Cardiometabolic health and the timing of habitual exercise in the All of Us Research Program

This study utilizing one year of minute-level heart rate data from 14,489 All of Us participants reveals that habitual morning exercise is independently associated with significantly lower odds of coronary artery disease, hypertension, type 2 diabetes, hyperlipidemia, and obesity compared to daytime exercise, suggesting that exercise timing is a distinct and underappreciated factor in cardiometabolic health.

The Gist

Imagine your body is a high-performance car. For years, mechanics (doctors) have told us that to keep the engine running smoothly, we need to drive a certain distance (exercise volume) and at a certain speed (exercise intensity).

But what if the time of day you choose to drive matters just as much?

This new study, involving nearly 15,000 people wearing Fitbits for a whole year, suggests that when you exercise might be a secret ingredient for heart health that we've been ignoring.

Here is the breakdown in simple terms:

1. The Big Discovery: The "Morning Rush" is Better

The researchers looked at heart rate data to see when people were actually doing serious exercise (not just walking to the fridge). They found three main groups of people:

• The Early Birds: People who worked out in the morning.
• The Mid-Day Crew: People who exercised around lunch or early afternoon.
• The Night Owls: People who waited until the evening.

The Result? The Early Birds had the best health outcomes. Compared to the Mid-Day Crew, those who exercised in the morning had significantly lower odds of developing:

• Heart disease
• High blood pressure
• Type 2 diabetes
• High cholesterol
• Obesity

It's as if the morning workout "tunes" the engine in a way that afternoon or evening workouts simply don't.

2. It's Not Just About "How Much"

You might think, "Maybe morning people just exercise more than everyone else?"
The study checked this. Even when they compared people who exercised the exact same amount of time, the morning exercisers still had better health.

• Analogy: Think of it like watering a garden. If you water the plants at 6:00 AM, the sun helps the water soak in and the plants drink it up. If you water them at 8:00 PM, the water might just sit on the surface or evaporate differently. The amount of water is the same, but the timing changes the result.

3. The "Sweet Spot"

The study didn't just look at "Morning" vs. "Evening." They looked at every single hour of the day.

• The Golden Hour: The lowest risk for heart disease was found between 7:00 AM and 8:00 AM.
• The Trend: As the day went on, the benefits slowly faded. By the time you hit late evening (after 9:00 PM), the protective effect was gone, and in some cases, the risk was slightly higher than the daytime average.

4. Why Does This Happen? (The Science Simplified)

The paper suggests a few reasons why morning exercise might be a superpower:

• The Body Clock: Your body has an internal clock (circadian rhythm) that controls how you process sugar and fat. Morning exercise seems to sync perfectly with this clock, helping your body burn fuel more efficiently.
• Habit Formation: It's easier to stick to a routine if you do it first thing in the morning before the chaos of the day (work, kids, stress) gets in the way.
• Dietary Effects: Morning exercise might naturally make you eat better or manage your weight better throughout the rest of the day.

5. What This Means for You

• Don't Panic if You're a Night Owl: This study shows a statistical trend, not a rule. If you can only exercise at night, doing it is still better than not doing it. The study emphasizes that moving is the most important thing.
• The "Bonus" Tip: If you have the flexibility to choose, try shifting your workout to the morning. It might give your heart and metabolism an extra boost that you aren't getting otherwise.
• It's a New Dimension: For a long time, we thought exercise was just about "how hard" and "how long." This study adds a third dimension: "When."

The Bottom Line

Think of exercise timing as a "secret sauce." You can't replace the main ingredients (moving your body), but adding the right timing (morning) might make the dish taste even better for your long-term health.

Note: This study is a "preprint," meaning it's a very fresh look at the data that hasn't been fully peer-reviewed by other scientists yet, but the findings are exciting and align with what we know about our body's natural rhythms.

Technical Summary

1. Problem Statement

While the benefits of exercise volume and intensity for cardiometabolic health are well-established, the clinical significance of exercise timing remains poorly understood. Previous research has been limited by:

• Methodological constraints: Reliance on self-reported data or short-term accelerometry (e.g., 7-day wear periods), which fail to capture long-term habitual patterns.
• Lack of physiological specificity: Inability to distinguish structured, purposeful exercise from incidental daily movement.
• Data gaps: A lack of large-scale studies utilizing continuous, objective physiological data to investigate the association between habitual exercise timing and major cardiometabolic diseases.

2. Methodology

Study Population & Data Source

• Cohort: 14,489 participants from the All of Us Research Program (AoURP).
• Inclusion Criteria: Participants aged ≥18 with linked Fitbit wearable data and Electronic Health Record (EHR) data.
• Quality Control: To ensure stable estimation of habitual patterns, participants were required to have at least 30 valid days within the first six months of monitoring. A valid day was defined as ≥600 minutes of non-missing heart rate data and ≥100 steps.
• Data Harmonization: Fitbit data (minute-level heart rate and daily steps) was linked with EHR data (ICD-10 codes) and survey data.

