Abstract
Purpose
Elevated thyrotropin (TSH) levels in critically ill extremely premature infants have been attributed to transient hypothyroidism of prematurity or non-thyroidal illness syndrome. We evaluated the hypothesis that relatively high TSH levels in the first two postnatal weeks follow recovery from systemic inflammation, similar to non-thyroidal illness syndrome.
Design/Methods
The study was conducted in 14 Neonatal Intensive Care Units and approved by each individual Institutional Review Board. We measured the concentrations of TSH and 25 inflammation-related proteins in blood spots obtained on postnatal days 1, 7, and 14. We then evaluated the temporal relationships between hyperthyrotropinemia (HTT), defined as a TSH concentration in the highest quartile for gestational age and postnatal day, and elevated levels of inflammation-related proteins.
Results
880 newborns less than 28 weeks of gestation were included. Elevated concentrations of inflammation-related proteins during the first or second week did not precede day-14 HTT. Systemic inflammation on day 7 was associated with day-14 HTT only if inflammation persisted through the end of the two week period. HTT frequently accompanied elevated concentrations of inflammation-related proteins on the same day.
Conclusions
The hypothesis that HTT follows recovery from severe illness, defined as preceding systemic inflammation, is weakly supported by our study. Our findings more prominently support the hypothesis that TSH conveys information about concomitant inflammation in the extremely premature newborn.
Keywords: thyroid, inflammation, prematurity, TSH, nonthyroidal illness, sick euthyroid
Introduction
Measurement of thyrotropin (TSH) is one of the preferred methods of screening for congenital hypothyroidism [1]. Nonetheless, interpretation of TSH levels in extremely preterm newborns remains a challenging task. Preterm newborns are more likely than infants born at term to screen positive for hypothyroidism. Most often, the elevated concentrations of TSH reflect transient hypothyroxinemia of prematurity [2, 3]. A positive screen result, however, may also reflect temporary factors that are prevalent in the preterm population, such as exposure to iodine or other medications [4].
Preterm newborns are also at increased risk of being critically ill and consequently having non-thyroidal illness (NTI) with an abnormal regulation of the thyroid axis [5, 6]. In NTI, neuroendocrine changes in the hypothalamic–pituitary–thyroid (HPT) axis during acute illness can be associated with inappropriately normal TSH concentrations in the setting of decreased peripheral thyroid hormone levels. Frequently TSH level rises during recovery, possibly as a compensatory mechanism [5].
The ELGAN Study (Extremely Low Gestational Age Newborns) sample provided us an opportunity to evaluate to what extent hyperthyrotropinemia (HTT) is a consequence of NTI. We selected systemic inflammation as an indicator of illness severity. In this sample, systemic inflammation as measured by the set of 25 inflammation-related proteins presented in this study, is associated with documented, culture-positive early and late bacteremia [7], prolonged mechanical ventilation [9], blood gas derangements [8], necrotizing enterocolitis and isolated intestinal perforation [10], which makes systemic inflammation an appropriate surrogate for severity of illness. We hypothesized that if HTT was a consequence of NTI, it would occur during recovery from severe illness, that is, when inflammation subsides.
Materials and Methods
The ELGAN Study
The ELGAN study was designed to identify characteristics and exposures that increase the risk of structural and functional neurologic disorders in extremely low gestational age newborns [13]. During the years 2002–2004, women who delivered before 28 weeks gestation at one of 14 participating institutions in 11 cities in 5 states were asked to enroll in the study. The enrollment and consent processes were approved by the individual institutional review boards. A full description of the methods is provided elsewhere [13]. Here we assess the relationships between TSH concentrations and characteristics of the preterm newborn, including the blood concentrations of indicators of systemic inflammation.
Newborn variables
The gestational age estimates were based on a hierarchy of the quality of available information. Most desirable were estimates based on the dates of embryo retrieval or intrauterine insemination or fetal ultrasound before the 14th week (62%). When these were not available, reliance was placed sequentially on a fetal ultrasound at 14 or more weeks (29%), last menstrual period (LMP) without fetal ultrasound (7%), and gestational age recorded in the log of the neonatal intensive care unit (1%). Of the 939 infants in the sample, 52% were male; 20% were born at 23–24 weeks gestation, 46% at 25–26 weeks and, 34% at 27 weeks; and 37% had a birth weight ≤ 750 grams, 44% were 751–1000 grams, and 19% were > 1000 grams.
Blood spot collection
After blood collected for clinical indications was prepared for laboratory analysis, remaining drops were collected on filter paper (Schleicher & Schuell 903) on the first postnatal day (range: 1–3 days), the 7th postnatal day (range: 5–8 days), and the 14th postnatal day (range: 12–15 days). Dried blood spots were stored at -70OC in sealed bags with desiccant until processed.
Protein measurement
Details about elution of the 25 inflammation-related proteins from blood spots and their measurement with the Meso Scale Discovery (MSD) electrochemiluminescence system are provided elsewhere [14]. Validated by comparisons with traditional ELISA [15, 16], this system has high content validity [14, 17–19], and inter-assay variations that are invariably less than 20%.
