Subclinical Hypothyroidism in Children and Adolescents: Is It Clinically Relevant?

Table 1

Subclinical hypothyroidism in children and adolescents, clinical studies 2008–2014.

Authors and study design	Subjects	Methods	TSH concentration mIU/L	SH: TSH cut-off mIU/L	Duration of follow-up (months)	Main results
Lazar et al. [8]^R	121,052 with SH aged 0.5–16 yrs	TSH, fT4, fT3, TPO-Abs, and TG-Abs	—	>5.5	60	() Initial TSH greater than 7.5 mIU/L and female gender constitute predictive factors for sustained SH; () the natural history of initial mild TSH elevations (5–7.5 mIU/L) in otherwise healthy children not treated with thyroid or antithyroid medication revealed spontaneous normalization of TSH values
Cerbone et al. [9]^CS,CCL	36 with SH^SG, 36 without SH^CG	BMI, fasting TSH, FT4, TG, TPO-Abs, TG-Abs, IGF-1, UIE, BA/CA ratio, and intellectual evaluation	6.4 ± 0.3^SG; 2.6 ± 0.2^CG	>4.2	24–108	() Persistent SH in children not associated with alterations in growth, bone maturation, BMI, or cognitive function or other complaints that could be ascribed to SH even after several years without therapeutic intervention
Wasniewska et al. [10]^CSL	92; (MA: 8.1 ± 3.0 yrs)	TSH, FT4, TPO-Abs, TG-Abs, thyroid USG, and BMI	6.1 ± 1.3 baseline	>5.0	24	() The natural course of TSH values in a pediatric population with idiopathic SH characterized by a progressive decrease over time; () the majority of patients (88%) normalized or maintained unchanged TSH; () TSH changes not associated with any changes in either FT4 values or clinical status or auxological parameters
Kuiper and van der Gaag [11]^CCL	27^SG; 27^CG	FT4, TSH, gender, diet, age at presentation, use of medication, and BMI	5.7^SG; 5.5^CG	>4.2	36	() In the diet group, 74% of the patients normalized TSH; () TSH at the end: 3.5 mIU/L^SG and 4.9 mIU/L^CG
Reinehr et al. [12]^CSL	100 obese girls, 32 normal-weight girls, and 20 girls with anorexia nervosa	TSH, FT3, FT4, BMI, leptin, insulin, and HOMA; girls with AN enrolled in a psychotherapy and nutritional rehabilitation program, obese girls in the 1-year obesity intervention program Obeldicks	—	—	12	() TSH and FT3 levels of girls with AN significantly lower compared to TSH concentrations of normal weight girls; () TSH and FT3 levels of the obese girls significantly higher; () the 21 obese females with weight loss >5% demonstrated a significant decrease in FT3 and TSH; () the 9 adolescents with AN and weight gain >5% showed a significant increase in FT3 and TSH; () insulin and HOMA not significantly correlated to TSH, FT3, and FT4; () leptin was correlated to TSH and FT3 in both cross-sectional and longitudinal analysis
Baş et al. [13]^CSL	150 obese children; (BMI > 95th pc)	TSH, FT3, FT4, TPO-Abs, TG-Abs, BMI, glucose, insulin, HOMA-IR, and weight reduction intervention	2.8 ± 1.4	>4.2	6	() At baseline, 23 (15.3%) subjects had elevated TSH, and 21 of these patients completed the weight reduction intervention; () out of 23 patients, 14 had a substantial weight loss and a significant decrease in TSH and FT3 level
Grandone et al. [14]^CS,CCL	938 obese children and adolescents (BMI > 95th pc)	BMI, waist circumference, SBP, DBP, TG, HDL, insulin, glucose, TSH, FT3, FT4, TG-Abs, TPO-Abs, and weight loss program	5.1 ± 1.0^SG; 2.3 ± 0.8^CG	>4.2	6	() Isolated hyperthyrotropinemia diagnosed in 120 patients (62 girls) of the remaining 938 patients enrolled (12.8%); () in obese children, the increase in TSH is not associated with metabolic risk factors; () hyperthyrotropinemia is reversible after weight loss and these data suggest that it should not be treated; () BMI -score and FT3 levels were significantly higher in patients with elevated TSH, while their age was significantly lower
