Main

Subclinical hypothyroidism is characterized by serum TSH levels above the upper limit of the reference range, in the presence of normal serum concentrations of total T4 and free T4 (Ref. 1). This condition is also known as isolated hyperthyrotropinaemia2 or biochemically compensated primary hypothyroidism.

The serum TSH concentration recorded among euthyroid individuals falls within a distinct range, from 0.4–0.5 mIU/l at the lower limit to 4.0–5.0 mIU/l at the upper limit. This reference (or 'normal') range is to some extent dependent on the method used to measure TSH, with large variations found between different TSH assays. In addition, isolated increases in TSH levels can be a transient finding owing to between-laboratory or intra-individual variations3. Consequently, a diagnosis of subclinical hypothyroidism should be made only when at least two independent TSH measurements fall above the upper limit of the reference range. Depending on the degree of serum TSH elevation, subclinical hypothyroidism can be defined as mild (TSH 4.5–10 mIU/l) or severe (TSH >10 mIU/l)1.

Increases in serum TSH levels might be indicative of mild hypothyroidism and, therefore, of a reduced effect of thyroid hormones on the peripheral tissues. Among adults, subclinical hypothyroidism shows a clear tendency to progress to overt hypothyroidism; furthermore, this condition is associated with several adverse effects, including dyslipidaemia4,5, insulin resistance6, diastolic dysfunction7, endothelial dysfunction7,8,9, coronary heart disease and heart failure10,11,12,13. Therefore, treatment with levothyroxine is recommended for adults with serum TSH levels >10 mIU/l or among individuals with TSH levels <10 mIU/l in the presence of symptoms suggestive of hypothyroidism14,15.

During infancy and childhood, functional abnormalities of the thyroid gland can adversely affect growth and brain maturation, with clinical consequences depending on age and the severity of thyroid impairment. If not adequately treated with levothyroxine, children with overt hypothyroidism are at increased risk of mental retardation, metabolic abnormalities and impairments in growth and skeletal maturation. By contrast, any association between subclinical hypothyroidism in childhood and adverse health outcomes remains controversial. The majority of studies indicate that subclinical hypothyroidism is a benign and remitting condition among children, with a low risk of progression to overt thyroid dysfunction16. Nonetheless, some evidence suggests that modest increases in the levels of TSH might be associated with subtle metabolic abnormalities in childhood17.

As in adults, there is a general agreement in recommending treatment for children who have serum TSH levels >10 mIU/l; however, the optimum management of the mild form of subclinical hypothyroidism remains a matter of debate18,19,20. Clearly, the decision on how to manage and treat children with subclinical hypothyroidism should primarily depend on the likelihood of thyroid function worsening over time and on the adverse effects associated with persistent subclinical hypothyroidism.

This Review will focus on the diagnosis, natural history and management of subclinical hypothyroidism among children. The latest data on health outcomes and ongoing scientific research on metabolic issues and the risk of developing cardiovascular disease will be summarized. Finally, the controversy regarding treatment will be discussed.

Prevalence

Subclinical hypothyroidism is a common occurrence among adults, with a prevalence of 4–20% and a tendency to progress to overt hypothyroidism7. The most frequent cause of subclinical hypothyroidism in adulthood is Hashimoto thyroiditis, with >50% of affected individuals testing positive for the presence of thyroid peroxidase (TPO) antibodies7.

Few data are available from epidemiological studies conducted among children and adolescents. Data from NHANES III indicated a prevalence of 1.7% for subclinical hypothyroidism among 1,327 adolescents aged 13–16 years21. Moreover, assessment of TSH and free T4 measurements in a database of 121,052 children aged 0.5–16.0 years suggested that mild subclinical hypothyroidism (TSH level, 5.5–10.0 mIU/l) might be observed during routine assessment of thyroid function among 2.9% of children22.

Aetiology

Several factors are implicated in the development of subclinical hypothyroidism during childhood (Box 1).

Hashimoto thyroiditis

Hashimoto thyroiditis is one of the most frequent causes of persistent subclinical hypothyroidism, particularly among older children and adolescents (age range, 8–18 years)23. Individuals with genetic conditions, such as Down syndrome or Turner syndrome24, are also susceptible to subclinical hypothyroidism associated with Hashimoto thyroiditis, as are those with other autoimmune diseases (for example, coeliac disease or type 1 diabetes mellitus)24,25. Patients with Hashimoto thyroiditis might develop subclinical or overt hypothyroidism, depending on the severity of the immunological damage24.

Persistent neonatal hyperthyrotropinaemia

During the past 10–15 years, increased sensitivity of the TSH assay — coupled with the intention to improve the diagnosis of mild forms of congenital hypothyroidism — have led to a decrease in the TSH threshold for neonatal screening. Since the 2000s, many countries have adopted a TSH cut-off value of <15 mIU/l, albeit with great variation (range, 5–12 mIU/l) even within different regions of the same country26,27. The use of lowered TSH cut-off values has resulted in a progressively increased incidence of both mild and potentially transient forms of thyroid dysfunction, including isolated hyperthyrotropinaemia28.

