A

K

I

mc[Ia

L

4

0d

Effect of micronutrient supplement on health and nutritional status ofschoolchildren: mental function

Shahnaz Vazir, Ph.D.a,*, Balakrishna Nagalla, Ph.D.a, Vijayapushpam Thangiah, M.A.b,Vijayaraghavan Kamasamudram, M.B.B.S.a, and Sivakumar Bhattiprolu, Ph.D.c

a Division of Field Studies, National Institute of Nutrition, Hyderabad, Indiab Extension and Training Division, National Institute of Nutrition, Hyderabad, India

c Division of Biophysics, National Institute of Nutrition, Hyderabad, India

Manuscript received April 6, 2004; accepted July 23, 2005.

bstract Objective: We evaluated the effect of a micronutrient-fortified beverage on mental function inschoolchildren.Methods: This double-blind, placebo-controlled, matched-pair, cluster, randomized feeding trialassessed baseline before supplementation and after 14 mo in a cohort of 608 children aged 6 to 15 ywho received a micronutrient-fortified beverage or placebo. Children in two classes (clusters) ineach of nine grades were assessed for intelligence, attention and concentration, memory, and schoolachievement.Results: Mean intelligence quotient scores of the micronutrient-fortified beverage group and theplacebo group at baseline and final follow-up were not significantly different. Mean verbal andperformance scores and increments were also not different. After supplementation, attention-concentration increment scores of the micronutrient group were significantly higher (P � 0.05) thanthose of the placebo group. Mean and incremental scores of the supplemented group on the memoryscale were not significantly different from those of the placebo group.Conclusions: Supplementation with a beverage fortified with a range of micronutrients signifi-cantly improved attention-concentration over 14 mo, but not intelligence quotient, memory, orschool achievement among middle-income schoolchildren. © 2006 Elsevier Inc. All rights reserved.

Nutrition 22 (2006) S26–S32www.elsevier.com/locate/nut

eywords: Micronutrient-fortified beverage; Children; Intelligence; Attention/concentration; Memory; School achievement

et

tcb6soD1ivn

ntroduction

Early stimulation, socioeconomic status, home environ-ent, nutritional status, and interactions between parent and

hild influence the mental function of growing children1,2]. Although studies among children and adolescents inndia and abroad have indicated that micronutrient statusnd cognitive functions are positively associated [3–6], the

This study was supported by GlaxoSmithKline Consumer Healthcaretd., India.

* Corresponding author. Tel.: �91-40-2700-8921, ext. 274; fax: �91-0-2701-9074.

bE-mail address: s_vazir@hotmail.com (S. Vazir).

899-9007/06/$ – see front matter © 2006 Elsevier Inc. All rights reserved.oi:10.1016/j.nut.2004.07.021

ffect of supplementing a beverage with multiple micronu-rients on mental function has not been previously reported.

Studies on supplementation with multiple micronutrientso children with or without subclinical micronutrient defi-iencies in developed countries are divided on whether itenefits intellectual performance [6–13]. Among a group of- to 15-y-old schoolchildren given a multiple micronutrientupplement in doses of 50%, 100%, and 200% of the rec-mmended dietary allowance (RDA) according to the U.S.epartment of Agriculture [14], only those provided with00% of the RDA demonstrated a significant improvementn non-verbal intelligence [11]. There was no effect onerbal or crystallized intelligence. A selective response toon-verbal tests was predicted due to the fact that they

etter represent biological functioning than do verbal tests.

Hipiitpac

fett

eptsiitsIwop[(w

M

t6snia

w

I

[gIaosav

dfmtmsaociMsd

M

PCmfcpsasffrStscstdcst

smrpswTe

A

tt

S27S. Vazir et al. / Nutrition 22 (2006) S26–S32

owever, only children with lower dietary micronutrientntake responded to supplementation [11], suggesting theresence of subclinical deficiencies. Interestingly, two stud-es that followed similar methodologies found no significantmprovement in intelligence quotient (IQ) [8,9]. Micronu-rient supplementation has also been associated with im-roved attention and concentration [7] and, in children withttention deficits, increased the time these children spentoncentrating [12].

