Preconception TSH and Pregnancy Outcomes

A Population-Based Cohort Study in 184 611 Women

Shi Chen; Xiang Zhou; Huijuan Zhu; Hongbo Yang; Fengying Gong; Linjie Wang; Man Zhang; Yu Jiang; Chengsheng Yan; Jianqiang Li; Qing Wang; Shikun Zhang; Hui Pan


Clin Endocrinol. 2017;86(6):816-824. 

In This Article

Abstract and Introduction


Background Whether subclinical hypothyroidism adversely impacts pregnancy outcomes is inconclusive, and limited data are available on the optimal TSH range in women planning pregnancy.

Objective To investigate the association between maternal preconception TSH levels and pregnancy outcomes.

Design Population-based cohort study.

Methods From 2010 to 2012, the free National Pre-pregnancy Checkups Project recruited 248 501 pairs of volunteer couples across China, of which 184 611 women, who subsequently became pregnant, were studied. Maternal TSH was measured within 6 months prior to conception. Participants were grouped according to TSH: 0·48–2·49 mIU/l (n = 133 232, 72%), 2·50–4·28 mIU/l (n = 44 239, 24%) and 4·29–10·0 mIU/l (n = 7140, 4%). Multivariable logistic regression models were used to study the association between TSH and pregnancy outcomes.

Main outcomes Pregnancy loss, gestational age (GA), delivery modes and birthweight.

Results The overall incidence of adverse pregnancy outcomes was 28·6%. Compared with TSH 0·48–2·50 mIU/l, TSH 2·50–4·29 mIU/l was associated with spontaneous abortion [aOR: 1·10, 95% confidence interval (CI): 1·03–1·18], preterm birth (aOR: 1·09, 95% CI: 1·04–1·15) and operative vaginal delivery (aOR: 1·15, 95% CI: 1·09–1·21), while TSH 4·29–10 mIU/l was correlated with spontaneous abortion (aOR: 1·15, 95% CI: 1·10–1·22), stillbirth (aOR: 1·58, 95% CI: 1·10–2·28), preterm birth (aOR: 1·20, 95% CI: 1·08–1·34), caesarean section (aOR: 1·15, 95% CI: 1·10–1·22) and large for gestational age (LGA) infants (aOR: 1·12, 95% CI: 1·04–1·21).

Conclusion Preconception TSH elevation was associated with increased risk of adverse pregnancy outcomes, even within the normal nonpregnant range.


Thyroid hormones are critical for substance metabolism, energy production, early placental formation and neurological development.[1] Maternal thyroid function undergoes dynamic changes to adapt to pregnant status, with thyroid-stimulating hormone (TSH) dropping sharply in the first trimester, and rising in subsequent trimesters. Appropriate maternal thyroid function is of great concern throughout all trimesters. Overt hyperthyroidism and hypothyroidism during pregnancy have a well-established detrimental impact on pregnant and obstetric consequences[2,3] and offspring neurological development.[4]

Subclinical hypothyroidism (SCH), defined as elevated TSH and normal free T4 (fT4) levels, with an incidence of 2–13·7%,[5] is the most common thyroid disorder during pregnancy. SCH has also been associated with adverse foeto-maternal outcomes,[6–8] although some studies have not confirmed this association.[9] Such disparity may be attributed to the difference in defining of TSH cut-offs and in the timing of TSH screening, as well as bias of subjects' enrolment and selection of end-point events.

Previous studies on European pregnant women led to the recommendation that the TSH upper limit in the first trimester should be 2·5 mIU/l, if a local trimester-specific reference range is unavailable.[1,5] Nonetheless, with more recent studies revealing a higher TSH upper limit during pregnancy, the American Thyroid Association now recommends that a cut-off of approximately 4·0 mIU/l may be applied in the first trimester.[10] In addition, among infertile women attempting to conceive, reproductive endocrinologists seemed more supportive of defining SCH as TSH >4 mIU/l, rather than >2·5 mIU/L.[11] Studies in China yielded a significantly higher TSH upper limit (4·5–5·3 mIU/l) for pregnant women.[12–14] Li et al.[13] proposed that the TSH reference range was comparable between nonpregnant women and 4–6 weeks of gestation.

Furthermore, although women with SCH planning to conceive are recommended to consider levothyroxine treatment, to achieve a TSH below 2·5 mIU/L,[1,5,10] such recommendations were based exclusively on studies that screened thyroid function during pregnancy, except for the Danish nationwide registers which contained data on thyroid status before conception.[15] The only other two small-sample size cohort studies[16,17] conducted before pregnancy concluded discrepantly with regard to maternal thyroid function and pregnancy outcomes. Other studies[18] were carried out in infertile women undergoing in vitro fertilization, mainly aiming to assess the pregnancy rate. All added to the ambiguity in the diagnosis and management of SCH among apparently euthyroid women who are planning a conception.

To the best of our knowledge, this study was the first large-scale study to investigate the association between maternal TSH levels within the 6 months before conception and the risk of adverse pregnancy outcomes in a population at low risk. The second aim was to determine whether the first-trimester specific reference range or nonpregnant reference range for TSH should be applied during preconception evaluation.