Factors associated with stunting among childrenaged 0 to 59 months from the central regionof Mozambique

nu ien s A icle Fac o s Associa ed wi h S un ing among Child en Aged 0 o 59 Mon hs om he Cen al Region o Mozambique Loida Ma ía Ga cía C uz 1,2, Glo ia González Azpei ia 1, Deside io Reyes Súa ez 1, Al edo San ana Rod íguez 2, Juan F ancisco Lo o Fe e 3and Lluis Se a-Majem 4,5,* 1Resea ch Ins i u e o Biomedical and Heal h Sciences, Uni e si y o Las Palmas de G an Cana ia, 35016 Las Palmas de G an Cana ia, Spain; [email p o ec ed] (L.M.G.C.); [email p o ec ed] (G.G.A.); [email p o ec ed] (D.R.S.) 2 Clinical Gene ics Uni , Complejo Hospi ala io Insula -Ma e no In an il, 35016 Las Palmas de G an Cana ia, Spain; [email p o ec ed] 3Depa men o Clinical Sciences, Uni e si y o Las Palmas de G an Cana ia, 35016 Las Palmas de G an Cana ia, Spain; [email p o ec ed] 4Cibe OBN (CB06/03), Ins i u o Ca los III, Spanish Go e nmen , 28029 Mad id, Spain 5 Resea ch Ins i u e o Biomedical and Heal h Sciences, Uni e si y o Las Palmas de G an Cana ia/Nu i ion Wi hou Bo de s, 35016 Las Palmas de G an Cana ia, Spain *Co espondence: [email p o ec ed]; Tel.: +34-928-453-454; Fax: +34-928-453-475 Recei ed: 3 Ma ch 2017; Accep ed: 4 May 2017; Published: 12 May 2017 Abs ac : The objec i e o his s udy was o iden i y he majo socio-demog aphic, heal h, and en i onmen al de e minan s o s un ing among child en aged 0–59 mon hs om he Te e p o ince (Mozambique) and o e ing use ul in o ma ion o u u e heal hca e s a egies and in e en ions. A case-con ol s udy was conduc ed among 282 (162 boys; 120 gi ls) child en unde i e yea s o age om he cen al egion o Mozambique be ween 1 May and 3 June 2014. Child en wi h s un ing ( HAZ < −2 SD acco ding o he WHO Child G ow h S anda ds in 2006) we e conside ed as cases and hose who had a Z-sco e < − 2 SD we e conside ed as con ols. We collec ed da a ela ed o mo he s and child en and hei en i onmen , and hey we e assessed in wo g oups o ind a possible associa ion. The so wa e used o da a analysis was he SPSS ® ( e sion, 21.0) using desc ip i e s a is ics, - es , ANOVA, chi-squa e analyses, bi a ia e compa isons, and s epwise mul iple logis ic eg ession analysis. The esul s showed ha bi h weigh , mo he ’s educa ional s a us, ma e nal occupa ion, li ing in a u al a ea, amily size, numbe o child en unde i e yea s o age in he household, cooking wi h cha coal, inhabi ing wooden o s aw housing o housing wi hou p ope loo s, o e all du a ion o b eas eeding as well as du a ion o exclusi e b eas eeding, and ime o ini ia ion o complemen a y eeding we e signi ican ly ela ed o s un ing. Thus, app op ia e nu i ional in e en ion p og ammes conside ing hese de e minan s and he dissemina ion o knowledge a he popula ion le el ela ed o unde nu i ion a e necessa y o amelio a e he child en´s nu i ional s a us. Keywo ds: unde nu i ion; g ow h e a da ion; s un ing; nu i ional assessmen ; mozambique 1. Backg ound Unde nu i ion is a majo public heal h p oblem ha inc eases he global heal h bu den o p ema u e mo ali y and mo bidi ies du ing childhood [ 1 , 2 ]. I accoun s o 45% o all dea hs in child en unde i e yea s o age. Mo e han wo million child en unde i e yea s o age die each yea due o unde nu i ion a ound he wo ld [3,4]. Nu ien s 2017,9, 491; doi:10.3390/nu9050491 www.mdpi.com/jou nal/nu ien s Nu ien s 2017,9, 491 2 o 16 The p e alence o s un ing (ch onic unde nu i ion) (< − 2 s anda d de ia ion (SD) o low-heigh - o -age) emains one o he main p oblems o public heal h, and a signi ican p opo ion o people su e om mode a e o se e e unde nu i ion du ing ea ly childhood, especially in de eloping coun ies, as is he case in Mozambique. The p e alence o s un ing is he con en ional an h opome ic measu e ha e lec s long- e m ch onic unde nu i ion, ailu e o linea g ow h and mul i ac o ial social dep i a ion, a long- e m esponse o he p olonged dep i a ion o ood and/o p esence o disease. Unde nu i ion e e s o a s a e esul ing om a ela i e o absolu e de iciency o one o mo e essen ial nu ien s [ 5 – 9 ]. The h ee main indica o s used o de ine unde nu i ion, ha is o say, s un ing, unde weigh , and was ing, ep esen di e en nu i ional p oblems o he child and a e measu ed by he indices o heigh - o -age, weigh - o -heigh , and weigh - o -age, espec i ely. The Lance se ies on ma e nal and child unde nu i ion epo ed c i ical associa ions be ween s un ing (leng h/heigh - o -age Z-sco e < −2 SD) a age wo yea s and long- e m consequences [10,11]. Unde nu i ion in he i s 