Lee et al have resurrected the free thyroxine index (FT4I) against modern analog-type (labeled analog/labeled antibody) free thyroxine (FT4) assays for diagnosis in pregnancy. This harkens back 40 years to a more innocent era, when shortcomings in thyroid hormone uptake methods were scarcely understood. All authentic FT4 assays must rigorously obey the law of mass action when measuring free hormone.

The authors insist direct analog-type FT4 results, rather than FT4I approximations, are flawed in pregnancy. In contrast, copious evidence from (1) theoretic calculation of expected FT4 values by mass action, (2) T4/thyroxine-binding globulin (TBG) ratios, (3) equilibrium dialysis (ED), and (4) 2-step FT4 assays and the analog-type 1-step assays concurs that FT4 values in the 3 trimesters of pregnancy typify the nonpregnant euthyroid range in the first trimester, with means 20–25% lower in the second and third trimesters. Evidence from modern ED-liquid chromatography/tandem mass spectrometry agrees with this.

Crucially, the thyroid hormone uptake part of FT4I does not follow the law of mass action in measuring FT4, and such methods consequently show results with a TBG dependency, with demonstrably poorer discrimination of thyroid dysfunction from normality. Analysis illuminates the invalidity of this approach and elucidates reasons for the observed deficiencies. It is piquant that the authors thereby support an assay with fundamental faults against a valid method, apparently because the faulty approach artificially lifts results from later trimesters into the comfort zone of the normal range (through the artifact of its “TBG effect”). If they believe their stance is correct, they must explain cogently why mass action should not be obeyed.

The authors maintain that modern FT4 assays are affected directly by serum T4-binding proteins. Studies that show this are also invalid, because they were conducted outside the regions of where FT4 can be measured, when artificial protein effects consequently appear.

We predict thyroid dysfunction will be least well served using T4 assays, which are marginally better with FT4I and distinctly better with the use of trimester-specific reference ranges for FT4, with the disparaged modern 1-step assays. An FT4 assay must work correctly in all conditions, and it is unscientific to use “pick and mix” to bolster preconceptions of what the results should be.