f184 disease in iodine content intake premature deficiencyfor fetal thyroid function, both in utero...

Archives of Disease in Childhood 1994; 71: F184-Fl91

Iodine content of infant formulas and iodineintake of premature babies: high risk of iodinedeficiency

Susana Ares, Jose Quero, Socorro Duran, Maria Jesus Presas, Rafael Herruzo,Gabriella Morreale de Escobar

AbstractAs part of a study of thyroid function inpremature babies, the iodine content oftheir mothers' breast milk, that of 32formulas from different brands used inSpain, and that of 127 formulas used inother countries was determined. Breastmilk contained more iodine - mean(SEM) 10 (1) ,ug/dl - than most of the for-mulas, especially those for prematurebabies. Iodine intakes were thereforebelow the recommended daily amount(RDA) for newborns: babies of 27-30weeks' gestational age took 3-1 (1.1)jig/day at 5 days ofage and 29'8 (2.7) jig by2 months of age.This problem is not exclusive to Spanish

premature babies as the iodine content ofmany of the formulas on sale in othercountries was also inadequate. It is con-cluded that preterm infants who areformula fed are at high risk of iodinedeficiency.(Arch Dis Child 1994; 71: F184-FI91)

Unidad deNeonatologia, HospitalLa Paz, InstitutoNacional de SaludS AresJ Quero

Unidad deEndocrinologiaMolecular, Instituto doInvestigacionesBiom6dicas y Facultadde Medicina, ConsejoSuperior deInvestigacionesCientificas andUniversidadAut6nomaS DuranM J PresasGM de Escobar

Unidad de MedicinaPreventiva, Facultadde Medicina de laUniversidadAut6noma, Madrid,SpainR Herruzo

Correspondence to:Dr G Morreale de Escobar,Instituto de InvestigacionesBiomedicas, Facultad deMedicina UAM, ArzobispoMorcillo 4, 28029-Madrid,Spain.Accepted 23 July 1994

The trace element iodine is essential for thesynthesis of thyroid hormones. Iodine defi-ciency leads to a wide variety of disorders(iodine deficiency disorders or IDD). Theirseverity is related not only to the degree of defi-ciency, but also to the developmental phaseduring which it is incurred.1 2 Thus severeiodine deficiency in the mother leads toincreased rate of abortion, perinatal death, andinfant mortality, and to babies born withcentral nervous system defects. These rangefrom severe mental retardation, deaf-mutism,and spasticity to a milder decrease in intellec-tual performance.

Iodine deficiency is recognised as a majorcause of preventable mental retardationthroughout the world. IDD disappear whenthe deficiency is corrected. The magnitude ofthe problem, and the ease with which it can becorrected, led the Forty Third World HealthAssembly to urge its elimination by the year2000.3 The United Nations convened a WorldSummit for Children in 1990; 129 countrieshave now signed the ensuing declaration,which starts with an initial universal appeal togive every child a better future. It noted that'enhancement of children's health and nutri-tion is the first duty, and also a task for whichsolutions are now within reach'. The Summit

adopted a plan of action, which includedvirtual elimination of IDD.3To reach such a goal, the minimum daily

dietary iodine allowance should start with anintake of 175 ,ug a day in pregnant women,rising to 200 ,ug a day during breast feeing.4The recommended minimum daily intake forpregnant women has recently been increasedto 200 ,ug.5 This ought to ensure that thematernal thyroid synthesises enough hormonefor her needs and those ofthe developing fetus,and that there is an adequate supply of iodinefor fetal thyroid function, both in utero andduring lactation. 'Breast milk is the best sourceof iodine for the infant and exclusive breastfeeding till 4-6 months should be encouraged',according to recommendations on iodinenutrition for mothers and babies in Europe.5Even if the mother's intake is adequate, the

infant might be iodine deficient when breastfeeding is not possible. The iodine intake ofnewborns is then entirely dependent on theiodine content of the formulas used to feedthem. The risk of iodine deficiency increasesfurther in the case of premature babies: theyhave been unable to accumulate the amount ofiodine found in term newborns, and they areoften not able to breast feed. The minimumrecommended daily amount (RDA) for iodinewas established by the United States NationalAcademy of Sciences as 40 ,ug for babies aged0-6 months, 50 jig for those aged from 6months to one year, 70 jig for childrenbetween 1 and 3 years (table 1).4 Similarrecommendations were made by the AmericanAcademy of Pediatrics6 and the EuropeanSociety for Paediatric Gastroenterology andNutrition (ESPGAN).7 To meet these RDAs,the American Academy of Pediatrics recom-mended that the minimum iodine content ofpreparations for term newborns should be5 j,g/lOO kcal, equivalent to 3-5 jig/dl, andconsidered the same amount to be adequatefor premature infants.6 ESPGAN consideredthat term newborn babies would require 45 jiga day and recommended a minimum iodinecontent for formula of 3-5 ,ug/dl.78 TheESPGAN recommendation for formulas forpreterm infants is 7 ,ug/dl.9 The EuropeanCommunities Commission also recommended3-5 ,ug/dl for start and follow up formulapreparations,1I but made no recommendationsfor premature babies. The RDAs for differentage groups have recently been revised5 to90 jig a day, from birth up to 6 years of life

