ORIGINAL ARTICLE: ENVIRONMENT AND EPIDEMIOLOGY
Thyroid antibodies and gestational diabetes mellitus: a meta-analysis Ying Yang, Ph.D.,a Qian Li, M.D.,a,b Qianqian Wang, Ph.D.,c and Xu Ma, M.D.a,b a National Research Institute for Family Planning; b Graduate School of Peking Union Medical College; and c Department of Molecular Orthopaedics, Center of Clinical Research and Evidence-Based Medicine, Beijing, People’s Republic of China
Objective: To determine whether thyroid antibodies are associated with an increased risk of gestational diabetes mellitus (GDM) in pregnant women. Design: Meta-analysis. Setting: Not applicable. Patient(s): Twenty cohort and case-control studies involving pregnant women with positive thyroid antibodies were the exposure of interest, and GDM was the outcome. Intervention(s): None. Main Outcome Measure(s): A fixed-effects model was used to evaluate the relationship between thyroid antibodies and the risk of GDM. Subgroup analyses were performed among different types of study design, different thyroid antibodies, and patients with specific thyroid dysfunction. Result(s): The search strategy identified 167 potentially relevant publications, of which 20 were included in the meta-analysis. A significant association between thyroid antibodies and GDM was observed. A meta-analysis of the 11 cohort studies with pregnant women with positive thyroid antibodies in their first trimester suggested no obvious risk of GDM compared with the reference group. In subgroup meta-analyses, no significant association between thyroid antibodies and GDM was found in euthyroid pregnant women, whereas a significant positive association was identified in women with a thyroid dysfunction. Conclusion(s): Based on the currently available evidence, there is a significant but not strong association between thyroid antibodies and the risk of GDM, and thyroid antibodies in the first trimester lack predictive value for the risk of GDM. In addition, thyroid antibodies may not increase the risk of GDM in euthyroid pregUse your smartphone nant women. (Fertil SterilÒ 2015;-:-–-. Ó2015 by American Society for Reproductive to scan this QR code Medicine.) and connect to the Key Words: Thyroid, autoimmunity, antibodies, gestational diabetes mellitus Discuss: You can discuss this article with its authors and with other ASRM members at http:// fertstertforum.com/yangy-thyroid-antibodies-gdm/
G
estational diabetes mellitus (GDM) is defined by the World Health Organization as having ‘‘any degree of glucose intolerance with onset or first recognition during pregnancy’’ (1). The prevalence of GDM in pregnant women varies substantially, ranging from 1.7% to 11.6% in advanced economies (2). In Asian countries, the prevalence varies
according to the screening strategy and diagnostic criteria and ranges from 1% to 20%, with an increasing trend in recent years (3). GDM is one of the most common pregnancy complications associated with a number of adverse outcomes, including miscarriage, delivery, excessive fetal growth, birth trauma, and neonatal metabolic abnormalities (4). In addition, women
Received January 22, 2015; revised and accepted June 4, 2015. Y.Y. has nothing to disclose. Q.L. has nothing to disclose. Q.W. has nothing to disclose. X.M. has nothing to disclose. Y.Y. and Q.L. should be considered similar in author order. Supported by Operation Expenses for Basic Scientific Research of Central Authorities (grant no. 2012GJSSJKC03), International Science and Technology Cooperation Program of China (grant no. 2012DFB30130), and National Natural Science Foundation (grant no. 81402757). Reprint requests: Xu Ma, M.D., Professor, National Research Institute for Family Planning, 12 Dahuisi Road, Haidian District, Beijing 100081, People's Republic of China (E-mail: maxutougao@163. com). Fertility and Sterility® Vol. -, No. -, - 2015 0015-0282/$36.00 Copyright ©2015 American Society for Reproductive Medicine, Published by Elsevier Inc. http://dx.doi.org/10.1016/j.fertnstert.2015.06.003 VOL. - NO. - / - 2015
discussion forum for this article now.*
* Download a free QR code scanner by searching for “QR scanner” in your smartphone’s app store or app marketplace.
who have been exposed to GDM during pregnancy are at an exceptionally high risk of developing type 2 diabetes, chronic hypertension, and vascular disease later in life (5, 6). GDM has undoubtedly become an important public health issue worldwide. Studies have shown that GDM is caused by b-cell dysfunction and insulin resistance (7). A high prevalence of elevated thyroid antibodies have appeared in patients with a history of insulin resistance, such as pregnant women with GDM (8). The link between autoimmune disorders and insulin resistance could be the inflammatory events that are associated with both conditions. Autoimmune thyroid dysfunction was found to be associated with insulin resistance, suggesting a potential relationship between 1
ORIGINAL ARTICLE: ENVIRONMENT AND EPIDEMIOLOGY autoimmune thyroid dysfunction and GDM. The prevalence of thyroid antibodies in pregnant women is approximately 10%–15% (8). In recent years, many epidemiologic studies have investigated the links between thyroid antibodies, including thyroid peroxidase antibody (TPOAb) and thyroglobulin antibody (TgAb), and risk of GDM (9–28). However, these studies have apparently yielded conflicting results. Both obstetricians and endocrinologists pay great attention to whether euthyroid pregnant women with positive thyroid antibodies have a higher risk than women with negative thyroid antibodies and how their condition should be managed. Therefore, to clarify the evidence, we conducted a meta-analysis of available case-control and cohort studies of thyroid antibodies and GDM.
MATERIALS AND METHODS Search Strategy and Study Selection We followed the Meta-analysis of Observational Studies in Epidemiology guidelines to explicitly report this metaanalysis (29). To study the association between thyroid antibodies (TPOAb or TgAb) and GDM, an electronic search was conducted with the use of Pubmed and two Chinese databases, Wanfang and China National Knowledge Internet (CNKI), from inception to October 2014. The following search words were used: (1) thyroid antibody, thyroid antibodies, thyroid autoimmune antibody, thyroid autoimmune antibodies, thyroid autoantibody, thyroid autoantibody, TPOAb, TPOAb, thyroid peroxidase antibody, thyroid peroxidase antibodies, thyroid globulin antibody, thyroid globulin antibodies, TG-Ab, or TgAb; and (2) gestational diabetes, pregnancy diabetes, maternal diabetes, gestational diabetes mellitus, or GDM. An initial screening was based on the titles and abstracts of published papers, and the full texts were then reviewed carefully during the second screening. In addition, reference lists of the original studies and review articles were considered. Irrelevant studies, reviews, case reports, and letters to the editor were ineligible. Overlapping studies were also excluded. All of the literature searches were reviewed independently by two authors.
Data Extraction and Assessment of Quality of Methods Relevant data were extracted from each paper with the use of standardized data collection forms. Any disagreements were resolved by discussion until an agreement could be found. We contacted the corresponding authors through email to obtain additional detailed information when necessary. The related data were extracted to calculate an odds ratio (OR) or relative risk (RR) for the association between thyroid antibodies and GDM. We also extracted study characteristics for further exploration in subgroup meta-analyses. The following characteristics were recorded: first author, year of publication, study period, study design, definition of GDM, definition of thyroid dysfunction, and exclusion criteria (Supplemental Table 1, available online at www.fertstert.org). The selected studies were assessed for quality with the use of the Newcastle-Ottawa scale (NOS) for cohort studies and 2
case-control studies (http://www.ohri.ca/Programs/clinical_ epidemiology/default.asp). Studies with NOS scores of 0–3, 4–6, and 7–9 were considered to be of low, moderate, and high quality, respectively. Two investigators extracted the data and independently assessed the quality. Disagreements were resolved by consensus or arbitration by a third investigator.
Statistical Analyses The RR was used as the common measure of association across the studies. ORs were transformed into RRs according to the following formula: RR ¼ OR/[(1 P0) þ (P0 OR)], where P0 is the incidence of GDM in the thyroid antibodies–negative group (30). In addition, the Miettinen test–based approach was used to calculate the variance of logRR (variance logRR ¼ variance logOR [logRR/logOR]) (31). The overall combined RR and 95% confidence interval (CI) was calculated with the use of logRR and variance of logRR. Heterogeneity across studies was assessed with the use of the Cochrane Q statistic (significance level at P< .10) and the I2 statistic. If I2 < 50%, a random-effects model (DerSimonian–Laird method) was used to pool the results; otherwise, a fixed-effects model (Mantel–Haenszel method) was used. In addition, subgroup analyses were conducted stratifying by the type of study design, type of thyroid antibody (TPOAb or TgAb), and thyroid dysfunction exclusions. Sensitivity analyses were performed to detect the effects of individual study on the pooled result by omitting one study in each turn. Publication bias was assessed with the use of funnel plots and the Egger regression test. We conducted all analyses with the use of Stata software, version 12.0. Except where otherwise specified, a P value of < .05 was considered to be statistically significant.
RESULTS Study Selection and Study Characteristics The study selection process is shown in Figure 1. Initially, 167 unique citations were retrieved from the Pubmed, CNKI, and Wanfang databases. After the first screening of the abstracts and titles, 130 citations were excluded, primarily owing to being reviews, clearly irrelevant studies, or studies not focusing on GDM. Then we reviewed full-text articles for a more comprehensive evaluation. Two articles were excluded because the exposure was not relevant, and 15 articles were excluded because they did not focus on GDM. Finally, 20 articles (9–28) were included in our meta-analysis. The meta-analysis included 10 cohort studies and 10 case-control studies conducted with 34,566 participants. The characteristics of the 20 articles are presented in Supplemental Table 1. The cohort sizes of the studies ranged from 49 to 7,641 and were published from 1997 to 2014. Among the studies, 11 (9–14, 19, 24–27), seven (15–17, 21– 23, 28), and two (18, 20) were conducted in Asia, Europe, and America, respectively. The ascertainment of thyroid antibodies varied across studies. Among the 20 articles included here, nine studies evaluated TPOAb only (9–11, 20, 21, 23–25, 27), ten studies evaluated both TgAb and TPOAb (one type of positive antibody was considered to be a VOL. - NO. - / - 2015
Fertility and Sterility® Synthesis Analysis
FIGURE 1 Potentially relevant studies through literature search (n=167)
identified
130 articles excluded Obvious irrelevant studies(n=89) Reviews(n=23) Not on gestational diabetes mellitus(n=18) Full-text articles reviewed for more detailed evaluation (n=37)
To study the relationship between thyroid antibodies and GDM, all eligible studies were first summarized to calculate a combined RR. Then meta-analyses of the 11 cohort studies were conducted to examine the relationship between positive thyroid antibodies in early pregnancy and the risk of GDM. Summaries of the pooled outcomes in the synthesis metaanalyses using all of the studies and cohort studies are presented in Table 1.
