Reference Intervals of Thyroid Hormones During Pregnancy Established by Three Statistical Methods.

Abstract

Monitoring the thyroid hormones during pregnancy is of great importance for fetal growth and development. Throughout the whole pregnancy, there is constant fluctuation in the thyroid hormone reference intervals (RIs). The purpose of this study is to determine method- and trimester-specific RIs for thyroid-stimulating hormone, free thyroxine, and free triiodothyronine in pregnant women in China. In this study, 2,167 women with normal pregnancies (first trimester, n = 299; second trimester, n = 1,032; third trimester, n = 836) and 4,231 healthy nonpregnant women subjects were recruited. Serum thyroid-stimulating hormone (TSH), free thyroxine (fT4), and free triiodothyronine (fT3) concentrations were measured using electrochemiluminescence immunoassays in Abbott Alinity i analyzer. Following the exclusion of outliers, the RIs were determined using three statistical techniques: the non-parametric method, the Hoffmann method, and the Q-Q plot method. Pregnant women's levels of these three thyroid hormones differ markedly from those of healthy non-pregnant women. In addition, the concentrations of these three hormones change considerably over the course of the three phases of pregnancy. The Q-Q plot method offered more comparable RIs with the non-parametric method in healthy non-pregnant women than the Hoffmann method. Three statistical techniques were used to construct the trimester-specific RIs of thyroid hormones in pregnant women, and there was little difference between them all. The RIs by the non-parametric and Q-Q plot methods indicated closer RIs, and the RIs by the Hoffmann approach were higher and wider than those by the other two methods. For thyroid hormones, trimester-specific RIs are required. The RIs determined by non-parametric and Q⁃Q plot indirect calculation could be used as the alternative method.