AIM

We aimed to investigate TT4 physiological aspects and associations with clinical end-points.

BACKGROUND

Total T4 (TT4) has been suggested as a marker for maternal thyroid function during pregnancy because as compared to FT4 (i) TT4 measurement is not affected by binding protein interference, (ii) TT4 is considered to be more stable from the second trimester onwards, and (iii) TT4 better reflects changes in the hypothalamic-pituitary-thyroid axis. However, this is based on data from small studies, and, more importantly, it is unknown whether TT4 is associated with adverse pregnancy or child outcomes.

METHODS

We selected 5647 mother-child pairs from a large population-based prospective cohort with data on maternal TSH, FT4 and TT4 during early pregnancy (median 13·2 weeks, 95% range 9·8-17·6). We used multivariable (non)linear and logistic regression models to study the association of maternal TT4 with pre-eclampsia, premature delivery, birthweight and offspring IQ and compare the results with previously obtained results for FT4.

RESULTS

The change of mean TT4 levels was 27·5% compared to 20·2% for FT4. There was a log-linear association of TT4 and FT4 with TSH, but the explained variability of TSH was much lower for TT4 than for FT4 (R-squared TT4: 2·5% vs 8·0% for FT4). In contrast to FT4, there was no independent association of maternal TT4 with pre-eclampsia, premature delivery, birthweight or offspring IQ.

CONCLUSION

Maternal TT4 levels are highly variable in the first half of pregnancy and are poorly related to maternal TSH. This study shows that maternal TT4 levels are either not associated, or not better associated as compared to FT4, with adverse pregnancy or child outcomes. This suggests that the maternal TT4 is inferior to FT4 in the assessment of maternal thyroid function during the first half of pregnancy.