Thyroid Antibodies and GDM
16 articles excluded Exposure not relevant (n=2) Not on gestational diabetes mellitus (n=15)
Articles included in meta-analysis (n=20) Cohort study(n=10) Case-control study(n=10)
Flow chart of study selection. Yang. Thyroid antibodies and GDM. Fertil Steril 2015.
positive thyroid antibody) (12–15, 17–19, 22, 26, 28), and one article (16) reported data on TPOAb and TgAb separately and was extracted as two individual studies. In addition, the definition of GDM, definition of thyroid dysfunction, and exclusion criteria did not conform across the studies (Supplemental Table 1). All 21 studies were of a moderate or high quality on the Newcastle-Ottawa scale.
We obtained 21 individual data sets from 20 articles. The synthesis meta-analysis revealed a significant association between thyroid antibodies and GDM, with a pooled RR of 1.12 (95% CI 1.03–1.22; Table 1; Fig. 2). The meta-analysis of the 11 cohort studies (9, 15, 16, 18, 19, 21, 24, 25, 27, 28) suggested that pregnant women with positive thyroid antibodies in the first trimester had no obvious risk of GDM compared with the reference group (pooled RR 1.07, 95% CI 0.96–1.19; Table 1; Fig. 2). For all of the aforementioned meta-analyses, fixed-effect models were used because no statistically significant heterogeneity was found. Sensitivity analyses suggested that the combined RRs were valid and credible.
TPOAb and GDM Fifteen studies (9–14, 16, 18–21, 23–25, 27) involving 23,718 pregnant women reported relative data on the association between TPOAb and GDM. The synthesis meta-analysis showed a significant association between TPOAb and GDM risk with a pooled RR of 1.11 (95% CI 1.01–1.21). The
TABLE 1 Summary of the meta-analyses of thyroid antibodies and gestational diabetes mellitus risk. Analysis items All of the studies Thyroid antibodies Total studies With euthyroid Including thyroid dysfunction Thyroid dysfunction TPOAb Total studies With euthyroid Including thyroid dysfunction TgAb Total studies Cohort studies Thyroid antibodies Total studies With euthyroid Including thyroid dysfunction TPOAb Total studies With euthyroid Including thyroid dysfunction TgAb Total studies
Studies (participants)
RR (95% CI)
I2 (P value)
Begg test
21 (34,502) 8 (20,533) 16 (18,401) 5 (561)
1.12 (1.03–1.22) 1.08 (0.97–1.22) 1.18 (1.06–1.31) 1.35 (1.06–1.71)
0.0% (.590) 5.3% (.390) 0.0% (.730) 0.0% (.539)
0.486 0.479 0.752 0.293
15 (23,718) 6 (15,878) 11 (11,210)
1.11 (1.01–1.21) 1.06 (0.94–1.20) 1.16 (1.04–1.30)
0.0% (.473) 1.7% (.405) 0.0% (.624)
0.704 0.192 0.834
6 (13,847)
1.34 (0.93–1.95)
0.0% (.974)
0.521
11 (31,546) 6 (18,947) 7 (15,545)
1.07 (0.96–1.19) 1.12 (0.98–1.28) 1.14 (0.97–1.34)
2.5% (.419) 4.4% (.388) 22.3% (.259)
0.548 0.916 0.988
8 (21,107) 4 (14,292) 5 (8,699)
1.06 (0.94–1.19) 1.08 (0.94–1.25) 1.08 (0.91–1.29)
0.0% (.542) 7.4% (.356) 0.0% (.420)
0.412 0.438 0.910
3 (13,402)
1.23 (0.78–1.95)
0.0% (.755)
0.212
Note: All of the above meta-analyses used fixed-effects models. CI ¼ confidence interval; RR ¼ relative risk; TgAb ¼ thyroglobulin antibody; TPOAb ¼ thyroid peroxidase antibody. Yang. Thyroid antibodies and GDM. Fertil Steril 2015.
VOL. - NO. - / - 2015
3
ORIGINAL ARTICLE: ENVIRONMENT AND EPIDEMIOLOGY
FIGURE 2
Forest plot showing the association between thyroid antibodies and gestational diabetes mellitus. CI ¼ confidence interval; RR ¼ relative risk. Yang. Thyroid antibodies and GDM. Fertil Steril 2015.
meta-analysis of the eight cohort studies (9, 16, 18, 19, 21, 24, 25, 27) suggested that compared with the reference group, TPOAb-positive status in the first trimester had no obvious impact on the risk of GDM in pregnant women (pooled RR 1.06, 95% CI 0.94–1.19; Table 1). No statistically significant heterogeneity was found, and fixed-effect models were used in the meta-analysis. Sensitivity analyses by sequential omission of individual studies did not materially alter the overall combined RR, suggesting that the combined RR was valid and credible.
TgAb and GDM Six studies (12–14, 16, 18, 19) involving 13,847 pregnant women reported relative data on the association between TgAb and GDM risk. The synthesis meta-analysis showed no significant association between TgAb and GDM, with a pooled RR of 1.34 (95% CI 0.93–1.95). The meta-analysis of the three cohort studies (16, 18, 19) demonstrated results similar to the association between TPOAb and GDM risk (pooled RR 1.23, 95% CI 0.78–1.95; Table 1). For all of these meta-analyses, no statistically significant heterogeneity was found, and fixed-effect models were used. Sensitivity analyses suggested that the combined RRs were valid and credible. 4
Subgroup Analysis Based on detailed clinical information provided by the articles included in this meta-analysis, three subgroup meta-analyses were conducted: excluding women with thyroid dysfunction, including women with thyroid dysfunction, and including only women with thyroid dysfunction. In studies that included euthyroid women (9, 10, 15, 19, 20, 24, 25, 28), the pooled RR of thyroid antibodies or TPOAb positive status on GDM incidence did not reach statistical significance (for thyroid antibodies, pooled RR 1.08, 95% CI 0.97–1.22; for TPOAb, pooled RR 1.06, 95% CI 0.94–1.20; Table 1; Fig. 3). Notably, significant associations were observed between thyroid antibodies or a positive TPOAb status and the risk of GDM in studies including women with thyroid dysfunction (for thyroid antibodies, pooled RR 1.18, 95% CI 1.06–1.31; for TPOAb, pooled RR 1.16, 95% CI 1.04–1.30; Table 1). Owing to the lack of relevant studies, a subgroup analysis of pregnant women with TgAb-positive status was not performed. The same grouping strategy was used to conduct metaanalyses of the cohort studies. The pooled RRs for the risk of GDM in pregnant women with thyroid antibodies or with TPOAb-positive status were not significant in euthyroid VOL. - NO. - / - 2015
Fertility and Sterility®
FIGURE 3
Forest plot showing the association between thyroid antibodies and the risk of gestational diabetes mellitus in euthyroid pregnancy. CI ¼ confidence interval; RR ¼ relative risk. Yang. Thyroid antibodies and GDM. Fertil Steril 2015.
pregnant women group (for thyroid antibodies, pooled RR 1.12, 95% CI 0.98–1.28; for TPOAb, pooled RR ¼ 1.08, 95% CI 0.94–1.25; Table 1; Fig. 3). No significant results were also observed in the association between thyroid antibodies or TPOAb-positive status and the risk of GDM in studies including women with thyroid dysfunction (for thyroid antibodies, pooled RR 1.14, 95% CI 0.97–1.34; for TPOAb, pooled RR 1.08, 95% CI 0.91–1.29; Table 1). A subgroup analysis of pregnant women with TgAb-positive status was absent owing to the lack of relevant studies. Furthermore, we obtained five data sets from four articles (10, 15, 18, 24) that involved 561 participates with thyroid dysfunction. Based on the data from those five studies, we conducted a meta-analysis to evaluate the associationbetween thyroid antibodies positive status and the risk of GDM in women with thyroid dysfunction. The pooled RR of the fixed-effects model showed a significant increase in the risk of GDM on women with positive thyroid antibodies (polled RR 1.35, 95% CI 1.06–1.71).
