Table 2.

Association of Metabolomic Fitness Score and Clinical Risk Factors With Peak VO₂

Variable	β coefficient±SE	P value
Age	−0.033±0.003	<0.0001
Female sex	−0.831±0.055	<0.0001
Cardiovascular disease	−0.084±0.106	0.43
Diabetes	−0.356±0.091	0.0001
Hypertension treatment	−0.099±0.058	0.09
Smoking	−0.193±0.043	<0.0001
Log body mass index	−1.144±0.100	<0.0001
Log systolic blood pressure	0.210±0.136	0.12
Log total cholesterol	−0.209±0.087	0.02
Log HDL cholesterol	0.210±0.059	0.0004
Metabolomic fitness score	0.303±0.025	<0.0001

This linear model was built in the Framingham Heart Study validation cohort (N=914 individuals). Logarithms are base 2. The dependent variable in this model is peak oxygen uptake (VO₂ ) (expressed as milliliters per kilogram per minute, log‐transformed, and standardized). The ratio of the estimated regression coefficient for the metabolomic fitness score (0.303 per ≈1 SD in the score) was divided by the regression coefficient for age (−0.033 per year) to yield the relative association effect size of 1 SD of the score relative to years of age (≈9.2 years lower age has an equivalent association with log‐peak VO₂ as ≈1 SD higher in the metabolomic fitness score).