Abstract
Introduction
Circadian disruption contributes to adverse effects on sleep, performance and health. One of the most practical methods to track continuous daily changes in circadian timing is to measure core body temperature (CBT) to identify the CBT minimum time (Tmin), usually via cosine-model fits to measured data in controlled studies. However, this method ignores large effects of activity, sleep and wake on CBT that confound and mask the circadian component of primary interest. This study introduces a novel physiology-grounded analytic approach to separate circadian from non-circadian effects on CBT.
Methods
The dataset comprised 33 healthy participants (mean±SD 32±13 years old) attending a 39-hour in-laboratory study with an initial overnight sleep followed by extended wake. CBT data were collected at 30-second intervals via ingestible capsules. A physiology-guided generalized additive model was constructed to model the combined circadian and non-circadian effects of sleep, wake, and activity on CBT. Model fits and estimated Tmin inferred from extended wake without sleep were compared with traditional cosine-based models fits.
Results
Compared to the traditional cosine model, the new model exhibited superior fits to CBT (Pearson R 0.90 [95%CI; [0.83,0.96] versus 0.81 [0.55-0.93]) and better estimation of Tmin (difference of 0.2 [-0.5,0.3] hours versus 1.4 [1.1 to 1.7] hours).
Conclusions
This new method provides superior demasking of non-circadian influences compared to traditional cosine fits, including the removal of a sleep related bias towards an earlier estimate of the circadian component of Tmin.