Reference Values for Thyroid Function Tests in Pregnant Women Living in Catalonia,Spain

Abstract

Dear Editor:

Early detection of thyroid dysfunction during pregnancy is crucial to avoid fetal disturbances. Maternal hypothyroidism and/or maternal hypothyroxinemia can lead to impairment of fetal cerebral development (1). The pregnant state induces important changes in thyroid physiology (2), including an increase in thyroid binding globulin levels, a transient decrease in free thyroxine, a decrease in thyroid-stimulating hormone (TSH) at first trimester related to β-human chorionic gonadotropin (HCG) rise, and an increase in renal iodine clearance. These modifications of thyroid homeostasis determine specific thyroid hormone circulating patterns according to gestational age; this changing scenario in thyroid gestational physiology challenges the diagnosis of thyroid dysfunction in pregnant women because of the lack of precise reference values (RVs) obtained in specific populations and for each of the trimesters. The RVs of the available commercial kits for thyroid hormones are almost always related to the general population and therefore are not valid for pregnant women. Although there is a necessity to have specific RVs for thyroid hormones during pregnancy and during each trimester, only a few studies have been published covering specific geographical areas (3 –14). Because of this relative lack of data, we conducted a study to determine the RVs of thyroid hormones in a population of pregnant women from Catalonia (Spain).

Two hundred and seventy-six pregnant women were recruited consecutively in the first trimester of gestation and a further 130 women were selected in their final trimester. Their mean age was 29 ± 4 years; 4.7% were multiparus and the rest were nulliparus or uniparus. Women previously treated with thyroxine or with a known thyroid disorder or had a family history of thyroid disorder were excluded. More than 90% of these women were born in Spain, and the rest in Morocco, Peru, and Ecuador. All were living in the same location for at least 2 years before the study. The study group covered different geographical areas of Catalonia, including the Pyrenees and a costal area where there is an acceptable iodinated status for the general population, as shown by recent epidemiologic studies aimed at investigating this condition (15). Recruitment was made through the gynecological services at primary healthcare centers and hospitals where pregnant women were being monitored. Of the 276 women initially contacted, samples were available from 220 in the first trimester and from 60 of those 220, for the third trimester measurement. Of these 220 women recruited in the first trimester, 43.1% were usual consumers of iodinated salt, while 13.4% had taken pharmacologic supplements of 150 μg potassium iodide (KI) during the first trimester and 53.3% thereafter. The study was approved by the Ethics Committee of Hospital Dos de Maig, Barcelona. Written consent was obtained from all the participants.

Samples were taken during a fasting state at week 9 for the first trimester and at week 32 for the third trimester. A quimioluminescent assay (Advia Centaur, Bayer Tarrytown, NY) was used to measure free thyroxine (FT4) (RVs for nonpregnant: 0.8–2 ng/dL; intrassay coefficient of variation (CV): 5.4%), TSH (RVs for nonpregnant: 0.4–4 mU/mL; intrassay CV: 5.1%), and antiperoxidase antibodies (TPO abs) (positive > 60 IU/mL; intrassay CV: 6.6%). Total thyroxine (TT4) was measured with a radioimmunoassay (RIA) method (Diagnostic Systems Laboratories, Webster, TX; inter- and intrassay CVs: 7.4% and 5.1%) only in the first trimester. Urinary iodine was determined by the method described by Pino et al. (16) with inter- and intrassay CVs of 15.5% and 12.6%.

For the calculation of RVs of FT4, TSH, and TT4, a nonparametric method was applied, in which the ordinal value despite the real value was used. The calculated reference interval obtained by these procedures also included 95% of the sample, although under these conditions the interval was calculated using the 2.5 and 97.5 percentiles of the n data of the sample ordered from the lowest to the highest value. The lower and higher limits of the interval are the ordinal values corresponding to 0.025 (N + 1) and 0.975 (N + 1), respectively. Initially the data obtained for the first and third trimesters of gestation were analyzed in all the samples, without consideration of TPO abs status and urinary iodine. Subsequently, the same analysis was performed using only those cases in which TPO abs were negative and a urinary iodine was higher than 150 and 200 μg/L (for the categorization of potential iodine deficiency/sufficiency).