1000 days pos concep ion ep esen s an impo an was ing o human po en ial. In hese c ucial days, he building blocks a e es ablished o he de elopmen o he b ain and o u u e g ow h. Any al e a ion in his s age has long- e m implica ions, and he damage caused by unde nu i ion in he ea ly yea s o li e is la gely i e e sible. In un a o able en i onmen al condi ions, he majo i y o child en will emain locked in hei acqui ed g ow h channel. The school pe o mance o unde -nou ished child en is educed. They ha e lowe wo k capaci y and p oduc i i y as adul s. La e in li e hey ha e an inc eased likelihood o being o e weigh and de eloping associa ed ch onic diseases such as ca dio ascula disease, diabe es, cance , and men al heal h diso de s. Al hough he e has been a dec ease in p e alence, i has no been accompanied by a commensu a e dec ease in he numbe o s un ed child en due o he inc easing popula ion [ 12 – 18 ]. O e a 10-yea pe iod (1990–2010), A ica was he only egion whe e he numbe o s un ed child en unde i e yea s inc eased. P ojec ions o 2025 show ha he inc easing end is likely o emain [ 3 , 8 ]. ‘The UNICEF’s Annual Repo s’ consis en ly show he p e alence o s un ing in Sub-Saha an A ica o be a ound 30–40%, a ying pe coun y o egion o u ban e sus u al en i onmen [ 17 , 19 – 22 ]. Acco ding o ‘The 2016 Global Nu i ion Repo ’, he p e alence o s un ing in Mozambique in child en unde he age o i e yea s is s ill high (43.3%) while he unde weigh (19%) and was ing (5.9%) ca ego ies a e dec easing [23]. E en hough he p oblem o s un ing in Mozambique has been well documen ed, i s speci ic de e minan s a e no clea ly unde s ood. In addi ion, case-con ol o coho s udies ha e no been conduc ed o iden i y he isk ac o s o g ow h e a da ion in he Te e p o ince. The e is also inconsis ency ac oss s udies ega ding he de e minan ac o s behind s un ing in childhood. The e o e, his s udy a emp s o in es iga e he majo socio-economic, demog aphic, heal h, and en i onmen al de e minan s o s un ing among in an s aged 0–59 mon hs om he cen al egion o Mozambique, i.e., Te e p o ince. The esea ch esul s ob ained would be help ul o heal hca e p o ide s in designing and implemen ing app op ia e in e en ions o imp o e he g ow h o child en in he cen al egion o Mozambique and o he simila egions in he coun , and will enable he go e nmen and non-go e nmen al agencies o o mula e app op ia e policies and ini ia e in e en ion s a egies o he wellbeing o he popula ion. 2. Ma e ials and Me hods 2.1. S udy A ea and Popula ion A case-con ol s udy was conduc ed among child en unde i e yea s o age be ween 1 May and 3 June 2014 in he Te e p o ince, Mozambique. The p o ince is loca ed in he cen al egion o Mozambique, bo de ing Zambia on he No h and Zimbabwe on he Eas , and is he loca ion o he huge Caho a Bassa Dam and he ci y o Te e. I co e s an a ea o 100,724 km 2 . Acco ding o ‘The 2014 Annual Na ional Ins i u e o S a is ics’s Repo ’ [ 24 ], he p o ince has a popula ion o 2,500,000 indi iduals; app oxima ely 20.6% a e unde he age o i e. Nu ien s 2017,9, 491 3 o 16 The heal h sys em in Mozambique is p o ided by he Minis y o Heal h h ough hospi als, heal h cen es, and heal h pos s. The e a e h ee le els o o ganiza ion o heal h; he na ional, p o incial, and, inally, he dis ic le el. The lowes le el o ca e is p o ided by heal h pos s. Be ween he cen al hospi als and he heal h pos s he e a e o he ypes o heal h dispensing uni s such as heal h cen e s, u al hospi als, and p o incial o gene al hospi als [ 25 ]. Unde nu i ion is a e y common p oblem in he egion, and he p e alence o s un ing is 52% [ 22 ]. The s udy popula ion included child en aged 0 o 59 mon hs om wo heal hca e cen es o dis ic s 2 and 3 a he da a collec ion pe iod. 2.2. Sample Size Calcula ion and Sampling P ocedu e The sample size was calcula ed using OpenEpi (Open Sou ce Epidemiologic S a is ics o Public Heal h) e sion 3.5.4 [ 26 ] using o mula o wo popula ion p opo ions by assuming he p opo ion o s un ing o be 43.3% based on esul s om ‘The UNICEF’s Annual Repo ’ (52% in he Te e p o ince) [ 22 ]. In his ega d, a 5% le el o signi icance, a powe o 80%, and a a io case o con ol o 1:2 was assumed. The o al sample size was 282 (102 cases and 180 con ols). Based on he abo e assump ions, wi h an addi ional 15% added o non- esponse, he o al sample size was 324 child en wi h 110 o cases and 214 o con ols. We used a conse a i e design e ec o wo o adjus o a complex su ey design. The eason o his is ha wo s eps we e equi ed o each o iden