Overtly abnormal amounts of thyroid hormones—either excessive because of Graves' thyrotoxicosis or deficient because of hypothyroidism—can have profound effects on glucose metabolism and insulin secretion. The resulting increased insulin resistance, glucose intolerance, and dyslipidemia are usually reversible when normal thyroid hormone levels are restored.1–3 Both subclinical hyperthyroidism and hypothyroidism are reported to have similar but correspondingly less marked metabolic effects in nonpregnant women.4,5 In pregnant women, however, virtually nothing is known regarding possible metabolic changes with either type of subclinical thyroid dysfunction.
It is well known that marked insulin resistance progressively increases with gestational age during pregnancy.6 Because of the similar effects of thyroid hormone and pregnancy on glucose metabolism, it seems reasonable to posit that women with either abnormally high or low levels of thyroid hormones would be more likely to develop gestational diabetes. Untreated overt hyperthyroidism or hypothyroidism in pregnant women is uncommon, and none of the reports of such pregnancies described any effects on glucose metabolism.7–10 Conversely, because subclinical thyroid dysfunction in pregnancy is relatively common and usually not treated, there recently has been a number of population-based studies of subclinical thyroid dysfunction during pregnancy. Two of the largest studies reported conflicting results concerning any associations with gestational diabetes.11,12 Because of this, we designed the currently reported secondary analysis of a larger cohort of women who presented for prenatal care at all gestational ages to further estimate any effects that subclinical thyroid dysfunction may have on the incidence of gestational diabetes.
MATERIALS AND METHODS
Excess serum obtained for prenatal serological screening, from November 1, 2000, to April 14, 2003, was delivered each weekday to our immunochemistry research laboratory for thyroid studies. The protocol for this study has been approved by the University of Texas Southwestern Medical Center institutional review board committee and found to be minimal risk research. Using a chemiluminescent assay, serum concentrations were determined for thyrotropin and serum free thyroxine (T4). Women identified to have abnormal thyrotropin levels had a serum free T4 reflexively determined, and, if it was either abnormally high or low, they were contacted and referred to an obstetrical complications clinic for further evaluation and possible treatment. These women were excluded from further study. Subsequent to this prospective evaluation, serum samples in women who presented for prenatal care in the first half of pregnancy were retrieved and serum free T4 levels were measured. The details of this process and assays have been previously described.11,13,14
For the purposes of the current study, normal values were considered to be those that comprised the range from the 2.5th to 97.5th percentiles, not corrected for gestational age, for the study population. For thyrotropin, these were 0.03–4.13 milliunits/L and for serum free T4, these ranged from 0.9 to 2.0 ng/dL. Using these thresholds, all women with abnormal thyrotropin values had reflexive serum free T4 values determined regardless of gestational age. Women screened and delivered of singleton neonates weighing 500 g or more at Parkland Hospital and without evidence of overt thyroid dysfunction were included in this analysis. Women with serum thyrotropin and serum free T4 values within the normal range were considered to be euthyroid. Those with an abnormally low thyrotropin but normal serum free T4 level were classified as having subclinical hyperthyroidism. Conversely, women with an abnormally high thyrotropin but a normal serum free T4 level were classified as having subclinical hypothyroidism.
Pregnancy outcomes were retrieved from a computerized perinatal database that has been previously described.11,13,14 These data are routinely entered for all women at the time of their delivery at Parkland Hospital. For this study, the incidence of gestational diabetes was compared for the three cohorts described. Gestational diabetes was defined as diabetes diagnosed during pregnancy regardless of the gestational age at diagnosis. This included both women with gestational diabetes treated with diet and those with gestational diabetes treated with insulin. The diagnosis of gestational diabetes was made using a 1-hour 50-g screening test and a 100-g 3-hour tolerance test. If two or more plasma levels using the National Diabetes Data Group criteria, and accepted by the American Congress of Obstetricians and Gynecologists, were abnormal, then the diagnosis of gestational diabetes was made.15
Pearson χ2, Armitage test for trend, analysis of variance with Dunnett's test for multiple comparisons to a control group, and Student's t test were used for group comparisons. Logistic regression was applied to examine the significance for selected pregnancy outcomes adjusted for maternal age, race, parity, and weight. The Hosmer-Lemeshow statistic was used to examine the goodness of fit for the logistic regression model. To further investigate the relationship between thyrotropin and the likelihood of gestational diabetes, we incorporated a logistic regression model including thyrotropin level as a continuous variable. We used the logarithm of thyrotropin as an independent predictor of gestational diabetes and adjusted for maternal age and weight as continuous measures through a cubic smoothing splines with knots at 5th, 25th, 50th, 75th, and 95th percentiles as well as categorical variables of race and parity. Statistical computations were performed with SAS 9. Two-sided P<.05 were judged statistically significant.