The optimum management of neonates with an isolated increase of their TSH levels is debated. In a prospective study, thyroid function and morphology were longitudinally evaluated among 44 children with a mildly elevated TSH level at neonatal screening29. Progressive normalization of TSH was observed with increasing age for the majority of these children (68%), whereas mildly increased TSH levels (4.1–8.2 mIU/l) persisted among 32% of the participants at 7.2–9.5 years of age. Most of the children with persistent subclinical hypothyroidism exhibited morphological defects (thyroid hypoplasia or hemiagenesis) or genetic abnormalities (mainly mutations and polymorphisms of the TSH receptor (TSHR)). Unfortunately, the neurocognitive outcomes of these children were not assessed29. Little information is available on mild neonatal elevation of TSH levels and cognitive outcome; thus, whether mild forms of subclinical hypothyroidism require treatment with levothyroxine to attain full intellectual potential remains to be determined26.

Gene defects and thyroid dysfunction

TSH receptor mutations. Heterozygous mutations in the gene encoding TSHR have been documented in 11.4–29.0% of children with nonautoimmune subclinical hypothyroidism30,31,32. Loss-of-function mutations are associated with a wide spectrum of thyroid dysfunction, ranging from severe congenital hypothyroidism with thyroid hypoplasia to a compensated state of mild hyperthyrotropinaemia33,34,35. This phenotypic variability depends on the degree of TSHR impairment, with homozygous mutations exerting a more severe phenotype than heterozygous mutations33, and on the number of mutated alleles34. Moreover, a differing severity of thyroid dysfunction might be observed for the same TSHR mutation among members of the same family35. To date, >60 different TSHR mutations have been reported in association with nonautoimmune hyperthyrotropinaemia35.

Dual oxidase 2 mutations. Mutations in the gene encoding dual oxidase 2 (DUOX2) are associated with partial iodide organification defects; these mutations have been also reported to cause mild persistent hyperthyrotropinaemia28,36.

Genetic syndromes

Persistent subclinical hypothyroidism is a frequent finding in several genetic syndromes.

Down syndrome. The association between thyroid dysfunction and Down syndrome is well-recognized but its natural history has yet to be defined37. A high prevalence of subclinical hypothyroidism has been reported among children with Down syndrome, ranging from 25.3% to 60.0%38.

The underlying cause of the isolated elevations in TSH levels among children with Down syndrome has not been elucidated. In addition to thyroid autoimmunity, possible mechanisms include a central disorder causing inappropriate release of TSH; production of TSH with lowered activity; and some degree of TSH insensitivity in the thyroid gland39. TSH bioactivity seems to be normal among affected individuals40. Furthermore, the genes encoding TSHR and the α subunit of the stimulatory G protein (Gsα) do not seem to harbour mutations that might be involved in resistance to TSH41. Further studies are, therefore, required to unravel the causes of mild thyroidal dysfunction among patients with Down syndrome.

Subclinical hypothyroidism associated with Down syndrome seems to be frequently self-limiting (that is, without the need for treatment). In a longitudinal cohort study, thyroid function was initially evaluated among 122 children with Down syndrome and then repeated among 103 of the participants 4–6 years later39. At the first testing, subclinical hypothyroidism was found among 20 patients. By contrast, TSH levels had normalized among 14 of these patients at the second testing, whereas hyperthyrotropinaemia persisted among five patients and only one patient developed hypothyroidism.

Although it is clear that thyroid function should be evaluated among children with Down syndrome, the optimum frequency of biochemical screening is still debated. Nonetheless, frequent testing seems not to be justified in this population. The effectiveness of levothyroxine administration among children with Down syndrome and subtle thyroid abnormalities is also a controversial issue that requires validation in randomized, double blind, controlled studies37.

Pseudohypoparathyroidism type 1a. Pseudo-hypoparathyroidism type 1a is a rare heterogeneous genetic disorder caused by deficiency of Gsα. This protein is a key regulator of the cAMP signalling pathway; deficiency leads to resistance to multiple hormones, including parathyroid hormone and TSH42. Pseudohypoparathyroidism type 1a can be associated with increased levels of TSH. The main clinical features include short stature, obesity, round face, shortening of the fingers and toes (brachydactyly), subcutaneous ossifications and intellectual impairment42.

Iodine deficiency

Iodine is an essential micronutrient for the production of thyroid hormones. Chronic low iodine intake might result in mild-to-severe subclinical hypothyroidism, as well as goitre and overt hypothyroidism43.

Medications

Several pharmacological agents can induce both subclinical and overt hypothyroidism; however, the mechanisms underlying these alterations are not completely understood.

Iodine-containing drugs. Drugs that contain iodine, such as amiodarone and its main metabolite desethylamiodarone, might block the ability of the type 2 iodothyronine deiodinase to mediate conversion of T4 to T3. Inhibition of type 2 iodothyronine deiodinase activity weakens feedback mediated by T4 at the level of the pituitary gland, thus elevating plasma concentrations of TSH44.

IFN-α. Treatment with IFN-α can induce thyroid dysfunction via immune mechanisms that result in the production of anti-thyroid antibodies45. These antibodies either promote destruction of the thyroid parenchyma or exacerbate pre-existing thyroid autoimmunity. In addition, IFN-α might induce thyroid dysfunction by a direct toxic effect on thyroid cells45.

