169

Acta Clinica Belgica, 2003; 58-3

IRON DEFICIENCY ANAEMIA DURING PREGNANCY IN MONS, BELGIUM

A SURVEY OF IRON DEFICIENCY ANAEMIA DURINGPREGNANCY IN BELGIUM:

ANALYSIS OF ROUTINE HOSPITAL LABORATORY DATAIN MONS

C. Massot1, J. Vanderpas2

–––––––––––––––1 Observatoire de Santé,

Rue Saint Antoine, 7021 Havré (Mons)2 CHU Ambroise Paré,

Bd Kennedy 2, 7000 Mons

Address for correspondence:Dr Christian MassotObservatoire de SantéRue Saint Antoine 1B - 7021 Havré (Mons)Tel. (32) 065 87 96 00

Original articles

ABSTRACT

Anaemia during pregnancy is a common problemworldwide. In industrialised countries, it is still frequentin some groups of population. This study is based onthe retrospective analysis of results of routine bloodanalysis performed on 15-44 years old women attend-ing prenatal clinics (study group) or other outpatient clin-ics (non pregnant group) in a public hospital in Mons,Belgium from 1997 to 1999.

In the non-pregnant group (2503 women), anaemia(haemoglobin < 12 g/dL) was present in 7.7% of thewomen. During pregnancy, anaemia was defined as hae-moglobin level < 11g/dL. In our sample, during the 1st

trimester of pregnancy, anaemia was present in 4.3% of887 pregnancies, among which 35% meeting CDC cri-teria (ferritin < 12 µg/L) for iron deficiency anaemia(IDA). Frequency of anaemia increases through preg-nancy. Among 1313 pregnancies, 31% suffer from anae-mia during the 3rd trimester, among which 75% meetcriteria for IDA.

Both low haemoglobin and low ferritin levels dur-ing the 1st trimester are good predictors of 3rd trimesteranaemia.

Systematic administration of iron supplement dur-ing pregnancy is matter of debate. In order to limit sup-plementation to pregnant women at risk of 3rd trimesteranaemia, we suggest to treat anaemia (haemoglobin level< 11g/d) detected at the first prenatal visit and to givesmall doses of iron (30mg per day) when haemoglobinlevel is between 11g/dL and 13g/dL or ferritin level isless than 20µg/dL. Low doses are generally well toler-ated and compliance is better.

INTRODUCTION

Pregnancy represents a major challenge to maintainan adequate iron status: the usual diet does not supplyenough iron to compensate for increased needs. Duringa normal pregnancy of 40 weeks duration, the require-ment for iron is 1,000 milligrams, and the negative bal-ance of iron is estimated to be around 500 milligrams(1). This is equivalent to an exhaustion of the reservesof a normal adult.

Iron deficiency anaemia during pregnancy is well-recognised worldwide (2), affecting developing (3) aswell as developed countries. Severe anaemia (< 8g/dL)(4;5;6) has been shown to be associated to low birth-weight and prematurity. The clinical consequences oflighter forms of anaemia are still a matter of debate.

Mons is a 75,000 inhabitants’ city located in theFrench speaking part of Belgium. The area suffers frommajor economic problems. Four general hospitals coverthe local population, two within the city (a private oneand a public one) and two at a small distance of the city.

During routine supervision of laboratory worksheets,the very frequent occurrence of low serum ferritin lev-els in women attending prenatal clinics raised concerns.

170

Acta Clinica Belgica, 2003; 58-3

IRON DEFICIENCY ANAEMIA DURING PREGNANCY IN MONS, BELGIUM

To study further this observation, the database of a three-year-recording of laboratory data was analysed, in rela-tionship with iron status.

OBJECTIVES OF THE STUDY

- Measurement of the frequency of anaemia in preg-nant women

- Determination of the haematological parameters thatbest predict anaemia during the third trimester.

SUBJECTS AND METHODS

Data collection

Haematological data, age, pregnancy status and preg-nancy age were extracted from the Laboratory Informa-tion System (SYSMEX-Molis, Barchon (Liège), Bel-gium). All data obtained between 1997 and 1999 fromwomen aged between 15 and 44 years was included inthe files. No clinical data was available.

