0021-972X/05/$15.
00/0
Printed in U.S.A.
The Effect of Thyroxine Treatment Started in the
Neonatal Period on Development and Growth of Two-
Year-Old Down Syndrome Children: A Randomized
Clinical Trial
A. S. Paul van Trotsenburg, Thomas Vulsma, Susanne L. Rutgers van Rozenburg-Marres,
Anneloes L. van Baar, Jeannette C. D. Ridder, Hugo S. A. Heymans, Jan G. P. Tijssen, and
Jan J. M. de Vijlder
Departments of Paediatric Endocrinology (A.S.P.v.T., T.V., S.L.R.v.R.-M., J.C.D.R., J.J.M.d.V.), Paediatrics (H.S.A.H.), and
Cardiology (J.G.P.T.), Academic Medical Center, University of Amsterdam, 1100 DE Amsterdam, The Netherlands; and
Department of Paediatric Psychology (A.L.v.B.), University of Tilburg, 5000 LE Tilburg, The Netherlands
Context: Young Down syndrome children appear to have a mild form
of congenital hypothyroidism that is rarely detected by neonatal
screening and usually left untreated.
Objective: To investigate the effects of thyroxine treatment on de-
velopment and growth of young Down syndrome children.
Design, Setting, and Participants: Single-center, randomized,
double-blind, 24-month trial (enrollment June 1999 to August 2001)
with nationwide recruitment, comparing thyroxine administration
with placebo in 196 Down syndrome neonates.
Intervention: Neonates were randomly assigned to treatment for 2
yr with either thyroxine (n ⫽ 99; initial dose 8 ␮g/kg) or placebo (n ⫽
97). Daily thyroxine doses were adjusted at regular intervals to main-
tain plasma TSH in its normal and free T4 concentrations in its
high-normal range. Placebo dose adjustments mirrored those of
thyroxine.
Main Outcome Measures: The primary outcome was mental and
T HIRTY YEARS AFTER the worldwide introduction of
neonatal screening for congenital hypothyroidism, it is
beyond doubt that early diagnosis and thyroxine treatment
prevent brain damage. Although benefits of early treatment
have never been proven in randomized placebo-controlled
trials, the changes in the natural course of congenital hypo-
thyroidism are considered sufficient evidence for its efficacy,
at least in severe forms (1, 2). A still unsolved question is
whether neonates with mild hypothyroidism benefit from
treatment, too (3–5). In the context of neonatal screening the
term “mild” is often used for the combination of a normal
plasma thyroxine concentration and plasma TSH elevation
on recall testing. Nevertheless, the current advice is to treat
these children to avoid any risk of brain damage due to
hypothyroidism.
First Published Online March 8, 2005
Abbreviations: BSID-II, Bayley Scales of Infant Development II; FT4,
free T4; KID, Kent Infant Development Scale.
JCEM is published monthly by The Endocrine Society (http://www.
endo-society.org), the foremost professional society serving the en-
docrine community.
3304
The Journal of Clinical Endocrinology & Metabolism 90(6):3304 –3311
Copyright © 2005 by The Endocrine Society
doi: 10.1210/jc.2005-0130
Downloaded from https://academic.oup.com/jcem/article-abstract/90/6/3304/2870560 by guest on 07 May 2020
motor development at age 24 months, assessed with the Bayley Scales
of Infant Development II.
Results: At age 24 months, the developmental testing results of 90
thyroxine-, and 91 placebo-treated children were available for anal-
ysis. The thyroxine-treated children had a 0.7-month smaller delay in
motor developmental age (95% confidence interval, ⫺1.4 to 0), cor-
responding to a difference of seven motor developmental index points.
The mental developmental age delay was also 0.7 month smaller in
the thyroxine group (95% confidence interval, ⫺1.5 to 0.2), but lacked
statistical significance. Thyroxine-treated children had greater gains
in length (1.1 cm; 95% confidence interval, 0.2 to 2.0) and weight (378
g; 95% confidence interval 55 to 701).
Conclusions: The data of our study provide evidence to support the
hypothesis that thyroxine treatment may improve development and
growth of young Down syndrome children. Thyroxine treatment
should be considered in Down syndrome neonates to maximize their
early development and growth. (J Clin Endocrinol Metab 90:
3304 –3311, 2005)
In one particular group of children, those with Down
syndrome, mild plasma TSH elevation is extremely preva-
lent: 80 –90% in early infancy, and 30 –50% thereafter (6 –10).
Most of these neonates escape detection by neonatal screen-
ing because their thyroxine levels are just above, or their TSH
levels are just below the screening cutoffs. In other words,
according to the current neonatal screening cutoffs, most
Down syndrome neonates are not at risk for brain damage
due to hypothyroidism.
Recently, we found that Down syndrome neonates as a
group have a lower thyroxine concentration than non-Down
syndrome neonates, caused by a shift to lower values of their
individual thyroxine concentrations (11). We hypothesized
that this might point to a Down syndrome-specific thyroid
(regulation) disorder affecting all children and that the lower
thyroxine tissue availability might be disadvantageous for
these children’s already severely compromised brain devel-
opment and somatic growth (12). We suggested that Down
syndrome infants might benefit from thyroxine treatment. To
test our hypothesis, we carried out a randomized clinical trial
