Background: Maternal thyroid dysfunction has been associated with a variety of adverse pregnancy outcomes. Laboratory measurement of thyroid function plays an important role in the assessment of maternal thyroid health. However, occult thyroid disease and physiologic changes associated with pregnancy can complicate interpretation of maternal thyroid function tests (TFTs).
Objective and methods: To 1) establish the prevalence of laboratory evidence for autoimmune thyroid disease (AITD) in pregnant women; 2) establish gestational age-specific reference intervals for TFTs in women without AITD; and 3) examine the influence of reference intervals on the interpretation of TFT in pregnant women. Serum samples were collected from 2272 pregnant women, and TFT performed. Gestational age-specific reference intervals were determined in women without AITD, and then compared with the non-pregnant assay-specific reference intervals for interpretation of testing results.
Results: Thyroid peroxidase antibodies (TPO-Ab) and thyroglobulin antibodies (Tg-Ab) were positive in 10.4 and 15.7% of women respectively. TPO-Ab level was related to maternal age, but TPO-Ab status, Tg-Ab status, and Tg-Ab level were not. Women with TSH > 3.0 mIU/l were significantly more likely to be TPO-Ab positive. Gestational age-specific reference intervals for TFT were significantly different from non-pregnant normal reference intervals. Interpretation of TFT in pregnant women using non-pregnant reference intervals could potentially result in misclassification of a significant percentage of results (range: 5.6–18.3%).
Conclusion: Laboratory evidence for thyroid dysfunction was common in this population of pregnant women. Accurate classification of TFT in pregnant women requires the use of gestational age-specific reference intervals.
During pregnancy, proper maternal thyroid function is important for both the mother and child (1). This is especially true during the first trimester, when the developing fetus is completely dependent on the mother for thyroid hormones that are critical for optimal development (2). Maternal thyroid dysfunction during pregnancy has been shown to be associated with a number of adverse outcomes. For example, elevated maternal thyroid-stimulating hormone (TSH) has been associated with an increased risk of pre-term birth, placental abruption, fetal death, and impaired neurological development in the child (3, 4, 5). Similarly, the presence of antibodies to thyroid peroxidase (TPO-Ab) has been associated with increased risk of miscarriage, pre-term birth, and maternal post partum thyroid disease (6, 7). These findings are significant, because laboratory evidence of thyroid dysfunction is common in women of reproductive age, with the prevalence of elevated TSH ranging from 4 to 9%, and the prevalence of TPO-Ab ranging from 11.3 to 18% in this population (8, 9). For these reasons, it is important to ensure optimal maternal thyroid function during pregnancy. However, pregnant women with thyroid disease do not always develop symptoms, and when they do, these symptoms can sometimes be attributed to the pregnancy itself (10). In these situations, accurate laboratory assessment of maternal thyroid function assumes an even greater importance.
Pregnancy produces a series of profound physiologic changes in the mother that have a significant affect on maternal thyroid function. These changes can, in turn, complicate the interpretation of maternal thyroid function tests (TFTs) (11). As a consequence, the United States National Academy of Clinical Biochemistry (NACB) recommends that ‘trimester-specific reference values should be used when reporting thyroid test values for pregnant patients’ (12). In order for these reference intervals to be relevant, they must be determined using well-characterized specimens, and the NACB guidelines call particular attention to the fact that specimens used for such studies should not contain thyroid auto-antibodies (i.e. TPO-Ab and Tg-Ab). The purpose of our study was to determine the prevalence of laboratory evidence for autoimmune thyroid disease (AITD) in pregnant women, establish gestational age-specific reference intervals for thyroid hormones in women without thyroid autoimmunity, and examine the implications of these reference intervals for the interpretation of TFTs in pregnant women.