Publication Bias Publication biases were not found in the included studies, as suggested by the Begg tests and Begg funnel plots. All of the P values given by Begg tests are listed in Table 1. VOL. - NO. - / - 2015
DISCUSSION Studies have demonstrated that the underlying mechanism of GDM is likely to involve increased insulin resistance more than decreased insulin secretion (7). Thyroid autoimmunity could be an indication of autoimmune dysfunction. Montaner et al. (21) suggested that insulin resistance is an important mechanism of thyroid antibodies for the development of GDM. The presence of thyroid antibodies would result in an increase of proinflammatory cytokines, inducing insulin resistance. Nevertheless, the mechanisms involved in the association between thyroid antibodies and GDM risk remain uncertain. Epidemiologic studies have reported controversial results regarding the association between thyroid antibodies and the incidence of GDM. These controversies may result from different study designs and small sample sizes among the studies. Meta-analysis has been considered to be a powerful tool for pooling the results from different studies, which can overcome the problems of small sample size and inadequate statistical power. Therefore, we performed the present meta-analyses to study the association between thyroid antibodies and the risk of GDM. In the meta-analysis of all of the studies, including both case-control and cohort studies, a significant but not very strong association between thyroid antibodies and the risk of GDM was observed, with an overall combined RR of 1.12 (95% CI 1.03–1.22). However, the results of the 5
ORIGINAL ARTICLE: ENVIRONMENT AND EPIDEMIOLOGY meta-analysis of cohort studies only showed no significant association between positive thyroid antibodies in the first trimester and the risk of GDM in pregnant women compared with the reference group (pooled RR 1.08, 95% CI 0.97–1.21). Based on the present meta-analysis, the predictive value of thyroid antibodies in the first trimester on the risk of GDM is low. Similar results were attained if TPOAb only was taken into consideration. In addition, the pooled RR of thyroid antibodies–or TPOAb-positive status of the risk of GDM did not demonstrate statistical significance in euthyroid women. A previous meta-analysis (32) on maternal subclinical hypothyroidism and GDM identified that the risk of GDM in pregnant women with subclinical hypothyroidism is substantially higher than in euthyroid pregnant women (pooled RR 1.39, 95% CI 1.07–1.79). Therefore, it should be noted that the combined estimation of the association between thyroid antibodies and the risk of GDM may be biased owing to different TSH levels or thyroid dysfunction status. Consequently, three subgroup meta-analyses were conducted according to participants' thyroid function status to fully consider the effect of confounding by thyroid dysfunction. One group excluded pregnant women with thyroid dysfunction, and the other did not. In all subgroups, no significant association was found in euthyroid pregnant women. Therefore, in euthyroid pregnancy women, thyroid antibodies had no relationship with GDM, and thyroid antibodies–positive status in the first trimester had no obvious risk on GDM incidence. This finding will assist doctors who are managing euthyroid pregnant women with positive thyroid antibodies. Thyroid dysfunction is one of the most common endocrine disorders affecting reproductive-age women. With the progress of medical research, gestational thyroid dysfunction is not only one type of common gestation complication, but it is also an important risk factor leading to severe obstetrical complications during pregnancy or adverse pregnancy outcomes (33, 34). Our subgroup metaanalysis showed a significant increase in the risk of GDM in thyroid dysfunction during pregnancy with positive thyroid antibodies (pooled RR 1.35, 95% CI 1.06–1.71). This finding indicated that thyroid dysfunction in pregnancy with positive thyroid antibodies might have a higher risk for GDM than thyroid dysfunction in pregnancy with negative thyroid antibodies. Our findings provide insights for clinical treatment strategies for pregnant women with positive thyroid antibodies. Although women with positive isolated thyroid antibodies may not need drug therapy during pregnancy, doctors should more heavily monitor their clinical treatments and maintain TSH levels within normal range in pregnant women with both thyroid dysfunction and positive thyroid antibodies compared with those with isolated thyroid dysfunction to effectively reduce the incidence of GDM. The present study is the first meta-analysis to examine the association between thyroid antibodies and the risk of GDM with a greatly increased sample size, enhancing the power to assess the association. The major strength of this analysis is that all of the qualified studies, not only case-control studies but also prospective cohort studies, were included, eliminating the possibility of reverse causation and mini6
mizing selection bias. The predictive value of thyroid antibodies in early pregnancy on the risk of GDM was further assessed. Moreover, subgroup analyses were conducted to explore the association between thyroid antibodies and the risk of GDM considering maternal thyroid function. Conversely, this study has several limitations that should be noted. First, the individual studies used obvious different test assay platforms for thyroid antibodies and thresholds for test positivity in addition to different definitions of thyroid dysfunction, which could have resulted in heterogeneity in the present meta-analysis. Second, although we tried our best to contact all of the authors by e-mail, not all of the confounding factors were considered, owing to limited information obtained from the published articles and material provided by authors—except for thyroid function status. Third, taking racial differences into consideration, the heavily weighted studies in the meta-analysis were mostly from China, which limited the generalizability, especially to non– east Asian countries. However, multicenter studies and comparisons between different ethnic groups remain nonexistent so far. Large cohort studies of different races are eagerly expected in the future. Finally, the test timing points of thyroid antibodies were not coincident between the casecontrol studies and cohort studies included in the analysis, particularly compared with the incidence of GDM, which could weaken the validity of combined estimation when making causal inferences. In summary, our meta-analyses suggested a significant but not strong association between thyroid antibodies and the risk of GDM, and thyroid antibodies in the first trimester lacked predictive value for the risk of GDM. In addition, thyroid antibodies may not increase the risk of GDM in euthyroid pregnant women.
REFERENCES 1. 2.
3. 4. 5.
6.
7. 8.
9.
10.
World Health Organization. Laboratory diagnosis and monitoring of diabetes mellitus. Geneva: World Health Organization; 2002. Schneider S, Bock C, Wetzel M, Maul H, Loerbroks A. The prevalence of gestational diabetes in advanced economies. J Perinat Med 2012;40: 511–20. Tutino GE, Tam WH, Yang X, Chan JC, Lao TT, Ma RC. Diabetes and pregnancy: perspectives from Asia. Diabet Med 2014;31:302–18. Philips JC, Emonts P, Pintiaux A, Kirkpatrick C, Scheen AJ. Management of gestational diabetes. Rev Med Liege 2013;68:489–96. Bellamy L, Casas JP, Hingorani AD, Williams D. Type 2 diabetes mellitus after gestational diabetes: a systematic review and meta-analysis. Lancet 2009; 373:1773–9. Nerenberg K, Daskalopoulou SS, Dasgupta K. Gestational diabetes and hypertensive disorders of pregnancy as vascular risk signals: an overview and grading of the evidence. Can J Cardiol 2014;30: 765–73. Catalano PM. Trying to understand gestational diabetes. Diabet Med 2014; 31:273–81. Bitterman O, Bongiovanni M, Giuliani C, Roma G, Toscano V, Napoli A. Anti thyroperoxidase and anti thyroglobulin antibodies in diabetic pregnancies. J Endocrinol Invest 2014;37:911–5. Wang CF, Xu H, Cheng HD, Li XT. Pregnancy outcomes in euthyroid pregnant women with positive thyroid peroxidase antibodies. Chin J Birth Health Hered 2014;22:67–70. Kang SY, Wang Y, Wu L. The study of thyroid function and pregnancy outcomes in the pregnant women with gestational diabetes mellitus. Prog Obstet Gynecol 2014;23:50–2. VOL. - NO. - / - 2015
Fertility and Sterility® 11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21. 22.
Lin QL, Lin XF, Mai SM. Thyroid function and thyroid autoantibodies changes in the pregnant women of gestational diabetes millitus. Guangdong Med J 2014;35:1355–6. Li YH, Xie QP, Ma JX, Wang XJ, Shang SY. The study of thyroid function tests index changes in the pregnant with high risk of gestational diabetes mellitus. Matern Child Health Care China 2013;28:5420–1. Wang XM. The study of thyroid autoantibodies and thyroid function changes in the pregnant women with high-risk of gestational diabetes mellitus. China Med Her 2011;8:20–1. Wang H, Luo W, Liao WJ, Liu HX, Xiang B. The study of thyroid autoantibodies and thyroid function in pregnant women with increased risk of gestational diabetes mellitus. Chin J Lab Diag 2011;15:273–6. Karakosta P, Alegakis D, Georgiou V, Roumeliotaki T, Fthenou E, Vassilaki M, et al. Thyroid dysfunction and autoantibodies in early pregnancy are associated with increased risk of gestational diabetes and adverse birth outcomes. J Clin Endocrinol Metab 2012;97:4464–72. Mannisto T, Vaarasmaki M, Pouta A, Hartikainen AL, Ruokonen 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. Olivieri A, Valensise H, Magnani F, Medda E, de Angelis S, d’Archivio M, et al. High frequency of antithyroid autoantibodies in pregnant women at increased risk of gestational diabetes mellitus. Eur J Endocrinol 2000;143: 741–7. Pascual Corrales E, Andrada P, Auba M, Ruiz Zambrana A, Guillen Grima F, Salvador J, et al. Is autoimmune thyroid dysfunction a risk factor for gestational diabetes? Endocrinol Nutr 2014;61:377–81. Chen LM, Zhang Q, Si GX, Chen QS, Ye EL, Yu LC, et al. Associations between thyroid autoantibody status and abnormal pregnancy outcomes in euthyroid women. Endocrine 2014;48:924–8. Ortega-Gonzalez C, Liao-Lo A, Ramirez-Peredo J, Carino N, Lira J, Parra A. Thyroid peroxidase antibodies in Mexican-born healthy pregnant women, in women with type 2 or gestational diabetes mellitus, and in their offspring. Endocr Pract 2000;6:244–8. Montaner P, Juan L, Campos R, Gil L, Corcoy R. Is thyroid autoimmunity associated with gestational diabetes mellitus? Metabolism 2008;57:522–5. Olivieri A, Magnani F, Valensise H, d’Archivio M, Medda E, Baccarini S, et al. Thyroid autoimmunity in pregnant women at risk for GDM. Ann Ist Super Sanita 1997;33:447–50.
VOL. - NO. - / - 2015
23.
24. 25.
26.
27.
28.
29.
30. 31. 32.
33.
34.
Velkoska Nakova V, Krstevska B, Dimitrovski C, Simeonova S, Hadzi-Lega M, Serafimoski V. Prevalence of thyroid dysfunction and autoimmunity in pregnant women with gestational diabetes and diabetes type 1. Prilozi 2010;31: 51–9. Zhang XL. Clinical epidemiology research of thyroid disease during pregnancy. Beijing: Capital Medical University; 2013. Tang L. The combined effect of maternal mild hyperthyroidism and emotional symptoms to adverse pregnancy outcomes during the first and second trimester: a population-based cohort study. Anhui: Anhui Medical University; 2013. Zhao Y, Liu XQ, Su Y, Lin Y, Liu W. Analysis of thyroid function and antibody changes in gestational diabetes mellitus and pregestational diabetes mellitus. Chin J Control Endem Dis 2013;28:232–3. Agarwal MM, Dhatt GS, Punnose J, Bishawi B, Zayed R. Thyroid function abnormalities and antithyroid antibody prevalence in pregnant women at high risk for gestational diabetes mellitus. Gynecol Endocrinol 2006; 22:261–6. Negro R, Schwartz A, Gismondi R, Tinelli A, Mangieri T, Stagnaro-Green A. Thyroid antibody positivity in the first trimester of pregnancy is associated with negative pregnancy outcomes. J Clin Endocrinol Metab 2011;96: E920–4. Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, et al. Meta-analysis of Observational Studies in Epidemiology (MOOSE) Group. Meta-analysis of observational studies in epidemiology: a proposal for reporting. JAMA 2000;283:2008–12. Zhang J, Yu KF. What’s the relative risk? A method of correcting the odds ratio in cohort studies of common outcomes. JAMA 1998;280:1690–1. Miettinen O. Estimability and estimation in case-referent studies. Am J Epidemiol 1976;103:226–35. Toulis KA, Stagnaro-Green A, Negro R. Maternal subclinical hypothyroidsm and gestational diabetes mellitus: a meta-analysis. Endocr Pract 2014;20: 703–14. Vandana, Kumar A, Khatuja R, Mehta S. Thyroid dysfunction during pregnancy and in postpartum period: treatment and latest recommendations. Arch Gynecol Obstet 2014;289:1137–44. Ong GS, Hadlow NC, Brown SJ, Mun Lim E, Walsh JP. Does the thyroidstimulating hormone measured concurrently with first trimester biochemical screening tests predict adverse pregnancy outcomes occurring after 20 weeks gestation? J Clin Endocrinol Metab 2014;99:E2668–72.
