| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Community Health Sciences (F.L.R.W., S.A.O.) and Maternal and Child Health Sciences (J.S., C.D., R.H.), University of Dundee, Ninewells Hospital and Medical School, Dundee DD1 9SY, United Kingdom; Department of Pediatrics (J.J.B.-S.), Division of Endocrinology, Erasmus University Medical Center/Sophia Children’s Hospital, 3015 GJ Rotterdam, The Netherlands; Department of Paediatrics (N.M.), The Rotunda Hospital, Parnell Square, Dublin 1, Ireland; Department of Internal Medicine (H.v.T., T.J.V.), Erasmus University Medical Center, 3015 GE Rotterdam, The Netherlands; and Nuclear Medicine Service (S.-Y.W.), Veterans Affairs Medical Center, University of California–Irvine Medical Center, Long Beach, California 90822
Address all correspondence and requests for reprints to: Professor Robert Hume, Maternal and Child Health Sciences, University of Dundee, Ninewells Hospital and Medical School, Dundee DD1 9SY, Scotland, United Kingdom. E-mail: r.hume{at}dundee.ac.uk.
 |
Abstract |
|---|
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The comparison of preterm infant cord and postnatal thyroid hormone sera levels with term infant and/or adult values has established that these groups have important differences in sera levels of iodothyronines and related parameters (1, 7, 8, 9, 10). The accepted definition of preterm birth is one at less than 37 wk completed gestation, but the lower limit of viability extends to around 23 wk gestation. It is therefore not surprising that thyroid hormone levels differ in preterm infants grouped by gestational age (11, 12). This makes it essential for observational studies, or comparative studies of treatment regimens, to analyze and interpret data from groups based on relatively narrow gestational age ranges (13, 14, 15). To date, longitudinal studies of postnatal thyroid hormone sera levels in gestationally grouped preterm infants are limited because they are restricted in number and also vary in the range and nature of the serum parameters analyzed (14, 16, 17, 18). Other studies are limited to the early postnatal period (12) or to measurement of a single iodothyronine (19). Indeed, the three seminal studies that established the statistical relationship between neurodevelopmental outcome and hypothyroxinemia were based on a single measurement of postnatal T4 level (3, 4, 5). The pathophysiological processes of transient hypothyroxinemia in preterm infants are complex, and to inform further understanding this requires a longitudinal postnatal study of sufficient power and with measurement of a comprehensive range of thyroid hormones in gestationally restricted age groups.
The definition of optimal ranges for serum thyroid hormones in preterm infants related to gestation and postnatal age has not been resolved. Preterm infant reference ranges should be derived from populations of gestationally restricted age groups (12, 20). But are these reference ranges optimal for the maturation of critical organs such as brain? Ideally, and in the longer term, measures of physiological function, such as long-term neurodevelopmental outcome, related to postnatal thyroid hormone levels should be used to determine the optimal postnatal serum thyroid hormone levels. Such studies have begun (18). In the interim, other means of informing the definition of optimal levels need to be explored. We propose that the relationship of postnatal to cord levels of iodothyronines, T4-binding globulin (TBG), and TSH may be assessed only once the cord levels have been corrected to equivalent gestational age had the fetus remained in utero.
|
Subjects and Methods |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
All infants had intensive care support as required, including intermittent positive ventilation and, where appropriate, correction of fluid, electrolyte, blood glucose, and acid-base abnormalities. Blood pressure was supported with inotropes, plasma, or other blood products as required. Infants with significant persistence of the ductus arteriosus were treated with diuretics and indomethacin, or surgical ligation if appropriate. Parenteral nutrition regimens, if required, were based on a solution of electrolytes, 10% dextrose, amino acids (Vaminolact, Fresenius Kabi, Cheshire, UK), a phosphate supplement (Addiphos, Fresenius Kabi), water-soluble vitamins (Solvito N, Fresenius Kabi), and trace elements (Peditrace, Fresenius Kabi) to levels recommended by the manufacturer. In tandem, a fat emulsion solution (Intralipid 20%, Fresenius Kabi) with added fat-soluble vitamins (Vitlipid, Fresenius Kabi) was used. Enteral feeds were started when the condition of the infant was stable. Thereafter, enteral feed volumes were gradually increased as determined by the infants’ clinical condition, with reciprocal reductions in the volume of parenteral nutrition infused.
