Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2017 May 1.
Published in final edited form as: J Abnorm Psychol. 2016 Mar 31;125(4):471–481. doi: 10.1037/abn0000152

Negative Emotionality and its Facets Moderate the Effects of Exposure to Hurricane Sandy on Children's Post-Disaster Depression and Anxiety Symptoms

Daniel C Kopala-Sibley 1, Allison P Danzig 1, Roman Kotov 2, Evelyn J Bromet 2, Gabrielle A Carlson 2, Thomas M Olino 3, Vickie Bhatia 1, Sarah R Black 1, Daniel N Klein 1,2
PMCID: PMC4850107  NIHMSID: NIHMS773167  PMID: 27030993

Abstract

According to diathesis-stress models, temperament traits, such as negative emotionality (NE), may moderate the effects of stressors on the development of symptoms of psychopathology, although very little research has tested such models in children. Moreover, there are few data on whether specific facets of NE (sadness, fear, or anger) may specifically moderate the effects of stress on depression versus anxiety. Finally, there is a paucity of research examining whether childhood temperament moderates the effect of disaster exposure on depressive or anxiety symptoms. Hurricane Sandy, which affected many thousands of people in New York State and surrounding regions in October 2012, offers a unique opportunity to address these gaps. Seven years prior to Hurricane Sandy, 332 three-year-old children completed lab-based measures of NE and its facets. Six years later, when they were nine years old, one parent rated their child’s depressive and anxiety symptoms. Approximately eight weeks post-Sandy (an average of one year after the age nine assessment), a parent again rated their child’s depressive and anxiety symptoms, as well as a measure of exposure to stress from Hurricane Sandy. Adjusting for symptom levels at age 9, higher levels of stress from Hurricane Sandy predicted elevated levels of depressive symptoms only in participants with high levels of temperamental sadness, and predicted elevated levels of anxiety symptoms only in participants high in temperamental fearfulness. These findings support the role of early childhood temperament as a diathesis for psychopathology, and highlight the importance of considering facets of temperament when examining their relationship to psychopathology.

Keywords: Temperament, Anxiety, Depression, Natural Disaster, Negative Emotionality

According to diathesis-stress models, temperament or personality traits may moderate the effects of stressors on the development of symptoms of psychopathology (Blatt & Zuroff, 1992; Ingram & Price, 2010; Klein, Kotov, & Bufferd, 2011).4 Natural disasters offer a unique opportunity to test such models, as the occurrence of a natural disaster is unrelated to the individual’s temperament. Hurricane Sandy affected New York State and surrounding regions in October, 2012, and was the most destructive hurricane in the region in many years and the second costliest to hit the United States after Hurricane Katrina, destroying over 100,000 homes (Neria & Shultz, 2012). A substantial number of youth experience significant psychological symptoms after natural disasters (Bonanno, Brewin, Kaniasty, & La Greca, 2010; Wang, Chan, & Ho, 2013). However, it is unknown why some children and not others develop symptoms following disasters, and whether these effects are moderated by childhood temperament. In the present study, we prospectively examined whether early childhood negative emotionality (NE) and its facets (sadness, fear, and anger) moderated the effects of Hurricane Sandy on increases in children’s symptoms of depression and anxiety.

Impact of life stress and natural disasters on depression and anxiety

In both adults and children, there is considerable evidence indicating that stress plays a causal role in the onset of depression and anxiety symptoms (see Grant et al., 2014; Hammen, in press; Monroe et al., 2014 for recent reviews). Much of this research has focused on relatively common life events or, in some studies, daily hassles (e.g., Brown & Rosellini, 2011; Hutchinson & Williams, 2007; Kendler et al., 2004). In parallel, a substantial literature examining both children and adults has confirmed that exposure to natural disasters, including hurricanes, confers risk for the subsequent development of psychological symptoms (Acierno et al., 2006, 2007; Bonanno et al., 2010; Boscarino et al., 2013; Galea et al., 2007; McLaughlin et al., 2009; McLaughlin et al., 2010; North & Pfefferbaum, 2013; Wang et al., 2013). However, even following major stressors and disasters, most children and adults do not experience a significant increase in symptoms (Hammen, 2005; Pfefferbaum, Noffsinger, Wind, & Allen, 2014; Vogel et al., 1993).

The considerable heterogeneity in individual responses to stress suggests that other factors may moderate the effect of stress on depression and anxiety. These findings have prompted research into how individual differences confer vulnerability to psychopathology in the presence of stressful life events (e.g., Blatt & Zuroff, 1992; Compas, Connor-Smith, & Jaser, 2004; Ingram et al., 1998).

Personality/temperament diatheses and vulnerability to depression and anxiety

Based on Clark and Watson’s influential tripartite model (Clark & Watson, 1991; Clark, Watson, & Mineka, 1994), high levels of NE have long been hypothesized to be a key vulnerability to depression and anxiety in children and adults (e.g., Bienvenu et al., 2004; Brown, 2007; Gershuny & Sher, 1998; Klein, Dyson, Kujawa, & Kotov, 2012). NE, which is closely related to the construct of neuroticism, is characterized by a propensity to experience dysphoric mood states such as fear, anger, and sadness, as well as being particularly susceptible to the effects of stress on mood (Watson, Kotov, & Gamez, 2006). Numerous studies have reported that NE significantly predicts increased depressive and anxiety symptoms and disorder onsets (Compas et al., 2004; Klein et al., 2011; Kotov et al., 2010; Ormel et al., 2013). However, only a handful of studies have examined whether NE buffers the effects of stress on depression in adults, and to our knowledge, none have examined anxiety. Most of these studies have reported that the effect of stress on subsequent depressive symptoms is greater in individuals with higher levels of NE or neuroticism (Brown & Rosellini, 2011; Hutchinson & Williams, 2007; Kendler et al., 2004; Kopala-Sibley et al., 2015; Ormel et al., 2001; Vinkers et al., 2014; although see Spinhoven et al., 2011 for an exception).

In children and adolescents, there is evidence that NE predicts symptoms of depression following stressors such as parental divorce (Lengua, Sandler, West, Wolchik, & Curran, 1999; Lengua, Wolchik, Sandler, & West, 2000; although Lengua, Long, Smith, and Meltzoff [2005] did not find an effect of child fearfulness on post-traumatic stress symptoms following the September 11 terrorist attacks). However, with one exception, we are unaware of studies that have examined whether NE moderates the effects of stress on depressive or anxiety symptoms in youth. In a study of 6th to 10th-graders, Wetter and Hankin (2009) reported that NE did not moderate the effects of dependent or independent stressors on subsequent anhedonic depressive symptoms.

