300 mg *oral* progesterone nightly during luteal phase of cycle...
These are the first randomized controlled trial data to show that treatment with luteal phase equivalent doses of oral micronized progesterone is associated with a significant increase in Free T4 values.
This 12-week randomized controlled progesterone trial documented that Free T4 levels were significantly elevated in progesterone-treated compared with placebo-treated women. There also was a trend, as previously documented,[15, 16]towards lower TSH levels although there was no TSH-Free T4 interaction.
The absolute increase in Free T4 on progesterone therapy (+2.5 pmol/l; 95% CI: 1.9–3.0) is greater than the longitudinal 13-year, population-based normal within-person change of 0.04 pmol/l (95% CI: −0.16, 0.24).
Likewise, although small, the adjusted mean difference in the Free T4 increase between progesterone and placebo of 0.8 pmol/l is also greater than that within-person 0.04 pmol/l longitudinal change.
Given that the coefficient of variation in the FreeT4 assay is three per cent and the mean Free T4 level is 13.7 pmol/l, the observed adjusted change is double the potential maximal analytical variability of 0.41 pmol/l. These data, therefore, suggest that the increase in Free T4 during oral micronized progesterone therapy compared with placebo therapy is neither likely to be due to endogenous variance nor is it likely related to lack of analytical precision.
The observation that there is a nonsignificant trend towards lower TSH values during progesterone therapy is consistent with results from a smaller and shorter controlled trial in which TSH was a significant 24% lower during progesterone than during placebo treatment. In that study, Free T4 did not change and Free T3 was not reported. It is also consistent with data showing a trend towards lower night-time TSH values during the luteal phase.
This trial is also the first to quantitatively examine a potential relationship between thyroid function and hot flushes and night sweats—no important relationships were observed. In a prospective observational Norwegian study, however, in 57 initially regularly menstruating women aged 51, as women became postmenopausal, higher TSH levels were noted in those with VMS compared with those without.
Our VMS–thyroid analysis ideally should have included women with minimal or no vasomotor symptoms. As ours was a treatment trial, women were required to have treatment-requiring, troublesome VMS (at least five mild, or fewer more intense, daytime hot flushes or at least one night sweat per week of sufficient intensity to awaken them). Further study of potential thyroid–VMS interactions is needed.
The strengths of this study are its prospective, randomized placebo-controlled design, and that blinding was maintained throughout thyroid hormone analysis. This trial also examined all three major hormones of the thyroid axis: TSH, Free T4 and Free T3. These data are limited, however, as a post hoc analysis of sera from a trial designed and registered for other purposes. It is puzzling that placebo as well as progesterone caused an increase in Free T4. This increase was smaller than that shown on progesterone, so that analysis of covariance documented a significantly greater Free T4 increase on progesterone therapy. Although it is regrettable that sera from more women in the primary trial were unavailable for thyroid hormone analysis, this should not bias results.
That we did not measure TBG may also be perceived as detrimental, however, the literature suggested no relationship to us; also, albumen levels did not change on progesterone therapy in the cardiovascular outcome analysis from the same trial (J.C. Prior; unpub. data). It would also perhaps have been ideal to assess deiodinase enzyme activity, but these analyses were not available to us.
Whether or not the observed relationship between progesterone therapy and increased Free T4 levels has physiological importance is currently unknown. It is possible that progesterone plays a role in preventing inappropriately low Free T4 levels during pregnancy. It is also possible that Free T4 mediates the progesterone-associated higher basal temperature and the increased luteal phase energy requirements.
Cigarette use and body mass index were associated with thyroid function in a recent large cohort study. Given that the women participating in this trial were carefully screened to be nonsmokers without diabetes or hypertension and to have normal cardiovascular function, this study needs to be repeated in an unselected community cohort of postmenopausal women.
Finally, a larger randomized controlled trial of progesterone vs placebo in postmenopausal women without VMS or sleep disturbances is needed to confirm the increase in Free T4 shown here.
In summary, these data from a blinded and randomized controlled 12-week progesterone treatment trial in healthy postmenopausal women with hot flushes and night sweats document for the first time that progesterone therapy appears to cause an increase in Free T4. The clinical importance of this observation remains to be determined.