Effects of iron supplementation to iron depleted and iron replete pregnant Danish women: Defining criteria for identification of women who can manage without supplements: A randomized,placebo-controlled study

2.1 Women

The women were consecutively attending the Birth Clinic at Herning Hospital and were included in the iron supplementation study within 14–18 weeks of gestation. Gestational age was determined by ultrasound. At inclusion and in the subsequent gestation and postpartum period, women had their iron status markers, S-ferritin, and percentage S-transferrin saturation (TSAT), Hb as well as S-erythropoietin (S-EPO) and S-human placental lactogen (S-HPL) monitored at regular intervals. The final criteria for inclusion in this subanalysis were: healthy ethnic Danish pregnant women with a normal single pregnancy and a normal delivery with an estimated blood loss of less than 500 mL [16].

The women were allocated into two treatment arms. In the iron arm, women were told to take one tablet containing ferrous fumarate 200 mg (66 mg elemental iron) daily and in the placebo arm women were recommended to take a similarly looking placebo tablet daily. Placental weight was recorded after delivery.

From the original series of 119 women [17] four groups were segregated as shown in Table 1; Group 0, having S-ferritin < 30μg/L, should be considered as a “control” group, while Groups I-III having S-ferritin≥30μg/L and Hb≥110 g/L were the principal study groups. From the original series [17], 6 women in Group I were excluded due to Hb levels < 110 g/L at inclusion –so the total number of women included in the present study is 113 (Table 1).

Cord blood was obtained immediately after birth and serum immediately frozen at –25°C for subsequent analysis of S-ferritin. Birth weight was recorded.

2.3 Statistics

Data were entered into the statistical program MedCalc^® [18] and checked for normality using the D’Agostino-Pearson test. In normally distributed variables, arithmetic means±SD (standard deviation) are quoted, and Student’s t-tests for paired and unpaired values and Pearson’s correlation coefficient r were used. In non-normally distributed variables, medians and 2.5–97.5 percentiles are quoted. Qualitative data were evaluated by Fisher’s exact test. Logarithmic (log₁₀) transformation was performed in the non-normally distributed variables S-ferritin and S-EPO, which became normally distributed after this procedure. After calculation of the arithmetic mean and SD of log₁₀ values, the geometric mean was calculated as antilog₁₀ values of arithmetic mean. A p-value of≤0.05 was considered statistically significant.

2.4 Iron status markers, serum erythropoietin and serum human placental lactogen

Blood samples were obtained by venipuncture in the non-fasting state. Hb was analyzed immediately on Coulter-S^®; the intra-assay coefficient of variation was 0.5–0.6% [19]. Conversion between Hb units is: Hb in g/L×0.0621 = Hb in mmol/L. Serum samples were frozen immediately at –25°C and analyzed together after closure of the study. S-ferritin was measured by a radioimmunoassay (Ferritin RIA Amersham®), which was calibrated according to the WHO Human Liver Ferritin International Standard 80/602; the intra-assay coefficient of variation was 6.8% at a S-ferritin concentration of 13μg/L and 4.5% at concentrations > 30μg/L [20]. S-iron, S-transferrin and TSAT were analyzed by previously described methods [16]. S-EPO was measured by an in-house radioimmunoassay [16] and S-HPL by a commercially available radioimmunoassay.

2.5 Definition of iron status

S-ferritin values < 15μg/L are considered to indicate depleted or absent body iron reserves and indicative of iron deficiency (ID) [21, 22]. The author has chosen S-ferritin values < 30μg/L to define the presence of small or depleted iron reserves, as bone marrow biopsies in persons having S-ferritin values < 30μg/L show no stainable hemosiderin iron, while S-ferritin values≥30μg/L are indicative of body iron reserves with stainable hemosiderin iron [21, 24]. Hb values < 110 g/L (6.8 mmol/L) are considered consistent with anemia during pregnancy [25–27]. The combination of Hb < 110 g/L and ferritin < 15μg/L is considered consistent with iron deficiency anemia (IDA). At 8 weeks postpartum the discriminatory Hb value for anemia is < 120 g/L (7.4 mmol/L) [26, 27]. A TSAT < 15% is considered indicative of insufficient supply of iron to the tissues, including the erythropoietic tissue [28].

