1Pesquisador Tit ular e Chef e, Laborat orio de Neuropsicologia Clínica, Set or de Neurociencias, Depart ament o de Pediat ria, Inst it ut o Fernandes Figueira, FIOCRUZ, Rio de Janeiro, RJ, e Prof essor Adjunt o Dout or, Universidade Federal Fluminense, Nit erói RJ, Brasil; 2M edical Off icer, Depart ment of Nut rit ion f or Healt h and Development , World Healt h Organizat ion, Geneva, Sw it zerland. Research support ed by FUNPENE.
Received 2 February 2004, received in f inal f orm 13 M ay 2004. Accept ed 13 July 2004.
Dr. Gilbert o N.O. Brit o - Caixa Post al 100846 - 24001-970 Nit erói RJ - Brasil. E-mail: ccsgnob@vm.uf f .br
Growth retardation - as measured by low
height-for-age - during infancy and early childhood has
be-en f ound t o have a st rong adverse eff ect on
cog-nit ive f unct ion in lat e childhood
1-3. How ever, less is
known about relationships between growth
impair-ment and behavior and neuropsychological perf
or-mance. In Brazil, t o a large ext ent , research in child
neuropsychology has been hindered by t he lack of
normat ive dat a f or neurobehavioral assessment
in children
4. Earlier research of t he f irst aut hor has
f ocused on t he development of normat ive dat a f or
neuropsychological inst rument s used in t he
asses-sment of children in Brazil such as t he Conners
Ab-breviat ed Teacher Rat ing Scale
5, t he Teacher Rat ing
Scale-Brazil (TRS-B)
6, t he Edinburgh Handedness
In-ventory
7-8, the Attention Deficit/Hyperactivity
Disor-der Teacher Rating Scale-Brazil (ADHD-DSM-IIIR-B)
9,
t he St eadiness Test and t he Purdue Pegboard
10. The
GROWTH STATUS, BEHAVIOR AND
NEUROPSYCHOLOGICAL PERFORM ANCE
A st udy of Brazilian school age children
Gilbert o N.O. Brit o
1, M ercedes de Onis
2ABSTRACT - Aim: To assess t he associat ion bet w een child grow t h and t eacher-report ed behavior and aca-demic st anding, and neuropsychological perf ormance. M et hod: The height s of 344 public-school children
w ere measured using st andard procedures. Teachers w ere request ed t o complet e t w o behavioral and one academic performance rating scales. Neuropsychological assessment consisted of the Edinburgh Handedness Inventory, the Steadiness Test and the Purdue Pegboard. Height-for-age z (HAZ) scores were calculated based on an int ernat ional ref erence. St andard (z) scores f or each f act or of t he f irst t w o behavioral rat ing scales and f or each neuropsychological variable w ere comput ed f or each child based on locally-derived norms.
Results: HAZ scores were negatively correlated with factors related to hyperactivity, conduct problem,
impulsi-vit y and inat t ent ion of t he behavioral rat ing scales. In addit ion, t he low er t he HAZ score t he w orse t he academic perf ormance. Conclusion: These result s add t o t he exist ing evidence indicat ing a st rong
associa-t ion beassocia-t w een grow associa-t h sassocia-t aassocia-t us and child behavior and academic perf ormance.
KEY WORDS: child grow t h, hyperact ivit y, inat t ent ion, conduct problem, neurobehavior.
Crescim ent o, com port am ent o e desem penho neuropsicológico: est udo de escolares brasileiros
RESUMO - Objetivo: Examinar a associação entre crescimento, comportamento na sala de aula, rendimento acadêmico e desempenho neuropsicológico em escolares. Método: A estatura de 344 escolares da rede
pú-blica de Niterói foi medida através de procedimentos padronizados. Os professores preencheram duas es-calas de avaliação comportamental e uma escala de avaliação do rendimento acadêmico. O exame neuro-psicológico incluiu o Inventário da Preferência Manual de Edimburgo, o Teste do Equilíbrio da Haste e a Pran-cha de Purdue. Escores padronizados da estatura para a idade (HAZ) foram calculados com base em referên-cia internacional. Escores padronizados (z) dos fatores das duas escalas de avaliação comportamental e das variáveis neuropsicológicas foram computados para cada criança com base em normas derivadas localmente.