Antiepileptic drugs. The use of antiepileptic drugs, such as phenobarbital, phenytoin, carbamazepine and valproic acid might affect thyroid function46. The underlying mechanisms for thyroid dysfunction with antiepileptic drugs are not completely understood. Enhancement of hepatic microsomal enzyme systems, accelerating clearance of thyroid hormones or interference with the regulation of pituitary TSH secretion by TSH-releasing hormone are possible mechanisms46,47.

Exposure to ionizing radiation

Subclinical hypothyroidism is increased among children following either therapeutic or environmental exposure to ionizing radiation.

Therapeutic exposure. A long-term follow-up study of thyroid function among patients who had received irradiation before undergoing bone-marrow transplantation during childhood and adolescence reported a 26.5% incidence of subclinical hypothyroidism48. Furthermore, thyroid dysfunction occurred most frequently among the participants who were aged younger than 9 years at the time of radiation therapy. Subclinical hypothyroidism can occur several years after therapeutic irradiation but resolves spontaneously in the majority of cases48.

Environmental exposure. Nuclear accidents represent the major source of information on the effects of environmental exposure to ionizing radiation on thyroid function49. For instance, childhood exposure to iodine-131 from the accident at the Chernobyl nuclear plant has been associated with the development of thyroid cancer49. Moreover, two large screening studies conducted on 11,853 children from Ukraine and 10,827 children and adolescents from Belarus found an association between iodine-131 exposure and the development of thyroid dysfunction, including subclinical hypothyroidism49,50.

Obesity

Subclinical hypothyroidism is a frequent finding among obese children, with moderate elevations in TSH concentrations documented among 10–23% of all such individuals51. Hyperthyrotropinaemia seems to be a consequence, rather than a cause, of weight gain. Accordingly, several studies have documented normalization of TSH levels after participation in weight-loss programmes52,53,54,55. A study of 1,010 children with obesity aged 4.5–15.9 years detected autoimmune subclinical hypothyroidism among 7.0% of the participants, whereas isolated hyperthyrotropinaemia was observed among 12.8%52.

The mechanisms underlying isolated hyperthyrotropinaemia among children with obesity are far from clear and several hypotheses have been put forward. One model suggests that increased levels of TSH represent a process of adaptation to increased energy expenditure. The moderate increase in free T3 and total T3 levels observed in obesity leads to increased energy expenditure and, in turn, to reduced availability of energy for conversion to adipose tissue, which leads to a reduced rate of weight gain51,56. A state of resistance to thyroid hormones has also been postulated. This theory is supported by the fact that the levels of T3 receptors are decreased and the negative feedback between TSH and peripheral thyroid hormones is altered in the obese state, which results in resistance of the pituitary gland to the effects of thyroid hormones57. Neuroendocrine dysfunction resulting in an abnormal secretion rate of TSH represents a further hypothesis to explain the elevated TSH concentrations present in obesity56. Furthermore, direct correlations between serum concentrations of TSH and leptin have been reported among obese children and adults, with leptin potentially modifying the hypothalamic production of TSH, either directly or through stimulation of other neurotransmitters and hormones57. Finally, mutations in TSHR have been reported to cause increased TSH levels among children with obesity30.

Idiopathic subclinical hypothyroidism

Children with no clear aetiology who have persistent increases in their TSH levels are classified as having idiopathic subclinical hypothyroidism. The incidence of this form of subclinical hypothyroidism remains unknown.

Natural history

The natural history of subclinical hypothyroidism varies substantially depending on its aetiology, with different courses noted for the autoimmune and nonautoimmune forms.

Nonautoimmune forms

In a retrospective study, TSH values normalized among 73.6% of individuals with mild subclinical hypothyroidism during a 5-year follow-up period, whereas subclinical hypothyroidism remained stable (TSH level, 5.5–10.0 mIU/l) in ∼25% of cases22. Predictive factors for sustained subclinical hypothyroidism in this study were an initial TSH level >7.5 mIU/l and female sex; however, the role of aetiology on the natural course of the disease was not considered. A multicentre prospective study included a 2-year follow-up of 92 children with mild and idiopathic subclinical hypothyroidism; TSH levels normalized in 41.3% of participants, remained stable in 46.7% and increased to overt hypothyroidism in 12.0%58.

A study published in 2015 provided new insights on the natural history of thyroid dysfunction associated with TSHR mutations33. A total of 27 individuals carrying TSHR mutations were followed up for a period of 11 years. Those with heterozygous TSHR mutations exhibited stable hyperthyrotropinaemia associated with normal levels of thyroid hormones that did not change over time. By contrast, a progressive decline in free T4 to levels below the normal range was observed among the individuals with homozygous TSHR mutations33.

Autoimmune forms

The natural history of children affected by autoimmune subclinical hypothyroidism is characterized by an increased rate of progression to overt hypothyroidism. Nonetheless, the outcome is highly heterogeneous among studies owing to variability in the number of patients enrolled and in the severity of subclinical hypothyroidism59. The evolution of subclinical hypothyroidism in seven long-term observational studies of children with Hashimoto thyroiditis was assessed in the 2014 European Thyroid Association guidelines59. A worsening of thyroid function was observed among 24.4% of children, whereas subclinical hypothyroidism persisted in 41.7%. In some studies, an increased risk of progression to overt hypothyroidism was associated with the presence of goitre, elevated anti-thyroid antibodies at presentation and/or a progressive increase in both TPO antibodies and the levels of TSH during follow-up.