The study was approved, as appropriate, by the Multicenter Research Ethics Committee (Edinburgh, UK) and the Tayside Committee on Medical Research Ethics; in all cases, written informed consent was obtained.
Cord blood (n = 695) was collected into a tube without anticoagulant as soon as possible after delivery of the live-born infants (23–42 wk gestation) from fetal vessels running over the placental surface. Blood was also collected at postnatal d 7, 14, and 28. The blood samples were allowed to clot for at least 15 min and then centrifuged at 4000 rpm for 5 min. If collected outside of normal laboratory hours, the blood was stored at 4 C (maximum 12 h) before processing. The serum was removed, stored, and transported at a maximum of –20 C for assays in one laboratory (T.J.V.).
Provided sufficient serum was available, T4, free T4 (FT4), TSH, T3, rT3, T4 sulfate (T4S), and TBG levels were determined. Serum T4, T3, and rT3 were measured by in-house RIA; FT4 by Vitros ECI technology (Ortho-Clinical Diagnostics, Amersham, UK); TSH by Dynotest immunoradiometric assay; and TBG by Dynotest RIA (Brahms, Berlin, Germany). T4S was prepared by the method of Eelkman Rooda et al. (22). The measurements of T4S in serum were done by a specific antibody, as described previously (23). Within-assay coefficients of variation were calculated as 2–8% for T4, 3–7% for FT4, 2–6% for T3, 3–4% for rT3, 6–17% for T4S, 2–5% for TSH, and 2–4% for TBG. Between-assay coefficients of variation were 5–10% for T4, 5–10% for FT4, 8% for T3, 9–14% for rT3, 4–19% for T4S, 2–14% for TSH, and 2–3% for TBG.
Because ratios tend not to follow a Gaussian distribution, all ratios were log transformed, so that the data approximated a normal distribution; this allowed the determination of group means and SE of the means. The median values are given in parentheses for TSH because it has a skewed distribution; however, the data are presented untransformed so that they can be readily compared with other published studies. The data were subdivided into four gestational age groups: 23–27, 28–30, 31–34, and 37 or more weeks. Infants born at or under 30 wk gestation constitute the extreme preterm infant.
All the graphs were constructed as follows. For individual gestational weeks between 23 and 42 wk, the mean and twice the SEM of cord blood levels was plotted for each iodothyronine, TSH, and TBG. This baseline graph shows the joined and smoothed lines of plus twice the SEMs and minus twice the SEMs. The data were grouped by gestational age at delivery into 23–27 (n = 101), 28–30 (n = 196), and 31–34 (n = 253) wk. The mean and ± twice the SEM were calculated for each iodothyronine, TSH, and TBG for cord and for postnatal d 7, 14, and 28. When the SE bars do not overlap, the difference between (any) two means may be assumed to be statistically different (P < 0.05).
The mean gestational age at birth was determined for each of the groups: 25.8 wk for the 23- to 27-wk group, 29.2 wk for 28- to 30-wk group, and 32.5 wk for the 31- to 34-wk group. The postnatal data at d 7, 14, and 28 were plotted onto the baseline graph starting at the appropriate mean birth gestation for each group. For example, the most immature group started at 26 wk gestation and d 7 was plotted against 27 wk, d 14 against 28 wk, and d 28 against 30 wk. Thus, data for any given particular postnatal age (i.e. 7, 14, or 28 d) was plotted and described in relation to the equivalent gestational age had the fetus remained in utero (referred to hereafter as equivalent gestational age). The assumption we make is that the cord levels of iodothyronines, TBG, and TSH reflect those of the normal fetus in utero of equivalent gestational age and who progress to a term delivery.
