Effect of 12-month intervention with lipid-based nutrient supplement on the physical activity of Malawian toddlers: a randomised, controlled trial

E ec o 12-mon h in e en ion wi h lipid-based nu ien supplemen on he physical ac i i y o Malawian oddle s: a andomised, con olled ial A. Pulakka 1 *, Y. B. Cheung 2,3 , K. Male a 4 , K. G. Dewey 5 , C. Kumwenda 6,7 , J. Bendabenda 4 , U. Asho n 7 and P. Asho n 7,8 1 Depa men o Public Heal h, Uni e si y o Tu ku and Tu ku Uni e si y Hospi al, 20014 Tu ku, Finland 2 Cen e o Quan i a i e Medicine, Duke-Na ional Uni e si y o Singapo e Medical School, Singapo e 169857, Singapo e 3 Depa men o Bios a is ics, Singapo e Clinical Resea ch Ins i u e, Singapo e 138669, Singapo e 4 School o Public Heal h and Family Medicine, Uni e si y o Malawi College o Medicine, Maha ma Gandhi Road, Blan y e, Malawi 5 Depa men o Nu i ion, Uni e si y o Cali o nia, Da is, Da is, CA 95616-5270, USA 6 Depa men o Human Nu i ion and Heal h, Lilongwe Uni e si y o Ag icul u e and Na u al Resou ces, PO Box 219, Lilongwe, Malawi 7 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, 30014 Tampe e, Finland 8 Depa men o Paedia ics, Tampe e Uni e si y Hospi al, 33521 Tampe e, Finland (Submi ed 29 Augus 2016 –Final e ision ecei ed 17 Decembe 2016 –Accep ed 24 Janua y 2017) Abs ac Physical ac i i y is beneficial o child en’s well-being. The e ec o die a y supplemen a ion on child en’s physical ac i i y in ood-insecu e a eas emains li le s udied. We examined he e ec s o a lipid-based nu ien supplemen (LNS) on child en’s objec i ely measu ed physical ac i i y in a andomised, con olled, ou come-assesso -blinded ial. Mo he s o he child en ecei ed one capsule daily o Fe- olic acid (IFA), one capsule con aining eigh een mic onu ien s (MMN) o one 20 g sache o LNS (con aining wen y- wo MMN, p o ein, ca bohyd a es, essen ial a y acids and 494 kJ (118 kcal)) du ing p egnancy and o 6 mon hs he ea e . Child en in he IFA and MMN g oups ecei ed no supplemen a ion, and hese g oups we e collapsed in o a single con ol g oup; child en in he LNS g oup ecei ed 20 g LNS om 6 o 18 mon hs. We measu ed physical ac i i y wi h accele ome e s o e 1 week a 18 mon hs. The main ou come was mean ec o magni ude coun s/15 s. O he 728 child en a he beginning o child in e en ion a 6 mon hs, 570 (78 %) p o ided su ficien da a o analysis. The mean accele ome e coun s o he 190 child en in he LNS g oup and o he 380 child en in he con ol g oup we e 303 (SD 59) and 301 (SD 56), espec i ely (P o di e ence =0·65). LNS, gi en o mo he s du ing p egnancy and 6 mon hs pos pa um and o hei in an s om 6 o 18 mon hs o age, did no inc ease physical ac i i y among 18-mon h-old child en. Key wo ds: Lipid-based nu ien supplemen s: Unde nu i ion: Physical ac i i y: Accele ome e : Sub-Saha an A ica S udies in Ghana (1,2) , Malawi (3,4) , Hai i (5) and Bu kina Faso ha e es ed he g ow h-p omo ing e ficacy o small-quan i y, lipid- based nu ien supplemen s (LNS) as a low-cos solu ion o ackling g ow h- al e ing in young child en. The s udies ha e demons a ed modes (1,2,5) o no e ec s (3,4) o LNS supple- men a ion on child g ow h. A leas h ee possible explana ions o limi ed g ow h esponse o supplemen a ion exis . Fi s , supplemen a ion s a ing a 6 mon hs o age migh come oo la e as child en in esou ce-poo se ings a e commonly a ec ed by in a-u e ine g ow h es ic ion and hus a e bo n s un ed (6) . Second, subclinical in ec ion o inflamma ion may block he g ow h esponse o nu ien supplemen s, sugges ing ha in e en ions combining nu i ion in e en ions wi h p e en ion and managemen o in ec ions migh inc ease he e ec (6–8) . Thi d, ene gy om he supplemen a y ood may be di e ed o o he needs o he child en, such as physical ac i i y (9,10) . Physical ac i i y, which in young child en is exhibi ed commonly in he o m o play (11) , is an impo an de e minan o young child en’s cogni i e and mo o de elopmen (12,13) and a con ibu o o hei heal h in la e li e (14,15) . Some s udies ha e ound ha unde nou ished child en a e less physically ac i e han well-nou ished child en, bu inc ease hei ac i i y le el as hei nu i ional s a us imp o es (16–18) . This is plausible, gi en ha he human body unde goes a se ies o physiological and beha iou al changes, including educ ion o physical ac i i y, as a esponse o lowe ed ene gy in ake (19,20) . In addi ion o ene gy Abb e ia ions: IFA, Fe- olic acid; iLiNS, In e na ional Lipid-Based Nu ien Supplemen s S udy G oup; LNS, lipid-based nu ien supplemen s; MMN, mic onu ien s; MVPA, mode a e- o- igo ous physical ac i i y; VM, ec o magni ude. *Co esponding au ho : A. Pulakka, email: anna.pulakka@u u.fi B i ish Jou nal o Nu i ion (2017), 117, 511–518 doi:10.1017/S0007114517000290 © The Au ho s 2017. This is an Open Access a icle, dis ibu ed unde he e ms o he C ea i e Commons A ibu ion licence (h p://c ea i ecommons.o g/licenses/by/4.0/), which pe mi s un es ic ed e-use, dis ibu ion, and ep oduc ion in any medium, p o ided he o iginal wo k is p ope ly ci ed. h ps:/www.camb idge.o g/co e/ e ms. h ps://doi.o g/10.1017/S0007114517000290 Downloaded om h ps:/www.camb idge.o g/co e. Tampe e Uni e si y and Uni e si y Hospi al, on 12 Jun 2017 a 07:01:59, subjec o he Camb idge Co e e ms o use, a ailable