A retrospective multicentre study compared the natural disease course among children with Hashimoto thyroiditis or isolated hyperthyrotropinaemia25. Progression from subclinical hypothyroidism to overt hypothyroidism was observed among 21.4% of the children with Hashimoto thyroiditis and 13.6% of those with isolated hyperthyrotropinaemia after 3 years of follow-up. Elevated levels of TSH and TPO antibodies at presentation and concomitant coeliac disease were associated with an increased risk of developing overt hypothyroidism. These data were confirmed by two prospective studies60,61, which documented an increased risk of progression to overt hypothyroidism among children with mild subclinical hypothyroidism owing to Hashimoto thyroiditis (53.1%) versus children with mild nonautoimmune subclinical hypothyroidism (11.1%)60. A high risk of deterioration in thyroid function was found among girls with Turner syndrome61. The natural history of subclinical hypothyroidism among children with Down syndrome is unknown; however, the condition seems to be self-limiting and the presence of anti-thyroid antibodies or goitre is an independent factor associated with the persistence of subclinical hypothyroidism in this population62.

Symptoms and clinical outcome

The definition of subclinical hypothyroidism refers to a biochemical dysfunction in the absence of clinical signs or symptoms of thyroid failure; however, it seems that mild thyroid dysfunction in adults could be associated with subtle clinical, biochemical or functional alterations7. Clinical features associated with modestly increased TSH levels in adulthood range from no manifestations to signs and symptoms of hypothyroidism. This clinical heterogeneity can be related to the differing severity, duration and aetiology of thyroid dysfunction.

By contrast, only a few studies have investigated the effects of untreated subclinical hypothyroidism among children. The majority of these studies have examined the effect of modestly increased TSH levels on growth, neurocognitive development, metabolic parameters and cardiovascular risk.

Linear growth

Adequate concentrations of thyroid hormones during childhood and adolescence are essential for normal growth, with short stature and bone retardation recognized as common clinical signs of untreated overt hypothyroidism. Thyroid hormones regulate linear growth, either by direct action on the bone or by influencing the activity of the growth hormone–insulin-like growth factor 1 (GH–IGF-1) axis18.

Several studies have evaluated linear growth among children with autoimmune and nonautoimmune subclinical hypothyroidism. A retrospective study assessed the growth pattern of 25 children with autoimmune subclinical hypothyroidism and type 1 diabetes mellitus63. Short stature was observed only among children with severe thyroid dysfunction (TSH >50 mIU/l and T4 levels at the low end of the reference range), with no growth impairment documented in children with mild subclinical hypothyroidism. Another study documented normal height and growth velocity among 55 children with autoimmune subclinical hypothyroidism (either isolated or associated with other autoimmune or genetic conditions) after 5 years of follow-up without any therapeutic intervention64. A further study that enrolled 88 children and adolescents with mild-to-severe nonautoimmune subclinical hypothyroidism identified idiopathic short stature among 19.3% of the participants; however, this result probably reflected selection bias given that all of the participants were previously referred for growth impairment30.

Two studies documented normal linear growth among selected children with mild, idiopathic and long-lasting subclinical hypothyroidism18,58. No changes in height were observed among 92 children with persistent idiopathic subclinical hypothyroidism after 2 years of follow-up58. Height, growth velocity, bone maturation and IGF-1 levels were normal (and similar to those of a matched control group) among 36 children with untreated subclinical hypothyroidism of long duration (2.0–9.3 years)18.

Neurocognitive outcome

Thyroid hormones have essential roles in brain development during fetal and postnatal life by influencing neuronal migration, differentiation, myelination and synaptogenesis65. Indeed, neonates and infants with hypothyroidism are at risk of permanent mental retardation if not adequately treated with levothyroxine. Conversely, mild thyroid dysfunction occurring after 3 years of age, when most of the thyroid-hormone-dependent brain development is complete, might result in only subtle neurocognitive impairment66.

Few studies have examined the effect of subclinical hypothyroidism on neurodevelopment in children, and they have yielded contrasting results. A survey among healthy children aged 9–11 years found that TSH levels in the upper portion of the reference range were associated with subtle negative effects on verbal comprehension and immediate and long-term recall67. Another study evaluated cognitive function among 17 children aged 9.9 ± 2.3 years with mild nonautoimmune subclinical hypothyroidism and 17 healthy control individuals68. The children with subclinical hypothyroidism had markedly lower scores in tests measuring attention compared with the control group; however, no differences were reported in verbal fluency and encoding tests. Attention problems have also been reported in a non-controlled study of 11 children69. In NHANES III, cognitive performance was normal among adolescents with subclinical hypothyroidism who scored even better than euthyroid participants in block design and reading21. A prospective case–control study evaluated neuropsychological outcome among 30 children aged 8.4 ± 0.7 years with mild but long-lasting idiopathic subclinical hypothyroidism18. No differences were found in IQ, degree of depression and behavioural problems between the children with subclinical hypothyroidism and a control group matched for socioeconomic status. Moreover, no relationship was observed between IQ and TSH levels or duration of subclinical hypothyroidism18.