Resultados: Escores HAZ apresentaram correlação negativa com fatores relacionados à hiperatividade,
pro-blemas de conduta, impulsividade e inatenção das escalas de avaliação comportamental. Ademais, quanto mais baixo o escore HAZ, pior se mostrou o rendimento acadêmico. Conclusão: Os resultados acrescentam
clinical applicat ions and relevance of t hese inst
ru-ments for research in child neuropsychology in Brazil
are described elsew here
11.
The aim of our st udy is t o build up on t he
earli-er research described above and explore the
associa-tion between child growth (i.e., height-for-age) and
neurobehavioral perf ormance in a sample of
Bra-zilian school age children as assessed by t he
TRS-B, the ADHD-DSM-IIIR-TRS-B, the Academic Performance
Teacher Rating Scale (AcPTRS-B), the Edinburgh
Han-dedness Invent ory, t he St eadiness Test and t he
Purdue Pegboard.
M ETHOD
Subject s
A comput er-generat ed unbiased sample of 346 chil-dren was the subject of this study. The chilchil-dren had a mean age of 9.4 years and were drawn from a large public school in the greater Rio de Janeiro area (Niteroi) with over 2000 regist ered children in preschool, element ary and junior high school classes. We select ed t his school not only due t o it s large number of children but also because it at -t rac-t s children f rom all e-t hnic groups and social s-t ra-t a, albeit most ly f rom low er social classes, w it h place of res-idence in communit ies dist ribut ed t hroughout Nit eroi, a city located across the Guanabara Bay from Rio de Janei-ro. Children w ere screened f or t he presence of minor physical disabilities, motor and vocal tics, speech disorders, and ot her behavioral deviances (e.g., nail bit ing, st reot y-pies) by t he use of a physical handicap and behavior che-cklist described elsew here11and t hose w it h impairment
deemed capable t o int erf ere w it h perf ormance of t he St eadiness Test and t he Purdue Pegboard (e.g., visuo-mot or impairment or severe visuo-mot or t ics) w ere excluded f rom t he st udy. Only t w o boys, one w ho ref used t o be t est ed and anot her boy w ho present ed w it h a severe f orm of st rabismus, w ere removed f rom t he st udy. The f inal sample comprised 344 children (173 boys and 171 girls). The social class dist ribut ion of children in t he pre-sent sample, according t o Hollingshead and Redlich12,
w as: I (N=2, 0.6% ), II (N=9, 2.6% ), III (N=34, 9.9% ), IV (N=119, 34.6% ) and V (N=176, 51.2% ). Social classes I and II include universit y-educat ed execut ives, business mana-gers in large concerns, and major and lesser professionals; social class III includes administ rat ive personnel, ow ners of small independent businesses and minor professionals; and social classes IV and V include skilled, semi-skilled and unskilled laborers with lower educational levels. One hun-dred and f ort y eight children (43.0% ) w ere Caucasian, 129 (37.5%) were of African ancestry and 67 (19.5%) were of a mult iple-race group.
The st udy w as approved by t he Board of Direct ors of t he Inst it ut o de Educação Prof . Ismael Cout inho (IEPIC) and performed under the auspices of the agreement cele-brated between IEPIC and the Instituto Fernandes Figuei-ra of t he Fundação Osw aldo Cruz (FIOCRUZ).
Ant hropomet ric measurement s
Height w as measured f ollow ing st andard procedu-res13. Tw o t rained individuals (t he f irst aut hor and a
re-search associat e) collect ed t he dat a on height under t he supervision of a nut rit ionist acquaint ed w it h ant hropo-met ric measurement procedures. Three measurement s of height , each of w hich required t o be w it hin 0.5 cm of t he ot hers, w ere t aken f or each child and t he mean height w as used f or dat a analysis. Age w as det ermined t hrough t he child’s birt h cert if icat e included in his/her school records. Height -f or-age z scores (HAZ) w ere calcu-lat ed based on t he Nat ional Cent er f or Healt h St at is-t ics/World Healis-t h Organizais-t ion (NCHS/WHO) ref erence using t he age-specif ic regression equat ions14and
imple-ment ed in t he ANTHRO sof t w are15. HAZ w as t hen used
as a cont inuous variable in dat a analysis.
Behavioral assessment
The t eachers of t he children w ere request ed t o com-plet e t he TRS-B6and t he ADHD-DSM -IIIR-B9. Bot h rat ing
scales w ere scored f rom not at all (0) and just a lit t le (1) t o pret t y much (2) and very much (3).