Facets of NE as unique diatheses for depression versus anxiety

NE is comprised of several correlated facets, including propensities to experience sadness, anger, and fear, with some investigators also including guilt and self-consciousness (Costa & McCrae, 1995; Tellegen et al., 1999; Watson et al., 2011). Because these facets share substantial variance, this often results in the higher order trait of NE showing largely non-specific associations with psychopathology (Griffith et al., 2010; Watson et al., 2011). However, it is conceivable that these facets have more specific effects, especially in interaction with stressors, on the development of depressive or anxiety symptoms (Klein et al., 2011). For example, temperamental sadness may be a stronger risk factor for depression, whereas temperamental fearfulness may be a more potent risk factor for anxiety (Watson et al., 2011).

Other investigators have demonstrated the incremental validity of the facet level of personality organization in psychopathology research. For example, Paunonen (2003) and Reynolds and Clark (2001) found that facets of personality typically show more robust associations with psychological symptoms than do higher order trait scores. Despite this, there is very limited research on temperament/personality and psychopathology at the facet level (Klein et al., 2011; Watson et al., 2011). As far as we are aware, only one study to date found that, in adults, sadness and guilt were more strongly associated with major depression compared to anger. In contrast, fear was more strongly associated with generalized anxiety than were sadness or anger, although it was also predictive of depression (Naragon-Gainey & Watson, 2014). Importantly, we are not aware of any research that has taken a diathesis-stress perspective and examined whether facets of NE moderate the effects of stress on depression versus anxiety.

Issues in the assessment of temperament in diathesis-stress research

The extant literature on the role of temperament traits as diatheses for psychopathology is commonly hampered by several limitations (Klein et al., 2011). First, the majority of studies rely on the same informant (self or parent) and method (questionnaires) to report on both temperament/personality and symptoms, raising the issue of informant and shared method effects. Second, most self- and informant-report measures of NE contain items which are very similar to depressive and anxiety symptoms, raising concerns about effects being due to shared content (Ormel et al., 2004). Third, consistent with stress-generation models (Hammen, 2006), traits such as NE predict subsequent dependent life events and chronic stressors (Stroud, Sosoo, & Wilson, 2015; Wetter & Hankin, 2009). Thus, many personality/temperament diathesis X stress studies may confound the diathesis and the stressor (Monroe & Simons, 1991). Finally, existing studies examining personality diatheses have generally assessed temperament/personality within a year of the stressor (e.g., Brown & Rosellini, 2011; Hutchinson & Williams, 2007; Kendler et al., 2004). As temperament/personality diatheses are generally considered to be stable, long-term vulnerability factors, designs using longer intervals between assessing traits and stress provide more conservative and compelling tests. We address these issues in the current study by using observational measures of early childhood temperament conducted approximately 7 years prior to Hurricane Sandy, while using mothers’ reports of symptoms pre- and post-Sandy. Examining a natural disaster as a form of stress reduces concerns about child temperament/personality influencing the occurrence of the stressors, thereby permitting stronger conclusions about the role of personality/temperament in moderating the effects of stress on symptoms.

Study overview and hypotheses

This paper takes advantage of a pre-existing longitudinal study (Olino, Klein, Dyson, Rose, & Durbin, 2010) to examine whether NE or its three core facets (sadness, anger, and fear), moderate the effects of Hurricane Sandy-related stress on parents’ reports of their children’s depressive and anxiety symptoms 8 weeks after the disaster, when children were approximately 10 years old. We hypothesized that hurricane-related stress would have a particularly strong impact on both depressive and anxiety symptoms in children who showed elevated NE. In addition, we hypothesized that temperamental sadness would specifically moderate the effects of stress on depression, while temperamental fear would specifically moderate the effects of stress on anxiety. We did not have specific hypotheses regarding the effects of temperamental anger on depressive and anxiety symptoms. All analyses adjusted for children’s depressive or anxiety symptoms at age 9 (prior to the hurricane) in order to examine change in symptoms over time, as well as to account for any effects of pre-Sandy symptoms on post-Sandy symptoms.

Method

Participants

Our sample consisted of 332 three-year olds (M age = 43.5 months, SD = 2.8) and their mothers from a larger longitudinal study of 559 children (see Olino et al., 2010 for details). Briefly, in 2004-2006, participants were recruited through commercial mailing lists and screened by phone to select children with no significant medical problems or developmental disabilities and who had at least one English-speaking biological parent who could participate. As part of this larger, ongoing study, mothers and children were assessed at three-year intervals, i.e., at ages 3 (2004-2006), 6 (2007-2009), and 9 (2010-2012). An additional 50 minority families were added at age 6 to increase diversity, bringing the total sample to 609.

Of the 609 participating families, 446 had completed age 9 assessments prior to the hurricane; these families were invited to participate in the post-Sandy assessment. Of these 446, 362 (81.2%) agreed to participate. Of these 362 families, 30 were excluded: 15 who were not on Long Island when Hurricane Sandy occurred; and 15 who entered the study at age 6, and therefore did not have laboratory measures of temperament at age 3. Thus, the effective sample size for the current study is 332 children and their mothers. Children who participated in the current study did not differ from children who completed the age 9 assessment but did not participate in the post-Sandy assessment on age 3 NE or on age 9 CBCL depressive or anxiety scores (all ps > .16). Finally, our effective sample also did not differ from the remaining children in the larger study on age 3 NE or age 9 depressive or anxiety scores.

Our effective sample (N = 332; 154 females; M age post-Sandy = 10.25 years; SD = .76) was primarily middle-class as measured by Hollingshead’s four-factor index of social status, which is based on a combination of parental education and occupational prestige (M = 45.33, SD = 10.99; Hollingshead, 1975). Most mothers were married at each assessment (>85%) while 55.15% had completed a 4 year college degree. Of the children, 92.5% were Caucasian and non-Hispanic. Following Hurricane Sandy, 96.0% of families in the current study lived in FEMA-declared disaster zones.

Procedure

During the age 3 assessment, all children completed the Laboratory Temperament Assessment Battery (Lab-TAB; Goldsmith, Reilly, Lemery, Longley, & Prescott, 1995) as a measure of temperamental NE. During the age 9 assessment, mothers completed the Child Behavior Checklist (CBCL; Achenbach & Rescorla, 2001) as a measure of their children’s depressive and anxiety symptoms.

Six weeks after Hurricane Sandy, mothers were contacted and asked to complete the CBCL, as well as a measure of hurricane related stress. Mothers completed their surveys an average of 8.4 weeks (SD = 1.49) after the hurricane, and 63.98 weeks after the age 9 assessment (SD = 31.26). This study was approved by the Stony Brook University Institutional Review Board.