Bone health

Thyroid hormones are crucial for normal postnatal skeletal development and bone remodelling, influencing both osteoclast-induced bone resorption and osteoblast-induced bone formation70. In vitro studies suggest that TSH also has a direct role in skeletal remodelling71.

To date, only two studies have evaluated the effects of untreated subclinical hypothyroidism on bone health72,73. Neither of these studies identified changes in biochemical markers of bone metabolism, lumbar BMD and bone quality (measured as phalangeal amplitude-dependent speed of sound and bone transmission time z-score) among individuals with subclinical hypothyroidism. These findings suggest that long-term untreated mild subclinical hypothyroidism does not affect bone health among children72,73.

Cardiovascular risk

Thyroid hormones influence heart and vascular function, as well as lipid and carbohydrate metabolism, through actions on the liver, adipose tissue and muscle74,75 (Fig. 1). As a consequence of these metabolic actions, thyroid hormones have a key role in the modulation of several atherosclerotic factors. The relationship between overt hypothyroidism and atherosclerosis has been clearly documented76. Furthermore, a potential link between subclinical hypothyroidism and cardiovascular risk factors, the metabolic syndrome and heart failure has been raised in adults1.

Figure 1: Functional effects of thyroid hormones.
figure 1

Thyroid hormones (TH) influence fatty acid turnover by increasing uptake, lipogenesis, lipolysis and β oxidation in the liver and by stimulating lipolysis in the adipose tissue. TH potentiate thermogenesis and energy expenditure in white and brown adipose tissue. TH reduce insulin sensitivity in the liver and increase hepatic glucose production; TH also increase insulin sensitivity in the muscle. The effects of TH on the heart involve increased velocity and strength of contraction and increased velocity of diastolic relaxation TH induces the synthesis of vasodilators (nitric oxide (NO) and atrial natriuretic peptide (ANP)) and an increase in the number of resistance arterial vessels, leading to reduced arterial resistance and increased venous tone.

PowerPoint slide

Full size image

Coronary heart disease and heart failure seem to occur most frequently among adults with subclinical hypothyroidism, particularly when the TSH levels exceed 10 mIU/l1,10,11,12. Nonetheless, individuals aged <50 years with TSH levels of 7.0–9.9 mIU/l might have an increased risk of stroke13. Subclinical hypothyroidism also seems to be associated with impaired left ventricular diastolic function at rest, systolic dysfunction on effort and impaired flow-mediated dilation and intima–media thickness in adults8,13,77,78.

Whether subclinical hypothyroidism is associated with an increased prevalence of atherosclerotic risk factors among children is still unclear. The atherosclerotic process is known to start in childhood and progress throughout adult life, eventually leading to cardiovascular adverse effects79. The Bogalusa Heart Study found that early atherosclerotic changes among children correlate with risk factors such as BMI, systolic blood pressure and diastolic blood pressure, as well as with the levels of triglycerides, total cholesterol, LDL cholesterol and HDL cholesterol80.

Studies that have focused on the relationship between subclinical hypothyroidism in children and early markers of cardiovascular risk are outlined in Table 1.

Table 1 The effects of subclinical hypothyroidism on cardiovascular risk factors among children
Full size table

Dyslipidaemia. Some data suggest that children with subclinical hypothyroidism might be at risk of proatherogenic abnormalities. Indeed, mild dyslipidaemia has been documented among children with untreated subclinical hypothyroidism17,81,82, and the levels of TSH seem to be positively correlated with triglycerides83 and non-HDL cholesterol (a measure of the cholesterol content of all the plasma atherogenic lipoproteins)84,85. A study of 315 children with subclinical hypothyroidism found reduced HDL cholesterol levels only among the participants with severe forms of subclinical hypothyroidism (TSH level >10 mIU/l)82. By contrast, a study of 17 children followed for a period of 4 months reported decreased concentrations of HDL cholesterol among those with mild subclinical hypothyroidism (TSH level <10 mIU/l)81. These findings have been confirmed in a longitudinal study that enrolled 49 children with mild idiopathic subclinical hypothyroidism17. Decreased levels of HDL cholesterol, an increased ratio of triglycerides to HDL cholesterol and an increased atherogenic index were observed in comparison with the euthyroid control group. Moreover, abnormalities in lipid parameters among the children with subclinical hypothyroidism correlated with the duration of TSH elevation17.

Blood pressure. A link between TSH levels and blood pressure in childhood has been demonstrated in two large cross-sectional observational studies86,87. One study, which included >12,000 children and adolescents, found an association between serum TSH concentrations above the reference range and hypertension86. Similarly, in a large cohort of Chinese children and adolescents, the TSH level positively correlated with systolic and diastolic blood pressure, both of which were appreciably higher among individuals with subclinical hypothyroidism than in euthyroid children87. Conversely, another study failed to identify differences in blood pressure levels between children with mild idiopathic subclinical hypothyroidism of long duration and healthy euthyroid children17.