in which either thyroxine or placebo was administered to
van Trotsenburg et al. • Thyroxine Therapy in Down Syndrome
Down syndrome children during their first 2 yr of life, with
mental and motor development as primary outcome
measures.
Patients and Methods
Patients
All Dutch Down syndrome neonates born between June 1999 and
August 2001 were potential candidates for the study. Exclusion criteria
were: abnormal congenital hypothyroidism screening, postnatal age
more than 28 d, gestational age less than 252 d, 5-min Apgar score less
than 7, or insufficient parental command of the Dutch language. Neo-
nates were referred by pediatricians and the Down Syndrome Founda-
tion. The study was conducted in the Academic Medical Center, Uni-
versity of Amsterdam, after approval by its ethics committee. Written
informed consent was obtained from all parents.
Study design
Down syndrome neonates were randomized to receive either thy-
roxine (8 ␮g/kg per day) or placebo, started within 24 h after random-
ization and continued until age 24 months. Thyroxine tablets contained
25 ␮g l-thyroxine, and were divisible in quarters. Placebo tablets were
identical in appearance. Starting doses were rounded off to the nearest
6.25 ␮g. Thyroxine doses were adjusted to reach and maintain normal
plasma TSH (0.4 – 4.0 mIU/liter) and high-normal plasma free T4 (FT4)
concentrations [1.40 –1.86 ng/dl (18 –24 pmol/liter)]. Dose adjustments
were made by a pediatric endocrinologist who did not have any contact
with participants during the study. To ensure blinding, 1) only this
endocrinologist had access to laboratory data; 2) placebo dose adjust-
ments were matched to the thyroxine dose adjustments of the preceding
week; 3) all dosing communication with principal investigators (pedi-
atric endocrinologist, research nurse, and developmental psychologist)
went through standard forms; and 4) parents were informed about new
doses only by principal investigators. The principal investigators and the
participants were blinded to treatment assignment throughout the
study. Besides the nonblinded endocrinologist and a hospital pharma-
cist, only members of the data monitoring committee were privy to data
by treatment, but none had contact with participants.
FIG. 1. Trial profile. CNS, Central nervous sys-
tem. *, Death, one child; trial too demanding,
three children; without reason, two children. †,
Death, one child; thyroxine prescription and
use outside the study, two children; trial too
demanding, one child. ‡, Trial too demanding.
J Clin Endocrinol Metab, June 2005, 90(6):3304 –3311 3305
The probability of assignment to the thyroxine or placebo group was
equal for all children. Balance of maternal educational levels (high,
medium, low) and numbers of children in both groups was ensured by
using a computer-generated randomization list with three strata (ma-
ternal educational levels high, medium, and low) and block size 10.
Assessments
Study visits were at randomization and at ages 2, 3, 6, 9, 12, 15, 18,
21, and 24 months. Data collected at randomization included parental
information on the perinatal period and their socioeconomic condition.
Measurements to guide dose adjustments (weight, plasma TSH, FT4, T4,
T3, and thyroxine-binding globulin) were carried out at all study visits.
TSH and FT4 were measured by time-resolved fluoroimmunoassay
Downloaded from https://academic.oup.com/jcem/article-abstract/90/6/3304/2870560 by guest on 07 May 2020
(Delfia hTSH and FT4 Ultra; PerkinElmer, Wallac Oy, Turku, Finland).
T4 and T3 were measured by in-house RIA methods. Thyroxine-binding
globulin was determined by RIA (Eiken Chemical Co., Tokyo, Japan). If
parents were unable to visit the Academic Medical Center, the weight
measurement and blood collection were performed by the child’s local
pediatrician.
At randomization and at ages 2, 6, 12, 18, and 24 months, length, head
circumference, anterior fontanel size, and heart rate were measured. In
addition, parents were interviewed about their child’s condition, hos-
pital admissions, and medical checks, diagnoses, and procedures in the
preceding period. The means of three consecutive length and head
circumference measurements were used. Length, weight, and head cir-
cumference sd scores were calculated using current Dutch standards
(13).
At ages 6, 12, and 24 months, development was assessed with the
Bayley Scales of Infant Development II (BSID-II), and the Dutch version
of the Kent Infant Development Scale (KID) (14, 15). All children were
tested at age 24 months, irrespective of premature discontinuation of
study medication. The BSID-II was adjusted by starting with items
younger than the actual age at testing (2, 5, and 9 months instead of 6,
12, and 24 months). All developmental tests were carried out by one
developmental psychologist. Parents completed the KID questionnaire
in the week before these study visits. Adherence to study medication
was assessed by counting tablets at all study visits. The effectiveness of
the blinding was assessed by telephone interview 3– 6 months after the
last study visit, in which parents were asked whether during the study
3306 J Clin Endocrinol Metab, June 2005, 90(6):3304 –3311 van Trotsenburg et al. • Thyroxine Therapy in Down Syndrome

