Patterns of thyroid dysfunction in children with trisomy 21: an observational study

Abstract

Background

Thyroid dysfunctions are the most common endocrine abnormalities observed in trisomy 21 (T21). The prevalence of thyroid disorders in children with T21 varies between 4% and 40%. Dysregulation of thyroid function, which is specific to T21, is a possible etiology. Thyroid dysfunction can have a significant impact on cognitive development and growth. Transiently abnormal thyroid function tests (TFTs) are frequently observed in T21 and may cause diagnostic confusion. There is a lack of literature on several specific thyroid conditions (such as CH, transient abnormal TFT and thyroid dysfunction) in children with T21 in resource-limited settings.

Methods

Records of children followed up at the Child Development Clinic dedicated to children with T21 at the university unit of Lady Ridgeway Hospital (LRH), Sri Lanka, were reviewed retrospectively. The prevalence and characteristics of different thyroid dysfunctions were studied.

Results

Among the 88 children with T21 (Male: Female ratio of 1.5:1), 35 (39.7%) had thyroid dysfunction. Seventeen (20%) children were managed as congenital hypothyroidism (CH), nine (10.2%) had transiently abnormal TFTs. Acquired hypothyroidism was noted in 8 (9%) children, and two children (2.3%) had hyperthyroidism.

Among the 17 children who were managed as CH, 12 (70.5%) were diagnosed during the neonatal period following newborn screening (NBS). The mean TSH level at diagnosis was 22.3 mIU/l (10.6–58). All 17 children had normal ultrasound scans of the thyroid. The average current thyroxine dose in the < 3-year-old group (n = 4) was 2.1 µg/kg/day, and in the ≥ 3-year-old group, it was 2.5 µg/kg/day. The interquartile range(IQR) for age at diagnosis of acquired hypothyroidism was between 7years 9months to 12years 8 months. In those managed as transiently abnormal TFTs, the time taken for TFTs to normalise (without any treatment) ranged widely between six weeks and eight months.

Conclusion

Thyroid dysfunction is a commonly encountered challenge in managing T21. Transiently abnormal TFTs are common and require close observation. Gland-in situ CH (GIS-CH) is increasingly being detected among this population. The cost-effectiveness of using genetic analysis in GIS-CH in resource-limited settings needs to be studied. Differentiating transient CH from permanent CH is quite challenging in clinical practice, although the low thyroxine dose requirement, especially after 3 years of age, may provide valuable information. Further longitudinal studies are needed to delineate the prevalence of transient vs. permanent CH.

Background

The spectrum of thyroid disorders in trisomy 21 (T21) consists of congenital hypothyroidism (CH), acquired hypothyroidism (subclinical or overt) and hyperthyroidism [1, 2]. Limited information is available about transiently abnormal thyroid function tests (TFTs), although they are very common in DS [3]. Differentiating between transient and permanent thyroid dysfunction remains challenging, partly because of the lack of awareness of this transient phenomenon.

Gland-in situ CH (GIS-CH) has also shown an increasing prevalence in recent studies. This rise in detection rates can be attributed to the gradual reduction in the cut-off values of newborn screening [4]. Approximately 35% of patients with GIS-CH have transient thyroid dysfunction and do not require lifelong thyroxine therapy [5].

CH can be transient or permanent. However, thyroid dysgenesis and most dyshormogenesis usually predict that CH will be permanent. Transient CH may be caused by transplacental transfer of maternal TSH receptor blocking antibodies (TRB-Ab), exposure to antithyroid drugs, and iodine deficiency, as well as some genetic causes of dyshormonogenesis, such as DUOX2 mutations [6, 7]. It may also occur following perinatal hypoxia or sepsis, and in some cases, the cause is idiopathic [8]. A study from a general paediatric cohort in the United States reported that 24–36% of CH cases identified by newborn screening (NBS) were later classified as transient [9].

In terms of disease burden, the lifetime prevalence of thyroid dysfunction in T21 is as high as 13–63% [1, 2]. However, further deterioration of cognitive function can be reduced via early detection and treatment of thyroid dysfunction [3, 10]. Research on the prevalence and patterns of thyroid dysfunction among children with T21 in resource-limited settings at both the local and regional levels is scarce.

Regarding screening practices in Sri Lanka, all newborns, including neonates with T21, are subjected to screening using thyroid stimulating hormone (TSH) concentrations. In those with T21, TFTs are then performed at six months, one year and annually thereafter in accordance withthe American Academy of Pediatrics (AAP) recommendations for T21 [11]. Studies conducted in Scotland have suggested that dried blood spot TSH is effective for follow-up screening in T21 to detect hypothyroidism and is easily performed at school. It is more feasible than a venepuncture [12].

