A Call to Validate the Stress Continuum Model

ABSTRACT Introduction Since the start of the Global War on Terrorism, exponential demands have been put on military personnel, their families, and the military health care system. In response to a Department of Defense Task Force on Mental Health, the U.S. military began developing and fielding programs to promote the psychological health of its personnel. As part of these initiatives, the Navy and Marine Corps developed the Stress Continuum model. The Stress Continuum is a stress classification system (“ready,” “reacting,” “injured,” and “ill”) that provides a common language for identifying, engaging, and intervening when stress reactions or stress injuries are present in military personnel. It is the foundation for resilience and prevention efforts across the Navy and Marine Corps. Although the Stress Continuum has strong face validity, is consistent with current theory, and has been agreed up by expert consensus, it has yet to be empirically validated. The goal of the current article is to begin to empirically validate the Stress Continuum using validated measures of psychological stress. Materials and Methods We conducted a retrospective analysis of Stress Continuum data (n = 2,049) collected as part of a program evaluation of two Navy operational stress control programs. Receiver operating characteristic (ROC) curves and analyses were conducted to determine the classification quality of the Stress Continuum using a validated measure of stress (a brief version of the Perceived Stress Scale [PSS-4]). Results For the first ROC curve, we used the “ill” category (vs. the other three categories) to identify the cut point on the PSS-4. PSS-4 cut point values of 9 and 10, respectively, maximized sensitivity and 1-specificity values. Using the chi-square test, we further found that a more accurate prediction for those in the “ill” category was using the cut point of 9 (79%) relative to 10 (71.8%). For the second and the third ROC curves, we used the “ill” and “injured” categories (vs. the other two categories) and “ill,” “injured,” and “reacting” categories (vs. the “ready” category), respectively. No optimal cut points on the PSS-4 were identified for these models, indicating that the PSS-4 could not reliably differentiate true-positive and false-positive rates. Conclusions We found that the “ill” category of the Stress Continuum was predictive of higher levels of stress on the validated measure of perceived stress. Thus, our findings strongly suggest that t