Though pregnancy and pregnancy age are normallyrecorded in the laboratory database, a list of all deliver-ies that took place at the hospital between 1/1/1997 and30/08/2000 was obtained from the hospital files. Thislist includes stillbirths, but does not include abortion anddeliveries taking place outside the hospital (in otherhospitals or infrequently at home). Dates of birth werematched with blood samples using the patient identifi-cation number. A theoretical age of pregnancy was com-puted using pregnancy duration of 280 days (40 weeks)after the first day of the last menstruation period. Thisintroduces an approximation in the age of pregnancy.However as preterm labour of less than 32 weeks areroutinely transferred to another hospital equipped witha neonatal intensive care unit, the error is limited. For141 samples, both the computed age and the recordedage were available. Comparison of both values showedthat in 90% of the samples, recorded age and computedage did not differ by more than 4 weeks. Whenever bothrecorded and computed ages of pregnancy were avail-able, only the youngest age was taken into account.

After matching and computing, all records were madeanonymous and the file was analysed with the SPSS –Statistical Package for Social Sciences, version 10, USA.In discriminant analysis, we use Wilk’s lambda param-eter as measure of the ability of the model to predict anoutcome (here anaemia). Wilks’s lambda varies from 0

to 1 (0 indicating a perfect model and 1 a poor model).

Non pregnant Women Group

It was not possible to have a true reference group asthe data available did not come from a random sampleof the population.

In order to have a rough estimate of the prevalenceof anaemia among non-pregnant women, we selected agroup on the following criteria:• women aged between 15 and 44 years• not pregnant• only one haematological sample taken during the 3-

year period• women attending outpatient clinics not related to pre-

natal careWe also compared that group with samples taken at pe-

riodic occupational health checks.

Study group (pregnant women)

Samples were included in the study if they satisfied thefollowing criteria:• Age between 15 and 44 years• Pregnant with at least 4 weeks of amenorrhea• Age of pregnancy known from the lab file or calcu-

lated from the date of delivery (when both data wereavailable, the lowest figure was taken into account)

• No stay in intensive care unit• Mean Cell Volume less than 105 µm3 (in order to ex-

clude cord blood samples attributed to the mother(clerical mistake)).

Pregnancies were divided in three periods (expressedin weeks of amenorrhea since the first day of the lastmenstrual period):1st trimester: 4 to 15 weeks2nd trimester: 16 to 27 weeks3rd trimester: 28 weeks or more

When a woman had several samples analysed dur-ing a trimester, the lowest value was taken into account.

Biochemistry and Haematology parameters

Erythrocyte count and the dimensions of the bloodcell components were measured on whole blood sam-ples taken on EDTA tubes with an automated device –COULTER STKS, Beckman, Los Angeles – based oncell count and impedance measurement of the variousblood cells components (cell volume, nuclear volume).

171

Acta Clinica Belgica, 2003; 58-3

IRON DEFICIENCY ANAEMIA DURING PREGNANCY IN MONS, BELGIUM

Haemoglobin concentration was measured spectropho-tometrically on the same device. The mean corpuscularvolume (MCV) and other indices of erythrocyte char-acteristics (mean cell haemoglobin (MCH)) were cal-culated automatically.

Serum iron and iron binding capacity were meas-ured spectrophotometrically on an automated biochem-istry device – HITACHI 717, Boehringer-Mannheim,Munich with the reagents of this company.

Serum ferritin was measured by immunoassay(Chiron reagents) on automated device (ACS-180,Chiron, Los Angeles).

All biological measurements were performed withinmaximum six hours after blood collection – except se-rum ferritin, which was measured twice a week.

Internal and external quality control: Daily, two orthree levels of whole blood or serum controls were ana-lysed. These controls were from commercial source (E-check from Sysmex Co. for haematological parameters;FERTC from Roche Co for iron and iron binding capac-ity; FRT from Chiron Co. for ferritin). When the inter-nal control was outside the accepted range, new calibra-tion was implemented. Besides the daily internal con-trols, the clinical laboratory takes part four times peryear in a mandatory national surveillance program forall the tests under the control of the Institut Scientifiquede Santé publique – Louis Pasteur, Brussels, Ministèrede la Santé publique.

Reference values

The reference values followed CDC Recommenda-tions (7).

Anaemia was defined as Haemoglobin lower than12 g/dL for non-pregnant women and 11 g/dL for preg-

nant women. Iron deficiency anaemia was consideredwhen serum ferritin was lower than 12 µg/L. This levelof ferritin provides a specificity of 100% and a sensitiv-ity of 61% for iron deficiency anaemia (7).

