Introduction
According to the Centers for Disease Control and Prevention (CDC, 2011), there were 34 598 suicide deaths in the USA in 2007, 11.5 per 100 000 people, making suicide the 11th leading cause of death in the USA. This rate varies dramatically across ethnic groups: the suicide rate per 100 000 people was 14.4 among Non-Hispanic Whites, 6.2 among Asians, 5.1 among Non-Hispanic Blacks and 5.4 among Hispanics. Differences related to nativity (US-born versus foreign-born) (Choi et al. Reference Choi, Rogers and Werth2007; Duldulao et al. Reference Duldulao, Takeuchi and Hong2009; Singh & Hiatt, Reference Singh and Hiatt2006), adverse social status (Crosby et al. Reference Crosby, Ortega and Stevens2011) and genetic endowment (Galfalvy et al. Reference Galfalvy, Huang, Oquendo, Currier and Mann2009) have been examined as potential explanations for these differences, without conclusive results (Nock et al. Reference Nock, Borges, Bromet, Cha, Kessler and Lee2008; Perez-Rodriguez et al. Reference Perez-Rodriguez, Baca-Garcia, Oquendo and Blanco2008).
Studies that focus on suicidal ideation and suicide attempts can offer some insight into the demographic patterns of suicide. Suicide ideation increases the likelihood of suicide attempt and suicide death (Goldstein et al. Reference Goldstein, Black, Nasrallah and Winokur1991; Kessler et al. Reference Kessler, Borges and Walters1999; Borges et al. Reference Borges, Angst, Nock, Ruscio and Kessler2008; Nock et al. Reference Nock, Borges, Bromet, Cha, Kessler and Lee2008). Suicide attempts are strongly associated with risk for suicide death (Suominen et al. Reference Suominen, Isometsa, Suokas, Haukka, Achte and Lonnqvist2004) and repeated suicide attempts (Borges et al. Reference Borges, Angst, Nock, Ruscio and Kessler2008). Whether suicidality differs by these ethnic groups in the USA in the same way as observed for suicide death is a matter of current discussion (Perez-Rodriguez et al. Reference Perez-Rodriguez, Baca-Garcia, Oquendo and Blanco2008), with inconsistent results across studies (Roberts et al. Reference Roberts, Roberts and Xing2007; CDC, 2009). Reports from national surveys of adults in the USA have reported lower prevalence of attempts only among Blacks, relative to Non-Hispanic Whites (Kessler et al. Reference Kessler, Borges and Walters1999; Borges et al. Reference Borges, Angst, Nock, Ruscio, Walters and Kessler2006, Reference Borges, Angst, Nock, Ruscio and Kessler2008), but no studies have focused systematically on possible differences across the main ethnic groups in the USA using comparable assessments of suicidality and mental disorders (Moscicki et al. Reference Moscicki, O'Carroll, Rae, Locke, Roy and Regier1988; Oquendo et al. Reference Oquendo, Lizardi, Greenwald, Weissman and Mann2004; Perez-Rodriguez et al. Reference Perez-Rodriguez, Baca-Garcia, Oquendo and Blanco2008).
Whereas most studies focus on a range of psychological, cultural and adverse risk factors among minorities groups to explain alleged differences in suicidality (Perez-Rodriguez et al. Reference Perez-Rodriguez, Baca-Garcia, Oquendo and Blanco2008), here we examined three other potential explanations for ethnic variation in suicide ideation and attempt. First, psychiatric disorders, which are known risk factors for suicidal behavior (Kessler et al. Reference Kessler, Borges and Walters1999), are more prevalent among Non-Hispanic Whites in the USA relative to other ethnic groups (Breslau et al. Reference Breslau, Aguilar-Gaxiola, Kendler, Su, Williams and Kessler2006). The potential that pre-existing psychiatric disorders and types of disorders (Nock et al. Reference Nock, Hwang, Sampson, Kessler, Angermeyer, Beautrais, Borges, Bromet, Bruffaerts, de Girolamo, de Graaf, Florescu, Gureje, Haro, Hu, Huang, Karam, Kawakami, Kovess, Levinson, Posada-Villa, Sagar, Tomov, Viana and Williams2009; Borges et al. Reference Borges, Nock, Haro, Hwang, Sampson, Alonso, Andrade, Angermeyer, Beautrais, Bromet, Bruffaerts, Girolamo, Florescu, Gureje, Hu, Karam, Kovess, Lee, Levinson, Medina-Mora, Ormel, Posada-Villa, Sagar, Tomov, Uda, Williams and Kessler2010) might account for the reported ethnic differences in suicidal behaviors has not been ruled out. Our hypothesis is that the differential distribution of psychiatric disorders among the main ethnic groups in the USA may partially explain possible differences in suicidality among these groups.
