nutrition and student performance at...

Nutrition and Student Performance at SchoolHoward Taras

ABSTRACT: This article reviews research from published studies on the association between nutrition among school-aged childrenand their performance in school and on tests of cognitive functioning. Each reviewed article is accompanied by a brief description ofits research methodology and outcomes. Articles are separated into 4 categories: food insufficiency, iron deficiency andsupplementation, deficiency and supplementation of micronutrients, and the importance of breakfast. Research shows that childrenwith iron deficiencies sufficient to cause anemia are at a disadvantage academically. Their cognitive performance seems to improvewith iron therapy. A similar association and improvement with therapy is not found with either zinc or iodine deficiency, according tothe reviewed articles. There is no evidence that population-wide vitamin and mineral supplementation will lead to improved academicperformance. Food insufficiency is a serious problem affecting children’s ability to learn, but its relevance to US populations needs tobe better understood. Research indicates that school breakfast programs seem to improve attendance rates and decrease tardiness.Among severely undernourished populations, school breakfast programs seem to improve academic performance and cognitivefunctioning. (J Sch Health. 2005;75(6):199-213)

Parents, educators, and health professionals have longtouted the association between what our children eat

and their school performance. Evidence for this correla-tion is not always apparent, and biases on both sides ofthe argument sometimes override data when this topicis discussed. Understanding existing evidence linking stu-dents’ dietary intake and their ability to learn is a logicalfirst step in developing school food service programs, pol-icies, and curricula on nutrition and in guiding parents ofschool-aged children.

The National Coordinating Committee on SchoolHealth and Safety (NCCSHS) comprises representatives ofseveral federal departments and nongovernmental organi-zations working to develop and enhance coordinatedschool health programs. The NCCSHS has undertaken aproject to enhance awareness of evidence linking childhealth and school performance and identifying gaps in ourknowledge. NCCSHS has conducted a search of peer-reviewed, published research reporting on the relationshipbetween students’ health and academic performance. Inaddition to nutrition, NCCSHS has sponsored research re-views of the association between academic performanceand asthma, diabetes, sickle cell anemia, sleep, obesity,and physical activity.

SELECTION OF ARTICLESArticles meeting the following specific characteristics

were selected. (1) Subjects were school-aged children (5 to18 years), (2) article was published after 1980 in a peer-reviewed journal, and (3) findings included at least 1 ofthe following outcome measures: school attendance, aca-demic achievement, a measure of cognitive ability (such asgeneral intelligence, memory), and attention. Students’level of attention was only acceptable as an outcome mea-sure for purposes of inclusion in this review, if attentionwas measured objectively in the school environment. Stud-

ies of the impact of nutritional intake in children prior toschool age were not included. Studies were identifiedusing MedLine and similar Internet-based searches. If a fullarticle could not be retrieved, but a detailed abstract wasavailable, the research was included. Outcomes other thanacademic achievement, attendance, and cognitive ability,although considered major by the authors, may not bedescribed at all or are only briefly alluded to in the tablesof research descriptions.

LITERATURE REVIEWThe body of literature on child nutrition and its poten-

tial effects on school performance are categorized in 4subject areas: (1) nutritional supplements and micronu-trients, (2) iron deficiency and supplementation, (3) foodinsufficiency, and (4) effect of eating breakfast. This cate-gorization is intrinsically artificial since nutritional prob-lems, chronic hunger, iron deficiency, and poor breakfastpatterns are frequently interrelated. Nevertheless, themajor variable of evaluation, as described by each article’sauthors, naturally led to this classification, with only afew articles falling into more than 1 category.

Nutritional Supplements and MicronutrientsArticles listed in Table 1 examine the relationship

between academic or cognitive performance and vitamin/mineral supplementation, zinc deficiency, and iodinedeficiency. Researchers in the United States and UnitedKingdom have investigated giving multivitamin and min-eral supplements to normal populations of school-agedchildren to determine their effect on intelligence and per-formance on standardized tests or levels of school achieve-ment. Results have varied. Studies indicate there are noeffects of multivitamin supplementation on the intelligenceor performance of most children. Improvements on nonver-bal intelligence of certain subpopulations of children arepossible, but there are no predictable characteristics ofthese subpopulations. Commenting on this body of litera-ture, Smith points out there may be a fundamental problemwith the research. The lack of a theoretic foundation as tohow minerals/vitamins improve intelligence, the smallnumber of researchers in this field, and the uncertaintyabout the potential connection between the vitamin-mineralsupplementation industry and the research further limitwhat can be concluded.1

Howard Taras, MD, Professor, School of Medicine, ([email protected]),Division of Community Pediatrics, University of California, San Diego,Gilman Drive #0927, La Jolla, CA 92093-0927. This article is 1 of 6 ar-ticles that are part of a project of the National Coordinating Committee onSchool Health and Safety (NCCSHS). This NCCSHS project was funded bythe US Department of Health and Human Services, Department of Educa-tion, and US Department of Agriculture. Opinions expressed in this articleare not necessarily shared by these federal agencies or other institutionsthat comprise NCCSHS membership.

Journal of School Health d August 2005, Vol. 75, No. 6 d 199

Zinc naturally occurs in foods such as meats, lentils,rice, potatoes, and grains. Marginal zinc deficiencies occurin low-income groups as well as among those eatingunbalanced diets that are deficient in fat as well as in pro-teins, especially those of animal origin. Zinc is not as wellabsorbed from carbohydrate-rich foods.2 Mild zinc defi-ciency is thought to be more common among endurance

athletes.3 In one Chinese study (Sandstead et al. 1998;Table 1), supplementation with zinc was helpful to improvedperformance only when it was combined with multiplenutrients. The study’s finding may be specific to the diets ofChinese children. Its relevance to school-aged children inthe United States is unknown. Zinc studies in Canada andGuatemala also suggest that zinc supplementation alone is

Table 1Articles That Address Micronutrients and Nutritional Supplements and School

Performance and/or Cognitive Ability* (Continued on next page)

Reference Experimental Design Outcome

van Stuijvenberg ME, Kvalsvig JD,Faber M, Kruger M, Kenoyer DG,Benade AJ. Effects of iron, iodine,and beta-carotene fortified biscuitson the micronutrient status ofprimary school children: arandomized controlled trial.Am J Clin Nutr. 1999;69:497-503.(South Africa)

115 children (age 6-11) were matchedwith a control group (115 children)and for 12 months given biscuitsfortified with iron, iodine, andbeta-carotene. Control childrenreceived nonfortified biscuits.Micronutrient status was assessedas was cognitive function,growth, and morbidity.

In addition to improved blood testsof nutrition, there was a significantbetween-group treatment effect(P < .05) in cognitive function withthe digit span forward task(short-term memory). Fewer schooldays were missed in the interventionthan in the control group becauseof respiratory- (P = .097) anddiarrhea-related (P = .013)illnesses.

Sandstead HH, Penland JG, Alcock NW,et al. Effects of repletion with zincand other micronutrients on neuro-psychologic performance and growthof Chinese children. Am J Clin Nutr.1998;68(suppl 2):470S-475S.(China)

740 children (age 6-9) were givenzinc therapy or zinc and othermicronutrients or micronutrientsalone (randomized) for a period of10 weeks. Tested with CognitionPsychomotor Assessment System.

In the group receiving both zinc andother micronutrients, there was asignificant improvement in fine andgross motor skills, sustainedattention, and capacity for conceptformation and abstract reasoning.