None of the studies reviewed had used food as a vehicleor providing micronutrients as per the RDA. The previousvidence concerned micronutrient supplementation throughhe provision of tablets or capsules. Moreover, studies tendedo evaluate development before and after intervention.

The present study tested the hypothesis that micronutri-nt supplementation of a fortified beverage would improverimary outcomes such as IQ, memory, attention-concen-ration, and school achievement over time in residentialchoolchildren who had no overt signs (i.e., signs not phys-cally visible) of deficiency and adequate macronutrientntake (�75% of the RDA). The secondary hypothesis washat micronutrient supplementation would improve the con-tituents of primary outcomes, i.e., verbal and performanceQs and school marks on math, science, and social studieshether or not overall IQ, memory, attention-concentration,r aggregate school marks improved. The secondary hy-othesis was considered important because previous studies4,13] have indicated that verbal IQ and fluid intelligenceperformance IQ) improves after supplementation evenhen IQ remains unchanged.

aterials and methods

A double-blind, placebo-controlled, matched-pair, clus-er randomized study was undertaken in children who were

to 15 y old and in grades 1 to 10 from two residentialchools that catered to middle-income semi-urban childrenear Hyderabad, India. The experimental design, random-zation, the sample covered for assessing mental function,nd statistical analysis have been published previously [22].

The following culture-appropriate standardized testsere administered to the children.

ntelligence

Malin’s Intelligence Scale for Indian Children (MISIC)15], which is an Indian adaptation of the Wechsler Intelli-ence Scale for Children [16], is currently in wide use inndia. Because of its appropriateness, this test was used tossess intelligence. It is suitable for use in children 6 to 15 yf age. The MISIC comprises Verbal and Performancecales with six subscales, i.e., information, comprehension,rithmetic, similarities, vocabulary, and digit span under the

erbal and five subscales, namely picture completion, block s

esign, object assembly, coding, and mazes under the Per-ormance scale. The test provides a verbal IQ, a perfor-ance IQ, and a full-scale IQ for each individual taking the

est. The test-retest reliability of the MISIC using productoment correlation was 0.91. The concurrent validity with

chool ranks was 0.61 and congruent validity from thedapted California Short Form Test of mental maturity forlder ages and Goodenough’s Draw-a-man Test for youngerhildren was similar at 0.63 [15]. The norms for comparisonn the MISIC are given separately for age and gender. The

ISIC is standardized with an average IQ of 100 and 15tandard deviation points. The items in the test increase inifficulty according to age from 6 to 15 y.

emory

The PGI Memory Scale, which was developed by theost Graduate Institute of Medical Education and Research,handigarh, India [17], was used to assess different types ofemory. Although this test was developed to assess dys-

unction in adults, some test items were pretested on school-hildren before beginning the study and modified for theurpose of the present study. During the pretest, items andcores were compared with similar items from other IQ testslready in use for children in India (memory being a con-tituent of IQ) such as the Binet-Kamat Test of Intelligenceor Indian Children and the MISIC and the scores obtainedrom them. Correlations (r) for the scores of these itemsanged from 0.7 to 0.9. Almost all items in the PGI Memorycale were similar to the memory items found in these IQ

ests. The difficulty levels were also similar. In the presenttudy, raw scores obtained on this test for each child wereonverted into Z scores for a given age and transformed intotandard scores, with a mean of 100 and a standard devia-ion of 15 [18]. Z scores were used to compare meanifferences between groups instead of against any norms forlassifying children. Because the same test and method ofcoring and transformation were used for both study groups,here was no scope for error.

