Effect of iodine supplementation in pregnancy on neurocognitive development on offspring in iodine deficiency areas: a systematic review

ABSTRACT Objective: To investigate the effect of iodine supplementation during gestation on the neurocognitive development of children in areas where iodine deficiency is common. Materials and methods: Based on the PRISMA methodology, we conducted the search for articles in the PubMed, LILACS and Scopus databases, between March and April 2020, without limitation of dates. We used descriptors in English, Portuguese, and Spanish, without filters. Four clinical trials and four cohort articles were included in the review. Results: The maximum supplementation was 300 μg of potassium iodide per day. The Bayley scale and Children’s Communication Checklist-Short were used to assess neurodevelopment in children. There was no significant improvement in the children’s mental development index and behavioural development index in the supplemented group; however, the psychomotor development index (PDI) showed improvement in the poorer gross motor skills. We found differences in the response time to sound in the supplemented group living in mild deficiency areas. Conclusion: Daily supplementation with iodine can improve poor psychomotor development of children living in mild to moderate iodine deficiency areas. Thus, it is necessary to perform further studies to assess the effect of supplementation on neurodevelopment before, during and after gestation in mild to moderate iodine deficiency areas.


INTRODUCTION
I odine deficiency affects almost 2 billion people worldwide (1). In 2017, 18 countries were identified in which women of reproductive age were iodine-deficient, whereas for pregnant women, this was found in 39 countries (2). At this stage, deficiency induces the occurrence of irreversible brain damage in children (1). In fact, inadequate iodine intake in the foetal period may cause dwarfism, cretinism, mental retardation, deafness, psychomotor defects, or congenital anomalies, and may lead to miscarriage or stillbirth (3). Throughout growth, it negatively affects physical and neurocognitive development, especially hippocampal development and memory functions, and in adult life, causes goiter and hypothyroidism (4).
The recommended daily intake of iodine is 90 μg in the age group 0-59 months, 120 μg in 6-12-yearolds, 150 μg in adolescents and adults, and 250 μg during gestation and lactation (5). To ensure sufficient iodine intake, women who are planning pregnancy, pregnant or lactating should be recommended by the American Thyroid Association and European Thyroid Association to ingest daily oral supplements containing 150 μg of iodine (6,7). The World Health Organization (WHO) affirm that this supplementation should be Iodine effect on neurocognitive development Arch Endocrinol Metab. 2021;65/3 undertaken when iodized salt does not reach over 90% of households (5).
Recent findings in mild iodine deficiency areas in Israel and Iceland report the improvement of iodine intake in pregnant women supplemented with iodine compared with those not taking iodine supplements (8,9). Other studies in mild iodine deficiency areas in Brazil showed that supplementation corrects maternal thyroid indices and avoids impairment of the neuropsychological development in the offspring (10).
However, the effectiveness of iodine supplementation in pregnant women at improving children's cognitive development is poorly explored and uncertain (11)(12)(13). Therefore, this review aimed to investigate the effect of iodine supplementation during gestation on children's neurocognitive development in iodine deficiency areas.

MATERIALS AND METHODS
This systematic review sought to answer the following question: "What is the effect of iodine supplementation during gestation on children's cognitive development?". The review protocol was registered in PROSPERO (International Prospective Register of Ongoing Systematic Reviews) with the identification number CRD42019116962.
We used the PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) (14) methodology to select articles. To identify the articles, we conducted the search in the PubMed, LILACS (Health Sciences in Latin America and the Caribbean) and Scopus databases, from March 1st to April 1st 2020, without limitations of dates. We used the descriptors: "iodine AND supplementation AND child AND development AND cognitive", provided by DeCS (Health Science Descriptors) (15), in English, Portuguese, and Spanish, without filters (Supplement Appendix 1).
After the searches and elimination of duplicates by database and between databases, we registered all articles in a spreadsheet in Microsoft Excel ® . Then, we recorded data from the articles, detailing the year, authorship, place of origin, type of study, target population, sample size, dose and time of supplementation, tests to assess neurocognitive development, and main results observed.