they had had knowledge of the kind of study medication their child TABLE 1. Baseline characteristics of the Down syndrome
received. children included in the primary outcome analysis

Statistical analysis Thyroxine Placebo

Characteristic group group
The primary outcome was development at age 24 months, assessed (n ⫽ 90) (n ⫽ 91)
with the BSID-II. All other results were secondary outcomes. The pri- Male sex [no. (%)] 46 (51.1) 53 (58.2)
mary outcome analyses were based on all available data, including Karyotyping [no. (%)]
measurements after premature discontinuation of the intervention. Sec- Trisomy 21 86 (95.6) 88 (96.7)
ondary outcome analyses were based on the data of children who did Translocation 4 (4.4) 1 (1.1)
not discontinue their treatment. Mosaic 0 2 (2.2)
In general, the BSID-II test result is expressed as mental and motor Age (d) 24.3 (3.3) 24.2 (3.3)
developmental indexes, with a mean of 100 and a sd of 15. However, Length (supine)
because of severe developmental delay, in the majority of Down syn- cm 50.0 (2.4) 50.3 (2.5)
drome children developmental indexes cannot be determined. There- SDS ⫺1.8 (1.1) ⫺1.6 (1.2)

Downloaded from https://academic.oup.com/jcem/article-abstract/90/6/3304/2870560 by guest on 07 May 2020

fore, developmental outcome was expressed as mental and motor de- Missing data [no. (%)] 1 (1.1) 1 (1.1)
velopmental age delay (which corrects for differences in age at testing). Weight
Mental and motor developmental index differences were deduced from g 3368 (488) 3384 (486)
the differences in mental and motor raw score between the thyroxine and SDS ⫺1.7 (1.1) ⫺1.7 (1.1)
placebo group (14). To detect a clinically relevant difference in mental Head circumference
and motor developmental index of 7.5 points (or 0.5 sd) at chronological cm 34.4 (1.2) 34.6 (1.3)
age 24 months, with a power of 90% and a two-tailed P value of 0.05, SDS ⫺1.8 (0.9) ⫺1.7 (0.9)
it was calculated that 84 Down syndrome children were required on each Gestational age (wk) 38.3 (1.4) 38.6 (1.2)
treatment (16). To compensate for anticipated loss to follow-up, 100 Birth weight (g) 3050 (573) 3061 (529)
children were planned on each treatment. 5-min Apgar score [median (range)] 10 (7–10) 9 (7–10)
A preplanned interim analysis was carried out after 100 children had Maternal age (yr) 33.0 (4.5) 33.7 (3.9)
undergone BSID-II testing at age 12 months. At this time, several chil- No. of older siblings [no. (%)]
dren were diagnosed with central nervous system disease, among which 0 33 (36.7) 36 (39.6)
children with infantile spasms. Thyroxine treatment could neither sta- 1 38 (42.2) 36 (39.6)
tistically nor causally be connected with the occurrence of infantile 2 15 (16.7) 12 (13.2)
spasms or other central nervous disease (17). Therefore, the data and 3 or more 4 (4.4) 7 (7.6)
safety monitoring committee recommended that the study be continued Mother’s educational level [no. (%)]a
as planned. Because central nervous system disease was considered a High 42 (46.7) 42 (46.2)
major negative influence on the primary outcome, the committee ad- Medium 36 (40.0) 36 (39.6)
vised additional analyses of the primary outcome, excluding affected Low 12 (13.3) 13 (14.2)
children. Hospital admission [no. (%)]
Primary and secondary outcomes were compared using the inde- Current 7 (7.8) 15 (16.5)
pendent samples t test. Plasma TSH, thyroid hormone, thyroxine-bind- Previous 71 (79.8) 77 (84.6)
ing globulin concentrations, and study medication doses were compared No. of admission days 12.7 (10.6) 15.1 (11.5)
using general linear model repeated measures analysis. All reported P Tube feeding [no. (%)]
values are two-sided. Comorbidity and cointerventions were compared Current 11 (12.2) 15 (16.5)
using the independent samples t test (numerical data), and Pearson ␹2 Previous 45 (50.0) 49 (53.8)
test or Fisher exact test (categorical data). All statistical analyses were Central nervous system disease
done with SPSS for Windows, Release 11.0.1 (15 Nov 2001). [no. (%)]
Neonatal convulsionsb 1 (1.1) 1 (1.1)
Results Abdominal surgeryc 2 (2.2) 8 (8.8)
Characteristics of the patients Data are mean (SD), unless noted otherwise. “Missing data” are
data that were not reported or recorded. SDS, SD score.
During the recruitment period, 335 Down syndrome ne- a
High, Higher general secondary education (passed exam) or
onates were assessed for eligibility, of whom 196 underwent preuniversity education (passed exam) ⫾ higher vocational education
randomization (Fig. 1). Fifteen children discontinued the in- and/or university; medium, lower general secondary education
(passed exam) or intermediate vocational education (passed exam);
tervention: 10 were lost to follow-up, five returned for pri-
low, primary school ⫾ lower vocational education.
mary outcome measurement. Treatment was never discon- b
Resulting from hypoxia or meningitis.
tinued because of adverse events. In five of the children who c
Esophageal atresia, duodenal atresia, annular pancreas, imper-
continued the intervention, the primary outcome was not forate anus, and omphalocele.
available; in four, the primary outcome could not be mea-
sured for technical reasons (hyperactive behavior obstruct-
ing developmental assessment, two children in both treat-
after (Fig. 2). Placebo doses and the total number of dose
ment groups), and one test report got lost (thyroxine group).
adjustments during the study (thyroxine group, 318; placebo
Thus, 90 patients in the thyroxine group and 91 patients in
group, 322) were similar.
the placebo group were included in the primary outcome
From age 2 months onward, median plasma TSH concen-
analyses. Clinical and laboratory data were similar in the two
trations were always normal in the thyroxine group and
treatment groups at baseline (Table 1 and Fig. 2). In both
elevated in the placebo group. Under thyroxine treatment,
groups one child had suffered from neonatal convulsions.
mean FT4 and T4 concentrations were always higher [differ-
ences, 0.51– 0.67 ng/dl (6.6 – 8.6 pmol/liter) and 2.41–3.50
Treatment ␮g/dl (31– 45 nmol/liter), respectively], whereas mean T3
The daily thyroxine dose decreased from 8 ␮g/kg at ran- concentrations were always lower [difference, 13.0 –39.0
domization to a little more than 4 ␮g/kg at age 9 months and ng/dl (0.2– 0.6 nmol/liter)]. Mean thyroxine-binding glob-
van Trotsenburg et al. • Thyroxine Therapy in Down Syndrome J Clin Endocrinol Metab, June 2005, 90(6):3304 –3311 3307

