American Journal of Orthopsychiatry 2014, Vol. 84, No. 4, 341–352

© 2014 American Orthopsychiatric Association http://dx.doi.org/10.1037/ort0000006

Family and Peer Social Support and Their Links to Psychological Distress Among Hurricane-Exposed Minority Youth Donice M. Banks and Carl F. Weems

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

University of New Orleans

Experiencing a disaster such as a hurricane places youth at a heightened risk for psychological distress such as symptoms of posttraumatic stress disorder (PTSD), anxiety, and depression. Social support may contribute to resilience following disasters, but the interrelations of different types of support, level of exposure, and different symptoms among youth is not well understood. This study examined associations among family and peer social support, level of hurricane exposure, and their links to psychological distress using both a large single-time assessment sample (N ⫽ 1,098) as well as a longitudinal sample followed over a 6-month period (n ⫽ 192). Higher levels of hurricane exposure were related to lower levels of social support from family and peers. Higher levels of family and peer social support demonstrated both concurrent and longitudinal associations with lower levels of psychological distress, with associations varying by social support source and psychological distress outcome. Findings also suggested that the protective effects of high peer social support may be diminished by high hurricane exposure. The results of this study further our understanding of the role of social support in hurricane-exposed youths’ emotional functioning and point to the potential importance of efforts to bolster social support following disasters.

T

here are a wide range of physical and mental health consequences that may result from experiencing a disaster, and posttraumatic stress reactions such as posttraumatic stress disorder (PTSD) symptoms, depression, and anxiety are among the most prevalent mental health outcomes postdisaster (e.g., Garrison et al., 1995; Goenjian et al., 2001; Pina et al., 2008). Theoretically, natural disasters may have a detrimental effect on individuals’ mental health resulting from threat to the satisfaction of basic human needs and goals (Sandler, 2001; Weems & Overstreet, 2008). However, there is variation in the levels of psychological distress (including symptoms of PTSD, depression, and anxiety) that are experienced among trauma-exposed individuals, and some individuals experience very little distress (Weems & Overstreet, 2008). Many youth will recover from any psychological distress symptoms within a year of the event (La Greca & Silverman, 2009), but others may experience chronic and persistent symptoms. One potential factor that may contribute to resilience in the face of disaster is social support.

Social support may consist of social resources perceived by an individual to be available or resources that are actually received by an individual through informal helping relationships and/or formal support groups (Cohen, Gottlieb, & Underwood, 2000). Social support is thought to have an effect on posttraumatic stress symptoms after the initial experience of the disaster has occurred (Vernberg, La Greca, Silverman, & Prinstein, 1996). Following a disaster, social support from others may alleviate physical and psychological distress, as this support demonstrates others’ awareness of the situation and empathy (Mohay & Forbes, 2009). Individuals who have stronger sources of social support tend to cope better with life stresses than do individuals who do not have such means of support (Cohen & Wills, 1985). The perceived availability of social support during stressful situations, such as following a natural disaster, is thought to aid an individual in the provision of psychological resources needed to cope with the situation, and therefore serve to buffer the effects of stress (Cohen, 2004). This may operate through individuals’ reappraisal of the highly stressful situation and therefore decrease the risk of negative mental health outcomes (Cohen, 2004; Mohay & Forbes, 2009; Thoits, 1986). The perception of available social support may also result in the reduction or elimination of affective and physiological responses to the stressful event, which in turn may alter potentially maladaptive behavioral responses (Cohen et al., 2000). Additionally, social support provides an outlet for individuals to voice their problems and concerns, which has been associated with a reduction in intrusive thoughts that can serve to maintain maladaptive responses to a stressful event (Cohen et al., 2000; Lepore, Silver,

Donice M. Banks and Carl F. Weems, Department of Psychology, University of New Orleans. This research was supported in part by a grant from the New Orleans Institute of Mental Hygiene awarded to Carl F. Weems. Correspondence concerning this article should be addressed to Carl F. Weems, Department of Psychology, University of New Orleans, 2001 Geology and Psychology Building, New Orleans, LA 70148. E-mail: [email protected] 341

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

342

BANKS AND WEEMS

Wortman, & Wayment, 1996). Research suggests that higher levels of social support are associated with lower levels of PTSD symptoms following a disaster in adults (e.g., Brewin, Andrews, & Valentine, 2000; Ozer, Best, Lipsey, & Weiss, 2003; Weems, Watts et al., 2007), as well as in youth (e.g., La Greca, Silverman, Vernberg, & Prinstein, 1996; La Greca, Silverman, Lai, & Jaccard, 2010; Moore & Varela, 2010; Pina et al., 2008; Vernberg et al., 1996). The most frequently examined sources of social support in the youth postdisaster literature are family (e.g., parents and relatives), peers (e.g., friends and classmates), and teachers. Parents have been cited as the most important source of social support to elementary-school-age children following disasters (Pynoos & Nader, 1988; Vernberg & Vogel, 1993). Parents may fulfill roles such as providing comfort, nurturing, and a sense of physical safety, and they may serve to model coping behavior (Compas & Epping, 1993; Pynoos & Nader, 1988; Vernberg & Vogel, 1993). Peers may function to help youth in responding to a traumatic event by assisting in coping efforts, decreasing isolation, and decreasing youth’s subjective feelings of responding abnormally (e.g., Pynoos & Nader, 1988; Vernberg & Vogel, 1993). Whereas a distinction may be made between youths’ friends and classmates, Dubow and Ullman (1989) found that youth did not distinguish between the friend and classmate support items when completing their measure of social support (the Survey of Children’s Social Support). Teachers, as a source of social support, may serve to provide factual information about a disaster and its consequences, and may aid in reestablishing familiar roles and routines for disaster-exposed youth (Vernberg & Vogel, 1993; Vernberg et al., 1996). To date, eight studies have examined the link between social support and posthurricane psychological distress in youth samples (Hardin, Weinrich, Weinrich, Hardin, & Garrison, 1994; Jaycox et al., 2010; La Greca et al., 1996, 2010; Moore & Varela, 2010; Pina et al., 2008; Self-Brown, Lai, Thompson, McGill, & Kelley, 2013; Vernberg et al., 1996). Across these studies the percentage of minority youth has ranged from 16% to 71%. Out of these studies, seven have examined PTSD symptoms as an outcome and each has shown a negative association between social support and symptoms of PTSD (i.e., higher social support is associated with fewer PTSD symptoms; Jaycox et al., 2010; La Greca et al., 1996, 2010; Moore & Varela, 2010; Pina et al., 2008; Self-Brown et al., 2013; Vernberg et al., 1996). Only two of the studies have examined the link between social support, anxiety, and depression, with both studies showing a significant link (Hardin et al., 1994; Pina et al., 2008). Three studies have demonstrated a longitudinal association in which earlier social support was associated with later PTSD symptoms (Jaycox et al., 2010; La Greca et al., 1996, 2010) but none with multiple outcomes. In addition, the findings have varied as a function of the source of social support and the timing of assessment. The pattern seems to be one with family and peers being important with two of seven studies showing a link to family social support (Jaycox et al., 2010; La Greca et al., 1996) and three of seven showing a link to peer social support (La Greca et al., 2010; Pina et al., 2008; Self-Brown et al., 2013). The majority of the studies that have examined the role of social support in predicting psychological distress among hurricane-exposed youth are cross-sectional, and PTSD symp-

