This letter is regarding the recent publication of ‘Assessment of thyroid function in children, adults and pregnant and lactating women after long-term salt iodisation measurements’ by Su et al. ⁽ Reference Su, Li and Liu ^{1 )}. The findings presented by the authors are both relevant and interesting to the iodine research conducted in China. The cross-sectional study covered six different provinces located in the northern, central and southern regions of China; however, some regions such as Xinjiang (northwest region), Qinghai (northwest region) and Yunnan (southwest region) were not included. For example, Xinjiang is classified as iodine excessive region. Overall, the findings by the authors have provided some valuable information especially the use of thyroglobulin (Tg) to assess iodine status in children, adults, pregnant and lactating women living in China.

I would like to highlight the use of Tg by the authors in the assessment of iodine status⁽ Reference Su, Li and Liu ^{1 )}. There are limited data on the use of Tg to assess iodine status in populations with adequate (median urinary iodine concentration (UIC) 100–199 µg/l), above requirements (median UIC 200–299 µg/l) and excessive iodine levels (median UIC ≥300 µg/l)⁽ Reference Ma and Skeaff ^{2 ,} Reference Ma and Skeaff ^{3 )}. Most of the data on Tg as a promising biomarker of iodine status are derived from iodine-deficient population (median UIC <100 µg/l). Therefore, Tg has been proposed to be a useful biomarker to assess iodine deficiency in populations based on these studies.

A median Tg cut-off of <13 µg/l has been proposed to indicate iodine sufficiency in children and possibly in adults⁽ Reference Ma and Skeaff ^{3 ,} Reference Ma, Venn and Manning ^{4 )}. However, there are no further specific Tg cut-offs to determine the adequate (median UIC 100–199 µg/l), above requirements (median UIC 200–299 µg/l) and excessive iodine levels (median UIC ≥300 µg/l). In this study, Su et al. reported that children and adults with a median UIC of 200–299 µg/l had a significantly lower median Tg concentration than children and adults with a median UIC of 100–199 µg/l (children: 2·79 v. 2·91 µg/l; adults: 5·56 v. 8·12 µg/l), although both groups were considered to have good iodine status and normal thyroid function (i.e. thyroid-stimulating hormone (TSH) and free thyroxine (FT4) of both groups were remained within normal range⁽ Reference Ma and Skeaff ^{2 )}). Therefore, it is suggested that Tg might be a sensitive biomarker to assess iodine sufficiency in populations⁽ Reference Ma, Venn and Manning ^{4 )}. In addition, Tg is a longer measure term of iodine status than UIC, which is affected by hydration status and high intra-individual variation⁽ Reference Ma and Skeaff ^{2 ,} Reference Ma, Venn and Manning ^{4 )}. Although Su et al. collected three spot urine samples to reduce the intra-individual variation in UIC⁽ Reference Ma and Skeaff ^{2 )}, at least ten spot urine samples are needed to reliably assess individual iodine status using UIC⁽ Reference König, Andersson and Hotz ^{5 )}. However, even with the collection of multiple spot urine samples (>10), the high intra-individual variation still cannot be eliminated⁽ Reference König, Andersson and Hotz ^{5 ,} Reference Ma, Venn and Manning ^{6 )}. Moreover, there is no consensus on the timing and period of collecting multiple spot urine samples⁽ Reference König, Andersson and Hotz ^{5 ,} Reference Ma, Venn and Manning ^{6 )}. It is unclear if an average UIC value obtained from three spot urine samples over a 10-d period could be comparable to the average UIC value from three spot urine samples over 30-d period.

One of the limitations of using Tg to assess iodine status is that Tg can be confounded by the presence of thyroglobulin antibody (TgAb), which can give falsely high or low Tg values⁽ Reference Ma and Skeaff ^{2 )}. Su et al. reported that adults had a relatively higher prevalence of TgAb than children (12·4 v. 2·8 %)⁽ Reference Ma and Skeaff ^{2 )}, suggesting that Tg might not be a suitable to assess iodine status of adult populations because Tg value obtained might be confounded by the presence of TgAb⁽ Reference Ma and Skeaff ^{3 )}. In addition, if the measurement of TgAb is required before the determination of Tg, the additional test of TgAb would increase the cost of using Tg to assess iodine status⁽ Reference Ma and Skeaff ^{3 ,} Reference Ma, Venn and Manning ^{4 )}. Therefore, if not corrected, this issue will affect future publications that measure Tg in populations.