The TRS-B is comprised of 62 it ems dist ribut ed among five factorial dimensions (Hyperactivity-Conduct Problem; Independent Funct ioningPosit ive Socializat ion; Inat -t en-t ion; Neuro-t icism/Anxie-t y; Nega-t ive Socializa-t ion). St andard (z) scores f or each of t hese f act ors w ere compu-t ed f or each child based on locally-derived norms6and
used as indices of behavioral f unct ioning. St andard (z) scores w ere t hen ent ered as cont inuous variables in t he st at ist ical analysis of t he dat a. Except f or t he behavioral dimension Independent Funct ioning-Posit ive Socializa-t ion, higher (and more posiSocializa-t ive) sSocializa-t andard z scores signi-f y t hat t he child is having more disigni-fsigni-f icult ies along t he be-havioral dimensions derived f rom t he TRS-B.
The ADHD-DSM IIIR-B includes t he ent iret y of t he diagnost ic crit eria f or At t ent ion Def icit /Hyperact ivit y Disorder (ADHD) according t o DSM -IIIR16. St andard (z)
scores f or each of t he f act ors ext ract ed f rom t his scale, Hyperact ivit y-Impulsivit y and Inat t ent ion, w ere compu-t ed f or each child based on locally-derived norms9and
also used as indices of behavioral f unct ioning. St andard (z) scores w ere t hen used as cont inuous variables in t he analysis. Higher (and more posit ive) st andard z scores signif y t hat t he child is having more dif f icult ies along t he t w o behavioral dimensions derived f rom t he ADHD-DSM IIIR-B.
Academic assessment
The t eachers w ere also request ed t o complet e t he AcPTRS-B11f or each child and inst ruct ed t o rat e his/her
Neuropsychological assessment
The procedures f or neuropsychological t est ing are described in det ail elsew here10. Brief ly, t he
neuropsycho-logical bat t ery used in t he present report w as administ e-red over t w o sessions (one session a day) by t he f irst au-t hor and a research associaau-t e. Children w ere au-t esau-t ed indi-vidually in a quiet and acclimat ized room of t he school.
The f irst session included t he f ollow ing t est s:
Edinburgh Handedness Invent ory. This inst rument
w as used t o assess t he manual pref erence of t he part ici-pant s. The procedures f or t he administ rat ion of t he Edin-burgh Handedness Inventory in children were as described before8. Each child was requested to perform the ten tasks
included in t he invent ory and t he hand used t o perf orm each act ivit y w as recorded. Lat eralit y quot ient s (LQs) could range f rom -100 (consist ent lef t -handedness) t o +100 (consist ent right -handedness)8.
St eadiness Test . This t est w as used t o assess mot or
persistence, motor coordination and attention17. The
pro-cedures f or t his t est w ere as report ed by Brit o and San-t os-M orales10and Gardner17. Three t rials of 60 sec w ere
administ ered w it h t he hand t he child pref erred t o use in t he t est as w ell as w it h t he non-pref erred hand. Trials f or each hand w ere administ ered alt ernat ely. St andard (z) scores f or t ot al number of cont act s and t ot al t ouch t ime across t he t hree t rials w ere derived f or each child based on Brazilian norms10.
The second session consist ed of t he f ollow ing t est :
Purdue Pegboard. This inst rument provides an
excel-lent test of fine motor coordination in children with neuro-development al disabilit ies17. Procedures f or t his t est w
e-re as described in Brit o and Sant os-M orales10and Tiff in18.
St andard (z) scores f or number of pegs placed w it h t he
Table 1. M ean (and SD) z scores, and minimum and maximum z scores f or height -f or-age, and t he behavioral and neu-ropsychological variables. N=344 children.
