Catch me if you can: do the Five-Factor Model personality traits moderate dropout and acute treatment response in post-traumatic stress disorder patients?

sulting in a sample size of n = 123. FFM traits were measured at baseline with the NEO-Five Factor Inventory [22, 23]ater - d f . Sl ePTS symptoms were measured at baseline and posttest with the Impact of Event Scale [24] . PTSD diagnostic status was estab-lished at baseline and posttest with the Structured Clinical Inter-view for DSM-IV Axis I disorders [25] . Treatment consisted of a maximum of 10 1.5-hour sessions of TFCBT or structured writing therapy, and was compared to waitlist control (WLC) (for com-plete study procedures and results, see Van Emmerik et al. [21] ).Following Kraemer et al. [26] , linear multiple regression models were constructed for each FFM trait with: (1) baseline PTS symp-toms entered first (step 1), (2) main effect of each FFM trait, main effect of study condition (active treatment or not), and interaction of FFM trait and study condition entered second (step 2), and (3) posttest PTS symptoms as the dependent variable. Of note, sig-nificant interaction terms indicate differential relationships be-tween FFM traits and baseline-to-posttest changes in PTS symp-toms after active treatment compared to WLC, and thus provide a stringent test of whether FFM traits moderate treatment out-come. Similarly, logistic multiple regression models were con-structed with dropout and posttest PTSD diagnostic status as (bi-nary) dependent variables. Forty-one randomized controlled trial participants (33.3%) dropped out and did not complete the posttest. None of the inter-actions between FFM traits and study condition were significant (all p 1 0.10), indicating that the FFM traits did not moderate dropout. For the 82 completers (66.7%), baseline PTS symptoms were strongly associated with treatment outcome in each FFM multiple regression model (all p ! 0.001), as was study condition (i.e. receiving active treatment; all p ! 0. 001). In contrast, no main effects were observed for any of the FFM traits (all p 1 0.10). Ta-ble 1 displays the regression model yielding the one significant FFM moderator, i.e. openness to experience (p ! 0.05). Lower scores on openness to experience were more strongly associated with posttest PTS symptoms after active treatment than after WLC, indicating that this FFM trait moderated PTSD treatment outcome. No significant interactions were observed between ac-tive treatment and neuroticism, extraversion, or agreeableness (all p 1 0.05), and a marginally significant interaction was observed between active treatment and conscientiousness (p = 0.05). Study condition (i.e. receiving active treatment) was also associated with posttest (no-PTSD) diagnostic status across the FFM logistic mul-tiple regression models. No significant interactions emerged be-tween study condition and Five-Factor Inventory traits, suggest-ing that FFM traits do not moderate change in PTSD diagnostic status.O ur main findings can be summarized as follows. First, base-line self-reported PTS symptoms were predictive of posttest levels of PTS symptoms. This is in line with previous studies, which in Research has identified trauma-focused cognitive-behavioral therapy (TFCBT) as an effective treatment for post-traumatic stress disorder (PTSD) [1] . However, a substantial minority of PTSD patients do not sufficiently benefit from TFCBT [2–10] . Many clinicians would explain differential PTSD treatment out-come at least partly in terms of their patients’ personality charac-teristics. This contrasts with the scarcity of empirical studies on personality characteristics as predictors of PTSD treatment out-come [11–16] . Furthermore, those studies that do exist employed heterogeneous operationalizations of personality characteristics, which, perhaps accordingly, failed to demonstrate consistent re-lationships with PTSD treatment outcome. This is unfortunate because moderator analyses of personality characteristics could potentially yield information on mechanisms of change or the tailoring of treatment to individuals with extreme personality rat-ings. We therefore investigated the Five-Factor Model (FFM) per-sonality traits as moderators of dropout and treatment response in PTSD patients. The FFM (neuroticism, extraversion, openness to experience, agreeableness, and conscientiousness) has estab-lished itself as a widely accepted and extensively researched per-sonality model [17] , and may potentially influence PTSD treat-ment outcome in various ways. For instance, a minimal degree of openness to experience is a likely prerequisite for involvement in emotionally aversive exposure procedures or in cognitive restruc-turing exercises, and some degree of conscientiousness may be essential for consistent attendance and homework compliance [18–20] . Sp ecifically, we hypothesized that neuroticism was asso-ciated with dropout, and that lower scores on openness to experi-ence, agreeableness, and conscientiousness, and higher scores on neuroticism, were associated with less favorable treatment out-come.D ata were obtained from patients meeting DSM-IV criteria for PTSD in a previous randomized controlled trial [21] . Two par-ticipants missing post-treatment PTSD data were excluded, re