i y he s udy pa icipan s. A consecu i e sampling echnique was used o selec he pa icipan s un il he calcula ed sample size was a ained. All child en aged 0 o 59 mon hs isi ing heal hca e cen es 2 and 3 du ing he da a collec ion pe iod we e measu ed o hei heigh o age z-sco e and ca ego ized as s un ed o no s un ed. Unde -nou ished child en we e i s iden i ied and hen selec ed as cases. The cases we e s un ed child en aged 0 o 59 mon hs wi h heigh - o age z-sco es (HAZ) below − 2 SD acco ding o he Wo ld Heal h O ganiza ion (WHO) Child G ow h S anda ds in 2006 [ 10 ]. They we e ca ego ized as se e ely s un ed i he HAZ was < − 3 SD. The con ols we e child en aged 0 o 59 mon hs wi hou s un ing, appa en ly heal hy and li ing in he communi y whe e cases esided. They a e conside ed o be appa en ly heal hy i he mo he s/ca e ake s pe cei ed he child o be heal hy and decla ed no symp oms o disease such as e e , coughing, di icul y o b ea hing, o dia hoea and wi hou unde nu i ion selec ed om he same heal hca e cen es. Child en wi h men al e a da ion, physical challenges, se ious illnesses, p e e m in an s (less han 37 weeks o ges a ion), child en su e ing om was ing and an exace ba ion o ch onic unde nou ishmen , and hose wi h mild ch onic unde nu i ion we e excluded om he s udy. A e measu emen s, mo he s o he esponden s we e in e iewed by he p incipal in es iga o based on a alida ed ques ionnai e. Only child en accompanied by hei own mo he s we e ec ui ed o a oid ecall bias. 2.3. Da a Collec ion and Quali y Assu ance A wo-day aining session was gi en o nine da a collec o s and one supe iso . The aining ocused on he ques ionnai e and he an h opome ic measu emen s. The ques ionnai e included he demog aphic cha ac e is ics o he child en and hei ca egi e s/mo he s such as he sex o he child, he age o he child and hei mo he , he le el o educa ion, occupa ion, and ca e p ac ices. To minimize ecall bias and measu emen bias we ha e o e ed ins uc ions o he da a collec o and a supe iso h ough a wo-day aining session ocused on he su ey and an h opome ic measu emen s. The ini ial design o he su ey was conduc ed in Spanish, based on he li e a u e de eloped o simila pu poses and inally ansla ed in o Po uguese ( he o icial language) a e being adap ed o he local con ex . I also included a physical examina ion and an an h opome ic measu emen . Bo h open and closed ques ions we e used. On a e age, he in e iew las ed 30–40 min. The da a collec ion ins umen s we e es ed p e iously and alida ed. The heigh and weigh o each child we e eco ded using he s anda d p ocedu e [27]. Nu ien s 2017,9, 491 4 o 16 The heigh o he subjec was eco ded wi h an accu acy o 0.10 cm wi h he help o a od ou o he an h opome e , wi h he head held in he F ank o ho izon al plane. Heigh was measu ed o child en abo e wo yea s o age, while leng h o child en below wo yea s using a leng h/heigh boa d. The weigh o he child en was aken wi h a po able weighing scale wi h an accu acy close o 0.1 kg, using minimum clo hing and wi h ba e ee , and he de ice was calib a ed equen ly. 2.4. Da a P ocessing and Analysis The da a we e analyzed s a is ically using he S a is ical Package o Social Sciences (SPSS ® Inc., Chicago, IL, USA; e sion 21.0). The desc ip i e s a is ical analysis o he da a ob ained is desc ibed in e ms o mean and s anda d de ia ion (SD). The homogenei y o a iance was es ed using Le ene’s es o equal a iances. Fo all an h opome ic a iables, i was obse ed ha he p- alue was s a is ically signi ican (p< 0.05). No mali y was es ed and also speci ies gende using he Shapi o-Wilk es o each one o he an h opome ic a iables, and i was no ed ha he p- alues we e no s a is ically signi ican (p> 0.05). An independen sample - es was pe o med o assess gende di e ences in he an h opome ic a iables. Analysis o a iance (ANOVA) using he Sche épos -hoc es was applied o e alua e he mean di e ences by age in he an h opome ic a iables. Chi-squa e analysis ( χ2 ) was used o assess he di e ences in he age-speci ic and gene al p e alence o s un ing be ween he sexes. A bina y logis ic eg ession analysis was conduc ed o es ima e he odds a ios (OR), and 95% con idence in e als (CIs) we e used o e alua e he possible di e ences and he associa ed isk ac o s among child en wi h heigh - o -age z-sco es less han − 2 SD (HAZ < − 2 SD) and no mal heigh - o -age z-sco es (HAZ > − 2 SD). The bina y logis ic eg ession analysis allowed he c ea ion o ca ego ical dependen a iables, and he odds we e ob ained by compa ing hese sco