At the heart of this work, Dr Tinatin Kutchukhidze, from the University of Oxford and New Anglia University, combines a
AI hormonal patch for testosterone in men and an AI score, theEndocrine Rhythm Integrity (ERI)for female cycles. His idea is simple to state and radical in practice: it is not only the quantity of hormones that counts, but their choreography over time. By combining her studies, she shows that men and women who are “normal” on paper nevertheless present broken hormonal rhythms, associated with impaired fertility. Enough to lay the foundations of fertility medicine focused on rhythms, and not just on dosages.
Succeed in elucidating the 15 to 30% of unexplained infertility
When a couple wanting a child hears that “everything is normal” when the pregnancy does not come, the frustration is immense. This situation concerns almost 15 to 30% of couplesclassified in unexplained infertility after reassuring classic hormonal assessments.
- For men suffering from infertility or hypogonadism, characterized by low testosterone levels, the usual tests consist of measuring serum testosterone in the morning;
- For women, fertility assessments focus on the analysis of the menstrual cycle and reproductive hormones, such as LH (luteinizing hormone), FSH (follicle-stimulating hormone), estradiol and progesterone.
However, these tests only give a fixed picture, whereas sexual hormones are very dynamic and follow a circadian rhythm, with regulated fluctuations throughout the day. A European team has just presented, at the 28th European Congress of Endocrinology in Prague, connected skin patches driven by AI capable of revealing these hidden disorders.
An AI-boosted hormonal patch tracks male rhythms
Dr Tinatin Kutchukhidze, from the University of Oxford and New Anglia University, followed 102 men aged 22 to 38, with normal morning total testosterone levels (12-35 nmol/l), with or without infertility or symptoms of hypogonadism. All received a transdermal patch developed by the team capable of measuring testosterone every 15 minutes for 96 hours, via a mobile app analyzed by AI.
They found that men with symptoms had significantly disrupted testosterone rhythms (with much lower average diurnal amplitude and shifted peaks), despite having normal testosterone levels on standard laboratory tests. Furthermore, these revealed rhythm abnormalities were associated with a decrease in sperm concentration and symptoms of androgen deficiency.
“For the first time, we were able to track androgen trends in real time over several days using an innovative, non-invasive, AI-driven continuous testosterone monitoring patch compatible with Android and iPhone mobile devices.said Dr. Kutchukhidze. “Previous research suggested that a normal morning testosterone level was sufficient to rule out clinically significant androgen deficiency. However, our results challenge this hypothesis by demonstrating that men with normal serum testosterone levels can nevertheless exhibit significant disruptions in hormonal rhythm associated with reproductive dysfunctions.
Rhythm indices calculated by the AI predicted these subclinical dysfunctions with high accuracy (AUC 0.87 versus 0.61 for isolated testosterone). In practice, this amounts to saying that a “normal” dosage in the morning can mask an androgen timing disorder which affects sperm quality and libido, which could explain why 25 to 30% of symptomatic men today maintain a reassuring assessment.
AI reveals “falsely normal” female cycles
On the women’s side, the team followed 312 participants aged 18 to 22, all with regular cycles of 24 to 35 days, divided between 162 fertile women and 150 with unexplained infertility. Researchers developed Endocrine Rhythm Integrity (an IRE score), an artificial intelligence-based metric, to analyze data related to key reproductive hormones during the luteal phase, basal temperature, heart rate and sleep cycles. The results showed that women with unexplained infertility had lower IRE scores, even with normal hormone levels, which was predictive of infertility. Lower IRE scores were also associated with a higher incidence of implantation failure.
“Our study reveals that a woman can have a seemingly healthy menstrual cycle and normal hormone levels, yet suffer from latent endocrine dysfunction that affects her fertility.”explains Dr. Kutchukhidze. “Rather than analyzing hormone levels in isolation, ERI assesses whether reproductive hormones change in the appropriate pattern, at the appropriate time, and in relation to each other throughout the menstrual cycle.”.
The results thus suggest that female and male endocrine disorders may not just be disorders of hormone quantity, but rather disorders of hormonal timing, synchronization and biological rhythm.
What These AI Hormone Patches Change for Unexplained Infertility
For a couple already labeled with unexplained infertility, such a portable hormonal chronodiagnostic could, in the long term, help target the best window for scheduled intercourse, insemination or embryo transfer, or even guide male care when traditional dosages are reassuring.
“Our AI-based rhythm analyzes were found to perform significantly better than conventional tests at identifying subclinical reproductive dysfunctions, suggesting that endocrine disorders in both women and men may not simply be disorders of hormone quantity, but rather disorders of hormonal timing, timing, and biological rhythm.”said Dr. Kutchukhidze.
For the moment, this work remains clinical trials presented at conferences, carried out on young and limited populations, and Dr. Kutchukhidze plans to extend them to larger and more diverse groups. The researcher is already planning ahead: “Our goal is to evolve fertility care toward predictive, rhythmic reproductive medicine, enabling clinicians to identify dysfunction earlier, personalize interventions, and improve outcomes even before infertility becomes clinically evident.“.
She also imagines “portable hormonal chronodiagnostics” useful in personalized endocrinology and transgender medicine, where real-time monitoring of hormonal fluctuations would allow more precise, adaptive and patient-centered care in various clinical contexts.