Some patterns of sleep quality and Daylight Saving Time across countries: a predictive and exploratory analysis

Bhanu Sharma, Eugene PinskyarXiv 2025 · DOI · arXiv · OpenReview

TL;DR

This study reveals that Daylight Saving Time (DST) is associated with better sleep quality and duration globally, but its effectiveness is significantly moderated by latitude. While DST appears to have minimal impact at low latitudes, it shows progressively greater benefits in mid to high latitudes, suggesting that geophysical factors should inform DST policy.

What Problem Does This Paper Address?

The debate surrounding Daylight Saving Time (DST) often centers on its potential impacts on health, productivity, and well-being. While many countries observe DST, there's a lack of systematic, data-driven evaluations of its actual effects on sleep. Existing research has highlighted potential disruptions to circadian rhythms and associations with sleep disturbances, but the overall impact and its variability across different regions remain unclear. This paper aims to bridge this gap by analyzing sleep duration data from a global sample of countries to understand how DST practices influence sleep quality and duration, and how these effects vary with geographical location, particularly latitude.

Methods

The researchers employed a multi-faceted approach:

  • Data Collection and Augmentation: They utilized publicly available sleep statistics from Sleep Cycle, covering 61 countries. This original dataset included sleep quality, sleep duration, snore duration, wake-up time, and bedtime. They augmented this with geophysical data such as latitude, hemisphere, and seasonal daylight metrics (longest night duration, equinox night duration, and their ratios).
  • Statistical Analysis: Pearson correlation coefficients were calculated to identify relationships between sleep features and geophysical variables. They also performed descriptive statistical analysis to compare sleep patterns between countries that observe DST and those that do not.
  • Machine Learning for Prediction: To assess the predictive power of geophysical variables for DST implementation, they developed binary classification models. This involved feature scaling (StandardScaler), model selection (Random Forest, K-Nearest Neighbors, Gaussian Naive Bayes, Logistic Regression), and cross-validation (10-fold) to ensure robust performance estimates.

Datasets

  • Sleep Cycle Dataset: This dataset, sourced from Sleep Cycle's open dataset API, provided average sleep statistics for 61 countries. The original variables included: Country, Sleep Quality, Sleep Duration, Wake-up Time, and Bedtime.
  • Augmented Dataset: The researchers created an augmented dataset by incorporating geophysical variables and seasonal daylight metrics. This expanded dataset included 36 countries practicing DST and 25 that do not, enabling comparative analysis.

Key Insights and Ideas

  • Global Sleep Advantage with DST: On average, countries observing DST tend to report better sleep quality and longer sleep durations compared to non-DST countries. This suggests a potential global benefit of DST.
  • Latitude is Key: The impact of DST is not uniform. At lower latitudes, DST's effect on sleep quality is minimal, with high variability suggesting other factors are more dominant. However, at mid and high latitudes, DST shows a progressively greater positive impact on sleep quality.
  • DST Restructures Sleep Dynamics: DST countries exhibit a stronger correlation between sleep quality/duration and bedtime, indicating that sleep timing becomes a more critical factor for sleep health in these regions. This suggests DST fundamentally alters sleep parameter relationships, creating heightened sensitivity to timing.
  • Geophysical Factors Predict DST: Latitude and seasonal daylight variation are significant predictors of whether a country observes DST, suggesting that geophysical realities, rather than solely historical or political factors, could better inform DST policy.

Key Results

  • Sleep Quality and Duration: Countries practicing DST showed higher average sleep quality (0.7681 vs. 0.7257) and longer sleep duration (7.6262 hrs vs. 7.1591 hrs) compared to non-DST countries.
  • Latitude-Specific Sleep Quality:
    • Low Latitudes (0-30°): DST had a marginal effect on sleep quality (0.741 for DST vs. 0.726 for non-DST).
  • Mid Latitudes (30-45°): DST countries showed significantly higher sleep quality (0.760) than non-DST countries (0.714).
  • High Latitudes (45-60°): DST countries demonstrated consistently superior sleep quality (0.774) compared to the single non-DST country (0.760).
  • Machine Learning Performance: Classification models predicting DST implementation based on geophysical variables achieved high performance. Random Forest achieved the highest AUC of 0.863, with all models demonstrating mean accuracies exceeding 78%.

Limitations and Future Work

  • Limited Sample Size: The study acknowledges that the sample size of 61 countries limits the generalizability of the conclusions.
  • Data Imbalance at High Latitudes: A significant limitation is the imbalance in the dataset at high latitudes, with only one non-DST country (Russia) represented compared to 24 DST countries. This hinders comparative claims for this region.
  • Causation vs. Prediction: While high classification accuracy suggests predictive strength, it does not imply causation. DST policies are influenced by complex socio-political factors, and the dataset reflects current political choices, which may not always align with optimal health or energy considerations.
  • Future Work: The authors suggest further investigation with larger datasets and complementary variables (e.g., economic activity, detailed sleep data, health outcomes) to build upon these findings. They also emphasize the need for more balanced representation of non-DST countries at high latitudes in future research.

Summary updated

Abstract

In this study we analyzed average sleep durations across 61 countries to examine the impact of Daylight Saving Time (DST) practices. Key metrics influencing sleep were identified, and statistical correlation analysis was applied to explore relationships among these factors. Countries were grouped based on DST observance, and visualizations compared sleep patterns between DST and non-DST regions. Results show that, on average, countries observing DST tend to report longer sleep durations than those that do not. A more detailed pattern emerged when accounting for latitude: at lower latitudes, DST-observing countries reported shorter sleep durations compared to non-DST countries, while at higher latitudes, DST-observing countries reported longer average sleep durations. These findings suggest that the influence of DST on sleep may be moderated by geographical location.
BibTeX
@misc{sharma2025some,
  author = {Bhanu Sharma and Eugene Pinsky},
  title = {{Some patterns of sleep quality and Daylight Saving Time across countries: a predictive and exploratory analysis}},
  year = {2025},
  eprint = {2509.03358},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2509.03358},
}
Figure