a deficiency, mic onu ien deficiencies, especially Fe deficiency, ha e been shown o esul in educed ac i i y le els (21–24) . Howe e , he e is limi ed in o ma ion on physical ac i i y, measu ed by mode n, objec i e me hods (18,24–26) , o young child en in esou ce-poo se ings, al hough some s udies om high-income coun ies exis (27–29) . Reduced physical ac i i y o unde nou ished child en is o conce n, because i can be one o he media o s leading o poo child de elopmen . This heo y is known as he unc ional iso- la ion hypo hesis: beha iou o unde nou ished child en – ha is, educed ac i i y and explo a ion and inc eased apa hy – esul s in child en ha ing less in e ac ion wi h hei en i onmen , and his in u n may es ic op imal de elopmen (23,30) . Mo eo e , in esponse o he unde nou ished child en’s beha iou , ca egi e s may o e hem less s imula ion, which again hinde s hei de elopmen (16,30) . S udies on he e ec o nu ien supplemen a ion on physical ac i i y, using accele ome e -measu ed coun s as an ou come, in a eas wi h high p e alence o unde nu i ion a e a e (17,31) . The In e na ional Lipid-Based Nu ien Supplemen s S udy G oup (iLiNS) has p e iously epo ed ha small-quan i y LNS did no inc ease he physical ac i i y o he child en in he P e en ion o Linea G ow h Failu e in In an s and Young Child en Wi h Lipid-based Nu ien Supplemen s (iLiNS-DOSE) ial (31) . Howe e , he iLiNS-DOSE in e en ion was pos na al only, om 6 o 18 mon hs o age, hus excluding he e al pe iod, which is c ucial o b ain g ow h and o la e g ow h (6) and de elopmen (32) o child en. The cu en ly epo ed s udy (Supplemen ing Ma e nal and In an Die Wi h High-ene gy, Mic onu ien Fo ified Lipid-based Nu ien Supplemen s; iLiNS-DYAD-M NCT01239693, h ps://clinical ials.go /c 2/ show/NCT01239693) was designed o assess he heal h impac s o small-quan i y LNS gi en o mo he s du ing p egnancy and 6 mon hs pos pa um and o hei in an s om 6 o 18 mon hs o age (33) . Ou hypo hesis was ha child en in he in e en ion g oup would be mo e physically ac i e a he age o 18 mon hs han child en in he con ol g oup. Me hods Se ing and pa icipan s The s udy was conduc ed in a semi-u ban and u al a ea o Mangochi Dis ic , Sou he n Malawi. This a ea has high p e alence o ch onic in an unde nu i ion (34) . The ac i i y sub- s udy was a pa o a la ge ial, iLiNS-DYAD-M, de ails o which ha e been published ea lie (35) . In b ie , iLiNS-DYAD-M was a andomised, single-blind, pa allel-g oup con olled ial es ing he heal h e ec s o supplemen ing ma e nal die du ing p eg- nancy and lac a ion and in an die om 6 o 18 mon hs o age wi h LNS. This s udy was conduc ed acco ding o he guidelines laid down in he Decla a ion o Helsinki, and all p ocedu es in ol ing human subjec s we e app o ed by he College o Medicine Resea ch and E hics Commi ee, Uni e si y o Malawi, and by he E hics Commi ee o Pi kanmaa Hospi al Dis ic , Finland. We pe o med he ial acco ding o Good Clinical P ac ice guidelines. W i en o humb-p in ed in o med consen was ob ained om all subjec s. The ma e nal en olmen ook place h ough he an ena al clinics in Mangochi dis ic hospi al, Malindi hospi al and Lungwena heal h cen e. We included mo he s whose ul a- sound scan confi med p egnancy o ≤20 comple ed ges a ion weeks. De ailed ma e nal inclusion and exclusion c i e ia we e published p e iously (35) . Fo he physical ac i i y sub-s udy, we ec ui ed all pa icipan s who came o he las clinic isi o he main ial a he age o 18 mon hs when hey we e s ill ecei ing he in e en ion. Those child en who had mo ed ou o he s udy a ea o whose gua dian did no gi e consen o he sub-s udy we e excluded. Da a collec ion was conduc ed be ween Janua y 2013 and Ma ch 2014. Randomisa ion and blinding The de ails o he andomisa ion p ocedu e can be ound in an ea lie publica ion (35) . In b ie , a esea che no in ol ed in he da a collec ion c ea ed andomisa ion slips in blocks o nine and sealed he slips in opaque, numbe ed en elopes. Eligible p egnan women we e eques ed o choose one o he op six en elopes in a s ack. The con en s o he en elope indica ed he pa icipan numbe and g oup alloca ion. We used single-masked p ocedu es o he LNS in e en ion; ha is, field wo ke s who deli e ed he supplemen s knew which mo he s we e ecei ing LNS. The esea ch assis an s measu ing ac i i y and an h opome ic ou comes we e kep blinded o he g oup alloca ion un il he end o da a collec ion. The esea che s doing he analyses we e blinded un il he da a we e cleaned and he s a is ical analysis plan published online (www.ilins.o g and online Supplemen a y Ma e ial S1). In e en ions The ial included h ee s udy g oups (35) , which we e collapsed in o wo g oups o he pu pose o his analysis. Child en in he LNS g oup ecei ed 20 g o small-quan i y LNS (20 g LNS) daily om 6 o 18 mon hs o age. Thei mo he s had ecei ed 20 g o LNS du ing p egnancy and lac a ion (20 g LNS-P&L), daily du ing p egnancy and o 6 mon hs he ea e . Bo h he 20 g milk con aining LNS and 20 g LNS-P&L included wen y- wo i amins and mine als, 10 g a (including linolenic acid and α-linolenic acid) and 2·6 g p o ein (33,36) . The main ing edien s o he LNS we e peanu s, ege able oil, milk powde , and a i amin and mine al mix. The 20 g daily a ion o LNS would p o ide he