Iodine supplementation for women during the preconception, pregnancy and postpartum period

Summary of findings for the main comparison. Any supplement containing iodine versus same supplement without iodine or no treatment/placebo (maternal outcomes)

Any supplement containing iodine versus same supplement without iodine or no treatment/placebo (maternal outcomes)
Population: women during the preconception, pregnancy and postpartum period Setting: Denmark, Germany, Morocco, New Zealand, Thailand, Zaire Intervention: any supplement containing iodine Comparison: same supplement without iodine or no treatment/placebo
Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Quality of the evidence (GRADE)	Comments
	Risk with same supplement without iodine or no treatment/placebo	Risk with any supplement containing iodine
Maternal hypothyroidism ‐ pregnancy	Study population		Average RR 1.90 (0.57 to 6.38)	365 (1 RCT)	⊕⊕⊝⊝ Low ^{1, 2}
	21 per 1000	40 per 1000 (12 to 134)
	Moderate
	21 per 1000	40 per 1000 (12 to 135)
Maternal hypothyroidism ‐ postpartum	Study population		Average RR 0.44 (0.06 to 3.42)	540 (3 RCTs)	⊕⊕⊝⊝ Low ^{2, 3}
	7 per 1000	4 per 1000 (1 to 34)
	Moderate
	11 per 1000	5 per 1000 (1 to 37)
Preterm birth	Study population		Average RR 0.71 (0.30 to 1.66)	376 (2 RCTs)	⊕⊕⊝⊝ Low ^{1, 2}
	111 per 1000	78 per 1000 (33 to 184)
	Moderate
	122 per 1000	87 per 1000 (37 to 202)
Maternal adverse effect: elevated thyroid peroxidase antibodies ‐ pregnancy	Study population		Average RR 0.95 (0.44 to 2.07)	359 (1 RCT)	⊕⊕⊝⊝ Low ^{1, 2}
	68 per 1000	65 per 1000 (30 to 142)
	Moderate
	68 per 1000	65 per 1000 (30 to 142)
Maternal adverse effect: elevated TPO‐ab ‐ postpartum	Study population		Average RR 1.01 (0.78 to 1.30)	397 (3 RCTs)	⊕⊕⊝⊝ Low ^{1, 4}
	301 per 1000	304 per 1000 (235 to 391)
	Moderate
	77 per 1000	78 per 1000 (60 to 100)
Maternal adverse effect: hyperthyroidism ‐ pregnancy	Study population		Average RR 1.90 (0.57 to 6.38)	365 (1 RCT)	⊕⊕⊝⊝ Low ^{1, 2}
	21 per 1000	40 per 1000 (12 to 134)
	Moderate
	21 per 1000	40 per 1000 (12 to 135)
Maternal adverse effect: hyperthyroidism ‐ postpartum	Study population		Average RR 0.32 (0.11 to 1.91)	543 (3 RCTs)	⊕⊕⊝⊝ Low ^{2, 4}
	49 per 1000	16 per 1000 (5 to 45)
	Moderate
	6 per 1000	3 per 1000 (1 to 7)
Maternal adverse effect: digestive intolerance ‐ pregnancy	Study population		Average RR 15.33 (2.07 to 113.70)	76 (1 RCT)	⊕⊝⊝⊝ Very low ^{2, 5}
	22 per 1000	333 per 1,000 (45 to 1,000)
	Moderate
	25 per 1000	383 per 1000 (52 to 1,000)
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: Confidence interval; RR: Risk ratio; TPO‐ab: thyroid peroxidase antibodies
GRADE Working Group grades of evidence High quality: We are very confident that the true effect lies close to that of the estimate of the effect Moderate quality: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different Low quality: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect Very low quality: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect
¹One study with design limitations (high risk for attrition bias). ² Wide confidence interval. ³Most studies contributing data had design limitations (high risk for attrition bias and unclear for selection bias). ⁴Most studies contributing data had design limitations (high risk for attrition bias and selection bias). ⁵One study with high risk of blinding and attrition bias.

Summary of findings 2. Any supplement containing iodine versus same supplement without iodine or no treatment/placebo (child outcomes)

Any supplement containing iodine versus same supplement without iodine or no treatment/placebo (child outcomes)
Population: women during the preconception, pregnancy and postpartum period Setting: Germany, New Zealand, Thailand, Zaire Intervention: any supplement containing iodine Comparison: same supplement without iodine or no treatment/placebo
Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Quality of the evidence (GRADE)	Comments
	Risk with same supplement without iodine or no treatment/placebo	Risk with any supplement containing iodine
Perinatal mortality	Study population		Average RR 0.66 (0.42 to 1.03)	399 (1 RCT)	⊕⊕⊝⊝ Low ^{1, 2}
	208 per 1000	137 per 1000 (87 to 214)
	Moderate
	208 per 1000	137 per 1000 (87 to 214)
Low birthweight	Study population		Average RR 0.56 (0.26 to 1.23)	377 (2 RCTs)	⊕⊕⊝⊝ Low ^{2, 3}
	90 per 1000	51 per 1000 (24 to 111)
	Moderate
	96 per 1000	54 per 1000 (25 to 118)
Neonatal hypothyroidism or elevated TSH	Study population		Average RR 0.58 (0.11 to 3.12)	219 (1 RCT)	⊕⊝⊝⊝ Very low ^{2, 3, 5}
	34 per 1000	20 per 1000 (4 to 106)
	Moderate
	34 per 1000	20 per 1000 (4 to 106)
Neonatal adverse effect: elevated TPO‐ab	Study population		Average RR 0.61 (0.07 to 5.70)	108 (1 RCT)	⊕⊝⊝⊝ Very low ^{2, 4, 5}
	43 per 1000	26 per 1000 (3 to 244)
	Moderate
	43 per 1000	26 per 1000 (3 to 245)
Child adverse effect: hyperthyroidism			See comments			No trials reported this outcome



*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: Confidence interval; RR: Risk ratio; TPO‐ab: thyroid peroxidase antibodies; TSH: thyroid‐stimulating hormone
GRADE Working Group grades of evidence High quality: We are very confident that the true effect lies close to that of the estimate of the effect Moderate quality: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different Low quality: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect Very low quality: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect
¹Design limitations (high risk of attrition bias and unclear for selection bias). ²Wide confidence interval. ³One study had design limitations (high risk for attrition bias). ⁴One study with design limitations (high risk for selection bias). ⁵Small number of events.

Background

Description of the condition

Iodine is an essential nutrient required for the biosynthesis of thyroid hormones thyroxine (T4) and triiodothyronine (T3), which are responsible for regulating growth, development and metabolism. Very little iodine is required to meet the body's needs, ranging from 90 µg to 290 µg per day depending on age and physiological status (IOM 2001). However it can be challenging to meet these needs given that naturally‐occurring iodine is low in most foods and beverages, with the exception of seafood, including fish, shellfish and plants from the ocean (Rohner 2014). For many people, the major source of dietary iodine is added, from salt fortified with iodine or from dairy products in settings where iodine is added to animal feed or where iodine‐containing antiseptics are used.

Iodine in preconception, pregnancy and breastfeeding

Throughout pregnancy there are major alterations in thyroid function as a result of metabolic demands and hormonal changes (Glinoer 1997). The concentrations of T4 and T3 rise significantly until approximately mid‐gestation and then remain relatively stable until the end of gestation at term. Iodine requirements increase substantially during pregnancy; initially as a result of a 50% increase in thyroid hormone production and a 30% to 50% increase in the renal excretion of iodine (Glinoer 2007), and later in gestation when iodine passes through the placenta for fetal production of thyroid hormones (Glinoer 1997). Maternal and fetal thyroid hormones are essential in regulating the development of the fetal brain and nervous system including the creation and growth of nerve cells, the formation of synapses, which are required for communication between nerve cells, and myelination, which is the formation of a fat‐based layer that allows for fast communication between nerve cells (Prado 2014). Some of these events begin in the second month of gestation (Prado 2014), so may be influenced by iodine status and thyroid hormone production prior to conception (or prior to knowledge of pregnancy).

During breastfeeding, thyroid hormone production and urinary iodine excretion return to non‐pregnancy levels, but iodine requirements remain elevated because iodine is concentrated in the mammary gland for excretion in breast milk (Leung 2011). Breastmilk iodine content varies with maternal dietary iodine intake and is lowest in iodine‐deficient areas and highest where additional iodine is routinely provided through supplements or universal salt iodization (Azizi 2009). As long as maternal iodine intake is adequate, breast milk can meet infant iodine needs.

Recommended iodine intakes

Minute amounts of iodine are required to meet iodine needs and prevent deficiency. Different agencies have recommended different intakes to meet iodine needs of non‐pregnant, pregnant and breastfeeding women, ranging from 150 μg to 290 μg a day. The International Council for Control of Iodine Deficiency Disorders (ICCIDD), World Health Organization (WHO), and United Nations Children's Fund (UNICEF) recommended a daily iodine intake of 250 μg for pregnant women and those breastfeeding their babies (WHO/UNICEF/ICCIDD 2007). The US Institute of Medicine established a recommended dietary allowance of 220 μg a day for women during pregnancy and 290 μg a day whilst breastfeeding (IOM 2001). The European Food Safety Authority set the adequate intake among pregnant women and women who are breastfeeding their infants at 200 μg/day (EFSA 2014). All three recommendations suggest 150 μg a day for non‐pregnant adult women though recommended intakes for the adolescent period differ.

Both the US Institute of Medicine and the European Commission Scientific Committee on Food have set tolerable upper intake levels for iodine, which are the highest levels of daily intake not likely to pose a risk of adverse health effects in the general population (IOM 2001; SCF 2006). These two organizations' recommendations vary greatly for non‐pregnant, pregnant and breastfeeding women though, with the US Institute of Medicine level at 1100 µg a day and the European Commission level set at 600 µg a day. These recommended intakes are not intended for iodine‐deficient populations however, as metabolic adaptations to deficiency could result in adverse health effects at lower intakes (Rohner 2014).

Burden of iodine deficiency

It is estimated that over 1.8 billion people worldwide have an insufficient iodine intake, putting them at risk of iodine deficiency (Andersson 2012). Europe is the region with the highest proportion of individuals with insufficient intake (44%), whereas Southeast Asia has the highest number (540 million). Given the elevated iodine requirements during pregnancy and breastfeeding and the importance of thyroid hormones for growth and development of the nervous system, ensuring adequate status in women and young children is critical. However, as only a limited number of countries have completed surveys in pregnant women and women of reproductive age on national or large sub‐national levels, there are insufficient data to directly estimate the regional or global prevalence of low iodine intake in these important target groups (Wong 2011).

Consequences of iodine deficiency

The consequences of iodine deficiency, which result from inadequate thyroid hormone production and can affect individuals of any age, sex, and physiological status, are collectively known as iodine deficiency disorders (Hetzel 1983; WHO/UNICEF/ICCIDD 2007). They are most serious and may be irreversible if they occur during pregnancy or early childhood, though the effects depend on the timing and severity of iodine deficiency (Zimmermann 2012a). If the increased iodine requirements during pregnancy are not met, the concentration of T4 diminishes (hypothyroxinaemia) in both mother and fetus, which can lead to irreversible brain damage with intellectual disability and neurologic abnormalities (Glinoer 2007; Williams 2008). Iodine deficiency in pregnancy has been associated with maternal and fetal goitre and hypothyroidism, increased pregnancy loss and infant mortality, and impairments in child development (Dunn 1993; Dunn 2001). Symptoms of hypothyroidism, or an underactive thyroid, include tiredness, weakness, poor memory and difficulty concentrating, weight gain, feeling cold, dry skin and hair loss (Jameson 2005). Globally, iodine deficiency is the most common cause of hypothyroidism (Jameson 2005) and the most common preventable cause of impaired brain development and mental function (WHO/UNICEF 2007).

The effect of iodine deficiency on child development can manifest itself as anything from mild intellectual blunting to cretinism, with a large proportion of the population experiencing intellectual impairment somewhere in between these extremes (Glinoer 2000). Endemic cretinism, the most serious iodine deficiency disorder, is a permanent condition of severely stunted physical and mental development that results from an untreated congenital deficiency of thyroid hormones caused by severe maternal iodine deficiency (Hetzel 1983). It is still unclear whether mild‐to‐moderate maternal iodine deficiency in humans produces changes in cognitive function. Some evidence suggests that children with chronic severe iodine deficiency have significantly lower intelligence quotient (IQ) scores than those with adequate iodine status (Zimmermann 2009). Some randomised clinical studies conducted in school‐age children indicate that cognitive performance may improve with iodine supplementation, even in mildly deficient areas (Gordon 2009; Melse‐Boonstra 2010). In the absence of iodine these effects are largely preventable by immediate thyroid hormone replacement, although deficits in memory and IQ may persist over time (Williams 2008). It is also thought that iodine deficiency could be associated with autism (Sullivan 2008) and with children's attention deficit and hyperactivity disorder (Vermiglio 2004).

A variety of factors affect iodine metabolism and thyroid function and can exacerbate iodine deficiency. Other nutritional deficiencies, including of selenium and iron, can lead to decreased thyroid hormone production and cause damage to the thyroid because key enzymes required for thyroid hormone synthesis and metabolism depend on these nutrients (Zimmermann 2002). Iron supplementation has been shown to improve the efficacy of iodine supplementation or salt iodization in iodine‐ and iron‐deficient children (Hess 2002; Zimmermann 2000).

Compounds found naturally in some foods such as cassava, sorghum, soy and millet, and pollutants in food and water such as perchlorate and nitrate can also negatively affect iodine metabolism and thyroid function (Rohner 2014). Collectively known as goitrogens, these substances can compete with iodine for uptake by the thyroid and impair the activity of key enzymes required to produce thyroid hormones. Infants and young children appear particularly vulnerable to the effects of goitrogens, and effects are generally only seen where there is pre‐existing iodine deficiency.

Consequences of excess iodine

Excess iodine exposure can also cause serious negative health effects, and can occur through ingestion of supplements, water or foods with high iodine content or via medical treatments or procedures. Acute iodine poisoning may cause gastrointestinal or cardiovascular symptoms, or even coma, after ingestion of many grams of iodine (Zimmermann 2008). Excess iodine can also cause the thyroid to become over or underactive (hyper or hypothyroidism). Hypothyroidism symptoms are described above. Symptoms of excess thyroid hormone production as a result of hyperthyroidism include hyperactivity, irritability, heat intolerance, palpitations, weakness, fatigue and weight loss (Jameson 2005).

In iodine sufficiency, the thyroid is able to adjust to a wide range of iodine intakes, so healthy individuals may tolerate up to 1 mg daily (Zimmermann 2008). In areas of chronic iodine deficiency however, individuals are less tolerant to high iodine intake, especially older adults with longstanding goitre. Iodine‐induced hyperthyroidism has been reported in the initial phases of salt iodization programmes, though it is nearly always temporary (WHO/UNICEF/ICCIDD 2007). It was estimated that 11 countries have excessive iodine intakes (up from seven in 2007) as presented at the Sixty‐sixth World Health Assembly in a progress report from the WHO in 2013.

Indicators of iodine status

Urinary iodine concentration (UIC)

In conditions of iodine sufficiency, over 90% of ingested iodine is excreted in the urine, whereas in chronic iodine deficiency the excretion may be less than 20%, making urinary iodine concentration (UIC) a good indicator of recent iodine intake, or short‐term iodine status (Rohner 2014). UIC has limited utility in assessing individual intake or status because of large variations within and between days (WHO 2013). These variations level out in large population samples though, making UIC a useful population‐level indicator. UIC is not a direct indicator of thyroid function, but low values suggest a greater risk of developing thyroid disorders (Rohner 2014).

UIC is commonly collected from school age children and extrapolated to the general population or other population groups; however, neither this group nor non‐pregnant women serve as an adequate proxy for pregnant women (Wong 2011). For pregnant women, a median UIC below 150 μg/L is indicative of insufficient iodine intake, 150 μg/L to 249 μg/L adequate iodine intake, 250 μg/L to 499 μg/L above iodine requirements, and 500 μg/L or higher concentrations suggest an excessive intake (beyond that needed for prevention and control of iodine deficiency) (WHO 2013). For women who are breastfeeding their infants and children under two years of age, a median UIC of below 100 μg/L indicates insufficient iodine intake, and 100 μg/L or more indicates adequate iodine intake. For these population groups, the category of iodine intake is not extrapolated to category of iodine status.

Goitre

The development of goitre, or enlargement of the thyroid gland, begins as an adaptive response when iodine available to the thyroid is not sufficient for adequate thyroid hormone production (Rohner 2014). Goitre responds slowly to changes in iodine intake and is therefore an indicator of longer‐term iodine status. In areas of chronic iodine deficiency it can take years for thyroid size to return to normal, and goitre may never completely disappear (Zimmermann 2012b).

The presence of goitre can be determined by neck inspection and palpation or by thyroid ultrasonography (Rohner 2014; WHO/UNICEF/ICCIDD 2007). This method however has poor sensitivity and specificity in areas of mild‐to‐moderate iodine deficiency. In these settings, assessment of thyroid size by ultrasonography is preferred and technology is available for use in field settings. International reference values for thyroid volume using ultrasound are available only for school age children though (Zimmermann 2004).

Thyroid‐stimulating hormone (TSH)

This hormone, also known as thyrotropin, is produced by the pituitary gland and stimulates thyroid hormone production and release by the thyroid gland (Rohner 2014). Serum TSH levels increase in response to low thyroid hormone concentration and decrease in response to high concentration; it is a very sensitive indicator of thyroid function and is the primary screening test for thyroid dysfunction (Jameson 2005). Though TSH levels may rise in response to iodine deficiency, they are often in the normal range, therefore TSH is not considered a sensitive indicator of iodine status (Rohner 2014; Zimmermann 2008). TSH is a useful indicator of iodine nutrition in neonates though because they have high thyroidal iodine turnover (Delange 1998). Moderately elevated levels (higher than 5 mIU/L in whole blood) indicate neonatal iodine deficiency, which is a direct reflection of iodine deficiency during pregnancy (WHO/UNICEF/ICCIDD 2007). A prevalence of less than 3% of infants with moderately elevated TSH is expected in iodine‐sufficient regions (Rohner 2014; WHO/UNICEF/ICCIDD 2007). Neonatal TSH screening for congenital hypothyroidism is standard practice in many developed countries (WHO/UNICEF/ICCIDD 2007). This condition, which has genetic causes, is relatively rare (one in 4000 births) and indicated by highly elevated TSH levels (20 mIU/L or higher), and requires immediate treatment to prevent permanent neurological damage.

Because of a physiological surge in newborns, neonatal TSH assessment must take place at least 48 hours following birth (WHO/UNICEF/ICCIDD 2007). In some countries elevated neonatal TSH may be attributed to the use of beta‐iodine‐containing antiseptics so results need to be interpreted cautiously in these contexts.

Thyroid hormones

Triiodothyronine (T3) and its prohormone, thyroxine (T4), are hormones produced and secreted by the thyroid gland (Rohner 2014). T4 is converted to T3, the active hormone, in peripheral tissues. A very small amount (less than 1%) of T4 and T3 in the blood is not bound to protein, and therefore biologically active. In the past, thyroid hormone assessments typically measured total amounts because the free levels were too low to detect. However, with the advent of newer techniques, free T4 and free T3, which better reflect the physiological effects of thyroid hormones than total hormone concentrations, are usually measured nowadays.

T3 and T4 levels are direct clinical indicators of thyroid function, but levels are protected in the early stages of thyroid dysfunction and changes occur only at later stages (Rohner 2014). Therefore these thyroid hormones are not good indicators of iodine status, except in cases of severe iodine deficiency. A diagnosis of subclinical hypothyroidism is based on elevated TSH with normal T4 and T3 levels. Overt hypothyroidism occurs when thyroid function further diminishes, as indicated by falling free T4 levels, and the diagnosis is based on high TSH and low T4. T3 levels are typically maintained even longer than T4 levels. Some research has shown that neonatal T4 levels are lower in iodine‐deficient compared to iodine‐sufficient areas; however, validated norms have not been established for comparison.

Thyroglobulin (Tg)

Tg is a protein matrix for thyroid hormone synthesis (Glinoer 1997). Tg levels rise early in pregnancy, and the increase is most pronounced towards the end of gestation. In newborns, Tg levels are normally increased in the first few days, possibly in response to the physiological TSH surge (Pezzino 1981).

Tg is an indirect indicator of thyroid function and Tg levels are positively correlated with thyroid volume (Rohner 2014). Tg is elevated in iodine‐deficient populations and is also an indirect indicator of iodine status. Levels respond more quickly (weeks to months) to iodine repletion than TSH or goitre (Zimmermann 2008). Assays can be performed on samples collected on dried blood spots and international reference ranges are available for school‐age children (Zimmermann 2004). The presence of anti‐Tg antibodies, however, is reported to complicate interpretation of Tg values (Rohner 2014).

Complex changes in thyroid physiology can make interpretation of thyroid function and iodine status in pregnancy and early infancy difficult, therefore special considerations should to be taken into account to avoid misinterpreting results (Laurberg 2007). Though Stinca and colleagues recently reported that dried blood spot Tg is a sensitive indicator of iodine status in pregnant women and that Tg antibodies may not be necessary in this assessment (Stinca 2016).

Description of the intervention

Guidance and recommendations

Since 1993, universal salt iodization, or the addition of iodine to all salt for human and animal consumption including food industry salt, has been recommended for preventing and controlling iodine deficiency (UNICEF/WHO 1994; WHO 2014). As a result of this long‐standing recommendation and the related support for its implementation, most countries have some form of salt iodization program in place to prevent and control iodine deficiency and its consequences (UNICEF 2008). Universal salt iodization is widely acknowledged as a cost‐effective, feasible, and sustainable approach to control iodine deficiency, and research suggests that successful salt iodization programmes can meet the needs of population groups susceptible to iodine deficiency and its consequences, specifically pregnant and breastfeeding women and infants (Zimmermann 2007). However, it has been recognized that these groups may need to be targeted with other iodine interventions (Untoro 2007). WHO and UNICEF recommend considering iodine supplementation in pregnant women, women breastfeeding their infants and children from six to 23 months of age, alongside efforts to scale up salt iodization, in settings where large proportions of the population do not have access to iodized salt (WHO/UNICEF 2007). In addition, where pregnant women are difficult to reach, WHO and UNICEF recommend that supplementation be extended to all women of reproductive age. In some countries, for example the USA, Canadian and Australian medical bodies have issued specific recommendations for iodine supplementation in women who are pregnant or breastfeeding (and women considering becoming pregnant in Australia) (ATA 2006; NHMRC 2010).

Supplement form, dose, and regimen

Oral supplements

Several different types of oral iodine supplements are available for public health purposes. These can be broadly divided into frequent low dose (such as daily or weekly), or infrequent high dose (such as annually or only once). The low‐dose formulations usually contain iodine as potassium iodide and come in the form of tablets or drops for oral consumption. Many commercially available multiple‐micronutrient supplements including prenatal formulations also contain iodine, often 150 μg for a daily dose (Leung 2009).

High‐dose iodine supplements usually come in the form of oral iodized oil capsules; the iodine is stored mainly in the thyroid gland and can meet needs for up to a year. The above‐mentioned WHO/UNICEF guidance recommends a single annual dose of 400 mg or a daily dose of 250 μg for pregnant and breastfeeding women (or 150 μg a day for non‐pregnant women) (WHO/UNICEF 2007), whereas the USA, Canadian and Australian recommendations suggest 150 μg/day (ATA 2006; NHMRC 2010).