The Laboratory of Genital Tract Biology of the Department of Obstetrics, Gynecology and Reproductive Biology at Brigham and Women’s Hospital, Boston measured TSH along with the following 25 proteins using multiplex assays designed for the Meso Scale Discovery platform: IL-1β (Interleukin-1β), IL-6 (Interleukin-6), IL-6R (interleukin-6 receptor), TNF-α (tumor necrosis factor- α), TNF-R1 (tumor necrosis factor- α-receptor1), TNF-R2 (tumor necrosis factor- α-receptor2), IL-8 (CXCL8) (interleukin-8), MCP-1 (CCL2) (monocyte chemotactic protein-1), MCP-4 (CCL13) (monocyte chemoattractant protein-4) (CCL13), MIP-1β (CCL4) (Macrophage Inflammatory Protein-1β) (CCL4), RANTES (CCL5) (regulated upon activation, normal T-cell expressed, and [presumably] secreted), I-TAC (CXCL11) (Interferon-inducible T cell α-Chemoattractant), ICAM-1 (CD54) (intercellular adhesion molecule-1), ICAM-3 (CD50) (intercellular adhesion molecule-3), VCAM-1 (CD106) (vascular cell adhesion molecule-1), E-SEL (CD62E) (E-selectin) (CD62E), MMP-1 (matrix metalloproteinase-1), MMP-9 (matrix metalloproteinase-9), CRP (C-Reactive Protein), SAA (serum amyloid A), MPO (myeloperoxidase), VEGF (vascular endothelial growth factor), VEGF-R1 (vascular endothelial growth factor-receptor1), VEGF-R2 (vascular endothelial growth factor-receptor2), and IGFBP-1 (insulin-like growth factor binding protein-1). The MSD prototype platform for measuring TSH had been validated by the protein standard provided by Scripps Laboratories.
Measurements were made in duplicate and the mean served as the basis for all analyses. Because the volumes of blood spots collected from leftover blood at bed site vary, concentrations of each protein were normalized to mg total protein quantified by the BCA total protein assays as previously reported [17]. The measurements of TSH were expressed as nano (International) Units/mg total protein. The measurements of all other proteins were expressed as pg/mg total protein.
The concentrations of TSH and the inflammation-related proteins were not normally distributed and varied with gestational age and with the day of blood collection. Consequently, we classified newborns by the quartile of their concentrations within gestational-age (23–24, 25–26, and 27 weeks) and collection-day (postnatal day 1, 7, and 14) strata. HTT was defined as a TSH concentration in the top quartile within a gestational-age and postnatal-day stratum. Consequently, by this definition, HTT does not vary with gestational age in Table 1.
Table 1.
The percentage of newborns whose day-14 TSH concentration was or was not in the top quartile who also had the characteristic listed on the left. Panel A shows characteristics of newborns and antenatal exposures. Panel B shows postnatal exposures. These are column percents.
| A. Characteristics of newborns and antenatal exposures | Top quartile TSH* | Row N | ||
|---|---|---|---|---|
| Yes | No | |||
| Antenatal steroid course | Complete | 63 | 66 | 512 |
| Partial | 26 | 24 | 194 | |
| None | 10 | 10 | 80 | |
| Pregnancy complication | Preterm labor | 40 | 47 | 356 |
| Preterm premature ROM | 19 | 23 | 172 | |
| Preeclampsia | 18 | 11 | 102 | |
| Abruption | 9 | 11 | 86 | |
| Cervical Insufficiency | 7 | 4 | 39 | |
| Fetal Indication | 6 | 3 | 31 | |
| Gestational age (weeks) | 23–24 | 21 | 21 | 163 |
| 25–26 | 46 | 47 | 365 | |
| 27 | 33 | 33 | 258 | |
| Birth weight Z-score* | < −2 | 9 | 5 | 46 |
| ≥ −2, < −1 | 20 | 12 | 108 | |
| ≥ −1 | 71 | 83 | 632 | |
| Maximum column N | 193 | 593 | 786 | |
| B. Postnatal exposures | Top quartile of TSH | |||||
|---|---|---|---|---|---|---|
| Day 1 | Day 7 | Day 14 | ||||
| Yes | No | Yes | No | Yes | No | |
| Steroid in first postnatal week† | 10 | 8 | 6 | 9 | 7 | 8 |
| Vasopressor in first postnatal week‡ | 26 | 34 | 40 | 31 | 37 | 31 |
Top quartile TSH within gestational age category on postnatal day 14
Hydrocortisone or dexamethasone
Dubutamine, dopamine, epinephrine
Data analysis
Our first null hypothesis is that HTT is not a consequence of recovery from a presumed non-thyroidal illness (NTI). We were especially interested in assessing the relationship between systemic inflammation on days 1 and 7 and HTT on postnatal day 14. To test this hypothesis we compared infants whose concentration of an inflammation-related protein on days 1 and/or 7 was in the top quartile (on each relevant day) to infants whose inflammation-related protein concentration was in the lower three quartiles (on the same day(s)) (Table 2). We chose the 95th percentile cut-off to identify a group of newborns large enough that would allow us to make comparisons and reach conclusions with reasonable confidence. Twenty-five logistic regression models (one for each inflammation-associated protein) were fit to estimate odds ratios and 99% confidence intervals for a day-14 TSH concentration in the top quartile. We selected the 99% confidence interval, rather than the conventional 95% interval, to minimize both type 1 and type 2 errors, while adjusting for multiple comparisons. Nonetheless, the 25 proteins selected represent broadly the mediators of inflammation including cytokines, chemokines, adhesion molecules, metalloproteinases, and growth factors that reflect all different steps of the inflammatory process [20] and not necessarily contributing unique information about a distinct or separate exposure, and in these sample their top quartile concentrations tend to vary together.[21]
Table 2.
