Should We Keep Changing the Clock? Characterizing Causal Effects of Daylight Saving Time on Behavior and Physiology

Using large-scale, objective wearable data from the NIH All of Us Research Program, this study employs a causal difference-in-differences design to reveal that Daylight Saving Time transitions do not alter total daily activity but instead reallocate steps across the day, with these shifts varying significantly by age, income, and activity phenotype due to structural constraints.

The Gist

Imagine the country is a giant orchestra, and the conductor (the government) decides to shift the music sheet by one hour twice a year. This is Daylight Saving Time (DST). The big debate is: Does this shift make us healthier and more active because we have more evening sunlight, or does it just mess up our sleep and rhythms?

This paper is like a massive, high-tech detective story that uses data from over 1,000 people wearing Fitbits to find the real answer. Here is the breakdown in plain English:

1. The Big Question: Do We Move More?

For years, the argument for keeping DST has been: "If we shift the clock, we get an extra hour of sunlight in the evening. People will use that light to go for walks, play with their kids, or run errands, making the whole population healthier."

The Verdict: The study says no.
Think of your daily energy like a bucket of water. The DST clock change didn't pour more water into the bucket. Instead, it just moved the water around. The total amount of walking people did every day stayed exactly the same. We didn't get fitter; we just changed when we walked.

2. The "Time Travel" Experiment

To figure this out, the researchers needed a control group—a group of people who didn't change their clocks. They found the perfect natural experiment: Arizona.

• The Control Group: People in Arizona (who don't do DST) kept their clocks steady.
• The Test Group: People in neighboring states like Colorado and Utah (who do do DST) changed their clocks.

By comparing these two groups, the researchers could see what happened only because of the clock change, not because of the changing seasons (like getting colder or hotter).

3. The Great "Step Shuffle"

Here is what they found when the clocks changed:

• In the Fall ("Fall Back"): When we gain an hour of morning light, people in the DST states started walking 202 more steps in the morning but took 180 fewer steps in the evening.

  • The Metaphor: Imagine you usually eat dinner at 6:00 PM. Suddenly, the sun sets later, so you decide to eat at 7:00 PM. You didn't eat more food; you just ate your meal at a different time. The same thing happened with walking. People shifted their morning walks to take advantage of the earlier sunrise, but they cut back on evening walks because it got dark earlier.

• In the Spring ("Spring Forward"): The opposite happened. People walked a bit more in the evening (when the sun stayed up later) but a bit less in the morning. Again, the total steps for the day didn't change.

4. Not Everyone Can "Shuffle" the Same Way

The study found that some people are better at adjusting their schedules than others. It's like a dance floor: some people can easily switch partners, while others are stuck in their spots.

• Older Adults (65+) & Lower-Income Groups: These groups were the most flexible. When the clock changed, they successfully shifted their walking to the new daylight hours. They were like dancers who could easily adapt to a new rhythm.
• Younger Adults & Higher-Income Groups: These groups struggled to change. Why? Because their lives are often more rigid. They have strict 9-to-5 jobs, school schedules, or commutes. Even if the sun is shining at 7:00 AM, they can't go for a walk because they are already at work or stuck in traffic. They are like dancers who are tied to a specific chair.
• Women vs. Men: In the spring, women took advantage of the extra evening light to walk more, while men didn't change their habits much. The researchers suggest this might be because women often feel less safe walking alone in the dark, so the extra evening sunlight removes a barrier for them.

5. The Heartbeat Check

The researchers also looked at resting heart rates (a sign of stress or sleep quality).

• The Result: The changes were tiny—so small they wouldn't matter to a doctor.
• The Takeaway: While our steps moved around, our bodies didn't get a massive shock. However, the study notes that the body's internal clock (circadian rhythm) did get slightly nudged, just like a clock that is a few minutes off.

The Bottom Line

The common belief that "more evening sunlight = more exercise" is a myth, at least for the general population.

The Analogy: Think of DST not as a "bonus hour" of activity, but as a traffic reroute. The government changed the traffic signs, and cars (people) took a different route to get to the same destination. They didn't drive more miles; they just drove at a different time of day.

Why does this matter?
If policymakers want to make people healthier, simply changing the clocks isn't the magic bullet. To actually increase physical activity, we need to fix the things that stop people from moving in the first place—like rigid work schedules, unsafe neighborhoods, or lack of time. The clock change just moves the problem around; it doesn't solve it.

Technical Summary

1. Problem Statement

Daylight Saving Time (DST) remains a contentious public policy issue. While proponents argue that shifting clocks to extend evening daylight promotes physical activity and public health, opponents cite risks to circadian rhythms, sleep, and cardiovascular health.

• Limitations of Existing Literature: Previous studies rely heavily on:
  • Aggregated administrative data (e.g., hospital admissions, traffic crashes).
  • Self-reported surveys (e.g., American Time Use Survey), which suffer from recall bias and low temporal resolution.
  • Descriptive wearable studies lacking robust control groups, making causal inference difficult due to unobserved confounders (e.g., seasonal weather changes).
• Research Gap: There is a lack of large-scale, causal evidence regarding how DST transitions affect individual-level physical activity patterns and physiological markers using continuous, objective data.