M ean SD M inimum M aximum
Height -f or-age 0.28 1.10 -2.92 4.61
Behavior TRS-B
Hyperact ivit y/Conduct Problem -0.21 0.99 -1.56 4.07
Independent Funct ioning/Posit ive Socializat ion 0.16 0.54 -1.21 1.69
Inat t ent ion -0.77 0.34 -1.56 0.22
Neurot icism/Anxiet y -0.81 0.32 -1.56 0.15
Negat ive Socializat ion -0.91 0.28 -1.56 -0.15
ADHD-DSM -IIIR-B
Hyperact ivit y/Impulsivit y -0.76 0.36 -1.56 0.42
Inat t ent ion -0.98 0.28 -1.71 -0.28
Neuropsychological Perf ormance Edinburgh Handedness Invent ory
Lat eralit y quot ient 75.40 53.13 -100.00 100.00
St eadiness Test
Tot al number of cont act s
Pref erred hand 0.0 0.95 -1.88 2.74
Non-pref erred hand 0.0 0.95 -2.19 2.80
Tot al t ouch t ime (s)
Pref erred hand 0.0 0.95 -1.42 2.85
Non-pref erred hand 0.0 0.95 -1.61 2.86
Purdue Pegboard Number of pegs placed
Pref erred hand 0.0 0.95 -2.42 2.06
Non-pref erred hand 0.0 0.95 -2.42 2.31
Tot al t ime t o place 10 pegs (s)
Pref erred hand 0.0 0.95 -1.94 2.38
Non-pref erred hand 0.0 0.95 -2.24 2.60
Tot al t ime t o t ransf er 10 pegs (s)
Pref erred hand 0.0 0.95 -2.17 2.54
Non-pref erred hand 0.0 0.95 -1.91 2.50
Bot h hands 0.0 0.95 -2.62 2.21
Assembly 0.0 0.95 -2.29 2.56
pref erred and non-pref erred hands, number of pegs placed w it h bot h hands, t ot al t ime t o place 10 pegs a-cross t hree t rials w it h t he pref erred and non-pref erred hands, t ot al t ime t o t ransf er 10 pegs across t hree t rials w it h t he pref erred and non-pref erred hands and t ot al number of it ems assembled in t he “ assembly” modali-t y w ere obmodali-t ained f or each child based on normamodali-t ive da-t a gada-t hered in Brazil10.
St at ist ical analysis
The data were analyzed using PC-SAS19. Proc CORR was
used t o det ermine w het her grow t h st at us correlat ed w i-t h neurobehavioral variables. Addii-t ionally, w e used i-t he Spearman’s rank-order correlat ions as t he assumpt ions
f or t he use of paramet ric procedures could not be met by t he neurobehavioral dat a. Furt hermore, t he indivi-dual correlat ion w as used as t he concept ual unit of error rat e due t o t he explorat ory nat ure of our st udy.
RESULTS
Table 1 show s mean (and SD) z scores, and
min-imum and maxmin-imum z scores for height-for-age and
the behavioral and neuropsychological variables.
On-ly nine children (2.6% ) had HAZ scores equal or
be-low -2.0. The dist ribut ion of scores on t he
AcPTRS-B w as t he f ollow ing: (1) much higher t han
avera-ge - N = 9 (2.6% ); (2) higher t han averaavera-ge - N = 64
(18.6% ); (3) average - N = 209 (60.8% ); (4) below
average - N = 44 (12.8% ); (5) much below average
- N = 18 (5.2% ). Lat eralit y quot ient s (LQs) w ere
dis-t ribudis-t ed as f ollow s: (1) -100&ldis-t; LQ&ldis-t;-50, N = 30 (8.7% );
(2) -50<LQ<0, N = 5 (1.5% ); (3) 0<LQ<+50, N = 0;
(4) +50<LQ<+100, N = 309 (89.8% ).
Table 2 shows the Spearman’s rank order
correla-t ion coeff iciencorrela-t s becorrela-t w een HAZ and correla-t he z-scores f or
t he behavioral and neuropsychological variables.
The dat a present ed in t he t able demonst rat e t hat
HAZ is signif icant ly and negat ively correlat ed w it h
t w o specif ic behavioral domains derived f rom t he
TRS-B: Hyperact ivit y/Conduct Problem and Inat t
en-t ion. The low er en-t he HAZ of en-t he child en-t he higher
his/her z-score in t hese t w o behavioral domains.
The signif icant negat ive correlat ion coeff icient s
also f ound bet w een HAZ and t he z-scores of t he
behavioral domains ext ract ed f rom t he
ADHDDSM IIIRB, i.e., Hyperact ivit y/Impulsivit y and Inat
-tention, provide further confirmation of the
specifi-cit y of t he relat ionship bet w een height -f or-age
and Hyperact ivit y and Inat t ent ion.
Teacher rat ings of academic perf ormance
accor-ding to the AcPTRS were also found to be
significan-tly and negatively correlated with HAZ, i.e., the low er
t he lat t er scores t he higher (and w orse) t he rat ing
of academic perf ormance.
