Table 2. Associations Between Body Mass Index Variability and Composite Adverse Cardiovascular Events.
| Harmonized ancestry and race and ethnicity | HR (95%CI)a | P value |
|---|---|---|
| Hispanic | ||
| Model 1b | 1.24 (1.14-1.35) | <.001 |
| Model 2c | 1.24 (1.14-1.35) | <.001 |
| Model 3d | 1.25 (1.10-1.40) | .003 |
| Non-Hispanic Black | ||
| Model 1b | 1.16 (1.10-1.21) | .001 |
| Model 2c | 1.16 (1.10-1.21) | <.001 |
| Model 3d | 1.13 (1.05-1.22) | .005 |
| Non-Hispanic White | ||
| Model 1b | 1.15 (1.12-1.19) | <.001 |
| Model 2c | 1.15 (1.12-1.19) | <.001 |
| Model 3d | 1.11 (1.07-1.16) | <.001 |
| Overall cohort | ||
| Model 1b | 1.16 (1.13-1.19) | <.001 |
| Model 2c | 1.16 (1.13-1.19) | <.001 |
| Model 3d | 1.13 (1.09-1.16) | <.001 |
Abbreviation: HR, hazard ratio.
a
Per 1 SD of body mass index variability.
b
Model 1 adjusted for age, sex, mean body mass index, diabetes, smoking history, high-density lipoprotein, systolic blood pressure, blood pressure therapy, total cholesterol level, and statin use.
c
Model 2 adjusted for model 1 covariates plus polygenic risk score for body mass index trait.
d
Model 3 adjusted for model 1 covariates plus physical activity (n = 45 038).