F184

on February 4, 2021 by guest. P

rotected by copyright.http://fn.bm

j.com/

Arch D

is Child F

etal Neonatal E

d: first published as 10.1136/fn.71.3.F184 on 1 N

ovember 1994. D

ownloaded from

http://fn.bmj.com/

Iodine content of infantformulas and iodine intake ofpremature babies: high risk of iodine deficiency

Table 1 Recommendations for minimum daily iodineintake (,ug/day)

Group Age A: before 1992 B:from 1992

Preterm infants >30 p.glkg/day0-5 months 35-40 905-12 months 50 901-3 years 70 904-6 years 90 907-10 years 120 120

Adults 150 150Pregnant women 175 200Lactating women 200 200

A: RDA from Food and Nutrition Board of the NationalAcademy of Sciences,5 American Academy of Pediatrics,7 andESPGAN Committee on Nutrition.89B: Recommendations on iodine nutrition made by ICCIDD.6

(table 1). Taking into consideration new infor-mation on iodine metabolism in prematureand term newborn babies,"1 12 the iodinecontent of formulas for premature newbornsneeds to contain 20 ,ug/dl and that of follow uppreparations 10 p,g/dl. We will refer here tothese new recommendations as those of theICCIDD (International Committee for theControl of Iodine Deficiency Disorders).The present study was performed to assess

both the iodine content of preparations whichare currently used to feed premature babies,and to determine the actual iodine intake of thebabies using them. Although most of the infor-mation has been obtained for babies from onecountry (Spain), formula preparations fromother countries have also been studied. Themain conclusions drawn for Spain also seemto pertain for premature babies from othercountries as well.

MethodsAs part of a study on the thyroid function ofpremature babies, performed on 1 15 infants inthe neonatology unit of the La Paz Hospital inMadrid, samples of the different formulas usedover a one year period were collected directlyfrom the bottle, when fed to babies of differentpostnatal ages. Infants exposed to externalsources of iodine as a result of radiologicalexaminations were excluded from the study.During the study period iodine-containingtopical antiseptics were banned from theneonatology unit. The exact volume taken overa 24 hour period was recorded in each case.The samples had been reconstituted in thehospital kitchen as 14% dilutions in boiledwater. All the different types of formula usedduring this period were also obtained inpowdered form and reconstituted in thelaboratory as 14% dilutions, using distilledwater.

Using Student's paired t test there was nodifference in iodine concentration between theready-to-feed samples and those prepared inthe laboratory, so results have been evaluatedas a whole. Data were obtained for 162samples of formulas for premature infants, 77samples of start formulas, and 55 samples ofspecial (medical) formulas. These samplesentailed nine different brands, comprising 32different formulas available in Spain. Thesewere manufactured by Abbott (n= 1), Alter(n=2), Mead-Johnson (n=2), Milupa (n=7),

Nestle (n=7), Nutricia (n=3), Ordesa (n=3),Uniasa (n=4), Wander-Sandoz (n= 3).

Formulas from the same and differentbrands were also obtained from 12 othercountries (Czech Republic, Finland, France,Greece, Italy, Japan, Norway, Poland,Switzerland, and the United States ofAmerica). There were 103 samples of formulas(65 samples from start formulas, 21 samplesfrom follow up formulas, 19 samples fromformulas for premature babies and 35 samplesof special formulas). Data on the iodinecontent of formulas used in Austria, Canada,Denmark and Germany were kindly madeavailable to us for this report.

Seventy one samples of breast milk werecollected in sterilised plastic bottles from 34 ofthe mothers of the premature babies (less than36 weeks of gestation) included in the study.Thirty four were obtained during the first weekafter delivery, 15 between the second andfourth weeks, and 14 between the first andsecond month. Twenty one samples were alsoobtained from 13 mothers of full term infants -13 samples at one week and eight samples atone month after delivery.The iodine concentration was determined

using a modification of the method describedby Benotti and Benotti'4 for serum, with dupli-cate 40 [LI aliquots of each preparation beingdigested in 2 ml of chloric acid. This methoddetermines the total iodine content of milk,irrespective of its initial chemical form. Theinterassay and intra-assay coefficients of varia-tion were 14-7 (SD) (6.7)% and 12-5 (1.8)%,respectively.Serum thyroid stimulating hormone (TSH)

(Dynotest TSH; Henning, Berlin GMBH),thyroxine (T4) and trio-iodothyronine (T3)(radioimmunoassay KIT; Clinical Assays), freethyroxine (FI4) (Two Step Gammacoat FreeT4 RIA KIT; Clinical Assays) activities weremeasured in duplicate by specific radio-immunoassays. Thyroglobulin was measured byIRMA (Dynotest Tg; Henning, Berlin GMBH).When a small blood volume was obtained, onlyFT4, T3, and thyroglobulin were measured.Data were subjected to one way analysis of

variance. The significance of the differencebetween the groups was identified using the

15-

10-

5.