Median urinary iodine was 163 μg/L for the first trimester and 172 μg/L for the third trimester. TPO abs at titers >60 U/mL were present in 17.6% during the first trimester and 9.7% during the third. Table 1 shows the RVs, mean, and median of FT4, TT4, and TSH obtained in our sample according to iodine urinary excretion and TPO abs status; no statistical differences in all hormonal values were found when the cutoff of 150 or 200 μg/L of urinary iodine, or positive TPO abs were considered. When TPO abs-positive women were excluded and no low cutoff for urinary iodine excretion was considered, the RVs for FT4 were similar to those found for the other conditions, although the upper cutoff of RVs for TSH was slightly lower during the first trimester, but with no statistically significant differences.

Table 1.

Reference Values of Thyroid Hormones in the First and Third Trimester of Pregnancy in Three Different Conditions

	First trimester					Third trimester
	n	RVs	P10	Median	Mean (SD)	n	RVs	P10	Median	Mean (SD)
FT4 (ng/dL)^a	220	0.80–1.67	0.86	1.09	1.11 (0.22)	60	0.70–1.20	0.79	0.90	0.92 (0.12)
FT4 (ng/dL)^b	121	0.79–1.86	0.88	1.07	1.10 (0.22)	42	0.67–1.18	0.79	0.90	0.90 (0.10)
FT4 (ng/dL)^c	178	0.80–1.60	0.86	1.08	1.10 (0.20)	57	0.69–1.2	0.80	0.90	0.91 (0.12)
TT4 (μg/dL)^a	220	5.87–14.3	7.26	9.7	9.82 (2.2)	60	6.38–15.9	8.78	10.8	11.1 (1.99)
TT4 (μg/dL)^b	121	5.30–18.1	7.55	9.8	10 (2.54)	42	6.20–16.1	8.52	10.9	11.3 (2.2)
TT4 (μg/dL)^c	178	5.78–16.5	7.76	9.8	10.06 (2.23)	57	6.27–16.03	8.9	11.1	11.17 (2)
TSH (mU/mL)^a	220	0.12–5.76	—	1.43	1.97 (2.37)	60	0.29–4.46	—	2.14	2.14 (1.03)
TSH (mU/mL)^b	121	0.15–5.49	—	1.42	1.79 (1.18)	42	0.27–4.57	—	2.25	2.19 (1.11)
TSH (mU/mL)^c	178	0.12–4.75	—	1.36	1.66 (1.14)	57	0.28–4.48	—	2.18	2.16 (1.04)

Total population of pregnant women.

Population with negative TPO abs and urinary iodine >150 μg/L.

Population with negative TPO abs independent of urinary iodine status.

RVs, reference values; P10, percentile 10; FT4, free thyroxine; TT4, total thyroxine; TSH; thyroid-stimulating hormone; TPO abs, antiperoxidase antibodies.

When the same women were studied in both first and third trimesters (n = 60), a significant decrease in FT4 was observed (median: 1.10 vs. 0.90 ng/dL; p < 0.001), whereas TSH showed a significant increase (median: 1.55 vs. 2.16 mIU/mL; p < 0.010). Correlation between FT4 and TSH was weak in the first trimester (r _s = –0.17; p = 0.004), whereas there was no correlation in the third (r _s = −0.21; p = 0.07).

The frequent lack of local thyroid hormone RVs for pregnant women in most countries does not allow an appropriate interpretation of thyroid function tests during pregnancy, with a potential threat for maternal and fetal health. In Europe, a region with heterogeneous iodine status (17) data has been investigated, in United Kingdom with Caucasian women (5,11,12), the Asian population (11), Sweden, Belgium, and Germany in a combined multinational study (18,19), and in France (20). There are a few publications related to these normative data in others populations with different ethnic origins: Arabs (9), diverse American studies targeting different geographical and ethnic-based ancestral backgrounds (10,21), Japanese (7), Hong Kong (3), Brazilians (4), and India (8). Table 2 shows the normative data and method of measurement used in these studies as well as our own data. Most of these studies have been performed in countries with acceptable iodine status in the population, but it is noteworthy the differences in the reference range published. Moreover, the number of women included in each of the different studies aimed at establishing RVs is variable. Taken together, the reported European data—with the exception of the recent study by Vaidya including 1147 women—comprised of around 100 pregnant women per group. The purpose of establishing diagnostic criteria of gestational hypothyroxinemia—of most relevance for the fetus—still seems to be an elusive goal that requires further trimester- and geographic-specific studies.

Table 2.

