Table 2.
Baseline characteristics | Overall (n=7532) | Hospital admissions | P value† | ||
---|---|---|---|---|---|
None (n=4954) | Any surgery (n=1863) | Medical only (n=715) | |||
Mean (SD) age (years) | 55.5 (5.99) | 55.0 (5.95) | 56.6 (5.88) | 56.4 (6.1) | <0.001 |
Women | 2235 (29.7) | 1465 (29.6) | 599 (32.2) | 171 (23.9) | <0.001 |
Race: | |||||
White | 6874 (91.5) | 4553 (92.0) | 1694 (91.1) | 627 (87.9) | 0.001 |
Non-white | 645 (8.6) | 394 (8.0) | 165 (8.9) | 86 (12.0) | |
Missing | 13 | 7 | 4 | 2 | |
Education level (years): | |||||
0-12 | 1958 (29.1) | 1297 (29.2) | 501 (30.1) | 160 (25.6) | 0.09 |
13-17 | 3000 (44.6) | 2000 (45.0) | 704 (42.3) | 296 (47.4) | |
≥18 | 1771 (26.3) | 1145 (25.8) | 458 (27.5) | 168 (26.9) | |
Missing | 803 | 512 | 200 | 91 | |
Maximum employment position: | |||||
High | 3413 (45.3) | 2305 (46.5) | 784 (42.1) | 324 (45.3) | 0.02 |
Intermediate | 3219 (42.7) | 2077 (41.9) | 835 (44.8) | 307 (42.9) | |
Low | 900 (12.0) | 572 (11.6) | 244 (13.1) | 84 (11.8) | |
Mean (SD) Framingham CVD risk score‡ | 0.20 (0.12) (n=6792) | 0.19 (0.11) (n=4578) | 0.21 (0.12) (n=1600) | 0.22 (0.12) (n=614) | <0.001 |
Diabetes | 889 (11.8) | 495 (10.0) | 277 (14.9) | 117 (16.3) | <0.001 |
Smoking: | |||||
Current | 684 (10.0) | 438 (9.7) | 180 (10.7) | 66 (10.4) | 0.21 |
Former | 2774 (40.7) | 1803 (40.1) | 716 (42.5) | 255 (40.3) | |
Never | 3353 (49.2) | 2255 (50.1) | 787 (46.8) | 311 (49.1) | |
Missing | 721 | 459 | 180 | 82 | |
Medical and surgical events: | |||||
Mean (SD) surgical admissions | 0.39 (0.85) | 1.6 (1.0) | |||
Stroke | 63 (0.8) | 28 (1.5) | 35 (4.9) | ||
Any medical admission | 1328 (18) | 613 (33) | 715 (100) | ||
Mean (SD) medical admissions | 0.29 (0.84) | 0.60 (1.2) | 1.5 (1.2) | ||
Mean (SD) follow-up: | |||||
No of assessments | 3.8 (1.4) | 3.6 (1.5) | 4.1 (1.1) | 3.9 (1.2) | <0.001 |
Duration (years) | 14.6 (5.56) | 13.8 (6.25) | 16.3 (3.25) | 15.7 (3.85) | <0.001 |
Categories are based on events occurring before each participant’s last cognitive assessment. Missing values are excluded from percentages.
P values are from I2 tests (categorical) or one way analysis of variance F tests (continuous) comparing across never surgery or never medical, any surgery, and medical only categories.
Averages and standard deviations are based on subset of participants owing to missing values.