During the 30-month study, a total of 24,883 women were delivered of a singleton fetus weighing more than 500 g and are included in this study. Of these, 23,771 (95.5%) had thyrotropin values within the normal range and were considered euthyroid; 528 (2.1%) had a thyrotropin level greater than 4.13 milliunits/L along with a normal serum free T4 level and therefore met criteria for subclinical hypothyroidism; and the remaining 584 (2.3%) had a thyrotropin level less than 0.03 milliunits/L along with a normal serum free T4 level meeting criteria for subclinical hyperthyroidism.
Maternal demographics of the three cohorts are shown in Table 1. Women with either subclinical hyperthyroidism or hypothyroidism are compared with euthyroid women. Those with subclinical hyperthyroidism were more likely to be African American or “other” ethnicity. Also, on average, women with subclinical hyperthyroidism weighed less than women in the other two cohorts. Conversely, women with subclinical hypothyroidism were less likely to be African American and, as a group, were significantly older than the other study groups.
As shown in Figure 1, the incidence of gestational diabetes is compared between the three study groups. In general, as screening thyrotropin level increased, the likelihood of diabetes complicating pregnancy also increased (P=.002). For example, 2.2% of women identified with subclinical hyperthyroidism had gestational diabetes as compared with 4.2% of euthyroid women and 6.3% of women with subclinical hypothyroidism. As shown in Figure 2, after adjustment for maternal age, weight, race, parity, and gestational age at screening, the risk for gestational diabetes in women with subclinical hyperthyroidism remained significantly lower (P=.022). However, the increased risk of gestational diabetes in women with subclinical hypothyroidism was no longer significant (P=.056).
To further investigate the relationship between thyrotropin level and the likelihood of gestational diabetes, we incorporated a logistic regression model including thyrotropin level as a continuous measure. After adjusting for maternal factors of age, weight, race, and parity, thyrotropin was a significant predictor of gestational diabetes (P=.001). There was no indication of a lack of fit for this model as indicated by the Hosmer-Lemeshow test (P=.31). To illustrate this effect, we provide Figure 3 with selected covariable levels of a 25-year-old nulliparous, Hispanic woman weighing 175 pounds. For this example, the probability of gestational diabetes complicating pregnancy increased from 1.9% if the thyrotropin was 0.001 milliunits/L to 4.9% if the thyrotropin was 10 milliunits/L in such women (P=.001). Because this is the result of the logistic regression model, this is representative of other levels of the covariates but with the likelihood of gestational diabetes increasing with thyrotropin.
This report is another in a series in which we further describe the relationships between a number of thyroid-associated analytes and pregnancies in nearly 25,000 women.11,13,14,16–18 This secondary analysis was designed to elucidate previously reported associations between subclinical thyroid dysfunction in these women and their risks for developing gestational diabetes during the same pregnancy. Our findings now presented indicate that there are two reasons to support such a relationship. The first and strongest of these is the direct correlation between increasing serum thyrotropin concentrations and the probability of gestational diabetes as depicted in Figure 3. Importantly, this association remained significant after adjustment for maternal age, race, and weight, factors known to have profound effects on both the incidence of subclinical hypothyroidism as well as gestational diabetes. The second supportive finding is that women with subclinical hypothyroidism were at significantly greater risk for gestational diabetes when compared with the cohort of euthyroid women (Fig. 1). Unlike the first observation, however, the significance of this finding was now of borderline significance after we adjusted for the same risk confounders (P=.056).
There are previous reports that reinforce a positive relationship between subclinical hypothyroidism and gestational diabetes. The biological plausibility of our findings finds support from studies in nonpregnant adults that describe the metabolic actions of T4 and insulin. Some of these observations are that both overt and subclinical hypothyroidism cause significantly increased insulin resistance.4,19–21 In fact, this was shown to induce abnormal glucose tolerance test values remarkably similar to those characteristic of normal pregnancy and also as a result of insulin resistance.4,6 When taken together, it seems reasonable to speculate that pregnant women with subclinical hypothyroidism have further amplified insulin resistance and thus an increased risk for gestational diabetes.