None of t he ot her behavioral and
neuropsycho-logical variables had a signif icant correlat ion w it h
HAZ.
DISCUSSION
The result s show t hat HAZ is negat ively and
spe-cif ically correlat ed w it h t he behavioral domains of
Hyperact ivit y/Conduct Problem and Inat t ent ion
derived f rom the TRS-B. The significant negative
cor-relations also f ound bet w een HAZ and t he t w o
be-havioral domains of Hyperact ivit y/Impulsivit y and
Inat t ent ion ext ract ed f rom t he ADHD-DSM -IIIR-B
Table 2. Spearman’s rank-order correlat ion coeff icient s bet w eenheight -f or-age z-scores and z-scores f or t he behavioral, academ-ic and neuropsychologacadem-ical variables. N=344 children.
R P
Behavior TRS-B
Hyperact ivit y/Conduct Problem* -0.139 0.009 Independent Funct ioning/
Posit ive Socializat ion 0.033 0.538
Inat t ent ion* -0.116 0.030
Neurot icism/Anxiet y -0.073 0.172
Negat ive Socializat ion -0.049 0.355 ADHD (DSM -IIIR)-B
Hyperact ivit y/Impulsivit y* -0.111 0.039
Inat t ent ion* -0.113 0.035
Academic perf ormance* -0.146 0.006
Neuropsychological perf ormance Edinburgh Handedness Invent ory
Lat eralit y quot ient 0.011 0.835
St eadiness Test
Tot al number of cont act s
Pref erred hand -0.004 0.940
Non-pref erred hand -0.047 0.375
Tot al t ouch t ime (s)
Pref erred hand -0.012 0.814
Non-pref erred hand -0.021 0.685
Purdue Pegboard Number of pegs placed
Pref erred hand -0.054 0.314
Non-pref erred hand -0.033 0.532
Tot al t ime t o place 10 pegs (s)
Pref erred hand 0.037 0.493
Non-pref erred hand -0.015 0.773
Tot al t ime t o t ransf er 10 pegs (s)
Pref erred hand -0.033 0.541
Non-pref erred hand -0.042 0.436
Bot h hands 0.014 0.785
Assembly 0.024 0.650
* Signif icant correlat ion coef f icient TRS-B, Teacher Rat ing Scale-Brazil;
conf irm t he specif icit y of t he relat ionship bet w een
grow t h st at us and Hyperact ivit y and Inat t ent ion.
These significant negative correlations indicat e t hat
low er HAZs are associat ed w it h higher degrees of
neuropsychopat hology along t hose t w o
behavio-ral domains. The results also show a significant
asso-ciat ion bet w een HAZ and t eacher’s rat ings of
aca-demic perf ormance; t he low er t he HAZ t he w orse
the rating of academic performance. No associations
w ere f ound bet w een HAZ and neuropsychological
variables derived from the Edinburgh Inventory, the
St eadiness Test and t he Purdue Pegboard.
The associat ion of HAZ and t eacher rat ings of
children’s behavior in the domains
Hyperactivity/Im-pulsivity and Inattention reported in the present st
u-dy is consistent with the finding that the birth weight
of children w it h generalized hyperkinesis is signif
-icantly lower than that of normal children
20.
Further-more, a study performed in Brazil demonstrated that
grow t h st at us represent s a f act or associat ed w it h
attention deficit-hyperactivity disorder
21. Moreover,
a recent st udy show ed t hat children st unt ed in
ear-ly childhood had more conduct problems at home,
regardless of their social backgrounds, and fewer
pro-social behaviors at school t han non-st unt ed
chil-dren
22.
The signif icant associat ion f ound in our st udy
between HAZ and teacher’s ratings of children’s
aca-demic perf ormance adds t o t he evidence f rom a
number of studies demonstrating a relationship bet
-ween growth status and academic performance and
cognitive functioning
22-26. Although our study could
not ascertain the effects of stunting in early
childho-od on later school achievement due to the small
num-ber of children who were stunted in our sample, t he
direct ion of t he associat ion bet w een height -f
or-age and academic perf ormance is in agreement w
i-t h i-t hese earlier f indings. Furi-t hermore, an
associa-t ion beassocia-t w een body-size measures and menassocia-t al associa-t esassocia-t
scores in a populat ion of w ell-nourished children
has also been report ed by ot her invest igat ors
27.