Reference Values of Free Thyroxine of Different Studies

Groups (ref. no.)	Location/country	n	Analysis method	RVs in nonpregnant population	RVs for first trimester (weeks)	RVs for second trimester (weeks)	RVs for third trimester (weeks)	Iodine sufficiency and negative TPO abs
Our results	Catalonia (Spain)	178 (1T); 57 (3T)	Immunoanalytic assay—Adviar Centaur Bayer	0.8–2 ng/dL	0.80–1.60 ng/dL (9)	—	0.69–1.2 ng/dL (35)	TPO abs negative
Panesar et al. (3)	Hong Kong (China)	343	Immunoassay kit on ACS 180 (Chiron Diagnostics Corporation)	0.93–1.74 ng/dL	1.01–2.09 ng/dL (11)	0.87–1.41 ng/dL (19)	0.71–1.25 ng/dL (31)	Not determined
Kurioka et al. (7)	Japan	119 (1T); 132 (2T); 136 (3T)	Electrochemiluminescence immunoassay (ECLusys Roche)	1.0–1.8 ng/dL	1.16–1.95 ng/dL (<14)	0.85–1.37 ng/dL (15–28)	0.77–1.27 ng/dL (>28)	Not determined
Dhatt et al. (9)	United Arab Emirates	Arabs/Asian: 97/79 (1T); 252/174 (2T); 52/21 (3T)	Chemiluminescent immunoassay (Abbott Architect i2000)	0.89–1.7 ng/dL	Arabs (UAE): 0.81–2.25 ng/dL; Asian: 1.03–2.0 ng/dL	Arabs (UAE): 0.77–1.76 L ng/dL; Asian: 0.88–1.69 ng/dL	Arabs (UAE): 0.73–1.64 ng/dL; Asian: 0.81–1.52 ng/dL	TPO abs negative
Deam et al. (29)	Australia	20 (3T)	Amerlex-M GammaCoat 2-Step RIA Amerlite MagicLite	0.82–2.47 ng/dL; 0.82–2.38 ng/dL; 0.91–2.29 ng/dL; 1.18–2.24 ng/dL	—	—	0.55–1.10 ng/dL; 0.75–1.37 ng/dL; 0.46–1.20 ng/dL; 0.36–1.64 ng/dL	Not determined
Vieira et al. (4)	Brazil	132	Auto DELFIA	0.7–1.5 ng/dL	0.61–1.21 ng/dL (?)	0.59–0.99 ng/dL (?)	0.43–0.91 ng/dL (?)	Not determined
Sapin and D'Herbomez (30)^a	France	29 (3T)	Direct Equilibrium Dialysis (Nichols ED RIA) and others^a	0.94–3.17 ng/dL	—	—	0.68–1.52 ng/dL (>25)	Not determined
Sheehan and Christofides (19)	United Kingdom Germany France Belgium Sweden	66 (1T); 72 (2T); 71 (3T)	Amerlex-MAB	1.04–2.23 ng/dL	1.09–2.07 ng/dL (?)	0.80–1.87 ng/dL (?)	0.7–1.56 ng/dL (?)	Not determined
Christofides and Sheehan (18)	United Kingdom Germany Sweden	84 (1T); 131 (2T); 128 (3T)	Amerlite MAB (chemiluminescent immunoassay)	1.12–2.53 ng/dL	0.94–2.11 ng/dL (?)	0.86–1.69 ng/dL (?)	0.75–1.47 ng/dL (?)	Not determined
Vaidya et al. (12)	United Kingdom	1147 (1T)	Electrochemiluminescent immunoassay (Roche Diagnostics)	1.09–2.1 ng/dL	0.97–1.86 ng/dL (<12)	—	—	TPO abs negative
Stricker et al. (31)	Switzerland	575 (1T); 193 (2T); 203 (3T)	Chemiluminescent immunoassay (Abbott ARCHITECT i2000_SR)	0.7–1.48 ng/dL	0.81–1.44 ng/dL (6–12)	0.73–1.09 ng/dL (18–24)	0.66–1.05 ng/dL	TPO abs negative
Cotzias et al. (32)	United Kingdom	307	Immunoanalytic assay—Adviar Centaur Bayer	0.7–2 ng/dL	0.75–1.25 ng/dL (9w)	0.69–1.18 ng/dL (20)	0.6–1.09 ng/dL	Not determined
Pearce et al. (13)	United States	585 (3T)	Free T4 index	1.0–4.0 ng/dL	1.5–1.9 ng/dL (<14)	—	—	TPO abs negative
Kahric-Janicic et al. (25)^b	United States	59 (1T); 35 (2T); 26 (3T)	Tandem mass spectrometry	0.44–1.42 ng/dL	0.68–1.58 ng/dL (8.7)	0.33–1.51 ng/dL (17.8)	0.45–1.27 ng/dL (28.7)	Not determined

For conversion, FT4: pmol/L = ng/dL × 12.82.

