Leisure-time physical activity on lifelong trajectories of body mass index and obesity risk throughout life: multivariable regression and Mendelian randomization analyses using real-world data from the CORDELIA-Catalunya Study

Using data from the CORDELIA-Catalunya Study, this research demonstrates through both multivariable regression and Mendelian randomization that higher leisure-time physical activity is causally associated with more favorable lifelong BMI trajectories and a reduced risk of obesity, with protective effects observed in both sexes but being stronger in men.

The Gist

The Big Picture: The "Body Battery" Study

Imagine your body is a car, and your Body Mass Index (BMI) is the fuel gauge. Usually, as we get older, this gauge tends to drift upward (we gain weight) until about age 60, and then it might start drifting downward as we lose muscle mass in our later years.

This study asked a simple but huge question: Does moving your body (exercise) change how that fuel gauge behaves over a whole lifetime?

The researchers wanted to know if being active doesn't just help you lose weight today, but actually changes the entire "road trip" of your life, keeping your weight healthier for longer.

How They Did It: The "Three-Tool" Detective Work

To get a clear answer, the researchers didn't just ask people, "Do you exercise?" and "Are you heavy?" They used three different detective tools to make sure they weren't being fooled by coincidence.

1. The Long-Term Diary (Observational Data)
They looked at the real-world medical records of nearly 15,000 people in Catalonia, Spain. They tracked these people's weight from age 30 all the way to age 90.

• The Analogy: Imagine watching a time-lapse video of a garden over 60 years. They didn't just take one photo; they took thousands of photos to see exactly how the plants (the people's weight) grew and changed over decades.

2. The Genetic "Blueprint" (Mendelian Randomization)
This is the tricky part. Sometimes, people who exercise also eat better or have more money, which makes it hard to know if the exercise is the hero or if it's just the lifestyle.
To solve this, the researchers looked at people's DNA. They found specific genetic "blueprints" that make some people naturally have better cardio-respiratory fitness (like having a naturally bigger engine).

• The Analogy: Think of genetics as the factory settings of a car. If you were born with a "high-performance engine" gene, you are naturally more active and fit, regardless of your choices. By seeing if people with these "high-performance genes" stayed lighter, the researchers could prove that the fitness itself (caused by the genes) was protecting them from weight gain, not just their good habits.

3. The "Real-World" Check
They combined the diary data with the genetic data. If both the diary and the genetic blueprint told the same story, they knew it was a fact, not a fluke.

What They Found: The "Slow-Motion" Effect

The results were fascinating and showed that exercise acts like a brake and a shock absorber for your weight.

• The "Brake" (Ages 30–60):
  As we get older, our weight usually creeps up. The study found that people who exercised a lot had a slower brake on this creep.

  • The Metaphor: Imagine two people walking up a hill. The person who doesn't exercise is walking up a steep, slippery slope, gaining weight quickly. The person who exercises is walking up a gentle, paved path. They are still going up, but much, much slower.

• The "Shock Absorber" (Ages 70–90):
  In old age, people often lose weight rapidly because they lose muscle. The study found that active people lost weight more slowly than inactive people.

  • The Metaphor: Inactive people are like a house with a leaky roof; the weight (and muscle) drains away fast. Active people have a sturdy roof; they still lose a little, but they hold onto their "structure" much better.

• The "Obesity Shield":
  People who exercised were significantly less likely to ever cross the line into "obesity" (BMI over 30).

  • The Metaphor: It's like having a forcefield. The more you move, the harder it is for the "obesity monster" to catch you.

A Surprising Twist: The "Men vs. Women" Engine

The study found that while exercise helped everyone, the "engine" seemed to run a bit more efficiently in men.

• The Analogy: Think of men and women as two different types of vehicles. Both benefit from a good tune-up (exercise), but the men's vehicles seemed to show a slightly more dramatic improvement in fuel efficiency. However, the women's vehicles still ran much better with exercise than without it!

The Bottom Line: Why This Matters

For a long time, we thought exercise was just a quick fix to burn off a cookie. This study says: No, exercise is a lifetime investment.

• It's not just about today: It changes the entire curve of your life.
• It's causal: It's not just that healthy people exercise; exercise actually makes you healthier and keeps your weight in check.
• It works for everyone: Whether you are 30 or 80, moving your body is the best way to keep your "fuel gauge" stable.

In short: If you want to keep your weight steady and avoid the "obesity trap" as you age, the best strategy is to keep moving. It's the only tool that slows down the weight gain in your middle years and cushions the weight loss in your later years.

Technical Summary

1. Problem Statement

High Body Mass Index (BMI) and obesity are critical global health challenges. While leisure-time physical activity (LTPA) is widely recommended for weight management, existing evidence regarding its lifelong impact on BMI trajectories is fragmented.

• Limitations of Current Evidence: Most data comes from short-to-medium-term randomized controlled trials (RCTs) or observational studies. Observational studies are prone to residual confounding (e.g., diet, socioeconomic status) and reverse causality.
• The Gap: There is a lack of robust evidence covering the entire life course (from early adulthood to old age) that establishes a causal link between physical activity and BMI evolution.
• Objective: To determine if higher levels of self-reported LTPA and moderate-to-vigorous LTPA (MVLTPA), as well as genetically determined cardiorespiratory fitness (CRF), are associated with favorable lifelong BMI trajectories and reduced obesity risk.