The PGI Memory Test comprises 10 subtests that con-titute total memory. Of these, seven, i.e., recent memory,ental balance, remote memory, delayed recall, immediate

ecall, retention of similar pairs, and retention of dissimilarairs, were used in this study because these were found to beuitable for children in the pretest. The remaining three testsere not found to be suitable for children 6 to 15 y of age.his was the only Indian standardized test available for thexclusive assessment of memory.

ttention and concentration

The Knox Cube Imitation Test and the Letter Cancella-ion Test [19] were used to assess attention-concentration ofhe study children. The Knox Cube is an imitation by the

ubject of an assorted series of taps on cubes in predeter-

mcsstiminiu

S

mf

acaIdtco

t

ccmbbpvvKmodtw

R

crfm

f�pct

TMa

V

MPPPMA

SHWEPHCVVPRVF

t

s

TI

CDP

ferent b

S28 S. Vazir et al. / Nutrition 22 (2006) S26–S32

ined sequences with increasing difficulty. The Letter Can-ellation Test requires a set of six specified letters to betruck off from a jumble of letters on a single page within apecified time limit. Both tests require eye-hand coordina-ion, speed, and sustained attention and concentration dur-ng performance. However, the Knox Cube Imitation Test isore difficult because it requires diverse skills such as

mmediate recall of the series of taps to be imitated, alert-ess, and ability to perceive and remember sequences ofncreasing difficulty. The Letter Cancellation Test only callspon recent memory and recall of the specified letters.

cholastic achievement

Science, mathematics, and social studies and aggregatearks of quarterly and final annual examinations were used

or analyzing school achievement.Eighteen psychologists were recruited and trained in the

dministration of the cognitive tests. The examiners (psy-hologists) and the entire research team and the schooluthorities were blinded to the content of the beverages.nterexaminer agreement and consistency were confirmeduring training and randomly checked during the course ofhe study before each assessment. The reliability coeffi-ients ranged from 0.85 to 0.90 on the IQ test and was 0.99n the Knox Cube and Memory tests.

After completion of baseline data, supplementation ofhe drink was started.

All tests were administered on the same day for eachhild. Psychologists completed the administration of tests tohildren in both treatment groups of one grade before com-encing to the next grade. Comparison of baseline values

etween groups was done to assess comparability or “bias”etween groups. To assess the effect of intervention, com-arisons were done between groups during and after inter-ention. Mean differences and increments in global IQ,erbal quotient, performance quotient, memory variables,nox Cube Test and Letter Cancellation Test, and schoolarks of the two groups were compared using paired t test

r weighted paired t test and, after adjusting for initialifferences, as covariates in a regression model. Details ofhe statistical design and analysis have been reported else-

able 1Q, VQ, PQ, memory, and Knox Cube scores of children who dropped ou

IQ VQ

ohort followed 100.2 � 4.51 99.8 � 4.37rop-outs 98.5 � 3.70 98.3 � 3.44

NS NS

IQ, intelligence quotient; NS, not significant; PQ, performance quotient* Mean � standard deviation. Mean values between groups are not dif

here [20]. t

esults

Results of follow-up data on mental functions from theohort of 680 children ages 6 to 15 y are presented in thiseport. Initially, 814 children who also had been measuredor their heights and weights completed the baseline assess-ent on mental tests.Of 814 children, 206 dropped out (�25%; graduation

rom school, n � 53; migration and missing follow-up, n153). Mean IQ scores of all those who did not com-

lete follow-up in the micronutrient (n � 100) and pla-ebo (n � 106) groups were similar to baseline values ofhe two cohorts (Table 1). Selection of the cohort (n �

able 2ean age, nutrient status, IQ, memory, attention-concentration score,

nd anthropometry of subjects at baseline

ariables Micronutrientgroup

Placebo group

Mean n Mean n

ean age (y) 10.8 421 10.6 393ooled IQ 101.7 421 99.7 393ooled VQ 101.3 421 99.3 393ooled PQ 102.1* 421 100.0 393emory 99.9 421 98.0 393ttention- concentration scoreKnox Cube Test 101.0 421 98.6 393Letter Cancellation Test 100.3 421 98.5 393

