This AI model claims it can predict diseases using data from just one night of sleep

What happened: Researchers have introduced SleepFM, an AI model that analyses a single night of sleep data to predict disease risks.

Why it matters now: The model suggests early health warning signs may appear during sleep, before symptoms are visible.

What changes for people: Sleep data from wearables could soon play a bigger role in preventive healthcare.

Who is affected: Patients, doctors, health-tech companies, and users of sleep-tracking devices.

A new artificial intelligence model called SleepFM is drawing attention for its ability to predict potential diseases using data from just one night of sleep. The development points to a future where routine sleep tracking could help flag health risks long before a doctor’s visit.

While AI-driven disease prediction is not new, SleepFM stands out for focusing entirely on sleep signals, turning overnight body patterns into meaningful medical insights.

How SleepFM works

Traditional medical assessments rely on symptoms, physical exams, and patient history. SleepFM takes a different approach by analysing physiological changes during sleep, a time when the body undergoes critical recovery and regulation processes.

According to researchers, sleep captures subtle signals related to:

Heart rate and variability

Breathing patterns

Sleep stages and disruptions

Autonomic nervous system activity

These markers, when processed through AI, may reveal early indicators of disease risk.

Why sleep is becoming a health goldmine

Sleep has long been linked to conditions such as cardiovascular disease, diabetes, neurological disorders, and mental health issues. What SleepFM highlights is the possibility that even a single night’s data could carry enough information to flag concerns.

Experts say this could be especially valuable for:

Early detection and prevention

Monitoring chronic conditions

Reducing dependence on invasive tests

What this could mean for healthcare

If validated at scale, models like SleepFM could:

Support doctors with data-driven risk assessments

Integrate with consumer wearables and home monitoring

Shift healthcare toward predictive and preventive care

However, specialists caution that such tools are not diagnostic on their own and must complement clinical judgment, not replace it.

One night of sleep as a health signal could redefine how medicine understands the body’s early warnings.