RESULTS

Non pregnant women group

The group includes 2,503 women. The age distribu-tion of this group is different from that of the city popu-lation (younger age groups are overrepresented in ourgroup). However, within the range of age 15 to 44 years,there was no effect of age as a confounding variable.So, we did not proceed to any adjustment for age.

Table 1: Haemoglobin distribution in the referencegroup.

Anaemia (haemoglobin level < 12g/dL) was present

Haemoglobin Frequency Percent Cumulative(g/dL) (n) (%) Percent (%)

8.0 – 8.9 4 0.2% 0.2%9.0 – 9.9 12 0.5% 0.6%10.0 – 10.9 32 1.3% 1.9%11.0 – 11.9 144 5.8% 7.7%

12.0 – 15.9 2 305 92.1% 99.8%16.0 – 16.9 6 0.2% 100.0%

Total 2 503 100.0%

Anaemia Normal Sample

8-10.9 g/dL 11-11.9 g/dL >11.9 g/dL Size

Mean CellVolume <85fL 81.3% out of 48 38.8% out of 139 8.3% out of 2 217 2.404

Mean Cell Haemoglobinconcentration <31.4g/dL 25.0% out of 48 5.8% out of 139 0.6% out of 2 217 2.404

Ferritin <12µg/L 40.0% out of 20 28.0% out of 25 6.4% out of 488 533

Serum Iron <40µg/dL 42.9% out of 35 23.1% out of 52 2.9% out of 904 991

in 7.7% of non-pregnant women. Microcytosis (MeanCell Volume (MCV) < 85µl) was present in respectively50% in anaemic women whereas it occurs only in 8%of non-anaemic subjects (chi2 = 292, p <0.001, N=2404).

Table 2: Haematological parameters according to haemoglobin levels among non-pregnant women.

172

Acta Clinica Belgica, 2003; 58-3

IRON DEFICIENCY ANAEMIA DURING PREGNANCY IN MONS, BELGIUM

Ferritin levels were available for approximately 40%of women with haemoglobin level of less than 11g/dLand for less than 20% of all other categories, suggestingthat iron deficiency was specifically sought by the cli-nician who requested the lab test.

Using CDC criteria (ferritin<12µg/L)(7), 40% ofwomen with haemoglobin < 11g/dL and 28% of womenwith haemoglobin between 11.0 g/dL and 11.9 g/dL haveiron deficiency anaemia. However, results must be in-terpreted with caution, as ferritin levels are not knownfor a majority of women.

Serum iron level was available for 73% of womenwith haemoglobin < 11g/dL and for 37% of the othercategories. Low serum iron level show a pattern amonganaemic women similar to low ferritin level.

In order to check the result of non-pregnant women,we compared them with 84 women having an occupa-tional health check. In that group, 8 subjects (9.5%) wereanaemic (haemoglobin < 12g/dL).

STUDY GROUP (PREGNANT WOMEN)

Sample characteristics

In 39% of the pregnancies, only one blood samplewas collected during pregnancy. For the other pregnan-cies, the number of samples varies from 2 to 14. In gen-eral, samples were taken during the first trimester (56%of the pregnancies) and the third trimester (83% of thepregnancies). A blood sample was analysed during thesecond trimester only in 19% of the pregnancies.

The age distribution shows a peak between 25 and29 years of age. In nearly 7% of pregnancies, womenwere aged between 15 and 19 years; this is more thanthe average proportion for the province (3.5% at the timeof delivery in 1996 (8)). In nearly 11% of pregnancies,women were 35 years or older.

Half of the samples came from the French speakingCommunity official organism in charge of mother andchild health (ONE – Œuvre de la Naissance et del’Enfance), a quarter from private practitioner, 15% fromthe maternity ward, 5% from neonatology and the re-mainder came from various hospital wards.

Haemoglobin levels

Figure 1: Frequency of anaemia as a function of theage of pregnancy (N= number of pregnancies forwhich data are available).

Figure 1 indicates for each trimester the proportionof pregnancies where anaemia was present. The propor-tion of women having anaemia (< 11 g/dL) increases withthe age of the pregnancy, reaching 31% in the 3rd trimes-ter. Nearly 9% of the women have an haemoglobin con-centration lower than 10 g/dL during the third trimesterof pregnancy. Haemoglobin < 8 g/dL occurs in only 0.5%of the pregnancies, mostly during the third trimester.