toms are the most commonly measured outcome. There is also a scarcity of research that has formally tested social support as a moderator of the link between hurricane exposure and posttraumatic stress symptoms in youth. Theory and findings in the extant literature imply not only main effect associations but also possible moderating or mediating relationships (with only two of eight studies examining moderation). La Greca et al. (2010) examined the relationship between hurricane-related exposure events, posthurricane major life events, social support, and posttraumatic stress symptoms. Participants (16% minority youth) in Grades 2 through 4 were assessed at 9 months (Time 1) and 21 months (Time 2) following Hurricane Charley. At Time 1, peer social support was found to moderate the relationship between youths’ disaster exposure and posttraumatic stress symptoms at the Time 1 assessment (9 months posthurricane). Specifically, youth who thought their lives were threatened by the hurricane and had greater perceived peer social support reported significantly fewer posttraumatic stress symptoms than did youth who thought their lives were threatened and had lower perceived peer social support. The experience of major life events posthurricane was also was found to decrease the amount of peer social support available to youth. When the participants were assessed at Time 2 (21 months posthurricane), youth with lower levels of perceived peer social support at that time reported higher levels of posttraumatic stress symptoms, although peer social support was not noted to moderate the relationship between disaster exposure and posttraumatic stress symptoms (La Greca et al., 2010). Moore and Varela (2010) also investigated social support as a moderator in the association between hurricane exposure and PTSD symptoms among 156 hurricane-exposed youth (71% minority) in Grades 4 through 6 assessed 33 months after Hurricane Katrina. Perceived social support from classmates was negatively related to symptoms of PTSD. However, the results did not show that social support moderated the relationship between disaster exposure and PTSD symptoms. In summary, research is needed to further understanding of the interrelations of different types of support, levels of exposure/time since exposure, and different symptoms, particularly among minority youth. This study therefore examined associations among both family and peer social support, level of hurricane exposure, intervening major life events, and multiple indices of psychological distress (specifically symptoms of PTSD, anxiety, and depression) among urban minority youth. Associations were tested in two studies. The first study used a large sample to examine concurrent associations across different cohorts assessed at different time frames from the event, and the second study used a longitudinal sample that involved reassessment of symptoms after a 6-month period. The following hypotheses were tested: Higher levels of hurricane exposure would be related to lower levels of social support from peers and family. Lower levels of social support from peers and family would be related to higher levels of psychological distress (i.e., symptoms of PTSD, anxiety, and depression). Social support would moderate the association between hurricane exposure and psychological distress such that the association between hurricane exposure and psychological distress would be stronger among youth with lower levels of peer and family social support and weaker among youth with higher levels of

SOCIAL SUPPORT

peer and family social support. Alternatively, we also tested if social support would mediate the association between hurricane exposure and psychological distress. Finally, the possibility of differential effects by age, gender, and time since the event on associations among hurricane exposure, social support, and psychological distress were also examined.

Study 1

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

Method Participants. The recruitment sample for the study was comprised of 1,198 youth in Grades 3 through 12 who were attending schools in New Orleans neighborhoods that received massive damage and almost total flooding following Hurricane Katrina. Consent was not obtained from the parents of 100 of the youth, and thus the final sample consisted of 1,098 participants aged 7 to 18 (median age ⫽ 14 years; 53% female; 90.8% African American) who were assessed at a single time point. Youth completed self-report measures to assess PTSD symptoms (PTSD-RI), anxiety and depression symptoms (modified version of the RCADS), hurricane exposure (survey of hurricane events and hurricane-related distress), major life events (LEC), and social support (SOCSS) approximately 36 to 65 months following Hurricane Katrina. Common Katrina-related events reported included: “thought someone might die” (65%), “clothes or toys ruined” (68%), “home badly damaged or destroyed” (65%), “witnessed others hurt during the storm” (39%), “thought you might die during the storm” (26%), and “had trouble getting food and water” (21%). Measures. Hurricane exposure. Hurricane Katrina exposure experiences were assessed via a survey of exposure to the hurricane and its aftermath based on the work of La Greca and colleagues (La Greca, Silverman, & Wasserstein, 1998; Vernberg et al., 1996). The survey consists of 16 items representing disaster exposure (i.e., exposure experiences). Sample items include: “Did you get hurt during the hurricane?” “Was your home damaged badly or destroyed by the hurricane?” Respondents indicated “Yes” (1) or “No” (0) to whether they were exposed to each event. Items are summed to create an exposure events score (i.e., number of hurricane-related exposure events experienced during and after the hurricane). This and similar measures of hurricane events have been shown to be valid indices of exposure (e.g., associated with PTSD symptoms; La Greca et al., 1996; Weems et al., 2013 [rs from .23 to .44]). Major life events. A short form of the Life Events Checklist (LEC; Johnson & McCutcheon, 1980) used in previous disaster research (e.g., La Greca et al., 1996) was administered to assess intervening life events (nondisaster events) associated with psychological distress symptoms. The short form contains 14 major life events that reflect personal stressors (e.g., serious illness or hospitalization; death of a family member or loved one), as well as other major stressful events (e.g., birth of a sibling). Respondents indicated “Yes” (1) or “No” (0) to whether each event occurred following the hurricane. The items endorsed are summed to yield a total number of major life events occurring since the hurricane. Studies have provided support for the validity of the measure and

343

shown that the 2-week test–retest reliability is .72 (e.g., Greenberg, Siegel, & Leitch, 1983). PTSD symptoms. Symptoms of PTSD were measured using a modified version of the Posttraumatic Stress Reaction Index for Children (PTSD-RI; Frederick, Pynoos, & Nadar, 1992). The PTSD-RI is a widely used instrument to assess posttraumatic stress reactions and thus allows for easy comparison among studies. The measure has been found to have good convergent validity (ranging from .70 to .82 in comparison other measures assessing PTSD symptoms) and internal consistency (many studies report Cronbach’s alpha in the range of 0.90; see Steinberg, Brymer, Decker, & Pynoos, 2004 for a review). The internal consistency for this study (Cronbach’s alpha ⫽ .87) was similar to reports from other studies (e.g., Moore & Varela, 2010; Vernberg et al., 1996). As done in previous research (Hensley & Varela, 2008; Vernberg et al., 1996; La Greca et al., 1996), the PTSD-RI administered to participants in the proposed study contains 20 items, with answer choices modified for ease of administration from the original five options to three options (none of the time, some of the time, most of the time; coded as 0, 2, 4, respectively) and youth were instructed to rate their symptoms in response to the Hurricane Katrina events. Total PTSD-RI scores thus range from 0 to 80. The system developed by Frederick et al. (1992) was used to classify the sample by the severity of symptoms, and includes the following categories: Doubtful (total score of 0 –11), Mild (12–24), Moderate (25–39), Severe (40 –59), and Very Severe (60 – 80). Anxiety and depression symptoms. Symptoms of anxiety (i.e., Generalized Anxiety Disorder, Separation Anxiety Disorder, Social Phobia, and Panic Disorder) and Major Depressive Disorder were assessed using a modified version of the Revised Child Anxiety and Depression Scales (RCADS; Chorpita, Yim, Moffitt, Umemoto, & Francis, 2000; Spence, 1997). The measure consists of 38 items that assess the symptoms of the aforementioned anxiety and depressive disorders based on the Diagnostic and Statistical Manual of Mental Disorders-IV (DSM–IV) criteria (American Psychological Association, 1994). Youth are asked to rate how often they experience these symptoms on a 4-point scale. Possible responses consist of Never (1), Sometimes (2), Often (3), and Always (4). In the current study, the internal consistency was .83 for the depression subscale and .94 for the overall anxiety subscale. The measure has additionally been demonstrated to have good convergent validity with other measures assessing youth anxiety and depression symptoms (Chorpita et al., 2000) Social support. A short form of the Survey of Children’s Social Support (SOCSS; Dubow & Ullman, 1989) was used to assess youths’ perceived social support. This short form of the SOCSS was created and utilized by La Greca et al. (2010). Five of the measure’s items assess perceived family support (e.g., “Can you count on your family for help or advice when you have problems?”), and six of the items assess perceived peer support (e.g., “Do you think your friends care about you?”). La Greca and colleagues (2010) selected items from the family and peer/friend SOCSS subscales because these two sources have been reported to provide the most support for youth following disasters (Prinstein, La Greca, Vernberg, & Silverman, 1996). The items selected from the family and peer/friend subscales were the ones with the highest factor loadings on each respective subscale (La Greca et al., 2010). Responses are rated on a 5-point scale ranging from 0 (never) to 4 (always). Relevant

BANKS AND WEEMS

344

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

items are averaged to calculate individual subscale scores. Dubow and Ullman (1989) found that the 3- to 4-week test– retest reliability for the total APP subscale of the SOCSS was .75, and internal consistencies for the subscales were between .78 and .83. In the current study, the internal consistency for the peer support subscale was .71, and the internal consistency for the family support subscale was .88. Procedures. Data collection was conducted as a part of the schools’ counseling curriculum. Parental active written informed consent for use of the data in research was obtained. Oral assent was also obtained from youth, and youth were not required to participate in the study or to complete the questionnaires. The procedures were reviewed by the UNO IRB and use of the de-identified data was approved. Participants completed the measures within a group classroom setting and were assisted by trained staff. All staff had taken a class in research methods and were trained on the assessment procedures and measures via a standardized presentation consisting of at least two hourlong meetings. Younger children were read the instructions along with each item, while trained assessors helped individual children as necessary (as done in previous research; La Greca et al., 1996). The teachers and/or counselors for each classroom were queried about the reading level of the class and shown the measures. Students were read the measures in classrooms where teachers or counselors indicated possible need or where assessors noticed difficulty with reading level.

