Published online February 29, 2008
PEDIATRICS
Vol. 121
No. 3
March 2008, pp.
e574-e580
(doi:10.1542/peds.2007-1327)
Luteinizing Hormone and Follicle-Stimulating Hormone Levels in Extreme Prematurity: Development of Reference Intervals
Ronda F. Greaves, BSc, MAACB, MappSca,
Rodney W. Hunt, FRACP, MMedb,
Angela S. Chiriano, BScc and
Margaret R. Zacharin, MBBS, FRACPd
a Departments of Complex Biochemistry
b Neonatal Medicine
c Core Laboratory
d Endocrinology and Diabetes, Royal Children's Hospital, Victoria, Australia
 |
ABSTRACT
|
|---|
OBJECTIVES. Establishing pediatric reference intervals has always been challenging, with most ranges used in pediatric laboratories developed over many years. The clinical interpretation of gonadotropins is important in the context of ambiguous genitalia. The aim of this study was to develop reference intervals for luteinizing hormone and follicle-stimulating hormone in infants born between 24 and 29 weeks’ gestation.
METHODS. Samples were collected at 0 to 43 days after birth from 82 premature infants born <30 weeks’ gestation for analysis of luteinizing hormone and follicle-stimulating hormone by automated immunochemiluminometric immunoassays.
RESULTS. The 43 male infants demonstrated a range of luteinizing hormone levels from 0.1 to 13.4 IU/L and of follicle-stimulating hormone levels from 0.3 to 4.6 IU/L. The 39 female infants demonstrated a range of luteinizing hormone levels from 0.2 to 54.4 IU/L and of follicle-stimulating hormone levels from 1.2 to 167.0 IU/L. The ratio of luteinizing hormone/follicle-stimulating hormone levels differed with males, ranging from 0.3 to 9.4, and females, at <0.5.
CONCLUSION. These data provide guidance for the interpretation of luteinizing hormone and follicle-stimulating hormone levels for the first 6 weeks of life in extremely premature infants born between 24 and 29 weeks’ gestation. The availability of age-appropriate reference intervals is essential for correct and timely interpretation of biochemical results to the clinician.
Key Words: extremely premature infants • follicle-stimulating hormone • luteinizing hormone • reference range
Abbreviations: LH—luteinizing hormone • FSH—follicle-stimulating hormone • hCG—human chorionic gonadotropin • CV—coefficient of variation
Establishment of pediatric reference intervals has always been a challenging and difficult process, with most ranges used in pediatric biochemistry laboratories developed over decades. There are known pitfalls with the reference intervals of a number of analytes, in particular, hormones,1 which can be attributed in part to assumptions routinely made in the assignment of ranges. Reference intervals are commonly assigned by laboratory information systems based on age, and correction is not often made for life events such as prematurity and puberty.
Improved survival rates associated with preterm deliveries of infants weighing <1.5 kg have increased to 85%.2 Successes in treatment and management of previously life-threatening diseases, such as hyaline membrane disease, have brought with them the opportunity to focus attention on other organ systems, including the endocrine system. New challenges are now evident with regard to the interpretation of findings in relation to the endocrine system of the extremely premature neonate. What defines "normal" from an underlying pathologic process is not always clear in the very premature population. This is highlighted with the question of ambiguous genitalia, where hormone levels are requested to aid in clinical decision-making.3
Along with steroid hormones, gonadotropins are often requested for the assessment of ambiguous genitalia. Because we had inadequate data available to interpret luteinizing hormone (LH) and follicle-stimulating hormone (FSH) levels in very premature infants (born between 24 and 29 weeks’ gestation), we sourced published data for interpretation. The early literature clearly described the ontogeny of LH and FSH in fetal pituitary glands from 68 days’ gestation to term.4 Serial measurement of serum LH and FSH have been reported in infants born as early as 28 weeks’ gestation,5 and higher levels of LH and FSH have been demonstrated in cord blood samples collected from infants born between 26 and 40 weeks’ gestation.6 A premature infant hormone replacement therapy study measured LH and FSH levels after supplementation with progesterone or estrogen in infants born between 24 and 29 weeks’ gestation.7
The data from the study by Massa et al6 have provided us with a guide for interpretation, because they demonstrated higher LH and FSH levels in cord blood collected from infants as young as 26 weeks’ gestation compared with term neonates. However, these levels did not reflect the even higher levels we were seeing in our 
24 weeks’ gestation infants, raising the question of whether this was associated with an inherent methodologic difference or a reflection of differing gonadotropin levels in even younger premature neonates. Overall, the interpretation of the published studies was limited for a number of reasons: (1) the samples were collected from cord blood and, therefore, could reflect maternal gonadotropins levels6; (2) the samples were from older premature infants and may not reflect levels in very premature neonates5; (3) the blood was collected after hormone replacement therapy and may, therefore, not reflect natural hormone levels7; or (4) the methods used for analysis were subject to significant interference from human chorionic gonadotropin (hCG)5 or were analyzed by now nonroutine methods (radioimmunoassay5 and solid-phase enzyme-amplified sensitivity immunoassays6). Because of these limitations, LH and FSH reference data could not readily be transposed to routine automated assays for our extremely premature male and female infants born as early as 24 weeks’ gestation.