Injectable supplements

Much of the early iodine supplementation research used high‐dose intramuscular iodized oil injections (e.g. Pharoah 1971 and Pretell 1972) and this approach was used in public health programmes especially in the 1970s and 1980s. This would currently be considered a medical intervention that should be provided under medical supervision. Other forms of medical iodine interventions are available including sodium iodide solution used in intravenous parenteral nutrition.

How the intervention might work

Iodine requirements increase during pregnancy because of increased thyroid hormone production and iodine excretion, and during breastfeeding because iodine is concentrated in the mammary gland for excretion in breast milk. If maternal iodine requirements are not met during this period, the production of thyroid hormones may decrease and be inadequate for maternal, fetal and infant needs (Glinoer 2007). Consequences may include maternal or child hypothyroidism or goitre, pregnancy loss, low birthweight, infant mortality, and developmental delays ranging from mild intellectual impairment to cretinism.

Additional iodine intake through iodine supplementation may help meet the increased iodine needs for thyroid hormone production and transfer to the fetus/infant during pregnancy and the postpartum period and prevent or correct iodine deficiency and its consequences. Iodine supplementation prior to conception could increase iodine stores and thyroid hormone production before the additional demands of pregnancy. This may be especially important in severely iodine‐deficient areas to allow time for correction of long‐standing deficiency. Even where iodine deficiency is less severe though, additional iodine intake prior to pregnancy may be warranted because thyroid hormones are important for brain and nervous system development events starting as early as the seventh week of gestation (Prado 2014), when women may not know or share with others that they are pregnant.

Previous evidence reviews reported that high‐dose iodine supplementation through intramuscular injection prior to, or early in pregnancy reduced the incidence of cretinism and improved child cognitive development scores in severely iodine‐deficient areas (Bougma 2013; Zimmermann 2012a). There is also evidence of improved birthweight, through oral iodine supplementation, and decreased child mortality, through injected iodized oil (Zimmermann 2012a).

From an implementation perspective, pregnant and postpartum women often have contact with the healthcare system, which generally provides or recommends prenatal and sometimes postnatal micronutrient supplementation ‐ usually iron and folic acid. Similarly, where indicated, iodine supplementation could be integrated into routine antenatal and postnatal care. To reach women prior to pregnancy, existing contacts with the healthcare system or other platforms could be used to provide or recommend iodine supplementation for those planning on becoming pregnant or to all women of reproductive age, or both, because pregnancies are often unplanned (as is done with folic acid supplementation recommendations in many settings).

Why it is important to do this review

It is important to assess the effects and safety of iodine supplementation in women before or during pregnancy and in the postpartum period for optimal maternal and child outcomes and to inform policy making towards the achievement of the WHO global targets for maternal, infant and young child nutrition by 2025 (WHA 2012).

This review will complement the findings of other existing reviews assessing the provision of iodine through a variety of interventions. Mahomed and colleagues conducted one of the first Cochrane Reviews on the topic (now withdrawn), examining maternal iodine supplementation in areas of iodine deficiency (Mahomed 2006). More recently a non‐Cochrane review examined the effect of prenatal or periconceptional iodine supplementation on child development, growth and other clinical outcomes (Zhou 2013). The effects of iodine supplementation for preventing iodine deficiency disorders in children (Angermayr 2004) and mortality and adverse neurodevelopmental outcomes in preterm infants (Ibrahim 2006) are addressed in other Cochrane Reviews. A review on salt iodization for prevention of iodine deficiency disorders is published elsewhere (Aburto 2014) and a Cochrane Review on fortification of foods and condiments (other than salt) with iodine for prevention of iodine deficiency disorders (Land 2013) is being conducted. Furthermore, a review on the effect of iodized salt or iodine supplements on prenatal and postnatal growth (Farebrother 2015) is also underway. Cochrane reviews have also been conducted on point‐of‐use fortification of foods with multiple‐micronutrient powders (Suchdev 2015) and multiple‐micronutrient supplementation (Haider 2015) for women during pregnancy.

Objectives

Methods

Criteria for considering studies for this review

Types of studies

We included randomised and quasi‐randomized controlled trials with randomisation at either the individual or cluster level. We intended to include eligible cross‐over trials (and use only the results from the first period) however we did not identify any such studies.

Types of participants

Women who become pregnant, or pregnant or postpartum women of any chronological age and parity (number of births), regardless as to the iodine status of the study population or setting. We included studies that randomised women to receive treatment starting at any point prior to conception, during pregnancy or within the first six weeks postpartum. We planned to exclude studies specifically targeting women screened for and diagnosed with thyroid disorders (as defined by trial authors) or other health problems (e.g. thyroid disease, HIV, tuberculosis) however we did not identify any such studies.

Types of interventions

Injected or oral iodine supplementation (such as tablets, capsules, drops) during preconception, pregnancy or the postpartum period irrespective of compound, dose, frequency or duration.

Specifically, where data were available we planned to assess the following comparisons:

any supplement containing iodine versus same supplement without iodine or no intervention/placebo;
any oral iodine supplement versus same supplement without iodine or no intervention/placebo;
oral iodine‐only supplement versus no intervention or placebo;
oral iodine supplement with other vitamins and/or minerals versus only other vitamins and/or minerals (exact same formulation of other vitamins/minerals, but no iodine);
any injected iodine supplement versus same supplement without iodine or no intervention/placebo;
injected iodine‐only supplement versus no intervention or placebo;
injected iodine supplement with other vitamins and/or minerals versus only other vitamins and/or minerals (exact same formulation of other vitamins/minerals, but no iodine).

We only identified data for comparisons 1 through 6. Because we did not identify any studies for comparison 7, comparisons 5 and 6 include the exact same data and we have only reported comparison 5. We have summarized comparisons 2, 3, 4 and 5 in comparison 1, and we have summarized comparisons 3 and 4 in comparison 2. We have summarized comparison 1, the main comparison, in the 'Summary of findings' tables (summary of findings Table for the main comparison and summary of findings Table 2).

We included interventions that combined iodine supplementation with co‐interventions (e.g. education), only if the co‐interventions were the same across study arms. We excluded studies that examined tube feeding, parenteral nutrition, or food‐based interventions (e.g. fortified or bio fortified foods, point‐of‐use fortification with micronutrient powders or lipid‐based nutrient supplements).

Types of outcome measures

We included studies which met the above‐mentioned criteria regardless of outcomes reported, though we only extracted the outcomes described below.

Primary outcomes

Maternal

Hypothyroidism (as defined by trial authors)
Preterm birth (as defined by trial authors)
Adverse effects
1. Elevated thyroid peroxidase antibodies (TPO‐ab) (as defined by trial authors)
2. Hyperthyroidism (as defined by trial authors)
3. Digestive intolerance (as defined by trial authors)

Infants and children ‐ to 23 months of age

Perinatal mortality (including stillbirth/fetal death and neonatal death, as defined by trial authors)
Low birthweight (less than 2500 g)
Hypothyroidism or elevated TSH (as defined by trial authors)
Adverse effects
1. Elevated TPO‐ab (as defined by trial authors)
2. Hyperthyroidism (as defined by trial authors)

Secondary outcomes

Maternal

Spontaneous miscarriage (as defined by trial authors)
Thyroid size (assessed by any method)
Thyroglobulin (Tg) (μg/L)
Insufficient iodine intake (pregnancy: median urinary iodine concentration (UIC) less than 150 μg/L, breastfeeding: median UIC less than 100 μg/L)
Excessive iodine intake (pregnancy only: median UIC greater than or equal to 500 μg/L)

Infants and children ‐ to 23 months of age

Small‐for‐gestational age (as defined by trial authors)
Congenital anomalies (including cretinism, as defined by trial authors)
Growth (anthropometric Z scores)
Thyroid size (assessed by any method)
Insufficient iodine intake (median UIC less than 100 μg/L)
Mental or motor development (as defined by trial authors)
Infant death (death in the first year of life)

For relevant maternal outcomes, we included both the pregnancy and postpartum period. If maternal outcomes were assessed at multiple time points, unless otherwise specified, we used the last assessment during pregnancy and the last assessment in the postpartum period. We planned to include the last assessment during the intervention period and the last assessment during the follow‐up period though did not because many of the trials were single dose and therefore this distinction was not relevant, and also because of the way data were presented in most of the trials. We did not include any outcomes during the preconception period. For infant and child outcomes assessed at multiple time points, where relevant we planned to include the neonatal period (to 28 days), infancy (under one year), and childhood (beyond one year). We summarized primary outcomes in 'Summary of findings' tables (summary of findings Table for the main comparison, summary of findings Table 2).

We planned to include any adverse event, such as iodine‐induced hyperthyroidism, in the mother and child as primary outcomes. None of the trials which reported on hyperthyroidism indicated whether it was or suspected to be iodine‐induced (none of these trials took place in severely‐deficient women so it is unlikely that the hyperthyroidism would have been caused by the supplemental iodine). The only other reported outcomes considered by the trial or review authors to be potentially adverse were elevated TPO‐ab and digestive intolerance. Therefore we replaced any adverse effects with hyperthyroidism and elevated TPO‐ab (and digestive intolerance for women), and extracted these outcomes separately to avoid losing potentially important information.

Urinary and breast milk iodine concentrations are probably the most common indicators of iodine status; they provide important information on the adherence and biological response to the intervention or potential contamination in the control group. However these indicators are generally highly skewed and are typically reported as median with some description of the range. As we did not have access to primary data, we have presented these outcomes in Table 1 but they were not included in a meta‐analysis and did not directly inform the conclusions of the review.

See: Summary of findings for the main comparison Any supplement containing iodine versus same supplement without iodine or no treatment/placebo (maternal outcomes); Summary of findings 2 Any supplement containing iodine versus same supplement without iodine or no treatment/placebo (child outcomes)

Table 1. Urinary iodine concentration and breast milk iodine concentration following prepregnancy, pregnancy or postpartum iodine supplementation

Study	Setting	Intervention timing	Intervention description	Timing of assessment	Urinary iodine concentration		Breast milk iodine concentration
Study	Setting	Intervention timing	Intervention description	Timing of assessment	Iodine	Control	Iodine	Control
Bouhouch 2014 (C)	Southern Morocco	Postpartum	Iodine: single dose oral 400 mg, control: single dose placebo	Baseline	Median: 37 μg/L, IQR: 22‐77 (n = 119)	Median: 30 μg/L, IQR: 18‐61 (n = 115)
Bouhouch 2014 (C)	Southern Morocco	Postpartum		Postpartum: 9 months	Median: 58 μg/L, IQR: 34‐135 (n = 94)	Median: 39 μg/L, IQR 24‐62 (n = 81)	Median: 39.4 μg/L, IQR: 23.5‐66.7 (n = 94)	Median: 26.2 μg/L, IQR: 17.7‐42.7 (n = 81)
Brucker‐Davis 2013	Nice, France	Pregnancy	Iodine: daily dose 150 μg of iodine, control: vitamin mix but no iodine	Baseline: before 12 weeks of amenorrhoea	Median: 111 μg/L, IQR: 28‐399 (n = 32)	Median: 103 μg/L, IQR: 14‐355 (n = 54)
				Pregnancy: 3rd trimester	Median: 160.5 μg/L, IQR: 18‐358 (n = 30)	Median: 76 μg/L, IQR: 16‐303 (n = 46)
				Postpartum: 3 months	Median: 58 μg/L, IQR: 34‐135 (n = 18)	Median: 58 μg/L, IQR: 34‐135 (n = 18)
Glinoer 1993	Brussels, Belgium	Pregnancy	Iodine: single dose 1 mL iodized oil, control: multivitamin injection no iodine	Neonatal: 3‐6 days	Mean: 77 μg/L, SEM: +/‐ 8 (n = 60)	Mean: 43 μg/L, SEM: +/‐ 4 (n = 60)	Mean: 61 μg/L, SEM: +/‐10, (n = 60)	Mean: 29 μg/L, SEM: +/‐2, (n = 60)
Liesenkotter 1996	Berlin, Germany	Pregnancy	Daily dose oral 300 µg KI, control: no intervention	Postpartum: mean 11 days	Median: 104.5 μg/dl, IQR: NR (n = NR)	Median: In fig only (n = NR)
Liesenkotter 1996	Berlin, Germany	Pregnancy	Daily dose oral 300 µg KI, control: no intervention	Neonatal: 5 days	Median: 8.3 μg/dl, IQR: In fig only (n = NR)	Median: 6.5 μg/dl, IQR: In fig only (n = NR)
Mulrine 2010	Dunedin, New Zealand	Postpartum	Iodine: daily 75 or 150 µg/d iodine, control: placebo	Postpartum: 24 weeks (75 µg)	Median: 78 μg/L, 25th 75th %: 50,126 (n = 18)	Median: 34 μg/L, 25th 75th %: 26, 57 (n = 21)
Mulrine 2010	Dunedin, New Zealand	Postpartum	Iodine: daily 75 or 150 µg/d iodine, control: placebo	Postpartum: 24 weeks (150 µg)	Median: 84 μg/L, 25th‐75th %: 60,157 (n = 19)	Median: 34 μg/L, 25th 75th %: 26,57 (n = 21)	In figure only	In figure only
Nohr 2000	Aalborg, Denmark	Pregnancy or pregnancy and postpartum	Iodine: daily oral 150 µg elemental iodine and micronutrient supplement, control: micronutrient supplement no iodine	Pregnancy: 35 weeks	Median: 86 μg/24 h, 25th‐75th %: 66–147 (n = 22) pregnancy and postpartum	Median: 88 μg/24 h, 25th 75th %: 68‐123 (n = 24)
					Median: 83 μg/24 h, 25th 75th %: 54–145 (n = 20) pregnancy only
					Median: 50 μg/L, 25th 75th %: 35‐101 (n = 22) pregnancy and postpartum	Median: 51 μg/L, 25th 75th %: 30–80 (n = 20)
					Median: 52 μg/L, 25th 75th %: 36–81 (n = 20) pregnancy only
Pedersen 1993	East Jutland, Denmark	Pregnancy	Iodine: daily drops 200 μg iodine, control: no treatment	Pregnancy: 37 weeks	Median: 106 μg/L, 95 %: NR (n = 28)	Median: 54 μg/L, 95 %: NR (n = 26)
				Postpartum: 12 months	Median: 56 μg/L, 95 %: NR (n = NR)	Median: 54 μg/L, 95 %: NR (n = NR)	Median: 41 μg/L, 95 %: 31‐74 (n = 27)	Median: 28 μg/L, 95 %: 19‐46 (n = 26)
				Neonatal: 5 days	Median: 64 μg/L, 95 %: 34‐70 (n = 27)	Median: 27 μg/L, 95 %: 21‐56 (n = 25)
Silva 1981	Eastern Greater Santiago, Chile	Pregnancy	Iodine: daily drops approximately 300 µg iodine, control: no intervention	Pregnancy: 15‐40 weeks	Mean: 376 μg iodine/g creatinine, SD: 456 (n = 36)	Mean: 56 μg iodine/g creatinine, SD: 25 (n = 10)
Silva 1981	Eastern Greater Santiago, Chile	Pregnancy		Postpartum: at delivery	Mean: 493 μg iodine/g creatinine, SD: 586 (n = not specified)	Mean: 127 μg iodine/g creatinine, SD: 94 (n = not specified)
Zhou 2015	Adelaide, New Zealand	Pregnancy	Iodine: daily tablet 150 μg iodine, control: placebo	Postpartum: 6 weeks			Median (IQR): 106.0 (84.0‐146.0), n = 19	Median (IQR): 124.0 (76.0‐155.0), n = 22
Gowachirapant 2014	Bangkok, Thailand	Pregnancy	Iodine: daily 200 µg iodine and multivitamin‐mineral tablets, control: multivitamin‐mineral tablets with no iodine	Pregnancy: 30 weeks	Median: 233.3 μg/L, IQR: 140.0, 314.8 (n = 171)	Median: 154.6 μg/L, IQR: 107.9, 212.4 (n = 185)
				Postpartum: 6 weeks	Median: 115.5 μg/L, IQR: 73.0, 155.3 (n = 129)	Median: 93.6 μg/L, IQR: 60.3, 159.8 (n = 135)
				Neonatal: 6 weeks	Median: 168.7 μg/L, IQR: 97.7, 305.7 (n = 83)	Median: 193.3 μg/L, IQR: 136.7, 303.5 (n = 76)

IQR: interquartile range
KI: potassium iodide
NR: not reported
SD: standard deviation
SEM: standard error of the mean

Search methods for identification of studies

The following methods section is based on a standard template used by Cochrane Pregnancy and Childbirth.

Electronic searches

We searched Cochrane Pregnancy and Childbirth’s Trials Register by contacting their Information Specialist (14 November 2016).

The Register is a database containing over 22,000 reports of controlled trials in the field of pregnancy and childbirth. For full search methods used to populate Pregnancy and Childbirth’s Trials Register including the detailed search strategies for CENTRAL, MEDLINE, Embase and CINAHL; the list of handsearched journals and conference proceedings, and the list of journals reviewed via the current awareness service, please follow this link to the editorial information about the Cochrane Pregnancy and Childbirth in the Cochrane Library and select the ‘Specialized Register ’ section from the options on the left side of the screen.

Briefly, Cochrane Pregnancy and Childbirth’s Trials Register is maintained by their Information Specialist and contains trials identified from:

monthly searches of the Cochrane Central Register of Controlled Trials (CENTRAL);
weekly searches of MEDLINE (Ovid);
weekly searches of Embase (Ovid);
monthly searches of CINAHL (EBSCO);
handsearches of 30 journals and the proceedings of major conferences;
weekly current awareness alerts for a further 44 journals plus monthly BioMed Central email alerts.

Search results are screened by two people and the full text of all relevant trial reports identified through the searching activities described above is reviewed. Based on the intervention described, each trial report is assigned a number that corresponds to a specific Pregnancy and Childbirth review topic (or topics), and is then added to the Register. The Information Specialist searches the Register for each review using this topic number rather than keywords. This results in a more specific search set which has been fully accounted for in the relevant review sections (Included studies; Excluded studies; Studies awaiting classification; Ongoing studies).

In addition, we searched the WHO International Clinical Trials Registry Platform (ICTRP) (17 November 2016) for unpublished, planned and ongoing trials using search terms as described in Appendix 1.

Searching other resources

We searched through the bibliographies of existing reviews on similar topics and of retrieved studies. We contacted experts in the field as well as authors of retrieved studies for lists of other studies that should be considered for inclusion.

For assistance in identifying ongoing or unpublished studies, we also contacted the WHO Departments of Reproductive Health and Research and Nutrition for Health and Development, the Iodine Global Network, the nutrition section of the United Nations Children's Fund (UNICEF), the World Food Programme (WFP), the US Centers for Disease Control and Prevention (CDC), the Micronutrient Initiative (MI), the Global Alliance for Improved Nutrition (GAIN), Hellen Keller International (HKI), and Sight and Life (28 November 2016).

We did not apply any date or language restrictions.

Data collection and analysis

The following methods section is based on a standard template used by Cochrane Pregnancy and Childbirth.

Selection of studies

Two review authors (KBH and LDR) independently assessed for inclusion all the potential studies we identified as a result of the search strategy. We resolved any disagreement through discussion and, if required, consulted with a third review author (ACW). For one study (Mohammed 2015), funded by the organization of the two authors doing the eligibility, another author (JPP) assessed and determined it was excluded. We included studies published as abstracts but if we could not assess quality and extract information (after attempting to contact study authors), these studies were marked as "awaiting classification".

Data extraction and management

We designed a form to extract data and record information requested from study authors. For eligible studies, data were extracted once using the agreed form by three review authors (KBH, BAP and CMYY). Two review authors (KBH and BAP) entered the data into Review Manager 5 (RevMan 5) software (RevMan 2014). We resolved discrepancies through discussion or, if required, consulted a third person.

When information regarding any of the above was unclear, we attempted to contact authors of the original reports to provide further details.

Assessment of risk of bias in included studies

Two review authors independently assessed risk of bias for each study using the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011a). We resolved any disagreement by discussion or by involving a third assessor.

(1) Random sequence generation (checking for possible selection bias)

We described for each included study the method used to generate the allocation sequence in sufficient detail to allow an assessment of whether it should produce comparable groups.

We assessed the method as:

low risk of bias (any truly random process, e.g. random number table; computer random number generator);
high risk of bias (any non‐random process, e.g. odd or even date of birth; hospital or clinic record number);
unclear risk of bias.

(2) Allocation concealment (checking for possible selection bias)

We described for each included study the method used to conceal allocation to interventions prior to assignment and assess whether intervention allocation could have been foreseen in advance of, or during recruitment, or changed after assignment.

We assessed the methods as:

low risk of bias (e.g. telephone or central randomisation; consecutively‐numbered sealed opaque envelopes);
high risk of bias (open random allocation; unsealed or non‐opaque envelopes, alternation; date of birth);
unclear risk of bias.

(3.1) Blinding of participants and personnel (checking for possible performance bias)

We described for each included study the methods used, if any, to blind study participants and personnel from knowledge of which intervention a participant received. We considered that studies were at low risk of bias if they were blinded, or if we judged that the lack of blinding would be unlikely to affect results. Where relevant, we planned to assess blinding separately for different outcomes or classes of outcomes.

We assessed the methods as:

low, high or unclear risk of bias for participants;
low, high or unclear risk of bias for personnel.

(3.2) Blinding of outcome assessment (checking for possible detection bias)

We described for each included study the methods used, if any, to blind outcome assessors from knowledge of which intervention a participant received. Where relevant, we planned to assess blinding separately for different outcomes or classes of outcomes.

We assessed methods used to blind outcome assessment as:

low, high or unclear risk of bias.

(4) Incomplete outcome data (checking for possible attrition bias due to the amount, nature and handling of incomplete outcome data)

We described for each included study, and for each outcome or class of outcomes, the completeness of data including attrition and exclusions from the analysis. We stated whether attrition and exclusions were reported and the numbers included in the analysis at each stage (compared with the total randomised participants), reasons for attrition or exclusion where reported, and whether missing data were balanced across groups or were related to outcomes. Where sufficient information was reported, or could be supplied by the trial authors, we planned to re‐include missing data in the analyses which we undertook.

We assessed methods as:

low risk of bias (e.g. no missing outcome data; missing outcome data were less than 20% and were balanced across groups);
high risk of bias (e.g. numbers or reasons for missing data imbalanced across groups; ‘as‐treated’ analysis done with substantial departure of intervention received from that assigned at randomisation);
unclear risk of bias (e.g. level of missing data was unclear).

(5) Selective reporting (checking for reporting bias)

We described for each included study how we investigated the possibility of selective outcome reporting bias and what we found.

We assessed the methods as:

low risk of bias (where it was clear that all of the study’s pre‐specified outcomes and all expected outcomes of interest to the review were reported);
high risk of bias (where not all the study’s pre‐specified outcomes were reported; one or more reported primary outcomes were not pre‐specified; outcomes of interest were reported incompletely and so could not be used; study failed to include results of a key outcome that would have been expected to have been reported);
unclear risk of bias.

(6) Other bias (checking for bias due to problems not covered by (1) to (5) above)

We described for each included study any important concerns we had about other possible sources of bias.