The inclusion criteria were that the studies should be randomized or non-randomized controlled trials or cohorts that evaluated the effect of iodine supplementation during gestation on the neurocognitive development of children living in moderate to severe, mild to moderate, severe, moderate, or mild iodine deficiency regions. We included all children in this study, without any age limit, provided that the study presented some scale of measurement of their neurodevelopment. Studies on the effect of intake of fortified foods, as well as literature reviews, crosssectional studies, animal model studies and studies that assessed supplementation in pregnant women with thyroid disease were discarded (Supplemental Table 1).
The PICO was defined, namely: Populationpregnant women; Intervention -iodine supplements (iodine supplement use, iodine supplement coverage, iodine content in supplements); Comparator -other children of mothers without iodine supplement use; and Outcomes -development index (mental, psychomotor and verbal), sound response time, IQ (Intelligence Quotient) score (verbal, performance, and reasoning), skills score (language, reading, and writing), mapping test, reading, mathematics and special education.
The scale used to assess neurodevelopment in children selected from the included articles was the Bayley and Children's Communication Checklist-Short (CCC-S).
The Bayley scale has three indices: mental, psychomotor, and behavioural development. The mental development index assesses the visual perceptual acuity, discrimination between objects, problem solving skills, language, and memory (16)(17)(18). The psychomotor development index (PDI) is assessed through postural control and appendicular motricity (16)(17)(18). The behavioural development index (BDI) assesses the follow-up of instructions, attitudes, and energy during the test, among other social behaviors (16)(17)(18). The Bayley score includes cognition and psychomotor skills with mental index (MDI), with a mean score of 100 (SD 15, range 55-155). The mean language (BDI) score was 100 (SD 15; range 45-155). Severe to moderate neurodevelopmental issues were defined as a mean MDI < 85 or BDI < 85, or both < 70; mild to moderate issues were defined as 85-100, and adequate function was defined as ≥ 100 (19).
However, the CCC-S is effective as a standardized assessment at identifying children with clinicallysignificant language impairment (20), containing 13 items that best discriminate typically-developing children from peers with language impairment in the validation study (21), with a high degree of internal consistency. Each item provides an example of language behaviour in everyday contexts and covers speech, vocabulary, grammar, and discourse. The items are scored as 0 -absent response, or 1 -present response, with final analysis using statistical methods.
The quality of the studies was assessed according to the checklist of Joanna Briggs Institute (JBI) Critical Appraisal Tools of the Faculty of Health and Medical Sciences at the University of Adelaide, South Australia (22,23). The checklist consider each question should be answered through four options: Yes (Y), No (N), Unclear (U) and Not Applicable (NA). The bias risk percentage calculation is done by the amount of "Y" that has been selected in the checklist. When "NA" was selected, this question was not considered in the calculation, according to the guidelines of JBI. This tool classifies the studies in: up to 49% is considered a high risk of bias. From 50% to 70% is moderate and above 70% is low risk of bias.

RESULTS
The search resulted in 136 articles, of which eight were included in the review (Figure 1 and Supplement Appendix 1). The studies dated from the year 2009 (24) to 2019 (28), four of which were performed in Spain (24,26,27,29), two in Norway (25,28), one in India or Thailand (30), and the other in Australia or New Zealand (31). Two studies were performed in mild to moderate iodine deficiency areas (24,31), five in mild iodine deficiency areas (25)(26)(27)(28)(29), and one in a severe iodine deficiency area (29).
Regarding the design, four studies were randomized clinical trials (RTC) (24,26,30,31) and four were cohorts (25,27,28,29 The maximum supplementation was 300 μg of potassium iodide (KI) per day (24,26) and one study did not specify supplementation dosages (28). Among the reviewed studies, five started supplementation in the first trimester (24,26,27,29), one in the 14th week (30), another between the 16th and 26th week (25), and one used four different start time categories (28). Only one study continued the supplementation in the lactation period (24); the others finished at the child's birth (Table 1). Most studies used KI (24,26,30,31); however, some studies did not specify the source of the supplementation (Table 1 and Supplemental Table 2).
