ethnic groups: Findings from the National Comorbidity Survey-Adolescent (NCS-A)

Accepted Manuscript Title: Differential Associations between Social Anxiety Disorder, Family Cohesion, and Suicidality Across Racial/Ethnic Groups: Findings from the National Comorbidity Survey-Adolescent (NCS-A) Author: Amy M. Rapp M.A. Anna Lau Ph.D. Denise A. Chavira Ph.D. PII: DOI: Reference:

S0887-6185(16)30128-1 http://dx.doi.org/doi:10.1016/j.janxdis.2016.09.009 ANXDIS 1880

To appear in:

Journal of Anxiety Disorders

Received date: Revised date: Accepted date:

1-7-2016 15-9-2016 16-9-2016

Please cite this article as: Rapp, Amy M., Lau, Anna., & Chavira, Denise A., Differential Associations between Social Anxiety Disorder, Family Cohesion, and Suicidality Across Racial/Ethnic Groups: Findings from the National Comorbidity Survey-Adolescent (NCS-A).Journal of Anxiety Disorders http://dx.doi.org/10.1016/j.janxdis.2016.09.009 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

RUNNING HEAD: Social Anxiety Disorder and Suicidality

1

Abstract Word Count: 200 Manuscript Word Count (excluding abstract, acknowledgements, and disclosures): 5,275 Table Count: 4 Figure Count: 3

Differential Associations between Social Anxiety Disorder, Family Cohesion, and Suicidality Across Racial/Ethnic Groups: Findings from the National Comorbidity Survey-Adolescent (NCS-A)

Amy M. Rapp, M.A.1, Anna Lau, Ph.D.1, Denise A. Chavira, Ph.D.1

1 Department of Psychology, University of California Los Angeles

Corresponding Author: Amy M. Rapp, M.A., Department of Psychology, University of California, Los Angeles, 2244B Franz Hall, Los Angeles, CA 90095. Email: [email protected].

Social Anxiety Disorder and Suicidality

2

Abstract The proposed research seeks to introduce a novel model relating Social Anxiety Disorder (SAD) and suicide outcomes (i.e., passive suicidal ideation, active suicidal ideation, and suicide attempts) in diverse adolescents. This model posits that family cohesion is one pathway by which suicide risk is increased for socially anxious youth, and predicts that the relationships between these variables may be of different strength in Latino and White subgroups and across gender. Data from a sample of Latino (n=1922) and nonHispanic White (hereafter referred to as White throughout) (n=5648) male and female adolescents who participated in the National Comorbidity Survey-Adolescent were used for this study. Analyses were conducted using generalized structural equation modeling. Results showed that the mediation model held for White females. Further examination of direct pathways highlighted SAD as a risk factor unique to Latinos for active suicidal ideation and suicide attempt, over and above comorbid depression and other relevant contextual factors. Additionally, family cohesion showed a strong association with suicide outcomes across groups, with some inconsistent findings for White males. Overall, it appears that the mechanism by which SAD increases risk for suicidality is different across groups, indicating further need to identify relevant mediators, especially for racial/ethnic minority youth.

KeyWords: social anxiety disorder; suicide; adolescent; Latino mental health; disparities

Social Anxiety Disorder and Suicidality

3

The majority of youth in community settings who contemplate or attempt suicide have a psychiatric disorder (approximately 70-90%)1,2. While most studies of suicidality in adolescents emphasize the importance of depression as a risk factor, anxiety disorders also show a significant and consistent association with suicide outcomes3-10. Social Anxiety Disorder (SAD; also referred to in the literature as Social Phobia) in particular, has received little attention even though there is a 19.9% prevalence of SAD among youth with a history of suicidal ideation and behaviors compared to a 7.0% prevalence rate of SAD among those without11. Some efforts have been made to identify mediators of the association of social anxiety and suicidality, such as loneliness12, however such investigations are rare and further research is required to fully understand why risk for suicide is increased in socially anxious teens. An important point of discussion for research investigating SAD as a risk factor for suicidality is the considerable comorbidity of SAD with Major Depressive Disorder (MDD). MDD is considered one of the most well-supported risk factors for suicide in adolescents13, and comorbid depression and anxiety confers the highest risk for suicidality14. Yet, in the context of comorbidity there is mixed evidence concerning the independent contribution of social anxiety to the development of suicide-related thoughts and behavior. For example, some studies from adult epidemiological samples have shown that SAD did not increase suicide risk above and beyond a diagnosis of MDD5,10. However, there is a small body of literature suggesting that social anxiety plays an independent role in increasing suicide risk after controlling for other psychopathology in both adolescent and adult samples9,12,15-19.

Social Anxiety Disorder and Suicidality

4

Importantly, there are racial/ethnic disparities in rates and severity of both anxiety symptoms and suicidality. Among teens, Latino adolescents self-report the highest severity of social anxiety compared to other racial/ethnic groups20. Findings in adult samples reveal that Latino adults report greater impairment due to social anxiety compared to Whites, with some evidence to suggest that risk for social anxiety may be influenced by cultural factors specific to Latinos such as language use21. Further, Latino female adolescents in particular have been shown to be at significantly higher risk for suicidal thoughts and behaviors. Results from the 2015 Youth Risk Behavior Surveillance convey that Latino female adolescents report higher rates of suicidal ideation (25.6%), planning (20.7%), and attempts (15.1%) than other racial and ethnic groups (e.g., 22.8% and 18.7% of African American and White females reported suicidal ideation, respectively; 18.4% and 17.3% of African American and White females reported planning, respectively; and 9.8% and 10.2% of African American and White females reported an attempt, respectively)22. Although these racial/ethnic disparities have been identified, investigation is needed to examine the cultural and contextual factors associated with the expression of SAD and possible links to suicidality among Latinos. There are likely numerous mechanisms by which suicidality risk is increased among those with social anxiety. One possible pathway is dysfunction in interpersonal relationships. Interpersonal models of suicidality suggest that the construct of thwarted belongingness, which refers to a sense of isolation, loneliness, and perceived lack of mutually supportive relationships, contributes to one’s risk for suicide23. This same construct has also been uniquely linked to social anxiety24, with consistent associations observed between social anxiety symptoms and perceptions of low social support in close

