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 |
---|
![]() ![]() ![]() |
---|
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 23 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 months JCEM (see page 1704), report a detailed study of iodine metabolism and iodine balance in 115 premature infants of 2736 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 23 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 3040 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 30100 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 217 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 714 days. Serum TSH levels did not increase in these infants, and serum FT4 values returned to near cord values by 34 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 2740 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 0.91.4 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 2830 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 1.01.5 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 (15002499g), 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 2529 weeks GA. However, a subgroup analysis of 13 treated infants vs. 18 control infants of 2526 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 2728 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 2728 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 2728 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 2728 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 46 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 1 shows FT4 concentrations in three
groups of premature infants (ranging in age from 2529 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. 1
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. 1
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 24 ug/kg/day. This compares with an
optimal absorbed replacement dose in term infants of 710 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.
|
Received March 19, 1997.
Accepted March 24, 1997.
![]() |
References |
---|
![]() ![]() ![]() |
---|