3.1 Pregnant women

The iron- and placebo subgroups were comparable concerning gestational age at inclusion and body mass index (BMI). The length of gestation in the series of 80 women in Group I, having S-ferritin≥30μg/L and Hb > 110 g/L, was mean 280±10 days, being similar in the iron- and placebo group. Likewise, placental weight and prepartum S-HPL were not significantly different between the two groups (Table 2).

Placental weight displayed a significant correlation with prepartum S-HPL in all the women in Group I (r = 0.53, p < 0.0001), as well as in the iron treated- (r = 0.52, p = 0.0003) and placebo treated subgroup (r = 0.50, p = 0.0028). Furthermore, prepartum S-HPL displayed a positive correlation with newborns body weight in all women in Group I (r = 0.33, p = 0.0029) as well as in the iron- (r = 0.25, p = 0.11) and placebo subgroup (r = 0.41, p = 0.0128). These correlations were essentially similar in Groups 0, II and III, but are not shown due to the small number of women. Also, due to the small number of women, the results for Groups 0 and III are displayed only in Tables 4, 6, 8 and 9.

In both the iron- and placebo treated subgroup of women, Hb declined to a nadir at gestational weeks 24–27 and subsequently rose again prepartum. In Groups I-III there were no significant differences between the mean Hb concentrations in the iron- and placebo treated subgroup of women at inclusion and at gestational weeks 24–27 (Table 3).

In Group I, Hb values prepartum and one week postpartum were significantly higher in the iron treated than in the placebo treated subgroup of women, while Hb values 8 weeks postpartum were similar in the iron- and placebo subgroup. In Group II, Hb values prepartum were significantly higher in the iron treated than in the placebo treated subgroup of women, while there were no significant differences one week and 8 weeks postpartum.

Comparison of Hb levels in the iron treated subgroups in Group I with Group II, and in the placebo treated subgroups in Group I with Group II, respectively, showed no significant differences either during pregnancy or postpartum (p-values not shown) (Table 3).

The occurrence of anemia is shown in Table 4. In Group 0, at inclusion, the frequency of anemia was not significantly different in the iron and placebo subgroup (p-value not shown). However, the frequency of anemia prepartum as well as 8 weeks postpartum was significantly higher in the placebo Group 0 compared to placebo Group I (p < 0.0106 and p < 0.0001, respectively).

In Group I, the frequency of anemia prepartum was significantly lower in the iron treated than in the placebo treated subgroup of women. In Groups II and III, there were no significant differences between the iron and placebo treated subgroups of women. Eight weeks postpartum, none of the women displayed anemia.

Comparison of the frequency of anemia in the iron treated subgroups in Groups I, II and III, and in the placebo treated subgroups in Groups I, II and III, respectively, showed no significant differences during pregnancy or postpartum (p-values not shown).

Prepartum, in Group I, the frequency of high Hb levels > 130 g/L (8.07 mmol/L) was significantly higher in the iron treated subgroup (14/44 –31.8%), than in the placebo treated subgroup (2/36 –5.6%), (p = 0.0229). However, there were no differences concerning length of gestation or newborns’ birth weight between women having high and “normal” Hb levels, neither at inclusion, weeks 24–27 or prepartum (data not shown).

There was no significant difference between the arithmetic mean log₁₀ S-ferritin concentrations in iron- and placebo treated women at inclusion in Groups I and II (Table 5), where all women had S-ferritin levels≥30–40μg/L indicating the presence of stainable hemosiderin iron reserves in the bone marrow.

Both Group I and Group II displayed a steady, significant decline in S-ferritin levels during pregnancy, and a subsequent significant increase one week postpartum; the changes were observed both in the iron- and placebo subgroups (Table 5). At every time point after inclusion, including one week and 8 weeks postpartum, S-ferritin values in Group I and II were significantly higher in the iron- compared with the placebo subgroups.

Comparison of the S-ferritin levels in the iron treated subgroups in Group I and Group II, and in the placebo treated subgroups in Group I and Group II, respectively, showed no significant differences either during pregnancy or postpartum (p-values not shown).

In both the iron- as well as in the placebo subgroups, S-ferritin levels at 8 weeks postpartum were significantly lower than levels at inclusion (Table 5), indicating that body iron status had not been restored to the levels observed in early pregnancy.

The frequency of ID is shown in Table 6. In Group 0, 6/33 women had ID at inclusion, not being significantly different in the iron- and placebo subgroup (p-value not shown).