Wolters et al. [15]^CSL	477 obese children (BMI > 97th pc)	TSH, FT3, FT4, BMI, and BMI-SDS	3.0 ± 1.4	>3.0	24	() 39% of the children demonstrated TSH levels above 3.0 mIU/L for TSH. () BMI-SDS reduction during the lifestyle intervention associated with a reduction in TSH and FT3 concentrations. () Moderately increased TSH and FT3 concentrations in obese children normalized in substantial weight loss
Hari Kumar et al. [16]^R	1: overweight () (BMI: 85th–95th pc); 2: obesity () (BMI > 95th pc)	TSH, FT3, FT4, and BMI	3.22 ± 3.12 (overweight); 3.63 ± 2.24 (obesity)	>4.5	—	() Elevated TSH level (between 4.5 and 10 mIU/L) seen in 4/20 overweight and 9/30 of obese children; () the mean TSH comparable in both the groups; () no correlation between TSH and BMI; () the preliminary data did not show any relation between severity of obesity and TSH level
Shalitin et al. [17]^CSL	207 obese children (BMI > 95th pc)^SG	TSH, FT3, FT4, age, BMI, waist circumference, TG, TC, LDL, HDL, glucose, insulin, CRP, IL-6, homocysteine, adiponectin, leptin, ghrelin, and 12-week weight loss program	5.08 ± 0.84 (boys) 5.20 ± 1.31 (girls)	>4.0	24	() At baseline, 46 participants (22.2%) had hyperthyrotropinemia; () baseline TSH significantly correlated with TG levels but not with age, anthropometric, or other laboratory variables; () Of the 142 participants who completed the intervention, 27 (19%) had hyperthyrotropinemia; () no significant relationship between changes in TSH level and changes in BMI-SDS; () a significant correlation was found between the final TSH level and TG level and between the decrease in TSH level and the decrease in waist circumference
Radetti et al. [18]^CS,CC	186 screened for thyroid disease^SG, 40 without SH^CG	TSH, FT3, FT4, TPO-Abs, TG-Abs, thyroid USG, percent of body fat, height-SDS, and BMI-SDS	A: 7.3 ± 10.6; B: 2.2 ± 1.4; C: 3.9 ± 2.0; D: 2.7 ± 1.7; Control: 2.1 ± 0.7 All: 2.6 ± 4.3	>3.6	—	() 23 children (12.4%) Abs (+) and an USG pattern-> Hashimoto’s thyroiditis (group A); () 20 (10.8%) Abs (+) and normal USG (group B); () 70 subjects (37.6%) Abs (−) and an USG pattern like in Hashimoto’s thyroiditis (group C); () 73 children (39.2%) Abs (−) with normal USG (group D); () groups A and B excluded due to diagnosed thyroiditis; () TSH serum levels significantly different in children with or without thyroid alterations at USG; () obese children may show a different degree of thyroid impairment
Rapa et al. [19]^CS,CC	88	Height, weight, family history of thyroid diseases, thyroid USG TSH, FT3, FT4, UIE, and genetic variations in the TSH-R gene	Overweight/obese versus normal-weight patients; (7.4 versus 5.7); Overweight/obese with hypoechogenicity versus patients with normal USG pattern; (8.5 versus 6.8)	>4.5	1 visit	() TSH levels tended to be higher in the 20 patients with hypoechogenicity; () overweight/obese status, hypoechogenicity at USG, and nonsynonymous mutations in TSH-R gene are characterizing features of a large portion of SH children