The aim of this pilot study was to develop reference intervals for LH and FSH, performed on a common autoanalyzer, for extremely premature infants born between 24 and 29 weeks’ gestation.
|
METHODS
|
|---|
Subjects
Samples (n = 82) were collected from premature infants with no suspicion of ambiguous genitalia who were born at <30 weeks’ gestation. This study was approved by the ethics in human research committee at Royal Women's Hospital, and informed consent was obtained from the parents before sampling.
Sample Collection
Samples were collected from cord blood and from infants 
6 weeks (43 days) of age. Sampling coincided with clinical sampling and referred to analysis of LH and FSH after routine sampling was completed. Specimens for the study were spun and the serum stored at –70°C until analysis. A total of 82 samples (43 male and 39 female) were analyzed between 0 and 43 days of subject age.
Serum Assays
The samples were assayed for LH and FSH by immunochemiluminometric immunoassays on the Centaur (Bayer Corporation, Elkhart, IN). The between-run imprecision, reported as coefficient of variation (CV), for LH at 4.9 IU/L is CV 2.9% (n = 32) and at 37.4 IU/L is CV 3.7% (n = 33). The between-run imprecision for FSH at 4.8 IU/L is CV 7.3% (n = 29) and at 60.4 IU/L is CV 3.5% (n = 31). The minimum volume of serum required for analysis of both gonadotropins is 225 µL (FSH: 100 µL; LH: 50 µL; dead volume: 75 µL). The functional sensitivity for both LH and FSH is <0.1IU/L. There is no significant cross-reactivity with hCG in the LH and FSH assays.
Analysis of Data
Gonadotropin results were presented as days after birth and relative or corrected gestational age (being the gestational age at birth plus the age of the infant at the time of sample collection). Graphs of LH and FSH concentrations related to age were generated with Microsoft Excel (Microsoft, Redmond, WA). Box-and-whisker plot and statistical analysis of the ratio of LH to FSH were performed with SPSS for Windows 11 (SPSS Inc, Chicago, IL). The Tietz textbook of clinical chemistry was enlisted for criteria to develop reference intervals.8
|
RESULTS
|
|---|
All of the infants were physically normal in appearance, with no suspicion of ambiguous genitalia or other endocrinopathy. Of the 82 (43 males and 39 females) samples included in this study, we obtained 66 LH and 51 FSH results across the corrected gestational age range of 25 to 35 weeks, representing the first 6 weeks of life.
Male Samples
The results obtained from the 43 male samples demonstrated a range of LH results from 0.1 to 13.4 IU/L (median: 1.7 IU/L; n = 39) and a range of FSH results from 0.3 to 4.6 IU/L (median: 1.2 IU/L; n = 29; Table 1). Of the 10 LH results >4 IU/L, 5 were collected at 0 days of age (cord blood samples), 3 were collected at 1 day of age, 1 from 7 days of age, and 1 from a 14-day-old infant. An equivalent rise is not evident for FSH with the ratio for these 10 samples all >2.0 (Fig 1). Therefore, because of the higher levels evident in cord blood samples, the results from cord blood and 1-day-old infants were separated and compared with the results from samples of older infants (4–43 days of age; Table 1).