Second, large portions of Hispanics and Asians in the USA are immigrants (about 66% and 75% respectively) (Grieco, Reference Grieco2010; Pew Hispanic Center, 2010). High rates for suicide death are visible for all US-born compared to foreign-born of the largest ethnic groups in the USA (Singh & Hiatt, Reference Singh and Hiatt2006). Part of the alleged differences in suicidality among ethnic groups might be caused by the large influx of new immigrants into the USA. There is evidence that, within these groups, immigrants have a much lower risk for psychiatric disorders and possibly suicide attempts (Breslau & Chang, Reference Breslau and Chang2006; Fortuna et al. Reference Fortuna, Perez, Canino, Sribney and Alegria2007; Borges et al. Reference Borges, Breslau, Su, Miller, Medina-Mora and Aguilar-Gaxiola2009; Breslau et al. Reference Breslau, Borges, Hagar, Tancredi and Gilman2009). Our hypothesis here is that the US-born population will show higher rates of suicidality compared to immigrants, and that among those US-born (second-generation), there are no ethnic differences in suicidality.
Third, ethnic comparisons of suicidal behavior in the USA may gain from a deeper analysis of possible differences between immigrants and US-born in suicidality, within each ethnic category, as has been suggested (Baca-Garcia et al. Reference Baca-Garcia, Perez-Rodriguez, Keyes, Oquendo, Hasin, Grant and Blanco2011). Our hypothesis here is that, within ethnic groups, the US-born would have higher prevalence of suicide ideation and attempts than foreign-born immigrants of the same ethnicity and that immigrants who arrived as children (at age ⩽12 years) would have higher risk than immigrants who arrived as adolescents or adults (at age ⩾13 years). Evidence for these hypotheses came from prior studies of immigration and mental disorders in the USA (Breslau et al. Reference Breslau, Aguilar-Gaxiola, Borges, Kendler, Su and Kessler2007, Reference Breslau, Borges, Hagar, Tancredi and Gilman2009).
The aim of this report was to compare the lifetime prevalence of suicide ideation and attempt among four main ethnic groups in the USA using data from a multi-ethnic national sample that assessed suicidality with the same survey instrument in five languages (English, Spanish, Chinese, Tagalog and Vietnamese). Analysis such as the one proposed here can have an impact on public health policies that delve into the health disparities in the treatment and prevention of suicide and suicide behaviors associated with race/ethnicity and migration patterns (Goldsmith et al. Reference Goldsmith, Pellmar and Kleinman2002).