Cavan KR,GibsonRS,Grazioso CF,Isalgue AM, RuzM, SolomonsNW.Growth and body compositionof periurbanGuatemalanchildren in relation to zinc status: alongitudinal zinc intervention trial.Am JClinNutr. 1993;57:344-352.(Guatemala)

In a randomized, double-blindcontrolled study, 162 children(age 7-8) were given zincdaily, for 25 weeks.

No difference on mental concentrationor short-term memory was founddespite changes in body compositionof zinc.

Gibson RS, Vanderkooy PD,MacDonald AC, Goldman A,Ryan BA, Berry M. A growth-limiting,mild zinc-deficiency syndrome insome South Ontario boys with lowheight percentiles. Am J Clin Nutr.1989;49:1266-1273.(Canada)

60 relatively short boys (<15percentile) whose height couldnot be explained by parentalheight were given either zinc or aplacebo for 1 year. All had zinchair measurements and theDetroit Test of Learning Abilities.

Although boys with zinc deficienciesgiven supplementation improved inheight, there was no improvement inattention span or other subscalesof this test.

Bautista A, Barker PA, Dunn JT,Sanchez M, Kaiser DL. The effectsof oral iodized oil on intelligence,thyroid status, and somatic growth inschool-age children from an areaof endemic goitre. Am J Clin Nutr.1982;35:127-134.(Bolivia)

100 children with goiter from a regionwith low iodine in diets weregiven iodized oil or plain mineral oil.Both groups were studied forintelligence (Stanford Binetand Bender tests).

After 22 months, urinary iodine hadincreased in the treated group. Thosewho showed goiter reduction(regardless of treatment or controlgroup) showed improvements inoverall IQ. The effect was strongestin girls. Correcting iodinedeficiency may improve mentalperformance.

* WISC, Wechsler Intelligence Scale for Children; WRAT, Wide Range Achievement Test; WISC-R, Wechsler Intelligence Scalefor Children-Revised.

200 d Journal ofSchool Health d August 2005, Vol. 75, No. 6


Performance and/or Cognitive Ability* (Continued from previous page)


Sankar R, Rai B, Pulger T, et al.Intellectual and motor functions inschool children from severely iodinedeficient region in Sikkim. IndianJ Pediatr. 1994;61:231-236.(India)

Four villages were randomly selectedfrom many known to have highprevalence of severe iodinedeficiency and 90 school childrenselected from these villages.Children were assessed for goiter,acuity of hearing, and signs ofcretinism, and urine iodine levels.Children with cretinism wereexcluded. The Bender Visual MotorGestalt Test, the Binet-Kamal Testfor Mental Ability (IQ), andRaven’s Coloured ProgressiveMatrices were used.

82 of the 90 children had goiter, and 87urinary iodine studies revealedsevere iodine deficiency. Test resultsshowed impairment in language,meaningful and nonmeaningfulmemory, conceptual thinking,nonverbal reasoning, numericalreasoning, motor skills, socialintelligence (from 69% to 91% ofpopulation found to be below normal)and 34% of population with an IQbetween 70 and 84. Correlations ofdegree of iodine deficiency with degreeof neuropsychological impairmentswere not studied.

Benton D, Roberts G. Effects of vitaminand mineral supplementation onintelligence of a sample of schoolchildren. Lancet. 1988;1:140-143.(United Kingdom)

90 school children (age 12-13) kepta dietary diary for 3 days. Amultivitamin/mineral supplement ora placebo was administered doubleblind for 8 months to 60 children; 30received nothing.

Average intake of vitamins was close tothe recommended daily allowance,although for a minority the intake waslow. The recommended daily allow-ance for minerals was rarely achieved.The supplement group, but not theplacebo group or the remaining 30who took no tablets, showed a signifi-cant increase in nonverbal intelligence.

Crombie IK, Todman J, McNeill G,Florey CD, Menzies I, Kennedy RA.Effect of vitamin and mineralsupplementation on verbal andnon-verbal reasoning of schoolchildren.Lancet. 1990;335:744-747.(United Kingdom)

86 school children (age 11-13) wererandomly given either vitamin andmineral supplements or placebo for7 months. A nonverbal performancetest of reasoning was studied.

A small, nonsignificant differencebetween the control and supplemen-tation groups was found in a nonverbaltest. Vitamin and mineral supplementa-tion does not improve the performanceof schoolchildren in tests of reasoning.

Nelson M, Naismith DJ, Burley V,Gatenby S, Geddes N. Nutrientintakes, vitamin-mineralsupplementation, and intelligence inBritish schoolchildren. Br J Nutr.1990;64(1):13-22.(United Kingdom)

227 children (7-12 years of age)consumed food and drinks, andweight was recorded for 7 consecutivedays. Nonverbal and verbal IQ testswere performed. Each child wasrandomly allocated to 1 of 2 groups;matched for age, sex, IQ, and height;and given either vitamin-mineralsupplement or placebo for 28 days.

No correlation was found betweentest scores and micronutrientsconsumed with weighted records.No significant differences werefound in performance betweenthose taking placebo andthose on supplements.

Schoenthaler S, Bier ID, Young K,Nichols D, Jansenns S. The effect ofvitamin-mineral supplementation onthe intelligence of Americanschoolchildren: a randomized,double-blind placebo-controlled trial.J Altern Complement Med. 2000;6(1):19-29, 31-35.(United States)

Vitamin-mineral supplementation(vitamins A, D, E, C, B1-6, B12,iron, folate, zinc, chromium,manganese, molybdenum,selenium, copper at 50% of dailyrecommended allowance) or aplacebo was given at school dailyfor 13 weeks to 245 children(6-12 years of age). NonverbalIQ (from WISC-R) was givenpre- and postintervention.

When studying the data by matchingpairs (placebo-supplement), there wasno difference in nonverbal IQ gain.However, because 24 children takingsupplements had a 16-point higher netgain in IQ than remaining 19 placebocontrols, an increase in overall relativeincrease of 2.5 IQ points in thosereceiving supplements (vs placebo)existed and was statistically significant.A significantly higher number of thosewith supplements gained.15 IQpoints compared to placebo group.


Journal ofSchool Health d August 2005, Vol. 75, No. 6 d 201




Schoenthaler S, Amos SP, Eysenck HJ,Peritz E, Yudkin J. Controlled trialof vitamin-mineral supplementation:effects on intelligence andperformance. Pers Individ Dif. 1991;12(4):351-362.(United States)

615 school children were assigned to1 of 4 treatment groups (placeboand 3 different strengths of vitamin-minerals) for 12 weeks. Pre- andpostmeasures included measureof intelligence (WISC-R), theEysenck Junior PersonalityQuestionnaire, and the MatrixAnalogies Test.

There were significant improvementson the nonverbal WISC-R(3.7 points), but no improvementsfor verbal tests. Supplementationwith 100% of recommended dailyallowances (RDA) was superiorto placebo 50% or 200% of RDA.

Schoenthaler S, Amos SP, Doraz WE,et al. Controlled trial of vitamin-mineralsupplementation on intelligence andbrain function. Pers Individ Dif.1991;2(4):343-350.(United States)

26 delinquent children from apsychiatrically oriented facilitywere given either a placebo ornutrition supplements (consistingof all vitamins and 11 minerals) for13 weeks. Measures included anutritional assessment (7-day foodrecord), blood analysis of vitaminconcentrations, intelligence testing(WISC-R), and brain electricalactivity mapping (computerizedelectroencephalogram/electroencephalogram).