Subjects and methods
This study was performed on surplus, de-identified serum specimens collected from pregnant women in the Geneva, Switzerland area as part of their routine antenatal care. Exclusion criteria were: miscarriage, currently undergoing treatment for thyroid disease, and any evidence of genetic abnormality in the fetus (e.g. trisomy). A total of 2272 samples were included in the study. For specimens from the first trimester, gestational age was calculated using ultrasound and crown-rump length (13); for the second and third trimesters, gestational age was established based on last menstrual period and ultrasound. For each sample, TSH, free thyroxine (FT4), total T4 (TT4), free tri-iodothyronine (FT3), total T3 (TT3), TPO-Ab, and thyroglobulin antibodies (Tg-Ab) were measured using the Abbott ARCHITECT i2000SR Analyzer (Abbott Diagnostics). The ARCHITECT i2000SR is a high-throughput immunoassay analyzer that uses paramagnetic microparticles and chemiluminescent detection technology (14). Each individual sample was tested a single time for each analyte. For TPO-Ab and Tg-Ab, results above the manufacturer’s reference limit were considered positive (TPO-Ab > 5.61 IU/ml; Tg-Ab > 4.11 IU/ml). Specimens positive for TPO-Ab and/or Tg-Ab were excluded from the reference interval analysis. TSH data were log transformed for analysis (12). Data analysis was performed using Analyse-It software, version 1.73 (Leeds, UK). ANOVA (continuous variables), χ2 or Fisher’s exact test (categorical variables), and Spearman rank correlation were used to analyze the data. A two-tailed P value < 0.05 was judged to be statistically significant. The study design and protocol were reviewed and approved by the Internal Institutional Review Board.
For the study population, median maternal age was 30.8 years (range = 17.8–44.3 years). TPO-Ab and Tg-Ab were positive in 10.4 and 15.7% of women respectively. For the first trimester, a total of 1014 women were tested. Mean maternal and median gestational age was 30.4 years and 7.6 weeks respectively. Of these women, 21.4% were positive for TPO-Ab and/or Tg-Ab; 10.8% were TPO-Ab positive and 17.3% were Tg-Ab positive. For the second trimester, a total of 661 women were tested. Mean maternal and median gestational ages were 30.5 years and 16 weeks respectively. Of these women, 19.8% were positive for TPO-Ab and/or Tg-Ab; 10.8% were TPO-Ab positive and 17.3% were Tg-Ab positive. For the third trimester, a total of 598 women were tested. Mean maternal age was 31.4 years (significantly different from the first and second trimester mean maternal age, P < 0.001); median gestational age was 32 weeks. Of these women, 15.7% were positive for TPO-Ab and/or Tg-Ab; 9.9% were TPO-Ab positive and 11.7% were Tg-Ab positive. Overall, TPO-Ab and Tg-Ab status were not related to maternal age (P = 0.7394 and P = 0.1758 respectively). TPO-Ab level was related to maternal age (P = 0.035), but Tg-Ab level was not (P = 0.7776). Mean thyroid hormone values were compared in women with and without thyroid auto-antibodies. TSH and FT3 were significantly higher in antibody positive women (mean TSH of 1.38 vs 0.98 mIU/l, P = 0.0003; mean FT3 of 4.72 vs 4.56 pmol/l, P = 0.0429). Mean FT4 and TT4 were not influenced by antibody status (P = 0.2577 and P = 0.8232 respectively).
Gestational age-specific reference intervals in antibody negative women for each assay are shown in Table 1. These reference intervals were used to classify TFT results (e.g. ‘high’ = above 97.5th confidence limit, ‘normal’ = within central 95% confidence interval, and ‘low’ = below 2.5th confidence limit), and were then compared with classifications determined using the non-pregnant assay-specific reference intervals provided by the assay manufacturer. The number and percentage of results potentially misclassified if non-pregnant reference intervals were used are summarized in Table 2. For TSH, a total of 82 (3.6%) women with elevated TSH would not have been identified, and 83 (3.7%) women would have been incorrectly classified as having a low TSH. Potential for misclassification of TSH results was greatest in the first trimester (10.4%). For FT4, 43 (1.9%) women with elevated results would not have been identified. In the first and second trimesters, 38 women (2.3%) with low FT4 would not have been identified. In the third trimester, five women (0.83%) would have been incorrectly identified as having low FT4. Potential for misclassification of FT4 results was greatest in the second trimester (5.2%). For TT4, 357 women (15.9%) would have been incorrectly identified as high, and 55 women (2.4%) with low TT4 would not have been identified. Potential for misclassification of TT4 results was greatest in the third trimester (28.5%). For FT3, 67 (3.0%) women would have been incorrectly identified as high, and 58 women (2.6%) would not have been identified as low. Potential for misclassification was greatest in the first trimester (8.3%). For TT3, 326 (14.5%) women would have been incorrectly identified as high, and 70 women (3.1%) would not have been identified as low. Potential for misclassification of TT3 results was greatest in the third trimester (25%). The relationship between maternal TFT and gestational age in women without AITD is shown graphically in Fig. 1.
