The Stress Bias in Mental Health Reporting: Death Anxiety Biases Mental Health Self-Assessments Amongst Deployed Soldiers

摘要

This study examined whether posing questions that create a stress state in respondents biases subsequent reporting of mental health (MH) symptomology. For instance, questions related to trauma exposure may activate thoughts about death (facing it, surviving it, being afraid of it), and these death anxieties can influence assessments of one's health and thus bias self-reports of MH measured in the same questionnaire context. A controlled randomized experiment was conducted as part of a longitudinal study of U.S. Army soldiers who deployed to Afghanistan between April 2013 and January 2014. Anonymous surveys administered middeployment (n = 464) included self-report MH screening instruments for posttraumatic stress disorder, depression, and generalized anxiety disorder. Combat exposures (CEs) were primed by randomly presenting the CE questions before or after MH instruments. The order of the measures significantly influenced reports of MH symptoms, and this stress bias is explained by differences in death anxiety. Moderated mediation analyses revealed that the well-documented relationship between the intensity of CEs and each MH screener severity score was significant when the CEs were primed but not vice versa (index of moderated mediation: for PTSD, 95% CI [.0000,.0216]; for depression, 95% CI [.0000,.0216]; and for generalized anxiety disorder, 95% CI [.0005,.0170]), and that death anxiety mediated these relationships. Health care providers and psychological researchers who use standardized, nonrandomized, clinical screening instruments should be aware of this stress bias: Asking respondents to report a potentially traumatic exposure can bias the self-reporting of MH symptomatology.