Preintervention showed no differencesin IQ scores of placebo andtreatment groups (IQ 100). At endof 13-week trial, placebo group(11 children) remained the sameand supplement group (15 children)IQ rose to 107. Gains in excessof normal test-retest variation(8 points) were limited to 6 children.

Carlton RM, Ente G, Blum L, Heyman N,Davis W, Ambrosino S. Rationaldosages of nutrients have aprolonged effect on learningdisabilities. Altern Ther. 2000;6(3):85-91.(United States)

19 learning-disabled children (age7-14) were recruited viaadvertisements and then testedbaseline at end of study:WISC, WRAT 6 additionalstandardized measures ofintelligence and/or achievement,blood and urine tests (chemistry,blood cell tests). Class grades andany level of mainstreaming were alsomeasured outcomes. In Year 1, allchildren were given 1 or morenutrition supplements (vitamin B,zinc, magnesium, etc.) at 2 to 10times recommended dailyallowance—only investigators knewwhich child received which nutrient.Child required evidence of academicimprovement during Year 1 (highergrades, increased mainstreaming) toremain in study. Year 2 consistedof a closed random trial of 4 rotations(2 months per rotation) with childreceiving either active nutrientsupplements or placebo. In Years 3and 4, all children received onlynutrition supplementation (nointermittent placebos) and allintakes became open labeled toinvestigators families andstudents.

Intelligence tests did not changesignificantly after Year 1 or Year 2,but all 19 improved academicallyor were mainstreamed after Year 1.Twelve children completed theclosed trial at Year 2. All hadimprovements in grade scores.During Year 2, no differencesexisted between months onplacebo versus months on nutrients.All students continued to improveacademically during Year 3, on or offthe nutrients. During Year 4, thoseremaining on nutrients furtherincreased grade averages, whilethose discontinuing nutrients hadlower grade averages. Authorsconcluded that (a) nutrientsupplementation improvesachievement, not intelligence, and(b) longer than 2 months off ofsupplements is necessary to losesupplements’ beneficial effect.


202 d Journal of School Health d August 2005, Vol. 75, No. 6

Table 2Articles That Address Iron Deficiency in School-Aged Children and School



van Stuijvenberg ME, Kvalsvig JD,Faber M, Kruger M, Kenoyer DG,Benade AJ. Effects of iron, iodine, andbeta-carotene fortified biscuits on themicronutrient status of primary schoolchildren: a randomized controlled trial.Am J Clin Nutr. 1999;69:497-503.(South Africa)

115 children (age 6-11) were matchedwith a control group (115 children)and for 12 months given biscuitsfortified with iron, iodine, andbeta-carotene. Control childrenreceived nonfortified biscuits.Micronutrient status was assessedas was cognitive function,growth, and morbidity.

In addition to improved blood tests ofnutrition, there was a significantbetween-group treatmenteffect (P < .05) in cognitive functionwith the digit span forward task(short-term memory). Fewer schooldays were missed in the interventionthan in the control group becauseof respiratory- (P = .097) anddiarrhea-related (P = .013)illnesses.

Walker SP, Grantham-McGregor SM,Himes JH, Williams S, Duff EM.School performance in adolescentJamaican girls: associations withhealth, social and behaviouralcharacteristics, and risk factorsfor dropouts. J Adolesc Health.1998;21:109-122.(Jamaica)

452 girls (age 13-14) were assessedfor school performance (WRAT),nutritional status (height/weight forage, hemoglobin), and healthbehaviors (questionnaire onhunger, sexual activity, violence).

Girls who were anemic were significantlymore likely to have poor academicachievement (multiple regressionanalysis). Poor school attendance,early sexual activity, and not livingwith both parents were alsoassociated with school dropoutin the subsequent year.

Halterman JS, Kaczorowski JM, AligneCA, Auinger P, Szilagyi PG. Iron defi-ciency and cognitive achievementamong school-aged children andadolescents in the United States.Pediatrics. 2001;107(6):1381-1386.(United States)

Data were taken from National Healthand Nutrition Examination Survey III(6-16 years); blood samples wereavailable to determine iron deficiencyin 3% of 5398 children. Standardizedtest scores for intelligence (WISC-2subtests) and achievement (WRAT-reading and math) are available.

Average math scores (scoreapproximately 87) lower for childrenwith iron deficiency with and withoutanemia compared with children withnormal iron status (score 94).Children with iron deficiency hadmore than twice risk of scoring belowaverage in math than those withnormal iron. No difference wasfound on other test scores foriron deficiency.

Hutchinson SE, Powell CA, Walker SP,Chang SM, Grantham-McGregor SM.Nutrition, anaemia, geohelminth infec-tion, and school achievement in ruralJamaican primary school children.Eur J Clin Nutr. 1997;51(11):729-735.(Jamaica)

809 fifth graders (age 9-13) wererandomly selected from 16 schools.WRAT achievement test, attendancerecords, a questionnaire todetermine socioeconomic status,height, weight, blood test for anemia(hemoglobin), and stool specimenswere measurements.

School achievement (arithmetic) wasassociated with height for age.School achievement was notassociated with body mass index(measure of relative weight). Anemicchildren and those with parasitesin their stool had lower attendanceand achievement scores (readingand spelling).

Lynn R, Harland P. A positive effect ofiron supplementation on the IQs ofiron-deficient children. Pers Individ Dif.1998;24:883-885.(United Kingdom)

208 children (12-16 years of age) werematched with 205 controls for age,sex, and IQ. Treated group receivediron and ascorbic acid for 16 weeks.The Ravens Progressive Matrixestest was used to measure IQ.

An initial correlation betweenhemoglobin levels and IQ wasfound to be small but significant.There was no significant differencebetween treatment and control groups,excepting for those with low ferritin(measure of iron). In this group,the treated group gained 3 IQ points,whereas the placebo group lost2 IQ points.

* WISC, Wechsler Intelligence Scale for Children; WISC-R, Wechsler Intelligence Scale for Children-Revised; WRAT, WideRange Achievement Test.





Abalkhail B, Shawky S. Prevalence ofdaily breakfast intake, iron-deficiencyanaemia, and awareness of beinganaemic among Saudi school students.Int J Food Sci Nutr. 2002;53:519-528.(Saudi Arabia)

800 students were selected for bloodtests for anemia (based onhemoglobin levels). Of these,42 boys and 42 girls (age 9-21years, mean 14) were surveyed fornutritional habits.

Anemia (20% of population) was mostprevalent among those greater than age12. Menstruating girls had twice the rateof anemia than males. Anemia wassignificantly more prevalent amongthose who failed school performanceexaminations. Of those surveyed,skipping breakfast was reported by 15%(same for all ages, sex, body massindex, and social classes). Skippingbreakfast was significantly associatedwith poor school performance.

Otero GA, Aguirre DM, Porcayo R,Fernandez T. Psychologicaland electroencephalographic studyin school children with irondeficiency. Int J Neurosci. 1999;99:113-121.(Mexico)

33 children (6-12 years of age) whowere iron deficient but not anemicand 33 controls with normal iron(matched for socioeconomicvariables) were given psychometrictests (WISC-R for IQ and a computer-ized test of learning, dynamicevaluation of learning as well as anelectroencephalogram (EEG).

Iron-deficient children had lower values inWISC items of information, comprehen-sion, and verbal performance, andfull-scale IQ scores than children withnormal iron levels. EEGs showedslower activity among the iron-deficientchildren as well, suggesting develop-mental lag and/or a central nervoussystem dysfunction.