Measures

Hurricane Sandy

Six weeks after Hurricane Sandy, mothers were asked to complete a 13 item web-based questionnaire concerning the effects of the storm on families and children (Table 1). The items were drawn from previous questionnaires developed for Hurricane Ike (Norris, Sherrieb, & Galea, 2010) and Hurricane Katrina (Galea et al., 2007). Items 1-8 were rated on a 5-point scale (1=not at all affected; 5=extremely affected); items 9-10 were rated on duration (1=0 days; 5=2 weeks or more); items 11-13 were rated as present/absent. To create an overall sum of exposure severity, non-dichotomous items were rescored such that 1=present and 0=absent. We selected cutoffs for each item using a combination of statistical and clinical considerations that distinguished a subgroup of participants with a high and clinically significant level of stress on that item. This involved examining the distribution of responses on each item and considering the nature of the stressor and the response options. For most items, we selected a cutoff of 4 (affected quite a bit; where 3 = moderately and 5 = extremely), thinking that for most of these experiences, responses of “quite a bit” and “extremely” would indicate a meaningful degree of stress, whereas the significance of a response of “moderately” was less clear. However, there were two exceptions. “Quite a bit” was the modal response for difficulty finding gasoline. Given the high frequency of this response, we chose to err on the conservative side and required a rating of 5 (“extremely”) for this item. In contrast, responses to the item on financial hardships were extremely skewed, with very few respondents choosing responses of 4 and 5. This may, in part, have been due to the phrasing of the item (“hardships”), which suggested a high threshold for endorsement. Hence, we selected a cutoff of 3 (“moderately”) for this item. Total scores therefore ranged from 0-13. This scale showed adequate internal consistency (α = .73).

Table 1.

Items comprising the Hurricane Sandy exposures composite

Item N (%) experienced exposure
1. Damage to home or possessions 40 (12.6%)
2. Own or family’s safety threatened 69 (21.7%)
3. Financial hardships 43 (13.5%)
4. Children fear for their own or their family’s safety 79 (24.8%)
5. Life disrupted by Hurricane Sandy 82 (25.8%)
6. Difficulty finding gasoline 57 (17.9%)
7. Difficulty getting food, water, or warmth 69 (21.7%)
8. Children quarreling or complaining more than usual 54 (17.0%)
9. At least a week without power 124 (39.0%)
10. Children’s school closed for more than two weeks 170 (53.5%)
11. Self, friends, family, pets, or coworkers robbed, injured, or got lost 11 (3.5%)
12. Apply to FEMA, government aid, or Red Cross or other aid agency 14 (4.4%)
13. Evacuated home 16 (5.0%)
Open in a new tab

N = 332. Note: Item scores dichotomized as present or not present. See methods for a full description.

Depressive and anxiety symptoms pre- and post-Sandy

At the age 9 and post-Sandy evaluations, depressive and anxiety symptoms were assessed via mothers’ reports on the CBCL (Achenbach & Rescorla, 2001). The CBCL is a 113-item parent-report checklist assessing emotional and behavioral problems in children. In the current study, we used the Diagnostic and Statistical Manual (DSM–IV; American Psychiatric Association (APA), 1994) Anxiety Problems (6 items) and Affective Problems (13 items) subscales derived from the CBCL. These scales show good internal consistency and convergence with interview-based symptom measures (Ebesutani et al., 2010; Nakamura et al., 2009). At the age 9 assessment, items were rated for the past 6 months; coefficient α was .68 for affective problems, and .73 for anxiety problems. In the post-Sandy assessment, mothers were asked to rate symptoms in the period since Sandy; coefficient α was .66 for affective problems, and .73 for anxiety problems.

Child NE

At age 3, children participated in 12 age-appropriate episodes from the Lab-TAB (Goldsmith et al., 1995) that were designed to elicit a range of temperament-relevant behaviours and emotions. Child NE was coded from videotapes; for full details regarding the episodes and coding, see Olino et al. (2010). During each episode, children’s bodily, vocal, and facial expressions of sadness, anger and fear were rated on a three-point scale. Ratings of sadness, anger, and fear were averaged to create a total NE scale. Coefficient alpha for overall NE was .82, and the intraclass correlation for interrater reliability (n = 35) was .74. Coefficient alphas/ICCs for fear, sadness, and anger, respectively, were .63/.64, .81/.79, and .68/.73. Male and female children were not significantly different in NE, t (158) = −0.67, p = .50.

Data Analyses

Primary data analyses consisted of simultaneous multiple linear regression. Parameters were tested simultaneously rather than hierarchically as we were primarily interested in the interaction terms in our model, rather than any main effects.5 All analyses initially adjusted for families’ socioeconomic status anCd age, as well as child’s age, sex, and ethnicity, and duration of time between the age 9 and post-Sandy assessments. However, these variables were unrelated to symptoms post-Sandy (all p’s > .30), and all effects remained significant at levels reported here after adjusting for these variables, so results below are reported without these covariates in our models. Results when controlling for demographic variables are available upon request. Each model controlled for age 9 levels of the symptom scale being predicted. We tested 4 models. In the first two models, one each predicting depressive and anxiety symptoms post-Sandy, post-Sandy symptoms were regressed onto the corresponding pre-Sandy symptom score, followed by the main effects of Hurricane Sandy-related stress and temperamental NE assessed at age 3, and finally by the interaction of stress and NE. A similar approach was used in our second pair of models, except that the three NE facets were entered into the model simultaneously. Thus, depressive or anxiety symptoms post-Sandy were regressed on their pre-Sandy levels, followed by Hurricane Sandy-related stress and the main effects of temperamental sadness, anger, and fear. In the final step, the three two-way interactions of stress from Sandy with each of the three facets of NE were entered into the model. All predictor variables were centered (Aiken & West, 1991). Interactions were interpreted by comparing simple slopes at high and low levels (±1 SD) of the moderator variable. 95% confidence intervals (CIs) are presented for all significant results.

Finally, one concern is that two items from our measure of stress from Hurricane Sandy, specifically “children fear for their own or their family’s safety,” and “children quarrelling or complaining more” may tap children’s subjective distress about Hurricane Sandy. This subjective distress may be confounded with their own temperamental NE, which may increase perceptions of or reactions to stressful situations. As such, we conducted ancillary analyses in which we re-analyzed our models after excluding these two items from our stress measure.

Results

Descriptive statistics and bivariate correlations

Rates of Hurricane Sandy experiences are presented in Table 1. The most common were prolonged school closing, losing electric power, life disruption, children fearing for their or family’s safety, difficulty getting food, water, or warmth, and own or family’s safety threatened. The average number of stressors (out of 13) was 2.17 (SD = 2.19). Scores for both depressive (M = .60, SD = 1.31) and anxiety symptoms (M = .72, SD = 1.33) ranged from 0-9.

Correlations and descriptive statistics are presented in Table 2. Age 9 depressive and anxiety symptoms were both associated with greater depressive and anxiety symptoms post-Sandy. Age 3 NE and its facets were largely unrelated to the psychopathology variables, either pre- or post-Sandy. Higher levels of exposure to stress from Hurricane Sandy were related to greater depressive and anxiety symptoms post-Sandy.

Table 2.

Descriptive statistics and bivariate correlations between all variables.