Social Anxiety Disorder and Suicidality

5

relationships25,26. Overall, the literature supports impairment in close relationships for individuals with SAD. First, there may be impairment related to clinical characteristics of the disorder such as avoidance and oversensitivity to criticism27. These behaviors, in addition to negative cognitive schemata that increase fear of threat and rejection in social situations, result in dysfunctional interpersonal styles28 and related sequelae such as social isolation, exclusion, and a lack of connectedness. Although findings support that SAD primarily causes interference in domains such as friendships, school, and work, there are also data to suggest that socially anxious individuals experience dysfunction attributable to SAD in their relationships with family members29. Socially anxious individuals exhibit conflict avoidance, avoidance of emotional expression, increased fear of rejection, and less assertion in intimate relationships often resulting in familial relationships characterized by interpersonal chronic stress29. Dysfunction in familial relationships, specifically conflict and poor communication, has been shown to be associated with suicide outcomes, and often, adolescent suicide attempts are precipitated by conflicts with parents or family members30-32. Conversely, connectedness with family has been found to be protective against suicide attempts33 and several family-based interventions for youth with suicidality have targeted family relationships with favorable effects34,35. Given the importance of culture in shaping family processes, it is likely that such mechanisms may operate differently across racial/ethnic groups. It is also plausible that family and other culture related variables may differentially impact risk for suicidality among anxious ethnic minority youth. As an example, family is given great importance in Latino culture and strain in family relations may be a particularly risky condition for Latino youth. Research has identified

Social Anxiety Disorder and Suicidality

6

high levels of familismo (attitudinal familism), defined as emphasis on family centrality, loyalty, and cohesion, as a distinctive cultural dimension of Latino families36-38. Evidence supports the protective role of perceived familial social support for Latinos for psychiatric disorders (e.g., depression) and general psychological distress, particularly in the presence of risk factors such as acculturative stress39-41. Family processes have more specifically been implicated in explaining the remarkably high rates of suicidality among Latino female adolescents. Findings show that Latino female adolescents who perceived low levels of parental caring were at 2.5-5 times greater risk of experiencing suicidal ideation and 3.5-10.0 times greater risk of attempting suicide compared to Latino females who reported high levels of perceived parental caring42. Still, the heightened suicide risk of Latino females remains relatively understudied, particularly from an empirical standpoint, and further testing of relevant familial processes is warranted. In light of evidence suggesting differential risk for suicidality for Latino females and the salience of family in this culture, it is possible that disruption in family cohesion may produce greater deleterious effects on psychological adjustment among Latino female adolescents. Given this evidence, we posit that SAD may confer risk for suicidal thoughts and behaviors to the extent that symptoms disrupt family cohesion. We expect that this risk pathway may be particularly strong for Latino youth, given the centrality of family relations in this cultural group. As such, we aim to test race/ethnicity (i.e., Latino vs. nonLatino White status) as a potential moderator in order to better understand factors that may contribute to disparities in rates of adolescent suicidality. Further, we examine whether this risk pathway is equally salient across adolescent males and females. While we predict that the proposed model will hold across Latino and White subgroups of male

Social Anxiety Disorder and Suicidality

7

and female adolescents, we believe that the associations between SAD, family cohesion, and suicide outcomes will be stronger for Latino female adolescents, a group who exhibits particularly high risk for suicide. Findings will provide data to inform meaningful and modifiable intervention targets for socially anxious youth with suicidal thoughts and behaviors. These data will also clarify whether there are unique risk factors for Latino female teens that need to be addressed in existing evidence based interventions. Data from a subsample of adolescents comprised of Latino and White females and males who participated in the National Comorbidity Survey Adolescent (NCS-A) was used to examine the following hypotheses: 1. Within each racial/ethnic and gender subgroup (i.e., Latino females, Latino males, White females, White males), family cohesion mediates the relationship between SAD and suicidal thoughts and behaviors (i.e., passive suicidal ideation, active suicidal ideation, and suicide attempt), controlling for factors known in the literature to be related to variables in the model (i.e., age, parent-reported total family income, and Diagnostic Statistical Manual-IV (DSM-IV) lifetime MDD diagnosis31,43,44). 2. Latino ethnicity and female gender will moderate the associations between SAD and family cohesion and between family cohesion and suicide outcomes, such that Latino female adolescents will show the strongest indirect effect from SAD to suicidality via family cohesion. Methods

Social Anxiety Disorder and Suicidality

8

Data from a subsample of adolescents who participated in the NCS-A were used for this study (n = 7543). The NCS-A was designed to assess prevalence, correlates, and patterns of service use for a nationally representative sample of American youth with psychiatric disorders45,46. The study was designed to estimate lifetime-to-date and current prevalence, age of onset distributions, course, and comorbidity of DSM-IV disorders in children and adolescents ages 13-17 (n = 10,148). Details of the sampling frame and weighting methods have been previously reported45,46.

Procedure Adolescent participants were interviewed in person by professional interview staff using a modified version of the World Health Organization (WHO) Composite International Diagnostic Interview Version 3.0 (CIDI)47. The interview was computerassisted in order to minimize administration error. Full details regarding collection of data has been reported elsewhere46. Participants The present study includes data from a subsample of adolescents (n = 7543) included in the NCS-A. The subsample was 50.8% female and included adolescents from the ages of 13-18. The subsample was predominately comprised of White adolescents (74.5%), with the remainder of the subsample (25.5%) being Latino. Latino females showed the greatest percentage of DSM-IV SAD (17.86%) and MDD (17.75%) diagnoses, as well as the highest degree of passive suicidal ideation (15.44%), active suicidal ideation (5.25%), and suicide attempts (5.67%). Of Latino females who meet

Social Anxiety Disorder and Suicidality

9

criteria for SAD, 33.14% had a comorbid MDD diagnosis. SAD and MDD comorbidity was greatest in this subsample followed by White males (31.97%), Latino males (28.75%), and White females (27.04%). Full subsample descriptives are reported in Table 1. Measures Demographics. The NCS-A 2001-2004 Data Collection Instrument included modules that assessed demographics. Participants were asked open-endedly if they identified as Hispanic/Latino, what they considered to be their ethnic origin, and race category. From this information, a variable was derived categorizing participants as “Hispanic,” “Black,” “Other,” and “White.” For the present study, only participants who were categorized as Hispanic or White were included in analyses. Information regarding age was also collected. Income. Family finances were assessed using a module from the NCS-A 20012004 Parent Self-Administered Questionnaire (PSAQ). Parents were asked to report estimated income from various sources including self and spouse earnings from employment, Social Security, and government assistance programs. One variable was derived from parent responses to represent total family income. Diagnostic variables. A modified version of the WHO CIDI used in the NCS-R was administered in the NCS-A47. Four broad DSM-IV disorder categories were assessed: anxiety disorders, mood disorders, substance disorders, and behavior disorders. Additional modules assessing neurasthenia and eating disorders were added, as well as assessment

of

emotional

functioning,

including

suicidality

and

non-specific

psychological distress. The present study utilized the SAD and MDD diagnostic modules.