0-

Iodine concentration in human milk

Premature delivery

| Term delivery

T T T

6

-1.2

0.9

06E

r0.3

0

Days after deliveryFigure 1 Mean (SEM) iodine concentrations in the milkfrom mothers ofpremature and term infants, obtainedabout one, three, and six weeks after delivery.

F185



j.com/

Arch D

is Child F

etal Neonatal E


ovember 1994. D

ownloaded from

http://fn.bmj.com/

Ares, Quero, Duran, Presas, Herruzo, de Escobar

Newman-Keuls test for multiple comparisons.All calculations were performed as describedby Snedecor and Cochran.15 Multiple regres-sion and partial correlation analysis wereperformed using the SPSS program; thecovariance analysis was done using theSIGMA program.

ResultsFigure 1 shows the concentration of iodine inthe milk from mothers of premature and termbabies at different postnatal ages. No signifi-cant differences were found between samplesfrom women after premature and termdelivery, or between different times during thebreast feeding period. The overall mean(SEM) value (10 0 (1 0) p.g/dl) will thereforebe used for comparisons with the iodinecontent of the different formulas.

IODINE CONTENT OF PREMATURE AND IN'FANTFORMULAS AND IODINE INTAKEFigure 2 shows the iodine content of differentformulas for premature infants, start, andfollow up formulas, and special (medical)

Formulas for preterms

pig/dI I I I I I I I a I I I

0 5 10 15 20

0 (1;20)

P (1;63)

N (1;25)

M(1;50)a :

W (1; 15) i ......

pmol/I 0.0 0.4

ESPGAN

A

1.2 1.6

formulas, used over a one year period com-pared with the iodine content of their mothers'milk. In most of the preparations the iodinecontent was significantly lower than that ofhuman milk. Although the iodine content ofmost start, follow up, and special formulaswas similar to that recommended by ESPGAN(3*5 FLg/dl), it was usually lower than theESPGAN recommendation for prematurebabies (7 f.g/dl). The iodine content of differ-ent brands of the same type of formulas (forpremature babies, follow up, etc) was, in mostcases, very similar. However, different types offormulas from the same brand may have differ-ent iodine contents: thus in some cases theiodine content of a start formula is adequate,whereas that of the formulas for prematureinfants might be very low.The iodine intake of the premature babies

was directly related to the iodine content of theformula and to the volume of milk ingesteddaily. The latter is related to the postnatal ageof the premature baby. We observed that verysmall premature babies (27-30 weeks' gesta-tional age) ingested about 70 ml/kg six daysafter birth, 105 ml/kg at 3 weeks of postnatalage, 150 ml/kg by 2 months, and about 175

Starting formulas

P (1;11)

N (2; 10)

M (3; 15)

A (1;6)

Nu (1;8)

M-J (1;8)

W (1; 13)

0 (2; 15)

0 1 5 1~ ~~~~015 20* . _ ~~~~~~~.....:.:. ::.

......~~~~~~~........_ _,,'~~~~~~..........e.........~ ~~~...............~~~~~~~~................................................................::::::::::::::::::

. ............

...............

_ . ..........~~~...............I....................

,,, , ~~~~~........00 j K 04 08 1 2 16.......

ESPGAN

Follow up formulaspg/dl .

0 1 5 10 15 20

Special formulas0 5 . . . . 1. . . . 2. . . . 0

0 1 5 10 15 20

P (1;2)

NU (1;2)

N (1;2)

M (1;2)

pmol/I 0.0 0.4

ESPGAN

... ............08........

Figure 2 Mean (SEM) iodine content of different formulas usedfor premature infants (panel A), start (panel B), followup (panel C), and special (panel D) formulas, used over a one-year period in Madrid. The numbers in parentheses at theleft of the data bars correspond to the number ofpreparations, followed by the number of different sanmples for the brandidentified by the following initials: A=Alter; AB=Abbott; M-Jr=Mead-Johnson; M=Milupa; N= Nestek; Nu=Nutricia;P= Puleva; 0= Ordesa; W= Wander-Sandoz. The shaded area corresponds to the 95% confidence interval for the iodinecontenzt of nmaternial nmilk, as assessed from the data shown in fig 1. The arroz! inldicates the nilininal iodinie conceettatioutrecommended by ESPGAN.