In our cross-sectional study, we have determined the prevalence of laboratory evidence for AITD in pregnant women, established gestational age-specific reference intervals for TFT in pregnant women without laboratory evidence of AITD, and examined the effect of reference intervals on the interpretation of maternal TFT. Iodine nutrition status was not assessed in this study, as it has been previously reported that pregnant women in Switzerland are iodine sufficient (15).
Our data show laboratory evidence of AITD is common in this population, with 19.4% of women testing positive for TPO-Ab and/or Tg-Ab. In women aged 18–44 years, 9.4–15.1% were TPO-Ab positive. Our data roughly agree with those of Hollowell et al. who reported a prevalence for TPO-Ab of 11.3–18% in non-pregnant women aged 18–45 years (9). In our population, women with TSH > 3.0 mIU/l were more likely to be TPO-Ab positive (RR = 4.6 (3.4–6.3), P < 0.0001). Regarding elevated TSH, the prevalence of women with TSH above the 97.5th percentile for gestational age was higher in our study (3.6%) than that of Dashe et al. (2.5%) (16). This could be due to the fact that reference intervals in the present study excluded women positive for TPO-Ab and Tg-Ab, and this would be expected to lower the upper reference limit for TSH (12).
Gestational age-specific reference intervals were established for TFT after exclusion of women positive for TPO-Ab and/or Tg-Ab tests (12). In manycases, these reference intervals are significantly different from those reported by the assay manufacturer for non-pregnant women. As our data show, utilization of non-pregnant reference intervals to interpret TFT in pregnant women has the potential to result in a large number of misclassified results, and could contribute to suboptimal patient care. Since many algorithms for diagnosing thyroid disease start with serum TSH measurement (12), this is of particular concern for interpretation of maternal TSH values, and our data demonstrate that a large number of these tests would be misclassified when using non-pregnant reference intervals. This is also important for women with established hypothyroidism prior to pregnancy, as these women may require adjustment of their T4 dose, and careful monitoring throughout pregnancy (10). Our data demonstrate that the application of non-pregnant reference intervals to the interpretation of TT4 and TT3 results in pregnant women is particularly problematic, and support the need for gestational age-specific reference intervals when assessing thyroid function in pregnant women using these tests. Direct comparison of our reference intervals to other published data is problematic for several reasons. Previous studies reporting gestational age-specific reference intervals for thyroid hormones (16–19) either did not exclude women with laboratory evidence of thyroid autoimmunity, or did not test the full panel of thyroid hormone assays. In addition, because there is no internationally recognized method for standardization of free thyroid hormone tests, assay results, as well as the influence of pregnancy on assay performance, varies among different assay manufacturers (20, 21). This situation means it is likely that method-specific reference intervals are required for free thyroid hormone assays (12).
Given the prevalence and adverse outcomes associated with maternal thyroid dysfunction, considerable discussion has focused on the possibility that screening pregnant women for thyroid disease could improve outcomes for both mother and child. However, screening pregnant women for thyroid dysfunction remains controversial (22). Opponents of screening point to the lack of controlled clinical trails demonstrating efficacy for screening and subsequent treatment, and suggest that aggressive case finding is more appropriate for identification of maternal thyroid dysfunction during pregnancy (16, 22). A recent report by Negro et al. has shown that T4 therapy in pregnant women with AITD improves outcomes, reducing the rate of miscarriage and pre-term birth in treated women (23). In addition, Vaidya et al. have reported that targeted thyroid function testing of only pregnant women at high risk for thyroid disease (e.g. family history of thyroid disease) would miss about one-third of women with overt and subclinical thyroid disease (18). When seen in the context of extensive literature reporting adverse outcomes in pregnant women with thyroid dysfunction, these recent studies have prompted further discussion about the best approach to the diagnosis and treatment of thyroid dysfunction in pregnant women, and a renewed focus on the topic of screening all pregnant women for thyroid disease (22). Whatever approach is taken to identify thyroid dysfunction in pregnant women, appropriate interpretation of TFT plays a critical role in this process.