7
Characteristics of the 20 studies.
Reference Wang 2014 (9)
Location
Period
Design
Definition of GDM
China
2011.9– 2013.1
Cohort
China Kang 2014 (10)
2011.1– 2011.12
Case75-g OGTT control fasting, 1 h, 2 h, R1 abnormal
Lin
China
2012.3– 2013.6
Case75-g OGTT control fasting, 1 h, 3 h, R1 abnormal
China
2008.6– 2011.6
CaseOGTT, no control detailed information
2014 (11)
VOL. - NO. - / - 2015
Li 2013 (12)
Yang. Thyroid antibodies and GDM. Fertil Steril 2015.
75-g OGTT fasting, 1 h, 2 h, R1 abnormal
Definition of thyroid dysfunction Serum TSH or FT4 out of reference range. Abbott TSH reference range 0.04–3.61 mIU/L (8–12 wk), 0.09–3.93 mIU/L (13–27 wk). Abbott FT4 reference range 12.32–21.66 pmol/L (8–12 wk), 11.19–19.14 pmol/L (13–27 wk). Roche TSH reference range 0.03–4.17 mIU/L (8–12 wk), 0.19–4.94 mIU/L (13–19 wk), 0.70–5.42 mIU/L (20–27 wk), 0.63–5.52 mIU/L (28–40 wk). Roche FT4 reference range 11.57–20.88 pmol/L (8–12 wk), 10.09–16.88 pmol/L (13–19 wk), 8.39–13.79 pmol/L (20–27 wk), 8.35–14.14 pmol/L (28–40 wk). Serum TSH out of reference range 0.3–2.5 mIU/L (20 wk) or serum FT4 out of reference range 9.01–19.05 pmol/L. Serum TSH out of reference range 0.05–5.17 mIU/L (T1), 0.39–5.22 mIU/L (T2), 0.60–6.84 mIU/L (T3) or serum FT4 out of reference range 12.91–22.35 pmol/mL (T1), 9.81–17.26 pmol/ mL (T2),12–15.71 pmol/ mL (T3). NG
Eligible subjects
Positive thyroid antibodies
Exclusion criteria
GDM
Control
Thyroid dysfunction, miscarriages, history of TD and DM
457
2,062
33.11% (TPOAb)
History of TD and DM
569
1,221
9.94% (TPOAb)
History of TD
86
120
T1 6.31%; T2 7.77%; T3 11.17% (TPOAb)
None
18
102
15% (TPOAb); 11.67% (TGAb); 20% (ThyAb)
ORIGINAL ARTICLE: ENVIRONMENT AND EPIDEMIOLOGY
8
SUPPLEMENTAL TABLE 1
VOL. - NO. - / - 2015
SUPPLEMENTAL TABLE 1 Continued.
Reference
Location
Period
Definition of GDM
Design
Definition of thyroid dysfunction
Eligible subjects Exclusion criteria
GDM
Control
Wang XM China 2011 (13)
2008.10– 2010.10
CaseOGTT, no detailed control information
NG
None
30
130
Wang H China 2011 (14)
NG
CaseOGTT, no detailed control information
None
32
133
Karakosta Greece 2012 (15)
2007.2– 2008.2
Cohort
Miscarriages
88
1053
Mannisto Finland 2010 (16)
1986– 2006
Cohort
Serum TSH out of reference range 0.34–5.6 mU/L or serum FT4 out of reference range 7.5–21.1 pmol/L Normal TSH 0.05–2.53 mIU/mL; high TSH >2.53 mIU/mL Serum TSH out of reference range 0.19–3.6 mU/L or serum FT4 out of reference range 11.96–20.5 pmol/L Serum TSH out of reference range 0.2–4 mUI/mL or serum FT4 out of reference range 8–18 pg/mL TSH R2.5 mU/mL
Olivieri Italy 2000 (17)
Corrales America 2014 (18)
Chen China 2014 (19)
Montaner Spain 2008 (21)
Case100-g OGTT, fasting, control 1 h, 2 h, 3 h, R2 abnormal
2008–2012
Cohort
2009.5– 2012.7
Cohort
O'Sullivan test, blood glucose >140 mg/dL, 50 g 1 h, 100 g 12 h Fasting >5.3 mmol/L, Serum TSH out of reference 1 h >10.0 mmol/L, range 0.09–3.47 mIU/L or 2 h >8.6 mmol/L (T1), 0.20–3.81 mIU/L (T2), 0.67–4.99 mIU/L (T3) or serum FT4 out of reference range 6–12.25 ng/L (T1), 4.30–9.74 ng/L (T2), 4.56–8.50 ng/L (T3).
1997.7.1– CaseNG 1997.12.31 control
2003.5– 2003.12
Yang. Thyroid antibodies and GDM. Fertil Steril 2015.
Cohort
50-g oral load
Serum TSH out of reference range 0.3–4.2 mU/L or serum FT4 out of reference range 9.5–25.0 pmol/L NG
Multiple pregnancy
12.5% (TPOAb); 6.88% (TGAb); 16.25% (ThyAb) 13.33% (TPOAb); 6.06% (TGAb); 15.76% (ThyAb)
15.78% (ThyAb)
50 (TPOAb); 5,713 (TPOAb); 4.50% (TPOAb); 50 (TGAb) 5,655 (TGAb) 4.50% (TGAb)
None
19
179
History of impaired carbohydrate metabolism, earlier diagnosis of thyroid dysfunction Thyroid dysfunction, history of treatments with thyroid hormones or antithyroid drugs, autoimmune diseases, congenital heart disease or heart failure, elevated serum transaminases or serum creatinine, multiple pregnancy Thyroid dysfunction, history of TD
15
41
33.93% (TPOAb); 12.5% (TGAb); 37.5% (ThyAb)
268
7,373
5.03% (TPOAb); 3.89% (TGAb); 7.04% (ThyAb)
50
50
43
576
Pre-pregnancy diabetes mellitus
15.66% (ThyAb)
45% (TPOAb)
10.02% (TPOAb)
9
Fertility and Sterility®
OrtegaMexico Gonzalez 2000 (20)
NG
100-g OGTT, fasting, 1 h, 2 h, 3 h, R2 abnormal OGTT, no detailed information
Positive thyroid antibodies
Continued.
Reference
Location
Olivieri Italy 1997 (22)
Period NG
Definition of GDM
Design
Case100-g OGTT, fasting, control 1 h, 2 h, 3 h, R2 abnormal Case100-g OGTT, fasting, control 1 h, 2 h, 4 h, R2 abnormal
Nakova Macedonia 2010 (23)
2009.5– 2009.11
Zhang China 2013 (24)
2010.10– 2012.12
Cohort
75-g OGTT, fasting, 1 h, 2 h, R1 abnormal
Tang China 2013 (25)
2008.10– 2010.10
Cohort
NG
Zhao China 2013 (26)
2010–2012
Agarwal UAE 2006 (27)
2004.1– 2004.5
Negro Italy 2011 (28)
2005
Case75-g OGTT fasting, control 1 h, 3 h, R1 abnormal Cohort 75-g OGTT, fasting, 1 h, 4 h, R1 abnormal Cohort
NG
Definition of thyroid dysfunction
Eligible subjects Exclusion criteria
Positive thyroid antibodies
GDM
Control
3
46
14.29% (ThyAb)
40
30
12.86% (TPOAb)
Multiple pregnancy, history of TD and DM, family history of TD
255
1,813
18.23% (TPOAb)
Thyroid dysfunction, multiple pregnancy History of TD
49
2,199
17.79% (TPOAb)
49
51
7% (ThyAb)
History of TD
186
69
20% (TPOAb)
Thyroid dysfunction, history of TD
135
3,458
NG
None
Serum TSH out of reference range 0.2–4.2 mIU/L or serum FT4 out of reference range 10.3–24.45 pmol/L SCH: serum TSH 3.67–10 mIU/L (T1), 3.74–10 mIU/L (T2), 4.45–10 mIU/L (T3); and FT4 in reference range 7.2–18.66 pmol/L (T1), 5.53–12.74 pmol/L (T2), 5.40–10.40 pmol/L (T3) Serum TSH or FT4 out of reference range, no detailed information Serum TSH or FT4 out of reference range, no detailed information Serum TSH out of reference range 0.3–4.32 mIU/L or serum FT4 out of reference range 9.8–18.6 pmol/L TSH R2.5 mIU/L or hyperthyroid
History of TD
6.82% (ThyAb)
Note: DM ¼ diabetes mellitus; FT4 ¼ free thyroxine; NG ¼ data were not provided in the original study; SCH ¼ subclinical hypothyroidism; T1 ¼ first trimester; T2 ¼ second trimester; T3 ¼ third trimester; TD ¼ thyroid disease; TGAb ¼ thyroglobulin antibody; ThyAb ¼ thyroid antibody; TPOAb ¼ thyroid peroxidase antibody; TSH ¼ thyroid-stimulating hormone. Yang. Thyroid antibodies and GDM. Fertil Steril 2015.