Results and Discussion Preliminary analyses. The distribution of the PTSD symptom severity categories for the participants was as follows: 36.5% Doubtful, 32.8% Mild, 20.2% Moderate, 9.3% Severe, and 1.3% Very Severe. Examination of the social support, life events, and hurricane events scores’ range and skew indicated acceptable levels for the planned analyses. Scores of the symptoms of PTSD, depression, and anxiety were somewhat positively skewed but were retained for all analyses. However, given the skew, main analyses were supplemented with identi-

cal analyses using logarithmic transformations to normalize the distributions or nonparametric equivalents. Overall, the parametric and transformed/nonparametric correlational analyses yielded very similar findings and mostly identical conclusions. Because of this, parametric analyses of the nontransformed outcome variables are emphasized in this report with differences reported as needed to highlight any different conclusions. Demographic variables were included in correlational analyses to determine which demographic variables to include in subsequent analyses (see Table 1; gender was coded as 1 ⫽ boy, 2 ⫽ girl). Gender and age were both significantly related to the outcome variables of symptoms of PTSD, depression, and anxiety. Age was negatively related to psychological distress symptoms such that older children reported fewer symptoms of distress than did younger children. Girls reported more symptoms of PTSD (M ⫽ 21.76, SD ⫽ 14.87) than boys (M ⫽ 15.96, SD ⫽ 13.65), t(1014) ⫽ ⫺6.50, p ⬍ .001, as well as higher levels of anxiety, M ⫽ 20.71, SD ⫽ 14.97 for girls; M ⫽ 15.47, SD ⫽ 13.95 for boys; t(998) ⫽ ⫺3.65, p ⬍ .001, and depression, M ⫽ 7.29, SD ⫽ 5.60 for girls; M ⫽ 6.03, SD ⫽ 5.24 for boys; t(991) ⫽ ⫺5.67, p ⬍ .001. Because of the significant relationship between age and gender with psychological distress symptoms, these demographic variables were controlled for in subsequent regression analyses. Associations between hurricane exposure, social support, and psychological distress. Correlational analyses are presented in Table 1. Higher levels of hurricane exposure were significantly related to lower levels of family social support. Social support from peers and family were both significantly negatively related to symptoms of PTSD, anxiety, and depression. Regression analyses were next conducted to examine whether lower levels of peer and family social support were related to higher levels of distress while controlling for hurricane exposure, the time since Hurricane Katrina, major intervening life events, age, and gender (see Table 2). When controlling for these additional variables, peer social support was still significantly negatively associated with symptoms of PTSD, anxiety, and depression. Family social support was

Table 1. Zero-Order Correlations for Predictor, Outcome, and Demographic Variables in Study 1 1

2

3

4

5

6

7

8

9

10

1. Peer social spport 2. Family social support .23ⴱⴱ 3. PTSD symptoms ⫺.30ⴱⴱ ⫺.10ⴱⴱ ⴱⴱ 4. Anxiety ⫺.34 ⫺.11ⴱⴱ .63ⴱⴱ 5. Depression ⫺.35ⴱⴱ ⫺.24ⴱⴱ .57ⴱⴱ .77ⴱⴱ ⴱ ⴱⴱ 6. Life events ⫺.01 ⫺.07 .29 .18ⴱⴱ .19ⴱⴱ 7. Hurricane exposure ⫺.02 ⫺.11ⴱⴱ .41ⴱⴱ .27ⴱⴱ .24ⴱⴱ .42ⴱⴱ † ⴱⴱ 8. Months since Katrina .06 ⫺.08 ⫺.05 ⫺.02 .01 .11ⴱⴱ .16ⴱⴱ 9. Age .19ⴱⴱ ⫺.14ⴱⴱ ⫺.17ⴱⴱ ⫺.14ⴱⴱ ⫺.08ⴱⴱ ⫺.01 .08ⴱⴱ .41ⴱⴱ ⴱⴱ ⴱⴱ ⴱⴱ ⴱⴱ ⴱⴱ 10. Gender .12 .05 .20 .18 .12 ⫺.01 .13 .04 .06ⴱ Mean (SD) 3.12 (.76) 3.07 (1.04) 19.08 (14.61) 18.18 (14.63) 6.67 (5.45) 2.69 (2.19) 6.49 (3.23) 47.33 (11.47) 13.51 (2.44) na Range 0–4 0–4 0–80 0–112 0–40 0–14 0–16 36–65 7–18 Note. Gender was coded 1 ⫽ boy, 2 ⫽ girl. r ⫽ .058, p ⫽ .056. ⴱ p ⬍ .05 (2-tailed).



ⴱⴱ

p ⬍ .01 (2-tailed).

SOCIAL SUPPORT

345

Table 2. Prediction of Psychological Distress Symptoms in Study 1 Outcome

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

PTSD

Anxiety

Depression

Variable Step 1 Gender Age Months since Katrina Life events Hurricane exposure Peer social support Family social support R2 ⫽ .33 Step 2 Peer support ⫻ exposure Family support ⫻ exposure R2 ⫽ .33 ⌬R2 ⫽ .002 Step 1 Gender Age Months since Katrina Life events Hurricane exposure Peer social support Family social support R2 ⫽ .24 Step 2 Peer support ⫻ exposure Family support ⫻ exposure R2 ⫽ .24 ⌬R2 ⫽ .001 Step 1 Gender Age Months since Katrina Life events Hurricane exposure Peer social support Family social support R2 ⫽ .24 Step 2 Peer support ⫻ exposure Family support ⫻ exposure R2 ⫽ .24 ⌬R2 ⫽ .001

B

SE B



t

5.89 ⫺.84 ⫺.05 .97 1.45 ⫺5.50 .02

.78 .18 .04 .20 .13 .53 .39

.20 ⫺.14 .04 .14 .32 ⫺.29 ⬍.01

7.59ⴱⴱ ⫺4.84ⴱⴱ ⫺1.36 4.95ⴱⴱ 10.91ⴱⴱ ⫺10.33ⴱⴱ .06

.23 ⫺.01

.15 .11

.04 ⬍.01

1.54 ⫺.12

5.76 ⫺.67 .00 .53 .98 ⫺6.42 ⫺.43

.85 .19 .04 .22 .15 .58 .43

.20 ⫺.11 .01 .08 .22 ⫺.33 ⫺.03

6.76ⴱⴱ ⫺3.52ⴱⴱ ⫺.10 2.47ⴱ 6.77ⴱⴱ ⫺11.07ⴱⴱ ⫺1.00

⫺.02 ⫺.10

.16 .12

.01 ⫺.03

⫺.11 ⫺.82

1.55 ⫺.16 ⫺.01 .27 .29 ⫺2.26 ⫺.87

.32 .07 .02 .08 .05 .21 .16

.14 ⫺.07 ⫺.02 .11 .17 ⫺.32 ⫺.16

4.90ⴱⴱ ⫺2.23ⴱ ⫺.61 3.37ⴱⴱ 5.37ⴱⴱ ⫺10.55ⴱⴱ ⫺5.42ⴱⴱ

.03 ⫺.04

.06 .04

.02 ⫺.03

.52 ⫺.96

Note. The full models associated with each of the interaction terms were significant, ps ⬍ .001. The F values associated with ⌬R2 were not significant. Gender was coded 1 ⫽ boy, 2 ⫽ girl. ⴱ p ⬍ .05. ⴱⴱ p ⬍ .01.

uniquely associated with depression symptoms, but not with symptoms of PTSD or anxiety. Social support as a moderator of hurricane exposure link to psychological distress. Building upon the regression analyses predicting symptoms of PTSD, anxiety, and depression noted previously, additional steps were added to investigate whether social support moderated the association between hurricane exposure and psychological distress. Prior to conducting the analyses, the variables of average hurricane events, family social support, and peer social support were centered to the mean to reduce the possibility of multicollinearity influencing the re-

sults. Family social support, peer social support, hurricane exposure, life events, months since Hurricane Katrina, age, and gender were entered in Step 1, and two-way interactions between hurricane exposure (centered) and family social support (centered) and between hurricane exposure (centered) and peer social support (centered) were entered in Step 2. The results of these analyses are also presented in Table 2. None of the interactions were significant in the parametric analyses. In analyses of the transformed outcome variables, the pattern of findings for the outcomes of anxiety and depression remained very similar with one exception: The hurricane exposure by peer social support interaction significantly predicted PTSD symptoms (␤ ⫽