To address the aforementioned gaps in knowledge, we conducted the present study at Lady Ridgeway Hospital for children, a tertiary care centerwhere children suspected of, or diagnosed with T21, are referred from all over the country for long-term follow-up by a multidisciplinary team. Our objectives were to describe the prevalence and spectrum of thyroid dysfunction in children with T21 in a resource-limited setting, in this case Sri-Lanka. Particular emphasis has been placed on exploring congenital hypothyroidism and transiently abnormal TFTs.

Methods

Study design

This was a retrospective, descriptive study.

Study setting

Study was carried out at the Child Development Clinic dedicated to children with T21, University Paediatrics Unit, Lady Ridgeway Hospital for Children (LRH), where holistic care is offered for children referred from all over the country. We serve them from birth up to 16years of age. The study was conducted between August 2024 and March 2025.

Study population

All children aging Day 1 up to 16 years of age with a diagnosis of T21 (phenotypically confirmed by a consultant paediatrician with or without a confirmatory karyotype) were recruited.

Sampling method

A nonprobability consecutive sampling method was used.

Objectives

Our aim was to characterise the underlying thyroid dysfunctions in children with T21.

Data collection

Data were collected via an interviewer-administered questionnaire and by reviewing health records during routine follow-up clinic visits. Parents have given written informed consent to participate in the study. Children more than 16 years were excluded from the study.

Statistical analysis

Standard descriptive statistics were used to analyse the data.

The operational definitions used to classify thyroid dysfunctions are included in Table 1 [13].

Table 1.

Definitions of various thyroid dysfunction using biochemical and clinical parameters

Thyroid dysfunction	Operational definition
Congenital hypothyroidism	1. Baby with an abnormal newborn screening, who on further review has a serum TSH > 20 mIU/L ± low fT4 or 2. Baby with an abnormal newborn screening, who on further review has a serum TSH- 6–20 mIU/L and USS showing small/ectopic thyroid gland ± low fT4 or 3. If newborn screening (NBS) not done, a serum TSH > 20 mIU/L ± low fT4 and < 3 years of age
Acquired hypothyroidism	Raised serum TSH ± low fT4 based on age specific reference ranges, above 3years of age (and not meeting criteria for congenital hypothyroidism)
Hyperthyroidism	Low thyroid-stimulating hormone (TSH) concentration in combination with high free thyroxine (fT4) concentration compared with age specific reference ranges
Transiently abnormal TFT	Abnormal serum TSH level compared with age specific reference ranges with normal fT4 and TSH normalises over time without any therapeutic intervention

Adapted from “Thyroid Disorders., by G.Alarcon, V. Figueredo, J. Tarkoff; Pediatr Rev 2021; 42 [12]: 604–618. 10.1542/pir.2020-001420, CC BY-NC [13]

Results

Prevalence

Eighty-eight children (F: M-1.5:1) were recruited into the study. Nondysjunction was the most common finding in terms of karyotype (96%). Thyroid function was abnormal in thirty-five participants (39.8%), the majority of whom were male (F: M- 1:2). The age of the participants ranged from 7 days to 16 years.

The most common thyroid dysfunction was CH (48.6%), followed by transiently abnormal TFTs (25.7%). Eight participants had acquiredhypothyroidism (9%), and two children had hyperthyroidism (5.7%). The remaining participants never had abnormal TFTs among the available and recorded test results. The spectrum of thyroid dysfunctions are summarised in Tables 2 and 3.

Table 2.

Demographic data

Age range	Day7- 16years
Male: Female ratio	1.5:1
Karyotype done	n = 59
− 47,XX/47,XY Non dysjunction	n = 57(96.6%)
- Robertsonian translocation	n = 2(3.3%)
-Mosaics	n = 0
Thyroid dysfunction	n = 35(39.7%)

Table 3.