Pollitt E. Iron deficiency andeducational deficiency. Nutr Rev.1997;55(4):133-141.(Indonesia, Thailand,Egypt, Guatemala)

This article reviewed 4 studies.Study 1: 41 children without anemiaand 78 with iron-deficiency anemia(8.2-13 years of age) were given eitheriron supplements or placebo at school.Cognitive tests (Raven test of nonverbalintelligence), Bourden-Wisconsin test forconcentration, and another standardizedtest of achievement were administered.Study 2: 1358 children were classified asiron-deficiency anemia (101), iron-depleted anemia (47), or iron repleted(1210). All children then received eitheriron supplements for 16 weeks or pla-cebo. IQ, language, and math tests weregiven at baseline and after treatment.Study 3: 28 children with iron deficiencyand 40 nonanemic children (age 8-11)(anemia was determined by response toiron therapy) had measures of hemoglo-bin, various iron studies, a cognitive-performance test (CPT), PeabodyPicture Vocabulary test, and MatchingFamiliar Figure Test. Iron treatment wasgiven. Study 4: 1203 children and adultshad iron determinations and weredivided into iron replete, iron deficiencybut no anemia, and iron-deficiencyanemia.Arithmetic, reading,othercomplexcognitive processes, simple reaction time,short-term memory, and other elementarycognitive processes were measured.

Study 1: No significant differences werefound on Raven IQ tests betweennonanemic and anemic students atbaseline, but anemic children performedsignificantly poorer in achievement teststhan nonanemic. Those who receivediron improved their hemoglobin countsand improved significantly more onachievement tests than those receivingplacebo. Study 2: At baseline the meanIQ was 94 (iron-repleted children), 91(iron-deficiency anemia). Those withjust iron depletion had IQs that did notdiffer significantly from either group. The3 groups differed significantly forlanguage scores, favoring those withmore iron stores. Iron treatment had nodifference on IQ, math, or languagescores. Study 3: At baseline, no differ-ences were found between iron-deficiency anemia and nonanemicchildren in CPT and Peabody tests, butmore errors on matching test amongthose with anemia. No major differenceswere found on these tests after treat-ment except that among anemicchildren, those treated with iron selectedinformation faster and with fewer errors.Study 4: Iron-depleted and iron-repletedgroups were similar in all but 1 test:faster responses in memory for ironrepleted.






Seshadri S, Gopaldas T. Impact of ironsupplementation on cognitivefunctions in preschool andschool-aged children: the Indianexperience. Am J Clin Nutr.1989;50:675-686.(India)

Four studies. Study 1: The studyconsisted of 94 children (age 5-8)of whom one third were randomlyassigned to control; Experimental groupwas given oral iron and folic acid for 60days, and pre- and posttested withblood tests (for anemia) and intelligence(WISC). Study 2: 14 children (age 5-6)with normal intelligence, weight for age.60% of median value, and blood workdemonstrating iron-deficiency anemiawere matched with 14 controls (sameintelligence, weight, and blood work).All children were treated withantiparasitic agents (parasites oftencause anemia), but only experimentalgroup was given iron and folic acidsupplementation. Study 3: 16 subjectswere matched for age, hemoglobinlevels, and baseline scores with 2 othergroups of 16 each. Double blinded, 1group received a placebo, another 30mg of iron, and the third group received40 mg of iron. Visual-recall tests, digitspan, maze cognitive function test, andtests of visual discrimination andperception were given preinitiation and4 months postinitiation of therapy. Study4: 65 pairs of subjects were matched forage, hemoglobin, and total cognitivefunction scores and 1 group receivediron, the other placebo. Cognitive testsgiven were same as Study 3.

Study 1: Anemic children had lower testscores at beginning of experiment.Nutrition therapy improved IQ/WISCscores for both anemic (11 points) andnonanemic students (4 points) for chil-dren aged 7 and 8 years. Study 2: Bloodtests for anemia (hemoglobin levels)improved for treatment group andworsened for controls. Both groupsshowed statistically significant improve-ment in IQ scores (WISC), but higherincreases for treatment (10 pointsverbal, 17 performance) than controls(5 points verbal, 7 performance). Irontherapy improved cognitiveperformance among anemic children.Study 3: Overall cognitive functionscores improved with either dose of iron,except maze test for the lower dose.The 40-mg group was significantlybetter than control, indicating higherdose had better impact on cognition.Study 4: On supplementation for 1school year, anemic, iron-supplementedgroup did significantly better on allcognitive tests than nontreated group.Nonanemic group only improved onmaze test, as compared to controls.

Sever Y, Ashkenazi A, Tyano S,Weizman A. Iron treatmentin children with attentiondeficit hyperactivity disorder.A preliminary report.Neuropsychobiology.1997;35:178-180.(Israel)

14 nonanemic boys (age 7-11) withattention deficit hyperactivity disorder(ADHD) were assessed by parent andteacher for symptoms of ADHD usinga standardized questionnaire (Conners)and for anemia (blood test) before and30 days after the intervention. All chil-dren received iron supplements daily.

Iron measures increased, and parents’scores for attention problems improved,but teachers’ scores remainedunchanged after iron therapy.

Sungthong R, Mo-suwan L,Chongsuvivatwong V. Effects ofhaemoglobin and serum ferritin oncognitive function in school children.Asia Pac J Clin Nutr. 2002;11(2):117-122.(Thailand)

Two schools thought to have highanemia-risk students were chosen.Children (427, average age 9.6) weretested for cognitive function (Test ofNonverbal Intelligence), hemoglobin,iron, and height/weight. Language andmath scores were collected.

Cognitive function increased withincreased hemoglobin concentration inchildren with iron deficiency but did notchange with hemoglobin concentrationin children with normal iron stores.Those with iron-deficiency anemiaconsistently had poorest cognitive func-tion (IQ 75) and below-average math/language achievement. Those with irondeficiency but not anemia had highercognitive function (IQ 86) and above-average math/language achievement.



unlikely to change cognitive ability or school performance,even in children with zinc deficiencies (Gibson et al. 1989,Cavan et al. 1993; Table 1).

Iodine deficiency in a population causes increased prev-alence of goiter and, more importantly, may increase therisk for intellectual deficiency in that population. Iodinedeficiency has been demonstrated to affect children’s intel-ligence and performance. Of the studies that have lookedat this issue, most were conducted outside the UnitedStates and in regions where iodine deficiency is endemic.Urinary iodine studied in large national samples (NationalHealth and Nutrition Examination Survey [NHANES]I and III) indicates that there is adequate iodine intake forthe overall US population, but that urinary iodine levels aredropping over time. A trend toward iodine deficiency mustbe monitored in school-aged populations.4 Iodine replace-ment through dietary supplementation appears to at leastpartially reverse cognitive adverse effects among foreignpopulations of children who are severely iodine deficient.

Iron Deficiency and SupplementationTable 2, which includes 12 articles, describes the results

of 18 studies. Knowledge regarding the link between irondeficiency and school achievement or cognitive ability isderived primarily from observations and evaluated inter-ventions of populations of youth who live outside theUnited States. Two studies were found that are based onUS populations of children.