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

346

BANKS AND WEEMS

.06, p ⬍ .05). The interaction is depicted in Figure 1. Post hoc probing (as recommended by Holmbeck, 2002; i.e., computation of two new conditional moderator variables using separate analyses) indicated that hurricane exposure was associated with PTSD symptoms among both those with low and high peer social support; however, the association was stronger among youth with high levels of peer social support, t(998) ⫽ 10.14, p ⬍ .001; ␤ ⫽ .39, than those with lower levels of peer social support, t(998) ⫽ 7.77, p ⬍ .001; ␤ ⫽ .29. Post hoc probing of the interaction with hurricane events as the moderator of the relation between peer social support and PTSD symptoms was consistent with this interpretation. Peer social support had a weaker association with PTSD symptoms among youth with high hurricane exposure, t(998) ⫽ ⫺5.92, p ⬍ .001, ␤ ⫽ ⫺.21, than youth with lower hurricane exposure, t(998) ⫽ ⫺8.50, p ⬍ .001; ␤ ⫽ ⫺.32. The stronger association between hurricane exposure and PTSD symptoms considered together with the weaker association between peer social support and PTSD symptoms for those with high peer social support suggests that peer social support is not buffering the effect of hurricane exposure against the development of PTSD symptoms. Rather, high hurricane exposure level is lowering the positive effect of peer social support against the development of PTSD symptoms. However, as depicted in Figure 1, this difference was rather small. Alternative explanation of social support as a mediator. As can be seen in Table 1 and noted previously, the first three conditions of mediation were met for the associations between hurricane exposure with psychological distress and with family social support, and between both sources of social support with psychological distress. However, as can be seen in Table 2, hurricane exposure remained a significant predictor of psychological distress while controlling for social support, so no evidence was found for mediation. Identical regression analyses were also conducted without the covariates of gender, age, months since Hurricane Katrina, and life events, and the pattern of findings remained the same. Differential effects by gender, cohort, and age. The above analyses were further examined by gender, cohort, and age to determine whether there were any differential effects. In the

Figure 1. Interaction between hurricane exposure and peer social support predicting PTSD symptoms.

analyses examining potential gender differences, peer social support significantly predicted symptoms of PTSD, anxiety, and depression for both boys and girls. Family social support significantly predicted depression for both genders with a stronger effect for girls, r(527) ⫽ ⫺.321, p ⬍ .001, than for boys, r(447) ⫽ ⫺.164, p ⬍ .05 (z ⫽ 2.59, p ⬍ .01). Overall, the pattern of findings was highly similar for both boys and girls. Next, potential differences in findings among three cohorts were examined. Cohort 1 included participants whose assessment occurred approximately 36 to 40 months following Hurricane Katrina. The participants in Cohort 2 were assessed approximately 48 to 50 months post-Katrina, and Cohort 3 was assessed approximately 60 to 65 months post-Katrina. The cohorts differed in terms of hurricane-exposure events reported, mean age, intervening life events reported, and level of family social support. Cohort 1 reported fewer hurricane-exposure experiences and intervening major life events than Cohorts 2 and 3, and reported significantly more family social support than Cohort 3. Participants’ mean age in Cohort 1 was younger than Cohorts 2 and 3, and the mean age of Cohort 2 was younger than Cohort 3. Each of the main analyses were run separately in each cohort. In general, the pattern of findings was similar across cohorts. Peer social support significantly predicted each dimension of psychological distress (PTSD, anxiety, and depression symptoms) among each of the three cohorts. Its effect in predicting depression symptoms was stronger for Cohort 1, r(492) ⫽ ⫺.42, p ⬍ .001, than for Cohort 3, r(296) ⫽ ⫺.27, p ⬍ .001 (z ⫽ ⫺2.31, p ⬍ .05). Family social support significantly predicted PTSD symptoms in Cohort 3, and also predicted depression symptoms in each of the three cohorts. Its effect in predicting depression symptoms was stronger for Cohort 3, r(296) ⫽ ⫺.32, p ⬍ .001, than for Cohort 1, r(492) ⫽ ⫺.18, p ⫽ .01 (z ⫽ 2.03, p ⬍ .05). Two differences emerged among the cohorts when social support and hurricane exposure interactions were tested. One difference was that a statistically significant interaction term was obtained between peer social support and hurricane exposure in Cohort 1 for the prediction of PTSD symptoms (␤ ⫽ .09, p ⬍ .05). Post hoc probing of the interaction (Holmbeck, 2002) indicated that the association between hurricane exposure and PTSD symptoms was positive and significant among both those with higher levels of peer social support, t(501) ⫽ 7.31, p ⬍ .001; ␤ ⫽ .39 and with lower levels of peer social support, t(501) ⫽ 4.29, p ⬍ .001; ␤ ⫽ .22. The second difference was that the interaction between hurricane exposure and peer social support also significantly predicted depression symptoms in Cohort 1 (␤ ⫽ .08, p ⬍ .05) when tested in the analyses of transformed outcome variables. Post hoc probing indicated that, again, hurricane exposure had a stronger association with depression symptoms for youth with higher levels of peer social support, t(492) ⫽ 4.16, p ⬍ .001; ␤ ⫽ .24, than those with low peer social support, t(492) ⫽ 3.25, p ⫽ .001; ␤ ⫽ .23. However, given the age differences across cohorts, the differential cohort effects may be confounded by age. Thus, in the final set of analyses, differential effects by age were tested. Age was examined as a continuous moderator. Building from the regression models in Table 2, an additional third step was added with interaction terms. The variable of cohort (i.e., a measure of time since Hurricane Katrina) was added to the list of variables in Step 1, and interaction terms of age by social support and cohort by social support were added in Step 2. Three-way interactions between age, hurricane exposure, and the two sources

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

SOCIAL SUPPORT

of social support were tested for the prediction of symptoms of PTSD, anxiety, and depression in Step 3, and none of the interactions was significant. A two-way interaction between peer social support and age significantly predicted symptoms of PTSD (␤ ⫽ .06, p ⬍ .05). Post hoc analyses showed that there was a stronger negative association between peer social support and PTSD symptoms among younger youth, t(998) ⫽ ⫺9.58, p ⬍ .001; ␤ ⫽ ⫺.31, than older youth, t(998) ⫽ ⫺5.84, p ⬍ .001; ␤ ⫽ ⫺.25. However, it should be noted that this interaction was not significant when tested in the analyses of the transformed outcome variables. Twoway interactions between family social support and age significantly predicted PTSD symptoms (␤ ⫽ ⫺.13, p ⬍ .001), anxiety (␤ ⫽ ⫺.09, p ⬍ .01), and depression symptoms (␤ ⫽ ⫺.09, p ⬍ .01). Again, post hoc analyses were conducted, showing that family social support had a negative association with PTSD symptoms among older youth, t(998) ⫽ ⫺3.56, p ⬍ .001; ␤ ⫽ ⫺.14, but a positive association with PTSD symptoms among younger youth, t(998) ⫽ 3.59, p ⬍ .001; ␤ ⫽ .14. Family social support had a negative association with anxiety symptoms for older youth, t(967) ⫽ ⫺2.87, p ⬍ .01; ␤ ⫽ ⫺.12, but the association was not significant among younger youth, t(967) ⫽ 1.38, p ⫽ .17; ␤ ⫽ .06. A similar pattern was obtained between family social support and depression symptoms, with a significant negative association between the two among older youth, t(974) ⫽ ⫺6.36, p ⬍ .001; ␤ ⫽ ⫺.26, and a nonsignificant association obtained for younger youth, t(974) ⫽ ⫺1.58, p ⫽ .11; ␤ ⫽ ⫺.06. Overall, peer and family social support were associated with lower psychological distress for older youth. Family social support, in particular, was associated with fewer symptoms of PTSD, anxiety, and depression for older youth. Peer social support was associated with fewer PTSD symptoms among younger youth, and perhaps more strongly so than for older youth. This finding must be interpreted with some caution since the results differed between the analysis of the transformed and nontransformed outcome variables. In contrast, family social support had a positive association with PTSD symptoms for younger youth, and was not significantly associated with anxiety or depression symptoms. Further, none of the cohort by social support interactions was significant in any of the analyses examining differential effects by age, suggesting that the cohort effect may better be explained instead as a function of age.