Downloaded from https://academic.oup.com/jcem/article-abstract/90/6/3304/2870560 by guest on 07 May 2020

FIG. 2. Thyroid function parameters and study medication doses at baseline and during the study. TSH is expressed as median and interquartile
range. FT4, T4, T3, thyroxine-binding globulin, and study medication doses are means ⫾ SD. At all ages, calculations were based on 84 –90
measurements in the thyroxine group and 84 –91 measurements in the placebo group. P values refer to the comparison of measurements in
the thyroxine and the placebo group from age 2–24 months by general linear model repeated measures analysis. SI conversion factors: to convert
FT4 to picomoles per liter and T4 to nanomoles per liter, multiply values by 12.87; T3 to nanomoles per liter, multiply by 0.0154; thyroxine-binding
globulin to nanomoles per liter, multiply by 185.19.

ulin concentrations were about similar. During the study,
none of the children developed overt hypothyroidism or
hyperthyroidism.
In both treatment groups, parents reported similar num-
bers of missed study medication administrations [thyroxine
group, 2.2 (3.6); placebo group, 3.3 (4.7)]. After completion of
the study, parents of two children (one in the thyroxine
group and one in the placebo group) indicated that they
knew about the kind of study medication their child received
through self-organized laboratory testing.

TABLE 2. Primary outcome: BSID-II at age 24 months

Thyroxine group Placebo group Difference (95% CI) P

Main analysis n ⫽ 90 n ⫽ 91
Age corrected for preterm birth (months) 23.6 (0.4) 23.7 (0.4)
Mental
Raw score 96.1 (10.9) 93.4 (13.7)
Developmental age delay (months) 9.5 (2.6) 10.2 (3.0) ⫺0.7 (⫺1.5 to 0.2) 0.12
Motor
Raw score 62.2 (5.9) 60.4 (7.4)
Developmental age delay (months) 12.3 (2.1) 13.0 (2.4) ⫺0.7 (⫺1.4 to 0.0) 0.042
a
Additional analysis n ⫽ 81 n ⫽ 87
Age corrected for preterm birth (months) 23.6 (0.4) 23.7 (0.4)
Mental
Raw score 97.9 (8.8) 95.0 (10.4)
Developmental age delay (months) 9.1 (2.1) 9.9 (2.6) ⫺0.8 (⫺1.5 to ⫺0.1) 0.032
Motor
Raw score 63.1 (4.6) 60.9 (6.7)
Developmental age delay (months) 12.1 (2.0) 12.9 (2.3) ⫺0.8 (⫺1.5 to ⫺0.2) 0.015
Data are mean (SD). Only the primary outcomes (developmental age delay) were compared. CI, Confidence interval.
a
After exclusion of all children with central nervous system disease (thyroxine group, nine; placebo group, four).
3308 J Clin Endocrinol Metab, June 2005, 90(6):3304 –3311 van Trotsenburg et al. • Thyroxine Therapy in Down Syndrome

TABLE 3. Secondary outcomes: BSID-II at ages 6 and 12 months, and KID and auxology at ages 6, 12, and 24 months

Thyroxine group Placebo group Difference (95% CI) P value

Age 6 months
Age corrected for preterm birth (months) 5.6 (0.3) 5.7 (0.3)
BSID-II
Mental (n ⫽ 89/83)
Raw score 42.7 (6.5) 42.5 (5.7)
Developmental age delay (months) 1.9 (0.7) 2.0 (0.6) ⫺0.1 (⫺0.3 to 0.1) 0.33
Motor (n ⫽ 89/84)
Raw score 27.4 (3.9) 26.9 (4.1)
Developmental age delay (months) 2.1 (0.8) 2.2 (0.7) ⫺0.1 (⫺0.4 to 0.1) 0.24
KID (n ⫽ 89/87)
Raw score, total 81.7 (18.6) 79.6 (21.8) 2.1 (⫺3.9 to 8.1) 0.50
Length (n ⫽ 88/86)