2. Methodology

The study employed a triangulation approach, combining multivariable regression on real-world data with one-sample Mendelian Randomization (MR) to strengthen causal inference.

Study Population and Data Sources

• Cohorts: Pooled data from three Catalan cohorts within the CORDELIA project: REGICOR-ACRISC, ILERVAS, and ARTPER.
• Sample Size: 14,993 adults (ages 30–90 at baseline; 52.7% women).
• Data Integration:
  • Exposure: Baseline self-reported LTPA (measured via Minnesota or IPAQ questionnaires) and MVLTPA (METs-min/day).
  • Outcome: Repeated BMI measurements (36,157 total measures) extracted from the Catalan electronic health records (PADRIS) spanning 1990–2024.
  • Genetics: Genotype data used to construct a Polygenic Risk Score (PRS) for Cardiorespiratory Fitness (CRF) based on a large UK Biobank GWAS.

Statistical Analysis

1. Multivariable Regression (Observational):

   • Model: Smoothed cubic spline mixed-effects models with participants as random effects.
   • Trajectory Analysis: Modeled BMI changes across age (30–90 years) stratified by quartiles of LTPA/MVLTPA.
   • Obesity Incidence: Analyzed the time to first obesity onset (BMI ≥30 kg/m²) using IPW-weighted Kaplan-Meier curves and Cox proportional hazards models.
   • Adjustments: Birth year, sex, baseline comorbidities (diabetes, hypertension, hypercholesterolemia), smoking, and socioeconomic status.

2. Mendelian Randomization (Causal Inference):

   • Instrument: A CRF-PRS (derived from UK Biobank GWAS) used as a proxy for habitual physical activity.
   • Approach: One-sample MR using the same trajectory and incidence models as the observational analysis, but stratified by quartiles of the CRF-PRS.
   • Assumptions Checked: Relevance (F-statistic), population stratification (adjusted for 10 genetic PCs), and horizontal pleiotropy (tested via negative controls and attenuation analysis).

3. Key Contributions

• Life-Course Perspective: Unlike previous studies focusing on specific age groups or short follow-ups, this study reconstructs BMI trajectories from age 30 to 90 using real-world data.
• Causal Triangulation: By combining observational regression with MR using a physiological proxy (CRF) rather than self-reported behavior, the study mitigates confounding and strengthens causal claims.
• Sex-Stratified Insights: Detailed analysis of how these associations differ between men and women.
• Real-World Validity: Utilization of electronic health records (EHR) provides a large-scale, ecologically valid dataset reflecting routine clinical practice.

4. Key Results

A. Observational Findings (LTPA/MVLTPA)

• BMI Trajectories (Ages 30–60): Higher LTPA was associated with a slower rate of BMI increase.
  • Lowest activity (Q1): +0.120 kg/m²/year.
  • Highest activity (Q4): +0.075 kg/m²/year.
• BMI Trajectories (Ages 70–90): Higher LTPA was associated with a slower rate of BMI decline (likely preserving lean mass).
  • Lowest activity (Q1): -0.143 kg/m²/year.
  • Highest activity (Q4): -0.123 kg/m²/year.
• Obesity Risk: Higher LTPA significantly delayed the onset of obesity.
  • Q4 vs. Q1 Hazard Ratio (HR): 0.83 (95% CI 0.72–0.96).
  • MVLTPA ≥200 METs-min/day showed even stronger protection (HR 0.66 for adequate levels).

B. Mendelian Randomization Findings (Genetically Determined CRF)

• Concordance: The MR results mirrored the observational findings, supporting a causal relationship.
• Trajectories: Higher genetic predisposition to CRF was associated with slower BMI increases in mid-life and slower declines in old age.
• Obesity Risk: Higher CRF-PRS quartiles significantly reduced obesity risk.
  • Q4 vs. Q1 HR: 0.66 (95% CI 0.56–0.78).
• Validity: No evidence of horizontal pleiotropy was found; the genetic instrument was robustly associated with CRF (F-statistic = 7.66).

C. Sex-Specific Differences

• Associations were present in both sexes but were stronger in men.
  • Men (LTPA Q4 vs Q1): HR 0.75.
  • Women (LTPA Q4 vs Q1): HR 0.85.
  • Men (MR Q4 vs Q1): HR 0.61.
  • Women (MR Q4 vs Q1): HR 0.73.

5. Significance and Conclusion

• Causal Evidence: The study provides strong evidence that physical activity causally shapes lifelong BMI trajectories, not just short-term weight loss.
• Mechanistic Insight: The finding that active individuals experience a slower decline in BMI during old age suggests that physical activity helps preserve lean muscle mass, counteracting sarcopenia.
• Public Health Impact: The results support physical activity promotion as a scalable, lifelong strategy for obesity prevention in primary care, with particular efficacy in men.
• Limitations: The study relies on self-reported baseline activity (potential misclassification), stitched EHR data (potential selection bias in older ages), and a modest genetic instrument strength. However, the convergence of observational and genetic evidence suggests residual confounding is unlikely to fully explain the results.

Final Conclusion: Higher leisure-time and moderate-to-vigorous physical activity are causally linked to more favorable BMI trajectories throughout life, characterized by slower weight gain in mid-adulthood, slower weight loss in old age, and a significantly delayed risk of developing obesity.