OBJECTIVE: To further evaluate the relationship between thyroid antibodies and preterm births.
METHODS: This is a prospective study of pregnancy outcome and demographic data combined with retrospective measurement of thyroperoxidase and thyroglobulin antibodies. Sera were obtained at 11–13 and 15–18 weeks of gestation from 10,062 women with singleton viable pregnancies (a subset from the First- and Second-Trimester Risk of Aneuploidy [FaSTER] trial).
RESULTS: Women with elevated levels of thyroperoxidase, thyroglobulin antibodies, or both in the first trimester have a higher rate of preterm delivery before 37 weeks of gestation than antibody-negative women (7.5% compared with 6.4%, odds ratio [OR] 1.18; 95% confidence interval [CI] 0.95–1.46). This is also the case for very preterm delivery before 32 weeks of gestation (1.2% compared with 0.7%, OR 1.70; 95% CI 0.98–2.94). Preterm premature rupture of membranes is also increased (2.0% compared with 1.2%, OR 1.67; 95% CI 1.05–2.44). These associations are less strong for second-trimester antibody measurements.
CONCLUSION: The present data do not confirm strong associations between thyroid antibody elevations and preterm birth found in three of five previously published reports. Preterm premature rupture of membranes appears to contribute to the thyroid antibody-associated early deliveries, possibly as a result of inflammation.
LEVEL OF EVIDENCE: II
Elevated thyroid antibody levels in early pregnancy are weakly associated with early delivery and preterm premature rupture of membranes.
From the Women and Infants Hospital and Alpert Medical School of Brown University, Providence, Rhode Island; the Savjani Institute for Health Research, Standish, Maine; Columbia University College of Physicians and Surgeons, New York, New York; the Royal College of Surgeons in Ireland, Dublin, Ireland; the University of Utah and Intermountain HealthCare, Salt Lake City, Utah; the Swedish Medical Center, Seattle, Washington; and Montefiore Medical Center/Albert Einstein College of Medicine, Bronx, New York.
Partially supported by grant RO1 HD 38652 from the National Institutes of Health and the National Institute of Child Health and Human Development.
Corresponding author: James E. Haddow, MD, Co-Director, Division of Medical Screening & Special Testing, Women and Infants Hospital, Alpert Medical School of Brown University, 70 Elm Street, 2nd Floor, Providence, RI 02903;e-mail: email@example.com.
Financial Disclosure Geralyn Lambert-Messerlian is a consultant to Beckman Coulter, Inc. The other authors did not report any potential conflicts of interest.
Stagnaro-Green1 recently summarized current knowledge about the relationship between thyroid antibodies and prematurity. Three of the reviewed studies reported large increases in the rate of preterm births among antibody-positive women in comparison to women without antibodies (odds ratios [ORs] of 2.22, 4.19, and 3.25, respectively),2–4 whereas two others documented only small increases that were not statistically significant (ORs of 1.33 and 1.39).5,6 Further investigation appears warranted, therefore, to better understand these widely varying associations between elevated thyroid antibody levels and preterm birth. The present study's objective is to examine that relationship retrospectively in a large cohort of pregnancies with thyroid-related measurements that was enrolled secondarily in conjunction with a multicenter observational study (the First- and Second-Trimester Risk of Aneuploidy [FaSTER] trial).7–9 The primary purpose of the FaSTER trial was to study newer screening strategies for Down syndrome.10
MATERIALS AND METHODS
As previously described, participants in the multicenter FaSTER trial were enrolled between October 1, 1999, and December 31, 2002.10 Enrollment was restricted to women with viable singleton pregnancies. Institutional review board approval was obtained at all of the recruitment centers. At five of these centers, supplementary consent was sought from a total of 14,554 participants to allow their residual serum samples to be used for additional research studies. Samples from the 10,990 women who consented at those centers (Montefiore Medical Center, Bronx, New York; Swedish Medical Center, Seattle, Washington; LDS Hospital, Salt Lake City, Utah; Utah Valley Regional Medical Center, Provo, Utah; and McKay Dee Hospital, Ogden, Utah) were eligible for inclusion in the present study. Women who did not consent and those whose fetuses were affected by Down syndrome were excluded.
Inclusion criteria also required documentation that thyroid-related measurements were available in both the first trimester (11–13 weeks) and second trimester (15–18 weeks), that gestational ages were established by ultrasonography, and that pregnancies were viable at the time when the second serum sample was obtained in the second trimester. From the 10,329 women who met those inclusion criteria, 267 were excluded because the question about whether the woman had hypothyroidism was not answered; 10,062 women remained for analysis, 389 of whom reported having an existing diagnosis of hypothyroidism. These self-reports were not individually confirmed, but the rate of antibody positivity and distribution of thyroid-stimulating hormone measurements for the group as a whole offered indirect evidence that this classification was generally reliable.8 Table 1 shows demographic characteristics of the study population according to enrollment site.