s wi h he e e ence ca ego y in he analysis o he uni a ia e independen model. To c ea e dicho omous dependen a iables (s un ed g ow h e sus no mal), he no mal code is coded as (0) and he delay o g ow h was coded as (1) in he eg ession models. The mul iple logis ic eg ession analysis by s ages (in ad ance) in he condi ional model was also ca ied ou o de e mine he mos e ec i e p edic i e a iables om he a iables conside ed in he bina y logis ic eg ession analysis. A s epwise backwa d elimina ion app oach was applied, and collinea i y was es ed in he inal model and epo ed. The odds a ios wi h 95% CIs we e calcula ed in o de o assess he adjus ed isk o independen a iables, and hose wi h p< 0.05 we e e ained in he inal model [28,29]. 2.5. E hical Conside a ions E hical app o al was p o ided by he dis ic heal h au ho i ies and he E hics Commi ee o he Te e Regional Hospi al, Mozambique. All he pa icipan s in ol ed in he s udy we e in o med abou he na u e o he s udy, he esea ch objec i es, and hecon iden iali y o he da a. The pa icipa ion o he subjec s was comple ely olun a y in na u e, and e bal consen was ob ained om each mo he be o e hei child en we e ec ui ed in o he s udy. All he subjec s we e ee om any physical de o mi ies and no su e ing om any disease a he ime o da a collec ion. To a oid any selec ion bias, he subjec s we e examined o any nu i ional de iciencies and ela ed diso de s. Any p e ious his o ies ela ed o medical and su gical episodes we e also aken in o conside a ion du ing he ime o examina ion. Pe sons eligible o pa icipa e in he s udy we e no o e ed a mone a y incen i e o pa icipa ion. In he cases whe e he esponden was illi e a e, we asked a li e a e pe son om he communi y o ead ou he consen o m and explain i o he head o he amily. Then we ob ained he humb p in o he esponden . In hose cases, he pe son who ead he consen o m also signed as a wi ness. The esea ch p ocedu es we e consis en wi h he Decla a ion o Helsinki [ 30 ]. In e iews we e adminis e ed a e ob aining in o med consen . The p o ocol was e iewed by a small g oup o expe s who had expe ience wo king in cen e s o eco e y and nu i ional educa ion and was amended based on hei ecommenda ions. Nu ien s 2017,9, 491 5 o 16 3. Resul s 3.1. Sociodemog aphic Cha ac e is ics The size o he sample was 282 pa ien s (n= 282) om 13 dis ic s. O hese, 69.2% (n= 195) li ed in an u ban a ea and 22.7% (n= 64) in a u al a ea. Rega ding dis ibu ion by sex, 57.4% o he sample and o he cases we e boys, wi h an a e age age o 42 mon hs ± SD o 18.3 ( ange: 0–59). Di e ences obse ed be ween pe cen ages o s un ing and no mally nou ished child en acco ding o sex we e s a is ically signi ican (p< 0.05). As o he child en’s age, i was highe in he g oup o s un ing. The mean ages ± SD o he cases and con ols was 43.3 ± 18.5 and 40.9 ± 18.2, espec i ely (p= 0.31). When hey we e g ouped by age in e als, a highe pe cen age o s un ing was shown in hose unde six and hose o e 24, bu he esul s we e no signi ican ei he (p> 0.05). The mean p egnancies ± SD o he cases and con ols we e 3.6 ± 1.8 and 1.6 ± 0.9, espec i ely, while hei mean bi h weigh s ± SD we e 2.8 ± 0.4 and 3.1 ± 0.3 kg, espec i ely (p< 0.001). The indings highligh ed ha he e we e signi ican associa ions be ween undesi able g ow h o he subjec s and hei mo he s’ educa ional le el (p< 0.001). The mean age o he mo he s in hei i s p egnancy ± SD o he cases and con ols was 16.8 ±1.8 and 20.5 ±2.5, espec i ely (p< 0.05) (Table 1). Table 1. Dis ibu ion o s un ing s a us by cha ac e is ics o mo he s and child en and hei en i onmen (N= 282). Da a is exp essed in equencies (%) o a e age ±SD (s anda de ia ion). Ch onically Malnou ished No mally Nou ished To al p-Value N= 102 N= 180 N= 282 n(%) n(%) n(%) X±σX±σX±σ Sex Male 79 (77.5) 83 (46.1) 162 (57.4) 0.001 *** Female 23 (22.5) 97 (53.9) 120 (42.6) Age (mon hs) 43.3 ±18.5 40.9 ±18.2 - 0.31 A ea o o igin U ban 20 (19.6) 175 (97.2) 195 (69.1) 0.001 *** Pe iu ban 20 (19.6) 3 (1.7) 23 (8.2) Ru al 62 (60.8) 2 (1.1) 64 (22.7) Reason o consul a ion Respi a o y in ec ion 17 (16.7) 10 (5.6) 27 (9.6) 0.001 *** Gas oin es inal in ec ion 43 (42.2) 90 (50.0) 133 (47.2) Fe e 33 (32.4) 78 (43.3) 111 (39.4) Weigh con ol and accina ion 9 (8.8) 2 (1.1) 11 (3.9) Rou ine checkups a16 (15.7) 157 (82.7) 173 (61.3) 0.001 *** Comple ed accina ion schedule b56 (54.9) 179 (99.4) 235 (83.3) 0.001 *** Age o beikos c <6 mon hs 75 (73.5) 54 (30) 129 (45.7) 0.001 *** ≥6 mon hs 27 (26.5) 126 (70) 153 (54.3) Time in mon hs Ce eals