HormoneBayes: A novel Bayesian framework for the analysis of pulsatile hormone dynamics

Area of Science:

• Endocrinology
• Computational Biology
• Reproductive Medicine

Background:

• The hypothalamus regulates reproductive hormones via pulsatile gonadotropin releasing hormone (GnRH) secretion.
• Altered GnRH pulsatility is linked to reproductive disorders like polycystic ovary syndrome (PCOS) and hypothalamic amenorrhea (HA).
• Luteinizing hormone (LH) is a surrogate marker for GnRH, but its pulsatility analysis is resource-intensive and lacks accessible software.

Purpose of the Study:

• To introduce hormoneBayes, a novel open-access Bayesian framework for analyzing serial LH measurements.
• To enable reliable assessment of LH pulsatility and hypothalamic GnRH dynamics.
• To provide clinicians with an accessible tool for reproductive hormone dysfunction diagnosis.

Main Methods:

• Utilized parsimonious models to simulate hypothalamic signals driving LH dynamics.
• Employed state-of-the-art sequential Monte-Carlo methods for parameter inference.
• Applied the framework to analyze serial LH measurements for pulsatility assessment.

Main Results:

• hormoneBayes accurately estimates key pulse parameters (inter-pulse interval, secretion/clearance rates) and identifies LH pulses.
• The method distinguishes LH pulsatility patterns across various clinical contexts (healthy, menopause, HA, PCOS) in men and women.
• Quantified estimation of uncertainty is a key advantage of the hormoneBayes framework.

Conclusions:

• hormoneBayes offers a reliable, open-access tool for analyzing LH pulsatility.
• The framework can differentiate LH pulsatility in diverse reproductive health and disease states.
• This tool has potential for real-time hormone monitoring and personalized, data-driven clinical care in reproductive endocrinology.