Study 2 The purpose of Study 2 was to replicate the concurrent analyses from Study 1 using a longitudinal design (i.e., to determine whether the findings hold over time to predict later distress).

Method Participants. Data for this study came from an existing data set collected by the Youth and Family Anxiety, Stress, and Phobia Lab at UNO consisting of a longitudinal sample of youth aged 8 to 15 years (median age ⫽ 11.5 years; 55% male; 97% African American) who were assessed twice during a 6-month period (further details about the sample are reported in Weems et al., 2010, 2013). The first assessment occurred approximately 24 months following Hurricane Katrina (Time 1), and the

347

second assessment occurred approximately 30 months postKatrina (Time 2). All of the participants attended a school in New Orleans that received massive damage and almost total flooding following Hurricane Katrina. The total number of participants for inclusion in the study was 192 following data cleaning and examination of missing data. Measures. Participants completed the same measures as those in Study 1. The PTSD-RI, modified version of the RCADS, survey of hurricane events and hurricane-related distress, LEC, and SOCSS were administered at the initial assessment (Time 1). The internal consistencies of the SOCSS subscales were .66 for peers and .84 for family. Additionally, participants were readministered the PTSD-RI and modified version of the RCADS approximately 6 months following the initial assessment to reassess symptoms of PTSD, anxiety, and depression (Time 2). Procedures. The procedures for Study 2 were the same as Study 1 except that students were reassessed at Time 2 with the PTSD-RI and RCADS.

Results and Discussion Preliminary analyses. The participants in Study 2 reported similar levels of hurricane exposure experiences as those in Study 1. At Time 1, the distribution of the PTSD symptom severity categories for the participants was as follows: 27% Doubtful, 32% Mild, 22% Moderate, 15% Severe, and 4% Very Severe. At Time 2, the distribution of the PTSD symptom severity categories was: 25% Doubtful, 36% Mild, 23% Moderate, 13% Severe, and 4% Very Severe. Descriptive analyses and examination of the scores’ ranges and skew indicated acceptable levels for the planned analyses. In this study, the parametric and nonparametric correlational analyses and the main analyses of both transformed and nontransformed outcome variables all yielded identical findings. Therefore, only the parametric and nontransformed variable analyses are reported. Demographic variables were included in correlational analyses to determine which demographic variables to include in subsequent analyses (see Table 3; gender was coded 1 ⫽ boy, 2 ⫽ girl). Age and gender were both significantly related to the Time 2 outcome variables of symptoms of PTSD, depression, and anxiety. Age was negatively related to psychological distress symptoms such that older children reported fewer symptoms of distress than did younger children. Girls reported more symptoms of PTSD (M ⫽ 28.81, SD ⫽ 16.15) than boys (M ⫽ 18.25, SD ⫽ 12.51), t(164) ⫽ ⫺5.02, p ⬍ .001, as well as higher levels of anxiety, M ⫽ 24.23, SD ⫽ 15.91 for girls; M ⫽ 16.73, SD ⫽ 12.76 for boys; t(168) ⫽ ⫺3.57, p ⬍ .001, and depression, M ⫽ 8.26, SD ⫽ 5.89 for girls; M ⫽ 5.71, SD ⫽ 4.65 for boys; t(166) ⫽ ⫺3.30, p ⫽ .001. Age and gender were therefore controlled for in subsequent regression analyses. Associations between hurricane exposure, social support, and psychological distress. Correlational analyses were used to examine whether higher levels of hurricane exposure were related to lower levels of peer and family social support at the Time 1 assessment. Results are shown in Table 3.

BANKS AND WEEMS

⫺.14ⴱ .26ⴱⴱ 7.06 (2.81) 0–16

⫺.07 11.50 (1.63) 8–15

na

There was a significant negative correlation between hurricane exposure and peer social support. However, the correlation between exposure and family social support was not significant. Next, the association between social support and psychological distress was examined to determine whether lower levels of peer and family social support (assessed at Time 1) were related to higher levels of psychological distress at both Time 1 and Time 2. Social support from peers was significantly negatively related to both Time 1 and Time 2 symptoms of PTSD, anxiety, and depression. In contrast, family social support was not significantly related to symptoms of PTSD, anxiety, or depression at Time 1 or Time 2. Regression analyses were then conducted to examine whether lower levels of social support were associated with higher Time 2 psychological distress while controlling for hurricane exposure, major life events, Time 1 psychological distress, age, and gender (see Table 4). Results showed that peer social support was significantly predictive of lower depression at Time 2, and that family social support was again not significantly related to symptoms of PTSD, anxiety, or depression at Time 2 while controlling for the additional variables.



Note. T1 ⫽ Time 1; T2 ⫽ Time 2; Exposure ⫽ Hurricane Exposure; Gender was coded 1 ⫽ boy, 2 ⫽ girl. p ⬍ .05 (2-tailed). ⴱⴱ p ⬍ .01 (2-tailed).

.42ⴱⴱ ⫺.16ⴱ .07 2.80 (2.07) 0–14 .19ⴱⴱ .29ⴱⴱ ⫺.15ⴱ .24ⴱⴱ 6.88 (5.40) 0–40 .81ⴱⴱ .18ⴱ .34ⴱⴱ ⫺.21ⴱⴱ .25ⴱⴱ 20.21 (14.75) 0–112 .46ⴱⴱ .55ⴱⴱ .51ⴱⴱ .22ⴱⴱ .31ⴱⴱ ⫺.10 .22ⴱⴱ 7.21 (6.12) 0–40 1. Peer social support 2. Family social support 3. PTSD (T1) 4. Anxiety (T1) 5. Depression (T1) 6. PTSD (T2) 7. Anxiety (T2) 8. Depression (T2) 9. Life events 10. Exposure 11. Age 12. Gender Mean (SD) Range

.13 ⫺.35ⴱⴱ ⫺.47ⴱⴱ ⫺.44ⴱⴱ ⫺.31ⴱⴱ ⫺.37ⴱⴱ ⫺.37ⴱⴱ ⫺.22ⴱⴱ ⫺.15ⴱ .21ⴱⴱ .06 2.97 (.80) 0–4

.03 .01 ⫺.02 .01 ⫺.08 ⫺.13 ⫺.07 .04 ⫺.07 .16ⴱ 3.32 (.88) 0–4

.69ⴱⴱ .61ⴱⴱ .57ⴱⴱ .54ⴱⴱ .46ⴱⴱ .35ⴱⴱ .48ⴱⴱ ⫺.22ⴱⴱ .35ⴱⴱ 24.15 (17.13) 0–80

.86ⴱⴱ .47ⴱⴱ .58ⴱⴱ .48ⴱⴱ .25ⴱⴱ .37ⴱⴱ ⫺.12 .28ⴱⴱ 21.71 (17.39) 0–112

.71ⴱⴱ .68ⴱⴱ .28ⴱⴱ .43ⴱⴱ ⫺.33ⴱⴱ .35ⴱⴱ 23.09 (15.20) 0–80

8 7 6 5 4 3 2 1

Table 3. Zero-Order Correlations for Predictor, Outcome, and Demographic Variables in Study 2