View larger version (10K):
[in this window]
[in a new window]
|
FIGURE 1 LH and FSH results for 43 samples collected from male infants born between 25 and 29 weeks’ gestation. Samples were collected from 0 to 43 days of age, presented as age after birth (A) and corrected gestational age (excluding cord blood and samples collected at 1 day of age) (B).
|
|
Despite the obvious difference in LH and FSH concentrations between the 2 sets of results, the ratios of LH/FSH were consistent between the cord day-1 and day-4 to -43 groups (Table 1). The ratio of LH/FSH was calculated (n = 25) and showed a median ratio of 2.3 with the ratio ranging from 0.3 to 3.9 with 2 high outliers demonstrated on the box-and-whisker plot (Fig 3).
View larger version (6K):
[in this window]
[in a new window]
|
FIGURE 3 LH/FSH ratio of serum collected in the first 43 days of life from infants born between 24 and 29 weeks’ gestation. Male (n = 25) corrected gestational age at the time of sampling ranged from 27 to 34 weeks’ gestation, and female (n = 9) corrected gestational age at the time of sampling ranged from 25 to 32 weeks’ gestation. The 2 outliers evident in the male plot are from 1 cord blood sample and one 7-day-old infant. A 2-tailed t test demonstrated a significant difference between the LH/FSH ratio between male and female extremely premature infants (P < .01). a Outlier from 7-day-old infant; b Outlier from cord blood sample.
|
|
Female Samples
The results obtained from the 39 female samples demonstrated a range of LH results from 0.2 to 54.4 IU/L (median: 4.7 IU/L; n = 27) and a range of FSH results from 1.2 to 167.0 IU/L (median: 15.7 IU/L; n = 21; Table 1). Unlike the male LH and FSH levels, there was no relative increase for the female cord blood samples. LH and FSH levels do not demonstrate an obvious peak at a specific gestational age. Eight FSH results were >40 IU/L; 2 samples were collected at 28 days of age and 1 each at 5, 7, 14, 21, 29, and 41 days of age. None of the high FSH levels (>40 IU/L) were obtained from cord blood samples (Fig 2). The results are also reported as "corrected gestational age" at the time of sampling (Fig 2). Applying corrected gestational age did not reveal additional information.
View larger version (10K):
[in this window]
[in a new window]
|
FIGURE 2 LH and FSH results for 39 samples collected from female infants born between 24 and 29 weeks’ gestation. Samples were collected from 0 to 43 days of age, presented as age after birth (A) and corrected gestational age (B).
|
|
The LH/FSH ratio was calculated (n = 9) and showed a median of 0.3 with the ratio ranging from <0.1 to 0.5, with no apparent outliers (Fig 3). There was insufficient data to calculate the ratio of LH/FSH for the high FSH samples (ie, the samples >40 IU/L).
|
DISCUSSION
|
|---|
We present here pilot reference intervals for the first 6 weeks of life in extremely premature infants born between 24 and 29 weeks’ gestation. Serum LH and FSH levels have been shown previously to be higher in the cord blood of premature infants6 compared with cord blood and peripheral blood from healthy term neonates.6,9 However, this is the first such report for extremely premature infants that uses routine automated assays that are not subject to cross-reactivity from the 
subunit of the gonadotropin peptides and looks at levels for the first 6 weeks of life.