ORIGINAL ARTICLE: ENVIRONMENT AND EPIDEMIOLOGY
10
SUPPLEMENTAL TABLE 1
VOL. - NO. - / - 2015
Thyroid antibodies and gestational diabetes mellitus: a meta-analysis Ying Yang, Ph.D.,a Qian Li, M.D.,a,b Qianqian Wang, Ph.D.,c and Xu Ma, M.D.a,b a National Research Institute for Family Planning; b Graduate School of Peking Union Medical College; and c Department of Molecular Orthopaedics, Center of Clinical Research and Evidence-Based Medicine, Beijing, People’s Republic of China
Objective: To determine whether thyroid antibodies are associated with an increased risk of gestational diabetes mellitus (GDM) in pregnant women. Design: Meta-analysis. Setting: Not applicable. Patient(s): Twenty cohort and case-control studies involving pregnant women with positive thyroid antibodies were the exposure of interest, and GDM was the outcome. Intervention(s): None. Main Outcome Measure(s): A fixed-effects model was used to evaluate the relationship between thyroid antibodies and the risk of GDM. Subgroup analyses were performed among different types of study design, different thyroid antibodies, and patients with specific thyroid dysfunction. Result(s): The search strategy identified 167 potentially relevant publications, of which 20 were included in the meta-analysis. A significant association between thyroid antibodies and GDM was observed. A meta-analysis of the 11 cohort studies with pregnant women with positive thyroid antibodies in their first trimester suggested no obvious risk of GDM compared with the reference group. In subgroup meta-analyses, no significant association between thyroid antibodies and GDM was found in euthyroid pregnant women, whereas a significant positive association was identified in women with a thyroid dysfunction. Conclusion(s): Based on the currently available evidence, there is a significant but not strong association between thyroid antibodies and the risk of GDM, and thyroid antibodies in the first trimester lack predictive value for the risk of GDM. In addition, thyroid antibodies may not increase the risk of GDM in euthyroid pregUse your smartphone nant women. (Fertil SterilÒ 2015;-:-–-. Ó2015 by American Society for Reproductive to scan this QR code Medicine.) and connect to the Key Words: Thyroid, autoimmunity, antibodies, gestational diabetes mellitus Discuss: You can discuss this article with its authors and with other ASRM members at http:// fertstertforum.com/yangy-thyroid-antibodies-gdm/
G
estational diabetes mellitus (GDM) is defined by the World Health Organization as having ‘‘any degree of glucose intolerance with onset or first recognition during pregnancy’’ (1). The prevalence of GDM in pregnant women varies substantially, ranging from 1.7% to 11.6% in advanced economies (2). In Asian countries, the prevalence varies
according to the screening strategy and diagnostic criteria and ranges from 1% to 20%, with an increasing trend in recent years (3). GDM is one of the most common pregnancy complications associated with a number of adverse outcomes, including miscarriage, delivery, excessive fetal growth, birth trauma, and neonatal metabolic abnormalities (4). In addition, women
Received January 22, 2015; revised and accepted June 4, 2015. Y.Y. has nothing to disclose. Q.L. has nothing to disclose. Q.W. has nothing to disclose. X.M. has nothing to disclose. Y.Y. and Q.L. should be considered similar in author order. Supported by Operation Expenses for Basic Scientific Research of Central Authorities (grant no. 2012GJSSJKC03), International Science and Technology Cooperation Program of China (grant no. 2012DFB30130), and National Natural Science Foundation (grant no. 81402757). Reprint requests: Xu Ma, M.D., Professor, National Research Institute for Family Planning, 12 Dahuisi Road, Haidian District, Beijing 100081, People's Republic of China (E-mail: maxutougao@163. com). Fertility and Sterility® Vol. -, No. -, - 2015 0015-0282/$36.00 Copyright ©2015 American Society for Reproductive Medicine, Published by Elsevier Inc. http://dx.doi.org/10.1016/j.fertnstert.2015.06.003 VOL. - NO. - / - 2015
discussion forum for this article now.*
* Download a free QR code scanner by searching for “QR scanner” in your smartphone’s app store or app marketplace.
who have been exposed to GDM during pregnancy are at an exceptionally high risk of developing type 2 diabetes, chronic hypertension, and vascular disease later in life (5, 6). GDM has undoubtedly become an important public health issue worldwide. Studies have shown that GDM is caused by b-cell dysfunction and insulin resistance (7). A high prevalence of elevated thyroid antibodies have appeared in patients with a history of insulin resistance, such as pregnant women with GDM (8). The link between autoimmune disorders and insulin resistance could be the inflammatory events that are associated with both conditions. Autoimmune thyroid dysfunction was found to be associated with insulin resistance, suggesting a potential relationship between 1
ORIGINAL ARTICLE: ENVIRONMENT AND EPIDEMIOLOGY autoimmune thyroid dysfunction and GDM. The prevalence of thyroid antibodies in pregnant women is approximately 10%–15% (8). In recent years, many epidemiologic studies have investigated the links between thyroid antibodies, including thyroid peroxidase antibody (TPOAb) and thyroglobulin antibody (TgAb), and risk of GDM (9–28). However, these studies have apparently yielded conflicting results. Both obstetricians and endocrinologists pay great attention to whether euthyroid pregnant women with positive thyroid antibodies have a higher risk than women with negative thyroid antibodies and how their condition should be managed. Therefore, to clarify the evidence, we conducted a meta-analysis of available case-control and cohort studies of thyroid antibodies and GDM.
MATERIALS AND METHODS Search Strategy and Study Selection We followed the Meta-analysis of Observational Studies in Epidemiology guidelines to explicitly report this metaanalysis (29). To study the association between thyroid antibodies (TPOAb or TgAb) and GDM, an electronic search was conducted with the use of Pubmed and two Chinese databases, Wanfang and China National Knowledge Internet (CNKI), from inception to October 2014. The following search words were used: (1) thyroid antibody, thyroid antibodies, thyroid autoimmune antibody, thyroid autoimmune antibodies, thyroid autoantibody, thyroid autoantibody, TPOAb, TPOAb, thyroid peroxidase antibody, thyroid peroxidase antibodies, thyroid globulin antibody, thyroid globulin antibodies, TG-Ab, or TgAb; and (2) gestational diabetes, pregnancy diabetes, maternal diabetes, gestational diabetes mellitus, or GDM. An initial screening was based on the titles and abstracts of published papers, and the full texts were then reviewed carefully during the second screening. In addition, reference lists of the original studies and review articles were considered. Irrelevant studies, reviews, case reports, and letters to the editor were ineligible. Overlapping studies were also excluded. All of the literature searches were reviewed independently by two authors.
Data Extraction and Assessment of Quality of Methods Relevant data were extracted from each paper with the use of standardized data collection forms. Any disagreements were resolved by discussion until an agreement could be found. We contacted the corresponding authors through email to obtain additional detailed information when necessary. The related data were extracted to calculate an odds ratio (OR) or relative risk (RR) for the association between thyroid antibodies and GDM. We also extracted study characteristics for further exploration in subgroup meta-analyses. The following characteristics were recorded: first author, year of publication, study period, study design, definition of GDM, definition of thyroid dysfunction, and exclusion criteria (Supplemental Table 1, available online at www.fertstert.org). The selected studies were assessed for quality with the use of the Newcastle-Ottawa scale (NOS) for cohort studies and 2
case-control studies (http://www.ohri.ca/Programs/clinical_ epidemiology/default.asp). Studies with NOS scores of 0–3, 4–6, and 7–9 were considered to be of low, moderate, and high quality, respectively. Two investigators extracted the data and independently assessed the quality. Disagreements were resolved by consensus or arbitration by a third investigator.
Statistical Analyses The RR was used as the common measure of association across the studies. ORs were transformed into RRs according to the following formula: RR ¼ OR/[(1 P0) þ (P0 OR)], where P0 is the incidence of GDM in the thyroid antibodies–negative group (30). In addition, the Miettinen test–based approach was used to calculate the variance of logRR (variance logRR ¼ variance logOR [logRR/logOR]) (31). The overall combined RR and 95% confidence interval (CI) was calculated with the use of logRR and variance of logRR. Heterogeneity across studies was assessed with the use of the Cochrane Q statistic (significance level at P< .10) and the I2 statistic. If I2 < 50%, a random-effects model (DerSimonian–Laird method) was used to pool the results; otherwise, a fixed-effects model (Mantel–Haenszel method) was used. In addition, subgroup analyses were conducted stratifying by the type of study design, type of thyroid antibody (TPOAb or TgAb), and thyroid dysfunction exclusions. Sensitivity analyses were performed to detect the effects of individual study on the pooled result by omitting one study in each turn. Publication bias was assessed with the use of funnel plots and the Egger regression test. We conducted all analyses with the use of Stata software, version 12.0. Except where otherwise specified, a P value of < .05 was considered to be statistically significant.
RESULTS Study Selection and Study Characteristics The study selection process is shown in Figure 1. Initially, 167 unique citations were retrieved from the Pubmed, CNKI, and Wanfang databases. After the first screening of the abstracts and titles, 130 citations were excluded, primarily owing to being reviews, clearly irrelevant studies, or studies not focusing on GDM. Then we reviewed full-text articles for a more comprehensive evaluation. Two articles were excluded because the exposure was not relevant, and 15 articles were excluded because they did not focus on GDM. Finally, 20 articles (9–28) were included in our meta-analysis. The meta-analysis included 10 cohort studies and 10 case-control studies conducted with 34,566 participants. The characteristics of the 20 articles are presented in Supplemental Table 1. The cohort sizes of the studies ranged from 49 to 7,641 and were published from 1997 to 2014. Among the studies, 11 (9–14, 19, 24–27), seven (15–17, 21– 23, 28), and two (18, 20) were conducted in Asia, Europe, and America, respectively. The ascertainment of thyroid antibodies varied across studies. Among the 20 articles included here, nine studies evaluated TPOAb only (9–11, 20, 21, 23–25, 27), ten studies evaluated both TgAb and TPOAb (one type of positive antibody was considered to be a VOL. - NO. - / - 2015
Fertility and Sterility® Synthesis Analysis
FIGURE 1 Potentially relevant studies through literature search (n=167)
identified
130 articles excluded Obvious irrelevant studies(n=89) Reviews(n=23) Not on gestational diabetes mellitus(n=18) Full-text articles reviewed for more detailed evaluation (n=37)
To study the relationship between thyroid antibodies and GDM, all eligible studies were first summarized to calculate a combined RR. Then meta-analyses of the 11 cohort studies were conducted to examine the relationship between positive thyroid antibodies in early pregnancy and the risk of GDM. Summaries of the pooled outcomes in the synthesis metaanalyses using all of the studies and cohort studies are presented in Table 1.
