**5. Results**

moderators was created by multiplying CTQ-SF five scales and the two dimensions of ECR-R. In the Step 1, predictors (childhood traumatic experiences) and moderators (avoidance and anxiety) were entered for direct effects analyses; in Step 2, variables of the previous step were considered along with the ten interaction variables for moderating effects, as suggested by

116 A Multidimensional Approach to Post-Traumatic Stress Disorder - from Theory to Practice

**Figure 1.** Moderation model of childhood traumatic experiences, romantic attachment and PTSD.

**Figure 2.** Mediation model of childhood traumatic experiences, romantic attachment and PTSD.

significant direct and/or mediated effects in the first model.

For the mediation effect hypothesized, a path analysis was performed using LISREL 8.8 [62]. PCL-5 total score has been used as dependent variable and the five scales of the CTQ-SF as independent variables. Avoidance and anxiety in romantic attachment have been considered as mediator variables. In a first saturated model, all the five scales of the CTQ-SF were supposed to have both direct and mediated effects on the post-traumatic symptoms. Secondly, from first results, a second model has been performed including the clinical scales of CTQ-SF presenting

Frazier and Barron [61].


Descriptive statistics of the variables are presented in **Table 2**.

**Table 2.** *Descriptive statistics of CTQ-SF, ECR-R, PCL-5 (N = 327*).

#### **5.1. Prevalence of PTSD symptoms**

**Table 3** shows percentile ranks of PCL-5 and the rate of subjects exceeding cutoff point for the overall PTSD measure.


C.O., cutoff.

\*The provisional diagnosis is calculated considering: at least 1 item of Cluster B scored above 2 point; at least 1 item of Cluster C above 2;, at least 2 items above 2 in Cluster D and 2 items above 2 in Cluster E.

**Table 3.** PCL-5 percentile scores.

#### **5.2. Moderating role of romantic attachment**

**Table 4** presents the hierarchical regression analyses results.

Results from the hierarchical regression analyses showed that PTSD symptoms are significantly predicted using childhood traumatic experiences.

In particular, in the Model 1, both the Emotional Abuse (β = 0.198; *p* = 0.01) and Emotional Neglect (β = 0.148; *p* = 0.03) show significant direct effects on PTSD symptoms; moreover, the physical categories of traumatic experiences (abuse and neglect), display negative effects on PTSD, respectively, with β = −0.136; *p* = 0.03 and β = −0.170; *p* = 0.01.

As regards the romantic attachment, anxiety shows a direct effect on the post-traumatic stress symptoms, β = 0.286; *p* = 0.00, while no effects have been found for the avoidance. Globally, the *R*2 of the first model indicates that childhood traumatic experiences and the romantic attachment account for the 22.7% of PTSD variance.


\**p* < 0.05; \*\**p* < 0.01; \*\*\*\**p* < 0.001.

**Table 4.** Hierarchical regression analyses: direct and moderating effect on PTSD symptoms.

In the Model 2, results show no moderating effects of anxiety and avoidance dimensions, with interaction variables presenting no significant weights. Among childhood traumatic experiences, Emotional Abuse (β = 0.262; *p* = 0.00), Physical Abuse (β = −0.147; *p* = 0.04) and Physical Neglect (β = −0.178; *p* = 0.01) remain significant predictors, while Emotional Abuse shows no significant association with post-traumatic symptoms. In particular, Emotional Abuse has a positive effect, while Physical Abuse and Neglects present negative influences. With regard to romantic attachment, in the second model, both anxiety and avoidance display direct significant effects on PTSD overall symptoms, respectively, β = 0.265; *p* = 0.00 for anxiety and β = −0.124; *p* = 0.04 for avoidance. The *R*<sup>2</sup> shows that Model 2 accounted a small (2.6%) nonsignificant percentage of additional variance of PTSD.

#### **5.3. Mediating role of romantic attachment**

In particular, in the Model 1, both the Emotional Abuse (β = 0.198; *p* = 0.01) and Emotional Neglect (β = 0.148; *p* = 0.03) show significant direct effects on PTSD symptoms; moreover, the physical categories of traumatic experiences (abuse and neglect), display negative effects on

As regards the romantic attachment, anxiety shows a direct effect on the post-traumatic stress symptoms, β = 0.286; *p* = 0.00, while no effects have been found for the avoidance. Globally, the

of the first model indicates that childhood traumatic experiences and the romantic attach-

**β SE B SE**

PTSD, respectively, with β = −0.136; *p* = 0.03 and β = −0.170; *p* = 0.01.

