Abstract P229: Cross-Sectional Relationship Between Posttraumatic Stress Disorder and Atherosclerotic Disease Assessed by Coronary Calcium Score: A Counterfactual Model of Older Male Twins

Abstract

Background: Posttraumatic stress disorder (PTSD) is a risk factor for cardiovascular disease. However, the causal relationship between PTSD and atherosclerotic burden is controversial, leaving uncertain whether atherosclerosis is implicated in the association between PTSD and cardiovascular events. Objective: To test the cross-sectional relationship between lifetime PTSD and coronary atherosclerotic burden assessed by calcium score by comparing PTSD-discordant twin pairs, a design that inherently controls for familial and early environmental factors. Methods: We examined 218 male twins from the Vietnam Era Twin Registry, who underwent coronary calcium score assessment with computed tomography as part of the Emory Twin Study Follow-Up study. Lifetime history of PTSD was obtained with the Structured Clinical Interview and PTSD symptom severity with the Clinician-Administered PSTD Scale (CAPS) for DSM-IV diagnostic criteria. Coronary calcium scoring was performed by the Agatston Method as a measure of atherosclerotic burden. The relationship between PTSD and calcium score was tested across individuals and within PTSD-discordant twin pairs as a counterfactual model for genetic and early environmental factors. Initially, 2 linear mixed models were built according to the exposure variable (PTSD clinical diagnosis or symptoms scale), with log-transformed coronary calcium score as the dependent variable. We defined a priori that the model with the lowest Akaike Information Criteria (AIC) would be the primary analysis to drive the conclusion. Based on a priori direct acyclic graph decision process, we defined the need to adjust for income and education as potential confounders and not to adjust for traditional risk factors, considering their potential roles as mediators. Results: Of the 218 twins analyzed, the median age was 68 (IQR 67-70) years, and the calcium score had a median of 114 (IQR 5.2-347). There were 88 twin pairs (30 monozygotic), 26 discordant for the binary diagnosis of PTSD, and 40 discordant for the ordinal PTSD symptom score (within-pair differences in scale points ranged from zero to 48.5). The model with the best AIC was the one with the PTSD score, and showed no increment of log-calcium score as the PTSD symptoms increased, both between individuals (Beta coefficient = - 0.005; 95% CI = -0.02 to + 0.03) and within twin pairs (Beta coefficient = - 0.010; 95% CI = -0.03 to + 0.01). Conclusion: The present analysis does not support a relationship between PTSD and coronary atherosclerotic burden. The link between PTSD and cardiovascular events may occur through other pathways, such as microvascular disease.