Hyperthyrotropinemia following elevated concentrations of inflammation-associated proteins in total sample.
Odds ratios (and 99% confidence intervals) for a day-14 TSH concentration in the top quartile among children who had a top quartile concentration of the inflammation-associated protein listed on the left on the day(s) listed at the top of each column. The referent group for each comparison consists of infants whose concentration of the inflammation-associated protein was in the lower three quartiles that day. The models are adjusted for gestational age. Statistically significant elevated values are in bold.
| Protein in top quartile | Days concentration of protein on left was in top quartile | ||
|---|---|---|---|
| Day 1 but not 7 | Day 7 but not 1 | Both Days 1 and 7 | |
| CRP | 0.9 (0.4, 1.8) | 1.6 (0.8, 2.9) | 1.3 (0.6, 2.6) |
| SAA | 0.8 (0.4, 1.6) | 1.3 (0.7, 2.4) | 1.9 (0.8, 4.3) |
| MPO | 0.6 (0.3, 1.3) | 0.8 (0.4, 1.7) | 0.6 (0.3, 1.4) |
| IL-1β | 0.5 (0.3, 1.2) | 1.5 (0.8, 2.8) | 1.5 (0.7, 3.0) |
| IL-6 | 0.8 (0.4, 1.6) | 1.5 (0.8, 2.7) | 1.8 (0.8, 3.9) |
| IL-6R | 1.1 (0.6, 2.2) | 0.9 (0.4, 1.7) | 1.1 (0.5, 2.3) |
| TNF-α | 0.6 (0.3, 1.3) | 1.2 (0.6, 2.4) | 1.3 (0.6, 2.6) |
| TNF-R1 | 0.9 (0.4, 1.9) | 2.1 (1.1, 4.0) | 1.7 (0.8, 3.3) |
| TNF-R2 | 0.9 (0.4, 1.9) | 2.1 (1.1, 3.7) | 1.9 (0.9, 4.0) |
| Il-8 (CXCL8) | 0.7 (0.3, 1.5) | 1.5 (0.8, 2.9) | 1.5 (0.7, 3.1) |
| MCP-1 (CCL2) | 1.1 (0.6, 2.2) | 2.2 (1.2, 4.1) | 2.6 (1.2, 5.4) |
| MCP-4 (CCL13) | 0.7 (0.3, 1.5) | 1.5 (0.8, 2.8) | 1.4 (0.7, 2.9) |
| MIP-1β (CCL4) | 0.7 (0.3, 1.5) | 2.1 (1.1, 4.0) | 1.2 (0.6, 2.3) |
| RANTES (CCL5) | 1.1 (0.5, 2.1) | 1.2 (0.6, 2.3) | 1.1 (0.5, 2.2) |
| I-TAC (CXCL11) | 0.7 (0.3, 1.5) | 1.6 (0.8, 3.2) | 1.6 (0.8, 3.3) |
| ICAM-1 (CD54) | 0.8 (0.4, 1.7) | 2.1 (1.1, 3.8) | 1.6 (0.8, 3.4) |
| ICAM-3 (CD50) | 1.0 (0.5, 2.1) | 0.8 (0.4, 1.8) | 1.0 (0.5, 1.9) |
| VCAM-1 (CD106) | 1.5 (0.7, 3.0) | 1.6 (0.8, 3.1) | 1.8 (0.9, 3.5) |
| E-SEL (CL62E) | 0.7 (0.3, 1.5) | 1.9 (1.01, 3.7) | 1.5 (0.8, 3.0) |
| MMP-1 | 1.0 (0.5, 2.0) | 1.2 (0.5, 2.5) | 0.7 (0.3, 1.4) |
| MMP-9 | 0.7 (0.4, 1.5) | 0.7 (0.4, 1.5) | 1.0 (0.5, 2.1) |
| VEGF | 0.8 (0.4, 1.6) | 1.2 (0.6, 2.3) | 0.7 (0.3, 1.4) |
| VEGF-R1 | 1.4 (0.7, 2.5) | 1.2 (0.6, 2.3) | 0.9 (0.4, 2.0) |
| VEGF-R2 | 0.8 (0.4, 1.7) | 1.5 (0.8, 2.9) | 1.6 (0.8, 3.2) |
| IGFBP-1 | 1.4 (0.7, 2.6) | 1.9 (1.02, 3.4) | 2.7 (1.2, 5.9) |
Every model included variables for gestational age category (i.e., 23–24, 25–26, and 27 weeks) and three exposure variables for the inflammation-related protein in the top quartile, one for a top quartile measurement on day 1 but not day 7, another for a top quartile measurement on day 7 but not day 1, and a third for a top quartile measurement on both days 1 and 7. These three exposure groups were compared to the referent group of children who did not have a top-quartile measurement on either day 1 or day 7. Because of the possibility that a top quartile measurement of an inflammation-associated protein on day 7 was an antecedent of systemic inflammation evident on day 14, we considered the possibility that HTT accompanies, rather than follows, systemic inflammation. To test this, we repeated the analyses in Table 2, this time excluding observations with a top-quartile, day-14 concentration of the inflammation-related protein (Table 3). This procedure minimized the confounding due to concomitant inflammation.
