Ethnic differences in first-trimester thyroid reference intervals.

Thyroid disease is common in women of reproductive age, and the physiological changes that occur during pregnancy complicate the diagnosis of thyroid dysfunction. Trimesterand methodspecific reference intervals (RIs) for thyroid function tests (TFTs) are necessary and should be determined with thyroid autoantibody– negative individuals (1 ). Ethnic differences in TFT results exist among nonpregnant individuals, but the data on ethnic differences in thyroid function during pregnancy are limited (2–5 ). We used the Abbott ARCHITECT i2000SR analyzer to determine RIs for TFTs by ethnic group for the first trimester of pregnancy. Serum samples submitted between 10 and 13 weeks gestation were collected in the US from 540 Asians, 549 blacks, 601 Hispanics, and 878 whites (age, 15– 46 years). Gestational ages were provided by the ordering physicians. Iodine status and medical histories were unknown, and only 1 sample was available from each individual. This study was approved by the University of Utah Institutional Review Board. The numbers of thyroid autoantibody–negative individuals were 174, 579, 798, and 626 for gestational weeks 10, 11, 12, and 13, respectively. Samples from 129 self-reported healthy, nonpregnant adult individuals were obtained from 5 Asians, 2 blacks, 2 Hispanics, 114 whites, and 6 individuals of other ethnic backgrounds (63 females and 66 males, 18 – 62 years of age). Samples were thawed, mixed, centrifuged at 2095g for 10 min, and checked for clots before analysis. The following tests were performed on the ARCHITECT i2000SR: thyrotropin (TSH), free thyroxine (FT4), total thyroxine (TT4), free triiodothyronine (FT3), total triiodothyronine (TT3), thyroglobulin autoantibodies, thyroid peroxidase autoantibodies, and T-uptake (measures the total binding capacity of thyroxine-binding proteins for thyroid hormones). The free thyroxine index (FTI) and the thyroid hormone binding ratio (THBR) were calculated per the package insert as TT4/T-uptake and 1/Tuptake, respectively. The establishment of RIs and the analysis of ethnic differences were performed as previously described (4 ). We used autoantibody-negative samples for the nonparametric determination of RIs (central 95%) for both pregnant and nonpregnant individuals. Individuals with nonpathologic TSH values were used to determine the RIs for FT4, FTI, TT4, THBR, FT3, and TT3. We used EP Evaluator Release 8 software (Data Innovations) to determine RIs. GraphPad Prism (version 5.03; GraphPad Software) was used for linear regression analysis, the unpaired t-test, the 2 test, and the Fisher exact test. We observed numerous ethnic differences for the established firsttrimester RI data (Table 1), in contrast to our findings for the second trimester (4 ). The RI for an ethnic group was considered significantly different if it did not fall within the 95% CI of the reference limits for the combined group. For every analyte, both the lower and upper reference limits for at least 1 ethnic group were significantly different from those for all ethnic groups combined. It is difficult to compare the ethnic differences we observed with those established by other studies, because RIs can be method dependent and are affected by the gestational weeks included. The use of thyroid autoantibody–negative individuals with a TSH value within the RI potentially narrows the RIs for other analytes. RIs for the other TFTs that were established by excluding only thyroid autoantibody– positive individuals were compared with RIs that were established with autoantibody-negative individuals with TSH results within the RI. No significant differences in lower reference limits were observed for any ethnic group; however, the upper reference limits for FT4, TT4, FTI, and FT3 were significantly higher for black individuals when all antibody-negative individuals were included. The lower reference limits for TSH were significantly lower in the first-trimester women compared with nonpregnant adults (Table 1). TSH suppression during pregnancy is associated with high concentrations of human chorionic gonadotropin subunit ( -hCG) (1 ). All samples with TSH concentrations 0.35 mIU/L were tested for -hCG on the ARCHITECT i2000SR. Of the 282 samples tested, 26 individuals had -hCG concentrations 200 000 IU/L (7 Asian, 12 black, 2 Hispanic, and 5 white individuals). Exclusion of these individuals with very high -hCG concentrations did not significantly alter the lower reference limit for TSH. RIs determined by gestational week for all groups combined showed an upward trend for TT4 and a downward trend for FTI, in that the slopes were significantly different from zero by linear regression (P 0.04 and 0.02, respectively). All other analytes showed no significant trends. RIs were also determined for nonpreg1 Nonstandard abbreviations: RI, reference interval; TFT, thyroid function test; TSH, thyrotropin; FT4, free thyroxine; TT4, total thyroxine; FT3, free triiodothyronine; TT3, total triiodothyronine; T-uptake, measurement of total binding capacity of thyroxine-binding proteins for thyroid hormones; FTI, free thyroxine index; THBR, thyroid hormone binding ratio; -hCG, human chorionic gonadotropin subunit. Clinical Chemistry 57:6 000 – 000 (2011) Letters to the Editor