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

9

10

11

12

348

Social support as a moderator between hurricane exposure and psychological distress. Similar to Study 1, hierarchical regression analyses were used to examine whether social support moderated the association between hurricane exposure and Time 2 psychological distress. Three separate hierarchical regressions were conducted with the outcome variables of PTSD, anxiety, and depression symptoms. The variables of gender, age, life events, hurricane exposure, Time 1 psychological distress, peer social support, and family social support were entered in Step 1, and two-way interactions between hurricane exposure (centered) and the two sources of social support (both centered) were entered in Step 2. The results of these analyses are also presented in Table 4. None of the interactions was significant, indicating that social support did not moderate the association between hurricane exposure and Time 2 psychological distress. Alternative explanation of social support as a mediator. Both peer and family social support were examined as mediators between hurricane exposure and Time 2 psychological distress. As noted above and seen in Table 3, hurricane exposure was associated with symptoms of PTSD, anxiety, and depression (Condition 1 of mediation), and was also associated with peer social support (Condition 2). Peer social support was associated with symptoms of PTSD, anxiety, and depression at Time 2 (Condition 3). Hurricane exposure remained a significant predictor of PTSD symptoms while controlling for social support, but was no longer a significant predictor of anxiety and depression symptoms (Condition 4; see Table 4). Therefore, Sobel tests were conducted to determine whether peer social support carried a significant proportion of the variance in predicting anxiety and depression symptoms (i.e., whether it had a significant indirect effect on anxiety and depression). Sobel test results showed that peer social support did not have a significant indirect effect on anxiety (Sobel test statistic ⫽ 1.20, SE ⫽ .07, p ⫽ .23) or depression (Sobel test statistic ⫽ 1.59, SE ⫽ .03, p ⫽ .11). Therefore, there was little evidence for social support mediating the hurricane exposure-psychological distress association. The nonsignificant relationship between hurricane exposure and the

SOCIAL SUPPORT

349

Table 4. Prediction of Time 2 Psychological Distress Symptoms in Study 2 Outcome

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

PTSD

Anxiety

Depression

Variable Step 1 Gender Age Life events Hurricane exposure PTSD Time 1 Peer social support Family social support R2 ⫽ .42 Step 2 Peer support ⫻ exposure Family support ⫻ exposure R2 ⫽ .44 ⌬R2 ⫽ .01 Step 1 Gender Age Life events Hurricane exposure Anxiety Time 1 Peer social support Family social support R2 ⫽ .39 Step 2 Peer support ⫻ exposure Family support ⫻ exposure R2 ⫽ .40 ⌬R2 ⫽ .01 Step 1 Gender Age Life events Hurricane exposure Depression Time 1 Peer social support Family social support R2 ⫽ .34 Step 2 Peer support ⫻ exposure Family support ⫻ exposure R2 ⫽ .34 ⌬R2 ⫽ .002

B

SE B



t

5.35 ⫺1.88 .25 .89 .31 ⫺2.28 ⫺.42

1.90 .54 .50 .37 .07 1.19 1.01

.18 ⫺.20 .03 .17 .34 ⫺.12 ⫺.02

2.82ⴱⴱ ⫺3.45ⴱⴱ .51 2.44ⴱⴱ 4.72ⴱⴱ ⫺1.92 ⫺.41

⫺.28 .34

.33 .38

⫺.05 .05

⫺.85 .90

3.31 ⫺.90 ⫺.09 .67 .38 ⫺2.06 ⫺1.55

2.05 .60 .55 .39 .07 1.42 1.11

.11 ⫺.09 ⫺.01 .13 .44 ⫺.11 ⫺.09

1.61 ⫺1.46 ⫺.16 1.72 5.64ⴱⴱ ⫺1.45 ⫺1.40

⫺.15 ⫺.61

.35 .43

⫺.03 ⫺.09

⫺.43 ⫺1.42

1.73 ⫺.20 .10 .15 .32 ⫺1.19 ⫺.76

.75 .22 .20 .14 .07 .51 .41

.16 ⫺.06 .03 .08 .37 ⫺.18 ⫺.12

2.30ⴱ ⫺.92 .47 1.03 4.86ⴱⴱ ⫺2.36ⴱ ⫺1.84

.01 ⫺.11

.13 .16

.01 ⫺.04

.09 ⫺.66

Note. The full models associated with each of the interaction terms were significant, ps ⬍ .001. The F values associated with ⌬R2 were not significant. Gender was coded 1 ⫽ boy, 2 ⫽ girl. ⴱ p ⬍ .05. ⴱⴱ p ⬍ .01.

outcomes of anxiety and depression when controlling for social support is likely because of the influence of the additional covariates (gender, age, life events, and Time 1 psychological distress) that were included in the regression analyses.

General Discussion The study increments the existing knowledge of the role of social support in youth’s emotional functioning following a natural disaster in a number of ways. As expected, lower social support was associated with higher psychological distress (i.e., symptoms of PTSD, anxiety, and depression) when examined

concurrently. This is consistent with findings from previous studies (e.g., Hardin et al., 1994; Jaycox et al., 2010; La Greca et al., 1996, 2010; Pina et al., 2008; Self-Brown et al., 2013) but adds to the literature by demonstrating the protective nature of both peer and family social support following hurricanes across multiple indices of distress at time frames more distal to the event than in any previous study of youth. Additionally, there was an association between lower social support and higher psychological distress longitudinally. This finding increments the few studies that have examined this (Jaycox et al., 2010; La Greca et al., 1996, 2010) and furthers the understanding by demonstrating a longitudinal association between social support

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

350

BANKS AND WEEMS

and other indices of psychological distress (anxiety and depression symptoms). Overall, high peer social support was consistently significantly associated with lower levels of PTSD, anxiety, and depression symptoms in the cross-sectional sample, and these associations held while controlling for factors that may affect distress such as major life events. Longitudinally, higher peer social support was also associated with lower levels of PTSD, anxiety, and depression symptoms, and remained uniquely associated with lower depression even while controlling for additional factors such as major life events and distress reported at the initial assessment. Importantly, however, findings suggest the protective effects of peer social support may be negatively impacted by high hurricane exposure. That is, the effect of high hurricane exposure may overwhelm the benefit of peer social support against the development of PTSD symptoms for some youth. To date, this is the first study to show a conditional association between the protective role of social support with lower PTSD symptoms based upon hurricane exposure. It should be noted that this finding emerged following the transformation of PTSD symptoms to normalize the distribution, and therefore should be interpreted with some caution. The results of the present study additionally inform us about the nature of social support’s association with anxiety and depression, two outcomes that have been studied less frequently than PTSD symptoms in the extant literature in relation to hurricane exposure and social support. Social support from peers and family were both related to anxiety and depression, demonstrating associations concurrently and longitudinally. The finding that girls reported higher rates of anxiety, depression, and PTSD symptoms than boys is consistent with previous findings in the literature that girls exhibit higher levels of posthurricane PTSD symptoms (e.g., La Greca et al., 2010; Kronenberg et al., 2010; Weems, Pina et al., 2007; Vernberg et al., 1996). Social support did show a protective effect against psychological distress for both genders. Higher peer support was associated with lower levels of PTSD symptoms, anxiety, and depression symptoms for boys and girls. Higher family social support was associated with lower depression for both genders, although this effect was stronger for girls than boys. While speculative, because girls reported higher levels of depression than did boys and family social support predicted lower depression for girls, this may point to the importance of family social support following a natural disaster in the development of depression among girls. In terms of age, there were variations among reported psychological distress and associations with social support. Peer social support was associated with lower PTSD symptoms for both older and younger youth. This negative association may have been stronger for younger youth, although this finding is somewhat tentative given that it did not hold in the analyses of transformed outcome variables. Family social support functioned differently between younger and older youth in the prediction of distress. Among younger youth, higher levels of family social support were associated with higher levels of PTSD symptoms, whereas higher levels of family social support among older youth were associated with lower levels of symptoms of PTSD, anxiety, and depression. While these findings may suggest that family social support was protective for older youth, they may also indicate that younger youth are more vulnerable to the effects of a disaster (see also Kronenberg et al., 2010). Indeed, results showed that, in general,