Other haematological parameters

During the first trimester, ferritin was measured inapproximately two thirds (64%) of the pregnancieswhere haemoglobin was measured, regardless of thehaemoglobin level. During the 3rd trimester, this pro-portion reaches only 50% and is also independent fromthe haemoglobin level.

Figure 2: Frequency of low serum ferritin level as afunction of the age of pregnancy (N= number of preg-nancies for which data are available).

173

Acta Clinica Belgica, 2003; 58-3

IRON DEFICIENCY ANAEMIA DURING PREGNANCY IN MONS, BELGIUM

The frequency of low ferritin levels (less than 12 µg/L)increases with the age of the pregnancy and reaches morethan 50% during the 3rd trimester indicating a depletionof iron stock.

Table 3: Low ferritin level (<12µg/L) as a function ofanaemia (haemoglobin < 11g/dL) in pregnant women.

Table 4 : Influence of 1st trimester anaemia on therisk of 3rd trimester anaemia

Haemoglobin level

Anaemia No-anaemiaFerritin (<12µg/L) (HGB<11g/dL) (HGB>=11g/dL)

1° trimester 34.6% out of 26 6.4% out 5453° trimester 75.6% out of 193 45.0% out of 464

In pregnancies without anaemia, iron deficiency ispresent in 6% of the pregnancies during the 1st trimesterand 45% of during the 3rd trimester. During the 1st tri-mester, approximately one third of the anaemic patientsmeet the criteria for iron deficiency anaemia (IDA). Thisproportion reaches three quarters in the 3rd trimester.Taking into account that at the threshold of 12µg/L, fer-ritin concentration has a sensitivity of only 61% forIDA(7), this would mean that most anaemia would bedue to iron deficiency.

During the 1st trimester, microcytosis (MCV < 85fL)is present in 87% when haemoglobin < 10g/dL and in36% when haemoglobin level is between 10.0 g/dL and10.9 g/dL; this proportion is lower during the 3rd trimes-ter (57% and 28% respectively). Hypochromia followsa similar pattern as microcytosis.

Factors predicting 3rd trimester anaemia

In order to prevent IDA at the 3rd trimester, it wouldbe interesting to be able to detect during the 1st trimesterwomen most at risk of anaemia. Presumably, they wouldbenefit most from iron supplementation prescribed atthe first prenatal consultation.

For 656 pregnancies, we had haemoglobin measure-ment both during the first and the third trimester and for413 of them, we also knew ferritin levels during the 1st

trimester.

Anaemia during the first trimester (HGB<11g/dL) isa strong predictor of anaemia during the third trimester(Fischer’s Exact Test: p<0.001). In this series, the posi-tive predictive value of 1st trimester anaemia for 3rd tri-mester anaemia is 79% and the negative predictive valueis 78%. However, the sensitivity reaches only 16% andwhereas specificity is almost perfect (99%). On a clini-cal point of view, it would be more interesting to in-crease sensitivity in order to prevent more 3rd trimesteranaemia. This means choosing a cut-off point higher than11g/dL.

Other haematological parameters were also includedin the analysis. Discriminant analysis was used to selectthe best combination of parameters to predict an hae-moglobin level less than 11g/dL during the 3rd trimester.Adding haematocrit, number of red cells did not im-prove the model compared to haemoglobin alone (Wilk’slambda = 0.84). Among ferritin, iron and transferrin lev-els, ferritin proved to best improve the model when itwas added to haemoglobin level (Wilk’s lambda = 0.78).Adding more than one of those parameters did not im-prove the model.

ROC curves were used to determine the cut-off pointsfor haemoglobin levels and ferritin levels during the 1st

trimester that best predict anaemia during the 3rd trimes-ter. We look for a sensitivity of at least 80% and aspecificity of at least 50%. This means that 80% of thewomen having anaemia could benefit from iron supple-mentation but that half the women who would not needit would actually receive it.

