Numerous studies have related neuroticism to negative emotional outcomes of adverse life events, including post-traumatic stress disorder (PTSD) symptoms. However, the nature of the relationship between neuroticism and post-trauma symptoms is unclear. The purpose of this study was to prospectively examine whether individuals high in neuroticism, relative to low neuroticism individuals, show a larger increase in symptoms after an adverse event. A sample of infantry troops completed questionnaires before deployment to Iraq (n = 214) and about five months (n = 170; 76%) thereafter. The findings showed that, after controlling for an indicator of trauma severity, (a) higher neuroticism individuals reported more PTSD symptoms, depression symptoms, and somatic problems after negative events, and (b) these relationships disappeared after controlling for pre-trauma symptoms. There were no significant differences between individuals high and low in neuroticism in the increase in symptoms from pre to post-trauma. This suggests that individuals high in neuroticism are not more reactive to adverse events.
Numerous studies have related neuroticism to negative emotional outcomes of adverse life events, such as distress (e.g., Ormel and Wohlfarth, 1991 and Van Os and Jones, 1999), grief (e.g., Janssen, Cuisinier, de Graauw, & Hoogduin, 1997), depression (e.g., Ormel, Oldehinkel, & Brilman, 2001), somatic problems (e.g., Chung, Berger, Jones, & Rudd, 2006), and (symptoms of) posttraumatic stress disorder (PTSD). When indicators of trauma severity or trauma type are held constant, individuals with elevated neuroticism scores are more likely to develop PTSD (e.g., Cox et al., 2004 and Kelly et al., 1998), and report more severe PTSD symptoms (e.g., Bramsen et al., 2000 and Holeva and Tarrier, 2001). These data have been taken to imply that high neuroticism individuals are more reactive to adverse events (e.g., Janssen et al., 1997 and Kelly et al., 1998).
However, this conclusion does not necessarily follow from the data. Neuroticism is associated with a wide range of complaints even before the occurrence of a negative life event (see Claridge and Davis, 2001, Ormel et al., 2004 and Watson and Pennebaker, 1989). Watson and Clark (1984) have argued that individuals who score high on neuroticism are more likely to experience distress at all times and across situations, even in the absence of overt stress. The increase in symptoms after an adverse event may not differ between individuals high and low in neuroticism. In other words, neuroticism may not have a moderator effect on the stressor-symptoms association. The impression of a larger symptom-increase in high neuroticism individuals may be created by the absence of mental health measures before a negative life event. Perhaps individuals high in neuroticism are more likely to be classified as a psychiatric patient after a negative life event (e.g., Breslau, Davis, Andreski, & Peterson, 1991), because they have more subclinical complaints before the event, and require less additional (event-related) symptoms to enter the clinical realm.
Such a pattern of a parallel increase in symptoms in high/low neuroticism individuals was recently suggested by Ormel et al. (2004), p. 907): “...prospective associations of neuroticism with mental health outcomes are basically futile, and largely tautological since scores on any characteristic with substantial within-subject stability will predict, by definition, characteristic and related variables at later points in time.” It is an empirical issue whether the net increase in symptoms after life events is higher among individuals high in neuroticism. It should be studied by measuring neuroticism and relevant emotional distress before a negative event and re-measuring distress afterwards. Such a design can determine whether neuroticism interacts with negative events, resulting in a greater post-event increase in distress for high neuroticism individuals. Engelhard, van den Hout, and Kindt (2003) reported a study with this design. In a sample of 1372 pregnant women, neuroticism and PTSD symptoms were assessed. In a subsample of 118 women who had a miscarriage or stillbirth, PTSD symptoms were re assessed. The data gave unequivocal support for the tautology account: relative to low neuroticism women, high neuroticism women had more symptoms after pregnancy loss, but they already had more symptoms before the loss, and the rise in symptoms did not differ between the groups. Further evidence comes from a prospective study by De Beurs and colleagues (2005) among a large nonclinical elderly sample. They found that neuroticism is associated with a decrease in emotional health over time, but showed that life events involving threat did not have a stronger impact on high compared to low neuroticism individuals.
The present study was carried out to test the generalizability of the findings of the Engelhard et al. (2003) study, using the same basic design: measuring distress and neuroticism before (an) adverse life event(s) and re-measuring distress after the event(s). Army soldiers (95% male) were tested before and after deployment to Iraq. Although the pre-deployment period may be stressful itself, ‘baseline’ symptoms were assessed during this period, because we were interested in increases in these symptoms after deployment. The nature of adverse events was obviously rather different from pregnancy loss, and included being shot at, being informed that a colleague got killed, being injured because of an accident, and seeing dead or injured individuals. The first goal of this study was to replicate the earlier findings of more PTSD symptoms after these events for higher neuroticism individuals, while controlling for trauma severity, which would disappear after controlling for PTSD symptoms before these events. A second goal was to extend the earlier findings by testing whether (non-PTSD) common complaints after stressful life events (i.e., depressive symptoms and somatic problems) would yield a comparable pattern. The data presented are part of a larger project, which is described in detail elsewhere (see Engelhard et al., 2007b and Engelhard and van den Hout, 2007).