The results found an association between supplementation with 150 µg of KI/day and poorer gross motor skills of the PDI standardized beta 0.18 (95% CI: -0.33, -0.03, p = 0.02) in one study (25), but in another four studies (24,26,27,29) supplementation with ≥ 150 µg of KI/day was associated with a 5.2-point decrease in PDI (95% confidence interval: -8.1, -2.2), decrease in PDI with < 85, odds ratio: 1.7 (95% confidence interval: 1.1, 2.6). The supplementation with 200 or 300 µg of KI/day was related to lower PDI than the iodized salt group. However, another outcome of our study showed that intake of 300 µg of KI/day in breastfeeding was associated with a mean 6.1 ± 0.9 -point increase in PDI compared to the control. Three other studies (28,30,31) did not find an association between iodine supplementation and neurodevelopment in children (Table 1 and  Supplemental Table 3).
Regarding the quality analysis of the studies, the authors observed some limitations in reporting the methods of all trials, leaving some uncertainty in the assessment of several bias criteria, because in two point assessed in RCT studies were high risk of bias (<50%) but, as the studies in many points were moderate or above low risk bias and evidenced a clear delineation of the intervention, as well as were published in good journals we assumed to use all studies include in our review (Figures 2 and 3).

DISCUSSION
The findings showed an association between iodine supplementation and poor psychomotor development of children aged between 3 and 18 months, living in mild to moderate iodine deficiency areas.    Although not significant, other studies have shown positive results, in which children of supplemented mothers presented higher values of the psychomotor development index (25,27,29,31), behavioural (25,27,29), mental (24,25,30) and communication skills (28), when compared to those who were not supplemented. On the other hand, supplementation of mothers with between 150 and 200 μg of KI per day had no positive effect on the neurocognitive development of their children, as much as in those living in mild as well as moderate iodine deficiency areas, and in some studies the scores were low PDI point and your chance assessed were worse in the treated group (Supplemental Table 2).
None of the RCTs show an association between supplementation and child neurodevelopment, except for a negative association between iodine supplementation and expressive language (BSID) at 1 year in a single trial. The non-RCT studies show mixed results: with a positive association in one case and a negative association in the second. Children in the treatment group were associated with a lower PDI score than in the control group, with a better speaker skills score, poorer skills in the languages domain, lower mapping test results in reading in school, and suboptimal or low scores in mathematics.
Recent evidence has demonstrated these outcomes presented above, showing that 18-month-old children of mothers supplemented with 220-390 μg of KI per day had lower cognitive, language and motor scores (32).
In addition, Gowachirapan and cols. (2017) identified all development scale in primary results with placebo group had higher scores than the treatment group (30) in children aged 12 to 24 months in mild iodine deficiency areas.
Our findings mostly covered children under 24 months old, and the poor psychomotor effect on the children of supplemented mothers was demonstrated in this age group. In our results, the mothers supplemented from the 14th gestational week had a negative association between supplementation and child neurodevelopment, at ages from 14th months.
However, the start of supplementation at the 14th gestational week how showed our findings seem to be Iodine effect on neurocognitive development Arch Endocrinol Metab. 2021;65/3 late to start supplementation, since the development of the nervous system occurs mainly between the 5-6th gestational weeks and birth, and between birth and 2 years, for infants and children (33,34).
Most of the mothers were supplemented from the 1st trimester of gestation, and in one study, the treatment continued during lactation. Through the results of this study, it was possible to verify that the psychomotor and behavioral development differed significantly among children of mothers supplemented with 300 μg of KI per day, living in areas with mild to moderate iodine deficiency (24). Recommendations from the American Thyroid Association and European Thyroid Association indicate that supplementation started in the pregestational period is more effective (6,7).