Study with other nine methods: Nichols ED/RIA; immunoassays: Elecsys (Roche, Madrid, Spain); VIDAS (bioMérieux, Madrid, Spain), Vitros ECi (Ortho-Clinical Diagnostics, Madrid, Spain), GammaCoat 2-Step RIA (DiaSorin, Madrid, Spain); Immulite (Diagnostic Product Corp./Siemens, Los Angeles, CA); Nichols Advantage, AxSYM (Abbott Diagnostics, Madrid, Spain); ACS (Bayer Diagnostic, Alcalá, Spain); AIA (Tosoh Bioscience, Tessenderlo, Belgium).

Data obtained after calculation of FT4 (mean ± 1.96 × standard error) from range values by Kahric-Janicic et al. (25).

1T, first trimester; 3T, third trimester; ?, no data on gestational age are available.

Our data were generated in a group of acceptably well-iodinated women with negative TPO abs, as positive TPO abs in iodine sufficient population are associated with overt and subclinical hypothyroidism, as indicated by important epidemiological studies, that is, the National Health and Nutrition Examination Survey III in which these two conditions were detected in 6.9% and 4%, respectively (22); however, probably because of the limited sample of our study, we did not find differences in TSH and FT4 in relation to TPO abs. The urinary iodine of our population seems to be sufficient for pregnancy, according to the recommendations of WHO (23), and indicates that Catalonia, a geographical area of northeast Spain, has improved its history of iodine deficiency over the last 20 years (median of urinary iodine in nonpregnant women population: 148 μg/L), which is attributable to multiple campaigns promoted by the regional Catalan government and educational efforts performed by devoted local endocrinologists (15,24). Despite a small sample size of the third trimester (low for National Academy of Clinical Biochemistry [NACB] guidelines recommendation), our RVs of the FT4 and TSH in the different situations evaluated (TPO abs −/+ and iodine sufficiency) are concordant. Roughly, our FT4 RVs are in a range similar to those published in other countries and geographical locations, and probably, methodological issues may explain most of the differences observed in relation to these published studies (Table 2). This warrants the necessity of generating local, trimester- and method-specific RVs, as there is no universal consensus on common methodology and the accepted gold standard procedures such as equilibrium dialysis or mass spectrometry are neither useful nor convenient for daily clinical practice. Maybe mass spectrometry is the most promising method to be used for a more accurate measurement of FT4 (25), overcoming all the inconsistencies of the immunoassays; however, until now, its use remains limited probably because of its recent development. It is difficult to recommend the most adequate hormonal parameter based on the robustness of the technique used, its physiological variability, and change over time, which is typical of the dynamic situation of pregnancy, where there is virtually no steady-state conditions because of the continuous changing influences of many factors on the thyroid physiology of pregnant women. Although TSH measurement is methodologically much more robust than the methods used for FT4, the lag time to attain a new steady state for TSH makes it not so real-time informative, and thus questions its usefulness for evaluation of thyroid function in the first trimester. TSH values may not fully reflect coetaneous FT4 maternal and fetal concentrations (1). Further, considering the TSH cutoff value of 2.5 mIU/L as suggested by the Endocrine Society Guideline (26), a surprisingly high number of our women (25%) were shown to be out of range, suggesting a hypothyroid status, while having a normal FT4 according to our normative values. For these reasons, we consider that TSH values should also be taken into consideration as a secondary important parameter for evaluating thyroid function in pregnant women. Although apparently well iodinated, this 25% of women with TSH values above 2.5 mIU/L is a matter of concern and poses the question if a normal urinary iodine for the general population is sufficient at the time of initiation of pregnancy. As we do not have yet the opportunity to explore IQ and other cognitive performance in the offspring of these women, we do not know if the low range of normal FT4 and supranormal (>2.5 mIU/L) TSH is of relevance for the progeny.