The closest previous study reported LH and FSH levels for infants as young as 28 weeks’ gestation but was subject to significant cross-reactivity from hCG in the LH assay.5 This study demonstrated higher female levels of FSH (up to 
300 IU/L compared with our upper levels of 167 IU/L) with a peak appearing at 2 to 4 weeks of age and comparatively very high male LH levels (up to 250 IU/L compared with our highest result of 13.4 IU/L) with a peak appearing at 2 to 4 weeks after birth. All of the results from this early report are higher than those obtained by us, although our studies overlap with results in both studies for infants born at 28 and 29 weeks’ gestation. The difference between this early study and our study may be attributed to a combination of factors including a difference in the international reference preparation used for standardization of the gonadotropin assays; lack of specificity of the antibody used resulting in higher than acceptable cross-reactivity between subunits from other related peptides; the older gestational age group selected for the study; or the imprecision of the assays.5
The study by Massa et al6 demonstrated levels of LH and FSH in cord blood samples from infants as young as 26 weeks’ gestation. The male FSH results are consistent with the results presented in our report, but the LH results are inconsistent with our findings with cord blood LH results up to 100 IU/L compared with an upper result of 13.4 IU/L reported here. The female LH and FSH cord blood results are consistent between the report by Massa et al6 and those reported here. The difference in male gonadotropin results between the studies is probably because of a calibration difference but alternatively may be because of a change in the management of preterm infants over time, resulting in more clinical stability in our recent study, decreasing the effect of stress on the production of hormones. The major limitation of the cord blood gonadotropin study is that it cannot be extrapolated for interpretation of LH and FSH results in blood samples collected in the days and weeks after birth.
The underlining mechanism of why the LH and FSH levels are higher in premature neonates compared with term neonates is not well defined. Possible reasons for these levels include the following: (1) immaturity of the hypothalamic-pituitary gonadal axis; (2) the persistence of fetal adrenal steroids10 causing an alteration to the axis; or (3) lack of negative feedback on the axis. Given that the premature levels of LH and FSH are significantly different between males and females, the mechanism of action may be gender dependant.
The results from our extremely premature neonates demonstrate a significant difference (P < .01) in the ratio of LH/FSH between the genders, with male infants having a higher ratio of LH/FSH (median: 2.3) compared with female infants (median: 0.3). This difference in the ratio of LH/FSH between male and female infants is supported by other studies of premature infants5,6 and also by studies of term neonates.9,11 Where questions of genital ambiguity exist, the calculation of the ratio of LH/FSH may provide additional useful information for all neonates regardless of gestational age at birth.
The definition of "normal" in the premature population has not been well defined. From the laboratory's perspective, a lower gestational age at delivery can have implications on the interpretation of the results of hormone assays. This places considerable stress on the clinician, parents, and the laboratory in endeavoring to interpret data in this population. To address the problem of "normal" in very premature infants, it is essential that laboratory data are interpreted in relation to the relative gestational age of the infant in the first 2 to 3 months of life to determine whether the infant has reached the equivalent age of a term neonate.3 In this study, at 6 weeks of age our infants were still extremely premature and had not reached the equivalent of term, and there was no discernible advantage in reporting age in weeks compared with relative gestational age (Figs 1 and 2). However, research into levels of other more closely studied hormones (eg, thyroxine) clearly show that consideration of both gestation and postnatal age are important in the interpretation of hormone levels in infants born at <30 weeks’ gestation.12
Concurrent with improvement in survival of the very premature neonate is the increased availability of endocrine tests on automated platforms. Accuracy and reliability of laboratory tests impact on the clinical decision cascade. There is constant pressure on laboratories to use automated assays on large-scale platforms. Automated assays are generally more cost-effective, require less expertise, and have significantly more efficient turnaround times than less automated assays. However, the choice of instrumentation for analysis should be considered for sample volume requirements in the premature population. Analysis of LH and FSH levels on the Centaur immunoassay analyser requires an absolute minimum serum volume of 225 µL (or 0.5 mL of whole blood). This is a trivial volume to collect for older patients, but in the low birth weight infants described here, multiple venipunctures can result in significant iatrogenic anemia.
Recent published data suggest that, with the current standardization of LH and FSH levels, the reference interval data presented here can be easily transposed to other automated platforms.13
|
CONCLUSIONS
|
|---|
Apparent "abnormal" hormone levels in premature infants may not be indicative of an underlying pathologic process but rather the inappropriate assignment of a reference interval. Interim reference intervals are now available for interpretation of LH and FSH levels for the first 6 weeks of life in extremely premature infants. Because the range of LH and FSH level results is broad, especially in the female premature infants, we recommend that the ratio of LH/FSH should be calculated and included for the interpretation of gonadotropin results. The selection of hormone assays that use a small sample volume in this population and availability of age-appropriate reference intervals are essential for correct and timely interpretation of biochemical results to the clinician.
