📄 public and global health

Methodological guidelines for circadian modeling of Daylight Saving Time: application to the United States

This paper critiques a recent study linking Daylight Saving Time to disease prevalence in the United States for a fundamental computational error that inverted the longitudinal axis, and subsequently provides corrected methodological guidelines for accurately modeling the circadian impacts of seasonal clock changes.

Original authors: Martin-Olalla, J. M., Mira, J.

Published 2026-06-22

📖 4 min read☕ Coffee break read

CC BY 4.0

Original authors: Martin-Olalla, J. M., Mira, J.

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ⚕️ This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer

The Big Picture: A Map That Got Flipped Backwards

Imagine you are trying to understand how changing the clocks for Daylight Saving Time (DST) affects people's health across the United States. A recent study tried to do this by creating a computer model.

This paper argues that the recent study made a massive, fundamental mistake. It's as if they tried to study the weather in New York but accidentally used the weather data for a location in the middle of the Pacific Ocean, then claimed their findings applied to New York.

Because of this error, the recent study's conclusions about health and clock changes are, according to the authors, unsupported.

The "East-West" Mix-Up

To understand the mistake, you have to think about how sunlight works:

The Sun rises in the East. If you live in New York, you see the sun before someone in Chicago.
The Sun sets in the West. If you live in Los Angeles, the sun stays up later than in New York.

The authors say the previous study got the direction of time wrong.

How it should work: If you move West, the sun rises later relative to the clock on your wall.
What the study did: They accidentally programmed the computer so that moving West made the sun rise earlier.

The Analogy:
Imagine you are watching a relay race. The runners (sunlight) start at the East end of the track.

Correct: The runner reaches the East side first, then the middle, then the West.
The Study's Error: They told the runners to start at the West and run backward to the East.

Because of this "flipped" map, the study took health data from real American counties and matched it with the wrong amount of sunlight. They essentially paired a county in Florida with the sunlight schedule of a place that doesn't exist, or is on the opposite side of the time zone.

The "Tiny Shift" Problem

Even if the map hadn't been flipped, the authors point out another issue: the effect of changing the clocks is incredibly small.

The Analogy: Imagine you are trying to lose weight. The study claims that changing your diet by eating a single crumb of bread every day for a year will cause massive weight gain or loss.
The Reality: The authors calculated that the "circadian burden" (the stress on your body clock) caused by these time changes is tiny—about 2.6 minutes per day.

They argue that trying to link such a tiny, 2-minute daily shift to big societal health problems (like obesity or stroke) is like trying to hear a whisper in a hurricane. The signal is too weak to be the main cause of major health trends.

What They Did Instead

The authors from the University of Seville and the University of Santiago de Compostela built a corrected model.

Fixed the Map: They ensured that the sun rises earlier in the East and later in the West, just like in real life.
Checked the Math: They looked at how the body clock reacts to different schedules (like waking up at 7 AM vs. 9 AM) and different latitudes (Florida vs. Maine).
The Result: They found that while changing the clocks does create a tiny amount of stress on the body, the effect is so small and inconsistent that it shouldn't be used as the main reason to change public policy.

The Bottom Line

The authors conclude with a simple rule for science: Always anchor your analysis to physical reality.

The sun doesn't care about our clocks; it follows the Earth's rotation.
Our social schedules (when we wake up and work) are the things that should adapt to the sun, not the other way around.

Because the previous study got the direction of the sun wrong, their "proof" that Daylight Saving Time causes specific health issues is invalid. The authors suggest that while the current seasonal clock changes are a "compromise," the data doesn't support the idea that they are causing the major health crises some people claim they do.

Technical Summary: Methodological Guidelines for Circadian Modeling of Daylight Saving Time

Problem Statement
This report addresses a critical methodological flaw in a recent study (Weed and Zeitzer, 2025) that attempted to associate disease prevalence across United States counties with seasonal clock exposure (Daylight Saving Time, DST). The authors identify that the original study contains a fundamental computational error: a sign reversal in the longitudinal offset calculation. This error effectively inverted the US East-West axis within the model, causing local health data to be cross-correlated with the hypothetical circadian burden of locations on the opposite side of a time zone. Consequently, the original study's conclusions regarding the health impacts of seasonal clock changes lack empirical support. Additionally, the report critiques the original study for potential ecological fallacy and the use of a metric (absolute yearly circadian burden) that is insensitive to the direction of phase drift.

