Table 1.
Baseline characteristics of the CARDIA study population
| Characteristic | Overall n = 2,238 | Derivation n = 1,569 | Validation n = 669 | P value |
|---|---|---|---|---|
| Age (years) | 51.0 (47.0, 53.0); 0% | 50.0 (47.0, 53.0); 0% | 51.0 (48.0, 54.0); 0% | 0.015 |
| Sex, n (%) | >0.9 | |||
| Male | 978 (44%); 0% | 686 (44%); 0% | 292 (44%); 0% | |
| Female | 1,260 (56%); 0% | 883 (56%); 0% | 377 (56%); 0% | |
| Race, n (%) | 0.3 | |||
| Black | 973 (43%); 0% | 670 (43%); 0% | 303 (45%); 0% | |
| White | 1,265 (57%); 0% | 899 (57%); 0% | 366 (55%); 0% | |
| CARDIA Field Center, n (%) | 0.7 | |||
| Birmingham | 531 (24%); 0% | 362 (23%); 0% | 169 (25%); 0% | |
| Chicago | 564 (25%); 0% | 403 (26%); 0% | 161 (24%); 0% | |
| Minnesota | 523 (23%); 0% | 368 (23%); 0% | 155 (23%); 0% | |
| Oakland | 620 (28%); 0% | 436 (28%); 0% | 184 (28%); 0% | |
| Body mass index (kg m−2) | 29 (25, 33); <0.1% | 29 (25, 33); <0.1% | 28 (25, 33); 0% | 0.8 |
| Lifetime smoking pack years | 0 (0, 5); 0% | 0 (0, 5); 0% | 0 (0, 7); 0% | 0.5 |
| Systolic blood pressure (mmHg) | 116 (108, 126); <0.1% | 116 (107, 126); 0% | 116 (108, 125); 0.1% | 0.7 |
| Diastolic blood pressure (mmHg) | 73 (66, 80); <0.1% | 73 (66, 80); 0% | 72 (66, 80); 0.3% | 0.8 |
| Treated for hypertension, n (%) | 583 (26%); 0% | 395 (25%); 0% | 188 (28%); 0% | 0.15 |
| Diabetes, n (%) | 313 (14%); 0% | 210 (13%); 0% | 103 (15%); 0% | 0.2 |
| History of CVD | 44 (2.0%); 0% | 36 (2.3%); 0% | 8 (1.2%); 0% | 0.5 |
| eGFR (ml min−1 1.73m−2) | 94 (82, 107); <0.1% | 93 (82, 106); <0.1% | 94 (83, 108); 0.1% | 0.087 |
| Total cholesterol (mg dl−1) | 190 (167, 215); 0% | 190 (167, 215); 0% | 190 (166, 215); 0% | 0.5 |
| High density lipoprotein (mg dl−1) | 55 (45, 67); 0% | 56 (45, 67); 0% | 54 (45, 67); 0% | 0.7 |
| Year 20 ETT time (s) | 420 (304, 539); 0% | 420 (304, 539); 0% | 420 (304, 539); 0% | >0.9 |
The study population was split into derivation/validation samples, balanced by Year 20 ETT time. Continuous variables are reported at median (25th, 75th percentile) with percentage missingness. Categorical variables are reported as n (%) with percentage missingness. Reported P values are from two-sided Wilcoxon tests (for continuous variables) and two-sided Chi-square tests (categorical variables).