d5.51 ±2.42 7.76 ±2 - 0.001 *** F ui 7.01 ±3.02 7.92 ±2.04 - 0.003 ** Vege ables 7.5 ±3.31 8.28 ±2.1 - 0.02 * Cow’s milk 0.93 ±4.22 1.04 ±4.56 - 0.85 Cow’s mea (bee ) 13.3 ±6.74 11.48 ±9.33 - 0.06 Fish 11.85 ±5.04 13.93 ±5.51 - 0.002 ** Chicken 13.4 ±5.75 15.32 ±5.34 - 0.01 * Eggs 9.97 ±5.37 10.39 ±4.43 - 0.48 Yoghu 0.9 ±3.24 1.06 ±2.76 - 0.66 Legumes 7.77 ±4.06 8.18 ±2.37 - 0.35 Juices 3.83 ±5.51 6.99 ±3.43 - 0.001 *** Nu ien s 2017,9, 491 6 o 16 Table 1. Con . Ch onically Malnou ished No mally Nou ished To al p-Value N= 102 N= 180 N= 282 n(%) n(%) n(%) X±σX±σX±σ Siblings < 5 yea s old 78 (76.5) 13 (7.2) 91 (32.3) 0.001 *** Type o home S aw and wood 45 (44.1) 35 (19.4) 80 (28.4) 0.001 *** Clay b icks 53 (52) 79 (43.9) 132 (46.8) O he s 4 (3.9) 66 (36.7) 70 (30.6) Type o loo Soil 98 (96.1) 63 (35) 161 (57.1) 0.001 *** Conc e e 4 (3.9) 117 (65) 121 (42.9) Access o d inking wa e - 52 (28.9) 52 (18.4) 0.001 *** Cooking uel * Coal 19 (18.6) 128 (71.1) 147 (52.1) 0.001 *** Wood 83 (81.4) 30 (16.7) 113 (40.1) Gas - 22 (12.2) 22 (7.8) *p< 0.05 (Chi-squa e), ** p< 0.01, *** p< 0.001; a Those who ha e isi ed he heal hca e cen e a leas once in he las yea ha e been conside ed; b Comple ed accina ion schedule o he child’s age acco ding o he Minis y o Heal h in Mozambique; c Child en aged six mon hs o o e we e conside ed; d Ce eals consumed in o m o ‘xhima’ o ‘papinha’. 3.2. Feeding Habi s and P ac ices and Findings in he Examina ion o Malnou ished Child en Bo h he du a ion o b eas eeding (17.4 ± 5.4 and 11.4 ± 6.1 mon hs) and he in oduc ion o beikos (7.6 ± 2.3 and 5.3 ± 2.3 mon hs) we e highe in he g oup o no mally nou ished agains he g oup o malnou ished, being signi ican in bo h cases (p< 0.001). The ea ly inco po a ion o hese oods was associa ed wi h a highe isk o su e ing s un ing, and he obse ed di e ences we e s a is ically signi ican , excep o eggs and legumes. Mo eo e , he equency o in ake o oods was also analysed. No ele an di e ences we e obse ed be ween g oups in ega ds o he consump ion o oods such as yoghu , pas eu ized milk, ege ables, po a oes, chicken, and ish. Only a small pe cen age had yoghu egula ly and pas eu ized milk wi h suga occasionally. The majo i y a e ege ables, po a oes, chicken, and ish e e y day on a egula basis. Howe e , he e we e s a is ically signi ican di e ences obse ed ega ding he in ake o o he oods. The esul s ob ained om he compa ison o housewi es wi h hose who wo k ou side he home, showed ha he condi ion o ha ing a job ou side he home inc eases he isk o a child su e ing unde nu i ion (p< 0.05). The mos equen indings in physical examina ion o malnou ished child en we e changes in he skin (60.8%), loss o muscle mass (57.8%), al e a ions o he o al mucosa (57.8%), al e a ions o he subcu aneous cellula issue (56.9%), al e a ions in hai and nails (48%), ocula al e a ions (47.1%), dis ended abdomen (30.4%), and neu ological diso de s (16.7%) (Table 1). 3.3. Cha ac e is ics o he En i onmen O he ela i es li ed a home wi h he malnou ished in 65.7% o he cases, whils only in 9.4% o he con ol g oup had ela i es li ing wi h hem (p< 0.001) (Table 1). 3.4. Simple and Mul iple Logis ic Reg ession Analyses The esul s o he uni a ia e bina y logis ic eg ession analysis and he associa ions o sociodemog aphic a iables on he p e alence o s un ing among child en a e shown in Table 2. The ollowing ac o s emained and we e inco po a ed in o he eg ession model; he child ´ s age in mon hs, bi h weigh , mo he ’s educa ional s a us, ma e nal occupa ion, li ing in a u al a ea, amily size, numbe o child en unde i e yea s o age in he household, cooking wi h cha coal, and inhabi ing wooden o s aw housing o housing wi hou p ope loo s (Table 2). Nu ien s 2017,9, 491 7 o 16 Finally, he ac o s associa ed wi h s un ing we e examined in a mul iple logis ic eg ession model. A s epwise backwa d end was e ained in he inal model. The odds a ios wi h 95% CIs we e calcula ed in o de o assess he adjus ed isk o independen a iables, and hose wi h p< 0.05 we e e ained in he inal model. I was obse ed ha he child’s sex, bi h weigh , li ing in a u al a ea, ha ing siblings unde i e yea s o age, li ing in houses made o s aw and wood wi h soil loo , and homes whe e o he ela i es li ed oo, main ain s a is ically signi ican di e ences (p< 0.05). Sociodemog aphic ac o s we e signi ican ly associa ed wi h ch onic unde nu i ion in child en. The ela ionship be ween gende and ch onic unde nu i ion was s a is ically signi ican [Adjus ed Odds Ra io (AOR) = 4.57, 95% CI = (2.06–10.12), p< 0.05]. Bi h weigh was signi ican ly associa ed wi h s un ing [AOR = 19.99, 95% CI = (5.8–68.85), p< 0.001] and epo ed a lowe isk o s un ing. Child en li ing in u ban a eas signi ican ly epo ed a lowe isk o s un ing han hose in u al a eas [AOR = 138.0, 95% CI = (32.38–587.80), p< 0.001]. Table 2. Bina y logis ic eg ession (BLR) analysis and associa ions o sociodemog aphic a iables wi h s un ing. Bina y Logis ic Reg ession Analysis Uni a ia e Model † p-Value Cases Con ols C ude OR 95% IC N= 102 N= 180 n(%) n(%) Sex Male 79 (77.5) 83 (46.1) 4.01 (2.32; 6.95) 0.001 *** Female ( e .) 