Spectrum of thyroid dysfunctions

Thyroid dysfunction	n	%	IQR
CH	17	19.3	6.25-71 days
Transiently abnormal TSH	9	10.2	5.5 months-7 year 4months
Acquired hypothyroidism	8	9	7years 9months − 12years 8 months
Hyperthyroidism	2	2.2	4years − 4years 5months

Congenital hypothyroidism, gland-in situ (GIS- CH)

Sixty-nine (78.4%) children had undergone NBS. Among the seventeen participants diagnosed with CH, twelve (70%) were detected by NBS. The other patients (n = 5) were diagnosed via TFTs performed during follow-up. The mean TSH level at the time of CH diagnosis was 22.3 mIU/L. Ultrasound scans of the thyroid were normal in all subjects with CH. The age at diagnosis of CH varied between Day 3 and 1 year. In 2 children (2/17, 11%), CH was diagnosed between 2 weeks and 6 months during the time gap in between first two TFT screenings (NBS and then at 6months) recommended in most guidelines.

Associated anomalies and comorbidities

Eighty-eight percent of T21 subjects with CH also had congenital heart disease, and ventricular septal defect (VSD) was the most common cardiac anomaly. None of the participants with CH had associated gastrointestinal or renal anomalies.

Transiently abnormal TFT

Transiently abnormal TFT were observed among participants at varying ages ranging from 2 weeks to 9 years. The majority (4/9) were observedin infancy, and one was observed during the neonatal period (Day 6). Importantly, none of the patients fulfilled the above-mentioned diagnostic criteria for CH (Table 1). The maximum TSH value among these patients was 21.37 mIU/L, which was associated with a normal fT4 for age (1.2 ng/dL). One of these participants who had transiently abnormal TFTs (elevated TSH) at seven years of age developed acquired hypothyroidism five years later. Among nine children with transiently abnormal TFT, only four have undergone USS and all four of them had normal USS thyroid.

The child in whom transient elevation of TSH was detected on Day 6 of age had abnormal NBS. She was started on thyroxine on Day 6 of life after blood was drawn for venous TSH. The venous TSH level was high (26.6 mIU/L). However, the TSH level after two weeks of starting thyroxine was very low (0.32 mIU/L), and thyroxine was withheld. The TSH results over three weeks of repeated testing without thyroxine treatment were normal.

Acquired thyroid dysfunctions

The F: M ratio of acquired hypothyroidism patients was 1:1, and the IQR for age at diagnosis was between 7years 9months − 12years 8 months.Ultrasound scan of the thyroid revealed evidence of thyroiditis in 6 (75%) participants. Hyperthyroidism was noted in two participants whose average age at diagnosis was 4 years and 2 months. Type 1 diabetes was noted in only one child with acquired hypothyroidism. One participant with hyperthyroidism subsequently developed Addison’s disease.

Among the 35 participants with thyroid dysfunction, only 3 (8.5%) had a positive family history of thyroid disorder in a first- or second-degree relative.

Discussion

Transiently abnormal TFT

One of the key findings of this cross-sectional descriptive study performed at a tertiary care center in Sri Lanka was that a significant proportionof children (10.2%) with T21 had transiently abnormal TFTs. Transiently elevated TSH has been reported to be a common thyroid dysfunction in children with T21 in several studies where mild isolated elevation of TSH is reported as high as 80 − 90% in early infancy, and 30 − 50% thereafter [3, 8, 14–19].

Although isolated mild elevation of TSH is a common finding in children with T21, the evidence supporting treatment remains controversial. Some studies suggest that thyroxine therapy should not be initiated unless the TSH level exceeds the 95th centile [12, 20]. The European Society for Paediatric Endocrinology recommends that in a well infant older than 21 days, a TSH concentration between 6 and 20 mU/L with a normal free T4 can be managed with close follow-up. This includes repeat thyroid function testing after two weeks, along with appropriate imaging, to establish a definitive diagnosis. When it comes to neonates with mild isolated elevation of TSH, families should be counselled regarding the available management options, which include initiating thyroxine treatment immediately with later reassessment and a trial off therapy, or alternatively, closely monitoring thyroid function without starting treatment at the outset [21].

This view is supported by observations that isolated TSH elevation may be an inherent feature of T21, rather than an abnormality affecting only a subset of children [12, 20]. Low fT4 compatible with left shift of normal distribution and mildly elevation of TSH raises the possibility of thyroid regulation disorder specific to T21 and hence the necessity of having diagnostic criteria specific for T21 [22].

Isolated TSH elevations tends to normalise spontaneously in DS. Therefore, the watch and wait policy can avoid chronic treatment with thyroxine. Anti-TPO antibodies may predict the risk of developing acquired thyroid disease later on [23]. A randomized trial reported that early thyroxine treatment in children with T21 who had normal newborn screening results was associated with improved motor development and modest growth benefits at two years of age.¹ However, longer-term follow-up of the same cohort (mean age 10.7 years) demonstrated no sustained benefit in motor or cognitive outcomes, although a small improvement in linear growth persisted [15, 24]. Overall, the study did not support a clear long-term developmental benefit of thyroxine therapy initiated during the first two years of life. A study has shown that overtreatment of CH is detrimental for cognitive outcome at 11years [25].