NU (1;15)

AB (1;2)

P (1;2)

W (2;4)

N (1;7)

M-J (2; 12)

M (2;6) PW- I-I00 0-4

ESPGAN

D

1.2 16

F186

0.8



j.com/

Arch D

is Child F

etal Neonatal E


ovember 1994. D

ownloaded from

http://fn.bmj.com/

Iodine content of infant formulas and iodine intake ofpremature babies: high risk of iodine deficiency

Iodine intake

U 27-30weeksGA U 31-33weeksGA U 34-36weeksGA

60- A

501T

4.8- B

4.01

3-2 -

2.4E

1.6

08

lD

Figure 3 PanelA shows the mean (SEM) 24 hour iodine intake ofpremature infants formula fed at different postnatalages. Panel B shows the corresponding values for premature infants on mixed breast andformula milks at the samepostnatal ages. Panel B includes only data from infants receiving more than half their intake as maternal milk. The shadedarea indicates the recommendations for iodine intake (35-40 ,ug a day).4 7 8 The 1992 ICCIDD recommendations wouldbe much higher, at 90 ,ug per day. 12 The data of the premature babies have been grouped according to their gestational age(GA)at birth.

ml/kg by three months. They do not ingest therecommended 150-200 ml/kg until they aremore than 1 month old. Less premature babies(31-36 weeks' gestational age) were drinkingabout 100 ml/kg six days after birth, 150 mI/kgby three weeks of age, and 150-200 ml/kgbetween 1 and 2 months of age. The volumesof milk ingested were mainly related to theirbody weight, and these were lower thanrecommended until the baby weighed about2 kg. Figure 3 shows the iodine intake ofpremature babies at different postnatal ages.The actual iodine intake is lower than the RDAin most groups of premature babies takingformulas alone for at least one month afterbirth, especially in the case ofthose aged 27-33weeks' gestation. However, the recommendediodine intake was reached sooner by prematurebabies on 50/o breast milk or more. Again, the27-30 weeks' gestational age group has thelowest intake, the daily volume ingested isdirectly proportional to their body weight

(p<0-0 1). The iodine intake of the sick prema-ture babies was even lower than that of weightmatched healthy babies. None of the prema-ture babies had an iodine intake approachingthe ICCIDD recommendation of 90 pug a day.

Figure 4 shows the mean iodine content fordifferent types of formulas in countries otherthan Spain. Again, preparations for prema-ture babies and special formulas fail moreoften to reach the ESPGAN or ICCIDDrecommendations for iodine content than thestart formulas.We did not find any significant differences in

the iodine content among 12 different brands:the variability around the mean values for eachbrand was very high. For 32 preparations forone brand, for instance, values ranged from 1 0to 15-0 p.g/dl. We have already commentedthat the iodine content of different types ofpreparations from the same brand may be dif-ferent, and this is also true of those sold inother countries. Another factor increasing vari-

Preterm formulas

GR (1;2)JP (1;2)IT (1;2)CH (2;4)AT (3;6)ES (4;8)CA (5; 10)DE (1;2)

11 -mI

.

ESPGAN ICCIDD

Starting formulas

FR (13;26)GR (2;4)USA (1;2)CA (8; 16)Fl (1;2)CK (3;6)DE (4;8)CH (7; 14)NO (2;4)JP (4;8)IT (13;26)ES (16;32)DK (5; 10)AT (8; 16)PO (2;4)

5 10 15

ESPGAN ICCIDD

USA (3;6)CH (6; 12)FR (4;8)CA (5; 10)ES (11;22)DK (2;4)AT (5; 10)IT (4;8)

20

Special formulas

--I

. .

0 5 10 15 20

ESPGAN

Iodine concentration (gg/dl)Figure 4 Mean (SEM) iodine content of different types offormulas used in Spain and other countries. AT=Austria; CA =Canada; CH= Switzerland;CK=Czech Republic; DE=Denmark; ES=Spain; FI=Finland; FR=France; GR=Greece; IT=Italy;J7P=_Japan; NO=Norway; PO=Poland;USA= United States ofNorth America. As formulas from the same brand may have different names in different countries, the data were grouped accordingto type. The numbers in parentheses at the left of the data bars correspond to the number ofpreparations, followed by the number of different samples. Thearrows indicate the minimal iodine concentration recommended by ESPGAN7 8 and ICCIDD. 12 ICCIDD has notformulated a recommendation forspecialformulas. To convert the iodine concentrations into jumoOl, multiply the tabular values by 0-08.

F187

40-

'a

5

T

vrFs \ I;--I



j.com/

Arch D

is Child F

etal Neonatal E


ovember 1994. D

ownloaded from

http://fn.bmj.com/


Table 2 Mean (SEM) iodine content of infantformulasused in different countries

No of Iodine contentCountry participants (pugldl) Range

Austria 16 6-6 (1-0) 33-17-0Canada 18 9 3 (0 8) 4-0-14-0Czech Republic 3 8-7 (2 2) 60-14-0Denmark 8 5-5 (1-6) 2-0-13-7Finland 3 9-0 (1-0) 8-0-10-0France 17 10-5 (0 8) 40-15-0Germany 5 7-9 (1-0) 5-9-10-4Greece 3 12-0 (1-8) 9-0-15-0Italy 16 7-5 (08) 2-0-12-0Japan 5 8-8 (1-6) 4-0-12-0Norway 2 8-0 (5-7) 4-0-12-0Poland 2 4-5 (0-7)Spain 32 6-5 (04) 1-5-11-0Switzerland 15 9-2 (0.8) 5-0-13-0United States ofAmerica 4 13-5 (1-6) 5-0-13-0