We assessed whether each study was free of other problems that could put it at risk of bias:

low risk of other bias;
high risk of other bias;
unclear whether there was risk of other bias.

(7) Overall risk of bias

We made explicit judgements about whether studies were at high risk of bias, according to the criteria given in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011a). With reference to (1) to (6) above, we assessed the likely magnitude and direction of the bias and whether we considered it was likely to impact on the findings. We planned to explore the impact of the level of bias through undertaking sensitivity analyses however we considered only one trial at low overall risk of bias therefore this analysis would not have been meaningful.

Assessing the quality of the evidence using GRADE

We assessed the overall quality of the evidence for primary outcomes using the GRADE approach as outlined in the GRADE handbook in order to assess the quality of the body of evidence relating to the primary outcomes for any supplement containing iodine versus same supplement without iodine or no intervention/placebo. Two review authors independently assessed the quality of the evidence for each of the maternal and child primary outcomes.

Maternal outcomes

Hypothyroidism (as defined by trial authors)
Preterm birth (as defined by trial authors)
Adverse effects
1. Elevated thyroid peroxidase antibodies (TPO‐ab) (as defined by trial authors)
2. Hyperthyroidism (as defined by trial authors)
3. Digestive intolerance (as defined by trial authors)

Infant/child outcomes

Perinatal mortality (including stillbirth/fetal death and neonatal death, as defined by trial authors)
Low birthweight (less than 2500 g)
Hypothyroidism or elevated TSH (as defined by trial authors)
Adverse effects
1. Elevated TPO‐ab (as defined by trial authors)
2. Hyperthyroidism (as defined by trial authors)

We used the GRADEpro Guideline Development Tool to import data from RevMan 5 (RevMan 2014) in order to create 'Summary of findings' tables (summary of findings Table for the main comparison; summary of findings Table 2). We produced a summary of the intervention effect and a measure of quality for each of the above outcomes using the GRADE approach. The GRADE approach uses five considerations (study limitations, consistency of effect, imprecision, indirectness and publication bias) to assess the quality of the body of evidence for each outcome. The evidence can be downgraded from 'high quality' by one level for serious (or by two levels for very serious) limitations, depending on assessments for risk of bias, indirectness of evidence, serious inconsistency, imprecision of effect estimates or potential publication bias.

Measures of treatment effect

Dichotomous data

For dichotomous data, we have presented results as summary risk ratio (RR) with 95% confidence intervals (CI).

Continuous data

For continuous data, we used the mean difference (MD) if outcomes were measured in the same way between trials. We planned to use the standardized mean difference to combine trials that measured the same outcome, but used different methods.

Where trials reported median, we included this information in additional tables and a narrative summary of the findings.

Unit of analysis issues

Cluster‐randomized trials

We included cluster‐randomized trials in the analyses along with individually‐randomized trials. Cluster‐randomized trials are labelled with a (C). We had planned to adjust their standard errors and determine effective sample size using the methods described in the Cochrane Handbook for Systematic Reviews of Interventions using an estimate of the intra cluster correlation co‐efficient (ICC) derived from the trial (if possible), from a similar trial or from a study of a similar population to take into account the design effect (Higgins 2011b). If we used ICCs from other sources, we planned to report this and conduct sensitivity analyses to investigate the effect of variation in the ICC. However we identified only two cluster‐randomized trials. In one, family was the unit of allocation, and it was not possible to know if there was more than one pregnant women per family and therefore any clustering effect (Pharoah 1971 (C)). Therefore we did not adjust data from this trial for clustering. In the other cluster‐randomized study, the reported analysis accounted for clustering therefore no additional adjustments were necessary (Bouhouch 2014 (C)).

We synthesized the relevant information from the cluster‐randomized and individually‐randomized trials.

We considered it reasonable to combine the results from both if there was little heterogeneity between the study designs and the interaction between the effect of intervention and the choice of randomisation unit was considered to be unlikely. We also acknowledged heterogeneity in the randomisation unit.

Cross‐over trials

We planned to include cross‐over trials which would be otherwise eligible for inclusion and use only data from the first period (if the data were presented in this way) and anticipated that additional methods for 'Risk of bias' assessment and analysis would not be needed. We did not identify any eligible cross‐over trials though.

Other unit of analysis issues

For studies with more than two intervention groups (multi‐arm studies, e.g. using different doses), we included only directly relevant arms. For studies with more than one relevant arm we combined the arms into a single pair‐wise comparison (Higgins 2011) and included the disaggregated data in the corresponding subgroup category. To avoid double counting participants, if the control group was shared by different study arms, we divided the control group (events and total population) over the number of relevant subgroup categories.

For studies that included non‐pregnant women, we only included those who became pregnant.

If a study examined supplementing both the mother and infant, and was otherwise eligible, we included that study arm and the maternal data but excluded the infant data, as it would have been affected by the infant intervention.

We included all relevant details in the Characteristics of included studies tables.

Dealing with missing data

For included studies, we noted levels of attrition in the Characteristics of included studies tables.

For all outcomes, we carried out analyses, as far as possible, on an intention‐to‐treat basis, that is, we attempted to include all participants randomised to each group in the analyses, and all participants were analysed in the group to which they were allocated, regardless of whether or not they received the allocated intervention. The denominator for each outcome in each trial was the number randomised minus any participants whose outcomes were known to be missing.

Assessment of heterogeneity

We assessed statistical heterogeneity in each meta‐analysis using the Tau², I² (Higgins 2003) and Chi² statistics (Deeks 2011). We considered heterogeneity as substantial if an I² statistic was greater than 30% and either the Tau² was greater than zero, or there was a low P value (less than 0.10) in the Chi² test for heterogeneity. We planned to explore any identified substantial heterogeneity (above 30%) by pre‐specified subgroup analysis.

Assessment of reporting biases

We planned to investigate reporting biases (such as publication bias) using funnel plots if there were 10 or more studies in the meta‐analysis. We planned to assess funnel plot asymmetry visually and if asymmetry was suggested by a visual assessment, we planned to perform exploratory analyses to investigate it. However there were no meta‐analyses with 10 or more studies so we were unable to assess reporting bias.

Data synthesis

We carried out statistical analysis using the RevMan 5 software (RevMan 2014). We intended to use fixed‐effect meta‐analysis for combining data where it would be reasonable to assume that studies were estimating the same underlying treatment effect: that is, where trials were examining the same intervention, and we judged the trials’ populations and methods sufficiently similar. Since we identified both clinical heterogeneity (sufficient to expect that the underlying treatment effects differed between trials) or substantial statistical heterogeneity, or both, we used random‐effects meta‐analysis to produce overall summaries if we considered an average treatment effect across trials to be clinically meaningful. We treated the random‐effects summary as the average of the range of possible treatment effects and we have discussed the clinical implications of treatment effects differing between trials. If the average treatment effect was not clinically meaningful, we did not combine trials.

As we used random‐effects analyses, we have presented the results as the average treatment effect with 95% CIs, and the estimates of Tau² and I².

Subgroup analysis and investigation of heterogeneity

If we identified substantial heterogeneity, we planned to attempt to investigate it using subgroup analyses and sensitivity analyses. We considered whether an overall summary was meaningful and, if it was, used random‐effects analysis to produce it.

We planned to carry out the following subgroup analyses, when information was available, and using only information reported by the trial.

By period of supplementation: only before pregnancy versus only during pregnancy versus only during the postpartum period versus mixed.
By supplementation regimen: daily versus weekly versus annual or single dose.
By access to iodized salt: less than 50% versus 50% to 69% versus greater than or equal to 70% versus not reported.
By iodine status of the participants at the start of the intervention: adequate versus mild or moderate deficiency versus severe deficiency versus unknown (as defined by trial authors, or following WHO‐recommended criteria).
By breastfeeding status: yes at any time versus never versus mixed/not reported. Note: this factor is not related to pregnancy outcomes therefore this analysis was performed only on maternal postpartum and child outcomes.

We planned to include only the primary outcomes in subgroup analyses.

If we found substantial heterogeneity for any primary outcome, we planned to assess subgroup differences by interaction tests available within RevMan 5 (RevMan 2014). We planned to report the results of subgroup analyses quoting the Chi² statistic and P value, and the interaction test I² value.

We decided to add subgroups for a given outcome only where there were at least four studies as we considered anything less to be insufficient to conduct meaningful analysis by the covariate. However no outcome with statistical heterogeneity included data from four or more trials, so we did not perform any subgroup analyses.

Sensitivity analysis

We planned to conduct sensitivity analysis based on the risk of bias of the studies. We considered a study to be of high quality if it was graded as 'low risk of bias' in both randomisation and allocation concealment and in either blinding or loss to follow‐up. However we only considered one trial at low risk of bias therefore did not conduct this sensitivity analysis. We planned to conduct other sensitivity analyses based on the studies included, for example, if we identified and included any cluster‐randomized trials we would carry out sensitivity analysis using a range of ICC values.

We planned to carried out sensitivity analysis for primary outcomes only.

Results

Description of studies

See: Characteristics of included studies; Characteristics of excluded studies; Characteristics of ongoing studies tables.

Results of the search

The search of Cochrane Pregnancy and Childbirth's Trials Register retrieved 43 reports and the additional search strategy identified another 53 reports. See: Figure 1 which shows the process for assessing and selecting the studies (Moher 2009). We have included 14 and excluded 48 trials; five are ongoing or have not yet been published and two are awaiting classification. Eleven trials involving over 2000 women contributed data for the comparisons in this review (three trials did not report any primary or secondary outcomes (Kämpe 1990; Mulrine 2010; Silva 1981).

Figure 1

Study flow diagram

Included studies

Design

Most of the included studies were randomised controlled trials, except for four which were quasi‐randomized (Kevany 1969; Liesenkotter 1996; Pharoah 1971 (C); Silva 1981). Individuals were the unit of allocation for all except for two cluster‐randomized trials which allocated clinic days (Bouhouch 2014 (C)) and families (Pharoah 1971 (C)). The Bouhouch 2014 (C) analysis adjusted for clustering. Seven of the trials were reportedly double‐blinded (Bouhouch 2014 (C); Glinoer 1993; Mulrine 2010; Nohr 2000; Pharoah 1971 (C); Zhou 2015; Gowachirapant 2014). Samples sizes were generally small (eight trials were 120 or less) and ranged from 54 to 671. Two trials which recruited families or communities did not specify how many pregnant women, or women who became pregnant, were initially included (Kevany 1969; Pharoah 1971 (C)). One trial stopped at a sample size of 59 women (planned n = 1098) and was unblinded (Zhou 2015). Three trials were assessed as eligible for inclusion but did not contribute data because they did not report on any prespecified primary or secondary outcomes (Kämpe 1990; Mulrine 2010; Silva 1981).

Settings

Six trials were conducted in Europe, in Belgium (Glinoer 1993), Denmark (Nohr 2000; Pedersen 1993), France (Brucker‐Davis 2013), Germany (Liesenkotter 1996) and Sweden (Kämpe 1990). The other trials took place in Australia (Zhou 2015), Chile (Silva 1981), Morocco (Bouhouch 2014 (C)), New Zealand (Mulrine 2010), Papua New Guinea (Pharoah 1971 (C)), Peru (Kevany 1969), Thailand (Gowachirapant 2014), and Zaire (Thilly 1978). The trials which reported when the research took place did so from 1966 through 2011.

Participants

Most of the trials recruited women during pregnancy through antenatal care contacts at clinics or hospitals (Brucker‐Davis 2013; Glinoer 1993; Kämpe 1990; Liesenkotter 1996; Nohr 2000; Pedersen 1993; Silva 1981; Thilly 1978; Zhou 2015; Gowachirapant 2014). Two trials recruited postpartum women at the infant's first vaccination visit (Bouhouch 2014 (C)) or via advertisements (Mulrine 2010). Two included entire families or communities and as a result, pregnant or non‐pregnant women who later became pregnant were included (Kevany 1969; Pharoah 1971 (C)).

Average age of participants, where reported, ranged from 23 to 32 years. Information on socioeconomic status was generally not reported. Five trials reported on anthropometric nutritional status at baseline, though different indicators were used. Women in the Danish trials were pregnant and average weight ranged from 62 kg to 68 kg (Nohr 2000; Pedersen 1993), body mass index (kg/m2) of postpartum women in the Moroccan trial was 25.6 (Bouhouch 2014 (C)) and of pregnant women in the French trial was 22 (Brucker‐Davis 2013). In the Thai trial, 14% to 17% of pregnant women were overweight or obese (Gowachirapant 2014).

Information on baseline iodine status of the participants or the study area was provided in all but one trial (Kämpe 1990). Where mean UIC was reported, it ranged from 30 µg/L to approximately 130 µg/L and women had insufficient iodine intake, as indicated by median UIC below 150 μg/L for pregnant women or below 100 μg/L for breastfeeding women (Bouhouch 2014 (C); Brucker‐Davis 2013; Glinoer 1993; Liesenkotter 1996; Mulrine 2010; Nohr 2000; Pedersen 1993; Silva 1981; Gowachirapant 2014; Zhou 2015). Three trials only reported goitre prevalence which ranged from 20% to 70% (Kevany 1969; Pharoah 1971 (C); Thilly 1978). One study area was reported as high in endemic cretinism (Pharoah 1971 (C)).

Only four trials reported information on household coverage of iodized salt. Coverage was above 80% in the three trials that used participant‐reported information (Brucker‐Davis 2013; Mulrine 2010; Gowachirapant 2014). Bouhouch 2014 (C) collected and analysed household salt samples in a sub‐sample and found that 57% contained some iodine (6% with "adequate" or at least 15 ppm iodine and 51% with iodine less than 15 ppm).

Two trials screened women for thyroid peroxidase antibodies (TPO‐ab) and recruited only women with elevated levels (Kämpe 1990; Nohr 2000).

Interventions

Seven trials included iodine supplementation during pregnancy only (Glinoer 1993; Nohr 2000; Pharoah 1971 (C); Silva 1981; Thilly 1978; Zhou 2015; Gowachirapant 2014), three included iodine supplementation only during the postpartum period (Bouhouch 2014 (C); Kämpe 1990; Mulrine 2010) and four gave iodine to the same women in both pregnancy and the postpartum period (Brucker‐Davis 2013; Liesenkotter 1996; Nohr 2000; Pedersen 1993). Two trials provided iodine supplementation to women who later became pregnant (Kevany 1969; Pharoah 1971 (C)).

Most of the included trials used daily oral iodine supplementation, mainly in the form of tablets (Brucker‐Davis 2013; Liesenkotter 1996; Mulrine 2010; Nohr 2000; Zhou 2015; Gowachirapant 2014) though one used capsules (Kämpe 1990), two used drops (Pedersen 1993; Silva 1981) and one trial reported using a daily supplement but did not specify the form (Glinoer 1993). The daily iodine dose ranged from 75 µg to 300 µg and the duration ranged from approximately one to 17 months. Where the iodine compound was reported, it was mostly potassium iodide (Glinoer 1993; Kämpe 1990; Liesenkotter 1996; Pedersen 1993; Silva 1981; Gowachirapant 2014) though one trial used potassium iodate (Mulrine 2010). One trial used a single high dose oral iodine supplement, in the form of two soft gel capsules containing iodized poppyseed oil with 400 mg iodine (Bouhouch 2014 (C)).

The three earliest trials, which took place in the 1960s and 1970s, used single high‐dose intramuscular injections of iodized oil (Kevany 1969; Pharoah 1971 (C); Thilly 1978). One trial repeated this intervention in the same communities three years later (Kevany 1969). The iodine dose in these trials ranged from 475 mg to 1600 mg, though in one study adult women with nodular goitre were given a lower dose of 95 mg (Kevany 1969). Thilly 1978 specified that the iodine was slowly resorbable.

Most of the included trials used iodine‐only supplementation, except for three which provided the same additional vitamins and minerals to the intervention and control groups, either as part of the supplement containing iodine (Brucker‐Davis 2013; Nohr 2000) or as a separate supplement (Gowachirapant 2014).

Comparisons

Most studies had two trial arms comparing an iodine intervention with a no‐iodine comparison in the same population group though two trials also included an additional intervention arm (one included two different iodine doses (Mulrine 2010) and one included iodine in pregnancy only and in pregnancy and postpartum (Nohr 2000)). Two trials also had a third arm with thyroxine (or thyroxine and iodine) supplementation (Glinoer 1993; Kämpe 1990) however these arms were not eligible and therefore were not included in this review.

Seven trials reported use of a placebo comparison group (Bouhouch 2014 (C); Glinoer 1993; Kevany 1969; Mulrine 2010; Pharoah 1971 (C); Zhou 2015; Gowachirapant 2014), four used no intervention (Kämpe 1990; Liesenkotter 1996; Pedersen 1993; Silva 1981), and two used other micronutrients (Brucker‐Davis 2013; Nohr 2000). For one study, we found conflicting information about the control supplement composition as some reports stated it was a placebo and some stated it included vitamins (Thilly 1978). For the purposes of this review, we have considered the control for this trial as placebo.

Comparison 1

Eleven studies compared the effects of any supplement containing iodine versus same supplement without iodine or no treatment/placebo (Bouhouch 2014 (C); Brucker‐Davis 2013; Glinoer 1993; Kevany 1969; Liesenkotter 1996; Nohr 2000; Pedersen 1993; Pharoah 1971 (C); Thilly 1978; Zhou 2015; Gowachirapant 2014). Three trials were otherwise eligible but did not contribute any data for this comparison (Kämpe 1990; Mulrine 2010; Silva 1981).

Comparison 2

Eight trials compared the effects of any oral iodine supplement versus the same supplement without iodine or no treatment/placebo (Bouhouch 2014 (C); Brucker‐Davis 2013; Glinoer 1993; Liesenkotter 1996; Nohr 2000; Pedersen 1993; Zhou 2015; Gowachirapant 2014). Three trials were otherwise eligible but did not contribute any data for this comparison (Kämpe 1990; Mulrine 2010; Silva 1981).

Comparison 3

Five trials compared an oral iodine‐only supplement versus no intervention or placebo (Bouhouch 2014 (C); Glinoer 1993; Liesenkotter 1996; Pedersen 1993; Zhou 2015). Three trials were otherwise eligible but did not contribute any data for this comparison (Kämpe 1990; Mulrine 2010; Silva 1981).

Comparison 4

Three trials compared the effects of oral iodine supplements with other vitamins and/or minerals versus only other vitamins and/or minerals (exact same formulation of other vitamins/minerals, but no iodine) (Brucker‐Davis 2013; Nohr 2000; Gowachirapant 2014).

Comparison 5

Three trials compared the effects of any injected iodine supplement versus same supplement without iodine or no treatment/placebo (Kevany 1969; Pharoah 1971 (C); Thilly 1978).

Comparison 6

Three trials compared the effects of any injected iodine‐only supplement versus no intervention or placebo (Kevany 1969; Pharoah 1971 (C); Thilly 1978).

Comparison 7

No studies compared an injected iodine supplement with other vitamins and/or minerals versus only other vitamins and/or minerals (exact same formulation of other vitamins/minerals, but no iodine). Because of this, the trials (and data) are the same for comparisons 5 and 6, therefore we have only reported comparison 5.

See Characteristics of included studies for a detailed description of the trials.

Excluded studies

We excluded 48 trials. The main reason for excluding studies was that they were not randomised or quasi‐randomized trials. Some randomised trials, or trials where it was unclear whether randomisation was used, were excluded because they compared different iodine doses (without a placebo or a no treatment control) or assessed iodized salt (Antonangeli 2002; Mohammed 2015; Romano 1991; Santiago 2013; Troshina 2010).

See the Characteristics of excluded studies tables for detailed descriptions of the excluded studies and the reasons for exclusion.

Risk of bias in included studies

Allocation

Random sequence generation

We considered five trials at low risk of selection bias for random sequence generation (Bouhouch 2014 (C); Brucker‐Davis 2013; Thilly 1978; Zhou 2015; Gowachirapant 2014). Four we deemed high risk because they used quasi‐randomization (Kevany 1969; Liesenkotter 1996; Pharoah 1971 (C); Silva 1981). The remaining trials were reportedly randomised but did not provide adequate information on how random sequences were generated, therefore we considered risk of bias in this domain unclear.

Allocation concealment

We considered three trials at low risk of selection bias for allocation concealment because clinic days were the unit of randomisation or were used to assign individuals (and study staff could not have reasonably influenced the day women showed up for clinic) (Bouhouch 2014 (C); Liesenkotter 1996) or because randomisation was performed centrally (Zhou 2015). We considered three trials at high risk because they used alternation to assign individuals, or families, to groups (Kevany 1969; Pharoah 1971 (C); Silva 1981). The remaining trials did not adequately report the methods used to conceal allocation so we considered the risk to be unclear.

Blinding

Blinding of participants and staff

We considered four trials at low risk of performance bias because a placebo was used for the control group and staff and participants were reportedly blinded to group assignment (Bouhouch 2014 (C); Mulrine 2010; Thilly 1978; Gowachirapant 2014). We determined six trials high risk because no blinding was reported and the control group did not receive any intervention (Kämpe 1990; Liesenkotter 1996; Pedersen 1993; Silva 1981) or because it was described as an "open study" (Brucker‐Davis 2013) or unblinded during the study due to ethical consideration to placebo group (Zhou 2015). It was not possible to assess risk of bias in this domain for the other four trials due to lack of information reported (Glinoer 1993; Kevany 1969; Nohr 2000; Pharoah 1971 (C)).

Blinding of outcome assessors

Risk of detection bias was mostly unclear because study authors did not adequately report this information, with the exception of two trials which we considered low risk because they specified that outcome assessors were blinded to group assignment (Pharoah 1971 (C);Gowachirapant 2014) and two trials which were considered high risk because they were described as an "open study" (Brucker‐Davis 2013) or because the trial was unblinded mid‐way through (Zhou 2015).

Incomplete outcome data

Two trials were at low risk of attrition bias because of low (less than 20%) and balanced incomplete outcome data (Nohr 2000, Zhou 2015). We determined seven trials to be a high risk because of high or imbalanced incomplete outcome data, or both (Bouhouch 2014 (C); Brucker‐Davis 2013; Mulrine 2010; Pharoah 1971 (C); Silva 1981; Thilly 1978; Gowachirapant 2014). For the remaining studies, lack of detailed information on attrition precluded judgement (Glinoer 1993; Kämpe 1990; Kevany 1969; Liesenkotter 1996; Pedersen 1993).

Selective reporting

For most of the included trials, we did not have access to the study protocol or clinical trial registration which stated the pre‐specified outcomes, therefore the risk of reporting bias due to selective reporting was unclear. In two cases we did have access to the trial registry and were able to determine the trial to be low (Zhou 2015) or high (Bouhouch 2014 (C)) risk of reporting bias, based on the information provided in the registries and the publications. In another case, we found the risk of bias to be high because baseline information was only provided for those who completed the study (Brucker‐Davis 2013).

Other potential sources of bias

Most of the included trials appeared to be free of other potential sources of bias, though this could not be confirmed, therefore we considered this risk of bias unclear. We did however consider the risk to be high in two trials which provided iodine to women before they became pregnant, potentially affecting their probability of becoming pregnant and therefore leading to differences between groups and bias (Kevany 1969; Pharoah 1971 (C)).

We have included full details of 'Risk of bias’ assessments in the Characteristics of included studies tables. We have also included figures which summarize our 'Risk of bias’ assessments (Figure 2; Figure 3), which we used to judge study quality in the 'Summary of findings' tables (see summary of findings Table for the main comparison; summary of findings Table 2).