Social Anxiety Disorder and Suicidality

10

The average interview length was two and a half hours with a range from 69-347 minutes. Suicide outcomes. Outcome variables were derived from the WHO CIDI. Participants were asked to indicate if in their lifetime they seriously thought about killing themselves (operationalized as passive suicidal ideation), made a plan for killing themselves (operationalized as active suicidal ideation), or tried to kill themselves (operationalized as a suicide attempt). All outcome variables were dichotomously coded yes/no. Family cohesion. The WHO CIDI Family Environment module contains 21 items that assess aspects of family structure, parental psychopathology, as well as parenting style47. Items were adapted from measures used frequently in the literature including the Family History Research Diagnostic Criteria Interview48,49 and the Parental Bonding Instrument50. For the present study, we sought to extract a scale that tapped the construct of family cohesion. To accomplish this, exploratory factor analysis (EFA) using principal-components factoring with orthogonal varimax rotation was conducted. The number of underlying factors was determined based on the eigenvalues of the rotated matrix (loadings > 0.40). EFA revealed four unique factors. The first factor appeared to map on most closely to the construct of family cohesion. This factor was comprised of ten items with value loadings ranging from 0.408-0.705. Items contained in this factor asked participants to respond to questions such as “How often did everyone compromise when there were disagreements?”, “ How often could family members talk to each other about their feelings?”, “ How often did family members share interests and hobbies with each other?”, and “ How often did family members feel very close to each other?”

Social Anxiety Disorder and Suicidality

11

Variance explained by the solution for this factor was 48.61%. Internal consistency for the overall factor was good (Cronbach’s  = 0.84), as well as for both the White ( = 0.85) and Latino ( = 0.84) subsamples. The other three factors revealed by EFA seemed to represent variables related to having one’s needs provided for (e.g., food, shelter) (five items accounting for 10.56% of the variance), physical danger (e.g., physical abuse, being asked to do developmentally inappropriate chores/tasks) (three items accounting for 10.31% of the variance), and avoidance (four items accounting for 6.86% of the variance). Data Analyses Data analyses were conducted using Stata 14 software to allow for the application of population weights. The primary mediational analyses were run using generalized structural equation modeling in order to examine the moderating effect of both Latino vs. non-Latino White status and gender on the associations between SAD (i.e., DSM-IV lifetime Social Anxiety Disorder), family cohesion, and binary suicide outcome variables (i.e., lifetime passive suicidal ideation, active suicidal ideation, and suicide attempts), over and above demographic, contextual and clinical factors known in the literature to be related to suicidality outcomes (i.e., age, total family income, and DSM-IV lifetime MDD diagnosis31,43,44), which were used as covariates in all analyses. Only participants who were categorized as Hispanic/Latino or White were included in data analyses. Three moderated mediational models were tested; each model contained SAD as a predictor, family cohesion as a mediator, and either passive suicidal ideation, active suicidal ideation, or suicide attempt as the outcome. In each model, the following three-way interactions were tested in order to examine moderating effects: 1) SAD by Latino vs.

Social Anxiety Disorder and Suicidality

12

non-Latino White status by gender as a predictor of family cohesion, 2) SAD by Latino vs. non-Latino White status by gender as a predictor of the suicide outcome in the model, and 3) family cohesion by Latino vs. non-Latino White status by gender as a predictor of the suicide outcome in the model. Regardless of the statistical significance of the interaction, point estimates, significance levels, and 95% Confidence Intervals (calculated using a general, asymptotically equivalent method) for nonlinear combinations of parameter estimates were computed for all four subsamples (i.e., Latino males and females, White males and females) in line with the theoretical aims of this study. Calculations were based on the “delta method” which yields an approximate variance for a nonlinear function and is appropriate given the large sample size included in analyses51. Point estimates for binary predictors were exponentiated to produce odds ratios. All analyses were run using survey weights. Results Descriptive Statistics Table 1 presents the percentage of participants who met DSM-IV criteria for SAD and MDD, and who endorsed lifetime passive suicidal ideation, active suicidal ideation, and suicide attempt. Logistic regression was used to test for significant differences in frequencies across subgroups. Results show that White males were less likely to meet criteria for SAD and MDD than White females (p = .01, p < .001, respectively) and Latino females (p < .0001, p < .0001, respectively). Latino females were significantly more likely to meet criteria for SAD and MDD compared to White females (p = .009, p = .002, respectively) and Latino males (p < .0001, p < .0001, respectively). White males reported less passive suicidal ideation than both White females (p < .0001) and Latino females (p < .0001), as did Latino males (p < .0001 for comparisons with both White and

Social Anxiety Disorder and Suicidality

13

Latino females). The same pattern was observed for active suicidal ideation and suicide attempt (p < .0001 for all comparisons). Table 1 contains the means and standard deviations for all continuous study variables by ethnic group and gender. Results indicate that there were significant differences in youth-rated levels of lifetime family cohesion. Pairwise comparisons between Latino vs. White and gender groups were conducted using Tukey’s Honest Significant Difference (HSD) test. White males and females reported significant higher family cohesion than Latino females (HSD-test = 8.76, p < .05, HSD-test = 7.47, p < .05, respectively). There were also significant differences in parent-reported levels of family income. Overall, income for White males and females was significantly greater compared to Latino males and females (i.e., White males vs. Latino males: HSD-test= 7.73, p < .05; White males vs. Latino females: HSD-test= 8.39, p < .05; White females vs. Latino males: HSD-test= 9.35, p < .05; White females vs. Latino females: HSD-test= 10.01, p < .05). Model 1: Passive Suicidal Ideation Family cohesion appeared to mediate the association between SAD and passive suicidal ideation for only White females, as evidenced by a significant indirect effect (see Table 2). We next conducted post-hoc point estimates of beta weights to examine significant moderating effects of Latino vs. non-Latino White status and gender on the individual associations of SAD with passive suicidal ideation and family cohesion with passive suicidal ideation. The three-way interactions of SAD, Latino vs. non-Latino White status, and gender, as well as family cohesion, Latino vs. non-Latino White status, and gender were nonsignificant ( = -0.87, p = 0.16,  = 0.03, p = 0.53, respectively).

Social Anxiety Disorder and Suicidality

14

Analyses revealed that across all subsamples, the direct effect of SAD and passive suicidal ideation was non-significant (see Table 3), however, the direct effect of family cohesion and passive suicidal ideation was negative and significant across all groups (see Table 4). Model 2: Active Suicidal Ideation The mediational model for active suicidal ideation was significant for only White females, as evidenced by a significant indirect effect for this subgroup (see Table 2). Further analyses were conducted to test for differential associations of SAD with active suicidal ideation and family cohesion with active suicidal ideation across subgroups. The three-way interaction of SAD, Latino vs. non-Latino White status, and gender was significant ( = -2.27, p = .007), and post-hoc exploration of this interaction revealed that the direct effect of SAD and active suicidal ideation was only significant in the Latino subgroups (see Table 3). The interaction of family cohesion, Latino vs. non-Latino White status, and gender was non-significant, however differential patterns of association between family cohesion and active suicidal ideation were observed. Specifically, the direct effect of family cohesion and active suicidal ideation was significant across all groups except for White males (see Table 4). Model 3: Suicide Attempt The indirect effect in Model 3 was non-significant across all groups (see Table 2). The three-way interaction of SAD, Latino vs. non-Latino White status, and gender approached but did not meet statistical significance ( = -1.78, p = .058). Post-hoc exploration revealed that the direct effect of SAD and suicide attempt was only significant in the Latino subsamples (see Table 3). The interaction of family cohesion,

