تفاوتهای قومی در شیوع و عوامل مرتبط با رفتارهای خودآسیبی در یک نمونه از جستجوی درمان بزرگسالان در حال ظهور

Abstract The present study examined differences between White and ethnic minority emerging adults in the prevalence of self-harm behaviors – i.e., non-suicidal self-injury (NSSI) and suicide attempts (SA) – and in well-documented risk (i.e., depressive symptoms, generalized anxiety symptoms, social anxiety symptoms, suicidal ideation (SI), substance use, abuse history) and protective factors (i.e., religiosity/spirituality, family support, friend support) associated with NSSI and SAs. Emerging adults (N=1156; 56% ethnic minority), ages 17–29 (M=22.3, S.D.=3.0), who were presented at a counseling center at a public university in the Northeastern U.S., completed a clinical interview and self-report symptom measures. Univariate and multivariate logistic regression models were used to examine the association between risk and protective factors in predicting history of NSSI-only, any SA, and no self-harm separately among White and ethnic minority individuals. Ethnic differences emerged in the prevalence and correlates of NSSI and SAs. Social anxiety was associated with SAs among White individuals but with NSSI among ethnic minority individuals. Substance use was a more relevant risk factor for White individuals, and friend support was a more relevant protective factor for ethnic minority individuals. These findings suggest differing vulnerabilities to NSSI and SAs between White and ethnic minority emerging adults.

1. Introduction Rates of self-harm, including both suicide attempts (SAs) and non-suicidal self-injury (NSSI), are higher in emerging adulthood (ages 18–29) than at any other time in adulthood (Kessler et al., 2005 and Prinstein, 2008). SAs involve potentially fatal, self-inflicted injury with intention to die, while NSSI involves self-inflicted injury without intention to die (Prinstein, 2008). Large-scale studies suggest a 17% lifetime prevalence of NSSI and 8% lifetime prevalence of SAs among college students in the U.S. (Whitlock et al., 2006; American College Health Association, 2012). However, there is a dearth of information about the prevalence of NSSI and SAs across racial/ethnic groups. According to national data, 11% of Latino high school students and 9% of Black students, compared to 6% of White students, made a SA within a 12-month period (Centers for Disease Control and Prevention, 2014). There are no comparable data available for emerging adults, although some smaller-scale research reported higher rates of SAs among racial/ethnic minority compared to White college students (Gutierrez et al., 2001). The racial/ethnic differences in SAs evident among adolescents may thus extend to emerging adults. Prior research has alluded to the role of social and environmental factors in risk for different types of self-harm (Gratz et al., 2002 and Dupéré et al., 2009). However, the role of culture has not been adequately investigated. The cultural theory and model of suicide suggests that culture affects the ways in which people experience and respond to stress and thus how self-harm-related thoughts and behaviors arise (Chu et al., 2010), but existing research findings are scarce and largely mixed. With regard to NSSI, some researchers report no significant racial/ethnic differences (Brausch and Gutierrez, 2010 and Serras et al., 2010), while others report higher rates among White compared to ethnic minority individuals (Whitlock et al., 2006, Muehlenkamp and Gutierrez, 2007, Gollust et al., 2008, Kuentzel et al., 2012, Swahn et al., 2012 and Chesin et al., 2013), and one study reported higher rates among ethnic minority compared to White adolescents (Taliaferro et al., 2012). Closer examination suggests that rates vary by race/ethnicity when the different types of self-harm behaviors are disaggregated. In a racially/ethnically diverse sample of urban adolescents in the U.S., White teenagers had greater odds of reporting a history of NSSI without a previous SA compared to Black and Hispanic teenagers, who had greater odds of having a history of SAs without previous NSSI (Swahn et al., 2012). These findings suggest that there may be racial and ethnic differences in the prevalence of different types of self-harm behaviors. Specifically, whereas White youth may be particularly vulnerable to engaging in NSSI, racial/ethnic minority youth may be particularly vulnerable to SAs.

