Key points are not available for this paper at this time.
Introduction There is a complex physiological interaction between the HPT axis and the HPG axis. However, previous studies have primarily focused on individuals with thyroid dysfunction, and our understanding of the relationship between these two axes in individuals with normal thyroid function remains limited. Methods This study based on multi-center cross-sectional data, included 1,025 healthy participants with no known endocrine disorders. It systematically assessed the association between thyroid hormones and sex hormones, with a focus on analyzing effects stratified by gender and age. Results The results showed that FT4 was positively correlated with T in men (Std β = 0.448, P = 0.002), while FT3 and TSH did not show a significant effect. No significant association was observed in the overall female population, but a clear trend emerged when the data were stratified by age. In the 18-year-old group, thyroid hormones and gonadotropins showed a negative correlation (e.g., FT3–LH: r = −0.372, P 0.001). In the 18–40-year-old group, a positive correlation was observed (FT4–FSH: r = 0.125, P = 0.001), whereas this correlation disappeared in the 40-year-old group. In women of childbearing age, multivariate analysis further confirmed the association between FT4 and FSH (Std β = 0.118, P = 0.005) and between TSH and LH (Std β = 0.136, P 0.001). Discussion The study findings indicate that the association between thyroid hormones and sex hormones exhibits significant gender differences and is age-dependent. Among women, the direction of this association varies. This suggests that pooled analyses may mask the true association due to population heterogeneity, thereby underestimating the actual association. This study reveals the potential role of normal fluctuations in thyroid hormone levels in the regulation of reproductive endocrinology and provides new epidemiological evidence for research into the mechanisms underlying related reproductive endocrine disorders and for the early identification of risks.
Yang et al. (Fri,) studied this question.