Type D personality predicts the occurrence of major cardiac events in patients with coronary artery disease: A multi-method analysis

Aim
Previous studies have shown that cardiovascular patients with a Type D personality (a combination of having high negative affectivity and high social inhibition), are at increased risk of recurring major cardiac events. However, some have criticized the methods used to arrive at this conclusion. This study aims to replicate earlier findings by assessing the robustness of the effect of Type D personality on cardiac outcomes using three commonly used statistical models.

Methods
In a sample of 541 patients with coronary artery disease, Type D personality was measured using the DS14 questionnaire and cardiac events were registered at a 5-year follow-up. Type D personality was modeled statistically as an interaction between its two subcomponents NA and SI. The effect of this interaction on cardiac events was modeled using three different logistic regression analyses: (1) logistic regression using the observed NA and SI scores, (2) logistic regression using the NA and SI factor scores, and (3) logistic regression using Structural Equation Modeling (SEM).

Results
The interaction between NA and SI significantly predicted the occurrence of cardiac events in each of the three logistic regression models, but the methods differed in the size of the estimated interaction, with odds ratio's varying from 2.47 (method 1), to 2.65 (method 2) and 3.71 (method 3). These differences reflect the assumptions underlying each of these statistical models.

Conclusion
The association between Type D personality and cardiac events appears robust across three commonly used statistical models to assess interaction effects on dichotomous outcomes.