Physiological Definition of Exercise

• Heart Rate Threshold: Exercise episodes were defined as sustained physiological exertion. An episode was identified if the mean heart rate was ≥60% of the participant's observed maximal heart rate for at least 15 consecutive minutes (using a sliding 15-minute window).
• Validation: Sensitivity analyses were performed using 50% and 70% thresholds. The 60% threshold was selected based on visual assessment of intensity distributions.
• Concordance Check: To filter out unreliable signals, the study calculated the log-ratio of heart rate-derived exercise minutes to total physical activity minutes. Only participants in the central 55% of this distribution were included in the final analysis.

Analytical Approaches

1. Unsupervised Clustering: A 1,440-element vector (representing exercise frequency per minute of the day) was constructed for each participant. K-means clustering (K=5, collapsed to 3 for interpretability) identified three distinct phenotypes: Morning, Daytime, and Evening.
2. Hour-of-Day Modeling: Exercise timing was analyzed as a continuous variable, calculating the percentage of total exercise episodes starting within each specific hour (e.g., 07:00–08:00).
3. Statistical Models:
   • Prevalence: Multivariable logistic regression to assess associations with existing conditions.
   • Incidence: Cox proportional hazards models for new-onset diseases (restricted to participants free of the condition at baseline).
   • Adjustments: Models were adjusted for age, sex, income, alcohol use, smoking, sleep duration, and total physical activity volume (to isolate the effect of timing from volume).

Outcomes
Primary outcomes included the prevalence and incidence of: Coronary Artery Disease (CAD), Hypertension, Type 2 Diabetes, Hyperlipidemia, Obesity, and Atrial Fibrillation. Negative controls (myopia, benign skin conditions) were used to test for specificity.

3. Key Contributions

• Scale and Duration: This is one of the first large-scale studies to utilize one year of continuous, minute-level heart rate data to define habitual exercise, overcoming the limitations of short-term accelerometry.
• Objective Phenotyping: By defining exercise based on physiological exertion (heart rate) rather than self-report or step counts, the study accurately distinguishes structured exercise from incidental movement.
• Novel Insight: It identifies exercise timing as an independent, modifiable dimension of cardiometabolic health, distinct from total activity volume.

4. Key Results

Phenotypic Characterization

• Three distinct clusters were identified: Morning (n=2,757), Daytime (n=5,728), and Evening (n=6,004).
• The Morning group was generally older, had higher income, and engaged in higher volumes of vigorous activity compared to the Evening group.

Associations with Cardiometabolic Disease
Compared to the Daytime reference group, Habitual Morning Exercise was associated with significantly lower odds of:

• Coronary Artery Disease: OR 0.69 (95% CI 0.55–0.87)

• Type 2 Diabetes: OR 0.70 (95% CI 0.58–0.85)

• Obesity: OR 0.65 (95% CI 0.55–0.77)

• Hypertension: OR 0.82 (95% CI 0.72–0.94)

• Hyperlipidemia: OR 0.79 (95% CI 0.69–0.90)

• Evening Exercise: Showed directionally higher odds of hypertension, diabetes, and obesity compared to the Daytime group, though not all reached statistical significance in the same magnitude as the morning benefits.

• Atrial Fibrillation: No significant associations were found between exercise timing and atrial fibrillation.

• Negative Controls: No associations were found with myopia or benign skin conditions, supporting the specificity of the findings to cardiometabolic pathways.

Temporal Gradients

• Hour-of-day analysis revealed a continuous gradient: risk decreased in the morning and increased progressively toward the evening.
• The nadir (lowest risk) for Coronary Artery Disease occurred specifically between 07:00 and 08:00.
• These associations remained robust after adjusting for total physical activity volume, indicating that the timing itself, not just the amount of exercise, drives the benefit.

Incident Disease Analysis

• Longitudinal analyses of incident disease showed similar trends (Hazard Ratios < 1.0 for morning exercise) but did not reach statistical significance, likely due to the smaller number of incident events compared to prevalent cases.

5. Significance and Implications

• Clinical Guidance: The findings suggest that current clinical guidelines, which prioritize volume and intensity, may be incomplete. Exercise timing represents a distinct, underappreciated dimension of behavior that could be optimized for disease prevention.
• Biological Plausibility: The results align with circadian biology, where metabolic and cardiovascular processes exhibit diurnal variation. Morning exercise may align better with natural circadian rhythms, enhancing insulin sensitivity, lipolysis, and blood pressure regulation.
• Future Directions: The study provides a strong rationale for randomized controlled trials to determine if shifting exercise timing to the morning can causally reduce cardiometabolic risk.
• Limitations: As an observational study, causality cannot be definitively proven. The cohort consists of active wearable users, which may limit generalizability to populations with lower digital health literacy. Resistance exercise with long rest periods may be under-detected by heart-rate thresholds.

Conclusion: Habitual morning exercise is independently associated with a lower likelihood of major cardiometabolic conditions. This suggests that strategic timing of exercise could be a powerful, modifiable tool for public health and clinical management of cardiometabolic disease.