In Group 0, the frequency of ID at weeks 24–27, prepartum, one week- and 8 weeks postpartum was significantly higher in the placebo- vs the iron group. Furthermore, at weeks 24–27 and prepartum the occurrence of ID was significantly higher in the placebo subgroup in Group 0 compared to the placebo subgroup in Group I (p < 0.0008 and p < 0.0004, respectively); 8 weeks postpartum the difference in ID was insignificant, but a significantly higher number of women in the placebo subgroup in Group 0 had S-ferritin < 30μg/L compared with the placebo subgroup in Group I (p = 0.0411).

In Group I, the occurrence of ID was significantly lower in the iron- vs. the placebo treated subgroup at weeks 24–27, prepartum, and one week postpartum; however, this difference had disappeared at 8 weeks postpartum (Table 6).

In Group II, the frequency of ID in the iron- and placebo subgroup was not significantly different at gestational weeks 24–27, but significantly lower prepartum and one week postpartum in the iron treated subgroup; this difference disappeared at 8 weeks postpartum (Table 6).

In Group III, the frequency of ID in the iron- and placebo subgroup was not significantly different at gestational weeks 24–27, but significantly lower prepartum in the iron treated subgroup; there was a tendency towards a lower frequency of ID in the iron treated- vs. the placebo treated subgroup at one week and 8 weeks postpartum, but the differences were not significant, probably due to the small number of women (Table 6).

Comparison of the frequency of ID in the iron treated subgroups in Groups I, II and III, and in the placebo treated subgroups in Groups I, II and III, respectively, showed no significant differences either during pregnancy or postpartum (p-values not shown).

Although all 80 women in Group I had S-ferritin levels≥30μg/L at inclusion, 40 (50%) women had levels < 30μg/L at 8 weeks postpartum. Furthermore, at 8 weeks postpartum, the occurrence of ID was significantly lower in the iron treated subgroup than in the placebo treated subgroup, while there was so such difference in Groups II and III.

There were no significant differences between TSAT levels in the iron- and placebo treated subgroups of women at inclusion either in Group I or Group II (Table 7). In both the iron- and placebo subgroups, TSAT declined significantly from inclusion to prepartum, reaching a nadir one week postpartum and subsequently increased to a higher level at 8 weeks postpartum (Table 7).

At weeks 24–27, prepartum and one week postpartum, TSAT values were significantly higher in the iron- compared with the placebo subgroups both in Group I and in Group II. At 8 weeks postpartum, TSAT values in the iron treated subgroups of women in Group I and Group II were not significantly different from levels at inclusion. In contrast, TSAT values 8 weeks postpartum in the placebo treated subgroups of women were significantly lower compared with levels at inclusion both in Group I (p = 0.031) and Group II (p = 0.0239).

Comparison of TSAT levels in the iron treated subgroups in Group I with Group II, and in the placebo treated subgroups in Group I with Group II, respectively, showed no significant differences during pregnancy or postpartum (p-values not shown).

At inclusion, none of the women in Groups 0–III had low TSAT values < 15% indicating an insufficient iron supply to the tissues (Table 8). In both Group 0, Group I and Group II, the frequency of low TSAT levels was significantly higher in the placebo- than the iron treated subgroups at weeks 24–27, prepartum and one week postpartum (Table 8). The same pattern was seen in Group III, where the placebo treated subgroup had an (insignificantly) higher frequency of low TSAT levels at weeks 24–27, and a significant higher frequency prepartum and one week postpartum (Table 8). Eight weeks postpartum values were not significant different in any of the three Groups and 6 subgroups (Table 8) and the frequency of low TSAT values in the 6 subgroups was not significantly different from the values at inclusion (p-values not shown).

Comparison of the frequency of TSAT values < 15% in the iron treated subgroups in Group I with Group II, and in the placebo treated subgroups in Group I with Group II, respectively, showed no significant differences either during pregnancy or postpartum (p-values not shown).

The frequency of IDA is shown in Table 9.

At inclusion, none of the women in Group 0 had IDA. Subsequently, in Group 0, none of the iron treated women displayed IDA, while placebo treated had a high frequency of IDA. Overall, the frequency of IDA in Groups I–III was low during pregnancy and postpartum, not being significantly different in the iron- and placebo treated subgroups, although there was a (nonsignificant) tendency towards slightly higher frequencies of IDA in the placebo treated subgroups compared with the iron treated subgroups in Groups I-II, but not in Group III (Table 9). In Groups I and II, the frequency of IDA during pregnancy in the placebo treated subgroups ranged between 2.8% at weeks 24–27 to 8.3% prepartum. One week postpartum, none of the women displayed IDA in any of the three Groups I–III, and 8 weeks postpartum the frequency of IDA was very low (Table 9).

