*2.3. Measures*

The Gra ffar Index was used to measure the socioeconomic status (SES) [37,38]. This Index classifies subjects into five classes, with 1 being the highest, and 5 the lowest SES class. The clinical history and other sociodemographic data were also collected.

The Clinician-Administered PTSD Scale (CAPS) [39] was used to characterize the participants in relation to PTSD diagnosis, and also to research the lifetime number and the type of TEs experienced. Lifetime PTSD was considered if participants had DSM-IV criteria, in accordance with Blake et al.'s [40] rule (frequency ≥1 and intensity ≥2) and a total CAPS score of 50 or more. TEs were assessed with the CAPS Life Events Checklist and were subsequently checked for DSM-IV A2 criterion, following the CAPS procedure. TEs occurring before and after the war were counted separately from each other, and also from war-related TEs. In our sample, the Cronbach's alpha (reliability) of the CAPS was superior to 0.90.

To determine participants' eligibility requirements for the study, current and past psychiatric disorders where investigated, using the Structured Clinical Interview for DSM-IV axis I (SCID-I) [41] except for the PTSD module.

The Childhood Trauma Questionnaire-Short Form (CTQ-SF) [42,43] is a retrospective, self-reported questionnaire which contains 28 questions about specific maltreatment experiences during childhood and adolescence. The items are classified into a 5-point ordinal scale, according to the frequency of

exposure to that specific experience. The Total CTQ-SF score provides a general ACE score, and not just TEs. It also provides scores for five di fferent types of maltreatment, namely: emotional abuse; physical abuse; sexual abuse; emotional neglect, and; physical neglect.

Attachment style was studied with the Revised Adult Attachment Scale (RAAS) [44,45]. This scale consists of 18 items, which were scored on a 5-point Likert scale. The scale contains three dimensions: anxiety, close, and depend. The close and depend dimensions are positively correlated and can be gathered as close–depend. Close and depend dimensions were averaged and were then reverse scored to yield the attachment-related avoidance dimension. According to the scores obtained for each dimension, Bartholomew's [46] attachment styles classification was adopted, namely: secure, dismissing, preoccupied, and fearful. These attachment styles are described as follows (adapted from [44]): secure—those participants who scored an average score below the midpoint (3.0) on avoidance, and below or equal to the midpoint on anxiety; dismissing—those whose scoring was above or equal to the midpoint on avoidance, and below or equal to the midpoint on anxiety; preoccupied—those scoring below the midpoint on avoidance, and above the midpoint on anxiety; fearful—those scoring above or equal to the midpoint on avoidance, and above the midpoint on anxiety. Those participants with dismissing, preoccupied, and fearful styles were also grouped in the insecure style [47].

In order to characterize and quantify the di fferent war-related experiences of the veterans, we constructed the War Exposure Questionnaire (WEQ; see Supplementary Materials)—which was adapted to the specificities of the guerrilla war where each veteran fought. This questionnaire was adapted from the "Severity of Exposure Index", which was used for the same purpose [32]. The WEQ has 38 items, which inquire about eight di fferent subdomains of war-related experiences (war-related experiences, physical conditions, injury or disease, witnessing casualties amongs<sup>t</sup> comrades, witnessing casualties amongs<sup>t</sup> the enemy, witnessing casualties amongs<sup>t</sup> civilians, actions on the enemy, and action against civilians). Each sub-domain is the result of the sum of positively-answered questions (e.g., have you been tortured?) which are related to that subject. The total sum provides a total WEQ score (ranging from 0 to 38), which represents war exposure, and which was used as a surrogate for war severity. In our sample, the Cronbach's alpha (reliability) of the total WEQ score was 0.81.