Thyroxine is the key hormone in terms of fetal development (1) and probably or possibly the better parameter to be used, either FT4 or TT4. FT4 is easily measured using automatized procedures in most clinical laboratories, but it only gives an estimation of true thyroxine pool, and for this reason its use has been strongly criticized by some authors as previously mentioned (27). TT4 has been recommended as an appropriate tool for evaluation of thyroid function during pregnancy, because it is technically more robust in terms of measurement accuracy at extremes of binding protein concentrations; however, the procedure is more time consuming and more expensive (28). Our data showed a low normal cutoff value for TT4 of 5.3 μg/dL for the first trimester, with a 10th percentile value of 7.5, close to the recommended 7.8 μg/dL cutoff value for TT4 by the Atlanta Conference for the first trimester. But, again, in our population, 24% would be hypothyroxinemic with this cutoff of 7.8 μg/dL. These observations indicate the importance of generating local normative values, as our population had an acceptable iodine nutrition status—at least for the general population—and yet, no differences were observed in the obtained RVs when either cutoff values of 150 or 200 μg/L of urinary iodine were considered.

Even when taking into account the criticism as to the usefulness of FT4 measured by immunoassay as a reliable indicator in pregnancy and that of mass spectrometry, when our data are compared with those obtained with mass spectrometry, the values are similar (25), with our low normal values slightly above those found by mass spectrometry. This implies that if we were to treat pregnant women with thyroxine according to our reference range, we would be overtreating only a very small number of women, which in fact would be of very minimal risk, but yet, on the other hand, no single hypothyroxinemic pregnant would not be treated. Taking into account all these considerations and with our current methodological limitations, the potential application of FT4 measurement with our immunoassay for the evaluation of thyroid function during pregnancy seems to be an acceptable option, and we think it is still recommendable provided that trimester-specific reference range data generated with local population variables are available.

Footnotes

Acknowledgment

This study was supported by “Fundación Sal y Salud.”

References

Morreale de Escobar

, Obregon

, Escobar del Rey

. 2000. Is neuropsychological development related to maternal hypothyroidism or to maternal hypothyroxinemia? J Clin Endocrinol Metab, 85:3975–3987.

Glinoer

. 2004. The regulation of thyroid function during normal pregnancy: importance of the iodine nutrition status. Best Pract Res Clin Endocrinol Metab, 18:133–152.

Panesar

, Li

, Rogers

. 2001. Reference intervals for thyroid hormones in pregnant Chinese women. Ann Clin Biochem, 38:329–332.

Vieira

, Kanashiro

, Tachibana

, Ghiringhello

, Hauache

, Maciel

. 2004. Free thyroxine values during pregnancy. Arq Bras Endocrinol Metab, 48:305–309.

Parker

. 1985. Amerlex free triiodothyronine and free thyroxine levels in normal pregnancy. Br J Obstet Gynaecol, 92:1234–1238.

Franklyn

, Sheppard

, Ramsden

. 1983. Serum free thyroxine, free triiodothyronine concentrations in pregnancy. Br Med J (Clin Res Ed), 287:394.

Kurioka

, Takahashi

, Miyazaki

. 2005. Maternal thyroid function during pregnancy and puerperal period. Endocr J, 52:587–591.

Kumar

, Gupta

, Nath

, Sharma

. 2003. Thyroid function tests in pregnancy. Indian J Med Sci, 57:252–258.

Dhatt

, Jayasundaram

, Wareth

, Nagelkerke

, Jayasundaram

, Darwish

, Lewis

. 2006. Thyrotrophin and free thyroxine trimester-specific reference intervals in a mixed ethnic pregnant population in the United Arab Emirates. Clin Chim Acta, 370:147–151.

10.

Haddow

, Knight

, Palomaki

, McClain

, Pulkkinen

. 2004. The reference range and within-person variability of thyroid stimulating hormone during the first and second trimesters of pregnancy. J Med Screen, 11:170–174.

11.

Price

, Obel

, Cresswell

, Catch

, Rutter

, Barik

, Heller

, Weetman

. 2001. Comparison of thyroid function in pregnant and non-pregnant Asian and western Caucasian women. Clin Chim Acta, 308:91–98.

12.

Vaidya

, Anthony

, Bilous

, Shields

, Drury

, Hutchison

, Bilous

. 2007. Detection of thyroid dysfunction in early pregnancy: universal screening or targeted high-risk case finding? J Clin Endocrinol Metab, 92:203–207.

13.

Pearce

, Oken

, Gillman

, Lee

, Magnani

, Platek

, Braverman

. 2008. Association of first-trimester thyroid function test values with thyroperoxidase antibody status, smoking, and multivitamin use. Endocr Pract, 14:33–39.

14.