Some studies make an important distinction betweenchildren who have iron deficiency but are not anemic andchildren with iron deficiency who have iron-deficiencyanemia. In the first group, children with iron deficiencyhave normal hemoglobin values (ie, not anemic) probablybecause the severity or duration of their iron deficiency isinadequate to cause anemia. National large sample data(NHANES III, 2001) indicate that 3% of school childrenin the United States have iron deficiency and that iron de-ficiency may still be relatively common in adolescentgirls (9% of adolescent girls of childbearing age).5 Studiesin Table 2 demonstrate that iron deficiency may be associ-ated with poorer academic outcomes (lower scores on

achievement were found for math only). Certainly forthose school-aged children with iron-deficiency anemia,a preponderance of evidence demonstrates an associationwith poorer cognition and/or lower academic achievement.However, it cannot yet be concluded that studentswith iron deficiency but no anemia are at greater riskfor diminished cognitive ability and/or poor academicperformance.

There is a need for large-scale studies that employcarefully matched socioeconomic control groups and mul-tiple standardized tests of cognition and academic abilityto better elucidate subtle effects of iron deficiency and tomore precisely define the effects on learning. It is encour-aging that when iron supplements are given to school-agedchildren as therapy for depleted stores, academic perfor-mance seems to improve. On the other hand, iron supple-ments do not appear to benefit those whose iron stores arealready normal.

School districts distribute medical forms to familiesthat elicit information on students’ health status. Someforms include inquiries for students’ hemoglobin or hem-atrocrit (blood tests that screen for anemia), and they callattention to the association between anemia and schoolperformance. The American Academy of Pediatrics rec-ommends that blood tests for iron-deficiency anemia onlyneed to be performed routinely if the student is a menstru-ating adolescent female or history reveals that the child isat risk for anemia.6 This recommendation is based on thebelief that a good medical history of children’s dietaryintake and/or blood loss is sufficient to identify those likelyto be anemic.

Evidence listed in Table 2 highlights the importance ofassessing school-aged children for iron-deficiency anemia.In any geographic region of the world with a high preva-lence of iron deficiency, public health programs that effec-tively prevent, detect, and treat iron deficiency likely helpto improve the academic performance of their school-agedpopulations.

Food InsufficiencyTable 3 lists published research studying whether food

insufficiency (ie, the limited availability of nutritionally




Bruner AB, Joffe A, Duggan AK,Casella JF, Brandt J. Randomizedstudy of cognitive effects of ironsupplementation in non-anaemic,iron-deficient adolescent girls. Lancet.1996;348:1789-1792.(United States)

98 adolescent girls (13-18 years) found tohave nonanemic iron deficiency wererandomly assigned to iron therapy orplacebo for 8 weeks. Study measuredBrief Test of Attention, Symbol DigitsModalities, Visual Search and Attention,and Hopkins Verbal Learning Test.

Level of ferritin (iron deficiency) was notcorrelated with results of girls’performance tests prior to intervention.Treatment group improved significantlymore than placebo group in 3 free recallitems, but not in other performancemeasures.



Table 3Articles That Address Food Insufficiency in School-Aged Children and School



Hall A, Khanh LN, Son TH, et al.An association between chronicundernutrition and educational testscores in Vietnamese children. EurJ Clin Nutr. 2001;55(9):801-804.(Vietnam)

Measured test result of 3055third-grade school children inmathematics and Vietnameselanguage, height for age, weightfor age, and weight for height.

Low test scores in both math andlanguage were significantlyassociated with low height for age,weight for age, but not weight forheight. Chronic undernutrition affectseducational achievement.

Ivanovic D, Vasquez M, Aguayo M,Ballester D, Marambio M, Zacarias I.Nutrition and education. III.Educational achievement and foodhabits of Chilean elementary and highschool graduates. Arch LatinoamNutr. 1992;42(1):9-14.(Chile)

550 school children (13-19 years of age)were chosen from variouscommunities. Food habits weredetermined through a questionnaireby trained interviewers. Educationalachievement was determined bytest results of various academic teststaken routinely at these ages. Multipleregression between academicachievement (dependent variable)and consumption of different foods.

Achievement was significantly andpositively correlated with frequency ofconsumption of dairy, meats, andeggs in both elementary and highschools. Significant and inversecorrelations were found betweenacademic achievement andconsumption of fruits and vegetables.Food habits explained 24%(elementary) and 17% (high school)of variance in academic achievement.Dairy products had greatestindependent influence on achievement.

Ivanovic D, Olivares MG, Castro CG,Ivanovic RM. Nutrition and learningin Chilean school-aged children:Child’s Metropolitan Region Survey1986-1987. Nutrition. 1996;12(5):321-328.(Chile)

More than 523,000 students in first,second, fourth, and eighth gradesand 2 grades of high school werestudied. Scholastic achievementwas measured using a previouslypiloted language and mathematicstest. This was compared to headcircumference for age, weight forage, and height for age.

Head circumference for age is the mostimportant anthropometric parameterassociated with scholasticachievement. Signs of chronicundernutrition are associated witheducational achievement.

Murphy JM, Wehler CA, Pagano ME,Little M, Kleinman RE, Jellinek MS.Relationship between hungerand psychosocial functioningin low-income American children.J Am Acad Child AdolescPsychiatry. 1998;37:163-170.(United States)

204 children in grades 3-8 (and theirparents) were interviewed. Ninety-sixwere reinterviewed 4 months aftera free breakfast program was initiated.Assessments included psychosocial,academic, and food sufficiency/hungermeasures.

Hungry and at risk for being hungrychildren were twice as likely to haveimpaired functioning (by parent orchild report). Teachers reportedhigher levels of hyperactivity,absenteeism, and tardiness amonghungry/at-risk than among nonhungrychildren. Intermittent experiences offood insufficiency are associated withpoor academic functioning inlow-income children.

Alaimo K, Olson CM, Frongillo EA Jr.Food insufficiency and Americanschool-aged children’scognitive, academic, andpsychosocial development.Pediatrics. 2001;108(1):44-53.(United States)

Data on 5344 children (age 6-16years) from a large database(NHANES III) were studied forfamilies who reported not gettingenough food to eat as comparedto those families who were foodsufficient. Regression analysis wasconducted to test this foodinsufficiency against cognitive(WISC-R), academic (WRATsubtests), school attendance,suspensions, and other measures.

Regression analysis (and adjustment forconfounding socioeconomic and otherfactors) showed that foodinsufficiency was significantlyassociated with WRAT arithmeticscores (0.4 points lower), graderepetition (1.44 times as likely), butnot associated significantly withcognitive outcomes, reading scores.It did not appear that diminishedhealth status was the cause of foodinsufficiency’s effect on theseoutcomes. Effect was found onchildren aged 6-11 years, not onthose aged 12-16 years.

* NHANES, National Health and Nutrition Examination Survey; WISC, Wechsler Intelligence Scale for Children; WRAT, WideRange Achievement Test.


adequate and safe foods) affects cognition in school-agedchildren. Of 10 articles found, 7 were conducted on popula-tions outside the United States. The majority describeimpoverished school-aged populations of foreign countries,where food scarcity was sufficiently severe to alter chil-dren’s height-weight ratios. Effects on cognitive ability

and/or reported school achievement were found in the for-eign studies. In some of the studied populations, povertyand parasites confounded the effects of undernutrition.

In 2 of the 3 studies conducted in the United States,food insufficiency was associated with significantly poorercognitive functioning, decreased school attendance, or

Table 3Articles That Address Food Insufficiency in School-Aged Children and School



Upadhyay SK, Agarwal KN,Agarwal DK. Influence ofmalnutrition on social maturity, visualmotor coordination and memory inrural school children [abstract].Indian J Med Res 1989;90:320-327.[Abstract](India)

1336 rural school children (age 6-8)were evaluated for malnutrition(grades I, II, III) and for visualmotor coordination andmemory function.

