TABLE III.
Model 1: Multivariable Logistic Regression Model Predicting a Clinically Significant Increase in PTSD Score Among Those without Full PTSD at Baselinea,b
| Characteristic | OR | 95% CI |
|---|---|---|
| PTSD symptoms at baseline | ||
| No symptoms | 1.0 | |
| Sub-PTSD | 6.4* | 3.2-8.1 |
| Combat exposure level | ||
| None/low | 1.0 | |
| Moderate | 2.1* | 1.2-3.6 |
| High | 1.4 | 0.8-2.5 |
| Positive coping behaviors | ||
| 0-1 | 2.1* | 1.3-3.2 |
| 2-3 | 1.3 | 0.7-2.6 |
| 4-5 | 1.0 | |
Abbreviations: OR, odds ratio; PCL-C, PTSD Checklist—Civilian Version; sub-PTSD, subsyndromal PTSD.
a
Additional variables that were originally entered into the backward elimination model include age, reason for separation, postdeployment social support, and depressive symptomatology.
b
This number includes only respondents who provided answers for all covariates entered into the final model (n = 899).
*
P < .05.