Summary: In two pooled randomised crossover trials, six weeks of mild sleep restriction did not change thyroid markers in the full sample, but a sex-by-condition interaction showed TSH fell significantly in women specifically, while free thyroxine and FGF-21 were unchanged.
PICO Summary
| Element | Detail |
|---|---|
| Population | 30 healthy adults (20 women) with adequate habitual sleep, pooled from two identical randomised crossover trials. |
| Intervention | 6 weeks of mild sleep restriction (1.5 h/night below adequate sleep). |
| Comparison | Adequate sleep (7–9 h/night), within-person crossover. |
| Outcome | No effect on FT4 or TSH in the full sample (both p>0.6). Significant sex-by-condition interaction for TSH (p=0.049): TSH fell in women (β -0.11; p=0.011) but not men. FGF-21 (women) unchanged. |
Sleep restriction and thyroid function
Pooled RCTs · crossover · 6 weeks
Mild sleep restriction lowered TSH in women but not men, with no effect overall and FT4 unchanged. The signal rests on a borderline interaction in 30 people, so it is hypothesis-generating only.
Expert Commentary
Chronic insufficient sleep is genuinely common and metabolically consequential, so asking whether it nudges the thyroid axis is a sensible question, and the crossover design with each person as their own control is a strength. My verdict is cautiously interested but firmly tempered, and the reason is in the numbers the headline skips: in the whole sample there was no effect on TSH at all. The signal is entirely a sex-stratified one, TSH falling in the twenty women but not the men, resting on a sex-by-condition interaction that scrapes significance. With thirty people total that is hypothesis-generating, not established, and the drop is within the normal range, so calling it subclinical axis suppression overstates a small statistical shift. TSH also has strong circadian variation, which sampling can confound. Can I use this with my patients? Only very softly. It is a reasonable nudge to ask about sleep when a woman has an unexplained low-normal TSH, and not to over-diagnose hyperthyroidism in someone sleep-deprived, but I would not attribute cardiometabolic risk to it or change management. I would want a larger, sex-balanced study with circadian-controlled sampling before drawing firmer conclusions.
References
Petrov ME, Zuraikat FM, Cheng B, et al. Impact of sleep restriction on biomarkers of thyroid function: two pooled randomized trials. Sleep Med. 2024;124:606–612. doi:10.1016/j.sleep.2024.10.035