Glucose metabolism. The relationship between parameters of glucose metabolism (insulin levels, glucose levels and insulin resistance) and subclinical hypothyroidism has been poorly investigated in children and the available studies provide contrasting results. A retrospective cross-sectional study of 1,146 children and adolescents reported a positive correlation between the levels of TSH and insulin, as well as with the presence of insulin resistance83. Another study reported that even TSH levels within the reference range are negatively associated with insulin sensitivity among non-diabetic, euthyroid male adolescents with obesity88. However, other paediatric studies found no association between subclinical hypothyroidism and insulin insensitivity17,84,89.

Obesity. Most of the studies evaluating the correlation between cardiovascular risk and elevated levels of TSH in childhood have been conducted in populations of children with obesity and subclinical hypothyroidism. However, such studies should be viewed with caution given that subclinical hypothyroidism might be a consequence, rather than a cause, of obesity. In addition, obesity is a condition that increases cardiovascular risk. Therefore, the relationship between obesity and subclinical hypothyroidism in childhood remains debated.

Convincing evidence suggests that elevation of TSH concentration might represent an adaptation of the hypothalamic–pituitary–thyroid axis to obesity56. In keeping with this hypothesis, TSH levels among children with obesity tend to decrease with weight loss induced by lifestyle changes54,55. Several paediatric studies failed to demonstrate an increased degree of overall adiposity among children with long-term untreated autoimmune and idiopathic subclinical hypothyroidism17,58,64. However, despite a normal degree of overall adiposity, children with subclinical hypothyroidism might exhibit increased visceral adiposity. Indeed, waist circumference and waist-to-height ratio were higher among children with mild subclinical hypothyroidism than among healthy euthyroid children matched for age, sex, height and pubertal status17. In this study, the waist-to-height ratio was above the 0.5 limit among children with subclinical hypothyroidism — a value considered to be the boundary indicating an increased metabolic risk for all ages and ethnic groups90 — and this parameter was also related to the degree of subclinical hypothyroidism17. Accordingly, a cross-sectional study that aimed to compare components of the metabolic syndrome by varying TSH levels found waist circumference to be larger among children with subclinical hypothyroidism than among their euthyroid counterparts and to correlate with the concentration of TSH84.

Adipokines and satiety hormones influence or contribute to cardiovascular risk91. Slight alterations in the levels of obestatin were reported for children with autoimmune subclinical hypothyroidism89. Conversely, no differences have yet been found in the levels of leptin, adiponectin, resistin and ghrelin17,89,92. Consequently, the role (if any) played by these markers of cardiovascular risk among children with subclinical hypothyroidism remains to be clarified.

Homocysteine. Elevated plasma levels of homocysteine are thought to be an independent risk factor for cardiovascular disease93. Only two studies have evaluated homocysteine concentrations in children and adolescents with subclinical hypothyroidism17,94. A small case–control study of 19 adolescents with subclinical hypothyroidism failed to identify any abnormalities in the levels of homocysteine94. By contrast, homocysteine levels were substantially elevated among children with subclinical hypothyroidism versus the healthy euthyroid control individuals in a study of 49 children; the levels of homocysteine also correlated with duration of subclinical hypothyroidism (3.2 ± 0.4 years)17. These observations suggest that only several years of untreated subclinical hypothyroidism might be required to affect the concentration of homocysteine.

Inflammation. Compelling evidence suggests a crucial role for inflammation in the initiation and progression of atherosclerosis. Nevertheless, one case–control study failed to demonstrate any association between mild subclinical hypothyroidism in childhood and increased serum levels of the inflammation markers C-reactive protein and fibrinogen17.

Endothelial dysfunction. Alterations in flow-mediated dilation and intima–media thickness are early markers of endothelial dysfunction and atherosclerotic changes. Data from adults suggest that subclinical hypothyroidism can impair both flow-mediated dilation8,78,95 and intima–media thickness9. To date, only one study has been published on the evaluation of vascular function by these markers among 39 children with mild but long-lasting idiopathic subclinical hypothyroidism96. This study did not demonstrate alterations in flow-mediated dilation and intima–media thickness; however, increased concentrations of asymmetric dimethylarginine were found among the children with subclinical hypothyroidism versus 39 healthy euthyroid children matched for age, sex, height and pubertal status96. Asymmetric dimethylarginine is an endogenous competitive inhibitor of endothelial nitric oxide synthase97, which is considered a novel and early marker of endothelial dysfunction. Indeed, the effects mediated by asymmetric dimethylarginine on nitric oxide availability might represent one of the mechanisms by which subclinical hypothyroidism affects vascular function. Consequently, increased levels of asymmetric dimethylarginine could be the first step in the atherosclerotic process that leads to structural abnormalities in adulthood.

Cardiac function. Although TSH and thyroid hormones are important for cardiac performance, very few studies have evaluated cardiac function among children with subclinical hypothyroidism. Ventricular abnormalities were reported in a population of 31 children with subclinical hypothyroidism displaying increased interventricular septum thickness and left ventricular mass thickness compared with healthy euthyroid children98. Conversely, a case–control study that enrolled children with both Down syndrome and subclinical hypothyroidism failed to document any abnormalities in systolic and diastolic function99. Studies of selected populations are required to clarify the presence of early ventricular dysfunction among untreated children with idiopathic subclinical hypothyroidism.