Postdelivery follow-up was performed by medical record review by the research coordinator at each site or by telephone interview with the patient. A single perinatologist and a pediatric geneticist reviewed detailed maternal and pediatric medical records for the following patient subsets: abnormal first-trimester screening, second-trimester screening, or both; adverse obstetric or pediatric outcomes, and 10% of healthy participants randomly selected at each site from the trial database. Further details of collection of follow-up data are published elsewhere.9
Samples were stored at −80°C and tested for thyroid antibodies between July 2004 and May 2005 (storage was for 3–6 years). Levels of thyroperoxidase and thyroglobulin antibodies were measured using the Immulite 2000 methodology (Siemens Medical Solutions Diagnostics, Tarrytown, NY). Normative data involving these analytes have been published separately for this cohort along with details of assay performance.8 Women's antibody measurements were considered positive if thyroperoxidase antibody concentrations were greater than 35 IU/mL or if thyroglobulin antibody concentrations were greater than 40 IU/mL (according to package inserts). Samples were thawed overnight before assay, and first- and second-trimester samples from each woman were assayed within 24 hours of each other.
Odds ratios and tests of significance for 2×2 tables were computed using Epistat (Epistat Services, Richardson, TX). Significance was two-tailed at the .05 level. All other data analyses were performed using SAS 9.1 (SAS Institute, Inc., Cary, NC). Stepwise logistic regression was used to adjust the observed ORs for possible covariates of the outcomes of interest (eg, preterm delivery). The baseline model included maternal age, body mass index, gravidity (primipara, multipara), and race (white, nonwhite).9 Stepwise regression was then used to identify the following four variables that also entered one (or more) of the models: smoking (yes, no), gestational diabetes, educational level (less than 12, 12, more than 12 years), and recruitment site. The final fixed models used for adjusting the ORs consisted of all eight variables.
Table 2 shows demographic characteristics of study participants according to thyroid antibody status and compares the rates of prematurity, low birth weight, and preterm premature rupture of membranes (PROM) among women without elevated thyroid antibody levels in the first trimester (neither thyroperoxidase nor thyroglobulin antibody elevated) with rates for antibody-positive women (elevated levels of thyroperoxidase, thyroglobulin antibody, or both). Among women with elevated levels of thyroperoxidase, thyroglobulin antibodies, or both in the first trimester, there is a 1.1% higher rate of births less than 37 weeks of gestation (6.4–7.5%; OR 1.18). Births less than 32 weeks of gestation increase by 0.5% (0.7–1.2%; OR 1.70). These associations are stronger in the first trimester than in the second (data not shown). The frequency of low birth weight deliveries is also slightly higher among antibody-positive women.
Preterm premature rupture of membranes increases by 0.8% (1.2–2.0%; OR 1.60) in the first trimester and also occurs more often among women with elevated thyroid antibody levels in the second trimester (data not shown). The higher rate of preterm PROM among all antibody-positive women in Table 2 is largely accounted for by excess cases among women with elevated levels of both antibodies, whereas the higher rates of preterm births at both less than 37 and less than 32 weeks of gestation are present in both categories of women with elevated antibody levels. Among three other late pregnancy complications (placenta previa, abruptio placentae, and preeclampsia), which might contribute to preterm delivery, none occur significantly more often in association with elevated levels of thyroid antibodies in the first trimester (data not shown).
Table 3 displays the rates of preterm birth, low birth weight, and preterm PROM according to first-trimester thyroid antibody status in women with and without a clinical diagnosis of hypothyroidism. Among the hypothyroid women, 57% are antibody-positive as opposed to 12.9% of women without a diagnosis of hypothyroidism. Antibody-related associations for the four variables in these subgroups of women are generally similar to the study population as a whole.
An earlier report of selected pregnancy outcomes in women enrolled in the FaSTER trial focused primarily on thyroid-stimulating hormone and free T4.9 The association of thyroid antibodies with preterm PROM was also briefly described. This report further explores this association among women with elevated levels of thyroperoxidase, thyroglobulin antibodies, or both in the first trimester. There is a 0.8% absolute increase in the rate of preterm PROM (a 66% relative increase), a 1.1% absolute increase in births less than 37 weeks of gestation, and a 0.5% absolute increase in births less than 32 weeks of gestation. This translates into relative increases of 17% and 71%, respectively, over the rates found among antibody-negative women. Preterm premature rupture of membranes likely contributes to these excess preterm births. Antibody-related rates are similar for women being treated for clinically diagnosed hypothyroidism as compared with women without hypothyroidism. Detailed descriptions of thyroid-stimulating hormone, free T4, and thyroid antibody measurements have been previously reported separately for these two groups.7,8
Directionally, our data on preterm births are consistent with published observations by others (Table 4) but are not consistent with the large increases in preterm births found in some of the studies. In 1994, Glinoer et al2 carried out a prospective observational study in which 87 euthyroid women with either thyroid peroxidase or thyroglobulin antibodies were enrolled during the first trimester and followed until delivery. The rate of prematurity among these women was 16.1% as opposed to 7.9% in control women. In Pakistan, Ghafoor et al3 reported a 26.8% preterm birth rate among 168 women with thyroperoxidase antibodies as opposed to 8.0% among 1,332 antibody-negative women. Negro et al4 carried out a randomized trial in Italy involving 115 euthyroid women with thyroperoxidase antibodies; 57 were given thyroid hormone and 58 were not treated; the control group was 869 antibody-negative women. Among the 58 untreated women, the rate of prematurity was 22.4% as opposed to 8.2% for control participants; this difference was highly significant.