Method
Data come from the National Institute of Mental Health (NIMH)-funded Collaborative Psychiatric Epidemiological Surveys (CPES), which includes the National Comorbidity Survey Replication (NCS-R), the National Latino and Asian-American Survey (NLAAS) and the National Study of American Life (NSAL) (Heeringa et al. Reference Heeringa, Wagner, Torres, Duan, Adams and Berglund2004). The CPES were conducted during the same time period, using the same diagnostic instrument, and field staff were certified by the Institute for Social Research at the University of Michigan (Kessler et al. Reference Kessler, Berglund, Chiu, Demler, Heeringa, Hiripi, Jin, Pennell, Walters, Zaslavsky and Zheng2004; Alegria et al. Reference Alegria, Mulvaney-Day, Torres, Polo, Cao and Canino2007; Takeuchi et al. Reference Takeuchi, Zane, Hong, Chae, Gong, Gee, Walton, Sue and Alegria2007; Williams et al. Reference Williams, Haile, Gonzalez, Neighbors, Baser and Jackson2007). The instrument used in these surveys, the World Mental Health version of the Composite International Diagnostic Interview (WMH-CIDI) was developed with NIMH support to assess DSM-IV mood, anxiety, substance use and impulse control disorders in addition to suicidality and a broad range of correlates and suspected risk factors. The CIDI is a fully structured computer-assisted diagnostic interview. After complete description of the study to the subjects, informed consent was obtained.
A total of 15 180 respondents were in one of the four ethnic groups in this study. Further analyses on 10 274 US-born subjects and 4853 immigrants, with complete information about nativity/migration status, were available for: 4179 Non-Hispanic Whites from the NCS-R; 3258 Hispanics from the NLAAS (2553), the NCS-R (527) and the NSAL (178); 2176 Asians from the NLAAS (2093) and the NCS-R (83); and 5514 Non-Hispanic Blacks from the NSAL (4799) and the NCS-R (715).
Measures of suicide-related outcomes
The WMH-CIDI contains a module that assesses history of suicide ideation (‘Have you ever seriously thought about committing suicide?’) and suicide attempts (‘Have you ever attempted suicide?’). These questions were printed in a booklet and referred to in the interview by letter, given that evidence suggests that participants' reports of such potentially sensitive behaviors are higher in self-administered than in interviewer-administered surveys (Turner et al. Reference Turner, Ku, Rogers, Lindberg, Pleck and Sonenstein1998). Respondents with a history of any suicide ideation or attempt were asked the age at which they first experienced that outcome. Timing of prior suicide ideation and attempt was obtained using a question sequence shown experimentally to improve recall accuracy (Knauper et al. Reference Knauper, Cannell, Schwarz, Bruce and Kessler1999).
Psychiatric disorders and other covariates
Psychiatric disorders were diagnosed according to DSM-IV criteria. The common diagnostic assessment in all surveys included mood disorders (major depressive disorder and dysthymia), anxiety disorders (panic disorder, agoraphobia without panic disorder, social phobia, generalized anxiety disorder and post-traumatic stress disorder) and substance use disorders (alcohol abuse, drug abuse, alcohol dependence with abuse, and drug dependence with abuse). All respondents were asked the age at which they had first experienced any of those disorders.
Demographics
Analyses also used sociodemographic information including gender, age, educational attainment, and age at first marriage and divorce (if any). Race/ethnicity was asked in all surveys and the original variables were recorded by CPES processors to harmonize data across the three studies (CPES, 2008). Twelve categories were identified: Vietnamese, Filipino, Chinese, All other Asian, Cuban, Puerto Rican, Mexican, All other Hispanic, Afro-Caribbean, African-American, Non-Latino Whites, and All other (excluded here). These categories were then collapsed into the following broader ethnic groups: Asian, Black, Hispanic and White (CPES, 2008; OMH, 2010).
Foreign- versus US-born and age at migration
Respondents were asked whether they were born in the USA, and, if not, they were asked the age at which they arrived at the USA. Because of possible differences in mental disorder risk by age at immigration (Breslau et al. Reference Breslau, Borges, Hagar, Tancredi and Gilman2009), we divided immigrants into two groups: immigrants who arrived as children (at age ⩽12 years) and immigrants who arrived as adolescents or adults (at age ⩾13 years).