RDA o mos mic onu ien s o a heal hy, b eas - eeding in an (36) . LNS was p oduced and packed in indi idual 20 g oil sache s by Nu ise S.A.S. Child en in he con ol g oup did no ecei e supplemen a- ion, bu hei mo he s had ecei ed ei he (a) one capsule daily o Fe- olic acid un il deli e y (60 mg Fe + 400 µg olic acid) and one daily able o Ca (200 mg), akin o placebo, om deli e y o 6 mon hs pos pa um (o iginal Fe- olic acid (IFA) g oup) o (b) one able o mul iple mic onu ien s (con aining Fe- olic acid and six een addi ional mic onu ien s (36) ) daily h ough p egnancy and 6 mon hs pos pa um (o iginal mic onu ien s (MMN) g oup). Because o he absence o di e ence in child ou comes be ween he o iginal IFA and MMN g oups in ou p e ious analyses (33,35) , and because child en in hese g oups 512 A. Pulakka e al. h ps:/www.camb idge.o g/co e/ e ms. h ps://doi.o g/10.1017/S0007114517000290 Downloaded om h ps:/www.camb idge.o g/co e. Tampe e Uni e si y and Uni e si y Hospi al, on 12 Jun 2017 a 07:01:59, subjec o he Camb idge Co e e ms o use, a ailable a ecei ed no in e en ion om 6 o 18 mon hs, we collapsed he IFA and MMN g oups in o a single con ol g oup. Follow up Da a collec o s made weekly home isi s o collec in o ma ion on supplemen use and o nigh ly isi s o deli e he supple- men s. Gua dians we e ad ised o di ide he daily LNS a ion in o wo equal p opo ions, mixed wi h po idge. As a esul o a new quali y assu ance p ocedu e in he supplemen p oduc- ion, he e was a b ie in e up ion in LNS deli e y du ing he ial implemen a ion. Because o his episode, 121 child en missed ecei ing LNS o a pe iod ha anged om 1 o 41 d be ween 1 Augus and 11 Sep embe 2012. Fu he de ails on his episode we e published ea lie (35) . Apa om mala ia, which was ea ed wi h lume an ine/a eme he a he s udy clinic, he pa icipan s’medical condi ions we e ea ed in Malawi’s na ional heal h sys em, wi h he s udy eam eimbu sing he pa icipan s o all medical cos s. Measu emen o ou comes Physical ac i i y was measu ed o e 1 week wi h he Ac iG aph GT3X+ (Ac iG aph LLC), a small accele ome e ha eco ds accele a ions in h ee di e en axes: e ical, an e o-pos e io and medio-la e al (26,37) . While a he clinic, esea ch assis an s ins uc ed he gua dians o secu e he accele ome e on he child’s igh hip using an elas ic bel and o allow he child o wea he de ice con inuously h oughou he day and nigh and o emo e i only i he child showed signs o discom o . Measu emen o co a ia es A he ma e nal en olmen isi , ained an h opome is s mea- su ed he mo he s’weigh wi h digi al scales (SECA 874 fla scale; Seca GmbH & Co.) and heigh wi h s adiome e s (Ha penden s adiome e ; Hol ain Limi ed). Resea ch assis an s ob ained ma e nal age and yea s o schooling by in e iewing mo he s a he en olmen isi . Household ood insecu i y was assessed wi h he Household Food Insecu i y Access Scale (HFIAS) (38) . On he basis o he leng h o ollow up and he numbe o supplemen doses deli e ed home and e u ned unused, he mean adhe ence (p opo ion o days when he child en consumed LNS supplemen s) was 77 % (33) . Child en’s da e o bi h was e ified by he esea ch assis an s who isi ed he child soon a e bi h (35) . A he clinic isi s a 6 and 18 mon hs o age, esea ch an h opome is s measu ed pa icipan s’weigh in iplica e o he nea es 20 g using an elec onic in an -weighing scale (SECA 381 baby scale; Seca GmbH & Co) and leng h o he nea es 1 mm using a high- quali y leng h boa d (Ha penden In an ome e ; Hol ain Limi ed). We calcula ed leng h- o -age z-sco e (LAZ) and weigh - o -leng h z-sco e (WLZ) using he WHO 2006 Child G ow h S anda ds (39) . Resea ch assis an s obse ed he child en o assess hei abili y o walk a 18 mon hs o age. Gua dians we e in e - iewed on how much he child en we e being ca ied by o he s, and child en who we e epo edly being ca ied 1 h o mo e e e y day o almos e e y day we e classified as ca ied. Da a p ocessing and analyses We analysed da a using S a a/IC so wa e, e sion 12.1 (S a a- Co p.). We se he le el o s a is ical significance a 0·05 o all analyses. All he analyses we e based on he p inciple o modified in en ion o ea ; ha is, we included all pa icipan s andomly assigned in he analyses, wi h he excep ion ha wo pa icipan s whose g oup alloca ions we e inco ec ly an- sc ibed and assigned du ing en olmen we e included in he g oup co esponding o he ac ual in e en ion hey ecei ed h oughou he ial. We conside ed physical ac i i y da a o be missing i he ac ual onse o measu emen was o e 30 d om he planned da e. The e we e wo easons o his choice: child en we e no longe ecei ing in e en ion a his s age and imp o ed mo o skills a olde age could esul in accele ome e eadings di e en om hose o child en a 18 mon hs o age. The da a educ ion was done simila ly o ou ea lie s udy (31) : we excluded he fi s and he las days o measu emen as incomple e days, nigh ime be ween 20.00 and 05.00 hou s, and s ings o ≥20 min o ze oes (11) . Pa icipan s wi h ≥4d (27) wi h ≥6h (40) o accele ome e da a we e included in he ana- lyses. We se epoch leng h a 15 s (11,41,42) . We used daily mean ec o magni ude (VM) coun s/15 s as he main ou come. The VM coun s we e calcula ed by aking he squa e oo o he sum o squa ed ac i i y coun s o each o he h ee axes. Seconda y ou