23 (22.5) 97 (53.9) - A ea o o igin U ban 40 (39.2) 178 (98.9) 137.95 (32.38; 587.65) 0.001 *** Ru al ( e .) 62 (60.8) 2 (1.1) - - Bi h weigh ≤2.50 kg 36 (35.3) 4 (2.2) 23.86 (8.18; 69.65) 0.001 *** >2.50 kg ( e .) 66 (64.7) 175 (97.8) - - Mo he s’ educa ion No educa ion/p ima y 102 (100) 123 (68.3) - - ≥Seconda y ( e .) - 57 (31.7) - - Mo he occupa ion Housewi e ( e .) 76 (74.5) 167 (92.8) - - O he s 26 (25.5) 13 (7.2) 0.23 (0.11; 0.47) 0.304 Ex ended amily household Yes 67 (65.7) 17 (9.4) 18.36 (9.63; 35) 0.001 *** No ( e .) 35 (34.3) 163 (90.6) - Siblings unde 5 yea s old Yes 78 (76.5) 13 (7.2) 41.75 (20.19; 86.33) 0.001 *** No ( e .) 24 (23.5) 167 (92.8) - - Type o home S aw and wood 45 (44.1) 35 (19.4) 3.27 (1.91; 5.60) 0.002 *** O he s ( e .) 57 (55.9) 145 (80.6) Type o loo Soil 98 (96.1) 63 (35.0) 45.5 (15.99; 129.46) 0.001 *** Conc e e ( e .) 4 (3.9) 117 (65.0) - - Access o d inking wa e Yes ( e .) - 52 (28.9) 0.56 (0.50; 0.625) - No 102 (100) 128 (71.1) - Cooking uel Coal ( e .) 19 (18.6) 150 (83.3) - - Wood 83 (81.4) 30 (16.7) 0.05 (0.024; 0.086) 0.001 *** *** p< 0.001; e .: e e ence ca ego y; OR—odds a io; CI—con idence in e al; † —bina y logis ic eg ession analysis conside ing e ec o one explana o y/p edic o a iable. Child en li ing in households wi h o he amily membe s [AOR = 17.3, 95% CI = (7.62–39.12), p= 0.001 ] and li ing in households wi h o he child en less han i e yea s o age we e mo e likely o de elop s un ing [AOR = 28.42, 95% CI = (11.93–67.70), p< 0.001]. Simila ly, s un ing was also ela ed o li ing in households which used cha coal o cooking, households made o s aw and wood [ AOR = 3.10, 95% CI = (1.53–6.26), p= 0.002 ], o hose households which did no ha e a p ope loo [AOR = 17.26, 95% CI = (5.87–50.75), p= 0.001] (Table 3). Nu ien s 2017,9, 491 8 o 16 Table 3. Unadjus ed and adjus ed odds a ios (OR) (95% CI) o s un ed child en aged 0–59 mon hs. Cases Con ols C ude Odds Ra io OR Adjus ed Odds Ra io aaOR p-Value N= 102 N= 180 n(%) n(%) 95% CI 95% CI Sex Male 79 (77.5) 83 (46.1) 4.01 (2.32; 6.95) 4.57 (2.06; 10.12) 0.001 *** Female ( e .) 23 (22.5) 97 (53.9) - - A ea o o igin U ban 40 (39.2) 178 (98.9) 137.95(32.38; 587.65) 138 (32.38; 587.80) Ru al ( e .) 62 (60.8) 2 (1.1) - - 0.001 *** Bi h weigh ≤2.50 kg 36 (35.3) 4 (2.2) 23.86 (8.18; 69.65) 19.99 (5.80; 68.85) 0.001 *** >2.50 kg ( e .) 66 (64.7) 175 (97.8) - - Mo he s’ educa ion No educa ion/1 a102 (100) 123 (68.3) - - - ≥Seconda y ( e .) - 57 (31.7) - - Mo he occupa ion Housewi e ( e .) 76 (74.5) 167 (92.8) - - O he s 26 (25.5) 13 (7.2) 0.23 (0.11; 0.47) 2.37 (0.46; 12.21) 0.304 Ex ended amily household Yes 67 (65.7) 17 (9.4) 18.36 (9.63; 35) 17.27 (7.62; 39.12) 0.001 *** No ( e .) 35 (34.3) 163 (90.6) - Siblings unde 5 yea s old Yes 78 (76.5) 13 (7.2) 41.75 (20.19; 86.33) 28.42 (11.93; 67.70) 0.001 *** No ( e .) 24 (23.5) 167 (92.8) - - Type o home S aw and wood 45 (44.1) 35 (19.4) 3.27 (1.91; 5.60) 3.10 (1.53; 6.26) 0.002 ** O he s ( e .) 57 (55.9) 145 (80.6) Type o loo Soil 98 (96.1) 63 (35.0) 45.5 (15.99; 129.46) 17.26 (5.87; 50.75) 0.001 *** Conc e e ( e .) 4 (3.9) 117 (65.0) - - Access o d inking wa e Yes ( e .) - 52 (28.9) 0.56 (0.50; 0.625) - No 102 (100) 128 (71.1) - Cooking uel Coal ( e .) 