Social Anxiety Disorder and Suicidality

15

Latino vs. non-Latino White status, and gender was significant ( = 0.163, p = .041). Differential patterns in this association were discovered such that the direct effect of family cohesion with suicide attempt was significant in the Latino subgroups. The direct effect of family cohesion and suicide attempt was marginally significant (p = .05) in the White female subsample, however the 95% confidence interval included zero, indicating that this result should be interpreted with caution (see Table 4). Models 1-3: Association of SAD and family cohesion When predicting family cohesion, the three-way interaction of SAD, Latino vs. non-Latino White status, and gender was non-significant. However, post-hoc estimates revealed differential associations of SAD and family cohesion across Latino and White males and females, such that the direct effect was only significant for White females (see Table 3). Discussion An initial aim of this study was to test a model in which family cohesion mediates the association between SAD and suicide outcomes within subgroups of Latino and White males and females. We found support for this model specifically within the White female subgroup for the outcomes of active and passive suicidal ideation. These findings suggest that for White female adolescents, SAD may interfere with maintaining family cohesion, likely in a similar manner as described in previous research focused on impairment in intimate relationships due to social anxiety29. It is possible that social anxiety symptoms contribute to deficits in family cohesion, which may impact an adolescent’s sense of isolation and impoverished support from family. This likely

Social Anxiety Disorder and Suicidality

16

increases risk for suicidal ideation and marks a possibly modifiable target for suicide intervention among a certain segment of socially anxious youth. The direct pathway of the mediational model that was most variable was the association between SAD and lowered family cohesion, which was only significant for White females. Thus, on the whole, there was little evidence for concurrent links between social anxiety and poorer family relations outside of one subgroup that has generally been the focus of most studies of adolescent psychopathology. Among White families, the salience of family cohesion in shaping adjustment may differ. Adolescent females are thought to be more sensitive to family conflict52,53, and a large-scale study of adolescents revealed that family conflict precipitated early alcohol use for females but not for males54. Psychopathology in the context of impaired family relationships may be more likely for White females than for White males, perhaps due in part to gender-related socialization pressures that encourage greater investment in and valuation of family connectedness for females than for males53. Among Latino adolescents, various explanations for the lack of an association between SAD and family cohesion are plausible. First, the family construct that was tested in the present study is most consistent with a measure of family functioning that captures behavioral and emotional family cohesion. It is possible that there are other elements of family process more impacted by SAD among Latino youth. For example, family obligation and family assistance have shown to be associated with declines in adolescent risk taking and a higher degree of effortful cognitive control55,56. In the presence of increased anxiety, the protective effects of these family-related cultural values and behaviors might be attenuated, increasing risk for adverse mental health outcomes. A second interpretation may be that the finding is

Social Anxiety Disorder and Suicidality

17

evidence of resilience. In line with a body of research that supports the protective role of family against mental health problems for Latinos39-41, it is possible that family cohesiveness

remains

unaffected

among

Latino

families

even

when

youth

psychopathology is present. Lastly, research supports that some behaviors indicative of anxiety (e.g., shyness) are often considered normative (i.e., consistent with specific cultural scripts) in collectivistic cultures57. Thus, it may be so that in Latino families, where parental socialization goals are in line with certain cultural views such as primacy of group over self, deference, obedience, and respect58,59, social anxiety symptoms may not be interpreted to be problematic by Latino parents or may even treated favorably, leaving family cohesion intact. Although the mediational model was not supported across all groups, examination of direct paths provides insights into racial/ethnic anxiety-related mental health disparities. Findings show that when controlling for MDD, social anxiety appears to uniquely confer risk for active suicidal ideation and suicide attempt in Latino adolescents. First, this is notable considering evidence that Latinos show a disproportionate degree of symptom severity and impairment due to SAD compared to other racial/ethnic groups20,21. Secondly, it is also of importance that we were able to demonstrate this finding while controlling for MDD. A limitation of the field has been mixed support for SAD as a factor that independently increases risk for suicidality. In the present analyses, we found support for this independent role. Third, these results also held while controlling for family income, which we found in this sample to be significantly higher in White families compared to Latino families. Socioeconomic status (SES) has been found to contribute to a host of psychological disorders, with a particularly strong association

Social Anxiety Disorder and Suicidality

18

with anxiety disorders60. In the present study, we show that SAD increases suicide risk for Latinos even when SES is accounted for. Overall, it is possible that social anxiety may incrementally increase risk for suicide among minority youth, whereas for Caucasian youth, this risk may be increased in the presence of a comorbid disorder such as MDD, similar to findings from the Great Smoky Mountain Study, which showed the greatest risk for suicidality conferred by comorbid anxiety disorder and MDD14. Further, this risk remains elevated for Latinos regardless of socioeconomic adversity, a factor known to unilaterally drive mental health disparities. Collectively, the present findings demonstrate that SAD represents a potent risk factor for Latinos independent of other psychiatric and contextual variables. Next, the direct path between family cohesion and suicidality was significant for both Latinos and Whites, although in line with our original hypotheses, this association was stronger for Latinos. As predicted, lack of family cohesion is a meaningful variable that increases suicide risk for Latino teens, however, it does not appear that disruption of family cohesion is driven by symptoms of SAD. The impact of family cohesion may be amplified in a cultural context that prioritizes family relationships. This has been found in some studies in which attitudinal variables related to family (e.g., familismo) have been linked to increased rates of internalizing psychopatholgy61-63 and suicide attempts among Latino teens64. Discrepancies in endorsement of attitudinal cultural values between parents and children, as may sometimes occur in response to differing rates of acculturation, have also been found to disrupt family functioning by increasing levels of familial tension and conflict which may contribute to poorer mental health outcomes65,66. Present results support the notion that family environment is a critical variable that may

Social Anxiety Disorder and Suicidality

19

enhance or attenuate suicide risk, and may be particularly relevant in the context of other corresponding cultural views of family. Beyond expanding our empirical understanding of the etiology of serious psychopathology among diverse groups of adolescents, the present findings also have clinical applications. When treating socially anxious White females, clinicians should recognize increased risk for suicide in the presence of impaired family cohesion. When treating socially anxious Latino youth, however, clinicians may not need to be as vigilant to family environment, or at least to the behavioral and emotional elements of family that were captured by the family cohesion construct used in this study. Instead, as suggested by previous studies, it may be worthwhile to assess other aspects of family environment (e.g., discrepant parent-child attitudinal cultural values), as these may potentially inform a youth’s clinical presentation. That being said, practitioners should be aware of the unique risk for suicide conferred to socially anxious Latino youth. Suicidality, especially active suicidal ideation and behaviors, should be assessed and monitored from the onset of treatment for SAD. Further, in light of the consistent association of poor family cohesion with suicidality across almost all subgroups, the inclusion of strategies aimed at addressing issues within the family should be considered in suicide prevention and intervention programs. Given that the present study used a nationally representative, nonclinical community sample, these recommendations might be particularly relevant for school or outpatient settings. Limitations and Future Directions This study is limited by several methodological considerations. First, we used dichotomous variables to indicate presence of DSM-IV SAD. This lack of continuous