chool achievement 62.3 421 57.5 393eight (cm) 137.9 446 136.3 423eight (kg) 29.97 446 29.25 423

nergy 1856 46 1874 44rotein (g) 53.1 46 54.7 44emoglobin (g/L) 117.6 155 117.4 173alcium (mmol/L) 2.46 154 2.48 17.1itamin C (�mol/L) 32.9 28 31.0 72itamin A (�mol/L) 0.72 150 0.88 160yridoxine (AC) 2.82 89 2.80 137iboflavin (AC) 1.55 153 1.51 164itamin B12 (pmol/L) 277.0 112 239.6 122olic acid (nmol/L) 264.8 147 257.1 163

IQ, intelligence quotient; PQ, performance quotient; VQ, verbal quo-ient.

* Mean values between groups were not different, except mean PQcores, which were significantly higher (P � 0.05) within the micronutrient

s scores of the cohort followed up*

Memory Attention-concentration score(Knox Cube Test)

� 4.73 99.5 � 3.26 99.7 � 4.03� 4.33 96.1 � 3.79 99.6 � 2.22

NS NS

erbal quotient.y paired t test.

t versu

PQ

100.698.5

NS

; VQ, v

han within the placebo group.

6mgecpRtabv

5fa(fi

c(fptmhH

I

bnovaMm

M

(

FagI

Fd

Fm(Tt

S29S. Vazir et al. / Nutrition 22 (2006) S26–S32

08) was therefore unbiased. The micronutrient-supple-ented group consisted of 321 children and the placebo

roup consisted of 287 children. Nutrient intake wasssentially similar in both groups and met 80% of thehildren’s calorie requirement [21] (Table 2). Intakes ofrotein, calcium, folate, and vitamin C were above IndianDA values. Limiting nutrients in the diet were iron,

hiamin, niacin, and vitamin A. The prevalence (percent-ge) of children with inadequate biochemical status ataseline (and after supplementation) in the supplementedersus placebo groups was as follows: 54.8 (62.7) versus

ig. 1. Mean IQ, VQ, and PQ scores of the two groups at baseline and finalssessments were not different, except baseline IQ scores of the micronutrientroup were significantly (*P � 0.05) higher than those of the placebo group.Q, intelligence quotient; PQ, performance quotient; VQ, verbal quotient.

ig. 2. Mean increments of intelligence quotient were not significantly

gifferent between groups after adjusting for initial values.

6.6 (69.0) for hemoglobin, 57.1 (0.9) versus 62.5 (3.0)or vitamin C, 98.0 (57.7) versus 100.0 (92.6) for foliccid, 54.7 (1.8) versus 33.1 (3.7) for vitamin A, and 66.09.1) versus 65.9 (35.6) for riboflavin. Details of thesendings have been discussed elsewhere [22].

Mean ages and key anthropometric measurements ofhildren (Table 2) across study groups were comparableinitial mean weight � standard deviation 29.3 � 8.69 kgor the placebo group versus 30.0 � 9.04 kg for the sup-lemented group, initial mean height 136.6 � 13.28 cm forhe placebo group versus 137.9 � 14.14 cm for the supple-ented group). None of the 608 children had a weight and

eight below �3 standard deviations of National Centerealth Statistics standards [21].

ntelligence

There were no significant differences between mean ver-al and performance quotients, but mean IQ of the micro-utrient group was significantly higher (P � 0.05) than thatf the placebo group at baseline. Mean differences in IQ,erbal quotient, and performance quotient between groupst final follow-up were not statistically significant (Fig. 1).ean increments of IQ (Fig. 2) were also comparable in theicronutrient (7.5) and placebo (8.5) groups.

emory

Mean memory scores of the two groups were comparableFig. 3). Moreover, increments across grades between

ig. 3. Mean memory scores were not different between groups; however,ean Knox Cube Test scores at final assessment were significantly higher

**P � 0.01) within the micronutrient group and mean Letter Cancellationest scores at final assessment were significantly higher (*P � 0.05) within

he placebo group.

roups were not significantly different (Fig. 6).