To convert the iodine concentration into p.mol/l, the tabularvalues should be multiplied by 0-08. Results of the one wayANOVA: F(13,127) 4-57; p<0-001.

ability is the country where the formula is sold(table 2). The large variability in the data forthe same country is related to the fact that dif-ferent types of formulas from different brandswere included in the calculation of the meanvalue: although the iodine content of specialformulas was not significantly different fromthat of start formulas sold in the same country,the iodine content of formulas for preterminfants was often lower than the recommendedamount. Despite this variability, there weredifferences between countries, irrespective ofbrand and type of formula. Thus as a whole,among European countries formulas containas much iodine in France, Greece, andSwitzerland as those used in Canada, Japan,and the United States of America; the iodinecontent is significantly lower in formulas fromAustria, Denmark, Germany, Italy, and Spainwhether or not the data offormulas for pretermbabies are included. This cannot only be dueto the preferential use of formula in eachcountry, because the iodine content offormulas from the same brand may varybetween countries.The iodine content of Mead-Johnson

preparations, for instance, is significantly lowerin Spain (2.2 (0-3) ,ug/dl) than in Canada (7.2(1-0) pug/dl), France (10 - (0 5) ,ug/dl) and theUnited States (12 0 (1P7) ,ug/dl); that ofNestlepreparations is significantly higher in France(14-0 (0 0) ,tg/dl) and Switzerland (10-6 (3 0)pug/dl) than in Austria (4 0 (05) ,ug/dl), Italy(7 0 (3 5) ,ug/dl) and Spain (6&4 (2-3) Rg/dl);and that of Milupa preparations is lower inAustria (7 3 (1-6) ,ug/dl), France (6-0 (1.7),ug/dl), and Spain (5 1 (2 0) ,ug/dl) comparedwith Greece (13-5 (2-1) ,ug/dl), Italy (9 0 (3 5),g/dl), Germany (9 9 (2-1) pug/dl) andSwitzerland (8-7 (3-2) pugtdl). The iodine con-tent of start and special formulas from differentbrands and different countries tends to besimilar and more often meets the ESPGANrecommendations than do the preparations forpremature babies. For these there is a muchgreater variability, both between brands andbetween countries. These differences might berelated to different legislation in differentcountries.

THYROID FUNCTION TESTSCorrelations were found between iodine intakeand circulating T3 (r=0-3; p<0001), FT4(r=0-80; p<0001), thyroglobulin (r= -0-3;p<000 1), and thyroid stimulating hormone(r=-0-3; p<0Q05); iodine intake and post-menstrual age (r=0-58; p<0 001); postmen-strual age and T3 (r=0-28; p<0-0 1), T4 (0-17;p<005), thyroglobulin (r=-0 30; p<0 01),but not between postmenstrual age and FT4 orthyroid stimulating hormone.The correlations between iodine intake and

postmenstrual age and between postmenstrualage and T3 (or T4, or thyroglobulin) couldobviously be influencing the correlationbetween the several parameters of thyroidfunction and iodine intake. To assess whetherthe correlations between iodine and thyroidindicators are merely a consequence of theinfluence of postmenstrual age on iodineintake and several parameters of thyroid func-tion, or whether there is an effect of the iodineintake on thyroid function independent of thedegree of maturation of the premature infant,we analysed the results by multiple regression,partial correlation, and by covariance.We used only the data from those cases in

which all variables had been measured; thosecases where one or more variables were miss-ing were omitted from the study. This avoideddifferent variables being assigned greaterweight. We also omitted the data from verysick infants to exclude any changes in thyroidfunction caused by different pathologies.These criteria reduced the study to 90 cases.

Results obtained by multiple regression andpartial correlation analysis were quite compar-able and showed that iodine intake is highlycorrelated to postmenstrual age, but that,independently of age, it is also correlated posi-tively with FT4 and T3, and negatively withthyroglobulin.

Covariance analysis was applied to iodineintakes below or above 40 ,ug/day. This iodineintake is the previous RDA. Postmenstrual agewas entered as a control variable. Parametersof thyroid function were classified into threegroups according to gestational age andcorrected by the program for postmenstrualage. The adjusted means (SEM) thus obtainedare shown in fig 5, as well as the significantdifferences obtained between the group ofsimilar gestational age with iodine intakesabove the RDA compared with the group withlower intakes. Most babies of the 31-36 weeksin gestational age groups with an iodine intakebelow 40 ,ug/day have decreased serum T3 andFT4, and higher serum thyroglobulin andTSH values, than those with intakes above theRDA. Circulating T3 and thyroglobulin werealso different in the 27-30 weeks in gestationalage premature babies.