Supplementation with ≥ 150 μg of KI per day in pregnancy can be improve poor psychomotor development in children. This outcome is observable in lactation if supplementation dose is doubled (300 μg of KI per day).
In another study, children of mothers living in mild iodine deficiency areas and supplemented with 200 μg of KI per day during gestation showed a better response time to sound at 60 to 72 months than their is not supplemented group, but there was no difference between the groups (Supplemental Table 2) (30). This was the only study that used other methods to assess child development beyond the Bayley scale (30), and was the only one that assessed children over two years old, showing that this may be a more interesting time to assess the children's development. However, use of the Children's Communication Checklist-Short showed to be better for the assessment of skills and knowledge, including the domains of writing, speaking, reading, mathematical calculations and all languages in older children (>3 years old). This method uses the mental and behavioural skills applicable to the Bayley scale (mental and behaviour development index), and we did not find an association between iodine supplementation and this score in our results (28). These findings were reported for other authors that used the CCC-S to assess older children and used the Bayley score to assess the infant group; they did not find clear differences between these groups (35).
Although the findings showed poor psychomotor development in the children of the supplemented mothers, it seems that this effect is more pronounced in younger children compared to older children using the Bayley scale. However, we observed a high score of the sound response time in children from 60 months, open in this age the children are keen senses.
The use of developmental scales requires caution, since they depend on the evaluator's observation (36). Despite this, the use of these scales seems to have good results for those living in areas with mild to moderate iodine deficiency. However other factors that may interfere in test results are family income, mother's education, inadequate urinary iodine concentration (UIC) of the mother, and the presence of siblings, since they directly influence the family stimulus that the child receives (7,11,30,36).
The lack of similarity between initial time, duration, dosage of supplementation, and the time of application of neurocognitive development tests were limiting factors. In addition, three of the seven studies did not assess behavioural development.
The authors observed that supplementation during lactation brings interesting results, which may be the starting point for future research. In areas with mild to moderate iodine deficiency, changes are more likely to develop in children's psychomotor, behavioural, and mental capabilities. The authors questioned whether the duration of supplementation may have a greater influence than the dose administered, since we did not find any studies with a longer time of supplementation with a lower dose of iodine content, nor did we obtain further assessments of lactation.
The best neurodevelopmental can be good in children with mother living in iodine adequate areas. However, in these results, the mothers in the control group had below adequate UIC, showing iodine deficiency for maternal group in region, which can affect the outcomes in their offspring. Additionally, according to Mao and cols. (2018), the supplementation of pregnant women living in areas of mild iodine deficiency did not have any effect on their children's neurocognitive development (35).
Improving some factors, such as the start and end times of supplementation, iodine sufficiency of the mothers and the iodine deficiency in the areas where the mothers live, as well as the age of the children and the type of scale used in the tests, can contribute to better results. Therefore, iodine supplementation, if well implemented, can reduce risks to the population and, consequently, reduce public health expenditure.
Final remarks: In general, in this study we did not find an association between iodine supplementation in pregnant women and the neurodevelopment of their children in mild to Iodine effect on neurocognitive development Arch Endocrinol Metab. 2021;65/3 moderate iodine deficiency areas. Despite this, supplementation in pregnancy and lactation can be improve poor psychomotor development in children. However, in older children, it seems to have a greater effect on the sound response time. Supplementation in pregnant women also improved urinary iodine concentration of the mother and her children, as well as leading to a high PDI score in young children. Thus, it is necessary to perform further studies using the Bayley scale or another scale alongside the Children's Communication Checklist-Short (CCC-S) or CCC-2 to assess the effect of iodine supplementation in pregnant women in iodine deficiency areas on the neurodevelopment of children before, during and after pregnancy.   The results are consistent in demonstrating no beneficial effects of iodine supplement use in pregnancy. Not have any associations between maternal use of iodine-containing supplements and child outcomes. But in sample, 6.9% of the 8year old were granted special education at school. According to the mother, 28% had suboptimal or low score on the mandatory mapping test in school in reading, and 18% had suboptimal or low score in mathematics.