comes included mean e ical axis coun s/15 s, pe cen age o ime in mode a e- o- igo ous physical ac i i y (MVPA), pe cen age o ime being seden a y and pe cen age o ac i e child en. We calcula ed mean VM and e ical axis coun s/15 s by a e aging mean coun s/15 s o each day o e all alid days o each o he pa icipan s. The pe cen age o ime spen in MVPA was defined using alida ed cu o poin s o e ical axis ac i i y coun s ≥419 coun s/15 s (25) and pe cen age o ime being seden a y as e ical axis ac i i y coun s ≤48 coun s/15 s (25) . The p opo ion o ac i e child en was calcula ed acco ding o he guidelines o he US Na ional Associa ion o Spo s and Physical Educa ion as hose child en whose mean ime in MVPA o e all alid days was ≥90 min/d (43) . We used Fishe ’s exac es o es o di e ences in he a e o loss o ollow up be ween g oups. We es ed he hypo hesis ha physical ac i i y o in an s in he in e en ion g oup would be g ea e han ha o in an s in he con ol g oup o each ac i i y ou come using S uden ’s es , and he hypo hesis ha a g ea e p opo ion o child en in he in e en ion g oup would each 90 min o MVPA/d wi h a log-binomial eg ession model. We also d ew ke nel densi y plo s o each o he ou comes. As a seconda y analysis o assessing he mean VM coun s in he in e en ion and con ol g oups, we buil a eg ession model adjus ing o eigh p e-specified a iables (LAZ a 6 mon hs, WLZ a 6 mon hs, sex, season o ac i i y measu emen , bi h o de , ma e nal educa ion, ma e nal age and HFIAS sco e) and o child ca ying (ca ied . no ). We also pe o med a sensi- i i y analysis by es ing he di e ences in mean VM coun s be ween he h ee o iginal g oups, LNS, IFA and MMN, wi h ANOVA. The sample size was o iginally calcula ed in acco dance wi h he main objec i e o he ial: 288/g oup o de ec an e ec size Nu ien supplemen s and physical ac i i y 513 h ps:/www.camb idge.o g/co e/ e ms. h ps://doi.o g/10.1017/S0007114517000290 Downloaded om h ps:/www.camb idge.o g/co e. Tampe e Uni e si y and Uni e si y Hospi al, on 12 Jun 2017 a 07:01:59, subjec o he Camb idge Co e e ms o use, a ailable a o 0·3 o LNS on child leng h. The sample size o abou 190 LNS and 380 con ol pa icipan s o his sub-s udy o e ed abou 80 % powe o de ec an e ec size o 0·25 SD in con inuous ou comes a 5 % wo-sided ype I e o a e. Resul s Fig. 1 p esen s he flow o he s udy pa icipan s and hei mo he s. O iginally, 869 p egnan mo he s we e en olled in o his s udy. The 786 mo he s who we e s ill in he s udy a he ime o deli e y ga e bi h o 790 in an s. O he 661 child en who we e measu ed o ac i i y a 18 mon hs o age, 570 had su ficien da a and we e included in he ac i i y analysis (78 % o he 728 child en who we e in ollow up a he beginning o child in e en ion a 6 mon hs o age). Backg ound cha ac e is ics o he child en we e simila in he in e en ion and con ol g oups a 6 mon hs o age (Table 1). O he child en who we e in ollow up a 6 mon hs o age, 21 % in he LNS g oup and 22 % in he con ol g oup (P=0·85) we e ei he los o ollow up o had insu ficien accele ome e da a. Child en who we e included in he analysis we e gene ally simila in hei mean backg ound cha ac e is ics compa ed wi h he 158 child en who we e a ailable a 6 mon hs o age bu we e no included in he analysis, excep ha hei mo he s had ewe yea s o schooling. The included pa icipan s had a mean o 5·9( SD 1·5, ange 4–10) d o ac i i y measu emen wi h a median measu emen ime o 13·5 h/d (in e qua ile ange 12·1–14·3 h). The main ou come –mean VM accele ome e coun s/15s – was 302 (SD 57) o all he included pa icipan s. The mean VM coun s/15 s was 303 (SD 59) o child en in he LNS g oup and 301 (SD 56) o child en in he con ol g oup (P=0·65) (Fig. 2 and Table 2). The e was no s a is ically significan di - e ence in he seconda y ou comes be ween he g oups (Table 2 and online Supplemen a y Fig. S1). In he LNS g oup, 38·4 % o he child en eached he ecommenda ion o 90 min o MVPA/d, whe eas he co esponding figu e o he con ol g oup was 35·8 ( isk a io 0·93; 95 % CI 0·74, 1·17) %. The esul s om he eg ession analysis also showed no associa ion be ween in e en ion and physical ac i i y (β-coe ficien 4·3; 95 % CI 15, 6; P=0·41). In he sensi i i y analysis, compa ing he h ee o iginal g oups, he mean VM coun s/15 s was 303 (SD 59) in he LNS g oup, 306 (SD 62) in he MMN g oup and 297 (SD 49) in he IFA g oup (P=0·28). Discussion The pu pose o his sub-s udy o he andomised, con olled iLiNS-DYAD ial was o measu e he e ec o die a y supple- men a ion wi h LNS, gi en o mo he s du ing p egnancy and 6 mon hs pos pa um and o hei in an s om 6 o 18 mon hs o age, on child en’s physical ac i i y in semi- u al Malawi. The mean accele ome e coun s we e sligh ly highe among he supplemen ed han among he non-supplemen ed pa icipan s in ou sample, bu he di e ences we e small and s a is ically insignifican . The esul s o his s udy a e e y simila o ou p e ious findings whe e LNS did no inc ease he physical ac i i y o Malawian child en when gi en in 10–40 g quan i ies/d om 6 o 18 mon hs o age (31) . In he cu en s udy, gi ing LNS o mo he s du ing p egnancy and lac a ion did no seem o o e addi ional benefi s in e ms o physical ac i i y o he child en. We a e no awa e o any o he s udies measu ing he e ec s o LNS sup- plemen a ion on