Analysis
After estimating the lifetime prevalence of suicidal ideation and attempts among various ethnic groups in the CPES, discrete-time survival analysis with time-varying covariates (Efron, Reference Efron1988; Willett & Singer, Reference Willett and Singer1993) was used to examine the associations between ethnicity, immigration and risk of suicidality, adjusting for sociodemographic variables. Time-varying indicators for DSM-IV disorders were further added to adjust for possible effects of mental health on the differential distribution of suicidality across ethnic and migrant groups. Discrete-time survival analysis was used instead of logistic regression to model the lifetime prevalence of suicide-related outcomes as this allows for incorporation of the temporal ordering of onset of mental disorders and suicide-related outcomes, time-varying sociodemographics (e.g. marital status) and, most importantly, immigration status among CPES respondents. This, in turn, allowed for comparison of the risk of the onset of suicide-related outcomes among foreign-born immigrants in the CPES before and after their immigration to the USA. We used weights developed by CPES biostatisticians (NIMH, 2010).
Survival coefficients were converted to odds ratios (ORs) for ease of interpretation; significance tests of sets of coefficients in the logistic regression equations were made using Wald χ2 tests based on design-corrected coefficient variance–covariance matrices. Statistical significance was consistently evaluated as p=0.05, two-tailed. We estimated standard errors and significance tests by the Taylor series method with SUDAAN version 9 (SUDAAN, 2005) to adjust for the weighting and clustering of the data. We also report the 95% confidence intervals (CIs) of the ORs, which were adjusted for design effects on stratification and clustering and for unequal weighting of the observations. Because of the multiple comparisons included in this report for each dependent variable, we emphasize findings with p<0.01.
Results
Suicidality among ethnic groups
Table 1a shows that 14.64% of the sample reported suicidal ideation and 4.58% reported a suicide attempt. Suicidal ideation was most common among Non-Hispanic Whites, and suicide attempts were most common among Hispanics. Asians had the lowest prevalence in both outcomes. The crude analyses (Table 1b) suggested significant differences in suicide ideation (χ2=33.04, p<0.001) and attempts (χ2=24.41, p<0.001) across ethnic groups.
Lifetime prevalence and crude association of suicidality among race/ethnicity in the CPES study (n=15 180)Footnote a
CPES, Collaborative Psychiatric Epidemiology Surveys; s.e., standard error; OR, odds ratio; CI, confidence interval; df, degrees of freedom.
a The total sample includes long-version National Comorbidity Survey Replication (NCS-R) respondents and excludes the following respondents from the National Study of American Life (NSAL): Non-Hispanic Whites (were not asked the suicidality section), respondents not willing to think before answering questions or who did not pass the Short Portable Mental Status Questionnaire (SPMSQ) test. There were 17 respondents with missing data on ideation and 21 on attempt.
b Frequencies are unweighted; percentages are weighted.
c ORs were estimated by discrete-time survival models controlling only for person-years.
p values were computed by the Wald χ2 test.
As expected, there were differences in the crude lifetime prevalence of any of the psychiatric disorders listed in the Method section. The lifetime prevalence of any disorder was: Asians (18%, 1.3 s.e.), Hispanics (31%, 1.3 s.e.), Blacks (31%, 0.9 s.e.) and Non-Hispanic Whites (40%, 1.2 s.e.) (χ2=125, 3 df, p<0.001). Discrete-time survival estimates of ORs adjusted by age, sex, marital status, educational attainment (Table 2a) and the series of demographic variables and psychiatric disorders (any mood, any anxiety and any substance use disorders) (Table 2b) show that, compared to Non-Hispanic Whites, all ethnic groups had lower risk of suicide ideation (χ2=33.5, OR range 0.67–0.74) but similar risk for suicide attempts (χ2=7.17, OR range 0.80–1.21) (Table 2b). The lack of association between ethnic groups and suicide attempt was mostly due to the further adjustment of psychiatric disorders because all ethnic groups had significantly lower ORs in models that adjusted only for basic demographic variables (χ2=16.88, OR range 0.53–0.92, Table 2a).