Thyroid Antibodies and GDM
16 articles excluded Exposure not relevant (n=2) Not on gestational diabetes mellitus (n=15)
Articles included in meta-analysis (n=20) Cohort study(n=10) Case-control study(n=10)
Flow chart of study selection. Yang. Thyroid antibodies and GDM. Fertil Steril 2015.
positive thyroid antibody) (12–15, 17–19, 22, 26, 28), and one article (16) reported data on TPOAb and TgAb separately and was extracted as two individual studies. In addition, the definition of GDM, definition of thyroid dysfunction, and exclusion criteria did not conform across the studies (Supplemental Table 1). All 21 studies were of a moderate or high quality on the Newcastle-Ottawa scale.
We obtained 21 individual data sets from 20 articles. The synthesis meta-analysis revealed a significant association between thyroid antibodies and GDM, with a pooled RR of 1.12 (95% CI 1.03–1.22; Table 1; Fig. 2). The meta-analysis of the 11 cohort studies (9, 15, 16, 18, 19, 21, 24, 25, 27, 28) suggested that pregnant women with positive thyroid antibodies in the first trimester had no obvious risk of GDM compared with the reference group (pooled RR 1.07, 95% CI 0.96–1.19; Table 1; Fig. 2). For all of the aforementioned meta-analyses, fixed-effect models were used because no statistically significant heterogeneity was found. Sensitivity analyses suggested that the combined RRs were valid and credible.
TPOAb and GDM Fifteen studies (9–14, 16, 18–21, 23–25, 27) involving 23,718 pregnant women reported relative data on the association between TPOAb and GDM. The synthesis meta-analysis showed a significant association between TPOAb and GDM risk with a pooled RR of 1.11 (95% CI 1.01–1.21). The
TABLE 1 Summary of the meta-analyses of thyroid antibodies and gestational diabetes mellitus risk. Analysis items All of the studies Thyroid antibodies Total studies With euthyroid Including thyroid dysfunction Thyroid dysfunction TPOAb Total studies With euthyroid Including thyroid dysfunction TgAb Total studies Cohort studies Thyroid antibodies Total studies With euthyroid Including thyroid dysfunction TPOAb Total studies With euthyroid Including thyroid dysfunction TgAb Total studies
Studies (participants)
RR (95% CI)
I2 (P value)
Begg test
21 (34,502) 8 (20,533) 16 (18,401) 5 (561)
1.12 (1.03–1.22) 1.08 (0.97–1.22) 1.18 (1.06–1.31) 1.35 (1.06–1.71)
0.0% (.590) 5.3% (.390) 0.0% (.730) 0.0% (.539)
0.486 0.479 0.752 0.293
15 (23,718) 6 (15,878) 11 (11,210)
1.11 (1.01–1.21) 1.06 (0.94–1.20) 1.16 (1.04–1.30)
0.0% (.473) 1.7% (.405) 0.0% (.624)
0.704 0.192 0.834
6 (13,847)
1.34 (0.93–1.95)
0.0% (.974)
0.521
11 (31,546) 6 (18,947) 7 (15,545)
1.07 (0.96–1.19) 1.12 (0.98–1.28) 1.14 (0.97–1.34)
2.5% (.419) 4.4% (.388) 22.3% (.259)
0.548 0.916 0.988
8 (21,107) 4 (14,292) 5 (8,699)
1.06 (0.94–1.19) 1.08 (0.94–1.25) 1.08 (0.91–1.29)
0.0% (.542) 7.4% (.356) 0.0% (.420)
0.412 0.438 0.910
3 (13,402)
1.23 (0.78–1.95)
0.0% (.755)
0.212
Note: All of the above meta-analyses used fixed-effects models. CI ¼ confidence interval; RR ¼ relative risk; TgAb ¼ thyroglobulin antibody; TPOAb ¼ thyroid peroxidase antibody. Yang. Thyroid antibodies and GDM. Fertil Steril 2015.
VOL. - NO. - / - 2015
3
ORIGINAL ARTICLE: ENVIRONMENT AND EPIDEMIOLOGY
FIGURE 2
Forest plot showing the association between thyroid antibodies and gestational diabetes mellitus. CI ¼ confidence interval; RR ¼ relative risk. Yang. Thyroid antibodies and GDM. Fertil Steril 2015.
meta-analysis of the eight cohort studies (9, 16, 18, 19, 21, 24, 25, 27) suggested that compared with the reference group, TPOAb-positive status in the first trimester had no obvious impact on the risk of GDM in pregnant women (pooled RR 1.06, 95% CI 0.94–1.19; Table 1). No statistically significant heterogeneity was found, and fixed-effect models were used in the meta-analysis. Sensitivity analyses by sequential omission of individual studies did not materially alter the overall combined RR, suggesting that the combined RR was valid and credible.
TgAb and GDM Six studies (12–14, 16, 18, 19) involving 13,847 pregnant women reported relative data on the association between TgAb and GDM risk. The synthesis meta-analysis showed no significant association between TgAb and GDM, with a pooled RR of 1.34 (95% CI 0.93–1.95). The meta-analysis of the three cohort studies (16, 18, 19) demonstrated results similar to the association between TPOAb and GDM risk (pooled RR 1.23, 95% CI 0.78–1.95; Table 1). For all of these meta-analyses, no statistically significant heterogeneity was found, and fixed-effect models were used. Sensitivity analyses suggested that the combined RRs were valid and credible. 4
Subgroup Analysis Based on detailed clinical information provided by the articles included in this meta-analysis, three subgroup meta-analyses were conducted: excluding women with thyroid dysfunction, including women with thyroid dysfunction, and including only women with thyroid dysfunction. In studies that included euthyroid women (9, 10, 15, 19, 20, 24, 25, 28), the pooled RR of thyroid antibodies or TPOAb positive status on GDM incidence did not reach statistical significance (for thyroid antibodies, pooled RR 1.08, 95% CI 0.97–1.22; for TPOAb, pooled RR 1.06, 95% CI 0.94–1.20; Table 1; Fig. 3). Notably, significant associations were observed between thyroid antibodies or a positive TPOAb status and the risk of GDM in studies including women with thyroid dysfunction (for thyroid antibodies, pooled RR 1.18, 95% CI 1.06–1.31; for TPOAb, pooled RR 1.16, 95% CI 1.04–1.30; Table 1). Owing to the lack of relevant studies, a subgroup analysis of pregnant women with TgAb-positive status was not performed. The same grouping strategy was used to conduct metaanalyses of the cohort studies. The pooled RRs for the risk of GDM in pregnant women with thyroid antibodies or with TPOAb-positive status were not significant in euthyroid VOL. - NO. - / - 2015
Fertility and Sterility®
FIGURE 3
Forest plot showing the association between thyroid antibodies and the risk of gestational diabetes mellitus in euthyroid pregnancy. CI ¼ confidence interval; RR ¼ relative risk. Yang. Thyroid antibodies and GDM. Fertil Steril 2015.
pregnant women group (for thyroid antibodies, pooled RR 1.12, 95% CI 0.98–1.28; for TPOAb, pooled RR ¼ 1.08, 95% CI 0.94–1.25; Table 1; Fig. 3). No significant results were also observed in the association between thyroid antibodies or TPOAb-positive status and the risk of GDM in studies including women with thyroid dysfunction (for thyroid antibodies, pooled RR 1.14, 95% CI 0.97–1.34; for TPOAb, pooled RR 1.08, 95% CI 0.91–1.29; Table 1). A subgroup analysis of pregnant women with TgAb-positive status was absent owing to the lack of relevant studies. Furthermore, we obtained five data sets from four articles (10, 15, 18, 24) that involved 561 participates with thyroid dysfunction. Based on the data from those five studies, we conducted a meta-analysis to evaluate the associationbetween thyroid antibodies positive status and the risk of GDM in women with thyroid dysfunction. The pooled RR of the fixed-effects model showed a significant increase in the risk of GDM on women with positive thyroid antibodies (polled RR 1.35, 95% CI 1.06–1.71).
