The Hypothroxinemia of Prematurity

Delbert A. Fisher, M.D.

Quest Diagnostics, Nichols Institute San Juan Capistrano, California and Harbor-UCLA Medical Center Torrance, California

Address all correspondence and requests for reprints: Delbert A. Fisher, M.D., 33608 Ortega Highway, San Juan Capistrano, California 92690.


    Introduction
 Top
 Introduction
 References
 
For premature infants the stresses of extrauterine transition are superimposed on an immature thyroid axis (1). Hypothalamic TRH production/secretion is relatively reduced, the thyroid gland response to TSH is not yet mature, the capacity of the thyroid follicular cell to iodinate the tyrosyl residues of thyroglobulin (organify iodine) remains inefficient, and the capacity to convert thyroxine (T4) to active triiodothyronine (T3) is low. Levels of thyroxine binding globulin (TBG) also are relatively low. Thus the premature infant in the extrauterine environment, like the intrauterine fetus of comparable gestational age (GA), is relatively hypothyroxinemic with normal to low levels of serum TSH and T3 concentrations. Serum free thyroxine (FT4) values also are low and thyroglobulin (Tg) levels are high, presumably reflecting increased thyroid gland production of poorly iodinated thyroid hormone precursor. The TSH and T4 responses to TRH are normal, supporting the view that the hypothyroxinemia reflects, in addition to low TBG concentrations, a relative hypothalamic immaturity. Thus the prevailing view has been that the hypothyroxinemia of prematurity is physiologic, not requiring hormone supplementation.

This view has been questioned recently because of several lines of evidence suggesting that the critical period of central nervous system dependency on thyroid hormone in humans, known to extend from birth through 2–3 yr of age, also extends to intrauterine fetal life and to the extrauterine premature infant. This evidence includes 1) the observation in areas of endemic iodine deficiency that iodine treatment of women before pregnancy or up to the end of the second trimester protects the fetal brain from the effects of iodine deficiency; third trimester or neonatal iodine treatment does not improve neurologic status (2); 2) feto maternal Pit-1 deficiency with severe maternal and fetal hypothyroidism is associated with impaired fetal neurological development (3); 3) in premature infants very low levels of blood total T4, measured in newborn thyroid screening programs using filter paper blot samples, are associated with an increased risk of neurologic dysfunction and a reduced IQ (4, 5, 6); and 4) transplacental maternal T4 has been shown to provide normal levels of brain T3 in the hypothyroid fetal rat, helping to explain why neonatal thyroxine treatment of sporadic congenital hypothyroidism is associated with normal mental development (7).

Susana Ares and coworkers (8) in this month’s JCEM (see page 1704), report a detailed study of iodine metabolism and iodine balance in 115 premature infants of 27–36 weeks postmenstrual age (PMA, a measure of GA) weighing 900-4200g. They demonstrate that the more immature (<32 weeks PMA) and sick infants were in negative iodine balance for the first 2–3 neonatal weeks, after which time iodine intake increased progressively. In 69 premature infants with adequate urine iodine excretion data, positive iodine balance (as percent of intake) increased from 30% at 30 weeks to 80% (comparable to term infants) at 42 weeks PMA. Iodine intake in these infants varied from less than 5 to about 40 µg/day for the premature and 30–40 µg/day for the term infants. The current RDA for iodine in infants is more than 30 µg/kg/day, so that optimal iodine intake in the studied infants would have varied from about 30–100 µg/day. Thus the infants in Madrid were somewhat iodine deficient. This would account, at least in part, for the highly positive iodine balance and would argue, as the authors suggest, for an increase in the iodine content of infant formulas, which currently range from 2–17 µg iodine/100 mL (9).

This transient period of negative iodine balance in the infants of less than 32 weeks GA correlates with observations of Rooman et al. in Antwerp (10) and van Wassenaer and coworkers in Amsterdam (11) that infants of less than 30 weeks GA usually manifest a transient fall in FT4 values, with a nadir at 7–14 days. Serum TSH levels did not increase in these infants, and serum FT4 values returned to near cord values by 3–4 weeks. This transient fall in FT4 concentrations was not observed in older premature infants. It seems that these very low birth weight (VLBW) infants are unable to adapt to the extrauterine environment with the augmentation of thyroidal iodine uptake and increased thyroxine secretion characteristic of the older premature and term infants (8).