Variable 1. 2. 3. 4. 5. 6. 7. 8. 9.
1. Age 3 NE -- .74** .64** .69** −.07 .06 .08 .05 .05
2. Age 3 Sadness -- .19** .45** −.02 .09 .08 .08 .03
3. Age 3 Fear -- .06 −.03 .01 .05 .03 .04
4. Age 3 Anger -- −.12* .04 .07 .01 .06
5. Sandy stress -- −.02 .01 .13* .17**
6. Age 9 depressive symptoms -- .54** .41** .34**
7. Age 9 anxiety symptoms -- .34** .60**
8. Post-Sandy depressive -- .58**
symptoms
9. Post-Sandy anxiety symptoms --
Mean .56 .55 .65 .57 2.17 1.14 1.30 .59 .72
SD .27 .31 .35 .34 2.19 1.82 1.80 1.31 1.32
Open in a new tab

Note:

**

p < .01,

*

p < .05.

N = 332.

NE = Negative emotionality.

Predicting post-Sandy anxiety and depressive symptoms – Overall NE as a moderator

Results from all regression models are presented in Table 3. Initially, post-Sandy anxiety symptoms were regressed on age 9 anxiety symptoms, then on NE and total stress scores, and then on their interaction. Higher levels of anxiety symptoms at age 9 predicted higher levels of anxiety symptoms post-Sandy. Higher stress from Sandy also predicted elevated anxiety symptoms post-Sandy, although the effect of NE was not significant. In addition, there was a significant interaction between NE and Sandy-related stress (Figure 1). The effect of stress on anxiety symptoms was significant at high levels of NE (β = .38, t = 4.41, p < .001, 95% CI [.21, .52]), but not at low levels (β = .08, t = 1.09, p = .27, 95% CI [−.07, .22]). The full model explained 40.17% of variance in post-Sandy anxiety symptoms.

Table 3.

Results of regression models predicting post-Sandy depressive or anxiety symptoms.

β SE T R2 95% CI
Model 1: Anxiety symptoms 40.17%
Age 9 anxiety symptoms .83** .06 13.92 .68, .91
Age 3 NE .02 .06 .39 −.09, .13
Stress .23** .06 3.97 .10, .34
NE*Stress .15** .06 2.62 .04, .25
Model 2: Depressive symptoms 19.3%
Age 9 depressive symptoms .53** .06 8.20 .41, .66
Age 3 NE .05 .07 .76 −.08, .18
Stress .20** .06 3.00 .06, .32
NE*Stress .11+ .07 1.73 .005, .21
Model 3: Anxiety Symptoms 41.1%
Age 9 anxiety symptoms .84** .06 13.96 .68, .92
Age 3 Sadness −.10 .06 -1.43 −.22, .03
Age 3 Fear .02 .06 .38 −.09, .14
Age 3 Anger .12* .06 1.96 .001, .25
Stress .24** .06 3.96 .11, .35
Sadness*Stress .10 .07 1.47 −.03, .22
Fear*Stress .11* .05 2.05 .04, .22
Anger*Stress −.01 .07 −.06 −.14, .13
Model 4: Depressive symptoms 20.0%
Age 9 depressive symptoms .53** .07 8.15 .41, .67
Age 3 Sadness .04 .08 .52 −.10, .17
Age 3 Fear .03 .07 .42 −.10, .16
Age 3 Anger .02 .07 .21 −.12, .15
Stress .20** .07 2.84 .06, .32
Sadness*Stress .16* .08 2.06 .01, .29
Fear*Stress .03 .06 .53 −.08, .15
Anger*Stress −.04 .08 −.54 −.19, .11
Open in a new tab
**

p < .01,

*

p < .05,

+

p < .10.

N = 332.

Note: NE = Negative emotionality. Sadness, anger, and fear refer to age three temperamental variables.

Stress = Exposure to hurricane Sandy-related stress.

Figure 1.

Figure 1

Open in a new tab

Relationship between stress from Hurricane Sandy and post-Sandy anxiety symptoms at high and low levels of age 3 negative emotionality (NE). Only the slope at high NE is significant.

The next model was similar, except that it predicted post-Sandy depressive symptoms, adjusting for age 9 depressive symptoms. Higher levels of depressive symptoms at age 9 predicted higher levels of depressive symptoms post-Sandy. Stressors also predicted elevated depressive symptoms post-Sandy, although the effect of NE was not significant. In addition, there was a trend (p = .085) towards an interaction between NE and Sandy-related stress (Figure 2). Although the interaction did not meet conventional levels of statistical significance, we examined the slopes knowing that any interpretation must be very tentative. The effect of stress on depressive symptoms was significant at high levels of NE (β = .32, t = 3.17, p <.01, 95% CI [.12, .48]), but not at low levels (β = .09, t = 1.02, p = .31, 95% CI [−.08, .26]). The full model explained 19.30% of variance in post-Sandy depressive symptoms.

Figure 2.

Figure 2

Open in a new tab

Relationship between stress from Hurricane Sandy and post-Sandy depressive symptoms at high and low levels of age 3 negative emotionality (NE). Only the slope at high NE trended towards significance.

Examining facet-level moderators

The above analyses were repeated using the three NE facets entered simultaneously in each model. In our model predicting anxiety symptoms, higher levels of anxiety symptoms at age 9 predicted higher levels of anxiety symptoms post-Sandy. Additionally, higher anger predicted elevated post-Sandy anxiety symptoms, although there was no main effect of sadness or fear. Higher levels of stress from Sandy also predicted elevated anxiety symptoms. Finally, there was a significant interaction between fear and Sandy-related stress (Figure 3), but not between stress and either sadness or anger. The effect of stress on anxiety symptoms was significant at high levels of fear (β = .35, t = 4.27, p < .0001, 95% CI [.18, .50]), but not at low levels (β = .12, t = 1.50, p = .13, 95% CI [−.04, .28]). The full model explained 41.1% of variance in anxiety symptoms.

Figure 3.

Figure 3

Open in a new tab

Relationship between stress from Hurricane Sandy and post-Sandy anxiety symptoms at high and low levels of age 3 temperamental fear. Only the slope at high fear is significant.

In our model predicting depressive symptoms, higher levels of depressive symptoms at age 9 predicted higher levels of depressive symptoms post-Sandy. There were no main effects of sadness, anger, or fear, although more exposure to stress from Sandy predicted elevated depressive symptoms. In addition, there was a significant interaction between sadness and Sandy-related stress (Figure 4), but no significant interactions between stress and either anger or fear. The effect of stress on depressive symptoms was significant at high (β = .35, t = 3.42, p < .001, 95% CI [.15, .54]), but not low levels of sadness (β = .03, t = .31, p = .75, 95% CI [−.16, .23]). The full model explained 20.0% of variance in post-Sandy depressive symptoms.

Figure 4.