Adjusted lifetime association of suicidality with race/ethnicity in the CPES study (n=15 180)
CPES, Collaborative Psychiatric Epidemiology Surveys; OR, odds ratio; CI, confidence interval; df, degrees of freedom.
a ORs were estimated by discrete-time survival models controlling for age and sex as fixed covariates, and person-year, marital status and education as time-varying covariates.
b ORs were estimated by discrete-time survival models controlling for age and sex as fixed covariates, and person-year, marital status, education, any mood, any anxiety and any substance DSM-IV disorders as time-varying covariates. Any mood includes major depressive disorder with hierarchy and dysthymia; any substance includes both alcohol and drug abuse and dependence; any anxiety includes panic disorder, agoraphobia without panic disorder, social phobia, generalized anxiety disorder and post-traumatic stress disorder.
p values were computed by the Wald χ2 test.
Suicidality and nativity
Table 3 presents the raw data on suicidality among ethnic groups and nativity. For the total population, the prevalence of suicide ideation is highest in the US-born (15.68%), followed by immigrants who migrated as children (⩽12 years) (12.64%), and immigrants who migrated at older ages (>12 years, adolescents or adults) (6.48%). Among the US-born, the prevalence of suicidal ideation was highest in Non-Hispanic Whites and the prevalence of suicide attempt was highest among Hispanics. The results of Table 3 suggest large differences between the US-born and the foreign-born of the same ethnicity in both ideation and attempts. To examine differences within the US-born population only, discrete-time survival models similar to those used in Table 2b were estimated in this subsample, adjusting for demographic characteristics and psychiatric disorders (results not shown). Only two significant differences emerged in this subsample: compared with the US-born Non-Hispanic Whites, risk for suicidal ideation was significantly lower among the US-born Non-Hispanic Blacks (OR 0.74, 95% CI 0.65–0.86) and risk for suicide attempt was significantly higher among the US-born Hispanics (OR 1.43, 95% CI 1.10–1.87).
Lifetime prevalence of suicidality among race/ethnicity and nativity in the CPES study by nativityFootnote a
CPES, Collaborative Psychiatric Epidemiology Surveys; s.e., standard error; USB, United States born; FB, foreign born.
a Frequencies are unweighted; percentages are weighted.
b There were 53 respondents with missing data on migrant status or age of migration.
In Table 3, within the four ethnic groups, immigrants who migrated as adolescents or adults had consistently lower prevalences for suicide ideation and attempts compared to US-born respondents of the same ethnic group. The immigrants who migrated as children of Asian and Black ethnicity showed higher lifetime prevalence of ideation and attempts than the US-born of the same ethnicity. By contrast, among Whites, older immigrants had prevalence of suicide ideation that was higher than younger immigrants.
Suicidality and immigration
A final set of models was estimated to compare immigrants with the US-born of the same ethnicity during years prior to and following their arrival in the USA. In the total sample, risk for both ideation and attempt is lowest during years prior to immigration, and intermediate in the years following immigration, after adjusting for age, sex, marital status and education (Table 4, model 1). After further adjustment for psychiatric disorders (Table 4, model 2), the risk remains lowest in years prior to immigration, but is no longer significantly different from the US-born in years after immigration in three out of four comparisons in the total sample. A similar pattern of results was found across all ethnic groups when we adjusted for psychiatric disorders, except for Non-Hispanic Whites. Focusing on the results from model 2, there are only two instances where risk is significantly lower among immigrants after they have arrived in the USA compared to their co-ethnic US-born population. Both occur in immigrants who arrived after age 12, Hispanics (suicidal ideation) and Non-Hispanic Blacks (suicide attempt). The sample size for the non-Hispanic whites, by migration status, is too small for any conclusive statement.
Adjusted lifetime association of suicidality among race/ethnicity and immigration in the CPES study by nativity
CPES, Collaborative Psychiatric Epidemiology Surveys; USB, United States born; FB, foreign born; PY, person-years; OR, odds ratio; CI, confidence interval.
a Estimated in discrete-time survival models controlling for age and sex as fixed covariates and person-year, marital status and education as time-varying covariates.
b Estimated in discrete-time survival models controlling for age and sex as fixed covariates and person-year, marital status, education, any mood, any anxiety and any substance DSM-IV disorders as time-varying covariates.
c Could not be estimated due to zero cell.