Effect of levothyroxine treatment

The management of subclinical hypothyroidism is still a debated issue, particularly regarding the mild forms. Current guidelines from the American Thyroid Association and the European Thyroid Association suggest offering levothyroxine treatment to adults with TSH levels >10 mUI/l, as well as to those with TSH levels of 4.5–10.0 mIU/l in the presence of symptoms or signs of hypothyroidism and/or anti-thyroid antibodies and/or evidence of atherosclerotic cardiovascular disease14,15.

The issue of treatment during childhood is even more controversial. Beneficial effects of levothyroxine on clinical signs and symptoms of hypothyroidism and goitre, especially among children with Hashimoto thyroiditis, have been documented in several studies100,101,102,103. However, long-term studies on the effects of levothyroxine on other clinical outcomes, in particular for children with nonautoimmune subclinical hypothyroidism, are lacking, with only a few nonrandomized short-term trials that had inconsistent results.

Linear growth

No appreciable differences in height were detected between children with type 1 diabetes mellitus and mild subclinical hypothyroidism and children with only type 1 diabetes mellitus either before or after treatment with levothyroxine63. However, an improvement in growth velocity was observed after 1 year of levothyroxine therapy only among the children with severe subclinical hypothyroidism63.

In a study conducted among 39 Turkish children with subclinical hypothyroidism and short stature, treatment with levothyroxine for 1 year seemed to improve growth pattern, particularly among the pubertal group104. However, the definition of subclinical hypothyroidism used in this study was arbitrary (an exaggerated response to the TSH-releasing hormone test with normal basal levels of TSH) and the participants were affected by short stature and delayed bone age, thus suggesting constitutional growth delay with possible spontaneous improvement of growth velocity over time, independent of the effects of levothyroxine.

Two prospective studies investigated the effect of 2 years of levothyroxine treatment on linear growth among children with mild idiopathic subclinical hypothyroidism and no symptoms of hypothyroidism; however, no statistically significant effects of treatment on height were found19,96.

Neurocognitive development

The effect of levothyroxine therapy on neurodevelopmental outcomes has been evaluated in one small study69. Levothyroxine was administered for a period of 6–8 weeks among eight children with congenital subclinical hypothyroidism and three children with acquired subclinical hypothyroidism. No changes in neurocognitive scores were observed after treatment69.

Bone health

The only study evaluating the effect of levothyroxine therapy on bone health was conducted among 13 children with autoimmune subclinical hypothyroidism72. No evidence was found for an effect of treatment (2–5 years) on BMD.

Metabolic and cardiovascular function

Few studies have evaluated the effect of levothyroxine treatment on metabolic parameters and cardiovascular function. Administration of levothyroxine reversed the mild abnormalities in left ventricular systolic and diastolic parameters detected among 31 children with subclinical hypothyroidism98. In another study, no appreciable changes in lipid profile were observed among the 27 participants who received levothyroxine treatment compared with the lipid profiles of 24 healthy euthyroid children105. A 2-year prospective study enrolled 39 prepubertal children with mild idiopathic subclinical hypothyroidism of long duration96. In this study, administration of levothyroxine was associated with beneficial effects on biochemical markers of cardiovascular risk and endothelial function. Substantial reductions in visceral adiposity (waist-to-height ratio) and asymmetric dimethylarginine levels were observed among children with subclinical hypothyroidism. Moreover, levothyroxine therapy was associated with a modest improvement in the lipid profile96.

Diagnosis and management

The management of subclinical hypothyroidism in childhood is a controversial issue. The first step in managing a child with a modest increase in TSH levels should be differentiation between persistent and transient forms of subclinical hypothyroidism (Fig. 2). Persistent subclinical hypothyroidism should be confirmed by re-evaluation of the TSH levels at 4–12 weeks after the first test to rule out abnormal values caused by laboratory problems, diurnal variation in TSH concentration and transient causes of subclinical hypothyroidism (recovery phase from nonthyroidal illness or subacute thyroiditis).

Figure 2: Proposed management of children with subclinical hypothyroidism.
figure 2

The initial step in the management of a child with a mild increase in TSH concentration should be confirmation of hyperthyrotropinaemia by re-evaluation of TSH levels 4–12 weeks after the first test. If an elevated TSH concentration persists, the diagnostic process should first include careful assessment of the child's history and a clinical evaluation focused on the detection of signs and symptoms suggestive of thyroid dysfunction. Thereafter, evaluation of anti-thyroid antibodies and thyroid ultrasonography will enable differentiation between autoimmune and nonautoimmune forms of subclinical hypothyroidism (SH). The subsequent management and follow-up will depend both on the aetiology and the degree of TSH elevation. The final decision on treatment should be based on the assessment of clinical symptoms or signs of mild thyroid impairment and on the risk of progression to overt hypothyroidism. FT4, free T4.

