Mandsager et al. published a study in JAMA Network Open in 2018 that should have ended the debate about exercise and longevity. They followed 122,007 consecutive patients who underwent treadmill stress testing at the Cleveland Clinic between 1991 and 2014. Compared to patients in the top quartile of cardiorespiratory fitness, those in the bottom quartile had a hazard ratio for death of 5.04. For context, the hazard ratio for current smoking versus never smoking is approximately 1.8. Low fitness was a stronger predictor of death than smoking, diabetes, or coronary artery disease.

The dose-response curve did not plateau. Moving from low fitness to below-average fitness cut mortality risk by approximately 50%. Moving from below-average to above-average cut it further. Contrary to the belief that extreme exercise is harmful, the lowest mortality rates were in the elite fitness group (top 2.3%). The U-shaped curve some earlier studies suggested did not survive adjustment for confounding variables.

VO2max measures the maximum rate at which your body can transport and use oxygen during exercise. An average 40-year-old man has a VO2max of about 35–40 mL/kg/min. Elite endurance athletes reach 70–85. By age 80, the average drops to 18–20, close to the threshold for independent living (approximately 15–18 mL/kg/min). VO2max declines approximately 10% per decade after age 30. If you start at average and decline at the standard rate, you cross the threshold of functional independence in your late 70s. If you build and maintain above-average fitness, you delay that crossing by 10–20 years.

The Arem et al. 2015 study in JAMA Internal Medicine analyzed pooled data from six prospective cohorts totaling 661,137 adults. Meeting US guidelines of 150 minutes/week of moderate activity was associated with a 31% mortality reduction compared to inactivity. Two to three times the guidelines (300–450 minutes/week) provided 37%. Even the most active group showed no increase in mortality risk.

Ekelund et al. (2019) conducted a harmonized meta-analysis of accelerometer data from 44,370 adults across four continents. Any intensity of movement was associated with reduced mortality. Replacing 30 minutes of sedentary time with moderate-to-vigorous activity was associated with a 35% mortality reduction. The largest marginal gains came from the first increments of activity added by the most sedentary individuals.

Resistance training contributes independently. A 2022 meta-analysis by Momma et al. found that 30–60 minutes of muscle-strengthening activity per week was associated with 10–20% lower risk of all-cause mortality. The optimal program combines aerobic and resistance training — VO2max for cardiovascular health, muscle mass for metabolic health and functional independence.

The most efficient method to improve VO2max is high-intensity interval training: 4×4 minutes at 85–95% of maximum heart rate, 2–3 times per week, has been shown to increase VO2max by 5–10% over 8–12 weeks in previously sedentary adults. Zone 2 training (sustained aerobic exercise at conversational intensity) builds mitochondrial density. Most longevity-focused programs recommend 3–4 hours of Zone 2 plus 1–2 HIIT sessions per week.

Exercise activates AMPK, increases PGC-1α (the master regulator of mitochondrial biogenesis), upregulates BDNF, promotes autophagy, reduces systemic inflammation, improves insulin sensitivity, and enhances endothelial function. These pathways overlap with the mechanisms targeted by metformin, rapamycin, and NAD+ precursors — except exercise activates all of them simultaneously, for free, with a side-effect profile of improved mood, sleep, and body composition.

No pharmaceutical intervention comes close to matching exercise's effect size on all-cause mortality. The hazard ratio reduction from elite versus low fitness (approximately 80%) dwarfs metformin (15–20% in observational data), rapamycin (no human mortality data), and NAD+ precursors (no mortality data). The evidence base spans millions of participant-years. The limiting factor is not evidence. It is behavior change.