Figure 4

Open in a new tab

Relationship between stress from Hurricane Sandy and post-Sandy depressive symptoms at high and low levels of age 3 temperamental sadness. Only the slope at high sadness is significant.

Ancillary analyses excluding subjective stress items

After excluding two items (“children fear for their own or their family’s safety,” and “children quarrelling or complaining more”) due to their subjective nature, results were virtually identical to those previously reported. All models were the same as those specified above. All main effects in all models were significant at levels reported previously. There was a significant interaction between overall NE and stress predicting anxiety symptoms post-Sandy (β = .17, t = 2.98, p = .003, 95% CI [.05, .29]), such that higher levels of stress predicted increases in anxiety symptoms at high levels of NE (β = .38, t = 4.43, p < .001, 95% CI [.20, .56]) but not at low levels of NE (β = .04, t = .57, p = .57, 95% CI [−.12, .20]).

Whereas the interaction between overall NE and stress predicting depression was previously significant at a trend level, after dropping two items, results showed a significant interaction between overall NE and stress on post-Sandy depression (β = .13, t = 1.98, p = .048, 95% CI [.002, .26]), such that higher levels of stress from Sandy predicted increases in depressive symptoms at high levels of NE (β = .30, t = 3.02, p = .003, 95% CI [.10, .49]), but not at low levels of NE (β = .04, t = .46, p = .64, 95% CI [−.12, .20]).

Examining effects of facet-level interactions, in our model predicting anxiety, there was a significant interaction between stress and fearfulness (β = .13, t = 2.33, p = .02, 95% CI [.21, .41]), but not with sadness (p = .24) or anger (p = .65). Higher levels of stress were related to higher levels of anxiety symptoms post-Sandy at high levels of fearfulness (β = .35, t = 4.29, p < .001, 95% CI [.19, .51]), but not at low levels (β = .09, t = 1.07, p = .28, 95% CI [−.07, .25]). In our model predicting depression, there was a significant interaction between stress and sadness (β = .16, t = 1.96, p = .05, 95% CI = [.005, .32]), but not with fear (p = .47) or anger (p = .71). Higher levels of stress were related to higher levels of depressive symptoms post-Sandy at high levels of sadness (β = .32, t = 3.02, p = .003, 95% CI [.11, .53]), but not at low levels of sadness (β = .01, t = .11, p = .92, 95% CI [−.19, .21]).

Discussion

This study provided a novel test of diathesis-stress models of depression and anxiety by examining whether temperamental NE in early childhood moderates the effect of exposure to a natural disaster on increases in depressive and anxiety symptoms in late childhood. Moreover, it is the first paper to our knowledge to examine whether specific facets of NE interact with stress to uniquely predict anxiety versus depressive symptoms. Results showed, initially, that stress from Sandy only predicted increases in anxiety symptoms in children who were higher in overall NE, although there was a trend towards a similar interaction between overall NE and stress predicting depressive symptoms. Subsequent facet-level analyses demonstrated that temperamental sadness specifically moderated the effects of stress from Sandy on depressive symptoms, while temperamental fear specifically moderated the effects of stress from Sandy on anxiety symptoms, with no interactions found between stress and temperamental anger. Remarkably, temperamental risk for anxiety and depression symptoms was detectable even in young children, indicting the enduring importance of NE for the development of internalizing psychopathology and the potential value of temperament assessment for both etiology and prevention research. Finally, these results were robust after removing two items from the stress measure which may tap stress perception rather than stress exposure. Taken together, these results provide evidence for specific facets of NE as diatheses for depression versus anxiety in the face of stress.

Facets of NE, disaster stress, and anxiety and depressive symptoms

Our findings are broadly consistent with the literature in adults supporting a moderating role for NE/Neuroticism in the relationship between life stress and depression (Hutchison & Williams, 2007; Kendler et al., 2004; Brown & Rosellini, 2011; Ormel et al., 2001; Vinkers et al., 2014), and with recent theory and evidence that specific facets of NE show specific relationships with depression versus anxiety (Naragon-Gainey & Watson, 2014; Watson et al., 2011). However, our results extend this nascent literature by showing the importance of facet-level analyses in temperament/personality diathesis-stress models of depressive and anxiety symptoms. Moreover, it appears that these diatheses can be observed as early as age 3; while it is often assumed that temperamental/personality diatheses are evident early in life, the bulk of literature on this topic is based on samples of older youth and adults (Brown & Rosellini, 2011; Hutchison & Williams, 2007; Kendler et al., 2004; Ormel et al., 2001; Vinkers et al., 2014). Additionally, the current results suggest that facets of NE not only moderate relatively common environmental adversities, such as life events, but also stress ensuing from a natural disaster, which is independent of participants and informants.

The present findings are also notable in that they overcome concerns about shared method variance or informant effects inflating associations between temperament and psychopathology, and cannot be due to item overlap in measures of temperament and symptoms. Our significant interaction between lab-based observations of temperament and stress approximately 7 years later are also unlikely to be due to the diathesis (i.e., temperament) and the stressor being confounded, a concern relevant to many prior studies on this topic.

Taken together, these results further our understanding of the etiology of depression versus anxiety. Specifically, it is well-known that depressive and anxiety symptoms are frequently comborbid, and indeed showed correlations of .54 and .58 at age 9 and post-Sandy, respectively, in the current study. Depression and anxiety have both shared and unique variance, which may explain why NE as a unitary construct is related to both. However, the facets of NE appear to predict the variance that is unique to depression versus anxiety, at least in the context of elevated levels of stress. Our facet-level analyses included the effects of all three facets together, thereby adjusting for their shared variance, which allowed us to examine the effects of each facet’s unique variance separately on depression and anxiety. Thus, while elevated NE overall appears to confer risk to both depression and anxiety, temperamentally sad children appear to be specifically at risk for depression, while fearful children appear to be specifically at risk for anxiety following stress.

Our results provide support for predisposition/vulnerability models of temperament in childhood (e.g., Clark et al., 1994; Compas et al., 2004; Klein et al., 2011) in which traits place individuals at risk for psychopathology. However, these findings are less readily explained by other models of the relationship between personality and psychopathology, such as spectrum/continuum and shared etiological factors models, given the paucity of main effects of temperament on post-Sandy symptoms (Clark et al., 1994; Compas et al., 2004; Klein et al., 2011).

The lack of main effects of temperament variables is somewhat surprising, given that NE and its facets, especially sadness and fear, are typically associated with both depressive and anxiety symptoms (e.g., Kotov et al., 2010; Klein et al., 2011). We suggest that symptoms at any given point are determined by many factors which may suppress or drown out the signal or effect from early life vulnerability factors. However, when focusing on the situation in which this vulnerability plays a central role, for instance following a serious threat such as a natural disaster, the impact of early life vulnerability becomes apparent. That said, more research into the long term effects of temperament on symptoms, both on its own and in interaction with stressors, is needed.