Comparing measured FT4 levels in their premature infants over the range of 27–40 weeks PMA to published values of Thorpe Beeston and colleagues (12) for the intrauterine fetus of comparable PMA, Ares and coworkers observed that the premature values approximated 50% of the intrauterine fetal levels, whereas term infant FT4 levels were comparable (8). The average FT4 measured by Ares et al. in 30 week PMA infants, as a reference age, approximated 0.5 ng/dL (8). These values were measured by direct nondialysis immunoassay methods that are protein dependent, and all such assays tend to underestimate FT4 concentrations relative to the direct dialysis method (13). Average FT4 concentrations measured by direct immunoassay of cordocentesis or cord blood samples in 30-week GA infants have varied from 0.9–1.4 ng/dL (12, 14, 15). Cord blood FT4 by equilibrium dialysis in the 30 week fetus averaged 2.9 ng/dL (16) In a recent study of 28–30 week premature infants measured at 1 week of postnatal age, the average FT4 value assessed by direct dialysis was 2.0 ng/dL (17). Immunoassay FT4 values in other studies of VLBW infants acclimated to the extrauterine environment have ranged from 1.0–1.5 ng/dL (11, 18).

The very low FT4 values in the Madrid infants are not likely the result of iodine deficiency-induced primary hypothyroidism because the TSH values were lower than those of the intrauterine fetus. Delange and coworkers (19) in Belgium and Frank et al. in New England (20) have shown that severe iodine deficiency in premature infants is associated with marked elevation of serum TSH concentrations. In New England the prevalence of transient hypothyroidism associated with serum TSH values of more than 40 mU/L is highest (0.4%) in VLBW infants (<1500g) (19). The prevalence is 0.2% in LBW infants (1500–2499g), and 0.02% in term infants; permanent hypothyroidism occurs in 0.026% of premature or term infants (19). Because the hypothalamic pituitary axis can respond to hypothyroxinemia with increased TSH secretion, secondary/tertiary hypothyroidism would seem an unlikely diagnosis in most premature infants. Ares and colleagues noted the expected increase in T4, FT4, and T3 levels with increasing PMA, but additionally observed that serum FT4 and T3 values correlated with iodine intake independently of PMA (8). However, the quantitative impact of milder iodine deficiency on FT4 levels in premature infants without elevated TSH levels is not clear, and criteria for possible secondary/tertiary hypothyroidism in the premature infant have not been defined.

Thus, the mechanism(s) for the hypothyroxinemia of prematurity and the very low FT4 levels observed by Ares et al. remains unclear. To some extent it reflects altered T4 metabolism in the extrauterine environment. As much as 30% of circulating T4 in the intrauterine fetus appears to be of maternal origin, and this contribution is terminated with parturition (21). Additionally, by analogy with the sheep model, more than 85% of T4 in the intrauterine fetus is metabolized to inactive metabolites (T4 sulfate, reverse T3, reverse T3 sulfate), whereas in the neonatal period production of these metabolities is markedly reduced and T3 production increased (22).

The characterization of premature infant hypothyroxinemia as physiologic has reflected its invariability and our inability to clearly demonstrate that supplemental thyroxine therapy is necessary or beneficial, particularly with regard to brain maturation. The effect of supplemental thyroxine treatment on central nervous system development of premature infants has been assessed in two prospective studies to date. There was no effect of thyroxine treatment in 8 infants in the 1984 study of Chowdry et al. (23) or in the more recent study of van Wassenaer et al. (11) of 100 treated and 100 control infants of 25–29 weeks GA. However, a subgroup analysis of 13 treated infants vs. 18 control infants of 25–26 weeks GA indicated an 18 point higher IQ in the treated infants (11). These cohorts are small, however, and should be confirmed.

The accumulated information to date suggests that the hypothyroxinemia of premature infants of 27–28 weeks GA or older does not require treatment unless associated with an increased TSH level (>20 mU/L). The iodine metabolism data of Ares and colleagues and the thyroxine supplementation study of van Wassenaer et al. support the view that marked hypothroxinemia, in premature infants 27–28 weeks GA and younger, reflects a transient hypothalamic-pituitary hypothyroidism characterized by very low FT4 and normal or low TSH levels due to failure of the thyroid gland to increase (auto regulate) iodine uptake and lack of a TRH/TSH response to hypothyroxinemia. All infants under 27–28 weeks do not require treatment, but the threshold FT4 for this diagnosis remains unclear and will vary with the FT4 assay method. The FT4 data of Ares et al. from Madrid (8) and Rooman et al. from Antwerp (10), employing two-step FT4 immunoassay methods, would suggest a value of 0.5 ng/dL (6.4 pmol/L). The paper of Adams et al. (17) suggests a threshold value of 1.3 ng/dL (17 pmol/L) by direct dialysis, but further study will be necessary to resolve this issue. FT4 screening of premature infants under 27–28 weeks GA would seem desirable at 2 weeks of age. Frank et al. (20) have recommended screening VLBW infants for transient hypothyroidism using T4 and TSH measurements during the first week, at 2 weeks, and 4–6 weeks.