118 A Multidimensional Approach to Post-Traumatic Stress Disorder - from Theory to Practice

**Measures Model 1 Model 2**

Emotional Abuse 0.198\*\* 0.390 0.262\*\* 0.434 Physical Abuse −0.136\* 0.576 −0.147\* 0.649 Sexual Abuse 0.087 0.530 0.083 0.706 Emotional Neglect 0.148\* 0.245 0.133 0.257 Physical Neglect −0.170\*\* 0.631 −0.178\*\* 0.671 Avoidance 0.105 0.044 0.124\* 0.046 Anxiety 0.286\*\* 0.042 0.265\*\*\* 0.044

Emotional Abuse × Anxiety −0.150 1.109 Emotional Abuse × Avoidance 0.058 1.051 Physical Abuse × Anxiety 0.104 1.042 Physical Abuse × Avoidance −0.025 1.383 Sexual Abuse × Anxiety 0.079 0.964 Sexual Abuse × Avoidance −0.103 1.287 Emotional Neglect × Anxiety −0.019 0.972 Emotional Neglect × Avoidance −0.032 0.989 Physical Neglect × Anxiety 0.094 1.148 Physical Neglect × Avoidance −0.027 1.174 Intercept 3.835 3.971 3.655 5.110

*R*<sup>2</sup> 0.227 0.253

In the Model 2, results show no moderating effects of anxiety and avoidance dimensions, with interaction variables presenting no significant weights. Among childhood traumatic experiences, Emotional Abuse (β = 0.262; *p* = 0.00), Physical Abuse (β = −0.147; *p* = 0.04) and Physical

**Table 4.** Hierarchical regression analyses: direct and moderating effect on PTSD symptoms.

ment account for the 22.7% of PTSD variance.

*R*2

*Step 1*

*Step 2*

\**p* < 0.05; \*\**p* < 0.01; \*\*\*\**p* < 0.001.

Results of the first saturated model tested show the childhood traumatic experiences to have mainly direct effects on PTSD symptoms' severity.

From the results of the first model, a second path analysis was performed, removing Sexual Abuse and testing the direct effects of Emotional Abuse, Emotional and Physical Neglect. Moreover, in the second model, the effect of Physical Abuse and Emotional Neglect on avoidance and the effect of Physical Neglect on anxiety were tested.

**Figure 3.** Results of mediation analyses.

For both models, the overall fit was provided by different goodness-of-fit indices, while the path coefficients estimated the relative effect of one variable on another. The goodness-of-fit indices, following Schermelleh and colleagues [63] recommendations, were: the Non-normed Fit Index (NNFI) and the Comparative Fit Index (CFI), both ranging from 0 to 1 with values close to 1 indicating good fit. The root mean square error of approximation (RMSEA) was considered, following Browne and Cudec [64] indications, that is: ≤0.05 considered as a good fit, between 0.05 and 0.08 as an adequate fit, and between 0.08 and 0.10 as a mediocre fit, whereas values >0.10 are considered not acceptable.

Chi-square of the second model was 6.84 df = 6; *p* = 0.34. Considering the definition of fit, all indices show a good fit, with NNFI and CFI equal to 1 and RMSEA 0.021.

Path coefficients indicate significant direct and mediating effects. In particular, in the second model, anxiety and avoidance mediate the effect of Emotional Neglect on PTSD (β = 0.12) and avoidance shows a tending to significance mediation of Physical Abuse on PTSD (β = −0.02; *p* = 0.091).

For the direct effects, Emotional Abuse shows significant direct effect (β = 0.14) on PTSD; Physical Abuse shows no direct influence on PTSD, but a significant negative effect on avoidance (β = −0.15). Emotional Neglect has direct effect on PTSD as well as on avoidance and anxiety, respectively 0.16, 0.34 and 0.27. The Physical Neglect only fits negatively to PTSD symptoms (−0.16).

Results of the mediation analyses are presented in **Figure 3**.