physical ac i i y; howe e , he e a e p e ious s udies examining physical ac i i y a e supplemen a ion wi h mul iple mic onu ien s (12,44–46) . These s udies ha e demon- s a ed a ying esul s. Se e al o hem epo ed posi i e e ec s on physical ac i i y om supplemen a ion wi h ei he Zn (47,48) , mul iple mic onu ien s (46) o milk and mul iple mic onu ien s in child en wi h Fe-deficiency anaemia (45) . One o he s udies ound no e ec (44) and one epo ed mixed esul s (12) o a milk- based in e en ion. All o he s udies used obse a ion as he me hod o measu ing ac i i y. The e a e se e al heo e ically plausible explana ions o why he LNS in e en ion had no significan e ec on child en’s physical ac i i y. One limi ing ac o could be ela i ely low (77 %) adhe ence o consuming LNS. The ene gy p o ided by 869 en olled mo he s in he comple e ollow up 288 women in he LNS g oup 581 women in he con ol g oup 241 child en in he LNS g oup a 6 mon hs 487 child en in he con ol g oup a 6 mon hs –Dea hs: 2 – Los o ollow up: 6 –Re used om sub- s udy: 6 –Measu ed >30 d a e planned: 5 –Dea hs: 11 –Los o ollow up : 11 –Re used om sub-s udy: 13 –Measu ed >30 d a e planned: 13 222 child en en olled o he ac i i y sub-s udy 439 child en en olled o he ac i i y sub-s udy –Da a los / unusable: 5 –Insu icien da a: 27 –Da a los / unusable: 11 –Insu icien da a: 48 Numbe analysed: 190 Numbe analysed: 380 – Abo ions o s illbi hs: 11 –Los o ollow up: 19 – – Abo ions o s illbi hs: 11 Los o ollow up: 46 264 li e-bo n in an s (including 3 win pai s) 526 li e-bo n in an s (including 1 win pai ) –Dea hs: 9 –Los o ollow up: 14 –Dea hs: 25 –Los o ollow up: 14 Fig. 1. Pa icipan low. LNS, lipid-based nu ien supplemen s. 514 A. Pulakka e al. h ps:/www.camb idge.o g/co e/ e ms. h ps://doi.o g/10.1017/S0007114517000290 Downloaded om h ps:/www.camb idge.o g/co e. Tampe e Uni e si y and Uni e si y Hospi al, on 12 Jun 2017 a 07:01:59, subjec o he Camb idge Co e e ms o use, a ailable a LNS migh eplace a he han be addi i e o ene gy om b eas milk o complemen a y oods. Howe e , ou p e ious s udies sugges ha p o ision o LNS does no educe b eas milk o complemen a y ood in ake a he age o 9–10 mon hs (49,50) . Acco ding o he accele ome e measu emen , child en (e en in he con ol g oup) we e a leas as ac i e as has been epo ed o oddle s in high-income coun ies (28,51) , sugges ing ha hey we e al eady as ac i e as migh be expec ed a his age despi e esiding in a ood-insecu e a ea. The e is conside able he e o- genei y in he esul s o s udies es ing LNS o imp o ing g ow h o ea ing mode a e malnu i ion (52) , implying ha he e could be unde lying, con ex -specific mechanisms a ec ing child en’s heal h and well-being and limi ing esponse o nu i ion in e en ions. The e o e, i is possible ha nu i ion in e en ions alone a e no su ficien , bu in e en ions ha aim o imp o e child en’s well-being also need o a ge o he isk ac o s, such as in ec ions (7,53,54) . The main s eng hs o his s udy include he ollowing: an- domisa ion, which con ols o con ounde s and dec eases selec ion bias, objec i e measu emen o physical ac i i y, which elimina es epo ing bias inhe en o subjec i e physical ac i i y measu emen me hods; and b oad inclusion c i e ia, which inc ease he gene alisabili y o esul s. The sample size was la ge enough o de ec an e ec size o 0·25 SD wi h abou 80 % powe . The main weakness o he s udy was ha he gua dians could no be blinded o he ea men alloca ion o hei child en. Howe e , because we used objec i e measu emen o he ou come, his is unlikely o ha e a ec ed he physical ac i i y measu emen . The weaknesses o accele ome e s in measu ing physical ac i i y include he ac ha hey do no p o ide in o ma ion on ac i i y ype o con ex , no do hey dis inguish be ween he causes o mo emen , which may lead o o e - es ima ion o ac i i y in in an s who a e equen ly being ca ied by hei ca egi e s (11,18,55) . Howe e , o e es ima ion o ac i i y while he child is being ca ied appea s o be mo e common among young in an s who canno walk (18) , whe eas only 17 % o he 18-mon h-old child en in bo h he con ol and in e en- ion g oups in ou s udy we e epo edly being ca ied ≥1h daily o almos daily. Fu he mo e, we did no pe o m pe - p o ocol analysis – ha is, including only he mos adhe en child en –because we did no ha e a placebo con ol and we e hus unable o iden i y and exclude non-adhe en child en in he con ol g oup. Al hough loss o ollow up was abou 22 %, Table 1. Backg ound cha ac e is ics o pa icipan s and hei mo he s (Mean alues and s anda d de ia ions; numbe s and pe cen ages) Con ol LNS Va iables Mean SD Mean SD Si ua ion a baseline, ma e nal en olmen Numbe o mo he s 581 288 Ma e nal age (yea s) 25 6 25 6 Ma e nal educa ion, comple ed yea s o schooling 3·94 4·04 Ma e nal BMI (kg/m 2 )22·02·622·23·1 Se e ely ood insecu e households n208 104 %3838 Si ua ion a 6 mon hs, beginning o child in e en ion Numbe o child en 487 241 Males n226 117 %4649 Leng h- o -age z-sco e −1·25 1·2−1·26 1·1 Weigh - o -leng h z-sco e 0·39 1·20·36 1·1 Si ua ion a 18 mon hs, en olmen o physical ac i i y measu emen Numbe o child en 439 222 Age (mon hs) 18·10·618·10·1 Season o ac i i y measu emen (%) I: 28 I: 26 II: 20 II: 23 III: 28 III: 28 IV: 25 IV: 23 Walking unassis ed a 18 mon hs (%) 95 97 Child en who we e ca ied (%)* 17 17 LNS, lipid-based nu ien supplemen s. * Ca ied 1 h o mo e e e y day o almos e e y day by ma e nal epo . 