Longevity and VO₂ max

VO₂ max, a measure of aerobic fitness, is one of the strongest known predictors of long-run mortality: in Mandsager et al. (2018), the mortality gradient across fitness levels is steeper than the one associated with end-stage kidney disease. This page converts my watch's VO₂ max estimates into an implied life expectancy. The exercise is deliberately back-of-the-envelope; the methodology is below.

Garmin estimates VO₂ max (mL/kg/min) from heart-rate measurements during exercise and recovery. Wrist-based heart rate is noisy, but these estimates have been validated against laboratory tests. I smooth the raw estimates with a rolling average.

Each VO₂ max level maps to a mortality hazard ratio relative to the median individual, using the gradient in Mandsager et al. (2018). I fit a natural cubic spline through their estimates so that any VO₂ max value yields a hazard ratio.

Survival by fitness level, Mandsager et al. (2018)

Survival by cardiorespiratory fitness group, from Mandsager et al. (2018).

Baseline survival curves by age and sex come from CDC mortality data (2018–2021, so the pandemic years push them down a bit). Applying the hazard ratio under a proportional-hazards assumption shifts the survival curve, and integrating the shifted curve gives period life expectancy. A higher VO₂ max means a hazard ratio below one, more area under the survival curve, and a longer implied life. Code and data are in the GitHub repository.

Methodology