A

nhTKti

Lhm5g

S

e

Fse

Fb

Fb

Fhg

S30 S. Vazir et al. / Nutrition 22 (2006) S26–S32

ttention and concentration

The mean Knox Cube Test score (Fig. 3) within the micro-utrient group at final follow-up (102.2) was significantlyigher (P � 0.01) than that within the placebo group (98.3).he mean increment (final value minus initial value) of thenox Cube Test scores (Fig. 4) was also significantly higher in

he micronutrient group than in the placebo group after adjust-ng for initial differences. The mean final score (Fig. 3) of the

ig. 5. Mean increments of Letter Cancellation Test scores were notignificantly different between groups after adjustment for initial differ-

ig. 4. Mean increments of Knox Cube Test scores were significantlyigher (*P � 0.05) within the micronutrient group than within the placeboroup after adjustment of initial values by regression model.

nces by regression model. g

etter Cancellation Test of the placebo group was significantlyigher than that of the micronutrient group. Although incre-ents on the Letter Cancellation Test were also higher (Figs.and 6) within the placebo group than within the micronutrientroup, the differences were not significant.

cholastic achievement

With regard to school marks of the quarterly and finalxaminations of children in the two study groups, results

ig. 6. Mean increment of memory scores were not significantly differentetween groups after adjusting for initial differences by regression model.

ig. 7. Mean school marks (percentage) were not significantly differentetween groups at quarterly and final examinations. A, micronutrient

roup; B, placebo group.

ilnfi

D

mFpass(peicIia

wgacntsistwpte

ictwmcbqaesvtd[fm

lsopviwtittwabdcifocci

cfl9ttSstfstCLdcwim

trtbshpws

nem

S31S. Vazir et al. / Nutrition 22 (2006) S26–S32

ndicated no significant differences between groups at base-ine and final follow-up (Fig. 7). However, there was aon-significant decrease in marks scored by both groups atnal examination, which was also comparable.

iscussion

The effect of supplementation on a psychological do-ain may not be observable by an assessment of global IQ.or example, a child might be better in numerical ability andoor in language but that child’s IQ may be average or evenbove average. However, the present results indicated noignificant differences in IQ and its subtest scores betweentudy groups. The large increase in IQ points in both groupsabout eight points) could be attributed in part due to theractice effect over time. The micronutrient-enriched bev-rage had little effect on improving IQ even after adjustingnitial differences between groups. It is important in thisontext to note that subjects in both groups had cluster meanQs in the normal range of 99 to 110 and it is difficult toncrease beyond that level through nutritional interventionlone.

However, mean increment scores on the Knox Cube Testere significantly higher within the micronutrient-enrichedroup at the end of the study after adjusting initial valuesnd considering the baseline differences as covariates toontrol for any baseline differences that were present butot significant [20]. Micronutrients have been implicated inhe attention-deficit hypothesis [12], and in the presenttudy, significant differences in increments and the interceptn the regression model are attributed to the effect of theupplement. In contrast, mean increment scores on the Let-er Cancellation Test, after adjusting for initial differences,ere not significant. It has to be borne in mind that thelacebo drink contained some macronutrients and micronu-rients and may have been a source of confounding to somextent.