19 (18.6) 150 (83.3) - - Wood 83 (81.4) 30 (16.7) 0.05 (0.024; 0.086) 0.055 (0.024; 0.124) 0.001 *** ** p< 0.01, *** p< 0.001; CI—con idence in e al; signi icance conside ed when p< 0.05 and i he OR and adjus ed odds a ios (aOR) es ima e did no c oss he null o 95% CI. aOR adjus ed o child‘s sex and a ea o o igin. 4. Discussion Linea g ow h is he bes o e all indica o o child en’s well-being and p o ides an accu a e ma ke o inequali ies in human de elopmen . This is agically e lec ed in he millions o child en h oughou he wo ld who do no achie e hei ull po en ial o linea g ow h due o subop imal heal h condi ions and inadequa e nu i ion and childca e [ 31 ]. The nu i ion assessmen o he ulne able segmen s o he popula ion should be emphasised, no only o he iden i ica ion o nu i ional isks bu also o he imp o emen o exis ing heal h si ua ions. Cu en ly, he nu i ional scena io o he de eloping coun ies has been changing adically o e he pas wo decades, expe iencing he double bu den o malnu i ion (bo h unde and o e nu i ion) due o changes in socio-economic and demog aphic ansi ion, die a y habi s, li es yle modi ica ion, and inc easing isks o non-communicable diseases [32,33]. This s udy assessed he child g ow h si ua ion in he cen al egion o Mozambique (Te e p o ince) and iden i ied sociodemog aphic, heal h, and en i onmen al a iables as impo an de e minan s o s un ing o child en unde i e in he Te e p o ince. The indings showed ha he o e all p e alence o s un ing was obse ed o be 36.2% using he p oposed WHO e e ence (WHO 2006). The a e o 52% s un ing in he Te e p o ince is much highe han he 40% c i ical public heal h h esholds. This es ima ion is almos simila o he 43% le el epo ed by he 2007 Demog aphic Heal h Su ey (DHS) o his age g oup, sugges ing ha no much has changed be ween 2007 and 2016 [34]. Nu ien s 2017,9, 491 9 o 16 The s udy indings depic ed he sex o child o be a s ong de e minan o childhood s un ing. P e ious s udies om Sub-Saha an A ica ha e epo ed mixed indings on he e ec s o sex, wi h some sugges ing ha males a e mo e a ec ed by unde nu i ion compa ed o emales, while o he s epo ed o he wise [ 35 – 38 ]. In his s udy, males we e mo e likely o be s un ed. This is in line wi h he indings o a me a-analysis o DHS o 16 Sub-Saha an A ican coun ies [ 39 ], as well as he 2011 Mozambique DHS [ 40 ] and he 2010 DHS in Eas e n A ica [ 41 ]. S udies in Sub-Saha a A ican coun ies ha e epo ed gi ls o be mo e ecognized, i s because o hei high alue in ag icul u e and secondly due o he ac ha hey a e seen as an in es men , especially among he low socio-economic class, hus leading o mo e ca e and die a y p e e en ial ea men [ 39 ]. Conside ing ha o e 80% o Mozambique is ag icul u ally based and ha he majo i y o he popula ion is socio-economically cons ained, i.e. an es ima ed 61% a e below he po e y line o $1.25 pe day [ 42 ], his no ion may explain why males a e mo e malnou ished. Addi ionally, epidemiological e idence depic s boys o be biologically mo e ulne able o mo bidi y [ 43 , 44 ], and, in a se ing like Mozambique whe e mo bidi y incidences a e high, his p obably exe s conside able e ec s on boys. Childhood s un ing was ound o p og essi ely ise wi h an inc ease in age up o he age o 24 mon hs. A simila end has been epo ed in a numbe o p e ious s udies in de eloping coun ies [18,37,45–48]. The dec easing immune p o ec i e e ec s o b eas milk coupled wi h inc easing exposu e o con amina ed complemen a y oods, culmina ing in he onse o in ec ious diseases along wi h inc easing nu ien equi emen s, explains he end [ 49 – 52 ]. In line wi h a numbe o s udies om se ings like Cen al-Eas e n A ica (Kenya) [ 36 ], Tu key [ 53 ], and Eas e n A ica (Bu undi) [ 41 ] along wi h he ecen lance se ies, which pu child en bo n wi h low bi h weigh a a 20% isk o s un ing [ 3 ], low bi h weigh child en in his s udy we e a a highe isk o s un ing. Resea ch shows ha low bi h weigh babies a e bo n wi h low ese es o i al g ow h nu ien s; i amin A, zinc, and i on [54]. The e o e, hey depend on b eas milk o co e hese de iciencies. Howe e , as he amoun o hese nu ien s p esen in b eas milk mainly depends on ma e nal in ake and nu i ional s a us, low bi h weigh child en a e a a isk o no mee ing he ecommended amoun s i ma e nal in ake is inadequa e. Addi ionally, smalle han a e age babies expe ience eeding p oblems, a ac o in ol ed in unde nu i ion, e en in his s udy [ 55 ]. A numbe o s udies ha e shown ma e nal educa ion o p o ide p o ec i e e ec s agains all unde -nu i ion indica o s in child en [ 36 , 45 , 46 , 56 – 60 ]. The indings o his s udy a e inconsis en wi h he esul s o he s udies in e ms o g ow h e a da ion and unde weigh child en, wi h which no associa ion was ound. Pa icula ly, he esul s disag ee wi h he 10 yea s schooling h eshold epo ed in a ecen s udy conduc ed ac oss he h ee coun ies o Malawi, Zimbabwe, and Tanzania [ 61 ], as p o ec i e e ec s we e only ound o seconda y/ e ia y educa ion wi h no e ec s o p ima y comple ion. As epo ed in o he se ings [ 59 , 62 ], 10 yea s o schooling may ep esen an oppo uni y o achie ing a highe paying job, meaning be e access o he ma ke and leading o be e eeding p ac ices. Addi