Figure 2

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies

Figure 3

Risk of bias summary: review authors' judgements about each risk of bias item for each included study

Effects of interventions

In this review we included 14 trials involving 3154 women, though one trial did not report the number of women randomised (Pharoah 1971 (C)). Three trials involving 417 women (Kämpe 1990; Mulrine 2010; Silva 1981) did not contribute data because they did not report any primary or secondary outcomes. We have organized the summary results by comparison, by primary and secondary outcomes and by maternal and child outcomes.

See Data and analyses for detailed results on primary and secondary outcomes. We have set out below five out of the seven prespecified comparisons. None of the included trials included an assessment of comparison 7, which meant that the originally‐planned comparisons 5 and 6 were the exact same and therefore we have not reported the results from comparison 6 separately.

For each of the comparisons and outcomes with data, we have indicated the number of studies, the number of studies contributing data and the total number of women or children included in that analysis (instead of the number randomised in case of attrition).

(1) Any supplement containing iodine versus same supplement without iodine or no treatment/placebo (11 trials, 2737 participants)

Eleven trials contributed data involving 2737 women (Bouhouch 2014 (C); Brucker‐Davis 2013; Glinoer 1993; Kevany 1969; Liesenkotter 1996; Nohr 2000; Pedersen 1993; Pharoah 1971 (C); Thilly 1978; Zhou 2015; Gowachirapant 2014). None were considered high quality. Three other trials did not contribute any data to our analyses (Kämpe 1990; Mulrine 2010; Silva 1981).

Maternal primary outcomes

Hypothyroidism ‐ pregnancy (as defined by trial authors)

Data from one trial (Gowachirapant 2014) involving 365 women showed no difference in risk of hypothyroidism in pregnancy between women who received and did not receive iodine supplementation (risk ratio (average RR) 1.90; 95% confidence interval (CI) 0.57 to 6.38, low‐quality evidence) (Analysis 1.1).

Hypothyroidism ‐ postpartum (as defined by trial authors)

Among 540 women in three trials (Bouhouch 2014 (C); Liesenkotter 1996; Gowachirapant 2014), those who received iodine supplementation were as likely to be diagnosed with hypothyroidism in the postpartum period as those who did not receive any iodine (average RR 0.44; 95% CI 0.06 to 3.42, low‐quality evidence) (Analysis 1.2). We did not find evidence of statistical heterogeneity (I² = 0%, Tau² = 0.00 and Chi² test for heterogeneity P = 0.41).

Preterm birth (as defined by trial authors)

Two trials (Zhou 2015; Gowachirapant 2014) involving 376 pregnant women reported on preterm birth, and showed no difference between groups (average RR 0.71; 95% CI 0.30 to 1.66, low‐quality evidence) (Analysis 1.3). We identified substantial heterogeneity for this outcome though (I² = 32%, Tau² = 0.13 and Chi² test for heterogeneity P = 0.23).

Maternal adverse effect: elevated thyroid peroxidase antibodies (TPO‐ab) ‐ pregnancy (as defined by trial authors)

One trial with 359 women reported on elevated TPO‐ab in pregnancy (Gowachirapant 2014) and found no difference in risk between the iodine and control groups (average RR 0.95; 95% CI 0.44 to 2.07, low‐quality evidence) (Analysis 1.4).

Maternal adverse effect: elevated TPO‐ab ‐ postpartum (as defined by trial authors)

This outcome was reported in three trials with 397 women (Liesenkotter 1996; Pedersen 1993; Gowachirapant 2014). Findings show no difference in the likelihood between the group that received iodine and the group that did not (average RR 1.01; 95% CI 0.78 to 1.30, low‐quality evidence) (Analysis 1.5). We did not find any evidence of statistical heterogeneity for this analysis (I² = 0%, Tau² = 0.00 and Chi² test for heterogeneity P = 0.63).

Maternal adverse effect: hyperthyroidism ‐ pregnancy (as defined by trial authors)

One trial with 365 women reported on this outcome (Gowachirapant 2014) and found no difference between groups (average RR 1.90; 95% CI 0.57 to 6.38, low‐quality evidence) (Analysis 1.6).

Maternal adverse effect: hyperthyroidism ‐ postpartum (as defined by trial authors)

Data from three studies involving 543 women (Bouhouch 2014 (C); Liesenkotter 1996; Gowachirapant 2014) showed that 1.6% of those who received iodine supplementation and 4.9% of those who did not had hyperthyroidism in the postpartum period (average RR 0.32; 95% CI 0.11 to 0.91, low‐quality evidence) (Analysis 1.7). There was no statistical heterogeneity (I² = 0%, Tau² = 0.00 and Chi² test for heterogeneity P = 0.67).

Maternal adverse effect: digestive intolerance ‐ pregnancy (as defined by trial authors)

One trial with 76 women reported on this outcome (Brucker‐Davis 2013) and found a higher likelihood of digestive intolerance in the iodine group whereby 33.3% of women who received iodine and 2.2% of those who did not dropped out of the study due to nausea or vomiting (average RR 15.33; 95% CI 2.70 to 113.7, very low‐quality evidence) (Analysis 1.8).

Infant/child primary outcomes

Perinatal mortality (including stillbirth/fetal death and neonatal death, as defined by trial authors)

Two trials with 457 assessments (Thilly 1978; Zhou 2015) reported on this outcome, though all of the deaths occurred in one trial (Thilly 1978). In this trial there was lower perinatal mortality in the group that received iodine supplementation compared to the control group, however the difference did not reach statistical significance (average RR 0.66; 95% CI 0.42 to 1.03, low‐quality evidence) (Analysis 1.9).

Low birthweight (less than 2500 g)

Two trials with 377 infants reported on this outcome (Zhou 2015; Gowachirapant 2014) and showed no difference between groups (average RR 0.56; 95% CI 0.26 to 1.23, low‐quality evidence) (Analysis 1.10). There was no evidence of statistical heterogeneity (I² = 0%, Tau² = 0.00 and Chi² test for heterogeneity P = 0.62).

Neonatal hypothyroidism or elevated thyroid‐stimulating hormone (TSH) (as defined by trial authors)

This outcome was reported in two trials (Zhou 2015; Gowachirapant 2014) with 260 infant assessments, although all of the cases occurred in one trial (Gowachirapant 2014). The average RR for neonatal elevated TSH in this trial was 0.58 (95% CI 0.11 to 3.12, very low‐quality evidence) (Analysis 1.11).

Neonatal adverse effect ‐ elevated TPO‐ab (as defined by trial authors)

One study with 108 assessments (Liesenkotter 1996) reported on this outcome and showed no difference between infants whose mothers received iodine and those who did not (average RR 0.61; 95% CI 0.07 to 5.70, very low‐quality evidence) (Analysis 1.12). In this trial, all of the neonates with elevated TPO‐ab were born to women with elevated TPO‐ab at baseline.

Neonatal adverse effect ‐ hyperthyroidism (as defined by trial authors)

No trials reported this outcome.

Maternal secondary outcomes

Spontaneous miscarriage (as defined by trial authors)

Three trials (Brucker‐Davis 2013; Zhou 2015; Gowachirapant 2014) with information on 645 pregnancies showed a similar likelihood of spontaneous miscarriage (average RR 1.31; 95% CI 0.64 to 2.69) and no evidence of statistical heterogeneity (I² = 0%, Tau² = 0.00 and Chi² test for heterogeneity P = 0.64) (Analysis 1.13).

Maternal goitre ‐ pregnancy (assessed by any method)

Two studies (Glinoer 1993; Gowachirapant 2014) with 486 women showed no difference in likelihood of goitre in pregnancy (average RR 1.00; 95% CI 0.33 to 3.06) however we found significant heterogeneity for this outcome (I² = 53%, Tau² = 0.35 and Chi² test for heterogeneity P = 0.14) (Analysis 1.14).

Maternal goitre ‐ postpartum (assessed by any method)

Postpartum goitre was reported in two trials (Liesenkotter 1996; Gowachirapant 2014). Data from 370 women showed no evidence of difference (average RR 1.17; 95% CI 0.64 to 2.16) though we did identify substantial heterogeneity (I² = 33%, Tau² = 0.10 and Chi² test for heterogeneity P = 0.22) (Analysis 1.15).

Maternal thyroid volume ‐ pregnancy (in mL) (assessed by any method)

This outcome was reported in three trials with 496 participants, all of which reported median thyroid volume so no meta‐analysis was performed (Brucker‐Davis 2013; Pedersen 1993; Gowachirapant 2014). In two trials, findings showed lower median thyroid volume in the iodine compared to the no‐iodine groups (Brucker‐Davis 2013; Pedersen 1993) (Table 2). Gowachirapant 2014, however, found similar thyroid volumes in both groups.

Table 2. Maternal thyroid volume ‐ pregnancy

Study	Thyroid volume (mL)
Study	Iodine	No iodine
Brucker‐Davis 2013¹	Median: 9.73, range: 5.4–15.5 (n = 30)	Median: 10.13, range: 4.2–25.5 (n = 46)
Pedersen 1993²	Median: 11.1 (n = 28)	Median: 12.4 (n = 26)
Gowachirapant 2014³	Median: 8.0, IQR: 7.0, 9.1 (n = 176)	Median: 8.1, IQR: 7.1, 9.4 (n = 190)

¹Assessed in the 3rd trimester.
²Assessed at 37 weeks' gestation. Median estimated from figure. No estimate of range or distribution was provided. Total was not provided so was assumed.
³Assessed at 30 weeks' gestation. Intention‐to‐treat analysis.

Maternal thyroid volume ‐ postpartum (in mL) (assessed by any method)

Four trials with 457 women reported on this outcome (Brucker‐Davis 2013; Liesenkotter 1996; Pedersen 1993; Gowachirapant 2014) and most used median thyroid volume, so we did not perform a meta‐analysis. Three of the studies showed similar postpartum thyroid volumes between the iodine and no iodine groups (Liesenkotter 1996; Pedersen 1993; Gowachirapant 2014). (Table 3). One trial, however, reported a lower median thyroid volume (by 1.2 mL) in the iodine group (Brucker‐Davis 2013).

Table 3. Maternal thyroid volume ‐ postpartum

Study	Thyroid volume (mL)
Study	Iodine	No iodine
Brucker‐Davis 2013¹	Median: 8.5, range: 3.8–13.8 (n = 18)	Median: 9.7, range: 6.6–18.9 (n = 18)
Liesenkotter 1996²	Mean: 20.5, SD: 8.5 (n = 38)	Mean: 20.0, SD: 10.5 (n = 70)
Pedersen 1993³	Median: 10.2 (n = 25)	Median: 10 (n = 24)
Gowachirapant 2014⁴	Median: 7.3, IQR: 5.8, 9.0 (n = 129)	Median: 7.2, IQR: 6.1, 8.4 (n = 135)

¹Assessed at 3 months postpartum.
²Assessed by ultrasound on average at 11 days postpartum. Mean and SD extracted from figure.
³Assessed at 12 months postpartum. Median estimated from figure. No estimate of range or distribution was provided. Total was not provided so was assumed.
⁴Assessed at 6 weeks postpartum.

Maternal thyroglobulin (Tg) ‐ pregnancy (in μg/L)

This outcome was reported in four studies with 562 women (Brucker‐Davis 2013; Nohr 2000; Pedersen 1993; Gowachirapant 2014) all of which reported median Tg, so we did not perform a meta‐analysis. In all four trials, median Tg in pregnancy was lower in the group which received iodine compared to the no‐iodine group; the difference ranged from 2.1 µg/L to 7.5 µg/L (Table 4).

Table 4. Maternal thyroglobulin ‐ pregnancy

Study	Thyroglobulin (in µg/L)
Study	Iodine	No iodine
Brucker‐Davis 2013¹	Median: 17.5, range: 2.5–56.2 (n = 30)	Median: 19.6, range: 3.8–98.6 (n = 46)
Nohr 2000²	Median: 14.1, IQR: 5.0, 21.5 (n = 42)	Median: 19.4, IQR: 8.2, 33.5 (n = 24)
Pedersen 1993³	Median: 9.2 (n = 28)	Median: 16.7 (n = 26)
Gowachirapant 2014⁴	Median: 9.78, IQR: 5.46, 18.10 (n = 176)	Median 11.90, IQR: 6.35, 22.30 (n = 190)

¹Assessed in the 3rd trimester.
²Assessed at 35 weeks' gestation.
³Assessed at 37 weeks' gestation. Median estimated from figure. No estimate of range or distribution was provided. Total was not provided so was assumed.
⁴Assessed at 30 weeks' gestation.

Maternal Tg ‐ postpartum (in μg/L)

Four trials involving 457 women (Brucker‐Davis 2013; Liesenkotter 1996; Pedersen 1993; Gowachirapant 2014) reported this outcome, mostly as median Tg, therefore we did not perform a meta‐analysis. All four trials showed that women who received iodine had lower postpartum Tg levels compared to those who did not (Table 5). The difference between groups ranged from 1.57 µg/L to 11.6 µg/L.

Table 5. Maternal thyroglobulin ‐ postpartum

Study	Thyroglobulin (in µg/L)
Study	Iodine	No iodine
Brucker‐Davis 2013¹	Median: 11.5, range: 4.4–37 (n = 18)	Median: 23.1, range: 3.0–45.5 (n = 18)
Liesenkotter 1996²	Mean: 8.3, SD: 10.9 (n = 38)	Mean: 13.5, SD: 19.3 (n = 70)
Pedersen 1993³	Median: 4.7 (n = 24)	Median: 10.3 (n = 24)
Gowachirapant 2014⁴	Median: 7.94, IQR: 4.89, 15.30 (n = 129)	Median: 9.51, IQR: 4.79, 15.90 (n = 135)

¹Assessed at 3 months postpartum.
²Assessed on average at 11 days postpartum.
³Assessed at 12 months postpartum. Median estimated from figure. No estimate of range or distribution was provided. Total was not provided so was assumed.
⁴Assessed at 6 weeks postpartum.

Insufficient iodine intake ‐ pregnancy (median urinary iodine concentration (UIC) less than 150 μg/L)

Two studies with 432 participants reported on this outcome (Brucker‐Davis 2013; Gowachirapant 2014). Results show that women who received iodine supplements had a lower likelihood of insufficient iodine intake in pregnancy compared to those who did not receive any iodine. In the iodine group, 33.3% of women had inadequate iodine intake in pregnancy compared to 53.2% in the control group (average RR 0.64; 95% CI 0.51 to 0.80) (Analysis 1.16). There was no evidence of statistical heterogeneity for this outcome (I² = 0%, Tau² = 0.00 and Chi² test for heterogeneity P = 0.56).

Insufficient iodine intake ‐ postpartum (breastfeeding: median UIC less than 100 μg/L)

Results from two trials (Bouhouch 2014 (C); Gowachirapant 2014) involving 425 women showed that iodine supplementation reduced the likelihood of having insufficient iodine intake in the postpartum period (average RR 0.81; 95% CI 0.70 to 0.93), with no evidence of statistical heterogeneity (I² = 0%, Tau² = 0.00 and Chi² test for heterogeneity P = 0.90) (Analysis 1.17).

Excessive iodine intake (pregnancy only: median UIC greater than or equal to 500 μg/L)

One trial (Gowachirapant 2014) with 356 women reported that 7.0% of women who received iodine supplements and 1.6% of women who did not had excessive iodine intake during pregnancy (average RR 4.33; 95% CI 1.24 to 15.07) (Analysis 1.18).

Infant/child secondary outcomes

Small‐for‐gestational age (as defined by trial authors)

This outcome was reported in two trials (Zhou 2015; Gowachirapant 2014) with 377 infants that showed no evidence of difference between groups (average RR 1.26; 95% CI 0.77 to 2.05) (Analysis 1.19). There was no substantial heterogeneity (I² = 0%, Tau² = 0.00 and Chi² test for heterogeneity P = 0.83).

Congenital anomalies (including cretinism, as defined by trial authors)

Two trials (Pharoah 1971 (C); Zhou 2015), which assessed 875 children, showed a lower frequency of congenital anomalies in the iodine group compared to the control group, 1.7% versus 6.4%, respectively, though all of the cases came from one trial (Pharoah 1971 (C)) (average RR 0.27; 95% CI 0.12 to 0.60) (Analysis 1.20). In six out of seven cases in the iodine group and five out of 26 cases in the control group, the mothers received the treatment during pregnancy. The other mothers received treatment before pregnancy.

Growth (anthropometric Z scores)

No trials reported on this outcome.

Neonatal goitre (assessed by any method)

This outcome was reported in three trials (Glinoer 1993; Kevany 1969; Liesenkotter 1996) that assessed 684 neonates. Findings showed a lower likelihood of enlarged thyroid gland in the neonates in the iodine group compared to the control group (average RR 0.11; 95% CI 0.02 to 0.56) (Analysis 1.21) with no evidence of statistical heterogeneity (I² = 0%, Tau² = 0.00 and Chi² test for heterogeneity P = 0.76).

Neonatal thyroid volume (in mL) (assessed by any method)

Three trials with assessments of 359 neonates (Glinoer 1993; Liesenkotter 1996; Gowachirapant 2014) showed lower thyroid volume in the iodine compared to the control group. The MD was ‐0.34 mL (95% CI ‐0.58 to ‐0.11) though there was evidence of statistical heterogeneity (I² = 93%, Tau² = 0.04 and Chi² test for heterogeneity P < 0.00001) (Analysis 1.22).

Insufficient neonatal iodine intake (median UIC less than 100 μg/L)

One trial (Gowachirapant 2014), which assessed 159 neonates, found a higher likelihood of insufficient iodine intake in the iodine compared to the control group (average RR 2.14; 95% CI 1.04 to 4.37) (Analysis 1.23).

Neonatal mental or motor development (as defined by trial authors)

One study (Gowachirapant 2014) reported scores from the Neonatal Behavioral Assessment Scales in median and IQR for 149 neonates (Table 6). Out of the eight sub‐scales, three showed higher median scores in the iodine group compared to the control group, one showed a higher score in the control group, and three showed equal scores in both groups.

Table 6. Neonatal behavioural assessment scales scores assessed at 6 weeks postpartum (Zimmerman 2009)

Outcome	Iodine			Control
Outcome	Median	IQR	n	Median	IQR	n
Habituation	7.0	(5.4, 8.3)	22	7.6	(6.8, 8.1)	18
Social‐interactive	7.4	(6.4, 8.0)	66	6.9	(6.0, 8.1)	68
Motor system	6.0	(5.3, 6.6)	73	5.8	(5.2, 6.4)	72
State organization	3.8	(2.8, 4.0)	48	3.8	(3.0, 4.1)	53
State regulation	4.0	(3.3, 4.5)	38	4.0	(2.6, 4.7)	40
Autonomic system	6.7	(6.0, 7.3)	61	6.7	(5.3, 7.3)	71
Reflexes	3.0	(1.8, 4.0)	74	3.0	(2.0, 4.0)	75
Supplementary items	6.7	(5.3, 8.3)	48	6.3	(5.1, 8.1)	55

IQR: interquartile range

Child mental or motor development (as defined by trial authors)

Data from two trials (Kevany 1969; Thilly 1978), which measured 174 children, showed a higher IQ score in the group whose mothers received iodine compared to those who did not. The MD was 11.21 points (95% CI 7.96 to 14.46) (Analysis 1.24) with no statistical heterogeneity (I² = 0%, Tau² = 0.00; Chi² test for heterogeneity P = 0.70). These findings were combined using MD because the same instrument, the Brunet‐Lézine scale, was used in both trials.

One trial (Zhou 2015) reported on child cognitive, language development, motor, social‐emotional and adaptive behaviour scores for 53 children at 18 months of age assessed by the Bayley scales of infant and toddler development (Bayley‐Ⅲ). Scores were similar in both groups (Analysis 1.25; Analysis 1.26; Analysis 1.27; Analysis 1.28; Analysis 1.29). The same trial also reported moderate and severe developmental delays based on the cognitive, language and motor scores and found no significant differences between groups (Analysis 1.30; Analysis 1.31; Analysis 1.32; Analysis 1.33; Analysis 1.34; Analysis 1.35).

Infant death (death in the first year of life)

No trials reported on this outcome.

Other outcomes

We also extracted data on urinary and breast milk iodine concentration following supplementation (Table 1). Average UIC was generally higher in the intervention than the control groups, however this was not the case for all studies and in some cases depended on the timing of assessment (e.g. higher in the iodine group during pregnancy but not postpartum Brucker‐Davis 2013 and Pedersen 1993) and the group assessed (e.g. higher among intervention women but lower among intervention neonates, compared to controls (Gowachirapant 2014)). With the exception of one trial (Zhou 2015), in each study where breast milk iodine concentration was assessed, levels were higher in the iodine compared to the control groups (Bouhouch 2014 (C); Glinoer 1993; Pedersen 1993).

(2) Any oral iodine supplement versus same supplement without iodine or no treatment/placebo (eight trials, 1274 participants)

Eight trials involving 1274 women were included in this comparison (Bouhouch 2014 (C); Brucker‐Davis 2013; Glinoer 1993; Liesenkotter 1996; Nohr 2000; Pedersen 1993; Zhou 2015; Gowachirapant 2014). None were considered high quality. Three other trials did not contribute any data to our analyses (Kämpe 1990; Mulrine 2010; Silva 1981).

Maternal primary outcomes

Hypothyroidism ‐ pregnancy (as defined by trial authors)

Data from one trial (Gowachirapant 2014) involving 365 women showed no difference in risk of hypothyroidism in pregnancy between women who received and did not receive iodine supplementation (average RR 1.90; 95% CI 0.57 to 6.38) (Analysis 2.1).

Hypothyroidism ‐ postpartum (as defined by trial authors)

Among 540 women in three trials (Bouhouch 2014 (C); Liesenkotter 1996; Gowachirapant 2014), those who received oral iodine supplementation were as likely to be diagnosed with hypothyroidism in the postpartum period compared to those who did not receive any iodine (average RR 0.44; 95% CI 0.06 to 3.42) (Analysis 2.2). We did not find evidence of statistical heterogeneity (I² = 0%, Tau² = 0.00 and Chi² test for heterogeneity P = 0.41).

Preterm birth (as defined by trial authors)

Two trials (Zhou 2015; Gowachirapant 2014) involving 376 women reported on preterm birth, and showed no difference between groups (average RR 0.71; 95% CI 0.30 to 1.66) though there was statistical heterogeneity ((I² = 32%, Tau² = 0.13 and Chi² test for heterogeneity P = 0.23) (Analysis 2.3).

Maternal adverse effect: elevated TPO‐ab ‐ pregnancy (as defined by trial authors)

One trial with 359 women reported on elevated thyroid peroxidase antibodies in pregnancy (Gowachirapant 2014) and found no difference in risk between the oral iodine and control groups (average RR 0.95; 95% CI 0.44 to 2.07) (Analysis 2.4).

Maternal adverse effect: elevated TPO‐ab ‐ postpartum (as defined by trial authors)

This outcome was reported in three trials with 397 women (Liesenkotter 1996; Pedersen 1993; Gowachirapant 2014). Findings show no difference in the likelihood between the group which received oral iodine and the group which did not receive any iodine (average RR 1.01; 95% CI 0.78 to 1.30) (Analysis 2.5). We did not find any evidence of statistical heterogeneity for this analysis (I² = 0%, Tau² = 0.00 and Chi² test for heterogeneity P = 0.63).