Social Anxiety Disorder and Suicidality

20

measurement does not allow us to examine subthreshold anxiety which research suggests may also increase risk for suicidal behaviors67. Further, while we are able to assume that participants experienced a certain degree of functional impairment and symptomatology in order to receive a SAD diagnosis, we were unable to determine if our results were driven by factors such as extent of role impairment or symptom severity. This is similarly true of our suicide outcomes which were measured dichotomously. For example, we were not able to control for symptom severity or examine degree of severity as a potential moderator. This study is also limited by its cross-sectional design. We were unable to determine temporal precedence of SAD and suicidality and thus, cannot make causal inferences. Future research should explore the relationship between SAD and suicidality using multiple time lags. Additionally, focus on discrete age ranges (e.g., early childhood vs. adolescence) would further elucidate the links between SAD and suicidality over time. An important future direction for the field is improved specificity in our measurement of family constructs. The present investigation used a measure of family cohesion that has not been validated in other studies. While this certainly may limit our findings, the scale that was constructed was derived using items from standardized measures and appears to have face validity as a measure of family cohesion. The scale was derived using EFA data analytic procedures and a large amount of variance was explained by the solution for the factor (48.61%). However, it is possible that the constructed scale does not definitively represent family cohesion and may require additional testing and refining. For example, this scale was derived using a sample that only included Latino and White adolescents. Given that family and parenting is known to

Social Anxiety Disorder and Suicidality

21

be influenced by culture, our derived measure may not generalize to other racial/ethnic groups such as African American and Asian adolescents. Thus, more nuanced study of family environment is needed. One recommendation for future research that aims to discern the ways in which cultural values shape the etiology and expression of psychopathology among diverse youth is the inclusion of measures of both behavioral (e.g., conflict; emotional expression) and attitudinal constructs (e.g., familismo; familial obligation) related to family. The identification of mediators of the direct path between SAD and suicidality among Latino adolescents remains a priority and could inform the development of novel interventions for this cultural group Characteristics related to the experience of being an ethnic minority youth may be relevant and might include constructs subsumed under the experience of acculturative stress such as perceived discrimination, intergenerational conflict, and adaptation. These acculturative stress processes are also dependent on a number of variables including level of acculturation (i.e., orientation to mainstream and origin cultures), geographic location, and ethnic identity. For example, Latino adolescents commonly report discriminatory treatment across settings68-70. Perceived general discrimination shows a unique link with social anxiety71, and emerging adults who report high levels of perceived discrimination have been shown to be at three times higher odds of a suicide attempt72. However, a Latino adolescent’s experience of perceived discrimination might be different if he/she lives in a community that is predominately Latino. In such a case, other cultural factors may need to be considered to elucidate how an acculturative stress related variable such as perceived discrimination increases risk for negative mental health outcomes among some racial/ethnic minority adolescents. Indeed,

Social Anxiety Disorder and Suicidality

22

acculturative stress is a multifaceted process that should be thoroughly unpacked in future research in order to understand culturally relevant mechanisms that contribute to internalizing disorders. Conclusions Despite these limitations, the current study contributes to the sparse literature examining the relationship between social anxiety and suicidality in a large, nationally representative sample and extends findings in this field by identifying the role of a potentially modifiable mechanism (i.e., family cohesion) associated with increased risk of suicidality. Despite promising findings that provide important data to inform potential modifications to treatment protocols for White female adolescents with anxiety, further efforts are required to understand the nature of suicidality within minority populations experiencing SAD, particularly among Latino females who demonstrate higher rates of ideation and attempts compared to other racial/ethnic groups of females. Additional research is needed to understand the impact of psychopathology on family factors within minority families, which may perhaps require more nuanced assessment of family constructs not fully represented by current measures. Tests of other culturally relevant mechanisms are also warranted and may inform the science that guides the development of culturally appropriate interventions designed for Latino youth, potentially reducing mental health disparities for an underserved and understudied population.

Social Anxiety Disorder and Suicidality

23

References 1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

Gould MS, King, R., Greenwald, S., Fisher, P., Schwab-Stone, M., Kramer, R., Flisher, A. J., Goodman, S., Canino, G., & Shaffer, D. Psychopathology associated with suicidal ideation and attempts among children and adolescents. Journal of the American Academy of Child and Adolescent Psychiatry. 1998;37:915-923. Fergusson DM, & Lynskey, M. T. Suicide attempts and suicidal ideation in a birth cohort of 16-year-old New Zealanders. Journal of the American Academy of Child and Adolescent Psychiatry. 1995;34:1308-1317. Cox BJ, Direnfeld, D. M., Swinson, R. P., & Norton, G. R. Suicidal ideation and suicide attempts in panic disorder and social phobia. The American Journal of Psychiatry. 1994;151:882-887. Ferrada-Noli M, Asberg, M., Ormstad, K., Lundin, T., & Sundbom, E. Suicidal behavior after severe trauma, part 1: PTSD diagnoses, psychiatric comorbidity and assessments of suicidal behavior. Journal of Traumatic Stress. 1998;11:103112. Kessler RC, Borges, G., & Walters, E. E. Prevalence of and risk factors for lifetime suicide attempts in the National Comorbidity Survey. Archives of General Psychiatry. 1999;56:617-626. Kotler M, Iancu, I., Efroni, R., & Amir, M. Anger, impulsivity, social support and suicide risk in patients with posttraumatic stress disorder. Journal of Nervous and Mental Disease. 2001;189:162-167. Bomyea J, Lang, A. J., Craske, M. G., Chavira, D., Sherbourne, C. D., Rose, R. D., Golinelli, D., Campbell-Sills, L., Welch, S. S.,Sullivan, G., Bystritsky, A., Roy-Byrne, & P., Stein, M. P. Suicidal ideation and risk factors in primary care patients with anxiety disorders. Psychiatry Research. 2013;209:60-65. O'Neil Rodriguez KA, & Kendall, P. C. Suicidal Ideation in Anxiety-Disordered Youth: Identifying Predictors of Risk. Journal of Clinical Child & Adolescent Psychology. 2014;43(1):51-62. Sareen J, Houlahan, T., Cox, B. J., & Asmundson, G. J. G. Anxiety Disorders Associated With Suicidal Ideation and Suicide Attempts in the National Comorbidity Survey. The Journal of Nervous and Mental Disease. 2005;193:450454. Strauss J, Birmaher, B., Bridge, J., Axelson, D., Chiappetta, L., Brent, D., & Ryan, N. Anxiety Disorders in Suicidal Youth. Canadian Journal of Psychiatry. 2000;45(8):739-744. Nock MK, Green, J. G., Hwang, I., McLaughlin, K. A., Sampson, N. A., Zaslavsky, A. M., & Kessler, R. C. Prevalence, Correlates, and Treatment of Lifetime Suicidal Behavior Among Adolescents Results From the National Comorbidity Survey Replication Adolescent Supplement. JAMA Psychiatry. 2013;70(3):300-310. Gallagher M, Prinstein, M. J., Simon, V., & Spirito, A. Social anxiety symptoms and suicidal ideation in a clinical sample of early adolescents: Examining loneliness and social support as longitudinal mediators. Journal of Abnormal Child Psychology. 2014;42(6):871-883.