|
Results |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
). The retention of these subdivisions is important in our analysis of this data set, because both Murphy et al. (24) and van Wassenaer et al. (15, 18) have described discontinuities in thyroid hormone responsiveness and metabolism with extreme prematurity.
|
and Fig, 1) (21). The postnatal elevation in serum T4 is attenuated in 31- to 34-wk infants, absent in 28- to 30-wk infants (although levels are similar to cord values of equivalent gestational age), and reversed in 23- to 27-wk infants at d 7 and 14 with hypothyroxinemia relative to other preterm and term groups and to cord values of equivalent gestational age (Table 1 and Fig. 1).
|
|
and Fig. 1).
The trend in postnatal TBG levels in all preterm infant groups is to increase and generally to follow, with minor variations, the equivalent gestational age limits (Fig 1). In term infants, postnatal TBG levels are similar to cord values at birth (Table 1 and Fig. 1).
The postnatal log T4:TBG ratios (Fig. 1) show similar trends, within the gestational age groups, to the pattern of postnatal T4 levels. In term infants, postnatal log T4:TBG ratios increase above cord values of equivalent gestational age. The postnatal elevation in serum log T4:TBG is similar in 31- to 34-wk infants, attenuated in 28- to 30-wk infants (although levels are similar to cord values of equivalent gestational age), and reversed in 23- to 27-wk infants (Fig. 1).
The TSH values were similar in all gestational groups in cord and at all ages postnatally; the values were highest in cord, fell by d 7, and remained constant thereafter (Fig. 1). TSH levels were lower (with two exceptions) at all ages postnatally than cord values of equivalent gestational age.
Cord log TSH:FT4 ratios are similar in all groups including term infants. At d 7 and 14, the 23- to 27- and 28- to 30-wk gestational groups have higher ratios than the 31- to 34-wk and term infants. The log TSH:FT4 ratios in nonpregnant women [mean, –1.13, and twice SEM, 0.04 (21)] are identical to the postnatal values at d 7, 14, and 28 in term infants (Fig. 1).
In all preterm infants, mean serum T3 levels rise appreciably from similar cord values in all gestational groups, but the magnitude of the increments are related to gestational and postnatal age. Mean serum T3 values show a remarkable continuity of apparent linear increments in postnatal levels extending through the gestational groups to term. In effect, all preterm infant groups from d 7 are exposed to increasingly higher serum levels of T3 than cord levels of equivalent gestational ages (except d 14 in the 23- to 27-wk group) (Table 1 and Fig. 1).
The rT3 values at all postnatal ages and in all groups are substantially below the cord values of the equivalent gestational ages (Table 1 and Fig. 1).
In all our preterm groups, postnatal T4S levels increase, with maximal values measured in d 7 sera. The 23- to 27-wk gestation group are distinctive because they not only have the highest d 7 mean serum level, but in contrast to the more mature preterm groups (28–30 and 31–34 wk gestation), mean T4S levels at postnatal d 14 and 28 remain generally elevated above cord levels of equivalent gestational ages.
|
Discussion |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Thyroid hormones are essential for adaptation to extrauterine life, including lung development and fluid control (27), adaptive thermogenesis (28), and a diverse range of metabolic processes in liver including gluconeogenesis (29). In thyroidectomized rats, only combined treatment with T4 and T3 ensures euthyroidism in all tissues (30). An adequacy of T3 and T4 may similarly be important for the maturation of essential functions in human infants.