PowerPoint slide

Full size image

If an elevated TSH level persists after re-testing, a diagnostic evaluation is recommended. The child's history should focus on the presence of neonatal hyperthyrotropinaemia; autoimmune and/or genetic conditions; use of medications known to interfere with thyroid function; previous exposure to ionizing radiation; and endemic iodine deficiency. Attention should be given to the presence of subclinical hypothyroidism, goitre, endocrine diseases, autoimmune diseases or genetic conditions in other members of the patient's family. Physical examination should focus on signs of hypothyroidism, goitre, weight gain and clinical features suggestive of specific genetic conditions.

Given that Hashimoto thyroiditis is the condition most often responsible for the onset of subclinical hypothyroidism in childhood, all patients with the persistent form should be screened for the presence of anti-thyroid antibodies and undergo ultrasonography of the thyroid gland. Moreover, based on the history and physical examination, further investigations can be considered for some children, including TSHR genotype for cases arising in familial settings, urinary iodine excretion for those living in endemically deficient areas or screening for resistance to parathyroid hormone, follicle-stimulating hormone or luteinizing hormone if pseudohypoparathyroidism type 1a is suspected.

The subsequent management and follow-up of persistent forms of subclinical hypothyroidism should depend both on the aetiology and degree of TSH elevation. The final decision on treatment should be made according to the assessment of clinical symptoms or signs of mild thyroid impairment and the risk of progression to overt hypothyroidism.

In the most common clinical scenario of autoimmune subclinical hypothyroidism, treatment with levothyroxine should be considered for all children affected by severe forms (TSH level >10 mIU/l) or among those with mild subclinical hypothyroidism in the presence of goitre or the signs or symptoms of hypothyroidism.

Untreated children should be monitored every 6 months (thyroid function tests) and every 1–2 years (anti-thyroid antibodies and ultrasonography). Careful monitoring is particularly recommended in the presence of chromosomal abnormalities (Turner syndrome and Down syndrome) or other autoimmune conditions to assess increased risk of progressive thyroid dysfunction.

Management of children with reversible causes of subclinical hypothyroidism should focus on modifiable factors. Diet and lifestyle changes are advisable for children who are overweight or obese; thyroid function should be checked after weight loss. Iodine supplementation is recommended among children living in areas with endemic iodine deficiency and/or with documented reduced iodine excretion. Thyroid function in such cases should be re-evaluated after iodine normalization.

The use of some medications, such as antiepileptic drugs and IFN-α, might interfere with thyroid function. Treatment with levothyroxine should be considered for children with a TSH level >10 mIU/l until medications are discontinued. Children with mild forms of subclinical hypothyroidism should be monitored every 6 months.

The management of children with genetic conditions and neonatal hyperthyrotropinaemia should be evaluated on an individual basis. Intervention should depend on the child's age, the degree of TSH elevation and the underlying genetic condition. However, as for other aetiologies, levothyroxine is recommended for severe forms of subclinical hypothyroidism and for symptomatic children, whereas careful monitoring is suggested for the mild and asymptomatic forms.

The management of idiopathic subclinical hypothyroidism is particularly challenging. Children with severe forms (TSH level >10 mIU/l), goitre or symptoms suggestive of hypothyroidism should receive treatment. For children with mild forms (TSH level <10 mIU/l), a trial of levothyroxine can be considered if there is a clinical suspicion of hypothyroidism. In the absence of signs and symptoms, regular clinical evaluation of TSH and free T4 levels, along with periodic re-evaluation of anti-thyroid antibodies, is advisable and should be tailored on the basis of the duration and degree of TSH elevation.

Repeated TSH monitoring can be avoided for children with stable but mild increases of TSH concentration after 2 years of follow-up, unless indicated by the onset of goitre or signs and symptoms suggestive of hypothyroidism.

Finally, it must be highlighted that all forms of subclinical hypothyroidism that resolve at any point during follow-up should be considered for re-evaluation of thyroid function later in life, particularly during adolescence and pregnancy.

Conclusions

Subclinical hypothyroidism in childhood seems to be a benign and remitting condition. The risk of progression to overt thyroid dysfunction depends on the underlying condition, with the risk increased for the autoimmune forms. The major concern regarding subclinical hypothyroidism in children is to establish whether the condition should always be considered an expression of mild thyroid dysfunction.

The current evidence suggests that both clinical signs and symptoms and long-term outcomes depend on the severity and duration of TSH elevation. Present recommendations support levothyroxine therapy for children with severe subclinical hypothyroidism, goitre or symptoms suggestive of hypothyroidism, whereas the management of mild subclinical hypothyroidism remains controversial and individually based. Thus far, no clear evidence has been found that mild untreated subclinical hypothyroidism is associated with alterations in growth or neurocognitive development. However, ongoing scientific investigations have highlighted the presence of subtle proatherogenic abnormalities among children with modest increases in TSH levels, suggesting an improvement of these parameters after treatment with levothyroxine.

These findings are not sufficient to recommend treatment for children with mild asymptomatic forms of subclinical hypothyroidism. Nevertheless, they highlight the need for cardiovascular risk assessment among individuals with subclinical hypothyroidism in childhood. Indeed, the evaluation of early cardiovascular risk biomarkers during childhood and adolescence could improve subsequent management and inform decisions regarding treatment. Adequately powered randomized controlled trials are now required to fully assess the long-term benefits and risks of levothyroxine treatment among children and adolescents with mild subclinical hypothyroidism.