Social Anxiety Disorder and Suicidality 13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23. 24.

25.

26. 27. 28.

24

Shaffer D, Gould, M. S., Fisher, P., Trautman, P., Moreau, D., Kleinman, M., & Flory, M. Psychiatric diagnosis in child and adolescent suicide. Archives of General Psychiatry. 1996;53:339–348. Foley DL, Goldston, D. B., Costello, E. J., & Angold, A. Proximal Psychiatric Risk Factors for Suicidality in Youth: The Great Smoky Mountain Study. Archives of General Psychiatry. 2006;63:1017-1024. Nelson EC, Grant, J. D., Bucholz, K. K., Glowinski, A., Madden, P. A. F., Reich, W., & Heath, A. C. Social phobia in a population-based female adolescent twin sample: co-morbidity and associated suicide-related symptoms. Psychological Medicine. 2000;30:797-804. Statham DJ, Heath, A. C., Madden, P. A. F., Bucholz, K. K., Bierut, L., Dinwiddie, S. H., Slutske, W. S., Dunne, M. P. & Martin, N. G. Suicidal behaviour: an epidemiological and genetic study. Psychological Medicine 1998;28(839-855). Nock MK, & Kessler, R. C. Prevalence of and risk factors for suicide attempts versus suicide gestures: analysis of the National Comorbidity Study. Journal of Abnormal Psychology. 2006;115:616-623. Valentiner DP, Gutierrez, P. M., & Blacker, D. Anxiety measures and their relationship to adolescent suicidal ideation and behavior. Anxiety Disorders. 2002;16:11-32. Nock MK, Hwang, I., Sampson, N. A., & Kessler, R. C. Mental disorders, comorbidity and suicidal behavior: results from the National Comorbidity Survey Replication. Molecular Psychiatry. 2010;18:868-876. McLaughlin KA, Hilt, L. M., & Nolen-Hoeksema S. Racial/Ethnic Differences in Internalizing and Externalizing Symptoms in Adolescents. Journal of Abnormal Child Psychology. 2007;35(5):801-816. Polo A, Alegria, M., Chen, C., & Blanco, C. The Prevalence, Comorbidity, and Age of Onset of Social Anxiety Disorder among U.S. Latinos. Jounral of Clinical Psychiatry. 2011;72(8):1096-1105. Kann L, McManus, T., Harris, W. A., Shanklin, S. L., Flint, K. H., Hawkins, J., et al. Youth Risk Behavior Surveillance- United States, 2015. Morbidity and Mortality Weekly Report Surveillance Summaries 2016. 2016;65(6):1-174. Joiner TE. Why people die by suicide. Cambridge, MA: Harvard University Press; 2005. Davidson CL, Wingate, L. R., Grant, D. M., Judah, M. R., & Mills, A. C. Interpersonal suicide risk and ideation: the influence of depression and social anxiety. Journal of Social and Clinical Psychology. 2011;30:842-855. Calsyn RJ, Winter, J. P., & Burger, G. K. The relationship between social anxiety and social support in adolescents: a test of competing causal models. Adolescence. 2005;40:103-113. LaGreca AM, & Lopez, N. Social anxiety among adolescents: linkages with peer relations and friendships. Journal of Abnormal Child Psychology. 1998;26:83-94. American Psychiatric Association. Diagnostic and statistical manual of mental disorders: DSM-5. Washington, D.C.: American Psychiatric Association; 2013. Alden LE, & Taylor, C. T. Interpersonal processes in social anxiety disorder. In: Beck JG, ed. Interpersonal processes in the anxiety disorders: Implications for

Social Anxiety Disorder and Suicidality

29. 30. 31.

32.

33. 34.

35.

36.

37. 38. 39.

40.

41.

42.

43. 44.

25

understanding psychopathology and treatment. Washington, DC: American Psychological Association; 2010. Davila J, & Beck, J.G. Is social anxiety associated with impairment in close relationships? A preliminary investigation. Behavior Therapy. 2002;33(427-336). Wagner B. Family and risk factors for child and adolescent suicidal behaviour. Psychological Bulletin. 1997;121(246-298). Brent DA, Greenhill, L. L., Compton, S., et al. The Treatment of Adolescent Suicide Attempters Study (TASA): predictors of suicidal events in an open treatment trial. Journal of the American Academy of Child and Adolescent Psychiatry. 2009;48:987-996. Gould MS, Greenberg, T., Velting, D. M., & Shaffer, D. Youth suicide risk and preventative interventions: a review of the past 10 years. Journal of the American Academy of Child and Adolescent Psychiatry. 2003;42(386-405). Qin P, & Mortensen, PB. The impact of parental status on the risk of completed suicide. Archives of General Psychiatry. 2003;60:797-802. Diamond GS, Wintersteen, M. B., Brown, G. K., et al. Attachment-based family therapy for adolescents with suicidal ideation: a randomized controlled trial. Journal of the American Academy of Child and Adolescent Psychiatry. 2010;49:122-131. Pineda J, & Dadds, M. R. Family Intervention for Adolescents With Suicidal Behavior: A Randomized Controlled Trial and Mediation Analysis. Journal of the American Academy of Child and Adolescent Psychiatry. 2013;52(8):851-862. Sabogal F, Marin, G., & Otero-Sabogal, R. Hispanic familism and acculturation: What changes and what doesn’t? Hispanic Journal of Behavioral Sciences. 1987;9:397-412. Zinn M. A new familism scale for use with Latino populations. Hispanic Journal of Behavioral Sciences. 2003;25:312-330. Steidel L, & Contreras, J. M. A new familism scale for use with Latino populations. Hispanic Journal of Behavioral Sciences. 2003;25:312-330. Rivera FI. Contextualizing the experience of young Latino adults: The Relationship between acculturation and depression. Journal of Immigrant and Minority Health. 2007;3:237-244. Vega WA, Kolody, B., Valle, R., & Weir, J. Social networks, social support, and their relationship to depression among immigrant Mexican women. Human Organization. 1991;50:154-162. Rivera FI, Guarnaccia, P. J., Mulvaney-Day, N., Lin J. Y., Torres, M., & Alegria, M. Family cohesion and its relationship to psychological distress among Latino groups. Hispanic Journal of Behavioral Sciences. 2008;30(3):357-378. Garcia C, Skay, C., Sieving, R., Naughton, S., & Bearinger, L. H. Family and racial factors associated with suicide and emotional distress among Latino students. Journal of School Health. 2008;78(9):487-495. Cash SJ, & Bridge, J. A. Epidemiology of youth suicide and suicidal behavior. Current Opinion in Pediatrics. 2009;21(5):613-619. Gunnell DJ, Peters, T. J., Kamerling, R. M., & Brooks, J. Relation between parasuicide, suicide, psychiatric admission, and socioeconomic deprivation. BMJ. 1995;311(6999):226-230.