Maternal adverse effect: hyperthyroidism ‐ pregnancy (as defined by trial authors)

One trial with 365 women reported on this outcome (Gowachirapant 2014) and found no difference between groups (average RR 1.90; 95% CI 0.57 to 6.38) (Analysis 2.6).

Maternal adverse effect: hyperthyroidism ‐ postpartum (as defined by trial authors)

Data from three studies involving 543 women (Bouhouch 2014 (C); Liesenkotter 1996; Gowachirapant 2014) showed that 1.6% of those who received oral iodine supplementation and 4.9% of those who did not had hyperthyroidism in the postpartum period (average RR 0.32, 95% CI 0.11 to 0.91) (Analysis 2.7). There was no statistical heterogeneity (I² = 0%, Tau² = 0.00 and Chi² test for heterogeneity P = 0.67).

Maternal adverse effect: digestive intolerance ‐ pregnancy (as defined by trial authors)

One trial with 76 women reported on this outcome (Brucker‐Davis 2013) and found a higher likelihood of digestive intolerance in the group which received oral iodine supplementation compared to those who did not. One‐third of women (33.3%) in the iodine group and 2.2% of women in the control group dropped out of the study due to nausea or vomiting (average RR 15.33; 95% CI 2.70 to 113.70) (Analysis 2.8).

Infant/child primary outcomes

Perinatal mortality (including stillbirth/fetal death and neonatal death, as defined by trial authors)

One trial with 58 assessments (Zhou 2015) reported no events in either group (Analysis 2.9).

Low birthweight (less than 2500 g)

Two trials with 377 infants reported on this outcome (Gowachirapant 2014) and showed no difference between groups (average RR 0.56; 95% CI 0.26 to 1.23) (Analysis 2.10).

Neonatal hypothyroidism or elevated TSH (as defined by trial authors)

This outcome was reported in two trials (Zhou 2015; Gowachirapant 2014) with 260 infant assessments. All of the cases though occurred in one trial (Gowachirapant 2014); the average RR for neonatal elevated TSH in this trial was 0.58 (95% CI 0.11 to 3.12) (Analysis 2.11).

Neonatal adverse effect ‐ elevated TPO‐ab (as defined by trial authors)

One study with 108 assessments (Liesenkotter 1996) reported on this outcome and showed no difference between infants whose mothers received oral iodine and those who did not (average RR 0.61; 95% CI 0.07 to 5.70) (Analysis 2.12). In this trial, all of the neonates with elevated TPO‐ab were born to women with elevated TPO‐ab at baseline.

Neonatal adverse effect ‐ hyperthyroidism (as defined by trial authors)

No trials reported this outcome.

Maternal secondary outcomes

Spontaneous miscarriage (as defined by trial authors)

Maternal goitre ‐ pregnancy (assessed by any method)

Maternal goitre ‐ postpartum (assessed by any method)

Postpartum enlarged thyroid gland or goitre was reported in two trials (Liesenkotter 1996; Gowachirapant 2014). Data from 370 women showed no evidence of difference (average RR 1.17; 95% CI 0.64 to 2.16) though we did identify substantial heterogeneity (I² = 33%, Tau² = 0.10 and Chi² test for heterogeneity P = 0.22) (Analysis 2.15).

Maternal thyroid volume ‐ pregnancy (in mL) (assessed by any method)

This outcome was reported in three trials with 496 participants, all of which reported median thyroid volume so no meta‐analyses was performed (Brucker‐Davis 2013; Pedersen 1993; Gowachirapant 2014). In two trials, findings showed lower median thyroid volume in the oral iodine compared to the no iodine groups (Brucker‐Davis 2013; Pedersen 1993) (Table 2). Zimmermann and colleagues however found similar thyroid volumes in both groups (Gowachirapant 2014).

Maternal thyroid volume ‐ postpartum (in mL) (assessed by any method)

Four trials with 457 women reported on this outcome (Brucker‐Davis 2013; Liesenkotter 1996; Pedersen 1993; Gowachirapant 2014) and most used median thyroid volume so we did not perform a meta‐analysis. Three of the studies showed similar postpartum thyroid volumes between the oral iodine and no iodine groups (Liesenkotter 1996; Pedersen 1993; Gowachirapant 2014). (Table 3). One trial however reported a lower median thyroid volume (by 1.2 mL) in the iodine group (Brucker‐Davis 2013).

Maternal Tg ‐ pregnancy (in μg/L)

This outcome was reported in four studies with 562 women (Brucker‐Davis 2013; Nohr 2000; Pedersen 1993; Gowachirapant 2014) all of which reported median Tg so we did not perform a meta‐analysis. In all four trials, median Tg in pregnancy was lower in the group which received oral iodine compared to the no iodine group; the difference ranged from 2.1 to 7.5 µg/L (Table 4).

Maternal Tg ‐ postpartum (in μg/L)

Four trials involving 457 women (Brucker‐Davis 2013; Liesenkotter 1996; Pedersen 1993; Gowachirapant 2014 reported this outcome, mostly as median Tg therefore we did not perform a meta‐analysis. All four trials showed that women who received oral iodine supplements had lower postpartum Tg levels compared to those who did not (Table 5). The difference between groups ranged from 1.57 to 11.6 µg/L.

Insufficient iodine intake ‐ pregnancy (median UIC less than 150 μg/L)

Two studies with 432 participants reported on this outcome (Brucker‐Davis 2013; Gowachirapant 2014). Results show that women who received oral iodine supplements had a lower likelihood of insufficient iodine intake in pregnancy compared to the comparison group; 33.3% of women in the iodine group had inadequate iodine intake compared to 53.2% in the control group (average RR 0.64; 95% CI 0.51 to 0.80) (Analysis 2.16). There was no evidence of statistical heterogeneity for this outcome (I² = 0%, Tau² = 0.00 and Chi² test for heterogeneity P = 0.56).

Insufficient iodine intake ‐ postpartum (breastfeeding: median UIC less than 100 μg/L)

Results from two trials (Bouhouch 2014 (C); Gowachirapant 2014) involving 425 women show that iodine supplementation reduced the likelihood of having insufficient iodine intake in the postpartum period (average RR 0.81; 95% CI 0.70 to 0.93) with no evidence of statistical heterogeneity (I² = 0%, Tau² = 0.00 and Chi² test for heterogeneity P = 0.90) (Analysis 2.17).

Excessive iodine intake (pregnancy only: median UIC greater than or equal to 500 μg/L)

One trial (Gowachirapant 2014) with 356 women showed that 7.0% of women who oral received iodine supplements and 1.6% of women who did not had excessive iodine intake during pregnancy (average RR 4.33; 95% CI 1.24 to 15.07) (Analysis 2.18).

Infant/child secondary outcomes

Small‐for‐gestational age (as defined by trial authors)

This outcome was reported in two trials (Zhou 2015; Gowachirapant 2014) with 377 infants which showed no evidence of difference between groups (average RR 1.26, 95% CI 0.77 to 2.05). There was no evidence of statistical heterogeneity (I² = 0%, Tau² = 0.00 and Chi² test for heterogeneity P = 0.83) (Analysis 2.19).

Congenital anomalies (including cretinism, as defined by trial authors)

One trial with 57 infants (Zhou 2015) reported this outcome, with no cases of congenital anomalies in either group (Analysis 2.20).

Growth (anthropometric Z scores)

No trials reported on this outcome.

Neonatal goitre (assessed by any method)

This outcome was reported in two trials (Glinoer 1993; Liesenkotter 1996) which assessed 228 neonates. Findings showed a lower likelihood of enlarged thyroid gland in the neonate following oral iodine supplementation in the mother (average RR 0.11; 95% CI 0.02 to 0.56), with no evidence of statistical heterogeneity (I² = 0%, Tau² = 0.00 and Chi² test for heterogeneity P = 0.76) (Analysis 2.21).

Neonatal thyroid volume (in mL) (assessed by any method)

Three trials with assessments of 359 neonates (Glinoer 1993; Liesenkotter 1996; Gowachirapant 2014) showed lower thyroid volume in the oral iodine compared to the control group. The MD was ‐0.34 mL (95% CI ‐0.58 to ‐0.11) though there was evidence of statistical heterogeneity (I² = 93%, Tau² = 0.04 and Chi² test for heterogeneity P = 0.00001) (Analysis 2.22).

Insufficient neonatal iodine intake (median UIC less than 100 μg/L)

One trial (Gowachirapant 2014) which assessed 159 neonates found a higher likelihood of insufficient iodine intake in the oral iodine compared to the control group (average RR 2.14; 95% CI 1.04 to 4.37) (Analysis 2.23).

Neonatal mental or motor development (as defined by trial authors)

One study (Gowachirapant 2014) reported scores from the Neonatal Behavioral Assessment Scales in median and IQR for 149 neonates (Table 6). Out of the eight sub‐scales, three showed higher median scores in the oral iodine group compared to the control group, one showed a higher score in the control group, and three showed equal scores in both groups.

Child mental or motor development (as defined by trial authors)

One trial (Zhou 2015) reported on child cognitive, language development, motor, social‐emotional and adoptive behaviour scores for 53 children at 18 months of age assessed by the Bayley scales of infant and toddler development (Bayley‐Ⅲ). Scores were similar in both groups (Analysis 2.24; Analysis 2.25; Analysis 2.26; Analysis 2.27; Analysis 2.28). This trial also reported moderate and severe developmental delays based on the cognitive, language and motor scores and found no significant differences between groups (Analysis 2.29; Analysis 2.30; Analysis 2.31; Analysis 2.32; Analysis 2.33; Analysis 2.34).

Infant death (death in the first year of life)

No trials reported on this outcome.

(3) Oral iodine‐only supplement versus no intervention or placebo (five trials, 580 participants)

Five studies involving 580 women were included in this comparison (Bouhouch 2014 (C); Glinoer 1993; Liesenkotter 1996; Pedersen 1993; Zhou 2015), all of which contributed data. None were assessed as high quality.

Maternal primary outcomes

Hypothyroidism ‐ pregnancy (as defined by trial authors)

No trials reported on this outcome.

Hypothyroidism ‐ postpartum (as defined by trial authors)

Among 340 women in three trials (Bouhouch 2014 (C); Liesenkotter 1996; Pedersen 1993), those who received oral iodine‐only supplementation were as likely to be diagnosed with hypothyroidism in the postpartum period as those who did not receive any iodine (average RR 0.91; 95% CI 0.10 to 8.60) (Analysis 3.1). We did not find evidence of statistical heterogeneity (I² = 0%, Tau² = 0.00 and Chi² test for heterogeneity P = 0.32).

Preterm birth (as defined by trial authors)

One trial with 58 assessments (Zhou 2015) showed no difference in the likelihood of preterm birth for women who received oral supplements with iodine only compared to those who received no iodine (average RR 1.25; 95% CI 0.37 to 4.19) (Analysis 3.2).

Maternal adverse effect: elevated TPO‐ab ‐ pregnancy (as defined by trial authors)

No trial reported on this outcome.

Maternal adverse effect: elevated TPO‐ab ‐ postpartum (as defined by trial authors)

This outcome was reported in two trials with 162 women (Liesenkotter 1996; Pedersen 1993). Findings showed no difference between the group which received oral iodine‐only and the group which did not receive any iodine (average RR 0.62; 95% CI 0.15 to 2.51) (Analysis 3.3). We did not find any evidence of statistical heterogeneity for this analysis (I² = 0%, Tau² = 0.00 and Chi² test for heterogeneity P = 0.52).

Maternal adverse effect: hyperthyroidism ‐ pregnancy (as defined by trial authors)

No trial reported on this outcome.

Maternal adverse effect: hyperthyroidism ‐ postpartum (as defined by trial authors)

Two trials involving 286 women (Bouhouch 2014 (C); Liesenkotter 1996) reported postpartum hyperthyroidism. This event occurred only in one trial though (Bouhouch 2014 (C)) and there was no difference in the likelihood of postpartum hyperthyroidism between groups (average RR 0.29; 95% CI 0.01 to 7.06) (Analysis 3.4).

Maternal adverse effect: digestive intolerance ‐ pregnancy (as defined by trial authors)

No trial reported on this outcome.

Infant/child primary outcomes

Perinatal mortality (including stillbirth/fetal death and neonatal death, as defined by trial authors)

One trial with 58 assessments (Zhou 2015) reported no events in either group (Analysis 3.5).

Low birthweight (less than 2500 g)

One trial (Zhou 2015) reported this outcome with 58 neonates, showing no difference between groups (average RR 0.33; 95% CI 0.04 to 3.02) (Analysis 3.6).

Neonatal hypothyroidism or elevated TSH (as defined by trial authors)

One trial with 41 assessments (Zhou 2015) reported on this outcome and showed no events in either group (Analysis 3.7),

Neonatal adverse effect ‐ elevated TPO‐ab (as defined by trial authors)

One study with 108 assessments (Liesenkotter 1996) reported on this outcome and showed no difference between groups (average RR 0.61; 95% CI 0.07 to 5.70) (Analysis 3.8).

Neonatal adverse effect ‐ hyperthyroidism (as defined by trial authors)

No trials reported this outcome.

Maternal secondary outcomes

Spontaneous miscarriage (as defined by trial authors)

This outcome was reported in one trial with 58 pregnant women (Zhou 2015), which showed no difference in likelihood of spontaneous miscarriage between the oral iodine‐only and no iodine groups (average RR 3.00; 95% CI 0.13 to 70.74) (Analysis 3.9).

Maternal goitre ‐ pregnancy (assessed by any method)

One study involving 120 women (Glinoer 1993) reported on goitre in pregnancy and found no difference between groups (average RR 0.60; 95% CI 0.23 to 1.55) (Analysis 3.10).

Maternal thyroid size ‐ goitre postpartum (assessed by any method)

Postpartum goitre was reported in one trial (Liesenkotter 1996). Data from 108 women showed no evidence of difference between groups (average RR 1.01; 95% CI 0.73 to 1.39) (Analysis 3.11).

Maternal thyroid volume ‐ pregnancy (in mL) (assessed by any method)

One trial with 54 women (Pedersen 1993) reported on this outcome. Median thyroid volume in the oral iodine‐only group was lower, by 1.3 mL, than in the no‐iodine group (Table 2).

Maternal thyroid volume ‐ postpartum (in mL) (assessed by any method)

Two trials including 157 women reported on this outcome (Liesenkotter 1996; Pedersen 1993). Findings show similar median thyroid volume in the iodine‐only supplementation group and the no‐iodine group (Table 3).

Maternal Tg ‐ pregnancy (μg/L)

This outcome was reported in one trial with 54 women (Pedersen 1993) which showed 45% lower postpartum Tg levels in the iodine‐only group compared to the group that received no iodine (Table 4).

Maternal Tg ‐ postpartum (μg/L)

Two trials involving 157 women (Liesenkotter 1996; Pedersen 1993) reported on this outcome. We did not perform a meta‐analysis because one trial reported median Tg. Results from these studies show lower postpartum Tg levels in the iodine‐only supplementation group compared to the no‐iodine group (Table 5). In one trial mean Tg was 5.2 μg/L lower (Liesenkotter 1996) and in the other median Tg was 5.6 μg/L lower (Pedersen 1993).

Insufficient iodine intake ‐ pregnancy (median UIC less than 150 μg/L)

No trial reported on this outcome.

Insufficient iodine intake ‐ postpartum (breastfeeding: median UIC less than 100 μg/L)

Results from one trial (Bouhouch 2014 (C)) with 175 women show that those who received oral iodine‐only supplements were less likely to have inadequate iodine intake in the postpartum period compared to those who received no iodine (average RR 0.81; 95% CI 0.69 to 0.96) (Analysis 3.12).

Excessive iodine intake (pregnancy only: median UIC greater than or equal to 500 μg/L)

No trial reported on this outcome.

Infant/child secondary outcomes

Small‐for‐gestational age (as defined by trial authors)

One trial with 58 neonates (Zhou 2015) reported this outcome (average RR 1.50; 95% CI 0.27 to 8.32) (Analysis 3.13).

Congenital anomalies (including cretinism, as defined by trial authors)

One trial with 57 infants (Zhou 2015) reported on congenital anomalies and showed no cases in either group (Analysis 3.14).

Growth (anthropometric Z scores)

No trial reported on this outcome.

Neonatal goitre (assessed by any method)

Two trials with 228 neonates (Glinoer 1993; Liesenkotter 1996) reported on this outcome and showed a lower likelihood of neonatal goitre in the oral iodine‐only supplementation group compared to the control group (average RR 0.11; 95% CI 0.02 to 0.56) (Analysis 3.15). We found no evidence of statistical heterogeneity (I² = 0%, Tau² = 0.00 and Chi² test for heterogeneity P = 0.76).

Neonatal thyroid volume (in mL) (assessed by any method)

This outcome was reported in two studies with 228 assessments (Glinoer 1993; Liesenkotter 1996) which showed a lower neonatal thyroid volume, by 0.53 mL, in the oral iodine‐only group in comparison to the control group (95% CI ‐1.02 to ‐0.03) (Analysis 3.16). However there was significant statistical heterogeneity (I² = 91%, Tau² = 0.12 and Chi² test for heterogeneity P = 0.0007).

Insufficient neonatal iodine intake (median UIC less than 100 μg/L)

No trial reported on this outcome.

Child mental or motor development (as defined by trial authors)

One trial (Zhou 2015) reported on child cognitive, language development, motor, social‐emotional and adaptive behaviour scores at 18 months of age for 53 children assessed by the Bayley scales of infant and toddler development (Bayley‐Ⅲ). Scores were similar in both groups (Analysis 3.17; Analysis 3.18; Analysis 3.19; Analysis 3.20; Analysis 3.21). The same trial also reported moderate and severe developmental delays based on the cognitive, language and motor scores and found no significant differences between groups (Analysis 3.22; Analysis 3.23; Analysis 3.24; Analysis 3.25; Analysis 3.26; Analysis 3.27).

Infant death (death in the first year of life)

No trial reported on this outcome.

(4) Oral iodine supplement with other vitamins and/or minerals versus only other vitamins and/or minerals (exact same formulation of other vitamins/minerals, but no iodine) (three trials, 694 participants)

Three studies involving 694 women were included in this comparison (Brucker‐Davis 2013; Nohr 2000; Gowachirapant 2014); all three contributed data. None were assessed as high quality.

Maternal primary outcomes

Hypothyroidism ‐ pregnancy (as defined by trial authors)

Data from one trial (Gowachirapant 2014) involving 365 women showed no difference in risk of hypothyroidism in pregnancy between women who received supplements containing iodine and other vitamins/minerals and those who received only the other vitamins/minerals (average RR 1.90; 95% CI 0.57 to 6.38) (Analysis 4.1).

Hypothyroidism ‐ postpartum (as defined by trial authors)

One study (Gowachirapant 2014) with 254 women reported on this outcome and found no evidence of difference in the likelihood of postpartum hypothyroidism between groups (average RR 0.97; 95% CI 0.06 to 15.32) (Analysis 4.2).

Preterm birth (as defined by trial authors)

Only one trial (Gowachirapant 2014) involving 318 women reported on preterm birth, and showed no difference between groups (average RR 0.51; 95% CI 0.23 to 1.14) (Analysis 4.3).

Maternal adverse effect: elevated TPO‐ab ‐ pregnancy (as defined by trial authors)

One trial (Gowachirapant 2014) involving 359 women showed no difference between groups (average RR 0.95; 95% CI 0.44 to 2.07) (Analysis 4.4).

Maternal adverse effect: elevated TPO‐ab ‐ postpartum (as defined by trial authors)

This outcome was reported in one trial (Gowachirapant 2014) with 235 women, which showed no difference between groups (average RR 1.03; 95% CI 0.79 to 1.33) (Analysis 4.5).

Maternal adverse effect: hyperthyroidism ‐ pregnancy (as defined by trial authors)

One trial with 365 women reported on this outcome (Gowachirapant 2014) and found no difference between groups (average RR 1.90; 95% CI 0.57 to 6.38) (Analysis 4.6).

Maternal adverse effect: hyperthyroidism ‐ postpartum (as defined by trial authors)

This outcome was reported in one trial involving 257 women (Gowachirapant 2014) (average RR 0.35; 95% CI 0.12 to 1.06) (Analysis 4.7).

Maternal adverse effect: digestive intolerance ‐ pregnancy (as defined by trial authors)

One trial with 76 women reported on this outcome (Brucker‐Davis 2013) and found a higher likelihood of digestive intolerance in the group that received oral iodine supplementation with other micronutrients compared to those who received only other micronutrients. One third of women (33.3%) in the iodine group and 2.2% of women in the control group dropped out of the study due to nausea or vomiting (average RR 15.33, 95% CI 2.07 to 113.70; participants = 76) (Analysis 4.8).

Infant/child primary outcomes

Perinatal mortality (including stillbirth/fetal death and neonatal death, as defined by trial authors)

No trial reported on this outcome.

Low birthweight (less than 2500 g)

One study with 319 infants reported on this outcome (Gowachirapant 2014) and showed no difference between groups (average RR 0.61; 95% CI 0.27 to 1.39) (Analysis 4.9).

Neonatal hypothyroidism or elevated TSH (as defined by trial authors)

This outcome was reported in one trial (Gowachirapant 2014) with 219 infant assessments; the average RR for neonatal elevated TSH was 0.58 (95% CI 0.11 to 3.12) (Analysis 4.10).

Neonatal adverse effect ‐ elevated TPO‐ab (as defined by trial authors)

No trials reported this outcome.

Neonatal adverse effect ‐ hyperthyroidism (as defined by trial authors)

No trials reported this outcome.

Maternal secondary outcomes

Spontaneous miscarriage (as defined by trial authors)

Two trials (Brucker‐Davis 2013; Gowachirapant 2014) with information on 587 pregnancies showed a similar likelihood of spontaneous miscarriage (average RR 1.26; 95% CI 0.60 to 2.62) and no evidence of statistical heterogeneity (I² = 0%, Tau² = 0.00 and Chi² test for heterogeneity P = 0.43) (Analysis 4.11).

Maternal goitre ‐ pregnancy (assessed by any method)

One trial with 366 women (Gowachirapant 2014) showed no difference between groups (average RR 1.89; 95% CI 0.56 to 6.34) (Analysis 4.12).

Maternal goitre ‐ postpartum (assessed by any method)

This outcome was reported in one study (Gowachirapant 2014) involving 262 women. Results showed no difference between women who received oral iodine with other micronutrients and those who received only other micronutrients (average RR 2.06; 95% CI 0.64 to 6.68) (Analysis 4.13).

Maternal thyroid volume ‐ pregnancy (in mL) (assessed by any method)

Two trials (Brucker‐Davis 2013; Gowachirapant 2014) with 442 participants reported median prenatal thyroid volume so we did not perform a meta‐analysis. Values were similar in both groups in one trial (Gowachirapant 2014) and 0.4 mL lower in the group receiving iodine with other micronutrients compared to the no‐iodine group in the other trial (Brucker‐Davis 2013) (Table 2).

Maternal thyroid volume ‐ postpartum (in mL) (assessed by any method)

Two trials (Brucker‐Davis 2013; Gowachirapant 2014) with 300 women reported median postpartum thyroid volume, so we did not perform a meta‐analysis for this outcome. Findings showed similar values in both groups in one trial (Gowachirapant 2014) and 1.2 mL lower thyroid volume in the group that received iodine with other micronutrients compared to the no‐iodine group in the other trial (Brucker‐Davis 2013)(Table 3).