DiscussionThe iodine intake ofnewborn babies is entirelydependent on the iodine content ofbreast milkor of fornula preparations. The criteria usedinitially to define the RDA ofterm newborns as30-40 ,ug a day were based on the amounts

F188



j.com/

Arch D

is Child F

etal Neonatal E


ovember 1994. D

ownloaded from

http://fn.bmj.com/

Iodine content of infantformulas and iodine intake ofpremature babies: high risk of iodine deficiency

* Iodine intake below 40 jg/day

0 Iodine intake above 40 jig/day

FreeT4 (pmol/l)

* *

ml

27-30 31-33 34-36

27-30 31-33 34-36

Gestational age (weeks)Figure 5 Mean values of circulating FT4, T3, thyroglobulin, and TSH adjusted by covariance analysis to postmenstrualage, for premature babies of different gestational ages with iodine intakes below or above 40 pg/day. *Differences which aresignificant atp<O 05.

which would be needed for the synthesis of theiodinated hormones metabolised by the termnewborn, the likely amount lost via the urine,faeces, and saliva, and the amount needed forthe build up of thyroid stores of iodinatedcompounds. To define the minimum iodineconcentration for formula preparations fornewborn babies, it was reasoned that theiriodine content should be the same as that ofbreast milk from normal women, a value of 7,ug/dl being taken as the mean value.9 This ledto a recommendation of 3-5 ,ug/dl of iodine inreconstituted formula preparations. On thisbasis, however, term newborns would notreach the recommended intake until theydrank 1000 ml or more of formula.

Moreover, mean values for the iodine con-tent of breast milk vary (table 3). Values mayvary from 1 2 (0 1) jig/dl in Jena (Germany),known as an iodine deficient area, to 17-8 (1 1),ug/dl in the United States, where the popula-tion is considered to receive an adequateiodine intake. The value of 7 ,ug/dl used byESPGAN9 as a reference for the iodine contentof breast milk is lower than the values actuallyfound in women from Madrid and Brussels.The latter have clearly been shown to havemoderate iodine deficiency,'2 23 especiallyduring pregnancy.24 Thus the recommenda-

tion of 3-5 ,ug/dl for term newborn formulaswould lead to an iodine intake equivalent tothat of babies born in areas with moderateiodine deficiency.The recommendation for premature formu-

las was increased to 7 ,ug/dl, no explanationbeing given as to how this value was reached.9On this basis, the 27-30 weeks in gestationalage preterm babies would have to ingest morethan 500 ml of formula daily to reach the RDAof 40 ,ug for iodine. The actual volumes, how-ever, which were fed to such babies were lowerthan expected at least at one month after birth,

Table 3 Mean (SEM) iodine content of breast milk asreportedfor women from different countries

No of Human milkCountry City samples (jig /dl) References

Germany Friburg 41 2-5 16Jena* 55 1-2 (0-1) 1617

Belgium Brussels 91 9-5 (0-6) 16 17Spain Madrid (1988) 69 7-7 (0-9) 18

Madrid (1991) 52 10-0 (1 0)France Paris 68 8-2 (0-5) 19Italy Verona 77 5-4 (0-69) 20

San Angelo* 59 2-7 (0-3) 16 17Sicily

Sweden Stockholm 60 9-3 16USA 61 17-8 (1 1) 21

15-3 (1.0) 21Canada 11 0 22

*Areas with endemic goitre.

F189



j.com/

Arch D

is Child F

etal Neonatal E


ovember 1994. D

ownloaded from

http://fn.bmj.com/


or longer in the case of the 27-30 weeks ingestational age premature babies. If theformula preparations contained the recom-mended amount of iodine their iodine intakewould be about 5 jig, 9 jig, and 22 [Lg a day,respectively. Moreover, as many of thepreparations for premature babies used in thepresent study did not contain the recom-mended iodine content, the actual iodineintake of the 27-30 weeks in gestational agepremature babies was 3 1 (1 1) jig/day at 6days of age, 7-4 (5-5) jig at 3 weeks, and 17-1(2 9) jig at one month, amounts which areclearly far below the RDA.4 The 40 jig RDAhad not yet been reached at 2 months of age,when the mean intake was still only 29-8 (2-7)jig. Findings were similar in premature babiesof 31-36 weeks' gestational age, who did notreach the above recommended intake untilthey were 1 month of age, or older.The previous recommendations for prema-

ture babies, apparently derived from thosemade for term newborns, implied that iodinerequirements would be proportional to theweight of the preterm infant, and would thusbe lower with increasing prematurity. This,however, has clearly been contested by severalsubsequent studies,11 1225 which show thatiodine metabolism in premature babies pre-sents special features compared with termneonates. For example, Delange et al 12 foundthat the urinary iodine excretion in 29 pretermbabies was 1-5 times that of 20 term neonates;the iodine in faeces were very low; and prema-ture babies were frequently in negative iodinebalance, attributed to immaturity both of thekidney and the thyroid. As a consequence, andtaking into consideration that the growinginfant has to build up iodine stores in itsthyroid, Delange proposed that to reach a posi-tive iodine balance, the preterm babies wouldneed a minimum of 30 jig/kg iodine per day,which is double the minimum amount (15jig/kg a day) for term babies.1' 12 To ensure apositive balance, the minimum RDA would be90 jim a day (table 1). The recent consensus isthat formulas for premature babies should con-tain at least 20 jig/dl iodine, and start formulasat least 10 jig/dl iodine.'2 Even with this higheriodine content, very small premature babiesreceiving formula feeds would not reach anintake of 90 jig a day, but they would at leastingest 30-40 jig. These new recommendationsare well below the limit (100 jig/kg/day) set byFisher as the maximum iodine intake compat-ible with normal thyroid function in prematureand term newborns.26The data presented here show that many