In conclusion, our study shows a high prevalence of laboratory evidence for AITD in pregnant women, and that reference intervals for TFTs in pregnant women can be significantly different from those in non-pregnant women. Application of non-pregnant reference intervals to the interpretation of TFT in pregnant women has the potential to result in misclassification of patient test results. Accurate classification of TFT in this population requires the use of gestational age-specific reference intervals.
Reagents used to conduct this study were supplied by Abbott Diagnostics.
Gestational age-specific reference intervals for thyroid hormones in women without thyroid autoimmunity.
|Gestational age (weeks)||N||Mean||Median||2.5th||97.5th|
|First trimester (< 6–12)||783||0.8666||1.0402||0.0878||2.8293|
|Second trimester (> 12–24)||528||0.9358||1.0214||0.1998||2.7915|
|Third trimester (> 24 to term)||501||1.1138||1.1390||0.3070||2.9028|
|> 36 to term||112||1.2349||1.3720||0.3327||2.8907|
|Manufacturer’s non-pregnant reference interval is 0.35–4.94 mIU/l|
|Free T4 (pmol/l)|
|First trimester (< 6–12)||783||13.96||13.79||10.53||18.28|
|Second trimester (> 12–24)||528||12.29||12.17||9.53||15.68|
|Third trimester (> 24 to term)||501||11.19||11.08||8.63||13.61|
|> 36 to term||112||11.11||10.94||9.26||13.61|
|Manufacturer’s non-pregnant reference interval is 9.01–19.05 pmol/l|
|Total T4 (nmol/l)|
|First trimester (< 6–12)||783||114.67||110.64||72.27||171.18|
|Second trimester (> 12–24)||528||136.15||134.84||94.77||182.51|
|Third trimester (> 24 to term)||501||137.97||136.65||94.88||193.35|
|> 36 to term||112||136.25||134.14||93.21||193.60|
|Manufacturer’s non-pregnant reference interval is 62.7–150.8 nmol/l|
|Free T3 (pmol/l)|
|First trimester (< 6–12)||783||4.73||4.67||3.52||6.22|
|Second trimester (> 12–24)||528||4.52||4.47||3.41||5.78|
|Third trimester (> 24 to term)||501||4.30||4.27||3.33||5.59|
|36 to term||112||4.23||4.23||3.30||5.43|
|Manufacturer’s non-pregnant reference interval is 2.63–5.70 pmol/l|
|Total T3 (nmol/l)|
|First trimester (< 6–12)||783||1.84||1.78||1.25||2.72|
|Second trimester (> 12–24)||528||2.16||2.15||1.43||3.16|
|Third trimester (> 24 to term)||501||2.21||2.19||1.40||3.16|
|> 36 to term||112||2.13||2.10||1.33||3.12|
|Manufacturer’s non-pregnant reference interval is 0.89–2.44 nmol/l|
Potential for misclassification of thyroid function tests in pregnant women if non-pregnant reference intervals are used.
|Results potentially misclassified,n (%)|
|N||1st trimester||2nd trimester||3rd trimester||Overall|
|TSH||2248||103 (10.4)||44 (6.7)||18 (3.0)||165 (7.3)|
|Free T4||2250||32 (3.2)||34 (5.2)||20 (3.3)||86 (3.8)|
|Total T4||2246||85 (8.6)||157 (23.8)||170 (28.5)||412 (18.3)|
|Free T3||2251||83 (8.3)||24 (3.6)||18 (3.0)||125 (5.6)|
|Total T3||2250||93 (9.4)||154 (23.4)||149 (25)||396 (17.6)|
LaFranchi SH, Haddow JE & Hollowell JG. Is thyroid inadequacy during gestation a risk factor for adverse pregnancy and development outcomes? Thyroid 2005 15 60–71.