Ergür et al. [20]^CC	17 with SH^SG; 17 without SH^CG	TSH, TT4, TRH-test, thyroid volume, urinary iodine level, neuropsychological tests, and BMI	7.2 ± 0.89^SG; 3.6 ± 0.89^CG	>5.0	1 visit	() The SH children obtained significantly lower scores on both the Digit Span subtest of the WISC-R and the Stroop subtests (sensitive to attention); () no significant differences found between the SH group and the healthy controls in verbal fluency and encoding tests
Chen et al. [21]^CS	880 school-aged children (7–18 yrs): 124^SG; 754^CG	TSH, FT4, FT3, TPO-Abs, SBP and DBP related to age, gender, height, and 90th percentile (DBP- and SBP-)	5.47 ± 1.27^SG; 2.50 ± 0.84^CG	>4.2	1 visit	() Serum TSH and FT3 (+) correlated with DBP -score and SBP -score; () DBP- and SBP- in subjects with SH were significantly higher than in euthyroid ones; () both DBP- and SBP- increased linearly in boys with TSH concentrations after adjusting BMI; however a similar linear trend was not observed in girls
Ittermann et al. [22]^CS	6,435 children aged 3–10 yrs; 5,918 adolescents	TSH, FT4, FT3, TPO-Abs, SBP and DBP, and hypertension defined by an increased SBP or DBP using age-, sex-, and height-specific reference values from the KiGGS study	—	—	2 visits	() Increased serum TSH levels significantly associated with continuous values of SBP and DBP in children and adolescents; () children with increased serum TSH levels had significantly lower pulse pressures and a higher risk of hypertension than euthyroid children
Cerbone et al. [23]^CS,CCL	49 with SH^SG; 49 without SH^CG	TSH, FT3, FT4, TPO-Abs, TG-Abs, SBP, DBP, BMI, waist-to-height ratio, lipid profile, homocysteine, hs-CRP, fibrinogen, adiponectin, and HOMA	6.28 ± 0.2^SG; 2.75 ± 0.1^CG	>4.5	38 ± 5	() Waist-to-height ratio, atherogenic index, TG to HDL-cholesterol ratio, and homocysteine levels significantly higher and HDL-cholesterol significantly lower in SH subjects compared with controls; () glucose, insulin, HOMA index, TG, TC, non-HDL-C, fibrinogen, hs-CRP, and adiponectin concentrations similar in the SH subjects and the controls
Zhang et al. [24]^CC,CS	24 with SHyperthyroidism; 866 euthyroid; 29 with SHypothyroidism	TSH, TPO-Abs, thyroid examination BMI, waist circumference, LDL, TG, TC, SBP, DBP, fasting plasma glucose, fasting insulin, and HOMA-IR	<1.01^SHyper; 1.02–5.72^Euthyroid; >5.73^SHypo	>5.73	1 visit	() Waist circumference and BMI were significantly greater among adolescents with SHypothyroidism compared with euthyroid subjects; () the risk of obesity in the SHypothyroid group was 3.444 times that in the euthyroid group (odds ratio = 3.444, 95% confidence interval (CI): 1.570–7.553); () TSH was significantly positively correlated with waist circumference, TC, LDL-C, and TG; () TSH level in the metabolic syndrome group was significantly higher than that in nonmetabolic syndrome group
di Mase et al. [25]^CS,CCL	25 with SH^SG; 25 without SH^CG	TSH, FT4, TPO-Abs, TG-Abs, DXA to evaluate lumbar spine BMD, QUS at proximal phalanges of the nondominant hand to assess bone quality, measured as Ad-SoS and BTT	6.39 ± 1.25^SG; 2.84 ± 0.92^CG	>4.5	36 ± 3	() SH children: normal bone density and structure as assessed by lumbar DXA and phalangeal QUS; () both Ad-SoS and BTT normal and comparable to the controls, despite the long-lasting SH
Fallah et al. [26]^CS	25^SG; 79^CG; (MA: 10.46 ± 2.72 yrs)	History, type, and severity of migraine, TSH, and FT4	—	>4.5	1 visit	() 24% of 5- to 15-year-old migraineurs had subclinical hypothyroidism; () the monthly frequency of headache and the duration of headache were more statistically significant in migraineur children with hypothyroidism