Strengths and limitations

This study overcame the limitations of much prior research by measuring temperament observationally and symptoms via mothers’ reports, thereby minimizing issues of informant and method effects, and by examining a stressor that was independent of temperament, with temperament measured well in advance, spanning the interval from early to late childhood. However, several limitations should also be noted. Perhaps the most substantial concerns are that depressive and anxiety symptoms were assessed at the same time as Hurricane Sandy-related stress, and both were reported by mothers. As such, it is possible that the correlations between Sandy-related stress and children’s depressive and anxiety symptoms may be inflated by shared methods and informant biases. However, this would not explain the significant interactions found between temperament, assessed observationally, and stress in predicting symptoms. Further work is also needed to determine if our results would extend to reports by other informants or diagnostic interviews.

Second, as we sought to assess Sandy’s impact as rapidly as possible after the disaster, we used self-report measures of hurricane-related stress and children’s symptoms. In the life events area, self-report measures show only modest convergence with interview-based measures, and may be prone to errors and biases in the recall and reporting of life stressors (see Monroe, 2008 for a discussion of these issues). We were also unable to conduct diagnostic interviews, and do not know if our results would extend to diagnoses of major depression or specific anxiety disorders.

Third, our measure of stress primarily reflected events that happened to the children’s families; we cannot be sure if experiences that were specific to the children would show similar interactions with temperament. To the extent that this is a factor, it likely reduced the effects of exposure, making our estimates conservative. We also did not determine if there were other stressors in children’s lives that were unrelated to Sandy which may have compounded its effects. However, this should be a source of error, making it more difficult to detect the effects that we observed.

Fourth, we did not ask mothers to report on posttraumatic stress disorder (PTSD) symptoms in their children. Thus, we do not know if these results would extend to PTSD symptoms, or if adjusting for PTSD symptoms would affect our results.

Fifth, the duration of the moderation effects we found is unclear; longer-term follow-up is necessary to address this question. Sixth, the most commonly reported stressors from Hurricane Sandy could arguably be considered to be major hassles, rather than major stressors; we do not know whether similar effects would be found in an even more serious disaster in which, for instance, most participants lost their homes, or there was a high rate of injury or death.

Seventh, CBCL affective symptoms and Lab-TAB temperamental sadness and anger showed only modest reliability. However, to the extent that this is a concern, it should have decreased the likelihood of finding significant effects by introducing more measurement error into the data.

Finally, our sample was relatively homogeneous in that it was largely Caucasian and middle class. While this is consistent with the census data from the area (see Olino et al., 2010), results may not generalize to a broader, more diverse sample.

Conclusion

This study provided a novel test of diathesis-stress models by showing that NE, assessed in early childhood using laboratory observations, moderated the effects of Hurricane Sandy exposure on increases in internalizing symptoms 7 years later. Even more importantly, the NE facets of sadness and fear exhibited very specific moderating effects on depressive and anxiety symptoms, respectively. Results reinforce the vulnerability view of NE and show that temperamental risk can be detected as early as age 3. They also support the assessment of NE at the facet level, rather than examining it solely as a higher order construct. Such an understanding may be important for future efforts to match children’s risk profiles with more precisely tailored preventive interventions.

General Scientific Summary.

Results suggest that a temperamental tendency at three years old towards being fearful increases the effects of a natural disaster (Hurricane Sandy), which occurred approximately eight years later, on anxiety symptoms following the disaster. However, a temperamental tendency at age three towards sadness specifically increases the effect of exposure to a natural disaster on subsequent depressive symptoms. This suggests we may be able to identify children who are specifically at risk for anxiety versus depression in the context of stress, as much as eight years prior to the stressor.

Acknowledgements

Supported by NIMH grant RO1 MH45757 (Klein) and NIMH grant R01 MH093479 (Kotov).

Footnotes

4

For the purposes of this paper, we use the concepts of temperament and personality interchangeably, given the lack of empirical evidence for a clear distinction between the two (See Klein et al., 2011 for a review).

5

A possible concern is that main effects of temperament may be obscured by inclusion of interaction terms of temperament with stress. All analyses were therefore repeated without inclusion of interaction terms. Only anger showed a trend towards predicting higher levels of anxiety symptoms (p = .07).