Lambert-Messerlian

, McClain

, Haddow

, Palomaki

, Canick

, Cleary-Goldman

, Malone

, Porter

, Nyberg

, Bernstein

, D'Alton

. 2008. First- and second-trimester thyroid hormone reference data in pregnant women: a FaSTER (First- and Second-Trimester Evaluation of Risk for aneuploidy) Research Consortium study. Am J Obstet Gynecol, 199:62–66.

15.

Vila

, Castell

, Wengrovicz

, de Lara

, Casamitjana

. 2006. Urinary iodide assessment of the adult population in Catalonia. Med Clin (Barc), 127:730–733.

16.

Pino

, Fang

, Braverman

. 1996. Ammonium persulfate: a safe alternative oxidizing reagent for measuring urinary iodine. Clin Chem, 42:239–243.

17.

Delange

. 2002. Iodine deficiency in Europe and its consequences: an update. Eur J Nucl Med Mol Imaging, 29,Suppl:S404–S416.

18.

Christofides

, Sheehan

. 1995. Enhanced chemiluminescence labeled-antibody immunoassay (Amerlite-MAB) for free thyroxine: design, development, and technical validation. Clin Chem, 41:17–23.

19.

Sheehan

, Christofides

. 1992. One-step, labeled-antibody assay for measuring free thyroxin. II. Performance in a multicenter trial. Clin Chem, 38:19–25.

20.

Sapin

, D'Herbomez

, Schlienger

. 2004. Free thyroxine measured with equilibrium dialysis and nine immunoassays decreases in late pregnancy. Clin Lab, 50:581–584.

21.

Soldin

, Hilakivi-Clarke

, Weiderpass

, Soldin

. 2004. Trimester-specific reference intervals for thyroxine and triiodothyronine in pregnancy in iodine-sufficient women using isotope dilution tandem mass spectrometry and immunoassays. Clin Chim Acta, 349:181–189.

22.

Hollowell

, Staehling

, Flanders

, Hannon

, Gunter

, Spencer

, Braverman

. 2002. Serum TSH, T(4), and thyroid antibodies in the United States population (1988 to 1994): National Health and Nutrition Examination Survey (NHANES III) J Clin Endocrinol Metab, 87:489–499.

23.

ICCIDD. 2007. Iodine requirements in pregnancy and infancy. A WHO Technical Consultation has produced new guidelines on iodine requirements and monitoring in these vulnerable age groups. IDD Newsl, 23:1–2.

24.

Serra Majem

, Lloveras

, Vila

, Salleras

. 1993. Strategies to prevent iodine deficiency disorders in Catalonia (1983–1992) Endocrinología, 40:273–277.

25.

Kahric-Janicic

, Soldin

, West

, Gu

, Jonklaas

. 2007. Tandem mass spectrometry improves the accuracy of free thyroxine measurements during pregnancy. Thyroid, 17:303–311.

26.

Abalovich

, Amino

, Barbour

, Cobin

, De Groot

, Glinoer

, Mandel

, Stagnaro-Green

. 2007. Management of thyroid dysfunction during pregnancy and postpartum: an Endocrine Society Clinical Practice Guideline. J Clin Endocrinol Metab, 92:S1–S47.

27.

Soldin

, Tractenberg

, Soldin

. 2004. Differences between measurements of T4 and T3 in pregnant and nonpregnant women using isotope dilution tandem mass spectrometry and immunoassays: are there clinical implications? Clin Chim Acta, 347:61–69.

28.

Mandel

, Spencer

, Hollowell

. 2005. Are detection and treatment of thyroid insufficiency in pregnancy feasible? Thyroid, 15:44–53.

29.

Deam

, Goodwin

, Ratnaike

. 1991. Comparison of four methods for free thyroxin. Clin Chem, 37:569–572.

30.

Sapin

, D'Herbomez

. 2003. Free thyroxine measured by equilibrium dialysis and nine immunoassays in sera with various serum thyroxine-binding capacities. Clin Chem, 49:1531–1535.

31.

Stricker

, Echenard

, Eberhart

, Chevailler

, Perez

, Quinn

, Stricker

. 2007. Evaluation of maternal thyroid function during pregnancy: the importance of using gestational age-specific reference intervals. Eur J Endocrinol, 157:509–514.

32.

Cotzias

, Wong

, Taylor

, Seed

, Girling

. 2008. A study to establish gestation-specific reference intervals for thyroid function tests in normal singleton pregnancy. Eur J Obstet Gynecol Reprod Biol, 137:61–66.