Table 3.
Hyperthyrotropinemia following elevated concentrations of inflammation-associated proteins in sub-samples that exclude observations associated with a top quartile day-14 concentration of the inflammation-associated protein.Odds ratios (and 99% confidence intervals) for a day-14 TSH concentration in the top quartile among children who had a top quartile concentration of the inflammation-associated protein listed on the left on the day(s) listed at the top of each column. The referent group for each comparison consists of infants whose concentration of the inflammation-associated protein was in the lower three quartiles that day. The models are adjusted for gestational age. Statistically significant elevated values are in bold.
| Protein in top quartile | Days concentration of protein on left was in top quartile | ||
|---|---|---|---|
| 1 but not 7 | 7 but not 1 | Both 1 and 7 | |
| CRP | 0.8 (0.3, 1.8) | 0.8 (0.3, 2.2) | 0.9 (0.3, 2.6) |
| SAA | 0.8 (0.4, 1.8) | 1.5 (0.7, 3.1) | 1.0 (0.3, 3.7) |
| MPO | 0.5 (0.2, 1.3) | 0.7 (0.3, 1.7) | 0.6 (0.2, 1.6) |
| IL-1β | 0.4 (0.1, 1.1) | 1.0 (0.4, 2.2) | 0.7 (0.2, 2.2) |
| IL-6 | 0.9 (0.4, 1.9) | 1.6 (0.7, 3.4) | 1.4 (0.4, 4.5) |
| IL-6R | 1.0 (0.4, 2.2) | 0.5 (0.2, 1.6) | 1.6 (0.5, 4.8) |
| TNF-α | 0.6 (0.2, 1.5) | 1.3 (0.6, 3.0) | 0.8 (0.3, 2.4) |
| TNF-R1 | 0.9 (0.3, 2.3) | 2.3 (1.03, 5.2) | 1.4 (0.5, 4.0) |
| TNF-R2 | 0.8 (0.3, 1.9) | 1.6 (0.7, 3.7) | 1.9 (0.7, 5.4) |
| Il-8 (CXCL8) | 0.9 (0.4, 2.0) | 1.5 (0.7, 3.4) | 1.3 (0.5, 3.5) |
| MCP-1 (CCL2) | 1.4 (0.7, 3.1) | 1.5 (0.6, 3.5) | 2.2 (0.7, 7.0) |
| MCP-4 (CCL13) | 0.6 (0.2, 1.5) | 1.0 (0.4, 2.7) | 1.6 (0.5, 5.2) |
| MIP-1β (CCL4) | 0.6 (0.2, 1.5) | 1.9 (0.9, 4.3) | 1.1 (0.4, 3.1) |
| RANTES (CCL5) | 1.1 (0.5, 3.2) | 1.1 (0.4, 2.7) | 0.9 (0.3, 2.6) |
| I-TAC (CXCL11) | 0.5 (0.2, 1.3) | 1.4 (0.6, 3.6) | 1.0 (0.3, 3.3) |
| ICAM-1 (CD54) | 0.8 (0.3, 1.9) | 1.4 (0.5, 3.4) | 1.4 (0.5, 4.3) |
| ICAM-3 (CD50) | 1.3 (0.5, 3.0) | 0.6 (0.2, 2.0) | 1.2 (0.5, 2.6) |
| VCAM-1 (CD106) | 1.2 (0.5, 3.0) | 1.0 (0.4, 2.7) | 0.8 (0.2, 2.7) |
| E-SEL (CL62E) | 0.7 (0.3, 1.8) | 1.6 (0.6, 3.9) | 0.9 (0.3, 2.6) |
| MMP-1 | 0.7 (0.3, 1.8) | 0.9 (0.3, 2.6) | 0.9 (0.3, 3.1) |
| MMP-9 | 0.5 (0.2, 1.3) | 0.5 (0.2, 1.3) | 1.0 (0.4, 2.3) |
| VEGF | 0.7 (0.3, 1.7) | 0.7 (0.2, 2.4) | 0.9 (0.3, 2.2) |
| VEGF-R1 | 1.5 (0.7, 3.1) | 1.1 (0.5, 2.9) | 1.6 (0.6, 4.6) |
| VEGF-R2 | 0.7 (0.3, 1.8) | 0.9 (0.3, 2.3) | 0.9 (0.3, 2.9) |
| IGFBP-1 | 1.3 (0.6, 2.9) | 2.1 (0.97, 4.4) | 2.8 (1.03, 7.8) |
Our second null hypothesis is that HTT is not associated with inflammation. We created logistic regression models that compared infants with HTT to infants without HTT to see which group was at higher risk of having concentrations of inflammation-related proteins in the top quartile for gestational age on the same day. All models were adjusted for gestational age category (i.e., 23–24, 25–26, 27 weeks) (Table 4).
Table 4.
Elevated concentrations of inflammation-associated proteins and their relation to hyperthyrotropinemia.