* p<0.05, ** p<0.001.
Discussion
This large national survey provided an opportunity to examine suicide ideation and attempt in relationship to ethnicity and immigration in population samples assessed with the same survey instrument. First, crude ethnic differences in suicide ideation and attempt were generally similar to ethnic differences in rates of suicide as compiled by the CDC (2011). Nevertheless, after adjustment for mental disorders, the ethnic differences in prevalence of suicide ideation was greatly reduced but still persisted. For suicide attempt, the ethnic differences were no longer significant when these adjustments were performed.
Second, immigration makes a large contribution to ethnic differences in the prevalence of suicide ideation and attempt, with the US-born more likely to report suicidality. Suicidality is less common among the large immigrant segments of the Hispanic and Asian-American populations, contributing to the low crude prevalence of suicide ideation and attempt in these groups. US nativity further attenuated the differences in suicide ideation and attempt across ethnic groups. When an ethnic comparison was performed among the US-born, no differences were found in suicide ideation and attempts with only two exceptions: US-born Blacks had lower prevalence of suicide ideation when compared to US-born Whites, and US-born Hispanics had higher prevalence of suicide attempt when compared to US-born Whites.
Third, within ethnic groups and using the US-born as the reference, the lowest ORs for both ideation and attempts were found among immigrants prior to their arrival in the USA. After arrival in the USA, risk among immigrants tended to increase to levels similar to that of the US-born population of the same ethnicity.
Prevalence of suicidality
The lifetime prevalence of ideation and attempt that we report here for Blacks, Asians and Hispanics are comparable to, but slightly different from, those reported by the individual studies from the CPES (Joe et al. Reference Joe, Baser, Breeden, Neighbors and Jackson2006; Fortuna et al. Reference Fortuna, Perez, Canino, Sribney and Alegria2007; Duldulao et al. Reference Duldulao, Takeuchi and Hong2009), as we included ethnic groups from the NCS-R in these analyses. The Epidemiologic Catchment Area (ECA) Study (Moscicki et al. Reference Moscicki, O'Carroll, Rae, Locke, Roy and Regier1988), conducted more than 25 years ago, is the only other large epidemiological study to report separate estimates of lifetime suicide attempt for Whites (3.0%), Blacks (2.3%) and Hispanics (3.3%). The ECA estimates are lower than those reported here, but are in the same direction. Whether there was an increase in lifetime prevalence of suicidality in the past 30 years is a matter of discussion beyond our results, but more limited evidence from examining the NCS and the NCS-R (Kessler et al. Reference Kessler, Berglund, Borges, Nock and Wang2005), a 10-year interval, and the National Longitudinal Alcohol Epidemiologic Survey (NLAES) and the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC; Baca-Garcia et al. Reference Baca-Garcia, Perez-Rodriguez, Keyes, Oquendo, Hasin, Grant and Blanco2010) showed no recent increase in suicide attempt. A recent study based on the NESARC reports data on suicide attempt by ethnicity (2.4% for Whites, 2.1% for Blacks, 2.1% for Hispanics and 2.0% for Asians), but unfortunately the data are limited to respondents who screened into the depression section and resulted in extremely low lifetime prevalences (Baca-Garcia et al. Reference Baca-Garcia, Perez-Rodriguez, Keyes, Oquendo, Hasin, Grant and Blanco2010). A combination of differences in sample representativeness, differential composition of the first and the second generation of immigrants in the 1980s and the 2000s, and, most importantly, the instruments used in the ECA and the NESARC to assess suicide attempt may account for the differences in the lifetime prevalence between these surveys and the CPES.
