 (4.2%) in the lifestyle ed ucation group
(1.23, 0.74 to 2.06).
In participants with an SPPB score of 8 or 9, 79/155
(51.0%) in the multic omponent interv ention group
and 79/159 (49.7%) in the lifestyle ed ucation group
experienc ed at least one ad verse e v ent durin g the trial
(1.03, 0.82 to 1.28). Serious ad verse ev ents occurred
in 45/155 (29.0%) participants in the multicomponent
interv ention group and 48/159 (30.2%) in the
lifestyle ed ucation group (0.96, 0.68 to 1.35). F alls
wer e recorded in 9/155 (5.8%) participants in the
multic omponent interv ention group and 16/159
(10.1%) in the lifestyle ed ucation group (0.58, 0.26 to
1.27). Deaths occurr ed in 5/155 (3.2%) participants
in the multic omponent interv ention group and 3/159
(1.9%) in the lifestyle ed ucation group (1.71, 0.42 to
7.03).
The proportion of participants who w ere admitted
to hospital or t o the emer g ency department was
comparable betw een intervention gr oups within
SPPB categ ories. Reasons for hospital admission and
emer g ency department or ur g ent care visits w ere
hig hly heter og eneous and were c onsidered unr elated
to stud y procedur es.
discussion
In the SPRINT T trial, an interv ention based on
ph ysical acti vity with technological support and
nutritional counsellin g in participants with ph ysical
frailty and sarc openia and an SPPB scor e of 3 to 7
was associated with a r eduction in the risk of incident
mobility disability d uring 36 months of follow- up,
compar ed with an interv ention comprising lifesty le
education. Participants with an SPPB scor e of 3 to 7
assigned t o the multicomponent interv ention showed
great er improv ements in phy sical performance than
participants assigned t o lifestyle ed ucation. W omen
with an SPPB scor e of 3 to 7 in the mul ticomponent
interv ention group lost less muscle stren gth and
appendicular lean mass than w omen in the lifestyle
education gr oup. In participants with an SPPB scor e of
8 or 9, the multic omponent interv ention did not affect
the risk of dev eloping mobility disability , had mar ginal
effects on ph ysical performance, and, in w omen,
attenuat ed the loss of appendicular lean mass .
c omparison with prev ious studie s
Sev er al in v estig ations hav e tested the impact of lifesty le
interv entions on fr ailty , disability , and other health
outc omes in community d wellin g older adul ts. In L IFE,
a ph ysical acti vity intervention was associat ed with
a red uction in the risk of mobility disability ov er 2.6
y ears of follow -up compar ed with a health education
pro gramme in 1635 older adults with an SPPB sc ore
of ≤9. 17 In participants with an SPPB score of <8 (731,
44.7%), mobility disability de veloped in 38.2% of
those in the ph ysical acti vity intervention gr oup and
46.8% in the contr ol group. In participants with an
SPPB scor e of 8 or 9, mobility disability oc curred in
No at risk
Time (months)
Cumulative incidence of events
0
0.4
0.6
1.0
0.8
0.2
03 6 12 18 24 30 36
605
600
547
537
474
452
413
378
346
299
259
196
106
98
18
13
0
0
28
26
93
103
141
163
195
222
231
270
267
299
281
313
42
Multicomponent intervention
Multicomponent intervention
Lifestyle education
No of events
Multicomponent intervention
Lifestyle education
Lifestyle education
Hazard ratio 0.78 (95% CI 0.67 to 0.92); P=0.005
Fig2 | Kap lan-Meier cu r ves for inc ident mobility di sability in participants with baseline
short phys ical performanc e batter y (sPPb) sc ore of 3-7. the graph i s t runc ated at
36 months, after which two addition al mob ility di sab ility ev ents wer e recor ded in
the multicomponent inter vention grou p and three in the life style educ ation group .