3. Results About one-third (N=354) of participants reported a history of any self-harm, with 228 (18%) reporting NSSI-only, 55 (4%) reporting SA-only, and 117 (9%) reporting both NSSI and SAs. Therefore, 13% (N=172) of respondents reported any SA history (with or without NSSI history). A higher proportion of females than males reported a history of NSSI and SA, Zadj=3.0, p<0.01, and NSSI-only, Zadj=2.3, p<0.05. There were racial/ethnic differences in self-harm history, χ2(12)=33.68, p<0.01, with White individuals having a lower proportion of SA-only than would be expected by chance, Zadj=3.8, p<0.01. Black individuals and individuals identifying as Other race/ethnicity had a higher proportion of SA-only than would be expected by chance, Zadj=2.3, p<0.05, and 2.6, p<0.01, respectively. Furthermore, individuals identifying as Other had a lower proportion of no self-harm, Zadj=2.9, p<0.01, and higher proportion of NSSI+SA, Zadj=2.0, p<0.05, than would be expected by chance. Since only a small subgroup of individuals reported SA-only (without NSSI), further analyses examining SA history as an outcome did not exclude individuals who also had a history of NSSI. Further, White emerging adults had a lower proportion of lifetime SA history, with or without NSSI, than would be expected by chance, Zadj=3.7, p<0.01, whereas individuals identifying as Other race/ethnicity reported a lower proportion of no history of self-harm, Zadj=2.9, p <0.01, and a higher proportion of SA history, with or without NSSI, Zadj=3.2, p<0.01, than expected by chance. For more details on ethnic differences in history of NSSI and SA, see Table 1. 3.1. Racial/ethnic differences in risk and protective factors There were significant racial/ethnic differences in reports of substance use, χ2(4)=18.15, p<0.001, and suicidal ideation, χ2(4)=10.01, p<0.04, but not in other risk factors. Specifically, a greater proportion of White individuals, Zadj=4.0, p<0.01, and a lower proportion of Asian individuals Zadj=2.1, p<0.05, reported high levels of substance use than would be expected by chance. Furthermore, a lower proportion of White individuals reported suicidal ideation than would be expected by chance, Zadj=2.9, p<0.01. Differences also emerged in protective factors – specifically in religiosity, F(4, 1132)=4.82, p<0.01, friend support, F (4, 1141)=4.55, p<0.05, and family support, F(4, 1142)=12.12, p<0.01. Post-hoc Bonferroni-corrected t-tests revealed that Black individuals reported significantly greater religiosity (M=3.43; S.D.=1.25) compared to White (M=3.01; S.D.=1.19), t(654)=3.84, p<0.01, and Latino (M=2.99; S.D.=1.16) individuals, t(367)=3.54, p<0.01. In addition, White individuals reported significantly higher family support (M=3.35; S.D.=1.34) than did Latino (M=2.82; S.D.=1.30), t(722)=4.96, p<0.01, Black (M=2.78; S.D.=1.27), t(659)=4.65, p<0.01, and Asian (M=2.75; S.D.=1.33) individuals, t(678)=5.09, p<0.01. For more details on racial/ethnic differences in risk and protective factors, see Table 1. For additional power, and given the similarities across the profiles of the ethnic minority groups, further analyses compared white individuals with ethnic minority individuals. 1 3.2. Correlates of NSSI and suicide attempts among White emerging adults In univariate analyses, high symptoms of depression and generalized anxiety were associated with higher odds of NSSI-only, versus no NSSI/SAs (O.R.dep=2.54; 95% C.I.=1.59, 4.07, Bonferroni-corrected p<0.01; O.R.genanx=1.93; 95% C.I.=1.21, 3.08, Bonferroni-corrected p<0.05), and depressive symptoms were also associated with higher odds of SAs, versus no NSSI/SAs (O.R.=2.51; 95% C.I.