Maternal Tg ‐ pregnancy (μg/L)

This outcome was reported in three studies (Brucker‐Davis 2013; Nohr 2000; Gowachirapant 2014) with 508 women. All three trials reported median Tg therefore we did not perform a meta‐analysis. In each of these trials, prenatal Tg concentrations were lower in women who received iodine supplements with other micronutrients compared to those who received no iodine (Table 4). Differences ranged from 2.1 μg/L to 5.3 μg/L.

Maternal Tg ‐ postpartum (μg/L)

Two trials involving 300 women (Brucker‐Davis 2013; Gowachirapant 2014) reported median postpartum Tg. In both of these studies, Tg concentrations were lower in the group receiving iodine plus other micronutrients than in the no‐iodine group (Table 5), in one study by 1.57 μg/L (Gowachirapant 2014) and in another by 11.6 μg/L (Brucker‐Davis 2013).

Insufficient iodine intake ‐ pregnancy (median UIC less than 150 μg/L)

Two studies with 432 participants reported on this outcome (Brucker‐Davis 2013; Gowachirapant 2014). Results show that women who received iodine supplements with other micronutrients had a lower likelihood of insufficient iodine intake in pregnancy compared to those who received only other micronutrients (average RR 0.64; 95% CI 0.51 to 0.80) (Analysis 4.14). There was no evidence of statistical heterogeneity for this outcome (I² = 0%, Tau² = 0.00 and Chi² test for heterogeneity P = 0.56).

Insufficient iodine intake ‐ postpartum (breastfeeding: median UIC less than 100 μg/L)

Findings from one trial (Gowachirapant 2014) with 250 women showed no clear difference of having insufficient iodine intake postpartum (average RR 0.80; 95% CI 0.61 to 1.04) (Analysis 4.15).

Excessive iodine intake (pregnancy only: median UIC greater than or equal to 500 μg/L)

One study with 356 women reported on this outcome (Gowachirapant 2014) and found a greater likelihood of excessive iodine intake in pregnancy in women who received iodine and other vitamins/minerals compared to those who received only other vitamins/minerals (average RR 4.33; 95% CI 1.24 to 15.07) (Analysis 4.16).

Infant/child secondary outcomes

Small‐for‐gestational age (as defined by trial authors)

This outcome was reported in one trial (Gowachirapant 2014) with 319 assessments which showed no evidence of difference between groups (average RR 1.24; 95% CI 0.74 to 2.06) (Analysis 4.17).

Congenital anomalies (including cretinism, as defined by trial authors)

No trials reported on this outcome.

Growth (anthropometric Z scores)

No trials reported on this outcome.

Neonatal goitre (assessed by any method)

No trials reported on this outcome.

Neonatal thyroid volume (in mL) (assessed by any method)

This outcome was reported in one trial (Gowachirapant 2014) which assessed 131 neonates and found a trend towards lower thyroid volume in the group whose mothers received iodine with other micronutrients compared to those who received only other micronutrients (MD ‐0.09; 95% CI ‐0.18 to ‐0.00) (Analysis 4.18).

Insufficient neonatal iodine intake (median UIC less than 100 μg/L)

Neonatal mental or motor development (as defined by trial authors)

One study (Gowachirapant 2014) reported scores from the Neonatal Behavioral Assessment Scales for 149 neonates in median and IQR (Table 6). Out of the eight sub‐scales, three showed higher median scores in the oral iodine group compared to the control group, one showed a higher score in the control group, and three showed equal scores in both groups.

Infant death (death in the first year of life)

No trials reported on this outcome.

(5) Any injected iodine supplement versus same supplement without iodine or no treatment/placebo (three trials,1463 participants)

Three trials were included in this comparison (Kevany 1969; Pharoah 1971 (C); Thilly 1978), all of which contributed data. One trial did not report the number of women randomised (Pharoah 1971 (C)). None of the studies were assessed as high quality.

Maternal primary outcomes

None of the trials reported on this review's maternal primary outcomes (hypothyroidism during pregnancy or postpartum; preterm birth or any adverse effect in pregnancy or postpartum).

Infant/child primary outcomes

Perinatal mortality (including stillbirth/fetal death and neonatal death, as defined by trial authors)

One trial with 399 assessments (Thilly 1978) reported on this outcome and found a trend towards lower perinatal mortality in the injected‐iodine compared to the control group (average RR 0.66; 95% CI 0.42 to 1.03) (Analysis 5.1).

No trials reported on this review's other infant primary outcomes (low birthweight, hypothyroidism or elevated TSH or any adverse effect).

Maternal secondary outcomes

No trials reported on the prespecified maternal secondary outcomes (spontaneous miscarriage, maternal goitre during pregnancy or postpartum, thyroid volume ‐ pregnancy or postpartum, Tg ‐ pregnancy or postpartum, insufficient or excessive iodine intake during pregnancy or postpartum).

Infant/child secondary outcomes

Congenital anomalies (including cretinism, as defined by trial authors)

One study (Pharoah 1971 (C)), which assessed 818 children, showed a lower frequency of congenital anomalies in the injected iodine group compared to the control group (average RR 0.27; 95% CI 0.12 to 0.60) (Analysis 5.2).

Neonatal goitre (assessed by any method)

One trial with 465 neonates (Kevany 1969) reported on this outcome; no events were found in either group (Analysis 5.3).

Mental or motor development (as defined by trial authors)

Data from two trials (Kevany 1969; Thilly 1978), which measured 174 children, showed a higher IQ score, by 11.21 points on average, in the injected iodine compared to the control group (95% CI 7.96 to 14.46), with no statistical heterogeneity (I² = 0%, Tau² = 0.00; Chi² test for heterogeneity P = 0.70) (Analysis 5.4).

No trials reported on any of the other infant/child prespecified secondary outcomes (growth, small‐for‐gestational age, thyroid volume, insufficient neonatal iodine intake or infant death).

(6) Injected iodine‐only supplement versus no intervention or placebo (three trials)

All of the trials included in comparison 5 reported using injected iodine‐only supplements (Kevany 1969; Pharoah 1971 (C); Thilly 1978). Therefore the findings for comparison 6 are the same as for comparison 5 and are not repeated here.

(7) Injected iodine supplement with other vitamins and/or minerals versus only other vitamins and/or minerals (exact same formulation of other vitamins/minerals, but no iodine) (no trials)

No studies were included in this comparison.

Discussion

Summary of main results

This review evaluates the effects of oral or injected iodine supplementation alone or in combination with other micronutrients prior to, during, or following pregnancy. It includes 14 trials, though only 11 (with over 2700 women) contributed data. Eight of these 11 trials compared iodine alone versus no treatment or placebo (five oral and three injected) and three compared iodine with other vitamins/minerals compared to only other vitamins/minerals (all oral). We did not identify any studies that used injected iodine with other vitamins/minerals compared to only other vitamins/minerals.

We have summarized the primary outcome findings along with an overall assessment of quality of evidence in 'Summary of findings' tables (summary of findings Table for the main comparison; summary of findings Table 2).

In comparison to the group that received no iodine, those who received iodine had a lower likelihood of postpartum hyperthyroidism (1.6% versus 4.9% in three trials, low‐quality evidence) and a higher likelihood of digestive intolerance in pregnancy (33.3% versus 2.2% in one trial, very low‐quality evidence). Both hyperthyroidism and digestive intolerance were considered potential adverse effects. Those who received iodine supplementation also had a lower likelihood of perinatal mortality (11.9% versus 18.2% in two trials, low‐quality evidence), though this finding was not statistically significant. For the remaining primary outcomes, there were no clear differences between groups.

For secondary outcomes, in comparison to the group which received no iodine, those who received iodine supplementation had lower likelihood of insufficient iodine intake in pregnancy (two trials) and the postpartum period (two trials), greater likelihood of excessive iodine intake in pregnancy (one trial), lower likelihood of congenital anomalies (two trials), lower likelihood of neonatal goitre (three trials), lower neonatal thyroid volume (three trials), higher likelihood of neonatal insufficient iodine intake (one trial) and higher child mental development score (two trials).

No trials reported on infant/child growth or infant death and we found evidence of no effect for the remaining secondary outcomes.

Our findings, both those showing an effect and those showing no effect, must be interpreted with caution due to the low number of studies contributing data and small sample sizes.

Overall completeness and applicability of evidence

This review included data from 11 randomised or quasi‐randomized controlled trials that have taken place in 13 countries since the 1960s. We consider this evidence base to be limited, both in the number of trials and included women and in the outcomes of interest reported. The majority of primary outcomes were reported in only one or two trials. Where we identified substantial statistical heterogeneity, we were unable to perform any of the planned subgroup analyses because of insufficient data. None of the included trials were assessed as high quality and the quality of evidence for all of the primary outcomes was considered low or very low.

Iodine supplementation in women who may become pregnant, and pregnant or breastfeeding women generally aims to prevent and/or treat iodine deficiency and its consequences, both in the woman and her child. Our review has shown that iodine supplementation may decrease the risk of postpartum hyperthyroidism and increase the likelihood of experiencing nausea or vomiting in pregnancy. Evidence for these outcomes came from trials which took place in areas with mild to moderate iodine deficiency. These findings must be interpreted cautiously; the postpartum findings were driven by findings from one study and the digestive intolerance results came from one trial. There was an indication of decreased risk of perinatal mortality, though the difference between groups was not statistically significant and all of the events occurred in one trial which took place in a setting with severe iodine deficiency.

We have also demonstrated that iodine supplementation may affect indicators of iodine intake and iodine or thyroid status in women and their children ‐ mostly positively with the exception of an increased risk of excessive intake in pregnancy and higher likelihood of insufficient neonatal intake. We also identified evidence for prevention of cretinism in a severely iodine‐deficient area and mixed findings on mental development indicators, with positive effects in studies that took place in settings with severe deficiency. We had planned on performing subgroup analyses, including by baseline iodine status, however we did not identify enough trials to perform these additional analyses for any of the included outcomes.

The importance of routine iodine supplementation for important outcomes in pregnancy and postpartum and in children under two still needs to be demonstrated. To our knowledge there are five planned, ongoing or unpublished trials which may be eligible for inclusion in this review, in addition to data not yet published for the Indian component of the Gowachirapant 2014 trial (see Characteristics of ongoing studies). We expect that when reported, these trials will help fill some of the evidence gaps identified in this review. This review should be updated to incorporate the new findings once they are published.

Quality of the evidence

The overall risk of bias was unclear in most of the trials. Methods for sequence generation and allocation concealment were either not specified or considered at high risk of selection bias for most trials. Similarly, most trials did not describe blinding of personnel, participants and outcome assessors or the methods were considered at high risk of performance or detection bias. For most of the outcomes however, lack of blinding would be unlikely to influence results, though we cannot rule out this possibility. All but two studies were considered at either high risk of attrition bias due to incomplete outcome data or unclear risk because this information was not provided. For reporting and other bias, we generally did not have enough information to classify trials as high or low risk.

We assessed the quality of the evidence using GRADE and judged the evidence for comparison 1, any supplement containing iodine compared with same supplement without iodine or no treatment/placebo, for maternal primary outcomes (summary of findings Table for the main comparison), and child primary outcomes (summary of findings Table 2) (no outcomes beyond the neonatal period were reported in the included trials). For maternal outcomes, the quality of evidence was judged as low for maternal hypothyroidism during pregnancy, postpartum, preterm, elevated TPO‐ab during pregnancy and postpartum, hyperthyroidism during pregnancy and postpartum. Digestive intolerance was judged as very low‐quality evidence. For child outcomes, the quality of evidence was judged as low to very low for perinatal mortality, low birthweight, neonatal hypothyroidism or elevated TSH, and neonatal elevated TPO‐ab. The main reasons for downgrading were design limitations, wide confidence intervals and small event size.

Potential biases in the review process

We aimed to minimize bias at each stage of the review process. Two review authors independently assessed eligibility for inclusion, carried out data extraction and assessed risk of bias. However we recognize that this type of review requires many subjective judgements and others undertaking the same review may have made different decisions, particularly regarding eligibility and risk of bias.

We attempted to use a systematic and transparent process to assess the quality of the evidence relating to specific outcomes and produce 'Summary of findings' tables. Two review authors independently assessed the evidence for each quality domain for each outcome. Any disagreements were discussed and we involved a third review author where necessary. We planned to use funnel plots to assess publication bias but we were not able to do this for any of the outcomes due to the limited number of studies contributing data. We also relied mostly on published literature and did not systematically search for grey literature.

Agreements and disagreements with other studies or reviews

Zhou and colleagues conducted a systematic review on the effects of iodine supplementation during pregnancy or the periconceptional period on child growth and development as well as pregnancy outcomes and thyroid function, and included randomised or quasi‐randomized controlled trials (Zhou 2013). Our findings are consistent with the authors' conclusions that there was a lack of high‐quality evidence on the effect of supplementation on child growth and cognition.

Taylor and colleagues undertook a systematic review on the impact of iodine supplementation in settings with mild‐to‐moderate iodine deficiency (Taylor 2013). They included observational studies and randomised controlled trials of maternal iodine supplementation in pregnancy. Of the trials, they found that most showed a positive impact on maternal Tg levels and maternal thyroid volume, no evidence of negative effect on maternal TPO‐ab, and lower thyroid volume in neonates of women who received iodine. We found similar results, with the exception of maternal thyroid volume, where our findings were mixed.

Zimmermann and colleagues conducted a narrative systematic review/systematic literature search on the effects of iodine deficiency in pregnancy and infancy, and included trials on supplementation before and during pregnancy (Zimmermann 2012a). The authors concluded that in moderate to severely iodine‐deficient areas, iodine supplementation prevented cretinism, increased birthweight, reduced perinatal and infant death and improved child cognition. They also concluded that mild deficiency could impair thyroid function but that effects on child cognitive or neurologic function were uncertain. This review reached different conclusions to ours, however this may be explained by its design; it did not specify criteria for study inclusion and included non‐randomized trials as well as studies with different comparisons.

An additional review and meta‐analysis looked at iodine and mental development of young children and included randomised and non‐randomized supplementation trials (in mothers or infants, or both) and prospective cohort studies stratified by mother or infant iodine status (Bougma 2013). Findings showed a positive impact of iodine, regardless as to the study design, ranging from 6.9 to 10.2 IQ points, whereas our findings on child development were mixed.

Angermayr and Clar conducted a Cochrane Review on the effects of iodine supplementation for preventing iodine deficiency disorders in children and found that most trials resulted in increased urinary iodine excretion, an overall tendency towards reduction of goitre and mixed results for cognitive, psychomotor and physical development (Angermayr 2004). Less than 2% of children studied showed adverse effects and these effects were largely minor and transient. Most trials included in this review used iodized oil and most were considered low quality.

In the Cochrane Review on iodine supplementation for preventing mortality and adverse neurodevelopmental outcomes in preterm infants, Ibrahim and colleagues found only one eligible trial to include (Ibrahim 2006). This trial was underpowered for mortality and did not assess neurodevelopment. It showed no effect of iodine on T3, T4 or thyrotropin or the incidence of chronic lung disease. Our review did not include any of these outcomes.

Aburto and colleagues performed a review on the effect and safety of salt iodization for prevention of iodine deficiency disorders, which included over 80 studies of various designs (Aburto 2014). Findings showed that iodized salt increased intelligence quotient and urinary iodine excretion and reduced the risk of goitre, cretinism, low intelligence and low urinary iodine excretion. There was no effect found on hypothyroidism, considered by authors as a potential adverse effect, and inconsistent results for hyperthyroidism.

A Cochrane Review on point‐of‐use fortification of foods with multiple‐micronutrient powders for women during pregnancy identified one eligible trial that included iodine in the formulation, along with six other nutrients (Suchdev 2015). The comparison was tablets with the same formulation, therefore the findings are not relevant for discussion here. Similarly, the Haider and Bhutta Cochrane Review on multiple‐micronutrient supplementation for women during pregnancy included trials that compared multiple‐micronutrient supplements with supplements containing iron and folic acid, iron only, or placebo (Haider 2015).

Some of our findings were unexpected. The evidence that iodine supplementation reduced the likelihood of hyperthyroidism in the postpartum period came mainly from one trial, which took place in women with mild iodine deficiency in Thailand (Gowachirapant 2014). We are not certain of a mechanism to explain this relationship. It is possible that iodine supplementation helped to prevent postpartum thyroiditis, a feature of which is often hyperthyroidism (Stagnaro‐Green 2012). Other available evidence however suggests that iodine has no effect on or aggravates postpartum thyroiditis. We also found, with data from the same trial, that iodine supplementation increased the likelihood of insufficient iodine intake in neonates (as indicated by UIC) (Gowachirapant 2014). This finding is also unexpected and we are not aware of a plausible explanation.

Figure 1

Study flow diagram

Figure 2

Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies

Figure 3

Risk of bias summary: review authors' judgements about each risk of bias item for each included study

Analysis 1.1

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 1 Maternal hypothyroidism ‐ pregnancy.

Analysis 1.2

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 2 Maternal hypothyroidism ‐ postpartum.

Analysis 1.3

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 3 Preterm birth.

Analysis 1.4

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 4 Maternal adverse effect: elevated thyroid peroxidase antibodies ‐ pregnancy.

Analysis 1.5

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 5 Maternal adverse effect: elevated thyroid peroxidase antibodies ‐ postpartum.

Analysis 1.6

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 6 Maternal adverse effect: hyperthyroidism ‐ pregnancy.

Analysis 1.7

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 7 Maternal adverse effect: hyperthyroidism ‐ postpartum.

Analysis 1.8

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 8 Maternal adverse effect: digestive intolerance ‐ pregnancy.

Analysis 1.9

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 9 Perinatal mortality.

Analysis 1.10

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 10 Low birthweight.

Analysis 1.11

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 11 Neonatal hypothyroidism or elevated TSH.

Analysis 1.12

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 12 Neonatal adverse effect: elevated thyroid peroxidase antibodies.

Analysis 1.13

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 13 Spontaneous miscarriage.

Analysis 1.14

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 14 Maternal goitre ‐ pregnancy.

Analysis 1.15

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 15 Maternal goitre ‐ postpartum.

Analysis 1.16

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 16 Maternal insufficient iodine intake ‐ pregnancy.

Analysis 1.17

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 17 Maternal insufficient iodine intake ‐ postpartum.

Analysis 1.18

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 18 Maternal excessive iodine intake ‐ pregnancy.

Analysis 1.19

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 19 Small‐for‐gestational age.

Analysis 1.20

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 20 Congenital anomalies.

Analysis 1.21

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 21 Neonatal goitre.

Analysis 1.22

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 22 Neonatal thyroid volume (in mL).

Analysis 1.23

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 23 Neonatal insufficient iodine intake.

Analysis 1.24

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 24 Child mental or motor development (IQ points).

Analysis 1.25

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 25 Child mental or motor development (cognitive score).

Analysis 1.26

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 26 Child mental or motor development (language score).

Analysis 1.27

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 27 Child mental or motor development (motor score).

Analysis 1.28

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 28 Child mental or motor development (social‐emotional score).

Analysis 1.29

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 29 Child mental or motor development (adaptive behaviour score).

Analysis 1.30

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 30 Child mental or motor development (cognitive score < 85).

Analysis 1.31

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 31 Child mental or motor development (cognitive score < 70).

Analysis 1.32

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 32 Child mental or motor development (language score < 85).

Analysis 1.33

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 33 Child mental or motor development (language score < 70).

Analysis 1.34

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 34 Child mental or motor development (motor score < 85).

Analysis 1.35

Comparison 1 Any supplement containing iodine versus same supplement without iodine or no intervention/placebo, Outcome 35 Child mental or motor development (motor score < 70).

Analysis 2.1

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 1 Maternal hypothyroidism ‐ pregnancy.

Analysis 2.2

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 2 Maternal hypothyroidism ‐ postpartum.

Analysis 2.3

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 3 Preterm birth.

Analysis 2.4

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 4 Maternal adverse effect: elevated thyroid peroxidase antibodies ‐ pregnancy.

Analysis 2.5

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 5 Maternal adverse effect: elevated thyroid peroxidase antibodies ‐ postpartum.

Analysis 2.6

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 6 Maternal adverse effect: hyperthyroidism ‐ pregnancy.

Analysis 2.7

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 7 Maternal adverse effect: hyperthyroidism ‐ postpartum.

Analysis 2.8

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 8 Maternal adverse effect: digestive intolerance ‐ pregnancy.

Analysis 2.9

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 9 Perinatal mortality.

Analysis 2.10

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 10 Low birthweight.

Analysis 2.11

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 11 Neonatal hypothyroidism or elevated TSH.

Analysis 2.12

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 12 Neonatal adverse effect: elevated thyroid peroxidase antibodies.

Analysis 2.13

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 13 Spontaneous miscarriage.

Analysis 2.14

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 14 Maternal goitre ‐ pregnancy.

Analysis 2.15

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 15 Maternal goitre ‐ postpartum.

Analysis 2.16

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 16 Maternal insufficient iodine intake ‐ pregnancy.

Analysis 2.17

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 17 Maternal insufficient iodine intake ‐ postpartum.

Analysis 2.18

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 18 Maternal excessive iodine intake ‐ pregnancy.

Analysis 2.19

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 19 Small‐for‐gestational age.

Analysis 2.20

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 20 Congenital anomalies.

Analysis 2.21

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 21 Neonatal goitre.

Analysis 2.22

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 22 Neonatal thyroid volume (in mL).

Analysis 2.23

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 23 Neonatal insufficient iodine intake.

Analysis 2.24

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 24 Child mental or motor development (cognitive score).

Analysis 2.25

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 25 Child mental or motor development (language score).

Analysis 2.26

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 26 Child mental or motor development (motor score).

Analysis 2.27

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 27 Child mental or motor development (social‐emotional score).

Analysis 2.28

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 28 Child mental or motor development (adaptive behaviour score).

Analysis 2.29

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 29 Child mental or motor development (cognitive score < 85).

Analysis 2.30

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 30 Child mental or motor development (cognitive score < 70).

Analysis 2.31

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 31 Child mental or motor development (language score < 85).

Analysis 2.32

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 32 Child mental or motor development (language score < 70).

Analysis 2.33

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 33 Child mental or motor development (motor score < 85).

Analysis 2.34

Comparison 2 Any oral iodine supplement versus same supplement without iodine or no intervention/placebo, Outcome 34 Child mental or motor development (motor score < 70).

Analysis 3.1

Comparison 3 Oral iodine‐only supplement versus no intervention or placebo, Outcome 1 Maternal hypothyroidism ‐ postpartum.

Analysis 3.2

Comparison 3 Oral iodine‐only supplement versus no intervention or placebo, Outcome 2 Preterm birth.

Analysis 3.3

Comparison 3 Oral iodine‐only supplement versus no intervention or placebo, Outcome 3 Maternal adverse effect: elevated thyroid peroxidase antibodies ‐ postpartum.

Analysis 3.4

Comparison 3 Oral iodine‐only supplement versus no intervention or placebo, Outcome 4 Maternal adverse effect: hyperthyroidism ‐ postpartum.

Analysis 3.5

Comparison 3 Oral iodine‐only supplement versus no intervention or placebo, Outcome 5 Perinatal mortality.