Social Anxiety Disorder and Suicidality 45.

46.

47.

48.

49.

50. 51. 52.

53.

54.

55.

56.

57.

58.

26

Kessler RC, Avenevoli, S., Costello, E. J., Green, J. G., Gruber, M. J., Heeringa, S., Merikangas, K. R., Pennell, B. E., Sampson, N. A., & Zaslavsky, A. M. Design and field procedures in the US National Comorbidity Survey Replication Adolescent Supplement (NCS-A). International Journal of Methods in Psychiatric Research. 2009;18(2):69-83. Kessler RC, Avenevoli, S., Costello, E. J., Green, J. G., Gruber, M. J., Heeringa, S., Merikangas, K. R., Pennell, B. E., Sampson, N. A., & Zaslavsky, A. M. National Comorbidity Survey Replication Adolescent Supplement (NCS-A): II, Overview and design. American Academy of Child and Adolescent Psychiatry. 2009;48(4):380-385. Merikangas K, Avenevoli, S., Costello, J., Koretz, D., & Kessler, R. C. National Comorbidity Survey Replication Adolescent Supplement (NCS-A): I, background and measures. Journal of the American Academy of Child and Adolescent Psychiatry. 2009;48(4):367-369. Andreasen NC, Endicott, J., Spitzer, R. L., & Winokur, G. The family history method using diagnostic criteria. Reliability and validity. Archives of General Psychiatry. 1977;34:1229-1235. Kendler KS, Davis, C. G., & Kessler, R. C. The familial aggregation of common psychiatric and substance use disorders in the National Comorbidity Survey: a family history study. The British Journal of Psychiatry. 1997;170(541-548). Parker G. The Parental Bonding Instrument: psychometric properties reviewed. Psychiar Dev. 1989;7:317-335. Dorfman R. A Note on the δ-Method for Finding Variance Formulae. The Biometric Bulletin. 1938;1:129-137. Unger JB, Ritt-Olson, A., Teran, L., Huang, T., Hoffman, B. R., & Palmer, P. Cultural values and substance use in a multiethnic sample of California adolescents. Addiction Research and Theory. 2002;10:257-279. Davies PT, & Lindsay, L. L. Interparental conflict and adolescent adjustment: Why does gender moderate early adolescent vulnerability. Journal of Family Psychology. 2004;18(160-170). Kelly AB, Tournbourou, J. W., O’Flaherty, M., Patton, G. C., Homel, R., Connor, J. P., et al. Family relationship quality and early alcohol use: Evidence for genderspecific risk processes. Journal of Studies on Alcohol and Drugs. 2011;72(3):399407. Telzer EH, Masten, C. L., Berkman, E. T., Lieberman, M. D., & Fuligni, A. J. Gaining while giving: An fMRI study of rewards of family assistance. Social Neuroscience. 2010;5:508-515. Telzer EH, Fuligni, A. J., Lieberman, M. D., & Galvan, A. Meaningful family relationships: Neurocognitive buffers of adolescent risk taking. Journal of Cognitive Neuroscience. 2013;25(3):374-387. Gudiño OG, & Lau, A. S. Parental cultural orientation, shyness, and anxiety in Hispanic children: An exploratory study. Journal of Applied Developmental Psychology. 2010;31:202-210. Marín G, & Marín, B. V. Research with Hispanic populations. Thousand Oaks, CA: Sage; 1991.

Social Anxiety Disorder and Suicidality 59.

60.

61.

62.

63. 64.

65. 66.

67.

68. 69.

70.

71.

72.

27

Leyendecker B, Harwood, R. L., Lamb, M. E., & Schölmerich, A. Mothers' socialisation goals and evaluations of desirable and undesirable everyday situations in two diverse cultural groups. International Journal of Behavioral Development. 2002;26:248-258. Kessler RC, McGonagle, K. A., Zhao, S., Nelson, C. B., Hughes, M., Eshleman, S., et al. Life time and 12-month prevalence of DSM-III-R psychiatric disorders in the United States: Results from the National Comorbidity Survey. Archives of General Psychiatry. 1994;41:8-19. Smokowski PR, Rose, R. A., Evans, C. B. R., Cotter, K. L., Bower, M., & Bacallao, M. Familial influences on internalizing symptomatology in Latino adolescents: An ecological analysis of parent mental health and acculturation dynamics. Development and Psychopathology. 2014;26:1191-1207. Smokowski PR, & Bacallao, M. Acculturation, internalizing mental health symptoms, and self-esteem: Cultural experiences of Latino adolescents in North Carolina. Child Psychiatry and Human Development. 2007;37:273-292. Bacallao M, & Smokowski, P. R. The costs of getting ahead: Mexican family systems after immigration. Family Relations. 2007;52:52-66. Nolle AP, Gulbas, L., Kuhlberg, J. A., & Zayas, L. H. Sacrifice for the sake of the family: Expressions of familism by Latina teens in the context of suicide. American Journal of Orthopsychiatry. 2012;82(3):319-327. Falicov CJ. Mexican families. In: M. McGoldrick JG, & J. K. Pearce, ed. Ethnicity and Family Therapy. New York, NY: Guilford Press; 1996:169-182. Miranda AO, Estrada, D., & Firpo-Jimenez, M. Differences in family cohesion, adaptability, and environment among Latino families in dissimilar stages of acculturation. The Family Journal: Counseling and Therapy for Couples and Families. 2000;8:341-350. Marshall R, Olfson, M., Hellman, F., Blanco, C., Guardino, M., & Struening, E. Comorbidity, impairment, and suicidality in subthreshold PTSD. American Journal of Psychiatry. 2001;158(9):1467-1473. Edwards LM, & Romero, A. J. Coping withdiscrimination among Mexican descent adolescents. Hispanic Journal of Behavioral Sciences. 2008;30:24-39. Martinez CR, Jr., DeGarmo, D. S., & Eddy, J. M. Promoting academic success among Latino youths. Hispanic Journal of Behavioral Sciences. 2004;26:128151. Rosenbloom SR, & Way, N. Experiences of dis-crimination among African American, Asian American,and Latino adolescents in an urban high school. Youth and Society. 2004;35:420-451. Levine DS, Himle, J. A., Abelson, J. M., Matusko, N. M., Dhawan, N., & Taylor, R. J. Discrimination and Social Anxiety Disorder Among African-Americans, Caribbean Blacks, and Non-Hispanic Whites The Journal of Nervous and Mental Disease. 2014;202(3):224-230. Gomez J, Miranda, R., & Polanco, L. Acculturative Stress, Perceived Discrimination, and Vulnerability to Suicide Attempts Among Emerging Adults. Journal of Youth and Adolescence. 2011;40(11):1465-1476.