Increasingly higher grades ofmalnutrition were associated withincreasingly lower scores on socialcompetence. Memory scores wereweakly associated with malnutrition,with some differences between boysand girls on memory for picture andmemory for words.

Walker SP, Grantham-McGregor SM,Himes JH, Williams S, Duff EM.School performance in adolescentJamaican girls: associations withhealth, social, and behaviouralcharacteristics, and risk factorsfor dropouts. J AdolescHealth. 1998;21:109-122.(Jamaica)

452 girls (age 13-14) wereassessed for school performance(WRAT), nutritional status (height/weight for age, hemoglobin), andhealth behaviors (questionnaire onhunger, sexual activity, violence).

Girls who were sexually active, anemic,or aggressive were significantly morelikely to have poor academicachievement (multiple regressionanalysis). Poor attendance, earlysexual activity, and not living withboth parents were associated withschool dropout in thesubsequent year.

Hutchinson SE, Powell CA, Walker SP,Chang SM, Grantham-McGregor SM.Nutrition, anaemia, geohelminthinfection, and school achievementin rural Jamaican primary schoolchildren. Eur J Clin Nutr.1997;51(11):729-735.(Jamaica)

809 fifth graders (age 9-13) wererandomly selected from 16 schools.WRAT achievement test, attendancerecords, a questionnaire todetermine socioeconomic status,height, weight, blood test foranemia (hemoglobin), and stoolspecimens were measurements.

School achievement (arithmetic) wasassociated with height for age.School achievement was notassociated with body mass index(measure of relative weight).Anemic children and those withparasites in their stool had lowerattendance and achievementscores (reading and spelling).

Kleinman RE, Murphy JM, Little M,et al. Hunger in children in theUnited States: potential behavioraland emotional correlates.Pediatrics. 1998;101(1):e3(United States)

328 families (parent with at least 1 childyounger than 12 years of age) wereidentified from a large CommunityChildhood Hunger IdentificationProject database. Parents completedquestionnaires (Pediatric SymptomsChecklist, standard food insufficiencyquestions, and questions onrepeating a grade and receivingspecial education).

Of this sample, 56 were describedas hungry, 157 described as atrisk, and 108 described as nothungry. Hungry children had higherscores than at-risk and not-hungrychildren for school problems andattention problems, but the findingsdid not reach statistical significance.Aggression and anxiety had thestrongest (and statically significant)degrees of association withexperiences of hunger.

Mukudi E. Nutrition status, educationparticipation, and school achievementamong Kenyan middle-schoolchildren. Nutrition. 2003;19:612-616.(Kenya)

Data collected from 851 students from5 elementary schools included height,weight, attendance records, andraw scores from standardized tests.

29% of students fell below 90% cutoffof acceptable weight for height(described as a measure of nutrition).The measure of "weight for height"was the second strongest predictor ofachievement on standardized tests(school attendance was the strongestpredictor).

* NHANES, National Health and Nutrition Examination Survey; WISC, Wechsler Intelligence Scale for Children; WRAT, WideRange Achievement Test.


Table 4Articles That Address Role of Breakfast in School Performance and/or Cognitive Ability*

(Continued on next page)

Reference Research Design Outcome

Abalkhail B, Shawky S. Prevalence ofdaily breakfast intake, iron deficiencyanaemia, and awareness of beinganaemic among Saudi schoolstudents. Int J Food Sci Nutr.2002;53:519-528.(Saudi Arabia)

800 students were selected for bloodtests for anemia (based on hemoglo-bin levels). Of these, 42 boys and 42girls (age 9-21 years, mean 14) weresurveyed for nutritional habits.

Of those surveyed, skippingbreakfast was reported by 15%of girls of all ages, sex, body massindex, and social classes. A signifi-cant association was foundbetween skipping breakfastand poor schoolperformance.

Chandler AK, Walker SP, Connolly K,Grantham-McGregor SM. Schoolbreakfast improves verbal fluency inundernourished Jamaican children.J Nutr. 1995;125:894-900.(Jamaica)

97 undernourished (low weight for age)and 100 adequately nourishedchildren (third and fourth grades)were given either a full breakfastor a quarter of an orange. A batteryof 4 cognitive tests were given.After a few weeks, the treatmentswere reversed and testsrepeated.

Undernourished children’s performanceimproved significantly on a test ofverbal fluency when they receivedbreakfast. Adequately nourishedchildren did not change. Other testscores did not change as a result ofbreakfast.

Cromer BA, Tarnowski KJ, Stein AM,Harton P, Thorton DJ. The schoolbreakfast program and cognition inadolescents. J Dev Behav Pediatr.1990;11:295-300.(United States)

34 adolescents (mean age 14.2) wereselected, of which 18 were randomlyassigned to get breakfast (doughnut,chocolate milk, orange juice); othersconsumed a low-energy breakfast(diet gelatin, powdered milk). Allwere given cognitive tests 4 hoursafter breakfast (Peabody PictureVocabulary, auditory-verbal learning,matching figure test, continuousperformance test, an anxietyinventory).

Cognitive tests did not discriminatebetween breakfast and low-energybreakfasts.

Edward HG, Evers S. Benefits andbarriers associated with participationin food programs in three, low-income Ontario communities. Can JDiet Pract Res. 2001;62:76-81.(Canada)

Focus groups were held afterinitiating a food program atschools in 3 communities.

Parents and teachers of childrenattending a ‘‘breakfast club’’ notedimprovements in behavior atschool: teachers felt childrenwho had eaten breakfast werebetter inclined to study and listenbetter, concentrate better. Parentsfelt the children were absentless often.

Grantham-McGregor SM, Chang S,Walker SP. Evaluation of schoolfeeding programs: some Jamaicanexamples. Am J Clin Nutr.1998;67(suppl):785S-789S.(Jamaica)

100 undernourished children (based onweight for age) aged 8-11 years in 4schools were matched for school,class, and sex with a child withnormal weight for age. Childrenwere randomly assigned to receivebreakfast or a slice of orange. Thiscontinued for 1 week. Two weekslater, the breakfast and nonbreakfastgroups were reversed. A battery ofcognitive function tests were givenand classroom behavior wasobserved.

Nutritional status had no significanteffect on the cognitive function test.Undernourished children performedbetter after they had receivedbreakfast, whereas adequatelynourished children showed nochange in scores. Children’sattention to task in schoolincreased significantly withbreakfast, and in 1 school, grossmotor behavior declined.Authors postulate that the quality ofa school may modulate theeffects of breakfast andno breakfast.

* WISC, Wechsler Intelligence Scale for Children; WRAT, Wide Range Achievement Test.

Journal of School Health d August 2005, Vol. 75, No. 6 d 209


(Continued from previous page)


Jacoby ER, Cueto S, Pollitt E. Whenscience and politics listen to eachother: good prospects from a newschool breakfast program in Peru.Am J Clin Nutr. 1998;67(suppl):795S-797S.(Peru)

More than 500,000 school childrenreceived school breakfast under thisprogram, but the numbers from 1town used for this study were notprovided in this published paper.

School breakfast significantly increasedenergy intake, protein, and iron.School attendance rates improvedsubstantially (not defined).Nutritionally at-risk children in theprogram improved their performancein the short term (not defined) on avocabulary test (1 of 4 tests given).

Lopez I, de Andraca I, Perales CG,Heresi E, Castillo M, Colombo M.Breakfast omission and cognitiveperformance of normal, wasted, andstunted school children. Eur J ClinNutr. 1993;47:533-542.(Chile)

279 children (age 8-10) from lowsocioeconomic backgrounds werecategorized as normal, wasted, orstunted. Children were randomized asreceiving breakfast and others ina fasting situation. Childrenwere given3 cognitive tasks on a microcomputer.