Downloaded from https://academic.oup.com/jcem/article-abstract/90/6/3304/2870560 by guest on 07 May 2020

cm 63.7 (3.5) 63.4 (2.7)
SDS ⫺1.5 (1.5) ⫺1.7 (1.0) 0.2 (⫺0.2 to 0.6) 0.31
Weight (n ⫽ 91/89)
g 6,554 (942) 6,431 (878)
SDS ⫺1.4 (1.2) ⫺1.6 (1.1) 0.2 (⫺0.1 to 0.6) 0.17
Head circumference (n ⫽ 89/85)
cm 40.9 (1.3) 40.9 (1.2)
SDS ⫺2.0 (0.9) ⫺2.1 (0.8) 0.1 (⫺0.2 to 0.3) 0.55
Age 12 months
Age corrected for preterm birth (months) 11.6 (0.4) 11.7 (0.3)
BSID-II
Mental (n ⫽ 87/88)
Raw score 65.8 (5.7) 64.7 (8.6)
Developmental age delay (months) 5.1 (1.2) 5.2 (1.2) ⫺0.1 (⫺0.5 to 0.2) 0.45
Motor (n ⫽ 87/88)
Raw score 41.9 (7.0) 42.0 (6.3)
Developmental age delay 5.5 (1.5) 5.6 (1.2) ⫺0.1 (⫺0.5 to 0.3) 0.65
KID (n ⫽ 86/90)
Raw score, total 145.5 (28.1) 141.3 (28.4) 4.2 (⫺4.2 to 12.6) 0.33
Length (n ⫽ 88/89)
cm 71.4 (2.6) 70.9 (2.9)
SDS ⫺1.7 (0.9) ⫺1.9 (1.0) 0.2 (⫺0.1 to 0.5) 0.16
Weight (n ⫽ 90/90)
g 8,635 (1059) 8,450 (1120)
SDS ⫺1.3 (1.0) ⫺1.5 (1.1) 0.2 (⫺0.1 to 0.6) 0.13
Head circumference (n ⫽ 89/90)
cm 43.7 (1.3) 43.6 (1.5)
SDS ⫺2.0 (0.8) ⫺2.1 (0.9) 0.1 (⫺0.1 to 0.4) 0.25
Age 24 months
Age corrected for preterm birth (months) 23.7 (0.4) 23.8 (0.4)
KID (n ⫽ 87/88)
Raw score, total 224.1 (23.1) 218.3 (28.6) 5.8 (⫺2.0 to 13.6) 0.14
Length (n ⫽ 91/90)
cm 81.3 (3.2) 80.4 (3.5)
SDS ⫺2.2 (1.0) ⫺2.5 (1.1) 0.3 (0.0 to 0.6) 0.046
Weight (n ⫽ 91/90)
g 10,923 (1219) 10,559 (1255)
SDS ⫺1.4 (1.0) ⫺1.7 (1.1) 0.3 (0.1 to 0.7) 0.022
Head circumference (n ⫽ 91/90)
cm 45.6 (1.4) 45.5 (1.4)
SDS ⫺2.0 (0.8) ⫺2.1 (0.8) 0.1 (⫺0.1 to 0.4) 0.30
Data are mean (SD). Numbers in parentheses (first column, n ⫽ xx/xx) refer to the number of patients (in the thyroxine group and placebo
group, respectively) with measurements available. Except for the KID raw score, only values corrected for (small) differences in age at testing
(developmental age delay and SD score) were compared. CI, Confidence interval; SDS, SD score.

Primary outcome
The thyroxine-treated children had a smaller mental de-
velopmental age delay than the placebo-treated children
(9.5 ⫾ 2.6 vs. 10.2 ⫾ 3.0; difference, 0.7 months) (Table 2). The
thyroxine-treated children also had a smaller motor devel-
opmental age delay (12.3 ⫾ 2.1 vs. 13.0 ⫾ 2.4; difference, 0.7
months). These differences correspond to approximately six
mental and seven motor developmental index points. Only
the difference in motor developmental age delay was statis-
tically significant. Similar results were obtained when the
analyses were repeated in the subset of 176 children who did
not discontinue their treatment (data not shown).
Secondary outcomes
Thyroxine-treated children had a greater length and
weight than placebo-treated children at all three measure-
ments, with differences increasing over time (Table 3). These
van Trotsenburg et al. • Thyroxine Therapy in Down Syndrome J Clin Endocrinol Metab, June 2005, 90(6):3304 –3311 3309

differences reached statistical significance at age 24 months Additional analyses

when corrected for age at measurement [length difference,
0.3 sd (⫽ 0.9 cm), P ⫽ 0.046; weight difference, 0.3 sd (⫽
364 g), P ⫽ 0.022]. Gains in length and weight between
randomization and age 24 months were also greater and
statistically significant in the thyroxine group [31.3 (3.0) vs.
30.2 (3.1) cm, difference, 1.1; P ⫽ 0.016; and 7550 (1101) vs.
7172 (1102) g, difference, 378; P ⫽ 0.022]. Throughout the
study, there were no significant differences in head
circumference.
Thyroxine-treated children had a somewhat higher KID
total raw score than placebo-treated children at all assess-
After exclusion of the 13 children with central nervous
system disease (at baseline and comorbidity during the
study), the differences in mental and motor developmental
age delay between the thyroxine and placebo-treated chil-
dren were 0.8 months (Table 2, additional analysis). These
differences correspond to approximately six mental and
eight motor developmental index points. Both developmen-
tal age delay differences were statistically significant.
Discussion