There are other findings that argue against the likelihood that subclinical hypothyroidism is related to an increased risk for gestational diabetes. First, as discussed previously, the relationship between subclinical hypothyroidism and gestational diabetes was only of borderline significance after adjustment for age, weight, race, parity, and gestational age (P=.056). A second argument is that such an association was not found by Cleary-Goldman et al12 in a similar study of thyroid analytes in more than 10,000 pregnant women One explanation for this is that one-third of women in our cohort with subclinical hypothyroidism were found to have abnormally elevated antithyroid antibodies, more than twice that of 15% in the Cleary-Goldman cohort.12,16 This observation is important in light of findings of Mannisto et al22 who reported that pregnant women with thyroid disease and abnormally high thyroid antibody levels were at inordinately high risk for later developing noninsulin-dependent diabetes mellitus. Also possibly related to the ethnic differences, Hispanic women are at much higher risk for gestational diabetes when compared with white women.23
Another interesting and unexpected finding was that women with subclinical hyperthyroidism had a significantly lower risk for developing gestational diabetes when compared with euthyroid women. This observation has not been reported in the previously cited population studies.12,14,22 This observation is counterintuitive because insulin resistance is a feature of hyperthyroidism, both overt and subclinical.5,24 Moreover, diabetes has a consistently positive association in patients with overt hyperthyroidism25–27 At this time, we can offer no biological explanation for the significant association between this possible “protective” effect of subclinical hyperthyroidism against gestational diabetes and await any confirmatory studies.
In summary, findings now reported include a consistent significant relationship between increasing serum thyrotropin levels, in women, with normal serum free T4 concentrations and a correspondingly increased risk for gestational diabetes in the same pregnancy. These findings suggest, albeit less strongly, that women with subclinical hypothyroidism are at higher risk for gestational diabetes when compared with euthyroid women. Taken together with reports that overt or subclinical hypothyroidism results in increased peripheral insulin resistance, it is biologically plausible that these effects magnify normal pregnancy-induced insulin resistance.4,6 At this time, these findings should not be considered an indication for routine prenatal serum thyroid-analyte screening nor for targeted screening in women identified to have gestational diabetes.
1. Handisurya A, Pacini G, Tura A, Gessl A, Kautzky-Willer A. Effects of T4 replacement therapy on glucose metabolism in subjects with subclinical (SH) and overt hypothyroidism (OH). Clin Endocrinol (Oxf) 2008;69:963–9.
2. Tene C, Zarate A, Basurto L, Islas S, Revilla C, Ochoa R, Galvan R, Santos P. Correction of insulin resistance in methimazole-treated patients with Graves disease. Rev Invest Clin 2001;53:531–5.
3. Al-Shoumer KA, Vasanthy BA, Al-Zaid MM. Effects of treatment of hyperthyroidism on glucose homeostasis, insulin secretion, and markers of bone turnover. Endocr Pract 2006;12:121–30.
4. Maratou E, Hadjidakis DJ, Kollias A, Tsegka K, Peppa M, Alevizaki M, et al.. Studies of insulin resistance in patients with clinical and subclinical hypothyroidism. Eur J Endocrinol 2009;160:785–90.
5. Maratou E, Hadjidakis DJ, Peppa M, Alevizaki M, Tsegka K, Lambadiari V, et al.. Studies of insulin resistance in patients with clinical and subclinical hyperthyroidism. Eur J Endocrinol 2010;163:625–30.
6. Lind T, Billewicz WZ, Brown G. A serial study of changes occurring in the oral glucose tolerance test during pregnancy. J Obstet Gynaecol Br Commonw 1973;80:1033–9.
7. Leung AS, Millar LK, Koonings PP, Montoro M, Mestman JH. Perinatal outcome in hypothyroid pregnancies. Obstet Gynecol 1993;81:349–53.
8. Davis LE, Leveno KJ, Cunningham FG. Hypothyroidism complicating pregnancy. Obstet Gynecol 1988;72:108–12.
9. Davis LE, Lucas MJ, Hankins GD, Roark ML, Cunningham FG. Thyrotoxicosis complicating pregnancy. Am J Obstet Gynecol 1989;160:63–70.