ci=c onfidence int erval
Overall
Sex
Men
Women
Race or ethnicity
White
Other
Age (years)
<80
≥80
History of CVD
No
Yes
History of diabetes
No
Yes
Gait speed (m/s)
<0.8
≥0.8
0.78 (0.67 to 0.92)
0.87 (0.65 to 1.17)
0.75 (0.61 to 0.91)
0.75 (0.63 to 0.90)
0.23 (0.03 to 2.15)
0.75 (0.59 to 0.96)
0.76 (0.61 to 0.95)
0.62 (0.44 to 0.89)
0.84 (0.69 to 1.01)
0.82 (0.68 to 0.99)
0.70 (0.50 to 0.97)
0.79 (0.66 to 0.95)
0.78 (0.54 to 1.12)
0.4 12
Study Hazard ratio
(95% CI)
Hazard ratio
(95% CI)
283/605 (46.8)
88/171 (51.5)
195/434 (44.9)
253/535 (47.3)
1/7 (14.3)
115/318 (36.2)
168/287 (58.5)
54/162 (33.3)
229/443 (51.7)
216/474 (45.6)
67/131 (51.1)
226/450 (50.2)
57/155 (36.8)
Multicomponent
intervention
316/600 (52.7)
97/175 (55.4)
219/425 (51.5)
290/526 (55.1)
4/8 (50.0)
154/336 (45.8)
162/264 (61.4)
80/177 (45.2)
236/423 (55.8)
236/461 (51.2)
80/139 (57.6)
248/444 (55.9)
68/156 (43.6)
Lifestyle
education
0.38
0.30
0.94
0.15
0.40
0.96
Interaction
P value
No with event/No in group (%)
Fig3 | Prespec ified su bgroup an alyses in par ticipants w ith baseline short phys ical
performance b attery (sPPb) sc ore of 3-7. cvD=cardiov ascular di sease; c i=confidence
interval
copyright. on 18 November 2022 at Biblioteca di Medicina e Chirurgia. Protected by http://www.bmj.com/ BMJ: first published as 10.1136/bmj-2021-068788 on 11 May 2022. Downloaded from
RESEAR CH
the bmj |
BMJ
20 22;377 :e0687 88 | doi: 10 .11 36/bmj- 2021 -0687 88 9
23.9% of those in the ph ysical acti vity intervention
group and 25.7% in the c ontrol gr oup. 17 In SPRINTT ,
the proportion of participants with an SPPB scor e of
3 to 7 w ho experienc ed mobility disability was 46.8%
(283/605) in the multic omponent interv ention group
(45.8% ex cluding deaths) and 52.7% (316/600) in the
lifestyle ed ucation group (51.5% ex cludin g deaths). In
those with an SPPB scor e of 8 or 9, incident mobility
disability oc curred in 29.7% of participants (46/155)
in the multic omponent interv ention group and 23.9%
(38/159) in the contr ol group. These findin gs su g gest
that in older adul ts with an SPPB scor e of <8 the presence
of red uced a ppendicular lean mass mig ht identify a
subset of mobility limited ad ults at especially hi gh risk
of disability . This observation mi ght also explain w h y
the effect size of the multic omponent interv ention was
lower than e xpected (22% v 25%). The estimation was
based on the resul ts of LIFE, in w hich onl y a portion
of participants presumabl y had low appendicular
lean mass. 19 In the e xploratory sample of older adul ts
with moderate red uction in ph ysical function, the
primary outc ome was observ ed more frequentl y in
those assigned t o the multicomponent interv ention
than those assigned t o lifestyle education. This finding
is unexpect ed and in contrast with resul ts from LIFE;
owing t o insufficient power and wide c onfidence
interv als, how ev er , no meaningful interpr etations can
be provided.
Participants with an SPPB score of 3 t o 7 assigned to
the multic omponent interv ention had a 2 point higher
scor e at 36 months relati ve t o baseline. The SPPB score
in those in the lifestyle ed ucation group had increased
by 1 point at 36 months. A 0.5 point incr ease in SPPB
scor e was observ ed at 36 months in par ticipants with
an SPPB scor e of 8 or 9, reg ardless of gr oup allocation.
