Table 1.
Descriptive statistics on the study sample (National Health and Nutrition Examination Survey, 1999–2010 linked to the 2011 National Death Index, n=41 990)
| Training data for model derivation | Test data for model evaluation | P value for difference* | |
| n=29 390 | n=12 600 | ||
| CVD death | |||
| No | 28 211 (96.0) | 12 093 (96.0) | 0.96 |
| Yes | 1179 (4.0) | 507 (4.0) | |
| Heart disease death | |||
| No | 28 507 (97.0) | 12 214 (96.9) | 0.76 |
| Yes | 883 (3.0) | 386 (3.1) | |
| Cerebrovascular death | |||
| No | 29 094 (99.0) | 12 479 (99.0) | 0.71 |
| Yes | 296 (1.0) | 121 (1.0) | |
| Time since interview (months) | 79.3 (±41.4) | 79.4 (±41.6) | 0.84 |
| Wave | |||
| 99–00 | 3810 (13.0) | 1633 (13.0) | 1.0 |
| 01–02 | 8853 (30.1) | 3795 (30.1) | |
| 03–04 | 3926 (13.4) | 1684 (13.4) | |
| 05–06 | 3891 (13.2) | 1669 (13.2) | |
| 07–08 | 4353 (14.8) | 1866 (14.8) | |
| 09–10 | 4557 (15.5) | 1953 (15.5) | |
| Age | 50.0 (±20.4) | 50.1 (±20.6) | 0.60 |
| Sex | |||
| Male | 13 924 (47.4) | 5887 (46.7) | 0.22 |
| Female | 15 466 (52.6) | 6713 (53.3) | |
| Black | |||
| No | 14 807 (50.4) | 6335 (50.3) | 0.94 |
| Yes | 5882 (20.0) | 2511 (19.9) | |
| Missing | 8701 (29.6) | 3754 (29.8) | |
| Hispanic | |||
| No | 21 871 (74.4) | 9359 (74.3) | 0.77 |
| Yes | 7519 (25.6) | 3241 (25.7) | |
| Education level | |||
| <9th | 3942 (13.4) | 1756 (13.9) | 0.087 |
| 9–11 | 4538 (15.4) | 1954 (15.5) | |
| HS degree | 6543 (22.3) | 2716 (21.6) | |
| Some college or Associate’s | 7138 (24.3) | 2986 (23.7) | |
| College degree | 5061 (17.2) | 2268 (18.0) | |
| Missing | 2168 (7.4) | 920 (7.3) | |
| Ratio of family income to poverty threshold | 2.5 (±1.6) | 2.5 (±1.6) | 0.59 |
| Missing | 2655 (9.0) | 1109 (8.8) | |
| Total cholesterol | 198.0 (±43.1) | 198.0 (±43.9) | 0.86 |
| Missing | 3641 (12.4) | 1484 (11.8) | |
| HDL | 45.5 (±23.0) | 45.6 (±23.0) | 0.36 |
| Missing | 3643 (12.4) | 1484 (11.8) | |
| SBP | 125.4 (±20.6) | 125.6 (±21.1) | 0.38 |
| Missing | 3175 (10.8) | 1348 (10.7) | |
| DBP | 69.9 (±12.6) | 69.8 (±12.7) | 0.50 |
| Missing | 3374 (11.5) | 1431 (11.4) | |
| Number of blood pressure medications | |||
| 0 | 19 892 (67.7) | 8436 (67.0) | 0.32 |
| 1 | 7851 (26.7) | 3452 (27.4) | |
| 2 or more | 1647 (5.6) | 712 (5.7) | |
| Type 2 diabetes | |||
| No | 10 537 (35.9) | 4541 (36.0) | 0.42 |
| Yes | 4783 (16.3) | 2008 (15.9) | |
| Missing | 14 070 (47.9) | 6051 (48.0) | |
| Smoking | |||
| No | 23 774 (80.9) | 10 185 (80.8) | 0.90 |
| Yes | 5615 (19.1) | 2414 (19.2) | |
| Missing | 1 (0.0) | 1 (0.0) | |
| HEI | 47.0 (±11.0) | 47.2 (±11.0) | 0.28 |
| Missing | 3277 (11.2) | 1361 (10.8) | |
| AHEI | 47.1 (±11.1) | 47.1 (±11.0) | 0.76 |
| Missing | 3263 (11.1) | 1353 (10.7) | |
| MDS | 5.1 (±1.2) | 5.1 (±1.2) | 0.095 |
| Missing | 3270 (11.1) | 1368 (10.9) | |
| DASH | 47.4 (±9.3) | 47.4 (±9.4) | 0.75 |
| Missing | 8835 (30.1) | 3661 (29.1) |
Mean (±SD) reported for continuous variables and N (%) reported for categorical variables.
Statistics are grouped to reflect participants in the training (n=29 390/41 990=70%) or test (n=12 600/41 990=30%) data subsets.
*Wilcoxon rank sum test for continuous variables, eg, age, and Fisher’s exact test for categorical variables, eg, black race.
AHEI, Alternative Healthy Eating Index; CVD, cardiovascular disease; DASH, Dietary Approaches to Stop Hypertension diet score; HDL, high-density lipoprotein; HEI, Healthy Eating Index; MDS, Mediterranean Diet Score.