Odds ratios (and 99% confidence intervals) for a concentration in the top quartile of the protein listed on the left among children whose TSH concentration that same day was in the highest quartile. The referent group for all comparisons consists of infants whose TSH concentration was in the lower three quartiles that day. The models are adjusted for gestational age. Statistically significant elevated values are in bold.
| Protein in top quartile | TSH concentration in top quartile | ||
|---|---|---|---|
| Day 1 | Day 7 | Day 14 | |
| CRP | 0.6 (0.4, 1.04) | 2.1 (1.3, 3.2) | 1.5 (0.9, 2.4) |
| SAA | 0.8 (0.5, 1.4) | 1.5 (0.96, 2.4) | 1.0 (0.6, 1.6) |
| MPO | 1.1 (0.7, 1.8) | 0.9 (0.6, 1.5) | 1.2 (0.7, 1.9) |
| IL-1β | 2.0 (1.3, 3.1) | 1.7 (1.1, 2.6) | 2.2 (1.4, 3.5) |
| IL-6 | 1.9 (1.2, 3.0) | 1.9 (1.2, 3.0) | 2.1 (1.3, 3.3) |
| IL-6R | 1.8 (1.1, 2.8) | 1.3 (0.8, 2.0) | 1.7 (1.1, 2.7) |
| TNF-α | 1.7 (1.1, 2.7) | 2.0 (1.3, 3.1) | 2.3 (1.5, 3.6) |
| TNF-R1 | 2.7 (1.7, 4.2) | 1.8 (1.2, 2.8) | 3.4 (2.1, 5.3) |
| TNF-R2 | 2.5 (1.6, 3.8) | 2.4 (1.6, 3.7) | 2.4 (1.5, 3.8) |
| Il-8 (CXCL8) | 0.9 (0.5, 1.4) | 2.1 (1.4, 3.3) | 1.8 (1.1, 2.8) |
| MCP-1 (CCL2) | 1.5 (0.98, 2.4) | 2.1 (1.4, 3.3) | 2.7 (1.7, 4.3) |
| MCP-4 (CCL13) | 1.9 (1.2, 3.0) | 2.2 (1.4, 3.4) | 1.9 (1.2, 3.0) |
| MIP-1β (CCL4) | 2.3 (1.5, 3.5) | 1.6 (1.03, 2.5) | 1.3 (0.8, 2.1) |
| RANTES (CCL5) | 1.5 (0.96, 2.3) | 1.4 (0.9, 2.2) | 1.2 (0.7, 1.9) |
| I-TAC (CXCL11) | 2.2 (1.4, 3.4) | 2.2 (1.4, 3.4) | 1.9 (1.2, 3.0) |
| ICAM-1 (CD54) | 1.5 (0.9, 2.3) | 3.1 (2.0, 4.8) | 2.2 (1.4, 3.4) |
| ICAM-3 (CD50) | 1.3 (0.8, 2.0) | 1.0 (0.6, 1.6) | 1.3 (0.8, 2.1) |
| VCAM-1 (CD106) | 1.5 (0.98, 2.3) | 1.7 (1.1, 2.6) | 3.5 (2.2, 5.5) |
| E-SEL (CL62E) | 2.0 (1.3, 3.1) | 2.2 (1.4, 3.5) | 2.6 (1.7, 4.1) |
| MMP-1 | 1.4 (0.9, 2.2) | 1.3 (0.8, 2.1) | 1.3 (0.8, 2.1) |
| MMP-9 | 1.3 (0.8, 2.0) | 0.9 (0.6, 1.5) | 1.3 (0.8, 2.1) |
| VEGF | 1.5 (0.96, 2.3) | 1.3 (0.8, 2.1) | 1.4 (0.9, 2.2) |
| VEGF-R1 | 2.8 (1.8, 4.4) | 1.7 (1.1, 2.6) | 1.9 (1.2, 3.0) |
| VEGF-R2 | 1.2 (0.8, 2.0) | 2.7 (1.7, 4.2) | 2.0 (1.3, 3.2) |
| IGFBP-1 | 2.8 (1.8, 4.3) | 2.2 (1.4, 3.4) | 2.0 (1.3, 3.2) |
Results
Characteristics and exposures of newborns with HTT (Table 1)
Newborns who had HTT on day 14 did not differ from newborns who did not have HTT in their birth weight Z-scores, pregnancy complications (preterm labor, premature rupture of membranes, placental abruption, cervical insufficiency), or exposure to antenatal steroids. However, newborns in the HTT group had a higher risk of being delivered because of a pregnancy complicated by preeclampsia or a fetal indication.
Elevated concentrations of inflammation-associated proteins preceding day-14 HTT: total sample (Table 2)
Children who had only a day-1 concentration of the inflammation-associated protein in the top quartile were not at increased risk of day-14 HTT. On the other hand, children who had only a top quartile day-7 concentration of TNF-R1, TNF-R2, MCP-1, MIP-1β, ICAM-1, E-SEL, or IGFBP-1 were at statistically significant risk of day 14-HTT. MCP-1, and IGFBP-1 were the only two proteins whose elevated concentrations on both day-1 and day-7 was associated with an increased risk of day-14 HTT.
Elevated concentrations of inflammation-associated proteins preceding day-14 HTT: excluding cases with elevated inflammation-related proteins on day-14 (Table 3)
While 7 proteins had top quartile concentrations on day 7 that predicted day-14 HTT (Table 2), only 1 of these proteins, TNF-R1, did so in the sub-set that excluded cases with a top quartile day-14 concentration of the same protein.