preparations presently on the market do noteven contain the minimum amount of iodinerecommended before 1992. This is especiallyso for preparations used for preterm infants.Only a few contain the 10 jig/dl now recom-mended for start formulas, none reaching the20 jig/dl recommended for premature babies.5Even if the preparations did contain the pre-viously recommended 7 jig/dl, prematurebabies in the present study did not ingest 40 jigiodine daily until they were more than 2months old. Like Delange et al, we

believe'1 12 16 17 that the premature babies inour study were iodine deficient, and that thiscould affect their thyroid function. The resultsobtained show that iodine intake is highlycorrelated with postmenstrual age, but that,independent of age, it is also positively corre-lated with TF4 and T3, and negatively withthyroglobulin. Most babies with an intakebelow 40 jig/day had lower circulating FT4and T3, and higher thyroglobulin and TSHthan babies with an iodine intake above theRDA. Again, this cannot be entirely attributedto their degree of immaturity. In the presentstudy increased thyroglobulin activities seemto be a clear sign of thyroid stimulation.27

Iodine deficiency in preterm newbornsincreases the risk of hypothyroidism.25Alterations in thyroid function in prematurebabies, leading to low circulating T4 or T3,have been associated with impaired neuralmaturation, as measured by nerve conductionvelocity28 and by lower scores in the Bayleymental and motor scales.29 Iodine deficiencymay well contribute to inadequate thyroidhormone activity in premature babies, butcould easily be avoided.

In conclusion, in view of more reliablerecent information on the thyroid function andphysiology of preterm babies, the iodine con-tent of many infant formulas seems to beinadequate. Most premature babies in manycountries might not be ingesting the previouslyrecommended amount of 40 jig a day, evenfewer the 90 jig recommended in 1992 at ameeting sponsored by the ICCIDD, the WorldHealth Organisation, and the EuropeanCommunity.5 Producers of such preparationsshould be urged to comply with the newrecommendations and to specify that theirproducts do so, irrespective of the type offormulas and the country where their productsare being used. Premature babies in manycountries are now iodine deficient, precisely ata stage of development that is highly sensitiveto changes in thyroid function. This may occurin countries where the iodine intake of the restof the population is adequate.23We are grateful for the invaluable collaboration of the nursingstaff of the Neonatology Unit of La Paz Hospital, Madrid, forcareful collection of formulas and breast milk samples.This work was supported in part by the Heinz-Koch

Foundation (Milupa, Spain) and by grants from Fondo deInvestigaciones Sanitarias (grant No 9210888), and by a fellow-ship (No 92/5351) to S Ares.We are grateful to the following colleagues for providing us

with formulas used in different countries: Dr P Langer fromBratislava, Dr B A Lamberg from Helsinki, Dr R Momex fromParis, Dr D Koutras from Athens, Dr P Vitti from Pisa, Dr SNagataki from Nagasaki, DrH Frey from Oslo, DrM Gembickifrom Warsaw, Dr Ruth ilmig from Zurich, Dr H Burgi fromSolothurn, and Dr J T Dunn from Charlottesville. Data forformulas used in Austria were provided by Dr 0 Eber fromGraz,13 in Canada by Dr J H Dussault from Quebec, inDenmark by Dr P Laurberg from Aalborg, and in Germany byMilupa.

1 Hetzel BS. Iodine deficiency disorders (IDD) and theireradication. Lancet 1983; ii: 1126-9.

2 Dunn JT. Iodine deficiency, the next target for elimination?N EnglJMed 1992; 326: 267-8.

3 Resolution WHA43.2 of the World Health Assembly,Geneva 1990. World declaration on the survival, protectionand development of children, and plan of action. New York:United Nations, 1990.

4 Food and Nutrition Board of the National Academy ofSciences and National Research Council. Recommendeddaily amounts. 10th Edn. Washington, DC: NationalAcademy Press, 1989: 213-17.

5 Recommendations on iodine nutrition for mothers andinfants in Europe. In: Delange F, Dunn JT, Glinoer D,

F190



j.com/

Arch D

is Child F

etal Neonatal E


ovember 1994. D

ownloaded from

http://fn.bmj.com/

Iodine content of infant formulas and iodine intake ofpremature babies: high risk of iodine deficiency

eds. Iodine deficiency disorders in Europe: a continuingconcern. New York: Plenum Press, 1993: 471-8.

6 American Academy of Pediatrics. Committee on Nutrition.Nutritional needs of low birth weight infants. Pediatrics1985; 60: 519-30.