In term infants, serum T3 levels increase approximately 3-fold from cord values by 3–4 wk postnatal age (11, 31) to values of 2.5–3.0 nmol/liter (31). Preterm infants of 32–33 wk gestation also increase postnatal T3 levels by approximately the same magnitude albeit from lower cord levels (7, 8, 19). In preterm infants less than 32 wk gestation, T3 changes are more variable; some series record no postnatal changes over 2 wk (32) or even 6 wk (16); in other studies, T3 levels increased postnatally (13, 14, 15). Serum T3 levels rise appreciably from similar cord values in all our preterm groups. Importantly, in our 23- to 27-wk group, postnatal increments in T3 occur at a time when mean T4 (but not FT4 levels) are decreased below cord levels of equivalent gestational age. T3 levels can be increased in preterm infants of less than 30 wk gestation by T3 supplementation, but this decreases FT4 levels (33). Even a single administration of T3 to infants of less than 28 wk gestation, although increasing T3, results in a transient decrease in FT4 (34). Maintaining serum FT4 levels in extreme preterm infants may be a priority for sustaining postnatal brain development and neurodevelopmental outcome (35) because sufficient circulatory FT4, rather than total T4, is necessary for T3 generation within brain cells (36).
Both T4 and TBG in cord serum increase over the range 15–42 wk gestation (21). Log T4:TBG ratios are low at early gestations but increase until late second trimester, as temporal increments in T4 are greater than those of TBG (21). This is in agreement with the data shown by Greenberg et al. (37). In the third trimester, cord T4 and TBG levels both increase, but the log T4:TBG ratio is now constant (21). In our 23- to 27-wk group, postnatal reductions in log T4:TBG ratios, to levels equivalent to earlier gestations, are the result of a larger decrease in T4 than TBG levels. In contrast, the 31- to 34-wk group, where T4 and TBG levels increase postnatally, the log T4:TBG ratios also increase substantially, a result of increased T4 levels to values above those equivalent to the corrected gestational ages.
Until the late second trimester, the developmental pattern of log T4:TBG ratios shows a remarkable similarity to that of FT4; both increase with gestational age. Thereafter, log T4:TBG ratios are constant, but FT4 levels peak at 31–34 wk and decrease to term (21). In our postnatal preterm infants, this relationship of log T4:TBG ratios to FT4 levels shows additional inconsistencies. The reasons for this may be multiple and vary in different gestational age groups and postnatal ages. In our term and 31- to 34-wk infants, T4 levels increase postnatally with almost constant TBG levels. As a consequence, in this group, log T4:TBG ratios increase above gestational age-related limits to levels greater than maternal and nonpregnant women (21), and as a result, FT4 levels also increase. In contrast, in the 23- to 27-wk group, T4 levels decrease postnatally against a background of postnatal TBG levels that are maintained within limits corrected for gestational age. As a result, postnatal log T4:TBG ratios decrease below gestational age-related limits. In these circumstances, it was expected that FT4 levels would also decrease as a feature of transient hypothyroxinemia (1, 15, 16). In fact FT4 levels in our 23- to 27-wk gestation group remain within cord values of equivalent gestational age, levels apparently adequate to allow normal brain development in utero assuming these cord iodothyronine levels reflect those of the normal fetus in utero of equivalent gestational age and progressing to a term delivery.
The depression of T4 levels and maintenance of FT4 levels at 2 wk postnatal age shown in our 23- to 27-wk group has been previously described in a group of very low birth weight infants (<1500 g), changes that were exaggerated by the most severely ill subgroup of five infants (38). FT4 levels in that study, as determined by the two-stage method of AutoDelfia, were 67% of those of equilibrium dialysis FT4 levels (38). The measurement of sera FT4 by commercial kits is problematic, overestimating the FT4 levels at high serum binding protein concentrations and underestimating FT4 at low protein concentrations (39), and there are differences in FT4 estimation between the kits (40). It is possible that FT4 levels in our study were also similarly underestimated, although the Vitros ECI FT4 method has one of the best concordances with the dialysis method (40) (van Toor, H., and T. J. Visser, unpublished data). This underestimation will also apply to other studies in preterm infants where postnatal FT4 levels have been measured by commercial kits (1, 15, 17). Calculation of the FT4 index using T4:TBG ratios is also problematic; TBG binding capacity can be altered independent of changes in concentrations of TBG protein (39) in agreement with our data.