The present study used a matched-pair, cluster, random-zation design in which the class under each grade was aluster. Individual randomization could have been at-empted, but under the residential school conditions, thisas not practical and could have affected children’s perfor-ance on tests. Because cognitive function depends on the

ombined activity of billions of neurons and numerousiochemical pathways, with their associated enzymes re-uiring several micronutrients, the present study was anttempt to study the effect of providing these micronutri-nts. Whenever the diet offers a wide variety of foods andupplies sufficient protein and calories, it is likely thatitamin and mineral intake will also be adequate. However,here is evidence suggesting that subclinical micronutrienteficiency can disrupt psychological function in children11]. Ten of 13 studies [3–7,10–13,25] that were reviewedound a significant response to vitamin and mineral supple-

entation, with an increase in non-verbal intelligence in at s

east a subsample of the supplemented group. However,ome studies have reported no association with non-verbalr verbal IQ [8,9,23,24]. All studies used double-blind,lacebo-controlled procedures. A selective response to non-erbal tests, which better represented biological function-ng, versus verbal tests was predicted. However, childrenith lower dietary micronutrient intake have been reported

o respond more readily to supplementation than to normalntake [11]. The results of the present study did not confirmhe findings of previous studies on the effect of supplemen-ation on non-verbal IQ. The children in the present studyere from middle-income groups and had above average IQ

t baseline. There were no significant differences in IQetween boys and girls; even after the samples were pooled,ifferences between groups were not significant. The mi-ronutrient-enriched beverage could not further stimulatemprovement in IQ because the children were already per-orming well. In addition, the placebo drink provided 7.3 gf protein, 2 g of fat, 208 kcal of energy, and 176 mg ofalcium per day and its influence on the improvement of theognitive scores in the placebo group over time cannot begnored.

It is also important to note that almost 50% of thehildren in the present study exhibited deficiencies of ribo-avin, pyridoxine, iron, and vitamins A, B12, and C, and8% had folic acid deficiency at baseline. After supplemen-ation, there was significant improvement in the micronu-rient status of the micronutrient-supplemented group [22].ome studies have suggested that response to micronutrientupplementation might reflect on improved attention. Ben-on and Cook [12] tested the attention-deficit hypothesis andound that supplementation increased the time that childrenpent in concentration. Of the two tests that measured at-ention and concentration in the present study, the Knoxube Test is an independent measurement, whereas theetter Cancellation Test is frequently used as an interme-iary diversion between two successive tests. The attention-oncentration results using the Knox Cube Test are in lineith the hypothesis that the micronutrient group would

mprove their attention-concentration scores after supple-entation.In the cited studies [13,25], the investigators indicated

hat “responders” were children who showed increased se-um micronutrient levels and improvement in mental func-ion after supplementation. In this context, iron status haseen associated consistently with cognitive performance ineveral studies [3,6]. Although the iron status indicated byemoglobin in general did not improve in children in theresent study, there was a significant increase in those whoere anemic to start with [22]. The attention-concentration

cores also increased over time in the supplemented group.There was no difference in school marks in the micro-

utrient-enriched and placebo groups at quarterly or finalxaminations. However, a decrease in their respectivearks from the quarterly to the final examination was ob-

erved. This finding may be explained by the fact that the

scBt

C

wecm

A

mINoMd

R

[

[[

[

[

[

[

[

[

[

[

[

[

[

[

[

S32 S. Vazir et al. / Nutrition 22 (2006) S26–S32

yllabi covered for the quarterly examination is much lessompared with the syllabi covered for the final examination.ecause the marks decreased in both groups, this explana-

ion seems most plausible.

onclusion

The results of this study suggest that supplementationith a beverage that is fortified with a range of micronutri-

nts could significantly improve mean incremental scores ofhildren on the attention-concentration test but not IQ,emory, or school achievement scores.

cknowledgments

The authors acknowledge the support and encourage-ent received from Kamala Krishnaswamy, M.D., National

nstitute of Nutrition (ICMR), the guidance received fromadamuni Naidu, M.Sc., National Institute of Nutrition, andur consultants Saroj Arya, Ph.D., National Institute of theentally Handicapped, Hyderabad, India and K. Ravichan-

ra, Ph.D., Osmania University.

eferences

[1] Vazir S, Kashinath K. Influence of ICDS on psychosocial develop-ment of rural school children in Southern India. J Indian Acad ApplPsychol 1999;25:11–24.