younger youth reported higher levels of PTSD, anxiety, and depression symptoms than older youth, and some evidence from prior studies examining postdisaster psychological distress support these findings (e.g., Giannopoulou et al., 2006; Kronenberg et al., 2010; Lonigan, Shannon, Taylor, Finch, & Sallee, 1994; McDermott & Palmer, 2002; Osofsky, Osofsky, Kronenberg, Brennan, & Hansel, 2009). It is also possible that younger youth had caregivers who experienced higher rates of distress which, in turn, affected youths’ distress. There is evidence to suggest that hurricane-exposed youth whose mothers report higher levels of distress posthurricane (including PTSD and depression symptoms) tend to report higher levels of distress, including PTSD and depression (Scheeringa & Zeanah, 2008). Additionally, maternal care has been found to have measurable effects on stress reactivity (i.e., susceptibility to anxiety disorders) in animal models (e.g., Meaney, 2001; Parker, Buckmaster, Sundlass, Schatzberg, & Lyons, 2006), and the evidence suggests this is true for humans as well (e.g., Bernard & Dozier, 2010). While speculative, it is possible that the younger youth in this sample were more susceptible to developing psychological distress because of environmental factors such as maternal care. As predicted, hurricane exposure was related to lower levels of social support. This finding is similar to the results from other studies of youth postdisaster (e.g., La Greca et al., 2010; Vernberg et al., 1996), and makes theoretical sense because disasters may disrupt social support networks through mechanisms such as family separation and difficulty maintaining peer connections, which may serve to confer risk by altering an individual’s sources of social support that satisfy the need of social relatedness (Weems & Overstreet, 2008). A possible direction for future research would be to examine youth’s predisaster levels of social support, as this could provide further information concerning the disruption to support networks following a hurricane and also how predisaster support levels may be related to psychological distress outcomes. Hurricane exposure was also a risk factor for the development of psychological distress. Higher levels of hurricane exposure were associated with higher levels of PTSD, anxiety, and depression symptoms in the cross-sectional sample, and predicted higher PTSD symptoms longitudinally. These findings provide further evidence that the number and intensity of hurricane events experienced is related to distress such as has been shown in previous research (e.g., La Greca et al., 1998; Weems et al., 2013). The results are also consistent with theory suggesting that negative effects to mental health may follow a natural disaster because of the associated threat to the satisfaction of basic human needs and goals (Sandler, 2001; Weems & Overstreet, 2008). This study’s contributions to the existing literature must be considered in light of its limitations. First, the study is limited by the fact that youth completed self-report measures to assess symptoms. It is possible that other sources such as parents or teachers would have provided different reports of youths’ levels of psychological distress. Future research that incorporates information from multiple sources (e.g., youth, parents, and teachers) could be an informative and important next step. Second, the measure of social support used in this study has not been validated and the reliability coefficient of the peer support subscale may be considered low, and so this study’s findings should be considered in terms of these limitations. Additionally, the cross-sectional nature of the data in

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

SOCIAL SUPPORT

Study 1 does not allow for causal inferences to be made. This limitation should be noted for Study 2 as well. While it is possible to infer predictive associations for the prediction of Time 2 psychological distress, one cannot conclude that any of the variables caused these outcomes and linkages may be because of preexisting conditions not assessed (Weems, Scott, Banks, & Graham, 2012). Finally, the large sample size of Study 1 leads to the possibility that some of the statistically significant correlations obtained were relatively small and would not have been significant utilizing a smaller sample size. Despite these limitations, the present study makes a number of contributions to the literature by furthering our understanding of posthurricane social support in relation to psychological distress outcomes. Given the potential for negative mental health outcomes such as symptoms of PTSD, anxiety, and depression following natural disasters such as a hurricane, efforts to enhance youths’ social support to the extent possible through mechanisms such as interventions may be especially important. Keywords: social support; disasters; PTSD symptoms; children; anxiety; depression

References American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author. Bernard, K., & Dozier, M. (2010). Examining infants’ cortisol responses to laboratory tasks among children varying in attachment disorganization: Stress reactivity or return to baseline? Developmental Psychology, 46, 1771–1778. doi:10.1037/a0020660 Brewin, C. R., Andrews, B., & Valentine, J. D. (2000). Meta-analysis of risk factors for posttraumatic stress disorder in trauma-exposed adults. Journal of Consulting and Clinical Psychology, 68, 748 –766. doi: 10.1037/0022-006X.68.5.748 Chorpita, B. F., Yim, L., Moffitt, C., Umemoto, L. A., & Francis, S. E. (2000). Assessment of symptoms of DSM–IV anxiety and depression in children: A revised child anxiety and depression scale. Behaviour Research and Therapy, 38, 835– 855. doi:10.1016/S0005-7967(99)00130-8 Cohen, S. (2004). Social relationships and health. American Psychologist, 59, 676 – 684. doi:10.1037/0003-066X.59.8.676 Cohen, S., Gottlieb, B. H., & Underwood, L. G. (2000). Social relationships and health. In S. Cohen, L. G. Underwood, & B. H. Gottlieb (Eds.), Social support measurement and intervention: A guide for health and social scientists (pp. 3–25). New York, NY: Oxford University Press. Cohen, S., & Wills, T. A. (1985). Stress, social support, and the buffering hypothesis. Psychological Bulletin, 98, 310 –357. doi:10.1037/00332909.98.2.310 Compas, B. E., & Epping, J. E. (1993). Stress and coping in children and families: Implications for children coping with disaster. In C. F. Saylor (Ed.), Children and disasters (pp. 11–28). New York, NY: Plenum Press. doi:10.1007/978-1-4757-4766-9_2 Dubow, E. F., & Ullman, D. G. (1989). Assessing social support in elementary school children: The Survey of Children’s Social Support. Journal of Clinical Child Psychology, 18, 52– 64. doi: 10.1207/s15374424jccp1801_7 Frederick, C. J., Pynoos, R. S., & Nadar, K. (1992). Reaction index to Psychic Trauma Form C (Child). Unpublished manuscript, University of California, Los Angeles. Garrison, C. Z., Bryant, E. S., Addy, C. L., Spurrier, P. G., Freedy, J. R., & Kilpatrick, D. G. (1995). Posttraumatic stress disorder in adolescents after Hurricane Andrew. Journal of the American Academy of Child &

351

Adolescent Psychiatry, 34, 1193–1201. doi:10.1097/00004583199509000-00017 Giannopoulou, I., Strouthos, M., Smith, P., Dikaiakou, A., Galanopoulou, V., & Yule, W. (2006). Post-traumatic stress reactions of children and adolescents exposed to the Athens 1999 earthquake. European Psychiatry, 21, 160 –166. doi:10.1016/j.eurpsy.2005.09.005 Goenjian, A. K., Molina, L., Steinberg, A. M., Fairbanks, L. A., Alvarez, M. L., Goenjian, H. A., & Pynoos, R. S. (2001). Posttraumatic stress and depressive reactions among Nicaraguan adolescents after Hurricane Mitch. The American Journal of Psychiatry, 158, 788 –794. doi:10.1176/ appi.ajp.158.5.788 Greenberg, M. T., Siegel, J. M., & Leitch, C. J. (1983). The nature and importance of attachment relationships to parents and peers during adolescence. Journal of Youth and Adolescence, 12, 373–386. doi: 10.1007/BF02088721 Hardin, S. B., Weinrich, M., Weinrich, S., Hardin, T. L., & Garrison, C. (1994). Psychological distress of adolescents exposed to Hurricane Hugo. Journal of Traumatic Stress, 7, 427– 440. doi:10.1002/jts .2490070308 Hensley, L., & Varela, R. E. (2008). PTSD symptoms and somatic complaints following Hurricane Katrina: The roles of trait anxiety and anxiety sensitivity. Journal of Clinical Child and Adolescent Psychology, 37, 542–552. doi:10.1080/15374410802148186 Holmbeck, G. N. (2002). Post-hoc probing of significant moderational and meditational effects in studies of pediatric populations. Journal of Pediatric Psychology, 27, 87–96. doi:10.1093/jpepsy/27.1.87 Jaycox, L. H., Cohen, J. A., Mannarino, A. P., Walker, D. W., Langley, A. K., Gegenheimer, K. L., . . . Schonlau, M. (2010). Children’s mental health care following Hurricane Katrina: A field trial of trauma-focused psychotherapies. Journal of Traumatic Stress, 23, 223–231. Johnson, J. H., & McCutcheon, S. M. (1980). Assessing life stress in older children and adolescents: Preliminary findings with the Life Events Checklist. In I. Sarason & C. Spielberger (Eds.), Stress and anxiety (pp. 111–125). Washington, DC: Hemisphere. Kronenberg, M. E., Hansel, T. C., Brennan, A. M., Osofsky, H. J., Osofsky, J. D., & Lawrason, B. (2010). Children of Katrina: Lessons learned about postdisaster symptoms and recovery patterns. Child Development, 81, 1241–1259. doi:10.1111/j.1467-8624.2010.01465.x La Greca, A. M., & Silverman, W. K. (2009). Treatment and prevention of posttraumatic stress reactions in children and adolescents exposed to disasters and terrorism: What is the evidence? Child Development Perspectives, 3, 4 –10. doi:10.1111/j.1750-8606.2008.00069.x La Greca, A. M., Silverman, W. K., Lai, B., & Jaccard, J. (2010). Hurricane-related exposure experiences and stressors, other life events, and social support: Concurrent and prospective impact on children’s persistent posttraumatic stress symptoms. Journal of Consulting and Clinical Psychology, 78, 794 – 805. doi:10.1037/a0020775 La Greca, A. M., Silverman, W. K., Vernberg, E. M., & Prinstein, M. J. (1996). Symptoms of posttraumatic stress in children after Hurricane Andrew: A prospective study. Journal of Consulting and Clinical Psychology, 64, 712–723. doi:10.1037/0022-006X.64.4.712 La Greca, A. M., Silverman, W. K., & Wasserstein, S. B. (1998). Children’s predisaster functioning as a predictor of posttraumatic stress following Hurricane Andrew. Journal of Consulting and Clinical Psychology, 66, 883– 892. doi:10.1037/0022-006X.66.6.883 Lepore, S. J., Silver, R. C., Wortman, C. B., & Wayment, H. A. (1996). Social constraints, intrusive thoughts, and depressive symptoms among bereaved mothers. Journal of Personality and Social Psychology, 70, 271–282. doi:10.1037/0022-3514.70.2.271 Lonigan, C. J., Shannon, M. P., Taylor, C. M., Finch, A. J., & Sallee, F. R. (1994). Children exposed to disaster: II. Risk factors for the development of post-traumatic symptomatology. Journal of the American Academy of Child & Adolescent Psychiatry, 33, 94 –105. doi:10.1097/ 00004583-199401000-00013