Downloaded from https://academic.oup.com/jcem/article-abstract/90/6/3304/2870560 by guest on 07 May 2020

ments, with differences increasing over time. However, these
differences did not reach statistical significance. At ages 6
and 12 months BSID-II developmental age delays were sim-
ilar in the two groups.
Comorbidity, cointerventions, and adverse events
Comorbidity and cointerventions did not significantly dif-
fer between the two treatment groups, except that in the
thyroxine group five children were diagnosed with “other
central nervous system disease” (Table 4). In both groups,
three children were diagnosed with infantile spasms.
During the study, stool frequency, number of sleeping
hours, heart rate, and anterior fontanel size were similar in
the two groups and showed a natural developmental pattern
(data not shown).
TABLE 4. Comorbidity and cointerventions during the study of the Down syndrome children included in the primary outcome analysis
In this randomized trial, we found that thyroxine treat-
ment of a representative sample of Down syndrome children
during their first 2 yr of life resulted in modest improvements
in motor development and growth. This supports our hy-
pothesis that the mild plasma TSH elevation encountered in
almost all young Down syndrome children and the lowered
thyroxine concentrations in Down syndrome neonates at the
end of their first week of life indicate genuine hypothyroid-
ism that is disadvantageous to their early development and
growth (11). Although the clinical significance of the im-
provement in growth may be disputed, the improvement in
motor development seems more relevant. The magnitude of
the improvement (0.7 months, corresponding to approxi-
mately seven motor developmental index points) compares
to the small developmental delays of four to 10 intelligence