=1.32, 4.76, Bonferroni-corrected p<0.05), among White emerging adults. However, after adjusting for other risk (i.e., suicidal ideation, abuse history, social anxiety symptoms, and substance use), and protective (i.e., religiosity, family support, and friend support) factors, depression and generalized anxiety symptoms were no longer significantly associated with higher odds of having a history of NSSI-only or history of any SA (compared to no NSSI/SA history) among White emerging adults (see Table 2). Similarly, while family support was associated with lower odds of NSSI-only history (O.R.=0.77; 95% C.I.=0.65, 0.91, Bonferroni-corrected p<0.05), versus no history of NSSI/SA, in a univariate analysis, it was no longer significant in the multivariate model (Adjusted O.R.=0.88; 95% C.I.=0.72, 1.09, p=0.25). The following findings also emerged in the multivariate models: Individuals reporting substance use had 2.8 times higher odds of endorsing SA history (95% C.I.=1.08, 7.20, p<0.05) and 2.1 times higher odds of endorsing NSSI-only history (95% C.I.=1.16, 3.84, p<0.05) compared to individuals with no self-harm history. Suicidal ideation was associated with 54.4 times higher odds of SA (95% C.I.=18.13, 162.88, p<0.01). It should be noted that only 9% (n=47) of White individuals reported any SA, which may influence the reliability of this estimate. Further, suicidal ideation was associated with 5.5 times higher odds of NSSI-only (95% C.I.=3.11, 9.67, p<0.01) versus no history of NSSI/SA, and 9.9 times higher odds of SA (95% C.I.=3.25, 30.15, p<0.01) versus NSSI-only. Social anxiety was associated with nearly 3.7 times higher odds of SA versus no history of NSSI/SA (95% C.I.=1.49, 9.11, p<0.05), and 4.3 times higher odds of SA versus NSSI-only history (95% C.I.=1.69, 10.78, p<0.01). Finally, abuse history was associated with 8.3 times higher odds of SA (95% C.I.=3.23, 21.43, p<0.01) and 1.8 times higher odds of NSSI (95% C.I.=1.05, 2.99, p<0.05) versus no history of NSSI/SA, and 4.7 times higher odds of SA versus NSSI-only (95% C.I.=1.80, 12.22, p<0.01). See Table 2 for more details on the unadjusted and adjusted odds ratios for White emerging adults. Table 2. Univariate and multivariate models for risk and protective factors predicting history of NSSI and SA, adjusting for gender, among White emerging adults. Unadjusted O.R. Adjusted O.R. NSSI only versus no NSSI/SA SA versus no NSSI/SA SA versus NSSI only NSSI only versus no NSSI/SA SA versus no NSSI/SA SA versus NSSI only Protective factors Religiosity/spirituality 0.99 (0.81–1.20) 1.06 (0.81–1.37) 1.07 (0.79–1.44) 1.00 (0.80–1.25) 0.95 (0.67–1.35) 0.95 (0.67–1.35) Friend support 0.83 (0.70–0.99) 1.03 (0.79–1.33) 1.23 (0.93–1.64) 0.96 (0.77–1.21) 1.34 (0.90–1.97) 1.39 (0.94–2.05) Family support 0.77 (0.65–0.91) 0.77 (0.61–0.98) 1.01 (0.78–1.32) 0.88 (0.72–1.09) 0.80 (0.58–1.11) 0.91 (0.65–1.26) Risk factors Depressive Sxs 2.54 (1.59–4.07) 2.51 (1.32–4.76) 0.99 (0.47–2.06) 1.59 (0.85–2.96) 1.04 (0.38–2.88) 0.65 (0.23–1.88) Generalized anxiety Sxs 1.93 (1.21–3.08) 1.58 (0.85–2.95) 0.82 (0.40–1.68) 0.95 (0.52–1.72) 0.54 (0.20–1.47) 0.58 (0.21–1.59) Social anxiety Sxs 1.15 (0.68–1.92) 4.18 (2.23–7.83) 3.65 (1.75–7.60) 0.86 (0.47–1.59) 3.68 (1.49–9.11) 4.26 (1.69–10.78) Substance use 1.97 (1.18–3.29) 1.73 (0.86–3.47) 0.87 (0.40–1.90) 2.11 (1.16–3.84) 2.79 (1.08–7.20) 1.32 (0.52–3.39) Suicidal ideation 6.04 (3.63–10.06) 51.15 (19.08–137.16) 8.47 (3.06–23.45) 5.49 (3.11–9.67) 54.35 (18.13–162.88) 9.91 (3.25–30.15) Abuse history 2.02 (1.28–3.19) 6.05 (2.89–12.67) 2.99 (1.33–6.74) 1.77 (1.05–2.99) 8.32 (3.23–21.43) 4.69 (1.80–12.22) Note: NSSI=non-suicidal self-injury; SA=suicide attempt; Sxs=symptoms. No NSSI/SA was entered as the reference group in the first model, while any NSSI only was entered as the reference group in the second model. p<0.05 (corrected p-Value for univariate models, p<0.006). Table options 3.3. Correlates of NSSI and suicide attempts among ethnic minority emerging adults In univariate analyses conducted with ethnic minority emerging adults, high depression and generalized anxiety symptoms were associated with higher odds of NSSI-only, versus no NSSI/SAs (O.R.dep=1.95; 95% C.I.