0 0 0.002 0.004 0.006 0.008 100 200 300 400 500 Vec o magni ude Fig. 2. Ke nel densi y plo s o mean ec o magni ude coun s/15 s by g oups. , Lipid-based nu ien supplemen s; , con ol. Nu ien supplemen s and physical ac i i y 515 h ps:/www.camb idge.o g/co e/ e ms. h ps://doi.o g/10.1017/S0007114517000290 Downloaded om h ps:/www.camb idge.o g/co e. Tampe e Uni e si y and Uni e si y Hospi al, on 12 Jun 2017 a 07:01:59, subjec o he Camb idge Co e e ms o use, a ailable a di e ences in backg ound cha ac e is ics be ween pa icipan s who we e en olled o no en olled in o his sub-s udy we e small, sugges ing ha a i ion bias was no likely. Taking in o accoun hese sho comings, we belie e ha he main esul s a e alid and su ficien ly ep esen a i e o he a ge popula ion, sugges ing ha LNS supplemen a ion does no ma kedly inc ease he mean physical ac i i y o young child en in he defined a ge g oup. In conclusion, ou esul s do no suppo he hypo hesis ha LNS, gi en o mo he s du ing p egnancy and 6 mon hs pos - pa um and o hei in an s om 6 o 18 mon hs o age, inc eases physical ac i i y among 18-mon h-old child en in semi- u al Malawi. Fu he esea ch is needed o iden i y in e - en ions ha could imp o e he heal h and well-being o chil- d en who a e li ing in ood-insecu e a eas. Gi en he bidi ec ional ela ionship be ween nu i ion and in ec ions, whe eby poo nu i ion inc eases he isk o in ec ions and in ec ions can wo sen nu i ional s a us, hese in e en ions migh need o a ge bo h nu i ion and in ec ions. Fu he mo e, cu en ecommenda ions o child en’s physical ac i i y a e mainly based on pa en al- epo ed physical ac i i y, a measu e ha is known o be biased. I would hus be impo an o es ablish ecommenda ions o su ficien physical ac i i y o suppo he heal h and de elopmen o child en based on esea ch using objec i e physical ac i i y measu es, such as accele ome e s. Acknowledgemen s The au ho s hank all he pa icipa ing amilies and he da a collec ion eams o hei pa icipa ion and con ibu ion o he s udy. The accele ome e da a we e collec ed by eams lead by Ma in Ndelemani. This publica ion is unded in pa by a g an o he Uni e si y o Cali o nia, Da is, om he Bill & Melinda Ga es Founda ion, wi h addi ional unding om he O fice o Heal h, In ec ious Diseases, and Nu i ion, Bu eau o Global Heal h, US Agency o In e na ional De elopmen (USAID) unde e ms o coop- e a i e ag eemen AID-OAA-A-12-00005, h ough he Food and Nu i ion Technical Assis ance III P ojec , managed by FHI 360. Fo da a managemen and s a is ical analysis, he eam ecei ed addi ional suppo om he Academy o Finland h ough g an no. 252075 and om he Medical Resea ch Fund o Tampe e Uni e si y Hospi al h ough g an no. 9M004. Y. B. C. was suppo ed by he Singapo e Minis y o Heal h’s Na ional Medical Resea ch Council unde i s Clinician Scien is Awa d. Lo a Hallamaa p o ided s a is ical suppo . The findings and conclusions con ained wi hin he a icle a e hose o he au ho s and do no necessa ily eflec posi ions o policies o he Bill & Melinda Ga es Founda ion, USAID, he US go e nmen , o he o he unde s. The unde s had no ole in s udy design, da a collec ion and analysis, decision o publish o p epa a ion o he manusc ip . The au ho s’con ibu ions we e as ollows: A. P., Y. B. C., K. M., K. G. D. and P. A. we e esponsible o he s udy concep and design. A. P., C. K., J. B., U. A and P. A. conduc ed he da a collec ion. A. P., Y. B. C. and P. A. designed he analysis. A. P. conduc ed he da a analysis, w o e he fi s d a o he manu- sc ip and had p ima y esponsibili y o he final con en . All he au ho s c i ically e iewed and app o ed he final manusc ip . The au ho s decla e ha he e a e no conflic s o in e es . Supplemen a y ma e ial Fo supplemen a y ma e ial/s e e ed o in his a icle, please isi h ps://doi.o g/doi:10.1017/S0007114517000290 Re e ences 1. Adu-A a wuah S, La ey A, B own KH, e al.(2007)Rando- mized compa ison o 3 ypes o mic onu ien supplemen s o home o ifica ion o complemen a y oods in Ghana: e ec s on g ow h and mo o de elopmen . Am J Clin Nu 86,412–420. 2. Adu-A a wuah S, La ey A, Ok onipa H, e al. (2016) Small- quan i y lipid-based nu ien supplemen s p o ided o women du ing p egnancy and 6 mo pos pa um, and o hei in an s om 6 mo o age, inc eases he mean a ained leng h o 18-mon h-old child en in semi-u ban Ghana: a andomized con olled ial. Am J Clin Nu 104, 797–808. 3. Male a KM, Phuka J, Alho L, e al. (2015) P o ision o 10-40 g/d lipid-based nu ien supplemen s om 6 o 18 mon hs o age does no p e en linea g ow h al e ing in Malawi. J Nu 145, 1909–1915. 4. Mangani C, Male a K, Phuka J, e al. (2015) E ec o com- plemen a y eeding wi h lipid-based nu ien supplemen s and co n-soy blend on he incidence o s un ing and linea g ow h among 6- o 18-mon h-old in an s and child en in u al Malawi. Ma e n Child Nu 11, Suppl. 4, 132–143. 