Analysis 3.6

Comparison 3 Oral iodine‐only supplement versus no intervention or placebo, Outcome 6 Low birthweight.

Analysis 3.7

Comparison 3 Oral iodine‐only supplement versus no intervention or placebo, Outcome 7 Neonatal hypothyroidism or elevated TSH.

Analysis 3.8

Comparison 3 Oral iodine‐only supplement versus no intervention or placebo, Outcome 8 Neonatal adverse effect: elevated thyroid peroxidase antibodies.

Analysis 3.9

Comparison 3 Oral iodine‐only supplement versus no intervention or placebo, Outcome 9 Spontaneous miscarriage.

Analysis 3.10

Comparison 3 Oral iodine‐only supplement versus no intervention or placebo, Outcome 10 Maternal goitre ‐ pregnancy.

Analysis 3.11

Comparison 3 Oral iodine‐only supplement versus no intervention or placebo, Outcome 11 Maternal goitre ‐ postpartum.

Analysis 3.12

Comparison 3 Oral iodine‐only supplement versus no intervention or placebo, Outcome 12 Maternal insufficient iodine intake ‐ postpartum.

Analysis 3.13

Comparison 3 Oral iodine‐only supplement versus no intervention or placebo, Outcome 13 Small‐for‐gestational age.

Analysis 3.14

Comparison 3 Oral iodine‐only supplement versus no intervention or placebo, Outcome 14 Congenital anomalies.

Analysis 3.15

Comparison 3 Oral iodine‐only supplement versus no intervention or placebo, Outcome 15 Neonatal goitre.

Analysis 3.16

Comparison 3 Oral iodine‐only supplement versus no intervention or placebo, Outcome 16 Neonatal thyroid volume (in mL).

Analysis 3.17

Comparison 3 Oral iodine‐only supplement versus no intervention or placebo, Outcome 17 Child mental or motor development (cognitive score).

Analysis 3.18

Comparison 3 Oral iodine‐only supplement versus no intervention or placebo, Outcome 18 Child mental or motor development (language score).

Analysis 3.19

Comparison 3 Oral iodine‐only supplement versus no intervention or placebo, Outcome 19 Child mental or motor development (motor score).

Analysis 3.20

Comparison 3 Oral iodine‐only supplement versus no intervention or placebo, Outcome 20 Child mental or motor development (social‐emotional score).

Analysis 3.21

Comparison 3 Oral iodine‐only supplement versus no intervention or placebo, Outcome 21 Child mental or motor development (adaptive behaviour score).

Analysis 3.22

Comparison 3 Oral iodine‐only supplement versus no intervention or placebo, Outcome 22 Child mental or motor development (cognitive score < 85).

Analysis 3.23

Comparison 3 Oral iodine‐only supplement versus no intervention or placebo, Outcome 23 Child mental or motor development (cognitive score < 70).

Analysis 3.24

Comparison 3 Oral iodine‐only supplement versus no intervention or placebo, Outcome 24 Child mental or motor development (language score < 85).

Analysis 3.25

Comparison 3 Oral iodine‐only supplement versus no intervention or placebo, Outcome 25 Child mental or motor development (language score < 70).

Analysis 3.26

Comparison 3 Oral iodine‐only supplement versus no intervention or placebo, Outcome 26 Child mental or motor development (motor score < 85).

Analysis 3.27

Comparison 3 Oral iodine‐only supplement versus no intervention or placebo, Outcome 27 Child mental or motor development (motor score < 70).

Analysis 4.1

Comparison 4 Oral iodine supplement with other vitamins and/or minerals versus only other vitamins and/or minerals but no iodine, Outcome 1 Maternal hypothyroidism ‐ pregnancy.

Analysis 4.2

Comparison 4 Oral iodine supplement with other vitamins and/or minerals versus only other vitamins and/or minerals but no iodine, Outcome 2 Maternal hypothyroidism ‐ postpartum.

Analysis 4.3

Comparison 4 Oral iodine supplement with other vitamins and/or minerals versus only other vitamins and/or minerals but no iodine, Outcome 3 Preterm birth.

Analysis 4.4

Comparison 4 Oral iodine supplement with other vitamins and/or minerals versus only other vitamins and/or minerals but no iodine, Outcome 4 Maternal adverse effect: elevated thyroid peroxidase antibodies ‐ pregnancy.

Analysis 4.5

Comparison 4 Oral iodine supplement with other vitamins and/or minerals versus only other vitamins and/or minerals but no iodine, Outcome 5 Maternal adverse effect: elevated thyroid peroxidase antibodies ‐ postpartum.

Analysis 4.6

Comparison 4 Oral iodine supplement with other vitamins and/or minerals versus only other vitamins and/or minerals but no iodine, Outcome 6 Maternal adverse effect: hyperthyroidism ‐ pregnancy.

Analysis 4.7

Comparison 4 Oral iodine supplement with other vitamins and/or minerals versus only other vitamins and/or minerals but no iodine, Outcome 7 Maternal adverse effect: hyperthyroidism ‐ postpartum.

Analysis 4.8

Comparison 4 Oral iodine supplement with other vitamins and/or minerals versus only other vitamins and/or minerals but no iodine, Outcome 8 Maternal adverse effect: digestive intolerance ‐ pregnancy.

Analysis 4.9

Comparison 4 Oral iodine supplement with other vitamins and/or minerals versus only other vitamins and/or minerals but no iodine, Outcome 9 Low birthweight.

Analysis 4.10

Comparison 4 Oral iodine supplement with other vitamins and/or minerals versus only other vitamins and/or minerals but no iodine, Outcome 10 Neonatal hypothyroidism or elevated TSH.

Analysis 4.11

Comparison 4 Oral iodine supplement with other vitamins and/or minerals versus only other vitamins and/or minerals but no iodine, Outcome 11 Spontaneous miscarriage.

Analysis 4.12

Comparison 4 Oral iodine supplement with other vitamins and/or minerals versus only other vitamins and/or minerals but no iodine, Outcome 12 Maternal goitre ‐ pregnancy.

Analysis 4.13

Comparison 4 Oral iodine supplement with other vitamins and/or minerals versus only other vitamins and/or minerals but no iodine, Outcome 13 Maternal goitre ‐ postpartum.

Analysis 4.14

Comparison 4 Oral iodine supplement with other vitamins and/or minerals versus only other vitamins and/or minerals but no iodine, Outcome 14 Maternal insufficient iodine intake ‐ pregnancy.

Analysis 4.15

Comparison 4 Oral iodine supplement with other vitamins and/or minerals versus only other vitamins and/or minerals but no iodine, Outcome 15 Maternal insufficient iodine intake ‐ postpartum.

Analysis 4.16

Comparison 4 Oral iodine supplement with other vitamins and/or minerals versus only other vitamins and/or minerals but no iodine, Outcome 16 Maternal excessive iodine intake ‐ pregnancy.

Analysis 4.17

Comparison 4 Oral iodine supplement with other vitamins and/or minerals versus only other vitamins and/or minerals but no iodine, Outcome 17 Small‐for‐gestational age.

Analysis 4.18

Comparison 4 Oral iodine supplement with other vitamins and/or minerals versus only other vitamins and/or minerals but no iodine, Outcome 18 Neonatal thyroid volume (in mL).

Analysis 4.19

Comparison 4 Oral iodine supplement with other vitamins and/or minerals versus only other vitamins and/or minerals but no iodine, Outcome 19 Neonatal insufficient iodine intake.

Analysis 5.1

Comparison 5 Any injected iodine supplement versus same supplement without iodine or no treatment/placebo, Outcome 1 Perinatal mortality.

Analysis 5.2

Comparison 5 Any injected iodine supplement versus same supplement without iodine or no treatment/placebo, Outcome 2 Congenital anomalies.

Analysis 5.3

Comparison 5 Any injected iodine supplement versus same supplement without iodine or no treatment/placebo, Outcome 3 Neonatal goitre.

Analysis 5.4

Comparison 5 Any injected iodine supplement versus same supplement without iodine or no treatment/placebo, Outcome 4 Child mental or motor development (IQ points).

Summary of findings for the main comparison. Any supplement containing iodine versus same supplement without iodine or no treatment/placebo (maternal outcomes)

Any supplement containing iodine versus same supplement without iodine or no treatment/placebo (maternal outcomes)
Population: women during the preconception, pregnancy and postpartum period Setting: Denmark, Germany, Morocco, New Zealand, Thailand, Zaire Intervention: any supplement containing iodine Comparison: same supplement without iodine or no treatment/placebo
Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Quality of the evidence (GRADE)	Comments
	Risk with same supplement without iodine or no treatment/placebo	Risk with any supplement containing iodine
Maternal hypothyroidism ‐ pregnancy	Study population		Average RR 1.90 (0.57 to 6.38)	365 (1 RCT)	⊕⊕⊝⊝ Low ^{1, 2}
	21 per 1000	40 per 1000 (12 to 134)
	Moderate
	21 per 1000	40 per 1000 (12 to 135)
Maternal hypothyroidism ‐ postpartum	Study population		Average RR 0.44 (0.06 to 3.42)	540 (3 RCTs)	⊕⊕⊝⊝ Low ^{2, 3}
	7 per 1000	4 per 1000 (1 to 34)
	Moderate
	11 per 1000	5 per 1000 (1 to 37)
Preterm birth	Study population		Average RR 0.71 (0.30 to 1.66)	376 (2 RCTs)	⊕⊕⊝⊝ Low ^{1, 2}
	111 per 1000	78 per 1000 (33 to 184)
	Moderate
	122 per 1000	87 per 1000 (37 to 202)
Maternal adverse effect: elevated thyroid peroxidase antibodies ‐ pregnancy	Study population		Average RR 0.95 (0.44 to 2.07)	359 (1 RCT)	⊕⊕⊝⊝ Low ^{1, 2}
	68 per 1000	65 per 1000 (30 to 142)
	Moderate
	68 per 1000	65 per 1000 (30 to 142)
Maternal adverse effect: elevated TPO‐ab ‐ postpartum	Study population		Average RR 1.01 (0.78 to 1.30)	397 (3 RCTs)	⊕⊕⊝⊝ Low ^{1, 4}
	301 per 1000	304 per 1000 (235 to 391)
	Moderate
	77 per 1000	78 per 1000 (60 to 100)
Maternal adverse effect: hyperthyroidism ‐ pregnancy	Study population		Average RR 1.90 (0.57 to 6.38)	365 (1 RCT)	⊕⊕⊝⊝ Low ^{1, 2}
	21 per 1000	40 per 1000 (12 to 134)
	Moderate
	21 per 1000	40 per 1000 (12 to 135)
Maternal adverse effect: hyperthyroidism ‐ postpartum	Study population		Average RR 0.32 (0.11 to 1.91)	543 (3 RCTs)	⊕⊕⊝⊝ Low ^{2, 4}
	49 per 1000	16 per 1000 (5 to 45)
	Moderate
	6 per 1000	3 per 1000 (1 to 7)
Maternal adverse effect: digestive intolerance ‐ pregnancy	Study population		Average RR 15.33 (2.07 to 113.70)	76 (1 RCT)	⊕⊝⊝⊝ Very low ^{2, 5}
	22 per 1000	333 per 1,000 (45 to 1,000)
	Moderate
	25 per 1000	383 per 1000 (52 to 1,000)
*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: Confidence interval; RR: Risk ratio; TPO‐ab: thyroid peroxidase antibodies
GRADE Working Group grades of evidence High quality: We are very confident that the true effect lies close to that of the estimate of the effect Moderate quality: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different Low quality: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect Very low quality: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect
¹One study with design limitations (high risk for attrition bias). ² Wide confidence interval. ³Most studies contributing data had design limitations (high risk for attrition bias and unclear for selection bias). ⁴Most studies contributing data had design limitations (high risk for attrition bias and selection bias). ⁵One study with high risk of blinding and attrition bias.

Summary of findings for the main comparison. Any supplement containing iodine versus same supplement without iodine or no treatment/placebo (maternal outcomes)

Summary of findings 2. Any supplement containing iodine versus same supplement without iodine or no treatment/placebo (child outcomes)

Any supplement containing iodine versus same supplement without iodine or no treatment/placebo (child outcomes)
Population: women during the preconception, pregnancy and postpartum period Setting: Germany, New Zealand, Thailand, Zaire Intervention: any supplement containing iodine Comparison: same supplement without iodine or no treatment/placebo
Outcomes	*Anticipated absolute effects^ (95% CI)**		Relative effect (95% CI)	№ of participants (studies)	Quality of the evidence (GRADE)	Comments
	Risk with same supplement without iodine or no treatment/placebo	Risk with any supplement containing iodine
Perinatal mortality	Study population		Average RR 0.66 (0.42 to 1.03)	399 (1 RCT)	⊕⊕⊝⊝ Low ^{1, 2}
	208 per 1000	137 per 1000 (87 to 214)
	Moderate
	208 per 1000	137 per 1000 (87 to 214)
Low birthweight	Study population		Average RR 0.56 (0.26 to 1.23)	377 (2 RCTs)	⊕⊕⊝⊝ Low ^{2, 3}
	90 per 1000	51 per 1000 (24 to 111)
	Moderate
	96 per 1000	54 per 1000 (25 to 118)
Neonatal hypothyroidism or elevated TSH	Study population		Average RR 0.58 (0.11 to 3.12)	219 (1 RCT)	⊕⊝⊝⊝ Very low ^{2, 3, 5}
	34 per 1000	20 per 1000 (4 to 106)
	Moderate
	34 per 1000	20 per 1000 (4 to 106)
Neonatal adverse effect: elevated TPO‐ab	Study population		Average RR 0.61 (0.07 to 5.70)	108 (1 RCT)	⊕⊝⊝⊝ Very low ^{2, 4, 5}
	43 per 1000	26 per 1000 (3 to 244)
	Moderate
	43 per 1000	26 per 1000 (3 to 245)
Child adverse effect: hyperthyroidism			See comments			No trials reported this outcome



*The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: Confidence interval; RR: Risk ratio; TPO‐ab: thyroid peroxidase antibodies; TSH: thyroid‐stimulating hormone
GRADE Working Group grades of evidence High quality: We are very confident that the true effect lies close to that of the estimate of the effect Moderate quality: We are moderately confident in the effect estimate: The true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different Low quality: Our confidence in the effect estimate is limited: The true effect may be substantially different from the estimate of the effect Very low quality: We have very little confidence in the effect estimate: The true effect is likely to be substantially different from the estimate of effect
¹Design limitations (high risk of attrition bias and unclear for selection bias). ²Wide confidence interval. ³One study had design limitations (high risk for attrition bias). ⁴One study with design limitations (high risk for selection bias). ⁵Small number of events.

Summary of findings 2. Any supplement containing iodine versus same supplement without iodine or no treatment/placebo (child outcomes)

Table 1. Urinary iodine concentration and breast milk iodine concentration following prepregnancy, pregnancy or postpartum iodine supplementation

Study	Setting	Intervention timing	Intervention description	Timing of assessment	Urinary iodine concentration		Breast milk iodine concentration
Study	Setting	Intervention timing	Intervention description	Timing of assessment	Iodine	Control	Iodine	Control
Bouhouch 2014 (C)	Southern Morocco	Postpartum	Iodine: single dose oral 400 mg, control: single dose placebo	Baseline	Median: 37 μg/L, IQR: 22‐77 (n = 119)	Median: 30 μg/L, IQR: 18‐61 (n = 115)
Bouhouch 2014 (C)	Southern Morocco	Postpartum		Postpartum: 9 months	Median: 58 μg/L, IQR: 34‐135 (n = 94)	Median: 39 μg/L, IQR 24‐62 (n = 81)	Median: 39.4 μg/L, IQR: 23.5‐66.7 (n = 94)	Median: 26.2 μg/L, IQR: 17.7‐42.7 (n = 81)
Brucker‐Davis 2013	Nice, France	Pregnancy	Iodine: daily dose 150 μg of iodine, control: vitamin mix but no iodine	Baseline: before 12 weeks of amenorrhoea	Median: 111 μg/L, IQR: 28‐399 (n = 32)	Median: 103 μg/L, IQR: 14‐355 (n = 54)
				Pregnancy: 3rd trimester	Median: 160.5 μg/L, IQR: 18‐358 (n = 30)	Median: 76 μg/L, IQR: 16‐303 (n = 46)
				Postpartum: 3 months	Median: 58 μg/L, IQR: 34‐135 (n = 18)	Median: 58 μg/L, IQR: 34‐135 (n = 18)
Glinoer 1993	Brussels, Belgium	Pregnancy	Iodine: single dose 1 mL iodized oil, control: multivitamin injection no iodine	Neonatal: 3‐6 days	Mean: 77 μg/L, SEM: +/‐ 8 (n = 60)	Mean: 43 μg/L, SEM: +/‐ 4 (n = 60)	Mean: 61 μg/L, SEM: +/‐10, (n = 60)	Mean: 29 μg/L, SEM: +/‐2, (n = 60)
Liesenkotter 1996	Berlin, Germany	Pregnancy	Daily dose oral 300 µg KI, control: no intervention	Postpartum: mean 11 days	Median: 104.5 μg/dl, IQR: NR (n = NR)	Median: In fig only (n = NR)
Liesenkotter 1996	Berlin, Germany	Pregnancy	Daily dose oral 300 µg KI, control: no intervention	Neonatal: 5 days	Median: 8.3 μg/dl, IQR: In fig only (n = NR)	Median: 6.5 μg/dl, IQR: In fig only (n = NR)
Mulrine 2010	Dunedin, New Zealand	Postpartum	Iodine: daily 75 or 150 µg/d iodine, control: placebo	Postpartum: 24 weeks (75 µg)	Median: 78 μg/L, 25th 75th %: 50,126 (n = 18)	Median: 34 μg/L, 25th 75th %: 26, 57 (n = 21)
Mulrine 2010	Dunedin, New Zealand	Postpartum	Iodine: daily 75 or 150 µg/d iodine, control: placebo	Postpartum: 24 weeks (150 µg)	Median: 84 μg/L, 25th‐75th %: 60,157 (n = 19)	Median: 34 μg/L, 25th 75th %: 26,57 (n = 21)	In figure only	In figure only
Nohr 2000	Aalborg, Denmark	Pregnancy or pregnancy and postpartum	Iodine: daily oral 150 µg elemental iodine and micronutrient supplement, control: micronutrient supplement no iodine	Pregnancy: 35 weeks	Median: 86 μg/24 h, 25th‐75th %: 66–147 (n = 22) pregnancy and postpartum	Median: 88 μg/24 h, 25th 75th %: 68‐123 (n = 24)
					Median: 83 μg/24 h, 25th 75th %: 54–145 (n = 20) pregnancy only
					Median: 50 μg/L, 25th 75th %: 35‐101 (n = 22) pregnancy and postpartum	Median: 51 μg/L, 25th 75th %: 30–80 (n = 20)
					Median: 52 μg/L, 25th 75th %: 36–81 (n = 20) pregnancy only
Pedersen 1993	East Jutland, Denmark	Pregnancy	Iodine: daily drops 200 μg iodine, control: no treatment	Pregnancy: 37 weeks	Median: 106 μg/L, 95 %: NR (n = 28)	Median: 54 μg/L, 95 %: NR (n = 26)
				Postpartum: 12 months	Median: 56 μg/L, 95 %: NR (n = NR)	Median: 54 μg/L, 95 %: NR (n = NR)	Median: 41 μg/L, 95 %: 31‐74 (n = 27)	Median: 28 μg/L, 95 %: 19‐46 (n = 26)
				Neonatal: 5 days	Median: 64 μg/L, 95 %: 34‐70 (n = 27)	Median: 27 μg/L, 95 %: 21‐56 (n = 25)
Silva 1981	Eastern Greater Santiago, Chile	Pregnancy	Iodine: daily drops approximately 300 µg iodine, control: no intervention	Pregnancy: 15‐40 weeks	Mean: 376 μg iodine/g creatinine, SD: 456 (n = 36)	Mean: 56 μg iodine/g creatinine, SD: 25 (n = 10)
Silva 1981	Eastern Greater Santiago, Chile	Pregnancy		Postpartum: at delivery	Mean: 493 μg iodine/g creatinine, SD: 586 (n = not specified)	Mean: 127 μg iodine/g creatinine, SD: 94 (n = not specified)
Zhou 2015	Adelaide, New Zealand	Pregnancy	Iodine: daily tablet 150 μg iodine, control: placebo	Postpartum: 6 weeks			Median (IQR): 106.0 (84.0‐146.0), n = 19	Median (IQR): 124.0 (76.0‐155.0), n = 22
Gowachirapant 2014	Bangkok, Thailand	Pregnancy	Iodine: daily 200 µg iodine and multivitamin‐mineral tablets, control: multivitamin‐mineral tablets with no iodine	Pregnancy: 30 weeks	Median: 233.3 μg/L, IQR: 140.0, 314.8 (n = 171)	Median: 154.6 μg/L, IQR: 107.9, 212.4 (n = 185)
				Postpartum: 6 weeks	Median: 115.5 μg/L, IQR: 73.0, 155.3 (n = 129)	Median: 93.6 μg/L, IQR: 60.3, 159.8 (n = 135)
				Neonatal: 6 weeks	Median: 168.7 μg/L, IQR: 97.7, 305.7 (n = 83)	Median: 193.3 μg/L, IQR: 136.7, 303.5 (n = 76)
IQR: interquartile range KI: potassium iodide NR: not reported SD: standard deviation SEM: standard error of the mean

Table 1. Urinary iodine concentration and breast milk iodine concentration following prepregnancy, pregnancy or postpartum iodine supplementation

Table 2. Maternal thyroid volume ‐ pregnancy

Study	Thyroid volume (mL)
Study	Iodine	No iodine
Brucker‐Davis 2013¹	Median: 9.73, range: 5.4–15.5 (n = 30)	Median: 10.13, range: 4.2–25.5 (n = 46)
Pedersen 1993²	Median: 11.1 (n = 28)	Median: 12.4 (n = 26)
Gowachirapant 2014³	Median: 8.0, IQR: 7.0, 9.1 (n = 176)	Median: 8.1, IQR: 7.1, 9.4 (n = 190)
¹Assessed in the 3rd trimester. ²Assessed at 37 weeks' gestation. Median estimated from figure. No estimate of range or distribution was provided. Total was not provided so was assumed. ³Assessed at 30 weeks' gestation. Intention‐to‐treat analysis.

Table 2. Maternal thyroid volume ‐ pregnancy

Table 3. Maternal thyroid volume ‐ postpartum

Study	Thyroid volume (mL)
Study	Iodine	No iodine
Brucker‐Davis 2013¹	Median: 8.5, range: 3.8–13.8 (n = 18)	Median: 9.7, range: 6.6–18.9 (n = 18)
Liesenkotter 1996²	Mean: 20.5, SD: 8.5 (n = 38)	Mean: 20.0, SD: 10.5 (n = 70)
Pedersen 1993³	Median: 10.2 (n = 25)	Median: 10 (n = 24)
Gowachirapant 2014⁴	Median: 7.3, IQR: 5.8, 9.0 (n = 129)	Median: 7.2, IQR: 6.1, 8.4 (n = 135)
¹Assessed at 3 months postpartum. ²Assessed by ultrasound on average at 11 days postpartum. Mean and SD extracted from figure. ³Assessed at 12 months postpartum. Median estimated from figure. No estimate of range or distribution was provided. Total was not provided so was assumed. ⁴Assessed at 6 weeks postpartum.