Thyroxine group (n ⫽ 90) Placebo group (n ⫽ 91) P

Comorbidity
Hospital admission [no. (%)] 45 (50.0) 48 (52.7) 0.65
No. of admission days 20.1 (20.1) 19.7 (25.4) 0.92
No. of admissions 2.0 (1.3) 2.1 (1.8) 0.81
Missing data 2 (2.2) 3 (3.3)
Central nervous system disease [no. (%)]a
Infantile spasms 3 (3.3) 3 (3.3) ⬎0.99
Otherb 5 (5.6) 0 0.059
Missing data 2 (2.2) 3 (3.3)
Heart surgery for [no. (%)]
AVSD or TOF 13 (14.4) 8 (8.8) 0.25
VSD or ASD or PDA 5 (5.6) 9 (9.9) 0.27
Missing data 0 1 (1.1)
Abdominal surgery [no. (%)]b 4 (4.4) 4 (4.4) ⬎0.99
Missing data 2 (2.2) 0
Other disease [no. (%)]
Celiac disease 0 2 (2.2) 0.50
Diabetes 1 (1.1) 0 ⬎0.99
Tube feeding after study visit one [no. (%)] 10 (11.1) 14 (15.4) 0.40
Total duration (months) 2.8 (4.8) 4.5 (8.2) 0.55
Cointerventions
Routine checks and/or medical procedures [no. (%)]
Eye 75 (83.3) 75 (82.4) ⬎0.99
Ear-nose-throat 79 (87.8) 81 (89.0) 0.60
Tympanostomy tube insertion 26 (28.9) 29 (31.9) 0.63
Missing data 2 (2.2) 3 (3.3)
Developmental stimulation [no. (%)]
Developmental stimulation program 61 (67.8) 66 (72.5) 0.40
Physical therapy 82 (91.1) 82 (90.1) ⬎0.99
Speech therapy 64 (71.1) 66 (72.5) 0.73
Missing data 2 (2.2) 3 (3.3)
Data are mean (SD), unless noted otherwise. Missing data are data that were not reported and were not compared. AVSD, Atrioventricular
septal defect; TOF, tetralogy of Fallot; VSD, ventricular septal defect; ASD, atrial septal defect; PDA, patent ductus arteriosus.
a
Arnold-Chiari malformation, prenatal cerebral infarction, postnatal cerebral infarction, hypoxia after cardiac surgery, meningitis.
b
Duodenal web and Hirschsprung disease.
3310 J Clin Endocrinol Metab, June 2005, 90(6):3304 –3311
quotient points found in (non-Down syndrome) children of
mothers who were hypothyroid or hypothyroxinemic at the
end of the first trimester of their pregnancy, and for which
(mass) screening and thyroxine treatment during pregnancy
has been proposed (18 –20). Because the earlier attainment of
(just) a few developmental milestones may be even more
important for a child with a substantial developmental delay
(like in Down syndrome) than for a normally developing
child, thyroxine treatment of young Down syndrome chil-
dren may be beneficial to maximize their early development
and growth.
Until now, it was actually unknown whether the mild
hypothyroidism that young Down syndrome children expe-
rience affects their thyroid hormone-dependent brain devel-
opment. Several observational studies and (thyroid hormone
vs. placebo) trials in Down syndrome lacked a proper design
to answer this question (no information on the thyroid hor-
mone state during the first months of life, triiodothyronine
instead of thyroxine, study participants too old, and no de-
velopmental assessment) (21–25). Well-designed random-
ized trials or observational studies in young non-Down
syndrome children with mild sporadic congenital hypothy-
roidism have not been conducted. So, the results of our trial
not only provide the first proof of harm done by untreated
mild congenital hypothyroidism in young Down syndrome
children, but they may also be a clue to the possible negative
effects of mild congenital hypothyroidism not detected by
neonatal screening in young children without Down
syndrome.
Our study had some limitations. Although our thyroxine-
dosing strategy resulted in plasma thyroid hormone and
TSH concentrations comparable to those of thyroxine-treated
children with thyroidal (primary) congenital hypothyroid-
ism (26), treatment could not be started until the mean age
of 24 d. If treatment could have been initiated directly after
the clinical diagnosis of Down syndrome, it is possible that
a greater effect could have been achieved (27). Furthermore,
the thyroid hormone (steady) state of the thyroxine-treated
children differed substantially from the placebo-treated chil-
dren at the moment of developmental testing, raising the
question whether this also might have influenced test per-
formances. However, to our best knowledge, abnormally
high as well as abnormally low thyroid hormone concentra-
tions only have a negative influence on performance on tests
of attention (28). Therefore, we feel that the differences in
development, like the differences in somatic growth, must be
ascribed to the long-term, central nervous system develop-
ment and maturation promoting effects of thyroxine rather
than to its direct metabolic effects. Finally, an important issue
to address is the duration of the developmental follow-up. A
criticism could be that 2 yr is too short to draw any conclu-
sions on the possible long-term benefits of our intervention.
However, several long-term follow-up studies in early-
treated children with moderate to severe congenital hypo-
thyroidism have shown that developmental outcome later in
life is strongly correlated to early developmental assessments
(29 –32). We could not think of any argument that this is
different in Down syndrome children with mild congenital
hypothyroidism. Yet, only thorough developmental fol-
low-up of this cohort of children will reveal the actual long-
van Trotsenburg et al. • Thyroxine Therapy in Down Syndrome
term benefits of early childhood thyroxine treatment. In this
follow-up, the use of neuropsychological tests that are more
sensitive than the BSID-II in detecting specific hypothyroid-
ism-related defects in early brain development may show
which neuropsychological domains benefit most from early
thyroxine treatment (28).
In summary, the data of our study are consistent with the
presumption that the mild hypothyroidism of young Down
syndrome children is not a harmless phenomenon and pro-
vide evidence to support the hypothesis that thyroxine treat-
ment may improve development and growth of young Down
syndrome children. Given the observed benefits, the absence
Downloaded from https://academic.oup.com/jcem/article-abstract/90/6/3304/2870560 by guest on 07 May 2020
of adverse events, and the low treatment costs, we feel that
thyroxine treatment should be considered in Down syn-
drome neonates to maximize their early development and
growth.
Acknowledgments
We are indebted to all Down syndrome children and their parents for
their participation in the study; to Organon Inc. for providing the study
medication; to Erik A. B. de Graaf and the Dutch Down Syndrome
Foundation, and the Dutch pediatricians for providing and caring for the
participating children; to Marjo J. Geerlings, Marijke Dekker-van der
Sloot, Marlies J. E. Kempers, Hanneke M. van Santen, Kamil Wojciec-
howicz, and Richard Schol for collecting data; and to Raoul C. Hen-
nekam for critically reviewing the study proposal.
Received January 20, 2005. Accepted March 1, 2005.
Address all correspondence and requests for reprints to: Paul van
Trotsenburg, Academic Medical Center, University of Amsterdam, De-
partment of Pediatric Endocrinology, P.O. Box 22700, 1100 DE Amster-
dam, The Netherlands. E-mail: a.s.vantrotsenburg@amc.uva.nl.
This work was supported by The Netherlands Organization for
Health Research and Development Grant 2100.0025.
References
1. Klein RZ, Mitchell ML 2000 Hypothyroidism in infants and children. Neo-
natal screening. In: Braverman LE, Utiger RD, eds. Werner, Ingbar’s the thy-
roid: a fundamental and clinical text. 8th ed. Philadelphia: Lippincott Williams
& Wilkins; 973–988
2. Derksen-Lubsen G, Verkerk PH 1996 Neuropsychologic development in
early treated congenital hypothyroidism: analysis of literature data. Pediatr
Res 39:561–566
3. Daliva AL, Linder B, DiMartino-Nardi J, Saenger P 2000 Three-year fol-
low-up of borderline congenital hypothyroidism. J Pediatr 136:53–56
4. Kohler B, Schnabel D, Biebermann H, Gruters A 1996 Transient congenital
hypothyroidism and hyperthyrotropinemia: normal thyroid function and
physical development at the ages of 6 –14 years. J Clin Endocrinol Metab
81:1563–1567
5. Rapaport R 2000 Congenital hypothyroidism: expanding the spectrum. J Pe-
diatr 136:10 –12
6. Sharav T, Collins Jr RM, Baab PJ 1988 Growth studies in infants and children
with Down’s syndrome and elevated levels of thyrotropin. Am J Dis Child
142:1302–1306
7. Toledo C, Alembik Y, Dott B, Finck S, Stoll C 1997 Anomalies du fonction-
nement thyroidien des enfants trisomiques 21: Anomalies of thyroid function
in children with Down syndrome. Arch Pediat 4:116 –120
8. Jimenez-Lopez V, Arias A, Arata-Bellabarba G, Vivas E, Delgado MC, Paoli
M 2001 Concentration of thyrotropic hormone and free thyroxin in children
with Down’s syndrome. Invest Clin 42:123–130
9. Rubello D, Pozzan GB, Casara D, Girelli ME, Boccato S, Rigon F, Baccichetti
C, Piccolo M, Betterle C, Busnardo B 1995 Natural course of subclinical
hypothyroidism in Down’s syndrome: prospective study results and thera-
peutic considerations. J Endocrinol Invest 18:35– 40
10. Crino A, Ciampalini P, DiGilio MC, Obregon MG, Giannotti A, Borrelli P
1993 Abnormal thyrotropin secretion in Down’s syndrome. In: Castells S,
Wisniewski KE, eds. Growth hormone treatment in Down’s syndrome. New
York: John Wiley and Sons; 101–109
11. van Trotsenburg ASP, Vulsma T, van Santen HM, Cheung W, de Vijlder JJM
2003 Lower neonatal screening thyroxine concentrations in Down syndrome
newborns. J Clin Endocrinol Metab 88:1512–1515
van Trotsenburg et al. • Thyroxine Therapy in Down Syndrome
12. Epstein CJ 2001 Down syndrome (trisomy 21). In: Scriver CR, Beaudet AL, Sly
WS, Valle D, Childs B, Kinzler KW, Vogelstein B, eds. The metabolic and
molecular bases of inherited disease. 8th ed. New York: McGraw-Hill; 1223–
1256
13. Fredriks AM, van Buuren S, Burgmeijer RJ, Meulmeester JF, Beuker RJ,
Brugman E, Roede MJ, Verloove-Vanhorick SP, Wit JM 2000 Continuing
positive secular growth change in The Netherlands 1955–1997. Pediatr Res
47:316 –323
14. Bayley N 1993 Bayley Scales of Infant Development: manual. 2nd ed. San
Antonio, TX: Psychological Corp.
15. Schneider MJ, Cools GMP, Reuter J 1990 KID-N. Kent Infant Development
Scale Nederlandse bewerking. Lisse, The Netherlands: Swets and Zeitlinger
16. Pocock SJ 1988 Clinical trials: a practical approach. Chichester, UK: John Wiley
and Sons; 123–138
17. Brent GA, Larsen P Reed 2000 Treatment of hypothyroidism. In: Braverman
LE, Utiger RD, eds. Werner, Ingbar’s the thyroid: a fundamental and clinical
text. 8th ed. Philadelphia: Lippincott Williams & Wilkins; 853– 858
18. Haddow JE, Palomaki GE, Allan WC, Williams JR, Knight GJ, Gagnon J,
O’Heir CE, Mitchell ML, Hermos RJ, Waisbren SE, Faix JD, Klein RZ 1999
Maternal thyroid deficiency during pregnancy and subsequent neuropsycho-
logical development of the child. N Engl J Med 341:549 –555
19. Pop VJ, Kuijpens JL, van Baar AL, Verkerk G, van Son MM, de Vijlder JJ,
Vulsma T, Wiersinga WM, Drexhage HA, Vader HL 1999 Low maternal free
thyroxine concentrations during early pregnancy are associated with impaired
psychomotor development in infancy. Clin Endocrinol (Oxf) 50:149 –155
20. Pop VJ, Brouwers EP, Vader HL, Vulsma T, van Baar AL, de Vijlder JJ 2003
Maternal hypothyroxinaemia during early pregnancy and subsequent child
development: a 3-year follow-up study. Clin Endocrinol (Oxf) 59:282–288
21. Pueschel SM, Pezzullo JC 1985 Thyroid dysfunction in Down syndrome. Am J
Dis Child 139:636 – 639
JCEM is published monthly by The Endocrine Society (http://www.endo-society.org), the foremost professional society serving the
endocrine community.
J Clin Endocrinol Metab, June 2005, 90(6):3304 –3311 3311
22. Selikowitz M 1993 A five-year longitudinal study of thyroid function in
children with Down syndrome. Dev Med Child Neurol 35:396 – 401
23. Koch R, Share J, Graliker B 1965 The effects of cytomel on young children with
Down’s syndrome (mongolism): a double-blind longitudinal study. J Pediatr
66:776 –778
24. Tirosh E, Taub Y, Scher A, Jaffe M, Hochberg Z 1989 Short-term efficacy of
thyroid hormone supplementation for patients with Down syndrome and
low-borderline thyroid function. Am J Ment Retard 93:652– 656
25. Sharav T 1993 Thyroid dysfunction and growth retardation in Down’s syn-
drome. In: Castells S, Wisniewski KE, eds. Growth hormone treatment in
Down’s syndrome. New York: John Wiley and Sons; 73–91
26. Fisher DA, Schoen EJ, La Franchi S, Mandel SH, Nelson JC, Carlton EI, Goshi
JH 2000 The hypothalamic-pituitary-thyroid negative feedback control axis in
children with treated congenital hypothyroidism. J Clin Endocrinol Metab
85:2722–2727
27. Van Vliet G 2001 Treatment of congenital hypothyroidism. Lancet 358:86 – 87
28. Rovet JF 2002 Congenital hypothyroidism: an analysis of persisting deficits
Downloaded from https://academic.oup.com/jcem/article-abstract/90/6/3304/2870560 by guest on 07 May 2020
and associated factors. Neuropsychol Dev Cogn Sect C Child Neuropsychol
8:150 –162
29. Simons WF, Fuggle PW, Grant DB, Smith I 1994 Intellectual development at
10 years in early treated congenital hypothyroidism. Arch Dis Child 71:232–234
30. Oerbeck B, Sundet K, Kase BF, Heyerdahl S 2003 Congenital hypothyroidism:
influence of disease severity and l-thyroxine treatment on intellectual, motor,
and school-associated outcomes in young adults. Pediatrics 112:923–930
31. Heyerdahl S, Oerbeck B 2003 Congenital hypothyroidism: developmental
outcome in relation to levothyroxine treatment variables. Thyroid 13:1029 –
1038
32. Simoneau-Roy J, Marti S, Deal C, Huot C, Robaey P, Van Vliet G 2004
Cognition and behavior at school entry in children with congenital hypothy-
roidism treated early with high-dose levothyroxine. J Pediatr 144:747–752