Elevated concentrations of inflammation-associated proteins and their relation to concurrent HTT (Table 4)
Newborns with HTT on each of the three days assessed were more likely than others to have elevated concentrations of inflammation-related proteins on the same day. On day 1, elevated concentrations of 12 proteins were significantly associated (p < 0.01) with HTT on the same day. Newborns with HTT on day 7 were more likely than others to have an elevated concentration of 16 proteins on day 7, while those who had HTT on day 14 were more likely than others to have an elevated concentration of 15 proteins that day.
Discussion
In our sample of extremely preterm newborns, HTT tended to be accompanied by systemic inflammation rather than follow recovery from systemic inflammation.
Systemic inflammation as a surrogate of severe illness and NTI
The 25 inflammation-related proteins analyzed in our study are a broad representation of cytokines, chemokines, adhesion molecules, metalloproteinases, and growth factors. These biomarkers represent various steps or aspects of the inflammatory process, and their peak concentrations may vary within the biological process of inflammation. Thus, all together they give a robust picture of the inflammatory process. [20] In the ELGAN sample, these 25 inflammation-related proteins are associated with documented, culture-positive early and late bacteremia [7], prolonged mechanical ventilation [9], blood gas derangements [8], necrotizing enterocolitis and isolated intestinal perforation [10], which are critical illnesses known to affect thyroid regulation in preterm infants.[22]
HTT is probably not a consequence of NTI
ELGAN infants with intense systemic inflammation are at heightened risk of NTI. We hypothesized that consequently to NTI, ELGANs would demonstrate higher levels of TSH after the inflammation subsided (i.e. during the “recovery phase”). We did find elevated TSH consequential to increase of inflammatory proteins but only for day-7 systemic inflammation, and only if it did not subside by the time the TSH was elevated. This indicates that concomitant inflammation on day 14 probably accounted for some or most of the day 7 inflammation that preceded day 14 HTT. To the extent that systemic inflammation can be viewed as a surrogate for disorders and processes that can result in NTI, the main implication of our findings is that day-14 HTT is unlikely to be a consequence of NTI but rather may be a lagging indicator of systemic inflammation.
HTT is concomitant with inflammation
To our knowledge, the only study of TSH-inflammation relationships in premature infants found that higher concentrations of IL-6 and CRP were not appreciably associated with TSH concentration, although they were associated with lower concentrations of total T3 [23]. Observational studies of adults document that those who have elevated TSH concentrations are more likely than others to have systemic inflammation [24, 25]. These studies, however, do not identify which phenomenon precedes the other, the thyroid dysfunction or the inflammation.
Our finding of concomitant inflammation and HTT has several plausible explanations. One is that TSH can potentially be a marker of an acute phase response independent of its role in thyroid axis regulation. In essence, our findings are compatible with the hypothesis that both HTT and systemic inflammation have a common set of antecedents.
Another explanation is that elevated TSH concentrations stimulate systemic inflammation. The inflammation-related proteins found to be concomitantly elevated with TSH are similar to the ones described in adult populations and in tissue-specific cells following TSH administration [26, 27]. For example, TSH promotes the release of TNF-alpha in vitro by one type of bone marrow cell [26], and release of IL-6 by human abdominal subcutaneous differentiated adipocytes [28, 29]. In the same manner, administration of recombinant human TSH to human adults increases blood concentrations of CRP and ICAM-1 [27]. In addition, successful treatment of hypothyroidism in dogs reduces circulating levels of SAA [30].
A third possible explanation is that inflammation promotes TSH synthesis and release. This is plausible in light of studies in mice that lipopolysaccharide, a strong inflammatory stimulus, decreases serum thyroxine and triiodothyronine concentrations [31]. The discrepancy between the findings of Tables 2 and 3 suggest that systemic inflammation might sometimes contribute to HTT.
Limitations
This study has several limitations. First, we were not able to measure thyroxine concentration at the time of TSH measurements. Additionally, due to the measurement technique involving eluents of dried blood spots, we did not have a volume for each specimen, and were, therefore, not able to calculate actual concentration in mIU/L. However, because they are standardized to protein concentrations in the blood spots, our measurements allow comparisons within our ELGAN cohort. In addition, it is evident from our results that some of the markers of inflammation where related to concomitant HTT while others were not. The selectivity of these findings could be explained by the kinetics of the systemic inflammation and the variability in the timing of each biomarker’s concentration peak; however, we cannot discriminate in this study whether these discrepancies are due to false-positive or false-negative results or represent the complexity of systemic inflammation. Last, like any other observational study, ours does not distinguish between causation and association.
Strengths
Our study has several strengths. We included a large number of infants, making it unlikely that we have missed important associations due to lack of statistical power, or claimed associations that might reflect the instability of small numbers. We selected infants based on gestational age, not birth weight, in order to minimize confounding due to factors related to fetal growth restriction [32]. All data were collected prospectively and our protein measurements are of high quality [15] and have high content validity [14, 17–19].
The future
Because our findings raise the possibility that HTT in preterm infants reflects ongoing inflammation, future studies should better define the nature of the relationship between systemic inflammation and elevated concentrations of TSH in very preterm newborns. Exploration is also encouraged to assess to what extent HTT constitutes an adaptive or maladaptive response.
Conclusion
Elevated TSH concentrations in extremely low gestational age newborns (ELGANs) are associated with concomitant systemic inflammation. The hypothesis that HTT follows recovery from severe illness in this cohort is minimally supported by our data. To what extent TSH elevation acts as an activator of an inflammatory response remains to be determined.
