https://www.nature.com/articles/s41398-026-04183-3
study suggests that chronic insomnia is linked to disruption of the body clock, visible in behavior, hormones and molecules. The effect is strongest in the short-sleep subtype, meaning under 6 hours measured by polysomnography.
Key findings:
- Cortisol was higher in the evening (9:30 PM) in insomnia patients, especially in the short-sleep subtype. This fits the idea of a hyperarousal state, and the less people slept, the higher their cortisol.
- Body temperature was lower in the morning, at both the wrist and the armpit, which points to a misaligned circadian rhythm.
- Clock genes in the blood were altered, mainly with higher PER1, PER2, REV-ERBα and REV-ERBβ, and again the changes were stronger in the short-sleep subtype.
- The more severe the insomnia, the weaker the circadian function measured by the activity wristband. Even so, most activity measures didn’t differ between groups.
- A machine-learning model using the average of just 3 clock genes told patients from controls apart with 92% accuracy. It also separated the two insomnia subtypes with 92% accuracy, using PER2, CLOCK and BMAL1.
The limits are big, though. The sample is very small (14 patients and 13 controls), almost all participants were women, and the study is observational, so it doesn’t prove cause and effect. It’s unclear whether the altered clock causes insomnia or results from it. The paper also contradicts itself on which 3 genes best identify insomnia, and a 92% result in such a small group needs confirming in larger studies. For now it’s a promising lead toward a future blood test for insomnia, not something ready for clinical use.