Ethnicity and immigration into the USA
Our analyses suggest that ethnicity, per se, has a more limited role in the current epidemiology of suicidality in the USA than that shown by previous studies (Joe et al. Reference Joe, Baser, Breeden, Neighbors and Jackson2006; Fortuna et al. Reference Fortuna, Perez, Canino, Sribney and Alegria2007; Duldulao et al. Reference Duldulao, Takeuchi and Hong2009). Differences in suicide attempt among ethnic groups were accounted for by differences in psychiatric disorders among these groups and we found only a moderate gap in suicide ideation between the four main ethnic groups in the USA, which mostly disappeared when we focused exclusively on the second-generation ethnic groups. Among the second generation, Blacks were at a lower risk for ideation, but Hispanics were at a higher risk for attempt. Moscicki et al. (Reference Moscicki, O'Carroll, Rae, Locke, Roy and Regier1988) found lower ORs for suicide attempt for both Blacks (OR 0.59) and Hispanics (OR 0.56, non-significant), contrary to our findings. These differences in results may be due to differences in the statistical models used in these studies and, importantly, differences in the composition of first- and second-generation ethnic groups in the ECA and in more recent surveys of the US population. As no prior study has reported on ethnic differences in suicidality among the second generation only, future research should confirm these findings and look into more details on protective and risk factors for the second generation of the ethnic groups in the USA before any firm conclusions can be drawn.
Previous studies suggest that differences between foreign-born and US-born individuals may be due to acculturation (Canino & Roberts, Reference Canino and Roberts2001; Joe et al. Reference Joe, Baser, Breeden, Neighbors and Jackson2006; Choi et al. Reference Choi, Rogers and Werth2007; Fortuna et al. Reference Fortuna, Perez, Canino, Sribney and Alegria2007; Smokowski et al. Reference Smokowski, David-Ferdon and Stroupe2009) and selectivity of immigration (Duldulao et al. Reference Duldulao, Takeuchi and Hong2009). Besides eroding basic culturally influenced coping strategies and cultural values (Baca-Garcia et al. Reference Baca-Garcia, Perez-Rodriguez, Keyes, Oquendo, Hasin, Grant and Blanco2011), which are core elements of the concept of acculturation, it is still unclear how assimilation into US society may influence suicide ideation and attempts. For example, Fortuna et al. (Reference Fortuna, Perez, Canino, Sribney and Alegria2007) suggested that English proficiency, parental US nativity, family values, and language spoken as a child may be risk factors among Hispanics, whereas Duldulao et al. (Reference Duldulao, Takeuchi and Hong2009) argued that acculturation can be seen as both a protective and a risk factor for Asians. It may be a protective factor because highly acculturated immigrants are more socially integrated; and a risk factor because higher acculturation may imply the presence of new risk factors, such as a loss in traditional values. Crawford et al. (Reference Crawford, Nur, McKenzie and Tyrer2005) found no significant effect of acculturation levels on suicidal behavior in a UK immigrant sample. In addition, our finding that the protection for suicidality decreases after moving to the USA is consistent with the relationship that Duldulao et al. (Reference Duldulao, Takeuchi and Hong2009) found between suicidality and number of years living in the USA for Asians. Similar to Fortuna et al. (Reference Fortuna, Perez, Canino, Sribney and Alegria2007), we also found among Hispanics a higher risk of suicidality in the second generation and among immigrants who arrived in the USA at a younger age. However, for the other ethnic groups, the difference between US-born and foreign-born individuals is mainly present in the years prior to immigration.
Future research should attempt to disentangle more precisely how specific mental disorders and sociodemographic factors may differentially affect ethnic groups or even nations (Borges et al. Reference Borges, Nock, Haro, Hwang, Sampson, Alonso, Andrade, Angermeyer, Beautrais, Bromet, Bruffaerts, Girolamo, Florescu, Gureje, Hu, Karam, Kovess, Lee, Levinson, Medina-Mora, Ormel, Posada-Villa, Sagar, Tomov, Uda, Williams and Kessler2010). Our results suggest that at least some of the differences in crude rates of suicide among ethnic groups in the USA may reflect baseline differences in the prevalence of psychiatric disorders and suicidality prior to immigration among these groups, factors that were not fully investigated previously.