7 ESPGAN Committee on Nutrition. Guidelines on infantnutrition. Recommendation for the composition of anadapted formula. Acta Paediatr Scand 1977; 262 (suppl):1-20.

8 ESPGAN Committee on Nutrition. Guidelines on infantnutrition. II Recommendations for the composition offollow-up formula and beikost. Acta Paediatr Scand 1981;287 (suppl): 1-25.

9 Bremer HJ, Brooke OG, Orzalesi M, Putet G, Raiha NCR,Senterre J, et al. Nutrition and feeding of preterm infants.Committee on Nutrition of the Preterm Infant, EuropeanSociety of Pediatric Gastroenterology and Nutrition(ESPGAN). Acta Paediatr Scand 1987; 336 (suppl): 1- 14.

10 Committee of the European Communities for the legisla-tion on infant formulas. Official Bulletin of the EuropeanCommunities 1986; C285: 5-19.

11 Delange F, Bourdoux J, Chanoine P, Ermans AM.Physiopathology of Iodine nutrition during pregnancy,lactation, and early postnatal life. In: Berger H, ed.Vitamins and minerals in pregnancy and lactation. NestleNutrition Workshop Series. Vol 16. New York:Vevey/Raven Press, 1988: 205-13.

12 Delange F. Requirements of iodine in humans. In: DelangeF, Dunn JT, Glinoer D, eds. Iodine deficiency in Europe. Acontinuing concern. New York: Plenum Press, 1993: 5-16.

13 Buchinger W, Langsteger W, Tiran B, Wawschinek 0,Ebber 0, Jodgehalt der Sauglingsnahrung in Oesterreich.Acta Med Austr 1993; 20: 131-3.

14 Benotti J, Benotti NA. A semiautomated method for thedetermination of the plasma PBI. Clin Chem 1963; 9:408-16.

15 Snedecor GW, Cochran WG. Statistical methods. 7th Edn.Ames, Iowa: Iowa State University Press, 1980.

16 Delange F. Iodine nutrition and risk of thyroid irradiationfrom nuclear accidents. In: Rubery E, Smales E, eds.Iodine profylaxis folowing nuclear accidents. Oxford:Pergamon Press, 1990: 45-53.

17 Delange F, Ermans AM. Iodine deficiency. In: BravemanLE, Utiger RD, eds. The thyroid. Afundamental and clinicaltext. Philadelphia: J B Lippincott, 1991.

18 Mallol J, Morreale G, Escobar del Rey F. Thyroid hormoneand iodine intakes of bottle fed versus breast-fed babies.Endocrinologia 1988; 35: 12-20.

19 Etling N, Gehin-Fouque F. Iodinated compounds andthyroxine binding to albumin in human breast milk.PediatrRes 1984; 18: 901-3.

20 Etling N, Padovani E, Gehin-Fouque F, Tato L. Iodine andthyroid hormone levels in serum and urine of fullterm newborn infants. Helv Paediatr Acta 1983; 38:117-22.

21 Gushurst CA, Mueller JA, Green JA, Sedor F. Breast milkiodine: Reassessment in the 1980s. Pediatrics 1984; 73:354-7.

22 Ross Laboratories. Composition of feeding for infants andyoung children in Canada. Montreal: Ross Laboratories,1989.

23 Delange F, Burgi H. Iodine deficiency disorders in Europe.Bull WHO 1989; 67: 317-25.

24 Glinoer D, Laboureur I, Delange F, Bourdoux P, KinthaertJ, Grun JP, et al. Maternal and neonatal thyroid functionsin conditions of marginally low iodine intake. In: GordonA, Gross J, Hennemann G, eds. Progress in thyroid research.Rotterdam: Balkema, 1991: 165-8.

25 Delange F, Dalhem A, Bourdoux P, Lagasse R, Glinoer D,Fisher DA, et al. Increased risk ofprimary hypothyroidismin preterm infants. J Pediatrics 1984; 105: 462-9.

26 Fisher DA. Upper limit of iodine in infants formulas. J Nutr1989; 119: 1865.

27 Pacini F, Lari R, La Ricca P, Grasso L, Taddei D, BardiniN, et al. Serum thyroglobulin in newborns cord blood, inchildhood and adolescence: a physiological indicator ofthyroidal status. J Endocrinol Invest 1984; 7: 467-71.

28 De Vries LS, Heckmatt JZ, Burrin JM, Dubowitz LMS,Dubowitz V. Low serum thyroxine concentrations andneural maturation in preterm infants. Arch Dis Child 1986;61: 862-6.

29 Lucas A, Rennie J, Baker BA, Morley R. Low plasma tri-iodothyronine concentrations and outcome in preterminfants. Arch Dis Child 1988; 63: 1201-6.

F191



j.com/

Arch D

is Child F

etal Neonatal E


ovember 1994. D

ownloaded from

http://fn.bmj.com/

f184 disease in iodine content intake premature deficiencyfor fetal thyroid function, both in utero...

Documents