Among the 113 included women, 36 (32%) (21 iron-, 15 placebo treated) had S-ferritin levels≥60μg/L at inclusion. Similar to Group III, none of these women, either iron-, or placebo treated, developed IDA during pregnancy or postpartum (data not shown).

At inclusion, log₁₀ S-EPO levels were not significantly different in the iron- and placebo treated subgroups, either in Group I or Group II (Table 10). In both the iron- and placebo treated subgroups in Group I and Group II, S-EPO increased significantly during gestation reaching a peak prepartum, and subsequently declined moderately one week postpartum, followed by a further, significant decline at 8 weeks postpartum.

In Group I, prepartum, iron treated women had significantly lower S-EPO levels than placebo treated women, while the levels in the two groups one week- and 8 weeks postpartum not were significantly different (Table 10). In Group II, iron treated women likewise had lower S-EPO levels than placebo treated women, but the difference was not significant. One week- and 8 weeks postpartum, S-EPO levels were not significantly different (Table 10).

Comparison of the S-EPO levels in the iron treated subgroups in Group I with Group II and in the placebo treated subgroups in Group I with Group II, respectively, showed no significant differences either during pregnancy or postpartum (p-values not shown).

In Group I and II, there were no significant differences between the newborns’ log₁₀ cord S-ferritin and body weight either in the iron- or placebo subgroups, respectively (Table 11).

Comparison of cord S-ferritin levels and birth weight in the iron treated subgroup in Group I with Group II and in the placebo treated subgroup in Group I with Group II, respectively, showed no significant differences (p-values not shown).

4.1 Hemoglobin

Group 0: The iron depleted women had clear benefits of iron supplements, which reduced the frequency of anemia (Table 4), ID (Tables 6 and 8) and IDA (Table 9), compared to placebo treated women. It is evident from these data that iron supplementation should be a must in women with S-ferritin levels < 30μg/L.

Groups I-II: The longitudinal changes in Hb levels during pregnancy and postpartum (Table 3) were in accordance with those reported in the entire series [17]. There were no differences between the iron/placebo subgroups at inclusion and weeks 24–27, but prepartum the iron subgroups had higher Hb than the placebo subgroups. In Group I, one week postpartum, the iron subgroup had higher Hb than the placebo subgroup, while there was no such difference in Group II. This suggests that Group II “tolerates” placebo treatment better than Group I. Eight weeks postpartum there were no differences between the Groups.

4.2 Anemia

Groups I–III: At weeks 24–27, all three Groups displayed a similar frequency of anemia in the iron- and placebo subgroups (Table 4). However, in Group I, prepartum, the iron treated subgroup had a significantly lower frequency of anemia than the placebo subgroup. This difference was not seen in Groups II and IIII. This suggests that Groups II and III “tolerate” placebo treatment better than Group I.

4.3 Serum ferritin

Groups I-II: The longitudinal changes in S-ferritin levels during pregnancy and postpartum (Table 5) were in accordance with those reported in the entire series [17]. There were no differences in S-ferritin levels between the iron/placebo subgroups in Groups I and II at inclusion. At weeks 24–27, levels in Group I were higher in the iron- than in the placebo subgroup in contrast to Group II. However, prepartum, one week- and 8 weeks postpartum, the iron subgroups had consistently higher S-ferritin levels than the placebo subgroups. This is a logical consequence of the iron treatment that it will increase iron status, e.g., S-ferritin, even in iron replete women, when the gestational needs for iron have been fulfilled. However, in the placebo treated groups, body iron reserves will gradually become more and more depleted as pregnancy progresses. Adequate iron reserves prepartum are important to secure a sufficient iron supply to the fetus and to prevent postpartum anemia in case of major blood losses at delivery [31].

In both the iron- as well as in the placebo subgroups, S-ferritin levels at 8 weeks postpartum were significantly lower than levels at inclusion (Table 5), indicating that body iron status had not been restored to the levels in early pregnancy. Multiple childbirths cause a significant decrease in iron reserves, which actually persists until the menopause [32].

4.4 Iron deficiency

Groups I–III: At weeks 24–27, prepartum and one week postpartum, the iron treated subgroup in Group I had a lower frequency of ID than the placebo treated subgroup, while the differences between iron- and placebo treated subgroups in Group II and especially in Group III were less pronounced (Table 6). Prepartum, the three iron treated subgroups in Groups I–III displayed a lower frequency of ID than the placebo treated subgroups (Table 6). One week postpartum, Group III displayed the lowest frequency of ID, being the same in iron- and placebo treated women. This suggests that Group III tolerates placebo treatment better than Group I and II due to higher body iron reserves in early pregnancy.

4.5 Transferrin saturation

Groups I-II: The longitudinal changes in TSAT levels during pregnancy and postpartum (Table 7) were in accordance with those reported in the entire series [17]. At weeks 24–27, prepartum and one week postpartum, TSAT levels were consistently higher in the iron treated- compared to the placebo treated subgroups indicating that iron treatment significantly influenced TSAT irrespectively of S-ferritin levels. It is not clarified whether a higher TSAT level (within the normal range) is beneficial for the tissue iron supply to the mother and fetus, as the transport capacity of iron in the blood also depends on the S-transferrin concentration, which increases steadily during gestation [16] to fulfill the fetal needs for iron. The blood samples were taken in the non-fasting state at different points in time during the day, but this seems of minor importance, since there are no consistent diurnal or day to day variations [33].

The frequency of low TSAT levels < 15% in Groups I and II was consistently higher in the placebo treated- than in the iron treated subgroups at weeks 24–27, prepartum and one week postpartum, (Table 8), while in Group III, this difference was present only prepartum and one week postpartum.

4.6 Iron deficiency anemia

IDA is presently probably the best “hard core” and weighty parameter to assess the presence and ultimate effect of ID. Groups I-III all had a relatively low frequency of IDA. The highest frequency of IDA was observed in the placebo treated subgroup in Group I, the second highest in the placebo treated subgroup in Group II, while none of the women in Group III, and none of the women with S-ferritin≥60μg/L, had IDA. This finding strongly suggests that the majority (approximately 95%) of women in Group II, and all women in Group III “tolerate” placebo treatment and can complete a normal pregnancy without taking iron supplements and without developing IDA.

This study also shows that a ferrous iron dose of 66 mg as fumarate appears adequate to prevent IDA, even in women who are iron depleted in early pregnancy (Table 9).

4.7 Serum erythropoietin

The longitudinal changes in S-EPO levels during pregnancy and postpartum (Table 10) were in accordance with those reported in the entire series [17]. During pregnancy, S-EPO increases significantly due to the accelerated erythropoiesis and there is a negative correlation between Hb- and S-EPO levels [17]. In Group I, prepartum, the placebo treated subgroup had higher S-EPO than the iron treated subgroup, indicating a “normal” physiological response to lower Hb levels, which may be due to a relative ID/IDA. In Group II, S-EPO was similar in the iron- and placebo treated subgroups, again suggesting that these women may be able complete pregnancy without iron supplements.

4.8 Newborns

In accordance with previous findings [17], there were no differences in birth weight in the four Groups 0–III (Table 11). Cord S-ferritin levels were consistently slightly higher in the iron treated-, than in the placebo treated subgroups, but the differences were not significant in Groups I–III, probably due to the low number of newborns in each group. There is no obvious explanation for the high geometric mean cord S-ferritin ferritin value in the iron treated subgroup in Group 0, it may be due to a statistical coincidence.

We have previously shown that placental weight in the entire series of 119 women is correlated to newborns birth weight [17]. HPL is produced in the placenta and is an indirect measure of the size and function of placenta [34] and prepartum S-HPL levels display a significant correlation with placental weight in the total series of 119 women (r = 0.47, p < 0.0001); this relationship explains why prepartum S-HPL is “indirectly” correlated to birth weight (r = 0.26, p = 0.0037). Another factor determining birth weight is the mothers body weight/body mass index (BMI) which is correlated to birth weight [17], probably in part because mothers BMI is correlated to placental weight (r = 0.26, p = 0.0052); these results have not previously been published.

Some obstetricians remain concerned that iron supplementation can induce high Hb levels that may lead to gestational complications with still birth, premature birth and small for gestational age birth weight [35].

In the total series of 119 women [17], the occurrence of high Hb values > 130 g/L was higher in the iron treated group, 23/62 (37.1%) than in the placebo treated group, 2/57 (3.5%) (p < 0.0001). However, we found no associations between high Hb levels, length of gestation or birth weight in any of our two previous studies [16, 30]. Furthermore, such an association has not been reported in any of the other published placebo controlled studies on iron supplementation in pregnancy [6].

In comparison with the high demands for iron during pregnancy, dietary iron intake in pregnant women in Denmark and in other European countries is inadequate and markedly below the national dietary recommendations [36, 37]. Therefore, iron supplementation in pregnancy should be considered and is recommended in several countries. However, general routine iron supplementation will also target women with adequate iron reserves who can complete a pregnancy without iron supplements and without developing ID/IDA [6, 38]. In these women, iron supplements are unnecessary and may have unwanted side effects. Furthermore, in Denmark approximately 400 women/year with (undiagnosed) mutations related to genetic hemochromatosis become pregnant and are advised unnecessary, and for these women even harmful routine iron supplements [10].

How do we identify women who do not need iron supplementation? The gross total need for extra iron during a normal pregnancy is approximately 1200–1400 mg [1–3]. After delivery, there is a marked reduction in the mothers red blood cell volume and about 500 mg iron will be returned to the iron reserves, yielding a net iron loss of 500–600 mg [1–3]. Based on these figures, there appears to be consensus that women with iron reserves≥500 mg are not in need of iron supplements during pregnancy [38–40] but should be capable of absorbing the remaining 600–700 mg iron from a normal diet during a median gestation period of 282 days [16, 17]. In healthy persons, S-ferritin is a good biomarker of body iron reserves, i.e., 1μg/L corresponds to iron reserves of 7–7.5 mg [41, 42]. This implicates, that women with S-ferritin levels above 60–70μg/L in early pregnancy, corresponding to iron reserves of 400–500 mg, should be able to complete a pregnancy without developing ID/IDA, but this has until now not been confirmed in any randomized study.

The present study demonstrates that the need for iron supplements gradually decline with increasing body iron reserves preconceptionally/in early pregnancy as estimated by the S-ferritin levels. Non-iron-supplemented pregnant women with S-ferritin < 30μg/L have a far higher incidence of ID/IDA than non-iron-supplemented women with S-ferritin levels,≥30,≥40,≥50, and≥60μg/L. It appears S-ferritin levels≥40μg/L are associated with a very low risk of ID/IDA and at levels≥50- and≥60μg/L, the chance of getting ID/IDA is close to nil.

According to the present new findings and the results from an earlier study [30], the suggested recommendation for iron supplementation in pregnancy with the “classical” ferrous iron salts, e.g., ferrous sulphate/ferrous fumarate based on S-ferritin levels at the first antenatal visit could be revised to: I. Ferritin≥50μg/L - no iron supplement; II. Ferritin 30–49μg/L –30 mg elemental iron as ferrous sulphate/ferrous fumarate daily; III. Ferritin < 30μg/L: 50–60 mg elemental iron as ferrous sulphate/ ferrous fumarate daily –this dosage will also be adequate for the 2–4% of women [5] who present with slight ID/IDA at the first visit.

Ferrous bisglycinate is an iron chelate, with a two-fold higher absorption rate than the standard reference ferrous sulphate. Therefore, a dose of 25 mg elemental iron/day as ferrous bisglycinate appears adequate to prevent IDA in more than 95% of pregnant women [43]. Using ferrous bisglycinate, the recommendation can be changed to a lower iron dose: I. Ferritin≥50μg/L - no iron supplement; II. Ferritin 30–49μg/L –25 mg elemental iron as ferrous bisglycinate daily; III. Ferritin < 30μg/L - 50 mg elemental iron as ferrous bisglycinate daily. All iron preparations should be taken at bedtime or between meals to optimize iron absorption.

4.9 Limitations of the study

The number of included women in the various Groups and subgroups are relatively small, which may weaken the statistical conclusions. The participants were not instructed to take the iron tablets between meals or at bedtime, which enhances the iron bioavailability [30]. The individual compliance to the intake of iron tablets was not assessed and non-compliers not excluded. Non-compliance may explain why some women in the iron treated subgroups displayed anemia (Table 4) and some had low S-ferritin levels (Table 6). However, in general, non-compliance appeared not to be a significant problem, as none of the women in the iron treated subgroups in Groups 0–III displayed IDA (Table 9).