These controversies may lead to unwarranted burdens through frequent investigations, mislabelling as permanent hypothyroidism or even unnecessary treatment. An important issue to be clarified is whether an isolated TSH elevation implies mild hypothyroidism that can affect developing brain [26].

Gland-in situ congenital hypothyroidism (GIS-CH)

Interestingly, the USS in all children with CH was normal in this group. This finding is similar to the results of a study performed in Thailand [27]. This finding contrasts with the prevailing literature, which mentions thyroid dysgenesis as the most common cause of CH [28]. A study performed in Turkey reported thyroid hypoplasia in 95% of children with CH among their study participants [3, 29]. At times, it may not be possible for USS to differentiate thyroid hypoplasia, normal size and thyroid enlargement clearly [30]. It is advised to be cautious in interpreting neonatal USS alone, since misdiagnosing non thyroidal tissue in thyroid fossa as thyroid dysplasia or hypoplasia is being reported [31, 32].Therefore, using ultrasound and isotope scanning in combination is more informative in the diagnosis of congenital hypothyroidism than USS alone [33].

Moreover, mean TSH level of 22.3 mIU/L at the time of diagnosis of CH is lower than typically reported in CH. This finding was an interesting observation in our cohort of children where all of them were having normal USS thyroid. It suggests the possibility that TSH levels at diagnosis may be lower in GIS-CH compared with CH due to thyroid dysgenesis. This observation highlights an area for future studies to further explore differences in biochemical presentation between these aetiologies.

Thyroid hormone organification defects and inactivating TSH-receptor gene mutations have been suggested as possible but rarer mechanisms of this entity [34, 35]. Therefore, the role of genetic testing in this group of children to reveal molecular mechanisms needs to be studied further.

An increasing number of studies have recently reported GIS-CH among the general population undergoing newborn screening. The recent increase in GIS-CH has been attributed to changes in NBS strategies, as well as the growing number of neonates at risk of developing transient CH such as prematurity, low birth weight, and iodine deficiency during pregnancy. With the rising incidence of mild or transient forms of CH, cases of GIS-CH require careful re-evaluation to distinguish between transient and permanent disease and to guide appropriate long-termmanagement decisions [4, 36]. A study done to re-evaluate children with GIS-CH after 3years of age has shown normalisation of TFT in more than one third of the cohort, upon withdrawing treatment after 1 year of follow-up [37].

Transient vs. permanent CH

In both the CH patient groups stratified by age in our study, those younger than three years and those ≥ 3 years, the average daily thyroxine requirement was relatively low (2.1 and 2.5 µg/kg/day, respectively). Comparing the prevalence of transient vs. permanent CH is beyond the scope of our cross-sectional study. However, according to a recent consensus guideline endorsed by the European Society of PediatricEndocrinology, interrupted treatment is recommended if the daily dose of L-thyroxine (LT4) is less than 25 µg (or < 3 µg/kg/day), with stable or decreasing dose requirements without an increase in TSH during therapy. If a child with no permanent CH diagnosis and a GIS- CH requires an LT4 dose less than 3 µg/kg per day at the age of 6 months, then re-evaluation can be performed at that time [4, 38].

Donaldson et al. have proposed an insightful diagnostic classification for congenital hypothyroidism (CH) in Scotland, in which transiently elevated TSH is recognised as a distinct entity. In this category, the NBS TSH is elevated, but subsequent TFTs are normal after withdrawal of thyroxine therapy. Additionally, preterm, unwell, or dysmorphic infants with persistently but mildly elevated TSH levels were classified under“Status Uncertain.” [39]. This approach reflects careful consideration of the fact that TSH elevation in certain groups, including infants with T21, may represent a transient physiological phenomenon rather than permanent CH.

In cases of TSH elevation of uncertain aetiology, such as in children with trisomy 21 or in those with mildly elevated TSH ± gland-in-situ (GIS), re-evaluation is recommended at 2–3 years of age. This involves gradually tapering thyroxine over a four-week period, followed by repeat TFT two weeks after discontinuation. Imaging may also be performed at that stage if it has not been undertaken previously [32].

TFTs during follow-up

Since there were children (n = 2, 11%) who were diagnosed with CH between 2 weeks and 6 months, the necessity of an additional TFT between birth and 6 months in this at-risk population needs to be evaluated further.