[2] Engle PL. Care and child nutrition. Theme paper for the InternationalConference on Nutrition. New York: UNICEF; 1992, p. 1.

[3] Seshadri S, Gopaldas T. Impact of iron supplementation on cognitivefunctions in preschool children: the Indian experience. Am J ClinNutr 1989;50:675–84.

[4] Agarwal, DK, Upadhyay SK, Tripathi AM, Agarwal KN. Nutritionalstatus, physical work capacity and mental function in school children.Nutr Found India 1987;6:1–86.

[5] Pollitt E, Gorman KS, Engle PL, Rivera JA, Martorell R. Nutrition inearly life and the fulfillment of intellectual potential. J Nutr 1995;125:1111S–8.

[6] Pollitt E. Iron deficiency and cognitive function. Annu Rev Nutr

1993;13:521–37.

[7] Benton D. Vitamin/mineral supplementation and non-verbal intelli-gence. Lancet 1988;1:407–9.

[8] Nelson M, Naismith DJ, Burley V, Atenby S, Geddes N. Nutrientintakes vitamin-mineral supplementation in British schoolchildren.Br J Nutr 1990;64:13–22.

[9] Crombie IK, Todman J, McNeill G, Florey D du V. Vitamin/mineralsupplementation and intelligence. Lancet 1990;336:175–6.

10] Benton D, Buts JP. Vitamin/mineral supplementation and intelli-gence. Lancet 1990;335:1158–60.

11] Schoenthaler SJ. Vitamins and IQ. Br Med J 1991;302:1020–1.12] Benton D, Cook R. Vitamin and mineral supplements improve the

intelligence scores and concentration of six-year-old children. PersIndiv Differ 1991;12:1151.

13] Schoenthaler SJ, Amos SP, Doraz WE, Kelly MA, Wakefield J.Controlled trial of vitamin-mineral supplementation: effect on intel-ligence and performance. Pers Indiv Differ 1991;12:1151–8.

14] National Research Council. Recommended dietary allowances. 10thed. Washington, DC: National Academy Press; 1989.

15] Malin AJ. Malin’s intelligence scale for Indian children. Indian JMental Retard 1971;4:15–25.

16] Wechsler D. Wechsler Intelligence Scale for Children. New York:The Psychological Corporation; 1958.

17] Pershad D, Wig NN. PGI Memory Scale. Agra: National Psycholog-ical Corporation; 1990.

18] Wechsler D. WISC-R manual. New York: The Psychological Corpo-ration; 1974.

19] Anastasi A. Psychological testing. New York: MacMillan Company;1976.

20] Sivakumar B, Vijayaraghavan K, Vazir S, Balakrishna N, ShatrugnaV, Sarma KVR, et al. Effect of micronutrient supplement on healthand nutrition status of schoolchildren: study design. Nutrition 2006;22:S1–7.

21] Sarma KVR, Uday Kumar P, Balakrishna N, Vijayaraghavan K,Sivakumar B, Effect of micronutrient supplement on health andnutrition status of schoolchildren: growth and morbidity. Nutrition2006;22:S8–14.

22] Sivakumar B, Madhavan Nair K, Sreeramulu D, Suryanarayana P,Ravinder P, Shatrugna V, et al. Effect of micronutrient supplement onhealth and nutrition status of schoolchildren: biochemical status.Nutrition 2006;22:S15–25.

23] Nelson M, Naismith DJ, Burley V, Gatenby S, Geddes N. Nutrientintakes, vitamin-mineral supplementation, and intelligence in Britishschoolchildren. Br J Nutr 1990;64:13–22.

24] Naismith DJ, Nelson M, Burley VJ, Gatenby SJ. Can children’sintelligence be increased by vitamin and mineral supplements? Lan-cet 1988;2(8606):335.

25] Benton D. Micronutrient supplementation and the intelligence of

children. Neurosci Biobehav Rev 2001;25:297–309.