10. Haddow JE, Palomaki GE, Allan WC, Williams JR, Knight GJ, Gagnon J, et al.. Maternal thyroid deficiency during pregnancy and subsequent neuropsychological development of the child. N Engl J Med 1999;341:549–55.
11. Casey BM, Dashe JS, Spong CY, McIntire DD, Leveno KJ, Cunningham FG. Perinatal significance of isolated maternal hypothyroxinemia identified in the first half of pregnancy. Obstet Gynecol 2007;109:1129–35.
12. Cleary-Goldman J, Malone FD, Lambert-Messerlian G, Sullivan L, Canick J, Porter F, et al.. Maternal thyroid hypofunction and pregnancy outcome. Obstet Gynecol 2008;112:85–92.
13. Casey BM, Dashe JS, Wells CE, McIntire DD, Byrd W, Leveno KJ, et al.. Subclinical hypothyroidism and pregnancy outcomes. Obstet Gynecol 2005;105:239–45.
14. Casey BM, Dashe JS, Wells CE, McIntire DD, Leveno KJ, Cunningham FG. Subclinical hyperthyroidism and pregnancy outcomes. Obstet Gynecol 2006;107:337–41.
15. Screening and diagnosis of gestational diabetes mellitus. ACOG Committee Opinion No. 504. American College of Obstetricians and Gynecologists. Obstet Gynecol 2011;118:751–3.
16. Abassi-Ghanavati M, Casey BM, Spong CY, McIntire DD, Halvorson LM, Cunningham FG. Pregnancy outcomes in women with thyroid peroxidase antibodies. Obstet Gynecol 2010;116:381–6.
17. Greer LG, Casey BM, Halvorson LM, Spong CY, McIntire DD, Cunningham FG. Antithyroid antibodies and parity: further evidence for microchimerism in autoimmune thyroid disease. Am J Obstet Gynecol 2011;205 471.e1–4.
18. Wilson KL, Casey BM, McIntire DD, Halvorson LM, Cunningham FG. Subclinical thyroid disease and the incidence of hypertension in pregnancy. Obstet Gynecol 2012;119:315–20.
19. Garduño-Garcia Jde J, Alvirde-Garcia U, López-Carrasco G, Padilla Mendoza ME, Mehta R, Arellano-Campos O, et al.. TSH and free thyroxine concentrations are associated with differing metabolic markers in euthyroid subjects. Eur J Endocrinol 2010;163:273–8.
20. Roos A, Bakker SJL, Links TP, Gans ROB, Wolffenbuttel BHR. Thyroid function is associated with components of the metabolic syndrome in euthyroid subjects. J Clin Endocrinol Metab 2007;92:491–6.
21. Dimitriadis G, Mitrou P, Lambadiari V, Boutati E, Maratou E, Panagiotakos DB, et al.. Insulin action in adipose tissue and muscle in hypothyroidism. J Clin Endocrinol Metab 2006;91:4930–7.
22. Männist ö T, Vääräsmäki M, Pouta A, Hartikainen AL, Roukonen A, Surcel HM, et al.. Thyroid dysfunction and autoantibodies during pregnancy as predictive factors of pregnancy complications and maternal morbidity in later life. J Clin Endocrinol Metab 2010;95:1084–94.
23. Clinical management guidelines for obstetrician-gynecologists. September 2001 (replaces Technical Bulletin No. 200, December 1994). Gestational diabetes. ACOG Practice Bulletin No. 30. American College of Obstetricians and Gynecologists. Obstet Gynecol 2001;98:525–38.
24. Rezzonico J, Niepomniszcze H, Rezzonico M, Pusiol E, Alberto M, Brenta G. The association of insulin resistance with subclinical thyrotoxicosis. Thyroid 2011;21:945–9.
25. Kadiyala R, Peter R, Okosieme OE. Thyroid dysfunction in patient with diabetes: clinical implications and screening strategies. Int J Clin Pract 2010;64:1130–9.
26. Potenza M, Via MA, Yanagisawa RT. Excess thyroid hormone and carbohydrate metabolism. Endocr Pract 2009;15:254–62.
© 2012 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.
27. Pisarev MA. Interrelationships between the pancreas and the thyroid. Curr Opin Endocrinol Diabetes Obes 2010;17:437–9.