Dataset and Code for Circadian ΔT Study

This repository contains anonymized data and analysis code from a 15-day prospective observational study investigating circadian body temperature dynamics under real-world conditions.

The dataset includes:

Daily axillary temperature measurements (morning and evening)

Derived thermal oscillation metric (ΔT = T_night − T_morning)

Intermittent cognitive–autonomic self-assessments (MiSBIE Brief-6)

Environmental exposure variables (morning light exposure and nighttime screen use)

The dataset comprises 239 valid daily observations from 16 adult participants.

Analysis scripts reproduce:

Linear mixed-effects models for longitudinal ΔT variation

K-means clustering applied to ΔT

Descriptive analyses relating ΔT to environmental and self-reported variables

All data have been fully anonymized.

This repository is intended to support reproducibility of analyses associated with a manuscript currently under peer review. The manuscript itself is not included to avoid prior publication conflicts.

VARIABLES:

Participant: Anonymous ID (P01–P16)

Day: Study day (1–15)

DeltaT_night_minus_morning: ΔT in °C (T_night − T_morning)

MiSBIE_delta: Change in MiSBIE Brief-6 score between consecutive assessments (when available)

Morning_Light_min: Estimated minutes of light exposure >1000 lux

Screen_Time_h: Nighttime screen exposure (hours)

Cluster: Cluster assignment derived from ΔT-based k-means clustering (k = 3)

METHODOLOGY:

Clustering was performed using ΔT alone (n = 239 complete observations).

MiSBIE Brief-6 was assessed on days 1, 7, and 15 (mandatory), with additional optional entries recorded when participants reported changes in symptoms, resulting in approximately 41.8% coverage.

MiSBIE-delta values were used descriptively to characterize cluster-level symptom and recovery patterns and were not included as clustering inputs.

This repository contains anonymized data and analysis code associated with a manuscript currently under peer review. The manuscript itself is not included to avoid prior publication conflicts