The TSH reference interval was defined by 95% confidence limits of the log-transformed values of at least 120 rigorously screened normal euthyroid volunteers, according to the National Academy of Clinical Biochemistry (NACB). Normal euthyroid subjects were defined as those with no detectable thyroid peroxidase antibody (TPOAb) or thyroglobulin antibody, no personal or family history of thyroid dysfunction, no visible or palpable goiter, and no use of any medications affecting thyroid function except estrogen [8]. However, the NACB guideline does not consider the iodine intake of the population when defining the reference population for TSH measurements. Nonetheless, many guidelines recommend using populations with adequate iodine intake to define TSH reference ranges [1910]. Both iodine deficiency and iodine excess can be associated with reduced thyroid function [11], and these possibilities should be considered as an important parameter for defining population-based trimester-specific reference ranges of TSH.

In this issue of Endocrinology and Metabolism, Castillo et al. [12] determined the thyroid hormone reference range of pregnant women in Chile. They recruited 1,022 pregnant women in their first trimester, checked their thyroid function and TPOAb status, and used NACB guidelines to select 670 subjects as the reference population. The reference range of TSH was 0.13 to 5.37 mIU/L, with a median level of 1.88 mIU/L. The reference upper limit proposed by the manufacturer of the assay for TSH for non-pregnant women was 4.2 mIU/L and the upper limit suggested by the 2011 ATA guideline was 2.5 mIU/L. Therefore, the prevalence of subclinical and/or overt hypothyroidism was 34.8%, 10.4%, or 5.7% according to the reference range suggested by the 2011 ATA guideline, the manufacturer's reference, and the present study, respectively (shown in Fig. 3 of Castillo et al. [12]). This discrepancy means that the number of pregnant women to be treated by levothyroxine varies significantly according to the choice of the TSH reference range. Levothyroxine is a quite safe medication for pregnant women, and some clinicians still follow the 2011 ATA guideline, which defines the TSH reference as 2.5 mIU/L due to the fear of undertreatment. However, several recent studies have repeatedly shown that a TSH level of 2.5 mIU/L, or even the kit manufacturer's reference range for TSH, might be too strict to be used as an indicator for levothyroxine treatment, because no definitive benefit of levothyroxine treatment was reported using those TSH reference values.

Castillo et al. [12] also evaluated the iodine status of pregnant women in Chile. They stated that the Chilean universal salt iodine fortification program started in 1979, and required each kilogram of salt to be fortified with 100 µg of iodine (100 ppm). In 2000, very high levels of iodine were found in school-aged children, so the level of iodine was reduced to 40 ppm. The median urinary iodine concentration was 173.45 µg/L (interquartile range, 108.11 to 249.35), reflecting adequate iodine status according to World Health Organization recommendations [13]. Nonetheless, some proportion of the population still showed iodine deficiency or more than adequate iodine status, as shown in Fig. 2 of the paper by Castillo et al. [12]. Those historical changes in the iodine fortification program and/or current iodine status might explain why the population of their study showed a right-shifted distribution of serum TSH during pregnancy.

It is crucial to establish population-based trimester-specific reference ranges to identify and treat hypothyroidism in pregnant women. However, the effects of iodine intake and the optimal level of during pregnancy remain to be determined.