The dose of thyroxine to treat such infants has not been defined. The thyroxine production rate in the premature infant has not been assessed directly, but recent studies of van Wassenaer and colleagues (11, 24) provide insight regarding thyroxine utilization in premature infants. Figure 1Go shows FT4 concentrations in three groups of premature infants (ranging in age from 25–29 weeks GA and weighing 650-1475 g) treated with 10, 8, or 6 ug/kg thyroxine for the first 6 weeks of postnatal life (24). An untreated control group was assessed for only 3 weeks, so that the control FT4 plot in Fig. 1Go was derived from their later, larger, more prolonged study (11). Plasma TSH levels were suppressed in all three treatment groups. It is clear from Fig. 1Go that the normal T4 production rate is less than 6 µg/kg/day. The thyroxine was given intravenously for about 14 days and orally thereafter. Thus, beyond 14 days, the absorbed T4 probably approximated 70% of the given dose, so that the 6 µg/kg/day dose approximated 4 ug/kg/day (25). A reasonable estimate of the absorbed T4 dose to reproduce the control FT4 plot would be 2–4 ug/kg/day. This compares with an optimal absorbed replacement dose in term infants of 7–10 ug/kg/day. It should be recalled that the low levels of serum T4 of maternal origin in the athyroid human fetus appear adequate to prevent phenotypic hypothroidism and protect the developing brain (21). An 8 µg/kg/day supplemental thyroxine dose in infants of more than 27 weeks GA in the study of van Wassenaer and colleagues was associated with a 10 point IQ deficit relative to the placebo counterparts (11). Thus, overtreatment might carry some risk.



View larger version (19K):
[in this window]
[in a new window]
 
Figure 1. Serum FT4 concentration patterns in premature infants supplemented with exogenous Na-L-thyroxine. Each treatment cohort included 10–13 infants of 26–29 weeks GA. The untreated groups included 100 infants of 26–30 weeks GA. The dose of thyroxine is indicated for each cohort. The three FT4 patterns for treated infants were redrawn from van Wassenaer et al. (24). The untreated control cohort pattern was redrawn from the authors’ later study (van Wassenaer et al., ref. 11). Thyroxine was given intravenously during the early period of parenteral feeding (average 14 days) and orally thereafter. All three supplemental thyroxine dosages produced relative hyperthyroxinemia and serum TSH levels were suppressed in all three treatment cohorts.

 

Received March 19, 1997.

Accepted March 24, 1997.