Prevalence

Our study confirmed a higher prevalence of thyroid dysfunction in T21 patients (39.8%). This varies among studies. While the reported prevalence in another Asian cohort was 40% [27], it was comparatively lower (24%) in a Caucasian cohort [1]. Wide variation in prevalence data can be due to population characteristics, differences in the diagnostic criteria used and the technique used to analyse thyroid hormones [40]. The most common thyroid dysfunction in our study cohort was CH.

The prevalence of CH in our T21 cohort was as high as 19.3% with 70% detected by NBS. Similarly, a high prevalence (21.5%) was reported in another study (n = 50) from Libya [41]. In contrast, a study performed in Thailand reported the prevalence of CH to be 3.4% in a cohort of 140 participants with T21 [27]. Local and regional prevalence data need further validation through studies involving substantially larger cohorts.

Hyperthyroidism

The prevalence of hyperthyroidism was greater in our study cohort than in the studies published to date. The highest reported prevalence of hyperthyroidism was 0.06% [42]. Two children (2.2%) in our cohort had hyperthyroidism. In a study conducted to define thyroid autoimmunity has reported the prevalence of TPOAb as 10.3% and TgAb as 6.4% among Sri –Lankan children and adolescents [43]. This is further supported by another study from Sri- Lanka demonstrating a rising aTG antibody levels in school children in the post iodisation era [44].

Hyperthyroidism in T21 is caused mainly by Graves’ disease. Thyroid autoantibodies were not done in these study participants because of concerns about cost-effectiveness in a resource-limited setting. However, screening children with T21 using TFT is essential for the early detection of autoimmune thyroid dysfunction, most commonly Hashimoto’s thyroiditis and, less frequently, Graves’ disease. The onset of Hashimoto’s thyroiditis is typically reported during the primary school years. Although it predominantly leads to hypothyroidism, it may initially present with a transient hyperthyroid phase before progression to hypothyroidism [45]. Thyroid antibody testing is recommended in all children with deranged thyroid functions [46].

Associated anomalies

Congenital heart diseases, particularly VSDs, are frequently observed among children with CH and in children with T21. None of the participants who were diagnosed with CH has undergone surgical procedures; including surgeries for congenital heart disease during the neonatal period, which itself can lead to elevation of TSH and diagnostic dilemmas.

One limitation of this retrospective descriptive study was the inability to observe the impact of abnormal TFTs on growth and cognition. Because there was no control group, we were unable to compare our study population with children from the general population who have thyroid disorders. Although this study was performed at the largest paediatric centre in the island, the study population may not ideally represent the national population with T21, as the majority of the participants were referred from urban areas situated on the outskirts in proximity to the hospital. Neonates who were enrolled at 7 days may have developed transient thyroid abnormalities or acquired hypothyroidism later in life which will not be captured by this study.

Conclusion

Transiently abnormal TFTs are common in children with T21 than expected. Studies addressing its precise pathophysiology and its potential impact on cognition will lay the background for establishing specific management protocols.

Our study further supports recent reports indicating a rise in GIS-CH. Utilisation of recent genetic advances to identify the aetiology of GIS-CH is a challenge in resource-limited settings.

To date, specific management of transient CH in the context of T21 lack clear recommendations in the guidelines. Since the evidence is inconclusive, decision to treat may be made on clinical judgement.

Abbreviations

T21

Trisomy 21/Down Syndrome

CH

Congenital hypothyroidism

TFT

Thyroid function tests

GIS-CH

Gland-in situ congenital hypothyroidism

TSH

Thyroid stimulating hormone

USS

Ultrasound scan

NBS

Newborn screening, LRH-Lady Ridgeway Hospital for children

VSD

Ventricular septal defect

LT4

L-thyroxine

IQR

Interquartile range

Authors' contributions

AW and RD designed the study, collected the data and analysed the patient data. SS and NA interpreted the data regarding thyroid dysfunction. All the authors contributed to the writing, read and approved the final manuscript.

Funding

No funding was received for this study.

Data availability

The datasets used during the current study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

Ethics approval was obtained from the ERC (Ethics Review Committee), Faculty of Medicine, University of Colombo, Sri Lanka. Written, informed consent to participate was obtained from the parents or legal guardians of all the participants under the age of 16. Our study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki.

Consent for publication

Written informed consent was obtained from the parent/legal guardian for the publication of all clinical details included in this article.

Competing interests

The authors declare no competing interests.