Displacement of T4 from TBG by drugs, metabolites, and free fatty acids (41, 42) is a possible explanation for the increases in FT4 relative to the T4:TBG ratio evident in our most immature group. Elevated levels of free fatty acids, particularly when serum albumin is low, as is common in preterm infants (43), can displace protein-bound T4 and elevate FT4 (44). In our study, parenterally fed newborns are infused with Intralipid at levels previously shown to generate average ratios of free fatty acids to albumin of 12 or more; the ratio needed to inhibit protein binding in vitro is approximately 3 (44, 45). The patency of arterial and venous catheters is maintained by heparin infusions; this activates lipoprotein lipase, which can generate free fatty acids in postvenesection serum and falsely augment FT4 levels in vitro (46). Extreme preterm infants are more likely to have conditions necessary for displacement of protein-bound T4, such as parenteral nutrition, lower serum albumin levels, heparin infusions, and other drug usage. In preterm infants, serum FT4 levels increase relative to those of T4 in the first hour of life and before exposure to heparin or iv triglyceride emulsions (24); this suggests endogenous factors influence FT4 levels. An additional explanation for increases in FT4 levels may be the result of polymorphonuclear leukocyte elastase cleavage of TBG during sepsis and inflammation liberating bound T4 (47). Polymorphonuclear leukocyte elastase activity is 8-fold higher in cord blood than maternal sera (48). Cleaved TBG is present in cord blood and in the absence of sepsis may be part of the physiological inflammatory response in the newborn (49).
Cord log TSH:FT4 ratios are similar in our preterm gestational groups and term infants. This is in contrast to the work of Fisher et al. (50), who used combined data from multiple individual studies and where cord values of very low birth weight infants were higher than those of term infants. Our postnatal log TSH:FT4 ratios in the 23- to 27- and 28- to 30-wk groups are significantly higher at 7 and 14 d compared with the 31- to 34-wk groups and term infants. This pattern is in contrast to the data interpreted by Fisher et al. (50), which is relatively constant in the first postnatal week in groups of infants from 25–42 wk gestation. The mean postnatal log TSH:FT4 of term infants is similar to that of our nonpregnant women (21), which is very different from the work of Fisher et al. (50), who suggest that TSH:FT4 does not reach adult levels until the early 20s. However, as with all ratios, changes in each parameter, provided they happen together because of a change in the set point, will not be reflected in a change in the ratio.
Cord T4S levels increase through the second trimester to reach a peak in late second to early third trimester before the last-trimester decline in T4S levels (21). In this study, T4S levels transiently increase postnatally before declining; the explanation for these changes is not obvious. For example, the contribution of different sulfotransferase isoforms and tissues to T4S serum levels is not known, or even whether the contribution changes through fetal life, or postnatally (51, 52). T4S is exclusively metabolized to inactive rT3S; elevation in serum T4S levels is associated with reduced type 1 iodothyronine deiodinase activity (53), but there is no supportive evidence for this scenario in studies of human hepatic deiodinase development (54).
We have previously shown that the late second/early third trimester is a critical transition period in fetal thyroid hormone metabolism, which may be interrupted by preterm birth and contribute to postnatal thyroid dysfunction (21, 24). In this present study, we provide more evidence that the 23- to 27-wk group of infants is particularly distinctive from other preterm groups; they are hypothyroxinemic on T4 levels, yet FT4 levels are equivalent to gestational age values, and T3 increases to levels above equivalent gestational age values.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
Abbreviations: FT4, Free T4; TBG, T4-binding globulin; T4S, T4 sulfate.
1 For a list of members, see Acknowledgments. 
Received May 10, 2004.
