| Characteristic | Included (N = 4316) | Excluded (N = 13997)* | |
| IQ at army entry, mean, SD | 106.1 (20.1) | 103.8 (20.4) | |
| Age at telephone interview†, years (% (N)) | ⩽34 | 12.2 (527) | 13.0 (1818) |
| 35–39 | 68.0 (2935) | 67.5 (9451) | |
| ⩾40 | 19.8 (854) | 19.5 (2728) | |
| Army income, US$/month (% (N)) | 145 | 38.0 (1640) | 35.0 (4896) |
| 120–144 | 49.5 (2135) | 48.2 (6751) | |
| 83–119 | 12.5 (541) | 16.8 (2350) | |
| Place of service (% (N)) | Served in USA only | 19.0 (818) | 21.8 (3051) |
| Other overseas posting | 25.7 (1110) | 28.7 (4010) | |
| Ever served in Vietnam | 55.3 (2388) | 49.6 (6936) | |
| Ethnic group (% (N)) | White | 81.8 (3529) | 83.0 (9104) |
| Black | 11.8 (510) | 10.8 (1184) | |
| Other | 6.4 (277) | 6.2 (684) | |
| Hazard ratio‡ (95% CI) | 0.71 (0.63 to 0.81) | 0.78 (0.73 to 0.84) |
*Numbers vary slightly as a result of missing data. †For those in the excluded group who did not complete a telephone interview, age is taken at the midpoint date (31 July 1985) of the telephone interviews. ‡Hazard ratios are for a 1 SD increase in IQ in relation to all-cause mortality and are adjusted for age. In these analyses, mortality follow-up is from the date of telephone interview or the midpoint date (31 July 1985) for those who did not participate in the telephone interview; the 499 subjects who died before 31 July 1985 are excluded from this analysis. With the exception of age at telephone interview, all data were extracted from military records.