The improv ement experienced by participants with
an SPPB scor e of 3 to 7 equals or e xc eeds clinically
meaningful chan ges of the test (1.0-1.5 points). 20 21
The between gr oup difference in SPPB scor e in fa vour
of the multic omponent interv ention (0.8 points at 24
months and 1.0 point at 36 months) is consist ent with
pre vious studies that test ed lifestyle int erv entions in
frail older people. 22-25
The multic omponent interv ention showed a
positi ve effect on appendicular lean mass in w omen,
irrespecti ve of SPPB cat egory . Studies ha ve shown that
sex infl uences bod y composition chang es in response
to e x ercise in old ag e, with women e xperiencing
great er benefits than men. 26 27 In addition, sex specific
associations between pr otein intake and longitudinal
chang es in appendicular lean mass ha ve been
described in older people. 28
strengths and limitations of thi s st udy
The SPRINT T trial has sev eral strengths. The ph ysical
frailty and sarc openia construct, albeit original,
relies on v alidated t ests and assessments. SPPB is a
compr ehensi ve test that ca ptures limitations in lower
extr emity function. 10 F or its validity , sensitivity to
chang es, repr oducibility , feasibility , and predicti ve
val ue for disability and mortality across heal thcare
setting s, the EMA indicated SPPB as the preferr ed
option to characterise ph ysical frailty for int erv ention
trials in older adul ts. 29 Indeed, chan g es in SPPB scor es
are incr easing ly used as k ey efficacy endpoints in
clinical trials on sarc openia, phy sical frailty , and other
ag e related conditions . 22 24 25 30 31 The presence of low
appendicular lean mass was det ermined accordin g to
the cut -points rec ommended by the Foundation for the
National Institutes of H ealth as the best predictors of
mobility disability . 11 Stud y par ticipants wer e followed
for up to 36 months , confirmin g the feasibility of
identifying , enrollin g, and retainin g fr ail older adul ts
on a lar g e scale. 17 22 32 The stud y sample included a
g eographicall y and culturally heter ogeneous c ohort
of frail older people across Eur ope. The ke y efficacy
endpoints are r eliable, standar dised, and well v alidated
outc omes in older people. 33 34 Dietary plans wer e tailored
to the nutritional needs of indi vidual participants
following e xpert recommendations for standar d practice
in g eriatrics. 16 The ph ysical acti vity routine, which can
be performed at home after a supervised familiarisation
phase, is included in int ernational guidelines for the
manag ement of fr ailty in older people. 35 36 Ret ention and
adherenc e to int erventions w ere hig h and comparable
with other major non-dru g trials in fr ail older adults . 17
22 32 Finall y , the multicomponent int ervention prov ed
to be feasible, safe, and effecti ve in a hi ghly vulnerable
population. The risk of ad v erse ev ents was, how ev er ,
t able2 | secondary outcome s in participants with baseline short phy sic al performance
battery (sPPb) sc ore 3-7 accor ding to group al location
Outcome s
Multic omponent
intervention
(n=605)
lifestyle educ ation
(n=600) effect s ize (95% ci) P value
No (%) of persi stent mobility
dis ability events
127 (21.0) 150 (25.0) 0.79 (0.62 to 1.01)* 0.06
Ch anges from ba seline in phys ical performanc e (SPPB su mmary scor e)
24 months 2.0 (0.1) 1.2 (0.1) 0.8 (0.5 to 1.1) <0.001
36 months 2.0 (0.2) 1.0 (0.2) 1.0 (0.5 to 1.6) <0.001
Ch anges from ba seline in handgrip (muscle) s trength (k g)
Men:
24 months −1.6 (0.5) −1.6 (0.5) 0.0 (−1.4 to 1.5) 0.97
36 months −2.8 (1.4) −4.4 (1.2) 1.6 (−2.2 to 5.4) 0.41
W omen:
24 months −0.3 (0.3) −1.1 (0.3) 0.9 (0.1 to 1.6) 0.028
36 months −0.3 (0.4) −1.3 (0.4) 0.9 (−0.2 to 2.1) 0.11
Ch anges from ba seline in appendicular le an mass
Men:
24 months −0.55 (0.12) −0.83 (0.12) 0.28 (−0.06 to 0.62) 0.11
36 months −0.55 (0.34) −0.78 (0.33) 0.23 (−0.72 to 1.18) 0.65
W omen:
24 months −0.13 (0.05) −0.37 (0.05) 0.24 (0.10 to 0.39) <0.001
36 months −0.19 (0.09) −0.68 (0.08) 0.49 (0.26 to 0.73) <0.001
Ch anges from ba seline in appendicular le an mass/body m ass index
Men:
24 months −0.01 (0.00) −0.01 (0.00) 0.00 (−0.01 to 0.02) 0.48
36 months −0.02 (0.02) −0.01 (0.01) −0.02 (−0.06 to 0.03) 0.53
W omen:
24 months 0.00 (0.00) −0.01 (0.00) 0.01 (0.00 to 0.02) 0.002
36 months 0.00 (0.00) −0.01 (0.00) 0.02 (0.01 to 0.03) 0.003
CI=confid ence interval.
V alues are lea st squared means (standard errors), exc ept for persi stent mobility dis ability.
*Hazard r atio (95% CI). For all other sec ondary outc omes, effect siz e is shown as leas t squared mean differenc e
(95% CI) between multic omponent intervention and lifestyle educ ation.
copyright. on 18 November 2022 at Biblioteca di Medicina e Chirurgia. Protected by http://www.bmj.com/ BMJ: first published as 10.1136/bmj-2021-068788 on 11 May 2022. Downloaded from