Additional work with a larger number of subjects, incorporating weekly serial time point measurements for 1 to 2 months after birth, to give longitudinal data would be of value to strengthen the reference interval data presented here.
|
ACKNOWLEDGMENTS
|
|---|
This work is supported by National Health and Medical Research Council grant 216757. We thank the biochemistry staff at the Royal Womens’ Hospital for performing the analysis of LH and FSH and the Royal Childrens’ Hospital Clinical Epidemiology and Biostatistics Unit for advice on the level of statistical analysis suitable for this data set.
|
FOOTNOTES
|
|---|
Accepted Aug 13, 2007.
Address correspondence to Ronda F. Greaves, BSc, MAACB, MappSc, Royal Children's Hospital, Complex Biochemistry Department, Parkville, Victoria 3052, Australia. E-mail: ronda.greaves{at}rch.org.au
The authors have indicated they have no financial relationships relevant to this article to disclose.
|
REFERENCES
|
|---|
- Verras P, Greaves R. Abnormal laboratory results: important issues to consider in the interpretation of paediatric biochemistry.
Aust Prescriber.2005; 28
(5):126
–129
- National Health and Medical Research Council.
Care Around Preterm Birth. Clinical Practice Guidelines. Canberra, Australia: Australian Government Publishing Service;1997
- Greaves R, Kanumakala S, Read A, Zacharin M. Transient 3 beta HSD deficiency and ambiguous genitalia in premature babies.
J Paediatr Child Health.2004; 40
(4):233
–236[CrossRef][Medline]
- Kaplan S, Grumbach M. The ontogenesis of human foetal hormones. II. Luteinizing hormone (LH) and follicle stimulating hormone (FSH).
Acta Endocrinol.1976; 81
(4):808
–829
- Shinkawa O, Furuhashi N, Fukaya T, Suzuki M, Kono H, Tachibana Y. Changes of serum gonadotropins levels and sex difference in premature and mature infants during neonatal life.
J Clin Endocrinol Metab.1983; 56
(6):1327
–1331[Abstract]
- Massa G, de Zegher F, Vanderschueren-Lodeweyckx. Serum levels of immunoreactive inhibin, FSH, and LH in human infants at preterm and term birth.
Biol Neonate.1992; 61
(3):150
–155[CrossRef][ISI][Medline]
- Trotter A, Maier L, Grill H-J, Kohn T, Heckmann M, Pohlandt F. Effects of postnatal estradiol and progesterone replacement in extremely preterm infants.
J Clin Endocrinol Metab.1999; 84
(12):4531
–4535[Abstract/Free Full Text]
- Solberg HE.
Establishment and Use of Reference Values in Tietz Textbook of Clinical Chemistry. 2nd ed. Philadelphia, PA: WB Saunders Company;1994
:454
–484
- Schmidt H, Schwarz HP. Serum concentrations of LH and FSH in the healthy newborn.
Eur J Endocrinol.2000; 143
(2):213
–215[Abstract]
- Midgley PC, Russell K, Oates N, Shaw JCL, Honour JW. Activity of the adrenal fetal zone in preterm infants continues to term.
Endocr Res.1996; 22
(4):729
–733[Medline]
- Corbier P, Dehennin L, Castanier M, Mebazaa A, Edwards D, Roffi J. Sex differences in serum luteinizing hormone and testosterone in the human neonate during the first few hours after birth.
J Clin Endocrinol Metab.1990; 71
(5):1344
–1348[Abstract]
- Mercado M, Yu VY, Fancis I, Szymonowicz W, Gold H. Thyroid function in very preterm infants.
Early Hum Dev.1988; 16
(2–3):131
–141[CrossRef][Medline]
- Sikaris K. McLachlan R, Kazlauskas R, de Kretser D, Holden C, Handelsman D. Reproductive hormone reference intervals for healthy fertile young men: evaluation of automated platform assays.
J Clin Endocrinol Metab.2005; 90
(11):5928
–5936[Abstract/Free Full Text]
PEDIATRICS (ISSN 1098-4275). ©2008 by the American Academy of Pediatrics
TOP
ABSTRACT
METHODS