5. Ianno i LL, Dulience SJL, G een J, e al. (2014) Linea g ow h inc eased in young child en in an u ban slum o Hai i: a andomized con olled ial o a lipid-based nu ien supple- men . Am J Clin Nu 99, 198–208. Table 2. Physical ac i i y a 18 mon hs o age (Mean alues and s anda d de ia ions; di e ence in mean alues and 95 % con idence in e als) T ial g oup Con ol (n380) LNS (n190) Compa ison be ween he g oups Va iables Mean SD Mean SD Mean 95 % CI P Vec o magni ude (accele ome e coun s/15 s) 301 56 303 59 2 −8, 12 0·65 Ve ical axis (accele ome e coun s/15 s) 140 32 144 36 3 −3, 9 0·27 Time in MVPA, by e ical axis (%) 11·13·511·44·00·3−0·4, 0·90·38 Time seden a y, by e ical axis (%) 53·48·252·38·6−1·1−2·5, 0·30·13 LNS, lipid-based nu ien supplemen s; MVPA, mode a e- o- igo ous physical ac i i y. 516 A. Pulakka e al. h ps:/www.camb idge.o g/co e/ e ms. h ps://doi.o g/10.1017/S0007114517000290 Downloaded om h ps:/www.camb idge.o g/co e. Tampe e Uni e si y and Uni e si y Hospi al, on 12 Jun 2017 a 07:01:59, subjec o he Camb idge Co e e ms o use, a ailable a 6. de Onis M, Dewey KG, Bo ghi E, e al. (2013) The Wo ld Heal h O ganiza ion’s global a ge o educing childhood s un ing by 2025: a ionale and p oposed ac ions. Ma e n Child Nu 9, Suppl. 2, 6–26. 7. Dewey KG & Maye s DR (2011) Ea ly child g ow h: how do nu i ion and in ec ion in e ac ? Ma e n Child Nu 7, 129–142. 8. Hess SY, Abbeddou S, Jimenez EY, e al. (2015) Small-quan i y lipid-based nu ien supplemen s, ega dless o hei zinc con en , inc ease g ow h and educe he p e alence o s un ing and was ing in young bu kinabe child en: a clus e - andomized ial. PLOS ONE 10, e0122242. 9. Becke C, Du nin JV, Ai chison TC, e al. (2000) E ec s o an ene gy and mic onu ien supplemen on an h opome y in unde nou ished child en in Indonesia. Eu J Clin Nu 54, S52–S59. 10. Mana y MJ, Ndkeha MJ, Asho n P, e al. (2004) Home based he apy o se e e malnu i ion wi h eady- o-use ood. A ch Dis Child 89, 557–561. 11. Cli DP, Reilly JJ & Okely AD (2009) Me hodological conside a ions in using accele ome e s o assess habi ual physical ac i i y in child en aged 0–5 yea s. J Sci Med Spo 12, 557–567. 12. Jaha i AB, Saco-Polli C, Husaini MA, e al. (2000) E ec s o an ene gy and mic onu ien supplemen on mo o de elopmen and mo o ac i i y in unde nou ished child en in Indonesia. Eu J Clin Nu 54,60–68. 13. Timmons BW, Naylo PJ & P ei e KA (2007) Physical ac i i y o p eschool child en –how much and how? Can J Public Heal h 98, Suppl. 2, 122–134. 14. Timmons BW, Leblanc AG, Ca son V, e al. (2012) Sys ema ic e iew o physical ac i i y and heal h in he ea ly yea s (aged 0–4 yea s). Appl Physiol Nu Me ab 37, 773–792. 15. Ekelund U, Luan J, She a L, e al. (2012) Mode a e o igo ous physical ac i i y and seden a y ime and ca diome abolic isk ac o s in child en and adolescen s. JAMA 307, 704–712. 16. G an ham-McG ego S & Bake -Henningham H (2005) Re iew o he e idence linking p o ein and ene gy o men al de elopmen . Public Heal h Nu 8, 1191–1201. 17. Fau hol -Jepsen D, Hansen KB, Van Hees VT, e al. (2014) Child en ea ed o se e e acu e malnu i ion expe ience a apid inc ease in physical ac i i y a ew days a e admission. J Pedia 164, 1421–1424. 18. Wo obey J (2014) Physical ac i i y in in ancy: de elopmen al aspec s, measu emen , and impo ance. Am J Clin Nu 99, 729S–733S. 19. Wa e low JC (1990) Ene gy-spa ing mechanisms: educ ions in body mass, BMR and ac i i y: hei ela i e impo ance and p io i y in unde nou ished in an s and child en. In: Ac i i y, Ene gy Expendi u e and Ene gy. Requi emen s o In an s and Child en, pp. 239–251 [B Schü ch and NS Sc imshaw, edi o s]. Lausanne: In e na ional Die a y Ene gy Consul ancy G oup. 20. She y PS (1999) Adap a ion o low ene gy in akes: he esponses and limi s o low in akes in in an s, child en and adul s. Eu J Clin Nu 53,14–33. 21. Abu o NJ, Rami ez-Zea M, Neu eld LM, e al. (2009) Some indica o s o nu i ional s a us a e associa ed wi h ac i i y and explo a ion in in an s a isk o i amin and mine al deficiencies. J Nu 139, 1751–1757. 22. Olney DK, Polli E, Ka ige PK, e al. (2007) Young Zanziba i child en wi h i on deficiency, i on deficiency anemia, s un ing, o mala ia ha e lowe mo o ac i i y sco es and spend less ime in locomo ion. J Nu 137, 2756–2762. 23. Lozo B, Klein NK, Nelson EC, e al. (1998) Beha io o in an s wi h i on-deficiency anemia. Child De 69,24–36. 24. Yaméogo CW, Cichon B, Fabiansen C, e al. (2017) Co ela es o physical ac i i y among young child en wi h mode a e acu e malnu i ion. J Pedia 181, 235–241. 25. T os SG, Fees BS, Haa SJ, e al. (2012) Iden ifica ion and alidi y o accele ome e cu -poin s o oddle s. Obesi y (Sil e Sp ing) 20, 2317–2319. 26. Pulakka A, Cheung Y, Asho n U, e al. (2013) Feasibili y and alidi y o he Ac iG aph GT3X accele ome e in measu ing physical ac i i y o Malawian oddle s. Ac a Paedia 102, 1192–1198. 27. Hna iuk J, Ridge s ND, Salmon J, e al. (2012) Physical ac i i y le els and pa e ns o 19-mon h-old child en. Med Sci Spo s Exe c 44, 1715–1720. 28. Wij zes AI, Kooijman MN, Kie e-de Jong JC, e al. (2013) Co ela es o physical ac i i y in 2-yea -old oddle s: he gene a ion R s udy. J Pedia 163, 791–799. 29. Johansson E, Hags ome M, S ensson V, e al. (2015) Objec i ely measu ed physical ac i i y in wo-yea -old child en –le els, pa e ns and co ela es. In J Beha Nu Phys Ac 12,3. 30. Ga dne JM & G an ham-McG ego SM (1994) Physical ac i i y, unde nu i ion and child de elopmen . P oc Nu Soc 53, 241–248. 