Morreale de Escobar G, Obregon MJ & Escobar del Rey F. Maternal thyroid hormone early in pregnancy and fetal brain development. Best Practice and Research. Clinical Endocrinology and Metabolism 2004 18 225–248.
Casey BM, Dashe JS, Wells CE, McIntire DD, Byrd W, Leveno KJ & Cunningham FG. Subclinical hypothyroidism and pregnancy outcomes. Obstetrics and Gynecology 2005 105 239–245.
Allan WC, Haddow JE, Palomaki GE, Williams JR, Mitchell ML, Hermos RJ, Faix JD & Klein RZ. Maternal thyroid deficiency and pregnancy complications: implications for population screening. Journal of Medical Screening 2000 7 127–130.
Haddow JE, Palomaki GE, Allan WC, Williams JR, Knight GJ, Gagnon J, O’Heir CE, Mitchell ML, Hermos RJ, Waisbren SE, Faix JD & Klein RZ. Maternal thyroid deficiency during pregnancy and subsequent neuropsychological development of the child. New England Journal of Medicine 1999 341 549–555.
Stagnaro-Green A & Glinoer D. Thyroid autoimmunity and the risk of miscarriage. Best Practice and Research. Clinical Endocrinology and Metabolism 2004 18 167–181.
Hollowell JG, Staehling NW, Flanders WD, Hannon WH, Gunter EW, Spencer CA & Braverman LE. Serum TSH, T4, and thyroid antibodies in the United States population (1988 to 1994): National Health and Nutrition Examination Survey (NHANES III). Journal of Clinical Endocrinology and Metabolism 2002 87 489–499.
Demers LM & Spencer CA. Laboratory medicine practice guidelines: laboratory support for the diagnosis and monitoring of thyroid disease. Thyroid 2003 13 3–98.
Robinson P & Fleming JEE. A critical evaluation of sonar ‘crown-rump length’ measurements. British Journal of Obstetrics and Gynaecology 1975 82 702–710.
Quinn FA. Architect i2000 and i2000SR analyzers. In The Immunoassay Handbook, edn 3, pp 406–411. Ed. D Wild, Amsterdam: Elsevier, 2005.
Zimmermann MB, Aeberli I, Torresani T & Burgi H. Increasing the iodine concentration in the Swiss iodized salt program markedly improved iodine status in pregnant women and children: a 5-y prospective national study. American Journal of Clinical Nutrition 2005 82 388–392.
Dashe JS, Casey BM, Wells CE, McIntire DD, Byrd EW, Leveno KJ & Cunningham FG. Thyroid stimulating hormone in singleton and twin pregnancy: importance of gestational age-specific reference ranges. Obstetricia et Gynecologica 2006 106 753–757.
Panesar NS, Li CY & Rogers MS. Reference intervals for thyroid hormones in pregnant Chinese women. Annals of Clinical Biochemistry 2001 38 329–332.
Vaidya B, Anthony S, Bilous M, Shields B, Drury J, Hutchison S & Bilous R. Detection of thyroid dysfunction in early pregnancy: universal screening or targeted high-risk case finding? Journal of Clinical Endocrinology and Metabolism 2007 92 203–207.
Soldin OP, Hilakivi-Clarke L, Weiderpass E & Soldin SJ. Trimester-specific reference intervals for thyroxine and triiodothronine in pregnancy in iodine-sufficient women using isotope dilution tandem mass spectrometry and immunoassays. Clinica Chimica Acta 2004 349 181–189.
Roberts RF, La’ulu SL & Roberts WL. Performance characteristics of seven automated thyroxine and T-uptake methods. Clinica Chimica Acta 2007 377 248–255.
Mandel SJ, Spencer CA & Hollowell JG. Are detection and treatment of thyroid insufficiency in pregnancy feasible? Thyroid 2005 15 44–53.
Brent GA. Editorial: diagnosing thyroid dysfunction in pregnant women: is case-finding enough? Journal of Clinical Endocrinology and Metabolism 2007 92 39–41.
Negro R, Formoso G, Mangieri T, Pezzarossa A, Sazzi D & Hassan H. Levothyroxine treatment in euthyroid pregnant women with autoimmune thyroid disease: effects on obstetrical complications. Journal of Clinical Endocrinology and Metabolism 2006 91 2587–2591.