=1.24, 3.07, Bonferroni-corrected p<0.05; O.R.genanx=2.31; 95% C.I.=1.48, 3.62, Bonferroni-corrected p<0.01) and of any SA history versus no NSSI/SAs (O.R.dep=2.86; 95% C.I.=1.77, 4.62, Bonferroni-corrected p<0.01; O.R.genanx=2.08; 95% C.I.=1.34, 3.24, Bonferroni-corrected p<0.05), and high substance use was also associated with higher odds of any SA history versus no NSSI/SA (O.R.substance=2.40; 95% C.I.=1.37, 4.19, Bonferroni-corrected p<0.05). However, after adjusting for other risk (i.e., suicidal ideation, abuse history, and social anxiety symptoms) and protective (i.e., religiosity, family support, and friend support) factors, substance use, depressive symptoms, and generalized anxiety symptoms were no longer associated with higher odds of NSSI-only history or any SA history (compared to no NSSI/SA history). In multivariate analyses conducted among racial/ethnic minority individuals, friend support was not associated with NSSI-only nor with any SA history, when no history of NSSI/SA was the reference group. However, it was associated with lower odds (Adjusted O.R.=0.75; 95% C.I.=0.58, 0.98, p<0.05) of any SA history compared to a reference group of NSSI-only, after adjusting for other risk and protective factors. Furthermore, social anxiety was associated with 1.8 times higher odds of NSSI versus no history of NSSI/SA (95% C.I.=1.05, 2.99, p<0.05). Suicidal ideation was associated with 17.5 times higher odds of endorsing SA (95% C.I.=9.00, 34.12, p<0.01) and 3.4 times higher odds of endorsing NSSI (95% C.I.=2.09, 5.66, p<0.01) versus no history of NSSI/SA, and 5.1 times higher odds of endorsing SA (95% C.I.=2.42, 10.74, p<0.01) versus NSSI-only history. Abuse history was associated with 5.7 times higher odds of endorsing SA versus no history of NSSI/SA (95% C.I.=3.06, 10.67, p<0.01) and 4.4 times higher odds of endorsing SA versus NSSI-only (95% C.I.=2.22, 8.80, p<0.01). For more details on the unadjusted and adjusted odds ratios for ethnic minority emerging adults, see Table 3. Table 3. Univariate and multivariate models for risk and protective factors predicting history of NSSI and SA, adjusting for gender, among ethnic minority emerging adults. Unadjusted O.R. Adjusted O.R. NSSI only versus no NSSI/SA SA versus no NSSI/SA SA versus NSSI only NSSI only versus no NSSI/SA SA versus no NSSI/SA SA versus NSSI only Protective factors Religiosity/spirituality 0.98 (0.82–1.18) 1.10 (0.91–1.32) 1.12 (0.89–1.41) 1.04 (0.85–1.28) 1.12 (0.88–1.43) 1.07 (0.82–1.41) Friend support 1.07 (0.89–1.28) 0.81 (0.68–0.96) 0.76 (0.61–0.95) 1.19 (0.96–1.47) 0.89 (0.71–1.13) 0.75 (0.58–0.98) Family support 0.89 (0.75–1.05) 0.82 (0.69–0.97) 0.92 (0.75–1.14) 0.93 (0.77–1.13) 0.99 (0.79–1.23) 1.05 (0.82–1.35) Risk factors Depressive Sxs 1.95 (1.24–3.07) 2.86 (1.77–4.62) 1.47 (0.80–2.68) 0.98 (0.54–1.78) 1.14 (0.57–2.27) 1.16 (0.53–2.55) Generalized anxiety Sxs 2.31 (1.48–3.62) 2.08 (1.34–3.24) 0.90 (0.51–1.59) 1.70 (0.98–2.94) 1.13 (0.61–2.08) 0.66 (0.33–1.34) Social anxiety Sxs 1.97 (1.26–3.08) 1.52 (0.97–2.40) 0.77 (0.44–1.35) 1.76 (1.04–2.99) 1.08 (0.58–2.02) 0.61 (0.31–1.22) Substance use 1.56 (0.84–2.90) 2.40 (1.37–4.19) 1.53 (0.76–3.11) 1.05 (0.53–2.08) 1.30 (0.64–2.64) 1.24 (0.56–2.74) Suicidal ideation 3.72 (2.36–5.86) 22.09 (12.15–40.15) 5.94 (3.03–11.62) 3.43 (2.09–5.66) 17.52 (9.00–34.12) 5.10 (2.42–10.74) Abuse history 1.58 (1.01–2.46) 7.39 (4.27–12.78) 4.68 (2.45–8.92) 1.29 (0.80–2.09) 5.71 (3.06–10.67) 4.42 (2.22–8.80) Note: NSSI=non-suicidal self-injury; SA=suicide attempt; Sxs=symptoms. No NSSI/SA was entered as the reference group in the first model, while NSSI only was entered as the reference group in the second model. p<0.05 (corrected p-Value for univariate models, p<0.006).