The precise mechanism linking physical grow t h
and poor neurobehavioral development is not
kno-wn. There has been a long standing interest in
whe-t her general or specif ic nuwhe-t riwhe-t ional def iciencies awhe-t
a sensit ive or crit ical period of brain grow t h or
ma-t urama-t ion could have a long lasma-t ing or permanenma-t
“ programming” influence on later cognitive
perfor-mance. One not ew ort hy example of a nut rit ional
inf luence on cognit ive f unct ion is mat ernal iodine
def iciency, w hich may result in reduced cognit ive
f unct ion or even cret inism
28,29. Furt hermore,
prena-t al nuprena-t riprena-t ional deprivaprena-t ion changes sleep/w ake
me-chanisms
30and int erf eres w it h
benzodiazepine-induced spat ial learning def icit s
31in rat s as adult s.
How ever, it is diff icult t o ext rapolat e f rom such st
u-dies in rats to human behavior. In humans, there are
at least t w o commonly suggest ed explanat ory
hy-pot heses. One of t hese hyhy-pot heses is t hat malnut
ri-t ion impairs brain grow ri-t h and neural developmenri-t
and hence f ut ure behavior, w hile t he ot her st at es
t hat t he associat ion bet w een nut rit ional st at us and
cognit ive development is mediat ed by mot or mat
-urat ion, act ivit y level and explorat ory behavior
3.
The associat ion bet w een grow t h st at us and
cogni-t ive developmencogni-t cannocogni-t be regarded as a simple
causal relat ionship because of t he complex
(proxi-mal and dist al) environment al f act ors t hat af f ect
bot h grow t h and development . It is possible t hat
more t han one mechanism act in concert t o impair
neurobehavioral f unct ioning. It might also be
hy-pot hesized t hat low height -f or-age int erf eres w it h
school achievement by it s ef f ect on t he
behavio-ral domains relat ed t o hyperact ivit y/impulsivit y and
inat t ent ion as indicat ed by t he present f indings.
The f act t hat w e did not f ind a signif icant
cor-relation between height-for-age and the
neuropsy-chological variables derived from the Edinburg
Han-dedness Invent ory, t he St eadiness Test and t he
Purdue Pegboard is possibly due t o t he small
num-ber of children w it h severe height -f or-age def icit s
in our sample.
In conclusion, w e f ound signif icant associat ions
bet w een height -f or-age, ext ernalizing behaviors
and t eacher rat ings of academic perf ormance in
a large sample of Brazilian school-age children.
Ex-t ernalizing behaviors can inEx-t erf ere w iEx-t h academic
perf ormance and are considered t o predict f ut ure
psychosocial adjustment
22. However, the mechanism
of t he development al def icit s in academic perf
or-mance induced by grow t h ret ardat iondef icit s has
not yet been elucidated. The use of instruments that
assess academic perf ormance direct ly rat her t han
t eacher rat ings of academic perf ormance might of
-fer a more precise view of the relationship between
growth status and achievement. Such research is now
underw ay in our laborat ory.
Acknow ledgments - The authors are grateful to
package, Rubem Goulart (NPD-UFF) f or t he maint enan-ce of comput er syst ems and Tat ianna R. Sant os f or assis-t ance in assis-t he preparaassis-t ion of assis-t he manuscripassis-t . The auassis-t hors also w ish t o t hank t he children w ho part icipat ed in t he st udy, t heir t eachers and t he st af f of IEPIC (Inst it ut o de Educação Prof . Ismael Cout inho).
REFERENCES
1. Berkman DS, Lescano AG, Gilman RH, Lopez SL, Black MM. Effects of stunting, diarrhoeal disease, and parasitic infection during infancy on cognition in late childhood: a follow-up study. Lancet 2002;359:564-571. 2. Grantham-McGregor SM. A review of studies of the effect of severe
mal-nutrition on mental development. J Nutr 1995;125:S2233-2238S. 3. Gorman KS. Malnutrition and cognitive development: evidence from
experimental/quasi-experimental studies among the mild-to-moder-ate malnourished. J Nutr 1995;125:S2239-S2244.
4. Brito GNO, Alfradique GMN, Pereira CCS, Porto CMB, Santos TR. Developmental norms for eight instruments used in the neuropsycho-logical assessment of children: studies in Brazil. Braz J Med Biol Res 1998;31:399-412.