    References
 Top
 Introduction
 References
 

  1. Fisher DA. 1990 Euthyroid low T4 and T3 states in prematures and sick infants. Pediatr Clin North Am. 37:1297–1312.[Medline]
  2. Xue-Yi C, Xin-Min J, Zhi-Hong D, et al. 1994 Timing of vulnerability of the brain to iodine deficiency in endemic cretinism. N Engl J Med. 331:1739–1944.[Abstract/Free Full Text]
  3. de Zegher F, Permasetti F, Vanhole C, Devlieger H, Van den Berghe G, Martial JA. 1995 The prenatal role of thyroid hormone evidenced by fetomaternal Pit-1 deficiency. J Clin Endocrinol Metab. 80:3127–3130.[Abstract]
  4. Meiger JJ, Verloove-Vanhorick SP, Brand R, van den Brande JL. 1992 Transient hypohyroxinemia associated with developmental delay in very preterm infants. Arch Dis Child. 67:944–947.[Abstract]
  5. Den Ouden AL, Kok JH, Verkerk PH, Brand R, Verloove-Vanhorick SP. 1996 The relation between neonatal thyroxine levels and neurodevelopmental outcome at age 5 and 9 years in a national cohort of very preterm and/or very low birth weight infants. Pediatr Res. 39:142–145.[Abstract]
  6. Reuss ML, Paneth N, Pinto-Martin JA, Lorenz JM, Susser M. 1996 The relation of transient hypothyroxinemia in preterm infants to neurologic development at two years of age. N Engl J Med. 334:821–827.[Abstract/Free Full Text]
  7. Morreale de Escobar G, Obregon MJ, Calvo R, Escobar del Rey F. 1993 Effects of iodine deficiency on thyroid hormone metabolism and the brain in fetal rats: the role of the maternal transfer of thyroxin. Am J Clin Nutr (Suppl).57:280S–295S.
  8. Ares S, Escobar-Morreale HF, Quero J, et al. 1997 Neonatal hypothyroxinemia: effects of iodine intake and of premature birth. J Clin Endocrinol Metab. 82:1704–1712.[Abstract/Free Full Text]
  9. Ares S, Quero J, Durans Presas MJ, Herruzo R, Morreale de Escobar G. 1994 Iodine content of infant formulas and iodine intake of premature babies. Arch Dis Child. 71:184–191.[Medline]
  10. Rooman RP, DuCaju MVL, Op De Beeck L, Docx M, Van Reempts P, Van Acker RT. 1996 Low thyroxinemia occurs in the majority of very preterm newborns. Eur J Pediatr. 155:211–215.[Medline]
  11. Van Wassenaer AG, Kok JH, DeVijlder JJM, et al. 1997 Effects of thyroxine supplementation on neurologic development in infants born at less than 30 weeks gestation. N Engl J Med. 336:31–26.[CrossRef]
  12. Thorpe Beeston JG, Nicolaides KH, Felton CV, Butler J, McGregor AM. 1991 Maturation of the secretion of thyroid hormone and thyroid stimulating hormone in the fetus. N Engl J Med. 324:532–536.[Abstract]
  13. Nelson JC, Weiss RM, Wilcox RB. 1994 Underestimates of serum free thyroxine (T4) concentrations by free T4 immunoassays. J Clin Endocrinal Metab. 79:76–79.[Abstract]
  14. Klein AH, Oddie TH, Parslow M, Foley Jr TP, Fisher DA. 1982 Developmental changes in pituitary-thyroid function in the human fetus and newborn. Early Hum Dev. 6:321–330.[CrossRef][Medline]
  15. Ballabio M, Nicolini V, Jowett T, Ruiz de Elvira MC, Ekins RP, Rodeck CH. 1989 Maturation of thyroid function in normal human fetuses. Clin Endocrinol (Oxf). 31:565–571.[Medline]
  16. Fisher DA, Dussault JH, Sacks J, Chopra IJ. 1977 Ontogenesis of hypothalamic pituitary thyroid function and metabolism in man, sheep, and rat. Rec Prog Horm Res. 33:59–116.
  17. Adams LM, Emery JR, Clark SJ, Carlton EL, Nelson JC. 1995 Reference ranges for newer thyroid function tests in premature infants. J Pediatr. 126:122–127.[Medline]
  18. Fuse Y, Shimizu M, Uga N, Fujii T, Irie M. 1990 Maturation of feedback control of thyrotropin in premature infants. J Dev Physiol. 14:17–22.[Medline]
  19. Delange F, Dalhem A, Bourdoux P, et al. 1984 Increased risk of primary hypothyroidism in preterm infants. J Pediatr. 105:462–469.[CrossRef][Medline]
  20. Frank JE, Faix JE, Hermos R, et al. 1996 Thyroid function in very low birth weight infants: effects on neonatal hypothyroid screening. J Pediatr. 128:548–554.[Medline]
  21. Vulsma T, Gons MH, de Vijlder JJM. 1989 Maternal-fetal transfer of thyroxine in congenital hypothyoidism due to a total organification defect or thyroid dysgenesis. N Engl J Med. 321:12–16.
  22. Fisher DA, Polk DH, Wu SY. 1994 Fetal thyroid metabolism: a pluralistic system. Thyroid. 4:367–371.[Medline]
  23. Chowdry P, Scanlon JW, Auerback R, Abbassi V. 1984 Results of controlled double-blind study of thyroid replacement in very low birth weight premature infants with hypothyroxinemia. Pediatrics. 73:301–305.[Abstract]
  24. VanWassenaer AG, Kok JH, Endert E, Vulsma T, de Vijlder JJM. 1993 Thyroxine administration to infants of less than 30 weeks gestational age does not increase plasma triiodothyronine concentrations. Acta Endocrinol (Copenh). 129:139–146.[Medline]
  25. Choe W, Hays MT. 1995 Absorption of oral thyroxine. Endocrinologist. 5:222–228.