Table 3. Maternal thyroid volume ‐ postpartum

Table 4. Maternal thyroglobulin ‐ pregnancy

Study	Thyroglobulin (in µg/L)
Study	Iodine	No iodine
Brucker‐Davis 2013¹	Median: 17.5, range: 2.5–56.2 (n = 30)	Median: 19.6, range: 3.8–98.6 (n = 46)
Nohr 2000²	Median: 14.1, IQR: 5.0, 21.5 (n = 42)	Median: 19.4, IQR: 8.2, 33.5 (n = 24)
Pedersen 1993³	Median: 9.2 (n = 28)	Median: 16.7 (n = 26)
Gowachirapant 2014⁴	Median: 9.78, IQR: 5.46, 18.10 (n = 176)	Median 11.90, IQR: 6.35, 22.30 (n = 190)
¹Assessed in the 3rd trimester. ²Assessed at 35 weeks' gestation. ³Assessed at 37 weeks' gestation. Median estimated from figure. No estimate of range or distribution was provided. Total was not provided so was assumed. ⁴Assessed at 30 weeks' gestation.

Table 4. Maternal thyroglobulin ‐ pregnancy

Table 5. Maternal thyroglobulin ‐ postpartum

Study	Thyroglobulin (in µg/L)
Study	Iodine	No iodine
Brucker‐Davis 2013¹	Median: 11.5, range: 4.4–37 (n = 18)	Median: 23.1, range: 3.0–45.5 (n = 18)
Liesenkotter 1996²	Mean: 8.3, SD: 10.9 (n = 38)	Mean: 13.5, SD: 19.3 (n = 70)
Pedersen 1993³	Median: 4.7 (n = 24)	Median: 10.3 (n = 24)
Gowachirapant 2014⁴	Median: 7.94, IQR: 4.89, 15.30 (n = 129)	Median: 9.51, IQR: 4.79, 15.90 (n = 135)
¹Assessed at 3 months postpartum. ²Assessed on average at 11 days postpartum. ³Assessed at 12 months postpartum. Median estimated from figure. No estimate of range or distribution was provided. Total was not provided so was assumed. ⁴Assessed at 6 weeks postpartum.

Table 5. Maternal thyroglobulin ‐ postpartum

Table 6. Neonatal behavioural assessment scales scores assessed at 6 weeks postpartum (Zimmerman 2009)

Outcome	Iodine			Control
Outcome	Median	IQR	n	Median	IQR	n
Habituation	7.0	(5.4, 8.3)	22	7.6	(6.8, 8.1)	18
Social‐interactive	7.4	(6.4, 8.0)	66	6.9	(6.0, 8.1)	68
Motor system	6.0	(5.3, 6.6)	73	5.8	(5.2, 6.4)	72
State organization	3.8	(2.8, 4.0)	48	3.8	(3.0, 4.1)	53
State regulation	4.0	(3.3, 4.5)	38	4.0	(2.6, 4.7)	40
Autonomic system	6.7	(6.0, 7.3)	61	6.7	(5.3, 7.3)	71
Reflexes	3.0	(1.8, 4.0)	74	3.0	(2.0, 4.0)	75
Supplementary items	6.7	(5.3, 8.3)	48	6.3	(5.1, 8.1)	55
IQR: interquartile range

Table 6. Neonatal behavioural assessment scales scores assessed at 6 weeks postpartum (Zimmerman 2009)

Comparison 1. Any supplement containing iodine versus same supplement without iodine or no intervention/placebo

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Maternal hypothyroidism ‐ pregnancy Show forest plot	1	365	Risk Ratio (M‐H, Random, 95% CI)	1.9 [0.57, 6.38]

2 Maternal hypothyroidism ‐ postpartum Show forest plot	3	540	Risk Ratio (M‐H, Random, 95% CI)	0.44 [0.06, 3.42]

3 Preterm birth Show forest plot	2	376	Risk Ratio (M‐H, Random, 95% CI)	0.71 [0.30, 1.66]

4 Maternal adverse effect: elevated thyroid peroxidase antibodies ‐ pregnancy Show forest plot	1	359	Risk Ratio (M‐H, Random, 95% CI)	0.95 [0.44, 2.07]

5 Maternal adverse effect: elevated thyroid peroxidase antibodies ‐ postpartum Show forest plot	3	397	Risk Ratio (M‐H, Random, 95% CI)	1.01 [0.78, 1.30]

6 Maternal adverse effect: hyperthyroidism ‐ pregnancy Show forest plot	1	365	Risk Ratio (M‐H, Random, 95% CI)	1.9 [0.57, 6.38]

7 Maternal adverse effect: hyperthyroidism ‐ postpartum Show forest plot	3	543	Risk Ratio (M‐H, Random, 95% CI)	0.32 [0.11, 0.91]

8 Maternal adverse effect: digestive intolerance ‐ pregnancy Show forest plot	1	76	Risk Ratio (M‐H, Random, 95% CI)	15.33 [2.07, 113.70]

9 Perinatal mortality Show forest plot	2	457	Risk Ratio (M‐H, Random, 95% CI)	0.66 [0.42, 1.03]

10 Low birthweight Show forest plot	2	377	Risk Ratio (M‐H, Random, 95% CI)	0.56 [0.26, 1.23]

11 Neonatal hypothyroidism or elevated TSH Show forest plot	2	260	Risk Ratio (M‐H, Random, 95% CI)	0.58 [0.11, 3.12]

12 Neonatal adverse effect: elevated thyroid peroxidase antibodies Show forest plot	1	108	Risk Ratio (M‐H, Random, 95% CI)	0.61 [0.07, 5.70]

13 Spontaneous miscarriage Show forest plot	3	645	Risk Ratio (M‐H, Random, 95% CI)	1.31 [0.64, 2.69]

14 Maternal goitre ‐ pregnancy Show forest plot	2	486	Risk Ratio (M‐H, Random, 95% CI)	1.00 [0.33, 3.06]

15 Maternal goitre ‐ postpartum Show forest plot	2	370	Risk Ratio (M‐H, Random, 95% CI)	1.17 [0.64, 2.16]

16 Maternal insufficient iodine intake ‐ pregnancy Show forest plot	2	432	Risk Ratio (M‐H, Random, 95% CI)	0.64 [0.51, 0.80]

17 Maternal insufficient iodine intake ‐ postpartum Show forest plot	2	425	Risk Ratio (M‐H, Random, 95% CI)	0.81 [0.70, 0.93]

18 Maternal excessive iodine intake ‐ pregnancy Show forest plot	1	356	Risk Ratio (M‐H, Random, 95% CI)	4.33 [1.24, 15.07]

19 Small‐for‐gestational age Show forest plot	2	377	Risk Ratio (M‐H, Random, 95% CI)	1.26 [0.77, 2.05]

20 Congenital anomalies Show forest plot	2	875	Risk Ratio (M‐H, Random, 95% CI)	0.27 [0.12, 0.60]

21 Neonatal goitre Show forest plot	3	684	Risk Ratio (M‐H, Random, 95% CI)	0.11 [0.02, 0.56]

22 Neonatal thyroid volume (in mL) Show forest plot	3	359	Mean Difference (IV, Random, 95% CI)	‐0.34 [‐0.58, ‐0.11]

23 Neonatal insufficient iodine intake Show forest plot	1	159	Risk Ratio (M‐H, Random, 95% CI)	2.14 [1.04, 4.37]

24 Child mental or motor development (IQ points) Show forest plot	2	174	Mean Difference (IV, Random, 95% CI)	11.21 [7.96, 14.46]

25 Child mental or motor development (cognitive score) Show forest plot	1	53	Mean Difference (IV, Random, 95% CI)	‐2.30 [‐7.88, 3.28]

26 Child mental or motor development (language score) Show forest plot	1	53	Mean Difference (IV, Random, 95% CI)	‐0.70 [‐7.08, 5.68]

27 Child mental or motor development (motor score) Show forest plot	1	53	Mean Difference (IV, Random, 95% CI)	1.5 [‐4.02, 7.02]

28 Child mental or motor development (social‐emotional score) Show forest plot	1	53	Mean Difference (IV, Random, 95% CI)	0.40 [‐8.25, 9.05]

29 Child mental or motor development (adaptive behaviour score) Show forest plot	1	53	Mean Difference (IV, Random, 95% CI)	1.70 [‐6.40, 9.80]

30 Child mental or motor development (cognitive score < 85) Show forest plot	1	53	Risk Ratio (M‐H, Random, 95% CI)	2.89 [0.12, 67.96]

31 Child mental or motor development (cognitive score < 70) Show forest plot	1	53	Risk Ratio (M‐H, Random, 95% CI)	2.89 [0.12, 67.96]

32 Child mental or motor development (language score < 85) Show forest plot	1	53	Risk Ratio (M‐H, Random, 95% CI)	0.96 [0.21, 4.35]

33 Child mental or motor development (language score < 70) Show forest plot	1	53	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]

34 Child mental or motor development (motor score < 85) Show forest plot	1	53	Risk Ratio (M‐H, Random, 95% CI)	0.39 [0.08, 1.81]

35 Child mental or motor development (motor score < 70) Show forest plot	1	53	Risk Ratio (M‐H, Random, 95% CI)	2.89 [0.12, 67.96]

Comparison 1. Any supplement containing iodine versus same supplement without iodine or no intervention/placebo

Comparison 2. Any oral iodine supplement versus same supplement without iodine or no intervention/placebo

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Maternal hypothyroidism ‐ pregnancy Show forest plot	1	365	Risk Ratio (M‐H, Random, 95% CI)	1.9 [0.57, 6.38]

2 Maternal hypothyroidism ‐ postpartum Show forest plot	3	540	Risk Ratio (M‐H, Random, 95% CI)	0.44 [0.06, 3.42]

3 Preterm birth Show forest plot	2	376	Risk Ratio (M‐H, Random, 95% CI)	0.71 [0.30, 1.66]

4 Maternal adverse effect: elevated thyroid peroxidase antibodies ‐ pregnancy Show forest plot	1	359	Risk Ratio (M‐H, Random, 95% CI)	0.95 [0.44, 2.07]

5 Maternal adverse effect: elevated thyroid peroxidase antibodies ‐ postpartum Show forest plot	3	397	Risk Ratio (M‐H, Random, 95% CI)	1.01 [0.78, 1.30]

6 Maternal adverse effect: hyperthyroidism ‐ pregnancy Show forest plot	1	365	Risk Ratio (M‐H, Random, 95% CI)	1.9 [0.57, 6.38]

7 Maternal adverse effect: hyperthyroidism ‐ postpartum Show forest plot	3	543	Risk Ratio (M‐H, Random, 95% CI)	0.32 [0.11, 0.91]

8 Maternal adverse effect: digestive intolerance ‐ pregnancy Show forest plot	1	76	Risk Ratio (M‐H, Random, 95% CI)	15.33 [2.07, 113.70]

9 Perinatal mortality Show forest plot	1	58	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]

10 Low birthweight Show forest plot	2	377	Risk Ratio (M‐H, Random, 95% CI)	0.56 [0.26, 1.23]

11 Neonatal hypothyroidism or elevated TSH Show forest plot	2	260	Risk Ratio (M‐H, Random, 95% CI)	0.58 [0.11, 3.12]

12 Neonatal adverse effect: elevated thyroid peroxidase antibodies Show forest plot	1	108	Risk Ratio (M‐H, Random, 95% CI)	0.61 [0.07, 5.70]

13 Spontaneous miscarriage Show forest plot	3	645	Risk Ratio (M‐H, Random, 95% CI)	1.31 [0.64, 2.69]

14 Maternal goitre ‐ pregnancy Show forest plot	2	486	Risk Ratio (M‐H, Random, 95% CI)	1.00 [0.33, 3.06]

15 Maternal goitre ‐ postpartum Show forest plot	2	370	Risk Ratio (M‐H, Random, 95% CI)	1.17 [0.64, 2.16]

16 Maternal insufficient iodine intake ‐ pregnancy Show forest plot	2	432	Risk Ratio (M‐H, Random, 95% CI)	0.64 [0.51, 0.80]

17 Maternal insufficient iodine intake ‐ postpartum Show forest plot	2	425	Risk Ratio (M‐H, Random, 95% CI)	0.81 [0.70, 0.93]

18 Maternal excessive iodine intake ‐ pregnancy Show forest plot	1	356	Risk Ratio (M‐H, Random, 95% CI)	4.33 [1.24, 15.07]

19 Small‐for‐gestational age Show forest plot	2	377	Risk Ratio (M‐H, Random, 95% CI)	1.26 [0.77, 2.05]

20 Congenital anomalies Show forest plot	1	57	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]

21 Neonatal goitre Show forest plot	2	228	Risk Ratio (M‐H, Random, 95% CI)	0.11 [0.02, 0.56]

22 Neonatal thyroid volume (in mL) Show forest plot	3	359	Mean Difference (IV, Random, 95% CI)	‐0.34 [‐0.58, ‐0.11]

23 Neonatal insufficient iodine intake Show forest plot	1	159	Risk Ratio (M‐H, Random, 95% CI)	2.14 [1.04, 4.37]

24 Child mental or motor development (cognitive score) Show forest plot	1	53	Mean Difference (IV, Random, 95% CI)	‐2.30 [‐7.88, 3.28]

25 Child mental or motor development (language score) Show forest plot	1	53	Mean Difference (IV, Random, 95% CI)	‐0.70 [‐7.08, 5.68]

26 Child mental or motor development (motor score) Show forest plot	1	53	Mean Difference (IV, Random, 95% CI)	1.5 [‐4.02, 7.02]

27 Child mental or motor development (social‐emotional score) Show forest plot	1	53	Mean Difference (IV, Random, 95% CI)	0.40 [‐8.25, 9.05]

28 Child mental or motor development (adaptive behaviour score) Show forest plot	1	53	Mean Difference (IV, Random, 95% CI)	1.70 [‐6.40, 9.80]

29 Child mental or motor development (cognitive score < 85) Show forest plot	1	53	Risk Ratio (M‐H, Random, 95% CI)	2.89 [0.12, 67.96]

30 Child mental or motor development (cognitive score < 70) Show forest plot	1	53	Risk Ratio (M‐H, Random, 95% CI)	2.89 [0.12, 67.96]

31 Child mental or motor development (language score < 85) Show forest plot	1	53	Risk Ratio (M‐H, Random, 95% CI)	0.96 [0.21, 4.35]

32 Child mental or motor development (language score < 70) Show forest plot	1	53	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]

33 Child mental or motor development (motor score < 85) Show forest plot	1	53	Risk Ratio (M‐H, Random, 95% CI)	0.39 [0.08, 1.81]

34 Child mental or motor development (motor score < 70) Show forest plot	1	53	Risk Ratio (M‐H, Random, 95% CI)	2.89 [0.12, 67.96]

Comparison 2. Any oral iodine supplement versus same supplement without iodine or no intervention/placebo

Comparison 3. Oral iodine‐only supplement versus no intervention or placebo

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Maternal hypothyroidism ‐ postpartum Show forest plot	3	340	Risk Ratio (M‐H, Random, 95% CI)	0.91 [0.10, 8.60]

2 Preterm birth Show forest plot	1	58	Risk Ratio (M‐H, Random, 95% CI)	1.25 [0.37, 4.19]

3 Maternal adverse effect: elevated thyroid peroxidase antibodies ‐ postpartum Show forest plot	2	162	Risk Ratio (M‐H, Random, 95% CI)	0.62 [0.15, 2.51]

4 Maternal adverse effect: hyperthyroidism ‐ postpartum Show forest plot	2	286	Risk Ratio (M‐H, Random, 95% CI)	0.29 [0.01, 7.06]

5 Perinatal mortality Show forest plot	1	58	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]

6 Low birthweight Show forest plot	1	58	Risk Ratio (M‐H, Fixed, 95% CI)	0.33 [0.04, 3.02]

7 Neonatal hypothyroidism or elevated TSH Show forest plot	1	41	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]

8 Neonatal adverse effect: elevated thyroid peroxidase antibodies Show forest plot	1	108	Risk Ratio (M‐H, Random, 95% CI)	0.61 [0.07, 5.70]

9 Spontaneous miscarriage Show forest plot	1	58	Risk Ratio (M‐H, Random, 95% CI)	3.0 [0.13, 70.74]

10 Maternal goitre ‐ pregnancy Show forest plot	1	120	Risk Ratio (M‐H, Random, 95% CI)	0.6 [0.23, 1.55]

11 Maternal goitre ‐ postpartum Show forest plot	1	108	Risk Ratio (M‐H, Random, 95% CI)	1.01 [0.73, 1.39]

12 Maternal insufficient iodine intake ‐ postpartum Show forest plot	1	175	Risk Ratio (M‐H, Random, 95% CI)	0.81 [0.69, 0.96]

13 Small‐for‐gestational age Show forest plot	1	58	Risk Ratio (M‐H, Fixed, 95% CI)	1.5 [0.27, 8.32]

14 Congenital anomalies Show forest plot	1	57	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]

15 Neonatal goitre Show forest plot	2	228	Risk Ratio (M‐H, Random, 95% CI)	0.11 [0.02, 0.56]

16 Neonatal thyroid volume (in mL) Show forest plot	2	228	Mean Difference (IV, Random, 95% CI)	‐0.53 [‐1.02, ‐0.03]

17 Child mental or motor development (cognitive score) Show forest plot	1	53	Mean Difference (IV, Random, 95% CI)	‐2.30 [‐7.88, 3.28]

18 Child mental or motor development (language score) Show forest plot	1	53	Mean Difference (IV, Random, 95% CI)	‐0.70 [‐7.08, 5.68]

19 Child mental or motor development (motor score) Show forest plot	1	53	Mean Difference (IV, Random, 95% CI)	1.5 [‐4.02, 7.02]

20 Child mental or motor development (social‐emotional score) Show forest plot	1	53	Mean Difference (IV, Random, 95% CI)	0.40 [‐8.25, 9.05]

21 Child mental or motor development (adaptive behaviour score) Show forest plot	1	53	Mean Difference (IV, Random, 95% CI)	1.70 [‐6.40, 9.80]

22 Child mental or motor development (cognitive score < 85) Show forest plot	1	53	Risk Ratio (M‐H, Random, 95% CI)	2.89 [0.12, 67.96]

23 Child mental or motor development (cognitive score < 70) Show forest plot	1	53	Risk Ratio (M‐H, Random, 95% CI)	2.89 [0.12, 67.96]

24 Child mental or motor development (language score < 85) Show forest plot	1	53	Risk Ratio (M‐H, Random, 95% CI)	0.96 [0.21, 4.35]

25 Child mental or motor development (language score < 70) Show forest plot	1	53	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]

26 Child mental or motor development (motor score < 85) Show forest plot	1	53	Risk Ratio (M‐H, Random, 95% CI)	0.39 [0.08, 1.81]

27 Child mental or motor development (motor score < 70) Show forest plot	1	53	Risk Ratio (M‐H, Random, 95% CI)	2.89 [0.12, 67.96]

Comparison 3. Oral iodine‐only supplement versus no intervention or placebo

Comparison 4. Oral iodine supplement with other vitamins and/or minerals versus only other vitamins and/or minerals but no iodine

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Maternal hypothyroidism ‐ pregnancy Show forest plot	1	365	Risk Ratio (M‐H, Random, 95% CI)	1.9 [0.57, 6.38]

2 Maternal hypothyroidism ‐ postpartum Show forest plot	1	254	Risk Ratio (M‐H, Random, 95% CI)	0.97 [0.06, 15.32]

3 Preterm birth Show forest plot	1	318	Risk Ratio (M‐H, Random, 95% CI)	0.51 [0.23, 1.14]

4 Maternal adverse effect: elevated thyroid peroxidase antibodies ‐ pregnancy Show forest plot	1	359	Risk Ratio (M‐H, Random, 95% CI)	0.95 [0.44, 2.07]

5 Maternal adverse effect: elevated thyroid peroxidase antibodies ‐ postpartum Show forest plot	1	235	Risk Ratio (M‐H, Random, 95% CI)	1.03 [0.79, 1.33]

6 Maternal adverse effect: hyperthyroidism ‐ pregnancy Show forest plot	1	365	Risk Ratio (M‐H, Random, 95% CI)	1.9 [0.57, 6.38]

7 Maternal adverse effect: hyperthyroidism ‐ postpartum Show forest plot	1	257	Risk Ratio (M‐H, Random, 95% CI)	0.35 [0.12, 1.06]

8 Maternal adverse effect: digestive intolerance ‐ pregnancy Show forest plot	1	76	Risk Ratio (M‐H, Random, 95% CI)	15.33 [2.07, 113.70]

9 Low birthweight Show forest plot	1	319	Risk Ratio (M‐H, Random, 95% CI)	0.61 [0.27, 1.39]

10 Neonatal hypothyroidism or elevated TSH Show forest plot	1	219	Risk Ratio (M‐H, Random, 95% CI)	0.58 [0.11, 3.12]

11 Spontaneous miscarriage Show forest plot	2	587	Risk Ratio (M‐H, Random, 95% CI)	1.26 [0.60, 2.62]

12 Maternal goitre ‐ pregnancy Show forest plot	1	366	Risk Ratio (M‐H, Random, 95% CI)	1.89 [0.56, 6.34]

13 Maternal goitre ‐ postpartum Show forest plot	1	262	Risk Ratio (M‐H, Random, 95% CI)	2.06 [0.64, 6.68]

14 Maternal insufficient iodine intake ‐ pregnancy Show forest plot	2	432	Risk Ratio (M‐H, Random, 95% CI)	0.64 [0.51, 0.80]

15 Maternal insufficient iodine intake ‐ postpartum Show forest plot	1	250	Risk Ratio (M‐H, Random, 95% CI)	0.80 [0.61, 1.04]

16 Maternal excessive iodine intake ‐ pregnancy Show forest plot	1	356	Risk Ratio (M‐H, Random, 95% CI)	4.33 [1.24, 15.07]

17 Small‐for‐gestational age Show forest plot	1	319	Risk Ratio (M‐H, Random, 95% CI)	1.24 [0.74, 2.06]

18 Neonatal thyroid volume (in mL) Show forest plot	1	131	Mean Difference (IV, Random, 95% CI)	‐0.09 [‐0.18, ‐0.00]

19 Neonatal insufficient iodine intake Show forest plot	1	159	Risk Ratio (M‐H, Random, 95% CI)	2.14 [1.04, 4.37]

Comparison 4. Oral iodine supplement with other vitamins and/or minerals versus only other vitamins and/or minerals but no iodine

Comparison 5. Any injected iodine supplement versus same supplement without iodine or no treatment/placebo

Outcome or subgroup title	No. of studies	No. of participants	Statistical method	Effect size
1 Perinatal mortality Show forest plot	1	399	Risk Ratio (M‐H, Random, 95% CI)	0.66 [0.42, 1.03]

2 Congenital anomalies Show forest plot	1	818	Risk Ratio (M‐H, Random, 95% CI)	0.27 [0.12, 0.60]

3 Neonatal goitre Show forest plot	1	456	Risk Ratio (M‐H, Random, 95% CI)	0.0 [0.0, 0.0]

4 Child mental or motor development (IQ points) Show forest plot	2	174	Mean Difference (IV, Random, 95% CI)	11.21 [7.96, 14.46]

Comparison 5. Any injected iodine supplement versus same supplement without iodine or no treatment/placebo