Diurnal Within-Person Coupling Between Testosterone and Cortisol in Healthy Men: Evidence of Positive and Bidirectional Time-Lagged Associations Using a Continuous-Time Model

Objective

The hypothalamic-pituitary-gonadal (HPG) and -adrenal (HPA) axes are traditional viewed as mutually inhibitory systems. However, several diurnal studies have reported positive within-person testosterone and cortisol relationships, as evidence of facilitative processes, but with some constraints (e.g., low-frequency sampling, use of static longitudinal models). Continuous-time (CT) models can help illuminate testosterone-cortisol “coupling” by testing for bidirectional, cross-lagged effects.

Methods

This study investigated diurnal testosterone and cortisol coupling in healthy males (n = 30) using high-frequency sampling protocols. Participants self-collected saliva at work or home using one of three sampling formats; every 10 mins for 9 h, 15 mins for 8 h, and 30 mins for 10 h. After detrending, daily within-person fluctuations in testosterone and cortisol concentration were modeled in a CT framework.

Results

Autoregressive effects for each hormone indicated moderate stability over a shorter period (~6 mins), as a mean-reverting process, and higher stability over longer time periods. Cross-lagged effects were also demonstrated, with testosterone showing a positive relationship to cortisol (.12 within-person standardized effect) and cortisol to testosterone (.08). Both linkages followed a non-linear trajectory, rising in strength from a zero-time lag to peak with a lag of ~8 mins before dissipation beyond this period.

Conclusion

We verified reports of positive within-person coupling between testosterone and cortisol across the day in healthy men. Added novelty comes from bidirectional and time-lagged associations on hormonal pulses, although the effect sizes were small. Hence, we offer a more nuanced understanding of HPG and HPA crosstalk within a CT framework.