Ekici and Cebeci [27]^CC,CS	93^CG; 5^SG	TSH, FT4, goiter, BMI-SDS, stage of puberty, glucose, TG, HOMA-IR, family history of migraine, and type and severity of migraine	1.96 ± 1.27^CG; 5.36 ± 1.02^SG	>5.0	1 visit	() None of the investigated parameters (sera FT4, TSH, TC, HDL, LDL, and HOMA-IR) had significant effect on the headache severity score
Cho et al. [28]^R	254^SG; 122^CG	TSH, FT3, FT4, TPO-Abs, TG-Abs, and thyroid USG	—	>5.5	—	() No significant difference in the incidence of thyroid disease between children and adolescents with vitiligo and the control group
Onal et al. [29]^CC,CS	10 newborns out of 98 in vitro babies with SH^SG, 10 naturally conceived babies^CG	TRH tests performed in all subjects in^SG,CG; peak TSH response at the 20th min	0 min of TRH-test: 8.08 ± 1.66^SG; 7.55 ± 0.80^CG 20th min of TRH-test: 55.57 ± 14.31^SG; 19.37 ± 3.47^CG	>6.5	1 visit	() An exaggerated TSH response to TRH in all of 10 IVF babies but in none of the control babies in study; () a significant difference between the two groups with respect to TSH levels at the 20th minute of the TRH stimulation test
Sakka et al. [30]^CC,CS	106 in vitro children^SG, 68 randomly selected non-in vitro children^CG; (aged 4–14 yrs)	TSH, FT3, FT4, TPO-Abs, and TG-Abs	—	>5.0	1 visit	() Seven IVF children but none of the controls had persistent elevations of circulating TSH; () TSH: significantly higher in the IVF group than in controls; () an increased incidence of SH in IVF children compared with controls not attributable to birth weight, gestational age, SGA-AGA status, breastfeeding duration, sex, current age, BMI-SDS, or pubertal status of IVF children or to BMI, thyroiditis, PCOS, GDM, parity, AH, or smoking of their mothers
Kratzsch et al. [31]^R	1004	TSH, FT3, FT4, T3, T4, t-uptake, TPO-Abs, TG-Abs, WBC, BMI, and gender	—	—	5	() BMI-SDS and WBC should also be considered when interpreting TSH and thyroid hormone measurements, whereas gender, TPO-Abs, or TG-Abs play a minor role; () BMI-SDS was significantly correlated with T3 and FT3; correlation with FT4 was inversely related; () WBC positively correlated with sera TSH, FT4, T3, and T4; () FT4 is significantly predicted by age and BMI-SDS and FT3 by and BMI-SDS and WBC; () T4 variation depends significantly on age, gender, and WBC while T3 variation was related to age, BMI-SDS, and gender

Ad-SoS: amplitude-dependent speed of sound, BA/CA ratio: bone age/chronological age ratio, BMD: bone mineral density, BTT: bone transmission time, CC: base-control study, CCL: base-control longitudinal study, CG: control group, CRP: C-reactive protein, CS: bross-sectional study, CSL: cross-sectional longitudinal study, DBP: diastolic blood pressure, DBP-: diastolic blood pressure -score, DXA: X-ray densitometry, FU: follow-up, HDL: high-density lipoprotein, hs-CRP: high-sensitive C-reactive protein, HOMA-IR: homeostatic model assessment-insulin resistance, IFG-1: insulin-like growth factor 1, IVF: in vitro fertilization, LDL: low-density lipoprotein, MA: mean age, QUS: quantitative ultrasound, R: retrospective, SBP: systolic blood pressure, SBP-: systolic blood pressure -score, SG: study group, TC: total cholesterol, TG: triglycerides, TG-Abs: anti-thyroglobulin antibodies, TPO-Abs: anti-thyroid peroxidase antibodies, TRH: Thyrotropin-releasing hormone, UIE: Urinary iodine excretion, and USG: ultrasonography.