Publication Bias Publication biases were not found in the included studies, as suggested by the Begg tests and Begg funnel plots. All of the P values given by Begg tests are listed in Table 1. VOL. - NO. - / - 2015
DISCUSSION Studies have demonstrated that the underlying mechanism of GDM is likely to involve increased insulin resistance more than decreased insulin secretion (7). Thyroid autoimmunity could be an indication of autoimmune dysfunction. Montaner et al. (21) suggested that insulin resistance is an important mechanism of thyroid antibodies for the development of GDM. The presence of thyroid antibodies would result in an increase of proinflammatory cytokines, inducing insulin resistance. Nevertheless, the mechanisms involved in the association between thyroid antibodies and GDM risk remain uncertain. Epidemiologic studies have reported controversial results regarding the association between thyroid antibodies and the incidence of GDM. These controversies may result from different study designs and small sample sizes among the studies. Meta-analysis has been considered to be a powerful tool for pooling the results from different studies, which can overcome the problems of small sample size and inadequate statistical power. Therefore, we performed the present meta-analyses to study the association between thyroid antibodies and the risk of GDM. In the meta-analysis of all of the studies, including both case-control and cohort studies, a significant but not very strong association between thyroid antibodies and the risk of GDM was observed, with an overall combined RR of 1.12 (95% CI 1.03–1.22). However, the results of the 5
ORIGINAL ARTICLE: ENVIRONMENT AND EPIDEMIOLOGY meta-analysis of cohort studies only showed no significant association between positive thyroid antibodies in the first trimester and the risk of GDM in pregnant women compared with the reference group (pooled RR 1.08, 95% CI 0.97–1.21). Based on the present meta-analysis, the predictive value of thyroid antibodies in the first trimester on the risk of GDM is low. Similar results were attained if TPOAb only was taken into consideration. In addition, the pooled RR of thyroid antibodies–or TPOAb-positive status of the risk of GDM did not demonstrate statistical significance in euthyroid women. A previous meta-analysis (32) on maternal subclinical hypothyroidism and GDM identified that the risk of GDM in pregnant women with subclinical hypothyroidism is substantially higher than in euthyroid pregnant women (pooled RR 1.39, 95% CI 1.07–1.79). Therefore, it should be noted that the combined estimation of the association between thyroid antibodies and the risk of GDM may be biased owing to different TSH levels or thyroid dysfunction status. Consequently, three subgroup meta-analyses were conducted according to participants' thyroid function status to fully consider the effect of confounding by thyroid dysfunction. One group excluded pregnant women with thyroid dysfunction, and the other did not. In all subgroups, no significant association was found in euthyroid pregnant women. Therefore, in euthyroid pregnancy women, thyroid antibodies had no relationship with GDM, and thyroid antibodies–positive status in the first trimester had no obvious risk on GDM incidence. This finding will assist doctors who are managing euthyroid pregnant women with positive thyroid antibodies. Thyroid dysfunction is one of the most common endocrine disorders affecting reproductive-age women. With the progress of medical research, gestational thyroid dysfunction is not only one type of common gestation complication, but it is also an important risk factor leading to severe obstetrical complications during pregnancy or adverse pregnancy outcomes (33, 34). Our subgroup metaanalysis showed a significant increase in the risk of GDM in thyroid dysfunction during pregnancy with positive thyroid antibodies (pooled RR 1.35, 95% CI 1.06–1.71). This finding indicated that thyroid dysfunction in pregnancy with positive thyroid antibodies might have a higher risk for GDM than thyroid dysfunction in pregnancy with negative thyroid antibodies. Our findings provide insights for clinical treatment strategies for pregnant women with positive thyroid antibodies. Although women with positive isolated thyroid antibodies may not need drug therapy during pregnancy, doctors should more heavily monitor their clinical treatments and maintain TSH levels within normal range in pregnant women with both thyroid dysfunction and positive thyroid antibodies compared with those with isolated thyroid dysfunction to effectively reduce the incidence of GDM. The present study is the first meta-analysis to examine the association between thyroid antibodies and the risk of GDM with a greatly increased sample size, enhancing the power to assess the association. The major strength of this analysis is that all of the qualified studies, not only case-control studies but also prospective cohort studies, were included, eliminating the possibility of reverse causation and mini6
mizing selection bias. The predictive value of thyroid antibodies in early pregnancy on the risk of GDM was further assessed. Moreover, subgroup analyses were conducted to explore the association between thyroid antibodies and the risk of GDM considering maternal thyroid function. Conversely, this study has several limitations that should be noted. First, the individual studies used obvious different test assay platforms for thyroid antibodies and thresholds for test positivity in addition to different definitions of thyroid dysfunction, which could have resulted in heterogeneity in the present meta-analysis. Second, although we tried our best to contact all of the authors by e-mail, not all of the confounding factors were considered, owing to limited information obtained from the published articles and material provided by authors—except for thyroid function status. Third, taking racial differences into consideration, the heavily weighted studies in the meta-analysis were mostly from China, which limited the generalizability, especially to non– east Asian countries. However, multicenter studies and comparisons between different ethnic groups remain nonexistent so far. Large cohort studies of different races are eagerly expected in the future. Finally, the test timing points of thyroid antibodies were not coincident between the casecontrol studies and cohort studies included in the analysis, particularly compared with the incidence of GDM, which could weaken the validity of combined estimation when making causal inferences. In summary, our meta-analyses suggested a significant but not strong association between thyroid antibodies and the risk of GDM, and thyroid antibodies in the first trimester lacked predictive value for the risk of GDM. In addition, thyroid antibodies may not increase the risk of GDM in euthyroid pregnant women.
REFERENCES 1. 2.
3. 4. 5.
6.
7. 8.
9.
10.
World Health Organization. Laboratory diagnosis and monitoring of diabetes mellitus. Geneva: World Health Organization; 2002. Schneider S, Bock C, Wetzel M, Maul H, Loerbroks A. The prevalence of gestational diabetes in advanced economies. J Perinat Med 2012;40: 511–20. Tutino GE, Tam WH, Yang X, Chan JC, Lao TT, Ma RC. Diabetes and pregnancy: perspectives from Asia. Diabet Med 2014;31:302–18. Philips JC, Emonts P, Pintiaux A, Kirkpatrick C, Scheen AJ. Management of gestational diabetes. Rev Med Liege 2013;68:489–96. Bellamy L, Casas JP, Hingorani AD, Williams D. Type 2 diabetes mellitus after gestational diabetes: a systematic review and meta-analysis. Lancet 2009; 373:1773–9. Nerenberg K, Daskalopoulou SS, Dasgupta K. Gestational diabetes and hypertensive disorders of pregnancy as vascular risk signals: an overview and grading of the evidence. Can J Cardiol 2014;30: 765–73. Catalano PM. Trying to understand gestational diabetes. Diabet Med 2014; 31:273–81. Bitterman O, Bongiovanni M, Giuliani C, Roma G, Toscano V, Napoli A. Anti thyroperoxidase and anti thyroglobulin antibodies in diabetic pregnancies. J Endocrinol Invest 2014;37:911–5. Wang CF, Xu H, Cheng HD, Li XT. Pregnancy outcomes in euthyroid pregnant women with positive thyroid peroxidase antibodies. Chin J Birth Health Hered 2014;22:67–70. Kang SY, Wang Y, Wu L. The study of thyroid function and pregnancy outcomes in the pregnant women with gestational diabetes mellitus. Prog Obstet Gynecol 2014;23:50–2. VOL. - NO. - / - 2015
Fertility and Sterility® 11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21. 22.
Lin QL, Lin XF, Mai SM. Thyroid function and thyroid autoantibodies changes in the pregnant women of gestational diabetes millitus. Guangdong Med J 2014;35:1355–6. Li YH, Xie QP, Ma JX, Wang XJ, Shang SY. The study of thyroid function tests index changes in the pregnant with high risk of gestational diabetes mellitus. Matern Child Health Care China 2013;28:5420–1. Wang XM. The study of thyroid autoantibodies and thyroid function changes in the pregnant women with high-risk of gestational diabetes mellitus. China Med Her 2011;8:20–1. Wang H, Luo W, Liao WJ, Liu HX, Xiang B. The study of thyroid autoantibodies and thyroid function in pregnant women with increased risk of gestational diabetes mellitus. Chin J Lab Diag 2011;15:273–6. Karakosta P, Alegakis D, Georgiou V, Roumeliotaki T, Fthenou E, Vassilaki M, et al. Thyroid dysfunction and autoantibodies in early pregnancy are associated with increased risk of gestational diabetes and adverse birth outcomes. J Clin Endocrinol Metab 2012;97:4464–72. Mannisto T, Vaarasmaki M, Pouta A, Hartikainen AL, Ruokonen 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. Olivieri A, Valensise H, Magnani F, Medda E, de Angelis S, d’Archivio M, et al. High frequency of antithyroid autoantibodies in pregnant women at increased risk of gestational diabetes mellitus. Eur J Endocrinol 2000;143: 741–7. Pascual Corrales E, Andrada P, Auba M, Ruiz Zambrana A, Guillen Grima F, Salvador J, et al. Is autoimmune thyroid dysfunction a risk factor for gestational diabetes? Endocrinol Nutr 2014;61:377–81. Chen LM, Zhang Q, Si GX, Chen QS, Ye EL, Yu LC, et al. Associations between thyroid autoantibody status and abnormal pregnancy outcomes in euthyroid women. Endocrine 2014;48:924–8. Ortega-Gonzalez C, Liao-Lo A, Ramirez-Peredo J, Carino N, Lira J, Parra A. Thyroid peroxidase antibodies in Mexican-born healthy pregnant women, in women with type 2 or gestational diabetes mellitus, and in their offspring. Endocr Pract 2000;6:244–8. Montaner P, Juan L, Campos R, Gil L, Corcoy R. Is thyroid autoimmunity associated with gestational diabetes mellitus? Metabolism 2008;57:522–5. Olivieri A, Magnani F, Valensise H, d’Archivio M, Medda E, Baccarini S, et al. Thyroid autoimmunity in pregnant women at risk for GDM. Ann Ist Super Sanita 1997;33:447–50.
VOL. - NO. - / - 2015
23.
24. 25.
26.
27.
28.
29.
30. 31. 32.
33.
34.
Velkoska Nakova V, Krstevska B, Dimitrovski C, Simeonova S, Hadzi-Lega M, Serafimoski V. Prevalence of thyroid dysfunction and autoimmunity in pregnant women with gestational diabetes and diabetes type 1. Prilozi 2010;31: 51–9. Zhang XL. Clinical epidemiology research of thyroid disease during pregnancy. Beijing: Capital Medical University; 2013. Tang L. The combined effect of maternal mild hyperthyroidism and emotional symptoms to adverse pregnancy outcomes during the first and second trimester: a population-based cohort study. Anhui: Anhui Medical University; 2013. Zhao Y, Liu XQ, Su Y, Lin Y, Liu W. Analysis of thyroid function and antibody changes in gestational diabetes mellitus and pregestational diabetes mellitus. Chin J Control Endem Dis 2013;28:232–3. Agarwal MM, Dhatt GS, Punnose J, Bishawi B, Zayed R. Thyroid function abnormalities and antithyroid antibody prevalence in pregnant women at high risk for gestational diabetes mellitus. Gynecol Endocrinol 2006; 22:261–6. Negro R, Schwartz A, Gismondi R, Tinelli A, Mangieri T, Stagnaro-Green A. Thyroid antibody positivity in the first trimester of pregnancy is associated with negative pregnancy outcomes. J Clin Endocrinol Metab 2011;96: E920–4. Stroup DF, Berlin JA, Morton SC, Olkin I, Williamson GD, Rennie D, et al. Meta-analysis of Observational Studies in Epidemiology (MOOSE) Group. Meta-analysis of observational studies in epidemiology: a proposal for reporting. JAMA 2000;283:2008–12. Zhang J, Yu KF. What’s the relative risk? A method of correcting the odds ratio in cohort studies of common outcomes. JAMA 1998;280:1690–1. Miettinen O. Estimability and estimation in case-referent studies. Am J Epidemiol 1976;103:226–35. Toulis KA, Stagnaro-Green A, Negro R. Maternal subclinical hypothyroidsm and gestational diabetes mellitus: a meta-analysis. Endocr Pract 2014;20: 703–14. Vandana, Kumar A, Khatuja R, Mehta S. Thyroid dysfunction during pregnancy and in postpartum period: treatment and latest recommendations. Arch Gynecol Obstet 2014;289:1137–44. Ong GS, Hadlow NC, Brown SJ, Mun Lim E, Walsh JP. Does the thyroidstimulating hormone measured concurrently with first trimester biochemical screening tests predict adverse pregnancy outcomes occurring after 20 weeks gestation? J Clin Endocrinol Metab 2014;99:E2668–72.