Limitations
The results of this study are limited in important ways. First, our analyses were based on retrospective self-reports, which may be affected by recall bias if some groups are systematically more likely to accurately recall and report past life events than others, especially recall over longer periods of time (pre-migration years). Second, the CIDI questions used to elicit suicide ideation and attempt are clearly limited and did not include any measure of chronicity or severity of suicide ideation or attempt that may also differ by ethnicity and immigration status. Furthermore, no reliability or validity data were obtained on the measures of ideation or attempt and we did not examine non-suicidal self-injury (i.e. self-mutilation) (Nock & Prinstein, Reference Nock and Prinstein2005); these restrict the inferences that can be drawn from this study and represent areas for improvement in future studies. Clinical studies of suicidal patients that include a more in-depth evaluation of the patient's state of mind and a larger number of risk factors (Mann et al. Reference Mann, Waternaux, Haas and Malone1999; Oquendo et al. Reference Oquendo, Dragatsi, Harkavy-Friedman, Dervic, Currier and Burke2005) are therefore needed. Third, the four ethnic groups examined here are internally heterogeneous and future studies with larger samples are needed to explore suicidality among important subgroups (Baca-Garcia et al. Reference Baca-Garcia, Perez-Rodriguez, Keyes, Oquendo, Hasin, Grant and Blanco2011). An example of this within-ethnic group difference was reported on suicide-related outcomes between African Americans and Caribbean Blacks, where Caribbean black men were mostly affected (Joe et al. Reference Joe, Baser, Breeden, Neighbors and Jackson2006), but this effect was not seem within Latinos (Fortuna et al. Reference Fortuna, Perez, Canino, Sribney and Alegria2007). All of the ethnic groups and their cultures studied here are richer and show a wider variability of norms, family structure or coping strategies among their members than this succinct description of findings or, sometimes, tentative conclusions may suggest. The careful description and analysis of the multiple cultural components that exist within a particular ethnicity and its constituent nationalities is a fruitful avenue that has been used before in analysis of, for example, suicidality among Hispanics in the USA (Fortuna et al. Reference Fortuna, Perez, Canino, Sribney and Alegria2007; Baca-Garcia et al. Reference Baca-Garcia, Perez-Rodriguez, Keyes, Oquendo, Hasin, Grant and Blanco2011). Finally, two sources of bias have been identified in studies of mental health and immigration: the selection of healthier immigrants and the return of sick immigrants. Without data from the sending countries, we can only speculate regarding whether these biases could be operating in our study. Nevertheless, the single study to date to test these biases in the context of immigration and suicidality among Mexicans (Borges et al. Reference Borges, Breslau, Su, Miller, Medina-Mora and Aguilar-Gaxiola2009) did not find such bias.
Conclusions
Despite these limitations, this large epidemiological study of four ethnic groups in the USA adds to the literature in showing that ethnic differences in suicidal behaviors are smaller in magnitude than those reported for suicide deaths. The findings that ethnic differences in suicidal behaviors are partly explained by differences in psychiatric disorders, nativity and risk prior to arrival in the USA are important. As most immigrants start with low rates of suicidality and these rates tend to increase as they become familiar with the US milieu and eventually settle as US nationals, the ethnic convergence of rates is not a good outcome. These findings suggest that, from a public health perspective, more preventive efforts should be directed to second-generation immigrants who have lost all protective factors, whatever they were. Pre-migration factors, such as baseline risks of suicidality, social and cultural conditions in the country of origin and selection factors (i.e. ways in which those who migrate differ from those who do not migrate), potentially play a large role. As the proportion of these ethnic groups born in the USA increases, the prevalence of suicide ideation and attempt, and potentially also of suicide, is likely to increase to levels similar to those currently found among Non-Hispanic Whites.
Acknowledgements
Funding for this study was provided by NIMH Grant R01 MH082023 (J. Breslau, PI).
Declaration of Interest
None.