31. Pulakka A, Asho n U, Cheung YB, e al. (2015) E ec o 12-mon h in e en ion wi h lipid-based nu ien supplemen s on physical ac i i y o 18-mon h-old Malawian child en: a andomised, con olled ial. Eu J Clin Nu 69, 173–178. 32. Black MM, Walke SP, Fe nald LCH, e al. (2017) Ea ly child- hood de elopmen coming o age: science h ough he li e cou se. Lance 389,77–90. 33. Asho n P, Alho L, Asho n U, e al. (2015) Supplemen a ion o ma e nal die s du ing p egnancy and o 6 mon hs pos pa um and in an die s he ea e wi h small-quan i y lipid-based nu ien supplemen s does no p omo e child g ow h by 18 mon hs o age in u al Malawi: a andomized con olled ial. J Nu 145, 1345–1353. 34. Na ional S a is ical O fice Malawi and ICF Mac o (2011) Malawi Demog aphic and Heal h Su ey 2010. Zomba and Cal e on, MD: Na ional S a is ical O fice and ICF Mac o. 35. Asho n P, Alho L, Asho n U, e al. (2015) The impac o lipid- based nu ien supplemen p o ision o p egnan women on newbo n size in u al Malawi: a andomized con olled ial. Am J Clin Nu 101, 387–397. 36. A imond M, Zeilani M, Jungjohann S, e al. (2015) Con- side a ions in de eloping lipid-based nu ien supplemen s o p e en ion o unde nu i ion: expe ience om he In e na- ional Lipid-Based Nu ien Supplemen s (iLiNS) P ojec . Ma e n Child Nu 11,31–61. 37. John D & F eedson P (2012) Ac iG aph and Ac ical physical ac i i y moni o s: a peek unde he hood. Med Sci Spo s Exe c 44, S86–S89. 38. Coa es J, Swindale A & Bilinsky P (2007) Household Food Insecu i y Access Scale (HIFAS) o Measu emen o House- hold Food Access: Indica o Guide (V.3). Washing on, DC: Food and Nu i ion Technical Assis ance P ojec , FHI 360. 39. Wo ld Heal h O ganiza ion (2006) WHO Child G ow h S an- da ds: Leng h/Heigh - o -Age, Weigh - o -Age, Weigh - o - Leng h, Weigh - o -Heigh and Body Mass Index- o -Age: Me hods and De elopmen . Gene a: WHO. 40. T os SG, Pa e RR, F eedson PS, e al. (2000) Using objec i e physical ac i i y measu es wi h you h: how many days o moni o ing a e needed? Med Sci Spo s Exe c 32, 426–431. 41. Oli e M, Schofield GM & Kol GS (2007) Physical ac i i y in p eschoole s: unde s anding p e alence and measu emen issues. Spo s Med 37, 1045–1070. Nu ien supplemen s and physical ac i i y 517 h ps:/www.camb idge.o g/co e/ e ms. h ps://doi.o g/10.1017/S0007114517000290 Downloaded om h ps:/www.camb idge.o g/co e. Tampe e Uni e si y and Uni e si y Hospi al, on 12 Jun 2017 a 07:01:59, subjec o he Camb idge Co e e ms o use, a ailable a 42. Pa e RR, O’Neill JR & Mi chell J (2010) Measu emen o physical ac i i y in p eschool child en. Med Sci Spo s Exe c 42, 508–512. 43. Na ional Associa ion o Spo and Physical Educa ion (2009) Ac i e S a : A S a emen o Physical Ac i i y Guidelines o Child en om Bi h o Age 5,2nded.Sewickley,PA:Ame ican Alliance o Heal h, Physical Educa ion, Rec ea ion and Dance. 44. Ga dne JMM, G an ham-McG ego SM, Chang SM, e al. (1995) Ac i i y and beha io al de elopmen in s un ed and nons un ed child en and esponse o nu i ional supple- men a ion. Child De 66, 1785–1797. 45. Ha ahap H, Jaha i AB, Husaini MA, e al. (2000) E ec s o an ene gy and mic onu ien supplemen on i on deficiency anemia, physical ac i i y and mo o and men al de elopmen in unde nou ished child en in Indonesia. Eu J Clin Nu 54, 114–119. 46. Abu o NJ, Rami ez-Zea M, Neu eld LM, e al. (2010) The e ec o nu i ional supplemen a ion on physical ac i i y and explo a o y beha io o Mexican in an s aged 8-12 mon hs. Eu J Clin Nu 64, 644–651. 47. Sazawal S, Ben ley M, Black RE, e al. (1996) E ec o zinc supplemen a ion on obse ed ac i i y in low socioeconomic Indian p eschool child en. Pedia ics 98, 1132–1137. 48. Ben ley ME, Caulfield LE, Ram M, e al. (1997) Zinc supple- men a ion a ec s he ac i i y pa e ns o u al Gua emalan in an s. J Nu 127, 1333–1338. 49. Kumwenda C, Dewey KG, Hemswo h J, e al. (2014) Lipid-based nu ien supplemen s do no dec ease b eas milk in ake o Malawian in an s. Am J Clin Nu 99, 617–623. 50. Hemswo h J, Kumwenda C, A imond M, e al. (2016) Lipid- based nu ien supplemen s inc ease ene gy and mac o- nu ien in akes om complemen a y ood among Malawian in an s. J Nu 146, 326–334. 51. Van Cauwenbe ghe E, Gubbels J, De Bou deaudhuij I, e al. (2011) Feasibili y and alidi y o accele ome e measu emen s o assess physical ac i i y in oddle s. In J Beha Nu Phys Ac 8, 67. 52. Len e s L, Wazny K, Webb P, e al. (2013) T ea men o se e e and mode a e acu e malnu i ion in low- and middle-income se ings: a sys ema ic e iew, me a-analysis and Delphi p ocess. BMC Public Heal h 13, Suppl. 3, S23. 53. Humph ey JH (2009) Child unde nu i ion, opical en e o- pa hy, oile s, and handwashing. Lance 374, 1032–1035. 54. Gough EK, Moodie EE, P ende gas AJ, e al. (2014) The impac o an ibio ics on g ow h in child en in low and middle income coun ies: sys ema ic e iew and me a-analysis o andomised con olled ials. BMJ 348, g2267. 55. Tsai SY, Bu RL & Thomas KA (2009) E ec o ex e nal mo ion on co espondence be ween in an ac ig aphy and ma e nal dia y. In an Beha De 32, 340–343. 518 A. Pulakka e al. h ps:/www.camb idge.o g/co e/ e ms. h ps://doi.o g/10.1017/S0007114517000290 Downloaded om h ps:/www.camb idge.o g/co e. Tampe e Uni e si y and Uni e si y Hospi al, on 12 Jun 2017 a 07:01:59, subjec o he Camb idge Co e e ms o use, a ailable a