ionally, hese mo he s may ha e be e child and heal hca e knowledge, mo e heal h seeking beha iou s, lowe e ili y a es, and access o be e medical ca e. Ce ainly, in his s udy, a signi ican p opo ion o child en whose mo he s had seconda y/ e ia y educa ion ollowed a die ha me he ecommended die a y di e si y, in compa ison wi h hei coun e pa s whose mo he s we e uneduca ed. The s udy indings depic no associa ion be ween ma e nal occupa ion and any o he unde nu i ion indica o s. This is con a y o a ecen sys ema ic e iew on de e minan s o unde nu i ion in Sub-Saha an A ica [ 38 ], as well as o he s udies om Malaysia [ 63 ] and E hiopia [ 64 ], bu in ag eemen wi h indings o Adekanmbi and Kayode [ 45 ] om Nige ia. As he majo i y o mo he s a e in he ag icul u al sec o , which is la gely on he subsis ence le el, i could be ha he e con ibu ion owa ds he household income is oo low o in luence household decisions like eeding p ac ices. Child en o unde nou ished mo he s a e mo e likely o be s un ed. Simila indings ha e been documen ed in Tanzania [ 56 ] and o he low-income coun ies such as Nige ia [ 45 ] and E hiopia [ 18 , 58 ]. Though a g oup o e idence shows ha ma e nal nu i ional s a us has no e ec on he composi ion Nu ien s 2017,9, 491 16 o 16 61. Makoka, D.; Masibo, P.K. Is he e a h eshold le el o ma e nal educa ion su icien o educe child unde nu i ion? E idence om Malawi, Tanzania and Zimbabwe. BMC Pedia . 2015 ,15, 96. [C ossRe ] [PubMed] 62. Limwa ananon, S.; Tangcha oensa hien, V.; P akongsai, P. Equi y in ma e nal and child heal h in Thailand. Bull. Wo ld Heal h O gan. 2010,88, 420–427. [C ossRe ] [PubMed] 63. Shuhaimi, F.; Muniandy, N.D. The associa ion o ma e nal employmen s a us on nu i ional s a us among child en in selec ed kinde ga ens in Selango , Malaysia. Asian J. Clin. Nu . 2012,4, 53–66. [C ossRe ] 64. Negash, C.; Whi ing, S.J.; Hen y, C.J.; Belachew, T.; Hailema iam, T.G. Associa ion be ween Ma e nal and Child Nu i ional S a us in Hula, Ru al Sou he n E hiopia: A C oss Sec ional S udy. PLoS ONE 2015 ,10, e0142301. [C ossRe ] [PubMed] 65. Wo ld Heal h O ganiza ion. Global Heal h Risks: Mo ali y and Bu den o Disease A ibu able o Selec ed Majo Risks; Wo ld Heal h O ganiza ion: Gene a, Swi ze land, 2009; A ailable online: h p://www.who.in /i is/ handle/10665/44203 (accessed on 10 May 2017). 66. Da eh, E.K.M.; Acquah, E.; Kumi-Kye eme, A. Co ela es o s un ing among child en in Ghana. BMC Public Heal h 2014,14, 504. [C ossRe ] [PubMed] 67. Co eia, L.L.; Sil a, A.C.; Campos, J.S.; And ade, F.M.; Machado, M.M.; Lindsay, A.C.; Lei e, Á.J.; Rocha, H.A.; Cunha, A.J. P e alence and de e minan s o child unde nu i ion and s un ing in semia id egion o B azil. Re . Saude Publica 2014,48, 19–28. [C ossRe ] [PubMed] 68. Dewey, K.G. Impac o b eas eeding on ma e nal nu i ional s a us. In P o ec ing In an s h ough Human Mil; Sp inge : Be lin, Ge many, 2004; pp. 91–100. 69. Dewey, K.G.; Cohen, R.J. Does bi h spacing a ec ma e nal o child nu i ional s a us? A sys ema ic li e a u e e iew. Ma e n. Child Nu . 2007,3, 151–173. [C ossRe ] [PubMed] 70. G eene, M.E.; Me ick, T. Po e y Reduc ion: Does Rep oduc i e Heal h Ma e ? A ailable online: h p://documen s.wo ldbank.o g/cu a ed/en/709821468137992945/Po e y- educ ion-does- ep oduc i e-heal h-ma e (accessed on 10 May 2017). 71. Cleland, J.; Be ns ein, S.; Ezeh, A.; Faundes, A.; Glasie , A.; Innis, J. Family planning: The un inished agenda. Lance 2006,368, 1810–1827. [C ossRe ] 72. Fink, G.; Gun he , I.; Hill, K. The e ec o wa e and sani a ion on child heal h: E idence om he demog aphic and heal h su eys 1986–2007. In . J. Epidemiol. 2011,40, 1196–1204. [C ossRe ] [PubMed] 73. Checkley, W.; Gilman, R.H.; Black, R.E.; Eps ein, L.D.; Cab e a, L.; S e ling, C.R.; Moul on, L.H. E ec o wa e and sani a ion on childhood heal h in a poo Pe u ian pe i-u ban communi y. Lance 2004 ,363, 112–148. [C ossRe ] 74. Humph ey, J.H. Child unde nu i ion, opical en e opa hy, oile s, and handwashing. Lance 2009 ,374, 1032–1035. [C ossRe ] 75. Uni ed Na ions Child en’s Fund (Execu i e Boa d). Mozambique-UNICEF Coun y P og amme o Coope a ion, 2017–2020. Economic and Social Council, 2016. A ailable online: h p://www.un.o g/ ga/sea ch/ iewm_doc.asp?symbol=E/ICEF/2016/13 (accessed on 10 Ap il 2017). © 2017 by he au ho s. Licensee MDPI, Basel, Swi ze land. This a icle is an open access a icle dis ibu ed unde he e ms and condi ions o he C ea i e Commons A ibu ion (CC BY) license (h p://c ea i ecommons.o g/licenses/by/4.0/).