Methodology
The authors propose a corrected framework for modeling circadian processes under seasonal time shifts, grounded in the synchronization between solar time ( $t_s$ ) and social time ( $t_c$ ).

Solar vs. Social Time: The model utilizes a 365-day time series with a 5-minute resolution. It distinguishes between fixed social schedules (e.g., 9 am–5 pm work) and variable solar conditions. The time offset ( $t_o$ ) is defined as the longitude difference relative to the time meridian, scaled by Earth's rotation period ( $\Omega = 15^\circ h^{-1}$ ): $t_o = (\lambda - \lambda_t) / \Omega$ .
Correction of the Sign Error: The original study incorrectly used $\Omega = -15^\circ h^{-1}$ , which reversed the physical reality that solar noon occurs later in the West. The corrected model ensures that for a fixed social time, solar time decreases westward, accurately reflecting that sunrises occur later in western longitudes.
Spatial Parametrization: The analysis groups US counties (N=3108) into bins based on two parameters: the time offset ( $t_o$ ) and the winter sunrise solar time (WSR), which serves as a synchronizer for human activity. The WSR is calculated based on latitude ( $\phi$ ) and Earth's axial tilt ( $\epsilon$ ).
Metric Refinement: The authors argue against using the sum of absolute daily phase shifts ( $Y = \sum |\Delta_i|$ ) as the primary metric, noting it cannot distinguish between a stable phase and a monotonically drifting phase. They suggest the standard deviation (SD) of the yearly distribution of phase shifts as a more physiologically and statistically robust alternative.
Simulation Scope: The corrected model simulates various scenarios, including locked clocks (no DST) and seasonal clocks, across different chronotypes (early, intermediate, late) and social schedules (e.g., 9-5, 8-4, 7-3).

Key Contributions

Identification of Geographical Inversion: The paper demonstrates that the original study's use of a negative rotation constant inverted the US East-West axis, creating a "flipped" distribution of counties (Figure 2) where western counties were modeled with eastern solar conditions and vice versa.
Methodological Correction: The authors provide the correct mathematical formulation for mapping social schedules to solar time, ensuring that boolean conditions for light exposure (e.g., work hours) are applied to the correct solar time values.
Re-evaluation of Metrics: The report highlights the limitations of the "yearly circadian burden" metric used in previous work, advocating for metrics that account for the stability of the circadian phase rather than just the magnitude of absolute shifts.
Data Density Analysis: The authors establish a "geographic h-index" for the analysis, identifying 29 distinct bins containing at least 30 counties each, ensuring sufficient data density for robust societal outcome analysis.

Results

Magnitude of Burden: In the corrected model, the yearly circadian burden ( $Y$ ) is found to be extremely low, averaging approximately 16 hours per year (2.63 minutes per day), which represents only 0.2% of a day.
Statistical Significance: The authors argue that such marginal variations are unlikely to yield statistically significant societal health outcomes given the inherent limitations of the model and the noise in epidemiological data.
Pattern of Burden: The burden generally scales with chronotype. Within a specific chronotype, shifting schedules earlier (moving left in the heat maps) generally decreases the burden. A diagonal band of low burden runs from the Southwest to the Northeast, indicating that southern populations can sustain earlier schedules without significant deterioration in $Y$ due to the lack of dark morning hours.
Impact of DST Transitions: The biannual clock transitions (ST to DST and back) inflate the burden immediately following transition dates. However, these transitions also shift the baseline schedule, which can decrease the overall burden. These two competing effects largely cancel out, resulting in small net differences between seasonal and locked clocks.
Health Correlation: The dispersed distribution of $Y$ across minor changes in time offset and WSR challenges the validity of $Y$ as a reliable predictor of health risks in the current model.

Significance and Claims
The paper concludes that the original study's findings are unsupported due to the geographical inversion error. The authors assert that in a scientific context, analysis must be anchored to physical reality: solar conditions are fixed, and social schedules are the flexible variable.

The authors modestly claim that while their corrected model suggests that permanent DST is not necessarily superior to permanent Standard Time or the current seasonal system (as it would deteriorate outcomes for individuals aligned with earlier schedules), the metric $Y$ is too low and unsmooth to serve as a robust deciding factor for public policy. Instead, the analysis may offer a subtle quantitative clue regarding current public preferences for DST. The primary significance of this work lies in establishing a methodologically robust framework for future circadian modeling, preventing the propagation of errors related to longitudinal offsets and ensuring that correlations between time policy and health outcomes are based on accurate physical modeling.