References

  1. Achenbach TM, Rescorla L. ASEBA school-age forms & profiles. Aseba; Burlington: 2001. [Google Scholar]
  2. Acierno R, Ruggiero KJ, Kilpatrick DG, Resnick HS, Galea S. Risk and protective factors for psychopathology among older versus younger adults after the 2004 Florida hurricanes. The American Journal of Geriatric Psychiatry. 2006;14:1051–1059. doi: 10.1097/01.JGP.0000221327.97904.b0. [DOI] [PubMed] [Google Scholar]
  3. Acierno R, Ruggiero KJ, Galea S, Resnick HS, Koenen K, Roitzsch J, Kilpatrick DG. Psychological sequelae resulting from the 2004 Florida hurricanes: implications for postdisaster intervention. American Journal of Public Health. 2007;97(Suppl 1):S103–S108. doi: 10.2105/AJPH.2006.087007. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Aiken LS, West SG. Multiple regression: Testing and interpreting interactions. Sage; Newbury Park: 1991. [Google Scholar]
  5. Bienvenu OJ, Samuels JF, Costa PT, Reti IM, Eaton WW, Nestadt G. Anxiety and depressive disorders and the five-factor model of personality: A higher- and lower-order personality trait investigation in a community sample. Depression and Anxiety. 2004;20:92–97. doi: 10.1002/da.20026. [DOI] [PubMed] [Google Scholar]
  6. Blatt SJ, Zuroff DC. Interpersonal relatedness and self-definition: Two prototypes for depression. Clinical Psychology Review. 1992;12:527–562. [Google Scholar]
  7. Bonanno GA, Brewin CR, Kaniasty K, La Greca AM. Weighing the costs of disaster consequences, risks, and resilience in individuals, families, and communities. Psychological Science in the Public Interest. 2010;11:1–49. doi: 10.1177/1529100610387086. [DOI] [PubMed] [Google Scholar]
  8. Boscarino JA, Hoffman SN, Kirchner HL, Erlich PM, Adams RE, Figley CR, Solhkhah R. Mental health outcomes at the Jersey shore after hurricane Sandy. International Journal of Emergency Mental Health. 2013;15:147–158. [PubMed] [Google Scholar]
  9. Brown TA. Temporal course and structural relationships among dimensions of temperament and DSM–IV anxiety and mood disorder constructs. Journal of Abnormal Psychology. 2007;116:313–328. doi: 10.1037/0021-843X.116.2.313. [DOI] [PubMed] [Google Scholar]
  10. Brown TA, Rosellini AJ. The direct and interactive effects of neuroticism and life stress on the severity and longitudinal course of depressive symptoms. Journal of Abnormal Psychology. 2011;120:844–856. doi: 10.1037/a0023035. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Clark DA, Beck AT. Scientific Foundations of Cognitive Theory and Therapy of Depression. Wiley; New York, NY: 1999. [Google Scholar]
  12. Clark LA, Watson D. Tripartite model of anxiety and depression: psychometric evidence and taxonomic implications. Journal of Abnormal Psychology. 1991;100(3):316. doi: 10.1037//0021-843x.100.3.316. [DOI] [PubMed] [Google Scholar]
  13. Clark LA, Watson D, Mineka S. Temperament, personality, and the mood and anxiety disorders. Journal of Abnormal Psychology. 1994;103(1):103–116. [PubMed] [Google Scholar]
  14. Compas BE, Connor-Smith J, Jaser SS. Temperament, stress reactivity, and coping: Implications for depression in childhood and adolescence. Journal of Clinical Child and Adolescent Psychology. 2004;33:21–31. doi: 10.1207/S15374424JCCP3301_3. [DOI] [PubMed] [Google Scholar]
  15. Costa Jr PT, McCrae RR. Domains and facets: Hierarchical personality assessment using the Revised NEO Personality Inventory. Journal of Personality Assessment. 1995;64:21–50. doi: 10.1207/s15327752jpa6401_2. [DOI] [PubMed] [Google Scholar]
  16. Ebesutani C, Bernstein A, Nakamura BJ, Chorpita BF, Higa-McMillan CK, Weisz JR, Research Network on Youth Mental Health Concurrent validity of the Child Behavior Checklist DSM-oriented scales: Correspondence with DSM diagnoses and comparison to syndrome scales. Journal of Psychopathology and Behavioral Assessment1. 2010;32(3):373–384. doi: 10.1007/s10862-009-9174-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Galea S, Brewin CR, Gruber M, Jones RT, King DW, King LA, Kessler RC. Exposure to hurricane-related stressors and mental illness after Hurricane Katrina. Archives of General Psychiatry. 2007;64(12):1427–1434. doi: 10.1001/archpsyc.64.12.1427. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Gershuny BS, Sher KJ. The relation between personality and anxiety: Findings from a 3-year prospective study. Journal of Abnormal Psychology. 1998;107:252–262. doi: 10.1037//0021-843x.107.2.252. [DOI] [PubMed] [Google Scholar]
  19. Goldsmith HH, Reilly J, Lemery KS, Longley S, Prescott A. Preschool Laboratory Temperament Assessment Battery. Unpublished instrument; University of Wisconsin: 1995. [Google Scholar]
  20. Grant KE, McMahon SD, Carter JS, Carleton RA, Adam EK, Chen E. The influence of stressors on the development of psychopathology. In: Lewis M, Rudolph KD, editors. Handbook of Developmental Psychopathology. 3rd Springer; New York: 2014. pp. 205–224. [Google Scholar]
  21. Griffith JW, Zinbarg RE, Craske MG, Mineka S, Rose RD, Waters AM, Sutton JM. Neuroticism as a common dimension in the internalizing disorders. Psychological Medicine. 2010;40(7):1125–1136. doi: 10.1017/S0033291709991449. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Hammen C. Stress and depression. Annual Reviews of Clinical Psychology. 2005;1:293–319. doi: 10.1146/annurev.clinpsy.1.102803.143938. [DOI] [PubMed] [Google Scholar]
  23. Hammen C. Stress generation in depression: Reflections on origins, research, and future directions. Journal of Clinical Psychology. 2006;62:1065–1082. doi: 10.1002/jclp.20293. [DOI] [PubMed] [Google Scholar]
  24. Hammen C. Depression and Stressful Environments: Identifying Gaps in Conceptualization and Measurement. Anxiety, Stress, & Coping. :1–43. doi: 10.1080/10615806.2015.1134788. in press. [DOI] [PubMed] [Google Scholar]
  25. Hollingshead AB. Four factor index of social status. Unpublished manuscript; Department of Sociology, Yale University, New Haven, CT: 1975. [Google Scholar]
  26. Hutchinson JG, Williams PG. Neuroticism, daily hassles, and depressive symptoms: An examination of moderating and mediating effects. Personality and Individual Differences. 2007;42:1367–1378. [Google Scholar]
  27. Ingram RE, Miranda J, Segal ZV. Cognitive Vulnerability to Depression. Guilford Press; New York, NY: 1998. [Google Scholar]
  28. Ingram RE, Price JM. Understanding psychopathology: The role of vulnerability. In: Ingram RE, Price JM, editors. Vulnerability to Psychopathology: Risk Across the Lifespan. 2nd Guilford Press; New York, NY: 2010. pp. 3–17. [Google Scholar]
  29. Kendler KS, Kuhn J, Prescott CA. The interrelationship of neuroticism, sex, and stressful life events in the prediction of episodes of major depression. American Journal of Psychiatry. 2004;161:631–636. doi: 10.1176/appi.ajp.161.4.631. [DOI] [PubMed] [Google Scholar]
  30. Klein DN, Kotov R, Bufferd SJ. Personality and depression: Explanatory models and review of the evidence. Annual Review of Clinical Psychology. 2011;7:269–295. doi: 10.1146/annurev-clinpsy-032210-104540. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Klein DN, Dyson MW, Kujawa AJ, Kotov R. Temperament and internalizing disorders. In: Zentner M, Shiner R, editors. Handbook of Temperament. Guilford Press; New York: 2012. pp. 541–561. [Google Scholar]
  32. Kopala-Sibley DC, Kotov R, Bromet EJ, Carlson GA, Danzig AP, Black SR, Klein DN. Personality diatheses and Hurricane Sandy: Effects on post-disaster depression. Psychological Medicine. doi: 10.1017/S0033291715002378. in press. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Kotov R, Gamez W, Schmidt F, Watson D. Linking “big” personality traits to anxiety, depressive, and substance use disorders: A meta-analysis. Psychological Bulletin. 2010;136:768–821. doi: 10.1037/a0020327. [DOI] [PubMed] [Google Scholar]
