Study design and participants
This cohort study recruited pregnant women in their first trimester who visited the Beijing Obstetrics and Gynecology Hospital, Capital Medical University between January 2018 and May 2020. All participants answered a questionnaire during early pregnancy(6-13+6 weeks of gestation) about demographic and obstetric characteristics (age, weight, parity, and history of adverse pregnancy outcomes), history of thyroid disease before pregnancy, history of hypertension, and history of diabetes. Written consent has been obtained from each patient or subject after full explanation of the purpose and nature of all procedures used. We continued to follow up these women through the middle and third trimesters of pregnancy until delivery. A total of 20365 women who had complete thyroid examination data and lipid data in the first and third trimesters of pregnancy( 6-13+6 weeks of gestation and 28-33+6 weeks of gestation) respectively were enrolled in this study.
We excluded pregnant women with TSH levels < 0.59 mIU/L (n = 3001), mild TSH elevation(TSH concentration between 2.5 and 4.0 mIU/L, n = 2414) and those who had thyroid disease before pregnancy (including thyroid cancer, thyroid nodule, and Hashimoto thyroiditis; n = 328). When all the women had given birth, we also excluded those who had twin or multiple pregnancies (n = 541). Thus, ultimately, 14081 pregnant women were included in this cohort study (Figure 1).
We grouped the women as follows: SCH treatment group (treatment begin at 6-13+6 weeks of gestation, n = 319), SCH non-intervention group (n = 103) and normal thyroid function group (n = 13659).
Ethical considerations
Our study protocol was approved by the Ethics Committee of Beijing Obstetrics and Gynecology Hospital, Capital Medical University (Review approval number: 2018-KY-003-01).
Definitions
SCH was defined as a normal FT4 level with TSH elevation (4.0–10 mIU/L). The reference values for the normal FT4 range during the first trimester were 11.8–18.4 pmol/L, and these values were determined using Roche Modular Analytics E170 (Roche Diagnostics, Mannheim, Germany). Electrochemiluminescence immunoassays (ADVIA Centaur XP, Siemens Healthcare Diagnostics, Tarrytown, NY, USA) were used to measure the serum FT4 and TSH concentrations. The methods and kits used for the FT4, TSH, and TPOAb tests in our hospital remained the same throughout the study period.
Serum lipids measurement
All pregnant women collected venous blood samples after overnight fasting. The concentrations of TC, TG, HDL-C and LDL-C were determined for each sample. Automatic biochemical analyser (AU5400, Beckman, US) were used to measure TC, TG, HDL-C and LDL-C detection kits (Beckman, US). All measurements were measured using continuous monitoring methods with appropriate quality control prior to measurement.
Adverse pregnancy outcomes
Adverse pregnancy outcomes included preterm delivery (defined as birth before 37 weeks of gestation), , gestational diabetes mellitus (screened for at 24–28 weeks, and diagnosed according to the criteria of the International Association of Diabetes and Pregnancy Study Groups, which require a 75-g oral glucose tolerance test and cut-off values of >5.1 mmol/L, >10.0 mmol/L, and >8.5 mmol/L for fasting blood glucose, blood glucose at 1 h after sugar intake, and blood glucose at 2 h after sugar intake, respectively), gestational hypertension and macrosomia (≥4000 g).
Statistical analysis
All statistical analyses were performed using the SPSS, version 24.0 (SPSS Inc., Chicago, IL, USA). In order to observe the normal trend of serum lipid levels during the first and third trimesters of pregnancy, we excluded those pregnant women with overweight/obesity, gestational hypertension, GDM, preterm birth and macrosomia. Then, using appropriate percentiles and median to descriptive analyses of TC, TG, LDL-C and HDL-C levels for this group (excluded adverse pregnancy outcomes) in first and third trimester of pregnancy. Finally, to establish the lipid reference ranges suitable for the first and third trimesters of pregnancy. Categorical variables were presented as frequency (percentage), and continuous variables were presented as mean (standard deviation [SD]) or median (interquartile range), as appropriate. Analysis of variance was used for normally distributed continuous variables in the multiple subgroup analyses. The Kruskal-Wallis test was used to evaluate continuous variables with non-normal distributions. To compare categorical variables, we used the chi square test or Fisher exact test. The Bonferroni correction was applied for pairwise comparisons. To evaluate the effect of L-T4 treatment on SCH with hyperlipidemia, we used generalized estimation equation. A P value < 0.05 or a P value < 0.05/n in the Bonferroni corrections was considered statistically significant.