Stunted children performed poorer thanothers in attention test. No associa-tions were found between consumingbreakfast and children’s short-termvisual memory, problem solving, orattention tasks in normal, wasted, orstunted children.

Meyers AF, Sampson AE, Weitzman M,Rogers BL, Kayne H. Schoolbreakfast program and schoolperformance. AJDC. 1989;143:1234-1239.(United States)

More than 1000 students in third throughsixth grades attending 6 elementaryschools were selected (in a schooldistrict with a new school breakfastprogram). The school tested all gradesin the spring using the ComprehensiveTest of Basic Skills (CTBS); tests weregiven in themornings. Attendance andtardiness data were collected fromdistrict administrative records. Datawere collected for the year before andyear after implementation of the schoolbreakfast program.

Participation in the school breakfastprogram contributed significantly tohigher CTBS scores (5 points on thetotal score) and to lower tardinessand absence rates.

Murphy JM, Pagano ME, Nachmani J,Sperling P, Kane S, Kleinman RE.The relationship of school breakfastto psychosocial and academicfunctioning: cross-sectional andlongitudinal observations in aninner-city school sample.Arch Pediatr Adolesc Med.1998;152:899-907.(United States)

In 3 elementary schools, parents andstudents (133 in grades 3 and higher)were interviewed and school datawere collected both prior to and afterimplementation of a school breakfastprogram.

Students who increased their participa-tion in the school breakfast programhad significantly greater increases intheir math grades (10.3 grade),decreases in rates of school absen-ces (�0.1 days), and decreases intardiness (�0.4 days). Students whodecreased their participation in schoolbreakfast had lower math grades(�0.9 of a grade), higher absences(11.6 days), and higher tardiness(10.9 days).

Kleinman RE, Hall S, Green H, et al.Diet, breakfast, and academicperformance in children. Ann NutrMetab. 2002;46(suppl 1):24-30.(United States)

Information was gathered from 97students (grades 4-6) prior to start ofa school breakfast program and afterthe program had been in place for6 months. Validated dietary recallinstruments were used, as werehunger indices and other symptomchecklists. Students’ daily atten-dance records and participation inbreakfast program were collected.

Prior to the school breakfast program,study children classified fromquestionnaires as being nutritionallyat risk had significantly poorerattendance, punctuality, and gradesat school. Six months after the freeschool breakfast program, thosestudents who decreased theirnutrition risk showed significantlygreater improvements in attendanceand improvements in math gradesthan children who did not decreasetheir nutritional risk.






Pollitt E, Leibel RL, Greenfield D.Brief fasting, stress, and cognitionin children. Am J Clin Nutr. 1981;34:1526-1533.(United States)

32 children (age 9-11) ate dinner andslept at a clinical research center for2 nights (1 week apart). Each wasrandomly assigned to either break-fast in the morning or no breakfast(reversed at next night’s stay).Blinded psychologists administeredmemory, matching, attention, andother standardized tests. Bloodsamples were tested for beta-hydroxybutyrate glucose, lactatefatty acids.

Interaction between having breakfastand IQ score accounted for asignificant portion of variance inerrors on a test of competence indiscrimination of similar visualstimuli. As measured glucoseconcentrations dropped, the numberof errors increased. Correct recall ofincidental objects in the memory testwas better among those who did nothave breakfast.

Pollitt E, Lewis NL, Garza C,Schulman RJ. Fasting andcognitive function. J PsychiatrRes. 1982-1983;17:169-174.(United States)

39 children (age 9-11) ate dinner andslept at a clinical research center for2 nights (1 week apart). Each wasrandomly assigned to eitherbreakfast or no breakfast (reversedat next night’s stay). Blindedpsychologists administered memory,matching, attention, and otherstandardized tests. Blood sampleswere tested for glucose and insulin.

Errors on a test of competence indiscrimination of similar visual stimuliwere greater among those withoutbreakfast. Correct recall of incidentalobjects in the memory testwas better among those whodid not have breakfast.

Pollitt E, Cueto S, Jacoby ER. Fastingand cognition in well- andunder-nourished school children:a review of three experimentalstudies. Am J Clin Nutr.1998;67(suppl):779S-784S.(Peru)

23 undernourished and 29 well-nourished boys (age 9-11) enrolled infourth or fifth grade were chosenbased on weight or height as a signof nourishment. They were eithergiven breakfast or no breakfast(sugar-free soda). Blood samplesand a battery of psychological testswere given (Number Discrimination,Peabody Picture Vocabulary Test,Raven Progressive Matrices, andcomputer tests of reaction time,stimulus discrimination, and memory).

Only performance on a memory searchtest (speed of scanning memory)was slower in the ‘‘no breakfast’’ thanin the ‘‘breakfast’’ group forundernourished boys. Nutritionallyat-risk boys were vulnerable toadverse effects of fasting ascompared to well-nourished children.

Powell C, Grantham-McGregor S,Elston M. An evaluation of giving theJamaican government school mealto a class of children. Hum Nutr ClinNutr. 1983;37C:381-388.(Jamaica)

Seventh-grade children (all in bottomone third of scholastic ability) weredivided into 3 groups, those receiving(1) a school lunch (milk and cake ormeat-filled pastry), (2) a syrup drink,and (3) nothing. Students weretracked for 2 semesters: baselinemeasured during first and interven-tion occurring during second. Schoolachievement was measured usingthe Wide Range Achievement Test.

Those receiving a school mealperformed better on arithmetictest and had better attendance.Provision of a meal in school wasdetermined to attract some studentsto attend. When attendancewas controlled, arithmeticscores were still higheramong those who hadreceived a meal.

Powell CA, Walker SP, Chang SM,Grantham-McGregor SM. Nutritionand education: a randomized trialof the effects of breakfast in ruralprimary school children. Am J ClinNutr. 1998;68:873-879.(Jamaica)

All students (grades 2-5) in 16 schoolswere weighed. The 405 studentsidentified as undernourished by lowweight for age were each matchedwith a same sex, age, schoolmatewho was not malnourished.

Breakfast group (compared to controlgroup) had improved height,weight, and attendance. Neitherbreakfast nor control group improvedin WRAT achievement test.






Attendance, a breakfast question-naire, and socioeconomic statusmeasures as well as tests forachievement (WRAT; reading, spell-ing, arithmetic) were studyinstruments. Researchers also madehome visits. All were assigned toeither control or breakfast group(at-school breakfast every day for1 year).

Richter LM, Rose C, Griesel RD.Cognitive and behaviouraleffects of a school breakfast.S Afr Med J. 1997;87:93-100.(South Africa)

108 children from a school known tohave undernutrition (second and thirdgrades) were divided into the experi-mental class (55 students, mean age10.5 years) and control class (53,mean age 8.3 years). Attention, dis-tractibility, and short-term memorywere measured (pre- and postinter-vention) using teacher ratings, stan-dard psychomotor tests, and videorecordings of classroom. All childrenin the experimental class receivedbreakfast for period of 6 weeks.

Between the pre- and postinterventionperiods, both experimental andcontrol groups improved in overallperformance (as well as coding andvigilance) on the WISC, but the exper-imental group improved more on thevigilance tests. Only the experimentalgroup improved significantly on digitspan. Changes in absenteeism werenot significant. School breakfast wasfelt to improve cognitive performanceof socially disadvantaged, undernour-ished children.

