We publish the methods behind the phase estimate, including the cases where it fails. Protocols and analysis code go to collaborating institutions on request.
Sixty-eight participants, paired lab DLMO sessions against ring-derived phase estimates. Median absolute error of 19 minutes, with degradation in participants reporting irregular ambient temperature exposure.
Two hundred and twelve participants tracked across full rotation cycles. Most participants partially re-entrained rather than inverting, which reframes what a prescription should aim for.
Three hundred and forty night-shift workers, prescribed twenty-minute light windows against usual behaviour. Primary endpoint is phase shift magnitude; adherence is reported as a secondary outcome and is the honest limit on effect size.
The dual-thermistor arrangement and the correction we apply to separate circadian vasodilation from a cold room. Includes the failure modes we could not correct.
We supply devices, raw 1 Hz series and phase estimates to labs working on circadian measurement, under a data use agreement. Participant consent covers research sharing in de-identified form only.
Contact the science team