2. Methodology

The authors employed a rigorous Difference-in-Differences (DiD) design to isolate the causal effect of DST transitions from seasonal trends and other confounders.

• Data Source:
  • Dataset: NIH All of Us Research Program (Fitbit data).
  • Cohort: 1,157 participants from Arizona (control group) and neighboring Mountain Time states (Colorado, New Mexico, Utah; treatment group) with valid intraday heart rate and step data during 2021–2023.
  • Timeframe: Two weeks surrounding Spring ("spring forward") and Fall ("fall back") DST transitions.
• Study Design (DiD):
  • Control: Arizona (does not observe DST).
  • Treatment: Neighboring Mountain Time states (observe DST).
  • Rationale: These regions share similar geography, latitude, and weather patterns, satisfying the parallel trends assumption (i.e., in the absence of DST, activity trends would have evolved similarly in both groups).
  • Model: A regression model controlling for year fixed effects, day-of-week fixed effects, and subject-level clustering.
  • Outcome Variables:
    • Behavioral: Total daily steps, morning steps (6–9 AM), and evening steps (5–7 PM).
    • Physiological: Average Resting Heart Rate (RHR) during restful periods.
• Subgroup Analysis:
  • Activity Phenotypes: Participants were classified as "Morning Walkers," "Evening Walkers," or "Neutral Walkers" based on step patterns in the 6 weeks prior to the transition.
  • Demographics: Stratified by age (18–45 vs. 65+), income, and sex.
  • Environmental Factors: Stratified by perceived neighborhood walkability.

3. Key Contributions

• First Large-Scale Causal Analysis: This is the first study to use continuous wearable data from a diverse, large-scale cohort to establish a causal link between DST and behavior/physiology, moving beyond simple before-and-after comparisons.
• Refutation of the "Activity Boost" Hypothesis: The study challenges the common assumption that DST increases total physical activity volume.
• Granular Temporal Reallocation: It demonstrates that DST does not change how much people move, but rather when they move, revealing a "reciprocal shift" in activity timing.
• Heterogeneity of Effects: It identifies that the ability to adapt to time shifts is not uniform, varying significantly by age, income, and activity phenotype, suggesting structural constraints (e.g., rigid work schedules) limit adaptation for certain groups.

4. Key Results

A. Behavioral Effects (Physical Activity)

• No Net Change in Total Steps: DST transitions did not result in a statistically significant increase or decrease in total daily steps for either Spring or Fall transitions.
• Temporal Reallocation:
  • Fall Transition: Activity shifted earlier. Morning steps increased by 202 steps (p=0.001), while evening steps decreased by 180 steps (p<0.001).
  • Spring Transition: Activity shifted later. Evening steps increased by 90 steps (p=0.021), with a non-significant reduction in morning steps.
• Phenotype-Specific Responses:
  • Morning Walkers: Significantly capitalized on earlier sunrise in the Fall (increased morning activity).
  • Evening Walkers: Did not significantly increase evening activity during the Spring transition, despite the alignment of daylight with their preferred time. This indicates an asymmetry in behavioral flexibility.
  • Neutral Walkers: Showed reciprocal shifts (morning up/evening down in Fall; vice versa in Spring), but only if they perceived their neighborhood as highly walkable.
• Demographic Stratification:
  • Age: Adults aged 65+ showed significant morning reallocation in the Fall; younger adults (18–45) showed limited adaptation, likely due to rigid social/work schedules.
  • Income: Lower-income individuals (≤$50k) compensated for reduced evening activity with increased morning activity in the Fall. Higher-income groups did not show this morning increase, suggesting occupational constraints.
  • Sex: Females increased evening steps in the Spring (potentially due to improved perceived safety in daylight), while males did not.

B. Physiological Effects (Resting Heart Rate)

• No Net Change: There was no significant change in the daily average RHR.
• Intraday Shifts: RHR mirrored behavioral shifts. In the Fall, morning RHR rose slightly (+1.3 bpm) while evening RHR declined (-0.8 bpm).
• Clinical Significance: These physiological shifts were statistically significant but clinically insignificant (changes < 2 bpm).

5. Significance and Implications

• Policy Implications: The study suggests that DST is unlikely to serve as an effective public health intervention for increasing total physical activity levels. The "extra" evening light does not generate new activity volume but merely displaces existing activity.
• Health Equity: The findings highlight that the capacity to adapt to time shifts is unequal. Vulnerable populations (older adults, lower-income individuals) may be more flexible, while those with rigid schedules (often higher-income or younger workers) cannot adapt, potentially exacerbating health disparities if DST is made permanent.
• Methodological Advancement: By utilizing a natural experiment (Arizona vs. Mountain Time) with wearable data, the study provides a template for generating real-world evidence (RWE) for health policy that avoids the pitfalls of self-reporting and strict modeling assumptions.
• Future Directions: The authors note that while total volume didn't change, the timing of activity has known metabolic implications. Future research should investigate whether these forced shifts in activity timing (e.g., moving exercise to darker mornings) have long-term circadian or metabolic consequences.

Conclusion: The paper concludes that DST transitions primarily cause a reallocation of activity timing rather than an increase in activity volume. This reallocation is heterogeneous across the population, constrained by structural factors like work schedules and neighborhood safety, and does not yield the net health benefits often attributed to extended evening daylight.