Simeon DT, Grantham-McGregor S.Effects of missing breakfast on thecognitive functions of school childrenof differing nutritional status.Am J Clin Nutr. 1989;49:646-653.(Jamaica)

Subjects included 3 groups of 30 childreneach (age 9-11) who were (1) hospital-ized for malnutrition during first 2 yearsof life, (2) stunted in growth (height forweight), and (3) nonstunted. All weregiven cognitive battery (WISC andPeabody), matching figures, andothers. Half received and half did notreceive breakfast after sleeping over-night at research center.

Performance on the fluency test andmental arithmetic were poorer forthe undernourished groups whenbreakfast was omitted. Omittingbreakfast made no difference onthe nonstunted group. Unexpectedly,the nonstunted group performedbetter in arithmetic when breakfastwas omitted.

Wyon DP, Abrahamsson L, Jartelius M,Fletcher RJ. An experimental studyof the effects of energy intake atbreakfast on the test performance of10-year-old children in school. Int JFood Sci Nutr. 1997;48:5-12.(Sweden)

195 families with 10-year-old childrenrecruited from 5 schools (10 classes)altered their children’s breakfast regi-men over a period of 4 successivedays for 1 week. Standard breakfastwas delivered to each boy’s home,with uneaten food quantitatively eval-uated. Children, not parents, wereblinded to purpose of study. Over 4days, different breakfasts wereoffered (high vs low nutritional value).Teachers blinded to nature of break-fasts assessed physical endurance,mathematical and creativity tests,and other measures of mood andbehavior.

Children’s energy intake when offereda higher energy breakfast (20% ofrecommended daily values) wasalways higher than when offereda lower energy breakfast (<10% of rec-ommended daily values). Amongthose who ate the higher energybreakfast, there were fewer errorsmade on the following tests: numberchecking (boys) and addition (girls andboys) and higher scores for creativity(boys). Girls worked faster on numberchecking. No differences were foundfor multiplication or grammatical rea-soning with high-energy breakfast.Physical endurance was also betteramong those with higher energybreakfasts. Intake at school lunch wasnot affected by type of breakfast.



diminished academic achievement. The 3 studies employeddifferent research methodologies and measures to formtheir conclusions; for example, 1 used parent report togauge academic achievement. The volume and quality ofdata from research in this field are modest. This may bepartially explained by the fact that prospective, controlledstudies on chronic undernourishment (which require thatclinicians not intervene in cases of hunger) are not ethi-cally feasible. Modest evidence in this small body of liter-ature should be sufficient to raise concern that foodinsufficiency exists among school-aged children and thatthe issue must be addressed. Academic outcomes appearto be among the important consequences of inaction.

BreakfastOf the 18 articles identified here (Table 4) that describe the

effects on school-aged children of eating, or not eating,breakfast, fewer than one-half studied populations in theUnited States.

Research studies of populations in Jamaica, Chile,Peru, and South Africa typically included children knownto have signs of undernourishment, such as low weightfor age. Populations studied in the United States, Sweden,and Canada were not reported to have signs or symptomsof undernourishment. Studies of absenteeism rates amongstudents offered breakfast at school found that childrenare more likely to attend school and have low rates of tar-diness when a school breakfast program exists.7

Positive short-term effects of eating breakfast on stu-dents’ school performance are almost consistently foundamong severely undernourished populations. The specificfunction that improved as a result of having had breakfastwas inconsistent. Improved functions found in studieswere verbal fluency, arithmetic, tests of attention, mem-ory, creativity, physical endurance, and general tests ofacademic achievement and cognitive functioning. Somestudies looked at the effects of missing breakfast on stu-dents’ performance on the same day the morning mealwas missed. Others studies investigated the positive ef-fects of instituting a school breakfast program over longerperiods of time. Others investigated breakfast versus fast-ing, while yet others investigated the quality of breakfasteaten. These vastly different study designs may explainthe variation found in cognitive functioning and schoolperformance associated with eating breakfast.

It is not clear from research on US populations thatbreakfast prior to school actually improves academic per-formance or cognitive ability. Those US studies that didfind improvements among students who had eaten break-fast were usually not carefully controlled. One Californiastudy found a strong association between percentages ofstudents eating breakfast and schools that had higher aca-demic scores.8 This study is purely retrospective, is withouta control group, is not yet published in a peer-reviewedjournal, and its research methodology has yet to be fullydescribed and exposed to objective critical analysis. How-ever, the information is derived from an impressive data-

base of 500,000 students, which is far bigger thanpublished studies that precede it in this field, and thereforecannot be ignored.

In summary, there appears to be a positive impact ofbreakfast on various cognitive skills in the short term.9

Long-term effects of healthy breakfasts on a student’sschool performance are still unknown. Long-term studiesare recommended as an important next step for researchersin the field of school nutrition.

CONCLUSIONChildren with iron deficiencies sufficient to cause ane-

mia are at a disadvantage academically, unless theyreceive iron therapy. Students with mild iron deficienciesand no anemia do not routinely exhibit problems in cog-nition or academic performance. There is no evidence thatmicronutrients such as zinc and iodine are deficient in USschool children in amounts that affect their academic per-formance. There is also no evidence that population-widevitamin and mineral supplementation will lead to anybenefits for academic performance in this country. Foodinsufficiency is a serious problem affecting children’sability to learn, but its relevance to US populations needsto be better understood. School breakfast programs seemto improve attendance and decrease tardiness. Offering ahealthy breakfast is an effective measure to improve aca-demic performance and cognitive functioning among un-dernourished populations. Eating breakfast, in contrast tofasting, may improve performance on the morning eaten.The long-term effects of eating breakfast on the perfor-mance of school children who do not have physical signsof severe undernourishment is less certain. j

References1. Smith WB. Commentary on Schoenthaler et al: vitamin and mineral

supplements—is the methodology sufficient to support the conclusions?J Altern Complement Med. 2000;6(1):31-35.

2. Egli I, Davidsson L, Zeder C, Walczyk T, Hurrell R. Dephytiniza-tion of a complementary food based on wheat and soy increases zinc, butnot copper, apparent absorption in adults. J Nutr. 2004;134(5):1077-1080.

3. Micheletti A, Rossi R, Rufini S. Zinc status in athletes: relation todiet and exercise. Sports Med. 2001;31(8):577-582.

4. Hollowell JG, Staehling NW, Hannon WH, et al. Iodine nutrition inthe United States. Trends and public health implications: iodine excretiondata from National Health and Nutrition Examination Surveys I and III(1971-1974 and 1988-1994). J Clin Endocrinol Metab. 1998;83(10):3401-3408.

5. Looker AC, Dallman PR, Carroll MD, Gunter EW, Johnson CL. Prev-alence of iron deficiency in the United States. JAMA. 1997;277(12):973-976.

6. American Academy of Pediatrics. Recommendations for preventivepediatric health care. Pediatrics. 2000;105(3):645-646. Available at:http://www.aap.org/policy/re9939.html.

7. Fernald L, Ani CC, Grantham-McGregor S. Does school breakfastbenefit children’s educational performance? Afr Health. 1997;19(6):19-20.

8. Hanson TL, Austin GA. Student Health Risks, Resilience, and Aca-demic Performance in California. California Healthy Kids Survey Fact-sheet 1. Los Alamitos, Calif: WestEd; 2002.

9. Cueto S. Breakfast and performance. Public Health Nutr. 2001;4(6A):1429-1431.


nutrition and student performance at...

Documents