7
Characteristics of the 20 studies.
Reference Wang 2014 (9)
Location
Period
Design
Definition of GDM
China
2011.9– 2013.1
Cohort
China Kang 2014 (10)
2011.1– 2011.12
Case75-g OGTT control fasting, 1 h, 2 h, R1 abnormal
Lin
China
2012.3– 2013.6
Case75-g OGTT control fasting, 1 h, 3 h, R1 abnormal
China
2008.6– 2011.6
CaseOGTT, no control detailed information
2014 (11)
VOL. - NO. - / - 2015
Li 2013 (12)
Yang. Thyroid antibodies and GDM. Fertil Steril 2015.
75-g OGTT fasting, 1 h, 2 h, R1 abnormal
Definition of thyroid dysfunction Serum TSH or FT4 out of reference range. Abbott TSH reference range 0.04–3.61 mIU/L (8–12 wk), 0.09–3.93 mIU/L (13–27 wk). Abbott FT4 reference range 12.32–21.66 pmol/L (8–12 wk), 11.19–19.14 pmol/L (13–27 wk). Roche TSH reference range 0.03–4.17 mIU/L (8–12 wk), 0.19–4.94 mIU/L (13–19 wk), 0.70–5.42 mIU/L (20–27 wk), 0.63–5.52 mIU/L (28–40 wk). Roche FT4 reference range 11.57–20.88 pmol/L (8–12 wk), 10.09–16.88 pmol/L (13–19 wk), 8.39–13.79 pmol/L (20–27 wk), 8.35–14.14 pmol/L (28–40 wk). Serum TSH out of reference range 0.3–2.5 mIU/L (20 wk) or serum FT4 out of reference range 9.01–19.05 pmol/L. Serum TSH out of reference range 0.05–5.17 mIU/L (T1), 0.39–5.22 mIU/L (T2), 0.60–6.84 mIU/L (T3) or serum FT4 out of reference range 12.91–22.35 pmol/mL (T1), 9.81–17.26 pmol/ mL (T2),12–15.71 pmol/ mL (T3). NG
Eligible subjects
Positive thyroid antibodies
Exclusion criteria
GDM
Control
Thyroid dysfunction, miscarriages, history of TD and DM
457
2,062
33.11% (TPOAb)
History of TD and DM
569
1,221
9.94% (TPOAb)
History of TD
86
120
T1 6.31%; T2 7.77%; T3 11.17% (TPOAb)
None
18
102
15% (TPOAb); 11.67% (TGAb); 20% (ThyAb)
ORIGINAL ARTICLE: ENVIRONMENT AND EPIDEMIOLOGY
8
SUPPLEMENTAL TABLE 1
VOL. - NO. - / - 2015
SUPPLEMENTAL TABLE 1 Continued.
Reference
Location
Period
Definition of GDM
Design
Definition of thyroid dysfunction
Eligible subjects Exclusion criteria
GDM
Control
Wang XM China 2011 (13)
2008.10– 2010.10
CaseOGTT, no detailed control information
NG
None
30
130
Wang H China 2011 (14)
NG
CaseOGTT, no detailed control information
None
32
133
Karakosta Greece 2012 (15)
2007.2– 2008.2
Cohort
Miscarriages
88
1053
Mannisto Finland 2010 (16)
1986– 2006
Cohort
Serum TSH out of reference range 0.34–5.6 mU/L or serum FT4 out of reference range 7.5–21.1 pmol/L Normal TSH 0.05–2.53 mIU/mL; high TSH >2.53 mIU/mL Serum TSH out of reference range 0.19–3.6 mU/L or serum FT4 out of reference range 11.96–20.5 pmol/L Serum TSH out of reference range 0.2–4 mUI/mL or serum FT4 out of reference range 8–18 pg/mL TSH R2.5 mU/mL
Olivieri Italy 2000 (17)
Corrales America 2014 (18)
Chen China 2014 (19)
Montaner Spain 2008 (21)
Case100-g OGTT, fasting, control 1 h, 2 h, 3 h, R2 abnormal
2008–2012
Cohort
2009.5– 2012.7
Cohort
O'Sullivan test, blood glucose >140 mg/dL, 50 g 1 h, 100 g 12 h Fasting >5.3 mmol/L, Serum TSH out of reference 1 h >10.0 mmol/L, range 0.09–3.47 mIU/L or 2 h >8.6 mmol/L (T1), 0.20–3.81 mIU/L (T2), 0.67–4.99 mIU/L (T3) or serum FT4 out of reference range 6–12.25 ng/L (T1), 4.30–9.74 ng/L (T2), 4.56–8.50 ng/L (T3).
1997.7.1– CaseNG 1997.12.31 control
2003.5– 2003.12
Yang. Thyroid antibodies and GDM. Fertil Steril 2015.
Cohort
50-g oral load
Serum TSH out of reference range 0.3–4.2 mU/L or serum FT4 out of reference range 9.5–25.0 pmol/L NG
Multiple pregnancy
12.5% (TPOAb); 6.88% (TGAb); 16.25% (ThyAb) 13.33% (TPOAb); 6.06% (TGAb); 15.76% (ThyAb)
15.78% (ThyAb)
50 (TPOAb); 5,713 (TPOAb); 4.50% (TPOAb); 50 (TGAb) 5,655 (TGAb) 4.50% (TGAb)
None
19
179
History of impaired carbohydrate metabolism, earlier diagnosis of thyroid dysfunction Thyroid dysfunction, history of treatments with thyroid hormones or antithyroid drugs, autoimmune diseases, congenital heart disease or heart failure, elevated serum transaminases or serum creatinine, multiple pregnancy Thyroid dysfunction, history of TD
15
41
33.93% (TPOAb); 12.5% (TGAb); 37.5% (ThyAb)
268
7,373
5.03% (TPOAb); 3.89% (TGAb); 7.04% (ThyAb)
50
50
43
576
Pre-pregnancy diabetes mellitus
15.66% (ThyAb)
45% (TPOAb)
10.02% (TPOAb)
9
Fertility and Sterility®
OrtegaMexico Gonzalez 2000 (20)
NG
100-g OGTT, fasting, 1 h, 2 h, 3 h, R2 abnormal OGTT, no detailed information
Positive thyroid antibodies
Continued.
Reference
Location
Olivieri Italy 1997 (22)
Period NG
Definition of GDM
Design
Case100-g OGTT, fasting, control 1 h, 2 h, 3 h, R2 abnormal Case100-g OGTT, fasting, control 1 h, 2 h, 4 h, R2 abnormal
Nakova Macedonia 2010 (23)
2009.5– 2009.11
Zhang China 2013 (24)
2010.10– 2012.12
Cohort
75-g OGTT, fasting, 1 h, 2 h, R1 abnormal
Tang China 2013 (25)
2008.10– 2010.10
Cohort
NG
Zhao China 2013 (26)
2010–2012
Agarwal UAE 2006 (27)
2004.1– 2004.5
Negro Italy 2011 (28)
2005
Case75-g OGTT fasting, control 1 h, 3 h, R1 abnormal Cohort 75-g OGTT, fasting, 1 h, 4 h, R1 abnormal Cohort
NG
Definition of thyroid dysfunction
Eligible subjects Exclusion criteria
Positive thyroid antibodies
GDM
Control
3
46
14.29% (ThyAb)
40
30
12.86% (TPOAb)
Multiple pregnancy, history of TD and DM, family history of TD
255
1,813
18.23% (TPOAb)
Thyroid dysfunction, multiple pregnancy History of TD
49
2,199
17.79% (TPOAb)
49
51
7% (ThyAb)
History of TD
186
69
20% (TPOAb)
Thyroid dysfunction, history of TD
135
3,458
NG
None
Serum TSH out of reference range 0.2–4.2 mIU/L or serum FT4 out of reference range 10.3–24.45 pmol/L SCH: serum TSH 3.67–10 mIU/L (T1), 3.74–10 mIU/L (T2), 4.45–10 mIU/L (T3); and FT4 in reference range 7.2–18.66 pmol/L (T1), 5.53–12.74 pmol/L (T2), 5.40–10.40 pmol/L (T3) Serum TSH or FT4 out of reference range, no detailed information Serum TSH or FT4 out of reference range, no detailed information Serum TSH out of reference range 0.3–4.32 mIU/L or serum FT4 out of reference range 9.8–18.6 pmol/L TSH R2.5 mIU/L or hyperthyroid
History of TD
6.82% (ThyAb)
Note: DM ¼ diabetes mellitus; FT4 ¼ free thyroxine; NG ¼ data were not provided in the original study; SCH ¼ subclinical hypothyroidism; T1 ¼ first trimester; T2 ¼ second trimester; T3 ¼ third trimester; TD ¼ thyroid disease; TGAb ¼ thyroglobulin antibody; ThyAb ¼ thyroid antibody; TPOAb ¼ thyroid peroxidase antibody; TSH ¼ thyroid-stimulating hormone. Yang. Thyroid antibodies and GDM. Fertil Steril 2015.
ORIGINAL ARTICLE: ENVIRONMENT AND EPIDEMIOLOGY
10
SUPPLEMENTAL TABLE 1
VOL. - NO. - / - 2015