Acknowledgments
Research reported in this publication was supported by the National Institute of Neurological Disorders and Stroke (5U01NS040069-05; 2R01NS040069-06A2), the National Institute of Child Health and Human Development (5P30HD018655-28), and the National Institute of Diabetes and Digestive and Kidney Diseases (T32DK007699-31S1). The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The funding institutions did not contribute to the collection, management, analysis, and interpretation of the data; nor preparation, review, or approval of the manuscript.
Footnotes
The authors declare that they have no conflict of interest or disclosures relevant to the subject matter or materials included in this work.
References
- 1.Rose SR, Brown RS, Foley T, Kaplowitz PB, Kaye CI, Sundararajan S, Varma SK. Update of newborn screening and therapy for congenital hypothyroidism. Pediatrics. 2006;117:2290–2303. doi: 10.1542/peds.2006-0915. [DOI] [PubMed] [Google Scholar]
- 2.Bhavani N. Transient congenital hypothyroidism. Indian J Endocrinol Metab. 2011;15:S117–120. doi: 10.4103/2230-8210.83345. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Fisher DA. Thyroid system immaturities in very low birth weight premature infants. Semin Perinatol. 2008;32:387–397. doi: 10.1053/j.semperi.2008.09.003. [DOI] [PubMed] [Google Scholar]
- 4.Aitken J, Williams FL. A systematic review of thyroid dysfunction in preterm neonates exposed to topical iodine. Arch Dis Child Fetal Neonatal Ed. 2013 doi: 10.1136/archdischild-2013-303799. [DOI] [PubMed] [Google Scholar]
- 5.Marks SD. Nonthyroidal illness syndrome in children. Endocrine. 2009;36:355–367. doi: 10.1007/s12020-009-9239-2. [DOI] [PubMed] [Google Scholar]
- 6.Srinivasan R, Harigopal S, Turner S, Cheetham T. Permanent and transient congenital hypothyroidism in preterm infants. Acta Paediatr. 2012;101:e179–182. doi: 10.1111/j.1651-2227.2011.02536.x. [DOI] [PubMed] [Google Scholar]
- 7.Leviton A, O’Shea TM, Bednarek FJ, Allred EN, Fichorova RN, Dammann O. Systemic responses of preterm newborns with presumed or documented bacteraemia. Acta Paediatr. 2012;101:355–359. doi: 10.1111/j.1651-2227.2011.02527.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Leviton A, Allred EN, Kuban KC, Dammann O, Fichorova RN, O’Shea TM, Paneth N. Blood protein concentrations in the first two postnatal weeks associated with early postnatal blood gas derangements among infants born before the 28th week of gestation. The ELGAN Study. Cytokine. 2011;56:392–398. doi: 10.1016/j.cyto.2011.07.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Bose CL, Laughon MM, Allred EN, O’Shea TM, Van Marter LJ, Ehrenkranz RA, Fichorova RN, Leviton A. Systemic inflammation associated with mechanical ventilation among extremely preterm infants. Cytokine. 2013;61:315–322. doi: 10.1016/j.cyto.2012.10.014. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Leviton A, Kuban K, O’Shea TM, Paneth N, Fichorova R, Allred EN, Dammann O. The relationship between early concentrations of 25 blood proteins and cerebral white matter injury in preterm newborns: the ELGAN study. J Pediatr. 2011;158:897–903. e891–895. doi: 10.1016/j.jpeds.2010.11.059. [DOI] [PubMed] [Google Scholar]
- 13.O’Shea TM, Allred EN, Dammann O, Hirtz D, Kuban KC, Paneth N, Leviton A. The ELGAN study of the brain and related disorders in extremely low gestational age newborns. Early Hum Dev. 2009;85:719–725. doi: 10.1016/j.earlhumdev.2009.08.060. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Hecht JL, Fichorova RN, Tang VF, Allred EN, McElrath TF, Leviton A. Relationship Between Neonatal Blood Protein Concentrations and Placenta Histologic Characteristics in Extremely Low GA Newborns. Pediatr Res. 2011;69:68–73. doi: 10.1203/PDR.0b013e3181fed334. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Fichorova RN, Richardson-Harman N, Alfano M, Belec L, Carbonneil C, Chen S, Cosentino L, Curtis K, Dezzutti CS, Donoval B, et al. Biological and technical variables affecting immunoassay recovery of cytokines from human serum and simulated vaginal fluid: a multicenter study. Anal Chem. 2008;80:4741–4751. doi: 10.1021/ac702628q. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Fichorova RN, Trifonova RT, Gilbert RO, Costello CE, Hayes GR, Lucas JJ, Singh BN. Trichomonas vaginalis lipophosphoglycan triggers a selective upregulation of cytokines by human female reproductive tract epithelial cells. Infect Immun. 2006;74:5773–5779. doi: 10.1128/IAI.00631-06. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Fichorova RN, Onderdonk AB, Yamamoto H, Delaney ML, DuBois AM, Allred E, Leviton A. Maternal microbe-specific modulation of inflammatory response in extremely low-gestational-age newborns. MBio. 2011;2:e00280–10. doi: 10.1128/mBio.00280-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Leviton A, Fichorova R, Yamamoto Y, Allred EN, Dammann O, Hecht J, Kuban K, McElrath T, O’Shea TM, Paneth N. Inflammation-related proteins in the blood of extremely low gestational age newborns. The contribution of inflammation to the appearance of developmental regulation. Cytokine. 2011;53:66–73. doi: 10.1016/j.cyto.2010.09.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.McElrath TF, Fichorova RN, Allred EN, Hecht JL, Ismail MA, Yuan H, Leviton A. Blood protein profiles of infants born before 28 weeks differ by pregnancy complication. Am J Obstet Gynecol. 2011;204:418 e411–418 e412. doi: 10.1016/j.ajog.2010.12.010. [DOI] [PubMed] [Google Scholar]
- 20.Namas R, Zamora R, Namas R, An G, Doyle J, Dick TE, Jacono FJ, Androulakis IP, Nieman GF, Chang S, Billiar TR, Kellum JA, Angus DC, Vodovotz Y. Sepsis: Something old, something new, and a systems view. J Crit Care. 2012;27(3):314.e1–11. doi: 10.1016/j.jcrc.2011.05.025. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Leviton A, Allred EN, Yamamoto H, Fichorova RN ELGAN Study Investigators. Relationships among the concentrations of 25 inflammation-associated proteins during the first postnatal weeks in the blood of infants born before the 28th week of gestation. Cytokine. 2012;57(1):182–90. doi: 10.1016/j.cyto.2011.11.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Williams FL, Ogston SA, van Toor H, Visser TJ, Hume R. Serum thyroid hormones in preterm infants: associations with postnatal illnesses and drug usage. J Clin Endocrinol Metab. 2005 Nov;90(11):5954–63. doi: 10.1210/jc.2005-1049. [DOI] [PubMed] [Google Scholar]
- 23.Dilli D, Dilmen U. The role of interleukin-6 and C-reactive protein in non-thyroidal illness in premature infants followed in neonatal intensive care unit. J Clin Res Pediatr Endocrinol. 2012;4:66–71. doi: 10.4274/jcrpe.625. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.La Vignera S, Condorelli R, Vicari E, Calogero AE. Endothelial dysfunction and subclinical hypothyroidism: a brief review. J Endocrinol Invest. 2012;35:96–103. doi: 10.3275/8190. [DOI] [PubMed] [Google Scholar]
- 25.Yu YT, Ho CT, Hsu HS, Li CI, Davidson LE, Liu CS, Li TC, Shih CM, Lin CC, Lin WY. Subclinical hypothyroidism is associated with elevated high-sensitive C-reactive protein among adult Taiwanese. Endocrine. 2013 Dec;44(3):716–22. doi: 10.1007/s12020-013-9915-0. Epub 2013 Mar 7. [DOI] [PubMed] [Google Scholar]
- 26.Wang HC, Dragoo J, Zhou Q, Klein JR. An intrinsic thyrotropin-mediated pathway of TNF-alpha production by bone marrow cells. Blood. 2003;101:119–123. doi: 10.1182/blood-2002-02-0544. [DOI] [PubMed] [Google Scholar]
- 27.Desideri G, Bocale R, Milardi D, Ghiadoni L, Grassi D, Necozione S, Taddei S, di Orio F, Pontecorvi A, Ferri C. Enhanced proatherogenic inflammation after recombinant human TSH administration in patients monitored for thyroid cancer remnant. Clin Endocrinol (Oxf) 2009;71:429–433. doi: 10.1111/j.1365-2265.2008.03485.x. [DOI] [PubMed] [Google Scholar]
- 28.Antunes TT, Gagnon A, Bell A, Sorisky A. Thyroid-stimulating hormone stimulates interleukin-6 release from 3T3-L1 adipocytes through a cAMP-protein kinase A pathway. Obes Res. 2005;13:2066–2071. doi: 10.1038/oby.2005.256. [DOI] [PubMed] [Google Scholar]
- 29.Antunes TT, Gagnon A, Langille ML, Sorisky A. Thyroid-stimulating hormone induces interleukin-6 release from human adipocytes through activation of the nuclear factor-kappaB pathway. Endocrinology. 2008;149:3062–3066. doi: 10.1210/en.2007-1588. [DOI] [PubMed] [Google Scholar]
- 30.Tvarijonaviciute A, Jaillardon L, Ceron JJ, Siliart B. Effects of thyroxin therapy on different analytes related to obesity and inflammation in dogs with hypothyroidism. Vet J. 2013;196:71–75. doi: 10.1016/j.tvjl.2012.08.005. [DOI] [PubMed] [Google Scholar]
- 31.van Zeijl CJ, Surovtseva OV, Wiersinga WM, Fliers E, Boelen A. Acute inflammation increases pituitary and hypothalamic glycoprotein hormone subunit B5 mRNA expression in association with decreased thyrotrophin receptor mRNA expression in mice. J Neuroendocrinol. 2011;23:310–319. doi: 10.1111/j.1365-2826.2011.02116.x. [DOI] [PubMed] [Google Scholar]
- 32.Arnold CC, Kramer MS, Hobbs CA, McLean FH, Usher RH. Very low birth weight: a problematic cohort for epidemiologic studies of very small or immature neonates. Am J Epidemiol. 1991;134:604–613. doi: 10.1093/oxfordjournals.aje.a116133. [DOI] [PubMed] [Google Scholar]