  34. Lengua LJ. The contribution of emotionality and self-regulation to the understanding of children's response to multiple risk. Child Development. 2002;73:144–161. doi: 10.1111/1467-8624.00397. [DOI] [PubMed] [Google Scholar]
  35. Lengua LJ, Long AC, Smith KA, Meltzoff AN. Pre-attack Symptomatology and Temperament as Predictors of Children's Response to the September 11th Terrorist Attacks. Journal of Child Psychology and Psychiatry. 2005;46:631–645. doi: 10.1111/j.1469-7610.2004.00378.x. [DOI] [PubMed] [Google Scholar]
  36. Lengua LJ, Sandler IN, West SG, Wolchik SA, Curran P. Emotionality and self-regulation, threat appraisal, and coping in children of divorce. Development and Psychopathology. 1999;11:15–37. doi: 10.1017/s0954579499001935. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Lengua LJ, Wolchik SA, Sandler IN, West SG. The additive and interactive effects of parenting and temperament in predicting adjustment problems of children of divorce. Journal of Clinical Child Psychology. 2000;29:232–244. doi: 10.1207/S15374424jccp2902_9. [DOI] [PubMed] [Google Scholar]
  38. McLaughlin KA, Fairbank JA, Gruber MJ, Jones RT, Lakoma MD, Pfefferbaum B, Kessler RC. Serious emotional disturbance among youths exposed to Hurricane Katrina 2 years post-disaster. Journal of the American Academy of Child & Adolescent Psychiatry. 2009;48:1069–1078. doi: 10.1097/CHI.0b013e3181b76697. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. McLaughlin KA, Fairbank JA, Gruber MJ, Jones RT, Osofsky JD, Pfefferbaum B, Kessler RC. Trends in serious emotional disturbance among youths exposed to Hurricane Katrina. Journal of the American Academy of Child & Adolescent Psychiatry. 2010;49:990–1000. doi: 10.1016/j.jaac.2010.06.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Monroe SM, Slavich GM, Georgiades K. The social environment and depression: The roles of life stress. In: Gotlib IH, Hammen CL, editors. Handbook of Depression and Its Treatment. 3rd Guilford Press; New York: 2014. pp. 296–314. [Google Scholar]
  41. Monroe SM, Simons AD. Diathesis-stress theories in the context of life stress research: implications for the depressive disorders. Psychological Bulletin. 1991;110(3):406–425. doi: 10.1037/0033-2909.110.3.406. [DOI] [PubMed] [Google Scholar]
  42. Nakamura BJ, Ebesutani C, Bernstein A, Chorpita BF. A psychometric analysis of the child behavior checklist DSM-oriented scales. Journal of Psychopathology and Behavioral Assessment. 2009;31(3):178–189. doi: 10.1007/s10862-009-9174-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Naragon-Gainey K, Watson D. Consensually defined facets of personality as prospective predictors of change in depression symptoms. Assessment. 2014;21(4):387–403. doi: 10.1177/1073191114528030. [DOI] [PubMed] [Google Scholar]
  44. Neria Y, Shultz JM. Mental health effects of Hurricane Sandy: Characteristics, potential aftermath, and response. Journal of the American Medical Association. 2012;308:2571–2572. doi: 10.1001/jama.2012.110700. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Norris FH, Sherrieb K, Galea S. Prevalence and consequences of disaster related illness and injury from Hurricane Ike. Rehabilitation Psychology. 2010;55(3):221–230. doi: 10.1037/a0020195. [DOI] [PubMed] [Google Scholar]
  46. North CS, Pfefferbaum B. Mental health response to community disasters: a systematic review. Journal of the American Medical Association. 2013;310:507–518. doi: 10.1001/jama.2013.107799. [DOI] [PubMed] [Google Scholar]
  47. Olino TM, Klein DN, Dyson MW, Rose SA, Durbin CE. Temperamental emotionality in preschool-aged children and depressive disorders in parents: associations in a large community sample. Journal of Abnormal Psychology. 2010;119:468–478. doi: 10.1037/a0020112. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Ormel J, Jeronimus JF, Kotov R, Riese H, Bos EE, Hankin B, Rosmalen JGM. Neuroticism and common mental disorders: Meaning and utility of a complex relationship. Clinical Psychology Review. 2013;5:686–697. doi: 10.1016/j.cpr.2013.04.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Ormel J, Oldehinkel AJ, Brilman EI. The interplay and etiological continuity of neuroticism, difficulties, and life events in the etiology of major and subsyndromal, first, and recurrent depressive episodes in later life. American Journal of Psychiatry. 2001;158:885–891. doi: 10.1176/appi.ajp.158.6.885. [DOI] [PubMed] [Google Scholar]
  50. Ormel J, Rosmalen A, Farmer A. Neuroticism: A non-informative marker of vulnerability to psychopathology. Social Psychiatry and Psychiatric Epidemiology. 2004;39:906–912. doi: 10.1007/s00127-004-0873-y. [DOI] [PubMed] [Google Scholar]
  51. Paunonen SV. Big Five factors of personality and replicated predictions of behavior. Journal of Personality and Social Psychology. 2003;84(2):411–424. [PubMed] [Google Scholar]
  52. Pfefferbaum B, Noffsinger MA, Wind LH, Allen JR. Children's coping in the context of disasters and terrorism. Journal of Loss and Trauma. 2014;9(1):78–97. doi: 10.1080/15325024.2013.791797. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Reynolds SK, Clark LA. Predicting dimensions of personality disorder from domains and facets of the five factor model. Journal of Personality. 2001;69(2):199–222. doi: 10.1111/1467-6494.00142. [DOI] [PubMed] [Google Scholar]
  54. Spinhoven P, Elzinga BM, Hovens JGFM, Roelofs K, van Oppen P, Zitman FG, Penninx BWJH. Positive and negative life events and personality traits in predicting course of depression and anxiety. Acta Psychiatrica Scandinavica. 2011;124:462–473. doi: 10.1111/j.1600-0447.2011.01753.x. [DOI] [PubMed] [Google Scholar]
  55. Stroud CB, Sosoo EE, Wilson S. Normal personality traits, rumination and stress generation among early adolescent girls. Journal of Research in Personality. 2015;57:131–142. doi: 10.1016/j.jrp.2015.05.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Tellegen A, Watson D, Clark LA. On the dimensional and hierarchical structure of affect. Psychological Science. 1999;10(4):297–303. [Google Scholar]
  57. Vinkers CH, Joëls M, Milaneschi Y, Kahn RS, Penninx BW, Boks MP. Stress exposure across the lifespan cumulatively increases depression risk and moderated by neuroticism. Depression and Anxiety. 2014;31:737–745. doi: 10.1002/da.22262. [DOI] [PubMed] [Google Scholar]
  58. Vogel JM, Vernberg EM. Part 1: Children's psychological responses to disasters. Journal of Clinical Child Psychology. 1993;22(4):464–484. [Google Scholar]
  59. Wang C-W, Chan CLW, Ho RTH. Prevalence and trajectory of psychopathology among child and adolescent survivors of disasters: A systematic review of epidemiological studies across 1987-2011. Social Psychiatry and Psychiatric Epidemiology. 2013;48(11):1697–1720. doi: 10.1007/s00127-013-0731-x. [DOI] [PubMed] [Google Scholar]
  60. Watson D, Clark LA, Stasik SM. Emotions and the emotional disorders: A quantitative hierarchical perspective. International Journal of Clinical and Health Psychology. 2011;11(3):429–442. [Google Scholar]
  61. Wetter EK, Hankin BL. Mediational pathways through which positive and negative emotionality contribute to anhedonic symptoms of depression: A prospective study of adolescents. Journal of Abnormal Child Psychology. 2009;37:507–520. doi: 10.1007/s10802-009-9299-z. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES