A Randomized Controlled Trial of Daily and Weekly Iron Supplementations for Improving Iron Status in the Infants

Background: Iron deciency anemia screening and iron supplementation in infants aged 6-12 months are recommended in the Guideline in Child Health Supervision. This study aims to evaluate the effectiveness of weekly iron supplementation compared with daily supplementation in improving the iron status in infants. Methods: A single-blind randomized controlled trial was conducted in infants aged 6 months visiting the Well Child Clinic between May 2019 and November 2020 at Burapha University Hospital, Chonburi, Thailand. The intervention consisted of either daily or weekly iron supplementation combined with iron-rich complementary food promotion for six months. The outcomes were the differences of serum ferritin and hematological variables before and after being iron supplemented. Results: Sixty-nine six-month-old healthy infants were randomized to receive either 10 mg Fe/day (daily group) or 30 mg Fe/week (weekly group). Forty-ve infants (daily group; n = 24 and weekly group; n = 21) completed the intervention. After the six-month period of iron supplementation, the mean differences of serum ferritin in the daily and the weekly group were 8.78±37.21 and -13.05 ±17.53 ng/mL, respectively (95%CI: 4.54, 39.12; P=0.015). The mean differences of hemoglobin in the daily and the weekly group were 0.58±0.82 and 0.08±0.59 g/dL, respectively (95%CI: 0.06, 0.93; P=0.026). Daily supplementation could prevent iron deciency more than weekly supplementation signicantly (P=0.029), particularly in the exclusive breastfeeding subgroup (P=0.032). Conclusions: Daily iron supplementation is more effective than weekly iron supplementation in improving iron status and hemoglobin level in infants, especially in the exclusively breast-feds.

(CONSORT) statements. The CONSORT ow diagram of participant's enrollment are shown in Figure 1. The study was registered at the Thai Clinical Trials Registry (code: TCTR20191107001).
Healthy 6-month-old infants who attended the Well Child Clinic at Burapha University Hospital were voluntarily enrolled to the study. Only infants who met the following criteria were considered eligible: (1) singleton birth, (2) gestational age more than 37 weeks and birth weight more than 2,500 g, (3) no history of serious perinatal complication including cardiopulmonary resuscitation or blood/exchange transfusion, (4) have not previously taken any iron supplementation or therapy (5) no chronic disease (such as malnutrition, hematologic, cardiologic, pulmonologic, neurologic and liver diseases). Eligible infants were screened for anemia by Hct screening as a routine well-child practice. Infants who were anemic (Hct <33%) were excluded from the study. At the enrollment, information about the aim of study was provided to the parents. A written informed consent was obtained from parents who agreed to participate. Baseline characteristics including gestational age, birth weight, gender, weight, length, dietary intake, and perinatal/pregnancy complications were obtained.
The determined dietary intake included breast milk, infant formula, and common Thai iron-rich complementary food (meat, liver, and yolk). Caregivers were advised to provide iron-rich complementary diets to their children.
Participants were randomized using a computer-generated randomization code. Allocation was concealed in opaque envelopes and numbered to mask the randomization code. The daily or weekly assignment was sealed accordingly. Envelopes were opened by research assistants only after the infant was sequentially enrolled by the investigators. All instructions were given by research assistants to blind the investigators from the allocation. Participants were distributed into two groups, one receiving daily (elemental iron 10 mg/day) and the other receiving weekly (elemental iron 30 mg/week) iron supplementation.
The given iron supplement was a commercially available formula of iron (III) hydroxide polymaltose complex (Eurofer ® -Iron, Osoth Inter Laboratories, Chonburi, Thailand). The caregivers were instructed to give the iron supplements to their infants 1 hour before breakfast for 6 months (1 ml every day in the daily group and 3 ml every Sunday in the weekly group) and record the dose taken/missed, side effects and any infection occurred in the calendar booklet provided. If the daily dose was missed, the iron supplement should be taken as soon as they remember within the same day. If the weekly dose was missed, the iron supplement should be administered within the same week. Apart from that, it should be recorded as a missed dose.
The average daily dietary intake was collected every three months by phone or during well-child visits at 9 months old and 12 months old. Portion sizes were estimated using tablespoons and sample pictures. Caregivers were asked to recall the daily average food monthly. Food intake data were converted into iron intakes using Thai Food Composition Database, Online version 2 (Institute of Nutrition, Mahidol University) [26]. Commercial infant formula available in Thailand contains iron 0.55-1.2 mg/dL. Conversion of 1 dL of infant formula to 0.8 mg of elemental iron was applied.
Venous blood sampling for CBC and SF was collected from infants aged 6 and 12 months. If there were any recent infection of the participants, the blood sampling would be postponed for 1 week to prevent the falsely high SF. CBC was performed by automated cell counter (CAL 8000, Mindray, China). SF was measured by electrochemiluminescence (ECLIA) method (Cobas ® pro 801, Roche, Germany).
A follow-up evaluation was scheduled at 12 months of age. The calendar booklets were collected back to the investigators. Side effects, any infection occurred, and adherence were checked. Non-adherent (lower than 80%) participants were withdrawn from the analysis. The primary outcome was the change in SF level from the beginning. The secondary outcome was the changes in CBC and side effects occurred.

Sample size and statistical analysis
Assuming that weekly iron supplementation would be as effective as daily iron supplementation. The formula for non-inferiority design was used. The sample size required to be able to distinguish a difference between groups in the ferritin level was 25 in each group (de ned the signi cant level at 0.05 and power of test 80%) [24]. Assuming a dropout rate of 40%, therefore a total of 70 participants was used in this study. Results were analyzed using SPSS Version 20.0.
Analysis of continuous data with a normal distribution was conducted by independent sample t-test. Categorical data was analyzed by Chi-square test or Fisher's exact test where applicable.

Results
A total of 149 six-month-old infants were eligible and screened for anemia. Eighty infants were excluded (42 infants were anemic and 38 declined to participate). Sixty-nine infants were randomized into daily group (n = 35) and weekly group (n = 34). The enrollment ended ahead of scheduled because of the di culty during the COVID-19 pandemic. Fourteen participants lost to follow-up and/or discontinued the intervention. Ten participants were withdrawn from the analysis due to non-adherent to medication. There were 45 participants remaining to the analysis (daily group; n = 24 and weekly group; n = 21).
Baseline characteristics including gender, birth weight, gestational age, percentage of exclusive breast-fed infants, average iron intake, weight, and length (both baseline and nal point) were not different between groups (Table 1). The baseline (aged 6 months) and nal (aged 12 months) iron status and hematological variables (Hb, Hct, and SF) were not signi cantly different between groups except for the initial mean corpuscular volume (MCV). The initial mean MCV of the weekly group was signi cantly larger than the daily group (75.80 ± 4.83 vs 71.62 ± 7.85 fL, P = 0.04). The mean difference of SF after six-month period of iron supplementation was 8.78 ± 37.21 ng/mL in the daily group and -13.05 ± 17.53 ng/mL in the weekly group (95%CI: 4.54, 39.12; P = 0.015). The mean difference of Hb was 0.58 ± 0.82 g/dL in the daily group and 0.08 ± 0.59 g/dL in the weekly group (95%CI: 0.06, 0.93; P = 0.026). The mean difference of Hct and MCV were not statistically signi cant (Table 2).

Variables
Daily (n = 24) mean ± SD Weekly (n = 21) mean ± SD P value # 95% CI # Numbers of infants with ID were observed at both baseline and nal point. Initially (at 6 months old), 29.2% of the daily group and 9.5% of the weekly group were found to have ID. When the infants were followed at age 12 months old, the number of iron-de cient infants decreased to 12.5% in the daily group but increased to 33.3% in the weekly group. The ID status change after iron supplementation were changed to categorical data by using the de nition as follows: -1; the non-ID have changed to ID/IDA, +1; the ID have changed to non-ID and 0; the ID and non-ID remained unchanged. The results have shown that the weekly-group infants subsequently developed ID more than the daily-group. The daily-group infants also subsequently improved iron su ciency more than the weekly-group signi cantly (P = 0.029). In the subgroup analysis (EBF vs non-EBF), ID status modi cation was statistically signi cant between daily and weekly groups only in the exclusively breast-fed subgroup (P = 0.032) ( Table 3). EBF exclusively breast-fed, ID iron de ciency # Frequency of ID status change was reported as follows: -n the non-ID have changed to ID, +n the ID have changed to non-ID * Chi-square test: P value of ID status changes between daily and weekly group The frequency of side effects between groups were compared in Table 4. The data was collected and analyzed from all the contactable participants (daily group; n = 30 and weekly group; n = 29). Seven (23.3%) infants in the daily group compared with nine (31.0%) infants in the weekly group were reported with side effects related to the iron supplements. The most serious side effect was constipation resulting in two participants withdrawing from the study (one from each group). Other side effects were vomiting/spitting (reportedly due to unfavorable taste) which affected the adherence of the participants. The total side effect rate was not different between groups (P = 0.506).

Discussion
In the hypothesis of the study, daily and weekly iron supplementations are equally effective in improving the infants' iron status. However, this study has shown that only daily iron supplementation is effective in improving the iron status while weekly iron supplementation is not enough. The weekly group initially tended to have a better iron status than the daily group, according to the higher mean SF level and lower percentage of ID. After 6 months of weekly iron supplementation, there were decremental change of the SF level and frequency of iron-su cient infants. In contrast, the SF level was increased and the frequency of iron-de cient infant was decreased in the daily group. This result of intermittent iron supplementation has not been observed in previous studies [21,[23][24][25]. The hematological variables of both groups were compared. The Hb level of both groups have shown improvement, whilst the daily group improved more than the weekly group signi cantly. The baseline MCV of the weekly group were signi cantly larger than the daily group. At the end point, the MCV (in terms of size and difference) between groups were not signi cantly different. This might be due to the red blood cells of the weekly group having reached their maximal size, weekly supplementation was less effective, or the infants had Thalassemia trait. Therefore, daily iron supplementation could somehow improve the MCV parameters better than in the weekly group, although not statistically signi cant. During subgroup analysis, we found that most of the iron-de cient infants were exclusively breast-fed which tended to continue with breast milk as the major formula through the age of 12 months. This may conclude that weekly iron supplementation of 30 mg is not enough to maintain the iron status in the infants who were breast-fed beyond 6 months.
The comparison between this study and previous studies of intermittent and daily iron supplementation were simpli ed and shown in Table 5. (There were variations of situation and limitation among studies. Some of the studies had more than 2 arms but this table was simpli ed to compare with this study.) The previous studies' population ranged from 4 to 60 months, the duration of iron supplementation ranged from 2 to 6 months and the iron formulation used were ferrous sulfate. Compared to previous studies, intermittent iron supplementation may not differ from daily supplementation in terms of preventing ID/IDA/anemia [22][23][24], increasing Hb level [21][22][23][24][25], and increasing ferritin level [21,[24][25]. Apart from these studies, Ermis et al [23] demonstrated a higher SF increment of the intermittent group than the daily group (the total iron dose of both groups in this study were equal). However, the iron cumulative dose per week of previous reports were mostly higher than our study. Some of them were similar between groups which could be the reason why intermittent and daily supplementation outcome were not signi cantly different. This study was carried out with a longer duration of iron supplementation which may be another additional factor. However, the general outcomes of previous studies were similar to this study of which the more iron supplemented the better improvement of iron/hematological status. Further studies comparing a higher dose of intermittent supplementation with the standard recommendation dose could be evaluated.
The limitations of this study were the dietary intake interview which was recalled every 3 months and the average iron intake during 6-12 months from complementary food may be underestimated. Only the common Thai iron-rich complementary food (not the commercial) intake was calculated as the iron intake from food. The results have shown that the main source of iron intake in 6-12 months infants was from infant formula rather than complementary food.
During enrollment period, we found that 28.2% of eligible infants were anemic. The prevalence of anemia in this study tended to be similar to previous reports in Thailand for the past 10-20 years [7][8]27]. Future studies should be aiming to ameliorate IDA in infancy which is the critical period of neurodevelopment.