3rd Trimester

1st Trimester Anaemia No Anaemia Total(HGB<11g/dL) (HGB>=11g/dL)

Anaemia(HGB<11g/dL) 27 (79%) 7(21%) 34 (100%)

No Anaemia(HGB>=11g/dL) 139 (22%) 483 (78%) 622 (100%)

Total 166 (25%) 490 (75%) 656 (100%)

174

Acta Clinica Belgica, 2003; 58-3

IRON DEFICIENCY ANAEMIA DURING PREGNANCY IN MONS, BELGIUM

Fig 3 : ROC curve of Ferritin and haemoglobin levelsduring the first trimester of pregnancy as predictorsof 3rd trimester anaemia (haemoglobin < 11g/dL).

More recent tests (such as serum transferrin receptorconcentration (9;10) erythropoietin, zinc protoporphy-rin (percentage of reticulocytes of various (12-14) ma-turity levels) have been suggested to better characteriseanaemia. However, few clinical laboratories are accus-tomed with these tests and it remains to be proven thatthose indicators perform better than more classical onesat a population level.

Frequency of anaemia.

In the group of non-pregnant women, the prevalenceof anaemia (haemoglobin < 12 g/dL) reaches 7.7%among which one third had a ferritin level of 12µg/L orless. This seems to be in the same range as the preva-lence reported by another study in the USA (16). Ournon-pregnant group was not a healthy control group, asthe data came from routine laboratory measurements inoutpatient clinics. Nevertheless, as the age distributionwas restricted to a relatively young population (15 -44 yrs) who only had one blood test in 3 years, we areconvinced that the majority of this population attendedan outpatient clinic for minor ailments.

During pregnancy, the prevalence of anaemia in-creases at each trimester and reaches 31% of the preg-nancies at the 3rd trimester (haemoglobin < 11g/dL). Thisis intermediate between the prevalence for Frenchwomen and migrant women reported by a French study(17).

Microcytic anaemia is not only due to iron deficiency.It may, also, reflect heterozygote thalassemia . In a pre-ceding study including more than three hundred sub-jects of our hospital, heterozygote thalassemia was foundin 2 percent of the subjects – a low proportion as com-pared to the 30.5 percent anaemic women during thethird trimester of pregnancy. So, most of the microcyticanaemia cases were due to iron deficiency, and this wascorroborated by ferritin measurement (11).

Using CDC criteria, iron deficiency anaemia occursin 23% of pregnancies. This is higher than the percent-age reported in an American low-income population18.8% (17) but in the same range as the values reportedin the Netherlands (15). The prevalence of anaemia inour study group seems to be in the high range whencompared to other population based studies in otherindustrialised countries.

The ROC curve shows an area under curve of 0.73for haemoglobin and 0.65 for ferritin, both significantlydifferent from 0.5 (p<0.001). Using only haemoglobinat a cut-off point of 13 g/dL provides a sensitivity of80% and a specificity of 45%. Using only ferritin at acut-off point of 60 µg/L provides a sensitivity of 80%and a specificity of 29%. Defining the at-risk group asthe women who had either an haemoglobin level of lessthan 13.0 g/dL or a ferritin level of less than 20 µg/L,provides a sensitivity of 85% and a specificity of 40%.These cut-off points provide a positive predictive value(PPV) of 36% and a negative predictive value of 87%.

This means that among the women who will developanaemia during the 3rd trimester, 85% were coming fromthe at risk group. In other words, if we supplement thatgroup with iron, we could potentially reduce 3rd trimes-ter by 85% and we will have too treat approximately 3women to prevent one anaemia.

DISCUSSION

Choice of biochemical parameters

This study was based on routine laboratory data. Thebiochemical indicators available included haemoglobinlevel, haematocrit, red cells count (and their calculatedindices mean cell volume, mean cell haemoglobin, meancell haemoglobin concentration), iron , ferritin and trans-ferrin levels.

175

Acta Clinica Belgica, 2003; 58-3

IRON DEFICIENCY ANAEMIA DURING PREGNANCY IN MONS, BELGIUM

Clinical importance of anaemia or iron deficiencyanaemia during pregnancy?

The level at which anaemia is noxious for the motherand the foetus is still a matter of debate (4;6;15;18). Somestudies showed that the haemoglobin level at whichpreterm labour and low birth weight are at their lowestis between 9.5 and 11 g/dL (4;6;15). Others argue thateven light anaemia has a deleterious effect when asso-ciated to iron deficiency (ferritin < 12µg/L).

It is reported that iron deficiency in the mother is arisk factor for iron deficiency in the young infant(5;18-21), with consequent more pronounced “physiological”anaemia around 10 to 12 weeks of age.

Both, plasma expansion and adequate erythropoiesisconcur to a favourable outcome of pregnancy. They actin opposite direction as far as haemoglobin level is con-cerned but act synergistically to lower ferritin level. This“physiological” haemodilution during pregnancy, witha maximum of 50 percent dilution around 32 weeks,can explain that erythrocyte count is lower in pregnantwomen than in paired-for-age not pregnant women. Evenif erythrocyte mass increases, it is insufficient to main-tain a “normal” haematocrit or haemoglobin, and it isaccepted that a 25 percent decrease of haemoglobin orhaematocrit to be considered as physiological duringpregnancy – even when iron supplement has been ad-ministered. Haemodilution does not explain all thevariation of biological parameters associated with preg-nancy: serum ferritin decreases more than expected fromthe 50 percent haemodilution, and previous studies haveshown that it is observed even when iron supply is suf-ficient. Serum ferritin recovers its “normal” range withina few weeks after delivery. This remarkable decrease ofserum ferritin during pregnancy, independent ofhaemodilution or iron status, has no clear explanation.

In our sample, haemoglobin less than 10 g/dL oc-curs in 10% of the pregnancies during the 3rd trimester.It is associated with a ferritin level of less than 12µg/Lin 70% of these pregnancies.

Bias and limits of the study

Our sample is not a random sample of the popula-tion of the region of Mons as the recruitment is basedon the women attending the public hospital of Mons. Itis thought that the public hospital has a greater chanceto recruit lower socio-economic women than privatehospitals – even if there are no objective local data avail-able.

The non-pregnant women group was selected onstringent criteria (outpatients and only one sample col-lected in 3 years). These criteria of selection involve atrend toward subjects in better health than the averagepopulation of the same age; if this bias of selection ispresent, prevalence of anaemia might be underestimated,especially in older age groups.

The study group (pregnant women) includes all thewomen who delivered at the hospital and all the womenwhose pregnancy was recorded when a blood samplewas taken. Half of the blood samples were from the staterun mother and child health service (ONE) and a quar-ter from private practitioners. Women of higher socialstatus are more likely to go to private practice for theirfollow-up. Some pregnancies may have been missed ifthe information about pregnancy was not recorded andif they did not deliver at the hospital. The study groupcan also be considered as a target population for an in-tervention. If self-selection of women at risk of anae-mia has indeed occurred it can be an advantage for anintervention programme.

Results indicate that this hospital-based populationshows a high prevalence of iron deficiency anaemiaduring pregnancy.

Expected benefit of correction of iron deficiency inpregnant women

Population impact of generalised supplementationhas been limited and the efficacy of such a measure isquestioned by several authors (18-23). However, irondeficiency anaemia (IDA) during pregnancy might havea negative impact on mother and child well being. In apopulation where IDA has a high prevalence, it wouldbe interesting to screen pregnant women during the firsttrimester in order to determine those at risk of anaemia.Women with a haemoglobin level of less than 13 g/dLwould most benefit from an iron supplementation. Fer-ritin level could also be measured though in our study itdid not improve sensitivity and specificity.

CONCLUSION

It is beyond the scope of a regional hospital to con-duct large studies on the effect of anaemia on pregnancyoutcomes.

Our study shows with reasonable certainty that irondeficiency anaemia among the pregnant women attend-

176

Acta Clinica Belgica, 2003; 58-3

IRON DEFICIENCY ANAEMIA DURING PREGNANCY IN MONS, BELGIUM

ing the hospital is rather frequent compared to otherpopulation in industrialised countries. It would be inter-esting to conduct similar studies in hospitals having dif-ferent recruitment. Combining the data available in vari-ous laboratories covering different populations couldprovide at low cost a reasonable estimate of the preva-lence of anaemia during pregnancy for a population.Actually, haematological and iron status in pregnantwomen constitute an adequate - and relatively easy toobtain - epidemiological marker of iron status in thewhole population, as part of a nutritional surveillanceprogram at regional or country level.

Iron supplement can clearly correct the biologicalindices of iron deficiency. Whether this will involve aclinical improvement of the mother or of the newborn isstill a matter of debate. A substantial part (9%) of thepregnancies in this hospital shows haemoglobin levelbelow the optimal range as far as prematurity andbirthweight are concerned (10g/dL).

In spite of all the uncertainties, prevention of irondeficiency and treatment of iron deficiency anaemiaseems reasonable in this population.

We suggest that haemoglobin and ferritin are meas-ured during the first trimester and that iron treatment isgiven to those with a haemoglobin lower than 11 g/dL.For pregnant women between 11 and 13g/dL, supple-ment with low doses (30 mg elemental iron per day)may be proposed in order to raise iron stocks. Low dosesare associated with a decreased risk of side effects andwith better compliance (24-29).

In the absence of a clear consensus, further studiesare needed on the best regimen (dose, timing of supple-mentation during pregnancy and duration) as far as com-pliance and effect on haematological parameters areconcerned.

RÉSUMÉ

L’anémie de la grossesse est un problème fréquentdans le monde. Dans les pays industrialisés, elle estfréquente dans certains groupes de population. Cetteétude se base sur l’analyse rétrospective des résultatsd’analyses de sang de femmes âgées de 15 à 44 ansfréquentant la consultation prénatale et d’autres serv-ices ambulatoires d’un hôpital public de Mons, Belgiquede 1997 à 1999.

Chez les femmes non gravides (2503 sujets), l’anémie(hémoglobine < 12g/dL) était présente chez 7,7% des

femmes. Pendant la grossesse, l’anémie était définiecomme un taux d’hémoglobine inférieure à 11g/dL. Dansnotre échantillon, l’anémie était présente au 1e trimestrechez 4,3% des 887 grossesses Parmi ces cas, 35%correspondait aux critères d’anémie ferriprive du CDC(ferritine <12 µg/L). La fréquence de l’anémie augmentedurant la grossesse. Sur 1313 grossesses, elle atteignait31% au troisième trimestre. Parmi ces anémies, 75%répondait aux critères de l’anémie ferriprive.

Des taux d’hémoglobine ou de ferritine bas au 1e

trimestre sont de bons indicateurs de l’anémie du 3e

trimestre.L’administration systématique de fer durant la

grossesse est un sujet controversé. Pour limiter lasupplémentation en fer aux femmes les plus à risqued’anémie du 3e trimestre, nous suggérons de traiter lesanémies (hémoglobine < 11g/dL) détectées lors de lapremière visite prénatale et de donner de faibles dosesde fer (30mg par jour) quand le taux d’hémoglobine estcompris entre 11g/dL et 13g/dL ou une ferritineinférieure à 20µg/L. A ces doses, le fer est généralementbien toléré et la compliance est meilleure.

REFERENCES

1. Fraipont V, Beguin Y, Fillet G. L’anémie ferriprive de lagrossesse. Rev Med Liege 1994; XLIX(8):436-45.

2. Baker WF, MD, FACP. Iron Deficiency in pregnancy, obstet-rics, and gynecology. Hematol Oncol Clin North Am 2000;14(5):1061-77.

3. Rush D. Nutrition and maternal mortality in the developing world.Am J Clin Nutr 2000; 72(suppl):212S-40S.

4. Steer P. Maternal haemoglobin concentration and birth weight.Am J Clin Nutr 2000; (7(suppl)):1285S-87S.

5. Blot I, MD, Diallo D, Tchernia G. Iron deficiency inpregnancy:effects on the newborn. Current Opin Hematol 1999;6:65-70.

6. Rasmussen K. Iron-Deficiency Anaemia : Reexaminig the Na-ture and Magnitude of the Public Health Problem - Is there aCausal Relationship between Iron Deficiency or Iron-DeficiencyAnaemia and Weight at Birth, Length of Gestation and PerinatalMortality? J Nutr 2001; (Supplement):590S-603S.

7. Recommendations to Prevent and Control Iron Deficiency inthe United States. Mor Mort Wkly Rep CDC Surveill Summ 1998;47(RR-3):1-29.

8. Massot C, Berghmans L, Godin I, Coppieters Y, Thauvoye N.Santé en Hainaut - Tableau de bord de la santé. Edition. 3, Mons,Observatoire de la Santé du Hainaut, 2000;: 1-60.

9. A. Akesson, P. Bjellerup, M. Berglund, K. Bremme, and M.Vahter. Serum transferrin receptor: a specific marker of iron de-ficiency in pregnancy. Am J Clin Nutr 1998; 68 (6):124-46.

10. U. Rusia, C. Flowers, N. Madan, N. Agarwal, S. K. Sood, andM. Sikka. Serum transferrin receptors in detection of iron defi-ciency in pregnancy. Ann.Hematol. 1999; 78 (8):358-63.

177

Acta Clinica Belgica, 2003; 58-3

IRON DEFICIENCY ANAEMIA DURING PREGNANCY IN MONS, BELGIUM

11. Tshilolo L, Gulbis B, Barlow P, Damis E, Denis-Song R, FersterA, Haumont D, Vanderpas JB, Pardou A, Schwers J, Vokaer A,Vertongen F. Dépistage néonatal des hémoglobinopathies dansla région bruxelloise. [Neonatal screening for hemoglobino-pathies in the Brussels region] Rev Med Brux. 1997;18(2):70-3.

12. N. Madan, U. Rusia, A. Manocha, and S. K. Sood. Detection ofiron deficiency in pregnancy by haemoglobin, serum ferritin &protoporphyrin/haem ratio. Indian J Med.Res 1988; 87:245-51.

13. Janet Stein and Paul Whittaker. Volume-corrected ferritin andzinc protoporphyrin/heme ratio during pregnancy: the role of anintracellular iron measure. Obstet Gynecol 1998; 725-725.

14. Bianco, F. Mastropietro, C. D’Asero, B. Graziani, P. Piergrossi,M. Mezzabotta, and G. Modiano. Serum levels of erythropoi-etin and soluble transferrin receptor in the course of pregnancyin non beta thalassemic and beta thalassemic women.Haematologica 2000; 85 (9):902-7.

15. Roodenburg A. Iron supplementation during pregnancy. Eur JObstet Gynecol Reprod Biol 1995; (61):65-71.

16. Walker A. Review article - The remedying of iron deficiency:what priority should it have? Br J Nutr 1998; 79:227-35.

17. Société française de pédiatrie-Comité de nutrition. Fer etgrossesse. Arch Pédiatr 1995; 2:1209-18.

18. Allen LH. Anaemia and iron deficiency:effects on pregnancyoutcome. Am J Clin Nutr 2000; 71(suppl):1280S-84S.

19. Harthoorn-Lasthuizen E, Lindemans J, Langenhuijsen MMAC.Does iron-deficient erythropoiesis in pregnancy influence fetaliron supply? Acta Obstet Gynecol Scand 2001; 80:392-96.

20. Preziosi P, Prual A, Galan P, Daouda H, Boureima H, HercbergS. Effect of iron supplementation on the iron status of pregnantwomen:consequences for newborns. Am J Clin Nutr 1997;66:1178-82.

21. Halvorsen S. Iron balance between mother and infant duringpregnancy and breastfeeding. Acta Paediatr 2000; 89:625-27.

22. Yip R. Significance of an abnormally low or high haemoglobinconcentration during pregnancy:special consideration of ironnutrition. Am J Clin Nutr 2000; 72(suppl):272S-79S.

23. Lindsay H, Allen PHD. Pregnancy and Iron Deficiency: Unre-solved Issues. Nutr Rev 1997; 55(4):91-101.

24. Eskeland B, Malterud K, Ulvik R, Hunskaar S. Iron supplemen-tation in pregnancy:is less enough?A randomized, placebo controlled trial of low dose iron supple-mentation with and without heme iron. Acta Obstet GynecolScand 1997; 76:822-28.

25. Marx J. Iron deficiency in developed countries: prevalance, in-fluence of lifestyl factors and hazards of prevention. Eur J ClinNutr 1997; 51:491-94.

26. Scholl TO, Reilly T. Symposium: Improving Adolescent IronStatus before Childbearing Anaemia, Iron and Pregnancy Out-come. American Society for Nutritional Sciences 2000; 443S-47S.

27. Beard JL. Effectiveness and strategies of iron supplementationduring pregnancy. Am J Clin Nutr 2000; 71(suppl):1288S-94S.

28. Yip R. Iron supplementation during pregnancy : is it effective?Am J Clin Nutr 1996;(63):853-55.Milman N, Bergholt T, Byg K-E, Eriksen L, Graudal N. Ironstatus and iron balance during pregnancy. A critical reappraisalof iron supplementation. Acta Obstet Gynecol Scand 1999;78:749-57.