In contrast to these large increases in prematurity rates, a cohort study from Japan reported a slightly increased rate of prematurity among women with antimicrosomal antibodies (4.0% as opposed to 3.0%) that was not statistically significant.5 Average birth weight, however, was significantly lower among infants of the 125 women with antithyroid microsomal antibodies. Similar findings were reported from Finland, where a nonsignificant increase in the rate of prematurity was observed among women with elevated thyroperoxidase antibodies (5.9% as opposed to 4.3% among antibody-negative women).6
The point estimate for increased prematurity among antibody-positive women in the present study is even smaller in magnitude than those reported in the studies from Japan and Finland. Those studies had the lowest prematurity rates among antibody-negative women, and the higher prematurity rates among antibody-positive women were not statistically significant. The higher rate of preterm PROM in our study suggests the possibility of antibody-induced inflammation as a precipitating event for the slightly higher rate of preterm delivery. Against this hypothesis is documentation from a randomized trial by Negro et al4 that thyroid hormone replacement in euthyroid antibody-positive women results in a preterm birth rate that is similar to control participants.
If effective treatment were discovered for women with thyroid antibody elevations in the first trimester, it would need to be highly effective, safe, and inexpensive, because a large number of antibody-positive women would require treatment for every preterm birth avoided. At present, there is no obvious candidate therapy. A case might be made, however, for further investigation of the association among thyroid antibodies, preterm PROM, and preterm delivery (especially, deliveries less than 32 weeks) with a view to gaining insight into a possible mechanism.
1.Stagnaro-Green A. Maternal thyroid disease and preterm delivery. J Clin Endocrinol Metab 2009;94:21–5.
2.Glinoer D, Riahi M, Grun JP, Kinthaert J. Risk of subclinical hypothyroidism in pregnant women with asymptomatic autoimmune thyroid disorders. J Clin Endocrinol Metab 1994;79:197–204.
3.Ghafoor F, Mansoor M, Malik T, Malik MS, Khan AU, Edwards R, et al. Role of thyroid peroxidase antibodies in the outcome of pregnancy. J Coll Physicians Surg Pak 2006;16:468–71.
4.Negro R, Formoso G, Mangieri T, Pezzarossa A, Dazzi D, Hassan H. Levothyroxine treatment in euthyroid pregnant women with autoimmune thyroid disease: effects on obstetrical complications. J Clin Endocrinol Metab 2006;91:2587–91.
5.Iijima T, Tada H, Hidaka Y, Mitsuda N, Murata Y, Amino N. Effects of autoantibodies on the course of pregnancy and fetal growth. Obstet Gynecol 1997;90:364–9.
6.Männistö T, Vääräsmäki M, Pouta A, Hartikainen AL, Ruokonen A, Surcel HM, et al. Perinatal outcome of children born to mothers with thyroid dysfunction or antibodies: a prospective population-based cohort study. J Clin Endocrinol Metab 2009;94:772–9.
7.Lambert-Messerlian G, McClain M, Haddow JE, Palomaki GE, Canick JA, Cleary-Goldman J, et al. 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 2008;199:62.e1–6.
8.McClain MR, Lambert-Messerlian G, Haddow JE, Palomaki GE, Canick JA, Cleary-Goldman J, et al. Sequential first- and second-trimester TSH, free thyroxine, and thyroid antibody measurements in women with known hypothyroidism: a FaSTER trial study. Am J Obstet Gynecol 2008;199:129.e1–6.
9.Cleary-Goldman J, Malone FD, Lambert-Messerlian G, Sullivan L, Canick J, Porter TF, et al. Maternal thyroid hypofunction and pregnancy outcome. Obstet Gynecol 2008;112:85–92.
10.Malone FD, Canick JA, Ball RH, Nyberg DA, Comstock CH, Bukowski R, et al. First-trimester or second-trimester screening, or both, for Down's syndrome. N Engl J Med 2005;353:2001–11.