5. Brito GNO. The Conners Abbreviated Teacher Rating Scale: develop-ment of norms in Brazil. J Abn Child Psychol 1987;15:511-518. 6. Brito GNO, Pinto RCA. A Composite Teacher Rating Scale: analysis in
a sample of Brazilian children. J Clin Exp Neuropsychol 1991;13:417-418. 7. Brito GNO, Lins MFC, Paumgartten FJR, Brito LSO. Hand preference in 4- to 7-year-old children: An analysis with the Edinburgh Inventory in Brazil. Dev Neuropsychol 1992; 8:59-68.
8. Brito GNO, Santos-Morales TR. Lateral preferences in 8- to 15-year-old Brazilian children assessed with the Edinburgh Inventory: different meas-ures of handedness and comparison with younger children and adults. Dev Neuropsychol 1999;16:433-453.
9. Brito GNO, Lins MFC, Pinto RCA. A behavioral assessment scale for Attention Deficit Disorder in Brazilian children based on DSM-III-R crite-ria. J Abn Child Psychol 1995;23:509-520.
10. Brito GNO, Santos-Morales TR. Developmental norms for the Gardner Steadiness Test and the Purdue Pegboard: a study with children of a metropolitan school in Brazil. Braz J Med Biol Res 2002;35:931-949. 11. Brito GNO, Pereira CCS, Santos-Morales TR. Behavioral and
neuro-psychological correlates of hyperactivity and inattention in Brazilian school children. Dev Med Child Neurol 1999;41:732-739.
12. Hollingshead AB, Redlich FC. Social class and mental illness. New York: Wiley; 1958.
13. World Health Organization. Physical status: the use and interpretation
of anthropometry: report of a WHO Expert Committee. WHO Tech Rep Series 1995;854:424-434.
14. Dibley MJ, Goldsby JB, Staehling NW, Trowbridge FL. Development of normalized curves for the international growth reference: historical and technical considerations. Am J Clin Nutr 1987;46:736-748. 15. World Health Organization and US Centers for Disease Control.
AN-THRO software. Version 2000. Geneva: WHO; 2002.
16. American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 3rd ed-revised (DSM-III-R). Washington, DC: American Psychiatric Association, 1987.
17. Gardner RA. The objective diagnosis of minimal brain dysfunction. Cres-skill, NJ: Creative Therapeutics, 1979.
18. Tiffin J. Purdue pegboard: examiner manual. Chicago, IL: Science Re-search Associates, 1968.
19. SAS Institute. SAS/STAT user’s guide. Vol 2, 4thEd. Cary, NC: SAS Institute; 1990.
20. Gillberg C, Carlström G, Rasmussen P. Hyperkinetic disorders in sev-en-year-old children with perceptual, motor and attentional deficits. J Child Psychol Psychiatry 1983;24: 233-246.
21. Guardiola A, Fuchs FD, Terra AR, et al. Importância dos aspectos nutri-cionais como fator associado à síndrome de hiperatividade com déficit de atenção. Arq Neuropsiquiatr 1997;55:598-605.
22. Chang SM, Walker SP, Grantham-McGregor S, Powell CA. Early childho-od stunting and later behaviour and school achievement. J Child Psy-chol Psychiatry 2002;43:775-783.
23. Clarke NMA, Grantham-McGregor SM, Powell C. Nutrition and health predictors of school failure in Jamaican children. Ecol Food Nutr 1991;55:955-958.
24. Moock P, Leslie J. Childhood malnutrition and schooling in the Terai region of Nepal. J Dev Economics 1986;20:33-52.
25. Martorell R, Rivera J, Kaplowitz H, Pollit E. Long-term consequences of growth retardation during early childhood. In Hernandez M, Argente J, (eds). Human growth: basic and clinical aspects. Amsterdam: Elsevier, 1992:143-149.
26. PAHO. Nutrition, health and child development. Scientific Publication No. 566. Washington, DC: PAHO; 1998.
27. Pollitt E, Mueller W. The relation of growth to cognition in a well-nou-rished preschool population. Child Development 1982;53:1157-1163. 28. Morley R, Lucas A. Nutrition and cognitive development. Br Med Bull
1997;53:123-134.
29. Susser M. The challenge of causality: human nutrition, brain develop-ment and develop-mental performance. Bull NY Acad Med 1989;65:1032-1049. 30. Datta S, Patterson EH, Vincitore M, Tonkiss J, Morgane PJ, Galler JR. Prenatal protein-malnourished rats show changes in sleep/wake behavi-or as adults. J Sleep Res 2000;9:71-79.