This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.

352

BANKS AND WEEMS

McDermott, B. M., & Palmer, L. J. (2002). Postdisaster emotional distress, depression and event-related variables: Findings across child and adolescent developmental stages. Australian and New Zealand Journal of Psychiatry, 36, 754 –761. doi:10.1046/j.1440-1614.2002.01090.x Meaney, M. J. (2001). Maternal care, gene expression, and the transmission of individual differences in stress reactivity across generations. Annual Review of Neuroscience, 24, 1161–1192. doi:10.1146/annurev .neuro.24.1.1161 Mohay, H. A., & Forbes, N. (2009). Reducing the risk of posttraumatic stress disorder in children following natural disasters. Australian Journal of Guidance and Counseling, 19, 179 –195. doi:10.1375/ajgc.19.2 .179 Moore, K. W., & Varela, R. E. (2010). Correlates of long-term posttraumatic stress symptoms in children following Hurricane Andrew. Child Psychiatry and Human Development, 41, 239 –250. doi:10.1007/ s10578-009-0165-6 Osofsky, H. J., Osofsky, J. D., Kronenberg, M., Brennan, A., & Hansel, T. C. (2009). Posttraumatic stress symptoms in children after Hurricane Katrina: Predicting the need for mental health services. American Journal of Orthopsychiatry, 79, 212–220. doi:10.1037/a0016179 Ozer, E. J., Best, S. R., Lipsey, T. L., & Weiss, D. S. (2003). Predictors of posttraumatic stress disorder and symptoms in adults: A meta-analysis. Psychological Bulletin, 129, 52–73. doi:10.1037/0033-2909.129.1.52 Parker, K. J., Buckmaster, C. L., Sundlass, K., Schatzberg, A. F., & Lyons, D. M. (2006). Maternal mediation, stress inoculation, and the development of neuroendocrine stress resistance in primates. Proceedings of the National Academy of Sciences of the United States of America, 103, 3000 –3005. doi:10.1073/pnas.0506571103 Pina, A. A., Villalta, I. K., Ortiz, C. D., Gottschall, A. C., Costa, N. M., & Weems, C. F. (2008). Social support, discrimination, and coping as predictors of posttraumatic stress reactions in youth survivors of Hurricane Katrina. Journal of Clinical Child and Adolescent Psychology, 37, 564 –574. doi:10.1080/15374410802148228 Prinstein, M. J., La Greca, A. M., Vernberg, E. M., & Silverman, W. K. (1996). Children’s coping assistance after a natural disaster. Journal of Clinical Child Psychology, 25, 463– 475. doi:10.1207/ s15374424jccp2504_11 Pynoos, R. S., & Nader, K. (1988). Psychological first aid and treatment approach to children exposed to community violence: Research implications. Journal of Traumatic Stress, 1, 445– 473. doi:10.1002/jts .2490010406 Sandler, I. (2001). Quality and ecology of adversity as common mechanisms of risk and resilience. American Journal of Community Psychology, 29, 19 – 61. doi:10.1023/A:1005237110505 Scheeringa, M. S., & Zeanah, C. H. (2008). Reconsideration of harm’s way: Onsets and comorbidity patterns of disorders in preschool children and their caregivers following Hurricane Katrina. Journal of Clinical Child & Adolescent Psychology, 37, 508–518. doi:10.1080/15374410802148178 Self-Brown, S., Lai, B. S., Thompson, J. E., McGill, T., & Kelley, M. L. (2013). Posttraumatic stress disorder symptom trajectories in Hurricane

Katrina affected youth. Journal of Affective Disorders, 147, 198 –204. doi:10.1016/j.jad.2012.11.002 Spence, S. H. (1997). Structure of anxiety symptoms among children: A confirmatory factor-analytic study. Journal of Abnormal Psychology, 106, 280 –297. doi:10.1037/0021-843X.106.2.280 Steinberg, A. M., Brymer, M. J., Decker, K. B., & Pynoos, R. S. (2004). The University of California at Los Angeles Post-Traumatic Stress Disorder Reaction Index. Current Psychiatry Reports, 6, 96 –100. doi: 10.1007/s11920-004-0048-2 Thoits, P. A. (1986). Social support as coping assistance. Journal of Consulting and Clinical Psychology, 54, 416 – 423. doi:10.1037/0022006X.54.4.416 Vernberg, E. M., La Greca, A. M., Silverman, W. K., & Prinstein, M. J. (1996). Prediction of posttraumatic stress symptoms in children after Hurricane Andrew. Journal of Abnormal Psychology, 105, 237–248. doi:10.1037/0021-843X.105.2.237 Vernberg, E. M., & Vogel, J. (1993). Interventions with children after disasters. Journal of Clinical Child Psychology, 22, 485– 498. doi: 10.1207/s15374424jccp2204_8 Weems, C. F., & Overstreet, S. (2008). Child and adolescent mental health research in the context of Hurricane Katrina: An ecological-needs-based perspective and introduction to the special section. Journal of Clinical Child and Adolescent Psychology, 37, 487– 494. doi:10.1080/ 15374410802148251 Weems, C. F., Pina, A. A., Costa, N. M., Watts, S. E., Taylor, L. K., & Cannon, M. F. (2007). Pre-disaster trait anxiety and negative affect predict posttraumatic stress in youth after Hurricane Katrina. Journal of Consulting and Clinical Psychology, 75, 154 –159. doi:10.1037/0022006X.75.1.154 Weems, C. F., Scott, B. G., Banks, D. M., & Graham, R. A. (2012). Is TV traumatic for all youths? The role of pre-existing posttraumatic stress symptoms on the link between disaster coverage and stress. Psychological Science, 23, 1293–1297. doi:10.1177/0956797612446952 Weems, C. F., Scott, B. G., Taylor, L. K., Cannon, M. F., Romano, D. M., & Perry, A. M. (2013). A theoretical model of continuity in anxiety and links to academic achievement in disaster exposed school children. Development and Psychopathology, 25, 729 –737. doi:10.1017/ S0954579413000138 Weems, C. F., Taylor, L. K., Cannon, M. F., Marino, R., Romano, D. M., Scott, B. G., . . . Triplett, V. (2010). Posttraumatic stress, context, and the lingering effects of the Hurricane Katrina disaster among ethnic minority youth. Journal of Abnormal Child Psychology, 38, 49 –56. Weems, C. F., Watts, S. E., Marsee, M. A., Taylor, L. K., Costa, N. M., Cannon, M. F., . . . Pina, A. A. (2007). The psychosocial impact of Hurricane Katrina: Contextual differences in psychological symptoms, social support, and discrimination. Behaviour Research and Therapy, 45, 2295–2306. doi:10.1016/j.brat.2007.04.013

Family and peer social support and their links to psychological distress among hurricane-exposed minority youth.

Experiencing a disaster such as a hurricane places youth at a heightened risk for psychological distress such as symptoms of posttraumatic stress diso...
203KB Sizes 0 Downloads 0 Views