Pregnancy induces fundamental changes in thyroid function and iodine metabolism leading to thyroid stimulation(Reference Glinoer1). The stimulation of the thyroid gland during pregnancy stems from the rise of oestrogen concentrations resulting in an increase of serum thyroxin-binding globulin, an increase in the renal clearance of iodide, the iodide transfer to the fetus, the direct stimulation of the thyroid by the human chorionic gonadotrophin and finally resulting in changes in the peripheral metabolism of maternal thyroid hormones under the influence of the placenta. Iodine-sufficient pregnant women meet their thyroid hormone requirement by increasing the thyroid iodide intake, maintaining a plentiful store of iodine in the thyroid. In iodine-deficient women, this adaptive mechanism fails to maintain adequate iodine stores, which can lead to thyroid dysfunction in the pregnant women and their newborn. Some studies have suggested that mild iodine deficiency during pregnancy may impair neuropsychological development(Reference Haddow, Palomaki and Allan2) and psychomotor(Reference Pop, Kuijpens and van Baar3) development in the offspring, preventing them from reaching their full intellectual potential(Reference Zimmermann4). However, the impact of mild-to-moderate maternal iodine deficiency on cognition of the offspring is still unclear as few controlled intervention studies have measured long-term clinical outcomes(Reference Zimmermann5). Correction of mild-to-moderate iodine deficiency via supplementation in primary school children was found to improve cognitive and motor function(Reference Gordon, Rose and Skeaff6).
Despite a worldwide successful implementation of iodine supplementation programmes over the last decades, iodine deficiency remains a public health problem in Europe(7, Reference De Benoist, McLean and Andersson8). According to the WHO, the median urinary iodine concentration (UIC) among pregnant women within the range 150–249 μg/l is indicative of iodine sufficiency(9). Although, since 2003, the number of European countries in which iodine deficiency remains a public health problem decreased from twenty-three to fourteen(Reference Zimmermann and Andersson10), it is a matter of concern that iodine deficiency has reappeared in countries thought to have sufficient iodine intake, such as the UK(Reference Vanderpump, Lazarus and Smyth11). Several surveys have indicated that Belgium is affected by mild iodine deficiency, and that this represents a substantial economic burden to the Belgian healthcare system(Reference Vandevijvere, Annemans and Van Oyen12–Reference Ciardelli, Haumont and Gnat14). Consequently, optimising iodine intake was chosen as a priority in the nutritional policy of the Belgian Ministry of Health.
An agreement was signed between the bakery sector and the Ministry of Health in April 2009, to encourage the fortification of bread with iodised salt (voluntary fortification)(Reference Moreno-Reyes, Van and Vandevijvere15).
A recent national survey on iodine status in Belgian school-aged children has shown a median UIC of 113·1 and 84·8 μg/l among their mothers, indicating iodine sufficiency among the children and suggesting that the voluntary fortification of bread with iodised salt may have contributed to the optimisation of iodine intake in school-aged children but not in the adult population(Reference Vandevijvere, Mourri and Amsalkhir16).
The aim of this first national representative study was to assess iodine status among pregnant women in Belgium during the first and third trimester of pregnancy, and to examine the determinants of iodine status in a mild iodine-deficient area, 1 year after the introduction of bread fortification with iodised salt.
Methods
Sampling
The target population of the survey comprised all pregnant women in Belgium during the first or third trimester of pregnancy in the period of the survey from September 2010 to June 2011. The women were selected according to a multistage proportionate-to-size stratified sampling design. Because some population-based data suggest that the prevalence of iodine deficiency is higher in the south than in the north of the country(Reference Vandevijvere, Dramaix and Moreno-Reyes17, Reference Brull and Dewart18), the hospitals were stratified by region. A thirty-cluster survey was performed in both regions with at least twenty women per cluster, as recommended(Reference Sullivan19). A sample size calculation based on an estimated 70 % prevalence of UIC < 150 μg/l(Reference Delange, Van Onderbergen and Shabana13), a 95 % CI for the true prevalence of UIC < 150 μg/l, a design effect of 2 and an absolute precision of 5 % resulted in a sample size of 645 women per region (1290 women, in total) or about twenty-two women per hospital: eleven in the first trimester and eleven in the third trimester of pregnancy.
In each region, the hospitals were ordered by province and size based on the number of deliveries during the past year and sixty clusters of four hospitals were selected using systematic sampling while accounting for the number of deliveries in order to have enough replacement hospitals in case hospitals refused to participate. Of these sixty clusters, thirty clusters were randomly selected and within each cluster, the first hospital was invited to participate. In each hospital, all gynaecologists were invited to participate in order to level out a possible gynaecologist effect.
Data collection
The present study was conducted according to the guidelines laid down in the Declaration of Helsinki and all procedures involving human subjects were approved by the medical ethical committee of the Erasme Hospital in Brussels. Written and verbal informed consent was obtained from all pregnant women. The fieldwork took place between September 2010 and June 2011. Urine samples were collected and a general questionnaire about sociodemographic and socio-economic characteristics, smoking and alcohol-drinking behaviour, thyroid diseases, use of iodine-containing supplements, consumption frequency of seaweed, fish and seafood, eggs, milk and dairy drinks, yoghurt, cottage cheeses and cheeses, consumption quantity of fish and seafood, milk and dairy drinks, yoghurt, cottage cheeses and cheeses, usual number of slices of bread consumed per d and use of iodised household salt was completed face to face with the study nurse. For one hospital (n 23 women), certain questions (mainly nationality, ethnicity and education level) needed to be omitted from the questionnaire upon decision of the ethical committee of this particular hospital. BMI was obtained from weight and height recorded by the gynaecologist during the first consultation for women in both the first and third trimester of pregnancy.
There were sixteen categories to report education level and they were recoded into six categories: secondary education or lower; higher education; university or higher; other education; no education; not known.
Analysis of samples
All urine samples were frozen and kept at − 80 °C until analysis. UIC were measured in duplicate at Erasme Hospital using a modification of the Sandell–Kolthoff reaction with spectrophotometric detection(Reference Pino, Fang and Braverman20). The sensitivity of the assay was 12 μg/l. The Erasme laboratory participated successfully in the Program to Ensure the Quality of Urinary Iodine Procedures of the US Centers for Disease Control and Prevention(Reference Caldwell, Makhmudov and Jones21). Urinary creatinine was determined by a colorimetric method based on the Jaffe reaction(Reference Hervey22).
Statistical analyses
Statistical analyses were carried out using STATA 10.1 (StataCorp). As UIC is not normally distributed, non-parametric methods were used. Differences in UIC between regions, trimesters and age groups were explored using the two-sample Wilcoxon rank-sum test or the Kruskal–Wallis equality-of-populations rank test. The odds of having a UIC lower than 150 μg/l v. an optimal iodine status were estimated by multiple logistic regression while controlling for age, trimester of pregnancy, region, BMI, smoking behaviour, alcohol consumption, use of iodine-containing food supplements, use of iodised household salt, bread consumption, fish consumption, milk and dairy drink consumption, education level, ethnicity and parity.
Results
Among the 1311 pregnant women participating in the survey, 214 were from Brussels, 640 from Flanders and 455 from Wallonia (Fig. 1). For two women, the region was missing. The mean age of women was 28·5 (sd 5·1) years (n 1305) and was similar in all three regions. For 1307 women, a general questionnaire was available. The mean BMI of women was 24·4 (sd 5·1) kg/m2. More than 20 % of the women in the sample were from non-Caucasian origin and more than 50 % of the women included had a lower education level. Of all women included, 15 % smoked during pregnancy and 12 % reported having drunk alcohol during pregnancy (Table 1).
Fig. 1 Geographical distribution of the fifty-five hospitals visited in Belgium and the number of pregnant women (n 1311) investigated by site (national survey on the iodine status of pregnant women, Belgium 2010–11).
Table 1 Characteristics of the pregnant women included in the study (Belgian national survey on micronutrient status in pregnant women, 2010–11) (Number of participants and percentages, n 1311)
* Values were significantly different from those of the first trimester (P< 0·001).
There were 640 women in the first trimester of pregnancy while 666 were in the third trimester of pregnancy and two were in the second trimester of pregnancy. For 41·7 % of the women, this was their first pregnancy, and 45·6 % of the women had been pregnant before. For the other women, this information was missing. Nearly 4·0 % of the women had had a miscarriage at least once. Interestingly, 77·6 % of the women stated that their pregnancy was planned. Of all women included, thirty-six (2·7 %) reported suffering from a thyroid disease: eighteen suffered from hypothyroidism; four from hyperthyroidism; six from goitre or nodules; one from thyroiditis; the disease was unspecified for the other women.
The median UIC among the pregnant women was 124·1 μg/l and even lower when corrected for urinary creatinine (122·6 μg/g creatinine). Of all pregnant women included, 59·3 % had a UIC below 150 μg/l, 37·8 % had a UIC below 100 μg/l while 18·4 % had a UIC above 249 μg/l and 3·7 % of the women had a UIC above 500 μg/l. The median UIC in the first trimester of pregnancy was significantly lower than that in the third trimester (118·3 and 131·0 μg/l, respectively). These differences were even more pronounced when the UIC was corrected for urinary creatinine (106·2 and 139·6 μg/g creatinine, respectively; Table 2). Compared with pregnant women, the median UIC in women of childbearing age was significantly lower, even when the UIC was corrected for urinary creatinine (Fig. 2).
Table 2 Distribution of urinary iodine concentrations (UIC, μg/l and μg/g creatinine) in pregnant women (n 1299) (national study on iodine status among pregnant women in Belgium, 2010) (Mean values and standard deviations; medians, inter quartile ranges (IQR), percentages and 95 % confidence intervals)
*** Mean values were significantly different from the first trimester (P< 0·001).
††† Mean values were significantly different from Wallonia (P= 0·040).
‡‡‡ Mean values were significantly different from Wallonia (P< 0·001).
Fig. 2 Median urinary iodine concentration (UIC) (in μg/l (
) and μg/g creatinine (
)) in women of childbearing age(Reference Vandevijvere, Mourri and Amsalkhir16) and during the first and third trimester of pregnancy in Belgium. The upper horizontal dotted line represents the lower threshold indicating iodine deficiency in pregnant women (150 μg/l). The lower horizontal dotted line represents the lower threshold indicating iodine deficiency in women of childbearing age (100 μg/l). * Median UIC values were significantly different from those of the first and third trimester pregnant women (P< 0·001).
The median UIC in Flanders and Wallonia were not significantly different. However, the UIC values corrected for urinary creatinine were significantly lower in Wallonia than in Flanders (112·0 and 124·2 μg/g creatinine, respectively; Table 2). The median UIC was highest in the oldest age group (>31 years, 132·2 μg/l), and lowest in the youngest age group ( ≤ 26 years, 116 μg/l) (P< 0·001).
Most of the women reported taking iodine-containing multivitamins during pregnancy (Table 1). The percentage of women taking iodine supplements was significantly higher in the third than in the first trimester of pregnancy. The majority of women started taking iodine supplements during pregnancy and only 12 % of them started taking iodine supplements before pregnancy (Table 2). There were no differences among the regions in the use of iodine-containing supplements and the frequency of intake; however, in Flanders, the percentage of women starting to take iodine supplements before pregnancy was significantly higher than in Wallonia.
Of all women included in the survey, 87·2 % reported using non-iodised salt or sea salt, while 11·2 % reported using iodised household salt and 1·6 % reported using a combination of iodised and non-iodised salt.
The risk of iodine deficiency was investigated by logistic multivariate analysis (Table 3). The risk of iodine deficiency was significantly higher in younger women, in women not taking iodine-containing supplements and in women consuming milk and dairy drinks less frequently. In addition, the risk of iodine deficiency was significantly higher in autumn compared with winter. When a second model was constructed to asses the risk of iodine deficiency using the UIC values corrected for urinary creatinine, BMI, trimester of pregnancy and alcohol consumption appeared as determinants of iodine status (Table 4). The risk of iodine deficiency was higher in women with a higher BMI and women in the first trimester of pregnancy. Pregnant women who reported alcohol consumption during pregnancy had a lower risk of iodine deficiency (Table 4). Age, use of iodine-containing supplements, milk and dairy drink consumption frequency and season remained as determinants of iodine status in pregnant women as in the first model using uncorrected UIC values.
Table 3 Risk of iodine deficiency during pregnancy in Belgium (urinary iodine concentration (UIC) <150 μg/l, n 1229) (Odds ratios and 95 % confidence intervals)
Table 4 Risk of iodine deficiency during pregnancy in Belgium (urinary iodine concentration (UIC) <150 μg/g creatinine, n 1202) (Odds ratios and 95 % confidence intervals)
Discussion
The present results indicate that pregnant women in Belgium are iodine deficient based on the median UIC, as recommended by the WHO. Although about 60 % of pregnant women reported taking iodine-containing multivitamins during pregnancy, the median UIC did not reach the lower recommended threshold of 150 μg/l, indicating iodine sufficiency in pregnant women. More than 50 % of urine samples had a UIC below 150 μg/l.
As this is the first national survey on iodine nutrition in Belgian pregnant women, the evolution of iodine status over time cannot be accurately estimated. However, a previous small-scale study in Brussels in 1990 reported a median UIC of 56 μg/l in pregnant women(Reference Glinoer1). The median UIC reported in the present national survey was 124·1 μg/l, suggesting that iodine intake in pregnant women has increased over time. However, unlike the population of school-aged children(Reference Vandevijvere, Mourri and Amsalkhir16), iodine intake remains insufficient in some pregnant women. Other studies have also shown that despite the optimal median UIC in school-aged children, the median UIC in their mothers, in pregnant women or weaning infants may indicate iodine deficiency(Reference Gowachirapant, Winichagoon and Wyss23, Reference Andersson, Aeberli and Wust24).
The median UIC was significantly higher in the third trimester compared with the first trimester or with women of childbearing age. The increase in UIC during pregnancy can be attributed to the intake of iodine-containing supplements. The percentage of women taking iodine-containing supplements was lower in the first trimester than in the third trimester of pregnancy. The failure of iodine supplements to optimise iodine intake in pregnant women stems from the fact that iodine deficiency was probably present before pregnancy as suggested by the median UIC of 84·8 μg/l in women of childbearing age(Reference Vandevijvere, Mourri and Amsalkhir16). During pregnancy, the daily thyroid hormone and iodine requirement increases because of the increase in the renal clearance of iodine and because of the transfer of iodine to the fetus, aggravating therefore a pre-existent iodine deficiency. There are only a few studies with enough power to classify median UIC by trimester of pregnancy and even fewer studies comparing UIC with UIC of women of childbearing age(Reference Zimmermann25).
In iodine-deficient areas, the median UIC during pregnancy remains constant or even decreases in the absence of iodine-containing supplements(Reference Limbert, Prazeres and Sao26, Reference Stilwell, Reynolds and Parameswaran27). In areas with iodine sufficiency, the UIC during pregnancy increases and the median UIC is generally above 150 μg/l(Reference Hollowell, Staehling and Hannon28, Reference Caldwell, Jones and Hollowell29). The iodine status in pregnant women in Belgium depicted a situation in which the recommendations to take iodine-containing supplements work relatively well as UIC are higher in the third trimester of pregnancy. However, because of the inadequate low iodine intake before pregnancy, the daily administration of 150 μg iodine is not sufficient to reach iodine sufficiency during pregnancy.
This situation illustrates the importance of monitoring the iodine status of pregnant women even when iodine supplements are routinely given to pregnant women.
Many factors may affect urinary iodine excretion during pregnancy. Obviously, the most important is iodine intake and whether or not iodine supplements are given. In the case of iodine sufficiency before pregnancy and in the absence of systematic iodine supplements, median UIC is likely to remain unchanged during pregnancy as suggested by a study from Switzerland(Reference Andersson, Aeberli and Wust24).
Because glomerular filtration rate increases during pregnancy, the UIC:creatinine ratio, which minimises the variation due to dilution and urine volume, may be more appropriate than UIC expressed as μg/l to assess iodine intake in pregnant women. However, there are only a few studies comparing UIC:creatinine during pregnancy with a sufficient number of subjects in the first and third trimester of pregnancy to really conclude(Reference Fuse, Ohashi and Yamaguchi30, Reference Brander, Als and Buess31). In Belgium, the UIC:creatinine ratio in women of childbearing age and among women in the first trimester of pregnancy was lower than the UIC expressed in μg/l. By contrast, in the third trimester, the UIC:creatinine ratio was higher than the uncorrected UIC. These findings are consistent with a study from an iodine-sufficient area from Japan showing lower median UIC:creatinine during the first trimester and higher during the third trimester compared with UIC not corrected for urinary creatinine(Reference Fuse, Ohashi and Yamaguchi30).
The current iodine status, particularly during early pregnancy, is a matter of concern as most of the women in Belgium are iodine deficient when they become pregnant and they remain deficient during pregnancy, even if the iodine status improves during the third trimester. The exacerbation of iodine deficiency during pregnancy may lead to maternal hypothyroxenaemia and enhanced stimulation of the thyroid gland(Reference Glinoer, De and Bourdoux32).
Thyroid volume of mildly iodine-deficient pregnant women and their newborn increases during pregnancy as shown in a study performed in Belgium(Reference Glinoer, De and Bourdoux32) and other European countries(Reference Smyth, Wijeyaratne and Kaluarachi33). In addition, goitrogenesis during pregnancy can be prevented by iodine supplementation during pregnancy(Reference Glinoer, De and Delange34). Furthermore, brain maturation of the fetus needs thyroid hormones.
However, fetal thyroid secretion occurs only during the second trimester of pregnancy; therefore, maternal thyroxine is the only source of thyroid hormones during the first trimester. Interestingly, several epidemiological studies and some clinical studies have reported an association between intelligence quotient and maternal mild iodine deficiency during pregnancy(Reference Azizi, Afkhami and Sarshar35–Reference Vermiglio, Lo Presti and Moleti37).
The median UIC increased with the age of the pregnant women, probably because of the fact that older pregnant women are better informed about food supplements or because of a previous pregnancy. Iodine status was significantly higher in Flanders than in Wallonia based on UIC:creatinine. This finding corroborated a previous study among another Belgian population group(Reference Vandevijvere, Dramaix and Moreno-Reyes17). The model investigating the risk of iodine deficiency was explained by the use of iodine-containing supplements and the consumption of milk and dairy drinks. Dairy products are the main source of iodine in Belgium as in many other industrialised countries(Reference Vandevijvere, Lin and Moreno-Reyes38). No significant associations were found between an optimal iodine status and fish intake. There were no significant differences in iodine status among pregnant women from different ethnic origin and socio-economic status, as shown previously in Brussels(Reference Moreno-Reyes, Carpentier and Macours39).
The risk of iodine deficiency was higher in autumn than in winter as reported in other European countries(Reference Nawoor, Burns and Smith40, Reference Als, Haldimann and Burgi41). The variation of iodine intake with season has been attributed to variations of the iodine content in milk. During the winter months, cattle are housed indoors and fed with iodine-containing food.
When the UIC:creatinine ratio was used in the multivariate analysis, all the determinants associated with iodine status in the first model using the uncorrected UIC values remained significantly associated with iodine status. Interestingly, the second model using the UIC:creatinine ratio yielded new determinants of iodine status: BMI and alcohol consumption. Pregnant women who reported alcohol consumption had a lower risk of iodine deficiency. This is the first time that such an association is reported in pregnant women. In a study among adults from Denmark, an association between a reduced prevalence of goitre and solitary nodules and alcohol consumption has been reported(Reference Knudsen, Bulow and Laurberg42). Finally, the higher risk of iodine deficiency in women with a higher BMI may also explain the association between thyroid function and BMI in a mildly iodine-deficient area(Reference Knudsen, Laurberg and Rasmussen43).
These associations have not been reported before, probably because this is the first study gathering at the same time enough statistical power, detailed nutritional questionnaires and a correction of UIC for urinary creatinine. Alcohol intake may relate to creatinine or an imprecision of the creatinine correction with a different alcohol intake(Reference Chung, Yang and Shieh44).
The association between BMI and iodine could be caused by differences in creatinine rather than iodine excretion. This is plausible as a higher BMI associates with a higher muscle mass and hence a higher creatinine clearance(Reference Gerchman, Tong and Utzschneider45).
The current policy in Belgium to optimise iodine status implemented since 2009 is based on the fortification of bread with iodised salt and the use of iodised table salt. This policy is voluntary and, currently, only 44 % of bread is fortified with iodised salt (source: ESCOSALT). In addition, the use of iodised table salt remains low in Belgium; only 37 % of households use iodised salt(Reference Vandevijvere, Mourri and Amsalkhir16). The use of iodised salt among pregnant women was even lower, although, in the present survey, the use was self-reported and some misreporting may have occurred. The current voluntary strategy seems, however, to work well for school-aged children as shown previously(Reference Vandevijvere, Mourri and Amsalkhir16) but not for the adult population. To optimise the iodine status in pregnant women, an increase of iodine intake in women of childbearing age is necessary in order to reach a daily intake of 150 μg iodine well before pregnancy. This objective may be reached by increasing the number of bakers using iodised salt in the production of bread and by increasing the utilisation of iodised instead of non-iodised table salt by the general population.
The study has several limitations. The individual iodine intake could not be determined as only one spot urine sample was collected for each woman. The use of iodised salt among pregnant women was self-reported and no salt sample was collected to determine the iodine content. In addition, although hospitals were selected using a proportionate-to-size stratified sampling design; women in each hospital were included until quotas were reached. Finally, the number of women who refused to participate in the study was not recorded. The main strengths of the study were the nation-wide representative sample, the high response rate of hospitals and the fact that women of lower socio-economic classes were not under-represented in the present study as is the case in many other studies.
In conclusion, iodine-containing multivitamins taken during pregnancy increase urinary iodine excretion in Belgian pregnant women, underlying the necessity to promote iodine supplements ideally starting before pregnancy. However, despite the fact that about 60 % of pregnant women reported taking iodine supplements, the median UIC during pregnancy indicated iodine deficiency. Furthermore, some groups of women are at a higher risk of iodine deficiency. The current low iodine intake in women of childbearing age in Belgium precludes the correction of iodine deficiency in pregnant women supplemented with multivitamins containing 150 μg iodine as recommended. The iodine status in women of childbearing age needs to be increased in order to meet the iodine requirement during pregnancy. A more generalised use of iodine-containing multivitamins during pregnancy, iodised instead of non-iodised table salt and bread fortified with iodised salt is necessary to optimise the iodine intake in women of childbearing age and in pregnant women.
Acknowledgements
The authors acknowledge the financial support from the Federal Public Service of Health, Food Chain Safety and Environment. We thank the hospitals and gynaecologists who agreed to participate and all participating pregnant women. The authors would also like to acknowledge B. Hauquier and D. Martin for the urinary iodine determinations. S. V. and R. M.-R. coordinated the study and drafted the manuscript. S. V. performed the statistical analyses. S. A. participated in the fieldwork. All authors contributed to the data interpretation of the results and critically revised the draft versions of the manuscript. The authors declare that there are no conflicts of interest with regard to this study.