Social Anxiety Disorder and Suicidality

28

Social Anxiety Disorder and Suicidality

29

Figure 1. Path model relating SAD, family cohesion, and passive suicidal ideation for Latino and White adolescents.

-1.5 (-2.0) -.92 (.32)

Family Cohesion

-.15 (-.21) -.12 (-.16)

Passive Suicidal Ideation

Social Anxiety Disorder .71 (.69) .02 (-.88)

Note: Coefficient values in parentheses represent associations among males and those outside parentheses represent associations among females; italicized coefficients represent associations among White adolescents and non-italicized coefficients represent associations among Latino adolescents; significant associations are bolded.

Social Anxiety Disorder and Suicidality

30

Figure 2. Path model relating SAD, family cohesion, and active suicidal ideation for Latino and White adolescents.

-1.5 (-2.0)

Family Cohesion

-.18 (-.31) -.12 (-.15)

-.92 (.32)

Active Suicidal Ideation

Social Anxiety Disorder 2.0 (3.6) .15 (-.56)

Note: Coefficient values in parentheses represent associations among males and those outside parentheses represent associations among females; italicized coefficients represent associations among White adolescents and non-italicized coefficients represent associations among Latino adolescents; significant associations are bolded.

Social Anxiety Disorder and Suicidality

31

Figure 3. Path model relating SAD, family cohesion, and suicide attempt for Latino and White adolescents.

-1.5 (-2.0) -.92 (.32)

Social Anxiety Disorder

Family Cohesion

-.21 (-.37) -.08 (-.07)

Suicide Attempt 2.6 (4.1) .37 (.07)

Note: Coefficient values in parentheses represent associations among males and those outside parentheses represent associations among females; italicized coefficients represent associations among White adolescents and non-italicized coefficients represent associations among Latino adolescents; significant associations are bolded.

Social Anxiety Disorder and Suicidality

32

Tables Table 1. Descriptive Statistics Latino

White

Female Male Female Male N=952 N=970 N=2882 N=2766 Dichotomous variables (percentage endorsed) DSM-IV Lifetime 17.86 11.08 14.37 11.45 SAD DSM-IV Lifetime 17.75 7.64 13.60 6.75 MDD Lifetime Passive 15.44 6.97 14.99 8.2 Suicidal Ideation Lifetime Active 5.25 1.72 4.68 1.95 Suicidal Ideation Lifetime Suicide 5.67 1.81 4.58 1.44 Attempt Continuous variables (Mean (SD)) Age 15.21(1.57) 15.13 (1.57) 15.24(1.51) 15.13 (1.57) Imputed Parent 93290.60 95740.57 130864.15 124777.00 Reported Total (99501.27) (116710.23) (157751.22) (145624.98) Family Income Note: DSM-IV= Diagnostic Statistical Manual IV; SAD= Social Anxiety Disorder; MDD= Major Depressive Disorder; SD= standard deviation

Social Anxiety Disorder and Suicidality

33

Table 2. Tests of indirect effects (a*b paths) by racial/ethnic and gender group for each suicide outcome Coefficient p-value 95% CI 95% CI (upper (lower limit) limit) Outcome: Passive Suicidal Ideation White Males -.05 0.70 -.31 .21 White Females .11 0.01 .02 .20 Latino Males .43 0.37 -.53 1.4 Latino Females .22 0.18 -.11 .55 Outcome: Active Suicidal Ideation White Males -.05 0.70 -.30 .21 White Females .11 0.05 -.001 .21 Latino Males .64 0.35 -.72 2.0 Latino Females .28 0.20 -.14 .69 Outcome: Suicide Attempt White Males -.02 0.74 -.16 .12 White Females .07 0.10 -.02 .17 Latino Males .76 0.32 -.74 2.2 Latino Females .32 0.14 -.12 .76 Note: CI= confidence interval; significant associations are bolded

Social Anxiety Disorder and Suicidality

34

Table 3. Associations of SAD with mediator and outcome variables Coefficient

OR

p-value

95% CI 95% CI (lower limit) (upper limit) Association of SAD and family cohesion White Males

.32

--

0.70

-1.3

1.9

White Females

-.92

--

0.006

-1.6

-.26

Latino Males

-2.0

--

0.30

-6.0

1.9

Latino Females

-1.5 -0.12 -3.4 Association of SAD and passive suicidal ideation

.41

White Males

-.88

2.4

0.13

-2.0

.26

White Females

.02

.05

0.95

-.56

.59

Latino Males

.69

1.8

0.50

-1.3

2.7

Latino Females

.71 1.9 0.14 SAD and active suicidal ideation

-.24

1.7

White Males

-.56

1.5

0.57

-2.5

1.4

White Females

.15

.41

0.74

-.76

1.1

Latino Males

3.6

9.8

< 0.0001

1.8

5.4

Latino Females

2.0 5.43 < 0.0001 SAD and suicide attempt

1.1

2.9

White Males

.07

.19

0.94

-1.7

1.8

White Females

.37

1.0

0.38

-.47

1.2

Latino Males

4.1

11.1

0.003

1.4

6.8

Latino Females 2.6 7.1 <0.0001 1.3 3.9 Note: OR= odds ratio; CI= confidence interval; SAD= Social Anxiety Disorder; significant associations are bolded

Social Anxiety Disorder and Suicidality

35

Table 4. Association of mediator with outcome variables Coefficient

OR

p-value

95% CI 95% CI (lower limit) (upper limit) Association of family cohesion and passive suicidal ideation White Males

-.16

.43

0.005

-.27

-.05

White Females

-.12

.33

< 0.0001

-.07

Latino males

-.21

.57

0.004

-.17 -.36

Latino females

-.15 .41 0.001 -.23 -.06 Association of family cohesion and active suicidal ideation

White Males

-.15

.41

0.09

-.33

.02

White Females

-.12

.33

0.004

-.20

-.04

Latino males

-.31

.84

0.009

-.55

-.08

Latino females

-.18 .49 0.002 -.30 Association of family cohesion and suicide attempt

-.07

White Males

-.07

.19

0.42

-.25

.11

White Females

-.08

.22

0.05

-.16

.001

Latino males

-.37

1.0

0.001

-.60

-.15

-.07

Latino females -.21 .57 0.001 -.34 -.08 Note: OR= odds ratio; CI= confidence interval; significant associations are bolded