Accepted July 26, 2004.
|
References |
|---|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  N. Forghani and T. Aye Hypothyroxinemia and Prematurity NeoReviews, February 1, 2008; 9(2): e66 - e71. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. J. Forhead, J. K. Jellyman, D. S. Gardner, D. A. Giussani, E. Kaptein, T. J. Visser, and A. L. Fowden Differential Effects of Maternal Dexamethasone Treatment on Circulating Thyroid Hormone Concentrations and Tissue Deiodinase Activity in the Pregnant Ewe and Fetus Endocrinology, February 1, 2007; 148(2): 800 - 805. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. J. Forhead, K. Curtis, E. Kaptein, T. J. Visser, and A. L. Fowden Developmental Control of Iodothyronine Deiodinases by Cortisol in the Ovine Fetus and Placenta Near Term Endocrinology, December 1, 2006; 147(12): 5988 - 5994. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. J. E. Kempers, C. I. Lanting, A. F. J. van Heijst, A. S. P. van Trotsenburg, B. M. Wiedijk, J. J. M. de Vijlder, and T. Vulsma Neonatal Screening for Congenital Hypothyroidism Based on Thyroxine, Thyrotropin, and Thyroxine-Binding Globulin Measurement: Potentials and Pitfalls J. Clin. Endocrinol. Metab., September 1, 2006; 91(9): 3370 - 3376. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. P. Peeters, M. H. A. Kester, P. J. Wouters, E. Kaptein, H. van Toor, T. J. Visser, and G. Van den Berghe Increased Thyroxine Sulfate Levels in Critically Ill Patients as a Result of a Decreased Hepatic Type I Deiodinase Activity J. Clin. Endocrinol. Metab., December 1, 2005; 90(12): 6460 - 6465. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. L. R. Williams, S. A. Ogston, H. van Toor, T. J. Visser, R. Hume, and with collaboration from the Scottish Preterm Thyro Serum Thyroid Hormones in Preterm Infants: Associations with Postnatal Illnesses and Drug Usage J. Clin. Endocrinol. Metab., November 1, 2005; 90(11): 5954 - 5963. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. G. van Wassenaer, J. Westera, B. A. Houtzager, and J. H. Kok Ten-Year Follow-up of Children Born at <30 Weeks' Gestational Age Supplemented With Thyroxine in the Neonatal Period in a Randomized, Controlled Trial Pediatrics, November 1, 2005; 116(5): e613 - e618. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. den Brinker, K. F. M. Joosten, T. J. Visser, W. C. J. Hop, Y. B. de Rijke, J. A. Hazelzet, V. H. Boonstra, and A. C. S. Hokken-Koelega Euthyroid Sick Syndrome in Meningococcal Sepsis: The Impact of Peripheral Thyroid Hormone Metabolism and Binding Proteins J. Clin. Endocrinol. Metab., October 1, 2005; 90(10): 5613 - 5620. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. L. R. Williams, G. J. Mires, C. Barnett, S. A. Ogston, H. van Toor, T. J. Visser, R. Hume, and with collaboration from the Scottish Preterm Thyro Transient Hypothyroxinemia in Preterm Infants: The Role of Cord Sera Thyroid Hormone Levels Adjusted for Prenatal and Intrapartum Factors J. Clin. Endocrinol. Metab., August 1, 2005; 90(8): 4599 - 4606. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Fisher Next Generation Newborn Screening for Congenital Hypothyroidism? J. Clin. Endocrinol. Metab., June 1, 2005; 90(6): 3797 - 3799. [Full Text] [PDF]
|
|
|
|
|
|
J. Simpson, F. L. R. Williams, C. Delahunty, H. van Toor, S.-Y. Wu, S. A. Ogston, T. J. Visser, R. Hume, and with collaboration from the Scottish Preterm Thyro Serum Thyroid Hormones in Preterm Infants and Relationships to Indices of Severity of Intercurrent Illness J. Clin. Endocrinol. Metab., March 1, 2005; 90(3): 1271 - 1279. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
